relative performance ranking: insights into competitive advantage · 2015-04-22 · relative...
TRANSCRIPT
Relative Performance Ranking: Insights into Competitive Advantage
Jared Jennings
Olin School of Business
Washington University
Hojun Seo
Olin School of Business
Washington University
Mark T. Soliman*
Marshall School of Business
University of Southern California
April 2015
*Corresponding author. We thank Alex Edwards, Richard Frankel, Miguel Minutti-Meza, Bob Holthausen, Dan
Taylor, John Donovan, Raffi Indjejikian, Pat O’Brien, Tatiana Sandino, Haresh Sapra, Regina Wittenberg, and
Franco Wong for helpful comments. We also thank workshop participants at the 8th annual Rotman Accounting
Research Conference at the University of Toronto, University of Pennsylvania, and Washington University in St.
Louis. We are grateful for financial support from the Olin Business School and the Leventhal School of Accounting.
Relative Performance Ranking: Insights into Competitive Advantage
Abstract:
We examine how changes to the firm’s competitive advantage are manifested in the firm’s
financial statements and how market participants react to these changes. We argue that changes
to the firm’s competitive advantage are manifested through changes to the firm’s intra-industry
performance ranking (measured using return on assets). We find that changes to the firm’s
performance ranking are positively associated with earnings persistence, consistent with the
sustainability of profits increasing when the firm’s competitive advantage improves. We then
provide evidence consistent with investors and analysts positively valuing improvements to the
firm’s performance ranking, especially when the firm’s ranking has been stable in the recent
past. In fact, our evidence suggests that investors and analysts react more strongly to a change to
the firm’s performance ranking than the firm’s earnings surprise. Our results suggest that the
firm’s performance ranking within the industry constitutes an additional and relevant
performance benchmark for investors and managers that has not been explored by prior research.
This evidence also suggests that investors use the entire distribution of earnings to evaluate a
firm’s performance and not just analyst expectations or the prior period’s performance.
Keywords: performance ranking, earnings, benchmarking
JEL classification: M40; M41
1
1. Introduction
In this paper we examine how changes to the firm’s competitive advantage are
manifested in the firm’s financial statements and how market participants react to these changes.
Saloner, Shepard, and Podolny (2001) suggest that the firm’s competitive advantage is
manifested by the firm’s ability to create and capture value. They suggest that “a firm creates
value only where there is a difference between what customers will pay for its products or
services and what the firm must pay to provide them.” Corporate managers pursue various
strategic activities to establish competitive advantages in order to differentiate themselves from
their rivals. We measure a change to the firm’s competitive advantage as a change to the firm’s
performance relative to that of other firms in the same industry.1 We first examine whether
changes to the firm’s performance ranking within the industry are positively associated with
more persistent earnings, reflecting the firm’s ability to sustain future shareholder profits.
Second, we examine how investors and analysts value and respond to a change to the firm’s
performance ranking.
A firm’s performance can be decomposed into a firm-specific and non-firm-specific
component (Waring, 1996; Rumelt 1991; McGahan and Porter 2002). The non-firm-specific
component of the firm’s performance is influenced by industry or market level shocks that affect
all firms and may be useful in evaluating the overall health of the industry or market. However,
the component of performance that is largely informative to the firm’s individual performance is
the firm-specific component, which reflects the firm’s intra-industry heterogeneity in
performance and the firm’s ability to capture value within the industry (Rumelt, 1991; Nelson,
1991; McGahan and Porter, 2002). As a result, the firm-specific component of performance is
1 In our primary empirical tests, we measure the firm’s performance using return on assets (ROA) because it is more
likely to identify changes to the competitive advantage of the firm that are a result of both cost and quality
advantages. Cost and quality advantages are two mechanisms by which firms can create and capture value.
2
useful in assessing the competitive advantages held by the firm.2 We measure the firm-specific
component of performance by ranking the firm’s performance within its industry, allowing us to
remove the effect of industry-specific and market-specific shocks from the firm’s performance.
Thus, a change to the relative performance ranking of the firm within its industry provides
market participants with a metric to assess the firm’s competitive advantage.
Using a sample of 198,467 firm/quarter observations from 1997 to 2013, we first
examine whether changes to the firm’s performance ranking (based on return on assets) within
the industry result in more persistent earnings.3 If the competitive advantage of the firm is not
readily substituted or imitated by rivals, then a change to the firm’s competitive advantage
should reflect a more sustainable level of future shareholder profits (e.g., Peteraf, 1993; Saloner
et al., 2001). To calculate changes in the firm’s performance ranking, we compare the initial
ranking of the firm’s performance within its industry, based on expected earnings at the
beginning of the period, to the realized ranking of the firm’s performance, based on realized
earnings released on the date of earnings announcement.4 Consistent with expectations, we find
that a change to the firm’s performance ranking in the current period is positively associated with
more persistence earnings in the future. We specifically find that the firm’s earnings persistence
increases by 22.77% when moving from the 1st to 10th decile for the change in performance
ranking variable.
2 These concepts are found in other areas of the literature and are not new. For example, the CAPM argues that only
firm-specific risk should be priced and all sources of risk can be diversified away and the compensation literature
argues that only firm-specific performance, controllable by the manager, should be compensated. 3 We examine the correlations between changes in performance ranking measured using return on assets, return on
equity, and profit margin. The correlation between changes in performance ranking based on return on assets and
that of return on equity is 0.907 and the correlation between changes in performance ranking based on return on
assets and that of profit margin is 0.836. These correlations suggest that we are picking up a similar performance
ranking when using these three performance measures. 4 When we examine the persistence of earnings, the change in the firm’s performance ranking is specifically defined
as the change in the firm’s performance ranking based on IBES-reported Actual EPS in quarter t compared to the
firm’s performance ranking based on expected earnings at the beginning of quarter t. Expected earnings are based on
analyst expectation measured at the beginning of quarter t. We discuss the construction of this variable more in
depth in Section 3.
3
After establishing that changes to the firm’s performance ranking within the industry are
associated with more persistent earnings, we examine how investors and analysts react to short-
window unexpected changes in the firm’s performance ranking within the industry.5 Consistent
with expectations, we find that a change in the firm’s performance ranking is positively
associated with buy-and-hold abnormal returns during the three-day period surrounding the
earnings announcement, after controlling for the firm’s earnings surprise and other firm
characteristics.6 In fact, surprisingly, the positive market reaction to an increase in the firm’s
performance ranking within the industry is greater than the market’s reaction to a similar change
in the firm’s earnings surprise after standardizing the coefficients. We find that one standard
deviation increase in the firm’s performance ranking is associated with a 0.187 standard
deviation increase in returns, while one standard deviation increase in the firm’s earnings
surprise is only associated with a 0.038 standard deviation increase in returns.7 This evidence
suggests that investors use the firms’ relative performance ranking within the industry to assess
the firm’s ability to generate future profits for shareholders.
We also examine how sell-side analysts respond to short-window changes in the firm’s
performance ranking within the industry. Consistent with expectations, we find that analysts
5 When we examine the investor and analyst reactions to a change in the firm’s performance ranking, we examine
the market and analyst reactions to short window performance ranking changes. The initial performance ranking
measured based on expected earnings two-days prior to the earnings announcement. If peer firms have already
announced earnings, we use the announced earnings when determining the performance ranking of the firm;
otherwise, we use analyst expectations. We discuss the construction of this variable more in depth in Section 3. 6 While we could examine the change in the firm’s performance ranking at various points during the fiscal period,
we choose the earnings announcement date for two reasons. First, the earnings announcement is a significant firm
event that reveals information about the firm, prompting investors to evaluate the firm’s performance, increasing the
power of our tests. Second, management announces earnings during the earnings announcement, allowing us to
evaluate the market’s response to the change in the firm’s performance ranking relative to the market’s response to
the earnings surprise. This allows us to evaluate the incremental informativeness of the change in the firm’s
performance ranking conveyed by the release of earnings. 7 Standardizing coefficients are used to compare the relative strength of each independent variable included in the
regression (e.g., Maddala, 1977; Palmrose, 1986; Waring, 1996). Because standardized coefficients are measured in
terms of standard deviations and not the underlying units for each independent variable, we can compare the relative
strength of each standardized coefficient across independent variables. Standardized coefficients are calculated by
multiplying the coefficient by the standard deviation of the independent variable and dividing by the standard
deviation of the dependent variable.
4
positively revise their forecasts in response to an increase in the firm’s performance ranking and
that this response is more sensitive than analysts’ response to the earnings surprise. We find that
one standard deviation increase in the performance ranking variable equates to a 0.139 standard
deviation increase in analyst forecast revisions, while a one standard deviation increase in the
earnings surprise equates to a 0.066 standard deviation increase in analyst forecast revisions. We
find similar results when examining how a change in the firm’s performance ranking impacts
changes to analyst stock recommendations. Interestingly, the earnings surprise has no significant
effect on the change in analyst stock recommendations after including the change in the firm’s
performance ranking in the regression. This evidence is consistent with analysts positively
valuing changes to the competitive advantage of the firm.
In additional cross-sectional results, we examine whether the stability of the firm’s past
performance ranking influences the investors’ and analysts’ reaction to a change in the
performance ranking of the firm in the current period. If the firm-specific component of
performance has been stable (volatile) in the recent past, a change in the firm’s performance
ranking in the current period is more (less) likely to provide an informative signal about the
change in the expected performance or competitiveness of the firm within the industry.
Consistent with our prediction, we find that the reaction of investors and analysts to a change in
the firm’s performance ranking is higher when the firm’s past performance ranking has been
more stable (above the sample median) over the preceding 16 quarters.
In each of our tests described above, we control for firm size, book-to-market ratio, sales
growth, earnings surprise, changes in return on assets, accruals, initial performance ranking of
the firm, and the volatility of the firm’s performance ranking over the past 16 quarters. We also
include industry-quarter fixed effects, which control for industry-time varying effects, such as
5
industry competition (Gormley and Matsa, 2014). We also cluster the standard errors by firm and
quarter to correct for potential serial and cross-sectional correlation (Peterson, 2009).
This paper contributes to the literature to the literature in two key ways. First, we provide
evidence on how changes to the firm’s competitive advantage are manifested in the firm’s
financial statements and how market participants react to these changes. We specifically provide
evidence that a change to the firm’s performance ranking is positively associated with more
persistent earnings, suggesting more sustainable future shareholder profits. We also provide
evidence consistent with shareholders and analysts positively valuing changes to the firm’s
performance ranking. Our evidence suggests that earnings convey useful information to investors
about unexpected earnings innovations as well as changes to the firm’s competitive position
within the industry.
