Journal of accounting, auditing finance, tập 26, số 03, 2011 7

Analysts’ Recommendation
Revisions and Subsequent
Earnings Surprises: Pre- and
Post-Regulation FD

Journal of Accounting,
Auditing & Finance
26(3) 475–501
Ó The Author(s) 2011
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0148558X11401556


Dan Palmon1 and Ari Yezegel2

Abstract
This study examines the extent to which analyst recommendations were useful in identifying earnings surprises during the pre- and post-Regulation Fair Disclosure (FD) periods. A
comparative analysis of the association between recommendation revisions and subsequent
earnings surprises suggests a significant decline in the predictive value of analysts’ recommendations after Regulation FD took effect. Recommendation revisions are roughly 55%
less useful in predicting earnings surprises in the post-Regulation FD period. Furthermore,
the average abnormal return earned by investors following analysts’ advice to exploit earnings surprises is approximately 70% lower in the post-Regulation FD period. Overall, this
article’s findings are consistent with Regulation FD having considerably reduced analysts’
comparative advantage in identifying earnings surprises.
Keyword
Regulation FD, analyst recommendations, earnings surprises, portfolio analysis

Introduction
Earnings-related selective disclosure was one of the most publicized cases of unfair disclosure that contributed to Regulation Fair Disclosure’s (FD) acceptance.1 The Securities and
Exchange Commission (SEC) received several thousand comment letters expressing frustration on the basis of the belief that corporations were giving earnings-related information

only to a select group of financial analysts and institutional investors. Relying on this information, analysts then made recommendations to their clients prior to earnings announcements, thus giving them an unfair advantage over other investors. Inevitably, such
disclosure policies helped certain selected investors earn profits or avoid losses at the
expense of other investors. In response, the SEC passed Regulation FD and listed earningsrelated disclosure on top of the list of potential material information that needs to be disclosed simultaneously to all market participants.

1

Rutgers Business School, NJ, USA
Bentley University, Waltham, MA, USA

2

Corresponding Author:
Ari Yezegel, Bentley University, 175 Forest Street, Waltham, MA 02452
Email:

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


476

Journal of Accounting, Auditing & Finance 26(3)

This study examines the extent to which analysts’ recommendations helped their clients
identify firms with earnings surprises before and after Regulation FD took effect.2 If analysts, as alleged, were privately communicating earnings-related information with managers
and providing consistent investment advice to their clients, then analysts’ recommendations
should have exhibited predictive power of upcoming earnings surprises in the preRegulation FD period. Furthermore, to the extent that Regulation FD was effective in curtailing selective disclosure, the predictive value of recommendations should have declined
after Regulation FD took effect.
We estimate recommendations’ predictive value of upcoming earnings surprises using
the association between recommendation revisions and unexpected earnings calculated
based on (a) time-series earnings expectations, (b) analysts’ earnings expectations, and

(c) earnings announcement returns. In addition, we examine the association between earnings surprises and recommendation revisions using regression analysis controlling for
postearnings announcement drift, return momentum, accruals anomaly, and institutional
trading. Finally, we construct a trading strategy designed to capture recommendation revisions’ predictive power of earnings surprises and compare the abnormal returns accrued by
this portfolio during the pre- and post-Regulation FD periods.
We find that prior to Regulation FD’s acceptance, upgraded firms exhibited 3-day
market-adjusted earnings announcement returns that were on average 0.93% higher than
downgraded firms. After Regulation FD took effect, the return differential between
upgraded and downgraded firms declined 55% to 0.43%. The regression analysis provides
similar results and supports the inference that the association between recommendation
revisions and subsequent earnings surprises declined after Regulation FD took effect.
Finally, the trading strategy analysis reveals an approximately 70% decline in the portfolio performance after Regulation FD took effect. Overall, our results are consistent with
Regulation FD having been effective in limiting selective disclosure and reducing analyst
recommendations’ predictive power of upcoming earnings surprises.
Regulation FD was preceded with intense objection that the rule would harm the level
of corporate disclosure. Consistently, prior studies on Regulation FD focused primarily on
whether the rule damaged corporate disclosure level, increased earnings volatility, and
reduced forecast accuracy. The literature suggests that Regulation FD did not have significant adverse effects on corporate disclosure. This article contributes to the extant literature
by comparing the extent to which recommendations were valuable to analysts’ clients in
identifying earnings surprises before Regulation FD and how much of this predictive value
was eliminated after Regulation FD took effect. We also examine the impact of Regulation
FD on the abnormal performance of investors who followed analysts’ recommendation
revisions with the intent of benefiting from analysts’ earnings-related private information.
Hence, our analysis provides important insights relating to the fundamental concern of certain investors earning abnormal profits at the expense of other investors based on selective
disclosure.

Literature Review and Hypotheses Development
In response to growing concerns of select individuals obtaining access to inside information, the SEC passed Regulation FD, which was concerned with the fair disclosure of nonpublic material information. Regulation FD required managers to disseminate any material
information simultaneously to all market participants and prohibited selective disclosure.
Many securities markets professionals and institutional investors argued that bringing

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

477

additional restrictions on corporate disclosure would reduce the quantity and quality of
information available to capital markets.
After Regulation FD took effect, there was increased concern by practitioners that
Regulation FD increased return volatility, adversely affected analysts’ earnings forecasts,
reduced corporate disclosure, and increased information asymmetry. However, academic
research investigating Regulation FD’s impact provided mixed inferences. Table 1 presents
studies that investigated Regulation FD’s impact and summarizes their findings.
On the return volatility aspect, while Heflin, Subramanyam, and Zhang (2003),
Eleswarapu, Thompson, and Venkataraman (2004), and Sinha and Gadarowski (2010)
found a significant decline after Regulation FD, Bailey, Li, Mao, and Zhong (2003) and
Francis, Nanda, and Wang (2006) found no change in volatility after Regulation FD took
effect. Analogous to studies on return volatility, no consensus was reached on whether
analysts’ forecasts suffered or improved after Regulation FD. Heflin et al., Bailey et al.,
and Francis et al. found no significant impact of Regulation FD, whereas Agrawal,
Chadha, and Chen (2006) and Mohanram and Sunder (2006) documented deterioration.
Similarly, Heflin et al. and Francis et al. reported no change in forecast dispersion after
Regulation FD, whereas Bailey et al., Irani and Karamanou (2003), and Agrawal et al.
found forecast dispersion to have increased. Results on corporate disclosure were also
mixed. Heflin et al. and Bailey et al. documented an increase in conference call frequency, and Irani (2004) found that conference calls became more useful in helping analysts increase forecast accuracy. In contrast, Bushee, Matsumoto, and Miller (2004) found
that corporate conference call policy did not change after Regulation FD.
Studies examining Regulation FD’s impact on information asymmetry reached inferences ranging from an increase to a decrease in information asymmetry after Regulation
FD took effect. On one hand, Eleswarapu et al. (2004), Chiyachantana, Jiang,
Taechapiroontong, and Wood (2004), and Ahmed and Schneible (2007) documented evidence consistent with an improvement after Regulation FD. On the other hand, Sidhu,

Smith, Whaley, and Willis (2008) and Gomes, Gorton, and Madureira (2007) reported evidence suggesting an increase in information asymmetry. Finally, Charoenrook and Lewis
(2009) and Collver (2007) found no change in information asymmetry.
Another strand of literature examined the informativeness of analyst reports before and
after Regulation FD took effect. Gintschel and Markov (2004) studied the value of analysts’ information outputs using the return volatility surrounding analysts’ announcements.
They found that the absolute price impact of financial analysts’ forecasts and recommendations declined by 28% after Regulation FD consistent with Regulation FD having curtailed
selective disclosure. Cornett, Tehranian, and Yalcin (2007) provided further evidence by
evaluating the impact of Regulation FD on affiliated versus unaffiliated analysts. Their
results suggested that the market reaction to affiliated analysts’ recommendation changes
decreased significantly after the passage of Regulation FD. Francis et al. (2006) provided
supporting evidence to Gintschel and Markov’s (2004) results using American Depositary
Receipt (ADR) firms to control for confounding events.
The prior literature provides extensive evidence on the effect of Regulation FD on the
preearnings announcement informational efficiency, analysts’ forecast accuracy and dispersion, the informativeness of analysts’ reports, and corporate disclosure. However, little is
known about Regulation FD’s impact on the usefulness of analysts’ advice in identifying
earnings surprises. The unfair disclosure of information causing some investors to take
positions prior to earnings announcements with advance knowledge of the outcome was
one of the central issues in the debate surrounding Regulation FD. We contribute to the

