Board Independence and Firm Performance
Name: Marsha van Hattem
Student Number: 10643230
Programme: Economie en Bedrijfskunde
Track: Financiering en Organisatie
Name supervisor: Ross Gardner
Date: 31/01/2018
Abstract
This thesis focuses on the effects of board independence on the performance of US-listed firms in the period 1999 to 2005. In this time period, the Sarbanes-Oxley Act (SOX) required companies to alter their board structure. The analyzed companies include 94 service industry firms within the S&P 1500. The board independence is defined as the number of independent directors divided by the total number of directors. The firm performance is expressed as Tobin’s Q and Return on Assets (ROA). Furthermore, board size, assets, leverage and firm size are considered as control variables. This study includes panel data and performs an Ordinary Least Square (OLS) regression analysis with fixed effects for ROA and random effects for Tobin’s Q. Although the influence of board independence on the firm performance is small, the findings show that the adoption of SOX has a positive influence on the firm performance of Tobin’s Q and ROA.
Keywords: Corporate governance, Board independence, Firm performance, Sarbanes Oxley Act
Statement of Originality
This document is written by student Marsha van Hattem who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document are 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.
§1 Introduction
In corporate governance board independence is an important issue, particularly when the board monitoring functions and associated agency problems of the infamous Enron and WorldCom scandals are considered. Partly as a result of these two scandals the need for independent boards gathered a lot of public attention. In 2001 the US energy company Enron was declared bankrupt because the board of directors failed to behave in a moral and ethical way (Zandstra, 2002, p. 16). Similarly the U.S. telecom firm, WorldCom announced its bankruptcy in 2002, largely due to fraud. These scandals increased the public interest in corporate governance matters and also contributed to the adoption of the Sarbanes-Oxley Act (SOX) in 2002, which was written by the U.S. politicians Paul Sarbanes and Michael Oxley. This act helps to ensure the alignment of interests between the directors and the shareholders (Nejadmalayeri et al., 2013). A requirement established in the act is that a firm needs a majority of independent board members (Dah et al., 2014, p. 97). Previous studies uncovered different conclusions about the effect of board independence on firm performance. Therefore, the purpose of this research, comparable to previous papers, is to clarify the relation between board independence and firm performance. The following research question will be examined: What has been the effect of the board independence on the firm performance of US-listed firms since the adoption of the Sarbanes-Oxley Act?
To examine the relation of board independence on firm performance both a theoretical and empirical exploration was undertaken. With the data obtained from The Wharton Research Data Services (WRDS) an Ordinary Least Square (OLS) regression analysis has been undertaken in STATA. The dependent variable of firm performance is expressed in two measures. Tobin’s Q ratio, which is a market-based measure, is a good guideline for board composition (Lefort & Urzúa, 2008, p. 616). Second, in order to verify the results an accounting-based measure will be used: the Return on Assets (ROA). The main independent variable ‘board independence’ is defined as the number of independent directors divided by the total number of directors. An interaction term for SOX with board independence is used to measure the effect of the adoption of the Sarbanes-Oxley Act in 2002. Furthermore, three control variables will be used to check the relationship of board independence and firm performance. The control measures are the board size, the leverage ratio and the firm size.
The article by Pathan and Faff (2013) focuses on the effect of board independence, gender diversity and board size on the firm performance of large US bank holding companies. Using this research for theoretical guidance, this paper focuses only on the potential effects of board independence on the firm performance in the period 1999 to 2005 and will contribute to
the field by using a sample of 94 US S&P 1500 firms within the services industry. Firms within the service industry include mainly Hotel, Engineering, Accounting, Research, Management firms, health services and business services. The service sector industry includes 7,762,981 companies and is therefore the largest division. So it would appear that these firms would make a suitable benchmark to measure the potential effects of board independent members on the firm performance for the US-listed firms. Moreover, in this thesis a defined explanation will be given about the question why there has been so much emphasis on board independence since the adoption of SOX? Furthermore, a clarification will be given about the way in which board independence is linked to the agency problem. In this thesis a regression analysis using fixed effects is performed to see the potential effect on ROA and a regression analysis using random effects is undertaken to see the potential effects on Tobin’s Q. It can be expected that the influence of board independence on firm performance is not high because many other factors are influencing the firm performance. However, it can be expected that the adoption of the SOX has a positive influence on the firm performance.
The remainder of the paper is structured as follows. In paragraph two the related literature is discussed. In addition to the research question, a hypothesis is formulated. In paragraph three a description of the data is given and the set up of the model used is specified, including a justification of the dependent, independent and control variables used. Next, in paragraph four the results are displayed. Finally, the discussion and the conclusion about the effects of board independence on firm performance are presented in paragraph five.