Second, we contribute to the literature by documenting another important and significant
benchmark that investors use to evaluate the firms’ performance – the relative ranking of the
firm’s performance within the industry. Put differently, our study suggests that investors evaluate
the firm’s performance based on the entire distribution of earnings in a given industry to shed
additional light on the firm’s competitive position within the industry. The prior literature,
however, has primarily focused on the costs and benefits of meeting or beating analyst
expectations (e.g., Degeorge et al., 1999; Matsumoto, 2002; Skinner and Sloan, 2002; Fischer et
al., 2014), increasing earnings or revenues from the prior period (e.g., Burgstahler and Dichev,
1997; Degeorge et al., 1999; Ertimur, Livnat, and Martikainen, 2003; Jegadeesh and Livnat,
2006), or reporting earnings greater than zero (e.g., Burgstahler and Dichev, 1997). We extend
the prior literature by providing evidence that the firm’s performance ranking is an equally
important benchmark that investors and analysts use to evaluate the firm’s performance. Market
6
participants, such as the media, often compare the performance of the firm with that of other
firms in the same industry and either implicitly or explicitly rank how firms compare to
competitors.8 Despite this common approach in the popular press, the academic literature has
done little to explore the notion of whether the firm’s performance ranking within the industry is
associated with changes to the firm’s competitive advantage.9
In the next section, we develop our hypotheses. In Section 3, we describe our empirical
tests. We discuss our results in Section 4 and Section 5. We discuss the results from additional
robustness tests in Section 6. We conclude our study in Section 7.
2. Hypothesis Development
Saloner, Shepard, and Podolny (2001) suggest that a firm’s competitive advantage is
created when “there is a difference between what customers are willing to pay for its products or
services and what the firm must pay to provide them.” Therefore, a firm’s competitive advantage
can either be characterized as a cost or quality advantage. That is, a firm can either produce a
product or service at a lower cost or produce a higher quality product or service than other
industry firms. The firm’s performance represents the gap between what customers are willing to
pay and what the firm must pay to provide the good or service. As a result, the firm’s relative
8 For example, The New York Times (2015) recently compared the operating incomes of Apple and Microsoft,
implicitly ranking each firm’s performance and the Wall Street Journal (2012) noted that Lenovo increased personal
computer shipments from 2010 to 2012, improving its ranking from 4th to 2nd in the industry. This type of analysis is
common. The Wall Street Journal (2014) compared the operating profit margins for several firms in the automotive
industry when evaluating the operating performance of Chrysler. Similar examples dot the financial press landscape. 9 A similar notion exists in the contracting literature, where Relative Performance Evaluation (RPE) theory has been
extensively studied. Since Holmstrom (1982), much of the empirical RPE literature examines weather the
compensation committee uses the performance of peer firms to filter out the effect of common external shocks from
the firm’s performance to evaluate CEOs (Aggarwal and Samwick, 1999; Gibbons and Murphy, 1990). According
to the RPE theory, the CEO’s exposure to uncontrollable risks can be eliminated by filtering out the effect of
common external shocks when determining the CEO’s compensation, leading to increased contracting efficiency
(Holmstrom, 1982).
7
performance within the industry represents a summary figure for the firm’s competitive
advantage (the compilation of cost and quality advantages) held within its industry.
We more formally describe the firm’s competitive advantage within the industry by
decomposing the firm’s performance into three components. Rumelt (1991) labels these three
components as 1) the overall business cycle component (i.e., market-specific component), 2) the
industry component, and 3) the business-specific component (i.e., firm-specific component).
Rumelt (1991) documents that the business-specific, or firm-specific, component of performance
is the most significant driver of the firm’s overall performance. Rumelt argues that the firm-
specific component of performance is mainly determined by “the presence of business-specific
skills, resources, reputations, learning, patents, and other intangible contributions to stable
differences among business-unit returns”, all of which are likely to affect the firm’s competitive
advantage.
By examining the intra-industry performance ranking of the firm, we are able to identify
and isolate the firm-specific component of firm performance and control for the portion of firm
performance attributable to industry and market effects, leaving the portion most likely to be
associated with the firm’s competitive advantage. If rivals cannot easily imitate the competitive
advantages held by the firm, then these competitive advantages are expected to be sustained,
allowing the firm to generate higher future profits for shareholders (e.g., Peteraf, 1993; Barney
1986; Barney 1991). Therefore, a change to the firm’s performance ranking ultimately conveys
information to investors about the firm’s ability to continue as a going concern. Therefore, we
anticipate an increase to the firm’s performance ranking to be positively associated with more
persistent earnings, reflecting the firm’s ability to sustain higher future profits due to its
competitive advantage. We state our first hypothesis below.
8
H1 – Firms have more (less) persistent earnings when the firm’s performance ranking within
the industry increases (decreases).
As stated above, we argue that investors obtain information regarding changes to the
firm’s competitive advantage by observing changes to the firm’s performance ranking within the
industry. Of course, managers are always pursuing various strategic activities to establish
competitive advantages in order to differentiate themselves from their rivals. However, the
effectiveness of those activities may not be directly observable and may not be properly
evaluated by investors when immediately implemented. Litov et al. (2012) argue that market
participants face significant information problems resulting from managerial proprietary insights
about the future value of the firm’s unique strategy. They also argue that “if managers do not
possess proprietary insights, and instead all opportunities are transparently obvious to the market,
replication of strategies will occur and arbitrageurs will buy the resources required by the
managers and sell them to the firms at prices near their value added in the manager’s strategy,
thereby dissipating any value to be created by the strategy (Barney, 1986).” In addition,
significant uncertainty exists as to whether the particular strategy can establish competitive
advantages that generate sustainable future profits. Therefore, market participants are less likely
to fully understand the implications of various strategic activities until earnings are released.
We anticipate that the firm’s performance ranking, which removes industry-specific and
market-specific information, functions as a summary statistic for the strategic activities
implemented by the manager, allowing investors to evaluate the firm’s competitive advantage
within the industry.10 We specifically predict that investors positively (negatively) value an
10 The competitive advantages that generate higher firm performance may not be sustainable in the future if the
strategy can be easily replicated by rivals. We discuss this issue in detail in Hypothesis 3.
9
increase (decrease) in the performance ranking of the firm within the industry. We state our
hypothesis in alternative form below.
H2 – Investors positively (negatively) value an increase (decrease) in the firm’s relative
performance ranking within the industry.
Depending on the firm and/or industry characteristics, the firm’s performance ranking
may fluctuate significantly, providing a less informative signal about changes to the firm’s
competitive advantage within the industry. For instance, if rivals are able to imitate the strategies
of the firm relatively easily, then an increase in the firm’s performance ranking is expected to
reverse quickly, leading to volatile performance ranking changes. In this case, investors are more
likely to view changes to the firm’s performance ranking in the current period as a temporary
fluctuation rather than an indication of a persistent shift in the firm’s competitive advantage
within the industry. However, if the past volatility of the firm’s performance ranking is low, we
anticipate that the change in the firm’s performance ranking in the current period provides
investors with a more informative signal of how the firm’s competitive position within the
industry has shifted, leading to a greater investor response. Therefore, we predict that investors
react more (less) strongly to a change in the firm’s performance ranking when the volatility of
the firm’s past performance rankings is lower (higher). Our hypothesis is stated in alternative
form below.
H3 – Investors react more (less) strongly to a change in the firm’s performance ranking when
the past volatility of the firm’s past performance ranking is low (high).
3. Empirical Design
3.1. Changes in Performance Rankings and Earnings Persistence
10
In Hypothesis 1, we predict that earnings are more (less) persistent if a firm experiences
an increase (decrease) in the firm’s relative performance ranking within the industry. In the main
analyses, we use the Global Industry Classification Standard (GICS) codes to define the industry
to which a firm belongs. Bhojraj et al. (2003) document that firms in the same GICS
classifications have higher profitability and growth correlations than firms that share the same
Standard Industrial Classification (SIC) codes, North American Industry Classification System
(NAICS) codes, and Fama-French classification codes. They conclude that GICS is a better
industry classification to identify industry peers that compete in similar product markets.
Using GICS codes to define the industry, we measure the change in the firm’s
performance ranking within the industry during fiscal quarter and estimate the following
regression model.
ROAi,t+4 = α + β1 ROAi,t + β2 D_∆Ranking_QTRi,t + β3 ROAi,t × D_∆Ranking_QTRi,t
+ β4 Surprise_QTRi,t + β5 Sizei,t + β6 Book-to-Marketi,t + β7 STD_∆Ranking_QTRi,t
+ β8 Initial Ranking_QTRi,t + β9 ∆ROAi,t + β10 SalesGrowthi,t + β11 Accrualsi,t + εi,t
(1)
All variable definitions can be found in Appendix A. The dependent variable is firm i’s
return on assets (ROAi,t+4) at quarter t+4. The coefficient on the ROAi,t variable reflects earnings
persistence. We interact the ROAi,t variable with the D_∆Ranking_QTRi,t variable to examine H1.
The D_∆Ranking_QTRi,t variable is the decile-ranked ∆Ranking_QTRi,t variable. The
∆Ranking_QTRi,t variable is measured as firm i’s realized ranking in quarter t, based on realized
earnings, minus firm i’s initial ranking measured at the beginning of quarter t, based on analysts’
expected earnings, divided by the number of firms in the same industry. Realized ranking is
equal to firm i’s performance ranking within its 6-digit GICS industry based on firms’ IBES-
reported actual earnings in quarter t plus interest expenses multiplied by one minus marginal tax
11
rates, divided by average total assets in quarter t (i.e., Return on assets = (net income + interest
expense× (1 – Marginal Tax Rate)) / Average Total Assets).11 Initial ranking is constructed in a
similar way except for the use of consensus analyst median forecast at the beginning of quarter
t.12 If the consensus analyst forecast is missing, the expected earnings of firm i is equal to the
IBES-reported actual earnings in quarter t-4, which assumes that expected earnings follow a
seasonal random walk (e.g., Freeman and Tse, 1989; Bernard and Thomas, 1990). All EPS
figures are multiplied by the number of shares outstanding. We create decile-ranked variables by
ranking the ∆Ranking_QTRi,t variable into deciles (i.e., 0 through 9) and dividing by 9. Hence,
the estimated coefficient on the interaction term represent an increase in earnings persistence as
the ∆Ranking_QTRi,t variable moves from its 1st to 10th decile.
We also include several control variables in the model that are likely to be simultaneously
associated with performance ranking changes and earnings persistence. The Surprise_QTRi,t
variable is calculated as the difference between IBES-reported actual EPS for firm i in quarter t
less the expected EPS at the beginning of quarter t, which is used in the construction of the
∆Ranking_QTRi,t variable, divided by stock price at the beginning of quarter t. The Sizei,t
variable is equal to the natural logarithm of the market value of equity at the end of quarter t.
The Book-to-Marketi,t variable is calculated by dividing the book value of equity by the market
value of equity at the end of quarter t. The STD_∆Ranking_QTRi,t variable is the standard
deviation of the ∆Ranking_QTRi,t variable over the previous 16 quarters (we require a minimum
11 Marginal tax rate is assumed as the top statutory federal tax rate plus 2% average state tax rate (Nissim and
Penman 2003). 12 We calculate the daily median EPS consensus analyst forecast using the I/B/E/S unadjusted detail file. We
specifically calculate the median EPS consensus based on individual analyst forecasts, which are required to be
reported within the 90-day window immediately preceding the consensus forecast date to ensure that our analyst
consensus is not based on stale forecasts. We exclude individual analyst forecasts if I/B/E/S excludes the forecasts
from calculating IBES-reported median EPS consensus. If the daily median EPS consensus analyst forecast is
missing, we supplement our data by using IBES-reported median EPS consensus forecasts (i.e., IBES item
MEDEST).
