The The The The influence influence influence influence of of of of compensation compensation compensation compensation committee committee committee committee and and and and ownership ownership ownership ownership structure

The The The The influence influence influence influence of of of of compensation compensation compensation compensation committee committee committee committee and and and and ownership ownership ownership ownership structure

structure structure structure on on on on CEO CEO CEO CEO compensation compensation compensation compensation in in in in China China China China’’’’ssss listed listed listed listed firms firms firms firms

October October October

October ,,,, 2008 2008 2008 2008

Author

Author Author Author Research Research Research Research Supervisor Supervisor Supervisor Supervisor

Lin He Dr. D.H.M.Akkermans

S1744798 Faculty of Economics & Business

L.He.3@student.rug.nl d.h.m.akkermans@rug.nl d.h.m.akkermans@rug.nl d.h.m.akkermans@rug.nl d.h.m.akkermans@rug.nl

Abstract Abstract Abstract Abstract

With the heated argument of the rising CEO compensation, the role of the board and the compensation committee had viewed as a vital influence.

This study examined the effect of China’s special board structure, ownership structure and compensation committee on CEO compensation and average salary level of executive directors from the social comparison perspective. After controlling the economic factors form principal-agent approach, I found the state-owned share is significantly and negatively related to the CEO compensation. And the relation between the average salary level of compensation committee and CEO compensation is positive, but weakly. No evidence demonstrated the influence of foreign membership in compensation committee on designing the compensation level.

Keywords: CEO compensation, executive directors, social comparison theory,

ownership structure, China

Contents Contents Contents Contents

1 Introduction ………4

2 Related theories ………5

2.1 CEO compensation and principal-agent theory ………5

2.2 CEO compensation and social comparison theory ………8

3 Institutional backgrounds ………10

3.1 The Structure of board system in China ………10

3.2 Ownership structure of listed firms in China ………13

4 Hypotheses ………14

5 Methodologies ………18

5.1 Data sources and sample selection ………18

5.2 Economic model and variables ………19

6 Analysis and Results ………22

7 Discussion and conclusion ………24 Appendix A

Appendix B

Reference

1 111 IIIINTRODUCTION NTRODUCTION NTRODUCTION NTRODUCTION

The issue of CEO compensation has attracted the attention of economist, sociologist and press world-wide. They did not only focus on how much should pay to CEOs, but also how to pay. As early as 1980s, scholars started to rely on principal-agent theory (Jensen and Murphy 1990), human capital theory (Hogan and McPheters, 1980), tournament theory (Lazear, 1989; Main et al., 1993) and social comparison theory to study CEO compensation.

Principal-agent theory studies the relation between CEO compensation and economic factors, such as firm performance, growth, firm size. The performance-based pay in managerial compensation plan tends to employ quite simple specifications of how firm performance influences compensation (Joskow and Rose, 1994). But today's CEOs won't work hard without lucrative incentives. And a CEO's job is riskier and harder than it used to be (Wessel, 2006).

In the reality, shareholders or board of directors can hardly observe the CEO’s effort and act to maximize shareholder’s interests rather than their own interests. The problem is CEOs still can get considerable amount of compensation and incentives, even when CEOs did not offer the effort or the performance is poor. At this point, social comparison theory can work to explain the setting of CEO’s compensation. Based on social comparison theory, individuals have a drive to compare themselves to others to evaluate their own ability or attribution. And since CEO compensation is determined by the compensation committee, and the effort and performance of CEOs is sometimes ambiguous, social comparison theory began to attract scholars’

attention.

My focus is consistent with the study of the influence of the board structure and ownership on CEO compensation. Most studies concern firms in the U.S.

or the U.K. which adopt a one-tier board system. However, China’s listed firms adopt a dual board system and state-owned enterprises (SOEs) hold dominant power. Thus it is a good notion to test the role of compensation committee in China’s listed firms under this environment. In this paper, I will study the effect of ownership structure, the role of compensation committee and some standard economic factors on CEO compensation and average salary level of executive directors.

Thus my main research questions concern the relation between CEO compensation and average salary level of members of compensation committee and the linkage between the average salary level of board executive directors and compensation committee. In addition, I also attempt to study whether different ownership (state-owned, foreign) affects the amount of CEO compensation. This paper proceeds as follows. Section 2 mentions two related theories. In section 3, institutional background, such as board system and ownership in China’s listed firms, will be discussed. And hypotheses of this study present in section 4. In section 5, it talks about methodology and detailed description of variables. Analyses and outcome present in section 6. The last part is discussion and conclusion.

2 222 R R R RELATED ELATED ELATED ELATED THEORIE THEORIE THEORIE THEORIES S S S

2.1 2.1

2.1 2.1 CEO CEO CEO CEO compensation compensation compensation compensation and and and and principal- principal- principal- principal-agen agen agen agentttt theory theory theory theory

As Murphy (1999) indicated, the modern history of executive compensation

research began in the early 1980s and paralleled the emergence and general

acceptance of agency theory (e.g., Fama, 1980).

The agency problem arises simply because there is a separation of ownership and control. The problems are most commonly found in the corporate form of business organization, where the shareholders can not closely monitor the CEOs (O’Reilly III and Main, 2007). There is a conflict of interest between shareholders and CEOs. And the shareholders know what actions they want the CEO to take and want such actions appropriate to maximize their interests,

but they cannot directly observe the CEO’s actions (Murphy, 1999). Only CEOs can observe their own actions and compare their private gain and cost when pursuing particular activities.

When the board of directors sets the compensation of CEO and other top executives, they had the choice of contacting either on future earnings or a combination of future earnings and stock price. The board is viewed as principal, and the CEO is its agent. Because the CEO’s effort choice is unobservable to the board (Milbourn, 1997), the board faces effort-uncertainty in designing the CEO’s contract. Meanwhile, CEOs face different managerial incentives under different compensation schemes (Larcker 1983; Lambert and Larcker 1987). Compensation schemes tend to be implemented to elicit high corporate financial performance and managerial changes consistent with the goals of the firm (Baek and Pagan, 2006).

