Microsoft Word - 12592-46134-1-SM-writer2-new Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 40 Effect of Corporate Social Responsibility Disclosure on Financial Performance Hang Thi Thuy Ta University of Labor and Social Affairs, Vietnam Ngoc Thi Bui (Corresponding author) University of Labor and Social Affairs, Vietnam Room No. 44, A8, 120 Hoang Quoc Viet Street, Cau Giay District, Hanoi, Vietnam E-mail: buithingoc.ldxh@gmail.com Received: Jan. 3, 2018 Accepted: Feb. 22, 2018 Published: June 1, 2018 doi:10.5296/ajfa.v10i1.12592 URL: https://doi.org/10.5296/ajfa.v10i1.12592 Abstract Corporate social responsibility is the commitment by business to contribute to sustainable development through the balance of three factors which are economic, social and environmental factor. Enterprises carry out corporate social responsibility through specific activities such as controlling environmental pollution, natural resource conservation, energy conservation, employment safety, especially supporting and developing community. Corporate social responsibility disclosure plays an important role in implementation of corporate social responsibility, promotion of business’s image and creation of good impression to stakeholders. This study was conducted using a two-step Generalized Method of Moment (GMM) technique with instrumental variables for balanced panel data through the annual reports and sustainable development reports of 43 enterprises listed on the Vietnam stock market from 2006 to 2016 (473 observations). The results showed that the level of corporate social responsibility disclosure has a positive effect on return on assets. This study has important implications for enterprises in terms of investing activities and corporate social responsibility disclosure. Keywords: Corporate social responsibility disclosure, financial effectiveness, return on assets (ROA) Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 41 1. Introduction Sustainable development is the trend of the times and businesses cannot afford to ignore this trend. The practice and disclosure of corporate social responsibility information is considered a tool for businesses to communicate the activities of sustainable development of society. Gray et al. (1995) provided that corporate social responsibility disclosure is the process of providing information about corporate activities affecting environment and society. Enterprises are accountable for such information to related parties, apart from financial information publicized to shareholders. Non-governmental organizations and nations have introduced standards and regulations to guide firms to establish and display corporate social responsibility information, including the United Nations Global Compact (UNGC), International Finance Corporation (IFC), Global Reporting Initiative (GRI), International Integrated Reporting Council (IIRC), The Sustainability Accounting Standards Board (SASB). Depending on conditions, legal regulations and needs, enterprises can disclose corporate social responsibility information in accordance with standard or legal requirements. According to the above standards, corporate social responsibility may include following key information: Environment: Environmental policies or environmental concerns of enterprises, environmental pollution control in business operations, conservation of natural resources, prevention and remedies for environmental consequences of the manufacturing process or natural resources, information about emissions, discharge of waste water, waste treatment. Energy: Information about a company's energy policies, energy conservation in the course of business operations, utilization of waste materials for energy production, research towards energy savings by reprocessing products. Employment: Information regarding regulations and standards on labor safety, working conditions, health care for workers, training, financial support for housing, information about the daily allowance for employees, maternity leaves, holidays, information regarding policies for seasonal/ contractual employees or plan of hiring seasonal/ contractual employees, provision of information regarding recruitment/ jobs for women or ethnic minorities or special interest groups. Supporting and developing community: Information about donations with money, artifacts or services of enterprises to support community activities, events, arts, sports, funding for health projects and providing information related to public health. Products: Researching and developing products to improve product quality and to meet safety standards. With increasing community interest in corporate social responsibility, many studies in the world have been conducted to verify the impact of corporate social responsibility disclosure to business efficiency such as Mahoney & Roberts (2007), Makni et al. (2009), Saleh et al. (2011), Tjia & Setiawati (2012), Vurro & Perrini (2011), Lu et al. (2015), Strouhal et al. (2015), Jitaree (2015). Impacts are expressed in a variety of indicators, such as the impact on a firm's share price, return on assets (ROA), return on