EFFECT OF BANK MONITORING ON EARNINGS MANAGEMENT OF THE BORROWING FIRM: AN EMPIRICAL INVESTIGATION

• Date: 01 June 2015
• Author: Anand Jha,Arun Prakash,Siddharth Shankar
• DOI: http://doi.org/10.1111/jfir.12055
• Published date: 01 June 2015

EFFECT OF BANK MONITORING ON EARNINGS MANAGEMENT OF THE

BORROWING FIRM: AN EMPIRICAL INVESTIGATION

Anand Jha and Siddharth Shankar

Texas A&M International University

Arun Prakash

Florida International University

Abstract

The literature on bank monitoring posits that strong bank monitoring can either increase

or decrease the earnings management of the borrowing firm. We test these two

competing hypotheses and find that earnings management (measured by the absolute

value of the discretionary accruals) is higher when monitoring is strong—unless the firm

is close to default. This result suggests that it might not be appropriate to view bank

monitoring as always curbing corporate misbehavior. Our findings are based on

examining 22,526 financial statements of private and public firms in India between 2001

and 2009.

JEL Classification: G21, G34, M41

I. Introduction

A substantial body of research documents that managers manipulate earnings. Such

manipulations are costly. Firms that manage earnings pay a higher interest rate on their

debt, face stringent collateral terms (Bharath, Sunder, and Sunder 2008), credit

constraints (Beatty, Liao, and Weber 2010), and a higher cost of capital (Lambert, Leuz,

and Verrecchia 2007; Sohn and Kim 2013). From a policy perspective, earnings

management erodes the credibility of financial statements and hinders the efficient

allocation of capital.

Expectedly, there is a strong interest in understanding the reasons why firms

manage earnings and how it can be curbed. The earnings management literature shows

that strong monitoring from the board, audit committee (Klein 2002; Xie, Davidson, and

DaDalt 2003), and analysts (Yu 2008) reduces earnings management. But there is little

knowledge about how strong monitoring from banks affects earnings management. We

fill this gap in the literature.

We investigate whether strong bank monitoring increases or decreases earnings

management. Our question is particularly interesting because banks are supposed to be

We are grateful to Larry Wall (associate editor) and Peter Dedalt (referee) for their valuable comments. We

appreciate the helpful comments from George Clarke, Abu Jalal, Greg Nagel (discussant), and the participants at

the 2013 annual meeting of the Financial Management Association. We also thank Jonathan Moore forcopyediting

our manuscript.

The Journal of Financial Research Vol. XXXVIII, No. 2 Pages 219–253 Summer 2015

219

© 2015 The Southern Finance Association and the Southwestern Finance Association

RAWLS COLLEGE OF BUSINESS, TEXAS TECH UNIVERSITY

PUBLISHED FOR THE SOUTHERN AND SOUTHWESTERN

FINANCE ASSOCIATIONS BY WILEY-BLACKWELL PUBLISHING

important monitors of the firm. But the direction in which bank monitoring affects

earnings management is not clear.

On the one hand, some studies in the banking literature suggest that the earnings

management of the borrowing firm is negatively associated with bank monitoring. For

example, Fama (1985) and Diamond (1984) propose that banks have informational

advantages and are the delegated monitors. They argue that banks have the ability to curb

the opportunistic misbehavior of the managers. According to their studies, because the

earnings management is an example of opportunistic misbehavior, strong bank

monitoring should limit it. Ahn and Choi (2009) find empirical evidence that earnings

management is indeed negatively associated with bank monitoring.

On the other hand, there are studies from different streams of literature that

suggest earnings management might be positively associated with bank monitoring.

Strong monitoring involves more covenants, tighter covenants, and better enforcement

of covenant violations (Rajan and Winton 1995). The debt covenant and earnings

management literature suggests that faced with a large number of tighter covenants, the

borrowing firms’managers might manipulate earnings to a greater extent (Beneish and

Press 1993; Chava and Roberts 2008; Dichev and Skinner 2002; Jha 2013). Also, there

are a number of studies in the banking literature that suggests that bank monitoring might

not limit earnings management. The main idea in these studies is that the banks are more

interested in preserving their business relation with the firm than in disciplining the

managers (Cornett et al. 2007; Harris and Raviv 1990). They argue that the banks are

only interested in monitoring the firm when it is consequential for them to do so—that is,

when the borrowing firm is close to default. What is more, the banks might have a

perverse incentive to let its borrowing firm manage earnings. When managers manage

earnings, they make their firms opaque; this decreases their ability to get loans from new

banks. The lack of the outside opportunity to get new loans makes it easier for the

existing banks to “arm-twist”the firms into paying higher rates (Rajan 1992).

