A STUDY OF OWNERSHIP CONCENTRATION, CONTROL AND EVOLUTION IN CHINESE IPOS

A STUDY OF OWNERSHIP CONCENTRATION, CONTROL AND EVOLUTION IN CHINESE IPOS Grace Jinying Hu Manchester School of Management UMIST PO Box 88 Manchester M60 1QD United Kingdom Phone: (44)161-2003529 Fax : (44)161-2003505 Email: mcyiijh2@stud.umist.ac.uk Marc Goergen Manchester School of Management UMIST PO Box 88 Manchester M60 1QD United Kingdom Phone: (44)161-2003456 Fax : (44)161-2003505 Email: mcyssmgg@umist.ac.uk ABSTRACT Chinese IPO companies that have a mixed ownership structure and the state still keeps majority control over them have been commonly recognized as their typical features by researchers worldwide. This paper gives a cross-sectional analysis on the way of concentration between different groups of shareholders and the evolution of state and private control for the 129 IPO companies that went public in 1993. First, we find that smaller and modern industrial companies are concentrated with indirect state shares (legal persons and employee shares) and A shares (domestic private investors) and are more widely held. Conversely, larger and traditional industrial companies are heavily concentrated with direct state shares. Second, smaller, growing and high risk companies has a greater reduction of state control over the 7 year period. By contrast, the state retains majority control in larger and strategically important companies even 7 years after their flotation. Lastly, the concentration of foreign shares displays the fact that the Chinese government has instituted favourable policies to support the development of larger companies and has given priority to them to acquire foreign fundings in the overseas market. Key words: Corporate governance, Ownership concentration, Corporate Control, Evolution

INTRODUCTION As Shleifer and Vishney (1997) define, corporate governance deals with the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment. Further, they also suggest that state ownership as an example of concentrated control with no cash flow rights and socially harmful objectives is ineffective. Privatization, a common response created as a result of the recognition of the enormous inefficiency of state firms coupled with the pressures on public budgets, creates concentrated private cash flow ownership to go along with control and brings more prominent ownership structures, thereby helping improve the performance of privatized firms. China, like the former communist country, Russia, undertook a series of economic reforms during the past two decades, gradually moving towards market economy from the central planning economy. One of the core tasks in the economic reforms is the reform of the vast state-owned companies that have made up a large part of the country s industrial sector. Two approaches have been adopted, the market approach and the ownership approach in the reform(xu and Wang, 1999). Unlike Russia and some Eastern European countries that focus on radical changes in corporate structures, China places emphasis on the market oriented approach. The development of the stock markets has provided an opportunity for larger and better-performing firms with good prospects acquiring funds from the general public. While allowing firms to go public, the state in total still ultimately holds more than half of the firms shares. As reported by the China Securities Regulatory Commission (CSRC), the average shareholding by the state, also called untraded shares in all listed companies at the end of 1998 is 68.11%. In other words, those that circulate and are traded on the stock exchanges are only accounted for 34.89% (CSRC, 2000). Due to the fact that the state, state-owned companies themselves and other institutions respectively hold a significant proportion of an IPO company s shares, the company has a complex ownership structure after their flotation. Usually, it consists of five major groups of shareholders - the state (shares owned by the government, SGV ), the legal persons (shares owned by institutions and the founder, SLP ), the employees (shares owned by employees, SEM ), domestic investors (A shares,

ASHR ) and foreign investors ( B and H shares, FORNSH ) (Xu and Wang 1999, Qi,et al 2000 and CSRC 2000). Among these five categories of shareholders, the state, the legal persons and domestic investors are the largest, each of them holding nearly one-third of a company s total shares. This brings us to another issue in the corporate governance landscape, that is, whether the privatization of the state-owned companies that have resulted in ownership concentration by different large shareholders is effective in terms of monitoring and management. As mentioned above, state-ownership as a form of control is inefficient and lacks in corporate governance. As Shleifer and Vishney (1997) suggest, privatization, which diversifies the former state-concentrated ownership, may help to improve corporate value. Hence, the question is how the situation is in China: has the partial privatization of state-owned companies and the change in corporate structures had impacts on companies performance. The rest of the paper is organized as follows. Section I describes the way of ownership concentration among different industry and size categories and Section II presents the evolution of state and private control 7 years after flotation. In Section III, we briefly study some exceptional cases regarding the state control issues in those companies that are either with extremely high or low state control in the year of IPO in 1993. Lastly, in Section IV, we summarize our results. Section I - Ownership Concentration To examine the concentration of different classes of ownership in Chinese IPO companies within different industry and size groups, we have classified our sample companies that went public in 1993 into 9 industry sectors and 5 size groups. Since the fraction of employee shares is small (accounted for only 1%) and are collectively owned by the company (in other words, indirectly owned by the state), instead of examining each category of shares separately, we classify the total state-owned shares into 2 broad types, namely, the direct state-owned shares( SGV ) and the indirect state-owned shares ( SLP and SEM ). The issues related to employee shares are examined in the next section. In short, the ownership groups that we are going to examine are shares directly owned by the state (SGV), shares indirectly owned by the

state (SLP and SEM), A shares (shares issued only to domestic investors) and foreign shares (shares issued to foreign investors). To survey the concentration of these 4 classes of ownership between different sectors and size groups, we undertake a parametric test, namely, the t-test on the mean of each ownership category within each industry and size group across the 7-year period. We compare our sub-sample mean (X i ) with population mean (µ o ) to estimate the way of concentration of different ownership types. In addition, since not all the companies include the full categories of ownership, i.e., some may have SGV, A-shares and foreign shares while others may have SLP and A shares, we also carry out a t-test between each sub-sample mean and zero. A non-parametric test, namely the signed rank sum test for a mean, is taken due to the fact that the number of companies that are in each industry and size category is small (less than 30) and some of the means are substantially different from the median. (The results for the signed rank sum test are shown in the appendix I.) Table 1 below presents information regarding the direct state ownership. The average percentage of the direct state ownership for all of our sample companies, i.e., the µ o is 37.4%. As displayed, the state holds a higher proportion of shares in companies manufacturing machinery and equipment ( I2 ). The average direct shares held by the state in I2 is 51%. This figure is significantly different from the average of the whole sample, the 37.4%. Similarly, the direct state shareholding is significant at the 1% level in the largest quintile as compared to the population mean. The high fraction of direct state ownership in the I2 and the largest quintile reflects that the larger the company, the greater the concentration of the state shareholding. Conversely, the state owns relatively less shares in companies manufacturing electric and electronic products. Average shareholding is 10% significantly lower than the population average.

