Board Independence and Company Performance in East Africa

The research is self-financed by the author Abstract Purpose : The main aim of this study was to examine influence of board independence on financial performance of listed companies in the East African Community (EAC) by examining the influence of board independence on performance of EAC’s listed companies and compare the influence of board independence on company’s financial performance before and after the operationalisation of the EAC- Common Market in 2010 to make recommendations on how board independence can enhance of hinder a company’s financial performance Design/Methodology/Approach : We adopted a positivist paradigm in a quantitative analysis using non-probability sampling to select forty-two EAC listed companies. We developed hypothesises basing on secondary data from databases (DataStream, Eikon, Mint Global Bureau Van Dijk) and company’s annual reports. SPSS was used to carry out regression diagnostic test and generate descriptive statistics, correlation, and regression results. Findings and Recommendation: Our findings indicated that to some extent board independence had a statistically significant influence on company financial performance measured by RoA, RoE, TbQ but not PeR. Secondly; we also discovered some negligible changes in board independence for most listed companies for the period before and after operationalisation of the EAC - Common Market. Based on our inconclusive statistical results and the extant literature review, we recommend that EAC listed companies adopt a code of best practice that emphasises an increase in board independence.

measures are strong measures of performance; they provide reliable results, especially during period of economic crisis, and are relatively free from speculation as compared to market-based measures. Furthermore, according to Baker and Anderson (2010), accounting-based performance measures have the advantage of being directly linked to a company's financial survival.
Market performance measurements on the other hand include market-based ratios such as market share, number of customers and the Tobin's q ratio (Kim 2015;Rossi, Nerino & Capasso 2015;Zagorchev & Gao 2015) and PeR (Aik, Hassan & Mohamad 2015;De Aguiar, Pinheiro & Oyadomari 2014;Shah, Haldar & Rao 2014). Unlike accounting-based performance measures, market-based measures use market value data to determine company financial performance (Eikelenboom 2005). The main advantage of market-based approaches is that they are less prone to managerial manipulations or creative accounting (Mulsow 2011). Market-based measures of performance are also risk-adjusted, especially where the Capital Asset Pricing Model (CAPM) is used in calculating the company's market value (Mans-Kemp 2014). Hence, it is believed that using market-based ratios like TbQ provides better measures of performance, which are more reliable, long term-oriented and risk-adjusted (Kim 2015;Li, SL & Tallman 2011). This study adopted a combination of accounting-based (RoA and RoE) and market-based (TbQ and PeR) to maximise the advantage of each of the methods. These measures have been adopted given their wide usage in previous business finance, corporate governance, finance and company performance researches (AdeBIite 2012; Ansong 2013; Bhagat, S. & Bolton 2008;Mans-Kemp 2014;Okiro 2014).

Aims of the study
The overarching aim of this study is to examine the influence of board independence on the companies' financial performance within the EAC by addressing the following specific objectives: i.
Examine the influence of board independence on performance of EAC's listed companies ii.
Compare the influence of board independence on company's financial performance before and after the operationalisation of the EAC-Common Market to make recommendations on how board independence can enhance of hinder a company's financial performance We hope this study will provide new and relevant knowledge about board independence in developing countries in general, and EAC countries in particular. The findings will also help in encouraging economic growth and development by identifying board independence as good corporate governance practices (OECD 1999(OECD , 2004(OECD , 2015.

LITERATURE REVIEW
A number of theories have been used in research to explain the relationship between board independence and company financial performance. These include the agency theory, the stakeholders' theory, the resource dependency theory, the legitimacy theory, and the institutional theory.

Resource dependency theory
According to resource dependency theory, the value of any board member is measured by his or her ability to attract and bring in external resources into the company for the benefits of the shareholders (AdeBIite 2012). Hence, companies appoint board members based on their ability to promote the company access to some critical external resources that can enhance company value (Johnson, Daily & Ellstrand 1996). These resources may include expertise, public influence, networks or influential professionals and industry experts (Abdullah, H & Valentine 2009). According to Pfeffer and Salancik (2003), recruitment of an external (independent) board member is likely create a favourable linkage between the company and its external environment. Such networks provide a company with new information and support from its operational environment (public or government) which legitimises a company with its environment. Hillman, Shropshire and Cannella (2007), hence posits that independent directors are the most critical resource needed by any company; their presence on the board portrays equality and better career opportunities for prospective employees. This helps an organisation to attract better quality employees, who will drive increases in the company's performance.

