Tài liệu The effect of corporate governance on firm performance in vietnam

Abstract The study focuses on the characteristics of Board of Directors (BOD) and their influences on the enterprise value in the data set, which consists of 100 non-financial enterprises listed on HOSE and HNX with 800 observations from 2008 to 2015. Furthermore, I use system GMM for generalized estimation which allows controlling endogenous problems. The two theories: agency theory and resource dependence theory help explain the research results. The characteristics of BOD include the board diversity, board size, non-executive directors, CEO duality, ownership structures and the qualifications of BOD. The firm performance is measured by Tobin’s Q. The results show that the gender diversification, CEO duality, ownership structure and qualifications have effects on firm performance while other variables do not. This study provides information about the characteristics of corporate governance by showing the empirical evidence for Vietnam market - one of the emerging markets in Asia. Key words: Board of directors, coporate governance, firm performance, endogeneity, GMM, panel data, Vietnam. 1. Introduction In current business environment, competitions have become increasingly competitive, especially companies in the same industry. Following movement trend of the economy and development process, many enterprises are now transforming into Joint Stock Company and people who invest capital in these enterprises (shareholder) establish a Board of Directors (BOD) to represent them. It is a group of many individuals who have excellent qualifications to implement corporate management functions. Adam and Ferreira (2009) suggested that the two most important functions of a BOD are consultancy and monitoring. Consulting function is performed by executive and non-executive board members, especially the role of non-executive ones for their business relationships with important outside resources. Moreover, monitoring function will help to reduce opportunity cost; it depends on the characteristics and performance of BOD. Both functions play a vital roles in firm performance. They ensure that managers always take action for the purpose of maximizing the benefit of shareholders. Therefore, we can see the importance of BOD. In addition, it holds many expectations about maximization of shareholder benefit because not just does it prevent poor management causing violation in business but also takes advantage of opportunities to increase enterprise value. According to Vietnam Enterprise Law (2005), BOD limits violations and influences the monitoring mechanism for managers. Moreover, it also affects the appointment, dismissal, suspension and compensation policies. Together with such roles, it is necessary to carry out research “The effect of corporate governance on firm performance in Vietnam”. In fact, although there are many researches on the characteristics of board of directors that affect enterprise value, they have different results. It may be due to different legal, social and economic conditions in countries as well as different stock markets in the developed countries and emerging markets. Stock market in Vietnam was established in 2000 and became increasingly exciting in 2007 when Vietnam integrated in WTO (World Trade Organization). Thus, the stock market in Vietnam is still young. In fact, many companies listed on Vietnam’s stock market do not understand well the corporate governance due to lack of theoretical foundation of corporate governance and the theory of role and characteristics of BOD. Therefore, they have problems about information, resulting in unreality reflection of enterprise value. From these problems, this research is very necessary. Non-financial companies mentioned in this research are obtained on Ho Chi Minh City Stock Exchange (HOSE) and Hanoi Stock Exchange (HNX) in the period of eight years from 2008. In this research, I exclude financial and banking firms because (i) their liquidity and management may be affected by different legal factors, (ii) these firms have particular business environment, which are shown in the significant difference in financial ratios and legal requirements. Thus, the research results will be misleading if research sample includes these enterprises (Bauer et al, 2008; Dittmar and Mahrt-Smith, 2007; Mak and Kusnadi, 2005; Schultz et al, 2010). In addition, 100 non-financial enterprises are selected on the ground of available financial statements, annual reports and the prospectus from 2008 to 2013. Moreover, industry classification in the research replies on GISC (Global Industry Classification Standard) developed by MorganStanley Capital and S&P. Therefore, the data includes 800 observations of enterprises by year in which relatively sufficient information about corporate governance are used. Regarding contribution, this research is implemented in Vietnam. Furthermore, the length of this is eight years from 2008, which is large enough to explain the influences of characteristics of the BOD on enterprise value. In addition to this, the results are very useful for policymakers, listed companies and investors and encourage researchers to concern about corporate governance in the stock market in Vietnam. The research includes five chapters. Chapter 1 is an overview of research which indicates its contribution. Chapter 2 is the theoretical foundation and overview of previous studies. Next are the data, research methods, models and variables measurement. Chapter 4 is the empirical results and the last one is conclusion and limitation. 2. Theories and literature This section shows the theories related to Board of Directors, which mainly emphasizes the role of Board of Directors and enterprise value. The layout is presented as follows: (i) role of BOD, (ii): related theory, (iii): the effect of the coporate governance on enterprise value. 