AN INVESTIGATION OF THE INFORMATION CONTENT OF THE FINANCIAL POLICIES OF CHINA’S LISTED COMPANIES LI NING (MA, Nankai University) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF ECONOMICS NATIONAL UNIVERSITY OF SINGAPORE 2004 i ACKNOWLEDGEMENTS Being in the Ph.D. program in Economics and writing this thesis have been a very special experience in my life. I am very grateful to my supervisor, Associate Professor Lu Ding, who has not only provided valuable opinions and advices on my research but also shown extraordinary kindness and patience throughout this period. Special thanks have to be given to Dr John Sequeira who provided many useful comments and suggestions on the revisions of this thesis. The ceaseless love, encouragement, and support that I received from my family, especially my parents and my husband, are the sources of my spirit, without which, it is unimaginable for me to finish this thesis finally. I also wish to say thank you to people who helped me and to my friends who shared the pains and happiness with me. I will bear all of these in my mind forever. i Table of Contents Page i ii iv vi Acknowledgments Table of content Summary List of Tables Content of Thesis 1 Chapter 1.1 1.2 1.3 Introduction Background Objectives Methodology Chapter 2.1 2.1.1 2.1.2 2.1.3 2.2 2.3 Institutional Feature of China’s Stock Market Features of China’s Stock Market A Unique Feature of China’s Stock Market Comparison with the Insider-based Model Comparison with the Outsider-based Model Problems of China’s Stock Market Solution to the Problems 11 11 11 14 18 22 33 Chapter 3.1 3.1.1 3.1.2 3.2 3.2.1 3.2.2 3.2.3 3.3 3.3.1 3.3.2 3.3.3 Literature Review Cash Dividends Introduction Information Content of Cash Dividends Stock Distributions Introduction Accounting Treatment of Stock Dividends in China Information Content of Stock Dividends Rights Issues Introduction Information Content of Rights Issues Institutions of Rights Issues in China 39 39 39 43 47 47 50 53 55 55 58 65 Chapter 4.1 4.2 4.2.1 4.2.2 4.2.3 4.3 4.4 4.4.1 4.4.2 4.4.3 Methodology Data Hypotheses Cash Dividends Stock Distributions Rights Issues Event Study Regression Models Cash Dividends Stock Distributions Rights Issues 68 68 71 71 79 85 91 96 96 99 102 Chapter Findings and Results 104 ii 5.1 5.1.1 5.1.2 5.2 5.2.1 5.2.2 5.3 5.3.1 5.3.2 5.4 Chapter Tables Bibliography Appendix Cash Dividends Event Studies Regression Analysis Stock Distributions Event Studies Regression Analysis Rights Issues Event Studies Regression Analysis Summary 104 104 108 113 113 116 120 120 122 126 Conclusions 129 The OLS Model of Event Study 137 178 190 iii Summary The thesis attempts to find out the reasons that caused the deteriorating share price performance and the unusual financial behavior of China’ listed companies. By comparing the institutions of China’s stock market with those of the stylized markets in matured capitalist economies, it reveals that the chief problem of China’s stock market is its failure to protect the interests of public minority shareholders and this is caused by the share structure of the stock market where most shares are state-owned or state-controlled which are non-tradable. The possible solution to the problems of weak corporate governance without changing the share structure is to connect the firm’s performance with shareholders’ wealth by enforcing regular cash dividend payment. Based on data collected for financial events of China’s listed companies over 1994-2003, the thesis investigated the information content of cash dividends, stock dividends, transferred capital stocks, and rights issues using event study and regression analysis under the framework of information signaling. The empirical results did not support the main hypotheses implied by the information signaling theory, suggesting that these hypotheses may not be applicable to China’s case due to the institutional differences, which are discussed in light of our test results. A series of regulatory policies associating the cash dividend distributions with the equity issues were released since 2000. The empirical findings indicate that the new policies, iv associating the equity issues with cash dividends, have impacts on the managerial behavior to a certain extent. These findings have brought optimism to the effect of the cash-dividend-enforcement policy on improving corporate governance. This study has implications for the relationship between corporate governance and corporate finance. On one hand, the thesis provides evidence about the influence of corporate governance on corporate finance by investigating the information content of various financial policies in a stock market with unique institutions. On the other hand, it provides evidence about the how corporate finance can affect corporate governance by examining the impact of the cash-dividend-enforcement policy on the managerial behavior of listed companies. v List of Tables Page Table 1-1 Number of Listed Companies (1993-2004) 137 Table 1-2 Features of All A-share Firms (1993-2004) 137 Table 1-3 China’s Stock Market in the National Economy (1992-2004) 138 Table 2-1 Share Structure of All Listed Firms (1992-2004) 139 Table 2-2 Ownership Structure of Listed Companies (1993-2001) 139 Table 2-3 Ownership concentration in terms of a single shareholder 140 Table 2-4 Overall Performance of Listed Companies (1992-2001) 140 Table 2-5 Losing Firms in China’s Stock Market (1995-2003) 140 Table 2-6 Changes of Capital Structure of Listed Companies (19922002) A summary of A-share Firms Issuing New Shares (19942003 A summary of A-share Firms Paying Dividends (1994-2003) 141 Table 2-7 Table 2-8 Table 2-9 142 142 143 Table 2-10 Turnover Rates of China’s Stock Market and Major World Markets Number of A-share Investors (1990-2001) Table 4-1-1 Sample Description of Cash Dividend Events (1994-2003) 145 Table 