Legitimacy Theory
According to Suchman (1995, p. 574), the legitimacy theory states "the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions". The theory assumes existence some beliefs which gives rise to a social contract between an organisation and the society. This gives approval to an organization to carry out business, and in turn an organisation is expected to provide accountability to the society (Deegan 2013). The legitimacy theory thus emphasises the need for an organisation to comply with the rights and expectations of its host community, failure of which would result in sanctions against the company in the form of restricted access to the market and other key resource (Deegan 2013).
In the East Africa, most companies' operations are controlled by governments via the fiscal policy (tax systems) and the general public who provide market and input resources of materials and labour. Other listed company's stakeholders (with interest and influence) include financial institutions and recently the security Research Journal of Finance and Accounting www.iiste.org ISSN 2222-1697 (Paper) ISSN 2222-2847 (Online) Vol.12, No.12, 2021 4 markets who provide investment capital and the media that influences public opinion about the company's products and services (Mendelow 1991;Tilling 2004). Therefore, listed companies have a legitimate responsibility to meet the expectations of the society (Ashforth & Gibbs 1990). The legitimate theory has a big influence on corporate governance in the EAC; listed companies often appoint outside (independent) directors from public, government, advocacy and special interest groups, and financial institutions to legitimise their operations.

Transactional cost Theory
This theory was developed from the agency theory (Coase 1937). According to Williamson (1985) and Pound (1993), like in agency theory firms will incur high transaction costs due to the principal -agent conflicts. Ordinarily, principals (shareholders) contracts an agent (manager) to run the company and maximise their wealth. But company performance is often influenced by the manager's behaviour. Due to lack altruism, a manager will often focus on personal pay, privileges and prestige at the expense of shareholder's wealth (Jensen, Michael C & Meckling 1976). The Transaction cost theory thus focuses on the agent's responsibility in managing (minimising) transaction costs such information gathering, negotiation, re-negotiation and enforcement (Williamson 1985). Williamson (1988), thus posits that a company's transaction costs are influenced by bounded rationality (the notion of homo economics) and opportunism. Because of management bounded rationality and opportunism, managers tend to make opportunistic decisions on some transaction costs for personal gains or due to limited information or time (Simon, 1982). The transaction cost theory hence calls for stringent measures in managing company transactions to reduce operational costs. The presence of independent directors thus increases company monitoring and supervision while mitigating the risk of excessive transaction costs by the management. Hence the presence of outside directors reduces transaction and agency costs (Benkler 2002;Coase 1937;Jensen, Michael C & Meckling 1976).
2.4 Regional Economic Integration and the EAC Regional economic integration refers to a joint trade policy by countries aimed at reducing or eliminating trade barriers amongst the member states (Salvatore 2004). It is achievable by linking the member country's economic, social and political infrastructures to enhance each other's economic growth and development (Kehoe 2006). According to Schiff and Winters (2003), there has been a growing trend towards regionalism, which makes economic integration a vital element among the global economies. The idea of regional economic integration was first emerged after World War II (Schiff & Winters 2003), it then expanded and was embraced by most developed countries globally. One of the best examples of such economic integration was the European Union (EU), which was formed by Luxembourg, Netherlands, United Kingdom, Italy, France and Germany in 1951. It later became the European Economic Community (EEC) in 1957 and the European Free Trade Association (EFTA) in 1960 (Radebaugh, Sullivan & Daniels 2015). The success of the EEC has encouraged the spread of integration to other regions of Latin America, Asia and Africa (Schiff & Winters 2003).
According to Mwasha, (2011), economic integration in Africa started in 1910s with the creation of the Southern African Customs Union with membership from South Africa, Botswana, Lesotho, Swaziland and, Namibia. Today Africa has many economic integrations such as; East African Community (EAC), Common Market for Eastern and Southern Africa (COMESA), Economic Community of West African States (ECOWAS), Southern African Development Community (SADC), Economic Community of Central African States (ECCAS) and the Arab Maghreb Union (AMU) among others. Despite the increase in numbers, most economic integration in Africa have not been very effective in delivering the expected economic growth and development (Mwasha 2011). Some of the reasons behind their poor performance include lack of commitment by member countries, political and economic instabilities, external debt problems, currency fluctuations, and narrow tax bases (Matthews 2003).
The EAC economic block is made up of six partner states, namely: Kenya, Uganda, Tanzania, Rwanda, Burundi and South Sudan. It has aggregate population of about 145.5 million, a total land mass of 1.82 million square kilometres and an aggregate Gross Domestic Product (GDP) of US$ 147.5 billion (EAC 2016). The EAC was first formed in 1917 between Uganda and Kenya and was later joined by Tanzania in 1927, before being dissolved in 1977 (EAC 2016). After its dissolution, the member countries came to a mediation agreement regarding the distribution of the community's assets and liabilities, which was concluded with the signing of the 1984 treaty amongst the then three EAC member states (Kasaija 2004). In the early 1990s, the East Africa region experienced changes in leadership which resulted in significant social, economic and political transformations and the need to revive the EAC as a regional economic integration and promote the development of security markets (EAC 2016;Kasaija 2004).
In the mid-1990s, Uganda, Kenya and Tanzania formed Capital Market Authorities to regulate the activities of the East African Community Security Markets and in 1997, a joint regulatory body known as the East African Securities Regulatory Authorities (EASRA) was formed and mandated with regulating the EAC Security Markets' activities, fast-tracking the EAC Security Markets' integration, harmonising the legal frameworks and market infrastructure as well as developing policy guidelines for the capital markets' growth incentives within the EAC (Yabara 2012). Meanwhile, in 2005, the EAC's economic integration had changed from a free trade zone to the East African Community Customs Union, which later became the East African Common Market (EAC-Common Market) in 2010. This transformation led to the advent of corporate governance in the early 2000s among all the EAC countries. Kenya was the first country to introduce a code of corporate governance in 2002 followed by Uganda andTanzania in 2003 (CMA, K 2002;CMA, U 2003). Rwanda introduced their code of corporate governance in 2010 (RSE 2015)