Table 1 Descriptive Statistics on the Direct State Ownership ( SGV ) Concentration for the 129 sample companies The first figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 37.4%); The second figure in the bracket under the mean is T-statistic between sub-sample mean (X I ) and zero (0); Variable Description Number Mean Std. Dev. Median Minimum Maximum sub-sample Panel A: Industrial Sectors CI Commercial Companies 16 0.32 0.22 0.35 0 0.75 (-0.93) (5.81***) DI Diversified 13 0.35 0.34 0.10 0 0.86 Industrial companies (-0.27) (3.96***) I1 Electrical and Electronic Products 17 0.27 (-1.59*) (4.41***) I2 Machinery and Equipment 16 0.51 (4.10***) (15.19***) I3 Food, Beverage and Pharmacy I4 Motor Vehicle and steel products I5 Chemical and Textile Products 11 0.41 (0.52) (5.38***) 15 0.35 (-0.36) (5.02***) 15 0.36 (-0.26) (5.95***) PU Public Utilities 13 0.38 (0.12) (5.41***) RE Real Estate 13 0.43 (0.75) (5.93***) Panel B: Size Quintiles Lowest Lowest Quintile 26 0.29 (-1.79**) (5.77***) 2 nd Quintile 2 26 0.36 (-0.28) (8.25***) 3 rd Quintile 3 26 0.33 (-0.80) (6.66***) 4 th Quintile 4 26 0.38 (0.12) (6.93***) Highest Highest Quintile 26 0.51 (3.24***) (11.89***) * 10% significant level ** 5% significant level *** 1% significant level 0.26 0.30 0 0.87 0.13 0.49 0.30 0.83 0.26 0.44 0 0.79 0.27 0.41 0 0.86 0.23 0.39 0 0.75 0.26 0.47 0 0.72 0.26 0.48 0 0.75 0.25 0.38 0 0.75 0.22 0.41 0 0.74 0.26 0.34 0 0.86 0.28 0.42 0 0.86 0.22 0.61 0 0.87

Table 2 below presents the descriptive statistics for the indirect state ownership concentration across industries and firm size. The indirect state ownership or socalled Legal Persons Shares ( SLP ) include shares owned by the founder, other institutions and employees. The average of the indirect state shares in all of our sample companies is 24.2% of the total share equity. Firms controlled by this type of shares are more similar to listed companies in more developed economies in the way they are formed, governed and managed (Qi, et.al., 2000). An average proportion which is significantly higher than the population average can be found in companies producing electrical and electronic appliances and in the lowest quintile. Sectors such as manufacturing of machinery and equipment, and real estate have a relative low fraction of indirect state ownership. Especially in the sector of heavy industry, the average of indirect state shareholding is 8%, severely lower than the population average. This is due to a higher fraction of direct state ownership in this sector. Regarding different size groups, the proportion of indirect state ownership is the highest in the lowest quintile and the lowest in the highest quintile. Ownership concentration is negatively related to firm size. This shows that the state has less direct involvement in smaller companies, which have more dispersed outside ownership.

Table 2 Descriptive Statistics on the Indirect State Ownership ( INDSO ) Concentration for the 129 sample companies The first figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 24.2%; The second figure in the bracket under the mean is T-statistic between sub-sample mean (X I ) and zero (0); Variable Description Number Mean Std. Dev. Median Minimum Maximum Sub-sample Panel A: Industrial Sectors CI Commercial Companies 16 0.29 0.25 0.19 0 0.86 (-0.82) (4.69***) DI Diversified 13 0.32 0.28 0.31 0 0.90 Industrial companies (-0.95) (4.04***) I1 Electrical and Electronic Products 17 0.35 (1.69*) (5.42***) I2 Machinery and Equipment 16 0.08 (-7.02***) (3.23***) I3 Food, Beverage and Pharmacy I4 Motor Vehicle and steel products I5 Chemical and Textile Products 11 0.21 (-0.70) (3.91***) 15 0.25 (0.19) (3.75***) 15 0.22 (-0.42) (3.57***) PU Public Utilities 13 0.29 (0.75) (4.80***) RE Real Estate 13 0.17 (-1.21) (2.91***) Panel B: Size Quintiles Low Lowest Quintile 26 0.34 (2.03**) (7.00***) 2 nd Quintile 2 26 0.24 (-0.01) (6.23***) 3 rd Quintile 3 26 0.26 (0.36) (5.15***) 4 th Quintile 4 26 0.24 (-0.03) (4.58***) Highest Highest Quintile 25 0.12 (-3.24***) (3.45***) * 10% significant level ** 5% significant level *** 1% significant level 0.27 0.29 0 0.86 0.09 0.05 0 0.46 0.17 0.19 0 0.75 0.26 0.06 0 0.75 0.23 0.11 0 0.76 0.22 0.21 0.05 0.75 0.21 0.06 0 0.77 0.25 0.26 0 0.77 0.20 0.20 0 0.77 0.26 0.15 0 0.85 0.27 0.10 0 0.86 0.18 0.05 0 0.90

