The Other Side of the Trade-Off: The Impact of Risk on Executive Compensation: A Revised Comment

Abstract:     
In contrast to a body of research starting with Demsetz and Lehn (1985) that predict and find a strong positive association between firm percent return variance and incentives, Aggarwal and Samwick (1999) predict and find a strong negative association between firm dollar return variance and incentives. A key assumption of Aggarwal and Samwick's analysis is that firm risk is the sole determinant of the pay-performance sensitivity, and that expected dollar return variance (the product of expected percent return variance and firm market value) is the correct proxy for risk. We demonstrate that dollar return variance is a noisy measure of firm market value and argue that A&S re-documents a size effect that is already well-known from prior literature. Because dollar return variance is shown to be a noisy proxy for firm size, the A&S empirical specification does not include an appropriate proxy for firm risk. The data consistently show that it is important to examine market value and percent return variance as separate determinants of the effects of size and risk on CEO incentives, as is done in the managerial ownership literature. In a model of CEO incentives that includes market value and risk as separate explanatory variables, we find that, contrary to the results in A&S, percent return variance is positively associated with incentives. The A&S empirical work cannot be interpreted as evidence of a negative relation between risk and incentives.

Number of Pages in PDF File: 34

Keywords: incentives, risk, uncertainty, standard agency model, managerial ownership

JEL Classification: J30, J33, G32, L14, L22

working papers series