The form of CEOs’ compensation consists of base salaries, annual bonus plan,

stock options, and other forms (restricted stock, long-term incentive plan, and

retirement plan). Virtually, every company offers an annual bonus plan based

on a single-year’s performance. Under a typical plan, no bonus is paid until a

certain performance is achieved. And all annual bonus plans provide

incentives to increase company profits. Clearly, if the board directors are

maximizing shareholder benefits they will be exercising independent

directors’ oversight by seeking minimum agency costs and linking incentives

to maximize shareholder value. This focus on shareholder interests requires that the board bargains ‘at arm’s length’ with top executives to best ensure outcomes favorable to shareholders.

Pay-for-performance was used to link CEO compensation with their performance. As Stroh et al. (1996) addressed, performance-based pay fits with agency theory in that remuneration contracts can be used as a means to reduce or eliminate potential conflicts of interest between the CEOs and shareholders. Empirically, most of firms use single performance measurement.

As earlier as 1982, Holmstrom argued that basing CEO compensation on performance measured relative to aggregate performance in the industry or market provides CEOs with incentives to increase shareholder wealth while filtering out the risk-increasing effects of industry wide and market wide factors beyond the control of executives. Murphy (1985) appointed that compensation must be tied to observed productivity. Moreover, growth of firm sales (Baker et al., 1988) – another measure of performance - is also strongly related to executive compensation. Abowd (1990) analyzed the effect of performance-based executive compensation on company performance, finding that companies with higher pay–performance sensitivity are more likely to achieve better economic returns in the following year. Pay- performance sensitivity is one measure of the firm’s commitment to linking pay to performance, and is becoming an increasingly popular measure in research.

However, results of researches on ‘pay-for-performance’ were insignificantly and negatively related to CEO compensation. It suggested that ‘pay-for- performance’ may be not reasonable to set CEOs’ and other top executives’

compensation. Jensen and Murphy (1990) argued that the pay-performance

sensitivities generally observed are too low to be consistent with principal-

agent theory. Another anomalous result from them is that the pay for

performance relation is independent of stock ownership. In fact, it might even make sense to have pay go up in bad years to compensate for some of the loss the CEO is taking in the stock market. Schaefer (1998) showed that the pay sensitivity (measured as the dollar change in CEO wealth per dollar change in firm value) falls with the square root of firm size. Bebchuk and Fried (2004) claimed that due to the powerful influence of the CEO, there is a lack of arm’s-length bargaining between the CEO and the board; therefore, a suboptimal executive compensation contract which does not link pay to performance results. Their evidence indicates that cash compensation has been at best weakly correlated with firms’ industry adjusted performance.

Such compensation has been generously awarded even to managers whose performance was mediocre relative to other executives in their industry.

Another analyses, by using a sample of 210 large listed firms in Japan, shown a 1 percent increase in company performance will lead to just a zero to 0.33 percent increase in directors’ pay. The link between pay and performance is weak in Japan, because the proportion of firms with negative pay- performance sensitivity is large (Kato, 2006).

2.2 2.2

2.2 2.2 CEO CEO CEO CEO compensation compensation compensation compensation and and and and social social social social comparison comparison comparison comparison theory theory theory theory

Several alternative theories emerged, such as relative-performance evaluation (Gibbons and Murphy, 1989), tournament theory (O’Reilly et al. ， 1988), human capital theory, and social comparison theory (Belliveau et al., 1996). It was Festinger（1954） that first put forward the social comparison conception.

The idea is that there is a drive with individuals to evaluate their abilities

through comparison with others. And similar others (people with similar

abilities and attributes) are preferred for comparison. Since the compensation

of CEOs is determined by the compensation committee, and the performance

of CEOs is sometimes ambiguous, social comparison theory began to attract

scholars’ attention. O’Reilly et al.’s early research paper (1988) examined CEO

compensation from two perspectives, one is tournament theory and the other is social comparison theory. Concerning social comparison, he explained why this theory is reasonable:

“The compensation committee consists of disinterested outside directors, frequently individuals who are CEOs of other firms……those selected to serve on the board and the compensation committee may be chosen because of their similarity to each other (Festinger, 1954), and these individuals may then anchor their judgments about appropriate CEO remuneration, that is, by comparing the CEO's salary to their own.”(1988: 261)

Later, Belliveau et al., (1996) stressed to employ social capital, social comparison and other social influences to explain CEO compensation. In their paper, they focused on the comparison between CEOs and chairman of compensation committee. The chairman is significant because the chairman can use his or her formal authority to control the flow of information to committee and frame the premises of the decision (Pfeffer, 1992). They conceptualized social capital in terms of both social similarity and the absolute and relative status between a CEO and compensation committee chairman and. Furthermore, they mentioned a global comparison which based upon attributes relevant to an equivalent set of people (for instance, CEOs comparing themselves to other CEOs). Additionally, CEOs of higher status have more influence on those who set their compensation than that of lower status. Similarly, compensation committee chairman may compare their own status to that of other compensation committee chairmen. And members of corporate boards often value their positions, gaining both personal prestige and remuneration with each directorship (Lublin, 1991).

More recently, O’Reilly and Main (2005) drew on a more psychology-based

interpretation of the mechanism of action (reciprocity and social influence)

that underlies how boards of directors may operate. These two mechanisms will affect the dynamics of CEO-board interactions and the executive wage setting process. “Whenever you find highly paid CEOs, you will find highly paid directors. It’s no accident,” Crystal (1991) said. The structure of the idea of reciprocity is that “when one party benefits another, an obligation is generated (Gouldner: 174).” In the context of corporate governance, there are both financial and status-related benefits from serving on a board. Insofar as the CEO is seen as at least partly responsible for aspects of their appointment, for example by serving on the nominating committee or paying generous fees, a board member can expect to feel some reciprocal obligation (Lorsch 1989).

When the chairman of the compensation committee receives comparatively higher fees, CEO cash compensation is significantly higher. Every $1,000 the board member receives is associated with an increase of $1,746 in CEO cash compensation.

3

333 INSTITUTIONAL INSTITUTIONAL INSTITUTIONAL INSTITUTIONAL BACKGROUDS BACKGROUDS BACKGROUDS BACKGROUDS

3.1 3.1

3.1 3.1 The The The The Structure Structure Structure Structure of of of of board board board board system system system system in in in in China China China China

Even in china, the issues of CEO’s and top executives’ compensation, appointment and dismissal decision, and the huge salary gap between top executives and employees are discussed by economists, businessmen, and public. But before studying CEO compensation, we should first understand China’s board system and ownership structure in listed firms.

Different firms adopt different board systems. Basically there are 3 types of board systems: one-tier board system, two-tier board system and dual board system.

