Compensation Rigging by Powerful CEOs: A Reply and Cross-Sectional Evidence

Wan (2013) argues that the statistical inferences in our Journal of Finance (2011) paper are not robust, as we do not prove that it is powerful CEOs that rig incentive contracts. Wan makes the theoretical claim that the rigging results are consistent with ex-post optimal re-contracting. However, optimal re-contracting cannot explain the loss in firm value from contract switching we show in the paper. Nor do we know of a theory that would predict that ex-post realignment could be tested using our contract switching term in the wage function, like Wan does. On the empirical front, Wan's critique has at least three flaws. First, his standardized performance measures — different than ours — result in accounting returns being 14 percentage points higher than stock returns. Consequently, switching between measures, necessary for identification, is infrequent and outlier-based, not surprisingly delivering regression estimates different from ours. Second, he interprets selectively among insignificant coefficients to make his claims. Third, regardless of interpretation, basic mathematics casts doubt on the premise of his estimation strategy. Wan makes one valid point: our original work could have provided more extensive cross-sectional empirical support for our rigging claims. We take this opportunity to present new cross-sectional (between-firm) evidence and conclude even more strongly that powerful CEOs sway boards to load their incentive pay on more favorably performing measures.