QUALITY OF GOVERNANCE AND FIRM PERFORMANCE: EVIDENCE FROM SPAIN

Eloisa Prez de Toledo

Universitat Autnoma de Barcelona Dept. Economia de lEmpresa
Eloisa.Perez@uab.es

Supervisor : Prof. Carles Gispert Pellicer

Universitat Autnoma de Barcelona Dept. Economia de lEmpresa
Carles.Gispert@uab.es

ABSTRACT
Corporate governance is a set of mechanisms relevant to economic efficiency since it can minimize agency problems. The question is to determine how governance and firm performance interact. Recent research shows that firm-level corporate governance mechanisms are more important in countries with low investor protection, suggesting that firms can partially compensate for ineffective legal environments. Within this context, the main objective of this paper is to construct a robust proxy for quality of governance for the Spanish public companies. A second objective is to verify which are the determinants of governance in the case of Spain, and to assess whether they influence the performance of the companies. Thus, after providing an extensive literature review on the field, I construct a governance index (GOV-I) for a sample of 97 Spanish non-financial public companies and, through simple and multiple OLS regressions, I assess the interaction between governance and performance. The results show a significant relationship between governance and performance, future growth opportunities and size, demonstrating that Spanish firms adopt better standards of governance to compensate for the low level of investor protection holding in the country. The results support the prevalent hypothesis of a positive relationship between corporate governance and performance. Keywords: corporate governance, governance index (GOV-I), firm performance, investor protection, Spain. JEL classification: G32, G34

TABLE OF CONTENTS

1. Introduction 2. Problem Statement And Objectives 2.1 2.2 Determinants of the Quality of Corporate Governance Quality of Governance and Performance

3 5 6 8 9 11 15 16 17 19 19 20 21 25 25 28 31 36 38

3. Governance And Performance: Theory And Practice 3.1 3.2 3.3 What Does the Literature Say? The Construction of Indexes as a Proxy for Quality of Governance Designing the Research: Questions and Hypotheses

4. Methodology 4.1 4.2 4.3 4.4 Sample Selection and Data Collection The Corporate Governance Index (GOV-I) definition and specifications Determinants of the Quality of Corporate Governance Relationship between Governance and Performance

5. Empirical Results 5.1 Descriptive Statistics 5.2 The Governance Index (I-GOV) description 5.3 Empirical Results 6. Discussion and Conclusions References

1. INTRODUCTION
In a capitalist economy, financing is fundamental to the viability of companies and to the persistence of the capitalism itself. The availability of funds depends on the efficient allocation of resources by the economic agents from financial markets to productive investments, e.g. for the creation of new ventures or to finance the growth process of established companies. An efficient allocation depends on the investors expected return, but also, on the investors belief that the firm will be managed in order to maximize the investment and that the cash flows promised in exchange for the investment will effectively be returned. The economic viability of investment projects can be assessed through capital budgeting techniques and risk-return trade-off analysis for asset allocation decisions. Nevertheless, investors trust depends on a broad set of factors as the legal, institutional and regulatory environment that guarantees the investor protection. In this sense, corporate governance surges to mitigate the agency problems derived from the relationship between principals and agents. Shleifer and Vishny (1997) define corporate governance as a set of mechanisms relevant to economic efficiency due to its influence over the decision of investors to provide finance, debt or equity, to the firm. The purpose of a governance structure is to assure a significant flow of capital to the financing of firms. The separation between ownership and control, as described by Berle and Means (1932), aggravated by the problem of information asymmetry between managers and providers of capital, can lead to the expropriation of the capital suppliers wealth. An efficient governance structure should be able to guarantee that the agent will undertake the optimal level of investment and minimize the amount of rent seeking behavior. In the presence of agency problems, it is necessary a mechanism that is able to govern the way in which decisions will be taken in the future in face of an event that was not contemplated in the contract established between agent and principal, as described by Hart (1995, p.679) () governance structure matters when some actions have to be decided in the future that have not been specified in an initial contract: governance structure provides a way for deciding these actions. A variety of governance mechanisms can be used in order to design efficient governance structures, for instance, the organization of a board of directors, the ownership structure and control, stock options and other incentives programs to management and employees, the capital structure, the market competition, the product competition, the presence of an active market for corporate control, among others. Another reason why corporate governance is relevant to economic growth is related with its possible impact on the performance of the companies. The basic idea is that in a population of companies, some can be distinguished as companies with good governance. These firms would become more attractive to investors, ceteris paribus, increasing their access to capital. As a result of such increment in the availability of credit, the cost of capital of these companies would be reduced, both the cost of debt and the cost of equity, which implies that companies with good governance should experience a reduction in their weighted average cost of capital (WACC). As a consequence of such reduction in the cost of capital, there would be an increment in the

market value of these companies1. Besides, the reduction in the required rate of return allows the firm to accept a greater number of investment projects which could increase its competitiveness. In a broad sense, research on corporate governance is justified by its contribution to the increase in the access to capital and, consequently, to the reduction of the cost of capital in a given economic system. As stated by Shleifer and Vishny (1997), the suppliers of finance use corporate governance structures to ensure that they will get a return on their investment. Moreover, according to Rajan and Zingales (2004, p.51) there are three obstacles in the way of broadening access to finance: (1) the degree to which risk is concentrated (in a developed system the risk is widely distributed and allocated to the players that can best hold it, which reduces the risk premium demanded by investors); (2) the limited information financiers or investors have about borrowers and their prospects; and, (3) the possibility that borrowers may not act in the best interest of the financiers. Research on corporate governance can reduce the third problem by analyzing and defining the mechanisms that assure that firms (managers) will use the funds in the best interest of the investors. There are some firm characteristics that are associated with the governance of the public companies, the so-called internal and external mechanisms of governance. Ownership concentration and board structure are pointed to be the primary internal mechanisms, while an active market for corporate control is the main external mechanism. These mechanisms are alleged to work together in a system to affect the governance of the firms (Cremers and Nair, 2005). In this paper, I try to provide some empirical evidence of how these mechanisms interact within the Spanish reality. For reaching this objective I ask two questions: Which observable factors make companies adopt different levels of governance under the same contracting environment? Does quality of governance affect firm performance? In order to proxy quality of governance, and following a new trend in governance studies, I construct a governance index for the Spanish public companies, namely GOV-I. This paper has two major contributions. First, I construct a governance index (GOV-I) for Spain. And, second, I assess the relationship between governance and performance through the use of OLS regressions. The results show a positive relationship between corporate governance and firm performance, for instance, firms with higher standards of governance receive higher market valuations, measured by Tobins q. Generally, the paper shows that ownership concentration (measured by the presence of a controlling shareholder and the presence of large blockholdings) and leverage are the significant governance mechanisms in the case of Spain. On the other hand, other traditional mechanisms such as board independence, board size or duality CEOChairman did not show a significant impact on the valuation of the firms. The paper is organized as follows. Part 2 states the problem, defines the objectives and puts forward the hypotheses. Part 3 presents the theoretical framework and offers a revision of the extant literature. In Part 4 there is a description of the research

The logic of such increment in the value of the firm is based on the fundamentals of capital budgeting. The value of a company is calculated discounting its expected free cash flows by the weighted average cost of capital (WACC).

methodology, data and design. Part 5 presents the empirical results. In Part 6, I discuss the results and conclude the paper.

2. PROBLEM STATEMENT AND OBJECTIVES

The theoretical discussion about corporate governance is based on two hypotheses, firstly that governance mechanisms influence the performance of the firms, and, second that firm performance also influences the governance system adopted by the firms. Essentially, the basic issue is to detect whether the performance is determined by internal or external mechanisms of governance. Gillan (2006, p.385) divide the internal mechanisms into 5 categories: (1) the board of directors; (2) managerial incentives; (3) capital structure, (4) bylaw and charter provisions (antitakeover measures); and (5) internal control systems. Similarly, the external mechanisms of governance are also divided into 5 categories: (1) law and regulation; (2) the markets (capital markets, market for corporate control, labor markets, and product markets); (3) the providers of capital market information (credit, equity, and governance analysts); (4) accounting, financial and legal services from parties external to the firm (auditing, insurance, and investment banks); and, (5) private sources of external oversight (media and external lawsuits). Hitherto, there is still no conclusive empirical evidence in the literature about whether and how governance mechanisms influence the performance of the firms; and, about how governance mechanisms interact (in a complementary or substitute way) (Bhren and degaard, 2003). According to Chi (2005), there are three possible causal relationships between quality of governance and firm performance (or market value proxy by Tobins q), as illustrated by Figure 1. The first possibility is that there is a direct causal relationship with governance enhancing firm performance. In the second possibility, causality runs in both ways and, finally, the third possibility is that governance and performance are not directly related, but they are spuriously connected through other variables (Chi, 2005 p.67). Most studies analyze exclusively the possible influence of specific governance mechanisms on specific corporate performance variables. In these studies, governance mechanisms are treated as independent variables and performance measures as dependent variables. In this sense, governance mechanisms are considered and treated as exogenous variables with no relation with other governance mechanisms or other firms characteristics. Himmelberg et al. (1999), however, argue that the ownership structure of the firm may be endogenously determined by the firms contracting environment, which differs across firms in observable and unobservable ways. For instance, if the scope for perquisite consumption is low in a firm, then a low level of management ownership may be the optimal incentive contract. The endogeneity of management ownership has also been noted by Jensen and Warner (1988, p.13): A caveat to the alignment/entrenchment interpretation of the crosssectional evidence, however, is that it treats ownership as exogenous, and does not address the issue of what determines ownership concentration for a given firm or why concentration would not be chosen to maximize firm value. Managers and shareholders have incentives to avoid inside ownership stakes in the range where their interests are not aligned, although managerial wealth constraints and benefits from entrenchment could make such holdings efficient for managers.

FIGURE 1 THREE POSSIBLE CAUSAL RELATIONSHIPS BETWEEN GOVERNANCE AND PERFORMANCE

Possibility I
Firm performance (P) causal Quality of governance (G)

Spurious correlation observed Firm performance (P) Firm performance (P) causal Quality of governance (G)

Quality of governance (G)

causal

Other factors (observable and unobservable)

causal

Possibility II Possibility III

Source: Adapted from Chi (2005, p.68).

