SlideShare ist ein Scribd-Unternehmen logo
1 von 25
Downloaden Sie, um offline zu lesen
Accounting Research Center, Booth School of Business, University of Chicago



An Empirical Investigation of the Effect of Quarterly Earnings Announcement Timing on
Stock Returns
Author(s): William Kross and Douglas A. Schroeder
Reviewed work(s):
Source: Journal of Accounting Research, Vol. 22, No. 1 (Spring, 1984), pp. 153-176
Published by: Blackwell Publishing on behalf of Accounting Research Center, Booth School of Business,
University of Chicago
Stable URL: http://www.jstor.org/stable/2490706 .
Accessed: 27/02/2012 13:15

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.




                Blackwell Publishing and Accounting Research Center, Booth School of Business, University of Chicago are
                collaborating with JSTOR to digitize, preserve and extend access to Journal of Accounting Research.




http://www.jstor.org
Journal of Accounting Research
  Vol. 22 No. 1 Spring 1984
      Printed in U.S.A.




  An Empirical Investigation of the
    Effect of Quarterly Earnings
  Announcement Timing on Stock
               Returns

     WILLIAM            KROSS     AND     DOUGLAS           A. SCHROEDER*




1. Introduction
  This research examines both the association between quarterly an-
nouncement timing (early or late) and the type of news (good or bad)
reported,and the relationship between stock returns and timing around
the earnings announcement date. Recent research on announcement
timing (Givoly and Palmon [1982], Patell and Wolfson [1982], Kross
[1981], and Whittred [1980]) has provided evidence that delayed an-
nouncements of annual earnings more often convey bad news (i.e., lower
than expected earnings) than do early announcements. However, we
know of no study which has reportedevidence of the same phenomenon
for quarterlyearnings.Furthermore,   there is a limited amount of evidence
regardingthe reaction of market participants to announcement timing.
While three studies (Givoly and Palmon [1982], Kross [1982], and
Chambersand Penman [1984]) find that early (late) announcementsare
associated with higher (lower) abnormal returns or high (low) stock
return variability, relative to late (early) announcments, only Kross
[1982] controlled for the sign of the earnings forecast error and none
controlled for the magnitudeof the earnings forecast error.
  It is well accepted that stock returns are associated with the sign of
the earnings forecast error (EFE). Since announcement timing is also
   * Associate Professor and Assistant Professor, Purdue University. We wish to thank the
members of the Purdue University and the University of Chicago accounting seminars for
their comments. [Accepted for publication June 1983.]
                                          153
                                         Copyright ?, Institute of Professional Accounting 1984
154      JOURNAL OF ACCOUNTING RESEARCH, SPRING 1984

associated with the EFE this variable must be controlled if one is to
examine market reaction to announcement timing. Similarly, Beaver,
Clarke, and Wright [1979] reported that stock returns are also associated
with the magnitude of the earnings forecast error, so it is necessary to
control for forecast error magnitudes as well. This is because early (late)
announcers could be releasing extremely good (bad) news. Finally, in the
light of recent research which shows that stock returns around the
announcement date are inversely related to the size (market value) of
the firm (Atiase [1980] and Ro [1983]), like Chambers and Penman
[1984], we decided to control for the potentially confounding size effect.1
   Our objective, then, was to determine whether the association between
announcement timing and stock returns persists after controlling for the
sign and the magnitude of the earnings forecast error and firm size.
Briefly, our results show that early quarterly earnings announcements
(1) contain better news and (2) were associated with larger abnormal
returns relative to late announcements. These findings hold both for
large and small firms, for positive and negative EFEs, and for small
absolute values of the EFE.
   In section 2 we describe the procedures and the data used in the tests.
The lag and earnings expectations models used to classify firms into
reporting and earnings categories are discussed in section 3. The results
appear in section 4, followed by a summary and conclusions (section 5).


2. Procedure and Data
2.1 PROCEDURE
   First, we computed a time lag forecast error for each of 12 quarters on
the basis of a comparison of the actual quarterly announcement date
with a forecasted date. Second, we computed an earnings forecast error
for each firm and each quarter on the basis of a comparison of actual to
forecasted quarterly earnings. We then examined the earnings forecast
errors for the earliest and latest quarterly announcements for each firm
with the expectation that the earliest announcements (relative to expec-
tations) would have a higher (larger positive or smaller negative) median
EFE than observed for the latest announcements. Next, we categorized
each firm on the basis of both its lag forecast error (early, on time, late)
and its earnings forecast error (good news, bad news). This process
resulted in six distinct groups of firms: early-good, early-bad, on time-
good, on time-bad, late-good, late-bad. Finally we examined the abnormal
stock return behavior on the days surrounding the quarterly announce-
ments for all six groups of firms in order to determine whether the
announcement timing conveyed or was associated with information other
than that contained in the earnings number.
  'We want to thank the reviewer for making this suggestion.
EFFECT OF EARNINGS ANNOUNCEMENT                TIMING       155
                                    TABLE         1
                       Median Autocorrelations at Lags One-Four

   Autocorrelation         Raw                Residuals                Residuals
      at Lag             Time Lag         RandomWalk Model                     Model
                                                                  Autoregressive

         1                -.1712               +.1909                  -.1015
         2                -.0939               +.0517                  -.0069
         3                -.1887               -.0086                  -.0302
         4                +.5809               -.2832                  -.1310


2.2 THE SAMPLE
   Our total sample consisted of 297 NYSE and American Stock Exchange
firms that met the following conditions: (1) daily stock price data were
available for the years 1977-80 on the daily ISL (Investment Statistics
Laboratory) tapes produced by Standard and Poor's and Chase Econo-
metrics; (2) quarterly earnings from the third quarter of 1968 through
the third quarter of 1980 (the latest available at the time this study was
conducted) were available on the quarterly COMPUSTAT tapes; (3)
quarterly earnings announcements dates were available from the second
quarter of 1971 through the third quarter of 1980 on the quarterly
COMPUSTAT tapes; and (4) the fiscal year ended in December.
   These filters resulted in 3,564 observations-12 quarters for each of
297 firms. All observations were used when we examined the relationship
between announcement timing and the earnings forecast error. However,
missing stock return data caused us to eliminate one firm, yielding 3,552
observations for the examination of stock price behavior.

3. Models
3.1 ANNOUNCEMENT           LAG FORECAST MODELS

   A firm was classified as reporting early or late based on a comparison
of the actual time of announcement with the expected time of announce-
ment. The expected time of announcement was formulated via a time-
series analysis of each firm's reporting history. We examined the auto-
correlation functions for the 26 quarterly report time lags from the
second quarter of 1971 through the third quarter of 1977 for the 297
firms in our sample.2 The median autocorrelations for lags one through
four are presented in table 1.
   As one would expect, there is clearly a spike in the autocorrelation
function at lag four. The autocorrelation function of the fourth differ-
ences (a random walk model) still produced small spikes at lags one and
four. Since we expected fourth-quarter (annual) earnings to be reported
at longer lags, on average, than interim reports, we estimated an auto-
   2
     A lag was measured by the number of days elapsing from the end of the reporting
period to the earnings announcement.
156      W. KROSS AND D. A. SCHROEDER

regressive model with an indicator variable associated with fourth-
quarter announcements for predicting report time lags for each firm.
Model (1) predicts the report lag as follows:
             Lagiq   =   ai +   fj(Lagiq-4)        +   Yi(Q4)       + Fi(Lagiq-)        (1)

where:
  Lagiq= forecast of the number of days spanning the end of the quarter
and the earnings announcement for firm i in quarter q;
  Lagi-4 = actual number of days spanning the end of the quarter and
the earnings announcement for the same quarter of the preceding year;
  Lagi,_q = actual number of days spanning the end of the quarter and
the earnings announcement for the quarter immediately preceding
quarter q;
  Q4=   1 if quarter q is the fourth fiscal quarter, 0 otherwise;
  a, y,y, r = firm-specific parameters.
  The medians of the autocorrelation function of the residuals for model
(1) (column 4 in table 1) indicate that they are nearly white noise. This
model was used to forecast the report time lag for each of the next 12
quarters with a reestimation of the parameters after each quarter's
forecast.
  A second lag forecast model, model (2), was chosen as a benchmark
for comparison to model (1). This model is a random walk in which
report lags are predicted as:
                                    La~gq= Lagiq-4                                      (2)
where the terms are defined as before. As reported in column 3, table 1,
small spikes appear in the autocorrelation function at lags one and four
for this model. Nevertheless model (2) seemed a reasonable and appro-
priate benchmark for purposes of comparison with model (1). Each of
these models was used to forecast the announcement lag for each of 12
consecutive quarters beginning with the fourth quarter of 1977.
3.2 EARNINGS FORECAST MODEL
   The type of earnings news reported was classified vis-A-vis an extrap-
olation of each firm's quarterly earnings. We utilized the premier quar-
terly forecasting model proposed by Griffin [1977] and Watts [1975], the
parameters of which were estimated for each firm:
            Zq = Zq-4 + Zq-i        -   Zq-5   -       Oaq1     -   yaq4   +   Oyaq-5   (3)

where:
  Zq =   one-quarter-ahead forecast of EPS in quarter q,
  Zq =   actual EPS in quarter q,
   0=    a regular first-order moving average parameter,
   y=    a seasonal moving average parameter,
   a=    a serially uncorrelated error term.
EFFECT OF EARNINGS ANNOUNCEMENT            TIMING       157
The choice of a "common model" structure follows from Jenkins [1979]
as well as empirical studies of earnings. Jenkins suggests that for rela-
tively short time series a common model may be appropriate where the
environments generating the data have common features. In addition,
empirical studies such as Foster [1977], Lorek [1979], and Schroeder
[1980] provide evidence that time-series models generate quarterly fore-
casts that are superior to naive martingalelike models.
   For each firm, the 37 quarterly EPS figures (adjusted for stock divi-
dends and stock splits) from the third quarter of 1968 through the third
quarter of 1977 were used to estimate the parameters of this model. All
forecasts of interim and annual (fourth-quarter) EPS were one-quarter-
ahead forecasts generated by updating the model for each of the next 12
quarters.

3.3 STOCK RETURN MODEL
  In the evaluation of stock return response to earnings announcements,
we used the traditional market model posited by Sharpe [1963]:
                          Rik = ai + O2iRmk Like
                                          +                              (4)
The parameters of this    model were estimated using one year of daily
returns (approximately    252 observations) prior to the quarter of the
earnings announcement.    The parameter estimates were used to compute
the abnormal return for   days -2 through +2.

4. Results
4.1 RESULTS-TIMING       VERSUS   TYPE OF NEWS

   In this section we report on the nature of the report timing for both
interim and annual earnings and the association of report timing with
the EFE.
   Figures 1 and 2 depict the distributions of the actual reporting lags
(the number of days between the end of the fiscal period and the earnings
announcement), while figures 3-6 depict the unexpected (actual-ex-
pected) lag over four-(calendar) -day reporting intervals. The distribution
of the raw reporting lag is quite similar to that reported by Chambers
and Penman [1984]. Interim reports, reported in figure 1, cluster between
22 to 30 days after the quarter and are characterized by a distribution
that is skewed to the right. However, the distribution of annual an-
nouncements, reported in figure 2, seems more symmetric, with some
clustering of announcements between 28 and 40 days after the end of the
year. The distributions of the lag forecast errors, reported in figures 3-6,
are almost symmetrical for both annual and interim announcements. As
expected, all four distributions are characterized by a sharp spike at zero
with a similar number of observations above (late) and below (early) the
mean.
158           W. KROSS AND D. A. SCHROEDER




bays        -6    10    14    18    22    26        30     34   38      42    46     50   54   58   -62

       FIG. 1.-Frequency     distribution of reporting time lags: interim announcements.




