1
BACHELOR’S THESIS
AUDIT COMMITTEE CHARACTERISTICS AND BANK PERFORMANCE: EVIDENCE FROM THE UNITED STATES DURING THE GLOBAL FINANCIAL CRISIS
Written by Fatimah Ahmad Munawar
10864644 Supervised by
2
Statement of Originality
This document is written by Student [Fatimah Ahmad Munawar] who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
This study aims to investigate the impact of various audit committee characteristics on US bank performance during the financial crisis of 2008. After recent accounting scandals and the global financial crisis, the role of corporate governance attributes related to the audit committee has received heightened scrutiny. The empirical approach includes panel data fixed effect regressions for 22 commercial banks listed on the S&P 500 index over the years 2005-2010. The agency theory is applied in this study to understand and predict the performance implications of audit committee characteristics in the banking industry. The main findings of this study suggest that audit committee characteristics have an impact on the financial performance of banks under stable economic
conditions but found limited evidence on its impact on bank performance during the financial crisis. The results suggest that there is a significant non-linear relationship between the audit committee size and tenure, and bank financial performance. Additionally, the results of the study present a significant negative relationship between the financial expertise of the audit committee and bank performance in both crisis and non-crisis periods. However, no relationship was found between the audit committee size, activity and tenure, and performance during the financial crisis. The study offers contributions to the growing body of research on the impact of audit committee governance attributes on bank
financial performance during the global financial crisis. Moreover, the present study shows that the impact of governance mechanisms regarding the audit committee on bank performance are different in crisis and non-crisis periods.
4
Table of Contents
1 Introduction………...5
2 Literature Review……….6
2.1 Theoretical Framework………...6
2.1.1 Corporate Governance and the Financial Crisis………6
2.1.2 Role of Audit Committee in the United States………...6
2.1.3 Agency Theory………..8
2.1.4 Firm Financial Performance……….10
2.2 Empirical Findings……….10
2.3.1 Audit Committee Size and Firm Performance……….10
2.3.2 Audit Committee Expertise and Firm Performance……….11
2.3.3 Audit Committee Activity and Firm Performance………...12
2.3.4 Audit Committee Member Tenure and Firm Performance………..12
3 Methodology and Data………...13
3.1 Methodology………..13 3.1.1 Regression Model……….13 3.2 Data………14 3.2.1 Control Variables………...15 4 Results………..15 4.1 Descriptive Statistics………...15 4.2 Correlation Analysis………..15
4.3 Regression Results and Implications………...16
4.4 Robustness Check………..22
5 Conclusion and Limitation………23
5
1 Introduction
Brunnermeier (2009) argues that the firms’ financing policies and risk management had a significant impact on the degree to which the financial crisis had impacted the firms. Because the financing policies and risk management of firms are fundamentally the results of cost-benefit trade-offs between corporate boards and shareholders (Kashyap et al., 2008), an important implication from this is that corporate governance had an impact on the financial performance of banks during the global financial crisis. An internal corporate governance mechanism designed to oversee a firm’s financial decision-making and risk management processes is the audit committee. Thus, the aim of this study is to answer the following: To what extent do audit committee characteristics such as size, expertise, activity and tenure impact the performance of American commercial banks during the Global Financial Crisis?
Four audit committee characteristics were identified namely audit committee size, expertise, tenure and activity to study their impact on firm financial performance represented by the return on assets (ROA) and Tobin’s Q. The research approach adopted in this study includes fixed effect panel regression for the period 2005 to 2010. The period of the global financial crisis is defined from 2007 to 2009. The sample used in the study is based on 22 commercial banks listed on the S&P 500 during the 2008 financial crisis. The findings from this research provide evidence that there is a significant non-linear relationship between the size and tenure of the audit committee, and the ROA and Tobin’s Q of banks sample under stable market conditions. The proportion of audit members with experience as a certified public accountant, controller or chief accounting officer over the total number of audit committee members showed a significant negative relationship with bank performance in both crisis and non-crisis periods. However, the results of this study do not find evidence that the size, activity intensity and tenure of the audit committee had an influence on the performance of banks during the 2008 financial crisis.
The remainder of this paper is structured as follows. Section 2 discusses the theoretical framework and reviews prior literature on audit committee characteristics and firm performance. Section 3 describes the research design which includes the empirical model, data collection and variable measures used to estimate the relationship between audit committee characteristics and firm performance. Section 4 presents the results and implications of the data analyses. Finally, Section 5 offers a conclusion evaluating the impact of audit committee characteristics on U.S. commercial banks performance during the global financial crisis.
6
2 Literature Review
2.1 Theoretical Framework2.1.1 Corporate Governance and the Financial Crisis
The role of corporate governance in the banking industry is considered as one of the decisive factors explaining the recent global financial crisis. Anecdotal evidence suggests that during periods of adverse economic shocks, the corporate governance structure of banks were ineffective at monitoring the quality of risk management processes which consequently result in an extreme vulnerability of the financial system. Thus, failing corporate governance systems are often related to poor performance of firms. As most of the prior research investigated the role of corporate governance under normal market conditions, this signifies the need for more empirical research on the influence of corporate governance structure on the performance of the banking industry during the 2008 financial crisis.
One of the core mechanisms of corporate governance is the audit committee. As the primary role of the audit committee is to oversee a firm's risk management practices, the present study focus on the characteristics of the audit committee and their influence over a firm's financial performance during the financial crisis periods. However, much of the empirical research (Leech and Leahy, 1991; Berger et al, 2007; Al-Hawary, 2011) on corporate governance structures were carried out on the impact of board composition, size and ownership structure on a firm's financial performance, very few considered the role of the audit committee. Accordingly, this study attempted to offer insights into the relationship between corporate governance mechanism described by the audit committee and bank performance during the global financial crisis. Additionally, the study shows that the influence of the audit committee characteristics on bank performance is different in crisis and non-crisis periods.
2.1.2 The Role of Audit Committee in the U.S.
The Sarbanes-Oxley Act (SOX, 2002) was instituted in the United States as a response to a series of corporate accounting scandals perpetrated by Enron and other companies that occur between 2000 to 2002. This act extends the roles and responsibilities of an audit committee. In a U.S. publicly traded company, an audit committee is a standing committee of a company’s board of directors that is charged with the oversight of the audit engagement and their company’s financial reporting process. In order to be listed on a stock exchange, all U.S. publicly traded companies are required to maintain a qualifying audit committee. The Sarbanes-Oxley Act 2002 specifies that listed companies must have
7
an audit committee comprising of at least three members who must be independent outside (non-executive) directors with at least one qualifying as a financial expert.