2.1. Role of Board of Directors According to the International Finance Corporation (IFC), corporate governance is defined as “the structures and processes by which companies are directed and controlled”. In 2004, following Organization for Economic Co-operation and Development (OECD), the definition was “Corporate governance involves a set of relationships between a company’s management, its Board of Directors, its shareholders and other stakeholders. Corporate governance also provides the structure through which the objectives of the company are established, and the means of achieving those objectives and monitoring performance are determined. Good corporate governance should provide proper incentives for the Board of Directors and management to pursue objectives that are in the interests of the company and its shareholders, thereby facilitating the effective use of resources.” Generally, it includes internal and external governance mechanisms, and the BOD is the former one representing the relationship between capital providers (shareholders) and capital users (director/general manager) to create and increase enterprise value (Fama and Jensen, 1983). Under Enterprise Law in Vietnam (2005), “Board of Directors is a corporate governance which has full authority to decide, implement the rights and obligations not under the control of the Annual General Meeting”. The Board of Directors works effectively when it completes its role successfully. According to Pfeffer and Salancik (1978), BOD has three main roles: (1) the mechanism for determining approaches to important resources for enterprises; (2) the marginal factor obstructing the work and (3) increasing legitimacy for the organization. Regarding Hillman and Dalziel (2003), BOD has two roles: (1) controlling business and (2) providing the methods approaching resources. According to Babic et al (2011), the role of Board of Directors includes: (1) controlling; (2) supporting and (3) strategizing. In this research, the role of BOD will be presented by the approach method of Babic et al (2011) by which the role of Board of Directors will be associated with the related theories such as agency theory and resource dependence theory. This connection is considered as the contribution to explanation the effect of the coporate governance on enterprise value. Controling role According to Fama and Jensen (1983), the BOD is an internal governance mechanism that controls the managers through processes of decision making, approval and monitoring. The process consists of four steps: (1) initiation; (2) confirmation; (3) implementation and (4) supervision; the first two steps are to make decision and the others are to monitor decision making. Therefore, controlling is the fundamental role of BOD. Regarding Zahra and Pearce (1989), controlling fulfills crucial function of BOD and also relates to the responsibilities of BOD’s members for appointing someone to be the director manager; monitoring; explaining the audit report and ratifying the decision. Furthermore, controlling is also illustrated through the other critical activities such as recruitment, appointment and dismissal of managers; confirmation objectives or monitoring the effectiveness of management level in order to ensure and increase the benefits to shareholders (Babic et al, 2011). According Baysinger and Hoskisson (1990), controlling includes two functions: (1) financial control and (2) strategic control. Financial control involves in short-term profit maximization and replies on evaluation of managers’ performance, whereas strategic control addresses a company’s long-term existence and profitability. However, BOD not only controls activities but also consults and gives a suitable orientation for managers, which is called supporting role. Supporting role According to Pfeffer and Salacik (1978), BOD’s members provide advices and expertise, resources approaches, and legitimacy. Simultaneously, they also contribute to making strategic decision by providing accessibility to key resources related to business activities. They select external influences, thereby offering reasonable approaches related to business activities; realize relationships between BOD’s members and relevant factors to ensure pivotal company resources. Additionally, they enhance the company’s reputation, and advise the management during the strategic decision-making process (Babic et al, 2011). Moreover, they can provide the managers with advice during strategic decision-making process and help them achieve the business goals through their knowledge (Carpenter and Westphal, 2001). Stratergic role Chandler (1962) defined strategy as the determination of the basic long-term goals and objectives of an enterprise, and the adoption of courses of action with the allocation of resources necessary for implementing these goals. Therefore, companies are required to identify their strengths, and then make the strategic decisions to increase competitiveness. The strategic decisions will determine the direction of future company’s objectives with many aspects such as finance, time, product manufacturing. These aspects will determine the success or failure of the companies, especially in the global integration. The BOD’s roles may also be explained via two different perspectives - an active one and a passive one. According to the passive perspective, the BOD has no significant influence on the company’s strategic development, and its responsibility is merely defined within the evaluation of the strategy adopted by top management. Otherwise, following a positive one, the BOD exerts influences on developing business strategy. Therefore, they are required to take more responsibility in that process as well as decision-making because it will affect the enterprise value (Babic et al, 2011). According to the OECD (2004), the corporate governance framework must ensure the strategic orientation. Then, the BOD must be responsible for both the strategy and contribution to the