Investigating the Impact of Board of Directors on Accounting Conservatism in Kuwait

This study aims to study the influence of boards of directors on the accounting conservatism of non-financial listed firms in Kuwait Stock Exchange (KSE). According to agency theory, accounting conservatism and boards of directors can be considered effective mechanisms for reducing agency conflicts, thereby protecting the interests of investors that used the financial information while also increasing firms’ performance and value. This study sought to build a link between accounting conservatism and boards of directors by using a sample of 87 non-financial firms listed on the KSE at the end of 2019. The study used three independent variables (i.e., board size, board independence, and family directors) and found that none affected accounting conservatism. This study contributes to the literature by identifying the impact of boards of directors on accounting conservatism in Kuwait as no work has been done in this area before.


Introduction
The purpose of this study is to examine a new area of corporate governance in Kuwait-namely, the impact of a board of directors on accounting conservatism. Agency theory argues that the board of directors is an effective mechanism that firms use to reduce conflicts of interest between managers and shareholders as well as among shareholders as such conflict may reduce the performance and value of the firms (Jensen & Meckling, 1976). Agency theory also argues that conservatism can be used to reduce agency conflict because it reduces the overpayment of contracts to directors. Firms with good corporate governance are likely to adopt a high level of conservatism (Mohammed et al., 2019). Basu (1997) argued that conservatism requires a higher level of verification to recognize good news compared to bad news. This situation leads to systematic differences between the news and the timeliness of earnings.
Empirically, many studies from developed countries (Ahmed & Duellman, 2007;Ahmed & Henry, 2012;Boussaid et al., 2015;Elshandidy & Hassanein, 2014;Lara et al., 2009;Lim, 2011) and developing countries (Abdul-Malik, 2017;Abdul-Manaf et al., 2014;Al-Absy et al., 2019;Al-Sraheen et al., 2014;Chi et al., 2009;Chi et al., 2019;El-Habashy, 2019;Foroghi et al, 2013;Kootanaee et al., 2013;Mohammed et al., 2017;Mohammed et al., 2019;Nasr & Ntim, 2018;Suleiman, 2014;Yunos et al. 2014) examined the impact of boards of directors on accounting conservatism. However, these studies may not be adequately relevant for a frontier market such as Kuwait. No work has been done in this area in Kuwait. Thus, studying the impact of boards of directors on accounting conservatism in Kuwait reduces the shortage of literature on this relationship in less developed countries. It also adds a significant contribution to the corporate governance field in Kuwait for practitioners and academic researchers. IASB (2001, p. 21) published Framework for Preparation and Presentation of Financial Statement, which stated that prudence is the inclusion of a degree of caution in the exercise of the judgment needed in the making of estimates required under conditions of uncertainty, such that assets or income are not overstated and liabilities and expenses are not understated. as the policy of expecting all losses and no gains, Watts (2003) classified conservatism into four groups: contracting, accounting regulation, shareholder litigation, and taxation. Previous studies have divided conservation into only two types-namely, conditional and unconditional conservatism. Conditional conservatism is measured by asymmetric timeliness whereas unconditional conservatism is measured by the price-to-book ratio. However, IASB did not distinguish between the two types of conservatism. Abdul-Malik (2017) argued that both types led to the same results: decreasing profits.
The current study seeks to determine whether a board of directors affects the accounting conservatism of non-financial listed firms. To this end, it used data from a sample of 87 firms collected at the end of 2019 and three board variables-namely, board size, board independence, and family directors. Kuwaiti listed firms on the Kuwait Stock Exchange (KSE) offer an appropriate context for this study for three reasons. First, the Kuwaiti government is planning to upgrade the KSE from a frontier market to an emerging market, although this plan has been postponed because of the coronavirus pandemic. Thus, examining the role of directors on accounting conservatism during this time is critical for the government and regulators to understand the roles related to the board of directors and its composition. Second, the KSE is one of the oldest stock exchange markets in the Gulf region and is located in one of the Gulf Cooperation Council (GCC) countries, which include Kuwait, Saudi Arabia, Bahrain, Qatar, United Arab Emirates (UAE), and Oman. Thus, studying the case of Kuwait is like studying six countries at one time. The results of this study could be useful for all six countries because they have the same business environment and rules as well as the same ownership structure. Finally, as no study has yet examined the impact of a board of directors on accounting conservatism, this study aims to fill the existing gap by considering the situation in Kuwait and providing meaningful recommendations for regulators, investors, listed firms, and academics.