METHODOLOGY
The main aim for this study was to examine the influence of board independence on the financial performance of listed companies within the EAC. To achieve this objective, we adopted the positivist paradigm under deductive approach, using quantitative techniques to identify the causes and effects of social phenomena (Collis & Hussey 2013). We adopted the quantitative approach as often used in corporate governance and company performance studies (Alagha 2016; Heenetigala 2011; Waduge 2011). We also adopted a deductive approach, in which hypotheses were developed from the review of existing literature, and data was collected and used to confirm or negate the proposed hypotheses. We based our hypothesis testing on secondary data from published statistics and annual reports. The use of a deductive approach and hypothesis testing method is a consistent with a quantitative research approach (Gill and Johnson, 2010) and was adopted in this study due to its advantages over the qualitative approach. For instance, the use of numerical measurement in the quantitative approach makes it easier for research analysis and presentation of results for explanatory purposes. Additionally, the quantitative approach has less bias error than the qualitative approach (Collis & Hussey 2013). According to Veal (2005), a qualitative approach does not often provide researchers with the same level of rigour as a quantitative approach.
We obtained our quantitative data from secondary sources, which is the most commonly used method for obtaining data in most business finance, accounting and company performance research studies (Adams, R, Hermalin & Weisbach 2008; Alagha 2016; Haniffa, R. & Hudaib 2006;Heenetigala 2011;Kiel & Nicholson 2003;Klein 1998;Laing & Weir 1999;Tshipa 2015;Waduge 2011). Our choice of secondary data was therefore based on its suitability in addressing the aforementioned aims of this study.

Dependent variables
We adopted Return on Assets (RoA), Return on Equity (RoE), Tobin Q ratio and Price Earnings Ratio (PeR) as our independent variables as commonly used performance measurement in business, finance, accounting and

.1 Return on Assets (RoA)
The RoA ratio is a measure of company efficiency in generating income from its total assets (Lesakova, 2007). It one of the commonly used accounting measure of performance when evaluating companies' economic health and the proficiency of investment portfolios (Basarab, 2010;Lesakova, 2007). According to Ingram and Albright (2006), the RoA ratio provides a links between company's annual operations and investment activities. The RoA ratio hence measures the efficiency management in using company assets (Lesakova, 2007). The RoA is calculated as: Annual Profits after Interest and Tax Total assets at the year-end A higher RoA ratio is an indication of management efficiency in utilising company assets to generate a higher value for the investors (Basarab, 2010; Lesakova, 2007).

