Master Thesis Corporate Value, Managers Proportion of Shareholdings and CEO Remuneration: The Case of Chinese listed Firms

Master Thesis

Corporate Value, Managers Proportion of Shareholdings

and CEO Remuneration:

The Case of Chinese listed Firms

April 2009

University of Groningen Faculty of Economics and Business MSc Business Administration – Finance

Abstract

Berle and Means (1932)’s Ownership Dispersion hypothesis suggests that CEOs’ compensation plans should be in line with the shareholders’ benefits. Fama and Jensen (1983) propose a manager shareholdings proportion incentive system on solving the manager self-interested behaviors which implies that CEOs’ remuneration policies should track their way based on managers’ proportion of shareholdings aspect. Furthermore, according to previous studies, corporate value, managers’ proportion of shareholdings and CEOs’ remuneration are mutually determined. We test for this mutual dependence by applying the two stage least square regression model. We collect our input data from the annual reports of Chinese firms listed on the Shenzhen stock market for the years 2006 and 2007. Our empirical results verify that corporate value and managers’ proportion of shareholdings have both a positive and significant influence on CEO remuneration in Chinese firms. In addition, we discover that most Chinese managers do not hold any firm shares in their personal portfolios.

JEL Classification: G32, G34

Preface

Throughout the last year I have been studying the MSc Finance at the University of Groningen. For the last part of this master I have been writing a master thesis. The subject of my master thesis is the Corporate Value, Managers Proportion of Shareholdings and CEOs Remuneration: A Case of Chinese listed Firms in Shenzhen Stock Market. During the process of writing my thesis several people helped me a lot. I would like to thank them for their support. In the first place, my supervisor, Dr. Lammertjan Dam, who is always willing to give me suggestions and gives useful insights in the subject of my thesis. And I also want to thank my second supervisor, Drs. Katzur, who is justice and honestly assess my thesis. Last but not less, I want to thank my father Xin Xu, as a businessman, provided me with real-life based advice that ensured that theory and practice stayed in touch.

Table of Contents

Introduction………5

Literature and Hypothesis………7

Methodology………..11

Data Description and Analysis………18

Results………23

Conclusions………33

Reference………...35

I Introduction

Fast economy development and industrial innovation in recent decade lead companies to extend in an amazing speed, and Chinese firms have no exclusion. The fast growing firms need to rely on CEOs’ experience and know-how to cope with the increasingly larger business. And the compensation contracts play an important role in motive the CEOs’ behaviors in operating the firms. Thus, the design on CEO remuneration structure is becoming a grabbing attention corporate organization topics discussing at all time. However, as some global famous

companies like Enron, Tyco, WorldCom and Merck collapse with financial frauds in 2002, the information asymmetric conflicts between external investors and internal stakeholders are more serious, leads the firms’ principle of good faith to be widely questioned (Bidwai, 2002). Investors always think about the following issues before they invest into a specific company. First and foremost, are the CEOs worth their high compensations? Moreover, whether the CEOs are efficiently operating the firms’ assets and create value for their shareholders. Last but not least, managers may attempt to maximum their own benefits by scarifying

shareholders’ benefits.However, most studies of the above considerations are focus on developed countries. Thus, in this paper, we research the above arguments based on Chinese listed firms.

The remuneration of Chinese corporations’ CEOs used to be a sensitive topic before 2006. It was especially a sensitive topic for Remuneration of CEOs in State-owned or State-controlled corporations. The implication is that external investors have difficulty to access this kind of relevant information before 2006.

In 2006, however, a new “Company Law of China” was introduced. This new law brings in mandatory requirement on disclosing CEOs’ remuneration details in annual reports for all the listed firms in Chinese stock market.

This new availability of data creates opportunities to further study topics which are associated with Chinese CEOs’ remunerations (WEI, 2000 and Song, 2002 and Hu, 2005).

At the beginning of the Chinese economic opening up in 1980’s, the setup of managers’ remuneration was always related to their administration level or craft peers’ level (Yang, Toms and Yurtoglu, 2004). The remuneration of managers in stated-owned or stated-controlled corporations is about three times that of non-skilled workers. Although this multiple is enlarged to seven times in recent years, and the level of remuneration increase is based on the economic conditions of the country (Liu and Otsuka, 2004), it is still far from the multiple level of its counterparties, like Japan, US and most western European countries. However, Li and Wang (2007) demonstrates that, as the value of firms grows, the managers still receive compensation contracts which are not properly in line with the economic conditions of the country. Thus, we attempt to discover some forms of variables that are related to the increment on the remuneration of managers and verify the results by empirical analysis in this paper.

Managers are proxies for the shareholders in operating the investments. If the managers do not hold any shares in firms, it is difficult to align their benefits with shareholders. However, contrary to this point of view, we find that most Chinese managers are usually not holding any shares in their firms. According to the statistical results of Zhang (2002), there are 864 firms listed on Shanghai Stock Exchange, and the CEOs of 454 firms do not hold any shares in their firms, which captures 52.57% of the total sample. Firms offer shares to managers in order to align their benefits with shareholders’. However, if the managers do not hold any shares in firms, they do not need to take any responsibility to the performance of the firms and the task of creating value for their shareholders. That is another important potential problem that we want to investigate in this paper.

are positively correlated with stock returns. Simultaneously, Gregg, Machin and Szymanski (1992) testify that there is a positive relationship between the top sum of CEOs remunerations and bonus, and stock returns of 288 British companies from 1983 to 1991. Conyon and Leech (1994) separately study the relationship between the CEOs compensation package and the returns of stock of UK companies into different periods, and the results of different periods all yield the same positive relationship between the two objects. Generally speaking, due to the fact that different researchers select different samples with different assessment standards, the conclusions yield different results, but there is some consensus that corporate performance has a positive relationship with CEOs’ remuneration. However, few literature attempts to study the relationship between CEOs’ remunerations and the corporate performance at developing countries level. Thus, in this paper, we use the Chinese listed firms’ performance as our observation to test whether the relationship between corporate performance and CEOs’ remuneration still keep their track in a developing country sample.

The following section of this paper refers to some prior researches to motivate my purpose on writing this paper and develop my empirical hypotheses. We describe the variables and generate the regression models in Section 3. And we present my data collection and statistics description in section 4. In the section 5, we report and analyze the statistical results from my regression models. The last section summarizes the findings, and gives advices on future studies in this dimension.