As a result of the passage of the Sarbanes-Oxley Act of 2002, the role of the audit committee has evolved and has progressively been redefined to improve the accuracy and reliability of corporate disclosures to shareholders. Pursuant to the Sarbanes-Oxley Act of 2002, the role of the committee is to assist the board in fulfilling its oversight responsibility on (i) the company’s financial reporting and internal control system; (ii) performance of the internal audit function; (iii) the implementation and effectiveness of accounting policies and principles; (iv) the evaluation of the independent auditors’ qualifications, independence and performance; and (v) the discussion of risk management policies and practices with the management. The committee is also required to prepare the report of the audit committee, which is included in the company’s annual proxy statement (SOX, 2002).
Four audit committee characteristics were identified namely audit committee size, audit committee financial expertise, activity and tenure to study their impact on US commercial banks listed on the S&P 500 index. Audit committee size is measured as the number of members serving on the committee (Bauer et al., 2009; Hsu & Petchsakulwong, 2010). Pursuant to the Sarbanes-Oxley Act of 2002, the Securities and Exchange Commission (SEC) mandates that audit committees consist of a minimum three directors.
Another audit committee requirement for listed publicly traded firms mandated by the Sarbanes-Oxley Act of 2002 is that the committee must be composed of at least one qualifying financial expert. Following the definition of the Sarbanes-Oxley Act of 2002, the variable audit committee expertise is included in this study as the proportion of audit committee members assigned as financial experts over the total number of audit committee members. This Sarbanes-Oxley Act of 2002 (SOX, 2002) defines audit committee financial expert as having an understanding of generally accepted accounting principles (GAAP) and financial statements, experience in (i) the preparation or auditing of financial statements of generally comparable issuers; and (ii) the application of such principles in connection with the accounting for estimates, accruals and reserves, experience with internal accounting controls and an understanding of audit committee functions.
The third proxy used in investigating the effectiveness of audit committee on bank
performance is the activity of the committee. The activity of audit committee is defined as the number of times Committee members meet in a year. Adams (2000) considers meetings as proxies of
8
diligence and an important resource in monitoring managerial performance. The Blue Ribbon Committee (1999) recommends that audit committees meet at least once a quarter.
Audit committee tenure is defined as the fourth characteristics in this study. It measures the average term length of all audit committee directors in the individual company in a given year. According to Spencer Stuart Board Index 2016, the average tenure of S&P 500 boards is 8.5 years. Only 3% of S&P 500 boards set an explicit term limit for non-executive directors. While 66% of boards explicitly state in their corporate governance guidelines that they do not have term limits and 31% do not mention term limits at all (Spencer Stuart, 2016).
2.1.3 Agency Theory
Studies examining the role of the audit committee in corporate governance originally arose out of the Agency Theory. Ross (1973) and Mitnick (1973) who independently developed the theory of agency, define it as the relationship between the principals such as the shareholders and the agents such as the company managers and executives. Agency theory suggests that the separation of ownership and control in a corporation can result in a principal-agent problem between the shareholders and management. This theory assumes that due to asymmetric information and conflicting interests between the principal and the agent, the principal cannot directly ensure that the agent acts in the best interest of the principal (Ross, 1973; Mitnick, 1973). However, the principal can reduce this agency dilemma by giving incentives to the agent and by enacting monitoring mechanisms to limit the self-serving nature of the agent (Fama and Jensen, 1983; Bonazzi and Islam, 2007).
One of the responsibilities of the management body of a large publicly held corporation is to suggest and implement major policies; however, shareholders do not always agree with these policies, which can lead to an agency problem between the management and the shareholders. As the role of the audit committee is to protect the shareholder’s interests in relation to financial oversight and control, implementing such corporate governance mechanism ensures that the interests of the managers are aligned with those of shareholders, which can, therefore, mitigate the agency conflicts within the corporation.
Agency theory suggests that there is a positive relationship between committee size and firm financial performance as agency costs reduce with larger committees that can provide more effective monitoring of management by reducing the domination of the CEO on the board. Consistent with the agency theory, Frooman (1999) and Donaldson (1999) argue that due to a greater diversity of
9
likely to effectively monitor management information systems on managers than that of a smaller committee. Although the agency theory expects a positive relationship between size and financial performance, a linear relationship between the two concepts should not be assumed; indicating that there is an inverted U-shaped relationship between size and firm performance. Jensen (1993) argues that as committee size increases, committees become less effective at monitoring management as free-riding problems amongst committee members increases. As committee size increases, difficulties in communication and coordination would arise, which may delay decision-making process; hence resulting in a decline of firm performance (Lipton and Lorsch, 1992).
In order to minimize information asymmetry, there is a need for governance mechanisms such as committees consisting of directors with attributes such as expertise and experience to prevent or reduce the self-serving and individualistic nature of the agent (Wiseman et al., 2012). The
appointment of audit committee members with advanced financial and accounting knowledge increases monitoring effectiveness of management processes such as internal accounting control and financial reporting as they have a better understanding and the experience of auditing and internal control evaluation. From an agency theory perspective, audit committee members who are competent in financial aspects are expected to adopt a high standard of accountability, which would ensure the transparency in financial information, and hence would alleviate the agency conflicts between the managers and shareholders. Consistent with the agency theory, DeZoort and Salterio (2001) argue that financially experienced members are more likely to detect material misstatements and are able to correct it in a timely manner, which would consequently prevent corporate fraud. In turn, this would lead to achieving firm financial performance with increased financial expertise within audit committee members.
Agency theory expects a committee that conducts more frequent meetings to have greater opportunity in performing their duties efficiently as they remain informed and knowledgeable about firm-specific information, and hence able to execute their monitoring role. An active and effective committee ensures that agents act in the best interests of shareholders as committee members are well-informed about the firm’s current position, and therefore are in better position to promptly respond to emerging critical issues (Mangena and Tauringana, 2008). According to Vafeas (1999a), a committee that meets less frequently may be an indicator of a lack of diligence and monitoring which can create incentives for agents to pursue selfish objectives. Lipton and Lorsch (1992) found committees that meet more frequently tend to generate higher financial performance.
