Corporate Governance and Protection of the Rights of Minority Shareholders in China

Transcription

1 Corporate Governance and Protection of the Rights of Minority Shareholders in China Center for China Financial Research (CCFR)* Faculty of Business and Economics The University of Hong Kong April 2002 Abstract The issue of corporate governance in China is examined in this study. We identify several important governance mechanisms for China s publicly listed companies and construct four sets of variables based on these mechanisms. The variables include measures of ownership structure, the board, managerial incentives, corporate control, and financial transparency. We then create a ranking index for corporate governance (G-index) based on these variables. Finally, we relate the G- ranking index to measures of corporate performance. We find that better-governed companies are associated with higher profitability as measured by ROA and ROE, higher stock market valuation as measured by the ratio of market value and book value of the net asset, and lower market turnover ratio. The results indicate that good corporate governance matters greatly in China s emerging stock market. Keywords: Corporate governance, Corporate performance, Stock valuation, Market premium *Project members: Chongen Bai, Qiao Liu, Joe Lu, Frank M. Song, and Junxi Zhang. We thank Li Chuntao for his excellent research support. 1

2 1.Introduction Corporate governance has always been an important issue of interest to academics, businesses, and policy makers. Policy makers in many countries have been encouraging companies to adopt higher governance standards. International organizations have also actively participated in the movement for better corporate governance. For example, the International Monetary Fund has demanded that governance improvements should be included in its debt relief program, especially during the most recent Asia financial crisis. The World Bank and the Organization for Economic Cooperation and Development (OECD) have also been working hard towards raising the standard of corporate governance. Notably, in 1998, OECD issued its influential OECD Principles of Corporate Governance, which are intended to assist member and non-member countries in their effort to evaluate and improve the legal, institutional and regulatory framework for corporate governance in their economies. In addition, many private companies, such as Standard & Poor, California Retirement Pension System, and McKinsey, are calling for sweeping reforms of corporate governance of companies, especially publicly listed firms in emerging economies. Corporate governance has also gained unparalleled importance in China. Although China has made remarkable progress in its development since the start of economic reform in 1978, some obstacles remain that hinder further economic development and growth. One of the obstacles seems to be the lack of a sustainable efficient and competitive market-based economic mechanism. A significant number of corporations, both state-owned enterprises (SOEs) and corporatized firms, are often fraught with serious problems of low economic efficiency, ambiguous property rights, weak corporate governance and generally poor financial discipline. For example, recent official statistics suggest that about one-third of all SOEs are loss-makers, another third either break even or are plagued with implicit losses while the remaining one-third are marginally profitable. Corporate China's lackluster performance is believed to be the direct consequence of ineffective governance. All these problems point to the crucial role of good corporate governance in sustaining the growth of the economy. In this project, we seek to study corporate governance of China s listed companies. It is believed that China s listed companies are among the best corporations in China. The government gives preferential treatment to better- managed SOEs to allow them to be listed and raise capital in the stock market. Over a relatively short period of time, the market value of China's listed companies has risen to over 4,000 billion Yuan, or roughly 50% of China's GDP. It is reported that there are over 60 million registered individual shareholders, 1/6 of them actively trade and hold stock. Therefore, the governance and performance of China's listed companies has enormous economic significance on both the economy as a whole and the wealth of individuals. Hence, we believe it is important to have a good understanding of corporate governance structure in China's listed companies. Our effort also echoes the main policy goal of the regulatory agency Chinese Securities Regulatory Commission (CSRC) for this year. Early this year, the CSRC designated year 2002 as the year for better corporate governance for China s listed companies. To conduct a test on the proposition that corporate governance behavior affects the performance and the market value of Chinese firms, one needs an objective and reliable dataset on corporate governance rankings for the publicly listed companies in China. To our best knowledge, such a dataset is not available. Thus, in this project we devote a considerable amount of time to construct an index on corporate governance, dubbed as the G index. In 2

