Sie sind auf Seite 1von 378

University of MANOUBA

HIGHER INSTITUTE OF ACCOUNTING AND BUSINESS


ADMINISTRATION (ISCAE)

A thesis submitted in partial fulfillment of the requirements for


the degree of

Doctoral Thesis in Accounting


THE EFFECT OF VOLUNTARY RISK
DISCLOSURE, CORPORATE GOVERNANCE
AND PROPRIETARY COST ON SHARE PRICE
ANTICIPATION OF FUTURE EARNINGS:
EVIDENCE FROM MENA EMERGING MARKETS
By:
NJIA MOUMEN
Research Supervisor:
Professor Hakim Ben Othman, Tunis Business School
Examination Committee:
President

: Mr. Hassouna Fedhila,

Examiner

Professor, ISCAE
: Mr. Chedli Baccouche

Examiner

Professor, ISCAE
: Mrs. Naila Boulila Taktak,

Member

Associate Professor, ISG Gabes


: Mr. Khaled Hussainey,
Professor, University of Plymouth
Co-Supervisor

2014-2015

ABSTRACT
This study has three main purposes: (i) to examine whether voluntary risk disclosure in the
annual reports contains value-relevant information to predict future earnings, to address (ii)
how governance structures interact with risk disclosure practice to impact share price
informativeness about future earnings and (iii) to examine whether proprietary costs have a
moderating effect on the value relevance of risk disclosures news in future earnings. This
study focuses on corporate annual reports as managers and researchers alike have recently
recognized the importance of narrative risk disclosures. The reason for this growing interest is
that information on risks and uncertainties in corporate business is becoming crucial for wellinformed investors and for an accurate corporate valuation and investment decision.
We used a content analysis approach to measure the amount of narrative risk disclosure across
different countries in the MENA region. As for share price informativeness with respect to
future earnings growth, we relied on Collins et al. (1994) model that became a benchmark for
many value relevance studies. We used a large-scale sample firms from MENA emerging
markets corresponding to a three year period: from 2007 to 2009. We retained 809; 785 and
789 firm years-observations for our three regression models and provided regression
estimates based on fixed effects and pooled OLS regressions.
This study offers several interesting results. First, we find a positive relationship between
voluntary risk information and the markets ability to anticipate two-years ahead of future
earnings growth. The positive association provides us with the first empirical evidence of the
value relevance of voluntary risk disclosure in annual reports. Second, we find that risk
disclosure associated with good board structure (large size and high proportion of nonexecutive directors) enhances share price anticipation of future earnings. In contrast,
ownership structure as examined through the level of ownership concentration, management
and institutional ownership decreases the association between risk disclosure and share price
anticipation of future earnings. Finally, we notice that the level of proprietary costs moderates
the perceived relevance of risk information, thereby making investors rely on other sources of
information in forecasting future earnings change. These results suggest that information
about corporate governance and product market competition is likely to shape the perceived
relevance of narrative risk disclosure in annual reports.

39

ACKNOWLEDGMENTS
This dissertation is the result of my journey in obtaining my Ph.D. It has been a laborious but
a highly rewarding path. This thesis has been supervised and fulfilled thanks to the invaluable
help and insightful advice of several people. At this moment of achievement, I would like to
express my deepest acknowledgement to all those who made this study successful and an
unforgettable experience for me.
First, I would like to thank my supervisor Prof. Dr. Hakim Ben Othman, who provided me
with continuous support, constructive critiques, comments, and constant guidance throughout
the period of my study. He generously contributed to the development of the thesis since
the initial stages and offered precious suggestions to improve my research. He inspired me
on how to pursue a successful research career and guided me toward proper methodologies
when addressing my research questions and structuring my PhD dissertation. Special thanks
go to Prof., Dr. Khaled Hussainey whose constant encouragement and patience were vital in
making this dissertation a reality. He spent hours proofreading my research and giving me
excellent suggestions which always resulted in improved versions of my thesis. His belief that
I could accomplish this research and belief in my effort thereafter has been a great
encouragement. Besides completing a PhD dissertation, Prof Hakim and Prof Khaled
stimulated and directed me in publishing conference and journal papers. I thank them both for
the long hours that we spent discussing various methodological issues, which enhanced my
understanding of this fascinating area of study.
I would like to thank the LIGUE lab for the funding support provided to me to participate
in the 2012 International Conference of the Association Global Management Studies
(ICAGMS) which has valuably benefited my PhD. I would also like to thank all PhD
candidates and colleagues for their help and beneficial discussions.
Last but not least, I would like to thank my parents for their encouragement and prayers for
success in my study. Then, I express my gratitude and appreciation to my brothers and
sisters for their interest and encouragement to finish my study especially during hard
times. Finally, I am thankful to my friends for their help and support and to all people who
contributed to this study.

CONTENTS
39

ABSTRACT_________________________________________________________________i
ACKNOWLEDGMENTS______________________________________________________ii
CONTENTS________________________________________________________________iii
TABLE OF TABLES_________________________________________________________v
TABLE OF FIGURES________________________________________________________vi
INTRODUCTION___________________________________________________________7
1. Background__________________________________________________________7
2. Research objectives__________________________________________________9
3. Research motivations and contributions_____________________________12
4.

Summary of key findings___________________________________________16

5.

Outline of the study________________________________________________17

Chapter I: MENA Background: The legal, financial and governance systems in MENA
emerging countries__________________________________________________________19
Introduction______________________________________________________________19
1. The origin of legal systems in the MENA region and its implication for their financial
disclosure environment._____________________________________________________20
2. The financial system in the MENA region____________________________________30
3. Corporate governance framework in the MENA region__________________________38
Summary________________________________________________________________42
Chapter II: Market based accounting research: an overview of the return-earnings
relationship________________________________________________________________43
Introduction______________________________________________________________43
1. The emergence of market based accounting literature___________________________43
2. A critical review of the current return-earnings association_______________________52
3. Lack of timeliness concern in the current return-earnings association______________64
Summary________________________________________________________________72
Chapter III: Corporate voluntary disclosures and their economic consequences: disclosure
theories & empirical literature_________________________________________________74
Introduction______________________________________________________________74
1. Attributes of corporate voluntary disclosure___________________________________75
2. Theoretical background___________________________________________________81
3. Capital market implications for corporate voluntary disclosure____________________91
Summary_______________________________________________________________114

39

Chapter IV: Corporate governance, proprietary costs and share price informativeness with
respect to future earnings news: the theoretical framework_________________________116
Introduction_____________________________________________________________116
1.

Corporate governance mechanisms, voluntary disclosure and share price anticipation

of future earnings_________________________________________________________117
2. Proprietary cost, voluntary disclosure and the return-future earnings relationship_____136
Summary_______________________________________________________________146
Chapter V: Voluntary risk disclosure and share price informativeness with respect to future
earnings: hypothesis development and research methodology.______________________147
Introduction_____________________________________________________________147
1. Hypothesis development: prior researches___________________________________148
2. Research methodology__________________________________________________156
3. Descriptive statistics____________________________________________________175
Summary_______________________________________________________________179
Chapter VI: Voluntary risk disclosure and share price informativeness with respect to
future earnings: empirical evidence.___________________________________________180
Introduction_____________________________________________________________180
1.

Multiple regression analysis____________________________________________180

2. Findings discussion and analysis__________________________________________188


3.

Robustness checks: controlling for potential endogeneity problems__________196

Summary_______________________________________________________________200
Chapter VII: Corporate governance, proprietary costs and share price informativeness with
respect to future earnings news: hypotheses development and research methodology____201
Introduction_____________________________________________________________201
1. Hypotheses development: prior researches___________________________________201
2. Research methodology__________________________________________________216
3. Descriptive statistics____________________________________________________229
Summary_______________________________________________________________239
Chapter VIII Corporate governance, proprietary costs and share price informativeness with
respect to future earnings news: empirical evidences______________________________240
Introduction_____________________________________________________________240
1.

Panel regression analysis______________________________________________240

2.

Results discussion and analysis_________________________________________246

Summary_______________________________________________________________272

39

CONCLUSION____________________________________________________________274
1. Summary_____________________________________________________________274
2. Implications___________________________________________________________276
3. Limitations and suggestions for future research_______________________________277
REFERENCES____________________________________________________________279
Appendix A: Main regulation on investment protection in MENA emerging countries
(Source: OECD, 2012)______________________________________________________314
Appendix B: Risk disclosure categories adopted from Linsley and Shrives (2006)_______316
TABLE OF CONTENT_____________________________________________________317
TABLE OF TABLES
Table1. Investor protection in MENA emerging economies__________________________________________26
Table 2. Corporate Governance Codes in some MENA emerging economies____________________________41
Table3: Sample composition by Country________________________________________________166
Table4: Sample composition by Sector__________________________________________________167
Table .5 Summary descriptive statistics: Panel data_____________________________________176
Table 6 Descriptive statistics for Risk Disclosure level from 2007-2009_______________________________176
Table .7 Pearson Correlations: Panel Data______________________________________________178
Figure 1 A Kernel density plot________________________________________________________________181
Table 8. Sktest for Normality of residuals_______________________________________________________182
Table .9 Breusch-Pagan / Cook-Weisberg test for heteroskedasticity____________________182
Table .10 VIF and Tolerance values_____________________________________________________183
Table.11: 2007 Cross Section Multiple Regression________________________________________________189
Table.12 2008 Cross Section Multiple Regression_______________________________________191
Table.13 2009 Cross Section Multiple Regression_______________________________________193
Table.14 Panel Multiple Regression___________________________________________________________194
Table 15 Instrumental multiple regression analysis_______________________________________________199
Table.16 Sample composition by Country_______________________________________________________218
Table.17 Sample composition by Sector________________________________________________218
Table .18 Panel A. Summary descriptive statistics for continuous variables (longitudinal
data)___________________________________________________________________________________230
Table .18 Panel B. Summary descriptive statistics for continuous variables (Pooled data)
________________________________________________________________________________________231
Table .18 Panel C. Summary descriptive statistics for dummy variable (longitudinal data)________________231
Table .18 Panel D. Summary descriptive statistics for Proprietary cost variable model (longitudinal data)___232
Table .19 Panel A: Pearson Correlations: Shareholding Structure (longitudinal Data)_____234
Table.19 Panel B: Pearson Correlations: Board Characteristics (Pooled Data)_________________________236
Table.19 Panel C: Pearson Correlations: Proprietary Cost (longitudinal Data)________________________238

39

Table.20 The Jacques Bera Test for normality of residuals__________________________________________241


Table.21 The Chow-test for fixed effect model___________________________________________________241
Table.22 TestParm for a time fixed effects regression______________________________________________242
Table 23 The Jacques Bera test for Normality of residuals__________________________________________243
Table 24 The Wald test for groupwise heteroskedasticity___________________________________________244
Table. 25 VIF and Tolerance values___________________________________________________________246
Table.26 the joint effect of voluntary risk disclosure and ownership concentration on share price anticipation of
future earnings.___________________________________________________________________________249
Table 27 the joint effect of voluntary risk disclosure and managerial ownership on share
price anticipation of future earnings.___________________________________________________252
Table .28 the joint effect of voluntary risk disclosure and institutional ownership on share price anticipation of
future earnings.___________________________________________________________________________255
Table 29 The joint effect of non-executive directors, risk disclosure and share price
anticipation of future earnings__________________________________________________________258
Table 30. The joint effect of board size, risk disclosure and share price anticipation of
future earnings_________________________________________________________________________261
Table 31 The joint effect of duality and risk disclosure on share price anticipation of future
earnings_______________________________________________________________________________263
Table 32 The impact of unmodeled variables___________________________________________269
Table 33: the joint effect of proprietary cost and risk disclosure on share price anticipation
of future earnings______________________________________________________________________271

TABLE OF FIGURES
Figure 1 A Kernel density plot..............................................................................................................................181

39

INTRODUCTION

INTRODUCTION
1. Background

ccounting information has been considered as one of the fundamental pillars


contributing to good corporate governance (Healy and Palepu, 2001). As
owners delegated to managers the leading function of their enterprises, agency
conflicts arose between insiders and outsiders. Typically, managers are more

informed about corporate business and activities than investors. The latter is mainly affected
since they must make the right choice for their investment portfolios. As such, availability of
information is essential to minimize the information asymmetry between both sides. It plays
an important role in the individual and corporate decision making process (Bogdan et al.,
2009). Firms reporting activity is expected to provide users with a great insight into the
amounts, timing, and uncertainty of its future cash flows (FASB, 2010). Adequate disclosure
by corporations helps to ensure the efficiency of capital markets.
While transparency and disclosure are socially desirable (Frolov, 2007; Diamond, 1985),
firms often trade off costs and benefits of revealing private information when determining the
optimal levels of their disclosure. According to Gibbins et al. (1990), corporate disclosure
strategies are a response to internal as well as external conditions. The firms disclosure
decisions are likely to be motivated by a variety of factors, e.g., agency costs (Leftwich,
Watts, and Zimmerman, 1981); litigation costs (Skinner, 1994); information asymmetries
(Hughes, 1989); proprietary costs (Verrecchia, 1983) and disclosure related costs (Ali, Ronen,
and Li, 1994). The interplay between costs and benefits may lead to partial or no disclosure.
Regulators had thereupon introduced rules to modify the content of, and the practices that
bring about, firms financial reports. Disclosure regulations seek to provide investors with the
minimum amount of information that would make effective investment decisions easier
(Griffin and Williams, 1960; Wolk et al., 1992). These minimum disclosure requirements
should reduce the information gap between a company and its stakeholders and provide a
level playing field for sophisticated and unsophisticated investors (Leftwich 1980; Watts and
Zimmerman, 1986; Beaver, 1998). At the same time accounting research suggests that the
information asymmetry wont be diminished and enhanced disclosure is preferred for it brings
more gains in economic efficiency (Hossain, 2008). Voluntary information, i.e., information
in excess of compulsory disclosure may promote managements accountability and decrease
the monitoring costs of investors.

INTRODUCTION
A rich information environment should foster a healthy relationship between a company and
the business community (e.g. Financial analysts, creditors), improves the market liquidity
and reduces the cost of capital and the cost of external financing. Nevertheless, the extent of
the gains and the ultimate effect on share prices may differ considerably depending on the
informativeness of corporate disclosure and on the ways the information is conveyed and
used (Hossain and Reaz, 2007). As the content of voluntary information is not explicitly ruled
through norms or laws for it is subject to managerial discretion, enhanced disclosure remains
a matter of biased selected information (Core, 2001).
Haniffa and Cooke (2002) point out that corporate voluntary disclosure reflects the underlying
environmental influences that affect a firms accounting policy. It is admitted that these
voluntary disclosures are driven by the changing economic incentives and the regulatory
environment. Indeed, with the increasing complexity of business operations, stakeholders
needs for information have become more sophisticated. Society and standard setters, now,
require a higher level of information disclosure, specifically, for publicly traded companies.
The unexpected successive collapses of large firms with a biased image of being low risk and
highly predictable (e.g., Enron) put into question the scope and the usefulness of corporate
reporting. This situation compelled accounting regulators to rethink the set of requirements
for financial reporting.
Meanwhile, there were considerable pressures from a wide range of market participants for
relevant and understandable information about corporate prospects and business uncertainties
(Deumes, 2008). As a result, narrative risk disclosure became increasingly required in
periodic reports by national GAAPS or formal codes of best practice in corporate governance
worldwide. Many developed countries such as USA, Canada, Germany, UK, Austria and
Finland etc..., mandate financial risk communication, however its current regulatory
framework reveals a piecemeal approach, focused, predominantly, on market risk associated
with the use of derivatives (e.g., FAS 119, FAS 133, IAS 32, and IAS 39). In this regard, a
wider set of risk information is hitherto not driven by rules and is offered on a voluntary basis.
Recently, accounting literature (e.g., Abraham and Cox, 2007; Beretta and Bozzolan, 2004;
Dobler, 2008; Dobler et al., 2011; Lajili and Zghal, 2005; Linsley and Shrives, 2006;
Elshandidy et al., 2013; 2014) showed an increasing interest in corporate risk reporting, yet
such practice is still at its early stage of development.

INTRODUCTION
The reason for this concern is that information on the risks and rewards of corporate business
is becoming crucial for well-informed investors and for an accurate investment decision. Such
information is believed to improve investors understanding of uncertainty, which is inherent
in every business, and enable them to make more accurate corrections to their early
expectations (Fuller and Jensen, 2002). From a corporate governance perspective, narrative
risk disclosure is expected to narrow the information gap between management and
stakeholders with respect to business uncertainties and opportunities. It may decrease the
firms perceived risk because enhanced information on corporate risk should result in a better
assessment of the firms future performance. Managers also benefit from their transparency
about the underlying risks in their strategic goals. Elshandidy et al., (2013) suggest that
managers may signal their quality in identifying and managing risk, hence differentiating
themselves from other corporate managers who may be perceived to measure and report on
risk less effectively. Linsley and Shrives, (2006) argue that risk disclosure provides managers
with a competitive advantage to attract a lower cost of capital. Skinner, (1994, 1997) contends
that timely and sufficient risk information protects managers from potential litigation and
reputation costs.
Despite the recent considerable focus on examining risk disclosure, information on corporate
risk exposure is still one of the most ambiguous areas of corporate disclosure choices. In
particular, we know too little about the value relevance of risk disclosure, though the growing
debate about its inadequacy. As risk disclosure activity exhibits a piecemeal approach, it relies
on voluntary risk reporting beyond specific mandatory disclosures and allows for managerial
discretion. Market forces, management incentives and governance attributes are likely to play
a major role in shaping the amount of risk disclosures and their informativeness (Dobler,
2008).
2. Research objectives
It is acknowledged that risk is an inherent part of business activities as companies become
more and more exposed to volatility and uncertainties. Corporations tend to develop
sophisticated ways to assess and manage risk (Dobler, 2008). Corporate risk management
aims then to identify the risk factors and to evaluate their potential impact on future outcomes.
Within an adequate risk management system, managers, unlike outside shareholders, have
access to more information on corporate uncertainties and prospects. Thus, a risk information
gap exists between companies and their stakeholders (Linsley and Shrives, 2006).

INTRODUCTION
Narrative risk information in annual reports may be an important tool to decrease this
information asymmetry (Linsley and Shrives, 2000; Lajili and Zghal, 2005). Risk disclosure
can include information on strategy, actions and performance, risk and forward-looking
information (ICAEW, 2002).
There is a considerable body of literature reflecting detailed academic work on risk
management, but and despite the increasing interest in corporate risk information there is still
limited research on risk disclosure (Woods et al. 2007). Most studies on risk information
raised concerns about the current state of risk disclosure activity. Some scholars argued that,
as it is, current risk disclosure does not contain reliable information, for it is, in most cases,
left to managers discretion, (Schrand and Elliot, 1998), others pointed to its lack of progress
(Abraham and Shrives, 2014), and others considered it unhelpful and devoid of real meaning
(Campbell and Slack, 2008; Davies et al, 2010; Moxey and Berendt, 2008). These criticisms
may be argued to be groundless, for they are not supported by any empirical evidence. The
few existing studies, investigated, mainly, the value relevance of mandatory risk disclosures
provided in line with the Securities and Exchange Commission's requirement FRR No.48.
Firms are required to provide quantitative and qualitative information about exposure to
market risk and to disclose how they account for derivatives (e.g. Rajgopal, 1999; Linsmeier
et al 2002; Jorion, 2002; Liu et al., 2004; Lim and Tan, 2007; Prignon and Smith, 2010).
Recently, Kravet and Muslu (2013) filled this gap in the literature by addressing how users
risk perceptions change around the filing dates in response to changes in textual risk
disclosures. This study provided compelling evidence on the informative nature of narrative
risk disclosures. Following Kravet and Muslu (2013), other recent studies (Bao and Datta,
2014; Campbell et al, 2014) addressed the information content of risk factor disclosure and
provided mixed findings. Campbell et al. (2014) showed that risk factor disclosures are not
boilerplate, but instead meaningfully reflect the risks a firm faces. They decrease the
information asymmetry among firms shareholders and are incorporated into stock prices.
Conversely, Bao and Datta (2014) find that around two-thirds of risk types lack of
informativeness and have no significant influence. It is important then to take these mitigated
results one step further and to document more evidence on whether narrative risk disclosures
convey useful information to investors. We basically conjecture that voluntary risk
information on future earnings in the annual reports will be impounded into current stock
prices.

INTRODUCTION
We explored the following research questions: First, does voluntary risk disclosures in firms
annual reports impact the return-future earnings relationship? Second, how governance
attributes interact with risk disclosures to impact share price informativeness with respect to
future earnings growth? Third, does the level of proprietary costs moderate investors ability
to anticipate revealed future earnings news?
The aim of this study is threefold. First, is to extend risk disclosure literature by providing
empirical evidence on whether voluntary risk disclosure in annual reports contains valuerelevant information for investors to predict future earnings growth. Given the recent
evidence on the usefulness of various risk management disclosures for investors in assessing
corporate risk, it is argued that the content of risk information shall satisfy an early-warning
function and more specifically a forecasting function for outsiders. Accordingly, our study is a
way forward in seeking an empirical assessment of the informativeness of risk information
conveyed on a voluntary basis.
Some studies addressed the possible incentives of high level of risk reporting in a regulated
environment, including corporate governance characteristics (Solomon et al. 2000; Taylor et
al. 2008). Studies like Zhang et al. (2013), Abraham and Cox (2007) and Bushee and Noe
(2000) evidenced that ownership structure is an important monitoring mechanism.
Institutional investors, as pressure groups, encourage managers to provide a high amount of
risk information. The existence of a dominant group of shareholders gives rise though to
possible collusion between dominant owners and managers that reduces the incentives and
ability of concentrated owners to monitor managerial actions (Ho and Wong, 2001; Chen and
Jaggi, 2000). While such possibility can lead to lower corporate disclosures, good board
composition can outweigh their entrenchment incentives (Abraham and Cox, 2007). Such
evidence remains limited on how corporate governance structure influences the extent of
corporate disclosure and little is known about how it might influence the investors perceived
relevance of voluntary risk information. The second aim of this study, therefore, is to model
and test how governance characteristics interact with risk disclosure to influence investors
ability to anticipate future earnings growth.

INTRODUCTION
Managerial decisions to reveal useful information about future earnings are altered by the
presence of high competition in the product markets (Hossain et al. 2006). A firm
management is less motivated to convey voluntarily sensitive information such as risk related
disclosure because competitors may use this proprietary information and threaten the firm
competitive position. The third aim of this study is hence to examine the moderating effect of
proprietary costs on corporate willingness to reveal timely and value relevant information in
their annual reports. The potential competing corporate governance characteristics and
product market competition highlight the importance of examining the factors influencing
managerial incentives with respect to informative risk disclosure through multi-theoretical
perspectives.
3. Research motivations and contributions
3.1. Research motivations
We addressed our research questions for several reasons. First, most evidence on the
informativeness of risk disclosure is based on US data where specific information on risk
factors (market risk) is mandated. Voluntary risk information covers, though, a broader
spectrum of risk factors (e.g. Operational and strategic risks). Their informational content has
been recently emphasized by Kravet and Muslu (2013). The differences in the firms sample
and of the empirical choices do not allow for a direct comparison between our study and
Kravet and Muslu (2013) findings. Our research undertook a cross-country analysis and
focused on an institutional setting where heavily concentrated shareholding is prevalent. In
the MENA region, investors protection and capital market development are considerably
lower compared to the U.S. context (World Bank, Doing Business 2012). This specific
institutional milieu suggests that disclosure policy is not as developed in MENA emerging
markets as in the U.S.A. There might be little chances for risk information to diminish the
information asymmetry between managers and market participants. Second, unlike the
recent decision usefulness studies which suggest that investors tend to
impound risk information into their pricing decision in a manner that it
either affects their perceived risk (e.g. Bao and Datta, 2014; Kravet and
Muslu, 2013) or improves the market liquidity and reduces the information
asymmetry (e.g. Miihkinen, 2013 and Campbell et al., 2014), we conducted an
association study that focuses on whether it improves share price anticipation of future
earnings changes.

INTRODUCTION

Third, since Li, (2010) found that risk related tone in forward-looking statements has a
predictive power for future earnings, we examine a specific type of information in annual
reports that is different from the notion of tone. Fourth, despite these recent evidence,
little attention has been paid to distinguish mandatory from voluntary risk
disclosure in either observing how the strength of a firms governance and
market forces influence the amount of risk disclosure or in how risk
disclosure practice impact on market indicators (Elshandidy and Neri,
2015). It is theorized indeed that these two forms of disclosure have
different incentives and distinct observed usefulness (Jorgensen and
Kirschenheiter, 2012). Given that voluntary risk reporting depends on managerial
willingness to reveal useful information, corporate governance mechanisms and market forces
as examined through ownership structure, board structure and proprietary costs are likely to
shape firms risk disclosure policy.
3.2. Research contributions
Reviewing risk disclosure research highlights some gaps in the literature. First, recent studies
addressed risk reporting in developed countries with established risk reporting legislations
and little is known about developing countries, such as MENA emerging economies. Our
study answers in this respect the call of Dobler et al. (2011) regarding the
current gaps in the body of risk reporting literature in countries with weak
risk reporting legislation. The approach towards risk disclosure in MENA
emerging markets relies largely on encouraging rather than requiring firms
to reveal risk information (risk reporting regulations is limited to the
financial risk factor) differing from other contexts where the usefulness of
risk disclosure has been recently tested. Making generalizations, based on
the recent evidence on the usefulness of risk disclosure might go against a
strong strand of literature that argues that accounting practices
(measurements and disclosures) within a specific context (such as MENA
countries) should reflect its underlying environmental factors (e.g., Nobes,
1998; Dobler et al., 2011). Second, despite their growing importance, stock markets in
the MENA region have been widely ignored by international investors and academic research
until very recently. This is mainly due to imposed restrictions on foreign stock ownership, the

INTRODUCTION
lack of common accounting standards and limited corporate transparency. Most MENA
countries are likely to opt for conservative and rigid accounting system whereby
transparency and the average of investor protection are lower than common law
countries (Gray, 1988; Ben Othman and Zeghal, 2008).
The cultural context within the region is by far characterized by a preference for
secrecy which is encouraged by weak law enforcement and low non compliance costs,
if any (Al-Akra et al., 2009; Al-Omari, 2010; Ismail et al., 2013; Samaha and
Dahawy, 2011; Hassaan, 2013). As a consequence, and due to the difculty of obtaining
sufcient and reliable market data, researchers did not heavily focus on these regional
nancial markets. Our study contributes then to the recent accounting literature using a
longitudinal approach to assess the value relevance of risk reporting in MENA emerging
markets. To the best of our knowledge, there is no study that provides an empirical assessment
of the effect of voluntary risk information in annual reports on investors ability to anticipate
future earnings growth. Third, in contrast to general disclosure studies, little is known about
the impact of competition, board size, board composition and role duality on the
informativeness of risk disclosure. We draw on multi theoretical perspectives to, empirically,
examine if some corporate governance attributes and monitoring mechanisms influence the
perceived relevance of risk disclosure.
We also investigate whether managers are likely to manipulate their risk disclosure to reduce
competitive costs or are, likely, to provide credible information to enhance their reputation
and to signal their effectiveness in dealing with business uncertainties and opportunities even
if they incur some proprietary costs. The agency theory and the proprietary costs
theory provide a way of explaining and understanding the current
problematic state of risk reporting and notably its reliability. They help to
better understand the extent to which, and in what manner, risk disclosure
influences the investors ability to anticipate future earnings in an
environment where overall country-level investor protection is relatively
poor. Inconsistent with the contention that accounting information is less relevant in the
emerging markets because stock prices may fail to reflect all available information due to
some market imperfections, our study suggests that narrative risk disclosure in corporate
annual reports is an important and powerful source of information for participants in making
investment decisions.

INTRODUCTION

Understanding the information conveyed by risk disclosures is important to regulators,


investors, and academic researchers.

Potential and current Investors

The role of voluntary risk disclosure in helping current and potential investors anticipate
firms future earnings growth is critical for the well-functioning of capital markets. A
confident and well-informed investor is necessary for an accurate valuation of one entity.
Recognizing then the long-run health of a company and the relevance of risk related
disclosure to investors are key elements in the investment decision process. Informative risk
disclosures about firm prospects in annual reports signal managerial forthright about the
underlying risks in the firms business. Sensible or unsophisticated investors will not
accordingly rely on other (costly) sources of information to make their own assessments. At
the same time, potential investors should make allowances for management discretion over
risk information that is influenced by governance characteristics and competition. They shall
recognize risk disclosure limitations and diversify their sources of information and look for
checks on the information provided by managers. Such disclosure may affect investors
decisions on future investments.

Regulators and standard setters

Regulators are often criticized for failing to mandate adequate and informative disclosure
regarding firms risk and uncertainties. In general, compulsory reporting provides historical
financial information about corporate business and does not recognize economic events in a
timely manner. Enhanced narrative disclosure such as forward looking and risk information
depend as such on managerial incentives to enhance corporate transparency. In light of the
recent financial crisis and growing market volatility, it is important to document whether risk
disclosures convey information that is useful to investors. Based on our findings, regulators
should focus more on how to improve the firms information environment; Regulators and
standard setters should assure an increased focus on the informational needs of investors in
accounting regulation. They should require firms to provide details about the specific risks

INTRODUCTION
they face, given their importance to investors in predicting companies future earnings. The
findings in this study are also expected to provide regulatory bodies with useful information
about factors that influences the perceived relevance of risk disclosure. As such, regulators
should not acknowledge the proprietary nature of risk disclosure when asking for more
corporate transparency.

Academic Research

Academic research is also interested in the effect of disclosure on investors ability to


anticipate future earnings growth. We provide compelling evidence that managers narrative
disclosures about risks and uncertainties are not boilerplate or lack for a real meaning.
Contrary to such critics assertions, we evidenced that it helps investors assess the specific
risks the firm faces and narrow the range of their uncertainties with respect to future
outcomes.

Corporations

As it was noticed over time, corporate decisions are grounded on internal as well as external
perceptions of the cause of their actions. Corporate risk disclosure policy, among others, has
to take into account the market reaction to conveyed risk information and to ensure that the
information needs of investors are satisfied. This study evidences empirically that
incorporating voluntary risk information in annual reports provides timely and reliable
information to investors in assessing a companys risk exposure and in anticipating future
earnings. In doing so, managers are likely to gain the trust of investors, improve their
reputation and acquire economic benefits that are critical to the survival of the entity in the
future. Corporations can also rethink the set of monitoring mechanisms and implement better
governance structures that proved to have an incremental effect on the value relevance of risk
disclosure.
4. Summary of key findings
Our study offers many appealing results. We report first a positive impact
of voluntary risk disclosure on share price anticipation of future earnings
for two years and three years ahead. These findings give support for our
first hypothesis. They suggest that the users of corporate annual reports
need additional information, other than current earnings, to predict future

INTRODUCTION
earnings growth and that risk disclosure is impounded into current stock
returns. Estimates are based on year by year and panel regressions
whereby we controlled for the cross-sectional effects of some early
documented determinants of voluntary disclosure as well as the earnings
response coefficient such as industry, risk, firm size and profitability. The
second set of findings considers the interaction effects of governance
mechanisms, proprietary cost level and risk disclosure on the return future
earnings relationship.
Some of these results dont give support for our predictions. We show that
risk information associated with good board characteristics (large size and high
proportion of non-executive directors) improves share price anticipation of future earnings
growth. Conversely, ownership structure as analyzed through ownership concentration,
managerial ownership and institutional investors decreases the association between risk
disclosure and share price informativeness with respect to future earnings. The CEO/chairman
duality has no impact on revealed future earnings news into current returns. Finally, we find
that the level of proprietary costs moderates the perceived relevance of risk information,
thereby making investors rely on other sources of information in forecasting future earnings
growth.
5. Outline of the study
Our study begins by providing the contextual background of our research. It gives an
overview on MENA emerging economies. It discusses the prevailing legal systems in the
region and the likelihood of legal traditions to impact the level of investor protection and
accounting systems in MENA countries. The chapter reviews additionally the main
characteristics of the banking sector and the key indicators of the financial market
development in the region. The chapter ends by examining the state of the corporate
governance framework in MENA countries. The second chapter reviews the academic
literature on the return-earnings relationship. The chapter starts by examining the main factors
that fostered the development of market based accounting research. It also provides a
discussion about early literature on the return-earnings relationship. The chapter ends by
underlying the existing anomalies in the contemporaneous return-earnings association. The
third chapter addresses the theoritical framework of corporate voluntary
disclosures and their economic consequences. The chapter discusses first

INTRODUCTION
the main attributes of corporate voluntary disclosure, emphasizes its
theoretical justification and reviews the economic consequences of
enhanced disclosure. The third chapter concludes with the main studies
that dealt with share price informativeness with respect to future earnings
news and highlights the main gaps in literature that we intend to fill in our
research. The fourth chapter extends the work in previous chapters by
examining how governance structures and proprietary costs interact with
voluntary risk disclosure to impact share price informativeness with
respect future earnings growth. The chapter develops mainly the theoretical
arguments and the empirical literature regarding the effect of ownership structure, the board
characteristics and proprietary costs on the association between voluntary risk disclosure and
share price anticipation of future earnings.
The fifth chapter is devoted to the research methodology through which we empirically
assess how voluntary risk disclosure influences share price anticipation of
future earnings in the context of MENA emerging markets. It describes the
hypothesis development, highlights the methodology of measuring
voluntary risk disclosure, develops the regression model and discusses
descriptive analysis. The methodology is based on content analysis of the
management discussion and analysis sections in annual reports and relies
on Collins et al. (1994) model in measuring share price anticipation of
future earnings. The sixth chapter provides an analysis of the main empirical findings and
considers their consistency with the developed hypothesis and the recent disclosure literature.
The seventh chapter discusses the research hypotheses with respect to corporate governance,
proprietary costs and share price anticipation of future earnings news. It sheds light on the
research methodology and reports the main descriptive statistics. The eighth chapter presents
the related empirical results. The conclusion summarizes the main findings, the implication of
the results, the limitations of the study and the suggestions for future research.

CHAPTER I
Chapter I: MENA Background: The legal, financial and governance
systems in MENA emerging countries

Introduction
The Middle East and North Africa (MENA) region has many elements- be
they legal, economic, geographical, and cultural- that offer a favorable
context for considerable regional interaction and economic integration.
The definition of the MENA region is often unclear. There is no single
definition that fully captures the different key historical and cultural factors
which are shared by most countries in this region. The World Bank defines
the MENA region as a wide variety of states, going geographically from
Morocco on the Western apex of North Africa to Iran across the Arabian
Peninsula. It ranges from resources scarce to rich and from small to large
nations. Geographically the Middle East and North African countries can be
broken down into three sub-regions: North Africa, Western Asia, and the
Gulf cooperation council countries. In some respects, the MENA region is
culturally homogenous. Arabic is the primary spoken language in most of
the MENA countries except Israel, Iran and Turkey. They share a rich
historic heritage. Monotheistic religions with a historical relation among
them dominate the region. Despite these important commonalities, the
MENA region has some dissimilarity with regard to their legal traditions:
they form a mixed group of civil law and common law systems. The
interplay of secular private law and religious law continues to be of
relevance (Schwenzer et al, 2012; p. 27). Economically, one defining
characteristic of the MENA region is the presence or the absence of
significant natural resource endowments in the form of oil and gas (The
2008 MENA Economic Developments and prospects, World Bank).
Countries in the MENA region exhibit mixed economic outlook. Most of the
MENA oil-rich countries are evolving at steady rates while the oil scarce
countries face subdued economic prospects. MENA countries, mainly the
GCCC, are still single-commodity (oil) economy dependent despite the

CHAPTER I
continuous and huge effort to diversify the economies and to open many
of their financial markets to foreign investors.
Recently, in line with their common trend toward modernization and
integration into the global economy, MENA region focused on improving
the climate for private investment, with opportunities for more trade and
efficient governance mechanisms and accountability (World Bank, 2007).
Many of the MENA countries were encouraged to establish a new legal
framework.
This is likely to ensure effective governance of specialized capital market
companies, strengthen corporate financial disclosure and transparency,
attract more foreign capitals into the financial markets, and increase
investors rights through provisions forbidding unfair market practices (Ben
Naceur et al., 2008).
In this chapter, we highlight, in the first section, the legal systems in the
MENA region and its implication for their financial disclosure environment.
We emphasize in the second section the financial sector development for
MENA countries. The third section stresses eventually corporate
governance framework in the MENA region.
1. The origin of legal systems in the MENA region and its implication for their financial
disclosure environment.
1.1 The origin of legal systems in the MENA region

In a series of papers, La Porta et al., (1997, 1998, 2006 and 2008; p. 287)
argued that law and legal systems codification did not start from scratch;
they are transmitted through various channels including trade, conquest,
colonization, missionary work, migration, and so on. In most cases, legal
traditions are transplanted involuntarily among a population, from
relatively few mother countries to the rest of the world. For that matter,
legal scholars believe that most of national legal systems all over the
world are sufficiently similar to permit their classification into major
families of law (Glendon et al., 2008).

CHAPTER I
Early studies identified two main secular legal systems: common law
versus civil law, and several sub-categories within the civil law tradition,
namely French, German, and Scandinavian (LaPorta et al, 2008). LaPorta
et al. (2008) added that while the basic legal systems were mostly
inherited from former colonizers, the national laws of many countries were
adapted to local cultural, political, and economic circumstances. National
laws are likely to reflect the specificities of every society. For instance legal
and judiciary systems of no two countries are exactly alike. David (1985)
noted conversely, that the process of individualization was incomplete.
Enough of the basic transplanted elements have persisted to allow the
classification into legal traditions (LaPorta et al, 2008; p. 288).
The common-law legal system is rooted in the British law tradition while
the civil law originates from the Roman law tradition. Both formed the
basis of secular laws in many European countries.
The law of England is almost synonymous with judicial independence
being central from both the executive and the legislature and with judges
having a broad interpretation power. The law is established by solving
specific legal disputes and molded as circumstances change (LaPorta et al,
2008). In contrast, most of the civil legal traditions reject jurisprudence
and rely instead on legal scholars to formulate rules and on statutes and
comprehensive codes as a primordial means of ordering legal material.
Judges play a mechanical role since they are not allowed to interpret the
law, lack of judicial discretion and are under the control of the State (Beck
et al, 2003). Accordingly, the British regulation is marked by juries
independence, weak dependence upon rules and the reliance on contracts
and private judicial proceedings as a way to handle social disputes. Civil
rules are characterized by state-employed judges, great reliance on legal
and procedural codes, and a preference for state regulation over private
litigation (LaPorta et al, 2006, p. 14). In comparison to the common law
legal system, the civil law tradition is the oldest, the most influential, and
the most broadly distributed around the world (LaPorta et al. 1998, p.
1118). It is noteworthy that while scholars occasionally compare between

CHAPTER I
common and civil laws, most of the studies compare common law to the
French civil law. This is mainly due to the fact that each legal family
includes a large number of countries and both of them exhibit the most
distinct regulations which provide interesting opportunities to undertake
meaningful comparisons. Common law tradition was transmitted to former
British colonies, including the United States, Canada, Australia, India,
South Africa, and many other countries, while France expanded its legal
influence-involuntarily or otherwise- to Eastern Europe, the Northern and
Sub-Saharan Africa, the Middle East and Central Asia with which it still
maintains cordial economic and business relations (La Porta et al, 2008).
Many scholars suggest that former French colonies inherited restrictive
civil rules under particular conditions that increase the likelihood of their
legal system being less flexible compared to Common and German civil
law countries (Beck et al. 2003).
The MENA region follows a mix of legal systems, where the body of laws is
a combination of several legal traditions blended into one (Farooq and
Derrabi, 2012). The distribution of their legal and commercial laws, as
emphasized by La Porta et al (2008, 1999), classify Egypt, Jordan, Kuwait,
Morocco, Mauritania, Qatar, Turkey and Tunisia as French civil law
countries, while Saudi Arabia (Islamic Law), Israel, United Arab Emirates
and Bahrain follow the English common law.
The ACRLI1 (2009) regional report indicated also that the civil law system
as compared to the common law one is relatively prevalent in MENA
countries, specifically in terms of form and context where legislative
authorities, whether elected or not, issue, develop, amend and change
laws. The laws in these systems are positive, with the impact of the Islamic
Sharia or religious legislation in some aspects related to personal status
and familys right. Kobeissi (2005) noted that within the MENA region, the
codified legal system originates from distinct mother legal families. They

1 The Arab Centre for the Development of the Rule of Law and
Integrity

CHAPTER I
range from countries that apply the sharia, to those which legal system is
developed far from it (Shaaban, 1999). Kobeissi (2005) explained that it is
possible to assort MENA countries in 3 categories based on their
commercial laws. The first group includes those that westernized their
legal system like Lebanon, Syria and Egypt. The second group
encompasses countries which codified laws were mostly shaped by the
sharia law, such as Saudi Arabia, Kuwait, Oman and Yemen. The third one
comprises hybrid countries. Those which westernized their commercial
laws, but drew from the Islamic law in some areas (e.g. Contracts,
banking) including Iraq, Jordan and Libya (Shaaban, 1999).
Various studies (La Porta et al, 1997; 1998; 2006; 2008) demonstrated
that legal origin is likely to influence the institutions that are even not
fundamentally linked to the legal system. Beck et al. (2003, p. 654)
reported that Britain has better institutions that protect shareholders than
does France. British colonies are likely to inherit better institutions than
French colonies with positive ramifications on financial development.
1.2 The investor protection environment in the MENA region

Political theories stipulate that differences in legal systems impact the


priority attached to (a) private property and to (b) State rights. The
protection of private contract rights forms the basis for financial funding
and for the development of capital markets around the world (La Porta et
al, 1999). Beck et al. (2003) note that English common law was developed
to protect private property owners against the attempts by the crown to
regulate and to expropriate them. The French civil law evolved as an
instrument used by the sovereign to strengthen the States building and to
control the economic life by placing the prince above the law (Beck al,
2003; p. 654).
Over time, State dominance promoted institutions and regulations that
concentrate more on State power and less on the protection of investors
rights. French civil law tradition is believed then to have adverse
implications on financial development (Mahoney, 2001).

CHAPTER I
Investor protection is seen as a basic element of a country legal
environment that is may influence the development of financial markets.
Investor protection includes not only shareholder rights written into the
laws and regulations, but also the effectiveness of their enforcement (La
Porta et al., 2000). Countries with commercial legal systems following the
English common law are preserving more property rights, have less
market regulation and are better at protecting shareholders and creditors
compared to civil law countries (LaPorta, et al., 1998, 1999, 2000; Djankov
et al., 2002). Common law countries exhibit also lower contracting costs
since their regulations tend to interpret the spirit rather than the letter of
the contract (Lang and So, 2002). Higher contract enforcement and a
greater security of property rights facilitate the development of capital
markets. They also provide common law countries with better investment
opportunities. Civil law regimes, in contrast, attract less investment
because they are associated with long trials, more fraud, less integrity and
impartiality and inferior access to justice (Globerman and Shapiro, 2003).
La Porta et al. (1997, 1998, 2000, and 2006) suggested that law
enforcement as measured by the efficiency of the judicial system, the rule
of law, and the control of corruption depends on legal origins and is likely
to vary across countries. They showed that greater investor protection
increases investors willingness to provide financing and results in a
greater dispersion of corporate ownership in these countries. They
concluded that the common law countries offer better legal protection to
creditors. Firms in common law countries seem to have better borrowing
conditions and higher debt financing. Archambault and Archambault
(2003) corroborated these findings and asserted that common law
countries are oriented towards more transparent markets which are the
dominant source of financing. Conversely, code-law countries have smaller
capital markets and financing is largely through banks.
Farooq and AbdelBari (2013) suggested that earlier arguments about the
significant difference between the common law and the civil law countries
with respect to investor protection cannot be taken for granted in most
MENA countries. The main reason is the disparities among them in the

CHAPTER I
enforcement of laws and regulations. Farooq and AbdelBari (2013)
believed that such arguments only hold in weak property rights regimes.
In such regimes, the enforcement of private contracts is low and often
lacking. Court systems are corrupt and unable to resolve private disputes.
They cant access to required resources to find out the facts pertaining to
private contracts between investors and firms. Farooq and AbdelBari
(2013, p. 228) argued that in such circumstances, when the enforcement
of private contracts is costly, other forms of protecting property rights,
such as judicially enforced laws or government-enforced regulations, may
become more important. Accordingly, governments in the civil law
regimes may enhance the enforcement of contracts by laws and
regulations in order to offset the weak regulatory protection for minority
investors. The civil law environment will not be then disadvantageous as
long as the property rights are effectively enforced and strongly protected.
In recent years, the MENA region has been engaged in strengthening their
investment legal frameworks through the extension of the scope to trade
liberalization, investment protection and promotion. Most of MENA
countries revised their investment codes to meet the international
standards and mainly the OECD guidelines. Their commitment to
protecting investors is also shown in the evolving number of bilateral
investment treaties (BITs) adopted in the recent decade as well as the
new provisions about investor protection incorporated into their
investment laws (OECD, 2012). By 2012, 582 BITs were signed by MENA
countries and 73 intra-MENA BITs were enacted, including agreements on
proprietary rights, transparency and dispute settlements. Egypt, Morocco,
Jordan and Tunisia are at the vanguard of investment treaties concluded
with OECD countries. Particularly, Egypt, Morocco and Tunisia adhered to
the OECD Declaration on International Investment and Jordan is in the
process. The Arab states of the Gulf adhered, but at a smaller degree, to
these bilateral investment treaties concluded either with OECD or with
other states in the MENA region.

CHAPTER I
MENA countries adopted also some reciprocal agreements earlier: The
1980 unified Agreement for the Investment of Arab Capital in the Arab
States and the 1993 Convention relative lencouragement et la
protection des investissements entre les pays de lUMA gave the basic
institutional framework for encouraging the intra-Arab capital movement
and establishing the Arab Investment Court. These regional investment
agreements aimed to clarify and harmonize the protection standards. It
aimed to provide investors with high enforceable guarantees and sent a
strong signal to the international community in terms of transparency,
predictability and accountability (see appendix A).
Notwithstanding these rich regulations on investment protection and law
enforcement, the inefficiency of the judicial system remains the main
concern in the MENA region. Kobeissi (2005) highlighted that property
rights, yet protected by the constitution in many countries, face many
deferments and barriers from the legal local rules. Dispute settlement is
likely to be difficult and even unsure, judgments enforcement is far from
being an easy task as it is often lacking, and judicial proceedings are a
long process. Kobeissi (2005) explained that particularly in Egypt a
commercial case can take 6 years for a decision and with appeals the
duration could exceed 15 years. In Qatar, the legal system is prejudicial
and supportive of citizens and the government. In Saudi Arabia, The U.S.
Department of State reported that in various cases, disputes caused
serious problems for foreign investors. It prevents their departure from the
country, blocks their access to exit visas, or imposes restraint of personal
property, pending then the adjudication of a commercial dispute (Kobeissi,
2005). As for Morocco and Tunisia, The U.S. Department of State argued
that the judicial system lacks of transparency and is inefficient. Delays are
long to the extent that they impede the use of the court system.
Corruption is also present and hinders the development of private
investment. The judiciary is influenced by governmental intervention and
proprietary rights expropriation is possible. Miteva (2007) asserted
likewise that for a wide variety of MENA countries, the judicial system is
under political pressure and exhibits long delays, resulting in poor

CHAPTER I
enforcement. Private dispute resolution mechanisms are newly
implemented and largely untested. Accordingly, property rights
enforcement tends to be weak in the region.
On the Heritage Foundations index2 of private property protection, only
two countries in the region (Israel and Qatar) have a rating of very high
protection (investors have not been involved in any expropriations in
recent years), while only Jordan and Bahrain have a rating of high
protection. As for the strength of the minority shareholders protection
against directors misuse of corporate assets, most of the MENA countries
exhibit a low investor protection index (except Israel and KSA). Despite
some improvement in recent years (as in Morocco and KSA), protections
are often weak because of limited access to corporate information during
litigation. Likewise enforcing contracts is difficult in most countries in the
MENA region. Turkey, Tunisia, Morocco and Israel show the best ranking
among the listed economies as they have been recently engaged in
simplifying and speeding up proceedings.

Table1. Investor protection in MENA emerging economies


MENA

Investor

Enforcing

Proprietary

Emerging

protection

contracts

Rights index

Markets

Index (WB)

rank

(HF)

5.3
5.3
8.3
4.3
6.3
7

113
152
94
129
117
124

55
35
70
55
50
45

5
5

88
107

40
50

Bahrain
Egypt
Israel
Jordan
Kuwait
KSA

Legal tradition

Common law
Civil Law
Common Law
Civil Law
Civil Law
Mixed
(Islamic

Morocco
Oman

/Common Law)
Civil Law
Mixed
(Islamic

2 2012 Heritage Foundation/Wall Street Journal Index of Economic


/Common Law)
Qatar
Civil Law
5
95
70
Freedom. http://www.heritage.org/index/explore?view=by-regionTunisia
Civil Law
6
78
40
country-year
Turkey
Civil Law
5.7
40
50
UAE
Common law
4.3
104
55

CHAPTER I
Country classification is based on LaPorta et al (2008). Data are adapted from the 2013
World Bank doing business report and from the Heritage foundation association. The
investor protection index distinguishes 3 dimensions of investor protections: the extent of
disclosure index, the extent of director liability index and the ease of shareholder suits
index. Enforcing contracts rankings are based on the average of the economys rankings
on the procedures, time and cost to resolve a commercial dispute through the courts. The
proprietary rights scores measure the degree to which a countrys laws protect private
property rights and the degree to which its government enforces those laws.

MENA countries also impose trade policies that restrain foreign investment
and heavily regulate the banking sector. These are signs of weak economic
freedom and indicate the various ways in which a government may take
away potential profits. This can be an obstacle to the development of
capital markets (Conklin, 2002). The ACRLI (2009, p. 23) regional report
pointed out that the main dilemma facing legal systems in the developing
countries in general and the MENA countries in particular does not reside
in texts and judicial regulations. It results from the extent of respect of
these texts, the transparency in its application and commitment to the
principle of separation of powers, their independence and cooperation.
They are lacking of some modern legal institutions that should be
introduced to fill a legislative gap or for the sake of transparency in
monitoring the business environment and institutions (ACRLI, 2009, p.
28).
Given the strong impact of legal systems on investor protection, corporate
ownership and debt financing, prior accounting studies suggested that the
legal environment is likely to be a surrogate either directly or indirectly for
the amount of political influence on accounting systems.

A typical example of direct influence is the development of accounting


rules in different countries, which prescribes general requirements for the
measurement and the disclosure of accounting information (Jaggi and Low,
2000; Meek and Saudagaran, 1990; Salter and Doupnik, 1992).

CHAPTER I
1.3 The financial disclosure environment in MENA countries

Relevant literature classified the categories of accounting system under


either the Anglo-Saxon cluster or the Continental Europe cluster. Gray
(1988) identified some accounting values that characterize the British and
the French business culture traditions. The former is marked by
professionalism, flexibility, optimism, and transparency while the latter is
characterized by state control, uniformity, conservatism, and secrecy.
Nobes and Parker (2000) contended that socioeconomic environment in
common law countries fosters corporate transparency and a high level of
disclosure. Conversely, the civil law countries exhibit socioeconomic
environment that is rather linked to secrecy and uncertainty avoidance,
which limits the level of disclosure in such countries. Saudagaran and
Biddle (1992) commented furthermore, that the disclosure requirements
and expectations in Anglo-American developed countries (e.g., the USA,
the UK, Canada) are higher than in Continental European countries (e.g.,
France, Germany, Switzerland). Ben Othman and Zghal (2008) inferred
that legal systems through both codified anti-director rights and their
judicial enforcement serve as crucial elements of securing property rights
and protecting investors which may offer a favorable environment to high
corporate voluntary disclosure. Doidge et al. (2007) argued in other
respects that poor investor protection would discourage corporate
stakeholders to invest in corporate governance and in enhancing
disclosure. Controlling shareholders who risk the expropriation by the state
will earn less from implementing best practices regarding corporate
governance and disclosure. Countries influenced by the civil law legal
system are likely to opt for the legalization of accounting standards and
procedures. In these countries, accounting rules are usually detailed and
comprehensive, leaving very small latitude for interpretation and no
possibility for improvising. In this type of conservative and rigid accounting
system, the role of the accountant resides in applying literally codified and
detailed legal requirements, with special emphasis on protecting creditors.
In contrast, for countries that are under the influence of the common law
legal system, and in which regulations are set based on individual

CHAPTER I
decisions, accounting rules and practices are implemented by
professionals, mostly private sector organizations.
Accounting systems in these countries are more flexible, more transparent
and focus on timely financial reports, as well as on the information needs
and investors protection (Cerne, 2009). Jaggi and Low (2000; p. 501)
suggested that a widely dispersed ownership and a high level of debt
financing in common law countries would place heavy demand on
corporation to display a high level of financial disclosures. Managers will
be constrained to issue more detailed disclosure to meet the information
needs of investors and to enable the debt holders to monitor their
compliance with debt covenants. Ball et al. (2000) indicated additionally
that for common law countries, information asymmetry would be resolved
by timely public disclosure compared to code law countries where the level
of information asymmetry would be higher. Managers tend to
communicate privately with main equity-holders which did not encourage
the disclosure of public information.
With respect to MENA countries, only few studies addressed the
association between the implemented accounting systems and their legal
origin. Gray (1988) focused on a few countries from the MENA region,
named as the Near Eastern countries, when developing the accounting
values. Arab countries, Iran and Turkey were grouped among countries
within which accounting systems reflect the preference for statutory
control, uniformity, conservatism and secrecy. Ben Othman and Zeghal
(2008) focused on investor rights, law enforcement and capital market size
as country-specific attributes that may impact corporate governance
disclosure (CGD) in emerging countries including some MENA countries.
Results indicated that the level of CGD is considerably higher in common
law emerging countries compared to the civil law ones. Findings revealed
also that CGD is significantly influenced by the size of the capital market
for both common and civil law regimes and is strongly affected by law
enforcement only in common law emerging markets. Anandarajan and
Hasan (2010) highlighted some factors that may influence the value

CHAPTER I
relevance of reported earnings with specific reference to some MENA
firms. They argued that in common law countries, managers have less
accounting flexibility to smooth reported earnings compared to code law
countries. They hypothesized that the low degree of discretion in common
law countries may be an indication that the numbers are perceived as a
reflection of economic reality. Anandarajan and Hasan (2010) findings
showed that the nature of the legal system in one country enhances (if
common law) or detracts (if code law) the message conveyed from
financial statement numbers with reference to earnings and book values.

Al-Akra et al. (2009) examined the environmental factors that influenced


accounting development in Jordan. They observed that political and
economic factors which led to privatization had an influence on Jordan's
accounting disclosure and practices. Privatization, in turn, influenced
accounting significantly by introducing governance and disclosure
regulation reforms in 1997 and 2002. Their study indicated also that
privatization influenced significantly the Jordan legal system, prompting
the government to improve legal investor protection and to enact the new
disclosure regulation which significantly improved the disclosure quality.
Farooq and Derrabi (2012) maintained that MENA countries follow a mixed
legal system, where the body of law is a combination of several legal
traditions, including the Islamic Sharia, Ottoman law, French civil codes
and English common law. Following La Porta et al. (1999), they asserted
that differences in investor protection between common law and civil law
countries in the MENA region would impact corporate governance
mechanisms and in turn their cost of debt. Farooq and Derrabi (2012) held
that one of the implications of superior corporate governance mechanisms
in common law countries is lower agency costs and information
asymmetry problems. In these countries, compared to civil law countries,
better accounting information and governance practices allow shareholder
to receive dividends, to vote for directors, to subscribe to new issues of
securities in the same terms as the insiders and to sue directors or the

CHAPTER I
majority for suspected expropriation etc. Creditors will be also exposed to
a lower risk and eventually to a lower cost of debt. Results showed a lower
cost of debt for firms headquartered in the common law countries relative
to firms headquartered in the civil law countries in the MENA region
(Morocco, Egypt, Saudi Arabia, United Arab Emirates, Jordan, Kuwait,
Qatar, and Bahrain). This result was particularly due to differences in
corporate governance mechanisms, namely ownership structure and
auditor size between the two clusters of MENA countries.
Farooq and AbdelBari (2013) examined whether common law countries in
the MENA region provide better investor protection and thus have better
governance mechanisms compared to civil law countries. They suggested
that the extent of property rights enforcement is likely to shape the
controlling/ minority shareholders relationship. It may also explain the
difference between corporate governance mechanisms across the two
legal regimes. Their results showed that the above-mentioned
relationships hold only in weak property rights regimes.
They documented lower ownership concentration, higher likelihood of
appointing one of the big four auditors and more transparency for firms in
the common law countries relative to firms in the civil law countries. As
the property rights become strong, the difference between corporate
governance mechanisms across the common law and the civil law
countries disappears.
Overall, relevant literature emphasized the likelihood of legal system to
impact capital market development, investor protection and countries
accounting systems. MENA region, yet sharing important commonalities
such as culture, religion, and language, exhibits also some dissimilarity
regarding in particular the origin of their legal traditions. As former British
and French colonies (except for Turkey) MENA countries are often clustered
in common law family and French civil law family. They transplanted
similar law and regulations and inherited their business milieu and
accounting values from mother legal countries. Major institutions
regarding the financial sector were accordingly established following the

CHAPTER I
Western style, but they present some interesting features that make them
a challenging fieldwork (TurkAriss, 2009). Historically, the MENA countries
differed significantly in terms of economic policy orientation, degree of
openness of the economy, availability of resources and economic
structure. While some of the countries followed economic policies that
were based on government intervention and the dominance of public
sector, other allowed larger role for the private sector and for market
forces in the process of economic production and resource allocation.
These differences had significant effects on the level of development of
the financial system, the types of reforms implemented, their speed and
comprehensiveness (El Safti, 2007).
2. The financial system in the MENA region
Several MENA countries have been embarking on economic reforms during
the late 1980s and 1990s in order to comply with the structural
adjustment programs of the international institutions such as the World
Bank or the international Monetary Fund... etc.. (Ben Othman and Zeghal,
2010). A series of plans were implemented to foster the economic
diversification in this region through the development of private sector
activities and the improvement of economic freedom in terms of trade and
property rights. There has been then growing opportunities in the MENA
region for more trade and efficient governance mechanisms and
accountability (World Bank, 2007).
The heart of the above-mentioned programs is the reform in the financial
sector, which allowed most of the MENA countries to mobilize savings,
encourage good corporate governance, an enable trading, hedging, and
risk diversification.
2.1. The banking sector

Notwithstanding the considerable variation in the level of the banking


sector development and sophistication among MENA countries, the sector
was historically, by far, the main channel of financial intermediation in this
region. Banks are considered as the main financial institutions since they

CHAPTER I
control the most of capital flows and possess most financial assets. Banks
often play an important role in the course of gathering savings, promoting
investment opportunities, monitoring managers, and diversifying risk
among MENA economies (BenNaceur and Omran, 2011).
Several MENA countries experienced restructuring initiatives that aimed to
enhance privatization, bank regulation, market-orientation, and integration
of privately owned banks across different organizational structure. MENA
governments improved macroeconomic management, simplified business
regulations, reduced tariffs, and started gradual opening of their financial
sectors (World Bank, 2011). While these reforms were less vibrant than
those taking place in Eastern Europe and parts of Asia, they triggered a
well-developed, profitable, and efficient banking sector, particularly in the
Gulf Cooperation Council (GCC), Lebanon, and Jordan. Others, such as
Egypt, Morocco, and Tunisia, made considerable improvement yet more
remains to be done (BenNaceur and Omran, 2011).
For decades, banking institutions in the MENA region were either mostly
family-owned businesses or dominated by state authorities. The economic
reforms altered bank ownership and activity from government to private
control and from domestic to foreign control, which animated the dormant
securities market. Such ongoing reforms occurred as governments
privatized many of their state-owned banks and went through several
bank laws which reduced barriers to foreign entry and enhanced
transparency in line with the World Trade Organization accession
requirements. Banks are hence increasingly required to adopt international
accounting standards and prudential guidelines of capitalization and
governance. Foreign banks are encouraged to become active players in
domestic markets (Turk-Ariss, 2009). By 2009, the structure of the banking
sector in the non-oil producing countries was still marked by the presence
of government ownership.
For example, the state owns around 48.5% of banking assets in Egypt
while it owns only 25% in Morocco, 38.5% in Tunisia and 32.2% in Turkey.
However, there is no state ownership of banks in both Jordan and Israel.

CHAPTER I
With respect to the gulf cooperation council (GCC) countries, they have a
fairly large number of banks, which are financially strong and wellcapitalized. Most of them are family-owned, with modest state ownership
despite the presence of some fully state owned in Kuwait and Saudi
Arabia. As for foreign ownership, oil producing countries such as Bahrain
and United Arab Emirates have respectively 52.7% and 19% of banks with
foreign controlled assets which are the highest numbers among the GCC
countries. Israel had the lowest number of foreign owned banks (2%) and
Jordan exhibits the highest one (48.9%) among non-oil MENA emerging
economies.
Despite the progress made, MENA countries efforts were eclipsed by
faster reform and growth in other parts of the world. The restructuring
process is overshadowed by a number of features regarding the MENA
region. First, the financial sector in most of MENA economies has been for
a long time under strict government control which delayed the desired
outcomes from the financial reform agenda. The Heritage Foundation
emphasized significant government embroilment in banking and finance in
the MENA region. Its index (which weighs state ownership, restrictions,
influence over credit allocation, regulations and freedom to offer services
in the financial sector) rates only a few countries (Bahrain, Jordan, Israel,
Lebanon) as having very low government restrictiveness in the financial
sector in 2009, while others such as Egypt, Kuwait, Oman, Saudi Arabia
and Tunisia as having notable government restrictiveness.
Second, Islamic motives (private sector may be reluctant to borrow from
the conventional banks due to interest prohibition) as well as the high
information and transaction costs seem to hinder the development in the
financial sector, which constrain their role in economic growth in these
countries (Kar et al, 2011). Third, Ben Naceur and Labidi (2009; p. 2)
highlighted that vulnerable banks dominate the financial sector in the
MENA region, () public sector ownership is high, and access to banking
services is low. The substantial state ownership, and the limited financial

CHAPTER I
freedom in some countries shaped the direction of credit in MENA
countries as well as their operating efficiency.
There is considerable variation in the degree of financial development
within the MENA region. Some countries have developing financial sectors,
while for others the headway in this area has been limited.
Kar et al. (2011, p. 690) argued that differences between MENA region
regarding legal origin, trade openness, financial integration, deposit
insurance, regulatory and supervisory framework, () may have an impact
on the development of financial system. Some of MENA countries are
witnessing growing securities markets, which provide another channel of
resources for the real sector.
2.2. Characteristics of the MENA stock markets

Security markets in the MENA region attracted the attention of


professionals and policy makers within the framework of the financial
reform agenda and the massive privatization plans introduced in the
region. Most MENA countries managed to revitalize their long-standing
markets, such as Egypt, Saudi Arabia, Kuwait and Amman or to establish
new markets, such as Dubai and Abu Dhabi. Stock exchanges in this
region became very meaningful to the worlds economy and their role in
the international financial system experienced a considerable increase
(Ben Naceur et al. 2008). Several MENA countries were encouraged to
establish a new legal framework in order to ensure effective governance of
specialized capital market companies, to strengthen corporate financial
disclosure and transparency, to attract more foreign capitals into the
market, and to increase investors rights through provisions forbidding
unfair market practices (Ben Naceur et al. 2008). As a result, there were a
tremendous flow of funds into these emerging financial markets especially,
after the financial crisis in East Asia and in Argentine.
The openness of MENA markets to local and foreign investments increased
by the sale of governments assets to private funds, and the number of
companies going public also soared. The stock markets of Egypt, Turkey,

CHAPTER I
Jordan and Morocco became important contributors to the diversification of
capital markets in the region (Bouri and Yahchouchi, 2014). MENA
countries can be therefore clustered into two distinct groups with respect
to their institutional characteristics and particularly their equity market
liberalization. The first group consists of the most open economies to
foreign investments. It includes Egypt, Bahrain, Israel, Jordan, Morocco,
Oman, Tunisia and Turkey. Securities markets in these countries are the
main vector of state owned firms privatization and trade liberalization.
The second group consists of the Gulf Cooperation Councils (GCCs)
capital markets. It includes Kuwait, Qatar, United Arab Emirates and Saudi
Arabia, which are rated as being relatively restrictive with regard to foreign
ownership. OPEC countries impose restrictions on foreign ownership, which
in turn led to relatively low foreign direct investment (FDI) inflows.
Only the UAE and Saudi Arabia have recently been successful in attracting
significant amounts of FDI (Hanouz et al., 2007). Bley (2011) noted that
Saudi Arabia stock market allows only its bank shares to be purchased by
Gulf Cooperation Council (GCC) members but not by other foreigners. In
contrast, Egypt, Lebanon, Morocco, and Turkey allow 100 percent
ownership by foreigners. The remaining MENA nations fall in the middle,
allowing only a partial ownership, ownership of certain firms, and/or
ownership by GCC members only. As of 2012, Saudi Arabia, the UAE and
Turkey account for 58.13% of the FDI (net inflows) stock in the region.
Regarding Saudi Arabia and the UAE, much of this FDI entered the
countries during recent years (World Bank Development Indicators, WDI
2012).
The development and maturity process among MENA financial markets
remains, though heterogeneous and slow. Turkey, Israel, UAE, and Egypt
are perceived as fast emerging markets and in some respect moving
closer to the standards of developed countries. Kuwait, Lebanon, Tunisia,
and Morocco are still considered as frontier markets (Rejichi et al, 2014;
Lagoarde-Segot and Lucey, 2008). Market capitalization in the Middle East
and North Africa (MENA) countries rose from 244$ billion in 2002 to more

CHAPTER I
than $1153$ billion in 2012 which represents 40.36 % of MENAs GDP and
about 2.17% of world market capitalization (World Development
Indicators, WDI 2012). By the end of 2012, Saudi Arabia and Turkey are
the most developed markets in the MENA region with approximately
374$ billion and 309$ billion of market capitalization. These two security
markets, followed by the Israeli market, have also the highest market
trading value and market turnover ratio while the Tunisian market is the
smallest for most performance indicators, including market capitalization
(8,886$ billion), the market trading value (1,251$ billion) and the market
turnover ratio (13). Measured by the number of listed companies, Israels
(532), Turkeys (405) and Jordans (243) markets have the highest number
of listed firms in this region while Qatar and Bahrain have the
smallest with respectively 42 and 43 listed companies. As for stock
markets dynamic, Saudi Arabia stock market has unsurprisingly the
highest stock traded value (turnover rate ratio) in the MENA region with
almost 514,42$ billion (144.45) followed by Turkey (348.5 $ billion) and
Israel (67.34$ billion) at the end of 2012. These figures are still modest
when we compare them with other equity markets in the developed world
such as Tokyo (3605$ billion) or the NYSE (21375$ billion). Bahrain and
Tunisia have the lowest values of stock traded among MENA emerging
markets with share turnover about 1.25$ billion and 0.38$ billion
respectively.
Accordingly, it is clear that the financial markets of Israel, Saudi Arabia and
Turkey stand out in the MENA region for their relatively high liquidity. The
rest of MENA countries have turnover rates that range from 1.85 in
Bahrain to 37.8 in Egypt.
Last, countries in the MENA region differ substantially in term of the priceearnings ratio. Tunisia stands out in the MENA region with a ratio of 16.47
in 2012. Qatar and Jordan stock markets, with ratios of 13.04 and
15.575respectively are closely followed by Saudi Arabia (14.4) and Abu
Dhabi (13.49). Conversely, Bahrain (12.46), Egypt (11.5), Oman (11.43)
and Turkey (11.20) have the lowest price earnings ratios in the region,

CHAPTER I
suggesting that companies in these nations are relatively cheap and thus
attractive for investment. It is worth noticing that the price earnings ratios
of the latter financial markets are in the range of those found in developed
economies. In terms of new equity introduction, the markets in Egypt,
Israel, Jordan, Saudi Arabia, and Turkey showed a sizable increase in
numbers of new issues in recent years, which may convey a positive signal
indicating the existence of growth potential in the region. By the end of
2012, both the sizes and the dynamics of the MENA stock markets, which
had a total of 2163 listed firms, differ substantially across the region.
In line with this recent growth, stock markets participants asked for more
relevant and reliable financial information as a way to ensure an effective
resource allocation process. Financial markets have been urged to be more
active and to move toward more sophisticated trading arrangements. The
World Bank doing business report (2012) indicated that most MENA
countries have been engaged in reinforcing investor protections. For
example, Morocco allowed minority shareholders to obtain any nonconfidential corporate document during trial and required greater
disclosure in companies annual reports. Similarly, Tunisian regulators
asked corporations for enhanced information disclosure and the Tunisian
Law on Economic Initiative allowed minority investors to request a judge to
rescind a prejudicial related-party transaction. Egypt and Saudi Arabia
have strengthened the protections for minority shareholders through new
listing rules and new provisions that prohibit interested parties from voting
on the approval of related-party transactions and increase sanctions
against directors for misconduct (World Bank, Doing Business 2012). By
2012, Saudi Arabia, Kuwait and Tunisia became the countries with the
strongest investor protection index among MENA region and scored
respectively 7.0, 6.3 and 6 while Jordan and UAE have both the same
lowest score of investor protection (4.3).
The scores recorded for Middle Eastern and North African countries (MENA)
are based on three indices: the extent of disclosure index, the extent of
director liability index and the ease of shareholder suits index. As for the

CHAPTER I
extent of disclosure, Saudi Arabia (9), Oman (8), Egypt (8) and Bahrain (8)
have the highest score of corporate transparency while Jordan, Qatar and
Tunisia are ranked at the bottom among MENA countries with a relatively
low score of disclosure (5) (World Bank, Doing Business 2012). It should be
pointed out that the International financial reporting standards (IFRS) are
required in most MENA emerging markets (Bahrain, Kuwait, Oman, Qatar,
UAE) and the rest of MENA countries (Tunisia, Morocco, Saudi Arabia,
Israel) are converging to IFRS to attract international investors and
enhance corporate transparency and disclosure (Pacter report, 2014).
Notwithstanding the considerable efforts devoted to enhancing the
transparency, efficiency, depth, integration, and liquidity of MENA equity
markets, emerging markets are, by definition, evolving and then in the
early stages of development. They are characterized by a thin trading, a
low amount of liquidity and a limited number of listed companies
compared to the developed and even the emerging markets of Asia and
Latin America (Bley, 2011). MENA stock markets exhibit in this respect a
weak-form efficiency and a capital market fragmentation due to the poorquality of information and to the weak competition. There exist relatively
limited corporate disclosure requirements and barriers to efficient
information transmission through these markets. Investors are likely to be
ill-informed and have often access to unreliable financial information
(Harrison and Moore, 2012).
Other common features of MENA security markets include relatively weak
regulatory frameworks, high concentration of ownership and limited role of
market forces. A significant proportion of corporate ownership is held by
the State, influential institutions and families. According to Aloui and
Rejichi (2012) the regions financial markets are still dominated by the
banking system. Also family-controlled firms (FOEs) and state-owned
enterprises (SOEs) both play a crucial role in MENA economies. Kuwait,
Egypt and Qatar show the highest institutional investor participation in the
region, which is estimated to be closer to 30% (Koldertsova, 2012). This
figure is though lower than the level of institutional ownership in

CHAPTER I
developed markets such as the UK, USA and France and even some
emerging markets (e.g. China). The low institutional participation can be
linked to the low market movements in most MENA markets. While the free
float accounts for no less than half of the market capitalization in the most
developed stock markets, it is only alike in few MENA markets.
Kuwait, Tunisia and the UAE are a few exceptions with free float exceeding
50% of market capitalization (World Bank, 2011). The low free float in the
MENA markets reflects the high ownership concentration and the state
active role as the owner of listed companies, either directly or through
sovereign investment vehicles (Koldertsova, 2012). BenNaceur et al.
(2008) explained that Egypt had historically the largest presence of firms
government ownership averaged at the 34% level compared to Jordan,
Oman, Morocco and Tunisia. Until recently, 32 of the top 100 largest listed
companies in the MENA region, corresponding to approximately 45% of
total market capitalization of these firms, have been, under partial control
of the state. This figure is still remarkable when compared with developed
and emerging markets, where listed SOEs rarely account for more than
30% of market capitalization (Koldertsova, 2012). Jordan and Oman appear
as the countries with the highest private ownership, having more than
80% of a firm's ownership in the hands of private institutions and
individuals and Tunisia is considered as the country with the largest foreign
participation in a firms ownership at 18%. The low foreign ownership in
MENA emerging markets is attributable to the existing investment
restrictions is some markets despite the ongoing liberalization process
over the past decade. For example, Tadawul, the biggest market in the
region, has the highest investment restrictions: non-GCC nationals can
currently invest through swaps only. In the UAE, foreign investment is
limited to 49% of equity and in Qatar to 25%, which has effectively been a
barrier for these markets to be qualified as "emerging markets" by the
Morgan Stanley Capital International (MSCI) indexes (Koldertsova, 2012).
Finally, family-owned companies are the dominant characteristic of the
MENA capital markets. A single family may be among the top five

CHAPTER I
shareholders and have controlling stakes in a number of companies,
whether directly or indirectly (Ben Othman and Zghal, 2010).
Overall, listed companies in the MENA region present four main
characteristics: concentration of ownership (due to large family/state
dominated companies), family ownership and control (dominance-of
control-oriented shareholders), bank-based corporate finance (banks are
the most important source of external funds) and underdeveloped capital
markets compared to the banking sector (foreign participation is limited).
Recent surveys (OECD, 2014; 2012) highlighted that the MENA region
lacks for best corporate governance practices compared to the OECDs
governance principles and to the implemented mechanisms in developed
countries.

Controlling shareholders is the dominant governance attribute in listed


MENA corporations. Despite their strong incentives to closely monitor firm
management, their interests might conflict with the interest of minority
shareholders. The conflict is evident when the controlling shareholders
abuse the company's resources by extracting private benefits to the
detriment of other shareholders. These aspects, among others, are likely
to have a negative impact not only on liquidity and trading, but also on
transparency and firms reporting (Koldertsova, 2011). When an owner
effectively controls a firm, he also controls the production of the firms
accounting information and reporting policies. This is likely to reduce the
credibility of the accounting information and to impact negatively
corporate governance practices in general (Fan and Wong, 2002). MENA
companies are characterized also by a major role of the board of directors,
the absence of call for a separation between the chair of the board and the
chief executive officer (CEO), limited protection of the shareholders rights
and the absence of board independence (Turki and Ben Sedrine, 2012).
Ben Naceur and Laabidi (2009; p. 4) Contend that while corporate
governance framework is already in place, there is still room for
improvement with respect to transparency, disclosure, protection of non-

CHAPTER I
controlling shareholders, directors independence. Corporate
governance in the MENA region should be recognized as a public policy
concern of rapidly increasing importance in the region. More efforts are
hence needed in order to reinforce the regulatory and the institutional
frameworks and to promote non-bank financial sector development.
3. Corporate governance framework in the MENA region
Corporate governance is relatively perceived as a nascent issue in MENA
countries. It is only in the last 10 years that an arabic word for corporate
governance, HAWKAMAH, emerged (Saidi, 2011). Despite its infancy in
the region, the concept is exhibiting a growing importance at the same
time that financial scandals resulted in demands for improved corporate
governance practices in developed as well as emerging economies
(Baydoun et al., 2013). The South East Asian crisis of 1997, the prominent
collapse of Enron, WorldCom, HiH, Tyco, Adelphia and the 2006 stock
market crash - particularly pronounced in the Gulf Cooperation Council
countries- were attributed in part to inefficient governance systems,
including poor investor protection and reporting practices (e.g. Kaufmann,
et al. 2000; Rahman, 1998; Furman and Stiglitz, 1998; Becht et al. 2003;
Koldertsova, 2011). Enhancing corporate governance of listed companies
in this regard was a logical starting point that aims to strengthen the role
of the private sector in economic development and to attract foreign
investment in the MENA region (Al Yafi, 2005).
Saidi (2004; p. 16) argued that the values of corporate governance
transparency, accountability, and responsibility offer the key for the
modernization of the countries of the Middle East and North Africa.
Notwithstanding its significant headway, information about corporate
governance systems in the MENA region is relatively scarce and what is
available is dispersed among a number of sources, some of them are
public and others are not. Among the most reliable and available sources
of information are the published working papers by the Organization for
Economic Co-operation and Development (OECD, 2005; Miteva, 2007;

CHAPTER I
Koldertsova 2011, 2012) and Hawkamah (2010) surveying corporate
governance practices in this region. While above surveys provide a
descriptive overview of corporate governance progress in MENA countries,
to date, a comprehensive empirical research on the MENA region remains
in somewhat scant.
The next subsection discusses the evolution of corporate governance as a
new concept in MENA region and an overview of recent development.
3.1. Corporate governance trend in the MENA region

MENA region-particularly the gulf cooperation council countries (GCC) witnessed in the last two decades a period of high economic growth. The
exploitation of the abundant oil resources created new opportunities for
investment and for raising savings (Baydoun et al, 2013). As a result,
significant flow of funds into the financial sector led to requirements of
lenders and investors to raise standards of corporate governance (Saidi,
2005). Despite the dominant role of the banking sectors within the MENA
region (over 60% of the financial structure) in channeling these funds,
their ability to meet corporate borrowing demands was moderated by their
risk-adverse profile and the opaqueness of corporate borrowers. This
together with the lack of corporate transparency and the reluctance of
controlling shareholders to reveal additional information (other than
financial statements) triggered the need for proper disclosure and
governance, which has slowly gained traction on the political agenda
(Koldertsova, 2011). Likewise, petrochemical resources scarce countries
engaged in corporate governance reforms as a way to attract foreign
investment and to raise considerable funds for large scale infrastructure
projects. Consequently, effective governance of the financial sector
seemed to be fundamental, especially as those markets are expected to
promote better investment opportunities and a vibrant private sector
(McNally, 2011).
The first steps toward the design of corporate governance practices in the
Middle East and North Africa began in early 2000. MENA countries were

CHAPTER I
notably affected by the stock market crash of 2006 whereby the
enforcement of a good governance system quickly became key priorities
of the regional authorities. The main weakness blamed to be responsible
for the 2006 crash was capital market uncertainty and in particular the
inadequacy of information held by a plethora of investors in relation to
their investments. The information revealed by listed companies was not
sufficiently adequate and not subject to analyst coverage (Koldertsova,
2011). Therefore, strengthening the corporate governance practices of
listed companies was a rational starting point. Enhancing the financial
literacy of investors and improving the information flow to the investing
public were considered as a crucial criterion for the corporate governance
reform agenda (McNally, 2011)
Meanwhile, there has been a need for access to international market to
fund growth and to benefit from the globalization of business. Somewhat
in parallel was the target to become a regional financial center and a
member of the World Trade Organization (WTO). Consequently,
governmental agencies in association with non-profit entities and
international organizations had a heightened awareness of the importance
to launch new regulatory frameworks to improve corporate governance
(Baydoun et al., 2013). Oman and Egypt were the pioneer countries in the
region, elaborating domestic governance codes related to listed
companies in 2002 and 2005 respectively, both based on the OECD
Guidelines on Corporate Governance. Concomitant with this was the
establishment of corporate governance centers providing educational and
training services as corporate governance was a nascent concept in the
region. The Egyptian Institute of Directors and the Hawkamah Institute in
the UAE were a response as much to the existing demand for governance
know-how as for a needed coordinated strategy towards the harmonization
of corporate governance standards in the MENA region. These initial steps
gave a strong appetite for corporate governance reforms and presaged
regional trend that gained an impetus around the middle of this recent
decade when most national regulators moved to adopt higher corporate
governance standards (Koldertsova, 2011).

