The Costs of Debt Contracts and the Repeated Use of Debt Covenants

• Published date: 01 December 2018
• Date: 01 December 2018
• DOI: http://doi.org/10.1111/ajfs.12240
• Author: Bong Hwan Kim

The Costs of Debt Contracts and the

Repeated Use of Debt Covenants

Bong Hwan Kim*

Graduate School of Public Administration and The Korea Institute of Public Affairs, Seoul National

University, Republic of Korea

Received 8 January 2018; Accepted 9 July 2018

Abstract

I examine the tendency to repeatedly use the same financial covenant in debt contracts and

the economic determinants of such repeated use. Using the Dealscan database, I find a ten-

dency to repeatedly use the same financial covenant. I also find that this tendency becomes

stronger when a lender has high search costs for an informative covenant or high negotiation

costs. I further find that this tendency becomes weaker when a borrower has a high risk of

default. My study provides evidence that the cost of writing a contract is economically mean-

ingful and needs to be considered to avoid overstating the effect of accounting informative-

ness on covenant choice.

Keywords Debt contracts; Financial covenants; Informativeness; Costs of contracts

JEL Classification: G14, G21, G24, G29

1. Introduction

I examine the tendency to repeatedly use the same financial covenant in debt con-

tracts for the same firm and the economic determinants of such repeated use. Using

the Dealscan database, I find a strong tendency to repeatedly use the same financial

covenant. I also find that this tendency becomes stronger when a lender has high

search costs for informative covenants or high negotiation costs. I further find that

this tendency becomes weaker when a borrower has a high risk of default.

Debt covenants as they pertain to the debt covenant hypothesis and the relation

between accounting conservatism and debt covenants have been extensively studied

in the accounting and finance literature (DeFond and Jiambalvo, 1994; Dichev and

Skinner, 2002; Beatty et al., 2008; Zhang, 2008). Recent studies expand this litera-

ture by investigating how a specific covenant is selected (Demerjian, 2007; Frankel

and Litov, 2007; Christensen and Nikolaev, 2012; Bradley and Roberts, 2015; De

Franco et al., 2015; Li, 2016). The literature argues that covenants that best measure

*Corresponding author: Graduate School of Public Administration and The Korea Institute

of Public Affairs, Seoul National University, Seoul 08826, Korea. Tel: +82-2-880-8083,

Fax: +82-2-877-2411, email: kimbong@snu.ac.kr.

Asia-Pacific Journal of Financial Studies (2018) 47, 858–880 doi:10.1111/ajfs.12240

858 ©2018 Korean Securities Association

a borrower’s credit risk are included in debt contracts (Smith and Warner, 1979;

Dichev and Skinner, 2002; Christensen et al., 2016). However, evidence on the role

of contracting costs in the selection of covenants is scarce. Understanding the selec-

tion process of a covenant is important because it guides the predictions for studies

related to debt covenants. My study enhances understanding of the covenant selec -

tion process by providing evidence on the costs of writing a contract.

Financial debt covenants are included in contracts to protect lenders from

unexpected deterioration of borrowers’ credit risk. Therefore, the informativeness of

borrowers’ credit risk is a key driver of covenant choice (Smith and Warner, 1979;

Dichev and Skinner, 2002; Demerjian, 2007; Dyreng et al., 2017). On the other hand,

the literature also argues that costs of debt covenants such as negotiation, monitoring,

and renegotiation are not negligible (El-Gazzar and Pastena, 1991; Frankel and Litov,

2007). Thus, I conjecture that the choice of covenants will reflect their benefits, i.e.,

informativeness of credit risk and the costs of including covenants in debt contracts.

Using the Dealscan database from Loan Pricing Corporation (LPC), I identify pri-

vate lending agreements containing all types of financial covenants. I then test

whether the choice of a covenant is explained by the existence of the same covenant

in prior loans after controlling for accounting informativeness. I test for the three

most commonly used covenants: Debt-to-EBITDA, Interest Coverage, and Fixed

Charge Coverage. I find a strong tendency to repeatedly use the same financial cove-

nants for the same firm in all three covenants. I conjecture and find that this tende ncy

will be even stronger when the cost to search for an optimal covenant, proxied by the

number of the borrower’s segments and the lender’s expertise in the industry, or the

cost to negotiate on a new covenant, proxied by the number of loan s a borrower has

at the time of loan contracts, is higher because the cost savings from the repeated use

of a covenant are greater. I also conjecture that a lender will strive to include an infor-

mative covenant if a borrower has a high risk of default. This will reduce the tendency

of repeated use of a covenant as lenders will choose more informative covenants at

the contract stage because the cost of uninformativenss of covenants is higher for this

type of borrower. I find repetition becomes weaker when a borrower has a lower

credit rating or a higher default risk or if a loan is collateralized.

My study contributes to the literature in several ways. Firstly, it sheds light on

the cost side of debt contracts by providing evidence on the costs of selecting debt

covenants. While prior studies suggest the benefits of debt covenants in signaling

credit risk and reducing agency problems, there is not much evidence on the cost

of writing contracts.

1

Prior studies on debt covenants either ignore the cost of debt

covenants or assume they are constant across contracts.

1

El-Gazzar and Pastena (1991)argue that multiple-lender loans have fewer financial restric-

tions because of high negotiation and renegotiation costs. Frankel and Litov (2007) argue

that firms minimize renegotiation, monitoring and administrative costs. The costs discussed

in these studies are mainly costs after contract while my study shows that costs before

contract affect covenant choice.

The Costs of Debt Contracts

©2018 Korean Securities Association 859