Does executive compensation reflect corporate productivity?

• DOI: http://doi.org/10.1111/jbfa.12437
• Date: 01 July 2020
• Published date: 01 July 2020
• Author: Yoon K. Choi

DOI: 10.1111/jbfa.12437

Does executive compensation reflect corporate

productivity?

Yoon K. Choi

Department of Finance, College of Business

Administration,University of Central Florida,

Orlando, USA

Correspondence

YoonK. Choi, Department of Finance, College of

BusinessAdministration, University of Central

Florida,PO Box 161400, Orlando, FL 32816, USA.

Email:ychoi@ucf.edu.

Abstract

Recent literature has given attention to the effect of CEO-specific

productivity on the structure of CEO compensation. Our paper

instead focuses on the effect of a different productivity factor—

which we call “corporate productivity”—on CEO compensation. In

particular,we show that corporate productivity affects the trade-off

between incentive and risk in a non-monotonic fashion, which the

literature has not yetrecognized. Using various empirical proxies for

corporate productivity,we show that our results are consistent with

the non-monotonic relation and thus contribute to the debates in

the incentive-risk trade-off literature. Second, our findings also con-

tribute to the internal capital marketliterature by exploring the rela-

tion between the structure of CEO compensation and excessvalue.

KEYWORDS

CEO compensation, corporate productivity, excessvalue, incentive-

risk trade-offs

JEL CLASSIFICATION

G30, G32, J33

1INTRODUCTION

A vast principal-agent literature examines the relation between managerial compensation and major parameters,

including agents’ risk aversion, firm risk (volatility), and firm size. Only a few papers, however,have paid due attention

to the importance of the production function parameter (i.e., the marginal product of effort) in influencing managerial

compensation, either theoretically or empirically. For example,Baker and Hall (2004) attempt to explain the impact

of firm size used as a proxy for CEO-specific productivity on the level of CEO compensation. Meanwhile, Gabaix and

Landier (2008) and Edmans, Gabaix, and Landier (2009) examine the relation between CEO pay sensitivity and firm

size, in terms of the functional form of the production function.

Given that the marginal product of effort (MPE) can be described as the marginal contribution of managerial effort

to firm value, we envision two factors that determine MPE. The first factor is related to CEO-specific characteristics

such as the CEO’s skill or quality, while the other factor is associated with a productiveenvironment such as the effi-

ciency of the internal capital marketor the c omplexityof organizations. Baker and Hall (2004) examine the first factor,

1012 c

2020 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2020;47:1012–1033.

CHOI 1013

whereas we focus on the second.1In other words, in our definition, two identical CEOs, with respect to their quality,

could face two different production environments that affect the MPE in their own firms. Tothe best of our knowledge,

the compensation literature has not fully examined the impact of this second factor based on the production environ-

ment, which we call a firm’s corporate productivity for convenience(e.g., the efficiency of internal capital markets or the

complexity of the organization) on the structure of CEO compensation.2,3

In this paper, we show that corporate productivity affects the trade-off between incentive and risk in a non-

monotonic fashion, which the literature has not yet recognized. Specifically,the incentive–risk trade-off is exacerbated

(i.e., becomes more negative) as productivity increases in a risk-dominant environment, where risk (uncertainty) is

greater than corporate productivity.At the same time, the trade-off is attenuated (i.e., becomes less negative) as pro-

ductivity increases in a productivity-dominant environment, where corporate productivity is greater than risk (uncer-

tainty). The intuition is as follows. Corporate productivity has two effects on incentives: On the one hand, firm perfor-

mance is expected to be higher with higher corporate productivity,and thus more equity incentives will be provided

accordingly. On the other hand, the higher equity incentives will also increase the risk-premium demanded byCEOs.

When a manager faces a sufficiently great level of risk (uncertainty) relative to productivity, the marginal impact of

productivity on the risk premium would be greater than its impact on incentives. Therefore, the incentive–risk relation

is exacerbated with a high levelof productivity in a risk-dominant environment.

Our major contribution is twofold. First, our results on the non-monotonic incentive–riskrelation thus contribute to

the debates in the incentive–risk trade-offliterature (e.g., Prendergast, 2002). Specifically, equity-based compensation

in our sample is found to be much more sensitive to risk in firms with higher productivity. We obtain these results

even after controlling for the effect of firm size on pay-performance sensitivity (PPS), and thus the relation between

corporate productivity and the incentive–risk trade-offseems to be strong, independent of firm size.

Second, our work also contributes to the internal capital market literature by exploring the relation between the

structure of CEO compensation and excessvalue. We employ three key proxy variables for the corporate productivity

parameter,based on the internal capital market literature. First, we examinetwo different forms of corporate structure

in terms of diversification: single-segment versus multi-segment firms. The literature recognizes that on average,the

excessvalue of multi-segment firms is lower than that of single-segment firms. To the extent that single-segment firms

systematicallydiffer from multi-segment ones in corporate productivity, we should be able to observe different optimal

compensation structures between the two types of firms. Second, we use excess value itself as a proxyfor corporate

productivity.4We interpret excessvalue as a reasonable manifestation of corporate productivity in the sense that it

measures the efficiency of production environments in which more efficient investment leads to greater firm value.

Finally, the third proxy for corporate productivity is the future growth opportunity measured by Tobin’s Q. In fact,

excessvalue can be viewed as an industry-adjusted Tobin’s Q, which reflects the sensitivity of firm value to CEO action

(Sloan, 1993).

This paper is organized as follows. Section 2 briefly describes an extensionof a standard principal–agent model with

corporate productivity,leaving the details to Appendix A. We also discuss empirical hypotheses based on the implica-

tions from the model. Section 3 describes the data, the methodology,and the summary statistics. Section 4 elaborates

1Infact, Baker and Hall (2004) themselves recognize the concern that differences in firms, apart from firm size, may explain the marginal product of CEOeffort.

Figures1 and 2 in their paper show a wide variation in the estimates of the marginal product given firm size. Our corporate productivity factor, measured after

controlling for firm size, may capture the different sources of variation in the MPE, including the level of diversification, different industries, and varying

organizationalstructures, as suggested by Baker and Hall (2004).

2The existing literature has focused on the CEO-specific MPE (e.g.,CEO skill or quality measured by firm size) in explaining the structure of compensation

(Baker & Hall, 2004; Jensen & Murphy,1990a). The labor market may recognize CEO quality and determine the level of CEO compensation accordingly.

Therefore,we may naturally observe a positive relation between firm size and total compensation. However, our paper is interested in exploringthe structure

ofcompensation as a function of corporate productivity that is influenced by internal operational efficiency, investment opportunity,or both.

3The efficiency of internal capital markets has been extensivelystudied (for the related literature, see Scharfstein & Stein, 2000; Stein, 1997). Meanwhile,

Coles,Daniel, and Naveen (2008) examine organizational complexity in explaining the relation between board size and firm performance. Conceptually,these

variablesare good proxies for corporate productivity. We appreciate comments from an anonymous referee in pointing out this insight.

4Excessvalue is the difference between the actual value of the firm and its imputed value (see Berger & Ofek, 1995). Excess value becomes positive (negative)

whenthe actual value is greater (less) than the imputed value. See Appendix B for a rationale for using excess value as a proxy for corporate productivity.