To achieve the aims of this study, the study is structured as follows. Section 2 presents the literature review and hypothesis development. Section 3 presents the research methodology, and Section 4 presents the results and discussion. Finally, Section 5 concludes the study, offering limitations and suggestions for future research. Fama and Jensen (1983) argued that using corporate governance mechanisms may encourage firms to adopt conservatism in their financial reports. Thus, a high level of conservatism and good corporate governance would increase the firm's financial reporting quality (El-Habashy, 2019). The current study investigates the impact of three variables of the board of directors (i.e., board size, board independence, and family directors) on conservatism to understand which of these variables firms can use to protect shareholders' interests and reduce agency problems, thereby improving firms' performance. Previous studies have asserted that an effective board of directors would use conservative accounting to assist it in reducing the agency conflict and increasing the transparency of financial reporting (Yunos et al., 2014).

Theoretical Framework and Previous Studies
The current literature review on the relationship between the board of directors and accounting conservatism has found a significant association between the two variables over many years. These studies can be divided into three categories-namely, studies from developed countries, studies from developing countries, and studies from Middle Eastern countries. In terms of developed countries, Ahmed and Duellman (2007) studied the situation in the United States and found that board size does not affect accounting conservatism while inside directors negatively impact accounting conservatism. Meanwhile, Lara et al. (2009) studied accounting conservatism from different perspectives and found that low CEO involvement negatively affects accounting conservatism. This is probably because the US market has effective board independence and such directors work to reduce the control of the CEO. Similarly, Elshandidy and Hassanein (2014) examined this issue in the United Kingdom and found that board size does not affect accounting conservatism, although board independence positively affects it. Meanwhile, Boussaid et al. (2015) studied the situation in France and found that only board size negatively affects accounting conservatism. However, Lim (2011) and Ahmed and Henry (2012) examined this relationship in Australia and found mixed results-probably because of the endogeneity issue that affects the relationship among board variables.
In terms of developing countries, Chi et al. (2009) found that board size negatively impacts accounting conservatism for Taiwanese listed firms whereas Chi et al. (2019) found that family directors positively affect accounting conservatism for the listed firms. Both studies' results are consistent with agency theory arguments and support the studies of Suleiman (2014) in Nigeria and Mohammed et al. (2019) in Turkey, whose results supported agency theory in developing countries (with the exception of board independence in Turkey). Moreover, many studies from Malaysia, such as Al-Absy et al. Finally, in terms of Middle Eastern countries, a number of studies have been conducted, including Nasr and Ntim (2018) and El-Habashy (2019) in Egypt, Al-Sraheen et al. (2014) in Jordan, Abdul.  Malik (2017) in Saudi Arabia, and Kootanaee et al. (2013) and Foroghi et al. (2013) in Iran. The majority of these studies have found that the board of directors did not have any impact on accounting conservatism. They also argued that the board variables are considered very important in developed countries, but not necessarily in Middle Eastern or less developed countries. For example, the majority of these studies found that board size does not impact accounting conservatism. Despite an increasing number of studies examining the impact of corporate governance on accounting conservatism in developed and developing countries, few studies are testing this issue in Middle Eastern countries. As presented in Table 1, although many studies have examined the impact of board of directors on accounting conservatism, no study has examined the situation in Kuwait. These previous studies demonstrated that studying this issue is very limited, even in developed and developing countries. With the exception of one study, no study examined the three board variables in a single study. Thus, this study attempted to fill this gap by examining the impact of three board variables-namely, board size, board independence, and family directors-on accounting conservatism in non-financial listed firms

Hypothesis Development
In terms of board size, agency theory asserts that a small board is more effective and discloses more information to the public (Jensen & Meckling, 1976). Yermack (1996) further argued that big board size makes it easier to control the CEO but reduces the board's ability to make decisions. Empirically, Chi et al. (2009), Suleiman (2014, Mohammed et al. (2019), Nasr and Ntim (2018), Boussaid et al. (2015), and Abdul-Manaf et al. (2014) studied the impact of board size on accounting conservatism and found a negative association. They argued that a small board size is better because it increases accounting conservatism. However, Ahmed and Henry (2012) studied the situation in Australia and found a positive relationship between board size and accounting conservatism. Meanwhile, Ahmed and Duellman (2007) 2014), and Elshandidy and Hassanein (2014) found no relationship between the two variables. Despite these mixed results about the impact of board size on accounting conservatism, the current study still followed the argument of agency theory that a small board is better and has more power in improving the level of accounting conservatism. Thus, the first hypothesis is: H1: Board size has a negative impact on accounting conservatism.
