Executive compensation and firm performance : evidence from China’s listed firms

Master’s thesis (6314M0044) 2013-2014 MSc Accountancy and Control – Control track University of Amsterdam

Amsterdam Business School Faculty of Economics and Business Roetersstraat 11

1018WB Amsterdam Student: Jennifer Poon

Student number: 10453822

Date: 22-06-2014

First supervisor: Second supervisor:

Dr. B. Qin M. Schabus MSc

University of Amsterdam University of Amsterdam

Executive compensation and firm performance: Evidence

from China’s listed firms

Abstract

This paper investigates the relation between firm performance and executive compensation of China’s firms listed on the Shanghai- and Shenzhen Stock Exchange between 2006 and 2007 using the firm performance proxies: return on assets, annual market return, and Tobin-Q. Furthermore, it examines the effect of foreign institutional investors on the amount of compensation, and whether it will strengthen the pay-performance sensitivity.

Three hypotheses are investigated in this research. Hypothesis 1: Executive

compensation depends on firm performance. Hypothesis 2: Executive compensation is higher when there are more foreign institutional investors. Hypothesis 3: The positive relation

between firm performance and executive compensation is stronger in firms with more foreign institutional investors. Based on the hypotheses the main research question: “Does firm performance determine the top executive compensation of China’s listed firms?” and sub-question “Does the foreign institutional investors influence the pay-performance sensitivity of China’s listed firms?” are answered.

Using multiple regression analyses to investigate the aforementioned relations, the data supports the hypothesis that firm performance does determine the amount of executive compensation. Furthermore, there is evidence that the presence of foreign institutional investors in the firm result in higher executive compensation. However, the data does not support that foreign institutional investors will strengthen the pay-performance sensitivity in China’s listed firms. It is believed that foreign institutional investors in firms are used as camouflage to create the illusion that the corporate governance and, in turn, the pay-performance sensitivity is strengthened, while in practice this is not the case.

This research contributes to the extant literature by including equity-based

compensation, which was legalized by the end of 2005, in the total executive compensation, and focusing on the influence of foreign institutional investors on the pay-performance sensitivity. In contrast, the extant literature is based on studies that were conducted before the legalization of equity-based compensation with a focus on the effects of State ownership on pay-performance sensitivity.

Keywords: Executive compensation, firm performance, pay-performance sensitivity, China’s listed firms, foreign institutional investors

Acknowledgements

I would like to express my appreciation to Dr. Qin for his valuable and constructive

suggestions during the development of this research work. I also wish to thank the University of Amsterdam for providing a strong theoretical foundation to successfully finish this

Table of content Abstract ...1 Acknowledgements ...2 1. Introduction ...5 1.1 Background information ... 5 1.2 Research question ... 7 1.3 Results ... 7 1.4 Motivation ... 8 1.4.1 Scientific relevance ... 8 1.4.2 Societal relevance ... 10

1.5 Structure of the thesis ... 10

2. Literature review and hypotheses ... 10

2.1 Economic Reform: Transition SOEs to privately held firms and executive compensation reform in China ... 11

2.1.1 Stage 1: Management reform, 1978 – 1984 ... 11

2.1.2 Stage 2: The Dual track-system, 1984 – 1992 ... 12

2.1.3 Stage 3: Ownership reform, post 1992 ... 12

2.2 Corporate governance in Chinese listed firms ... 13

2.2.1 Performance assessments and incentive systems ... 13

2.2.2 Ownership structure A-shares dominated ... 14

2.2.3 Ownership structure influence of foreign ownership ... 16

2.3 Information environment A- and B-shares... 17

2.4 Pay-performance sensitivity in China ... 19

2.5 Hypotheses and theory ... 20

3. Data and research method ... 24

3.1 Data ... 24

3.2 Measurement theoretical construct ... 26

3.3 Research method ... 30

4. Results ... 32

4.1 Sample ... 32

4.2 Descriptive statistics ... 32

4.4 Regression analysis ... 35

4.4.1 Firm performance, foreign institutional investors, and executive pay ... 36

4.4.2 Foreign institutional investors and pay-performance sensitivity ... 39

4.5 Robustness tests ... 41

4.6 Pay-performance sensitivity in firms with and without foreign institutional investors.. 42

5. Discussion and conclusion ... 45

5.1 Limitations ... 49

5.2 Recommendations ... 49

References ... 50

1. Introduction

This paper aims to examine the relationship between firm performance and executive

compensation in China’s publicly traded firms. Furthermore, the effect of foreign institutional investors affecting the relation between the executive compensation and firm performance will be investigated.

1.1 Background information

For over 20 years, policymakers, shareholders, academics, and other parties have expressed criticism regarding executive pay according to Harvard Magazine (Lorsch & Khurana, 2010). Two major issues emerge based on the criticism: the amount of incentive pay and the flaws between the relation of firm performance and executive pay. This subject has also been investigated by many researches over the years (Coughlan et al., 1985; Mehran, 1995; Core et al., 1999; Murphy, 2004; Brick et al., 2006; 2012). However, the researchers have not proven a clear correlation between firm performance and executive compensation.

Murphy (2004) showed that the majority of corporations apply performance-based compensation to incentivize the management of companies to reduce agency problems. Agency problems arise when there is a conflict of interest between the principal

(shareholders of the company) and agents (management of the company). Furthermore, in later research, Murphy (2012) showed that CEOs might be influenced by the opportunities to earn bonuses; however, this will not necessarily give them the motivation to increase the firm value. For instance, executives might apply earnings management to earn the compensation without adding value to the company. Core et al. (1999) showed that a weak corporate governance will lead to higher CEO compensation. This does not necessary reflect that better firm performance was the provocation to receive the higher compensation. Furthermore, Brick et al. (2006) stated that excessive compensation is related with firm-underperformance and is not dependent on poor corporate governance.

Scientific researches on the aforementioned topics are mostly conducted in Anglo-American countries, due to the availability of data for example (Coughlan et al., 1985; Mehran, 1995; Core et al., 1999; Murphy, 2004; Brick et al., 2006; 2012). However, related scientific researches in emerging economies like China are underrepresented, with researches mainly using datasets between 2001 and 2005 (Mengistae & Xu, 2004; Firth et al., 2006; Kato & Long, 2006a; Fan et al., 2007; Buck et al., 2008). This is the period before the legalization of

equity-based compensation for the firm’s managers, directors, and executives. Furthermore, China Daily USA reported an increasing rate of executive pay in China compared to the United States (U.S.). By 2017, it is expected that the executive pay of China’s listed firms will exceed the U.S. firms’ pays (Barris, 2013). It is therefore of interest to investigate whether firm performance determine the growing executive compensation in China’s listed firms.