12
of 8 quarter observations. The Initial Ranking_QTRi,t variable is the initial performance ranking
of the firm measured as the beginning of quarter t. The ∆ROAi,t variable is measured as changes
in firm i’s return on assets between quarter t and quarter t-4. The SalesGrowthi,t variable is equal
to net sales in quarter t divided by net sales in quarter t-4. Accrualsi,t is measured as firm i’s
GAAP EPS less cash flows from operations per share in quarter t divided by the stock price at
the end of quarter t. Once again, industry is defined as firms in the same GICS code. For all our
tests, we include industry-quarter fixed effects to control unobserved time-varying industry-
specific factors such as industry competition (Gormley and Matsa, 2014). We cluster the
standard errors by calendar/quarter and firm to correct for cross-sectional and serial-correlation
in the standard errors (Petersen, 2009).
3.2. The Market’s Reaction to Changes in Performance Rankings
In Hypothesis 2, we predict that market participants such as investors and analysts
positively (negatively) react to an increase (decrease) in the firm’s relative performance ranking
within the industry. To test this hypothesis, we measure the change in the firm’s performance
ranking within the industry on the date of the firm’s earnings announcement. While we could
examine the change in the firm’s performance ranking at various points during the fiscal period,
we choose the earnings announcement date for two reasons. First, the earnings announcement is
a significant firm event that reveals information about the firm, prompting investors to evaluate
and revise their expectations about the firm’s performance ranking. Second, earnings are released
during the earnings announcement, allowing us to examine the market’s response to the change
in the firm’s performance ranking relative to the market’s response to the earnings surprise.
Therefore, we can control for the firm’s earnings surprise in the regression analyses, allowing us
to isolate the incremental effect of the change in the firm’s performance ranking. We employ the
13
following regression model to examine market participants’ responses to changes in the firm’s
performance ranking.
DEPVARi,t = α + β1 ∆Ranking_EAi,t + β2 Surprise_EAi,t + β3 Sizei,t + β4 Book-to-Marketi,t
+ β5 STD_∆Ranking_EAi,t + β6 Initial Ranking_EAi,t + β7 ∆ROAi,t
+ β8 SalesGrowthi,t + β9 Accrualsi,t + εi,t
(2)
The dependent variable is equal to either the 3DayReti,t variable, the AF_Revisioni,t
variable, or the ∆RECi,t variable. The 3DayReti,t variable is measured as the three-day market-
adjusted buy-and-hold abnormal return centered on the earnings announcement for firm i in
quarter t. The AF_Revisioni,t variable is measured as the consensus median EPS forecast five
days after the announcement of earnings minus the consensus median EPS forecast five days
prior to the announcement of earnings divided by the absolute value of the consensus forecast
five days prior to the announcement of earnings for firm i in quarter t. The consensus median
EPS forecast is calculated based on individual analyst forecasts, which are required to be
reported within a 90-day window preceding the daily date. The ∆RECi,t is measured as the
difference between the consensus median stock recommendation five days after the
announcement of earnings less the consensus median stock recommendation five days prior to
the announcement of earnings for firm i in quarter t. The consensus median stock
recommendation is calculated based on individual analyst recommendations, which are required
to be reported within the preceding 90-days. A standard set of recommendation maintained by
IBES is as follows: Strong Buy (1), Buy (2), Hold (3), Underperform (4), and Sell (5). Hence, we
multiply changes in stock recommendation by -1 to be consistent with the AF_Revisioni,t variable.
The main independent variable of interest in this regression is the ∆Ranking_EAi,t
variable, which is measured as firm i’s realized performance ranking within the industry on the
14
earnings announcement date in quarter t less firm i’s initial performance ranking within the
industry two days prior to the earnings announcement date in quarter t, divided by the total
number of firms in the industry. Initial Ranking of the firm two days prior to the announcement
of earnings is based on expected earnings of firm i and calculated using the median analyst
forecast calculated two days prior to the earnings announcement in quarter t. If the consensus
EPS forecast is missing, the expected earnings of firm i is equal to the IBES-reported actual EPS
in quarter t-4 (Freeman and Tse, 1989; Bernard and Thomas, 1990). We then rank the expected
earnings for firm i with the realized or expected earnings for all other firms sharing the same
GICS code (i.e., peer firms) on the same date. If peer firms have already announced earnings, we
use peer firms’ realized earnings. If peer firms have not already announced earnings, we
calculate the expected earnings for peer firms following the same procedure described above.
Prior to calculating the initial ranking of the firm i in quarter t, we standardize the realized and
expected earnings for all firms in the industry by multiplying each EPS figure by the number of
shares (depending on the IBES basic/diluted flag) adding total interest expense multiplied by one
less marginal tax rates, and dividing by the average total assets for each firm. We calculate the
realized ranking of firm i in quarter t at the earnings announcement date similarly to how we
calculated the initial ranking of the firm with the one exception. Instead of using expected
earnings for firm i, we use the realized earnings for firm i that are announced at the earnings
announcement date for quarter t. We then rank firm i’s realized earnings relative to peer firms’
realized or expected earnings, depending on whether or not peer firms have announced their
earnings on the earnings announcement date of firm i in quarter t. We anticipate finding a
positive coefficient on the ∆Ranking_EAi,t variable in equation (2), which is consistent with
15
investors and analysts positively valuing an improvement to the firm’s performance ranking in
the industry.
We also include several control variables in the model. The Surprise_EAi,t variable is the
earnings surprise for firm i in quarter t, which is equal to the IBES-reported Actual EPS figure
less the expected earnings, which is used in the construction of the ∆Ranking_EAi,t variable
divided by stock price at the end of quarter t. The STD_∆Ranking_EAi,t variable is the standard
deviation of the ∆Ranking_EAi,t variable over preceding 16 quarters (we require a minimum of 8
quarter observations). The Initial Ranking_EAi,t variable is firm i’s performance ranking
measured at two days prior to earnings announcement in quarter t. All other variables are
previously defined and described in the Appendix A.
4. Data and Descriptive Statistics
Data for our empirical tests were obtained from the intersection of I/B/E/S,
COMPUSTAT, and CRSP. We start our analysis in 1995 because individual analyst forecasts are
relatively sparse prior to 1995 (Clement et al., 2011). Since one of our main control variables,
STD_∆Ranking_QTRi,t, requires at least past 8 quarters of data, our sample period ranges from
1997 to 2013. We retrieve quarterly financial statement data from COMPUSTAT and daily stock
return data from CRSP. We require at least 10 firm-quarter observations in each industry for
each quarter to calculate the performance ranking changes for each firm in the industry. We only
keep firm/quarter observations with fiscal quarter ends of March, June, September, and
December.13 We also require firm/quarter observations to have sufficient data to calculate the
13 We include only firms that have calendar/quarter fiscal period ends to ensure that the earnings windows are the
same for each firm in the industry. For example, we do not want to compare earnings that are generated from
September to November for one firm to earnings that are generated from October to December of another firm
16
independent and dependent variables in each regression. Our final sample consists of 198,467
firm-quarter observations ranging from 1997 to 2013. The number of observations in any
particular test varies depending on the availability of data necessary for each test. All continuous
variables are winsorized at 1% and 99%.
Table 1 presents descriptive statistics for the full sample. We note that average changes in
the firm’s performance ranking in fiscal quarter (∆Ranking_QTRi,t) or at earnings announcement
(∆Ranking_EAi,t) are 0.001 and 0.004 that are reasonably close to zero. We expected this given
its construction. The average (median) earnings surprise at the announcement of earnings
(Surprise_EAi,t) is -0.004 (0.000). The mean and median values for the Surprise_EAi,t variable
suggests that analyst forecasts are relatively unbiased. Mean and the median values of all other
variables are similar to those reported in prior research. For instance, the mean (median) book-to-
market ratio is 0.665 (0.515) and the mean (median) sales growth is 1.151 (1.075), which are
both similar values to those found in prior studies (e.g., Doyle et al., 2013).
Table 2 reports the Pearson and Spearman correlations for the primary variables in our
study. We provide preliminary support for our hypothesis by finding a positive correlation
between the 3DayReti,t and ∆Ranking_EAi,t variables, suggesting that investors positively value
changes in the firm’s performance ranking. We also find a positive correlation between the
3DayReti,t and Surprise_EAi,t variables, which is consistent investors positively responding to
unexpected earnings. The ∆Ranking_EAi,t variable is also positively correlated with the
Surprise_EAi,t variable (Pearson correlation 0.53), the ROAi,t variable (Pearson correlation 0.21),
and the ∆ROAi,t variable (Pearson correlation 0.40), suggesting that a firm with an increase in its
because there could be an industry or market shock in December that make the earnings of the two firms less
comparable.
17
change in performance ranking have higher earnings surprises and have better performance.14
This further highlights the need to control for various measures of the firm’s performance in our
multivariate regressions analyses.
Panel A of Table 3 presents descriptive statistics on the likelihood that firms maintain
their performance ranking during fiscal quarter. Using the initial ranking of the firm measured at
the beginning of fiscal quarter, we divide the firm/quarter observations into quintiles, which are
denoted in the first column of Panel A. We also divide the firm/quarter observations into
quintiles using the realized ranking of the firm measured at the end of fiscal quarter, which are
denoted in the first row of Panel A. The number and percentage of firm/quarter observations that
are categorized into each Initial Ranking_QTRi,t quintile and Realized Ranking_QTRi,t quintile
are found at the intersection of these two quintile categories. As we expected, we find that firms
are more likely to maintain their performance rankings. For example, approximately 62% of the
firm/quarter observations that are initially ranked to be in the 1st performance quintile at the
beginning of fiscal quarter end up having a realized ranking in the same 1st performance quintile
at the end of fiscal quarter. We also note that the likelihood of a firm changing its performance
ranking decreases as we move away from the diagonal (i.e., firms maintaining the same initial
and realized performance ranking quintiles). For example, only 4% of the firm/quarter
observations are initially ranked as being in the 1st performance quintile and end up having a
realized ranking in the 4th performance quintile. This evidence suggests that performance
rankings are relatively stable during fiscal quarter.
In Panel B, we also examine the transition likelihood at the date of earnings
announcement. If analysts update earnings forecasts during the period and the updated consensus
14 Given the higher correlation between the ∆Rankingi,t variable and other performance controls, we check the
Variance Inflation Factor (VIF) in all subsequent regression analyses and find that multicollinearity problem is not
present in our analyses.
18
forecast better reflects the firm’s position within the industry, we expect to observe a higher
percentage of firms staying the same initial and realized ranking quintile (e.g., we expect a firm
in the 1st initial ranking quintile to also be in the 1st realized ranking quintile after earnings are
announced). Consistent with expectation, we find that, approximately 76% of the firm/quarter
observations that are initially ranked to be in the 1st performance quintile two days prior to the
announcement of earnings end up having a realized ranking in the same 1st performance quintile
at the date of earnings announcement. In addition, we note that 2% of the firm/quarter
observations are initially ranked as being in the 1st performance quintile and end up having a
realized ranking in the 4th performance quintile at the date of earnings announcement. This
evidence suggests that analysts’ most recent earnings forecasts prior to earnings announcement
reflect some changes in firms’ competitive position within the industry. However, we still
observe variations in changes in performance rankings at the date of earnings announcement. As
a result, a change in the performance ranking of the firm is more likely to provide information to
investors about a notable change to the performance of the firm.
