Corporate governance and its effects on financial performance of banks evidence from selected private commercial banks in Ethiopia

Corporate governance is not an abstract goal but exists to serve corporate purposes by providing a structure within which stockholders, directors, and management can pursue most effectively the objectives of the corporation and thereby it maximize the financial performance of the company for its shareholders. This paper investigates and examines the corporate governance and its effects on financial performance from the evidenced data gathered from selected privately owned commercial banks in Ethiopia. By adopting panel data, Purposive sampling technique has been utilized to select the respondents from the existing target population. This paper will include independent corporate governance variables such as Board size, Meeting frequency of board, Audit Committee Size, Liquidity Ratio, Existence of outside director, Board Educational Qualification, Board Gender Diversity, chief executive compensation and controlling variable of bank age and bank growth. Random Effect Model GLS regressions analysis was adopted to analyze the annual data generated from the annual statements of the selected private commercial banks of Ethiopia and annual reports of National Bank of Ethiopia covering a period from 2010 to 2018G.C.In addition, primary data of structured open ended questionnaires used to support the finding of secondary data. The presence female director, chief compensation, audit committee size, Educational Qualification, meeting frequency of board, liquidity ratio was found to be a positive and significant relationship with return on asset. In the other side, board size has a negative and insignificant association with return on asset. As a recommendation, Private commercial banks have to have a moderate number of board size and they have to hold meeting and discussion for the future fate of banks. Audit committee in private commercial banks should be given as a major determinant. As much as possible, banks have to hire and train female board and it is better in that banks should encourage and motivate executive with compensation.


INTRODUCTION
Corporate governance refers to the way in which a corporation is directed, administered, and controlled; it concerns mainly the relationships among the various internal and external stakeholders involved as well as the governance processes designed to help a corporation achieve its goals. Of prime importance are those mechanisms and controls that are designed to reduce or eliminate the principal-agent problem (Aldamen et al., 2003). Corporate governance is a means used to maximize the financial performance of banks and concerned stakeholders such as employees, shareholders, the board of directors and executive management (Ashenafi and Kelifa, 2013). Many researchers have worked on this area (Brooks, 2008;Creswell, 2009;Freeman et al., 2004;Imam and Malik, 2007). Even though all these studies were on the effects of Corporate Governance on banks' performance most of the researchers found a mixed and contradicted result for measuring financial performance. Effective corporate governance is mandatory to enhance financial performance, that is financial institution generally and banking sector particularly. This would ensure the management (the agent) runs the firm for the benefit of one or several stakeholders (principals). According to Imam and Malik (2007); Kelifa (2012); Kyereboah-Coleman (2007); Khatab et al. (2011); Kim and Rasiah (2010) and Kothari (2004), there is a key regulatory framework for the financial institution as well for banks such as interest rate regulation, minimum capital requirement, foreign exchange currency controls, inflation which are the concern of government. Malik (2007); Kelifa (2012); Kyereboah-Coleman (2007); Khatab et al. (2011); Kim and Rasiah (2010) and Kothari (2004) stated governance issue ensures performance in banks; and maintaining the government requirement has different problems such as non-existence of organized stock exchange, high government intervention and control, absence of national standards of corporate governance, absence of effective accounting and auditing and lack of educational qualification of board. Maintaining, ensuring directing and supervision of good corporate governance for financial institution specifically for banks is the responsibilities of the National Bank of Ethiopia because it controls all aspect in relation to money supply and other financial issues. In developing countries especially in Ethiopia, corporate governance issues need critical concentration so as to maintain financial performance as different studies conducted in Ethiopia provided. Mohammed et al. (2014); Melvin and Hirt (2005); Naser (2012); Olajide (2013);O'Connell and Cramer (2010) and Rose (2007) focused on corporate governance effects on microfinance institution located in Ethiopia from 2005 to 2014 by using the Corporate governance variable such as educational qualification, board gender diversity, business management experience of board member, board size and by quantitative research approach by multiple linear regression model with proxy of financial performance of banks return on equity. They concluded gender diversity, board educational qualification, business management experience were negatively related to banks' financial performance of return on equity. Even though research was performed on this area corporate governance on the financial performance of micro finance institution was not fully assessed as the preliminary objective was stated and the result or findings were very contradictable. The reason for indicating that the finding is contradictable is first its real board educational qualification and the expertise they invest on the company depends and has positive effects on financial performance MFI of developing economies than developed countries. The second reason is that individual feeling or attitude towards corporate governance may not be only concluded by quantitative approach rather it was more advanced and the researcher may acquire as well may be saved from contradictable findings, if descriptive study with qualitative study approach were employed. The other study was (Rose, 2007;Vafeas, 1999) conducted on the areas of ownership structure and corporate governance effects on financial institution especially on microfinance institutions. Behind these studies, there are many limitations that the variables used to show corporate governance were opposite. From them almost all the studies show ownership structure, management capability and board structure on financial performance as determining variables for corporate governance. The second limitation is that only MFIs were assessed without corporate governance on banks using quantitative study data from 2004-2010; 6 year data were used and qualitative study is not included. Still additional variable should be considered to show corporate governance effects on financial performance of banks (Vafeas, 1999;Wooldridge, 2006). For the sake of current study, corporate governance variables such as board size, meeting frequency of board, existence of audit committee, liquidity ratio, existence of outside director performing internally, board gender diversity, executive compensation, and other effects towards financial performance of selected private commercial bank by using return on asset were proxy variables, dependent variable. The control variables were bank growth or net asset and bank size or the number of private banks branches.