enterprise. Table 1: general role of Board of Directors - - Controling role Maximize of owners’ wealth Monitoring managers’ activities to ensure that they are acting for the interest of owners; Selecting, electing dismissing Board of Directors; Approving major decisions; Reducing agencies’ costs. - Supporting role Representing a selection mechanism of affecting external resource; Giving advice and supporting the management; Company representation; Ensuring critical resources; Guiding managers to accomplish company’s objectives; Participating in strategy formulation and implementation. Strategic role Negative perspective - No significant influence on company’s strategic development; - Implementing the strategy of Board of Directors – the Board of Directors only plays role of review and ratification Positive perspective - Board of Directors is an important strategic mechanism in governing the company; - Strategic decision-making: the Board of Directors and top management are strategic partners in the strategic decision-making process. Source: Babic et al (2011) 2.2. Theories This section presents an overview of the agency theory and resource dependence theory. These theories are considered to be related to the role of the BOD. Moreover, these theories can be used as a foundation to explain the research findings on the effect of governance structures on performance for companies. 2.2.1. Agency theory One of the first studies on agency theory was proposed by Berle and Means (1932). Agency theory developed by Alchian and Demsetz in 1972; Jensen and Meckling further developed it in 1976. According to this theory, shareholders are the owners or the head of the business. They hire and authorize to the administrators who are the shareholder representatives. Shareholders expect representatives will act in their interests. However, the representative may decide to bring their own interests, not necessarily for the best interests of the shareholders (Eisenhardt, 1989). Therefore, the employer must supervise representatives and pay monitoring costs. At the same time, the representatives will accept to bear the expenses of constrain. Total of monitoring costs, expenses of constrain and loss expenses is called the cost of representation. In addition, a form of the cost of representation derives from conflicts of interest between shareholders and corporate executives due to the separation between management and ownership (Davis, Schoorman and Donaldson, 1997). This cost is called agency cost. When shareholders are restricted or lost control to the administrator, then the administrator will have motivation to participate in the beneficial activities for them. Moreover, another form is demonstrated by debt costs, arising from the conflicts of interests between shareholders and creditors. Shareholders may possess the property of creditors by increasing investment risk (Jensen and Meckling, 1976) in case enterprises have loan. In addition, when companies are in financial distress, conflicts of interest between shareholders and bondholders could make shareholders acting for their own interests to profit. Losses are incurred by creditors by game theory to reduce the total value of enterprise. There are many mechanisms to control representative costs. Jensen and Meckling (1976) recommended the use of financial leverage, design of incentive contracts as share ownership and stock options to focus the efforts of the administrator on the interests of shareholders. Rozeff (1982) and Easterbrook (1984) made recommendations that the dividend can be considered as the tool to reduce representative expense; because with dividend, on the one hand, it provides the encouragement, on the other hand, it avoids excessive investment and creates external oversight mechanisms, thereby limiting the abuse of the free cash flow of administrators. 2.2.2. Resource dependence theory Resource dependence theory is started from the studies in economics and sociology, while focusing on the association role of member of BOD. The BOD’s members can both operate in host enterprise, and have positions of management in other ones. These members will act as a bridge between enterprise and competitors and other stakeholders (Zahra and Pearce, 1989). According to this theory, the BOD is considered as a voting mechanism in the enterprise to shape the linkages with the environment outside the enterprise, while receiving the influential and important resources to make stepping stones for enterprises to handle the adverse changes in the business environment (Pfeffer, 1972; Pfeffer and Salancik, 1978; Pearce and Zahra, 1991). These resources are general, unique and are essential for the growth of the enterprise. Specifically, the resources include all sources of asset formation, operation capital, business processes, business type and the ability to receive and respond to information as well as the application of the human knowledge of the enterprise, thereby increasing the enterprise value (Daft, 2008). The keynote of this theory states that the enterprises make efforts to control business environment by selecting the necessary resources for survival (Pfeffer and Salancik, 1978). Accordingly, the BOD plays a role as a bridge between the enterprise and the basic resources; at the same time, enterprise will receive external resources to operate more efficiently. With the role of resource constraint, external BOD’s members will provide resources for the enterprise such as receiving and responding to information, skills and knowledge, accessibility to key components (such as suppliers, customers, policy makers, social groups) and the legitimacy of business operation (Hillman et al, 2000). Pfeffer and Salancik (1978) stated that resource dependence theory replies on the idea that the environment provides scarce resources. Thus, the organization depends on that limited resources to survive. To ensure the survival and development of their own, they have to develop methods for using of resources effectively. In addition, they also said that the role of the Board of Directors is to link the enterprise with other external organizations to address the dependence on the environment. Therefore, as a mechanism for internal management of the enterprise, in order to improve and enhance enterprise value, the BOD clearly defines and ensures the effectiveness of its role. This requires that the Board should have specific characteristics. The next section presents the characteristics of the Board of Directors that affects the value of the enterprise. 