China has risen to the second largest economy and Asia’s largest economy in 2010 (Hamlin & Li, 2010), and it is expected that by 2030, China will be the world’s largest economy (China Daily, 2013).Along with the high pace of economic developments in China, the country has undergone drastic changes since the economic reforms in 1978. This was also the turning point from state-owned enterprises (SOEs) to privately owned enterprises in China (KPMG, 2013). In recent years, regulations for foreign investment in Chinese listed firms became less strict. In 1990 and 1991 Shanghai Stock Exchange (SSE) and Shenzhen Stock Exchange (SEZ) were opened and started with trading A- and B-shares for China’s listed firms according to a report by KPMG (2013). In a later stage, there were also H-shares traded in Hong Kong and N-shares traded in New York. The main difference between A- and B-, N-, and H-N-shares is that the latter is quoted in foreign currencies and open to foreign investors. The A-shares, on the other hand, are only quoted in the Chinese currency Renminbi (RMB) and mainly available for mainland citizens (Abdel-Khalik et al., 1999; Chui & Kwok, 1998). According to Abdel-Khalik et al. (1999) and Siu (1996), B-shares are mainly dominated by foreign institutional investors, while A-shares are mainly dominated by mainland individual investors.

As of 2006, China’s listed firms were allowed to grant equity incentives to the firms’ managers, directors, and executives for the first time (CSRC, 2005b). Croci et al. (2012) showed that foreign institutional investors in Continental Europe have a positive influence on CEO-compensation level. Moreover, the authors have proven that foreign institutional investors will lead to higher equity-based compensation. Consequently, it can be questioned whether foreign institutional investors in China’s listed firms will enhance the executive compensation and, in turn, the pay-performance sensitivity, as in the research conducted in Continental Europe.

Based on the aforementioned extant literature and findings, China is deliberately chosen for this research. The reasons for this are its rapidly developing economy, the economic reform from SOEs to privately held firms, which led to trading of A- and foreign shares (B, H, and N-shares), and the legalization of equity-based compensation. This research

focuses mainly on A- and B-shares, due to their listing on the Shanghai- and Shenzhen Stock Exchange.

In short, this research aims to investigate two relations: the relation between firm performance and the growing executive compensation in China’s listed firms, and the effect of foreign institutional investors on the pay-performance sensitivity in China’s listed firms. The relations are graphically represented in figure 1.

(?)

(?)

Figure 1: Effect of foreign institutional investors on the relation between firm performance and executive compensation

1.2 Research question

The research question is formulated as follows: Does firm performance determine the top executive compensation of China’s listed firms?

This paper will also examine the effect of a potential moderator which is formulated as: Does the foreign institutional investors influence the pay-performance sensitivity of China’s listed firms?

1.3 Results

Multiple regression analyses are conducted to test the relation between firm performance and executive compensation. Furthermore, the effect of foreign institutional investors on the executive compensation and pay-performance sensitivity is also provided. The data supports that executive compensation is determined based on firm performance. Moreover, when foreign institutional investors are present in the firm, the data supports that this will lead to a higher amount of executive compensation. However, strengthened pay-performance

sensitivity in the presence of foreign institutional investors in the firm is not supported by the data. Executive compensation Firm performance Foreign institutional investors

1.4 Motivation

1.4.1 Scientific relevance

Scientific research regarding firm performance and executive compensation in emerging economies has become more common nowadays (Mengistae & Xu, 2004; Firth et al., 2006; Kato & Long, 2006a; Fan et al., 2007; Buck et al., 2008). However, extant researches tend to use datasets drafted from 2001 to 2005, because as of 2001, the China Securities Regulatory Commission (CSRC) (CSRC, 2001) required listed firms to report the three highest executive pay. Moreover, no equity-based compensation was included in pay-performance sensitivity researches in this period, using evidence of China’s listed firms. This executive compensation of the researches mainly consists of cash compensation and bonuses. The main reason is due to the fact that companies listed on the Shanghai- and Shenzhen Stock Exchange were not allowed to offer restricted stocks or stock options to managers, directors, and executives of the firm before 2006 (CSRC, 2005b). Research by Hu and Zhou (2008) showed that companies with managerial ownership in the form of equity shares are more positively related to higher firm value than companies without managerial ownership. Due to the difficulties in gathering data for listed firms, the authors have chosen to use 1,500 Chinese non-listed companies as their sample. The research focused on the period between 1998 and 2000, and the

companies represented ten different industries in the service- and manufacturing-sector. Eventually, 83 companies of the total sample had managerial ownership.

Negative publication regarding overcompensation and corruption dominated the media regarding Chinese companies over the years (Pei & Kaufmann, 2007; Rein, 2009; Barboza, 2012). Examples of factors influencing the inappropriate compensation are, for instance, cultural factors, family connections with the shareholders, the power of the executives, cronyism, and indirect compensation. Magnan and Li (2008) mentioned a potential factor that could lead to the problem regarding inappropriate incentives, as they believe that the lack of equity-based compensation before 2006 has led to incentive problems. Also, Liu and Lu (2007) concluded that the misalignment of interest between shareholders and investors is a significant factor in the company’s earnings management. This is in line with the agency-theory, which argues that aligning the interest of agent and principle using long-term incentives can reduce the agency problem of the company (Jensen &

Meckling, 1978). The assumption of Magnan and Li (2008) is supported by the research of Firth et al. (2006), who did not found statistical evidence between changes in cash

compensation and firm performance. However, the research by Buck et al. (2008) did find a positive relation between firm performance and executive compensation: both as a reward and motivation, solely based on short-term compensation without equity-compensation. As a result, it can be questioned whether cash compensation is the most suitable method in granting compensation in China’s listed firms.

By the end of 2005, the CSRC issued the regulation “Guidelines for Equity-Based Compensation”, which states that the Chinese listed firms on the Shanghai- and Shenzhen Stock Exchange are allowed to issue restricted stocks and stock options to the firm’s managers, directors (excluding independent directors), and executives with a maximum amount of 10% of the shares (CSRC, 2005b). Jensen and Meckling (1976) mentioned that agency costs of equity are a sum of (1) monitoring expenditure by the principle; (2) the bonding expenditure by the agent; and (3) the residual loss. Furthermore, the authors explain that agency costs of equity arises with the delegation of decision rights from the shareholders towards executives. Moreover, they have shown that if managers hold 100 percent of the shares, a one-dollar expenditure on perks will equal a one-dollar decrease in wealth.

However, if the manager only holds 95 percent of the shares, the manager can still consume the perks to its full extent, but experience only 95 percent decrease in their wealth.

Therefore, it is argued that managers will most likely take riskier decisions and consume more perks since they only experience a fraction of the total decrease in wealth. As a result, these costs are also incorporated in the agency costs. The agency theory argues that in order to lower the agency costs, there should be a proper balance between the firm performance and related compensation. However, as mentioned previously, the “Guidelines for Equity-Based Compensation” issued by the CSRC (2005b) have set a maximum amount of 10 percent of the total shares as an incentive for company´s management, directors, and executives. According to the view of Jensen and Meckling (1976), questions may arise whether this will result in higher agency costs since the executives can consume 100 percent of the perks while this will decrease their wealth by only 10 percent.

Thus, this research contributes to extant literature by using a more recent dataset from 2006 and 2007, in which the equity-based compensation is included as part of the total remuneration of executives in testing the pay-performance sensitivity. This compensation pattern is more comparable to Western best practices, as it applies for instance the agency theory on a wide scale and expects a positive relation between firm performance and

executive compensation using long-term incentives. Furthermore, extant literature mainly focuses on the State agency as the major stakeholder and its influence on the

pay-performance sensitivity. In contrast, this research aims to investigate the effect of foreign institutional investors on the pay-performance sensitivity.

