Journal of Economics and Business

This paper examines the determinants of board independence in the banking sector of Bangladesh. The study applies a multivariate panel regression analysis for the thirty banks listed with DSE covering the period from 2006 to 2016. We use the proportion of independent directors to total number of directors on the board to measure board independence. Findings of the empirical analysis show that a board is likely to be more independent when independent directors have relevant skill and knowledge as required by prevailing corporate governance regulation. Besides, this paper finds that the boards of larger firms are more independent than those of smaller firms. Moreover, the boards of the levered firms tend to be more independent because debt holders exert intense pressure to implement stringent monitoring mechanisms. Another vital finding of this paper is that board independence is lower for older firms. A plausible reason might be that matured firms devise stricter and more efficient internal control mechanisms that offset the need for independent board. This study also reports findings that are contrary to the traditional belief. For example, we find that longer tenure and prior relationship of independent directors with the firm increases board independence by developing a sense of dedicative attachment of the independent directors with the firm. On the other hand, larger and active boards tend to be an impediment for board independence. Finally, we find no association of board independence with growth opportunity, profitability, and board gender diversity. The findings will have policy implications in designing the role of board independence.

the goal of strengthening the effectiveness of corporate governance mechanisms in both listed and non-listed firms.
In line with such a movement, the incorporation of the concept of Independent Directors (ID) in the board to safeguard the broader interest of the stakeholders works as an important tool to ensure sound corporate governance in firms. Thus, regulators and other stakeholders have, from time to time, expressed their concerns regarding the independence and expertise of the independent directors serving in the board.
Independent directors have become a vital instrument in representing the interests of all stakeholders ranging from the tax authority to the minority shareholders. In fact, it is often said that independent directors are the first line of defense for minority shareholders (Juan Ma and Tarun Khanna, 2013). Moreover, being in a position to act without the undue influence of the management, independent directors provide third-party advice and oversight in ensuring the proper governance of the firms. It has been found that shareholders react favorably to the appointment of outside directors (Rosenstein and Wyatt, 1990). Consistent with this, by studying the stock market reaction to the sudden death of an outside director, Nguyen and Nielsen (2010) found that shareholders react negatively to the loss of outside directors. Although the impact of the presence of independent directors on performance is less clear, Knyazeva, Knyazeva, and Masulis (2013) found that independent directors are positively associated with firm value (market to book ratio) and operating performance (ROA). Again, Duchin, Matsusaka, and Ozbas (2010) observed that if the cost of acquiring information is low, performance increases when outside directors are added to the board (Tobin's Q and ROA). In addition, independent directors work as agents of negotiating conflicts of interests between management and the board due to their independence from both. Besides mitigating conflict, independent directors have other advantages as well. They supposedly have more general knowledge compared with an executive director who has more firm-specific knowledge. Thus, the independent directors very often possess a clear view of the environment external to the business and this enables him to identify and focus on various external issues bearing significance to the firm in the long run. This way, independent directors are capable of standing back from each issue and contemplating them in light of their expertise without bias.
The effectiveness of independent directors is limited by the extent to which they are truly independent. Historically not all independent directors have been truly independent. Hwang and Kim (2009) tested whether NYSE accurately measures independence and found that 87% are NYSE independent, but only 62% are both NYSE and socially independent. Independent directors need true independence from the board and management in order to ensure a harmless tie between the controlling and minority interest, sound governance and promote a culture of transparency as well as accountability. In Times of Malta, Christmas (2014) accurately emphasizes the independence of an independent director from managers and other service providers meaning the independent director and his close relatives must not be the employees or even owners of the service providers. He mentions that the reason of such a drive towards independence is the avoidance of conflicts of interests. Christmas (2014) adds that the reason behind an independent director's ability to add value to the firm is his capability to raise questions and criticize the top management executives should any conflict arises and therefore an independent director is effective only if he has such strength and character that enable him to do so. This implies that independence is the underlying factor that makes independent directors an effective tool to better governance.
All these evidences provide a strong ground to assess the factors affecting the board independence in Bangladesh. In doing so, we choose the banking sector as the area of our study because this is a vastly important sector having massive impact on the economy of Bangladesh. Moreover, the banks in Bangladesh are subject to tight regulatory control because of its involvement with broader public interest. However, this sector is currently facing severe liquidity crisis due to escalating amount of non-performing loan and resulting bad debt. This crisis is often attributed to the lack of good corporate governance practiced by the board. Allegations are that the board itself has overridden internal credit approval process to sanction loan to politically powerful but likely to default loan applicant. The present dismal state of corporate governance in the banking sector underpins the significance of a strong independent board which can serve as a deterrent to the sponsors' family influence on the board. Despite such potential, no extensive research has been conducted to date to investigate board independence in this highly critical sector. The demand for board independence generally originates from regulators, management, shareholders, and creditors of a firm. The limited work concerning board independence in Bangladesh motivates our study and strengthens the empirical significance of the determinants of board independence to different stakeholders among which the primary are shareholders, managers and regulatory authorities such as Bangladesh Bank or Bangladesh Securities and Exchange Commission.
In this paper, we investigate the determinants of board independence in the banking sector of Bangladesh by using multivariate regression analysis for 30 scheduled banks that provide the required information for the period 2006 to 2016. We exclude specialized banks in this analysis. The main objective of this study is to find out the relevant factors that influence board independence in the banking sector of Bangladesh. The results we report are expected to greatly benefit the shareholders, and regulators. Shareholders can apply the outcome of this paper in deciding what factors to consider before approving the appointment of an independent director in the AGM while regulators can use the findings to assess the effectiveness of existing regulation relating to the improvement of governance through appointing independent directors.
We organize the rest of this paper as follows: section two provides a brief review of prior studies on independent directors which are very inadequate till date. Section three presents the overview of independent directors, their role and relevant attributes. The development of hypotheses and rationale for their development are discussed in section four. Section five deals with various aspects of research design issues. The results of this study are presented, analyzed and discussed in section six and finally, section seven concludes the paper with some suggestions for future research opportunities.