Table 3 below presents the state of A share concentration within different sectors and size groups. A shares are shares owned by domestic individual investors in China. The average proportion of total A shares in our sample companies is 30.9%. However, each individual holds an extremely small fraction of shares in companies. As reported by Wang and Xu (1999), even in the top 10 largest shareholders in a company, a single domestic individual investor holds normally less than 0.5%. Compared with the proportion of shares held by the state and the legal persons, which hold the majority of the shares, 0.5% held by a single individual is negligible. In other words, A shares are widely dispersed among individual investors in the public and the likelihood that a single investor will be presented on the board is small. They have no incentive and no capability to monitor the firm operations and are mostly looking for short-term speculative profits. Therefore, the extent of A share concentration may imply the investors preference in particular industrial sectors and sector development. As shown, the proportion of A shares is significantly high in commercial companies as well as chemical and textile companies as compared to that of the average of the whole sample. The sectors that are on steady growth and profitable. Moreover, real estate sector is also with high proportion of A shareholders. Besides, the smaller the companies, the higher is the concentration of A shares. Most companies in the lowest and 2 nd quintiles are commercial, textile, chemical and real estate companies. This high concentration of domestic individual investors in the small-company groups that include sectors of commercial, textile, chemical and real estate may imply that the smaller the companies, the higher is the growth opportunity and therefore companies tend to be more widely held. On the other hand, it may also show the fact that the Chinese government provides less support to small firms and the transformation from state ownership to private owners is faster than that in large firms.

Table 3 Descriptive Statistics on the Domestic Individual Ownership ( ASHR ) Concentration the 129 sample companies The first figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 30.9%); The second figure in the bracket under the mean is T-statistic between sub-sample mean (X I ) and zero (0); Variable Description Number Mean Std. Dev. Median Minimum Maximum Sub-sample Panel A: Industrial Sectors CI Commercial Companies 16 0.35 0.13 0.37 0.09 0.62 (1.43*) (11.03***) DI Diversified 13 0.30 0.18 0.25 0.03 0.71 Industrial companies (-0.09) (6.06***) I1 Electrical and Electronic Products 17 0.31 (-0.11*) (9.80***) I2 Machinery and Equipment 16 0.26 (-1.30***) (6.26***) I3 Food, Beverage and Pharmacy I4 Motor Vehicle and steel products I5 Chemical and Textile Products 11 0.30 (-0.22) (7.40***) 15 0.28 (-0.60) (5.35***) 15 0.37 (1.64*) (9.63***) PU Public Utilities 13 0.26 (-1.27) (6.94***) RE Real Estate 13 0.35 (1.20) (10.35***) Panel B: Size Quintiles Low Lowest Quintile 26 0.37 (2.23**) (14.29***) 2 nd Quintile 2 26 0.35 (1.69*) (13.04***) 3 rd Quintile 3 26 0.35 (1.18) (11.00***) 4 th Quintile 4 26 0.26 (-1.80) (9.68***) Highest Highest Quintile 25 0.22 (-3.00***) (6.92***) * 10% significant level ** 5% significant level *** 1% significant level 0.13 0.29 0.07 0.61 0.16 0.28 0 0.61 0.13 0.29 0.08 0.55 0.20 0.28 0.05 0.62 0.15 0.39 0.08 0.61 0.14 0.20 0.08 0.50 0.12 0.34 0.13 0.58 0.13 0.37 0.11 0.61 0.14 0.34 0.08 0.62 0.16 0.36 0.05 0.71 0.14 0.25 0 0.53 0.16 0.15 0.03 0.60

Table 4 presents the situation regarding the concentration of foreign shares. Foreign shares consist of B shares, which are available only to foreign investors and are priced as well as traded in U.S. Dollars on the Shanghai Stock Exchange and in Hong Kong Dollars on the Shenzhen Stock Exchange. Foreign shares also include H shares, which are shares issued and traded on the Hong Kong Stock Exchange. Compared to the previous three categories of ownership, the proportion of foreign ownership is small and accounts on average for only 7.4% across all sample companies. Foreign ownership is concentrated in the sector of machinery and equipment and the proportion of foreign shares increases with the size of the company. In the lowest and the 2nd quintile, there was little foreign ownership in the companies. The 4 th and the highest quintile, in which are the foreign ownership concentrates. Commercial and diversified industrial companies consist of very few foreign investors. The way of foreign ownership concentration reflects different strategies developed by the Chinese government toward large and small companies, strategically important and less important industries. Large companies, which are in important industries, such as shipbuilding, steel making, home appliance, electronics and pharmaceutical industries, have priority to issue shares to foreign investors, absorb foreign funds and access to the overseas market. Moreover, those that involve in infrastructure and public sectors are also mostly likely to get approval for issuance of foreign shares. Such special regulations and issuing criteria regarding offering of foreign shares result in the concentration of foreign ownership in the sector of machinery and equipment and the larger quintiles that include companies in the important industries (as mentioned above).

