THE EFFECTS OF ANALYST FORECAST PROPERTIES AND COUNTRY‐LEVEL INSTITUTIONS ON THE COST OF DEBT

• Author: Omrane Guedhami,Anis Samet,Narjess Boubakri,Sadok El Ghoul
• DOI: http://doi.org/10.1111/jfir.12081
• Date: 01 December 2015
• Published date: 01 December 2015

THE EFFECTS OF ANALYST FORECAST PROPERTIES AND

COUNTRY-LEVEL INSTITUTIONS ON THE COST OF DEBT

Narjess Boubakri

American University of Sharjah

Sadok El Ghoul

University of Alberta, Edmonton

Omrane Guedhami

University of South Carolina

Anis Samet

American University of Sharjah

Abstract

We investigate the link between analyst forecast characteristics and the cost of debt

financing ininternational markets, and the influenceof country-level institutions. Using a

sample of 3,768 bond issues from 42 non-U.S. countries from 1996 to 2014, we find

statistically and economically significant evidence that analysts lower bond yield spreads.

Furthermore, this relation is stronger in firms operating in countries with weak institutions

governing property rights, creditor protection, and disclosure standards. Overall, our findings

imply that financial analysts play an important role as information intermediaries, and show

that this relation is especially important in countries with weak institutional environments.

JEL Classification: G32, G34

I. Introduction

The cost of debt issue has been the subject of considerable attention in the literature,

particularly in U.S. markets, where corporations have issued $4.6 trillion in corporate

bonds, compared to $1.5 trillion in equity over 1997–2008 alone (Bessembinder and

Maxwell 2008). Among scholars, there is consensus that the cost of debt is primarily

related to firm-level characteristics such as firm size, leverage, asset volatility, corporate

disclosure (Sengupta 1998; Yu 2005), corporate governance (Anderson, Sattar, and Reeb

2003; Mansi, Maxwell, and Miller 2004; Pittman and Fortin 2004; Ashbaugh-Skaife,

Collins, and LaFond 2006; Francis et al. 2012), earnings quality (Francis et al. 2005;

Francis, Nanda, and Olsson 2008), and individual stock liquidity (Huang, Huang, and

Oxman 2015). Recent evidence further shows that earnings’news conveys information

to credit markets (Callen, Livnat, and Segal 2009; Easton, Monahan, and Vasvari 2009;

We thank Najah Attig, Ruiyuan Chen, Larry Fauver (the referee), Andy Naranjo (the associate editor),

Mohsen Saad, and Walid Saffar for constructive comments. We appreciate generous financial support from

Canada’s Social Sciences and Humanities Research Council.

The Journal of Financial Research Vol. XXXVIII, No. 4 Pages 461–493 Winter 2015

461

© 2015 The Southern Finance Association and the Southwestern Finance Association

DeFond and Zhang 2014) and that analyst forecast properties affect the cost of debt and

credit ratings of U.S. firms (G€

untay and Hackbarth 2010; Mansi, Maxwell, and Miller

2011).

1

Yet little is known about the determinants of corporate debt pricing and the role

of financial analysts outside the United States.

2

Extending this line of research, we offer

the first evidence on the subject by looking at the role of analysts in international debt

markets and the conditioning impact of a country’s institutional environment on the

relation between analyst forecast properties and the cost of debt capital.

We begin by examining the link between analyst forecast properties (i.e.,

forecast dispersion, forecast accuracy, forecast revision, and an index using principal

component analysis based on the three previous characteristics) and both credit ratings

and credit spreads in a large sample of 3,768 bonds issued by 850 firms from 42 countries

between 1996 and 2014. We find statistically significant evidence that analyst forecast

characteristics affect firms’cost of debt financing and credit ratings. Economically, a

one-standard-deviation increase in analyst factor index decreases the credit spread by 31

basis points, which is material given that the mean (median) yield spread in our sample is

161 (132) basis points. These findings obtain after controlling for an extensive set of

bond- and firm-specific characteristics, and are robust to controlling for potential

endogeneity, employing alternative samples, and including additional country-level

factors. Taken together, the results imply that the scrutiny of financial analysts play an

informational role around the world.