Literature Review
There is a huge research gap till date concerning the empirical testing of the determinants of board independence in both local and international arena. The inadequacy of prior works in this field can be attributed to the difficulty of constructing a model that truly captures such a subjective concept as independence. Yet, the same inadequacy inspires me to take this as my topic of study. In this section we provide a succinct summary of prior studies that investigate the independence of the board.
There is a vast literature to improve the corporate governance of firms done by both academics and practitioners. Many such studies have either directly suggested or indirectly inferred that board independence or the existence of independent directors are associated with higher quality of governance while others have mixed findings often suggesting that board independence is neither a necessary nor a sufficient condition for strengthening firm governance practices. Weisbach (1988) found that in case of poor performance, a board dominated by outside independent directors is more likely to fire the top management. Moreover, Brickley, Coles, and Terry (1994) found evidence in favor of their hypothesis that outside directors are more likely to vote in the interests of the shareholders when the firm considers adopting poison pill strategy to avoid unfavorable acquisition. Byrd and Hickman (1992) showed that, a board dominated by outside directors realized much higher abnormal returns during announcement periods of tender offers. Besides, Rosenstein and Wyatt (1990) tested the effect of an event altering a specific governance mechanism. They found that the market reacted quite favorably to the inclusion and addition of outside directors to the board. Again, Gupta and Lee (2004) showed that payments for director premium and officer liability insurance are smaller in firms that have better governance. A general conclusion that follows from the results of these studies is that a board dominated by outside directors mitigate agency costs in various ways. However, two issues need to be mentioned here. First, there is a difference between outside directors and independent directors even though the two terms are often used interchangeably. An independent director is an outside director without any material interests in the business of the firm. This is where the question of determining true independence of independent directors comes. As D Clarke (2006) clearly puts that the concept of nonmanagement directors (NMDs) originate with the objective of assisting stockholders in solving agency problems and if NMDs are to effectively overlook management, they must be independent of the management. Second, contrary findings that no statistically significant relationship exists between the independence of board and firm performance are also in vogue. For example, Bhagat and Black (2001) didn't find any significant relationship between firm performance and independent directors' presence in the board.
Actually, there are two conflicting schools of thoughts regarding the effect of board composition on firm performance-agency theory and stewardship theory (Dalton et.al., 1998;Ramdani and Witteloostuijn, 2010). These theories provide contradictory predictions about independence of board and firm performance. The approach of agency theory can be used to predict that proportion of independent directors in the board is positively associated with firm performance as a way to mitigate agency costs. On the other hand, stewardship theory hypothetically predicts the same variables to be negatively associated as it assumes that executive directors are more informed about the current and future prospects of the firm than do independent directors and have a greater role to play in the operations of the firm. The findings of various researches regarding the representation of independent directors on the board and firm performance are, therefore, mixed. There are a good number of studies that have empirically examined the relationship between board independence (proportion of IDs in the board of a firm) and firm performance measures such as return on assets (ROA) or return on equity (ROE) and discovered a positive relationship (Hutchinson and Gul, 2002;Bonn, 2004). Knyazeva, Knyazeva, and Masulis (2013) also examined the relation between independent directors and firm performance using a sample of 900 small and mid-sized firms in S&P 1500 index for the period 1996 to 2006. They reported that independent directors positively affect both profitability and operating performance of the firms. On the contrary, others couldn't confirm a robust conclusion about any such statistically significant relationships (Christensen et.al., 2010;Azim, 2012).Some studies have tested the relationship between independence of the board and market-to-book ratios and have been able to display some convincing evidence of a positive and statistically significant relationship (Henry, 2008) while some others have provided a negative and statistically significant relationship (Kiel and Nicholson, 2003). Meanwhile, some researchers cannot confirm that any statistically significant relationship exists between board independence and firm performance (Azim, 2012;Matolcsy et.al., 2004;Bonn, 2004;Pham, Suchard and Zein, 2013;Christensen et.al., 2010). While the balancing evidence suggests that board independence or in other terms, the higher extent of presence of independent directors on the board, improves firm performance, the findings of these studies are obviously mixed.
While it comes about the perceptions of shareholders about independent directors, a large number of studies present findings that favor the independent directors. Examining 1,251 outside directors announcements for the period 1981 to 1985, Rosenstein and Wyatt (1990) found statistically significant positive reaction of stock price in the two-day trading windows around the announcements. They concluded that shareholders had favorable reaction towards the appointment of independent directors and the results were strongest for the small firms. Another study with somewhat similar conclusion put in a different way was undertaken by Nguyen and Nielsen (2010) who examined how the stock market reacts to the sudden death of an independent director. They took into consideration a sample of 229 deaths from 1994 to 2007 and reported that shareholder tend to react negatively to unanticipated loss of an independent director. They also found that the negative reaction was more intense when the directors served key roles such as the chair of the board or head of audit committee or when the representation of independent directors on the board was low and the reaction was comparatively less negative when the directors had already served for longer period in the board. Another striking role played by outside independent directors in enhancing firm value, as reported by Cotter, Shivdasani, and Zenner (1997), is their ability to negotiate higher takeover premiums. That is, independent directors play a role which is both statistically and economically significant in enhancing firm value during tender offers. Moreover, Byrd and Hickman (1992) report empirical findings that support the hypothesis that a more rational merger activity is encouraged by independent directors. However, both Boyd (1994) and Finkelstein and Hambrick (1989) have documented evidence that independent directors don't lead to tighter or lower CEO compensation.
After critically examining the pros and cons of independent directors, an important question to ask is how important it is to determine whether independent directors are truly independent. Realizing the importance of independent directors being free form any managerial influence, D Clarke (2006) nicely puts that directors expected to perform their designated functions cannot actually do so as long as they are not systematically independent of management. As mentioned before, Chiristian (2018) puts in Times of Malta that it should be definite that an independent director be free not only form managers and executives but also from the service providers, that is he and his close relatives should not have any sort of tie with the employees or even the shareholders of the firm so as to avoid conflict of interests. Recognizing the importance of true independence of independent director, it is worthwhile to have a check on the factors that drive the extent of board independence.
To date, both in local and global arena, there has not been much extensive empirical research about the factors that directly or indirectly determine or even influence the level of board independence. Therefore, the study is motivated to address the research gap regarding the empirical testing of the determinants of board independence in the context of the banking sector of Bangladesh.

