Earnings Management, Corporate Governance and the Market Performance of Seasoned Equity Offerings

Ken M.L. Ching, Michael Firth, Oliver M. Rui Department of Accountancy The Hong Kong Polytechnic University Hung Hom, Kowloon Hong Kong Current Version September 2002

Abstract This paper examines whether pre-issue discretionary current accruals predict post-issue earnings performance and returns. We find evidence suggesting that offering firms borrow future income to manage earnings in pre-issue years and consequently earnings decrease in post-issue year 2. The information content of pre-issue discretionary current accruals is reflected in one-year buy and hold abnormal returns for the rights offering firms. Our paper also examines whether the incidence of earnings management around SEOs depends on corporate governance structures. Our results show that SEO firms that have larger board size have a higher degree of earnings management around SEOs. This result is consistent with Jensens (1993) view that smaller boards provide more of a controlling function than do larger boards. JEL classification: G32; G34 Keywords: Earnings management; Corporate governance; Seasoned equity offerings

Corresponding author. Tel.: (852) 2766 7062; Fax.: (852) 2330 9845. Email address: acmaf@inet.polyu.edu.hk

Earnings Management, Corporate Governance and the Market Performance of Seasoned Equity Offerings in Hong Kong
Abstract This paper examines whether pre-issue discretionary current accruals predict post-issue earnings performance and returns. We find evidence suggesting that offering firms borrow future income to manage earnings in pre-issue years and consequently earnings decrease in post-issue year 2. The information content of pre-issue discretionary current accruals is reflected in one-year buy and hold abnormal returns for the rights offering firms. Our paper also examines whether the incidence of earnings management around SEOs depends on corporate governance structures. Our results show that SEO firms that have larger board size have a higher degree of earnings management around SEOs. This result is consistent with Jensens (1993) view that smaller boards provide more of a controlling function than do larger boards. JEL classification: G32; G34 Keywords: Earnings management; Corporate governance; Seasoned equity offerings

Earnings Management, Corporate Governance and the Market Performance of Seasoned Equity Offerings

1. Introduction Seasoned equity offerings (SEOs) provide a direct incentive to manage earnings. To the extent managers can undetectably increase reported earnings, they can improve the terms on which their firms shares are sold to a third party. In support of this, Rangan (1998) and Teoh et al. (1998) have shown that managers manipulate earnings in periods around SEOs. Their studies argue that managers overstate earnings before SEOs because of opportunism. Shivakumar (2000) also reports earnings management by firms making SEOs but he argues that investors are able to identify these discretionary accruals and price the stock according to fundamentals. He argues that earnings management at the time of an SEO is rational behavior by the managers.1 Most research on earnings management, and all the studies cited above, use data from the U.S. In comparison, there is a dearth of research in other countries. Examining whether earnings management takes place around SEO announcements is important from the viewpoint of investors, analysts, and regulators. Teoh et al. (1998) and Rangan (1998) show that poor post-issue performance can be explained partially by the pre-issue earnings management of seasoned new issuers. 2 Investors and analysts can therefore avoid investing in issuers that dramatically underperform their non-issuing peers by choosing firms with negative or low discretionary current accruals. Thus they use information contained in the pre-offering accounting accruals to discriminate among seasoned equity issuers. 3 Regulators should pay attention to accounting practices that misrepresent or otherwise induce a false picture about firms performances at the time of SEOs. If investors cannot see through the accounting practices, they will be induced to buy the stocks of those firms with higher false reported earnings at the time of SEOs. Regulators have a responsibility to identify situations where earnings management has a greater likelihood of occurring. Investors, analysts, and regulators can also take corporate governance structures, such as Big 6 auditors and family control,

Shivakumar (2000) calls this argument the managerial response hypothesis. This contrasts with the overvaluation Empirical studies in the U.S. have documented poor post-issue earnings performance (Hansen and Crutchley, 1990;

hypothesis that says managers have incentives to issue new equity when the firm is overvalued.
2

McLaughlin et al., 1996) and poor post-issue stock returns (Loughran and Ritter, 1995; Spiess and Affleck-Graves, 1995). Brous (1992) and Jain (1992) report that security analysts significantly lower their forecasts on short-term earnings of issuing firms in response to SEO announcements.
3

Discretionary accruals in the period surrounding the offering predict a portion of the subsequent poor earnings and stock

price performance.

into consideration as these characteristics appear to affect the magnitude of earnings management that accompany SEOs.4 In this study we use data from Hong Kong to explore whether discretionary accruals is used to boost reported earnings before the issue of SEOs and whether any identified earnings management is associated with future stock returns. If the market fails to understand that earnings management only temporarily increase the earnings, offering firms will be overvalued. When high pre-issue earnings are not maintained, the market subsequently revalues the firm down to a level justified by fundamental factors. We examine whether the situation in Hong Kong is consistent with the earnings management hypothesis.5 The prevalence and extent of earnings management is likely to be a function of the specific environment that the firm operates in. For example, corporate governance mechanisms and monitoring and oversight activities will affect the use of earnings management, and the importance of these factors vary across firms and across national jurisdictions. Because corporate governance, the legal environment, and monitoring activities are far different in Hong Kong than in the U.S., research studies using American data have limited relevance for Hong Kong. Firms in Hong Kong are predominantly controlled by families. This control is enhanced through pyramid structures and cross-share holdings among firms (Claessens et al., 2000). When an owner effectively controls a firm, he or she also controls the production of the firms accounting information and its reporting policies. The owner reports accounting information more for self-interest purposes than to reflect the firms true underlying economic performance. In this study, we investigate whether the family-controlled firm manipulates earnings to a greater extent than the non-family-controlled firm in order to influence the market perceptions of the value of the firm around SEOs. Boards structured to be independent of management are best able to perform their independent oversight functions (Fama, 1980; Fama and Jensen, 1983). We examine whether independent boards have the ability to withstand pressure from the firm to manipulate earnings around SEOs. U.S. evidence (Shleifer and Vishny, 1986; Jensen, 1993; Pushner, 1995) shows that large block shareholders have incentives to monitor management and serve as an additional control mechanism. We examine whether the presence of a blockholder is negatively related to the degree of earnings management around SEOs. Empirical studies in the U.S. (Francis et al., 1999; Becker et al., 1998) show that Big-6-audited firms have lower amounts of estimated discretionary accruals. This finding is consistent with Big 6 auditors constraining aggressive and potentially opportunistic reporting of accruals.
4 5

Following the 1998 merger of Coopers & Lybrand and Price Waterhouse, the Big 6 are now the Big 5 accounting firms. The earnings management hypothesis (Teoh et al., 1998) predicts that issuers have unusually high income-increasing

accounting adjustments pre-issue and unusually poor earnings and stock return performance post-issue. This pattern of earnings management may induce the overvaluation of SEOs.

We therefore examine whether Big-6-audited firms have lower amounts of earnings management around SEOs than non-Big-6-audited firms. 2. Institutional Background 2.1 Seasoned Equity Offerings Seasoned equity offerings include placings, rights issues, open offers, consideration issues, warrants exercised and share option schemes. Of these, rights issues and placings constitute the vast majority of seasoned equity issuance in Hong Kong. The huge amounts of funds raised through Hong Kongs equity markets in recent years reflect the popularity of both the IPO and SEO. Seasoned equity offerings have received comparatively little attention especially when viewed against the copy and airtime devoted to some of the larger IPO flotations. McGuinness (1992) notes that seasoned equity offerings feature significantly in the local market and raise more funding than unseasoned equity offerings in aggregate terms. He also notes that 20-25% of IPO firms return to the market to make capital-raising seasoned offerings in each of the three years following the IPO. Nearly half of IPO firms return to the market within three years of initial listing. 2.1.1 Placings Placings include both private and public placings. In a public placement (also called firmcommitment underwritten offers or underwritten public offers), the entire issue is sold directly to the underwriter, who, in turn, resells the shares to investors. In a private placement, the firm sells a block of securities to a single or small group of investors, usually fewer than five. McGuinness (1992) notes that over 51% of all seasoned issue funds, and over 38% of all equity funds (including IPOs) raised over the 1995-1997 period were through placings. Young and Chiang (1997) state that it is common for listed issuers in Hong Kong to initiate private placings to independent third parties or a select group of investors. Wruck (1989) finds that the four-day (-3, 0) abnormal announcement period return of U.S. firms making private placements is 4.41% and the two-day (-1, 0) abnormal return averages 1.89%. Using a different sample of U.S. SEOs, Hertzel and Smith (1993) find positive and significant abnormal returns of 1.72% for the four-day time period (-3, 0). Expanding the Myers and Majluf (1984) model to include private equity issues, Hertzel and Smith argue that private placements will be used if the net present value of the investment opportunity is greater than the cost of informing private investors about the true firm value. They suggest that the willingness of private placement investors to commit funds to a firm, together with managements decision to forgo a public issue, conveys managements belief that the firm is undervalued. In cross-sectional regression analyses, Hertzel and Smith find evidence that is consistent

