The Role of Luck in CEO Performances

Abstract

Firm performance is a crucial factor in how CEOs are evaluated. However, a CEO can be repeatedly lucky or unlucky, adding noise to performance outcomes as a measure of managerial ability. In this study, I examine how much of the observed cross-sectional dispersion in outcomes can be attributed to differences in luck as opposed to differences in skill. Using bootstrap simulations, I show that, even if all CEOs were equally skilled, we can expect substantial differences in performance outcomes. When comparing the simulated distribution of outcomes to the actual empirical distribution, I find that the best performing CEOs perform too well relative to the median to be completely explained by luck alone. However, the true underlying differences in skill are substantially smaller than suggested by simply looking at the raw performance differential.