Finally, the set for exploring these matters is the Spanish governance system which is characterized, according to Demirg-Kunt and Maksimovic (1996), as having an underdeveloped capital market both in terms of market capitalization and in volume of traded shares, the banking sector is of greater importance in financing firms, and, according to La Porta et al. (1998), the degree of investor protection is low since it is based on the Civil law system. Within this context, the main objective of this paper is to assess whether the mechanisms of governance are exogenous and influence the performance of the Spanish public companies. To reach this objective the study is divided into two parts: 1. Determinants of the quality of corporate governance to assess the possible factors that make companies adopt different levels of governance under the same level of investor protection (legal, institutional and regulatory environment). 2. Relationship between corporate governance and corporate performance to assess the influence of the quality of governance on the performance of the Spanish public companies through the use of OLS simple and multiple regressions. 2.1. Determinants of the Quality of Corporate Governance

First of all, it is preemptive to define quality of corporate governance. Durnev and Kim (2005, p.1463) define the quality of governance as (1 d), where d is the proportion of firm value diverted for private gains. Thus, a high level of d implies poor governance practices, where d is broadly defined to include a wide range of value-decreasing activities from what Jensen and Meckling (1976) define as excessive evasion and corporate benefits to direct stealing of tangible and intangible corporate resources. This

definition of the quality of governance captures various governance and managerial practices in place that may or may not be legally compulsory. Recent research has been focused in analyzing the quality of corporate governance among firms operating in different country-level investor protection. It is possible, however, that, due to some observable characteristics, not all firms operating in the same country (with the same legal environment) offer the same degree of protection to their investors. As hypothesized by La Porta et al. (1998), the legal system is fundamental to corporate governance. In particular, they argue that the extent to which a countrys laws protect investor rights and the extent to which those laws are enforced are the most basic determinants of the ways in which corporate finance and corporate governance evolve in that country. Within this framework, Klapper and Love (2004) provide a cross-country study of firmlevel corporate governance practices and they conclude that companies operating in the same level of investor protection show different levels in the quality of corporate governance. They found firms with a high level of corporate governance provisions in countries with weak legal environments and vice-versa. This approach, developed by Himmelberg et al. (1999), Himmelberg et al. (2002) and Klapper and Love (2004), states that investor protection has an external component related to the legal environment and an internal component related to the activity developed by the firm and other characteristics (endogenous protection). According to Himmelberg et al. (2002, p.2) () investor protection refers collectively to those features of the legal, institutional and regulatory environment and characteristics of firms or projects that facilitate financial contracting between insider owners (managers) and outside investors. Thus, it is probable that firms operating in the same country offer different degrees of investor protection, due to specific operational characteristics and to particular interests. It is corroborated by the research of La Porta et al. (2000). They find that firms in common law countries where investor protection is stronger make higher dividend payouts when the firms investment opportunities are poor than do firms in countries with weak legal protection. According to Klapper and Love (2004) corporate governance is likely to be endogenously determined and they point out three sources of endogeneity that in theory could be associated with firms adopting better governance mechanisms: (1) the composition of a firms assets; (2) unobservable growth opportunities; and, (3) firm size. The composition of a firms assets will affect its contracting environment because it is easier to control and harder to steal fixed assets (equipments, etc.) than soft capital (intangibles, R&D, etc.). In that sense, a firm with a high level of intangibles may find optimal to adopt a higher level of corporate governance (and avoid possible misuse of these assets). The variable unobservable growth opportunities is related with the fact that firms with good growth opportunities will need capital to finance the expansion process and they can find optimal to improve their level of governance in order to reduce the cost of capital. And finally, firm size has ambiguous effects because large firms may have greater agency problems due to destination of their free cash flows and small firms may have better growth opportunities and greater need for external finance, thus, both have incentives to adopt better governance mechanisms. Besides these three variables, other variables will be introduced, for instance, ownership structure, corporate performance, and issuance of stocks in an American or European

(non-Spanish) stock market2. As a proxy for quality of governance, it will be built a governance index (GOV-I). For the construction of the index, the approach to be used is the one developed by Gompers, Ishii and Metrick (2003) and strengthened by Brown and Caylor (2004), but departing from the determinants of governance detected by Klapper and Love (2004). The use of indexes in the field of corporate governance is relatively new, and the authors that have built and/or used governance indexes for analyzing the reality of different countries are Black (2001) for Russia, Gompers et al. (2003) for the US, Klapper and Love (2004) for emerging markets, Brown and Caylor (2004) also for the US (they strengthen the index developed by Gompers Ishii and Metrick 2003), Leal and Carvalhal-da-Silva (2004) and Silveira (2004) for Brazil, Black et al. (2005) for Korea, Durnev and Kim (2005) for emerging markets, Cremers and Nair (2005) for the US, and, Beiner et al. (2006) for Switzerland. 2.2. Quality of Governance and Firm Performance

Ownership structure is one of the most important mechanisms of governance. The vast majority of the initial studies in the field tried to capture the influence of this mechanism in the performance of the firms, for instance Demsetz and Lehn (1985), Mrck et al. (1988), McConnell and Servaes (1990) and Hermalin and Weisbach (1991). In sum, the results of these studies pointed to a positive and significant relationship between ownership concentration and corporate performance. Recently, a research line initially developed by La Porta et al. (1998) try to assess how ownership structure vary across countries, hypothesizing that the decisive factor to explain differences among countries is the degree of investor protection. From this perspective, firms ownership structure is an equilibrium response to the legal environment in which they operate. The research of La Porta et al. (2000), La Porta et al. (1999), Claessens et al. (2002) and Beck et al. (2001) suggest that country-level differences in law systems and law applicability cause differences in ownership structure, dividend policies, availability of external finance and in the valuation of corporate bonds. In this sense, most of the literature on ownership structure try to assess whether differences in the level of investor protection within distinct contractual environments (different countries) promotes a greater ownership concentration. Demsetz (1983, p.377) was the first to propose an approach in which ownership structure is an endogenous result of a corporate efficiency maximization process. To test this hypothesis, Demsetz and Lehn (1985) developed an empirical model in which specific firm or industry characteristics, such as size, riskiness and regulation could be assessed as the determinants of ownership concentration. The authors argue that the causal relationship between ownership structure and performance is spurious, since ownership concentration can be considered an endogenous variable. Himmelberg et al. (1999), in the same line of research as Demsetz and Lehn (1985), broaden their results through the introduction of other independent variables for explaining ownership concentration and the use of panel data. The authors propose that investor protection, besides having an external component related to the legal
The idea is that companies that issue stocks in an American or European stock exchange (besides listing in the Madrid Stock Exchange) are likely to present higher standards of governance.
2

environment (legal protection), also has an internal component related to the nature of the firms operation and to other firm characteristics (endogenous protection). In this sense, ownership concentration is a function of legal protection, but also of firms intrinsic protection, that is different for firms operating in distinct contracting environments. Among the endogenous aspects mentioned by Himmelberg et al. (1999) as the determinants of ownership concentration, three can be pointed as the most important: size, managers discretion, and idiosyncratic risk. According to the authors, size has a priori a dubious effect on ownership concentration. On the one hand, agency costs and monitoring could be more severe in big companies increasing the need for ownership concentration. On the other hand, big companies can make use of economies of scale in monitoring management, for instance, using rating agencies. In relation with management discretion, they argue that as fixed assets are observable and easier to monitor, firms with high level of tangible assets should present a lower optimal level of ownership concentration. In companies with a high level of intangibles, instead, high ownership concentration can improve monitoring. Finally, with relation to the idiosyncratic risk, the authors point out that, ceteris paribus, high ownership concentration implies to the investor a less diversified portfolio. Thus, as higher a firm systematic risk the lower its optimal level of ownership concentration. De Miguel et al. (2003) analyze ownership concentration in Spain, in order to test the monitoring and expropriation hypotheses, and to analyze insider ownership seeking evidence of the convergence-of-interest and entrenchment hypotheses. Their results confirm not only the monitoring but also the expropriation effect for the very high levels of ownership concentration of Spanish firms. The authors conclude that due to the fact that Spanish majority shareholders manage to expropriate the wealth of minority shareholders, unlike UK, US, Germany or Japan, differences in corporate governance systems lead to different value-ownership relations.

3. GOVERNANCE AND PERFORMANCE: THEORY AND PRACTICE

The conceptual framework for the present study is given by a combination of approaches to the theory of the firm: the neoclassical theory of the firm, the principalagent theory, the transaction cost economics, the property rights approach and the institutional theory. The theory of the firm is the classical theoretical framework for the studies in the field of corporate governance. The neoclassical theory is the one that provides the first notions of the firm. Nevertheless, despite its formalism and rigor in the construction of economic models, the neoclassical theory portrays the firm in a rudimentary way. In the words of Hart (1996 p.200), (n)eoclassical theory describes in rudimentary terms how firms function, but contributes little to any meaningful picture of their structure. Trying to fulfill the gap, important theories were developed in the twentieth-century by academics like Knight (1921), Coase (1937), Alchian and Demsetz (1972), Williamson (1975), Jensen and Meckling (1976), among others, that aimed to incorporate characteristics of the real world in a new theory of the firm. The agency theory surges to explain the agency problem and the costs associated with it. The discussion about the need for improving the governance of the firms is a response to many cases of expropriation of shareholders wealth by the top executives,

10

but also by the majority shareholders at the expense of the minority shareholders. This phenomenon describes quite well the agency problem, when the agents take decisions in order to maximize their own best interests rather than those of the shareholders (the same apply to the appropriation by the majority shareholders of the private benefits of control). The agency problem is an essential element within the contractual view of the firm, developed by Coase (1937), Alchian and Demsetz (1972) and Fama and Jensen (1983). The theme of corporate governance is inserted within this context and the development of governance mechanisms aims to mitigate this problem. According to Jensen (2001) the contractual view is based on the idea that the firm is a nexus of contracting relationships among clients, workers, executives and suppliers of goods and capital. In line with this view, executives and shareholders sign a contract specifying how firm resources must be managed. In theory, a complete contract should be designed, specifying all actions the agent must undertake in face of any possible situation or contingency. The problem is that, since many contingencies are hard to predict, complete contracts are unviable. Due to this problem, investors have to allocate their residual control rights or their right to decide in circumstances not predicted in the original contract. The institutional theory provides the conceptual basis for the discussion about the degree of investor protection holding in each country. The property rights are the explicative variables of the level of economic development in a given institutional environment, and of the governance model adopted by the companies, since it has an impact on their ownership structure. Moreover, the new institutional economics visualize the firm as a nexus of contracts and, for North (1990) the institutions are the rules of the game in a society. The institutions determine not only the economic performance, but also the governance structure and the governance model adopted by the firms. It is necessary to understand the evolution of the institutions in a given environment (e.g. country) to understand its patterns of corporate governance. Nevertheless, all these approaches to the theory of firm that try to develop a more realistic picture of the famous black box have weaknesses as pointed by Hart (1996): the (p)rincipal-agent theory enriches neoclassical theory significantly, but still fails to answer the vital questions of what defines a firm and where the boundaries of its structure are located. The transaction cost economics provides the basis for introducing the idea of planning and contracting costs that was neglected by the neoclassical theory, but despite the contribution of Williamson (1975, 1985), the precise nature of these costs are unclear (Hart, 1996 p.204). The same reasoning can be applied to the property rights approach to the firm, even though being more complete than the previous approaches (property rights contains common features of all described approaches)3, it does not consider the separation between ownership and control that is actually present in the large public corporations. As a consequence, Hart (1996) concludes that a formal model of the firm that incorporates all these features, including an explanation of firms financial structure is an important (but not impossible) task for future research.
It is based on maximizing behavior (like neoclassical approach); it emphasizes incentives issues (like the principal-agent approach); it emphasizes contracting costs (like transaction costs approach); it treats the firm as a standard form contract (like the nexus of contracts approach); and it relies on the idea that a firms owner has the right to alter membership of the firm () (Hart, 1996 p.210).
3

11

The theoretical framework could be improved by the inclusion of an approach developed by Jensen (2001), the corporate objective function. Actually there are two corporate objective functions discussed by the literature: the value (or shareholders) maximization proposition and the stakeholder theory. The stakeholder theory has been gaining great acceptance among professionals and governments but also among academics whose argument is that managerial decisions impact investors but also a number of stakeholders who have an intrinsic relationship with the firm: employees, clients, suppliers, the community were the firm is located, etc, the so-called externalities by the economists. In the words of Tirole (2001, p.4), Why should institution design ignore the natural stakeholders, and favor the investors, who are stakeholders by design, by giving them full control rights and by aligning managerial compensation with their interests?, and goes further, defining corporate governance as the design of institutions that induce or force management to internalize the welfare of stakeholders. Nevertheless, neither the theory of the firm with the value (shareholder) maximization paradigm nor the stakeholder theory with its multiple objectives offers a clear picture of the exact direction of the causality between governance and firm value. Governance theories suggest that strong shareholder rights can mitigate agency problems and, as a consequence, increase firm value. However, shareholders rights can be restricted by the managers. Therefore, no causal inferences can be drawn from the theory since it is not clear that there is a causal relationship and its direction. Due to this lacuna in the theoretical framework, many researchers have been showing empirically that governance drives performance. However, they point out the limitations of their results warning that they may not be robust to some unobservable firms characteristics (Chi, 2005). In the sequence, a literature review on the field of corporate governance is provided, giving special attention to the relationship between governance and performance. 3.1. What Does the Literature Say?