equity (ROE), market value compared to book value Tobin'Q, revenues of enterprises. Although these studies had different results, the majority Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 42 concluded that corporate social responsibility disclosure has a positive impact on business efficiency. Several studies had reported negative results Makni et al (2009) and in some cases there were not statistically significant Strouhal et al. (2015), Khlif et al. (2015), Tjia& Setiawati (2012). In Vietnam, some studies have investigated the current state and benefits of corporate social responsibility disclosure. The results showed that Corporate Social Responsibility disclosure increases the value of enterprises Tobin'Q Nguyen et al. (2015), Nguyen & Trinh (2016), increases the return on assets Ho & Ho (2017). However, research time was short and these studies used panel data with the Fixed Effects Model (FEM) and the Random Effects Model (REM) for the survey. Using these estimation methods will not be reliable if endogenous phenomena exists in the model. To overcome this situation, it is necessary to use other estimation methods and extend research time. It is critical to ensure reliability of the study on the effect of corporate social responsibility disclosure on corporate financial performance. For this purpose, in this study the authors used a two-step Generalized method of moment (GMM) for balanced panel data with 473 observations of 43 firms over a period of 11 years from 2006 to 2016. 2. Theoretical foundations and hypothesis From an academic point of view, many researchers have investigated the theoretical basis to explain the relationship between practice and disclosure of corporate social responsibility and financial performance of firms in different contexts. In this study, the authors used two theories to explain the relationship between corporate social responsibility disclosure and corporate financial performance. Stakeholder theory Edward (1984) was the first to propose the stakeholder theory. He supposed that the goal of a business is to meet the needs of stakeholders who may influence or be influenced by the attainment of their goals. If this is done a firm’s profits will be generated. Based on the stakeholder theory in strategic management, Ullmann (1985) presented 8 scenarios to explain the differences in the results of the relationship between Corporate social responsibility disclosure and corporate financial performance. According to him, the relationship between corporate social responsibility disclosure and corporate financial performance depends on other factors such as the power of stakeholders and management strategies. He provided that corporate financial performance has a positive impact on social efficiency and corporate social responsibility disclosure. Legitimacy theory Legitimacy theory is derived from the concept of organizational legitimacy, Dowling & Pfeffer (1975) defined: “An entity can exist when its value system is congruent with the value system of the larger social system. When there is a real or potential disparity between two value systems, the legitimacy of that entity is threatened”. Developing legitimacy theory, Deegan (2002) supposed that enterprises and society have a continuous and interconnected relationship. For example, companies buy human resources, raw materials from society, conversely, companies provide products and services for society, Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 43 discharge waste from the production process of enterprises to environment and society has to pay for the consequences. Enterprises try to control their legitimate existence to ensure capital flow, labor and customers, while avoiding the management activities of the government harming the business and boycotting products or other disruptive actions of competitors. Stakeholder theory and legitimacy theory indicate that practice and disclosure of corporate social responsibility will lead to better financial performance. Conversely, in the case of good corporate financial performance, corporate social responsibility disclosure is also greater. Based on these theoretical foundations, the author constructed the following hypothesis: H1: There is a positive relationship between social responsibility disclosure and financial performance. 3. Methodology 3.1. Data collection Samples selected by the authors were non-financial companies listed on the stock market from 2006 to 2016. The reason why the author chose this period is because 2006 Vietnamese stock market boom with the appearance of the Hanoi Securities Trading Center in March 2005. The number of listed companies increased sharply. As of 31 December 2006, there were 193 companies listed on both exchanges (106 companies listed on Hochiminh Stock Exchange and 87 companies listed on Hanoi Stock Exchange).Financial and banking organizations were not chosen because information provided by these companies is significantly different from other types of enterprises. The last sample was 43 listed companies on the Vietnam stock market randomly selected by the elimination