CHAPTER I
Since 2005, MENA countries have launched 11 corporate governance
codes with special guidance for state owned corporations, banks and
familyowned businesses. The Egyptian Institute of Directors was the
regional pioneer of a corporate governance code aimed particularly at
state owned enterprises, based upon the OECD Guidelines. The Emirati
and Jordanian regulators introduced specific codes for banks.
The Moroccan Corporate Governance Taskforce and the Lebanese
Transparency Association introduced specialist guidance for family owned
and small and medium sized businesses. By 2011, only Iraq, Kuwait and
Libya have no corporate governance code or guidelines (Koldertsova,
2011).
Notwithstanding corporate governance codes were mostly governmentinitiated, they were recommended as a voluntary guidance, providing the
basis of best practices consistent with international standards and
notably those developed by the OECD. Over the years, voluntary
recommendations were in some instances incorporated into the regulatory
and legal frameworks. For example, in Egypt, board independence
requirements were introduced into the listing rules of its financial
market and securities commissions in the UAE, KSA, Jordan, Oman and
Qatar all shifted to mandate corporate disclosure about compliance with
local corporate governance codes.
Table 2. Corporate Governance Codes in some MENA emerging economies

Country

Egypt

Jordan
Kuwait
Morocco

General

Date of

Corporate

issuanc nce

Governance

required

Code
Yes

2005

No

Yes
No
Yes

Complia

Other
Codes/Guidelines

Code for state-owned

Under

enterprise

revision

Code for banks being

2008
2008

Yes
No

drafted
Code for banks
Code for small and

CHAPTER I
medium-size
enterprises
Code for state-owned
enterprises being
drafted
Code for banks/credit
Oman
KSA

Yes
Yes

2002
2006

Yes
Yes

institutions
Guidelines for banks

amende
d 2009
Tunisia
Yes
2008
No
UAE
Yes
2007
Yes
Code for banks
(Source: adapted from Koldertsova (2011); OECD Journal: Financial Market
Trends)

Summary
This chapter gives an overview of the prevailing legal system in The MENA region
and its implication for their financial disclosure environment. Relevant
literature emphasizes the likelihood of legal system to impact capital
market development, investor protection and countries accounting
systems. Irrespective of whether they are former British or French colonies,
most MENA countries have a civil law legal system based on French legal
family. They are likely to opt for conservative and rigid accounting system
whereby transparency and the average of investor protection are lower
than common law countries.
The second section in this chapter sheds light on the financial system in
the MENA region. The characteristics of the banking sector as well as the
development of capital markets in these countries were discussed.
Notwithstanding the considerable role played by banking institutions in
channeling financial resources, MENA countries witnessed a heightened
interest in securities markets, which are considered as the main vector of

CHAPTER I
state owned firms privatization and trade liberalization. The third section
highlighted corporate governance framework in the MENA region. In the
recent decade, the trend in the MENA region was leaning towards
introducing corporate governance codes as voluntary guidelines for best
international standards, based upon the OECD Principles. All these efforts
have borne fruit: at 2011, only 3 out of the 17 MENA emerging countries
do not have any corporate governance code or guidelines.
The next chapter addresses the impetus of capital market research in
accounting and serves as a helpful guidance to understand the framework
of our study. Based on a historical perspective, the following chapter
provides an overview of the return-earnings relationship. It discusses
factors that sustained the development of association studies and
highlights the continual refinement in the earnings response coefficient
research. As the security prices, themselves, are influenced by financial
information, a new area of research examining long-horizon returnearnings relationship emerged. Such association continues to be a topic of
primary interest to researchers on issues surrounding market efficiency
and the value relevance of corporate disclosure.

CHAPTER II
Chapter II: Market based accounting research: an overview of the
return-earnings relationship

Introduction
Capital market research in accounting is a broad area of studies that emerged in the mid-1960s
along with the development of market efficiency hypothesis and the event study methodology.
It examines the relationship between published accounting information (accounting earnings,
cash flows from operations, financial statement) and key capital market metrics such as
stock returns, share prices and trading volume...etc. for a specific time horizon (Walker,
2004).
The significant growth of market based accounting literature in the last three decades was a
response to the increasing demand for fundamental analysis and valuation research. Market
based accounting research (MBAR) is intended to assess firms market value, to recognize the
mispriced securities, to find out whether security markets are informationally efficient, to
examine the economic consequence of political costs and accounting choices and to
investigate the value relevance of financial numbers prepared according to a new accounting
standard...etc. (Kothari, 2001)
This chapter reviews the existing literature relevant to the empirical analysis in the thesis and
highlights certain gaps in the previous literature. It focuses on three streams of research. The
first stream relates the factors that fostered the market based accounting research. The second
stream reviews earlier literature on the current return-earnings relationship. The third stream
underlines the existing anomalies in this contemporaneous association.
1. The emergence of market based accounting literature
In recent decades, financial accounting literature was the subject of a profound mutation. Till
the mid of the 1960s, accounting research was generally normative and essentially seeking to
provide a conceptual framework for best accounting practices. Accounting policies and
techniques were recommended based on their ability to best fulfill a set of accounting
objectives and their assumed relevance by researchers.

CHAPTER II
Since the late of the 1960s, accounting research has steadily evolved toward an empirical
approach in order to emphasize more on the validity of these normative studies. Analyses of
the theoretical and the methodological corpus of the market based accounting research show
its attachment to three main developments. First, the emergence of economic sciences and
positive approach (Friedman, 1953) drew a fundamental distinction between a positive
science described as a body of knowledge about how things are as opposed to how they
should be in a normative approach. The move forward this new trend of studies aimed to
provide a scientific basis for accounting research. In fact, alike exact sciences, positive
research link the usefulness of any theoretical proposition to its empirical validation.
Embracing such approach in accounting research helped assess the accounting policies and
methods that were assumed to be relevant (Casta, 2000). Besides, the development of the
efficient market hypothesis (EMH), the event study methodology and the capital asset pricing
model (CAPM) led to reconsider the object of accounting studies toward observing practices
instead of producing methods and techniques (Walker, 2004). Ball and Brown (1968) and
Beaver (1968), considered as two founding papers from which MBAR emerged, were the
pioneer in testing the decision usefulness of accounting information (accounting earnings) for
investors.
The next subsections develop the efficient markets hypothesis (EMH) research, the capital
asset pricing model (CAPM) and the Ball and Brown (1968) and Beaver (1968) studies with a
particular emphasize on the event study methodology.
1.1 The efficient market hypothesis

Stock market behavior has been for many years ago subject to various theoretical and
empirical researches in Finance. The efficient market hypothesis (EMH), popularly known as
the random walk theory was a fundamental and a controversial issue in these studies. Fama
(1965, 1970) made a major contribution in refining the conceptual framework and the
empirical testing of the efficient market hypothesis. According to Fama (1965, p. 4) in an
efficient market, on the average, competition will cause the full effect of new information on
intrinsic values to be reflected "instantaneously" in actual prices. Under this assumption,
market participants (investors, managers) would be rational and financial asset prices
(shares, bonds) would be equal to their fundamental values either because of the rationality of
investors or because the tradeoff that is likely to eliminate any price anomalies.

CHAPTER II
Shleifer (2000) argues that, from a theoretical perspective, the efficient market hypothesis is
grounded on three main arguments. First, since investors are assumed to be rational, they are
likely to better value the financial securities and each asset price will reflect its fundamental
value. As soon as asset price deviates from its intrinsic value, market participants will react
instantaneously to this new information by selling or buying assets depending on whether
they are under or overvalued. Second, to the extent that some investors are irrational, their
transactions are random and therefore mutually compensated, which is likely to cancel any
effect on prices. This argument relies on the absence of correlation between irrational investor
strategies which is quite difficult to observe. Hence, even though investors are irrational in
similar ways and their actions are correlated, they are met in the market by rational
arbitrageurs who eliminate their influence on prices.
Market efficiency is concerned with the extent to which security prices fully reflect all the
available information at any point in time. When news about a securitys value hits the
market, its price should incorporate, quickly and correctly, this information in order to meet
the new securitys intrinsic value. Quickly refers to the fact that investors who received late,
this new information should not be able to realize any profit. Correctly refers to the fact that
price should respond accurately and on average to the new information without under or
overreaction (Shleifer, 2000). The efficient market hypothesis predicts that prices should not
react to any information if it doesnt bring news about the security fundamental value. Fama
(1970) points out that different kind of information seems to be available in the financial
markets. Consequently, it gives rise to three forms of efficient market hypothesis:
-

The weak form efficiency: Current prices in the weak form of efficient market
hypothesis fully reflect available information concerning past prices and returns only.
Given that historical information about stock returns and volume is arguably the most
easily available pieces of information, it is impossible to make money or to earn a
superior profit based on such information. Under this version of the EMH, past prices
and returns cant be used to predict future prices.

The semi-strong form efficiency: The semi-strong version of the EMH states that
current prices fully reflect all publicly available information in financial markets. As
soon as the information becomes public, it is immediately incorporated into the price
and investors cannot gain by using this information to predict future prices.

CHAPTER II
Publicly available information includes not only the past prices, but other financial
and non-financial data, such as corporate reports, earnings and dividend
announcements, announced merger plansetc.
-

The strong form efficiency: The strong form of the EMH suggests that security price
fully reflects both the publicly available and private information. As a result market
participants should not be able to systemically earn superior risk- adjusted returns
even if they are trading based on private information. In other words, in a strong-form
of market efficiency, rational investors anticipate, in an unbiased manner, privileged
information and future developments. Stock prices would incorporate and evaluate
therefore the information in a much more objective and informative way than do the
insiders (Clarck et al., 2001).

In summary, efficient market hypothesis highlighted the securities behavior with respect to the
available information in financial markets. It formed the basis for the development of positive
accounting research. Hussainey (2004) argued that as financial reporting is an important
source of information about companies, it was natural for accounting research to examine
market efficiency regarding corporate disclosure. Beaver (1968) and Ball and Brown (1968;
p. 160) contended that it provides justification for selecting the behavior of security prices as
an operational test of usefulness.
1.2 The capital asset pricing model (CAPM)

Seeking whether and how the risk of an investment affects its expected return was a
fundamental issue in the modern finance. The capital asset pricing model (CAPM) provided
the first coherent framework to answer this question. Originally proposed by Sharpe (1964)
and Lintner (1965), the CAPM builds on the portfolio model developed by Markowitz (1959)
in which he argued that investors invest in a portfolio at t-1 that produces a random return at t.
They are assumed to be risk-averse and they only care about the mean and the variance of
their one period investment return. Accordingly, investors are likely to choose an efficient
mean-variance portfolio that minimizes the variance and maximize the return of an
investment. The CAPM improved the model of portfolio by introducing a testable prediction
about the relation between risk and expected return. The main implicit assumptions in the
CAPM model are: (1) all investors are likely to choose an efficient mean-variance portfolio
for one period horizon.

CHAPTER II
(2) The market is fully efficient in that there are no transaction costs and there are borrowing
and lending at a risk-free rate which is the same for all market participants. (3) Investors have
homogenous predictions for means, variances and covariance of returns. They have the same
subjective expectations for future assets behavior. In its simplest form, the CAPM assumes
that the expected return of an asset above the risk-free rate is a function of the nondiversifiable risk which is measured by the covariance of the asset return with the market
portfolio return (Bollerslev et al., 1988). The pricing kernel should be according to the
CAPM, a linear function of the market return.
The most important implication of the capital asset pricing model is that it distinguishes
between macro and micro economic events which are likely to influence stock returns. While
the market return reflects the impact of global economic events, the systematic risk shows the
sensitivity of financial assets to these events. The residual parameter indicates the impact of
corporate specific information on stock returns. This residual is qualified as an abnormal
return as it represents the unsystematic risk which will be eliminated when holding a broadly
diversified portfolio. Consequently, the risk related variation in returns is generally not of
interest for market based accounting researchers who focused first on accounting information
and its effect on the specific component of stock return (Brown and Warner, 1980). The
CAPM facilitated therefore the estimation of stock returns and the development of
information content studies relying mainly on the event study methodology.
1.3 The event study methodology

According to Campbell et al. (1997) the roots of events studies go back to the years of 1930s.
Perhaps the first published study is Dolley (1933) who examined the price effects of stock
split. Based on a sample of 95 splits from 1921 to 1931, he showed that prices increase and
decline are observed respectively in 57 and 26 cases. Until the late 1960s, consecutive studies
such as Myers and Bakay (1948), Barker (1956, 1957,1958) and Ashley (1962) produced
more sophisticated event studies taking into account the market evolution and isolating the
impact of confounding events. With the seminal works of Ball and Brown (1968), Beaver
(1968) and Fama et al. (1969), event studies emerged as a powerful and a standard tool in
assessing how new information imparts a change in equity market value. Event studies
empower the detection of an abnormal stock price effect with regard to unanticipated events.
Such events should enable investors to reassess their expectations about future cash flows.

CHAPTER II
Event studies have been employed in numerous accounting and finance research, including
earnings announcements, issues of a new debt or equity, stocks split, acquisitions and
mergers, change of management etc. The usefulness of these studies comes from the fact
that, assuming the rationality of market participants and capital market efficiency, the effects
of an event will be fully reflected in security prices.
Implementing an event study requires the adoption of a particular methodology and structure.
The specific event should be initially defined, then an event window is identified to fix the
period over which the equity price is examined. It is important to notice that the event
window should be larger than the period of interest in order to capture the presumed effect on
prices that occurs prior or lately to the events announcement (Campbell et al. 1997). The next
step aims to assess the impact of the event of interest on securitys price. The standard
approach is based on estimating a market model for each firm and on developing a measure of
abnormal returns that should reflect the stock market reaction to the arrival of new
information (McWilliams and Siegel, 1997). The abnormal return represents the difference
between actual security returns over the event window and the expected returns unconditioned
on the event taking place. Accordingly, a definition of an estimation period is required. The
most common choice is to use the period prior to the event window so that such event wont
influence the normal performance estimation.
Events studies are usually used by accounting researchers to assess the information content of
accounting event and to draw typically causal inferences from earnings announcement. If the
level of price changes around the event date, then accounting event conveys new information
about future cash ows which yields to a revision in markets previous expectations (Kothari,
2001).
Overall, it was obvious that the impetus of market based accounting research came mainly
from financial and economic disciplines. Accounting literature benefited from the emergence
of the EMH, the CAPM and the development of event study methodology. Ball and Brown
(1968) and Beaver (1968) studies are often cited as pioneer papers in this area.

CHAPTER II
1.3.1 The Ball and Brown (1968) study
The seminal work of Ball and Brown (1968) provided challenging evidence on the
information content of the earnings announcement event. Assuming the usefulness of
accounting earnings, Ball and Brown (1968) associated the sign of the unexpected annual
earnings changes with the residual stock returns in the month where the event announcement
took place. Empirical results demonstrated that a significant association exists between the
unexpected earnings and abnormal stock returns. The statistical method used for analyzing the
stock market reaction to earnings announcement corresponds to the abnormal performance
index (API). Ball and Brown (1968) traced out the value of one dollar invested in a portfolio
at the end of month -12 (12 months prior to the month of the annual report) and held till the
end of some arbitrary holding period T. The use of the API metric is intended to capture any
anticipation of information during this period, the impact of the income numbers
announcement at the period 0 and any drift after the date of announcement. Ball and Brown
(1968) noticed that most of the information (about 85 percent) contained in reported earnings
is anticipated by the market before the release of annual reports. Consequently, only 10 to 15
percent of the earnings announcement conveys new information to the market. Finding
revealed also that there is no significant association between earnings and stock returns
following the disclosure of annual reports. Ball and Brown (1968) concluded that though
annual income numbers are useful as they are correlated with stock returns, they are not the
only source of information (quarterly earnings) available to investors. Annual income
numbers are considered as lacking of timeliness in the capital market.
The Ball and Brown statistical tests were built on a specific earnings expectation model in
which earnings follow a random walk. That is, earnings should be adjusted to have zero mean
and constant variance. Current earnings are assumed to serve as a relevant proxy for expected
future earnings. The first difference in earnings reflects the surprise in earnings. Accordingly,
the new information conveyed by actual earnings can be approximated by the difference
between the actual change in income and its conditional expectation. The emphasis on
unexpected earnings yields, however, to methodological limitations since the random walk
process ignores naively all the information published after the last announcement.
Furthermore, the primary Ball and Brown (1968)s measure of unexpected earnings relied
on ex post knowledge of earnings for all firms in the market which introduces measurement
error in the unexpected earnings variable.

CHAPTER II
Ball and Brown (1968) extended their research by examining the markets reaction to good
and bad news incomes announcement. Results suggested that the relationship between the
sign of the unexpected earnings (in particular bad news) and that of stock returns residual
persisted for two months after the disclosure of annual reports which witnesses for a delay in
market adjustment. The authors undertook additional tests to assess the informativeness of
cash flows as approximated by operating incomes and whether the accrual process makes
earnings more informative than cash flows. Their finding suggested that the cash flow
variable wasnt as successful in influencing the residual stock return as the net income.
Following the Ball and Brown (1968) study, subsequent papers improved the methodological
estimation of unexpected earnings. For example, Patell (1976) used the management earnings
forecasts instead of the Ball and Brown earnings expectation model. Empirical findings
corroborated the Ball and Brown results. Patell (1976) noticed an upward price change for
firms which voluntarily issue forecasts of earnings per share. Price behavior was in the same
direction as the change in the expectation embodied in the forecast. Other researchers
extended and replicated the Ball and Brown study in different contexts such Australia (Brown,
1970), UK (Firth, 1976) and Tokyo stock exchanges (Deakin et al., 1974). These studies
examined both the sign and the magnitude of quarterly earnings announcement compared to
annual earnings reporting.
1.3.2 The Beaver (1968) study
The information content of earnings announcement was also the subject of investigation in
Beaver (1968) but with respect to both stock prices and trading volume reactions in the weeks
surrounding the announcement date. Beaver (1968) stated that corporate earnings report is
considered to be useful mainly if two new situations occur in the capital market: a change in
the equilibrium value of the current stock price and a shift in the optimal holding of that firm
stock in shareholders portfolio. The disclosure of accounting earnings is likely to help
investors revise their expectations of firms future returns which lead obviously to a new
market price equilibrium. Beaver (1968) contended that the change in expectations should be
sufficiently large to give rise to a change in investors trading behavior. As such, trading
volume is likely to be higher in the period surrounding the earnings report compared to other
periods during the year.

CHAPTER II
Beaver (1968) argued that addressing a trading volume study is due to the fact that volume
and price are two correlated variables in that the former is a natural consequence of a lack of
consensus regarding the price. In other words, since earnings report conveys new information
to the market, its interpretation may be different among investors. This is likely to increase the
trading volume until a consensus is reached. Accordingly, the volume reflects the change of
expectations of individual investors while the price reflects the change of market expectation
as a whole and in short volume will be more sensitive to the earnings announcement
compared to the stock price reaction (Beaver, 1968).
Based on a sample of 143 American firms during the period of 1961 to 1965, Beaver (1968)
findings were consistent with the information content of the earnings announcement. The
unadjusted volume analysis of market influences indicated that an important increase in
trading volume was observed during the announcement week (week 0). Investors responded
very quickly to the new information conveyed by corporate earnings and shifted their
portfolio position at the time of incomes announcement. A slightly above average trading
activity was also noticed four weeks after the announcement date. In contrast, trading volume
was below the normal average eight weeks before the earnings reports were released. These
findings suggested that investors exhibit a conservative trading strategy during this period and
that they conceive incomes numbers as useful information. Similar conclusions were
furthermore noticed in the adjusted volume analysis for market influences. The volume
residual behavior seems to have the largest peak during the week 0 compared to the nonreport periods where the volume is abnormally low prior to the earnings disclosure and
slenderly above average trading following this event. In sum, empirical tests supported the
contention that individual investors perceive earnings reports as useful information.
Beaver (1968) focused also on the security price response to the earnings announcement
whereby he conducted a price change analysis adjusted for the influence of market wide
factors. Results indicated that the magnitude of the price changes (without respect to the sign)
at the announcement week (week 0) was much higher than the average during the non-report
period. The observed change in the equilibrium price corroborated the assumption of the
informational content of incomes report. Beaver (1968) detected also a price change above
average with a lower magnitude in the week immediately prior to the earnings report. This
could be referred to the fact that earnings report was not the only source of information in the
New York Stock exchange (NYSE).

CHAPTER II
Alike these preceding conclusions, Beaver (1968) noticed price changes above the normal
magnitude for two weeks after the announcement.
In short, Beaver (1968) provided consistent evidence that investors and market expectation
were altered by the disclosure of corporate earnings as reflected respectively in the volume
and the price behavior. Nevertheless, one of the major practical limitations of this research is
that these two variables might often be positively correlated which trigger the question of how
to interpret price and volume reactions in conjunction with each other. In other words, the
raised issue is whether to attribute the increased price activity to more trading in the security
or to changes in equilibrium prices. Furthermore, Beaver (1968) founded his analysis of the
change in trading volume around earnings announcements based on the assumption that
market participants have heterogeneous expectations of future prospects. The reported
earnings would narrow these expectations and in process higher trading activity would be
observed. Recent models of investors belief revision showed in contrast that increased
heterogeneity is also possible as a consequence of new events and notably the incomes
announcement (Kothari, 2001).
Overall, the Ball and Brown and Beavers (1968) studies sparked a growing body of positive
research in accounting and provided early evidence on the usefulness of accounting data. The
decades following these researches witnessed the bulk in MBAR through which researchers
raised a wide range of questions. The informativeness of reported earnings, among other
issues, attracted major interest and the literature on the return-earnings association underwent
nowadays major developments.
The next section provides a critical review of empirical return-earnings studies.
2. A critical review of the current return-earnings association
Besides the focus on the information content of earnings announcements, another stream of
market based accounting research examined whether investors perceive accounting
information as value relevant and if it improves equity valuation outcomes. Unlike event
studies that focus on the market response to accounting reports over a short time interval,
value relevance studies consider the contemporaneous relationship between equity returns and
accounting information over a relatively long time period (one year). Since accounting reports
are presumed to provide investors with useful data, they are likely to help predict firms future
cash flows and assess future securitys risk and returns.

CHAPTER II
This is certainly why the relationship between stock returns and accounting earnings has been
for more than thirty years, one of the most widely researched topics in accounting mainly in
USA and Europe. First documented in the pioneering work of Ball and Brown (1968), the
aforementioned relationship tests for a positive association between accounting performance
measures (earnings, earnings per share or cash flows from operations) in one hand and prices
or stock returns in the other hand. The main objective of this study is to examine whether
accounting figures summarize and catch timely changes in the set of information that will be
reflected in stock returns (Kothari, 2001). Association studies do not infer a causal connection
between stock price movement and accounting figures. This is because accounting data are
not assumed to be the only timely and relevant information for investment decision and
valuation process. Market participants may have access to richer and various sources of
information in addition to accounting earnings (Dumentier and Raffounier, 2002).
Subsequent to Ball and Brown (1968), the literature on the returns-earnings relationship was
characterized by a continual refinement and a considerable methodological sophistication.
The nave earnings expectation model exhibited notable ingenuity and improvement. In
general, empirical studies on the returns-earnings association measured the intensity of the
correlation between earnings changes and security returns to determine how earnings changes
summarize the information incorporated into stock prices. The underlying logic in these early
returns-earnings association studies is that any revision in stock prices provides evidence on
earnings usefulness. The informational contribution of earnings to investors is perceived to be
significant if earnings exhibit a considerable explanatory power with respect to stock price
change.
The economic intuition of the relationship between accounting data and market value is based
on a benchmark model in which share price is the discounted present value of future net cash
flows. This valuation model relies on the assumption that current earnings provide
information about expected net cash flows (e.g. Watts and Zimmerman, 1986, Ch.2;
Kormendi and Lipe. 1987, Ohlson, 1991). Nevertheless, since market expectations about
future cash flows are unobservable and difficult to assess, empirical specification of the
prices-earnings association refers often to current earnings as a proxy for market expectation
about future cash flows. Basically, the specified model produces an estimation of an earnings
response coefficient (ERC) and an R- square as a measure of the magnitude of the correlation
between earnings and returns.

CHAPTER II
The ERC is used as an indicator of accounting earnings timeliness with regard its covariance
with market values. The R2 measures the extent to which that accounting number is related to
market value and hence its value relevance. Timeliness can be described as timely financial
information that reflects value relevant events as early as possible. Improved timeliness,
results in stronger relationship between accounting data and market values. This leads to
higher value relevance of accounting information (Luberrink, 2000).
Research on earnings response coefficients and value relevance was devoted, among others,
to test for contracting and political cost hypotheses, voluntary disclosure and signaling
hypotheses in financial accounting. A long line of empirical research demonstrated that
accounting earnings are related to share returns (Easton and Harris, 1991; Das and Lev, 1994;
Liu and Thomas, 2000) though a general problem in such association is still of concern for
accounting researchers. Lev (1989) conducted a comprehensive review of market based
accounting research on the quality of reported earnings and concludes that the explanatory
value of earnings for share returns, and subsequently the usefulness of earnings disclosures,
tends to be very low and sometimes negligible.
The next subsections are devoted to a review of this early earnings response coefficient
literature with a particular emphasize on the several explanations that was suggested for these
disappointing results.
2.1 The earnings response coefficients research

In earnings response coefficient (ERC) studies, researchers make inferences about the
usefulness of earnings from the slope coefficient on earnings in the returns-earnings
regression. The ERC is sometimes referred to as the constant marginal response of prices to
earnings in a linear regression (Freeman and Tse, 1992). As Collins and Kothari (1989)
explained, these studies can be either short-window or long-window tests. The short-window
tests are basically event studies in which researchers infer whether the earnings announcement
causes investors to revise their expectations about future cash flows over a short time period
(typically, 2-3 days) around the earnings announcement. This is because in event studies, it
seems reasonable that a very narrow window ensure that the price change around the event
announcement will be mostly related to the new information conveyed by reported earnings.
The estimated slope coefficient from regressing abnormal returns on unexpected earnings
(ERC) is considered as a test of information content in the market based accounting literature.

CHAPTER II
Examples include Foster (1977), Hagerman et al (1984), Wilson (1986, 1987), Ball and
Kothari (1991), Amir and Lev (1996), and Vincent (1999) etc..
Other studies investigated the time series of long-window returns-earnings associations to
measure changes in the usefulness of accounting earnings and to check whether annual
earnings summarizes timely information that is reflected in stock prices. Collins and Kothari
(1989) suggested that in an association study, returns over relatively long periods are
regressed on unexpected earnings or other performance measures (e.g. Cash flows) estimated
over a forecast horizon that corresponds roughly with the fiscal period of interest. The authors
added that association studies recognize that market agents learn about earnings and
valuation-relevant events from many non-accounting sources of information throughout the
period. Thus, these studies investigated whether accounting earnings measurements are
consistent with the underlying events and with the information set reflected in stock prices.
This is the accounting relevance literature initiated by Kormendi and Lipe (1987), and
continued in papers such as Lev (1989), Collins and Kothari (1989), Francis and Schipper
(1999) and Lev and Zarowin (1999). Lev (1989) argued that the issue of time interval
(window) over which stock returns are cumulated is important. The usefulness of earnings
may be understated when performing a regression of returns cumulated over a narrow
window around earnings announcements. This is particularly the case if the narrow window
fails to detect earnings-induced price changes beyond the windows (e.g. A delayed investor
reaction after the announcement-post announcement drifts). Regressions using wide windows
may also overstate the incremental information contribution of earnings as price changes
within the window is likely to reflect the investors' reaction to a myriad of other timely, nonearnings information (e.g., industry-wide events or stock splits and repurchases), which are
correlated with earnings. Accordingly, the window length in the returns-earnings studies helps
focus on either the unique information content of earnings over other sources of information
or on the extent to which earnings are just correlated like other, perhaps more timely, valuerelevant information items with stock returns (Lev, 1989). Kormendi and Lipe (1987), an
earlier paper on the earnings response coefficients research, built on the strong evidence
provided by a large body of accounting literature in support of the value relevance of reported
earnings. Kormendi and Lipe (1987) implemented new tests of the information contained in
accounting earnings by mainly focusing on the effect of earnings time-series properties
(permanent and transitory earnings) on the magnitude of the relationship between stock
returns and earnings.

CHAPTER II
They addressed the valuation implications of the persistent component of earnings as an
attempt to refine seminal studies in this area. The authors drew on the standard valuation
model under which a firm's stock price equals the present value of the expected future
benefits accruing to its equity holders. They argued that if earnings innovations are permanent
they should urge investors to revise their expectations about future earnings and cash flows.
Accordingly, the earnings response coefficient can be estimated as the discounted value of the
earnings innovation perpetuity at the risk adjusted rate of return on equity.
According to Lev (1989) time series properties of earnings play a role in making easier the
revisions of earnings forecasts based on current earnings though the absence of rigorous
theory. The focus on earnings persistence in the returns-earnings association aims to cope with
the constant response coefficient assumption in cross sectional studies. Investors are likely to
react differently to reported earnings as long as they perceive permanent component of such
earnings as distinct cross sectionally and over time. Easton and Zmijewski (1989) ascertained
empirically the above mentioned assumptions. Their study revealed a positive correlation
between earnings persistence and the ERC in the returns-earnings regression. Such results
were whereas subject to few limitations due to some bias in defining the research variables.
Alike previous research, the use of a nave market expectation model of future earnings
may lead to some measurement errors as it doesnt take into account the presence of
alternative information (other than currentperiod earnings) that can be more relevant for an
investment decision. Furthermore, the time series properties of earnings and notably the
magnitude of earnings innovation (persistence) may depend on some firms economic
characteristics such as technology, competition, innovation, etc. that their inclusion in the
multivariate regression may influence the earning response coefficient estimation.
2.2 The economic determinants of the earnings response coefficient

Collins and Kothari (1989) extended Kormendi and Lipe (1987)s study by examining
additional factors that explain both cross-sectional and inter-temporal variation in
ERCs. Based on a discounted dividend valuation model, the authors tested whether
differences in the information environment in the stock market understate significantly the
magnitude of correlation between price changes and earnings changes. Empirical results
ascertained that the earnings response coefficients exhibited a significant increase in growth
opportunities and persistence and a considerable decrease in risk.

CHAPTER II
Their findings suggested then that the return-earnings relationship varies cross sectionally and
over time as a function of some economic characteristics.
Easton and Zmijewski (1989) argued that a greater risk implies a larger discount rate, which
reduces the discounted present value of the revisions in expected future earnings. They
asserted that the greater persistent the time series properties of earnings, the larger will be the
earnings response coefficients. Empirical evidence supported these assumptions: the
magnitude of earnings innovation and the degree of the systematic risk lead the earnings
response coefficients to vary between firms and inter-temporally.
Kothari (2001) claimed that various studies reproduced these findings and it became a
benchmark to the market based accounting research to control for the effects of risk, growth,
persistence. Other researchers focused on the incremental effect of additional factors such
as auditor change, financial leverage, earnings management and firm features such as size,
age, foreign currency reliance, government ownership Wikil (1990) and Teoh and Wong
(1993) studied respectively the effect of audit change and auditor reputation on the quality of
accounting earnings as approximated by the earnings response coefficients. They assumed
that firms changing their auditors as well as big eight audited firms would exhibit a
significant correlation with the ERC. Results revealed a significant relationship between
auditor quality and ERC whereas auditors changes didnt impact significantly firms earnings
response coefficients. Their findings were robust to the inclusion of other explanatory factors
of the ERC that were suggested in previous studies such as growth, persistence, risk, firm
size, and pre-disclosure information environment. Dhaliwal, Lee and Fargher (1991) and
Dhaliwal and Reynolds (1994) focused on the relationship between unexpected earnings and
stock returns and the effect of default risk on it. They demonstrated that financial leverage as
well as bond ratings are significantly and negatively related to the earnings response
coefficients. Specifically, they showed that ERC of firms having low financial leverage was
larger than firms with high financial leverage. Feltham and Pae (1999) examined carefully the
impact of accounting accrual items on earnings uncertainty and the ERC. Their analysis
showed that depending on the impact of earnings management on the persistent element of
unexpected earnings, unexpected earnings will vary which would cause the earnings response
coefficient to change.

CHAPTER II
According to Watts (1992) a main criticism of researches on earnings response coefficient
determinants is that it did not explore adequately economic variables based on contracting or
accounting choice theory literature. Notably, ERC studies did not take into account the
differences in accounting methods that is likely to jeopardize the ability of accounting
earnings to proxy for current and future cash flows.
Earlier research on the return-earnings relationship raised additionally the issue of the time
interval over which returns are cumulated. Both narrow and large windows were practically
investigated in order to point out which case will trigger a perfect association between the two
variables of interest. Nevertheless, varying the window length from less than a week to one
year (Beaver et al., 1980; Hopwood and McKeown, 1985; Bowen et al., 1987; Wilson, 1987)
yielded to a very low correlation between unexpected earnings changes and returns as
measured by the R2 and the slope coefficient (Lev, 1989). Typically empirical findings
revealed a much lower explanatory power of earnings with an ERC around 2 and a R- square
value that rarely exceeds 10%. Lev (1989) ascribed this low statistical association to the loss
of relevance and the low quality of accounting earnings and challenged researchers to identify
the potential explanations for the weak contemporaneous return-earnings relationship.
2.3 Competing explanations for the low association in the returns-earnings studies

Based on prior MBAR, Lev (1989), Kothari (2001) and Walker (2004) identified at least four
potential reasons that may explain the low magnitude of the earnings response coefficient and
the lack of a perfect fit in the returns-earnings relationship. These are: (1) market inefficiency,
(2) stock markets react to value-relevant information that is not observed by the company, (3)
noise in reported earnings, and (4) earnings may lack of timeliness.
2.3.1 The market responds inappropriately to the earnings information
This first explanation refers to market efficiency issue with regard to accounting information
and particularly reported earnings. This concept is related to the extent to which the market is
able or fails to take into account differences in accounting methods and choices when valuing
companies. Walker (2004) argued that accounting earnings measurement relies on some
assessment procedures such as the choice between different inventory valuation and
depreciation methods. Such choices should not in a rational market influence the firms future
cash flows.

CHAPTER II
Otherwise, if the market does not take into account differences in accounting methods, then
differences in firm value which are not related to a rational estimate or to changes in future
cash flow will be obvious. This phenomenon, often known as the functional fixation
hypothesis, maintains that individual investors interpret earnings numbers the same way,
regardless of the accounting procedures used to calculate them. For example, they do not see
through the earnings effects of different accounting methods such as LIFO versus FIFO and
straight-line versus accelerated depreciation.
Considerable studies raised the issue of market efficiency with respect to the functional
fixation hypothesis. Kothari (2001) provided a brief review of the earlier literature in this
area. He suggested that differences in accounting methods are likely to produce differences in
reported financial statement numbers, but without any impact on a firms cash flow. Empirical
research, such as Beaver and Dukes (1973), Lee (1988) and Dhaliwal et al. (2000) did not
notice systematic or large differences in the prices of companies using different accounting
methods (accelerated versus straight line depreciation methods, LIFO versus Non LIFO
firms). Their findings rejected the market fixation hypothesis and provided significant
evidence congruent with the efficient capital market hypothesis.
Notwithstanding these early evidences, Vincent (1997) and Jennings et al. (1996) struck
discordant notes. They compared the price to earnings ratios of firms using the pooling
method with those using the purchase method for mergers and acquisitions. They observed
measurable differences in risk-adjusted stock returns between these two types of firms. The
authors linked these abnormal returns to market fixation hypothesis on accounting numbers
consistent with some degree of market inefficiency.
Another stream of research in the accounting and the finance literature focused on the
predictability of abnormal returns following earnings announcements and documented various
capital market anomalies. Known as the post-earnings announcement drift literature, these
studies revealed that the market under/over reacts to earnings news. This result suggested
possible trading rules that benefited from the knowledge of the mis-priced securities and
generated significant abnormal returns. The pioneering works in this area are Bernard and
Thomas (1989, 1990) and Debondt and Thaler (1985, 1987). Their findings showed that the
drift is positively (negatively) correlated with the good (bad) earnings news portfolio,
consistently it occurs surrounding the earnings announcement and lasts up to a year.

CHAPTER II
These properties pose a serious challenge to the efficient capital market hypothesis and to the
equilibrium asset pricing model. Many subsequent studies attempted to find out whether the
drift is incremental to other anomalies, including size, trading volume, industry, past sales
growth. They concluded that these factors do not subsume the drift. Overall, the survival of
the anomaly for more than 30 years after its discovery still puzzling researchers. Evidence
based on a rational explanation seems to be yet elusive.
2.3.2 Noise in reported earnings and deficient GAAPs
Considerable studies investigated the return-earnings relationship under the assumption that
accounting earnings are the aggregate value of true earnings and value-irrelevant noise
component. The former is assumed to be perfectly correlated with current and lagged returns,
whereas the latter is uncorrelated with stock returns in all periods.
The noise in earnings argument has gained currency in the academic literature since Beaver et
al. (1980). They intuitively proposed a model in which accounting earnings Xt contain a
garbling component ut that is unrelated to firm value. Within the developed framework,
Beaver et al. (1980) expressed the predictive ability of prices with regard to future earnings by
assuming that the value-relevant component of earnings xt (= Xt - ut) is observable and reflects
the same information set that is in the stock prices. Following Beaver et al. (1980), subsequent
papers (Choi and Salamon, 1990; Kothari, 1992; Khotari and Zimmerman, 1995; and Ryan
and Zarowin, 1995) ascertained that the earnings response coefficient is weakened due to the
presence of a price- irrelevant component in reported earnings. This latter differs typically
from permanent earnings for a number of reasons. Some of these could be linked to the
managers discretion over annual reports. Others follow from the correct application of
accounting principles known often as the deficient GAAPs argument. For example, many
GAAP regimes are against the anticipation of future revenues and the recognition of certain
types of assets. Rather than a perfect matching of expenses with revenues, some accounting
standards tend to recognize some expenses earlier than revenues. Firms are required to treat
their investments in R&D as revenue expenditure. Consequently, reported earnings for
growing firms with a high level of R&D would be biased downward compared to permanent
earnings. Such conservative attitude imparts a scrambled version of true earnings and an
otherwise signal about firm value (Walker, 2004).

CHAPTER II
Therefore, the deficient GAAPs argument suggests that financial statements are slow to
incorporate information while it is already reflected in contemporaneous market values. The
higher the association of incomes with stock returns, the more desirable the accounting
GAAPs are (Kothari, 2001). In addition to the delayed recognition in accounting earnings,
noise in reported earnings could be also related to the presence of transitory items. Some
business activities such as asset sales or the disposal of a subsidiary are likely to produce one
time losses or gains. Under some GAAP requirements, companies will report these gains or
losses at once in the income statement at the end of the financial year which means that
reported earnings will include large transitory items in a specific year.
A large body of accounting literature (e.g. Beaver et al, 1979; Hayn, 1995; Basu, 1997)
attempted to infer the transitory earnings effect on the earnings response coefficient. Results
revealed a smaller magnitude of the ERC if non-recurring items were disclosed in financial
statements. They showed that the extent of abnormal returns associated with extreme earnings
changes is not proportionally as large as those associated with non- extreme earnings change.
These results implied a non-linear return-earnings relationship. Kothari (2001) argued that the
market does not expect extreme earnings change to be permanent, so the price adjustment
would be smaller. This suggests a negative correlation between the transitory earnings and the
earnings response coefficient. Managerial discretion over reported earnings through
discretionary accruals might be one of the reasons behind the difference between accounting
earnings and true or permanent incomes. According to Holthausen (1990), discretionary
accruals management can be linked to a number of conceptual motives: opportunistic
behavior, efficient contracting, and information provision. First, the opportunistic behavior
perspective holds that managers benefit from the set of alternatives permitted by GAAP.
Corporate management tends to manipulate accounting earnings in order to maximize their
own utilities at the expense of the contracting parties and stakeholders. Investors are thereby
misled by the unreliable reported information. Second, efficient contracting theory suggests
that accounting method choices are selected to minimize agency costs among firms
stakeholders. This optimization will theoretically maximize the value of the firm and brings
reported earnings closer to periodic performance. The information perspectives, finally,
indicate that managers would be compensated when providing information about firms future
prospects. Accounting methods are chosen thereby to reveal managers expectations about
companies future cash flows which are informative about economic earnings.

CHAPTER II
Considerable studies (e.g. Barth et al., 1996; Beaver and Angel, 1996; Nelson, 1996)
investigated the efficient contracting and the opportunism hypotheses by correlating earnings
component (discretionary and non- discretionary accruals) with stock returns. Their main
purpose is to test the association of new mandated accounting standards (disclosure,
recognition criteria) with stock returns and to identify earnings components that relate to
managerial accounting choices. Empirical findings experienced first difficulty in
distinguishing which of these motivations provides additional reasons why reported earnings
didnt accurately reflect firms permanent earnings. They showed besides that the market did
not correctly price total accruals and the mis-pricing is particularly attributable to the
discretionary components. Despite these interesting achievements, such conclusions seem to
be at a first glance counterintuitive. A body of research, including Dechow (1994) and
Subramanyam (1996) tested the discretionary accruals association with security returns. They
concluded that discretionary accruals are on average informative and not opportunistic. There
is also no economic intuition suggesting that a transaction based approach would generate true
income that capture all the information set in security returns rather than accounting based
approach.
2.3.3 The market price movements respond to value relevant information not observed by the
company
The assertion that share price movements may be related to some information that is observed
by the market participants and not available to firm management was initially demonstrated in
the theoretical paper of Dye and Sridhar (2002). The authors argued that the information
flows between firms and capital markets not only in one way (from firms to the capital
market) but also occurring in the reverse sense. Managers are likely to build their strategic
decision based on shareholder valuation of firms activities which is reflected via a rational
share price movement. Dye and Sridhar (2002) paper was premised on a number of
assumptions. They first postulated that the resource allocation process is affected by prices in
the capital market where participants are experimented in predicting the valuation implication
of firms' past and anticipated decisions. The authors suggested second that the informationaggregating properties of capital market prices make them more likely to contain information
not known to managers.

CHAPTER II
Dye and Sridhar (2002) contended that firms may assess their strategic decision by disclosing
a project of strategy change that would be implemented if only the size of the price reaction to
this announcement was sufficiently favorable. This strategic directing disclosure will reduce
hence, if it works, the frequency of implementing undesirable strategy changes. Such
reflection posits that capital market prices perform simultaneously a double role: the
conventional role of predicting future cash flows implied by management activities and
directing these actions toward the highest cash flows generating decisions. Dye and Sridhar
(2002) showed that managers seek guidance from capital markets mainly in initiating a new
strategy proposal or in deciding whether to continue or to abandon a previously initiated
project. Such analysis applies well when the financial viability of a project is sensitive to
macroeconomic or industry-wide trends that capital market participants are especially good at
assessing. Making a firm's resource allocation contingent on the price reactions to its
disclosure may be therefore beneficial since capital markets possess an informational
advantage over managers. It is worth mentioning that the early reveal of a firms strategy will
be dropped out if it might reveal proprietary information to competitors or if the expected
returns for the new project are so high. Managers may also renounce to the strategic directing
disclosure if they are entrenched or if it may cause damage to their reputation due to reversing
previously announced strategic initiatives.
Overall, accounting information might be irrelevant for capital market participants since they
are likely to be more informed and to possess more sophisticated information compared to
corporate managers. As such, the informational contribution of reported earnings might be
questioned and a lower earnings response coefficient will be observed compared to its
theoretical value. To date, despite these interesting observations and analyses, there is no
substantial empirical evidence on this issue. The disclosure directing strategy is still an
intriguing area for market based accounting research.

CHAPTER II
3. Lack of timeliness concern in the current return-earnings association
Earnings lack of timeliness phenomenon was over the last two decades a critical subject of
investigation in a number of published papers. Taken together, these studies provided
overwhelming evidence that reported earnings are estimated in a way that they differ
considerably from permanent earnings (Walker, 2004).
The focus on earnings timeliness by academic researchers intended to identify potential
reasons for the low quality of accounting earnings as challenged by Lev (1989)s survey.
Accounting literature was particularly concerned with the extent to which stock market
participants have access to value-relevant information other than reported earnings. Earnings
lack of timeliness may be related to the fact that these value-relevant events are reflected in
stock prices as soon as the information reaches the stock market, while their impact on
reported earnings often occurs with a time lag (Schleicher, 1996). This is parsimoniously
referred to as prices lead earnings. Many empirical studies attributed this lagged reaction of
earnings to some accounting conventions such as objectivity and conservatism and
recognition criteria of some economic events and revenues.
The next subsections develop the underlying reasons and the main empirical works that dealt
with lack of timeliness in reported accounting earnings.
3.1 Lack of timeliness and recognition in reported earnings

The conceptual framework of the international financial reporting standards (IFRS) defines
accounting recognition as the process of incorporating in the balance sheet or income
statement an item that meets the definition of an element and satisfies some criteria for
recognition [F 4.37 and F 4.38]. The decision as to when a transaction will be recognized in
corporate financial statements depends therefore upon various measurements and interrelated
recognition concepts. In particular, accounting for an economic event should respect some
recognition criteria and namely realizability, reliability, verifiability, objectivity, matching and
conservatism. Applying these fundamental concepts assumes that corporate transactions
should be completed and the realized benefits should be measured without any uncertainty.
This may not be the case in many situations, particularly for some events related to
contingencies, long term guaranteed contracts, short term marketable securities, inventory or
human capital. For example, economic events which existence is uncertain require a particular
accounting treatment. In case of loss contingencies, no expenses are recorded unless the
accuracy of loss is probable and the amount can be measured with reliability.

CHAPTER II
Gain contingencies are, in contrast, rarely recognized prior to their realization and are only
disclosed when the probability of their accuracy is very high. The accounting recognition
process has also an important role in evaluating firms assets. The accounting measurement
bases for both short-term marketable securities and inventories deviate from the historical cost
principals with a conservative manner. Notably, assets are generally reported at historical cost
unless their market values are lower than cost. In this case holding losses are reflected in
accounting earnings. In contrast, in a situation where assets historical costs exceed their
market values, holding gains are not recorded until they are sold. Unlike accounting
recognition rules that emphasize on historical cost measurement and transaction based
revenue, the market takes into account these events and revises more timely its expectations
for firms future cash flows. Therefore, earnings numbers will exhibit a recognition lag
compared to stock prices (Warfield and Wild, 1992). Other economic events entail also a
delay in the accounting numbers picking up changes in the value of net assets or in the
expectation of future benefits. These events include the discovery of an oil reserve, Food and
drug administrations approval of a new drug, advertising and research and development
expenditure in the current period. While they have not met the condition for accounting
recognition, such events affect investors prediction of future cash flows and hence stock
returns (Collins et al., 1994).
Overall, the traditional accounting accrual system often trades off and even scarifies
timeliness in recognizing changes in net asset values in favor of objectivity, verifiability or
conservatism. Some economic transactions that do trigger market revisions about future
earnings are not captured in current periods earnings and will be reflected in future periods
when the conditions of recognition are satisfied. This implies that price changes for a given
period will be linked to future periods as well as current period, thereby weakening the
contemporaneous association in the return-earnings relationship. Earnings timely loss
recognition concept gets more at the heart of the distinction between the quality of the
fundamental earnings process and the ability of the accounting system to measure the process
than the other earnings quality proxies (Dechow et al., 2009 p. 13). A strong association of
earnings with negative stock returns seems to be linked to the use of financial reporting
standards rather than to be a characteristic of the process itself.

CHAPTER II
The use the returns-earnings relationship as a measure of earnings quality requires the
consideration of two significant issues. First, market efficiency hypothesis that underlies the
interpretation of returns-based metrics as a measure of earnings informativeness didnt seem
to hold equally across markets and as a consequence induce omitted variable bias. Second,
stock returns tend to reflect all information, not just information in accounting earnings,
which may also lead to some omitted correlated variables problem. Various empirical studies
attempted to avoid the above mentioned problems by introducing major improvement in the
statistical return earnings relationship. Collins et al. (1994) is a notable study in this area.
3.2 Empirical evidence on the low current return-earnings relationship and suggested
refinements

Assessing earnings timeliness was approached in most capital market researches via two
main methods depending on which issue is under consideration. The first approach was
developed initially in Beaver et al. (1980) whereas the second approach for measuring
earnings timeliness is developed in Collins et al. (1994).
Beaver et al. (1980) was among the seminal papers that raised theoretically the idea that the
information set reected in US stock prices is richer than that is in the current accounting
earnings. They empirically used a reverse regression under which current deflated earnings
are regressed on current and past stock returns. Their empirical findings showed a significant
coefficient on the lagged price variable consistent with the concept of prices leading earnings.
Following Beaver et al. (1980), there was a voluminous work on the implication of prices
leading earnings providing important insights on this issue. A list of these papers includes
Beaver, Lambert and Ryan (1987), Collins and Kothari (1989), Kothari (1992), Kothari and
Sloan (1992), Easton et al. (1992), Wareld and Wild (1992) and Basu (1997) etc.. Beaver
et al. (1987) and Freeman (1987) built upon the empirical analysis in Beaver et al. (1980) and
reported evidence broadly consistent with Beaver et al. (1980). Their regression results
indicated that stock returns in one year contain information about earnings in the following
years and ascertained that the stock prices of large firms anticipate future earnings earlier than
the stock prices of small firms. Collins and Kothari (1989) extended these early researches by
investigating the cross-sectional and inter-temporal determinants of the earnings response
coefficient. They applied a reverse regression to a number of assumptions and factors
contributing to cross-sectional and inter-temporal differences in the ERC.

CHAPTER II
Empirical findings revealed that the earnings response coefficient is positively correlated with
earnings persistence, economic growth opportunities and firms size (large versus small firms)
as a proxy for the firms information environment. They also showed that the ERC is
negatively related to the risk-free interest rate and CAPM beta risk. Moreover, they found that
the ERC is negatively associated with the interest rate through time. Collins and Kothari
(1989) argued that the conventional regression which models returns over the 12-months
period seriously underestimates the extent of the association between returns and earnings
news. They also demonstrated an improvement in this relation, by starting the return
measurement period earlier than the contemporaneous fiscal period.
Kothari (1992) examined alternative specifications in the price-earnings relationship when
prices lead earnings. He investigated the association between prices and earnings in levels and
changes depending on whether the earnings variables are deflated by price or earnings. The
empirical analysis in Kothari (1992) demonstrated that if one assumes prices do not lead
earnings, then the degree of bias in the earnings response coefficient estimates and the
explanatory power of the specified model (in level or change) would rank similarly. In the
presence of price leading earnings, the level specification yields a less biased earnings
response coefficient estimates and higher R2 compared to the change specification.
Kothari (1992) suggested that an accurate proxy for earnings markets expectation may be
difficult to obtain, though, earnings level deflated by price might be the best available
variable for use in the price-earnings regression. To test empirically the ideas in Kothari
(1992), Kothari and Sloan (1992) analyzed the presumed inconsistency between the time
series behavior of annual earnings and the market response to earnings changes. The authors
assumed that stock returns over a period reflect a rich information set which is likely to
induce a revision in market expectation about future cash flows. Accounting earnings have in
contrast a limited ability to contemporaneously reflect such revised expectations. They
attributed this limited ability to some accounting conventions that underlie the application of
accounting standards such as conservatism, objectivity and verifiability etc.. Consequently,
the authors expected returns to lead earnings changes. They suggested that including the
leading-period return in the statistical regression should ameliorate the mis-specification and
increases the earnings response coefficients. Findings indicated that the earnings response
coefficient is sensitive to the inclusion of the leading period returns which is in such case
more than doubles.

CHAPTER II
They also found that US stock prices anticipate future earnings changes up to four years ahead
and that there is an improve in ERC when measured over long intervals.
Warfield and Wild (1992) and Jacobson and Aaker (1993) were also among the early few
papers that tried to improve the weak statistical return-earnings association. Based on price
lead earnings assumption, the authors included next period earnings as additional explanatory
variables in the regression model. They aimed to assess the incremental contribution of these
future period earnings in explaining current returns. Empirical findings showed that the
adjusted R2 from their regression increases substantially when the immediate future period
earnings is added.
In a further attempt to investigate the lack of timeliness of accounting earnings, Basu (1997)
used reverse regressions in the analysis. He suggested that test statistics are better specified
when the leading variable is specified as independent and the lagging variable as dependent.
Basu (1997) developed four main predictions. He argued first that accounting earnings reflect
bad news more quickly than good news. For instance, unrealized losses are typically
recognized earlier than unrealized gains. He asserted second that the concurrent earningsreturns association is relatively stronger than the concurrent cash flows-returns association for
publicly available bad news compared to good news. Basu (1997) predicted third that good
news is likely to be persistent while bad news is more likely to be temporary. Finally, he
presumed that the short window earnings response coefficients are smaller for bad earnings
news than good earnings news. Empirical findings corroborated all four predictions. Basu
(1997) showed that earnings, the dependent variable, contain more timely information for 'bad
news' firms. This result implies that earnings are contemporaneously more sensitive to
negative unexpected returns than positive unexpected returns, as measured by the slope
coefficient and the R square. Findings revealed also that earnings recognize timely bad news
through accounting accruals compared to cash flows. Basu (1997) demonstrated finally that
positive earnings changes tend to persist whereas negative earnings changes show a
marked tendency to reverse. As a consequence, such difference in persistence implies that
positive earnings changes have higher ERCs than negative earnings changes.
Despite the compelling evidence on earnings lack of timeliness in recognizing some
accounting events, earlier studies in this area did not hold out a robust improvement for the
contemporaneous return-earnings association.

CHAPTER II
The failure to use an accurate proxy for unexpected earnings in the presence of prices leading
earnings could not be considered as a plausible explanation for the low magnitude of the
ERC. Collins et al. (1994) argued when discussing Warfield and Wild (1992) and Jacobson
and Aaker (1993) papers that including future period earnings in the return-earnings
regression is subject to an errors-in-variables problem that biases the ERCs and explanatory
power downward.
Collins et al. (1994) investigated two potential factors contributing to the weak association
between returns and current earnings. These factors fall into two main broad categories:
earnings lack of timeliness in capturing value-relevant events (e.g., Kothari, 1992; Kothari
and Sloan, 1992) and (2) the presence of irrelevant noise in reported earnings (e.g., Beaver et
al., 1980). To discriminate between these competing hypotheses, Collins et al. (1994) argued
that if earnings lack of timeliness, then stock returns of one time period would be related to
earnings growth rates in the current and future time periods. Value irrelevant noise would be
in contrast uncorrelated with stock returns in the current, past and future periods. To highlight
the intuition that prices are leading earnings, Collins et al. (1994) contended that, in an
efficient market, stock returns are driven by the unexpected current earnings and any new
information that leads the market to revise its expectations about future earnings. Earnings
lack of timeliness implies therefore, that some information in current earnings is already
anticipated by past prices and hence uncorrelated with current returns. Earnings change in
period t is considered as a noisy proxy for the surprise component of the current earnings.
This is certainly why current returns-earnings relationship is weakened and biased downward.
According to Collins et al (1994), this problem can be corrected by introducing lagged
earnings yield as a proxy for expected current earnings growth. The authors also suggested
that current earnings are not likely to reflect all the news received in the current period that is
likely to change market expectation about future earnings. The contemporaneous returnearnings association would be once again weakened. Hence, in order to improve the goodness
of fit of such relationship, there is a need to introduce proxies for any news the market
receives about future earnings changes during period t. Collins et al. (1994) explained how
this can be achieved by introducing future earnings changes and future returns as additional
variables in the return-earnings model. Such improvement will stand consequently only if the
reason for the poor performance of the simple return-earnings regression is prices leading
earnings.

CHAPTER II
In other words, if the value-irrelevant noise is the reason behind the poor statistical
performance of the standard return-earnings model, then the goodness of fit of this model will
not be improved by adding these proxies.
The above discussed proxies are intended to mitigate the various measurement errors in the
traditional return-earnings model. Collins et al (1994) relied on some economic rationale that
upholds the relevance of these explanatory variables. The first measurement error that proxies

for expected next period earnings change is the lagged earnings yield variable

EPt 1

EPt 1

is defined as the period t1s earnings over price at the start of the return window for period t.
Collins et al. (1994) asserted first that price impounds information about future earnings and

EPt 1
the

should proxy for the markets expectation of earnings growth. Second as prices

EPt 1
incorporate information about future earnings, a low

signals high expected earnings

growth for the current and future years. Third, given that the earnings yields variable and the
expected earnings growth (the measurement error) are negatively associated, the coefficient

EPt 1
on

should be positive. This is true because this measure serves to subtract the noise

element from realized earnings growth. The additional proxy is the growth in a firms book

AGt
value of assets,

. This variable serves as another measure for expected future earnings

growth. In fact, if managers have expanded recently their production capacity in expectation
of higher demands for their product in the future, then it should lead to higher expected
earnings growth. Asset growth and expected future earnings changes are as such positively

AGt
associated and the coefficient on

is forecasted to be negative. The last measurement

Rt k
error that proxies for unexpected future earnings is the future periods returns,

. In an

efficient market, some unanticipated events that will occur in future periods and will affect

CHAPTER II
Rt k
future earnings will be reflected in future returns

. If these unanticipated events lead to

higher (lower) earnings growth in period t+k, they should also lead to positive (negative)
returns in the period when the news becomes available to the market. Consequently, a positive
relation between unexpected future earnings and future returns is expected to yield negative
coefficients on the return variables. This is likely to provide a better approximation to the
change in expectations of future earnings growth that occurred in period t.

Collins et al. (1994) used the expanded return-earnings model to discriminate between the two
potential reasons (noise versus lack of timeliness) for the weak contemporaneous association
at three levels of aggregation: economy level, industry level and firm level. The intuition
behind is that the implied effect of data aggregation on the current return-earnings relationship
is different under the two hypotheses. The authors argued that the contemporaneous
association between returns and earnings should strengthen under the noise-in-earnings
hypothesis particularly when data are aggregated. This is due to the fact that, as data are
aggregated, the weakly cross-sectional correlation noise will be diversified away.
Accordingly, only the value-relevant earnings component in the aggregated data will remain
and this would be highly correlated with contemporaneous stock returns. In contrast, under
the lack-of-timeliness hypothesis, cross-sectional data aggregation would not be helpful in
improving current earnings timeliness.
Collins et al. (1994) expansion of the traditional return-earnings association yielded a
considerable increase in the explanatory power of the regression model. The explanatory
power of the current return-earnings regression (R2) increased from less than 15% to
approximately 35-50% at all levels of aggregation when including future earnings growth
variables. The authors showed that earnings lack of timeliness is the most important factor
contributing to the low contemporaneous return-earnings relationship. As such, since current
and future earnings measures are useful in explaining current returns, Collins et al. (1994)
results confirmed that accrual-based earnings measures do not capture value-relevant events
in a timely manner. Collins et al. (1994) found also that the noise in earnings does not seem to
be a major factor in explaining the low contemporaneous return-earnings association.

CHAPTER II
Overall, the difference in the adjusted R2 between the traditional and the modified returnearnings model in Collins et al (1994) paper provided a direct measure of the extent to which
the market had access to other relevant information beyond revealed earnings. Walker (2004)
argued that, holding constant the quality of earnings, one should observe a correlation
between the lack of timeliness in earnings and the quality of disclosed information which is
relevant for predicting future earnings. A few recent attempts to exploring this idea (using the
model of Collins et al, 1994) includes Schleicher and Walker (1999), Gelb and Zarowin
(2002), Lundholm and Myers (2002), Hussainey et al. (2003), Schleicher et al. (2007) and
Hussainey and Walker (2009) etc. They found some evidence that firms with high level of
disclosure exhibit low level of earnings timeliness.
The next chapter sheds light on this issue and provides further details on the recent
achievement in this area.
Summary
This chapter reviews the core literature on the contemporaneous return-earnings association
with a particular emphasize on early theoretical and empirical assessments. The first part of
this review sheds light on the main factors that contributed to the emergence of market based
accounting research. Specifically, the EMH, the CAPM and the event study methodology
were discussed as a major academic development that provided the basis for implementing
empirical researches in accounting. We refer also to the pioneering studies of Ball and Brown
(1968) and Beaver (1968) as providing seminal and compelling evidence on the information
content and the usefulness of accounting earnings.
The second part of the review discusses previous earnings response coefficient literature that
was devoted to evaluate the magnitude of association between current corporate earnings and
returns. Particularly, we highlight the basic achievement in this area by focusing on early
papers that dealt with accounting earnings timeliness and its covariance with stock returns.
We also feature some studies on economic determinants that are likely to influence the
contemporaneous return-earnings relationship. We finally underline the weak
contemporaneous return-earnings relationship and provided a brief summary on the potential
and competing factors that might be responsible of such weakened association. Mainly, three
alternatives were investigated: the weak association could be attributable to market
inefficiency, to stock market reaction to value-relevant information that is not observed by the

CHAPTER II
company, or to noise in reported earnings. The last part in this chapter discusses earnings lack
of timeliness factor as a considerable explanation for the low returns-earnings relationship.
We examine Collins et al (1994) and other key papers that are the closest to this area of
research within which a further statistical refinement to the traditional returns-earnings
association was introduced.
As early accounting literature noticed a decline in the value relevance of
accounting earnings and some financial statement items (e.g. Lev and
Zarowin, 1999), other studies (e.g. Gelb and Zarowin, 2002; Lundholm and
Myers, 2002) suggest that an increase in corporate disclosure activity might fill the gap
by revealing information about future earnings. Voluntary disclosures are widely
regarded as important in supporting investment decisions.
Narrative information might be basically useful in enhancing the returnearnings association. The next chapter develops accordingly the theoretical framework
for corporate reporting practice and emphasizes the capital market implications for voluntary
disclosure. It discusses eventually the empirical literature on corporate
disclosure and the market's ability to anticipate future earnings growth.

CHAPTER II

CHAPTER III
Chapter III: Corporate voluntary disclosures and their economic
consequences: disclosure theories & empirical literature

Introduction
Corporate disclosure is defined as a communication process within which a
corporation shares its financial situation with investors and related parties
through the most common way of disclosure: the financial reporting and in
particular the annual reports. With the separation of ownership and
control, corporate disclosure became a critical element in the well
functioning of an efficient capital market. It is considered as the most
important source of information for investors and analysts. Demand for
corporate disclosure can arise from the information asymmetry and
agency problems between insiders and outsiders (Healy and Palepu,
2001). Enhanced corporate disclosure is believed to mitigate these
problems. Investors may better understand accounting information and
make better predictions for the future, which positively affects the market
value of the company and the liquidity of its shares (Francis et al., 2008).
Healy and Palepu (2001) paper discussed several significant forces in the
economic environment that may influence the nature of corporate
reporting. The rapid technological innovation, the changes in the business
economics of audit and financial analyst firms, and globalization are the
main factors with a potential to shape corporate financial information.
Accordingly, it is suggested that the traditional financial reporting model
wont be able to capture the economic implications of many of these
changes in a timely manner. Various studies, with different research
designs, reported a decline in the value relevance of accounting earnings
and other financial statement items (Lev and Zarowin, 1999; Amir and Lev,
1996; Collins et al., 1997; Francis and Schipper, 1999; Brown et al., 1999;
Ely and Waymire, 1999). As a response to the decrease in the usefulness
of financial information, managers undertook major improvements in their
reporting practices by increasing disclosure frequency and enhancing

CHAPTER III
information publicity. Corporate disclosure policy covered hence two
different aspects: a compulsory and a voluntary aspect. The distinction
between voluntary and required information relies on the obligation of
disclosure resulting from national law and GAAP (Depoers, 2000).
Voluntary disclosure can be defined as any disclosure outside the legal
requirements representing the deliberate choices of managers to reveal
financial and non-financial information that is useful for the investment
decision making process (Meek et al., 1995).
Over the past two decades, the issue of voluntary disclosure attracted much interest in the
accounting literature. What lies behind this interest is the desire to identify factors
underpinning corporate voluntary reporting and the expected benefits of such practice. Most
studies were related to the developed western countries, while little attention was given to
other developing countries. Those studies explored various aspects of voluntary disclosure
with a particular emphasize on the characteristics of issuing firms and the economic
consequences of such choices on revealing firms. One of these consequences is the increase in
the markets ability to anticipate future earnings changes.
The purpose of this chapter is to review the research on corporate voluntary disclosures in the
annual report. Section1 starts by presenting the main attributes of corporate voluntary
disclosure. Section 2 emphasizes the theoretical justification for corporate disclosure.
Specifically the review discusses the concept of information asymmetry (Akerlof, 1970), the
signalling theory (Ross, 1977) and the agency theory (Jensen and Meckling, 1976), that form
the basis of information economics. Section 3 highlights the economic consequences of
enhanced disclosure. Section 4 summarizes studies that examined the association between
voluntary disclosure and the share price anticipation of future earnings.
1. Attributes of corporate voluntary disclosure
The changes in the international business environment imposed reviews of the nature and the
content of corporate reporting. In fact, along with the capital market development and the
increase in the financial activities all over the world, the corporate communication process
evolved from the compulsory financial information through the voluntary disclosure of
myriad types of information. Voluntary reporting refers to the fact that owners delegated to
managers the leading function of their enterprises inducing information asymmetry between

CHAPTER III
these two parties. Typically, managers are more informed about corporate business and
activities than investors. This information asymmetry created distortion between corporation
and investors. Investors are mostly affected since they must make the right choice for their
own resources allocation (Bogdan et al., 2009). Accounting information was considered hence
as a crucial source of information for market participants.
It helps ensure effective decision making and company valuation process (Breton and Taffler,
1995). Disclosure helps also investors predict the future since its primary objective is to
inform market participants about the amount, timing, and uncertainty of future cash flows
(FASB, 2010).
Nowadays, the modern stakeholders needs for information are more sophisticated. Society
advances higher requirements of information publicity specifically for publicly traded
companies. This drives them to provide voluntarily more information in addition to financial
statements. Corporate reporting policy covers henceforward two different aspects: a
compulsory aspect as it transmits periodic and legal information (ex. Financial statements)
and a voluntary aspect which is related to the non-mandatory disclosure (Declerck and
Martinez, 2004).
1.1 Definitions of voluntary disclosure

Despite the widespread of early studies on corporate reporting practices, the concept of
voluntary disclosure lacks for an explicit and precise definition. Most prior work presumed
that the voluntary feature is either implicitly accepted or partly highlighted according to the
mandated national laws and GAAP.
According to Pourtier (2004), empirical research in this area may be split into two categories:
the first category of papers refers mainly to studies on earnings forecasts. Firms predictions
are voluntary par excellence which is likely to make superfluous a thorough definition (e.g.
Penman, 1980; Pownal and Waymire, 1989; Lev and Penman, 1990; Frankel, McNichols and
Wilson, 1995; Baginsky and Hassel, 1997). This interest can be linked to the flourishing
market based accounting studies that examined the relationship between management
earnings forecast, realized earnings and stock returns. Similarly, studies dealing with
corporate information on the internet focused mainly on disclosure tools. So far, little
attention is given to define the discretionary practice (Deller, Stubenrath and Weber, 1999;
Bushee, Matsumoto and Miller, 2003; Trueman, Wong and Zhang, 2003). Additional

CHAPTER III
theoretical works addressed information publicity in an auxiliary way. Voluntary information
is implicitly examined when focusing on corporate annual reports as a whole (Brown and
Kim, 1993; Botosan and Plumlee, 2002; Miller, 2002). Others were solely concerned with the
amount of information managed and disclosed voluntarily by corporate managers (Verrechia,
1990; King and Wallin, 1991).
The second category of research referred explicitly to the requirements of accounting
standards in order to decide with the mandatory or the voluntary nature of the disseminated
information (Depoers, 2000). These studies were interested in examining segment information
(Hayes and Lundholm, 1996; Harris, 1998 and Botosan and Harris, 2000) and in emphasizing
the voluntary nature of corporate forecasts and forward looking information with respect to
national GAAP and compulsory data (Penman, 1980; Dye, 1985 and Pownall and Waymire,
1989). Disclosure frequency was also addressed by Leftwich, Watts and Zimmerman (1981)
and Bradbury (1992) before mandating interim reporting in New Zeland. Dumontier and
Raffournier (1998) and Ashbaugh and Pincus (2001) addressed an alternative conceptual
framework by investigating the voluntary use of International standards as opposed to local
GAAPs.
Meek et al. (1995, p. 555) argued that voluntary information can be defined as disclosures in
excess of requirements representing free choices on the part of company management to
provide accounting and other information deemed relevant to decision needs of users of their
annual reports. In other words, the voluntary release of information represents the excess of
information, depending on the free choice of corporate management, the national regulations
and the pressures of the capital market participants (investors, financial analysts, consulting
firm). The firm leadership decides then which amount and type of information needs to be
disclosed and how relevant is the information to the person who will use it to make decisions.
The FASB (2001) report contended that voluntary disclosure is related to the process of
managing information and is presented mostly outside the financial statements. Pourtier
(2004) suggested that voluntary information is a set of multi-attribute which is likely to make
difficult its rigorous definition. He identified three dimensions, whereby corporate financial
information can be considered as voluntary. First, if the content of corporate disclosure is not
established, exceeding, or simply derogating from the requirement of accounting standard.
Second, the chronological frequency of information release may be considered as voluntary if
it is anticipated, delayed or more frequent. Accounting interim reporting follows this

CHAPTER III
direction. Providing more detailed information may increase disclosure frequency beyond
what is mandated. Finally, other disclosure tools such as internet, financial magazine or press
conference when added voluntarily to traditional means will disseminate further noncompulsory information and may improve firms transparency.
1.2. The nature of corporate voluntary disclosure

Typically, studies on discretionary disclosures were interested in apprehending corporate


reporting practice under a multidimensional aspect. They showed that information publicity
emerged from disseminating merely financial information to a wide range of disclosure
(qualitative, quantitative, financial or non-financial information) to meet the information need
of a wide range of users. A study by the American Institute of Certified Public Accountants
(AICPA, Jenkins Report) in 1994 focused on users of financial information and revealed that
they ask in addition to the traditional financial reporting for forward looking and nonfinancial information (Labelle, 2001). The Jenkins report highlighted new information needs
for the users of financial information. It proposed a series of recommendations such as
mandating management information in order to enhance business reporting. Nonfinancial
information conveyed voluntarily is intended, therefore, to satisfy the information needs of
investors as well as corporate stakeholders and partners. Unlike financial information that is
mostly expressed in monetary term, non-financial information could be either qualitative or
quantitative and may be expressed in different units (Teller, 2004). Voluntary information
aims to account about firms future prospect, strategic, social and environmental aspects of
corporate business etc.
Schuster and OConnell (2006) identified according to the FASB (2001) report five categories
of discretionary disclosure. The first class of information deals with business data,
which refers to the major performance measures and operating data that
management frequently uses for the purposes of control (e.g., sales and
financial performance). The second set of information considers the
managements analysis of business data that reports on key trends and
their effects on the business (e.g. Reasons for changes in the operating
and performance-related data). The third category is concerned with the
forward-looking information that reveals forecasts of key trends and the
opportunities and threats associated with these developments. Forwardlooking disclosure enables shareholders and other investors to assess a companys future

CHAPTER III
financial performance. Hussainey (2004) argued that forward-looking information
involves, but is not limited to, anticipated operating results, anticipated financial resources,
changes in revenues, cash flows and profitability. Forward-looking information embraces
also risks and uncertainties that could significantly affect actual results and cause them to
differ from projected results. The fourth class involves information about
management and shareholders. It introduces the board of directors and
management and includes information about their compensation.
The principal shareholders and creditors are additionally identified, and
data about the percentage ownership of major individual and institutional
shareholders is presented. The fifth order comprehends background about
the company that broadly describes the companys tangible and intangible
assets. Information about intangible assets provides details on areas such
as R&D, human resources, innovations, patents, brands, and trademarks.
Another clustering is presented by Meek et al. (1995). They classified voluntary disclosures
into three groups: strategic, nonfinancial and financial information. Strategic and financial
information are included, since previous studies showed that they are relevant to investors and
are widely discussed in the literature. Nonfinancial information is more directed towards a
firms social accountability and is aimed at a broader group of stakeholders than the
owners/investors. The strategic information refers first to a company's strategy and investment
decisions. The main purpose of this type of information is to provide users with
information about general corporate characteristics, corporate strategy, acquisitions and
disposals, research and development, and future prospects. Nonfinancial information refers
second to information about employee matters and social policy. It includes information about
directors, employee information, social responsibility and value added disclosures. Financial
information refers finally to quantitative issues which improve the understanding of the
factors that play a role in the performance and the future growth of the company. It
encompasses segment information, financial review information, foreign currency
information, and stock price information (Meek et al. 1995).
A recent growing accounting literature (e.g., Abraham and Cox, 2007; Beretta and Bozzolan,
2004; Dobler, 2008; Dobler et al., 2011; Lajili and Zghal, 2005; Linsley and Shrives, 2006)
focused on risk information as a relatively new scope of corporate disclosure and as a serious
topic for research. In fact, while risk communication has been recently regulated in some

CHAPTER III
developed countries such as the USA, Canada, Germany, UK, Austria and Finland etc..,
their current regulatory framework reveals a piecemeal approach, focused predominantly on
market risk associated with the use of derivatives (e.g., FAS 119, FAS 133, IAS 32, and IAS
39). In this regard, a wider set of risk reporting is hitherto not driven by rules and is offered on
a voluntary basis.

May be the first attempt to identify a framework for corporate risk disclosure was provided by
Beretta and Bozzolan (2004, p. 269) who defined it as the communication of information
concerning firms strategies, characteristics, operations, and other external factors that have
the potential to affect expected results. Based on professional bodies guidance on voluntary
risk reporting, Beretta and Bozzolan (2004) clustered risk information in three related groups.
It encompasses strategy (goals for performance, mission, broad objectives, and way to achieve
objectives); company characteristics, such as financial structure, corporate structure (changes
in ownership, mergers, and acquisitions), technological structure (core and support
technologies), organization (organizational structure and human resources management), and
business processes (concerning the way operations are managed); and environment around
the company (legal and regulatory, political, economic, financial, social, natural, and
industry). Despite analyzing the semantic properties of corporate risk reporting, Beretta and
Bozzolan (2004) definition did not cover all facets of risk. Particularly, they did not explicitly
address the fact that risk disclosure can be related to both natures of information that are
opportunity and danger. This gap was filled by Linsley and Shrives (2006) who offered a
broad definition of risk that includes both good and bad estimations of risks and
uncertainties. They argued that revealed information is considered as risk disclosure if the
reader is informed of any opportunity or prospect, or of any hazard, danger, harm, threat or
exposure that has already impacted upon the company or may impact upon the company in
the future or of the management of any such opportunity, prospect, hazard, harm, threat or
exposure Linsley and Shrives (2006, p. 389). Similar to Beretta and Bozzolan (2004),
Linsley and Shrives (2006) identified six categories of risk disclosure based on the risk model
developed by one of professional accountancy firms. These categories consist of financial
risks, operations risks, empowerment risks, information processing and technology risks,
integrity risks and strategic risks.

CHAPTER III
Dobler (2008) provided briefly some attributes of corporate risk disclosure based on two
notions of risk. He posits that risk information is considered as disclosure about the
randomness or the uncertainty of the distribution in corporate future outcomes which shall
satisfy an early-warning function for outsiders. Companies seek to reduce information
asymmetry between managers and outsiders regarding various external and internal risk
factors that comprise, for example, politics, regulation, and market, as well as finance,
business process, and personnel, any and all of them potentially affect an entity's future
performance (Dobler, 2008, p. 187).
Kravet and Muslu (2013) addressed corporate risk disclosure compared to other type of
corporate disclosure that is forward looking information. They explained that while forwardlooking disclosures are likely to resolve corporate uncertainties, risk disclosures serve rather
to explain these uncertainties and have the potential to decrease or increase users risk
perceptions. Finally, Miihkinen (2012) stated that risk disclosure is any information provided
in annual reports that outlines the firms' major risks and their expected economic impact on
future performance. It embraces forward-looking information that should help investors
predict future cash flows, information on the sources of uncertainty surrounding forecasts of
the firm's future cash flows, information on the sources of non-diversifiable risk that should
be included in the cost of capital and historical information as well as forward-looking
information about programs planned and actions taken to face risks (Miihkinen 2012, p.
442).
Overall, the voluntary feature of information is strictly recognized with regard to accounting
standards that are likely to shape managerial discretionary latitude. Corporate voluntary
practice outstrips then the traditional financial information by providing qualitative and
quantitative information over and above the mandatory requirements (Gibbins et al. 1992).
They are generally diffused via formal and informal channels such as press releases, websites,
conferences, meetings with the financial analysts and annual reports etc. (Healy and Palepu,
2001).
The next section highlights theoretical background and motivation for such practice.
2. Theoretical background
Leventis and Weetman (2004) argued that apprehending corporate disclosure policy is not an
easy task. There are many reasons why firms may provide information beyond what is

CHAPTER III
mandated by regulation. Some theories try to explain those reasons within a coherent
theoretical framework. However, no single theory can explain disclosure phenomena
completely.
2.1. The information or lemons problem

Akerlof (1970) argued that information or lemon problem arises from information
asymmetry and conicting incentives between buyers of a used car and the sellers which
could create an adverse selection problem and leads to a breakdown in the functioning of the
market.
If consumers cannot tell the quality of a product and are willing to pay only an average price
for it, then this price is more attractive for sellers who have bad products than to seller who
have good products (hence the term adverse selection). Information asymmetry may give rise
to opportunistic behavior such as misrepresentation of product quality, which could lead to
mistrust or even market failure. If consumers are rational and yet uninformed, they are mostly
unable to distinguish between the two types of products. In such situation, they would predict
this adverse selection and expect that at any given price, a randomly chosen product is more
likely to be a lemon than a good product. Being risk averse, buyers will show consequently a
lower willingness to pay for products and so the proportion of good products that is actually
offered falls further. Eventually, this process may imply a complete breakdown of the market
(Akerlof 1970).
In capital markets, the lemon problem scenario can be observed when a corporation goes
public. The sellers (managers) who initial public offering is more informed about the firm
product compared to potential buyers (investors). Therefore, if the information asymmetry
problem is not fully resolved, investors would be willing to pay only an average price for
firms equity. Given that decision, bad firms owners would be more attracted by equity price
than owners of good firms. This implies that firms market value that is actually issuing equity
should be below average, which urges investors to renounce buying offered securities (Jeny
and Jeanjean, 2007).
Beyer et al. (2010) suggested that information plays two particular roles in the functioning of
the market-based economies. As an ex-ante role, investors and market participants rely on
publicly available information to involve the decision making process and to assess the return
potential of investment opportunities. Second, thanks to the ex-post role, capital providers

CHAPTER III
will be allowed and based on the available amount of information, to decide which
monitoring mechanisms to implement in order to control for the use of their capital resources
once those have been allocated. Nevertheless, it is commonly known that managers, given
their proximity to firm activities and business environment, are more informed about future
prospects and expected returns regarding investment opportunities than outsiders. This
information asymmetry will potentially prevent investors and stock market participants at
large to adequately assess investment opportunities with high profitability or those with low
profitability. Consequently, financial capital providers will be more likely to overprice low
profitable securities, and under-price high profitable ones resulting eventually in market
failure (Beyer et al, 2010).
The economic and financial literatures identified several well-known solutions to the lemon
problem. Kreps (1990) stated that optimal contracts between entrepreneurs and investors
could mitigate the mis-evaluation problem by including incentives for full private information
disclosure. The information asymmetry problem could be also alleviated if regulations
mandate the full disclosure of private information by managers. A third potential solution for
lemon problem is the presence of information intermediaries such as rating agencies and
financial analysts whose mission is to engage in private information production to satisfy the
information needs of investors and to uncover managers superior information (Healy and
Palepu, 2001).
With the recent development in the market based economies, ownership
and control were typically separated in modern companies whereby a
clear distinction exists between decision-making rights and the capital
provided being at risk. Investors while providing funds are not implicated
in their resource allocation process and are unable to anticipate the
returns on their investments. Optimal contracting between the two parties,
including accounting information, is likely to protect investors funds from
expropriation by monitoring and aligning management interests with those
of investors. Agency and Signalling theories are often cited as the main
theoretical basis for research on managers-investors relationship and corporate
disclosure. The next subsections review these theories and their
applicability to the research question of this study.

