Did Regulation Fair Disclosure, SOX, and Other Analyst Regulations Reduce Security Mispricing?

• DOI: http://doi.org/10.1111/1475-679X.12051
• Date: 01 June 2014
• Author: NORMAN STRONG,EDWARD LEE,ZHENMEI (JUDY) ZHU
• Published date: 01 June 2014

DOI: 10.1111/1475-679X.12051

Journal of Accounting Research

Vol. 52 No. 3 June 2014

Printed in U.S.A.

Did Regulation Fair Disclosure,

SOX, and Other Analyst Regulations

Reduce Security Mispricing?

EDWARD LEE,

∗NORMAN STRONG,

∗

AND ZHENMEI (JUDY) ZHU

†

Received 11 March 2011; accepted 9 February 2014

ABSTRACT

Between 2000 and 2003 a series of disclosure and analyst regulations curbing

abusive financial reporting and analyst behavior were enacted to strengthen

the information environment of U.S. capital markets. We investigate whether

these regulations reduced security mispricing and increased stock market ef-

ficiency. After the regulations, we find a significant reduction in short-term

stock price continuation following analyst forecast revisions and earnings an-

nouncements. The effect was more pronounced among higher information

uncertainty firms, where we expect security valuation to be most sensitive to

regulation. Analyst forecast accuracy also improved in these firms, consistent

with reduced mispricing being due to an improved corporate information

environment following the regulations. Our findings are robust to controls

for time trends, trading activity, the financial crisis, analyst coverage, delist-

ings, and changes in information uncertainty proxies. We find no concurrent

effect among European firms and a regression discontinuity design supports

our identification of a regulatory effect.

∗The University of Manchester; †Fudan University.

Accepted by Philip Berger. We acknowledge the comments of Hans Christensen, Gerald

Lobo, Richard Taffler, Martin Walker, Eric Yeung, and seminar participants at the University

of Manchester and Xian Jiaotong University. We thank Abbie Smith, the original Editor, and

the anonymous referee for constructive comments and suggestions. We thank Peter Iliev for

sharing valuable research data. We acknowledge financial support from The Financial Re-

search Center at Fudan University.

733

Copyright C, University of Chicago on behalf of the Accounting Research Center,2014

734 E.LEE,N.STRONG,AND Z.ZHU

1. Introduction

October 2000 to April 2003 witnessed the enactment of a series of disclo-

sure and analyst regulations affecting U.S. capital markets, designed to im-

prove the corporate information environment and restore investor confi-

dence. The regulations, which included Regulation Fair Disclosure (Reg

FD), NASD Rule 2711, the amended NYSE Rule 472, the Sarbanes–Oxley

Act (SOX), the Global Research Analyst Settlement, and Regulation Ana-

lyst Certification, were intended to improve capital market transparency by

restraining analyst forecast and financial reporting biases. Previous studies

have largely evaluated the benefits and costs of these regulations through

their impact on firm and analyst behavior. We examine whether the regula-

tions increased informational efficiency by reducing security mispricing in

U.S. stock markets.