2.3. The corporate governmances affect the firm performance 2.3.1. Board diversity Board diversity is the proportion of female members in total members of BOD. Women contribute to the diversity of the Board of Directors. In agency theory, the board diversity is more independent (Jensen and Meckling, 1976). In resource dependence theory, the board diversity helps to ensure important resources of the secrets of the enterprise and creates the condition for connection between enterprise and the external environment including the credibility and legitimacy (Pfeffer et al, 1972). However, the empirical reseaches have different results. Particularly, Erhardt et al (2003) did research on 127 large enterprises in the United States during the period of 1993-1998 to consider the diversity of the Board of Directors (as measured by the proportion of women in the Board) and financial performance (as measured by ROA). The results presented a positive relationship between the diversity of the Board of Directors and enterprise value or Dezso and Ross (2012) used a large observation sample include 21,790 observations of enterprises publishing financial information on CompuStat of S&P in the period of 1992- 2006. The study results demonstrated that the presence of women in the Board of Directors innovates the business strategies and thereby, having positive influence on the enterprise value. By constract, Bohren & Strom (2010) did a research in non-financial Norwegian listed compamies in the period from 1989 – 2002 and this result illustrated the negative relationship between gender diversity of BOD and the value of firms measured in ROA and Tobin’s Q. By similarity, Ahern & Dittmar (2012) showed the proportion of female in BOD have negatively effect on firm performance measured in Tobin’s Q. Furthermore, Tuan Nguyen et al (2015) did a research on 257 Singaporean non-financial listed companies from 2008 – 2011 and showed the gender diversity has a negative effect on firm performance. More interestingly, Rose (2007) studied on the Copenhagen Stock Exchange in Denmark from 1998 to 2001, the result presented there is no relationship between gender diversity and firm performance (as measured in Tobin’s Q). Moreover, Smith et al (2008) used a sample of 10000 observations in the period from 1994 – 2003 and illustrated the percentage of female in BOD does not have effect on listed enterprises. With positive influence of the board diversity on enterprise value which is predicted by agency theory and resource dependence theory, I propose the following hypotheses: Hypothesis 1: Board diversity positively affects the value of the listed enterprise in Vietnam 2.3.2. The size of board of directors Board size is the member of directors in BOD, including executive directors and non-executive directors. Dalton et al (1999) argued that the size of the BOD is one of the most important characteristics of its functions. However, there is uneven between agency theory and the resource dependence theory of the affect of the size of the BOD on enterprise value. Agency theory predicts a negative effect (Jensen, 1993). This may be due to large-scale BOD that has advantages in improving the functions of management board such as the consulting support, autocratic reduction of the administrators. However, the increase of the size of the BOD to a certain level will be ineffective due to increasing dependence and subjectivity within the BOD. Additionally, he recommended that the BOD size should not exceed seven people for all kinds of enterprises as this is the optimal level. While resource dependence theory suggests a positive effect (Dalton et al., 1999), the greater size of BOD will positively affect the enterprise value and administrators will receive better advices, better orientation with large-scale BOD. Regarding to resource dependence theory, the big size of BOD supports and increases in the value of enterprises in the limited resources (Pfeffer & Salacncik, 1978) Results of practical researches, Beiner et al (2006) did a research on 275 firms quoted at Swiss stock market (SWX) in 2002 and demonstrated positive relationship between the sizes of the BOD with Tobin’s Q whereas Yermack (1996) used Tobin’s Q index to measure enterprise value, with the number of observations was 452 enterprises of major industries in the U.S. over the period of 1984-1991. The results portrayed that board size negatively affect Tobin’s Q index. Similarity, Tuan Nguyen et al (2015) presented the negative relationship betwwen board size and firm performance measured in Tobin’s Q. However, research by Beiner et al (2004) with the data of 165 companies listed on the Swiss stock market demonstrated no positive relationship between the sizes of the BOD with Tobin’s Q or Schultz et al. (2010), Wintoki et al. (2012) concluded that this influence was not statistically significant after controlling for endogenous problems. Due to the conflict between agency theory and resource dependence theory of the effect of the size of BOD on enterprise value, I propose the second hypothesis is as follows: Hypothesis 2: Board size affects the value of the listed enterprises in Vietnam 2.3.3. The non-executive of board of directors According to Clifford and Evansm (1997), non-executive directors are the ones that not implement the execution. Non-executive directors are not the