10
Agency theory expectations regarding audit committee member tenure can be interpreted both positively and negatively on the degree of monitoring. For long tenure members, increased experience allows them to develop their monitoring competencies and gain a deeper knowledge of the firm’s operations and its executives (Beasley, 1996). Thus, with increased experience, long tenure members become more effective at monitoring the firm’s financial reporting process. On the contrary,
entrenchment may increase with director’s tenure. According to Vafeas (2003), friendly relationships are more likely to develop between long tenure members and the management. As a result, long tenure members are more likely to inadequately monitor managers, and thus are more likely to be ineffective. Herewith, it may be suggested that there is an inverted U-shaped relationship between tenure and firm financial performance.
2.1.4 Firm Financial Performance
Proxies to measure firm financial performance are categorized in two classifications: accounting and market-based. To measure a firm's accounting performance, Adjaoud, Zeghal and Andaleeb (2007), Bauer et al. (2009) and Zeitun and Gang Tian (2007) use return on assets (ROA), return on equity (ROE) and return on investment (ROI). In this research, ROA is used as a proxy to measure the accounting-based performance of banks. ROA is calculated by the ratio between net income and total book value assets. According to Ibrahim and Samad (2011), ROA reflects the firm's effective use of its assets in generating profits.
Tobin’s Q has been widely adopted in many empirical studies on corporate governance and firm performance such as Al-Matari et al. (2012), Reddy et al. (2010) and Bauer et al. (2009). Hence, Tobin’s Q is used in this study as a proxy in estimating the market-based performance of banks. In this research, Tobin’s Q is calculated by the total market value divided by the total assets (book) value of a firm.
2.2 Empirical Findings
2.3.1 Audit committee size and firm performance
Several studies have established that audit committee size affect firm performance. It is argued that increased in the number of committee members provide more effective monitoring of management and thus lead to better firm performance. Consistent with this notion, in the research of Bauer et al. (2009), they found that audit committee size is positively related to the ROA, ROE and Tobin’s Q of 113 real estate investment trust firms in the US during 2004 and 2006. Additionally, Reddy et al. (2010) conducted a study of 50 companies in New Zealand over the period of 1999 to 2007 and found
11
a positive relationship between audit committee size and performance; measured in terms of ROA and Tobin’s Q.
On the other hand, scholars such as Jensen (1993) and Lipton and Lorsch (1992) support small committees suggesting that in larger committees, free riding increases, which reduces the efficiency of the committee. Bozec (2005) conducted a study of 500 large firms that were listed on the Canadian Stock Exchange during the period of 1976 to 2000 and his result showed that audit
committee size has a negative impact on the ROA of the firms. Supporting Bozec (2005) finding, Al-Materi et al. (2012) also found evidence that there is a significant negative relationship between audit committee size and Tobin’s Q of 135 firms listed on the Saudi Stock Market in 2011. . Taken
together, the literature remains inconclusive on the relationship between board size and company performance. Nonetheless, the literature seems to suggest that there is a trade-off between the benefit and the cost of instituting a large committee. Therefore, the following hypothesis is proposed:
H1: There is an inverted U-shaped relationship between audit committee size and bank performance during the global financial crisis.
2.3.2 Audit committee expertise and firm performance
The next characteristic examined refers to the financial expertise of audit committee members. Pursuant to the Sarbanes-Oxley Act of 2002, it is mandated that the committee is composed of at least one qualifying financial expert. Krishnan and Visvanathan (2008) confirm that the presence of members with financial experience reduces financial reporting and enhances quality monitoring, improving the effectiveness of audit committees. Hamid and Aziz (2012) conducted a multiple regression analysis on audit committee member financial expertise and ROA of 33 Malaysian government-linked companies. Their result showed that there is a significant positive relationship between audit committee financial expertise and firm performance.
Consistent with the findings of Hamid and Aziz (2012), Rahmat, Iskandar and Saleh (2009), concluded that financial performance of 73 public listed companies in Malaysia improves with increased financial experts on the committee. They explained that audit committee with accounting financial experts (directors with experience as a certified public accountant, controller or chief accounting officer) will monitor financial and operational reports more efficiently. Thus, the following hypothesis is formulated:
12
H2: There is a positive impact of financial expertise of the audit committee on bank performance during the financial crisis
2.3.3 Audit committee activity and firm performance
The number of audit committee meetings in a year is considered as a proxy for audit committee activity. It is expected that audit committees that meet frequently will be more effective monitors that will provide more reliable information to the shareholders. Consistent with this expectation, Khanchel (2007) studied the determinants of good corporate governance in 624 US firms for the period of 1994 to 2003 and found that higher activities of committees strengthen governance quality, which results in a positive correlation with the Tobin’s Q of firms. He explains that an active audit committee that meets more frequently have more time to oversee the financial reporting process and monitor internal controls. As a result, firm performance increases with audit committee activity.
In the opposite direction, a study conducted by Hsu and Petchsakulwong (2010) found that over the period 2000 to 2007, audit committee meetings have a negative impact on the performance efficiency of 18 Thai non-life insurance companies. An explanation for this could be that due to the limited time committee members spend together, much of the meetings are occupied by routine tasks such as presenting management reports, which reduces the opportunities for directors to effectively monitor management, and hence negatively impair the financial performance of firms (Vafeas, 1999a). This suggests that the frequency of meetings held by the committee does not indicate their efficiency during their meetings. Therefore, the following hypothesis is proposed:
H3: There is a positive impact on the activity intensity of the audit committee on bank performance during the financial crisis
2.3.4 Audit committee tenure and firm performance
The fourth audit committee characteristics examined is the average tenure of the committee members. Consistent with the notion that due to greater expertise and experience of long tenure committee members, Liu and Sun (2010) conducted a research using a large sample size of 7,700 firm-year observations over the period 1998 to 2005 in U.S. that provide strong evidence that an audit committee with a high proportion of long tenure directors is positively correlated with firm performance, measured in terms of discretionary accruals. Their finding implies that long tenure members with greater expertise, experience and reputation are more effective in overseeing the
13
financial reporting process of firms. As a result, the likelihood of financial statement fraud decreases, improving the financial performance of firms.
