Do Analysts Say Anything About Earnings Without Revising Their Earnings Forecasts?
Philip G. Berger
Booth School of Business
University of Chicago
[email protected]
Charles G. Ham
John M. Olin School of Business
Washington University in St. Louis
[email protected]
Zachary R. Kaplan
John M. Olin School of Business
Washington University in St. Louis
[email protected]
February 2017
* We appreciate helpful comments from, and discussions with, Rich Frankel, Jared Jennings, Kevin
Koharki, Alastair Lawrence (discussant), Xiumin Martin, Volkan Muslu (discussant), Doug Skinner,
Sorabh Tomar and workshop participants at Carnegie Mellon, Washington University in St. Louis, the
2016 AAA Annual Meeting, the 2016 EAA Annual Congress, and the 2016 Nicholas Dopuch Accounting
Conference, as well as research assistance from Aadhaar Verma and Yuqing Zhou. Any remaining errors
or omissions are ours. Berger acknowledges financial support from the University of Chicago’s Booth
School of Business.
Do Analysts Say Anything About Earnings Without Revising Their Earnings Forecasts?
Abstract
We identify a novel bias in analyst forecasts, after-revision bias, which we identify by examining
an analyst’s reports after his final earnings forecast of the quarter. We document that (i)
qualitative predictions from the text of reports, (ii) share price target revisions, and (iii) revisions
to next quarter’s earnings forecast predict error in the current quarter’s earnings forecast (CQE).
Consistent with analysts maintaining a beatable benchmark for managers, we find analysts are
more likely to disseminate positive (negative) news by revising a forecast other than the CQE
(revising the CQE). Consistent with analysts minimizing deviations from the consensus, we
demonstrate analysts revise the CQE (revise another forecast) more frequently when a revision to
the CQE would move their forecast towards (away from) the consensus. Market returns are slow
to impound the information in qualitative predictions and share price target revisions, as both
predict earnings announcement window returns. Our results demonstrate that the value of the
current quarter’s earnings forecast to managers and investors distorts the flow of information into
the forecast.
Keywords: sell-side analysts, analyst incentives, earnings forecasts, forecast bias.
1. Introduction
Analysts’ clients value numerous attributes of their research product, with recent research
suggesting the most valuable attribute is qualitative insights and earnings forecasts are among the
least valuable attributes.1 We examine whether analysts disseminate earnings information
without revising their current quarter earnings forecasts to better understand whether the low
value clients assign to forecast accuracy distorts the flow of information into forecasts. Analysts
may incorporate all information into the earnings forecast if doing so enhances analysts’ ability
to communicate information to clients. Alternatively, omitting information from the earnings
forecast may enable analysts to better service key stakeholders if the associated benefits
outweigh the relatively low cost of forecast inaccuracy. For instance, if revisions impose
processing costs on clients and/or lead the analyst to articulate an inconsistent message, analysts
may limit the frequency with which they revise their forecasts (Bernhardt et al. 2016). Similarly,
incentives to maintain relationships with, and access to, company management (Lin and
McNichols 1998; Lim 2001; Ke and Yu 2006) can affect the decision to issue a forecast,
particularly one the firm may not beat. Analysts may compensate for the costs of
communicating earnings information through the current quarter earnings forecast by instead
disseminating information through other channels, at least partially explaining why analysts issue
a preponderance of forecasts early in the quarter (Ivkovic and Jegadeesh 2004).
We test whether analysts omit information from their earnings forecast using information
analysts disseminate after the final forecast of the quarter. We capture analysts’ outputs released
after the final forecast using two approaches. First, we capture quantitative information using the
sign of revisions to the share price target (SPT) and future quarter earnings (FQE). Second, we
use a textual approach to capture qualitative predictions about earnings. Specifically, we define
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1
Bagnoli et al. 2008; Groysberg et al. 2011; Brown et al. 2015.
1
qualitative predictions as positive (negative) if the analyst uses a synonym for “beat” (“miss”)
and a synonym for “earnings expectations” within the same sentence.2 Collectively, we refer to
the qualitative predictions, SPT revisions, and FQE revisions as “non-earnings forecast signals.”3
We document that all non-earnings forecast signals strongly predict the analyst’s own
earnings surprise. Specifically, a qualitative prediction of miss (beat) increases the probability of
the firm missing (meeting or beating) the earnings estimate by 5.5% (6.1%). When the analyst
revises a share price target negatively (positively), the probability of the firm missing (meeting
or beating) the earnings estimate increases 1.7% (4.3%). The FQE revisions have a lower,
though still statistically significant, association with the probability of the firm missing or
beating the analyst’s prior forecast.4
The preceding results imply that analysts do not always issue a revised earnings forecast
when they have information with earnings implications (“after-revision bias”). The evidence
from our textual analysis dismisses concerns that the issuance of non-earnings forecast signals is
unintentional, because the analyst would have to not understand the meaning of his own words.
Prior studies, which identify bias at the time of the forecast, have had difficulty distinguishing
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2
We examine the text of a subset of analyst reports published after the final forecast of the quarter and find
numerous examples of analysts publishing explicit predictions that the company will beat or miss the analyst’s own
forecast. Predicting performance relative to a previous forecast rather than revising the forecast implies the analyst
does not always issue a revised earnings forecast when he has significant earnings information. For example, one
week before the earnings announcement, a Wedbush Securities analyst published a report with the following phrase:
“We expect Garmin to report significant upside to our expectations for revenue, margins, and EPS given the
tremendous growth in the PND market and strong results from its closest rival, TomTom.” We validate our measure
by hand-coding a subset of the sentences identified by our algorithm as containing qualitative predictions to
ascertain whether the analyst report includes an explicit qualitative prediction that the firm will beat or miss earnings
expectations. We document a significantly positive correlation between our algorithmic coding scheme and the
hand-coded sentences, suggesting the text algorithm captures qualitative predictions, albeit with noise.
3
An advantage of the SPT and FQE setting is that revisions to these forecasts perfectly capture the intentions of the
analyst, whereas the textual analysis will capture those intentions with error. An advantage of the textual setting is
that the textual comments relate explicitly to the current quarter’s earnings forecast, while the SPT and FQE
revisions could relate to other information.
4
Our main specification compares the firm’s actual earnings to the analyst’s own forecast. We also document that
non-earnings forecast signals are associated with the earnings surprise if we compare the firm’s actual earnings to
the consensus forecast.
2
between intentional and unintentional bias (Francis 1997), because both intentional (incentivebased)
and
unintentional
(behavioral-based)
theories
can
predict
under-reaction
to
contemporaneously released information.
Next, we investigate cross-sectional variation in after-revision bias to shed light on why
analysts issue non-earnings forecast signals.5 First, we examine whether analysts revise SPT and
FQE forecasts for good news more or less frequently than for bad news. Incorporating good
(bad) news into an SPT or FQE forecast, rather than the current quarter earnings (CQE) forecast,
increases the likelihood that actual earnings meet or beat (miss) the earnings forecast. Because
managers prefer to meet or beat forecasts (Richardson et al. 2004), and managers reward analysts
who publish favorable forecasts by granting greater access to management (Lim 2001; Ke and
Yu 2006; Brown et al. 2015), analysts have incentives to omit positive news from (incorporate
negative news into) CQE forecasts. We find analysts revise SPT and FQE (CQE) forecasts more
frequently for good (bad) news than bad (good) news. We also find positive SPT and FQE
revisions provide more information content about the earnings surprise relative to negative SPT
and FQE revisions. Our findings contribute to understanding the mechanism analysts use to
walk-down forecasts – they incorporate negative news into the CQE forecast and omit positive
news from the CQE forecast, yet communicate positive news through other channels.
We also document that analysts with a “buy” recommendation walk-down their forecasts
more aggressively, suggesting a positive disposition toward a company leads the analyst to
incorporate more negative news into the CQE forecast. Consistent with this, we demonstrate that
analysts with a buy recommendation (i) are more likely to issue negative CQE forecast revisions,
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
5
In our setting, after the initial forecast of the quarter, the analyst has three options: (i) issue a non-earnings forecast
signal, (ii) revise the current quarter earnings forecast, or (iii) issue no forecasts. We exploit scenarios (ii) and (iii)
as counter-factuals to identify the determinants and consequences of after-revision bias. We exclude the qualitative
predictions from these analyses because the textual analysis is conducted on a small subset of the analyst reports,
which limits the amount of cross-sectional variation we can exploit.
3
and (ii) are more likely to issue positive SPT forecast revisions. Thus, analysts with a buy
recommendation distort the flow of information into the CQE forecasts in a way that increases
the probability of the firm beating these forecasts.
Second, we examine whether analysts respond to earnings information with a
non-earnings forecast signal more frequently when the information moves the analyst forecast
toward the consensus (herding) or away from the consensus (bold). Investors judge analyst
forecasts relative to those of competing analysts. Prior research documents that analysts who
issue herding forecasts are less likely to be fired, suggesting career consequences from issuing
extreme forecasts (Hong et al. 2000; Clement and Tse 2005). We find that (i) analysts are more
(less) likely to issue a non-earnings forecast signal for bold (herding) forecasts, and (ii) bold
(herding) non-earnings forecast signals provide more information content about the earnings
surprise. Our result that analysts suppress bold forecasts suggests that measures of forecast
dispersion may understate the amount of disagreement, because analysts suppress their private
information when it moves them away from the consensus. We provide further validation for
this assertion by demonstrating that analyst disagreement decreases when more analysts issue
non-earnings forecast signals.
Third, we examine how the magnitude of earnings information after the initial CQE
forecast affects the decision to issue a non-earnings forecast signal.
When there is little
information not already in the analyst’s forecast, frictions may limit the analyst’s incentive to
revise the CQE forecast (Bernhardt et al. 2016). Conversely, incentives for accuracy may
encourage the analyst to revise the CQE forecast when it omits substantial information (Clement
and Tse 2005; Groysberg et al. 2011). Consistent with this prediction, we find that the absolute
4
value of initial forecast error is lower when the analyst revises the SPT or FQE forecast relative
to when the analyst revises the CQE forecast.
Finally, we study the consequences of non-earnings forecast signals by examining the
association between the issuance of non-earnings forecast signals and earnings announcement
returns. If the market does not fully impound the earnings implications of non-earnings forecast
signals at the time of their issuance, these signals may be associated with returns at the earnings
announcement date. Consistent with this, we document that when the analyst issues a qualitative
prediction of miss (beat), average earnings announcement returns are 0.7% lower (0.5% higher)
than when the analyst does not. The large predictable returns at earnings announcements suggest
that the information in qualitative predictions is not fully impounded into returns prior to the
earnings announcement. We document less (no) return predictability for SPT (FQE) forecast
revisions.
Our results provide several contributions. First, we demonstrate both the importance of
CQE forecasts and the related danger of interpreting them as a proxy for market expectations.
Finding that analysts hesitate to incorporate changes in their own CQE expectations into their
CQE forecasts is consistent with analysts viewing CQE forecasts as prominent benchmarks, and
with modifications to the benchmark imposing costs on managers or the analyst’s clients. Our
findings suggest a danger of interpreting forecasts as a proxy for investor expectations, and
suggest an asymmetry in the quality of those expectations based on recent news. Finally,
researchers hoping to infer the role of analysts by examining the timing of their forecast
revisions may ignore the timing of qualitative analysis, which has substantial value to investors
(Frankel et al. 2006; Chen et al. 2010).
5
Second, we contribute to the literature on forecast bias by identifying a novel source of
error, for which we have increased power to distinguish between intentional and unintentional
bias.6 The ability to distinguish between intentional and unintentional bias is challenging at the
time of the forecast because both intentional (incentive-based) and unintentional (behavioralbased) theories can predict under-reaction to contemporaneously released information. In our
textual analysis, when analysts make bullish (bearish) statements about future earnings, the firm
tends to beat (miss) the unrevised forecast. Under unintentional explanations, the analyst would
have to not understand the meaning of his own words.
Our results are subject to a significant caveat. While we document that incentives drive
cross-sectional variation in after-revision bias, this does not fully explain the bias. For example,
analysts use non-earnings forecast signals more frequently for positive news which is consistent
with incentives to walk-down forecasts, but analysts also use non-earnings forecast signals for
negative news. One potential explanation is that analysts use forecasts not only as an expectation
of future earnings, but also as a benchmark to judge subsequent performance against. To
facilitate the forecast’s role as a benchmark, the analyst may omit some news from the earnings
forecast. For example, if negative news materializes during the quarter and the analyst fully
incorporates the information into the forecast, the firm may beat the updated forecast even if
reported earnings fall below a fair ex-ante benchmark. Omitting the information from the
earnings forecast, but incorporating it into other outputs, allows the analyst to judge the firm
relative to a static benchmark yet still provide timely information to clients.
Less timely
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
6
Prior studies document predictable errors in analyst forecasts, including: (i) forecasts under-react to past earnings
information (Abarbanell and Bernard 1992), (ii) forecasts under-react to returns (Lys and Sohn 1990), (iii) forecast
revisions have a positive correlation with subsequent forecast errors (Shane and Brous 2001), (iv) forecasts underincorporate information from other analyst forecasts (Stickel 1992), and (v) forecasts under-incorporate information
from firm characteristics (Hughes et al. 2008). We elaborate further on the state of the literature as to the cause of
the biases in section 2.
6
updating of the forecast may also allow the analyst to disseminate a more consistent signal (i.e.,
avoiding a downward revision followed by a positive earnings surprise), which evidence from
social psychology suggests facilitates persuasion.
2. Literature Review and Hypothesis Development
First, we review the literature on the consequences of meeting or beating analyst
estimates as well as evidence on biases in these estimates. Second, we review the prior literature
on the incentives of analysts, with a particular focus on their incentives to issue accurate
forecasts and beatable forecasts. We conclude by motivating our research question, which asks
whether analysts’ outputs after the final current quarter earnings forecast predict error in the
forecast.
2.1 Properties of Analyst Forecasts
2.1.1 Determinants and Consequences of Meeting or Beating
A lengthy stream of literature evaluates whether firms manage earnings to exceed
analysts’ forecasts of earnings, and the consequences of meeting or beating these expectations.
DeGeorge et al. (1999) provide evidence that a higher fraction of firms just meet or beat analyst
forecasts than just miss the threshold, suggesting managers take actions to reach this benchmark.
Evidence documenting analysts’ walk-down forecasts during the quarter suggests expectations
management may drive a portion of the propensity for firms to meet or beat analyst forecasts
(Kang et al. 1994; Matsumoto 2002; Cotter et al. 2006). Other studies document firms may use
real earnings management to meet or beat estimates (Roychowdhury 2006; Bhojraj et al. 2009).
