Corporate governance in China : the moderating effect of internationalization and board diversity on the relation between ownership structure and corporate performance

Corporate Governance in China: The moderating effect of

internationalization and board diversity on the relation between

ownership structure and corporate performance

Master Thesis Business Administration – Track International Management

Name: Zhi Li

Student number: 11089660

Date: 23-06-2016

Thesis supervisor: Dr. Ilir Haxhi

Second reader: Dr. Niccolo Pisani

State of Originality

This document is written by Zhi Li. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

Ownership structure has become an important topic in corporate governance field. However, the previous findings toward the relationship between ownership structure and corporate performance are inconsistent. Therefore, in this study, we aim to test the moderating effect of internationalization degree and board diversity on this relationship. The sample in this study consists of 400 listed corporations in China. By collecting and analyzing related data from different sources, we find that the percentage of state-owned shares in the corporate has a significantly negative influence on return on equity (ROE). Moreover, internationalization degree and board diversity also have different moderating effects on the relationship. Internationalization degree strengthens the negative relationship between state ownership and corporate performance and the positive relationship between family ownership and corporation performance. Board diversity strengthens the positive relationship between family ownership and corporate performance.

In this study, we provide a more concentrated research of the corporate governance in Chinese market. Based on the empirical results, we suggest that Chinese public corporations be more aware of the effects of ownership structure, internationalization degree and board diversity on corporate performance.

Keywords: Corporate governance, agency theory, ownership structure, corporate performance,

Table of Contents

2.2 Relation Between Ownership Structure and Corporate Performance ... 10

2.3 Dimensions of Ownership Structure ... 12

2.4 Moderators ... 14 2.4.1 Internationalization ... 14 2.4.2 Board Diversity ... 15 3. Theoretical Framework ... 17 3.1 Theory Background ... 17 3.2 Hypotheses Development ... 18

3.2.1 Ownership Structure and Corporate Performance ... 18

3.2.2 Internationalization ... 20 3.2.3 Board Diversity ... 21 3.3 Conceptual Model ... 23 4. Method ... 24 4.1 Sample ... 24 4.2 Data Collection ... 24 4.3 Dependent Variable ... 24 4.3.1 ROE ... 25 4.3.2 Tobin’s Q ... 25 4.4 Independent Variables ... 26 4.5 Moderating Variable ... 26 4.6 Control Variables ... 26 4.7 Method ... 29

5.Analyses and Results ... 31

5.1 Descriptive Statistics ... 31 5.2 Correlations ... 35 5.3 Regression Analysis ... 37 5.3.1 ROE regression ... 37 5.3.2 Tobin’s Q regression ... 39 5.4 Moderation ... 41

5.4.1 ROE as Dependent Variable ... 43

5.4.2 Tobin’s Q as Dependent Variable ... 48

6. Discussion ... 54 6.1 Findings ... 54 6.2 Limitations ... 58 6.3 Future Research ... 59 7. Conclusion ... 60 8. Reference List ... 63

1. Introduction

With the deepening of economic globalization and the continuous technological progress, the internal and external market environments have undergone tremendous changes for enterprises in the fierce market competition, which force the enterprises to face unprecedented impact and challenges in the corporate governance process. Corporate performance becomes a direct reflection of the potential for business development and growth as an evaluation criterion for the listed companies. There have been many studies between corporate governance and corporate performance up to now. Most of these empirical studies have focused on the particular aspects. For example, Karpoff et al. (1996) have studied the shareholders’ activism; Sundaramurthy et al. (1998) have studied the anti-takeover provisions; Millstein and McAvoy (1998) have put the focus on the board characteristics; Bhagat et al. (1999) have studied compensation to outside directors. They have come up with inconsistent conclusions on the relationship between the particular aspect and corporate performance. However, most of the relevant studies focus on corporate governance in developed markets, especially in the US equity markets. While the studies on corporate governance in emerging markets, especially the equity market in China, still remain underexplored. According to Bai et al. (2004), corporate governance is paramount in China. Specifically, Chinese firms have unique characteristics of ownership structure, which is, according to Bai et al. (2004), one of the important facets of corporate governance, providing interesting settings to study corporate governance and the effects on firm performance. Bai et al. (2004) find that ‘both high concentration of non-controlling shareholding and issuing shares to foreign investors have positive effects on market valuation, while a large holding by the largest shareholder, the CEO being the chairman or vice chairman of the board of directors, and the largest shareholder being the government have negative effects.’ Domestically, Sun (1999), Xu and Wang (1997) have also conducted

empirical studies based on evidence from China and have come up with the conclusion that ownership structure has significant influence on the performance of listed firms in China.

From previous domestic literatures we find that in order to study the corporate governance in China, it’s important for us to firstly learn the evolution of Chinese market. In 1950s, Chinese enterprises carried out the socialism reform of the non-state-owned enterprises and began to implement the centralized planned economy under the strict management of the country. In 1980s, Chinese government began to carry the reform of the contract system of state-owned enterprises in a large scale by encouraging self-employed, which was the embryonic form of private economy, and started to open to the outside world and introduced a huge amount of foreign investments into China. In 1990s, Chinese market changed from planned economy to planned commodity economy and, in the end, to socialist market economy. Ownership structure changed from only state-owned companies to a mutual growth of state-owned, private-owned and foreign-owned companies. In 1998, over 800 companies were listed on the two national stock exchanges, Shenzhen Stock Exchange and Shanghai Stock Exchange. These companies are typically owned by five groups of shareholders: the state, legal persons (institutions), tradable A-share holders (mostly individuals), employees, and foreign investors (Xu and Wang, 1999). The first three groups are the main shareholders, each controlled roughly 30% of the whole shares (Xu and Wang, 1999). The principal aims of the reforms include the modernization of industry, stimulation of growth, reduction of poverty, and improvement in economic efficiency (Chen et al., 2006). To implement these reforms policies, China has moved towards a free-enterprise system that includes, among other things, the privatization of state-owned companies, the formation of joint stock companies, and the development of stock markets (Chen et al., 2006). Hence, in order to learn corporate governance and firm performance in Chinese market, it’s significant for us to focus on the ownership structure of the public companies in China.