A firm’s information environment is the equilibrium outcome of a com-

plex interplay of factors. Foremost among these are the firm’s own disclo-

sures through official filings, press releases, and briefings, including finan-

cial signaling and capital market transactions. Reports and commentaries

by analysts and commentators external to the firm condition and enhance

a firm’s disclosures. Internal and external corporate governance arrange-

ments and the regulatory system discipline the quality of the corporate

information environment. Beyer et al. [2010] characterize the corporate

information environment in terms of three decisions that shape it, namely,

managers’ voluntary disclosure decisions, disclosures mandated by regula-

tors, and analysts’ reporting decisions. Disclosure regulations, such as Reg

FD and SOX, potentially affected disclosures mandated by regulators di-

rectly and the other two decisions indirectly. Analyst regulations, such as

NASD Rule 2711, the amended NYSE Rule 472, the Global Research An-

alyst Settlement, and Regulation Analyst Certification, potentially affected

reporting decisions by analysts directly and firms’ voluntary disclosure de-

cisions indirectly. The influence of these regulations on the corporate in-

formation environment, therefore, was potentially substantial. While the

literature tends to suggest that this influence should have been beneficial

(e.g., Ke, Petroni, and Yu [2008], Kothari, Shu, and Wysocki [2009], Amir,

Guan, and Livne [2010]), some studies acknowledge that the regulations

may have had adverse effects (e.g., Carney [2006], Sidhu et al. [2008]). We

examine the effect of the regulations on informational efficiency.

An inverse relation between security mispricing and the quality of the

corporate information environment is well-established theoretically and

empirically (e.g., Merton [1987], Brav and Heaton [2002], Zhang [2006]).

Thus, to the extent that the disclosure and analyst regulations improved

the corporate information environment, they should have reduced se-

curity mispricing. But the impact on mispricing is unlikely to have been

homogeneous across firms. As it is harder for investors to anticipate the

future earnings of firms with greater information uncertainty, we expect

the valuation of these firms to have benefited more from an improved

REG FD,SOX,AND OTHER ANALYST REGULATIONS 735

corporate information environment. Therefore, if the regulations reduced

security mispricing, we should observe a greater impact among firms

with greater information uncertainty, after controlling for risk and other

confounding effects.

To test these predictions, we examine two security mispricing effects,

namely, the short-term stock price continuation effects following new infor-

mation based on analyst forecast revisions and earnings announcements.

The analyst forecast revision effect captures the speed and efficacy of the

price discovery process associated with information that analysts dissem-

inate (Gleason and Lee [2003]). The postearnings announcement drift

(PEAD) effect captures the delayed price response due to investors’ fail-

ure to appreciate the full implications of earnings information (Bernard

and Thomas [1989]). Changes to the corporate information environment

due to the regulations potentially influenced both mispricing effects. An-

alysts act as information intermediaries and are sophisticated end users of

financial statement information. Changes in their reporting decisions af-

fect their forecast revisions directly, while changes in managers’ voluntary

disclosures and disclosures mandated by regulators affect their revisions in-

directly. Information disclosed in earnings announcements influences in-

vestor anticipation of a firm’s future earnings performance. Changes in

managers’ voluntary disclosures, mandated disclosures, and analysts’ de-

cisions affect investor earnings expectations, directly and indirectly. Stock

price adjustment delays can be due to information imperfections (Verrec-

chia [1980], Callen, Khan, and Lu [2013]) and reduced delays imply an

increase in stock market efficiency.

We use eight proxies to capture cross-sectional variation in firms’ in-

formation uncertainty. These are accruals quality, based on Francis et al.

[2007], firm size, firm age, analyst coverage, analyst forecast dispersion,

cash flow volatility, and stock return volatility, based on Zhang [2006], and

an aggregate measure that considers the joint effect of these seven indi-

vidual proxies. Empirical studies widely apply the seven individual prox-

ies to capture cross-sectional variation in firms’ information uncertainty.

High information uncertainty firms are likely to be more sensitive, and low

information uncertainty firms less sensitive, to the regulatory changes we

consider. Therefore, if the net effect of the disclosure and analyst regula-

tions on the corporate information environment was beneficial, we expect

a greater reduction in security mispricing among firms with lower accruals

quality, smaller size, shorter listing history, lower analyst coverage, higher

analyst forecast dispersion, higher cash flow volatility, and higher stock re-

turn volatility.

Our findings are as follows. From pre- to postregulation, we observe

significant declines in short-term price continuation effects based on an-

alyst forecast revisions and earnings announcements. The postregulation

decrease in risk-adjusted returns associated with these effects is consistent

with reduced security mispricing and an increase in the informational effi-

ciency of U.S. stock markets. Crucially, the reduced mispricing effects are