Clawback條款、權益基礎薪酬和審計委員會之監督效率性 - 政大學術集成

全文

(1)Clawback Provisions, Equity-based Compensation and Audit Committees’ Oversight Effectiveness. 立. 政治大 Yu-Chun Lin. ‧ 國. 學. Under the supervision of Professor Hung-Chao Yu. ‧ er. io. sit. y. Nat. n. A dissertation of a submitted in partial fulfillment v. i l C n h edegree i U of Philosophy the requirement for the n g cofhDoctor. Department of Accounting College of Commerce National Chengchi University July 30, 2012.

(2) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i Un. v.

(3) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v.


(5) 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v.


(7) ABSTRACT. This study first examines whether equity-based compensation (i.e., stocks and options) is associated with audit committees’ oversight failures. I then examine whether this association between equity-based compensation and oversight failures is affected when firms initiate the clawback provisions in their compensation contracts. I use the likelihood of restatements, the incidence of internal control weaknesses (ICW), and earnings management measures to proxy for audit committees’ oversight failures. Based on a sample of 129 firms that voluntarily adopt the clawback provisions during 2003-09 and a matched sample created from the propensity score matching technique, I find several important results. First, larger amounts and portions of stocks and options are associated with higher restatement and ICW likelihood and greater earnings management. Second, equity-based compensation appears to harm audit committees' oversight effectiveness. However, the adoption of the clawback provisions significantly mitigates such negative effect. Finally, the clawback provisions are effective in reducing restatements, ICW, and earnings management only when these provisions are triggered by "bad faith" rather than restatements. Overall, my empirical results bear policy implications on audit committees’ compensation practice and the mandatory adoption of the clawback provisions.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v.


(9) TABLE OF CONTENTS CHAPTER 1 INTRODUCTION .....................................................................................................1 1.1 Equity-based Compensation and Audit Committees’ Oversight Effectiveness .....................1 1.2 The Clawback Provisions .......................................................................................................3 1.3 The Interaction of Equity-based Compensation and Clawback Provisions ...........................5 CHAPTER 2 BACKGROUND, LITERATURE AND HYPOTHESIS DEVELOPMENT.......8 2.1 Clawback Provisions ..............................................................................................................8 2.1.1 Institutional background ...............................................................................................8 2.1.2 The nature of clawbacks ...............................................................................................9. 政治大. 2.1.3 Evidence of clawback provisions ...............................................................................12. 立. 2.2 Audit Committee Compensation ..........................................................................................14. ‧ 國. 學. 2.3 Hypothesis Development ......................................................................................................15 CHAPTER 3 RESEARCH DESIGN .............................................................................................19. ‧. 3.1 Data and Sample Selection ...................................................................................................19. Nat. sit. y. 3.2 The Endogenous Biases ........................................................................................................20. n. al. er. io. 3.2.1 Propensity score matching ..........................................................................................21. i n U. v. 3.2.2 Voluntary adoption model ..........................................................................................22. Ch. engchi. 3.3 Regression Model .................................................................................................................24 3.3.1 Restatement likelihood ...............................................................................................24 3.3.2 Incidence of internal control weaknesses ...................................................................29 3.3.3 Earnings management.................................................................................................31 3.3.4 Control for CEO compensation ..................................................................................35 CHAPTER 4 EMPIRICAL RESULTS .........................................................................................36 4.1 Descriptive Statistics ............................................................................................................36 4.2 Regression Results...............................................................................................................39 4.2.1 The voluntary adoption model ...................................................................................39 i.

(10) 4.2.2 Audit committees’ oversight failure – Proxied by restatement likelihood ................40 4.2.3 Audit committees’ oversight failure – Proxied by the incidence of ICW ..................42 4.2.4 Audit committees’ oversight failure – Proxied by accruals quality ...........................45 4.2.5 Audit committees’ oversight failure – Proxied by real earnings management ..........46 4.3 Additional Tests...................................................................................................................48 4.3.1 Trigger effects ............................................................................................................48 4.3.2 Excluding special restatement period ........................................................................49 4.3.3 Panel data results for restatements..............................................................................49 4.3.4 Excluding certain industries .......................................................................................50. 政治大. 4.3.5 Excluding cash compensation ratio variable ..............................................................50. 立. 4.3.6 Corporate governance effects .....................................................................................51. ‧ 國. 學. 4.3.7 Effects of board compensation ...................................................................................51 4.3.8 Heckman two stage results .........................................................................................52. ‧. 4.3.9 Alternative measures in firm performances................................................................53. Nat. sit. y. 4.3.10 Excluding financial crisis period ..............................................................................53. n. al. er. io. CHAPTER 5 SUMMARY AND CONCLUSIONS ......................................................................54. i n U. v. REFERENCES ................................................................................................................................55. Ch. engchi. TABLE ..............................................................................................................................................68 APPENDIX A .................................................................................................................................122 APPENDIX B .................................................................................................................................126. ii.

(11) LIST OF TABLES Table 1 Sample Selection Procedure .................................................................................................68 Table 2 Variable Definitions ..............................................................................................................69 Table 3 Distribution of Clawbacks ....................................................................................................72 Table 4 Analysis of Audit Committee Compensation ........................................................................74 Table 5 Descriptive Statistics .............................................................................................................76 Table 6 Distribution of Restatements, ICW, and Earnings Management ..........................................79 Table 7 Regression Result of the Voluntary Adoption Model ...........................................................81 Table 8 Clawback Provision Effects on the Association between Restatement Likelihood and Audit Committee Compensation ....................................................................................................82. 政治大 Table 9 Clawback Provision Effects 立on the Association between Incidence of ICW and Audit. Committee Compensation ....................................................................................................84. ‧ 國. 學. Table 10 Clawback Provision Effects on the Association between Accruals Quality and Audit Committee Compensation ..................................................................................................86. ‧. sit. y. Nat. Table 11 Clawback Provision Effects on the Association between Real Earnings Management and Audit Committee Compensation ........................................................................................88. al. er. io. Table 12 Trigger Effects on Audit Committees’ Oversight Failures..................................................90. v. n. Table 13 Clawback Provision Effects on Restatement Likelihood – Excluding Restatements Spanning Initiated Clawback Provision Adoption .............................................................93. Ch. engchi. i n U. Table 14 Clawback Provision Effects on Restatement Likelihood – Panel Data Results .................95 Table 15 Clawback Provision Effects – Excluding Certain Industries ..............................................97 Table 16 Clawback Provision Effects - Excluding Cash Compensation Ratio Variables ................100 Table 17 Clawback Provision Effects – Controlling for Corporate Governance Effects ................102 Table 18 Clawback Provision Effects – Compensation for Other Board Committees ....................105 Table 19 Clawback Provision Effects – Using Heckman Two-Stage Selection Model (Second-Stage Results) ............................................................................................................................. 110 Table 20 Clawback Provision Effects – Alternative measures in firm performances...................... 116 Table 21 Clawback Provision Effects – Excluding financial crisis period ...................................... 119 iii.