Return on Equity (RoE)
The RoE evaluates company performance based on investment returns. It focuses largely on the management's ability to make an investment return for equity holders in form of financial surplus after deducting all expenses. RoE was calculated in this study as: Annual Profits after Interest and Tax Total equity at the Year-end A higher RoE ratio is an indicator of management's ability to generate extra income to the owners/shareholders 3.1.3 Tobin's Q ratio (TbQ) Unlike RoA and RoE, the TbQ ratio uses market values when evaluating company performance. It's computed as the ratio of company market value to the asset's book value (Chorafas, 2004).. A lower TbQ ratio is an indicator of poor market confidence in the company shares, which could be attributed to poor governance that reduces company profits (Weir, Laing, & McKnight, 2002). According to Gross, (2007), TbQ ratio is a hybrid measure of performance, that is based on both accounting and market-based data. TbQ was calculated as: Research Journal of Finance and Accounting www.iiste.org ISSN 2222-1697 (Paper) ISSN 2222-2847 (Online) Vol.12, No.12, 2021 6 Year-end market capitalisation Total assets at the Year-end According to Leng (2004), the TbQ ratio measures the company's growth prospects using the company's asset base. A TbQ value of one indicates that the company's market value is equal to the total book value of its assets. If the ratio is greater than one, the company's market value is greater than its asset's book value, and hence management is deemed to have created more value for owners (Chorafas, 2004). On the other hand, a TbQ less than one indicates that the company's market value is lower than the total book value of its assets, which may suggest that the company's share price undervalued (Chorafas, 2004).

Price earnings Ratio (PeR)
The PeR is used to estimate company's market value using the earnings per share (EPS) at the year-end (Bernstein & Wild, 1993). The value of the company's PeR depends on its governance policies, the past performance future performance potential, and industry risks (Bernstein & Wild, 1993). If a company posts good past performance results (profitability), high future growth potential (positive net present value projects), it would have a higher PeR than a similar company with poor past performance and low growth potential (Bernstein & Wild, 1993). Equally, a company with good governance policies will attract affirmative market perceptions which makes it appear less risky than its peers within a same industry. Likewise, high and stable dividend pay-outs will influence a company's PeR because of its market-signalling effect. Consistent dividend pay-outs are a good signal to the market that a company is both financially strong and committed to rewarding its owners (Lease, John, Kalay, Loewenstein, & Sarig, 1999). This study calculated PeR using the following formula: Company's year-end share price Earnings per share (EPS) PeR is influenced by the company risk, particularly the finance risk or the risk of having debt capital within its capital structure. The presence of debt capital affects the company's bottom line and EPS, hence reducing earnings growth. This also increases the financial risks such as bankruptcy, tax exhaustion, leverage associated with higher PeR ratio (Bernstein & Wild, 1993).

Hypothesis Development
Our hypotheses were uncured of the assumption that higher proportion of external / independent directors improve the company financial performance (Daily & Dalton 1992;Schellenger, Wood & Tashakori 1989). It's a common practice in the EAC for companies to have not less than one-third of the board composed external independent directors (CMA, K 2002;CMA, U 2003;USE 2014). This is in compliance with the codes of corporate governance and good practice in EAC which recommends that any listed company should have at least one-third of directors as non-executive directors. According to Abdullah, (2004), there is a positive and significant correlation between board independence and company financial performance measured by returns on assets, profit margin and earnings per share. Similarly, Fauzi and Locke (2012) studied listed companies in New Zealand from the period 2007-2011 and discovered a significant association between the number of non-executive directors and company performance.
In many developed and emerging economies, board independence is considered as paramount contributor to effective company monitoring and performance (Siddiqui 2015), hence the company laws and regulations mandates all companies to have independent boards for company monitoring and control. The United States of America (USA), for instance introduced the Sarbanes-Oxley Act in 2002 to enforce the need for companies to have a majority of independent directors on their board (Li, X 2014). The choice of the board independence this study was therefore based on the assumption that board independence enhances a board's monitoring capacity, hence it was deemed useful to analyse the extent to which it may or may not influence the EAC-listed companies' financial performance before and after operationalisation of the EAC common market in 2010. The researcher thus hypothesised that: i. There is a significant relationship between board independence and company financial performance (H1). ii.
There has been a significant change in board independence after the operationalisation of the EAC-Common Market (H2).