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


478

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012

1995Q4–2001Q3

Irani and Karamanou
(2003)
Bushee, Matsumoto,

and Miller (2004)

10/1999–7/2001

10/1999–10/2001

1998–2001

3/1995–6/2004

Chiyachantana, Jiang,
Taechapiroontong,
and Wood (2004)

Gintschel and Markov
(2004)

Irani (2004)

Agrawal, Chadha, and
Chen (2006)

3/1999–10/2001

1999Q4–2001Q2

1999Q4–2001Q2

Sample period

Heflin, Subramanyam,
and Zhang (2003)

Bailey, Li, Mao, and
Zhong (2003)

Study

Table 1. Review of Prior Studies

- Conference call relevance measured
as the call’s ability to improve
forecast and consensus accuracy
- Forecast accuracy
- Forecast dispersion

- Return volatility surrounding earnings
forecasts and recommendations

- Abnormal return volatility and
trading volume
- Absolute consensus and time-series
forecast error
- Analyst information advantage
- Forecast dispersion
- Market reaction to earnings
announcements
- Forecast accuracy
- Forecast dispersion
- Voluntary disclosure

- Analyst following
- Forecast dispersion
- Timing of conference calls
- Number of conference calls
- Price volatility and trading volume
- Liquidity
- Adverse selection component
- Retail and institutional trades

Variables of interest

- After Regulation FD, forecast accuracy decreased and
dispersion increased particularly for early forecasts
(continued)

- Decrease in analyst following
- Increase in analyst forecast dispersion
- No substantial impact on conference call policies
- The information content of conference calls does not appear
to have declined
- Increase in liquidity
- Decrease in information asymmetry
- Decrease in preannouncement period institutional trading and
increase in retail trading activity after earnings announcements
- Return volatility associated with analyst announcements is on
average 28% lower in the in the post-Regulation FD period
- The difference in price impact between ‘‘The Leaders’’ and
other brokerage houses is 65% smaller in the post-Regulation
FD period
- Relevance of conference calls to analyst forecasts increased in

the post-Regulation FD period

- No change in return volatility around earnings announcements
- No change in forecast accuracy and increase in forecast
dispersion
- No evidence that Regulation FD prevents or enhances
information leakage during the preannouncement period
- Increase in voluntary disclosure
- No significant change in forecast accuracy or dispersion
- Some improvement in stock price efficiency
- Substantial increase in voluntary disclosure frequency

Main findings


479

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012

10/1998–11/2002

1997–2002

1999Q4–2001Q2

Cornett, Tehranian,
and Yalcin (2007)

Gomes, Gorton, and
Madureira (2007)

Ahmed and Schneible
(2007)
Charoenrook and
Lewis (2009)

- Analyst coverage
- Forecast error
- Return volatility on earnings
announcements
- Cost of capital
- Trading volume
- Earnings announcement return
Analysts’ responsiveness to:
- Earnings guidances
- Press releases
- Earnings announcements

- Hasbrouck summary informativeness
statistic
- Market reaction to analysts’
recommendation changes

Variables of interest

- No change in the aggregate amount of firm-specific information
conveyed by public disclosure events

- Reduction in information quality differences among investors

- No substantial change in informed trading that can be solely
attributed to Regulation FD
- Return volatility surrounding analysts’ recommendation changes
decreased significantly after the passage of Regulation FD
- Investors’ differential reaction to downgrades by affiliated and
unaffiliated analysts declined; no significant change was evident
for upgrades
- Increase in forecast error and earnings announcement return
volatility particularly for small firms

Main findings

Note: This table provides a summary of the prior studies that investigate the impact of Regulation FD. The sample period, variables of interest, and main findings are presented
for each study.

2000–2002

8/1999–1/2002

Sample period

Collver (2007)

Study

Table 1. continued


480

Journal of Accounting, Auditing & Finance 26(3)

literature by examining the extent to which analysts’ recommendation revisions were useful
in predicting earnings surprises in the pre- and post-Regulation FD periods and estimating
the level of abnormal returns that investors could have earned by following analysts’
recommendations preceding earnings announcements.
Analysts disclose forecasts and recommendations in their reports to their clients.
Forecasts represent analysts’ predictions of various financial statement line items (e.g.
earnings, sales) and do not necessarily convey information about analysts’ assessments of
companies’ intrinsic values relative to their stock prices (e.g., overvalued or undervalued). Conversely, recommendation ratings represent direct indication of analysts’ assessment of the valuation of the company. Consistently, investors react more strongly to
recommendation revisions than they do to earnings forecast revisions.3 Hence, analysts
who intend to give early warnings to their clients about earnings announcements are
more likely to communicate this through their recommendation ratings rather than their
earnings forecasts.
Furthermore, analysts can more effectively communicate information they received
through selective disclosure via recommendation ratings rather than earnings forecasts
because they may not have received a precise forecast from management. Many analysts
argued against Regulation FD on the basis that they only receive ‘‘soft’’ information from
managers, which could not be communicated to the public in a form other than selective
disclosure. Managers are likely to be reluctant to give precise figures about upcoming
earnings to analysts via selective disclosure because they themselves may not have precise
knowledge of upcoming earnings at the time. Anecdotal evidence suggests that managers
often limited their communications to general directional guidance such as ‘‘earnings
are likely to be better than expected’’ or ‘‘the current earnings expectations are unrealistic.’’4 Analysts, when not provided a point-forecast from managers about upcoming earnings,
are likely to limit their earnings forecast revisions. In contrast, through recommendations,
analysts can signal upcoming negative or positive earnings news without giving precise information about the degree of the earnings surprise.
If analysts, as alleged, were privately communicating earnings-related information with
managers and providing consistent investment advice to their clients, then analysts’ recommendations should have possessed predictive power of upcoming earnings surprises during
the pre-Regulation FD period. Furthermore, to the extent that Regulation FD was successful
in reducing selective disclosure, the association between recommendation revisions and

earnings surprises should have weakened in the post-Regulation FD period.
Hypothesis 1 (H1): The association between changes in analysts’ recommendations
and subsequent earnings surprises declined after Regulation FD took effect.
Supporters of Regulation FD argued that selective disclosure helped select analysts and
their clients reap economic benefits at the expense of other investors. If, as alleged, analysts
guided their clients to earn profits based on the private earnings guidance they received
from management, we should observe significant abnormal returns associated with implementing a trading strategy that follows analysts’ recommendations and liquidates after earnings announcements. Furthermore, to the extent that Regulation FD limited earnings-related
selective disclosure, we should observe a reduction in the abnormal performance of the
trading strategy after Regulation FD took effect.

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

481

Hypothesis 2 (H2): The profitability of a trading strategy intended to capture the private earnings-related information conveyed by recommendation revisions declined
after Regulation FD took effect.

Research Design
The period before an earnings announcement corresponds to a time in which firms are
likely to have prepared financial statements and managers have the greatest knowledge of
the current quarter’s earnings. If firm executives selectively disclose earnings-related information to analysts, then this information is most likely to be privately communicated to
analysts during the period before earnings announcements.
To isolate recommendations that may be associated with analysts’ communications with
managers about upcoming earnings, we limit our sample to analysts’ recommendation revisions made during the 3-week period ending 2 days before the earnings announcement.5
We then examine Regulation FD’s impact on these recommendations’ predictive value of
upcoming earnings surprises by estimating and comparing the association between recommendation revisions and subsequent earnings surprises during the pre- and post-Regulation
FD periods using univariate and regression analysis.6 Figure 1 provides an illustration of

the timeline and the main tests of this study.