In a one-tier board system which is common in the firms of the U.S. and U.K.,

executive and non-executive directors are directly elected by the shareholders.

This setting ensures the equal rights of each directors and their responsibility for codetermination. Moreover, the supervisory board does not set up in this board system.

In a two-tier board system, existing in Germany and the Netherlands, the supervisory and managerial functions were separated. The supervisory board is comprised of non-executive directors whom elected by shareholders. The main function of supervisory board is to exercise supervision over management. However, the board of management consist of executive directors whom elected by the supervisory board. A key feature of this system is that the supervisory board has authority to appoint and change the members in the board of management.

The dual board system also separates the function of supervision and management. The role of the board of directors is to exercise the business activity and management. It involves in strategic decisions and appointment of senior managers, meanwhile it is also responsibility to exercise supervision over senior managers. Although the supervisory board has parallel status to the board of directors, it is unable to directly veto the decisions made by the board of directors. It is solely responsible for supervision of the board of directors and mangers. The member of the supervisory board could be shareholder, employee or specialist. Unlike the two-tier board system, all the members in the supervisory board and the board of directors are elected by shareholders. Neither the supervisory board nor the board of directors can appoint or dismiss the members in each board setting except the general meeting of shareholders. China’s and Japanese firms adopt this system.

As to the independent director system, it has been incorporate in a one-tier

board system. It originated from American firms in 1930s for curing corporate

governance problems of public companies which have widely dispersed shareholders. In china, independent director system was first introduced to companies which listed in the off shore stock exchange market from 1999.The initial purpose of establishment of independent directors system had targeted on controlling shareholders, rather than management. Because the ownership structure of listed firms is highly concentrated, corruption and expropriation by these controlling shareholders really cause big problems. Thus, in a word, the current composition of the board of directors is formed from executive directors, non-executive directors and independent directors ¹ . And at the very least the independent directors should account for one third of total directors.

Additionally, compensation committee established, as well as the audit committee and nomination committee. The compensation committee normally is comprised of 3 to 7 directors. Its function is to review and recommend compensation policies, as well as the compensation and incentive level for executives, directors and employees. The compensation plans also need to be approved by the consentience of the board of directors and the general meeting of shareholders.

3.2 3.2

3.2 3.2 Ownership Ownership Ownership Ownership structure structure structure structure of of of of listed listed listed listed firms firms firms firms in in in in China China China China

China’s listed firms have special ownership structure. Normally, shares owned by largest shareholder can be classified into four types: state-owned share, legal person shares, individual share and foreign share. State-owned shares are held by central government, provincial and municipal government or state-owned enterprises ² (SOEs). These shares are not allowed to circulate

1

The term "Independent Director" means a director who does not hold any position in the company other than director and who has no relationship with the listed company or its shareholders, so that it could not influence on his option and decision-making (from Establishment of Independent Director Systems by Listed Companies Guiding Opinion, 2001).

2

State-owned enterprises (SOEs), means these companies which all property of the enterprise belongs to the state.

They are subject to the State Owned Asset Supervision Administration Commission of the State Council and

supervised by local State Owned Asset Supervision Administration Commissions.

publicly. Legal person shares (also called institutional shares) are owned by legal entities, such as investment institutions and other enterprises excluding SOEs. These shares also cannot be circulated on stock exchanges of Shanghai and Shenzhen, though they can be sold to other legal entities by agreement and upon approval by the government. Individual share are owned by citizens. Foreign shares are owned by foreign individual or foreign company.

Moreover, based on its listing location and investment object, the stock can be distinguished into 3 categories: 1) A share, which is issued and sold with CNY (Chinese Yuan) by Chinese firms. And only companies in China can buy and sell this kind of stock in stock market. 2) B Share must be bought and sold by foreign currency, like U.S. dollar. The purpose of this share is to help Chinese firms collecting foreign capital; meanwhile provide a chance for foreign company to buy shares in China. 3) H Share means the issuing company registers at mainland of China, and listed at Hong Kong. Besides, B shares in Shanghai Stock Exchange quoted with U.S. dollar, yet quoted with Hong Kong Dollar in Shenzhen Stock Exchange. Recently individuals in China have been allowed to buy foreign shares if they have account with foreign currency.

In addition to having both A- and B-shares, some Chinese firms also are listed on Hong Kong and foreign stock exchange.

The majority of listed firms in China are SOEs. Based upon the rank of Top

500 enterprises of China in 2007, there are 349 state-owned enterprises, which

account for 69.8% of the total number; while private enterprises just account

for 17.8%. In addition, on average, state-owned shares accounts for at least

one-third of a firm’s issued share. Li Huitian (2001) argued that the state

shareholder in control is the main feature of these Chinese firms. With the

panel dataset of five years, it was found that firms under the control of the

state shareholder are valued lower than the comparable firms under the

control of a non-state shareholder.

4

444 HYPOTHESES HYPOTHESES HYPOTHESES HYPOTHESES

Ownership structures are important in corporate governance because they affect the compensation and incentives of managers and thereby the efficiency of the board system. In almost all of firms in China there is a dominant shareholder who has significant influence over the way a firm is run and on the appointment and pay of the CEO (Firth et al., 2006). And the dominant shareholder is always a SOE. Under this highly concentrated ownership, executives and managers are directly appointed and controlled by dominant shareholder. Based on interviews with government officials, stock exchange regulators, security and corporate lawyers, and officials at both listed and non-listed firms, Kato and Long (2004) ^c oncluded that corporate governance in listed firms which owned by SOEs is of very low quality, characterized by excessive powers of the CEO and insider control, inadequate safeguards for outsiders, weak managerial incentives, and inadequate transparency and disclosure. They also argued another influence of the control by SOE on compensation level. That is, the controlling shareholder normally doesn’t allow the listed firm to set its executive compensation level higher than that of the controlling shareholder even if the listed firm performs far better.

Additionally, SOEs were always indirectly controlled by the state. And as stressed by La Porta et al. (1998), the state has been characterized as being a poor monitor of a firm’s financial performance because they are too detached from the firm. The state may pressure the firm to pursue objectives other than profit maximization (e.g., to increase employment).

Hypothesis 1a: when the listed firm is controlled by SOEs, it will decrease the CEO compensation.

Hypothesis 1b: when the listed firm is controlled by SOEs, it will also decrease the

average salary level of executive directors.