The relationship between ownership structure and corporate performance has always been an important debate in the corporate finance literatures and has been the subject of numerous studies dating back to the work of Berle and Means, The Modern Corporation and Private Property (1932), which suggests that there is an inverse correlation between the diffuseness of shareholdings and firm performance. Demsetz (1983) challenges this result by considering that market discipline forces managers to adhere to value maximization at very low levels of ownership. He argues that we should see corporate ownership structure as an endogenous outcome of the decisions, which reflect the influence of shareholders and trading on the market for shares. Demsetz and Lehn (1985) provide evidence of the endogeneity of a corporate ownership structure argued by Demsetz (1983). In their later study, they assess the validity of the Berle and Means thesis (1932): ‘A linear regression of an accounting measure of profit rate on the fraction of shares owned by the five largest shareholding interests (and on a set of control variables), in which ownership structure is treated as an endogenous variable, gives no evidence of a relation between profit rate and ownership concentration.’ In line with the study of Demsetz and Lehn (1985), Morck et al. (1988) ignore the endogeneity issue and re-explore the relationship between ownership structure and corporate performance and find no significant relation in the linear regressions by using Tobin's Q and accounting profit rate as alternative measures of corporate performance. Nevertheless, they estimate a piecewise linear regression of Tobin's Q on insider ownership, which provides an evidence of a non-monotonic relation. According to Demsetz and Villalonga (2001): ‘The ownership structure that emerges, whether concentrated or diffuse, ought to be influenced by the profit-maximizing interests of shareholders, so that, as a result, there should be no systematic relation between variations in ownership structure and variations in firm performance.’ This finding is consistent with the view that ownership is diffused, while it may exacerbate some agency problems, and it also yields compensating advantages that generally offset such problems. Hence, the empirical

studies about the relation between ownership structure and corporate performance seem to yield the conflicting results.

The above views are supported by many empirical studies, most of the which have been performed using data from the developed markets, such as United States, few others are based on Singapore or Hong Kong evidence. Due the the lack of study in Chinese market field and inconsistent research outcomes, the purpose of this study is to empirically examine the relation between ownership structure and firm performance in Chinese market based on the studies we mentioned before. Hence, in order to fill this gap, we raise the research question of this study:

How can ownership structure influence the corporate performance of the listed firms in China?

In order to gain a clarification on the inconsistent results of previous studies, different researchers suggest that moderator variables between ownership structure and corporate performance should be examined to find out when and how ownership structure influence corporate performance. According to Glaum and Oesterle (2007), understanding the relation between internationalization and corporate performance remains a core issue. Some researchers, such as Rugman (1981), and Tallman and Li (1996) indicate that internationalization is linked to producing competitive advantages at the firm level and to superior financial success.

According to Kling and Weitzel (2011), a conspicuous issue in the internationalization of Chinese corporates is that many of them are state-owned enterprises (SOEs) and corporate governance in China is highly idiosyncratic. Hence, internal and external corporate governance mechanisms can be very different. Based on the introduction to the evolution of Chinese market previously and the gap that there are so few of researches on internationalization of Chinese public companies, we consider the degree of internationalization as an important element when

studying the relation between ownership structure and corporate performance. Thus, we raise the first sub-question based on the main research question:

How does the internationalization degree affect the relation between ownership structure and corporate performance in Chinese listed companies?

As Hermalin and Weisbach (1991) have discussed, the board of directors is the most important decision-making body in a corporation and there seems to be a trend of increasing diversity within boards. Board diversity can be accomplished at many dimensions like ethnicity, age, gender, experience, education and background. However, few of the previous literatures have discussed the influence of board diversity on the relationship between ownership structure and corporate performance in Chinese market. In order to fill this gap, we raise the second sub-question based on the main research sub-question:

How does board diversity affect the relation between ownership structure and corporate performance in Chinese listed companies?

In this study, we select the sample of 400 listed firms from Shanghai, Shenzhen, Hong Kong and Taiwan stock exchange markets for the top 100 listed companies respectively in year 2014 to study the relation between ownership structure and corporate performance.

The empirical results show that when company has a higher degree of internationalization, the higher the percentage of state ownership it has, the poorer performance it makes when the performance is measured in Tobin’s Q. When company has a lower degree of internationalization, the higher the percentage of family ownership it has, the better performance it makes when the performance is measured in both ROE and Tobin’s Q. And when board of directors of a company has a higher level of education background, the higher percentage of family ownership it has, the better performance it makes when the performance

the moderating effect of education background on the relation between state ownership and corporate performance.

This study makes several contributions as follows. First, we extend the previous corporate governance literatures by intentionally analyzing the the corporate governance in emerging market, specifically Chinese market. Second, we add to the previous Chinese corporate governance literatures by analyzing the moderating effects of internationalization and board diversity. In particular, the empirical results of this study prove that internationalization and board diversity do have different moderating effects on the relation between ownership structure and corporate performance.

We structure this paper as follows. First, we elaborate the previous literatures on ownership structure and corporate performance. Then, based on the literature review, we develop the hypotheses and give the method of this research. We will analyze and make discussion in the following section. In the final section, we show the limitations of this study and conclude this paper.

2. Literature Review

2.1.1 Corporate Governance

The broad definition of corporate governance is that it broadly refers to the mechanisms, processes and relations by which corporations are controlled and directed. La Porta (1999) concludes corporate governance as ‘a set of mechanisms through which outside investors protect themselves against expropriation by the insiders.’

In recent years, China has made a great progress on improving its governance system and has pushed the governance to a higher level for the listed companies. Meanwhile, the foundation of the capital market has been consolidated, promoting a steady development. According to the report from Organization for Economic Co-operation and Development (OECD) in 2011, corporate governance in China has been explored and established in the process of state-owned enterprises reform and private enterprises growth. Although China has started the creation of a legal system for corporate governance rather lately, the system has developed fairly quickly and increasingly full-fledged (OECD, 2011).

Corporate governance in China has been explored and established in the process of state-owned enterprises reform and private enterprises growth. The experience and the model of corporate governance with Chinese characteristics have come into being in light of the actual situation in China (OECD, 2011). It has developed under the joint effort of the government and market participants, with the former playing a leading role in the construction and improvement of the corporate governance legal framework (OECD, 2011). However, in the transition period of China’s economy, limited by the specific political, legal and economic system of internal and external environments, there are still many companies which are only in the form to meet company system reform requirements instead of establishing incentive and supervision of corporation governance structure. Thus, Chinese corporations still need to pay attention to

corporate governance. Berle and Means (1932) view corporate governance as a classical agency problem: ‘How could corporate managers, as agents of the shareholders, be induced to manage corporate assets in the best interests of their principals?’ Hence, another way to understand corporate governance is to figure out the what agency theory is, which we will discuss in the following section.

2.1.2 Agency Theory

Agency theory is always used to explain when two parties have diverging interests and asymmetric information. According to Jensen and Meckling (1976), and Fama and Jensen (1983), the agency problem is an essential element of the so-called contractual view of the firm. The essence of the agency problem is the separation of management and finance, or of ownership and control in a more standard terminology. Jensen and Meckling (1976) define an agency relation in their study as ‘a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating some decision making authority to the agent and if both parties to the relation are utility maximizers, there is good reason to believe that the agent will not always act in the best interests of the principal.’ The principal can limit divergences from the interest by establishing appropriate incentives for the agent and by incurring monitoring costs designed to limit the aberrant activities of the agent (Jensen and Meckling, 1976). In most agency relations, the principal and the agent will incur positive monitoring and bonding costs (non-pecuniary as well as pecuniary), and in addition there will be some divergence between the agent’s decisions and those decisions which would maximize the welfare of the principal (Jensen and Meckling, 1976). Hence, agency cost is defined by Jensen and Meckling (1976) as the sum of:

1. the monitoring expenditures by the principal, 2. the bonding expenditures by the agent, 3. the loss of residual.