with their information cost theory. They also provide results that are interpreted as being consistent with a signaling motivation for SEOs. Hence, firms signal undervaluation by placing equity privately. 2.1.2 Rights Offerings In an uninsured rights offering (also called a non-underwritten rights offering), short-lived warrants are issued to current shareholders on a pro-rata basis. The firm avoids paying underwriter fees by marketing the new shares through current shareholders, but the gross proceeds of the offering are not guaranteed. In rights with standby underwriting (also called an underwritten rights offering), the firm also issues rights but a standby underwriter guarantees the gross proceeds by agreeing to purchase and resell all shares not marketed through the rights offer. In Hong Kong, rights issues are usually fully underwritten. The share price reaction to rights issue announcements is generally found to be negative in U.S. studies. Hansen (1989) finds significant negative two-day announcement abnormal returns of -1.21% for utilities and -2.61% for industrials underwritten rights offerings. Eckbo and Masulis (1992) find that the two-day announcement period abnormal return is a statistically significant 1.03% for industrial issuers and 0.53% for utility issuers in standby rights offerings. Singh (1997) finds negative two-day announcement abnormal returns of -1.07% for underwritten rights offerings. Slovin et al. (2000) indicate that rights offerings by British firms are a negative signal of firm value. They report a statistically significant two-day average excess return of 2.9% for British insured rights offerings for the 1986-1994 period while the two-day return for British uninsured rights offerings is a statistically significant 5%. The negative price reaction to the rights issue announcement is consistent with the asymmetric information release model of Myers and Majluf (1984). Hansen (1989) states that the underwritten rights offerings abnormal price behavior is consistent with the transaction-cost hypothesis proposed by Kraus and Stoll (1972). The transaction-cost hypothesis assumes that security demand is perfectly elastic but that security sales entail transaction costs. 2.2 Earnings Management Schipper (1989) states that earnings management is a purposeful intervention in the external financial reporting process, with the intent of obtaining some private gain. Healy and Wahlen (1999) state that earnings management occurs when managers use judgment in financial reporting and in structuring transactions to alter financial reports to either mislead some stakeholders about the underlying economic performance of the company or to influence contractual outcomes that depend on reported accounting numbers. Because auditing is imperfect, managements use of judgment creates opportunities for earnings management in which managers choose reporting methods and estimates that do not

accurately reflect their firms underlying economics.6 Earnings management threatens the credibility of financial reporting. The objective of earnings management is to mislead stakeholders about the underlying economic performance of the firm. This can arise if managers believe that at least some stakeholders cannot decipher or undo the earnings management. Several earnings management methods are available to managers such as real decisions like asset sales (Bartov, 1993) and changes in R&D expenditure (Bushee, 1998); and financial reporting decisions like accounting method changes (Watts and Zimmerman, 1986) and accruals choices (Healy, 1985). Manipulation of operating accruals is likely to be a favored instrument for opportunistic earnings management because they have no direct cash flow consequences and are relatively difficult to detect. In this paper, we use discretionary current accruals as our proxy for earnings management. Research in U.S. has identified the presence of earnings management in a number of settings. These settings include management buyouts, takeovers, IPOs, taxation, government controls, and SEOs. In the case of takeovers, IPOs, and SEOs, managers want to portray the financial standing and performance of the firm in the best possible light and to this end they make judicious use of accounting accruals. 2.3 Corporate Governance 2.3.1 Corporate Governance and Board of Directors According to Cadbury (1992), corporate governance is the system by which companies are directed and controlled. Corporate governance centers on processes designed to ensure that directors and managers of companies are held properly accountable to their shareholders. They must perform their duties with integrity and be subject to checks and balances which prevent abuse of power. Directors must keep shareholders informed of matters affecting shareholder interests. The shareholders role in governance is to appoint the directors and the auditors, and to satisfy themselves that an appropriate governance is in place. The board of directors should adequately reflect and represent the interests of different shareholder groups. They are responsible for the governance of their companies. The responsibilities of the board include setting the companys strategic aims, providing the leadership to put them into effect, supervising the management of the business and reporting to shareholders on their stewardship. The boards actions are subject to laws, regulations, and the shareholders approval at general meetings. For good governance, the quality and composition of the companys board of directors are the most important key elements. The quality of the board depends on a number of factors, such as its size,
6

Audits can lessen earnings management behavior. The auditing process is not perfect and there is some asymmetry

between what the manager knows about accruals and what the auditor is able to cost-effectively glean and report. The auditor has a limited influence on accounting choice because earnings management is not illegal.

its composition in terms of number and proportion of independent non-executive and executive directors, the chairman of the board, powers and position of the CEO, and the merits and qualifications of individual directors, nominee directors, and non-executive directors. The key to good corporate governance is a well functioning board of directors with a core of independent and reputable nonexecutive directors. Boards will be better served by professionals (e.g. accountants and lawyers) whose expertise matches the companys strategic requirements. Mok et al. (1992) show that high concentration of stocks in the hands of large family groups is common in the Hong Kong stock market. They show that information on family ownership of Hong Kong companies is not directly available because of interlocking and pyramid shareholdings. Many listed companies are family controlled and some are closely connected among themselves. A high percentage of their directors are also the major or controlling shareholders. It is the very nature of these familycontrolled companies that the management function is confined to a small group of individuals. Frequently, the chairman of the board is also the controlling shareholder and founder of the listed group. While family-managed firms display greater entrepreneurial activity, they exercise unchallenged control and have the power to garner personal enrichment at the expense of the firm. The degree to which these infractions occur depends on the level of integrity of the controlling groups. In companies where there is a single controlling shareholder or family group, corporate governance is closely bound up with the protection of minority shareholders. 2.3.2 The Stock Exchanges Initiatives The Stock Exchange of Hong Kongs (SEHK) effort to promote corporate governance culminated in the introduction of regulations in 1992 that govern the conflict of interests between connected parties and the company. To reflect the influence of directors of listed companies and to highlight their responsibilities, a Code of Best Practice was established in 1993 to enhance the accountability of directors to shareholders.7 In order to improve the effectiveness of corporate boards, the Exchange required listed companies to have at least two independent non-executive directors on their boards by December 31, 1994. Rules regarding disclosure of directors emoluments and senior management compensation were introduced in the Exchange Listing Rules in 1994. The Exchange also issued some guidelines for independent non-executive directors including their qualifications, appointments, and roles. To strengthen the observance of the Code of Best Practice, listed companies have been required since 1996 to include a statement in their interim and annual reports on compliance with the Code. These
7

The Code appears in Appendix 14 to the Exchange Listing Rules and is a set of voluntary code and guidelines concerning

the board of directors to which listed issuers should aim. All listed issuers are encouraged to devise their own code of practice taking into account the interests of stakeholders.

provisions increase transparency and enable the Exchange to ensure greater accountability of listed companies to the investing public. The Stock Exchange requires firms to disclose the presence or otherwise of an audit committee and to detail the directors education and background. Under the revised Code of Best Practice, an audit committee should have a minimum of two members and they should be appointed from among the non-executive directors.8 Listed issuers have to report in both their interim and annual reports their compliance with the setting up of audit committees, or any non-compliance therewith, for accounting periods commencing on or after 1 January 1999. The corporate governance system in Hong Kong relies on self-regulation through market selfdiscipline and peer pressure to ensure compliance. Self-regulation will be more effective if compliance comes from a recognition by market participants that it is in their long-term interests to maintain good standards of conduct. There are other legislative and regulatory provisions designed to encourage or enforce good corporate governance. These include the Takeover Code (administered by the Securities and Futures Commission (SFC)), the Securities Ordinance, the Securities (Disclosure of Interests) Ordinance, and the Securities (Insider Dealing) Ordinance. 3. Theory and hypotheses development 3.1 Earnings Management around SEOs Earnings management is typically accomplished by shifting income from future periods to the present. In particular, without violating current accounting rules, firms can accelerate the recognition of revenues and defer the recognition of certain expenses. If offering firms borrow future income to manage earnings in pre-issue years, then earnings will increase in pre-issue years and decrease subsequently. Consequently, we hypothesize that there is a negative relation between pre-issue earnings management and post-issue earnings performance. If investors regard the earnings growth of SEO firms in pre-issue periods as permanent, they will overvalue these firms. However, when high pre-issue earnings are not maintained in post-issue periods, the market subsequently revalues the firm down to a level justified by fundamental factors. In a more general context, Sloan (1996), using U.S. data, finds that investors overvalue firms with high levels of accounting accruals and this overvaluation is reversed in future periods. He therefore argues that investors are unable to see through the earnings management at the time it is done. Based on the above

The independent non-executive directors should be independent of management and free from any business or other

relationship which could materially interfere with the exercise of their independent judgment as a Committee member. According to Rule 3.11(1) the independent non-executive directors fees and shareholdings should not amount to more than 1% of the total issued share capital.

