CEO Compensation: Does Performance Matter?

Abstract:     
This paper analyzes CEO compensation in years around and including exceptionally good and poor performance. Using compensation data from 1993 through 2003, the results suggest CEOs are able to increase their compensation before exceptionally bad performance through the timing of option exercises. We also find as the percentage of compensation from exercised options increases, the probability of exceptionally poor performance the following year almost doubles. These results complement those of Yermack (1997), Chauvin and Shenoy (2001), and Lie (2005). These papers provide evidence that CEOs manipulate either timing or reporting of option grants to increase gains from options. By disentangling the components of compensation and conditioning on performance, we also test Bolton, Scheinkman, and Xiong’s (2005) hypothesis that excess compensation is the result of short term opportunism in speculative markets. Contrary to Bolton, et al., we find CEOs had lower gains from exercised options in the bubble (speculative market) years than in non-bubble years. We also examine Bebchuk and Fried’s (2003) CEO power hypothesis. CEOs of poorly performing firms are able to gain such large amounts from option exercise prior to poor performance that even after two years; their cumulative pay is as high as those who performed in the highest stock return decile. Finally, except for the option component of compensation, our results from tests of other compensation components are consistent with optimal contracting theory.

Number of Pages in PDF File: 41

Keywords: executive compensation, option exercises

JEL Classification: J33, M52, G3

working papers series