The Impact of Corporate Governance on Dividend Policy: An Empirical Evidence from Listed Companies in Sri Lanka

Purpose: This study aims to measure the relationship between corporate governance and dividend policy of Sri Lankan listed companies with the highest market capitalization, using the Agency theory. Methodology: The sample is based on the listed companies with the highest market capitalization at the Colombo Stock Exchange for a four-year period. The independent variables of this research include the board size, board independence, board gender diversity, board meetings, independent directors in audit committee, audit committee meetings, independent directors in remuneration committee and remuneration committee meetings. The dependent variables are dividend per share and dividend payout ratio. Descriptive analysis and Panel regression analysis were conducted to analyze the data. Findings: Independent directors in audit committee and return on assets have a significant positive impact on the dividend policy. Remuneration committee meetings have a significant negative impact on the dividend policy. However, board size, board meetings, board independence, board gender diversity, audit committee meetings, independent directors in remuneration committee, firm size and leverage have no significant impact on the dividend policy. According to the findings, corporate governance has an influence on the dividend policy of the listed companies during the period. Originality: This study fills the research gap in the local context, and this can be recommended for further research, changes in the academic concepts, and modifications in the accepted theories.


I. Introduction
The separation between owners and directors magnified, and the power shifted towards the directors as the listed companies grew and their shareholders became more diverse. Moreover, some of the directors and managers of those listed companies have been abused (Berle and Means 1932). Being the managers of other people's money, the directors and the management of companies cannot be expected to watch over it with the same vigilance they watch over their own (Smith A. 1776). Agents will tend to act in their own interests and not always in the best interest of their principal if both parties are utility maximizers (Jensen and Meckling 1976). This was the actual situation in the companies, and it led to the Agency problem. Well-known companies that collapsed because of the malpractices of governors (Toms 2019) showed the necessity of good governance for companies and corporate governance codes. The dividends mitigate agency conflicts between the managers and the shareholders. Dividends act as a component of corporate governance to align the interests of management with the interests of the shareholders. It shows that the dividend is a solution to the agency problem. Hence, there is a connection between the corporate governance and the dividend policy. It is important to study the relationship between corporate governance and dividend policy of the organizations to remain in the market and maximize the shareholders' wealth. International and local researchers have studied the impact of corporate governance on the firm dividend policy. The foreign researchers Setiawan and Phua (2013), Shamsabadi, Min and Chung (2016), Elmagrhi et al. (2017), Yarram (2015), Rajput and Jhunjhunwala (2019) have discussed the relationship between the corporate governance and the dividend policy. Ajanthan (2013), Ekanayake and Paranthaman (2016), Kulathunga, Weerasinghe and Jayarathne (2017) have studied the impact of corporate governance on the dividend policy in the listed companies of Sri Lanka. Previous literature concludes with positive, negative, mix and no relationship between the corporate governance and the dividend policy, and they are inconclusive. In Sri Lankan context, Kulathunga, Weerasinghe and Jayarathne (2017), have pointed out that only the companies listed under manufacturing sector are considered only the board characteristics of board size, board independence and board meetings have been addressed as the corporate governance variables to show the relationship between the corporate governance and the dividend policy, but not all the characteristics of the corporate governance. Corporate governance practices and their impacts on corporate performance in an emerging market: the case of Sri Lanka written by Manawaduge (2012) shows that the study is limited to industrial sector companies of the CSE and the sample includes only publicly listed companies in Sri Lanka. Ekanayake and Paranthaman (2016) in their research, state that all the listed companies in Colombo Stock Exchange were not considered but only the companies listed in S&P SL20 index. There is a dearth of studies within the local context regarding the relationship between corporate governance and the dividend policy of listed companies with the highest market capitalization in Sri Lanka. To bridge the research gap within the Sri Lankan context, this research considers more corporate governance variables such as the board gender diversity, the number of audit committee and remuneration committee meetings, independent directors in audit committee and remuneration committee other than the board characteristics (board size, board independence and board meetings) to achieve the main objective of examining the relationship between the corporate governance and the dividend policy of the companies listed in the Colombo Stock Exchange, which have the highest market capitalization. This study includes the sections of literature review and hypothesis development, methodology, findings, discussion, and conclusion.

