DOES BOARD INDEPENDENCE
MATTER DURING AND AFTER A
CRISIS PERIOD
By Rob Spoor
MSc Finance thesis
Supervisor: dr. R.O.S. Zaal
University of Groningen
February 2021
Abstract:
This study investigates the impact of the corporate board characteristics on firm performance during the 2020 corona crisis and the subsequent rebound period. Using the buy and hold abnormal returns as our main measure of firm performance, this study finds evidence of a negative relationship between board independence and firm performance
Introduction
There is no doubt that the recent corona crisis of 2020 had a significant impact on financial markets. The crisis hit relatively unexpected, causing a period of severe distress for
companies. This period however provides an interesting opportunity to study the effects of different characteristics of corporate boards on a firm performance.
The reason why this study is focused on corporate boards is because corporate boards play a key role within the corporate governance mechanism in a company. The board of directors should advice the management of the company and protect the interests of the
shareholders of the company. These two tasks are also especially important in a time of crisis, where decisions sometimes must be made on a short notice. The main hypothesis this study will test is whether it is beneficial for firm performance to have a more dependent or more independent board during and after a period of crisis.
One of the reasons why it could be beneficial to have a less dependent board of directors in times of a crisis, according to Bhagat & Black (2001), is that inside directors on the one hand are conflicted, but well informed on the firm, while outside directors on the other hand are not conflicted but are relatively ignorant about what is happening inside the company. The authors argue that when things go wrong, inside directors will be more likely to make the right decision for the firm since they possess more knowledge on the company’s operations. After the downfall period of the crisis, a strong rebound of the market can be observed. Therefore, besides investigating the impact of corporate governance during the crisis period, therefore, this paper also investigates whether board independence has any effect in the subsequent rebound period.
In line with the research of Francis, Hasan, and Wu (2012) on the impact of corporate boards on firm performance during the financial crisis of 2007/2008, this study will also be focused on the impact of several other characteristics of the board of directors. Although lots of research has been done to investigate the relationship between corporate boards and firm performance, this paper differentiates itself from prior research by using a systematic exogenous shock in the form of the 2020 corona crisis. This exogenous event helps to better avoid endogeneity issues and as a result provide more evident results on the effect of corporate boards on firm performance.
experience different returns compared to firms with a less independent board in the rebound period following the crisis.
Furthermore, this paper will investigate differences between industries when it comes to the relationship between board independence and firm stock performance.
The remainder of this paper is structured as follows. Chapter 1 will give an overview of the existing literature on the subject. Chapter 2 will describe the research method &
methodology that will be used to test the various hypotheses. Chapter 3 will give an
overview of the data and the summary statistics. Chapter 4 reports the results of the model. Finally, chapter 5 will provide the conclusion and discussion of the results.
1. Literature review
Previous research on the effect of corporate governance on firm performance (e.g. Johnson, Boone, Breach, & Friedman (2000) and Mitton (2002)) provides evidence that corporate governance is indeed an important factor in determining a firm’s performance during a crisis period.
Johnson, Boone, Breach, & Friedman (2000) investigated the effect of minority shareholder protection in twenty-five emerging markets during the Asian crisis of 1997 and 1998. They find that the effectiveness of minority shareholder protection explains more of the variation in stock market performance during this crisis.
Similarly, La Porta, Lopez‐de‐Silanes, Shleifer, & Vishny (2000) observe a higher Tobin’s q (a measure of firm value) in countries with better shareholder protection compared to
countries where shareholder protection is weaker.
Previous literature describes mainly two reasons why corporate governance is of the essence in times of a crisis. Firstly, the incentive of controlling shareholders to expropriate minority shareholders increases in times of a crisis, since the expected return on investment
decreases. Secondly, any preexisting weaknesses in the corporate governance of a firm will become more visible. During a crisis, the quality of corporate governance is likely to attract more scrutiny. As a result, any preexisting weaknesses might be revealed, leading to a possible decline in the firm’s stock price.
Baek, Kang & Park (2004) investigate the same hypothesis using data from the Korean
financial crisis of 1997. Their results state that the change of firm performance during a crisis is a function of firm-level differences in corporate governance measures.
Francis, Hasan & Wu (2012) conduct a similar research but use data from the financial crisis of 2007/2008. They find a positive and significant relationship between strong independence of directors and firm performance. Furthermore, they find a positive and significant
relationship between having outside financial experts and firm performance. Overall, they conclude that firm-level differences play an important role in determining a firm’s
performance in a time of crisis.
Furthermore, Francis, Hasan & Wu (2012) also find evidence for significant differences between industries when it comes to the effect of board independence on firm
performance. Therefore, this paper will also investigate if there are any difference in results for certain industries. Considering the 2020 corona crisis, it could be the case that certain industries suffer more from this crisis compared to other industries.
Previous research outlines two main categories of measures for firm performance. Cochran & Wood (1984) outline these two main categories. First there is investor returns with the underlying idea that returns should be measured from the perspective of the shareholders. Second there is accounting returns, a primary method for firms themselves to measure their financial performance. The basic idea behind this is to focus on how a firm’s earnings
respond to a crisis period and a following rebound period. This paper will use an investor returns measure as well as an accounting measure to evaluate firm performance.
Baysinger & Butler (1985) empirically analyzed the effect between board composition in terms of independency on firm performance during the 1970’s. They find that firms with a higher proportion of independent directors early in the decade showed, on average, superior performance records later in the decade. Therefore, the authors argue that there might be a presence of a lagged effect in the relationship between board independence and firm performance.
well informed on the firm, while outside directors on the other hand are not conflicted but are relatively ignorant about what is happening inside the company. The authors argue that when things go wrong, inside directors will be more likely to make the right decision for the firm since they possess more knowledge on the company’s operations. This paper will investigate whether this line of reasoning also holds during the volatile period of the corona crisis.
Moreover, Noe & Rebello (1996) find evidence suggesting that social ties between the CEO and other board members could increase board effectiveness. Therefore, inside directors could indirectly have a positive effect on firm performance.
Adams and Ferreira (2009) find that there is an effect of the gender diversity of board on firm performance. The authors find a negative relation between gender diversity of the board and firm performance, and therefore argue that well-governed firms which introduce a gender quota could see a reduction in firm performance. This paper will investigate the effect of gender diversity of boards on firm performance in the period of the 2020 corona crisis. The results will then be compared to the results found by Adams & Ferreira.
Apart from board independence, there could be other board characteristics which have an impact on firm performance. Frist of all, Yermack (1996) provides evidence that small boards of directors are more effective than large boards of directors. The author argues that a large board of directors might lead to poor communication and decision making. Since
communication and decision making are especially important in a volatile situation like the corona crisis, this paper will investigate whether there is a relationship between board size and firm performance.