arguments and the empirical evidence provided by Teoh et al. (1998) and Rangan (1998), we hypothesize that there is a negative relation between pre-issue earnings management and post-issue stock returns. 3.2 Family Control and Earnings Management around SEOs The SEHK (1994c) states that most board members of Hong Kong firms either belong to the management group or to the family owning the firm. The Hong Kong Society of Accountants (HKSA) (1997) reports that a substantial portion of Hong Kong firms is family owned. Further evidence comes from Mok et al. (1992) and Lam et al. (1994), who show that a substantial number of Hong Kong firms are family owned and controlled. Mok et al. (1992) find that about 55.88% of the market value of firms listed on the SEHK belong to a few large family groups. A family controls the enterprises by appointing members of the family as directors in the companies that are family owned. The appointment of outside directors in family controlled companies is likely to be influenced more by the personal ties of prospective directors to the controlling family and the likelihood of their support for the managements philosophy and policies. In the presence of family ownership and control, outside directors independence and monitoring roles are impaired. Though the employment of the CEO must be approved by the board of directors, the family, by occupying a majority of seats on the board, can easily appoint a family member or someone with a close connection with the family, to this important position. When a family owner effectively controls a firm, he or she also controls the production of the firms accounting information and its reporting policies. The owner reports accounting information more for self-interest purposes rather than to reflect the firms true underlying economic performance. Due to the larger stake invested, we hypothesize that the owner of the family-controlled firm manipulates earnings to a larger extent than the manager of the non family-controlled firm in order to influence the market perceptions of the value of the firm at the time of a SEO. In this study, we hypothesize that the relation between family-controlled firms and the degree of earnings management around SEOs is positive. 3.3 Corporate Governance and Earnings Management around SEOs 3.3.1 Board Independence Fama and Jensen (1983) argue that the board of directors is the highest internal control mechanism responsible for monitoring the actions of top management. Fama (1980) suggests that the domination by top management on the board of directors can lead to collusion and a transfer of shareholder wealth. Williamson (1984) states that managers have a huge informational advantage due to their full-time status and insider knowledge, so that the board of directors can easily become an instrument of management, thereby sacrificing the interests of shareholders. Fama (1980) and Fama and Jensen (1983) state that outside directors have incentives to carry out their monitoring tasks and to avoid

colluding with top managers to expropriate shareholder wealth. They also state that outside directors have incentives to develop reputations as experts in decision control because the external market for their services prices them according to their performance and independence as outside directors. Therefore the inclusion of outside directors increases a boards independence and its ability to monitor top management effectively in agency settings arising from the separation of corporate ownership and decision control. problems. Empirical research from the U.S. provides evidence about the importance of including outside directors on the board for purposes of monitoring management in acute agency settings. Lee et al. (1992) state that in management buyouts, shareholder wealth increases when boards are dominated by outside directors. Kosnik (1987, 1990) states that firms resisting greenmail payments have more outside directors relative to boards of firms not resisting greenmail payments. Weisbach (1988) argues that the turnover of CEOs of poorly performing firms is greatest when the proportion of non-executive directors is high. These findings are consistent with the hypothesis that outside directors help alleviate agency conflicts between shareholders and upper management. Note, however, the contrasting evidence of Klein (2000); she finds an insignificant relation between board independence and earnings management. In this study, we hypothesize that there is a negative relationship between board independence and the degree of earnings management around SEOs. 3.3.2 Size of the Board Jensen (1993) believes that a smaller board of directors plays a controlling function whereas a larger board of directors is easier for the CEO to control. In contrast, Chaganti et al. (1985) believe that large boards are valuable for the breadth of their services. They suggest that a larger board is more effective in preventing corporate failure. Beasley (1996) indicates that as board size increases, the likelihood of financial statement fraud increases. This result is consistent with Jensens (1993) view that smaller boards provide more of a control function than do larger boards. Following these arguments, we hypothesize that board size is positively related to the degree of earnings management around SEOs. 3.3.3 Blockholders Shleifer and Vishny (1986) and Jensen (1993) show that large block shareholders have incentives to monitor management and serve as an additional control mechanism. Pushner (1995) states that the presence of such shareholdings may lead to monitoring, intervention, or some other disciplinary influence on management, which serves to align managements and shareholders interests. Large blockholders Corporate boards generally include outside members who act as arbiters in disagreements among internal managers and ratify decisions that involve potentially serious agency

also affect board of director composition by influencing the selection of members. Gilson (1990) finds that increases in outside director representation on the board of directors is associated with increases in blockholder ownership in periods subsequent to a firms poor performance. We hypothesize that a blockholder is negatively related to the degree of earnings management around SEOs. 3.4 Big 6 Auditors and Earnings Management around SEOs Accounting earnings provide value-relevant information to investors about the firm. Earnings are not directly observable by outsiders. What is observed is managements disclosure of reported earnings. Reported earnings are uncertain because of noise (flexibility and subjectivity in accounting rules) and potential bias due to opportunistic and self-interested behavior. Therefore, the economic role of audits is to provide credibility to managements reported earnings. In the U.S., Francis et al. (1999) show that Big-6-audited firms have lower absolute amounts of estimated discretionary accruals. This finding is consistent with Big 6 auditors constraining aggressive and potentially opportunistic reporting of accruals. Becker et al. (1998) also report lower discretionary accruals for Big-6-audited companies. Chung et al. (2000) demonstrate that Big 6 auditors restrain the use of discretionary accruals when managers have incentives to boost earnings such as at the time of making equity issues. SEOs provide a direct incentive for managers to manage earnings (Rangan, 1998). If managers can undetectably increase reported earnings before offerings, they can raise more capital at more favorable terms than if earnings were not managed. In this study, we hypothesize that there is a negative relation between Big-6-audited firms and the degree of earnings management around SEOs. 4. Data and Methodology 4.1 Sample Selection The sample of SEO firms is identified using the Stock Exchange Fact Book and the Securities Journal over the period from 1993 through 1998. We obtain stock price and accounting data from the PacificBasin Capital Markets (PACAP) Databases and annual reports. The announcement dates of SEOs are identified from the Securities Journal and the Wise News. As we base our earnings management measures on cross-sectional regressions using industry data, industry-year portfolios with less than six observations are dropped from the sample. For the sample, we only include firms with enough data to calculate accruals measures, earnings performance, and stock returns. If a firm has multiple issues in a year, we include only the earliest issue to avoid using overlapping data in the analysis. Firms in the financial services industry are not included in our sample because discretionary accruals estimation is problematic for these firms. Firms in the utilities industry are excluded to avoid the possible confounding

10

effect of regulation on managers discretionary accrual choices. Finally, firms are excluded if there are changes in fiscal year end during the sample period. Panel A, Table 1 shows the sample construction. ********************* Please Insert Table 1 here ********************* 4.2 Earnings Management Measures and Tests We proxy earnings management by the signed discretionary current accruals. Current accruals are used by Teoh et al. (1998), Rangan (1998), and, in supplementary tests, by Shivakumar (2000). Current accruals are used as managers have greater discretion over them (Guenther, 1994), vis--vis long term accruals, and most of the variation in total accruals is driven by current accruals (Sloan, 1996).9 Long term accruals are less likely to be manipulated as they are more visible to market participants. To estimate discretionary current accruals, we first compute current accruals for SEO firms i in year t as follows:
CAi ,t [ ARi ,t + INV i ,t + OCAi ,t ] [ APi ,t + TPi ,t + OCLi ,t ]

(1)

where CAi,t is current accruals; ARi,t is accounts receivables; INVi,t is inventory; OCAi,t is other current assets; APi,t is accounts payable; TPi,t is tax payable; OCLi,t is other current liabilities. We then estimate a modified Jones model10 to decompose current accruals (CAi,t) into discretionary current accruals (DCAi,t) and non-discretionary current accruals (NDCAi,t):
CA j , t TA j , t 1 SALES j , t 1 ) + a1 ( ) + j ,t TA j , t 1 TA j , t 1

= a0 (

(2)

where SALESj,t is the change in sales and TAj,t-1 is the beginning of years total assets. We estimate the modified Jones model on a cross-sectional basis in order to maximize our sample size and to avoid the survivorship bias problem inherent in the time-series approach. For each year and industry, we estimate

Jones (1999) argues that current accruals provide a more accurate basis for estimating discretionary earnings management

than aggregate accruals because the estimated discretionary portion of noncurrent accruals is less likely to reflect year-specific discretion.
10

Bartov et al. (2000) and Dechow et al. (1995) argue that the modified Jones model is the best model to date for measuring

earnings management.