4-1-2 145 Table 4-1-3 Sample Description of Pure Cash Dividend Events (19942003) Expected Results of Regression Models of Cash Dividends 146 Table 4-2-1 Sample Description of Stock Dividend Events (1994-2003) 147 Table 4-2-2 Sample Description of Pure Stock Dividend Events (19942003) Expected Results of Regression Models of Stock Distributions Sample Description of Transfer Capital Stocks Events (1994-2003) Sample Description of Pure Transferred capital stocks Events (1994-2003) Sample Description of Pure Rights Issues Events (19942003) Expected Results of Regression Models of Rights Issues 147 Table 4-2-3 Table 4-3-1 Table 4-3-2 Table 4-4-1 Table 4-4-2 Table 5-1-1 Table 5-1-2 Announcement Effect of Cash Dividends Events (19942003) Announcement Effect of Cash Dividends Changes (19942003) 144 148 149 149 150 151 152 153 vi Table 5-1-3 Announcement Effect of Cash Dividends Initiations and Resumptions (1994-2003) 154 Table 5-1-4 Announcement Effect of cash dividends by Year 155 Table 5-1-5 Announcement Effect of cash dividends by Period 156 Table 5-1-6 Statistical Description of Explanatory Variables in Regression Models of Cash Dividends 157 Table 5-1-7 Cross-sectional Regressions of Cumulative Abnormal Return of Cash Dividends 158 Table 5-1-8 Cross-sectional Regressions of Percentage Change in Cash Dividend Distribution Size 159 Table 5-1-9 160 Table 5-2-1 Cross-sectional Regressions of Cash Dividend Distribution Size Announcement Effect of Stock Dividends (1994-2003) Table 5-2-2 Announcement Effect of Stock Dividends by Year 162 Table 5-2-3 Announcement of Stock Dividends by Period 163 Table 5-2-4 Statistical Description of Explanatory Variables in Regression Models of Stock Distributions 164 Table 5-2-5 Cross-sectional Regressions of Cumulative Abnormal Return of Stock Dividends 165 Table 5-2-6 Cross-sectional Regressions of Distribution Size of Stock Dividends Announcement Effect of Transferred Capital Stocks (19942003) Announcement Effect of Transferred capital stocks by Year 166 Table 5-3-1 Table 5-3-2 Table 5-3-3 161 167 168 Announcement Effect of Transferred Capital Stocks by Period Statistical Description of Explanatory Variables in Regression Models of Transferred Capital Stocks 169 Table 5-3-5 Cross-sectional Regressions of Cumulative Abnormal Return of Transferred Capital Stocks 171 Table 5-3-6 Cross-sectional Regressions of Distribution Size of Transferred Capital Stocks 172 Table 5-4-1 Announcement Effect of Rights Issues (1994-2003) 173 Table 5-4-2 Announcement Effect of Rights Issues by Year 174 Table 5-4-3 Announcement Effect of Rights Issues by Period 175 Table 5-4-4 Statistical Description of Explanatory Variables in Regression Models of Rights Issues 176 Table 5-4-5 Cross-sectional Regressions of Cumulative Abnormal Return of Rights Issues 177 Table 5-3-4 170 vii CHAPTER INTRODUCTION 1.1 Background The rapid development of stock market is one of the major achievements that China has made in the past two decades. Two national securities exchanges Shanghai Securities Exchange (SHSE) and Shenzhen Securities Exchange (SZSE) were established in Shanghai and Shenzhen at the end of 1990 and 1991 respectively. At the beginning, there were only a few companies publicly listed and little amount of transactions in these two national exchanges, but China’s stock market grew rapidly. The number of listed companies increased from 183 in 1993 to 1377 in 2004 (see Table 1-1). The listed firms raised a large amount of funds from the equity market, the total market capitalization increased from 331.8 billion Yuan in 1993 to 3630.9 billion Yuan in 2004 and the total tradable market capitalization increased from 68.3 billion in 1993 to 1099.8 billion in 2004 (see Table 1-2). As a result, the stock market has played a very important role in China’s economy. According to Table 1-3, the market capitalization as a percentage of GDP has been increasing until it reached the peak in 2000 which is 53.79 percent, after that, it began to decrease and reached 27.14 percent in 2004. The stock market was established due to practical needs of improving state-owned enterprises. Although the enterprise reform, which began in mid 1980s, had made many trials to replace the old management system administered by bureaucrats with more effective management system, including the Director Responsibility System, and the Contract System, the efficiency of State Owned Enterprises (SOEs) was not greatly improved. The SOE sector which employed over 70% of the urban workforce still occupied a predominant status in the economy, but the overall performance of SOEs was unsatisfactory because over one thirds of SOEs were in chronic financial distress and being supported by the state budget grants and government-directed bank credits. It was getting worse because the state treasury and the state-owned banks themselves were facing financial difficulties, the former of which had a constantly decreasing income since the taxation reform, and the latter of which had already been implicated by the huge amount of bad debts. How to solve the financing problems and to improve the efficiency of SOE became urgent issues of Chinese government at the early 1990s. The development of stock market was regarded as a remedy for the above issues and was thus greatly supported by the Chinese government. Undoubtedly, by directing the huge amount of personal savings and even the foreign funds to the investment needs, stock market can solve the financing problem faced by many enterprises including both state-owned enterprises and many non-SOEs, and to relieve the burden of state budget and state-owned banking sector as well. Moreover, it was believed that stock market can function as a governance system that would help establish a truly competitive corporate sector in China as argued by some economists. Firstly, companies going public have to be transformed to shareholding companies. In this process, the company’s ownership would become clearer based on the shareholdings that the ownership was not clear was considered as a key factor leading to the inefficiency of SOEs; secondly, by issuing shares to the public, the ownership of shares will be diversified. It helps improve the performance of SOEs and the public shareholders can directly or indirectly participate in the corporate management through the stock market. 