3 constructing the G index, we follow the corporate governance mechanisms suggested by the theoretical literature and consider four groups of variables. The first group of variables is related to the tunneling activities by the controlling shareholder; Tunneling refers to the diversion of corporate resources from the corporation (or its minority shareholders) to its controlling shareholder. The second is related to monitoring, while the third is related to managerial incentives; Finally, we consider several measures of financial transparency. The paper is organized as follows. Section 2 reviews the theoretical and empirical literature on corporate governance studies and summarizes major mechanisms of corporate governance. Section 3 discusses the variables used in our construction of the ranking of corporate governance and the methodology of the ranking analysis. Section 4 presents the ranking results and relates the corporate governance ranking with corporate performance and stock valuations. Section 5 concludes the report. 2. The literature Over three hundred years ago, in his masterwork The Wealth of Nations, Adam Smith raised the issue of the separation of ownership and stewardship in joint-stock corporations. Effective corporate governance has always been an important issue in market economies. Modern academic literature on corporate governance stems from the seminal book by Berle and Means (1932), who argued that, in practice, managers of a firm pursued their own interests rather than the interests of shareholders. The contractual nature of the firm and the principal-agent problem highlighted by Berle and Means led to the development of the agency approach to corporate finance. Over the years, in particular in the last quarter of the 20 th century, there has seen rapid growth in both the theoretical and empirical studies. In this section, we briefly review some of the main developments in the literature. The agency approach to corporate governance attempts to provide answers to the key question How can shareholders ensure that non-owner managers pursue their interests? The literature describes seven important corporate governance mechanisms that encourage managers to act in the interests of the shareholders. They are, respectively, the board of directors, executive compensation, the market for corporate control, concentrated holdings and monitoring by outside investors, debt, product market competition, and financial transparency and information disclosure. Besides, there may also exist a potential conflict among shareholders. It is well known that even in developed economies, tunneling the diversion of corporate resources from the corporation (or its minority shareholders) to the controlling shareholder can be substantial and value-destructive. Although some tunneling (especially in emerging markets) takes the form of theft or fraud, legal tunneling does take place in developed countries as well. Tunneling usually takes two forms. First, a controlling shareholder can simply transfer resources from the firm for his own benefit through self-dealing transactions. Second, the controlling shareholder can increase his share of the firm without transferring any assets through dilutive share issues, minority freezeouts, insider trading, creeping acquisitions, or other financial transactions that discriminate against minorities. Thus, a good corporate governance structure requires directors to follow two broad principles. First, the duty of care requires a director to act as a reasonable, prudent, or rational person would act in his position. Second, the duty of loyalty, or fiduciary duty, requires that insiders do not profit at the expense 3

4 of shareholders, or of the corporation as the case may be depending on whom they legally owe loyalty to. Next, we review each of the seven mechanisms in turn. (1) The board of directors In theory at least, the board of directors is the first instrument through which shareholders can exert considerable influence on the behavior of managers in order to ensure that the company is run in their interests. Empirical studies, however, are complicated by the fact that due to the well-known historical, political, social, economical, cultural and legal differences across countries, the structure of boards is significantly different. Nevertheless, the evidence available suggests that countries share common features with regard to this mechanism. The empirical literature on the relationship between board composition and firm performance obtains the following findings: (1) Firms with boards containing a majority of independent directors do not perform better than firms without such boards; (2) A moderate number of inside directors is associated with greater profitability; (3) In Japan, although the presence of outside directors on the board has no effect on the sensitivity of CEO turnover to either earnings or stock-price performance, concentrated equity ownership and ties to a main bank do have a positive effect; and (4) There is a strong inverse relationship between CEO turnover and firm performance in some countries. (2) Executive compensation The second method of ensuring that managers pursue the interests of shareholders is to structure compensation appropriately, where the measures used to motivate managers include both stock prices and accounting based performance measures. Although most of the empirical studies are constrained by data availability, the limited finding seems to suggest that there is a positive relationship between executive pay and performance in the US, Germany and Japan. (3) The market for corporate control It is generally believed that the existence of an active market for corporate control is essential for the efficient allocation of resources. It allows inefficient managers to be removed and replaced with able managers who can gain control of large amounts of resources in a short period of time. The market for corporate control can operate in three ways: proxy contests, friendly mergers and hostile takeovers. Proxy fights do not usually unseat the existing board of directors successfully because share holdings are often spread among many shareholders. Friendly mergers and takeovers occur in all countries and account for most of the transaction volume that occur. In some developed countries, it ranges from 60% to 90%. For hostile takeovers, they do occur fairly frequently in the US and UK, however, much less so in Germany, France and Japan. Empirical studies suggest that takeovers in the past did significantly increase the market value of target firms, although the increase in value for bidding firms was zero and possibly even negative. Studies using accounting data find that changes and improvements in operations can at least partially explain takeover premia. (4) Concentrated holdings and monitoring by outside investors 4

