文献出处：Scott Duellman. Equity Incentives and Earnings Management[J].
Account. Public Policy ,2014(32):495–517.
原文
Equity Incentives and Earnings Management
Scott Duellmana Abstract
Prior studies suggest that equity incentives inherently have both an interest alignment effect and an opportunistic financial reporting effect. Using three distinct proxies for earnings management we find evidence consistent with the incentive alignment (opportunistic financial reporting) effect of equity incentives increasing as monitoring intensity increases (decreases). Furthermore, using the accrual-based earnings management and meet/beat analyst forecast models we find that the opportunistic financial reporting effect of equity incentives dominates the incentive alignments effect for firms with low monitoring intensity. Using proxies for real earnings management, we find that the incentive alignment effect dominates the opportunistic financial reporting effect for high and moderate monitoring intensity firms. However, for low monitoring intensity firms the opportunistic reporting effect mitigates, but does not completely offset, the benefits of the incentive alignment effect. Overall, these findings are
consistent with the level of monitoring affecting the relation between equity incentives and earnings management. 1. Introduction
Classical agency theory suggests that equity incentives align
’’managers
interests
with
shareholders
interests
(see
for
example, Mirlees, 1976, Jensen and Meckling, 1976 and Holmstrom, 1979). However, recent theoretical papers suggest that equity incentives may also motivate managers to boost short term stock prices by manipulating accounting
numbers
(see
for
example, Bar-Gill
and
Bebchuk,
2003 and Goldman and Slezak, 2006). Empirical studies examining the effect of equity incentives on earnings management, a proxy for opportunistic reporting, yield mixed results. For example, Gao and Shrieves, 2002,Bergstresser and Philippon, 2006 and Weber, 2006, and Cornett et al. (2008) document a positive relation between equity incentives and accrual-based earnings management; while Hribar and Nichols (2007) find that after controlling for cash flow volatility the relation between equity incentives and earnings management becomes insignificant.1 Furthermore, Cohen et al. (2008) find a negative relation between equity incentives and real earnings management. Thus, whether equity incentives are associated with opportunistic financial reporting is an open empirical question that warrants further study.
We view equity incentives as one element of the firm
’s governance
structure and argue that equity incentives inherently have both an interest alignment effect and an opportunistic financial reporting effect. We investigate how the relation between equity incentives and earnings
’management changes with respect to the intensity of firms
monitoring
systems. More specifically, we expect that when monitoring intensity is relatively high, equity incentives will have more of an incentive alignment effect leading to lower earnings management in comparison with low monitoring intensity firms. Conversely, when monitoring intensity is relatively low, equity incentives will have more of an opportunistic financial reporting effect leading to higher earnings management in comparison to high monitoring intensity firms. Thus, we predict that the incentive alignment (opportunistic financial reporting) effect of equity incentives increases as monitoring intensity increases (decreases).
Using a sample over the time period 2001–2007, we proxy for earnings management using three different measures common in the literature: (i) absolute abnormal accruals, (ii) real earnings management measures, and (iii) the likelihood of meeting/beating an analyst forecast. We measure equity incentives, in a manner consistent with prior studies such as Bergstresser and Philippon (2006) as the percentage of total CEO compensation for the year that would come from a 1% increase in the companys stock as of the end of the previous fiscal year.
To measure the intensity of monitoring mechanisms, we focus on three
’financial reporting decisions (board of directors, external auditors, and institutional investors). We identify six board characteristics, one auditor characteristic, and two institutional investor characteristics that could potentially affect monitoring effectiveness. Using principal component analysis we collapse these nine characteristics into two monitoring intensity measures (principal components) which capture 51.1% of the variance in these characteristics.2 We classify firms as high (low) monitoring intensity firms if both monitoring intensity measures are above (below) median values while firms with only one monitoring factor above the median are classified as moderate monitoring intensity firms. We use this approach as different monitoring attributes may be substitutes or complements to one another and principal component analysis effectively reduces the redundancy in these variables.
We regress our measures of earnings management on lagged equity incentives, monitoring intensity classifications (moderate and low), the interaction between them, and a set of control variables. Our findings can be summarized as follows. First, we find evidence consistent with the incentive alignment (opportunistic financial reporting) effect of equity incentives increasing as monitoring intensity increases (decreases) across all three earnings management measures. Second, in tests using accrual based earnings management and meet/beat analyst forecasts, we find that for
’mechanisms that are most directly involved in monitoring managers
low monitoring intensity firms, the opportunistic reporting effect dominates the incentive alignment effect of equity incentives; and equity incentives and earnings management are unrelated when monitoring intensity is moderate or high.
Third, with respect to real earnings management, we find a negative relation between equity incentives and real earnings management for high and moderate monitoring intensity firms. Furthermore, for low intensity monitoring firms the negative relation is mitigated, but not completely offset, by the incentive alignment effect. In contrast with our abnormal accrual results, these findings suggest that the incentive alignment effect dominates the opportunistic financial reporting effect with respect to real earnings management. A potential explanation for these findings is that both monitors and managers are aware of the higher potential long-term costs of real earnings management and thus tend to avoid cuts to discretionary expenses (research and development) or increase production.
Our primary contribution to the literature on the relation between equity incentives and earnings management is that we provide evidence on how this relation varies with the level of oversight by monitoring mechanisms. This is in contrast with most prior studies in this area that either overlook the effects of monitoring (or governance) mechanisms or simply use one or more governance characteristics as control variables (Bergstresser and Philippon, 2006 and Cornett et al., 2008).3