of financial firms, creditors, and those were not listed companies during the research period as well as companies which the authors could not collect annual reports or sustainable development reports. Thus, 43 enterprises were surveyed for 11 years with the balanced panel data with a total observations of43 x 11 = 473.Shown in table1. Table 1. List of companies surveyed Code Company's name 1 AGF An Giang Fisheries 2 BHS Bien Hoa Sugar 3 BMC Binh Dinh Minerals 4 BMP Binh Minh Plastics 5 CAN Halong Canfoco 6 CII Ho Chi Minh Infrastructure 7 CLC Cat Loi Tobacco 8 CYC Chang Yih Ceramic 9 DHA Hoa An Stones and Materials 10 DHG Hau Giang Pharmaceutical 11 DMC DOMESCO Medical 12 DTT Do Thanh Technology Corp. 13 FPT FPT Group Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 44 14 GIL Binh Thanh Im-export 15 GMD Gemadept 16 HTV Ha Tien Transport 17 IMP Imexpharm Pharmaceutical 18 ITA Tan Tao Industrial Park 19 KDC Kinh Do Corporation 20 KHA Khanh Hoi Investment and Services 21 KHP Khanh Hoa Power 22 MCP My Chau Printing and Packaging 23 PGC Petrolimex Gas 24 PJT Petrolimex Tanker 25 PNC Phuong Nam Cultural 26 PPC Pha Lai Thermal Power 27 REE Refrigeration Electrical Engineering 28 SFC Saigon Fuel Co. 29 SFI Sea and Air Freight International 30 SGH Saigon Hotel Corp. 31 SJD Can Don Hydro Power 32 SMC SMC Trading and Invm't 33 SSC Southern Seed JSC 34 TAC Tuong An Vegetable Oil 35 TDH Thu Duc House 36 TNA Thien Nam Trading and Exim 37 TS4 Seafood No 4 38 TYA Taya (Vietnam) electric wire & cable 39 VGP Vegetexco Port 40 VIP Viet Nam Petroleum Transport 41 VIS Vietnam - Italy Steel 42 VNM Vinamilk 43 VTC VTC Telecom 3.2. Determining and measuring variables 3.2.1. Corporate social responsibility disclosure variable In this study, the author measured corporate social responsibility disclosure according to content analysis method based on annual reports and sustainable development reports of companies. The analyzing process was based on the list of indicators categorized in 4 groups shown in table2: information about environment (10 indicators), information about employment (12 indicators), information about supporting and developing community (8 indicators) and information about the products (5 indicators). The indicators of information were inherited from studies of Gunsanan et al. (2009), Jitaree (2015), Nguyen (2016) and according to circular 155/2015issued on 06/10/2015 by Vietnamese Ministry of Finance on Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 45 guidelines on information disclosure on the stock market. Table 2. List of indicators of information about corporate social responsibility Symbol Indicators (observed variables) ENV I. Information about environment responsibility ENV 1 The enterprise complies with the law and regulations on prevention of environmental pollution (including number of times and amount of fine for violation of laws and regulations on environment) ENV 2 Information about pollution control activities in the course of business / The report indicates that the polluting activities of enterprise has been and will be reduced. ENV 3 Information about conservation of natural resources such as the use of recycled materials, recycled glass, metal, oil, water, recycled paper ENV 4 Information about preventing and dealing with environmental consequences due to production process or exploitation of natural resources, such as soil improvement or reforestation. ENV 5 Information about strategies and supporting activities of enterprises towards environmental protection / improving environment ENV 6 Information about environmentally friendly design of infrastructure ENV 7 Report information on emissions ENV 8 Report information on discharging waste water ENV 9 Report information on solid waste disposal ENV 10 Report on information on environmental protection awards EMP II. Information on responsibilities to employees EMP 1 Information about enterprises complying with safety standards and working conditions for employees EMP 2 Information about health care for employees EMP 3 Information about training, financial support of training courses for employees EMP 4 Information about recreational activities for employees EMP 5 Information about homestay accommodation or plan of homeownership, food and other benefits for employees EMP 6 Information on compensation for employees, maternity leave, holidays EMP 7 Information about salaries, bonuses and other benefits for employees Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 46 EMP 8 Information about employee stock purchase plan EMP 9 Information about qualifications and experiences of employees EMP 10 Information on job stability of current and future employees EMP 11 Information about the relationship of enterprise with the union or workers through the movement of culture, sports, emulation of production EMP 12 Information about discrimination at workplace and jobs COM III. Information about responsibilities with community COM 1 Information about charitable activities with money, products / services or corporate employees participating in these activities, community events such as sports, culture