Chapter 5: Conceptual framework
This study aims to cover 99 enlisted organizations which are locally incorporated. The research provides an overview of relationship between corporate governance and financial performance of an organization. Concepts of ownership structure, impact of ownership structure, effect of organization’s various shareholders on decision making, Agency theory, components of financial performance indicator, framework variables (dependent, independent), hypothesis development, data collection, analysis, interpretation of data and suggestions have been discussed.
5.1 Effect of Directorial Ownership
Internal governing mechanism defines boards which shape the governance of a firm with the other two axes in corporate governance: shareholders and managers. It is argued by Fama (1980) that presence of non-executive directors ensure proper action monitoring of the executive directors and also ensuring that policies pursued by executive directors are aligned with the one of shareholders’ interest. Much of power within the organization has been provided to the board of directors for corporate governance and also to increase executive accountability of non-executive directors. As per Fernández-Rodríguez, Gómez-Ansón and Cuervo-García (2004) there are strong perceptions that independent directors lead to better governance. However the existing empirical shows mixed result between the relationship of firm performance and board independence for the higher expectations from the role of the non-executive board members (e.g. Dalton, Daily, Ellstrand and Johnson, 1998; Dulewicz and Herbert, 2004; Peng, 2004; Weisbach and Hermalin, 2003). On the contrary, few researches suggest that firm performance will worsen if independent directors is in majority (Bhagat and Black, 1999).
Cost is incurred whenever management tends to pursue their own interest rather than the shareholders’, therefore monitoring by the board is important to minimize this potential cost. Monitoring by board is also important in order to reduce agency costs required to separate ownership and control. Thus, firm performance improves (Fama, 1980; Mizruchi, 1983; Zahra and Pearce, 1989).
As suggested by researchers who studied on the monitoring function that preference shall be for board dominated by outside directors (Barnhart, Marr and Rosenstein, 1994; Baysinger and Butler, 1985; Daily, 1995; Daily and Dalton, 1994; Weisbach, 1988). According to their argument, boards that consist of insiders primarily or outsiders who are not independent of the firm or management tend to not have incentive for monitoring management due to their dependency on the CEO of the organization. However, boards dominated by nonaffiliated directors are expected to monitor better as they do not have any disincentive for monitoring. But this hypothesis is not supported by any statistical data (Dalton, Daily, Certo and Roengpitya, 2003; Dalton et al., 1998).
Firm performance gets better with increasing managerial ownership (Jensen ; Mecking, 1976), but according to another research substantial ownership provides manager with the power to utilize the opportunity for personal benefits (Ruan, Tian ; Ma, 2009). These suggests that level of determining the level of ownership of manager is necessary to ensure firm performance.
5.2 Effect of Government Ownership
It is assumed that government ownership decreases concentration on profit maximization as government has both of the political and economic objectives which is not similar to the objectives of commercial firms (Estrin and Perotin, 1991). According to Shleifer and Vishny (1998), private ownership is preferred over government ownership as it is expected to be more efficient and profitable improving the performance of the firm. Megginson and Netter (2001) states that change of ownership that reduces government share leads to improved efficiency.
Wang (2004) did not find any noticeable negative relationship between government ownership and firm performance. On the contrary, it is argued by Bos (1991) that when government has significant ownership procession, they have incentive to monitor closely which increases profitability by reducing agency costs. Outcome of researches on relationship between government ownership and firm performance in numerous economies have contradicting results. The outcome indicates and encourages the necessity for additional studies on the topic.
5.3 Effect of Institutional Ownership
There is a possibility of institutional ownership to have an impact on the activities of managers directly through ownership. Transformation in corporate behavior could result from existence of institutional ownership due to better monitoring from investors. The aspect of having a considerable percentage of ownership in firm’s share affects the ability to monitor policies and activities of managers. Therefore, it is rational to consider a relationship between institutional ownership and performance of a firm.
The “black box” theory from the study of Berle and Means (1932) states that separation of ownership and control of modern companies reduces management motivation to ensure efficiency of the firm. The studies of Pound’s (1988), Brickley, Lease, and Smith (1988); and Kocchar and David (1996) analyzed various aspects and impacts of institutional ownership and firm performance. Institutional ownership had been classified into two groups: pressure-resistant institutional investors and pressure sensitive institutional investors. The research resulted negative relationship between pressure sensitive institutional investor and firm performance, whereas it resulted inversely for pressure resistant institutional investors.
Relationship between institutional ownership and firm performance vary from country to country as per researches of Gilson and Roe (1993), and Roe (1994). To add, findings of Shleifer and Vishny (1986) also support the greater effect of monitoring managers by large shareholders like institutional investors, whereas monitoring by board of directors have little or no ownership in firm.