We test these two competing possibilities. Specifically, we ask the following

questions: How is the earnings management of the borrowing firm associated with (1)

whether its main bank is a private bank instead of a government bank, (2) its number of

bank relations, and (3) its proportion of bank debt? Based on the literature, we consider

private banks, the large number of bank relations, and the larger proportion of bank debt

to mean strong bank monitoring.

To examine these questions, we use the Prowess data set from India. We do so

because this data set provides us with a large sample of firms with more precise measures

of bank monitoring than those used in the literature. In addition to the financial data, the

data set provides the names of all of the banks, in their order of importance, with which

the firm has a banking relation. This feature gives us a unique opportunity to test the

impact of the ownership structure (government owned vs. privately owned) of the main

bank and the number of banking relations on the earnings management. Although the

idea that government banks are poor monitors compared to private banks is well

established, no study has used this measure of bank monitoring to test how bank

monitoring affects earnings management. The data set also provides the total debt

obtained from the banks as a separate category. Therefore, we are able to use the ratio of

bank debt to total debt, a much more precise measure for the importance of banking

220 The Journal of Financial Research

relations. These features are not available in the firm-level data in the United States. What

is more, India is among the set of countries where investor protection is weak and

earnings management is severe (Bhattacharya, Daouk, and Welker 2003; Leuz, Nanda,

and Wysocki 2003). This condition means that if bank monitoring affects earnings

management, it should be more easily observed in India.

Using the financial data available in Prowess, we calculate the absolute value of

the discretionary accruals—our measure for the firm’s earnings management. Then in a

multivariate framework, we examine the associations between the earnings management

and the three measures of bank monitoring.

Overall, our results suggest that the banks cannot be relied on as monitors to curb

earnings management. Rather, the results suggest that strong monitoring might actually

increase the incentives to manipulate earnings. Specifically, our results show that, ceteris

paribus, a firm with a private bank as the main bank uses about 13.9% more discretionary

accruals. We find that when the number of bankers increases from one to four, the use of

the discretionary accruals increases by about 5.3%. Furthermore, when the ratio of bank

debt to total debt increases from the 25th percentile to the 75th percentile, the use of the

discretionary accruals increases by about 5.7%. These results are robust to controlling for

a number of firm-level characteristics such as size, leverage, profitability, institutional

ownership, industry, and year fixed effects.

To understand why bank monitor ing does not curb earnings management, we

delve deeper. As additional tests, we examine if, as discussed in the literature, the

incentives to monitor can alte r depending on whether the firm is in di stress. That is,

we test whether the associat ion between earnings management and bank monitoring is

different for financially distressed firms compared to nondistressed firms. We

construct a measure of financia l distress based on Altman’sz-score and interact this

variable with the three measures of bank monitoring. Consistent with the literature,

we find that when a firm is close to default, bank monitoring seems to reduce the

extent of earnings managemen t. But when they are far from defau lt, the relation is

positive. Overall, this result suggests that the banks are indeed not interested in

monitoring the earnings man agement behavior of the firm unles s it is consequential to

do so.

We also investigate if how the borrowing firm’s earnings management affects

the interest rate charge by banks. To do so, we construct a measure of opacity based on

the extent of the earnings management. Then, we examine if this measure is associated

with a higher interest rate charged by the banks. We find, as in prior studies (Bharath,

Sunder, and Sunder 2008; Francis et al. 2005), a positive association between opacity and

the ratio of interest expense to total debt.

Based on these results, there are two explanations for the positive association

between the strength of bank monitoring and the amount of earnings management that

are not mutually exclusive. One is that the borrowing firm’s bank might benefit when that

firm manages its earnings. As mentioned earlier, when the borrowing firm manages

earnings, the firm becomes opaque. Because banks that have a relation with the firm can

“see through”the earnings management, at least to some extent, the increased opacity

translates into more risk only for new banks that the firm could potentially court—

creating an advantage that the current bankers can exploit.

Effect of Bank Monitoring 221