Definition of independent director
An independent director is a non-executive member of the board of directors of an incorporated company, who, apart from receiving director's remuneration, does not have any material interest in the company whether pecuniary or not. According to D Clarke (2006), an independent director is "one who has no need or inclination to stay in the good graces of management, and who will be able to speak out, inside and outside the boardroom, in the face of management's misdeeds in order to protect the interests of shareholders." The definition of independent director is entangled by the term "independent" as it is subject to interpretation rather than any objectively determinable criteria. Still as agreed by most regulators and academics, the term independent here means the absence of any pecuniary or material relationship with the firm in consideration and its insiders such as management and sponsor directors. The concept of independent director thus evolves to deal primarily with the agency problem. As a result, it is expected that independent director should not have any material relationship with firms' insiders whether financial or not that impedes them from undertaking the role they are supposed to play in the board-mitigating agency problems. But the presence of approved sitting charge or fees doesn't necessarily preclude the independence of independent directors. Thus in India, clause 49 of the listing agreements defined independent directors as "directors who apart from receiving director's remuneration, do not have any other material pecuniary relationship or transactions with the company, its promoters, its management or its subsidiaries, which in judgment of the board may affect independence of judgment of the directors". Similarly emphasizing the importance of independence of independent directors, NYSE states that "no director qualifies as 'independent' unless the board of directors affirmatively determines that the director has 'no material relationship' with the listed company, either directly or indirectly as a partner, shareholder or officer of an organization that has a relationship with the company." In line with the definitions above, regulators in Bangladesh also refer to independent directors as directors who are 'independent of material monetary and other relationships from management, sponsors, sponsor directors, and controlling shareholders".
Another important part of the definition of independent directors concerns their expertise. An independent director must possess relevant business acumen so as to enable him to act as a bridge between the management and different stakeholders of the firm ranging from minority shareholders to regulators. In fact, most legislation explicitly mentions that an independent director must be a person of high integrity and sound business knowledge. Moreover, some countries have vested the role of independent directors to a specific class of professionals such as business leaders, law, economics, finance, and accounting professionals and academics of the same areas of expertise, government officials serving higher ranks etc. Closely related with knowledge is the experience of independent directors and it has also been remarked as a distinguishing quality of them. Thus, taken together independent directors must combine integrity, independence, and relevant business knowledge along with experience.