Theoretical literature review
Corporate governance is the manner in which the power of a corporation is exercised in the stewardship of the corporation's total portfolio of assets and resources with the objective of maintaining and increasing shareholders' value and satisfaction of other stakeholders in the context of its corporate mission (O'Connell and Cramer, 2010). In this regard, there are a number of theoretical perspectives which are used in explaining the impact of corporate governance mechanisms on firms' financial performance. The challenges of corporate governance could help to align the interests of individuals, corporation, and society through a fundamental ethical basis and fulfill the long term strategic goal of the owners (Mohammed et al., 2014).

Agency theory
Agency theory was proposed by Imam and Malik (2007); Khatab et al., (2011) and Rose (2007); in field of Economics, directed at the agency relationship, in which one party (principal) delegates work to another (agent) who performs that work. Its focus is on shareholders' interests that necessitates security by split-up incumbency of the role of board; the segregation of management role and ownership lead to a serious matter of control over the risk attitude. This theory is on mechanism where the board of directors and owners act as the monitoring authority whereas agents are the managers (Wooldridge, 2006).
In order to prevent the managers to abuse their position and protect their interests, the stockholders may use several different mechanisms. Another way to overcome this problem is rewarding managers financially. The best way is to calculate their bonuses as a percentage of the realized profit of the company (Soludo, 2004).

Empirical review
The study conducted in developing economies (Melvin and Hirt, 2005). examined the relationship between corporate governance practices and performance of commercial banks in Kenya. The population of the study was the 45 banks licensed by the Central Bank of Kenya as at the end of 2010. The study adopted a census study approach because of the small population and the banks are easily accessible. Two methods of data analysis were employed, the description analysis which provides some average of relevant variables and regression analysis to establish the relationship between the corporate governance variable (explanatory variable) and firms' performance/dependent variable/ over the period of study. The research concluded that corporate governance practices, directors' effectiveness, management effectiveness have a positive relationship with banks' performance. The study investigated the effect of corporate governance on the performance of commercial banks in Kenya; 37 commercial banks were studied in Kenya over a period of 2005-2009. The finding is not compatible with the objective; all the variables were negatively related and the objective was not accomplished in that corporate governance has no relations with board size, CEO duality, and directors' independence as well on ROE and ROA.
From this study, the limitation is that the variables adopted such as board size, board independence and ownership concentration have a contradictable result on financial performance proxy of return on equity significantly and the opposite on return on asset, which is unthinkable (Olajide, 2013). The researchers used lodgit model analysis with a sample of 120 firms listed on the S&P300 during the period of 2008 and 2009. The study revealed that smaller audit committees with more experience and financial expertise are more likely to be associated with positive firm performance in the market. It is also found that longer serving chairs of audit committees negatively impact accounting performance. In addition to the above review conducted by different study on different time, the other study conducted in emerging countries (Naser, 2012;Soludo, 2004) and other study was done on developing countries on corporate governance and most of them found contradicting result with the stated objective.
Even if there is limited number of research work on the evaluation of the effect of corporate governance on private commercial banks' financial performance, there were mixed results from previous studies pertaining to the relationship between corporate governance mechanisms and financial performance; some of the reviewed papers were on the existing regulation and study conducted as follows.
O'Connell and Cramer (2010) and Rose (2007) argue that the regulation of NBE is non-sufficient to protect minority rights because the main objective of NBE is financial regulation and which is just one aspect of governance.
In Ethiopia also most of the study was done on microfinance institution and from them the study conducted by Olajide (2013); O'Connell and Cramer (2010); Rose (2007) and Soludo (2004) as well as other adequately performed on corporate governance effects on profitability of insurance companies for different time span and almost they used the variable such as board size, CEO duality, meeting frequency, and board independence as well as capital structure as Explanatory variable of corporate governance and almost all have reached on the same result in which the variable stated affect corporate governance positively.