In terms of board independence, Fama and Jensen (1983) and Jensen and Meckling (1976) argued that more independent directors lead to a reduction of the agency conflicts and monitoring managers; thus, board directors may be considered a good mechanism when using accounting conservatism to reduce this conflict. Empirically, the majority of the previous studies supported the agency theory perspective. For example, El-Habashy (2019) (2011), found that board independence does not impact accounting conservatism. Thus, consistent with agency theory, board independence is more likely to provide firms with high monitoring and control for managers, thereby producing more conservatism. Thus, the study's second hypothesis is: H2: Board independence has a positive impact on accounting conservatism.
Regarding the last variable, agency theory argues that, when more family directors are present, agency conflicts would be minimized because the owners and managers would be the same (Jensen & Meckling, 1976). In other words, family directors have a strong relationship with firms' large shareholders (usually from the same family), and such a relationship reduces the agency conflict between managers and shareholders or among shareholders while increasing the firms' value. This view is consistent with the study results of Chi et al. (2019), who examined the impact of family directors on accounting conservatism and found a positive impact. Mohammed et al. (2019) reached a similar conclusion. Both studies argued that family directors on boards have more motivation to implement conservative financial reporting in Turkey's and Taiwan's listed firms, respectively. Meanwhile, Ahmad and Duellman (2007) and Al-Absy et al. (2019) found that firms dominated by insiders (family directors) are more likely to adopt less conservative methods. Yet the current study assumes that family directors provide a high level of accounting conservatism based on the agency theory perspective. Thus, the third hypothesis is as follows: H3: Family directors have a positive impact on accounting conservatism.

Research Methodology
The study sample was collected from non-financial firms on the KSE. These firms were selected because they are the top firms in Kuwait and, therefore, are likely to have the most qualified individuals with the best potential on the board of directors as well as a strong reputation and influence in Kuwait's society. These firms also have strong access to capital and connections with the government to improve their performance and their competitive position. The Kuwait government introduced new rules related to boards of directors, effective from the beginning of 2017. These rules require a minimum of five directors, but have no maximum limit. These rules also require firms to have at least one independent director, but no more than 50% of all directors on the board can be independent. Thus, studying the impact of these rules on accounting conservatism is likely to produce interesting results. The KSE's official website revealed that, on 31 December 2019, the total number of listed firms was 174; 70 financial firms listed were excluded from the sample because they have different regulations and capital structures, and 17 non-financial firms were excluded because they did not have enough data. Thus, 87 firms remained. These firms equal 83% of the total non-financial listed firms and 50% of the total listed firms on the KSE at the end of 2019. This study reclassified the KSE according to the old classification scheme to increase the significance of this study. Table 2 presents more details about the study's sample. The study examined the relationship between three variables and three control variables and accounting conservatism by Kuwaiti non-financial firms listed on the KSE at the end of December 2019. In terms of dependent variables, the literature review identified two measures. The first one was conditional conservatism, based on Basu (1997), which argues that profit has a stronger relationship with bad news (negative stock return) than with good news (positive stock return). This measure indicates the timeliness of good news (G-score) and bad news (C-score) at a linear function of firm characteristics of the market-to-book ratio, firm size, and debt. The second measure is unconditional conservatism, based on Givoly and Hayn (2000). It measures conservatism by adding depreciation expense and subtracting operational cash flow from the income before extraordinary items and dividing the result by total assts. Then the result is divided by 3 and multiplied by -1 to control the effect of large accruals.