0.17 percent, and significant at the level of 0.1 percent and percent respectively. The cumulative abnormal return of days [0, +1] is also negative at -0.44 percent and significant at the level of 0.1 percent. The cumulative abnormal return of days [-1, +1] and [-2, +2] are also significantly negative. The overall effect of the announcement of rights issues are negative means that the shareholders not favor the announcement of rights issues. This result seems to be consistent with the findings of the US but different from the finding in many European and Pacific-basin countries. However, positive announcement effect of rights issues has been reported in some existing literature about China’s case, in order to interpret whether the mixed findings are due to the difference in the sample period, the announcement effect of rights issues for each year is then examined and presented in Table 5-4-2. It is found that the negative price effects towards the announcement of rights issues are not stable across the years. The abnormal return of day -1 is significantly positive for years 1994 and 1997, and the cumulative abnormal returns of days [-1, 0] for year 1994 and year 1997 are both positive and significantly at the level of 0.1 percent. The significant negative abnormal returns of the announcement day are found for years 2001-2003, which are -1.16 percent, -0.80 percent and 1.74 percent respectively. The implication is that, even if the same methodology is applied, the results may be inconsistent due to the difference in the sample period. It has been explained that the change from 2000 to 2001 may be due to the new policies regarding the equity issues. While many firms take the last opportunity to issue rights in 2000, more companies began to pay cash dividends in 2001 and less companies issued new equities through rights. In order to further test whether there is any difference in the market reaction before 2000 and after, the announcement effect 121 is tested by dividing the whole sample into two periods. The result of the announcement effect for each period is presented in Table 5-4-3. It shows that, for the period 1994-1999, the abnormal return of day -1 is positive at the value of 0.36 percent which is significant at the level of 0.1 percent, and the cumulative abnormal return of days [-1, 0] is also positive at 0.33 percent which is significant at the level of percent. In contrast, for the period 2000-2003, the abnormal returns are significantly negative for days -1 and are at the values of -0.24 percent, -0.64 percent respectively, and the cumulative abnormal returns ranging from -0.75 percent to -1.36 percent for the tested event windows within [-2, +2] are all significant at the level of 0.1 percent. It indicates that the market reacts more unfavorably during 2000-2003 than during 1994-1999. 5.3.2 Regression Analysis The statistical description of explanatory variables in the model (4.3) is presented in Table 5-4-4. The results of cross-sectional regression model (4.3) for rights issues are shown in Table 5-4-5. The dependent variable is CAR2 which is the cumulative abnormal return over days [0, +1]. The CAR2 is firstly regressed against all the explanatory variables in model (4.3), and the result is shown in the first column of the table. It shows that except for RUNUP and MRUNUP, all the other coefficients are not significant. The coefficients for OFFERSIZE, OFFERPRICE, RUNUP and MTB are negative whereas the coefficients for MRUNUP and SDVR are positive. Since OFFERSIZE and OFFERPRICE are highly correlated as both are the proxies for the information signal, so either OFFERSIZE or OFFERPRICE is dropped out of the model. The result of OLS regression without OFFERSIZE is shown in column of the table. Compared with the result in column 1, the coefficient of OFFERPRICE 122 becomes significantly negative. The regression result without OFFERPRICE in the model is shown in column 3. Compared with the result in columns of the table, the coefficient of OFFERSIZE becomes significantly negative. It seems that both OFFERSIZE and OFFERPRICE can be used as the proxy of information signal, although the adjusted R-square suggests that the model without OFFERPRICE is better than the model without OFFERSIZE. Therefore, both Hypothesis and Hypothesis are supported. While the negative relationship between the announcement effect and the offer size can be explained by the agency theory of Jensen, the signaling theory of Leland and Pyle, or the leverage hypothesis, it may have other implications for China’s investors. Because their investment return has to rely on the capital gains instead of cash dividends, China’s investors will try very hard to find any chance to capture the capital gains. The offer size is defined as the dollar value of new equity raised under the rights issue divided by the market value of the firm 10 day prior to the announcement day. The larger the value of offer size means the larger the shareholders need to pay to subscribe the additional shares relative the market value they own, the more difficult for the investors to capture the capital gains in the market. Therefore, the negative relationship can be attributed to the speculative motive of China’s investors instead of a signal of information asymmetry. The negative relationship between the announcement effect and the relative