Existing literature summary, limitations and identification of gaps
From the study conducted there almost all have performed on focused on the level of compliance with national codes of corporate governance and international principles, has examined the level of corporate governance disclosure in companies and has analyzed the state of the implementation of systems of regulation in developing nations while Some studies have investigated the difficulties that affect the improvement of corporate governance, and recommendations have been made to solve these problems. In addition to above, most of the literature that has examined the relationship between corporate governance and firm performance has utilized a few elements of corporate governance as well as the major point in all research conducted on different time is that investigation of the relationship between corporate governance and firm performance shows a significant concentration on developed countries, but this issue has attracted only limited research on developing countries especially in Ethiopia.

Board size and financial performance
Some of the studies, such as (Olajide, 2013) and others, state that the size of the board in bank operations affects the performance of bank positively and significantly. In consideration of this, some other such as (Vafeas, 1999) found that the larger board size will reduce the performance and results in easier CEO control.
Ha 1 : Board size has negative effects on banks Return on Asset.

Meeting frequency of board and financial performance
The empirical study conducted a different individual such as (Vafeas, 1999;Wooldridge, 2006) and other concluded that the minutes and meeting frequency that the board waste is negatively related to the financial performance of banks as well most of the study is contradictable and resulted in inconclusive findings.
Ha 2 : Meeting frequency of board has negative effects on banks Return on Asset.

Audit committee size on financial performance
The existence of an audit committee composed of external board members in the firm will create a transparent and credible environment between management, external auditors and the board members.
Ha 3 : Existence of a large number of the audit committee has negative effects on banks Return on Asset.

Liquidity ratio and financial performance
Most of the empirical study conducted by Kyereboah-Coleman (2007) and Yenesew (2012) found that the liquidity ratio requirement from NBE to private commercial banks in Ethiopia have a negative effect on their profitability while some other study by Olajide (2013) and Rose (2007) was found positive relationship thereof.
Ha 4 : Regulation of Liquidity requirement has negative effects on banks Return on Asset.

Board composition (outside board member) and financial performance
According to agency theory, advocators, outside directors are more likely to show objectivity in their deliberations and are willing to consider diverse groups in making their decisions. Contrary to this, there is some other study which defends this that the existence of outside director does not influence the operation at all and there is no consensus behind the existing literature.
Ha 5 : Proportion of outside director on the board has negative effects on banks Return on Asset.

Board of educational qualification and financial performance
The most study conducted by different researcher support this and finds a positive relation thereof. Creswell (2009) and Kelifa (2012) suggested in the study that directors educational qualification has the power to make banks operations more effective than that which have not and he found that positive and significant effects thereof.
Ha 6 : Board member educational qualifications have positive effects on banks Return on Asset.

Board gender diversity and financial performance
Empirical studies such as Brooks (2008) and Yenesew (2012) support this, in that the existence of female director in board positively affects the profitability because controlling function and could be one of several tools used to minimize potential agency issues; while some other study (Brooks, 2008) found the opposite, that the presence of board diversity influence none on profitability.
Ha 7 : Board gender diversity measured by the presence of female director in the board has positive effects on banks Return on Asset.

Director or chief executive compensation and financial performance
The agency-based theory generally supports that there is a positive relationship between executive compensation and firm performance (Naser, 2012;Soludo, 2004; Freeman et al., 2004). Only a few of the study concluded that the presence of compensation for chief executives and director motivate as well as encourage them for better performance and it has a positive relationship with profitability (Khatab et al., 2011;Melvin and Hirt, 2005) and while most defend this. While in emerging countries, this was not given concern. This study considers this variable of corporate governance mechanisms. Ha 8 : Chief executive Compensation has negative effects on banks Return on Asset.

METHODOLOGY
In order to achieve the objective of this research, the study adapts Explanatory Research Design. The rational for choosing explanatory research design is; it used to examine the cause and effect relation between dependent and independent variable. The required data to analyze the effects of Corporate Governance on Financial Performance was Mixed Approach in nature. In addition, Primary data was qualitatively adopted. The primary and secondary data for this study (the financial statements) was collected from private commercial banks head office majorly from Board, Board secretary and delegated staff. The target population of the study was from Private Commercial Banks in Ethiopia totaled 18. Among them, 12 banks were selected as respondents. Judgmental or Purposive sampling technique was adopted to select Twelve banks by making criteria and purpose (based on the age and availability data for the study period), which fitted as well make the study effective from objective perspective. Panel data comprising both Cross-sectional and time series data were chosen over the unbalanced one, thus periods covering Nine Years from 2010-2018 of 12 sampled banks were taken as respondents for the study.