Furthermore, other board characteristics such as board duality (Coles, McWilliams, & Sen, 2001; Dalton, Daily, Ellstrand, & Johnson, 1998), director age (Shivdasani & Yermack, 1999), and director tenure (Vafeas, 2003) have been studied in the past. Therefore, this paper will also provide additional tests to investigate if the results of these studies also hold in the corona crisis.
• The main hypothesis to be tested in this paper is that board independence affects firm performance during the 2020 corona crisis period and during the subsequent rebound period.
• The second hypothesis to be tested in this paper is that several board characteristics, like the ones mentioned above, affect firm performance during the 2020 corona crisis or in the subsequent rebound period.
• The third hypothesis to be tested in this paper is that there are differences between industries in terms of the relationship between board independence and firm performance during the 2020 corona crisis and/or the subsequent rebound period.
2. Research method & Methodology
2.1 Sample selection
The sample used in this paper consists of nonfinancial Standard and Poor 500 companies. First, financial companies with an SIC code between 6000 and 6999 are filtered out of the sample. Financial companies are filtered out, because their business model is highly different from other companies. According to Fama & French (1992), the high leverage that is normal for financial companies does not have the same meaning as for non-financial firms, where high leverage tends to indicate distress. Afterwards, some companies for which the appropriate data was not available were also removed from the sample. The final sample size used in the analysis consists of 375 non-financial S&P500 companies.
The accounting data of the firms within the sample was collected from the Compustat database. CEO information was collected from the ExecuComp database. Information on board characteristics was collected from the BoardEx database. All financial and board characteristic data dare measured at the end of the 2019 fiscal year if not specified otherwise.
2.2 Defining the crisis and rebound periods
a rebound period of the index. Figure 2 also suggests a case of a crisis period and a rebound period, the figure shows the number of firms with positive and negative cumulative returns during both periods. During the crisis period (1st January 2020 – 23rd March 2020) only nine
firms of our sample managed to report positive cumulative returns. During the rebound period (24th March 2020 – 30th June 2020) all the firms within our sample report positive
cumulative returns. Overall, these figures suggest that we can make the differentiation between the crisis period and the rebound period.
Figure 1
S&P500 index during the specified crisis and rebound periods.
Source: Financial Times Figure 2
Comparison of firm performance during both the crisis period (1) and the rebound period (2), by observing whether firms have either positive or negative cumulative returns.
9 375 366 0 0 100 200 300 400 1 2 Fi rms Period
Cumulative returns in the crisis period
(1) and rebound period (2)
As mentioned in the literature review, multiple starting and ending dates of both the crises and rebound periods will be used to test the robustness of the model. By analyzing the S&P500 index in figure 1, the following periods have been selected.
Crisis period 1: 1st January 2020 until 23rd March 2020.
Crisis period 2: 20th February 2020 until 23rd March 2020.
Crisis period 3: 4th March 2020 until 23rd March 2020.
Rebound period 1: 24th March 2020 until 30th June 2020.
Rebound period 2: 24th March 2020 until 29th April 2020.
Rebound period 3: 24th March 2020 until 26th March 2020.
All crisis periods end on the 23rd of March 2020, the lowest point on the S&P500 index during
the measuring period. Following the same reasoning, all rebound periods start from the 24th
of March 2020, the moment the S&P500 index started increasing again. Crisis period 1 represents the overall crisis period, starting on the first day of the year and ending on the lowest point on the index on the 23rd of March of the year 2020. In this period, the market is
stable at first, but slowly starts falling off. The second crisis period represents the period in which the market is decreasing, starting at the 20th of February, and ending on the lowest
point on the index on the 23rd of march, both in the year 2020. The third crisis period
represents a shorter period in which the index is plummeting the most. This period starts on the 4th of March 2020 and like the other crisis periods ends on the 23rd of March 2020.
All the rebound periods start on the 24th of March 2020, the day after the lowest point of the
market has been reached. The first rebound period ends on the 30th of June 2020 and
represents are longer-term rebound period. The second rebound period ends 29th of April
2020 and represents a shorter-term rebound period. The third rebound period ends on the 26th of March, only two days after the start of this period. This period captures the
2.3 variable description
2.3.1 Firm performance measure
As mentioned in the literature review, this paper will use one investor returns measure of firm performance, and one accounting measure of firm performance. For our investor returns measure of firm performance, we chose the buy and hold abnormal return. This variable captures the investor returns during defined periods. Since this paper uses different periods to estimate firm performance, the variable of buy and hold returns fits well into this model. As our accounting measure for firm performance, we used the return on assets during the first quarter of 2020 to capture the effects of the crisis period, and the return on assets during the second quarter of 2020 to capture the effects of the rebound period. According to Cochran & Wood (1984), the most common accounting measures used to evaluate firm performance are earnings per share (EPS) or price/earning (P/E) ratios. Since both of these variables require earnings calculations based on the previous year, both variables did not fit in with the model used in this paper. Therefore, the ROA during the first two quarters of 2020 were chosen as a measure of firm performance, this variable captures the firm performance exactly during the crisis and rebound periods and therefore fits the model.
Similar to Francis, Hasan & Wu (2012), the buy and hold abnormal return (BHAR) will be used as the investor returns measure of firm performance during the crisis and rebound periods. The BHAR will be computed as follows:
𝐵𝐻𝐴𝑅(𝑐𝑟𝑖𝑠𝑖𝑠)
= (1 + 𝑅𝑖 𝑑𝑎𝑦 1)(1 + 𝑅𝑖 𝑑𝑎𝑦 2)(1 + 𝑅𝑖 𝑑𝑎𝑦 3) … (1 + 𝑅𝑖 𝑑𝑎𝑦 𝑁) − (1 + 𝑅𝑚 𝑑𝑎𝑦 1)(1 + 𝑅𝑚 𝑑𝑎𝑦 2)(1 + 𝑅𝑚 𝑑𝑎𝑦 3) … (1 + 𝑅𝑚 𝑑𝑎𝑦 𝑁)
(1) Where Ri is the return of stock i at time t; and Rm is the return of the market at time t.
Similarly:
𝐵𝐻𝐴𝑅(𝑟𝑒𝑏𝑜𝑢𝑛𝑑)
= (1 + 𝑅𝑖 𝑑𝑎𝑦 1)(1 + 𝑅𝑖 𝑑𝑎𝑦 2)(1 + 𝑅𝑖 𝑑𝑎𝑦 3) … (1 + 𝑅𝑖 𝑑𝑎𝑦 𝑁) − (1 + 𝑅𝑚 𝑑𝑎𝑦 1)(1 + 𝑅𝑚 𝑑𝑎𝑦 2)(1 + 𝑅𝑚 𝑑𝑎𝑦 3) … (1 + 𝑅𝑚 𝑑𝑎𝑦 𝑁)
As mentioned in chapter 2.2, the BHAR will also be computed using different starting and ending dates of both the crisis and the rebound period.