Since the relation between the stockholders and the managers of a corporation fits the definition of a pure agency relation, it should come as no surprise to discover that the issues associated with the “separation of ownership and control” in the modern diffuse ownership corporation are intimately associated with the general problem of agency (Jensen and Meckling, 1976).

2.2 Relation Between Ownership Structure and Corporate Performance

Ownership structure is the basis of the internal control system arrangement, which is the instrument to maintain the benefits of shareholders. The analysis of the effectiveness of ownership structure for the internal control mechanisms in depth finds the effectiveness from three aspects, which are the shareholders to the operators, the operators to managers and managers to employees. The stability of internal control system is one of the determinisms of corporation performance, hence there is a strong relation between ownership structure and corporation performance. Some researchers analyze the relation between ownership structure and corporation performance in China. Shleifer and Vishny (1994) state ownership structure as the concentration in a firm’s ownership and the identity of a firm’s owners, and has an influence on firm’s corporate governance, its strategies, and its performance.

According to Xu and Wang (1999), a typical listed stock company in China has a mixed ownership structure with three predominant groups of shareholders—the state, legal persons (institutions), and individuals—each holding approximately 30% of the stock. Ownership is heavily concentrated. In their study, there is a positive and significant correlation between ownership concentration and profitability. And firm profitability is positively correlated with the fraction of legal person shares, but it is either negatively correlated or uncorrelated with the

fractions of state shares and tradable A-shares are held mostly by individuals. Then, labor productivity tends to decline as the proportion of state shares increase. Thus, as a conclusion, ownership structure significantly affects the performance of publicly listed companies in China within the framework of corporate governance.

By examining the cross-sectional relation between ownership structure and corporate performance of a sample of 434 manufacturing firms listed on the Chinese stock exchange, Chen (2001) finds a strong relation between ownership concentration and corporate performance that the percentage of state-owned shares plays a negative role in corporate governance, while domestic institutional and managerial shareholdings improve the firm performance. According to Chen and Cheung (2005), concentrated ownership is not associated with better operating performance or higher firm valuation in Hong Kong. They also find a negative relation at higher level of ownership concentration, which can attribute to managerial entrenchment, since managerial shareholdings confer to management, among other benefits, protection against hostile takeovers (Morck et al.,1988; McConnell and Servaes, 1990; Hermalin and Weisbach, 1991). Demsetz (1983) states that the separation of ownership and control in the modern corporation retains a central position of the economic theory of the corporation. Following Jensen and Meckling (1976), interest in the relation between corporate value and the allocation of shares among managers and non-managers has continued to evolve on both the theoretical and the empirical front. Hence, we can say that there is a relation between ownership structure and corporate performance.

However, in the study of Demsetz (1983), he suggests that there is no relation between ownership structure and profitability. He argues the study of Berle and Means (1932), who state that the separation of ownership and control, or specialized ownership, which characterizes the modern corporation, impairs the ability of the profit motive to encourage the most efficient use of resources, that this idealization seldom holds true for firms in general,

even for those owner-managed firms because ‘real world businesses often permit on-the-job consumption by managers and employees.’ In competitive markets, this permitted consumption is traded for reduced monetary compensation because ‘on-the-job consumption will take place only when off-the-job consumption would cost more; on-the-job consumption may not represent the dissipation of resources that critics of specialized ownership have contended’ (Demsetz, 1983).

Based on previous studies, it’s hard to state that there is a relation between ownership structure and corporate performance when there are no certain limitations of country, company size, industry and other standards during the study. Thus, it’s necessary to conduct different variables to restrict the study on different firms.

2.3 Dimensions of Ownership Structure

According to Delios (2008), the growth and heightened competitiveness of listed companies in China share several central features, which include ‘the gradual transition of state-owned assets to private investors, a rapid pace of product diversification, and impending rapid growth into international markets.’ Delios (2008) put the focus on measuring and identifying the implications of the ownership structure and diversification strategy of listed companies in China.

As mentioned before, a listed company in China has several types of shareholders: the state, legal persons, employees, individual domestic owners of A shares, and foreign private owners of B shares. Companies can issue shares in Hong Kong exchange, where such shares are called H shares. Chinese law has defined the categorization of shareholders in official documents and regulations. It is designed to seize the shifts in ownership of those companies from state shareholders to private shareholders of three types: legal person, employee, and private investors (A share, B share, and H share owners). More specifically, state shares are

the shares held by the central government, local governments, or solely-government-owned enterprises (Delios, 2008). Legal person shares, which are roughly analogous to the institutional share-holder identity as recognized in the United States, are shares owned by domestic institutions which are either independent from or partially owned by the central or local government (Delios, 2008). A shares are similar to ordinary equity shares as generally accepted in other equity markets, except that they are exclusively available to Chinese citizens and domestic institutions. B shares are also similar to ordinary equity shares, but they are available to foreign individuals and institutions. For most of the listed companies, the top 10 shareholders are normally the state and legal persons (Delios, 2008). This ownership categorization scheme is always used as the official one mandated by regulators in China. This scheme effectively classifies the shareholders of a firm into three dominant groups – state, legal person, and tradable A and B shares – which each hold approximately 30% of the total stock in these listed companies (Sun et al., 2002). However, this official categorization does not point out exactly who owns this corporation. Instead, it provides an approximation of the shift in ownership from the state to private individuals. Moreover, the owner categories have substantially ambiguous definitions and memberships, which make it difficult to determine both the identity of owners and what might be their consequent motivations and skills to manage the firm (Delios, 2008).

According to Chu (2009), by investigating the influence of founding-family ownership on the return on assets and Tobin’s Q of 341 public small- and medium-sized enterprises (SMEs) in Taiwan during the period of 2002–2006, the author finds that the influence of family ownership on SMEs performance is positive and significant so that family ownership is an effective organizational structure for SMEs in Taiwan. We can also find in the previous literatures that older firms are more likely to be family controlled, as are smaller firms in Taiwan (Stijn and Simeon, 2000). There is also a number of case studies that describe the

ownership and control structures of some of the largest business groups in East Asian countries: Taylor (1998) for the Li Ka-shing group in Hong Kong, and Numazaki (1993) for the Tainanbang Group in Taiwan. Anderson and Reeb (2003) investigate the relation between founding-family ownership and firm performance and find out that family firms perform better than nonfamily firms, suggesting that family ownership is an effective organizational structure.