Model specification
In examining the effects of corporate governance on financial performance from the evidence of selected commercial banks in Ethiopia, the econometric model developed from ground and gap in the existing literature is as follows: ROA=α+β1Bszeit+β2MfreBit+β3ACsZit+β4Rglqit+β5ExouDit+β6Be dQit+β7BgedVit+β8CexcoPit+LnBgrz+LnBasZ+ €it

RESULTS AND DISCUSSION
As above revealed that, regarding dependent variable, Return on asset mean value is 3.19 and with 6.7 and -3.2 maximum and minimum value respectively. Board Size in which the maximum number of board size in private commercial banks is 12 and the minimum Board size is 7. The meeting frequency held in commercial banks has a maximum of 60 and above times and the minimum meeting held was 10-20 times. As far as audit committee size is concerned, the results show above those banks have a maximum of 3 audit committee and a minimum of 1 audit committee for the past nine consecutive years. As far as board gender diversity, measured by the female directorate in the board, is concerned, the Commercial Banks have 5 female board maximum and there are also the banks having without the female board in the banks. Chief executive Compensation for the board was given a maximum of 3 times and a minimum of 1 time in a year for past nine consecutive years (Table 1).

Assumption: The average value of error term is zero (E (ut) = 0)
According to (9 and 15), the existence of the constant term in the regression equation will prevent the violation of the assumption the average value of errors is zero. Multiple regression model developed in this study included the constant term the assumption of the mean value the error term is zero was not violated.

Assumption: Test for normality assumption
For checking the presence of normality the following hypothesis is developed to support the null hypothesis in that there is normality at 5% significance level and the P   Table 2, it was concluded that violation of the normality assumption of CLRM does not occurred in all residuals Shapiro wilk test P value is above 0.05 (P value from the output of this tests ranges between (0.07-1.00).

Assumption: Test of autocorrelation
The Durbin-Watson test examines the hypothesis that H 0 :  = 0 (implying that the error terms are not autocorrelated with a first-order scheme) against the alternate, according to Soludo (2004) and Wooldridge (2006) (Table 3).
Ho: Residuals are not linearly auto-correlated Ha: Residuals are not linearly auto-correlated The DW test statistic value was 1.709884 a total observation of 108 (9*12) were used in the model also 8 Variables and an intercept term were modeled. So, the null hypothesis of no autocorrelation is within the inconclusive region. Therefore, the study can now conduct model selection for further regression. The rule of thumb provided by Creswell (2009); Kelifa (2012) and Rose (2007) revealed that the P-value from the Hausman test is statistically significant (less than five percent) the fixed-effect model is preferred in favor of random effect, otherwise the random effect model is selected.

Ho: Random Effect Model is appropriate Ha: Fixed Effect Model is appropriate
In this case, random effect model is appropriate because the p-value is above five percent (0.05) which is 0.3986 as shown on appendix part four in that Hausman specification test of random effect model is the best estimator of outcomes.

Random effect model regression results
In this study Adjusted R square value more than 50% which is 0.7734 to mean that, 77% of the explanatory variable of corporate governance stated on the model determines the financial performance of private commercial banks in Ethiopia while the other determining factor out of this model scores 23%. The null hypothesis of F-statistic (the overall test of significance), which states the Adjusted R-squared is equal to zero was rejected at a 5% significance level (Table 4). The model of explanatory and controlling variable determining the return of asset the regression results of coefficient and standard error were entered on the equation depending on the results and it is the model is hereunder; ROA=1.06817+0.0193914Bszeit+0.0156882MfreBit+0.1 641561ACsZit+0.0197362Rglqit+0.0724136ExouDit+0. 1402747BedQit+0.1736035BgedVit+0.2354656CexcoPi t+0.4067403LnBgrz+0.3631693LnBasZ+ €it

Board size and return on asset
From the regression results revealed on the above Table  4, the coefficients of a board size-0.0193914 and the P value is 0.721 which is negatively and insignificantly related with banks financial performance. This means that the larger number of board size has lower financial performance and vice versa.
Not only the negative relationship between Board size and return on an asset but also the relationship is insignificant. Most of the study conducted previously such as (Kim and Rasiah, 2010;Melvin and Hirt, 2005;Rose, 2007) found in their studies that the Board size is negatively and insignificant relationship with the financial performance of the banks. These findings are also consistent with the corporate governance theory in that it paves the way for reducing the agency cost.