and arts. COM 2 Information about employment opportunities for students, special people such as ethnic minorities, children of war invalids, disabled people, victims of Agent Orange and those in difficult circumstances. COM 3 Information about sponsoring projects of community health and providing health information to the Community / Supporting medical research COM 4 Information about establishing education funds or scholarships or educational conferences COM 5 Information about supporting victims of natural disasters, family under preferential treatment policy, victims of Agent Orange, contributions to charity, sponsoring disabled children and orphan COM 6 Information about supporting the development of industries, economic projects bringing benefits to the local COM 7 Creating jobs and developing skills for local people COM 8 Contribution to building electricity infrastructure, roads, schools and stations for localities CUS IV. Information about responsibilities to customers CUS 1 Information about product safety of enterprise CUS 2 Information about product quality CUS 3 Information about production and product development CUS 4 Commitment to quality, product safety and supporting when having trouble using the products CUS 5 Building customer relationships, collecting contributions and resolving customer complaints Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 47 Steps of the process: Corporate social responsibility is a complex field and there are many guiding standards, selecting ofstandards depends on economic conditions, laws and environment of each country. Therefore, in order to ensure the objectivity of the standard selection process, data collection and survey were conducted by two steps. Step 1 was conducting a survey of the 57 largest listed companies on the Vietnamese stock market, the survey year in2015. The purpose of this step was to adjust and select the appropriate indicators with conditions and characteristics of Vietnam. Step 2 was based on the indicators of information selected and corrected through step 1, the authors conducted the survey on 43 listed companies with 473 observations. This approach has been carried out by several authors, such as Branco & Rodrigues (2006), Gunawan et al (2009), Saleh et al. (2011), Tjia & Setiawati (2012), Bayoud et al (2012), Jitaree ). Technical process: First the authors read reports and find information related to the indicators in the evaluation list. If an enterprise did not disclose the ith indicator, the label is “0”, if the indicator was disclosed with general presentation or only with quantitative presentation without specific explanation, the label is “1”,if the indicator was disclosed with specific information about activities k, the label is “2”. After determining the score for each indicator for each enterprise in each year, the level of corporate social responsibility disclosure (CSRD) under the average number specified is determined as follows: CSRDj = ∑ ij 35 In which: CSRDj: indicator of information disclosure of jth enterprise Xij = 0 ifith indicator of information is not disclosed in enterprise j Xij = 1 ifith indicator of information disclosed in enterprise j is general information or quantitative information without specific explanation Xij = 2 ifithindicator of information disclosed in enterprise j is detailed information about specific activities 3.2.2. Financial performance variable Studies of authors such as Mahoney & Roberts (2007), Makni et al. (2009), Platonova et al. (2016), Cheung & Mak (2010), Jitaree (2015), Mohammed et al. (2016) used ROA as a measure of financial performance and verify the relationship between corporate social responsibility disclosure and corporate financial performance. Inheriting this research model, the author selected ROA as a measure of financial performance. ROA is determined by the following formula: The reason why profit before taxes was used is because the corporate income tax rate in ROA = Profit Before Tax Average Total Assets Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 48 Vietnam during the period 2006 - 2016 varied from 28%, 25%, 22% and 20%. Thus, to eliminate the effect of the corporate income tax rate on ROA, the author used profit before tax for analysis. 3.2.3. Controlling variables Following variables are controlling variables of the model: corporate size, leverage (debts/equity), foreign ownership, government ownership, revenue growth. Those are variables used by many authors when analyzing corporate profits. They are important factors influencing the estimation of variables affecting corporate financial performance. Corporate size (SIZE). Bayoud et al. (2012) argue that large-scale enterprises are likely to generate more profits than smaller ones. Therefore, the size of enterprises is considered as an important factor affecting the relationship between corporate social responsibility disclosure and corporate financial performance. Leverage (debts/equity) (LEV). Financial leverage is used by research as a factor affecting corporate financial performance. Nguyen & Dang (2017) argued that when a company