Based on previous studies, this sturdy will mainly focus on two types of ownership: state ownership and family ownership.

2.4 Moderators

2.4.1 Internationalization

Glaum and Oesterle (2007) find that the question of whether there is a systematic relationship between the internationalization of corporations and their performance is central to the field of international business. The empirical studies have come to heterogeneous, sometimes contradictory results.

The early researchers find a positive internationalization-performance relationship (Chen and Tan, 2012; Buhner, 1987; Geringer et al., 1989; Grant, 1987; Rugman et al., 1985), but later studies state a negative relationship (Michel and Shaked, 1986; Collins, 1990; Shaked, 1986; Kumar, 1984). Addressing those contradictory findings, researchers from the 1990s elaborate both the benefits and the costs, and suggest a curvilinear relationship (e.g. Geringer et al., 1989; Gomes and Ramaswamy, 1999; Hitt et al., 1997; Mauri and Sambharya, 2001). Ruigrok and Wagner (2003) then obtain a standard U-shape relationship between internationalization and firm performance. Contractor et al. (2003), and Lu and Beamish (2004) have independently suggested that all these contradictory findings might be reconciled in a new three-stage theory of international expansion in which positive, negative and U-shaped relationships can be found at different stages. In stage 1, the relationship is negative slope as

costs and barriers to initial international expansion outweigh the benefits. In stage 2, the relationship becomes positive as geographical expansion makes possible efficiencies that improve resource utilization, better scanning for market opportunities, the ability of some companies to exercise global market power and to extend the product cycle. In stage 3, the relationship becomes negative again as some firms expand beyond an optimal threshold.

2.4.2 Board Diversity

Loden and Rosener (1991) define diversity into two dimensions, in which they define the primary as: ‘Those exerting primary influences on our identities, are gender, ethnicity, race, sexual orientation, age and mental or physical abilities and characteristics.’ In their theory the primary dimensions shape the fundamental self-image and the views toward the world. Moreover, primary dimensions usually have the most significant impact on groups in both the workplace and the society. Loden and Rosener (1991) define another dimension of diversity defined as secondary dimension, which is less visible and, according to them, ‘exert a more variable influence on personal identity and add a subtler richness to the primary dimensions of diversity.’ Secondary dimension includes the following elements: educational background, geographic location, religion, first language, family status, work style, work experience, military experience, organizational role and level, income and communication style. The secondary dimension has an impact on self-esteem and self-definition. Within the empirical corporate governance research done by them, we can confirm that the board diversity plays a significant role regarding to the economic effects. Cross et al. (1994) declare the definition of board diversity in his study as: ‘Diversity is viewed as focusing on issues of racism, sexism, heterosexism, classism, ableism, and other forms of discrimination at the individual, identity group, and system levels.’ Cox (1993), instead, put the focus on cultural diversity, for which he defines as ‘the representation, in one social system, of people with distinctly different group

affiliations of cultural significance.’ In line with these studies, O’Reilly et al. (1998) define that: ‘A group is diverse if it is composed of individuals who differ on a characteristic on which they base their own social identity.’ Examples of very broad definitions include the definition by Thomas (1991): ‘Diversity includes everyone and it extends to age, personal and corporate background, education, function, and personality, instead of just something that is defined by race or gender.’

3. Theoretical Framework

The central question of this study is: How can ownership structure influence the performance of the listed firms in China? In order to answer this question, we develop several different hypotheses, and will test and analyze them in the following sections. By studying on the analysis results of these hypotheses, we would like to answer the research question and its sub-questions in a later section of this thesis. Like we have mentioned before, the agency theory is the most important theory in this study. The link between ownership structure and corporate performance can be explained by agency theory, which is also used to explain when two parties have diverging interests and asymmetric information (Fama and Jensen, 1983). According to Jensen and Meckling (1976), the agency problem is an essential element of the so-called contractual view of the firm. The essence of the agency problem is the separation of management and finance, or of ownership and control in a more standard terminology. They also define an agency relation in their study as ‘a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating some decision making authority to the agent and if both parties to the relation are utility maximizers, there is good reason to believe that the agent will not always act in the best interests of the principal.’ As discussed before, for the past few years, China has made huge progress on improving its governance system and has pushed the governance to a higher level for listed companies. At the same time, the foundation of the capital market has been consolidated, which promotes a steady development. According to OECD (2011), corporate governance in China has been explored and established in the process of state-owned enterprises reform and private enterprises growth. Corporate governance experience and model with Chinese characteristics have developed under the joint effort of the government and market participants (OECD, 2011). Based on the study of Berle and Means (1932), corporate

governance is a classical agency problem and is discussed as: ‘How could corporate managers, as agents of the shareholders, be induced to manage corporate assets in the best interests of their principals?’ Hence, based on the theoretical foundations, we analyze the ownership from two different aspects: state ownership and family ownership. And we will discuss the relationship between these two ownership structures and corporate performance in the following sections.

3.2 Hypotheses Development

3.2.1 Ownership Structure and Corporate Performance

By reviewing the previous literatures, this relationship between ownership structure and corporate performance is not consistent. Berle and Means (1932) observe an inverse correlation between the diffuseness of shareholdings and firm performance. However, Demsetz (1983) argues that the ownership structure of the firm should be seen as an endogenous outcome of the decisions. Then, in the study of Morck et al. (1988), they find no significant relation in the linear regressions of this relation by using Tobin's Q and accounting profit rate as alternative measures of corporate performance. Thus, the empirical studies on the relation between ownership structure and corporate performance yield conflicting results. However, most of the previous studies are about developed markets, such as United States. For the recent two decades, the strong economic development of China has been a dominant story in the business press. According to Delios (2008), what has been underlying and spurring this growth is the transition of state ownership to private ownership. Via these avenues, formerly state-owned firms begin to shift ownership shares to various types of private owners, ostensibly independent from the state. While in many transition economies the shift from state to private ownership has been accomplished in one swift, wholesale transfer of formerly state-owned assets, the ownership transition in China has been gradual, albeit inexorable, according to Shleifer and Vishny (1994).

In their study, they discuss that what underlies this shift is the fact that the motivations, goals, and capabilities of a company are strongly related to the identity of its owners and how widely-held or dispersed is the shareholding in the company. They also conclude that companies that have widely-held shares by private investors, or those owned by institutional shareholders, tend to have a greater focus on shareholder wealth maximization strategies than those companies that have a dominant majority owner, or are state-owned (Shleifer and Vishny, 1994).

Some researchers have tested the impacts of state ownership on the performance of listed corporations in China. Those studies find that there is a U-shape relation between state ownership and firm performance, which reveals in most situations that state ownership has negative influences on firm performance (Xu and Wang, 1997). In this study, ROE and Tobin’s Q are selected to measure the corporate performance. These two measurements respectively indicate the ability of a firm to generate profits from its shareholders’ investments in the company and the value of the whole market in ratio to the aggregate corporate assets. Therefore, it’s expected that state ownership negatively affects the corporate performance. Hence, we conduct the first hypothesis:

H1. State ownership has a negative effect on corporate performance.