§2 Theoretical Framework
In the theoretical framework a description of the most relevant literature will be discussed. More specifically the expected effect of board independence on firm performance will be given. Finally, based on previous literature, the expected effect on firm performance is formulated in the hypothesis.
As previously mentioned, in this thesis two firm performance measures will be utilized: Tobin’s Q and the Return on Assets (ROA). Tobin’s Q is a marketing-based performance measure and the ROA is an accounting-based performance measure. Previous studies have discovered different effects of the influence of board independence on Tobin’s Q and the ROA. Beginning with the effects concerning Tobin’s Q ratio: the study by Reddy et al. (2008) who investigated small cap enterprises in New Zealand finds a positive effect of board independence on the firm performance expressed in Tobin’s Q ratio. Pathan and Faff (2013), on the contrary, found some evidence in their research of a negative relation of board
independence and bank performance. This bank performance is expressed in Tobin’s Q and ROA. The cause of this negative relationship can be attributed to a rise in asymmetric information. The authors mention that financial firms are located in an industrial branch were high information asymmetry originates. The study by Song and Van Hoof and Sungbeen Park (2017) reports an insignificant negative relationship between ROA and board independence in the restaurant industry. Moreover the writers report a negative relationship between Tobin’s Q and board independence.
In view of the potential relationship between board independence and firm performance, it is essential to give a defined explanation concerning board independence. Within an enterprise, the board structure is an important feature of good governance says the agency theory. This theory is based on the separation between ownership and control (Dalton et al., 1998). Since the managers of the company run the day-to-day business, they have and need authority and firm-specific knowledge. They are in a position to make their own decisions without considering the interest of the shareholders (Dalton et al., 1998, p. 270). This phenomenon is also known as the principal-agent problem. This agency problem may partly be solved by the introduction of boards, which are expected to have a monitoring and advisory function. The board also includes external directors that ought to monitor the managers and prevent, for example, fraud or criminal activities and therefore an independent board is expected to have a positive influence on the firm performance. However, if the number of external directors is too extensive this may have a negative influence on the guidance the board members could give the managers since the external directors are not familiar with the company’s objectives and do not have the necessary firm-specific knowledge (Fama & Jensen, 1983).
Several pieces of legislations have attempted to define board independence: First Sarbanes-Oxley (2002) has set specific standards affecting the board structure and requiring a board member to be independent within the audit committee (Coates, 2007, p111). The act gives a well-defined definition for an independent board member: “a person who does not accept any fee from the issuer (other than as the director) and is not an affiliated person of the issuer or any subsidiary” (Duchin et al., 2010, p. 198). Consequently, The New York Stock Exchange (NYSE) gave guidelines for the definition of a board as independent in 2003. Cited from Duchin et al. (2010) “a director is independent if the director does not have any material relationships within the enterprise”. Likewise in 2003, the NASDAQ changed this paraphrase of board independence to a person who “has no material relationship with the listed company (either directly or as a partner, shareholder or officer of an organization that has a relationship
with the company). Both NYSE and NASDAQ require a majority of independent board members. A director is not considered independent if he or she is a direct family member or was an employee of the company in the three years prior to the directorship or has a relationship with the auditor of the company. Moreover, a director is not considered independent if he or she works for a company that does business with the company he is the director of (Duchin et al., 2010).
Researchers have repeatedly investigated the influence of board independence and have arrived at different conclusions. Yeh (2013) concluded that board independence has a positive influence on the firm performance of hotel firms in Taiwan. Duchin et al. (2010) say that the effect of board independence on firm performance is influenced by the information costs of a company. The authors discover that when the costs of information are high, an increased number of external directors have a negative influence on firm performance. Therefore, when the costs are low, an increased number of external directors should have a positive influence on firm performance. Ahmed et al. (2017) investigated the influence of board independence in Malaysia on corporate social responsibility (CSR) reporting of publicly listed companies. They concluded that the efficiency of the board does not improve by introducing more independent members. Overall, the conclusions concerning the influence of board independence on firm performance are highly diverse. Therefore, it is not possible to already derive a conclusion concerning the effect on firm performance.