Director (General Director), Deputy Director (Deputy General Director), Chief accountant and other administrative officers appointed to the Board (Item 2, Article 2 of Circular 121/2012 - Vietnam). Agency theory suggests a higher proportion of non-executive directors will bring higher better supervision (Fama and Jensen, 1983; Jensen and Meckling, 1976). It is assumed that the nonexecutive directors can perform their supervision functions better than the executives when they are less dependent on business management (Fama and Jensen, 1983). Nicholson and Kiel (2007) argued that if the supervision function of the BOD is effective, the interests of shareholders will be then increased because of the minimized cost of representation. This conception is considered appropriately for the conception of resource dependence theory. In emperical researches, Hermalin and Weisbach (1998), when they conducted a study of 142 companies listed on the U.S Stock Exchanges during the period of 1971-1983, it portrayed that the number of non-executive directors tended to increase when the enterprises are in a difficult situation, because they expect these members will improve enterprise value over that period. Besides, Bhagat and Bolton, 2008 used the sample on CompuStat of S&P, IRRC and TCL, GIM, BCF indices in the period from 1990 to 2004 and indicated the negative relationship between the non-executive directors of BOD and firm performance measure in ROA. By constract, Tuan Nguyen et al (2015) presented there is no relationship between non-executive directors in BOD and firm performance. With the positive effect of the ratio of non-executive directors on the enterprise value predicted by agency theory and resource dependence theory, I propose the following hypothesis: Hypothesis 3: The proportion of non-executive directors has positive influence on the value of the listed enterprises in Vietnam. 2.3.4. CEO duality CEO duality happens when Chairman of Board also takes the position of management in an enterprise (director/general director). On the ground of agency theory, CEO duality cause the difficulties for Board (Donaldson and Davis, 1991) due to supervision and administration functions tends to be abused for personal benefits and representation costs tend to rise higher. The focus over the management and control of an individual’s decision will negatively affect enterprise value and reduce the role of the Board in the supervision to the managers (Fama and Jensen, 1983). Practical research on the influence between CEO duality and enterprise value has conflicting results. Gill and Mathur (2011) conducted the study of 91 listing listed manufacturing companies on the Toronto Stock Exchange of Canada during period of 2008-2010 and discovered the positive relationship between the CEO duality and enterprise value of manufacturing industry while Daily and Dalton (1994) studied the stage of enterprises at their year 5 th and year 3rd before they went bankrupt in 1990, they discovered that the enterprise that not separate CEO duality have high possibility of bankruptcy or Haniffa, R & Hudaib, M., (2006) did research 347 Malaysian listed enterprises from 1996 and 2000 and showed the negative relationship between CEO duality and firm performance. Interestingly, Chen et al (2008) used the sample in Compustat method from 1993 – 2003 and showed there is no effect between CEO duality and firm value or Tuan Nguyen et al presented the CEO duality does not affect the firm performance. On the ground of agency theory and documents on the influence of the CEO duality on the enterprise value, I propose the fourth hypothesis as follows: Hypothesis 4: CEO duality has the negative influence on the value of the listed enterprises in Vietnam. 2.3.5. The ownership of the board of directors Agency theory illustrates that the percentage of ownership of the BOD’s members is one of the important mechanisms to monitor the behavior of management. One of the authors that have quite early research on this relationship is Berle and Means (1932). The authors stated that the percentage of equity ownership really affects corporate profitability. Therefore, it is needful to separate ownership and control right when the company is transformed into the joint stock company. Such separation between ownership and control will increase the professionalism and expertise in enterprise through expert management skills and professional of business. However, because of the separation between ownership and control rights generating the costs of shareholder representative and administrators, shareholders should establish the BOD, which performs monitoring functions to reduce agency costs (Fama and Jensen, 1983). Moreover, in order to operate effectively, it is necessary to stick to the benefits of shareholders and the BOD through capital ownership ratio (Jensen and Meckling, 1976). The percentage of ownership of the BOD help shareholders mitigate problems arising from the separation between ownership and control rights; thereby increasing enterprise value. However, the practical evidences on the effects of equity ownership percentage of board members on the enterprise value were unclear and inconsistent between different markets. Tuan Nguyen et al (2015) illustrated the possitive relationship between ownership of BOD’s members and Singaporean firm performance listed companies. By constract, Farooque et al (2007) studied on the companies listed in the Dhaka stock exchanges (Bangladesh) in the period of 1995-2001 with 660 observations. The authors used OLS method and discovered the opposite effect of the percentage of ownership of the Board on the enterprise value (measured by Tobin’s Q) or Hu et al (2010) did research on 304 publicly listed companies from 2003 to 2005 in China and presented negative effect between owership of BOD’s members and firm performance. On the ground of agency theory of the influence of equity ownership of BOD’s members on the value of enterprise, I propose the fifth hypothesis as follows: Hypothesis 5: Ownership structures have the positive effect on the value of the listed enterprises in Vietnam. 