CEO Personal Wealth, Equity Incentives and Firm Performance

• Date: 01 January 2013
• Author: Juha‐Pekka Kallunki,Anna Elsilä,Henrik Nilsson,Petri Sahlström
• DOI: http://doi.org/10.1111/corg.12001
• Published date: 01 January 2013

CEO Personal Wealth, Equity Incentives and

Firm Performance

Anna Elsilä, Juha-Pekka Kallunki*, Henrik Nilsson, and

Petri Sahlström

ABSTRACT

Manuscript Type: Empirical

Research Question/Issue: In this paper, we explore the determinants and performance effects of a novel measure of

executive incentives, that is, the elasticity of a CEO’s total wealth to f‌irm performance, computed as a CEO’s ownership

value relative to her total wealth.

Research Findings/Insights: Using unique data on the total personal wealth of the CEOs of listed Swedish f‌irms, we f‌ind

that while CEOs typically own only a very small fraction of the shares of their f‌irms, this ownership often constitutes a

signif‌icant part of their total wealth. We also f‌ind that a CEO’s total wealth elasticity is negatively associated with f‌irm size

and a CEO’s age is positively associated with a f‌irm risk environment. As for the effect on f‌irm performance, we f‌ind that

higher CEO’s equity incentives enhance future accounting f‌irm performance, when we estimate panel data regressions by

taking into account the dynamicnature of endogeneity between a CEO’s incentives and f‌irm performance. However, we do

not f‌ind any evidence that a CEO’s equity incentives to be related to future stock returns.

Theoretical/Academic Implications: Our analysis suggests that executives’ incentives when measured as price-

performance elasticity are economically more signif‌icant than has been often argued in the literature. Moreover, due to

several advantages of price-performance elasticity over traditionally used fractional ownership and dollar-at-stake mea-

sures, the inferences drawn from this study with respect to determinants and performance implications of a CEO’s

incentives may be superior to those from prior studies. The result on the relation between f‌irm size and a CEO’s incentives

is especially interesting, because, unlike fractional ownership, the elasticity of a CEO’s total wealth to f‌irm performance is

not spuriously related to f‌irm size. With respect to the relation between incentives and subsequent f‌irm performance, our

results are consistent with the view that higher CEO’s incentives, measured as price-performance elasticity, enhance f‌irm

accounting performance. Overall, future studies should consider price-performance elasticity as an alternative measure of

executive incentives in addressing various research questions.

Practitioner/Policy Implications: The evidence in this paper suggests that executive incentives calculated relative to their

non-f‌irm wealth are economically more meaningful if compared to the traditionally used measures.As such, compensation

consultants and boards of directors can consider this alternativemeasure when designing executive compensation plans and

determining an appropriate level of ownership, when the information on executive non-f‌irm wealth is available.

Keywords: Corporate Governance, Incentives, Firm Performance, Sweden, Price Performance Sensitivity

INTRODUCTION

In the executive compensation literature, much research

effort has been devoted to how to measure managerial

incentives and whether these incentives improve f‌irm per-

formance (e.g., Benson & Davidson, 2009; Demsetz & Lehn,

1985; Hall & Liebman, 1998; Himmelberg, Hubbard, & Palia,

1999; Jensen & Murphy, 1990; McConnell & Servaes, 1990;

Morck, Shleifer, & Vishny, 1988). However, the empirical

results from this research are still largely inconclusive. Many

studies argue that these mixed results are because of the

limitations of the two frequently used proxies for executive

incentives, that is, a dollar-to-dollar (price-performance

sensitivity, hereafter PPS) or dollar-to-percentage (price-

performance semi-elasticity, hereafter PPSE) change in the

wealth of executives due to changes in the market value of a

*Address for correspondence: Juha-Pekka Kallunki, Department of Accounting,

University of Oulu, P.O. Box 4600, FIN-90014 Oulu, Finland; Tel: +358 40 567 2508;

E-mail: Juha-Pekka.Kallunki@oulu.f‌i

26

Corporate Governance: An International Review, 2013, 21(1): 26–41

© 2012 Blackwell Publishing Ltd

doi:10.1111/corg.12001

f‌irm (e.g., Baker & Hall, 2004; Hall & Liebman, 1998). In this

paper, we extend the literature on executive compensation

by empirically analyzing a novel and theoreticallyappealing

measure of executive incentives, that is, the elasticity of

a CEO’s total personal wealth to f‌irm performance, often

referred to as total price-performance elasticity (PPE). Spe-

cif‌ically, we utilize unique data on Swedish CEOs’ total per-

sonal wealth to explore the determinants and performance

implications of the PPE incentive measure. To the best of our

knowledge, there are no prior studies to empirically explore

PPE due to the data limitations.

Our empirical analyses using the data on CEOs’ outside

wealth obtained from the Swedish tax authorities can be

summarized as follows. First, we directly compare the mag-

nitude of PPE and PPS and f‌ind strong evidence that a

CEO’s equity incentives are much greater than suggested in

the literature (Jensen & Murphy, 1990). In particular, while

CEOs’ mean (median) fractional ownership of their f‌irms is

3 percent (0 percent), the mean (median) ratio of this own-

ership to their total wealth is 45 percent(38 percent). In other

words, the PPE measure shows that CEOs typically invest

economically a much greater proportionof their total wealth

in the f‌irm they are managing than one would assume based

on the traditional PPS measure.