CHAPTER III
2.2. The Agency theory

The agency theory flourished in the organizational studies since Jensen


and Meckling (1976) paper. It postulates that organizations should be seen
as a nexus of implicit and explicit contracts between the owners of
economic resources (the principals) and managers (the agents) who are
charged with using and controlling those resources. In fact, with the
separation between firms equity ownership and its management, modern
organizations have widely dispersed ownership, in the form of shareholders and debt holders,
who are not normally involved in companies management. An agent is appointed to manage
the daily operations of the company. Such agency relationship was defined
according to Jensen and Meckling (1976, p. 308) as a contract under
which one or more persons (the principals) engage another person (the
agent) to perform some service on their behalf which involves delegating
some decision making authority to the agent.
This distinction between ownership and control leads to potential conflicts of interests
between agents and principals and results in costs associated with resolving these conflicts
(Jensen and Meckling, 1976).
The main premise of agency theory is that principals and agents, as
rational contracting parties, will act in a way to maximize their wealth.
Managers, usually motivated by their personal gains, will act against the
interests of owners rather than considering the shareholders interests and
maximizing their value. If the goals of the principal and agent conflict, it
will be difficult and expensive for the principal to verify what the agent is
actually doing and if he behaved appropriately. Hence, because this pursuit of
self-interest increases firms costs, agency theory aims to find the most appropriate contract
managing the principal-agent relationship. Notably, the principals will establish a set of
incentives and governance mechanisms to monitor the agent behavior and to protect their
interests from being compromised or expropriated. Managers will likewise incur bonding
costs in order to signal to owners that they are acting responsibly and in a manner to co-align
their interests. Jensen and Meckling (1976) postulated that agency costs do not include only
monitoring and bonding costs, but also an opportunity cost considered as a residual loss:

CHAPTER III
-

Monitoring expenditures incurred by shareholders to curb agent opportunism include


costs associated with information management, controlling and incentives for
implementing governance mechanisms that are likely to improve the communication

process.
Bonding costs that are supported by the agent in order to signal to principal that he
behaves adequately to serve the owners interest. Watts (1988) suggested that
examples of bonding costs include expenditure on audit committees, non-executive

directors and internal auditors.


Residual loss that corresponds to the reduction in welfare experienced by the principal
as a result of the divergence between the contracting parties decisions which would
maximize the welfare of the agent.

Another type of agency problem which arises is adverse selection. This occurs when the
principal/owner(s) did not have access to all available information at the time decisions is
made by managers. They are unable to determine whether managers actions are in the best
interests of the firm. Accordingly, it is argued that the presence of information system is likely
to curb the agent opportunism.
Since information systems inform the principal about what the agent is actually doing, they
are likely to curb agent opportunism, because the agent will realize that he or she cannot
deceive the principal.
Jensen and Meckling (1976) modelled the financial communication strategy of a corporation
that initiate for the first time a public share offering. They demonstrated that the likelihood of
investors to buy new shares will be downward since they predict potential managerial
opportunism in the security issuing process. Such attitude will increase firms cost of capital.
To mitigate these problems, managers will be likely to extend their information disclosure and
enhance its credibility by incurring for example, costs of subjecting financial statements to
external audit scrutiny. Information level will serve as an implicit contract by which managers
reassure potential investors that they are acting in a diligent way. For instance, accounting
information is placed at the core of the agency relationship and plays a considerable role in
the context of managing optimal contracting (compensation agreements) between managers
and shareholders. This contract frequently requires entrepreneurs to disclose relevant
information that helps investors monitor compliance with contractual agreements and find out
whether firms resources were managed in the interest of external owners.

CHAPTER III
Watson et al. (2002) added that managers could reduce investor uncertainty and increase
shareholders confidence by disclosing more corporate information in annual reports. In other
words, agency cost is reduced through managers incentives to reduce information asymmetry
by disclosing voluntary information. Corporate voluntary disclosure is apprehended hence as
an important solution for the agency problem. It will be less costly for corporation to
communicate directly with investors rather than investors engaging in private information
produced by financial analyst to uncover any manager misuse of firm resources (Healy and
Palepu, 2001).
Under the agency theory framework, accounting literature raised considerable issues with
respect to managerial discretion over corporate financial reporting and voluntary disclosure in
particular. The main research on the voluntary disclosure decision focused on the information
role of reporting for capital market participants and stressed notably on management
incentives, the credibility and the determinants of voluntary disclosure (Healy and Palepu,
2001).
Six factors and benefits may restrain/encourage managers to disseminate
voluntary information:
-

Decreases the information asymmetry and hence the cost of


external financing in a stock transaction case (Lang and Lundholm,

1996; Botosan and Harris, 2000; Leuz and Verrechia, 2000).


Reduces the risk of change in the management team and explains

the poor performance of a corporation (Morck et al, 1990).


Improves shares liquidity and increases managers compensation in

stock option (Aboody and Kasznik, 2000; Nagar et al, 2003).


Reduces the risk and the cost of litigation with stockholders caused
by a delay in information disclosure and leading to a change in share

price (Skinner, 1994; 1997).


Reveals managers talent and quality to the market (Trueman,

1986).
Voluntary disclosures will be constrained if managers perceive that
disclosure could be competitively harmful (The proprietary cost
hypothesis) (Verrecchia, 2001)

Another stream of researches dealt with the perceived credibility of


voluntary reporting given that managers have incentives to make self-

CHAPTER III
serving voluntary disclosures which may alter their credibility. As the
resource allocation in the capital market depends on the degree of
credibility of corporate economic information, studies in this area focused
on the accuracy of management forecasts and their effects on stock
prices. Waymire (1984) and Ajinkya and Gift (1984) studies revealed that
stock prices positively and negatively react to management forecasts of
respectively earnings increases and decreases. Other evidences justified
the perceived credibility of management forecasts of new information by
market participants. Accuracy studies ascertained that these forecasts are
unbiased and more accurate than contemporaneous analysts forecasts
and that financial analysts tended to revise their expectations as a
reaction to management disclosure of such information. Piotroski (1999)
expanded these researches by examining firms that increase segment
reporting disclosures and noticed an increase in analysts forecast
accuracy and a decline in their dispersion suggesting that voluntary
disclosures other than management forecasts are also reliable. Amir and
Lev (1996) evidenced also the credibility of voluntary disclosures by
showing that information about the size of market population and market
penetration has a more significant association with stock prices than
required financial statement.
Other studies examined whether voluntary disclosure was intended to
control conflicts of interest between shareholders and managers and
figured out if the level of information publicity depends on some corporate
characteristics such as governance structure. Given that governance
mechanisms are meant to promote corporate transparency and
accountability, they are predicted to significantly impact the level of
voluntary disclosure. For example, Jensen (1993) argued that board
composition and board leadership structure are associated with the board
monitoring incentives and with the extent of voluntary information. Ho and
Wong (2001) found that the independence of the audit committee impacts
the level of voluntary disclosure in Hong Kong. As for ownership structure,
prior literature on agency theory (Jensen and Meckling, 1976; Vishny and
Shleifer, 1986; Jensen, 1993) suggested that the extent of executive/ non-

CHAPTER III
executive directors shareholding and outside block holders are associated
with higher monitoring incentives and corporate disclosure. Nasir and
Abdullah (2004) documented strong and consistent findings with regard to
ownership patterns (outside ownership, the extent of executive directors
shareholdings and the extent of government-linked enterprises
shareholdings) in the context of financially distressed firms.
Overall, agency theory is a robust framework within which a wide spread
of empirical accounting studies addressed the contractual relationship
between managers and investors and discussed managerial discretion
over voluntary disclosure.
2.3. The Signaling theory

The signalling theory is a further explanatory theory that addresses


problems arising from information asymmetry between two parties in any
social setting. It stipulates that information asymmetry will be decreased if
the person owning more information send signals to other interested
parties. Akerlof (1970) paper was the first study to illustrate that a classic
signalling model occurs in a general product market setting between the
seller and the buyer. First, he argues that product sellers are usually more
informed about the quality of their products and services than buyers
leading probably to market breakdown. Buyers will be unable to
distinguish between the qualities of different products and will be likely to
price them based on products average quality value. Consequently, the
seller possessing products with a quality above average incurs an
opportunity loss considering that his products would be evaluated at a
superior cost if the buyer knows about their higher quality. However the
seller with a quality below average will obtain an opportunity gain
(Eggleton et al, 2011).
Strong and Walker (1987) propose a potential position the seller will take
in such case: the seller of particularly high quality products will tend to
signal favorable information about his products and their quality in order
to justify the higher price. To be useful, the signal should not be easy to
imitate by the merchants of low quality products. Signalling will be an

CHAPTER III
iterative process which continues as long as the higher in price obtained
exceeds the signalling costs (Morris, 1987).
Accounting research relied also on signalling theory to discuss managers
incentives to disclose more information in the annual report. Eggleton et
al. (2011) argued that information asymmetry usually persists between
the managers of a firm and investors with respect to firms operations and
future prospects (e.g. The growth of a project, the expected returns on
investment, or risk profile). Due to this asymmetric information, it will be
more difficult for investors to assess firm value and to distinguish the
quality of various firms. In such situation, high quality firms will wish to
distinguish themselves from lower quality firms and tend to signal their
advantages to the market through voluntary disclosure. The signaling will
help then potential investors make decisions more favorable to the
company and may reduce their costs of raising capital (Whiting and Miller,
2008).
Spence (1973) argued that firm management attempt to alter the
investors beliefs about firms future cash flows. She claims that the
information flows are the output of a signalling game and that to be
effective they should obey to three main conditions: (1) Incentives are an
integral part of the signalling game; (2) the distinguishing characteristic of
the firm must be subject to firm manipulation and (3) the costs of the
signal are negatively correlated with the quality of the firm. According to
the first condition, expanded disclosure is likely to increase managers
reputation in the eyes of financial analysts as well as the investment
community. This is evidenced by an increase in analysts following and a
lower dispersion among analyst forecasts which implies a lower
information asymmetry and a reduced cost of capital. Another direct
benefit to managers would include compensation contracts (stock option)
which relate to corporate share performance. The direct benefit provides
supporting evidence that managers, who make the disclosure decision,
have incentives to influence the subjective beliefs of the market. The
second condition outlined by Spence (1973) is that the distinguishing

CHAPTER III
quality is subject to managers manipulation. In this context, discretionary
accounting disclosures are important to market participants. They are the
subject of deliberate choices of managers with the purpose to influence
the beliefs of the investment community. The last condition of Spences
signalling game is that the cost of the signal is negatively related to
corporate quality. High quality firms are represented by those firms which
are undervalued by the market due to information asymmetry. These firms
would signal their superior quality, relative to the markets belief, by
providing more informative disclosure. In contrast, low quality firms are
either accurately valued or overvalued by the market. They have no
incentive to expand disclosures since the market would recognize its
pricing error and value the firm downward. In this situation the costs of
increased disclosure exceed the benefits and hence negatively correlated
with the firms quality. Such condition will prevent low quality firms from
mimicking high quality firms.
Watson et al. (2002) ascertained that for firms, the success of signalling
quality depends mainly on signal credibility. In other words, if managers
falsely try to signal that they are of high quality, when in fact they are of
low, no subsequent disclosures will be seen as credible once this was
revealed. Credibility will be therefore achieved as ultimately the true
quality of the firm is verifiable. Eccles et al. (2001) argued that enhanced
disclosure is likely to increase the markets perception of manager's
credibility. A management team that has confidence in both their own
abilities and their strategy will not shy away from telling the market, their
plans for the future and how well it is doing today.
A large number of papers investigated how corporate voluntary disclosure
could be apprehended as a signal to the financial market. Prior studies,
like Grossman (1981) and Milgrom (1981) for example, postulated that
corporate managers are motivated to behave in the best interest of firms
current owners. They are not likely to disclose all value-relevant
information, regardless of whether this information is good or bad.
Managers decision to disclose or not, value-relevant information depends

CHAPTER III
on whether they have learned the value of the signal and after bearing in
mind the effect of the disclosure on the wealth of current shareholders
(Walker, 1997). Grossman (1981) and Milgrom (1981) predicted that
managers truthfully disclose all value-relevant information except in a
situation where the manager receives the worst possible signal. Consistent
with the Signalling theory, Hughes (1986) assumed that information
disclosure practice as well as ex post information credibility can be seen as
a signal of corporate value in the presence of high information asymmetry
between outsiders (investors) and insiders (managers) regarding share
price. Verrecchia (1983) posited also that in order to receive a higher
valuation from investors and shareholders, corporate management is
induced to disseminate good news forecasts.
Analogically, Trueman (1986) put forward that managers may benefit from
a more positive talent assessment by investors if they release good news
that are likely to be translated into a higher market value. Managers are
thus motivated to disclose more detailed information to support the
continuance of their position and remuneration (Singhi and Desai, 1971)
and to signal institutional confidence (Ross, 1979).
Other studies suggested that firms with good news are likely to publish
their earnings forecasts in order to differentiate themselves from firms
with poor performances, particularly when an initial public offering (IPO) is
envisaged. For example, Lev and Penman (1990) exhibited that, on
average, voluntary disclosure forecasts are frequently observed for firms
with good news compared to those with bad news since the former
seek to differentiate themselves from rivals in the financial market. The
argument behind is that management earnings forecasts are likely to
influence firms value in two ways. First, the issue of earnings forecasts
might result in a revision of expected future earnings. Second,
management forecasts can increase the proportion of investors holding
corporate stocks, thereby decreasing the required rates of return and the
information asymmetry between interest-related parties in the financial
market. Ruland et al. (1990) and Mak (1996) papers provided evidence

CHAPTER III
supporting these contentions. They showed that firms increasing their
capital with low retained ownership by pre-IPO shareholders are more
inclined to disclose management forecasts. Such publicity can attenuate
adverse selection and moral hazard problems, making it easier for firms to
attract new investment. Corporate managers have therefore an incentive
to issue earnings forecasts by which they signal their firms value and
distinguish their IPO from those of lesser quality firm (Jeny and Jeanjean,
2007).
Signalling theory was applied likewise to various accounting studies
dealing with the association between corporate disclosure and some firms
characteristics. According to Foster (1986), profitable and well-managed
corporations have incentives to differentiate themselves from less
profitable firms. By providing discretionary disclosures, they seek to raise
capital on the best available terms. Thus, highly profitable companies can
be expected to reveal more voluntary information. Haniffa and Cooke
(2002) reported a positive and significant correlation between the firms
profitability and the level of voluntary disclosure. This result provided
consistent evidence supporting information signalling hypothesis which
claims that firms with good news are more likely to disclose additional
information (Ross, 1979).
Nassir and Abdallah (2004) argued that, under the principal-agent
relationship, voluntary disclosure is one of the means for the principals to
monitor their economic interests, and for the agents to signal that they are
acting in the best interest of the owners. Hossain et al. (1994) and Chau
and Gray (2002) provided support for this contention. They revealed an
association between the ownership structure and the extent of voluntary
disclosure of listed firms in Malaysia, Hong Kong and Singapore,
respectively.
From the above mentioned discussion of agency and signalling theories it
can be seen that there is considerable overlap between the two theoretical
justifications of corporate practices and particularly of voluntary

CHAPTER III
disclosure. Indeed, Morris (1987) explored whether these two theories are
consistent, equivalent or competing, by examining the necessary and
sufficient conditions for them both. Morris (1987) suggested that as the
sufficient conditions for signalling theory are consistent with those of
agency theory (rational behavior is common to both), the two theories are
consistent. However, a necessary condition for signalling theory,
informational asymmetry, is not shared by agency theory and therefore
they are not equivalent, i.e. one is not implied by the other. Morris (1987)
highlighted that given this consistency between agency and signalling
theory, it is possible to combine them to yield predictions about
accounting choices. He concluded (p. 52 the prediction of accounting
choices can at least be improved by adding together the predictions from
each theory. It seems, therefore, that greater insight can be gained into
why managers voluntarily disclose credible information by drawing from
both theories (Watson et al, 2002).
3. Capital market implications for corporate voluntary disclosure
The incentives and the economic consequences for disclosing high quality information were a
long-standing research question. Along with the stem of corporate discretionary reporting, a
growing body of literature was interested in investigating the capital market effects of
voluntary information disclosure. These studies asserted that there are potentially three types
of economic consequences of improved information publicity: reducing the cost of capital,
improving the liquidity of the firms stock in the capital market and increasing financial
analysts following.
This section reviews the available evidence on the above-mentioned impacts for firms that
make extensive voluntary disclosures.
3.1 Voluntary disclosure and the cost of capital

The link between corporate voluntary disclosure and the cost of capital gained considerable
attention among academics and regulators. Cheynel (2009) defined the cost of capital as the
minimum return demanded by investors to invest in a new project and contends that it is
often used to measure the impacts of disclosures. Botozan (1997) argued that whether firms
benefit from greater disclosure while minimizing the cost of capital is an important and

CHAPTER III
controversial question for managers, capital market participants, and standard setters. The
report of the American Institute of Certified Public Accountants' Special Committee on
Financial Reporting (AICPA Report) (1994) and the Financial Accounting Standards Board
(FASB) member, Foster (2003) suggested that high disclosing firms profit from a lower cost
of equity capital is intuitive. The Financial Executives Institute (FEI) asserted in contrast that
enhanced disclosure will increase share price volatility and consequently lead to an increase
in the cost of equity capital (Botozan 2006).
In light of this on-going debate, theoretical studies provided insights into the relationship
between disclosure and the cost of capital. Academic research supporting a negative
connection between disclosure level and the cost of equity capital were based on two
directions: a liquidity-based approach and an estimation risk perspective. The first suggested
that firms try to overwhelm the reluctance of shareholders and potential investors to hold
shares in illiquid markets by revealing private information and thereby reducing their cost of
capital (Hail, 2002). If information asymmetry between a firm and its shareholders or among
potential buyers and sellers of a firms shares is high, then demands for corporate securities
will decrease and transaction costs will be higher. This, in turn, would induce market
illiquidity and render a higher cost of equity capital. Earlier studies, like Amihud and
Mendelson (1986, 1988) linked higher transaction costs to market illiquidity. They contended
further that securities with large bid-ask spreads would encounter a greater cost of equity
capital and recommend managers to disclose private information as a way to reduce it. King
et al. (1990) and Diamond and Verrecchia (1991) confirmed such recommendation. They
stated that firms willingness to disclose private information will reduce the amount of
information revealed by a trade and hence the adverse prices impact of large trades. This
triggers potential buyers in the capital market to gather larger stock holdings than they
otherwise would, which increases demand and stock price, and reduces the cost of equity
capital (Botozan, 2006).
The second approach suggested that increased disclosure decreases the risk associated with
investors assessments of the parameters of an assets future return or with the payoff
distribution and, thereby, reduces the cost of capital (e.g., Coles et al., 1995 and Clarkson et
al., 1996). As investors assess securities parameter based on revealed information, their
confidence level depends on the characteristics of their information set. For instance, when
disclosure is imperfect, investors bear risks in forecasting the future payoffs from their
investment.

CHAPTER III
According to Botozan (2006) this line of studies reaches two main conclusions with regard
corporate disclosure research. First, if estimation risk is non-diversifiable, investors will
demand an incremental return for bearing such risk for poor disclosure securities. This will in
turn engender a higher cost of equity capital. Second, estimation risk is ignored in the
traditional analysis of optimal portfolio choice and equilibrium pricing since they treat the
estimated parameters as if they are true. As a result, estimation risk is not captured by market
beta. Botozan (2006) illustrated these basic ideas with the following example: investors are
confronted to two distinct situations in the capital market. Firm A and firm B, have the same
expected profit, but diverge in terms of the amount of information revealed to investors.
Shareholders and potential buyers are assured of their anticipation regarding firm A's future
payoff because ample information is available about the firm. In contrast, investors are quite
uncertain about their predictions for firm B since little information is available about this
firm. Botozan (2006) argued that the capital asset pricing model (CAPM) treats the expected
returns for both firms as if 'true' and ignores investors' differential information publicity and
uncertainty with regard to their anticipations. Investors' uncertainty has then no role in
determining their optimal portfolio choice or the equilibrium pricing of securities.
Inconsistent with such view, the above mentioned literature explicitly integrates investors'
uncertainty into the model and concludes that in equilibrium, securities with greater (lower)
estimation risk parameter have lower (greater) stock prices, all else equal. It is worth noting,
that no consensus has been reached on the diversification of risk estimation (Clarkson et al.
1996).
In brief, positive effects of enhanced financial reporting on the cost of equity capital are
maintained by accounting theory in two respects: (1) investors preferences for holding shares
with eventually low transaction costs while performing future trades and (2) investors
preferences for securities with relatively smaller uncertainty about future returns or payoffs
(Hail, 2002).
Various empirical researches sustained this theory by investigating the relation between the
voluntary release of information and a firms cost of capital. For example, Welker (1995) and
Sengupta (1998) examined corporate disclosure rankings by nancial analysts. They
documented that firms rated as more transparent have, on average, lower bid-ask spreads and
lower cost of debt at the time of issue, respectively. Botosan (1997) found a negative relation
between her self-constructed disclosure index and firms cost of equity capital, but only for
firms with a low analyst following. Healy et al. (1999) investigated the effect of a maintained

CHAPTER III
increase in corporate disclosure on a number of variables considered to be associated with the
cost of equity capital. They showed that firms enhancing their disclosure levels experience
improvements in the bid-ask spread. Piotroski (1999) ascertained that a contemporaneous
increase in the market capitalization of corporate earnings is associated with companies
providing additional segment disclosures. This finding is consistent with rms having a lower
cost of capital. Rashid (2000) focused on the association between disclosure and the cost of
debt for retail banks. He showed that firms with higher disclosure scores experience a lower
cost of debt capital. Zhang (2001) modeled a situation where a company selects its disclosure
level, and showed that the relation between cost of capital and voluntary disclosure may be
positive or negative depending on what causes variation in disclosure levels. When the
variation is primarily driven by factors like earnings variability, the variability of liquidity
shocks or the cost of informations analysis, disclosure levels are positively related to the cost
of capital. However, when variation in disclosure is driven by disclosure costs, the cost of
capital is negatively associated with disclosure levels.
Botosan and Plumlee (2002) re-examined the relation between corporate disclosure level and
the cost of equity capital and found some mixed results. Unlike what is expected, they
revealed that higher corporate total disclosure is associated with a greater cost of equity
capital. The authors investigated additionally whether the association between disclosure and
the cost of capital differs by type of disclosure. They noticed a negative relation between
analyst rankings of annual report disclosures and the cost of capital. They showed also that
rms cost of capital is positively related to rankings of quarterly disclosures, and
uncorrelated with investor relations activities. Hail (2002) explored in a similar way the
relation between disclosure quality and the cost of equity capital. The results showed that
companies on the Swiss market have a strong negative association between disclosure and the
cost of capital. Unlike Botosan (1997), Hail (2002) findings are not confined to low analyst
following only.
Francis et al. (2004, 2005) and Ecker et al. (2006) measured accounting quality using residual
accruals volatility and found similar results. Chen et al. (2003) provided evidence that more
rm- specic information in stock returns is related to a lower cost of equity. Finally, a more
recent strand of literature connects more precise information (taken as exogenous) to the cost
of capital. Francis et al. (2008) found that the more the voluntary disclosure is the less the cost
of capital will be. Lambert et al. (2007, 2012) suggested that firms with more precise
information about future cash flows have lower conditional covariances with the market, and

CHAPTER III
as a consequence, lower conditional betas and lower expected returns for all rms in a pure
exchange economy. These recent analytical papers did not provide implications for crosssectional results. In general this literature did not claim that any kind of signal by one rm
may or may not reduce ex-post cost of capital for the same rm compared to other rms. It
did not examine whether a rm in a given economy with any kind of information as opposed
to an economy without information will have unambiguously a lower cost of capital and a
higher market price. Their implications are not at the rm-specic level, but rather at the
aggregate level.
Overall, the above studies showed an inverse relation between disclosure and the cost of
capital. The current study examines the benefits of corporate disclosure from a different angle.
It tests whether increased level of voluntary disclosure in corporate annual reports allows a
better informed stock market. The concept of a better informed stock market refers to the
markets ability to anticipate future earnings changes more accurately.
3.2 Voluntary disclosure and stock market liquidity

Different theoretical models (Diamond and Verrecchia, 1991; Kim and Verrecchia, 1994;
Bushman and Smith, 2001) addressed how the level of corporate disclosure affects
information asymmetry among informed and uninformed investors who trade in the firm's
shares which, under other circumstances, may lead to market inefficiencies and to the mispricing of companies securities. They argued that voluntary information plays a key role in
the trading process, impacts the bid-ask spread and increases stock market liquidity. Firms
might follow a disclosure strategy as a reply to perceived illiquidity of their shares in the
market (Lakhal, 2008). By mitigating information asymmetry, investors will be relatively
assured that any stock market transaction occurs at a fair price, increasing trading activity
and hence liquidity in the firms stock (Healy and Palepu, 2001).
Several empirical papers put forward a relationship between corporate disclosure level and
stock market liquidity through reducing the information asymmetry among traders in
the financial markets. They noticed that the measurement of market liquidity is complex and
even subject to measurement problems. Leuz and Verrecchia (2000) and Welker (1995) for
example, suggested that market liquidity could be measured by both trade-based and orderbased measures i.e. transaction volumes and bid-ask spreads. Amihud and Mendelson (1986)
analyzed the effects of liquidity on asset pricing, and used for such the spread between
bidding and asking prices. They suggested that companies could decrease the adverse

CHAPTER III
selection of the bid-ask spread by revealing private information and thereby increasing their
stock liquidity. They showed that a greater cost of equity capital is associated with wider bidask spreads because investors demand a higher return to compensate for added transaction
costs. Welker (1995) focused on a sample of U.S companies over eight year period between
1983 to 1990 when examining the cross sectional association between disclosure score and
market liquidity. He relied on analyst ratings of corporate disclosures as provided by the
Association for Investment Management and Research (AIMR) and measured market
liquidity by the size of proportional bid-ask spread. Main findings revealed that an enhanced
disclosure policy reduces the information asymmetry and increases the liquidity in the stock
markets. He showed that the relative bid-ask spreads for firms with a disclosure ranking
in the bottom 33% were approximately about 50 percent higher than spreads for firms
with a disclosure ranking in the top third of the empirical distribution. Healy et al. (1999)
investigated also whether companies benefit from enhanced voluntary disclosure by
examining changes in capital market factors in relation to the increase in the AIMR disclosure
rankings of 97 firms. They found that the increases in disclosure level are accompanied by
increases in firms stock returns, institutional ownership, analyst following and stock liquidity.
Their findings were consistent with the disclosure model hypotheses that increased disclosure
drives market participants to reconsider the firms intrinsic value, improves shares liquidity,
attracts additional institutional owners and increases analyst interest in corporate shares. Leuz
and Verrecchia (2000) were interested in studying a sample of German firms that changed
from national GAAP to international accounting regime (IAS or US-GAAP) with a greater
disclosure requirement in consolidated financial statements. They argued that these
switching firms were committing themselves to increased levels of disclosure. Leuz and
Verrecchia (2000) demonstrated that firms revealing more information when adopting
international reporting standards were associated with lower bid-ask spreads and higher
trading volume than the ones keeping to the German reporting regime.
Coller and Yohn (1997) used a sample of 278 quarterly earnings forecasts to investigate
whether the decision to issue a management earnings forecast reduces the information
asymmetry in the market. They claimed that specialists widen the bid-ask spread when they
perceive greater information asymmetry. Results revealed that forecasting firms have wider
bid-ask spreads than a matched sample of non-forecasting firms prior to the forecast
release. Nevertheless, this difference disappears after the release of the management forecast.
Coller and Yohn (1997) showed also that forecasting firms experience a gradual increase

CHAPTER III
in spreads over the twelve months leading up to the forecast while the spread is reduced to
below the pre-forecast level after the forecast is released. Such a result is explained by the fact
that management forecasts are unanticipated by investors. Espinosa et al. (2005) examined
whether the quality of annual report disclosures is significantly associated with a crucial
dimension of the functioning of the stock market: liquidity. Using a sample of firms quoted on
the Spanish stock exchange between the period of 1994 and 2000, they showed that the
direction of the relationship depends crucially on the used liquidity measures. They claimed
that more transparency reduces the relative spread and the Amihud (2002) measure of market
illiquidity, but it also reduces the depth and the market quality index proposed by Bollen and
Whaley (1998). Heflin et al. (2002) suggested that information quality is important for market
liquidity. It serves a mean for reducing information asymmetries across equity traders and
increasing their ability to effectively execute stock trades when needed and at reasonable
costs. The authors examined the AIMR disclosure quality ranking of 221 american firms from
1989 to 1998 and used two measures of liquidity; bid-ask spread and depth. They found that
a firm with high quality of accounting disclosure enhances its market liquidity through
reducing information asymmetries across traders (increasing quoted depth and decreasing
effective spreads).
Most of these empirical papers used event study methodology to investigate the effect of
firm's publicly available information around the day of information releases on stock
market liquidity (e.g. Healy et al., 1999; Coller and Yohn, 1997; Marquardt and Wiedman,
1998). This may indicate that revealed information has an impact on stock liquidity on a
given or around the day of the particular event. Also, most of the empirical evidence on
the impact of enhanced disclosure on stock market liquidity relied on financial analysts'
scores of firm disclosure (e.g. AIMR). These scores are based on sell-side analysts'
perceptions of overall disclosure policy, including both voluntary and mandatory disclosure.
Therefore, there is a potential bias related to analysts' perceptions and the inclusion of
mandatory disclosure (Lang and Lundholm, 1993; Hossain et al., 2006).
In summary, these theoretical and empirical research suggested that increasing the
disclosure level may reduce information asymmetries in the stock market and increase
thus the stock market liquidity.

CHAPTER III
3.3 Voluntary disclosure and analyst behavior

Another economic consequence of corporate disclosure which received


much focus in early literature refers to increased information
intermediaries. Financial analysts are important actors in the capital
market. Their main role is to provide earnings forecasts, buy and sell
recommendations and other information to brokers and to institutional
investors. For the purpose of the evaluation process, financial analysts rely
mainly on information revealed directly by firms (Lang and Lundholm,
1996). Evidence on how financial analysts use firms revealed information
persists to be of interest to those concerned with apprehending analysts
behavior as information processors and the effect of corporate disclosure
practices.
Nichols (1989) and Schipper (1991) suggested that the behavior of
analysts may provide indirect insights into the unobservable beliefs of
investors. Bhushan (1989) is a landmark paper that yielded a significant
understanding of factors influencing analyst following. He considered that
analyst coverage could proxy for the equilibrium total expenditure on
analyst supplied services in the economy for that firm. Bhushan (1989)
suggested that the number of analysts following is assessed competitively
and that the equilibrium total expenditure by investors on analysts
services depends on various attributes. The latters are likely to impact
either the request for, or the cost of providing analysts services. Five key
factors are examined: ownership structure, firm size, stock return
volatility, task complexity and the strength of correlation between firm
return and market return. Bhushan (1989) modeled and interpreted the
relationship between corporate voluntary disclosure level and the extent
of financial analysts following in two ways: the first one is a function of
demand for analyst services while the second is a function of analyst
services supply to investors. Corporate information publicity is likely to
decrease costs associated with analysts information acquisition. In this
case, analysts will not be engaged in acquiring information from other
sources. For instance, when companies reveal additional information,

CHAPTER III
financial analysts service supply will increase and the number of analysts
following will increase at the equilibrium.
The effect of corporate information disclosure on analyst services demand
depends mainly on the financial analyst role in the capital market. As
financial intermediaries, analysts are likely to produce more relevant
investment recommendations for their client when more corporate private
information is available. Accordingly, the more voluntary information is
revealed, the more will be the demand of analysts services and the more
will be the number of analysts following at the equilibrium for those firms.
As information providers to investors, financial analysts are in the business
of generating transactions business for their companies. Investors are
likely to acquire information for free when firms decide to reveal it
voluntarily. Corporate disclosure could be seen, therefore as a substitute to
analysts reports. So far, demand for analyst services will decrease and
likewise the number of analysts following at the equilibrium. In sum,
voluntary information disclosure is likely to increase analyst service supply.
Such practice may induce an increase or a decrease in analysts services
demand depending on their role in the financial market.
Despite these theoretical ambiguities with respect the net effect of
corporate disclosure on analysts following, early empirical researches
showed a positive association between voluntary information and financial
analyst behavior. Lang and Lundholm (1996) was the first study that
focused on the direct relationship between corporate disclosure strategies,
the number of analysts following each company and the properties of
analysts earnings forecasts. They started first by examining the analysts
selection of which firms to follow as measured by the number of analysts
supplying earnings forecasts for a firm. Lang and Lundholm (1996) argued
that the extent of corporate disclosure is expected to influence the
provision and the demand of analysts services. If analysts act mainly as
professional information intermediaries who process firms provided
information for ordinary investors, then enhanced information disclosure
will ensure that analysts will provide more valuable reports to sell. In this

CHAPTER III
case, increased information publicity leads to an increase in the demand
for analysts services. Whereas, if analysts are acting mainly as primary
information providers competing with the provided disclosures made
directly to investors, then an increase in the level of disclosure will
substitute for the analysts reports. Increased disclosure leads then to a
decline in the demand for analysts services. The second issue of analyst
behavior in Lang and Lundholm (1996) was related to the characteristics
of analysts earnings forecasts as proxied by forecast accuracy, the degree
of dispersion among individual analyst forecasts and the variability of
forecast revisions during the year.
Lang and Lundholm (1996) suggested that the impact of increased
disclosure on the dispersion of analyst forecasts is a function of whether
differences in forecasts are due to differences in information or differences
in forecasting models. They argued that if analysts have the same
forecasting model and examine the same firm-provided disclosures, but
process distinct private information, they will rely less on their private
information. Considering that the informativeness of corporate disclosure
increases, the consensus among their forecasts will increase. If analysts
have the same firm-provided and private information but differ in the
weights they place on components of firm-provided disclosure in
forecasting earnings, then additional disclosure might increase the
dispersion of analyst forecasts (Walker and Tsalta, 2001). Unlike the
unclear direction of the effect of corporate disclosure on analysts
forecasts dispersion, the likely relationship between corporate disclosure
and analyst forecast accuracy is easier to predict. Lang and Lundholm
(1996) stated that if firms provided information is informative about future
earnings, then analysts forecast accuracy would increase. They even
affirmed that it is difficult to imagine scenarios in which additional
disclosure will systematically reduce the accuracy of analysts earnings
forecasts. Finally, they tested if the volatility of forecast revisions in the
period up to an earnings announcement is likely to be reduced by a more
forthcoming disclosure policy. Lang and Lundholm (1996) main results
indicated that corporate disclosure policy is significantly and positively

CHAPTER III
related to analysts following. They also showed that increases in disclosure
quality are ensued by higher levels of analyst following. They did not find
evidence that increased analyst following leads to an increase in corporate
disclosure. With respect to analyst earnings forecast dispersion, their
findings revealed that firms with more forthcoming disclosures have more
accurate consensus forecasts, less dispersion among individual analyst
forecasts and fewer extreme forecast revisions.
Following Lang and Lundholm (1996), a number of studies investigated the
association between disclosure quality and analyst behavior. For example,
Byard and Shaw (2003) examined in the United State context, how the
quality of corporate disclosures influences the accuracy of annual
earnings forecasts by financial analysts. They found that corporate
disclosure quality is positively associated with the precision of such
forecasts. Walker and Tsalta (2001) tested the extent to which the
publication of forward looking information in the annual reports of UK
companies is related to analysts following. They reported a strong and
positive relationship between the quality of forward-looking information in
the UK annual report discussion section and analysts following.
Walker and Tsalta (2001) also showed a positive causal impact of analysts
following on the page length of the annual report. Eng and Teo (2000)
investigated the effect of annual report disclosures on analyst forecasts for
a sample of firms listed on the stock exchange of Singapore. They
analyzed the association between the extent of corporate disclosure and
three characteristics of analyst behavior, namely the accuracy of analysts'
earnings forecasts, dispersion in analysts' earnings forecasts, and the size
of analysts following. Eng and Teo (2000) found that the level of annual
report disclosures is positively related to the accuracy of analysts
earnings forecasts and to analysts following. They also documented that
the level of corporate disclosure is negatively related to dispersion in
analysts' earnings forecasts. So far they showed that more corporate
disclosures by Singaporean firms lead to more accurate and less dispersed
earnings forecasts among analysts. Greater corporate disclosure can also
lead to greater analyst interest in the firm.

CHAPTER III
Based on an international sample, Hope (2003) studied whether the extent
of firms disclosure of their accounting policies in the annual reports is
associated with properties of analysts earnings forecasts. Results revealed
that the level of accounting policy disclosure is significantly and negatively
correlated with analysts forecast dispersion and positively related to
analysts forecast accuracy. Hope (2003) found in particular and after
controlling for firm- and country-level variables that accounting method
disclosures are incrementally useful to analysts over and above all other
annual report disclosures which suggests that accounting policy
disclosures reduce uncertainty about forecasted earnings. Lakhal (2009)
investigated the relationships between voluntary earning disclosures
made by 154 French-listed firms and financial analysts behavior.
Results indicated that the disclosure decision influences financial analysts
coverage. They suggested that analysts choose to follow firms with high
voluntary disclosure practices. Additional findings showed that voluntary
earnings disclosures improve analysts forecast accuracy and reduce
the dispersion among the financial analysts forecasts suggesting that
these disclosures reduce market uncertainty about forecasted earnings.
Vanstraelen et al. (2003) focused on the relationship of Jenkins Committee
non-financial disclosure levels with the accuracy and the dispersion of
financial analysts earnings forecasts across three continental European
countries (Belgium, Germany and the Netherlands). They revealed that
higher levels of forward looking non-financial disclosures are associated
with lower dispersion and higher accuracy in financial analysts earnings
forecasts.
In summary, empirical evidence indicated that enhanced disclosure is
associated with more accurate analysts forecasts and more analysts
coverage. Despite these clear cut results, the current study seeks to test
the relationship between prices and future earnings directly rather than
relying on proxies such as analyst forecasts. This is because more accurate
analyst forecasts might be evidence of firms managing their analyst
relationships better rather than evidence of more informative prices (Gelb
and Zarowin, 2002).

CHAPTER III
The following section reviews the literature related to the effect of
voluntary information on share price anticipation of future earnings.
3.4. Voluntary disclosure and the return future earnings relationship

The association between discretionary disclosure and share price


anticipation of future earnings is a topic of considerable interest among
accounting researchers. Schleicher and Walker (1999) were among the
pioneer studies that investigated the effect of voluntary information in
published annual reports on the return-future earnings relationship.
Particularly, they tested whether the quality of managements discussion
of operational and financial aspects influences the informativeness of
stock prices. They constructed three different disclosure indices based on
several sources, including recommendations provided in the ASB
statement operating and financial review, The corporate report (ASSC,
1975), making corporate reports valuable (ICAS, 1988) and previous
disclosure studies. The final outcome combines 82 items in the operating
and financial review (DOFR) index, 64 items in the operating and financial
projections (DOPF) index, and 34 items in the segmental reporting (DSEG)
index. The authors argued that constructing a separate forward looking
index is motivated by the fact that they expect such information to be
more useful in predicting future earnings changes. Schleicher and Walker
(1999) exploited the theoretical framework proposed by Kothari (1992)
and Donnelly and Walker (1995) which assumes that the capital market
can partially anticipate permanent shocks. Their work was restricted to a
small sample of firms. From the 94 primarily involved companies, they
selected only 20 non-financial UK companies (220 company-years) from
three industry sectors namely engineering, electronic and electrical
equipment. The main findings indicated that higher voluntary disclosure
enables capital market participants to better forecast future earnings
changes. Share price informativeness with respect to future earnings
seemed to be stronger (up to one third) when Schleicher and Walker
(1999) employ the forward-looking disclosure index, the DOPF and in

CHAPTER III
models that examine one-period- ahead and two-period-ahead of share
price anticipation.
Despite these pioneering evidences on the association between forward
looking information and share price anticipation of future earnings,
Schleicher and Walker (1999) study suffered from some caveats. First the
sample size was relatively small. Second, they concentrated on only three
related industry-sectors. For instance, it was difficult to generalize these
findings on other sectors.
Lundholm and Myers (2002) studied the effect of corporate disclosure
strategy on the association between current stock returns and the
contemporaneous and future earnings. They argued that stock return over
one year is not only a function of the unexpected portion of the current
year realized earnings, but also a function of changes in expectation about
future earnings. The authors postulated that corporate disclosure has the
potential to shape expectations about a companys future performance.
They stated that if firms tend to reveal useful news for predicting future
earnings, then such news will be reflected in current stock returns. In this
case, the coefficient on the future earnings variable will be positive in the
returns-earnings regression model. In contrast, if firms do not publish
information about future earnings, then such information wont be
revealed to the market. So far, the coefficient on the future earnings proxy
will be closer to zero. The authors suggested an interactive relationship
between future earnings and the level of a firms disclosure activity.
Lundholm and Myers (2002) examined mainly whether high levels of
disclosure activity bring the future forward by revealing relevant news
about future earnings while low levels of disclosure do not (p. 814). They
also investigated whether the quality of corporate disclosure is likely to
influence the importance of current earnings news. They expected that
current earnings news would become less relevant when corporate
disclosure cause returns to depend more heavily on the future earnings
news. However, current earnings are likely to be value-relevant for low
disclosure firms since current earnings would proxy for changes in

CHAPTER III
expectations about future earnings. This implies that corporate
information disclosure may cause a substitution away from current
earnings and toward future earnings (p. 816). Lundholm and Myers
(2002) third hypothesis analyzed the degree to which variation in
corporate disclosure is associated with changes in earnings
informativeness. They suggested that if the firms disclosure level affects
the extent to which future earnings are incorporated in current returns,
then the extent of change in current returns that is related to information
about future earnings should also be increasing in the level of disclosure.
To proxy for corporate disclosure quality, they built their analysis on the
published AIMR ranking (the report of the Association for Investment
Management and Research) for a sample of 724 american firms related to
33 industries from the year of 1980 to 1994 (a total of 4,478 firm-years).
This index is based on analyst perceptions of value-relevant information from both mandatory
and voluntary disclosure aspects of financial reporting. With respect to the regression
model, the authors used the Collins et al. (1994) model which regresses
current stock return on current and future earnings and stock returns plus
control variables. The future earnings response coefficient assesses so far
the ability of stock returns to predict future performance. The main
findings showed first a positive association between a firms disclosure
activity, as measured by the AIMR ratings of corporate disclosures, and the
amount of future earnings news incorporated in the current annual return.
Inconsistent with their expectations, Lundholm and Myers (2002) found
second no decline in the relevance of current earnings. Results indicated
finally that changes in corporate disclosure activity are positively
associated with changes in the extent of future earnings information
impounded into current stock returns.
Gelb and Zarowin (2002) investigated likewise the association between the
level of voluntary corporate information and share price anticipation of
future earnings. They postulated that high disclosure firms have more
informative stock prices than low disclosure firms. Stock price
informativeness refers to the extent to which current stock return reflects

CHAPTER III
changes in future earnings. Similar to Lundholm and Myers (2002), the
authors measured corporate disclosure level using the annual AIMR-FAF
rating reports for the period of 1980 to 1993. The total sample comprised
821 non-bank firms which are categorized into 22 separate industry
sectors. Gelb and Zarowin (2002) constructed subsequently two sub
samples. The two categories include firms with high and low AIMR-FAF
disclosure ranking as defined in terms of top versus bottom quartile within
their industry for two consecutive years. To implement their tests, the
authors assessed the differential informativeness of current stock prices
for future earnings using the multiple regression model of Collins et al
(1994). They regressed current returns against current and future earnings
for both groups and compared the coefficient on the future earnings
changes. Empirical findings indicated that high disclosure firms have
significantly higher future earnings response coefficients than low
disclosure firms. The authors conducted additional tests whereby they
examined the relationship between certain types of corporate disclosures
channels and share price anticipation of future earnings changes.
To address this issue, Gelb and Zarowin (2002) redefined their disclosure
rating based on three subcategories: the annual report score, the other
publications score and the investor relations score. Results revealed first
and surprisingly, that enhanced annual report disclosures do not improve
stock price informativeness. In contrast, with respect to other publications
and investor relations score, the future earnings response coefficient
(FERC) of high disclosing firms is significantly different from zero. Gelb and
Zarowin (2002) analyzed finally whether their empirical results are
sensitive to the length of the forecasting horizon. The authors investigated
earnings timeliness by using a four year horizon and suggest that if the
higher FERC of high disclosure forms are due to their earnings greater
timeliness within a three year horizon then expanded horizon would
weaker stock price informativeness. The results supported the conclusion
that the association between enhanced disclosure and share price
anticipation of future earnings is not related to differential timeliness
between the two groups of companies but to a genuine difference in

CHAPTER III
informativeness. Gelb and Zarowin (2002) provided first direct evidence
that differences in disclosure quality are the most likely explanation for the
extent to which share price anticipates future earnings. The authors
emphasized also that greater disclosure provides useful information for
investors.
Subsequent to Lundholm and Myers (2002) and Gelb and Zarowin (2002)
studies, recent papers extended these researches by focusing on other
contexts. Among them, Hussainey et al (2003) implemented their study in
the UK context whereby the AIMR-FAF ratings are not available. They
presented a new methodology for assessing the relationship between
corporate voluntary disclosure and price leading earnings. Particularly, the
authors automated the generation of disclosure scores through the use of
standard text analysis software. Unlike, the AIMR-FAF disclosure rating, an
important feature of their new scoring process is the emphasis on forward
looking information. Their scoring system helps scoring a very large
number of annual reports at lower marginal costs, comparing the
disclosure score across firms and over time and replicating this
methodology easily in subsequent disclosure studies. Hussainey et al
(2003) study aimed to test whether voluntary narrative disclosure in the
annual report discussion section influences the strength of the association
between current stock returns and future earnings changes. To measure
the informativeness of such information, the authors proceeded in three
steps: first they identified key words that are related with forward-looking
information in the annual report discussion sections; second, they checked
relevant topics for forecasting process and they finally count for each firm
the number of sentences that are forward looking in nature and include
relevant topics.
Hussainey et al. (2003) argued that forward looking information will enable
the market to better forecast firms future earnings. This should lead to a
greater share price anticipation of earnings for high disclosure firms
compared to low disclosure firms. To examine their prediction, the authors
followed Collins et al (1994) methodology by regressing current returns on

CHAPTER III
current and future earnings. Consistent with Lundholm and Myers (2002)
they interacted firms disclosure quality with all the independent variables
in the model. Based on a sample of 917 firm-year observations for the
period of 1996 to 1999, Hussainey et al (2003) results provided evidence
that forward looking information on profits helps the market predict future
earnings more accurately. Additional tests were also performed in order to
control for eventual measurement errors. The authors controlled for cross
sectional differences in the predictability and the timeliness of earnings by
assigning firms to nine broadly defined industry sectors. The re-estimation
of the multiple regression analysis, holding the industry classification
constant, yields larger and more significant coefficients than their earlier
results. The noticed incremental increase in the FERC suggests that the
failure to control for accounting and business factors is likely to weaken
the voluntary narrative disclosure implication for the return future earnings
relationship.
Hossain et al (2006) investigated the effect of voluntary disclosure,
ownership structure and proprietary costs on the association between
current stock returns and future earnings for Singaporean companies. The
authors extended the Lundholm and Myers (2002) paper of bringing the
future forward by using a direct measure of voluntary information
disclosure. They introduced both ownership structure and proprietary costs
into the current return-future earnings specification. They argued that the
level of corporate disclosure varies with the ownership structure and that
firms with higher managerial ownership are likely to have a weaker
association between current stock return and future earnings. They stated
also that external block ownership may mitigate the weakened association
between current stock return and future earnings due to managerial
ownership. As for the effect of proprietary cost, the authors emphasized
that the disclosure level decreases in relation to proprietary cost and
suggested that firms with higher proprietary cost are likely to have a
weaker association between current stock return and future earnings.

CHAPTER III
Hossain et al (2006) assessed the extent of corporate voluntary disclosure
using a direct and reliable empirical measure that helps differentiate firms
according to their disclosure level. Unlike early research that used
analysts' perceptions of firm total disclosure (mandatory and voluntary
information) such as the AIMR and the CIFAR3 indices, the authors relied on
a self-constructed index for the voluntary disclosure variable. They
developed their index based on a review of relevant sections of disclosure
requirements for Singaporean companies, the annual report award
preliminary screening criteria used by ICPAS (Institute of Certified Public
Accountants of Singapore), the report of business reporting research
project by FASB (2001) and a review of the relevant literature.
The final disclosure checklist included 82 discretionary items related to
both financial and non-financial information. The disclosure index
reflected three categories of information, namely business data,
management analysis of business data and forward-looking information. To
test empirically their assumptions, Hossain et al (2006) selected a sample
of 172 (516 firm-year observations) Singaporean firms pertaining to 8
industries and periodically listed from 1994 to 2000. They relied on Collins
et al (1994) and Lundholm and Myers (2002) empirical model that
regresses current stock return on current and future earnings. Results
revealed first a significant and positive association between voluntary
disclosure level and the amount of future earnings news incorporated into
the current stock return. Findings showed second that firms with high
managerial ownership have a significantly weaker relationship between
current stock return and revealed future earnings than those with lower
management ownership. Nevertheless, this weaker association can be
significantly mitigated by higher outside block ownership in a widely held
company. The presence of government ownership and insider ownership
weakens the current return-future earnings relationship, but not
significantly. Finally, consistent with their expectation, Hossain et al.
(2006) indicated that with regard to proprietary costs, a significantly

3 Centre for International Financial Analysis Research index published


in periodic editions of International Accounting and Auditing Trends.

CHAPTER III
weaker association between current stock returns and revealed future
earnings is noticed when proprietary costs are high. These findings
remained significant even after controlling for the usual determinants
(size, growth, risk, etc.) of the return-earnings relationship. Additional
sensitivity tests were also performed: the authors investigated the relation
between each of these three components of disclosure activity (business
data, management analysis of business data and forward-looking
information) and the return-future earnings relationship. The results
indicated that the significant and positive association between aggregate
disclosure level and the return-future earnings relationship is obvious in
only two of the three categories of corporate disclosure: business
information and forward-looking information. The disclosure of
management discussion of business information was insignificant which
may suggest that investors are more concerned with forward-looking
information and non-financial business information.
Schleicher et al. (2007) contributed to the on-going debate on the
association between voluntary disclosure quality and share price
anticipation of future earnings by discriminating between profitable and
loss firms. They expected that narrative information in the annual reports
will be an important source of information for loss making firms based on
two main reasons.
They argued first, that losses cannot prevail indefinitely in surviving firms
which may indicate that current incomes would not be considered as a
good proxy for loss firms future earnings. Investors will need -to assess
future earnings- further information that explains why the losses occur and
when the firm will eliminate it in the future. Schleicher et al. (2007)
contended second that according to Hayn (1995) the extent to which
annual stock returns are associated with same-period earnings changes
varies considerably between loss-making firms (lower) and profitable
firms. This suggests that in the short-term, the market is responding more
strongly to non-earnings information. The authors expected that this
value-relevant non-earnings information will be also reflected, sooner or

CHAPTER III
later, in reported earnings. The authors adopted the Hussainey et al.
(2003) scoring technique when measuring the quality of corporate
narrative disclosure. They identified and counted automatically the
number of forward-looking earnings words contained in the annual report
narratives section of 2446 UK firms-year observations. Among them, 324
(2122) firm- years are loss-making (profit-making). To measure the extent
to which current stock returns predict future earnings changes, Schleicher
et al. (2007) made three changes to the original model of Collins et al
(1994). First, they included only two years of future earnings growth
variables in the regression model in order to preserve a maximum of
observations with respect to loss-making firms. Second, in calculating the
current and future earnings growth variables, they deflated earnings
change by price and not by lagged earnings. Finally, Schleicher et al.
(2007) excluded the earnings yield variable, EPt-1, from the regression
model. The authors introduce subsequently two dummy variables in
Collins et al (1994) modified model, one for disclosure quality and one for
the existence of a loss.
To test their assumptions, the authors allowed the coefficients on the
earnings variables to vary across profit and loss firms and between high
and low disclosure firms. For loss firms Schleicher et al. (2007) found that
the ability of stock returns to forecast next period earnings change is
significantly greater when the firm provides a large number of earnings
predictions in annual report narratives. Such result wasnt observed for
profitable firms with high quality of voluntary disclosure. In addition, when
controlling for variations in the intrinsic leadlag relation between returns
and earnings across industries, the observed difference between profit and
loss firms became statistically significant.

Banghoj and Plenborg (2008) shed light on the trends in corporate


voluntary disclosure and its value relevance across the Danish capital
market. The authors argued that Denmark is a different context from USA

CHAPTER III
and UK in that it is characterized by high ownership concentration, lower
litigation costs, low enforcement of accounting information and a flexible
accounting regime. They tried to find out whether companies in a setting
with a modest level of accounting regulation fill out the information gap
through voluntary disclosure and thereby improve investor protection.
Based on a comparison between the US and the Danish voluntary
disclosure environments, Banghoj and Plenborg (2008) hypothesized that
in a setting like the Danish with a high level of ownership concentration,
voluntary disclosure is used to reduce the information asymmetry and
large shareholders are likely to discipline the management to produce
value-relevant information. The authors suggested that it is not possible to
predict a priori how the ownership concentration affects the relationship
between the level of disclosure and the prediction of future earnings.
Banghoj and Plenborg (2008) measured the extent to which current and
future earnings news are reflected in current returns using the future
earnings response coefficient (FERC) model. Their approach followed the
work of Kothari and Sloan (1992), Collins et al. (1994), Lundholm and
Myers (2002) and Gelb and Zarowin (2002). Since it is argued that the
level of disclosure is correlated with ownership concentration, size and
leverage, the authors included these variables in the current return-future
earnings regression model as control variables. Additionally, past growth,
risk and the presence of accounting loss were also included as in
Lundholm and Myers (2002) so that to rule out that their disclosure score
index would be merely a proxy for these determinants of the earnings
response coefficient. Similar to Botozan (1997) and Hossain et al (2006),
Banghoj and Plenborg (2008) developed a disclosure index in order to
assess the extent of corporate voluntary information in annual reports.
Their disclosure index was inspired from earlier studies and reports,
including Jenkins (1994), Botosan (1997), PricewaterhouseCoopers (1999)
and the Norby (2001) report. The final checklist included 62 indicators
referring to the following five subcategories: (I) strategy; (II) competition
and outlook; (III) production; (IV) sales and marketing; and (V) human
capital. Banghoj and Plenborg (2008) focused on a sample of 36 industrial

CHAPTER III
firms listed on the Copenhagen stock exchange for the period of 1996 to
2000 which totals 152 firm-year observations. Results documented first an
increased level of corporate voluntary disclosure across time. Particularly
the authors noticed that the level of voluntary disclosure increased by
approximately 40 per cent during the period 19962000.
With respect to the multiple regression analysis, findings showed that
disclosures published in the annual report do not reveal information about
future earnings. This contradicts the findings in Lundholm and Myers
(2002) and suggests that current returns in Denmark are less informative
about future earnings than current returns in the USA. Banghoj and
Plenborg (2008) investigated further whether the type of revealed
information influences its value relevance. For this purpose, they
decomposed the disclosure index according to the five categories
(strategy, competition and outlook, production, sales and marketing, and
human capital) and run a regression for each sub disclosure index. The
coefficients on the disclosure index and interaction terms were not
significantly different from zero. Similar insignificant results were also
detected when including the above mentioned control variables. For
instance, these results showed that voluntary disclosure does not make
current returns more informative about future earnings.
Hussainey and Walker (2009) studied the simultaneous effect of corporate
narrative disclosure in annual reports and firms propensity to pay
dividends on price leading earnings. Notably, the authors examined
whether paying dividends and disclosing voluntarily are substitutes or
complements in corporate communication activity and whether there is
difference in such effect between high- and low-growth firms. They
suggested that dividend policy may constitute an alternative way to
convey value-relevant information to the users of financial information.
Hussainey and Walker (2009) maintained, furthermore, that for high
growth firms, the financial reporting system is unable to capture the valuerelevance of intangible investments on a timely basis. They argued that
high growth and intangible asset intensity are factors that reduce the

CHAPTER III
predictive value of current earnings for future earnings. To perform their
analysis, the authors followed the approach of Hussainey et al. (2003) in
automating the generation of forward-looking earnings disclosure scores.
They used the regression model of Collins et al. (1994) to assess the
impact of voluntary disclosure and dividend payouts on share price
anticipation of future earnings. They selected a sample of 3503 nonfinancial observations for UK firms during the period of 1996 to 2002. Of
those, 1,770 firm-years are high-growth firms and 1,733 firm-years are
low-growth firms. Based on a pooled regression, empirical results revealed
that for companies with high growth, voluntary reporting practices and
dividend payouts significantly influence current share price anticipation of
future earnings. Results showed also some substitution between these two
practices for such firms.
Empirical findings indicated additionally that for high-growth-low disclosing
firms which do not distribute dividends, the market is unable to predict
future earnings change. There was besides strong evidence of share price
anticipation of earnings for three years ahead for low-growth, lowdisclosure firms that do not pay dividends. This suggests that investors
ability to predict future earnings change is neither related to nor increased
by narrative forward-looking information in corporate annual reports.
Hussainey and Mouselli (2010) built on prior research (Schleicher et al.
2007 and Hussainey and Walker, 2009) that investigates the importance of
disclosure quality (DQ) for stock market participants. In particular, the
authors re-examined the value relevance of future-oriented earnings
statements in the annual report narratives section and the degree to which
it improves investors ability to better anticipate future earnings. They also
expanded the market-based accounting literature by constructing a DQ
factor and add it to Fama-French three-factor model, in order to investigate
the usefulness of such factor in explaining the time-series variation of UK
portfolio returns. Hussainey and Mouselli (2010) suggested a different
proxy for information risk built on the basis of DQ that uses the number of
future-oriented earnings statements in annual report narratives as a

CHAPTER III
measure of DQ. They argued that the DQ factor is a systematic risk factor
that is likely to capture information risk and hypothesized that such factor
would be significant in pricing stock returns. To measure the quality of
forward looking information, the authors contended that the concept of DQ
is very difficult to assess since it refers to the degree to which current and
potential investors can read and interpret the information easily.
Hussainey and Mouselli (2010) acknowledge additionally that disclosure
quantity alone, as measured by the disclosure index, is not a satisfactory
proxy for the DQ. For this purpose, based on the framework proposed by
Beretta and Bozzolan (2004), they used the quantity and the richness of
future-oriented disclosures as a proxy for the quality of future-oriented
disclosures. Forward looking information quality was determined by
counting the number statements containing a future prospectus as
developed in Hussainey et al. (2003) and good news information is
considered as a proxy for the information richness criterion. Hussainey and
Muselli (2010) investigated subsequently the effect of corporate DQ on the
association between current annual stock returns and current and future
annual earnings. The authors focused on UK annual reports for the years
between 1996 and 2002. The final sample comprises 3,528 non-financial
firm-years observations.

Results provided evidence that high-disclosure firms exhibit higher levels


of share price anticipation of earnings three years ahead compared to lowdisclosure firms which is in line with previous research (i.e. Hussainey and
Walker, 2009). These findings suggested that future-oriented earnings
statements in corporate annual report narratives as a measure of DQ
contain value-relevant information for the stock market participants.
Finally, Athanasakou and Hussainey (2014) explored managerial incentives
in revealing narrative forward looking performance information in annual
reports. They also examined whether the credibility of forward looking
performance disclosures (FLPDs) depends on corporate reputation with
regard to earnings quality. The authors suggested that given the

CHAPTER III
qualitative and the less easily verifiable natures of forward looking
disclosures (FLPDs), it is difficult for users to assess their accuracy.
Consequently, investors tend to use safeguards when relying on such
voluntary disclosures in anticipating future earnings growth. Athanasakou
and Hussainey (2014) focused on a sample of UK firms and drew from
Hussainey et al. (2003) and Collins et al. (1994) methodologies in
measuring the frequency and the credibility of forward looking information.
They showed first, that managers disseminate more FLPDs when issuing
debt or disclosing bad news in the financial statements (e.g. Earnings
declines, falling short of analyst forecasts or underperforming industry
peers) after controlling for proprietary costs and the firms information
environment. Results indicated second, that despite the average increase
of future earnings news in current returns, investors do not rely on FLPDs
of companies that raise debt or report bad news. Their findings suggested
third that investors decision of whether to rely or not on FLPDs depends
on management credibility cues. In the presence of situational incentives
(debt, bad news), FLPDs seem to help investors to re-assess information in
contemporaneous earnings and to anticipate future earnings only when
reported earnings are of a high quality. Athanasakou and Hussainey (2014)
provided then evidences that managers reporting reputation reflected in
earnings quality is informative and affects the perceived credibility of
corporate forward looking performance disclosures.
Overall, these recent studies highlighted the effect of voluntary disclosure
on share price anticipation of future earnings mainly in the European and
the American context. This is may be due to the availability of subjective
analyst ratings such as the AIMR-FAF ratings for a large sample of firms.
This is likely to provide an opportunity to undertake large-scale disclosure
studies. The Collins et al (1994) future earnings response coefficient was
employed as a standard technique in measuring such effect.
Most of these studies focused merely on the effect of either narrative
forward looking information (e.g. Athanasakou and Hussainey, 2014;
Hussainey and Walker 2009; Hussainey et al. 2003) or business data,

CHAPTER III
management analysis and strategic information (e.g. Hossain et al, 2006)
on the return- future earnings relationship. They were mostly undertaken
in the context of developed countries (US, UK) which present a specific
institutional milieu in term of ownership structure, investor protection and
capital market development and where the scope of voluntary disclosure is
being more and more narrowed. These studies may also suffer from some
omitted variable problems. Indeed, theoretical research suggests that
returns and future earnings may be affected by both governance
mechanisms at a firm level (Bushman et al., 2004) and at a country level.
The quality of corporate voluntary reporting and the effectiveness of its
use by stock market participants is an outcome of a country's governance
regime (Roe, 2003) and its level of investor protection that is related to a
countrys legal system (common law versus civil law) (LaPorta et al. 2000).
In the context of emerging markets, researches on this topic are scarce
despite the wide scope of discretionary disclosure which still covers other
types of information such as corporate risk, corporate governance and
social and environmental information. While Hossain et al (2006) study
was a notable exception, their firm level study is limited to the degree of
forward-looking and business information provided by management as a
proxy of voluntary disclosure, focused solely on the effect of ownership
structure as an external corporate governance mechanism and restricted
to only one country.
Our study contributes then to the empirical literature in this area by
addressing first a different scope of corporate disclosure which is raising
recently a significant interest in the accounting literature that is risk
information. In fact, despite the considerable body of literature reflecting
detailed academic work on risk management, there is still limited research
on corporate risk disclosure (Woods et al. 2007). Moreover, since risk
information is relatively a new field for accounting research, there is
piecemeal evidence on narrative risk disclosures attributes and
determinants and only scant empirical evidence on its economic
consequences. Until to date, the existing few studies are limited to
investigating the value relevance of mandatory risk disclosures provided in

CHAPTER III
line with the Securities and Exchange Commission's requirement FRR
No.48. Firms are required to provide quantitative and qualitative
disclosures about exposure to market risk and to disclose how they
account for derivatives (see, Rajgopal, 1999; Linsmeier et 2002; Jorion,
2002; Liu et al., 2004; Lim and Tan, 2007; Prignon and Smith, 2010).
Accordingly, the lack of empirical evidence on the informative nature of
voluntary risk disclosures motivates us to explore this issue, especially
when Li (2010) evidenced that tone in forward-looking statements in the
MD&A section is associated with future earnings where statements related
to risk and uncertainty are a component of tone. We focus second on other
regions in the world and particularly the MENA emerging markets. Our
study will give some insight into the return-future earnings relationship
among MENA emerging countries where the related literature is relatively
silent. Third, we extend prior research by investigating simultaneously the
effect of corporate voluntary risk disclosure and corporate governance on
the return-future earnings relationship, especially when recent studies
(Wang and Hussainey, 2013) found adherent evidence that the forwardlooking statements of well governed firms improve the stock markets
ability to anticipate future earnings. Fourth, we consider the effect of
proprietary cost on the willingness of management to disclose voluntarily,
especially when the costs and benefits of reducing information asymmetry
through specific voluntary disclosure are not the same across all firms
(Hossain et al. 2006).
Summary
This chapter reviews the core literature on corporate voluntary disclosure
and the relation between revealed information and share price anticipation
of future earnings. The first section of this review gives some insight on
the concept of voluntary information publicity and discusses in detail the
lack of an explicit and precise definition in early papers. Discretionary
disclosure attributes are also highlighted in order to recognize the multidimensional aspect of the corporate communication process. The second
section of the review sheds light on the theoretical background and

CHAPTER III
justifications for corporate voluntary practice. It discusses the information
or lemon problem and management-investors relationship and corporate
disclosure based on agency theory and signalling theory. Drawing from
both theories is likely to give us a greater insight into why managers
voluntarily disclose additional information compared to the financial
compulsory information. The third section discusses the effects of
corporate disclosure on various dimensions. Main capital market
implications for voluntary information are then emphasized. These include
bid-ask spreads, the cost of capital, analyst forecasts and analyst
following. To the extent that the current study is concerned with the effect
of voluntary risk information on the return-future earnings association, we
discussed key papers that are the closest to this area of research in more
details in this chapter.
This chapter ends by discussing the extent to which our study contributes
to the existing market based accounting research literature. It also
discusses the main differences between our study and those published in
other contexts.
The next chapter extends further this literature review by focusing on empirical work related
to corporate governance mechanisms and to proprietary costs, which are likely to shape
corporate reporting activity, and accordingly the market's expectation about firms future
prospects. We believe that apprehending the value relevance of discretionary risk disclosure
should not neglect the costs and benefits of reducing the information asymmetry as well as the
governance structures that may enhance or reduce the usefulness of such information. The
link between corporate governance and the information content of accounting information is
based on the view that corporate governance influences shareholders perception of disclosure
reliability through its influence over management activities, opportunistic behavior and
accordingly the integrity of the financial reporting process. Competition in the product
market, despite the monitoring mechanisms, may conversely discourage managers to reveal
reliable information that should help investors anticipate corporate future earnings growth.

CHAPTER IV
Chapter IV: Corporate governance, proprietary costs and share price
informativeness with respect to future earnings news: the theoretical
framework

Introduction
Recent market based accounting research (MBAR) provided challenging evidence on how
corporate disclosure influences the returns future earnings relationship. It was emphasized
that the ability of stock returns to predict future earnings growth is positively related to the
level of voluntary information. Apprehending the role of discretionary risk disclosure in
improving stock price informativeness with respect to future earnings should also take into
account the costs and benefits of reducing the information asymmetry as well as corporate
governance structures which are not the same across all firms (Hossain et al, 2006).
Proprietary costs and governance mechanisms are shown to shape companies reporting
practices. This would in turn affect the market's expectation about firms future prospects.
Healy and Palepu (2001) argue that the changes in corporate disclosure are unlikely to be
random events: they coincide with changes in firm economics and governance mechanisms.
Limited corporate transparency increases the demands for corporate governance systems. The
latter should alleviate moral hazard problems resulting from a severe information gap between
managers, shareholders and potential investors. Corporate governance mechanisms are
suggested, hence to affect corporate returns and future earnings (Bushman et al, 2004).
Kothari (2001) contends that the most promising area of research in the earnings response
coefficient literature is to relate time-series properties of earnings to economic determinants
like competition, technology, innovation, risk, effectiveness of corporate governance,
incentive compensation policies, etc. For instance, the purpose of this chapter is to suggest
more refinements in the valuation models that we believe to be incrementally fruitful in
furthering our understanding of the returns-future earnings relationship. Basically, we point
out how proprietary costs and differences in corporate governance structures will interact with
risk disclosure and influence the strength of the relation between security returns and revealed
future earnings information.

CHAPTER IV
The organization of this chapter is as follows. The first section develops the literature review
with respect to corporate governance concept and addresses the ownership structure and board
characteristics effect on voluntary disclosure as well as the return-earnings relationship.
Section 2 provides the literature review that relates to the proprietary costs theory and
examines the costs and benefit arguments regarding the voluntary disclosure practice of
relevant proprietary information.
1. Corporate governance mechanisms, voluntary disclosure and share price anticipation
of future earnings
Jensen and Mecklings (1976) agency theory provides a framework linking corporate
disclosure activity to corporate governance. Corporate governance attributes are introduced to
control the agency problem and to ensure that managers act in the interest of shareholders.
Corporate disclosure is also considered as a tool of controlling managers and protecting
shareholders against managers opportunism, making them more likely to release frequent and
regular disclosures. Voluntary disclosures contribute, therefore, to the decrease of agency
costs stemming from the emergence of information asymmetry (Lakhal, 2006).
1.1. The corporate governance concept

The term corporate governance emerged in the 1970s as a framework to explain


contemporary corporate scandals and to solve monitoring problems that arose due to the
separation between corporate ownership and its control. With the word increasing in usage,
it became an institutionalized field of corporate activity and a central topic of academic
research in accounting (Ocasio and Joseph, 2005). Because of its breadth, research on
corporate governance is characterized by the lack of a unifying theory which is obvious in the
variety of definitions and in the heterogeneous studies dealing with the implemented
governance structures. Perhaps the well-known definition of corporate governance is
proposed by Shleifer and Vishny (1997) who argued that from an agency perspective,
corporate governance deals mainly with the issue of how shareholders can ensure that
managers will achieve outcomes that are in their interests. Specifically, they described
corporate governance as the way in which suppliers of finance to corporations assure
themselves of getting a return on their investment (p. 737). The main purpose of governance
mechanisms is to bring the interest of managers in line with those of shareholders. Aggarwal
et al. (2007) backed this view and stated that governance depends both on country-level
mechanisms and firm-level mechanisms.

CHAPTER IV
The country-level governance mechanisms include a countrys laws and the institutions that
enforce the laws, its culture and norms, and the various formal and informal monitors of
corporations. Firm-level or internal governance mechanisms are the mechanisms that operate
within the firm. These mechanisms are heavily influenced by a firms choice of governance
attributes through its charter and policies. Roe (2003) maintained that corporate governance
refers to the relationships among various stakeholders such as shareholders, management,
board of directors, lenders, regulator etc and corporate goals. The author added that issues
of corporate governance are related to corporate accountability, to whom and for what the
company is responsible, and by whom and what standards it is to be governed.
Governance refers accordingly to the structure of rights and responsibilities among the
parties with a stake in the firm (Aoki, 2000: p. 11). It includes controls mechanisms which
ensure that shareholders wealth is not expropriated and reduce the probability that a
manager acts in his self-interest (Kanagaretnam, et al. 2007). Brown et al. (2011)
explained that corporate governance relates to the direction and performance of corporations.
Such statement may not be a particularly revealing from a definitional point of view, but it
points out that corporate governance is about determining the activities in which corporations
are properly engaged. The Cadbury Committee4 suggested that corporate governance is the
system by which companies are directed and controlled (p. 14). The report basically
addressed the respective roles of the board of directors, the shareholders and the external
auditor. While the board of directors is responsible for the governance of their companies,
shareholders' purpose in governance is to appoint the directors and the auditors and to ensure
that an appropriate governance structure is established (para. 2.5)With respect to the
auditor, his role is to provide the shareholders with an external and objective opinion on
firms financial statements (para. 2.7). Brown et al. (2011, p. 99) held that corporate
governance is generally known as a set of rules, relationships, systems and processes within
and by which authority is exercised and controlled in corporations. They believed in the
same way that, corporate governance issues should be confined to the control and the decision
of the shareholders and the board.

Naciri (2008) asserted that the significance of governance system may differ from one context
to another and relates it to the legal framework, the ruling institutions, along with disclosure

4Cadbury Report, 1992, para. 2.5

CHAPTER IV
requirements which impacts the way organizations are managed, not forgetting the financial
market, dictating the behavior of organizations (p. 2). Corporate governance deals as such
with managerial behavior inside corporations and addresses power issues among governing
bodies and those who are governed by them. Naciri (2008) put forward that governance
remains a wide concept given the diversity of practices around the world which nearly defies
a common definition. Up to now, corporate governance is often referred to both internal and
external mechanisms to corporations and linked to realizing precise objectives. The main ones
stress firms ownership, how corporate decisions affect the wealth creation and even how such
wealth will be fairly allocated among corporate stakeholders.
In summary, taking various forms of agency problems and actors into account, corporate
governance has a more comprehensive meaning. As highlighted by Denis and McConnell
(2003), corporate governance is the best set of mechanisms both institutional and market
based that induce the self-interested controllers of a company (those that make decisions
regarding how the company will be operated) to make decisions that maximize the value of
the company to its owners (suppliers of capital) (p. 32). Governance mechanisms aim to
ensure that minority shareholders rights are not usurped, managers actions are monitored,
and poorly performing managers are replaced (Gibson, 2003). They are then meant to serve
several purposes, among them preserving management accountability to owners, monitoring
performance via providing efficient structure to key actors and encouraging the efficient use
of resources so that wealth creation is maximized in the benefit of shareholders.
Effective corporate governance mechanisms have attracted recently increasing attention from
both academics and professionals in response to the recent accounting scandals including
Enron, Tyco, WorldCom, and Global Crossing. In the wake of these scandals, investors
confidence was shaken and companies took steps to strengthen their governance not only
by making boards more independent but also by explicitly charging directors to enhance
corporate transparency through the adoption of higher disclosure standards (Hauswald
and Marquez, 2006). Studies like Ho and Wong (2001) and Patel and Dallas (2002) suggested
that good corporate governance mechanisms are accompanied with good disclosure practices.

To improve transparency and accountability, corporate governance must include a vigilant


board of directors, separate CEO from board chairman, timely financial disclosure and a

CHAPTER IV
transparent ownership structure identifying any conflicts of interests between mangers,
directors, shareholders, and other related parties. Forker (1992) believed that the adoption
of internal control devices such as audit committees and non-executives directors
improves the monitoring quality and reduces the benefits of withholding information. As a
consequence, the quality of corporate disclosure is improved. Conversely, a poor disclosure
practice within a company is associated with poor corporate governance. Improving
transparency and disclosure practice will lead hence to better corporate governance (Chen et
al., 2006).
Overall, Prior academic studies when focusing on the relationship between some corporate
governance attributes and the characteristic of accounting information, refer to two main lines
of research: the first suggests that governance mechanisms are likely to increase the quality of
accounting information whereas the second considers a substitution effect between the two
variables. In other words, corporate governance practices are intended to substitute the low
reliability of corporate financial reporting.
The next subsection addresses how elements of both internal and external corporate
governance mechanisms are likely to influence firms voluntary disclosure and consequently
the ability of stock returns to predict future earnings. Recall that internal governance
characteristics are those that result from the decisions and actions of the shareholders and the
board, such as the constitution and membership of the board of directors, and the structure of
equity ownership. External characteristics include monitoring by outside parties such as
institutional investors.
1.2. Internal corporate governance characteristics

Internal corporate governance mechanisms are basically meant to limit the discretionary
latitude available to managers and to ensure the quality of the financial reporting process. Of
governance practices, the ownership structure, the board of directors structure and the
dissociation between CEO/Chairman duality appear to be important mechanisms that could
influence corporate disclosure policy.
1.2.1. Ownership structure and voluntary disclosure
Several previous studies confirmed the link between corporate decisions to disclose voluntary
information and its ownership structure (Chau and Gray, 2002, Gelb, 2000, Ho and Wong,

CHAPTER IV
2001). Eng and Mak (2003) argued that the structure of ownership determines the level
of monitoring and thereby the level of disclosure(p. 326). We analyze next how
shareholding structure and managerial ownership are related to the level of voluntary
disclosure and accordingly to share price anticipation of future earnings.
1.2.1.1 Shareholding structure

Shareholding structure refers to the fact that companies have a diffuse or a concentrated share
ownership. Fama and Jensen (1983) argued that if the ownership is dispersed, the conflicts of
interest between the principal and the agent would be significant. Financial disclosure seems
to be a major mean, which can efficiently protect shareholders interests against managers
latitude. Demsetz and Lehn (1985) added that shareholders who own a small portion of
corporate equity are less engaged in monitoring activities since they will bear a higher cost
compared to the potential benefits. Those who own larger proportion are encouraged to bear
such costs, especially when benefits are important. As a result, corporate disclosure is likely to
be extensive in widely held firms so that information asymmetry and agency costs will be
reduced (Craswell and Taylor, 1992).
Raffounier (1995) focused on the relation between the level of voluntary disclosure and the
shareholding structure within the entrenchment theory framework. The author argued that in
a widely dispersed ownership, managers will seek to entrench themselves and to convince
shareholders that their economic interests are optimized by revealing more voluntary
information. The incentive behind such attitude is that in a diffused ownership, the costs of
collecting information are very expensive for minority shareholders yielding to information
asymmetry between both sides. However, companies with less dispersed (more concentrated)
ownership are influenced by the ability and motivation of large stockholders to monitor their
interests directly which is likely to influence the expected degree of corporate accountability
to minority shareholders (Shleifer and Vishny, 1997). Ho and Wong (2001) stated that in a
situation of more concentrated ownership the impact on voluntary disclosure is more
complicated.