11

regression parameters in Eq. (2) using an estimation sample j based on industry.11 We use the regression

0 and a 1 , from Eq. (2) to calculate non-discretionary current accruals. Thus: parameters, a
SALES i ,t ARi ,t 1 1 ( )+a ) + i ,t TAi ,t 1 TAi ,t 1

0 ( NDCAi ,t = a

(3)

Non-discretionary current accruals (NDCAi,t) is defined as the fitted value from Eq. (3). Following Dechow et al. (1995), we subtract the increase in accounts receivable from sales growth. Discretionary current accruals (DCAi,t), our measure of earnings management, is defined as the residual of Eq. (4). Thus:
CAi , t TAi , t 1

DCAi , t

NDCAi , t

(4)

4.2.1

Cross-Sectional Regressions of Post-Issue Earnings Performance to Pre-issue Discretionary Accruals To evaluate the ability of pre-issue accruals to explain the observed earnings performance in post-

issue years, we estimate two multivariate cross-sectional regressions of changes in return on assets (ROAi,1,2) in years 1 to 2, on year 1 variables. 12 Operating cash flows (OCFi,t-1) are scaled by beginning-period total assets. DCAi,t-1 is the experimental variable of interest. Control variables are preissue log market capitalization (LNMCi,t-1), log book-to-market (LNBMi,t-1), offering frequency dummy (ISSUEi) and security type dummy (SEOi) variables. ISSUEi is coded one (1) if the issuer makes more than one SEO in the six year period 1993 to 1998 and SEOi is coded one (1) if the issue is made by way of a placement and coded zero (0) if it is a rights offering. If SEO firms use discretionary accruals to shift income from the future, we expect a negative relation between pre-issue discretionary accruals and subsequent earnings changes. The regression model is:13
ROAi ,1, 2 = 0 + 1 DCAi ,t 1 + 2 NDCAi ,t 1 + 3OCFi ,t 1 + 4 LNMC i ,t 1 + 5 LNBM i ,t 1 + 6 ISSUE i + 7 SEO i + (5)
11 12

The SEO firm and other firms conducting an SEO in that year are not included in the regression. ROAi,t is computed as first differences in annual ROA. ROA is defined as income before extraordinary items scaled by We also include a complete set of industry and year dummies for regressions used in this paper. All results are

beginning assets less the proceeds from the SEO.

13

qualitatively the same as those without the dummies. Diagnostics (e.g. variance inflation factors and condition indices) indicate that multicollinearity is not a major problem in the regressions.

12

4.2.2

Cross-Sectional Regressions of Post-Issue Returns to Pre-Issue Discretionary Accruals In this section, we examine whether pre-issue discretionary accruals can predict post-issue returns.

We measure abnormal returns of the sample firms over one-year and two-year holding periods following the SEO announcement. These abnormal returns are calculated by using monthly value-weighted market returns of PACAP stocks with cash dividends reinvested. We compute the 12- and 24-month abnormal holding period returns for firm i, BHARi,12,24, by subtracting the holding period return of the benchmark portfolio from that of the sample firm over the corresponding holding horizon.
T T

BHARi ,12,24 [ (1 + ri ,k ) (1 + mi ,k )]
k =0 k =0

(6)

where T is the length of the holding period measured by number of months, which has a value of 12 and 24, and ri,k is the monthly rate of return on stock i for event month k relative to the announcement month (k=0). mi,k is the monthly value-weighted market returns of PACAP stocks with cash dividend reinvested. If the sample firm is delisted before the end of the holding period, its return for the rest of the holding period is assumed to be equal to the return of the benchmark portfolio to avoid survival biases in the analysis. To evaluate the ability of pre-issue accruals to explain the observed returns in post-issue years, we estimate a multivariate cross-sectional regression. The one- and two-year post-issue BHARi,12,24 are regressed on DCAi,t-1, the pre-issue discretionary current accruals. We predict a negative relationship as the discretionary accruals are expected to reverse in the years immediately after the issue. 14 Nondiscretionary accruals, NDCAi,t-1, and operating cash flow, OCFi,t-1, are also added to the model, along with four control variables. Control variables are pre-issue log market capitalization (LNMCi,t-1), log book-to-market (LNBMi,t-1), offering frequency dummy (ISSUEi), and security type dummy variables (SEOi).
BHAR i ,12 , 24 = 0 + 1 DCAi ,t 1 + 2 NDCAi ,t 1 + 3OCFi ,t 1 + 4 LNMC i ,t 1 + 5 LNBM i ,t 1 + 6 ISSUE i + 7 SEO i + 1

(7)

14

Note, Rangan (1998), Teoh et al. (1998), and Shivakumar (2000) find that accruals dont reverse in the first few years (or

quarters) after the SEO. They suggest that the threat of litigation influences firms to maintain positive earnings management for as long as possible. The need to keep accruals from reversing in the first few years is less urgent in Hong Kong due to the less litigious environment.

13

4.2.3

Cross-Sectional Regressions of Discretionary Current Accruals Both univariate and multivariate models are conducted to relate discretionary current accruals

(DCAi,t) to pre-issue family control (FAMCi,t-1), Big 6 auditors (AUDi,t-1), and corporate governance structures which include board independence (BDINDi,t-1), board size scaled by total assets (BSIZETAi,t-1) and blockholders (BLKi,t-1). Thus:
DCAi ,t = b0 + b1 FAMC i ,t 1 + b2 BDINDi ,t 1 + b3 BSIZETAi ,t 1 + b4 BLK i ,t 1 + b5 AUDi ,t 1 + b6 EEPi ,t 1 + b7 LEVi ,t 1 + b8 LSIZE i ,t 1 + b9 DCAi ,t 1 + b10 ISSUE i + b11 SEOi +

(8)

where FAMCi,t-1 is a dummy variable coded 1 if the firm is classified as family controlled. Firms are classified as family controlled if 10% or more of their outstanding common shares were held by a family and at least one family member was on the corporate board.15 BDINDi,t-1 is the percentage of the total number of board members who are outside directors. BSIZETAi,t-1 is the number of directors on the board scaled by total assets. BLKi,t-1 is a dummy variable taking the value of 1 if the firm has an external stockholder owning at least 10 percent of the outstanding shares and who is not affiliated with management. AUDi,t-1 is a dummy variable that takes the value of 1 if the auditor is a Big 6 accounting firm. We add six control variables to the models. Extreme earnings performance (EEPi,t-1) is the absolute change in the previous years income before extraordinary items divided by total assets. Financial leverage (LEVi,t-1) is total debt divided by total stockholders equity. Firm size (LSIZEi,t-1) is the log of beginning years assets. Discretionary current accruals (DCAi,t-1) is the previous years discretionary current accruals. ISSUEi is an offering frequency dummy and SEOi is a security type dummy variable. Firms that frequently issue equity may be less tempted to engage in earnings management because they realize they cannot repeatedly fool the market and, in fact, they may be penalized by such behaviors. Extreme earnings performance (EEPi,t-1), financial leverage (LEVi,t-1), and firm size (LSIZEi,t-1) are found to be significantly related to earnings management in U.S. studies (DeFond and Jiambalvo, 1994; Warfield et al., 1995; Dechow et al., 1995; Dechow et al., 1996; Becker et al., 1998; Bartov et al., 2000). Previous years discretionary current accruals (DCAi,t-1) is used to control the mean-reverting behavior of discretionary accruals. Since these six variables are used only as control variables, we do not predict their expected signs.

15

Prior studies (Mok et al., 1992; Lam et al., 1994) adopted the 10% or more threshold used by the SEHK to describe the

shareholdings of an individual as substantial. Hence, these studies used the 10% cut off point for classifying firms as family owned and controlled.

14

5. Empirical Results 5.1 Summary Statistics Panel B of Table 1 shows the distribution of placings and rights offerings by year of announcement. The number of placings issued is higher than the number of rights offerings issued in each sample year. The total funds raised for the sample period is HK$170,889.51m for placings and HK$43,007.16m for rights offerings. Panel C of Table 1 shows the distribution of placings and rights offerings by year of announcement and by industry. In summary, Table 1 indicates that a placing is more popular than a rights offering in terms of the number of issues. Furthermore, placings raise more funds than rights offerings in the sample period although the average proceeds per issue (HK$298.70 million for placements and HK$286.70 for rights issues) are about the same. Descriptive statistics for accruals are reported in Panel A of Table 2. Pre-issue discretionary current accruals have a mean (median) of 1.80% (0.00%). The large standard deviation of pre-issue discretionary current accruals relative to its mean suggests that the level of pre-issue discretionary current accruals varies widely over time and across firms. Panel B of Table 2 reports descriptive statistics for earnings performance and returns. The change in return on assets in year 1 and 2 (from year 1) have a mean (median) of 1.50% (3.30%) and 4.70% (1.50%). One-year and two-year buy and hold abnormal returns have a mean (median) of 16.30% (-29.30%) and 27.20% (-49.50%). The mean (median) of the sales growth in year 0 and 1 are 21% (7.90%) and 0.50% (3.50%) respectively. Panel C of Table 2 reports descriptive statistics for the corporate governance variables. About 73% and 81% of our sample firms are family-controlled firms and clients of Big 6 auditors, respectively. The average board in our sample contains eight directors, of whom approximately 38% are outsiders. Panel C also shows that 19% of the sample has at least one external blockholder whose stake exceeds 10%. Panel D reports descriptive statistics for control variables. The mean (median) of pre-issue market capitalization is $3355.84m ($478.50m), while the mean (median) of beginning years assets is $3934.16 ($672.44). Pre-issue financial leverage has a mean (median) of 1.20 (0.82) and a standard deviation of 1.57. The mean and median of pre-issue book-to-market ratios are 1.04 and 0.74. Panel D shows that 70% of our sample firms conducted more than one SEO in the six year period, and 79% of the sample SEOs are by way of placings. ********************* Please Insert Table 2 here *********************

