چه مسائلی و برای کدام شرکت ها برای اداره امور شرکت ها در بازارهای نوظهور پیش می آید؟ مدارک و شواهد از برزیل (و دیگر کشورهای حوزه بریک)

A central issue in corporate governance research is the extent to which “good” governance practices are universal (one size mostly fits all) or instead depend on country and firm characteristics. We report evidence that supports the second view. We first conduct a case study of Brazil, in which we survey Brazilian firms' governance practices at year-end 2004, construct a corporate governance index, and show that the index, as well as subindices for ownership structure, board procedure, and minority shareholder rights, predicts higher lagged Tobin's q. In contrast to other studies, greater board independence predicts lower Tobin's q. Firm characteristics also matter: governance predicts market value for nonmanufacturing (but not manufacturing) firms, small (but not large) firms, and high-growth (but not low-growth) firms. We then extend prior studies of India, Korea, and Russia, and compare those countries to Brazil, to assess which aspects of governance matter in which countries, and for which types of firms. Our “multi-country” results suggest that country characteristics strongly influence both which aspects of governance predict firm market value, and at which firms that association is found. They support a flexible approach to governance, with ample room for firm choice.

Capital market development has been linked to improved resource allocation (Wurgler, 2000) and economic growth (e.g., Levine and Zervos, 1998). In turn, capital market development has been related to protection of minority investors (e.g., Gleaser et al., 2001; La Porta et al., 1997, La Porta et al., 1998a and La Porta et al., 1998b). A number of articles also link firm-level corporate governance practices to firm value (e.g., Black et al., 2006a and Durnev and Kim, 2005). Overall these studies support the importance of firm-level corporate governance, especially in countries with weak legal protections for investors (e.g., Klapper and Love, 2004). How to improve corporate governance, however, is not clear. There are different approaches with distinct consequences. One approach treats legal rules as central. Good governance is achieved principally through rules that protect minority investors. (Examples of this approach include the Sarbanes-Oxley Act in the U.S.; New York Stock Exchange listing rules (requiring, for example, a majority of independent directors and an audit committee composed entirely of independent directors), and the OECD principles of corporate governance (OECD, 2004). This approach can be effective if many corporate governance practices are universal, so that a common set of rules can be applied to a broad spectrum of countries, and a broad spectrum of firms within each country. In contrast, if good corporate governance is often “local” — varying across countries, and across firms within a country, a more flexible approach will often be appropriate. Examples of this approach include comply or explain rules, such as the UK Combined Code on Corporate Governance (Financial Reporting Council, 2006), and multiple governance stock exchange listing tiers, exemplified by the Brazilian stock exchange, Bovespa, discussed below. There is, by now, substantial evidence that one size does not always fit all firms in all countries. Optimal governance likely differs between developed and emerging markets (e.g., Bebchuk and Hamdani, 2009), and potentially also between different emerging markets ( Durnev and Fauver, 2007). Within a given country, optimal governance may depend on firm characteristics (e.g., Arcot and Bruno, 2006, Bruno and Claessens, 2007 and Demsetz and Lehn, 1985). But we still know relatively little about the extent to which broad corporate governance principles can be applied across countries, or across firms within a country. If there is sufficient commonality, it could make sense to adopt “across the board” rules, both within and across countries, even if they do not perfectly fit every firm or every country. After all, there is also evidence that adoption of mandatory rules can be beneficial in some instances (e.g., Atanasov et al., 2010, on Korea; Black et al., 2006a, on Bulgaria). We address two principal questions. For both, we first study Brazil, then extend our analysis to the other BRIK countries, and evaluate our results in light of other existing studies. For both, we focus on emerging markets. The additional differences that surely exist between developed and emerging markets are outside our scope. Question 1: Which corporate governance rules are likely to be beneficial in emerging markets? One can readily compile a list of items that plausibly reflect good corporate governance and test whether, combined into an index, they predict firm market value (or performance). One can also test whether specific aspects of overall corporate governance (for example, board independence, disclosure, an audit committee, or cross-listing in the U.S.) predict firm market value on average, over many firms in many countries. These approaches are useful, but have important limits. Most centrally, they tell us little about which practices matter, for which firms and in which countries. One core problem is that different aspects of corporate governance are correlated. Thus, if one measures the overall predictive power of a list of governance measures, one does not know which elements drive the overall power. For instance: Gompers et al. (2003) develop for the U.S. a corporate governance index based on twenty-four provisions (G-index) and show that it predicts firm value. But Bebchuk et al. (2009) report that only six of these provisions fully drive the Gompers–Ishii–Metrick results. A related problem arises for studies that focus on a particular subset of governance measures. One then faces a classic omitted variables problem — one does not know whether the subset is truly important, or merely proxies for omitted aspects of governance. For example, a number of corporate governance studies rely on a 2002 survey by Standard and Poor's, which covers only disclosure. To overcome this problem, one needs a broad index that captures multiple aspects of corporate governance. One can then test the relevance of each aspect, controlling for the others. Moreover, what matters in corporate governance may vary from country to country, in ways not well captured by multi-country indices. As Bebchuk and Weisbach (2010) point out, the Gompers et al. (2003) and Bebchuk et al. (2009) indices