Conflicts of interest are no more between managers and shareholders, but between large and
small shareholders: large shareholders could exercise their absolute controlling rights,
exerting a powerful influence on managers to maximize their own benefits at the cost of small

CHAPTER IV
shareholders (Makhija and Patton, 2004). Large shareholders and managers can find it
mutually advantageous to work together and to retain corporate information. This cooperation would reduce the ability of other shareholders to monitor managers activities.
Consequently, information is likely to be asymmetric and agency costs are likely to rise
significantly (Brown and Higgins, 2002).
Ho and Wong (2001) assertion was supported by an important empirical research whereby
they have made a comparative analysis of earnings announcement surprises between the USA
and 12 other countries including France. They showed that American companies manage
more earnings surprises by issuing frequently earning disclosures than their counterparts. This
finding might be explained by corporate governance differences, mainly in concern with
ownership structure, which is largely diffused in the US firms. Nasir and Abdallah (2004)
maintained that large shareholders have greater incentives to monitor a firms management
since preserving their welfare depends on corporate performance. The authors predicted that
outside blockholders (concentrated ownership structure) will ask for an increased information
publicity to reduce the information asymmetry among the small shareholders. Hossain et al.
(2006) showed that the existence of outside block ownership is associated with improved
corporate disclosure behavior. Lakhal (2006) found, in contrast, that firms with more
concentrated ownership were less likely to provide voluntary earnings disclosures.
It is obvious from both theoretical arguments and empirical evidence that the shareholding
structures (low levels versus high levels) influences firms willingness to reveal voluntary
information. Shareholding structure matters because closely held companies do not have to
rely on external disclosures to the same extent as companies with dispersed ownership which
have to compensate for lack of owner proximity through higher levels of disclosure (Holm
and Scholer, 2010).

1.2.1.2 Managerial ownership

The association between managerial ownership and corporate voluntary disclosure can be
addressed from two different approaches. According to the agency theory, Jensen and
Meckling (1976) maintain that managerial ownership is positively related to firm value. They

CHAPTER IV
argued that the higher is the percentage of equity held by managers, the more the deviation
from the traditional goal of value maximization is low and the more the firm is profitable. In
such situation, conflicts of interests are resolved and information asymmetry is almost
inexistent. A positive relationship between increased managerial ownership and the level or
the quality of voluntary disclosure is expected. In the opposite way, the entrenchment theory
suggests that in situations of high managerial ownership -where managers obtain effective
control of the firm- managerial ownership and firm value are negatively related because of
entrenchment (Hossain et al, 2006). This theory suggests that actors develop strategies to
maintain their place in the organization and crowd out potential competitors. In doing so, they
make their replacement costly for the organization where they belong, allowing them to
increase their power and to obtain more latitude in determining corporate strategy. Thus,
managers are likely to extract higher wages and larger perquisites from shareholders
(Alexandre and Paquerot, 2000). Managers with important share ownership are more risk
averse for their personal wealth and may initiate decisions inconsistent with a growthoriented, risk-taking objective of enhancing shareholder value (Wright et al., 1996).
Accordingly, managers are likely to disclose less information when managerial ownership is
high because of their intention to evade shareholders' monitoring and/or appropriate the
wealth of minority shareholders (Hossain et al, 2006).
Several empirical studies provided mixed results regarding the effect of managerial ownership
on corporate voluntary disclosure. Warfield et al (1995) and Nagar et al. (2003) provided
evidence supporting the agency theory. They found that the extent of management ownership
(the existence of stock option plans) is positively associated with respectively the amount of
information given about earnings and the level as well as the quality of corporate disclosure.
Nagar et al. (2003) argued that managers are aware about the information needs of firms
shareholders and investors, but are unwilling to publicly reveal it unless they are provided
with appropriate incentives.

They presumed that stock price-based incentives in the form of stock-based compensation and
aggregate share ownership mitigate this agency problem (disclosure gap). Consistent with this
prediction, they found a positive relationship between CEO ownership and disclosure, as
measured by both management earnings forecast frequency and analysts disclosures ratings in
the USA. In Malaysia, Nasir and Abdullah (2004) predicted that, based on the entrenchment

CHAPTER IV
theory, a high managerial ownership is associated with a low extent of voluntary disclosure
given that disclosure policies are left to the discretion of controlling owners. They are
motivated to hold up the minority shareholders by reducing the amount of disclosure in the
annual reports. Inconsistent with these assumptions, the direction of the association between
management interest and the extent of voluntary disclosure supported the agency theory:
executive directors shareholdings are likely to have a positive influence on the level of
voluntary disclosure.
Considerable empirical findings were also in support of the entrenchment effect hypothesis.
Among them, Ruland et al. (1990) suggested that the percentage of equity owned by insiders
explains to some extent managerial behaviors and notably disseminating non-mandatory
information. The author predicted that managerial ownership will be positively related to the
level of management earnings forecast. Ruland et al. (1990) measured managerial ownership
by the percentage of voting stock owned by officers and directors. Inconsistent with their
prediction, results showed a negative and significant relationship between the two variables.
As inside ownership increases, firms provide less management earnings forecast. Eng and
Mark (2003) argued that when managerial ownership is low, there is a higher agency problem.
That is, the manager has greater incentives to expropriate shareholder wealth. In response,
outside shareholders will increase their monitoring of managers behavior which yields to an
increase in firms costs. Voluntary disclosure aims to reduce these costs and should act as a
substitute for monitoring. Eng and Mark (2003) predicted accordingly that voluntary
disclosure increases with the decrease in managerial ownership. Results revealed that higher
managerial ownership is associated with lower disclosure among listed firms in Singapore,
consistent with their prediction. Leung and Horwitz (2004) focused on listed companies in
Hong Kong, which are characterized by a high concentrated director ownership. They
presumed that an increase in directors ownership could align their interests with those of
shareholders resulting in a higher level of voluntary disclosure and in a lower information
asymmetry between both sides.
Leung and Horwitz (2004) contended that the relationship between directors ownership and
discretionary disclosure is nonlinear. When directors ownership became highly concentrated,
agency problems will be rather between directors and minority shareholders. The former are
likely to influence corporate decisions, including the reporting process. The authors expect a
negative impact of highly concentrated directors ownership on corporate willingness to

CHAPTER IV
reveal voluntary information. Leung and Horwitz (2004) found that high (concentrated) board
ownership explains the extent of low voluntary segment disclosure. They noticed that the
contribution of non-executive directors in enhancing voluntary segment disclosure is effective
for firms with low directors ownership but not for concentrated-ownership firms. Finally,
Karamanou and Vafeas (2005) studied how corporate governance structure, as expressed by
corporate boards, audit committees, and ownership characteristics, is associated with
voluntary financial disclosure decisions as measured by management earnings forecasts.
Particularly, they expected that high insider ownership is likely to be related to a better
likelihood of high-quality disclosure and accordingly to the likelihood of management
earnings forecasts. Inconsistent with these predictions, Karamanou and Vafeas (2005) showed
that a high managerial ownership is related to low management earnings forecast frequency
and precision.
In sum, these studies showed that the relationship between managerial/board ownership and
the decision to disseminate voluntary information cannot be considered straightforward.
Managerial incentives with respect to voluntary disclosure are likely to be a function of the
extent of their ownership in corporate equity.
2.

The board characteristics and voluntary disclosure

Of governance practices, the characteristic of the board of directors seems to be an important


mechanism that could influence corporate activities. Fama and Jensen (1983) highlighted that
the board of directors is responsible for setting objectives and monitoring the decision making
process within the firm. The boards size and composition influence its ability to function
effectively. Despite the variability of boards structures between countries and region all over
the world, the objectives of the boards are the same: constrain managers to improve their
transparency and to work in accordance with shareholders expectations and thus enhance
stakeholder investment.
After the recent financial scandals, corporate governance codes in various countries require
increasing non- executive members in the board and the separation between CEO and
chairman to improve its monitoring efficiency.

CHAPTER IV
1.

Board size

In prior studies the effectiveness of the board is often scrutinized with reference to its size and
the frequency of meetings (Gillan, 2006). There are opposite views and arguments on the
relevance of board size in board operations. Resource dependency theory suggests that large
board size has a rich knowledge and more ability to ensure the management of corporate
resources (Pfeffer, 1972). As the board size grows, the boards monitoring ability increases
too. Another view argues that benefits from large size boards may be overwhelmed by the
incremental cost of less effective communication and the reduced decision making
capabilities. Contradiction in thinking or objectives among directors is likely to be associated
with large groups (John and Senbet, 1998). Actually, with dispersed opinions and noncohesiveness in viewpoints, a board that is too large may have diminished monitoring
capabilities (Cheng and Courtenay, 2006). Lipton and Lorsch (1992) state that a large size
board is dysfunctional because a large number of directors are easy to be controlled by the
CEO and, therefore, they cannot reprehend the policies of the top managers or evaluate the
performance of the company truthfully. Conversely, smaller boards are considered to be more
effective in promoting better decision-making (Yermack, 1996). They are expected to be more
active and dynamic, so it is easy to reach consensus.
There has been increasing attention in the accounting literature about the role of the board of
directors in enhancing the quality of financial reporting (e.g., Beasley, 1996). This interest
arises from the board responsibility to monitor the quality of corporate financial reporting.
Linck et al. (2008) suggested that, given firm growth opportunities, board structure reflects
the need for monitoring of activities and the transparency of the firms earnings. Vafeas
(2000) argued that effective board structures (small size) may lead to desirable managerial
behavior proactively. That is, in fear of scrutiny by their firm's effective board, managers may
voluntarily report more accurate earnings information, thereby fending off the unwelcome
monitoring by the board. John and Senbet (1998) put forward that limiting the size of the
board might improve efficiency and improve corporate governance.
The empirical findings of the extant research are conflicting. Although Al Janadi et al (2013),
Samaha et al. (2012), Donnelly and Mulcahy (2008) reported a positive relationship between
board size and extent of voluntary disclosure, Cheng and Courtenay (2006) documented the
absence of a significant association between them and Vafeas (2000) and Arcay and Vazquez
(2005) found that small board size leads to quality monitoring and better disclosure.

CHAPTER IV
The presence of non-executive directors

Both Lefwich et al. (1981) and Fama and Jensen (1983) explain that boards effectiveness in
monitoring managerial decisions depends on its characteristics and its composition in inside
and outside directors: the first ones are likely to have a good knowledge of the firm and the
specificity of its activities, while the second are distinguished by their independence,
competence, reputation and experience in monitoring activities (Elouafa, 2007).
Some previous studies showed how these board characteristics compel managers to meet
shareholders needs and expectations by revealing frequent information and thus, keeping
them steadily informed. They identified three main attributes of a boards effectiveness in
corporate governance, i.e. the proportion of non-executive directors, board size and board
leadership structure (John and Senbet, 1998). The presence of non-executive directors
attracted the most attention of researchers with two theories being relied upon: the agency
theory and the resource dependency theory. Chen and Jaggi (2000) posited that including nonexecutive directors into corporate board was emphasized since the 1980s and was supported
by two main arguments: non-executive directors are likely to assist boards in taking strategic
decisions which may improve firms financial and economic performance. Non-executive
directors serve also to monitor management decisions and board activities. They are
considered to improve the boards ability to reduce agency conflicts between owners and
managers which may occur in the decision to voluntarily disclose information in the
annual report (Barako et al., 2006). Cheng and Courtenay (2006) stated that in the US
context, non-executive directors are shown to play more considerable role in monitoring
managers than do inside board directors. Kanagaretnam et al. (2007) indicated that effective
boards monitoring of corporate management enhances the quality and the frequency of
disclosing information and suggest that information asymmetry is on average lower for
companies which boards are more effective.
Resource dependency theory (Tricker 1984, p. 171) suggests that non-executive directors
provide additional windows on the world. These outside directors bring to the board their
expertise, vast experience, contact and prestige. Their perceived importance as a powerful tool
for constraining management behavior is largely attributed to their independence toward
corporate management (Nasir and Abdullah, 2004). Pincus et al. (1989, p. 246) put forward
that the presence of outside directors on the board should increase the quality of monitoring
because they are not affiliated with the company as officers or employees, and thus are

CHAPTER IV
independent representatives of the shareholders interests. Since outside directors are less
aligned to management, they may be more inclined to encourage companies to disclose more
voluntary information to outside investors (Eng and Mak, 2003).
There is consensus in the literature that the proportion of outside directors to the total number
of directors on the board is related to higher monitoring of financial disclosure quality and to
lower incentives to withhold information in a number of international studies (e.g., Forker,
1992; Ho and Wong, 2001; Chau and Gray, 2002; Haniffa and Cooke, 2002; Eng and Mak,
2003; Makhija and Patton, 2004; Nasir and Abdullah, 2004; Arcay and Vazquez, 2005;
Ajinkya et al., 2005; Barakoet al., 2006; Cheng and Courtenay, 2006; Lim et al., 2007). The
empirical evidences are, however, mixed. For example Ajinkya et al. (2005), Arcay and
Vazquez (2005), Cheng and Courtenay (2006), and Lim et al. (2007) found a significant
positive relationship between boards composed largely of independent directors and voluntary
disclosure in the USA, Spain, Singapore, and Australia, respectively. Ho and Wong (2001)
and Haniffa and Cooke (2002) reported an insignificant relationship between outside directors
on the board and voluntary disclosure in the Hong Kong and Malaysian markets respectively.
Eng and Mak (2003), Gul and Leung (2004), and Barako et al. (2006) showed a significant
negative relationship between non-executive directors and voluntary disclosure in Singapore,
Hong Kong, and Kenya, respectively.
In sum, despite these unclear findings, there are interesting theoretical arguments supporting
the contention that outside directors are a crucial factor in board composition. Greater board
independence seems to be linked to more transparency, better monitoring, and increased
voluntary disclosure.

1.2.2.3. CEO/Chairman duality

Brown et al. (2011) argue that the roles of chief executive officer and chairperson are
fundamental. The chairpersons role includes ensuring the boards activities are carried out
with diligence and information is provided to directors on a timely basis. The CEO is
responsible to the board for overall strategy and investment, and for managing day-to-day
affairs. These roles differ clearly in their nature and emphasis. Governance codes recommend
that they should be assumed by different individuals (see, e.g. Cadbury Report, 1992).
Shareholder activists and key legislators advocate that dissociating the titles of CEO and

CHAPTER IV
chairman will help maintain impartiality, reduces agency costs in corporations, and improves
performance. The Higgs report (2003) adds that the chairman should be an independent
nonexecutive director in order to minimize the possible abuse of power by the CEO.
The idea that concentrated decision-making power as a result of CEO duality may diminish
board independence and reduce management voluntarily release of information is grounded in
the agency theory (Fama and Jensen, 1983). It maintains that, without the direction of an
independent leader, it is much more difficult for the board to be effective when performing its
functions. Fama and Jensen (1983; p. 314) theorize that the organization suffers in the
competition for survival, when decision management and decision control are dominated by
the CEO who also serves the position of chairperson of the board. They contend that the board
of directors is not an effective device for controlling managements decisions unless it limits
the decision discretion of top management. Much of the literature supports strongly this point
of view. Forker (1992) suggested that the separation of the CEO and the chairman position
reduces the benefits of withholding information and promotes information transparency. The
presence of a dominant personality mitigates the role of the audit committee and the
independent directors in controlling managerial decisions and enhancing corporate voluntary
disclosure. The author believed, then that firms characterized by the presence of a
CEO/chairman duality would be less likely to integrate an audit committee and independent
directors and to communicate high quality information. Consistent with Forker (1992),
several studies (Chau and Gray, 2010; Ho and Wong 2001, Cheng and Courtenay 2006, Nasir
and Abdullah 2004, Gul and Leung 2004) maintained that the presence of a dominant
personality in both roles compromises the board independence. One individual possessing a
great amount of power and authority puts in jeopardy the monitoring quality and negatively
impacts the quality of corporate disclosure.
These studies expected that firms in which the chief executive officer holds also the board
chairman position tend to disclose less voluntary information. They ended, however, to some
mixed results: Ho and Wong (2001), Cheng and Courtenay (2006) and Nasir and Abdullah
(2004) found an insignificant relationship between the extent of voluntary disclosure and
CEO duality and accordingly it does not matter whether or not the two positions are
combined. Gul and Leung (2004) and Chau and Gray (2010) found, in contrast a significant
and negative relationship between CEO/chairman duality and the extent of disclosure. This

CHAPTER IV
indicated that firms with non-independent chairman tend to withhold information to outsiders
and reveal a low extent of voluntary information.
3.

External corporate governance characteristics

Brown et al. (2011) explain that external governance characteristics are those beyond the
control of the shareholders and the board. They are considered as either a complement or a
substitute to the internal characteristics, but on both views, external governance characteristics
are likely to influence overall outcomes. Institutional investors are often considered as an
external governance mechanism with the potential to shape the voluntary release of
information.
1.3.1 Institutional ownership and voluntary disclosure
Institutional investors are a special group of shareholders with a relatively concentrated stake
of shares in the equity of large companies. Given their block ownership they are likely to be a
main actor in corporate governance structures (Lakhal, 2006). Compared to individual
investors, institutional investors are both more sophisticated and more important price-setters
in capital markets (e.g., Hand, 1990; Chan and Lakonishok, 1995; Walther, 1997; Sias et al.,
2006). Large institutional shareholders have greater incentives to monitor management
performance in order to secure their substantial investment within companies. Donnelly and
Mulcahy (2008) documented two main reasons why institutional investors apply more easily
their monitoring role on firm management teams. First, institutional investors are
professionals, and so their cost of monitoring when compared with other small shareholders is
significantly lower. Second, institutional investors are better able to evaluate the financial
decisions of management and enjoy greater voting power, making it easier to take corrective
action when it is deemed necessary (Chung et al., 2002).
The role of institutional ownership in reducing the information asymmetry is grounded in the
agency theory which assumes that institutional owners are able to reduce managerial
discretionary power over corporate voluntary disclosure, especially when there is high
managerial ownership (Healy et al. 1999). They have sufficient resources and incentives to
control managers attitudes and especially their performance (Vishny and Shleifer, 1986). For
instance, they can constrain managers to make frequent disclosures in order to adjust their
predictions and minimize the risks related to their investments.

CHAPTER IV
Empirically, Carson and Simnett (1997) found that there is a significant positive
relationship between the percentage of ownership by institutional investors and voluntary
disclosure of corporate governance practices of listed companies in Australia. Similarly,
Elgazzar (1998) set forth that institutional investors can increase the voluntary release of
information. The author observed that a high level of institutional investors is correlated with
low market responses around the date of mandatory releases, and hence, with frequent
earnings pre-emptive disclosures. Bushee and Noe (2000) documented also a significant
positive association between institutional shareholdings and corporate disclosure
practices, as measured by the Association for Investment Management and Research
(AIMR) index. Karamanou and Vafeas (2005) expected that firm characterized by the
presence of a high institutional ownership will disseminate frequently relevant information.
Findings partially ratified this assumption: institutional investors impacted positively
management earnings forecast only in cases where disclosure reports good news. Ajinkya et
al. (2005) showed that block holders affect adversely the frequency and the accuracy of
managerial earnings forecasts. Ajinkya et al. (2005) explained that this observed negative
relationship may be due to their interest alignment with managers and to their ability to
acquire more private and timely pre-disclosure information than small shareholders. Jiang and
Habib (2009) reported that, at levels of high ownership concentration, financial institutioncontrolled companies tend to discourage corporate disclosures. Financial institutions are
likely to have superior information on trading; uninformed investors may feel insecure with
regard to their investment. With limited information available in the financial market and high
information asymmetry, adverse selection problems are expected to be severe. Jiang et al.
(2011) corroborated these findings and showed that information asymmetry as measured by
the bid-ask spread is positively related to financial institutional ownership. This information
asymmetry is attenuated by greater corporate voluntary disclosures which in turn enhance
market liquidity.
In sum, the prominent effect of institutional investors on corporate transparency was well
documented in early and most recent literature suggesting that the voluntary release of
information is a positive function of the amount of the stake held by outside institutional
owners.

CHAPTER IV
4.

Corporate governance and the return-earnings relationship

Aside the compelling empirical evidence on the likelihood of governance structure to shape
the extent of corporate voluntary disclosure, the relationship between earnings
informativeness and corporate governance structures has been also a fashion topic among
academics from different disciplines (Gul and Wah, 2002; Klein, 2002, Beekees, Pope and
Young, 2004, Bushman, et al, 2004, Ahmed et al, 2006). Corporate governance is considered
as pivotal in ensuring the integrity of corporate financial statements by limiting the ability of
management to manipulate earnings. So far, many aspects of corporate governance structure
and its impact on the earnings response coefficient have been scrutinized such as ownership
structure, board size, outside directors or board independence, board leadership or CEOchairperson duality etc. More and more mixed empirical results were reported.
1.4.1 Ownership structure and the return-earnings relationship
There are both theoretical and empirical evidences suggesting that different ownership
structures are associated with different incentives to monitor corporate management, to limit
their discretionary latitude and to ensure the quality of financial information. From an agency
theory perspective, it is believed that corporate performance increases if the level of the
insider ownership increases as the cost of the separation between ownership and control is
reduced. When ownership is dispersed, the principal-agent problems increase due to the
asymmetric information and uncertainty. The entrenchment theory suggests conversely, that
increased levels of insider ownership can result in reduced corporate performance (Demsetz,
1983). As the ownership of controlling owner increases, it is more likely that the controlling
owner uses his power to gain extra benefits at the expense of the minority shareholders (Firth
et al, 2007).

Both academics and practitioners increasingly recognize the importance of effective


monitoring over management in ensuring a useful earnings information. Fan and Wong (2002)
and Donnelly and Lynch (2002) argued, for example, that concentrated shareholding structure
influences the level of information asymmetry between managers and outside investors. This
would in turn influence managers accounting choices and consequently the informativeness
of accounting earnings. Firth et al. (2007) hypothesized that when most of corporate shares

CHAPTER IV
are concentrated in the hands of a few investors, the probability of the financial statements
being managed to meet these investors needs and expectations increases. This reduces the
informativeness of earnings in the eyes of the individual (or minority) investors. Firth et al.
(2007) found that concentration of ownership is negatively associated with abnormal earnings
in China consistent with the entrenchment effect hypothesis. Yeo et al. (2002) and Sarikhani
and Ebrahimi (2011) showed in contrast a positive and significant relationship between the
ownership concentration and earnings informativeness in the Singaporean and the Iranian
context, respectively, while Snchez-Ballesta and Garca-Meca (2007) reported an
insignificant relationship between the two variables based on a sample of Spanish companies.
With respect to managerial ownership, the empirical evidence of Lichtenberg and Pushner
(1994), Mehran (1995), Warfield et al. (1995), Xu and Wang (1999) and Mitton (2002)
showed a positive correlation between firm performance and managerial ownership consistent
with the convergence of interest hypothesis. Agrawal and Mandelker (1990), Jensen and
Murphy (1990), Slovin and Sushka (1993), Loderer and Martin (1997) and Gabrielsen et al.
(2002) etc provided, though evidence of a negative correlation between both variables
consistent with the entrenchment hypothesis. Sarikhani and Ebrahimi (2011) observed no
significant relationship between the managerial ownership and earnings informativeness for a
sample of Iranian non-financial companies.
Zhao and Millet-Reyes (2007) investigated the effect of family ownership and bank
ownership on earnings informativeness. They provided evidence that family ownership
reduces the information content of current reported earnings, and bank-owned firms report
highly persistent earnings through the use of accruals. Institutional investors have been shown
also to have an impact on accounting quality in the U.S. and Iranian context (Chung et al.,
2002; Sarikhani and Ebrahimi, 2011). In particular, consistent with the active monitoring
hypothesis, firms with institutional investors are likely to engage in less opportunistic
earnings management improving hence the informativeness of accounting earnings.
In sum, empirical studies provided mixed results regarding the impact of the
multidimensional ownership structure on the return earnings relationship in different contexts
which mitigate a general evidence and show that the influence of ownership differs across
national jurisdictions.

CHAPTER IV
1.4.2. Board structure and composition and the return-earnings relationship
The boards of directors are also considered to play a considerable role in corporate
governance. They are responsible for controlling and ensuring the disclosure of high quality
of information in the financial statements. They monitor the behavior of managers to
guarantee that their actions are aligned with the interests of shareholders, investors, debtors,
and any other stakeholders. They bear substantial reputation costs if they fail in their duties
(Fama and Jensen, 1983; Srinivasan, 2005; Sarikhani and Ebrahimi, 2011).
The empirical evidences on the association between board structure and composition and the
return-earnings relationship are inconclusive. For example, Beasley (1996) found that firms
with boards dominated by outside directors are less likely to engage in accounting fraud and
Karamanou and Vafeas (2005) concluded that U.S. companies with a high proportion of
outside directors have a higher quality of financial disclosure. Vargas (2000) found, though no
statistically significant evidence of a positive association between outside directors and the
informativeness of reported earnings in the U.S.A. Similarly, Klein (2002) reported no
significant relation between board independence and earnings management. In the U.K.
context, Peasnell et al. (2005) noticed that opportunistic earnings management is reduced as
the proportion of non-executive directors to the total number of directors, increases. Park and
Shin (2004), based on Canadian data, ascertained that outside directors are able to deter
earnings management if these directors have accounting or financial expertise. Firth et al.
(2007) noticed that the presence of independent directors affects the earnings response
coefficient of Chinese companies. Particularly, Firth et al. (2007) showed that firms with
higher proportion of independent directors have greater earnings informativeness and cleaner
audit opinions. They also observed that independent non-executive directors are negatively
associated with the magnitude of absolute discretionary accruals. Dimitropoulos and Asteriou
(2010) and Petra (2007) revealed similarly a positive and significant relationship between
board composition and earnings informativeness.

The duality of the board chairman and CEO positions is another variable that affects the
effectiveness of the board control function. Jensen (1993) argues that when the CEO holds the
position of chairman of the board, internal control systems fail as the board cannot effectively
perform its key control functions. Empirical studies failed to confirm such contention. Firth et
al (2007), Petra (2007) and Sarikhani and Ebrahimi (2011) found that the duality of the

CHAPTER IV
chairman and CEO have no significant impact on discretionary accruals and the earnings
response coefficients respectively. They showed accordingly that the duality of the CEO and
chairman roles cause no decrease in earnings informativeness.
In sum, it was emphasized that the credibility of corporate earnings as measured by the return
earnings relationship is strongly influenced by corporate ownership and board of directors
structures, suggesting that corporate governance mechanisms are linked to higher earnings
quality and informativeness. Taken together, corporate governance mechanisms are involved
in monitoring and determining a firms overall information disclosure policy and quality. Ho
and Wong (2001) suggested that governance mechanisms and corporate disclosure policies
may be either complementary or substitute. They are complementary when implemented
governance structures strengthen the internal control of the firm and make managers unlikely
to retain financial information for their own welfare. This should improve the quality of
corporate disclosure. They are substitute when governance mechanisms decrease the
opportunistic behaviors within the firm, and consequently the need for more monitoring and
corporate reporting (Ho and Wong, 2001). It is worth noting that this substitute relationship
between the level of voluntary disclosure and corporate governance mechanisms is likely to
hold particularly when proprietary cost exists.
While it is argued that the increase in investors ability to anticipate future earnings change is
a desirable consequence of corporate transparency and a higher level of voluntary release of
information, such influence may be moderated when managers realize that revealing private
and proprietary information is likely to weaken the firm competitive position. The next
section highlights the extent to which proprietary costs influence corporate voluntary
disclosure and accordingly investors ability to predict future earnings growth.

2. Proprietary cost, voluntary disclosure and the return-future earnings relationship


Early studies on corporate discretionary disclosure investigated whether there were any
economic forces preventing firms from fully disclosing information and providing high
quality information. Early explanation in the full disclosure models was emphasized by
Grossman (1981) and Milgrom (1981). They hinged on the fact that the information could be
conveyed with little or no cost presuming that there are no conflicts of interest between
managers and shareholders. Subsequent papers showed that the benefits of reducing

CHAPTER IV
information asymmetry and potentially decreasing the cost of raising capital via voluntary
disclosure could be offset by the costs of the disclosure.
Verrecchia (1983) and Bhattacharya and Ritter (1983) modeled an informational equilibrium
where a manager holds additional information and decides whether to disclose this
information or not depending on his objective of maximizing firm's market capitalization.
Their models provided a full disclosure result for low-cost information and a threshold level
of disclosure when information production seems to be costly. Verrecchia (2001; p. 141)
argued that while there are a variety of costs that can support the withholding of information
in equilibrium, arguably the most compelling is the cost associated with disclosing
information that is proprietary in nature.
The next subsections provide an overview of proprietary cost theory and highlight its impact
on the extent of voluntary disclosure and subsequently on the return future earnings
relationship.
2.1. The proprietary costs theory

The voluntary disclosure literature sheds light on theories that explain firms decision to not
reveal information. Dye (2001) posits that the theory of voluntary disclosure can be addressed
from a wider perspective that is the game theory based on the following central premise: if
disclosure is discretionary, firms will decide to reveal favorable information to the capital
market and withhold unfavorable information. Proprietary costs arise when, private
information, if it is released, may damage firms competitive position in the product markets.
Given the presence of competitors and the likelihood of potential entrants to the product
market, there are proprietary costs involved in the decision to disclose or not information that
may be proprietary in nature.
Verrecchia (1983), in an analytical study, explains that the degree of competition faced by a
company may influence corporate incentive and disincentive to reveal information. In his
model, Verrecchia (1983) argues that firms reduce their discretionary disclosure when
proprietary costs arise from it. These costs involve not solely the costs of preparing and
revealing information, but also the costs generated from disseminating strategic information.
Such disclosure is likely to be used by rivals and other stock market participants in a way that
it is potentially damaging firms market position. Verrecchia (1983) adds that the tradeoffs of
disclosure are based on an expected reaction by traders to the disclosure or the non-disclosure.

CHAPTER IV
When proprietary costs exist and firms withhold information, traders are unsure whether the
company retained it because it represents bad news or just because it is not sufficiently good
news to warrant incurring the proprietary costs. The full disclosure premise is no longer valid
which supports the threshold level of disclosure whereby a manager is motivated to withhold
information in certain cases. For instance, Verrecchia (1983) predicts that the market reacts
less negatively to the withholding of information than it would be in the absence of such
costs. Dye (1986) suggests that firms tend to limit the disclosure of both proprietary and nonproprietary information when the latter may partially reveal the former, e.g. when there are
known statistical association between the two types of information. Other analytical studies
examine corporate incentives to disclose relevant information to both investors and
opponents. For example, Wagenhofer (1990) asserts that strategic costs can still be incurred
even in a situation where information is withheld because a competitor may take an adverse
action based on the information signaled by non-disclosure. Wagenhofer (1990) proves that
companies tend to move from full to partial disclosure when proprietary costs are high and the
risk of adverse action by a competitor is rather low ex-ante. In other words, companies may
actually succeed in deterring the opponent from taking action through a lower level of
disclosure. Darrough and Stoughton (1990)s model suggests an additional implication
concerning the disclosure policy of a firm facing a threat of a potential entry as a form
of competition. Their model predicts that this form of competition encourages disclosure,
therefore predicting a positive association between the threat of entry and disclosure.
Darrough and Stoughton (1990) conclude that as the number and the size of rivals increase,
disclosure becomes more costly and the voluntary provision of proprietary information is
discouraged. Their finding seems to contradict Verrecchia's (1983) conclusion that the less
competitive the industries are, the more disclosure will occur.
Notwithstanding the substantial analytical researches that consider the importance of the
proprietary costs as a mechanism modeling disclosure practice, empirical evidences are
limited. Saudagaran and Meek (1997) noticed that proprietary costs have been modeled
analytically, but empirical research on their effects on disclosure is notably absent. Healy and
Palepu (2001) came to a similar conclusion in a review of the empirical disclosure literature
and indicated that there is little direct evidence on the proprietary cost hypothesis (p. 424).
Scott (1994) was one of the first papers to explicitly test the proprietary cost hypothesis using
voluntary disclosures of pension plan information by Canadian firms. He argued that this
information is proprietary with respect to the labor unions. Scott (1994) stated that the

CHAPTER IV
discretionary disclosure equilibrium depends on the favorableness of news and is an
increasing function of proprietary costs. He explained that if the manager holds good news, he
has a great incentive to disclose this information due to the discounted positive effect on share
price. In contrast, when proprietary costs are high, firm value upon disclosure will be lower
and managers have a greater incentive to withhold information. Scott (1994) predicted, hence
that: (1) firms likelihood to reveal pension plan information is negatively associated with
proprietary costs related to this disclosure and (2) the probability of disseminating additional
pension plan information is positively related to the favorableness of the reported news.
Consistent with the proprietary cost hypothesis, he found that the presence of proprietary
costs, reduces the firms propensity to disclose pension information and that additional plan
details is conditioned by the favorableness of such disclosure.
Bamber and Cheon (1998) expected that manager's choice of forecast venue is related to
proprietary information cost. The authors suggested that managers facing higher proprietary
costs will issue earnings forecasts in more reactive venues with narrower immediate
audiences such as meetings with analysts as opposed to special press releases. Two indicators
of proprietary information were examined: growth opportunities and product market
concentration ratios. Bamber and Cheon (1998) highlighted first that growth opportunities
refer to the availability of profitable investments and to the creation of barriers to entry of a
new opponent. They predicted that the higher is the growth opportunities, the lesser is a
corporate propensity to disclose information that would reveal the value of these
opportunities. With respect to product market concentration, the authors indicated that when
firms acquire a strategic competitive advantage, revealing earnings forecasts is likely to
jeopardize such advantage.
Bamber and Cheon (1998) hypothesized accordingly that earnings forecasts for firms
experiencing concentrated product-markets are likely to be more reactive and perhaps even
reluctant to disclose information to small audiences of analysts and reporters. Finding showed
that managers are more likely to issue the forecasts in special press releases when they
face low proprietary information costs as measured by lower growth opportunities. They
also found that high product market competition is related to a low probability of a firm
offering a forecast in a venue with more visibility. This negative relationship extends to the
specificity of the forecast.

CHAPTER IV
Harris (1998) examined the relationship between competition and industry segment reporting
decision. She argued that managers are more reluctant in providing segment disclosures for
operations in highly competitive industries. These disclosures may provide commercially
valuable information to rivals which are not available elsewhere. In fact, segment reporting
gives details about the companys operating margins, return on assets and growth rate in its
different lines of business. It may reveal to competitors and other parties (like customers) the
existence of weaknesses or opportunities to be exploited to their own advantage (Prencipe,
2004). Harris (1998) suggested also that firms in less competitive industries are unlikely to
disclose profitable operations as industry segments. Her findings were consistent with the idea
that high margin operations are proprietary information. Firms in less competitive industries
are reluctant to disclose detailed segment reporting in order to protect their abnormal profits
and market share. Piotroski (2001) investigated likewise firms decisions to provide
additional segment disclosures. He documented an association between voluntary increases
in segment reports and capital market benefits and confirmed that consistent with the
proprietary cost hypothesis, firms with a decreasing and less volatile profitability across
industry segments tend to enhance their segment disclosures.
Overall, proprietary costs arise when private information can benefit to opponents which is
likely to reduce the corporate competitive advantage. The release of favorable news may
encourage potential competitors to enter the market, while information relating to operations
processes and research could be used by the current competitors to increase their relative
market capitalization. Such proprietary costs are likely to be higher, and, likewise, the
disclosure of good news is likely to be lower, in more competitive and concentrated markets.

In support of these evidences, it has been assumed that traders tend to respond less negatively
to undisclosed information within a more competitive industry because they are rationally
aware of the high associated proprietary costs (Kent and Ung, 2003). Proprietary information
is considered as a form of indirect disclosure costs and is therefore difficult to quantify.
Accordingly, proxies are used to account for its important impact on voluntary disclosure.
2.2. Proxies for proprietary costs

Analytical models which exhibited proprietary costs as a moderating force with regard to
corporate disclosure incentives, contend that the probability of adverse action by competitors

CHAPTER IV
and the potential costs of such action, are known to the incumbent. Managers seem to lack, in
contrast, for a solid grasp of these costs and researchers find it difficult to estimate these
parameters (Guo et al., 2004). For instance, proxies for the product market competition are
strongly needed. Tang (2010) asserts that prior analytical models estimate competitive costs,
conditioned on how competition is defined. One group of studies model competition in the
context of an entry game where a firm decides whether to enter a particular new product
market (e.g., Darrough and Stoughton, 1990) while the others model competition in the
context of a post-entry game (e.g., Darrough, 1993). According to Monk (2011), the type of
competition has an impact on the disclosure equilibrium outcomes. The difference between
product market competition and the threat of entry has been shown to be enough to change the
effect of competition on voluntary disclosures. The variability of proxies for proprietary costs
across industries and firms can be drastically different. Although a number of proxies for
proprietary costs were proposed, no proxy has gained general acceptance.
2.2.1. Firm- Level Measures
Prior studies (e.g. Harris, 1998; Piotroski, 2001; Leuz, 2004; Botosan and Stanford, 2005;
Verrecchia and Weber, 2006) used firm level abnormal profits to capture potential strategic
information costs. Cohen (2005) highlighted that corporations which are more profitable and
have higher profit margins attract future competition and exhibit higher forewarning of
potential entrants. In fact, high profit margin reflects a corporate successful strategy that
yields a considerable competitive advantage compared to other existing rivals. Nichols (2009)
sustained that such competitive advantage could be only maintained if competitors cannot
reproduce the companys activities. Consequently, the presence of a competitive advantage
should discourage disclosure of the companys product market activities and business
expansion plans.
Tang (2010) suggested in a similar way, that proprietary disclosure on firm-specific values is
likely to trigger more adverse reactions from rivals when firms are more profitable. Dedman
and Lennox (2009) concluded that if managers face a strong threat of competition in the
product market or if their firms are highly profitable they are likely to withhold sensitive
information from the public sphere. Thus, higher proprietary cost creates a higher threshold
for disclosure, leading to a stronger good news bias associated with disclosure by high profit
margin firms.

CHAPTER IV
Barriers to entry to a given product market were also used as a proxy for corporate
competitive environment costs (Piotroski, 2001; Hou and Robinson, 2005). If the barriers to
entry are relatively high (low), the associated proprietary costs will be reduced (increased)
and accordingly firms propensity to reveal a high level/quality of information will be higher
(lower). Cohen (2005) suggested that high capital intensity is generally considered as an
important barrier of entry for future competitors in the product market. High requirements and
entry costs to a market are decision relevant to potential competitors and are likely to deter
them from entering the given product market. Cohen (2005) added that capital-intensity has a
significant positive effect on reporting quality and that in such situation, incumbent firms will
incur fewer costs when providing additional information that are more informative regarding
future performance.
Another measure of proprietary costs relates to the firms growth opportunities. Growth
opportunities indicate the availability of profitable investments such as new product
introductions, capacity expansion projects, or creation of barriers to entry (Gaver and
Gaver, 1993). Actually, as product markets being more and more innovative, firms are
strongly concerned with investing in intangible knowledge and engaging R&D expenditures
leading to new or improved products (Wang, 2007). King et al. (1990) appraised that property
rights associated with innovations are a major source of proprietary costs. Firms are supplying
considerable efforts toward retaining its unique status and preserving future opportunities by
withholding information from existing and potential opponents. Bamber and Cheon (1998)
posited that firms with higher growth opportunities may have a lower incentive to disclose but
this relationship could be in the opposite direction if a firm looks to create barriers to entry by
signaling that a particular industry or a product market has lower opportunities.
2.2.2. Industry level measure
The industry concentration ratio is the most widely used proxy of market competition in early
and recent theoretical and empirical studies (Verrecchia, 1983; Stiglitz, 1987; Sutton, 1991;
Harris, 1998; Dedman and Lennox, 2009; Ali et al., 2014). Verrecchia (1983) suggests that
competition in the product market discourages full disclosure. He posits that the higher the
industry concentration ratio, the less competitive the industry structure, and the lower the
proprietary information costs are, leading hence to a higher level of disclosure. Sutton (1991)
contends otherwise that intense competition can be related to a higher concentration. In a
cut-throat pricing environment, the survival of some inefficient competitors is endangered

CHAPTER IV
leading to fewer remaining corporations in the product market. Stiglitz (1987) indicates that
competition is likely to be higher in a concentrated market compared to atomistic markets.
Based on a customer research model, he notices that in a concentrated market, it is less
expensive for customers to search for all available prices since there are few suppliers, which
means that customers are better informed about price differentials. Accordingly, competition
among suppliers with respect to prices offer would be more intense in concentrated markets.
Recently, Ali et al. (2014) point out that concentrated industries are less competitive and
indicate that more concentrated industries include a small number of larger firms that hold
higher price-cost margins due to their greater market power, which is consistent with the
theoretical Cournot model of oligopolistic industries. Collectively, these controversial
theoretical models infer that high industry concentration does not necessarily correspond with
low competition intensity.
From an empirical perspective, Harris (1998) argued that since it is known that managers
mitigate proprietary costs in competitive industries through non-disclosure, it is suggested that
such competitive costs are higher in less concentrated industries. Harris (1998) found,
however, that managers are less likely to reveal information about segment operations in less
competitive industries. Considering that when one industry is dominated by a few large firms,
they are likely to realize abnormally high profits, which they seek to preserve by deciding to
restrain their disclosure. Dedman and Lennox (2009) argued however that the relation
between industry concentration and competition is still ambiguous in the economic literature.
Adding to this, empirical researches relying on this type of measure ended with mixed results.
The authors suggested, hence that the relation between industry concentration and
competition could be positive rather than negative as is commonly presumed by empiricists.
Based on managers perceptions of the level of market competition, they found that
managerial insights about existing competitors and the threat of new entry have similar effects
in deterring disclosure.
In sum, it is obvious that proprietary information costs are associated with the product market
competition that is influenced by some firm level as well as industry level characteristics.
2.3. Proprietary costs and voluntary disclosure

As raised in Fields et al. (2001) regarding the costs associated with disclosure choices, prior
empirical studies that investigate the effects of proprietary costs on corporate disclosure
practice addressed mainly one dimension of the disclosure decision. Harris (1998), Botosan

CHAPTER IV
and Harris (2000), and Piotroski (2001) are examples of empirical studies that examined the
effect of competition on the discretionary segment disclosure while Clarkson et al. (1994) and
Bamber and Cheon (1998) investigated mainly the association between proprietary costs and
earnings forecast venue and specificity. Nevertheless, as Shin (2001) correctly pointed out,
considering segment disclosure decision as accurately representing a voluntary disclosure
decision is likely to be impaired specially when dealing with firms that offer one product and
do not have accordingly much discretion regarding the disclosure of information.
Shin (2001) seems to be the first study that investigated empirically the effect of different
types of product market competition on the levels of voluntary disclosure. Shin (2001) argued
that the level of voluntary disclosure depends on the strategic interaction setting in the product
market. He assumed that firms are willing to disclose more when they compete on capacities
as the benefit of achieving a lower cost of capital in capacity competition exceeds the
proprietary cost associated with disclosure for these firms. In contrast, for firms competing on
prices, high proprietary costs of disclosure outweigh the benefits of attaining a lower cost of
capital which drives firms to disclose less. Shin (2001) referred to the AIMR scores to
measure the level of voluntary disclosure. Empirical findings corroborated his theoretical
prediction: it was evidenced that firms engaged in capacity competition disclose relatively
more information than firms engaged in price competition. Notwithstanding the interesting
results, this paper suffers from some limitations. Among them, The AIMR measure of
voluntary disclosure, as mentioned in Botosan (1997), introduces noise and measurement
errors in the research design which make the results difficult to generalize.
Leuz (2004) compared two types of voluntary disclosures, segment reports and cash flow
statements, which differ in their usefulness to competitors, and examined whether these
disclosures were more likely when proprietary costs were low in the German context. He
suggested that cash flow and segment information can be both useful to firms competitors.
Leuz (2004) considered additionally that segment data are most competitively sensitive
compared to cash flow statements. The former may reveal the operating margins and
investments in different lines of business while the latter may inform about firms liquidity at
an aggregated level. Empirical findings support the proprietary cost hypothesis. Leuz (2004)
found that German firms voluntarily release business segment data when the proprietary costs
are low, i.e., when entry barriers are relatively high, segment information is highly aggregated
and firm profitability is low. Cash flow statement disclosures are basically related to measures

CHAPTER IV
that capture capital-market considerations such as cost savings in private information
acquisition and high analyst following rather than proprietary cost concerns.
Prencipe (2004) identified also some proprietary costs factors that are potentially able to
explain the voluntary segment disclosure for a sample of Italian firms. She asserted that
segment reporting provides relevant information to existing and potential opponents. The
disclosure of details about the companys operating margins, return on assets and growth rate
in its different lines of business is likely to be exploited to their disadvantage. Prencipe (2004)
hypothesized that the quality of voluntary segment disclosure is associated with segments and
legally identifiable subgroups of companies, the level of details in segment definition, listing
status age and growth rate. Empirical results showed that, except for growth rate, all the other
above-mentioned variables proved to be significant determinants of segment reporting quality.
Accordingly, it was evidenced that competitive costs related to the risk that disclosed
information may be used by competitors or other parties to the disadvantage of the company
are particularly relevant for segment reporting, thus reducing the incentive for the companies
to provide this information to the market.
Hossain et al. (2006) investigated whether the product market competition is likely to mitigate
the value relevance of corporate voluntary disclosure for a sample of Singaporean companies.
They define proprietary cost as the costs associated with strategic decision-making by a
competitor using all available information, including the private information provided via
voluntary disclosure (p. 506). Hossain et al. (2006) suggested that the proprietary costs
related to the voluntary release of information can impact corporate disclosure practice.
Companies tend to balance their desire to disseminate private information against their need
to protect proprietary information from existing and potential competitors. Empirical results
showed that firms with higher proprietary costs choose to disclose less future earnings
information. Accordingly the value relevance of voluntary disclosure is less for firms with
high proprietary costs because less information is available.
More recently, Dedman and Lennox (2009) examined whether managers perceptions of
product market competition influence their voluntary disclosure decision in the absence of
capital market incentives for a sample of private UK companies. They conjectured that
competitive considerations are likely to be an important determinant of the decision to
withhold information from the public. Dedman and Lennox (2009) suggested additionally that
successful companies have potentially more to waste by disclosing proprietary information

CHAPTER IV
and informing the market about their business operations. Based on managers survey about
their firms competitive environments they found that managers tend to restrain information
about sales and costs if they observe that current or potential competition is likely to be high.
Particularly, they brought into evidence that corporate propensity to reveal such information is
less likely when managers consider that: (a) the company has more current competitors; (b)
there is a high threat of entry into its main product market. Consistent with disclosure being
costlier for successful firms, they also found that the most profitable companies are more
likely to withhold information.
Nichols (2009) focused on factors that are likely to impact firms decisions to disclose nonfinancial information through press releases. He contended that nonfinancial disclosures and
particularly product-related information and business expansion plans may provide
information about a firms strategic action in the product markets in which it competes.
Therefore, firms are likely to face proprietary costs in making these disclosures. Nichols
(2009) predicted that nonfinancial disclosures will occur less frequently, but with strong good
news bias when proprietary costs are high. Consistent with this contention, findings revealed
that when proprietary costs are low, as measured by the firm market share rank within the
industry, disclosure occurs with a greater likelihood but with a weak good news bias. In
contrast, when the firm faces the threat of losing its competitive advantage (measured by
profit margin rank within the industry), disclosure is less likely but occurs with a strong good
news bias.
Overall, the theoretical and empirical literature draws the same conclusion that companies
attempt to withhold information from their competitors since it is argued that releasing private
information is likely to weaken their competitive position. Firms will balance their
willingness to convey more voluntary information under a tradeoff between costs and
benefits. Consequently, it is not clear if better corporate governance mechanisms will always
improve the quality and the frequency of information provided by managers and reduces the
information asymmetry mainly if proprietary costs exist.
Summary
This chapter reviews the core literature on factors that are likely to influence the association
between voluntary disclosure and the returns-earnings relationship. The first section of the
review sheds light on how corporate governance mechanisms are likely to influence the

CHAPTER IV
association between voluntary disclosure and the returns-earnings relationship. Internal and
external governance structures are exhibited as important mechanisms that are likely to shape
the level of corporate voluntary disclosure and the value relevance of accounting numbers.
The second section discusses the effects of proprietary costs on corporate disclosure policy
and notably corporate willingness to issue private information. We reviewed analytical and
empirical studies in order to emphasize the moderating effects that competition potentially
exerts on the level and the quality of corporate disclosure. To the extent that the current study
also aims to identify the simultaneous effect of voluntary risk disclosure, corporate
governance, and proprietary costs on the ability of stock returns to predict future earnings, key
papers that are the closest to this area of research were discussed in more detail in this chapter.
The next chapter develops the research hypothesis and describes the
research design to examine empirically the effect of voluntary risk
disclosure on the returns-future earnings relationship for a sample of MENA
emerging markets.

CHAPTER V
Chapter V: Voluntary risk disclosure and share price informativeness
with respect to future earnings: hypothesis development and research
methodology.

Introduction
This chapter sets out the hypothesis which will be investigated using a quantitative approach
and presents an outline of the research design. The main elements in corporate disclosure
practice are the informational needs of market participants. In order to understand the
usefulness of such disclosure, the relevant literature (chapter III) discusses a variety of
theories that are employed to explain the disclosure activity (e.g. Information asymmetry,
agency theory, signaling theory). Furthermore, as has been noted in the literature review
(chapter III), researchers have widely examined the economic consequences of such practice.
This chapter seeks then to extend these studies and to provide a framework for an empirical
investigation of the effect of voluntary risk disclosure on share price anticipation of future
earnings in the context of MENA emerging markets.
This empirical chapter is organized as follows: The first section discusses our hypothesis
development. It attempts to address one of our main research questions by examining
thoroughly recent empirical evidence on the value relevance of risk disclosure. Section 2
provides an overview of the research design and methods employed in the study. The data
collection process is also highlighted in this section. Section 3 lastly discusses the descriptive
analysis of our sample observations.

CHAPTER V
1. Hypothesis development: prior researches
The focus of the hypothesis is to test whether any relationship exists
between the number of risk disclosure sentences being made within
company annual reports and the ability of stock returns to predict future
earnings. The rationale underlying the development of our hypothesis is
set out below.
1.1 The effect of voluntary risk disclosure on share price anticipation of future earnings

A major motivation for market based accounting research (MBAR) is


providing evidence on how accounting information is perceived as value
relevant by a wide range of users in making economic decisions. The
extent to which market participants are likely to impound relevant
information about a firms prospects into corporate stock prices, (e.g.,
Collins et al.1994; Kothari and Sloan 1992; Warfield and Wild, 1992) has
been of particular interest over the last two decades. Within this
framework, a series of accounting research (e.g. Hussainey et al., 2003;
Hossain et al, 2006; Shleicher et al, 2007; Hussainey and Walker, 2009;
Athanasakou and Hussainey, 2014) examined the cross-sectional variation
in price informativeness with respect to future earnings. These studies
identified a number of corporate attributes and incentives (e.g.
Profitability, growth, earnings quality and ownership structure) as well as
the level of forward-looking disclosure as the main reasons for such
variation. They suggested that forward-looking information enhances
investors ability to predict future earnings, and to make better investment
decisions. Empirical evidences (Hussainey et al., 2003; Hussainey and
Walker, 2009; Hussainey and Mouselli, 2010; Athanasakou and Hussainey,
2014) showed that the level of narrative forward-looking information in the
management discussion and analysis section (MD&A) of annual reports
reduces market uncertainty about corporate future earnings. They also
showed that any revealed news that changes market expectation about
future performance is impounded into the stock price.

CHAPTER V
Recently, along with the increasing sophistication of the international
business environment, stakeholders have asked listed companies to reveal
additional information about their uncertainties and outlooks. In response,
public companies have been improving their communication on risks faced
and their expected impact on future profits (Beretta and Bozzolan, 2004).
Risk disclosure gained accordingly interest in financial reporting activity,
regulation, and international research.
Kravet and Muslu (2013) argued that narrative risk disclosures differ from
other corporate reporting and notably forward looking information, for they
inform users about the scope of future earnings rather than the level of
future earnings. They are likely to elucidate, but do not necessarily unravel
corporate uncertainties. Linsley and Shrives (2000, 2005) suggested that
the disclosure of forward-looking risk information would be particularly
useful to investors. Dietrich et al.s (2001) experiments provided support
for the usefulness of revealing forward-looking risk information, concluding
that overt risk disclosures increase market efficiency. Informative risk
disclosures are believed to influence (increase/decrease) users risk
perceptions, i.e., the range of users forecast of future earnings as well as
users reliance on their predictions (Kravet and Muslu, 2013).
Notwithstanding recent regulatory efforts (IFRS7, Item 1A & Item7 USGAAP, CICA guidance in Canada, FRS7 and FRS32 in Malaysia and
Singapore respectively etc.) to require or encourage corporations to
reveal their assessments about risks and uncertainties in their annual
reports, risk disclosures remain largely discretionary and the value
relevance of such disclosures, to our knowledge, is largely unknown. Few
recent papers addressed this question from different aspects. Li (2008)
focused on the implications of corporate annual reports risk sentiment
(i.e., its emphasis on risk and uncertainty) for future earnings and stock
returns. He documented that an increase in risk sentiment in annual
reports (as captured by count of words risk and uncertainty) is
associated with lower future stock returns and poor future earnings,
suggesting that the stock market does not fully reflect the information

CHAPTER V
contained in the texts of annual reports about future profitability. Kothari
et al. (2009) examined some economic consequences of risk disclosure
disseminated through three main channels. In particular, they analyzed
the impact of risk information provided by corporations, financial analysts,
and business press, on the firms capital market environment. Their
empirical findings showed that favorable total risk disclosures are
associated with a significant decrease in the firms cost of capital, stock
return volatility, and dispersion in analysts earnings forecasts. In contrast,
unfavorable total risk information is accompanied by a significant rise in
the cost of capital, stock return volatility, and analysts earnings forecast
dispersion. Kothari et al. (2009) findings also emphasized that
managements favorable disclosure does not significantly impact the
firms cost of capital, which suggests that companies communications
with investors are not credible.
Kravet and Muslu (2013) examined whether annual changes in narrative
risk-based disclosures in 10-K filings have an impact on users risk
perceptions, as measured by investor and financial analyst activities within
the immediate two months before and after the release of 10-K filings.
Results showed that the annual changes in narrative risk information are
positively associated with variations in daily stock return volatility, trading
volume, the dispersion of outstanding forecasts and the volatility of
forecast revisions. Kravet and Muslu (2013) research provided hence evidence that
risk disclosure reveals new information about corporate risk and uncertainty. Corporate
risk disclosures appear, however, to introduce future contingencies and
new uncertainties rather than solely revising information about known
risks. Risk information is likely to expand the scope of investors forecasts
of future profitability and to reduce investors reliance on their predictions.
Campbell et al. (2014) investigated similarly whether mandatory risk
disclosures are informative about corporate risk factors, and whether
investors impound this public information into the firm value. Their
findings suggested that risk reporting informs about the risks facing the
business, reduces the information asymmetry, and is incorporated into the

CHAPTER V
stock price in a timely manner. Particularly, they found that qualitative
risk factor disclosures in the 10-K filings (in which US firms are asked to
reveal their risk factor) are not boilerplate as they are significantly
associated with the pre-disclosure firm specific risks. They showed also
that textual risk information is associated with low bid-ask spreads, high
beta and stock return volatility following the disclosure of such
information. These results suggested that managers provide useful risk
factor disclosures and investors impound this information into market
values. Finally, Bao and Datta (2014) proposed a new approach to identify
the types of risk information in section 1A of the 10-K form and examined
whether the extent of these textual risk disclosures affect users risk
perceptions. Inconsistent with Kravet and Muslu (2013) and Campbell et al
(2014), they found that around two-thirds of risk types (22 out of 30 risk
types) are lacking of informativeness and have no significant influence on
investors risk perception. Such result provides support to the earlier
claims that risk disclosures in 10-K forms are by and large boilerplate.
Managers tend to reveal generic risk factor disclosures that are not useful
in assessing corporate uncertainties and future cash flows.

Bao and Datta (2014) showed indeed that few types of risk information are
informative. In particular, the disclosure of three types of systematic and
liquidity risks (systematic, funding and credit risks) increases the risk
perceptions of investors, whereas the other five types of firm/industry
specific risks (i.e. Human resources, infrastructure, and regulation
changes) decrease the stock return volatility. Bao and Datta (2014)
explained that these contradictory findings with recent researches are due
to their method that seeks to discover one risk type specific impact on
stock return volatility rather than mix them together.

CHAPTER V
Overall, these recent studies lend support to the usefulness argument of
risk disclosure in annual reports. Based on the above discussion, we
hypothesize that:
H1. Share price anticipation of future earnings is significantly and
positively associated with the level of firms voluntary risk
disclosure.
A rich empirical accounting literature also confirmed some economic and
institutional firm characteristics that are related to reporting quality and
have an influence on the earnings-returns relationship. Past growth, risk,
earnings persistence, firm size, profitability and the presence of an
accounting loss have all been shown to be significantly related to the
coefficient on current earnings. To rule out that our disclosure score is
merely a proxy for these determinants of the earnings response
coefficient, we also include some of them as control variables. Additionally,
to the extent that future earnings and future returns capture expectations
about future earnings imperfectly, the variables identified in prior research
may still be relevant.

1.2 Control variables

Consistent with prior literature on the association between corporate


disclosure and the returns future earnings relationship, we control for the

CHAPTER V
cross sectional effects of some previously documented determinants of
voluntary disclosure as well as the earnings response coefficients on our
hypothesis testing.
1.2.1 Financial leverage
The literature on corporate finance widely recognizes the role of debt as an important
mechanism for addressing the agency problems in corporations characterized by the
separation of ownership and control and for influencing firm value. In fact, the degree of
financial leverage is an important component when assessing corporate risk, stock return
characteristics and firms future prospects and cash flows. Prior studies (Jensen, 1986; Lang
and Lundholm, 1993; Malone et al., 1993; Patton and Zelenka, 1997; Wallace et al, 1994)
considered leverage as a proxy for firms financial risk which can influence firms willingness
to disclose as well as outside parties demand for monitoring information. They predicted that
since firms with higher leverage are financially riskier, greater disclosure may be required to
assure investors confidence, thereby reducing the cost of financing. In that regard, when
leverage is important, creditors will ask for more information disclosure in order to estimate a
firms probability to keep its commitments and to assess the risk of wealth transfer to
shareholders. Similarly, shareholders will require a higher rate of stock return to compensate
this risky situation. A high indebted company will thus disseminate voluntarily and timely
more useful information to reassure both investors and creditors.
Based on the signalling theory, firms with higher debt to equity ratio may be encouraged to
signal their financial health and their strict respect to debt covenants by conveying more
private information. This is likely to influence the confidence of investors and creditors on
firm value. Higher corporate disclosure quality influences, hence not only investors
perception of future expected returns; it also enables them to affect firms future decisions and
cash flows (Lambert et al, 2007). The decrease of information asymmetry will consequently
reduce the uncertainty around firms' performance and future prospects. This would in turn
reinforce firm's credibility and creditability which should lead to improving investors ability
to forecast future growth rate and future earnings (Iatridis, 2008).
Some market based accounting research (Ball et al., 1993; Dhaliwal et al., 1991; Dhaliwal and
Reynolds, 1994) suggested also that the information content of earnings and cash flows is
significantly affected by contextual firm-specific factors and notably the level of financial
leverage. Raising debt would increase default risk and would lead to a lower earnings

CHAPTER V
response coefficient, while paying off debt would decrease default risk and would lead to an
increase in the ERC. Intuitively, this is consistent with the fact that higher levels of default
risk should be related to market perceptions of lower persistence of unexpected reported
earnings. Indeed, for high indebted firms, the increased uncertainty attaching to company
survival would cause investors to revise downward their expectation of future cash flows and
dividends. Watts and Zimmerman (1986), Habib (2008) and Saeedi and Ebrahimi (2010)
evidenced furthermore, that managers of highly leveraged firms, have incentives to
manipulate earnings in order to avoid debt covenant violations. This is likely to decrease the
reliability of current accounting earnings regarding firms future prospects and its ability to
inform investors about corporate future debt paying.
1.2.2 Profitability
Profitability is another factor that is hypothesized to impact positively
corporate disclosure level. Singhvi and Desai (1971) argued that higher
profitability encourages managers to provide greater information because
it enhances investors confidence, which in turn, increases management
compensation. Other studies (Cooke, 1989a, 1989b; Wallace et al., 1994;
Wallace and Naser, 1995) suggested that a highly profitable firm is more
likely to signal to the market its superior performance by disclosing more
information in its annual report.
Regarding risk disclosure Elzahar and Hussainey (2012) expected that
managers of companies with high profitability tend to provide more risk
information in the interim reports, in order to justify their present
performance to the shareholders. Corporate Management is likely to reveal
good news to the market to avoid any undervaluation of their shares
(Giner, 1997). High-profitability firms, accordingly, have relatively greater
incentives to signal their quality and their ability to manage risks
successfully. They may have additionally more resources available to
invest in systems in order to assess and manage their risks (Deumes and
Knechel, 2008).
Vandemele et al. (2009) stated that for poorly performing firms, managers
may experience increased pressure for high risk disclosures. Investors

CHAPTER V
may feel an increased urgency to learn about the drivers of company
performance and risk and when the firm will outstrip their difficulties in the
future. Finally, regarding the earnings response coefficient studies,
previous evidence in Hayn (1995) found that the strength of the relation
between annual stock returns and current earnings changes is
considerably lower for loss-making firms than for profitable firms. This
evidence suggests that in the short-term the market is responding more
strongly to non-earnings information. Schleicher et al (2007) reported that
for profit firms, there is evidence of prices anticipating next periods
earnings change, but there is no such anticipation for loss firms.
Accordingly, this suggests that share price informativeness with respect
current as well as future earnings depends on firm profitability.
1.2.3 Corporate size
Firm size has been always a robust variable in explaining cross-sectional
variations of accounting choices and disclosure. Some studies argued that
firm size may proxy for political costs. Others suggested that analysts and
investors follow more closely larger firms and should have higher
disclosure levels. Riahi-Belkaoui (2001) and Ho and Taylor (2007)
summarized the reasons for a positive association between corporate size
and corporate disclosure as follows. First, larger companies are more likely
to have higher agency costs because higher information asymmetry
between managers and shareholders. Therefore, larger companies are
likely to disclose more information. Second, disclosure costs may generally
be lower for larger companies because the economies of scale. The news
media are more likely to report stories about larger companies and
analysts are more likely to attend their meetings. Third, large companies
are generally more exposed to public scrutiny and to demands for greater
transparency and regulation. With respect to risk disclosure, Elzahar and
Hussainey (2012) contended that large firms rely on external finance; they
have incentives to disclose more risk information to send a good signal to
investors and creditors about their ability to manage risk. Large firms have
an incentive to disclose higher level of risk information in order to clarify

CHAPTER V
their level of return, increase investors confidence and decrease political
sensitivity (Abraham and Cox, 2007; Linsley and Shrives, 2006).
Additionally, large firms have sufficient resources to cover the cost of
additional risk disclosures while small firms are more sensitive to the
disadvantages of revealing risk information to their rivals in the market.
Finally, regarding the return earnings relationship, the differential
information hypothesis suggests that information is impounded into the
stock prices of large firms earlier than in those of small firms (Freeman,
1987). Hodgson and Stevenson (2000) argued that small firms earnings
tend to be more volatile and less predictable than large firms earnings
due to their high transitory components. This may yield to a higher
absolute magnitude of unexpected earnings. Small firms are suggested to
be more conditioned on earnings information because of the lesser
availability of information from alternative sources which may yield to low
price informativeness with respect to future earnings.
1.2.4 Industry sector
Gelb and Zarowin (2002) ascertained that the return future earnings
relationship is likely to be influenced by the cross-sectional differences in
the predictability and the timeliness of earnings regardless of corporate
voluntary disclosure level. Firms with more timely earnings should exhibit
a stronger association between current returns and current earnings. In
contrast, firms with less timely earnings should have a higher relationship
between current returns and future earnings. Besides, if earnings
timeliness is correlated with corporate voluntary disclosure, then the share
price anticipation of future earnings would be driven by differences in the
fundamental lead-lag relation between returns and earnings rather than
disclosure score. It is important then to control for this potential omitted
variable. Gelb and Zarowin (2002) and Hussainey (2004) suggested that
grouping firms by industry serves as an effective control. The industry
grouping is believed to isolate the effect of disclosure level on share price
anticipation of future earnings and to remove inter-industry differences in
reported accounting numbers. This is because firms pertaining to the

CHAPTER V
same industry sector are likely to share common business environment
and to use the same accounting methods. Therefore, their earnings
timeliness and forecastability should be similar.

2. Research methodology
In this section we discuss our approach in measuring the level of voluntary
risk disclosure, the sample selection and our regression model.
2.1 Measuring voluntary risk disclosure

A reliable empirical measure of voluntary disclosure is necessary to assess the level of


corporate reporting practice. Beattie et al. (2004) summarize main approaches
to analyze narrative disclosure in annual reports and these include subjective
analyst ratings, disclosure indices, content analyses, readability studies and linguistic
analyses. In this section, we briefly discuss main approaches, then we select one approach to
apply in our research.
2.1.1 Subjective ratings
Subjective ratings provide an aggregated proxy for corporate disclosure quality. This method
involves the use of analysts scores. The most widespread ratings are the US ratings provided
by the annual survey of the Association of Investment Management and Research (AIMRFAF). Based on this rating approach, disclosure quality is assessed according to three media:
disclosures in annual reports and required published materials, disclosures in quarterly reports
and other non-required published materials, and information provided through investor
relations programs. The overall score of corporate disclosure quality is estimated as a
weighted average of the three individual category ratings (Hussainey, 2004).

CHAPTER V
The AIMR-FAF ratings were used in prior studies on share price anticipation of future
earnings and notably in Lundholm and Myers (2002) and Geld and Zarowin (2002). The
AIMR-FAF rating is considered as a reliable measure of disclosure quality as it covers various
disclosures channels used by firms, including verbal information given during analysts
meetings and conference calls. This rating is based on analyst perceptions of value-relevant
information from both mandatory and voluntary disclosure aspects of financial reporting. The
AIMR-FAF rating is not a direct measure of the voluntary disclosure level, which could be
noisy proxies for voluntary disclosure (Hossain et al, 2006). It is worth noticing also
that a sample based on this rating may be biased towards large firms in
each industry with extensive analyst coverage since it is intimately related
to the American context. The AIMR-FAF rating does not provide then
sufficient variation to conduct meaningful statistical analysis (Botosan,
1997).
Finally, as discussed in Healy and Palepu (2001), it is unclear how seriously
AIMR analysts take these ratings, how firms are chosen to be included in
the rating and how much bias in the ratings is brought. One way to
overcome these basic problems with subjective ratings is the use of selfconstructed disclosure index that is based on a list of disclosure items
when assessing the quality of corporate disclosure.
2.1.2 Self-constructed disclosure indices
Disclosure index studies refer to the assumption that the amount of
disclosure equals the quality of disclosure (see Beattie et al., 2002). This is
because assessing corporate disclosure quality is not an easy task. This
concept refers to the extent to which investors and market participants
can read and interpret the information easily (see Hopkins, 1996).
Measuring investors perception of the qualitative attributes of annual
reports such as reliability, diagnostics, value and the interaction between
different report items would be difficult (Hussainey, 2004). Accordingly,
disclosure index studies assume that the quantity of corporate disclosure
proxy for the quality of corporate disclosure, suggesting that they are
positively related (Botosan, 1997).

CHAPTER V
Beattie et al. (2004) describe that coding can be done binary/ordinal,
weighted or nested (grouping items hierarchical). First, the binary
measurement checks the presence/absence of an item, while the ordinal
measures the frequency of the items. Second, the value weighted index
emphasizes the relative importance of the items and finally, the nested
index groups items into hierarchical categories. Disclosure indices
normally measure the extent of corporate disclosure by allocating a score
of 1 to the presence of disclosure topics and a score of 0 otherwise
(Hussainey, 2004). Healy and Palepu (2001) argue that the selfconstructed indices are a reliable measure that truly captures what is
intended. Nevertheless, the main disadvantage relates to the judgment of
the researcher in the construction of the proxy and the selection of the
items. This would make the research hard to replicate and disclosures
provided outside public documents are not involved in the analysis. Often
the Jenkins report (AICPA, 1994) and the Business Reporting Research
Project by FASB (2001) are used as a basis for disclosure index studies
(Beattie et al., 2004). These reports are basically meant to help companies
improve their business reporting and to provide a guide for an extensive
voluntary disclosure.

Researchers have extensively used disclosure indices to evaluate,


compare and explain differences in the amount of information disclosed in
corporate annual reports. Among them, Banghoj and Plenborg (2008),
Botosan (1997) and Hossain et al. (2006) who examine the economic
consequences of corporate voluntary disclosure.
2.1.3 Content analysis
2.1.3.1 Definitions

The content analysis approach of a written communication involves coding words, phrases,
and sentences against a particular schema of interest (Bowman, 1984). Hussainey (2004)
argues that scholars provided many definitions of content analysis, but the most commonly
adopted definition is the one presented by Berelson (1952). Berelson (1952, p.18) defines

CHAPTER V
content analysis as a research technique for the objective, systematic and quantitative
description of the manifest content of communication. Carney (1972, p. 21) describes the
content analysis as a research technique for making inferences by objectively and
systematically identifying specified characteristics of messages. Abbot and Monsen (1979, p.
504) suggest that content analysis is a technique for gathering data that consists of
codifying qualitative information in anecdotal and literary form into categories in order to
derive quantitative scales of varying levels of complexity. Krippendorff (2004, p. 21)
presents it as a research technique for making replicable and valid inferences from data.
Content analysis based on the above mentioned approaches is pursued as a meaningful tool in
the social sciences. It entails the clustering of text units into different categories and this
classification procedure should be reliable and valid for appropriate inferences. Content
analysis can be conducted either automatically or manually with the latter having the
advantage that it permits the quantitative assessment of achieved reliability (Beattie et al.
2004).
Described as a textual analysis, content analysis exhibits a number of advantages. First, it is
carried out directly on texts and, thus, focuses on a central aspect of social communication.
Second, it is easy to implement and it allows both quantitative and qualitative investigations.
Third, it is relevant for analyzing trends and patterns in written communications over time as
it offers valuable historical insights and permits the examination of textual content for
different periods (Hussainey, 2004). While human-coded content analysis is basically more
persuasive than computer-aided analysis in identifying certain themes in the texts, it has been
the subject of many criticisms.
Notably, this approach is a labor-intensive data collection and extremely time-consuming
process which inevitably restricts the sample size employed by prior studies (Beattie and
Thomson, 2007). Human coders are likely to make mistakes during their analyses. They may
overlook some texts of relevant content, potentially affecting the validity of the measure.
Therefore, the use of automated content analysis emerged in 1980s and it has been developing
ever since with different content analysis software (Hussainey, 2004).
Automated content analysis presents several advantages over the traditional content analysis.
First, it is relatively easy to implement if the materials are archived in an electronic database.
Second, it is an economic technique in terms of time, effort and money, especially when used
to conduct a comprehensive content analysis and to cover sizable samples (Hassan and

CHAPTER V
Marston, 2010). Computer-based content analysis suffers though from some limitations. First,
since the unit of analysis is often a word and as the study is concerned with word frequency,
one should include all possible synonyms and words with multiple meanings (Weber, 1990).
The use of words or keywords in isolation of the meaning of the whole sentence either
electronically or manually does not provide a sound unit of analysis (Milne and Adler, 1999;
Beattie and Thomson, 2007) and can yield misleading results (Hassan and Marston, 2010).
For example, Bontis (2003) and Beattie and Thomson (2007) notice the inferiority of using
automated word search when examining intellectual capital (IC) disclosure. Finally, some
limitations are inherent to the technical characteristics of qualitative software: for Nudist
software, documents have to be strictly in the form of text files while GI software is limited to
English text applications only.
2.1.4 The present study
To assess the level of risk disclosure across different countries, we use the traditional
approach in corporate-reporting studies that is content analysis. Krippendorff (1980) believes
that content analysis ensures replicability and appropriate inference about data in its contexts.
One advantage of the manual approach over an automated content analysis is that humans can
better assess the meaning of words and phrases within their context.

In content analysis, content analyst should make a distinction between mainly two kinds of
units: sampling units and recording/coding units (Krippendorff, 2004). First, regarding the
sampling units, we collect data from companys annual reports, an important source of
information on corporate activities because of its wide coverage and availability to users of
financial information (Adams et al., 1998). Marston and Shrives (1991) argue that the annual
report is the main disclosure vehicle, and it is the most comprehensive financial report
available to the public. Besides, despite the noticed decline in the value relevance of some
financial information (Francis and Schipper, 1999; Collins et al, 1994), the annual report
provides investors with information (in addition to financial statements) that explains
accounting numbers, sketches and presents firms future prospects (Beattie et al., 2002). Lang
and Lundholm (1993) show that the disclosure level in annual reports is positively correlated
with the amount of corporate disclosure communicated to the market and stakeholders using

CHAPTER V
other media. Annual reports provide also a good opportunity for relevant comparisons and
analysis as they are produced on a regular basis (Neimark, 1992). They reflect an improved
reporting activity within which corporations reveal both financial and nonfinancial
information, show leadership and vision and highlight values and positions (Abeysekera,
2007).
We performed the risk disclosures analysis of the sample companies in all the narrative
sections in the annual report and notably the management discussion and analysis (MD&A)
section and board reports. MD&A sections are meant to provide investors with a greater
understanding of corporate business, corporate strategy and performance, as well as how it
manages risk and capital resources. They serve as a tool for managers to reveal their
perspectives of the firm to investors, such as why earnings have changed, what liquidity needs
the firm faces, what capital resources have been or are planned to be used, what material
market risks the firm is exposed to (Brown and Tucker, 2011, p. 310) and what are the future
events and trends that may affect future operations. Brown and Tucker (2011) assert that such
information would help investors assess firms past performance and current financial
condition as well as predict future cash flows.
Second, with respect to recording/coding units, Holsti (1969, p.116) define it as the specific
segment of content that is characterized by placing it in a given category. Unlike
sampling units which assume including or excluding some units from available content for
coding pool, recording units are distinguished to be separately classified.
Thus recording units are typically contained in sampling units, at most coinciding with them,
but never exceeding them (Krippendorff, 2004 pp.99-100). Researchers have at times
employed a variety of different approaches to coding and measurement units, namely words,
sentences and proportion of pages. Drawing from the state-of-the-art content analysis in
narrative reporting and consistent with recent risk disclosure studies (Beretta and Bozzolan
2004; Linsley and Shrives 2006; Amran et al. 2009; Dobler et al, 2011; Elzahar and
Hussainey, 2012) we used the number of risk-related sentences as a measure of risk disclosure
levels. Even though Unerman (2000) argues that words can be counted accurately, they cant
be coded into different risk categories without reference to the sentence. That is, the meaning
of a word typically depends on its syntactical role within a sentence (Krippendorff,
2004 p. 101). Gray et al. (1995, p. 84) assert that sentences are to be preferred if one is
seeking to infer meaning. Particularly in our international study, annual reports are

CHAPTER V
disclosed and/or available in different languages (Arabic, French and English) depending on
the country of investigation. For instance, less bias can be expected when referring to
sentences compared to words.
A further requirement from a content analyst is to specify how to construct the coding
instrument. Krippendorff (2004, p.105) suggests that units in content analysis can be
identified based on five distinctions: physical, syntactical, categorical, propositional, and
thematic. Categorical and thematic distinctions seem to be more relevant to corporate
reporting studies. In particular, categorical distinctions which cluster units by their
membership in a class or category by their having something in common are by far the
most frequently employed instrument in early studies.
There are few coding grids for risk disclosure activity due to the limited academic research
(e.g. Beretta and Bozzolan, 2004; Linsley and Shrives, 2006). In our study, we relied on the
coding instrument used by Linsley and Shrives (2006) and Kajter (2001) which is developed
by one of the professional accountancy firms. Since our focus is on corporate voluntary risk
disclosure, we removed financial risk category from the initial coding grid. Most MENA
emerging markets adopted the international financial reporting standards (IFRS) and are
providing the mandatory information about their market risk (Currency, liquidity and credit
risks). The disclosure index (appendix B) reflects five categories of voluntary risk
information: operations risk, empowerment risk, information processing and technology risk,
integrity risk and strategic risk whereby 32 items were identified.
Following Linsley and Shrives (2006), a broad definition of risk is adopted to identify risk
disclosures. We coded risk disclosures any sentence that informs the reader about any
opportunity or prospect, or of any hazard, danger, harm, threat or exposure, that has already
impacted or may impact upon the company, as well as the management of any such
opportunity, prospect, hazard, harm, threat, or exposure.
Similar to Linsley and Shrives (2006) and Elzahar and Hussainey (2012), we had to adhere to
some decision rules. First, because the definition of risk is broad, disclosures had to be
explicitly mentioned and not merely implied. Subsequently, vague disclosure was not
recorded as a risk disclosure. Moreover, we coded risk disclosure sentence any disclosure that
is repeated each time it is discussed. Any sentence with more than one possible classification
was classified into the category most emphasized within the sentence. We then generated an

CHAPTER V
aggregated score for risk disclosure for each firm by counting the number of risk-related
sentences in corporate annual reports. It should be noted that the instrument we used captures
risk disclosure quantity and not necessarily its quality. Yet, our study does not seek to use
content analysis to assess the quality of risk disclosure rather we aim to investigate the level
of risk disclosure and its impact on the return future earnings relationship.
Since content analysis is inevitably subjective, the coding method has to be reliable and
agreement procedures should be adopted to draw valid conclusions (Bowman, 1984).
Krippendorff (2004) identified three different types of reliability: stability, accuracy and
reproducibility. Stability refers to the likelihood of the data to be coded consistently over time
and can be tested by coding the data more than once (Jones and Shoemaker, 1994, p. 165).
Milne and Adler (1999) report that stability is of least concern when looking to attain the
reliability of coding. It is hence insufficient as a sole criterion for assessing data reliability.
Accuracy is concerned with how well the coding compares to a pre-set standard, in other
words, how much the performance of the coding process is consistent with the performance of
a set of rules that are taken to be correct. Reproducibility issue arises when more than one
coder is introduced. It is often conceptualized through the level of inter-coder reliability,
which measures the extent to which the coding process can be replicated by other analysts,
working independently of each other and applying the same instructions to the same
observations. Krippendorff (2004) contends that reproducibility is the most important
interpretation of reliability in content analysis.
To ensure reproducibility, we used one single coder to perform content analysis. Bowman
(1984) argues that to increase confidence that the interpretation of a written document
correspond to objective reality, there should be more than one person to read and code the
written document. Accordingly, we used an experienced researcher familiar with the
technique of content analysis, to independently, code a sub sample of 5 firms. The sub sample
was selected randomly from the yearly pool of observations from 2007 to 2009 for a total of
15 observations. Before we began the study, we discussed risk disclosure and research
objective with the coder to familiarize him with relevant literature. Additionally, we provided
the coder with a set of rules to replicate the pretest coding and in the process, we clarified and
refined the rules as needed. As the coder and the researcher independently coded the initial
sample, we used tests of inter rater reliability to check for consistency in coding, a proxy for
accuracy.