The second measure of firm performance to be used in this paper is the return on assets of a firm over the first and second quarter of 2020. This variable is calculated by dividing the net income in the specified quarter by the total assets in the specified quarter. Cochran & Wood (1984) argue that one defect of using accounting measures to evaluate firm performance is that firms themselves can influence these variables to their own interest, which might lead to possible biases in the model.
2.3.2 Measure of board characteristics
Following multiple prior studies on the topic of board independence, an independent
director is defined as a board member who is not related to the company in terms of being a past or present employee of the firm, and who is not affiliated with the firm through family or business ties. The ratio of independent directors to board size will be used as the measure of board independence. Apart from board independence, previous literature also finds relationships between firm performance and other board characteristics. These
characteristics include board size (Coles et al. (2008); Yermack (1996), gender diversity (Adams & Ferreira (2009)), board duality (Coles, McWilliams, & Sen, 2001; Dalton, Daily, Ellstrand, & Johnson, 1998), director age (Shivdasani & Yermack, 1999), and director tenure (Vafeas, 2003).
This paper will also provide additional test on the relationship between the abovementioned board characteristics and firm performance during the crisis and the rebound periods of the 2020 corona crisis.
2.3.3 Control variables
year 2019, measured as the ratio of net income to total firm assets. The fourth control variable, Beta, is the five-year monthly Beta of the firm. The sixth control variable is CEO tenure, measured by how many years the CEO of the firm has been active for. This variable is used to control for CEO entrenchment. The last control variable is an industry control
variable which consist of the firm’s primary one-digit SIC code. This variable is used to control for difference between industries.
2.4 Multiple regression model
To investigate the effect of the board characteristics on firm stock price performance, a multiple regression model will be used in which the multiple BHAR’s and ROA’s will act as the dependent variables, and the independent variables will consist of the board
characteristics and multiple control variables.
Using the above overviewed data, we will estimate the following multiple regression model to investigate the effect of board independence on firm stock price performance during a crisis.
𝐵𝐻𝐴𝑅(𝑐𝑟𝑖𝑠𝑖𝑠) = 𝛼 + 𝛽1(𝐵𝑜𝑎𝑟𝑑 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒) + 𝛽2(𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒) + 𝛽3(𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒) + 𝛽4(𝑅𝑂𝐴) + 𝛽5(5𝑦 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝐵𝑒𝑡𝑎) + 𝛽6(𝐶𝐸𝑂 𝑇𝑒𝑛𝑢𝑟𝑒) + 𝛽7(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝜀
(3) Where the variables represent the abovementioned characteristics.
The same model will be used to investigate the effect of board independence on firm stock price performance in a rebound period.
𝐵𝐻𝐴𝑅(𝑟𝑒𝑏𝑜𝑢𝑛𝑑)
= 𝛼 + 𝛽1(𝐵𝑜𝑎𝑟𝑑 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒) + 𝛽2(𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒) + 𝛽3(𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒) + 𝛽4(𝑅𝑂𝐴) + 𝛽5(5𝑦 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝐵𝑒𝑡𝑎) + 𝛽6(𝐶𝐸𝑂 𝑇𝑒𝑛𝑢𝑟𝑒) + 𝛽7(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝜀
(4) Where the variables represent the abovementioned characteristics.
𝑅𝑂𝐴(𝑄1/2020) = 𝛼 + 𝛽1(𝐵𝑜𝑎𝑟𝑑 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒) + 𝛽2(𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒) + 𝛽3(𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒) + 𝛽4(𝑅𝑂𝐴) + 𝛽5(5𝑦 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝐵𝑒𝑡𝑎) + 𝛽6(𝐶𝐸𝑂 𝑇𝑒𝑛𝑢𝑟𝑒) + 𝛽7(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝜀
(5) And in the second quarter:
𝑅𝑂𝐴(𝑄2/2020) = 𝛼 + 𝛽1(𝐵𝑜𝑎𝑟𝑑 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑐𝑒) + 𝛽2(𝐹𝑖𝑟𝑚 𝑆𝑖𝑧𝑒) + 𝛽3(𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒) + 𝛽4(𝑅𝑂𝐴) + 𝛽5(5𝑦 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝐵𝑒𝑡𝑎) + 𝛽6(𝐶𝐸𝑂 𝑇𝑒𝑛𝑢𝑟𝑒) + 𝛽7(𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 𝐷𝑢𝑚𝑚𝑖𝑒𝑠) + 𝜀
(6) Where the variables represent the abovementioned characteristics.
2.5 Industry differences
In line with the findings of Francis, Hasan & Wu (2012), the effects of a crisis might vary among different industries, some industries might suffer more from a specific crisis compared to other industries. Furthermore, board characteristics might differ among industries. Therefore, the data sample will be split up into ten different industry groups based on the Fama French ten-industry classifications, to further investigate the possible differences among industries.
The Fama French ten-industry classifications consist of the following categories: 1. Consumer Nondurables
2. Consumer Durables 3. Manufacturing 4. Energy
5. Business Equipment
6. Telephone and Television Transmission 7. Wholesale, Retail and Services
3. Data & Summary statistics
Table 1 provides statistics of firm performance during the crisis periods in terms of means, standard deviations, the minimum reported value, and the maximum reported value. From the table we observe that firm performance during all three crisis periods is poor. This is in line with what we would expect. In all the crisis periods, we observe a negative average BHAR. Furthermore, we observe negative average cumulative returns in all crisis periods, most severe in crisis period 2 nearing an average negative cumulative return of -50%. Furthermore, we observe a large difference between the return on assets in the fourth quarter of 2019 and the return on assets in the first two quarters in 2020, this difference shows the magnitude of the effect of the crisis on the return on assets of the firms.
Table 1
Firm performance in crisis periods
This table shows the performance of the stock price of all the firms used in the analysis during the crisis periods. BHAR measures the buy and hold abnormal return of the firm’s stock price during the specified crisis period. Cum. returns shows the sum of all returns of the firm’s stock price during the specified crisis period. Return volatility is the standard deviation of the returns of the firm’s stock price over the specified crisis period.
Variable Obs. Mean Std. Min. Max.