1.4.2 Societal relevance

From a societal point of view, this research is interesting due to the problems regarding family-companies. Family companies make up for 40 percent of private companies. However, Reuters (2013) stated that the next generations of current family-company owners prefer to work for outside non-family companies. Reuters also showed that the Chinese Academy of Social Sciences expects that over three million private businesses will deal with succession issues in the upcoming three to eight years, as China’s first generation, where majority of the Chinese couples have only one child due to the one-child policy, have reached their

retirement age; it is expected that the next generation is unwilling to take over the family-company.

In other words, these private family-companies are forced to hire outside executives to run their businesses. In turn, this might lead to agency problems. This research contributes to this matter by investigating the effectiveness of performance-based compensation of executives including long-term compensation and the effect of issuing shares to foreign institutional shareholders on pay-performance sensitivity of the company.

1.5 Structure of the thesis

The remaining chapters of the thesis are as follows. Chapter 2 provides the literature review and hypothesis. In chapter 3, the research method and data used for the research are introduced. Chapter 4 will present the results based on the data and research method as mentioned in the preceding chapters. This is followed by the discussion and conclusion, limitations, and recommendations in chapter 5.

2. Literature review and hypotheses

This chapter will elaborate on more specific topics for this research, such as the different stages of China’s economic reform, corporate governance topic regarding the incentive

system and ownership structure, information environment, and pay-performance structure. Afterwards, the formulated hypotheses based on the literature review are presented.

2.1 Economic Reform: Transition SOEs to privately held firms and executive compensation reform in China

China has faced remarkable changes since the Deng Xiaoping economic reforms in 1978. Prior to 1978, the state-owned enterprises (SOEs) faced great control of the State controlling the enterprises’ activities according to Firth et al. (2006). Moreover, the research showed that executive compensation was fixed based on industry, region, age, and title. Furthermore, the profits were transferred to the government, whereas deficiencies were covered by the government as well. The starting point of China’s economic reform in 1978 can be divided in three stages (Wang, 2004, Firth et al., 2006; Brandt & Rawski, 2008; KPMG, 2013). These stages will be elaborated on more in detail in the next subparagraphs. Furthermore, the executive compensation reform in China will also be taken into consideration in the subparagraphs.

2.1.1 Stage 1: Management reform, 1978-1984

During the first stage of reform, in October 1978, the Sichuan provincial government introduced a pilot project to delegate greater autonomy to the management in six selected enterprises (Wang, 2004; Brandt & Rawski, 2008). As opposed to transferring the profits to the government, these six enterprises were allowed to keep a part of the total profit. Another remarkable change is the usage of the saved money: this could now be used to re-invest in the enterprise, for instance in production or to grant the employees with a bonus payment. Since the enterprises were allowed to produce more than the state-set quota, these changes in usage of saved money can in turn foster the production numbers. By the end of 1980, the project expanded to 6,600 large- and medium-sized SOEs. This made up for 60 percent of the national budgeted industrial output and 70 percent of national industrial profit (Wu, 2004).

In this stage, signs of a shift from a highly imbalanced to a more proper balanced relation between the inputs: hard work, commitment, effort and skill, and outputs: financial rewards, stimulus, sense of achievement and recognition became visible in privately held firms (Adams, 1965). This development also changed the perception of fairness of employees towards the firm and might foster the mitigation of agency problems in the long run.

2.1.2 Stage 2: The Dual-track system, 1984 – 1992

The second stage of reform was fostered by proclamation of the Provisional Regulations on Expanding the Autonomy of Enterprises in May 1984 (Wang, 2004, Firth et al., 2006; Brandt & Rawski, 2008; KPMG, 2013). TheDual-track system allowed enterprises to sell their products, if it has exceeded the production quotas, outside the state plan at prices to maximum 20 percent above the state price. Moreover, the enterprises were granted authority to hire and fire mid-level functions, grant rewards and bonuses, and set up direct links with suppliers. Furthermore, in 1984, as a replacement of profit remittance, the profit tax was introduced.

At this stage, the enterprises were not allowed to hire or fire staff solely based on business considerations. However, there is a continuous positive change in the granted authority to management and their perception towards a greater balance between their effort to enhance firm performance and interlinked rewards. In turn, this will add positive stimulus to keep up their hard work in the future (Premack, 1965).

2.1.3 Stage 3: Ownership reform, post 1992

The last stage of the economic reform started early in 1992, as the reformation process was further fostered by Deng Xiao Peng by the issuance of Regulations on Transforming the Operational Mechanism of State-Owned Industrial Enterprises (Wang, 2004, Firth et al., 2006; Brandt & Rawski, 2008; KPMG, 2013). He emphasized that officials should not focus solely on ideological correctness but focus more on economic development. In the same year, the pilot experiment of a ‘yearly salary system’ was introduced. This was the period in which the executive compensation really started to reform according to Kato & Long (2006a). Their research stated that the ‘yearly salary system’ was divided in two parts: the base salary that depends on the average wage for employees and the firm size; and risk salary, which can be interpreted as bonus that is interlinked with the base salary and the firm performance of the year. The base salary was distributed every month to the executives, while the risk salary is distributed on a yearly basis. In 1995, ownership reform was further developed in a high pace, because central government retained ownership of approximately 500 to 1,000 large-scale SOEs, and smaller SOEs could be leased or sold (Wang, 2004). Since 1997, the ‘yearly salary system’ became a very important form of executive compensation. Eventually, in 2001, around 70 percent of SOEs was partially or fully privatized (Garnaut et al., 2007).

At this stage, there is a more balanced relation between firm performance and related compensation, perception of fairness by executives, and positive stimulus to act in the best interest of the firm. Ultimately, this can reduce the agency problems in the future, when more advanced developments take place, such as the introduction of equity-compensation usage in 2006 as elaborated on in the scientific relevance section.

2.2 Corporate governance in Chinese listed firms

Lin (2001) showed that the corporate governance of China’s listed firms are of low quality. This is based on interviews with different parties such as stock exchange regulators and government officials between 1993 and 1997. The research highlights major problems: excessive power of the CEO and insider control, inadequate safeguard for outsiders, weak managerial incentives, and inadequate transparency and disclosure.

However, the corporate governance systems of the listed firms have undergone several changes over the years. Remarkable changes are, for instance, the mandated two-tier structure, codified independent directors, greater power granted to the supervisory board, introduction of an audit- and remuneration-committee that mainly consists of independent directors, and the introduction of equity-based incentives. Based on the aforementioned changes over the years, it becomes clear that the institutional context changes of China’s listed firms are developing towards the Anglo-Saxon economies (Core et al. 2003). However, Xi (2006) argued that even though the objective to hire independent directors is to safeguard the small- and middle shareholders and to reorganize for inside control, it is the State that ultimately appoints the independent directors in order to create stronger ties to the State. In the next sub-paragraphs, more specific corporate governance subjects are elaborated on.