Independent directors in Bangladesh: A regulatory framework
Although the origin of the concept of independent directors can be traced back to the 1950s even before legislation mandated it, this concept of independent director in Bangladesh is only about a decade old.  2. Partners or executives of the concerned company's statutory audit firm had not been considered disqualified to be the ID of the company.
3. Being a defaulter was not mentioned to be a reason of disqualification for the post of an ID.
1. It has been clearly mentioned that spouse, son, daughter, father, mother, brother, sister, son-in-law, and daughterin-law shall be considered as family members. 2. A current partner or an executive or one who had been the partner or an executive during the preceding 3 (three) years of the concerned company's statutory audit firm is disqualified to be the ID of the company. 3. One being convicted by a court of competent jurisdiction as a defaulter in payment of any loan to a bank or a Non-Bank Financial Institution (NBFI) is disqualified for the post of an ID.

Tenure
Nothing was mentioned about the tenure of an ID.
The tenure of office of an independent director shall be for a period of 3 (three) years, which may be extended for 1 (one) term only.

Minimum Number of ID in the Board
At least one tenth (1/10) of the total number of the company's board of directors, subject to a minimum of one shall be independent directors.
At least one fifth (1/5) of the total number of directors in the company's board shall be independent directors.

Appointment of ID
The independent director(s) should be appointed by the elected directors.
The independent director(s) shall be appointed by the board of directors and approved by the shareholders in the Annual General Meeting (AGM).

Vacancy of the post of the ID
Nothing was mentioned about the maximum days for which the post of an ID can remain vacant.
The post of independent director(s) cannot remain vacant for more than 90 (ninety) days.

Maximum number of companies an ID can serve
Nothing was mentioned about the maximum number of companies an ID can serve.
One shall not be an independent director in more than 3 (three) listed companies.

Chairman of the audit committee
May or may not be an ID.
Must be an ID.
By analyzing the two versions of CG guidelines, it has been found that the latest issue of 2012 takes the matter of independent directors as a mechanism of ensuring sound corporate governance quite seriously unlike the past one and has a stronger emphasis on the role and qualifications of independent directors. Yet, it is not all conclusive as it leaves some confusion such as-• Whether the size of the board is inclusive or exclusive of the IDs while calculating the minimum number of IDs in the board. • Moreover it has not specified the perquisites and other financial benefits an ID might expect from the company he serves. • It has not mentioned what is the extent of benefits received by the ID that will not pose a threat to his independence and objectivity. • The extent to which an ID will be liable if he fails to discharge his responsibilities properly.
• Who are the primary stakeholders (i.e. Shareholders or public interest) the IDs should strive to secure should a conflict of interest occurs etc. Taken together, it can be said that despite meaningful contribution till date, the introduction of IDs as a solution to corporate governance problems needs to be improved in order to make it viable and BSEC has a significant role to play in this regard.

Hypotheses Development
Agency theory implies that information asymmetry would exist in a situation where the ownership of a business is separate from its management. Moral hazard and adverse selection are two major problems that might arise because of such a separation of management from ownership. In such a situation, one would anticipate a conflict of interests between the principals (owners) and agents (management) and the resulting loss is the agency costs. Independent directors can be a useful instrument in mitigating such agency costs by acting as a watchdog in the board with a view to balancing the conflicting interests of the competing groups such as owners and management. But, as one would easily predict, their role will be effective only if they, themselves, are not caught up by any material interests in the business under consideration. Here is where the significance of independent directors' true independence comes into action. Only a truly independent director can add value to the organization as long as it is a matter of displaying a balance of conflicting interests by reducing information asymmetry.
In this section, we develop a set of testable hypotheses for each tentative variable that is likely to have an influence on the independence of the board. We identified some variables from extant literature and suggested some new variables based on our analysis of prior theoretical work on board independence.