A very important role in firm performance is played by institutional ownership which is done by reducing agency problem and increased monitoring. Firms with less institutional shareholding does not have the best governance structure or firm performance. For firms with higher institutional investor shareholding, the share prices are usually less marketable as they are held for longer periods.
The study paper of Dematz and Villolonga (2001) stated the relationship between institutional ownership and performance, with an outcome of ownership affecting firm performance in many ways.
5.4 Effect of Foreign Ownership
As per Gorg and Greenaway (2004), in international business strategy the outcome gained from foreign ownership of firms is the most challenging concern. Existing business literatures states that companies invest abroad to take advantage of the advantages which are not available in domestic companies (Dunning, 1993; Markusen, 1995; Caves, 1996).
It is widely known that foreign ownership plays a vital role in performance of the firm, especially in developing and transition economies. According to researchers (Aydin, Sayin ; Yalama, 2007), usually it is the multi-national institutions that performs better compared to domestically owned ones.
5.5 Effect of Public Ownership
When ownership of a company is private, the company is said to be private company. It can be expected that private ownership provides a higher guarantee of corporate governance by the role of external owners who monitor managerial performance and ensures complete focus on profitability as firm’s objective (Estrin, 2002).
Various outcomes have been generated by analysts who reviewed the performance differences between private and state owned enterprises. Most of the outcomes state that private owned enterprises has better financial performance. Among these, Andrews and Dowling (1998) and Parker (1997) can be highlighted as the analysis uncovered positive influence of private ownership with better financial performance. As corporate governance mechanisms vary around the world, change in ownership can effect performance of the companies.
Public ownership may represent personal claim to firm’s cash flow individually and from an agency perspective can be explained as principal directly monitoring the agent. On the contrary, government or institutional ownership is likely to have indirect monitoring where there are layers of agents acting on behalf of the principal. It is argued by Pound (1988) that institutional ownership is more efficient than public ownership due to the lower cost of monitoring. But public ownership is less risky as the number of shareholder is larger in number.
The larger number of shareholders interest, priorities and objectives has various effect to performance of the organization. It was observed by Thomas and Pedersen (2000) that greater number of shareholders has significantly higher impact on organization’s performance. Larger amount of shareholders relationship to better organization performance was examined by Nora and Rejab (2013).
Study of Wu and Cui (Zeitum & Tian, 2007) states the effect of ownership structure on a firm’s performance. Positive relation was found between ownership concentration and return on assets (ROA) and return on equity (ROE).
The research is aimed to provide an understanding regarding the relationship of ownership structure and different financial indicators of an organization. It can be more specified that the research is done to analyze if banks owned by public sector, directors, government, private, institutional owned organization differ significantly in performance.
5.6 Empirical Model
Usually three types of data are used for analysis: time series data, cross sectional data and panel data. In this study, data are collected from the time period of 2012 to 2016 on multiple variable to analyze the impact/change across the timeframe. Panel data is used for this research to examine the impact of ownership structure on firms’ performance in Bangladesh.
Random effects models and Fixed effects models are used for analyzing panel data in econometrics. In econometrics, Random Effects models and Fixed Effects models are used in the analysis of hierarchical or panel data when one assumes no fixed effects.
Two of the most common assumptions for panel data analysis are:
? Random effects assumption
? Fixed effects assumption
Random effects assumption states that specific individual effects are not correlated with independent variables. Whereas, fixed effect assumption states that specific individual effects are correlated with independent variables. Random effects model has been used to conduct the research.
I wanted to explore the nature of relationship between changes in ownership structure and its impact on financial performance. Various statistical methods (Chi-square, F-test) were used to analyze the association and extent of contribution within the variables. In order to assess the factors related to variables of ownership structure and variables of financial performance, multiple regression analysis was performed. The model used for this analysis can be explained as follows:
PER = ? + ?1 x1 + ?2 x2 + ?3 x3 + ?4 x4 + ?5 x5 + ?
PER (Performance): is a measure of financial performance or profitability of the organization. It consists of two variables in this research.
• ROA (return on assets): measured as ratio of net income after interest and tax to total assets. It is also known as the capacity of earning profit by a bank on the total asset employed.
ROA = NI/TA
NI: Net Income
TA: Total Assets
• ROE (return on equity): measured as ratio of net income after interest and tax to total equity. It can be also stated as the direct measure of return to the shareholders.