There is a gap in the research on the relation between family ownership and corporate performance since there are few related researches in Chines market. According to Luo (2014), in the study on whether family-concentrated ownership will enhance the corporate performance, they find a significant inverse U-shaped relation between the controlling family’s ultimate cash-flow rights and corporate value, as measured by Tobin’s Q. They conclude that as family-ownership concentration increases, corporate value first increases and then decreases. Anderson and Reeb (2003) investigate the relation between family-owned corporation and corporate performance, and then find out that family ownership is prevalent and substantial.

By studying the family-owned corporations of the S&P 500 and discovering that family corporations perform better than nonfamily corporations, they suggest that family ownership is an effective organizational structure. Miller and Le Breton-Miller (2011) also test and verify this view in their study. Therefore, we may expect that family ownership positively affects the corporate performance. Hence, we propose the second hypothesis:

H2. Family ownership has a positive effect on corporate performance.

3.2.2 Internationalization

Welch and Luostarinen (1988) describe internationalization as the process of increasing involvement of international activity across borders. According to Oesterle et al. (2013), the amount of cross border activities of a firm can be described by its degree of internationalization (DOI). Hence, in this study, we measure the degree of internationalization for corporates, as a moderator, in the number of foreign subsidiaries. In the previous studies, the degree of internationalization is proved to have influence on the ownership structure and therefore influence the corporate performance. One school of thought argues that the distribution of ownership has important implications on corporate efficiency and strategic development (Marris, 1964; Pfeffer and Salancik, 1978; Williamson, 1964).

According to Ribeiro (2014), the public management theory that is always used to explain the socio-political existence of state owned enterprises (SOEs) doesn’t have a focus on the explanation of the internationalization strategies of SOEs. Also, according to him, efficiency-based international business theories of internationalization, such as Uppsala internationalization model and Dunning’s Eclectic Paradigm, don’t have a focus on the explanations of the internationalization strategies of enterprises like SOEs. This gap still exists, especially in Chinese market. Based on few previous studies, the main school of international business theories have assumed that state ownership reduces the likelihood that a company will

expand abroad (Vernon, 1979). This literature review provides a path for the development of the third hypothesis:

H3. When company has a higher degree of internationalization, the higher the percentage of state ownership it has, the poorer performance it will make.

According to Lien et al. (2005), the corporate governance characteristics of Taiwanese firms, such as different forms of ownership structure and the composition of the board of directors, affect their FDI strategies. Bhaumik, et al. (2010) conclude that family firms and firms with concentrated ownerships are less likely to invest overseas than other types of firms. Hence, we conduct the fourth hypothesis:

H4. When company has a lower degree of internationalization, the higher the percentage of family ownership it has, the better performance it will make.

3.2.3 Board Diversity

According to Loden and Rosener (1991), and Berghe and Levrau (2004), the consideration of diversity when selecting the board of directors is essential since board diversity can represent an efficient corporate governance (CG) mechanism of realizing efficient management and monitoring within corporations. Thomas (1991) states in his study that diversity includes everyone and it extends to age, personal and corporate background, education, and personality, instead of just something that is defined by race or gender. According to Eulerich (2014), in the nowadays global market with an intense and competitive environment, the role of education background of the board of directors has become more and more important since ‘it is proved that personnel with high level of education is a determining factor of the research and innovation process, thus facilitating improvements in productivity and competitiveness.’ Hence, in this study we will mainly study the influence of education background on the relation

We divide education level into five levels, which are secondary school and less, Applied College, Bachelor, Master and Doctor. We assume that in sate-owned enterprise, when the board of directors have a higher level of education background, the enterprise will have a better performance. Because a higher education level can contribute to more beneficial decisions for both shareholders and market value. Thus, we conduct the fifth hypothesis:

H5. When board of directors of a company has a higher level of education background, the higher percentage of state ownership it has, the poorer performance it will make.

In the literature studying period, we have found it difficult to develop the hypotheses about the influence of education level on the relationship between family ownership and corporate performance because of the lack of the evidence from China. However, according to Hermalin and Weisbach (1991), the board of directors is the most important decision-making body in a corporation. Board diversity can be accomplished at many dimensions like ethnicity, age, gender, experience, education and background. Hence, we assume that in family-owned enterprise, when the education level of board of directors is higher, the corporate performance will be better since it’s more likely for board of directors to make beneficial decisions for shareholders and for the long-term development of the enterprise. Thus, we conclude the last hypothesis:

H6. When board of directors of a company has a higher level of education background, the higher percentage of family ownership it has, the better performance it will make.

3.3 Conceptual Model

We summarize the hypotheses developed in the previous section into a conceptual framework which is presented below:

In graph 1, as discussed in previous sections, the conceptual framework shows the main research question of this study: How can ownership structure influence the performance of the listed firms in China? By reviewing the previous literatures, we put the focus on state ownership and family ownership. Since the previous empirical results are inconsistent, we add internationalization and board diversity as two moderators and study how these elements influence the relation between ownership structure and corporate performance.

State Ownership

Family Ownership

Internationalization Board Diversity

Performance H2 H3 H4 H5 H6 Graph1.Theoretical Framework

4. Method

We select top 100 listed firms from Shanghai Stock Exchange (SHSE), Shenzhen Stock Exchange (SZSE), Hong Kong Stock Exchange (HKEx) and Taiwan Stock Exchange(TSEC) respectively for a period of year 2014. The top listed firms have the highest market value among all other listed firms in each stock exchange and are the representatives of each industry, which is necessary to include since the industry will also be criteria when studying the relation between ownership structure and corporate performance.

4.2 Data Collection

Through the official online database of Shanghai Stock Exchange (SHSE), Shenzhen Stock Exchange (SZSE), Hong Kong Stock Exchange (HKEx) and Taiwan Stock Exchange(TSEC), we collected the data of ownership structure from the company annual financial reports for the independent variables, which include the percentage of state ownership and family ownership. Also, we collected the financial data related to ROE and Tobin’s Q from the official databases of the stock exchange mentioned above. To obtain data related to board diversity, foreign subsidiaries, industry and firm size, we have consulted the Bloomberg Database, Wind Database and Orbis Database. These databases provide the data for the number of foreign subsidiaries, education background of board of directors, number of foreign subsidiaries and the firm size, which is indicated by the number of employees.

4.3 Dependent Variable

In this study, corporate performance is the dependent variable. Corporate performance is defined as companies follow the law of value, by value as the core of management, all business stakeholders can obtain satisfactory return ability. Obviously, the higher the value of the

enterprise, the enterprise ability to give its stakeholders return is higher. Business performance is an enterprise management during certain operating efficiency and operator performance. The level of operating efficiency of enterprises mainly in terms of profitability, asset management levels, solvency and subsequent development capabilities. Hence, we use Tobin’s Q and ROE as measurements to estimate the dependent variable: corporate performance.