   Days      .56 104   14_     18
                                -   224   264       30           1-38    42    -46   50   54   58     62


       FIG. 2.-Frequency     distribution of reporting time lags: annual announcements.

  Tests of the relationship between the unexpected lag and EFEs are
reported in table 2 for both interim and annual announcements. The raw
forecast error was first deflated by its standard error (se). Thus:
                                         EFE    =        (Zq    Zq)/se.
When EFE is positive (negative), earnings are greater (less) than ex-
pected.
   Panel A of table 2 shows the average earnings forecast error across all
firms from the earliest to the latest announcement (relative to expecta-
tions). The Wilcoxon statistic on the difference between the averages for
EFFECT OF EARNINGS ANNOUNCEMENT                             TIMING     159




                  A               *           (4)
               ,~,                I
                                  I
                                  *           (3)
               co                 I
              -I
D                  I              *           (1)


     0~~~~(0 I
     * v *H(4)
             I
     .
         L
         F I                      I
                                  *           (04)
                   I              **           (10)

                                  I


                                                (3)
                                  I**
                                  I

                                  ***********(125)
                            o     I
                                  ********************* (405))
                                  I
                                  I
                            )     ***********************                                (1214)
                                  I
               CI
0.                 o              ***********                     (533))
               iI
                                  I
L,           o'                   *****                       (235)
          ..I
o                                 I
         ..                       **                  (616)
              -I


                                  I
         D                        *




                                  I
                                  I
                        v         *           (4)
                            esI


                   T'                 IIILI


                                                        40)0               800   1200)    1600     200
                                  Freque-ncy
160        W. KROSS AND D. A. SCHROEDER




                       a (0          *         (4)
                                A*             (8)
                                     I (8I
                                     *

                                               (8)
                                     I
               4                     * (4I     (3)
      I'
           n               ,         *         (4)
                   U,                I

      S                              *         (18)

                                     I
      e.               '             **        (44)
                   CD                                                                    (1221
                                     I
           C               '             *****          (166)
                                     IFr
       l                             1
                       I             **************                     (5 23)
                                     I
      :L                             I
            5
                                     ********************************                    (1 221)
                                           ~~~(17

           o
                   ~                 I
                                     **********                   (376)
           S                         I
      a                              I
            t      '                 *******                  (237)
           S                         I
           -                         I
           e           o**                           ( 34 )
      S.                             I
      eD                             I
                   ,                 *         (1 7)
           ~'                        1

                : @                  *       ~(7)
                                     I

                   m                 *         (4 )
      ?u'                            I
                                     I
                                     *         (3)
                                     I
                                     I
                               IV*             (4)


                                                        400               800    12900        1600   2000
                                     F'requenlcy
EFFECT OF EARNINGS ANNOUNCEMENT                TIMING     161


                0,




                     A    **     (5)
                          I
                          I
                    I W   *      (2)
                H
cn         l

 I,                       **     (5)

     ::                   I
                          **     (8)
                          I

    _-~'-7                1      (87)
      -.                  I
..              I         **** (27)
                          I
                          I
                          ********            (71)
                          I
                          I
o               j         ***~*************            17
          cLI
                          I
               o          ************************************          ( 353)
                          I
               ?I
               ' '        *****************             (155)

< 00
                           I
                          ~Fq
               o?         *******           (59 )
0                         I
          cLI
                     e    ****         (25)
                          I
               ,_         *         (13)
                          I
o                         I
      :         -         **        (6)
                          I

                          I

               8v         *         (1)
     *                    I



                     I IllII
                                           100       200)        300   400        500
                          Frequency
162       W. KROSS AND D. A. SCHROEDER



                   (0




                            A            I
                                         **        (5)
                    tQ                   I
                                         I
                                X            **    (6)
                                    HI
                                         I
                        I
                        @                *         (3)

      I                                  I
               t                |        **        (8)
                                         I
           w                    I            **    (9)
           CLs
                                         I
               t                         ****      (25)
          _.                             1
                                                         2
               .                         I
      02                |                *******               (57)
      0                 Io
                                         I
                        I                *****************                    (163)

      ~0            121
                    o                    **t**************************************                (388)
                                         I
                                         I
               r            ,            ************                 (110)
                                         I
                                         I
                    0'o              ********                      (73)
          20~                       ~~~2
               I
      o 00     I
        'L ,o, ***                                  ( 20)
                                         FeuI
      02                                 I
                    H                    **         (12)
           02
                                         I
           02                            I
      0
      0             H_                   *         (2)
           v)                            I
                                         I
           02                            *         (2)
          *                              I
                                         I
                    v                    **        (2 )
                                         I


                                                             100              200     300   400
                                                                                                          500(
                                         Frequenc y
EFFECT OF EARNINGS ANNOUNCEMENT                                          TIMING   163




              t     *q




 ,-~~~       L bC --b    LO




                              cl                   '1' 1         '



'-~~~~~~~~~~~~~~~~~~~~~~~~-



      CTc"
             CCC         *          c         0   C)       Z**

cq~                ~          ~~~~.

  o                                 ~~~~~~~~~~~~~~~-o
  o                                     C~~~~~~~~~~~~~0C
            CC CeZCCC-




       LO    CCCI0



       2           ~                              CI                 e
             CCCCC




                         CI)Q           C4I       C)    1
             LO    M                                   Cl
              -q   cq          l2                 -

 cl                                     l

                   m~~~~~~~
164        W. KROSS AND D. A. SCHROEDER

the earliest and the latest announcing quarters is significant at a = .01
for both the autoregressive and the random walk models. The results of
the Page L test (which tests the null against the ordered alternative a,1>
  2>   **. *> 12 and thereby uses more of the sample information) is also
significant at a = .01.
   Panel B of table 2 shows the results when the three annual announce-
ments are omitted for each firm. Again, the results are significant and
consistent with the results based on all quarters. We conclude, therefore,
that earlier (later) quarterly announcements are characterized by higher
(lower) unexpected earnings. We next investigate whether this relation-
ship had any effect on stock returns.

4.2 RESULTS-ANNOUNCEMENT              TIMING AND STOCK
RETURNS

   Using daily data, the possibility of cross-sectional correlation in the
residuals existed due to earnings announcement clustering. For example,
as many as 39 firms announced their earnings on the same calendar date.
In order to mitigate the problem of cross-sectional correlation we gener-
ated controlled residuals by subtracting the residuals of a randomly
selected nonannouncing sample firm from that of each announcing firm
during the earnings announcement time period (day -2 through day +2).
The controlled residual was computed as follows:

                                  Vik =   fik -   fjk

where:
  Vik =    controlled residual for treatment firm i, day k;
   cik =   market model residual for treatment firm i, day k;
   cjk =   market model residual for control firm j, day k.
The only restriction applied in the selection of the control firm was that
the quarterly earnings announcement of the treatment firm did not occur
within seven calendar days of the announcement of its control firm. This
process was repeated for each firm for each quarter tested. The results
of all succeeding tests are reported using these controlled residuals
(hereafter residuals).3
   In order to assess the relationship between announcement timing and
stock returns, we had to control for the announcements' news effects on
those returns. We did so using a two-factor analysis of variance. One
factor classified firms by the type of news (good or bad) reported, while
the other represented the timing (early, on time, or late) of the announce-
ment. This approach allowed us to examine the timing effect indepen-
dently of the type of news reported. For test purposes we classified firms
as reporting on time if the actual announcement date was within ? one

  3We also conductedtests on the raw residualsof the treatment firms. Our conclusions
were not altered.
EFFECT OF EARNINGS ANNOUNCEMENT           TIMING        165
day of the expected announcement date. Announcements arriving earlier
(later) were categorized as early (late).
   The results for all observations are presented in figure 7 and table 3
for the autoregressive lag expectation model. The residuals are reported
as percentages. A glance at figure 7 reveals immediately that announce-
ment timing as well as the type of news had a distinct association with
the stock return residuals.
   Since the interaction between the two factors was never significant we
do not report it. Consistent with expectations, good news firms, with a
five-day cumulative average residual (CAR) of .83%, outperformed bad
news firms whose five-day CAR was -.97%. A stratification on timing
alone yielded a CAR on early firms of .43%, while late or on time
announcements had CARs of -.16% and -.27% respectively. Since earlier
announcements typically contain good news the better performance for
early firms is not surprising. However, when the sample is stratified into
news/timing categories the effect of announcement timing still persists.



   1.200 -
            .900   -~~-
                                           //

     .900                            /

                              A/




    -.300}                                             e



                          a:.QOO'                 '~~~~~~~~~~~~~~~~-            '      '




 -1.200



 -1.600-,
        -2.000                           -1.000            DRYS        1.()                2.000

  FIG. 7.-Cumulative averageresiduals (in percentages)aroundthe announcementdate
(day 0): good/early (GE), good/late (GL), good/on time (GOT), bad/early (BE), bad/late
(BL), bad/on time (BOT).
166   W. KROSS AND D. A. SCHROEDER



                                                                                                                                *
                                                                                                                                      LO
                                 cqc            t
                                                r    cv          cv    r,   cs             cv        Cq M         Lo   t        O     ce
                    >        4                       00                              f-              =
                             H                  m                      cq            r                    cv           u        o




                        T C        C             5               CD    CeD Ce        t5    CD        CD   t5      Ce            Ce    00

                                             I                    I I I               I        I+             I         10




              .)        +          O                                   00 CD          _     O        Ce C- C- CD OC
                        c          +                              I ++                   I + I+                 + +Lc
                        U                           I1+

              2              >          O   ~~~~~~
                                                 ~~~             O     O     O       n       ~~ ~~~~~~
                                                                                          ~~~~~~~~        O       O             L
                                                                                                                                >          _




                        >)                      cq    cq                                             cq   cll     cq   _q        00       t-




         E
        -<                                           +           .:II                    III              +++t-O-


                              Clz o,             t~-C~           m                   ClcO LC         LO    c      C
                                                                                                                  m    cT        c             l
                        CZ
         Z!
         C.                                                                                                                      *    *



                   Ce              +l-           ClC +C
                                                  I:                                 +         II+

          C.)