Corroborating this empirical evidence, Graham et al. (2005) survey managers and find (i) a
majority of managers would sacrifice value increasing projects to avoid missing expectations,
and (ii) analyst expectations are the most important financial reporting threshold to exceed.
7
Studies document that firms face negative consequences upon failing to meet or beat
earnings forecasts. Firms that miss estimates have negative share price performance (Skinner
and Sloan 2002; Bartov et al. 2002; Bhojraj et al. 2009), suggesting capital market consequences
upon missing expectations. In addition, managers suffer career penalties for failing to meet or
beat earnings forecasts (Matsunaga and Park 2001; Farrell and Whidbee 2003). While most
studies investigate the importance of meeting the consensus forecast, recent research documents
that the percentage of forecasts met has better predictive power for earnings announcement
returns, suggesting that investors evaluate firm performance relative to multiple benchmarks
(Kirk et al. 2014).
While the prior literature has established that firms have incentives to meet or beat and
earnings expectations decline as the earnings announcement approaches, one unaddressed
question is how analysts walk-down forecasts.
Our research design will address whether
analysts simply adjust their forecasts downward, or whether analysts respond differentially based
on the sign of the news (e.g., by omitting positive news from forecasts while revising the forecast
in response to negative news).
2.1.2 Predictable Errors in Earnings Forecasts
Numerous studies document information available to analysts at the time they issue
earnings forecasts predicts forecast error. Past earnings surprises (Abarbanell and Bernard 1992)
and past returns (Abarbanell 1991) have a positive correlation with forecast error.
Firm
characteristics such as book-to-market ratios, dividend payout ratios (Hughes et al. 2008; So
2013), accruals (Bradshaw et al. 2001), and book-tax differences (Weber 2009) also predict
forecast errors. Individual forecasts under-weight information from the consensus (Chen and
Jiang 2006) and the sign of the analyst’s revision to the current quarter’s earnings forecast has a
8
positive correlation with the error in the revised forecast (Shane and Brous 2001; Raedy et al.
2006).
Although short-horizon forecasts contain significant biases, they still provide more
accurate forecasts of future earnings than time-series models (Brown et al. 1987; Bradshaw et al.
2012). In contrast, long-horizon forecasts provide less accurate forecasts than time-series models
(Bradshaw et al. 2012) and contain significant bias (DeBondt and Thaler 1990).
2.1.3 Share Price Target Forecasts
Prior studies on share price targets have documented: (i) share price target revisions have
large announcement window returns (Brav and Lehavy 2003), (ii) forecasts of share price
appreciation contain some information about future returns, but also contain significant bias
(Bradshaw et al. 2013), and (iii) analysts do not exhibit persistent ability to forecast share price
appreciation (Bradshaw et al. 2013). No prior study that we are aware of has suggested the share
price target may have information content for future earnings beyond that in the current quarter’s
earnings forecast.
2.2 Analyst Incentives and Explanations for Forecast Bias
Explanations for analyst forecast bias fall under two broad categories: (i) intentional bias:
analysts have incentives to issue forecasts that satisfy key stakeholders such as managers or
clients, and (ii) unintentional bias: analysts make errors transforming information into forecasts
of earnings. In the ensuing discussion we focus on intentional bias, while acknowledging
numerous studies propose unintentional explanations (DeBondt and Thaler 1990; Friesen and
Weller 2006; Hilary and Menzly 2006).
Under intentional explanations for forecast bias, analysts face a trade-off between their
incentives to issue accurate forecasts and the benefit to biasing forecasts to please key
stakeholders (Lim 2001). Research demonstrates forecast accuracy is not valued as highly as
9
several other analyst forecast attributes.
Groysberg et al. (2011) find no evidence analyst
compensation relates to forecast accuracy after conditioning on institutional investor votes.
Bagnoli et al. (2008) investigate the attributes investors value most highly using survey evidence
and find that investors rank “written reports” and “industry insight” highest, above the
importance of earnings forecasts.
Groysberg et al. (2011) also find analyst compensation
increases with the analyst’s rating on investor surveys, suggesting analysts have incentives to
provide information institutional investors demand. Survey evidence confirms analysts believe
their primary objective is to provide information their clients perceive to be valuable (Brown et
al. 2015).
Research demonstrates a number of channels through which analysts can increase the
value of their information by potentially biasing forecasts.
First, analysts value access to
management (Soltes 2014; Brown et al. 2015), suggesting analysts might be willing to bias
forecasts to procure access (Lim 2001; Ke and Yu 2006).
Confirming the credibility of
managerial threats to withhold access, Soltes (2014) documents that managers refuse to interact
with analysts who issue negative recommendations. Brown et al. (2015) also provide survey
evidence suggesting analysts bias their earnings forecasts to obtain access to management.
Second, if forecast accuracy is evaluated relative to the accuracy of other analysts, this
creates incentives for analysts to omit information from forecasts in certain circumstances. Hong
et al. (2000) demonstrate that inexperienced analysts are more likely to be fired for inaccurate
forecasts and bold forecasts, whereas being bold and accurate does not improve the analysts’
career outcomes.
They argue this creates incentives for analysts to forecast strategically.
Clement and Tse (2005) further demonstrate that herding forecasts, or forecast revisions that
10
move the forecast toward the consensus, contain less new information than bold forecasts, which
they interpret as further evidence of strategic forecasting.
Third, the analyst could bear some cost when revising the forecast (i.e., frictions).
Empirical evidence that analysts revise their forecast infrequently, a median of once per quarter,
is consistent with frictions impeding the flow of information into forecasts. Frictions can impact
forecast frequency through several channels, including: (i) the analyst incurs costs from adjusting
the model and/or explaining the changes in the model to clients.7 Evidence that analysts make
mistakes preparing their models suggests the risk of error is non-negligible (Green et al. 2016).
(ii) The analyst endogenizes the processing cost his revisions impose on clients and only issues
revisions from which a client would obtain a benefit that exceeds the client’s processing cost.8
For the frictions explanation to be plausible in our setting, in at least some instances frictions
must prevent a revision to the CQE forecast while not preventing either the publication of a
report or the revision of another forecast.
Fourth, when analysts publish forecasts, they are broadly disseminated through I/B/E/S,
Bloomberg, and other financial networks. Thus, updating information in earnings forecasts
potentially allows non-clients to capture some of the informational rents created by an analyst’s
research. Inserting information into the text of the report could limit the number of non-clients
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
7
Costs also arise because analysts receive questions about the materials they disseminate. By issuing a report the
analyst alerts clients to his report, and these clients might follow up with questions or demands for additional
analysis.
8
When an analyst revises a forecast the reports and/or notes are circulated to clients. If the information content of
the reports is low, the client may feel the cost incurred from processing the information was not worth the benefit
from obtaining the information in the report. The client may then penalize the analyst on his survey evaluations for
wasting his time. Investors can condition their decision to process an analyst report on (i) the brokerage of the
analyst, (ii) the company covered, and/or (iii) the timing of the report. Investors will often find it difficult to decide
whether to process information based on the content of the information, because obtaining the content involves
incurring costs processing the information (Sims 2003).
11
consuming the information.9 This should increase the benefit clients obtain from an analyst’s
research by limiting the number of informed investors (Grossman and Stiglitz 1980).
2.3 Research Question
Ultimately, analysts must provide valuable information to their clients.
Empirical
evidence documents that the final revision to the current quarter earnings forecast often occurs
near the prior quarter earnings announcement date (Ivokovic and Jegadeesh 2004). If the analyst
disseminates earnings information only through revisions to the current quarter earnings forecast,
this implies that in many firm-quarters the analyst goes three months without providing earnings
information to clients.
Not disseminating information over such a long time frame likely
decreases the value of his research to clients (Groysberg et al. 2011). Alternatively, the analyst
may disseminate earnings information via other channels. In other words, the analyst may
respond to the incentives to omit information from the CQE forecast, yet incorporate the
information into other forecasts. Thus, we study whether the aforementioned non-earnings
forecast signals have information content about earnings.
RQ: Do analysts’ non-earnings forecast signals after the final current quarter earnings forecast
provide information content about the earnings surprise? If so, why do analysts issue nonearnings forecast signals after the final current quarter earnings forecast?
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
9
Prior research investigates “tipping” by studying whether institutional investors trade on analyst recommendations
before they have been publicly disseminated (Irvine et al. 2007). Although providing information to clients before
publishing it offers the clients a significant information advantage, it exposes the analyst to potential discipline from
his brokerage, many of which proscribe the behavior. This form of selective disclosure is also against FinRa
regulations.
12
3. Sample Selection and Descriptive Statistics
3.1 Sample Selection: Revision Sample
The sample begins with all analyst-firm-quarters on the I/B/E/S unadjusted detail file
over the period 1999Q1 to 2015Q2. The start date is the first year share price targets are
available on I/B/E/S and the end date is the last quarter with data availability (when
downloaded). We keep analyst-firm-quarters for which (i) the analyst issued a quarterly earnings
forecast in the current and prior quarters, (ii) the firm’s actual earnings figure and earnings
announcement date are available for the current and prior quarters, (iii) the firm’s share price,
outstanding shares, and the cumulative factor to adjust price are available on CRSP for the prior
quarter’s earnings announcement, and (iv) the cumulative factor to adjust price is available on
CRSP for the current quarter. This includes 2,293,778 analyst-firm-quarters.
We drop analyst-firm-quarters (i) without a current quarter earnings forecast after the
prior quarter earnings announcement date and before the current quarter earnings announcement
date (917,272), (ii) without a share price target issued either with the final current quarter
earnings forecast or in the 365 days before the final current quarter earnings forecast date
(359,227), (iii) without a next quarter earnings forecast either with the final current quarter
earnings forecast or in the 365 days before the final current quarter earnings forecast date
(149,856), (iv) with a stock split or a stock dividend after the prior quarter earnings
announcement date and before the current quarter earnings announcement date (13,648), and (v)
without CRSP or Compustat data to calculate control variables (134,402). The final revision
sample includes the remaining 719,373 analyst-firm-quarters corresponding to 8,504 unique
analysts and 7,509 unique firms. These sample selection procedures are detailed in Appendix A.
13
3.2 Descriptive Statistics: Revision Sample
Panel A of Table 1 reports descriptive statistics for the 719,373 analyst-firm-quarters in
the revision sample. The analyst revises a share price target after the final current quarter
earnings forecast in 13.1% of the analyst-firm-quarters, including 8.3% upward revisions and
4.8% downward revisions (thus 63.5% of the share price target revisions are revised upwards).
The analyst revises the next quarter earnings forecast after the final current quarter earnings
forecast in 11.7% of the analyst-firm-quarters, including 6.0% upward revisions and 5.6%
downward revisions (thus 51.8% of the next quarter earnings forecast revisions are revised
upwards). The analyst revises either a share price target or next quarter’s earnings forecast in
19.5% of analyst-firm-quarters. The high proportion of positive SPT and FQE revisions relative
to negative SPT and FQE revisions suggests analysts more frequently disseminate positive news
via non-earnings forecast signals after the final current quarter earnings forecast.
We also report descriptive statistics for a number of control variables. RET_QTR, the
firm’s return over the ninety days before the analyst’s final current quarter earnings forecast, has
a negative median, suggesting quarterly earnings forecasts tend to be issued when prior
economic news is negative. REV_CQE is an indicator variable set equal to one if the analyst
revises the current quarter earnings forecast after the first current quarter earnings forecast
following the prior quarter earnings announcement date, and to zero otherwise. NEG_CQE
(POS_CQE) is an indicator variable set equal to one if REV_CQE equals one and the revision is
negative (positive), and to zero otherwise. These variables help us evaluate the determinants of
an analyst’s decision to (i) revise the current quarter earnings forecast, (ii) revise the share price
target or next quarter earnings forecast, or (iii) revise no forecasts. The analyst revises the
current quarter earnings forecast in 32% of the analyst-firm-quarters in our sample and 59.6% of
14
these revisions are negative (12.9% upward revisions and 19.1% downward revisions). The
predominance of negative CQE revisions contrasts with the results for SPT and FQE revisions,
which are predominantly positive. Finally, the firm meets or beats the analyst’s forecast of
earnings (MEET_BEAT) in 70.6% of analyst-firm-quarters.
3.3 Sample Selection: Text Sample
The following sample selection procedures are detailed in Appendix A. The qualitative
prediction variables (MISS_TEXT, BEAT_TEXT) require the text of analyst reports. We collect
analyst reports from ThomsonOne Banker for 175 randomly selected analysts in the revision
sample detailed above.10 The 175 randomly selected analysts correspond to 65,539 analyst-firmquarters from the revision sample.
We search for the analyst name from the I/B/E/S
recommendation file in ThomsonOne Banker and if we cannot identify the analyst by name in
ThomsonOne Banker, we discard the analyst. We are unable to identify 33 of the 175 analysts in
ThomsonOne Banker. The text sample is thus comprised of the 53,765 analyst-firm-quarters for
the remaining 142 analysts we can identify in ThomsonOne Banker.
We collect 44,237 analyst reports corresponding to the 142 analysts and 958 unique
firms.
We match each analyst report to an analyst-firm-quarter based on its issue date
(classifying a report as belonging to an analyst-firm-quarter if the report is issued after the prior
quarter earnings announcement date and before the current quarter earnings announcement date).
We also drop analyst reports (i) issued before the analyst’s final current quarter earnings
forecast, and (ii) issued more than 30 days before the current quarter earnings announcement
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
10
We collect the analyst reports over three iterations. In the first iteration, we collect all analyst reports for 15
analysts randomly selected from the revision sample for the year 2010. In the second (third) iteration, we randomly
select 60 (100) analysts with at least twenty firm-quarters of observations for a minimum of three firms in the
revision sample.
15
date. The final text sample includes the remaining 7,593 analyst reports corresponding to 130
unique analysts and 509 unique firms.
3.4 Descriptive Statistics: Text Sample
Panel B of Table 1 reports descriptive statistics for the text sample of 53,765 analystfirm-quarters. In this sample of firms, the firm meets or beats the analyst’s forecast in 70.6% of
the observations, identical to that in the revision sample. The mean value of analysts’ positive
qualitative predictions (BEAT_TEXT) is 0.106 and the mean value of analysts’ negative
qualitative predictions (MISS_TEXT) is 0.084, suggesting analysts make positive qualitative
predictions more frequently than negative qualitative predictions.11 The revisions to the share
price target, next quarter earnings, and current quarter earnings forecasts are comparable to those
in the revision sample, but the analysts in the text sample are slightly more active.