5 It is believed that one of the most important ways through which a firm maximizes its value is through concentrated ownership of the firm s shares. The reason is that more wealth commitment by owners encourages monitoring and thus improves firm performance. Therefore, in general, concentrated equity ownership is regarded as a positive mechanism in corporate governance. But, it should also be noted that there are costs associated with-such monitoring by large shareholders: it may restrict the misuse of resources ex post but may also blunt ex ante managerial initiative. Moreover, large share-blocks result in incentives to monitor but also lead to a lack of liquidity. Similar arguments can be made with regard to outside investors, in particular institutional investors. In Japan and Germany, because of the lack of a market for corporate control, financial institutions act as outside monitors for large corporations. It is generally agreed among economists, although the main-bank system in Japan has problems, it is still effective to some degree in achieving good corporate governance. For Germany, until recently the view has been that banks provide effective monitoring of firms. Some economists argue that close relationships between banks and firms in Germany involve costs as well as benefits for the firm. Nevertheless, widespread equity ownership should increase firm monitoring and eventually improve performance. (5) Debt It has been argued that debt is a useful force in ensuring managerial discipline in corporations. The argument is that default on interest or principal can be personally costly to managers, therefore legally required interest payments force managers to focus on generating positive cash flows and prevent them from squandering resources. However, debt can have undesirable effects on managers behavior. When a firm has significant debt outstanding, managers would have an unduly incentive to take risks and may even accept projects that destroy value. The well-known debt overhang problem in the US in the 1980s is a good example. (6) Product market competition Another powerful force for solving agency problem is competition in product markets. If the managers of a firm waste resources, the firm will eventually fail in product market. Hence, increased competition reduces managerial slack and may be helpful in limiting efficiency losses. (7) Financial transparency and adequate information disclosure There is no doubt that financial transparency and adequate information disclosure are of ultimate importance in all countries, particularly developing ones. Managers play a vital role in securing the interests of not only the existing owners but also potential investors. Honest managers will attempt to provide sufficient, accurate and timely information regarding the firm s operations, financial status, and external environment. In sum, maximizing shareholders' value should be managers' mission; good corporate governance helps value-maximization. Hence an interesting question arises as to whether improved corporate governance actually pays off. For US firms, evidence that governance 5

6 practices matter is scarce. For example, the proportion of independent directors on a company s board (or whether the company has a majority-independent board) has no statistically significant effect on performance. Similarly, neither overt activism by institutional investors, nor insider share ownership, nor ownership by outside blockholders, nor a firm s committee structure, has a measurable effect on performance. Economists argue that the problem does not lie in the proposition that firms corporate governance behavior affects their market values, but lies in the data. A number of studies are made on countries with weak laws governing the behavior of firms and their insiders (managers and large shareholders), and weak norms for insider conduct. It is then found that good governance is actually rewarded with a higher market valuation. Shareholders in emerging markets are willing to pay a premium for good governance standard, averaging around 10% to 12%. For countries like Russia, this premium is even higher. 3. Construction of the G Index In this project, we try to quantify and evaluate the relative quality of corporate governance for each of the listed company in Shanghai and Shenzhen Stock Exchange. To accomplish this, we construct a corporate governance index, the G Index, for each of the listed company. We then rank the companies by the value of the G Index. In order to construct a quantitative index of corporate governance for every listed company, we need to quantify various theoretical aspects of corporate governance. In this study, we consider several factors that potentially affect the quality of corporate governance. The choice of the variables is guided by our understanding of the corporate governance mechanisms discussed in Section 2. However, due to data availability, some of the important variables are missing. In the mean time, we have added a few variables to capture the uniqueness of China s listed companies. 3.A. Definition of variables The first group of variables is related to tunneling (or expropriation of small shareholders by the controlling shareholder). Some of them also affect governance through other channels. (1) Shareholding of the largest shareholder top1 This variable potentially has two conflicting effects on the quality of corporate governance. One is related to the monitoring role of the largest shareholder, who has a stronger incentive to monitor the manager when his stake in the firm is high. The other effect arises out of the potential tunneling by the largest shareholder. When he can control the firm, he may use his control to benefit himself at the expense of other shareholders. The larger is his shareholding, the weaker is the threat to his control and hence the worse is his tunneling. (2) The firm has a parent company parent dummy When the largest shareholder is another firm, the scope for tunneling is wider. There are many more channels for a company than an individual to tunnel. The parent company can expropriate other shareholders of the concerned firm through various business dealings between them, or connected transactions. The most commonly used tool is transfer pricing. This is the reason for us to consider the next variable. 6