         -U              I                       Iz-Cl
                                                 O               O-                  >      %                          O              C'S
                                                                                                                                      h-

                                   0q            0U-~            U-~        00       L-00            C    0            Cl~0



                                   L
                                   O                  L+O        + 1         c       + 1
                                                                                     c                    c                     C?
                            r-- r--
           x~~~~~~~~~~~~~~~~~~~ L LO LO t- (s s.:
                                  ~O
          C.)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C




                                                                                          >)


                                            ,,*             ,E5                                                                      E-


                                       -H   z               E-                   Z
                                                                                 m                                          C
EFFECT OF EARNINGS ANNOUNCEMENT              TIMING       167
The CAR for good news announced early was 1.39%, but only .54% for
good news announced late. Early bad news produced a five-day CAR of
-.79%, compared to a -1.02% CAR when announced late. Thus, even
after controlling for the sign of the earnings forecast error, the timing
effect was still significant at day -2 through day 0 and for the five-day
CAR.
   The results of separate tests on annual and interim announcements
are reported in panels A and B of table 4. Both timing and news effects
were significant for the five-day CAR for both annual and interim
announcements, at day -2 for annual announcements and at day -2
through day 0 for interim announcements. These results lend strong
support to the notion that the timing effect is independent of the sign of
the EFE.
   Although the ANOVA results indicate that the timing effect was
distinct from the sign of the EFE, it is still plausible that it was a function
of the magnitude of the EFE. That is, "good news" firms might have
reported very good news when they announced early, but only moderately
good news when they announced late. Similarly, late "bad news" might
have been very bad compared to early bad news announcements. This is
a reasonable alternative hypothesis, particularly in the light of the study
by Beaver, Clarke, and Wright [1979] which reported a positive relation-
ship between stock returns and the magnitude of EFE.
   In order to address this issue, we subclassified each main group into a
moderate and an extreme subgroup-for example, moderately bad (neg-
ative EFE less than one standard error) and very bad (negative EFE
greater than one standard error in magnitude), with a similar dichotomy
for "good news" firms. If the report timing was a surrogate for magnitude
of EFE, we would not expect a relationship between announcement
timing and stock returns for moderate news classifications. As before, an
ANOVA was used, but only on the firms that reported moderately good
or bad news in early and late announcements.
   The test results in table 5 are not consistent with the proposition that
the timing effect was a proxy for the magnitude of the EFE. The timing
effect persisted at day -1 and day 0 and for the five-day CAR even
though, by construction, the magnitude of the EFE was small. Separate
examinations on annual and interim announcements, reported in panels
A and B of table 6, yielded similar results. Again the timing effect was
significant for the five-day CAR and for one or more of the days around
the announcement date. The news effect was very weak, as would be
expected since all extreme EFEs were omitted for these tests. These
results provide strong evidence that report timing was not a surrogate
for the magnitude of the earnings forecast error, but rather conveyed or
was correlated with information distinct from the sign and the magnitude
of EFE.
168   W. KROSS AND D. A. SCHROEDER




                                      _    o                  oo       ooo C                                                 o             oo c
                                                                                                                                              C                                C
                                      *         I             +I       +                       I                I I                        + I I



                                           o                       -
                               +
                                   CS
                                                     M
                                                                   .
                                                                       r-L(

                                                                            .
                                                                                L
                                                                                      .
                                                                                              m
                                                                                                   .            .
                                                                                                                    -        O m
                                                                                                                             .         .        .
                                                                                                                                                    r1 LC-o
                                                                                                                                                       .
                                                                                                                                                                     ~~~~~~~~~~~~~~~~
                                                                                                                                                                       C;      .
                ."        ..                                                                                                                                          .c           00



                               C                     OH                               fCf                   C                     Cf       o                                         o
                                                                                                                                                               Cf      c




                               +
                                                     00 .)
                                                      U0r              0                  '                                                                    (1-         (       a)
                                            I            II            ++ I                                     I'+ I                      4I                  I o6

                     o
                               CZ                   co        ol       o        co            Cm            m           (M        q        M          o-       M       cq           oo
         -N
                     =

          4.                                                                                                                                                          *
                     CE




                     0                                   1+
                     <U                    +             I +           + I + + I+ I +

                     ?
                     e         >
                               ;;>.        Cf        t        C9D      Ce        C)           C)            m           cq        q        L-         o         C9             1LO




                IS                                                                                                                                                             *




         _ A,                              C)            -             c         Nil          C             0           'I'd, r      M
                                                                                                                                     C                          -C1o
                                                                                                                                                                 t
           t                                CI           C1                     Co                                               C CC
                                            I



                                                    C9                                                 o
                                           o0                 CeD      0C         0O                   Zo               0M C)               0         0        0



                X                     CS                                                                                                                                            *~~~~~~~~~~
                                                    ~~~~~~.                           .            .            .            .         .        .                      .....
                                                                                                                                                           . . .. . .......

                                                    :~~~~~                            .            .            .                  .
                                                                                                                                  ::            .          .        ... . .             :    :::::




                                                    ~~~~~~.                           .            .            .            . . . . .. . . . . .. . . ..-




                               ;<               X z=          c H e a Rz i                                              4 i                 o         s ?                  E            EN
                                                :    >, O               ,        o                     >,
                                                                                                       zNO              m                  O                   O
t>C         C9           O cq               00
                                                     o                        mU~L
                                                                               5)CC&
                                                                                                     -              ~.4
                                                                                                                     C00O
                                  m~                                                                       rt--~00
                                                                                                         H o+ o+ c
                     0.
     .           o
                     4
                         :
                              +    I'       +:
                                           1+ + +*I* I'
                                                   I'                             I' I' r-                      t
                                                                                  II*-~+                       +     c+




         +           Iz-
                     C             M                  M
                                                     CO    -4M               L                      C
                                                                                                    oo         0~cm
         +~                       CO;                -i        CCCCO                          r-                          CIO          Ct

              :>         I        II II                   1+                      I         I+1           I*I




         +                                                                                    *+                                       oo
         >           C                               t-
                                                     CeD 00         4        C1 00            Ce         Cl0            0 Cl
      o~ CZ                                               o                  CeD o                  o                CeDc" C
 0~~~~~~~~~~~~~~~q0
      +
      4   I' +' I' +' +'
                                  r--     r--        r-        r-                             r--
                                                                                  I' +: I' + + +s -
                                                                                                               o)                           )




         0                                                                                          O
             0o
         ;>.         C'1           C9     C9         C-   Cq   r--           CM                     L
                                                                                                    O      C11       o~ o              Cf
CO




 E                           +     I' +              + I' I'                      I' I I' + +                                 a;

             I       O            C       z     ~~~~~~
                                                     r~   oo                 C1        LO     Loo C1       Cf        Cz       r        C




                                                                                                                              *        *        J
             CZ                                      r--~ CZ
                                                           C                                cq      m CO
         IZ                                                                                                               C       Cf
                     4:            C 4:              4:
                                                     oo         I'
                                                               C9            Cf O O                            o o C9




                      q
                             C9       I   I'd        c"   t- .CC9                 Cf   o~ o          z      . .C9C
                                                      m   oo 00%             -t        -t     -t    -t     -t    -t


             CS                                                                                                               *        *~~~~~


                                                                                                                                   .E               .O



                             *                  -
                                                E.                      Z.........*......
c
                           E-               cc.




I!       Z                      +           C    r+C-1
                                                  -
                                                                        C'-1
                                                                            +1                   + I +I |                 0o00
          +1                    0           c                      fo   m                             0    Lo             Ct


R~~~~~~~~~~~~~~~~~                                                                                                                *

   X~~~~~~~~~~~~~~~                                                                                                               *
                                I*L          I*I*                       LI*                      I-           660~. o I


                                            c1                cl            l                    m
                      >




         '-                          mttt             0                     m
                                                                            C                C-1 0         Lo      U:CT"t-

         >n                     +            I1+                            ++I               +I+IC

    C)            z                                                                                                           O
                                            cot-                        O        co          c0       C.   m       0      t




                      +t        oo               ms                                                   HC
                                                                                                      L        o
                                                                                                              CT




                                 I*          I            I                 I.I II               I            I       I



Q                     n         +            I +                            +I              1 + +I
                                                                                                 +                            ,L-




              I                 Q           t-        C%                    z C




                                o~                                               Lo                                       o
                                                                                             C

                                            oC~                         CC                           Co1oCC




                                       *~~~~               .                            ~    .             ~~~~                          O. O.


                                      Cl)                                                                                     I       I~~~~~~~~~~a)
                                                                                                                                         L




                                       .                  ..)                     ...


                                a,~E~c~ceZ
                                    m                                                   .Cm           m           O           3
>           o                       ~       o.
                                                            ;                  q   L-               >%
                                                                                                    L- 00                Lo
                       >            I             I1+               + 1                 .i              +I6oc




                   z       +~~~~
                               CT                CI        m        C-T c                               m C1             m      o




                   Q                I             I            1'    1-        1       I- 1             1           .?          ?



                               CT                          C                            so
         4CD       >                              m                        m
                                                                           LoC
                                    I            II II                         I       I     I +                 I




tk




                   >1                                                )Cl           t         'lC:               ?             t-s
                                                                                                                         00




                   I1                            cqc                (C         T   co               t       c            -0~.




L.                 n                         I    1+                I       II             I I              +oo

     X             l           ~~~C-1
                                    oo                     m        Cm: m               t- Lo               t:            e




                           z            o.       H o                o      o       o         C~~t(t
                                                                                               o



               -               -N                          -               n       >e




                               .'S
                   c                .              ,                                    *,-..'

                                H                c o                on
                                                       '                                                t C
Ev                 I+
                                                  1             +I                 +mI + +m1



                                                                                                  o            mT
                 +




                                         o    o             1                      Lo     0       o.    o   o

                                I
                                0                  I I          Lo+                     I +             ,Ioo



 a) C            >1                                             m         U1C1



           cn             +q                  00    C
                                                    t-          O     q              -
                                                                                   U:O             Q
                                                                                                  Lo


                                oI            C-                                              I       I+o       C
    Q CD         U              o~                  o.          LC                        m       0         o


                    +   ++
                <:~~~~~~~                                       +l                 +1             +1            *

 o          I;                                    U:1-1m                                                4   C
                                                                                                            m   L




r           =    o              %             H        e        e     o            t-     L       m o0




 I~~~~
  C2~                                                           CCL                t     t-C




                          >~~~~~~    o        CCHHC                                -U>




Eq                    l         o             t     cs          t     u:           ~~~~~~
                                                                                   u: Cl 't o                   bku


                                     H~~~ ~~
                                       Cfz C                              :        >U




                 '1                  W n               ()                     bs E n ? W
                                                                                   ,W W c
                      C                  *2                i3                 i=
                                                                              *2          m        1*                 C
                 ~~~ S                             ~            W~
                                                                _     _       Z                   O
                                     v        m                                                             CY. CY
4.3 RESULTS-ANNOUNCEMENT         TIMING AND FIRM SIZE

   Previous studies (Chambers and Penman [1984] and Ro [1983]) found
an inverse relationship between firm size and the absolute value of stock
returns around the earnings announcement date. A question naturally
arises about whether the timing effect phenomenon would be observable
for both large and small firms. In order to provide evidence on this issue
we took our "moderate news" subsample and split it into small and large
firms and conducted tests on each size substratum. All firms whose
market value was below (above) the medium market value at the end of
the second quarter of 1978 were defined as small (large) for the entire
sample period. The test results on large (small) firms are reported in
panel A (panel B) of table 7.
   As shown there, the timing effect persisted across both size categories.
For large firms (panel A), announcement timing was significantly related
to stock returns at day -2, day -1, (weakly) at day 0, and for the five-
day CAR. Late announcers of moderate news saw their share prices drop
by .89%, on average, as opposed to an increase of .30% for their early-
announcing counterparts. Announcement timing also affected moderate
good news firms, which had residual returns of .72% if they announced
early, but only .24% if they announced late.
   The test results on the small firms, reported in panel B of table 7, tell
a similar story. The announcement timing effect was significant at day
-1, day 0, and (weakly) for the five-day CAR. The share prices of firms
that announced moderately bad news late fell by .90%, on average, over
the five-day announcement period, whereas the early-announcing coun-
terparts fell by only .03%. Moderately good news announced early was
rewarded by a .31% abnormal return, whereas later-announced good news
was greeted with a .32% negative return. Thus, it appears that the "timing
effect" persisted for both large and small firms.

