Family Business Ownership, Corporate Governance and Firm Performance in Bangladesh

Abstract

Corporate governance is a crucial issue in achieving the corporate wealth maximization objective of the corporation. With the changes in the business environment through globalization, stakeholders are more concerned and demanding more clearly and reliably accounts, neutral behaviors, safeguards, and a clear view of the company’s progress. It is observed that as long as wealth maximization responsibilities are assigned mainly to investors holding stock companies, every country faces the challenge of how to minimize the cost of equity to a reasonable level. Besides, the ownership structures around the world reflect differences in several countries based on the theoretical development and financial environments. Available literature suggests that the ownership control mechanism is one of the most debated issues in corporate governance activities. However, different procedures have been used to categorize firms by control type. Furthermore, as the stock possession has become more delicate over time and with the size of the company, the amount of stock required for active control may decrease. The available literature documented that the extent to which family members continue to exercise control in the boardroom varies widely as their ownership of the firm decreases (Mace, 1971). However, there is an overall consensus that the concept of control envisions the ability to choose the board of directors for the corporations, either by voting power inherent in stock ownership or by position control attained by the organization when there is extensive diffusion of stock ownership. Different decision criteria for categorizing firms have resulted from changes in the amount of ownership required for control, shifts in ownership structures and board configurations among firms, and changes in the perceived control advantage over time. The fractions of the board of directors who perform and the extent to which ownerships are dispersed to ii different clusters have also been examined as elements of corporate control and governance devices. With the above background in mind, this thesis examines the relationship among family ownership, corporate governance practices, and firm performance within publicly listed companies in Bangladesh from 2015 to 2024. With structural changes in guidelines and principles in both domestic and international arenas, there have been numerous aspects to be addressed for CG practices and firm performance. This study, entrenched in CG theories such as agency theory, stewardship theory, stakeholder theory, institutional theory, and resource dependency theory, investigates the impact of family business characteristics and governance mechanisms on financial performance, assessed through return on assets (ROA), return on equity (ROE), and net worth (NW). Chapter two of this thesis provides an extensive literature review on family ownership, corporate governance practices, and firm performance across various international and domestic contexts, illustrating the diversity of methodologies and findings. Existing literature suggests that an in-depth examination of both earlier and recent studies focusing on the concentration of family ownership, the role of corporate governance, and how these factors are linked to a firm's financial performance is needed. Family-owned businesses make-up a significant portion of the global economy, as revealed in different studies and research. For example, Anderson and Reeb (2003) noted that over one-third of S&P 500 firms are family-owned. In 2018, a Global survey on Family Business by PwC (Price Waterhouse Coopers), London, UK, highlighted that the concentration of family ownership has a strong presence, with 64% of Indonesian businesses falling into this category. The way family firms are governed plays a crucial role in their success, often differing considerably from non-family enterprises. Globally, over two-thirds of iii businesses are family-owned, and their importance is increasingly acknowledged (London Economics, 2002). Jensen and Meckling’s (1976) ownership structure theory suggests that there is a positive association between managerial rights and firm worth. Supporting this view, studies by McConaughey et al. (1998) and Barontini and Caprio (2004) also find that control of the family firm is positively associated with firm efficiency, suggesting that higher levels of family ownership can enhance business outcomes. Anderson and Reeb (2003) and Miller et al. (2007) showed that family ownership may have a significant effect on business success. McConaughey et al. (1998) and Barontini and Caprio (2004) also find that control of the family firm is positively associated with firm efficiency, suggesting that higher levels of family ownership can enhance business outcomes. The earlier studies on Japanese family businesses (Yoshikawa, T. et al., 2010; Morikawa, M., 2013; Arikawa et al., 2019; Koji et al., 2020) have found consistent results with the parameters indicated above. The literature presents mixed findings: Pindado and Requejo (2015) identify a positive link, other studies—such as those by Fauzi and Locke (2012) and Wang and Shailer (2015)—report adverse outcomes in developing countries. Similarly, some scholars (Young et al., 2008; Miah, M.S., et al., 2023) argue that listed family firms do not outperform non-family counterparts, while others (Chahal & Sharma, 2020; Koji, K. et al., 2020) suggest that family involvement enhances firm value. Alves and Gama (2020) further contend that family influence on firm success is complex and