Knowledge and Power
It is usually expected that a positive association exists between the knowledge and skills an individual possesses with his independence in actions. World's leading credit rating organization Moody's in its report titled 'Criteria for Assessing Director Independence' mentions that "We expect that a high level of board independence is matched with sufficiently knowledgeable and engaged directors who can hold management to account." Fogel, Ma, and Morck (2014) examined whether "powerful" directors behave independently in the board. They defined powerful directors to be those with large professional networks. After examining a sample of 19,223 unique directors from 1998-2010, they found that Powerful directors were associated with more valuable mergerand-acquisition activities and stricter oversight of CEO performance. In addition, powerful directors were associated with less earnings management. They also found that shareholders tend to react negatively to the sudden death of any powerful directors. Taken together, powerful directors exhibit greater independence. On the other hand, Moody's put forth in 'Criteria for Assessing Director Independence' that sometimes directors with greater sector experience appear to have potential conflicts and this may impair their independence. However, knowledge of industry or expertise and independence are addressed as factors that are not necessarily mutually exclusive (Alex Frino, 2016).

Firm size
It is often argued that larger firms generally adopt stricter monitoring mechanisms, implying more demand on governance tools like audit committee and board independence (Klein 2002). Fama and Jensen (1983) show that larger and more hierarchical firms results from process that are either larger or more complex or both and as a result, larger firms need larger board to provide CEO appropriate advice and necessary access to information as well as resources. This finding suggests that firms with a relatively complex form of operations require a bigger and more independent board. Lehn et al. (2004) also suggest that as the scale and complexity of the firm increase, board independence has to increase as well in order to cater to the greater information needs. Also, it is argued by Anderson et al. (2000) and Coles et al. (2005) that more diversified firms tend to have more IDs in order to carry out the monitoring role expected of them. Considering all these findings, we hypothesize a positive relationship between firm size and board independence.

Financial leverage
A firm with leverage tends to have greater pressure from the creditors to put more transparent governance mechanisms in place so as to safeguard the interest of the creditors and lenders. In Japanese organizations, there is often representation in the board from the creditors while other countries often require greater outside representation (i.e. greater independence) in the board. This is an implication of the debt covenant hypothesis that managers in levered firms with debt covenants tend to manipulate earnings to either delay or even avoid potential violation of debt contracts. For example, Dechow et al. (1996) report that managers have shown tendency to overstate reported earnings in the year before violations of debt covenants. Thus creditors demand greater board independence in a firm with financial leverage. Another view consistent with this hypothesis is that a more complex business is usually a levered one. And as we have mentioned in the previous hypothesis, more complex business requires more independent directors. Thus the third hypothesis becomes: H3: Board independence is positively related to financial leverage.

Growth opportunities
Firms that have experienced rapid growth may, at one point, outgrow the infrastructure as well as the internal control system they have. According to Klein (2002) firms that are growing are associated with more uncertainties and greater complexities. He reports that such growing firms generally rely less on independent directors and more on insider directors. Consistent with his findings, Klein (2002) expects lower demand on board independence from both management and shareholders of firms with high growth opportunities. Linck et al. (2008) also show that firms with higher growth opportunities tend to have less independence in the board. This study uses market to book ratio as a proxy for growth. Consequently the hypothesis becomes: H4: Board independence is negatively related to firm's growth opportunities.

Profitability
This study recognizes profitability as having an important link with board independence. Adhikary and Mitra (2016) hypothesized that when there is a possibility of private gain for insiders, firms should employ more independent directors in audit committee. From that hypothesis, it is reasonable to conclude that insiders have incentives to manipulate firm profitability to serve self-interests. As a result, greater internal control and stricter governance mechanism should be put in place to curb such self-interested behavior and obviously this brings back the role of independent directors as a control mechanism. So, our next hypothesis is as follows: H5: Board independence is positively related to the profitability of a firm.

Firm age
Linck et al. (2008) demonstrate that when insiders have greater opportunity to derive much of private benefit, firms employ greater number of outside independent directors. Generally older firms develop stricter and more efficient internal mechanisms to bar insiders from extracting such private benefits in comparison to younger firms with low initial investments in developing sophisticated monitoring mechanisms. Thus, we expect older firms to have much less demand for independent directors as a tool to enhance governance and the related hypothesis is: H6: Board independence is negatively related to firm age.