2.2.1 Performance assessments and incentive systems

In the Code of Corporate Governance for listed companies in China (CSRC, 2001) it was stated that the companies should have a two-tier system that consists of a management- and

supervisory-board. The duties of the supervisory board are to monitor the management board on behalf of the shareholders and represent the voices of the shareholders. Moreover, the Code highlights the duties of the supervisory board and board of directors in the

In 2001, the China Securities Regulatory Commission (CSRC) mandated that all Chinese listed firms should have independent directors (CSRC, 2001). This was eventually codified in the guidelines of 2003, which state that at least one third of the board members are

mandatory to meet the independence director requirements. Requirements are, for instance, that director’s parents, children, and spouses are prohibited to fulfill an independent director function, as stated in the CSRC (CSRC, 2001). An important task of the independent directors is to provide an opinion on the remuneration of directors and senior officers.

Furthermore, in chapter 5, “Performance Assessments and Incentive and Disciplinary Systems” of the Code of Corporate Governance for listed companies in China, section 77 and 78 specifically state that there should be a performance-based incentive system in order to maintain and satisfy qualified employees (CSRC, 2001). The sections emphasize that

performance assessment should be the main tool in determining the compensation of management personnel. Also, the board of directors of the company should approve the performance assessment.

Xi (2006) highlighted that, in 2005, the National People’s Congress (NPC) made remarkable amendments on the Company Law of 1993. Besides that, independent directors were now compulsory for all Chinese listed firms and codified. The function of the supervisory board was also strengthened by granting greater power to them to independently obtain more information, should the board doubt the information provided by the firm.

Furthermore, listed firms are mandated to have an audit- and remuneration-committee that consists mainly of independent directors. All these developments in corporate governance are signs of changes to a more Anglo-American governance.

2.2.2 Ownership structure A-shares dominated

According to Firth et al. (2006), China’s firms can be categorized in three classes of shares as stated in the law: shares owned by the State, shares owned by legal entities, and tradable shares owned by individuals and private institutions. Moreover, they believe that the individual and institutional investors’ main objective is to maximize the profit and stock market value, while state ownership focuses on political and/or multiple objectives, such as employment growth, rather than profit maximization. Researches have shown that on average, one-third of the listed firm’s shares are owned by the State, one-third are owned by legal entities, and the remaining one-third of shares are tradable shares that are sold to

private investors only in small proportions (Firth et al., 2006; Li et al., 2008). This way, the State holds sufficient shares to remain control over the enterprise. Qiang (2003) have shown that by the end of 2001, the State holds ultimate control over 84 percent of shares in China’s listed firms.

Due to the economic reform, the government introduced a pilot project in 2005, aiming that all shares are tradable as of April 2005, which resulted in a high growth in tradable A-shares (Yeh et al., 2009). Nowadays, listed companies can issue four types of shares as listed below (Aharony et al., 2000; Chan et al., 2004):

- A-shares: Chinese companies listed on the Shanghai- and Shenzhen Stock Exchange issuing shares to domestic investors and are valued in Renminbi (RMB);

- B-shares: Chinese companies listed on the Shanghai Stock Exchange are valued in United States Dollar (USD) and companies listed on the Shenzhen Stock Exchange are valued in Hong Kong Dollar (HKD) issuing shares to foreign investors;

- H-shares: Chinese companies listed on the Hong Kong Stock Exchange are valued in HKD; and

- N-shares: Chinese companies listed on the New York Stock Exchange and NASDAQ are valued in USD.

The ownership structure of the listed firms, which consist of solely A-shares, are still highly concentrated notwithstanding the economic reform, which means that the firm is possibly dominated by one single shareholder according to Xu (2004). Moreover, the research showed that the dominant shareholder is normally either the State or a legal entity. The researcher also showed that it is not rare that the largest shareholder holds 40 percent of the shares, the second-largest shareholder approximately 9 percent of the shares and the third largest shareholder 4 percent of the shares. Therefore, there is a large difference between the largest- and second-largest shareholders. As a result, the largest and most powerful shareholder can exert power regarding the determination of executive compensation. The highly concentrated ownership structure of China’s listed firms is in contradiction with Anglo-Saxon economies, which is characterized by diffuse ownership to prevent too powerful shareholders. However, the author found an increasing rate in private owners as the largest shareholder of the firm who focus on the usage of performance-related pay schemes, market values, and profit of their invested firm.

Research by Core et al. (1999) showed that the concentration of ownership highly

influences the executive compensation. The same opinion is shared by Firth et al. (2007) and Chen et al. (2010); they believe that the ownership structure is also a potential factor that influences the level of cash pay. According to the agency theory, when the ownership is diffused, the owners are less likely to monitor and exercise influence over corporate decisions (Jensen & Meckling, 1976; Fama & Jensen 1983), which is applicable to Anglo-American economies. On the other hand, highly concentrated ownership leads to stronger incentives to provide supervision on managerial activities (Jensen & Warner, 1988), which is applicable to the Chinese listed firms. Therefore, it is expected that, due to the high concentration of ownership in China, shareholders can do better and are more willing to safeguard the interest of the firm; this reflects their own interest. Shivdasani (1993) expects that when there is high concentration of ownership, an A-shares firm, the ideal contract will have fewer financial incentives as a motivator for executives.

2.2.3 Ownership structure influence of foreign ownership

This research will mainly focus on shares listed on the Shanghai- and Shenzhen Stock Exchange. Consequently, B-shares are the only foreign shares among the total amount of foreign shares included in this research.

Since 1991, the category B-shares were introduced strictly for foreign investors and traded in foreign currency. According to Chakravarty et al. (1998), B-shares’ owners have the same rights and obligations as A-shares’ owners; however, an individual is not allowed to hold more than 25 percent of the total B-shares of the firm. Furthermore, the authors have

mentioned that the total issuance of B-shares is restricted at a maximum of 49 percent of the firm’s total shares. The maximum of 25 percent of B-shares per individual and a maximum amount of 49 percent of B-shares issued by the firm is mandated by the CSRC (Ernst & Young, 2012). Furthermore, the maximum issuance of B-shares of a firm, before the maximum of 49 percent, was 33 percent. Thus, this increased the influence of foreign institutional investors, but ultimately the domestic investors will remain in control of the listed firms (Ernst & Young, 2012).

Research by Firth et al. (2006) showed that, in order to attract foreign investors to buy the shares, China’s listed firms would more likely adopt international standards as well as international governance and business practices. Moreover, the authors expect that the

foreign investors will put pressure on the Chinese listed firms to hire executives with international experience and who possess higher qualities. Core et al. (1999) showed that complex firms need sophisticated managers; in turn, this will result in higher incentive pay to compensate for their input. Moreover, Firth et al. (2006) also expect that the better-qualified executives with international experience are able to negotiate higher incentive pay. This reflects the employees’ perception of fair distribution of resources, which plays an important role in finding a proper balance between inputs: hard work, commitment, effort and skill, and outputs: financial rewards, stimulus, sense of achievement and recognition, to motivate the executives to work hard and in the best interest of the firm (Adams, 1965). Ding et al. (2006) conducted a survey among 465 firms located in Shanghai, Nanjing, and Guangzhou, and found a positive relation between foreign investors and performance-related compensation

packages. The authors have mentioned that this result is consistent with a survey conducted in 2003 by an institute under China’s State Council’s Development Research Centre among 1,883 firms across China: this survey showed that, due to the higher pressure to attract potential management, companies with foreign shares will provide the local managers with higher pay, better benefits, and more performance-related compensation packages (China Daily, 2004). Furthermore, the foreign investors will put pressure on the firms to use

performance-related schemes (Firth et al., 2006). This will result in a positive reinforcement, which motivates to maintain the behavior in performing good in the future, while a negative reinforcement will result in the opposite effect (Premack, 1965). In short, it is believed that the presence of foreign institutional investors in the firm will foster a higher executive compensation and stronger pay-performance sensitivity.