BHAR (crisis 1) 375 -0.050 0.185 -0.598 0.512
Cum. returns (crisis 1) 375 -0.478 0.317 -1.806 0.199 Return volatility (crisis 1) 375 0.044 0.014 0.023 0.133
BHAR (crisis 2) 375 -0.042 0.161 -0.579 0.465
Cum. returns (crisis 2) 375 -0.498 0.274 -1.888 0.128 Return volatility (crisis 2) 375 0.064 0.021 0.028 0.190
BHAR (crisis 3) 375 -0.051 0.156 -0.654 0.399
Cum. returns (crisis 3) 375 -0.372 0.230 -1.758 0.137 Return volatility (crisis 3) 375 0.079 0.028 0.030 0.235 ROA (Q4-2019) ROA (Q1-2020) ROA (Q2-2020) 375 375 375 0.038 0.008 0.008 0.303 0.059 0.029 -0.591 -0.681 -0.118 1.305 0.383 0.140
Table 2
Firm performance in rebound periods
This table shows the performance of the stock price of all the firms used in the analysis during the rebound periods. BHAR measures the buy and hold abnormal return of the firm’s stock price during the specified crisis period. Cum. returns shows the sum of all returns of the firm’s stock price during the specified crisis period. Return volatility is the standard deviation of the returns of the firm’s stock price over the specified crisis period.
Variable Obs. Mean Std. Min. Max.
BHAR (rebound 1) 375 0.003 0.216 -0.671 1.107 Cum. returns (rebound 1) 375 0.350 0.166 -0.142 1.145 Return volatility (rebound 1) 375 0.037 0.014 0.016 0.114 BHAR (rebound 2) 375 0.041 0.174 -0.411 1.336 Cum. returns (rebound 2) 375 0.314 0.134 -0.085 1.111 Return volatility (rebound 2) 375 0.047 0.016 0.011 0.136 BHAR (rebound 3) 375 0.028 0.098 -0.240 0.466 Cum. returns (rebound 3) 375 0.187 0.088 -0.069 0.536 Return volatility (rebound 3) 375 0.056 0.031 0.003 0.218
Table 3 and table 4 provide a summary of statistics of the board characteristics and firm financials, respectively. The average board size in the sample is 10.675. The average board independence is 0.880, meaning that on average 88% percent of the board members of the firms in our sample are independent from the company. An independent board member is defined as a member of the board without any affiliation with the company in terms of employment, business, or family ties. The average duality dummy variable tells us that within our sample 43.2% of firms have a CEO who is also the chairman of the board of directors. On average the CEO’s in our sample have been in place for about 7.7 years. The average time individual directors have been on the board for is roughly 8 years. The average gender ratio we observe is 0.731, meaning that on average 73.1% of board members are male, and 26.9% of board members are female.
Table 3
Summary statistics board characteristics
of years the current CEO of the firm has been in place for. Time on the board is the average time (in years) for which the directors have been active on the board for. Director age is the standard deviation of the average age of all the directors on the board of the firm. Gender diversity measures the percentage of male directors on the board to female directors. All the firm and board variables are measured at the end of fiscal year 2019.
Variables Obs. Mean Std. Min. Max.
Board size 375 10.675 1.843 5 16
Board independence 375 0.880 0.053 0.625 0.933
Duality (dummy) 375 0.432 0.495 0 1
CEO tenure (in years) 375 7.710 7.271 1 41
Time on the board (in years) 375 8.041 3.369 0.5 31.163 Director age (std.) 375 6.899 2.008 1.8 14.5 Gender diversity 375 0.731 0.090 0.455 1 Table 4 Firm statistics
This table provides an overview of the financial situation of the firms included in the analysis. Assets is the total assets of a firm. Log Assets is the natural logarithm of the total assets of the firm, this variable will be used in the regression model to control for firm size. Liabilities is the total liabilities of a firm. Leverage is the ratio of total liabilities to total assets. Net Income measures the Net Income of the firm. ROA is the ratio of Net Income to total assets. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. All the firm and board variables are measured at the end of fiscal year 2019.
Variable Obs. Mean Std. Min. Max.
4. Results
4.1 Board independence and firm performance
The results in table 5 show a negative coefficient for board independence in all three crisis periods. We can observe that the coefficients are all negative, however insignificant. Meaning that we cannot prove that the coefficients are statistically different from zero. However, if we assume that there is a negative effect like the results of Bhagat & Black (2001) show, and we perform a one-sided t-test, we find that the coefficient of the third crisis period is statistically less than zero at the 10% level. While the coefficients of the first and second crisis periods are still insignificant. Interestingly, the negative coefficients indicate that a higher independence percentage of the board decreases the buy and hold abnormal return during a crisis period. The -0.164 Board Independence coefficient of the third crisis period indicates that an increase of a board’s independence from the 25th
percentile to the 75th percentile leads to a decrease of the buy and hold abnormal return of
about 8.2 percentage points.
Furthermore, the results in table 5 show a positive coefficient significant at the 1% level for the firms ROA across the three crisis periods. The economic meaning of this finding is that firms which had a higher ROA over the year 2019 experienced a higher buy and hold abnormal return during the 2020 corona crisis. The coefficients of beta are all negative and statistically significant at the 1% level. This tells us that there is a negative relationship between the 5-year monthly beta of the firm and their buy and hold abnormal return during the 2020 corona crisis. This is in line with regular finance theory which suggest that a lower beta indicates a relatively less volatile stock, which would on average outperform a more volatile stock in a time of crisis.
Table 5
This table provides the results of an OLS regression for the effect of board independence on firm stock performance during the crisis periods. The BHAR of the respective crisis period is the dependent variable, which is the buy and hold abnormal return in the specified period. Board
classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
BHAR (crisis 1) BHAR (crisis 2) BHAR (crisis 3) Board Independence -0.119 (1.09) -0.087 (0.87) -0.164 (1.59) Log Assets -0.009 (1.51) 0.007 (1.21) -0.004 (0.77) Leverage -0.036 (0.98) -0.027 (0.72) -0.053 (2.00) ** ROA 0.387 (3.97) *** 0.400 (4.72) *** 0.446 (5.35) *** Beta -0.228 (14.29) *** -0.195 (15.06) *** -0.176 (15.35) *** CEO Tenure 0.001 (0.27) -0.001 (0.36) -0.001 (0.73) Industry Dummy -0.001 (0.44) 0.004 (2.31) ** -0.003 (1.76) * R2 0.61 0.61 0.57 Adjusted R2 0.60 0.60 0.56 N 375 375 375
In table 6 we observe that the opposite of the crisis periods holds for rebound periods. The coefficients for board independence in rebound periods are positive. Furthermore, the coefficients in the first and second rebound period are statistically significant at the 10% and 5% level, respectively. In these periods an increase in board independence by 10 percentage points results in an increase of the buy and hold abnormal return of 2.95 percentage points in the first rebound period, and 3.30 percentage points in the second rebound period. If we apply the same principle of a one-sided test as in the previous section, the first, second, and third rebound periods would all have a positive and statistically significant coefficient at the 5%, 1% and 1% level, respectively.
between beta and the buy and hold abnormal in the rebound periods. As the market moves upwards, one would expect that on average firms with a higher beta will outperform firms with a lower beta, which is exactly what we observe from the beta coefficients in table 6. From these observations we can thus conclude that in a period of crisis it is more beneficial for the firm’s buy and hold abnormal return to have a more dependent board, and in
rebound periods it is more beneficial for the firm’s buy and hold abnormal return to have an independent board. One of the reasons behind this observation might be that during a time of crisis, it is beneficial to have a higher percentage of board members who are also involved in the day-to-day business of the firm. These board members might have a better view on what is happening within the company during a period of crisis compared to outside directors on the board.