cannot be labeled as positive or negative. Chapter Three of this thesis provides the concept regarding the institutional environment of corporate governance with its systems and practices, nature and style of the family firm that exists around the world and in Bangladesh. It also provides a detailed sketch of the iv financial ecosystem and the pattern of market governance, along with its performance in Bangladesh. Financial market structure includes the formal, informal, and semi-informal financial sectors of Bangladesh. It also encompasses the money market, capital market, forex market, and the regulators-Bangladesh Bank (BB), Bangladesh Securities and Exchange Commission (BSEC), Micro-Credit Regulatory Authority (MRA), and Insurance Development Regulatory Authority (IDRA) of the financial market of Bangladesh. Under the money market infrastructure, bank- like- State-owned Commercial Banks (SCBs), Private Commercial Banks (PCBs), Specialized Banks (SBs), Foreign Commercial Banks (FCBs), etc., non-bank financial institutions (NBFIs)-like- House Building Finance Corporation (HBFC), Bangladesh Development Bank Limited (BDBL), Palli Karma-Sahayak Foundation (PKSF), Grameen Bank (GB), etc., micro-financial institutions (MFIs), and leasing companies are noted. Similarly, under the capital market, stock exchanges – like; Dhaka Stock Exchange PLC (DSE) and Chittagong Stock Exchange PLC (CSE), the depository- Central Depository Bangladesh Limited (CDBL), merchant banks (MBs), asset management companies (AMCs), stock- brokers, stockdealers, venture capital, etc., are also discussed broadly. It is observed that as of fiscal year 2024, 61 scheduled banks and 5 non-scheduled banks were operating in the country. The aggregate ROA & ROE showed significant variability over the years, with ROA remaining slightly above 0%. Generally, below 1% in most years from 2012 to 2024, and the aggregate ROE rate began at approximately 8% in 2012, reached a peak of about 11–12% 2013, 2015, and 2017. The Amount of Non-performing Loans (NPLs) by major types of banks rose from 501.6 billion to 2113.91 billion from 2014 to the end of June 2024. The highest DSE index was 7329.00 in 2021, and the lowest was 4898.52in 2024. The highest-ranking aggregate market cap of DSE was 45.20 billion v in 2023, and the lowest was 2.54 billion in 2008. The market capitalization (Market cap)- to-Gross Domestic Product (GDP) ratio peaked at 50% in 2010, before the crash, and dropped to its lowest point at 4.2% in June 2006. Chapter four of the thesis has focused on the adopted research methods for this study and the reasons behind such selection. This chapter has also shed light on the conceptual framework employed to understand the relationship between family ownership and firmlevel profitability. Positivism was the guiding philosophy of the study. Secondary data collected from the books of account and the equity market could be taken at face value, and the insights stemming from these data points can perfectly represent the research area of interest. Choosing 'positivism' as the guiding philosophy is considered a mainstream paradigm in corporate finance research. While conducting the research, a deductive research approach was followed, whereby null hypotheses were constructed and tested. Determinants of firmlevel performance have already been well-identified in the empirical papers. So, as per the established literature, this research phenomenon is well-documented. It was empirical research; the established theory was tested in the context of Bangladesh. This study employed a gamut of quantitative research tools, which is consistent with the positivist philosophy. It is a multi-method quantitative study that encompassed correlation analysis, regression analysis, and other statistical methods. This research is based on archival research. The data collection is based on an in-house constructed Excel template. Secondary data was used for this research. The data was collected from annual reports of the listed firms of the Dhaka Stock Exchange PLC. As per the current regulations, annual reports are available in the public domain; there is no need to seek prior permission. Likewise, macroeconomic data were downloaded from the vi Bangladesh Bank and the Bangladesh Statistical Bureau websites. Again, this information is also available in the public domain and freely downloadable. One hundred four (104) firms representing both the financial and manufacturing sectors were included in the final sample. These firms were chosen based on convenience sampling. In order to manage 'survivorship bias’, newly listed and relatively young enterprises were eliminated from the sample. Companies that had been listed before 2015 and had been in continuous operation during the whole research period were taken into account. ‘Financial performance’, ‘Family ownership’ and ‘Corporate governance' were used as dependent variables, key independent variables and key moderating variables, respectively. Firms were classified as family businesses if the founder is in charge, or if family members hold important executive positions, or if family members are among the top ten shareholders, or if at least 50% of the board is made up of family members, or if a privately held family business retains ownership. The dependent variable - financial performance was measured using three primary metrics: net worth (NW), return on equity (ROE), and return on assets (ROA). The level of 'corporate governance' was measured through standard parameters such as board size, board independence, and board committees. The study has employed a multiple linear regression model in