3.1.1. Ownership Structure and Performance

The relationship between ownership structure and performance was firstly approached by Berle and Means (1932). They suggest that, due to the separation between ownership and control in the American big corporations, there is an inverse relation between disperse ownership and performance. Four decades after, Jensen and Meckling (1976) and Stulz (1988) developed theoretical models trying to formalize the relationship between ownership and performance, arguing that ownership influences performance. The model of Stulz (1988) predicts a concave relationship between managerial ownership and firm value. In the model, as managerial ownership and control increase, the negative effect on firm value associated with the entrenchment of manager-owners starts to exceed the incentive benefits of managerial ownership. The first empirical studies in the field aimed to test this hypothesis assessing the impact of ownership structure on performance through the use of linear regressions with ownership structures as the independent variables. Among the first empirical studies, the most important are Mrck et al. (1988), McConnell and Servaes (1990) and Hermalin and Weisbach (1991). In all these studies, the authors find a significant relationship between ownership structure and firm value that can be interpreted as consistent with the theoretical hypothesis formulated by Jensen and Meckling (1976)

12

and Stulz (1988). Mrck et al. (1988) find an inverse U-shaped relationship between managerial equity ownership and firm valuation for a sample of US firms. One interpretation is that firms' performance improves with higher managerial ownership, but after a point, managers become entrenched and pursue private benefits at the expense of outside investors. McConnell and Servaes (1990) also provide empirical support for this relationship for US firms. Demsetz and Lehn (1985) formulate an alternative hypothesis assuming that ownership structure is endogenously determined under the assumption of equilibrium. To test this hypothesis, more recent studies make use of sophisticated econometric techniques and consider the ownership structure variables as endogenous rather than exogenous. These studies have not been corroborating the hypothesis that ownership structure is an exogenous variable and that it influences performance (Cho, 1998; Himmelberg et al. 1999; Demsetz and Villalonga, 2001). Cho (1998) examines the relationship between ownership structure, investment and corporate value in the United States. According to the author, common sense says that ownership structure must influence corporate investment decisions, and that the last must influence corporate value. In the first part of the study, the author uses the method of ordinary least squares (OLS) to test the hypothesis. The initial results suggest that ownership concentration (considered as the independent variable) has a significant impact on corporate investment (proxy by capital investments and investments in R&D). Thus, these results corroborate the assumption that ownership structure influences firm value. In the sequence, the author changes the method to simultaneous equations systems and considers ownership structure as an endogenous variable. Cho (1998) concludes that causation is reversed, running from performance to ownership structure rather than in the opposite way, with investments influencing corporate value and corporate value, in its turn, influencing the ownership structure. The author presents empirical evidence showing that probably ownership structure is not an exogenous variable, and questions previous research that tried to demonstrate the causal relationship between ownership structure and performance. Himmelberg et al. (1999) analyze the determinants of insider ownership and the relationship between ownership structure and performance in the US. The study follows the methodology proposed by Demsetz and Lehn (1985) and tries to find evidence that insider ownership is endogenously determined by other corporate variables like size and industry. After introducing other possible variables as capital intensity, R&D expenses, free cash flow (FCF) and investment rate, the authors, through the use of panel data analysis, try to isolate unobservable firm characteristics that did not vary across the time period under analysis. They conclude that insider ownership and performance are endogenously determined by some characteristics of the legal environment and that they are only partially observable. Demsetz and Villalonga (2001) analyze the relationship between ownership and performance primarily using the traditional approach of isolated regressions (OLS) considering ownership structure as the independent variable. Their results indicated that ownership structure has a significant influence on performance so as the results obtained by Mrck et al. (1998) and McConnel ans Servaes (1990). Then, the authors run some tests using a two simultaneous equation system through the procedure of 2SLS (two stages least square). The results produced by this approach showed that ownership structure has no statistically significant influence on performance.

13

De Miguel et al. (2003) investigate the relationship between ownership structure and value in the Spanish firms. They provide new evidence on this relation, since the Spanish corporate governance system differs from the ones considered in previous theoretical and empirical research (e.g. US, UK, Germany). The authors use panel data methodology and control for potential endogeneity using instruments. Their results confirm the monitoring and the expropriation effect for the very highest concentration values in Spanish firms. The fact that Spanish majority shareholders manage to expropriate the wealth of minority shareholders confirms the idea that differences in corporate governance systems are related to the legal environment as described by Klapper and Love (2004) and Himmelberg et al. (2002). More recently, De Miguel et al. (2005) examine how different control mechanisms relate to one another in the Spanish corporate governance system. The authors propose a new empirical approach that consists in analysing control mechanisms according to the non-linearity of the value-ownership relation. They conclude that the Spanish corporate governance system is very different from the US one. Moreover, their results show that control mechanisms (especially insider ownership, debt and dividends) are used in a complementary way by Spanish firms.

3.1.2. Board Composition and Firm Performance

The Board of Directors has a fundamental role in a corporate governance system, considered a major internal mechanism along with ownership concentration, is used to reduce agency costs between shareholders and executives, and between controlling and minority shareholders. The work of Hermalin and Weisbach (1991) offers a summary of the United States evidence on the board mechanism: (a) higher proportions of outside directors are not associated with superior firm performance, but are associated with better decisions concerning issues such as acquisitions, executive compensation, and CEO turnover; (b) board size is negatively related to both general firm performance and the quality of decision-making; and, (c) poor firm performance, CEO turnover, and changes in ownership structure are often associated with changes in the membership of the board. Evidence from other countries is offered by Wymeersch (1998) for Europe, Rodriguez and Anson (2001) for Spain, Mak and Yuanto (2002) for Malaysia and Singapore, Eisenberg, Sundgren, and Wells (1998) for Finland, and Dahya, McConnell, and Travlos (2002) for the UK. Rodriguez and Anson (2001) examine the market reaction to announcements of compliance to the Olivencia Code by Spanish firms. They observe that the stock prices react positively to announcements of compliance when such announcements imply a major restructuring of the board. Besides, this reaction is stronger for firms that have been performing poorly. Wymeersch (1998) provides a wide narrative of the composition of European boards of directors. The author reports that, in most European countries, the role of the board of directors is not defined by law. So, the shareholder wealth maximization has not been the primary goal for European boards. This varies across countries, with the British, Swiss, and Belgian systems being closer to the American model. European Boards are most often unitary, as in the United States. However, in some European countries the two-tiered system is the rule. A two-tiered structure is compulsory in some countries,

14

e.g. Germany and Austria, and optional in others, e.g. France and Finland. Two-tier boards generally consist of a managing board, composed by executives of the firm, and a supervisory board. In the case of Germany, employees are represented in the supervisory board, a system called co-determination, and it is mandatory for firm with more than 500 employees. Dahya, McConnell, and Travlos (2002) address the effect of the Cadbury Committee (UK Code of Best Practice) on board effectiveness. Among other things, the Code recommends that boards of UK corporations include at least three outside directors and that the positions of chairperson and CEO is held by different individuals. The compliance to the Code is voluntary. Nevertheless, the London Stock Exchange requires that all listed companies explicitly indicate whether they are in compliance with the Code, and if not, an explanation is required. As consistent with US evidence, there is some evidence of a negative relation between board size and firm performance in several non-US countries. Mak and Yuanto (2002) find evidence of an inverse relationship between board size and Tobins Q in Singapore and Malaysia, while Eisenberg, Sundgren, and Wells (1998) document an inverse relation between board size and profitability for SMEs in Finland. Carline, Linn, and Yadav (2002) find that board size is negatively related to operating performance improvements after UK mergers.

3.1.3. Large Shareholdings and Firm Performance

The presence of large shareholders can have positive or negative effects on firm performance. Many researches have been conducted in order to assess the positive effects of large shareholdings in the maximization of firms value. However, less attention has been given to the costs associated with the presence of large investors, as pointed out by Claessens, Djankov, Fan and Lang (2002). The authors conducted an analysis with 1,301 publicly traded corporations from eight East Asian economies4. They find that relative firm value (measured by the market-to-book ratio of assets) increases with the share of cash-flow rights in the hands of the largest shareholder. This result is consistent with previous studies on the positive incentive effects associated with increased cash-flow rights in the hands of one or a few shareholders. But, on the other hand, they find that the entrenchment effect5 of control rights has a negative effect on firm value. The findings of Claessens et al. (2002) complement the findings of Mrck, Stangeland, and Yeung (2000) for Canada. The authors show that large shareholdings control impedes the growth of Canadian public companies, because entrenched controlling shareholders take many advantages in maintaining the current value of the firm. Dick and Zingales (2004), estimate the private benefits of control for 39 countries in an international comparison. They conclude that higher private benefits are associated with less developed capital markets, more concentrated ownership and more private negotiated privatizations. The authors point out the lack of empirical evidence on this topic despite the importance of this concept. It is related, according to the authors, to the
Hong Kong, Indonesia, South Korea, Malaysia, Philippines, Singapore, Taiwan, and Thailand. The entrenchment effect occurs when large controlling shareholders pursue the private benefits of control at the expense of other groups (minority shareholders and other stakeholders). This effect was originally formulated by Stulz (1988).
5 4

15

nature of the phenomenon and its difficult observation. In general, the international evidence indicates that the accumulation of control rights in excess of cash flow rights reduces the observed market value of firms. As summarized by Denis and McConnell (2003, p.26), a number of conclusions can be drawn from the international literature on the ownership of publicly-traded firms. First, ownership is, on average, significantly more concentrated in non-US countries than it is in the US. Second, ownership structure appears to matter more in non-US countries than it does in the US i.e. it has a greater impact on firm performance. Overall, private ownership concentration appears to have a positive effect on firm value. Third, there are significant private benefits of control around the world, and they are more significant for most non-US countries than they are for the US. Structures that allow for control rights in excess of cash flow rights are common, and generally value-reducing. 3.2. The Construction of Indexes as a Proxy for Quality of Governance The main objective of an empirical study in the field of corporate governance is to assess whether governance drives performance. To reach this objective, recent studies have constructed corporate governance indexes that put together in only one measure all relevant information about a series of governance mechanisms. Black (2001) analyzes the hypothesis that good governance practices affect firms market value in Russia. As a proxy to quality of corporate governance, the author uses a corporate governance ranking created by the Brunswick Warburg Investment Bank. Gompers Ishii and Metrick (2003) use the incidence of 24 different provisions to build a governance index for about 1,500 firms per year, and then they study the relationship between the index and several performance measures during the 1990s. The authors find a strong relationship between corporate governance and stock returns. They also find that weaker shareholder rights are associated with lower profits, lower sales growth, higher capital expenditures, and a higher amount of corporate acquisitions. Klapper and Love (2004) evaluate the differences in the governance practices of fourteen companies in emerging markets through the use of a corporate governance index developed by the Credit Lyonnais Securities Asia (CLSA), an investment bank. The authors verified a huge variation in the quality of corporate governance among companies, and the average quality of corporate governance was superior in countries with more efficient legal systems. Bhren and degaard (2003) analyze the relationship between corporate governance and performance in Norway. The authors find that corporate governance matters for economic performance, insider ownership is the most important, outside ownership destroys market value, and direct ownership is superior to indirect. Their results persist across a wide range of single-equation models, suggesting that governance mechanisms are independent and may be analyzed one by one. The authors conclude that the lack of significant relationships in the tests allowing for endogeneity may not reflect optimal governance, but rather an underdeveloped theory of how governance and performance interact. Leal and Carvalhal-da-Silva (2004) analyze the relationship between quality of corporate governance and the performance of the Brazilian public companies. The

16

authors also construct an overall governance index composed by fifteen questions divided into four categories: disclosure, board composition, ownership structure and shareholder rights. They found evidence that companies with best governance practices have a higher market value (proxy by Tobins Q). Brown and Caylor (2004) develop an index called Corporate Governance Quotient (CGQ). They found that firms with weaker corporate governance are less profitable, since they have lower return on assets, lower return on average equity, lower return on equity, and lower return on investment than do firms with stronger governance (measured by the CGQ). Besides, the authors also find evidence that firms with weaker corporate governance are riskier, have lower dividend payouts and lower dividend yields than firms with stronger corporate governance. They examine four factors: board composition, managerial compensation, takeover defenses, and audit. Board composition is the most important factor while takeover defenses is the least important for the firms quality of corporate governance. Black et al. (2005) report evidence that corporate governance is an important factor for predicting the market value of South Korean firms. The authors construct a corporate governance index for 515 Korean companies listed in the Korea Stock Exchange. The study offers evidence consistent with a causal relationship between an overall governance index and higher share prices in emerging markets. Finally, the authors allege that they report the first evidence consistent with greater board independence causally predicting higher share prices in emerging markets. In the case of Spain, the literature on corporate governance is concentrated on the analysis of ownership structure and the effects of ownership concentration on the performance of the companies, which is justified given that ownership concentration is the main control mechanism in the Spanish corporate governance system. Nevertheless, the construction of an index that considers other dimensions of governance can offer a more complete picture of the Spanish reality. 3.3. Designing the Research: Questions and Hypotheses In the first part of the empirical research the objective is to assess the possible factors that make companies adopt different levels of governance under the same level of investor protection (legal, institutional and regulatory environment), what takes us to the following research question and hypothesis: Part 1: Determinants of the quality of governance Research question: Which observable factors make companies adopt different levels of governance under the same contracting environment? H1 (Hypothesis 1) There is a significant relationship between the variables selected as possible determinant factors and the level of corporate governance adopted by the companies of the sample. Moreover, the direction of the relationship is the one proposed by the literature.