These results are in line with the findings of Bhagat & Black (2001), they also find evidence to support the hypothesis that in a time of crisis, a less independent board increases firm performance. The authors argue that when things go wrong, inside directors will be more likely to make the right decision for the firm since they possess more knowledge on the company’s operations.
Table 6
This table provides the results of an OLS regression for the effect of board independence on firm stock performance during the rebound periods. The BHAR of the respective crisis period is the dependent variable, which is the buy and hold abnormal return in the specified period. Board
Independence is the ratio of independent board directors to board size. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Industry Dummy is the respective Fama French ten-industry
classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
ROA 0.456 (2.45) ** 0.033 (0.22) -0.073 (1.02) Beta 0.164 (5.98) *** 0.174 (7.50) *** 0.067 (6.44) *** CEO Tenure -0.001 (0.12) -0.001 (1.36) -0.001 (0.36) Industry Dummy 0.001 (0.01) 0.001 (0.30) 0.004 (2.53) ** R2 0.22 0.36 0.19 Adjusted R2 0.20 0.35 0.17 N 375 375 375
Table 7 presents the results of the regression with the firm’s ROA in the first and second quarter of 2020 as the dependent variables. This model finds no significant evidence of a relation between board independence and the firm’s ROA in the first or second quarter of 2020. In the first quarter of 2020 we do however find the same effect compared to the crisis periods with respect to the firm’s ROA over 2019 and beta. Firms with a higher ROA over 2019 on average also observed a higher ROA during the first quarter of 2020. Firms with a lower beta outperformed firms with a higher beta in terms of ROA during the first quarter of 2020, similar to the results found in table 5.
Table 7
Leverage 0.022 (1.44) 0.004 (0.85) ROA 0.314 (7.12) *** 0.139 (4.74) *** Beta -0.024 (3.52) *** -0.015 (5.40) *** CEO Tenure -0.001 (0.79) -0.001 (1.19) Industry Dummy 0.001 (0.53) -0.001 (0.46) R2 0.30 0.34 Adjusted R2 0.29 0.33 N 375 375
4.2 Other board characteristics and firm performance
Table 8 presents the results of the OLS regression of firm performance on several board characteristics. The coefficient of board size is negative and significant at the 1% level, this indicates that during the crisis period a smaller board size had a positive impact on the BHAR of the firm. Furthermore, the coefficient of director age is positive and significant at the 10% level, indicating that during the crisis period having a higher average age of directors had a positive impact on firm performance. The coefficients of gender diversity, duality and CEO tenure are not significant, and thus have an ambiguous impact on firm performance during the crisis period.
Table 8
This table provides the results of an OLS regression for the effect of multiple board characteristics on firm stock performance during the first crisis period. Only the first crisis period is tested in this model. The BHAR of the crisis period is used as the dependent variable, which is the buy and hold abnormal return in the specified period. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the chairman of the board of directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, *** respectively.
Table 9 presents the results of the OLS regression of firm performance on several board characteristics. Contrary to the result in table 8, average age of directors had a negative and significant coefficient in the rebound period. This negative coefficient means that there is a negative relationship between the average age of directors on the board and firm
performance during the rebound period. This result suggests that firms with younger boards perform better during the rebound period, while the result form table 8 suggests that older boards perform better during the crisis period. The coefficients of board size, gender
diversity, duality and CEO tenure are not significant, and thus have an ambiguous impact on firm performance during the rebound period.
For the purpose of robustness checks, all regressions are also ran using the second and the third rebound periods as dependent variables. These results can be found in table 18 and table 19 in appendix A. Both tables indicate that the abovementioned results are robust across the three different crisis periods.
Table 9
This table provides the results of an OLS regression for the effect of multiple board characteristics on firm stock performance during the main rebound period. Only the first rebound period is tested in this model. The BHAR of the rebound period is used as the dependent variable, which is the buy and hold abnormal return in the specified period. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the chairman of the board of directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, *** respectively.
Director age -0.018 (3.59) *** CEO Tenure -0.001 (0.39) Log Assets -0.013 (1.40) -0.014 (1.57) -0.014 (1.58) -0.016 (1.86) * -0.014 (1.52) Leverage -0.085 (3.00) *** -0.094 (3.31) *** -0.085 (2.99) *** -0.089 (3.12) *** -0.088 (3.12) *** ROA 0.430 (2.30) ** 0.438 (2.37) ** 0.441 (2.38) ** 0.452 (2.44) ** 0.438 (2.37) ** Beta 0.164 (5.96) *** 0.166 (6.14) *** 0.166 (6.08) *** 0.166 (6.21) *** 0.165 (6.06) *** Industry Dummy 0.001 (0.00) -0.001 (0.06) 0.001 (0.08) 0.001 (0.11) 0.001 (0.05) R2 0.21 0.22 0.22 0.24 0.21 Adjusted R2 0.20 0.20 0.20 0.23 0.20 N 375 375 375 375 375
Table 10 shows the results of the same regression model with the firm’s ROA in the first quarter of 2020 as the dependent variable instead of the firm’s BHAR. We observe that none of the board characteristics have a significant relationship with ROA in the first quarter of 2020. We do however observe the expected significant results of the ROA over 2019 and beta, which we also find earlier in the paper.
Table 10
French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, *** respectively.
ROA (Q1-2020) ROA (Q1-2020) ROA (Q1-2020) ROA (Q1-2020) ROA (Q1-2020)
Board size 0.001 (0.59) Gender diversity -0.046 (1.31) Duality -0.008 (1.53) Director age 0.001 (0.72) CEO Tenure -0.001 (0.75) Log Assets -0.002 (0.88) -0.002 (0.99) -0.001 (0.55) -0.010 (0.71) -0.002 (0.92) Leverage 0.021 (1.49) 0.019 (1.45) 0.021 (1.50) 0.022 (1.53) 0.021 (1.48) ROA 0.316 (7.19) *** 0.314 (7.40) *** 0.310 (7.36) *** 0.312 (7.48) *** 0.315 (7.14) *** Beta -0.024 (3.46) *** -0.024 (3.53) *** -0.026 (3.58) *** -0.025 (3.55) *** -0.024 (3.51) *** Industry Dummy -0.001 (0.44) 0.001 (0.31) 0.001 (0.26) 0.001 (0.46) 0.001 (0.51) R2 0.30 0.31 0.31 0.30 0.30 Adjusted R2 0.29 0.30 0.30 0.29 0.29 N 375 375 375 375 375
Table 11
This table provides the results of an OLS regression for the effect of multiple board characteristics on firm performance during the second quarter of 2020. The measure used for firm performance in these regressions is the return on assets (ROA) of the firm during the second quarter of 2020. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the chairman of the board of directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets over the year 2019. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, *** respectively.