When a foreign company or institution become a firm’s shareholder, the firm is more likely to adopt international standards of corporate governance and business practices. Foreign investors are likely to pressure firms to hire better qualified CEOs who have international experience (Firth et al., 2007). They also seek to shape compensation through proxy voting and by formulating voting guidelines (Khan et al., 2005). Thus when foreign investors own certain shares in a firm, they gradually have the voting, electing and monitoring rights, so they can affect the corporate governance and compensation plans.

Hypothesis 2a: the presence of foreign ownership has positive effect on CEO compensation.

Hypothesis 2b: the presence of foreign ownership also positively impact on the average salary level of executive directors.

The compensation committee is responsible of decide the compensation of CEO and other top executives. According to the prediction of social comparison theory, these individuals judge an appropriate CEO compensation by comparing their own compensation. The compensation committee usually is comprised of independent directors. Some studies find that independent directors inflate a CEO’s compensation (Firth et al., 1999).

This might arise because the independent directors use the high pay for the

CEO as a comparison benchmark when they negotiate or renegotiate their

compensation at the firms or organizations where they work fulltime. Thus

they have an incentive for average CEO and top executive pay to rise (Firth et

al., 2007). Additionally, foreign board members are one step forward in a

firm’s globalization process. For improving corporate governance, more

foreign experts and executives get involve in Chinese company. As I said

above, if social comparison actually worked in here, foreign members in

compensation committee will increase the CEO compensation, because they will compare their compensation in foreign company with that of in China’s company.

Hypothesis 3a: there exists a positive relation between the compensation of CEO and compensation committee.

Hypothesis 3b: the relationship between the average salary level of executive directors and compensation committee is also positive.

Hypothesis 4a: when a foreign member in the Compensation Committee, the CEO compensation will higher.

Hypothesis 4b: the presence of foreign member in compensation committee also raises the average salary level of executive directors.

Internationalization is becoming a vital impact on CEO compensation. As mentioned by Oxelheim and Randoy (2004), such international influence does enhance CEO compensation and explain differences in levels of CEO compensation within countries. It may reflect a risk premium justified by the harsher monitoring conducted by international owners and board members, when it is an internationally oriented firms. Foreign stock exchange listing usually is a proxy to analyze internationalization. It implies that firms will be supervised by international owners or investors. And firms will be exposed to new regulatory authorities and need to comply with international standards as regards disclosure or accounting.

Hypothesis 5a: if a firm is listed on a foreign stock market (including Hong Kong

stock exchange), the CEO compensation will higher.

Hypothesis 5b: the average salary level of executive directors increases with a firm whom listed on a foreign stock market.

A great deal of literatures had shown the influence of firm size on CEO compensation. Large and more complex firms have more economic power and more capable to pay higher compensation for attracting more talent CEO and top executives. This prediction is consist with the research study by Ke et al. (1999) and Anderson et al. (2000) which found a positive relation between executive pay and firm size.

Hypothesis 6a: CEO in large firm gets higher compensation.

Hypothesis 6b: average salary level of executive directors will rise when the company is large.

A large number of scholars advocate that the setting of CEO companion should base on firm performance. From prediction of the principal-agent theory, when the board (principal) can not observe the effect of CEO (agent), the information asymmetries are high. Thus the board of directors depends on firm performance as the benchmark for setting CEO and other executives’

compensation, because firm performance is observable. A quite large amount of specifications of firm performance were employed, such as accounting performance, stock market value, and change in performance sensitivity.

Relied on 1009 CEOs in 678 firms during 1970 to 1990, Joskow and Rose (1994) found CEO compensation is influenced by both accounting profits and shareholder returns. More recently, Kato et al. (2004) observed 251 firms from 1998-2001 in Korean capital market and found that cash compensation of executives is statistically significantly related to firm stock performance.

Hypothesis 7a: firm performance is positively related to CEO compensation.

Hypothesis 7b: firm performance is positively related to average salary level of executive directors.

5 555 METHODOLOGIES METHODOLOGIES METHODOLOGIES METHODOLOGIES

5.1 5.1

5.1 5.1 Data Data Data Data sources sources sources sources and and and and sample sample sample sample selection selection selection selection

50 firms listed in the stock exchange of Shanghai, Shenzhen and Hong Kong had selected. But I exclude the firms who have inadequate information about remuneration of CEO and independent non-executive directors, and that have no compensation committee. According to the rules of stock exchange, my sample belongs to 3 categories of industry: manufacturing, financing &

insurance and wholesaling & retailing. I use firm’s annual reports in 2007 as my source of information for CEO compensation, salary level of executive directors and compensation committee, firms’ ownership, foreign stock exchange listing. The information about firm size, ROE and industry was obtained from the website CNINF and REUTERS. Note that the ROE was collected in year 2006.

5.2 5.2

5.2 5.2 Economic Economic Economic Economic model model model model and and and and variables variables variables variables

Basically, I will use following equations to do regression analysis.

CEOPAY _i =α ₀ +α ₁ CC _i +α ₂ FCC+α ₃ SOE _i +α ₄ FOR _i +α ₅ FSEL _i +α ₆ SALES _i +α ₇ ROE _t-

1 , _I +α ₈ IND _i +ε _i

ASD _i =β ₀ +β ₁ CC _i +β ₂ FCC+β ₃ SOE _i +β ₄ FOR _i +β ₅ FSEL _i +β ₆ SALES _i +β ₇ ROE _t-1 , _i +β 8 IND i +ε i

Dependent Dependent

Dependent Dependent variables variables variables variables

Owing to incapability to collect data of all forms of compensation package of CEO in annual report, I only collect CEO’s total cash compensation (CEOPAY) – the sum of base salary and annual bonus - as my dependent variable. And, average salary level of board executive directors (ASD) will also take into consideration, treated as dependent variable too. ASD calculates the total cash compensation of executive directors divided by the number of board executive directors.

Independent Independent

Independent Independent variables variables variables variables

The measurement for testing social comparison relation among CEO, board executive directors and compensation committee follows the way that O’Reilly et al. did in their 1988’s paper. I collect data about average salary of compensation committee (CC) as my main independent variable. It calculates the sum of total cash compensation paid to compensation committee divided by the number of committee members. To capture the effect of foreign members on designing CEO compensation, dummy variable FCC coded 1 if there are foreigners in compensation committee and 0 otherwise. I expect that the average salary of compensation committee will positively related to CEO compensation and average salary level of executive directors, namely α 1 >0 and β 1 >0. And the presence of foreign members in compensation committee will raise the compensation paid to CEO and executive directors, that is to say α ₂ >0 and β ₂ >0.