II Literature and Hypothesis

A large and growing body of prior studies is devoted to the contradiction between

control rights, reversely, limits the shareholders to capture their actual benefits. Thus, in order to mitigate this agency problem in corporate governance, relevant monitoring mechanisms should be imported into the compensation contracts for CEOs (Frank, 2006).

Along with the size expand, firms need to further rely on managers to manage corporate operating and make future investment strategies decisions. Since the managers need to take the biggest responsibility on operating the whole company, and their investment decision making errors may lead to serious loss on firms. Therefore, comparing to ordinary staffs, firms need to compensate the managers with higher remunerations in order to prevent them from tracing their own benefits at shareholder expense. Nevertheless, firms are desired to pay higher compensations to the CEOs, who lead the companies’ investment strategies planning and operating development, is because they wish the excellent managers can friendly use their experience and management skills to create future growth opportunities and value for them.

In a contrary point of view, Berle and Means (1932) propose an Ownership Dispersion hypothesis, and claim that after a company has separated management decision from ownership, more ownership rights can distribute to minority shareholders’ hands. But, simultaneously, managers might formulate remuneration contracts grounded on their self-interested in order to maximize their wealth. Then, they take their charges against the companies’ investment strategies, as a result, the operating performance fail to meet the target of maximum corporate wealth and value and expectation of shareholders.

In light of the above issues, it is necessary for the board of directors to build up a suit of consummate and efficient remuneration institution, which can motivate managers to genuinely serve their firms and allow managers to pursue their personal benefits together with optimal benefits of shareholders, then, conducing their benefits identical to corporate goals.

projects rather than the investments that can increase the shareholders’ interests. Therefore, in order to array managers’ behaviors in line with the principle of maximize shareholder value, the board of directors set up supervisory mechanisms and formulate compensation contracts to standardize their behaviors.

On the other hand, firms use enviable high compensation to motivate the managers efficiently managing the assets of firms and operating the firms, creating outstanding

accomplishments in the competitive market. Nevertheless, contingent on the favorable rewards derived from the proud performance, managers have more incentive to create higher value for the firms. Therefore, the compensation of managers and corporate value are supplementary to each other. Only when the managers master professional knowledge, besides being familiar with the procedures of corporate operating, can they take up significant influence on the board of directors. However, contrary to this point of view, the managers intend to design their favorite remuneration contracts through the control of board of directors, and ultimately hurt the substantial benefit of the shareholders. At the practical level, Mehran (1995) believes that managers’ remunerations are determined by board of directors. Speaking for shareholders, they wish to maximize the corporate value and drive the managers in line with their targets by paying managers with proper remunerations. At the corporate value level, in term of Chung and Pruitt (1996), William (2001), managers should implement most valuable investment strategy in the firms’ shoes, then, advance their compensations. Kerr and Bettis (1987) extend the notion in a reversing side, they demonstrate that the change in corporate value is not relative to the change of manages’ remunerations. The above arguments raise my first motivation on exploring whether the remunerations of managers are influenced by the corporate value.

the rights the more advantageous for them to determine the combination and contents in their compensations, thus, the shareholdings proportion of managers has a positive relationship with the remuneration of managers. However, Mehran (1995) validates that an increment on

shareholdings proportion of managers reduces the board of directors’ monitors on them. Thus, in order to obtain higher compensation from the corporate stocks, managers can increase the net profit of the firms by reducing their fixed salaries. And the above focal points form my second research motivation in this paper.

Coughlan and Schmidt (1985), Galbraith and Merrill (1991), Singh and Harianto (1989), and Ryan and Wiggins (2001) discuss the influence of shareholdings proportion of managers or the corporate value on CEOs’ remuneration, respectively. Our third motivation is based on

eliminating the endogenous problems in prior relevant researches and studying the managers’ remuneration problem by joining the managers’ shareholdings and corporate value together.

2.2 Hypothesis development

2.2.1 Corporate Value to Increase Remuneration of CEOs

For many firms, scholars use Tobin’s Q as one of the most important variables for testing corporate value and whether managers’ investment strategies can bring in profits to companies. Thus, managers intend to use their management know-how to efficiently manage current

resources in place for the improvement of profitability in future investment plans. Along with the change in economic circumstance, managers are able to explore better investment strategies and growth opportunity by relying upon their cumulative experience and management skills, and create higher value for firms as well. Therefore, it is reasonable for companies to create their value by offering more favorable compensation contracts to CEOs. Mehran (1995), Chung and Pruitt (1996) and William (2001) also assert that the managers’ remuneration is raised with the increment of corporate value. In term of this, we build up our first hypothesis as follow:

2.2.2 Managers’ Proportion of Shareholdings to increase Remuneration of CEOs The managers’ proportion of shareholdings stands for a certain degree of their influence in firms, especially in board of directors. The larger the ownership owned by managers, the bigger their impact on the investment decisions in the firms. As a result, managers attempt to control the board of directors to provide them with better compensation contracts and increase their personal wealth by holding more shares in hands. Ungson and Streets (1984) and Core et al. (1999) illustrate that it is much easier for managers to enhance their decision on their compensation contracts by holding a higher proportion of shares in firms. According to the above evidence, we generate my second hypothesis:

Hypothesis 2: Managers’ Proportion of Shareholdings has a positive relationship with CEOs’ Remuneration.

III Methodology

3.1 Definition of variables

Based on the previous relevant studies, we use the Pearson test to avoid the multicollearity problem among variables and increase the reliability of the empirical test results. And we give out the definitions of the variables that we use to build up our research models in table 1.

Table 1 Definition of Variables

This table gives out definition and calculation procedure for the variables. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities Shareholdings (BES) and Leverage (LV).

CR TQ MPS Control Variable: SIZE ROA TERM BSIZE LEV BES

Salary + Rewards + Traffic Allowance + Extraneous Charges + Bonus Pool (Equity Market Value + Liabilities Book Value) / Replacement Cost of Capital

Managers Shareholdings / Outstanding Stock

Logarithm of Total Asset EBITDA / Total Asset Accumulative Terms of Managers Number of members in Board of Directors

Debt/ (Debt + Equity)

Business Entities’ Shareholdings/ Outstanding Stock

3.2 Control variables

The remaining control variables in this paper are not only capture the other most possible information that has signaled the conditions of the CEOs’ remuneration, but some of them also have significant influence on the two main independent variables TQ and MPS on my main arguments. These variables include firm size, size of board of directors, ROA, managers’ terms, growth of total assets and leverage.