7 (3) The number of connected transactions ct_num reported by the firm in (4) Concentration of shareholding in the hands of the second to the tenth largest shareholders cstr2_10 = sum of squares of the percentage shareholding by the 2 nd to the 10 th largest shareholders This factor should have a positive effect through three channels. Firstly, these other large shareholders are the obstacles to the tunneling activities by the largest shareholder. Secondly, they enhance the efficiency of the market for corporate control. When the management under-performs, these large shareholders can either initiate a fight for corporate control or help an outsider s fight for control. Thirdly, these large shareholders also serve as monitors of the management. The higher the concentration of shareholding in the hands of these large shareholders, the stronger these roles. The second group of variables is related to monitoring. (5) The CEO is the chairman or a vice chairman of the board of directors ceo_is_top_dir dummy The board of directors should play a role of a monitor of the management. When the top manager, the CEO, controls or partially controls the board, it is hard for the board to play an active monitoring role. (6) The proportion of outsider directors out_in_ratio the ratio of the number of directors without pay with respect to the number of paid directors. Paid directors are often members of the management team. If they dominate the board, the board is not expected to play an effective monitoring role. The third group of variables is about managerial incentives. (7) Shareholding by the top five managers of the firm top5 The interests of the top managers are better aligned with the interests of other shareholders if the former have more stakes in the firm. (8) Dividend payout ratio div_earning_ratio the ratio of dividend paid in year 2000 with respect to year 2000 distributable earnings When the firm only pays out a small proportion of distributable earnings, the firm has more free cash flow a la Jensen. Free cash flow gives the management room for slack. For example, they may waste the free cash flow on unprofitable projects or on perks. The final group of variables concerns financial disclosure. (9) auditor opinion - audit Auditors conduct independent examinations of the financial statements prepared by a company. An audit report should state whether the financial statements are prepared in 7

8 accordance with the generally accepted accounting principles. It should also identify areas in which the generally accepted accounting principles are not followed. Auditors express an opinion regarding the financial reports taken as a whole. This variable represents auditor's opinion on a company's financial reports. It is a dummy variable that takes the value of 1, if the auditor issues an unqualified or clean opinion; i.e. the auditor believes that the financial reports are presented fairly, in all material aspects, in conformity with the generally accepted accounting principles. 2, if the auditor issues a qualified opinion; i.e. the auditor believes that the financial statements are presented fairly, except for the parts that the qualification relates to. 3, if a disclaimer of an opinion is issued by the auditor; i.e. the auditor does not have sufficient information to express an opinion. 4, if an adverse opinion is expressed by the auditor; i.e. the auditor believes that the financial statements taken as a whole are not prepared in accordance with the generally accepted accounting principles. We notice that most companies in our study received clean opinions from their auditors. In recent years, the media and the regulatory agency have exposed several companies who colluded with their auditors to publish false and misleading financial statements. We believe that the exposed cases were the most notorious ones; there could be many more cases. However, in our opinion, the auditor's opinion still provides important information regarding a company's corporate governance. Companies with strong corporate governance are less likely to issue financial reports that are not in accordance with the generally accepted accounting principles. (10) The number of law suits a company involves lawsuits This variable measures the number of lawsuits a company involves at the end of the fiscal year. The lawsuit information is gathered from the Wind database. We include all lawsuits in which a company involves, either as the plaintiff or the defendant. In developed economies, businesses constantly deal with lawsuits. The number of lawsuits a firm involves in may not provide information on its corporate governance. However, in China, the legal and business environments are unique. We believe that generally companies with strong corporate governance are less likely to entangle in lawsuits. We notice that there is no database, which comprehensively covers lawsuits information of China s publicly traded companies. Thus, this variable only serves as a crude measure. (11) Debt-to-equity ratio d_e This variable measures a company's debt to equity ratio. Debt is a company's year-end total liabilities; equity is its total shareholders' year-end equity. We employ this measure mainly because the debt-to-equity ratio captures a company's financial distress. Companies with strong corporate governance and good management are less likely to face financial distress. (12) 1/Accounts payable turnover ratio acc_index The accounts payable turnover ratio measures how quickly a company pays trade credit. It is computed as follows: Cost of goods sold Average accounts payable 8