This study used the second measure computed and developed by Givoly and Hayn (2000), leaving the first measure for further study because it easy to use. A positive value of accounting conservatism means higher accounting conservatism whereas a negative value means that firms have less accounting conservatism. This study also used three independent variables (board size, board independence, and family directors) and three control variables (debt, firm size, and industry type in KSE). Table 3 presents more details about the study variables. Following the previous studies, this study used an ordinary least square (OLS) regression analysis to examine the impact of boards of directors on accounting conservatism. Haniffa and Hudaib (2006) tested five OLS assumptions-namely, multicollinearity, autocorrelation, homoscedasticity, normality, and linearity-before conducting their analysis. Similarly, the current study used various tests and techniques to test these assumptions, as discussed in the next section. For the purposes of this study, the following regression was employed: AC = α +β1BS + +β2BI +β3FD + β4DT +β5FS +β6IN + ε

OLS Assumptions and Descriptive Analysis
The first step in the study's analysis was to test the OLS assumptions to ensure that our model is relevant for examining the impact of board variables on accounting conservatism; if not, the results would be misleading. As presented in Table 4, the relationship among variables is less than 80%; thus, there were no multicollinearity problems in this study (Brooks, 2014). However, as shown in Table 5, checking the analysis of the statistics for skewness and kurtosis indicated that normality was not met. In addition, testing the plots of the studentized residuals against predicted values, residuals analysis, and Q-Q plots indicated that the assumptions on autocorrelation, homoscedasticity, and linearity were not met. Thus, three variables-accounting conservatism (AC), board independence (BI), and firm size (FS)-were transformed using normal scores. This is consistent with the study of Haniffa and Hudaib (2006), who argued that the data are normal when the skewness is ±1.96 and kurtosis is ±3. Meanwhile, Brooks (2014) argued that the data are normal when skewness is zero and kurtosis is ±3. Notes: ***, **, and * significant at the 0.01. 0.05, and 0.10 levels, respectively (two-tailed). For the definition of the variables, see Table 3. Table 5 indicates that the mean value of accounting conservatism in Kuwait is 0.013; the maximum value is 0.19, and the minimum value is -0.03. This is less than accounting conservatism in Saudi Arabia. Abdul-Malik (2017) found the mean value for accounting conservatism in Saudi Arabia to be about 2.1, although this value is more than the values in Egypt and Jordan. Nasr and Ntim (2018) and El-Habashy (2019) found that accounting conservatism in Egypt was -0.01 and -0.07, respectively. In Jordan, Al-Sraheen et al. (2014) found that the mean value of accounting conservatism was -0.005. Notes: For the definition of the variables, see Table 3.
In terms of the independent variables, the study found that the mean value of the board size was 6.19, with a maximum value of 10 members and a minimum value of five members. Al-Shammria and Al-Sultan (2009) found that board size was 6.39 in non-financial firms listed on the KSE from 2004 to 2007. Meanwhile, the mean value for board independence in the current study was 0.19, with a maximum value of 0.80 and a minimum value of 20%. This is lower that the results of Al-Shammria and Al-Sultan (2009), who found that board independence equaled 82%. The reason for this difference is that previously there were no truly independent directors in Kuwait, so earlier studies measured this variable using non-executive directors; however, current Kuwaiti laws require listed firms to have at least one independent director (but no more than 50% of all directors should be independent). The current study also found that the mean value of the family directors was 0.19, with a maximum value of 0.63 and a minimum value of zero. For the control variables, the study found that the mean values of debt and firm size were 0.13 and KD232999 respectively (1 US dollar = 0.310 Kuwaiti dinar). Table 6 presents the results from the OLS regression by investigating the impact of boards of directors on accounting conservatism using a sample of non-financial firms listed on the KSE. The study found that the F-value was significant and equal to 2.297, the R-square was 0.198, and the Adjusted-R-square was 0.123. In addition, board size (BS) did not significantly impact accounting conservatism. Thus, Hypothesis 1, which predicted that board size negatively impacts accounting conservatism, was not supported. The coefficient of board size was not significant (β = -0.086; p > 0.10), which is inconsistent with agency theory's argument that a small board is more effective and can be used to reduce the agency conflict and improve firm performance and, thus, accounting conservatism. This result is probably due to the endogeneity issue. Many variables affect board size but not accounting conservatism, as also demonstrated in Ahmed and Duellman (2007), Kootanaee et al. (2013), Yunos et al. (2014), Elshandidy and Hassanein (2014), Mohammed et al. (2017), and Lim (2011  Notes: ***, **, and * significant at the 0.01. 0.05, and 0.10 levels, respectively (two-tailed). The excluded sector is the food sector. For the definition of the variables, see Table 3.