principally measure take-over defenses, which are of limited relevance in countries in which most firms have controlling shareholders. The RiskMetrics (formerly ISS) measure focuses on features that are common in the US but often not found in other capital markets (Bebchuk and Hamdani, 2009). Variation across countries in ownership patterns and background legal rules limits what one can learn by assessing whether a particular governance measure or index matters on average across all countries. One needs to examine individual country results (a step often not taken in cross-country studies), to determine whether the results are driven by a subset, perhaps a small one, of the studied countries, and to which countries they apply. A third concern for cross-country studies is that the available indices are limited. The S&P index (e.g., Durnev and Kim, 2005) covers only disclosure, and is available only for 2002. The Credit Lyonnais Securities Asia index (Durnev and Kim, 2005 and Klapper and Love, 2004) includes subjective elements and is available only for 2001. The RiskMetrics (formerly ISS) index covers only developed countries (e.g., Aggarwal et al., 2009).3 Summing up, to identify what matters in corporate governance, in which countries, one needs a broad index that is (i) tailored to the nuances of particular countries; yet (ii) has sufficient commonality across countries to permit cautious generalization. One then needs to assess both the predictive power of the overall index, and the importance of different aspects of governance, controlling for other aspects of governance. Question 2: What aspects of corporate governance matter to which firms? A second, often understudied question involves which firms can benefit from which aspects of corporate governance. A number of hypotheses have been suggested in prior work. Firm size. Large firms could need “better” (more formal) governance to respond to their more complex operations. They could have greater potential for agency costs due to greater financial resources or less concentrated ownership. Conversely, small firms might face greater information asymmetry and accompanying agency costs. Investors could also be more attentive to how governance affects value at larger firms. Smaller firms, with lower institutional ownership could “fly under the radar.” ( Black et al., 2006a and Black et al., 2006b). To assess these possibilities, one must begin with a dataset that includes both large and small firms, yet the principal cross-country datasets that cover emerging markets cover only the largest firms in each country. Profitability. Highly profitable firms could need less “external” governance, or could have lesser need for external funds and therefore less need to improve governance to attract investors ( Black et al., 2006b and Durnev and Kim, 2005). Growth: Faster growing firms need external capital to sustain growth, and therefore might choose better governance to attract investors ( Bennedsen et al., 2012, Doidge et al., 2004 and Durnev and Kim, 2005). They also might have greater need for governance, as suggested by Hutchinson and Gul (2004). Asset tangibility (manufacturing): Firms which have substantial tangible assets are more amenable to external oversight, including creditor monitoring. They may therefore have less need for “equity” governance, and benefit less from governance than other firms ( Klapper and Love, 2004). Moreover, many corporate governance studies examine only manufacturing firms (e.g., Bertrand et al., 2002), leaving open the question whether one would obtain similar results for other firms. This discussion suggests that to address what aspects of governance matter to whom, one needs a broad sample of firms in each country. In this article we seek to address these two questions, using in-depth hand collected data on corporate governance practices in Brazil, and then extending prior studies of Russia, India, and Korea. These “BRIK” countries together comprise the four major “BRIC” emerging markets, plus Korea but minus China, which is unique due to government control of most major firms.4 Together, they provide a representative sample of the results one might expect in moderately developed, emerging markets. The BRIK countries differ in many ways, including different legal traditions, language, culture, geographic location, and important background legal rules (for example, use of non-voting shares). This increases the credibility of the pattern we find or, more often, fail to find. We seek here to find a middle ground between single-country studies, from which it is hard to generalize; and studies covering many countries, from which it is hard to determine what matters in which countries or to which firms. For each country, we build broad indices covering six distinct aspects of corporate governance: board structure, board procedures, disclosure, ownership structure, related party transactions and minority shareholders rights. These indices differ in some details across these countries to reflect local laws, but share substantial common features. Brazil is an important country to study for several reasons. It is one of the largest emerging market economies. Private benefits of control have historically been high and legal rules and firm-level governance have been weak.5 Weak legal rules leave more room for firm-level governance to vary in economically significant ways (Durnev and Kim, 2005). At the same time, firm-level governance has been rapidly changing. Finally, prior research on firm-level governance in Brazil has been limited. We are aware of three other articles that study the relation between corporate governance and firm value in Brazil — Carvalhal-da-Silva and Leal, 2005 and Leal and Carvalhal-da-Silva, 2007 and a contemporaneous study by Braga-Alves and Shastri (2011). All use governance indices based solely on public information. The first two studies did not find a robust association between firm-level governance and market value. Braga-Alves and Shastri find a positive association, and have the advantage of panel data, which permits firm-fixed-effects estimation. However, their sample and index raise concerns. They include government controlled firms and subsidiaries of foreign companies in their sample.6 Their governance index is based on a subset of the original Bovespa rules for Novo Mercado listing, and includes only 6 elements, several of which are problematic.7 They do not report which elements or which types of firms drive their results; thus, their study suffers from the weaknesses discussed above. We