5. Summary and Conclusions
   Generally, we found that earnings announcement timing was associ-
ated with abnormal stock returns around the earnings announcement
date. Abnormal returns of firms that announced early (late) were sig-
nificantly higher (lower) than the returns of firms that announced late
(early). This general result is consistent with previous research by Givoly
and Palmon [1982], Kross [1982], and Chambers and Penman [1984];
however, these other studies did not completely control for potentially
confounding factors regarding the "timing effect." After controlling for
these, our results are remarkably consistent. The "timing effect" persisted
whether the earnings announcement (1) contained good news or bad
news, (2) was an annual or interim announcement, (3) was made by a
large or small firm, or (4) contained moderately good or moderately bad
news.
~~.o                 c~~~~~l
                                                                        0C
                                                                                                                                             C




                        ._c
                                  Cl~~~~~~~~~~~iCc!
                                  1       c                                                            cl     oO       c   cs                     l




                              +                                   m
                                                                  LC       Lo    CC                    O      m             m               LO




                              CS>)d                                             ~ls                    C)S            0O   cq       cq




                                              +                   Lo+               1+             I         +                     I,        CD
                   -      n                                                                                                                  *




                          cQ                  +                       I++           I +1                   I + I++             I O




         z             0~~~~~~~~~~~~~~~0
                        I >,0
                       C3 I               o
                                          CO                                    ~        0Cc           V           O       oo o              co
               8                          +                       1+             +1                    +I++                   N



^        Z
        L.~            ~ ~~~> ~ ~ ~ ~ ~ ~~~~~~~~L
                         ~~ ~               N N1l
                                              1               0         C<:         q    O                    q                     LNO




          Z~~~
                                      Q   +
                                          O                       + +            +) I'                 +      I + +




    *   _.>~           ~~ ~ ~ ~~~~~~~>
          a;~~~~~~~~~~~r
        COD                                                   -                                                    CJC5*



                                      %   t                   e         -;m
                                                                       CD n O                          mP


                                                                                                              q                         *


                                                  *       z                 *                                                  z        * cz




                              :           4-
                                          0
                                          ;               )m 4
                                                             z                                 b m- -j
                                                                                               = 0 )                       -e       3.C
                                                      _           z    r         >             3
                                                                                                                  O        C            CO
o                        LO           -            LCO                    C0
                                     o                                                                           m>C1C00
         >        +Lo                                          N,           o                    C.              o
                                                                                                                 0                    m            m




             Cl                       I                            I        +                                                                                o
                                                                                                                              CIIA CD
             +                       0                         C            CD           C
                                                                                         CD           D          0                CC
                                                                                                                             0 0 C1-
                                                                                                                                                                        O

                                      I+                                       II+                               ++                   II                     oo




             +                       C                         0o00o                     Ct-00                                       CC
                                                                                                                                      O            o                -




                  >                  CO                    ~~~COq                            l
                                                                                         CO 00                   m               CIA                   t - I-


    00                                                                                                                                                                  *

         3CL                                          o0                       0                 0                           cO       c10                    LO
                                                                                                                                                              -
             >                        4                            fq                    +1                      +?++ m 1- CDt




0                                                                                                                                                          *            *
                                     CO                        t                         LO          C
                                                                                                     0                       C0       ClLO                          Cl
                                                                               O0

                                          I                        l+                    +1                      +i4:ic~i




                                                           )                            *)C                                                                             O


                                     O
                                     L                         t~-00                     LO 0>1                      00C>>O                            lC9
                                      +                        OCI +                      I l0 +                     I           I+        +CL


                                                                                                                 s           00C9
                                     CO                        CIC C0                     O O                                                                               .

                                                  s                        CS CO         CO CS                  ~ ~~                                           ~*           *   II Ii I
                                          *.                           .                             ....-*                                                                         *   *


                  Cl~~~~~~~~~~~~~~~C

                            ..
                                              .                        .-e.......
                      ..-
                        .        -
                                                  .                        .       .         .           .               .        .        .                    *
                                                                                                                                                               0)..         *


                                              .           ~            .           .~        .           .               .                             .            *
                                                  z                    z           z         z           z      C            C        CO               <


                  m                                   '        t            ro               |           |    bD r           t         ?           ?           ''       *       CO
                                                          _Z                            E~                                       _             0           OC                   0

                                                                   3 Q                    >                   3Q                 Q                 Q v vD0
should affect stock returns. Indeed, we believe that the announcement
timing itself should have no effect, but rather is probably associated with
some other event that either is typically associated with a reporting delay
or is usually viewed as "bad" news. Loss contingency disclosures or CPA
switches could be two such types of events. Of course, additional research
is needed to explore this possibility.
   Because "timing" (or other events associated with it) can greatly
influence stock returns, we suggest that future studies on announcements
incorporate adjustments which control for announcement timing when
examining stock return responses to the announcements. Failure to do
this could bias, or induce noise into, the test results.
                                    REFERENCES
ATIASE, R. K. "Predisclosure Informational Asymmetries, Firm Capitalization, Financial
  Reports, and Security Price Behavior." Ph.D. dissertation, University of California,
  Berkeley, 1980.
BEAVER, W., R. CLARKE, AND W. WRIGHT. "The Association Between Unsystematic
  Security Returns and the Magnitude of Earnings Forecast Errors." Journal of Accounting
  Research (Autumn 1979): 316-40.
CHAMBERS, A. E., AND S. H. PENMAN. "Timeliness of Reporting and the Stock Price
  Reaction to Earnings Announcements." Journal of Accounting Research (Spring 1984):
  21-47.
FOSTER, G. "Quarterly Accounting Data: Time-Series Properties and Predictive Ability
  Returns." The Accounting Review (January 1977): 1-21.
GIVOLY, D., AND D. PALMON. "Timeliness of Annual Earnings Announcements: Some
  Empirical Evidence." The Accounting Review (July 1982): 486-508.
GRIFFIN, P. A. "The Time-Series Behavior of Quarterly Earnings: Preliminary Evidence."
  Journal of Accounting Research (Spring 1977): 71-83.
JENKINS, G. M. Practical Experiences with Modeling and Forecasting Time Series. San
  Francisco: Holden-Day, 1979.
KROSS, W. "Earnings and Announcement Time Lags." Journal of Business Research
  (September 1981): 267-81.
       . "Profitability, Earnings Announcement Time Lags, and Stock Prices." Journal of
  Business Finance and Accounting (Autumn 1982): 313-28.
LOREK, K. S. "Predicting Annual Net Earnings with Quarterly Earnings Time-Series
  Models." Journal of Accounting Research (Spring 1979): 190-204.
PATELL, J., AND M. WOLFSON. "Good News, Bad News, and the Timing of Corporate
  Disclosures." The Accounting Review (July 1982): 509-27.
Ro, B. "Firm Size and the Informational Asymmetry of Annual Earnings Announcements."
  Working paper, Purdue University, February 1983.
SCHROEDER, D. A. "Expectations Data and the Stochastic Properties of Accounting
  Numbers: A Multiple Time Series Methodology." Ph.D. dissertation, University of
  Kansas, 1980.
SHARPE, W. F. "A Simplified Model for Portfolio Analysis." Management Science (January
  1963): 277-93.
WATTS, R. L. "The Time Series Behavior of Quarterly Earnings." Working paper, Depart-
  ment of Commerce, University of Newcastle, 1975.
WHITTRED, G. P. "Audit Qualification and the Timeliness of Corporate Annual Reports."
  The Accounting Review (October 1980): 563-69.

Weitere ähnliche Inhalte

Andere mochten auch

Finance project
Finance projectFinance project
Finance projectZain Ali
 
Effect of operating, financial and total leverage on expected stock return an...
Effect of operating, financial and total leverage on expected stock return an...Effect of operating, financial and total leverage on expected stock return an...
Effect of operating, financial and total leverage on expected stock return an...Shoaib Lalani
 
Dubai stock market development and its effect on economic growth
Dubai stock market development and its effect on economic growthDubai stock market development and its effect on economic growth
Dubai stock market development and its effect on economic growthChenoy Ceil
 
Empirical study of on effect of corporate announcement on stock nehaviour
Empirical study of on effect of corporate announcement on stock nehaviourEmpirical study of on effect of corporate announcement on stock nehaviour
Empirical study of on effect of corporate announcement on stock nehaviournitishrocks143
 
Cause & effect relationshipt research paper qta -kashif ahmed saeed
Cause & effect relationshipt   research paper qta -kashif ahmed saeedCause & effect relationshipt   research paper qta -kashif ahmed saeed
Cause & effect relationshipt research paper qta -kashif ahmed saeedKashif Ahmed Saeed
 
Tagging methods for stock assessment and research in fisheries
Tagging methods for stock assessment and research in fisheriesTagging methods for stock assessment and research in fisheries
Tagging methods for stock assessment and research in fisheriesNazmul Ahmed Oli
 
A survey of day of the month effect in world stock markets 2
A survey of day of the month effect in world stock markets 2A survey of day of the month effect in world stock markets 2
A survey of day of the month effect in world stock markets 2IAEME Publication
 
Recruitment, mortality and exploitation rates estimate and stock assessment o...
Recruitment, mortality and exploitation rates estimate and stock assessment o...Recruitment, mortality and exploitation rates estimate and stock assessment o...
Recruitment, mortality and exploitation rates estimate and stock assessment o...sarmodou
 
Apostila matematica financeira
Apostila matematica financeiraApostila matematica financeira
Apostila matematica financeiraJ M
 

Andere mochten auch (10)

Finance project
Finance projectFinance project
Finance project
 
Effect of operating, financial and total leverage on expected stock return an...
Effect of operating, financial and total leverage on expected stock return an...Effect of operating, financial and total leverage on expected stock return an...
Effect of operating, financial and total leverage on expected stock return an...
 
Dubai stock market development and its effect on economic growth
Dubai stock market development and its effect on economic growthDubai stock market development and its effect on economic growth
Dubai stock market development and its effect on economic growth
 
Empirical study of on effect of corporate announcement on stock nehaviour
Empirical study of on effect of corporate announcement on stock nehaviourEmpirical study of on effect of corporate announcement on stock nehaviour
Empirical study of on effect of corporate announcement on stock nehaviour
 
Cause & effect relationshipt research paper qta -kashif ahmed saeed
Cause & effect relationshipt   research paper qta -kashif ahmed saeedCause & effect relationshipt   research paper qta -kashif ahmed saeed
Cause & effect relationshipt research paper qta -kashif ahmed saeed
 
Tagging methods for stock assessment and research in fisheries
Tagging methods for stock assessment and research in fisheriesTagging methods for stock assessment and research in fisheries
Tagging methods for stock assessment and research in fisheries
 
A survey of day of the month effect in world stock markets 2
A survey of day of the month effect in world stock markets 2A survey of day of the month effect in world stock markets 2
A survey of day of the month effect in world stock markets 2
 
Fdc economic bulletin - November 9, 2015
Fdc economic bulletin - November 9, 2015Fdc economic bulletin - November 9, 2015
Fdc economic bulletin - November 9, 2015
 
Recruitment, mortality and exploitation rates estimate and stock assessment o...
Recruitment, mortality and exploitation rates estimate and stock assessment o...Recruitment, mortality and exploitation rates estimate and stock assessment o...
Recruitment, mortality and exploitation rates estimate and stock assessment o...
 