The ownership of China’s listed firms is highly concentrated. State or state-

owned enterprises hold largest shares and get the control rights. Directors

who directly appointed by these large shareholders may abuse the power and

cause unbalance and ineffectiveness of the board of directors. By testing 549

companies from 1998 to 2000, Firth et al. (2006) concluded that state-

controlled firms pay less to CEO and foreign-invested firms pay more. I will use the percentage of shares owned by SOEs to measure the effect of state ownership. As to the foreign ownership, it was expected to have more effect when they are large shareholders. So I will use dummy variable (FOR) to measure the presence of foreign ownership and set 5% of shares as a cut-off for the presence of foreign ownership. Thereby FOR coded 1 if foreign ownership exists in the rank of ten largest shareholders in firm’s annual report and the shares is not less than 5%, 0 otherwise. I expect that SOE share reduce CEO compensation and average salary level of board executive directors which means α 3 <0 and β 3 <0, whereas the presence of foreign share increase CEO compunction and average salary level of board executive directors, namely α 4 >0 and β 4 >0.

A firm's degree of internationalization reflects its dependence on foreign markets for customers and factors of production, and the geographical dispersion of this dependence (Sullivan, 1994). Export and foreign sales, foreign stock exchange listing, and foreign board membership are used in most studies as proxy to analyze internationalization (Oxelheim and Randoy, 2003). In this paper, I choose foreign stock exchange listing (FSEL) as proxy for internationalization influence. Foreign stock exchange listing is a generally recognized way of breaking away from a domestic capital market (Doidgeet al., 2004). It also implies that the firm will be scrutinized by a new international investor. And it will be exposed to new regulatory authorities, and need to comply with new standards as regards disclosure and accounting.

Foreign stock exchange listing is coded 1 if the firm is listed outside the domestic stock market and 0 otherwise. And I expect that internationalization lead firms to pay CEO and executive directors higher compensation, that is to say α 5 >0 and β 5 >0.

Control Control

Control Control variables variables variables variables

Due to the prediction of principal-agent theory, firm size, firm performance and industry are always tied to CEO compensation. Hence, I have to control for variables representing these economic factors. Firm size is measured by firm sales (SALES). Firm performance is measured by return on equity (ROE). It reveals how much profit a company earned in comparison to the total amount of shareholder equity found on the balance sheet. So it is calculated by dividing net income by shareholders’ equity. In addition, annual bonus is paid based on a single-year’s performance (Murphy, 1999). And the compensation package shows a lag effect that compensation in 2007 is determined by results of firm performance over 2006. Although there are other measures for CEO performance, I assume that ROE is a good representative and that all performance measures are correlated with ROE.

Therefore, ROE in 2006 is collected. I expect that CEO and board executive directors in larger firm get higher compensation, α6 > 0 and β6 >0. And I also expect that it will have similar effect of firm performance, that is to say α7 > 0 and β7 >0.

Since performance may vary across industries, dummy variable are used to control each of these industries (IND). DUMc coded 1 if firm belongs to manufacturing, 0 otherwise. DUMi coded 1 if firm belongs to financial industry, 0 otherwise. DUMh coded 1 if firm belongs to wholesaling &

retailing industry, 0 otherwise. Setting dummies for industry like that will cause perfect multicollinearity, but it can be solved by deleting the constant term from the regression equation. Thus the research models will change by dropping constant terms of each equation - α ₀ and β ₀ (Baum, Christopher F., 2006).

CEOPAY i =α 1 CC i +α 2 FCC+α 3 SOE i +α 4 FOR i +α 5 FSEL i +α 6 SALES i +α 7 ROE t-1 , I

+α 8 IND i +ε i

(1)

ASD _i =β ₁ CC _i +β ₂ FCC+β ₃ SOE _i +β ₄ FOR _i +β ₅ FSEL _i +β ₆ SALES _i +β ₇ ROE _t-1 , _i

+β ₈ IND _i +ε _i (2)

6 666 ANALYSIS ANALYSIS ANALYSIS ANALYSIS AND AND AND AND RESULTS RESULTS RESULTS RESULTS

Two equations used in this study. One is to test the relation between CEO compensation and average salary level of compensation committee. Another is to examine the linkage between the average salary level of executive directors and compensation committee. Both equations included full set of control variables and ownership factors. And I relied on OLS procedure to analyze.

Table1 reports the overall description of variables. Although in the rank of ten larger shareholders in firms’ annual report the largest shareholder is a SOE, in most cases the ultimate control is the state. In my sample, 78% firms are ultimately controlled by the state. When the largest shareholder is SOEs, it owned 41.5% shares averagely. And over half of firms own foreign shares.

48% firms listed outside domestic stock market. But a small proportion of firms have foreign members in compensation committee.

Insert table 1 about here

Table2 reveals the regression results of two equations. The coefficient of SOE in equation 1 was negative and significant in 10% level that indicates the less percentage of shares owned by SOEs the more compensation will be paid to CEO. Moreover, 1% increases in state-owned shares only lead to about 0.23%

decrease in CEO pay. But this variable was insignificantly correlated with the

average salary level of executive directors. The outcome also gave no evidence of that the presence of foreign shares increases the compensation of CEO and average salary level of executive directors. Because the sign of coefficients of FOR in two equations is negative.

The results showed a very weak effect of compensation committee. As table 2 illustrated, coefficients of CC in two equations were positive and significant.

It appears that if the compensation committee receives a higher pay in a firm, then the CEO and executive directors in that firm will be paid higher. But the outcome also suggested that for an increment of 1,000 CNY (Chinese Yuan) per year in the average salary of compensation committee, the expected compensation of CEO will only raise by 1.26 CNY. Similarly, an extra 1,000 CNY in average salary of compensation committee was associated with an increase in the average salary of executive directors of 1.47 CNY.

Unfortunately, foreign membership in compensation committee showed an opposite way of my hypothesis 4a and 4b, where the coefficients were negative and statistically insignificant. The presence of foreign membership has no effect on compensation level of CEO and executive directors. Moreover,

the internalization factor (FSEL) supported my hypothesis 5a and 5b. CEOs in firms which also listed in foreign stock market receive a higher compensation.