4.3.1 ROE

Return on equity (ROE) is the amount of net income returned as a percentage of shareholders’ equity. Return on equity measures a corporation profitability by revealing how much profit a company generates with the money shareholders have invested.

Net Income ROE=

Shareholder’s Equity

Net income is for the full fiscal year (before dividends paid to common stock holders but after dividends to preferred stock.) Shareholders equity does not include preferred shares. In this study, ROE is used as an indicator of corporation performance.

4.3.2 Tobin’s Q

Tobin’s Q is a ratio of company’s total market value and its total asset value, which was devised by James Tobin in 1969. It hypothesizes that the combined market value of all the companies on the stock market should be equal to their replacement cost.

Total Market Value of Firm Q Ratio=

Total Asset Value

Tobin’s Q is usually used as an indicator of market based performance for a corporation. Market measures refer to those measures, which incorporate the market value of the equity.

Tobin’s Q is forward looking and reflects the shareholders’ expectations regarding future performance of the firm, which is based on past or current performance. In studying ownership structure, this market measure is meant for testing the market value of the firm. In this study, we assume that valuation of firm is linked with firm’s ownership structure and its performance.

4.4 Independent Variables

The independent variable in this study is ownership structure. As mentioned in previous sections, the percentage of state ownership and family ownership are two dimensions we use, which respectively indicate the percentage of shares held by state and by family member, in order to discover whether different ownership structures will have different influences on firm performance.

4.5 Moderating Variable

We take the degree of internationalization of the corporations and the board diversity as two moderators in this study. We consider the number of foreign subsidiaries of the corporations as an indicator towards the degree of internationalization, and consider the level of education background of board of directors as the measurement of board diversity.

4.6 Control Variables

In this study, we use different levels of control variables. The control variable should include the aspects of three levels: country level, industry level, and firm level. However, in this study only one country is included, which is China. Hence, we will not discuss the control on country level in this paper.

On the industry level, different industries are consolidated into sectors to act as control variables. Industry is an important control variable since in the numerator of Tobin’s Q formula

the perception of the stock market is included which means that also intangible assets are covered. Some industries are more knowledge intensive than others, therefore, this study will have to control the industry.

Table 1 shows the way we divide the industry and the score we give to each category. As table 1 shows, we divide the industry into 16 sectors based on the NAICS code. In order to have a clear view on the control effect of industry, we then sort 16 sectors into 7 categories. We consolidate Agriculture, Forestry, Fishing and Hunting into Agriculture sector, which takes 0.5% out of total 400 firms; consolidate Mining, Quarrying, Oil and Gas Extraction, Utilities, Construction, and Transportation and Warehousing into Energy sector, which accounts for 19.75% of the total amount; Manufacturing occupies 48% of the total and is the largest group; Wholesales and Retail Trade takes 7.25%, and Information Technology (IT) takes 4.25% of the total amount respectively; Finance, Insurance, Real Estate, Rental and Leasing are put into Finance sector and occupies 14% of the total amount; Services, including Professional, Scientific, Technology Services, Administrative and Support and Waste Management and Remediation Services, Arts, Entertainment, Recreation, Accommodation, Food Serviced and Other services (except Public Administration), take 6.25% of the total amount. Hence, the sample consists of 0.5% agriculture firms, 19.75% energy firms, 48% manufacturing firms, 7.25% whole sales and retail trade firms, 4.25% IT firms, 14% finance firms and 6.25 % services firms. All firms in the sample get different scores which represent the industry they operate in.

On the firm level, we use firm size as a control variable since Kogut and Singh (1988) have proved that it to have effect on firm performance. And firm size is used in this study in line with what Marinova et al. (2010) have conducted in their studies as well, for which the number of employees is a measurement.

Score Industry NAICS Code Description Company Sum

1 Agriculture 11 Agriculture, Forestry, Fishing and Hunting 2 2

21 Mining, Quarrying, and Oil and Gas Extraction 21

22 Utilities 19

23 Construction 19

48-49 Transportation and Warehousing 20

3 Manufacture 31-33 Manufacturing 192 192

42 Wholesale Trade 22

44-45 Retail Trade 7

5 Information Technology 51 Information 17 17

52 Finance and Insurance 23

53 Real Estate and Rental and Leasing 33

54 Professional, Scientific, and Technical Services 10

56 Administrative and Support and Waste Management

and Remediation Services 5

71 Arts, Entertainment, and Recreation 5

72 Accommodation and Food Services 4

81 Other Services (except Public Administration) 1

7 6 4 2

Whole Sales and Retail Trade Finance Services Energy 79 29 56 25

4.7 Method

We test the proposed hypotheses via a regression analysis. And we use regression analysis when one or several independent variables are hypothesized to affect one dependent variable. Thus, regression analysis finds the best fitting straight line through a set of points. This relation can be expressed in the following equation:

! = "

+ "

#

+ "

*"

#

+ "

*"

#

+$

In this case, Y represents the dependent variable corporate performance measured by the number of ROE or measured by the number of Tobin’s Q.

The regression coefficients are represented by "0. "1 represents the independent variable of ownership structure. "1*"2 represents the interaction effect between ownership structure and the degree of internationalization of the firm (moderator 1), and "1*"3 represents the interaction effect between ownership structure and the level of board diversity (moderator 2). Moreover, $ stands for the difference between the estimated #% and the actual #% (Field, 2009). Table 2 shows a summary of the combinations of variables used in regression analyses. First, we conduct a simple regression to test the relation between control variables and dependent variables in Model 1. Then, in Model 2 and Model 3, we add independent variables respectively to test proposed H1 and H2. Later, we conduct a multiple regression to test the remaining H3-6. In the last model, we put all variables into the test to show the relation among them. This regression consists of eight regression models which are build in a hierarchical way.

The first model is composed of the control variables number of employees to measure the size of firm and the seven dummy variables that determine the industry. In Models 2, 4 and 6, we add state ownership to test the effect on the dependent variables without moderating effect and with moderating effect of number of foreign subsidiaries and the education level of board of directors respectively. In Models 3, 5, and 7, we add family ownership to test the effect on the dependent variables by using the same method as Models 2, 4, and 6. The final model

includes all variables. The regression tests are carried out for both ROE and Tobin’s Q since they represent different ways of firm performance.

Table 2. Summary of Regression Analyses

Company Size Industry State Ownership Family Ownership Foreign Subsidiaries Education Level

Model 1 X X Model 2 X X X Model 3 X X X Model 4 X X X X Model 5 X X X X Model 6 X X X X Model 7 X X X X Model 8 X X X X X X Model

5.Analyses and Results

We dedicate the following chapter to analyze the results from the statistical variables, which will be reported separately for both variables. In the first part, we will present and analyze descriptive statistics. Following this, we will conduct the statistical analyses needed for the hypotheses in the second part.