4. Research Design and Empirical Results
4.1 The Earnings Surprise and Non-Earnings Forecast Signals
4.1.1 The Analyst’s Own Earnings Surprise
We first seek to understand whether research an analyst publishes after the final current
quarter earnings forecast predicts the analyst’s own forecast error. We address this question by
regressing the analyst’s own earnings surprise on information from research the analyst publishes
after his final current quarter earnings forecast, but before the current quarter earnings
announcement date. We estimate the following model.
EarningsSurprise = β0 + β1(NonEarningsForecastSignals) + Σβi(Controls) + ε
(1)
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
11
To validate our qualitative prediction variables, a research assistant read a subsample of 1,905 sentences in which
the textual algorithm classified the analyst as having issued a qualitative prediction. The research assistant assessed
whether the analyst clearly indicated an expectation that the firm would beat or miss expectations. We find a 0.46
(0.48) correlation between the research assistant’s coding of the sentences and the textual algorithm for those
sentences classified as beat (miss), suggesting we capture the intentions of the analyst, albeit with noise.
16
We use two measures of EarningsSurprise: (i) an indicator variable set equal to one
(zero) if the firm meets or beats (misses) the analyst’s own forecast (MEET_BEAT), and (ii) the
analyst’s earnings surprise, equal to the firm’s actual reported earnings per share less the
analyst’s final quarterly earnings per share forecast, scaled by price (CURR_SURP).
We use three measures of NonEarningsForecastSignals. First, we capture the analyst’s
revisions to the share price target. We set NEG_SPT (POS_SPT) equal to one if the analyst
issues a negative (positive) revision to the share price target after the final current quarter
earnings forecast date and before the current quarter earnings announcement date, and to zero
otherwise. Second, we capture the analyst’s revisions to the next quarter earnings forecast. We
set NEG_FQE (POS_FQE) equal to one if the analyst issues a negative (positive) revision to the
next quarter earnings forecast after the final current quarter earnings forecast date and before the
current quarter earnings announcement date, and to zero otherwise.
Third, we capture the analyst’s qualitative non-earnings forecast signals by analyzing the
text of analyst reports issued after the final current quarter earnings forecast date and before the
current quarter earnings announcement date. MISS_TEXT (BEAT_TEXT) is a count variable
equal to the number of times the analyst uses a synonym for “miss” (“beat”) and a synonym for
“earnings expectations” within the same sentence.12 The synonyms for miss include miss, fall
short, below, lower, worse, downside, and underperform. The synonyms for beat include beat,
exceed, outperform, better, higher, above, top, and upside.
The synonyms for earnings
expectations include earnings, EPS, estimates, and expectations.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
12
Our inferences remain unchanged if MISS_TEXT (BEAT_TEXT) is a count variable that equals the number of
times the analyst uses a synonym for “miss” (“beat”) and a synonym for “earnings expectations” within five words
of each other. Our inferences also remain unchanged if MISS_TEXT (BEAT_TEXT) is an indicator variable equal to
one if the analyst uses a synonym for “miss” (“beat”) and a synonym for “earnings expectations” within the same
sentence or within five words of each other.
17
We also include a series of control variables. PREV_SURP is the prior quarter earnings
surprise, equal to the firm’s actual prior quarter earnings per share less the analyst’s final prior
quarter earnings per share forecast, scaled by price. RET_QTR is the firm’s return for the period
beginning ninety days before the final current quarter earnings forecast date and ending at the
final current quarter earnings forecast date. MVE is the natural log of the market value of equity.
BTM is the book value of equity scaled by the market value of equity. ROA is the firm’s return
on assets, calculated as earnings before interest and taxes scaled by total assets. MVE, BTM, and
ROA are calculated as of the most recent fiscal year-end before the prior quarter earnings
announcement. FOLLOW is the natural log of analyst following. #DAYS is the natural log of the
number of days between the analyst’s final current quarter earnings forecast and the current
quarter earnings announcement date. #FORE is the natural log of the number of current quarter
earnings forecasts the analyst issued in the previous quarter. All variable definitions are detailed
in Appendix B.
Panel A of Table 2 reports the results for the revision sample from estimating equation
(1), which regresses the analyst’s own earnings surprise on the analyst’s non-earnings forecast
signals.
The dependent variable is MEET_BEAT in columns (1)-(3) and CURR_SURP in
columns (4)-(6). In column (1), we estimate equation (1) without any control variables. We find
positive coefficients on POS_SPT (β=0.043) and POS_FQE (β=0.018), both significant at the
1% level. We also find a significantly negative coefficient on NEG_SPT (β=-0.017) and an
insignificant coefficient on NEG_FQE. The results indicate SPT and FQE forecasts convey
economically significant information about the probability of a firm meeting or beating the
analyst’s own earnings forecast. For example, an analyst-firm-quarter with a positive share price
18
target revision will meet or beat the analyst’s earnings forecast 6% more frequently than an
analyst-firm-quarter with a negative share price target revision.
In column (2), we include control variables for factors previously shown to affect the
probability of a firm meeting or beating earnings expectations such as firm characteristics, prior
returns, and the prior quarter earnings surprise.
We find that three (four) of the four
non-earnings forecast signals are statistically significant at the 5% (10%) level, but we observe
substantial attenuation in the coefficient estimates. For instance, the coefficient estimate on
POS_SPT decreases from β=0.043 to β=0.026 after including the control variables. It is unclear
whether the coefficient estimates with or without control variables included in the model better
capture the information analysts convey to clients with non-earnings forecast signals because
investors may not incorporate predictable errors into analyst forecasts. In column (3), we also
include analyst and quarter fixed effects and obtain similar inferences to those in column (2).
In columns (4)-(6), we estimate analogous models to those in columns (1)-(3), but the
dependent variable is CURR_SURP, the analyst’s signed earnings surprise.
In all three
specifications, the coefficients on POS_SPT, NEG_SPT, and POS_FQE remain statistically
significant at the 5% level, whereas the coefficient on NEG_FQE is insignificant in two
specifications and significantly negative in one specification.
Thus, the evidence remains
consistent with the analyst’s non-earnings forecast signals providing information content for the
analyst’s own earnings surprise.
Panel A of Table 3 reports the results for the textual sample from estimating equation (1),
which regresses the analyst’s own earnings surprise on the analyst’s qualitative predictions. We
estimate analogous models to those presented in Panel A of Table 2, except the non-earnings
forecast signals are the qualitative predictions, BEAT_TEXT and MISS_TEXT. Thus, in column
19
(1), we regress MEET_BEAT on BEAT_TEXT and MISS_TEXT without control variables. The
coefficient on MISS_TEXT is significantly negative (β=-0.048) and the coefficient on
BEAT_TEXT is significantly positive (β=0.050). The results indicate qualitative predictions also
convey economically significant information about the probability of a firm meeting or beating
the analyst’s earnings forecast. For example, an analyst-firm-quarter with a positive qualitative
prediction will meet or beat the analyst’s earnings forecast 9.8% more frequently than an analystfirm-quarter with a negative qualitative prediction.
In column (2), we include the same series of control variables as in Panel A of Table 2.
We find similar coefficient estimates on MISS_TEXT and BEAT_TEXT after including these
control variables and both coefficient estimates remain statistically significant at the 1% level.
In column (3), we also include analyst and quarter fixed effects and find similar coefficient
estimates and inferences. In columns (4)-(6), we estimate analogous models to those in columns
(1)-(3), but the dependent variable is CURR_SURP, the analyst’s signed earnings surprise. The
coefficient on MISS_TEXT (BEAT_TEXT) remains negative (positive) and significant at the 1%
level in all three specifications.
4.1.2 The Consensus Earnings Surprise
In our main analysis, we examine firm performance relative to the analyst’s own forecast,
because an investor evaluating an analyst’s research product will judge the analyst on the
information in the forecasts he publishes. In this section, we examine performance relative to the
consensus forecast to examine whether non-earnings forecast signals provide incremental
information relative to the market’s expectation of earnings. To evaluate the information in
non-earnings forecast signals, we replace the analyst’s own earnings surprise in model (1) with
20
the consensus earnings surprise calculated using earnings forecasts issued after the previous
quarter’s earnings announcement. We estimate the following model:
ConsensusEarningsSurprise = β0 + β1(NonEarningsForecastSignals) + Σβi(Controls) + ε
(2)
We use two measures of ConsensusEarningsSurprise: (i) an indicator variable set equal
to one (zero) if the firm meets or beats (misses) the consensus earnings forecast on the day of the
analyst’s final current quarter earnings forecast (MEET_BEAT_CON), and (ii) the consensus
earnings surprise, equal to the firm’s actual earnings per share less the consensus earnings per
share forecast on the day of the analyst’s final current quarter earnings forecast, scaled by price
(CURR_SURP_CON).
In Panel B of Table 2, we estimate analogous models to those in Panel A of Table 2, but
the dependent variable is the consensus earnings surprise rather than the analyst’s own earnings
surprise. In column (1), we regress MEET_BEAT_CON on the analyst’s non-earnings forecast
signals without control variables. We find positive coefficients on POS_SPT (β=0.084) and
POS_FQE (β=0.034), and negative coefficients on NEG_SPT (β=-0.045) and NEG_FQE
(β=-0.018), all significant at the 1% level. Moreover, the coefficients are all economically larger
than the corresponding coefficients in Panel A of Table 2. In column (2), we include control
variables and in column (3) we also include analyst and quarter fixed effects. We find similar
results, all eight coefficients are statistically significant with the sign predicted by after-revision
bias and seven of the eight coefficients have a larger magnitude coefficient than the
corresponding coefficient in Panel A of Table 2. In columns (4)-(6), we continue to find all
coefficients are significant and ten of the twelve coefficient estimates are larger than the
corresponding coefficient in Panel A of Table 2.
21
In Panel B of Table 3, we estimate analogous models to those in Panel A of Table 3, but
the dependent variable is the consensus earnings surprise rather than the analyst’s own earnings
surprise.
Across all six columns, we find that both BEAT_TEXT and MISS_TEXT have
coefficients statistically significant at the 1% level with coefficient magnitudes larger than the
corresponding estimates using the analyst’s own earnings surprise as the dependent variable.
Collectively, the evidence indicates that analysts’ non-earnings forecast signals also provide
information content about the consensus earnings surprise.
4.2 When Do Analysts Issue Non-Earnings Forecast Signals?
Next, we exploit cross-sectional variation in the frequency of non-earnings forecast
signals to provide evidence on the incentives that generate after-revision bias. Specifically, we
test three hypotheses: (i) analysts issue non-earnings forecast signals more for positive news to
avoid increasing earnings expectations (“catering to management”), (ii) analysts issue
non-earnings forecasts more frequently for news that would have moved the analyst’s forecast
away from the consensus (“strategic herding”), and (iii) analysts issue non-earnings forecast
signals more frequently for smaller magnitude information (“frictions”).
4.2.1 Do Analysts Issue Non-Earnings Forecast Signals More Frequently for Good News?
Incorporating good news into an SPT or FQE forecast, and bad news into a CQE forecast,
increases the likelihood that management meets or beats the earnings forecast. Given analysts’
incentives to both publish forecasts that managers will meet or beat (Ke and Yu 2006; Brown et
al. 2015) and communicate information to clients (Groysberg et al. 2011; Brown et al. 2015),
mapping good and bad news into different forecasts may help analysts satisfy multiple
stakeholders.
22
We report forecast frequencies for positive and negative forecast revisions in Panel A of
Table 4. Row (1) indicates that analysts issue a positive non-earnings forecast signal (i.e., SPT
or FQE forecast revision) in 12.2% of analyst-firm-quarters, whereas they issue a negative nonearnings forecast signal in 8.9% of analyst-firm-quarters. The 3.3% difference is statistically
significant, indicating that analysts are more likely to issue positive than negative non-earnings
forecast signals. Rows (2) and (3) indicate this relation holds separately for SPT and FQE
forecast revisions. Row (4) documents that analysts issue a positive CQE forecast revision in
12.9% of analyst-firm-quarters, whereas they issue negative CQE forecast revisions in 19.1% of
analyst-firm-quarters. The 6.2% difference is also statistically significant and indicates that
analysts are more likely to revise CQE forecasts downwards. In rows (5)-(8), we exclude
analyst-firm-quarters in which the analyst revises the current quarter earnings forecast
(REV_CQE=1). The frequency of both positive and negative non-earnings forecast signals
increases relative to rows (1)-(3), but the ratio of positive to negative revisions also increases,
thus corroborating the above results.
The evidence is consistent with analysts responding to incentives to issue forecasts that
managers will meet or beat. Conditional on positive news, analysts are less likely to revise the
CQE forecast so earnings will be compared to a static benchmark. Conditional on negative
news, managers will be compared to a benchmark that varies with performance. The systematic
correlation between the sign of the news and the forecasts analysts choose to revise suggests one
mechanism analysts use to walk-down forecasts is omitting positive news from the CQE
forecast, yet incorporating positive news into other forecasts.
23
4.2.2 Do Analysts Issue Non-Earnings Forecast Signals More Frequently for Bold News?
Second, we examine whether analysts issue SPT and FQE forecast revisions more
frequently when a corresponding revision to the CQE forecast would move the analyst’s CQE
forecast towards the consensus (herding) or away from the consensus (bold). Prior research
suggests investors judge analyst forecasts relative to those of competing analysts, and analysts
face career penalties for issuing more inaccurate forecasts than their peers (Hong et al. 2000;
Groysberg et al. 2011). We argue these incentives may lead analysts to avoid issuing extreme
forecasts relative to those of their peers.13
The results are reported in Panel B of Table 4. Row (1) indicates that analysts issue a
negative non-earnings forecast signal (i.e., SPT or FQE forecast revision) in 9.4% (8.3%) of
analyst-firm-quarters when the analyst’s final CQE forecast is below (above) the consensus
forecast. Row (2) indicates that analysts issue a positive non-earnings forecast signal in 11.1%
(13.2%) of analyst-firm-quarters when the analyst’s final CQE forecast is below (above) the
consensus forecast. Thus, analysts are more likely to issue non-earnings forecast signals if a
corresponding revision to the CQE forecast would move the analyst’s forecast away from the
consensus. This finding is opposite to that predicted by the theory that analysts’ non-earnings
forecast signals are a function of their forecast error. The consensus forecast contains substantial
information about future earnings, omitted from the analyst’s own forecast. If non-earnings
forecast revision frequency increased in the prevalence of news omitted from the analyst’s
earnings forecasts, we would predict results opposite to those we find (i.e., more herding than
bold non-earnings forecast signals). Rows (3)-(6) indicate that the same relations hold for the
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
13
In a private conversation, an analyst at a prestigious bulge bracket firm commented that issuing a forecast which
departs substantially from consensus draws attention. The analyst commented that without (with) a convincing
reason for the deviation, the attention can damage (enhance) the analyst’s reputation.