SIZE

Chung and Pruitt (1996) convince that the size of corporation can be treated as a representative variable for managers’ task complexity, the expansion of corporate

organizational structure and additional complexity of works can lead to higher requirement on managers’ professional skills and responsibility. As the firms need to finance the managers’ professional skills with proper remunerations, hence, the expected task complexity and importance of managers have significant impact on managers’ remuneration. Due to the fact that managers’ task complexity is highly correlated with firm sizes, the size of the firms should have a positive relationship with managers’ remuneration.

changes of the industry in the future. However, when the sizes of firms remain in a small scale, it prevent them from rising in their profitability, dealing with the transitional economical environments and competing in a fast growing trend in globalization. Thus, the small size firms are more likely to expand their sizes and enlarge their market shares by mergers or acquisitions. Demsetz and Lehn (1985) demonstrate that the managers can capture more abundant resource and information to make efficient investment decisions, which maximize corporate value, by further enlarging their firm sizes.

At the Managers’ proportion of shareholdings level, the managers need to continuously invest marginal capital into their firms’ stocks as the sizes of firms are expanding, when they have a mind to keep a sustainable proportion of shareholdings. However, this kind of

investment activity is violated from the principle of risk diversification and brings additional cost to their increment of wealth. Thus, the proportion of shareholdings is likely to decrease with the expansion of the firms.

Business Entities Shareholdings

Pound (1988) firstly brings in three important hypotheses that broaden the horizon of agency theories. The first one is the Efficient Monitoring Hypothesis, which illustrates a picture that the institute investors, who are mostly the state in China, are easier to access to information of the firms and possess professional knowledge than common investors. Thus, managers can be more efficiently monitored by institute investors and give rise to the value of the firms.

Secondly, the Conflict of Interest Hypothesis assumes that when the institute investors become the majority shareholders of the firms, they are prone to let the managers to focus on the investment strategies in line with their benefits. However, their benefits may be conflict to the benefits of the firms and leads to reduction on firms’ value. Finally, the Strategy Alignment Hypothesis, which predicts that the institute investors can build up a cooperation relationship with the managers. Thus, the benefits of both parties can be flowed into a stream, and the operating of firms should be more efficient, then, it gives rise to the corporate value.

that the increment of business entities’ shareholdings can better monitor the managers’ inaction and self-interested behaviors. Reversely, the more intensely monitoring will naturally reduce the managers’ aspiration on holding more shares in firms.

Return on Assets

Agency problems are a result of the separation between management and ownership. Hence, corporations adopt the Pay-for performance program to design their remuneration mechanism to mitigate the conflict between shareholders and managers. Thus, managers are willing to seek their personal optimal utility to accompany with shareholders’ wealth maximum, and drive their benefit in line with the shareholders. Banker and Datar (1989) testify that firms utilize the operating performance as a benchmark to formulate their managers’ remuneration contracts by evaluating the contribution of managers to the firms. The theories in financial statements analysis consider return on total asset (ROA) as the aggregative indicator for evaluate the performance of corporate operation, as a result, the ROA can proxy for profitability. Thus, ROA is considered to have a correlation with CEOs’ remuneration level.

TERM

Managers need to aggregate their operating rights with a certain period of time, besides, as the growth of their serving time in the companies, they accumulate abundant experience and management skills, then, gradually control the firms and formulate self-advantage

remuneration contracts to themselves through the board of directors. Thus, the operating rights of the managers are correlated with their cumulative terms. Moreover, Cline (1975) attests the fact that the longer the managers serve in the companies, the deeper they get straight into the characteristics of the firms, and become more and more skillful on their works, then, creating higher operating performance for the firms with higher productivities, and in the train of higher remunerations. Nevertheless, some scholars treat managers’ terms as the human capital factor of one of the market portfolio factors. Finkelstein and Hambrick (1989) and Chung and Pruitt (1996) demonstrates that terms of managers are positive correlated with their remunerations.

experience are likely to control the board of directors by holding higher proportion of shareholdings. In light of this, the longer the managers serve in the firms, the higher they willing to hold the corporation shareholdings.

Chung and Pruitt (1996) also verify that the managers are more likely to hold more corporate shares as their terms accumulated. And they become more loyal and sincerely utilize their cumulative management know-how to discover growth opportunity and create value for firms.

BSIZE

Board of directors is one of the most important sectors in the corporation organizational structure, which plays a role of the representative of shareholders to monitor the behavior of managers, and urges the managers to create optimal value for its shareholders. In term of the company law, the assignment, layoff and remuneration of managers should be determined by the board of directors. Thus, the managers can maximize their remuneration by controlling the board of directors. Generally speaking, the corporate governance mechanism is weakened while the size of board of directors is extended and lead to diversification of interests among board members. Thus, the skillful managers can easily get trust from the board of directors, and upgrade their remuneration contracts by controlling the board (Singh and Harianto, 1989).

Board size also has an impact on corporate value. Jensen (1993) points out that the board of directors can monitor the managers’ behaviors more efficiently with smaller size. Besides, the interests among the board members are more diverged while number of directors in board is larger. Thus, managers should operate the firms with more flexibility. Yermack (1996) tests a sample of board size in firms from different industries, and finds that boards of directors with smaller size perform more efficient than larger ones. A larger board of directors reduces

efficiency on decision making in lieu of communicate and balance the benefits among different parties, leads to reduction on corporate value.

usually very small. Hence, we can imagine that the proportion of shareholdings of the manager is tremendous at the beginning, then, as the extending of the board, monitoring on manager’s behaviors becomes more and more strict, and reducing the aspiration for manager to maintain the original proportion of shareholdings in the firm. Thus, the size of board of directors should be correlated with the managers’ proportion of shareholdings (Barnhart and Rosenstein, 1998).

LEV

The leverage level is one of the most important indicators of firms operating conditions. On the one hand, firms can use the tax-saving advantage of debt to finance their investment strategies and create higher value for firms (Koller, Goedhart and Wessels, 2005). On the other hand, the higher the leverage ratio in place, the more active for creditors to monitor the

investment strategies of the firms. However, McConnell and Servaes (1990) testify that the firms should face huge amount of interest costs and higher bankruptcy risk when they finance their investment opportunities by high level of leverage. And they also point out that the firms with more health capital structure sustain a lower level of leverage, thus, can generate higher value for shareholders.