CHAPTER V
There's no general consensus in the literature on how to do and report inter-coder reliability
tests. In academic research, a variety of agreement and correlation measures are presented
such as Scotts p, Krippendorffs alpha and Cohen's kappa. In our study, consistent with
Hayes and Krippendorff (2007) and Krippendorff (2010), we relied on Krippendorffs alpha
test, which is considered to be the most appropriate test of inter rater reliability. The test
generated a Kalpha of 0.889, a satisfactory level of inter-rater reliability for this intra-class
agreement coefficient. It is customary to require Kalpha= 0.80 as the cut off point for a good
reliability test, with a minimum of 0.67 (Krippendorff, 2004).
It is worth noting finally that we included the pilot companies within the entire company
sample. Our motivation for this is to prevent a material reduction in sample size as a result of
excluding the pilot company documents following Linsley and Shrives (2006).
2.2 Sample selection and data collection

The purpose of this chapter is to investigate the effect of voluntary risk disclosure on share
price anticipation of future earnings in a number of Middle Eastern and North African
countries. As far as we know, the countries studied in this research have not been examined
extensively, despite the growing importance of the region with respect to commerce and
foreign direct investment. Over the last two decades, MENA countries have engaged in
building their capital markets institutional and legal frameworks, which is expected to
develop into healthy investment opportunities.
MENA stock markets are good examples of newly emerging capital markets, with significant
growth potential after the massive privatization plans introduced in the region. These markets
have recently become of significant interest to world investors and policymakers. There has
been an important flow of funds into these emerging financial markets, especially, after the
financial crisis in East Asia and the Argentinean crisis (Ben Naceur et al, 2008). The 2012
world doing business report indicates that most MENA emerging markets are involved in
reinforcing investors protection and corporate transparency. International financial reporting
standards (IFRS) are required in Bahrain, Kuwait, Oman, Qatar, and UAE. Other MENA
countries such as Tunisia, Morocco, Saudi Arabia, Israel are converging to IFRS (Pacter
report, 2014) to attract international investors and to enhance corporate disclosure.
We decided to focus on the emerging markets in the MENA region since it is argued that
investors pressure and demand for additional corporate disclosure are positively related to the

CHAPTER V
level of capital market development. Financial markets are one of the key factors in a
countrys economic development given their critical roles in the process of mobilizing
savings, funding investment opportunities and optimal resource allocation among the different
economic sectors (Ben Othman and Zeghal, 2008). In developed capital markets, investors
demands are more sophisticated. They are exerting considerable pressures toward more
corporate transparency and a high quality of financial information in order to make optimal
choices when analyzing investment opportunities (McSweeney et al. 1984; Adhikari and
Tondkar, 1992). Doupnik and Salter (1995) and Jaggi and low (2000) suggested that a strong
equity market is generally associated with better production and disclosure of relevant
information. Holmstrom and Tirole (1993) asserted that firms with a higher market
capitalization enjoy more confidence from investors, which will in turn impact positively the
value relevance of their financial information.
Our original sample covered twelve MENA emerging markets. We applied some filtering
rules to ensure data availability and sample homogeneity. Consequently, we dropped Israel
from our initial sample because in this country firms are dual-listed and provide annual
reports in conformity with the SEC requirements (10-K form). We also dropped Bahrain and
Qatar because their capital markets include mostly financial and investment corporations.
Our final filtered sample had only nine MENA emerging capital markets:
Egypt, Jordan, Kuwait, Morocco, Oman, Saudi Arabia, Tunisia, Turkey and
UAE. The companies in our countries sample are periodically listed from
2007 to 2012.
We chose this period of analysis for the following main reasons. One,
MENA emerging countries have witnessed a growing GDP per capita during
the last decade, which remained relatively notable and close over the
years before and after the international financial crisis. In fact, except for
2009 (-0.048%) and 2012 (2.229%) the GDP per capita growth is on
average above the 3 % for 2007, 2008, 2010 and 2011, respectively.
Foreign direct investment jumped also from $18.36 billion in 2005 to more
than $33 billion in 2007 and 2008 before falling slightly by nearly 16% and
13% during the next years respectively (2009, 2010 and 2011). Two,
market capitalization exhibited a considerable growth reaching 91.32% of
GDP in 2007 and represented on average 55.29 % of MENA region GDP

CHAPTER V
between 2007 and 2012. This growth is considered the most important
rise during the last decade compared to Latin America (46.61% of GDP)
emerging markets. Three, the notable growth of the capitalization ratio
coincided with a sizable number of listed companies reaching 2163 by
2012 and representing approximately two times the number of listed
companies in Latin America. Four, the total value traded as a share of GDP
(which gives the value of stock transactions relative to the size of the
economy) is on average 41.5%, representing about 4.394% of stocks
traded in East Asia and the Pacific and nearly 82.21% of stock traded in
Latin America (World Bank Development Indicators, WDI 2012).
Since in the context of emerging market disclosure channels are mostly limited, we used
corporate annual reports as the main source for information about corporate risk. We collected
annual reports for companies in nine MENA emerging capital markets from their home pages
if available or from financial market websites. We envisaged also a direct contact and via
email (mainly for Tunisian firms) if annual reports werent available online but there were no
issue for our request.
The firms included in our sample, had to satisfy the three conditions: First,
it had to belong to a non-financial sector. Financial firms such as banks,
insurance firms and investment firms were excluded because their reports
are not comparable to those of non-financial firms. Second, this study
focuses on annual reports and not on other media of financial
communications such as interim reports. Third, the non-financial firm, had
to have at least one annual report, from 2007-2009. We chose 2009 as the
end year for the study because the level of corporate risk disclosure is
linked to share price anticipation of earnings and accounting and return
data are required for at least three years ahead (Year 2012).
The initial number of available annual reports varies from year to year. For example, the total
number of firms in 2007 is 328. This number increases in 2008 to 335 firms, and then it is
reduced to 327 firms in 2009.
We matched the selected companies with the Thomson one database codes from which we
gathered financial information such as stock prices, earnings per share, asset growth, and

CHAPTER V
profit margin ratio. Firms that did not have Thomson one code were excluded from our
sample because they have no accounting and return data. Further observations are
deleted because of missing data and outliers are censored to avoid any
undue influence of extreme observations. Outliers are defined in the
present study as the top and bottom 1% of observations for the
distribution of any of the regression variables.5

After

performing these series of sample-filtering we ended up with a sample reduced from 990 to
809 firm-year observations. It corresponds to 254 firms for 2007, 264 for 2008 and 270
companies for 2009. The following tables 3 & 4 summarize the composition of
our sample by country and by economic sector.
Table3: Sample composition by Country

Country/Year
Egypt
Jordan
Kuwait
Morocco
Oman
Saudi Arabia
Tunisia
Turkey
UAE
Total

2007
19
38
17
16
45
27
24
49
19
254

2008
23
35
18
15
49
28
24
51
21
264

2009
23
50
19
17
42
31
8
51
29
270

Panel
68
122
57
48
146
87
56
154
71
809

Table4: Sample composition by Sector

Economic Sector/Year
Consumer Discretionary
Consumer Staples

2007
51
38

2008
54
39

2009
53
40

Panel
161
122

Excluding extreme observations is consistent with prior literature (e.g., Kothari and

Zimmerman, 1995). In addition, censoring the top and bottom 1% of observations is one
of the acceptable methodologies to reduce departures from normality (see Foster, 1986).

CHAPTER V
Energy
Healthcare
Industrials
Information Technology
Telecommunication Services
Materials
Utilities
Total

11
8
54
6
12
65
9
254

13
8
57
5
12
65
11
264

13
5
59
8
12
69
11
270

36
20
177
19
36
205
33
809

2.3 Research design

2.3.1 Regression model specification


Our study follows the recent trend in the accounting studies (e.g. Hossain
et al, 2006; Gelb and Zarowin, 2002; Hussainey and Walker, 2009;
Lundholm and Myers, 2002; Schleicher et al 2007) to measure the value
relevance of voluntary risk information. We used Collins et al. (1994)
regression model to assess the mix of revealed current and future earnings
news in current stock returns. Based on the observation that accounting
recognition lags stock returns in measuring value creation, this paper
added future earnings and control variables into the regression of current
returns on current earnings.
Collins et al (1994) applied the following specification:
N

k 1

k 1

Rt b0 b1 X t bk 1 X t k bk N 1 Rt k b2 N 2 EPt 1 b2 N 3 AGt
Where:
Rt: stock return for year t
Rt+1, Rt+2, Rt+3: stock returns for year t+1, t+2, t+3 respectively.
Xt, Xt+1, Xt+2, Xt+3: are defined as the earnings change for year t, t+1, t+2,
t+3 respectively.
AGt: is the growth rate of the total book value of assets for period t.
EPt-1: is the period t-1s earnings over price at the start of period t.
Collins et al. (1994, p. 295) worked up their multiple regression model by assuming that the
current return is determined by the unanticipated current earnings growth, the cumulative

CHAPTER V
change in expectations about future earnings and noise. The current annual stock return is
then specified as the sum of the following three components:

Rt 0 1UX 1

k 1

k 1

Et ( X t k ) et
(1)

Rt
Where

is the annual stock return for period t, UXt is the unanticipated

earnings defined as the growth rate of earnings in period t less the prior

periods expectation

(UX X E
( X ))
t
t
t 1 t

Et
,

is the change in market

expectations from the beginning of period t about future earnings in year


t+k and

et

is random noise, unrelated to current or future earnings. Kothari

and Sloan (1992) and Collins et al. (1994) found that share price
anticipation of earnings change is not significant beyond three years and
thus k was limited to three years ahead. To carry out the first specification
empirically, it is fundamental to replace the unobservable independent
variables with observable proxy variables. Warfield and Wild (1992)
suggested to rely on realized earnings as an observable measure for
markets expectations to explain stock returns. The following specification
shows the Warfield and Wilds regression modeling:

Rt b0 b1 X t

bk 1 X t k

k 1

et

(2)

However, as highlighted by Collins et al. (1994), the use of realized


earnings growth rates in equation (2) relative to the equation (1) is likely
to

introduce

errors-in-variables

problem

and

bias

the

regression

explanatory power downward. This problem becomes obvious when


rewriting equation (2) with independent variables decomposed into the
variables of interest from the equation (1) and measurement errors
(Collins et al., 1994; p. 296):

CHAPTER V
N

Rt b0 b1[UX t Et 1 ( X t )] bk 1[Et ( X t k ) UX t k Et 1 ( X t k )] et
k 1

(3)

UX t
Where

Et 1 ( X t )

at t1,

is the unanticipated component of current earnings growth,

is the prior anticipated portion of current periods earnings growth

Et 1 ( X t k )

in period t1,

is the portion of future earnings growth that is anticipated

UX t k

is the unanticipated component of period t+ks

earnings growth generated by news in periods t+1 through t+k.

A closer look at the equation (3) raises a number of measurement error

Xt
problems in the equation (2). Firstly,

expectation from

Et 1 ( X t )

UX t
differs from

by the

which represents old news about Xt that has been

already impounded into past returns and accordingly cannot explain the

current period return. Secondly,

X tk

differs from

UX t k

X tk
information about

on two bases. First,

( E t 1 ( X t k ))
as expressed by the term

may be

already available in the market at time point t1 and thus unrelated to

current period return. Second, new information about

the term

UX t k

X t k

as indicated by

reflects surprises that may be available to the market after

time point t and obviously cannot explain the current return. These last

CHAPTER V
two terms are acting therefore as measurement errors in Xt+k. To mitigate
these measurement error problems, Collins et al. (1994) suggested to add
three proxies for the various measurement errors into the regression
model that includes Xt and Xt+k. These are earnings to price ratio, the

AGt
current growth in book value of assets,

Rt k

and future periods returns,

. Involving these proxies in equation (2) yields the following

augmented regression model6:


N

k 1

k 1

Rt b0 b1 X t bk 1 X t k bk N 1Rt k b2 N 2 EPt 1 b2 N 3 AGt


(4)

Collins et al. (1994) provided the following economic rationale for using
these three proxies. The first measurement error proxy for market

anticipation of earnings growth is the past years earnings variable,

EPt 1

which is defined as the earnings of period t-1 over price at the beginning
of the return window for period t. The authors argue that since price

incorporate information about future earnings,

EPt 1

proxies for the

markets forecast of further earnings growth [i.e., proxies for

Et 1 ( X t k )

Et 1 ( X t )

and

]. In this regard, they suggest that a high price relative to past

years earnings signals a high expected earnings growth for the current
and future periods. Collins et al. (1994) add also that given the earnings to
price ratio and expected earnings growth (the measurement error) are

negatively related, the coefficient on

EPt 1

should be positive. This is

6 Equation (4) corresponds to Collins et al.s (1994:297) Equation (6).

CHAPTER V
plausible because this proxy would serve to subtract the measurement
error from the realized growth rate. The second proxy is the growth in

AGt
asset variable,

, and serves as another proxy for expected future

earnings growth.
Collins et al. (1994) assert that a high level of asset growth indicates that
managers expand their production capacity due to an anticipation of a
greater purchase of their product in the future and subsequently a higher
earnings growth. The authors contend that given that asset growth and
expected future earnings changes are positively related, the coefficient on

AGt
is forecasted to be negative.

The last measurement error proxy for

UX t k

is future periods returns

Rt k

This measurement error is generated by events occurring in the future


periods that affect Xt+k but were not anticipated at the end of period
t. Collins et al. (1994) indicate that if these unanticipated future events
lead to higher earnings growth in period t+k they should also lead to
higher returns in the period when the news becomes known to the market.

Accordingly, future returns should be positively associated with

the coefficient on the

Rt k

UX t k

and

variable is expected to be negative in Equation

(4).
Based on Collins et al. (1994) model, Lundholm and Myers (2002) suggest
that corporate disclosure activity is a significant source of changing
expectations about a firms future earnings. They argue that if a firm
reveals relevant information related to its future performance through its
disclosure activity, then the association between current returns and
future earnings will be improved. They find that if a firm does not disclose
its news about future earnings, then the news will not be revealed to the

CHAPTER V
market and hence not impounded into current returns. It follows that
there exists an interaction effect between future earnings and the
current level of a firms disclosure labeled by Lundholm and Myers
(2002) revealed future earnings. Subsequently, and consistent with
Lundholm and Myers (2002) and Raonic et al. (2011) approach, we
examine the effect of risk disclosure (RDit) on share price
anticipation of future earnings by interacting all right-hand side
variables in Collins et al. (1994) regression model with RDi,t variable
as follows:
3

Rt b0 b1 X t bk 1 X t k bk 4 Rt k b8 EPt 1 b9 AG t b10 RDi, t b11 [ RDi, t * X t ]


k 1

k 1

k 1

k 1

bk 11 [ RDi, t * X t k ] bk 14 [ RDi, t * Rt k ] b18 [ RDi, t * EPt 1 ] b19 [ RDi, t * AGt ] e t


(5)

We made two changes relative to the Collins et al. (1994) regression model. First, we
controlled for the cross sectional variation among corporate characteristics.
A considerable number of studies on the returns-earnings relationship
documented many determinants (past growth, risk, earnings persistence,
firm size, firm profitability and the presence of an accounting loss) of the
earnings response coefficient. In our study, we examine our hypothesis
with some of these determinants. We used firm size, profitability and risk
included in the regressions. We included one control variable in the model
at a time due to the limited number of observations7 (degrees of freedom).
When calculating the current and future earnings growth variables we
deflated earnings change by price and not by lagged earnings. Hussainey
and Walker (2009) argued that it is difficult to define earnings growth
when lagged earnings are negative or zero

7 This corresponds to Lundholm and Myers (2002) and Banghoj and Plenborg
(2008). By including control variables we expect to reduce the residuals,
although they will also affect our degrees of freedom. Controlling variables are
actually interacted with each predictor in the model in the same way as our
variables of interest.

CHAPTER V
Predictions for the signs of the model (5) parameters are summarized in
the following paragraphs according to prior literature (Lev, 1989;
Lundholm and Myers, 2002; Hussainey, 2004). First, we expect the

coefficient

b1

on the current earnings growth, Xt to be positive. Also, we


b2

predict that the future earnings response coefficients,

b3

and

b4

on Xt+k

will be positive. This is may be explained by the fact that industry-wide


and economy-wide effects should allow the market predict some portion of
the firms future earnings growth even though the annual report discussion
sections do not involve risk disclosure information. Besides, we anticipate

b5
negative coefficients

b6
,

b7
and

on the future returns Rt+k (the

measurement error proxy) since any unanticipated future events in period


t would lead to higher earnings change and to positive returns in the t+k
periods when the news becomes available to the market. The coefficient

b8

EPt 1
on lagged earnings

is expected to be positive because of the

negative association between the price to earnings ratio and expected

b9
earnings growth for the current and future years while

is predicted to be

negative because of the positive association between asset growth AGt


and the expected growth in future earnings. The second part of the
equation (5) interacts the variables in the first part with the risk disclosure

measure RDi,t. First, there is no particular prediction for

b10

. In fact,

consistent with Lundholm and Myers (2002) we included RD i,t in the


regression because it is part of the interaction terms [RDXt+k] and to
avoid that the interaction terms would inadvertently proxy for the level of
corporate disclosure.

CHAPTER V
Lundholm and Myers (2002) anticipated also a negative coefficient on [
RDi, t * X t
]

as it is suggested that the level of corporate disclosure may

cause a substitution effect away from current earnings and toward future
earnings. In contrast, Hussainey (2004) argued that enhanced voluntary
disclosure would make earnings announcements more credible and thus,
the sign for b11 will be difficult to forecast. Since [RD i,t Xt+k]8 proxies for the
future earnings news that is revealed by the firms risk disclosures, we

predict the coefficients,

b12

b13
,

and

b14

to be positive. This is because we

postulated that share price anticipation of future earnings changes should


increase with the level of risk disclosure. In other words, a high level of risk
disclosure is likely to reveal information about future earnings. Finally,

bk 14 b18
there are no particular predictions for the coefficients

b19
and

. It is

noteworthy that Rt+k and Xt+k measure the change in expectations about
future earnings. Then it looks plausible that we interact both sets of
variables with RDi,t to proxy for the revelation of future performance news
through risk disclosure.
2.3.2. Variables definition
As discussed above, we tested the association between the level of risk
disclosure and the return-future earnings relationship using Collins et al.
(1994) regression model. In this model, current stock returns is the
dependent variable, while the independent variables include current and
future earnings, future returns, past earnings to price ratio and asset
growth. We collected earnings and accounting data from Thomson one
database. We present the definition of each variable in the next
paragraphs.

8 Note that RDXt+k proxies for the future earnings news that is
revealed by the firms risk disclosures, whereas Xt+k and Rt+k proxy
for all future earnings news, regardless of the source.

CHAPTER V
2.3.2.1. Stock returns

In our study, we calculated annual returns based on monthly stock prices


collected from Thomson one database. Monthly returns are measured as
the growth in the price of a share over months, assuming that dividends
are re-invested to purchase additional units of equity at the closing price
applicable on the ex-dividend date. After computing monthly returns, they
are 'geometrically linked' to produce an annual return using these
formulas:
-

Monthly Returns : r1 (m0, m1) = (Pricem1 - Pricem0) / Price m0


Annual Return: Ryear (2007) = [(1+ r1)*(1+r2)*(1+r3).(1+r12)] - 1

In this regard, we defined return of the year t as the return for the 12
month period starting six months after the financial year-end of year t1.
The chosen return window extends from 6 months prior to the financial
year-end to 6 months after the financial year-end. We chose the six month
lag to ensure that the information in annual report narratives have been

read by the market. On this basis, we calculated

Rt 1 Rt 2
,

and

Rt 3

as the

buy-and-hold returns for the one-year period starting six, eighteen and
thirty months after the firms current financial year-end. Accordingly, the

return windows for

Rt 1 Rt 2
,

and

Rt 3

do not overlap with the current return

window (see Hussainey, 2004).

9 At an early stage, we identified the dates under the chairmens


statements. In most cases, corporations publish the annual reports
within four to five months of the firms financial year-end. We
assumed that one further month is sufficient for market participants
to read and process the information.

CHAPTER V
2.3.2.2. Earnings variables

The measure for earnings per share is provided by Reuters fundamentals


available on Thomson one database which is calculated by dividing
earnings for ordinary-full tax by the number of shares outstanding. We

defined

X t X t 1 X t 2
,

and

X t 3

as earnings change deflated by share price.

We deflated both current and future earnings changes by price at the start
of the return window for period t (see Lundholm and Myers, 2002). For
example, we calculated earnings variables for the year 2007 as follows:

Xt
= (EPS2007-EPS2006)/Price2006,

X t 1
X t 2

= (EPS2008-EPS2007)/Price2006,
= (EPS2009-EPS2008)/Price2006.

X t 3
And

= (EPS2010-EPS2009)/Price2006.

Where: Price2006 is the price six months after the year-end for 2006, in
other words at the start of return window for 2007 year.
2.3.2.3. Past earnings to price ratio

We defined past earnings,

EPt 1

as earnings for the period t1 divided by

price six months after the financial year-end of period t1. For example, we

calculated

EPt 1

for a December 2007 observation by dividing EPS for 2006

by stock price at the end of June 2007.


2.3.2.4. Asset growth

AGt
Asset growth,

, is the growth rate of the total book value of assets for

the year t. We estimated growth of assets as the change in book value of


assets in year t, divided by the book value of assets at the end of t1. For

CHAPTER V
example, asset growth in 2007 equals total assets in 2007 minus total
assets 2006 divided by total assets in 2006.
2.3.2.5. Disclosure level

We defined risk disclosure variable (RDi,t) as the natural logarithm of the number of riskrelated sentences, respectively for operations risk, empowerment risk, information processing
and technology risk, integrity risk and strategic risk information. Logged values improve the
approximation of the independent variables and mitigate potential problems with residuals
such as heteroskedasticity.
2.3.2.6. Controlling variables

Financial leverage

Consistent with Banghoj and Plenborg (2008), Elshandidy et al (2013), we defined leverage
as the book value of equity scaled by total liabilities.

Firm size

We measured the firm size by the natural logarithm of corporate revenues congruent with
Abraham and Cox (2007) and Linsley and Shrives (2006).

Profitability

We measured the firm profitability by the natural logarithm of Return on Equity (ROE) which
is defined as [Net profit after tax/Shareholders funds]*100% in accordance with Elshandidy et
al (2013) and Elzahar and Hussainey (2012).

Industry sector

Industry sector is a dummy variable defined as 1 if the firm is classified


into one of the nine broadly defined industry sectors and 0 otherwise.
3. Descriptive statistics
Table 5 provides summary statistics for our sample with observations coming from the year
2007 to 2009. The mean current return is 3.38 percent. The mean current earnings per share
change is 0.47 percent of the price. The mean of future earnings change is respectively 0.28,
0.45 and 0.3 percent of the price for the three periods ahead t+1, t+2 and t+3. These statistics

CHAPTER V
suggest that the mean accumulated 3 year future earnings is roughly three times the
size of current earnings. Similarly, we report a lower mean future return with respect to t+1
periods compared to current return. We notice a reverse and an increase in the mean stock
returns for period t+2 and t+3 which indicate changes in the returns over the sample time
period10. As also reported in table 5, the mean of corporate asset growth rate is about 12.84
percent and the standard deviation is about 0.179 suggesting that there is little variation in the
corporate asset growth rate among our sample firms.
Descriptive statistics provided in table 6 for risk disclosure level (RD) indicate that the mean
(median) risk disclosure level is increasing over time from 23.41 (19) sentences in 2007 to
29.76 (24) sentences in 200911. This trend is considerably different among firms with an
increasing standard deviation from 19.61 in 2007 to 22.27 in 2009. On a pooled data basis, the
level of risk disclosure is on average 27.26 sentences reflecting a fairly low score over time.
Regarding control variables, first, the mean of corporate financial leverage, as measured by
debt to equity ratio is 41.52 percent and the standard deviation is about 43.16 suggestive that
our sample includes moderately geared firms but high dispersion exists among observations
with regard to their financial leverage level. Second, there is a considerable variation in our
sample firm size. In fact, corporate revenue for the first quartile is $22.19 million, while it is
more than $437 in the third quartile. Thus the sample used in the present study does not
include only large firms in MENA emerging markets. Finally, the mean of corporate
profitability as measured by the return on equity (ROE) is at 25.96 and the standard deviation
is about 149 indicating that firms in our sample are relatively well performing during the
period of analysis though the substantial dispersion among them in term of profitability.

Table .5 Summary descriptive statistics: Panel data

Variabl

Mea

S.D

Min.

25%

Media

75%

Max.

Obs.

10 The paired samples t-tests for difference in means between the current period
and the t + 2 future period as well as thecumulated 3 year future returns are
significant at the level of 5%
11 The paired samples t-tests for difference in means between the level of risk
disclosure in 2007, 2008 and 2009 are significantat the level of 5%

CHAPTER V
es
Rt
Xt
Xt+1
Xt+2
Xt+3
Rt+1
Rt+2
Rt+3
AGt
Ept-1
RD
Lev
Size (M$)
Profit

n.
0.0338 0.464
0.0047 0.046
0.0028 0.042
0.0045 0.043
0.003
0.034
0.0051 0.4421
0.1083 0.3631
0.0447 0.2850
0.1284 0.1796
0.8855 1.3427
27.265 21.181
41.527 43.16
660.55 1930.02
25.964 149.132

-0. 597
-0.078
-0.073
-0.070
-0.0617
-0.604
-0.3538
-0.3478
-0.0894
-0.02
0
0.000
0.83
0.05

-0.340
-0.0173
-0.018
-0.0179
-0.0107
-0.3402
-0.1718
-0.1854
-0.0108
0.03
13
3.164
22.19
0.12

n
-0.044
0.002
0.0001
0.0002
0.000
-0.0679
0.0350
0
0.082
0.17
22
25.892
117.07
14.59

0.317
0.027
0.023
0.024
0.025
0.2816
0.3125
0.2519
0.2327
1.18
35
71.209
437.39
184.3

0.945
0.0925
0.079
0.0875
0.064
0.8514
0.8139
0.5532
0.4919
4.05
145
127.23
25101.77
3216.189

809
809
809
809
809
809
809
809
809
809
809
809
809
809

Table 5 reports the summary statistics for the sample firms using data pooled across the three year sample period.
The earnings per share measure is a Reuters fundamental item, calculated by dividing earnings of ordinary-full
tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price.
Both current and future earnings changes are deflated by the price at the start of the return window for period t.
Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period,
starting six months after the end of the previous financial year. EPt1 is defined as period t1s earnings over
price six months after the financial year-end of period t1. AGt is the growth rate of the total book value of assets
for period t. RD is the total number of risk related sentences, respectively for Operations risk, Empowerment
risk, Information processing and technology risk, Integrity risk and Strategic risk information. Unlogged values
are reported. In subsequent regressions the natural logarithm is used. Financial leverage Lev is defined as book
value of equity scaled by total liabilities. Size is measured by corporate revenues. Profit is measured by the
Return on Equity (ROE). Likewise, unlogged values are reported and in subsequent regressions the natural
logarithm is used.
Table 6 Descriptive statistics for Risk Disclosure level from 2007-2009

2007
2008
2009

N
254
264
270

Mean
23.411
27.640
29.762

Median
19
22
24

Min
0
1
1

Max
140
145
144

S.D
19.616
21.092
22.276

Table 7 presents Pairwise Pearson Correlations for all regression variables. P values are given
in parentheses. Correlations are estimated using data pooled across the three-year sample
period. As documented in previous studies, the correlation between current Returns (Rt)
and current earnings growth (Xt) is strong and significant at the 1 % level suggesting that
current earnings are perceived as value relevant. Besides, the correlation between Rt and
Xt+1 is weaker, but still significant at the 5 % level. The correlation between Rt and Xt+2 and
Xt+3 is not significant. This may provide evidence of prices leading earnings by only one
period.
Current return is correlated with future returns regarding period t+1 while Rt is uncorrelated
with Rt+2 and Rt+3. Future returns (Rt+1, Rt+2 and Rt+3) are significantly correlated with
future earnings growth (Xt+1, Xt+2 and Xt+3), consistent with Collins et al. (1994). These

CHAPTER V
correlations indicate that future returns should not influence the results except through
their role as a proxy for the measurement error in future earnings. Partial correlation shows
also a significant and negative correlation between current earnings (Xt) and future earnings
growth for period t+1 and t+2. Similarly, Xt+1 is significantly and negatively related to Xt+2 and
Xt+3. This may suggest potential multi-collinearity problems within the independent variables.
Nevertheless, the variance inflation factor (VIF) did not raise significant problems among the
explanatory variables. In fact the mean VIF is about 3.24 and the computed VIF for
each predictor variable is under 5. There is additionally some evidence that AGt as well as
Ept-1 are a good estimation error proxy. In theory, an errors-in-variables proxy should be
highly correlated with the measurement error but uncorrelated with the dependent variable.
This is the case for these two controlling variables. In table 5, the correlation coefficients
between Rt in one hand and AGt as well as Ept-1 in the other hand, are respectively 0.035(0.32)
and -0.024(0.489) both insignificant. Finally, risk disclosure level is positively and
significantly associated with current stock returns at the 10% level. This may indicate that risk
disclosure practice for firms in MENA emerging markets is driven by stock price
performance. It is worth noting that Lundholm and Myers (2002) find in the same way a
positive and significant correlation between current returns and the disclosure score.

CHAPTER V
Table .7 Pearson Correlations: Panel Data

Rt
Rt
Xt

Xt+1

Xt+2

Xt+3

Rt+1

Rt+2

1.000
0.143**
*
(0.000)
0.0843
**
(0.016)

Xt

Xt+1

Xt+2

Xt+3

Rt+1

Rt+2

1.000

1.000
0.200**
*
(0.000)
0.0132
-0.059* 1.000
(0.706) (0.093) 0.144**
*
(0.000)
0.0132
-0.035
-0.062* (0.708) (0.308) (0.077) 0.138**
*
(0.000)
-0.036
0.170** 0.153**
0.249** (0.301) *
*
*
(0.000) (0.000)
(0.000)
0.0175
0.032
0.004
0.067*
(0.619) (0.358) (0.907) (0.054)

1.000

0.061*
(0.081
)
0.132*
**
(0.000
)

1.000

1.000
0.169**
*
(0.000)

Rt+ 3

EPt1

AGt

RD

CHAPTER V
Rt+3

0.054
(0.123)

0.018
(0.599)

EPt

-0.024
(0.489)

-0.040
0.073** (0.244)
(0.036)

AGt

0.035
(0.320)

0.073** -0.034
(0.036) (0.322)

RD

0.0622 -0.021
*
(0.544)
(0.076)

-0.002
(0.974)

-0.005
(0.880)

0.116**
*
(0.001)
-0.060*
(0.087)

0.067*
(0.053
)
-0.038
(0.273)

0.041
(0.171)

0.039
(0.256)

1.000

0.014
(0.686)

0.003
(0.914)

-0.002
(0.951)

0.020
(0.558)

-0.056
(0.110)

0.099*
**
(0.004
)

0.027
(0.431)

-0.020
(0.569)

0.085**
(0.015) 0.064*
(0.065
)

0.072*
*
(0.040
)
0.083*
*
(0.017
)
0.002
(0.955)

1.000

0.063*
(0.070
)

1.000

0.162*
**
(0.000
)

0.030 1.00
(0.38 0
5)

Table 7 presents Pearson Correlations for all regression variables using panel data across the three year sample period. P-values are given
in parentheses. The number of observations is 809. The earnings per share measure is a Reuters fundamentals item, calculated by
dividing earnings for ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated
by price. Both current and future earnings changes are deflated by the price at the start of the return window for period t. Rt, Rt+1, Rt+2 and
Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting six months after the end of the
previous financial year. EPt1 is defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is the
growth rate of the total book value of assets for period t. RD is the natural logarithm of the total number of risks related sentences,
respectively for Operations risk, Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk
information.

CHAPTER V
Summary
The main hypothesis in this chapter is based on risk disclosure literature and on value
relevance studies within the developed literature review (chapter III). It predicts that a
high level of narrative risk disclosure is likely to enhance the share price informativeness with
respect to future earnings. This chapter explains also the research methodology to adopt for an
empirical assessment of this contention. We collected our data for a sample of firms listed in
nine MENA emerging markets. We applied content analysis, defined the measures of our
variables and developed the regression model. Content analysis has been widely criticized by
the researchers because of its subjectivity. Scholars suggest however many steps to address
this shortness, steps which are followed in this thesis in order to achieve acceptable
reliability and enhance the validity of this study. We relied on the regression model developed
by Collins et al (1994) and considered as a benchmark for recent studies on the return-future
earnings relationship. The chapter concludes with presenting the descriptive statistics of our
data and with highlighting the pairwise correlations between our predictors.
The next chapter provides a discussion of our main findings regarding the value relevance of
voluntary risk disclosure in corporate annual reports.

CHAPTER VII
Chapter VI: Voluntary risk disclosure and share price informativeness
with respect to future earnings: empirical evidence.

Introduction
This chapter presents the main findings regarding the effect of risk disclosure on share price
anticipation of future earnings using both cross sectional and panel data
analysis. It is divided into two sections. Section1 examines the regression diagnostics.
Given that the cross sectional analysis provides a snapshot of the relation between the tested
variables and the observed outcomes at one time point, our investigation is complemented
with a panel regression analysis consistent with a large body of accounting and financial
literature. Recall that it is important for researchers to make sure that all the regression
assumptions are met and valid before drawing any conclusion. Accordingly, statistical tests
are applied to support the findings and draw valid inference. Section 2 discusses the main
empirical results for this chapter. Section 3 provides additional robutness tests. It is
worthnoting that this chapter should be reviewed in line with the discussion provided in
the methodology (chapter V) which explains the process and methods applied to end up
with these findings.
1. Multiple regression analysis
To examine the suggested framework of the economic consequences of
voluntary risk disclosure, we run year-by-year regressions in addition to a
panel regression with all firm-year observations. We used two types of
regression estimates: an OLS linear regression for the cross-sections 2007
to 2009 and a fixed effect regression for the panel data.
1.1 Linear regression diagnostics

The linear regression model is considered to be the most common method in the
disclosure literature. Without verifying that the data meet some assumptions underlying
the OLS regression, the results may be misleading. These assumptions deal mainly with
the normality, the homogeneity of variance (homoscedasticity) and the collinearity issues. We
examine these assumptions based on the data for 2007, 2008 and 2009.

CHAPTER VII
1.1.1 Normality of residuals
This assumption refers to the fact that the residuals (errors) of a regression should be
normally distributed. Normality of residuals is mainly required for assurance that the Pvalues for t-tests and F-test are valid. The normality assumption is not required in
order to obtain unbiased estimates of the regression coefficients. Besides, there is no
requirement for the explanatory variables to be normally distributed. This assumption can
be examined using the Stata program as follows: After conducting the regression
analysis, we used the predict command to generate residuals (predict r, resid). Then, we
used the kdensity command to create a kernel density plot with the normal option
(kdensity r, normal).
In addition to the graphical test, the normality of residuals can be checked by numerical tests.
One of the available tests on Stata program is the Jacques Bera test which is more suitable
than Shapiro-Wilk test for relatively large observations (N> 88).
To perform the Jacques Bera test we run the sktest command. The sktest provides a test for
normality based on skewness and another based on kurtosis and then combines the two tests
into an overall test statistic. The p-value is based on the assumption that the distribution is
normal.
The following figure provides a kernel density plot.
Figure 1 A Kernel density plot

Normality of residuals (2007)

Normality of residuals (2008)

CHAPTER VII

Normality of residuals (2009)


The previous figures indicate that the cross sectional data for the three years of investigation
exhibits a minor and trivial deviation from normality. The sub tabulated Jacques Bera formal
test failed also to reject the null hypothesis H0 that r is normally distributed. We can accept
then, that the residuals are close to a normal distribution.
Table 8. Sktest for Normality of residuals

r (2007)
r (2008)
r (2009)

Obs.
254
264
270

Pr.(Skewness)
0.2849
0. 0180
0.2661

Pr. (Kurtosis)
0.3424
0.8811
0.0361

Joint
Adj. chi2(2) Prob.>chi2
2.06
0.3568
5.59
0.0610
5.61
0.0606

1.1.2. Homoscedasticity
One of the main assumptions of the OLS regression is the homogeneity of the variance
of residuals. If the variance of the residuals is non-constant, then the residual variance
is said to be heteroskedastic. There are graphical and non-graphical methods for detecting
heteroskedasticity. To examine the heteroskedasticity problem using the Stata program, we
used the hettest command after conducting a multiple regression analysis. The null
hypothesis of the Breusch-Pagan test is that the variance of residuals is homogenous. If
the p-value is small, we reject the null hypothesis and we accept the alternative
hypothesis that the variance is heteroskedastic. The following table presents the results of the
Breusch-Pagan test; the significant chi2 statistic indicates that we should reject the null
hypothesis and accept the alternative hypothesis that the variance is not homogenous.
Table .9 Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

chi2(1)
Prob > chi 2

2007
13.19
0.0003

2008
7.68
0.0056

2009
12.65
0.0004

CHAPTER VII
The use of robust standard errors addresses the problem of heteroskedastic residuals. Robust
standard errors do not change the coefficient estimates, but change the standard errors and the
tests of significance.
1.1.3 Multi-collinearity
The collinearity issue is addressed when there is a perfect linear relationship between two
explanatory variables which cause problems in estimating the regression coefficients. When
more than two variables are involved, it is often called multi-collinearity. The main concern is
that as the degree of multi-collinearity augments, the estimation of the regression coefficients
will be unstable and the standard errors for the coefficients can get seriously inflated.
To examine the multi-collinearity problem, we use the VIF command on STATA after running
the multiple regression analysis. VIF stands for variance inflation factor. As a rule of thumb,
a variable whose VIF value is greater than 10 may need further investigation. Tolerance,
defined as 1/VIF, is employed also in several researches to check on the degree of collinearity.
A tolerance value that is lower than 0.1 is comparable to a VIF of 10. It means that the
variable is considered as a linear combination of other predictors (Hair et al., 2006). The
following table provides the values of VIF as well as the tolerance values which indicate
a limited problem of multi-collinearity.
Table .10 VIF and Tolerance values

Xt
Xt+1
Xt+2
Xt+3
Rt+1
Rt+2
Rt+3
AGt
Ept-1
RD
RD* Xt
RD* Xt+1
RD* Xt+2
RD* Xt+3
RD* Rt+1
RD* Rt+2
RD* Rt+3
RD*AGt
RD* Ept-1

2007
VIF
3.71
2.15
2.76
2.09
4.71
5.31
4.52
1.30
1.57
1.25
3.71
2.70
3.87
2.11
4.60
5.41
4.63
1.31
1.53

1/VIF
0.2699
0.4645
0.3623
0.4788
0.2123
0.1884
0.2213
0.7712
0.6373
0.8019
0.2694
0.3698
0.2582
0.4732
0.2176
0.1848
0.2162
0.7629
0.6517

2008
VIF
2.49
2.09
3.36
2.07
9.57
5.48
5.49
1.34
2.89
1.49
3.01
2.34
3.34
2.42
9.93
5.44
5.27
1.36
1.89

1/VIF
0.4021
0.4792
0.2979
0.4829
0.1044
0.1826
0.1820
0.7439
0.3465
0.6700
0.3321
0.4279
0.2990
0.4136
0.1007
0.1838
0.1898
0.7330
0.5277

2009
VIF
3.71
2.85
3.33
2.03
5.15
5.65
2.74
1.19
1.47
1.17
4.17
2.96
3.98
2.55
5.65
5.59
2.85
1.17
1.39

1/VIF
0.2693
0.3512
0.3002
0.4915
0.1943
0.1770
0.3644
0.8404
0.6802
0.8576
0.2399
0.3379
0.2511
0.3922
0.1770
0.1788
0.3507
0.8537
0.7212

CHAPTER VII
Mean VIF

3.12

3.75

3.14

Table 10 provides the variance inflation factor (VIF) as a measure of multicollinearity between predictors. The
number of observations is 809. The earnings per share measure is a Reuters fundamentals item, calculated by
dividing earnings for ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as
earnings change deflated by price. Both current and future earnings changes are deflated by the price at the start
of the return window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of
dividends) over a 12-month period, starting six months after the end of the previous financial year. EPt1 is
defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is the
growth rate of the total book value of assets for period t. RD is the natural logarithm of the total number of risks
related sentences, respectively for Operations risk, Empowerment risk, Information processing and technology
risk, Integrity risk and Strategic risk information.

1.2. Panel regression specification tests

Panel data analysis is seldom employed in the accounting literature, despite the number of
advantages it has over the traditional cross-sectional (CS) or time series (TS) data analysis.
The first advantage of panel data regression is that it takes the time effect into account
that is not detectable in pure cross-section (CS) data.
According to Cormier et al. (2005) if an OLS regression for one period provides a
picture, panel data provide a sequence of pictures. It gives the researcher a larger number of
observations, increasing thus the degrees of freedom and reducing any collinearity problems
among the explanatory variables. This is likely to improve the estimation efficiency and
provides more reliable and stable parameter estimates (Baltagi, 2005). The use of panel data
reduces eventually the impact of any omitted-variable problem that might arise and control for
unobservable individual heterogeneity and dynamics which is not possible in TS (N=1) and
CS (T=1) (Hsiao, 2013).
Unlike pooled cross-section data an important attribute of panel data is that we cannot
assume that the observations are independently distributed across time and accordingly error
terms are likely to be correlated for a particular individual across different time periods. This
correlation between the error terms is a violation of one of the main assumptions of the OLS
method suggesting that OLS is no longer the best estimator.
Before running the regression analysis based on panel data, several specification tests and
procedures should be performed in order to insure that the regression model specification fits
the data. Three main concerns should be addressed: the question of whether to pool the data or
not, the tests for individual and time effects and the heteroskedasticity of error terms.
1.2.1. Pooling test
A pooling test, also known as a test for heterogeneity, seeks whether or not the intercepts take
a common value . An important advantage of panel data models, compared to cross-sectional

CHAPTER VII
regression models, is that we can allow for heterogeneity among individuals, generally
through individual-specific parameters. Accordingly, an important first procedure is to justify
the need for individual-specific effects. The null hypothesis of homogeneity can be expressed
as H0: 1= 2= 3= t= .
Beck (2001) proposes to use the Chow test to compare the pooled and unpooled estimates
under the assumption that the error term uit N (0, 2). To make sure that the
underlying assumption of the Chow test is correct, we run in a first step a
Pantest2 on Stata which provides the Jacques Bera Test for normality of
residuals. The Skewness/Kurtosis test distributed as a chi2 with 2 degrees
of freedom returns a value of 4.63 and a joint-prob > chi2=0.0986 failing
to reject thus the null of normality at the 5 % level and assuming then that
error terms are normally distributed.
The next step is to perform the Chow test statistic which follows an F
distribution with (N- 1, NT-N-K-1) degrees of freedom. In Stata, this statistic
is generated automatically after running a fixed effect regression
(xtreg, fe). In our case, the Chow test returns an F (321,468) =1.22
and a prob> F= 0.0241 rejecting accordingly the null hypothesis of
homogeneity among individuals at the level of 5%.
Baltagi (2005) recommends to rely on the Roy-Zellner test as a more
appropriate test for poolability because it was evidenced that the Chow
test lacks for robustness under non-spherical disturbances. To run a RoyZellner test in Stata, we follow a procedure in three steps. First, we
interact the regressors with a set of individual dummies to obtain the
unpooled data. Second, we run a random effect regression on the
unpooled data. Finally, we test for the equality of the individual
coefficients. Stata uses a Wald test distributed as a 2 with 72 degrees of
freedom, which returns a value of 242.97 rejecting the null of poolability
across individuals at the level of 1%.
1.2.2. Tests for individual and time effects
The Chow and the Roy-Zellner tests confirmed the presence of specific
effects in our cross sectional time series data. The OLS estimator is no

CHAPTER VII
longer best unbiased linear estimator. The panel data models provide
commonly two ways to deal with these problems: fixed effects and random
effects models.
The main difference between fixed and random effect models resides in
the role of dummy variables. A parameter estimate of a dummy variable is
a part of the intercept in a fixed effect model and an error component in a
random effect model. Slope coefficients remain the same across subjects
or time period in either fixed or random effect model.
To decide which model provides accurate inference from our panel data,
we run a Hausman specification test which compares a random effect
model to its fixed counterpart. If the null hypothesis that the individual
effects are uncorrelated with the other predictors is not rejected, a random
effect model is favored over its fixed counterpart. The test distributed as a
Chi2 with 19 degrees of freedom yields a value of 263.72 and a Prob>
Chi2=0.000 rejecting hence the null hypothesis of absence of correlation
between regressors and individual effects at the 1 % level. For instance,
random effects estimation is inconsistent and the fixed effects model is
more appropriate for our panel data.
As the Hausman test indicates that we should use a fixed effects model it
is important to see if time fixed effects are needed when running a fixed
effects regression. For this purpose, we use the Stata command testparm
after running a fixed effect regression with year dummies. The test
produced an F statistic with a value of 71.78 and a Prob > F = 0.000. We
then reject the null hypothesis that all years coefficients are jointly equal
to zero at the level of 1% and therefore time fixed effects are needed.
1.2.3. Residual diagnostics
The fixed-effects regression model estimated by xtreg, fe invokes the least
squares estimator for point and interval estimates under the classical
assumptions that the error terms are homoscedastic with the same
variance across time and individuals. However, this may be a restrictive
assumption for cross sectional time series data where the cross-sectional

CHAPTER VII
units may be of varying size and as a result may exhibit different variation
(Baltagi, 2005). Baltagi (2005) argues that assuming homoscedastic
disturbances when heteroskedasticity is present will affects the efficiency
of the regression coefficient though these estimates remain unbiased. The
standard errors of these estimates will be biased and one should compute
robust standard errors, correcting for the possible presence of
heteroskedasticity.
One method for detecting heteroskedasticity in Stata is the xttest3
command. This test calculates a modified Wald test for groupwise
heteroskedasticity in the residuals for a fixed-effects regression model.
The null hypothesis specifies that i2 =2 for i =1...Ng, where Ng is the
number of cross-sectional units. The modified Wald test distributed as a
Chi2 statistic with 322 degrees of freedom produced a value of 43.513 and
a prob.> Chi2=0.000 rejecting hence the null hypothesis of homoscedastic
disturbances. Consistent with Baltagi (2005) we handle heteroskedasticity
problem by computing standard errors that are robust to the
homoscedasticity specification.
Another issue in the cross sectional time series data is the
contemporaneous and serial correlation. According to Baltagi (2005),
cross-sectional dependence (contemporaneous correlation) is a problem in
macro panels with long time series (over 20-30 years). This is not much of
a problem in micro panels (a few years and a large number of cases) like
ours. It is argued conversely, that a model with individual effects has
composite errors that are serially correlated by definition. The presence of
the time-invariant error component gives rise to serial correlation which
does not die out over time (Hsiao, 2013). It is noteworthy that ignoring
such correlation when it is present, results in consistent, but inefficient
estimates of the regression coefficients and in biased standard errors.
Stata provides a test of serial correlation for a fixed effect model. The
xtserial., fe command performs a Wooldridge test for autocorrelation in

CHAPTER VII
panel data. The test rejects the null hypothesis of no first order
autocorrelation at the 1% level (F (1,196) = 65.36 and Prob> F=0.000).
To sum off, in order to correct for heteroskedasticity and a first order (AR1)
serial correlation in error terms, individual and time fixed effects
regression is modeled with the cluster option which would provide robust
estimates of the regression parameters according to Hoechle (2007).

2. Findings discussion and analysis


The following tables provide the results of the cross-sections 2007 to 2009
regression estimates as well as the cross sectional time series regression
with observations from all three years. These results highlight the impact
of voluntary risk disclosure on share price anticipation of future earnings.
The regression results in table 11 show that, unlike what was predicted
earlier, the coefficients on Xt are, though positive, insignificant for the
three regressions with each control variable. This provides evidence that
current earnings are not a good proxy for markets prior expectation about
future earnings in 2007. There is also strong evidence of share price
anticipation of future earnings for one period ahead. This is consistent with
accounting recognition lags stock returns in measuring value creation. This
phenomenon applies to the three regressions and is independent of the
amount of voluntary risk information in annual report narratives. All the
three coefficients on Xt+1 are positive and two of these coefficients are
significant at the level of 5%. The significant estimates for this coefficient
range from 1.175 to 1.415. There is however no evidence that share prices
anticipate future earnings changes with more than one period ahead. The
coefficients on Xt+2 and Xt+3 are insignificant in all cases.

CHAPTER VII
Table 11 shows that the coefficients on AGt, though significant, have the
wrong sign. This suggests that AGt might not be a satisfactory
measurement error proxy. Findings show also that the coefficients on Ept-1
are negative but insignificant. The negative sign on Ept-1 suggests that
earnings follow a random walk. The coefficients on the future stock return
variables (Rt+1, Rt+2 and Rt+3) are expected to be negative. The negative
coefficients on future stock returns should confirm that realized future
earnings contain a measurement error that future returns remove (Collins
et al., 1994). This is the case for two cross sectional regressions mainly
when controlling for financial leverage level and corporate size. Despite
the right signs in general, all coefficients are largely insignificant.
The coefficients of primary interest in the current study are the coefficients
on RD*Xt+1, RD*Xt+2, and RD*Xt+3. We predicted positive coefficients on
these interacted variables. Unfortunately, the coefficients on the
interaction term RD* Xt+1 is negative (with one exception) and insignificant
at conventional levels in the three regression models (for each control
variable apart).
At the same time, the coefficient on RD* Xt+2 is positive and significant at
the 10% level when controlling for financial leverage. In the same way, the
coefficients on RD* Xt+3 are positive and significant when controlling for
financial leverage as well as corporate size at respectively 5% and 10%
levels. There is accordingly evidence that the amount of risk disclosure
influences positively share prices ability to anticipate future earnings
changes for two years and three years ahead. Yet, no such significant
effect has been noticed when controlling for corporate profitability.
Table.11: 2007 Cross Section Multiple Regression

Independe

Expect

Financial

Firm

nt Variable ed Sign
Leverage
Intercept
(?)
-0.173

Size

Xt

(+)

(0.267)
1.690

Xt+1

(+)

(0.111)
1.415**

Profitability

-0.265

-0.226

(0.114)
1.580

(0.167)
1.496

(0.159)
1.340**

(0.192)
1.175*

CHAPTER VII
Xt+2

(+)

(0.038)
0.240

Xt+3

(+)

(0.739)
-1.238

(0.411)
-0.730

(0.212)
-0.078

Rt+1

()

(0.165)
-0.125

(0.405)
-0.153

(0.932)
-0.068

()

(0.410)
-0.022

(0.300)
-0.047

(0.686)
0.023

Rt+3

()

(0.833)
0.114

(0.638)
0.109

(0.849)
0.032

AGt

()

(0.400)
0.539***

(0.429)
0.595***

(0.834)
0.537***

Ept-1

(+)

(0.000)
-0.0076

(0.000)
-0.009

(0.000)
-0.012

(?)

(0.781)
0.040

(0.739)
-0.002

(0.665)
0.036

RD*Xt

(?)

(0.140)
0.066

(0.926)
0.075

(0.200)
-0.018

RD*Xt+1

(+)

(0.730)
-0.084

(0.692)
-0.037

(0.921)
0.036

RD*Xt+2

(+)

(0.322)
0.249*

(0.496)
0.163

(0.770)
0.188

(+)

(0.066)
0.380**

(0.136)
0.170*

(0.303)
0.005

RD*Rt+1

(?)

(0.021)
-0.026

(0.080)
-0.053*

(0.969)
0.000

RD*Rt+2

(?)

(0.449)
-0.037

(0.087)
-0.018

(0.983)
-0.021

RD*Rt+3

(?)

(0.187)
-0.023

(0.422)
-0.038

(0.436)
-0.060*

(?)

(0.459)
-0.005***

(0.161)
-0.004**

(0.090)
-0.001

(?)

(0.001)
-0.001

(0.032)
0.000

(0.696)
-0.003**

(0.110)
Yes

(0.938)
Yes

(0.015)
Yes

Dummies
Country

Yes

Yes

Yes

Dummies
Adj R2

38.1%

36.5%

37%

Rt+2

RD

RD*Xt+3

RD*AGt
RD*Ept-1
Industry

(0.044)
0.576

(0.069)
0.919

CHAPTER VII
Obs

254

254

254

Table 11 presents OLS regression estimates for 2007 cross section data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period return, Rt. Rt,
Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month
period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2 and Xt+3 are
defined as earnings change deflated by price. Both current and future earnings changes are
deflated by the price at the start of the return window for period t. EPt1 is defined as period t1s
earnings over price six months after the financial year-end of period t1. AGt is the growth rate of
the total book value of assets for period t. RD is the natural logarithm of the total number of risks
related sentences, respectively, for Operations risk, Empowerment risk, Information processing and
technology risk, Integrity risk and Strategic risk information. Industry and Country dummies are
used to control for cross sectional differences among firms. Financial leverage, firm size and
profitability are controlling variables (their regression estimates are not tabulated). The significance
levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

Table 12 provides regression findings for 2008. Unlike 2007 estimates, the
coefficient on Xt is as predicted significantly positive at the 10% level for
the regression when we control for financial leverage variable. This
provides weak evidence that current earnings reflect prior anticipation of
future earnings. There is also strong evidence of prices leading earnings by
one period, regardless of the extent of corporate risk disclosure. The future
earnings response coefficients on Xt+1 extend from 2.327 to 2.919 and are
considerably higher than the current earnings coefficients. The coefficients
on Xt+2, while negative, are not significant at conventional levels in all
cases. It is worth noting also that current returns appear to be significantly
and positively associated with future earnings changes for t+3 period,
suggesting that the market is likely to impound future earnings news into
current returns more than one year ahead when they are informed about
corporate profitability.
Inconsistent with 2007 findings, the coefficients on future stock returns
(Rt+1, Rt+2 and Rt+3) are mixed though they are significant for t+1 period. In
addition, the coefficients on AGt have the wrong sign and are insignificant.
Moreover, table 12 shows that the coefficients on Ept-1 are, as expected,
positive and significantly different from zero at the 10 % level in only two
cases. Additionally, unlike 2007 estimates, finding revealed a positive and
significant association between current stock returns and the interacted
term of current earnings with the risk disclosure level when controlling for
corporate profitability. This provides low evidence that information about
risk and uncertainties increase the credibility of current earnings when
investors are informed about the returns on their equities.

CHAPTER VII
The 2008 findings are likewise different from those of 2007 regarding the
interacted term of risk disclosure with future earnings. While positive
coefficients on RD* Xt+1, RD* Xt+2, and RD*Xt+3 are expected, results
indicate first negative and significant coefficients on the interacted term
RD* Xt+1 for the three regressions at varying acceptance levels. This
suggests that a higher level of voluntary risk disclosure reduces the
association between current returns and future earnings for t+1 period.
Still, because RD*Xt+i and RD*Rt+i together proxy for the revealed future
earnings, and because the individual years of future earnings are highly
correlated, a more powerful test examines the joint significance of RD*X t+1
and RD*Rt+1. The partial F-tests of the joint significance of these variables
have p-values of 0.0628 and 0.6298 which rejects the basic idea that risk
disclosures published in the annual report reveal information about future
earnings for one year ahead depending on firm size and profitability.
It is noteworthy, although that this partial F-test is significant when
controlling for risk as measured by financial leverage level with a p value
of 0.0229. Nevertheless, there is a strong and positive effect of corporate
risk disclosure level on share price informativeness with respect to future
earnings for two years ahead for the three main regressions. Accordingly,
it seems that investors take more time to assess the amount of risk
information disclosed in 2008 and available to the financial market at the
start of the following year. They are not capable to incorporate such
information in the firm value estimates in t+1 period. As a result the
benefits mainly accrue in the following t+2s financial year.
Table.12 2008 Cross Section Multiple Regression

Independ

Expect

Financial

Firm

Profitability

ent

ed Sign

Leverage

Size

Variable
Intercept

(?)

-0.213**

-0.209*

-0.211*

Xt

(+)

(0.044)
1.097*

(0.085)
1.012

(0.056)
0.848

Xt+1

(+)

(0.093)
2.327***

(0.114)
2.472***

(0.243)
2.919***

CHAPTER VII
Xt+2

(+)

(0.000)
-1.682

Xt+3

(+)

(0.132)
1.097

(0.420)
1.048

(0.173)
1.488*

Rt+1

()

(0.128)
-0.276***

(0.127)
-0.284***

(0.076)
-0.272***

()

(0.002)
0.177

(0.002)
0.187

(0.001)
0.188

Rt+3

()

(0.131)
0.217*

(0.123)
0.171

(0.121)
0.141

AGt

()

(0.075)
0.008

(0.210)
0.015

(0.277)
0.000

Ept-1

(+)

(0.932)
0.032*

(0.871)
0.035*

(0.993)
0.030

(?)

(0.072)
0.002

(0.059)
0.011

(0.102)
0.009

RD*Xt

(?)

(0.894)
0.092

(0.621)
0.088

(0.656)
0.253**

RD*Xt+1

(+)

(0.170)
-0.106***

(0.341)
-0.154**

(0.030)
-0.104

RD*Xt+2

(+)

(0.006)
0.509***

(0.021)
0.455*

(0.451)
0.515**

(+)

(0.004)
-0.156*

(0.071)
0.011

(0.018)
-0.161

RD*Rt+1

(?)

(0.067)
-0.003

(0.849)
0.015

(0.440)
-0.017

RD*Rt+2

(?)

(0.874)
-0.050**

(0.567)
-0.007

(0.459)
-0.059**

RD*Rt+3

(?)

(0.038)
-0.026

(0.855)
0.021

(?)

(0.259)
-0.000

431)
-0.001

(0.725)
-0.000

(?)

(0.941)
-0.000

(0.452)
-0.000

(0.756)
-0.000

(0.887)
Yes

(0.884)
Yes

(0.682)
Yes

Dummies
Country

Yes

Yes

Yes

Dummies
Adj R2

48.52%

48.01%

47.60%

Rt+2

RD

RD*Xt+3

RD*AGt
RD*Ept-1
Industry

(0.000)
-0.931

(0.000)
-1.596

(0.

(0.048)
-0.013

CHAPTER VII
Obs

264

264

264

Table 12 presents OLS regression estimates for 2008 cross section data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period return, Rt.
Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12month period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2 and
Xt+3 are defined as earnings change deflated by price. Both current and future earnings changes
are deflated by the price at the start of the return window for period t. EPt1 is defined as period t
1s earnings over price six months after the financial year-end of period t1. Act is the growth rate
of the total book value of assets for period t. RD is the natural logarithm of the total number of risks
related sentences, respectively, for Operations risk, Empowerment risk, Information processing and
technology risk, Integrity risk and Strategic risk information. Industry and Country dummies are
used to control for cross sectional differences among firms. Financial leverage, firm size and
profitability are controlling variables (their regression estimates are not tabulated). The significance
levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

Table 13 emphasizes cross sectional regression estimates for 2009. The


results are mostly the same as in 2007. Unlike Collins et al. (1994), the
coefficients on Xt are while positive, insignificant for the three expanded
regression models. This suggests that investors do not consider current
earnings as a good prior estimation for future earnings. We find also that
current returns contain information about future earnings for one period
ahead. In fact, all future earnings response coefficients for t+1 period are
positive and significant at the level of 5% and 10% consistent with earlier
evidence that prices lead future earnings changes irrespective of firm risk
reporting activity. The significant estimates for this coefficient range from
1.546 to 2.039. Nevertheless, there is no proof that prices lead future
earnings changes by more than one year ahead. The coefficients on Xt+2
and Xt+3 are negative and insignificant in all cases except for the expanded
model with firm size as a control variable. This result contradicts the
findings in Collins et al. (1994) and suggests that current returns in MENA
emerging markets are less informative about future earnings. Table 13
shows besides that the sign of the coefficients on Ept-1 is mostly as
expected positive despite the insignificant association with current returns.
This is not the case for the coefficients on AGt which are significant for two
regression models but have the wrong sign. This result questions once
again the role of AGt as a suitable measurement error proxy. Furthermore,
inconsistent with 2007 and 2008 findings, the coefficients on future stock
returns (Rt+1, Rt+2 and Rt+3) have the wrong sign (a negative sign is
predicted) and are largely insignificant.

CHAPTER VII
Our variables of interest exhibit approximately the same trend as in the
2007 cross sectional regression. First, the coefficients on the interacted
term RD* Xt+1 is negative and insignificant at conventional levels for the
three expanded regression models (for each control variable apart). In the
same way, the coefficient on RD* Xt+2 is positive and significant at the 10
% level when controlling for financial leverage. Besides the coefficients on
RD* Xt+3 are positive and significant in all cases at respectively 10% and
1% levels. For instance, results demonstrate that corporate risk reporting
practice influences positively share prices ability to anticipate future
earnings changes for two years and three years ahead.

Table.13 2009 Cross Section Multiple Regression

Model

Expect

Financial

Intercept

ed Sign Leverage
(?)
-0.189

Firm

Profitability

Size
-0.247

-0.253

Xt

(+)

(0.315)
-0.839

(0.179)
-0.765

(0.164)
-0.979

Xt+1

(+)

(0.215)
2.039**

(0.244)
1.546*

(0.169)
1.952**

(+)

(0.027)
-1.183

(0.093)
-0.565

(0.048)
-0.734

Xt+3

(+)

(0.177)
-1.028

(0.484)
-1.550*

(0.410)
-1.286

Rt+1

()

(0.314)
-0.078

(0.098)
0.017

(0.201)
0.020

Rt+2

()

(0.718)
0.269

(0.932)
0.240

(0.925)
0.213

()

(0.239)
0.111

(0.235)
0.175

(0.326)
0.176

()

(0.505)
0.384*

(0.246)
0.329

(0.287)
0.394*

Xt+2

Rt+3
AGt

CHAPTER VII
Ept-1

(+)

(0.079)
0.020

RD

(?)

(0.453)
0.040

(0.712)
-0.006

(0.855)
0.027

RD*Xt

(?)

(0.263)
0.202

(0.857)
0.021

(0.451)
0.181

(+)

(0.104)
-0.171

(0.877)
-0.002

(0.269)
-0.036

RD*Xt+2

(+)

(0.217)
0.241*

(0.986)
0.036

(0.832)
0.226

RD*Xt+3

(+)

(0.053)
0.176*

(0.778)
0.580***

(0.240)
0.538***

RD*Rt+1

(?)

(0.085)
0.061

(0.001)
0.081

(0.000)
0.076

(?)

(0.207)
-0.041

(0.192)
-0.032

(0.263)
-0.017

RD*Rt+3

(?)

(0.278)
-0.006

(0.464)
-0.002

(0.718)
-0.016

RD*AGt

(?)

(0.830)
0.005

(0.954)
0.004

(0.584)
-0.001

(?)

(0.275)
-0.000

(0.316)

RD*Ept-1

-0.000

(0.804)
-0.000

(0.190)
Yes

(0.910)
Yes

(0.967)
Yes

Dummies
Country

Yes

Yes

Yes

Dummies
Adj R2
Observati

40.74%
270

43.53%
270

41.74%
270

RD*Xt+1

RD*Rt+2

Industry

(0.133)
-0.010

(0.065)
0.005

ons
Table 13 presents OLS regression estimates for 2009 cross section data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period return, Rt. Rt,
Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month
period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2 and Xt+3 are
defined as earnings change deflated by price. Both current and future earnings changes are
deflated by the price at the start of the return window for period t. EPt1 is defined as period t1s
earnings over price six months after the financial year-end of period t1. Act is the growth rate of
the total book value of assets for period t. RD is the natural logarithm of the total number of risks
related sentences, respectively, for Operations risk, Empowerment risk, Information processing and
technology risk, Integrity risk and Strategic risk information. Industry and Country dummies are
used to control for cross sectional differences among firms. Financial leverage, firm size and
profitability are controlling variables (their regression estimates are not tabulated). The significance
levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

CHAPTER VII
Overall, main estimates of the cross sectional analysis confirm our first
hypothesis H1 in that they indicate that current returns are increasing in
future earnings as voluntary risk disclosure increases. It seems then that
voluntary risk information in the annual reports contains value-relevant
information about future performance that investors impound into stock
prices. This disclosure enables investors to correct for corporate risk and
to make more accurate forecasts when they value their investments. We
suggest accordingly that risk information decreases investors uncertainty
and increase market expectations about firm prospects by reflecting more
future earnings information in the current return.
Panel regression estimates corroborate to some extent the year by year
regressions. We notice a positive and significant association between stock
returns and future earnings of period t+1 at the 1% level. There is also
some evidence on the ability of stock returns to anticipate future earnings
changes for three years ahead. The coefficients on Xt+3 extend from 1.054
to 1.395 and are significantly different from zero at 5% and 10% levels
respectively when controlling for corporate size and profitability. The
coefficients on future stock returns exhibit the right sign and are
significantly different from zero for t+1 as well as t+2 period.
Regarding the second part of the regression models, we observe first a
strong and significant impact of risk disclosure on the return current
earnings association at the 1 % level. This suggests that voluntary risk
disclosure makes financial data more credible in that a price reaction to
the unanticipated portion of current earnings is documented mainly when
the market is informed about firm profitability. Finally, the voluntary risk
information seems to bring the future forward in that it impacts
positively share price informativeness with respect to future earnings for
t+2 period, particularly when information about the level of financial
leverage and corporate profitability are available to the market.
Table.14 Panel Multiple Regression

Model

Expect

Financial

Firm

ed Sign

Leverage

Size

Profitability

CHAPTER VII
Interce

(?)

-0.107

0.073

-0.018

pt
Xt

(+)

(0.257)
0.370

(0.607)
0.257

(0.859)
0.430

(+)

(0.546)
2.961***

(0.672)
2.634***

(0.488)
2.516***

Xt+2

(+)

(0.000)
0.695

(0.000)
1.005

(0.000)
0.833

Xt+3

(+)

(0.339)
1.054

(0.156)
1.550**

(0.238)
1.395*

Rt+1

()

(0.166)
-0.579***

(0.038)
-0.607***

(0.076)
-0.551***

()

(0.000)
-0.319**

(0.000)
-0.313**

(0.000)
-0.264*

Rt+3

()

(0.020)
-0.191

(0.019)
-0.187

(0.071)
-0.219

AGt

()

(0.168)
0.194

(0.180)
0.168

(0.121)
0.160

Ept-1

(+)

(0.134)
-0.018

(0.179)
-0.009

(0.200)
-0.011

(0.823)
0.061**

(0.774)

(?)

(0.630)
0.041

(?)

(0.155)
0.114

(0.037)
-0.011

(0.412)
0.271***

RD*Xt+1

(+)

(0.243)
-0.129

(0.914)
-0.163

(0.005)
0.118

RD*Xt+2

(+)

(0.178)
0.191*

(0.199)
0.091

(0.352)
0.302**

RD*Xt+3

(+)

(0.064)
0.035

(0.524)
0.054

(0.014)
0.104

(?)

(0.746)
-0.020

(0.713)
-0.003

(0.468)
0.023

RD*Rt+2

(?)

(0.423)
-0.064*

(0.904)
-0.030

(0.427)
-0.047

RD*Rt+3

(?)

(0.076)
-0.031

(0.468)
-0.023

(0.291)
-0.020

RD*AGt

(?)

(0.324)
0.002

(0.465)
0.002

(0.583)
0.003

(?)

(0.666)
-0.000

(0.587)
0.000

(0.469)
0.002

Xt+1

Rt+2

RD
RD*Xt

RD*Rt+1

RD*Ept-

0.024

CHAPTER VII
1

Year

(0.445)
Yes

(0.741)
Yes

(0.114)
Yes

39.31%
809

40.14%
809

39.96%
809

Dummi
es
Adj. R2
Obs.

Table 14 presents fixed effects regression results for panel data. Heteroscedasticity-consistent pvalues are given in parentheses. The dependent variable is current period return, Rt. Rt, Rt+1, Rt+2 and
Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting
six months after the end of the previous financial year. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings
change deflated by price. Both current and future earnings changes are deflated by the price at the
start of the return window for period t. EPt1 is defined as period t1s earnings over price six
months after the financial year-end of period t1. Act is the growth rate of the total book value of
assets for period t. RD is the natural logarithm of the total number of risks related sentences,
respectively, for Operations risk, Empowerment risk, Information processing and technology risk,
Integrity risk and Strategic risk information. Financial leverage, firm size and profitability are
controlling variables (their regression estimates are not tabulated). The significance levels (two-tail
test) are: *= 10 %, ** =5 % and *** = 1 %.

In summary, cross-sectional as well as panel regression analyses provide


evidence on the informational content of the voluntary release of risk
information in MENA emerging markets. Through the disclosure of
information about opportunities and/or potential threats to the companys
business, managers seem to give investors some idea of the level of value
creation to expect within the business. In this respect, risk disclosures
appear to unravel corporate uncertainties and introduce new information
about risk factors in the financial market, which is likely to narrow the
range of users predictions of future prospects. By revealing risk
information, managers tend to reduce the information asymmetry in the
market and to assist the user in gauging the future profitability of a
company. At the same time, investors seem to react positively to this
signal and are likely to incorporate relevant information about company
risks and prospects into corporate stock prices.
The empirical results in this chapter support to a large extent our
suggested economic based framework (the agency theory and the
signaling theory) for voluntary risk information. They provide clear
evidence that narrative risk disclosure serves as a monitoring and a
signaling mechanism and have quantitative effects on share prices. These
results reflect the role of voluntary risk disclosure in reducing the
information asymmetry about firm future performance between insiders

CHAPTER VII
and outsiders. This is consistent with the explanation that in a lowregulation environment, managers voluntarily report on risk and
uncertainties to reduce the agency problems and to allow investors to
make better-informed decisions. Our findings are also consistent with the
belief that discretionary risk disclosure sends a good signal to investors
about managers ability to disclose relevant information that reflects more
accurately the companys risk profile. Our results highlight then a trend
toward improving corporate disclosure activity and ensuring the well
functioning of financial markets within the scrutinized context.
Firmstransparency about their risk profile is somehow the outcome
of the recent corporate governance reforms in the region.
Regulators and professional agencies targeted to enhance monitoring
of management, to gain the trust of foreign investors and to
enforce their protection by encouraging the adoption of international
recommended practices. Investors financial literacy seems to be, on
the other side, in progress as they are aware about the
informational needs despite the unique cultural context in the region.
Their accurate understanding of business risks, improves their ability
to predict future earnings. The understanding of business risks by
investors and other users of corporate reporting improves the ability of
returns to predict future earnings. In related work, Gelb and Zarowin
(2002), Lundholm and Myers (2002), Banghoj and Plenborg (2008) and
Hussainey and Walker (2009) show that FERCs increase with the
informativeness of firm disclosures. While these studies address the
informativeness of either disclosure in general or forward looking
information in particular, to our knowledge, no studies have demonstrated
a relation between risk information and FERCs. Our results provide hence
evidence to the recent debate on the informativeness of corporate risk
disclosure. They complement the indirect evidence of a number of recent
studies (e.g. Bao and Datta, 2014; Campbell et al, 2014; Kravet and Muslu,
2013) which are based on statistical correlations between risk disclosures
and changes in share price volatility, trading volumes, bid-ask spreads, or
analysts forecasts. While many of them have been a response to new

CHAPTER VII
disclosure requirements, our results provide a direct evidence that
voluntary risk information is being used by investors in forecasting future
earnings.
3. Robustness checks: controlling for potential endogeneity problems
Accounting literature addressed an important question of whether
corporate disclosure has a first order effect on some outcome
variables (e.g. Firm value, cost of capital, share price
informativeness, etc.). It follows that the association between capital
market variables and corporate reporting may be driven by firm
performance rather than information per se. It is also possible that
disclosure policy changes are not random events: they may coincide
with changes in firm economics and governance characteristics (Healy
and Palepu, 2001).
According to Beyer et al. (2010) given the endogenous nature of the
corporate information environment and some observed outcomes, it is
hard to make an assessment of the causal connection and recognize
the exact impact that one mechanism would have on another one.
As a sensitivity test, we acknowledge in this section two potential
sources of endogeneity between share price informativeness and risk
disclosure. Intuitively, our reasoning is that differences exist in the
share price informativeness that are correlated with the firms risk
disclosure level, but that are not necessarily caused by this activity.
Instead, these differences are caused either by: (1) unobservable
heterogeneity among firms in a cross-sectional sample; or (2)
observable determinants of the return earnings relationship which are
correlated with disclosure but omitted from the analysis. All of these
problems imply a correlation between the error term and one or
some of the independent variables (e.g. risk disclosure and the
interaction terms) in the regression model. Controlling for timeinvariant, unobserved heterogeneity by adding firm fixed effects and
some appropriate control variables can help to mitigate this concern.

CHAPTER VII
However, this method implies a strong exogeneity hypothesis. It
assumes that current observations of the explanatory variable (e.g.
voluntary risk disclosure) are independent of past values of the
dependent variable (e.g. stock returns). As firms choose their level
of risk disclosure based on their perceived cost-benefit analysis, this
assumption is not likely to hold in the specific context of voluntary
disclosure.
Following Larcker and Rusticus (2010), we use an instrumental
regression as an alternative approach to deal with the above
mentioned concerns. These authors consider this method as the
most promising econometric approach to addressing the endogeneity
problem in social sciences. Instrumental variables are used to cut
correlations between the error term and independent variables. When
more instrumental variables are available, two stage least squares
regression (2SLS) provides a way of obtaining the optimal linear
combination of instruments (Wooldridge, 2010).
Given that previous studies documented that the extent and the
quality of corporate disclosure are associated with ownership
structure, size, leverage, profitability and industry sector, we
estimate the relationship between the level of voluntary risk
disclosure and these firms features in the first stage of the
regression. The endogenous dependent variable (the level of
voluntary risk disclosure) is as such explained by these instruments
and the rest of exogenous variables of our model are also included
in this first equation.

In the first stage we estimate:


M(1):

CHAPTER VII
RDit b0 b1OwCit b2 MOw it b3 INsOw it b4 Size it b5 Levit b6 Pr ofit
3

k 1

k 1

b7 Secti b8 X it bk 8 X it k bk 11Rit k b15 EPt 1 b16 AGit eit


In the second stage, we estimate the following regressions using the fitted
values (estimates) of risk disclosure (RDit) derived from the first stage:
M(2):
3

Rit b0 b1 X it bk 1 X it k bk 4 Rit k b8 EPit 1 b9 AGit b10 Fit ( RD it ) b11[ Fit ( RD it ) * X it ]


k 1

k 1

k 1

k 1

bk 11[ Fit ( RD it) * X it k ] bk 14 [ Fit ( RDit ) * Rit k ] b18 [ Fit ( RDit ) * EPit1 ] b19 [ Fit ( RDit) * AGit ] eit
Where: Rit : stock return for year t is the buy-and-hold returns from six months before the financial year-end
to six months after the financial year-end. Xit : earnings change per share deflated by the share price at the start of
the return window for period t. Xit+1 , Xit+2 , Xit+3: the earnings change per share for year t+1, t+2, t+3 respectively
deflated by the price at the start of the return window for period t. Rit+1, Rit+2, Rit+3: stock return for year t+1, t+2,
t+3 are the buy-and-hold returns for the 12-months period. AGit: The growth rate of total book value of assets for
period t. EPt-1: the period t1s earnings over price six months after the financial year-end of period t1. RDit: is
the natural logarithm of the number of risk-related sentences. Levit : Firm risk is the book value of equity scaled
by total liabilities. Sizeit: Firm size is the natural logarithm of corporate revenues. Profit : Firm profitability is the
natural logarithm of Return on Equity (ROE). Secti: Industry sector 1 if the firm is classified into one
of the nine commonly defined industry sectors and 0 otherwise. OwCit : the proportion of
equity held by the top-five largest shareholders, including financial institutions, firms
inside shareholders (directors and executives) and other outside block shareholders.
MOwit : the proportion of equity held by managers and executive directors. InsOwit : the
total number of shares held by institutions to the total number of shares outstanding per
sample firm.

According to Larcker and Rusticus (2010), it is common for accounting


researchers to report the classic Hausman (1978) test for endogeneity. As
we are dealing with panel data, we rely on the endog test (known as
difference-in-Sargan-Hansen" statistic) which is robust to various
violations of conditional homoscedasticity. If the probability of the test is
below 0.05, then we reject the hypothesis of the exogeneity of the
instrumented variable (related-risk disclosure, RDit). The endog test distributed as chi
(2) statistic with 10 degrees of freedom reports a value of 19.067 and a probability of 0.0394.
Thus, it is necessary to apply an instrumental variable approach to correct
for endogeneity problem. 2SLS procedure is used in the form of Baltagi's EC2SLS
random-effects estimator12
12 The test of overidentifying restrictions (orthogonality conditions) failed to reject the the
null hypothesis that the excluded instruments are valid instruments, i.e., uncorrelated
with the error term. Accordingly a RE estimation is preferred over a FE regression
(Wooldridge, 2010).

CHAPTER VII
Table 15 reports the main findings for our regression model. The instrumental approach
yielded slightly different results compared to our initial analysis.
The coefficients on Xit and Xit+k become insignificant suggesting that
current stock price is not associated with current and future earnings
change. So there is no evidence of prices leading earnings regardless of
corporate risk disclosure. There is also a weak positive association
between risk disclosure and current stock returns at the 10% level. This
indicates that risk reporting activity for our sample firms is driven by stock
price performance. Finding revealed furthermore a positive and significant
association between current stock returns and the interacted term of
current earnings with the risk disclosure at the level of 5%. This provides
evidence that information about risk and uncertainties increase the
credibility of current earnings. These results indicate that if earnings are not
complemented with risk information, they may not be believed by stock market participants.
Finally, unlike the initial findings, instrumental analysis showed that firm
risk disclosure level is incrementally relevant for anticipating future
earnings growth only for one year ahead. The coefficient on the interaction
term RDit*Xit+1 is positive and significant at the level of 5%. There is
accordingly evidence that corporate risk disclosures are value relevant to
market participants.
Table 15 Instrumental multiple regression analysis

Model
Intercept
Xit
Xit+1
Xit+2
Xit+3
Rit+1
Rit+2
Rit+3
AGit
Epit-1
RDit
RDit*Xit
RDit*Xit+1
RDit*Xit+2
RDit*Xit+3
RDit*Rit+1

Exp. Sign
(?)
(+)
(+)
(+)
(+)
()
()
()
()
(+)
(?)
(?)
(+)
(+)
(+)
(?)

M2
-0.788
-3.796
-6.894
4.294
4.625
0.429
-0.056
-0.450
0.416*
0.009
0.311*
0.904**
1.533**
-0.393
-0.512
-0.249*

(0.132)
(0.108)
(0.111)
(0.349)
(0.393)
(0.470)
(0.946)
(0.556)
(0.093)
(0.875)
(0.062)
(0.018)
(0.019)
(0.613)
(0.582)
(0.094)

CHAPTER VII
RDit*Rit+2
RDit*Rit+3
RDit*AGit
RDit*Epit-1
Year Dummies
Wald chi2
Observations

(?)
(?)
(?)
(?)

