How do clawbacks and managerial ability affect the association between audit committees’ equity-based compensation and fraud risk?

The second objective of our study is to examine whether managers’ managerial ability and clawbacks may mitigate the adverse effect of equity-based compensation on audit committees’ oversight effectiveness. In this section, we discuss the empirical models we use to address this research question.

Endogenity of voluntary clawbacks adoption

Because firms’ voluntary adoption of clawbacks is endogenously determined, our empirical analyses are subjected to potential self-selection bias. To control for this issue, we adopt Heckman’s (1979) two-stage procedure. At the first stage, we use the following probit model (2-1) that includes determinants expected to influence a company’s propensity to voluntarily adopt the clawbacks. Note that all independent variables are lagged by one year so that the likelihood of adopting clawbacks during each year depends on the firms.

4We use CEOs’ and CFOs’ equity-based compensation to proxy for boards’ equity compensation for two reasons. First, SOX requires that both CEOs and CFOs assume the responsibilities of establishing, maintaining, and reporting on the effectiveness of internal controls over financial reporting. Even though CFOs are directly responsible for overseeing internal controls (McConnell and Banks 2003; COSO 2004; Sinnett 2007), CFOs’ activities are subjected to CEOs’

monitoring (Hoitash et al. 2014). Second, directors’ independence threat may arise from how their compensation is determined (Brick et al. 2006) and directors’ and CEOs’ compensation is positively related (Vafeas 2003).

,

where the definitions of all variables are summarized in the Appendix.

Drawn on recent studies, we include several important variables that may affect firms’ decision to voluntarily adopt clawbacks. For example, we consider MV because having a formal clawback policy will enhance a firm’s reputation in the stock market (Brown et al. 2011, 2013) and firms attracting more attention in the stock market are more likely to be larger (Aboody et al. 2004; Barton 2005). Therefore, we predict the coefficient of MV to be positive. In addition, we include ROA and predict its coefficient to be positive because firms with good financial performance are more likely to adopt clawbacks (Armstrong et al. 2010; Brown et al.

2013; DeHaan et al. 2011). Moreover, since the Dodd-Frank Act explicitly stipulates that restatement is the trigger for exercising clawbacks and recent studies show that more prior restatements increase the likelihood of firms’ adopting the clawbacks (e.g., Addy et al. 2011; Brown et al. 2013), we consider PRIOR_REST as a control variable and predict its coefficient to be positive. We also include Debt_Issue and Equity_Issue because prior studies indicate that firms adopting the clawbacks are able to establish good reputation about the credibility of their governance and, therefore, enjoy lower cost of capital (e.g., Aboody et al. 2004; Barton 2005). This is consistent with Brown et al.’s (2011, 2013) finding that firms issuing new shares or debts are more likely to adopt the clawbacks. Thus, we predict the coefficients of these two variables to be positive.

We further consider CEOs’ and boards’ power in determining whether firms should voluntarily adopt clawbacks. First, early studies report that executives having longer tenures with the firms (Bushman et al. 2004) or serving as the chairmen of the boards (e.g., Jensen 1993, Core et al. 1999; Bebchuk et al. 2002; Adams et al.

2005) have larger influence over major decision-making. Addy et al. (2011) and Brown et al. (2011, 2013) show that CEOs with longer tenure or serving as the chairmen generally have greater power to reduce the likelihood of adopting clawbacks. We thus consider CEO_Tenure and CEO_Chair to capture these effects and predict their coefficients to be negative. Second, Core et al. (1999) show a negative association between CEO compensation and corporate governance. They argue that CEO compensation should be tied directly with CEO’s performance so that firm’s value can be maximized. Therefore, if the observed board ownerships represent optimal

contracting in equilibrium, CEO’s compensation should be related to firm size and business risk rather than any governance characteristics. We follow Brown et al. (2013) by considering CEO_Own and predict its coefficient to be negative. Finally, since it is more difficult for firms to recover cash bonus if such bonus is earned by the CEO in a fraudulent way (Leone et al. 2006), it is possible that firms whose CEOs receive larger amount of cash bonuses are less likely to adopt clawbacks. Therefore, we include CEO_Bonus and predict its coefficient to be negative. Brown et al. (2013) find that, when more of the CEO’s salary is placed at risk, the firm is less likely to adopt clawbacks.

Test of the clawbacks effect

To test our second research question, we include CLAWBACK, CLAWBACKEQUITY%, and inverse Mill’s ratios (denoted by LAMDA) estimated from model (2-1) into our model (1). This gives rise to the following model (2-2):

where the definitions of all variables are summarized in the Appendix.

Test of the managerial Ability effect

To examine whether managerial ability mitigates the effect of equity-based compensation on audit committees’ oversight effectiveness, we include MGR_Ability and MGRAbilityEQUITY% into model (1) to form the following model (3):

where the definitions of all variables are summarized in the Appendix.

Test of the relative efficacy of clawbacks and managerial Ability

While we can use the coefficients of the interaction terms (i.e., 18) in models (2-2) and (3) to examine whether clawbacks and managerial ability can individually mitigate the adverse effect of equity-based compensation on audit committees, we are also interested in testing which of these two forces is more effective in motivating audit committee members to remain independent. Therefore, we include CLAWBACK, MGR_Ability, and their interactions with EQUITY% into model (1). We then use the following model (4) to test the relative efficacy of clawbacks and managerial ability:

,

where the definitions of all variables are summarized in the Appendix.