Second, by employing a standard methodology in the

executive compensation literature, we explore the determi-

nants of PPE by regressing PPE against the economic deter-

minants of incentives such as f‌irm size, volatility, CEO age,

growth opportunities, and CEO changes, and compare the

results to those with PPS and PPSE as dependent variables

(e.g., Core & Guay, 1999; Demsetz & Lehn, 1985; Himmel-

berg et al., 1999). Our main interest is in exploring how f‌irm

size affects PPE compared to PPS and PPSE, because prior

studies criticize PPS for being sensitive to f‌irm size (Hall &

Liebman, 1998), while PPSE represents unscaled incentives.

Our results indicate that the CEOs of large f‌irms are less

incentivized in terms of their PPE. This evidence is consis-

tent with the view that larger f‌irms exhibit economies of

scale in monitoring costs, as they are also monitored by

credit rating agencies and media (Himmelberg et al., 1999).

We also f‌ind that, similar to other measures of incentives,

PPE is higher for more volatile f‌irms, which is consistent

with the prediction that CEOs of f‌irms operating in

noisier business environments require stronger monitoring

(Demsetz & Lehn, 1985). In addition, we document that PPE

becomes lower as a CEO gets older and for newly appointed

CEOs.

Third, we explore the performance implications of PPE.

We estimate performance regressions by using PPE as an

incentive measure, and compare the resultswith those based

on PPS and PPSE measures. Our results show that after

controlling for the dynamic nature of endogeneity, the rela-

tion between a CEO’s incentives and future accounting per-

formance is signif‌icantly positive.

The remainder of the paper is structured as follows. The

next section discusses the related literature and presents the

hypotheses. Following this, we outline our data and specify

empirical models followed by empirical results together

with the robustness tests. The f‌inal section concludes

the paper and summarizes its theoretical and practical

implications.

RELATED LITERATURE

Different Approaches to Measuring

Executive Incentives

According to the agency theory, f‌irms incur agency costs

arising from less-than-optimal alignment of the agent’s and

principal’s interests (Jensen & Meckling, 1976). This agency

conf‌lict can be mitigated by tying executives’ wealth to f‌irm

performance through grants of shares and stock options,

and by establishing minimum equity ownership require-

ments for management.

Several authors, exploring the incentives of top execu-

tives, empirically conclude that these incentives are too

small to motivate managers to create shareholder value (e.g.,

Bebchuk & Fried, 2004; Jensen & Murphy, 1990). In their

seminal paper, Jensen and Murphy (1990) report that the

mean change in wealth consisting of cash pay, options, stock,

and dismissal probability of US executives is only $3.25 per

thousand dollar change in shareholders’ wealth. Jensen and

Murphy (1990) measure executives’ incentives using PPS,

that is, the fraction of the f‌irm owned by the executives. This

approach has several shortcomings. In particular, Hall and

Liebman (1998) point out that, since executives are risk-

averse and f‌inancially constrained from buying a signif‌icant

portion of their f‌irms, especially in large f‌irms, the relation

between PPS and f‌irm size is spurious.1As an alternative to

PPS, they propose a dollar change in executive holdings per

percentage change in the market value of equity, or PPSE, as

a measure of incentives. Baker and Hall (2004) theoretically

analyze and compare the implications of PPS and PPSE

for f‌irms of different sizes. They conclude that PPSE is an

adequate measure of incentives if managerial decisions

affect percentage returns on f‌irm value, and that PPS can be

used in situations where the managerial effort is constant

across f‌irm sizes. In light of their results, one can conjecture

that PPSE is a more relevant incentive measure to induce

value creation, for example, through strategic projects, since

it is related to changes in f‌irm value multiplicatively.

A limitation of the PPSE measure is that it only accounts

for the absolute change in an executive’s wealth, therefore

assuming explicitly that the dollar at stake represents her

total wealth. This is a signif‌icant caveat, since it is obvious

that, e.g., a payoff of $1 million would be evaluated differ-

ently by individuals with, e.g., $100,000 of outside wealth

versus those with $100 million (Core & Guay, 2010; Core,

Guay, & Larcker, 2003). This problem can be resolved by

measuring incentives as a percentage change in executive

total wealth relative to percentage change in the market

value of the f‌irm, that is, by using the PPE measure. If a

CEO’s wealth consists solely of equity holdings in her f‌irm,

the PPE equals one, meaning that the CEO’s total wealth

changes by one percent in response to a one percent change

in the stock price.

Recently, several papers have recognized that PPE is the

desired measure of executive incentives, because it is not

affected by f‌irm size and accounts for executives’ key per-

sonal characteristics, that is, their outside wealth (Core &

Guay, 2010; Edmans, Gabaix, & Landier, 2009).An empirical

challenge in applying PPE is that in many countries, includ-

ing the United States, information about executives’ total

CEO PERSONAL WEALTH AND FIRM PERFORMANCE 27

Volume 21 Number 1 January 2013© 2012 Blackwell Publishing Ltd