Table 4 Descriptive Statistics on the Foreign Ownership ( FORNSH ) Concentration for the 129 sample companies The first figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 7.4%); The second figure in the bracket under the mean is T-statistic between sub-sample mean (X I ) and zero (0); Variable Description Number Mean Std. Dev. Median Minimum Maximum Sub-sample Panel A: Industrial Sectors CI Commercial Companies 16 0.03 0.11 0 0 0.47 (-1.56*) (1.03) DI Diversified 13 0.03 0.07 0 0 0.26 Industrial companies (-2.20**) (1.63*) I1 Electrical and Electronic Products 17 0.07 (-0.22) (2.40**)) I2 Machinery and Equipment 16 0.16 (1.91**) (3.64***) I3 Food, Beverage and Pharmacy I4 Motor Vehicle and steel products I5 Chemical and Textile Products 11 0.08 (0.16) (1.91**) 15 0.12 (1.22) (3.24***) 15 0.05 (-0.68) (1.73*) PU Public Utilities 13 0.07 (-0.14) (2.13**) RE Real Estate 13 0.05 (-0.73) (1.69*) Panel B: Size Quintiles Low Lowest Quintile 26 0.01 (-6.47***) (0.76) 2 nd Quintile 2 26 0.04 (-1.56*) (2.02**) 3 rd Quintile 3 26 0.06 (-0.53) (2.28**) 4 th Quintile 4 26 0.12 (1.55*) (4.12***) Highest Highest Quintile 25 0.15 (2.46**) (5.01***) * 10% significant level ** 5% significant level *** 1% significant level 0.12 0 0 0.35 0.17 0 0 0.45 0.14 0 0 0.39 0.14 0 0 0.34 0.12 0 0 0.39 0.12 0 0 0.34 0.11 0 0 0.50 0.05 0 0 0.35 0.11 0 0 0.41 0.13 0 0 0.47 0.15 0 0 0.50 0.15 0.12 0 0.45

Another interesting fact observed from our descriptive statistics is the result from the t-test between sub-sample mean and zero. As shown from the t-test, the averages for direct state ownership, indirect state ownership and domestic individual ownership are all remarkable away from zero across different industries and size groups. This reflects that most of the companies have these three kinds of shares. However, some of the sub-sample averages regarding foreign ownership are indifferent to zero. Typical cases can be found in sector such as commercial companies and the lowest quintile group. That is, the averages of foreign ownership in the sector of commercial companies and the lowest quintile group are equal to zero (please refer to Table 4). The indifference of the averages of foreign ownership to zero in the sector of commercial companies and the lowest quintile group shows that commercial companies and companies in the lowest quintile, which mostly involve in real estate and diversified industries are less likely to issue foreign shares. This again evidences our previous explanations, that is, large companies, especially those engage in strategically important industries and infrastructures are allowed to offer shares to foreign investors. Therefore, sectors such as machinery and equipment as well as the 4 th and largest quintile are concentrated with foreign ownership. To conclude this part, state ownership is concentrated in mature industries and larger, stable, low risk companies. Companies that engage in manufacturing of machinery and equipment or are in the highest quintile have concentrated direct state ownership as well as foreign ownership. This reflects that the state retains majority control in larger and important industrial companies and supports them to offer shares in the overseas markets. The concentration of indirect state ownership is close to that of private domestic ownership. Both of indirect state ownership and private ownership are concentrated in smaller and modern industrial companies. For example, indirect state ownership is concentrated in companies producing electrical and electronic appliances and private domestic ownership is in those producing chemical and textile products.

Section II - Corporate Control and Evolution This part presents the analysis on the state of control and evolution in Chinese IPO companies from 1993 to 1999 within different industry and size categories. Controlling a company is associated with the proportion of voting shares held by shareholders. In China, shares issued are all ordinary shares and one share gives one vote. There is no discrimination between shares held by different types of shareholders, such as the non-voting shares issued by German companies because the initial owner wants to stay control of the company after flotation and multi-voting shares issued by companies in Belgium. Therefore, in principle, the different categories of shareholders, namely, the government, the legal persons, A- and foreign shareholders have equal rights to vote in the shareholders meeting and have equal chance to be on the board of directors in a company. However, as previously mentioned, individual investors are hardly involved in monitoring the management of their companies in practice. To recapitulate, this problem is mainly due to three reasons. First, Chinese IPO companies are mainly former state-owned companies and even if they go public, the state still controls over 60% of the shares of a company on average. As shown in our sample, the mean of total state shares over the 7 years is 61.6%. Second, private investors are too diffused. Even in the top 10 largest shareholders in the company, a single investor only holds about 0.5% of shares (Xu and Wang 1999). This fraction of individual shareholding is too small to be capable of controlling the company. Third, the board is mainly occupied by the large shareholders, either the officers appointed by the state or from other state-owned institutions. Individuals with their small shareholding are unlikely to be on the board of directors and be involved in the decision making on the company s managerial matters. The significant proportion of shareholding by the state in total undermines the effectiveness of corporate governance. There is still no separation of ownership and control and there are also agency problems associated with widely held ownership. Briefly examined, there are three kinds of agency problems existed in Chinese IPO companies. They are the conflict of interests between large shareholders and creditors (mainly bankers), between large shareholders and small shareholders as well as

between shareholders and managers. Table 5 and 6 below show the respective information on state voting rights and private voting rights. Within different industry and size groups, the average of total state voting shares is extremely high in the public utility sector, in which companies are mainly set up by local government authorities. Regarding other sectors, the average of state-owned shares is indifferent to the population average of 61.6% and the right of votes is high enough to dominant the board. The average of private voting rights is too low to constitute a control power. The test on total mean of private voting rights is statistically equal to the assumed population mean (38.4%).