Jensen (1976) figures out that financing with debt proxies for the promise of firms’ future payment. Due to the fact that interest payment can cause a reduction on operating cash flow, managers have fewer resources to invest in profitable projects and lead to a decline on annual net profit. Furthermore, firms are threatening with a higher risk of bankruptcy by higher level of leverage, and managers are likely to reduce their proportion of shareholdings since they have perceived this hazard. Thus, leverage level can substitute for debt financing agency cost and has a negative relationship with managers’ proportion of shareholdings (Crthchley et al., 1989).

3.3 Models Construction

Chung and Pruitt (1996) and William (2001) demonstrate that corporate value has a

Chung and Pruitt (1996) convince that when the managers become the shareholders of

companies, they attempt to promote their proportion of shareholdings in firms, apparently, this behavior mitigates the agency problem and leads to better performance. In opposite to this point of view, Demsetz and Villalonga (2001) demonstrate that corporate value is significant and negative correlated with the managers’ proportion of shareholdings. Ungson and Steers (1984), Finkelstein and Hambrick (1989) and Mehran (1995) believe that managers’

shareholdings behavior has a significant positive relationship with CEOs’ remuneration. In light of the previous studies, I find that there are mutual determination effects among

Managers’ Proportion of Shareholdings, CEOs’ remuneration and Corporation value. In order to test the mutual determination and exclude multicollinearity problems, I use the two stages least square (2SLS) to build up the models in this paper.

Model 1: CEOs’ Remuneration Models

0 1 2 3 4 5 6 7

CR= +

α α

α

α

α

α

α

α

ε

0 1 2 3 4 5 6 7

TQ= +β β ×CR+β ×MPS+β ×SIZE+β ×BSIZE+β ×TERM+β ×LEV+β ×BES+ν

Where CR model denotes the influence of Corporate Value and Managers’ Proportion of Shareholdings on CEOs’ Remuneration, as we mention above, the relationships between TQ and CR, as well as MPS and CR are mutual determinate. Thus, we use the TQ model in model 1 as an instrument to eliminate the possible multicollearity problem on CR model. And we also introduce five control variables into CR model, which are based on theoretical evidences to prevent the statistical results of three main variables from biasing. Based on this point of view, we also introduce the LEV variable into TQ model to further capture accurate results.

Model 2: Corporate Value Models

0 1 2 3 4 5 6 7

TQ= +β β ×CR+β ×MPS+β ×SIZE+β ×BSIZE+β ×TERM+β ×LEV+β ×BES+ν

0 1 2 3 4 5 6

Model 3: Managers’ Proportion of Shareholdings Models

0 1 2 3 4 5 6

MPS= +λ λ ×TQ+λ ×SIZE+λ ×BSIZE+λ ×TERM+λ ×LEV+λ ×BES+ϕ

0 1 2 3 4 5 6 7

TQ= +β β ×CR+β ×MPS+β ×SIZE+β ×BSIZE+β ×TERM+β ×LEV+β ×BES+ν

We use Model 2 and Model 3 to model subtle discussions to our topics. In term of the previous studies, we find out that the Corporate Value also has a relationship with Managers’ Proportion of Shareholdings, in order to illustrate an all-around picture of my general topic, we use MPS model as an instrument to monitor its impact on TQ model and use TQ model as an instrument on MPS model. Based on the purpose of enhancing our accuracy of statistical results on model 2 and 3, control variables are imported as well.

3.4 Robustness Tests for 2SLS Regressions

We validate quality of our empirical results by process two tests: period and panel data econometrics. We generate period robustness tests by dividing the sample into two equally small samples by 2006 and 2007 and regressions of the two models by 2SLS. Moreover, we perform robustness tests by estimating both the Fixed Effects model and Random Effects model to verify the results Period robustness test should refer to table 8 and 9, while panel data robustness test should refer to table 5, 6 and 7 in every column 3 and 4. Besides, in order to test the fixed effect and random effect robustness of the panel data for the three 2SLS models, we drop variables TERM and BSIZE, which are likely to be constant in each companies in each test year and lead to perfect multicollearity problems in our test.

IV Data Description and Analysis

total market value of 2.88 trillion RMB (about EUR 0.33 trillion).

We first randomly pick out 200 companies which belong to different industries in Shenzhen stock exchange. Then, we collect firm-specific financial information, such as CEOs’ remuneration summary, market and book value of equity, book value of debt, total asset, size of board of directors, manager terms, return on weighted average asset and proportion of managers’ shareholdings and Corporate Body Shareholdings, which are correlated to my study from 2004 to 2007. Then according to this small data sample, we do figure out that most of companies do not present most of my tested variable items in their annual report before 2006, which may lead the research for earlier years less comparable. After 2006, due to the new listed criteria of Shenzhen Security Exchange, firms tend to disclosure more and more information, however, the difference between firms is still huge, which make research more meaningful to compare different firms and may lead to stronger conclusions. Hence, in this paper, we only choose information in 2006 and 2007, the most recent years. The database is analyzed separately by applying the two stage regression models which is mentioned in the following part.

Thus, in total, we have 200 firms from different industries excluding financial sector. Each firm will have two annual reports from 2006 and 2007. And the total sample will be 400 panel observations.

4.2 Data Statistics Analysis

4.2.1 Total sample statistics analysis

with different size. Another interesting thing is that the average MPS is close to zero, which means most CEOs in Chinese firms that listed on Shenzhen Stock Exchange are not holding any company’s shares as well as the listed firms in Shanghai Stock Exchange that we have mentioned in the introduction. Comparing to MPS, BES who is proxy for the institute shareholders is far more active to concentrate their shares in a firm and capture an average proportion of 37.5% of shares in each firms. Besides, we want to mention that government in this case is always playing an important role on holding large fraction of shares in firms. Except for MPS, the standard deviation of TQ, ROA are much smaller than other variables for 0.625, 0.306 respectively. Contrary to the Chinese CEOs’ remuneration policy before 2006 discussed in the first section, we can imagine that the Chinese firms are no longer design the remuneration policy just in light of their peers’ since the level of remuneration assessment indicators of TQ, ROA are close in different firms, but the remunerations in different firms are fiercely fluctuating with more factors. The Median of LEV is about 0.11 above its average, which is due to many Chinese firms prefer a stable level of capital structure rather than a dynamic one. Lastly, the period of TERM and the BSIZE in Chinese firms are much more stable on the average level of 3.88 and 9.385, the new Chinese Firms’ Law in 2006 has obliged the firms to maintain a steady managers’ term and the size of board of directors contribute a lot to these conditions.