Univariate Analysis
The univariate analysis examines the mean earnings surprise that follows upgrades and
downgrades and tests whether a significant change is evident after Regulation FD took
effect. For robustness, earnings surprise is computed using four alternative methods:
(a) standardized unexpected earnings based on time-series expectations, (b) standardized
unexpected earnings based on analyst expectations, (c) 3-day earnings announcement
abnormal returns, and (d) 2-day earnings announcement abnormal returns.
Standardized unexpected earning based on time-series expectation is computed as
follows:
SUEit ¼

et À etÀ4
st;tÀ8

where et is firm i’s earning (excluding extraordinary items) for quarter t and st,t–8 is the
standard deviation of unexpected earnings (et 2 et–4) in the past eight fiscal quarters. SUE
deciles are constructed and transformed to range between 20.5 and 10.5. The construction
of SUE deciles controls for common marketwide effects and reduces the influence of
extreme values on the results.
Standardized unexpected earning based on analysts’ expectations is computed as
follows:
ASUEit ¼

et À e^t
;
pt

where et is firm i’s actual earnings for quarter t, e^t is the earnings forecast consensus

defined as the median of all analysts’ latest earnings forecasts made after the previous quarter’s earnings announcement, and pt is the firm’s stock price at the end of quarter t 2 1.
ASUE deciles are constructed and transformed to range between 20.5 and 10.5.

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


482

Journal of Accounting, Auditing & Finance 26(3)
Panel A

Fiscal
Fiscal
Quarter
Quarter
Begins
Begins

Fiscal
Fiscal
Quarter
Quarter
Ends
Ends

Firm
generates
revenues
Firm
Firmgenerates

generatesrevenues
revenues
and
incurs
expenses
and
andincurs
incursexpenses
expenses

Management
obtains
rough
Management
Managementobtains
obtainsaaarough
rough
idea of
of upcoming
upcoming earnings.
earnings.
idea
idea of upcoming earnings.
Potential period
period for earningsearningsPotential
Potential periodfor
for earningsrelated selective
selective disclosure.
disclosure.
related

related selective disclosure.

Analysts
Analystsreceiving
receivingearningsearningsrelated selective disclosure
related selective disclosure
revise recommendations.
revise recommendations.
(–22,–2)
(-22,-2)
Upgrades and Downgrades
Upgrades & Downgrades

Earnings
Earnings
are
are
reported
reported

Earnings
Earnings
Surprise
Surprise
1.1.CAR
CAR
2.2.SUE
SUE
3.3.UE
UE

Panel B
Earnings Surprise

Pre-Regulation FD

Upgrade
Upgrade
Association

Return
Difference
CAR (0,1)
CAR (–1,1).

Unexpected
Earnings
Difference
(Time-Series)

Unexpected
Earnings
Difference
(Analyst)

Downgrade
Downgrade

Post-Regulation FD

Earnings Surprise

Upgrade
Upgrade
Association
Downgrade
Downgrade

Return
Difference
CAR (0,1)
CAR (– 1,1).

Unexpected
Earnings
Difference
(Time-Series)

Unexpected
Earnings
Difference
(Analyst)

Figure 1 Research design
Panel A: Timeline of the empirical analysis
Panel B: Association test for the pre- and post-Regulation periods
Note: This figure provides a visual overview of the research design. Panel A illustrates which analyst
recommendation revisions are included in the analysis in relation to the firm’s quarterly cycle. Panel B
shows the main test of the article and lists which earnings surprise measures are used.

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

483

Earnings announcement abnormal returns are computed as follows:
CARit ¼

1
X

ðRmit À Mktm Þ;

m¼À1

where m is equal to 0 on firm i’s quarter t earnings announcement date. Rm,i,t is firm i’s
daily return on the mth day of quarter t’s earnings announcement, and Mktm is the Center
for Research in Security Prices (CRSP) New York Stock Exchange (NYSE)/American
Stock Exchange (AMEX)/NASDAQ value-weighted daily return for day m.7
For each quarter, we compute the mean earnings surprise for upgraded and downgraded
firms. As recommendation revisions can be driven by marketwide information that affects
numerous firms, earnings surprises that follow revisions may be correlated across firms.
Therefore, we follow the Fama and Macbeth (1973) procedure and first compute average
quarterly earnings surprises and then calculate time-series averages and t-statistics of quarterly average surprises.

Regression Analysis
The regression analysis allows us to examine the change in recommendations’ predictive
value while controlling for confounding factors. To measure the change in the association

between recommendation revisions and earnings surprises, we estimate the following
regression models:
SSUEit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit
þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit þ b8 ACCRit
þ b9 CHNG IOit þ eit :
SASUEit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit þ b5 REVit
3 SOXit þ b6 LANCRETit þ b7 LRETit

ð1Þ

ð2Þ

þ b8 ACCRit þ b9 CHNG IOit þ eit :
CARðÀ1; þ1Þit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit
þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit þ b8 ACCRit
þ b9 CHNG IOit þ eit :

ð3Þ

CARð0; þ1Þit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit
þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit þ b8 ACCRit

ð4Þ

þ b9 CHNG IOit þ eit :
In the regression analyses above, we regress alternative earnings surprise measures on
the recommendation revisions made during the 3-week period before earnings announcements (REV), a Regulation FD indicator variable that takes a value of 1 for calendar quarters during the post-Regulation FD period (FD) and the interaction of REV and FD
variables (REV 3 FD). The coefficient of the REV variable tests whether recommendation

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012

484

Journal of Accounting, Auditing & Finance 26(3)

revisions have predictive power of subsequent earnings surprises. The interaction of the
REV and FD variables, labeled REV 3 FD, tests whether the association between analysts’
revisions and subsequent earnings surprises declined after Regulation FD.
In addition, we examine whether the Sarbanes-Oxley Act had a significant impact on
the association between recommendation revisions and earnings surprises by including
a dummy variable labeled SOX that takes a value of 1 for calendar quarters ending after
December 31, 2003, and the interaction of this variable with the recommendation revision
variable (REV 3 SOX). The coefficient of REV 3 SOX tests whether the predictive value
of recommendation revisions changed after the Sarbanes-Oxley Act took effect.
There are several potential confounding factors that may be correlated with analysts’
revisions. Bernard and Thomas (1989) found that stock prices underreact to earnings
announcements and this leads to a postearnings-announcement drift that is concentrated
around subsequent earnings announcements. Therefore, it is possible that analysts revise
their recommendation ratings in reaction to the previous quarter’s earnings results rather
than to their own information acquisition efforts or selective disclosure. To control for the
potential effect of postearnings announcement drift, we include the previous quarter’s earnings announcement market-adjusted returns (LANCRET) in the empirical model. To the
extent that there is an underreaction to the previous quarter’s earnings announcement,
LANCRET will be positively correlated with the current quarter’s earnings surprise.
Furthermore, Jegadeesh and Titman (1993) found that past winners outperform past
losers. Analysts aware of the positive association between past and future returns may
revise their recommendations accordingly, and this can result in an association between
revisions and subsequent earnings surprises. To control for the momentum effect, we measure the 3-month buy-and-hold return ending in the 2nd month of the firm quarter (LRET)
and include this variable in our regression model.8
In addition, Sloan (1996) found evidence suggesting that investors fixate on bottom

line earnings and ignore the accruals component of earnings. He finds that accruals are
negatively associated with the subsequent year’s abnormal returns and demonstrates that
the mispricing is mainly corrected during subsequent quarterly earnings announcements.
As previously announced accruals are negatively associated with subsequent earnings
announcement returns, analysts may be revising their recommendation ratings in response
to past accruals rather than to their private communication with management. Therefore,
we also control for the accrual component in the regression analysis. We compute
accruals (ACCR) as in Sloan (1996) using the most recently announced annual reports.9
The results are similar when we alternatively use discretionary accruals estimated from
the modified Jones model (Dechow, Sloan, & Sweeney, 1995), or from the performanceaugmented modified Jones model (Kothari, Leone, & Wasley, 2005), or using statement
of cash flows data.
Finally, Ali, Durtschi, Lev, and Trombley (2004) document that institutional investors
have superior knowledge of upcoming earnings results and that institutional trades are positively correlated with future earnings announcement returns. Again, analysts’ revisions may
be correlated with subsequent earnings surprises because analysts simply respond to contemporaneous institutional trading. To control for this possibility, we measure the change
in institutional ownership during the most recent calendar quarter (CHNG_IO) and include
it in the regression model. Table 2 provides detailed descriptions of the variables used in
the regression analysis.