24
SPT and FQE forecast revisions separately – for both positive and negative news, non-earnings
forecast signal frequency is higher for bold forecasts than herding forecasts.
In rows (7)-(8), we present similar statistics for CQE forecasts. Row (7) documents that
analysts issue a negative CQE forecast revision in 14.7% (23.1%) of analyst-firm-quarters when
the analyst’s initial CQE forecast is below (above) the consensus forecast. Row (8) indicates
that analysts issue a positive CQE revision in 17.4% (9.4%) of analyst-firm-quarters when the
analyst’s initial CQE forecast is below (above) the consensus forecast. Thus, analysts tend to
revise the CQE forecast towards the consensus, whereas analysts are more likely to issue
non-earnings forecast signals in lieu of CQE forecast revisions that would move the analyst away
from the consensus.
Documenting that analysts suppress bold forecasts suggests that measures of forecast
dispersion may understate the amount of disagreement, because analysts suppress their private
information when it moves them away from the consensus. Consistent with this, in untabulated
analyses we find that: (i) analyst forecast volatility is significantly lower in firm-quarters with at
least one analyst issuing a non-earnings forecast signal, and (ii) analyst forecast volatility is
significantly negatively correlated with the percentage of analysts following the firm who issue a
non-earnings forecast signal.
4.2.3 Do Analysts Issue Non-Earnings Forecast Signals More Frequently for Less Substantial
News?
Third, we examine whether the amount of earnings news during the quarter affects the
analyst’s decision to issue a non-earnings forecast signal relative to a current quarter earnings
forecast revision. If frictions limit revisions to the CQE forecast, we expect less (greater)
earnings news subsequent to the initial CQE forecast increases the likelihood the analyst issues a
25
non-earnings forecast signal (revises the CQE forecast) instead of revising the CQE forecast
(issuing a non-earnings forecast signal). Frictions can arise because (i) revising a forecast
imposes costs on the analyst, (ii) revising a forecast imposes costs on the analyst’s clients, which
the analyst endogenizes and thus only revises the forecast when the information is sufficiently
large, and/or (iii) the analyst ultimately judges the firm’s performance relative to his earnings
forecast, and the analyst would prefer to judge the firm relative to a static benchmark rather than
one that varies with performance. We interpret evidence of analysts revising the CQE forecast
(SPT of FQE forecast) when there is significant (little) earnings news during the quarter as
consistent with frictions impeding the flow of information into the CQE forecast.
The results examining how forecast frequency varies with the magnitude of the news are
reported in Panel C of Table 4. Row (1) documents that analysts revise an SPT or FQE forecast
in 21.2% (17.8%) of analyst-firm-quarters when the absolute value of the analyst’s initial
earnings surprise is below (above) the sample median. Rows (2) and (3) document similar
findings for SPT and FQE forecast revisions separately. Contrary to this, row (4) indicates that
analysts revise the CQE forecast in 25.4% (38.6%) of analyst-firm-quarters when the analyst’s
absolute earnings surprise is below (above) the median. Thus, consistent with frictions impeding
(not impeding) CQE forecasts when a small (large) amount of earnings news occurs after the
analyst’s initial CQE forecast, the analyst is motivated to revise the SPT or FQE (CQE) forecast.
4.2.4 Determinants of Non-Earnings Forecast Signals: Regression Results
Next, we examine whether our univariate results continue to hold when including other
determinants of non-earnings forecast signals as well as a series of control variables. We
estimate the following model:
NonEarningsForecastSignals = β0 + Σβi(Determinants) + Σβj(Controls) + ε
(3)
26
We use four different independent variables. First, we use an indicator variable set equal
to one (zero) if the analyst revises (does not revise) either the SPT forecast or the FQE forecast
after the final CQE forecast (REV_TOT). Second, the indicator variable captures revisions to the
CQE forecast (REV_CQE). Third, we set an indicator variable equal to one if the analyst issues
a positive revision to either an SPT or FQE forecast (POS_TOT). Fourth, we set an indicator
variable equal to one if the analyst issues a negative revision to either an SPT or FQE forecast
(NEG_TOT).
We use different specifications and variables to test each of our three hypotheses. We
test whether analysts revise non-earnings forecast signals (CQE forecasts) more for good (bad)
news by including as a regressor returns in the ninety days before the final CQE forecast of the
quarter.
Because analysts do not incorporate all of the news from returns into forecasts
(Abarbanell and Bernard 1992), these returns predict error in forecasts. We expect that analysts
will be more likely to respond to the error by issuing a non-earnings forecast signal when it is
positive, leading to the prediction of a positive coefficient on QTR_RET in column (1), when
REV_TOT is the dependent variable. We expect analysts will be more likely to issue a CQE
revision when the news is negative, leading to a prediction of a negative coefficient on
QTR_RET in column (2), when REV_CQE is the dependent variable.
We test whether analysts issue more non-earnings forecast (CQE) revisions for bold
(herding) revisions, those that would move the CQE forecast away from (towards) the consensus,
by regressing the signed indicators (POS_TOT, NEG_TOT) on an indicator for whether the
analyst forecast is above the consensus (ABOVE_CON). If analysts forecast strategically to
avoid bold CQE revisions, we expect being above the consensus will make the analyst less
27
(more) likely to issue a negative (positive) non-earnings forecast signal. Thus, we predict a
negative (positive) coefficient on ABOVE_CON in column (3) (column (4)).
We test whether analysts are more likely to issue CQE forecast revisions in response to
more significant news by including the absolute value of the initial earnings surprise
(ABS[CURR_SURP_IN]).
The frictions hypothesis predicts ABS(CURR_SURP_IN) will be
positively associated with REV_CQE in column (2) and negatively or insignificantly associated
with REV_TOT in column (1).
In addition to the control variables included in Table 2, we also include a series of
variables to capture analyst characteristics. ABS(PREV_SURP) is the absolute value of the
analyst’s prior quarter earnings surprise. EXPERIENCE is the natural log of the number of years
since the analyst’s first earnings forecast in I/B/E/S.14 BROKER_SIZE is the natural log of the
number of analysts at the corresponding analyst’s firm. ALLSTAR is an indicator variable set
equal to one if the analyst finished first, second, third, or honorable mention in the Institutional
Investor all-star rankings during the calendar year of the quarterly forecast, and to zero
otherwise.15
The results from estimating equation (3) are reported in Table 5. Consistent with analysts
using non-earnings forecasts to issue positive news, we find a statistically significant positive
coefficient on QTR_RET in column (1), when the dependent variable is REV_TOT. Conversely,
analysts issue more CQE forecasts in response to negative news as evidenced by the significantly
negative coefficient on QTR_RET in column (2). This suggests that at least a portion of the
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
14
Our inferences remain unchanged if we define EXPERIENCE as the analyst’s firm-specific experience as opposed
to the analyst’s total experience.
15
To avoid a substantial reduction in sample size due to missing observations, we apply zero-order regression by
setting ALLSTAR equal to zero if missing, and including an indicator variable set equal to one for these observations,
ALLSTAR_DUM.
28
walk-down arises because analysts are less likely to revise the CQE forecast in response to
positive news.
Consistent with analysts using non-earnings forecast signals to avoid bold forecasts, the
coefficient on ABOVE_CON is negative and statistically significant in column (3), when the
dependent variable is NEG_TOT, and is positive and statistically significant in column (4), when
the dependent variable is POS_TOT.
Also consistent with the frictions explanation, the coefficient on ABS(CURR_SURP_IN)
is positive and statistically significant in column (2), when the dependent variable is REV_CQE,
and it is negative and marginally significant in column (1), when the dependent variable is
REV_TOT. Thus, analysts issue CQE revisions (non-earnings forecast signals) more frequently
when the absolute value of the initial forecast error is higher (lower).
We also note that a number of analyst forecast characteristics have significant
associations with REV_TOT. The coefficients on ABS(PREV_SURP) and EXPERIENCE are
significantly negative and the coefficients on BROKER_SIZE and ALLSTAR are significantly
positive.
Thus, non-earnings forecast signals tend to be issued by analysts who are more
accurate, less experienced, at larger brokerages, and all-star analysts. If issuing a non-earnings
forecast signal reduces accuracy by omitting information from the forecast, these relations
suggest correlations between forecast accuracy and analyst characteristics may be affected by
strategy in addition to skill (Clement and Tse 2003).
4.3 When Do Earnings Forecast Signals Convey More Information Content?
In the prior section, we test for cross-sectional variation in non-earnings forecast signal
frequency to study when after-revision bias is most pronounced. However, if more frequent
revisions also have less information content, the increased frequency would not necessarily
29
imply greater after-revision bias. To test for incremental information content, we estimate
modified versions of equation (1):
EarningsSurprise = β0 + β1(NEG_TOT) + β2(POS_TOT) + Σβi(Controls) + ε
(4a)
EarningsSurprise = β0 + β1(NEG_TOT_ABOVE) + β2(NEG_TOT_BELOW) +
β3(POS_TOT_ABOVE) + β4(POS_TOT_BELOW) + Σβi(Controls) + ε
(4b)
In equation (4a), the independent variables include NEG_TOT and POS_TOT, which
capture negative and positive non-earnings forecast signals, respectively. We then test whether
the absolute value of the coefficient on POS_TOT is significantly greater than the absolute value
of the coefficient on NEG_TOT.
In equation (4b), the independent variables also capture
whether the analyst’s final CQE forecast was above or below the consensus earnings forecast.
NEG_TOT_ABOVE (NEG_TOT_BELOW) is an indicator variable set equal to one if the analyst
issues a negative SPT or FQE revision and the analyst’s final CQE forecast was above (below)
the consensus earnings forecast.
POS_TOT_ABOVE (POS_TOT_BELOW) is an indicator
variable set equal to one if the analyst issues a positive SPT or FQE revision and the analyst’s
final CQE forecast was above (below) the consensus earnings forecast. We then test whether the
absolute value of the coefficient on NEG_TOT_BELOW is significantly greater than the absolute
value of the coefficient on NEG_TOT_ABOVE, as well as whether the absolute value of the
coefficient on POS_TOT_ABOVE is significantly greater than the absolute value of the
coefficient on POS_TOT_BELOW.
As in equation (1), the dependent variable is either
MEET_BEAT or CURR_SURP in equations (4a) and (4b).
Panel A of Table 6 reports the results from estimating equation (4a). In column (1), the
dependent variable is MEET_BEAT. As expected, the coefficient on NEG_TOT is significantly
negative (β=-0.009) and the coefficient on POS_TOT is significantly positive (β=0.040). More
30
importantly, the absolute value of the coefficient on POS_TOT is significantly greater than the
absolute value of the coefficient on NEG_TOT.
This indicates that positive non-earnings
forecast signals provide more information content about the analyst’s earnings surprise relative
to negative non-earnings forecast signals. We include control variables in column (2), as well as
analyst and quarter fixed effects in column (3).
The absolute value of the coefficient on
POS_TOT remains significantly greater than the absolute value of the coefficient on NEG_TOT
in both specifications. In columns (4)-(6), the dependent variable is CURR_SURP. The absolute
value of the coefficient on POS_TOT is greater than the absolute value of the coefficient on
NEG_TOT in all three columns, but significantly so only in column (4). Overall, the evidence is
consistent with positive non-earnings forecast signals providing more information content about
the analyst’s earnings surprise relative to negative non-earnings forecast signals.
Panel B of Table 6 reports the results from estimating equation (4b). In column (1), the
dependent variable is MEET_BEAT.
The absolute value of the coefficient on
NEG_TOT_BELOW (β=-0.016) is greater than the absolute value of the coefficient on
NEG_TOT_ABOVE (β=-0.012), but the difference is not statistically significant. The absolute
value of the coefficient on POS_TOT_ABOVE (β=0.048) is greater than the absolute value of the
coefficient on POS_TOT_BELOW (β=0.043), but the difference also is not statistically
significant. In column (2), we include a series of control variables, and in column (3) we also
include analyst and quarter fixed effects. The coefficient on POS_TOT_ABOVE is marginally
statistically greater than the coefficient on POS_TOT_BELOW in both specifications, and the
absolute value of the coefficient on NEG_TOT_BELOW is greater than that on
NEG_TOT_ABOVE in both specifications, but only statistically greater in column (3).
columns (4)-(6) the dependent variable is CURR_SURP.
In
In all three specifications, the
31
coefficient
on
POS_TOT_ABOVE
is
significantly
greater
than
the
coefficient
on
POS_TOT_BELOW, and the absolute value of the coefficient on NEG_TOT_BELOW is
significantly greater than that on NEG_TOT_ABOVE.
Collectively, the evidence is consistent with analysts’ non-earnings forecast signals
having more information content about the earnings surprise when a corresponding CQE forecast
would move the analyst away from the consensus.
4.4 Non-Earnings Forecast Signals and the Analyst’s Stock Recommendation
Next, we examine how an analyst’s stock recommendation affects (i) the degree to which
the analyst walks down the CQE forecast, and (ii) how the analyst maps positive and negative
information into the CQE forecast. If an analyst has a positive recommendation on a stock, we
might expect the analyst to have a positive outlook on the firm’s earnings. However, optimistic
earnings expectations could lead the firm to miss earnings estimates, which could have negative
consequences for both the share price and the manager’s human capital (Matsunaga and Park
2001; Skinner and Sloan 2002). Given the importance of meeting or beating earnings estimates,
analysts with buy recommendations may have a vested incentive in helping the firm meet or beat
earnings expectations by walking down their forecasts.
More specifically, one mechanism
analysts could use to walk down their forecasts is to omit positive information from the CQE
forecasts and include such positive information in the SPT or FQE forecasts.
In Table 7, we examine the effect of the analyst’s stock recommendation on (i) the
likelihood that the firm meets or beats the analyst’s earnings forecast, (ii) the relative frequency
of positive vs. negative CQE forecast revisions, and (iii) the relative frequency of positive vs.
negative SPT and FQE forecast revisions. In column (1), we regress MEET_BEAT on an
indicator variable set equal to one if the analyst’s stock recommendation is buy or strong buy
32
(BUY). The coefficient on BUY is positive and statistically significant (β=0.019). In column (2),
the dependent variable is the earnings surprise, calculated using the analyst’s first CQE forecast
during the quarter (MEET_BEAT_IN). The coefficient on BUY is still positive and statistically
significant (β=0.013), but only 68% of the coefficient in column (1), suggesting analysts with a
buy rating walk down their forecasts more aggressively during the quarter. In columns (3)-(5),
we capture the relative frequency of positive vs. negative CQE, SPT, and FQE forecast revisions.