2.3.6. Qualification of board of directors Many researchers consider that the qualifications of the members of the BOD are one of the very important characteristics because it creates value of the execution function of the board (Carpenter and Westphal, 2001). The quality of each BOD’s member has a very important role when making decisions. According to resource dependence theory, the BOD’s members who have the skills and knowledge will be considered as important strategic resources in the connection with external resources (Ingley & Van der Walt, 2001). Fairchild and Li (2005) investigated 121 companies which disclosed information on data system of merger and acquisition of Thomson in the period of 1990 - 1993. These authors concluded that the quality of the BOD’s members has positive effect on business share price. Bathula (2008) on 61 companies listed on the New Zealand stock exchange during period of 2004-2007. The author discovered the negative effect of the BOD’s members with PhDs on the enterprise value because members who hold doctorate and have skills of research and analysis do not bring the enterprises any value. On the ground of resource dependence theory of the effect of the qualifications of BOD’s members on the enterprise value, I propose the sixth hypothesis as follows: Hypothesis 6: Qualifications of Board members have the positive effect on the value of the listed enterprises in Vietnam. Table 2: Expected expectation between variables Agency theory Board diversity Board size Non-executive directors CEO duality Ownership structures Qualifications BOD’s members + + Resource dependence theory + + + + of Expectatio n + +/+ + + + In summary, this section discusses three important roles of the BOD including (1) the controlling role, (2) the supporting role and (3) the strategic role. These roles have linkages with the development of the business management theory such as agency theory, theory of constraint of resource. These theories help to explain the influence of the characteristics of the BOD on enterprise value. Generally, there are many different viewpoints of the influence. The next section presents research methodology to measure the effect of the characteristics the BOD on the value of the listed enterprises in Vietnam. 3. Research method 3.1. Emperical model As metioned above, the effect of governance structures on enterprise value creates the problem of endogeneity. I inherit the approaches of Carter, D’Souza, Simkins and Simpson (2010) which use lag of dependent variable and fixed effect model for enterprises to control the endogeneity. In case of lagged variable, theories could not estimate the time necessary for the effect of characteristics of BOD on enterprise value. The number of lags will depend on experience from empirical researches. Referring to previous research models of Wintoki et al. (2012) and Tuan Nguyen et al (2014), I add the qualification variable of BOD’s members into the model and proposes the model for this research as follow: ln q it α   β 5 ownit γ ln + q it−k β 1 female it + β 6 edu it + δ 1 lnfageit + + β 2 nonexe it δ 2 fsize it + + β 3 dual it δ 3 levit + + β 4 lnbsize it ηi + + ε it where:; i denotes observational firms t denotes time ηi denotes unobserved time-invariance firm effects k is lag of dependent variable. On the ground of previous experimental researches which used lag of k = 1 (Adams and Ferreira, 2009; Dezsö and Ross, 2012), I use k = 1 to do research on the effect of characteristics of Board of Directors on value of listed enterprises in Vietnam in this research. 3.2. Variables 3.2.1. Dependent variable Based on other previous studies (Coles et al 2012; Khatab et al 2011; Reddy et al, 2008), I use Tobin’s Q as a dependent variable in this research. According to Ammann et al. (2011), Tobin’s Q ratio is total market value of equities and the book value of debt divide by the book value of total assets. Logarithmic form of Tobin’s Q ratio is used to reduce the effect of outlier (one observation point that is distant from other observations abnormally). 3.2.2. Explanatory variables and control variables Independent variable in this study include: (i) board diversity; (ii) the percentage of nonexecutive directors of Board of Directors, (iii) CEO duality, (iv)Board size, and (v) the ownership of BOD’s member, (vi) Qualification of BOD’s member. According to the previous researches (Adams and Ferreira, 2009; Reddy et al, 2010, 2011; Rustam et al, 2013; Tariq and Abbas, 2013), I use the enterprise size, age of enterprise, leverage, dummy variable, lags of dependent variable as control variable. Table 3: Definitions of variables in the research Variable Abreviations Independent variable Tobin’s Q ratio lnq Dependent variable Board diversity female Non-exercituve directors CEO duality Non-exe Board size lnbsize Ownership of Board of Directors’ member Qualification of Board of Directors’ member Control variables Enterprise size Leverage Lag of dependent variable Own Dual Edu Lnfsize Lev Lglnq Definitions Tobin’s Q = total market value of equities  book value of debts book value of total asset The percentage of female members in total members of Board of Directors The percentage of non-exercutive directors in members of Board of Directors As the dummy variable, value of this variable will be 1 if CEO is chairman and will be 0 on the contrary Total member of Board of directors, its logarithm is used in model Total common share owned by Board of Directors divides by total outstanding common share and then multiply by 100% The number of Board of Directors’ members have master’s degree of higher Logarithm of total book value of enterprise Ratio of total debt to total asset 1 year lag of variable lgq Age of enterprise Lnfage Industries dummies d1 Year