offer price is inconsistent with theoretical prediction of Heinkel and Schwartz (1986) that higher quality firms use the subscription price in a rights issue as a signaling mechanism to distinguish them from lower quality firms, but consistent with the model of Myers and Majluf (1984) stating that the firms resort to a rights issue with a 123 subscription price low enough to warrant a successful offering to prevent wealth transfers from old stockholders to new stockholders. For China’s case, the possibility of setting subscription price low enough to warrant a successful offering is not precluded, but the explanation provided by Myers and Majluf (1984) may not be enough. A higher subscription price relative to the market price means the price is less discounted, and therefore less chances for the investors to capture capital gains. The relative offer price may serve as an indication of speculative chances rather than an information signal sounds to be more applicable to the China’s case. The coefficient of RUNUP is consistently negative and significant at the level of percent in all the results shown in the Table 5-4-5. As a proxy for the preannouncement share price increase, RUNUP can be regarded as information leakage. Hence, Hypothesis is supported as the announcement excess return responds negatively to the cumulative stock return prior to the announcement. This can be explained by the argument of Masulis and Korwar (1986) that a higher preannouncement stock price increase may be viewed as an indication of firms not making a stock offerings and thus the stock offering announcement is less anticipated resulting a lower price effect. However, it is also likely that a higher stock return prior to the offer is viewed as a market reassessment of the firm’s value based on the information leakage, resulting in a less price effect towards the announcement of the offer. As expected, the coefficient sign of MRUNUP is consistently positive at the significance level of percent for all the results shown in Table 5-4-5. This is consistent with Hypothesis that the market reaction is more favorable to the announcement of rights issues when the market is in a favorable condition. This is 124 consistent with the study of Zhang and Han (1997) who found that the announcement day abnormal return in China was significantly positive in bull market and was significantly negative in bear market. Many findings suggest that the market reacts more favorably to the rights issue announcement of firms with greater investment opportunities. The rationale is that those firms with more investment opportunities are lack of free cash flow and tend to issue equity to finance more profitable projects. However, the view of this study is that this argument may not be valid in China because the investor are not so concern about the information related to the cash flow of the firms, and it is also uncertain whether the true incentive of firm’s issuing equity is the need to invest in more profitable projects. As expected, the coefficient of the proxy of the investment opportunities is negative and insignificant. Hypothesis that the market tends to react more favorably to the announcement of firms with greater investment opportunities is rejected. Hypothesis is not supported by the result because the coefficient of SDVR is positive but insignificant. It is inconsistent with the prediction of Myers and Majluf (1984) and Merton (1987) that the relationship between the announcement excess return and the preannouncement return variance should be negative. The positive coefficient sign is consistent with the prediction made for China that, since China’s shareholders are speculators and risk lovers, they are more willing to invest in stocks with greater uncertainty and higher stock return variance. Although the coefficient is insignificant, the finding does not contradict with the hypothesis of speculative motive. In order to see that whether the new equity issue policy released by the CSRC has some impacts on the market reactions, the whole sample is divided into two sub- 125 samples and dummy variables are introduced. Dummy variable PD00 is equal to if the event is within the period 2000-2003, and zero otherwise. Dummy variable SIZE_00 is the multiplication of PD00 and OFFERSIZE. The result of model including PD00 is presented in column of the table. The coefficient of PD00 is negative but insignificant. By adding both PD00 and SIZE_00 into the model, the result in the column shows that the coefficient of both PD00 and SIZE_00 are not significant. Although the signs of the coefficients suggest that the market reaction is more unfavorably to and less dependent on the OFFERSIZE during period 2000-2003, but a significant structural change is not detected. The possible reason is, while it may have impact on the managerial incentives and thus on the firm’s performance sooner or later, the new policies may not have immediate effect on the investors’ speculative behavior. 