Data collection and sample selection
We adopted secondary data source mainly because the data required for this study was readily available in company annual reports and databases. The use of secondary data is consistent with other accounting, finance and corporate governance research studies, in which researchers clearly stated that they used secondary data to save time and money for data collection (Beasley, MS 1996;Heenetigala 2011;Ngwenya & Khumalo 2012;Okiro 2014;Waduge 2011). The type of secondary data used in this study includes journal articles, e-books, press releases and websites. The researcher also obtained financial data from well-established databases, namely Orbis Research Journal of Finance and Accounting www.iiste.org ISSN 2222-1697 (Paper) ISSN 2222-2847 (Online) Vol.12, No.12, 2021 Bureau Van Dijk, DataStream, Eikon and Mint Global Bureau Van Dijk. Additional data sources, such as published companies' annual reports and company websites was used in the current study. Microsoft Excel and Statistical Package for the Social Science (SPSS) version 23 was used for data handling and analysis. Excel was used to manage and format data before exporting it into SPSS for statistical annalysis. SPSS was used to carry out the preliminary diagnostic tests, descriptive statistics, correlation, and regression analyses. We adopted non-probability sampling to select the forty-two listed companies used in ours study. We used listed companies instead of private companies because their information is publicly available (Okiro 2014). We initially intended to use as a population all the listed companies on the EAC Security Markets in 2008/2009 and 2013/2014. A total of 108 companies were listed on the EAC Security Markets as at 31st June 2016. Unfortunately, not all the listed companies qualified to be included in the sample because they were either not listed on the EAC stock market for the full financial years 2008/2009 and 2013/2014, or their annual reports were not available from the DataStream, Eikon and Mint Global Bureau Van Dijk databases.

Diagnostics statistical tests
We adopted a number of diagnostic statistics to ascertain the validity of the regression models used to test the relationship between board independence and company financial performance. We tested normality of the error distribution, homoscedasticity and coefficient of variation tests.

Normality of the error distribution
We tested the normality of error distribution using the Shapiro-Wilk test, which is considered as the best measure for identifying any departures from normality distribution due to skewness or kurtosis (Baty et al. 2015). Under the Shapiro-Wilk test, the null hypothesis states that there is no difference between the observed distribution of survey scores and a normally distributed sample error. Hence, if the critical alpha is larger than the obtained pvalue, the null-hypothesis is rejected (Shapiro & Wilk 1965). The results of the normality test are discussed in section 4 below.

Heteroscedasticity test
We tested homoscedasticity using the Koenke tests in the SPSS macro developed by Garcia-Granero (2002). According to Watson and Teelucksingh (2002), Koenker test results are easy to understand and the test can be carried out with small data sample, which makes it suitable for this study's small data sample. The null hypothesis of the heteroscedasticity test implies that there is no conditional heteroscedasticity and thus the individual-specific or time-specific variance error components are equal to zero (Park 2011). The homoscedasticity assumptions of regression were met, and consequently the current study adopted the regression analysis to test the study's hypotheses.

Coefficient of variation
The coefficient of variation is calculated as the ratio of the standard error of the estimate to the mean of the dependent variable (Cook & Krishnan 2015) According to Aiken, West and Pitts (2003), there is no acceptable threshold for a coefficient of variation since it vary across different disciplines but the lower the coefficient of variation, the smaller the residuals relative to the predicted value. The lower coefficient suggests a relatively good model fit compared to a higher coefficient of variation. Nonetheless, there is a drawback in using the coefficient of variation to assess model fit when a variable contains a positive and negative coefficient of variation values and the mean is close to zero, which gives a misleading result (Cook & Krishnan 2015). The results of the coefficient of variation are discussed below

Correlation analysis
Our diagnostic tests revealed no approximate normal distribution. Hence, this study used Spearman's rank correlation analysis to examine the relationship between board independence company performance indicators. Spearman's rank correlation has been used in previous research on corporate governance and company performance studies (Abdullah, SN 2004;Alagha 2016;Heenetigala 2011).

Ordinary least squares (OLS) regression
We adopted OLS regression which is considered a straightforward method of statistical analysis (Zikmund et al. 2012). We adopted our model using the following equation: Yi = βo + β1 X +Ɛ 1. Where: Yi = the dependent variable X = the independent variable β0 = intercept β1 = slope and Ɛ = error term. From the above equation, we derive equation 2 -6 used in this study: Yt = βo + β1Bi +Ɛt 2.

RESULTS
Below the findings of our study represented by the diagnostic test, descriptive statistics, correlation and regression tests. Table 4.1 presents the test results of the normality, heteroscedasticity, and Coefficient of variation used in this study These values are all above the significance value of 0.05, which therefore confirms that the residual is normally distributed in this study (Shapiro & Wilk 1965).