-0.072
0.078
0.000
-0.002

(0.726)
(0.656)
(0.991)
(0.458)

Yes
240.81(p-value<0.05)
785

Table 15 presents IV regression results for panel data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is
current period return, Rit. Rit, Rit+1, Rit+2 and Rit+3 are calculated as buy-andhold returns (inclusive of dividends) over a 12-month period, starting six
months after the end of the previous financial year.
Xit, Xit+1, Xit+2 and Xt+3 are defined as earnings change deflated by price. Both
current and future earnings changes are deflated by the price at the start of
the return window for period t. EPit1 is defined as period t1s earnings over
price six months after the financial year-end of period t1. AGit is the growth
rate of the total book value of assets for period t. RDit is the natural
logarithm of the total number of risk related sentences. The significance
levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

Summary
A major contribution of our study is the focus on a new form of voluntary
reporting activity in corporate annual reports that is risk disclosure. This
chapter aimed to examine empirically the economic consequences of this
information in the context of MENA emerging markets following Lundholm
and Myers (2002), Gelb and Zarowin (2002) and Hussainey (2004). The
main hypothesis in this chapter predicts that there is a positive association
between voluntary risk disclosures and share price anticipation of future
earnings. The empirical findings are mostly in line with this assumption.
The results suggest that the ability of the market to anticipate two-year
and three years ahead future earnings changes is positively related to the
level of risk information.
Estimates are generated from year by year and panel regressions whereby
we controlled for the cross-sectional effects of some previously
documented determinants of voluntary disclosure as well as earnings
response coefficients such as industry, risk, firm size and profitability. As a
sensitivity test we re-estimate our regression models based on an
instrumental variables approach. The results confirm to some extent our
initial findings in that risk disclosure seems to be useful to investors in
anticipating future earnings growth for one year ahead. Our findings are of
interest because they suggest that risk information are perceived as value

CHAPTER VII
relevant in that it helps investors anticipate the amount of future net
income changes of one entity.
In choosing a disclosure strategy, prior literature suggests that corporate
governance plays a more significant role in increasing transparency and
encourages informed trading while managers have to trade off the
benefits of expanded disclosure against the costs of disclosing potentially
damaging information. Given the specificity of this form of disclosure, we
investigate in the following chapters how firm-specific governance
mechanisms as well as proprietary costs might interact to affect stock
price informativeness with respect to revealed future earnings.
Accordingly, the next chapter is consecrated to the hypotheses development and to the
research methodology which will be implemented to empirically assess how voluntary risk
disclosure, corporate governance mechanisms and proprietary costs impact the return future
earnings relationship for a sample of MENA emerging markets.

Chapter VII: Corporate governance, proprietary costs and share price


informativeness with respect to future earnings news: hypotheses
development and research methodology

Introduction
The influence of governance mechanisms and proprietary costs on
corporate transparency is well established in the literature (see chapter
IV for a review). The separation of ownership and control leads to agency
conflicts between corporate managers and shareholders (Jensen and
Meckling, 1976). Insiders and controlling shareholders -better informed
about corporate opportunities- have greater incentives to expropriate
wealth from minority investors and to withhold private information from
outsiders (Shleifer and Vishny, 1997). Such opportunistic behavior should
reduce when corporate governance becomes more effective. Insiders may
however have other incentives to withhold sensitive information,
particularly when this information would harm the firms competitive
position. This should impact the information content and the credibility of

CHAPTER VII
firms' disclosure, causing more/less firm-specific risk information to be
included in the share price.
The purpose of this chapter is to set out the hypotheses and to highlight
the research design which will be used to implement our empirical
investigation. The focus of our hypotheses is to examine whether there is any effect of
corporate governance structure and proprietary costs on the association between the extent of
risk disclosures being made within corporate annual reports and the ability of stock returns to
predict future earnings.
1. Hypotheses development: prior researches
The first group of hypotheses is developed to examine corporate
ownership effect on share price informativeness with respect to future
earnings news. The second set of hypotheses examines the influence of board
characteristics on the association between risk disclosure and share price anticipation of future
earnings. The third set of hypothesis investigates the moderating effect of proprietary costs on
the ability of stock returns to forecast revealed future earnings. The rationale underlying the
development of our hypotheses is set out below.
1.1. The effect of corporate governance and voluntary risk disclosure on share price anticipation
of future earnings

Agency theory suggests that conflicts of interest arise between managers and outside
shareholders due to the dissociation between ownership and control (Jensen and Meckling,
1976). Ways to mitigating and preventing these conflicts of interests include control
mechanisms and corporate governance. Both are likely to counter managerial power and to
improve accountability (OSullivan, 2000). With respect to corporate control, Useem (1996)
presents ownership structure and in particular ownership concentration as a main mechanism,
for it has considerably grown in recent decades, largely due to the growth of financial
intermediaries such as pension and mutual funds. Regarding corporate governance, disclosure
activity stems from the board of directors (Gul and Leung, 2004), and corporate annual
reports are prepared by the board, so that its governance structure can be expected to
influence disclosure policy. These two aspects -corporate ownership and board structurebring forth expectations concerning relationships with the level and the value relevance of
disclosure in the annual reports (Abraham and Cox, 2007).

CHAPTER VII
Within the risk disclosure literature, Taylor et al. (2010) assert that firms with strong corporate
governance mechanisms are more effective in managing their financial risk; that is reflected
in enhanced risk management disclosures. Abraham and Cox (2007, p231) argue that
ownership and governance may play a vital role in firms risk reporting because within the
joint stock company, owners may ask for a broad range of potentially relevant risk
information that management would otherwise choose to withhold. Directors (as agent) are
expected to improve managers accountability and hence corporate risk disclosure.
To examine the impact of governance mechanisms and narrative risk disclosure in annual
reports on share price informativeness with respect to future earnings growth, we focus on the
following corporate control and boards characteristics.
1.1.1. Corporate ownership structure
The relationship between corporate disclosure and ownership structure is mainly analyzed
according to ownership diffusion/concentration and the presence of managerial/institutional
ownership. Verrecchia (1990) argues that managers may have incentives to reveal more
information to outside investors in order to fill the information gap and to decrease
information costs arising from varied ownership.
1.1.1.1 Ownership concentration

The agency theory suggests that ownership-diffusion is positively related to the level of
corporate disclosure. Fama and Jensen (1983) propose that firms tend to reveal more
information when corporate ownership is diffused as disclosure should decrease the agency
costs. In a widely held company the conflicts of interest are significant and the amount of
monitoring costs is greater than that of a closely held company (Eng and Mak, 2003).
Widely held companies are then encouraged to issue more voluntary information in their
annual reports to show that they are acting in the best interest of shareholders and to
insure owners that their economic interests are optimized (Chau and Gray, 2002; Depoers,
2000; Ghazali and Weetman, 2006). Dumontier and Raffournier (1998) state that small
investors urges corporate management to comply with international accounting standards
because of their need for extensive disclosure in annual reports and the unaffordable
costs of collecting private information. Conversely, Ho and Wong (2001) argue that in a more
concentrated ownership, the impact on voluntary disclosures is more complicated. Conflicts
of interests are no more between managers and shareholders, but between large and small

CHAPTER VII
shareholders. Under very high ownership concentration, managers are encouraged to behave
against the interests of small shareholders by retaining corporate information (Ho and Wong,
2001).
Empirically, the findings appear to be mixed. For example, Gelb (2000) find that the quality
of annual reports increases when the ownership is less concentrated. Similarly, Chau and
Gray (2002), Haniffa and Cooke (2002) and Ho and Wong (2001) document a positive
association between wide ownership and the extent of corporate disclosure while Raffournier
(1995) and Eng and Mak (2003) report no significant association between ownership
diffusion and the level of voluntary disclosure. Hossain et al. (1994), Kelton and Yang
(2008), Mitchell et al., (1995), Samaha et al. (2012) and Schadewitz and Blevins (1998) end
in contrast to a negative relationship between ownership concentration and disclosure. This
implies that a greater percentage of substantial shareholder-ownership, leads to less
need for monitoring and transparent disclosure. Block holders seem not encouraged or
are not bothered with disclosure matters, either overall voluntary disclosure or specific
disclosure.

The growing body of risk disclosure literature invokes the agency theory to
draw the ownership structure as a determinant of the extent of risk
disclosure. For example Konishi and Ali (2007) and Mohobbot (2005)
contend that in concentrated ownership structures, risk information
will not be somehow disclosed via annual reports, but rather
communicated in the board room meetings or privately shared with
financial analysts. This is consistent with the contention that high
ownership concentration probably decreases firms' willingness to provide
risk disclosure to outside investors. As a result, there will be less
material risk disclosure available to the public. Abdul Hamid and Nordin
(2013) explain that in a concentrated ownership owners might not need
additional disclosure, as they are more closely related to board members,
as well as the management and hence couldnt be bothered by the extent
of any type of disclosure. Empirical risk disclosure literature provides
inconclusive results. Bauwhede and Willekens (2008) and Kajter (2006)
find a negative association between ownership concentration and risk
reporting, while Mohobbot (2005), Konishi and Ali (2007), Miihkinen

CHAPTER VII
(2012), Mokhtar and Mellett (2013) and Oliveira et al. (2011) find no
relationship between the two variables.
In a related vein, Fan and Wong (2002) and Donnelly and Lynch (2002)
suggest that ownership concentration influences the level of information
asymmetry between managers and outside investors, which in turn
influences managers accounting choices and consequently the
informativeness of accounting information. Empirically, the findings seem
to be mitigated. For example, Firth et al. (2007) find that concentration of
ownership is negatively associated with the informativeness of earnings in
the eyes of the minority investors while Yeo et al. (2002) and Sarikhani and
Ebrahimi (2011) show a positive relationship between the two variables
and Snchez-Ballesta and Garca-Meca (2007) reported an insignificant
impact on earnings informativeness.
Drawing from the agency theory, we formulate the following hypothesis:
H2. The association between voluntary risk disclosure and share price anticipation of
future earnings is significantly weaker for companies with concentrated ownership.

1.1.1.2. Managerial ownership

Managerial ownership refers to the proportion of shares owned by corporate management.


The agency theory describes managerial ownership as a possible solution for the conflicting
interests between the owners and agents (managers). Management holding is believed to
reduce the agency costs associated with the separation of ownership and control. This should
reduce the need to closely monitor the firms' managers as well as investors' information needs
(Gelb, 2000). Jensen and Meckling (1976)s interest alignment hypothesis suggests that
outside shareholders may ask for increased monitoring and accountability as managerial
ownership (i.e., ownership by executive directors) decreases. This monitoring constrains
managers to reveal more information than what is required by law or regulation. Drawing
from the entrenchment theory, Hossain et al. (2006) add that managers tend to disclose less

CHAPTER VII
information when managerial ownership is high because of their intention to evade outside
shareholders' monitoring and/or appropriate the wealth of minority shareholders.
Numerous empirical studies provide evidence consistent with Jensen and Meckling (1976).
Ruland et al. (1990) examine the relationship between management earnings forecasts and
managerial holdings. They show that as managerial ownership increases, firms are less likely
to issue management earnings forecasts. Gelb (2000) investigates the association between
insiders ownership and disclosure quality rating. The results show that annual and interim
disclosures of companies with low insider ownership are more likely to be highly rated. This
suggests that investors' information requirements, and thus corporate disclosures, are more
extensive for firms with low levels of managerial ownership. Eng and Mak (2003) predict that
voluntary disclosure is a substitute for outside monitoring and so is negatively related to
managerial ownership. They find evidence consistent with this prediction. Ghazali (2007)
evidences also that companies in which the directors hold a higher proportion of equity
shares disclose significantly less voluntary social information.
Recent risk disclosure literature provides evidence consistent with Jensen and
Meckling (1976). Marshall and Weetman (2007) assert that high levels of insider control, as
measured by the percentage of closely held shares, are associated with low extent of risk
disclosure. Brown et al. (2011) and Deumes and Knechel (2008) suggest also that firms with
greater (lower) inside (outside) ownership are likely to be more (less) secretive and less
(more) transparent. Their findings support their arguments in that they showed that managers
of firms with high levels of insider control have less incentive to disclose more risk
information.
In addition, managerial ownership affects the value relevance of accounting information.
Warfield et al. (1995) maintain that because accounting-based constraints are less prevalent
for firms with great level of managerial ownership (and thus low agency costs), managerial
manipulation of accounting numbers is less likely for such firms. Their findings indicate that
the information content of accounting disclosures increases with the level of managerial
ownership. The empirical evidence of Lichtenberg and Pushner (1992), Mehran (1995), Xu
and Wang (1999) and Mitton (2002) indicate also a positive association between disclosure
quality and management ownership consistent with the convergence of interest hypothesis.
Agrawal and Mandelker (1990), Jensen and Murphy (1990), Slovin and Sushka (1993) and
Loderer and Martin (1997), Gabrielsen et al. (2002) provide, in contrast, evidence of a

CHAPTER VII
negative association consistent with the entrenchment hypotheses. Sarikhani and Ebrahimi
(2011) find no significant relationship between the management ownership and earnings
informativeness.
Based on the interest alignment hypothesis, we state the following hypothesis:
H.3: The association between voluntary risk disclosure and share price anticipation of
future earnings is significantly weaker for companies with a high level of management
ownership.
1.1.1.3. Institutional ownership

Institutional investors hold a large block of shares in the equities of large companies. The
primary types of institutional investors include public and private pension funds, mutual
funds, insurance companies, investment funds and funds managed by banks or foundations
(Kochhar and David, 1996; p. 458). Unlike individual investors, institutional shareholders are
more sophisticated and manage their investment professionally. Their participation in
corporate ownership is believed to improve firm value due to more effective monitoring
(Vishny and Shleifer, 1986). Chung and Zhang (2011) notice that institutional
shareholders play an important role in improving corporate governance worldwide; firms
with large shareholders are more likely to be monitored than those with diffuse ownership.
With the significant resources and expertise they have, institutional owners have strong
capabilities to monitor corporate managers and to prevent them from making non-value
maximizing decisions or behaving opportunistically (Tihanyi et al., 2003; Velury and Jenkins,
2006).
With respect to voluntary disclosure, the agency literature suggests that institutional owners
could reduce the abuse of discretion over corporate disclosure arising from high managerial
ownership. Since monitoring is costly, it is done only when ownership is large enough to
motivate the acquisition of better information on managerial performance (Eng and Mak,
2003). Thus, managers may voluntarily disclose information to meet the expectations of large
shareholders. Bushee and Noe (2000) argue that the importance of corporate reporting to
institutional investors depends on their investment planning horizon, information gathering
capabilities, and governance activities. As a result, firms with a large ownership stake are
more likely to be surrounded by a much richer information environment (Rajgopal et al.,
1999).

CHAPTER VII
Empirical evidences on the relationship between institutional ownership and disclosure are
however mixed. For example Diamond and Verrecchia, (1991), Ajinkya et al., (2005),
Karamanou and Vafeas, (2005) and Barako, (2007) document a positive association between
institutional shareholdings and corporate disclosure practices while other studies report
different results. Makhija and Patton (2004) notice a nonlinear relationship between
corporate disclosure and the level of investment fund ownership. That is, the two variables
are positively associated at smaller levels of fund ownership, and negatively associated
at higher levels of ownership. Haniffa and Cooke (2002) and Eng and Mak (2003) find an
insignificant relationship between the two variables, suggesting that the main shareholders
can easily obtain information since they usually have access to private information via other
channels.
Within the risk disclosure literature, institutional ownership is expected to curb corporate
managers from restraining risk information disclosure. Prior work of Chalmers and Godfrey
(2004) and Taylor et al. (2008) argue that increased risk information may be a result of
institutional owners pressure, which behaves as a substitute for effective corporate
governance. Empirical studies did not, though support this agency perspective on the role of
institutional holding. For example, Bushee and Noe (2000) and Solomon et al. (2000) find no
relationship between the extent of long-term horizon institutional ownership and corporate
risk disclosure in annual reports. These findings show that institutional investors hold a
neutral view toward the need of risk disclosure. Abraham and Cox (2007) find that the
relationship between institutional investors and risk reporting depends on the category of
institutional owners.
Results indicate that in-house managed pension funds are negatively related to risk disclosure
supporting previous argument that long-term institutions can benefit from a mosaic of
information from private channels (Guercio and Hawkins, 1999; Bushee and Noe, 2000). On
the other side, institutional ownership by life assurance funds is positively related to risk
disclosure, suggesting that short term institutions prefer firms that report more risk
information because risk disclosure allows more accurate securities valuation and therefore
frequent trading strategies. Recently Zhang et al. (2013) provide some conflicting evidence
with Abraham and Cox (2007) findings. They report no relationship between long term
institutional ownership and risk disclosure categories. Only a positive association is found
between short term institutional owners and some risk categories, but not those that seem
relevant for investors, namely future-oriented risk and negative risk performance disclosures.

CHAPTER VII
Institutional investors are also shown to have an impact on accounting information quality
(Chung et al., 2002; Sarikhani and Ebrahimi, 2011). In particular, consistent with the active
monitoring hypothesis, large institutional investors are better informed and firms are likely to
engage in less opportunistic earnings management improving, hence, the informativeness of
accounting earnings. The presence of institutional investors will also deter managers from
performing strategic behavior towards earnings forecast information and constrains them to
disclose relevant financial information (Ajinkya et al., 2005).
Given shareholder activism and the monitoring potential of institutional shareholders, we
formulate the following hypothesis:
H.4: The association between voluntary risk disclosure and share price anticipation of
future earnings is stronger for firms with high institutional ownership.
1.1.2. Board characteristics
The board of directors plays an important role in enforcing governance standards within
publicly held companies. It is suggested to be the most important internal control device
seeking to control and monitor management in order to deter management from
opportunistic behavior Rose, (2007, p. 404). Previous accounting literature suggests that
several board characteristics may influence corporate disclosure. As already emphasized in
the literature, board characteristics are examined extensively in terms of size, composition,
diversity, or expertise. Fama (1980) and, more recently, Chau and Leung (2006) suggest that
the board of directors was an endogenously designed mechanism to unravel agency problems.
1.1.2.1. Board size

Corporate governance researchers argue that board size is an important market -based factor
which ensures that all major stakeholders receive reliable information about firm value. It
should also motivate managers to maximize firm value instead of pursuing personal
objectives (Luo, 2005). On one hand, Bassett et al. (2007) contend that a small board lacks for
sufficient expertise and may suffer from chief executive officer (CEO) dominance. This
impairs the board's ability to meet corporate governance responsibilities and yields in high
agency costs. On the other hand, the agency theory predicts that large boards, enjoy wide
expertise and more diversified knowledge which results in more effective boards monitoring
role (Luo, 2005; Singh et al., 2004). Based on this argument, Elzahar and Hussainey (2012)
suggest that large boards with wide expertise and particularly those with financial and

CHAPTER VII
accounting background, are more motivated to screen their efforts in risk management and
to signal this information to shareholders.
Cheng and Courtenay (2006) point out the lack of a dominant theory and a strong empirical
evidence to uphold an association between board size and the extent of corporate disclosure.
Some findings suggest that a large board size has a negative impact on the effectiveness of the
board. It may lead to less effective coordination, communication and decision-making (Jensen
1993). It may cause members to be less motivated to be involved in the strategic decision
making process (Goodstein et al. 1994).
The empirical evidence on the association between board size and voluntary disclosure
is inconclusive. Whilst Arcay and Vazquez (2005), Bassett et al., (2007), Cheng and
Courtenay, (2006) and Donnelly and Mulcahy, (2008) report a non-significant association
between the two variables, Khlif et al. (2014), Hussainey and Al-Najjar (2011) and Byard et
al. (2006) find positive and negative associations, respectively, between board size and
disclosure
Based on risk disclosure, some studies (Elzhar and Hussainey, 2012) find no relationship
between board size and risk reporting while others find a positive association (Laksmana,
2008; Hussainey and Al-Najjar, 2011; Mokhtar and Mellett, 2013). This reflects the increase
of board members' awareness regarding their duties to enhance corporate disclosure. Large
board size is likely to increase the diversity of the members expertise and guarantees a high level
of compliance with the accountability paradigm.
With respect to the earnings response coefficient studies, results seem to be inconclusive. For
example, while Vafeas (2000) ends to a negative relationship between the board size and
earnings informativeness, Dimitropoulos and Asteriou (2010), Sarikhani and Ebrahimi (2011)
and Firth et al. (2007) find no significant association between these two variables.
Based on these arguments and congruent with the agency theory and the resource dependence
theory we formulate the following hypothesis:
H5. The association between voluntary risk disclosure and share price anticipation of
future earnings is stronger for firms with a large board size.

CHAPTER VII
1.1.2.2. Board composition

Board composition is also emphasized as an important corporate governance mechanism in


the accounting literature. The Agency theory proposes that board composition may have an
impact on corporate disclosure activity. Abraham and Cox (2007) argue that the board of
directors includes a mix of inside and outside directors with variant tendencies toward
information disclosure. On the one hand, inside directors are full time corporate employees
and occupy an executive position within the firm. Because of the nature of their tasks, it may
be difficult for the insiders-board members to monitor managers actions. They lack sufficient
incentives regarding enhanced corporate disclosure because of their stewardship within the
firm and their behavior which may be open to more scrutiny (Abraham and Cox, 2007;
Leftwich et al., 1981). On the other hand, outside non-executive directors are expected to
provide independent advices to executive directors. They are believed to play a crucial role in
monitoring managers performance and limiting their opportunism which may lead to reduced
agency conflicts between managers and owners (e.g., Fama, 1980; Fama and Jensen, 1983;
Walsh and Seward, 1990). Because they aim to signal their competence to other potential
employers and to maintain their reputational capital, non-executive directors are expected to
be more effective in fulfilling shareholder preferences for accountability and transparency.
Therefore more disclosure is expected if they are actually carrying out their greater control
and monitoring of managerial decisions. Their dominance would provide more power to
force managers to disclose more private information (Haniffa and Cooke, 2002; Eng and
Mak, 2003).

Empirically, the association between board composition and disclosure is controversial. Some
studies find a positive association between independent board of directors and the level or the
quality of voluntary corporate disclosure (Boesso and Kumar, 2007; Chen and Jaggi, 2000;
Donnelly and Mulcahy, 2008; Forker, 1992; Gul and Leung, 2004; Cheng and Courtenay,
2006; Khlif et al. 2014; Wang and Hussainey, 2013) while others end to an insignificant
relationship between the two variables (Deumes and Knechel, 2008; Ho and Wong, 2001;
Haniffa and Cooke, 2002; Vandemele et al., 2009). Focusing on risk reporting, Abraham and
Cox (2007) reveal that different types of non-executive directors fulfil different functions
regarding corporate disclosure policy. First, the coefficient on dependent non-executive
directors is negative and insignificant, suggesting that despite the benefits (knowledge,

CHAPTER VII
specific expertise) they may bring to the firm, grey directors do not promote risk disclosure.
Second, findings show a significant positive relationship between the number of independent
non-executive directors and the level of risk disclosure, suggesting that independent directors
in great numbers are important in the transmission of risk information to investors. Barakat
and Hussainey, (2013) and Elshandidy et al. (2013) find similar results with respect to the
positive effect of the proportion of independent non-executive directors relative to board size
on the voluntary release of risk information. Allini et al (2014) and Elzahar and Hussainey
(2012) find, though insignificant association between the two variables.
With respect to value relevance literature, Dimitropoulos and Asteriou (2010), Firth et al.
(2007) and Petra (2007) observe that independent non-executive directors are positively
associated with earnings quality. Sarikhani and Ebrahimi (2011) report in contrast, no
significant association between outside participation on the board and the return earnings
relationship.
Based on the above arguments and consistent with the agency theory and the signaling theory,
we postulate that:
H6. The association between voluntary risk disclosure and share price anticipation of
future earnings is stronger for firms with high proportion of non-executive directors.

1.1.2.3. CEO/Chairman duality

Role duality occurs if the chief executive officer (CEO) holds the chairman position of the
board at the same time resulting in a unitary leadership structure. According to the agency
theory, concentration of decision-making power due to the unitary leadership structure can
significantly reduce the monitoring function of the board. Fama and Jensen (1983) suggest
that this combination of positions signals the absence of separation of decision management
and decision control and may erode board independence. Barako et al. (2006) argue that this
unitary leadership structure may facilitate opportunistic behavior by the CEO because of his
dominance over the board. Therefore, it is posited that for the board to be effective, it is
important to separate the CEO and chairman positions. Forker (1992) suggests that duality
may be detrimental to the quality of disclosure. Actually, the dominant personalities may
resist some governance and control mechanisms such as audit committee and non-

CHAPTER VII
executive directors, which may place pressure on the board and impair their governance role
regarding disclosure policies. Ho and Wong, (2001) believe that separating the CEO and the
board chairman position deters managers from withholding unfavorable information.
Ghazali and Weetman (2006) note likewise that separating CEO responsibilities and roles
from those of the board chair should support transparency and adequate disclosure in financial
reporting. Empirical studies provide mixed results. Some studies report a negative association
between role duality and corporate voluntary disclosure (Forker, 1992; Haniffa and Cooke,
2002; Gul and Leung, 2004; Khlif et al., 2014; Wang and Husainey, 2013). Other studies find
an insignificant association between CEO duality and voluntary disclosure (Cheng and
Courtenay, 2006; Ho and Wong, 2001)
Within the risk disclosure literature, Mokhtar and Mellett (2013) find a negative association
between role duality and mandatory risk disclosure which confirms that role duality is a
potential threat to disclosure quality. Elzahar and Hussainey (2012) and Vandemele et al.
(2009) show, however a non-significant association between role duality and voluntary
risk reporting.
Regarding the literature that examined the economic consequences for CEO/Chairman role
duality Firth et al (2007), Petra (2007) and Sarikhani and Ebrahimi (2011) find that the duality
of the chairman and CEO have no significant impact on discretionary accruals and the
earnings response coefficients respectively. They show that the duality of the CEO and
chairman roles cause no decrease in earnings informativeness.
Based on the arguments above and consistent with the agency theory, we formulate our
seventh hypothesis as follows:
H7. The association between voluntary risk disclosure and share price anticipation of
future earnings is weaker for firms with CEO/Chairman duality.
Notwithstanding the governance mechanisms that can positively influence corporate
willingness to provide voluntary risk information, managers may tend to not reveal risk
disclosure that would negatively affect their competitive advantage. Companies seem then to
adjust their reporting policy based on a tradeoff between cost and benefits. Proprietary costs
are likely to moderate the association between risk disclosure and the return future earnings
relationship.

CHAPTER VII
1.2. The effect of proprietary costs and voluntary risk disclosure on share price anticipation of
future earnings

The agency theory and the signaling theory suggest that voluntary risk disclosure is intended
to reduce information asymmetry about corporate uncertainties and future prospects and to
signal management performance with respect to business and financial targets. The
proprietary costs theory can explain why voluntary risk reporting may be unhelpful to users of
financial information. Since this theory addresses the costs and benefits of disclosure, it is
argued that managers may be undetermined of which position to adopt in relation to risk
disclosure (Abraham and Shrives, 2014). Marshall and Weetman, (2007) suggest that while
most firms are likely to own or to implement risk management systems, they may be
discouraged to disclose related information when they feel it is commercially sensitive. This
is mainly because outside parties such as corporate opponent, may use the information in
ways that it damages firms competitive position (Cormier et al., 2005). Abraham and Shrives
(2014) add that in such a situation, managers would find it difficult to decide how much and
what sort of information to disclose. On one hand, if they choose to limit their disclosure with
respect to corporate risk management system, investors may perceive it as a weakness in
analyzing and monitoring areas of such risk in their business. On the other hand, if managers
decide to disclose transparent risk information with respect to corporate exposure, such
information may be used in ways that are harmful to their interests. Managers should trade off
the benefits of expanded disclosure against the costs of disclosing potentially damaging
information (Abraham and Shrives, 2014).
Skinner (1994) and Healy et al. (1999) believe that managers should incur some upfront
costs by disclosing some proprietary information in order to enhance their reputation as a
credible discloser. According to Verrecchia (1983), proprietary costs are incurred when any
decrease in corporate future cash flows is likely to occur following the disclosure of particular
information. In a typical way, when a company reveals some good news, it would be more
attractive to capital market participants resulting in higher future cash flows. However, such
favorable information may also attract potential competitors and encourage them to enter the
product market reducing thereby those future cash flows. On the opposite side, the disclosure
of bad news may discourage outside shareholders and potential investors from (re)-investing
their resources into the firm leading to lower future cash flows. This unfavorable news may
also discourage potential competitors entering the market and as a result future cash flows
may increase. To face this dilemma, Abraham and Shrives (2014) suggest that corporations

CHAPTER VII
tend to shape their disclosure activity (e.g. Restrict or provide boilerplate) in order to reduce
any proprietary costs. Solomon et al. (2011) note also that managers may prefer to disclose
sensitive information to large shareholders or potential investors through other, more private,
reporting mechanisms such as special meetings. Prior empirical disclosure literature drew the
same evidence that companies decision to voluntarily disclose financial and non-financial
information ( Dedman and Lennox, 2009; Nichols, 2009) as well as segment information
(Harris, 1998; Botosan and Harris, 2000; Piotroski, 2001; Leuz, 2004; Prencipe, 2004),
management forecast of earnings (Bamber and Cheon, 1998; Clarkson et al, 1994; King and
Wallin , 1995; Li, 2010) and the AIMR disclosure ranking (Shin, 2001) are a negative
function of competitive costs.
As for risk disclosure literature, Abraham and Shrives (2014) found that managers do not
signal good risk management systems via comprehensive disclosures. Managers prefer
general nonspecific information, which could apply to any company within the same industry
consistent with the proprietary costs theory. The authors also found that companies tend to
manipulate disclosures to arrange themselves and do not wish to disclose sensitive
information that might be exploited by others including industry competitors. Alternatively,
Abraham and Shrives (2014) noticed that while a manager should prefer to disclose
information in order to screen his good risk management practice, thereby decreasing the cost
of capital, in practice proprietary costs may cause disclosure to be restrained. Corporations
consider risk information as something, which is useful for internal management only.
As an alternative, they may choose to disclose information privately and to provide risk
reporting, which is not useful to fill the disclosure gap.
With respect to disclosure informativeness studies, Hossain et al. (2006) suggested that the
product market competition is likely to reduce the value relevance of corporate voluntary
disclosure. Findings showed that firms with higher proprietary costs choose to disclose less
future earnings information suggesting that share price informativeness with respect to future
earnings is lower for firms with high proprietary cost because the information environment is
less rich. Ali et al. (2014) argued that companies in more concentrated industries face a
greater risk of revealing company secrets to their rivals through discretionary
disclosures. Such disclosure would be harmful to the competitive position of the firm because
rivals would take actions to enhance their profits at the expense of the firm. Results indicated
that the firms quality disclosure ratings are negatively related to industry concentration. Ali et

CHAPTER VII
al. (2014) contended additionally that if a firm discloses less, then analysts would find it more
difficult to make earnings forecasts for the firm. Findings showed that for firms in more
concentrated industries dispersion in analysts earnings forecasts is high, analyst
earnings forecast errors are large, and there is a greater volatility in analysts forecast
revisions. These results suggested that firms in more concentrated industries (i.e. Higher
proprietary costs) disclose less reliable voluntary information.
Based on what is discussed above, we formulate the following hypothesis:
H8. The association between voluntary risk disclosure and share price anticipation of
future earnings is weaker for firms with high proprietary costs.
As we discussed in the fifth chapter, we control likewise for the cross sectional effects of
some previously documented determinants of voluntary disclosure as well as earnings
response coefficients of our hypothesis testing. Namely, financial leverage, corporate size and
profitability are considered because it was also emphasized that differences between firms
corporate governance mechanisms could be due to some of their observable characteristics.

2. Research methodology
In this section we briefly review how we measured corporate risk disclosure, our sample
selection process and the multiple regression models.
2.2. Sample selection and data collection

The second aim of our study is to examine the simultaneous effect of voluntary risk
disclosure, corporate governance and proprietary costs on the ability of stock returns to
predict future earnings in a number of MENA emerging markets. As discussed in the second
chapter, the countries studied in our research were the focus of scarce studies in the
accounting and market based literature despite their particular legal, financial and governance
systems.
The MENA region is exhibiting a growing importance internationally with significant
external commercial exchanges, ongoing trade liberalization and inflow of foreign
investment. This is especially after the series of institutional, legal and financial reforms in the

CHAPTER VII
region as well as the financial crisis in East Asia and Argentina. Capital markets in the MENA
region gathered also increasing interest from world investors and policymakers due to their
growth potential compared to other emerging markets after the massive privatization plans
introduced in the region. They witnessed a recent market bubble and became a main channel
for substantial national and foreign funds. In line with this recent rapid growth, stock market
participants have been asking for more relevant and reliable financial information as a way to
ensure an effective resource allocation process. The World Doing Business (2012) report
indicates that most MENA countries and notably Tunisia, Morocco, Saudi Arabia, Kuwait
have been engaged recently in reinforcing investor protections and corporate transparency.
International financial reporting standards (IFRS) are required in most MENA emerging
markets (Bahrain, Kuwait, Oman, Qatar, UAE) and the rest of MENA countries (Tunisia,
Morocco, Saudi Arabia, Israel) are converging to IFRS (Pacter report, 2014) to attract
international investors and to enhance corporate disclosure.
The original sample covered twelve MENA countries; however, we applied some filtering
rules. Mainly, we dropped Israel from our initial sample because in this country firms are
dually listed and provide annual reports in conformity with the SEC requirements (10-K
form). Bahrain and Qatar were also dropped because their capital markets include mostly
financial and investment corporations and due to severe issues of data availability.
Accordingly, our final sample comprises companies from nine MENA emerging capital
markets, including Egypt, Jordan, Kuwait, Morocco, Oman, Saudi Arabia, Tunisia, Turkey
and UAE that are periodically listed from 2007 to 2012. The selection of this period of
analysis is triggered by the steady growth in GDP per capita in the region during these recent
years as well as the dynamism in their stock markets with a considerable increase in market
capitalization and total value traded.
The firms included in our sample, had to satisfy the three conditions: First,
it had to belong to a non-financial sector. Financial firms such as banks,
insurance firms and investment firms were excluded because their reports
are not comparable to those of non-financial firms. Second, this study
focuses on annual reports and no other media of financial communications
such as interim reports. Third, the non-financial firm, had to have at least
one annual report, from 2007-2009. We chose 2009 as the end year for
the study because the level of corporate risk disclosure is linked to share
price anticipation of earnings and accounting and return data are required

CHAPTER VII
for at least three years ahead (Year 2012). The initial number of available annual
reports varies from year to year. For example, the total number of firms in 2007 is 328. This
number increases in 2008 to 335 firms, and then it is reduced to 327 firms in 2009.
We matched the selected companies with the Thomson one database codes from which we
gathered financial information such as stock prices, earnings per share, asset growth, and net
profit margin ratio. Some firms have no Thomson one code. Thus, they have no accounting
and return data. These firms are excluded from the selected sample. In the second stage we
collected governance variables with respect to ownership structure and board characteristics
from Thomson one database as well as corporate annual reports and we dropped other firms
from our sample because of some missing information. Before we perform the regression
analysis, we deleted further observations because of missing data and outliers are censored to
avoid any undue influence of extreme observations. Outliers are defined in the present study
as the top and bottom 1% of observations for the distribution of any of the regression
variables.13

After these series of sample-filtering steps the sample is reduced from 990 to: 785
observations regarding the regression model that examines ownership structure hypotheses;
789 observations related board characteristics hypotheses and 809 observations for the
regression model that addresses the proprietary costs hypothesis. The resulting panel includes
320 companies. The average number of annual reports per company is about 2.5. Table 16 and
17 summarizes the composition of the sample by country and by economic sector.
Table.16 Sample composition by Country

Country/Model
M1
M2
M3
Egypt
68
66
68
Jordan
118
124
122
Kuwait
58
58
57
Morocco
48
48
48
Oman
143
143
146
Saudi Arabia
81
86
87
Tunisia
55
53
56
Turkey
151
145
154
UAE
63
66
71
Total
785
789
809 1995). In
13 Excluding extreme observations is consistent with prior literature (e.g., Kothari and Zimmerman,
addition, censoring the top and bottom 1% of observations is one of the acceptable methodologies to reduce
departures from normality (see Foster, 1986).

CHAPTER VII
Table.17 Sample composition by Sector

Economic Sector/Model
Consumer Discretionary
Consumer Staples
Energy
Healthcare
Industrials
Information Technology
Telecommunication Services
Materials
Utilities
Total

M1
156
115
36
19
173
19
34
201
32
785

M2
166
114
34
22
172
19
34
194
34
789

M3
161
122
36
20
177
19
36
205
33
809

2.1. Measuring voluntary risk disclosure

As debated earlier in the fifth chapter, many approaches were used in the accounting literature
to assess the level of the narrative disclosure. In our research, we rely on the content analysis
to examine the extent to which companies tend to issue voluntary risk information. We opted
for manual approach content analysis to ensure a better judgment of words and phrases
meaning within a context specially when dealing with data in different languages.

Another advantage of this manual approach over a computer-aided content analysis is that,
though time consuming, it enables the differentiation between voluntary and mandatory
statements, the identification of topics or themes associated with voluntary risk disclosures
and their scoring separately.
Consistent with the commonly adopted sampling unit, we retrieved risk information from
corporate annual reports since they are considered as a main vehicle of financial and nonfinancial information for outside investors. In particular, we performed the analysis of
corporate risk disclosures on main narrative sections in the annual reports, including the
management discussion and analysis (MD&A) section and board reports. This choice is
motivated by the fact that the management discussion and analysis (MD&A) section as well
as board reports are providing more and more detailed information about corporate business,
their strategy and performance. Through these narrative sections, corporate management
looks to inform investors about corporate activity, its perspectives and uncertainties. Such
information is meant to help market participants in their investment decision process.

CHAPTER VII
Researchers employed a variety of coding and measurement units such as words, sentences
and proportion of pages in analyzing narrative reporting. We refer to the number of riskrelated sentences as a reliable measure of risk disclosure levels. This is consistent with the
recent literature on corporate risk disclosure activity and with the claim that words cannot be
coded into different risk categories without reference to the sentence. Since categorical and
thematic distinctions are required in content analysis approach, we rely additionally on
Linsley and Shrives (2006) grid as a coding instrument in our research.
We retained five categories of risk information and we dropped the financial risk category.
They are related to operations risk, empowerment risk, information processing and
technology risk, integrity risk and strategic risk whereby we identified 32 items. We also
referred to Linsley and Shrives (2006) definition of risk disclosure. We coded sentences as
risk disclosures if the reader is informed of any opportunity or prospect, or of any hazard,
danger, harm, threat or exposure, that has already impacted or may impact upon the company,
as well as the management of any such opportunity, prospect, hazard, harm, threat, or
exposure. We generated an aggregated score for risk disclosure for each firm by counting the
number of risk-related sentences in corporate annual reports.

2.3. Research design

2.3.1. Regression model specification


To assess the simultaneous effect of voluntary risk disclosure, corporate governance and
proprietary costs on share price informativeness with respect to future earnings, we follow
prior studies and in particular Kothari and Sloan (1992) and Collins et al. (1994) papers.
Based on the concept that earnings lack of timeliness for recognizing economic events
compared to stock returns, these papers, add future earnings and some control variables into
the regression of current returns on current earnings. This regression model is used as a
benchmark in many recent studies (e.g. Banghoj and Plenborg, 2008; Gelb and Zarowin,
2002; Hussainey and Walker, 2009; Lundholm and Myers, 2002; Schleicher et al 2007) when
examining the value relevance of current earnings as well as the ability of stock returns to
predict future performance.
Collins et al (1994) applied the following specification:

CHAPTER VII
N

k 1

k 1

Rt b0 b1 X t bk 1 X t k bk N 1 Rt k b2 N 2 EPt 1 b2 N 3 AGt

Where:
Rt: stock return for year t
Rt+1, Rt+2, Rt+3: stock returns for year t+1, t+2, t+3 respectively.
Xt, Xt+1, Xt+2, Xt+3: are defined as the earnings change for year t, t+1, t+2, t+3 respectively.
AGt: is the growth rate of the total book value of assets for period t.
EPt-1: is the period t-1s earnings over price at the start of period t.
To test our hypotheses we interact all right-hand side variables in Collins et al (1994)
regression model with risk disclosure (RDi,t) variable and all corporate governance variables
(Govi) each one a part so that we detect the simultaneous effect of these explanatory variables
on the association between stock return and future earnings. We verified subsequently the
supposed moderating effect of proprietary cost on share price anticipation of earnings by
interacting it with the voluntary risk disclosure variable (RDi,t). This yields to the following
regression models:

M1.
3

R t b 0 b1 X t b k 1 X t k b k 4 R t k b 8 EPt 1 b 9 AG t b10 RDi, t b11[ RDi, t * X t ]


k 1

k 1

k 1

k 1

b k 11[RDi, t * X t k ] b k 14 [RDi, t * R t k ] b18 [RDi, t * EPt 1 ] b19 [RDi, t * AG t ]


3

b 20 Gov i b 21[Gov i * X t ] b k 21[Gov i * X t k ] b k 24 [Gov i * R t k ] b 28 [Gov i * EPt 1 ]


k 1

k 1

b 29 [Gov i * AG t ] b 30 [Gov i * RDi, t * X t ] b k 30 [Gov i * RDi, t * X t k ]


k 1

b k 33[Gov i * RDi, t * R t k ] b 37 [Gov i * RDi, t * EPt 1 ] b 38 [Gov i * RDi, t * AG t ] e t

k 1

CHAPTER VII
M2.

R t b 0 b1X t b k 1X t k b k 4 R t k b 8 EPt 1 b 9 AG t b10 RDi, t


k 1

k 1

k 1

k 1

b11[ RDi, t * X t ] b k 11[ RDi, t * X t k ] b k 14 [ RDi, t * R t k ] b18 [ RDi, t * EPt 1 ]


3

b19 [RDi, t * AG t ] b 20 PCi, t b 21[ PCi, t * X t ] b k 21[ PCi, t * X t k ]


k 1

b k 24 [PCi, t * R t k ] b 28 [PCi, t * EPt 1 ] b 29 [PCi, t * AG t ] b 30 [PCi, t * RDi, t * X t ]

k 1

k 1

k 1

b k 30 [ PCi, t * RDi, t * X t k ] b k 33 [PCi, t * RDi, t * R t k ] b 37 [ PCi, t * RDi, t * EPt 1 ]


b 38 [ PCi, t * RDi, t * AG t ] e t

We made two changes relative to the Collins et al. (1994) regression model. First, we control
for the cross sectional variation among corporate characteristics. In fact, as a considerable
number of studies on the returns-earnings relationship documented many determinants (past
growth, risk, earnings persistence, firm size, firm profitability and the presence of an
accounting loss) of the earnings response coefficient, we examine our hypothesis with some
of these variables (Firm size, profitability and risk) included in the regressions. We included
one control variable at a time in the model due to the limited number of observations (degrees
of freedom). Second, in calculating the current and future earnings growth variables we
deflated earnings change by price and not by lagged earnings since it is argued that it will be
difficult to define earnings growth when lagged earnings are negative or zero (Hussainey and
Walker, 2009).
Predictions for the signs of the regression models (M1 and M2) parameters are summarized
in the following paragraphs according to prior literature (Lev, 1989; Lundholm and Myers,

2002; Hussainey, 2004). First the coefficient

b1

on current earnings growth, Xt is expected to


b2

be positive. Also, the future earnings response coefficients

b3

and

b4

on Xt+k are expected

to be positive. This is may be explained by the fact that industry-wide and economy-wide

CHAPTER VII
effects should allow the market predict some portion of the firms future earnings growth even
though the annual report discussion sections do not involve risk disclosure information.

b5 b6
The coefficients

b7
and

on the future returns Rt+k (the measurement error proxy) are

expected besides to be negative as any unanticipated future events in the period t would lead
to higher earnings change and to positive returns in the t+k periods when the news becomes

b8
available to the market. Additionally, the coefficient

EPt 1
on lagged earnings

is expected to

be positive because of the negative association between the price to earnings ratio and

b9
expected earnings growth for the current and future years.

is predicted to be negative

because of the positive association between asset growth AGt and the expected growth in
future earnings.
The second part of the equations M1 & M2 interacts the variables in the first part with the risk

b10

i, t

disclosure measure (RD ). First, there is no particular prediction for

. Consistent with

i, t

Lundholm and Myers (2002) we included RD

in the regression because it is part of the

i, t

interaction terms [RD Xt+k] and to avoid that the interaction terms would inadvertently
proxy for the level of corporate disclosure. Lundholm and Myers (2002) predict also a
RDi, t * X t

negative coefficient on [

] as

it is suggested that the level of corporate disclosure may

cause a substitution effect away from current earnings and toward future earnings. However
Hussainey (2004) argues that enhanced voluntary disclosure would make earnings
announcements more credible and thus, the sign for b11 will be difficult to forecast. Since [RD
i, t

Xt+k]14 proxies for the future earnings news that is revealed by the firms risk disclosures,

14 Note that RDXt+k proxies for the future earnings news that is revealed by the
firms risk disclosures, whereas Xt+k and Rt+k proxy for all future earnings news,
regardless of the source.

CHAPTER VII
b12 b13
b14
the coefficients ,
and
are predicted to be positive. This is because it is postulated that

share price anticipation of future earnings changes should increase with the level of risk
disclosure.
In other words, a high level of risk disclosure is likely to reveal information about future

bk 14 b18
earnings. Finally, there are no particular predictions for the coefficients

b19
and

. It is

noteworthy indeed that Rt+k and Xt+k measure the change in expectations about future earnings.
It looks plausible that both sets of variables would be interacted with RD to proxy for the
revelation of future performance news through risk disclosure.
The last part of the equation (M1) interacts simultaneously the variables in the second part
with the governance mechanisms variables (Govi). First, there is no particular prediction for

b20
. Consistent with Lundholm and Myers (2002) approach we included Govi in the
i, t

regression because it is part of the interaction terms [Govi*RD *Xt+k] and to avoid that the
interaction terms would inadvertently proxy for the governance structure. The coefficient b30
i, t

on [Govi*RD *Xt] is difficult to predict because some governance mechanisms associated


with a high level of risk disclosure may enhance the credibility of current earnings while
others are likely to decrease the amount of risk information.
i, t

To the extent that [Govi*RD *Xt+k] proxies for the simultaneous effect of some governance
structures and risk disclosure level on share price anticipation of future earnings, the
b 31 b 32
b 33
coefficients
,
and
are predicted to be either positive or negative. This is because

while it is postulated that a high level of risk disclosure is likely to reveal information about
future earnings, ownership structure or board characteristics may lead to an increase or a
decrease in such association. Finally, there are no particular predictions for the coefficients
b k 33 b 37
b 38
,
and
. It is noteworthy that Rt+k and Xt+k measure the change in expectations about

future earnings. It looks plausible that both sets of variables would be interacted with RD and

CHAPTER VII
Govi to proxy for the effect of governance structures on the revelation of future performance
news through risk disclosure.
The last part of the equation (M2) interacts simultaneously the variables in the second part
i, t

with the proprietary costs variable (PC ). Similar to what is stated above, there is no

b20
particular prediction for

. In fact, consistent with Lundholm and Myers (2002) approach

i, t

we included PC

i, t

i, t

in the regression because it is part of the interaction terms [PC *RD

*Xt+k] and to avoid that the interaction terms would inadvertently proxy for the presence of
proprietary cost.
i, t

i, t

The coefficient b30 on [PC *RD *Xt] is difficult to predict because the presence of
proprietary costs may influence the extent of risk disclosure and accordingly the credibility of
current earnings while it is also suggested that the level of corporate disclosure may cause a
substitution effect away from current earnings and toward future earnings.
i, t

i, t

To the extent that [PC *RD *Xt+k] proxies for the simultaneous effect of proprietary costs
b 31 b 32
and risk disclosure level on share price anticipation of future earnings, the coefficients ,

b 33

and

are predicted to be negative. This is because while we postulated that a high level of

risk disclosure is likely to reveal information about future earnings, the presence of
proprietary costs may lead to a decrease in such association. Finally, there are no particular
b k 33 b 37
b 38
predictions for the coefficients
,
and . It is noteworthy that Rt+k and Xt+k measure

the change in expectations about future earnings. It seems then plausible that both sets of
i, t

variables would be interacted with RD

i, t

and PC

to proxy for the effect of proprietary costs

on the revelation of future performance news through risk disclosure.

CHAPTER VII
2.3.2. Variables definition
As discussed above, we tested the association between the level of risk disclosure, corporate
governance mechanisms and proprietary costs, and the return-future earnings relationship
using Collins et al. (1994) regression model. In this model, current stock returns is the
dependent variable, while the independent variables include current and future earnings,
future returns, past earnings to price ratio and asset growth. We collected earnings and
accounting data from Thomson one database while we gathered governance variables either
from Thomson one database or from corporate annual reports. We present the definition of
each variable in the next paragraphs.
2.3.2.1 Stock returns

We calculated annual returns based on monthly stock prices. We measured monthly returns as
the growth in the price of a share over months, assuming that dividends are re-invested to
purchase additional units of equity at the closing price applicable on the ex-dividend date.
After computing monthly returns, they are 'geometrically linked' to produce an annual return
using these formulas:
-

Monthly Returns : r1 (m0, m1) = (Pricem1 - Pricem0) / Price m0


Annual Return: Ryear (2007) = [(1+ r1)*(1+r2)*(1+r3).(1+r12)] - 1

We defined return of year t as the return for the 12 month period starting six months after the
financial year-end of year t1. In other words, the chosen return window extends from 6
months prior to the financial year-end to 6 months after the financial year-end. We chose the
six month lag to ensure that the information in annual report narratives have been read by the

market.

15

Rt 1 Rt 2
We calculated

Rt 3
and

as the buy-and-hold returns for the one-year

period starting six, eighteen and thirty months after the firms current financial year-end.

15 At an early stage, we identified the dates under the chairmens statements. In most
cases, the annual reports are published within four to five months of the firms
financial year-end. We assumed that one further month to be sufficient for market
participants to read and process the information.

CHAPTER VII
Rt 1 Rt 2
Accordingly, the return windows for

Rt 3
and

do not overlap with the current return

window (see Hussainey, 2004).


2.3.2.2 Earnings variables

The measure for earnings per share is provided by Reuters fundamentals available on
Thomson one database which is calculated by dividing earnings for ordinary-full tax by the

X t X t 1 X t 2
number of shares outstanding. We defined

X t 3
and

as earnings change

deflated by share price. Both current and future earnings changes are deflated by the price at
the start of the return window for period t (see Lundholm and Myers, 2002). For example, we
calculated earnings variables for the year 2007 as follows:

Xt
= (EPS2007-EP2006)/Price2006,

X t 1
= (EPS2008-EPS2007)/Price2006,

X t 2
= (EPS2009-EPS2008)/Price2006.

X t 3
And

= (EPS2010-EPS2009)/Price2006.

Where: Price2006 is the price six months after the year-end for 2006, in other words at the
start of return window for 2007 year.
2.3.2.3 Past earnings to price ratio

EPt 1
We defined past earnings,

as earnings for the period t1 divided by price six months

EPt 1
after the financial year-end of period t1. For example, we calculated

for a December

2007 observation by dividing EPS for 2006 by stock price at the end of June 2007.

CHAPTER VII
2.3.2.4 Asset growth

AGi ,t
Asset growth

is the growth rate of the total book value of assets for the year t. We

estimated growth of assets as the change in book value of assets in year t, divided by the book
value of assets at the end of t1. For example, asset growth in 2007 equals total assets in 2007
minus total assets 2006 divided by total assets in 2006.
2.3.2.5 Disclosure level

We measured risk disclosure variable (RDi,t) as the natural logarithm of the number of riskrelated sentences, respectively, for operations risk, empowerment risk, information processing
and technology risk, integrity risk and strategic risk information. Logged values improve the
approximation of the independent variables and mitigate problems with residuals such as
heteroskedasticity.
2.3.2.6 Governance structure

Ownership structure

Ownership concentration

We defined ownership concentration variable (OwCi,t) as the proportion of equity held by the
top-five largest shareholders, including financial institutions, firms inside shareholders
(directors and executives) and other outside block shareholders consistent with previous
studies such as Jiang et al (2011); Hossain et al (2006); Lakhal, (2006) and Makhija and
Patton, (2004).
-

Managerial ownership

We measured managerial ownership variable (MOw i,t) as the proportion of equity held by
managers and executive directors consistent with extant studies such as Gul et al (2002) and
Hossain et al (2006).
-

Institutional ownership

We estimated institutional ownership variable (InOw i,t) as the total number of shares held by
institutions to the total number of shares outstanding per sample firm, consistent with
Abraham and Cox (2007), Barako (2007), Elzahar and Hussainey (2012) and Zhang et al.
(2013).

CHAPTER VII
Board characteristics
-

Board size
i, t

Board size variable (BS ) is the number of directors sitting on the board at the end of each
year. This measure is consistent with Abdel-Fattah et al. (2008), Elzahar and Hussainey
(2012), Hussainey and Al-Najjar, (2011), Lakhal (2005) and Singh et al. (2004).
-

Board composition
i, t

The percentage of non-executive directors (NED ) is the proportion of non-executive


directors relative to the board size. It measures board composition consistent with Abraham
and Cox (2007), Elshandidy et al (2013), Elzhar and Hussainey (2012) and Vandemele et al
(2009).
-

CEO duality

The CEO/Chairman role duality (Dual) is a dummy variable that we defined as 1 if CEO is
the Chairman and 0 otherwise.
2.3.2.7 Proprietary costs

A reliable measure of proprietary costs is still challenging scholars. Although a


number of proxies of proprietary costs were proposed, no proxy has gained general
acceptance. Prior analytical models estimate competitive costs, conditioned on how
competition is defined. One group of studies model competition in the context of an
entry game where a firm decides whether to enter a particular new product market
(e.g., Darrough and Stoughton, 1990) while the others model competition in the
context of a post-entry game (e.g., Darrough, 1993). According to Monk (2011), the
type of competition has an impact on the disclosure equilibrium outcomes. The
difference between product market competition and the threat of entry was shown to
be enough to change the effect of competition on voluntary disclosures.
The industry concentration ratio is the most widely used proxy of the product market
competition in the accounting literature (Verrecchia, 1983; Stiglitz, 1987; Sutton, 1991;
Harris, 1998; Botosan and Stanford, 2005; Verrecchia and Weber, 2006; Ali et al.,
2014). In these studies, it is believed that highly concentrated industries are less
competitive. This involves a higher barrier to entry and a lower proprietary information
costs, which leads to a higher level of disclosure.

CHAPTER VII
Ali et al. (2014) and Dedman and Lennox (2009) suggest however that relying on the
industry concentration variable is problematic. They infer that high industry
concentration does not necessarily correspond with low competition intensity and that
an industry may have relatively few participants, but low barriers to entry. The use of
industry concentration ratio may lead then to incorrect conclusions and inferences,
especially that it assumes that all companies within a given industry face the same
level of competition.
For the purpose of our research we estimate the proprietary costs variable (PCi,t) using
a firm specific measure that is the Profit Margin ratio (Revenue/Income) consistent
with Cohen (2005), Dedman and Lennox (2009) and Tang (2010). According to these
studies, corporations with high profit margins attract future competition and exhibit
higher forewarning of potential entrants. In fact, high profit margin reflects a corporate
successful strategy that yields a considerable competitive advantage compared to other
existing rivals. Such competitive advantage can be only maintained if competitors
cannot reproduce the companys activities. Consequently, the presence of a competitive
advantage should discourage disclosure of business opportunities and expansion plans.
Highly profitable company has then an incentive to hide proprietary information in
order to prevent adverse reactions from less successful rivals.
2.3.2.8 Controlling variables

Financial leverage
Consistent with Banghoj and Plenborg (2008), Elshandidy et al (2013) Leverage is the book
value of equity scaled by total liabilities.
Firm size
Firm size is the natural logarithm of corporate net sales (turnover) congruent with Abraham
and Cox (2007) and Linsley and Shrives (2006).
Profitability
Firm profitability is the natural logarithm of the return on equity (ROE) which is defined as
the [Net profit after tax/Shareholders funds]*100% in accordance with Elshandidy et al
(2013) and Elzahar and Hussainey (2012).

CHAPTER VII
Industry sector
Industry sector is a dummy variable defined as 1 if the firm is classified into one of the nine
broadly defined industry sectors and 0 otherwise.
3. Descriptive statistics
We provide in Table 18 summary statistics for our samples with observations coming from the
year 2007 to 2009. The mean current return ranges between 2.9 and 3.38 percent. The mean
current earnings per share change varies between 0.47 and 0.9 percent of the price. The
aggregated mean of future earnings change is respectively 0.23, 0.5 and 0.34 percent of the
price for the three periods ahead t+1, t+2 and t+3. These statistics suggest a decline in future
performance in t+1, t+2 and t+3 compared to the performance of the current period. We find
also a lower mean future return with respect to t+1 periods compared to current return. There
is a reverse and an increase in the mean stock returns for period t+2 and t+3 (compared to t+1)
which indicate changes in corporate performance over the sample time period. As additionally
reported in table 18, the mean of corporate asset growth rate extends from 12.2 to 12.87
percent and the standard deviation is about 0.17 suggesting that there is little variation in the
asset growth rate among our sample firms.
The level of risk disclosure is on average 27 sentences and the standard deviation is on
average 21.3 reflecting a fairly low disclosure score over our period of analysis as well as
considerable dispersion among our sample firms. As for governance variables, first ownership
concentration ranges from 12.14 percent to 94.98 percent with a mean (median) of 61.8 (65)
percent and a standard deviation about 19.38. We argue that corporate ownership is highly
concentrated for our sample of Middle Eastern and North African firms though the
considerable variation among them. Second, managerial ownership ranges from 13.37 percent
to 79.93 percent. It has a mean of 13.37 percent and a standard deviation about 19.97. This
shows that managers do not hold large stakes in corporate ownership, despite the large spread
among the observations. Similarly, we notice that the extent of institutional owners is low
with a mean of 13.29 and a considerable dispersion of 15.44.

CHAPTER VII
Table .18 Panel A. Summary descriptive statistics for continuous variables
(longitudinal data)

Var.

Mean
S.D
Min.
.
Rt
0.029
0.457
-0.597
Xt
0.005
0.045
-0.075
Xt+1
0.001
0.040
-0.073
Xt+2
0.004
0.042
-0.069
Xt+3
0.002
0.034
-0.064
Rt+1
0.008
0.449
-0.605
Rt+2
0.100
0.363
-0.363
Rt+3
0.044
0.284
-0.346
AGt
0.127
0.178
-0.089
Ept-1
0.888
1.346
-0.01
RD
27.483
21.307
1
OwC
61.808
19.386
12.14
MOw
13.373
19.979
0
InOw
13.299
15.449
0
Lev
40.487
42.353
0.000
Size (M$) 662.338 1949.595
0.83
Profit
26.606 152.193 0.05

25%

Medi
an
-0.044
0.002
0.000
0.000
0.000
-0.067
0.028
-0.004
0.0833
0.17
22
65
3.309
9
25.191
117.07
14.71

-0.340
-0.015
-0.018
-0.016
-0.010
-0.339
-0.174
-0.185
-0.009
0.03
13
50.04
0
0.56
2.949
22.43
7.33

75%

Max.

0.316
0.025
0.021
0.023
0.019
0.284
0.303
.25
0.228
1.14
36
77.1
18.5
19.61
69.458
437.39
25.6

0.926
0.090
0.072
0.084
0.062
0.869
0.812
0.554
0.491
4.05
145
94.98
78.93
77.422
124.028
25101.77
92.90

Obs
.
785
785
785
785
785
785
785
785
785
785
785
785
785
785
785
785
785

Table 18 reports the summary statistics for the sample firms using data pooled across the three year sample
period. The earnings per share measure is a Reuters fundamentals item, calculated by dividing earnings for
ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change
deflated by price. Both current and future earnings changes are deflated by the price at the start of the return
window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a
12-month period, starting six months after the end of the previous financial year. EPt1 is defined as period t1s
earnings over price six months after the financial year-end of period t1. AGt is the growth rate of the total book
value of assets for period t. RD is the total number of risk related sentences, respectively, for Operations risk,
Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk information.
Unlogged values are reported. In subsequent regressions the natural logarithm is used. OwC is the proportion of
equity held by the top-five largest shareholders. MOw is the proportion of equity held by managers and
executive director. InOw is the total number of shares held by institutions to the total number of shares
outstanding per sample firm. Profit is measured by the Return on Equity (ROE). Likewise, unlogged values are
reported and in subsequent regressions the natural logarithm is used.
i, t

With respect to board characteristics, Board size (BS ) ranges from 2 members of directors
to 23 members of directors with a mean of 8.20 and a standard deviation of 2.42. Board
composition non-executive directors ranges from 0 to 100 percent with an average of 68.10
percent. This refers to a relatively good level of independence for the board in listed MENA
companies. The CEO/chairman role duality exists among 29.78 percent of our sample firms.
70.22 percent of observations opted for a separate position which is likely to foster board
monitoring role.

CHAPTER VII

Table .18 Panel B. Summary descriptive statistics for continuous variables


(Pooled data)

Var.

Mean
S.D
.
Rt
0.030
0.527
Xt
0.009
0.111
Xt+1
0.004
0.114
Xt+2
0.005
0.103
Xt+3
0.006
0. 086
Rt+1
0.062
0.609
Rt+2
0.154
0.565
Rt+3
0.084
0.474
AGt
0. 122
0.173
Ept-1
0.90
1.374
RD
27.343
21.255
BS
8.209
2.426
NED
68.108
23.643
Lev
41.418
43.168
Size (M$) 658.017 1936.599
Profit
26.792 151.942

Min.

25%

-0.921
-0.310
-0.521
-0.441
-0.401
-0.943
-0.860
-0.816
-0.090
-0.01
1
2
0
0
5.48
0.05

-0.342
-0.016
-0.018
-0.016
-0.010
-0.329
-0.171
-0.181
-0.014
0.04
13
7
50
3.753
21.75
7.25

Medi
an
-0.047
0.002
0.000
0.000
0
-0.061
0.024
0
0.081
0.19
22
8
70
25.673
117.015
14.44

75%

Max.

0.312
0.025
0.021
0.022
0.018
0.299
0.304
0.252
0.223
1.08
35
9
87.5
70.857
422.37
25.59

1.815
0.659
0.566
0.507
0.426
0.878
0.851
0.558
0.470
4.17
145
23
100
127.23
25101.77
98.729

Obs
.
789
789
789
789
789
789
789
789
789
789
789
789
789
789
789
789

Table 18 reports the summary statistics for the sample firms using data pooled across the three year sample
period. The earnings per share measure is a Reuters fundamentals item, calculated by dividing earnings for
ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change
deflated by price. Both current and future earnings changes are deflated by the price at the start of the return
window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a
12-month period, starting six months after the end of the previous financial year. EPt1 is defined as period t1s
earnings over price six months after the financial year-end of period t1. AGt is the growth rate of the total book
value of assets for period t. RD is the total number of risk related sentences, respectively, for Operations risk,
Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk information.
Unlogged values are reported. In subsequent regressions the natural logarithm is used. BS is the number of
directors sitting on the board at the end of each year. NED is the proportion of non-executive directors relative to
the Board size. Financial leverage Lev is defined as book value of equity scaled by total liabilities. Size is
measured by corporate revenues. Profit is measured by the Return on Equity (ROE). Likewise, unlogged values
are reported and in subsequent regressions the natural logarithm is used.
Table .18 Panel C. Summary descriptive statistics for dummy variable (longitudinal data)

DUAL
0
1

Frequency

Percent

Cum.

Obs

554
235

70.22
29.78

70.22
100.00

789
789

Dual is defined as 1 if CEO is the Chairman and 0 otherwise.

Proprietary costs level ranges from -34.454 percent to 95.2 percent. Corporate net profit
margin has a mean of 0. 35 and a standard deviation of 5.744 may indicate relatively high
competition among firms in the MENA region.

CHAPTER VII
Regarding control variables, first, the mean of corporate financial leverage, as measured by
debt to equity ratio differs slightly between the three regression models. On average, it runs
from 41.418 to 42.499 percent with high standard deviations.
This suggests that our sample includes moderately geared firms with high spread among their
financial leverage level. Second, there is a considerable variation in our sample firm size. In
fact, corporate revenue for the first quartile is about $22 million, while it is more than $420 in
the third quartile. Thus the present study does not focus only on large firms in MENA
emerging markets. Finally, the mean of corporate profitability as measured by the return on
equity (ROE) goes from 25.96 to 26.79 indicating that firms are relatively well performing
during the period of analysis though the substantial dispersion among them.
Table .18 Panel D. Summary descriptive statistics for Proprietary cost variable model (longitudinal data)

Variable

Mean

S.D

Min.

25%

Median

75%

Max.

Obs.

s
Rt

.
0.033

0.464

-0.

-0.340

-0.044

0.317

0.945

809

Xt

8
0.004

0.046

597
-0.078

0.002

0.027

0.0925

809

Xt+1

7
0.002

0.042

-0.073

0.0173
-0.018

0.0001

0.023

0.079

809

Xt+2

8
0.004

0.043

-0.070

0.0002

0.024

0.0875

809

Xt+3

5
0.003

0.034

0.0179
-

0.000

0.025

0.064

809

0.0107
-

-0.0679

0.2816

0.8514

809

Rt+1

0.021

0.4421

0.0617
-0.604

Rt+2

0.108

0.3631

0.3402
-

0.0350

0.3125

0.8139

809

Rt+3

3
0.044

0.2850

0.3538
-

0.1718
-

0.2519

0.5532

809

AGt

7
0.128

0.1796

0.3478
-

0.1854
-

0.082

0.2327

0.4919

809

Ept-1

4
0.885

1.3427

0.0894
-0.02

0.0108
0.03

0.17

1.18

4.05

809

RD

5
27.26

21.181

13

22

35

145

809

PC

5
0. 350

5.744

0.033

0.093

0.198

95.2

809

CHAPTER VII
Lev

42.49

194.06

34.454
-0.059

3.164

25.892

71.209

411.943

809

Size

9
660.5

1930.0

0.83

22.19

117.07

437.39

25101.7

809

(M$)
Profit

5
25.96

2
149.13

184.3

7
3216.18

809

0.05

0.12

14.59

Table 18 Panel D reports the summary statistics for the sample firms using longitudinal data across the three year sample
period. The earnings per share measure is a Reuters fundamental item, calculated by dividing earnings for ordinary-full tax
by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both current and
future earnings changes are deflated by price at the start of the return window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated
as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting six months after the end of the previous
financial year. EPt1 is defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is
the growth rate of the total book value of assets for period t. RD is the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk information.
Unlogged values are reported. In subsequent regressions the natural logarithm is used. Proprietary cost PC is defined as net
profit margin. Financial leverage Lev is defined as book value of equity scaled by total liabilities. Size is measured by
corporate revenues. Profit is measured by the Return on Equity (ROE). Likewise, unlogged values are reported and in
subsequent regressions the natural logarithm is used.

Table 19 presents pairwise Pearson correlations for all regression variables. P values are given
in parentheses. Correlations are estimated using longitudinal and pooled data across the threeyear sample period. As documented in the former empirical chapter, the correlation between
current Returns (Rt) and current earnings growth (Xt) is strong and significant at the 1% level
suggesting that current earnings are perceived as value relevant. The correlation between Rt
and Xt+1 is weaker, but still significant at 1% and 5% levels. The correlations between current
returns Rt and future earnings change for t+2 and t+3 (Xt+2 and Xt+3) are not significant. These
results may provide evidence of prices leading earnings by only one period.
Current return is also correlated with future returns of period t+1while Rt is uncorrelated with
Rt+2 and Rt+3. Future returns (Rt+1, Rt+2 and Rt+3) are significantly correlated with future
earnings growth (Xt+1, Xt+2 and Xt+3), consistent with Collins et al. (1994). These
correlations indicate that future are good proxies of the measurement error in future
earnings.
We noticed significant and negative correlation between current earnings (Xt) and future
earnings growth for 3 periods of analysis. Similarly, Xt+1 is significantly and negatively related
to Xt+2 and Xt+3. This may suggest potential multi-collinearity problems within the independent
variables. The variable inflation factor (VIF) did not raise significant problems
among the explanatory variables. The mean VIF is about 1.12 and the
computed VIF for each predictor variable is largely under 5. AGt and
Ept-1 seem to be also good error measurement proxies. The theory

CHAPTER VII
indicates that an errors-in-variables proxy should be highly correlated with
the measurement error but uncorrelated with the dependent variable. This
is the case for these two control variables. As reported in table 3, the
correlation coefficients between Rt in one hand and AGt as well as Ept-1 in
the other hand are insignificant. Risk disclosure level is positively and
significantly associated with current stock returns at the level of 10%. This
may suggest that corporate risk disclosure activity in MENA emerging
markets is induced by stock price performance. Lundholm and Myers
(2002) and Banghoj and Plenborg (2008) find in the same way a positive
and significant correlation between current returns and the disclosure
score.
i, t

As for ownership structure, institutional ownership (InOw ) is negatively


correlated with future returns of t+2 and t+3 periods at the level of 10%,
suggesting that institutional demand exercises a pressure on stock prices
which is likely to reduce future stock returns. Pairwise testing reveals that
Institutional ownership is positively correlated with ownership
concentration and negatively related to managerial ownership at 1% level.
These correlations suggest that firms with high level of institutional owners
have concentrated ownership but less management ownership. Table 18
panel A reports a positive association between managerial ownership
i, t

(MOw ) and future returns (Rt+2) at the 5% level consistent with the view
that management ownership positively impact firm value. Ownership
concentration is positively associated with the level of risk disclosure
indicating that large shareholders pressure is likely to improve corporate
reporting practice.

CHAPTER VII
Table .19 Panel A: Pearson Correlations: Shareholding Structure (longitudinal Data)

Rt
Rt
Xt

Xt+1

1.000
0.156**
*
(0.000)
0.108**
*
(0.002)

Xt

Xt+1

Xt+2

Xt+3

Rt+1

Rt+ 3

EPt1

AGt

RD

InsOw

MOw

Ow
C

CHAPTER VII

1.000

Xt+2

-0.023
(0.505)

0.203**
*
(0.000)
-0.044
(0.210)

Xt+3

0.020
(0.569)

-0.035
(0.327)

Rt+1

0.254**
*
(0.000)
-0.004
(0.904)

-0.025
(0.472)

0.137**
*
(0.000)

1.000

0.026
(0.466)

0.017
(0.628)

0.106**
*
(0.002)

Rt+3

0.055
(0.119)

0.040
(0.253)

-0.007
(0.835)

EPt1

-0.024
(0.489)

0.079**
(0.025)

AGt

0.042
(0.233)

RD

0.146**
*
(0.000)
0.075**
(0.034)

1.000

0.138**
*
(0.000)
0.157**
*
(0.000)

1.000

1.000

-0.030
(0.389)

0.104**
*
(0.003)
-0.063*
(0.075)

0.060*
(0.093
)
0.118*
**
(0.000
)
0.065*
(0.066
)
-0.039
(0.266)

0.077**
(0.030)

-0.021
(0.543)

-0.009
(0.784)

0.067*
(0.057
)

-0.019
(0.587)

0.005
(0.886)

InsO
w

-0.005
(0.869)

-0.000
(0.988)

MO
w

-0.023
(0.515)

-0.000
(0.981)

Rt+2

Rt+2

0.171***
(0.000)

1.000

0.039
(0.265)

0.046
(0.192)

1.000

0.015
(0.662)

0.004
(0.906)

1.000

0.026
(0.464)

-0.051
(0.150)

0.112**
*
(0.001)

0.061*
(0.087
)
0.091*
*
(0.010
)

0.077**
(0.029)

1.000

0.031
(0.373)

-0.018
(0.608
)

-0.054
(0.126)

-0.001
(0.957
)

1.000

0.025
(0.472)

-0.033
(0.355
)

0.025
(0.481)

1.000

-0.034
(0.336)

-0.016
(0.642
)

0.064
*
(0.06
9)
0.024
(0.489
)

0.052
(0.13
8)

0.052
(0.142)

0.0590
*
(0.098
)
0.088*
*
(0.012

0.164*
**
(0.000
)
-0.008
(0.809)

0.014
(0.68
9)

0.008
(0.821)

0.072*
*
(0.043
)
0.004
(0.896)

0.038
(0.283)

0.056
(0.11
5)

-0.048
(0.175)

0.122**
*

0.053
(0.133)

1.000

CHAPTER VII
Table 19 presents Pearson Correlations for all regression variables using panel data across the three year sample period. P-values are given
in parentheses. The number of observations is 785. The earnings per share measure is a Reuters fundamental item, calculated by
dividing earnings for ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated
by price. Both current and future earnings changes are deflated by the price at the start of the return window for period t. Rt, Rt+1, Rt+2 and
Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting six months after the end of the
previous financial year. EPt1 is defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is the
growth rate of the total book value of assets for period t RD is the natural logarithm of the total number of risk related sentences,
respectively, for Operation risk, Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk
information. OwC is the proportion of equity held by the top-five largest shareholders. MOw is the proportion of equity held by managers
and executive director. InsOw is the total number of shares held by institutions to the total number of shares outstanding per sample firm.

i, t

With respect to board characteristics, pairwise testing in table 19 panel B showed that board size (BS ) is
significantly and positively related to the earnings of period t-1. In contrast, there is negative correlation between
i, t

BS

and current and future earnings growth for t+2s period at the levels of 5%. These univariate results may

indicate that the larger is corporate board size the smaller is firms current and future earnings changes. Board size
seems to be likewise significantly and negatively associated with the amount of non-executive directors sitting on
the board. This shows that large sized board in our sample includes a low amount of non-executive directors. Nonexecutive directors are significantly and positively associated with the future returns for t+2s period, though it is
negatively correlated with future earnings growth for the same period. This suggests that firms with high proportion
of independent members on board have higher future returns despite the lower earnings growth for t+2. In
contrast, firms with high number of independent directors have more important future earnings change for t+3. As
for the CEO/Chairman role duality, it is negatively correlated with future returns (Rt+1, Rt+3) and the proportion of
i, t

NED

showing that firms with role duality exhibit lower stock performance and involve less proportion of outside

directors.

Rt
Rt
Xt

Xt+1

1.000
0.134*
**
(0.000)
0.097*
**
(0.006)

Xt+2

0.008
(0.820)

Xt+3

-0.018
(0.603)

Rt+1

0.223*
**
(0.000)
0.010
(0.776)

Rt+2

Xt

Xt+1

Xt+2

Xt+3

Rt+1

Rt+2

Rt+ 3

EPt1

AGt

RD

BS

1.000

0.287*
**
(0.000)
0.017
(0.632)

0.105*
**
(0.003)
-0.019
(0.586)

NED

Dua
l

CHAPTER VII
1.000

0.203*
**
(0.000)
-0.058*
(0.098)

0.195*
**
(0.000)

1.000

0.219*
**
(0.000)
0.111*
**
(0.001)

0.071*
*
(0.045)

-0.014
(0.687)

0.083*
*
(0.019)

1.000

0.099**
*
(0.005)
0.056
(0.114)

1.000

1.000

-0.019
(0.576)

0.150*
**
(0.000)
0.032
(0.368)

0.037
(0.289)

1.000

-0.021
(0.548)

-0.038
(0.279)

-0.052
(0.138)

0.085*
*
(0.016)
-0.012
(0.731)

1.000

-0.013
(0709)

1.000

0.024
(0.48
8)
0.047
(0.17
8)

1.000

0.050
(0.15
6)

1.000

0.024
(0.49
8)

0.010
(0.76
2)

0.136*
**
(0.000)

Rt+3

0.014
(0.682)

0.036
(0.305)

-0.026
(0.461)

EPt1

-0.017
(0.618)

-0.036
(0.306)

-0.012
(0.730)

0.141*
**
(0.000)
-0.036
(0.300)

AGt

-0.036
(0.312)

0.007
(0.839)

0.006
(0.866)

0.041
(0.247)

-0.007
(0.833)

-0.000
(0.979)

0.061*
(0.086)

RD

0.064*
(0.071
)

-0.003
(0.922)

-0.005
(0.869)

0.021
(0.543)

0.003
(0.920)

0.144*
**
(0.000)

-0.030
(0.385)

-0.051
(0.151)

-0.019
(0.585)

0.182*
**
(0.000)

0.041
(0.241)

0.071*
*
(0.044
)
0.069*
(0.051
)

-0.023
(0.519)

NE
D

0.086*
*
(0.015
)
0.026
(0.465)

0.083*
*
(0.018)
0.098*
**
(0.005)

0.003
(0.931)

BS

0.078*
*
(0.028
)
0.082*
(0.020
)
0.056
(0.111)

0.061*
(0.083)

0.026
(0.449)

0.122*
**
(0.000)

0.028
(0.433)

0.102*
**
(0.003
)

1.000

CHAPTER VII
Table.19 Panel B: Pearson Correlations: Board Characteristics (Pooled Data)

CHAPTER VII
Table 19 presents Pearson Correlations for all regression variables using panel data across the three year sample period. P-values are given in parentheses. The number of
observations is 789. The earnings per share measure is a Reuters fundamental item, calculated by dividing earnings for ordinary-full tax by the number of shares
outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both current and future earnings changes are deflated by the price at the start of the
return window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting six months after the end of
the previous financial year. EPt1 is defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is the growth rate of the total book
value of assets for period t. RD is the natural logarithm of the total number of risk related sentences, respectively, for Operations risk, Empowerment risk, Information
processing and technology risk, Integrity risk and Strategic risk information. BS is the number of directors sitting on the board at the end of each year. NED is the proportion
of non-executive directors relative to the Board size. Dual is defined as 1 if CEO is the Chairman and 0 otherwise.

i, t

Table 19 Panel C reports descriptive statistics for regression model with proprietary cost variable. The level of proprietary cost (PC ) as
measured by the net profit margin is positively and significantly correlated with the earnings yield of prior period (t-1). This suggests that the
proprietary costs are increasing in function of the earnings to price ratio. As the earnings yield serves for many investment managers to determine
optimal asset allocations, a high level is likely to attract more competitors into the product market. Pairwise testing reveals also that proprietary
cost level is significantly and negatively correlated with risk disclosure pointing out that corporations tend to decrease their risk disclosure when
the proprietary costs are high.