5.1 Descriptive Statistics

By descriptive analysis, we list a descriptive statistics of dependent variables, independent variables, moderating variables and control variables in table 3. There are some general characteristics of the firms listed (Tab.3) in order to provide a clear overview of the sample. The mean and standard deviation for each variable can also be found in the table.

The samples in this study consists of 400 public firms from four stock exchanges located in Shenzhen, Shanghai, Taiwan and Hong Kong. We collected the data of a period of year 2014. The valid number of sample is 390 instead of 400 since there are 10 missing values for the number of employees of 10 firms. The number of employees varied from 12 to 1110000, therefore the analysis uses the natural logarithm of the number of employees. In table 3, the mean of company size is 9.43 and the standard deviation of company size is 1.51.

The industry, as a control variable, is set into 7 dummy variables. Number 1-7 represent an industry respectively as mentioned in the previous section.

State ownership and family ownership are two types of ownerships that are mainly discussed in this study. In the analysis, these two kinds of ownerships are set as dummy variables. From table 3, the highest percentage of shares that held by state is 92% and by family is 94%. The average percentage of state ownership is 21.65%, while the average percentage of family ownership is 14.07% which is much lower than the average amount for state ownership. The number of foreign subsidiaries and education level of the board of directors are taken

as moderating variables. The minimum number for having foreign subsidiaries is 0 and the maximum number is 52, so the average number of foreign subsidiaries is 4.0 and the standard deviation is 6.64. In the analysis, level of education of board of directors are set into 5 dummy variables. Number 1-5 each represents a level of education. Number 1 represents the education level of Secondary School and less, which takes 0% of total; number 2 represents the education level of Applied College, which takes 13.5% of total; number 3 represents the education level of Bachelor, which takes 18.25; number 4 represents the education level of Master, which takes 65.75% of total; number 5 represents the education level of Doctor, which takes 2.5% of total. The average education level is 3.57, and the standard deviation is 0.75.

ROE and Tobin’s Q are two aspects that reflect the dependent variable firm performance. The minimum value of ROE is -45% and the maximum value is 92%. The average ROE is 13.82% and the standard deviation is 11.27%. A score for ROE that is larger than 15% is considered to be well-performed, a score that is larger than 20% is considered to be excellent-performed.

The value of Tobin’s Q varied from 0.02 to 68.3, having an average value of 3.49 and standard deviation is 6.61. When the value of Tobin’s Q is larger than 1, it means the value created by the firm is larger than the cost of creating, which means the firm contributes the value to society.

N Minimum Maximum Mean Std. Deviation Company Size 390 2.48 13.92 9.4300 1.50737 Industry 400 1 7 3.62 1.53018 State Ownership 400 0 0.92 0.2165 0.27640 Family Ownership 400 0 0.94 0.1407 0.20452 Foreign Subsidiaries 400 0 52 4.035 6.64184 Education Background 400 2 5 3.5725 0.75244 ROE 400 -0.45 0.92 0.1382 0.11273 Tobin's Q 399 0.02 68.3 3.4886 6.60974 Valid N (listwise) 390

Correlations Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Firm Size 35824.76 84192.67 1 Agriculture .0050 .07053 -.029 1 Energy .1970 .39823 .106* -.035 1 Manufacture .4788 .50017 -.079 -.068 -.475** ₁

Whole Sales and Retail

Trade .0723 .25934 .141 ** _{-.020 -.138}** _-.268** ₁ IT .0424 .20174 .088 -.015 -.104* _-.202** _{-.059 1} Finance .1397 .34706 -.106* -.029 -.200** -.386** -.112* -.085 1 Services .0623 .24208 -.074 -.018 -.128* _-.247** _{-.072 -.054 -.104}* ₁ State Ownership .2625 .44054 .186** _{-.042 .346}** _-.118* _{-.079 .015 -.142}** _{-.037 1} Family Owndership .0875 .28292 -.041 -.022 -.109* .057 -.087 -.021 .054 .103* -.185** 1 Foreign Subsidiaries 4.0350 6.64184 .064 -.038 .088 -.066 .026 -.009 -.004 -.012 -.001 .017 1 Applied College .1347 .34179 -.014 -.028 -.104* .017 -.026 -.010 .115* .019 -.136** .059 .178** 1 Bachelor .1820 .38636 -.028 .058 .043 .026 -.007 .093 -.078 -.095 -.105* _{-.032 -.047 -.186}** ₁ Master .6559 .47568 .038 -.023 .055 -.062 .040 -.056 .004 .057 .179** _{-.019 -.070 -.545}** _-.651** ₁ Doctor .0249 .15613 -.013 -.011 -.039 .103* -.045 -.034 -.064 .025 .014 .007 -.059 -.063 -.075 -.221** 1 ROE .1382 .11273 -.086 -.021 -.062 .037 -.031 -.033 -.083 .212** _-.188** _{.048 -.073 .038 .117}* _-.121* _{-.003 1} Tobin's Q 3.4886 6.60974 -.093 .005 -.020 .076 -.072 .039 -.143** .125* .014 .024 -.080 -.135** -.005 .096 .016 .154** 1 *. Correlation is significant at the 0.05 level (2-tailed).

**. Correlation is significant at the 0.01 level (2-tailed). Table 4. Correlation Analysis

5.2 Correlations

In this section, we conduct a correlation analysis in order to assess the correlation among the different variables and report the results in table 4.

We can derive from table 4 whether there are problems of multicollinearity among the different variables this study contains. There is no multicollinearity among all the variables according to the table since only the absolute value of a correlation score between variables is higher than 0.8 that implies a multicollinearity, and in this table the highest correlation score is -0.651.

As can be seen from table 4, the first control variable firm size has negative correlations with both ROE (-0.086) and Tobin’s Q (-0.093), which indicates that when firm size gets bigger, the performance of the firm will decline on two aspects of ROE and Tobin’s Q. The second control variable is industry which, in this study, is divided into 7 sectors, including agriculture, energy, manufacture, whole sales and retail trade, IT, finance and services. Among these 7 different sectors, energy, whole sales and retail trade and finance industry have negative correlations with both ROE (-0.062, -0.031, -.083) and Tobin’s Q (-0.020, -0.072, -0.143**), and the correlation between finance industry and Tobin’s Q (-0.143**) is significant at 0.01 level. Agriculture and IT industry both have negative correlations with ROE (-0.021, -0.033), while they have positive correlations with Tobin’s Q (0.005, 0.039). Manufacture and services industry have positive correlations with both ROE (0.037, 0.212**) and Tobin’s Q (0.076, 0.125*). The correlations that services industry has with ROE (0.212**) and Tobin’s Q (0.125*) are significant at the 0.01 level and at the 0.05 level, respectively. Hence, it’s necessary to set the control variables since they have effects on firm performance as well.

