We empirically examine bank CEOs’ compensation. We find that bank CEOs (a) are paid less than their nonfinancial counterparts, an effect driven by the CEOs of small bank; (b) experienced declining compensation during 2007–2009 (the hardest-hit banks cut compensation more) but pay is now 24% higher than precrisis levels; (c) are paid more at larger banks, those with less nonperforming loans, those with a higher proportion of noninterest income, and those with less demand-deposit dependence; and (d) have pay highly sensitive to ROA and ROE, but not stock returns. Tail risk is higher when compensation depends more on short-term measures of performance. (JEL, F34, G32, G33, G38, K42)

Using a large data set of performance goals employed in executive incentive contracts, we find that a disproportionately large number of firms exceed their goals by a small margin as compared to the number that fall short of the goal by a similar margin. This asymmetry is particularly acute for earnings goals, when compensation is contingent on a single goal, when the pay-performance relationship around the goal is concave-shaped, and for grants with non-equity-based payouts. Firms that exceed their compensation target by a small margin are more likely to beat the target the next period and CEOs of firms that miss their targets are more likely to experience a forced turnover. Firms that just exceed their Earnings Per Share (EPS) goals have higher abnormal accruals and lower Research and Development (R&D) expenditures, and firms that just exceed their profit goals have lower Selling, General and Administrative (SG&A) expenditures. Overall, our results highlight some of the costs of linking managerial compensation to specific compensation targets.