5. Empirical Results
5.1. Results for H1: Performance Ranking Changes and Earnings Persistence
Hypothesis 1 predicts that a change to the relative performance ranking is positively
associated with future earnings persistence. Table 4 presents the estimation results of the
equation (1). In column (1), the dependent variable is specified as the one-year-ahead same
quarter return on assets (ROAi,t+4) with the ROAi,t variable as the sole independent variable. In
our sample, we find that average earnings persistence is 0.669 in column (1), which is similar to
the level of earnings persistence reported in the prior studies (e.g., Dichev and Tang, 2008). In
19
column (2), we include the decile-ranked changes in ranking variable (D_∆Ranking_QTRi,t)
variable along with its interaction with the ROAi,t variable. We find a positive and significant
(1% level) coefficient on the ROAi,t variable equal to 0.576 and on the interaction term,
suggesting that the persistence of earnings increases for firms as the change to the firm’s
performance ranking increases from its 1st to 10th decile. This result holds when we include other
control variables in column (3). Regarding the magnitude of coefficients, firms moving from the
lowest to the highest change in performance ranking decile have earnings that are 22.77% (=
0.143 / 0.628) more persistent. In column (4) and column (5), we use two-year-ahead same
quarter return on assets (ROAi,t+8) and three-year-ahead same quarter return on assets (ROAi,t+12),
respectively. We continue to find consistent evidence supporting our prediction. Overall, the
results in Table 4 support our hypothesis that changes in the firm’s performance ranking provide
information concerning the firm’s competitive advantages in the product market, which are
associated with greater sustainable future profits.
5.2. Results for H2: Market participants’ responses to changes in performance rankings
5.2.1 Univariate Analyses
Hypothesis 2 predicts that market participants such as investors and analysts positively
value changes in the firm’s performance ranking because it provides information regarding the
improvement in the firm’s competitive advantage. Table 5 provides univariate results examining
investor and analyst reactions to changes in the firm’s performance ranking. Similar to Panel B
of Table 3, we include each Initial Ranking_EAi,t quintile in the first column and each Realized
Ranking_EAi,t quintile in the first row of the table. The values that lie at the intersection of each
Initial Ranking_EAi,t and Realized Ranking_EAi,t quintile represent the average reaction of
market participants for those firms that fall into this category. Panel A of Table 5 reports
20
averages of three-day market-adjusted stock returns surrounding earnings announcement. We
find that firms with an initial ranking in the 1st quintile that end with a realized ranking in the 5th
quintile on the date of earnings announcement experience market returns of 0.05 on average,
which is significant at the 1% level. We also note that average stock returns are increasing
monotonically as we move from the 1st realized ranking quintile to 5th realized ranking quintile
holding the initial ranking quintile constant.15
To further check the overall average effect on returns as the performance ranking
changes, we examine the average returns when the performance ranking of the firm increases or
decreases by one, two, three, and four quintiles. For example, we compute the average return for
all firms that increase one quintile (e.g., initial ranking is equal to 1 and the realized ranking is
equal to 2) and include it under the category “+1”. Similarly, we compute the average return for
all firms that increase two quintiles (e.g., initial ranking is equal to 2 and the realized ranking is
equal to 4) and include it under the category “+2”. Figure 1 provides graphical evidence that
market reactions increase monotonically as firms experience greater ranking changes relative to
their initial performance rankings, providing additional univariate support for our second
hypothesis.
Panel B and C of Table 5 report average analyst forecast revisions and average analyst
stock recommendation changes surrounding the earnings announcement, respectively. Consistent
with the results based on stock returns in Panel A, we continue to find that increases in
performance ranking on the date of earnings announcement are positively associated with
upward analyst forecast revisions and improvements in stock recommendations. Figure 2 is
15 We also note that nearly monotonic relations in average returns are observed as we move from the 1st to 5th initial
ranking quintile within each realized ranking quintile. However, we believe that examining how the average return
changes as we move from the 1st to 5th realized ranking quintile within each initial ranking quintile is more intuitive
because the initial ranking is first observed by investors and the realized ranking is subsequently realized.
21
calculated similarly to Figure 1 and provides some evidence that changes in performance
rankings are positively associated with favorable revisions in analyst forecasts and stock
recommendations.
Taken together, the results in Table 5, Figure 1, and Figure 2 are consistent with our
second hypothesis that investors and analysts positively (negatively) value an increase (decrease)
in the firm’s relative performance ranking. However, we note that the market and analyst
reactions examined in the univariate analyses do not include control variables and other
performance measures that are correlated with the change in the firm’s performance ranking;
therefore, the result may not indicate the incremental effect of the change in the firm’s
performance ranking. In the next section, we attempt to estimate the incremental effect on the
market reaction after controlling other performance measures and firm characteristics in the
multivariate regression analysis.
5.2.2. Market Reactions to Changes in Performance Rankings
In Hypothesis 2, we predict that investors and analysts positively respond to changes in
performance rankings within the industry surrounding earnings announcement after controlling
the effect of the earnings surprise along with other specific firm and industry characteristics.
Table 6 presents the results from equation (2) with the three-day buy-and-hold abnormal returns
(3DayReti,t) surrounding the date of earnings announcement as the dependent variable and the
firm’s performance ranking changes occurring on the date of earnings announcement
(∆Ranking_EAi,t) as the primary independent variable of interest. In column (1) and (2), we omit
the control variables from the analysis and include the ∆Ranking_EAi,t and Surprise_EAi,t,
variables in separate regressions. We find a positive and significant (1% level) coefficient on
both the ∆Ranking_EAi,t variable as well as the Surprise_EAi,t, variable in column (1) and (2),
22
respectively. This evidence is consistent with investors positively (negatively) valuing an
unexpected increase (decrease) in the firm’s performance ranking and an unexpected increase
(decrease) in earnings. We note that the R2 is equal to 3.8% when the ∆Ranking_EAi,t variable is
the independent variable in column (1) and 1.8% when the Surprise_EAi,t, variable is the
independent variable in column (2). Using a Vuong (1989) test, we find that the R2 calculated
when the ∆Ranking_EAi,t variable is the sole independent variable is statistically greater (Z-
statistics = 21.49) than the R2 calculated when the Surprise_EAi,t, variables is the sole
independent variable, suggesting that changes in performance rankings explain more of the
variation in the three-day market reaction around the earnings announcement than the earnings
surprise.16
In column (3), we include both the ∆Ranking_EAi,t and Surprise_EAi,t, variables together
in the same regression. We find positive and significant coefficients on both variables, consistent
with investors positively valuing an improvement in the firm’s relative performance ranking as
well as unexpected earnings. This evidence also suggests that the change in the firm’s
performance ranking conveys incremental information to the capital markets over the earnings
surprise. It is also worth noting that the magnitude of coefficient on the Surprise_EAi,t variable is
substantially reduced in column (3) relative to column (2) while the magnitude of coefficient on
the ∆Ranking_EAi,t variable remains relatively the same in column (3) relative to column (1).
This evidence further suggests that the ∆Ranking_EAi,t variable provides information distinct
from information contained in the earnings surprise. In column (4) of Table 4, we include control
variables along with industry-quarter fixed effects to reduce the likelihood of correlated omitted
variables. We continue to find that the coefficients on the ∆Ranking_EAi,t and Surprise_EAi,t,
variables are significantly positive at the 1% level.
16 We find the same result when we include control variables in each regression.
23
To gauge the relative effect of a change in the relative performance ranking of the firm
and the earnings surprise, we use the standardized coefficients. Standardizing coefficients are
used to compare the relative strength of each independent variable included in the regression
(e.g., Maddala, 1977; Palmrose, 1986; Waring, 1996). Because standardized coefficients are
measured in terms of standard deviations and not the underlying units for each independent
variable, we can compare the relative strength of each standardized coefficient across
independent variables. Standardized coefficients are calculated by multiplying the coefficient by
the standard deviation of the independent variable and dividing by the standard deviation of the
dependent variable. We find that one standard deviation increase in the ∆Ranking_EAi,t variable
leads to 0.187 standard deviation increase in the 3DayReti,t variable (0.187 = 0.103 * (0.158 /
0.087)). We find that one standard deviation increase in the Surprise_EAi,t variable leads to 0.038
standard deviation increase in the 3DayReti,t variable (0.038 = 0.076 * (0.043 / 0.087)).17
Surprisingly, this evidence suggests that the market’s reaction to a change in performance
ranking is greater than that for the earnings surprise.
In Table 7, we examine analysts’ responses to changes in performance rankings
surrounding earnings announcement. In Panel A of Table 7, we use revisions of the consensus
median analyst EPS forecast surrounding earnings announcement, i.e., the AF_Revisioni,t
variable, as the dependent variable in equation (2). Similar to Table 6, we include the
∆Ranking_EAi,t variable as the only independent variable in column (1), the Surprise_EAi,t
variable as the only independent variable in column (2), only the ∆Ranking_EAi,t and
Surprise_EAi,t variables in column (3), and all independent variables in column (4). The sign and
17 In untabulated result, we only include the Surprise_EAi,t variable along with control variables (i.e., without the
∆Ranking_EAi,t variable) and find that one standard deviation increase in the Surprise_EAi,t variable leads to 0.115
standard deviation in the 3DayReti,t variable. This evidence suggests that changes to the firm’s performance ranking
is an important variable that is correlated with the earnings surprise and market returns.
24
significance of the ∆Ranking_EAi,t and Surprise_EAi,t variables in each column are similar to
those in Table 6. The coefficients in column (4) suggest that one standard deviation increase in
the ∆Ranking_EAi,t variable leads to 0.139 standard deviation increase in the AF_Revisioni,t
variable while one standard deviation increase in the Surprise_EAi,t variable leads to 0.066
standard deviation increase in the AF_Revisioni,t variable. Once again, this evidence suggests that
analysts are responding more to a change in the firm’s performance ranking than the earnings
surprise.
In Panel B of Table 7, we investigate stock recommendation changes in response to the
firm’s performance ranking change surrounding earnings announcement. Interestingly, when we
include the Surprise_EAi,t variable with the ∆Ranking_EAi,t variable in column (3), the
coefficient on the Surprise_EAi,t variable becomes insignificant. However, in column (2) we find
a positive and significant coefficient on the Surprise_EAi,t variable when it is the only variable
included in the regression. We also find an insignificant coefficient on the Surprise_EAi,t variable
in column (4) when the control variables are included. This evidence continues to suggest that
changes to the firm’s performance ranking is more important to analysts than the firm’s earnings
announcement.
In summary, the results in Table 6 and 7 suggest that improvements to the firm’s
performance ranking represent an important metric that market participants use to assess the
firm’s performance. These results also suggest that investors evaluate firm performance based on
the entire distribution of earnings in the industry rather than just based on analysts’ expectations
of performance.
5.3. Results for H3: Market reactions to performance ranking changes conditioning on the
volatility of past performance ranking changes
25
We next examine Hypothesis 3, which predicts that investors’ response to a change in the
firm’s performance ranking in the current period more (less) when the firm’s volatility in past
performance ranking changes is low (high). To examine this hypothesis, we divide the full
sample into two subsamples based on the sample median of the standard deviation in the change
to the firm’s performance ranking over the past 16 quarters (STD_∆Ranking_EAi,t), estimate two
separate regressions (i.e. Low and High groups), and test the difference between the estimated
coefficients for the two groups. Table 8 presents the estimation results and the results from the
coefficient difference tests are reported at the bottom of Table 8.