Table 2 Sample statistics of variables

This table gives out summary statistics for the total sample. The database comprises 400 firm-year observations from 200 firms covering the period from 2006 to 2007. The descriptive statistics include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Board of directors size (BSIZE), Leverage (LEV) and Business Entities Shareholdings (BES).

CR 400 19,000 7,105,300 403,371 243,300 72.1325 TQ 400 0.0487 7.3955 1.3742 1.3283 0.6254 MPS 400 0 0.07 0.002575 0 0.009816 SIZE 400 14.1919 26.2664 21.2046 21.1796 1.4057 ROA 400 -3.81 2.51 0.037925 0.04 0.306321 TERM 400 0 12 3.88 4 1.8094 BSIZE 400 3 18 9.385 9 2.1699 LEV 400 0.0108 10.4277 0.6285 0.5317 0.7186 BES 400 0 1 0.3760 0.36 0.1829

4.2.2 Comparison of the samples between 2006 and 2007

increase, they begin to reduce their holdings in firms in order to give more flexibility to firms for their operating. Finally, the LEV level seems much fluctuate in 2007 than the year before, the leverage level of largest firm is 1.44 times higher than in 2006, and the standard deviation 1.77 times as high as the level in 2006, which implies that the risk of debt of Chinese firms in 2007 is more serious.

Table 3 Sample statistic of main variables between 2006 and 2007

This table gives out summary statistics for samples of 2006 and 2007. Each sample includes 200 observations in 2006 and in 2007. The descriptive statistics include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Board of directors size (BSIZE), Leverage (LEV) and Business Entities Shareholdings (BES).

V Results

5.1 Univariate correlation analysis

In this section, we use the Pearson test to verify the possible multicollearity problems between our variables.

First and foremost, TQ and SIZE are more strongly correlated with CR than other variables, following with MPS, which already suggest that SIZE and TQ are relevant inputs, though there are few managers holding shares in Chinese firms, the ones who holding shares reflect a relationship with CR. Moreover, we discover two variables, BSIZE and BES negatively correlated with CR. However, BSIZE is insignificant while BES is strongly significant. Not only such, they are both have negative and insignificant correlation with MPS and SIZE. Furthermore, LEV is significantly correlated with TQ, MPS, and the correlation level between LEV and TQ is close to the level between TQ and CR, but the sign is converse. However, LEV is only significantly correlated with two control variables, SIZE and BES. Finally, the results on the matrix says that, used alone, TERM, BES and BSIZE have little power to explain TQ, and SIZE, BSIZE and BES show weak relationships with MPS as well. And ROA shows insignificant correlation with TQ, MPS and CR. We therefore conclude that the data is adequate to interpret the analysis in our focus.

Table 4 Correlation matrix of all variables used in the model

This table gives out correlation matrix results for Pearson test of the inputs. The variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Board of directors size (BSIZE), Leverage (LEV) and Business Entities Shareholdings (BES).

CR TQ MPS SIZE LEV TERM BSIZE ROA BES

BSIZE -0.07 0.09 -0.01 0.11** -0.04 -0.04 1

ROA 0.02 0.09 -0.09 0.00 0.07 0.07 0.06 1

BES -0.14*** -0.06 0.04 -0.14*** 0.12** -0.03 -0.02 -0.01 1

N 400

*** Correlation is significant at the 0.01 level (2-tailed). ** Correlation is significant at the 0.05 level (2-tailed). * Correlation is significant at the 0.1 level (2-tailed).

5.2 Test Results and Multivariables Analysis

Before we use the two-year panel data to build up our database and adopt the two stage least square model to form our tests, we adopt the methodology from (Shema, 2008) that we comparing the 2SLS results in table 5, 6 and 7 with the OLS results to convince that 2SLS is more applicable than OLS in this case.

We then put our eyes on the results of our main analysis, the relationship between TQ, MPS and CR. We separately discuss the empirical test results of three two stage least square models that presents on table 5, 6 and 7 to give out our original understandings between our findings. Table 5, 6 and 7 below summarize the three 2SLS regression results of CR, TQ, MPS and all the relevant control variables, and in each model we let CR become endogenous variable in the CR model and exogenous variable in the TQ model, treat TQ as the exogenous variable in the CR and MPS model and endogenous variable in TQ model, MPS is consider to be the exogenous variable in the CR and TQ model, but become the endogenous variable in MPS model. Besides, for each model, among other control variables, we find that the result of R-squared in each model is 0.0572, -11.4257 and 0.3119, respectively. We should mention in this case the limitation of using R-squared in 2SLS models. Because in a 2SLS model R-squared can vary from plus one to the negative infinity, which prevent us from using this indicator from measuring the goodness of fit of our results. Thus, R-squared has no statistical meaning in the 2SLS context.

5.2.1 CEO Remuneration Model

discover TQ and MPS are significant predictors of CR at one percent and ten percent level, respectively, which are in line with the theoretical conclusions and our hypotheses. At the control variables level, SIZE and TERM are highly significant and positive correlated with CR at 1 percent level. BSIZE and ROA have weaker significant and negative related to CR at 10 percent and 5 percent level, respectively. However, institution shareholding has no significant correlation with CEOs’ remuneration. In term of the significant results of LM-test and F-test in table 5, which are reject the null hypothesis of applying the OLS model in the tests.

Table 5 2SLS and OLS Regression Results for CEO Remuneration model This table gives out statistical results for the variables of the CEO Remuneration model in 2SLS, OLS and 2SLS of Fixed Effect and Random Effect. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities’ Shareholdings (BES) and Leverage (LV).

Included observations: 400 CR (2SLS) CR (OLS) CR (2SLS) Fixed Effect CR (2SLS) Random Effect

R-squared 0.057 0.33

Hausman-test 0.36 0.2438

F-test 32.64457*** LM-test -2199.660***

*** Correlation is significant at the 0.01 level (2-tailed). ** Correlation is significant at the 0.05 level

* Correlation is significant at the 0.1 level (2-tailed). t-statistics in parentheses

SIZE has a significant and positive relationship with CR which is in line with the conclusions of Finkelstein and Hambrick (1989). SIZE can be interpreted as the task complexity of managers. Along with the structure of a firm becomes more and more

complicated, the requirement for responsibility and professional capacity of managers becomes higher and higher. Thus, due to the task of managers become more complicated, they are more likely to ask for a higher remuneration, while the size of firms is expected to expand.