(13) CHAPTER 1. INTRODUCTION. 1.1 Equity-based Compensation and Audit Committees’ Oversight Effectiveness Due to recent financial reporting scandals, regulators and researchers have emphasized the importance of corporate governance to enhance financial reporting quality and restore investors’ confidence (e.g., Karamanou and Vafeas 2005; Srinivasan 2005). The Sarbanes-Oxley Act (hereafter, called SOX) marks a significant milestone for corporate governance and imposes numerous provisions on executives (Section 302, 305), board of directors (Section 301, 407), auditors (Section 201, 203), and internal control over financial reporting (Section 404). One major governance mechanism emphasized by the SOX is audit committee.1 According to Section 301,. 政治大. audit committees should be fully independent so that they can effectively oversee corporate. 立. financial reporting, internal controls addressing key risks, and auditor activities such as. ‧ 國. 學. appointment, dismissal, and the determinant of audit fees. In addition, Section 407 requires that audit committees contain at least one financial expertise. Prior auditing studies have examined. ‧. whether and how audit committee characteristics (e.g., independence, expertise, composition, and. Nat. sit. y. diligence) affect firms’ financial reporting (e.g., Bronson et al. 2009; Beasley 1996; Beasley et al.. n. al. er. io. 2000; Dhaliwal et al. 2010; Klein 2002a, 2002b), auditor changes (e.g., Abbott et al. 2000; Carcello. i n U. v. and Neal 2003; Chen and Zhou 2007; Lennox and Park 2007; Menon and Williams 2004, 2008;. Ch. engchi. Naiker and Sharma 2009), and audit committees’ oversight effectiveness (e.g., Beasley et al. 2009; Bédard et al. 2004; Srinivasan 2005). More recent studies turn attention to audit committee compensation (e.g., Archambeault et al. 2008; Cullinan et al. 2010; Engel et al. 2010; Magilke et al. 2009). There are two possible reasons. 1. Before the SOX regulation, the role of audit committees had received much attention. Since 1940, the Securities and Exchange Commission (SEC) has recognized that an audit committee could serve an important, and ultimately necessary, function in ensuring that a publicly traded company’s financial reporting is accurate. In the 1970s, the New York Stock Exchange (NYSE) required boards of directors of listed companies to appoint an audit committee; in the 1980s, the National Association of Securities Dealers (Nasdaq) and American Stock Exchange (AMEX) subsequently followed suit. In February 1999, audit committees received attention when a committee composed of individuals from the NYSE, Nasdaq, public companies, and CPA firms issued the Report and Recommendations of the Blue Ribbon Committee on Improving the Effectiveness of Corporate Audit Committees (available at www.nyse.com or www.nasd.com). The report recognized that the audit committee has a crucial role in ensuring high-quality financial reporting. 1.

(14) for the increased importance of audit committees’ compensation. First, audit committees are subjected to many SOX previsions, leading to higher workload and liability exposures than other board members (Ward 2009). Therefore, board compensation has become more individualized to reflect members' different efforts and responsibilities. According to Hay’s (2003) survey, 58% of the audit committee chairs and 19% of the audit committee members receive higher retainers than their counterparts on other board committees. Second, the equity-based compensation (including stocks and options) may be a potential source that may threaten audit committee independence. Stocks and options to compensate non-executive directors have been used by large U.S. firms (Hay survey. 2003;. Taub. 2005;. Winikoff. 2006).. Although. the. National. Association of. 政治大. Corporate Directors (NACD 2001, 2003) supports the use of equity-based compensation,. 立. compensating audit committees with stocks and options could be problematic because equity. ‧ 國. 學. ownership may increase the affiliation between audit committee members and the firms (Carcello and Neal 2003), leading to more financially dependent audit committees.. ‧. Prior studies use the pre-SOX sample and find that option grants reduce directors’ monitoring. Nat. sit. y. of earnings management (e.g., Bédard et al. 2004) and are associated with accounting restatements. n. al. er. io. (e.g., Archambeault et al. 2008) and internal control weaknesses (e.g., Cullinan et al. 2010). Engel. i n U. v. et al. (2010) focuses on the determinants of audit committee compensation and finds that cash. Ch. engchi. compensation is positively associated with the demand for monitoring financial reporting process. The first purpose of this study is to test whether equity-based compensation is associated with audit committees’ oversight effectiveness. Using S&P 500 firms and controlling for self-selection bias, I find that firms compensating their audit committees with equity-based compensation are more likely to incur restatements, internal control weaknesses, and engage in earnings management. 2 I also find that paying more cash compensation strengthens audit committees’. 2. I use both accruals quality and real earnings management index to proxy for earnings manipulation. Audit committees are responsible for firms’ financial reporting quality, Cohen et al. (2008) find that, while managers tend to use traditional accruals to manage earnings before SOX, they switch to the real activities after SOX. Because this research used post-SOX sample (i.e., year 2003 - 2009), I use real earnings management to proxy for audit committee’ oversight failure. 2.

(15) oversight effectiveness. This study differs from prior studies in four aspects. First, instead of focusing on option compensation, I extend Bédard et al. (2004) and Archambeault et al. (2008) by examining the differential effects cash, stocks, and options on audit committees’ oversight effectiveness. Second, Cullinan et al.’s (2010) results may not be convincing because ICW in 2004-2005 are not suitable to proxy for ineffective audit committees’ oversight. Firms could probably need more time and audit committees’ effort to remedy those ICW in the starting two years of Section 404. My evidence regarding ICW is obtained from 2004 to 2009, and it could be better to test the research question. Third, while Engel et al. (2010) shows that firms having higher demand for monitoring. 政治大. financial reporting process pay more cash retainers to audit committees, I extend Engel et al. (2010). 立. by showing that firms paying more cash compensation are likely to have less audit committees’. ‧ 國. 學. oversight failure. Finally, once the Public Company Accounting Oversight Board (PCAOB) is steering its attention to audit committees, my empirical results suggest that regulators may need to. ‧. put more restrictions on audit committees’ compensation.3. Nat. sit. y. 1.2 The Clawback Provisions. n. al. er. io. Another important provision in SOX is the recoupment of executives’ compensation or. i n U. v. bonuses (i.e., the clawback provisions). Section 304 stipulates that certain bonuses previously paid. Ch. engchi. to the executives could be forfeited or repaid to the issuer when restatements occur due to material noncompliance or misconduct. Since SOX authorizes the Securities and Exchange Commission (SEC) to recoup these bonuses, the clawback provisions under Section 304 is enforced at the SEC level (Fried and Shilon 2011). The SEC rarely enforces this provision due to the difficulty in assessing and proving managerial misconduct (Chan et al. 2012b).4 Even though Section 304 is 3. An opening speech by Franzel (March 28, 2012): We heard from the participants a wide range of potential actions that could help improve the objectivity, credibility and reliability of financial audits, including: strengthening audit committee oversight and evaluation of the audit firm and audit process, including disclosures about the audit committee's activities;…. 4 The recoupment at SEC level is triggered in two cases. UnitedHealth Group recently recouped more than $450 million in compensation from its CEO, Dr. William McGuire, as a result of a stock options backdating scandal that was disclosed in 2006 (SEC 2007). In addition, Maynard Jenkins, former CEO of the car parts manufacturer CSK Auto Corp., was ordered to repay over $4 million in connection with firm financial reports for 2004 and 2007 that were subsequently restated (SEC 2009). It has been reported that the agency is contemplating similar action in a case 3.