15

The upper (lower) triangle of Table 3 reports Pearson (Spearman) correlation coefficients. Discretionary current accruals in year -1 is negatively correlated with discretionary current accruals in year 2 (rp=-0.135) and discretionary current accruals in year 0 is negatively correlated with discretionary current accruals in year 2 (rp=-0.143). The results indicate that discretionary current accruals reverses in year 2. Among other associations, the change in return on assets in year 1 is positively correlated with pre-issue operating cash flows (rp=0.176 and rs=0.198). The dummy variable representing familycontrolled firms is positively correlated with board independence (rp=0.174 and rs=0.166) but is negatively correlated with board size (rp=-0.175 and rs=-0.167). As board size is positively correlated with firm size which is measured by total assets (rp=0.471 and rs=0.351), we scale board size by total assets (BSIZETAi,t-1) in the regression equations. ********************* Please Insert Table 3 here ********************* 5.2 Pre-Issue Discretionary Accruals and Post-Issue Earnings Performance Table 4 shows the results of regression equation (5). Panel A relates to the overall results while panels B and C relate to the placing and rights issue samples, respectively. DCAi,t-1, our main experimental variable, has a negative and statistically significant (at the 0.10 level) coefficient in the model of ROAi,2. The negative sign implies that the transient effects of the positive discretionary current accruals pre the SEO reverse by two years after the issue. This result is consistent with our hypothesis of earnings management being used immediately prior to an SEO so as to boost reported earnings. There is no evidence of reversal of earnings management in the year after the SEO; Rangan (1998) and Teoh et al. (1998) also report that reversals take some time to occur. OCF, LNMC, and LNBM are significantly associated with ROA in year one while none of the control variables are significant in explaining ROA in year two. In panel B, there is no evidence of earnings management reversing in the first two years after the SEO. Other variables are positive and sometimes significant except for ISSUE, which has a negative sign. Panel C shows that DCAi,t-1 is negative and significant when explaining ROA in year 2. Our results suggest that some discretionary accounting accruals reverse in year 2 and so the oft-cited evidence that SEO firms suffer poor post-issue profitability is partly due to over-optimistic pre-issue earnings. By the very nature of earnings management, discretionary accruals will be expected to reverse and our findings

16

suggest that it takes about two years for this to happen. Note that the significant earnings management result is confined to the rights issue sample.16 **************************** Please Insert Table 4 and 5 here **************************** 5.3 Pre-Issue Discretionary Accruals and Post-Issue Returns Table 5 shows the regression results from regression equation (7). The main variable of interest, DCAi,t-1, has a negative coefficient in the six regressions. However, only in the rights issue subsample, and only for the one year returns, is DCAi,t-1 significant at conventional levels. Apart from this one instance, there is no statistically significant evidence that pre-issue discretionary current accruals fool the stock market and result in poor after-issue price performance. The stock market appears to price the impact of discretionary current accruals in share prices at the time of the SEO. This evidence contrasts with the results of Rangan (1998) and Teoh et al. (1998). Most of the control variables, as well as nondiscretionary current accruals, are not significant. ********************* Please Insert Table 6 here ********************* 5.4 Models of Discretionary Current Accruals In Panel A of Table 6, models 1 to 5 examine the univariate relationship between corporate governance characteristics and earnings management. Model 3 shows that board size scaled by total assets is significantly (at the 0.01 level) and positively related to discretionary current accruals. Model 6 shows the regression results from regression equation (8). Model 6 (6a) shows that the relation between board size scaled by total assets and discretionary current accruals is positive and significant at the 0.10 level for the overall sample (placing sub-sample). The results support our hypothesis that there is a positive relation between board size and the degree of earnings management around SEOs. A large board size appears to enhance the opportunistic use of discretionary accruals by management when they issue new equity. None of the other variables are significant in explaining the level of earnings management.

16

As a sensitivity test we include sales growth as an additional independent variable in equation (5). Issuing firms that

experience rapid sales growth are likely to attract new entrants into their industries and this may cause pressure on profit margins and hence profits. We find, however, that the inclusion of a sales growth variable, does not significantly change the results shown in Table 4.

17

Unlike studies in the U.S., there is no evidence that Big 6 auditors constrain the use of positive discretionary accruals. In Panel C, Model 6b shows that the relation between blockholders and discretionary current accruals is positive and significant at the 0.05 level for the rights offering sub-sample. Rights offering firms that have blockholders have a higher degree of earnings management around SEOs. This evidence does not support our hypothesis that blockholders act to constrain earnings management. ********************* Please Insert Table 7 here ********************* 6. Conclusions and Implications This study examines whether pre-issue discretionary current accruals predict post-issue earnings performance and returns. A number of interesting results emerge from our analyses. First, we find evidence of positive discretionary accruals being made in the year prior to the SEO. Second, we find a significant and negative relation between pre-issue discretionary current accruals and the change in return on assets in year 2 for both the overall sample and the rights offering sub-sample. These results imply that offering firms borrow future income to manage earnings in pre-issue years and then earnings decline in year 2. Third, we find a significant and negative relation between pre-issue discretionary current accruals and one-year buy and hold abnormal return for the rights offering firms. This finding shows that the information content of pre-issue discretionary current accruals is reflected in one-year buy and hold abnormal return for the rights offering firms. Fourth, our results show that SEO firms that have larger board have a higher degree of earnings management around SEOs. The results are consistent with Jensens (1993) view that smaller boards provide more of a controlling function than do larger boards. Our results indicate that managers engage in earnings management prior to announcing seasoned equity offerings. They do this in order to increase the proceeds from the SEO, for a given number of shares issued. Whether this earnings management fools the stock market is questionable. In the case of rights issues, earnings management is negatively associated with one-year stock returns; however, there is no significant relationship for the placing sub-sample. Overall, our findings suggest that investors need to be aware of possible earnings management by firms that make seasoned equity offerings.

18

REFERENCES Bartov, E. 1993. The Timing of Asset Sales and Earnings Manipulation. The Accounting Review 68 (4):840-855. Bartov, E., F. A. Gul, and J. S. Tsui. 2000. Discretionary-Accruals Models and Audit Qualifications. Journal of Accounting and Economics 30:421-452. Beasley, M. S. 1996. An Empirical Analysis of the Relation between the Board of Director Composition and Financial Statement Fraud. The Accounting Review 71 (4):443-466. Becker, C. L., M. L. DeFond, J. Jiambalvo, and K. R. Subramanyam. 1998. The Effect of Audit Quality on Earnings Management. Contemporary Accounting Research 15:1-24. Brous, P. A. 1992. Common Stock Offerings and Earnings Expectations: A Test of the Release of Unfavorable Information. Journal of Finance 47:1517-1536. Bushee, B. J. 1998. The Influence of Institutional Investors on Myopic R&D Investment Behavior. The Accounting Review 73 (3):305-333. Cadbury Committee Report. 1992. Report of the Cadbury Committee on the Financial Aspects of Corporate Governance. London: Gee. Chaganti, R. S., and S. S. Mahajan. 1985. Corporate Board Size, Composition, and Corporate Failures in Retailing Industry. Journal of Management Studies 22:400-417. Chung, R., M. Firth, and J. B. Kim. 2000. Auditor Conservatism, Asymmetric Monitoring and Discretionary Accrual Choices. Working paper, The Hong Kong Polytechnic University. Claessens, S., S. Djankov, and L. H. P. Lang. 2000. The Separation of Ownership and Control in East Asian Corporations. Journal of Financial Economics 58:81-112. Dechow, P. M., R. G. Sloan, and A. P. Sweeney. 1995. Detecting Earnings Management. The Accounting Review 70:193-226. Dechow, P. M., R. G. Sloan, and A. P. Sweeney. 1996. Causes and Consequences of Earnings Manipulation: An Analysis of Firms Subject to Enforcement Actions by the SEC. Contemporary Accounting Research 13 (1):1-36. DeFond, M. L., and J. Jiambalvo. 1994. Debt Covenant Violation and Manipulation of Accruals. Journal of Accounting and Economics 17:145-176. Eckbo, B. E., and R. W. Masulis. 1992. Adverse Selection and the Rights Offer Paradox. Journal of Financial Economics 32:293-332. Fama, E. F. 1980. Agency Problem and the Theory of the Firm. Journal of Political Economy 88:288308. Fama, E. F., and M. C. Jensen. 1983. Separation of Ownership and Control. Journal of Law and Economics 26:301-325.