2.3 Information environment A- and B-shares

A-shares are dominated by mainland domestic investors, while foreign institutional investors mainly dominate B-shares. This is also reflected in the information environment. Research showed that the differences in the information environment directly influence the pay-performance structure of firms (Abdel-Khalik et al., 1999; Gul et al., 2010).

Abdel-Khalik et al. (1999) identified the environment of B-shares with the following characteristics:

- Financial reporting of firms must follow International Accounting Standards (IAS) to ensure the credibility of the reported numbers;

- Firms must use international recognized audit firms to audit the financial statements in order to ensure a high quality reports which is compliant with IAS;

- State officials are more likely to avoid domestic issues to foster and ensure the continuous inflow of foreign investment;

- There is a segregation of duties between the financial analysis, investors, auditor, and state officials in monitoring the activities and performance during financial analysis to ensure the quality of the work performed;

- The majority of the foreign institutional investors will act as an external monitor in firm’s decisions and activities.

In contrast, the authors have identified that the A-shares environment is less formal. This was also reflected in the procedures of making initial public offerings of A-shares, as the authors characterize the initial years of the public offerings by the communication networks of relatives and friends which was further fostered by the institutional arrangements.

Gul et al. (2010) mentioned that regulations are different for firms consisting of solely A-shares or both A- and B-A-shares. The authors have used a sample of 1,142 non-financial listed firms between 1996 and 2003, and showed that firms issuing solely A-shares prepared

financial reports according to the Domestic Accounting Standards (DAS), while firms issuing A- and B-shares were required to prepare financial reports in accordance with International Financial Reporting Standards (IFRS). Hou et al. (2014) have found a more positive relation between the adoption of IFRS and executive compensation, in contrast to the adoption of DAS, using data of listed firms on the Shanghai- and Shenzhen Stock Exchange from 2004 to 2009. In other words, companies that consist of both A- and B-shares are expected to have higher executive pay, as companies with B-shares are mandated to adopt IFRS instead of DAS. Furthermore, Gut et al. (2010) showed that foreign investors are less associated with

information asymmetry compared to firms issuing solely A-shares. Consequently, lower information asymmetry results in lower agency costs according to the agency theory.

Abdel-Khalik et al. (1999) conclude that the B-share market is more formal and has a similar information environment as developed markets, such as the United States. The A-shares market is less formal, and therefore less comparable to the information environment of developed markets.

2.4 Pay-performance sensitivity in China

The executive pay in China started to reform in 1992 as described in the preceding part. However, the executive pay is differently disclosed in China than in developed countries, where the total compensation is divided in separate elements. China’s listed firms were mandatory to report the compensation as a total number of the top three directors’ and executives’ total compensation as of 2001 (CSRC, 2000; CSRC, 2002). Even though it was not mandatory to enclose the executive pay before 2001, some firms did it on a voluntary basis. Additionally, equity-based compensation was not allowed before the end of 2005 (CSRC, 2005a), which is in contrast with the Anglo-American studies.

A widely accepted economic theory of executive pay is the agency theory of Jensen and Meckling (1976). The agency theory advocates that top management salaries should be based on stock returns and accounting profitability, and is widely used in scientific papers among the Anglo-American studies (Jensen and Murphy, 1990; Murphy et al., 1999; 2012; Goh and Gupta, 2010). The agency theory of Jensen and Meckling (1976) argues that long-term incentive is an important motivator for employees to work in the best interest of the firm. Moreover, the authors claim that agency costs arise due to moral hazards and adverse selection problems. However, determining an efficient compensation package in order to align the interest of the executive with the interest of the firm and shareholders can reduce agency costs. As a result, the agency theory predicts that the executive pay is positively related to the firm performance. However, Kato and Long (2006a) have investigated 937 Chinese listed firms using data from 1998 to 2002, and found a positive relation between short-term compensation and firm performance. Moreover, Firth et al. (2007) have also found a positive relation between cash compensation and firm performance using a sample of 549 listed firms between the period of 1998 and 2000. Furthermore, research by Buck et al. (2008) showed a positive relation between firm performance and executive pay even though no equity compensation was observed using data of 601 listed companies between 2000 and 2003.

Finkelstein and Hambrick (1988) stress that the effectiveness of performance-pay schemes is low when the base salary is already relatively high. Furthermore, due to the personal reputation of executives, they are inclined to maximize the profitability of the firm notwithstanding the performance-related pay. This way, executives can have a higher chance to rotate to other firms with the same or a higher salary.

However, since 2006, equity-based compensation was legalized and public listed firms were mandated to report the total compensation of each members as the sum of salary, bonus, stock options and other benefits (CSRC, 2005a). The most remarkable change is the issuance of stock options and restricted stocks, which makes the compensation structure more comparable to the Anglo-American studies. Firth et al. (1999) argue that compensation is a motivator towards executives to align their goals with the goals of the company.

Therefore, it is suggested that executives should be rewarded for a desired behavior:

increasing stock prices with the proper risk appetite and penalized for inappropriate behavior (Premack, 1965).

In short, theories used in Anglo-American studies are better applicable in investigating the pay-performance sensitivity in China’s listed firms after 2006 due to the legalization of long-term incentives, which play an important role in the Anglo-American studies.

2.5 Hypotheses and theory

Based on the literature review, three hypotheses are developed and are presented in the next section.

Firm performance and executive pay

In the third stage of the economic reform, the ownership reform after 1992, there have been remarkable changes in the relation between executive compensation and firm performance compared to the first stage in 1978. It is expected that there is a more proper balance in the pay-performance structure in stage three, and that this is still developing.

According to Jensen and Meckling (1976), the agency theory focuses on the divergent interests and goals of the organization between the principals (the shareholders), and agents (the firm executives). The agency theory addresses two problems: the goals of the principle and agent diverge, and when the principal and agent have a different attitude towards risk. Moreover, it focuses on ways that employee long-term compensation can be used to

encourage the agents to act in the best interests of the organization. Therefore, it is predicted that firm performance is perfectly aligned with executive pay in order to mitigate agency problems and consequently, agency costs.

However, Bebchuk and Fried (2003) introduced two alternative views on the linkage between agency problems and executive compensation: the optimal contracting method, and

the managerial power approach. Both views are seen as mutually adjoining and not as a replacement of the other view.