Board size
Some authors reported that a reasonably larger board combines varying degree of knowledge and experience of directors and add more value to the firm making a larger board more desirable. In addition, both Collier (1993) and Beasley and Salterio (2001) reported that keeping pace with the increase in board size, the ability of a firm to employ more independent directors increase. Moreover, Adhikary and Mitra (2016) found that firms with larger board had more independent audit committee. On the other hand, it is also argued that monitoring role is stronger in a smaller board and it is less probable that the board be captured by the CEO when it is small. Besides, a general reasoning follows from the field of Origination Behavior (OB) that people tend to be less independent and show more conformity while surrounded by a large number of people having opposite ideas. Following these opposite ideas, we hypothesize that board independence has association with board size without predicting about the direction of association. H7: Board independence is related to board size.

Board meetings
Board meeting is a very important part of the overall functioning of many governance mechanisms. In fact, Van den Berghe & Levrau (2003) show that directors themselves put tremendous importance on the quality of board meetings followed only by board composition while judging the elements of good BoD. More frequency in board meetings is a sign that issues of corporate governance are being reviewed more frequently. This also keeps the role of independent directors' a rolling stone that has little chance to gather moss. Thus it might be a positive sign for the governance and work well to the advantage of an ID to demonstrate his independent views regarding different issues. Contrary to the notion imbibed above, more board meetings mean more sitting charge to IDs and the number of board meetings might thus latently be used by the insiders to transfer significant pecuniary benefits to the independent directors even within the legal bounds. Thus the number of board meetings might be used by insiders as a loophole to financially capture the independence of IDs. Considering this view, we hypothesize the following relationship to exist between the number of board meetings and board independence without any specific direction. Bohren and Staubo (2015) reported that a 40% mandatory female quota in the board was associated with an increase in board independence. Empirically they found that after such a quota, average fraction of IDs jumped from 67% from 46%. They conveyed that female directors are, indeed, much more often independent directors in comparison to male directors. Thus, I hypothesize that the fraction of female directors in the board is positively related to board independence.

Tenure
As IDs tend to have longer tenure in the board, the information gap generally subsides and IDs become more acquainted with the operations and governance of the firm by getting more exposure to inside information. Thus over time, IDs can pile up their firm specific knowledge and information (Vance, 1983) which, in turn, enhances their commitment and dedication to the firm as well as improves ID competence (Buchanan, 1974). Thus longer tenure may be expected to have favorably affected independence. Contrary to this view, (Vafeas, 2003) argues that the longer an ID serves in a board, the more personal connections he develops with management and their relationship becomes cozy. This perspective holds the view that the longer the tenure, the more the possibility of independence being compromised and the easier it is for management to capture the ID. Consequently, regulators, advocate of governance and many large institutions hold a skeptical view of IDs serving longer tenure. Consistent with this 'Compromise view', CG guidelines issued by SEC in 2012 calls for rotation of IDs after only one renewal of the 3 year serving period. So, a more rational hypothesis is the possibility of a positive or negative association between board independence and the tenure of independent director. H10: Board independence is associated with tenure of the independent directors.

Relationship with the firm
Whether or not an ID is already related with the firm is an important factor that tentatively is expected to affect the degree of actual independence an ID can subsequently demonstrate. Moody's considers that prior connection with firm or even with other directors or management of the firm is a sign of a potential compromise with freedom for independent directors. But this doesn't necessarily always constitute a hindrance to independence. Rather, it is argued that some sort of relationship with the firm, as long as it is not extensive enough to capture the independence of the directors, is good for the improvement of monitoring activity on the part of an ID. This is backed by the fact that firm specific knowledge is important for an ID to ensure that the loopholes management uses to maximize self-interest at the cost of owners are better mended. And such expertise often requires a mild connection of independent directors with the firm. Moreover, if an ID is a total alien to the firm, chances are there that he will find it monotonous to execute his duties with genuine and diligent interest. As a consequence, different acts and governance guidelines often allow a small stake of the ID in the firm (i.e. 2012 SEC CG notification allows an ID to hold less than 1% of the paid up capital of the firm). Thus, we don't ascertain the sign but instead hypothesize the following with respect to ID prior relation with the firm in general: H11: Board independence is related to prior relationship of independent directors with firm.