19

Francis, J. R., E.L. Maydew, and H. C. Sparks. 1999. The Role of Big 6 Auditors in the Credible Reporting of Accruals. Auditing: A Journal of Practice & Theory 18 (2):17-34. Gilson, S. C. 1990. Bankruptcy, Boards, Banks, and Blockholders: Evidence on Changes in Corporate Ownership and Control when Firms Default. Journal of Financial Economics 27:355-387. Guenther, D. 1994. Measuring Earnings Management in Response to Corporate Tax Rate Changes: Evidence from the 1986 Tax Reform Act. The Accounting Review 69:230-243. Hansen, R. S. 1989. The Demise of the Rights Issue. Review of Financial Studies 1 (3):289-309. Hansen, R., and C. Crutchley. 1990. Corporate Earnings and Financings: An Empirical Analysis. Journal of Business 63:347-371. Healy, P. M. 1985. The Effects of Bonus Schemes on Accounting Decisions. Journal of Accounting and Economics 7:85-107. Healy, P. M., and J. M. Wahlen. 1999. A Review of the Earnings Management Literature and its Implications for Standard Setting. Accounting Horizons 13 (4):365-383. Hertzel, M., and R. L. Smith. 1993. Market Discounts and Shareholder Gains for Placing Equity Privately. Journal of Finance 48 (2):459-485. Hong Kong Society of Accountants (HKSA). 1997. Second Report of the Corporate Governance Working Group. Hong Kong: HKSA. Jain, P. C. 1992. Equity Issues and Changes in Expectations of Earnings by Financial Analysts. Review of Financial Studies 5 (4):669-684. Jensen, M. C. 1993. The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems. Journal of Finance 48:831-880. Jones, C. L. 1999. An Analysis of the Effectiveness of Discretionary Accrual Measures. Working paper, George Washington University. Jones, J. 1991. Earnings Management during Import Relief Investigation. Journal of Accounting Research 29 (2):193-228. Klein, A. 2002. Audit Committee, Board of Director Characteristics, and Earnings Management. Journal of Accounting and Economics 33:375-400. Kosnik, R. D. 1987. Greenmail: A Study of Board Performance in Corporate Governance. Administrative Science Quarterly 32:163-185. Kosnik, R. D. 1990. Effects of Board Demography and Directors Incentives on Corporate Greenmail Decisions. Academy of Management Journal 33:129-150.

20

Kraus, A., and H. R. Stoll. 1972. Price Impacts of Block Trading on the New York Stock Exchange. Journal of Finance 27:569-588. Lam, K., H. M. Mok, I. Cheung, and H. C. Yam. 1994. Family Groupings on Performance of Portfolio Selection in the Hong Kong Stock Market. Journal of Banking and Finance 18 (4):725-742. Lee, C. I., R. N. Rangan, and W. N. Davidson. 1992. Board Composition and Shareholder Wealth: The Case of Management Buyouts. Financial Management 21:58-72. Loughran, T., and J. Ritter. 1995. The New Issues Puzzle. Journal of Finance 50 (1):23-51. McGuinness, P. B. 1992. An Examination of the Underpricing of Initial Public Offerings in Hong Kong: 1980-90. Journal of Business Finance & Accounting 19 (2):165-186. McLaughlin, R., A. Safieddine, and G. Vasudevan. 1996. The Operating Performance of Seasoned Equity Issuers: Free Cash Flow and Post-Issue Performance. Financial Management 25 (4):41-53. Mok, H. M., K. Lam, and I. Cheung. 1992. Family Control and Return Covariation in Hong Kongs Common Stocks. Journal of Business Finance & Accounting 19 (2):277-293. Myers, S., and N. Majluf. 1984. Corporate Financing and Investment Decisions when Firms have Information that Investors do not have. Journal of Financial Economics 13:187-221. Pushner, G. M. 1995. Equity Ownership Structure, Leverage, and Productivity: Empirical Evidence from Japan. Pacific-Basin Finance Journal 3:241-255. Rangan, S. 1998. Earnings Management and the Performance of Seasoned Equity Offerings. Journal of Financial Economics 50:101-122. Schipper, K. 1989. Commentary: Earnings Management. Accounting Horizons 3 (4):91-102. Shivakumar, L. 2000. Do Firms Mislead Investors by Overstating Earnings before Seasoned Equity Offerings? Journal of Accounting and Economics 29 (3):339-371. Shleifer, A., and R. W. Vishny. 1986. Large Shareholders and Corporate Control. Journal of Political Economy 95:461-488. Singh, A.K. 1997. Layoffs and Underwritten Rights Offers. Journal of Financial Economics 43:105130. Sloan, R. 1996. Do Stock Prices Fully Reflect Information in Accruals and Cash Flows about Future Earnings? The Accounting Review 71:289-315. Slovin, M. B., M. E. Sushka, and K. W. L. Lai. 2000. Alternative Flotation Methods, Adverse Selection, and Ownership Structure: Evidence from Seasoned Equity Issuance in the U.K. Journal of Financial Economics 57:157-190. Spiess, K., and J. Affleck-Graves. 1995. Underperformance in Long-Run Stock Returns Following Seasoned Equity Offerings. Journal of Financial Economics 38:243-267.

21

Stock Exchange of Hong Kong Limited (SEHK). 2000. Rules Governing the Listing of Securities. Hong Kong: SEHK. Stock Exchange of Hong Kong Limited (SEHK). 1994c. The Way forward for the Stock Exchange of Hong Kong. Research and Planning Department, SEHK. Teoh, S. H., I. Welch, and T. J. Wong. 1998. Earnings Management and the Underperformance of Seasoned Equity Offerings. Journal of Financial Economics 50:63-99. Warfield, T. D., J. J. Wild, and K. L. Wild. 1995. Managerial Ownership, Accounting Choices, and Informativeness of Earnings. Journal of Accounting and Economics 20:61-92. Watts, R., and J. L. Zimmerman. 1986. Positive Accounting Theory. New York: Prentice-Hall. Weisbach, M. S. 1988. Outside Directors and CEO Turnover. Journal of Financial Economics 20:431460. Williamson, O. E. 1984. Corporate Governance. The Yale Law Journal 93:1197-1230. Wruck, K. H., 1989. Equity Ownership Concentration and Firm Value. Journal of Financial Economics 23:3-28. Young, L., and R. Chiang. 1997. The Hong Kong Securities Industry, 3rd ed. Hong Kong: The Stock Exchange of Hong Kong: Asia-Pacific Institute of Business, The Chinese University of Hong Kong.

22

TABLE 1 Descriptive Statistics for the Sample and the Distribution of Placings and Rights Offerings by Year of Announcement and Industry. Panel A: The numbers of placings and rights offerings in the sample from 1993-1998. Placings Rights offerings Total No. of events 572 150 722 No. of firms represented 294 124 418 Subsequent SEO announcement within one year 84 27 111 Duplicate offers 6 0 6 Combination offers 2 2 4 Missing company identity 2 0 2 Missing announcement date 62 0 62 Missing data in database 64 26 90 Firms in the financial and utility 40 13 53 industries No. of "clean" events 312 82 394 No. of firms represented 236 77 313 Panel B: Distribution of placings and rights offerings by year of announcement. Type of Offering Total funds Funds raised Rights Funds raised Year All Raised (HK$mil) Placings (HK$mil) offerings (HK$mil) 1993 138 34370.37 114 23725.39 24 10644.98 1994 59 15948.98 41 11406 18 4542.98 1995 35 7799.11 25 6542.3 10 1256.81 1996 114 42360.79 90 37145.18 24 5215.61 1997 270 92452.02 222 75916.5 48 16535.52 1998 106 20965.4 80 16154.14 26 4811.26 Total 722 213896.67 572 170889.51 150 43007.16 Panel C: Distribution of placings and rights offerings by year of announcement and industry. Year All Finance Utilities Properties Enterprises Industrials Hotels Others 1993 114 (24) 9 (2) 0 (2) 12 (7) 43 (7) 48 (5) 1 (0) 1 (1) 1994 41 (18) 2 (0) 0 (1) 11 (1) 13 (4) 15 (12) 0 (0) 0 (0) 1995 25 (10) 2 (0) 0 (0) 2 (0) 12 (3) 8 (7) 1 (0) 0 (0) 1996 90 (24) 6 (3) 0 (0) 15 (2) 37 (11) 30 (7) 0 (1) 2 (0) 1997 222 (48) 15 (2) 2 (0) 32 (6) 75 (23) 94 (17) 1 (0) 3 (0) 1998 80 (26) 3 (3) 1(0) 13 (1) 30 (14) 29 (8) 0 (0) 4 (0) Total 572 (150) 37 (10) 3 (3) 85 (17) 210 (62) 224 (56) 3 (1) 10 (1) Distribution of rights offerings is in brackets.