9 The ratio proxies a company's liquidity. A high value normally indicates that a company pays its trade credit timely. Defaults on trade credits are a serious problem in China. We expect that better-managed companies are less likely to default. However, this ratio may not reflect the fact that a company pays some of its creditors on time but is late with others, which is quite common in China. The second set of variables is related to corporate performance. We use the following two measures for corporate profitability. (1) Return on asset (ROA) is measured by the ratio of net profit in year 2000 and the average asset of year 1999 and (2) Return on equity (ROE) is measured by the ratio of net profit and the market value at the end of year ROA and ROE are two most used measures of corporate profitability. In addition, we also use the measure of market value to book value of company s net asset, Market/Book, to gauge the market valuation of the company. The turnover in the trading of company s stock, turnover, is measured by averaging monthly turnover rate for the year 2000 where the monthly turnover rate is approximated by total A shares traded during the month divided by the total number of A shares. Finally, we measure the volatility of the stock returns, Std, by calculating its standard deviation of the 12 monthly returns in year 2000 and divide the standard deviation by the mean of monthly returns in year B. Summary Statistics Table 1 presents the summary statistics of the variables defined above. In panel A, we present the 12 variables used in forming the corporate governance ranking. The mean for the dummy variable parent is 0.78, indicating on average, 78% of the sample firms have a parent company. The mean for the variable audit is 1.08 with lower standard deviation, indicating most firms received unqualified or clean opinion from their auditors. The mean of the top largest shareholder, top1, is 0.43, with highest value equaling This suggests that on average the largest shareholders hold a significantly large portion of shares in China s listed companies. The mean of the variable ceo_is_top_dir indicates that more than 30% of CEOs in China s listed companies are also major directors in the board. The number of lawsuits in which a company involved, lawsuits, has a mean of 0.12, with the highest number being 15. The mean for the variable, ct_num, is 0.38 with the highest number being 6. The mean and the standard deviation for the concentration of the second to the tenth largest shareholders, cstr2_10, are and 28.34, suggesting a big variation of the concentration ratio. The mean of top5 variable is % with standard deviation %, indicating that top managers own very little of their company. The number of outsider directors in the sample companies is not low, with mean of 1.19 and standard deviation The mean of dividend/net earnings variable div is 14%, with the minimum being and the maximum The negative number indicates that some companies pay out dividend even in loss while the large positive number suggests that some companies distribute a lot of dividend out of their net earnings. In either case, we believe they are not signs of responsible governance. The mean for the measure of short-term liquidity, acc_index, is 0.19 with a standard deviation of 0.58, indicating big 9

10 variation among sample firms. Finally the mean for the debt/equity ratio d_e has the value of 0.23 and standard deviation of The highest d_e ratio is In panel B, we report summary statistics of several corporate performance variables. The mean for the first measure of profitability, ROA, is 3.23% with a very high standard deviation 9.20%. Another measure of profitability, ROE, has the mean of 1.03% and a standard deviation of 4.28%. The mean of market-to-book value of company s asset, market/book, is 3.56, with the standard deviation being The relatively high mean value for market/book suggests a booming stock market in the sample period. The mean for the measure of turnover of trading, turnover, is 0.43 and the standard deviation is Finally, the mean of the volatility measure, std, is 3.42 with a standard deviation of C. Ranking Methodology According to our theoretical analysis, we divide the variables used in empirical ranking analysis into two broad groups. The first group of variables are: (1) Shareholding of the largest shareholder top1; (2) the firm has a parent company parent dummy; (3) the CEO is the chairman or a vice chairman of the board of directors ceo_is_top_dir; (4) the number of connected transactions ct-num; (5) the number of law suits a company involves sue-num; (6) auditor opinion audit;(7) debt-to-equity ratio d_e; (8) the inverse of accounts payable turnover ratio acc_index. The higher the value of each variable, the lower the rank of corporate governance. The second group of variables are: (1) concentration of shareholding in the hands of the second to the tenth largest shareholders cstr2_10; (2) shareholding by the top five officials of the firm top5; (3) the proportion of outside directors out_in_ratio; and (4) dividend payout ratio div_earning_ratio. The lower the value of each variable, the lower the rank of corporate governance. We sort the variables in the first group in descending order, and the variables in the second group in ascending order. And the ranking of the companies is generated accordingly. Specifically, we rank each company according to each of the 12 variables. After obtaining the ranking according to each variable, we divide it by the total number of available observations in the study and multiply the resulting measure to obtain a normalized value from 0 to 100. Finally, the G index is constructed in such a way that we summarize the individual rankings for each company according to the following weighted average index: G = Σ i (w i * rank_v i ) (1) Where w i is the weight assigned to variables v i. The detailed description and construction of the weights are available upon request. 4. Empirical Results on Corporate Governance, Performance and Valuations We rank all companies according to the G formula (1). The company with the highest G index is ranked number one, while the company with the lowest G is ranked as 788. The details of all the rankings of individual variables and the ranking according to the total score are available upon request. In theory, good corporate governance should be related to good corporate performance. A number of empirical studies on emerging markets have found that investors are willing to 10