5.4 Summary The findings about cash dividend events shows that the cash dividend announcement is on average positive although the price effect of any single day around the announcement is insignificant. Because the unexpected change of cash dividends is not significantly related to the change of expected future earnings, the information signaling hypothesis is not supported. But the results cannot reject the agency cost theory, as the level of cash dividend depends positively on the expected earnings. A structural change for cash dividend events around 2000 is detected for both the market behavior and the managerial behavior. The change of managerial behavior probably can be explained by the new policy of rights issues. Under the new policy, firms expecting to issue rights in the future have to pay cash dividends now. 126 Therefore, they would concern more about the current ability to pay the cash dividend and consider less about the future performance. Event studies of stock dividends show that the market reaction towards the announcement of stock distributions is on average significantly positive and is much higher than that towards the cash dividend. Consistent with the information signaling hypothesis, the level of cash dividend can be regarded as an information signal, and the level of cash dividend does convey positive information about the current earnings and the future earnings. The dividend substitution theory and the cash increase theory are not supported by the results, which is probably because the investors in China not care much about the cash dividend due to the low dividend yield. The trading range hypothesis, which predicts that firms paying stock dividend due to the intention of reducing the share price within a reasonable range, cannot be rejected either. A structural break is detected for the market reaction which reacts more positively to the level of cash dividend before 2000 than after. Since 2000, the relationship between the level of cash dividend and the expected future earnings also changed from positive to negative. This is probably because firms that intend to pay stock dividends now would rely more on the current earnings than on the future expected earnings as the future expected earnings is more likely to be used to pay future cash dividends. The announcement effect of transferred capital stocks is also significantly positive suggesting the similarity between the stock dividend and the transferred capital stock. Similar to the case of stock dividends, the result is consistent with the trading range hypothesis. As expected, hypotheses suggested by the dividend substitution theory 127 and cash increase theory are consistently rejected, suggesting that the cash flow is not the main concern of China’s investors. The information signaling hypothesis is not supported by the whole sample although the level of transferred capital stock is positively related to the current earnings. The level of transferred capital stock can only be regarded as a signal conveying information of current earnings for the sample of 1994-1999. This result is apparently different from the results of stock dividend events and inconsistency may be explained by the retained earnings hypothesis. A structural change is also detected for the managerial behavior as the size of transferred capital on average becomes less during 2000-2003 than during 1994-1999 holding all others constant. The results of event studies show that on average the market reacts unfavorably to the announcement of rights issues. As expected, the magnitude of the market reaction is negatively related to the offer size and relative offer price, and positively related to the contemporary market return which is a proxy for the market condition. However, the hypothesis that the market reacts more favorably to the rights issue announcement of firms with greater investment opportunities is rejected. It is also found the price effect is positively but insignificantly related to the stock return variance prior to the offer. While the above findings can be explained by the information signaling hypothesis, the agency theory, or some other hypotheses, the speculative motive hypothesis cannot be rejected as one possible explanation, although further studies are necessary to substantiate this hypothesis. A structural change around 2000 is not detected for the market reaction which is probably because the new policies of rights issue not have immediate effect on the investors’ speculative behavior. 128 CHAPTER CONCLUSIONS The thesis attempts to find out the reasons that caused the deteriorating performance and the unusual financial behavior of China’s listed companies. The study starts with the discussion of the institutions of China’s stock market by comparing them with those of stylized markets. It reveals that the chief problem of China’s stock market is its failure to protect the interests of public minority shareholders. However, the cause of the problem is the dominance of state ownership structure in China’s stock market where most of the firms are either state-owned or state-controlled. With the intention to maintain the dominant status of state ownership in the stock market, T the shares are classified as non-tradable and tradable based on the ownership property, aggravating the complexity of the agency relationships of state-owned or statecontrolled firms. The direct consequence is, China’s stock market failed to be an effective market for corporate control as its US counterpart. Despite the similarities in many aspects, in China’s stock market the institutions that patterned after the US market could not serve the same function. Under this circumstance, the managers have the discretion to pursue their own interests or common interests with those state asset representatives at the cost of the public minority shareholders, leading to the deteriorating performance and many unusual market behaviors including the high volatility of share prices. The impacts on the corporate finance are that the firms frequently omitted cash dividends, to substitute cash dividends with stock dividends or transferred capital stocks, or to frequently issue new equities. 