With regard to economic control variables, all of them showed robust and

significant results except sales. Return to equity (ROE) of 2006 was positively

and significantly related to CEO compensation and average salary level of

executive directors indicate that firms which had a good performance in prior

year will pay more to CEO and executive directors in next year. So

compensation is significantly sensitive to firm performance. Industry

influence also significantly and positively related to CEO compensation and

average salary level of executive directors. All of three industries

(manufacturing, financing & insurance and wholesaling & retailing) positively correlated with CEO compensation and average salary level of executive directors. And, firms which belong to financing & insurance will pay higher to CEO rather than manufacturing and wholesaling & retailing industry. But this study didn’t demonstrate any relation between firm size and compensation level.

Note that no outcomes of F-test for two equations were reported. In the econometric software (EViews), It would not report F-statistic if the OLS regression do not include a constant term, because it's invalid (according to a content of online econometrics textbook).

Insert Table 2 about here

7 777 Discussion Discussion Discussion Discussion and and and and conclusion conclusion conclusion conclusion

As the results showed, the percentage of state-owned shares was negatively related to CEO compensation, meanwhile the influence of the presence of foreign shares didn’t reveal positive correlation with CEO compensation as I expected. It suggested that state-owned shares weaken the link between compensation committee and CEO compensation. But according to a research of performance of listed firms in 2007, compensation level of CEOs and top executives in large SOEs exceed the compensation level in private enterprises.

Press and the public though it is unjustifiable that such high compensation

was paid to CEO in these large SOEs even in nonprofit year. For instance, the

CEO of PingAn Insurance-Ma Mingzhe-received 25,794 thousand CNY after

taxes and donation and his compensation was raised 92.75% by comparing

last year. Whereas the stock price of the company slumped by 93.22 CNY on

stock market. But the CEO of PingAn argued that they employed two

different systems for designing compensation. One is adopted for their foreign employees and executive directors to accommodate an international level of compensation. The other is adopted to adjust the domestic compensation level for the rest of employees and directors. Therefore, with the reform and the study a western way to running a company, SOEs is developing and growing well. It is probably an advantage that firms are under the control of government.

In a modern company which separates the control and ownership, the role of compensation committee is to design the appropriate level of compensation, while the board of directors is the ultimate authority to approve the compensation plan. Besides, the board of directors has the power to assess whether the effort and actions of CEO aimed maximize the shareholder’s interests and they often subjected to government influence due to ownership.

As Core et al. (1999) addressed, excessive CEO compensation is associated with relatively weaker corporate governance. Thus, corporate governance in SOEs is supposed to be improved.

My finding manifested the influence of compensation committee on setting the compensation level of CEO and executive directors. Although this influence is weak, it still suggested the important role of compensation committee. However, a few reports indicate that the members in compensation committee are not specialist in designing compensation, sometimes they lack of basic knowledge in this filed and unfamiliar with the board member for whom they need to design the salary level. Therefore, it is priority that authorities optimize the composition and function of compensation committee, so that the CEO and board directors can receive a reasonable compensation.

An alternative interpretation for this weak influence is the overlap between

the mechanism of independent directors and supervisory board due to their similar role of supervision. Shao Shaoming (2004) studied the relation among independent director, board composition and ownership structure of listed firms in Zhejiang province. They argued that the introduction of independent director into China’s board system brings about the conflict with the existing supervisory board system and meanwhile a high proportion of state-owned shares constrain the role of independent director.

The results also indicated a high association between firm performance and the compensation level of CEO and board executive directors. That is to say, in some degree, it is also reasonable to depend on the firm performance to setting the compensation level. No matter to the public, investors, the board of directors or the CEO, they thought it is reasonable if the compensation of CEO and executives increases with good firm performance and decrease with firm losses. This suggested that the transparency of compensation disclosure should put on the agenda.

As regards the industry factor, a boom in financing & insurance industry can explain a higher compensation level in this industry. The boom caused a rat race between firms to attract and to keep competent people by using attractive compensation and incentive plan.

This study exist several limitations. First of all, by comparing to the China

capital market, my research sample is quite small. And neither firms without

disclosure of compensation of CEO and board executive director, nor firms

without compensation committee are selected. Thus it is possible that

unselected firms may associate with CEO compensation by hindering vital

information to which lead selection bias. Secondly, unlike western countries,

the released compensation of CEO and executive directors presents in a total

number in firm’s annual report, it’s not broken down into base salary, annual

bonuses, stock-options and other forms of compensation. It is a shortcoming that considers total cash compensation (base salary and annual bonuses), because it can not provide a completely research results of the role of compensation committee in setting CEO compensation. If the compensation disclosure will be more transparent, then using stock option or long-term incentive plan may get more detailed information. Last but not least, the results may differ with the different index for one factor. For instance, using return on asset (ROA), return on equity (ROE) and stock return to proxy firm performance will obtain different outcomes.

For the further expansion, I plan to use a panel database, collecting changed

CEO compensation and the compensation paid to compensation committee in

3 or 5 years from more than 50 firms which I hope to make my results more

precise. Moreover, an interaction between ownership and firm performance

should be included to better control for the effect of performance. Besides,

when I look at the article by Ozerturk (2005), it is interesting that they use

monitor intensity from the board to the CEO as a measure of effective

corporate governance. They argued, before offering an incentive contract, the

board of directors engages in monitoring the CEO and observes an

information signal correlated with CEO’s unknown ability. They provide an

alternative way to test the role of the board of directors on CEO compensation.

Appendix

Appendix Appendix Appendix A A A A:::: List of selected firms China Pertroleum & Chenmical Corporation

PetroChina Company Limited ³

Industrial and Commercial Bank of China Limited China Mobile（HK） Limited

Bank of China Limited

China Life Insurance Company Limited China Telecom Corporation Limited China Construction Bank Corporation Lenovo Group

China COSCO Holdings Company Limited Ping An of China

Gome Electrical Appliances Holding Limited CITIC Securities Company Limited

Suning Appliance Co., Limited

Bank of Communications Co., Limited

GuangDong Midea Electric Appliances Co., Limited China International Marin Containers (Group) CO., LTD Shanghai Pudong development Bank Co., Limited Wumart Stores, Inc.