In column (1) and (2), we find a positive and significant (1% level) coefficient on the
∆Ranking_EAi,t variable in both Low and High group but the coefficient in High group (column
2) is significantly lower than the coefficient in Low group (difference -0.214, p-value 0.000).
This evidence is consistent with our hypothesis that the market reaction to changes in the
performance ranking is greatest when the volatility of past performance ranking changes is low.
We also note that the coefficients on the Surprise_EAi,t variable are not statistically different
across two subsamples.
In column (3) and (4), we examine analyst forecast revisions. In column (5) and (6), we
examine changes in stock recommendation surrounding earnings announcement. We continue to
find results that a change to the firm’s performance ranking is more highly associated with
analyst forecast revisions and analyst stock recommendations when the standard deviation of
past performance rankings is lower. We also find that the coefficients on the Surprise_EAi,t
variable are not significantly different between the high and low groups when the dependent
variable is the AF_Revisioni,t or ∆RECi,t variable. Overall, the results in Table 8 corroborate our
26
argument that market participants respond to changes to the firm’s competitive advantage that
are manifested through changes to the firm’s performance ranking within the industry.
6. Additional tests
6.1. Industry-adjusted buy-and-hold abnormal returns
We argue that changes to performance ranking provide information about the firm-
specific component of performance as compared with the industry-wide or market-wide
component of performance. In our main empirical tests, we use market-adjusted buy-and-hold
abnormal returns to examine whether or not market participants respond to the firm’s
performance ranking changes. However, there is a possibility that the significantly positive
coefficient on the ∆Ranking_EAi,t variable might contain industry-wide information and
investors respond to the information. To rule out this possibility, we calculate industry-adjusted
buy-and-hold abnormal stock returns at the earnings announcement date of each firm and
reperform our tests found in Table 6, which are found in Table 9.18 Consistent with expectations,
we continue to find a significantly positive coefficient on the ∆Ranking_EAi,t variable in all
specifications at the 1% level. We also note that the coefficients are very similar in Table 6 and
9. This evidence is consistent with the ∆Ranking_EAi,t variable capturing the firm-specific
component of performance and not the industry-specific component of performance. and that
market participants respond to the changes in the firm’s performance ranking.
6.2. Additional Alternative Specifications
18 We specifically calculate the equal-weighted average stock returns using stock returns of all firms in the same
industry (GICS codes) as firm i surrounding the firm i’s earnings announcement (excluding firm i from the
calculation) and subtract it from firm i’s 3-day buy-and-hold stock returns, and use this industry-adjusted abnormal
return as a dependent variable.
27
We perform several additional tests to examine the robustness of our results. In our first
test, we delete all observations in which there is no change between the initial and realized
performance ranking and rerun our main results found in Table 6. In untabulated results, we find
qualitatively similar results to those results reported in Table 6. Our results do not seem to be
driven by firm/quarter observations with no change the firm’s performance ranking. Second, we
drop all observations that experience an increase or a decrease in the performance ranking of
more than one quintile. In other words, we delete all firm/quarter observations in which the
absolute value of the difference between the initial ranking and realized ranking quintiles is
greater than one. The untabulated results suggest that our results are qualitatively similar to those
in Table 6. Our results do not appear to be driven by firm/quarter observations with extreme
changes in the firm’s performance ranking. Third, we assess whether small firms are driving our
results. We eliminate all observations with stock price less than $3 or total assets less than $5
million and find that the results are qualitatively similar. Based on the above, it does not appear
that firms with little economic significance are driving our results.
7. Conclusion
In this study, we examine how changes to the firm’s competitive advantage are
manifested and in the firm’s financial statements and how market participants react to these
changes. We measure changes to the firm’s competitive advantage using changes to the firm’s
performance ranking within the industry, which allows us to eliminate industry and market level
shocks that affect the firm’s performance. We measure the change in the firm’s performance
ranking within the industry by considering how a realized ranking based on the released EPS
figure at the earnings announcement date is different from an initially expected ranking based on
28
expected earnings either at the beginning of fiscal quarter (in earnings persistence test) or
immediately prior to the earnings announcement (in market response tests). Using these
measures, we find that earnings persistence is higher when the firm experiences an increase in
performance ranking in the industry. We also find that the buy-and-hold market-adjusted
abnormal returns, the consensus analyst forecast revisions, and changes in stock
recommendations surrounding the earnings announcement are positively associated with the
change in the firm’s performance ranking after controlling the earnings surprise and other control
variables. We also find that firms with a history of stable (volatile) performance ranking changes
have a higher (lower) market and analyst response to changes in the firm’s performance ranking,
which is less indicative of a change to the firm’s competitive advantage when it fluctuates
significantly.
This study provides evidence on how changes to the firm’s competitive advantage are
reflected in the firm’s financial statements. Earnings realizations convey useful information to
investors regarding the firm’s ability to compete within the industry. The prior research has
examined the information conveyed at specific points in the earnings distribution, such as around
analyst expectations, zero net income, and increases in earnings relative to the prior fiscal period.
We believe that our study provides evidence that investors use the entire distribution of earnings
within the industry to assess the firm’s performance.
29
References
Abarbanell, J. and Lehavy, R. 2003. “Biased Forecasts or Biased Earnings? The Role of
Reported Earnings in Explaining Apparent Bias and Over/underreaction in Analysts’
Earnings Forecasts.” Journal of Accounting and Economics 36 (1): 105–46.
Aggarwal, R. K., and Samwick, A. A. 1999. The Other Side of the Trade-Off: The Impact of
Risk on Executive Compensation. Journal of Political Economy,107(1), 65-105.
Angrist, J. D., and Pischke, J. S. 2008. Mostly harmless econometrics: An empiricist's
companion. Princeton university press.
Bernard, V. L., and Thomas, J. K. 1990. Evidence that stock prices do not fully reflect the
implications of current earnings for future earnings. Journal of Accounting and
Economics, 13(4), 305-340.
Bhojraj, S., Hribar, P., Picconi, M., McInnis, J. 2009. Making Sense of Cents: An Examination
of Firms that Marginally Miss or Beat Analyst Forecasts. The Journal of Finance 64(5):
pg. 2361-2388.
Bhojraj, S., Lee, C., and Oler, D. K. 2003. “What’s My Line? A Comparison of Industry
Classification Schemes for Capital Market Research.” Journal of Accounting Research
41 (5): 745–74.
Boni, L. and Womack, K. L. 2006. “Analysts, Industries, and Price Momentum.” Journal of
Financial and Quantitative Analysis 41 (01): 85–109.
Bowen, R. M., Davis, A. K., and Matsumoto, D. A.. 2005. “Emphasis on pro forma versus
GAAP earnings in quarterly press releases: Determinants, SEC intervention, and market
reactions.” The Accounting Review 80, no. 4: 1011-1038.
Bradshaw, M and Sloan, R. 2002. “GAAP versus The Street: An Empirical Assessment of Two
Alternative Definitions of Earnings.” Journal of Accounting Research 40 (1): 41-66.
Burgstahler, D., and Dichev, I., 1997. “Earnings management to avoid earnings decreases and
losses.” Journal of Accounting and Economics 24, no. 1: 99-126.
Burgstahler, D. and Eames, M. 2006. “Management of Earnings and Analysts’ Forecasts to
Achieve Zero and Small Positive Earnings Surprises.” Journal of Business Finance &
Accounting 33 (5‐6): 633–52.
Calia, M., 2014. “Fiat Chrysler, GM sales jump, Ford sales slip”, The Wall Street Journal, 1 May
Clement, M. B., Hales, J., and Xue, Y. 2011. Understanding analysts' use of stock returns and
other analysts' revisions when forecasting earnings. Journal of Accounting and
Economics, 51(3), 279-299.
30
Dichev, I. D., and Tang, V. W. 2008. Matching and the changing properties of accounting
earnings over the last 40 years. The Accounting Review, 83(6), 1425-1460.
Doyle, J. T., Lundholm, R. J., and Soliman, M. T., 2003. “The predictive value of expenses
excluded from pro forma earnings.” Review of Accounting Studies 8, no. 2-3: 145-174.
Degeorge, F., Patel, J., and Zeckhauser, R. 1999. “Earnings Management to Exceed
Thresholds*.” The Journal of Business 72 (1): 1–33.
Doyle, J., Jennings, J., and Soliman, M. 2013. “Do Managers Define Non-GAAP Earnings to
Meet or Beat Analyst Forecasts?” Journal of Accounting and Economics 56 (1): 40–56.
Ertimur, Y., Livnat, J., and Martikainen, M. 2003. Differential market reactions to revenue and
expense surprises. Review of Accounting Studies, 8(2-3), 185-211.
Fischer, P., Jennings, J., and Soliman, M. 2014. “Meeting, Beating, and Bubbles” Working
paper.
Freeman, R. N., and Tse, S. (1989). The multiperiod information content of accounting earnings:
Confirmations and contradictions of previous earnings reports. Journal of Accounting
Research, 49-79.
Gibbons, R., Murphy, K., 1990. Relative Performance Evaluation for Chief Executive Officers.
Industrial and Labor Relations Review: 30S–51S.
Gormley, T. A., and Matsa, D. A. 2014. Common errors: How to (and not to) control for
unobserved heterogeneity. Review of Financial Studies, 27(2), 617-661.
Graham, J. R., Harvey, C. R., and Rajgopal, s., 2005. “The economic implications of corporate
financial reporting.” Journal of Accounting and Economics 40, no. 1: 3-73.
Harris, M. S. 1998. The association between competition and managers' business segment
reporting decisions. Journal of Accounting Research, 111-128.
Healy, P., Serafeim, G., Srinivasan, S., and Yu, G. 2013. Market competition, earnings
management, and persistence in accounting profitability around the world. Review of
Accounting Studies, 1-28.
Holmstrom, B., 1982. Moral Hazard in Teams. The Bell Journal of Economics: 324–340.
Jegadeesh, N., and Livnat, J. 2006. Revenue surprises and stock returns. Journal of Accounting
and Economics, 41(1), 147-171.
Kadan, O., Madureira, L., Wang, R., and Zach., T. 2012. “Analysts’ Industry Expertise.” Journal
of Accounting and Economics 54 (2): 95–120.
31
Kasznik, R. and McNichols, M. F., 2002. “Does Meeting Earnings Expectations Matter?
Evidence from Analyst Forecast Revisions and Share Prices.” Journal of Accounting
Research 40 (3): 727–59.
Kolev, K., Marquardt, C. A., and McVay, S. E. 2008. SEC scrutiny and the evolution of non-
GAAP reporting. The Accounting Review, 83(1), 157-184.
Lee, C., Ma, P., Wang, C. 2014. “Search Based Peer Firms: Aggregating Investor Perceptions
through Internet Co-searches.” Journal of Financial Economics, forthcoming
Levitt, A. 1998 “The ‘Numbers Game” – Remarks of Chairman Arthur Levitt at the N.Y.U.
Center for Law and Business, New York, N.Y.
Litov, L. P., Moreton, P., and Zenger, T. R., 2012. “Corporate strategy, analyst coverage, and the
uniqueness paradox.” Management Science 58, no. 10: 1797-1815.
Maddala. G.S.. 1977, Econometrics, McGraw-Hill. New York.