Apostila matematica financeira
Apostila matematica financeiraApostila matematica financeira
Apostila matematica financeira
 

Ähnlich wie An empirical investigation of the effect of quarterly earnings announcement timing on stock returns

Stock liquidity 2
Stock liquidity 2Stock liquidity 2
Stock liquidity 2ZNiazi2
 
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...Business, Management and Economics Research
 
Criterion Feedback2 Mar Answers and essay were sloppy with no a
Criterion Feedback2 Mar Answers and essay were sloppy with no aCriterion Feedback2 Mar Answers and essay were sloppy with no a
Criterion Feedback2 Mar Answers and essay were sloppy with no aMargenePurnell14
 
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docxdrennanmicah
 
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growth
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growthFaj jan feb_2003_surprise_higher_dividends_higher_earnings_growth
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growthSarayut Wanasin
 
corporate ValuationModels for service.pdf
corporate ValuationModels for service.pdfcorporate ValuationModels for service.pdf
corporate ValuationModels for service.pdfshubhampandey429275
 
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...HafizArslan19
 
Predicting corporate business failure in the nigerian manufacturing industry
Predicting corporate business failure in the nigerian manufacturing industryPredicting corporate business failure in the nigerian manufacturing industry
Predicting corporate business failure in the nigerian manufacturing industryAlexander Decker
 
Determinants of audit delay in nigerian companies
Determinants of audit delay in nigerian companiesDeterminants of audit delay in nigerian companies
Determinants of audit delay in nigerian companiesAlexander Decker
 
Dotcom bubble and underpricing: conjectures and evidence
Dotcom bubble and underpricing: conjectures and evidenceDotcom bubble and underpricing: conjectures and evidence
Dotcom bubble and underpricing: conjectures and evidenceFGV Brazil
 
M&A Event Study: Men’s Wearhouse & Gap
M&A Event Study: Men’s Wearhouse & GapM&A Event Study: Men’s Wearhouse & Gap
M&A Event Study: Men’s Wearhouse & GapAnh Ho
 
Forecasting Economic Activity using Asset Prices
Forecasting Economic Activity using Asset PricesForecasting Economic Activity using Asset Prices
Forecasting Economic Activity using Asset PricesPanos Kouvelis
 
Statistics in Finance - M&A and GDP growth
Statistics in Finance - M&A and GDP growthStatistics in Finance - M&A and GDP growth
Statistics in Finance - M&A and GDP growthJean Lemercier
 
Case 2 final input
Case 2  final inputCase 2  final input
Case 2 final inputYifan Cui
 
Markit dividend forecasts and their value
Markit dividend forecasts and their valueMarkit dividend forecasts and their value
Markit dividend forecasts and their valueThomas Matheson
 
IRJET - Stock Recommendation System using Machine Learning Approache
IRJET - Stock Recommendation System using Machine Learning ApproacheIRJET - Stock Recommendation System using Machine Learning Approache
IRJET - Stock Recommendation System using Machine Learning ApproacheIRJET Journal
 

Ähnlich wie An empirical investigation of the effect of quarterly earnings announcement timing on stock returns (20)

Stock liquidity 2
Stock liquidity 2Stock liquidity 2
Stock liquidity 2
 
Event_studies
Event_studiesEvent_studies
Event_studies
 
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...
Undercover Boss: Stripping Away the Disguise to Analyze the Financial Perform...
 
Criterion Feedback2 Mar Answers and essay were sloppy with no a
Criterion Feedback2 Mar Answers and essay were sloppy with no aCriterion Feedback2 Mar Answers and essay were sloppy with no a
Criterion Feedback2 Mar Answers and essay were sloppy with no a
 
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx
16Case 1.3 Just for Feet, Inc.Prepared byIvette Must.docx
 
Dfa and the error term 12-2001
Dfa and the error term 12-2001Dfa and the error term 12-2001
Dfa and the error term 12-2001
 
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growth
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growthFaj jan feb_2003_surprise_higher_dividends_higher_earnings_growth
Faj jan feb_2003_surprise_higher_dividends_higher_earnings_growth
 
corporate ValuationModels for service.pdf
corporate ValuationModels for service.pdfcorporate ValuationModels for service.pdf
corporate ValuationModels for service.pdf
 
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...
IPO long-run underperformance slides 2020-2021 U30273_9dd73850423bc1a9e6124aa...
 
Shareholder Activism
Shareholder ActivismShareholder Activism
Shareholder Activism
 
Predicting corporate business failure in the nigerian manufacturing industry
Predicting corporate business failure in the nigerian manufacturing industryPredicting corporate business failure in the nigerian manufacturing industry
Predicting corporate business failure in the nigerian manufacturing industry
 
Determinants of audit delay in nigerian companies
Determinants of audit delay in nigerian companiesDeterminants of audit delay in nigerian companies
Determinants of audit delay in nigerian companies
 
Dotcom bubble and underpricing: conjectures and evidence
Dotcom bubble and underpricing: conjectures and evidenceDotcom bubble and underpricing: conjectures and evidence
Dotcom bubble and underpricing: conjectures and evidence
 
M&A Event Study: Men’s Wearhouse & Gap
M&A Event Study: Men’s Wearhouse & GapM&A Event Study: Men’s Wearhouse & Gap
M&A Event Study: Men’s Wearhouse & Gap
 
247958[1]
247958[1]247958[1]
247958[1]
 
Forecasting Economic Activity using Asset Prices
Forecasting Economic Activity using Asset PricesForecasting Economic Activity using Asset Prices
Forecasting Economic Activity using Asset Prices
 
Statistics in Finance - M&A and GDP growth
Statistics in Finance - M&A and GDP growthStatistics in Finance - M&A and GDP growth
Statistics in Finance - M&A and GDP growth
 
Case 2 final input
Case 2  final inputCase 2  final input
Case 2 final input
 
Markit dividend forecasts and their value
Markit dividend forecasts and their valueMarkit dividend forecasts and their value
Markit dividend forecasts and their value
 
IRJET - Stock Recommendation System using Machine Learning Approache
IRJET - Stock Recommendation System using Machine Learning ApproacheIRJET - Stock Recommendation System using Machine Learning Approache
IRJET - Stock Recommendation System using Machine Learning Approache
 

Kürzlich hochgeladen

Call Girls in Gomti Nagar - 7388211116 - With room Service
Call Girls in Gomti Nagar - 7388211116  - With room ServiceCall Girls in Gomti Nagar - 7388211116  - With room Service
Call Girls in Gomti Nagar - 7388211116 - With room Servicediscovermytutordmt
 
John Halpern sued for sexual assault.pdf
John Halpern sued for sexual assault.pdfJohn Halpern sued for sexual assault.pdf
John Halpern sued for sexual assault.pdfAmzadHosen3
 
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Pune Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service AvailableDipal Arora
 
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...rajveerescorts2022
 
0183760ssssssssssssssssssssssssssss00101011 (27).pdf
0183760ssssssssssssssssssssssssssss00101011 (27).pdf0183760ssssssssssssssssssssssssssss00101011 (27).pdf
0183760ssssssssssssssssssssssssssss00101011 (27).pdfRenandantas16
 
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒anilsa9823
 
HONOR Veterans Event Keynote by Michael Hawkins
HONOR Veterans Event Keynote by Michael HawkinsHONOR Veterans Event Keynote by Michael Hawkins
HONOR Veterans Event Keynote by Michael HawkinsMichael W. Hawkins
 
Insurers' journeys to build a mastery in the IoT usage
Insurers' journeys to build a mastery in the IoT usageInsurers' journeys to build a mastery in the IoT usage
Insurers' journeys to build a mastery in the IoT usageMatteo Carbone
 
Famous Olympic Siblings from the 21st Century
Famous Olympic Siblings from the 21st CenturyFamous Olympic Siblings from the 21st Century
Famous Olympic Siblings from the 21st Centuryrwgiffor
 
M.C Lodges -- Guest House in Jhang.
M.C Lodges --  Guest House in Jhang.M.C Lodges --  Guest House in Jhang.
M.C Lodges -- Guest House in Jhang.Aaiza Hassan
 
Monte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMMonte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMRavindra Nath Shukla
 
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLMONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLSeo
 
Regression analysis: Simple Linear Regression Multiple Linear Regression
Regression analysis:  Simple Linear Regression Multiple Linear RegressionRegression analysis:  Simple Linear Regression Multiple Linear Regression
Regression analysis: Simple Linear Regression Multiple Linear RegressionRavindra Nath Shukla
 
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best Services
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best ServicesMysore Call Girls 8617370543 WhatsApp Number 24x7 Best Services
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best ServicesDipal Arora
 
Pharma Works Profile of Karan Communications
Pharma Works Profile of Karan CommunicationsPharma Works Profile of Karan Communications
Pharma Works Profile of Karan Communicationskarancommunications
 
How to Get Started in Social Media for Art League City
How to Get Started in Social Media for Art League CityHow to Get Started in Social Media for Art League City
How to Get Started in Social Media for Art League CityEric T. Tung
 
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...lizamodels9
 
Dr. Admir Softic_ presentation_Green Club_ENG.pdf
Dr. Admir Softic_ presentation_Green Club_ENG.pdfDr. Admir Softic_ presentation_Green Club_ENG.pdf
Dr. Admir Softic_ presentation_Green Club_ENG.pdfAdmir Softic
 
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service BangaloreCall Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangaloreamitlee9823
 
Monthly Social Media Update April 2024 pptx.pptx
Monthly Social Media Update April 2024 pptx.pptxMonthly Social Media Update April 2024 pptx.pptx
Monthly Social Media Update April 2024 pptx.pptxAndy Lambert
 

Kürzlich hochgeladen (20)

Call Girls in Gomti Nagar - 7388211116 - With room Service
Call Girls in Gomti Nagar - 7388211116  - With room ServiceCall Girls in Gomti Nagar - 7388211116  - With room Service
Call Girls in Gomti Nagar - 7388211116 - With room Service
 
John Halpern sued for sexual assault.pdf
John Halpern sued for sexual assault.pdfJohn Halpern sued for sexual assault.pdf
John Halpern sued for sexual assault.pdf
 
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service AvailableCall Girls Pune Just Call 9907093804 Top Class Call Girl Service Available
Call Girls Pune Just Call 9907093804 Top Class Call Girl Service Available
 
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...
👉Chandigarh Call Girls 👉9878799926👉Just Call👉Chandigarh Call Girl In Chandiga...
 
0183760ssssssssssssssssssssssssssss00101011 (27).pdf
0183760ssssssssssssssssssssssssssss00101011 (27).pdf0183760ssssssssssssssssssssssssssss00101011 (27).pdf
0183760ssssssssssssssssssssssssssss00101011 (27).pdf
 
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒VIP Call Girls In Saharaganj ( Lucknow  ) 🔝 8923113531 🔝  Cash Payment (COD) 👒
VIP Call Girls In Saharaganj ( Lucknow ) 🔝 8923113531 🔝 Cash Payment (COD) 👒
 
HONOR Veterans Event Keynote by Michael Hawkins
HONOR Veterans Event Keynote by Michael HawkinsHONOR Veterans Event Keynote by Michael Hawkins
HONOR Veterans Event Keynote by Michael Hawkins
 
Insurers' journeys to build a mastery in the IoT usage
Insurers' journeys to build a mastery in the IoT usageInsurers' journeys to build a mastery in the IoT usage
Insurers' journeys to build a mastery in the IoT usage
 
Famous Olympic Siblings from the 21st Century
Famous Olympic Siblings from the 21st CenturyFamous Olympic Siblings from the 21st Century
Famous Olympic Siblings from the 21st Century
 
M.C Lodges -- Guest House in Jhang.
M.C Lodges --  Guest House in Jhang.M.C Lodges --  Guest House in Jhang.
M.C Lodges -- Guest House in Jhang.
 