Huang (2013) found that tenure of US firms listed on the S&P 1500 exhibits an inverted U-shaped relation with firm value, measured by Tobin’s Q. Covering the period of 1998 to 2010, his research deduced that firm value increases from average tenure of three years and reaches a peak value in Tobin’s Q around nine years tenure. After its peak, adding one year to tenure decrease firm value by an average of 0.52% (Huang, 2013). These results suggest that when tenure is shorter, the marginal value of learning exceeds the marginal cost of entrenchment, but as tenure continues to increase, the entrenchment effect dominates the learning effect. Thus the following hypothesis is constructed:
H4: There is an inverted U-shaped relationship between audit committee tenure and bank performance during the global financial crisis
3 Methodology and Data
3.1 Methodology
3.1.1 Regression Model
In order to test the hypothesized relationship between audit committee characteristics and bank performance over the period of 2005 to 2010, two fixed effect panel data regression models, one for each dependent variable were estimated using Stata. Table 3.1 summarizes the definition,
measurement and expected sign for each independent variable along with the dependent and control variables. The two regression models utilized to test the relationship between the audit committee (AC) characteristics and bank performance are as follows:
ROAi,t = β0+ β1(AC Characteristics)i,t + β2(Crisis)i,t + β3(AC Characteristics*Crisis)i,t + β4(Control
Variables)i,t + ɛi,t
(1)
Tobin’s Qi,t = β0+ β1(AC Characteristics)i,t + β2(Crisis)i,t + β3(AC Characteristics*Crisis)i,t +
14
3.2 Data
The present study analyzes 22 commercial banks listed on Standard & Poor’s 500 index (S&P 500) with a total of 132 observations. Financial data are extracted from Compustat North America
database. Data on audit committee characteristics are manually collected from DEF 14A forms (proxy statement) of the selected companies for the years 2005-2010. The crisis period is defined as the years 2007-2009. The dependent variables used in this study are return on assets (ROA), an accounting measure and Tobin’s Q, a market performance measure. For the independent variables, four measures of audit committee characteristics are considered: committee size, financial expertise (members with experience as a certified public accountant, controller or chief accounting officer over a total number of audit committee members), tenure and activity intensity. Bank size and leverage are employed as control variables of the empirical model. Financial data used to compute bank size and leverage are obtained from Compustat.
3.2.1 Control Variables
When studying the impact of audit committee characteristics on bank performance, bank size, leverage and credit default swaps (CDS) are chosen as control variables. Bank size is the book value of total assets in millions of dollars (Klein, 2002; Sharma et al., 2009). As firms with high leverage
Table 3.1 Dependent, Independent and Control Variables
Variable Measurement Predicted
sign
Dependent variables
ROA Return on assets (in percentage points) = Net income divided by
total book value of assets
Tobin’s Q Total market value of assets divided by total book value of assets
Audit Committee characteristics
Size Number of people serving on the audit committee +
Size2 Squared term of the variable Size -
Expertise Proportion of audit committee members with experience as certified public accountant, controller or chief accounting officer over total number of audit committee members
+
Activity Number of meetings reported in a given year +
Tenure Average number of years that directors serve on the audit
committee +
Tenure2 Squared term of the variable Tenure -
Dummy variable
Crisis Crisis = 1 if Years = 2007, 2008, 2009, 0 otherwise +
Control variables
BSize Bank size = the book value of assets (in million dollars) +
15
are more likely to be impacted by the financial crisis, a control for leverage, LEV has been included. In this study, leverage is total debt divided by total shareholder’s equity.
4 Results
4.1 Descriptive Statistics
The results of descriptive statistics for the entire sample are reported in Table 4.1. On average, there are 5 members on the audit committees of S&P 500 commercial banks. The minimum number of the AC members is 3 as it is the legal requirement by the Sarbanes-Oxley Act of 2002 and the maximum is 9. Audit committee members meet approximately 11 times in a year on average. However, it is interesting to note that some audit committees do not meet at all in one year while some meet on a bi-monthly basis. 50% of audit committee members are considered as financial experts. The average tenure of audit committee members is approximately 7.5 years. The average return on asset was found to be 0.67% and the average Tobin’s Q ratio was 1.04 over the years 2005 to 2010.
Table 4.1 Descriptive statistics for the years 2005-2010
Variable Mean Median Std. Dev. Min Max
Size 5.01 5.00 1.28 3.00 9.00 Expertise 0.51 0.40 0.29 0.13 1.00 Activity 11.15 12.00 3.44 0.00 23.00 Tenure 7.46 7.50 3.34 0.67 15.60 Bank size 0.45 0.15 0.62 0.01 2.26 Leverage 3.11 2.05 3.47 0.01 18.02 ROA 0.67 0.91 1.07 -6.00 2.91 Tobin’s Q 1.04 1.03 0.07 0.93 1.47 4.2 Correlation Analysis
In order to discover possible correlation among the variables, Pearson correlation analysis was performed. To obtain unbiased results of the regression, it is necessary that the variables do not correlate with each other. If independent variables are highly correlated, it may raise multicollinearity issue which will lead to biased overall result. From table 4.2, apart from Bank Size and Leverage, none of the variables are highly correlated as the Pearson correlation coefficients are less than 0.4.
Table 4.2 Correlation matrix for independent variables
Size Expertise Activity Tenure Bank size Leverage
Size 1 Expertise 0.260 1 Activity -0.026 0.219 1 Tenure -0.013 -0.213 -0.002 1 Bank size 0.086 0.339 0.162 -0.253 1 Leverage 0.180 0.384 -0.043 -0.380 0.426 1
16
4.3 Research Findings and Implications
The independent variables represented by four audit committee characteristics: size, expertise, tenure and activity were regressed against the dependent variable, ROA. To further test the robustness of the results, regression on market performance (Tobin’s Q), modeled as a function of the audit committee characteristics was performed. The relationship between the audit committee characteristics and bank performance were estimated using panel regression. After running the Hausman specification test on both ROA and Tobin’s Q regressions, the null hypotheses were rejected and thus the fixed effect specification was applied to the models. The results for the accounting (ROA) and market performance (Tobin’s Q) analyses are reported in Tables 4.3 and 4.4 respectively.