(16) enforceable only by the SEC, a few listed firms began to establish their clawback provisions since early 2005. On March 26, 2006, the Council of Institutional Investors recommended to the SEC that firms should include policies for recapturing incentive pay following restatements in the Compensation Discussion and Analysis of their proxy statements. In response to this suggestion, the SEC revised the 2006 Disclosure Provision of Regulation S-K, stating that clawbacks constitute a material element of public firms’ compensation of named executive officers and, therefore, should be disclosed. Two recent Acts further reinforce the implementation of the clawback provisions. The first one is Section 111(b)(2)(B) of the Emergency Economic Stabilization Act (EESA) of 2008. 政治大. (enacted on October 3, 2008), which require financial institutions have the repayment of. 立. executives’ bonuses; the second one is Section 954 of the 2010 Dodd-Frank Act (signed on July 21,. ‧ 國. 學. 2010), which rules that all listed firms to implement a policy to recover incentive compensation from any current or former executive after restatements that occur due to material noncompliance. ‧. with financial reporting rules (U.S. Congress, 2010). 5 Different from SOX Section 304, the. Nat. sit. y. Dodd-Frank Act authorizes the board of directors to recoup the compensation.. n. al. er. io. While the SEC has decided to postpone the implementation of Section 954 to Middle 2012, a. i n U. v. notable trend in the development of the clawback provisions is that many listed firms other than. Ch. engchi. financial institutions voluntarily adopted their own provisions to recover bonuses before the Dodd-Frank Act.6 Using firms voluntarily adopting clawbacks from 2007 to 2011, recent research examines the economic determinants of firms’ voluntary adoptions (e.g., Brown et al. 2011; Addy et al. 2011; Gao et al. 2010), the impacts of clawback provisions on financial reporting quality and auditor behavior (e.g., Chan et al. 2012a; DeHaan et al. 2011), and market’s reaction to such. against Ian McCarthy, CEO of Beazer Homes (Esme 2009). The SEC must direct national stock exchanges to require each listed firm to adopt a policy to require the clawback of incentive compensation erroneously awarded to current and former executive officers during the three-year period preceding the date on which a firm is required to prepare an accounting restatement. 6 The Corporate Library (2010) indicates that the number of firms with clawback provisions is increasing, albeit slowly, and this rise has continued. A total of 638 firms (18.9 percent) in The Corporate Library’s coverage universe of 3,380 firms had a clawback provision as of January 19, 2010. About 39.8 percent of the S&P 500 have clawback provisions (194 firms), while 28.4 percent of the Russell 1000 have such a policy (284 firms). 4 5.

(17) voluntary adoptions (e.g., Gao et al. 2010). A well-crafted clawback policy can enhance a firm’s overall compensation strategy by establishing a viable disincentive against fraud, misconduct, and excessively risky or otherwise harmful acts. Other regulatory agencies and the press explicitly recognize clawbacks as one of the tools a firm can use to manage compensation related risks (e.g., Leaders’ statement 2009; Scott et al. 2010). The firms voluntarily adopt clawback provisions by showing clawback information in the disclosure in firms’ Compensation Discussion and Analysis (CDA), a portion of firms’ definitive proxy statement (Form 14A-DEF). The clawback provisions are generally applied to all executives and board members. Since Caskey et al. (2010) points out that audit committees’ penalties for not. 政治大. diligently collect information during their oversight process are too low, it is reasonable to expect. 立. that clawback provisions shall constitute a form of monetary penalty that may affect audit. ‧ 國. 學. committees’ ex ante effort in monitoring firms’ financial reporting.. 1.3 The Interaction of Equity-based Compensation and Clawback Provisions. ‧. Equity-based compensation and clawback provisions create two counterbalancing forces on. Nat. sit. y. audit committees’ oversight effectiveness. On the one hand, prior auditing studies show that stock. n. al. er. io. options could potentially weakens audit committee independence, leading to oversight failurs. i n U. v. reported in restatements, internal control weakness and earnings management (e.g., Archambeault. Ch. engchi. et al. 2008; Bédard et al. 2004; Cullinan et al. 2010). On the other hand, clawback provisions create one form of monetary penalties on audit committees’ oversight failures (e.g., Caskey et al. 2010).7 Therefore, whether firms’ clawback provisions mitigate the adverse influence of equity-based compensation on audit committees’ effectiveness is an empirical question and deserves more in-depth examination.8 Currently, such evidence is rare, if exists.. 7. While Caskey et al. (2010) concluded that audit committee parameters affect not just the ex post reporting process, but also the ex ante information-collection process, I extend this notion to suggest that misreporting penalties (i.e., clawback provisions) affect audit committees’ ex ante effort on collecting due-diligence information. 8 This empirical question seems to the second-order effect of clawback provisions. Although the first-order effect (i.e., the association clawback provisions and financial reporting quality) is tested in prior literature, I attempt to consider the role of audit committees on financial reporting process by showing the second-order effects. Considering audit committees and CEOs can help understand how clawback provisions affect firm participants’ quality, leading to high financial reporting quality. 5.

(18) Using all S&P 500 firms and controlling for endogenous biases, I find that clawback provisions significantly mitigate the adverse effects of equity-based compensation on audit committees’ oversight effectiveness. These findings enrich the growing literature that examines the association between equity-based compensation and audit committees’ effectiveness (e.g., Archambeault et al. 2008; Bédard et al. 2004; Cullinan et al. 2010). To the best of my knowledge, my study is the first one that investigates the issue of whether firms’ voluntary adoption of clawback provisions mitigates the unfavorable effects of the use of equity-based compensation for their audit committees. Because a observed decrease in audit committees’ oversight failure could be driven by the. 政治大. decreased CEO manipulation in reported earnings, I thus control for CEO compensation by. 立. classifying all sample into two subsamples: high and low CEO equity-based compensation groups.. ‧ 國. 學. Desai et al. (2006) finds that, if managers understand that their fraudulent behavior will be penalized ex post, they should have less incentive to engage in earnings manipulation ex ante. The. ‧. control for this possibility does not affect my empirical results notably, and clawback provisions. Nat. n. al. er. io. committees, no matter what type of firms’ CEO compensation.. sit. y. themselves mitigates the adverse effects of the equity-based compensation for their audit. i n U. v. While the content of clawback provisions may vary widely with the language of any particular. Ch. engchi. contract, Chrry and Wong (2009) suggests that clawback provisions could be triggered under three circumstances: misconduct or fraud, the restatement of financial results, or the event of employees’ bad faith. I use Chrry and Wong’s (2009) categorization to analyze the possibly differential influences of these three triggers. Since regulators require firms to adopt recoupment policy on firms’ compensation agreement from 2012, the evidence of differential trigger effects from voluntary clawback adoption provides implications for the mandatory clawback provisions. I find that clawback provisions strengthen audit committees’ oversight effectiveness, especially when firms use bad faith events as the trigger. This finding makes three contributions to the literature and the practice. First, Caskey et al. (2010) models a financial reporting process to define 6.

(19) audit committees’ ex ante information-collection process. However, it is hard to observe audit committees’ diligence during the financial reporting monitoring. I extend Caskey et al. (2010) by providing empirical evidence of the association between clawback provisions and the monitoring outcome of audit committees. Second, I show the differential effects of ckawback triggers. Although prior literature suggests that clawback provisions have a beneficial impact on financial reporting quality (e.g., Chan et al. 2012a; DeHaan et al. 2011), the diversity of clawback provision has not been discussed. The result in this paper indicates that the benefits of clawback provisions are driven by specific “bad faith” triggers. Third, my empirical results are expected to have useful implications for the mandatory clawbacks that will become effective in July 2012. Under Section. 政治大. 954 of the Dodd-Frank Act, clawback provisions will be triggered only when firms incur. 立. restatements. I find no significant association between restatement triggers and audit committees’. ‧ 國. 學. effectiveness. Since CEOs and directors could engage in bad faith actions which probably do not cause restatements, regulators shall impose extra restrictions on the triggers of clawback provisions. ‧. to increase the recoupment effects.. Nat. sit. y. The remainder of this dissertation is organized as follows. Section 2 discusses the background,. n. al. er. io. relevant literature and hypothesis development. Section 3 describes the sample selection procedure. i n U. v. and research design. Section 4 reports the descriptive statistics and empirical results. Section 5 contains summary and conclusion.. Ch. engchi. 7.