11 pay a premium averaging 10% to 12% for good corporate governance. It is interesting to see whether better-governed Chinese companies, measured by our measure of corporate G index, are associated with good corporate performance and whether Chinese investors reward good corporate governance. 4.A. Regression Results As mentioned before, we measure corporate performance by using two popular measures, ROA and ROE. Market value/book value of net asset of a company is used as a measure of investor s valuation of the company. Finally, the turnover ratio and stock return volatility is used to gauge the impact of corporate governance on market performance of the company. Table II reports the regression results of ROA, ROE, Market/Book, turnover, and standard deviation measures on the corporate rank. First, the two profitability measures ROA and ROE are both negative, with the significance level of 1%, related to the corporate governance ranking. The negative coefficient in the regressions indicate that highly ranked companies are indeed associated with higher corporate performance as measured by ROA and ROE. In the regression of Market/book value on the ranking index, we report a significant (at 1% level) and negative coefficient. This result strongly suggests that the investors do value companies with good governance. In the last two regressions, we report the impact of governance on market performance of a company. In the turnover regression, the coefficient for the ranking index is positive and significant at 5% level. The lower turnover for the better-governed companies suggests that investors tend to hold stocks of better-governed companies and trade more on companies with weaker corporate governance. Finally, in the last regression, we report the relationship between the volatility of stock returns with that of governance ranking. The relationship seems to be positive but insignificant. Overall, the results from Table II suggest that good corporate governance is associated with better corporate performance, higher valuation, and lower turnover of the stock market. Table III reports regression results of corporate performance and stock valuation variables on all 12 variables used in forming the ranking. The purpose of running these regressions is to check if our theoretical prediction about the effect of the 12 variables on corporate governance is indeed valid. Of course, running regressions with so many regressors would encounter the problem of multicolinearity. However, for illustrative purposes, we still conduct such regressions but interpret the result with special caution. We interpret our results as follows. The first exercise regresses ROA on all 12 variables used in our ranking analysis. In addition, we add one variable, top1_sq, which is the square of top1, to reflect our concern on the nonlinear effect of the shareholding of largest shareholder on corporate governance. Among the variables that significantly affect ROA, Audit, top5, d_e, and acc_index have the expected signs, while ct_num has the opposite sign as expected. In the second exercise on regression of ROE, among the statistically significant coefficients, Audit, lawsuits, top5, and d_e have the expected signs while the sign for ct_num is the opposite of what we expected. It should not be surprising that ct_num has the positive sign in both regressions as connected transactions are mostly used to shore up profitability of the listed companies. Both regressions of profitability measures have relatively high R-sqaures, 0.25 and 0.32 respectively. The third exercise regresses the variable market/book on the same set of explanatory variables. Among the statistically significant coefficients, the variables top1, top1_sq, 11