On the one hand, the authors argue that the optimal contracting method is in line with the agency theory: they provide incentives to managers to maximize the shareholders’ value. Moreover, equity-based incentives strengthen the pay-performance relation, while the link between cash compensation and firm performance is weak. However, this method argues that there are no “perfect” contracts due to constraints in designing optimal contracts. The authors have highlighted factors that might negatively influence the pay-performance relation, such as: executives prefer stock options to cash compensation due to the higher value. It is hard to determine whether the stock prices rise due to the performance of executives or because of general market trends which occur in all firms of that certain industry, and executives can easily play around with the equity incentives by unloading it in order to earn more options to keep the pay-performance sensitivity in balance. Main factors that influence the executive compensation design in arm’s length contract mentioned by Bebchuk et al. (2002) and Bebchuk and Fried (2003) are the directors’ dependence on the executives and market forces that might distort optimal contracting outcomes. Consequently, the incentives that were originally introduced to align the interest of the principle and agent might be violated. As a result, it is argued that it is impossible to completely remove agency costs. This approach therefore does not try to completely remove agency costs, but instead focuses on minimizing the agency costs.

On the other hand, Bebchuk and Fried (2003) introduce a contrasting view as opposed to the agency theory and the optimal contracting method, namely the managerial power approach. This approach argues that the compensation scheme design itself forms part of the agency problem as mentioned by the other two views. According to the authors, powerful managers can exert power on the directors that determine their compensation to infer their own compensation design. Consequently, this results in “renting” from the company. In other words, the powerful managers can steal from the company; this distorts the link between firm performance and compensation. Bebchuk and Fried (2003) perceive the “outrage costs” as the most important factor in the managerial power approach. Because the approval of executive compensation ultimately depends on how outsiders, notwithstanding whether it favors the executives and does not favor the shareholders, perceive the executive

Thomas and Martin (1999). Furthermore, “camouflage” is also an important factor in the managerial power approach to reduce the “outrageous” view of outsiders. Consequently, this might lead to a suboptimal incentive scheme and reduced firm performance due to other forces influencing the compensation design. Therefore, a link between firm performance and compensation is not predicted by the managerial power approach. Furthermore, the theory predicts that the pay will be less sensitive to firm performance in the following situation according to Bebchuk and Field (2003): when there is an ineffective and/or weak board; small outside shareholders; low amount of institutional shareholders and; there is protection for the managers due to antitakeover arrangements.

As shown in the literature review, China’s listed firms have undergone positive changes towards strengthening the board of the firm such as introducing mandated independent directors on the board (CSRC, 2001). As a result, the dependence of the directors on the executives as mentioned in the optimal contracting method diminishes, which in turn fosters arm’s length contracts and minimized agency costs. Furthermore, there is a growth in private owners as the largest shareholder of the firm who focus on increasing the firm value according to Xu (2004). There is also a growth in foreign institutional investors in China’s listed firms due to the removal of restrictions in issuing foreign shares. Specifically, the increasing number of foreign institutional investors might act as a countervail factor in the managerial power approach. Since the foreign institutional investors are profit maximizers, they will therefore monitor the activities of the board of the firms more strictly. Eventually, this might lead to an improved corporate governance, and ultimately the pay-performance sensitivity will improve. This perception has been proven by the research of Hartzell and Starks (2002): they have shown that higher ratio of institutional investors providing greater oversight on the board, using a sample of 1,914 firms between 1991 and 1997 listed on the S&P 500 index, S&P Midcap 400 index, and S&P Smallcap 600 index, have ultimately lead to stronger corporate governance and mitigation of agency problems. Consequently, this results in stronger pay-performance sensitivity according to the managerial power approach.

Based on the aforementioned theory and interpretations, extant researches that proven that in China’s listed companies cash compensation is related to firm performance (Kato & Long, 2006a; Firth et al., 2007; Buck et al., 2008; Chen et al., 2010), and The Code of Corporate Governance for Listed Firms in China where it is specifically stated in section 77 and

78 that the listed company should have a performance-based incentive in order to attract and maintain qualified employees, the following hypothesis is developed:

- Hypothesis 1: Executive compensation depends on firm performance.

Influence of foreign ownership structure on firm-performance sensitivity

According to the agency theory, diffuse ownership leads to less monitoring and influence over corporate decisions (Jensen & Meckling, 1976; Fama & Jensen 1983). Highly concentrated ownership will have the opposite effect and leads to stronger incentives to provide

supervision on managerial activities (Jensen and Warner, 1988). Shivdasani (1993) expects that firms with high concentration ownership, dominated by A-shares, will have fewer performance-based financial incentives as a motivator for CEOs. On the other hand, Firth et al. (2006) showed that in order to attract foreign investors to buy the shares, firms will more likely adopt international standards as well as international corporate governance and

business practices. This will foster a higher executive compensation as shown in the literature review. Furthermore, foreign investors will urge the firms to use performance-related

schemes, as is common in Western countries according to the authors. As mentioned in the literature review, equity-based compensation is legalized as of 2006 (CSRC, 2005b).

Researches of Murphy (1999) and Core et al. (2003) conducted in Anglo-American studies showed the importance of equity-based compensation in aligning the diverse interest of the shareholders and management of companies, which is consistent with the agency theory and optimal contracting method. The authors define incentive pay as the change in managerial wealth as a result of a change in the shareholder value, which is related to the share or equity of the individual. Both researches argue that equity paid to the executives is more important than cash pay: incentives directly related to the firm’s stock price gives the executive a greater incentive to act in best interest of the company to increase firm value related to their own wealth. Consequently, foreign institutional investors interpret equity compensation as a highly important motivator for executives to increase the firm value. Furthermore, as

mentioned in managerial power approach in the previous section, the foreign institutional investors are expected to strengthen the corporate governance of China’s listed firms by providing stricter oversight on the board. This will eventually weaken the managerial power to exert influence on their own pay-performance design. In turn, this will strengthen the link

between firm performance and executive compensation. The foreign investors can therefore act as a proxy for good corporate governance and, in turn, a more balanced pay-performance sensitivity.

Based on the aforementioned findings and the factors highlighted in the optimal contract method and managerial power approach that strengthen the pay performance sensitivity due to the improved board quality, higher amount of outside shareholders, and oversight

provided by foreign institutional investors as mentioned in the “Firm performance and executive pay” section, hypothesis two and three are as follows:

- Hypothesis 2: Executive compensation is higher when there are more foreign institutional investors.

- Hypothesis 3: The positive relation between firm performance and executive compensation is stronger in firms with more foreign institutional investors.

3. Data and research method 3.1 Data

To test the hypotheses, original data across all non-financial industries of firms listed on the Shanghai- and Shenzhen Stock Exchange are retrieved from China Stock Market and

Accounting Research (CSMAR) database developed by Shenzhen GTA Information Technology Company.

For this research, the firm performance at time t-1 is used, while the executive compensation is measured at time t. Since the compensation is a response of firms’ previous performance, this method will give more accurate results and is consistent with the research of Firth et al. (2007). Consequently, firm performance data is retrieved for the years 2005 and 2006, while the executive compensation data are retrieved for the years 2006 and 2007. Furthermore, data of the ownership structure are also retrieved over the years 2006 to 2007.