Monte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSMMonte Carlo simulation : Simulation using MCSM
Monte Carlo simulation : Simulation using MCSM
 
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRLMONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
MONA 98765-12871 CALL GIRLS IN LUDHIANA LUDHIANA CALL GIRL
 
Regression analysis: Simple Linear Regression Multiple Linear Regression
Regression analysis:  Simple Linear Regression Multiple Linear RegressionRegression analysis:  Simple Linear Regression Multiple Linear Regression
Regression analysis: Simple Linear Regression Multiple Linear Regression
 
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best Services
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best ServicesMysore Call Girls 8617370543 WhatsApp Number 24x7 Best Services
Mysore Call Girls 8617370543 WhatsApp Number 24x7 Best Services
 
Pharma Works Profile of Karan Communications
Pharma Works Profile of Karan CommunicationsPharma Works Profile of Karan Communications
Pharma Works Profile of Karan Communications
 
How to Get Started in Social Media for Art League City
How to Get Started in Social Media for Art League CityHow to Get Started in Social Media for Art League City
How to Get Started in Social Media for Art League City
 
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...
Call Girls In DLf Gurgaon ➥99902@11544 ( Best price)100% Genuine Escort In 24...
 
Dr. Admir Softic_ presentation_Green Club_ENG.pdf
Dr. Admir Softic_ presentation_Green Club_ENG.pdfDr. Admir Softic_ presentation_Green Club_ENG.pdf
Dr. Admir Softic_ presentation_Green Club_ENG.pdf
 
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service BangaloreCall Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
Call Girls Hebbal Just Call 👗 7737669865 👗 Top Class Call Girl Service Bangalore
 
Monthly Social Media Update April 2024 pptx.pptx
Monthly Social Media Update April 2024 pptx.pptxMonthly Social Media Update April 2024 pptx.pptx
Monthly Social Media Update April 2024 pptx.pptx
 

An empirical investigation of the effect of quarterly earnings announcement timing on stock returns