12 cstr2_10, out_in_ratio, and d_e have the expected signs while Audit has the opposite sign. The R-square, at 0.27, is again very high. The fourth regression regresses Turnover on these same variables. ct_num has a marginal positive impact on stock turnover ratio while top5 has the negative coefficient. The positive coefficient associated with connected transaction variable may reflect active market trading over connected transaction news. The R-square for this regression, however, is quite small. Finally, in the fifth regression, we regress the volatility variable, std, on these explanatory variables. Only Audit has a positive and significant coefficient. The R-square is again quite small. Overall, the results from Table III suggest that our corporate governance variables as a whole have significant relationship with the corporate performance and stock market valuation variables. 4.B. Correlation Results Based on Ranking Groups In the following analysis, we divide all sample firms into quintiles according to our governance-ranking index (G). Group 1 indicates the lowest corporate governance ranking while Group 5 indicates the highest. In each group we calculate the mean, standard deviation and other summary statistics of each corporate performance and stock valuation measures in each group. Panel A of Table IV reports the summary statistics of ROA for each group. It is shown that the lowest ranked companies have an average return to asset of 1.47%, while the highest is 4.92%. Clearly, ROAs rises as corporate governance index rises. Companies in the highest group have an average of ROA more than three times higher than those in the lowest group. Panel B tests whether these differences in ROAs for each group are statistically significant. Group 1 companies has statistically significantly different ROAs from their counterparts in groups of 3, 4, and 5. Similarly, the mean ROA of Group 5 is significantly different from those of other groups. Figure I compares the mean of ROA for each group. It is apparent that there is an increasing trend in the ROAs of companies from Group 1 to Group 5. Once again, it illustrates the proposition that better-governed companies tend to have higher profitability. Table V reports summary statistics and tests for another measure of corporate profitability, ROE. Patterns as shown in Table V are similar to those observed in Table IV, with the former having even more striking magnitude. For example, the best-ranked companies now have an average of ROE of 1.49%, more than 7 times higher than the lowest-ranked companies, (0.2%). The statistics presented in Panel B also suggest that these discrepancies in ROEs for each group are statistically significant. In addition, Figure II plots the increasing ROEs as company s governance ranking index improves. Table VI reports summary statistics and tests for the variable Market/book. The average market/book value of net asset is 4.84 for the best-ranked companies while it is 2.73 for the lowest group. In other words, there is a market premium of 77% for the best-governed companies relative to the worst companies. The differences among the five groups are again statistically significant as indicated by test results shown in Panel B. Figure III plots the mean 12

13 values of Market/book for the five groups. It is clear that better-governed companies are rewarded by investors by commanding high market value relative to the book value. Table VII presents summary statistics and tests for the market turnover ratio for the five different groups of the companies. There is a slightly higher turnover ratio for the lowest group. Group 2 has significant difference in turnover from Groups 3,4, and 5. The two lowest groups, 1 and 2, also have statistically higher turnover than the higher groups. Figure IV graphically supports the conclusions of Table VII. Finally, in Table VIII, we present the summary statistics of the volatility measure, the normalized standard deviation std for each group of companies. There seems to be a slightly higher volatility in the lowest group than the highest group, but the statistical level is insignificant. Figure V supports those conclusions. Overall, the results from the analysis of different ranking groups of companies suggest that good corporate governance matters. It improves the performance of the company, reduces the market turnover, and investors seem to reward these companies by paying market premium as high as 77%. 5. Conclusion The paper has studied an important issue of corporate governance for China s publicly listed companies. The main findings of the study can be summarized as the following: (1) We identify several important corporate governance mechanisms stemming from the agency theory and the more recent - tunneling - theory of corporate governance. Among others, the ownership and the board structure, executive compensation, the market for corporate control, and the financial transparency are found to be the most important factors in influencing corporate governance. (2) Based on our theory and understanding of China s capital market, we construct four different set of variables to explain corporate governance. The first set of variables is related to - tunneling - theory. Variables such as whether the company has a parent company, the number of connected transactions, and concentration of the largest shareholders are in this set. The second set is related to monitoring. It includes whether the CEO is the chairman or a vice chairman of the board of directors and the proportion of outside directors to inside directors. The third set is related to managerial incentives, including shareholdings by the top five officials of the firm and the dividend payout ratio. Finally, the fourth set includes several measures of financial transparency. The variables include auditor opinion of the company and the number of lawsuits the company is involved in. (3) We rank our sample companies according to the above four sets of variables. After assigning weights to these variables, we obtain an index called G index which is a score constructed with weights based on the above-mentioned four sets of variables. All sample firms are then divided into five nearly equal groups. (4) We explore the possible links among corporate governance, corporate performance 13

14 and stock valuations. We find that better-governed companies are associated with higher profitability as measured by ROA and ROE, higher stock market valuation as measured by the ratio of market value and book value of the net asset, and lower market turnover ratio. Key Reference: Berle, Adolf and Gardiner Means, 1932, The modern corporation and private property, New York: Macmillan. 14