State ownership and family ownership are two independent variables that may have effect on dependent variable. Hence, a further check for multicollinearity is conducted. In table 5, the results of the VIF tests are listed to check if state ownership and family ownership have

multicollinearity problems. From the table we can see that there are no multicollinearity problems among these two independent variables since the VIF scores are all 1 and only when a VIF score is above 3 a multicollinearity problem can be indicated.

Back to the correlation matrix, state ownership has a negative correlation with ROE (-0.188**) which is significant at the 0.01 level and it has a positive correlation with Tobin’s Q (0.014). Family ownership has positive correlations with both ROE (0.048) and Tobin’s Q (0.024). Number of foreign subsidiaries and education background are two moderating variables. During the analysis, education background is divided into 4 sectors, which are Applied College, Bachelor, Master and Doctor. As can be observed from the correlation matrix, number of foreign subsidiaries has negative correlations with both ROE (-0.073) and Tobin’s Q (-.080). Both Applied College and Bachelor have positive correlations with ROE (0.038, 0.117*) and negative correlations with Tobin’s Q (-0.135**, -0.005). Master and Doctor both have negative correlations with ROE (-0.121*, -0.003) and have positive correlations with Tobin’s Q (0.096, 0.016). The correlations that Bachelor and Master have with ROE (0.117*, -0.121*) are significant at 0.05 level, and the correlation between Applied College and Tobin’s Q (-0.135**) is significant at 0.01 level. Therefore, moderating variables have different effects on dependent variable. However, in this study, we conduct the moderating test in order to figure out how moderators can influence the relation between independent variable and dependent variable, hence the further tests will be conducted in the following study.

Tolerance VIF State Ownership Family Ownership 1.000 1.000 Family Ownership State Ownership 1.000 1.000 Collinearity Statistics Model Dependent Variable

In the end of the matrix, it can be seen that ROE has a positive correlation with Tobin’s Q (0.154**) which is significant at 0.01 level. It means when ROE goes up, Tobin’s Q will go up accordingly, which is logical because both of them are used to indicate the firm performance.

5.3 Regression Analysis

A hierarchical regression analysis was used to demonstrate the effect of ownership initiatives on firm performance, when controlling the variables as discussed above.

5.3.1 ROE regression

In this part, we report the results of the hierarchical regression of firm size, industry, state ownership and family ownership on ROE in table 6. This analysis is aimed to check whether hypotheses 1 and 2 hold. Hypothesis 1 expects a negative relation between state ownership and firm performance and hypothesis 2 expects a positive relation between family ownership and firm performance.

As can be seen in the table 6, the hierarchical regression model is significant in each step of the analysis due to the F-Statistic (3.223*, 4.223**, 3.749**). R-square, which stands for the explained variation, is almost on the same level with other studies (0.056, 0.081, 0.082). This means that a big part of firm performance measured in ROE can be explained by ownership structure. When we add state ownership after the first step, R-Square increases from 0.056 to 0.081, which can also be explained by the significant F-Change in step 2 (10.651**). When we add family ownership, R-Square doesn’t have a significant change, which can also be explained by the non-significant F-Change (0.041). As control variables, the services industry (0.058*, 0.059*, 0.058*) is significant in each step of the model, while the firm size (-0.01**) is only significant in step 1.

that the negative correlation that state ownership has with ROE (-0.044**, -0.043**) is not high, however the influence is strongly significant. Thus, state ownership is significant negatively related to ROE in both step 2 and step 3. However, from the hierarchical regression in table we can observe that although family ownership doesn’t have a significant relation with ROE (0.004), the correlation it has with ROE is positive. Hence it’s hard to say whether family ownership is significantly related to ROE.

Step 1 Step 2 Step 3

Firm Size -0.01** -0.007 -0.007

Agriculture -0.062 -0.062 -0.062

Energy -0.016 -0.001 -0.001

Whole Sales and Retail

Trade -0.012 -0.017 -0.016 IT -0.023 -0.019 -0.019 Finance -0.028 -0.031 -0.031 Services 0.058* 0.059* 0.058* State Ownership -0.044** -0.043** Family Ownership 0.004 (Constant) 0.239** 0.214** 0.214** R-Square 0.056 0.081 0.082 F-Statistic 3.223* 4.223** 3.749** F-Change 3.223* 10.651** 0.041

Dependent Variable: ROE

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Table 6. ROE Regression

5.3.2 Tobin’s Q regression

In this part, we report the results of the hierarchical regression of firm size, industry, state ownership and family ownership on Tobin’s Q in table 7. This regression model is similar to the previous regression model, only that the dependent variable has changed into Tobin’s Q as a measurement for firm performance. These two regression analyses can be compared to find out if different measures for firm performance can lead to different results.

In table 7, the hierarchical regression model is significant in each step of the analysis due to the F-Statistic (4.255**, 3.961**, 3.522**). R-Square, which stands for the explained variation, is almost on the same level with other studies (0.073, 0.077, 0.077). This means that a big part of firm performance measured in Tobin’s Q can be explained by ownership structure. When we add state ownership after the first step, R-Square increases (0.073 vs 0.077). However, F-Change is not significant in step 2 (1.836). The same condition goes with F-Change in step 3 (0.09) after adding family ownership. As control variables, the finance industry (3.097*, -3.007*, -3.008*) and firm size (-0.79**, -0.883**, -0.886**) are significant at different levels in each step of the model and the correlation they have with Tobin’s Q are all negative. State ownership (1.139, 1.18) and family ownership (0.36) don’t show significant relations with Tobin’s Q.

The results showed in these hierarchical regressions imply that no support can be found for hypothesis 2, which means there is no evidence to prove that family ownership can influence firm performance from both aspects of ROE and Tobin’s Q. There is a difference in the effects that state ownership and family ownership have on corporate performance. State ownership has a significant effect on firm performance only when it’s measured as ROE, hence hypothesis 1 is supported from ROE aspect.

Step 1 Step 2 Step 3

Firm Size -0.79** -0.883** -0.886**

Agriculture -1.915 -1.899 -1.857

Energy -0.636 -1.027 -1.013

Whole Sales and Retail

Trade -1.81 -1.688 -1.646 IT 0.586 0.477 0.49 Finance -3.097* -3.007* -3.008* Services 2.615 2.599 2.559 State Ownership 1.139 1.18 Family Ownership 0.36 (Constant) 11.503** 12.147** 12.12** R-Square 0.073 0.077 0.077 F-Statistic 4.255** 3.961** 3.522** F-Change 4.255* 1.836 0.09

Dependent Variable: Tobin's Q

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Table 7. Tobin’s Q Regression