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

485

Table 2. Variable Definitions
Variable
SSUE

SASUE

CAR (21, 11)

CAR (0, 11)

REV
FD
SOX
LANCRET

LRET
ACCR

CHNG_IO

Definition
Quarterly decile of unexpected earnings defined as (et 2 et–4) scaled by the
standard deviation of unexpected earnings (st,t–8). e is basic earnings per share
excluding extraordinary items (epsfxq), adjusted for stock splits and stock
dividends. The variable is transformed to range between 20.5 and 10.5
Quarterly decile of unexpected earnings defined as actual earnings reported by IBES
minus median earnings estimate of analysts scaled by the price at the end of the
previous fiscal quarter. The deciles are transformed to range between 20.5 and
10.5
Cumulative market-adjusted returns during the 3-day period centered on the
earnings announcement date. The CRSP NYSE/AMEX/NASDAQ value-weighted
index return is used as the market return. Returns are adjusted for delisting
Cumulative market-adjusted returns during the 2-day period beginning on the
earnings announcement date. The CRSP NYSE/AMEX/NASDAQ value-weighted
index return is used as the market return. Returns are adjusted for delisting

Recommendation revisions during the 3-week period ending 2 days before the
earnings announcement date
An indicator variable that takes a value of 1 for fiscal quarters ending after
October 23, 2000, and 0 for prior fiscal quarter
An indicator variable that takes a value of 1 for fiscal quarters ending after
December 31, 2003, and 0 for other quarters
Previous fiscal quarter’s earnings announcement return, which is defined as the
cumulative market-adjusted return during the 3-day period centered on the
earnings announcement date. The CRSP NYSE/AMEX/NASDAQ value-weighted
index return is used as the market return. Returns are adjusted for delisting
The 3-month buy-and-hold return ending at the end of the fiscal quarter’s 2nd
month
DCA – DCL – DEP scaled by average total assets where DCA is the change in
current assets (act) minus the change in cash and short-term investments (che),
DCL is the change in current liabilities (lct) minus the sum of changes in debt in
current liabilities (dlc) and income taxes payable (txp), and DEP is depreciation
and amortization (dp)
The change in institutional ownership percentage compiled from the ThomsonReuters Institutional Holdings (13F) database

Note: This table lists and defines the variables used in the study. The first column reports the variable name and
the second column provides the definition.

Trading Strategy Analysis
The trading strategy analysis involves constructing a portfolio that aims to capture abnormal returns earned by investors who followed recommendations with the intent of exploiting analysts’ earning-related private information. The trading strategy focuses on
recommendation revisions made after the fiscal quarter-end date and before the earnings
announcement date. The hedge portfolio purchases (sells) shares of upgraded (downgraded)
firms 1 day after the revision date and holds these shares until 1 day after the earnings are
announced. In this strategy, all dates are known in event time; hence, no hindsight bias is
introduced to the analysis.
To estimate abnormal returns, we first compute the daily raw returns that accrue to

upgrade and downgrade portfolios. Each firm that is upgraded after the fiscal quarter-end

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


486

Journal of Accounting, Auditing & Finance 26(3)

enters the upgrade portfolio 1 day after the revision date and remains in the portfolio until
1 day after the firm announces its quarterly earnings results. Similarly, each firm that is
downgraded after the fiscal quarter-end enters the downgrade portfolio 1 day after the revision date and remains in the portfolio until 1 day after the firm announces its quarterly
earnings results. We calculate value-weighted daily returns for each portfolio as follows:
Rpt ¼

nX
p;tÀ1

xjtÀ1 Rj;t ;

j¼1

where Rj,t is the day t return on security j, np,t-1 is the number of firms in the portfolio, and
xj,t–1 is the day t 2 1 market capitalization of firm j divided by the sum of day t 2 1
market capitalization of all the firms in the portfolio. The daily portfolio returns are then
compounded to monthly returns:
"
#
nt
Y

ð1 þ Rpt Þ À 1
Rpq ¼
t¼1

where nt is the number of trading days in the month q and Rpt is the raw monthly return for
the portfolio on day t.
We construct a hedge portfolio that goes long on the upgrade portfolio and short on the
downgrade portfolio. The hedge portfolio’s return is equal to the difference between the
returns of the upgrade and downgrade portfolios. Finally, we subtract the risk-free rate
from the upgrade and downgrade portfolios to compute the excess returns of the two
portfolios.10
We estimate the average monthly abnormal return associated with each portfolio by estimating the four-factor model represented by the equation below. The intercept of this equation (Jensen’s a) serves as an estimate of the average monthly abnormal return:
Rpt À Rft ¼ ap þ bp Mktt þ sp SMBt þ hp HMLt þ up UMDt þ ept ;
where Mkt is the market risk premium that is equal to the market return minus the risk-free
rate, SMB is the average return on three small-market capitalization portfolios minus the
average return on three large-market capitalization portfolios on day t, HML is the average
return on two high book-to-market equity portfolios minus the average return on two low
book-to-market equity portfolios for day t, and UMD is the average of the returns on two
(big-sized and small-sized) high prior return portfolios minus the average of the returns on
two low prior return portfolios, where a big-sized company is identified as being larger
than the median NYSE market cap.

Sample
The initial sample consists of all firms traded in the NYSE and AMEX and NASDAQ that
have data available in both CRSP and Compustat files. The sample spans the fiscal quarters
between 1995Q1 and 2009Q2. The sample begins in 1995Q1 because the Institutional
Brokers’ Estimate System (IBES) recommendation file is sparse for the period before 1994,
and we require 1 year of prior recommendation data to compute analysts’ previous recommendations. The sample ends in 2009Q2 because that is the latest fiscal quarter for which
we have data on Compustat files. We exclude closed-end funds, investment trusts, units,

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

487

Table 3. Descriptive Statistics

MV ($MM)
PRC ($)
BETA
B/M
CAR (21, 11)
CAR (0, 11)
SSUE
SASUE
REV
FD
REV 3 FD
SOX
REV 3 SOX
LANCRET
LRET
ACCR
CHNG_IO

Mean

1st quartile

Median

3rd quartile

SD

8,254.079
30.191
1.326
0.481
0.002
0.001
0.005
0.004
20.144
0.632
20.099
0.391
20.042
0.002
0.020
20.045
0.003

499.320
14.320
0.661
0.241
20.037

20.034
20.278
20.167
21.000
0.000
21.000
0.000
0.000
20.039
20.126
20.081
20.022

1,598.876
25.205
1.122
0.397
0.001
0.000
0.056
20.056
21.000
1.000
0.000
0.000
0.000
0.001
0.012
20.042
0.004

6,034.567
40.250
1.752
0.616
0.041
0.037
0.278
0.278
1.000
1.000
1.000
1.000
0.000
0.044
0.145
20.002
0.030

20,209.528
22.083
1.026
0.400
0.085
0.079
0.335
0.281
0.990
0.482
0.789

0.488
0.624
0.090
0.286
0.202
0.071

Note: This table reports descriptive statistics of the final sample, which consists of 41,833 firm quarter observations for the period between 1995Q1 and 2009Q2. The mean, 1st quartile, median, 3rd quartile, and standard
deviation statistics are reported for each variable. MV is market value computed as price times shares outstanding
at fiscal quarter-end, PRC is share price at fiscal quarter- end, and BETA is the market model beta computed using
60 months of prior return data. All other variables are defined in Table 2.