Even though Chinese firms fall under specific information disclosure regulations of the Shanghai- and Shenzhen Stock Exchange, there might be limited financial information

disclosure, especially in early years of the introduction of equity-based compensation. As a result, companies with missing data for more than one variable are filtered out. This method will reduce the data set; however, it will also increase the reliability and quality of the data for

the research. The original random sample size consists of 1,375 firms in 2006 and 1,456 firms in 2007. As shown in table 1, 52 and 122, the observations over 2006 and 2007 respectively with no data available on executive compensation, firm performance, and corporate

governance are excluded. Furthermore, a firm in the finance and insurance industry is excluded from the total sample in 2007. The random sample size consists of 1,323

organizations in 2006 and 1,333 organizations in 2007, which ultimately resulted in a total of 2,656 firm-year observations. Furthermore, all the variables used in this research are

winsorized to retain as much valuable information as possible, instead of trimming the data. Using 98 percent Winsorization, the top and bottom percentile are winsorized.

Within a timeframe of two years and using firms listed on the Shanghai- and Shenzhen Stock Exchange represented in different non-financial industries, reliable data can be

gathered that represent the relation between firm performance, executive compensation and the effect of foreign institutional investors on the pay-performance sensitivity. Moreover, this timeframe covers the period of high growth of A-shares trading from 2005 to 2007 due to the removal of restrictions as mentioned by Yeh et al. (2009), and the introduction of equity-based compensation by the end of 2005 (CSRC, 2005a).

In short, this research includes the effect of the period of the A-shares growth after the loosened restrictions and the introduction of equity-based compensation, which is more common to Anglo-American studies, by using original data of non-financial firms listed on the Shanghai- and Shenzhen Stock Exchange retrieved from CSMAR database over the years 2006 and 2007.

Sample selection Numbers of observation

Firms listed on the Shanghai- and Shenzhen Stock Exchange 2006 1,375

Firms listed on the Shanghai- and Shenzhen Stock Exchange 2007 1,456

Total 2,831

Excluding observation with 1 or more missing variable 2006 (52)

Excluding observation with 1 or more missing variable 2007 (122)

Excluding observation in the Finance and insurance industry (1)

Total 2,656

3.2 Measurement theoretical construct

On the basis of the research question, three theoretical constructs will be measured

empirically. The theoretical constructs are elaborated on in the next section, followed by the description of control variables.

Independent variable: Firm performance

The firm performance will be measured using return on assets (ROA), which is consistent with the paper of Core et al. (1999). ROA is measured as (operating profit + financial

expense)/total assets and is an accounting-based measure of performance.

Next, the stock-based performance measure, annual return with cash dividend

reinvested, is included using the previous closing price of the stock prior to its first trading day consistent with the research of Conyon and He (2011) and as stated by the CSMAR database where the data is retrieved from. This market-based performance measure is a forward looking measure, and more difficult to manipulate compared to an accounting measure, ROA, which is a backward looking measure according to Firth et al. (1999).

Lastly, firm performance proxy, Tobin-Q, is added. This included both the accounting- and market-based performance measure. This is measured as the ratio of market value of the firm to the book value of total assets (Chung & Pruitt, 1994; Yermack, 1996; La Porta, Lopez-de-Silanes & Shleifer, 1999).

Dependent variable: Executive compensation

Executive compensation is the dependent variable measure using the natural logarithm of compensation. The natural logarithm is used to reduce the influence of outliers, which may influence the results of the research. From 2006, it is mandatory to report the compensation of the top three executives as the sum of base salary, performance commissions, stock options, and other benefits according to the China Securities Regulatory Commission (2002; 2005). Therefore, the executive compensation is measured as the total compensation of the top three executives of the company. This is consistent with the method of Kato and Long (2006a).

Moderator: Foreign institutional ownership

The moderator might influence the strength of the relation between the independent and dependent variable. In this research, the effect of foreign institutional ownership on the pay-performance structure is investigated. Foreign institutional investors are represented by the ratio of B-shares of the total amount of shares of the company; this is a comparable

measurement to the paper of Croci et al. (2012).

Control variables

Control variables are added, which are kept constant during the research, in order to eliminate effects on the independent variable. In turn, this will increase the accuracy of the research.

First, the control variable state ownership is included. As mentioned in the research of Xu (2004), firms solely issuing A-shares are mostly dominated by one shareholder: the State. As a result, the State has a significant influence over the determination of executive

compensation. Therefore, it is expected that the executive compensation is lower in firms dominated by the State (Kato & Long, 2006a).

Second, the control variable firm size is controlled for. This is a widely used control variable where it is stated that a larger firm size is more positively related to higher executive compensation (Smith and Watts, 1992; Schaefer, 1998; Hall and Liebman, 1998; Hermalin and Wallace, 2001; Goh and Gupta, 2010). The firm size is measured as the market value of the total assets of the firm. Moreover, the natural logarithm transformation to control for potential outliers and biases will also be used; this is consistent with the aforementioned researches.

Third, directors’ independence is measured as the ratio of independent directors on the board, which is also used in the paper of Kato and Long (2006a). In the Code of Corporate Governance for listed companies in China (CSRC, 2001) it is stated that the performance assessment should be the main tool in defining the compensation of management personnel, while the performance assessment should be approved by the board of directors of the company. Consequently, it is important that the board of directors represent the

management and shareholders’ interests. In other words, too many dependent directors might only represent the shareholders’ interest, while too many independent directors might

only represent the management interest. Therefore, it is necessary to control for the independence of the directors.

Fourth, firm financial risk represents the degree of combined leverage (DCL), which consists of financial- and operating-leverage. A higher level of DCL reflects more fixed costs of a firm and is perceived as a riskier firm than firms with a lower level of DCL. According to Lippert and Moore (1994), and Lippert and Porter (1997), when there is a higher risk within companies, the executive compensation will be higher as well.

Fifth, industry of the firm is included as a dummy variable, because according to Duffhues and Kabir (2008), and Goh and Gupta (2010), the type of industry in which the firms operates significantly influences the executive compensation. The sample is segmented following the method of the paper of Hou et al. (2014). The authors have segmented the industries as presented by the China Securities Regulatory Commission (CSRC) (CSRC, 2006) in the “Guidelines on industry classification of listed firms”. Even though the Fama-French 12 industry classification is widely applied in Anglo-American studies, it is difficult to apply this to China’s listed firms due to the differently encoded industry codes compared to the four-digit codes (SIC codes). Therefore, this research will use the industry classification as provided in the guidelines on industry classification of listed firms issued by the CSRC (CSRC, 2006), which is specifically designed for China’s listed firms. Since a complete set of dummy variables will be collinear with the intercept in a linear regression, 11 industry dummies are included to control for industry fixed effects of 12 industry groups: Manufacturing; Agriculture, forestry, livestock farming, fishery; Mining; Electric power, gas and water production and supply; Construction; Transport and storage; Information Technology; Wholesale and retail trade; Real estate; Social service; Communication and Cultural Industry and; Comprehensive.

Lastly, a year dummy variable is included to control for year-level fixed effects of 2006 and 2007 (Kato & Long, 2006a), because temporary fluctuations in the dependent variable that are not captured by the independent variables might lead toprecluded variable bias.

An overview of the variables used in this research is presented in table 1. Please note that the Renminbi (RMB) is used as the currency for this research.