BSIZE is correlated with CR, which yields the same result Core et al. (1999). Board of directors is an important section of a firm’s organization structure, the board aims at monitoring the behaviors of CEOs for shareholders, and urging the CEOs to fight for their shareholders’ optimal benefits. According to the relevant regulations of the corporate law, the assignment, removal and compensation policies of managers are determined by the board of directors. That is to say, if the managers can control the board, they can easily maximum their personal benefits by operating the decisions of the board members. Besides, interests of board members become diverged when the size of the board is expanded, and it is more difficult for them to reach a consensus within the board, and leads to a weaker corporate governance

rights in board, however, as the BSIZE become larger and lager, the force of monitoring the behaviors of CEOs become more intense. Thus, CEOs find it harder to maximum their remunerations while their control rights are weakened by additional board members.

At the corporate governance level, business entities shareholders can play part of the role in monitoring the behaviors of managers. They are propensity to raise their shareholdings in order to enhance the monitoring on the firms’ investment decisions base on considering their benefits. Thus, business entities are more likely to work their way through prevent the board of directors from adopting negative investment decisions on ongoing business. However, in this case, BES is negative and insignificant correlated with CR, which is similar to Ryan and Wiggins

(2001)’s empirical result but with a reverse sign. In many cases, the largest institute holder in Chinese firms is the government, and its purpose on monitoring the behaviors of CEOs in firms is mostly based on guaranteeing a better operating environment in firms when they are at their junior, comparing to their counterparties, like US, UK and Japanese firms with mature

structures. Thus, the BES is less likely correlated with the CEOs’ remuneration. Last but not least, when the firms become stronger, the government would like to relieve their monitoring on CEOs operating activities, therefore, tend to reduce their shares in hand.

ROA is positive correlated with CR which yields the same result with the empirical test of Attaway (2000). Operating performance is usually treat as one of the profitability indicators of a firm, in order to settle down the agency problems following by the separation of management and operating rights, board of directors go on designing the remuneration contracts in term of the profitability indicator. Thus, the CEOs can acquire better self-interests along with

maximum wealth of shareholders by devoting themselves to a probability promotion strategy. Generally speaking, the better performance it gains, the higher compensation they receive.

5.2.2 Tobin’s Q Model

variables level, TERM, SIZE and LEV are negative, but only TERM and SIZE are significant, correlated with TQ. BSIZE and BES have weak positive and significant relationship with TQ at ten percent level. In term of the significant results of LM-test and F-test in table 6, which are reject the null hypothesis of applying the OLS model in the tests.

Table 6 2SLS and OLS Regression Results for Tobin’s Q model

This table gives out statistical results for the variables of the Tobin’s model in 2SLS, OLS and 2SLS of Fixed Effect and Random Effect. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities’ Shareholdings (BES) and Leverage (LV).

Included observations: 400 TQ (2SLS) TQ (OLS) TQ (2SLS) Fixed Effect TQ (2SLS) Random Effect

*** Correlation is significant at the 0.01 level (2-tailed). ** Correlation is significant at the 0.05 level

* Correlation is significant at the 0.1 level (2-tailed). t-statistics in parentheses

LEV is another performance indicator of firms, in term of the previous studies we have mentioned about, LEV should be positive and significant correlated with TQ. Because of the state-control characteristic, many so-called large firms in China, which have 20 years operating histories at least, are still predominated by the government and leading to lack of competition advantages and operating efficiencies in the global market. Recall from table 2, the average level of leverage ratio in the sample is 0.37, thus, we can judge that CEOs in this kind of firms are still pursuing conservative growing strategies. Namely, when the corporate value is

abnormally growing, the CEOs prefer to keep down the pace of growth by reducing their LEV. Furthermore, due to the fact that most of these firms are consistently controlling by the state, the CEOs are usually propensity of gazing at industrial LEV but ignore the real conditions of economies. Hence, LEV seems much difficult to measure the relationship with corporate value.

TERM is found to be negatively correlated with TQ which is consistent with the empirical results of Chung and Pruitt (1996). CEOs with long continuous tenures in a firm are more likely to trend the more conservative investment opportunities in order to sustain their reputations or positions in their last few years tenures and that they loss many better growth opportunities, then, miss many chances to create higher value for their firms.

Managers can capture more abundant information, resources and make more efficient

investment decisions to maximum their corporate value as the size of their firms become larger. However, in light of the empirical results on table 6, the SIZE presents a negative relationship with TQ, this fact probably due to the board members attempt to use the raised funds to expand their SIZE. This action decentralizes their shareholdings speedily and leads to a further

Contrary to Eisenberg et al. (1998)’s point of view, BES in this sample is positive correlated with TQ. However, ironically, Chinese government has issued an act from the Chinese

corporate law in 2006 which obliges all the listed firms to fix their board members with nine persons within three years. In term of table 3, the average number of board members of 2006 and 2007 are identically to 9.385, that is to say, most of firms have already implemented the act on the issue year. Thus, except for the firms with BSIZE larger than nine or equal to nine, we conclude that the remaining observations whose BSIZE are smaller than nine and corporate values are growing during the sample periods contribute to the weak relationship between BSIZE and TQ.

Institution shareholding is positive correlated with TQ and in opposite to Pound (1988)’s Conflict of interest hypothesis, which believes that institute investors can utilize the

convenience of a shareholder to acquire information or benefits for themselves but contradict the benefits of firms. As we analyzed in the above parts, we get to know Chinese government play an important part in monitoring the operation of firms and giving additional assistant for firms in industries with potential growth, for instance, IT, finance and automobile. As a result, government attempts to hold more shares of firms with potential growth and high return on investment capital, simultaneously, reduce its holdings in industries with poor growing opportunities. Thus, BES has a positive relationship with TQ.

5.2.3 Manager’s Proportion of Shareholdings Model

Table 7 2SLS and OLS Regression Results for Manager’s Proportion of Shareholdings model

This table gives out statistical results for the variables of the Manager’s Proportion of Shareholdings model in 2SLS, OLS and 2SLS of Fixed Effect and Random Effect. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities’ Shareholdings (BES) and Leverage (LV).

