The Effect of CEO Option Compensation on the Capital Structure: A Natural Experiment

• Author: Onur Kemal Tosun
• Published date: 01 December 2016
• Date: 01 December 2016
• DOI: http://doi.org/10.1111/fima.12116

The Effect of CEO Option Compensation

on the Capital Structure: A Natural

Experiment

Onur Kemal Tosun∗

Firms simultaneously chooseboth their capital and their executive compensation structure. Using

the Internal Revenue Code 162(m) tax law as an exogenous shock to compensation structure in a

natural experiment setting, I identify firm leverage changes as a result of chief executive officer

(CEO) option compensation changes. The evidence provides strong support for debt agency

theory. Firmsappear to decrease leverage when CEOs are paid with moreoptions and when CEO

options become a higher percentage of future cash flows. The findings are robust to controlling

for corporate governanceand convertible debt.

Numerous studies have examined the relation between chief executive officer (CEO) option

pay and firm capital structure choice. Some examine the relation between CEO option pay and

leverage using leverage as the dependent variable,explicitly assuming that pay structure variation

causes differences in observed firm leverage. Others use option compensation as the dependent

variable and investigate how it varies with the leverage decision. The empirical challenge is that

these are both firm choices that are arguably made simultaneously. Therefore, it is difficult to

identify the cause of this relation from the existing literature. In this article, I use an exogenous

shock that influences only CEO option compensation. Subsequently, I examine how changes

in option pay resulting from that shock affect the capital structure. This allows me to identify

whether changes in compensation structure cause changes in leverage ratios.

Specifically, my main research question is how CEO option compensation influences the

capital structure decision. There is substantial disagreement over the nature of this relation.

Bryan, Hwang, and Lilien (2000), Hassan and Hosino (2008), Andrikopoulos (2009), and Sepe

(2010) claim there is a negative relation between option pay and the leverage decision. Relying

on debt agency theory, these studies report a decreasing effect of CEO option compensation on

leverage. Jensen and Meckling (1976) suggest that stock options align CEO interests with those

of shareholders to mitigate agency problems between managers and shareholders. Debtholders,

however, demand higher returns because CEOs with high option payments have incentives to

increase stock price volatility by investing in riskier projects (see, e.g., Agrawal and Mandelker,

1987; DeFusco, Johnson, and Zorn, 1990), potentially transferring wealth from bondholders to

shareholders. Hence, to optimize the cost of debt, firms paying a large proportion of option-based

compensation optimally reduce leverage.

I appreciate the valuable comments from Raghu Rau (Editor) and an anonymous referee. I gratefully acknowledge the

contribution of Michael Faulkender, Gordon Phillips, Lemma Senbet, Gerard Hoberg, and seminar participants at the

University of Maryland, University of Kentucky,University of Warwick, University of Cambridge, Norwegian School of

Economics, Norwegian Business School, and Ozyegin University.All errors are my own.

∗Onur Kemal Tosun is an Assistant Professor of Finance in Warwick Business School at University of Warwick, United

Kingdom.

Financial Management •Winter 2016 •pages 953 – 979

954 Financial Management rWinter 2016

In contrast, relying on managerial agency theory arguments, Lewellen, Loderer, and Martin

(1987), Berger,Ofek, and Yermack (1997), MacMinn and Page (2006), Coles, Daniel, and Naveen

(2006), and Tchistyi, Yermack, and Yun(2011) argue in favor of a positive relation between CEO

option pay and leverage.If debtholders are not fully aware of CEO pay composition and its wealth-

shifting implications, CEOs with large option pay components choose to increase leverage.

Finally, other studies argue there is no relation between CEO option payand leverage. Yermack

(1995) and Mehran (1995), for example, do not find evidence of a significant relation between

equity-based pay and leverage. Hayes, Lemmon, and Qui (2012) examine the link between option

pay and risk-taking behaviorby using Financial Accounting Standards (FAS) 123R and the change

in accounting treatment of options, but they do not find a strong relation between the decline in

option pay and less risky investments.

I analyze an exogenous shock to CEO compensation structure, specifically, Section 162(m)

of the Internal Revenue Code (IRC 162(m)). The Revenue Reconciliation Act of 1993 added

Section 162(m) to restrict the corporate tax deduction for executive compensation to $1 million.

It contains an exception for performance-based compensation:

. . . In the case of any publicly held corporation, no deduction shall be allowedunder this chapter

for applicable employee remuneration with respect to any covered employee to the extent that

the amount of such remuneration for the taxable year with respect to such employee exceeds

$ 1,000,000. . . .

Consequently,beginning January 1, 1994, companies have largely adjusted their compensation

packages so that pay over $1 million qualifies under the performance-based exception. That

change primarily occurs in the form of increased option compensation. IRC 162(m) should

have no direct influence on the capital structure. This tax deduction limitation and the link

to performance-based compensation should not alter the tax benefits, f inancial distress costs,

information asymmetry, or market timing motivations of a firm when determining its optimal

capital structure. Hence, I use IRC 162(m) as a valid instrument for the exogenous shock in the

natural experiment. Moreover, IRC 162(m) is not a binding constraint for all companies because

only firms paying the CEO a cash salary of $1 million or more are affected. This binding constraint

enables me to compare these treatment firms (those paying at least $1 million in salary) with the

control companies before and after the exogenous shock using a triple difference-in-difference

analysis. I also compare firms in the pre-period with fir ms in the post-period.

I find that the increase in CEO option compensation has a mitigating effect on the general

trend of increasing leverage across firms. In particular, firms decrease leverage as the proportion

of CEO option compensation increases and as those options become a higher percentage of

the firm’s future cash flows. Moreover, firms reduce leverage as the value of the CEO option

compensation goes up. The findings are consistent with debt agency theory arguments.

I also address possible concerns about the natural experiment and alternative explanations for

the findings. Although the 162(m) tax law is an exogenous event for firms, it might be argued

that companies anticipated its occurrence and behaved accordingly. For example, firms might

have increased option pay and issued more debt before the tax law change. I show this is not the

case because the treatment firms (those with CEO salary of at least $1 million) and control f irms

show no rapid increase in option pay before the law. In addition, leverage appears to increase not

only for treatment firms but also for control fir ms before the shock. Additional placebo tests for

the period before the exogenous event should capture any significant anticipation by firms and

show there is no such anticipation.