23

TABLE 2 Descriptive Statistics for Seasoned Equity Offerings Variable Mean Panel A: Earnings management DCAt-1 0.018 NDCAt-1 0.043 0.033 OCFt-1 Std Dev 0.234 0.099 0.281 Median -0.000 -0.022 -0.030 0.033 0.015 -0.293 -0.495 0.079 0.035 1.000 0.375 8.000 0.000 1.000 Minimum -1.308 -0.283 -1.117 -3.190 -4.864 -1.165 -1.502 -1.122 -19.513 0.000 0.000 3.000 0.000 0.000 Maximum 1.248 0.853 3.813 0.849 5.056 7.751 8.350 6.020 2.632 1.000 0.800 19.000 1.000 1.000 133080.00 8.293 1.077 15.649 129000.00 1.000 1.000

Panel B: Earnings performance and returns -0.015 0.325 ROA1 -0.047 0.539 ROA2 BHAR12 -0.163 0.774 -0.272 0.933 BHAR24 SG 0.210 0.643 0.005 1.212 SG1 Panel C: Corporate governance FAMCt-1 0.730 BDINDt-1 0.380 8.250 BDSIZEt-1 0.190 BLKt-1 0.810 AUDt-1 0.450 0.146 2.780 0.390 0.400

Panel D: Control variables MCt-1 3355.84 13297.37 478.50 52.00 BMt-1 1.038 0.998 0.736 -2.245 0.095 0.130 0.046 0.000 EEPt-1 1.202 1.572 0.821 0.025 LEVt-1 3934.16 12995.47 672.44 20.74 SIZEt-1 0.700 0.460 1.000 0.000 ISSUEi 0.790 0.410 1.000 0.000 SEOi _______________________ = Pre-issue discretionary current accruals. DCAi,t-1 = Non-discretionary current accruals. NDCAi,t-1 = Operating cash flow scaled by prior total assets. OCFi,t-1 ROAi,1,2 = Post-issue earnings performance. BHARi,12,24 = Post-issue returns. SG = Sales growth in year 0. = Sales growth in year 1. SG1 = Pre-issue family control. FAMCi,t-1 = Board independence. BDINDi,t-1 = Board size. BDSIZEi,t-1 = Blockholders. BLKi,t-1 = Big 6 auditors. AUDi,t-1 = Market capitalization in $million. MCi,t-1 = Book-to-market. BMi,t-1 = Extreme earnings performance. EEPi,t-1 = Financial leverage. LEVi,t-1 = Firm size measured by assets in $million. SIZEi,t-1 = Offering frequency dummy. ISSUEi = Security type dummy. SEOi

24

TABLE 3 Pearson and Spearman Correlations

ROA1 ROA1 ROA2 DCAt-1 DCA DCA1 DCA2 NDCAt-1 OCFt-1 SG SG1 BHAR12 BHAR24 FAMCt-1 BDINDt1 BDSIZEt1 BLKt-1 AUDt-1 0.514 (0.000) -0.015 (0.808) 0.077 (0.189) 0.132 (0.023) 0.184 (0.004) 0.094 (0.117) 0.198 (0.001) 0.202 (0.000) 0.431 (0.000) 0.318 (0.000) 0.194 (0.002) 0.066 (0.280) 0.074 (0.286) 0.036 (0.563) -0.114 (0.062) -0.087 (0.156) -0.070 (0.313) 0.071 (0.294) 0.040 (0.554) 0.180 (0.007) -0.115 (0.094) 0.156 (0.022) 0.042 (0.534) 0.184 (0.005) 0.230 (0.000) 0.298 (0.000) 0.180 (0.012) 0.036 (0.677) 0.131 (0.069) -0.024 (0.737) -0.085 (0.236) ROA2 0.224 (0.001) DCAt-1 -0.031 (0.610) -0.101 (0.141) -0.019 (0.717) -0.068 (0.259) -0.054 (0.412) -0.198 (0.000) -0.332 (0.000) 0.089 (0.091) 0.125 (0.036) 0.034 (0.553) -0.018 (0.781) -0.030 (0.590) -0.102 (0.090) 0.036 (0.517) 0.030 (0.587) 0.101 (0.067) DCA 0.108 (0.063) -0.014 (0.831) -0.067 (0.200) -0.046 (0.427) -0.073 (0.254) 0.050 (0.337) 0.079 (0.131) 0.154 (0.002) 0.065 (0.266) -0.113 (0.041) -0.095 (0.135) 0.003 (0.961) 0.007 (0.910) 0.111 (0.039) 0.058 (0.282) 0.073 (0.171) DCA1 0.039 (0.506) -0.009 (0.894) -0.038 (0.523) 0.016 (0.786) -0.100 (0.116) 0.035 (0.560) 0.030 (0.621) 0.057 (0.325) 0.122 (0.035) 0.081 (0.166) -0.108 (0.091) -0.174 (0.005) 0.117 (0.093) -0.062 (0.317) -0.068 (0.272) 0.138 (0.025) DCA2 -0.257 (0.000) 0.614 (0.000) -0.135 (0.039) -0.143 (0.025) -0.059 (0.352) -0.073 (0.267) -0.005 (0.945) -0.117 (0.066) 0.085 (0.184) 0.216 (0.001) 0.183 (0.004) 0.012 (0.856) 0.043 (0.587) 0.043 (0.534) 0.025 (0.716) -0.004 (0.957) NDCAt1 0.033 (0.587) -0.011 (0.872) -0.109 (0.037) -0.043 (0.412) -0.019 (0.754) 0.081 (0.215) 0.160 (0.002) 0.177 (0.001) 0.116 (0.053) -0.142 (0.013) -0.177 (0.007) -0.134 (0.015) 0.007 (0.910) -0.018 (0.739) -0.098 (0.075) -0.015 (0.779) OCFt-1 0.176 (0.003) -0.032 (0.638) -0.471 (0.000) 0.012 (0.815) 0.023 (0.696) -0.155 (0.017) 0.069 (0.188) 0.128 (0.014) 0.008 (0.892) 0.048 (0.398) 0.027 (0.673) 0.106 (0.052) -0.018 (0.761) -0.009 (0.872) -0.132 (0.015) -0.079 (0.147) SG 0.073 (0.210) -0.098 (0.140) 0.164 (0.002) 0.279 (0.000) 0.047 (0.416) -0.059 (0.352) 0.072 (0.168) -0.047 (0.364) 0.245 (0.000) 0.027 (0.623) 0.013 (0.835) -0.061 (0.250) 0.054 (0.355) -0.026 (0.628) -0.092 (0.081) 0.100 (0.060) SG1 -0.087 (0.131) 0.127 (0.056) 0.163 (0.006) 0.031 (0.593) -0.051 (0.379) 0.003 (0.959) 0.069 (0.252) 0.109 (0.066) 0.009 (0.872) 0.270 (0.000) 0.084 (0.184) -0.016 (0.792) 0.119 (0.084) -0.037 (0.547) -0.142 (0.020) 0.051 (0.412) BHAR1 2 0.117 (0.042) 0.057 (0.388) -0.026 (0.645) -0.037 (0.504) 0.163 (0.005) 0.027 (0.668) -0.022 (0.702) 0.000 (0.994) -0.019 (0.731) 0.073 (0.205) 0.628 (0.000) 0.058 (0.320) 0.051 (0.425) 0.060 (0.302) 0.077 (0.182) -0.041 (0.482) BHAR2 4 0.039 (0.538) -0.115 (0.089) -0.059 (0.369) -0.064 (0.316) 0.018 (0.784) 0.035 (0.592) -0.040 (0.536) 0.019 (0.771) -0.032 (0.618) 0.019 (0.768) 0.323 (0.000) 0.041 (0.547) -0.041 (0.600) 0.250 (0.000) 0.129 (0.056) -0.085 (0.212) FAMCt1 0.088 (0.153) 0.061 (0.398) -0.038 (0.491) -0.029 (0.584) -0.085 (0.171) -0.023 (0.738) -0.072 (0.192) 0.062 (0.256) -0.066 (0.213) 0.017 (0.779) -0.026 (0.660) -0.030 (0.663) 0.166 (0.004) -0.167 (0.002) 0.004 (0.933) -0.084 (0.113) -0.022 (0.709) 0.047 (0.420) -0.006 (0.913) 0.238 (0.000) -0.015 (0.775) BDINDt -1 0.052 (0.450) -0.010 (0.904) -0.068 (0.260) -0.027 (0.647) 0.036 (0.604) -0.115 (0.145) 0.052 (0.383) -0.045 (0.447) 0.020 (0.724) -0.043 (0.538) -0.038 (0.552) 0.046 (0.556) 0.174 (0.003) BDSIZEt1 0.060 (0.333) 0.016 (0.829) 0.021 (0.710) 0.028 (0.602) -0.102 (0.099) 0.006 (0.926) 0.022 (0.695) 0.000 (0.996) -0.042 (0.428) -0.025 (0.685) 0.004 (0.943) -0.027 (0.691) -0.175 (0.001) 0.045 (0.442) BLKt-1 -0.057 (0.356) -0.019 (0.789) 0.023 (0.671) 0.044 (0.411) -0.071 (0.249) -0.044 (0.516) -0.083 (0.133) -0.115 (0.034) 0.014 (0.798) -0.134 (0.028) 0.069 (0.231) 0.127 (0.060) 0.004 (0.933) 0.063 (0.279) 0.256 (0.000) 0.001 (0.980) AUDt-1 -0.029 (0.64) 0.071 (0.322) 0.054 (0.326) 0.062 (0.246) 0.101 (0.103) 0.032 (0.639) -0.047 (0.393) -0.033 (0.543) 0.110 (0.037) -0.014 (0.817) 0.031 (0.595) 0.028 (0.684) -0.084 (0.113) -0.001 (0.985) -0.004 (0.935) 0.001 (0.980)

____________________ P-values are in parentheses. Pearson correlation coefficients are in the upper triangle. Spearman correlation coefficients are in the lower triangle.