It is important to note that the R-squared value for the ROA crisis model, model (6) in Table 4.3 is around 54.6% indicating that only 54.6% of variations in ROA is determined by the audit committee characteristics used in the regression. Whereas the remaining 45.4% of ROA variations would be further discussed by other variables which have not been included in this study. The value of R-squared for Tobin’s Q crisis model was much higher than that of ROA. It indicates that 71.4% of Tobin’s Q variations are determined by the audit committee characteristics used in this study while 28.6% of variations in Tobin’s Q is further discussed by other variables which have not been included in the present study. Given that the probability of F-statistics value is less than 0.1, the overall model is considered to be statistically significant. Thus, the variables used in the regression models can jointly estimate the performance of U.S. banks in our sample during the financial crisis.
Hypothesis 1. The first hypothesis (H1) predicts that there is an inverted U-shaped
relationship between the size of the audit committee and bank performance during the global financial crisis. However, the results of the regression models are inconsistent with the hypothesis. The results in Table 4.3, model (6) indicates that audit committee size has a convex relationship with ROA, which can be interpreted by a negative and significant association between Size and ROA (coefficient = -.633, p<0.1) and a positive and significant association between Size2 and ROA (coefficient = .031). Holding all other variables constant, the results of panel data indicate that smaller audit committee impairs bank performance, but after a certain size of the audit committee is reached, with each additional number of members on the audit committee, ROA increases changes with 0.571%, during non-crisis periods.
Table 4.4 reports the regression analyses of audit committee size and Tobin’s Q. Consistent to the regressions on ROA, the regression models of Tobin’s Q estimate a convex relationship between
17
the two concepts. Holding all other variables constant, the results of Table 4.4 indicate that an audit committee of a smaller size performed worse, but by appointing an extra member on the audit committee will change the value of Tobin’s Q with 0.039. These results are inconsistent with the findings of Andres and Vallelado (2008) who found an inverted U-shaped relation between size and performance of U.S. commercial banks. One possible reason for the convex relationship is that a larger committee size usually consists of more directors with diversified education and expertise background. With more diversified perspectives and greater quality of opinions, this increases the resources platform and increases capabilities in problem-solving under exogenous financial pressure, which in turn leads to better firm performance during non-crisis periods.
Although the direction of the linear and quadratic coefficients of audit committee size for the both ROA and Tobin’s Q regressions suggest a convex relationship between the two concepts, the coefficients on these variables are insignificant when interacted with Crisis. This implies that the size of the audit committee did not have an influence on bank performance measured in terms of ROA, during the crisis.
Hypothesis 2. The second hypothesis (H2) states that there is a positive relationship between the audit committee financial and bank performance during the global financial crisis. However, the result of the regressions in Table 4.3 contradicts this hypothesis. Models (2)-(6) present that the coefficients are negative and significant at the 5 percent level, which implies that with increased proportion of financial experts (members with experience as a certified public accountant, controller or chief accounting) on the committee, ROA decreases. When interacting the variable Expertise with
Crisis, the coefficient estimates for Expertise are also negative and significant. Model (6) shows that
the overall effect of Expertise on ROA during the financial crisis is negative. Holding all other variables constant, model 6 of Table 4.3 predicts that ROA changes with -1.227% for every unit change in the proportion of audit committee members with financial qualifications. These results imply that banks with more financial expertise among audit committee members performed worse during the global financial crisis.
Similarly, as observed in Table 4.4, when Expertise and Expertise x Crisis are regressed against Tobin’s Q, the direction and significance of the coefficients are the same as when Expertise is regressed with ROA. Model (6) presents that the overall effect of Expertise on Tobin’s Q during the financial crisis. Holding all other variables constant, this indicates that for every unit change in the proportion of financially qualified audit committee members, Tobin’s Q decreases by 0.118. These results correspond with the findings of Minton, Taillard and Williamson (2010) and Erkens, Hung and
18
Matos (2012). A possible explanation for the negative relationship between the two concepts could be that banks with audit committee consisting of financially expert directors who were also outside directors were more willing to undertake risk taking activities taken by the banks prior to the financial crisis. Due to their knowledge of the financial instruments and valuation techniques, audit committee directors with more financial expertise provided their support in expanding the risk-taking profile taken by the banks. As a result, the performance of the banks with more financial expertise on the audit committee was impaired during the financial crisis.
Hypothesis 3. The third hypothesis (H3) assumes that audit committee activity will positively contribute to bank performance during the global financial crisis. Based on Table 4.3, there is no evidence to support the third hypothesis on ROA. Models (1)-(3) indicate that there is a negative but non-significant association between the number of audit committee meetings and ROA without year fixed effects but models (4)-(6) show that the two concepts are positively related with year fixed effects. The same pattern is observed when a crisis effect is added to the regressions. Controlling for year fixed effects, model (6) indicates that although the effect of the number of audit committee meetings on ROA during the crisis period is positive, the effect is small and insignificant (coefficient = .014, p>0.1).
Based on the results of Table 4.4, the same conclusion can be reached when Activity is regressed against Tobin’s Q. Controlling for year fixed effects, model (6) indicates that there is a positive association between Activity and Tobin’s Q (coefficient = 0.003) but a negative association when a crisis effect is added (coefficient = -0.002). However, the coefficient estimates are small and insignificant. Contradictory to the findings of Khanchel (2007) who found a positive significant relationship between audit committee meetings and Tobin’s Q, our findings suggest that the frequency of audit committee meetings do not explain the performance of banks during the financial crisis. A possible explanation for this result is that how a firm performs under negative shocks depends most closely to the quality of day-to-day management activities of the bank and is less affected by the activity intensity of the audit committee.
Hypothesis 4. The last hypothesis tested (H4) predicts that there will be an inverted U-shaped relationship between the tenure of audit committee members and bank performance during the global financial crisis. However, our results are inconsistent with this hypothesis. The results of Table 4.3, model (6) show that there is a positive, significant association between Tenure and ROA (coefficient = .026, p-value< .1) and a negative, significant association between Tenure2 and ROA (coefficient = -.001, p-value<.05). This indicates that there is an inverted U-shaped relationship between the tenure
19
of the audit committee members and ROA during non-crisis periods. Holding all other variables at a constant, during the non-crisis period, the results of the model (6) indicate that the change in ROA is a negative function of Tenure, for every additional year of director tenure on the audit committee, ROA will change with 0.024%. This effect is positive up to Tenure = 13, after that it is negative.
