CEO Bonuses, Relative Performance Evaluation, and Pay-for-Luck

This study investigates whether the inconsistent findings on the implementation of relative performance evaluation (RPE) can be explained by CEOs' control over their pay process and by the interaction of RPE with pay-for-luck. Using a sample of CEO bonus compensation awards from 1992 to 2008, we test whether there is asymmetric use of RPE and pay-for-luck that indicates CEO power over the compensation process. Specifically we identify the conditions under which a CEO would or would not prefer, ex post, the implementation of RPE and pay-for-luck and analyze whether the sensitivity of bonus compensation to industry-adjusted firm performance varies depending on industry performance and firm performance relative to the industry. Our results suggest that CEO bonus compensation is more sensitive to industry-adjusted performance when the firm outperforms its industry benchmark (asymmetric RPE) and when the industry benchmark is positive (asymmetric pay-for-luck), consistent with CEO manipulation of the pay process to extract rents from the firm. We also find that there is a limit to the magnitude of this rent extraction, however, since CEOs are not paid for good luck when they are already being paid for beating the industry benchmark (that is, there is no evidence of double-dipping). In sum, our findings are consistent with a weak implementation of RPE (which may be efficient contracting), where the manager is shielded from extremely bad outcomes but at the same time not excessively rewarded when the combination of firm and industry performance (i.e., talent and luck) is extremely good