As previously mentioned this thesis focuses on the potential effects of board independence on firm performance since the adoption of the Sarbanes-Oxley (SOX) Act. Therefore, this paragraph will focus on the SOX. On the 25th of July 2002, Congress adopted the Sarbanes-Oxley Act. This act has been put into action to increase the quality of financial statements and to mainly concentrate on corporate disclosure and governance practices (Akhigbe & Martin, 2006). Moreover, the act was adopted to restore the audit practices of public enterprises in the U.S after the occurrence of a few corporate scandals in connection with fraud and unethical activities. Therefore, the SOX have included by-laws for “the auditor–firm relationship, auditor rotation, auditor provision of non-audit services, and corporate whistle-blowers” (Coates, 2007, p. 91). Not complying with SOX rules may result in lawsuits.
Literature on the subject has investigated the effects of the adoption of SOX on the firm performance. In one of these studies, the effect of the SOX on the monitoring abilities of firms compliant to the SOX expressed in Chief Executive Officer (CEO) turnover is analyzed (Dah et al., 2014). The writers concluded that the CEO turnover decreased when a compliant
firm, which already had a majority of independent directors, reduced the number of independent directors due to the SOX. Reversely, when these firms added more independent directors the CEO turnover increased because an increased number of independent board members will result in better supervision and an improved monitoring of CEO’s with an unsatisfactory performance. The paper by Akhigbe and Martin (2006) finds a positive influence on the portfolio returns of financial services firms since the adoption of the SOX. The papers of Akhigbe and Martin (2006) and Dah et al. (2014) distinguish between compliant and non-compliant firms. For the U.S. banking industry, the study by Pathan and Faff (2013) finds a positive effect on the bank performance in the post-SOX period. Based on the above findings a hypothesis concerning the effect of the SOX can be formulated: an increase in independent board members in the post-SOX period has a positive effect on the firm performance of US-listed firms within the service industry.
In this research three control measures are included: board size, leverage, and firm size. These variables are believed to have an influence on the two performance measurements. First, the determinant board size is discussed. Cheng (2008) studies a negative relation between board size and Tobin’s Q and ROA. Jensen (1993) states that this negative relationship can be explained by the idea that “a larger board has less communication, coordination and supervisory abilities and improves the possibilities for the CEO to influence and dominate the board’s decisions”. Cheng (2008) mentions that with larger boards the decision-making abilities of the board decline and agency problems arise. Hermalin and Weisbach (2001) also stated that board size is negatively related to firm performance. Additionally, they state that board size is connected to board independence. However, other researchers discover a positive relationship between the board size and firm performance. For example, Setia-Atmaja (2008) finds this positive relation between board size and Tobin’s Q ratio for Australian companies. Kalsie and Shrivastav (2016) also examine a positive relation between the board size and the two firm performance measurements. This positive relation can be contributed to the resource dependence theory (RDT), which says that a larger number of board members will improve possibilities to give guidance to the managers (Klein, 1998). Taking empirical literature into account, the expectation of the effects of board size on firm performance is positive.
The second control variable considered is leverage. Leverage is believed to have an influence on the performance of the enterprise. The Chief Executive officers use the degree of debt as a signaling method towards the shareholders. If the CEO’s expect favorable future firm returns they accept a higher debt level (Wanzenried, 2002). The agency problem and the
risk that the enterprise faces bankruptcy are consequences of higher debt levels and they imply a negative relation between leverage and firm performance (Fama & French, 2002). Holding a higher leverage ratio can reduce moral hazard because this increases the need for managers to operate more efficiently since leverage implies more risk for the enterprise (Jensen, 1986). Shahzad et al. (2015) examine a negative relationship between leverage and the ROA and a positive relationship between leverage and Tobin’s Q for the textile sector in Pakistan. The same findings are conducted in the study by Salim and Yadav (2012). So it would appear that the expectation considering the relationship between the leverage and Tobin’s Q is positive and the relationship between leverage and ROA is negative.
The last control variable is the firm size. It is believed that a larger firm size leads to a reduction in production costs by economies of scale (Miller, 1978). Beccalli et al. (2015) define economies of scale, as the decrease of unit costs, through the expansion of the volume produced. However, Lee (2011) reports that when the firm size is large diseconomies of scale could arise. The cause of these inefficiencies can be a decrease in motivation, resource misallocation, higher planning costs and communication difficulties between workers and divisions. The study by Bangun et al. (2017) finds a positive relationship between the firm size and the ROA for manufacturing companies in Indonesia. Campbell and Verbreke (1994) argue that firms in the service industry focus on increasing the firm size. The economies of scale in this industry will arise in the marketing segment since that sector focuses on providing services instead of products. Taking the previous empirical findings into account the expectation concerning the influence of firm size on the performance is therefore negative on Tobin’s Q and positive on ROA.