(21) CHAPTER 2 BACKGROUND, LITERATURE AND HYPOTHESIS DEVELOPMENT 2.1 Clawback Provisions 2.1.1 Institutional background Firm’s incentive contracts can require that awards be cancelled or “clawed back” (i.e., must be repaid to the firm) after financial statements are restated in order to enhance board’s monitoring. Although some firms utilize clawbacks as a tool prior to 2002, Section 304 of SOX is the first federal statute to introduce that certain bonuses previously paid to the executives could be forfeited or repaid to the issuer. Under this provision, such forfeiture or repayment obligation applies only to the issuer’s CEO and CFO and is triggered upon a restatement of the issuer’s financial statements. 政治大. due to material noncompliance or misconduct. While SOX authorizes the Securities and Exchange. 立. Commission (SEC) to recoup bonuses, the SEC rarely enforce this provision due to its limited. ‧ 國. 學. recourses and difficulty in proving managerial misconduct (Chan et al. 2012b). Afterward, under the Emergency Economic Stabilization Act (EESA) of 2008 and the. ‧. American Recovery and Reinvestment Act (ARRA) of 2009, the federal bail-out program. Nat. sit. y. re-introduced the concept of executive repayment of bonuses related to inaccuracies of financial. n. al. er. io. statements.9 These new rules apply to a broader group of executives and are widely viewed as. i n U. v. critical to the Congressional approval of the federal bail-out program. In 2009, the Troubled Asset. Ch. engchi. Relief Program (TARP) implements the bailout program’s executive compensation provisions.10 The interim final rule provides that any bonus payment with respect to certain executives of bail-out recipients must be subjected to a recovery or clawback provision, which is triggered under certain circumstances relating to materially inaccurate financial statements or performance metric criteria. A bonus payment for this purpose is broadly defined to include retention and incentive payments and will be deemed to have been made when the employee obtains a legally binding right 9. Under the employment agreement, the CEOs are required to repay certain bonus and incentive- or equity-based compensation they receive if firms are required to restate their financial statements as a result of CEOs’ misconduct, consistent with Section 304 of the Sarbanes-Oxley Act of 2002. 10 On June 10, 2009, the Department of the Treasury issued an interim final rule entitled“TARP Standards for Compensation and Corporate Governance.” See 31 CFR Part 30, RIN 1505-AC09; scheduled for Federal Register publication on June 15, 2009. 8.

(22) to the payment. In addition, the TARP delegates the Secretary of the Treasury as the enforcement authority and requires each bailout recipient to meet appropriate standards for executive compensation and corporate governance. A key clarification in TARP is that firms are required to enforce the measures unless doing TARP clawbacks is demonstrably unreasonable.11 Firms’ clawback provisions may bear little resemblance to the similar statutory measures found in the SOX and TARP legislation. Those clawback provisions are applied only to CEOs and CFOs of public firms under SOX, and to a certain number of top executives and highly paid employees under TARP. Importantly, only financial restatements arising from misconduct gives rise to the SOX clawback, which may be enforced solely by and at the discretion of the SEC. In. 政治大. contrast, Section 954 of the Dodd-Frank Act in 2010 requires all listed firms adopt and implement a. 立. policy on the recovery of incentive compensation based on erroneous financial statements that are. ‧ 國. 學. later restated due to material noncompliance with financial reporting requirements. Since the clawback provisions are mandatory under the Dodd-Frank Act, and enforced by the boards,. ‧. directors will become increasingly concerned with the compliance with Dodd-Frank Act. In light of. Nat. sit. y. the importance of the clawback provisions to firms’ compensation strategy, shareholder groups,. n. al. er. io. legislators, and compensation reform advocates are endorsing clawbacks as an effective tool to. i n U. v. prevent undeserved windfalls by mitigating compensation-related risk. 2.1.2 The nature of clawbacks. Ch. engchi. The design and implementation of firms’ clawback policy can dramatically affect how it is perceived, its efficacy as a deterrent against misconduct or harmful acts, and the extent to which it may be enforced. The clawback design is usually be informed by firms’ goals for the clawback policy within the context of its existing compensation, recruitment and retention policies, and practices. Some firms and institutions specifically note the use of clawbacks as a factor to manage risks arising from incentive-based compensation.12 Also, the corporate culture and the degree of 11. For example, the cost of enforcing the rights would exceed the amount to be recovered. The Group of 20 (G20) nations specifically endorsed clawbacks as a risk-management tool for the financial industry at their September 2009 Summit in Pittsburgh (Leaders’ statement 2009). Specifically, Leaders' Statement (2009) stipulates that (emphasis added): “Reforming compensation practices to support financial stability: Excessive 9. 12.

(23) shareholder commitment to good governance initiatives play a vital role in shaping the policy. One important feature of the clawback provisions is the “trigger” to enforce the recoupment. Since firms employ voluntary clawback provisions for myriad purposes, the purpose of the clawback may affect the choice of triggers. Different triggers result in varying administrative responsibilities when the clawbacks are applied. The clawback provisions could be triggered at the time of the event of fraud or misconduct, the restatement of financial results, or the incidence of employee bad faith (Cherry and Wong 2009). In this study, I also analyze the trigger of clawback provisions using S&P 500 firms. Examples of selected clawback provisions are listed in Appendix A. Clawback Triggers. 立. 政治大. The most prevalent trigger is a restatement due to fraud or misconduct. Many clawbacks are. ‧ 國. 學. drafted to be triggered by violations of firm policy or misconduct actions. This type of clawbacks is perceived as a strategic policy against misconduct or harmful acts. For instance, Monsanto. ‧. Company will return performance-based compensation in the event of a material restatement in the. Nat. sit. y. company’s financials as a result of the misconduct or fraud on the part of the executive officers (see. n. al. er. io. Appendix A). From the clawbacks lists in Corporate Library database, more than 40% of the. i n U. v. clawback adopters employ this type of clawbacks. In fact, the “fraud or misconduct” trigger is consistent with Section 304 of SOX.. Ch. engchi. Clawback provisions can also be triggered by any material mistake or incorrect data found to be in applicable financial statements or performance criteria, regardless of an individual’s knowledge. Some observed firms make no reference to misconduct, and the clawbacks could be triggered when the firm is required to “restate” its financial statement. For example, International. compensation in the financial sector has both reflected and encouraged excessive risk taking. Reforming compensation policies and practices is an essential part of our effort to increase financial stability. We fully endorse the implementation standards of the FSB aimed at aligning compensation with long-term value creation, not excessive risk-taking, including by (i) avoiding multi-year guaranteed bonuses; (ii) requiring a significant portion of variable compensation to be deferred, tied to performance and subject to appropriate clawback and to be vested in the form of stock or stock-like instruments, as long as these create incentives aligned with long-term value creation and the time horizon of risk;……We task the FSB to monitor the implementation of FSB standards and propose additional measures as required by March 2010.” 10.

(24) Paper Company adopted a policy regarding the adjustment and recapture of compensation in the event of a significant restatement of financial results for errors, omission, or fraud (see Appendix A). Under this type of clawbacks, if the financial statements are required to be restated as a result of errors, omission, or fraud, the board may, in its discretion, based on the facts and circumstances surrounding the restatement, direct that firm recover all or a portion of the equity award. Since the trigger to enforce clawbacks is not limited to restatements due to fraud or misconduct, it is closer to the Dodd-Frank Act. Finally, some clawback provisions may have broader trigger which is called “bad faith” conduct. Some firms impose a clawback on cash bonuses paid to certain managing directors, and. 政治大. the provision is triggered if the employee takes a job with a competitor in the specific period. 立. following payment (Kleinman and Harrington 2009). “Bad faith” would include clawbacks. ‧ 國. 學. triggered by a breach of a non-competition clause. Such performance-based triggers raise similar questions of accountability to those noted above. Some firms, like Automatic Data Processing Inc,. ‧. recoup certain amounts of compensation awarded since that time when the participants engage in. Nat. sit. y. activity that is in conflict with or adverse to firms’ interests (see Appendix A). Specifically, these. n. al. er. io. clawbacks have triggers unrelated to erroneous or fraudulent financial data.. i n U. v. In practice, firms choose one or more triggers to strengthen enforceability of the clawback. Ch. engchi. provisions. Although there are a variety of triggers to enforce clawbacks, Section 954 of the Dodd-Frank Act stipulates that firms can initiate their clawbacks only when restatements occur due to material noncompliance with financial reporting rules. The triggers result in different clawbacks enforceability for adopted clawbacks. Enforcement Authority For voluntary clawback adopters, each firm’s unique circumstances, pay schemes, compensation strategy, and culture will determine the features of clawback provisions. In practice, there is significant uncertainty due to a lack of litigation involving enforcement of clawbacks for voluntary adopters. Another enforcement issue is whether the enforcement authority is definite (i.e., 11.

