CEO inside debt and convertible bonds

• Author: Carl Hsin‐han Shen,Wei‐Hsien Li,S. Ghon Rhee
• Published date: 01 January 2018
• DOI: http://doi.org/10.1111/jbfa.12285
• Date: 01 January 2018

DOI: 10.1111/jbfa.12285

CEO inside debt and convertible bonds

Wei-Hsien Li1S. Ghon Rhee2,3 Carl Hsin-han Shen1

1Departmentof Finance, National Central

University,TaoyuanCity, Taiwan

2ShidlerCollege of Business, University of

Hawai'i,Honolulu, Hawai'i, USA

3Departmentof Finance, National Central

University

Correspondence

CarlHsin-han Shen, Department of Finance,

NationalCentral University, No. 300, Zhongda

Rd.,Zhongli District, TaoyuanCity 32001, Taiwan.

Email:hshen@ncu.edu.tw

JELClassification: G30, M12

Abstract

The question whether convertible bonds are issued to combat the

risk-shifting problem is a subject of debate in the literature,primarily

because of the unavailability of clear measures regarding manage-

rial risk-shifting incentives. Takingadvantage of recently developed

inside debt-holding measures for CEOs, we find strong evidence in

support of the risk-shifting hypothesis. When a CEO holds a large

amount of inside debt, three distinct patterns emerge: (i) the firm

exhibits a lower ratio of outstanding convertibles to total debt;

(ii) the firm is less likely to issue convertibles than straightdebt; and

(iii) the firm devisescontract terms to decrease the chance of conver-

sion when it issues convertibles.

KEYWORDS

convertible bond, inside debt, risk shifting

1INTRODUCTION

Financial economists havebeen challenged with the task of explaining the issuance of convertible bonds because these

hybrid securities do not fit perfectly into the classical capital theories on the equity-debt choice. The risk-shifting

hypothesis (RSH), first introduced by Jensen and Meckling (1976), is often proposed as a reason that convertiblesare

issued. The RSH posits that controlling shareholders pursue a level of asset risk that is excessive from debtholders’

point of view because the value of an equity claim is analogous to a call option written on asset value. Risk-shifting

firms thus experience relativelystrict financing constraints in the debt market.

The issuance of convertible bonds rather than straight bonds is a financing strategy that firms can use to offer

assurance to their debt investors. Jensen and Meckling (1976) argue that by issuing a bond-warrant combined secu-

rity,shareholders effectively share part of the proceeds of the increased asset risk with the firm's debtholders, in turn

curbing shareholders’ incentiveto pursue risk. Green (1984) presents a theoretical model and formalizes that intuition.

However,the RSH has received limited empirical support in the literature despite its theoretical validity. The main

challenge confronted by any empirical test of the RSH is the difficulty of finding clear empirical proxiesfor managerial

risk-shifting incentives. Earlier studies use firm characteristics as their empirical proxiesfor the degree of risk-shifting

problems. Lewis,Rogalski, and Seward (1999, 2003) propose that firms with fewer profitable investment opportunities

(measured by the market-to-book ratio)and greater idiosyncratic risk have a higher degree of risk-shifting concern. In

a similar vein, King and Mauer (2014) relate a firm's set of investment opportunities to the extent of its agency con-

cerns. Although these variables are arguably associated with agency costs, they do not exclusivelycapture managerial

risk-shifting incentives. For instance, substantial investmentopportunities and high risk may occur together with high

232 c

2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2018;45:232–249.

LI ET AL.233

information asymmetry, which increases both the cost of equity and the cost of debt. Convertibles maytherefore be

issued to resolve the informational problem associated with equity issuance (i.e., the backdoor equity hypothesispro-

posed by Stein (1992), rather than aproblem associated with debt issuance, as suggested by the RSH.1

Furthermore, managerial incentive measures based on equity-based contracts, such as restricted stocks and exec-

utive stock options, are also not appropriate measures of risk-shifting incentives. The relationship between equity-

basedcompensation and corporate decisions can be driven by both the incentive alignment effect and the risk-aversion

effect (Chava & Purnanandam, 2010; Low,2009). On the one hand, equity-based compensation helps align managers’

and shareholders’ interests and thus enhances risk-shifting incentives. On the other hand, if managers cannot freely

adjust their equity holdings because of trading restrictions associated with their contracts (e.g., lock-up periods for

restricted stocks and stock options), they are forced to hold an under-diversified portfolio and become more risk-

averse than their shareholders (Brockman, Martin, & Unlu, 2010; Guay,1999). Therefore, equity-based compensation

holding might capture conflicting managerial incentives and is not an ideal variable for the purpose of testing the RSH.