§3 Methodology and Data
This paragraph is separated into three sections. First, a summary is provided of the data collection. Second, a description of the methodology used to test the hypothesis is formulated. Finally, the different hypotheses are stated.
In order to find the influence of board independence on the firm performance data for the model have been obtained from The Wharton Research Data Services (WRDS). The financial variables have been collected from the Compustat-Capital IQ database of North America for fundamental annuals. The currency used is the dollar since the sample refers to US-listed firms. In order to collect data for the board characteristics as board independence and board size, the Institutional Shareholder Service (ISS) is consulted. In this research the sample consists of U.S. S&P 1500 companies within the Services industry with standard
industrial classification (SIC) codes 7000 to 8999. The time sample used is from 1999 to 2005. First, the financial variables are obtained from the Compustat-Capital IQ database. Observations that missed information about the variables required for the model were excluded from the sample. Next, the ISS database is utilized to collect the data for the board characteristics. Since the ISS database is much smaller than the Compustat-Capital IQ database many firms had to be excluded from the sample, as there was no information available about the necessary board characteristics. In the end 94 companies and 658 firm-year observations remained.
To test the hypothesis the following empirical model is defined:
𝐹𝑖𝑟𝑚 𝑝𝑒𝑟𝑓! = 𝛽! + 𝛽! 𝐵𝐼𝑁 + 𝛽!𝑆𝑂𝑋 + 𝛽! 𝐵𝐼𝑁𝑆𝑂𝑋 + 𝛽! 𝐵𝑆𝐼𝑍𝐸 + 𝛽! 𝐿𝐸𝑉
+ 𝛽!𝐹𝑆𝐼𝑍𝐸+ 𝜀!
To express the independent variable firm performance this study wishes to concentrate on two measures: Tobin’s Q and Return on Assets (ROA). First, Tobin’s Q is defined as the sum of the market value of equity plus the book value of liabilities divided by the book value of total assets. Second, the ROA is expressed in the net income divided by the book value of total assets. To measure the explanatory variable board independence the number of independent directors is divided by the total size of the board. Another board characteristic is the board size, which is defined as the total number of board members. The control variables in the regression are the leverage ratio, calculated as the total liabilities divided by total assets and the firm size measured as the ln from the book value of assets. In this thesis, an interaction variable for the Sarbanes-Oxley Act and the percentage of independent board members is included in the regression to investigate the effect of the adoption of the legislation. Therefore, the variable SOX is one for the years 2002 to 2005 and zero for the years prior to the SOX.
To test the model, two regression analyses in the program STATA are undertaken following a T-test to conclude if the variables of interest are significantly different from zero and have the predicted effects. The robust standard errors are calculated to exclude the existence of heteroscedasticity and to obtain unbiased standard errors. In this thesis the data investigated is strongly balanced panel data with cross-sectional time-series from 1999 to 2005 and could therefore be affected by time-invariant aspects and autocorrelation. To regress panel data two models could be used: fixed effects and random effects. With fixed effects the time-invariant effects are excluded. The characteristics of all firms included are influenced by one specific constant effect. With random effects there is not one specific constant effect but a
random intercept is included. In order to choose between the two methods, the Hausman test and the Breusch-Pagan Lagrange multiplier (LM) are performed. Therefore, the regressions include the OLS regression with fixed effects for ROA and random effects for Tobin’s Q. By evaluating the outcome of the output a prediction about the potential effects of board independence on firm performance since the adoption of the Sarbanes-Oxley Act can be made.
The first testable hypothesis is formulated concerning the potential effects of board independence on Tobin’s Q since the adoption of SOX.
H0: Board independence does not have an effect on Tobin’s Q in the post-SOX period. (𝛽! = 0)
H1: Board independence does have a positive effect on Tobin’s Q in the post-SOX period.
(𝛽! > 0)
The second testable hypothesis is formulated:
H0: Board independence does not have an effect on ROA in the post-SOX period. (𝛽! = 0)
H1: Board independence does have a positive effect on ROA in the post-SOX period.
§4 Regression Results
In this paragraph the outcomes are represented and discussed. First, the descriptive statistics and the correlation matrix for the regression model are represented. Thereafter the OLS regression models are given and discussed. Next, the results are compared with the expectations expressed in the theoretical framework. The regression models will be investigated based on the hypothesis made.
Table 1.
The table shows the Descriptive Statistics of the main variables. Including the number of observations, mean, standard deviation, the minimum and the maximum. The variable descriptions are represented in the Appendix: table 1.
Variable Obs. Mean. Std. Dev. Min. Max.