Matsumoto, D. A. 2002. “Management’s Incentives to Avoid Negative Earnings Surprises.” The
Accounting Review 77 (3): 483–514.
McGahan, A. M., and Porter, M. E., 2002. “What do we know about variance in accounting
profitability?” Management Science 48, no. 7: 834-851.
Nelson, R. R., 1991. “Why do firms differ, and how does it matter?” Strategic management
journal 12, no. S2: 61-74.
Nissim, D., and Penman, S. 2001. Ratio analysis and equity valuation: From research to practice.
Review of Accounting Studies, 6, 109–154.
Nissim, D., and Penman, S. 2003. Financial statement analysis of leverage and how it informs
about profitability and price-to-book ratios. Review of Accounting Studies, 8(4), 531-560.
Orlik, T., 2012. “Lenovo’s Uncertain Prize for Taking First Place in Global PCs”, The Wall
Street Journal, 28 October,
Palmrose, Z. V. (1986). Audit fees and auditor size: Further evidence. Journal of Accounting
Research, 97-110.
Peteraf, M. A., 1993. “The cornerstones of competitive advantage: A resource‐based view.”
Strategic management journal 14, no. 3: 179-191.
Petersen, M. A., 2009. “Estimating standard errors in finance panel data sets: Comparing
approaches.” Review of financial studies 22, no. 1: 435-480.
32
Rogers, C., 2013. “Chrysler Earnings To Weigh On an IPO”, The Wall Street Journal, 28
October
Roychowdhury, S. 2006. “Earnings Management through Real Activities Manipulation.” Journal
of Accounting and Economics 42 (3): 335–70.
Rumelt, R. P., 1991. “How much does industry matter?” Strategic Management Journal 12, no.
3: 167-185.
Saloner, G., Shepard, A., and Podolny, J. 2001. “Strategic Management” New York: Wiley
Skinner, D. J. and Sloan, R. G. 2002. “Earnings Surprises, Growth Expectations, and Stock
Returns or Don’t Let an Earnings Torpedo Sink Your Portfolio.” Review of Accounting
Studies 7 (2-3): 289–312.
Sloan, R. G. 1996. “Do Stock Prices Fully Reflect Information in Accruals and Cash Flows
about Future Earnings?” The Accounting Review, 289–315.
Stewart, J. 2015. “How, and Why, Apple Overtook Microsoft”, The New York Times, January
29
Stickel, S. E. 1985. “The Effect of Value Line Investment Survey Rank Changes on Common
Stock Prices.” Journal of Financial Economics 14 (1): 121–43.
Stigler, G. 1961. The economics of information. Journal of Political Economy, 69, 213–225.
Waring, G. F., 1996. “Industry differences in the persistence of firm-specific returns.” The
American Economic Review: 1253-1265.
33
Appendix A. Variable Definitions
Variables Descriptions
∆Ranking_QTRi,t ∆Ranking_QTRi,t is measured by firm i’s realized ranking at the end of
quarter t, based on realized earnings, less firm i’s initial ranking at the
beginning of quarter t, based on expected earnings, divided by the number of
firms in the same industry. Realized ranking is equal to firm i’s performance
ranking within 6-digit GICS industry based on firms’ IBES-reported actual
earnings in quarter t plus interest expenses multiplied by one minus marginal
tax rates, divided by average total assets. Initial ranking is constructed in the
same way except for the use of consensus analyst median forecast at the
beginning of quarter t. If the consensus analyst forecast is missing, the
expected earnings of firm i is equal to the IBES-reported actual earnings in
quarter t-4, which assumes that expected earnings follow a seasonal random
walk (e.g., Freeman and Tse, 1989; Bernard and Thomas, 1990). The
consensus median EPS forecast is calculated based on individual analyst
forecasts, which are required to be reported within a 90-day window
preceding the daily date to ensure that our analyst consensus is not based on
stale forecasts. We exclude individual analyst forecasts if I/B/E/S excludes
the forecasts from calculating IBES-reported median EPS consensus. If the
daily median EPS consensus analyst forecast is missing, we supplement our
data by using IBES-reported median EPS consensus forecasts. Marginal tax
rate is assumed as the top statutory federal tax rate plus 2% average state tax
rate (Nissim and Penman, 2003).
Initial Ranking_QTRi,t Initial Ranking_QTRi,t is measured as firm i’s initial ranking within the same
industry at the beginning of quarter t, and described in detail above.
Realized Ranking_QTRi,t Realized Ranking_QTRi,t is measured as firm i’s performance ranking within
the same industry based on firms’ IBES-reported actual earnings in quarter t
and described in detail above.
STD_∆Ranking_QTRi,t STD_∆Ranking_QTRi,t is measured by the standard deviation of the
∆Ranking_QTRi,t variable using preceding 16 quarters observations (a
minimum of 8 quarters observations is required).
∆Ranking_EAi,t ∆Ranking_EAi,t is measured by firm i’s realized ranking at the earnings
announcement in quarter t less firm i’s initial ranking two days prior to the
earnings announcement in quarter t, divided by the number of firms in the
same industry. Realized (Initial) ranking is equal to firm i’s performance
ranking within 6-digit GICS industry based on firms’ realized (expected)
earnings plus interest expenses multiplied by one minus marginal tax rates,
divided by average total assets. Realized (expected) earnings are measured by
IBES-reported Actual EPS (either the consensus IBES median analyst
forecast or IBES-reported Actual EPS in quarter t-4 if the consensus IBES
median forecast is missing) multiplied by the number of shares outstanding. If
industry peer firms have already announced earnings, peer firms’ announced
earnings are used to determine firm i’s realized and initial rankings; otherwise
34
peer firms’ expected earnings are used.
Initial Ranking_EAi,t Initial Ranking_EAi,t is measured as firm i’s initial ranking based on expected
earnings within the same industry two days prior to the announcement of
earnings and described in detail above.
Realized Ranking_EAi,t Realized Ranking_EAi,t is measured as firm i’s performance ranking within
the same industry based on firms’ IBES-reported actual earnings in quarter t
and described in detail above.
STD_∆Ranking_EAi,t STD_∆Ranking_EAi,t is measured by the standard deviation of the ∆Rankingi,t
variable using past 16 quarters observations (a minimum of 8 quarters
observations is required). This variable is converted to range between 0 and 1.
Surprise_QTRi,t Surprise_QTRi,t is measured by firm i’s IBES-reported Actual EPS in quarter
t less expected EPS, which is used in the construction of the ∆Ranking_QTRi,t
variable, divided by stock price.
Surprise_EAi,t Surprise_EAi,t is firm i’s earnings surprise in quarter t as measured by firm i’s
IBES-reported Actual EPS less expected earnings, which is used in the
construction of the ∆Ranking_EAi,t variable divided by stock price.
Sizei,t Sizei,t is equal to the natural logarithm of firm i’s market value of equity at the
end of quarter t.
Book-to-Marketi,t Book-to-Marketi,t is measured by firm i’s book value of equity divided by the
market value of equity at the end of quarter t.
SalesGrowthi,t SalesGrowthi,t is equal to firm i’s net sales in quarter t divided by net sales in
quarter t-4.
Accrualsi,t Accruals is measured as firm i’s GAAP EPS less cash flows from operation
per share in quarter t divided by stock price.
3DayReti,t 3DayReti,t is firm i’s three-day buy-and-hold stock returns centered on the
earnings announcement in quarter t less three-day value-weighted CRSP
market returns over the same window.
ROAi,t ROAi,t is calculated as firm i’s IBES-reported Actual EPS in quarter t
multiplied by the number of shares outstanding plus interest expenses
multiplied by one minus marginal tax rates, divided by average total assets.
∆ROAi,t ∆ROAi,t is measured as firm i’s return on assets in quarter t less firm i’s return
on assets in quarter t-4.
AF_Revisioni,t AF_Revisioni,t is measured as the consensus median EPS forecast five days
after the announcement of earnings less the consensus median EPS forecast
five days prior to the announcement of earnings divided by the absolute value
of the latter for firm i in quarter t. The consensus median EPS forecast is
calculated based on individual analyst forecasts, which are required to be
reported within a 90-day window preceding the daily date.
∆RECi,t ∆RECi,t is measured as the difference between the consensus median stock
recommendation five days after the announcement of earnings less the
consensus median stock recommendation five days prior to the announcement
of earnings for firm i in quarter t. The consensus median stock
35
recommendation is calculated based on individual analyst recommendations,
which are required to be reported within a 90-day window preceding the daily
date. A standard set of recommendation maintained by IBES is as follows:
Strong Buy (1), Buy (2), Hold (3), Underperform (4), and Sell (5). Hence, we
multiply changes in stock recommendation by -1 to be consistent with the
AF_Revisioni,t variable.
36
Figure 1. Average Stock Returns for Performance Ranking Changes by Quintile
The figure above represents average stock returns for firms that experience increases and decreases from
the initial ranking to the realized ranking quintile surrounding earnings announcement. The “-4”, “-3”, “-
2”, and “-1” categories represent the average market responses for all firms that have a decrease of 4, 3, 2,
and 1 quintiles from the initial ranking to realized ranking quintiles. The “0” category represents the
average market responses for all firms that experience not change from the initial ranking to realized
ranking quintile. The “1”, “2”, “3”, and “4” categories represent the average market responses for all
firms that have an increase of 1, 2, 3, and 4 quintiles from the initial ranking to realized ranking quintiles.
-0.050
-0.040
-0.030
-0.020
-0.010
0.000
0.010
0.020
0.030
0.040
0.050
0.060
-4 -3 -2 -1 0 +1 +2 +3 +4
Average Stock Return
37
Figure 2. Average Analyst Responses for Performance Ranking Changes by Quintile
The figure above represents the average consensus analyst forecast revisions (left) and the average stock
recommendation changes (right) for firms that experience increases and decreases from the initial ranking
to the realized ranking quintile surrounding earnings announcement. The “-4”, “-3”, “-2”, and “-1”
categories represent the average market responses for all firms that have a decrease of 4, 3, 2, and 1
quintiles from the initial ranking to realized ranking quintiles. The “0” category represents the average
market responses for all firms that experience not change from the initial ranking to realized ranking
quintile. The “1”, “2”, “3”, and “4” categories represent the average market responses for all firms that
have an increase of 1, 2, 3, and 4 quintiles from the initial ranking to realized ranking quintiles.