Empirical studies on Corporate governance and dividend policy
Agency theory is a statistically robust theoretical approach to corporate governance. As Jensen and Meckling (1976) stated, agency theory can be defined as a contract with shareholders and directors by delegating decision-making authority on behalf of the shareholders. Agents act for their own benefit and not always on behalf of the principles. Berle and Means (1932) explained that the agency problem arises whenever the owner of the wealth (the principal) contracts with someone else (the agent) to manage their affairs. The separation between owners and directors magnified, and the power shifted towards the directors, where some of the directors have been abused, as the listed companies grew and their shareholders became more diverse.
As a response to the agency dilemma, a portion of companies' profit is called as the dividend, which is paid to the shareholders in each financial year according to the type and the amount of the shares they own. It is the sum of money paid regularly by a company to its shareholders out of its profits. Dividends can be considered as a corporate governance component to align the directors' interests with the interests of the shareholders. Dividends are the distribution of earnings allocated to the shares and other forms of participation in the equity of the incorporated private enterprises, public corporations and the co-operatives (OECD 2001).
According to Kulathunga, Weerasinghe and Jayarathne (2017),dividends are a signal of the firm's prospects due to the asymmetric information. Not only that, dividends also act as a corporate governance component to align the management's interests with the expectations of the shareholders. Shamsabadi, Min and Chung (2016) point out that, to reduce the existing conflicts between the corporate insiders and the outsiders or decrease the agency conflict between the majority shareholder and the minority shareholders, the dividend payments can be used. Yarram (2015) shows that dividend policy provides important monitoring and disciplinary roles. Rajput and Jhunjhunwala (2019) define dividend as the distribution of the residual profits among the shareholders, who are the suppliers of finance to the listed companies. The OECD Principles of Corporate Governance described that corporate governance associates a set of relationships among the management, the board, the shareholders and other stakeholders of a company.
Corporate governance accomplishes the structure through which the company's objectives are assigned, the methods of achieving those objectives and determines the ways to monitor the performance. According to Manawaduge (2012), the corporate governance means the authority to direct, organize and control the corporate entity. Kulathunga, Weerasinghe and Jayarathne (2017) define corporate governance as a mechanism, a process and the relations by which the organizations are monitored and directed. According to Rajput and Jhunjhunwala (2019), the governance role of the board of directors is the 'soul' of corporate governance as the shareholders have a delegated authority on the board to supervise and control the decisions made by the upper management. Setiawan and Phua (2013) point out that good corporate governance practice is one mechanism to protect minority shareholders. Benjamin and Zain (2015) show that corporate governance converses the agency problems that are emerged by the separation of the ownership and the control in the modern corporation. There are several empirical evidences in Sri Lanka regarding the relationship between corporate governance and the dividend policy. Ajanthan (2013) has examined the association between the corporate governance variables and the dividend payout of the hotels and the restaurant companies in Sri Lanka. According to that, board size, board independence, ROA, and debt-to-total assets are not significantly related to the dividend payout ratio. Kulathunga et al.Weerasinghe and Jayarathne (2017) have mentioned their aim of the study was to ascertain the relationship between the corporate governance variables and the dividend policy of the listed manufacturing companies at the Colombo Stock Exchange in Sri Lanka. The researchers claim that the board independence and ROA have a significant positive impact on the dividend policy and the board size has a negative influence on the dividend policy in the listed manufacturing companies in Sri Lanka. The objective of the study of Ekanayake and Paranthaman (2016) was to recognize the impact of corporate governance on the dividend policy of the listed S&P SL20 companies in the Colombo Stock Exchange. According to Ekanayake and Paranthaman (2016), the impact of the board size and the board independence is considered to be insignificant. In addition, they have shown that the firm size and profitability also explain the firm dividend policy. Several empirical evidence can be found in other countries except for Sri Lanka regarding the relationship between corporate governance and the dividend policy. Elmagrhi et al. (2017) have examined the degree to which the corporate board characteristics reign the level of the dividend payout ratio using a sample of small and medium-sized enterprises in the United Kingdom, from 2010 to 2013, which are listed on the Alternative Investment Market. Elmagrhi et al. (2017) prove that the board size, the frequency of board meetings, board gender diversity and audit committee size have a significant relationship with the level of the dividend payout ratio. The audit committee size and the board size have a positive connection with the level of the dividend payout when the frequency of the board meetings and the board gender diversity has a significant negative relationship with the level of the dividend payout in listed companies of UK. Board independence does not have any considerable influence on the level of the dividend payout.