17

In the second part, the objective is to assess the influence of the quality of governance on the performance of the Spanish public companies, thus the question to be answered and the correspondent hypothesis are: Part 2: Relationship between corporate governance and performance Research question: Does corporate governance influence corporate performance? H2 (Hypothesis 2) There is a significant relationship between the quality of governance and the performance of the companies. Besides, companies with higher quality of governance present better performance, ceteris paribus.

4. METHODOLOGY
Empirical research in the effectiveness of corporate governance mechanisms are subject to the problem of endogeneity and reverse causality. Nevertheless, endogeneity and reverse causality are under-explored theoretically and empirically (Bhren and degaard, 2003). According to the authors, endogeneity occurs when mechanisms are internally related, () and reverse causation is when performance drives governance (Bhren and degaard, 2003, p.2). And, according to Gompers Ishii and Metrick (2003, p.4) the governance structures of a firm are not exogenous, so it is difficult in most cases to draw causal inferences. For this reason, we make no claims about the direction of causality between governance and performance. Similarly, Bhagat and Jefferis (2002, p.3) analyze the efficiency of antitakeover mechanisms and state: (w)e argue that takeover defenses, takeovers, management turnover, corporate performance, capital structure, and corporate ownership structure are interrelated. Hence, from an econometric viewpoint, the proper way to study the relationship between any two of these variables would be to set up a system of simultaneous equations that specifies the relationships between these six variables. However, specification and estimations of such a system of simultaneous equations are nontrivial. Empirical research on the evidence about the relationship between governance mechanisms and performance can be classified according to the methodology used. Bhren and degaard (2003, p.8) propose the following classification presented in Table 1: TABLE 1 MECHANISM INTERACTION AND MECHANISM-PERFORMANCE CAUSALITY Causation Mechanisms Exogenous Endogenous One-way 1 2 Two-way 3 4

Source: Bhren and degaard (2003, p.8)

18

According to the authors, almost all studies can be classified in cell 1, where the econometric approach considers governance mechanisms as externally given, causation is supposed to run from insider ownership to performance, and where the singleequation regression typically contains one or two mechanisms. Among the most tested mechanisms of governance are the internal mechanisms: ownership structure, Board composition and managers compensation; and, the external mechanisms: active market for corporate control, capital structure and the presence of institutional investors. Himmelberg et al. (1999) are closed to cell 2. They analyze one-way causation from insider ownership to performance, similarly to the studies classified in cell 1. However, they point out that ownership structure is determined by the country-level investor protection, and in that sense, they treat ownership structure as an endogenous variable. According to Bhren and degaard (2003), research in cell 3 is not viable since twoway causation cannot be modeled without considering at least one governance mechanism as endogenous related to performance. Finally, in cell 4 are the empirical researches that estimate the coefficient of the governance mechanisms and performance measures through the use of simultaneous equations. Cell 4 methodology has been used in a series of studies: Agrawal and Knoeber, 1996; Barnhart and Rosenstein, 1998; Cho, 1998; Demsetz and Villalonga, 2001; Bhagat and Jefferis, 2002; Claessens et al. 2002; Klapper and Love, 2004; Bhren and degaard, 2003; Silveira, 2004, and Beiner (2006). In all these studies, most of the significant results disappear. Due to the occurrence of endogeneity and reverse causality, research in corporate governance should be developed in cell 4 methodology, more specifically with simultaneous equations systems. Nevertheless, Bhren and degaard (2003) argue that successful implementation of this method depends on whether corporate governance theory can offer well-founded restrictions on the equation system. In other words, we cannot find in the theoretical framework of corporate governance reference about whether and how governance mechanisms interact, which are the exogenous variables (not related to governance) driving two-way causation or the nature of the equilibrium in terms of an optimal combination of governance mechanisms for a given set of exogenous variables. The opened question is to know whether the results obtained using cell 4 methodology is reliable. Recent research and the extant literature in the field consider the use of different econometric approaches as very important for capturing the reverse causality between governance and performance and the potential endogeneity among the mechanisms of governance, as corroborated by all quoted studies. Nevertheless, still in line with the research on the field, the objective of this paper is to answer the research questions proposed through the use of statistics, concretely, through the use of OLS simple and multiple regressions (cross-sectional analysis). The use of more a sophisticated methodology, for instance simultaneous equations aims a database with a longer horizon of time for the development of a panel data, for example, in order to be robust and to avoid misspecifications. Thus, since our governance data were collect only for the year 2005, the use of this methodology is a suggestion for a future study. For this, our study belongs to cell 1 methodology. The vast majority of studies in corporate governance are classified in cell 1 methodology, and since my study is a first approximation to the dynamics of the Spanish non financial firms governance system, the methodology adopted is justified.

19

4.1.

Sample Selection and Data Collection

The sample is composed by all Spanish non-financial listed companies in the Madrid Stock Exchange (Mercado Continuo). The main data source is the web pages of the companies, but also the Spanish Securities Exchange Commission (Comisin Nacional del Mercado de Valores CNMV) and the Madrid Stock Exchange (Bolsa de Madrid) databases for all governance related information (board composition, ownership structure and control, etc.). All financial and accounting information (balance sheets, income statements, capital structure, industry/sector, book values, stock prices, etc.) were obtained from COMPUSTAT. The data refers to the year 2005 for all governance related information (GOV-I and board composition) and for the years 2004 and 2005 for all financial information (sales, assets, ROA, Tobins q, etc.) except for sales growth that captures the growth in sales in the last three years, covering the period from 2002 to 2005. The final sample is composed by 97 firms and the selection criteria are (1) to be a Spanish firm and (2) to not belong to the financial and real estate sectors. During the collection of the data, one firm has presented insufficient information to construct the index, so it was excluded from the sample. The rational for such selection criteria is that the foreign companies listed in the Madrid Stock Exchange are not located and operating in Spain, thus they are not exposed to the legal, institutional and regulatory environment holding in Spain. The financial and real estate sectors are regulated by specific rules which influence their governance model directly, for this reason they were excluded from the sample. 4.2. The Corporate Governance Index (GOV-I) and the quality of governance

The governance index is created for proxy quality of governance. It is constructed based on a questionnaire with binary objective questions and the answers must be obtained exclusively from secondary data. Since the ultimate objective is to measure the degree of transparency of the companies, the use of secondary data is justified (annual reports, companies webpage, and the securities exchange commission webpage CNMV). The questions of the governance index (GOV-I) were developed based on the Credit Lyonnais Securities Asia (CLSA) questionnaire used by Klapper and Love (2004) and in the questions developed by Brown and Caylor (2004) when building their Gov-Score. For a detailed description of provisions and antitakeover measures, the reference was the work of Gompers et al. (2003). The construction of the index is straightforward, I first code the 25 variables as 1 or 0 depending on whether the firm has satisfactory corporate governance standards or not. Each positive answer adds one point to the index, and the companies present a corporate governance level that ranges, in theory, from 0 to 25. The main source of information is the Annual Report on Corporate Governance prepared by the companies for the year 2005. The index is composed by four dimensions in order to assess good governance practices: (1) access and content of the information; (2) structure of the board; (3) ownership structure and control; and, (4) progressive practices. Table 9 provides the questions compounding the index. The governance index (GOV-I) is one proxy for quality of governance. Other proxies used in the study are: board independence (BIN), board size (BSZ), and duality CEO-

20

Chairman. These mechanisms are also contemplated and measured by the GOV-I. Table 2 provides a description of each dimension and the question that assess each of them. These dimensions could also be called subindices. TABLE 2 I-GOV DIMENSIONS DIMENSION QUESTIONS 1-7 8-16 17-20 21-25

GOV-I 1 GOV-I 2 GOV-I 3 GOV-I 4

Access and content of the information Structure of the board Ownership structure and control Progressive practices

4.3.

Determinants of the Quality of Governance

Based on the work of Himmelberg et al. (1999), Himmelberg et al. (2002) and Klapper and Love (2004), the governance determinants to be tested are: future growth opportunities, firm size, composition of firms assets, ownership structure, corporate performance, and listing in an American or European (non-Spanish) stock market, besides the control variable industry. Future growth opportunities are measured following Klapper and Love (2004) through the average annual sales growth over the past three years (2002-2005). There are three measures of performance, Tobins q, return on assets (ROA) and EBITDA. Tobins q reflects firm performance and also firm profitability, the other two accounting variables are used to proxy operating profitability (EBITDA) and net profitability (ROA). Firm size is proxy by the logarithm of the total assets. Finally, composition of firms assets is proxy by the ratio of fixed assets to net sales. Table 3 describes the variables, the rational for introducing each variable in the analysis explaining its possible influence in the governance of the companies and the code attributed to each of them. The governance index (GOV-I) is constructed for proxy quality of governance, as described in the previous section, and a cross-sectional OLS model is used to directly examine the relation between governance and the selected operational characteristics measured by the above mentioned variables. The general model to be tested is the following: GOV-Ii = + 1 GROWTHi +2 SIZEi +3 TANGi +4 INTSMi +5 IBEX-35i + +6 OWNCONi +7 PERFi +j INDji + i
EQUATION 1

21

The coefficients are expected to be statistically significant and to present the following signals: 1, 4, 5, 7 > 0; 3 < 0; since the effect of firm size and ownership structure on governance are ambiguous, in theory, no signal is expected for 2 and 6; j is the coefficient for the binary variable of industry, no signal is expected for it. 4.4. Relationship between Governance and Performance

This part of the study aims to contribute to the body of knowledge in the field of corporate governance by answering the following question: Does corporate governance influence corporate performance of the Spanish public companies? As described earlier, the direction of the causality is not clearly defined by the theory, nevertheless, it is conceived that causality may run both ways. Nevertheless, assuming that all relations are linear, in this second part of the study, I first run a series of regressions as an initial approach to assess the relationship between governance mechanisms and performance measured by Tobins q (Q) (equation 2). The regressions are run considering performance as an endogenous variable and the governance variables as exogenous but including other variables to control for observable firm heterogeneity. In order to assess the impact of quality of governance on the valuation of the firms, another equation is estimated with Tobins q (Q) as the dependent variable. Since Tobins q must also reflect firms profitability, ROA is included in equations (2) and (3) in order to capture a possible interrelation between operating profitability and firmspecific governance. Besides, the introduction of ROA is based on simple valuation models: Q may depend on ROA and Beta (Chi, 2005; Beiner et al., 2006). In this sense, ROA is introduced in equation 2 to capture a possible influence of operational profitability on firm valuation. Finally, as the calculation of the traditional Tobins q is costly both in terms of its data requirements and computational efforts (p.70), I propose the use of the approximation of Tobins q proposed by Chung and Pruitt (1994)6.