Table 5 Descriptive Statistics on the State Voting Rights for the 129 sample companies The figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 61.6%); Variable Description Number Mean Std. Dev. Median Minimum Maximum Sub-sample Industrial Sectors CI Commercial Companies 16 0.62 0.12 0.63 0.38 0.86 (-0.01) DI Diversified 13 0.66 0.15 0.72 0.29 0.93 Industrial companies (1.14) I1 Electrical and Electronic Products 17 0.63 0.14 0.65 0.36 0.93 (0.32) I2 Machinery and Equipment 16 0.59 0.12 0.60 0.39 0.91 (-0.93) I3 Food, Beverage and Pharmacy 11 0.62 0.15 0.62 0.28 0.89 (0.08) I4 Motor Vehicle and steel products 15 0.60 0.12 0.63 0.38 0.94 (-0.40) I5 Chemical and Textile Products 15 0.57 0.13 0.57 0.29 0.78 (-1.25) PU Public Utilities 13 0.67 0.11 0.64 0.50 0.90 (1.72*) RE Real Estate 13 0.60 0.15 0.65 0.20 0.83 (-0.42) Firm Size Low Lowest Quintile 26 0.63 0.13 0.62 0.39 0.89 (0.39) 2nd Quintile 2 26 0.60 0.12 0.62 0.38 0.88 (-0.56) 3rd Quintile 3 26 0.59 0.14 0.60 0.29 0.94 (-0.82) 4th Quintile 4 26 0.62 0.14 0.63 0.28 0.92 (0.19) Highest Highest Quintile 25 0.64 (0.80) 0.14 0.65 0.20 0.93 * 10% significant level ** 5% significant level *** 1% significant level

Table 6 Descriptive Statistics on Private Voting Rights for 129 sample companies The first figure in bracket under the mean is T-statistic between sub-sample mean (X I ) and full-sample mean (µ o, which is 38.4%) ; Variable Description Number Mean Std. Dev. Median Minimum Maximum Sub-sample Industrial Sectors CI Commercial Companies 16 0.38 0.12 0.37 0.14 0.62 (0.01) DI Diversified 13 0.34 0.15 0.28 0.07 0.71 Industrial companies (-1.14) I1 Electrical and Electronic 17 0.37 0.14 0.35 0.07 0.64 Products (-0.32) I2 Machinery and Equipment 16 0.41 0.12 0.40 0.09 0.61 (0.93) I3 Food, Beverage and Pharmacy 11 0.38 (-0.08) 0.15 0.38 0.11 0.72 I4 Motor Vehicle and steel 15 0.40 0.12 0.37 0.06 0.62 products (0.40) I5 Chemical and Textile Products 15 0.43 0.13 0.43 0.23 0.71 (1.25) PU Public Utilities 13 0.33 0.11 0.36 0.10 0.50 (-1.72*) RE Real Estate 13 0.40 0.15 0.35 0.17 0.80 (0.42) Firm size Low Lowest Quintile 26 0.37 0.13 0.38 0.11 0.61 (-0.39) 2 nd Quintile 2 26 0.40 0.12 0.38 0.12 0.62 (0.56) 3 rd Quintile 3 26 0.41 0.14 0.40 0.06 0.71 (0.82) 4th Quintile 4 26 0.38 0.14 0.37 0.08 0.72 (-0.19) Highest Highest Quintile 25 0.36 (-0.80) 0.14 0.35 0.07 0.80 * 10% significant level ** 5% significant level *** 1% significant level Even though the state still controls most IPO companies in general, there is a declining trend over the years. The state diminishes its shareholding gradually in the companies and there is an increasing proportion of private shareholdings. As shown in Table 7 below, starting with a very high concentration of state voting shares, 73% in 1993, it has be reduced to 56.8% in 1999, decreased by 16.2%. In comparison, the individual shares increase by 16.2% to 43.2% in 1999 from 27% in 1993. Table 7 Evolution of State and Private Control of All Sample Companies from 1993 to 1999 No. 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 129 73% 65.7% 61.1% 59.5% 58.3% 57.2% 56.8% Private Shares Voting 129 27% 34.3% 38.9% 40.5% 41.7% 42.8% 43.2%

The rise of A-shares and the decline of state shares over time may be due to first, the flotation of the employee shares and second, the subsequent issuance of shares to the public. Further, this may also be due to the allocation of share dividends or rights offerings of private shares. For liquidity reasons, the state including legal person shareholders prefers cash dividends to stock dividends or rights offerings (Xu and Wang 1999, page 81). Therefore, the amount of private shares will augment because of the distribution of share dividends or rights offerings. Employee shares are shares sold to the staff at a discounted price as a kind of benefits when a company makes an initial public offering. The initial purpose of the state to set up an employee share scheme in the early stage of the stock market is to promote confidence in the market and to create managerial incentives. Not all the companies have employee shares at the time of IPO. In our sample, there are 109 companies out of 129 with employee shares. Different from employee shares in developed economies, the employee shares in China are not tradable and are collectively owned by the company s trade union. A single employee does not have voting rights on his/her shares. After the shares have been held by the employee in the company for 6-12 months after the IPO date, the company may apply to the China Security Supervisory Commission ( CSRC ) for allowing its employees to sell or trade the shares in the open market. According to the Standard Opinions on Limited Company By Shares, the maximum amount of employee shares must not exceed 10% of the total shares that the company offers in the open market instead of the total share equity of the company. However, after the implementation of the Company Law of the People s Republic of China on 1 st July 1994, companies can no longer be sanctioned and allowed to distribute employee shares. Hence, most companies obtained approval from the CSRC for trading of employee shares in the market in 1995. Table 8 presents the state of employee shares over the 7 years.

Table 8 The Average of Employee Shares from 1993 to 1999 No. 1993 1994 1995 1996 1997 1998 1999 Employee Shares 109 5.3% 1.5% 0.2% 0.2% 0.1% 0 0 * Please note that the proportion of employee shares is taken as a percentage to the total shares in the companies, instead of to the total private shares. In order to study the changes of the state and private control within the context of different industrial sectors and firm size, we make a cross-sectional comparison between the averages of the sectors and size groups for each year from 1993 to 1999. A total percentage of reduction over the 7 year is also calculated. Table 9 below reports the way of evolution of the state shares. The fall in the state shares is greater in industrial companies, such as those producing electronic and electrical products, machinery an equipment, food, beverage and pharmaceutical products as well as chemical and textile products. Regardless of the fact that there is a gradual reduction in the state shares overtime, the proportion of state shares decreases faster in the just two years after the IPO than during later years. For instance, companies in the electrical and electronic sector decrease their state shares by 13.8% from 1993 to 1995. Similarly, those in machinery and equipment sector reduce their state shares by 15.7% and in food, beverage and pharmacy decrease by 12.5%. The same evolution trend can be found in the different size groups. Those in quintile 3 drop by 13.2% while those in quintile 4 drop by 14%. The significant decrease during these two years may be because of the flotation of the employee shares in 1995. Moreover, it may also be due to the increase of the total share capital as a result of the share dividends paid to individual shareholders, whilst the state shareholders receive cash dividends instead. Regarding firm size, the largest quintile has the lowest rate of reduction in state shares. In other words, the state retains majority control in largest companies.