CHAPTER VII

Rt
Rt
Xt
Xt+1

Xt

Xt+1

Xt+2

Xt+3

Rt+1

Rt+2

Rt+ 3

EPt1

AGt

RD

PC

1.000
0.143***

1.000

(0.000)
0.0843**

(0.016)

0.200*

CHAPTER VII
1.000

**

Xt+2

0.0132

(0.000)
-0.059*

(0.706)

(0.093)

0.144*

1.000

**

Xt+3

0.0132

-0.035

(0.000)
-0.062*

(0.708)

(0.308)

(0.077)

0.138*

1.000

**

Rt+1

Rt+2

Rt+3

EPt1

-0.249***

-0.036

0.170*

(0.000)
0.153*

-0.061*

(0.000)

(0.301)

**

**

(0.081)

0.0175

0.032

(0.000)
0.004

(0.000)
0.067*

0.132*

(0.619)

(0.358)

(0.907)

(0.054)

**

0.169*

(0.000)

**

1.000

1.000

0.054

0.018

-0.002

0.116*

0.067*

(0.000)
0.041

0.039

(0.123)

(0.599)

(0.974)

**

(0.053)

(0.171)

(0.256)

-0.024

-0.040

(0.001)
-0.060*

-0.038

0.014

0.003

-0.072**

(0.489)

0.073*

(0.244)

(0.087)

(0.273)

(0.686)

(0.914)

(0.040)

1.000

1.000

AGt

RD

PC

0.035

(0.036)
0.073*

-0.034

-0.002

0.020

-0.056

0.099***

-0.083**

0.063*

(0.320)

(0.322)

(0.951)

(0.558)

(0.110)

(0.004)

(0.017)

(0.070)

0.0622*

(0.036)
-0.021

-0.005

0.027

-0.020

0.085*

0.064*

0.002

0.162*

0.030

(0.076)

(0.544)

(0.880)

(0.431)

(0.569)

(0.065)

(0.955)

**

(0.385)

0.0341

0.022

-0.016

-0.0087

-0.025

(0.015)
-0.027

0.001

-0.044

(0.000)
0.162*

0.043

(0.332)

(0.518)

(0.645)

(0.805)

(0.467)

(0.437)

(0.978)

(0.211)

**

(0.221)

0.092**

1.000

1.000

1.000

CHAPTER VII
Table.19 Panel C: Pearson Correlations: Proprietary Cost (longitudinal Data)
Table 19 presents Pearson Correlations for all regression variables using panel data across the three year sample period. P-values are given in parentheses. The number of observations is 809. The earnings per share
measure is a Reuters fundamental item, calculated by dividing earnings for ordinary-full tax by the number of shares outstanding. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both current and
future earnings changes are deflated by the price at the start of the return window for period t. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12-month period, starting six
months after the end of the previous financial year. EPt1 is defined as period t1s earnings over price six months after the financial year-end of period t1. AGt is the growth rate of the total book value of assets for
period t. RD is the natural logarithm of the total number of risk related sentences, respectively, for Operations risk, Empowerment risk, Information processing and technology risk, Integrity risk and Strategic risk
information. Proprietary cost PC is the net profit margin defined as Net Income/Revenue.

CHAPTER VII
Summary
The hypotheses in this chapter are developed based on voluntary and risk disclosure
literature and on value relevance studies within the literature review (chapter V). The
main hypotheses in this chapter predict that ownership structure, board characteristics and
proprietary costs are likely to enhance/reduce the association between voluntary risk
disclosures and share price informativeness with respect to future earnings. This chapter
explains also the research methodology employed in this second empirical chapter. We
collected our data for a sample of firms listed in nine MENA emerging markets. We applied
content analysis, defined the measures of our variables and developed the two regression
models. Content analysis has been widely criticized by the researchers because of its
subjectivity. However, the researchers suggest many steps to minimize the limitations and
subjectivity of the content analysis approach, steps which are followed in this thesis in
order to achieve acceptable reliability and enhance the validity of this study. We derived all
the regression models from Collins et al (1994) and subsequent studies on the return future
earnings relationship. The chapter concludes with presenting the descriptive statistics of our
data and with highlighting the pairwise correlations between our predictors.
The last chapter discusses our empirical findings regarding the interaction effect that
governance structure and proprietary costs may exert on share price informativeness with
respect to revealed future earnings.

CHAPTER VIII
Chapter VIII Corporate governance, proprietary costs and share price
informativeness with respect to future earnings news: empirical
evidences

Introduction
This chapter presents the main findings of the regression analysis of our panel data. It is
divided into two sections. Section1 examines the regression diagnostics. Panel regression
analysis is more and more applied within the accounting and financial literature. It is
important that researchers make sure all the regression assumptions are met and valid before
drawing any conclusion. Statistical test and techniques are applied in this study in order to
support the findings and draw valid conclusion. Section 2 discusses the main empirical results
for this chapter. This last chapter should be reviewed in line with the discussion provided
in the methodology (chapter VI) which explains the process and techniques applied to
end up with these findings.
1. Panel regression analysis
Before we run the multiple regressions based on panel data, we perform several specification
tests in order to insure that the regression specification fits the data. We address five main
concerns: the question of whether to pool the data or not, the tests for individual and time
effects, the normality, the heteroskedasticity and the serial correlation of error terms.
1.1. Pooling test

A pooling/ heterogeneity test seeks to find out whether the intercepts in the regression take on
a common value . An important advantage of panel data models, compared to cross-sectional
regression models, is that we can allow for heterogeneity among individuals, generally
through individual-specific parameters. The main purpose of the first procedure is to justify
the use of individual-specific effects. The null hypothesis under the pooling test can be
expressed as H0: 1= 2= 3 = t= . Beck (2001) proposes to use the Chow test to compare
the pooled and the unpooled estimates under the assumption that the error term uitN
(0, 2).

CHAPTER VIII
To make sure that the underlying assumption of the Chow test is correct,
we run in a first step a Pantest2 for the three regression models which
examine the ownership structure (M1), the board characteristics (M2) and
the proprietary cost (M3) hypotheses.
The Stata command provides the Jacques Bera Test for normality of
residuals. The Skewness/Kurtosis test distributed as a chi2 with 2 degrees
of freedom returns the following results.
Table.20 The Jacques Bera Test for normality of residuals

Obs
785
789
809

M1
M2
M3

Adj Chi2 (2)


3.81
5.56
4.63

Joint
Prob>chi2
0.148
0.062
0.098

Based on these findings, we failed to reject the null of normality at the 5%


level and we assume then that error terms are normally distributed. In the
next step, we ran the Chow test statistic which follows an F distribution
with (N- 1, NT-N-K-1) degrees of freedom. In Stata, this statistic is
generated automatically after running a fixed effect analysis (xtreg, fe)
for our three regression models. This yields the following results:
Table.21 The Chow-test for fixed effect model

M1
M2
M3

Obs
785
789
809

F Statistics
1.21
0.97
1.22

Prob > F
0.029
0.596
0.024

The Chow-test which compares between a fixed effect model and an OLS
regression rejected the null of homogeneity among individuals for the
regression models that analyze the ownership structure (M1) and the
proprietary costs (M3) hypotheses at the level of 5%. In contrast, it failed
to reject the null of homogeneity at the level of 5% for the regression
model (M2) that tests the board characteristics hypotheses.
Baltagi (2005) recommends the Roy-Zellner test as a more suitable test for
poolability as it is argued that the Chow test lacks for robustness under

CHAPTER VIII
non spherical disturbances. To perform a Roy-Zellner test in Stata, we
followed a three step procedure. First, we interacted the regressors with a
set of individual dummies to generate the unpooled data. Second, we
perform a random effect regression on the unpooled data. Third, we test
for the equality of the individual coefficients. Stata uses a Wald test
distributed as a 2 with 72 degrees of freedom, which returns a value of
256.5 and 242.97 for M1 and M3 rejecting the null of poolability across
individuals at the level of 1%. With respect M2, the Roy-Zellner test returns
a value of 73.25 failing to reject the null of poolability at the level of 5% (p
value= 0.436).
1.2. Tests for individual and time effects

The Chow and the Roy-Zellner tests rejected the homogeneity hypothesis
of our longitudinal data which is used to examine the ownership structure
and the proprietary costs hypotheses. The OLS estimator is no longer the
best unbiased linear estimator. The panel data models provide two main
ways to deal with these specific effects: fixed effects and random
effects regression models. The main difference between fixed and random
effects models resides in the role of dummy variables. A parameter
estimate of a dummy variable is a part of the intercept in a fixed effect
model and an error component in a random effect model. Slope
coefficients remain the same across subjects or time period in either fixed
or random effect model.
We use the Hausman specification test to compare between a fixed effect
model and a random effect model. Random effects (RE) are preferred
under the null hypothesis due to higher efficiency, while under the
alternative fixed effects (FE) is at least consistent and thus preferred. The
test distributed as a Chi2 with 19 degrees of freedom yields a value of
210.97 and a Prob. > Chi2=0.000 for M1 and a value of 263.72 with a
Prob. > Chi2=0.000 for M3. The obtained results indicate that random
effects estimation is inconsistent and that the fixed effects model is more
appropriate for the panel data dealing with M1 and M3 regression models.

CHAPTER VIII
As shown earlier, the Chow and the Roy-Zellner tests both failed to reject
the null hypothesis of homogeneity in our cross sectional time series data
for M2 testing. That is, individual effects ui (cross-sectional or time specific
effects) do not exist (ui=0) and ordinary least squares (OLS) produces
efficient and consistent parameter estimates.
After choosing between an OLS, a fixed and a random effects model it is
important to see if we need to run time fixed effects regression. We use
the Stata command testparm after performing a fixed effects regression
with year dummies. The test produced the following F statistics.
Table.22 TestParm for a time fixed effects regression

M1
M3

Obs.
785
809

F Statistics
44.72
71.78

Prob > F
0.000
0.000

We then reject the null hypothesis that years coefficients are jointly equal
to zero at the level of 1%. Unit and time fixed effects are therefore needed
when testing for ownership structure and proprietary cost assumptions.
1.3. Residual diagnostics

The pooled OLS and the (individual/time) fixed effects regressions invoke
the least squares estimator for point and interval estimates under the
classical and core assumptions that the error terms are normally
distributed and homoscedastic. A specific feature for panel effects is that
error terms should be serially uncorrelated with the same variance across
time and individuals. The pooled OLS regression requires further no multicollinearity among independent variables.
1.3.1. Normality of the residuals
The normality of the residuals is important for assurance that the pvalues for t-tests and F-test are valid. The normality assumption is
not required neither in the independent variables distribution, nor to
obtain unbiased estimates of the regression coefficients. We examine
this assumption using the Stata program as follows: we first conduct a
fixed effects regression, then we implement the Jacques Bera test which is

CHAPTER VIII
provided via the command Pantest2. For the pooled data analysis, we
conducted an OLS regression, generated residuals then ran the sktest
command.
The Skewness/Kurtosis test distributed as a chi2 with 2 degrees of freedom
returns the following values:
Table 23 The Jacques Bera test for Normality of residuals

Obs
785
789
809

M1
M2
M3

Adj Chi2(2)
3.81
5.56
4.63

Joint
Prob>chi2
0.148
0.062
0.098

The test failed to reject the null hypothesis of normality of residuals at the
5% level.

1.3.2.

Homoscedasticity

A good regression model must account for the variancemean relationship


adequately. As raised by Baltagi (2005), assuming homoscedastic
disturbances while heteroskedasticity is present affects the efficiency of
the regression coefficient though these estimates remain unbiased. Their
standard errors will be in contrast biased. One should computes robust
standard errors to correct for the possible presence of heteroskedasticity.
One way for detecting heteroskedasticity in Stata is the xttest3 command.
This test calculates a modified Wald test for groupwise heteroskedasticity
in the residuals for a fixed-effects regression model. The null hypothesis
specifies that i2 =2 for i =1...Ng, where Ng is the number of crosssectional units. The modified Wald test distributed as a Chi2 statistic with
322 degrees of freedom produced the following results:
Table 24 The Wald test for groupwise heteroskedasticity

Regression
Models
M1

Obs.

Chi2 Statistic

Prob. > Chi2

785

43.873

0.000

CHAPTER VIII
M3

809

43.513

0.000

We reject accordingly the null hypothesis of homoscedastic disturbances


for our panel data. We handle heteroskedasticity problem by computing
standard errors that are robust to the homoscedasticity specification
consistent with Baltagi (2005).
With respect to the pooled OLS regression, we examine the potential
heteroskedasticity problem using the hettest stata command after
conducting an OLS regression analysis. This command provides the
Breusch-Pagan / Cook-Weisberg test under the null hypothesis that the
variance of residuals is homogenous. The test returns a chi2 statistic of
1.72 with a Prob. > chi2 = 0.1899 failing to reject thus the null of constant
variance in residuals.
1.3.3. Serial correlation
The contemporaneous and the serial correlation among panel data is
another issue that we should deal with. According to Baltagi (2005),
contemporaneous correlation is a problem in macro panels with long time
series (over 20-30 years). This is not much of a problem in micro panels (a
few years and a large number of cases) like the present study. Baltagi
(2005) argues that a model with individual effects has composite errors
that are serially correlated by definition. The presence of the timeinvariant error component gives rise to serial correlation which does not
disappear over time (Hsiao, 2013).
Disregarding these correlations will result in inefficient estimates (though
consistent) of the regression coefficients and biased standard errors.
Stata provides a test of serial correlation for a fixed effect model. The
xtserial., fe command performs a Wooldridge test for autocorrelation in
panel data. The test rejects the null hypothesis of no first order
autocorrelation at the 1% level (F (1,181) = 38.83 and Prob.> F=0.000).
In summary, our multiple regression analysis of panel data (M1 & M3) is
based on individual and time fixed effect models. To correct for

CHAPTER VIII
heteroskedasticity and for a first order (AR1) serial correlation in error
terms, we model fixed effects regression with the cluster option which
provides robust estimates of the regression parameters according to
Hoechle (2007). For pooled model (M2) we examine board characteristics hypotheses
based on the traditional ordinary least square regression.
1.4. Multi-collinearity

Multi-collinearity is a common problem when estimating linear or generalized linear models.


It occurs when there are high correlations among predictor variables, leading to unreliable and
unstable estimates of regression coefficients. The most widely-used diagnostic for multicollinearity is the variance inflation factor (VIF). We calculated the VIF for each predictor in
our M2 testing via the Collin command in Stata. As a rule of thumb, a variable which VIF
value is greater than 10 may need further investigation. Likewise, a tolerance value that is
lower than 0.1 is comparable to a VIF of 10. It is worth noting that high VIF is commonly
noticed in interaction models. Recent studies (e.g. Brambor et al., 2006; Echambadi
and Hess, 2007) argue that this issue has been overstated since the coefficients in
interaction models no longer indicate the average effect of a variable as they do in an
additive model. consequently, they are almost certain to change with the inclusion of
an interaction term and this should not be interpreted as a sign of multicollinearity.
Allison (1999) suggests also that we can safely ignore multicollinearity as long as the
variables of interest and the overall mean VIF in interaction models are below the
critic threshold. Recall that Echambadi and Hess (2007) showed the shortness of the
main applied method (mean centering) in addressing the collinearity issue in moderated
multiple regression models.
The following table provides the values of VIF as well as tolerance values which reveal a
limited problem of multi-collinearity.
Table. 25 VIF and Tolerance values

Variables
Rt
Xt
Xt+1
Xt+2
Xt+3
Rt+1
Rt+2

VIF
1.12
1.19
1.27
1.22
1.12
1.22
1.10

1/VIF
0.892
0.839
0.786
0.818
0.896
0.816
0.907

CHAPTER VIII
Rt+3
AGt
Ept-1
RD
BS
NED
CEO
Lev
Size
Profit
Mean VIF
2.

1.03
1.02
1.17
1.14
1.13
1.14
1.13
1.03
1.25
1.12

0.966
0.985
0.858
0.858
0.885
0.873
0.883
0.974
0.798
0.896
1.14

Results discussion and analysis


This subsection discusses the empirical findings regarding the effect of ownership structure,
board characteristics and proprietary costs on share price informativeness with respect to
future earnings news as conveyed via corporate risk disclosure.
2.1. Ownership structure, risk disclosure and share price anticipation of future earnings.

We mainly focused on three corporate ownership characteristics. We hypothesized that


concentrated ownership, managerial ownership and institutional ownership are likely to shape
(increase/decrease) the current return- revealed future earnings relationship.
2.1.1. Concentrated ownership, risk disclosure and the return future earnings relationship
Table 26 provides panel regression estimates for the simultaneous impact of risk information
and the level of ownership concentration on the markets ability to anticipate future earnings
growth. It reports main findings of H2 testing that predicts a weaker value relevance of
voluntary risk disclosure when interacted with the ownership concentration level.
The coefficient on Xt is, though positive, insignificant in all cases (the three
regressions with each control variable apart). The future earnings response
coefficients on Xt+1, Xt+2 and Xt+3 are, unlike what is expected, negative and
insignificant in the three regression models with one exception. When we
control for financial leverage level, the coefficient on Xt+1 is -8.065 with a pvalue of 0.075. These results suggest that current stock price is not
associated with future earnings change. So there is no evidence of prices
leading earnings regardless of corporate risk disclosure and their
ownership concentration level. The coefficients on future stock returns

CHAPTER VIII
(Rt+1, Rt+2 and Rt+3) have, as predicted, the right sign and are significantly
different from zero mainly for t+1 and t+2 periods. These findings indicate
that realized future earnings contain measurement error that future
returns remove. Findings revealed that risk disclosure level is positively
and significantly associated with current stock returns at the 10% level.
This indicates that risk reporting activity for our sample firms is driven by
stock price performance.
The incremental predictive value of risk disclosure level in anticipating
3

X t+K
future earnings is given by the coefficient on RD
K=1

. The

i, t

coefficients on RD *Xt+2 are positive and significant. They are about 0.365
and 0.233 and significantly different from zero at the level of 1% and 10%
when controlling for corporate risk and profitability respectively. Since RD
i, t

i, t

*Xt+i and RD *Rt+i together proxy for the revealed future earnings, and
because the individual years of future earnings are highly correlated, a
i, t

i, t

more powerful test examines the joint significance of RD *Xt+2 and RD


*Rt+2 for the regression that controls for corporate size. The partial F-test of
the joint significance of these variables have p-values of 0.0125 which
confirms the anticipated idea that risk disclosures published in the annual
report reveal information about future earnings for two years ahead. There
is, accordingly, evidence on share price informativeness with respect to
future earnings change when interacted with risk disclosure level.
Table 26 reports a positive and significant association between current
returns and current and future earnings when interacted with the level of
i, t

ownership concentration. The coefficient on the interacted term OwC *Xt


is 2.109 and is significantly different from 0 at the level of 10% (a p-value
of 0.053) when we control for financial leverage level. This finding

CHAPTER VIII
suggests that investors perceive the contemporaneous earnings, driven by
the presence of block ownership, as value relevant.
i, t

In the same way, the coefficients on the interacted terms OwC *Xt+1 range
from 1.650 to 2.348 and are positive and significant at the level of 5%
mainly when we control for corporate risk and profitability. Besides, the Fi, t

test on the sum of the coefficients of interaction terms OwC *Xt+1 and
i, t

OwC *Rt+1, yields a p-value of 0.0003 when controlling for corporate size.
There is hence strong evidence on share price informativeness with
respect future earnings of period t+1 driven by the presence of block
ownership. Increased concentrated ownership enhances the relative
amount of firm-specific information about future earnings in stock prices
as block holders do have superior access to private information.
Consequently, their trading can convey a timely and an accurate firmspecific information into stock prices (Carlton and Fischel 1983).
We turn to our hypothesis testing, which is concerned with the joint effects
of risk disclosure and concentrated ownership on share price anticipation
of future earnings. The incremental predictive value of both variables for
i, t

i, t

anticipating future earnings is given by the coefficients on OwC *RD *


3

X t+K

K =1

i, t

. The coefficients on the interaction terms OwC


i, t

and OwC

i, t

*RD *Xt+1

i, t

*RD *Xt+2 are negative and significantly different from zero at


i, t

the levels of 1% and 5% in all cases. The joint significance test of OwC
i, t

i, t

i, t

*RD *Xt+2 and OwC *RD *Rt+2 returns a value of 10.89 with a p-value of
0.000 in the regression model that controls for corporate size. Besides, the
i, t

i, t

coefficient on OwC *RD *Xt+3 is negative and significant at the level of


10% when we control for corporate profitability. These findings indicate

CHAPTER VIII
that a high level of ownership concentration moderates the association
between voluntary risk disclosure level and the returnfuture earnings
relationship as predicted in H2. The negative coefficients on the interacted
terms show that the amount of voluntary risk information impounded into
the current stock price is lower for firms with less diffused ownership
because investors perceive it as less relevant. Increased ownership
concentration is likely to intensify agency problems and worsen the
credibility of firms information environments. Investors prefer to rely on
other sources of information such as block owners trading to anticipate
future earnings growth. These results give support to our hypothesis that
predicted a significantly weaker association between voluntary risk
disclosure and share price anticipation of future earnings for companies
with concentrated shareholding structure. Our findings are consistent with
Yu (2011) and Piotroski and Roulstone (2004) who show that a high
ownership concentration enhances the relative amount of firm-specific
information in stock prices.
As substantial shareholders, they do have superior access to private
information and consequently, their trading can convey a timely and
accurate firm-specific information into stock prices. In contrast, our results
are not in line with Hossain et al. (2006) and Mak and Li (2001) who
indicate that outside block ownership has a significant and positive effect
on the level of disclosure, and this positive effect is reflected in current
stock returns. Their results imply that outside block ownership reduces
agency and informational risks and strengthens corporate governance.
Table.26 the joint effect of voluntary risk disclosure and ownership concentration on share price
anticipation of future earnings.

CHAPTER VIII

CHAPTER VIII

CHAPTER VIII

CHAPTER VIII

CHAPTER VIII
Model
Intercept
Xt
Xt+1
Xt+2
Xt+3
Rt+1
Rt+2
Rt+3
AGt
Ept-1
RD
RD*Xt
RD*Xt+1
RD*Xt+2
RD*Xt+3
RD*Rt+1
RD*Rt+2
RD*Rt+3
RD*AGt
RD*Ept-1
OwC
OwC *Xt
OwC *Xt+1
OwC *Xt+2
OwC *Xt+3
OwC *Rt+1
OwC *Rt+2
OwC *Rt+3
OwC *AGt
OwC *Ept-1
OwC *RD*Xt
OwC *RD*Xt+1
OwC *RD*Xt+2
OwC *RD*Xt+3
OwC *RD*Rt+1
OwC *RD*Rt+2
OwC *RD*Rt+3
OwC *RD*AGt
OwC *RD*EPt-1
Year Dummies
Adj R2
Observations

Exp.
Sign
(?)
(+)
(+)
(+)
(+)
()
()
()
()
(+)
(?)
(?)
(+)
(+)
(+)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
()
()
()
(?)
(?)
(?)
(?)
(?)

Financial
Leverage
0.143

(0.63

9)
-8.065*
(0.075)
-6.625
(0.142)
-4.427
(0.366)
4.717
(0.357)
-1.742*** (0.000)
-0.511
(0.378)
-0.212
(0.726)
0.287
(0.792)
-0.200
(0.224)
0.064
(0.106)
0.203
(0.156)
0.182
(0.198)
0.365*** (0.008)
0.100
(0.528)
0.080
(0.145)
0.076
(0.302)
0.034
(0.344)
0.014*** (0.001)
0.001
(0.417)
-0.002
(0.973)
2.109*
(0.053)
2.348**
(0.032)
1.281
(0.276)
-0.832
(0.511)
0.269**
(0.014)
0.010
(0.941)
-0.041
(0.778)
-0.052
(0.844)
0.038
(0.290)
-0.020
(0.313)
-0.062*** (0.001)
-0.054*** (0.002)
-0.011
(0.421)
-0.018** (0.033)
-0.023*
(0.061)
-0.000
(0.905)
-0.000** (0.012)
-0.000** (0.018)
Yes
39.33%
785

Firm
Size
0.052

Profitability
(0.866) 0.225

-4.948
(0.272)
-3.489
(0.473)
-3.008
(0.543)
2.693
(0.594)
-2.137
(0.000)
-0.898
(0.097)
0.462
(0.478)
0.232
(0.830)
-0.107
(0.503)
0.070*
(0.061)
0.094
(0.471)
0.075
(0.618)
0.183
(0.131)
0.057
(0.718)
0.081
(0.165)
0.159
(0.017)
0.003
(0.968)
0.013*** (0.004)
0.000
(0.997)
0.016
(0.811)
1.450
(0.180)
1.650
(0.162)
0.964
(0.411)
-0.364
(0.769)
0.365*** (0.001)
0.087
(0.496)
-0.196
(0.200)
-0.033
(0.900)
0.022
(0.511)
-0.007
(0.669)
-0.039** (0.037)
-0.013
(0.411)
-0.009
(0.556)
-0.023** (0.018)
-0.043*** (0.000)
0.006
(0.675)
-0.000
(0.107)
-0.000
(0.315)
Yes
42.73%
785

(0.443)

-4.523
(0.288)
-5.572
(0.192)
-4.292
(0.344)
6.623
(0.139)
-1.787*** (0.000)
-0.874*
(0.094)
-0.048
(0.929)
-0.126
(0.907)
-0.099
(0.520)
0.067*
(0.072)
0.058
(0.653)
-0.036
(0.777)
0.233*
(0.100)
0.154
(0.307)
0.105**
(0.031)
0.108
(0.114)
0.037
(0.295)
0.010**
(0.015)
0.000
(0.602)
0.003
(0.959)
1.419
(0.170)
2.137**
(0.041)
1.233
(0.251)
-1.391
(0.215)
0.250*** (0.004)
0.120
(0.336)
-0.063
(0.625)
0.053
(0.837)
0.014
(0.674)
-0.017
(0.230)
-0.047** (0.022)
-0.035** (0.041)
-0.036*
(0.062)
-0.032*** (0.000)
-0.033*** (0.008)
-0.002
(0.654)
-0.000
(0.186)
-0.000
(0.240)
Yes
43.97%
785

Table 26 presents fixed effects regression results for panel data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period
return, Rt. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of
dividends) over a 12-month period, starting six months after the end of the previous
financial year. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both
current and future earnings changes are deflated by the price at the start of the return
window for period t. EPt1 is defined as period t1s earnings over price six months after

CHAPTER VIII
the financial year-end of period t1. AGt is the growth rate of the total book value of
assets for period t. RD is the natural logarithm of the total number of risk related
sentences, respectively, for Operations risk, Empowerment risk, Information processing
and technology risk, Integrity risk and Strategic risk information. OwC is the proportion of
equity held by the top-five largest shareholders. Financial leverage, firm size and
profitability are control variables (their regression estimates are not tabulated). The
significance levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

2.1.2. Managerial ownership, risk disclosure and the return future earnings relationship
Table 27 provides panel regression estimates of the joint impact of risk
disclosure and the level of managerial ownership on investors ability to
forecast future earnings change. It highlights the main findings of H3
testing that predicts a weaker share price anticipation of revealed future
earnings when interacted with the level of management ownership.
The coefficient on Xt is, while positive, not significantly different from zero
for the three regressions with each controlling variable. Current earnings
do not reflect investors early expectations about future earnings. There is
besides strong evidence on price leading earnings for one year ahead
regardless of the extent of corporate risk disclosure and the level of
management ownership. The future earnings response coefficients on Xt+1
are positive and significant at the level of 1%. They range from 2.552 to
3.043 and are substantially higher than the coefficients on Xt. These
results emphasize investors ability to impound information about future
earnings of the period t+1 into stock prices. There is no evidence on share
price anticipation of future earnings for t+2 and t+3 periods. The
coefficients on Rt+1, Rt+2 and Rt+3 exhibit the expected sign and are
significant in most cases at the levels of 1% and 5%.
Results in table 27 show a positive relation between current returns and
i, t

the extent of voluntary risk disclosure. The coefficients on RD

are

significant at the 10% level, mainly when we control for corporate size and
profitability. As stated earlier, risk disclosure activity of our sample of
MENA firms is driven by corporate financial performance. There is also
some evidence of share price anticipation of future earnings driven by the
i, t

level of corporate risk disclosure. The coefficients on RD *Xt+2 are positive

CHAPTER VIII
and significant at the levels of 5% and 10 % in the presence of financial
leverage and firm size variables.
These results suggest that risk disclosure convey relevant information
about corporate prospects which are impounded into stock prices.
Management ownership is negatively and significantly associated with
current returns at the level of 5%. This suggests that the higher is
managements stake in firm equity the lesser is firm performance.
Now we focus on our hypothesis testing H3 which addresses the
simultaneous effects of risk disclosure and managerial ownership on share
price anticipation of future earnings. The incremental predictive value of
both variables for anticipating future earnings is given by the coefficients
3

i, t

i, t

on MOw *RD *
i, t

X t+K

K =1

. The coefficient on the interaction term MOw

i, t

*RD *Xt+1 are negative and significantly different from zero at the levels
of 5% in all specifications. We notice also a negative and significant
i, t

i, t

coefficient on the interaction term MOw *RD *Xt+3 at the level of 5%


when we control for corporate profitability. Management ownership seems
to weaken the association between current earnings and investors ability
to anticipate revealed future earnings for one year and three years ahead.
Investors do believe that managers are most likely to withhold or manage
firm risk information and divert private benefits from outside investors.
They consider that owner-managers and directors make value reducing
decisions in order to safeguard their positions in the firm, therefore
expropriating wealth from outsiders and decreasing the quality of
information reported. Market participants react adversely by questioning
the credibility of conveyed risk information and prefer to rely on other
sources of information to predict future earnings growth other than those
driven by an increased managerial ownership. These findings corroborate
our third hypothesis (H3) and are consistent with Karamanou and Vafeas
(2005), Hossain et al. (2006) and Yeo et al. (2002) who found a significant

CHAPTER VIII
negative relationship between managerial ownership and the quality of
financial disclosure practice using US and Singaporean data. Our results
are also in line with those of Wang and Hussainey (2013) which indicate that
inside owners are more likely to have advance access to forward-looking information, hence
reducing the firm managements incentive to disclose such information voluntarily in annual
reports.

Table 27 the joint effect of voluntary risk disclosure and managerial ownership
on share price anticipation of future earnings.

Model

Exp

Financial

Firm

Leverage

Size

Profitability

CHAPTER VIII

Sig
Intercept

n
(?)

0.194

(0.12

0.180

(0.14

0.276**

(0.02

Xt

(+)

0.542

7)
(0.55

0.857

6)
(0.33

1.114

6)
(0.19

Xt+1

(+)

2.494**

8)
(0.00

2.868**

0)
(0.00

3.043**

9)
(0.00

Xt+2

(+)

*
0.263

9)
(0.77

*
0.309

3)
(0.73

*
0.080

1)
(0.92

1.070

4)
(0.21

0.599

7)
(0.44

Xt+3

(+)

1.081

9)
(0.25

Rt+1

()

8)
(0.00

0)
(0.00

9)
(0.00

0.743**

0)

0.652**

0)

0.760**

0)

*
-

(0.00

*
-

(0.01

*
-

(0.02

0.400**

9)

0.370**

8)

0.322**

2)

(0.02

(0.03

-0.208

(0.19

0.347**
0.120

3)
(0.42

0.104

6)
(0.48

-0.072*

4)
(0.09

Rt+2

()

Rt+3

()

*
-

AGt

()

0.375**
0.100

6)
(0.51

-0.056

2)
(0.23

Ept-1

(+)

-0.074*

5)
(0.09

RD

(?)

0.062

0)
(0.10

0.073**

9)
(0.05

0.066*

1)
(0.07

RD*Xt

(?)

0.189*

9)
(0.06

0.131

0)
(0.15

0.021

5)
(0.86

-0.002

4)
(0.97

-0.165

2)
(0.12

RD*Xt+1

(+)

0.041

1)
(0.63

RD*Xt+2

(+)

0.255**

8)
(0.03

0.215*

8)
(0.08

0.186

5)
(0.14

RD*Xt+3

(+)

0.195

5)
(0.13

0.132

0)
(0.28

0.204

5)
(0.13

-0.023

7)
(0.47

-0.035

7)
(0.23

-0.057

4)
(0.14

-0.036

7)
(0.34

-0.018

2)
(0.67

RD*Rt+1

(?)

0.013

8)
(0.68

RD*Rt+2

(?)

-0.037

0)
(0.33

0.013

4)
(0.70

RD*Rt+3

(?)

0.015

8)
(0.65

RD*AGt

(?)

0.011**

9)
(0.00

0.010**

7)
(0.00

0.008**

0)
(0.04

RD*Ept-1

(?)

*
-0.000

3)
(0.82

*
-0.000

3)
(0.68

-0.000

1)
(0.91

1)
(0.02

-0.002

9)
(0.15

0.003**
-0.015

9)
(0.63

-0.012

1)
(0.70

MOw

(?)

2)
(0.03

MOw*Xt

(?)

0.003**
-0.007

5)
(0.81

CHAPTER VIII
Table 27 presents fixed effect regression results for panel data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period
return, Rt. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of
dividends) over a 12-month period, starting six months after the end of the previous
financial year. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both
current and future earnings changes are deflated by the price at the start of the return
window for period t. EPt1 is defined as period t1s earnings over price six months after
the financial year-end of period t1. AGt is the growth rate of the total book value of
assets for period t.
RD is the natural logarithm of the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity
risk and Strategic risk information. MOw is the proportion of equity held by managers and
executive directors. Financial leverage, firm size and profitability are control variables
(their regression estimates are not tabulated). The significance levels (two-tail test) are:
*= 10 %, ** =5 % and *** = 1 %.

2.1.3. Institutional ownership, risk disclosure and the return future earnings relationship
Table 28 reports main findings regarding the fourth hypothesis (H4) testing
that predicts a stronger association between voluntary risk disclosure and
share price anticipation of future earnings for firms with high institutional
ownership.
The coefficients on Xt are, despite the positive sign, insignificant in the two
specifications. This coefficient became significant at the level of 10% when
we control for corporate profitability. In the same way the coefficients on
future earnings change (Xt+1, Xt+2, Xt+3) are insignificant in all cases except
in the regression specification which controls for firm profitability. In
particular, the coefficient on Xt+1 is positive and significant at the level of
10%. This provides a weak evidence of price leading for one year ahead
regardless of corporate risk disclosure and the level of institutional
ownership mainly for highly profitable firms. We did not notice such
evidence for big sized and risky corporations. The coefficients on future
stock returns Rt+1, Rt+2 and Rt+3 have the expected sign and are in most
cases significantly different from zero. These findings indicate that future
returns are a good measurement error proxy for the unexpected portion of
realized future earnings.
Results in table 28 show a positive and significant association between
current returns and the extent of firm risk disclosure at the level of 10%
and 5% which is consistent with our earlier findings. Risk disclosure seems
to influence positively the information content of current earnings in that

CHAPTER VIII
i, t

the coefficient on the interaction term RD *Xt is significantly different


from zero at the levels of 1% and 5%. Risk disclosure seems to influence
likewise the informativeness of stock prices with respect future earnings
i, t

for two years ahead. The coefficients on the interaction term RD

*Xt+2 are

positive and significant in all the models at the levels of 1% and 5%. This
suggests that corporate risk reporting convey value relevant information
to market participants that are useful for anticipating future earnings
growth. Share price anticipation of future earnings appears to be driven by
the presence of institutional owners. The coefficients on the interacted
i, t

variables InOw *Xt+1 are positive and significant at the levels of 1% and
5%.
These findings may be attributed to the informational advantage of
institutional investors. They are acting as informed traders, and their
trading activities are conducive for integrating firm-specific information
into stock prices. These results are in line with Jiambalvo et al. (2002) who
found that for firms with higher levels of institutional ownership
relatively more future earnings information is impounded in stock
prices in comparison to firms with lower institutional ownership. It is
also believed that firms with institutional owners are likely to engage in
less opportunistic earnings management, hence improving the
credibility of financial information and future earnings forecast.
We turn to our hypothesis testing H4 which measures the simultaneous
effects of risk disclosure and institutional ownership on share price
informativeness with respect future earnings. The incremental predictive
value of both variables for anticipating future earnings growth is given by
3

i, t

i, t

the coefficients on interacted variables InOw *RD *

X t+K

K =1

. Findings

in table 27 reject our fourth hypothesis in that it indicates that the value
relevance of risk disclosure is a decreasing function of the level of

CHAPTER VIII
i, t

i, t

i, t

institutional ownership. The coefficients on InOw *RD *Xt+1, InOw *RD


i, t

i, t

i, t

*Xt+2 and InOw *RD *Xt+3 are negative and significantly different from
zero at the levels of 1%, 5% and 10%. These results provide evidence on
the lower informativeness of current returns with respect to revealed
future earnings. While agency theory suggests that institutional
shareholders fulfil a monitoring function, this does not appear to include
pressuring firms to increase relevant risk disclosure. This weakened
association occurs because investors believe that institutional owners are
sophisticated and have advantages in acquiring and processing
information. Such information is reflected in future earnings regardless of
the extent of risk information conveyed to the market. This situation yields
a less rich information environment and a difficult transfer of information
as long as these investors' ownership increases. Consequently, market
participants seem to question the usefulness of conveyed risk information
driven by a high concentration of institutional ownership. Our findings are
consistent with Abraham and Cox (2007) argument and evidence that
these institutions can benefit from a mosaic of non-public non-material
information and reveal a preference for firms with a lower than average
level of risk reporting. Our results are in contrast inconsistent with Wang
and Hussainey (2013) who reported an insignificant relationship between
institutional investors and the level of voluntary disclosure regarding
future earnings. They suggest that, as powerful investors, institutional
owners might have other more efficient means of communicating with the
firms management, for example, one-to-one meetings.
Table .28 the joint effect of voluntary risk disclosure and institutional ownership on share price
anticipation of future earnings.

Model

Exp

Financial

Firm

Leverage

Size

Profitability

CHAPTER VIII

Sig
Intercept

n
(?)

0.151

(0.24

0.109

(0.38

0.246

(0.05

Xt

(+)

0.457

5)
(0.65

0.725

1)
(0.42

1.540*

3)
(0.09

Xt+1

(+)

1.458

5)
(0.16

1.640

9)
(0.11

1.723*

6)
(0.08

Xt+2

(+)

0.119

8)
(0.90

0.198

8)
(0.84

0.151

5)
(0.86

0.554

0)
(0.55

0.097

9)
(0.91

Xt+3

(+)

0.633

2)
(0.53

Rt+1

()

8)
(0.00

9)
(0.00

4)
(0.00

0.674**

0)

0.618**

0)

0.761***

0)

*
-

(0.00

*
-

(0.00

-0.

(0.00

0.458**

2)

0.419**

5)

382***

6)

(0.00

*
-

(0.02

-0.237

(0.10

0.347**
0.041

5)
(0.81

0.042

8)
(0.79

-0.030

6)
(0.46

Rt+2

()

Rt+3

()

*
-

AGt

()

0.415**
0.029

9)
(0.86

-0.04

0)
(0.93

Ept-1

(+)

-0.026

5)
(0.55

RD

(?)

0.064*

9)
(0.09

0.073**

0)
(0.04

0. 067*

3)
(0.06

RD*Xt

(?)

0.307**

2)
(0.00

0.227**

3)
(0.02

0.124

9)
(0.32

RD*Xt+1

(+)

*
0.208*

8)
(0.07

0.139

5)
(0.19

0.051

7)
(0.70

RD*Xt+2

(+)

0.331**

2)
(0.00

0.292**

3)
(0.00

0.285**

7)
(0.01

RD*Xt+3

(+)

*
0.147

1)
(0.22

*
0.162

5)
(0.15

0.105

7)
(0.42

0.008

0)
(0.82

0.002

8)
(0.94

-0.034

5)
(0.33

-0.009

3)
(0.77

-0.000

9)
(0.98

RD*Rt+1

(?)

0.032

2)
(0.33

RD*Rt+2

(?)

-0.020

8)
(0.56

0.021

3)
(0.49

RD*Rt+3

(?)

0.029

6)
(0.35

RD*AGt

(?)

0.009**

3)
(0.04

0.009**

6)
(0.02

0.007*

6)
(0.08

RD*Ept-1

(?)

-0.000

3)
(0.84

-0.000

5)
(0.86

0.001

9)
(0.36

-0.000

8)
(0.80

-0.000

9)
(0.79

0.026

8)
(0.49

0.003

3)
(0.92

InOw

(?)

-0.001

3)
(0.59

InOw*Xt

(?)

0.029

1)
(0.53

CHAPTER VIII
Table 28 presents fixed effects regression results for panel data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period
return, Rt. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of
dividends) over a 12-month period, starting six months after the end of the previous
financial year. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both
current and future earnings changes are deflated by the price at the start of the return
window for period t. EPt1 is defined as period t1s earnings over price six months after
the financial year-end of period t1. AGt is the growth rate of the total book value of
assets for period t.
RD is the natural logarithm of the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity
risk and Strategic risk information. InOw is the total number of shares held by institutions
to the total number of shares outstanding per sample firm. Financial leverage, firm size
and profitability are control variables (their regression estimates are not tabulated). The
significance levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

To summarize, Ownership structure as examined through the level of


ownership concentration, managerial ownership and institutional investors
decrease the association between risk disclosure and share price
anticipation of future earnings. These findings corroborate our second and
third hypotheses and reject our fourth assumption. This may be explained
by the fact that ownership in our sample firms is on average highly
concentrated (the mean is 61.80%), with lower presence of institutional
investors (the mean is 13.29%) and inside owners (13.37%). In such
situation corporation tend to issue less relevant risk disclosure as block
holders do not have to rely on external disclosures to the same extent as
companies with dispersed ownership, management interest are not
aligned with shareholders and the low presence of institutional owners
may prevent them from fulfilling their monitoring role.
2.2. Board characteristics, risk disclosure and share price anticipation of future earnings.

We basically address three corporate board characteristics. We


hypothesized that the amount of non-executive directors, board size and
the CEO/Chairman duality are governance mechanisms that may impact
(increase/decrease) the association between risk disclosure and share
price informativeness with respect to future earnings. We provide
coefficients estimates based on pooled OLS regression.

CHAPTER VIII
2.2.1. Non-executive directors, risk disclosure and share price anticipation of future
earnings
Table 29 covers main regression findings for our fifth hypothesis. It
provides estimates of the joint impact of the proportion of non-executive
directors sitting on the board and the level of risk disclosure on the return
future earnings relationship.
The coefficients on Xt are negative and insignificant in the three
specifications. In the same way the coefficients on future earnings change
(Xt+1, Xt+2, Xt+3), while mostly positive, are insignificant in all cases. Current
returns are less informative about future earnings regardless of corporate
risk disclosure and the proportion of non-executive directors. The
coefficients on future stock returns Rt+1 have the expected sign and are
significantly different from zero at the level of 5%.
The coefficients on Rt+2 and Rt+3 are mostly insignificant despite the
negative sign. These findings suggest that future returns are a good
measurement error proxy for the unexpected portion of realized future
earnings.
Results in Table 29 show a positive and strong association between current
returns and the extent of firm risk disclosure at the level of 1% which is
consistent with our earlier findings. Risk disclosure does not seem to
impact share price informativeness with respect current and future
i, t

i, t

earnings. The coefficients on the interaction terms RD *Xt and RD *


3

X t+k
k=1

i, t

are positive (except for RD *Xt+1) but insignificant. The


i, t

coefficient on RD *Rt+1 is positive and significant at the levels of 10% and


5%, mainly when controlling for financial leverage and firm size. The
i, t

i, t

partial F-test of the joint significance of RD *Xt+1 and RD *Rt+1 is only

CHAPTER VIII
significant in the specification model that controls for corporate size at the
level of 10%.
Current return is positively associated with the proportion of non-executive
directors sitting on the board at the end of every year. The coefficients on
i, t

NED

are significant at the levels of 10% and 5%. This suggests that the

stock price performance is an increasing function of the amount of


independent members in the boards. The value relevance of current
earnings and share price anticipation of future earnings is not driven by
the proportion of non-executive directors. The coefficients of the
i, t

interaction terms on NED *Xt are insignificant in most cases, while the
i, t

coefficients on NED *Xt+1 are negative and significant at the level of 1%


particularly when we control for corporate risk and size. These results
indicate that board independence is not important in enriching the
information content of stock prices.
3

i, t

i, t

The coefficients of interest are NED *RD *

X t+k
k=1

. These terms

emphasize how the amount of non-executive directors interacts with


corporate risk disclosure in influencing the return future earnings
association. In support of hypothesis 5, the coefficients on the interaction
i, t

i, t

term NED *RD *Xt+1 are significantly positive at the levels of 1% and
5%. Our findings suggest that outside directors sitting on the board
enhance the value relevance of risk disclosure in that their interactions
impact positively share price anticipation of future earnings for one year
ahead. There is no evidence of such impact for t+2 and t+3 periods. It is
i, t

i, t

worth noting also that the coefficient on NED *RD *Xt is significantly
positive at the levels of 10% and 1% particularly for the specification
models that control for financial leverage and firm size. Accordingly, a high

CHAPTER VIII
number of non-executive directors provides investors with assurance over
annual report narratives.
They are believed to be more influential in terms of the board decision
making and particularly in enhancing the quality of corporate voluntary
disclosure. Risk reporting is considered hence as a credible source of
information for investors to predict future earnings growth. These findings
corroborate our fifth hypothesis and are consistent with Wang and
Hussainey (2013). They showed that higher proportions of non-executive
directors are more likely to disclose relevant information related to future
earnings, suggesting that greater financial reporting expertise exists on
such boards. Our results are whereas inconsistent with Yu (2011) findings,
which indicate that the information about expected future earnings growth
do not appear to be effectively incorporated into the stock price for firms
with optimal board structure (e.g. Board independence, board size, role
duality). The insignificant association suggests that the quality of the
board is a form of internal oversight that does not urge managers to offer
information on a voluntary basis.
Table 29 The joint effect of non-executive directors, risk disclosure and share
price anticipation of future earnings

Model

Exp.

Financial

Firm

Leverage

Size

Profitability

CHAPTER VIII

Intercept

Sign
(?)

Xt

(+)

0.278*** 4
-1.309
0.15

Xt+1

(+)

0.888

3
0.31

0.00

0.00

0.285*** 4
-1.187
0.19
0.633

9
0.47

0.00

0.323*** 1
-1.308
0.15
0.543

6
0.55

0.127

3
0.89

0.196

8
0.83

Xt+2

(+)

0.119

5
0.89

Xt+3

(+)

-0.741

9
0.48

-0.363

1
0.72

-0.527

2
0.61

Rt+1

()

-0.381**

2
0.02

-0.404**

9
0.01

-0.386**

8
0.02

0.242

6
0.15

0.217

2
0.20

Rt+2

()

0.320*

3
0.06

Rt+3

()

-0.017

1
0.95

-0.123

3
0.54

-0.020

2
0.92

AGt

()

0.067

2
0.11

0.034

3
0.46

0.062

7
0.12

Ept-1

(+)

0.007

2
0.28

0.005

5
0.53

0.005

9
0.49

RD

(?)

9
0.066*** 0.00

2
0.070*** 0.00

8
0.064*** 0.00

RD*Xt

(?)

0.355

6
0.34

0.091

4
0.81

0.440

8
0.27

RD*Xt+1

(+)

-0.287

0
0.40

-0.471

4
0.20

-0.128

9
0.73

RD*Xt+2

(+)

0.424

7
0.21

0.189

2
0.60

0.411

1
0.24

0.294

7
0.51

0.447

7
0.29

RD*Xt+3

(+)

0.470

9
0.25

RD*Rt+1

(?)

0.097*

8
0.09

0.137**

3
0.02

0.102

1
0.10

RD*Rt+2

(?)

-0.087

4
0.16

-0.026

7
0.68

-0.104

5
0.11

RD*Rt+3

(?)

0.031

6
0.75

0.025

5
0.75

0.022

7
0.79

RD*AGt

(?)

-0.027

4
0.15

-0.014

7
0.40

-0.034

3
0.17

-0.001

9
0.56

-0.000

3
0.66

RD*Ept-1

(?)

-0.001

0
0.38

NED

(?)

0.001*

6
0.05

0.001**

1
0.04

0.001**

7
0.04

NED*Xt

(?)

0.001

1
0.76

-0.008*

6
0.06

0.003

7
0.50

5
0.00

9
0.00

-0.006

2
0.20

0.011*** 0
-0.004
0.13

-0.003

7
0.43

NED*Xt+1

(?)

NED*Xt+2

(?)

0.010*** 0
-0.003
0.34

CHAPTER VIII
Table 29 presents OLS regression results for pooled data. Heteroscedasticity-consistent pvalues are given in parentheses. The dependent variable is current period return, Rt. Rt,
Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12month period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2
and Xt+3 are defined as earnings change deflated by price. Both current and future
earnings changes are deflated by the price at the start of the return window for period t.
EPt1 is defined as period t1s earnings over price six months after the financial year-end
of period t1. AGt is the growth rate of the total book value of assets for period t. RD is
the natural logarithm of the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity
risk and Strategic risk information. NED is the proportion of non-executive directors
relative to the board size. Financial leverage, firm size and profitability are control
variables (their regression estimates are not tabulated). The significance levels (two-tail
test) are: *= 10 %, ** =5 % and *** = 1 %.

2.2.2. Board size, risk disclosure and the return future earnings relationship
In table 30 we present multiple regression estimates for our sixth
hypothesis which examine the simultaneous effect of board size and risk
disclosure on the return future earnings relationship. The coefficient on Xt
is insignificant in the presence of leverage and firm size while it is
negatively and weakly associated with the current return when we control
for profitability. Similarly, the future earnings response coefficients on Xt+1
and Xt+2 are mostly insignificant despite the right sign. There is a weak
evidence of price leading earnings for one year ahead when controlling for
corporate profitability. Results show also a strong and positive association
between current return and future earnings of period t+3. The coefficients
on Xt+3 extend from 9.423 to 13.063 and are significant at the levels of 5%
and 1%. There is accordingly strong evidence of share price anticipation of
future earnings by three years ahead regardless of the extent of risk
information and board size. The level of risk disclosure is positively related
i, t

to stock returns. The coefficients on RD

are significantly different from

zero at the level of 1%. Consistent with earlier findings, firms with higher
stock returns issue more voluntary risk disclosure. The coefficients on the
i, t

interaction terms RD *Xt is positive and significant at the levels of 10%


and 5%, except for the specification model that control for financial
leverage.

CHAPTER VIII
This suggests that risk information improves the value relevance of current
3

i, t

earnings. The coefficients on RD *

X t+k
k=1

exhibit a negative sign and


i, t

are mostly insignificant except for the interaction term RD *Xt+3 which are
significantly different from zero at the levels of 5% and 1%. These findings
indicate that risk information which are not driven by the number of board
directors appear to reduce the share price forecast of future earnings. The
i, t

coefficients on BS

are negative, but insignificant except for the

regression model that control for financial leverage. There is hence weak
evidence that board size lower equity cost (i.e., share return). Information
about board size reduces investors ability to anticipate future earnings
i, t

growth given the negative and significant coefficients on BS *Xt+3 at the


levels of 1% and 5%.
Estimates of primary interest are given by the coefficients on the
3

i, t

i, t

interaction terms BS *R *

X t+k
k=1

. Consistent with our prior

expectation, results revealed a positive and significant effect of risk


disclosure driven by board size on the markets ability to anticipate future
earnings growth for three years ahead. Such results indicate that firms
with large board size are more likely to disclose significantly useful
voluntary information about corporate opportunities and risks. Investors
seem to trust risk information conveyed by firms with a large board size as
it reflects a diverse expertise and is likely to guarantee a higher level of
compliance with the accountability paradigm. These findings, therefore,
support our hypothesis 6. They are consistent with Wang and Hussainey
(2013) who showed that large board size is associated with more relevant
information about future earnings in corporate annual reports. On the
contrary, our results are inconsistent with Ferrreira et al. (2011) and Yu

CHAPTER VIII
(2011) who failed to provide evidence that stock prices are more
informative about future earnings growth in firms with a large board size.
Their studies suggest that firms with more informative stock prices have
less demanding board structures.

Table 30. The joint effect of board size, risk disclosure and share price
anticipation of future earnings

Model

Exp

Financial

Firm

Leverage

Size

Profitability

CHAPTER VIII
Sig
Intercept
Xt

n
(?)
(+)

-0.093
-4.297

0.373
0.260

-0.088
-4.892

0.398
0.191

-0.154
-6.449*

0.160
0.09

Xt+1

(+)

5.231

0.200

4.270

0.292

7.549*

6
0.07

Xt+2
Xt+3

(+)
(+)

3.750
10.780*

0.351
0.013

4.350
9.423**

0.275
0.032

5.209
13.063*

2
0.193
0.00

Rt+1
Rt+2

()
()

*
0.128
-0.543

0.806
0.223

0.162
-0.404

0.755
0.368

**
0.078
-0.889*

4
0.881
0.06

Rt+3
AGt
Ept-1
RD

()
()
(+)
(?)

0.104
-0.130
0.032
0.081**

0.877
0.321
0.383
0.001

-0.083
-0.057
0.027
0.079***

0.901
0.655
0.462
0.002

0.179
-0.021
0.038
0.071**

5
0.791
0.86
0.291
0.00

RD*Xt

(?)

*
1.837

0.119

1.952*

0.090

*
2.430**

5
0.03

RD*Xt+1
RD*Xt+2
RD*Xt+3

(+)
(+)
(+)

-1.266
-1.185
-

0.339
0.362
0.015

-1.106
-1.620
-3.106**

0.404
0.211
0.032

-1.717
-1.611
-

7
0.207
0.213
0.00

3.941**

*
-0.131
0.200
-0.057
0.010
-0.010
-0.013
0.573
-0.965*

0.413
0.183
0.789
0.825
0.285
0.110
0.197
0.06

-0.619
-

3
0.214
0.00

1.451**

1.662**

0.482
0.088

-0.044
0.091

0.467
0.125

*
-0.039
0.134**

0.519
0.02

3.475**
RD*Rt+1
RD*Rt+2
RD*Rt+3
RD*AGt
RD*Ept-1
BS
BS*Xt
BS*Xt+1
BS*Xt+2
BS*Xt+3

(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)
(?)

-0.143
0.123
-0.047
0.047
-0.008
-0.014*
0.274
-0.802
-0.494
-

0.375
0.408
0.824
0.348
0.356
0.096
0.538
0.117
0.326
0.006

-0.128
0.132
-0.009
0.025
-0.007
-0.014
0.304
-0.681
-0.564
-1.297**

0.424
0.373
0.966
0.606
0.430
0.102
0.485
0.179
0.257
0.015

BS*Rt+1
BS*Rt+2

(?)
(?)

*
-0.043
0.101*

BS*Rt+3
BS*AGt
BS*Ept-1
BS*RD*Xt
BS*RD*Xt+

(?)
(?)
(?)
(?)
(+)

-0.040
0.019
-0.002
-0.106
0.244

0.612
0.271
0.413
0.434
0.142

-0.020
0.014
-0.001
-0.126
0.216

0.791
0.393
0.511
0.340
0.194

-0.038
0.007
-0.002
-0.183
0.289*

9
0.628
0.651
0.309
0.171
0.09

(+)

0.174

0.282

0.198

0.216

0.226

1
0.160

(+)

0.480**

0.006

0.440**

0.012

0.534**

0.00

BS*RD*Xt+
2

BS*RD*Xt+

CHAPTER VIII
Table 30 presents OLS regression results for pooled data. Heteroscedasticity-consistent pvalues are given in parentheses. The dependent variable is current period return, Rt. Rt,
Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12month period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2
and Xt+3 are defined as earnings change deflated by price. Both current and future
earnings changes are deflated by the price at the start of the return window for period t.
EPt1 is defined as period t1s earnings over price six months after the financial year-end
of period t1. AGt is the growth rate of the total book value of assets for period t.
RD is the natural logarithm of the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity
risk and Strategic risk information. BS is the number of directors sitting on the board at
the end of each year. Financial leverage, firm size and profitability are control variables
(their regression estimates are not tabulated). The significance levels (two-tail test) are:
*= 10 %, ** =5 % and *** = 1 %.

2.2.3. CEO/Chairman duality, risk disclosure and the return future earnings relationship
Table 31 summarizes main findings regarding our hypothesis 7 testing. We
predicted a weaker association between voluntary risk disclosure and
share price anticipation of future earnings for firms with CEO/Chairman
duality.
The coefficients on Xt are negatively significant at the level of 1%. Current
earnings growth does not serve as a good estimation for prior anticipation
of future earnings growth. There is also no evidence of price leading
3

X t+k

earnings given the insignificant coefficients on

k=1

. Similar to prior

results risk disclosure is positively and significantly associated with current


returns at the levels of 5% and 1%.
Risk disclosure enhances the value relevance of current earnings in that
i, t

the coefficients on the interacted variables RD *Xt are positive and


significantly different from zero at the level of 1%. There is, in addition,
evidence on improved share price informativeness with respect future
i, t

earnings of period t+1 given the significant coefficients on RD *Xt+1. The


i, t

coefficients on the interaction term RD *Rt+3 are likewise positive and


significant at the levels of 5% and 10% and the joint significance test for

CHAPTER VIII
i, t

i, t

the slope coefficients on RD *Xt+3 and RD *Rt+3 yielded a significant F


statistic at the level of 5%. These findings suggest that risk reporting
activity which is not related to the presence of the CEO/chairperson duality
is informative about future earnings growth. The presence of a dominant
person in the board of directors as measured by the role duality (DUAL)
variable affects positively shareholders perception of earnings reliability.
The coefficients on DUAL*Xt are significant at the levels of 1% and 5%.
Investors believe that combining the two positions of chairman and CEO
facilitates the decision-making process and increase firm performance.
Now we turn to our hypothesis H7 which looks to find out the joint effects
of risk disclosure and CEO/Chairman duality on share price informativeness
with respect to future earnings. The incremental predictive value of both
variables for anticipating future earnings growth is given by the
3

i, t

coefficients on interacted variables DUAL*RD *

X t+K

K =1

Findings rejected our hypothesis H7 in that the coefficients on the


interaction terms are insignificant in all the specifications even though
they exhibit the predicted sign. There is hence no evidence that risk
disclosure driven by role duality reduces share price anticipation of future
earnings.
Risk reporting related to role duality impacts negatively and significantly
the credibility of current earnings at the level of 1%. Firms with the same
person occupying the roles of chairman and CEO have then significant
credibility problems with their financial reports and in particular with
information about corporate risk. The stock market places lower credibility
on such information, presumably because the CEO/chairperson is likely to
compromise the board independence and to seek private interests.
Thereby, this situation can lead to conflicts of interest that encompass the
accounting system and the voluntary release of useful information. Our
results are in line with Yu (2011) findings, which indicate that there is no

CHAPTER VIII
effect of the board leadership structure (as examined through board
quality index) on the earnings information content of stock prices. They
are also consistent with some recent voluntary and risk disclosure
literature such as Cheng and Courtenay (2006), Elshandidy et al (2013)
Elzahar and Hussainey (2012) and Mokhtar and Mellet (2013) which failed
to confirm that role duality is a potential threat to disclosure quality. These
results in turn do support neither agency theory nor the signaling theory.
Table 31 The joint effect of duality and risk disclosure on share price
anticipation of future earnings

Model

Exp

Financial

Firm

Leverage

Size

Profitability

CHAPTER VIII
Sig
Intercept

n
(?)

0.04

0.02

-.213***

0.01

Xt

(+)

0.158**
-

1
0.00

0.173**
-

8
0.00

0
0.00

3.871**

3.773**

3.485**

*
-1.412

0.28

*
1.603

0.22

*
-0.768

0.55

-0.555

0
0.66

-0.854

6
0.51

Xt+1

(+)

Xt+2

(+)

-1.179

7
0.34

Xt+3

(+)

-1.764

7
0.16

-1.692

1
0.18

-1.317

2
0.32

Rt+1

()

-0.240

2
0.15

-0.187

9
0.27

-0.259

3
0.13

0.214

6
0.16

0.067

0
0.67

-0.414*

8
0.05

-0.387*

6
0.06

0.043

6
0.17

Rt+2

()

0.128

6
0.38

Rt+3

()

9
0.02

AGt

()

0.467**
0.015

9
0.51

0.057

1
0.16

Ept-1

(+)

0.000

2
0.96

0.004

3
0.83

0.000

9
0.97

RD

(?)

0.061**

6
0.01

0.066**

7
0.00

0.054**

6
0.02

*
1.492**

8
0.00

1.624**

7
0.00

RD*Xt

(?)

1.608**

3
0.00

RD*Xt+1

(+)

*
0.759*

0
0.07

*
0.793*

0
0.06

*
0.775*

0
0.05

RD*Xt+2

(+)

0.473

0
0.23

0.091

8
0.83

0.442

8
0.26

0.691

5
0.12

0.593

5
0.13

0.005

1
0.91

0.004

7
0.92

-0.049

6
0.32

RD*Xt+3

(+)

0.633

3
0.11

RD*Rt+1

(?)

0.005

1
0.91

-0.018

5
0.72

RD*Rt+2

(?)

-0.048

5
0.33

RD*Rt+3

(?)

0.176**

5
0.01

0.139*

4
0.05

0.161**

2
0.02

RD*AGt

(?)

-0.006

5
0.44

-0.018

8
0.13

-0.014

6
0.13

-0.001

3
0.75

-0.000

1
0.93

RD*Ept-1

(?)

-0.000

2
0.98

DUAL

(?)

0.026

2
0.55

0.024

4
0.57

0.023

8
0.58

DUAL*Xt

(?)

5.327**

1
0.00

4.320**

2
0.01

6.065**

9
0.00

CHAPTER VIII
Table 31 presents OLS regression results for pooled data. Heteroscedasticity-consistent pvalues are given in parentheses. The dependent variable is current period return, Rt. Rt,
Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of dividends) over a 12month period, starting six months after the end of the previous financial year. Xt, Xt+1, Xt+2
and Xt+3 are defined as earnings change deflated by price. Both current and future
earnings changes are deflated by the price at the start of the return window for period t.
EPt1 is defined as period t1s earnings over price six months after the financial year-end
of period t1. AGt is the growth rate of the total book value of assets for period t. RD is
the natural logarithm of the total number of risk related sentences, respectively, for
Operations risk, Empowerment risk, Information processing and technology risk, Integrity
risk and Strategic risk information. Dual is a dummy variable defined as 1 if CEO is the
Chairman and 0 otherwise. Financial leverage, firm size and profitability are control
variables (their regression estimates are not tabulated). The significance levels (two-tail
test) are: *= 10 %, ** =5 % and *** = 1 %.

In summary, our results evidenced that risk disclosure associated with


good board structure (large size and high proportion of non-executive
directors) enhances share price anticipation of future earnings. Investors
believe that board diversity increases members expertise, eliminates
environmental uncertainties and enhances the richness of information
environment. Market participants also think that non-executive directors
can restrain the managerial self-serving behavior and ensure useful risk
reporting in anticipating future earnings growth. Regarding board
leadership, risk disclosure driven by role duality does not seem to impact
stock price informativeness about future earnings. Even more, investors
seem to question the credibility of voluntary risk information and perceive
current earnings as less value relevant for firms characterized by the
presence of a dominant person.

2.3. The endogenous nature of corporate governance choice and value-relevance of


riskinformation

Recent academic research addressed an important question of whether


good corporate governance has a first order effect on some outcome
variables (e.g. Firm performance, firm valuation, financial reporting quality,
share price informativeness). According to Beyer et al. (2010) given the
endogenous nature of the corporate information environment, the firms
governance mechanisms, and some observed outcomes, it is hard to make

CHAPTER VIII
an assessment of the causal connection and recognize the exact impact
that one mechanism would have on another one.
Our main concern arises from the potential endogeneity problem between
share price informativeness and firm-level governance structure. The
causality might run from price informativeness to corporate governance
improvement, or other omitted firm characteristics could impact both
stock price informativeness and corporate governance. For example,
Durnev et al. (2004) suggest that information conveyed by share prices
can affect firm governance mechanisms as it signals to the capital markets
on the need to react when management decisions are inadequate. More
informative stock prices can serve as a disciplining mean in that it
enhances external monitoring mechanisms (shareholder lawsuits,
institutional investor pressure, and the market for corporate control) as
well as the internal monitoring role of the board (Ferreira et al. 2011;
Holmstrm and Tirole, 1993). If governance structure regressors are
inferred to be endogenously determined, failure to incorporate exogenous
determinants of the association between share price informativeness and
corporate governance choices will result in correlated omitted variables
problem. Standard OLS regression will provide inconsistent parameters
due to the correlated omitted variables problem.
The common econometric solution to endogeneity issue is the use of
instrumental variables specification procedure. Instrumental variables
should be associated with endogenous regressors but unrelated to the
error term in the structural equation (Habib and Azim 2008). Winship and
Morgan (1999) and Larker et al (2007) argue that while this approach is
theoretically strong, in practice, it is difficult to find an instrumental
variable that is correlated with assignment to treatment level but not the
outcome.

Larcker and Rusticus (2010) showed that OLS estimates provide better
parameter estimates compared to two-stage least square approach if the

CHAPTER VIII
selected instrumental variables do not comply with the standard definition
of instrumental variables. The selection of adequate instruments in the
present study is challenging because we use multiple endogenous
variables as measures of corporate governance, that is, ownership
concentration, institutional ownership, insider ownership, board size, board
composition and role duality. In addition, since our board structure data
have mostly no time variation, there is no appropriate way to address the
issue of causality directly.
We addressed the potential endogeneity problems using first time and firm
fixed effects regressions that adjust for hidden sources of corporate
heterogeneity. Fixed effects approach deals with the joint determination
issues in which an unnoticed time invariant variable influences
simultaneously share price informativeness and ownership structure. The
fixed effects estimations help decline omitted variable justification as a
source of endogeneity (Ferreira et al, 2010). Given that only the impact of
within variations in corporate ownership structure is considered, firm
specific omitted variables cannot explain the observed relationship
between share price informativeness about future earnings news and
ownership structure.
The second way to mitigate the problem of causality is to add appropriate
control variables consistent with Lehn et al., (2007) and Larcker et al.,
(2007). Specifically, we included some of the joint exogenous
determinants of firm-level corporate governance choices and earnings
response coefficients determinants which are firm financial leverage level
(debt to equity ratio), firm size (natural logarithm of corporate revenue)
and corporate profitability (natural logarithm of the return on equity ROE)
in all of our regression equations.
As for our pooled OLS regression that examines board characteristics we
followed the Larcker and Rusticus (2010) and Frank (2000)s alternative
approach. Their method involves assessing how large the endogeneity
issue (unmodeled variable) has to be to change the OLS coefficient

CHAPTER VIII
estimates and, in particular, how large it has to be to make the coefficient
statistically insignificant. Unlike the instrumental variable approach that is
meant to reduce bias in the coefficient estimates, Frank (2000) and Frank
et al. (2008) method is used to assess the sensitivity of a coefficient and
its standard error to the inclusion of a confounding variable.
They specified a minimum threshold necessary for an omitted confounding
variable to invalidate the significant results of a variable of interest in an
ordinary least square regression model. Frank (2000) offered an
improvement over previous applications of sensitivity analysis and
suggests that for a confounding variable to impact the statistical
inference, it should be correlated with both the independent variable and
the dependent variable (controlling for the other variables). He defined the
impact threshold (ITCV) as the lowest product of the partial correlation
between the confounding variable and the outcome and the partial
correlation between the endogenous variable of interest and the
confounding variable.
For the purpose of our sensitivity test we calculate the impact of the
confounding variables on the significant coefficients of two endogenous
independent variables with respect board characteristics (BS and NED).
While, by definition, we do not have access to the unobserved and
unmodeled variable, we do have other control variables: financial leverage
ratio, firm size and profitability. We are able, therefore, to calculate the
impact of the inclusion of each control variable on the significant
coefficient on the interaction terms NED*RD*Xt+1 and BS*RD*Xt+3. Likewise,
the impact of each control variable is the product of the partial correlation
between the endogenous board characteristics measures and the control
variables and the correlation between the dependent variable (Rt) and the
control variable (taking out partially the effects of the other control
variables). Frank et al. (2013) suggest that for a valid inference, the
impact of a confounding and unmodeled variable should not exceed the
estimated impact threshold, otherwise the coefficients on the explanatory
endogenous variables would be considered as fragile.

CHAPTER VIII
Table 32 reports the results of our sensitivity analysis. The threshold value
for the proportion of non-executive directors is 0.0364 which suggests the
correlation between the dependent variable (Rt) and the endogenous
independent variable (NED*RD*Xt+1) with the unobserved confounding
variable each only need to be about 0.190 ( 0.036 ) for the OLS result to
be overturned. Given that our control variables are interacted with
predictors in the augmented Collins et al (1994) regression of current
returns on future earnings, we follow Frank (2000)s approach in
addressing the specific case of multiple confounds. Frank (2000) used the
square root of the multiple correlation between x and cv (r(x.cv)) and the
square root of the multiple correlation between y and cv.(r(ycv)) to assess
the impact of confounding variables.
It is worth noting that in our case r(ycv) and r(x.cv) are the r2 statistics
from the regressions of current returns and the predictor of interest on
controlling variables16. Firm size (0.21) and corporate profitability (0.175)
have the largest impact on our regression coefficients. Indeed, the impact
of financial leverage, firm size and corporate profitability are respectively
58 percent, 476 percent and 382 percent of the ITCV for the interacted
variable NED*RD*Xt+1. These results suggest that these control variables
are important covariates to be included in the model, although
comparable covariates, in and of themselves, would not alter the inference
with regard to the variable of interest. Accordingly, our statistic inference
with respect to the joint effect of the proportion of non-executive directors
and risk disclosure on share price anticipation of future earnings is robust
to the problem of omitted variable.
The threshold single value for Board size (BS) is about 0.0345 suggesting
each of the relevant correlations needs to be about 0.185 ( 0.0345 ) for
the OLS result to be overturned. Similarly, firm size (0.227) and
16 In comparing the ITCV to the distribution of impact scores for the control variables we
implicitly assume that the confounding variable is similarly correlated with the other
control variables.

CHAPTER VIII
profitability (0.167) exhibit the largest impact on our OLS regression
coefficients. Moreover, the impact of financial leverage, firm size and
corporate profitability are 135 percent, 558 percent and 386 percent
respectively of the ITCV for the interacted predictor BS*RD*Xt+3. These
control variables are therefore important covariates and have to be
included in the model, although comparable covariates, in and of
themselves, would not alter the inference with regard to the variable of
interest. Accordingly, our statistic inference with respect to the joint effect
of the board size and risk disclosure on share price anticipation of future
earnings is robust to the problem of unmodeled variables. These findings
cleared, then the concern with regard to causal inference in our pooled
OLS regression and confirmed that the estimated coefficients are valid as
we controlled for the observed strong covariate.

Table 32 The impact of unmodeled variables

Explanatory(endogenous)/

Financi

Firm

Profitabi

Control Variables

al

size

lity

ge
-

0.278*

0.285*

0.323***

BC1: NED*RD*Xt+1

**
0.009*

**
0.010*

0.007**

BC2: BS*RD*Xt+3

**
0.480*

**
0.440*

0.534***

**

levera

OLS REG

Constant

Impact thresholds (BC1)

0.0364

CHAPTER VIII
Impact thresholds (BC2)
Impact of control (BC1)
Reliability (Impact/ITCV)
Impact of control (BC2)
Reliability (Impact/ITCV)

0.0345
0.057
1.58
0.081
2.35

0.21
5.76
0.227
6.58

0.175
4.82
0.167
4.86

The number of firm-year observations (n) is 789. For the two endogenous independent
variables, an impact statistic is calculated (ITCV) indicating the minimum impact of an
unobserved and unmodeled variable that is needed to render the coefficient statistically
insignificant. The ITCV is defined as the product of the correlation between the
endogenous independent variables (BC1 and BC2) and the unmodeled variable and the
correlation between the dependent variable (current stock returns) and the control
variables (partialling out the effect of the other control variables). The sign of the impact
measure indicates how the inclusion of the control variable affects the coefficient for the
endogenous independent variables (BC1 and BC2) respectively. The impact results also
help in assessing the likelihood that such an unmodeled variable exists. The sign of the
impact score indicates how the inclusion of each control variable affects the coefficient of
each endogenous independent variable. A positive impact score indicates that inclusion
of the control variables makes the coefficient on the endogenous independent variable
more positive or less negative. A negative impact score has the opposite effect.
Impact/ITCV is the reliability of the control variable.

Notwithstanding the empirical evidence on governance mechanisms that


enforce the voluntary release of useful risk information, higher stock price
informativeness seems to be associated also with more intense
competitive business environment (Irvine and Pontiff, 2009). In such
situation, the corporation may incur proprietary costs. A decrease in
corporate future cash flows is likely to occur following the disclosure of
such information. Typically, actual as well as potential competitors would
use risk information in a way it harms corporate market position.
Corporations look to limit their disclosure on their risk exposure and balance
their desire to convey relevant information under a tradeoff between costs and benefits.
Investors may consider this practice as a weakness in analyzing and
monitoring areas of such risk in their business. The lack of transparency
may lead to a lower perceived credibility of risk disclosure.

CHAPTER VIII
2.4.

The effect of proprietary costs and voluntary risk disclosure on share price anticipation of future
earnings

Table 33 reports main findings of H8 testing that predicts a weaker value


relevance of voluntary risk disclosure when interacted with proprietary
costs level. First, the coefficient on Xt is as predicted significantly positive
at the 10% level for the regression when we control for financial leverage
variable. This result provides weak evidence that current earnings reflect
prior anticipation of future earnings. There is also strong evidence of prices
leading earnings by three periods ahead regardless of the extent of
corporate risk disclosure and proprietary costs level. The future earnings
response coefficients on Xt+1, Xt+2 and Xt+3 are, with one exception, positive
and significant in the three regression models.
When we control for corporate risk, the coefficient on Xt+3 is insignificant
suggesting that share price informativeness with respect to future
earnings is limited to two years ahead. Future earnings response
coefficients extend from 1.062 to 2.831 and are considerably higher than
the current earnings coefficients. The coefficients on future stock returns
(Rt+1, Rt+2 and Rt+3) have as predicted the right sign and are significantly
different from zero mainly for t+1 and t+2 periods. These findings indicate
that realized future earnings contain measurement error that future
i, t

returns remove. Table 33 also shows that the coefficients on RD

*Xt as

well as

RD X t+K
K=1

are mostly insignificant in the three regression

models. There is only a positive and significant association between


current returns and current and future earnings for period t+2 when the
market is informed about corporate profitability. There is, accordingly, little
evidence on share price informativeness with respect to current and future
earnings when interacted with risk disclosure level.
The coefficients of our primary interest in this section indicate that
proprietary costs affect investors ability to reflect more future earnings

CHAPTER VIII
news in the current return. Consistent with Darrough and Stoughton
(1990), Gelb (2000) and Hossain et al (2006) s findings that when
proprietary costs are high, firms tend to restrain value relevant information
about future cash flows, we find that there is a significantly negative
3

relationship between

PCRD X t+K
K =1

and current returns. The negative

coefficients on the interacted terms indicate that the amount of voluntary


risk information impounded in the current stock price is less for firms with
high proprietary cost because investors perceive it as less relevant.
Investors will reconsider the perceived relevance of risk information as
long as they are aware that firms management would restrain useful
information about their business uncertainties.
Capital market participants are likely to rely on other sources of risk
information other than corporate annual rather than hold a short-term
view. One possible interpretation of these results is that a significant net
-profit margin (the mean is about 0.35) may attract future competition
and reveal higher forewarning of potential entrants (Cohen, 2005).
Investors may believe that corporations need to protect their future
opportunities and are likely to provide relevant risk information about its
future prospects via private channels.
Actually, the development of private sector and the improvement of
economic freedom through trade liberalization and property rights
promotion within the MENA region reduced the barriers to market
entry and created new business opportunities for international
investors. As risk management and disclosure strategies reflect the
underlying environmental influences (e.g. economic incentives, the
regulatory environment, competition, etc.), it seems natural that
successful MENA companies tend to withhold information about their
opportunities and prospects. Meanwhile, market participants seem to
make allowances for management discretion over risk information

CHAPTER VIII
and tend to diversify their sources of information when making their
economic decision.
Table 33: the joint effect of proprietary cost and risk disclosure on share price
anticipation of future earnings

Model

Exp.

Intercept

Sign
(?)

Xt

(+)

Financial

Firm

Leverage
Size
0.208**
(0.03 0.389
(0.12
0.043*

8)
(0.06 -0.167

4)
(0.79

Profitability
0.236**
(0.04VIII
CHAPTER
0.181

6)
(0.842

2)
(0.03

2.292*

)
(0.08

1.256*

3)
(0.06

Xt+1

(+)

2.831**

0)
(0.04 2.565*

Xt+2

(+)

1.062*

8) *
(0.09 1.106*

8)
(0.04

3) *
(0.11 2.050*

0)
(0.02

1.706*

1)
(0.08

Xt+3

(+)

1.364

Rt+1

()

-0.545*

2) *
(0.05 -

5)
(0.05

-0.531*

8)
(0.06

Rt+2

()

7) 0.542*
(0.00 -

9)
(0.01

1)
(0.01

3)

0.366**

3)

*
(0.16 -0.169

(0.17

-0.221

(0.118

0.423**

7) 0.369*

Rt+3

()

*
-0.184

AGt

()

0.112**

5)
(0.02 0.104*

0)
(0.02

0.076**

)
(0.02

Ept-1

(+)

-0.023*

6) *
(0.08 -0.010

2)
(0.41

-0.011

8)
(0.347

4)
(0.16

0.021

)
(0.309

0.288*

)
(0.09

0.180

8)
(0.272

RD

(?)

0.035

4)
(0.21 0.057

RD*Xt

(?)

0. 149

4)
(0.17 0.024

9)
(0.76

-0.146

6)
(0.22 -

4)
(0.04

RD*Xt+1

(+)

2) 0.176*

9)

RD*Xt+2

(+)

0.168

*
(0.13 0.095

(0.53

0.347*

(0.05

RD*Xt+3

(+)

-0.001

0)
(0.98 -0.048

7)
(0.54

0.135

1)
(0.385

8)
(0.24 -0.006

3)
(0.62

-0.031

)
(0.405

9)
(0.09 -0.020

6)
(0.38

-0.029

)
(0.217

RD*Rt+1

(?)

-0.023

RD*Rt+2

(?)

-0.035*

RD*Rt+3

(?)

-0.031

7)
(0.29 -0.016

6)
(0.69

-0.028

)
(0.177

RD*AGt

(?)

0.007

1)
(0.16 0.007

3)
(0.25

0.008

)
(0.111

RD*Ept-1

(?)

0.000

7)
(0.98 0.001

8)
(0.11

0.002**

)
(0.00

6)
(0.11 0.085

6)
(0.15

*
0.079**

3)
(0.05

PC

(?)

0.079

PC*Xt

(?)

0.036**

6)
(0.02 0.029*

8)
(0.08

0.032**

0)
(0.01

PC*Xt+1

(?)

0. 029*

1)
(0.07 0.022*

4)
(0.07

0.026*

8)
(0.06

CHAPTER VIII
Table 33 presents fixed effects regression results for panel data. Heteroscedasticityconsistent p-values are given in parentheses. The dependent variable is current period
return, Rt. Rt, Rt+1, Rt+2 and Rt+3 are calculated as buy-and-hold returns (inclusive of
dividends) over a 12-month period, starting six months after the end of the previous
financial year. Xt, Xt+1, Xt+2 and Xt+3 are defined as earnings change deflated by price. Both
current and future earnings changes are deflated by the price at the start of the return
window for period t. EPt1 is defined as period t1s earnings over price six months after
the financial year-end of period t1. AGt is the growth rate of the total book value of
assets for period t. RD is the natural logarithm of the total number of risk related
sentences, respectively, for Operations risk, Empowerment risk, Information processing
and technology risk, Integrity risk and Strategic risk information. Proprietary cost PC is the
net profit margin defined as Net Income/Revenue. Financial leverage, firm size and
profitability are control variables (their regression estimates are not tabulated). The
significance levels (two-tail test) are: *= 10 %, ** =5 % and *** = 1 %.