15 TABLE I. Summary Statistics Panel I, Corporate Governance Variables Variable Obs Mean Std. Dev. Min Max parent audit top ceo_is_top_dir lawsuits ct_num cstr2_ e top out_in_ratio div acc_index d_e Panel II, Corporate Performance Variables Variable Obs Mean Std. Dev. Min Max roa roe Market/book Turnover Std TABLE II. Regression of ROA (return on asset), ROE (Return on equity), market/book value on rank and turnover. The table reports slop coefficients, t-statistics (in ()), and adjusted-r 2 s from regressions of ROA and ROE on rank variable. Symbols *, ** and *** represents significance level of 10%, 5% and 1% respectively. Dependent variable Coeff. For rank Intercept r-square Roa *** ***.0252 ( ) ( ) Roe *** ***.0148 ( ) ( ) Market/book *** ***.101 ( ) ( ) Turnover ** ***.0051 ( ) ( ) Std *** (0.04) (4.61) 15

16 TABLE III. Regression of ROA (return on asset), ROE (Return on equity) and market/book value, turnover and Std on all the index variables Independent Variables Parent Audit top1 top1_sq ceo_is_top_dir lawsuits ct_num cstr2_10 top5 out_in_ratio div acc_index d_e Intercept Adj-R 2 ROA ( ) *** ( ) ( ) ( ) ( ) ( ) ** ( ) ( ) *** ( ) ( ) ( ) * ( ) *** ( ) ( ) ROE Market/Book value Turnover Std ( ) *** ( ) ( ) ( ) ( ) *** ( ) * ( ) ( ) ** ( ) ( ) ( ) ( ) *** ( ) *** ( ) ( ) *** ( ) * ( ) * ( ) ( ) ( ) ( ) *** ( ) ( ) ** ( ) ( ) ( ) *** ( ) *** ( ) ( ) ( ) ( ) ( ) ( ) ( ) * ( ) ( ) *** ( ) ( ) ( ) ( ) ( ) *** ( ) ( ) *** ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. 16

17 TABLE IV. Summary statistics of ROA Panel A, Summary by ranked grade grade Obs Mean Std. Dev. Min Max total summary Panel B, test if the mean with grade=i is larger than that with grade=j (where I>J) T-stat. I=2 I=3 I=4 I=5 J= ** *** *** J= ** *** J= * J= * Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. Figure I. Mean of ROA for different grades Grade 17

18 TABLE V. Summary statistics of ROE (by Grade) Panel A, Summary by ranked grade grade Obs Mean Std. Dev. Min Max total summary Panel B, test if the mean with grade=i is larger than that with grade=j (where I>J) T-stat. I=2 I=3 I=4 I=5 J= ** ** *** J= * * ** J= J= Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. Figure II. Mean of ROE for different grades Grade 18

19 TABLE VI. Summary statistics of Market/book value ratio Panel A, Summary by ranked grade grade Obs Mean Std. Dev. Min Max total summary TABLE VI. Panel B, test if the mean with grade=i is larger than that with grade=j (where I>J) T-stat. I=2 I=3 I=4 I=5 J= *** *** *** *** J= ** *** J= *** *** J= *** Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. Figure III. Mean of Market/Book for different grades Grade 19

20 TABLE VII. Panel A, Summary statistics of Turnover rate. Panel A, Summary by ranked grade grade Obs Mean Std. Dev. Min Max Total summary Panel B, test if the mean with grade=i is larger than that with grade=j (where I>J) T-stat. I=2 I=3 I=4 I=5 J= J= * * * J= J= Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. t-test whether the mean with grade<=2 is larger than the mean of the remaining groups, t= = with p-value Figure IV. The mean of Turnover rate for different grades Grade 20

21 TABLE VIII. Summary statistics of normalized standard deviation Panel A, Summary by ranked grade* Grade Obs Mean Std. Dev. Min Max Total summary * We have dropped four extremely large outliers TABLE VIII. Panel B, test if the mean with grade=i is larger than that with grade=j (where I>J) T-stat. I=2 I=3 I=4 I=5 J= J= * J= J= Symbols *, ** and*** represents significance level of 10%, 5% and 1% respectively. t-test whether the mean with grade = I is larger than that with grade = J (where I>J) Figure V, normalized standard deviation Grade 21