Table 9 Evolution of the State Shares (the statistics for state shares in the table are in percentage) Description Number 1993 1994. 1995 1996 1997 1998 1999 Changes Across Sub-sample years Industrial Sector Commercial Companies 16 73 63.9 59.5 59.2 58.8 58.4 58.1 15.1 Diversified 13 70.5 70.4 65.6 66.6 64.3 64.1 63.6 6.9 Industrial companies Electrical and Electronic 17 76.6 70.1 62.8 58.1 58 57.4 55.8 20.8 Products Machinery and Equipment 16 73.7 61.7 58 55.7 55.5 53.6 53.6 20.1 Food, Beverage and Pharmacy 11 74.5 67.1 62 60.1 57.8 56.3 55.8 18.7 Motor Vehicle and steel products 15 69.4 61.6 60.6 58.9 57.5 57.3 57.3 12.1 Chemical and Textile Products 15 69.7 62.1 57.8 55.7 54.9 50.9 50.9 18.8 Public Utilities 13 78.7 73.2 65.3 64.5 62.4 62.1 62.3 16.4 Real Estate 13 70.5 62.6 59.4 58.6 57.1 55.9 55.1 15.4 Firm Size Lowest Quintile 26 72.4 69.1 62.7 60.4 58.9 57.7 57.2 15.2 Quintile 2 26 71 63.6 60.1 58.4 57.8 55.7 55.8 15.2 Quintile 3 26 71.6 62.8 58.4 57.1 56 54.8 54.8 16.8 Quintile 4 26 76.3 64.6 61.3 59.4 58.6 57.6 57 18.7 Highest Quintile 25 73.5 68.5 63 62.2 60.5 60.3 59.2 14.3 As a mirror to the evolution of the state shares, there is a tendency of private shares rising in the companies during the 7 years. The evolution of private shares is shown in Table 10 Below. Remarkable increases in private shares can be found in sectors, such as electrical and electronic products, machinery and equipment, food, beverage and pharmaceutical products as well as chemical and textile products. Regarding firm size, obvious increases are shown in the 4 th and 3 rd quintiles. The largest quintile has the lowest changes in private control.

Table 10 Evolution of the Private Shares (the statistics for private shares in the table are in percentage) Description Number 1993 1994 1995 1996 1997 1998 1999 Changes Across Sub-sample years Industrial Sector Commercial Companies 16 27 36.1 40.5 40.8 41.2 41.6 41.9 14.9 Diversified 13 29.5 29.6 34.4 33.4 35.7 35.9 36.4 6.9 Industrial companies Electrical and Electronic 17 23.4 29.9 37.2 41.9 42 42.6 44.2 20.8 Products Machinery and Equipment 16 26.3 38.3 42 44.3 44.5 46.4 46.4 20.1 Food, Beverage and Pharmacy 11 25.5 32.9 38 39.9 42.2 43.7 44.2 18.7 Motor Vehicle and steel products 15 30.6 38.4 39.4 41.1 42.5 42.7 42.7 12.1 Chemical and Textile Products 15 30.3 37.9 42.2 44.3 45.1 49.1 49.1 18.8 Public Utilities 13 21.3 26.8 34.7 35.5 37.6 37.9 37.7 16.4 Real Estate 13 29.5 37.4 40.6 41.4 42.9 44.1 44.9 15.4 Firm Size Lowest Quintile 26 27.6 30.9 37.3 39.6 41.1 42.3 42.8 15.2 Quintile 2 26 29 36.4 39.9 41.6 42.2 44.3 44.2 15.2 Quintile 3 26 28.4 37.2 41.6 42.9 44 45.2 45.2 16.8 Quintile 4 26 23.7 35.4 38.7 40.6 41.4 42.5 43 19.3 Highest Quintile 25 26.5 31.5 37 37.8 39.5 39.7 40.8 14.3 In summary, even if the state still controls the companies, there is a gradual reduction in the state shares and a tendency of a rising proportion of private shares over the years. To analyse the results from the control and evolution part with those from ownership concentration, there are some stylized facts. Companies that are concentrated with indirect state shares (legal persons and employee shares), A shares and foreign shares, namely, companies manufacturing electrical and electronic products, machinery and equipment, chemical and textile products have a higher reduction in the state shares over 1993 to 1999, whilst, the increase of private shares in those sectors are more substantial. The largest quintile has the lowest decrease of the state shares. In summary, larger, stable and low risk companies have larger proportion of state shares and highly controlled by the state. Smaller and growing companies have more private shares and thereby, more widely held.