PERFi = + 1 GOVi +2 GROWTHi +3 SIZEi + 4 TANGi + 5 ROAi + +j INDji + i

EQUATION 2

Defined as: Tobins q (MVE + PS + DEBT)/TA. Where: MVE is the product of a firms share price and the number of common stock shares outstanding (or the market value); PS is the liquidating value of the preferred stocks; DEBT is the value of the firms short-term liabilities net of its short-term assets, plus the book value of the firms long term debt; and, TA is the book value of the total assets of the firm.

22

Where: PERFi is represented by Tobinss q (Q) and is measuring performance and firm profitability. GOVi represents the governance mechanisms: GOV-I (the Governance Index), BSZ (Board Size), BIN (Board Independence) and CEO-CHAIR (duality between CEO and Chairman of the board), ownership structure (OWNCON and OWNmain), and LEVER (Leverage). And finally, a final equation is estimated considering all governance mechanisms simultaneously. This regression allows for the adoption of different governance mechanisms at the same time, which is conceivable with the practice where companies adopt a series of mechanisms together. PERFi = + j GOVji +7 GROWTHi +8 SIZEi + 9 TANGi + 10 ROAi + + 11 LEVERi +j INDji + i
EQUATION 3

Where: GOVji represents the governance variables: GOV-I (the Governance Index), BSZ (Board Size), BIN (Board Independence) and CEO-CHAIR (duality between CEO and Chairman of the board). Table 4 provides a description of each research variable used in the study as well as the descriptive statistics.

23

TABLE 3 POSSIBLE DETERMINANTS OF GOVERNANCE

GOVERNANCE DETERMINANT Future Growth Opportunities REASONING CODE

A growing firm with large needs for outside financing has more incentive to adopt better governance practices in order to lower its cost of capital (Klapper and Love, 2003). The effect of size is ambiguous as large firms may have greater agency problems (because it is harder to monitor them, so they need to adopt better standards of governance to compensate. On the other hand, small firms may have better growth opportunities and greater need for external finance, so they may present better governance mechanisms (Klapper and Love, 2003). Managers and shareholders have incentives to avoid inside ownership stakes in the range where their interests are not aligned, although managerial wealth constraints and benefits from entrenchment could make such holdings efficient for managers (Jensen and Warner, 1988). Besides, several studies describe a positive and significant relationship between ownership concentration and corporate performance. The best is the performance of the company, the higher the governance standards we could expect, due to the lower external shareholders expropriation. Besides, the performance variables are used to assess the possible occurrence of reverse causality with corporate governance. The composition of a firm's assets will affect its contracting environment because it is easier to monitor and harder to steal fixed assets than intangibles. Hence, the firm operating environment will affect its governance system. (Himmelberg et al., 1999). Companies that are included in the Spanish Market Index (IBEX-35) are expected to adopt higher standards of governance. Companies that are listed in an American or European stock market (non-Spanish) must adopt higher standards of governance. Industry is expected to influence governance. Mostly because there are more regulated economic sectors than others, but also due to the competition holding in certain sectors.

GROWTH

Firm Size

SIZE

Ownership Structure (OWN)

OWNCON OWNmain

Q ROA EBITDA

Performance (PERF)

Composition of Firms Assets

TANG

IBEX-35

IBEX-35

Listing in another stock market

INTSM

Industry

IND

TABLE 4 SUMMARY OF THE RESEARCH VARIABLES

CODE VARIABLE Future Growth Opportunities Firm Size Corporate Governance Index Board Size Board Independence CEO and Chairman are not the same person Ownership Concentration Ownership of the main shareholder Tobins q Return on Assets Earnings Before Interests Taxes Depreciation and Amortization Composition of firms assets Capital Structure Belong to the IBEX-35 Listing in an International Stock Market DEFINITION Average sales growth in the last 3 years Log of net sales Index composed by 25 binary questions Total number of board members % of independent and external directors 1 if the CEO and the Chairman are not represented by the same person; 0 otherwise % of shares owned by the controlling shareholders (shareholders with more than 5% stake in the firm) % of shares owned by the main shareholder Ratio of the market value of equity plus the book value of debt to book value of total assets Net income / Total assets EBITDA / Total assets 96 Fixed assets / Net Sales Total debt / Total assets 1 if the company belongs to the IBEX-35; 0 otherwise 1 if the company is listed in an American or European (nonSpanish) stock market; 0 otherwise 95 96 97 97 0,10 0,85 0,23 0,26 0,33 0,09 0,56 0,25 0 0 0,11 1,03 0,18 0,44 0,47 -0,46 0,04 0 0 0 0,47 7,24 0,75 1 1 N 95 96 97 97 97 97 97 97 93 96 Mean 0,09 5,76 13,38 10,67 0,36 0,41 0,62 0,32 1,45 0,04 Median 0,07 5,74 14,00 10,00 0,33 0 0,66 0,25 1,12 0,04 Std Dev 0,16 0,85 2,77 3,78 0,19 0,49 0,22 0,23 1,02 0,09 Min -0,49 3,78 4,00 3,00 0,00 0 0,03 0,01 0,18 -0,54 Max 0,71 7,68 21,00 20,00 0,87 1 0,97 0,94 5,61 0,33

GROWTH SIZE GOV-I BSZ BIN CEOCHAIR OWNCON OWNmain Q ROA EBITDA TANG LEVER IBEX-35 INTSM

This table provides descriptive statistics for all variables included in the empirical analysis. The initial sample consists of 97 companies and the total number of observations for each variable is included in the table (N). The data is for the year 2005.

5. RESULTS
5.1. Descriptive Statistics The descriptive statistics for the governance index (GOV-I) are presented in Table 5 and the histogram and the normal curve of the distribution are drawn in Figure 2. The mean for the GOV-I is 13,38 and the median is 14, indicating a relatively symmetric distribution. Appendix 1 provides histograms for the total sample and for the firms that belong to the Spanish stock market index (IBEX-35). As one could expect, the companies that compose the IBEX-35 present a significantly higher mean for the GOV-I than the other firms compounding the sample. Besides, the histogram can reveal that there is a wide distribution for the GOV-I between the firms compounding the sample, the minimum value is 4 and the maximum is 21 (16% and 84%, respectively) which mitigates possible sample selection bias (Beiner et al., 2006). TABLE 5 THE GOVERNANCE INDEX (GOV-I)
2005 Governance Index (GOV-I) Minimum Mean Median Mode Maximum Standard Deviation Number of Firms GOV-I 10 GOV-I = 11 GOV-I = 12 GOV-I = 13 GOV-I = 14 GOV-I = 15 GOV-I = 16 GOV-I = 17 GOV-I 18 TOTAL Means by Dimension Access to information (GOV-I 1) Board structure (GOV-I 2) Ownership and control (GOV-I 3) Progressive practices (GOV-I 4) absolute 4,0 13,381 14,0 14,0 21,0 2,770 14 3 15 13 20 10 13 6 3 97 absolute 4,948 4,186 1,268 2,979

%
16,00 53,53 56,00 56,00 84,00 11,08

%
79,69 46,51 31,70 59,59

This table provides summary statistics on the distribution of GOV-I, the Governance Index for the Spanish public companies, and the dimensions (Access to information, Board structure, Ownership and control, and Progressive practices) for the year 2005. GOV-I was constructed from the questionnaire presented in Table 9 as described in Part 4.

Table 6 presents the GOV-I per industry with aerospace and defense being the sector with the higher score, 16, followed by transport/airlines with 15,5. Then, we have the vast majority of companies in the range between 14 and 12 and, finally, the sector presenting the lowest average is textile/apparel.

26

TABLE 6 THE GOVERNANCE INDEX (GOV-I) PER INDUSTRY

INDUSTRY N mean std dev 10 12,6 2,836 Food 5 10,6 4,393 Textil & Apparel 6 11,7 4,033 Paper Products 6 12,7 3,327 Pharmaceuticals 3 13,7 1,528 Leisure 5 14,2 2,049 Media Entertainment 2 15,5 2,121 Transports/Airlines 3 12,7 2,517 Logistics 7 14,3 1,890 IT & Telecom 6 14,5 2,588 Other Services 9 14,6 2,128 Gas & Utilities 8 13,8 3,412 Steel & Metals 9 12,9 3,060 Industrial Machinery 10 14,3 1,059 Construction 2 13,0 1,414 Chemicals 5 12,6 2,510 Engineering 1 16,0 0 Aerospace & Defense This table provides summary statistics on the distribution of the Governance Index GOV-I by industry.

Table 7 shows the correlation coefficients between Tobins q (Q) and the governance mechanisms used in the study. Despite not being significantly correlated with any variable, what is frequent in governance studies, we can analyze the nature of the relationship, if it is positive, negative or nonexistent.Thus, we find a positive correlation between Q and GOV-I as was expected, so as with the other governance mechanisms (BIN, BSZ, CEO-Chair and OWNmain), except for ownership concentration (OWNCON) with what we find a negative correlation. On the other hand, the GOV-I is significantly positive correlated with the other three mechanisms of governance (BIN, BSZ and CEO-Chair) and negatively correlated with the variables of ownership (OWNCON and OWNmain). It is also expected since the GOV-I is constructed based on the recommendations of the Spanish code of best practices (Aldama and Olivencia Codes) and is composed by the aforementioned four dimensions of governance that are also reflected in the other variables. The two ownership variables (OWNCON and OWNmain) present a significant and positive correlation between them and a negative correlation with board size and board independence. Nevetheless, the ownership variables present a positive correlation with CEO-Chair. Board independence is negatively correlated with board size and with CEO-Chair. And, finally board size and CEO-Chair are positively correlated. Table 8 provides descriptive statistics for the board structure variables. In relation with the size of the boards (BSZ) the average number of directors in the Spanish Boards is 10,67. This is an acceptable average size since the consensus is in something between 5 and 15. Nevertheless, it is clear that the size of the board depends largely on the number of influent shareholders that can nominate external directors as their representatives7. The more influent the shareholder the large the number of directors he/she indicates.

In the case of Spain, the external directors indicated by large shareholders are called dominicales.

27

TABLE 7 CORRELATION MATRIX BETWEEN GOVERNANCE MECHANISMS AND TOBINS Q

I-GOV Tobin's Q OWN CON OWNmain BIN BSZ 0,0271 1 (0,7962) -0,1109 -0,0059 1 OWN CON (0,2796) (0,9555) -0,0617 0,1294 0,6443*** 1 OWNmain (0,5486) (0,2163) (0,0000) -0,1098 1 0,2516** 0,0207 -0,2082** BIN (0,0407) (0,2843) (0,0129) (0,8440) 0,3036*** 0,0089 -0,0793 -0,0876 -0,0649 1 BSZ (0,0025) (0,9326) (0,4402) (0,3933) (0,5277) 0,0363 0,1407 -0,0588 0,0958 0,2716*** 0,0748 CEO-CHAIR (0,7239) (0,1693) (0,5671) (0,3507) (0,0071) (0,4758) This table reports Pearson correlation coefficients between Tobins q and the governance mechanisms for the year 2005. The variables are described in Table 4. Significance at the 10, 5 and 1 percent levels is indicated by *, ** and *** respectively. Tobin's Q

Board independence (BIN) is related to the number of independent directors in the board and the percentage is obtained considering all independent and external directors over the total number of directors. The best practices in corporate governance recommend that at least 50% of the directors compounding the board be independent. And, independent means with no relation with any shareholder or executive of the firm. In the case of Spain, the companies must inform if the director is executive, external indicated by a shareholder, or external and independent. This separation is extremely important, because investors are able to understand the dynamics of ownership and control within the target company just by reading the Corporate Governance Annual Report issued by the firms. From Table 8, we can appraise the independence of the Spanish boards with an average independence of 36,28%. It means that, on average, 4 out of 10 directors are independent in sample firms. It can and should be improved by the companies, mainly because there are firms with no independent directors and because the Aldama code strongly recommends the increase in the number of independent directors in the boards of the public companies. Finally, in 41,24% of the companies compounding the sample the CEO and the Chairman are not the same person, which means that in 58,76% of the inquired companies they are represented by the same person. TABLE 8 DESCRIPTIVE STATISTICS FOR THE BOARD STRUCTURE VARIABLES
BOARD STRUCTURE BSZ Mean Std Dev Median Min Max
10,67 3,78 10,00 3,00 20,00

