CEO Overconfidence and Relative Performance Evaluation.

Building upon a broad stream of research in psychology, incentive theory and strategy, this paper offers one of the first empirical tests of how performance evaluation mechanisms affect risk-taking by biased CEOs and how rational principals design optimal performance evaluation mechanisms to account for this effect. We test if the risk-taking behavior of overconfident chief executive officers (``CEO""s) change when they are forced to focus on their relative standing, rather than solely absolute skills. Using an options-based measure of CEO overconfidence in the period of 1998-2014, we show that firms with overconfident CEOs have higher return volatility and that the use of explicit relative performance evaluation (``RPE"") contracts significantly increases return volatility associated with the overconfident CEO. Finally, anticipating this, rational principals are less likely to offer RPE contracts to overconfident CEOs. By showing a previously unknown adverse effect of RPE contracts, we establish the first link between CEO attitudes and the design of performance evaluation mechanisms. [ABSTRACT FROM AUTHOR]