(25) giving directors discretion to waive the clawbacks). It is generally preferable that the compensation committee, given its independence requirements, to be delegated with enforcement responsibility because the inherent complexity offers compensation committees the opportunity to tailor firms’ clawback provisions to its individual needs, goals and existing arrangements (Scott et al. 2010).13 With adequate authority, compensation committees could assess the appropriateness of implementing a clawback policy, and, just as importantly, consider modification of clawback measures that are already in place. Clawback provisions that provide firms considerable discretion in applying and enforcing the provisions will generally leave firms with more options in such cases.14 It is a good practice to. 政治大. determine default procedures to be followed in order to ensure consistent and fair implementation. 立. of the clawback. Arbitrary exercise of discretion in applying a clawback could open the door to. ‧ 國. 學. accusations of bias, retaliation or other bad faith in enforcement actions (Scott et al. 2010). Many firms, like Rockwell Automation, Inc., International Paper Company, Automatic Data Processing,. ‧. Inc., and Rowan Companies, Inc., state clearly that the compensation committee or the board have. Nat. n. al. 2.1.3 Evidence of clawback provisions. Ch. engchi. er. io. I will also consider the enforcement authority in my analyses.. sit. y. the discretion to enforce the clawbacks. Appendix A reports some of these examples. In this study,. i n U. v. Clawback provisions became an important issue in executive compensation in the wake of the 2007-2008 credit crisis (Brown et al. 2011). Therefore are four research lines for clawback provisions adoption. First, some studies discuss the economic determinants of firms’ voluntarily adopting clawback provisions in executive compensation contracts. Addy and Yoder (2011) reports that 29% of the S&P 500 firms have adopted clawback provisions. They conjecture that firms 13. If inside directors or other officers are involved, enforcement decisions could be constrained due to associations with those subject to the clawbacks. Therefore, Dodd-Frank Act in 2010 requires U.S. public firms to have fully independent compensation committee members. 14 However, some reports stated that too much discretion could call into question the firm’s dedication to the clawback’s effectiveness, lessen its deterrence effect and invite shareholder proposals for stricter measures (e.g., Fried and Shilon 2011; Scott et al. 2010). In my opinion, because neither the firm nor any of its shareholders can sue to enforce section 304, SEC has rarely enforced the clawbacks. It is recommended that the enforcement authority should be more definite and the enforcement body has sufficient discretion to take into account unforeseen extenuating circumstances as may be appropriate. 12.

(26) voluntarily adopt the provisions because they are less costly to enforce than equity claims based on unjust enrichment. Prior studies find that firms with more independent governance (Addy et al. 2011), larger firm size (Brown et al. 2011), and previous financial restatements (Gao et al. 2010) are more likely to voluntarily adopt clawback provisions. In addition, influential CEOs (i.e., those where the CEOs are also the chairman of the board) reduce the likelihood that a firm will adopt a clawback provision (e.g., Addy et al. 2011; Brown et al 2011). Second, some research suggests that clawbacks are effective governance mechanisms that improve financial reporting quality and affect auditor behavior. For example: Chan et al. (2012a) uses all firms covered in the Russell 3000 from 2005 to 2009 and shows that voluntary clawback. 政治大. adoptions lead to a reduction in financial misstatements. Also, market reacts favorably to such. 立. voluntary adoption by higher earnings response coefficients, implying that firm-initiated clawbacks. ‧ 國. 學. appear to be an effective deterrent of financial misstatements. Notably audit fees are lower after clawback provisions are adopted because auditors may perceive clawback adopters as associated. ‧. with lower control risk, leading to lower audit risk. In another study, Chan et al. (2012b) finds that,. Nat. sit. y. while clawback-adopting firms reduce accruals management, they increase real transactions. n. al. er. io. management (e.g., reduce R&D expenditures), especially when firms have pressure to meet or beat. i n U. v. earnings benchmarks. Chan et al.’s (2012b) empirical results further show that engaging in more. Ch. engchi. real transactions after clawbacks mainly occurs in firm-years in which the actual earnings meet or just beat earnings benchmarks. These results imply that, although clawbacks deter managers from engaging in financial misreporting, clawbacks do not fully eliminate earnings management. The third research issue is to test market reaction. By adopting clawback policies, firms may signal the quality of their governance is that they can access to more capital with lower costs (Brown et al. 2011). Gao et al. (2010) finds a significantly positive market reaction to the announcement of clawback adoption, as well as a reduction in bid-ask spreads following clawback adoption, particularly in firms with previous restatements. Finally, some studies focus on changes of CEO compensation. Prior literature finds that adopting of clawback provisions appears to 13.

(27) increase executive compensation (e.g., Babenko et al. 2012; Chen et al. 2012; DeHaan et al. 2011) and deduce CEO tenure (e.g., Babenko et al. 2012). Clawbacks adoption is also associated with smaller higher CEO pay-performance sensitivity (e.g., Chen et al. 2012). Appendix B summarizes current studies that have examined the clawback provisions from different aspects. Although prior studies show that clawbacks adoption strengthens earnings quality (e.g., Chan et al. 2012a; DeHaan et al. 2011) and investors have positive reaction to clawback provisions adoption (Gao et al. 2010), little research has discussed how clawback provisions interact with other governance mechanisms to improve financial reporting quality. I attempt to consider the influence of clawback provisions on audit committees’ monitoring process of financial reporting.. 立. 政治大. 2.2 Audit Committee Compensation. ‧ 國. 學. Even though the agency theory suggests that equity-based compensation can align board members’ monitoring incentives with the shareholders (Dalton et al. 2003; Hillman and Dalziel. ‧. 2003; Monks and Minow 2001), the use of stocks and options has became more problematic in. Nat. sit. y. recent years as boards recruit more independent outside directors who are less financially. n. al. er. io. independent than executives who traditionally serve on firms’ boards (Zong 2004).. i n U. v. The stock ownership, which makes audit committees economically dependent on the firms,. Ch. engchi. may increases the affiliation between audit committee members and the firms and, thus, the likelihood that the audit committees side with management (Carcello and Neal 2003). Since audit committees play conflicting roles to manage business operations and oversee board decisions simultaneously (Ezzamel and Watson 1997), equity-based compensation could affect audit committees’ monitoring effectiveness. Two studies use the pre-SOX data and provide similar evidence. Bédard et al. (2004) shows that stock options may reduce audit committees’ monitoring of earnings management to increase either current earnings (positive earnings management) or those of future years (negative earnings management). This is because audit committee members may have a short-term perspective with respect to their ownership stake (Leonhardt 2002), and 14.