Next, we investigate the extent to which country-level institutions affect the

importance of financial analysts to debt pricing. More specifically, we examine the link

between analyst forecast properties and credit spreads in countries with weak and strong

institutions governing property rights, creditor protection, and disclosure standards.

Although the relation between analyst forecast characteristics and firms’debt financing

costs is statistically significant in both sets of countries, we find it to be economically

more important in countries with weak institutions. This result suggests that the

informational and monitoring roles of financial analysts compensate for weak

governance institutions in these countries.

Our study contributes to the literature in three ways. First, by focusing on the link

between analyst forecast properties and the cost of debt in firms’home countries, we

complement recent evidence on a sample of bonds issued by U.S. firms (e.g., G€

untay and

Hackbarth 2010; Mansi, Maxwell, and Miller 2011). Extending this evidence to an

international context is important because institutional differences in legal and financial

systems could condition the availability of information and hence the role that financial

analysts play as information intermediaries and monitors. Prior studies show that the

extent of information asymmetry, which is at the root of the wedge between the internal

1

This evidence is consistent with Lang, Lins, and Miller (2004, p. 592), who argue that by providing

“potentially important scrutiny over management’s actions,”financial analysts play a dual role of (1) reducing

agency problems arising from managerial opportunism and (2) reducing information asymmetry problems between

managers and outside investors through information gathering and reporting.

2

According to Gozzi et al. (2012), bond markets constitute a larger and more internationalized source of

capital for firms than equity markets. Analyzing data for nearly 14,000 firms from 99 countries over 1991 to 2008,

they show that the aggregate amount raised by firms in bond markets around the world nearly doubled over that

period.

462 The Journal of Financial Research

and external costs of funding, depends on the institutional context (e.g., Love 2003;

Khurana, Pereira, and Martin 2006; Islam and Mozumdar 2007). Compared to other

countries, the United States has a well-developed “frictionless”financial market with

strong investor protection and contract enforcement, and with relatively better access to

outside financing (Francis et al. 2005). Consequently, the results documented for the

United States may not generalize to other countries with different legal and institutional

environments. Using an institutionally diverse sample of countries provides a natural

laboratory in which to explore whether analysts play as significant a role in debt markets

around the world as they do in the United States.

Second, by assessing the roleof analysts around the world, our article could have

important policy implications for environments where information disclosure and other

corporate governance mechanisms are weak, as doing so advances our understanding of

the potentialsubstitutability/complementaritybetween firm- and country-level monitoring

mechanisms.Our study therefore answers the callby Ramnath, Rock, and Shane (2008) for

further research on the institutional and regulatory factors that lead to cross-country

differences in the role of analysts.Third, according to Chang, Khanna, and Palepu (2001),

the financial crises in Asia and other emerging markets highlight the importance of

information intermediaries (such as analysts) to the functioning of capital markets. By

considering cross-country differences, we add to our understanding of how capital

markets function around the world and how bondholders assess the default risk of firms.

Overall, our study helps in the identificationof factors that drive the cost of debt financing

and hence firms’financing and growth potential, across countries as well as firms. Such

identificationis important given that firm growth directly affects overalleconomic growth.

II. Hypothesis Development

Analyst Forecast Properties and Debt Pricing

In this article, we explore the relations between credit ratings/spreads and properties of

analyst forecasts. The analyst forecast variables most widely used in the literature are: (1)

forecast accuracy, defined as the negative absolute value of the difference between the

median forecast and actual earnings per share (EPS) deflated by the stock price; (2)

forecast dispersion, defined as the standard deviation of the most recently revised

forecasts by all analysts deflated by the stock price;

3

(3) forecast revision, defined as the

standard deviation of the difference between the median forecast and the lagged median

forecast by all analysts deflated by the stock price; and (4) analyst factor index, an index

using principal component analysis based on the three previous analyst forecast

variables, as those variables are highly correlated.

Although the literature devotes much attention to the link between some or all of

these properties and firm performance, or the cost of equity capital, researchers have only

recently turned their attention to the potential spillover effects of these properties on debt

3

According to Collin-Dufresne, Goldstein, and Martin (2001), forecast dispersion largely captures future

cash-flow uncertainty in their sample of corporate bonds.

Effects of Analyst Forecast Properties 463