BIN
36,28% 19,32% 33,33% 0,00% 86,67%

CEOChairman
41,24% -

N (Companies)

97

28

5.2. The Governance Index (I-GOV) description As described in Table 5, the mean for the GOV-I is 13,38 with a standard deviation of 2,77. A mean of 13,38 out of 25 (53,53%) can be interpreted as low if we consider that many questions compounding the index were related to basic concepts of corporate governance, such as access to information and transparency. Besides, one can observe that the questions with higher percentage of firms with a yes are those recommended by the Spanish codes of best practices on corporate governance (Aldama and Olivencia Codes). The company with the higher score is Arcelor which is actually a company that operates within the regulation of Netherlands8, in this sense it is expected that Arcelor shows a higher standard of governance. Table 9 describes in detail the percentage of yes obtained in each question. The results showed in Table 9 show that virtually all companies are concerned about governance issues, since 94,85% provide information in the corporate website about its governance system. On the other hand, only 64,95% provide the same information in English, many times we can see the icon English in the home page of the company, but after accessing the link what appears is a message such as information not available or page under construction. Question 4 is related to the disclosure of information about the companys future strategies and the projection of the results, it also shows a low rate (59,79%) of positive answers, mainly if taken into account that the sample is composed by the biggest publicly traded corporations in the country that are supposed to disclosure this kind of information. FIGURE 2 HISTOGRAM AND NORMAL CURVE OF THE I-GOV

30

25

20

15

10

5 Mean = 13,381 Std. Dev. = 2,770 N = 97 3 6 9 12 15 18 21

I-GOV

This table shows the distribution of the GOV-I for the 97 listed companies in the mercado continuo of the MSE. The index was constructed based on a questionnaire with 25 binary questions that were designed to proxy four categories of governance: (1) Access to information,(2) Board structure, (3) Ownership and control, and (4) Progressive practices. Better-governed firms have higher index scores.
8

The headquarters of Arcelor are located in the Netherlands and the company is subject to the legislation of this country.

29

The publication of the Corporate Governance Annual Report (question 22) is a question with a high percentage of positive answers, but one important point to mention is that many companies produce a very low quality report; it seems that they are just complying with an obligation, not to be out of the market. Moreover, question 5 Does the company disclosure information about its next or tree-year ROA or ROE targets? presents a disappointing 0%. This is the type of information that should be disclosed if the companies were interested in attract new capital, or even to renew the outstanding debt or equity. Since the main function of the managers is precisely to create value for the company, they were expected to disclose information about the creation or destruction of value. In question 9, Is the Chairman an independent, non executive director? only 9% of the companies have an independent director as the Chairman of the board, which is actually a progressive practice in Western Europe and highly recommended as a good governance practice. As a contrast, in question 17, The Chairman and the CEO are not represented by the same person, 41,24% of the companies have nominated the CEO as also the Chairman of the board. Considering the traditional ownership structure of the Spanish companies, as to say familiar ownership structures, the CEO as Chairman can clearly create favorable conditions to the appropriation of the benefits of control by the majority shareholders. The same is valid for question 8, Are the audit committee and the nominating committee exclusively composed by independent outside directors? , for which only a quarter of the companies said yes. Question 10, Does the CEO serve on no more than one additional board of other public company?, shows an interesting phenomenon: the cross participation of the same individuals in many boards. Analyzing the composition of the boards, one can figure out that the same names appear many times in different boards. There are Chairmen appearing in more than twenty five boards, and a question that naturally rises from this fact is if they have time to manage their own company if they are so busy participating in so many board meetings during the year. Another point is that, in question 10, I only consider the participation in boards other than those of the companies same group, which aggravates the problem. Questions 13 (Golden Parachutes) and 20 (Casting Vote) also show the power of the Chairman/CEO with 45,36% and 39,18% of the companies presenting these provisions, respectively. In relation to the board composition and functioning, question 11 Is the board composed by no less than 5 and more than 15 members? shows a 85,57% of the companies compounding the sample with an acceptable board size. This question is complemented by question 12 Is shareholder approval required for changing the board size?, which is actually dictated by the bylaws of the firms and, as a consequence, presents a 100% of positive answers. Question 14 shows a 80,41% of the companies not having representatives of banks in the board. This is an important question because having representatives of banks in the board is very negative and 20% of the companies present such problem. Nevertheless, in the Spanish case, banks play an important role as shareholders and, in order to adapt the research to this reality, the companies that have a no in question 14 are only those who inform that besides the equity relationship they also maintain commercial relations with the bank. In any case, this situation should be avoided by the firms, since there is a clear conflict of interests between the companies and the banks. Finally, question 15 shows that the Spanish public companies are far from having an independent board, since in only 31,96% of the firms compounding the sample the independent directors account for more than 50% of the board. In the item are board members elected annually? (question 16) another disappointing 1,03%.

30

TABLE 9 I-GOV QUESTIONS

Dimension of Governance # QUESTION OF THE GOVERNANCE INDEX (GOV-I) Does the company website provide information about its governance system? Does the company have an English version of its website where results and corporate governance related information are promptly updated (no later than one business day)? Does the company have an Investors Relation Department? Does the company disclosure enough information or analysts presentations with what any investor can make projections for the company? Does the company disclosure information about its next or tree-year ROA or ROE targets? Does the company publish/announce quarterly reports within two months of the end of the quarter? Has the public announcement of results promptly published in the web page of the company? Are the audit committee and the nominating committee exclusively composed by independent outside directors? Is the Chairman an independent, non executive director? Does the CEO serve on no more than one additional board of other public company? Is the board composed by no less than 5 and more than 15 members? 11 Board Structure 12 13 Is shareholder approval required for changing the board size? Have the Board approved any Golden Parachute Provision for the senior executives? Does the board include no direct representative of banks and other large creditors of the company? (having any representatives is negative) Do independent, non-executive directors account for more than 50% of the board? Are board members elected annually (they have a unified mandate of one year and the reelection is not automatic?) The Chairman and the CEO are not represented by the same person. Do directors receive part of their remuneration in stocks/stock options? Is directors stock ownership at least 1% but not over 30% of total outstanding shares? Does the Chairman have Casting Vote? Does the company offer tag along to the minority shareholders? Does the company publish the Corporate Governance Annual Report (as stated by the Aldama Code)? Does the board have outside advisors? Do directors term limits exist? Does mandatory retirement age for directors exist? Percentage of firms with a "YES" in the Question 94,85%

2 3 Access to Information 4 5 6 7 8 9 10

64,95% 87,63% 59,79% 0,00% 90,72% 96,91% 27,84% 9,28% 40,21% 85,57% 100,00% 45,36%

14

80,41%

15 16 17 Ownership Structure and Control 18 19 20 21 Progressive Practices 22 23 24 25

31,96% 1,03% 41,24% 16,49% 29,90% 39,18% 3,09% 90,72% 84,54% 71,13% 48,45%

31

In terms of the directors remuneration, question 18 points that only 16,49% of the firms believe in the stock options as a mechanism of governance capable of aligning the interests of managers and shareholders. Question 19 deserves deeper research on the causes of such problem, since only 29,90% of the boards have a percentage ranging between 1% and 30% of the total outstanding shares in their hands. Progressive practices in corporate governance were assessed through questions 21 to 25. Three questions reached a high rate of positive answers (questions 22, 23 and 24) showing that the companies are progressively incorporating the Aldama Code suggestions and are now publishing the Corporate Governance Annual Report (90,72%), establishing directors term limits (71,13%), and hiring outside advisors to assist the directors when necessary (84,54%). On the other hand, only half of the companies have mandatory retirement age for the directors (48,45%), for the rest it is still a problem, mainly in the case of family companies. Nevertheless, the critical point within this dimension is definitely the fact that practically no company (3,09%) offers tag along to the minority shareholders. 5.3. Empirical results The empirical analysis proceeds as follows: First, I estimate equation (1) using ordinary least squares (OLS) regressions in order to assess whether there is a relationship between the variables selected as possible determinant factors and the level of governance adopted by the sample firms. Second, I estimate equations (2) and (3) using OLS regressions to verify whether there is a significant relationship between quality of governance and performance. The equations include a series of control variables as described in section 4.3 and 4.4.

5.3.1. Determinants of governance

Table 10 presents the results from OLS regressions of GOV-I on the selected variables as its determinants. The results of the regressions show that future growth opportunities and composition of firms assets have a statistically significant positive effect on the quality of governance. Size also presents a positive effect on governance for the sample firms. However, it is not statistically significant. Moreover, performance also has a positive effect on the quality of governance showing that firms with higher performance adopt higher standards of governance9. These findings support Hypothesis 1, since these are the factors described in the literature as the possible determinants of governance. The effect of ownership concentration is ambiguous, so no signal was expected for this variable. For instance, a firm with a more concentrated ownership structure could present poorer quality of governance due to the high level of ownership concentration of the main shareholder; nevertheless, this fact could also take the company to adopt better governance practices to compensate for the higher likelihood of expropriation of the minority shareholders. I found a negative effect of both proxies of ownership concentration (OWN concentration and OWN main shareholder) on the quality of governance, this result is maintained for the three proxies of performance: Q, ROA and EBITDA. These results suggest that the quality of governance could be a function of the probability of expropriation of the minority shareholders due to the companys ownership structure and control. Besides, the coefficients indicate that the negative effect on governance is more accentuated for the

This result could indicate a problem of reverse causality since performance can improve governance but governance can also enhance performance.

32

ownership concentration in the hands of the main shareholder than for the ownership concentration of blockholdings. In summary, Table 10 presents six columns. Column (1) shows the results for the regressions of GOV-I on the selected variables along with Q and OWNCON. Column (2) shows the results for the regressions of GOV-I on the selected variables along with Q and OWNmain. Columns (3) and (4) show the results for the EBITDA and OWNCON and OWNmain respectively. Finally, Columns (5) and (6) show the results for the ROA as a proxy of performance and OWNCON and OWNmain respectively. TABLE 10 RESULT FROM OLS REGRESSIONS POSSIBLE DETERMINANTS OF THE QUALITY OF GOVERNANCE (GOV-I)
Dependent variable = GOV-I Independent variable (1) 11,876*** (0,003) 3,901*** (0,003) 0,677 (0,251) 0,547* (0,010) 0,387 (0,736) -0,374 (0,765) 1,284** (0,030) (2) 11,818*** (0,002) 4,016* (0,056) 0,695 (0,230) 0,630* (0,059) 0,224 (0,844) -0,518 (0,676) 1,444** 0,015 (3) 13,552*** (0,001) 3,719* (0,089) 0,511 (0,383) 0475 (0,157) 0,383 (0,739) 0,132 (0,910) (4) 13,530*** (0,000) 3,721* (0,082) 0,500 (0,389) 0,536 (0,111) 0,231 (0,841) 0,147 (0,899) (5) 14,093*** (0,000) 3,788* (0,077) 0,462 (0,428) 0,465 (0,160) 0,498 (0,664) -0,042 (0,971) (6) 13,871*** (0,000) 3,768* (0,074) 0,468 (0,418) 0,525 (0,114) 0,324 (0,779) 0,006 (0,996)

Constant GROWTH SIZE TANG IBEX-35 INTSM Q EBITDA ROA OWNCON OWNmain

2,055 (0,368)

2,372 (0,296) 3,835 (0,187) -1,137 (0,430) 3,911 (0,176)

-0,721 (0,613)

-0,853 (0,557)

-2,028 -1,590 -1,633 (0,155) (0,296) (0,251) Included Included Included Included Included Included Industry 0,198 0,220 0,151 0,162 0,162 0,171 Adjusted R2 0,018 0,011 0,048 0,038 0,037 0,031 Probability F 90 90 92 92 92 92 Companies (N) This table reports the results from OLS regressions of GOV-I on its determinants. The definition of the variables is provided in Table 4. Control variables for 17 industries (IND) were included in the regressions but do not appear here due to the limitation of space. The data is relative to the year 2004 and the GOV-I was constructed based on information relative to the year 2005. The numbers in parentheses are probability values for two-sided F test. ***, **, * denotes statistical significance at the 1%, 5% and 10% level respectively.