(28) equity-based compensation makes outside members become insiders, which contradicts the traditional definition of independence. 15 In another study, Archambeault et al. (2008) uses restatements announced during 1999 and 2002 to show that both short-term and long-term options are associated with higher restatement likelihood. In contrast, Cullinan et al. (2010) uses 456 post-SOX firms that have ICW during 2004 and 2005 to examine the association between the likelihood of ICW and option compensation for audit committees. They report a marginally significant association between stock compensation and the incidence of ICW, and conclude that firms with a stock option plan for their audit committees are significantly more likely to report an internal control weakness. Engel et al. (2010) shows that,. 政治大. firms facing a higher demand for monitoring of the financial reporting process pay higher total. 立. compensation and cash retainers to their audit committees.. ‧ 國. 學. My study differs from the above studies in three aspects. First, since the level of audit committee compensation increases substantially after SOX (Linck et al. 2009), the findings. ‧. reported in Bédard et al. (2004) and Archambeault et al. (2008) may be different in the post-SOX. Nat. sit. y. period. Second, these two studies focus on the analysis of option compensation, and there is a lack. n. al. er. io. of evidence of differential compensation components. It is not clear that, how cash and stock. i n U. v. compensation affect audit committees’ oversight effectiveness. Finally, the implication from. Ch. engchi. Cullinan et al.’s (2010) findings is limited because the reported ICW do not necessarily mean that the audit committees are not effective. It is possible that the audit committees effectively identify the weakness and thus initiate the auditor (and management) to report ICW. Rather than using ICW alone, I use outcome measures to capture the audit committees’ oversight effectiveness. 2.3 Hypothesis Development Prior experimental research has shown that cash compensation does not create direct incentives for audit committees to prefer biased reporting. Magilke et al. (2009) finds that students serving as audit committee members are least biased when there is no stock-based audit committee 15. Prior studies have defined audit committee independence using whether audit committee members are employees or affiliators of the firm (e.g., Klein 2002a; Abbott and Parker 2004; Lennox and Park 2007). 15.

(29) compensation. This suggests that if the audit committees are paid in cash, financial reporting quality shall improve. In a recent empirical research, Engel et al. (2010) shows that firms having higher demand for monitoring financial reporting process pay more cash retainers to audit committees. The cash compensation for audit committee might be a tool to strengthen the effectiveness of monitoring financial reporting. Based on the above findings, I assume that increasing the amounts and portions of cash compensation may improve audit committee independence. I posit the following hypothesis: H1a: The amounts and portions of the cash compensation are negatively associated with audit committees’ oversight failure.. 政治大. Before SOX, the exchanges’ listing requirements provide for the appointment of certain. 立. affiliated directors if the board determines it is in the best interests of the firm for these individuals. ‧ 國. 學. to serve on its audit committee (Klein 2002b; NYSE Rule §303.01[B][3][b]; NASDAQ Rule 4310[c][26][B][ii]). Therefore, many audit committees did not have fully independent outside. ‧. directors before SOX (Klein 1998, 2002b; Vicknair et al. 1993). The pre-SOX studies use the. Nat. sit. y. percentage of outsiders (who may or may not own firms’ shares) on the audit committees (e.g.,. n. al. er. io. Bédard et al. 2004; Klein 2002a; Abbott and Parker 2000) and whether audit committees do not. i n U. v. include employees (e.g., Abbott et al. 2000; Lennox and Park 2007) to measure audit committee. Ch. engchi. independence. Because the inside-affiliated directors probably hold firms’ shares before they became members, the issue of equity-based compensation is not important in the pre-SOX period. However, Section 301 of SOX mandates that audit committees be composed entirely of independent directors, and thus, audit committee members’ stock ownership exists almost from the equity-based compensation. Therefore, an investigation of the association between equity-based compensation and audit committee independence becomes important. When firms reward audit committees by equity-based compensation, an affiliation relation occurs between the audit committees and the firms because audit committee members’ wealth is. 16.

(30) tied closely with firms’ reported performance.16 Archambeault et al. (2008) shows that firms with option compensation for their audit committees are significantly more likely to have higher likelihood of restatements. However, they use the pre-SOX data, and their finding may be different in the post-SOX period because the pre- and post-SOX restatements tend to be caused by different types of misstatements (Hennes et al. 2008). Archambeault et al. (2008) focuses on options purely, and only use restatements as the measure of audit committees’ oversight effectiveness. Cullinan et al. (2010) uses ICW in 2004-2005 as the measure of audit committees’ oversight failures only and finds that firms with option compensation are more likely to report ICW. The concerns about their findings are that ICW may reflect effective audit committees and firms are still in their learning. 政治大. stage in 2004-2005. Thus, Cullinan et al.’s (2010) results may not be convincing. In light of the. 立. potential problem of audit committees compensated by firms’ stocks and options, I consider both. ‧ 國. 學. stock and option compensation in this study and predict a positive association between equity-based compensation and audit committees’ oversight failure. These give rise to the. ‧. following hypotheses:. Nat. n. al. er. io. audit committees’ oversight failure.. sit. y. H1b: The amounts and portions of the stock compensation are positively associated with. i n U. v. H1c: The amounts and portions of the option compensation are positively associated with. Ch. engchi. audit committees’ oversight failure.. Recently, shareholders, legislators, and compensation reform advocates have endorsed clawback provisions as a tool to prevent executives and employees from retaining undeserved windfalls and to enhance pay-for-performance initiatives (Scott and Bradley 2010). While. 16. With respect to the qualification of the independence, the SOX states that an audit committee member cannot accept any fees from the company other than for serving as a director, and cannot be an affiliated person of the company or any of its subsidiaries. Under NYSE rules approved on August 1, 2002, audit committees must consist of a minimum of three members. To be independent, a director must not have any relationship with the company that interferes with the exercise of independent judgment, and must not have worked for the company within the past three years. NASDAQ’s board of directors approved similar rules on July 24, 2002, and amended them on August 21, 2002, to reflect certain provisions of the Sarbanes-Oxley Act. AMEX’s board of directors also approved new corporate governance rules conforming to the Act in September 2002. Accordingly, companies should ensure that current and new audit committee members have no potential conflicts of interest that may interfere with their ability to act independently from management (Buchalter et al. 2003). 17.

(31) clawback provisions can be employed for a variety of purposes, many clawbacks are drafted to be triggered under violations of firm policy or ethical misconduct. Desai et al. (2006) points out that, if managers understand that their fraudulent behavior will be penalized ex post through a loss in wealth, job, and/or reputation, they should have less incentive to engage in earnings manipulation ex ante. Similarly, Caskey et al. (2010) indicates that the misreporting penalty affects not only audit committees’ ex post reporting process, but also their ex ante information-collection process. Therefore, clawback provisions may serve as an effective tool in strengthening audit committees’ oversight effectiveness. Since clawback provisions impose restrictions against illusory gains for board compensation, it. 政治大. is reasonable to assume that clawbacks could improve board governance.17 Specifically, firms may. 立. be more interested in ensuring that incentive pay is based on appropriate time horizons because. ‧ 國. 學. clawbacks can offer a backstop against payments of bonuses (including cash, stocks, and options) based on fleeting or illusory gains in performance metrics (Scott and Bradley 2010).18 Since the. ‧. equity-based compensation may weaken audit committees’ independence while clawback. Nat. sit. y. provisions could improve audit committees’ oversight effectiveness, the issue of whether the. n. al. er. io. positive effect of clawback provisions outweighs the negative effect of equity-based compensation. i n U. v. becomes an empirical one. Currently, no studies have ever examined this issue. I thus posit the following hypothesis:. Ch. engchi. H2: Firms’ voluntary adoption of the clawback provisions will change the association between equity-based compensation and audit committees’ oversight failure, if exist.. 17. The litigation risk due to firms’ recoupment enforcement is expected to affect board governance. The growth of shareholder litigation against directors coupled with the media attention and reputational damage to the directors who are sued, and to some extent to all directors. Lipton (2008) indicates that, to compound pressures on boards, shareholder litigation and other public attacks on board members have been undermining the willingness of some of the most qualified individuals to serve as directors. 18 Engel et al. (2010) indicate that cash compensation for audit committee is positively associated with firms’ demand for financial reporting monitoring. Chan et al. (2012a) suggest that clawbacks are effective governance mechanisms that improve financial reporting quality. Therefore, it is expected that firms using more cash compensation are less likely to have audit committees’ oversight failure when firms adopt clawback provisions. 18.