ROA (Q2-2020) ROA (Q2-2020) ROA (Q2-2020) ROA (Q2-2020) ROA (Q2-2020)
4.3 Industry differences
The coefficients on the industry dummy variable in table 5 and table 6 indicate that there might be different effects between industries in terms of the relationship between board independence and firm performance. This section will dive deeper into these possible differences and show the regression results per industry.
Table 12 gives the results of the regression ran in all of the ten Fama French industry classifications in the crisis period. In this case, for the ninth (Utilities) and tenth (other) classifications we find significant negative relationships between board independence and the buy and hold abnormal return in the crisis period. The negative coefficient of
classification 10 is in line with the results we found before, that during the 2020 corona crisis, a more independent board decreases the buy and hold abnormal return of a firm. However, the coefficient is significantly higher compared to the result in the original regression, indicating that the negative effect of having a more independent board on the buy and hold abnormal return in a time of crisis is larger in industry classification 10 (other). For classification 9 however, we observe a significant positive coefficient. This positive coefficient indicates that classification 9 firms benefit from having a more independent board, contrary to the results we found for the entire sample size. These two observations give reason to believe that there are differences between industries when investigating the relationship between board independence and firm performance during the 2020 corona crisis. This result is in line with the findings of Francis, Hasan & Wu (2012) on the financial crisis of 2008-2009. They also find evidence of differences between industries, as a possible reason they argue that some industries might be closer related to financial institutions compared to other industries. Therefore, industries closer related to financial institutions might suffer more from a crisis period.
The fact that we observe a significant coefficient at only two of the ten industry
classifications could be because the divided sample sizes are very small, which reduces the explanation power of the model. We also observe some positive but not significant
coefficients in the crisis periods, so there could still be differences between the other eight industry classifications, only the sample size might be too small to detect it.
classifications. Furthermore, we find a negative relationship between the measure for firm size, the natural logarithm of assets, and the firm’s BHAR. These results indicate that in these industries (manufacturing, energy and utilities) bigger firms experienced a lower BHAR during the crisis periods.
For the purpose of robustness checks, all regressions are also ran using the second and the third crisis periods as dependent variables. These results can be found in table 20 and table 21 in appendix A. Both tables indicate that the abovementioned results are robust across the three different crisis periods.
Table 12
This table provides the results of an OLS regression for the effect of board independence on firm stock performance during the first crisis period for different industries. Only the first crisis period is tested in this model. The industry classifications are based on the Fama French ten-industry classifications. The BHAR of the respective crisis period is the dependent variable, which is the buy and hold abnormal return in the specified period. Board Independence is the ratio of independent board directors to board size. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. CEO Tenure
describes the number of years the current CEO of the firm has been in place for. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
BHAR (crisis 1) Fama French ten industry classification
BHAR (crisis 1)
Fama French ten industry classification
6 7 8 9 10 Board Independence 0.610 (0.96) -0.023 (0.05) -0.162 (0.24) 0.526 (2.06) ** -0.791 (2.55) ** Log Assets -0.042 (1.31) 0.035 (1.58) -0.009 (0.43) -0.028 (1.97) * 0.004 (0.16) Leverage -0.190 (0.47) 0.027 (0.45) -0.029 (0.24) -0.283 (2.18) ** 0.017 (0.16) ROA -2.132 (1.64) 0.500 (1.24) 0.239 (0.59) 0.158 (1.02) 0.084 (0.24) Beta -0.446 (6.47) *** -0.338 (5.14) *** -0.293 (3.82) *** -0.287 (15.54) *** -0.248 (8.43) *** CEO Tenure 0.011 (3.39) ** 0.002 (0.76) -0.001 (0.22) 0.003 (1.22) -0.004 (1.87) * R2 0.96 0.60 0.37 0.92 0.73 Adjusted R2 0.89 0.52 0.26 0.90 0.69 N 10 41 39 30 48
Table 13 shows the results of the regression ran in all of the ten Fama French industry classifications in the first rebound period. None of the different industry classifications shows a significant relationship between board independence and the firm’s buy and hold abnormal return.
The fact that none of the ten industry classifications has a significant coefficient leaves us in the middle as to whether there are significant differences between industries when
evaluating the relationship between board independence and firm performance in the 2020 corona rebound period.
For the purpose of robustness checks, all regressions are also ran using the second and the third rebound periods as dependent variables. These results can be found in table 22 and table 23 in appendix A. Both tables indicate that the abovementioned results are robust across the three different crisis periods.
Table 13
This table provides the results of an OLS regression for the effect of board independence on firm stock performance during the first rebound period for different industries. Only the first rebound period is tested in this model. The industry classifications are based on the Fama French ten-industry classifications. The BHAR of the respective crisis period is the dependent variable, which is the buy and hold abnormal return in the specified period. Board Independence is the ratio of independent board directors to board size. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. CEO Tenure
describes the number of years the current CEO of the firm has been in place for. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
BHAR (rebound 1) Fama French ten industry classification
BHAR (rebound 1)
Fama French ten industry classification
6 7 8 9 10 Board Independence -3.851 (1.97) 0.374 (0.59) 0.589 (0.67) -0.491 (0.96) 0.615 (1.03) Log Assets 0.016 (0.16) -0.007 (0.22) -0.119 (4.26) *** -0.007 (0.25) -0.067 (1.56) Leverage 2.674 (2.16) -0.038 (0.44) 0.157 (1.03) -0.236 (0.92) -0.439 (2.16) ** ROA -2.993 (0.75) 0.400 (0.69) -0.273 (0.51) 0.285 (0.92) 0.432 (0.64) Beta 0.614 (2.88) * 0.338 (3.55) *** 0.091 (0.90) 0.223 (6.06) *** 0.098 (1.74) * CEO Tenure -0.014 (1.38) -0.001 (0.25) 0.001 (0.12) 0.006 (1.28) -0.004 (1.08) R2 0.83 0.37 0.47 0.71 0.25 Adjusted R2 0.48 0.25 0.37 0.64 0.14 N 10 41 39 30 48
Table 14 provides the results of the regressions per industry classification with the firm’s ROA in the first quarter of 2020 as the dependent variable. From this table we can only observe a positive significant relationship between board independence and ROA in industry classification 3 (manufacturing). This result differs from the result obtained in table 12, where we found significant relationships between board independence and firm performance in industry classifications 9 and 10.