129 The thesis suggests one solution to the problem of weak corporate governance based on the analysis of the mechanism of the US market for corporate control. It is found that the US stock market relied on the share price serving as an indication of the firm’s value to realize effective governance. However, this condition that the change of share price reflects the change in the expectation about firm’s value was absent for China’s stock market. The public minority shareholders would not benefit more with the enhancement of firm’s value because firms not pay regular cash dividends based on their permanent earnings. From the viewpoint of this thesis, one solution to the improvement of corporate governance is to connecting the firm’s performance with shareholders’ wealth by enforcing regular cash dividend payment. This cannot help solve the radical problem caused by the institutions of ownership structure, but it is supposed to help improve the corporate governance to some extent. This is coincident with the ideas of China’s policy makers who started to enact a series of regulatory policy aiming at improving the corporate governance and protecting the public shareholders’ interests since 2000. The current institutions of stock market suggest that the firm’s management is likely to pursue their own interests instead of the shareholders’ interests, and the shareholders invest in such a risky market in the hope of capturing capital gains. In order to provide some empirical supports to the above argument, the thesis examines the financial polices of China’s listed companies under the framework of information signaling hypothesis. The main idea of the information signaling hypothesis is that, due to information asymmetry, financial instruments can be used by the management to convey inside information to outside investors. But information signaling theory was developed in countries with different institutions, and the assumptions implied in the information signaling hypothesis are (i) that the share price is changed with the 130 shareholders’ assessment about the firm’s value; and (ii) that the falling share price tends to oblige managers to maximize the shareholders’ interests. The institutional analysis suggests that the above two assumptions may not hold in China’s stock market, and thus the hypotheses related to the information signaling may not be able to explain the case of China. Based on data collected for financial events of China’s listed companies over 19942003, the thesis investigated the information content of cash dividends, stock dividends, transferred capital stocks, and rights issues using event study and regression analysis. The main conclusions are as follows. On average, the market reacts favorably to the announcement of cash dividends. However, compared with cash dividend announcement effect, the announcements of stock dividends and transferred stock dividends are much more welcomed by the market. The average announcement effect of rights issues is negative. The announcement of each type of events is found to be instable across the years over 1994-2003, suggesting that the selection of different sample periods is probably one reason of the mixed findings of existing literatures other than the methodologies applied. The information signaling hypothesis is not supported by the study of cash dividend events because the market reaction is not related to the unexpected change of cash dividends, although the unexpected change of cash dividends is positively related to the unexpected change of future earnings. The agency cost theory is weakly supported as the announcement effect is positively related to the cash dividend per share, and cash dividend per share depends positively on the current earnings, expected future earnings, and the expected free cash flow but the explanatory power of the cash 131 dividend per share on the announcement effect is very small. The event studies also contradict with the information signaling hypothesis in that the market reaction of cash dividend initiations and resumptions are not significantly higher than the market reaction of cash dividend increases. The study of stock dividend events supports the information signaling hypothesis because the market reaction is positively related to the stock dividend per share, and the stock dividend per share positively depends on the current earnings and expected future earnings. The information signaling hypothesis is not supported by the events of transferred capital stocks during 1994-2000 because the announcement effect is positively related to the size of transferred capital stock, and the size of transferred capital stock positively depends on the current earnings and negatively depends on the expected future earnings. The different findings about the events of stock dividends and transferred capital stocks are consistent with the retained earnings hypothesis. It is uncertain whether the information signaling hypothesis is supported by the study of rights issues. The price effect is negatively related to the relative offer size and the offer price suggesting the relative offer size and offer price may serve as signals of information. However, considering the China’s institutions, these findings can also be explained by the speculative motive of public shareholders. The price effect is negatively related to the pre-announcement stock return and positively related to the pre-announcement market return. All the findings are not contradicting with the hypothesis of speculative motive. Due to the time constraints, it is not possible to test all the hypotheses related to the rights issue, and the managerial incentives of rights issues is a direction for the further research. 