Hypotheses Development
The hypotheses built for independent variables can be presented accordingly. Kulathunga et al. (2017) have stated that the size of the board has a negative influence on the dividend policy in the listed manufacturing companies in Sri Lanka. But Elmagrhi et al. (2017), Rajput and Jhunjhunwala (2019), Bokpin (2011) and Abor and Fiador (2013) have claimed that there is a positive relationship between board size and dividend payout ratio. Ajanthan (2013) and Ekanayake and Paranthaman (2016) have shown that the impact of board size on the dividend policy is deemed to be insignificant. Accordingly, the hypothesis first hypothesis can be formulated as; -H1: There is a positive association between negative association between the board size and the dividend policy. Kulathunga et al. (2017), Rajput and Jhunjhunwala (2019) and Abor and Fiador (2013) stated that there is a positive relationship between the board independence and the dividend payout ratio. Elmagrhi et al. (2017), Benjamin and Zain (2015) and Mehdi, Sahut and Teulon (2017) have stated that there is a negative relationship between board independence and dividend payout rate. But Ajanthan (2013), Ekanayake and Paranthaman (2016) and Bokpin (2011) have pointed out that the board independence is deemed to be insignificant related to the dividend payout. Accordingly, the second hypothesis can be formulated as; H2: There is a positive association between the board independence and the dividend policy. Elmagrhi et al. (2017) have stated a negative relationship between the board gender diversity and the dividend payout ratio. Accordingly, the third hypothesis is -

H3: There is a negative association between the board gender diversity and the dividend policy
According to Elmagrhi et al. (2017) and Benjamin and Zain (2015) there is a negative relationship between the frequency of board meetings and the dividend payout ratio. However, Mehdi, Sahut and Teulon (2017) claimed a positive relationship between the frequency of board meetings and dividend payout ratio. Bokpin (2011) has shown that there is no significant relationship between board intensity and dividend payout. Accordingly, the fourth hypothesis is; H4: There is a negative association between the board meetings and the dividend policy.
According to Elmagrhi et al. (2017) and Shamsabadi, Min and Chung (2016), there is a positive relationship between the independent directors in the audit committee and the dividend payout ratio. Thus, the fifth hypothesis is; Hypothesis H5: There is a positive association between the independent directors in the audit committee and the dividend policy.
Hypothesis H6: There is a positive association between the audit committee meetings and the dividend policy. Shamsabadi, Min and Chung (2016) have claimed a positive relationship between the independent directors in the remuneration committee and the dividend payout rate. Hence, the seventh hypothesis is; Hypothesis H7: There is a positive association between the independent directors in the remuneration committee and the dividend policy.
According to Shamsabadi, Min and Chung (2016), there is a positive relationship between the remuneration committee's frequency and the dividend payout. With that eighth hypothesis can be formulated as; Hypothesis H8: There is a positive association between the remuneration committee meetings and the dividend policy.

III. Methodology
This research is conducted on the Positivism paradigm to identify the association between corporate governance and the dividend policy in the listed companies in Sri Lanka. The deductive methodology is used in this research as the methodology which involves the collection of quantitative data. A theory and the hypotheses are built, and a research strategy is created to test the hypotheses. This study is based on the Quantitative approach which is related to the aspects that can be expressed in terms of quantity. Descriptive analysis and regression analysis are adopted for the analysis in this study.  Figure I illustrates the independent variables' direction towards the dependent variables and the influence of control variables towards the dependent variables. This research has 50 cross-sections for the period of 4 years from 2015 to 2018. Hence, the analysis includes 200 observations for this study. Since data is available for all the cross-sections for all periods, this is a Balanced panel. To analyze the data, the statistical software 'EViews 9' was used.   The Panel regression is the data analysis method and it uses hypothesis testing. Two alternative model developing approaches were followed to develop the regression model for panel data, namely, the Fixedeffect model and the Random-effect model. The Hausman test is conducted in order to test the most appropriate model from these two models, fixed-effect or random-effect. The initial regression equation can be developed as follows. The coefficients of the regression models indicate the change of the dependent variable when the independent variable is increased by one unit. The amount of the variance of the dependent variable that is explained by the independent variables is measured by the coefficient of determination which is denoted by R2. It indicates how well do the independent variables explain the dependent variable. The adjusted R2 measures the same as R2 but adjusted by the number of the variables and the observations. When the number of the variables is small, and the number of observations is large, the adjusted R2 is closer to R2.