is dominated by creditors, its financial performance may be affected. Therefore, financial leverage is necessary in the model to test the correlation between corporate social responsibility disclosure and corporate financial performance. Foreign ownership (FRO). Foreign ownership is an important variable affecting corporate fianncial performance (Zeitun, 2014). Zeitun assumed that foreign shareholders coming from developed countries with higher profit orientation, they have expertises, experiences and management skills which help boostingcorporate performance. Similarly, a study of Nguyen & Dang (2017) also concludeda positive relationship between foreign ownership and corporate fiancial performance Government ownership (GRO). According to Doan & Nguyen (2017) state-owned enterprises tend to have less incentive to maximize profits because their profitability and investment are guaranteed by the government. Therefore, GRO variable is necessary to consider the relationship between corporate social responsibility disclosure and corporate financial performance. Revenue growth (GRW). In a study of Zeitun's (2014) on capital structure and corporate performance of Jordanian countries demonstrated that revenue growth has a positive effect on corporate performance. The author explained that corporates with high growth rate will have better results because they can have more investment opportunities and more profits. 3.3. Data analysis To examine whether next year's financial performance has a positive relationship with corporate social responsibility disclosure of previous year, the author used the regression equation of dependent variable which is ROA and independent variable which is the level of CSRD and controlling variables as follows: ROAi,t= βo + β1CSRDi,t+ β2SIZEi,t + β3LEVi,t+ β4FROi,t + β5GROi,t + β6GRWi,t+ ui,t Method of data analysis used in this study is GMM. The reason why the author chose GMM Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 49 is that according to stakeholder theory and legitimacy theory, there is a two-way relationship between social responsibility disclosure and corporate financial performance. This statement was also demonstrated in the study of Ullmann (1985), Jitaree (2015). This results in an endogenous relationship between CSRDand ROA. Endogenous phenomena makesestimation unstable. To overcome this phenomenon, the author usedGMM developed by Arellano & Bover (1995) and Blundell & Bond (1998). 4. Results 4.1. Descriptive statistics According to survey on the level of corporate social responsibility disclosure of 43 listed companies during the period from 2006 to 2016, the average level of corporate social responsibility disclosure has increased steadily from 2006 to 2016. This indicates that more corporate concern about corporate social responsibility disclosure as well as focus on long-term development goals and bring benefits to the community. Source: created by authors Figure 1. Average level of CSRD of enterprises during the period from 2006 -2016 Detailed information on average level of corporate social responsibility disclosure on environmental responsibility (ENV), employment responsibility (EMP), community responsibility (COM), responsibilities to customers (CUS) showed that the level of environmental information disclosure is the lowest. This means that Vietnamese enterprises do not pay much attention to environmental responsibility. The activities of waste discharge, pollution control in business process, strategies to support environmental protection activities have not received enough care from enterprises. This is the difference between Vietnam and developed countries in Europe, where enterprises focus on information disclosure on environment, energy, recycling, pollution due to pressure from customers and investors on the environmental protection actions of enterprises. Information about responsibilities to customers announced by most enterprises includes information on product quality, product safety, production process, product commitment. Enterprises want to use this information to 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 CSRD CSRD Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 50 promote their products to customers and also to show corporate responsibility towards products they provide. Source: created by authors Figure 2. Level of information disclosure according to indicators of enterprises during the period from 2006 -2016 4.2. Correlation analysis Table 3. Correlation coefficient among variables in the model Variables ROA CSRD FRO GRO SIZE LEV GWR ROA 1 CSRD 0.4385 1 FRO 0.3525 0.488 1 GRO 0.1005 -0.0941 -0.1842 1 SIZE 0.0202 0.46 0.3797 -0.1636 1 LEV -0.4182 -0.2684 -0.2768 -0.1024 0.18 1 GWR 0.1569 0.0148 0.0716 -0.0352 0.0344 -0.0414 1 Source: calculation of authors based on Stata 12 Table 3 shows that corporate social responsibility disclosure in previous year has a relatively high correlation with ROA. This indicates that corporates with more information disclosure 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 0.800 0.900 1.000 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 ENV EMP COM CUS Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 51 have better asset usage efficiency. In addition, table 2 also shows that return on assets is positively correlated with firm size, the rate of government ownership, the rate of foreign ownership and growth rate but it is negatively correlated with the ratio of debt on equity. This proves that high margin firms are large-scale enterprises with a high proportion of government ownership, high proportion of foreign ownership, high revenue growth rates and low debt-to-equity ratio. 