اثر کاستی های کنترل داخلی بر سودمندی درآمدهای حاصل از جبران خسارت اجرایی

Since SOX 404 disclosures are informative about earnings, and due to the widespread practice of using earnings-based measures in executive compensation, this study examines whether reports of internal control material weaknesses (ICMW) under SOX 404 influence firms' reliance on earnings in tying executive pay to performance. Using 391 (366) firm-year observations with reported ICMW and 3648 (3138) firm-year observations for CEOs (CFOs) reporting NOMW under SOX 404, we find a decreased strength in the association between earnings and executives' (CEO and CFO) compensation when the firm reports an ICMW, and as the number of reported ICMW increases. In addition, we find this decreased weight on earnings for the more severe Company-Level than Account- Specific material weaknesses. Our study suggests that the ICMW report under SOX 404 provides incremental information for executive compensation beyond that contained in reported earnings.

The accounting scandals at firms such as Enron and WorldCom highlighted deficiencies in corporate governance that were characterised by low financial reporting quality and disproportionate pay-for-performance.2 To discipline firms and restore investor confidence, legislative authorities enacted the Sarbanes–Oxley Act. Among the reforms is Section 404 of SOX (SOX 404) which requires both the management and the external auditor to report on the adequacy of a firm's internal control over financial reporting. Prior research shows that, relative to non-disclosing firms, firms reporting material weaknesses in internal control (ICMW) have inferior accruals and earnings quality (Ashbaugh-Skaife et al., 2008, Bedard, 2006 and Doyle et al., 2007b), and lower earnings–returns coefficients (Chan, Farrell, & Lee, 2008). Given the fact that SOX 404 disclosures are informative about earnings, and due to the widespread practice of using earnings-based measures in executive compensation, this study examines whether reports of ICMW under SOX404 influence firms' reliance on earnings in tying executive pay to performance. A long line of research shows that earnings-based performance measures are commonly used to motivate and reward executives because such measures correspond to manager actions (Gjesdal, 1981). However, there are two drawbacks to using earnings to evaluate executive performance. First, because executives know how their actions impact earnings, they can manipulate this measure to increase their wealth. Second, earnings do not fully reflect the long-term implications of recent executive decisions. Based on these factors, firms place varying weights on earnings in compensating their executives, and the weights are determined by how sensitive earnings are to effort and on the precision, or lack of noise, with which it reflects executives' actions ( Banker and Datar, 1989 and Lambert and Larcker, 1987). However, there is evidence that CEOs are shielded from certain negative events, such as firm restructuring (Dechow, Huson, & Sloan, 1994), or above the line losses (Gaver & Gaver, 1998). Our main focus of inquiry is significant because it builds on this line of research by showing that the sensitivity of compensation–performance relations varies cross-sectionally with the quality of the system producing the earnings information. Weak internal controls potentially permit accounting errors to occur and go undetected, increasing unintentional errors in accrual estimation and/or facilitating intentional earnings management (Doyle et al., 2007b). A report of an internal control deficiency, therefore, signals that the manager is unable to provide reasonable assurance regarding the quality of reported earnings (Ashbaugh-Skaife et al., 2008, Bedard, 2006, Chan et al., 2008 and Doyle et al., 2007b). Using the sensitivity-precision framework, the errors introduced by weak internal controls are likely to result in earnings that capture executives' effort with low precision, diminishing its use as an assessment tool for evaluating managers' performance. Motivated by the optimal contracting hypothesis, we posit that firms with ICMW report earnings with lower precision-sensitivity. The purpose of this study is to examine whether firms reporting ICMW place relatively less weight on earnings compared to firms reporting no ICMW (NOMW). In other words, whether ICMW reports influence compensation contracts is an empirical issue because, although ICMW firms report lower earnings–returns coefficients (Chan et al., 2008), prior research does not suggest any direct association between the valuation role of earnings and its usefulness in compensating executives (Bushman, Engel, & Smith, 2006). Under the null hypothesis, the sensitivity of executive compensation to earnings is unaffected by internal control deficiency. Consistent with prior research, we also examine the weight placed on earnings for firms reporting two different types of ICMW: Account-Specific and Company-Level weaknesses (Doyle et al., 2007a and Doyle et al., 2007b). Account-Specific (AS) material weaknesses arise from routine firm operations and may be resolved by additional substantive auditing procedures. When AS material weaknesses are identified, executives or auditors can easily audit around them by performing additional substantive procedures. Company-Level (CL) material weaknesses, on the other hand, are less easily resolved by auditor involvement and result from lack of resources or inexperience in maintaining an effective control system. Due to the pervasiveness of CL material weaknesses, the scope of audit efforts needs to be frequently expanded to deal with these more serious concerns regarding the reliability of financial statements (Moody's, 2006 and Public Company Accounting Oversight Board (PCAOB), 2004). The extent to which auditors are able to mitigate the negative effect on earnings of these two types of weaknesses would suggest less noise/greater precision in earnings from Account-Specific relative to Company-Level weaknesses. The impact of precision times sensitivity on the weight of earnings for Account-Specific vs. Company-Level material weaknesses is therefore the second empirical question. Using 391 (366) firm-year observations with reported ICMW and 3648 (3138) firm-year observations for CEOs (CFOs) reporting NOMW under Section 404 of SOX, we find a decreased strength in the association between earnings and CEO compensation when the firm reports an ICMW, and as the number of reported ICMW increases. Our results are also robust to controls for various firm characteristics that prior studies have found to influence the role of earnings in compensation contracts, including earnings quality proxies such as earnings persistence (Baber, Kang, & Kumar, 1998) and corporate governance characteristics (Chhaochharia & Grinstein, 2009). In addition, for CL material weaknesses, we find evidence of a lower strength in the earnings–compensation relation for the CEOs. We find no such result with AS material weaknesses suggesting that only CL weaknesses affect the weight placed on earnings in compensating CEOs. This study makes two contributions. First, it contributes to existing literature by making an examination of the role of earnings as a performance measure in executive compensation contracts (Bushman et al., 2006 and Sloan, 1993; among others), and by examining how information on the quality of a firm's internal controls influences the earnings–compensation relation. We confirm that weak internal controls result in a diminished role for accounting measures in the CEO compensation relation, consistent with optimal contracting. Specifically, it is the firms with CL weaknesses that reduce the weight on earnings in CEO cash compensation. Overall, our findings suggest that the information in the ICMW report is incremental to, or more timely than, that provided by discretionary accruals or earnings persistence measures. Second, our study extends a growing body of literature on the relation between executive compensation and ICMW in the post-SOX era. Carter, Lynch, and Zechman (2009) show that the implementation of SOX in 2002 led to a decrease in earnings management, and that firms responded by placing more weight on earnings in bonus contracts for CEOs and CFOs in the post-SOX period. Another study by Hoitash, Hoitash, and Johnstone (2009) suggests that the compensation of the CFO, who has primary responsibility for the quality of the firm's internal controls, is penalized for reports of ICMW. Since prior evidence shows, and stresses the importance of, a performance-based compensation penalty for internal control quality as a non-financial performance measure in the evaluation of executives, our study further investigates whether an ICMW impacts the weight of earnings in compensation contracts under the mandate of SOX 404. We show that firms consider the strength of an earnings generation system and specifically choose to reduce emphasis on earnings-based performance measures in determining CEOs' cash compensation. The next section of this paper provides background information on the internal control disclosure practices required by the Sarbanes–Oxley Act, discusses the usefulness of earnings as a performance measure and further develops our hypotheses. The third section describes our sample and research design. The fourth section presents our descriptive statistics, results and sensitivity analyses. The fifth section concludes the paper.