OLS Results and Discussion
Board independence did not significantly impact accounting conservatism. Hypothesis 2 was rejected as the coefficient of board independence was not significant (β = -0.147; p > 0.10), which is inconsistent with agency theory's argument that board independence leads to a reduction in agency conflict and improves a firm's performance and value. The reason for such a result is probably Kuwait's laws, which require listed firms to have at least one independent director; however, one independent director is not enough to play an effective role inside the boardroom. This result is consistent with the results of studies by Kootanaee et al. (2013), Yunos et al. (2014), Lim (2011), Boussaid et al. (2015, and Mohammed et al. (2019). However, other studies, such as El-Habashy (2019), found that board independence positively affects accounting conservatism; she argued that more independent directors are likely to adopt greater accounting conservatism.
The results also indicated that family directors do not affect accounting conservatism; thus, the third hypothesis was rejected. The coefficient of board family directors was not significant (β = 0.275; p > 0.10), which is inconsistent with agency theory's argument that family directors reduce the agency conflict and improve a firm's performance and value. Ahmed and Duellman (2007) studied the role of inside directors on accounting conservatism and found a negative impact. Al-Absy et al. (2019) found similar results for family directors in Malaysia. Both of these studies concluded that family directors on the board are not an effective mechanism for improving accounting conservatism. However, Chi et al. (2019) and Mohammed et al. (2019) found opposite results in Taiwan and Turkey, respectively. In terms of the control variables, this study found that debt positively affects accounting conservatism (β = 1.740; p < 0.01), meaning that debt leads to more accounting conservatism and financial obligations from Kuwaiti banks, although the study found that firm size negatively impacted accounting conservatism (β = -0.385; p < 0.01) and industry type did not play a significant role in accounting conservatism. El-Habashy (2019) found that, in Egypt, banks and firm size do not affect accounting conservatism. Foroghi et al. (2013) found that greater debt and larger firms produce greater accounting conservatism.
In short, the results of this study differ from those conducted in less developed countries for three major reasons. First, the governance rules are still new in Kuwait and will probably need to be in effect longer before studying the situation. In addition, many Kuwaiti listed firms still do not fully follow these rules. Second, there is no single corporate governance code or rule relevant for all countries; thus, differences in results is common because of the different rules, ownership structures, legal systems, shareholders' protections, and cultures. Third, the issue of endogeneity could impact the findings of this study. The endogeneity issue refers to the interrelationship among board variables. Thus, finding that board variables do not work effectively in terms of accounting conservatism does not necessarily mean that agency theory is not applicable in Kuwait.

Conclusion
This study investigated the impact of a board of directors on accounting conservatism. It used three board variables (board size, board independence, and family directors), and the study results indicated that none of the board variables affected accounting conservatism. However, debt positively impacted accounting conservatism while firm size negatively impacted it. The findings may shed new light related to corporate governance on the board variables in listed firms through the relationship between the board of directors and accounting conservatism. The results may also contribute to current corporate governance studies by providing a new data set from boards of directors in Kuwait's listed firms. For listed firms and investors, this study highlighted the importance of accounting conservatism and what factors improve this figure.
For academics, the results of this study have highlighted that the possible impact of board variables on accounting conservatism found in previous studies probably stemmed from the fact that they used many omitted variables, more sample years, or an econometrics testing technique. The current study found no empirical relationship between all of the board variables and accounting conservatism for the sample of non-financial listed firms at the end of 2019. Thus, further studies could examine this relationship in more detail by including more variables, such as ownership concentration and board diversity, or by increasing the sample size to three years (e.g., from 2019 to 2021) and compare their results with the situation before 2017. Such panel data will give researchers more opportunity to examine the impact of boards of directors on accounting conservatism more deeply because, before 2017, there were no governance rules in Kuwait, and many listed firms delayed following these rules until 2019. Thus, using time periods prior to 2019 will be not relevant. Researchers can also use unconditional conservatism instead of conditional conservatism to examine the relationship between boards of directors and accounting conservatism. Finally, this study ignored the issues of endogeneity stemming from the relationship among board variables. Thus, researchers can conduct further research by considering the problem of endogeneity.