Do Opinions on Financial Misstatement Firms Affect Analysts’ Reputation with Investors? Evidence from Reputational Spillovers

• Published date: 01 September 2016
• DOI: http://doi.org/10.1111/1475-679X.12119
• Author: LIAN FEN LEE,ALVIS K. LO
• Date: 01 September 2016

DOI: 10.1111/1475-679X.12119

Journal of Accounting Research

Vol. 54 No. 4 September 2016

Printed in U.S.A.

Do Opinions on Financial

Misstatement Firms Affect Analysts’

Reputation with Investors? Evidence

from Reputational Spillovers

LIAN FEN LEE

∗AND ALVIS K. LO

∗

Received 31 October 2014; accepted 26 February 2016

ABSTRACT

We examine whether opinions on firms subsequently revealed to have mis-

stated earnings affect analysts’ reputation with investors. We find that positive

opinions by bullish analysts hurt their reputation, leading investors to react

less to their research on non-misstatement firms after the misstatement revela-

tion (i.e., negative spillovers). We also find that bearish analysts issuing more

∗Boston College.

Accepted by Douglas Skinner. We thank the Carroll School of Management at Boston

College for financial support. We are grateful to an anonymous reviewer for helpful sug-

gestions that greatly improved this paper. We would also like to thank Bill Baber, Joy Be-

gley, Mark Bradshaw, Shijun Cheng, Qiang Cheng, Weili Ge, Rebecca Hann, Gilles Hilary,

Amy Hutton, Edmund Keung, Marcus Kirk, Jerry Martin, Zoe-Vonna Palmrose, Chul Park,

Kyle Peterson, Sugata Roychowdhury, Philip Shane (discussant), Susan Shu, Billy Soo, Is-

abel Wang, Joanna Wu, Jerry Zimmerman, and the workshop participants at Boston Col-

lege, Boston University, Georgetown University, INSEAD, Michigan State University, Na-

tional University of Singapore, Singapore Management University, University of Florida,

University of Hong Kong, University of Maryland, University of Oregon, University of

Rochester, the 2014 UBCOW Research Conference, and the 2015 FARS mid-year meet-

ing for their useful comments and suggestions. We also thank Bill Mayew for sharing the

All-Star rankings by Institutional Investor. Last, we thank Susie Cao, Bobby Graf, Karen Jin,

Joseph Hong, Sean Li, Xiaohui Luo, Jiada Tu, David Shang, and Tiffany Wang for their

excellent research assistance. An online appendix to this paper can be downloaded at

http://research.chicagobooth.edu/arc/journal-of-accounting-research/online-supplements.

1111

Copyright C, University of Chicago on behalf of the Accounting Research Center,2016

1112 L.F.LEE AND A.K.LO

negative opinions gain reputation and experience positive spillovers. Finally,

for analysts who dropped coverage of the misstatement firm before the mis-

statement revelation, we find no spillovers, which suggests that analysts expe-

rience limited reputational gains when they did not issue a public negative

opinion.

JEL codes: G10; M40; M41

Keywords: analyst reputation; financial misstatement; reputational

spillover

1. Introduction

Analysts are sophisticated information intermediaries who form opinions

on a firm’s investment value by analyzing its fundamentals. Research sug-

gests that, for firms misstating earnings, some analysts question the fi-

nancial reporting process or business model before the misstatement is

revealed, and thus cease coverage or issue negative opinions (e.g., De-

chow, Sloan, and Sweeney [1996], Miller [2006], Dyck, Morse, and Zin-

gales [2010], Peng and Young [2013]).1An unaddressed issue is whether

these actions affect analysts’ reputation with investors. A bearish analyst may

gain reputation, or, because he was not bullish, minimize reputational loss.

Anecdotally, Healy and Palepu [2003, p. 19] highlight that “[bullish] an-

alysts have received considerable criticism for failing to provide an earlier

warning of problems at Enron.” However, Brown et al. [2015] interviewed

analysts who stated that they pay limited attention to potential misstate-

ments, as they believe the reputational consequences are minimal. In this

study, we examine whether an analyst’s opinions on misstatement firms af-

fect his reputation. This will further our understanding of how analysts gain

or lose reputation and whether reputational effects are important for them

to consider when covering firms with potential misstatements.