Sample Description and Data Collection
In order to test the hypotheses, we have used a sample of 30 (thirty) listed banks from the banking sector of Bangladesh that constitutes with a total 56 banks classified into 6 state-owned banks, 2 specialized banks, 39 private commercial banks and 9 foreign commercial banks. Even though our sample constitutes 53.57 % of the total number of banks (excluding Bangladesh Bank), it covers all the banks that are listed in Dhaka Stock Exchange (DSE).
We resort to secondary source while collecting the data required for the study. All data come from the published annual reports of the banks which are publicly available. In order to collect the data we have extensively analyzed the annual reports of the banks from the period 2006 (when for the first time SEC issued the corporate governance guidelines) to 2016 (which is the latest year for which annual reports are available online for all the banks).
As mentioned earlier, several factors motivate us to choose the banking sector as the field of analysis in this study. Among them, the most important factor was the stringent regulation that the banking sector is subject to. Banks in Bangladesh are under extensive regulatory pressure and strict monitoring as they, unlike most other business, employ greater amount of public wealth and trust. In order to retain the confidence and faith of the greater mass of the country, the regulatory agencies and the government must find some path to ensure superior governance, transparency, and accountability in this sector. Thus we expect banks to have a more sound set of governance mechanisms than do other sectors. With this expectation, the results of this study must have significant bearing for several parties including the government, regulators (Bangladesh Bank, BSEC etc.), management, as well as the general public who engage in day to day transactions with and take services of the banks. Finally, the results of the analysis done with the most strictly controlled banking sector might be a benchmark for deducing how well or bad the other sectors might be doing in appointing independent directors.

Variable Definition
Table II, summarizes the dependent and all the independent variables of the model we develop and also includes their definition as well as the expected nature of relationship with board independence:

Regression model and estimation method
Using all the identified variables we develop and estimate an econometric model to determine the significant factors likely to affect the board independence in the banking sector of Bangladesh. The regression model is as follow: + % + ) + . + 1 + 4 + 8 + 9 + ; + = + %> + %% + Before estimating the multivariate regression model, we examine the "multicollinearity effect" using Pearson's correlation matrix and VIF. Both methods ensure that the model has not been unnecessarily extended by considering variables that capture homogenous concepts. Since data structure of this study corresponds to panel data covering thirty listed banks over eleven years period, we run both pooled OLS regression and robust regression with observations being clustered into firm level and year level (known as two-way clustering) to take into consideration both the cross-sectional and the time series nature of the data and to control for heteroskedasticity. The two-way clustered method adjusts the standard error of the pooled OLS method to reflect cross-sectional and time-series relation of the data. However, the robust clustered regression doesn't differ significantly from the pooled OLS multivariate regression except for two variables. The results of estimation are described in the next section.

Descriptive Statistics
Table III summarizes the descriptive statistics for the variables used in the regression model. According to the table, on an average 62% of the independent directors serving in the board of the banking companies in Bangladesh have the educational qualifications required by BSEC guidelines for corporate governance. The average total asset of the firms is found to be 99,707.88 million BDT and the mean leverage of the firms is 93%, a very high one and is justified because of the nature of the industry. On average, a share is traded at 2.5 times high compared to its book value and return on asset is 1.3 percent. Around 18% of the IDs are somehow related to the firms they are serving indicating a probable and harmful nexus between management and ID under the compromise view of CG. But viewed from another angle, this relationship can improve the IDs knowledge about the firm and thus they have more scope to come up with value addition services to the firm by the incorporation of their firm specific knowledge and experience as well as a better sense of dedication and attachment to the welfare of the firm. In regard to board meetings, the average number is around 18 which is a decent number. We also find that average presence of female member in the board is 10% which has run to a highest of 75%. Another important finding of this study is that 8% of the IDs have either reached or exceeded the prescribed tenure mentioned by SEC that is six years at a stretch. The rest 92% of the IDs have not been in appointment for more than 2 consecutive periods. Finally, the degree of board independence as measured by their proportion in the board is found to be 10% on average which runs as high as 43% while there are boards with no ID and apparently no board independence. This phenomenon actually was prevalent in the banking sector during the period of 2006 to 2008 in most of the banks due to the lack of a stringent rule requiring independence in the board. But the situation dramatically improved after the 2012 revised notification on CG issued by BSEC. Still, Bangladesh is lagging behind in ensuring the proper representation of independence from the ID mechanism as long as the suggestion of Raghunandan and Rama (2007) is concerend that calls for at least 50% of the board to be represented by non-executive directors to ensure sound governance of the firm.