25

ROA1 = Post-issue earnings performance in year 1. ROA2 = Post-issue earnings performance in year 2. DCAt-1 = Discretionary current accruals in year 1. DCA = Discretionary current accruals in year 0. = Discretionary current accruals in year 1. DCA1 = Discretionary current accruals in year 2. DCA2 NDCAt-1 = Non-discretionary current accruals in year 1. = Operating cash flow in year 1 scaled by prior total assets. OCFt-1 SG = Sales growth in year 0. = Sales growth in year 1. SG1 BHAR12 = 12-month abnormal holding period return. BHAR24 = 24-month abnormal holding period return. FAMCt-1 = Pre-issue family control. BDINDt-1 = Board independence. BDSIZEt-1 = Board size. = Blockholders. BLKt-1 = Big 6 auditors. AUDt-1

26

TABLE 4 Cross-Sectional Regression of Post-Issue Earnings Performance

ROAi ,1, 2 = 0 + 1 DCAi ,t 1 + 2 NDCAi ,t 1 + 3OCFt 1 + 4 LNMC i ,t 1 + 5 LNBM i ,t 1 + 6 ISSUE i + 7 SEO i +

Panel A: Overall sample DCAi,t-1 NDCAi,t-1 0.262 ROAi,1 0.172 (1.62) (1.34) ROAi,2 -0.372* (-1.81) -0.089 (-0.25) 0.342** (1.96) 0.021 (0.07)

OCFi,t-1 0.378*** (3.22) -0.309 (-1.43) 0.240** (2.08) 0.328 (1.60) 0.774* (1.91)

LNMCi,t-1 0.031** (2.25) 0.048 (1.62) 0.029** (2.49) 0.048* (1.95) 0.058 (0.83)

LNBMi,t-1 0.060** (2.30) -0.041 (-0.77) 0.044 (1.59) 0.114** (2.10) 0.053 (0.68)

ISSUEi -0.066 (-1.54) -0.100 (-1.19) -0.061 (-1.53) -0.118 (-1.61) -0.104 (-0.71)

SEOi 0.072 (1.51) -0.119 (-1.27)

Adj R 0.075

0.012

Panel B: Placings ROAi,1 0.157 (1.57) ROAi,2 0.142 (0.76)

0.048

0.045

Panel C: Rights offerings -0.683 ROAi,1 -0.08 (-0.23) (-0.79) ROAi,2

0.067

-1.275* 0.784 -1.128 0.022 -0.189 0.035 0.104 (-1.95) (0.52) (-1.51) (0.16) (-1.23) (0.12) _________________ t-statistics are reported in parentheses. *, **, and *** indicate statistical significance at the 0.10, 0.05, and 0.01 levels, respectively. ROAi,1,2 = Post-issue earnings performance. DCAi,t-1 = Pre-issue discretionary current accruals. NDCAi,t-1 = Non-discretionary current accruals. = Operating cash flow scaled by prior total assets. OCFi,t-1 LNMCi,t-1 = Log market capitalization. LNBMi,t-1 = Log book-to-market. = Offering frequency dummy. ISSUEi = Security type dummy. SEOi

27

TABLE 5 Cross-Sectional Regression of Post-Issue Returns

BHAR i ,12 , 24 = 0 + 1 DCAi ,t 1 + 2 NDCAi ,t 1 + 3OCFi ,t 1 + 4 LNMC i ,t 1 + 5 LNBM i ,t 1 + 6 ISSUE i + 7 SEO i + 1

Panel A: Overall sample NDCAi,t-1 DCAi,t-1 BHARi,12 -0.219 -0.070 (-0.89) (-0.15) BHARi,24 -1.736 (-0.71) -0.325 (-0.07) 0.100 (0.19) -0.058 (-0.01)

OCFi,t-1 -0.140 (-0.52) -3.626 (-1.35) 0.060 (0.18) -4.380 (-1.19) -0.554 (-1.35)

LNMCi,t-1 -0.004 (-0.11) -0.004 (-0.01) 0.015 (0.41) 0.031 (0.07) -0.200** (-2.42)

LNBMi,t-1 0.056 (0.90) 0.386 (0.63) 0.084 (1.04) 0.388 (0.43) 0.034 (0.38)

ISSUEi 0.081 (0.79) 0.894 (0.86) 0.056 (0.47) 1.075 (0.82) 0.238 (1.36)

SEOi 0.022 (0.19) 1.474 (1.26)

Adj R 0.009 0.017

Panel B: Placings BHARi,12 -0.015 (-0.05) BHARi,24 -2.097 (-0.66)

0.006 0.016

Panel C: Rights offerings BHARi,12 -0.869** -1.073 (-2.08) (-1.06) BHARi,24

0.114

-0.647 -1.391 -0.755* -0.220*** 0.053 0.266 0.187 (-1.55) (-1.47) (-1.73) (-2.69) (0.62) (1.53) _________________ t-statistics are reported in parentheses. *, **, and *** indicate statistical significance at the 0.10, 0.05, and 0.01 levels, respectively. = One- and two-year holding period post-issue returns. BHARi,12,24 = Pre-issue discretionary current accruals. DCAi,t-1 = Non-discretionary current accruals. NDCAi,t-1 = Operating cash flow scaled by prior total assets. OCFi,t-1 = Log market capitalization. LNMCi,t-1 = Log book-to-market. LNBMi,t-1 = Offering frequency dummy. ISSUEi = Security type dummy. SEOi

28

TABLE 6 Cross-Sectional Regressions of Discretionary Current Accruals

DCAi ,t = 0 = b0 + b1 FAMC i ,t 1 + b2 BDINDi ,t 1 + b3 BSIZETAi ,t 1 + b4 BLK i ,t 1 + b5 AUDi ,t 1 + b6 EEPi ,t 1 + b7 LEVi ,t 1 + b8 LSIZE i ,t 1 + b9 DCAI ,t 1 + b10 ISSUE i + b11 SEOi +

Panel A: Overall sample

FAMCi,t-1 BDINDi,t-1 BSIZETAi,t-1 BLKi,t-1 AUDi,t-1 EEPi,t-1 LEVi,t-1 LSIZEi,t-1 DCAi,t-1 ISSUEi SEOi Adj R

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

-0.089 (-0.25) -0.126 (-0.80) 2432*** (3.41) -0.025 (0.68) 0.057 (1.12) -0.046 (-1.00) -0.018 (-0.14) -0.059 (-0.38) 2098* (1.70) 2385* (1.80) 0.026 (0.54) -0.033 (-0.58) 0.034 (0.68) 0.031 (0.51) -0.040 (-0.28) -0.132 (-0.70) -0.002 (-0.24) 0.006 (0.56) 0.010 (0.54) 0.008 (0.42) -0.065 (-0.88) -0.054 (-0.62) 0.001 (0.03) -0.012 (-0.26) 0.007 (0.16)

0.009 0.002 0.029 0.001 0.001 0.023

Panel B: Placings Model 6a -0.014 (-0.26)

0.028

Panel C: Rights offerings Model 6b -0.153 0.335 417.12 0.187** -0.008 0.106 -0.012 0.005 (-0.99) (1.40) (0.09) (2.39) (-0.10) (0.53) (-0.90) (0.07) ___________________ t-statistics are reported in parentheses. *, **, and *** indicate statistical significance at the 0.10, 0.05, and 0.01 levels, respectively. = Discretionary current accruals. DCAi,t=0 = Pre-issue family control. FAMCi,t-1 = Board independence. BDINDi,t-1 BSIZETAi,t-1 = Board size scaled by total assets. = Blockholders. BLKi,t-1 = Big 6 auditors. AUDi,t-1 = Extreme earnings performance. EEPi,t-1 = Financial leverage. LEVi,t-1

-0.103 (-0.77)

0.156* (1.65)

0.152

30

LSIZEi,t-1 DCAi,t-1 ISSUEi SEOi

= Firm size measured by assets. = Pre-issue discretionary current accruals. = Offering frequency dummy. = Security type dummy.

31