We identify financial misstatement firms as those associated with Ac-

counting and Auditing Enforcement Releases (AAERs) from the Secu-

rities and Exchange Commission (SEC).2Our test focuses on whether

misstatement revelations affect investors’ perceptions of the analyst, lead-

ing investors to change their reaction to the analyst’s research on non-

misstatement firms (i.e., spillover effects). For these spillovers to occur,

1For example, before the misstatement revelation of First Merchants Acceptance Corp., a

bearish analyst identified significant differences between trends at the company and those in

the industry. He cautioned that “First Merchants must contend with some serious problems:

it is underreserved, it has not demonstrated a consistent underwriting policy, and it lacks a

proven business model.” These opinions formed the basis of the analyst’s sell recommenda-

tion.

2Prior research uses AAERs to identify firms alleged to have accounting frauds (Beasley

[1996], Dechow, Sloan, and Sweeney [1996]). These cases are often prompted by a restate-

ment or lead to a restatement after the first revelation, and the misstated amount is typically

material. See, for example, Dechow et al. [2011] for further details.

ANALYSTS’REPUTATION WITH INVESTORS 1113

investors must judge that the analyst’s skills affect his analyses of both the

misstatement and the non-misstatement firms. This is the case when the

two firms are in the same industry, as the analyst’s industry knowledge is

critical in the assessment of the investment values of both firms (Hilary and

Shen [2013]). In addition to industry knowledge, skills such as gathering

and synthesizing information from multiple sources also affect the analyst’s

research on both firms. Hereafter, we use analyst ability to refer to an ana-

lyst’s proficiency in these skills.

When a misstatement is revealed, investors are likely to doubt bullish

analysts’ ability. They will, for example, question the analysts’ knowledge

of industry trends or whether they can effectively assess the reasonable-

ness of company-provided information when formulating their investment

opinions. These doubts lead investors to downplay the analysts’ research

on non-misstatement firms, which results in negative spillovers. Conversely,

investors are likely to deem the bearish analysts’ negative opinions as a pos-

itive reflection of their ability, which leads to positive spillovers. Regarding

the analysts who dropped coverage of the misstatement firm before the

misstatement revelation, prior studies suggest that they might have antici-

pated the misstatement (Dechow, Sloan, and Sweeney [1996]). As a result,

investors might view them favorably, but the positive spillovers could be

limited as the analysts did not issue a public negative opinion on the mis-

statement firm.

For each misstatement firm, we follow prior research (Dechow, Sloan,

and Sweeney [1996]) and verify when the misstatement was revealed to

the public. We identify analysts who followed the misstatement firm before

the misstatement revelation and measure their bullishness relative to other

analysts covering the same firm. We construct a summary measure that con-

siders analyst bullishness across three key research outputs: (1) stock rec-

ommendations, (2) target prices, and (3) annual earnings forecasts. In an

additional test for a hand-collected subsample, we also consider bullishness

in the analyst’s qualitative comments on the misstatement firm.

To test for spillovers, for all analysts covering misstatement firms (i.e.,

treatment analysts), we identify all the non-misstatement firms they cover.

Next, for each non-misstatement firm, we use all other analysts who do not

cover misstatement firms (i.e., control analysts) as a control to mitigate con-

founding time-varying factors that affect all analysts covering the same firm.

After identifying treatment and control analysts for each non-misstatement

firm, we obtain their annual and quarterly earnings forecast revisions in the

period from two years before to six months after the misstatement revela-

tion. Our test compares the change in market reactions to these forecast

revisions for a treatment analyst to that for a control analyst, after control-

ling for concurrent changes in both analyst and forecast characteristics.

Our main results are as follows.3First, for the bullish analysts, we find

a differential decline in the market reactions to their forecast revisions.

3For expositional convenience, bullish and bearish analysts refer to those with higher and

lower relative bullishness, respectively.