DEV_CQE (DEV_SPT, DEV_FQE) is set equal to positive one (negative one) if the analyst
positively (negatively) revises the CQE (SPT, FQE) forecast, and zero otherwise.
The
coefficient on BUY is insignificant when DEV_CQE and DEV_FQE are the dependent variables,
but it is significantly positive when DEV_SPT is the dependent variable, suggesting analysts with
a buy recommendation are more likely to issue positive SPT forecast revisions.
In columns (6)-(10), we estimate analogous models to those in columns (1)-(5), but we
include a series of control variables as well as analyst and quarter fixed effects. The coefficient
on BUY is significantly positive in column (1), when the dependent variable is MEET_BEAT, but
it is insignificant in column (7), when the dependent variable is MEET_BEAT_IN. This again
suggests that analysts with a positive stock recommendation walk down the forecast more
aggressively during the quarter. The coefficient on BUY is significantly negative in column (8),
when the dependent variable is DEV_CQE, consistent with these analysts issuing more negative
CQE forecast revisions. The coefficient on BUY is significantly positive in column (9), when the
dependent variable is DEV_SPT, consistent with these analysts issuing more positive SPT
forecast revisions. The coefficient on BUY is insignificant in column (10), when the dependent
variable is DEV_FQE.
33
4.5 Non-Earnings Forecast Signals and Returns
4.5.1 Do Non-Earnings Forecast Signals Predict Earnings Announcement Returns?
In our next set of tests, we examine whether non-earnings forecast signals predict
earnings announcement returns. We conduct these tests for two reasons. First, examining the
effect of non-earnings forecast signals on price formation illustrates a potential consequence of
after-revision bias. Second, one possible explanation for the existence of after-revision bias is
analysts wanting to limit the dissemination of their information to clients (“tipping”).
A
necessary but not sufficient condition for the tipping explanation is a finding of significant
returns, enabling clients to capture a larger share of the analyst’s informational rents. We
estimate the following model:
RET_ANN = β0 + β1(NonEarningsForecastSignals) + Σβi(Controls) + ε
(5)
RET_ANN is the firm’s market-adjusted returns compounded over the three-day trading
window beginning the trading day before the earnings announcement date and ending the trading
day after the earnings announcement date. For earnings announcements occurring after 4:00
P.M. Eastern Time, we use the next trading day as the earnings announcement date. We include
the same series of control variables as in equation (1).
Table 8 reports the results from estimating equation (5). In columns (1)-(2), we present
estimates using the revision sample. In column (1), we present our base model without control
variables. The coefficient on POS_SPT is significantly positive (β=0.002), but the coefficients
on NEG_SPT, NEG_FQE, and POS_FQE are insignificant. Our results suggest that positive
share price target revisions predict earnings announcement returns, but the other revisions do not.
In column (2), we include a series of control variables and the results are very similar. Columns
(3)-(4) report the results from estimating equation (5) on the text sample. In column (3), we
34
regress earnings announcement returns on BEAT_TEXT and MISS_TEXT without control
variables. The coefficient on BEAT_TEXT is significantly positive (β=0.005) and the coefficient
on MISS_TEXT is significantly negative (β=-0.007). In column (6), we include the same series
of control variables and the results are very similar.
4.5.2 The Market Response to Analysts’ Non-Earnings Forecast Signals
Finally, we examine whether the market responds to analysts’ non-earnings forecast
signals. This analysis provides descriptive evidence on the amount of the market reaction to the
announcement of non-earnings forecast signals. Table 9 reports the firm’s returns when the
analyst issues a non-earnings forecast signal. Specifically, RET_SPT (RET_FQE, RET_TEXT) is
the firm’s market-adjusted return cumulated over the three-day period centered on the date of the
analyst’s SPT forecast revision (FQE forecast revision, qualitative prediction). Rows (1) and (2)
report that the market responds negatively to downward share price target revisions
(RET_SPT=-0.014) and positively to upward share price target revisions (RET_SPT=0.015). It is
worth noting that these average returns are much lower than returns found in the extant literature,
which do not exclude revisions accompanying earnings announcements. Rows (3) and (4) report
that the market responds negatively to downward next quarter earnings forecast revisions
(RET_FQE=-0.005) and positively to upward next quarter earnings forecast revisions
(RET_FQE=0.006). Rows (5) and (6) report that the market responds negatively to downward
qualitative predictions (RET_TEXT=-0.006) and positively to upward qualitative predictions
(RET_TEXT=0.009).
35
5. Conclusion
We identify a novel bias in analyst forecasts, after-revision bias, which we identify by
examining an analyst’s reports after his final earnings forecast of the quarter. Reviewing the text
of analyst reports, we find explicit examples of analysts stating that they expect firms to beat or
miss their own earnings forecast.
We regard this as anecdotal evidence that analysts
communicate news about current quarter earnings without revising their current quarter
forecasts.
We provide large sample evidence on the means, determinants, and consequences of
disseminating news about current quarter earnings after the final forecast of the current quarter.
Specifically, we document that qualitative predictions in the text of analyst reports, share price
target revisions, and revisions to future quarter earnings forecasts (collectively, non-earnings
forecast signals) predict the analyst’s own forecast error. We also find non-earnings forecast
signals are (i) more prevalent for positive news than negative news, consistent with analysts
responding to incentives to issue forecasts managers will meet or beat, (ii) more prevalent when
a corresponding revision to the CQE forecast would move the analyst away from the consensus,
consistent with herding behavior, and (iii) issued more frequently when the absolute value of
forecast error is low, consistent with frictions limiting revisions to the current quarter earnings
forecast when the amount of news is low. Overall, the evidence is more consistent with an
intentional, as opposed to unintentional, explanation for after-revision bias.
Our findings demonstrate that the importance of the earnings forecast leads to a decrease
in the information flowing through that forecast.
Such behavior by analysts has critical
implications for academic researchers using analysts’ current quarter forecasts (either
individually or in a consensus) to proxy for investor expectations in a variety of widely studied
36
contexts. Our findings that analysts respond to positive news by revising an SPT or FQE
forecast more frequently than revising a CQE forecast suggests a managerial preference for
meeting or beating earnings estimates decreases the flow of positive news into the CQE forecast.
As a result, a portion of the walk-down likely arises out of a hesitancy to revise the forecast
upward rather than a willingness to revise the forecast downward.
Though we present evidence suggesting after-revision bias is intentional, we have limited
ability to identify the ultimate cause of the bias. A potential explanation is that omitting
information from forecasts can lead analysts to deliver a more consistent message to clients.
Numerous studies have documented that analysts are slow to update forecasts (Lys and Sohn
1990; Abarbanell and Bernard 1992; Brous and Shane 2001), which allows the analyst to
increase the frequency of revisions with the same sign (Smith-Raedy et al. 2006). Analysts may
strive to issue a consistent message if clients have limited attention to the message of any one
analyst. Although updating a forecast positively one quarter and negatively another is not
necessarily inconsistent, it may appear inconsistent to an investor not paying sufficient attention.
Evidence from social psychology suggests that individuals discount evidence that appears
inconsistent, so appearing consistent may have value.
37
References
Abarbanell, J. S. "Do analysts' earnings forecasts incorporate information in prior stock price
changes?" Journal of Accounting and Economics 14.2 (1991): 147-165.
Abarbanell, J. S., and L. Bernard. "Tests of analysts' overreaction/underreaction to earnings
information as an explanation for anomalous stock price behavior." Journal of Finance
47.3 (1992): 1181-1207.
Bagnoli, M., S. G. Watts, and Y. Zhang. "Reg-FD and the competitiveness of all-star analysts."
Journal of Accounting and Public Policy 27.4 (2008): 295-316.
Bartov, Eli, Dan Givoly, and Carla Hayn. "The rewards to meeting or beating earnings
expectations." Journal of Accounting and Economics 33.2 (2002): 173-204.
Bernhardt, Dan, Chi Wan, and Zhijie Xiao. "The reluctant analyst." Journal of Accounting
Research (2016).
Bhojraj, Sanjeev, et al. "Making sense of cents: An examination of firms that marginally miss or
beat analyst forecasts." The Journal of Finance 64.5 (2009): 2361-2388.
Bradshaw, M. T., S. A. Richardson, and R. G. Sloan. "Do analysts and auditors use information
in accruals?" Journal of Accounting Research 39.1 (2001): 45-74.
Bradshaw, M. T., M. S. Drake, J. N. Myers, and L. A. Myers (2012). A re-examination of
analysts’ superiority over time-series forecasts of annual earnings. Review of Accounting
Studies, 17(4), 944-968.
Bradshaw, M. T., L. D. Brown, and K. Huang. "Do sell-side analysts exhibit differential target
price forecasting ability?" Review of Accounting Studies 18.4 (2013): 930-955.
Brav, A., and R. Lehavy. "An Empirical Analysis of Analysts' Target Prices: Short-term
Informativeness and Long-term Dynamics." Journal of Finance 58.5 (2003): 1933-1968.
Brown, L. D., A. C. Call, M. B. Clement, and N. Y. Sharp. "Inside the “Black Box” of Sell!Side
Financial Analysts." Journal of Accounting Research 53.1 (2015): 1-47.
Brown, L. D., Hagerman, R. L., Griffin, P. A., & Zmijewski, M. E. "Security analyst superiority
relative to univariate time-series models in forecasting quarterly earnings." Journal of
Accounting and Economics 9.1 (1987): 61-87.
Chen, Qi, and Wei Jiang. "Analysts’ weighting of private and public information." Review of
Financial Studies 19.1 (2006): 319-355.
Chen, X., Q. Cheng, and K. Lo. "On the relationship between analyst reports and corporate
disclosures: Exploring the roles of information discovery and interpretation." Journal of
Accounting and Economics 49.3 (2010): 206-226.
Clement, Michael B., and Senyo Y. Tse. "Do investors respond to analysts' forecast revisions as
if forecast accuracy is all that matters?." The Accounting Review 78.1 (2003): 227-249.
Clement, Michael B., and Senyo Y. Tse. "Financial analyst characteristics and herding behavior
in forecasting." The Journal of Finance 60.1 (2005): 307-341.
Cotter, Julie, Irem Tuna, and Peter D. Wysocki. "Expectations management and beatable targets:
How do analysts react to explicit earnings guidance?." Contemporary Accounting
Research 23.3 (2006): 593-624.
DeBondt, Werner FM, and Richard H. Thaler. "Do security analysts overreact?." The American
Economic Review (1990): 52-57.
Degeorge, Francois, Jayendu Patel, and Richard Zeckhauser. "Earnings management to exceed
thresholds." The Journal of Business 72.1 (1999): 1-33.
38
Farrell, Kathleen A., and David A. Whidbee. "Impact of firm performance expectations on CEO
turnover and replacement decisions." Journal of Accounting and Economics 36.1 (2003):
165-196.
Frankel, R., S. P. Kothari, and J. Weber. "Determinants of the informativeness of analyst
research." Journal of Accounting and Economics 41.1 (2006): 29-54.
Francis, Jennifer. "Discussion of self-selection and analyst coverage." Journal of Accounting
Research 35 (1997): 201-208.
Graham, John R., Campbell R. Harvey, and Shiva Rajgopal. "The economic implications of
corporate financial reporting." Journal of Accounting and Economics 40.1 (2005): 3-73.
Grossman, S., and J. Stiglitz. "On the impossibility of informationally efficient markets." The
American Economic Review (1980): 393-408.
Groysberg, B., P. M. Healy, and D. A. Maber. "What Drives Sell!Side Analyst Compensation at
High!Status Investment Banks?" Journal of Accounting Research 49.4 (2011): 969-1000.
Hong, Harrison, Jeffrey D. Kubik, and Amit Solomon. "Security analysts' career concerns and
herding of earnings forecasts." The Rand journal of economics (2000): 121-144.
Hughes, J., J. Liu, and W. Su. "On the relation between predictable market returns and
predictable analyst forecast errors." Review of Accounting Studies 13.2-3 (2008): 266291.
Irvine, P., M. Lipson, and A. Puckett. "Tipping." Review of Financial Studies 20.3 (2007): 741768.
Ivkovic, Z, and N. Jegadeesh. "The timing and value of forecast and recommendation revisions."
Journal of Financial Economics 73.3 (2004): 433-463.
Kang, Sok-Hyon, John O'Brien, and K. Sivaramakrishnan. "Analysts' interim earnings forecasts:
evidence on the forecasting process." Journal of Accounting Research 32.1 (1994): 103112.
Ke, B., and Y. Yu. "The effect of issuing biased earnings forecasts on analysts' access to
management and survival." Journal of Accounting Research 44.5 (2006): 965-999.
Kirk, Marcus P., David A. Reppenhagen, and Jennifer Wu Tucker. "Meeting individual analyst
expectations." The Accounting Review 89.6 (2014): 2203-2231.
Lim, T. "Rationality and analysts' forecast bias." Journal of Finance 56.1 (2001): 369-385.
Lin, H., and M. F. McNichols. "Underwriting Relationships, Analysts’ Earnings Forecasts and
Investment Recommendations." Journal of Accounting and Economics 25.1 (1998): 101127.
Lys, T. and S. Sohn. "The association between revisions of financial analysts' earnings forecasts
and security-price changes." Journal of Accounting and Economics 13.4 (1990): 341-363.
Matsumoto, Dawn A. "Management's incentives to avoid negative earnings surprises." The
Accounting Review 77.3 (2002): 483-514.
Matsunaga, Steven R., and Chul W. Park. "The effect of missing a quarterly earnings benchmark
on the CEO's annual bonus." The Accounting Review 76.3 (2001): 313-332.
Raedy, J., P. Shane, and Y. Yang. "Horizon!Dependent Underreaction in Financial Analysts'
Earnings Forecasts." Contemporary Accounting Research 23.1 (2006): 291-322.
Richardson, S. A., S. H. Teoh, and P. D. Wysocki. "The Walk-Down to Beatable Analyst
Forecasts: The Role of Equity Issuance and Insider Trading Incentives." Contemporary
Accounting Research 21.3 (2004): 885-924.
Roychowdhury, Sugata. "Earnings management through real activities manipulation." Journal of
Accounting and Economics 42.3 (2006): 335-370.
39
Shane, P., and P. Brous. "Investor and (Value Line) Analyst Underreaction to Information about
Future Earnings: The Corrective Role of Non-Earnings-Surprise Information." Journal of
Accounting Research 39.2 (2001): 387-404.
Sims, C. "Implications of rational inattention." Journal of Monetary Economics 50.3 (2003):
665-690.
Skinner, Douglas J., and Richard G. Sloan. "Earnings surprises, growth expectations, and stock
returns or don't let an earnings torpedo sink your portfolio." Review of Accounting Studies
7.2-3 (2002): 289-312.