order to understand the relationship between family ownership and financial performance. Regression parameters can be estimated using different frameworks – the OLS (Ordinary Least Squares) framework, the MLE (Maximum Likelihood Estimation) framework, and the GMM (Generalized Method of Moments) framework. In this study, the researcher used the ordinary least squares technique to estimate the regression parameters. Since the research time frame covered a period of 10 years [2015-2024], in order to understand the nature of vii causality, panel regression was used. The baseline model was a fixed effects one, as suggested by the Hausman test. The researcher has reported the pooled OLS, Randomeffects, and panel corrected standard error model results, as well as the baseline research. Endogeneity concerns in the estimated effects were mitigated through the omitted variable channel. A gamut of tests for robustness was also run to assess the validity of the estimated effects in different contexts. Chapter -5 deals with the empirical analysis of the dissertation. It at first presented the descriptive statistics of the research variables, followed by a correlation matrix. For most of the research variables, the range is substantial, indicating a wide cross-sectional variation. The correlation between ‘Ownership Concentration’ and ‘ROA’ is weakly positive; on a similar line, the correlation between ‘Ownership Concentration’ and ‘ROE’ is also positive. By employing statistical tools like the Condition index and VIF, probable multicollinearity concerns were identified. It was revealed through these tests that multicollinearity did not pose any significant concern in this database. Likewise, probable outlier concerns were identified through standard tests like Cook's distance, COV ratio, and hat matrix. It was found that the outlier did not pose any significant concern in this database. By employing statistical tools like the White test, B-P test, and graphical technique, probable heterogeneity concerns were identified. It was revealed through these tests that the error variance is not constant. This problem was later mitigated by using robust standard errors while estimating coefficients. By summarizing the key conclusions of the A-D test and JB tests, and by graphically plotting the errors, it was concluded that the errors are not normally distributed. So, the researcher tried to estimate consistently by using a bigger sample size. Pesaran CD test indicated that there exists significant cross-sectional viii dependence in the data. Likewise, it was evident from the Lagrange multiplier test that there is statistically significant heteroskedasticity in the panel data model. There was robust statistical evidence to suggest that the random effects model is inconsistent. Therefore, the fixed effects model was the preferred specification for the panel regression model. The positive coefficient suggests that family firms tend to have superior performance compared to non-family firms. However, the effect is not statistically different from zero. As per the panel corrected standard error model, the beta coefficient concerning family ownership is positive and statistically significant. This suggests that family firms tend to have higher return on assets compared to non-family firms, holding other variables constant. The coefficients for Ownership Concentration, Age, Total Assets, and Board Committee are statistically significant. On a similar note, the beta coefficient concerning family ownership is positive and statistically significant as per the pooled OLS model. Before running the regression models, the preconditions were met. Moreover, the pre-conditions were as follows: the anticipated association between the factor and the independent variable was linear; data were gathered by random sampling; covariance between independent variables and the error term was zero; and there was no perfect multicollinearity among the pairs of independent variables. Endogeneity is a key statistical concern in regression analysis, which can lead to biased and inconsistent estimates. Three models were run to check whether the inclusion of initially omitted variables in the regression changed the results. In the 1st model, the GDP growth rate variable was added with the pre-specified set of control variables. Model 2 included all the standard firm-level controls along with the inflation variable. The baseline effects remained unperturbed when omitted variables were introduced in the model. Robustness of the model was evaluated by using winsorized data and alternative ix definitions of the dependent variable. The baseline effects remained unperturbed in these tests of robustness. Finally, the existence of cross-sectional heterogeneity was checked out. The hypothesized positive relationship between firm performance and family ownership concentration is supported in the cross-sectional regression, but the magnitude of the regression coefficient changes in the small-in-large firm case and the mature-in-young firm case. In a nutshell, the estimated effects demonstrated cross-sectional heterogeneity. In conclusion, this thesis documents the understanding of how family ownership, agency conflicts, and corporate governance interact to influence firm performance in Bangladeshi firms, using the model employed in the study. Moreover, the findings of the study indicate diverse results for family business ownership and corporate governance issues. It is argued that greater board independence, better regulatory execution, and increased transparency are critical for addressing agency problems and enhancing corporate performance.