The generalist CEO pay premium and CEO risk aversion

• Published date: 01 October 2023
• Author: Leila Zbib,Kwadwo Asare
• Date: 01 October 2023
• DOI: http://doi.org/10.1002/jcaf.22638

Received: 21 November 2022 Revised: 2 April 2023 Accepted: 14 April 2023

DOI: 10.1002/jcaf.22638

RESEARCH ARTICLE

The generalist CEO pay premium and CEO risk aversion

Leila Zbib1Kwadwo Asare2

1Department of Finance, Bryant

University, Smithfield, Rhode Island,

USA

2Department of Accounting, Bryant

University, Smithfield, Rhode Island, USA

Correspondence

Leila Zbib, Department of Finance, Bryant

University, Smithfield, Rhode Island,

USA.

Email: lzbib@bryant.edu

Abstract

We investigate whether and how managerial risk aversion influences the struc-

turing of the generalist pay premium. We use three candidate variables for

managerial risk aversion: the CEO’s equity portfolio vega, decision horizon, and

wealth. Wefind that equity-based pay differs between generalist CEOs (who have

broader experience spanning several industries) and non-generalist CEOs (who

tend to have narrower, specialized industry-specific experience) based on their

risk appetite. Further, how the generalist paypremium is structured depends on

the risk aversion measure. Because generalist CEOs tend to be better paid than

their counterparts, understanding their risk aversion level will help companies

improve the structure of these CEOs’ compensation packages, and ultimately

those of all CEOs. When vega is the risk aversion measure, CEOs get compen-

sated for bearing risk. However, generalist CEOs with high vegas suffer a “risk

discount” in their compensation. These generalist CEOs are associated with

lower cash, equity, and total pay. CEOs with longer decision horizons get more

equity and less cash, but this structuring does not extend to the generalist CEOs.

Finally, wealthier CEOs get more cash and total pay. However, generalist CEOs

who are wealthier get less cash pay.

KEYWORDS

CEO decision horizon, CEO general ability, CEO pay premium, CEO vega, CEO wealth,

managerial risk aversion

JEL CLASSIFICATION

G35, J24, J33

1 INTRODUCTION

We evaluate the effect of differences in managerial risk

aversion on how the generalist pay premium is struc-

tured. Custódio et al. (2013) find that the generalist CEOs

command a premium in compensation due to several fac-

tors, including generalizability of managerial skills and a

broader variety of professional experiences.

Weconjecture that the effect of risk aversion would man-

ifest in differences in CEO compensation structures across

different settings of measures that are sensitive to man-

agerial risk aversion. Consistent with results of analytical

and empirical models of agency theory, we expect that

firms will reduce the riskier equity portion of pay the more

risk aversion the CEO demonstrates. Similarly, we expect

less risk averse managers to negotiate for greater portions

of equity in their compensation packages. See for exam-

ple, Hall and Murphy 2003; Laffont and Martimort, 2009;

Bolton and Dewatripont 2005.

Custódio et al. (2013) find that the generalist pay pre-

mium is robust to CEO risk aversion. They find that

the generalist CEO pay premium exists for the major

J Corp Account Finance. 2023;34:89–107. © 2023 Wiley Periodicals LLC. 89wileyonlinelibrary.com/journal/jcaf

90 ZBIB and ASARE

FIGURE 1 This figure reports the

coefficients of the components of the generalist

CEO pay premium and equity as a percentage

of total pay by Vegaquintiles from Panel B of

Tabl e 4.The right y-axis is associated with the

bar graphs.

components of pay and for total pay.They use equity-based

pay as a component of total pay as a control variable for

CEO risk aversion and find that their results hold.

In this paper, we add to Custódio et al. (2013) and the

extant literature on CEO compensation by testing how

variations in risk aversion affect the structuring of the gen-

eralist CEO pay premium relative to non-generalists. For

example, it would be informative to know if the effect

of risk aversion on the structuring CEO pay generally,

and the generalist premium particularly, is monotonic, so

that the proportion of equity pay always decreases in risk

aversion.

A related issue in the empirical CEO compensation liter-

ature is the absence of generally accepted measures of CEO

risk aversion. Research and development as well as the

extent of diversification have been used as measures of risk

aversion (e.g., Liu et al., 2021). However, these measures

are structural and tend to persist beyond CEOs.This makes

them unsuitable as measures of individual CEO risk aver-

sion in some settings. While one CEO may alter a firm’s

R&D expenditures or change the extent of diversification

to match their level of risk aversion, others may demon-

strate their risk aversion in more personal ways such as

how much deferred compensation versus cash they nego-

tiate for in their compensation package. We triangulate

on the CEO risk aversion construct with three measures:

volatility in the CEO’s equity options or vega, the CEO’s

decision horizon, and finally the CEO’s wealth.

This paper makes several contributions to the CEO

compensation literature. Since generalist CEOs constitute

some of the highest paid CEOs, understanding how the

generalist premium is structured will enhance researchers’

and policymakers’ understanding of how a significant por-

tion of highly paid CEOs’ pay packages are structured. In

turn, this understanding will help researchers and policy-

makers disentangle the extent to which CEO pay is driven

by efficient contracting (e.g., Baker & Hall, 2004; Gabaix &

Landier, 2008) versus rent extractionand poor governance

(Bebchuk et al., 2002; Bebchuk & Fried, 2003).

Given the important role that equity plays in addressing

the risk aversion problem in designing managerial com-

pensation (e.g., Conyon et al., 2011; Edmans & Gabaix,

2011), differences in risk aversion wouldinfluence how the

pay premium accruing to generalist managers is structured

(e.g., Laffont & Martimort, 2009).

Furthermore, we contribute to the literature on CEO

compensation by providing further empirical evidence

that CEOs’ portfolio vegas and decision horizons can

serve as proxies for their risk tolerances/risk aversion1

(see for example, Coles et al. 2006). Investors and regula-

tors can use these measures as inputs to inferences about

the efficacy of CEO compensation packages. Our results

demonstrate that given the difficulty of measuring risk

aversion, it is best that researchers triangulate to capture

the construct with several measures. Our results suggest

that two of our risk aversion candidates, vega (change in

share price per unit change in volatility) and wealth, do

capture some aspects of risk aversion, while a CEO’s deci-

sion horizon, also captures different dimensions of risk

aversion.

We add to the literature on CEO characteristics, more

precisely on those of generalist CEOs and the frameworkof

their compensation packages. Our results show that gen-

eralist CEOs get less value from their generalist skills the

greater the vega in their compensation. This in turn can

drive a generalist CEO to negotiate for a much greater

proportion of equity pay in their compensation (Hall &

Murphy, 2002, 2003). The empirical evidence depicted in

Figure 1is consistent with this theory.

Because vega evaluates how the share price changes

with a unit change in volatility, it necessarily incorpo-

rates a temporal perspective into an evaluation of a CEO’s

compensation package. Along with the CEO’s decision

horizon, these two measures of risk aversion introduce an