When comparing table 12 and table 14, we observe several differences. When evaluating which of the two is the better performing model, one can look at the adjusted R2, which tells
us how well the model fits the data. When comparing the adjusted R2 of both table 12 and
Table 15 provides the results of the regressions per industry classification with the firm’s ROA in the first quarter of 2020 as the dependent variable. From this table we can observe positive significant relationships between board independence and ROA in industry
classifications 1 and 4.
Table 14
This table provides the results of OLS regressions for the effect of board independence on firm performance during the first quarter of 2020 for different industries. The industry classifications are based on the Fama French ten-industry classifications. The measure used for firm performance in these regressions is the return on assets (ROA) of the firm during the first quarter of 2020. Board Independence is the ratio of independent board directors to board size. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
ROA (Q1/2020) Fama French ten industry classification
ROA (Q1/2020)
Fama French ten industry classification
6 7 8 9 10 Board Independence 0.033 (1.35) -0.038 (1.28) 0.229 (0.97) 0.101 (1.03) -0.001 (0.00) Log Assets -0.001 (0.22) 0.002 (0.65) 0.002 (0.20) -0.001 (0.45) -0.003 (0.82) Leverage 0.013 (1.03) 0.003 (0.67) 0.040 (0.66) -0.039 (1.52) 0.012 (0.69) ROA 0.132 (1.39) 0.264 (5.57) *** 0.411 (1.73) 0.055 (2.95) *** 0.200 (3.14) *** Beta -0.005 (1.58) 0.005 (0.68) 0.068 (0.93) -0.012 (3.84) *** -0.022 (1.91) * CEO Tenure 0.001 (0.33) -0.001 (1.08) 0.001 (0.92) -0.001 (0.74) -0.001 (0.68) R2 0.78 0.71 0.17 0.49 0.49 Adjusted R2 0.34 0.66 0.01 0.35 0.42 N 10 41 39 30 48 Table 15
This table provides the results of an OLS regression for the effect of board independence on firm performance during the second quarter of 2020 for different industries. The industry classifications are based on the Fama French ten-industry classifications. The measure used for firm performance in these regressions is the return on assets (ROA) of the firm during the second quarter of 2020. Board Independence is the ratio of independent board directors to board size. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets. Beta measures the 5-year monthly stock beta of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, and ***, respectively.
ROA (Q2/2020) Fama French ten industry classification
ROA 0.123 (1.36) 0.412 (1.63) 0.075 (1.55) 0.097 (1.01) 0.178 (4.47) *** Beta -0.043 (3.94) *** -0.042 (2.52) -0.010 (1.79) * -0.009 (1.00) 0.001 (0.07) CEO Tenure -0.001 (0.16) 0.001 (1.41) -0.003 (1.44) -0.001 (0.73) 0.001 (0.75) R2 0.71 0.95 0.39 0.37 0.33 Adjusted R2 0.64 0.62 0.32 0.10 0.28 N 31 8 61 21 86 ROA (Q2/2020)
Fama French ten industry classification
6 7 8 9 10 Board Independence -0.050 (1.32) 0.019 (0.23) -0.035 (0.37) 0.057 (1.06) 0.025 (0.36) Log Assets -0.002 (0.45) -0.006 (0.88) -0.009 (2.29) ** -0.001 (0.29) -0.010 (1.81) * Leverage 0.109 (2.70) * 0.002 (0.19) 0.021 (0.78) 0.015 (0.69) -0.009 (0.42) ROA 0.122 (1.04) 0.110 (0.97) 0.118 (1.57) 0.047 (3.13) *** 0.185 (2.37) ** Beta -0.002 (0.21) -0.031 (2.24) ** -0.032 (1.87) * 0.001 (0.28) -0.018 (3.63) *** CEO Tenure 0.001 (1.51) -0.001 (0.55) 0.001 (0.47) -0.001 (0.80) -0.001 (2.62) ** R2 0.76 0.29 0.26 0.35 0.49 Adjusted R2 0.27 0.17 0.13 0.18 0.42 N 10 41 39 30 48 4.4 Robustness checks
In order to make this study more robust, every OLS regression ran in this study was ran using robust standard errors. Using the Breusch-Pagan test, we did not find evidence for
multicollinearity, the results of these test show a mean variance inflation factor of 1.08, showing no signs of multicollinearity.
Furthermore, we checked the results by selecting different starting and ending dates of the crisis and rebound periods. Most of the results showed to be robust across the different crisis and rebound periods, increasing the statistical strength of the model used in this paper.
Additionally, a second measure of firm performance was used to check if the results were similar. As this second measure we chose the firm’s ROA during the first and second quarter of 2020. In this case we have one measure of firm performance related to investor returns, in the form of the buy and hold abnormal return, and on the other hand we have an accounting measure of firm performance in the form of the ROA during the crisis and rebound period.
The relative unexpected shock of the 2020 corona crisis created an opportunity to optimally analyze the effect of board independence and other corporate governance characteristics on firm performance. This exogenous event helps to better avoid endogeneity issues and as a result provide more evident results on the effect of corporate boards on firm performance. Usually, one would argue that board variables are not exogenous, and are jointly determined with firm value, causing potential endogeneity problems. However, since the crisis is an unexpected exogenous shock, and the board characteristics and accounting data is collected at the end of 2019 before the beginning of the crisis, while changes in firm performance are measured during and after the crisis. Thus, any concerns about endogeneity should be mitigated.
5. Conclusion and discussion
5.1 Hypotheses conclusion and discussion
board. A possible explanation could be that during times of crises, in-company board directors have better knowledge of how to get the company through the crisis compared to outside directors (Bhagat & Black, 2001).
Moreover, this paper tested the hypothesis whether several other board characteristics would have any impact on firm performance during the 2020 corona crisis and the
subsequent rebound period. The results show that a smaller board size and a higher average age of directors had a positive effect on firm performance during the 2020 corona crisis. This result is in line with the findings of Yermack (1996), who also provides evidence which suggests that smaller boards might be more effective compared to larger boards. Gender diversity, duality and CEO tenure did not have any significant effect on firm performance during the 2020 corona crisis period. During the rebound period, the results showed a negative relationship between average age of directors and firm performance, while board size, gender diversity, duality and CEO tenure did not have any significant effect on firm performance. Interestingly, the results indicate that on average in the crisis period it is beneficial to have an older board, while in the rebound period it is beneficial to have a younger board. One could argue that the experience of older board directors is of more importance in crisis periods, thus increasing firm performance.
Additionally, this paper tested the hypothesis whether there was presence of differences between industries when it comes to the relationship between board independence and firm performance. The results show that there are some different effects for certain
industries, we find a reversed result for industry classification 9 (utilities) in the crisis period. Contrary to the main finding, in this industry it was beneficial to have a more independent board during the crisis period. Furthermore, we found a larger magnitude of the main results for industry classification 10 (other) during the crisis period. This presented evidence of differences between industries is in line with the research of Francis, Hasan & Wu (2012), who also found industry differences during the 2008/2009 financial crisis when investigating the relationship between board independence and firm performance.