TOBINQ 658 1.678 1.632 0.083 16.264 ROA 658 0.050 0.102 -1.282 0.482 BIN 658 0.636 0.183 0 1 BSIZE 658 9.366 2.651 1 19 LEV 658 0.202 0.174 0.000 1.179 FSIZE 658 8.164 1.623 5.088 13.469
In table 1 an overview is given, concerning the descriptive statistics of the main variables. The table shows that the mean of board independence is 63,60%. This outcome is comparable to the study by Duchin et al. (2010). The writers find a mean of 60,36% of independent directors within a board. The mean of the board size (BSIZE) is 9.37 (9) with a minimum of 1 and a maximum of 19 board members. This observation is once again consistent with the study of Dah et al. (2014). The authors report a mean of 9.73. The mean of Tobin’s Q is 1.678 with a minimum of 0.083 and a maximum of 16.264 and the mean of the ROA is 0.050.
Graph 2.
Graph 2 represents a histogram of the percentage of compliant and non-compliant service sector industry firms over the different years in the sample. A firm is labeled compliant if it has a majority of 50 percent of independent directors within the board. The histogram shows that from 1999 to 2002 a large part of the service industry firms already were compliant to the SOX. The same conclusion had been drawn in the study of Dah et al. (2014). After the adoption of SOX in 2002 the degree of non-compliant firms decreased almost to zero.
Table 3.
Correlation Matrix: Pearson pairwise sample correlations. The variable descriptions are represented in the Appendix: table 1. The asterisks indicate the level of statistical significance * = 5%.
TOBINQ ROA BIN BSIZE LEV FSIZE
TOBINQ 1.000 ROA 0.539* 1.000 BIN -0.174* -0.047 1.000 BSIZE -0.195* -0.075* 0.023 1.000 LEV 0.079* -0.038 0.010 0.001 1.000 FSIZE -0.403* -0.153* 0.210* 0.448* -0.165* 1.000
Table 3 represents the Pearson pairwise sample correlation matrix of the regression variables. The correlation matrix indicates the relation between two variables. The asterisks represent the correlations with statistical significance at a 5% level. Next, The bold correlations are in more detail described. The correlation between board independence (BIN) and Tobin’s Q
0% 20% 40% 60% 80% 100% 120% 1999 2000 2001 2002 2003 2004 2005
Degree of compliance with SOX over the years within the service industry.
compliant non-‐compliant
(TOBINQ) is negatively and significantly related. Therefore, the conclusion can be drawn that board independence does not increase Tobin’s Q. The correlation between board independence (BIN) and ROA is not statistically significant. Furthermore, board independence (BIN) and board size (BSIZE) are positively and significantly related to the firm size (FSIZE). Board size (BSIZE) is negatively related to both performance measures and is statistically significant. Finally, there is no appearance of multicollinearity since there are no considerable correlations between the dependent and control variables.
In Appendix Table 2 the Hausman test is performed to choose between using fixed effects or random effects. The test concerning ROA is statistically significant and therefore is the fixed effects method the right regression model to use. However, the test for Tobin’s Q is rejected and therefore the random effects seem to be the appropriate model. In order to check if the random effects or the pooled OLS regression is the appropriate model to use, the Breusch-Pagan Lagrange multiplier (LM) (Appendix Table 3) is performed. The conclusion can be drawn that the regression with random effects is the right model for Tobin’s Q.
Table 4.
The table represents the OLS regression results of board independence on Tobin’s Q and ROA. The firms referred to have SIC codes: 7000-8999. The time sample is 1999-2005. The dummy variable SOX is included with value 1 for the years 2002-2005. The interaction term board independence and SOX (BINSOX) is included. Regression 1 indicates the regression results of board independence on Tobin’s Q using random effects. Regression 2 indicates the regression results using fixed effects for ROA. The numbers in parentheses are the robust standard errors. The variable descriptions are represented in Appendix: table 1.