dummies d2 Logarithm of years since the first apperance of enterprise on HOSE/HNX Dummy variables of one of eight industries under GISC 8 year dummies for each of the eight years from 2008-2015 3.3. Econometric method 3.3.1. OLS OLS method is the BLUE because of some assumptions. In OLS method, we assume that there is no any multicollinearity between independent variables of model, which means that there is no effect between independent variables). Furthermore, it will have no deviation and minimum variance. One of these hypotheses is that variance is constant because estimations on the ground of OLS method are not effective due to difference and tests of the significance and confidence intervals are no longer meaningful in case of variance of the residuals. Moreover, Autocorrelation refers to the correlation between observations at the same data table. When autocorrelation occurs, OLS estimation shall be ineffective method because of inaccuracy of estimation regression coefficient. 3.3.2. FEM/REM Fixed-effects method and Random-effects method are used in panel data analysis. Aside from advantages of panel-data such as increasing sample size significantly, through researches on repeated cross observation, we know that it is suitable for researches on dynamics. Panel data supports the researches on complex behavior pattern and raises some issues in which data is related to both space and time; thus, troubles in cross data (heterogeneous variance), time-based data (autocorrelation) and other issues need to be solved (Ramu Ramanathan, 2002). There are some estimation techniques to handle these issues. Two of the most outstanding techniques are FEM and REM. FEM can control and separate the influence of unique characteristics which is constant over time and is unique for one enterprise and does not affect characteristics of other enterprises such as corporation business or corporation culture from explanatory variables to estimate the real effect of explanatory variable on dependent variable (Pooled OLS connecting with dummy variable). REM assumes that individual characteristics of enterprise are random and do not affect explanatory variable. It also considers the residual of enterprises that do not affect explanatory variable as a new explanatory variable (Ramu Ramanathan, 2002). 3.3.3. GMM GMM system method was proposed by Blundell and Bond (1998). In the context of research on corporate governance, the estimation of GMM system is considered as a feasible solution to control the problem of endogeneity (Antoniou et al, 2008; Nakano and Nguyen, 2012). Specifically, it allows using available tools in panel data (Blundell and Bond, 1998). Moreover, it also helps to handle panel data which has not enough time for collection, lack of external instrumental tools and has changes of dependent variable over time (Roodman, 2009b). GMM system can be briefly described as a system of two simultaneous equations with different lag (k). While lag of explanatory variable is considered as an instrumental variable in one differential equation or the lag of difference which is used as instrumental variable. According to Wintoki et al. (2012), I assume that age of enterprise (Infage) and industry dummy variable are exogenous variables. Endogeneity is a phenomenon in which independent variables have effect on residuals. It means that E (ε|X) is different from 0. This breaks the assumption of OLS. In this case, independent variable is endogenous variable. The appearance of endogenous variable will create missing of variable, error of variable, or simultaneous identification through explanatory variables. In order to solve this problem, we need instrument variables which must meet two criteria (i) instrument variables must be correlated with explanatory variabes and they must be uncorrelated with the error. When the dependent variables are exogenous, the OLS and FE/RE methods are more effective solutions than GMM method. Therefore, it is important to inspect the endogeneity of regression variable in this model. I conducted to test Durbin – Wu – Hausman (DWH) about the endogeneity of regression variable. These tests are assumed that the regression variable is exogenous one (Baum et al., 2007). The inspection is conducted on the ground of variable of enterprise value and variables regarding corporate governance (Schultz et al., 2010). The instrumental variables are included Δlnqit−1; Δfemaleit −1; Δnonexeit −1; Δdualit −1; Δlnbsizeit −1; Δownit −1 ; Δfsizeit −1 ; Δedu it and Δlevit −1 because they are strongly correlated to historical performance and lagged values of other explanatory variables as well as under weak rational expectations, if the the board structure that we observe today is one that trades off the anticipated costs and benefits of particular board structures, then current shocks to performance must have been unanticipated when the boards were chosen (Wintoki et al, 2012). 3.4. Data Data is obtained from enterprises listed on HOSE and HNX. These above data include 100 enterprises with full information about all variables used on the research that are collected from annual reports, financial statements and prospectus. Collected variables mostly are obtained from vietstock.vn. However, some information is missing. Therefore, I use the missing information from cafef.vn. In addition, foundation of industry classification in this research replies on standards of GISC such as finance, real estate, manufacture, energy, material, industry, consumer goods, medical and pharmacy, information technology, telecommunication, water and electricity. Otherwise, I eliminates some specific industries as mentioned in Section 1.5. 