-0.400
-0.300
-0.200
-0.100
0.000
0.100
0.200
-4 -3 -2 -1 0 +1 +2 +3 +4
Average Forecast Revisions
-0.080
-0.060
-0.040
-0.020
0.000
0.020
0.040
0.060
-4 -3 -2 -1 0 +1 +2 +3 +4
Average Recommendation Changes
38
Table 1 Descriptive statistics
N Mean Std P25 Median P75
Ranking variables (Fiscal Quarter)
∆Ranking_QTRi,t 198,467 0.001 0.222 -0.082 0.001 0.089
Initial Ranking_QTRi,t 198,467 0.522 0.281 0.284 0.527 0.764
Realized Ranking_QTRi,t 198,467 0.523 0.280 0.287 0.529 0.763
STD_∆Ranking_QTRi,t 198,467 0.188 0.116 0.102 0.161 0.246
# of Peer Firmsi,t 198,467 126.016 80.828 67 116 165
Ranking variables (Earnings Announcement Window)
∆Ranking_EAi,t 198,467 0.004 0.158 -0.024 0.000 0.041
Initial Ranking_EAi,t 198,467 0.516 0.278 0.281 0.518 0.753
Realized Ranking_EAi,t 198,467 0.520 0.288 0.270 0.528 0.771
STD_∆Ranking_EAi,t 198,467 0.115 0.098 0.045 0.084 0.155
# of Peer Firmsi,t 198,467 121.619 78.719 64 111 159
Dependent variables
ROAi,t+4 178,590 0.004 0.043 0.002 0.011 0.022
3DayReti,t 198,467 0.001 0.087 -0.040 -0.001 0.040
AF_Revisioni,t 151,200 -0.101 0.480 -0.100 0.000 0.017
∆RECi,t 111,027 -0.006 0.391 0.000 0.000 0.000
Control variables
Surprise_QTRi,t 198,467 -0.003 0.048 -0.006 0.001 0.006
Surprise_EAi,t 198,467 -0.004 0.043 -0.002 0.000 0.003
Sizei,t 198,467 6.199 1.985 4.779 6.159 7.528
Book-to-Marketi,t 198,467 0.665 0.631 0.294 0.515 0.837
SalesGrowthi,t 198,467 1.151 0.477 0.955 1.075 1.228
Accrualsi,t 198,467 -0.057 0.182 -0.082 -0.029 0.001
ROAi,t 198,467 0.003 0.045 0.002 0.011 0.022
∆ROAi,t 198,467 -0.001 0.032 -0.006 0.000 0.005
This table reports descriptive statistics for all sample firms with available information. The sample period ranges
from 1997 to 2013. All variables are defined in Appendix A and all continuous variables are winsorized at the 1 and
99th percentiles.
39
Table 2 Pearson (Above) / Spearman (Below) correlations
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
(1) ∆Ranking_QTRi,t -0.01 0.53 -0.01 0.50 0.30 0.01 -0.03 0.11 0.09 0.14 0.21 0.32 0.08 0.13 0.02
(2) STD_∆Ranking_QTRi,t 0.01
-0.01 0.70 -0.04 -0.05 -0.32 0.09 -0.01 -0.02 0.00 -0.04 -0.01 -0.06 -0.05 0.00
(3) ∆Ranking_EAi,t 0.46 0.00
-0.04 0.37 0.53 0.04 -0.06 0.12 0.10 0.20 0.21 0.40 0.06 0.17 0.04
(4) STD_∆Ranking_EAi,t 0.00 0.72 -0.02
-0.05 -0.08 -0.44 0.11 0.01 -0.01 -0.01 -0.13 -0.01 -0.14 -0.03 0.00
(5) Surprise_QTRi,t 0.72 -0.02 0.50 -0.03
0.67 0.09 -0.15 0.10 0.26 0.12 0.28 0.38 0.09 0.16 0.01
(6) Surprise_EAi,t 0.43 0.00 0.88 -0.02 0.55
0.12 -0.16 0.10 0.26 0.13 0.29 0.46 0.07 0.16 0.01
(7) Sizei,t 0.01 -0.32 0.06 -0.46 0.05 0.04
-0.36 0.05 0.06 0.01 0.34 0.06 0.30 0.10 -0.01
(8) Book-to-Marketi,t -0.02 0.15 -0.04 0.17 -0.08 -0.02 -0.33
-0.15 -0.27 0.02 -0.04 -0.06 -0.04 -0.10 0.00
(9) SalesGrowthi,t 0.12 -0.07 0.15 -0.07 0.18 0.15 0.14 -0.23
0.10 0.05 0.03 0.27 -0.05 0.07 0.00
(10) Accrualsi,t 0.08 -0.02 0.06 0.00 0.11 0.06 -0.06 -0.22 0.12
0.01 -0.02 0.09 -0.08 0.08 0.00
(11) 3DayReti,t 0.15 -0.01 0.26 -0.03 0.16 0.27 0.03 0.01 0.10 -0.01
0.12 0.12 0.10 0.20 0.14
(12) ROAi,t 0.24 -0.10 0.27 -0.20 0.35 0.27 0.38 -0.30 0.27 -0.03 0.14
0.41 0.71 0.10 0.01
(13) ∆ROAi,t 0.37 0.00 0.45 0.00 0.43 0.48 0.08 -0.12 0.34 0.06 0.16 0.31
0.13 0.12 0.01
(14) ROAi,t+4 0.10 -0.11 0.12 -0.19 0.20 0.12 0.33 -0.28 0.16 -0.06 0.12 0.72 0.15
0.10 0.01
(15) AF_Revisioni,t 0.17 -0.04 0.30 -0.01 0.26 0.32 0.06 -0.08 0.12 0.05 0.27 0.14 0.22 0.14
0.06
(16) ∆RECi,t 0.02 0.00 0.05 0.00 0.01 0.05 -0.01 0.00 0.00 -0.01 0.12 0.01 0.01 0.02 0.07
This table presents Pearson (Above) / Spearman (Below) correlations. Correlations that are significant at 1% level are bolded. The sample period ranges from
1997 to 2013. All variables are defined in Appendix A and all continuous variables are winsorized at the 1 and 99th percentiles.
40
Table 3 Transition Matrix
Panel A Ranking Change Transition Matrix (Fiscal Quarter)
Realized Ranking_QTRi,t
Initial Ranking_QTRi,t 1st (Low) 2nd 3rd 4th 5th (High) Sum
1st (Low) 24,455 8,783 2,719 1,740 1,993 39,690
(0.62) (0.22) (0.07) (0.04) (0.05)
2nd 7,961 17,573 9,314 3,200 1,651 39,699
(0.2) (0.44) (0.23) (0.08) (0.04)
3rd 3,154 8,447 16,709 8,924 2,461 39,695
(0.08) (0.21) (0.42) (0.23) (0.06)
4th 2,092 3,276 8,398 18,310 7,650 39,726
(0.05) (0.08) (0.21) (0.46) (0.19)
5th (High) 2,011 1,645 2,542 7,520 25,939 39,657
(0.05) (0.04) (0.06) (0.19) (0.65)
39,673 39,724 39,682 39,694 39,694 198,467
Panel B Ranking Change Transition Matrix (Earnings Announcement Window)
Realized Ranking_EAi,t
Initial Ranking_EAi,t 1st (Low) 2nd 3rd 4th 5th (High) Sum
1st (Low) 30,014 6,704 1,249 764 948 39,679
(0.76) (0.17) (0.03) (0.02) (0.02)
2nd 5,759 25,381 6,506 1,354 708 39,708
(0.15) (0.64) (0.16) (0.03) (0.02)
3rd 1,826 5,456 25,396 5,830 1,117 39,625
(0.05) (0.14) (0.64) (0.15) (0.03)
4th 1,060 1,510 5,511 27,121 4,557 39,759
(0.03) (0.04) (0.14) (0.68) (0.11)
5th (High) 1,043 636 1,030 4,624 32,363 39,696
(0.03) (0.02) (0.03) (0.12) (0.82)
39,702 39,687 39,692 39,693 39,693 198,467
This table presents the transition matrix based on the initial and realized rankings of the firm. In Panel A, the first
column denotes the initial ranking quintiles calculated at the beginning of quarter t, based on expected earnings, and
the first row denotes the realized ranking quintiles of the firm at the end of the quarter t, based on realized earnings.
In Panel B, the first column denotes the initial ranking quintiles, calculated two days prior to the announcement of
earnings based on expected earnings, and the first row denotes the realized ranking quintiles of the firm at the
earnings announcement date based on the reported earnings. Transition likelihoods are presented in parentheses and
the diagonal is bolded.
41
Table 4 Changes in Performance Rankings and Earnings Persistence
ROAi,t+4 ROAi,t+8 ROAi,t+12
Variables (1) (2) (3) (4) (5)
ROAi,t 0.669*** 0.576*** 0.628***
0.552***
0.482***
(56.552) (34.582) (39.063)
(27.836)
(23.210)
D_∆Ranking_QTRi,t - -0.008*** 0.003***
0.002**
0.001
- (-16.063) (4.988)
(2.516)
(1.586)
ROAi,t × D_∆Ranking_QTRi,t - 0.254*** 0.143***
0.109***
0.113***
- (10.543) (6.837)
(4.198)
(4.598)
Surprise_QTRi,t - - -0.057*** -0.057*** -0.055***
- - (-12.684) (-10.576) (-8.772)
Sizei,t - - 0.001***
0.002***
0.002***
- - (15.313)
(12.672)
(11.592)
Book-to-Marketi,t - - -0.002***
-0.001***
-0.001**
- - (-6.072)
(-3.183)
(-2.142)
STD_∆Ranking_QTRi,t - - -0.009***
-0.010***
-0.011***
- - (-6.003) (-4.680) (-4.278)
Initial Ranking_QTRi,t - - 0.007*** 0.004*** 0.003**
- - (7.515)
(3.691)
(2.349)
∆ROAi,t - - -0.160***
-0.148***
-0.151***
- - (-18.469)
(-16.517)
(-13.946)
SalesGrowthi,t - - -0.001***
-0.002***
-0.002***
- - (-2.852)
(-3.573)
(-2.931)
Accrualsi,t - - -0.009***
-0.009***
-0.008***
- - (-8.927)
(-8.297)
(-8.249)
Constant 0.001*** 0.004*** -0.010***
-0.008***
-0.008***
(4.651) (15.902) (-8.418)
(-5.451)
(-4.253)
Industry-Quarter FE Yes Yes Yes
Yes
Yes
Observations 178,590 178,590 178,590
153,556
132,217
Adjusted R-squared 0.524 0.530 0.556 0.460 0.405
This table presents estimation results from the regression of return on assets in quarter t+τ (ROAi,t+τ) on return on
assets in quarter t (ROAi,t), the decile-ranked change in the firm’s performance ranking in fiscal quarter
(D_∆Ranking_QTRi,t), the interaction of the ROAi,t variable with the D_∆Ranking_QTRi,t variable, and control
variables. The regression also include industry-quarter fixed effects. In column (1), (2), and (3) the dependent
variable is return on assets in quarter t+4. In column (4) and (5), return on assets at quarter t+8 and t+12 are used as
the dependent variable, respectively. All variables are defined in the Appendix and all continuous variables are
winsorized at the 1 and 99th percentiles. Standard errors are clustered by quarter and firm. *, **, and *** represent
significance level at the 10%, 5%, and 1% respectively.