The asterisks indicate the level of statistical significance *=10%, ** = 5%, *** =1%. Dependent Variable (1) TOBINQ RE (2) ROA FE BIN -0.530 (0.755) -0.005 (0.031) BINSOX 1.230* (0.731) 0.069** (0.038) SOX -1.148** (0.547) -0.067** (0.030) BSIZE 0.321 (0.254) 0.059** (0.025) LEV 0.138 (1.297) -0.180 (0.161) FSIZE -0.433*** (0.093) 0.007 (0.024) Constant 4.998*** (0.994) -0.085 (0.180)
Year FE Yes Yes
Firm FE Yes Yes
Obs 658 658
Adjusted –
R square 0.658 0.464
In table 4 random effects regression results for Tobin’s Q and fixed effects regression results for ROA are displayed, the values in parentheses are the robust standard errors. The regressions represent the entire sample period from 1999 till 2005 for the service industry sector with SIC codes 7000-8999. The regressions display potential effects of the adoption of
the Sarbanes-Oxley Act by the introduction of the interaction variable: BINSOX. Regression 1 presents the OLS regression of board independence on Tobin’s Q using random effects. Regression 1 shows that board independence (BIN) is negatively and statistically insignificantly related to Tobin’s Q. This result is consistent with the study by Pathan and Faff (2013). Regression 1 shows that the interaction term BINSOX is statistically significant and positively related to Tobin’s Q. So it would appear that the adoption of SOX has a positive effect on Tobin’s Q. Therefore, the negative effect of board independence on Tobin’s Q for the service industry was reduced in the post-SOX period from 2002 to 2005. This positive relationship ensures the acceptance of hypothesis 1. The SOX (SOX) is negative and statistically significant indicating that the total effect of the adoption of the Sarbanes-Oxley Act has a negative impact on Tobin’s Q. The board size (BSIZE) is positive and not statistically significant on Tobin’s Q. The firm size (FSIZE) is negatively related to Tobin’s Q and statistically significant at a 1% level. This negative relation can be explained by an increase in diseconomies of scale due to a larger firm size.
Regression 2 displays OLS regression results of the potential effects on ROA using fixed effects. In Regression 2 the board independence (BIN) is negatively and insignificantly related to ROA. These findings are consistent with the study by Song and Van Hoof and Sungbeen Park (2017). The variable (BINSOX) is positively related to ROA and significantly different from zero at a 5% level, causing an acceptance of hypothesis 2. This acceptance means that the adoption of SOX has a positive influence across time on the second firm performance measurement: ROA. This result is in line with the study by Pathan and Faff (2013). The board size (BSIZE) is positive statistically significant on ROA. This indicates that a larger board size increases ROA. This finding is consistent with the idea that a larger board size has more advisory powers than a smaller board. Furthermore, the firm size (FSIZE) is positively and insignificantly related to ROA. Additionally, in Appendix (Table 5) the regression results are displayed for Tobin’s Q pooled and fixed effects and the regression results for ROA pooled and random effects. In comparison with Appendix (Table 5), the adjusted R-square is a lot higher in the estimates for ROA with fixed effects 0.464 than using the pooled OLS regression model 0.024. Indicating that the model using fixed effects has more explanatory power than the pooled OLS regression model. The same conclusion can be drawn for Tobin’s Q.
§5 Conclusion
In this research different panel regression analyses are undertaken in STATA to conclude if board independence has a positive influence on the firm performance of 94 service industry firms in the U.S. since the adoption of the Sarbanes-Oxley Act. This Act requires a majority of independent members of the board and should reduce fraud and unethical activities after the occurrence of a few corporate scandals. This study shows that a lot of the service sector industry firms in the U.S. were already compliant with the SOX prior to the adoption of the legislation. The firm performance in this study is expressed in Tobin’s Q and ROA. The overall conclusion can be drawn that the hypotheses formulated are accepted, therefore the adoption of the SOX has a positive effect on Tobin’s Q and ROA for the service industry sector in the US. These findings imply that an increase in independent board members has improved the firm performance in the post-SOX period. However, the total effect of board independence on firm performance is negative. This negative relation could be explained by the fact that in the sample many boards have more than 50% independent board members. As a result, the advising abilities of the board can decrease because of a shortage of firm-specific knowledge. In this study the data utilized is panel data of the years 1999 to 2005. This panel data is used to give a good impression of the effects of the adoption of the SOX. The effects are not high, which results in the conclusion that the effects of SOX are not of really high influence on the firm performance. This presents the conclusion that there are many other variables that influence the firm performance.
The limitation exists that some coefficients are insignificant; this limitation could be explained by the size of the sample. Furthermore, the data shows that some service industry firms increased their number of independent board members. This increase in independent board members has not been undertaken to comply with the Sarbanes-Oxley Act. This suggests that an omitted variable is inflicting this increase in independent board members, which results in omitted-variable bias (Duchin et al., 2010). The literature uses different regression methods to show the potential effects of the SOX. These different regression methods could reduce the endogeneity problem as the GMM-estimation method or the difference-in-difference method. In future research, these different methods can be used for the service industry. Furthermore, in order to increase the external validity, the effects of the SOX could be investigated for more industry sectors and countries.