Correlation Analysis
Before moving into the main analysis, we check the Pearson correlation coefficient among the variables. The following table (Table IV) reports the presence of no independent variables in the model that has significant relationship with any other variables here. This univariate correlation suggests that board independence is positively and significantly correlated with knowledge of independent directors, firm size, firm age, gender diversity, tenure of independent directors and their relationship with the firm while board independence has significant negative relationship with market to book ratio and board size. However, no significant relation is found for financial leverage, operating performance and number of board meetings.

Result of Multivariate Analysis
We estimate the regression model using both pooled OLS method and two-way clustered method. While the results do not differ significantly, we report only the results of clustered regression analysis since this method corrects the standard errors for possible relation among variables across firms and over time commonly detected in panel data setting, and produces more robust and conservative standard errors. The results are presented in table V. We find the F-statistic of the model developed in this study to be highly significant (F=44.94, P < 0.0001) indicating overall significance of the model. Both R 2 and adjusted R 2 have reasonably high values respectively being 0.63 and 0.59. This is an indication that around 59% of the total variation in board independence can be explained by the set of independent variables used in the model. Moreover, the low value of Root MSE indicates greater accuracy of the model. Adj. R-square 0.59 ***, **, * represents significance level at 1, 5 and 10 percent respectively.
The results indicate that board independence has significant positive relationship with ID knowledge, firm size, leverage, tenure and ID relationship with the firm. These results are in line with the hypotheses that -• A board is likely to be more independent when independent directors have relevant skill and knowledge as required by prevailing corporate governance regulation. [Fogel, Ma, and Morck (2014)] • Bigger firms employ more independent directors in order to ensure that a stringent monitoring role is carried out by them (Klein, 2002) • Greater leverage turns out to be a reason for demanding stringent monitoring by creditors which in turn requires more indepepndence in the board.
• Longer tenure enhances the firm specific knowledge of the independent directors (Vance, 1983) and this in turn improves the dedication and commitment of the ID to the firm as well as improves competence (Buchanan, 1974). • The positive association between board independence and the existence of any relationship between the ID and the firm might be explained by fact that some sort of attachment with the firm is necessary for the part of the ID in order for them to have a greater degree of care for the sound governance of the firm. If IDs do not at all have any sort of relationship or attachment with the firm they serve in, it would be a dull responsibility for them to carry out and they would apparently lose interest in executing their duty properly.
On the other hand, this study finds significant negative relationship of board independence with firm age, board size and board meetings. Over times, older firms tend to develop more efficient internal control and monitoring systems that are apt in ensuring sound governance. The capability of older firms to demonstrate much maturity in controlling management's ability to extract private benefit (i.e. in the form of using up free cash flow for office amenities) from the firms leads to a lower demand on the board independence as a mainstream governance mechanism. Thus, matured firms develop stronger substitutes to ID and the negative association of firm age with board independence seems a fairly justified one. Similarly, the negative association of board size and board activity with board independence suggests that these board characteristics seem to work against board independence. One possible explanation might be that a firm is less likely to appoint new independent directors in response to increase in the number of board members or board meetings.
However, this study doesn't find any significant relationship of board independence with growth opportunities, profitability and gender diversification (presence of female members) in the board. The coefficients on these three variables are not statistically significant at the conventional level. Thus, this study indicates that growth opportunities (as measured by market to book ratio), firm profitability (as measured by ROA) and gender diversity (as measured by the proportion of female directors on the board) do not materially affect board independence in the banking sector of Bangladesh.

Conclusion
This study has attempted to identify the determinants of board independence measured by proportion of independent directors on the board in the banking sector of Bangladesh. This examination has applied a panel data analysis for DSE listed 30 scheduled banks that deliver the information necessary for analysis from 2006 to 2016. The results of this study show that boards tend to be more independent when they have independent directors with relevant skill and educational background. Firm specific factors that positively influence the independence of the board are larger size of firms and a relatively higher financial leverage. This study reports that board structure plays an important role in determining the level of board independence. Another vital finding conveys that the greater the board size, the lower is its independence. Contrary to the tentative expectation, longer tenure and relationship of independent directors with the firm have been found to enhance independence instead of decaying it. But it would be wise to be cautious in interpreting this result because too much increase in these two variables might create hindrance to independence. Finally, this study doesn't find any statistically significant relationship of board independence with profitability, growth opportunities of the firm and gender diversification in the board. Finally this paper confines its focus to the determinants of board independence without attempting to explain or predict the consequences of such. The concept of independent director has attained extensive popularity and importance in ensuring proper functioning of the board and other internal governance mechanisms in a firm. Thus this study might be helpful in undertaking future research such as determining various regulatory and performance implications of this concept, testing the attributes of independent directors with firm performance, examining the usefulness of having more independent boards and so on.