4.3. Regression analysis Detecting multi-collinearity Authors used command if with model OLS and command vif, uncentered with model REM and FEM. Tables 4showed that in the model REM and FEM variables SIZE have coefficient VIF> 10, with presence of multi-collinearity. SIZE variable will be discarded from the model. Table 4. Detecting multi-collinearity of variables in the model Variables Model OLS Model REM Model FEM VIF 1/VIF VIF 1/VIF VIF 1/VIF SIZE 1.58 0.611736 10.77 0.092864 10.77 0.092864 CSRD 1.63 0.631599 3.8 0.263049 3.8 0.263049 FRO 1.52 0.658508 3.5 0.285404 3.5 0.285404 LEV 1.35 0.738728 3.11 0.322046 3.11 0.322046 GRO 1.07 0.937104 2.59 0.386327 2.59 0.386327 GWR 1.01 0.992073 1.16 0.86241 1.16 0.86241 Mean VIF 1.36 4.15 4.15 Source: Calculation of authors based on Stata 12 Auto-correlation test Command xt serial was used to implement Wooldridge test with hypothesis Ho: there is no auto-correlation. The result showed Prob > F = 0.5122, there is no evidence to reject Ho. There is no auto-correlation in the model. Heteroskedasticity test Tác giả dùng kiểm định Breusch-Pagan test was used with hypothesis Ho: there is no heteroskedasticity in the model OLS. Breusch and Pagan Lagrangian test with hypothesis Ho: there is c in the model REM. Modified Wald test with hypothesis Ho: there is no heteroskedasticity in the model FEM. The results in table5showed that heteroskedasticity exists in all models. Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 52 Table 5. Summary of heteroskedasticity tests Model Test Chi-squar e statistic Pro>chi- square Test result OLS Breusch-Pagan 83.40 0.0000 There is heteroskedasticity REM Breusch and Pagan Lagrangian 140.87 0.0000 There is eteroskedasticity FEM Modified Wald 11435.69 0.0000 There is eteroskedasticity Source: Calculation of authors based on Stata 12 Generalized Method of Moment(GMM) As discussed above, there exists an endogenous relationship between CSRD and ROA. Endogenous phenomena makes estimation unstable. To overcome this phenomenon, regression method GMM was used which was developed by Arellano & Bover (1995) and Blundell & Bond (1998) to solve this problem. In addition, GMM can solve problems of heteroskedasticity and auto-correlation. In GMM, it is necessary to distinguish between instrumented variable and instrument variable. If the variables are predicted to be endogenous, then they are classified as instrumented variables according to GMM and then only the last values of these variables are appropriate tools. If the explanatory variables are defined as exogenous extrinsic variables, they are grouped into the instrument variable group (iv_instrument variable). The rationale for the GMM test is shown in the Hansen test results to determine whether the equation has used enough variables to measure the difference. The hypothesis in Hansen's test is as follows: Ho is an exogenous variable, meaning that it is not correlated with the error of the model. The p value of Hansen statistic> 0.05 is not sufficient basis to reject Ho means that the tool is exogenous. Then the model with instrument variables are appropriate. In addition, if Hansen test is strong, the number of instruments must be less than or equal to number of groups. Arellano Bond test about the auto-correlation with hypothesis Ho: there is no auto-correlation. In Arellano-Bond test AR (2) if Pr>0.05, we can include that there is no auto-correlation. In order to implement GMM, authors used command: Xtabond2 with endogenous variable CSRD and selecting two step to make two-step estimation. The results are shown in the table 6as follows: Asian Journal of Finance & Accounting ISSN 1946-052X 2018, Vol. 10, No. 1 ajfa.macrothink.org 53 Table 6. Regression result of GMM Variables Coefficient β P-value ROA (-1) 0.618209 0 CSRD 0.033219 0 FRO -0.00273 0.583 GRO 0.026129 0 LEV -0.0134 0 GWR 0.050793 0 _cons 0.02455 0 Number of obs= 430 Number of groups =43 Number of instruments = 31 Sargan test of overid. restrictions: chi2(34)=49.82Prob > chi2 =0.039 Hansen test of overid. restrictions: chi2(34)=38.44Prob > chi2 =0.275 Arellano-Bond test for AR(2) in first differences: z =-0.18Pr > z =0.859 Source: Calculation of authors based on Stata 12 The results of GMM in table 6are appropriate. Table 6 shows that there is no auto-correlation as Arellano-Bond test AR (2) Pr > z which is greater than α (5%). Furthermore, Hansen test shows that there is no endogenous phenomenon in the model as Prob > chi 2 of Hansen test which is greater than α (5%). Besides, the number of instrument variables = 41