Preliminary Reference List
Ahmad, N.B.J., Rashid A., & Gow, J. (2017). Board Independence and Corporate Social Responsibility (CSR) Reporting in Malaysia. Australasian Accounting, Business and Finance Journal
11, 61-85.
Akhigbe, A., & Martin, A.D. (2006). Valuation impact of Sarbanes–Oxley: Evidence from disclosure and governance within the financial services industry. Journal of Banking & Finance 30, 989–1006. Bangun, N., Tjakrawala, F.X.K., Andani, K.W., & Santioso. L. (2017). The Effect of Financial Leverage, Employee Stock Ownership Program and Firm Size on Firm Performance of Companies Listed in Indonesia Stock Exchange. International Business and Accounting Research Journal 1, 82-98.
Beccalli, E., Anolli, M., & Borello, G. (2015). Are European banks too big? Evidence on economies of scale. Journal of Banking & Finance 58, 232–246.
Bhagat, S., & Black, B. (1999). The uncertain relationship between board composition and firm performance. Bus. Lawyer 54, 921–963.
Campbell, A.J., & Verbeke, A. (1994). The globalization of service multinationals. Long Range
Planning 27, 95–102.
Cheng, S. (2008). Board size and the variability of corporate performance. Journal of Financial
Economics 87, 157–176.
Coates, J. (2007). The goals and promise of the Sarbanes-Oxley Act. Journal of Economic
Perspectives 21, 91–116.
Dah, M.A., Frye, M.B., & Hurst, M. (2014). Board changes and CEO turnover: The unanticipated effects of the Sarbanes–Oxley Act. Journal of Banking & Finance 41, 97-108.
Dalton, D. R., Daily, C. M., Ellstrand, A. E., & Johnson, J.L. (1998). Meta-analytic reviews of board composition, leadership structure, and financial performance. Strategic Management Journal 19, 269– 290.
Financial Economics 96, 195–214.
Fama, E.F., & French, K.R. (2002). Testing trade-off and pecking order predictions about dividends and debt. Review of Financial Studies 15, 1-33.
Fama, E.F., & Jensen, M.C. (1983). Separation of ownership and control. Journal of Law and
Economics 26, 301–326.
Hermalin, B.E., & Weisbach, M.S. (2001). Boards of directors as an endogenously determined institution: A survey of the economic literature. Economic Policy Review 9, 7–26.
Jensen, M.C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American
Economic Review 76, 323-329.
Jensen, M. (1993). The modern industrial revolution, exit and failure of internal control systems.
Journal of Finance 48, 831-880.
Kalsie, A., & Shrivastav, S.M. (2016). Analysis of Board Size and Firm Performance: Evidence from NSE Companies Using Panel Data Approach. Indian Journal of Corporate Governance 9, 148–172. Klein, A. (1998). Firm performance and board committee structure. Journal of Law and Economics
41, 275-303.
Lee, S. (2011). Does an optimal firm size exist for publicly traded US hotels? Tourism Economics 17, 359–372.
Lefort, F., & Urzúa, F. (2008). Board independence, firm performance and ownership concentration: Evidence from Chile. Journal of Business Research 61, 615–622.
Liu, Y., Miletkov, M.K., Wei, Z., & Yang, T. (2015). Board independence and firm performance in China. Journal of Corporate Finance 30, 223–244.
Miller, E. (1978). The extent of economies of scale: the effects of firm size on labor productivity and wage rates. Southern Economics Journal 44, 470–487.
Nejadmalayeri, A., Nishikawa, T., & Rao, R.P. (2013). Sarbanes-Oxley Act and corporate credit spreads. Journal of Banking & Finance 37, 2991–3006.
Pathan, S., & Faff, R. (2013). Does board structure in banks really affect their performance? Journal
Reddy, K., Locke, S., Scrimgeour, F., & Gunasekarage, A. (2008). Corporate governance practices of small cap companies and their financial performance: an empirical study in New Zealand.
International Journal of Business Governance and Ethics 4, 51–78.
Salim, M., & Yadav, R. (2012). Capital Structure and Firm Performance: Evidence from Malaysian Listed Companies. Procedia-Social and Behavioral Sciences 65, 156-166.
Setia-Atmaja, L.Y. (2008). Does board size really matter? Evidence from Australia. Gadjah Mada
International Journal of Business 10, 331–352.
Shahzad, S.J.H., Ali, P., Ahmad, T., & Ali, S. (2015). Financial Leverage and Corporate Performance: Does Financial Crisis Owe an Explanation? Pakistan Journal of Statistics and Operation Research 11, 67-90.