Additionally, holding all other variables constant, the results of the model (6) of Table 4.4 indicate that the change in Tobin’s Q is a negative function of Tenure, for every additional year of director tenure on the audit committee, Tobin’s Q will change with 0.014. This suggests that an increase in average term length of directors on the audit committee is associated with the financial performance of banks, but when the tenure of the members gets too long, this increase will impair the performance of banks during non-crisis periods. The results are consistent with the findings of Vafeas (2003) who suggest a possible reason for this is that over time, long-tenured directors are more likely to develop friendly relationships with the management and lose their ability to objectively monitor the management’s actions and thus decreases the level of the committee independence. As the level of audit committee independence is associated with the quality of financial reporting processes, this would deteriorate their monitoring mechanism and in turn lead to the erosion of firm value (Klein, 2002).
However, the impact of Tenure on ROA and Tobin’s Q is not significant during the global financial crisis. When adding a crisis effect to the regressions, model (6) of Tables 4.3 and 4.4 show a nonlinear but insignificant relationship between the tenure of audit committee members and bank performance during the financial crisis. This suggests that at times of financial distress, the tenure of audit committee members do not influence the accounting and market performance of banks.
In term of the control variables, bank size shows a non-significant relationship, while bank leverage suggests a negative significant relationship with both ROA and Tobin’s Q. As expected in an economic downturn, leverage is significantly related to lower financial performance.
20
Table 4.3 Audit committee characteristics and ROA
(1) (2) (3) (4) (5) (6) Size -.041** (.305) -.055** (.304) (.276) -.254 (.515) -.564 -.615* (.519) -.633* (.508) Crisis .589 (1.316) (1.501) .267 (1.56) .165 (1.613) -.052 (1.58) .087 (1.547) -.382 Size x Crisis -1.01* (.548) -.905* (.563) (.556) -.845 (.602) -.83 (.584) -.874 (.563) -.787 Size2 -.004* (.029) -.004 (.03) (.027) -.023 (.039) .022* .027* (.039) .031** (.038) Size2 x Crisis .091* (.05) (.052) .081 (.052) .077 (.054) .078 (.052) .083 (.049) .072 Expertise -.540 (.346) -.631* (.379) -.747** (.373) -.98** (.373) -1.003*** (.345) -1.016** (.372) Expertise x Crisis .696 (.541) (.558) .792 (.541) .876* -.265* (.511) -.248* (.495) -.211* (.504) Activity -.026 (.019) (.021) -.028 (.021) -.022 (.036) .023 (.036) .026 (.032) .013 Activity x Crisis -.013 (.044) (.043) -.015 (.041) -.02 (.042) .006 (.042) .001 (.041) .001 Tenure .126* (.069) (.074) .141* .193** (.072) (.096) .085 (.086) .061* .026* (.081) Tenure x Crisis .168 (.245) (.245) .176 (.244) .157 (.196) .082 (.198) .087 (.202) .123 Tenure2 -.007 (.004) (.005) -.008 -.01** (.004) (.005) -.003 -.002* (.004) -.001* (.003) Tenure2 x Crisis -.005 (.013) (.013) -.005 (.013) -.005 -.001 (.01) (.011) -.001 (.011) -.002 Bank size .124 (.169) (.158) .087 (.358) -.378 (.285) .006 Leverage .045* (.023) -.106** (.039) Constant 1.077 (.879) (.861) 1.09 (.769) .187 3.187* (1.649) 3.531** (1.597) 4.022** (1.623)
Year fixed effects No No No Yes Yes Yes
N 132 132 132 132 132 132
R2 0.269 0.267 0.259 0.528 0.531 0.546
Adjusted R2 0.203 0.237 0.235 0.276 0.456 0.469 Notes: Return on Assets (ROA) is defined in terms of percentage points. *significant at the 10% level; **significant at the 5% level; ***significant at the 1% level. Each number represent the coefficient estimate of the corresponding variable. The numbers in the brackets are the robust standard errors of the coefficient estimates.
21
Table 4.4 Audit committee characteristics and Tobin’s Q
(7) (8) (9) (10) (11) (12) Size -.029 (.036) (.035) -.035 (.032) -.026 -.041** (.017) -.04** (.016) -.041** (.016) Crisis -.069 (.11) (.105) -.037 (.095) -.024 (.073) -.051 (.076) -.054 (.084) -.065 Size x Crisis -.005 (.036) -.015* (.034) (.031) -.016 (.022) -.012 (.024) -.009 (.025) -.007 Size2 .002 (.003) (.003) -.002 (.003) .001 .001** (.001) .001** (.001) .001** (.001) Size2 x Crisis .001 (.003) (.003) .001 (.003) .002 (.002) .003 (.004) .003 (.002) .001 Expertise -.059 (.041) (.042) -.054 (.04) -.06 (.042) -.065 (.041) -.06 -.061 (.04) Expertise x Crisis .073 (.032) -.067** (.003) -.068** (.03) (.031) -.06* -.058* (.03) -.057* (.031) Activity -.001 (.002) (.002) -.001 (.002) .001 (.002) .003 (.002) .003 (.032) .003 Activity x Crisis -.001 (.001) (.002) -.001 (.002) -.001 (.002) -.002 (.002) -.002 (.002) -.002 Tenure .021** (.011) .018* (.011) .02** (.01) .016** (.008) (.008) .016* (.008) .002* Tenure x Crisis -.003 (.008) (.007) -.003 (.007) -.005 (.005) -.002 (.005) -.002 (.006) -.001 Tenure2 -.001** (.001) -.001* (.001) -.001** (001) -.001* (.001) (.001) -.001* -.001* (.001) Tenure2 x Crisis .001 (.001) (.013) .005 (.001) .002 (.001) -.001 (.001) -.001 (.001) -.001 Bank size -.021*** (.008) -.033*** (.007) (.018) .007 (.022) .016 Leverage .005*** (.002) (.002) -.002 Constant 1.142*** (.102) 1.171*** (.101) 1.12*** (.097) 1.16*** (.037) 1.157*** (.018) 1.17*** (.037)
Year fixed effects No No No Yes Yes Yes
N 132 132 132 132 132 132
R2 0.376 0.396 0.436 0.712 0.712 0.714
Adjusted R2 0.34 0.361 0.401 0.669 0.667 0.666 Notes: *significant at the 10% level; **significant at the 5% level; ***significant at the 1% level. Each number represent the coefficient estimate of the corresponding variable. The numbers in the brackets are the robust standard errors of the coefficient estimates.