Included observations: 400 MPS (2SLS) MPS (OLS) MPS (2SLS) Fixed Effect MPS (2SLS) Random Effect

TQ 0.0018*** (3.22) 0.001786*** (3.22) 6.50E-05 (0.21) 4.51E-05 (0.14) SIZE 0.0001 (0.34) 0.001786 (0.34) 0.00105 (1.38) 0.000957 (1.29) BSIZE -4.97E-05 (-0.26) 0.001786 (-0.26) TERM 0.0025*** (10.71) 0.002485*** (10.71) ROA BES 0.0011 (0.50) 0.001146 (0.50) -0.000654 (-0.22) 9.17E-05 (0.03) LEV 0.0037*** (5.92) 0.003681*** (5.92) -0.000662 (-1.43) -0.00069 (-1.51) R-squared 0.3119 0.3119 Hausman-test 39.26*** 39.11*** F-test 29.69504*** LM-test 1357.184***

*** Correlation is significant at the 0.01 level (2-tailed). ** Correlation is significant at the 0.05 level

* Correlation is significant at the 0.1 level (2-tailed). t-statistics in parentheses

BES is contrary to Pound (1988)’s Efficient Monitoring Hypothesis and SIZE is also biased from the previous empirical results. Both variables found to be weakly correlated with MPS. As we mentioned before, BES is always belong to the state, hence, its changes have nothing to do with the MPS.

BSIZE as well as BES and SIZE yield no close relationship with MPS. Most of the CEOs in Chinese firms do not hold any shares in firms. In other words, these managers find it difficult to pursue their benefits by the mean of controlling the board of directors. Thus, no matter how large the changes in the size of the board of directors, it does not give any interpretation on MPS’ change. In term of Demsetz and Lehn (1985), as the SIZE expanded, managers attempt to sustain the proportion of shares in order to maintain their controls by investing much more personal capital. In this case, managers do not need to consider about their proportion of shares due to their holding nothing. Thus, MPS has noting to do with the changes of SIZE.

the reformation policies are successfully implemented.

5.4 Summary for the Robustness Test

In term of Table 8 and 9 in the appendix, we can not find large differences regarding the main variables in 2006 and in 2007, respectively, from the two-year panel data’s empirical results. Qualititavely we can draw the same conclusions. Thus, we can conclude that the results are robust to model specification. In the fixed and random effect tests, due to the fact that size of board and the CEO’s term are constant through time, we drop the two variables of BSIZE and TERM in order to avoid the multicollearity problems and generate our fixed and random effect robustness tests for the variables left, respectively. Because we just use two years observations which are not sufficient to generate applicable results for the fixed and random effect tests, besides, most of the statistical results are not significant at all. Thus, the results of the fixed and random effect tests results in conlumn 3 and 4 of the table 5, 6 and 7 are for reference only.

VI Conclusion

On one hand, firms, which are enjoying fast growing returns, are more willing to demand on professional know-how of managers to operate their properties and make investment decisions. On the other hand, managers, who are compatible with crucial operating responsibilities in firms, are more likely to hold higher composition of firm shares and establish value for firms in order to influence the decision of board on remuneration policies. Thus, the framing of CEOs’ remuneration contract is playing an important role in corporate managerial organization.

Empirical results reveal that corporate value is positive and significantly influence the CEOs’ remuneration but CEOs’ remuneration also has a counterforce on propelling the grow of corporate value, and managers’ proportion of shareholdings are found to be positive significant with CEOs’ remuneration, though half of the managers in our sample have zero holdings in firms, and it also has an negative impact on corporate value. Besides, size of firms and tenures of managers are also positive significant correlated with CEOs’ remuneration. And operating performance, size of board are found to be significant and negative related to CEOs’ remuneration, but business entities’ shareholdings yields no correlation with CEOs’ remuneration.

References

Attaway, M. C. 2000. A Study of the Relationship Between Company Performance and CEO Compensation. American Business Review. Vol. 1: pp77-85.

Banker, R., and Datar, S., 1989. Sensitivity, precision, and linear aggregation of signals for performance

evaluation. Journal of Accounting Research. Vol. 27, No.1: pp21-39.

Barnhart, S. W., and Stuart Rosenstein. 1998. Board Composition, managerial ownership, and firm

performance: an empirical analysis. The Financial Review. Vol. 33, No.4: pp1-16.

Berle, A. and G. Means. 1932. The Modern Corporate and Private Property, Macmillan. New York. Micllan.

Bidwai, Praful, 2002. Global Corporate Meltdown: Market, the Fallen Angel. India-syndicate.com (July) Website:http://www.tni.org/detail_page.phtml?act_id=1144&username=guest@tni.org&password=9999&pu blish=Y (access date: 2009-02-17)

Chung, K. H. and S. W. Pruitt, 1996. Executive ownership, Corporate Value, and Executive Compensation:

A Unifying Framework. Journal of Banking & Finance. Vol. 20: pp1135-1159.

Cline, H., 1975. Wage Determination of Executive Compensation. Harvard Senior Honors Thesis. Conyon, Martin J & Leech, Dennis, 1994. Top Pay, Company Performance and Corporate Governance, Oxford Bulletin of Economics and Statistics, Department of Economics, University of Oxford, vol. 56(3), pp 229-247

Core., J. E., R. W. Holthausen, and D. F. Larcker. 1999. Corporate Governance, Chief Executive Officer

Compensation, and Firm Performance. Journal of Financial Economics. Vol. 51: pp371-406.

Coughlan, A. T. and R. M. Schmidt. 1985. Executive Compensation, Management Turnover, and Firm

Performance: An Empirical Investigation. Journal of Accounting & Economics. Vol. 7: pp43-66.

Crutchley, C. E., M. R. H. Jensen, J. S. Jahera, Jr. and J. E. Raymond. 1999. Agency Problems and The

Simultaneity of Decision Making: The Role of Institutional Ownership. International Review of Financial

Analysis. Vol. 8: pp177-197.

Crutchley, E. C. and R. S. Hansen. 1989. A Test of the Agency Theory of Managerial Ownership, Corporate

Leverage and Corporate Dividends. Financial Management. Vol. 18: pp36-46.

Demsetz, H. and K. Lehn. 1985. The Structure of Corporate Ownership: Causes and Consequences. Journal of Political Economy. Vol. 93: pp1155-1177.

Eisenberg, T., S. Sundgren, and M.T. Wells. 1998. Larger board size and decreasing firm value in small

firms. Journal of Financial Economics. Vol. 48: pp35-54.

Fama, E. F., and M. C. Jensen. 1983. Separation of Ownership and Control. Journal of Law and Economics. Vol.26: pp301-325.

Finkelstein, S. and D. C. Hambrick. 1988. Chief Executive Compensation: A Synthesis and Reconciliation. Strategic Management Journal. Vol. 9: pp543-558.

Finkelstein, S. and D. C. Hambrick. 1989. Chief Executive Compensation: A Study of the Intersection of

Markets and Political Process. Strategic Management Journal. Vol. 10: pp121-134.

