Performance peer groups in CEO compensation contracts

• Author: Tor‐Erik Bakke,Hamed Mahmudi,Ashley Newton
• Published date: 01 December 2020
• DOI: http://doi.org/10.1111/fima.12296
• Date: 01 December 2020

DOI: 10.1111/fima.12296

ORIGINAL ARTICLE

Performance peer groups in CEO compensation

contracts

Tor-ErikBakke1Hamed Mahmudi2Ashley Newton3

1UIC College of Business Administration,

University of Illinois at Chicago, Chicago, Illinois

2Alfred LernerCollege of Business and

Economics, University of Delaware,Newark,

Delaware

3Michael F.Price College of Business, University

of Oklahoma, Norman, Oklahoma

Correspondence

HamedMahmudi, Alfred Lerner College of Busi-

nessand Economics, University of Delaware, 20

OrchardRd, Newark, DE 19716.

Email:hmahmudi@udel.edu

Abstract

How do firms choose performance peer groups used in chief exec-

utive officer (CEO) relative performance evaluation contracts? We

find that while firms, for the most part, choose performance peers

to better identify their CEOs’ impact on firm performance, they also

tend to select underperforming peers. Dynamically, we find that

peers that are added and retained every year are weaker than ones

that were not chosen. These findings suggest managers may have

some influence on the choice of performance peers. Finally, using a

quasi-natural experiment, we find that enhanced disclosure did not

affect the tendency of firms to select underperforming peers.

KEYWORDS

executive compensation, performance peer group, relative perfor-

mance evaluation

1INTRODUCTION

The use of relative performance evaluation (RPE) in chief executiveofficer (CEO) pay packages is on the rise. Its popu-

larity among S&P 500 firms has roughly doubled from 17% in 2006 to 34% in 2012. Furthermore, among RPE granting

firms, the RPE component of executivecompensation is relatively large representing approximately 32% of total exec-

utive compensation in 2012 (Bettis, Bizjak, Coles, & Young,2014). The rationale for RPE is that a more precise measure

of CEO performance can be obtained bymaking the CEO’s compensation contingent on how the firm performs relative

to a group of peer companies (i.e., the performance peer group). However,the effectiveness of RPE contracts hinges on

selecting an appropriate performance peer group to filter out exogenousshocks to firm performance that are outside

the CEO’s control.

In this paper, we exploit the 2006 Securities ExchangeCommission (SEC) mandated disclosure of the contractual

termsof RPE contracts to provide the first comprehensive panel data analysis regarding how firms choose performance

peer groups used in CEO RPE contracts. We obtain data on the contractualdetails of CEOs’ RPE contracts from proxy

statements (DEF 14A) for the 750 largest publicly listed companies in the United States from 1998 to 2012.1This rich

c

2019 Financial Management Association International

1In our main analysis, we focus on the postregulation disclosure regime (2006–2012) to avoid the selection bias associated with voluntary disclosure in the

pre-2006subsample.

Financial Management. 2020;49:997–1027. wileyonlinelibrary.com/journal/fima 997

998 BAKKE ET AL.

panel data allow us to better explain,and rule out alternative hypotheses for, how firms choose their performance peer

groups and to investigate the causal impact of mandatory disclosure on the peer selection process.

Optimal contracting theory suggests that compensation contracts should rely on performance measures that are

more accurate signals of CEO actions (Holmstrom, 1979, 1982). RPE allows firms to increase the accuracy of perfor-

mance measures by taking away any component of the performance that is not due to CEO actions. Thus, properly

designed RPE contracts should include performance peers that best identify the CEO’s impact on firm performance.

To that end, RPE grantingfirms should choose performance peers that face similar risk exposures to their own. This

implies selecting peers whose performance is likely to be highly correlated with the RPE grantingfirm’s performance.2

However,managerial power theory suggests that executives,by controlling their own boards, maximize their own com-

pensation by choosing performance peers strategically at the cost of shareholders.

Totest these competing theories in the context of performance peer selection, we first investigate the variation in

the choice of index versus custom peer group in RPE contracts.In our sample, we find that 40% of firms compare their

performance to stock indices, whereas 65.3% compare to a custom peer group. Consistent with optimal contracting

theory, firms that are more complex(i.e., large and multi-segment firms) tend to avoid using a custom peer group and

instead compare their performance to an index. We also find that firms with fewer independent directors and firms

with more busy directors, who may not be able to construct well-designed custom peer groups, tend to rely on indices

as performance peers in their RPE contracts.

Next, we study the composition of custom peer groups. Wefind that RPE granting firms tend to select performance

peers that operate in similar industries and have historical stock returns that are highly correlated with the returns

of the RPE granting firm. Further,performance peers are similar in size, operate in similar industries, and have similar

operational structure and geographical presence relative to the RPE granting firm. Overall, consistent with optimal

contractingtheory, these findings suggest that firms tend to choose performance peers that better filter out exogenous

common factors in an effort to identify the true impact of CEO actions on performance.

Do factors supported by managerial power theory also contribute to predicting the choice of performance peer

firms? We find that the historical stock returns of a potential peer company are negatively associated with the likeli-

hood of a firm being chosen as a performance peer after controlling for other determinants of peer selection, such as

industry,historical correlations, and size. In other words, compensation committees seem to be endorsing performance

peer groups that include companies that underperformed historically,possibly because these peer companies may be

easier to outperform in the future. Interestingly,we find similar results when replacing historical returns with contem-

poraneous returns. This indicates that, on average,companies tend to select peer firms that underperform relative to

other potential peer firms. These results are consistent with the managerial power theory of executivecompensation

in which CEOs have incentives to influence the selection of performance peers in order to select peers that are eas-

ier to beat (Dye, 1992; Gibbons & Murphy,1990; Jensen & Murphy, 2011; Murphy, 2001). Our results haveimportant

implications for shareholders as underperforming peer groups could lead to lower exante incentives and higher ex post

compensation that cannot be justified by CEOs’ actions.

The negative peer choice to performance sensitivity remains unchanged with the inclusion of firm fixedeffects sug-

gesting that this result is driven by within firm variation in the choice of performance peers. To further investigate

what drivesthis finding, we study the dynamics of performance peer selection. Peers change frequently: 31.9% of firms

change their peers annually and, among these, 31.7% of peers are changed. We find that RPE granting firms actively

added and retained peers that were weak compared to peers that could have been added. Thus, our findings are not

merely due to selected performance peers persistently underperforming potential peers. However,we do not find evi-

dence that peers that were dropped outperform retained peers (e.g., drops are less frequent than adds). Nonetheless,

the strategic retention and addition of underperforming peers, combined with an overall increase in the number of

2Thisprocess of selecting performance peers is distinct from the selection of compensation peer groups (Albuquerque, De Franco, & Verdi, 2013; Faulkender

& Yang,2010, 2013). Although compensation peers are chosen to attract and retain talented CEOs, performance peers are expected to filter out the effect

of common shocks on performance measures to ensure that CEOs are not paid for luck. Consistent with this analogy,in our sample, only 17.8% of firms have

identicalcustom performance and compensation peer groups.