132 A series of regulatory policies about the equity issues, associating the equity issues with cash dividend payment, were released since 2000. The tests of the structural stability of the market reaction and managerial behavior were conducted and the hypothesis of a structural break cannot be rejected. For example, the announcement effect of cash dividends is less positively related to the level of cash dividend since 2000; the level of cash dividend is less positively related to the current earnings and expected cash flow; the size of cash dividend is positively related to the expected future cash flow over 1994-1999 but negatively related to the expected cash flow over 2000-2003; and the size of transferred capital stock is only positively related to the sample over 1994-1999 but not to the sample over 2000-2003. The above findings suggest that the new policies have impacts on the managerial behavior to certain extent. Since 2000, more firms started to paid cash dividends, but firms paying stock dividends or transferred capital stocks and those issuing rights became much less. The change of managerial incentives can be explained either by the motive to issue equities in the future or by the motive to pay cash dividends in the future. For example, with the motive to issue equities in the future, more firms began to pay cash dividends, but the level of cash dividend depended less on the expected future cash flow and more on the current cash flow; with the motive to pay cash dividends in the future, less firms paid stock dividend and the size of cash dividend depended less on the expected future earnings but more on the current earnings. At its early stage, it is difficult to tell whether the cash-dividend-enforcement policy is effective to improve the corporate governance as the investors’ behavior could not be changed immediately. But it is optimistic to expect that this policy will work in the future. If all the firms are forced to pay cash dividends regularly based on their 133 permanent earnings, and the shareholders’ wealth is associated with the firm’s performance, eventually shareholders will assess the stock value based on the information related to the firm’s value and exert some control power on the corporate management. As a part of the reform on the stock market, the above equity issuing policy gives one example of how the corporate finance can affect the corporate governance. However, there are many other policies which could have direct impacts on the corporate governance of listed companies. It was noticed that a series of regulations aiming at improving the corporate governance have been promulgated after 2000: the CSRC released the “Directives on the Establishment of the Institution of Independent Board of Directors” in August 2001; it also issued the “Code of Corporate Governance for Listed Companies in China” in January 2002, which specified the compositions of corporate governance. In October 2005, the China’s Company Law and the Securities Law were revised for the fourth time since 1997 by incorporating many institutions of corporate governance which was effective since January 2006. Furthermore, the “Regulations on Shareholders’ Meeting of Listed Companies” and the revised “Guidelines on the Statute of Listed Companies” were promulgated in March 2006 with immediate effect. Since 1997, the “Guidelines on the Statute of Listed Companies” was revised for the first time based on the new editions of the Company Law and the Securities Law. The main amendments were reflected in the sections of Shares, Shareholders and Shareholders’ Meeting, and the Board of Directors. The above new regulations gave explicit directives about the corporate governance and would certainly have positive effects on the firms’ performance. However, as the above regulations not change the dominance of state ownership structure in 134 China’s stock market, the problem of weak corporate governance cannot be solved completely. The policy makers of China’s stock market also realized that the split share structure34 had become an obstacle to the further development of stock market. The attempt of share structure reform started in June 2001 when the State Council issued the “Temporary Management Measures of Reducing State-owned Shares and Raising Social Security Fund”. It finally failed one year after in June 2002 due to the objection of tradable-share shareholders whose interests would be sacrificed by this reform. The State Council, in its “Guidelines on Promoting Reform, Opening-up and Steady Development of China's Capital Market” issued in February 2004, specified that the share structure reform must be conducted under the principle of protecting the legal rights of investors especially the public investors. The share structure reform restarted in April 2005 whose objective is to eliminate the non-tradable shares of listed companies by converting non-tradable shares into tradable shares through negotiations between non-tradable share shareholders and tradable share shareholders 35 . The issuance of the “Management Measures of Share Structure Reform for Listed Companies” in September 2005 symbolized the pilot reform on a small portion of listed companies had been finished and the share structure reform had entered into a steady propulsion phase. As suggested by this thesis, the share structure reform would have significant impacts on the corporate governance of listed companies. This is certainly a progress of China’s stock market, despite my doubts on whether the state share shareholders would be willing to give up their dominance in the listed companies because the high concentration of state-owned or state-controlled shares, even if they are tradable, 34 Split share structure refers to the classification of shares as tradable and non-tradable shares. There are certain limitations on the timing and the proportion of shares to be sold within certain period. For example, the shares, which are former non-tradable shares, are not allowed to sell in the stock exchange within 12 months since the reform scheme is carried out. 35 135 would dampen the effects of the reform. However, at least it can be sure that, once all non-tradable shares are converted into tradable shares, the “voting by feet”, the takeover, the incentive reward scheme for management, and many other governance devices can make real sense in aligning the interests of shareholders with the interests of the management. Since the share price