The T-statistic measures the significance of each of the individual independent variables of the regression model. The t-values test the hypothesis that each coefficient is different from 0. The t-value of the independent variable has to be higher than 1.96 (95% confidence level) to reject this. If this is the case, then the independent variable has a significant influence on the dependent variable. The higher the value, the higher the relevance of the independent variable. Twotail P-values test the hypothesis that each coefficient is different from 0. To reject this, P-value has to be lower than 0.05 (95%). If this is the case, it can be said that the variable has a significant influence on the dependent variable.

IV. Findings and Discussion
Descriptive statistics were used to measure the level of corporate governance in the listed companies for the 4 years. To analyze the level of corporate governance, a comparison of the average values of the mean, standard deviation, maximum and the minimum for the four years is presented for each variable in the Table 3.
The mean value of the board size (LBS) is 0.95, which means the board size of the companies has remained the same during the years 2015-2018. The mean value of the board independence (BIND) also does not show a considerable change during the period, and it is 0.44. The mean of the board gender diversity (BGD) has increased from 0.21 to 0.24 within the years, which shows an increase in female directors' presence within the companies' director boards. The mean of the board meetings (LBM) has increased up to 0.83 but again decreased to 0.81. The mean value of the independent directors in the audit committee (AIND) increased up to 0.76 but again decreased to 0.76 during the years. The audit committee meetings (LAUDM) have increased throughout the periods, and the mean is 0.78. The remuneration committee independent directors (RIND) have also decreased to 0.74 during the period, but the remuneration committee meetings (LREMM) have increased up to 0.32 from 0.29. The mean values of both dividends per share (DPS) and the dividend payout ratio (DPO) have increased during the period, and the respective values are 0.10 and 0.59. The standard deviation of the board size (LBS), board independent directors (BIND) and the board gender diversity (BGD) are respectively, 0.12, 0.12 and 0.11 within the listed companies considered during the years 2015-2018. For the board meetings (LBM), the standard deviation has been increased to 0.25 during the years. The standard deviation of the independent directors in the audit committee (AIND) has decreased during the years and it is 0.18. The standard deviation of the audit committee meetings (LAUDM) has increased throughout the periods, and it is 0.20. While the standard deviation of the remuneration committee independent directors (RIND) has decreased and increased again during the period, the standard deviation of the remuneration committee meetings (LREMM) has increased and decreased again, and they are 0.19 and 0.30, respectively. The standard deviation of both dividends per share (DPS) and the dividend payout ratio (DPO) have increased during the period, and the respective values are 0.27 and 0.84. The maximum number of the board size (LBS) is 1.20 for all the four years within the listed companies. The maximum of the board independent directors (BIND) is 0.80. The maximum of the board gender diversity (BGD) is 0.50 within the years, which shows an increase in the female directors' presence within the companies' director boards. The maximum of the board meetings (LBM) is decreased to 1.23 during the years. The maximum number of independent directors in the audit committee (AIND) is the same during the years, and it is 1.0. The audit committee meetings (LAUDM) have increased throughout the periods, and the maximum is 1.20. The maximum of the remuneration committee independent directors (RIND) remained constant during the period, and it is 1.0, but the maximum meetings of the remuneration committee (LREMM) have increased up to 1.0. The maximum values of both dividends per share (DPS) and the dividend payout ratio (DPO) have increased during the period, and the respective values are 1.67 and 5.56. The minimum number of the board size (LBS) is 0.70 for all four years within the listed companies. The minimum of the board independent directors (BIND) is 0.25. The minimum of the board gender diversity (BGD) is 0.06 within the years. The minimum of the board meetings (LBM) remains as 0.30 during the years. The minimum number of independent directors in the audit committee (AIND) is the same during the years, and it is 0.33. The audit committee meetings (LAUDM) have increased throughout the periods, and the minimum is 0.48. The minimum of the remuneration committee independent directors (RIND) has also increased during the period, and it is 033, but the minimum meetings of the remuneration committee (LREMM) are zero for the period of four years. The minimum value of both dividends per share and the dividend payout ratio is zero. The Hausman test and the Panel Regression analysis were conducted using 'EViews 9', the statistical software, to analyze the association between corporate governance and the dividend policy. According to the Correlated Random effects-Hausman test, the probability value is greater than 0.05. Thereby, the null hypothesis is accepted, which states