and foreign companies. We also exclude firms with share prices less than US$1 at the end
of the previous quarter to avoid outliers from biasing the results.
Analyst recommendations ratings are obtained from IBES (ibes.recddet), and recommendations issued by anonymous analysts are eliminated.11 We identify a recommendation
revision as the action of a particular analyst to change his or her prior recommendation
rating. If the recommendation is revised to a more favorable one, we identify it as an
upgrade. If the recommendation is revised to a less favorable one, we identify it as
a downgrade.
We collect accounting data and earnings announcement dates from the Compustat quarterly files (compq.fundq). Security return data is obtained from CRSP files. For each firm,
we collect daily return data (crsp.dsf) adjusted for dividends, stock splits, and delisting
(using the delisting return provided in CRSP).
Analysts’ quarterly earnings forecasts are obtained from the IBES-unadjusted detail file
(ibes.detu).12 We adjust earnings estimates in the unadjusted detail file using CRSP’s
adjustment factors when necessary. To compute analyst expectations, we retain the last
quarterly earnings forecast made by each analyst before the earnings announcement date
and calculate the median of all earnings estimates.13
Finally, we obtain institutional ownership data from the Thomson-Reuters Institutional
Holdings (13f) database. We use the Wharton Research Data Services (WRDS) re-created
shares outstanding to compute institutional ownership and exclude observations where the

filing and reporting dates are not equal to avoid erroneous observations from entering the
sample.
Table 3 reports the descriptive statistics of the sample employed in this study. The final
sample consists of 41,833 firm quarters. Due to numerous data requirements (CRSP,

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


488

Journal of Accounting, Auditing & Finance 26(3)

Compustat, IBES, and TFN) imposed by the research design, the final sample is composed
of relatively large firms. The average firm in the sample has a market capitalization of
US$8.3 billion with a mean share price of US$30 and an average book-to-market ratio of
0.481.
Table 4 reports the correlation matrix of the variables used in the multiple regression
analysis. The four earnings surprise measures, SSUE, SASUE, CAR (-1, 1), and CAR (0, 1),
are positively correlated because they all intend to capture the earnings surprise component.
The REV variable, which represents the recommendation revisions during the 3-week
period before the earnings announcement period, is positively correlated with earnings surprise measures. This is consistent with recommendation revisions having predictive power
of subsequent earnings surprises. The interaction of REV and FD is also positively correlated with earnings surprise measures but at a weaker level, which suggests that the association between recommendations and earnings surprises declined after Regulation FD took
effect. The LANCRET variable, which is the previous quarter’s 3-day earnings announcement return, is positively correlated with the four earnings surprise measures consistent
with the postearnings announcement drift documented in Bernard and Thomas (1989).
Furthermore, as suggested by the momentum effect (Jegadeesh & Titman, 1993), LRET is
positively correlated with the earnings surprise measures. The accrual component (ACCR),
as shown in Sloan (1996), is negatively correlated with subsequent earnings surprises. The
CHNG_IO variable is positively correlated with subsequent earnings surprises with the
exception of CAR (0, 1). In short, the control variables employed in the multiple regression
analysis are correlated with the earnings surprise measures consistent with the results documented in the prior literature. Finally, we do not find a strong correlation among the independent variables that could suggest multicolinearity.14

Empirical Results
Univariate Analysis
The univariate analysis results suggest that recommendation revisions were useful in predicting upcoming earnings surprises during the pre-Regulation FD period. On average,
upgraded firms reported earnings above expectations and downgraded firms reported earnings below expectations. Table 5 Panel A reports that before Regulation FD took effect, the
earnings surprise differential between upgraded and downgraded firms was 0.0531 based
on time-series expectations (2nd column) and 0.0855 based on analyst expectations (3rd
column). The average earnings announcement return differential between upgraded and
downgraded firms was 0.93% based on 3-day returns (4th column) and 0.67% based on
2-day returns (5th column). These results suggest that investors who followed analysts’
advice in the pre-Regulation FD period were able to benefit from subsequent earnings surprises. Overall, the results suggest that some form of information acquisition or interpretation either through selective disclosure or effort was taking place. These results are
consistent with pre-Regulation FD concerns that analysts were receiving early peeks at
earnings results.
In the post-Regulation FD period, a weaker association between recommendation revisions and earnings surprises is evident. The mean earnings surprise differentials between
upgraded and downgraded firms based on time-series and analyst expectations decline to
0.0279 and 0.0399, respectively. Similarly, the mean 3- and 2-day earnings announcement
return differentials between upgraded and downgraded firms decline to 0.42% and 0.24%,

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


489

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012

1.000
0.285
0.113
0.104
0.060

0.013
0.037
0.053
0.021
0.120
0.152
20.025
0.043

1.000
0.245
0.239
0.092
0.044
0.056
0.067
0.032
0.094
0.108
20.037
0.050

SASUE

1.000
0.913
0.033
0.005
0.013
20.001

0.014
0.015
0.020
20.009
0.016

CAR (21,1)

1.000
0.025
0.009
0.005
20.003
0.009
0.007
0.011
0.005
0.016

CAR (0,1)

1.000
20.015
0.829
0.025
0.662
0.023
0.074
20.020
0.024

REV

1.000
20.076
0.594
20.044
20.003
20.065
20.088
0.043

FD

1.000
20.004
0.798
0.012
0.062
0.005
20.016

REV 3 FD

1.000
20.074
0.017
20.028
0.008
0.043

SOX

1.000
0.007
0.030
20.007
20.021

REV 3 SOX

1.000
0.312
0.003
0.113

LANCRET

1.000
20.013
0.215

LRET

1.000
20.012

ACCR

1.000

CHNG_IO

Note: Pearson correlations are reported for the four earnings surprise measures and the independent variables used in the multiple regression analysis. All variables are defined
in Table 2.

SSUE
SASUE
CAR (21,1)
CAR (0,1)
REV
FD
REV 3 FD
SOX
REV 3 SOX
LANCRET
LRET
ACCR
CHNG_IO

SSUE

Table 4. Correlation Table


490

Journal of Accounting, Auditing & Finance 26(3)

Table 5. Recommendation Revisions and Subsequent Earnings Surprises

Standardized

Market-adjusted earnings

Unexpected earnings

Announcement return

Time-series
expectations

Period

Analyst
expectations

CAAR (21, 11)

CAAR (0, 11)

0.0531
(5.71)***
0.0279
(4.89)***
20.0252
(22.32)**

0.0855
(13.79)***
0.0399

(8.21)***
20.0456
(25.78)***

0.93%
(8.67)***
0.42%
(2.94)***
20.51%
(22.84)***

0.67%
(6.95)***
0.24%
(1.89)*
20.43%
(22.64)***

0.0339
(4.35)***
0.0185
(3.09)***
20.0154
(21.57)*

0.0394
(9.08)***
0.0333
(8.28)***
20.0061

(21.04)

0.63%
(4.73)***
0.47%
(4.1)***
20.16%
(20.92)

0.30%
(3.46)***
0.29%
(2.63)***
20.01%
(20.07)

20.0192
(22.6)***
20.0094
(21.78)*
0.0098
(1.08)

20.0461
(27.9)***
20.0067
(21.38)
0.0394
(5.2)***

20.29%
(22.4)**
0.05%
(0.35)
0.34%
(1.78)*

20.37%
(23.93)***
0.05%
(0.37)
0.42%
(2.58)***

Panel A: Upgrades–Downgrades
Pre-Regulation FD
Post-Regulation FD
Pre- vs. post-Regulation FD
Panel B: Upgrades
Pre-Regulation FD
Post-Regulation FD
Pre- vs. post-Regulation FD
Panel C: Downgrades
Pre-Regulation FD
Post-Regulation FD
Pre- vs. post-Regulation FD

Note: This table reports the mean earnings surprise that follows analysts’ recommendation revisions made in the
3-week period prior to earnings announcements (223, 22). Earnings surprise is measured and reported based on
(a) the time-series earnings expectation, (b) analyst expectations, (c) the market-adjusted, 3-day earnings

announcement return centered on the report date, and (d) the market-adjusted 2-day earnings announcement
return beginning on the report date. Earnings announcement returns are winsorized at the bottom and upper onepercentile. Mean and standard error statistics of earnings surprises are computed using the Fama and MacBeth
(1973) procedure based on quarterly means. Panel A reports results for the earnings surprise differential between
upgrades and downgrades for the pre- and post-Regulation FD periods and the final row tests for the difference in
means, assuming unequal variances. Panels B and C report results separately for upgrades and downgrades,
respectively.
***, **, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

respectively. The decline in the strength of the relationship between recommendation revisions and earnings surprises is statistically significant. The final row of Panel A reports that
changes in time-series and analyst-based earnings surprise differentials are 20.0252 and
20.0456, respectively. Both changes are statistically and economically significant. Finally,
results based on earnings announcement returns also indicate a significant decline after
Regulation FD. Table 5 Panel A reports that the mean 3-day market-adjusted earnings
announcement returns declined 55% from 0.93% to 0.42% and the mean 2-day marketadjusted earnings announcement returns declined 64% from 0.67% to 0.24%.