will indicate the value of the whole company instead of only the tradable shares of the company, the change of share price will reflect the expectations of all the shareholders about the firm’s prospect, and will adjust the rewards of the management accordingly. The takeover would also be a credible threat to the firm’s management where all the shares can be traded in the stock exchange. It is expected that, the share structure reform, combined with those new institutions of corporate governance, would have great effects on the performance of listed companies and the development of stock market. As the focus of this thesis is on the relationship between the corporate finance and corporate governance, the discussions about the effects of the share structure reform and other institutions on the listed companies were not covered. It is interesting to find that the overall performance of China’s listed companies has been in improving since 2001 in terms of debt ratio, net asset per share, return on equity, and earnings per share 36 . That whether the improvement of the firm performance is due to the new policies of corporate governance enacted can be a topic for the further studies. 36 See Table 2-4. 136 [...]... companies are organized around a major bank and form a network group called “keiretsu.” There are long standing business relationships between the group companies The banks and other financial institutions own shares in most of the group companies and those companies may in turn hold the bank’s shares or each other’ shares The power in Keiretsu is split among the main bank, the largest companies, and... towards the announcement of specific financial decisions made by the listed companies Regression models are then used to investigate the possible factors that cause the announcement effect of a specific event and the managerial incentives There are some features of the empirical tests of this thesis in comparison to those in the literature of the information content theory Firstly, the aim of the thesis... between the directors and managers as both are in nature the ‘agents’ of the state Moreover, the directors and the managers are not only the “agents” of the state; they are at the same time the “agents” of the minority shareholders of the company The outside shareholders cannot exert any direct control on the corporation, and have to rely on the directors and managers to manage their wealth The above... in addition to the information asymmetry, there also exists more complicated agency relationship in China’s listed companies The second objective is to verify whether the behavior of the shareholders and management in China’s listed companies are consistent with the predictions of conventional information signaling theory The investigation of the information content of the financial policies is conducted... membership The supervisory board supervises the activities, and can scrutinize and veto the management’s investment plans, appoints or removes members of the executive board Members of the executive board cannot sit on the supervisory board Board members are required to represent the interest of the company, not of the group they represent The advantage of Germany-Japan model is that, the banks and other... manage the state-owned companies And certainly these directors on behalf of the state will nominate some managers to undertake the routine management of the companies 22 But these directors of the state-owned companies are the actual “agents” of the state in charge of the state-owned shares and state-owned legal person’s shares Therefore, in a state-owned company, there is not much difference in the. .. shareholders, they may also pursue their common interests at the cost of minor shareholders Therefore, the objectives of this thesis are as follows It is hypothesized that the financial behavior and the overall performance of China’s listed companies are the outcome of the institutional arrangements, which shape the relationship between shareholders and managers, and thus shape the behavior of shareholders and... find out the nature of the information that can be used to interpret the market reactions and the managerial incentives, no matter whether it is suggested by the information signaling theory or any other theories Secondly, the thesis concerns whether the market reaction towards and the managerial incentives of a specific policy are consistent The empirical studies of information content in the literature... sample companies, an individual shareholder holds more than 30 percent of the total shares of a firm The degree of ownership concentration is comparable to that of the Germany and Japan Although the ownership structure is highly concentrated, the governance system of the China’s public companies was not similar to that of the Germany-Japan model which is bank-centered Banks in China cannot control the. .. is another feature of Germany-Japan model The large shares of a company are often held by other companies – a cross-holding of shares – or by holding companies for families In smaller German companies, the family exerts control through majority ownership, in which the owner controls 51% of a company, which in turn controls 51 percent of its subsidiaries and so on (Franks and Mayer, 2001) In Japan, many . AN INVESTIGATION OF THE INFORMATION CONTENT OF THE FINANCIAL POLICIES OF CHINA’S LISTED COMPANIES LI NING (MA, Nankai University) A THESIS SUBMITTED FOR THE DEGREE OF. listed companies are consistent with the predictions of conventional information signaling theory. The investigation of the information content of the financial policies is conducted through the. supposed that these policies would have some impacts on the financial policies of the listed companies. Therefore, the third objective of this thesis is to confirm whether the new policies have