Summary
This chapter aims to examine whether corporate governance structure and the competitiveness
of the firm business environment are likely to influence the value relevance of corporate risk
disclosure for a sample of MENA emerging markets. The empirical findings are
mostly in line with these predictions. They suggest that board composition
and board size enhance investors ability to anticipate revealed future
earnings

growth.

In

contrast

concentrated

ownership,

managerial

ownership and institutional owners are likely to weaken the returnrevealed future earnings association while such impact is not detected in
the presence of CEO/Chairperson role duality. As for proprietary costs,
corporate profit margin is likely to reduce the value relevance of risk
disclosure. Statistic estimates are generated from a panel and pooled OLS
regressions whereby we controlled for the cross-sectional effects of some
previously documented determinants of voluntary disclosure as well as the
earnings response coefficients such as risk, firm size and profitability.
The findings are of interest because they suggest that the perceived
relevance of risk information is significantly influenced with governance
structure and the product business environment. While board
characteristics seem to encourage corporate transparency and informed
trading, ownership structure and the presence of proprietary costs had the
opposite effect on the interaction between risk disclosure and share price
informativeness.

CHAPTER VIII

CONCLUSION
CONCLUSION
The present study aims to test whether narrative risk disclosure in annual reports helps
investors predict future earnings growth. It examines also how corporate governance and
proprietary costs impact the perceived relevance of risk disclosure for a sample of MENA
emerging markets. We provide next a summary of the main findings followed by the main
implications of these findings. We also discuss the limitations of the study and end by
suggesting several avenues for future research.
1. Summary
This study addresses the question of whether the voluntary risk information in corporate
annual reports contains value-relevant information about future performance in the
context of MENA emerging economies. We examine also how corporate
governance characteristics and product market competition influence
investors ability to anticipate future earnings news. Our main hypotheses
predict that there is a positive association between voluntary risk
disclosures and share price anticipation of future earnings. Such
association might be weaker for concentrated ownership, in the presence
of managerial ownership, CEO/Chairman duality and proprietary costs. In
contrast, the return revealed future earnings relationship might be
positively influenced by the presence of institutional owners, the amount
of non-executive directors and large-sized boards. Our empirical findings
are based on large samples of annual reports electronically available for
roughly 320 non-financial firms listed in nine MENA emerging markets. We
collected on average 264 annual reports per year for the time period
2007-2009. We gathered the corresponding financial information (EPS,
share prices) from Thomson one database. We measured the extent of
risk disclosure in the narrative sections of annual reports and particularly
in the management discussion and analysis section using a manual based
content analysis methodology. We relied upon Linsley and Shrives (2006)
definition of risk information and referred to their grid as a coding
instrument to generate an aggregate score of corporate voluntary risk
information. We retained five categories of risk information as we dropped
the financial risk category. They are related to operations risk,

CONCLUSION
empowerment risk, information processing and technology risk, integrity
risk and strategic risk whereby we identified 32 items. We examined the
reliability of the disclosure scores by referring to a second independent
coder. The kalpha yielded a satisfactory level of reproducibility.
We investigated subsequently the association between the amount of risk
disclosure and prices leading earnings using the Collins et al. (1994)
regression model and following the methodology in Lundholm and Myers
(2002). We generated our estimates from year by year OLS, pooled and
panel regressions whereby we controlled for the cross-sectional effects of
some determinants of voluntary disclosure as well as the earnings
response coefficients such as risk, industry, firm size and profitability.
Consistent with some recent studies, we highlighted the cross sectional
time series dynamics in risk disclosure by adopting a longitudinal
approach.
This study offers several interesting results. First, we found a positive
relationship between voluntary risk information and the markets ability to
anticipate two-years ahead of future earnings growth. The positive
association provides us with the first empirical evidence of the value
relevance of voluntary risk disclosure in annual reports. These findings support
the view that current earnings alone have a limited ability to communicate a firms value to
the market. The market uses additional disclosures to anticipate future earnings.
The second set of results considered the interaction effect of governance
structure, proprietary costs and risk disclosure level on share price
informativeness with respect to future earnings. Some empirical results for
the sample period 2007-2009 are not in line with the hypothesized
predictions. We found that risk disclosure associated with good board
structure (large size and high proportion of non-executive directors)
enhances share price anticipation of future earnings. In contrast,
ownership structure as examined through the level of ownership
concentration, managerial ownership and institutional investors seems to
decrease the association between risk disclosure and share price

CONCLUSION
anticipation of future earnings. The presence of CEO/chairman duality had
no impact on share price informativeness with respect to future earnings
news. Finally, we found that the level of proprietary costs tends to
moderate the perceived relevance of risk information, thereby making
investors rely on other sources of information in forecasting future
earnings change. These results suggest that information driven by some
corporate governance mechanisms and product market competition is
likely to shape the perceived relevance of narrative risk information in
annual reports. We dealt with the concern for potential endogeneity
problems between share price informativeness and firm-level governance
structure using first time and firm fixed effect regressions. We added
additionally appropriate control variables to mitigate the potential
causality issue consistent with Lehn et al., (2007) and Larcker et al.,
(2007).
As for our pooled OLS regression, we followed the Larcker and Rusticus
(2010) and Frank (2000)s alternative approach which investigates
whether regression is robust to the unmodeled/confounding variables
issue. This sensitivity test confirmed that the estimated coefficients are
valid as we controlled for strong observed covariate.
2. Implications
For corporate disclosure researchers, a potential implication of this study is
that it requires rethinking the role of corporate risk communication and
revising the markets perceptions of uncertainty. The findings are of
interest because they emphasize the benefits of corporate voluntary risk
disclosures to investors and to the disclosing firm. As regulators tighten
corporate reporting standards all over the world in response to the recent
economic crisis, we documented evidence that narrative risk disclosure
reveals new information about corporate future performance. Our results
challenge the claims made by some recent studies, that risk information is
ambiguous, with limited use (Schrand and Elliot, 1998) and does not
contain reliable information (Campbell and Slack, 2008; Davies, Moxey,
and Welch, 2010; Moxey and Berendt, 2008). This study also provides

CONCLUSION
empirical evidence on the part and the channel of information that are
needed for an accurate cost-benefit analysis of enhanced disclosures.
Our results have also managerial implications. They revealed that investors and market
participants are better informed when conveying information about uncertainties and
opportunities in annual report narratives. Therefore, to ensure adequate financial
communication with the stock market, managers should give high priority to appropriate and
effective disclosure policy. This disclosure strategy can protect both companies and
managers long term reputation. The findings provide additionally assistance to managers
looking to understand more precisely how risk disclosures affect investors ability to
anticipate future earnings growth. In particular, the results demonstrate that a good board
structure increases the perceived relevance of corporate disclosure while some ownership
attributes and high competition in the product market have the opposite effect. Managers
could then take active measures to influence investors by carefully choosing the quantity of
risk disclosure and the characteristics of their governance structures.
Our results have furthermore considerable implications for unsophisticated investors who did
not have access to private information through other sources in the same way that
sophisticated investors such as financial analysts or institutional investors do.
The findings suggest that small investors need additional information other than reported
earnings to anticipate future earnings growth and guide their investment decisions.
Finally, this study has implications for the efficient market hypothesis. Our findings suggest
that the amount of risk disclosures in the narrative sections of the annual report provides
investors with valuable relevant information. This information enables them to better
anticipate future earnings. Accordingly, this leads to more efficient capital markets.
3. Limitations and suggestions for future research
Limitations in this study are outlined, together with suggestions for further research.
Although this study is one of the pioneering researches that investigated the usefulness of
voluntary risk disclosure in anticipating future earnings, it still suffers from some caveats.
First, the study calculates risk disclosure scores by simply adding up the number of sentences
in the narrative sections of annual reports. This approach ignores the fact that the usefulness
of risk information can vary from sentence to sentence and from a risk category to another.

CONCLUSION
Beretta and Bozzolan (2004, 2008) suggest that the quantity of disclosure alone cannot serve
as a good proxy for disclosure quality. Second, despite the measures that
followed to thoroughly understand companies'
classification process suffers from inherent

annual reports,

judgment

have been

the scoring and

limitations and subjectivity,

which cannot be entirely eradicated. We believe then that the process of analyzing annual
report narratives has scope for further refinement. While our current methodology equates
disclosure quality with the amount of information provided, we think that future studies
should differentiate risk disclosure that provides favorable from those that provide
unfavorable earnings news. Another refinement is to discriminate between the sub-categories
of risk disclosure which could be extremely useful for studying the benefits of each type of
risk related disclosures. A third limitation is that our study does not cover all the managerial
incentives/disincentives for useful risk disclosure in anticipating future earnings growth. It is
important to acknowledge that the use of only one proxy of competition due to a lack of
proper data is one of

the limitations of

this study. Also, given that the corporate

governance definition of good best practices is still ambiguous and unresolved (Brickley
and Zimmerman, 2010), the internal and external governance measures might suffer from
measurement bias. As risk reporting is still a fertile area not only for empirical but also for
conceptual and analytical research, we think that there are many refinements and other
dependent variables that could be investigated in future research.
In particular, corporate audit environment (audit committee composition, external auditor
independence, etc.), categories of institutional investors (short term/long term investors),
different types of block-holding (i.e., in institutions, in financial firms, in non-financial firms,
and by individuals) and the presence of litigation costs may also have differing effects on
the perceived relevance of risk disclosure.
An additional methodological issue is related to dealing with the claim that the corporate
governance structure (ownership structure and board characteristics) is endogenous. It is
important to highlight that endogeneity is a highly subjective topic and to note that
researchers currently do not have enough knowledge to resolve this issue. Coles et al.
(2012) and Van Lent (2007) claim that the available solutions to endogeneity so far
(including the simultaneous system of equation, instrumental regressions and fixed
effects) all fail to provide a pure solution to the problem of endogeneity. It is highly
uncertain, then to what extent this problem is actually solved in the model. Future

CONCLUSION
research should consider using lagged data for all financial data. Li and Srinivasan (2011)
and Lo et al. (2010) suggest that the use of lagged data controls for endogeneity.
Finally, while our study provides interesting results about the cross sectional time series effect
of corporate risk disclosure on share price anticipation of future earnings in the context of
MENA emerging markets, cross-country differences in price informativeness about future
earnings are uncovered by this study. Accordingly, future research may fill this gap and
empirically examines some country-level structural factors that explain these differences.

REFERENCES
REFERENCES
Abbott. W.F, Monsen R.J (1979). On the Measurement of Corporate Social
Responsibility: Self-Reported Disclosures as a Method of Measuring
Corporate Social Involvement. Academy of Management Journal.
22(3), 501-515.
AbdelFattah, T., R. Dixon, A. Woodhead, (2008), Corporate governance
and voluntary disclosure in emerging capital markets: evidence from
Egypt, Working Paper, Durham University, Durham.
Abdul Hamid M., N. Nordin (2013) Corporate Risk Management, Firm
Characteristics, Ownership Structure, and Governance Attributes of
Banks: A Case of Malaysia. Prosiding Perkem, 8(2), 655 660.
Abeysekera I., (2007) "Intellectual capital reporting between a developing
and developed nation", Journal of Intellectual Capital, 8(2), 329 345
Aboody D. Kasznik R. (2000) CEO stock options awards and the timing of
corporate voluntary disclosures. Journal of Accounting and
Economics 29, 73100
Abraham S., Cox, P (2007) Analysing the determinants of narrative risk
information in UK FTSE 100 annual reports The British Accounting
Review , 39, 227248
Abraham S. Shrives P.J. (2014) Improving the relevance of risk factor
disclosure in corporate annual reports The British Accounting
Review, 46(1), 91107.
Accounting Standards Board (ASB) (1993) Operating and Financial
Review in Institute of Chartered Accountants in England and Wales.
Accounting Standards 1997-98.
Accounting Standards Steering Committee (ASSC) (1975) The Corporate
Report, London: ASSC.
ACRLI (2009) Regional Report on the Status of the Commercial Laws in
Lebanon, Yemen, Tunisia and United Arab Emirates.
http://www.arabruleoflaw.org/compendium/Files/MENA-CLSRegionalReport-En.pdf
Adams, C.A., Hill W.Y. and Roberts C.B. (1998) Corporate social reporting
practices in Western Europe: Legitimating corporate behaviors?
British Accounting Review, 30(1), 1-21.
Adhikari, A., R.H. Tondkar, (1992) Environmental Factors Influencing
Accounting Disclosure Requirements of Global Stock Exchanges,
Journal of International Financial Management and Accounting, 4(8),
75-105
Aggarwal, R. E., S. Isil, M. Rene, R. Williamson, (2007) Do U.S. Firms Have
the Best Corporate Governance? A Cross-Country Examination of the
Relation Between Corporate Governance and Shareholder Wealth
Working Paper. Fisher College of Business
Agrawal A., G. N. Mandelker (1990). Large Shareholders and the

REFERENCES
Monitoring of Managers: The Case of Antitakeover Charter
Amendments The Journal of Financial and Quantitative Analysis,
25(2), 143-161.
Ahmed, K., M. Hossain, M. Adams, (2006). The effects of board
composition and board size on the informativeness of annual
accounting earnings. Corporate Governance: An International
Review, 14 (5), 418431.
AICPA (1994) Improving Business Reporting A Customer Focus: Meeting
the Information Needs of Investors and Creditors, Comprehensive
Report of the Special Committee on Financial Reporting (The Jenkins
Report), New York, NY: American Institute of Certified Public
Accountants.
Ajinkya, B. Gift, M. (1984) Corporate managers earnings forecasts and
symmetrical adjustments of market expectations. Journal of
Accounting Research 22, 425444.
Ajinkya, B. S. Bhojraj, P. Sengupta, (2005). The association between
outside directors, institutional investors and the properties of
management earnings forecasts. Journal of Accounting Research,
43(2), 343-376.
Akerlof G. (1970) The market for lemons: quality uncertainty and the
market mechanism. Quarterly Journal of Economics 90, 629650.
Al Janadi Y., R. Abdul Rahman, N. Haj Omar (2013) Corporate Governance
Mechanisms and Voluntary Disclosure in Saudi Arabia. Research
Journal of Finance and Accounting, 4(4), 25-35.
Al-Akra, M., M. J. Ali, O. Marashdeh, (2009). Development of accounting
regulation in Jordan. The International Journal of Accounting, 44 (2),
163-189.
Alexandre H., M. Paquerot, (2000). "Efficacit des structures de contrle et
enracinement des dirigeants," Revue Finance Contrle Stratgie,
3(2), 5-29
Ali A. Klasa S., Yeung E. (2014) Industry concentration and corporate
disclosure policy. Journal of Accounting and Economics
(Forthcoming).
Ali, A., J. Ronen, S.H Li, (1994) Discretionary Disclosures in Response to
Intra-Industry Information Transfers. Journal of Accounting, Auditing
and Finance, 9(2), 265282.
Allini, A, Hussainey, K., Manes Rossi, F. (2014). The boards role in risk
disclosure: An exploratory study in Italian listed government-owned
companies. ERRN 9th European Risk Conference, Naples University.
Allison, P.D., (1999). Multiple Regression: A Primer . Pine Forge Press.
Al-Omari, AM (2010), The Institutional Framework of Financial Reporting in
Jordan, European Journal of Economics, 22, 32-50

REFERENCES
Aloui C., I. Z. Rejichi (2012) Hurst exponent behavior and assessment
of the MENA stock markets efficiency. Research in International
Business and Finance, 26, 353 370
Al-Yafi, M. (2005), AFCPA surveys publicly listed firms in the MENA region,
Corporate Governance Trends, No. 9, Fall.
Amico, A. (2014), Corporate Governance Enforcement in the Middle East
and North Africa: Evidence and Priorities, OECD Corporate
Governance Working Papers, No. 15, OECD Publishing.
http://dx.doi.org/10.1787/5jxws6scxg7c-en
Amihud Y., H. Mendelson (1988) Liquidity and asset prices: financial
management implications. Financial Management 17, 515
Amihud Y. H. Mendelson (1986) Asset pricing and the bidask spread.
Journal of Financial Economics 17, 223249.
Amihud Y. (2002). Illiquidity and stock returns: cross-section and timeseries effects. Journal of Financial Markets, 5, 3156.
Amir and Lev (1996) Value-relevance of nonfinancial information: the
wireless communications industry. Journal of Accounting and
Economics 22, 330.
Amran, A., A. M. R. Bin, and B. C. H. M. Hassan. (2009). Risk reporting-An
exploratory study on risk management disclosure in Malaysian
annual reports. Managerial Auditing Journal, 24 (1), 3957.
Anandarajan, A, I. Hasan (2010) Value relevance of earnings: Evidence
from Middle Eastern and North African Countries. Advances in
Accounting, incorporating Advances in International Accounting, 26,
270-279.
Aoki M., (2000) Information, Corporate Governance, and Institutional
Diversity: Competitiveness in Japan, the USA, and the Transitional
Economies. Oxford University Press, Inc. New York, NY, USA.
Arcay, M., M. Vazquez, (2005) Corporate Characteristics Governance
Rules and The Extent of Voluntary Disclosure in Spain. Advances in
Accounting, 21, 299-331.
Archambault, J.J., M.E. Archambault, (2003) A Multinational Test of
Determinants of Corporate Disclosure, The International Journal of
Accounting, 38.
Ashbaugh, Pincus (2001) Domestic accounting standards, international
accounting standards, and the predictability of earnings. Journal of
Accounting Research 39, 417434.
Ashley J. W. (1962) "Stock Prices and Changes in Earnings and Dividends:
Some empirical results." Journal of Political Economy, 82-85.
Athanasakou V., K. Hussainey (2014) The perceived credibility of forward
looking performance disclosures. Accounting and Business
Research, 44 (3), 227-259.
Baginsky S. P. Hassel J. M. (1997) The determinants of management

REFERENCES
forecast precision. The Accounting Review, 72, 303-312.
Ball R. Brown P. (1968) An empirical evaluation of accounting income
numbers. Journal of Accounting Research 6, 159177.
Ball R. Kothari S. (1991) Security returns around earnings
announcements. The Accounting Review 66, 718738.
Ball, R., S. Kothari, A. Robin, (2000), The effect of international
institutional factors on properties of accounting earnings, Journal of
Accounting and Economics., 29(1), 1-51.
Ball, R., S. Kothari, R. Watts, (1993). Economic determinants of the
relation between earnings changes and stock returns. The
Accounting Review, 68, 622638.
Baltagi B.H. (2005) Econometric analysis of panel data. John Wiley &
Sons Ltd. The Atrium, Southern Gate, Chichester
Bamber L., Cheon Y. (1998). Discretionary Management Earnings Forecast
Disclosures: Antecedents and Consequences of Forecast Venue and
Forecast Specificity Choices. Journal of Accounting Research: 36(2):
167-190.
Banghoj J. Plenborg T. (2008) Value relevance of voluntary disclosure in
the annual report, Accounting and Finance, 48, 159-180.
Bao Y., A. Datta (2014) Simultaneously Discovering and Quantifying Risk
Types from Textual Risk Disclosures. Management Science, 60(6),
1371-1391
Barakat A., Hussainey K. (2013) Bank governance, regulation,
supervision, and risk reporting: Evidence from operational risk
disclosures in European banks. International Review of Financial
Analysis, 30, 254273.
Barako, D. G. (2007). Determinants of voluntary disclosure in Kenya
companies annual reports. African Journal of Business Management,
1(5), 113-128.
Barako, D.G., P. Hancock, H.Y. Izan (2006) Factors influencing voluntary
corporate disclosure by Kenyan companies Corporate Governance:
An International Review, 14 (2), 107125.
Barker C. A. (1956) Effective stock splits. Harvard Business Review 34:1,
101-106.
Barker C. A. (1957) Stock splits in a bull market. Harvard Business
Review 35:3, 72-79.
Barker C. A. (1958) Evaluation of stock dividends. Harvard Business
Review, 36:4, 99-114.
Barth, M, Beaver, W. and Landsman W. (1996), Value-Relevance of Banks
Fair Value Disclosures under SFAS No. 107, The Accounting Review,
513-537.
Bassett M., Koh, P. S., Tutticci, I. (2007), The association between
employee stock option disclosures and corporate governance:

REFERENCES
evidence from an enhanced disclosure regime, British
Accounting Review, 39(4), 303-322.
Basu, S. (1997) The conservatism principle and the asymmetric
timeliness of earnings. Journal of Accounting and Economics 24, 3
37.
Bauwhede H. V., M. Willekens, (2008). Disclosure on Corporate
Governance in the European Union, Corporate Governance: An
International Review, 16(2), 101-115
Baydoun, N. Maguire, W. Ryan, N. and Willett, R. (2013). Corporate
Governance in Five Arabian Gulf Countries, Managerial Auditing
Journal, 28(1),7-22.
Beasley, M. (1996). An Empirical Analysis of the Relationship between the
Board of Director Composition and Financial Statement Fraud, The
Accounting Review, 71(4), 443-465.
Beattie, V. A., McInnes, B., S. Fearnley, (2002) Through the eyes of
management: A study of narrative disclosure. An interim report.
London: Centre for Business Performance, The Institute of Chartered
Accountants in England and Wales.
Beattie, V., B. McInnes, S. Fearnley, (2004). A Methodology for Analyzing
and Evaluating Narratives in Annual Reports: A Comprehensive
Descriptive Profile and Metrics for Disclosure Quality Attributes,
Accounting Forum, 28(3),.205-236.
Beattie, V., S. Thomson, (2007). Lifting the Lid on the Use of Content
Analysis to Investigate Intellectual Capital Disclosures, Accounting
Forum, 31(2), 129-163.
Beaver W.H., Angel E.E. (1996) Discretionary behavior with respect to
allowance for loan losses and the behavior of security prices.
Journal of Accounting and Economics 22, 177206.
Beaver W. H., R. Dukes (1973) Interperiod tax allocation and delta
depreciation methods: some empirical results. The Accounting
Review 48, 549559.
Beaver W.H., Clarke R., Wright F. (1979). The association between
unsystematic security returns and the magnitude of earnings
forecast errors. Journal of Accounting Research 17, 316340.
Beaver W.H., Lambert and Morse D. (1980) The Information Content of
Security Prices. Journal of Accounting and Economics 2, 3-28.
Beaver W.H., Lambert R. and Morse D. (1987) The information content of
security prices: a second look. Journal of Accounting and Economics
9, 139157.
Beaver, W.H., (1968). The information content of annual earnings
announcements. Journal of Accounting Research, 6, 6792.
Beaver, W.H. (1998) Financial Reporting: An Accounting Revolution.

REFERENCES
Upper Saddle River, NJ: Prentice Hall. 3rd ed.
Becht, M., Bolton, P., & Rell, A. (2003). Corporate governance and control.
Handbook of the Economics of Finance, 1, 1-109.
Beck, N. (2001), Time-series-cross-section data, Statistica Neerlandica,
55, 111133.
Beck, T., A., Demirguc-Kunt, R. Levine, (2003) Law and Finance: why Does
Legal Origin Matter? Journal of Comparative Economics, 31(4), 653676.
Beekees W., Pope P., Young, S. (2004). The link between earnings
timeliness, earnings conservatism and board composition: evidence
from the UK. Corporate Governance: An international review, 12,
47-59.
Beekes, W. A., P. Brown, (2006). Do better-governed Australian firms make
more informative disclosures?. Journal of Business Finance and
Accounting, 33(3-4), 422-450.
Belkaoui-Riahi, A. (2001) Level of Multinationality, Growth Opportunities
and Size as Determinants of Analysts Ratings of Corporate
Disclosures. American Business Review 19, 115-220.
Ben Naceur S, S. Ghazouani, M. Omran (2008) Does stock market
liberalization spur financial and economic development in the MENA
region? Journal of Comparative Economics 36, 673693.
Ben Naceur S., L. Chiraz (2009) Middle East and North Africa Region:
Financial Sector and Integration. Resource, 5(6.5), 5-7
Ben Othman, H., D. Zeghal, (2008) A study of corporate governance
disclosure and its country-level determinants in the emerging
markets. Research in Accounting in Emerging Economies, 8, 125155.
Ben Othman, H. D. Zghal, (2010). Investigating Transparency and
Disclosure Determinants at Firm-Level in MENA Emerging Markets.
International Journal of Accounting, Auditing and Performance
Evaluation, 6(4), 368-396.
BenNaceur S., M. Omran, (2011) The effects of bank regulations,
competition, and financial reforms on banks' performance.
Emerging Markets Review, 12, 120
Berelson B. (1952) Content Analysis in Communications Research.
Glencoe, IL: Free Press.
Beretta S. Bozzolan S. (2004) A framework for the analysis of firm risk
communication. The International Journal of Accounting, 39(3), 265288.
Beretta, S., S. Bozzolan, (2008). Quality versus Quantity: the Case of
Forward-Looking Disclosure. Journal of Accounting, Auditing and
Finance, 23 (3), 333-375.

REFERENCES

Bernard V. Thomas J. (1990) Evidence that stock prices do not fully


reflect the implications of current earnings for future earnings.
Journal of Accounting and Economics 13, 305-340.
Bernard V. Thomas J. (1989) Post-earnings-announcement drift: delayed
price response or risk premium? Journal of Accounting Research 27,
1-48.
Beyer, A, D. Cohen, T. Z. Lys, and Walther B. R., (2010) The financial
reporting environment: review of the recent literature. Journal of
Accounting and Economics, 50, 296-343.
Bhattacharya, S., Ritter, J.R. (1983), Innovation and Communication:
Signaling with Partial Disclosure, Review of Economic Studies, 50,
331346.
Bhushan (1989) Firm characteristics and analyst following. Journal of
Accounting and Economics, 11, 255-274.
Bley, J. (2011). Are GCC stock markets predictable? Emerging Markets
Review, 12 (3): 217237.
Boesso G., K. Kumar, (2007) Drivers of corporate voluntary disclosure: A
framework and empirical evidence from Italy and the United States.
Accounting, Auditing & Accountability Journal, 20(2), 269 296.
Bogdan V., Madalina P. C., Nicoleta P. D., Carmen S. (2009) Voluntary
Internet Financial Reporting and Disclosure A New Challenge for
Romanian Companies Annals of the University of Oradea, Economic
Science Series; 18(3), 770-778.
Bollen N. Whaley R. (1998) Are Teenies Better? Journal of Portfolio
Management, 25, 10-24.
Bollerslev T., E. Robert, F. Wooldridge, M. Jeffrey, (1988). "A Capital Asset
Pricing Model with Time-Varying Covariances," Journal of Political
Economy, University of Chicago Press, 96(1), 116-31.
Bontis, N. (2003). Intellectual Capital Disclosure in Canadian
Corporations, Journal of Human Resource Costing and
Accounting, 7(1/2), 9-20
Botosan C. A., M. A. Plumlee (2002) A Re-examination of disclosure
quality and the expected cost of capital. Journal of Accounting
Research, 40, 21-40.
Botosan C. A, M. S. Harris (2000) Motivations for a Change in Disclosure
Frequency and Its Consequences: An Examination of Voluntary
Quarterly Segment Disclosures. Journal of Accounting Research,
38(2), 329-353.
Botosan C. A., M. Stanford, (2005) Managers' Motives to Withhold
Segment Disclosures and the Effect of SFAS no. 131 on Analysts'
Information Environment. The Accounting Review, 8(3), 751-771.
Botosan, C. A. (2004). Discussion ofA framework for the analysis of firm

REFERENCES
risk communication. The International Journal of Accounting, 39 (3):
289295.
Botozan C. A (1997) Disclosure Level and the Cost of Equity Capital. The
Accounting Review, 72(3), 323-349.
Botozan C. A (2006) Disclosure and the cost of capital: what do we know?
Accounting and Business Research, 36(1), 31-40.
Bouri, E. I., Yahchouchi, G. (2014). Do return and volatility traverse the Middle Eastern and
North African (MENA) stock markets borders? Journal of Economic Studies, 41(2),
317-344.
Bowen, R., Burgstahler, D., Daley, L., (1987) The incremental information
content of accrual versus cash flows. The Accounting Review 62,
723747.
Bowman, E. H. (1984). Content analysis of annual reports for corporate
strategy and risk. Interfaces, 14, 61-71.
Bradbury, M.E. (1992) Voluntary disclosure of financial segment data:
New Zealand Evidence, Accounting and Finance, 32(1):15-26.
Brambor, T., Clark, W.R., Golder, M., ( 2006). Understanding
interaction models: Improving empirical analyses. Political
Analysis. 14 (1), 6382.
Breton, G., R.J. Taffler, (1995) Creative accounting and investment analyst
response. Accounting and Business Research, 25(98), 81-92.
Brickley, J. A., Zimmerman, J. L. (2010). Corporate governance myths: Comments on
Armstrong, Guay, and Weber. Journal of Accounting and Economics, 50(23),
235-245.
Brickley, J., Zimmerman, J., Smith, J., Clifford W. (2008). Managerial Economics &
Organizational Architecture (5 ed.). Boston: McGraw-Hill/Irwin.
Brown P. (1970) The Impact of Annual Net Profit Reports on the Stock
Market, Australian Accountant , 277 283
Brown L. D., K. J. Kim (1993) The association between non-earnings
disclosures by small firms and positive abnormal returns. The
Accounting Review 68, 668680.
Brown P, W. Beekes, P.Verhoeven (2011) Corporate governance,
accounting and finance: A review, Accounting and Finance, 51, 96
172
Brown, L. D., & H. Higgins, (2002). Managing earnings surprises in the US
versus 12 other countries. Journal of Accounting and Public
Policy, 20(4), 373-398.
Brown, S. V., J. W. Tucker, (2011). Large-sample evidence on firms yearover-year MD&A modifications. Journal of Accounting Research, 49,
309-346
Brown, S., K. Lo, T. Lys, (1999) Use of R2 in accounting research:

REFERENCES
measuring changes in value-relevance over the last four decades.
Journal of Accounting and Economics, 28(2), 83-115.
Brown S.J., J.B. Warner (1980) Measuring security price performance.
Journal of Financial Economics, 8(3), 205-258.
Bushee, B. J., Noe, C. F. (2000). Corporate Disclosure Practices, Institutional Investors, and
Stock Return Volatility. Journal of Accounting Research, 171-202
Bushee, B., D. Matsumoto, G. Miller (2003) "Open versus Closed
Conference Calls: The Determinants and Effects of Broadening
Access to Disclosure". Journal of Accounting and Economics, 34 149180
Bushman, R., A. Smith (2001) Financial Accounting Information and
Corporate Governance. Journal of Accounting and Economics 32(13), 237-333.
Bushman, R., J. Piotroski, A. Smith (2004) What Determines Corporate
Transparency? Journal of Accounting Research. 42 207-252.
Byard, D., K.W. Shaw (2003) Corporate disclosure quality and properties
of analysts Information environment. Journal of Accounting
Auditing and Finance 18 (3): 355378.
Byard, D., Li, Y. Weintrop, J. (2006) Corporate Governance and the Quality
of Financial Analysts Information, Journal of Accounting and Public
Policy, 25, 609-625
Cadbury Report: (1992), The Financial Aspects of Corporate Governance
(Gee and Co. Ltd., London: ISBN 0 85258 915 8).
Callahan, C., Smith, R. (2004). Firm performance and managements
discussion and analysis disclosures: An industry approach. Working
paper, University of Arkansas.
Campbell, D., R. Slack, (2008). Narrative reporting: Analysts perceptions
of its value and relevance. London: The Association of Chartered
Certified Accountants. RR104.
Campbell, J. L., Chen, H., Dhaliwal, D. S., Lu, M., & Steele, L. B. (2014).
The information content of mandatory risk factor disclosures in
corporate filings. Review Accounting Study,19, 396455
Campbell, J., Lo, A., MacKinlay, A., (1997) The Econometrics of Financial
Markets. Princeton University Press, Princeton, NJ.
Carlton, D. W., Fischel, D. R. (1983). The Regulation of Insider Trading. Program in Law and
Economics, University of Chicago Law School.
Carney, T. F. (1972). Content analysis: a technique for systematic inference from
communications. University of Manitoba Press.
Carson E., R. Simnett (1997) Voluntary disclosure of corporate
governance information. University of New South Wales, School of
Accounting
Casta J. F. (2000), Thorie Positive, In Encyclopdie Comptabilit,

REFERENCES
contrle, audit, Ed Economica
erne, K. (2009). Influential Factors Of Countrys Accounting System Development.
Economic Research, 22(2), 66-97.
Chalmers, K., Godfrey, J. M. (2004). Reputation costs: the impetus for voluntary derivative
financial instrument reporting. Accounting, Organizations and Society, 29(2), 95-125.
Chan, L.K.C., J. Lakonishok, (1995), The behavior of stock prices around
institutional trades, Journal of Finance, 50, 1147-1174.
Chau, G., P. Leung, (2006). The impact of Board Composition and Family
Ownership on Audit Committee Formation: Evidence from Hong
Kong. Journal of International Accounting Auditing & Taxation, 15, 115.
Chau, G., S. Gray, (2010). Family ownership, board independence and
voluntary disclosure: Evidence from Hong Kong. Journal of
International Accounting Auditing & Taxation, 19 (2), 93-109.
Chau, G.K., S.J. Gray (2002). Ownership structure and corporate
voluntary disclosure in Hong Kong and Singapore. The International
Journal of Accounting, 37(2), 247-265.
Chen, C., B. Jaggi, (2000) Association Between Independent NonExecutive Directors, Family control and Financial Disclosures in Hong
Kong, Journal of Accounting and Public Policy, 19, 285-310.
Chen, G., M. Firth, D. N. Gao, O. M. Rui al (2006) Ownership structure,
corporate governance, and fraud: Evidence from China. Journal of
Corporate Finance, 12, 424448
Chen, K. C. W., Z. Chen, K .C. J. Wei, (2003) Disclosure, Corporate
Governance, and the Cost of Equity Capital: Evidence from Asias
Emerging Markets. Unpublished working paper. Hong Kong University of
Science and Technology.
Cheng, C., M. Courtenay, (2006) Board Composition, Regulatory Regime
and Voluntary Disclosure. The International Journal of Accounting,
41, 262-289.
Cheynel E. (2009) A Theory of Voluntary Disclosure and Cost of Capital.
Review of Accounting Studies, 18(4), 987-1020.
Choi, S., G. Salamon, (1990) Accounting information and capital asset
prices. Working paper, Vanderbilt University, Nashville, TN
Chung K. H., H. Zhang. (2011) "Corporate Governance and Institutional
Ownership." Journal of Financial and Quantitative Analysis, 46, 247273.
Chung, R., Firth, M., Kim, J. (2002). Institutional monitoring and opportunistic earnings
management. Journal of Corporate Finance, 8(1), 29-48.
Clarke, J., T. Jandik, G. Mandelker (2001) The efficient markets
hypothesis. In Expert Financial Planning: Advice from Industry

REFERENCES
Leaders, ed. R. Arffa, 126-141.
Clarkson P. Guedes, J., R. Thompson (1996). On the diversification,
observability, and measurement of estimation risk. Journal of
Financial and Quantitative Analysis, 31:69-84.
Clarkson, P., J.L. Kao, G. Richardson (1994). The Voluntary Inclusion of
Forecasts in the MD&A Section of Annual Reports. Contemporary
Accounting Research, 11,423-450.
Cohen, D. A. (2005). Quality of financial reporting choice: Determinants and economic
consequences (Doctoral dissertation, Northwestern University).
Coles, J. L., Lemmon, M. L., & Felix Meschke, J. (2012). Structural models and endogeneity
in corporate finance: The link between managerial ownership and corporate
performance. Journal of Financial Economics, 103(1), 149-168.
Coles, J., U. Loewenstein, J. Suay, (1995) On equilibrium pricing under
parameter uncertainty. Journal of Financial and Quantitative
Analysis, 30:347-364.
Coller, M., T. Yohn (1997) Management forecasts and information
asymmetry: An examination of bid-ask spreads, Journal of
Accounting Research, 181-191.
Collins, D., E. Maydew, I.Weiss, (1997) Changes in the value-relevance of
earnings and book values over the past forty years. Journal of
Accounting and Economics 24 (1), 3968.
Collins, D., S. Kothari (1989) An analysis of inter-temporal and crosssectional determinants of earnings response coefficients. Journal of
Accounting and Economics, 11, 143181.
Collins, D.W., Kathari, S.P., Shanken, J., R.G. Sloan, (1994). Lack of
timeliness and noise as explanations for the low
contemporaneous return earnings association. Journal of
Accounting and Economics, 18, 289324
Conklin, D. (2002). Analyzing and managing country risks. Ivey Business
Journal, 66 (3): 36-41.
Cooke, T. E, (1989a),"Disclosure in the Corporate Annual Reports of
Swedish Companies", Accounting and Business Research, 19(74).
Cooke, T.E. (1989b), "Voluntary Corporate Disclosure by Swedish
Companies", Journal of International Financial Management and
Accounting, (1)2,171-195.
Core, J. E. (2001). A review of the empirical disclosure literature: discussion. Journal of
Accounting and Economics, 31(1-3), 441-456.
Cormier, D., M. Magnan, B. Van Velthoven, (2005). Environmental
disclosure quality in large German companies: economic incentives,
public pressures or institutional conditions? European Accounting
Review, 14(1), 339.
Craswell, A.T., Taylor, S. L. (1992). Discretionary Disclosure of Reserves by Oil and Gas
Companies: An Economic Analysis. Journal of Business Finance & Accounting,

REFERENCES
19(2), 295-308.
Darrough, M. (1993). Disclosure policy and competition: Cournot
vs.Bertrand. Accounting Review. 68 534561.
Darrough, M., N. Stoughton, (1990) Financial disclosure policy in an entry
game. Journal of Accounting and Economics 12, 219244.
Das, S., B. Lev, (1994) Nonlinearity in the returnearnings relation: tests
of alternative specifications and explanations. Contemporary
Accounting Research 11, 353379.
David, R. (1985). Major Legal Systems in the World Today: An
Introduction to the Comparative Study of Law, 3rd edn. (translated
by John E.C. Brierley), London: Stevens.
Davies, J., Moxey, P., I. Welch, (2010). Risk and reward: Tempering the
pursuit of profit. London: ACCA.
Deakin, E.B., Norwood, G.R. and Smith, C.H. (1974). The Effect of
Published Earnings Information on the Tokyo Stock Exchange
Trading. International Journal of Accounting, 9, 199-205.
DeBondt, W., Thaler, R., (1985) Does the stock market overreact? Journal
of Finance 40, 793805.
DeBondt, W., Thaler, R., (1987) Further evidence of investor overreaction
and stock market seasonality. Journal of Finance 42, 557581.
Dechow, P. M., Ge W., Larson C. R., Sloan R. G. (2009) Predicting Material
Accounting Misstatements. Working Paper, University of California.
Dechow, P., (1994) Accounting earnings and cash flows as measures of
firm performance: the role of accounting accruals. Journal of
Accounting and Economics 18, 342.
Declerck F., I. Martinez (2004), Informations volontaires et microstructure
des marchs : Quel impact sur le march des titres
NextEconomy et NextPrime?, 21me congrs international de lAssociation
Franaise de Finance (AFFI), juin, Paris.
Dedman, E., C. Lennox (2009) Perceived competition, profitability and the
withholding of information about sales and the cost of sales. Journal
of Accounting and Economics, 48(2-3), 210230.
Deller, D., M. Stubenrath, C. Weber, (1999), A Survey on the Use of the
Internet for Investor Relations in the USA, the UK and Germany, The
European Accounting Review, 8 (2) 351 -364.
Demsetz, H., (1983). "The Structure of Ownership and the Theory of the
Firm," Journal of Law and Economics, University of Chicago Press,
26(2), 375-90
Demsetz, H., Lehn, K. (1985). The Structure of Corporate Ownership: Causes and
Consequences. Journal of Political Economy, 93(6), 1155-77.
Denis, D. K., McConnell, J. J. (2003). International Corporate Governance. Journal of
financial and quantitative analysis, 38(01), 1-36.

REFERENCES
Depoers F. (2000) A cost benefit study of voluntary disclosure: Some
empirical evidence from French listed companies. European
Accounting Review, 9, 245-263.
Deumes R., (2008) Corporate risk reporting: A content analysis of
narrative risk disclosures in Prospectuses. Journal of Business
Communication, 45(2), 120-157.
Deumes R., W. R. Knechel, (2008) Economic Incentives for Voluntary
Reporting on Internal Risk Management and Control Systems
Auditing: A Journal of Practice & Theory, 27(1), 3566
Dhaliwal D., K. Lee, N. Fargher, (1991) The association between
unexpected earning and abnormal security returns in the presence
of financial leverage, Contemporary Accounting Research, 8, 20-41.
Dhaliwal D., S. Reynolds, (1994) The effect of the default risk of debt on
the earning response coefficient, The Accounting Review, 69 (2),
412-419.
Dhaliwal, D., D. Guenther, M. Trombley, (2000). Inventory accounting
method and earningsprice ratios. Contemporary Accounting
Research, 16(3), 419-436.
Diamond, D. W. (1985). Optimal Release of Information By Firms. The Journal of Finance,
40(4), 1071-1094.
Diamond, D., R. Verrecchia, (1991) Disclosure, liquidity, and the cost of
capital. The Journal of Finance, 66, 13251355.
Dietrich, J.R., S.J. Kachelmeier, D.N. Kleinmuntz, T.J. Linsmeier, (2001).
Market efficiency, bounded rationality, and supplemental business
reporting disclosures. Journal of Accounting Research 39 (2), 213
268.
Dimitropoulos, P E., A.D. Asteriou (2010). "The effect of board composition
on the informativeness and quality of annual earnings: Empirical
evidence from Greece," Research in International Business and
Finance, 24(2), 190-205
Djankov, S., R. La Porta, F. Lopez-de-Silanes, A. Shleifer. (2002). The
regulation of entry. Quarterly Journal of Economics, 117(1): 1-37.
Dobler, M. (2008). Incentives for risk reporting-A discretionary disclosure
and cheap talk approach. The International Journal of Accounting,
43(2), 184206.
Dobler, M., K. Lajili, D. Zeghal, (2011). Attributes of corporate risk
disclosure: An international investigation in the manufacturing
sector. Journal of International Accounting Research, 10(2), 122.
Doidge C., G. A. Karolyi, R. M. Stulz (2007) Why do countries matter so
much for corporate governance?, Journal of Financial Economics,
86, 139
Dolley, J. C., (1933) Common Stock Split-Ups Motives and Effects,
Harvard Business Review, 12(1), 7081.

REFERENCES
Donnelly, R., C. Lynch, (2002). The Ownership Structure of UK Firms and
the Informativeness of Accounting Earnings. Accounting and
Business Research, 32, 245-257.
Donnelly, R., M. Walker, (1995) Share Price Anticipation of Earnings and
the Effect of Earnings Persistence and Firm Size. Journal of Business
Finance and Accounting, 22 (1):5-18.
Donnelly, R., Mulcahy, M. B. (2008). Board Structure, Ownership, and Voluntary Disclosure
in Ireland. Corporate Governance: An International Review, 16(5), 416429.
Doupnik T.S., S.B. Salter (1995), External environment, culture and
accounting practice: A preliminary test of a general model of
international accounting development, International Journal of
Accounting, 30, 189-207.
Dumontier, P. B. Raffournier, (1998) Why Firms Comply Voluntarily with
IAS: An Empirical Analysis with Swiss Data. Journal of International
Financial Management and Accounting, 9(3), 216-45.
Dumontier, P., B. Raffournier, (2002) Accounting and capital markets: a
survey of the European evidence, The European Accounting
Review, 11(1), 119 151
Durnev, A., R., Morck, B. Yeung, (2004). Value enhancing capital
budgeting and firm specific return variation, Journal of Finance,
59(1), 65-105.
Dye, R. A. (1985). Disclosure of Non-proprietary Information. Journal of
Accounting Research, 23(1), 123-145
Dye, R. A. (1986). Proprietary and Non-proprietary Disclosures. The
Journal of Business, 59 (2), 331-366.
Dye, R. A. (2001). An Evaluation of Essays on Disclosure and the Disclosure Literature in
Accounting. Journal of Accounting and Economics, 32(1-3), 181235
Dye, R., S. Sridhar, (2002). Resource Allocation Effects of Price Reactions
to Disclosures, Contemporary Accounting Research19, 385-410.
Easton, P., Harris, T., (1991). Earnings as an explanatory variable for
returns. Journal of Accounting Research 29, 1936
Easton, P., Harris, T., Ohlson, J., (1992). Aggregate accounting earnings
can explain most of security returns: the case of long event
windows. Journal of Accounting and Economics 15, 119142.
Easton, P., Zmijewski, M., (1989). Cross-sectional variation in the stock
market response to accounting earnings announcements. Journal of
Accounting and Economics 11, 117141.
Eccles, R.G., Hertz, R.H., Keegan, M. Phillips, D.M.H. (2001). The Value
Reporting Revolution: Moving Beyond the Earnings Game, Wiley,
USA
Echambadi, R., Hess, J.D., (2007). Mean-centering does not alleviate
collinearity problems in moderated multiple regression models.
Marketing Science, 26 (3), 438445.

REFERENCES

Ecker, F., Francis, J., Kim, I., Olsson, P. and Schipper, K., (2006). A returnsbased representation of earnings quality. The Accounting Review
81 (4), 749-780.
Eggleton I.R.C., An Y., H. Davey, (2011) "Towards a comprehensive
theoretical framework for voluntary IC disclosure", Journal of
Intellectual Capital, 12 (4), 571 585
El Safti A. (2007). Financial sector reforms in the Arab countries.
Economic and Technical Dept. AMF. Abu Dhabi. UAE.
El-Gazzar, S. M., (1998), Pre-disclosure Information and Institutional
Ownership: A Cross-Sectional Examination of Market Revaluations
During Earnings Announcement Periods, The Accounting Review,
73(1), 119-129.
Elouafa, K. (2007). La publication volontaire des flux de trsorerie par les dirigeants:
incitations et consquences sur la valeur de la firme. Thse en Sciences de gestion,
Universit De Paris 8, Vincennes, Saint Denis.
Elshandidy, T., I Fraser, K. Hussainey, (2014). What drives mandatory and
voluntary risk reporting variations across Germany, the UK and the
US?. British Accounting Review, forthcoming
Elshandidy T., I. Fraser, K. Hussainey, (2013). Aggregated, voluntary, and
mandatory risk disclosure incentives: Evidence from UK FTSE allshare companies. International Review of Financial Analysis, 30,
320333
Elshandidy, T., Neri, L., (2015), Corporate Governance, Risk Disclosure
Practices, and Market Liquidity: Comparative Evidence from the UK
and Italy. Corporate Governance: An International Review.
Forthcoming
Ely, K., Waymire, G., (1999). Accounting standard-setting organizations
and earnings relevance: longitudinal evidence from NYSE common
stocks, 19271993. Journal of Accounting Research 37, 293318.
Elzahar H., Hussainey K, (2012) Determinants of narrative risk disclosures
in UK interim reports. The Journal of Risk Finance, 13 (2), 133 - 147
Eng, L. L., Mak, Y. T. (2003). Corporate governance and voluntary disclosure. Journal of
Accounting and Public Policy, 22(4), 325-345.
Eng, L.L. H.K. Teo (2000), The Relation between Annual Report
Disclosures, Analysts Earnings Forecasts and Analyst Following:
Evidence from Singapore, Pacific Accounting Review, 11(2), 219239
Espinosa, M., M. Tapia, M. Trombetta, (2005), Disclosure and liquidity,
Working Paper, Business Economics Series 02
Fama, E. (1965),"The Behavior of Stock Market Prices". Journal of Business
38: 34105
Fama, E. (1970). "Efficient Capital Markets: A Review of Theory and

REFERENCES
Empirical Work". Journal of Finance 25 (2): 383417.
Fama, E. F. (1980). Agency Problems and the Theory of the Firm. Journal of Political
Economy, 88(2), 288-307.
Fama, E. F., Jensen, M. C. (1983). Separation of ownership and control. Journal of
Finance, 25383.
Fama, E., Fisher, L., Jensen, M., Roll, R., (1969). The adjustment of stock
prices to new information. International Economic Review 10, 1-21.
Fan, J. P. H., Wong, T. J. (2002). Corporate ownership structure and the informativeness of
accounting earnings in East Asia. Journal of Accounting and Economics
33(3),401-425.
Farooq, O., R. Abdel Bari, (2013) Effect of property rights on the
relationship between legal traditions and corporate governance:
evidence from the MENA Region, Int. J. Business Governance and
Ethics, 8(3), 224241.
Farooq O., M. Derrabi, (2012). Effect of corporate governance mechanisms on the
relationship between legal origins and cost of debt. African Journal of Business
Management, 6(7), 2706-2715.
FASB (2001). Improving business reporting: insights into enhancing
voluntary disclosure steering committee report. Business Reporting
Research Project.
Feltham, G A, Pae J, (1999), "Analysis of the Impact of Accounting Accruals
on Earnings Uncertainty and Response coefficient", Journal of
Accounting, Auditing & Finance, 199-220.
Ferreira, D. A. Ferreira, M. Raposo, Clara C., (2011). "Board structure and
price informativeness," Journal of Financial Economics, 99(3), 523545.
Ferreira, M., M. Massa, P. Matos, (2010). Shareholders at the gate?
Institutional investors and cross-border mergers and acquisitions.
Review of Financial Studies, 23, 601644.
Fields T. D., T. Z. Lys, L. Vincent (2001). Empirical research on accounting
choice. Journal of Accounting and Economics, 31, 255307.
Financial Accounting Standards Board. (2010). Conceptual framework of
financial reporting. Chapter 1, The objective of general purpose
financial reporting, and Chapter 3, Qualitative characteristics of
useful financial information. Statement of Financial Accounting
Concepts No. 8. Norwalk, CT.
Firth, M. (1976) The impact of earnings announcements on the share
price behaviour of similar type firms, The Economic Journal, 86
(342), 296-306.
Firth, M., P. M. P. Fung, O. M. Rui, (2007), Ownership, two-tier board
structure, and the informativeness of earnings-Evidence from
China, Journal of Accounting and Public Policy, 26, 463-496
Forker, J. J. (1992). Corporate Governance and Disclosure Quality. Accounting and Business

REFERENCES
Research, 22(86), 111-124.
Foster, G. (1977) Quarterly accounting data: time-series properties and
predictive-ability results, The Accounting Review, 52(1), 1-21.
Foster, G.R. (1986), Financial Statement Analysis, 2nd ed., New-Jersey,
Prentice Hall.
Foster, J. (2003). The FASB and the Capital Markets. Stamford, CT: FASB.
Francis, J., K. Schipper, (1999). Have financial statements lost their
relevance? Journal of Accounting Research, 37, 319353.
Francis, J., Nanda, D., P. Olsson (2008) Voluntary disclosure, earnings
quality and cost of capital. Journal of Accounting Research, 46(1),
53-99.
Francis, J., Ryan L., Olsson P. K., Schipper (2004). Costs of equity and
earnings attributes. The Accounting Review, 79: 967-1010.
Francis, J., Ryan L., Olsson P., K. Schipper (2005) The market pricing of
accruals quality. Journal of Accounting and Economics, 39: 295327.
Frank, K. A., Sykes, G., Anagnostopoulos, D., Cannata, M., Chard, L., Krause, A., &
McCrory, R. (2008). Does NBPTS Certification Affect the Number of Colleagues a
Teacher Helps With Instructional Matters? Educational Evaluation and Policy
Analysis, 30(1), 3-30.
Frank, K.A., (2000). Impact of a confounding variable on a regression
coefficient. Sociological Methods and Research, 29, 147194.
Frank, K.A., Maroulis, S., Duong, M., and Kelcey, B. (2013). What would it
take to Change an Inference?: Using Rubins Causal Model to
Interpret the Robustness of Causal Inferences. Education,
Evaluation and Policy Analysis. 35: 437-460.
Frankel, R., M. McNichols, G.P.Wilson (1995) Discretionary Disclosure and
External Financing, The Accounting Review, 70(1), 135-15.
Freeman, R., (1987). The association between accounting earnings and
security returns for large and small firms. Journal of Accounting and
Economics 9, 195228.
Freeman, R., S. Tse, (1992) A nonlinear model of security price responses
to unexpected earnings. Journal of Accounting Research 30, 185
209.
Friedman, M. (1953) The methodology of positive economics. In Essays
in positive Economics, University of Chicago Press, 3-43.
Frolov, M. (2007). Why do we need mandated rules of public disclosure for banks? Journal
of Banking Regulation, 8(2), 177-191.
Fuller, J., M. C., Jensen, (2002). Just say no to Wall Street. Journal of
Applied Corporate Finance, 14(4), 41-46.
Furman, J., Stiglitz, J. E. (1998). Economic Crises: Evidence and Insights from East Asia.
Brookings Papers on Economic Activity, 29(2), 1-136.

REFERENCES
Gabrielsen, G., Gramlich, J. D., Plenborg, T. (2002). Managerial Ownership, Information
Content of Earnings, and Discretionary Accruals in a NonUS Setting. Journal of
Business Finance & Accounting, 29(7-8), 967-988.
Gaver, J. J., Gaver, K. M. (1993), Additional Evidence on the Association
Between the Investment Opportunity Set and Corporate Financing,
Dividend, and Compensation Policies. Journal of Accounting and
Economics, 16, 125-160.
Gelb, D. S., P. Zarowin (2002). Corporate disclosure policy and the
informativeness of stock prices. Review of Accounting Studies, 7
(1): 3352.
Gelb, S.D., (2000). Corporate signaling with dividends, stock repurchases,
and accounting disclosures: an empirical study. Journal of
Accounting, Auditing & Finance, 99121
Ghazali, N. A. M. (2007). Ownership structure and corporate social
responsibility disclosure: some Malaysian evidence. Corporate
Governance, 7(3), 251-266.
Ghazali, N.A.M., Weetman P., (2006). Perpetuating traditional influences:
Voluntary disclosure in Malaysia. Journal of International Accounting
Auditing and Taxation, 15 (2), 226-248.

Gibbins, M. Richardson A., J. Waterhouse (1992), The management of


Financial discoure: theory and perspectives, Fondation de recherche
de lassociation des comptables gnraux licencis du Canada,
Monographie de recherche n20.
Gibbins, M., Richardson, A., Waterhouse J., (1990), The management of
corporate financial disclosure: opportunism, ritualism, policies, and
processes, Journal of Accounting Research, 28(1), 121-143.
Gibson, M. S. (2003). Is Corporate Governance Ineffective in Emerging Markets? The
Journal of Financial and Quantitative Analysis, 38(1), 231-250
Gillan, S. L. (2006). Recent Developments in Corporate Governance: An Overview. Journal
of Corporate Finance, 12(3), 381-402.
Giner, B. (1997) The influence of company characteristics and accounting
regulation on information disclosed by Spanish firms, European
Accounting Review, 6, 45-68.
Glendon M. A., P. Wright-Carozza, C. Picker (2008) Comparative Legal
Traditions in a Nutshell. 3rd edn, St Paul, MN: West.
Globerman S., Shapiro D., (2003). "Governance infrastructure and US
foreign direct investment," Journal of International Business Studies,
34(1), 19-39.
Goodstein, J., Gautam, K., Boeker, W. (1994). The effects of board size and diversity on
strategic change. Strategic Management Journal, 15(3), 241-250.
Gray, R.H., Kouhy, R., S. Lavers, (1995), Constructing a research database

REFERENCES
of social and environmental reporting by UK companies: A
methodological note, Accounting, Auditing and Accountability
Journal, 8(2), 78-101.
Gray, S. J. (1988). Towards a Theory of Cultural Influence on the Development of Accounting
Systems Internationally. Abacus, 24(1), 1-15.
Griffin, C.H., Williams, T.H. (1960). Measuring adequate disclosure,
Journal of Accountancy, 109, 43 48.
Grossman S. (1981), The informational Role of Warranties and Private
Disclosure about Product Quality, The Journal of Law and
Economics, p. 461-485
Guercio, D. D., J. Hawkins, (1999). "The motivation and impact of pension
fund activism," Journal of Financial Economics, 52(3), 293-340.
Gul, F. A., L.K. Wah (2002). Insider Entrenchment, Board Leadership Structure and
Informativeness of Earnings. Working Paper, City University of Hong Kong.
Available at: http://papers.ssrn.com/abstract=304399
Gul, F. A., Leung, S. (2004) Board leadership, outside directors expertise
and voluntary corporate disclosures. Journal of Accounting and
Public Policy, 23(5), pp. 351379.
Gul, F. A., S. G. Lynn, J. S. L. Tsui. (2002). Audit quality, management
ownership and the informativeness of accounting earnings. Journal
of Accounting, Auditing and Finance, 25-49.
Guo, R., B. Lev, N. Zhou. (2004) Competitive costs of disclosure by
biotech IPOs. Journal of Accounting Research, 42 (2), 319-355.
Habib, A. (2008). The role of accruals and cash flows in explaining security returns: Evidence
from New Zealand. Journal of International Accounting, Auditing and Taxation,
17(1), 51-66.
Habib, A., Azim, I. (2008). Corporate governance and the value-relevance of accounting
information: Evidence from Australia. Accounting Research Journal, 21(2), 167-194.
Hagerman, R.L., M.E. Zmijewski, P. Shah, (1984) The Association between
the Magnitude of Quarterly Earnings Forecast Errors and Risk
Adjusted Stock Returns, Journal of Accounting Research, 22(2), 526540.
Hail, L. (2002). The impact of voluntary corporate disclosures on the exante cost of capital for Swiss firms. European Accounting Review,
11 (4): 741 773
Hair, J., Black, W., Babin, B., Anderson, R.,Tatham, R. (2006), Multivariate Data Analysis,
6 ed, Pearson Educational International -Prentice Hall, Upper Saddle River- New
Jersey.
Hand, J. (1990). A Test of the Extended Functional Fixation Hypothesis.
Accounting Review, 65:740-763.
Haniffa, R., T. Cooke (2002), Culture, corporate governance and disclosure
in Malaysian corporations, ABACUS, 36(3), 317- 773.
Hanouz M.D., S. El Diwany, T. Yousef (2007) The Arab World

REFERENCES
Competitiveness Report: Sustaining the Growth Momentum. the
World Economic Forum available at:
https://fr.scribd.com/doc/103020678/Arab-World-CompetitivenessReport-2007
Harris, M.S. (1998) The association between competition and managers
business segment reporting decisions, Journal of Accounting
Research, 36(1): 111-28.
Harrison, B., Moore, W. R. (2012). Stock Market Efficiency, Non-Linearity, Thin Trading and
Asymmetric Information in MENA Stock Markets. Economic Issues Journal Articles,
17(1), 77-93.
Hausman, J. A. (1978). Specification tests in
econometrics. Econometrica: Journal of the Econometric Society,
1251-1271.
Hassan, O., Marston, C. (2010). Disclosure Measurement in the Empirical Accounting
Literature - A Review Article. Economic.
Hassaan M. (2013) Corporate Governance and Compliance with IFRSs: MENA Evidence.
Cambridge Scholars Publishing
Hauswald, R., Marquez, R. (2006). Competition and Strategic Information Acquisition in
Credit Markets. Review of Financial Studies, 19(3), 967-1000.
Hawkamah (2010) SOE Governance Reforms in the MENA Region. CG
Report. http://www.oecd.org/daf/ca/corporategovernanceofstateownedenterprises/44824001.pdf
Hayes, A. F., K. Krippendorff, (2007).Answering the call for a standard
reliability measure for coding data. Communication Methods and
Measures, 1, 77-89.
Hayes, R.M. and Lundholm, R. (1996) Segment reporting to the capital
market in the presence of a competitor, Journal of Accounting
Research, 34(2): 261-79
Hayn, C., (1995). The information content of losses. Journal of
Accounting and Economics 20, 125153.
Healy, P., Hutton, A., Palepu, K., (1999). Stock performance and
intermediation changes surrounding sustained increases in
disclosure. Contemporary Accounting Research, 16, 485- 520.
Healy, P., Palepu, K., (2001). Information asymmetry, corporate disclosure
and the capital markets: A review of the empirical disclosure
literature. Journal of Accounting & Economics 31, 405440.
Heflin, F., K.W Shaw. J.J. Wild (2002). Disclosure quality and market
liquidity. Working Paper, Robert H. Smith School of Business,
College Park, MD.
Higgs (2003). Independent Review of Non-Executive Directors. ICAEW. available at:
http://www.icaew.com/en/library/subject-gateways/corporate-governance/codes-andreports/higgs-report
Ho, L. J., Taylor, M. E. (2007). An empirical analysis of triple bottom-line

REFERENCES
reporting and its determinants: evidence from the United States and
Japan. Journal of international Financial Management and
Accounting, 18(2), 123-150.
Ho, S.S.M., K.S. Wong, (2001). A study of the relationship between
corporate governance structure and the extent of voluntary
disclosure. Journal of International Accounting Auditing & Taxation,
10 (2), 139-156.
Hodgson, A., Stevenson C. P. (2000). Accounting Variables and Stock Returns: The Impact of
Leverage. Pacific Accounting Review, 12(2), 37-64.
Hoechle, D. (2007). Robust standard errors for panel regressions with
crosssectional dependence. Stata Journal, 7(3), 281-312.
Holm, C., Schler, F. (2010). Reduction of Asymmetric Information Through Corporate
Governance Mechanisms The Importance of Ownership Dispersion and Exposure
toward the International Capital Market. Corporate Governance: An International
Review, 18(1), 32-47.
Holmstrm, B., Tirole, J. (1993), Market liquidity and performance
monitoring, Journal of Political Economy, 101 (4), 678 709.
Holsti, O R. (1969), Content Analysis for the Social Sciences and
Humanities, Addison-Wesley Publishing Company, Reading, MA.
Holthausen, R., (1990). Accounting method choice: opportunistic
behavior, efficient contracting, and information perspectives.
Journal of Accounting and Economics 12, 207218.
Hope, O. (2003). Accounting policy disclosures and analysts forecasts.
Contemporary Accounting Research, 20 (2): 295321.
Hopkins, P. (1996) The effect of financial statement classification of
hybrid financial instruments on financial analysts stock price
judgments. Journal of Accounting Research, 34 (Supplement): 33
50.
Hopwood, W., S., J.C. McKeown, (1985) The incremental informational
content of interim expenses over interim sales, Journal of
Accounting Research, 23(1), 161 174.
Hossain M, S. M. Courtenay., S. Luo (2006), The effect of voluntary
disclosure, ownership Structure and proprietary cost on the return
future earnings relation. Pacific-Basin Finance Journal, 14,501521
Hossain, M. (2008). The extent of disclosure in annual reports of banking
companies: The case of India. European Journal of Scientific
Research, 23(4), 660681.
Hossain, M., Reaz, M. (2007). The determinants and characteristics of voluntary disclosure
by Indian banking companies. Corporate Social Responsibility and Environmental
Management, 14(5), 274-288.
Hossain, M., Tan, L.M., M. Adams, (1994) Voluntary disclosure in an
emerging capital market: Some empirical evidence from companies

REFERENCES
listed on the Kuala Lumpur Stock Exchange. The International
Journal of Accounting, 29, 334351.
Hou, K., Robinson, D. T. (2005). Industry Concentration and Average Stock Returns. The
Journal of Finance, 61(4). 19271956
Hsiao, C. (2013). Analysis of Panel Data, third edition. Cambridge
university press.
Hughes, J. S. (1989). Discussion of The valuation of initial public offerings .
Contemporary Accounting Research, 5(2), 519-525.
Hughes, P. (1986), Signalling by direct disclosure under asymmetric
information, Journal of Accounting and Economics, 119-142
Hussainey and Mouselli (2010) Disclosure quality and stock returns in the
UK. Journal of Applied Accounting Research, 11(2), 154-174.
Hussainey K. (2004) A study of the ability of (partially) automated
disclosure scores to explain the information content of annual report
narratives for future earnings. PhD, Manchester University, England.
Hussainey K., M. Walker, (2009) The effects of voluntary disclosure and
dividend propensity on prices leading earnings. Accounting and
Business Research, 39(1), 37-55.
Hussainey, K., Al-Najjar, B. (2011) Future-oriented narrative reporting:
determinants and use. Journal of Applied Accounting Research 12
(2), pp. 123-138
Hussainey, K., T. Schleicher, M. Walker, (2003). Undertaking large-scale
disclosure studies when AIMR-FAF ratings are not available: the case
of prices leading earnings. Accounting and Business Research, 33
(4): 275294.
Iatridis, G. (2008). Accounting disclosure and firms' financial attributes:
Evidence from the UK stock market. International Review of
Financial Analysis, 17(2), 219-241.
ICAS (1988). Making Corporate Reports Valuable, Edinburgh: Institute of
Chartered Accountants of Scotland.
Institute of Chartered Accountants In England and Wales (ICAEW). (2002).
Financial reporting of risk: Proposals for a Statement of Business
Risk. London.
Irvine, P. J., Pontiff, J. (2009). Idiosyncratic Return Volatility, Cash Flows, and Product
Market Competition. Review of Financial Studies, 22(3), 1149-1177.
Ismail, W. A. W., Kamarudin, K. A., Van Zijl, T., and Dunstan, K. (2013).
Earnings quality and the adoption of IFRS-based accounting
standards: Evidence from an emerging market. Asian Review of
Accounting, 21(1), 53-73.
Jacobson, R., Aaker, D., (1993). Myopic management behavior with
efficient, but imperfect, financial markets: a comparison of
information asymmetries in the U.S. and Japan. Journal of
Accounting and Economics, 16(4), 383-405

REFERENCES
Jaggi, B., Low, P. Y. (2000). Impact of Culture, Market Forces, and Legal System on
Financial Disclosures. The International Journal of Accounting, 35(4), 495-519.
Jenkins E. (1994). An information highway in need of capital
improvements. Journal of Accountancy. 177, 77-82.
Jennings, R.J., J.R. Robinson, R.B. Thompson, L. Duvall, (1996), The
Relation between Accounting Goodwill Numbers and Equity Values,
Journal of Business Finance and Accounting, 23, 513- 533.
Jensen M.C. (1993) The Modern Industrial Revolution, Exit, and the Failure
of Internal Control Systems, Journal of Finance, 48 (3), 831-880
Jensen, M. C. (1986). Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers.
American Economic Review, 76(2), 323-29.
Jensen, M. C., Murphy, K. J. (1990). Performance Pay and Top-Management Incentives.
Journal of Political Economy, 98(2), 225-64.
Jensen, M.C., W.H. Meckling, (1976). Theory of the firm; managerial
behaviour, agency costs and ownership structure, Journal of
Financial Economics, 3, 305-360.
Jeny A. C. Jeanjean T, (2007). Levels of voluntary disclosure in IPO
prospectuses: an empirical analysis, Review of Accounting and
Finance 6(2), 131-149.
Jiambalvo, J., Rajgopal, S., Venkatachalam, M. (2002). Institutional Ownership and the
Extent to which Stock Prices Reflect Future Earnings. Contemporary Accounting
Research, 19(1), 117-145.
Jiang, H., Habib, A. (2009). The impact of different types of ownership concentration on
annual report voluntary disclosures in New Zealand. Accounting Research Journal,
22(3), 275-304.
Jiang, H., Habib, A., Hu, B. (2011). Ownership concentration, voluntary disclosures and
information asymmetry in New Zealand. The British Accounting Review, 43(1),
39-53.
John, K., Senbet, L. W. (1998). Corporate governance and board effectiveness. Journal of
Banking and Finance, 22(4), 371-403.
Jones, M. J. and Shoemaker, P. A. (1994), "Accounting narratives: a review
of empirical studies of content and readability", Journal of
Accounting Literature, 13, 142-184.
Jorgensen, B. N. & Kirschenheiter, M. T. (2012). Interactive discretionary
disclosures. Contemporary Accounting Research, 29(2), 382-397.
Jorion, P. (2002). How Informative Are ValueatRisk Disclosures? The Accounting Review,
77(4), 911-931.
Kajter, P. (2006). Risk disclosure of listed firms in Germany: a
longitudinal study. Paper presented at the 10th Financial Reporting
& Business Communication Conference Cardiff Business school
Kajuter, P., (2001). Risikoberichterstattung: Empirische Befunde und der
Entwurf des DRS 5. Der Betrieb 54 (3), 105111.

REFERENCES
Kanagaretnam, K., Lobo, G. J., Whalen, D. J. (2007). Does good corporate governance reduce
information asymmetry around quarterly earnings announcements? Journal of
Accounting and Public Policy, 26(4), 497-522.
Kar, M., Nazlolu, ., Ar, H., 2011. "Financial development and economic
growth nexus in the MENA countries: Bootstrap panel granger
causality analysis," Economic Modelling, 28(1), 685-693.
Karamanou, I., Vafeas, N. (2005). The Association between Corporate Boards, Audit
Committees, and Management Earnings Forecasts: An Empirical Analysis. Journal of
Accounting Research, 43(3), 453-486.
Kaufmann, D., Kraay, A., Zoido-Lobaton, P. (2000), Governance Matters: From Measurement
to Action, Finance and Development, 37 (2), Washington DC: International Monetary
Fund.
Kelton, A., Yang, Y. (2008) The impact of corporate governance on Internet
financial reporting. Journal of Accounting & Public Policy, 27 (1), 6287.
Kent, P., Ung, K. (2003). Voluntary Disclosure of Forward-Looking Earnings
Information in Australia, Australian Journal of Management, 28 (3):
273-286.
Khlif, H., Samaha, K., Hussainey, K. (2014). The impact of board characteristics and audit
committee on voluntary disclosure: a meta-analysis
Journal of International Accounting, Auditing and Taxation, forthcoming.
Kim, O., R. Verrecchia, (1994). Market liquidity and volume around
earnings announcements. Journal of Accounting and Economics,
17:41-67.
King R. R., D. E. Wallin (1995) Experimental tests of disclosure with an
opponent. Journal of Accounting and Economics 19, 139-167.
King, R., D. Wallin, (1991). Voluntary Disclosures When Sellers Level of
Information is unknown Journal of Accounting Research, 29(1), 96108.
King, R., G. Pownall, G. Waymire. (1990). Expectations adjustment via
timely earnings forecast disclosure: Review, synthesis, and
suggestions for future research. Journal of Accounting Literature 9,
113-44.
Klein, A. (2002). Audit Committee, Board of Director Characteristics, and Earnings
Management. Journal of accounting and economics, 33(3), 375-400.
Kobeissi, N. (2005). Impact of Governance, Legal System and Economic Freedom on Foreign
Investment in the MENA Region. Journal of Comparative International Management,
8(1).
Kochhar, R., David, P. (1996). Institutional Investors and Firm Innovation: A Test of
Competing Hypotheses. Strategic Management Journal, 17(1), 73-84.
Koldertsova A. A. (2011) The second corporate governance wave in the
Middle East and North Africa. Organization for Economic Co-

REFERENCES
Operation and Development (OECD) Journal: Financial Market
Trends, 10(2), 219-226.
Koldertsova A. A. (2012) The Role of MENA Stock Exchanges in Corporate
Governance. Organization for Economic Co-Operation and
Development (OECD), Paris. Available at:
http://www.oecd.org/corporate/ca/theroleofmenastockexchangesinco
rporategovernance.htm

Konishi, N., Ali, M. M. (2007). Risk reporting of Japanese companies and its association with
corporate characteristics. International Journal of Accounting, Auditing and
Performance Evaluation, 4(3), 263-285.
Kormendi, R., Lipe, R., (1987). Earnings innovations, earnings persistence
and stock returns. Journal of Business 60, 323345.
Kothari S. P., X. Li, J. E. Short (2009) The Effect of Disclosures by
Management, Analysts, and Business Press on Cost of Capital,
Return Volatility, and Analyst Forecasts: A Study Using Content
Analysis. The Accounting Review, 84(5), 1639-1670.
Kothari, S. P. (2001). Capital markets research in accounting. Journal of
Accounting and Economics 31 (13): 105231.
Kothari, S. P., J. L. Zimmerman. (1995). Price and return models. Journal
of Accounting and Economics, 20 (2): 155192.
Kothari, S., (1992). Priceearnings regressions in the presence of prices
leading earnings: earnings level versus change specifications and
alternative deflators. Journal of Accounting and Economics 15, 173
302.
Kothari, S., R. Sloan, (1992). Information in prices about future earnings:
implications for earnings response coefficients. Journal of
Accounting and Economics 15, 143171.
Kravet T.D, V. Muslu (2013) Textual Risk Disclosures and Investors Risk
Perceptions. Review of Accounting Studies, 18(4), 1088-1122
Kreps, D., (1990). A course in microeconomic theory. Princeton
University Press, Princeton, NJ.
Krippendorff, K. (2010). Krippendorff's alpha. In N. Salkind
(Ed.), Encyclopedia of research design. (pp. 669-674). Thousand
Oaks, CA: SAGE Publications, Inc
Krippendorff, K., (2004). Content Analysis: An Introduction to Its
Methodology, second ed. Sage, Beverly Hills.
La Porta R., Lopez-de-Silanes F., Shleifer A., (1999), Corporate ownership
around the world. Journal of Finance, 54(2), 471-518.
La Porta R., Lopez-de-Silanes F., Shleifer A., (2008). "The Economic
Consequences of Legal Origins," Journal of Economic Literature,

REFERENCES
American Economic Association, 46(2), 285-332
La Porta R., Lopez-de-Silanes F., Shleifer A., and Vishny R.W. (2000).
Investor Protection and Corporate Governance, Journal of Financial
Economics 58(1), 3-27.
La Porta R., Lopez-de-Silanes F., Shleifer A., Vishny R., (1997), Legal
determinants of external finance. Journal of Finance, 52, 1131
1150.
La Porta R., Lopez-de-Silanes F., Shleifer A., Vishny R., (1998), Law and
finance. Journal of Political Economy, 106, 11131155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., (2006). What works in
securities laws? Journal of Finance, 61, 133
Labelle (2001). la recherche en comptabilit financire et la recherche
en comptabilit de gestion In faire de la recherche en
comptabilit financire coordonn par Pascal Dumontier et Robert
Teller, dit avec la FNEGE, Vuibert.
Lagoarde-Segot, T., Lucey, B. M. (2008). The Capital Markets of the Middle East and North
African Region: Situation and Characteristics. Emerging Markets Finance and Trade,
44(5), 68-81.
Lajili, K., and D. Zeghal. (2005). Content analysis of risk management
disclosures in Canadian annual reports. Canadian Journal of
Administrative Sciences, 22 (2): 125142.
Lakhal F., (2008) "Stock market liquidity and information asymmetry
around voluntary earnings disclosures: New evidence from France",
International Journal of Managerial Finance, 4(1), 60 - 75
Lakhal F., (2009) "Does corporate disclosure policy change financial
analysts' behavior? Evidence from France", Journal of Accounting &
Organizational Change, 5 (3), 344 361
Lakhal, F. (2005). Voluntary earnings disclosures and corporate
governance: Evidence from France, Review of Accounting and
Finance, 4(3), 64-85.
Lakhal, F. (2006). Les publications volontaires des rsultats et le comportement des analystes
financiers en France. 27me Congrs de LAFC. Available at http://halshs.archivesouvertes.fr/halshs-00558062
Laksmana, I. (2008). Corporate Board Governance and Voluntary Disclosure of Executive
Compensation Practices. Contemporary Accounting Research, 25(4), 1147-1182.
Lambert, R., Leuz C., R. Verrecchia (2007). Accounting information,
disclosure, and the cost of capital. Journal of Accounting Research
45: 385420.
Lambert, R. A., Leuz, C., & Verrecchia, R. E. (2012). Information
asymmetry, information precision, and the cost of capital. Review of
Finance, 16(1), 1-29.
Lang, L. H.P., So, R.W. (2002). Ownership structure and economic

REFERENCES
performance. Working Paper.
Lang, M. H., R. J. Lundholm, (1993). Cross-sectional determinants of
analyst ratings of corporate disclosures. Journal of Accounting
Research, 31 (2): 24671.
Lang, M. H., R. J. Lundholm, (1996). Corporate disclosure policy and
analyst behaviour. The Accounting Review, 71 (4): 46792.
Larcker, D. F., Richardson, S. A., Tuna, I. (2007). Corporate governance,
accounting outcomes, and organizational performance. The
Accounting Review, 82(4), 963-1008.
Larcker, D. F., Rusticus, T. O. (2010). On the use of instrumental variables in accounting
research. Journal of Accounting and Economics, 49(3), 186-205.
Lee C.W.J. (1988), Inventory accounting and earnings/price ratios: A
puzzle. Contemporary Accounting Research, 5: 371388.
Leftwich R . (1980). Market Failure Fallacies and Accounting Information."
Journal of Accounting and Economics, 1980, 2(3), pp. 193-211.
Leftwich, R. W.,. R. L, Watts, J. L. Zimmerman (1981): "Voluntary Corporate
Disclosure: The Case of Interim Reporting", Journal of Accounting
Research, 19(3), 50-77.
Lehn, K., Patro, S., Zhao, M. (2007). Governance indexes and valuation:
Which causes which? Journal of Corporate Finance, 13, 907928.
Leung S., B. Horwitz (2004). Director ownership and voluntary segment disclosure: Hong
Kong evidence. Journal of International Financial Management & Accounting, 15
(3): 235-260.
Leuz, C. (2004) Proprietary versus Nonproprietary Disclosures: Evidence
from Germany in C. Leuz, D. Pfaff and A. G. Hopwood: The
Economics and Politics of Accounting: International Perspectives on
Research, Trends, Policy, and Practice. Oxford University Press, pp.
164-97.
Leuz, C., Verrecchia, R. (2000). The Economic consequences of increased
disclosure. Journal of Accounting Research, 38 (Supplement): 91
124.
Lev, B. (1989) On the usefulness of earnings: lessons and directions from
two decades of empirical research. Journal of Accounting Research,
27 (Supplement): 15392.
Lev, B., Penman S.H. (1990), Voluntary Forecast Disclosure, NonDisclosure, and Stock Prices, Journal of Accounting Research, 28(1),
49-76.
Lev, B., Zarowin, P., (1999). The boundaries of financial reporting and how
to extend them. Journal of Accounting Research 37, 353385.
Leventis, S., & Weetman, P. (2004). Voluntary disclosure in an emerging
capital market: some evidence from the Athens Stock Exchange.

REFERENCES
Advances of International Accounting, 17, 227-250.
Li, F. (2008). Do stock market investors understand the risk sentiment of
corporate annual reports? Unpublished paper, University of
Michigan.
Li, F. (2010). The information content of forward-looking statements in
corporate filings-A nave Bayesian machine learning approach.
Journal of Accounting Research, 48, 1049-1102.
Li, F., Srinivasan, S. (2011). Corporate governance when founders are
directors. Journal of financial economics, 102(2), 454-469.
Li, J., Pike, R., Haniffa, R. (2008). Intellectual capital disclosure and
corporate governance structure in UK firms. Accounting and
Business Research, 38(2), 137-159.
Li, X., (2010). The impacts of product market competition on the quantity
and quality of voluntary disclosures. Review of Accounting Studies
15, 663-711.
Lichtenberg, F R., Pushner, G. M., (1994). "Ownership structure and
corporate performance in Japan," Japan and the World Economy,
6(3), 239-261.
Lim, C. Y., P. M. Tan, (2007) Value relevance of value-at-risk disclosure.
Review of Quantitative Finance and Accounting, 29(4), 353370.
Linck, J. S., Nett