Recent studies on CEO inside debt, however, present an opportunity to resolve this empirical dilemma. Because

the risk-shifting problem represents a conflict between shareholders and debtholders, an appropriate proxy for man-

agerial risk-shifting incentives should be directly related to incentive alignment between managers and debtholders.

Both executive pension and deferred compensation plans represent debts that a company owes its managers, to be

paid upon those managers’ retirement. Sundaram and Yermack(2007) therefore denote these vehicles as ‘inside debt’,

as opposed to financial debt owed to outsiders. These authors suggest that inside debt aligns managers’ incentives

with debtholders’ incentives by uniting managers and debtholders in their risk of incurring losses should the company

default. Edmans and Liu (2011) formally model this insight.2

Recent studies also find empirical evidence to support the predictions set forth above. Cassell, Huang, Sanchez,

and Stuart (2012) document a positive association between inside debt holdings and corporate conservatism. Weiand

Yermack(2011) document positive abnormal bond returns when companies follow the SEC rule and file their first-time

disclosure of inside debt positions. Anantharaman, Fang, and Gong (2014) present evidence showing that loans issued

to companies with more CEO inside debt are associated with lower promised yields and fewer covenants. Srivastav,

Armitage, and Hagendorff (2014) find that the CEOs of banks with higher inside debt relativeto inside equity are more

likely to cut bank pay-outs and follow more conservative bank pay-out policies. Dang and Phan (2016) document a

positive relation between CEO inside debt holding and short-maturity debt issuance.

This line of research provides direct evidence that inside debt is an effective mechanism for aligning managers’

and debtholders’ incentives. Tothe extent that inside debt can curb managers’ risk-shifting incentives, debt investors

should be reassured when they observe that a CEO holds a large amount of inside debt. We therefore expect a nega-

tive association between a CEO's inside debt holding and the firm's preference for convertibles according to the RSH.

Conversely,if convertibles are issued primarily for other reasons, such as backdoor equity, inside debt holdings should

have little (orweak) explanatory power regarding firms’ decision to issue convertible securities.

Our empirical evidence strongly supports the RSH. Using a dataset consisting of 3,558 firm-year observations and

1,338 straight debt and convertible debt issues from 2006 to 2011, we find that firms with more CEO inside debt

holdings show a significantly lower preference for convertible bonds. Specifically, firms with more CEO inside debt

holdings havefewer outstanding convertible debts relative to total debts on their balance sheets, are less likely to issue

convertibles than to issue straight debts, and devise contract termsthat result i na lower chance of conversion when

they do issue convertiblebonds. The economic significance of our findings is also strong. As CEO inside debt measures

increase from the first to the third quartile in our sample, we estimate that the proportion of outstanding convertible

1Stein(1992) theorizes that convertible bonds can be used as a mechanism to reduce the informational discount imposed by outside investors on newly issued

shares. Therefore, convertibles are issued as ‘backdoor’equities in the sense that they are meant to be converted if the manager performs well in pursuing

shareholderinterests, as promised to the stock investors.

2Sometheoretical models on compensation design also focus on improving incentive alignment between CEOs and debtholders. Bolton, Mehran, and Shapiro

(2015)suggest that excess risk-taking in the banking industry can be addressed by a compensation contract based on both stock price and credit defaultswap

(CDS) spreads. Waltherand Klein (2015) suggest that the impact of contingent convertible bonds (CoCo bonds) on excessive risk-taking can be neutralized

onceCoCo bonds are considered in the compensation contract.