that the Random effect model is appropriate. Accordingly, the Random effect regression model is selected to measure the connection between corporate governance and dividend policy. When considering each of the variables individually as presented by Model 1 in Table   4, the return on assets (ROA) has a significant positive impact on the dividend per share (DPS) because the two-tail P-value is lower than 0.05. Kulathunga et al. (2017) and Elmagrhi et al. (2017) stated a significant positive relationship between the return on assets and the dividend payout rate. The number of remuneration committee meetings (LREMM) has a significant negative impact on the dividend per share (DPS) in the listed companies as the two-tail P-value is lower than 0.05. Shamsabadi, Min and Chung (2016) have stated a positive relationship between the remuneration committee's frequency and the dividend payout. As the adjusted R2 is greater than 60% and the probability value of Fstatistics is lesser than 0.05, the overall model is appropriate to test the relationship between corporate governance and the dividend policy. As presented in Model 2, the number of independent directors in the audit committee (AIND) has a significant positive impact on the dividends to earnings ratio (DPO) because the two-tail P-value is lower than 0.05. Elmagrhi et al. (2017) and Shamsabadi, Min and Chung (2016) have stated a positive relationship between the independent directors in the audit committee and the dividend payout ratio. The number of board meetings (LBM) has a significant negative impact on the dividends to earnings ratio (DPO) in the listed companies as the two-tail P-value is lower than 0.05. Elmagrhi et al. (2017) and Benjamin and Zain (2015) have pointed out a negative relationship between the frequency of the board meetings and the dividend payout ratio. According to the probability value of F-statistics, Model 2 is appropriate to test the association between corporate governance and the dividend policy as its value is less than 0.05.
According to the above findings, there is a significant connection between corporate governance and dividend policy. It means that corporate governance influences the dividend policy of the listed companies with the highest market capitalization in CSE during the period. Table 5 summarizes the study objectives, hypotheses and the results. Examine the relationship between independent directors in the audit committee and dividend policy H5: There is a positive association between independent directors in the audit committee and dividend policy Accepted Examine the relationship between audit committee meetings and dividend policy H6: There is a positive association between audit committee meetings and dividend policy Accepted Examine the relationship between independent directors in the remuneration committee and dividend policy H7: There is a positive association between independent directors in the remuneration committee and dividend policy Rejected Examine the relationship between remuneration committee meetings and dividend policy H8: There is a positive association between remuneration committee meetings and dividend policy Rejected Source: Constructed based on the outcome of Analysis

V. Conclusions
The main objective of this research is to examine the relationship between corporate governance and the dividend policy of listed companies with the highest market capitalization. According to the findings, the independent directors in the audit committee and the return on assets have a significant positive influence on the dividend per share. The board size, board gender diversity and the firm size have an insignificant positive relationship with dividend payout. The board meetings and the remuneration committee meetings have a significant negative impact on the dividend payout. The board independence, independent directors in the remuneration committee and the leverage have an insignificant negative impact on the dividend payout. As the theoretical and practical implications, this study is significant for further research and changes in academic concepts and knowledge regarding corporate governance and the dividend policy. The analysis and findings of this research can be used to compare with the actual policies of listed companies in Sri Lanka, scholarly understanding of good governance of corporates, theory building and for local and international future research studies. This can be used to implement modifications in accepted theoretical constructs and concerning changes in procedures. Some limitations of this study also can be identified. This research considers only 50 listed companies from the Colombo Stock Exchange, but there are 283 companies listed in the CSE in Sri Lanka. Even though there are 20 business sectors according to Global Industry Classification Standard, the sample of the study includes only 14 industries. The time period of the research is only four years, from 2015 to 2018. Future researches can be conducted using a higher number of listed companies considering all the industrial groups of GICS in Sri Lanka. The study has considered the listed companies with the highest market capitalization. Hence, future studies can be conducted with another classification of listed companies. The researchers can use more corporate governance variables other than the variables included in this study for more than a period of four years.