Section III - An analysis of extreme cases of control and evolution issues in the IPO companies This part aims at shedding light on some special cases relating to the issue of control and evolution in our sample companies. As mentioned previously, the state shareholding accounts for 61.6% on average. However, some companies may have as high as 93.8% of state shares in the year of the IPO whereas some may have very low state shareholding, such as 28.6%. This raises our interest to investigate what the nature of these companies is, what sectors and size groups they belong to and how they perform during the period of study. Moreover, since the study in this paper is going to be the foundation for our further investigation on linking corporate governance issues with financial performance, the brief analysis of these exceptional companies gives an overview on whether concentrated state control is effective and whether there are improvements on performance along with the reduction of state control. The company that has the lowest state shareholding (20.1%) is Shenzhen Gintian Industry Company Limited, which is engaged in real estate development business and belongs to the highest quintile group. Shenzhen Gintian Industry Company Limited is the first company that shed solely state-owned status and transformed into limited company by shares in 1988. It is also one of the earliest companies that was privatized within the sector of real estate business in 1993. The changes of its state shares, private shares, accounting and share price performance from 1993 to 1999 are shown in Table 11 below. Table 11 Shenzhen Gintian Industry Company (0003) 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 28.6% 23% 20.1% 20.1% 20.1% 20.1% 20.1% Private Voting Shares 71.4% 77% 79.9% 79.9% 79.9% 79.9% 79.9% Return on Equity 14% 14% 4.6% -2.5% 4% -59% -178% (Net income/equity) Return on Asset 5.8% 5.2% 1.6% -0.8% 8.2% -8.4% -6.1% (Net income/total Assets) Share Price ( ) 89.7 30 21 35 50 35 31 * Accounting data, i.e., return on equity and return on asset are obtained from the company s annual reports. Share prices are obtained from the Datastream. The annual share prices are the average of its monthly prices.

As shown above, the state losses control in Shenzhen Gintian Industry Co. Ltd. at the time of the initial public offering. Total private shares in the company even further increase in the following two years to 77% and 79.9% respectively, a rise of 8.5%. However, such high degree of autonomy in management doesn t seem to bring improvements to the company s performance. With an average of the monthly share price of 89.7 in 1993, it has been dropped to 31 in 1999. On the company s accounting performance side, it was not very satisfactory as well. Starting with 14% of ROE and 8.5% of ROA in 1993, the company suffered losses since 1996 (except for the slight improvements in 1997). The poor performance of the company during recent years is mainly due to the slump of real estate business. After Chairman Deng Xiaoping s Southern Tour in Guangdong Province in 1992 and the fast economic development in Southern China, the demand for residential and commercial buildings was huge. The period of 1992 and 1993 was the booming time for real estate business. After 1995, because of over supply of properties in the market, real estate business began to slump down and this in turn directly influenced the profitability and stock price of the company. By contrast, the following 4 companies have an extremely high state ownership at the time of IPO. Even if the shares have been reduced over the years, but the fraction of the state shares in the companies is still over 50%. The first representative company is the Jiangsu Yueda Company (details of the company are shown in Table 12 below), which manufactures tractors and automobile components. Firm size is medium in our sample companies. Jiangsu Yueda Company is one of the oldest state-owned companies in China that specialises in tractor manufacturing. However, since the increasing competition in the market, its products is less marketable recently and the firm has been forced to develop some other new business activities, such as pharmacy and infrastructure construction. The penetration into new business areas have slightly improved the company s overall performance. As shown in Table 12 below, its ROE and ROA have reduced from 1993 to 1998. However, since the company diversifies its business, there is a slight improvement in 1999 in the profitability.

The evolution of the company may also be reflected from the significant reduction of state control and increase of the private control in the company. State shareholding largely decreases from 93.8% in 1993 to only 60.2% in 1994 and 1995. This fraction reduces again to 50.8% in 1996. Between 1993 and 1999, state votes are reduced by 46% while private votes are increased by more than 7 times. Market response to the reduction of state control in the company is positive and this can be reflected from the increase value of its share price over the years. Table 12 Jiangsu Yueda Company (600805) 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 0.938 0.602 0.602 0.508 0.508 0.507 0.507 Private Voting Shares 0.062 0.398 0.799 0.492 0.492 0.493 0.493 Return on Equity 16% 17% 14% 10% 10% 5.8% 12% (Net income/equity) Return on Asset 7% 8% 6.4% 6.4% 5.8% 3.5% 5.8% (Net income/total Assets) Share Price ( ) 14.8 5.7 9.8 11.4 15.5 19.7 19.9 * Accounting data, i.e., return on equity and return on asset are obtained from the company s annual reports. Share prices are obtained from the Datastream. The annual share prices are the average of its monthly prices. Table 13 below shows information regarding the 7-year control and performance of Shanghai Material Trade Center Co. Ltd., which has a similar evolution trend to Jiangsu Yueda Company. Shanghai Material Trade Center Co. Ltd. engages in trading, import and export business. The state keeps control in the company after the public offering and reduces its proportion of control in 1995, thereby increasing the private shares. The fraction of private shares rises by more than 4 times between 1993 and 1999. Regarding the firm s size, it belongs to the 3 rd Quintile. Regarding its performance, the company s profitability seems to negatively relate to the increase of private control. Moreover, its market price also declines along with the decrease of state control. This may be due to the fact that since 1995, the import and export business has become more competitive in the country and the profit margin has been squeezed.