4.4 Robustness Check
To further test the robustness of the results, an additional control variable, credit default swap (CDS) spread of bank sample was added to control for risk measure. The probability that financial
institutions default together is known as systemic default risk and at times of financial distress, this probability can abruptly increase. Due to the high interconnection between financial intermediaries and exposure to common stocks, it is important to construct systemic risk measures such as CDS spread to capture the joint default risk (Adrian and Brunnermeier, 2009). As data on CDS spread of our bank’s sample were limited, 5-year CDS spread (in basis points) were collected from Datastream, with a total of 48 observations. Table 4.5 reports the regression results for both ROA and Tobin’s Q
22
with the addition of CDS as a control variable. The results for the model (2) in Table 4.5 shows that the coefficient estimates of Expertise for both ROA and Tobin’s Q are negative and significant. This indicates that increased in the proportion of directors who have financial qualifications do not improve the performance of banks during the financial crisis. Furthermore, the quadratic coefficient estimates of Tenure for the regression on both ROA and Tobin’s Q are negative and significant at the 5 percent level. The results show that the control variable, credit default swap (CDS) do not have an impact on ROA and Tobin’s Q. These results confirm the prior regression results of Table 4.3 and 4.4.
Table 4.5 Robustness test with the control variable, CDS spread
ROA Tobin’s Q (1) (2) (3) (4) Size -.221** (.505) (.515) -.598 (.062) .008 (.025) -.025 Crisis 1.871 (2.815) (2.074) .294 (.169) -.085 (.118) -.131 Size x Crisis -1.109 (.892) (.537) -.654 (.071) -.002 (.038) .042 Size2 .011 (.038) (.039) .027 (.005) -.001 (.002) .002 Size2 x Crisis .098 (.074) (.042) .059 (.005) .001 (.003) -.002 Expertise -.228 (.633) -.276* (.654) (.053) .027 -.059* (.002) Expertise x Crisis .631 (.463) .584** (.466) (.046) .042* (.032) .001* Activity -.047 (.039) (.031) -.013 (.002) -.006 (.002) .006 Activity x Crisis .021 (.057) (.092) -.011 (.002) .005 (.003) -.003 Tenure .302 (.241) (.209) .208 (.018) .022 (.018) .018 Tenure x Crisis -.05 (.205) -.122** (.241) (.002) -.002 -.04** (.015) Tenure2 -.023 (.019) (.017) -.002 (.002) -.002 (.001) -.001 Tenure2 x Crisis .001 (.012) (.018) .009 (.002) .002 .003** (.001) Bank size -.169 (.189) (.172) .013 (.007) -.001 (.021) .015 Leverage -.027* (.015) (.049) -.055 (.002) -.001 (.004) -.003 CDS -.001 (.001) (.002) .003 (.001) -.001 (.001) .001 Constant 1.916 (1.201) 3.045*** (.877) 1.033*** (.162) 1.072*** (.052)
Year fixed effects No Yes No Yes
N 48 48 48 48
R2 0.515 0.565 0.481 0.673
Adjusted R2 0.482 0.522 0.401 0.669
Notes: Return on Assets (ROA) is defined in terms of percentage points. *significant at the 10% level; **significant at the 5% level; ***significant at the 1% level. Each number represent the coefficient estimate of the corresponding variable. The numbers in the brackets are the robust standard errors of the coefficient estimates.
23
5 Conclusion and Limitation
Consistent with prior research, the present study estimates panel data fixed effect regression to identify the impact of the size, financial expertise, activity and tenure of the audit committee on the accounting (ROA) and market performance (Tobin’s Q) of U.S. commercial banks during the 2008 financial crisis. The models were estimated using a sample of 22 commercial banks listed on the S&P 500 index in 2008. Further analysis included an additional control variable, credit default swap spread in the same model to check the robustness of the estimated models. The results indicate that the size, expertise and tenure of the audit committee have an impact on bank financial performance during non-crisis periods, but a limited number of audit committee characteristics are associated with the financial performance of the bank’s sample during the financial crisis.
Consistent with the agency theory, this study provides evidence that there is an inverted U-shaped relation between the tenure of audit committee members and firm performance during non-crisis periods. However, inconsistent with previous findings, the relationship between size and performance is convex, indicating that the more the number of members serving on the audit
committee, the greater the banks perform at times of stable market conditions. Also inconsistent with the agency theory and previous research, it is found that evidence that financial expertise committee (members with experience as a certified public accountant, controller or chief accounting officer) negatively impacts both accounting and market performance of banks for both non-crisis and crisis periods. One reason for this is that financially qualified members of the audit committee took on greater risk-taking activities prior to the crisis, which in turn impair the performance of banks. However, no evidence was found on the impact of audit committee size, activity and tenure at times of a financial distress. Thus, we provide evidence that the interaction between audit committee characteristics and bank performance is different under normal market conditions and during periods of financial distress.
As each study, this one also has its limitations. First, this study concentrates only on certain corporate governance mechanisms regarding four audit committee characteristics that cover size, expertise, activity and tenure of the audit committee to determine any relationship with bank performance during the global financial crisis. Hence, further studies are encouraged to explore additional independent variables such as board size, composition and diversity, CEO duality and ownership which may possibly have an influence on bank performance. Furthermore, further research can incorporate other audit committee attributes that were not incorporated in this study such as the
24
number of grey directors and the average number of directorships held by the audit committee members. As this study was only theoretically constructed based on the agency theory, further study should apply various methodologies to explore different measures of corporate governance
25
7 References
Adams, R. B. (2003). What do boards do? Evidence from board committee and director compensation data.
Adjaoud, F., Zeghal, D., & Andaleeb, S. (2007). The effect of board's quality on performance: A study of Canadian firms. Corporate Governance: An International Review, 15(4), 623-635. Aebi, V., Sabato, G., & Schmid, M. (2012). Risk management, corporate governance, and bank performance in the financial crisis. Journal of Banking & Finance, 36(12), 3213-3226.