(33) CHAPTER 3. RESEARCH DESIGN. 3.1 Data and Sample Selection The list of clawback adopters is available from the Corporate Library database. There are 195 S&P 500 firms adopting clawback provisions. Twenty-six financial institutions (SIC codes 6000-6999) are excluded because of their unique operating environment and accounting practices. To ensure that firms’ clawback provisions are applied to audit committees, I review the content of clawback provisions and further eliminate 38 firms whose clawback provisions are applied only to CEOs or executive board members but not the audit committees.19 Two firms changing their clawback provision coverages are also deleted from the sample. The final clawback adopters. 政治大. consist of 129 firms. Panel A of Table 1 presents the sample selection procedure.. 立. [Insert Table 1 here]. ‧ 國. 學. The original sample consists of 3,500 S&P 500 firm observations during fiscal years from 2003 to 2009.20 Five hundred and ninety-five financial institutions (SIC codes 6000-6999) are. ‧. excluded because of their unique operating environment and accounting practices. Three hundred. Nat. sit. y. and fifty-three observations are excluded due to incomplete financial and corporate governance. n. al. er. io. data. I further hand-collect audit committees' compensation and background information from. i n U. v. firms’ definitive proxy statement (Form DEF 14A) using the EDGAR database. One hundred and. Ch. engchi. forty-five observations are eliminated due to missing compensation data. The above procedures give rise to the sample consisting of 2,407 firm-year observations, which are used to estimate firms’ propensity scores. I use a matched-pair procedure and obtain matched sample by identifying the pairings that result in observations with the smallest propensity score differences. Because there are 281 firm-year observations with voluntary adoption, 562 observations are used in the analysis. Thirty-eight observations are excluded due to missing ICW information. Panel B of Table 1. 19. The survey results from TheCorporateCounsel.net report that about 25.9% of clawbacks only apply to executive officers. 20 I use S&P 500 firms in this research because of the greater disclosures of audit committee compensation for larger size firms. Since firms with high market values are more likely to disclose corporate information voluntary (Bamber et al. 2010), this compensation data is available for most S&P 500 firms. 19.

(34) presents this sample selection procedure. I specifically identify the years of adopting the clawback provisions from firms’ proxy statements. I use restatements coded in Audit Analytics to identify firms that restate their financial statements during the sample period. I also identify ICW firms from Audit Analytics if firms reported internal control weaknesses in their SOX Section 404 report. 21 Each sample firm’s financial data are collected from Compustat. Other corporate governance information is collected from ExecuComp, CRSP, and Audit Analytics. 3.2 The Endogenous Biases Firms’ decisions to voluntarily adopt clawback provisions are endogenous prior to the. 政治大. enactment of Dodd-Frank Act. To ensure that the expected correlation between clawback. 立. provisions and audit committees’ effectiveness is not driven by the determinants of this voluntary. ‧ 國. 學. adoption feature, I use propensity score matching to control for this endogenous (self-selection) problem.. ‧. I employ propensity score matching but not Heckman’s (1979) two-stage model for two. Nat. sit. y. reasons. 22 First, Heckman et al. (1997) selection model is more likely to suffer from. n. al. er. io. multicollinearity problems when there are no exclusion restrictions (i.e., which of the independent. i n U. v. variables in the first stage model should be excluded from the second stage model). If there are no. Ch. engchi. exclusion restrictions, the inverse mills ratio is correlated with the independent variables in the second stage (Manning et al. 1987; Puhani 2000; Li and Prabhala 2007).23 Because there is no inverse mills ratio variable under propensity score matching and so it is not required to impose exclusion restrictions (Heckman et al. 1997; Heckman and Navarro-Lozano 2004). Second, the 21. I focus on ICW under Section 404 because firms with ineffective internal controls may not discover or disclose their ICW under Section 302 if these firms have weak governance systems. Therefore, Section 302 reporting may give rise to weak association between corporate governance quality and internal control quality (Hoitash et al. 2009). See Hoitash et al. (2009) for discussions of these two regulatory regimes. Also ee Schneider et al. (2009) for prior studies examining internal control issues after SOX. 22 Previous studies examining the existence of endogenous biases typically used Heckman’s (1979) two-stage model (e.g., Fan and Wong 2005; Khurana and Raman 2004; Mansi et al. 2004; Louis 2005; Rajan and Servaes 1997, Weber and Willenborg 2003). Heckman (1979) include the inverse mills ratio estimated from the first-stage logistic model as an additional explanatory variable in the second-stage regression model. However, Lennox et al. (2012) analyzes the inherent limitations and fragility of Heckman (1979) two-stage model. 23 The inverse mills ratio is nonlinear in Heckman et al.’s (1997) arguments. 20.

(35) propensity score matching mitigate the selection biases due to observables but the Heckman two-stage model address the selection biases due to unobservables (Lennox 2012; Tucker 2010).24 Because the determinants of voluntary adoption of clawback provisions obtained from prior research are observable variables, I use propensity score matching to control for potential self-selection bias. While Heckman’s (1979) two-stage model uses a specific functional form to provide an indirect estimate of the treatment effects, the propensity score matching does not rely on a specific functional form but provide a more direct estimate of the treatment effects (Li and Prabhala 2007). 3.2.1 Propensity score matching. 政治大. Rosenbaum and Rubin (1983) develops the propensity score matching as a way to address. 立. matched pair problem and assess hidden bias within a broader sample. The propensity score is a. ‧ 國. 學. conditional probability of receiving some level of treatment given the observable covariates (Armstrong et al. 2010). In the case where a binary treatment is present (i.e., treatment or no. ‧. treatment), matched pairs are formed by selecting an observation that received the treatment and. Nat. sit. y. selecting another observation with the closest propensity score that did not receive the treatment.. n. al. er. io. Therefore, propensity score matching models match observations based on the probability of. i n U. v. undergoing the treatment, which is the probability of adopting clawback provision in this paper.. Ch. engchi. Since I use clawbacks adoption as the treatment, matching sample becomes an optimization problem of minimizing a function of the probabilities to adopt clawbacks between the propensity scores of the broader sample (Armstrong et al. 2010). Identifying the control firms via propensity score matching helps ensure that the observable characteristics that affect both the probability of clawback implementation and the probability of correlated omitted events are controlled in the analysis (Lemmon and Roberts 2010). According to Armstrong et al. (2010) and Lawrence et al. (2011), I use propensity score 24. Selection bias due to “observables” results from a failure to control for differences researchers can observe. Examples of observable differences are firm size and growth. Examples of unobservables are information revealed during a financial audit that is known to some market participants or other information that is publicly disclosed by the company but is too costly for researchers to collect (Tucker 2010). 21.