ROE = NI/TA
NI: Net Income
TA: Total Equity
x1: is the percentage of ownership by directors
x2: is the percentage of ownership by government
x3: is the percentage of ownership by institutions
x4: is the percentage of ownership by foreign investors
x5: is the percentage of ownership by public
?: the coefficients of the independent variables (explanatory variables)
After the formation of regression equation, research was forwarded to investigate correlation, R-squared (R2), meaningfulness and its coefficient. R-Square is the measure for determining relationship between dependent and independent variables. The value of R-Square coefficient determines the extent of dependent’s variability which is explained by independent variable. F-test is conducted to determine meaningfulness and T-test would be done for equation coefficients. To evaluate the hypothesis, two separate models were defined and were estimated for each dependent variable i.e. ROA and ROE. Below mentioned are the two models developed to conduct the research:
ROA = ? + ?1 x1 + ?2 x2 + ?3 x3 + ?4 x4 + ?5 x5 + ?1
ROE = ? + ?1 x1 + ?2 x2 + ?3 x3 + ?4 x4 + ?5 x5 + ?1
Chapter 6: Analysis and Results
This section is intended to provide details of empirical the study conducted. The evidences includes descriptive statistics, various tests and respective results.
6.1 Descriptive Statistics
In order to assure the accuracy of data and also to summarize and describe data meaningfully descriptive statistics test is conducted. Stated below are the behavior of data including dependent and independent variables.
Variable Observations Mean Std. Dev. Min Max
director 494 0.4338 0.1795 0 0.9
Government 494 0.0179 0.1013 0 0.9019
institute 494 0.1421 0.1024 0 0.5706
foreign 494 0.0276 0.0827 0 0.6318
public 494 0.3764 0.1759 0.0065 0.7694
roa 494 3.72% 0.0553 0.04% 0.4927
roe 494 11.73% 0.1110 0.18% 0.8869
Table 1: Statistical behavior of the data for the period of 2012 to 2016.
Table above contains descriptive statistics of the study, which indicates that mean of ROA for the period of 5 years is 3.72% with a standard deviation of 0.0553 and mean of ROE is higher equaling to 11.73% with standard deviation of .1110. This could be due to the fact that equity share in financial institutions is comparatively small which consists half of the sample. Primary source of funding in financial institutions are deposits which are used to finance investments and advances.
Mean values state that directors have the highest percentage of ownership among the other dependent variables, amounting to 43.38% with also highest standard deviation of .1795. Followed by public, institution, foreign and government ownership amounting to 37.64%, 14.21%, 2.76% and 1.79% respectively. It is also observed that foreign ownership has the least standard deviation with 0.0827.
6.2 Regression Analysis
Ownership Structure and Financial Performance (Dependent Variable: ROA)
Number of observations 494
F (5, 488) 11.96
Probability ; F 0.0000
Adjusted R-squared 0.1000
Root MSE 0.0524
Source SS df MS
Model 0.1643 5 0.0329
Residual 1.3414 488 0.0027
Total 1.5057 493 0.0031
roa Coefficient Std. Err. t P;|t| 95% Conf. Interval
director 0.1237 0.0967 1.28 0.201 -0.0662 0.3136
government 0.0375 0.0997 0.38 0.707 -0.1584 0.2335
institute 0.0110 0.0976 0.11 0.910 -0.1807 0.2027
foreign 0.1838 0.0997 1.84 0.066 -0.0121 0.3797
public 0.0432 0.0972 0.45 0.656 -0.1477 0.2342
_cons -0.0400 0.0962 -0.42 0.677 -0.2290 0.1490
Table 2: Ownership Structure and Financial Performance (Dependent Variable: ROA)
From Table 2, it can be inferred that all ownership variables have positive relationship with the profitability variable ROA. Majority of the resulted P-values of dependent variables are greater than 0.01, 0.05, 0.10 indicating that ownership structure does not have any significant effect on ROA. But foreign ownership has significant relationship with ownership concentration where the p-value i.e. .066 is less than 0.1, where 1% increase in foreign ownership will result in increase of ROA by 0.1838%. Organizations with maximum ownership to a single family can influence agency conflict that could result in reduced performance. Owners of institutions usually has short term objectives which might restrict organization from long term profitability. Greater government ownership could impact organizations’ financial performance negatively due to intervention in internal work process. Results are supported by Lehn and Demsetz (1985); McConnell and Servaes (1990) and in contrast to findings of Mehran and Cole (1998).
P-value of the model is 0.0000 (value less than 0.01, 0.05, 0.1) stating that at least one independent variable is significant. R-Squared value is 0.1091 which is not good but acceptable meaning that 10.91% of ROA is explained by all dependent variables together.