Meeting frequency of board and return on asset
The meeting frequency has a positive and significant relationship with banks financial performance because as the meeting frequency increase, board exerts much time with the board resulted in an increment in the performance of operational activities of the management. So, in the regression and questionnaires result, the results of the alternative hypothesis (Ha 2 ) stating that Meeting frequency of board has a negative impact on banks return on asset is rejected and the null hypothesis stating Board Meeting frequency is positively and significantly related with return on asset is accepted at 5% of the significance level.

Audit committee size and return on asset
As per the respondents, the existence of audit committee have a direct positive relationship with financial performance in that the level of fraud and misstated value on the financial statements will be assessed and resulted to minimize the level of non-corrective activities on the banks as well as the segregation of duties rested on financial statement understatement and overstatement. The hypothesis (Ha 3 ) states that Existence of a large number of the audit committee has a negative impact on banks Return on Asset is rejected from this study and the null hypothesis stating there is the positive and significant relationship among audit committee size and return on an asset will be accepted at 5% significance level.

Liquidity ratio and return on asset
The random effects regression results from Table 4 shows that the correlation coefficient and P value between liquidity ratio and banks return on asset is 0.0197362 and 0.893 respectively. This implies that as the liquidity ratio and financial performance are positively and insignificantly related. According to the respondent's reflection on primary data, liquidity ratio has a direct relationship with the financial performance of banks because the banks having the liquid asset in their banks results to opportunity to mature their liability. This resulted in better performance and growth of the banks because the liquidity ratio must be kept so as to cope up with current liability. At 5% level of significance, the null hypothesis is failed to reject in that liquidity ratio and return on asset is positively and insignificantly related to the regression and questionnaires result reflects.

Existence of outside director performing internal task and return on asset
As the respondents view, existence of outside board member but performing internally have a positive effects on financial performance of private commercial bank measured by return on asset in that, the task that they take as concerned body in each of the decision boosts or enhance the value of banks and thus results in increment of return on asset in best achieving the goal. The null hypothesis accepted indicates that the existence of outside board performing an internal task in the bank has a positive and significant correlation with return on asset.

Board educational qualification and return on asset
Table 4 random effect model showed that the board educational qualification and banks return on the asset are having the correlation coefficient and P value of 0.1402747 and 0.00 respectively. In addition, above 70% of the respondents answered that the level of education is determining issues because the board experience, know-how and transparency matters the corporate governance in enhancing financial performance. Depending on the regression results, discussion and questionnaires analysis conducted, the alternative hypothesis (Ha 6 ) indicating that Board education qualification has positive effects on return on asset is accepted.

Presence of female board in the bank and return on asset
The regression model showed on the Table 4 of board gender diversity measured by the female board in the bank and private commercial banks financial performance proxy of return on asset is conducted. The correlation coefficient and table value of P is 0.1736035 and 0.010 respectively. As revealed by the 75% respondents, the presence of female board in the banks increases financial performance because banks with diversified board gender composition, especially the existence of female board members, enable to see at the operations of the banks from different perspective. Accordingly, the alternative hypothesis (Ha 7 ) indicating that Board gender diversity measured by the presence of female director in the board has positive effects on banks return on asset is accepted at 5% of significance level

Executive compensation and return on asset
The correlation coefficient of 0.2354656 shows that there is a positive relationship between executive compensation and return on asset of the banks. In relation to this, the Pvalue results 0.017 indicate that there is a significant relationship between them. These findings really support theory of corporate governance which is agency theory in that, manager who has direct responsibility must be compensated in order to reduce the principal-agent problem; thus, to come up with better financial performance. From the above results of the analysis, the null hypothesis indicating chief executive compensation has positive effects on return on banks is accepted

CONCLUSION AND RECOMMENDATIONS
Moderate Board Size can enable them make adequate and quick decision to be forwarded. Accordingly, they minimize the number of Board Size in that the cost incurred for each will be decreased and results in high profitability. Banks' Board should meet and deal with future fates and growth of the banks at least three and above times in a month since they have positively and significantly related statistically. This study advice the banks to encompass different audit committee as much as possible since it is positively and significantly related to banks financial performance. Also, independent auditor needs too. Banks should also encourage board members to participate in the tasks of banks internally as revealed from the findings .This study recommends banks to train and hire more qualified board that have the knowledge of banking sector, management capability as far as Board Educational Qualification of CG is concerned. This study advice the bank should hire, train, and participate Female Board as much as possible so as to have high financial performance and then enhance CG. This study recommends that banks should maintain the Liquidity position and make cash ready to recover the liability and escape from the relevant mandatory cost of borrowing. It is good that banks encourage and motivate managers by