  • 1. Accounting Research Center, Booth School of Business, University of Chicago An Empirical Investigation of the Effect of Quarterly Earnings Announcement Timing on Stock Returns Author(s): William Kross and Douglas A. Schroeder Reviewed work(s): Source: Journal of Accounting Research, Vol. 22, No. 1 (Spring, 1984), pp. 153-176 Published by: Blackwell Publishing on behalf of Accounting Research Center, Booth School of Business, University of Chicago Stable URL: http://www.jstor.org/stable/2490706 . Accessed: 27/02/2012 13:15 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Blackwell Publishing and Accounting Research Center, Booth School of Business, University of Chicago are collaborating with JSTOR to digitize, preserve and extend access to Journal of Accounting Research. http://www.jstor.org
  • 2. Journal of Accounting Research Vol. 22 No. 1 Spring 1984 Printed in U.S.A. An Empirical Investigation of the Effect of Quarterly Earnings Announcement Timing on Stock Returns WILLIAM KROSS AND DOUGLAS A. SCHROEDER* 1. Introduction This research examines both the association between quarterly an- nouncement timing (early or late) and the type of news (good or bad) reported,and the relationship between stock returns and timing around the earnings announcement date. Recent research on announcement timing (Givoly and Palmon [1982], Patell and Wolfson [1982], Kross [1981], and Whittred [1980]) has provided evidence that delayed an- nouncements of annual earnings more often convey bad news (i.e., lower than expected earnings) than do early announcements. However, we know of no study which has reportedevidence of the same phenomenon for quarterlyearnings.Furthermore, there is a limited amount of evidence regardingthe reaction of market participants to announcement timing. While three studies (Givoly and Palmon [1982], Kross [1982], and Chambersand Penman [1984]) find that early (late) announcementsare associated with higher (lower) abnormal returns or high (low) stock return variability, relative to late (early) announcments, only Kross [1982] controlled for the sign of the earnings forecast error and none controlled for the magnitudeof the earnings forecast error. It is well accepted that stock returns are associated with the sign of the earnings forecast error (EFE). Since announcement timing is also * Associate Professor and Assistant Professor, Purdue University. We wish to thank the members of the Purdue University and the University of Chicago accounting seminars for their comments. [Accepted for publication June 1983.] 153 Copyright ?, Institute of Professional Accounting 1984
  • 3. 154 JOURNAL OF ACCOUNTING RESEARCH, SPRING 1984 associated with the EFE this variable must be controlled if one is to examine market reaction to announcement timing. Similarly, Beaver, Clarke, and Wright [1979] reported that stock returns are also associated with the magnitude of the earnings forecast error, so it is necessary to control for forecast error magnitudes as well. This is because early (late) announcers could be releasing extremely good (bad) news. Finally, in the light of recent research which shows that stock returns around the announcement date are inversely related to the size (market value) of the firm (Atiase [1980] and Ro [1983]), like Chambers and Penman [1984], we decided to control for the potentially confounding size effect.1 Our objective, then, was to determine whether the association between announcement timing and stock returns persists after controlling for the sign and the magnitude of the earnings forecast error and firm size. Briefly, our results show that early quarterly earnings announcements (1) contain better news and (2) were associated with larger abnormal returns relative to late announcements. These findings hold both for large and small firms, for positive and negative EFEs, and for small absolute values of the EFE. In section 2 we describe the procedures and the data used in the tests. The lag and earnings expectations models used to classify firms into reporting and earnings categories are discussed in section 3. The results appear in section 4, followed by a summary and conclusions (section 5). 2. Procedure and Data 2.1 PROCEDURE First, we computed a time lag forecast error for each of 12 quarters on the basis of a comparison of the actual quarterly announcement date with a forecasted date. Second, we computed an earnings forecast error for each firm and each quarter on the basis of a comparison of actual to forecasted quarterly earnings. We then examined the earnings forecast errors for the earliest and latest quarterly announcements for each firm with the expectation that the earliest announcements (relative to expec- tations) would have a higher (larger positive or smaller negative) median EFE than observed for the latest announcements. Next, we categorized each firm on the basis of both its lag forecast error (early, on time, late) and its earnings forecast error (good news, bad news). This process resulted in six distinct groups of firms: early-good, early-bad, on time- good, on time-bad, late-good, late-bad. Finally we examined the abnormal stock return behavior on the days surrounding the quarterly announce- ments for all six groups of firms in order to determine whether the announcement timing conveyed or was associated with information other than that contained in the earnings number. 'We want to thank the reviewer for making this suggestion.
  • 4. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 155 TABLE 1 Median Autocorrelations at Lags One-Four Autocorrelation Raw Residuals Residuals at Lag Time Lag RandomWalk Model Model Autoregressive 1 -.1712 +.1909 -.1015 2 -.0939 +.0517 -.0069 3 -.1887 -.0086 -.0302 4 +.5809 -.2832 -.1310 2.2 THE SAMPLE Our total sample consisted of 297 NYSE and American Stock Exchange firms that met the following conditions: (1) daily stock price data were available for the years 1977-80 on the daily ISL (Investment Statistics Laboratory) tapes produced by Standard and Poor's and Chase Econo- metrics; (2) quarterly earnings from the third quarter of 1968 through the third quarter of 1980 (the latest available at the time this study was conducted) were available on the quarterly COMPUSTAT tapes; (3) quarterly earnings announcements dates were available from the second quarter of 1971 through the third quarter of 1980 on the quarterly COMPUSTAT tapes; and (4) the fiscal year ended in December. These filters resulted in 3,564 observations-12 quarters for each of 297 firms. All observations were used when we examined the relationship between announcement timing and the earnings forecast error. However, missing stock return data caused us to eliminate one firm, yielding 3,552 observations for the examination of stock price behavior. 3. Models 3.1 ANNOUNCEMENT LAG FORECAST MODELS A firm was classified as reporting early or late based on a comparison of the actual time of announcement with the expected time of announce- ment. The expected time of announcement was formulated via a time- series analysis of each firm's reporting history. We examined the auto- correlation functions for the 26 quarterly report time lags from the second quarter of 1971 through the third quarter of 1977 for the 297 firms in our sample.2 The median autocorrelations for lags one through four are presented in table 1. As one would expect, there is clearly a spike in the autocorrelation function at lag four. The autocorrelation function of the fourth differ- ences (a random walk model) still produced small spikes at lags one and four. Since we expected fourth-quarter (annual) earnings to be reported at longer lags, on average, than interim reports, we estimated an auto- 2 A lag was measured by the number of days elapsing from the end of the reporting period to the earnings announcement.
  • 5. 156 W. KROSS AND D. A. SCHROEDER regressive model with an indicator variable associated with fourth- quarter announcements for predicting report time lags for each firm. Model (1) predicts the report lag as follows: Lagiq = ai + fj(Lagiq-4) + Yi(Q4) + Fi(Lagiq-) (1) where: Lagiq= forecast of the number of days spanning the end of the quarter and the earnings announcement for firm i in quarter q; Lagi-4 = actual number of days spanning the end of the quarter and the earnings announcement for the same quarter of the preceding year; Lagi,_q = actual number of days spanning the end of the quarter and the earnings announcement for the quarter immediately preceding quarter q; Q4= 1 if quarter q is the fourth fiscal quarter, 0 otherwise; a, y,y, r = firm-specific parameters. The medians of the autocorrelation function of the residuals for model (1) (column 4 in table 1) indicate that they are nearly white noise. This model was used to forecast the report time lag for each of the next 12 quarters with a reestimation of the parameters after each quarter's forecast. A second lag forecast model, model (2), was chosen as a benchmark for comparison to model (1). This model is a random walk in which report lags are predicted as: La~gq= Lagiq-4 (2) where the terms are defined as before. As reported in column 3, table 1, small spikes appear in the autocorrelation function at lags one and four for this model. Nevertheless model (2) seemed a reasonable and appro- priate benchmark for purposes of comparison with model (1). Each of these models was used to forecast the announcement lag for each of 12 consecutive quarters beginning with the fourth quarter of 1977. 3.2 EARNINGS FORECAST MODEL The type of earnings news reported was classified vis-A-vis an extrap- olation of each firm's quarterly earnings. We utilized the premier quar- terly forecasting model proposed by Griffin [1977] and Watts [1975], the parameters of which were estimated for each firm: Zq = Zq-4 + Zq-i - Zq-5 - Oaq1 - yaq4 + Oyaq-5 (3) where: Zq = one-quarter-ahead forecast of EPS in quarter q, Zq = actual EPS in quarter q, 0= a regular first-order moving average parameter, y= a seasonal moving average parameter, a= a serially uncorrelated error term.
  • 6. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 157 The choice of a "common model" structure follows from Jenkins [1979] as well as empirical studies of earnings. Jenkins suggests that for rela- tively short time series a common model may be appropriate where the environments generating the data have common features. In addition, empirical studies such as Foster [1977], Lorek [1979], and Schroeder [1980] provide evidence that time-series models generate quarterly fore- casts that are superior to naive martingalelike models. For each firm, the 37 quarterly EPS figures (adjusted for stock divi- dends and stock splits) from the third quarter of 1968 through the third quarter of 1977 were used to estimate the parameters of this model. All forecasts of interim and annual (fourth-quarter) EPS were one-quarter- ahead forecasts generated by updating the model for each of the next 12 quarters. 3.3 STOCK RETURN MODEL In the evaluation of stock return response to earnings announcements, we used the traditional market model posited by Sharpe [1963]: Rik = ai + O2iRmk Like + (4) The parameters of this model were estimated using one year of daily returns (approximately 252 observations) prior to the quarter of the earnings announcement. The parameter estimates were used to compute the abnormal return for days -2 through +2. 4. Results 4.1 RESULTS-TIMING VERSUS TYPE OF NEWS In this section we report on the nature of the report timing for both interim and annual earnings and the association of report timing with the EFE. Figures 1 and 2 depict the distributions of the actual reporting lags (the number of days between the end of the fiscal period and the earnings announcement), while figures 3-6 depict the unexpected (actual-ex- pected) lag over four-(calendar) -day reporting intervals. The distribution of the raw reporting lag is quite similar to that reported by Chambers and Penman [1984]. Interim reports, reported in figure 1, cluster between 22 to 30 days after the quarter and are characterized by a distribution that is skewed to the right. However, the distribution of annual an- nouncements, reported in figure 2, seems more symmetric, with some clustering of announcements between 28 and 40 days after the end of the year. The distributions of the lag forecast errors, reported in figures 3-6, are almost symmetrical for both annual and interim announcements. As expected, all four distributions are characterized by a sharp spike at zero with a similar number of observations above (late) and below (early) the mean.
  • 7. 158 W. KROSS AND D. A. SCHROEDER bays -6 10 14 18 22 26 30 34 38 42 46 50 54 58 -62 FIG. 1.-Frequency distribution of reporting time lags: interim announcements. Days .56 104 14_ 18 - 224 264 30 1-38 42 -46 50 54 58 62 FIG. 2.-Frequency distribution of reporting time lags: annual announcements. Tests of the relationship between the unexpected lag and EFEs are reported in table 2 for both interim and annual announcements. The raw forecast error was first deflated by its standard error (se). Thus: EFE = (Zq Zq)/se. When EFE is positive (negative), earnings are greater (less) than ex- pected. Panel A of table 2 shows the average earnings forecast error across all firms from the earliest to the latest announcement (relative to expecta- tions). The Wilcoxon statistic on the difference between the averages for
  • 8. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 159 A * (4) ,~, I I * (3) co I -I D I * (1) 0~~~~(0 I * v *H(4) I . L F I I * (04) I ** (10) I (3) I** I ***********(125) o I ********************* (405)) I I ) *********************** (1214) I CI 0. o *********** (533)) iI I L, o' ***** (235) ..I o I .. ** (616) -I I D * I I v * (4) esI T' IIILI 40)0 800 1200) 1600 200 Freque-ncy
  • 9. 160 W. KROSS AND D. A. SCHROEDER a (0 * (4) A* (8) I (8I * (8) I 4 * (4I (3) I' n , * (4) U, I S * (18) I e. ' ** (44) CD (1221 I C ' ***** (166) IFr l 1 I ************** (5 23) I :L I 5 ******************************** (1 221) ~~~(17 o ~ I ********** (376) S I a I t ' ******* (237) S I - I e o** ( 34 ) S. I eD I , * (1 7) ~' 1 : @ * ~(7) I m * (4 ) ?u' I I * (3) I I IV* (4) 400 800 12900 1600 2000 F'requenlcy
  • 10. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 161 0, A ** (5) I I I W * (2) H cn l I, ** (5) :: I ** (8) I _-~'-7 1 (87) -. I .. I **** (27) I I ******** (71) I I o j ***~************* 17 cLI I o ************************************ ( 353) I ?I ' ' ***************** (155) < 00 I ~Fq o? ******* (59 ) 0 I cLI e **** (25) I ,_ * (13) I o I : - ** (6) I I 8v * (1) * I I IllII 100 200) 300 400 500 Frequency
  • 11. 162 W. KROSS AND D. A. SCHROEDER (0 A I ** (5) tQ I I X ** (6) HI I I @ * (3) I I t | ** (8) I w I ** (9) CLs I t **** (25) _. 1 2 . I 02 | ******* (57) 0 Io I I ***************** (163) ~0 121 o **t************************************** (388) I I r , ************ (110) I I 0'o ******** (73) 20~ ~~~2 I o 00 I 'L ,o, *** ( 20) FeuI 02 I H ** (12) 02 I 02 I 0 0 H_ * (2) v) I I 02 * (2) * I I v ** (2 ) I 100 200 300 400 500( Frequenc y
  • 12. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 163 t *q ,-~~~ L bC --b LO cl '1' 1 ' '-~~~~~~~~~~~~~~~~~~~~~~~~- CTc" CCC * c 0 C) Z** cq~ ~ ~~~~. o ~~~~~~~~~~~~~~~-o o C~~~~~~~~~~~~~0C CC CeZCCC- LO CCCI0 2 ~ CI e CCCCC CI)Q C4I C) 1 LO M Cl -q cq l2 - cl l m~~~~~~~