33

5.3.2. Governance and performance

Table 11 reports the correlation matrix between the I-GOV and the performance variables, as well as the control variables TANG, LEVER, GROWTH and SIZE. I-GOV has a positive significant correlation with EBITDA and ROA, and a positive correlation with Q. It is also positive and significantly correlated with GROWTH and SIZE, which could indicate that bigger firms present higher future growth opportunities and adopt higher standards of governance. In the same sense, LEVER is positively correlated with TANG and SIZE but negatively correlated with ROA as well as TANG is negatively correlated with ROA. These signals are expected since, otherwise being equal, firms with higher levels of leverage and tangibles should present lower ROA than if the same firm was all equity financed, due to the interest and depreciation expenses. Nevertheless, the EBITDA should be equal for both firms, all equity financed and the leveraged firm, and this is not the case for the companies compounding the sample. There is a negative correlation also between EBITDA, TANG and LEVER. This result should be better explored before concluding anything. Table 12 presents the results from the OLS regressions of performance (Q, EBITDA and ROA) on individual governance mechanisms along with the exogenous variables included in equations (2) and (3). Column (1) shows that the governance index (GOV-I) has a positive impact on firms performance measured by EBITDA and ROA and on firms valuation measured by Tobins q, nevertheless the effect is not statistically significant. This result partially support the hypothesis that firms adopting higher standards of governance are better valued by the market, ceteris paribus. The only governance mechanism that shows a significant coefficient is LEVER in all specifications; nevertheless, the results show a negative effect of leverage on firms performance10 which notably contradicts the mainstream hypothesis that leverage can improve governance due to the debt discipline. This is a typical result from a cross-sectional study that aims a deeper scrutiny through the use of a panel data and a simultaneous equation system to capture the reverse causality of leverage and performance, but also the endogeneity of leverage since the legal and institutional environment has a clear influence in the capital structure decisions. An important result from the regressions is that ownership structure, more specifically the ownership concentration of blockholdings (OWNCON) has a negative statistically significant impact on Tobins Q which means that more concentration of ownership is associated with lower firm valuation. This observation is consistent with the results obtained by De Miguel et al. (2005) for Spain that reports a negative impact of high levels of ownership concentration11 on the valuation of the firms since a concentrated ownership structure allows the expropriation of minority shareholders. Another important finding is that a higher shareholding of the largest shareholder (OWNmain) is associated with higher firm valuation which could be a consequence of the more efficient monitoring provided by concentrated shareholding.

Regressions of ROA and EBITDA on LEVER not shown here also exhibit a negative statistically significant relationship between the variables. 11 They report a quadratic relationship between firm value and ownership concentration and that beyond the breakpoint firm value is negatively affected by ownership concentration.

10

34

In columns (9), (10) and (11), all variables are considered together disregarding the interdependence of different governance mechanisms. The results remain unchanged, with LEVER appearing as the most significant mechanism of governance for the Spanish public companies. In that sense the degree of leverage of the firms is negative related with their valuation, but also with their performance. Ownership concentration of blockholdings is also statistically significant and has a negative impact on the valuation of the firms, while the concentration of ownership in the hands of the main shareholder has a positive impact on the value of the firms. Nevertheless, ownership structure has no significant impact on corporate performance. Finally, the coefficients of the exogenous variables (ROA, SIZE, GROWTH and TANG) show, generally, the expected signs. The adjusted R2 is ranging from 0,631 to 0,674 for the regressions on Q and are significantly lower for the performance variables EBIT and ROA (0,276 and 0,232 respectively).

TABLE 11

CORRELATION MATRIX BETWEEN I-GOV AND THE PERFORMANCE VARIABLES

I-GOV 0,003 (0,975) TANGIBLES -0,038 (0,712) 0,240** GROWTH (0,019) 0,344** SIZE (0,001) 0,027 Tobin's Q (0,796) 0,181* EBITDA (0,077) 0,175* ROA (0,089) LEVERAGE LEVER 1 0,366*** (0,000) 0,131 (0,204) 0,317*** (0,002) -0,169 (0,105) -0,141 (0,170) -0,191* (0,063) TANG GROWTH SIZE Tobin's Q EBITDA

1 -0,231** (0,025) -0,047 (0,652) 0,001 (0,995) -0,146 (0,158) -0,206** (0,045) 1 0,268*** (0,009) 0,249** (0,016) 0,282*** (0,006) 0,354*** (0,000) 1 -0,176* (0,091) 0,335*** (0,001) 0,299*** (0,003) 1 0,396*** (0,000) 0,349*** (0,001) 1 0,906*** (0,000)

This table reports Pearson correlation coefficients between I-GOV and performance variables (Tobins q, EBITDA and ROA) besides the control variables TANG, SIZE, GROWTH and LEVER for the year 2005. The variables are described in Table 4. Significance at the 10, 5 and 1 percent levels is indicated by *, ** and *** respectively.

TABLE 12 RESULT FROM OLS REGRESSIONS OF PERFORMANCE ON GOVERNANCE MECHANISMS

Dependent variable Independent variable (1) 1,203 (0,271) 0,010 (0,729) (2) 1,235 (0,259) (3) 1,326 (0,211) (4) 1,247 (0,254) (5) 1,716 (0,123) Q (6) 0,923 (0,386) (7) -0,034 (0,975) (8) 0,704 (0,562) (9) 0,676 (0,581) 0,008 (0,801) 0,046 (0,833) -0,226 (0,551) -0,063 (0,677) -0,891** (0,042) 0,871* (0,054) -1,145* (0,054) 1,997 (0,177) 0,028 (0,833) 1,566*** (0,003) 0,229** (0,041) Included 0,669 0,000 92 EBITDA (10) -0,419** (0,026) 0,003 (0,536) 0,011 (0,754) -0,028 (0,630) 0,001 (0,962) -0,028 (0,671) 0,028 (0,666) -0,192** (0,017) ROA (11) -0,268* (0,092) 0,004 (0,370) 0,003 (0,926) -0,055 (0,265) -0,009 (0,643) 0,011 (0,845) -0,007 (0,901) -0,173** (0,011)

Constant GOV-I Ln(BSZ) BIN CEO-CHAIR OWNCON OWNCON-Main LEVER ROA SIZE GROWTH TANG Industry
Adjusted R2 Probability F Companies (N)

0,050 (0,816) -0,123 (0,744) 0,029 (0,849) -0,394 (0,260) 0,526 (0,125) -1,589*** (0,004) 1,328 (0,333) 0,108 (0,377) 1,411*** (0,005) 0,268** (0,014) Included 0,672 0,000 92

3,179** (0,017) -0,088 (0,432) 1,302** (0,014) 0,151 (0,162) Included 0,631 0,000 92

3,239** (0,014) -0,089 (0,440) 1,320** (0,012) 0,152 (0,157) Included 0,631 0,000 92

3,227** (0,014) -0,081 (0,467) 1,338** (0,011) 0,156 (0,145) Included 0,631 0,000 92

3,267** (0,012) -0,078 (0,487) 1,320** (0,013) 0,155 (0,148) Included 0,631 0,000 92

3,276** (0,011) -0,106 (0,344) 1,412** (0,008) 0,147 (0,167) Included 0,637 0,000 92

3,487** (0,007) -0,054 (0,627) 1,322** (0,011) 0,165 (0,118) Included 0,643 0,000 92

0,060 (0,773) -0,200 (0,581) -0,053 (0,716) -0,892** (0,040) 0,864* (0,053) -1,157** (0,049) 2,054 (0,157) 0,033 (0,798) 1,580*** (0,002) 0,232** (0,036) Included 0,674 0,000 92

0,058*** (0,005) 0,096 (0,199) -0,009 (0,615) Included 0,276 0,003 94

0,034** (0,045) 0,144** (0,026) -0,010 (0,478) Included 0,232 0,009 94

This table reports the results from OLS regressions of Q on each governance mechanism along with the exogenous control variables. The definition of the variables is provided in Table 4. Control variables for 17 industries (IND) were included in the regressions but do not appear here due to the limitation of space. The data is relative to the year 2005. The numbers in parentheses are probability values for two-sided F test. ***, **, * denotes statistical significance at the 1%, 5% and 10% level respectively.

6. DISCUSSION AND CONCLUSIONS

This paper addresses two questions: (1) whether the quality of governance is determined by some firm specific and observable characteristics, and (2) whether quality of governance influences firm performance. In order to create a measure for the quality of governance, I construct a governance index (GOV-I) composed by twenty-five questions covering four dimensions, (1) access and content of the information; (2) structure of the board; (3) ownership structure and control; and, (4) progressive practices. The population under scrutiny is the Spanish non financial publicly traded companies, which resulted in a final sample of 97 firms. The GOV-I first dimension, access and content of the information, intends to capture the relevance firms put on transparency and the results show that the Spanish public companies are paying great attention to this issue. Over 85% of the companies in the sample comply with at least 4 questions out of 6 (approximately 70% of the total). The only exception is for question 5, Does the company disclosure information about its next or tree-year ROA or ROE targets?, with 0% as the rate of positive answers. It indicates that the firms do not disclosure information about their projections of future creation or destruction of value, but also about their strategic plans. The second and third dimensions can be grouped into a broader dimension, the decision making process of the top executives through the use of the Board and the control and ownership of shares. And, in this scenario, the Spanish companies show a divergent behavior from the one previously described. Compared with the first dimension, only 16,5% of the companies in the sample comply with at least 6 questions out of 9 (approximately 70% of the total) for the second dimension (board structure) and 10,3% of the companies comply with at least 3 questions out of 4 (75% of the total) for the third dimension (ownership and control). In contrast with those transparent companies offering a lot of information in their websites and showing themselves as really concerned about governance matters, after reading the procedures, bylaws and annual reports, one finds out the other face of these companies: a very tied structure of ownership, a high level of cross-participation on the boards, and the control centralized in the person of the CEO-Chairman. Furthermore, after a quick scrutiny, it is clear that the questions with higher percentages of yes in the I-GOV are those recommended by the Spanish codes of best practices (Aldama and Olivencia Codes). It could indicate that the companies are just following the normative to be adjusted to the market, and raises concern about whether they are really committed to governance matters. In the second part of the paper, I empirically assess (1) the determinants of the quality of governance and (2) the relationship between quality of governance and firm performance in Spain. The vast majority of studies in the field of corporate governance are focused in the US and emerging markets sets. Only recently, we are witnessing the appearance of this type of study for the European countries. The main contribution of this paper is to provide a picture of the Spanish corporate governance system. Spain is a very interesting set for exploring governance matters due to its particularities: (1) high levels of ownership concentration; (2) the boards are inefficient; (3) the capital markets are underdeveloped; (4) the market for corporate control is practically nonexistent; and, (5) there is a low degree of investor legal protection. My most important result for the first part supports the hypothesis that there is a significant relationship between the selected factors and the level of governance adopted by the