Zhejiang Orient Holdings CO., Limited SAIC Motor Corporation Limited

Brilliance China Automotive Holdings Limited ZheJiang Zhongda Group Co., Limited

Wuhan Zhongbai Group Co., Limited

China Netcom Group Corporation (Hong Kong) Limited ZheJiang Holley Technolgy Co., Limited

Hua Xia Bank Co., Limited

Jiansu Guotai International Group Guomao Co. Limited JLF Investment Company Limited

Intime Department Store Company Limited CNOOC Limited

Sichuan Changhong Electrtic Co., Limited Bosideng International Holdings Limited

3 In the Company Law (revised in 2005), the term "company" as mentioned refers to a limited liability company or a

The term "company" as mentioned in this Law refers to a limited liability company or a joint stock limited company

established within the territory of the People's Republic of China in accordance with the provisions of this law. For a

limited liability company, a shareholder shall be liable for the company to the extent of the capital contributions it

has paid. And a limited liability company shall be established by no more than 50 shareholders that make capital

contributions. For a joint stock limited company, a shareholder shall be liable for the company to the extent of the

shares it has subscribed to.

Industrial Bank Co, Limited

China Resources Enterprises Limited Jiangling Motors Corporation, Limited Qingdao Aucma Company Limited

Baolilai Investment CO., Limited，Guangdong Nanjing Texitiles Import & Export Co. Limited Haima Investment Group Co. Limited

Beijing Wangfujing Department Store Co. Limited Chengshang Group CO. Limited

Tsingtao Brewery Company Limited Shanghai Highly Co. Limited

Sichuan Hongda Co. Limited

Hangzhou Silan Microelectronic Co. Limited Dashang Group Co. Limited

Xiamen ITG Group Co. Limited Jiangsu Skyrun Corporation

Jiangsu Guotai International Group Guomei Co. Limited

Xinjiang Youhao (Group) Co., Limited

Appendix Appendix Appendix Appendix B B B B

Table Table Table Table 1 111 CEOPAY

(thousand CNY)

ABC (thousand

CNY)

CC (thousand

CNY)

SOE (%)

ROE (%)

SALES (thousand

CNY) Mean 2244.640 1813.036 169.8400 0.415085 0.131968 14822542 Median 879.0000 733.0000 99.00000 0.353000 0.130400 822996.5 Maximum 46161.00 39767.33 965.0000 0.862900 0.424700 1.43E+08 Minimum 20.00000 64.00000 12.00000 0.084300 -0.531100 -559011.0 Std. Dev. 6468.452 5533.752 179.9285 0.210762 0.135316 29200723 Sum 112232.0 90651.81 8492.000 16.18830 6.598400 7.41E+08 Sum Sq.

Dev. 2.05E+09 1.50E+09 1586339. 1.687976 0.897211 4.18E+16 Observatio

ns 50 50 50 39 50 50

Table Table Table Table 2 222

* p<=10%

** p<=5%

***p<=1%

Equation 1 Equation 2

CEO PAY Average

directors pay

State-owned shares -0.225856* 0.755290

The presence of foreign shares -0.283452 -0.167163*

Average salary level of compensation

committee 0.001264** 0.001476*

Foreign membership in compensation

committee -0.426278 -0.653524

Foreign stock exchange listing 0.763781* 0.827445**

ROE 4.570513*** 2.906004***

Sales -5.55E-09 -8.43E-09

DUMc (manufacturing) 5.854997*** 5.335003***

DUMi (financing & insurance) 6.979216*** 6.895838***

DUMh (wholesaling & retailing) 5.854997*** 5.600297***

R-squared 0.578277 0.573017

Adjusted R-squared 0.447398 0.440505

R R

R REFERENCES EFERENCES EFERENCES EFERENCES

Abowd, J. M. 1990. Does performance-based managerial compensation affect corporate performance?. Industrial Industrial Industrial Industrial and and and and Labor Labor Labor Labor Relations Relations Relations Relations Review Review Review Review , 43:52-73.

Anderson, M. C., Banker R. D. and Ravindran S. 2000. Executive compensation in information technology industry. Management Management Management Management Science, Science, Science, Science, 46:1411–1438.

Baek, H. Y. and Pagan, J. A. 2006. Pay-performance Sensitivity and High Performing Firms. International International International International Research Research Research Research Journal Journal Journal Journal of of of of Finance Finance Finance Finance and and and and Economics Economics Economics Economics , 5:78-94.

Baker,G. P., Jensen, M. C. and Murphy K. J. 1988. Compensation and Incentives: Practice vs. Theory. The The The The Journal Journal Journal Journal of of of of Finance Finance Finance Finance, 43(3): 593-616.

Baum, Christopher F. 2006. An An An An Introduction Introduction Introduction Introduction to to to to Modern Modern Modern Modern Econometrics Econometrics Econometrics Econometrics Using Using Using Using Stata

Stata

Stata Stata. Stata Press

Bebchuk, L. and Fried, J.2004. Pay Pay Pay Pay without without without without performance: performance: performance: performance: The The The The unfilled unfilled unfilled unfilled promise

promise

promise promise of of of of executive executive executive executive compensation compensation compensation compensation. Cambridge, MA: Harvard University Press.

Belliveau, M. A., O'Reilly III, C. A. and Wade, J. B. 1996. Social Capital at the Top: Effects of Social Similarity and Status on CEO Compensation. The The The The Academy

Academy

Academy Academy of of of of Management Management Management Management Journal Journal Journal Journal, 39(6): 1568-1593.

Berle, A. A. and Means, G. C. 1932. The The The The Modern Modern Modern Modern Corporation Corporation Corporation Corporation and and and and Private Private Private Private Property

Property

Property Property. NY: Macmillan Company.

Crystal, G. S. 1991. In In In In Search Search Search Search of of of of Excess: Excess: Excess: Excess: The The The The Overcompensation Overcompensation Overcompensation Overcompensation of of of of American American American American Executives.

Executives.

Executives. Executives. New York: Norton.

Doidge,C., Karoly, A. and Stulz, R. 2004. Why are foreign companies listed in the US worth more?. Journal Journal Journal Journal of of of of Financial Financial Financial Financial Economics, Economics, Economics, Economics, 71: 205-238.

Fama, E. 1980. Agency problems and the theory of the firm. Journal Journal Journal Journal of of of of Political

Political

Political Political Economy, Economy, Economy, Economy, 88: 288-307.