42
Table 5 Univariate analyses: Average Market Responses Surrounding Earnings Announcement
Panel A Average Stock Returns
Realized Ranking_EAi,t
Initial Ranking_EAi,t 1st (Low) 2nd 3rd 4th 5th (High)
1st (Low) -0.015 0.018 0.036 0.042 0.050
2nd -0.031 -0.001 0.027 0.044 0.048
3rd -0.032 -0.026 0.003 0.035 0.047
4th -0.040 -0.030 -0.022 0.005 0.039
5th (High) -0.037 -0.033 -0.026 -0.018 0.008
Panel B Average Consensus Analyst Forecast Revisions
Realized Ranking_EAi,t
Initial Ranking_EAi,t 1st (Low) 2nd 3rd 4th 5th (High)
1st (Low) -0.263 -0.042 0.125 0.114 0.037
2nd -0.453 -0.163 0.003 0.062 0.105
3rd -0.338 -0.255 -0.072 0.031 0.023
4th -0.250 -0.224 -0.121 -0.037 0.029
5th (High) -0.317 -0.224 -0.170 -0.101 -0.020
Panel C Average Stock Recommendation Changes
Realized Ranking_EAi,t
Initial Ranking_EAi,t 1st (Low) 2nd 3rd 4th 5th (High)
1st (Low) -0.008 0.014 -0.003 -0.013 0.035
2nd -0.037 -0.005 0.022 0.036 0.004
3rd -0.050 -0.032 -0.003 0.024 0.031
4th -0.077 -0.026 -0.019 0.006 0.018
5th (High) -0.056 -0.064 -0.027 -0.018 -0.002
This table presents averages of market responses to performance ranking changes surrounding earnings
announcement for firm i in quarter t. In all panels, the first column denotes the initial ranking quintiles, calculated
two days prior to the announcement of earnings based on expected earnings, and the first row denotes the realized
ranking quintiles of the firm at the earnings announcement date based on the reported earnings. Panel A reports
average stock returns. Panel B presents average consensus analyst forecast revisions. Panel C reports average stock
recommendation changes. Significance level at 5% is bolded.
43
Table 6 Investor response to changes in performance rankings
3DayReti,t
Variables (1) (2) (3) (4)
∆Ranking_EAi,t 0.108*** - 0.095*** 0.103***
(31.342) - (24.990) (27.595)
Surprise_EAi,t - 0.271*** 0.087*** 0.076***
- (19.144) (8.081) (6.658)
Sizei,t - - - -0.001***
- - - (-2.984)
Book-to-Marketi,t - - - 0.008***
- - - (13.199)
STD_∆Ranking_EAi,t - - - -0.005
- - - (-1.390)
Initial Ranking_EAi,t - - - 0.027***
- - - (19.062)
∆ROAi,t - - - 0.057***
- - - (5.202)
SalesGrowthi,t - - - 0.006***
- - - (8.321)
Accrualsi,t - - - -0.003
- - - (-1.348)
Constant 0.001 0.002*** 0.001 -0.021***
(0.766) (2.633) (1.253) (-11.453)
Industry-Quarter FE No No No Yes
Observations 198,467 198,467 198,467 198,467
Adjusted R-squared 0.038 0.018 0.040 0.063
Vuong test
Column (1) 3DayReti,t = α + β ∆Rankingi,t + ԑ
Z-Stat
Column (2) 3DayReti,t = α + β Surprisei,t + ԑ 21.49
This table presents estimation results from the regression of three-day buy-and-hold abnormal returns surrounding
earnings announcement (3DayReti,t) on the firm's performance ranking changes surrounding earnings announcement
(∆Ranking_EAi,t), and control variables. All variables are defined in the Appendix and all continuous variables are
winsorized at the 1 and 99th percentiles. Standard errors are clustered by quarter and firm. *, **, and *** represent
significance level at the 10%, 5%, and 1%, respectively.
44
Table 7 Analyst Response to Changes in Performance Ranking
Panel A Analyst Forecast Revisions
AF_Revisioni,t
Variables (1) (2) (3) (4)
∆Ranking_EAi,t 0.707*** - 0.501*** 0.591***
(25.788) - (17.406) (22.024)
Surprise_EA i,t - 2.724*** 1.874*** 1.112***
- (21.948) (16.105) (8.320)
Sizei,t - - - 0.008***
- - - (4.968)
Book-to-Marketi,t - - - -0.023***
- - - (-4.872)
STD_∆Ranking_EAi,t - - - -0.043*
- - - (-1.740)
Initial Ranking_EAi,t - - - 0.222***
- - - (21.526)
∆ROAi,t - - - 0.299***
- - - (3.880)
SalesGrowthi,t - - - 0.025***
- - - (5.619)
Accrualsi,t - - - 0.049**
- - - (2.210)
Constant -0.105*** -0.096*** -0.101*** -0.282***
(-17.585) (-16.508) (-17.862) (-18.917)
Industry-Quarter FE No No No Yes
Observations 151,200 151,200 151,200 151,200
Adjusted R-squared 0.028 0.026 0.038 0.098
Panel B Stock Recommendation Changes
∆RECi,t
Variables (1) (2) (3) (4)
∆Ranking_EAi,t 0.133*** - 0.140*** 0.147***
(10.803) - (9.494) (9.416)
Surprise_EA i,t - 0.197*** -0.068 -0.022
- (4.141) (-1.265) (-0.397)
Sizei,t - - - -0.002
- - - (-1.631)
Book-to-Marketi,t - - - 0.004
- - - (1.112)
STD_∆Ranking_EAi,t - - - -0.006
- - - (-0.229)
Initial Ranking_EAi,t - - - 0.014***
- - - (2.770)
∆ROAi,t - - - -0.055
- - - (-1.092)
45
SalesGrowthi,t - - - -0.001
- - - (-0.284)
Accrualsi,t - - - -0.018**
- - - (-2.534)
Constant -0.007** -0.005** -0.007** -0.006
(-2.509) (-2.023) (-2.569) (-0.627)
Industry-Quarter FE No No No Yes
Observations 111,027 111,027 111,027 111,027
Adjusted R-squared 0.001 0.000 0.001 0.006
This table presents estimation results from the regression of either consensus analyst forecast revision
(AF_Revisioni,t) in Panel A or changes in stock recommendation (∆RECi,t) in Panel B on the firm's performance
ranking changes surrounding earnings announcement (∆Ranking_EAi,t), and control variables. All variables are
defined in the Appendix and all continuous variables are winsorized at the 1 and 99th percentiles. Standard errors
are clustered by quarter and firm. *, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.
46
Table 8 The effect of volatility of ranking changes on market responses
3DayReti,t AF_Revisioni,t ∆RECi,t
Variables (1) (2) (3) (4) (5) (6)
Low High
Low High
Low High
∆Ranking_EAi,t 0.307*** 0.093***
1.396*** 0.534***
0.452*** 0.117***
(28.956) (25.736)
(21.743) (19.580)
(7.288) (8.038)
Surprise_EA i,t 0.063*** 0.084***
1.083*** 1.156***
-0.063 -0.028
(2.758) (6.363)
(3.513) (9.284)
(-0.600) (-0.434)
Sizei,t 0.000 -0.002***
0.009*** 0.005**
-0.001 -0.002
(0.440) (-4.985)
(5.485) (2.042)
(-0.795) (-1.539)
Book-to-Marketi,t 0.007*** 0.008***
-0.022*** -0.025***
0.003 0.004
(9.049) (9.918)
(-3.346) (-3.981)
(0.514) (0.969)
STD_∆Ranking_EAi,t 0.022 -0.007
-0.565*** 0.004
0.200* -0.014
(1.569) (-1.575)
(-4.725) (0.112)
(1.897) (-0.425)
Initial Ranking_EAi,t 0.020*** 0.034*** 0.190*** 0.255*** 0.011 0.018***
(13.222) (18.352) (14.953) (18.334) (1.396) (2.681)
∆ROAi,t 0.097*** 0.034**
0.508*** 0.136
-0.045 -0.054
(4.747) (2.540)
(3.749) (1.344)
(-0.416) (-0.869)
SalesGrowthi,t 0.004*** 0.007***
0.030*** 0.017***
0.001 -0.003
(5.232) (7.984)
(5.704) (2.884)
(0.300) (-0.858)
Accrualsi,t -0.004 -0.002
0.106*** 0.006
-0.040*** -0.005
(-1.380) (-0.919)
(4.352) (0.234)
(-2.618) (-0.429)
Constant -0.022*** -0.020***
-0.266*** -0.277***
-0.020 -0.000
(-9.695) (-7.656)
(-14.358) (-14.144)
(-1.498) (-0.001)
Industry-Quarter FE Yes Yes
Yes Yes
Yes Yes
Observations 99,234 99,233
75,600 75,600
55,514 55,513
Adjusted R-squared 0.068 0.076 0.112 0.095 0.006 0.010
Coefficient Diff Diff. p-Value
Diff. p-Value
Diff. p-Value
∆Ranking_EAi,t -0.214*** 0.000
-0.862*** 0.000
-0.334*** 0.000
Surprise_EAi,t 0.021 0.412 0.073 0.807 0.035 0.759
This table presents estimation results from the cross-sectional analyses based on the volatility of past performance
ranking changes, the STD_∆Ranking_EAi,t variable. The full sample is divided into two subsamples based on the
median value of the STD_∆Ranking_EAi,t variable and estimate two separate regressions using each subsample. The
statistical difference of estimated coefficients are presented at the bottom of the table. In column (1) and column (2),
the dependent variable is three day buy-and-hold abnormal stock returns surrounding earnings announcement
(3DayReti,t). In column (3) and (4), the dependent variable is consensus analyst forecast revisions surrounding
earnings announcement (AF_Revisioni,t). In column (5) and column (6), the dependent variable is changes in stock
recommendation surrounding earnings announcement (∆RECi,t). All variables are defined in the Appendix and all
continuous variables are winsorized at the 1 and 99th percentiles. Standard errors are clustered by quarter and firm.
*, **, and *** represent significance level at the 10%, 5%, and 1%, respectively.
47
Table 9 Robustness Check: Industry-Adjusted abnormal Stock Returns
Industry-Adjusted 3DayReti,t
Variables (1) (2) (3)
∆Ranking_EAi,t 0.111*** - 0.104***
(30.410) - (27.685)
Surprise_EA i,t - 0.231*** 0.074***
- (15.912) (6.367)
Sizei,t -0.001*** -0.000 -0.001***
(-3.332) (-1.081) (-3.571)
Book-to-Marketi,t 0.007*** 0.007*** 0.008***
(12.162) (11.516) (12.720)
STD_∆Ranking_EAi,t -0.004 -0.004 -0.003
(-1.135) (-1.180) (-0.788)
Initial Ranking_EAi,t 0.027*** 0.012*** 0.027***
(18.789) (10.561) (19.145)
∆ROAi,t 0.086*** 0.168*** 0.057***
(8.391) (14.136) (4.961)
SalesGrowthi,t 0.006*** 0.007*** 0.006***
(8.150) (9.798) (8.611)
Accrualsi,t 0.000 -0.006*** -0.003
(0.189) (-2.611) (-1.348)
Constant -0.021*** -0.017*** -0.022***
(-11.856) (-9.480) (-11.948)
Industry-Quarter FE Yes Yes Yes
Observations 193,804 193,804 193,804
Adjusted R-squared 0.051 0.029 0.052
This table presents estimation results from the regression of three-day buy-and-hold abnormal industry-adjusted
returns surrounding earnings announcement (Industry-Adjusted 3DayReti,t) on the firm's performance ranking
changes (∆Ranking_EAi,t), and control variables. The Industry-Adjusted 3DayReti,t variable is defined as the
cumulated three days equal-weighted average stock returns of all firms in the same GICS industry as firm i at the
earnings announcement date of firm i (excluding firm i in the industry) subtracted from three-days cumulative stock
returns of firm i. All variables are defined in the Appendix and all continuous variables are winsorized at the 1 and
99th percentiles. Standard errors are clustered by quarter and firm. *, **, and *** represent significance level at the
10%, 5%, and 1%, respectively.