Song, S., & Van Hoof and Sungbeen Park, H B. (2017). The impact of board composition
on firm performance in the restaurant industry. International Journal of Contemporary Hospitality
Management 29, 2121-2138.
Wanzenried, G. (2002). Signaling with Capital Structure Revisited. University of Bern, Department of
Economics, Switzerland.
Yeh, C. (2013). Board Governance and Tourism Firms’ Performance: The Case of Taiwan. Journal of
Quality Assurance in Hospitality & Tourism 14, 123–141.
Zandstra, G. (2002). Enron, board governance and moral failings, Corporate Governance. The
Appendix
Table 1: Variable Descriptions Variable Description
TOBINQ The sum of the market value of equity plus the book value of liabilities divided by the book value of total assets
ROA Net income divided by the book value of total assets
BIN The number of independent directors divided by the total number of directors.
BINSOX Interaction term between BIN and the dummy variable SOX
SOX Dummy variable equals one if the year is 2002–2005, otherwise zero BSIZE Total directors
LEV Total liabilities divided by the total assets
FSIZE Ln assets
Table 2: Hausman Test
The asterisks indicate the level of statistical significance *** =1%.
Dependent Variable Chi-Square Degree of freedom P-value
TOBINQ 2.85 6 0.8274
ROA 17.56 6 0.0074***
Table 3: Breusch-Pagan Lagrange multiplier (LM) for Tobin’s Q
This table indicates if to use random effects or pooled OLS regression. The asterisks indicate the level of statistical significance *** =1%.
Dependent Variable Chi-square P-value
Table 4
The table represents the OLS regression results of board independence on Tobin’s Q and ROA. The firms referred to have SIC codes: 7000-8999. The time sample is 1999-2005. Regression indicates the pooled OLS regression of board independence on ROA and Tobin’s Q. Regressions 2 and 4 include firm and year fixed effects. The numbers in parenthesis represent the robust standard errors. The variable descriptions are represented in the Appendix: table 1.
The asterisks indicates the level of statistical significance *=10%, ** = 5%, *** =1%. Dependent Variable (1) TOBINQ (2) TOBINQ FE (3) ROA (4) ROA FE BIN -0.865* (0.443) -0.237 (0.420) -0.007 (0.016) -0.005 (0.032) BSIZE -0.018 (0.019) 0.234 (0.202) -0.000 (0.002) 0.049** (0.022) LEV 0.181 (0.504) 0.260 (1.124) -0.037 (0.051) -0.173 (0.206) FSIZE -0.369*** (0.040) -0.813*** (0.817) -0.010*** (0.004) -0.012 (0.020) Constant 5.369*** (0.515) 7.900*** (1.520) 0.145*** (0.020) 0.074 (0.142)
Year FE … Yes … Yes
Firm FE … Yes … Yes
Obs 658 658 658 658
Adjusted
Table 5.
The table represents the OLS regression results of board independence on Tobin’s Q and ROA. The firms referred to have SIC codes: 7000-8999. The time sample is 1999-2005. The dummy variable SOX is included with value 1 for the years 2002-2005. The interaction term board independence and SOX (BINSOX) is included. Regression 1 indicates the pooled OLS regression results of board independence on Tobin’s Q. Regression 2 indicates the regression results of board independence on Tobin’s Q using fixed effects. Regression 3 indicates the pooled OLS regression results of board independence on ROA. Regression 4 indicates the regression results of board independence on ROA using random effects. The numbers in parentheses are the robust standard errors. The variable descriptions are represented in the Appendix: table 1.
The asterisks indicate the level of statistical significance *=10%, ** = 5%, *** =1%. Dependent Variable (1) TOBINQ (2) TOBINQ FE (3) ROA (4) ROA RE BIN -1.146* (0.690) -0.292 (0.500) -0.026 (0.020) -0.003 (0.026) BINSOX 1.093* (0.658) 1.248** (0.513) 0.059* (0.033) 0.055** (0.055) SOX -1.008* (0.547) -1.157** (0.377) -0.049** (0.024) -0.050** (0.023) BSIZE -0.014 (0.019) 0.397** (0.05) 0.000 (0.002) 0.037 (0.025) LEV 0.137 (0.502) 0.139 (1.117) -0.039 (0.051) -0.111 (0.137) FSIZE -0.369*** (0.041) -0.517*** (0.201) -0.010*** (0.004) -0.012 (0.007) Constant 5.680*** (0.661) 5.365*** (1.611) 0.161*** (0.023) 0.095** (0.05)
Year FE … Yes … Yes
Firm FE … Yes … Yes
Obs 658 658 658 658
Adjusted –