Al-Hawary, S. (2011). The Effect of banks governance on banking performance of the Jordanian commercial banks Tobin’s q model. International Research Journal of Finance and Economics, I. 71, p. 34-47.
Al-Matari, Y. A., Al-Swidi, A. K., Fadzil, F. H. B., Fadzil, H., & Al-Matari, E. M. (2012). Board of Directors, Audit Committee Characteristics and the Performance of Saudi Arabia Listed
Companies. International Review of Management and Marketing, 2(4), 241.
Bauer, R., Eichholtz, P., & Kok, N. (2010). Corporate governance and performance: the REIT effect. Real Estate Economics, 38(1), 1-29.
Beasley, M. S. (1996). An empirical analysis of the relation between the board of director composition and financial statement fraud. Accounting review, 443-465.
Berger, A. N., Clarke, G. R., Cull, R., Klapper, L., & Udell, G. F. (2005). Corporate governance and bank performance: A joint analysis of the static, selection, and dynamic effects of domestic, foreign, and state ownership. Journal of Banking & Finance, 29(8), 2179-2221.
Bhardwaj, M. N., & Rao, C. D. B. R. (2015). Role of audit committee in corporate governance. International
Bonazzi, L., & Islam, S. M. (2007). Agency theory and corporate governance: A study of the effectiveness of board in their monitoring of the CEO. Journal of Modelling in Management, 2(1), 7-23.
Bozec, R. (2005). Boards of directors, market discipline and firm performance. Journal of Business
Finance & Accounting, 32(9‐10), 1921-1960.
Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007–2008. The Journal
of economic perspectives, 23(1), 77-100.
DeZoort, F. T., & Salterio, S. E. (2001). The effects of corporate governance experience and financial-reporting and audit knowledge on audit committee members' judgments. Auditing: A
Journal of Practice & Theory, 20(2), 31-47.
Donaldson, L., 1999. Making stakeholder theory whole. Academy of Management. The Academy of Management Review, 24 (2), 237-241.
Fama, E. F., & Jensen, M. C. (1983). Separation of ownership and control. The journal of law and
Economics, 26(2), 301-325.
Frooman, J. (1999). Stakeholder influence strategies. Academy of management review, 24(2), 191-205.
26
Gunes, N., & Atilgan, M. S. (2016). Comparison of the Effectiveness of Audit Committees in the UK and Turkish Banks. International Journal of Financial Research, 7(2), 18.
Hamid, A., & Aziz, R. (2012). Impact of the amendments of Malaysian code of corporate governance (2007) on governance of GLCs and performance. World Academy of Science,
Engineering and Technology, International Journal of Social, Behavioral, Educational, Economic, Business and Industrial Engineering, 6(11), 1622-1627.
Hsu, W. Y., & Petchsakulwong, P. (2010). The impact of corporate governance on the efficiency performance of the Thai non-life insurance industry. The Geneva Papers on Risk and Insurance
Issues and Practice, 35(1), S28-S49.
Huang, S. (2013). Zombie boards: Board tenure and firm performance.
Ibrahim, H., & Samad, F. A. (2011). Corporate governance mechanisms and performance of public-listed family-ownership in Malaysia. International Journal of Economics and Finance, 3(1), 105. Jensen, M.C., 1993. The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance 48, 831–880.
Kashyap, A. K., Rajan, R., & Stein, J. C. (2008). Rethinking capital regulation. publisher not identified.
Khanchel, I. (2007). Corporate governance: measurement and determinant analysis. Managerial
Auditing Journal, 22(8), 740-760.
Klein, A. (1998). Firm performance and board committee structure 1. The Journal of Law and
Economics, 41(1), 275-304.
Klein, A. (2002). Audit committee, board of director characteristics, and earnings management.
Journal of accounting and economics, 33(3), 375-400.
Leech, D., & Leahy, J. (1991). Ownership structure, control type classifications and the performance of large British companies. The Economic Journal, 101(409), 1418-1437.
Lipton, M., Lorsch, J., 1992. A modest proposal for improved corporate governance. Business Lawyer 48, 59–77
Liu, G., & Sun, J. (2010). Director tenure and independent audit committee effectiveness.
International research journal of finance and economics, 51.
Mangena, M., & Tauringana, V. (2008). Audit committees and voluntary external auditor involvement in UK interim reporting. International Journal of Auditing, 12(1), 45-63.
Mitnick, B. M. (1973). Fiduciary rationality and public policy: The theory of agency and some consequences.
Mohid Rahmat, M., Mohd Iskandar, T., & Mohd Saleh, N. (2009). Audit committee characteristics in financially distressed and non-distressed companies. Managerial Auditing Journal, 24(7), 624-638.
27
Peni, E., & Vähämaa, S. (2012). Did good corporate governance improve bank performance during the financial crisis? Journal of Financial Services Research, 41(1-2), 19-35.
Poudel, R. P., & Hovey, M. (2012). Corporate governance and efficiency in Nepalese commercial banks.
Reddy, K., Locke, S., & Scrimgeour, F. (2010). The efficacy of principle-based corporate governance practices and firm financial performance: An empirical investigation. International
Journal of Managerial Finance, 6(3), 190-219.
Ross, S. A. (1973). The economic theory of agency: The principal's problem. The American
Economic Review, 63(2), 134-139.
Sarbanes-Oxley Act (2002).Public company accounting reform and investor protection: (US Congress, Washington).Washington Press.
Sharma, V., Naiker, V., & Lee, B. (2009). Determinants of audit committee meeting frequency: Evidence from a voluntary governance system. Accounting Horizons, 23(3), 245-263.
Vafeas, N. (1999). Board meeting frequency and firm performance. Journal of financial economics,
53(1), 113-142.
Vafeas, N. (2003). Length of board tenure and outside director independence. Journal of Business
Finance & Accounting, 30(7‐8), 1043-1064.
Wild, J. J. (1996). The audit committee and earnings quality. Journal of Accounting, Auditing &
Finance, 11(2), 247-276.
Wiseman, R. M., Cuevas‐Rodríguez, G., & Gomez‐Mejia, L. R. (2012). Towards a social theory of agency. Journal of Management Studies, 49(1), 202-222.
Zeitun, R., & Gang Tian, G. (2007). Does ownership affect a firm's performance and default risk in Jordan?. Corporate Governance: The international journal of business in society, 7(1), 66-82.