So, E. C. "A new approach to predicting analyst forecast errors: Do investors overweight analyst
forecasts?" Journal of Financial Economics 108.3 (2013): 615-640.
Soltes, Eugene. "Private Interaction Between Firm Management and Sell!Side
Analysts." Journal of Accounting Research 52.1 (2014): 245-272.
Stickel, Scott E. "Reputation and performance among security analysts." The Journal of Finance
47.5 (1992): 1811-1836.
Weber, D. P. "Do Analysts and Investors Fully Appreciate the Implications of Book!Tax
Differences for Future Earnings?" Contemporary Accounting Research 26.4 (2009):
1175-1206.
40
Figure 1: Timeline
365 days before last
CQE forecast
FQEinitial
EAt-1
SPTinitial
CQEfirst
CQEsecond
NonEarnings
Forecast
Signalst
EAt
Black text denotes revisions required for inclusion in the sample
Blue text denotes revisions not required for inclusion in the sample
Red text denotes date windows
41
Appendix A: Sample Selection
Revision Sample
Criteria
All analyst-firm-quarters in the I/B/E/S unadjusted detail file, subject to the following
constraints:
(1) The analyst issued a forecast of the prior quarter’s earnings.
(2) Actual earnings and the earnings announcement date are available for current quarter
earnings and prior quarter earnings.
(3) The firm's share price, outstanding shares, and the cumulative factor to adjust price at
the prior quarter earnings announcement are available on CRSP.
(4) The firm's cumulative factor to adjust price at the current quarter earnings
announcement is available on CRSP.
(5) The firm's fiscal quarter end date occurs during the period 1999Q1 to 2015Q2. I/B/E/S
began coverage of share price target forecasts in 1999 and I/B/E/S data was available
through 2015Q2 at the time of download.
Less: Analyst-firm-quarters without a current quarter earnings forecast between the prior
quarter earnings announcement date and the current quarter earnings announcement date.
Less: Analyst-firm-quarters without a share price target forecast either with the final
current quarter earnings forecast or in the 365 days before the final current quarter
earnings forecast date.
Less: Analyst-firm-quarters without a future quarter earnings forecast either with the final
current quarter earnings forecast or in the 365 days before the final current quarter
earnings forecast date.
Less: Analyst-firm-quarters with a stock split or a stock dividend between the prior quarter
earnings announcement date and the current quarter earnings announcement date.
Less: Analyst-firm-quarters without CRSP or Compustat data for control variables.
Text Sample
Criteria
All Revision Sample analyst-firm-quarters.
Less: Analyst-firm-quarters relating to analysts other than the 175 randomly selected
analysts. We collect analyst reports from ThomsonOne Banker for 175 randomly selected
analysts in the revision sample detailed above. We collect the analyst reports over three
iterations. In the first iteration, we collect all analyst reports for 15 randomly selected
analysts in the revision sample for the year 2010. In the second (third) iteration, we
randomly select 60 (100) analysts with at least twenty firm-quarters of observations for a
minimum of three firms in the revision sample.
Less: Analyst-firm-quarters for analysts we could not identify in ThomsonOne Banker.
We are able to identify 142 of the 175 analysts in ThomsonOne Banker.
Observations
2,293,778
917,272
359,227
149,856
13,648
134,402
719,373
Observations
719,373
(653,834)
(11,774)
53,765
42
Appendix B: Variable Definitions
Non-Earnings Forecast Signal Variables and Current Quarter Forecast Revision Variables
Name
Definition
NEG_SPT
Indicator variable that equals one if the analyst issues a negative
(POS_SPT)
(positive) revision to the share price target forecast at least two days
after the analyst’s final forecast of the current quarter’s earnings and
before the firm’s quarterly earnings announcement, zero otherwise.
REV_SPT
Indicator variable that equals one if NEG_SPT=1 or POS_SPT=1, zero
otherwise.
DEV_SPT
POS_SPT less NEG_SPT.
NEG_FQE
Indicator variable that equals one if the analyst issues a negative
(POS_FQE)
(positive) revision to a future quarter earnings forecast at least two days
after the analyst’s final forecast of the current quarter’s earnings and
before the firm’s quarterly earnings announcement, zero otherwise.
REV_FQE
Indicator variable that equals one if NEG_FQE=1 or POS_FQE=1, zero
otherwise.
DEV_FQE
POS_FQE less NEG_FQE.
NEG_TOT
Indicator variable that equals one if NEG_SPT=1 or NEG_FQE=1, zero
otherwise.
POS_TOT
Indicator variable that equals one if POS_SPT=1 or POS_FQE=1, zero
otherwise.
REV_TOT
Indicator variable that equals one if NEG_TOT=1 or POS_TOT=1, zero
otherwise.
NEG_CQE
Indicator variable that equals one if the analyst's final current quarter
(POS_CQE)
earnings forecast negatively (positively) revises a previous current
quarter earnings forecast made after the prior quarter’s earnings
announcement, zero otherwise.
REV_CQE
Indicator variable that equals one if NEG_CQE=1 or POS_CQE=1, zero
otherwise.
DEV_CQE
POS_CQE less NEG_CQE.
NEG_TOT_ABOVE
Indicator variable that equals one if NEG_TOT=1 and
ABOVE_CON=1, zero otherwise.
NEG_TOT_BELOW
Indicator variable that equals one if NEG_TOT=1 and
ABOVE_CON=0, zero otherwise.
POS_TOT_ABOVE
Indicator variable that equals one if POS_TOT=1 and
ABOVE_CON=1, zero otherwise.
POS_TOT_BELOW
Indicator variable that equals one if POS_TOT=1 and
ABOVE_CON=0, zero otherwise.
MISS_TEXT
Count variable that equals the number of times the analyst uses a
(BEAT_TEXT)
synonym for miss (beat) and a synonym for earnings expectations
within the same sentence. The synonyms for miss include miss, fall
short, below, lower, worse, downside, and underperform. The synonyms
for beat include beat, exceed, outperform, better, higher, above, top, and
upside. The synonyms for earnings expectations include earnings, EPS,
estimates, and expectations. Included reports must meet the sample
selection criteria detailed in Appendix A.
!
Data Source
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
ThomsonOne
Banker
!
43
Earnings and Return Variables
Name
Definition
MEET_BEAT
Indicator variable that equals one (zero) if the firm meets or beats
(misses) the analyst’s final forecast of the current quarter’s earnings.
MEET_BEAT_CON
Indicator variable that equals one (zero) if the firm meets or beats
(misses) the consensus earnings forecast as of the analyst's final forecast
of the current quarter's earnings.
MEET_BEAT_IN
Indicator variable that equals one if the firm meets or beats the analyst’s
first forecast of the current quarter’s earnings after the prior quarter's
earnings announcement, zero otherwise.
CURR_SURP
Current quarter earnings surprise, equal to the firm’s quarterly earnings
per share less the analyst’s final forecast of the current quarter's
earnings per share, scaled by price.
CURR_SURP_CON
Current quarter earnings surprise, equal to the firm’s quarterly earnings
per share less the consensus earnings forecast as of the analyst’s final
forecast of the current quarter's earnings per share, scaled by price.
CURR_SURP_IN
The initial current quarter earnings surprise, equal to the firm’s
quarterly earnings per share less the analyst’s first forecast of the
current quarter's earnings per share after the prior quarter’s earnings
announcement, scaled by price.
PREV_SURP
Previous quarter earnings surprise, equal to the firm’s quarterly earnings
per share less the analyst’s final forecast of the previous quarter's
earnings per share, scaled by price.
ABS(CURR_SURP_IN) Absolute value of CURR_SURP_IN.
ABS(PREV_SURP)
Absolute value of PREV_SURP.
RET_ANN
The firm’s returns compounded over the three day trading window
beginning the trading day before the earnings announcement date and
ending the trading day after the earnings announcement date less the
market return over the same period. For earnings announcements
occurring after 4 pm, we use the next trading day as the earnings
announcement date.
RET_QTR
The firm's returns compounded over the period beginning ninety days
before the final current quarter earnings forecast and ending at the final
current quarter earnings forecast date less the market return over the
same period.
RET_SPT
The firm’s returns compounded over the three day trading window
beginning the trading day before the share price target forecast revision
date and ending the trading day after the share price target forecast
revision date less the market return over the same period.
RET_FQE
The firm’s returns compounded over the three day trading window
beginning the trading day before the future quarter earnings forecast
revision date and ending the trading day after the future quarter earnings
forecast revision date less the market return over the same period.
RET_TEXT
The firm’s returns compounded over the three day trading window
beginning the trading day before the analyst's text report date and
ending the trading day after the analyst's text report date less the market
return over the same period.
!
!
Data Source
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S, CRSP
I/B/E/S, CRSP
I/B/E/S, CRSP
I/B/E/S, CRSP
I/B/E/S, CRSP
I/B/E/S, CRSP
CRSP
CRSP
CRSP
CRSP
CRSP
44
Other Variables
Name
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
ABOVE_CON
EXPERIENCE
BROKER_SIZE
BUY
ALLSTAR
Definition
Natural log of the market value of equity.
Book value of equity scaled by the market value of equity.
Earnings before interest and taxes scaled by total assets.
Natural log of analyst coverage.
Natural log of the number of days between the analyst's final forecast of
the current quarter's earnings and the current quarter's earnings
announcement.
Natural log of the number of times the analyst forecasted the previous
quarter's earnings.
Indicator variable that equals one (zero) if the analyst’s final forecast of
the current quarter's earnings is above (below) the consensus earnings
forecast.
Natural log of the number of years since the analyst's first forecast in
I/B/E/S.
Natural log of the number of analysts at the analyst's brokerage.
Indicator variable that equals one if the analyst's stock recommendation
is buy or strong buy, zero otherwise.
Indicator variable that equals one if the analyst finished first, second,
third, or honorable mention in the institutional investor all-star rankings
during the calendar year of the quarterly forecast.
Data Source
Compustat
Compustat
Compustat
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
I/B/E/S
Hand
Collected
45
Table 1: Descriptive Statistics
Panel A: Revision Sample Descriptive Statistics
Variable
N
Mean
StdDev
P25
Median
P75
NEG_SPT
719,373
0.048
0.213
0.000
0.000
0.000
POS_SPT
719,373
0.083
0.276
0.000
0.000
0.000
REV_SPT
719,373
0.131
0.337
0.000
0.000
0.000
NEG_FQE
719,373
0.056
0.230
0.000
0.000
0.000
POS_FQE
719,373
0.060
0.238
0.000
0.000
0.000
REV_FQE
719,373
0.117
0.321
0.000
0.000
0.000
NEG_TOT
719,373
0.089
0.284
0.000
0.000
0.000
POS_TOT
719,373
0.122
0.327
0.000
0.000
0.000
REV_TOT
719,373
0.195
0.396
0.000
0.000
0.000
NEG_CQE
719,373
0.191
0.393
0.000
0.000
0.000
POS_CQE
719,373
0.129
0.335
0.000
0.000
0.000
REV_CQE
719,373
0.320
0.466
0.000
0.000
1.000
MEET_BEAT
719,373
0.706
0.456
0.000
1.000
1.000
CURR_SURP
719,373
0.048
0.769
-0.045
0.052
0.211
MEET_BEAT_CON
687,458
0.642
0.479
0.000
1.000
1.000
CURR_SURP_CON
687,458
0.008
0.788
-0.074
0.045
0.200
PREV_SURP
719,373
0.017
0.824
-0.057
0.050
0.205
RET_QTR
719,373
0.002
0.191
-0.099
-0.002
0.096
MVE
719,373
7.908
1.721
6.655
7.837
9.117
BTM
719,373
0.494
0.368
0.243
0.412
0.659
ROA
719,373
0.019
0.034
0.007
0.020
0.035
FOLLOW
719,373
17.400
10.448
9.000
16.000
24.000
#DAYS
719,373
61.812
33.708
28.000
76.000
91.000
#FORE
719,373
4.964
3.567
2.000
4.000
7.000
This table reports descriptive statistics for the revision sample. N is the number of observations, StdDev is the
standard deviation, P25 (P75) is the 25th (75th) percentile of the variable's distribution. Sample selection
procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
46
Panel B: Text Sample Descriptive Statistics
Variable
N
Mean
StdDev
P25
Median
P75
MISS_TEXT
53,765
0.084
0.491
0.000
0.000
0.000
BEAT_TEXT
53,765
0.106
0.579
0.000
0.000
0.000
NEG_SPT
53,765
0.054
0.226
0.000
0.000
0.000
POS_SPT
53,765
0.089
0.285
0.000
0.000
0.000
NEG_FQE
53,765
0.061
0.240
0.000
0.000
0.000
POS_FQE
53,765
0.064
0.245
0.000
0.000
0.000
NEG_CQE
53,765
0.215
0.411
0.000
0.000
0.000
POS_CQE
53,765
0.153
0.360
0.000
0.000
0.000
MEET_BEAT
53,765
0.706
0.455
0.000
1.000
1.000
CURR_SURP
53,765
0.054
0.678
-0.041
0.054
0.210
MEET_BEAT_CON
52,165
0.644
0.479
0.000
1.000
1.000
CURR_SURP_CON
52,165
0.010
0.723
-0.068
0.046
0.197
PREV_SURP
53,765
0.025
0.720
-0.050
0.051
0.202
RET_QTR
53,765
0.003
0.174
-0.090
0.001
0.092
MVE
53,765
8.207
1.668
6.996
8.170
9.384
BTM
53,765
0.503
0.358
0.258
0.426
0.669
ROA
53,765
0.022
0.029
0.008
0.021
0.035
FOLLOW
53,765
18.802
10.671
10.000
17.000
25.000
#DAYS
53,765
57.829
34.508
22.000
67.000
90.000
#FORE
53,765
5.660
4.171
3.000
5.000
8.000
This table reports descriptive statistics for the text sample. N is the number of observations, StdDev is the
standard deviation, P25 (P75) is the 25th (75th) percentile of the variable's distribution. Sample selection
procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
47
NEG_SPT
POS_SPT
NEG_FQE
POS_FQE
PREV_SURP
RET_QTR
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
CONS
Table 2: Relation Between the Earnings Surprise and the Analyst's Non-Earnings Forecast Signals
Panel A: The Analyst's Own Earnings Surprise
MEET_BEAT
MEET_BEAT
MEET_BEAT
CURR_SURP
CURR_SURP
-0.017***
-0.010**
-0.009**
-0.038***
-0.020***
(0.001)
(0.028)
(0.017)
(0.000)
(0.007)
0.043***
0.026***
0.021***
0.042***
0.020***
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
0.000
-0.006*
-0.001
-0.007
-0.011***
(0.963)
(0.093)
(0.642)
(0.133)
(0.005)
0.018***
0.012***
0.015***
0.023***
0.016***
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
0.054***
0.045***
0.188***
(0.000)
(0.000)
(0.000)
0.114***
0.115***
0.158***
(0.000)
(0.000)
(0.000)
0.005**
0.019***
0.006**
(0.048)
(0.000)
(0.041)
-0.061***
-0.034***
-0.044*
(0.000)
(0.000)
(0.050)
0.625***
0.340***
-0.243*
(0.000)
(0.000)
(0.065)
0.051***
0.026***
0.022***
(0.000)
(0.000)
(0.001)
0.002
-0.008***
-0.006***
(0.228)
(0.000)
(0.009)
-0.019***
-0.010***
0.001
(0.000)
(0.000)
(0.730)
0.702***
0.565***
0.652***
0.045***
-0.019
(0.000)
(0.000)
(0.000)
(0.000)
(0.471)
CURR_SURP
-0.013**
(0.048)
0.014***
(0.000)
-0.006
(0.133)
0.015***
(0.000)
0.169***
(0.000)
0.146***
(0.000)
0.008**
(0.014)
-0.056***
(0.009)
-0.313**
(0.011)
0.015**
(0.041)
-0.008***
(0.001)
-0.001
(0.843)
0.079**
(0.020)
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
719,373
719,373
719,373
719,373
719,373
719,373
Adjusted R-squared
0.001
0.033
0.071
0.000
0.047
0.063
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings equal or exceed the
analyst's own earnings forecast (MEET_BEAT) or the signed earnings surprise relative to the analyst's own earnings forecast (CURR_SURP). The
explanatory variables include the analyst's share price target revisions (NEG_SPT, POS_SPT) and future quarter earnings forecast revisions
(NEG_FQE, POS_FQE). Robust standard errors are clustered by firm and quarter. P-values are reported in parentheses. ***, **, and * denote
statistical significance at the 1%, 5%, and 10% levels for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A.