4. Empirical results 4.1. Descriptive statistics The table 4 shows the data of descriptive statistics for 800 observations on the explanatory variables and control variables used in this research during 8 years from 2008. The average of Tobin’s Q is 0.89. It portrays the average value created for shareholders during sampling period. Furthermore, an average percentage of female members in BOD is 13.41% which is 7 percentage point higher than that of Asia zone (about 6% according to Sussmuth Dyckerhoff et al.2012) whereas it is lower than that of other markets in developed countries such as Finland, France, United State and Australia (International Labor Organization – ILO). It demonstrates the gap between female representatives in BOD in listed companies in Vietnam and those in developed countries. Regarding percentage of non-executive directors and CEO duality, they account for roundly 47.48% and 37.13%, respectively. Besides, the mean of the number of directors on the board is 6. The mean of shares held by BOD’s member is approximately 31.77%, which is quite high. This is compatible with the conclusion of Claessens et al. (2000) regarding the percentage of ownership of board members in Asian market. When it comes to qualification, on an average, the number of people who have master degree or higher is nearly one person. Regarding control variables, the number of years from the time the company first appears on the HOSE/HNX is around 6.05. Natural logarithm of the book value of total asset is approximately 27.18. An average percentage of leverage is nearly 47.11%, which means that one VND of enterprise asset is financed by 0.4711 VND of loan. When it comes to correlation, the table 5 illustrates the correlation coefficients. Based on the results, the variables such as board diversity, CEO duality, board size and the ownership of BOD’s member have a positive effect on the Vietnam’s corporate governance while others such as percentage of non-executive directors and the qualification of BOD’s members have negative effect. More importantly, the one year lag of Tobin’s Q has positive effect on its present value (0.66). This demonstrates that the previous time has positive effect on present value of listed enterprised in Vietnam, which supports the hypothesis of Wintoki et al. (2012). 4.2. Multivariable regression analysis 4.2.1. OLS method Regarding multicollinearity test, the result from the table 7 presents that VIFS are 1.36 less than 10 (Pham Tri Cao et al., 2009). Thus, there is no any multicollinearity when using OLS method for estimation. When it comes to heteroskedasticity test, I use LM test of Breusch-Pagan/Cook-Weisberg which assumes that there is no heteroskedasticity. Results from the table 8 present the statistic data with χ 2 =70.20, p-value = 0.0000. It means that the hypotheses are rejected. Hence, I reach conclusion that there is variance of the residuals when using OLS regression model with lnq dependent variable Regarding autocorrelation test, I use Wooldridge (2002) test which assumes that there is no any autocorrelation in data researched. Based on the results in the table 9, it is concluded that there is autocorrelation in this data when using OLS regression model with lgq dependent variable, and F(1, 99) = 60.219, p-value = 0.0000. Thus, there is autocorrelation when estimating by OLS. 4.2.2. Fixed-effect method and random-effect method Purpose of Hausman test is to select whether FEM method or REM method is suitable to sample data regression on the ground of assumption that there is no any correlation between explanatory variable and random factor. The result in the table 11 shows 2 χ 10 = 126.78=, p-value = 0.0000, which illustrates that FEM is sutable for this research. In addition to this, from the table, F(15,99) = 42.21, p-value = 0.0000, which shows that FEM is acceptable/feasible. 4.2.3. GMM estimation method To identify the suitability of GMM model. Firstly, I implemented the envogenous test with the null hypothesis is that the regression variable is exogenous. With the result of the table 12, χ 2 = 1.24, p-value = 0.2651. Therefore, we reject the null hypothesis and there is endogeneity between coporate governance and enterprise value listed in Vietnam stock market. Then, I implemented Sargan test of overidentification tests the valid of instruments with the assumption that the defined number of instrumental variables are valid. With the result in this table 13, 2 χ  26 = 33.49, p-value = 0.148. Therefore, the instrumental variables used in GMM model are valid. In addition to this, I implemented the difference in Hansen tests of exogeneity of instrument subsets with the assumption that instruments used are exogenous. The result of this test is that 2 χ 6 = 6.79, p-value = 0.341. Hence, this implies that we cannot reject the hypothesis that the additional subset of instruments used in the system GMM estimates is indeed exogenous. 4.3. Emperical rerult In the table 14, regarding the diversity, when using Pooled OLS model, considering column (2) and (3), there is no statistical significance. But when control and separate the effects of individual characteristics which do not change over time from the explanatory variables, considering column (4) and (5) to estimate, the result has a positive effect with the level of 10%. However, when controlling the endogenous issues, it has a negative effect on the on the value of listed enterprises in Vietnam at the significance level of 5%. This result supports the first hypothesis. Although this result does not support the agency theory and resource dependence theory. This experimental result is supported by the previous study of Bohren & Strom (2010), Ahern & Dittmar (2012) and Tuan Nguyen et al (2015). The proportion of non-executive directors has a negative effect on the on the value of listed enterprises in Vietnam at the significance level of 10%. When using Pooled OLS model and FEM, the results have no statistical significance. However, when controlling the endogenous issues, the results reveal negative effects at the level of 10% illustrated in column (6), (7) in the table 14. This experimental result is supported by the previous study of Bhagat and Bolton (2008) and Tuan Nguyen et al (2015).