We examine whether disclosing an internal control deficiency under Section 404 of the Sarbanes–Oxley Act influences the usefulness of accounting earnings in executive compensation contracts. We argue that when firms disclose internal control deficiencies, earnings are less useful measures of executive performance than are the earnings of non-reporting firms. Using a sample of executive firm-years with and without reported ICMW under SOX 404, we find less strength in the relationship between executive pay and accounting measures of performance for CEOs and CFOs of firms reporting internal control deficiencies. Our results hold after controlling for earnings persistence, discretionary accrual use, governance oversight and controls for endogeneity bias. Therefore, our findings suggest that the ICMW report under SOX 404provides incremental information for executive compensation beyond that contained in reported earnings. We further analyze the effect of the type of ICMW (Company-Level vs. Account-Specific), as defined by Doyle et al., 2007a and Doyle et al., 2007b, on the executive earnings–compensation relation. The results show that firms place less weight on earnings for compensation for Company-Level weaknesses. Our study is subject to three caveats. First, we use the disclosure of ICMW under Section 404 to proxy for the actual presence of an internal control problem. Section 404 does not mandate that firms improve their internal controls; rather it suggests that firms devote adequate resources and attention to the maintenance of effective controls. Thus, to the extent that there is a systematic bias in the choice to identify and disclose ICMW, our sample may not represent the true underlying population of firms with ICMW problems. Second, the sample used in this study also suffers from a potential survivorship bias problem. Because the empirical test requires compensation data from ExecuComp, small firms or those with very low executive compensation may not have made the final sample used in empirical tests, biasing the results in favour of the hypotheses. Third, CFOs are only included in the ExecuComp data, and in our sample, if they are among the top five highest-paid executives of the firm. To the extent that CFO pay reflects their importance within a firm, the results may not apply to firms whose CFOs are not among the five highest-paid executives. Notwithstanding these caveats, our findings have implications for practitioners and researchers. With the potential to make a direct impact on practice, the findings support the notion put forward by the SEC: that firms' internal control environment is important for contracting and decision making. In particular, our paper provides further evidence that the quality of a firm's reporting system is useful for compensation contracting.