The economic implication of the results found in this paper are that firms should be aware of the characteristics of their board members, since it could have an impact their performance. Firms might want to favor some characteristics instead of others depending on the industry in which they operate.
Some of the results found in this paper were not consistent across the two different
measures used for firm performance. The model which used the investor returns measure of the buy and hold abnormal return as the dependent variable tended to have a higher
model, but in this paper, we tended to favor the investor returns measure in the form of the buy and hold return. Not solely based on the fact that it had on average a higher adjusted R2
compared to the model with an accounting-based measure for firm performance, but also because an investor returns measure tends to be purer compared to an accounting based measure which can be manipulated by firms in their favor.
5.2 Limitations and future research
Due to the recentness of the 2020 corona crisis, this paper was limited to a short-term evaluation of the effects of the crisis. Although the crisis was not as long as for example the financial crisis of 2008/2009, it would still be interesting to see if there are any long-term or lagged effect like Baysinger & Butler (1985) found in their research. Furthermore, it would have been useful to have a larger sample size, especially when testing for differences
between industries. The sample sizes per industry classification might have been too small to detect possible further significant differences between industries like Francis, Hasan & Wu (2012) managed to find during the 2008/2009 financial crisis with a larger sample size. A future research on this topic with a larger sample size might find more significant results compare to the only weak results found in this paper. Another interesting topic for future research is to compare the relationship between board independence and firm performance during a crisis with the relationship between board independence and firm performance during a regular period. This paper compared the results of the crisis to the following rebound period, however most of the previous research on this topic is done solely in the period of the crisis, while it would also be interesting to compare the results to a ‘regular’ period.
References:
Adams, R., & Ferreira, D. (2009). Women in the boardroom and their impact on governance and performance. Journal of Financial Economics, 94, 291–309.
Baek, J. -S., Kang, J.-K.,& Park, K. S. (2004). Corporate governance and firm value: Evidence from the Korean financial crisis, Journal of Financial Economics, 71,265–313.
Baysinger, B. D., & Butler, H. N. (1985). Corporate governance and the board of directors:
Bhagat, S., & Black, B. S. (2001). The non-correlation between board independence and long-term firm performance. Journal of Corporation Law, 27, 231–273.
Cochran, P., & Wood, R. (1984). Corporate Social Responsibility and Financial Performance. The Academy
of Management Journal, 27(1), 42-56.
Coles, J. W., McWilliams, V. B., & Sen, N. (2001). An examination of the relationship of governance mechanisms to performance. Journal of Management, 27, 23–50.
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.
Fama, E. F., & French, K. R. (1992). The cross‐section of expected stock returns. the Journal of
Finance, 47(2), 427-465.
Francis, B. B., Hasan, I., & Wu, Q. (2012). Do corporate boards matter during the current financial crisis?. Review of Financial Economics, 21(2), 39-52.
Johnson, S., Boone, P., Breach, A., & Friedman, E. (2000). Corporate governance in the Asian financial crisis. Journal of Financial Economics, 58, 141–186.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., (2000). Investor protection and corporate governance. Journal of Financial Economics 58, 3–27.
Mitton, T. (2002). A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis. Journal of Financial Economics, 64, 215–241.
Noe, T. H., Rebello, M. J. (1996). Asymmetric information, managerial opportunism, financing, and payout policies. Journal of Finance, 51, 637-660.
Shivdasani, A., & Yermack, D. (1999). CEO involvement in the selection of new board members: An empirical analysis. Journal of Finance, 54, 1829–1853.
Vafeas, N. (2003). Length of board tenure and outside director independence. Journal of Business Finance and Accounting, 30, 1043–1064.
Yermack, D. L. (1996). Higher market valuation of companies with a small board of directors. Journal of Financial Economics, 40, 185–211.
Appendix A: additional tables
Table 16
directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets over the year 2019. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, ***
respectively. BHAR (crisis 2) BHAR (crisis 2) BHAR (crisis 2) BHAR (crisis 2) BHAR (crisis 2) Board size -0.009 (2.90) *** Gender diversity -0.003 (0.05) Duality -0.002 (0.20) Director age 0.005 (1.94) * CEO Tenure -0.002 (0.25) Log Assets -0.003 (0.54) -0.007 (1.32) -0.007 (1.28) -0.006 (1.13) -0.007 (1.32) Leverage -0.022 (0.61) -0.030 (0.79) -0.030 (0.80) -0.029 (0.79) -0.030 (0.82) ROA 0.374 (4.65) *** 0.404 (4.77) *** 0.403 (4.76) *** 0.399 (4.62) *** 0.405 (4.77) *** Beta -0.200 (15.60) *** -0.195 (15.35) *** -0.196 (15.51) *** -0.196 (15.41) *** -0.195 (15.08) *** Industry Dummy -0.005 (2.50) ** -0.004 (2.40) ** -0.004 (2.41) ** -0.004 (2.35) ** -0.004 (2.34) ** R2 0.62 0.61 0.61 0.61 0.61 Adjusted R2 0.61 0.60 0.60 0.61 0.60 N 375 375 375 375 375 Table 17
model. The BHAR of the crisis period is used as the dependent variable, which is the buy and hold abnormal return in the specified period. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the chairman of the board of directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets over the year 2019. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, ***
Table 18
This table provides the results of an OLS regression for the effect of multiple board characteristics on firm stock performance during the second rebound period. Only the second rebound period is tested in this model. The BHAR of the rebound period is used as the dependent variable, which is the buy and hold abnormal return in the specified period. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the
R2 0.35 0.35 0.35 0.37 0.35
Adjusted R2 0.34 0.34 0.34 0.35 0.34
N 375 375 375 375 375
Table 19
This table provides the results of an OLS regression for the effect of multiple board characteristics on firm stock performance during the third rebound period. Only the third rebound period is tested in this model. The BHAR of the rebound period is used as the dependent variable, which is the buy and hold abnormal return in the specified period. Board size is the number of directors present on the board. Gender diversity measures the percentage of male directors on the board to female directors. Duality is a dummy variable which is equal to one if the CEO of the firm is also the chairman of the board of directors and zero otherwise. Director age is the standard deviation of the average age of all the directors on the board of the firm. CEO Tenure describes the number of years the current CEO of the firm has been in place for. Log assets is the logarithm of total assets of a firm, this variable is used as a measure of firm size. Leverage is the ratio of total liabilities to total assets. ROA is the ratio of net income on total assets over the year 2019. Beta measures the 5-year monthly stock beta of the firm. Industry Dummy is the respective Fama French ten-industry classification code of the firm. Robust t-statistics are in parentheses. Significance at the 10%, 5%, and 1% is indicated by *, **, ***