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

491

Overall, the univariate results in Table 5 Panel A reveal that the association between
revisions and earnings surprises weakened substantially after the Regulation FD period.
These results are consistent with Regulation FD having significantly reduced analysts’
power in predicting earnings surprises. In the post-Regulation FD period, analyst recommendations appear to be less useful in guiding investors to firms that later experience earnings surprises.
Analysis of revisions separately by upgrades and downgrades yields similar results.
Table 5 Panels B and C report that the mean time-series-based earnings surprise declined
45% from 0.0339 to 0.0185 for upgraded firms and increased 51% from 20.0192 to
20.0094 for downgraded firms. Although neither change is statistically significant in its

own, the combination of the two changes is significant. Results based on analyst expectations and market-adjusted returns are similar and support the conclusion that recommendation revisions’ usefulness in identifying earnings surprises declined after Regulation FD.

Regression Analysis
The regression analyses examining the association between recommendation revisions and
earnings surprises controlling for confounding factors also reveal a significant decline in the
predictive value of recommendation revisions after Regulation FD took effect. Table 6
reports the estimation results of Equation 1, which investigates the impact of Regulation FD
on the association between recommendation revisions and time-series-based unexpected earnings. The coefficient of REV is positive and significant in Model I, suggesting recommendation revisions to have predictive value of upcoming earnings surprises. Consistent with the
downturn in the economy, the FD indicator variable, which takes a value of 1 for fiscal quarters after Regulation FD took effect, is negative. The hypothesis variable, REV 3 FD, is
20.013 and significantly negative, implying that the association between recommendation
revisions and earnings surprises declined significantly after Regulation FD took effect. The
SOX indicator variable is significantly positive, indicating an improvement in earnings surprises as markets recovered from the tech bubble burst starting in 2004. However, different
from the results based on the REV 3 FD variable, REV 3 SOX is estimated to be insignificant, suggesting that the information dynamics relating to analysts’ recommendation revisions
did not change significantly after the Sarbanes-Oxley Act took effect.
In Model II, we include the LANCRET and LRET variables to control for postearningsannouncement drift and momentum effects. The LANCRET variable is the earnings
announcement return in the previous quarter, and it is estimated to have a positive coefficient. This is consistent with the prior literature and suggests that firms that experienced
earnings surprises in the prior quarter continue to do so in the next quarter. The positive
coefficient on the LRET variable that controls for the past 3-month buy-and-hold return
suggests that firms that experienced superior market performance during the recent months
exhibited positive earnings surprises in the subsequent quarter. In Model II, REV, FD, and
REV 3 FD coefficients are estimated to be 0.022, 20.008, and 20.013, respectively. REV
and REV 3 FD are both statistically significant. The negative coefficient on REV 3 FD
indicates that the extent to which recommendation revisions predicted earnings surprises
declined after Regulation FD took effect. The REV 3 SOX variable is insignificant and
does not point to a significant impact on the association between recommendation revisions
and earnings surprises that can be attributed to the Sarbanes-Oxley Act.
In Model III, we control for accruals with the ACCR variable. The ACCR variable is statistically significant, and its coefficient is estimated to be 20.089. This is consistent with the

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012

492

Journal of Accounting, Auditing & Finance 26(3)

Table 6. Regression Analysis of Subsequent Earnings Surprises (Based on Time-Series Expectations)

REV
FD
REV 3 FD
SOX
REV 3 SOX

Model I

Model II

Model III

Model IV

0.026
(6.82)***
20.015
(21.96)**
20.013
(22.29)**
0.022
(2.89)***
0.001

(0.14)

0.022
(5.67)***
20.008
(21.08)
20.013
(22.34)**
0.021
(2.83)***
0.003
(0.50)
0.254
(7.59)***
0.171
(13.57)***

0.023
(5.61)***
20.015
(21.76)*
20.012
(22.06)**
0.046
(5.84)***
20.001
(20.12)
0.288
(8.33)***
0.152

(11.62)***
20.089
(21.70)*

20.001
(20.27)

20.016
(22.71)***

0.023
(5.62)***
20.015
(21.77)*
20.012
(22.04)*
0.046
(5.84)***
20.001
(20.12)
0.288
(8.29)***
0.151
(11.19)***
20.089
(21.70)*
0.017
(0.32)
20.016
(22.70)***

27,792
.036
.035

27,792
.036
.035

LANCRET
LRET
ACCR
CHNG_IO
Constant

0.008
(1.44)

Observations
R2
Adjusted R2

32,684
.004
.004

32,579
.032
.032

Note: This table reports the estimation results of the empirical model:
SASUEit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit
þ b8 ACCRit þ b9 CHNG IOit þ eit :
where SSUE is the standardized unexpected earnings decile based on a time-series earnings expectation model.
REV is the recommendation revision during the preearnings announcement period, FD is a post-Regulation FD indicator variable, and REV 3 FD is the interaction of REV and FD variables. SOX is a Sarbanes-Oxley indicator variable
and REV 3 SOX is the interaction of REV and SOX variables. LANCRET is the previous quarter’s 3-day earnings
announcement return (market adjusted), and LRET is the past 3-month buy-and-hold return ending a month before
the fiscal quarter-end. ACCR is total accruals scaled by average total assets as in Sloan (1996), and CHNG_IO is the
change in percentage ownership during the most recent calendar quarter. Four specifications of the above model
are estimated. t-statistics based on firm clustered standard errors are reported in parentheses.
***,**, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

prior literature and suggests that firms that had income increasing accruals during the previous fiscal year reported disappointing earnings results during the subsequent fiscal quarter.
Results relating to the impact of Regulation FD on the predictive value of recommendation
revisions are unchanged by the inclusion of the ACCR variable. The REV, FD, and REV 3
FD variables are all statistically significant, and their coefficients are estimated to be 0.023,
20.015, and 20.012, respectively. These results indicate a significant decline in the predictive value of recommendation revisions after the enactment of Regulation FD.
Finally, in Model IV, we incorporate the change in institutional ownership to our empirical model to control for the positive association between institutional trading activity and
earnings surprises. While the CHNG_IO variable is estimated to have a positive coefficient,

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

493

Table 7. Regression Analysis of Subsequent Earnings Surprises (Based on Analyst Expectations)

REV

FD
REV 3 FD
SOX
REV 3 SOX

Model I

Model II

Model III

Model IV

0.042
(15.51)***
0.003
(0.46)
20.017
(24.01)***
0.022
(3.91)***
20.008
(21.91)*

0.039
(14.36)***
0.007
(1.36)
20.018
(24.20)***

0.020
(3.60)***
20.006
(21.40)
0.167
(6.81)***
0.088
(10.64)***

0.039
(13.13)***
0.009
(1.46)
20.016
(23.44)***
0.031
(5.29)***
20.007
(21.58)
0.180
(7.13)***
0.080
(9.23)***
20.049
(22.65)***

20.002
(20.64)

20.007

(22.02)**

20.012
(23.13)***

0.038
(12.91)***
0.008
(1.36)
20.015
(23.27)***
0.031
(5.23)***
20.007
(21.54)
0.175
(6.97)***
0.074
(8.48)***
20.050
(22.78)***
0.111
(3.91)***
20.012
(23.06)***