37

companies. This result is also in consonance with the literature. Specifically, Tobins q has a statistically significant positive impact in governance, what can indicate that performance drives governance. Another conclusion is that companies with higher future growth opportunities present better governance standards; this result is also in accordance with the literature. Firm size has a positive effect in the quality of governance, which means that bigger firms adopt higher standards of governance. Nevertheless, composition of firms assets showed a divergent relationship than the hypothesized one. The results indicate that, in the case of Spain, firms with a more hard assets structure tend to present better governance structures. This result can indicate two things: (1) a sample selection bias, since the companies compounding the sample are the biggest companies in the country with a high participation of fixed assets in its assets structure; or, (2) intangible assets can also be proxy by R&D investments, which means that there is a close relationship between intangibility and the extent to which firms invest in research and development, and Spain is known by the low rate of investments in R&D. Due to this fact, it is conceived that the results show a inverse relationship between tangibility of assets and quality of governance. In the second part of the empirical study, I assess the relationship between performance and the governance structure of the firms. The most important result in the second part of the study is the positive relationship between quality of governance and firm performance which supports the hypothesis that companies with higher quality of governance present better performance. Besides, I found evidence that high levels of ownership concentration of blockholdings have a significantly negative impact on the valuation of the firms (Q), but not in the corporate performance measured by ROA and EBITDA. For providing a more complete picture, I use the governance index (GOV-I) and six other governance mechanisms: board independence, board size, duality CEO-Chairman, blockholdings ownership, ownership of the main shareholder, and leverage, in order to assess the impact of governance on performance. The empirical results have important limitations. The most important limitation is the problem of time, since in the second part of the empirical research I am using only data for the year 2005. This is the only year for which I have governance related information, and the performance variables (Tobins q, EBITDA and ROA) are calculated based on the financial reports also for the year 2005. The solution for this problem of time is to collect data for the year 2006 and reassess the relationship, what is my intention for a future study. Another important limitation is the well described problem of endogeneity and reverse causality inherent to the research in corporate governance and that cannot be captured by OLS regressions. But could be also a problem of substitution, since I am using all mechanisms of governance together, as a consequence the results could be spurious. Nevertheless, the index is not directly related to the other governance mechanisms and the overlaps that cannot be controlled must not be causing major problems. Even so, the solution to all these problems either the reverse causality or the misspecifications is the use of simultaneous equations and panel data, what is my suggestion for a (near!) future study. Overall, the results confirm the positive relationship between governance and the determinant factors: performance, future growth opportunities and size what can be interpreted as evidence that the Spanish firms adopt better standards of governance to compensate for the low level of investor protection holding in the Spanish institutional environment. On the other hand, as was expected, our result show a negative relationship between quality of governance and the extremely high levels of ownership concentration holding in Spain. With regard to performance, the major result is the positive relationship between quality of governance and

38

firm performance (both in terms of valuation and profitability), but also the negative relationship between value and blockholdings ownership concentration and the positive relationship between shareholdings of the main shareholder and firms valuation. The study has also important implications for the practitioners that want to improve the governance of its companies and to investors and analysts that now have a more complete picture of the Spanish corporate governance system.

REFERENCES
Agrawal, A. and R. Knoeber (1996) Firm performance and mechanisms to control agency problems between managers and shareholders, Journal of Financial Quantitative Analysis, (31) 3 pp.377-398. Alchian, A. and H. Demsetz (1972) Production, information costs and economic organization, American Economic Review, (62) pp. 777-795. Barnhart, S. and S. Rosenstein (1998) Board composition, managerial ownership and firm performance: en empirical analysis, The Financial Review, (33) pp.1-16. Beck, T., A. Demirguc-Kunt and R. Levine (2001) Law, politics and finance, World Bank Police Research Working Paper n2585. Beiner, S. Drobetz, W., Schmid M. and H. Zimmermann (2006) An Integrated Framework of Corporate Governance and Firm Valuation, European Financial Management, Vol. 12 (2) pp. 249283. Berle, A. Jr. and G. Means (1932) The Modern Corporation and Private Property, New York: Macmillan. Bhagat, S. and R. H. Jefferis (2002) The Econometrics of Corporate Governance Studies, Cambridge, MA: The MIT Press. Black, B. (2001) The corporate governance behavior and market value of Russian firms, Emerging Markets Review, (2) pp. 89-108. Black, B., H. Jang and W. Kim, (2005) Does corporate governance predict firms' market values? Evidence from Korea, Journal of Law, Economics and Organization (forthcoming). Bhren, . and B. degaard (2003) Governance and Performance Revisited, ECGI Working Paper Series in Finance n28/2003. Brown, R. and L. Caylor (2004) Corporate Governance and Firm Performance, Working Paper, Georgia State University. Chi, J. (2005), Understanding the endogeneity between firm value and shareholder rights, Financial Management, Financial Management Association, Winter 2005 pp. 65-76. Cho, M. H. (1998) Ownership structure, investment and corporate value: An empirical analysis, Journal of Financial Economics, (47) pp.103121. Chung K. and S. W. Pruitt (1994) "A Simple Approximation of Tobin's q," Financial Management, Financial Management Association, (23) 3 pp.70-74. Claessens, S, S. Djankov, J. Fan, and L. Lang, (2002) Disentangling the incentive and entrenchment effect of large shareholdings, The Journal of Finance, (57) n7 pp.2741-2771. Cremers, M.K. and V.B. Nair (2005) Governance Mechanisms and Equity Prices, The Journal of Finance, Vol. 60 (6) pp. 2859-94. Coase, R.H. (1937) The Nature of the Firm, Economica, IV, New Series pp.386-405.

39

Dahya, J., J.J. McConnell, and N. Travlos (2002) The Cadbury Committee, Corporate Performance, and Top Management Turnover Journal of Finance, Vol.57 (1): 461483. De Andrs, P.; Lpez, F. J. & Rodrguez, J. A. (2002) Financial decisions and growth opportunities: A Spanish firms panel data analysis, European Financial Management Association (EFMA) Meeting, London. De Miguel, A., J. Pindado and C. de la Torre (2003) Ownership Structure and firm value: New evidence from the Spanish corporate governance system, EFMA 2003 Helsinki Meetings. http://ssrn.com/abstract=393464 access 02/5/2006. De Miguel, A., J. Pindado and C. de la Torre (2005) How do Entrenchment and Expropriation Phenomena Affect Control Mechanisms?, Corporate Governance: An International Review, (13) 4 pp.505-516. Demirg-Kunt, A. and V. Maksimovic (1996) Stock Market Development and Firm Financing Choices World Bank Economic Review, (10) pp.341-369. Demirg-Kunt, A. and V. Maksimovic (1998) Law, Finance and Firm Growth. Journal of Finance, (53) pp.2107-0137. Demsetz, H. (1983) The structure of ownership and the theory of the firm, Journal of Law and Economics, (26) pp. 375-390. Demsetz, H. and B. Villalonga (2001) Ownership structure and corporate performance, Journal of Corporate Finance, (7) pp.209-233. Demsetz, H. and K. Lehn (1985) The structure of corporate ownership: Causes and Consequences, Journal of Political Economy, (93) 6 pp. 1155-1177. Durnev, A. and H. Kim, (2005), To steal or not to steal: firm attributes, legal environment, and valuation, The Journal of Finance, (60) 3 pp.1461-1493. Eisenberg, T., S. Sundgren, and M.T. Wells (1998) Larger Board Size and Decreasing Firm Value in Small Firms, Journal of Financial Economics, 48: 35-54. Fama, E. and M. Jensen (1983), Separation of ownership and control, Journal of Law and Economics, (26) pp. 301-327. Gillan, S.L. (2006) Recent Developments in Corporate Governance: An Overview, Journal of Corporate Finance (12) 381-402. Gompers, P., J. Ishii, and A. Metrick (2003) Corporate governance and equity prices, Quarterly Journal of Economics, (118) pp.107-155. Hart, O. (1995) Corporate governance: Some theory and implications, The Economic Journal, (105) May pp. 678-689. Hart, O. (1996) An economists perspective on the theory of the firm, In: Buckley, P.J. and J. Michie (Eds.) Firms, organizations and contracts, 1st edition, Oxford: Oxford University Press. Hermalin, B.E. and M.S. Weisbach (1991) The effects of board composition and direct incentives on firm performance, Financial Management, (21) 4 pp.101-112. Himmelberg, C., G. Hubbard and I. Love (2002) Investor Protection, ownership and the cost of capital, World Bank Policy Research Working Paper n 2834, April 2002. Himmelberg, C., G. Hubbard, and D. Palia (1999) Understanding the determinants of managerial ownership and the link between ownership and performance, Journal of Financial Economics, (53) pp.353384. Jensen, M. C. (2001) Value maximization, stakeholder theory, and the corporate objective function, Journal of Applied Corporate Finance, (14) 3, pp. 8-21. Jensen, M. C. and J. Warner (1988) The distribution of power among corporate managers, shareholders and directors, Journal of Financial Economics, (20) pp. 3-24. Jensen, M. C. and W. Meckling (1976) Theory of the firm: Managerial behavior, agency costs and ownership structure, Journal of Financial Economics (3) pp. 305-360.

40

Klapper, L.F., and I. Love (2004) Corporate governance, investor protection, and performance in emerging markets, Journal of Corporate Finance, (10) 20328. Knight, F. (1921) Risk, uncertainty and profit, 1st edition. La Porta, R., F. Lopez-de-Silanes and A. Shleifer (1999) Corporate ownership around the world, The Journal of Finance, (54) pp. 471-517. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny (1998) Law and Finance, Journal of Political Economy, (106) pp.1113-1155. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny (2002) Investor protection and corporate valuation, The Journal of Finance, (57) 3 pp. 1147-1170. La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R. Vishny (2000) Investor Protection and Corporate Governance, Journal of Financial Economics (58) pp.3-27. Leal, R. and A. Carvalhal-da-Silva (2004) Corporate governance index, firm valuation and performance in Brazil, Working Paper COPPEAD/UFRJ. Lpez, F. J. and Rodrguez, J. A. (2001) Ownership structure, corporate value and firm investment: A simultaneous equations analysis of Spanish companies, Journal of Management and Governance, Vol. 5, pp. 179-204. Mak, Y.T. and K. Yuanto (2002) Size Really Matters: Further Evidence on the Negative Relationship Between Board Size and Firm Value, Working Paper. McConnell, J. and H. Servaes (1990) Additional evidence on equity ownership and corporate value, Journal of Financial Economics, (27) pp.595-612. Mrck, R., A. Shleifer, and R. Vishny (1988) Management ownership and market valuation: an empirical analysis, Journal of Financial Economics, (20) pp.293-315. North, D. (1990) Institutions, Institutional Change and Economic Performance, 1st edition, Cambridge, MA: Cambridge University Press, 152p. Rajan, R.G. and L. Zingales (2004) Saving capitalism from the capitalist, London: Random House. Rodriguez, E.F. and S.G. Anson (2001) Wealth Effects Associated with the Compliance with the Code of Best Practice: The Spanish Experience, Working Paper. Shleifer, A. and R. Vishny (1997) A Survey of Corporate Governance, The Journal of Finance, (52) 2 pp.737-783. Silveira, A. (2004) Governana Corporativa e Estrutura de Propriedade: Determinantes e Relao com o Desempenho das Empresas no Brasil, Doctoral Dissertation, Universidade de So Paulo, Brasil. Stulz, R.M. (1988) Managerial control of voting rights: Financing policies and market for corporate control, Journal of Financial Economics, (20) pp.25-54. Tirole, J. (2001), Corporate Governance, Econometrica, Vol.69 (1): 1-35. Williamson, O.E. (1975) Markets and Hierarchies: Analysis and Antitrust Implications, 1st edition, New York: Free Press. Williamson, O.E. (1985) The Economic Institutions of Capitalism, 1st edition, New York: Free Press. Wymeersch, E. (1998) A Status Report on Corporate Governance Rules and Practices in Some Continental European States, In Comparative Corporate Governance. The State of the Art and Emerging Research, Editors: Hopt, K.J.; Kanda, H.; Roe, M.J.; Wymeersch E. and Prigge, S., 1st edition, Oxford: Clarendon Press.

41

APPENDIX 1 HISTOGRAM AND NORMAL CURVE OF THE I-GOV - COMPARATIVE

30 25 20 15 10 5 0 30 25 20 15 10 5 0 3 6 9 12 15 18 21

IBEX-35 firms

All sample firms

I-GOV

IBEX-35 0 1 N 72 25 Mean 12,889 14,800 Std. Deviation Std. Error Mean 2,934 0,346 1,555 0,311

I-GOV