first demonstrate an economically important relationship between a broad Brazil Corporate Governance Index (BCGI) and lagged firm market value. We rely on hand-collected data from an early 2005 survey of Brazilian firms covering 2004 corporate governance practices. This allows us to go beyond public information in constructing our indices. A worst to best change in the index predicts almost a doubling in Tobin's q, from 1.16 to 2.13. We then assess which aspects of governance explain this overall association, by regressing Tobin's q against each of our six subindices, controlling for the remainder of the overall index. The overall index results derive mostly from subindices for ownership, board procedure, and minority shareholder rights. A disclosure subindex is significant by itself but loses significance when we control for the rest of BCGI, confirming the real-world importance of the omitted variable problem. Board structure, especially board independence, is widely seen as a central aspect of corporate governance. In contrast to the principal cross-country study of board independence (Dahya et al., 2008) and country studies of Korea (Black and Woochan, 2012 and Choi et al., 2007), we find in Brazil a significant negative association between board independence and firm market value. Thus, our results highlight the dangers in generalizing too readily concerning what matters in corporate governance. We then investigate for what types of firms the overall index, and each subindex, predicts higher firm value. We study the four broad firm characteristics discussed above: size, profitability, growth rate, and manufacturing versus non-manufacturing firms. We find a significant association between BCGI and market value for nonmanufacturing (but not manufacturing) firms, small (but not large) firms, and high-growth (but not low-growth) firms. Next we compare Brazil to the other BRIK countries. We obtain and then extend datasets for each other country (see Black et al., 2006a and Black et al., 2006b, for Korea; Balasubramanian et al., 2010, for India; Black et al., 2006c, for Russia). The governance indices in each country are similar, but reflect the rules and data limitations in each country. We find both common themes and differences across the BRIK countries. Across all four countries, governance predicts higher market value in small firms and high-profitability firms. The result for small firms is an important new finding — these firms are not included in the available multicountry indices, and thus are not part of the datasets for other multicountry studies. The result for high-profitability firms suggests that one cannot simply leave good managers alone, to run their businesses. A smaller gap between voting rights and cash flow rights predicts higher market value in Brazil and Korea, the two countries where we have this measure. Turning to differences, board independence predicts higher market value in Korea, lower market value in Brazil, and is insignificant in India. We also find major differences across countries on for which firms governance predicts higher market value. Overall, our results provide some common themes, but also underscore how much we do not yet know about what matters for corporate governance in emerging markets. This paper proceeds as follows. Section 2 discusses data. Section 3 describes the governance indices. Section 4 develops our methodology. Section 5 examines which aspects of governance predict firm market value, for which firms, in Brazil. Section 6 compares our Brazil results to the other BRIK countries. Section 7 concludes.

This article examines which aspects of governance matter and for which type of firms. We first conduct a case study of Brazil. We then assess commonalities and differences across four major emerging markets – Brazil, India, Korea, and Russia. For Brazil, we find an economically important relationship between an overall governance index and firm market value: a worst to best change in the index predicts almost a doubling in Tobin's q. Subindices for ownership, board procedure, and minority shareholder rights predict Tobin's q; while subindices for disclosure and related party transactions are insignificant. Strikingly we find a negative association between board structure, especially board independence, and market value. We find a significant association between corporate governance and market value for nonmanufacturing (but not manufacturing) firms, small (but not large) firms, and high-growth (but not low-growth) firms. Across the BRIK countries, we find both important commonalities and differences. Across countries, governance predicts higher market value in small firms and high-profitability firms. The small firm result is important because small firms are often unstudied. The result for high-profitability firms suggests that governance is at least as important for good performers as for poorly performing firms. A smaller wedge between voting rights and cash flow rights predicts higher market value in Brazil and Korea, the two countries where we have this measure. Board structure (independence): predicts higher market value in Korea, lower market value in Brazil, and is insignificant in India. Our analysis underscores the limits of broad cross-country analysis in assessing which aspects of corporate governance matter, for which firms. Use of a common index across many countries narrows the governance aspects and control variables that can be considered, making omitted variable biases likely. Moreover; an average effect across many countries doesn't tell us for which countries, and which firms, the aspect matters. Turning to policy implications, our results are not inconsistent with some mandatory minimum rules adding value. But in large part, they cast doubt on the wisdom of high regulatory minima, and on the extent to which different countries should adopt the same rules. Moreover, even if there are useful mandatory rules to be found, one can have little confidence as to what they are. An often better approach, our results suggest, will be to provide regulatory flexibility, coupled with sufficient disclosure so that investors can assess a company's governance choices. That flexibility could come through a comply-or-explain governance code, or as in Brazil, through firms choosing among different governance levels offered by the stock exchange. Overall, our results underscore how little we know about what matters for corporate governance in emerging markets and the core firm and country characteristics that predict when governance matters.