Executive compensation Abbreviation Definition

Executive compensation LNPAYit Natural log of the total compensation of top three executives

Firm performance

Return on assets ROAit-1 (Total profits + financial expenses)/total assets Annual market return (%) RMKTit-1 Annual return with cash dividend reinvested Tobin-Q TQit-1 Ratio of market value of the firm to the book

value of total assets Governance structure

Foreign institutional ownership FOREIGNit Ratio of B-shares of the total amount of shares of the company

Interaction terms

ROA*FOREIGN ROAit-1 *FOREIGNit (ROA-ROA MEAN)*(FOREIGN-FOREIGN MEAN) RMKT*FOREIGN RMKTit-1 *FOREIGNit (RMKT-RMKT MEAN)*(FOREIGN-FOREIGN

MEAN)

TQ*FOREIGN TQit-1 *FOREIGNit (TQ-TQ MEAN)*(FOREIGN-FOREIGN MEAN) Control variables

State ownership STATE Proportion of shares held by the State Firm size LNSIZE Natural log of market value total assets

Board size BOARD Number of directors on the board

Director independence INDEPENDENCE The ratio of the board comprised of independent directors

Firms financial risk LEVERAGE The degree of combined leverage (DCL) which consist of financial- and operating-leverage Industry 1 INDUSTRY1 Equals 1 for Agriculture, forestry, livestock

farming, fishery

Industry 2 INDUSTRY2 Equals 1 for Mining

Industry 3 INDUSTRY3 Equals 1 for Electric power, gas and water production and supply

Industry 4 INDUSTRY4 Equals 1 for Construction

Industry 5 INDUSTRY5 Equals 1 for Transport and storage Industry 6 INDUSTRY6 Equals 1 for Information Technology Industry 7 INDUSTRY7 Equals 1 for Wholesale and retail trade

Industry 8 INDUSTRY8 Equals 1 for Real estate

Industry 9 INDUSTRY9 Equals 1 for Social service

Industry 10 INDUSTRY10 Equals 1 for Communication and Cultural Industry

Industry 11 INDUSTRY11 Equals 1 for Comprehensive

Year YEAR Equals 1 for year 2007 and 0 otherwise

Table 2: Research variables

3.3 Research method

The relation between firm performance, executive pay, and foreign ownership variables are tested through six models and measured using multiple regression analysis with IBM SPSS software as presented next.

First, hypothesis 1 and 2 are translated as the relation between the independent variable firm performance (ROA, RMKT, TQ) and corporate governance (FOREIGN), dependent variable executive compensation (LNPAY), and the control variables (STATE, LNSIZE, BOARD, INDEPENDENCE, LEVERAGE, INDUSTRY DUMMIES, and YEAR DUMMY) using multiple

regression analysis. It is expected that the coefficient is positive and significant on the performance measures ROA, RMKT, and TQ and corporate governance measure FOREIGN. Therefore, model 1, 2, and 3 are formulated as:

Model 1:

Ln(PAY)it = α + β1 ROAit-1 + β2 FOREIGNit + β3 STATE + β4 LNSIZE + β5 BOARD + β6

INDEPENDENCE + β7 LEVERAGE + (IndustryDummies) + (YearDummy) +

ε

Model 2:

Ln(PAY)it = α + β1 RMKTit-1 + β2 FOREIGNit + β3 STATE + β4 LNSIZE + β5 BOARD + β6

INDEPENDENCE + β7 LEVERAGE + (IndustryDummies) + (YearDummy) +

ε

Model 3:

Ln(PAY)it = α + β1 TQit-1 + β2 FOREIGNit + β3 STATE + β4 LNSIZE + β5 BOARD + β6

INDEPENDENCE + β7 LEVERAGE + (IndustryDummies) + (YearDummy) +

ε

Lastly, hypothesis 3 regarding the influence of corporate governance, in the form of foreign institutional investors, on the pay-performance sensitivity is investigated including the control variables (STATE, LNSIZE, BOARD, INDEPENDENCE, LEVERAGE, INDUSTRY DUMMIES, and YEAR DUMMY) using the multiple regression analysis. In order to incorporate the

moderating effect of foreign institutional investors in the first three models, the interaction term “(Xit-1 * FOREIGNit)” is added to model 1 – 3, where “X” is represented by either ROA, RMKT, or TQ subsequently. Moreover, based on hypothesis 3, it is expected that the

coefficient on the interaction term is positive and significant. Consequently, the following models are formulated:

Model 4:

Ln(PAY)it = α + β1 ROAit-1 + β2 FOREIGNit + β3 (ROAit-1 * FOREIGNit) + β4 STATE + β5 LNSIZE + β6 BOARD + β7 INDEPENDENCE + β8 LEVERAGE + (IndustryDummies) + (YearDummy) +

ε

Model 5:

Ln(PAY)it = α + β1 RMKTit-1 + β2 FOREIGNit + β3 (RMKTit-1 * FOREIGNit)+ β4 STATE + β5

LNSIZE + β6 BOARD + β7 INDEPENDENCE + β8 LEVERAGE + (IndustryDummies) +

(YearDummy) +

ε

Model 6:

Ln(PAY)it = α + β1 TQit-1 + β2 FOREIGNit + β3 (TQit-1* FOREIGNit) + β4 STATE + β5 LNSIZE +

β6 BOARD + β7 INDEPENDENCE + β8 LEVERAGE + (IndustryDummies) + (YearDummy) +

ε

In all six models the following holds: α: stands for the constant

i: stands for the company at time t

β: stands for the coefficient of the independent variable, control variable, and interaction

terms

ε: stands for the error term; part of the dependent variable not explained by the independent variables in the model

4. Results 4.1 Sample

The sample of the research consists of 2,656 firm year observations, which is divided over 12 industries. As shown in table 3, the majority of the observations are represented in the manufacturing industry with 58.60 percent of the total observations. This is comparable to the industry distribution of the research of Kato and Long (2006a), and Conyon and He (2011), where 62 percent and 57 percent respectively of the sample are represented in the

manufacturing industry. In this research, the manufacturing industry is the reference group of the industry control.

Year Observations Percentage

2006 1,323 49.80%

2007 1,333 50.20%

Total 2,656 100%

Industries Observations Percentage

Agriculture, forestry, livestock farming, fishery 62 2.33%

Mining 43 1.62%

Electric power, gas and water production and supply 116 4.37%

Construction 56 2.11%

Transport and storage 115 4.33%

Information Technology 161 6.06%

Wholesale and retail trade 177 6.66%

Real estate 117 4.41%

Social service 76 2.86%

Communication and Cultural Industry 17 0.64%

Manufacturing 1,557 58.62%

Comprehensive 159 5.99%

Total 2,656 100%

Table 3: Sample distribution

4.2 Descriptive statistics

Table 4 represents the descriptive statistics of dependent-, independent- and control-variables. The sample minimum and sample maximum are replaced with the 1st _{and 99}th

percentile, since 98 percent Winsorization is applied to all the variables used in the research. Furthermore, the 25th_{, 50}th_{, and 75}th _{percentile are also presented in the table, as well as the}

mean and standard deviation.

As shown in table 4, the top three executive pay of the sample of the research is RMB 798,420 on average between 2006 and 2007. Compared to the research of Kato and Long (2006a), and Conyon and He (2011), where the average compensation of the top three