Variable definitions are reported in Appendix B.
48
NEG_SPT
POS_SPT
NEG_FQE
POS_FQE
MEET_BEAT_CON
-0.045***
(0.000)
0.084***
(0.000)
-0.018***
(0.000)
0.034***
(0.000)
PREV_SURP
RET_QTR
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
CONS
0.636***
(0.000)
Panel B: The Consensus Earnings Surprise
MEET_BEAT_CON
MEET_BEAT_CON
CURR_SURP_CON
-0.037***
-0.024***
-0.057***
(0.000)
(0.000)
(0.000)
0.035***
0.029***
0.077***
(0.000)
(0.000)
(0.000)
-0.034***
-0.030***
-0.015**
(0.000)
(0.000)
(0.013)
0.014***
0.010***
0.036***
(0.000)
(0.001)
(0.000)
0.048***
0.040***
(0.000)
(0.000)
0.330***
0.319***
(0.000)
(0.000)
0.015***
0.026***
(0.000)
(0.000)
-0.058***
-0.036***
(0.000)
(0.000)
0.386***
0.239***
(0.000)
(0.002)
0.041***
0.008
(0.000)
(0.107)
0.040***
0.031***
(0.000)
(0.000)
-0.003
-0.003
(0.432)
(0.266)
0.281***
0.301***
0.003
(0.000)
(0.000)
(0.817)
CURR_SURP_CON
-0.034***
(0.001)
0.022***
(0.000)
-0.026***
(0.000)
0.018***
(0.000)
0.176***
(0.000)
0.419***
(0.000)
0.015***
(0.000)
-0.119***
(0.000)
-0.221
(0.128)
0.028***
(0.001)
0.023***
(0.000)
0.004
(0.519)
-0.231***
(0.000)
CURR_SURP_CON
-0.013
(0.150)
0.012***
(0.008)
-0.022***
(0.000)
0.010**
(0.034)
0.159***
(0.000)
0.395***
(0.000)
0.020***
(0.000)
-0.125***
(0.000)
-0.288**
(0.046)
0.010
(0.274)
0.023***
(0.000)
-0.004
(0.335)
-0.203***
(0.000)
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
687,458
687,458
687,458
687,458
687,458
687,458
Adjusted R-squared
0.004
0.050
0.085
0.001
0.057
0.079
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings equal or exceed the consensus earnings
forecast (MEET_BEAT_CON) or the signed earnings surprise relative to the consensus earnings forecast (CURR_SURP_CON). The explanatory variables include the
analyst's share price target revisions (NEG_SPT, POS_SPT) and future quarter earnings forecast revisions (NEG_FQE, POS_FQE). Robust standard errors are clustered by
firm and quarter. P-values are reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels for two-tailed tests, respectively. Sample
selection procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
49
Table 3: Relation Between the Analyst's Own Earnings Surprise and the Analyst's Qualitative Predictions
Panel A: The Analyst's Own Earnings Surprise
MEET_BEAT
MEET_BEAT
MEET_BEAT
CURR_SURP
CURR_SURP
CURR_SURP
MISS_TEXT
-0.048***
-0.048***
-0.048***
-0.034***
-0.036***
-0.033***
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
BEAT_TEXT
0.050***
0.045***
0.043***
0.043***
0.037***
0.036***
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
PREV_SURP
0.058***
0.052***
0.174***
0.165***
(0.000)
(0.000)
(0.000)
(0.000)
RET_QTR
0.122***
0.123***
0.189***
0.170***
(0.000)
(0.000)
(0.000)
(0.000)
MVE
0.009***
0.018***
0.006*
0.004
(0.004)
(0.000)
(0.082)
(0.354)
BTM
-0.063***
-0.029***
-0.020
-0.030
(0.000)
(0.002)
(0.422)
(0.289)
ROA
0.398***
0.293**
-0.105
0.055
(0.005)
(0.044)
(0.671)
(0.837)
FOLLOW
0.046***
0.026***
0.029***
0.019
(0.000)
(0.002)
(0.004)
(0.212)
#DAYS
0.005
-0.002
-0.003
-0.001
(0.162)
(0.543)
(0.422)
(0.734)
#FORE
-0.017***
-0.006
0.006
0.006
(0.001)
(0.147)
(0.444)
(0.416)
CONS
0.705***
0.534***
0.603***
0.052***
-0.068*
-0.011
(0.000)
(0.000)
(0.000)
(0.000)
(0.083)
(0.794)
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
53,765
53,765
53,765
53,765
53,765
53,765
Adjusted R-squared
0.004
0.031
0.059
0.001
0.042
0.053
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings equal or
exceed the analyst's own earnings forecast (MEET_BEAT) or the signed earnings surprise relative to the analyst's own earnings forecast
(CURR_SURP). The explanatory variables include the analyst's qualitative predictions (MISS_TEXT, BEAT_TEXT). Robust standard
errors are clustered by firm and quarter. P-values are reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%,
and 10% levels for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A. Variable definitions are
reported in Appendix B.
50
MISS_TEXT
BEAT_TEXT
MEET_BEAT_CON
-0.055***
(0.000)
0.061***
(0.000)
PREV_SURP
RET_QTR
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
CONS
0.642***
(0.000)
Panel B: The Consensus Earnings Surprise
MEET_BEAT_CON
MEET_BEAT_CON
CURR_SURP_CON
-0.053***
-0.053***
-0.048***
(0.000)
(0.000)
(0.000)
0.052***
0.048***
0.053***
(0.000)
(0.000)
(0.000)
0.053***
0.046***
(0.000)
(0.000)
0.362***
0.351***
(0.000)
(0.000)
0.013***
0.023***
(0.000)
(0.000)
-0.072***
-0.042***
(0.000)
(0.001)
0.113
0.058
(0.451)
(0.696)
0.051***
0.020**
(0.000)
(0.032)
0.033***
0.027***
(0.000)
(0.000)
0.002
0.004
(0.741)
(0.405)
0.291***
0.238***
0.008
(0.000)
(0.000)
(0.534)
CURR_SURP_CON
-0.045***
(0.000)
0.042***
(0.000)
0.163***
(0.000)
0.469***
(0.000)
0.015***
(0.000)
-0.103***
(0.003)
-0.040
(0.869)
0.047***
(0.000)
0.020***
(0.000)
-0.003
(0.823)
-0.274***
(0.000)
CURR_SURP_CON
-0.040***
(0.000)
0.039***
(0.000)
0.153***
(0.000)
0.434***
(0.000)
0.017***
(0.001)
-0.111***
(0.003)
0.047
(0.862)
0.016
(0.330)
0.025***
(0.000)
-0.003
(0.763)
-0.336***
(0.000)
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
52,165
52,165
52,165
52,165
52,165
52,165
Adjusted R-squared
0.006
0.050
0.076
0.002
0.054
0.077
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings equal or exceed the consensus earnings
forecast (MEET_BEAT_CON) or the signed earnings surprise relative to the consensus earnings forecast (CURR_SURP_CON). The explanatory variables include the
analyst's qualitative predictions (MISS_TEXT, BEAT_TEXT). Robust standard errors are clustered by firm and quarter. P-values are reported in parentheses. ***, **, and *
denote statistical significance at the 1%, 5%, and 10% levels for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A. Variable definitions
are reported in Appendix B.
51
Table 4: Cross-Sectional Variation in Non-Earnings Forecast Signals
Panel A: Positive vs. Negative Non-Earnings Forecast Signals
Sample
Variable
N
% Positive
% Negative
Difference
t-stat
Full
REV_TOT
719,373
12.20%
8.90%
3.30%
66.982
Full
REV_SPT
719,373
8.30%
4.80%
3.50%
83.183
Full
REV_FQE
719,373
6.00%
5.60%
0.40%
10.572
Full
REV_CQE
719,373
12.90%
19.10%
-6.20%
-92.648
REV_CQE=0
REV_TOT
489,400
14.60%
9.90%
4.70%
71.819
REV_CQE=0
REV_SPT
489,400
10.30%
5.50%
4.80%
85.931
REV_CQE=0
REV_FQE
489,400
6.90%
6.10%
0.80%
14.755
REV_CQE=0
REV_CQE
489,400
0.00%
0.00%
0.00%
N/A
This table reports the frequencies of positive vs. negative share price target revisions, future quarter earnings forecast revisions, and
current quarter earnings forecast revisions (SPT, FQE, and CQE, respectively). TOT refers to either an SPT or FQE revision. Rows
1-4 include the full revision sample. Rows 5-8 include all quarters without a current quarter earnings forecast revision
(REV_CQE=0). N is the number of observations. % is the percentage of observations for which the analyst issues the
corresponding forecast revision. Two-tailed t-statistics report whether the corresponding difference is significantly different from
zero. Sample selection procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
Panel B: Above vs. Below Consensus
N Below
N Above
% Below
% Above
Sample
Variable
Consensus
Consensus
Consensus
Consensus
Difference
t-stat
Full
NEG_TOT
343,339
344,119
9.40%
8.30%
1.10%
15.418
Full
POS_TOT
343,339
344,119
11.10%
13.20%
-2.10%
-26.217
Full
NEG_SPT
343,339
344,119
5.10%
4.30%
0.80%
16.006
Full
POS_SPT
343,339
344,119
7.40%
9.20%
-1.80%
-26.893
Full
NEG_FQE
343,339
344,119
6.00%
5.40%
0.60%
11.434
Full
POS_FQE
343,339
344,119
5.60%
6.50%
-0.90%
-15.311
Full
NEG_CQE
316,651
358,861
14.70%
23.10%
-8.40%
-88.582
Full
POS_CQE
316,651
358,861
17.40%
9.40%
8.00%
97.655
This table reports the frequencies of positive and negative share price target revisions, future quarter earnings forecast revisions,
and current quarter earnings forecast revisions (SPT, FQE, and CQE, respectively) for observations above and below the consensus
earnings forecast. TOT refers to either an SPT or FQE revision. In rows 1-6 the consensus is as of the analyst's final earnings
forecast during the quarter and in rows 7-8 the consensus is as of the analyst's initial earnings forecast during the quarter. N is the
number of observations. % is the percentage of observations for which the analyst issues the corresponding forecast revision. Twotailed t-statistics report whether the corresponding difference is significantly different from zero. Sample selection procedures are
reported in Appendix A. Variable definitions are reported in Appendix B.
52
Panel C: Above vs. Below Median of the Analyst's Initial Earnings Surprise
N Below Median
N Above Median
% Below Median
% Above Median
Variable
ABS(CURR_SURP_IN)
ABS(CURR_SURP_IN)
ABS(CURR_SURP_IN)
ABS(CURR_SURP_IN)
Difference
t-stat
REV_TOT
359,692
359,681
21.10%
17.80%
3.30%
36.062
REV_SPT
359,692
359,681
14.30%
11.80%
2.50%
31.334
REV_FQE
359,692
359,681
12.60%
10.70%
1.90%
26.050
REV_CQE
359,692
359,681
25.40%
38.60%
-13.20% -121.473
This table reports the frequencies of share price target revisions, future quarter earnings forecast revisions, and current quarter earnings forecast revisions (SPT, FQE,
and CQE, respectively) for observations above and below the median of the absolute value of the analyst's initial earnings surprise (as of the analyst's initial earnings
forecast during the quarter). TOT refers to either an SPT or FQE revision. N is the number of observations. % is the percentage of observations for which the analyst
issues the corresponding forecast revision. Two-tailed t-statistics report whether the corresponding difference is significantly different from zero. Sample selection
procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
53
Table 5: Determinants of Non-Earnings Forecast Signals
REV_TOT
REV_CQE
NEG_TOT
ABOVE_CON
0.001
-0.043***
-0.009***
(0.352)
(-10.193)
(-6.300)
ABS(CURR_SURP_IN)
-0.001*
0.048***
0.001
(-1.678)
(16.326)
(0.690)
ABS(PREV_SURP)
-0.005***
-0.019***
-0.001
(-6.058)
(-9.976)
(-1.588)
EXPERIENCE
-0.020***
-0.022***
-0.011***
(-16.932)
(-10.076)
(-11.697)
BROKER_SIZE
0.020***
0.009***
0.010***
(15.615)
(8.907)
(11.650)
ALLSTAR
0.014***
0.005
0.007*
(2.900)
(1.163)
(1.782)
ALLSTAR_DUM
-0.027***
-0.020**
-0.017*
(-3.222)
(-2.344)
(-1.804)
PREV_SURP
0.002**
0.003***
0.001*
(2.105)
(3.093)
(1.744)
RET_QTR
0.031***
-0.044***
-0.103***
(3.055)
(-3.875)
(-10.656)
MVE
0.010***
0.006***
0.004***
(7.748)
(3.390)
(4.767)
BTM
0.002
0.028***
0.000
(0.582)
(5.640)
(0.133)
ROA
0.024
0.338***
0.005
(0.656)
(6.137)
(0.169)
FOLLOW
0.020***
0.022***
0.013***
(6.648)
(4.240)
(5.790)
#DAYS
0.109***
-0.317***
0.047***
(35.898)
(-139.810)
(16.035)
#FORE
0.049***
0.066***
0.023***
(15.028)
(18.426)
(9.071)
CONS
-0.466***
1.342***
-0.204***
(-24.283)
(73.103)
(-12.835)
POS_TOT
0.010***
(9.394)
-0.002***
(-3.331)
-0.004***
(-5.676)
-0.013***
(-11.513)
0.013***
(12.000)
0.005
(1.334)
-0.012
(-1.546)
0.001
(1.350)
0.135***
(16.322)
0.007***
(6.746)
0.001
(0.194)
-0.005
(-0.196)
0.010***
(4.050)
0.073***
(21.989)
0.034***
(12.268)
-0.321***
(-16.817)
Observations
680,172
680,172
680,172
680,172
Adjusted R-squared
0.066
0.411
0.030
0.051
This table reports OLS regression results. The dependent variables include: a dummy variable set equal to one if the analyst revises
a share price target or a future quarter earnings forecast (REV_TOT), a dummy variable set equal to one if the analyst revises the
current quarter earnings forecast (REV_CQE), and dummy variables set equal to one for negative and positive revisions to a share
price target forecast or a future quarter earnings forecast (NEG_TOT and POS_TOT, respectively). Robust standard errors are
clustered by firm and quarter. P-values are reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and
10% levels for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A. Variable definitions are
reported in Appendix B.