Table 13 Shanghai Material Trade Center Co. Ltd. (600822) 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 0.929 0.929 0.684 0.684 0.684 0.684 0.684 Private Voting Shares 0.071 0.071 0.316 0.316 0.316 0.316 0.316 Return on Equity 14% 10% 2.6% 3% 1.2% 0.7% -2.8% (Net income/equity) Return on Asset 6% 4% 1.4% 1.8% 0.7% 0.4% -1.3% (Net income/total Assets) Share Price ( ) 10.6 6.7 5.5 4.1 4 4.7 6.3 * Accounting data, i.e., return on equity and return on asset are obtained from the company s annual reports. Share prices are obtained from the Datastream. The annual share prices are the average of its monthly prices. To the opposite of the above two companies, in which the state significantly decreases its control in the company after the IPO, the reduction rate in Shanghai Dajiang (Group) Stock Co. Ltd. and Wenergy Co.Ltd. is less significant. Table 14 and 15 below show details of the respective companies. Shanghai Dajiang (Group) Stock Co. Ltd belongs to the sector of diversified industrial companies and also the largest size quintile. Between 1993 and 1999, the state keeps a constant level of control of the company with the state shares only dropping by 3.4%. The proportion of private shares is small. This is consistent with fact that traditional, stable and larger companies have high state control. Similarly, Wenergy Co. Ltd., which is in the utility sector, is extremely state shares concentrated. Regarding the accounting performance of Shanghai Dajiang (Group) Stock Co. Ltd. and Wenergy Co Ltd., their profitability decreases since 1996. This may associate with the macro-economic environment, i.e. the fluctuation of the national economy influenced by the Asian crisis in 1997. Their share prices are relatively stable as compared to the previous three companies. The market performance of these two companies implies that individuals are confident in investing in firms substantially controlled by the state. Furthermore, it may also reflect that the state supports large, strategically important companies, especially those in the utility and infrastructure sectors.

Table 14 Shanghai Dajiang (Group) Stock Co. Ltd. (600695) 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 0.904 0.899 0.873 0.873 0.873 0.873 0.873 Private Voting Shares 0.096 0.101 0.127 0.127 0.127 0.127 0.127 Return on Equity 15% 20% 20% 12% 3.2% -13% 0.7% (Net income/equity) Return on Asset 9.2% 9% 9% 5.7% 1.3% -5.6% 0.3% (Net income/total Assets) Share Price ( ) 11.8 6.2 7.5 6.8 6.1 9.2 8.5 *Accounting data, i.e., return on equity and return on asset are obtained from the company s annual reports. Share prices are obtained from the Datastream. The annual share prices are the average of its monthly prices. Table 15 Wenergy Co.Ltd (0543) 1993 1994 1995 1996 1997 1998 1999 State Voting Shares 0.886 0.886 0.886 0.856 0.856 0.818 0.818 Private Voting Shares 0.114 0.114 0.114 0.144 0.144 0.182 0.182 Return on Equity 13% 21% 20% 15% 13% 9.3% 8% (Net income/equity) Return on Asset 13% 17% 16.5% 13% 12% 8.7% 6.8% (Net income/total Assets) Share Price ( ) 0.53 0.265 0.24 0.44 0.63 0.52 0.45 * Accounting data, i.e., return on equity and return on asset are obtained from the company s annual reports. Share prices are obtained from the Datastream. The annual share prices are the average of its monthly prices. To summarize, the purposes of looking at the five companies are threefold. Firstly, we intend to shed light on the evolution of these companies, which are either substantially state controlled or private controlled at the time of flotation. Secondly, we aim at investigating the plausible link between their post issuance financial performance and the evolution of control. Thirdly, the results we find in this part would be a foundation for our future cross-sectional analysis that associates corporate governance issues with financial performance in Chinese IPO companies. The five companies above can be further classified into three general groups. The first group includes Shanghai Gintian Industry Company, which is privately controlled after public offering. Its performance is unsatisfactory over the years, which may imply that private ownership concentration and control are not helpful for the company s financial performance. The second group consists of Jiangsu Yueda Company and Shanghai Material Trade Center Co. Ltd, which belong to the third quintile group and have high concentration of state shares in the year of IPO. However, state control reduces very fast in the subsequent years. This can be

explained by the government s policy that let smaller size companies being privatized first and growing under the market competitive. The third group includes Shanghai Dajiang (Group) Stock Co. Ltd. and Wenergy Co. Ltd.. The former is a diversified industrial company and belongs to the largest quintile while the latter is a utility company and belongs to the lowest quintile. As the state supports larger companies and also companies in the utility sector, it does not reduce control in these two companies. The performance of these two companies are also more optimistic than the other three companies. Please also refer to the Appendix II for the market performance of each company and Appendix III for the comparative performance of the five companies. We adjust monthly returns of each company with the market composite index. Therefore, in the diagrams, they present the accumulated market adjusted returns of the companies over the 7-year period. (Method and data for calculation can be provided on request.) Section IV - Conclusion In summary, large and mature industrial companies that involve in shipbuilding, metallurgy, motor and vehicles as well as pharmaceuticals have higher concentration of state ownership. Smaller and growing companies, such as department stores, information technology companies and those manufacturing chemical and textile products are more widely held by individual investors in the market. This may be because large and mature industrial companies have less growing opportunities than smaller companies. Moreover, their share prices are more stable and less speculative profits can be made than smaller companies. Therefore, individual investors are more likely to purchase shares in smaller, growing companies than in larger, stable companies. On the other hand, indirect state-owned shares are concentrated in companies manufacturing electrical and electronic products and smaller companies. This may reflect that smaller companies may be more efficient in corporate control than larger ones because of more outside directors on the board. In contrast, foreign ownership is concentrated in larger and mature industrial companies. This may mainly due to the issuing requirements set by the state. Only companies that are complied with certain criteria, such as requirement on minimum total assets, past performance and social development objectives, can issue foreign shares and then list

on foreign stock exchanges. On the corporate control side, we find that the state is still the largest shareholder for most companies (except for the special cases that we analyze) 7 years after flotation and retains majority control. The public utility sector, in which companies are mainly supported by local government authorities, displays higher state control than other sectors. Private investors with their minimal individual shareholding can hardly involve in either monitoring or supervising the activities in the company. However, the proportion of state control in the companies has been decreased during the 7-year period. Substantial decline is shown in the sectors of electrical and electronic products, machinery and equipment, food, beverage and pharmacy as well as chemical and textile products. The largest quintile, which comprises utility, infrastructure and traditional industrial companies that receive supports from the government in terms of capital and policies, has the lowest reduction in state control.