(36) matching to control for the endogenous biases in the following way. First, I estimate a logistic propensity score model, which provides the probability that firms will adopt clawback provisions (i.e., the treatment) conditional on observable features of the contracting environment. The propensity score is estimated by including determinants to adopting clawback provisions into this propensity score model. Second, the developed propensity score model is then used to calculate firms’ probabilities of adopting clawback provisions. Particularly, I find matched firms by identifying the pairings that result in observations with the smallest propensity score differences (i.e., the most similar observed contracting environments). Using observations collected by propensity score matching, the effect of clawback provisions on the audit committees’ effectiveness. 政治大. is then inferred from the estimated coefficient on clawback adoption while other control variables. 立. are included in the regression estimation.. ‧ 國. 學. 3.2.2 Voluntary adoption model Incentives to Adopt Clawback Provisions. ‧. Some research investigated the economic determinants of firms’ decisions to voluntarily adopt. Nat. sit. y. clawback provisions (e.g., Addy et al. 2011; Brown et al. 2011). These determinants include. n. al. er. io. firm-specific incentives, the CEO’s influence, and firm’s governance characteristics. Under. i n U. v. propensity score matching, the voluntary adoption model is used to calculate each observation’s. Ch. engchi. propensity score. I then match each clawback firm to a non-adopting control firm using firms’ propensity score. The voluntary adoption model is as follows: P(CLAWBACK ) i ,t   0  1 LnASSET i ,t-1   2 PROFIT i,t-1   3 PRIORSTAT i ,t -1   4 EQUITY _ Issuei ,t -1   5 DEBT _ Issuei ,t -1   6 EXTRA _ Bonus i ,t -1   7 CEO _ Tenurei ,t -1   8 CEO _ Chairi ,t -1   9 Bonus to cash. (CLAW). i ,t -1.  10 CEO _ Ownershipi ,t -1  11 BSIZE i ,t -1  12 INSIDE _ % i ,t -1   i ,t. where the definitions of all the variables are summarized in Table 2. All independent variables are lagged by one year so that the likelihood of adoption during each 22.

(37) year depends on the firm. I include firm size (denoted by LnASSET) because larger firms, which are more likely to attract more attention in the capital markets (Aboody et al. 2004; Barton 2005), are more likely to adopt clawback provisions to enhance their reputation (Brown et al. 2011). I predict the coefficient on LnASSET be positive. Since more profitable firms are more likely to adopt clawbacks (Brown et al. 2011; DeHaan et al. 2011), and I use firms’ net income (denoted by PROFIT) to proxy for firms’ profitability and predict its coefficient to be positive. Recent studies indicate that prior restatements significantly increase firms’ likelihood to adopt clawback provisions (e.g., Brown et al. 2011; Gao et al. 2010) because past restatements are more salient to the boards (Addy et al. 2011). I control for restatements in the past five years (denoted by. 政治大. PRIORSTAT) and predict its coefficient to be positive. Moreover, since adopting clawback policy. 立. would establish firms’ reputation about the credibility of its governance (Aboody et al. 2004;. ‧ 國. 學. Barton 2005) to ensure a lower cost of capital, firms issuing equity and debts in the capital market are more likely to adopt the clawback provisions to send credible signals to the market (Brown et al.. ‧. 2011). I include firms’ issuance of equity (denoted by EQUITY_Issue) and debts (DEBT_Issue) in. Nat. sit. y. the prior year as two indicator variables and predict their coefficients to be positive.. n. al. er. io. While Bliss and Rosen (2001) reports that CEOs are often rewarded for engaging in. i n U. v. acquisitions, even if these activities are value-destroying, Brown et al. (2011) suggests that firms. Ch. engchi. that have paid significant bonuses related to a merger or acquisition might adopt clawback provisions to rescind bonuses for a merger or acquisition that is subsequently unsuccessful. I control for extraordinary CEO compensation in M&A bonuses (denoted by EXTRA_Bonus) in the logistic regression model and predict its coefficient to be positive. I also control for CEOs’ power on firms’ likelihood of adopting the clawback provisions. The executives having longer tenures with the firm could have significant influence over firms’ major operation decisions (Bushman et al. 2004). In addition, if CEOs are the chairman of firms’ board, the executives seem to strengthen CEO power (e.g., Jensen 1993, Core et al. 1999; Bebchuk et al. 2002; Adams et al. 2005). Higher relative compensation and higher CEO stock ownership reduce 23.

(38) the influence of the board and thus increase CEO power (Lisic et al. 2011). I use four proxies for CEO power: the number of years the executive has served as CEO for the firm (denoted by CEO_Tenure), whether the CEO is the chairman of the board (denoted by CEO_Chair), the ration of CEO bonus to cash compensation (denoted by Bonus to cash) and percentage of firm’s shares owned by the CEO (denoted by CEO_Ownership). I predict their coefficients to be negative because CEO power is likely to reduce the likelihood that firms adopt clawback provisions (Addy et al. 2011; Brown et al. 2011). Browen et al. (2011) indicates that managerial power, which is measured by the number of directors on the board, is negatively associated with firms’ likelihood of adopting clawback. 政治大. adoptions. Therefore, I include board size (denoted by BSIZE) in the CLAW model and predict its. 立. coefficient to be negative. Addy et al. (2011) finds that firms with more independent governance. ‧ 國. 學. are more likely to adopt clawback provisions. I thus include the percentage of inside directors (denoted by INSIDE_%) in the CLAW model and predict its coefficient to be negative.. ‧. 3.3 Regression Model. Nat. sit. y. This study extends the audit committee compensation literature by investigating the link. n. al. er. io. between clawback provisions and audit committees’ effectiveness. I use four measures to proxy for. i n U. v. audit committees’ oversight failure: the likelihood of restatements, the incidence of ICW, and the. Ch. engchi. level of accruals quality and real earnings management. Each of these oversight failure measures is discussed below. 3.3.1 Restatement likelihood I use restatement likelihood to proxy for audit committees’ oversight failure because SOX expends audit committees' responsibilities to assure that financial statement accurately portray companies’ economic activities (Laux and Laux 2009). To test whether the clawback provisions enable compensation policy more efficient for audit committees, leading to less likelihood of restatements, I estimate the following logistic model following Archambeault et al. (2008), Efendi et al. (2007), and Palmrose et al. (2004): 24.

(39) RESTATED i ,t   0  1 LnASSET i ,t   2 BIG 4 i ,t   3GOINGi ,t   4 M & Ai ,t   5 ROA_ind i ,t   6 MBi ,t   7 ACSIZE i ,t   8 OVERLAPCOM. i ,t.   9 ACCEXPERT i ,t  10 MEETING i ,t. (REST).  11CLAWBACK i ,t  12 Compensation _ Variables i ,t   13CLAWBACK  Compensation _ Variables i ,t   i ,t. where the definitions of all the variables are summarized in Table 2. Note that I include industry fixed effects and year fixed effects as controls for unobserved firm-level heterogeneity over time (Bowen et al. 2010; Linck et al. 2009). The fixed-effect model helps alleviate the endogeneity problem caused by the omitted variables (Campa and Kedia 2002). [Insert Table 2 here] Dependent variable. 立. 政治大. The dependent variable, RESTATED, is a dummy variable that equals 1 if a firm’s year t. ‧ 國. 學. financial statements are restated and 0 otherwise. Instead of using whether or not firms announce. ‧. restatements in year t, variable RESTATED provides a more appropriate test of the association between audit committees’ compensation and restatement likelihood because outside directors. y. Nat. io. sit. serving on year t’s audit committees are responsible for overseeing year t’s financial statements and. n. al. er. receive year t’s compensation. The use of restatement announcement year will mismatch the year. Ch. i n U. v. audit committees exercise their oversight responsibility and the year they receive compensation. I. engchi. thus use RESTATED to proxy for audit committees’ oversight failure and predict that the association between audit committee compensation and financial reporting failure is a moderated by clawback provisions. Control Variables In the REST model, I include major firm characteristics that are likely to affect the likelihood of restatements. Similar to previous studies (e.g., Dechow et al. 1996; Richardson et al. 2002; Desai et al. 2006), I control for firm size (denoted by LnASSET) and predict its coefficient to be negative because size might capture firm-specific risk (Fama and French 1995) and larger firms are more likely to be subjected to closer scrutiny by regulators and investors (Balsam et al. 2003; Romanus 25.