!
!
54
Table 6: Relation Between the Earnings Surprise and the Analyst's Non-Earnings Forecast Signals
Panel A: Positive vs. Negative News
MEET_BEAT MEET_BEAT MEET_BEAT CURR_SURP CURR_SURP CURR_SURP
NEG_TOT
-0.009**
-0.009**
-0.005*
-0.026***
-0.017***
-0.010**
(0.024)
(0.029)
(0.060)
(0.000)
(0.001)
(0.016)
POS_TOT
0.040***
0.024***
0.022***
0.042***
0.023***
0.018***
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
PREV_SURP
0.054***
0.045***
0.188***
0.169***
(0.000)
(0.000)
(0.000)
(0.000)
RET_QTR
0.115***
0.116***
0.158***
0.147***
(0.000)
(0.000)
(0.000)
(0.000)
MVE
0.005**
0.019***
0.006**
0.008**
(0.048)
(0.000)
(0.041)
(0.014)
BTM
-0.061***
-0.034***
-0.044**
-0.056***
(0.000)
(0.000)
(0.050)
(0.009)
ROA
0.626***
0.341***
-0.243*
-0.313**
(0.000)
(0.000)
(0.065)
(0.011)
FOLLOW
0.051***
0.026***
0.022***
0.015**
(0.000)
(0.000)
(0.001)
(0.042)
#DAYS
0.002
-0.008***
-0.006***
-0.008***
(0.232)
(0.000)
(0.008)
(0.001)
#FORE
-0.019***
-0.010***
0.002
-0.001
(0.000)
(0.000)
(0.717)
(0.849)
CONS
0.702***
0.566***
0.652***
0.045***
-0.019
0.079**
(0.000)
(0.000)
(0.000)
(0.000)
(0.474)
(0.019)
F-test
p=0.000
p=0.001
p=0.000
p=0.006
p=0.315
p=0.137
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
719,373
719,373
719,373
719,373
719,373
719,373
Adjusted R-squared
0.001
0.033
0.071
0.000
0.047
0.063
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings
equal or exceed the analyst's own earnings forecast (MEET_BEAT) or the signed earnings surprise relative to the analyst's
own earnings forecast (CURR_SURP). The explanatory variables include negative/positive revisions to the analyst's share
price target or future quarter earnings forecast (NEG_TOT, POS_TOT). P-values are reported to test whether the sum of the
coefficients on NEG_TOT and POS_TOT are significantly different from zero. Robust standard errors are clustered by firm
and quarter. P-values are reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and 10%
levels for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A. Variable definitions are
reported in Appendix B.
55
NEG_TOT_ABOVE
NEG_TOT_BELOW
POS_TOT_ABOVE
POS_TOT_BELOW
ABOVE_CON
MEET_BEAT
-0.012**
(0.030)
-0.016***
(0.000)
0.048***
(0.000)
0.043***
(0.000)
-0.137***
(0.000)
PREV_SURP
RET_QTR
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
CONS
0.773***
(0.000)
NEG_TOT F-test
POS_TOT F-test
p=0.467
p=0.200
Panel B: Above vs. Below the Consensus
MEET_BEAT MEET_BEAT CURR_SURP
-0.009*
-0.004
-0.016**
(0.083)
(0.296)
(0.017)
-0.016***
-0.013***
-0.043***
(0.000)
(0.000)
(0.000)
0.030***
0.027***
0.060***
(0.000)
(0.000)
(0.000)
0.024***
0.021***
0.034***
(0.000)
(0.000)
(0.000)
-0.142***
-0.146***
-0.166***
(0.000)
(0.000)
(0.000)
0.053***
0.044***
(0.000)
(0.000)
0.156***
0.156***
(0.000)
(0.000)
0.005**
0.020***
(0.048)
(0.000)
-0.065***
-0.035***
(0.000)
(0.000)
0.626***
0.327***
(0.000)
(0.000)
0.047***
0.018***
(0.000)
(0.000)
0.006***
-0.004***
(0.001)
(0.004)
-0.018***
-0.010***
(0.000)
(0.000)
0.631***
0.702***
0.131***
(0.000)
(0.000)
(0.000)
CURR_SURP
-0.006
(0.348)
-0.035***
(0.000)
0.039***
(0.000)
0.011**
(0.014)
-0.171***
(0.000)
0.186***
(0.000)
0.202***
(0.000)
0.005*
(0.098)
-0.043*
(0.052)
-0.266**
(0.047)
0.017**
(0.020)
-0.001
(0.629)
0.002
(0.619)
0.070***
(0.009)
CURR_SURP
0.001
(0.824)
-0.027***
(0.000)
0.034***
(0.000)
0.004
(0.229)
-0.176***
(0.000)
0.168***
(0.000)
0.189***
(0.000)
0.008**
(0.016)
-0.053**
(0.012)
-0.348***
(0.007)
0.006
(0.430)
-0.003
(0.165)
-0.000
(0.967)
0.151***
(0.000)
p=0.133
p=0.062
p=0.059
p=0.000
p=0.000
p=0.000
p=0.065
p=0.001
p=0.000
p=0.000
Analyst &
Analyst &
Fixed Effects
None
None
Quarter
None
None
Quarter
Observations
687,458
687,458
687,458
687,458
687,458
687,458
Adjusted R-squared
0.023
0.055
0.095
0.012
0.058
0.075
This table reports OLS regression results. The dependent variable is a dummy set equal to one if the firm’s reported earnings equal
or exceed the analyst's own earnings forecast (MEET_BEAT) or the signed earnings surprise relative to the analyst's own earnings
forecast (CURR_SURP). The explanatory variables include negative/positive revisions to the analyst's share price target or future
quarter earnings forecast apportioned by whether the analyst's final earnings forecast is above or below the consensus earnings
forecast (NEG_TOT_ABOVE, NEG_TOT_BELOW, POS_TOT_ABOVE, POS_TOT_BELOW). P-values are reported to test
whether the coefficients on NEG_TOT_ABOVE and NEG_TOT_BELOW are significantly different, as well as whether the
coefficients on POS_TOT_ABOVE and POS_TOT_BELOW are significantly different. Robust standard errors are clustered by
firm and quarter. P-values are reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels
for two-tailed tests, respectively. Sample selection procedures are reported in Appendix A. Variable definitions are reported in
Appendix B.
!
!
56
!
BUY
CONS
Table 7: Analysts' Stock Recommendations and Non-Earnings Forecast Signals
MEET_BEAT
MEET_BEAT_IN
DEV_CQE
DEV_SPT
0.019***
0.013***
-0.004
0.006**
(5.505)
(3.415)
(-0.931)
(2.306)
0.695***
0.645***
-0.057***
0.031***
(123.354)
(87.722)
(-5.580)
(3.979)
Fixed Effects
Observations
Adjusted R-squared
BUY
PREV_SURP
RET_QTR
MVE
BTM
ROA
FOLLOW
#DAYS
#FORE
CONS
None
582,626
0.000
MEET_BEAT
0.009***
(4.444)
0.044***
(17.318)
0.114***
(16.639)
0.018***
(8.251)
-0.036***
(-5.153)
0.371***
(5.741)
0.026***
(6.443)
-0.007***
(-4.287)
-0.010***
(-4.914)
0.632***
(34.136)
None
582,626
0.000
MEET_BEAT_IN
-0.001
(-0.387)
0.042***
(15.068)
0.288***
(22.430)
0.021***
(8.918)
-0.044***
(-5.866)
0.279***
(4.230)
0.014***
(2.950)
0.049***
(17.907)
-0.010***
(-4.809)
0.291***
(17.249)
None
582,626
0.000
DEV_CQE
-0.019***
(-8.195)
0.008***
(3.650)
0.492***
(23.289)
0.008***
(2.918)
-0.035***
(-5.044)
-0.122*
(-1.857)
-0.008*
(-1.770)
0.051***
(5.782)
0.004
(1.079)
-0.350***
(-14.594)
DEV_FQE
-0.001
(-0.520)
0.006
(1.583)
None
582,626
0.000
None
582,626
-0.000
DEV_SPT
0.008***
(3.266)
0.001
(1.180)
0.204***
(17.361)
0.003***
(3.007)
-0.001
(-0.264)
0.095***
(3.954)
-0.002
(-1.414)
0.026***
(4.261)
0.001
(0.791)
-0.090***
(-3.800)
DEV_FQE
-0.001
(-0.915)
-0.001
(-1.438)
0.064***
(12.559)
-0.001
(-1.041)
-0.004*
(-1.833)
-0.027
(-1.007)
-0.002
(-1.005)
0.003
(1.214)
-0.003**
(-2.161)
0.041***
(3.683)
Analyst &
Analyst &
Analyst &
Analyst &
Analyst &
Fixed Effects
Quarter
Quarter
Quarter
Quarter
Quarter
Observations
582,626
582,626
582,626
582,626
582,626
Adjusted R-squared
0.072
0.089
0.068
0.053
0.009
This table reports OLS regression results. The dependent variables include: a dummy set equal to one if the firm’s
reported earnings equal or exceed the analyst's final earnings forecast (MEET_BEAT), a dummy set equal to one
if the firm’s reported earnings equal or exceed the analyst's initial earnings forecast (MEET_BEAT_IN), the
difference between POS_CQE and NEG_CQE (DEV_CQE), the difference between POS_SPT and NEG_SPT
(DEV_SPT), and the difference between POS_FQE and NEG_FQE (DEV_FQE). The explanatory variables
include a dummy variable set equal to one if the analyst's stock recommendation is buy or strong buy (BUY).
Robust standard errors are clustered by firm and quarter. P-values are reported in parentheses. ***, **, and *
denote statistical significance at the 1%, 5%, and 10% levels for two-tailed tests, respectively. Sample selection
procedures are reported in Appendix A. Variable definitions are reported in Appendix B.
57
Table 8: Relation Between Earnings Announcement Returns and the Analyst's Non-Earnings Forecast Signals
RET_ANN
RET_ANN
RET_ANN
RET_ANN
NEG_SPT
0.000
0.000
(0.778)
(0.653)
POS_SPT
0.002***
0.002***
(0.006)
(0.001)
NEG_FQE
0.001*
0.001
(0.087)
(0.130)
POS_FQE
0.001
0.001
(0.186)
(0.195)
MISS_TEXT
-0.007***
-0.007***
(0.000)
(0.000)
BEAT_TEXT
0.005***
0.005***
(0.000)
(0.000)
PREV_SURP
-0.001***
-0.002***
(0.000)
(0.000)
RET_QTR
0.001
0.002
(0.846)
(0.638)
MVE
0.000
0.000
(0.776)
(0.720)
BTM
0.001
-0.001
(0.293)
(0.635)
ROA
0.059***
0.035*
(0.000)
(0.053)
FOLLOW
-0.000
0.000
(0.911)
(0.920)
#DAYS
-0.000
0.000
(0.660)
(0.985)
#FORE
-0.000
0.000
(0.619)
(0.697)
CONS
0.001
-0.001
0.001*
-0.001
(0.132)
(0.823)
(0.071)
(0.714)
Sample
Full
Full
Full
Full
Observations
716,566
716,566
53,581
53,581
Adjusted R-squared
0.000
0.001
0.002
0.003
This table reports OLS regression results. The dependent variable is the earnings announcement return (RET_ANN). The
explanatory variables include the analyst's share price target revisions (NEG_SPT, POS_SPT) and the analyst's future quarter
earnings forecast revisions (NEG_FQE, POS_FQE), or the analyst's qualitative predictions (MISS_TEXT, BEAT_TEXT).
Robust standard errors are clustered by firm and quarter. P-values are reported in parentheses. ***, **, and * denote statistical
significance at the 1%, 5%, and 10% levels for two-tailed tests, respectively. Sample selection procedures are reported in
Appendix A. Variable definitions are reported in Appendix B.
58
Table 9: Non-Earnings Forecast Signal Announcement Returns
Sample
Variable
N
Mean
t-stat
NEG_SPT=1
RET_SPT
33,490
-0.014
-39.122
POS_SPT=1
RET_SPT
58,028
0.015
75.439
NEG_FQE=1
RET_FQE
39,371
-0.005
-17.833
POS_FQE=1
RET_FQE
42,472
0.006
28.423
MISS_TEXT=1
RET_TEXT
2,429
-0.006
-4.604
BEAT_TEXT=1
RET_TEXT
2,927
0.009
8.238
This table reports mean announcement returns at the issuance of a share price target revision, a future quarter earnings forecast
revision, and a qualitative prediction (RET_SPT, RET_FQE, RET_TEXT, respectively). Row 1 (2) reports mean returns for
negative (positive) share price target revisions. Row 3 (4) reports mean returns for negative (positive) future quarter earnings
forecast revisions. Row 5 (6) reports mean returns for negative (positive) qualitative predictions. Two-tailed t-statistics are
reported to test whether the mean returns are significantly different from zero. Sample selection procedures are reported in
Appendix A. Variable definitions are reported in Appendix B.
59