Festinger, L. 1954. A theory of social comparison processes. Human Human Human Human Relations, Relations, Relations, Relations,

7: 117-140.

Firth, M. and Fung, Peter M.Y. and Rui, Oliver M. 2006. Corporate performance and CEO compensation in China. Journal Journal Journal Journal of of of of Corporate Corporate Corporate Corporate Finance, Finance, Finance, Finance, 12: 693– 714.

Firth, M. and Fung, Peter M.Y. and Rui, Oliver M. 2007. Ownership, two-tier board structure, and the informativeness of earnings -Evidence from China.

Journal Journal

Journal Journal of of of of Accounting Accounting Accounting Accounting and and and and Public Public Public Public Policy, Policy, Policy, Policy, 26: 463–496.

Firth, M., Tam, M., and Tang, M. 1999. The Determinants of Top Management Pay. Omega Omega Omega Omega, 27: 613-635.

Gibbons G. and Murphy K. J. 1989. Relative Relative Relative Relative Performance Performance Performance Performance Evaluation Evaluation Evaluation Evaluation for for for for Chief

Chief

Chief Chief Executives Executives Executives Executives Officers Officers Officers Officers. Working paper no. 248, Industrial Relations Section Princeton University,.

Gouldner, A.W. 1960. The norm of reciprocity. American American American American Sociological Sociological Sociological Sociological Review

Review

Review Review, 25: 161-178.

Hogan, T. D. and McPheters, L. R. 1980. Executive Compensation:

Performance versus Personal Characteristics. Southern Southern Southern Southern Economic Economic Economic Economic Journal Journal Journal Journal, 46(4): 1060-1068.

Holmstrom, B. 1982. Moral Hazard in Teams. Bell Bell Bell Bell Journal Journal Journal Journal of of of of Economics Economics Economics Economics 13:

324-340.

Jensen, M. C. and Murphy, K. J. 1990. CEO Incentives—It’s Not How Much You Pay, But How. Harvard Harvard Harvard Harvard Business Business Business Business Review, Review, Review, Review, (3): 138-153.

Jensen, M. C. and Murphy, K. J. 1990. Performance Pay and Top-Management Incentives. The The The The Journal Journal Journal Journal of of of of Political Political Political Political Economy Economy Economy Economy, 98(2): 225-264.

Jensen, M. C. and Meckling, W. 1976. Theory of the firm: managerial behavior, agency costs, and ownership structure. Journal Journal Journal Journal of of of of Financial Financial Financial Financial Economics, Economics, Economics, Economics, 3:

305- 360.

Joskow, P. L. and Rose, N. L. 1994. CEO CEO CEO CEO pay pay pay pay and and and and Firm Firm Firm Firm Performance: Performance: Performance: Performance:

Dynamics, Dynamics,

Dynamics, Dynamics, Asymmetries, Asymmetries, Asymmetries, Asymmetries, and and and and Alternative Alternative Alternative Alternative Performance Performance Performance Performance Measures Measures Measures Measures. Working Paper, No. 4976, NBER.

Kato, T., Kim, W. and Lee, J. H. L. 2004. Executive Executive Executive Executive Compensation Compensation Compensation Compensation and and and and Firm Firm Firm Firm Performance

Performance

Performance Performance in in in in Korea. Korea. Korea. Korea. Working Paper, 04-01, KDI School of Public Policy and Management.

Kato, T. and Kubo K. 2006. CEO compensation and firm performance in Japan:

Evidence from new panel data on individual CEO pay. Journal Journal Journal Journal of of of of Japanese Japanese Japanese Japanese international

international

international international Economies, Economies, Economies, Economies, 20: 1–19.

Kato T. and Long, C. 2004. Executive Compensation, Firm Performance, and Corporate Goverance in China: Evidence from Firms Listed in Shanghai and Shenzhen Stock Exchanges

(http://www.aeaweb.org/annual_mtg_papers/2005/0107_1015_0605.pdf) Khan, R. and Dharwadkar, R. and Brandes, P. 2005. Institutional ownership and CEO compensation: a longitudinal examination. Journal Journal Journal Journal of of of of Business Business Business Business Research,

Research,

Research, Research, 58: 1078– 1088.

Ke, B., Petroni K. and Safieddine A. 1999. Ownership concentration and sensitivity of executive pay to accounting performance measures: evidence from publicly and privately-held insurance companies. Journal Journal Journal Journal of of of of Accounting Accounting Accounting Accounting and and

and and Economics Economics Economics Economics 28: 185–209.

La Porta, R., Florencio Lopez-de-Silanes and Shleifer, A. 1998. Corporate Corporate Corporate Corporate Ownership

Ownership

Ownership Ownership Around Around Around Around the the the the World World World World. Working Paper no.W6625, National Bureau of Economic Research.

Lambert R .A. and Larcker D. F. 1987. An analysis of the use of accounting and market measures of performance in executive compensation contracts . Journal

Journal

Journal Journal of of of of Accounting Accounting Accounting Accounting Research Research Research Research, 25: 85-125.

Larcker, D. F. 1983. The Association between Performance Plan Adoption and Corporate Capital Investment . Journal Journal Journal Journal of of of of Accounting Accounting Accounting Accounting and and and and Economics Economics Economics Economics 5: 3-30.

Lazear, E. P. 1989. Pay Equality and Industrial Politics. The The The The Journal Journal Journal Journal of of of of Political

Political

Political Political Economy, Economy, Economy, Economy, 97(3): 561-580.

Li Huitian. 2001. Government Government Government Government Shareholding Shareholding Shareholding Shareholding and and and and the the the the Value Value Value Value of of of of China's China's China's China's Modern

Modern

Modern Modern Firms Firms Firms Firms. Working paper no. 395, William Davidson Institute.

Lorsch, J. 1989. Pawns Pawns Pawns Pawns or or or or potentates: potentates: potentates: potentates: The The The The reality reality reality reality of of of of America's America's America's America's boards boards boards boards.

Harvard Business School Press.

Lublin, J. 1991. While outside directors' pay increases, independence from managers may fade. Wall Street Journal, April 22: B1.

Milbourn, T. T. 1997. The The The The executive executive executive executive compensation compensation compensation compensation puzzle: puzzle: puzzle: puzzle: theory theory theory theory and and and and evidence.

evidence.

evidence. evidence. Working paper, London business school.

Murphy, K. J. 1985. Corporate Performance and Managerial Remuneration, an