Heteroscedasticity test
We carried out heteroscedasticity test using the Koenker tests in the SPSS macro developed by Garcia-Granero (2002). The null hypothesis in the Koenker test assumes that homoscedasticity is present (or heteroscedasticity is not present).  Cook and Krishnan (2015), the lower the coefficient of variation, the smaller the residuals relative to the predicted value, the above results thus suggests a relatively good model-fit.

Descriptive statistics
We adopted the mean, median, maximum, minimum and standard deviation to identify the statistical characteristics of the study's independent and dependent variables as indicated in   Table 4.2, indicates a mean of 40.58%, median of 40.11%, standard deviation of 4.26, minimum of 32.76% and maximum of 50.57%. For 2013/2014, the mean RoA was 40.01%, median 39.08%, standard deviation 9.09, minimum 19.11% and maximum 63.86%. There is a positive mean value in both periods, which indicates that the EAC-listed companies in this study on average generated a similar positive return of about 40% on assets for their shareholders before (2008/2009)

Tobin's q ratio (TbQ)
The TbQ descriptive statistics results in Table 4

Board independence
During 2008/2009, board independence had a mean value of 76.87%, median of 81.82%, standard deviation of 14.55, and minimum of 33.33% and maximum of 91.67%. Four years after operationalisation of the EAC-Common Market, the mean board independence was 76.11%, the median was 80.91%, the standard deviation was 15.55, the minimum board independence remained at 33.33% and the maximum was 93.33%. The above results indicate that the majority of companies in the EAC had more external than internal directors on their boards. The reason for the minimum board independence of 33.33% in the study data is that the corporate governance codes in all the EAC countries recommend that at least one-third of directors should be non-executive directors (CMA, K 2002;CMA, U 2003;USE 2014). Hence, the 33.33% minimum percentage for board independence in /2009.2) demonstrate that all companies maintained the degree of board independence recommended by the EAC's code of governance and good practice.
Research Journal of Finance and Accounting www.iiste.org ISSN 2222-1697(Paper) ISSN 2222-2847(Online) Vol.12, No.12, 2021 Table 4.3 indicate Spearman's rank correlation for 2008/2009 had the following pair of variables with significant correlation at 1% significance: RoA and RoE with correlation coefficient of 0.662, PeR and TbQ with correlation coefficient of 0.571. Those with a significant e pairs of variables that displayed significant correlation at 5% significance were: PeR and RoA with correlation coefficient of -0.325, PeR and RoE with correlation coefficient of -0.393, RoE and BI with correlation coefficient of -0.338. While in 2013/2014, the following pair of variables showed significant correlation at 1% significance: PeR and TbQ with correlation coefficient of 0.409, TbQ and RoA with correlation coefficient of 0.603, TbQ and RoE with correlation coefficient of -0.43, RoA and RoE with correlation coefficient of 0.687, The above correlation figures indicate lack or lower correlations between the dependent and independent variables which is common with many business, accounting and finance research using non independent variables with no linear relation.

Ordinary least squares (OLS) regression results
We interpreted OLS regression results considering the coefficients, the R-squared values and the P values. According to Bowerman et al. (2003) the variable coefficient shows the size of effect that one variable has over the dependent variable. Positive coefficients indicates how much the dependent variable is expected to increase when the independent variable increases by one unit, holding other variables constant and the reverse is true for the negative coefficient (Tabachnick & Fidell 2006;Tabachnick, Fidell & Osterlind 2001). The R-squared is the fraction of the variation in the dependent variable that is predictable by an independent variable. The P value for each independent variable tests the null hypothesis that the coefficient is equal to zero. If the null hypothesis is not rejected, the respective independent variable has no effect on the dependent variable (Field 2009

The influence of Board Independence on RoE
The RoE represents the net amount of profits created by the company using shareholders' funds (Khatab et al., 2011). Table 4. 5, below, presents a summary of the regression results on the relationship between board independence and RoE. 0.000*** *** Significant at 1% level, **Significant at 5% level, * Significant at 10% level Source: Own Source According to the OLS regression results in Table 4.5, the adjusted R-squared value in 2008/2009 was 0.41, which suggests that about 41% of the total variability in TbQ can be explained by board independence but board independence does not statistically and significantly influence on TbQ. The 2013/2014 results, shows an adjusted R-squared value of 0.74, which demonstrates that about 74% of the total variability in TbQ can be explained by board independence. Board independence had a statistically significant negative influence on TbQ (p=0.00<0.01), hence, an increase in board independence by one percent would result in a decrease in the TbQ of 64% holding other variables constant.