CEOs are rewarded through either relative performance evaluation or pay-for-luck contract agreements when one or the other leads to increased compensation. But CEOs are not rewarded through both simultaneously. CEO pay is shielded from the worst performance outcomes. These compensation patterns are even stronger when firms have strong corporate governance mechanisms in place.
What’s Inside?
The authors analyze both the level of CEO compensation and the sensitivity of CEO compensation to firm performance. The goal is to determine whether there is evidence that CEOs influence compensation committees to their benefit in order to expropriate shareholder wealth. The authors overlap two strands of research. The first addresses relative performance evaluation (RPE), to which a CEO is subject when his or her compensation depends on firm performance relative to a reference group or benchmark. The second strand investigates pay-for-luck (PFL), which exists if CEO compensation is affected by broad industry performance.
How Is This Research Useful to Practitioners?
The authors contribute to the debate regarding whether executive compensation is too big and/or too unconnected to performance. The authors build on previous research on PFL by analyzing the sensitivity of executive compensation to the systematic component of firm performance. They also add an industry performance dimension to the test of the asymmetrical implementation of RPE. In addition, the authors test whether CEOs extract more compensation when they beat their industry benchmark at a time when industry performance is positive.
They show that CEOs are rewarded through either RPE or PFL when one or the other leads to increased compensation. Importantly, CEOs do not receive additional pay for beating their industry benchmark when luck is good. Moreover, the authors find that firms with strong governance compensation committees prevent CEOs from double-dipping to a larger extent than firms with weak governance.
The authors also show that CEOs are shielded from the worst performance outcomes. Although they find that there is an asymmetrical implementation of RPE, this asymmetry does not indicate CEO rent extraction.
There are two key reasons this research should be of interest to those involved in the executive compensation debate, including shareholders, activists, regulators, and academics. First, the evidence shows that CEO pay is tied to firm-specific performance and CEOs are protected from only the worst outcomes. Second, the results show that CEOs’ ability to expropriate shareholder wealth through compensation is limited. Specifically, CEOs do not double-dip by receiving extra compensation for riding strong industry performance and beating their industry benchmark.
How Did the Authors Conduct This Research?
Data from the Compustat ExecuComp database are used, which includes data through the end of 2008. Although 27,278 CEO years of data were available initially, the authors eliminate almost 17,000 observations that would obscure the analysis (for example, observations that include a non-December fiscal year-end). This cleaning of the data leaves 10,475 CEO years available for the regression analyses.
The authors argue that because firms may make changes to such things as target performance measures and payout thresholds from one year to the next, they focus on compensation levels (rather than changes) and control for unobservable variation by using a fixed-effects specification.
Return on assets (ROA) is used as the performance measure. To gauge the implementation of RPE, ROA is decomposed into unsystematic (i.e., firm-specific) and systematic (i.e., industry-wide) components by regressing ROA on two industry-wide measures: equally weighted ROA and equally weighted 12-month cumulative stock returns.
To identify whether there is an asymmetrical execution of RPE and PFL, the authors separate the firm years in the sample by whether firm performance was above or below an industry benchmark and by whether the industry benchmark was positive or negative.
Data from the ExecuComp and RiskMetrics databases are merged to produce a measure of “Strength of Corporate Governance,” which is composed of eight attributes.
The authors acknowledge limitations to their research, such as the existence of relative performance measures other than industry accounting performance and the possibility that measurement error can misclassify some systematic performance as firm specific and vice versa.
Abstractor’s Viewpoint
A key contribution of this research is the investigation into the interactive effect of RPE and PFL. On this front, the authors show that CEOs are rewarded for either beating their benchmark or for good luck, but not both. The results indicate that compensation committees of firms with stronger governance already prevent CEOs from double-dipping to a larger extent than that observed among firms with weaker governance. As such, recommendations to strengthen corporate governance should be focused on current firms with weak governance, which would help avoid inefficient investment by some firms to comply with any new regulations.