Frank L. Winfrey, 2006. The Control of CEO Compensation by Institutional Investors: An Empirical

Study of Agency Theory in Large U.S. Industrial Corporations. Corporate Governance: An

International Review. Vol 2 No.4, pp211 - 224

Galbraith, C.S. and G. B. Merrill. 1991. The Effect of Compensation Program and Structure on SBU

Competitive Strategy: A Study of Technology-Intensive Firms. Strategic Management Journal. Vol. 12:

pp253-370.

Gregg, Paul. Stephen, Machin. and Stefan, Szymanski., 1992, The Disappearing Relationship Between

Directors' Pay and Corporate Performance. British Journal of Industrial Relations. Vol.31 No.1(September):

pp1-9.

Hall, B., Liebman, J., 1998, Are CEOs Really Paid Like Bureaucrats?, Quarterly Journal of Economics. Vol.1133: pp653–691.

Hu, Yifang, Min, Song and Junxi, Zhang., 2005. The Relevant Importance and Mutual determination of

Competition, Property and Corporate Governance, Economics Research. Vol. 9.

Jensen, M. C. 1993. The Modern Industrial Revolution, Exit and the Failure of Internal Control System. Journal of Finance. Vol. 58: pp831-880.

Jensen, M. C. and M. H. Meckling. 1976. Theory of the firm: managerial behavior-agency cost and

ownership structure. Journal of Financial Economics. Vol.3: pp305-360.

Kerr, J. L. and R. A. Bettis. 1987. Boards of directors, top management compensation, and shareholder

returns. Academy of Management Journal. Vol. 30, No.4: pp645-664

Liu, Deqiang and Keijiro, Otsuka., 2004. A Comparison of Management Incentives, Abilities, and Efficiency

between SOEs and TVEs: The Case of the Iron and Steel Industry in China. Economic Development and

Cultural Change, Vol.52: pp759-780

Li, Jin and Feng, Wang, 2007. The Corporate Value and Top Managers Compensation Policies in China. Economists. Vol. 12: pp 79-92

Main, B. 1992. Top Executive Pay and Company Performance. University of Edinburgh Mimeograph. McConnell, J. J. and H. Servaes. 1990. Additional Evidence on Equity Ownership and Corporate Value. Journal of Financial Economics. Vol.27: pp595-612.

Mehran, H., 1995. Executive compensation structure, ownership, and firm performance. Journal of Financial Economics. Vol.38: pp163-184.

Pound, J. 1988. Proxy Contests and the Efficiency of Shareholder Oversight. Journal of Financial Economics. Vol. 20: pp237-265.

Ryan Jr. H. E., and Wiggins. III. R. A. 2001. The Influence of Firm and Manager Specific Characteristics on

the Structure of Executive Compensation. Journal of Corporate Finance. Vol.7: pp101-123.

Singh, H. and F. Harianto. 1989. Top Management Tenure, Corporate Ownership Structure and the

Magnitude of Golden Parachutes. Strategic Management Journal: pp143-156.

Song, Zengji and Zongyi, Zhang. 2002. The Empirical Research between Managers’ Holdings and

Corporate Performance, Chongqing University Transaction. Vol.6: pp22-36

Shema, Amihood. 2008. Corporate Ownership Structure and Performance: The Reversed Causality

Argument Revisited. Faculty of Economics and Business, University of Groningen. Master Thesis.

Ungson, G. and R. M. Steers. 1984. Motivation and Politics in Executive Compensation. Academy of Management Review. Vol.9: pp313-323.

Wei, Gang, 2000. Top management incentive and listed firms operating performance, Economics Research. Vol.3

William, B. J., 2001. Return and Reward: Bank Performance and CEO Compensation. American Business Review. Vol.6: pp93-99.

Yang, Qing, Steven, Toms and Besim, Burcin, Yurtoglu, 2004. CEO Compensation, Corporate Governance

Yermack, D. 1996. Higher Market Valuation of Companies with a Small Board of Directors. Journal of Financial Economics. Vol.40, No.2: pp185-211.

Zhang, Xiaoning, 2002. Analysis between Operator Remuneration, Staff Holdings and Corporate

Appendix – Period Robustness Tests for the 2SLS Regressions

This table gives out statistical results for the variables of the CEO Remuneration model, Tobin’s Q model and the Manager’s Proportion of Shareholdings model with observations in 2006. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities’ Shareholdings (BES) and Leverage (LV).

Included observations: 200 CR model TQ model MPS model

CR 0.02*** (6.86) TQ 90.40*** (4.24) 0.006031*** (5.97) MPS -2889.12* (-1.68) -13.91 (-1.32) SIZE 22.67*** (2.99) -0.23*** (-3.12) -0.000431 (-1.00) BSIZE -0.80 (-0.20) 0.08* (2.26) -0.000184 (-0.75) TERM 19.38*** (2.74) -0.12** (-2.55) 0.002457*** (7.90) ROA -387.69*** (-3.00) BES -76.13 (-1.62) 0.62 (1.56) 0.001227 (0.42) LEV -0.43*** (-3.11) 0.007098*** (6.35) R-squared -1.705240 -1.298312 0.407088

Table 9 Robustness Test for 2SLS Regression Results of 2007

This table gives out statistical results for the variables of the CEO Remuneration model, Tobin’s Q model and the Manager’s Proportion of Shareholdings model with observations in 2007. The descriptive variables include: CEOs’ Remuneration (CR), Tobin’s Q (TQ), Managers’ proportion of shareholding (MPS), Corporation size (SIZE), Corporation operating performance (ROA), Manager Terms (TERM), Growth of total asset (GA), Board of directors size (BSIZE), Business Entities’ Shareholdings (BES) and Leverage (LV).

Included observations: 200 CR model TQ model MPS model

CR 0.08** (2.31) TQ 5.25 (0.75) 0.000523 (0.79) MPS 1595.78*** (2.95) -132.01* (-1.74) SIZE 22.42*** (5.67) -1.69** (-2.08) 0.000263 (0.61) BSIZE -4.20* (-1.95) 0.34 (1.53) 6.25E-05 (0.23) TERM 3.83 (1.21) -0.33 (-1.06) 0.002907*** (8.37) ROA 48.00 (1.28) BES -43.08 (-1.57) 2.90 (1.15) -0.000773 (-0.23) LEV -0.21 (-0.43) 0.002984*** (3.96) R-squared 0.278904 -32.150952 0.335026