36,895
.027
.027

31,841
.031
.031

31,841
.032
.031

LANCRET
LRET
ACCR
CHNG_IO
Constant
Observations
R2
Adjusted R2

37,323
.013
.013

Note: This table reports the estimation results of the empirical model:
SASUEit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit
þ b8 ACCRit þ b9 CHNG IOit þ eit ;
where SASUE is the unexpected earnings decile based on the consensus analyst expectation. REV is the recommendation revision during the preearnings announcement, FD is a post-Regulation FD indicator variable, and REV 3
FD is the interaction of REV and FD variables. SOX is a Sarbanes-Oxley indicator variable, and REV 3 SOX is the
interaction of REV and SOX variables. LANCRET is the previous quarter’s 3-day earnings announcement return
(market adjusted). LRET is the past 3-month buy-and-hold return ending a month before the fiscal quarter-end.
ACCR is total accruals scaled by average total assets as in Sloan (1996), and CHNG_IO is the change in percentage
ownership during the most recent calendar quarter. Four specifications of the above model are estimated.

t-statistics based on firm clustered standard errors are reported in parentheses.
***,**, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

it is not statistically significant. The REV, FD, and REV 3 FD variables confirm the inferences based on Models I to III. The predictive value of recommendation revisions appears
to be significantly lower after Regulation FD took effect.
We repeat the regression analysis using unexpected earnings based on analyst expectations and report the estimation results in Table 7. The REV variable is estimated to have
a coefficient of 0.042 and suggests a significantly positive association between recommendation revisions and earnings surprises. The FD variable is estimated to be 0.003, which is
insignificant, suggesting no change in the average level of earnings surprises after
Regulation FD. The interaction variable REV 3 FD is estimated as 20.017, which is

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


494

Journal of Accounting, Auditing & Finance 26(3)

statistically significant, implying a significant reduction in the association between recommendation revisions and earnings surprises. The SOX and REV 3 SOX variables reveal
a further decline in the predictive value of recommendation revisions after the SarbanesOxley Act.
In Model II, we control for postearnings-announcement drift and momentum effects and
find largely similar results. Both factors, LANCRET and LRET, have the expected signs.
The interaction of REV and FD is significantly negative, suggesting a decline in the association in the predictive value of recommendation revisions. Different from Model I, in
Model II we do not observe evidence suggesting a significant reduction in the predictive
value of recommendation revisions after the Sarbanes-Oxley Act took effect.
In Model III, we control for the accruals anomaly by incorporating the ACCR variable
into the empirical model. The ACCR variable is estimated to be significantly negative. The
REV 3 FD is negative, implying a significant decline in the predictive value of recommendation revisions. The REV 3 SOX variable is insignificant, suggesting no substantial
change in the predictive value of recommendation revisions after the Sarbanes-Oxley Act.
In Model IV, we control for changes in institutional ownership. The estimated coefficient on the CHNG_IO is positive, as expected. Firms with increased institutional ownership appear to experience positive earnings surprise. This is consistent with institutional
investors anticipating earnings surprises. The REV 3 FD coefficient indicates a significant

decline in the predictive value of recommendations after Regulation FD. We do not identify
any change in the association between recommendation revisions and earnings surprises
after the Sarbanes-Oxley Act.
Table 8 reports the estimation results of Equation (3), where 3-day market-adjusted earnings announcement returns are regressed on recommendation revisions and confounding
factors. The results echo the previous findings from the time-series and analyst-based earnings surprise measures. The REV 3 FD interaction variable is significantly negative, suggesting a substantial decline in the association between recommendation revisions and
subsequent earnings surprises. Finally, Table 9 reports the regression analysis of 2-day
market-adjusted earnings announcement returns. The results are consistent with prior results
and confirm the existence of a significant decline in the association between recommendation revisions and earnings surprises in the post-Regulation FD period.
Overall, the regression analysis results suggest a significant reduction in recommendation revisions’ ability to identify earnings surprises after Regulation FD took effect. The
results show that the level and change in the predictive value of recommendation revisions
cannot be explained by confounding factors that were also shown to exhibit predictive
power of earnings surprises. In conclusion, during the post-Regulation FD period, analysts
appear to be less successful in guiding their clients to firms that experience positive earnings surprises and warning their clients away from firms that experience negative earnings
surprises. These results are consistent with a reduction in selective disclosure, which was
a major source of information for analysts during the pre-Regulation FD period.

Trading Strategy Analysis
During the pre-Regulation FD period, the trading strategy that followed analysts’ advice to
capture earnings surprises appears to have earned significant abnormal returns controlling
for market risk, size, book-to-market, and momentum effects. Table 10 reports the performance of the hedge, upgrade, and downgrade portfolios. The results suggest that investors
who followed analysts’ recommendation revisions with the intent of capturing earnings

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

495

Table 8. Regression Analysis of Subsequent Earnings Surprise (Based on 3-Day Market-Adjusted

Returns)
Model I

Model II

Model III

Model IV

0.005
(6.29)***
20.000
(20.03)
20.003
(22.45)**
0.001
(0.81)
0.001
(0.57)

0.005
(6.20)***
0.000
(0.06)
20.003
(22.40)**
0.001
(0.75)
0.001
(0.53)

0.014
(1.41)
0.002
(0.64)

0.005
(5.92)***
20.000
(20.25)
20.004
(22.52)**
0.002
(0.89)
0.001
(0.64)
0.012
(1.09)
0.002
(0.63)
20.020
(23.68)***

Constant

0.002
(2.15)**

0.002
(2.02)**

0.001
(1.38)

0.005
(5.87)***
20.001
(20.30)
20.003
(22.46)**
0.002
(0.87)
0.001
(0.65)
0.011
(1.04)
0.001
(0.39)
20.020
(23.67)***
0.014
(1.34)
0.001
(1.42)

Observations
R2
Adjusted R2

41,820
.002

.001

41,324
.002
.002

35,382
.004
.004

35,382
.004
.004

REV
FD
REV 3 FD
SOX
REV 3 SOX
LANCRET
LRET
ACCR
CHNG_IO

Note: This table reports the estimation results of the empirical model:
CARðÀ1; þ1Þit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit
þ b8 ACCRit þ b9 CHNG IOit þ eit ;
where CAR(21,1) is the 3-day market-adjusted earnings announcement return. REV is the recommendation revision during the preearnings announcement period, FD is a post-Regulation FD-indicator variable, and REV 3 FD is
the interaction of REV and FD variables. SOX is a Sarbanes-Oxley indicator variable, and REV 3 SOX is the interaction of REV and SOX variables. LANCRET is the previous quarter’s 3-day earnings announcement return (market
adjusted), and LRET is the past 3-month buy-and-hold return ending a month before the fiscal quarter-end. ACCR is

total accruals scaled by average total assets as in Sloan (1996), and CHNG_IO is the change in percentage ownership during the most recent calendar quarter. Four specifications of the above model are estimated. t-statistics
based on firm clustered standard errors are reported in parentheses.
***, **, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

surprises earned an average monthly abnormal return of 4.6% before transaction costs
during the pre-Regulation FD period. The average abnormal returns associated with
upgrade and downgrade portfolios are 1.9 and 22.7%, respectively. Both abnormal return
estimates are statistically significant and consistent with investor complaints relating to certain select investors receiving privileged access to management via financial analysts.
During the post-Regulation FD, a considerable reduction in the performance of the portfolios is evident. The performance of the hedge portfolio declines 70% from 4.6% to 1.4%.
Although the hedge portfolio’s abnormal performance is statistically significant in the postRegulation FD period, it is substantially lower. Similarly, we see a 21% reduction in the

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


496

Journal of Accounting, Auditing & Finance 26(3)

Table 9. Regression Analysis of Subsequent Earnings Surprise (Based on 2-Day Market-Adjusted
Returns)

REV
FD
REV 3 FD
SOX
REV 3 SOX

Model I

Model II

Model III

Model IV

0.003
(4.97)***
0.002
(1.13)
20.003
(22.32)**
0.001
(0.44)
0.000
(0.44)

0.003
(4.96)***
0.002
(1.24)
20.003
(22.33)**
0.001
(0.45)
0.001
(0.46)
0.001
(0.15)
0.001
(0.52)

0.004
(4.79)***
0.002
(1.15)
20.003
(22.51)**
0.001
(0.49)
0.001
(0.83)
0.000
(0.02)
0.001
(0.34)
20.004
(20.83)

20.000
(20.39)

