Management Accounting Section Midyear Meeting

Why Do Companies Use Different Peers for Compensation Benchmarking than for Relative Performance Evaluation?

• In Event: Relative Performance Evaluation

Abstract

The literature on relative performance evaluation (RPE) has largely ignored that there are two related, but different sets of peers used in two connected stages of the executive pay setting process: pay benchmarking and RPE. The peers used for benchmarking may differ from those used for RPE. In this paper, we explore why companies use different sets of peers for compensation benchmarking and RPE. We document that out of firms that use peers for both compensation benchmarking and RPE, only 7% of them use the same set of peers, while the remaining of them have at least one different firm in the two peer sets. We find that for RPE purposes, companies tend to drop and/or add peers if the degree of commonality between the benchmarking peers and subject firms are low. We also find that if companies change peer composition they increase the degree of commonality between the subject firm and RPE peers. Finally, abnormal CEO pay is lower if there is higher commonality between the subject firm and RPE peers. In cross-sectional analyses, we find that CEO tenure moderates the efficient choice of peers for RPE while more compensation consultants enhance the efficiency of peer choice for RPE. Overall, our results are consistent with efficient contracting theory in the peer selection for CEO performance evaluation.

Authors

• James Gong, Cal State University - Fullerton

• Anthony Chen, California State University, Fullerton

• Chuchu Liang, University of California-Irvine