20.000
(20.49)

20.000
(20.35)

0.004
(4.73)***
0.002

(1.10)
20.003
(22.44)**
0.001
(0.46)
0.001
(0.85)
20.000
(20.04)
0.000
(0.10)
20.004
(20.86)
0.013
(1.41)
20.000
(20.31)

LANCRET
LRET
ACCR
CHNG_IO
Constant
Observations
R2
Adjusted R2

41,820
.001
.001

41,324
.001
.001

35,382
.001
.001

35,382
.001
.001

Note: This table reports the estimation results of the empirical model:
CARð0; þ1Þit ¼ a þ b1 REVit þ b2 FDit þ b3 REVit 3 FDit þ b4 SOXit þ b5 REVit 3 SOXit þ b6 LANCRETit þ b7 LRETit
þ b8 ACCRit þ b9 CHNG IOit þ eit ;
where CAR(0, 11) is the 2-day market-adjusted earnings announcement return beginning on the earnings
announcement date. REV is the recommendation revision during the preearnings announcement period, FD is
a post-Regulation FD-indicator variable, and REV 3 FD is the interaction of REV and FD variables. SOX is
a Sarbanes-Oxley indicator variable, and REV 3 SOX is the interaction of REV and SOX variables. LANCRET is the
previous quarter’s 3-day earnings announcement return (market adjusted). LRET is the past 3-month buy-and-hold
return ending a month before the fiscal quarter-end. ACCR is total accruals scaled by average total assets as in
Sloan (1996), and CHNG_IO is the change in percentage ownership during the most recent calendar quarter. Four
specifications of the above model are estimated. t-statistics based on firm-clustered standard errors are reported
in parentheses.
***,**, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

performance of the upgrade portfolio, decreasing from 1.9% to 1.5%. Finally, the most dramatic reduction is evident in the performance of the downgrade portfolio. During the postRegulation FD period, abnormal returns associated with downgrades vanish.
Overall, our results are consistent with analysts’ recommendation revisions conveying
less information about upcoming earnings surprises in the post-Regulation FD. The

reduction in the private information transmitted via recommendation revisions appears to
manifest itself as a reduction in the association between recommendation revisions and
a reduction in the performance of the portfolios constructed surrounding recommendation
revisions and earnings announcements. The results support the conclusion that analysts’

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


Palmon and Yezegel

497

Table 10. Portfolio Performance
Portfolio

Period

Intercept

b

SMB

HML

UMD

R2

0.046

20.121
20.321 20.331
20.267 6.0%
(4.4)*** (20.42)
(21.23) (20.91)
(21.51)
Hedge
Post-Regulation FD 0.014
20.433
0.583 20.268
20.086 9.1%
(1.95)* (22.44)**
(2.15)** (21.28)
(20.67)
Upgrade Pre-Regulation FD
0.019
1.501
20.072 20.365
20.249 55.4%
(2.05)**
(5.78)*** (20.31) (21.12)
(21.57)
Upgrade Post-Regulation FD 0.015
1.256
0.410 20.487
20.074 64.7%
(2.74)*** (9.13)*** (1.95)* (23.01)*** (20.75)
Downgrade Pre-Regulation FD 20.027
1.622
0.249 20.034

0.018 47.4%
(22.47)*** (5.43)*** (0.92) (20.09)
(0.1)
Downgrade Post-Regulation FD 0.001
1.689
20.174 20.219
0.012 64.3%
(0.21)
(10.79)*** (20.73) (21.19)
(0.1)

Hedge

Pre-Regulation FD

Adjusted
R2
0.8%
5.6%
52.9%
63.4%
44.4%
62.9%

Note: This table reports the performance of the portfolio that purchases shares of firms that were upgraded and
sells shares of firms that were downgraded after the fiscal quarter-end and liquidates positions at the end of the
1st day after the earnings announcement day. The portfolio performance and characteristics reported are based on
a time-series regression of each portfolio’s excess monthly return on the four-factor returns: excess market return
(b—4th column), a zero-investment size portfolio (SMB—5th column), a zero-investment book-to-market portfolio (HML—6th column), and a zero-investment momentum portfolio (UMD—7th column). The intercept of this
equation, which is an estimate of the abnormal return, is reported for each portfolio in the third column. The R2

and adjusted R2 values of the time-series regressions are reported in the last two columns. t-statistics are reported
in parentheses.
***, **, and * denote significance at the 1%, 5%, and 10% significance levels, respectively.

recommendation revisions are less useful in helping analysts’ clients earn abnormal
profits.

Sensitivity Analysis
In addition to confounding factors, analysts may also be revising their recommendations
in response to key corporate developments. To the extent that the impact of corporate
developments is less during the post-Regulation FD period, there may be a change in the
association between revisions and earnings surprises that is due to the impact of corporate
developments, rather than due to the information that analysts convey through their recommendation revisions. To control for this possibility, we use a hand-collected public
information arrival database and exclude revisions that took place within a 3-trading-day
window surrounding dates of Wall Street Journal news articles during the period 19952006 and find that the results are largely unaffected when we add this additional
control.15

Conclusion
This article examines the extent to which analyst recommendations are useful in identifying
firms that experience earnings surprises during the pre- and post-Regulation FD periods.
The empirical analysis suggests a significant decline in the predictive value of recommendation revisions after Regulation FD took effect. Prior to Regulation FD, upgraded firms

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012


498

Journal of Accounting, Auditing & Finance 26(3)

outperformed downgraded firms by 0.93% on earnings announcements. After Regulation

FD, the earnings announcement return differential between upgraded and downgraded
firms declined by 55% to 0.42%. In addition, we find that the abnormal returns associated
with following analysts’ recommendation revisions made shortly in advance of earnings
announcements declined considerably following the enactment of Regulation FD. The portfolio analysis reveals that the trading strategy yielded an average monthly abnormal return
of 4.6% during the pre-Regulation FD period compared with 1.4% during the postRegulation FD period.
The passage of Regulation FD was preceded with intense objection and frustration by
investors that certain analysts and institutional investors were enjoying privileged access to
material nonpublic information. Disclosure of earnings-related information was one of the
most scrutinized practices of selective disclosure. Investors complained that managers were
giving private information to analysts about upcoming earnings results and analysts
were then advising their clients accordingly. The empirical results provided in this study
suggest a considerable decline in analysts’ ability to guide their clients to and away from
firms that are likely to experience earnings surprises. These results are consistent with
a decline in selective disclosure and suggest that Regulation FD was effective in achieving
its objective.
Authors’ Note
We are grateful to the editor Kashi R. Balachandran, Jean Bedard, Yoel Beniluz, Mahendra R.
Gujarathi, Rani Hoitash, James Hunton, Bikki Jaggi, Jay Thibodeau, an anonymous reviewer, and
seminar participants at Bentley University and Rutgers University for their valuable comments and
suggestions. We thank Tesfalidet Tukue for excellent research assistance. We also thank the
Whitcomb Center for Research in Financial Services for providing research support through use of
the WRDS system. All errors are the authors’ responsibility.

Declaration of Conflicting Interests
The author(s) declared that they had no conflicts of interests with respect to their authorship or the
publication of this article.

Funding
The author(s) received no financial support for the research and/or authorship of this article.

Notes
1. Selective Disclosure and Insider Trading, Release No. 33-7881, August 15, 2000.
2. Forecasts also correspond to an important aspect of analysts’ reports; however, they do not pro3.
4.
5.
6.

vide investors direct advice regarding buy or sell decisions. In contrast, recommendation revisions are a direct way in which analysts advise their clients to buy or sell shares.
Francis and Soffer (1997) find that stock recommendation revisions contain information incremental to the information in earnings forecast revisions.
Our interviews with analysts and cases of Regulation FD violations (e.g., />litigation/admin/34-48461.htm and confirm
this conclusion.
The results remain qualitatively similar when we use a 2-week or 1-month period.
In early 2002, the National Association of Securities Dealers (NASD) proposed Rule 2711 and
the New York Stock Exchange (NYSE) proposed a modification to its Rule 472,

Downloaded from jaf.sagepub.com at Taylor's University on December 2, 2012