  • 13. 164 W. KROSS AND D. A. SCHROEDER the earliest and the latest announcing quarters is significant at a = .01 for both the autoregressive and the random walk models. The results of the Page L test (which tests the null against the ordered alternative a,1> 2> **. *> 12 and thereby uses more of the sample information) is also significant at a = .01. Panel B of table 2 shows the results when the three annual announce- ments are omitted for each firm. Again, the results are significant and consistent with the results based on all quarters. We conclude, therefore, that earlier (later) quarterly announcements are characterized by higher (lower) unexpected earnings. We next investigate whether this relation- ship had any effect on stock returns. 4.2 RESULTS-ANNOUNCEMENT TIMING AND STOCK RETURNS Using daily data, the possibility of cross-sectional correlation in the residuals existed due to earnings announcement clustering. For example, as many as 39 firms announced their earnings on the same calendar date. In order to mitigate the problem of cross-sectional correlation we gener- ated controlled residuals by subtracting the residuals of a randomly selected nonannouncing sample firm from that of each announcing firm during the earnings announcement time period (day -2 through day +2). The controlled residual was computed as follows: Vik = fik - fjk where: Vik = controlled residual for treatment firm i, day k; cik = market model residual for treatment firm i, day k; cjk = market model residual for control firm j, day k. The only restriction applied in the selection of the control firm was that the quarterly earnings announcement of the treatment firm did not occur within seven calendar days of the announcement of its control firm. This process was repeated for each firm for each quarter tested. The results of all succeeding tests are reported using these controlled residuals (hereafter residuals).3 In order to assess the relationship between announcement timing and stock returns, we had to control for the announcements' news effects on those returns. We did so using a two-factor analysis of variance. One factor classified firms by the type of news (good or bad) reported, while the other represented the timing (early, on time, or late) of the announce- ment. This approach allowed us to examine the timing effect indepen- dently of the type of news reported. For test purposes we classified firms as reporting on time if the actual announcement date was within ? one 3We also conductedtests on the raw residualsof the treatment firms. Our conclusions were not altered.
  • 14. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 165 day of the expected announcement date. Announcements arriving earlier (later) were categorized as early (late). The results for all observations are presented in figure 7 and table 3 for the autoregressive lag expectation model. The residuals are reported as percentages. A glance at figure 7 reveals immediately that announce- ment timing as well as the type of news had a distinct association with the stock return residuals. Since the interaction between the two factors was never significant we do not report it. Consistent with expectations, good news firms, with a five-day cumulative average residual (CAR) of .83%, outperformed bad news firms whose five-day CAR was -.97%. A stratification on timing alone yielded a CAR on early firms of .43%, while late or on time announcements had CARs of -.16% and -.27% respectively. Since earlier announcements typically contain good news the better performance for early firms is not surprising. However, when the sample is stratified into news/timing categories the effect of announcement timing still persists. 1.200 - .900 -~~- // .900 / A/ -.300} e a:.QOO' '~~~~~~~~~~~~~~~~- ' ' -1.200 -1.600-, -2.000 -1.000 DRYS 1.() 2.000 FIG. 7.-Cumulative averageresiduals (in percentages)aroundthe announcementdate (day 0): good/early (GE), good/late (GL), good/on time (GOT), bad/early (BE), bad/late (BL), bad/on time (BOT).
  • 15. 166 W. KROSS AND D. A. SCHROEDER * LO cqc t r cv cv r, cs cv Cq M Lo t O ce > 4 00 f- = H m cq r cv u o T C C 5 CD CeD Ce t5 CD CD t5 Ce Ce 00 I I I I I I+ I 10 .) + O 00 CD _ O Ce C- C- CD OC c + I ++ I + I+ + +Lc U I1+ 2 > O ~~~~~~ ~~~ O O O n ~~ ~~~~~~ ~~~~~~~~ O O L > _ >) cq cq cq cll cq _q 00 t- E -< + .:II III +++t-O- Clz o, t~-C~ m ClcO LC LO c C m cT c l CZ Z! C. * * Ce +l- ClC +C I: + II+ C.) -U I Iz-Cl O O- > % O C'S h- 0q 0U-~ U-~ 00 L-00 C 0 Cl~0 L O L+O + 1 c + 1 c c C? r-- r-- x~~~~~~~~~~~~~~~~~~~ L LO LO t- (s s.: ~O C.)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C >) ,,* ,E5 E- -H z E- Z m C
  • 16. EFFECT OF EARNINGS ANNOUNCEMENT TIMING 167 The CAR for good news announced early was 1.39%, but only .54% for good news announced late. Early bad news produced a five-day CAR of -.79%, compared to a -1.02% CAR when announced late. Thus, even after controlling for the sign of the earnings forecast error, the timing effect was still significant at day -2 through day 0 and for the five-day CAR. The results of separate tests on annual and interim announcements are reported in panels A and B of table 4. Both timing and news effects were significant for the five-day CAR for both annual and interim announcements, at day -2 for annual announcements and at day -2 through day 0 for interim announcements. These results lend strong support to the notion that the timing effect is independent of the sign of the EFE. Although the ANOVA results indicate that the timing effect was distinct from the sign of the EFE, it is still plausible that it was a function of the magnitude of the EFE. That is, "good news" firms might have reported very good news when they announced early, but only moderately good news when they announced late. Similarly, late "bad news" might have been very bad compared to early bad news announcements. This is a reasonable alternative hypothesis, particularly in the light of the study by Beaver, Clarke, and Wright [1979] which reported a positive relation- ship between stock returns and the magnitude of EFE. In order to address this issue, we subclassified each main group into a moderate and an extreme subgroup-for example, moderately bad (neg- ative EFE less than one standard error) and very bad (negative EFE greater than one standard error in magnitude), with a similar dichotomy for "good news" firms. If the report timing was a surrogate for magnitude of EFE, we would not expect a relationship between announcement timing and stock returns for moderate news classifications. As before, an ANOVA was used, but only on the firms that reported moderately good or bad news in early and late announcements. The test results in table 5 are not consistent with the proposition that the timing effect was a proxy for the magnitude of the EFE. The timing effect persisted at day -1 and day 0 and for the five-day CAR even though, by construction, the magnitude of the EFE was small. Separate examinations on annual and interim announcements, reported in panels A and B of table 6, yielded similar results. Again the timing effect was significant for the five-day CAR and for one or more of the days around the announcement date. The news effect was very weak, as would be expected since all extreme EFEs were omitted for these tests. These results provide strong evidence that report timing was not a surrogate for the magnitude of the earnings forecast error, but rather conveyed or was correlated with information distinct from the sign and the magnitude of EFE.
  • 17. 168 W. KROSS AND D. A. SCHROEDER _ o oo ooo C o oo c C C * I +I + I I I + I I o - + CS M . r-L( . L . m . . - O m . . . r1 LC-o . ~~~~~~~~~~~~~~~~ C; . ." .. .c 00 C OH fCf C Cf o o Cf c + 00 .) U0r 0 ' (1- ( a) I II ++ I I'+ I 4I I o6 o CZ co ol o co Cm m (M q M o- M cq oo -N = 4. * CE 0 1+ <U + I + + I + + I+ I + ? e > ;;>. Cf t C9D Ce C) C) m cq q L- o C9 1LO IS * _ A, C) - c Nil C 0 'I'd, r M C -C1o t t CI C1 Co C CC I C9 o o0 CeD 0C 0O Zo 0M C) 0 0 0 X CS *~~~~~~~~~~ ~~~~~~. . . . . . . ..... . . .. . ....... :~~~~~ . . . . :: . . ... . . : ::::: ~~~~~~. . . . . . . . .. . . . . .. . . ..- ;< X z= c H e a Rz i 4 i o s ? E EN : >, O , o >, zNO m O O
  • 18. t>C C9 O cq 00 o mU~L 5)CC& - ~.4 C00O m~ rt--~00 H o+ o+ c 0. . o 4 : + I' +: 1+ + +*I* I' I' I' I' r- t II*-~+ + c+ + Iz- C M M CO -4M L C oo 0~cm +~ CO; -i CCCCO r- CIO Ct :> I II II 1+ I I+1 I*I + *+ oo > C t- CeD 00 4 C1 00 Ce Cl0 0 Cl o~ CZ o CeD o o CeDc" C 0~~~~~~~~~~~~~~~q0 + 4 I' +' I' +' +' r-- r-- r- r- r-- I' +: I' + + +s - o) ) 0 O 0o ;>. C'1 C9 C9 C- Cq r-- CM L O C11 o~ o Cf CO E + I' + + I' I' I' I I' + + a; I O C z ~~~~~~ r~ oo C1 LO Loo C1 Cf Cz r C * * J CZ r--~ CZ C cq m CO IZ C Cf 4: C 4: 4: oo I' C9 Cf O O o o C9 q C9 I I'd c" t- .CC9 Cf o~ o z . .C9C m oo 00% -t -t -t -t -t -t CS * *~~~~~ .E .O * - E. Z.........*......
  • 19. c E- cc. I! Z + C r+C-1 - C'-1 +1 + I +I | 0o00 +1 0 c fo m 0 Lo Ct R~~~~~~~~~~~~~~~~~ * X~~~~~~~~~~~~~~~ * I*L I*I* LI* I- 660~. o I c1 cl l m > '- mttt 0 m C C-1 0 Lo U:CT"t- >n + I1+ ++I +I+IC C) z O cot- O co c0 C. m 0 t +t oo ms HC L o CT I* I I I.I II I I I Q n + I + +I 1 + +I + ,L- I Q t- C% z C o~ Lo o C oC~ CC Co1oCC *~~~~ . ~ . ~~~~ O. O. Cl) I I~~~~~~~~~~a) L . ..) ... a,~E~c~ceZ m .Cm m O 3
  • 20. > o ~ o. ; q L- >% L- 00 Lo > I I1+ + 1 .i +I6oc z +~~~~ CT CI m C-T c m C1 m o Q I I 1' 1- 1 I- 1 1 .? ? CT C so 4CD > m m LoC I II II I I I + I tk >1 )Cl t 'lC: ? t-s 00 I1 cqc (C T co t c -0~. L. n I 1+ I II I I +oo X l ~~~C-1 oo m Cm: m t- Lo t: e z o. H o o o o C~~t(t o - -N - n >e .'S c . , *,-..' H c o on ' t C
  • 21. Ev I+ 1 +I +mI + +m1 o mT + o o 1 Lo 0 o. o o I 0 I I Lo+ I + ,Ioo a) C >1 m U1C1 cn +q 00 C t- O q - U:O Q Lo oI C- I I+o C Q CD U o~ o. LC m 0 o + ++ <:~~~~~~~ +l +1 +1 * o I; U:1-1m 4 C m L r = o % H e e o t- L m o0 I~~~~ C2~ CCL t t-C >~~~~~~ o CCHHC -U> Eq l o t cs t u: ~~~~~~ u: Cl 't o bku H~~~ ~~ Cfz C : >U '1 W n () bs E n ? W ,W W c C *2 i3 i= *2 m 1* C ~~~ S ~ W~ _ _ Z O v m CY. CY
  • 22. 4.3 RESULTS-ANNOUNCEMENT TIMING AND FIRM SIZE Previous studies (Chambers and Penman [1984] and Ro [1983]) found an inverse relationship between firm size and the absolute value of stock returns around the earnings announcement date. A question naturally arises about whether the timing effect phenomenon would be observable for both large and small firms. In order to provide evidence on this issue we took our "moderate news" subsample and split it into small and large firms and conducted tests on each size substratum. All firms whose market value was below (above) the medium market value at the end of the second quarter of 1978 were defined as small (large) for the entire sample period. The test results on large (small) firms are reported in panel A (panel B) of table 7. As shown there, the timing effect persisted across both size categories. For large firms (panel A), announcement timing was significantly related to stock returns at day -2, day -1, (weakly) at day 0, and for the five- day CAR. Late announcers of moderate news saw their share prices drop by .89%, on average, as opposed to an increase of .30% for their early- announcing counterparts. Announcement timing also affected moderate good news firms, which had residual returns of .72% if they announced early, but only .24% if they announced late. The test results on the small firms, reported in panel B of table 7, tell a similar story. The announcement timing effect was significant at day -1, day 0, and (weakly) for the five-day CAR. The share prices of firms that announced moderately bad news late fell by .90%, on average, over the five-day announcement period, whereas the early-announcing coun- terparts fell by only .03%. Moderately good news announced early was rewarded by a .31% abnormal return, whereas later-announced good news was greeted with a .32% negative return. Thus, it appears that the "timing effect" persisted for both large and small firms. 5. Summary and Conclusions Generally, we found that earnings announcement timing was associ- ated with abnormal stock returns around the earnings announcement date. Abnormal returns of firms that announced early (late) were sig- nificantly higher (lower) than the returns of firms that announced late (early). This general result is consistent with previous research by Givoly and Palmon [1982], Kross [1982], and Chambers and Penman [1984]; however, these other studies did not completely control for potentially confounding factors regarding the "timing effect." After controlling for these, our results are remarkably consistent. The "timing effect" persisted whether the earnings announcement (1) contained good news or bad news, (2) was an annual or interim announcement, (3) was made by a large or small firm, or (4) contained moderately good or moderately bad news.
  • 23. ~~.o c~~~~~l 0C C ._c Cl~~~~~~~~~~~iCc! 1 c cl oO c cs l + m LC Lo CC O m m LO CS>)d ~ls C)S 0O cq cq + Lo+ 1+ I + I, CD - n * cQ + I++ I +1 I + I++ I O z 0~~~~~~~~~~~~~~~0 I >,0 C3 I o CO ~ 0Cc V O oo o co 8 + 1+ +1 +I++ N ^ Z L.~ ~ ~~~> ~ ~ ~ ~ ~ ~~~~~~~~L ~~ ~ N N1l 1 0 C<: q O q LNO Z~~~ Q + O + + +) I' + I + + * _.>~ ~~ ~ ~ ~~~~~~~> a;~~~~~~~~~~~r COD - CJC5* % t e -;m CD n O mP q * * z * z * cz : 4- 0 ; )m 4 z b m- -j = 0 ) -e 3.C _ z r > 3 O C CO
  • 24. o LO - LCO C0 o m>C1C00 > +Lo N, o C. o 0 m m Cl I I + o CIIA CD + 0 C CD C CD D 0 CC 0 0 C1- O I+ II+ ++ II oo + C 0o00o Ct-00 CC O o - > CO ~~~COq l CO 00 m CIA t - I- 00 * 3CL o0 0 0 cO c10 LO - > 4 fq +1 +?++ m 1- CDt 0 * * CO t LO C 0 C0 ClLO Cl O0 I l+ +1 +i4:ic~i ) *)C O O L t~-00 LO 0>1 00C>>O lC9 + OCI + I l0 + I I+ +CL s 00C9 CO CIC C0 O O . s CS CO CO CS ~ ~~ ~* * II Ii I *. . ....-* * * Cl~~~~~~~~~~~~~~~C .. . .-e....... ..- . - . . . . . . . . * 0).. * . ~ . .~ . . . . * z z z z z C C CO < m ' t ro | | bD r t ? ? '' * CO _Z E~ _ 0 OC 0 3 Q > 3Q Q Q v vD0
  • 25. should affect stock returns. Indeed, we believe that the announcement timing itself should have no effect, but rather is probably associated with some other event that either is typically associated with a reporting delay or is usually viewed as "bad" news. Loss contingency disclosures or CPA switches could be two such types of events. Of course, additional research is needed to explore this possibility. Because "timing" (or other events associated with it) can greatly influence stock returns, we suggest that future studies on announcements incorporate adjustments which control for announcement timing when examining stock return responses to the announcements. Failure to do this could bias, or induce noise into, the test results. REFERENCES ATIASE, R. K. "Predisclosure Informational Asymmetries, Firm Capitalization, Financial Reports, and Security Price Behavior." Ph.D. dissertation, University of California, Berkeley, 1980. BEAVER, W., R. CLARKE, AND W. WRIGHT. "The Association Between Unsystematic Security Returns and the Magnitude of Earnings Forecast Errors." Journal of Accounting Research (Autumn 1979): 316-40. CHAMBERS, A. E., AND S. H. PENMAN. "Timeliness of Reporting and the Stock Price Reaction to Earnings Announcements." Journal of Accounting Research (Spring 1984): 21-47. FOSTER, G. "Quarterly Accounting Data: Time-Series Properties and Predictive Ability Returns." The Accounting Review (January 1977): 1-21. GIVOLY, D., AND D. PALMON. "Timeliness of Annual Earnings Announcements: Some Empirical Evidence." The Accounting Review (July 1982): 486-508. GRIFFIN, P. A. "The Time-Series Behavior of Quarterly Earnings: Preliminary Evidence." Journal of Accounting Research (Spring 1977): 71-83. JENKINS, G. M. Practical Experiences with Modeling and Forecasting Time Series. San Francisco: Holden-Day, 1979. KROSS, W. "Earnings and Announcement Time Lags." Journal of Business Research (September 1981): 267-81. . "Profitability, Earnings Announcement Time Lags, and Stock Prices." Journal of Business Finance and Accounting (Autumn 1982): 313-28. LOREK, K. S. "Predicting Annual Net Earnings with Quarterly Earnings Time-Series Models." Journal of Accounting Research (Spring 1979): 190-204. PATELL, J., AND M. WOLFSON. "Good News, Bad News, and the Timing of Corporate Disclosures." The Accounting Review (July 1982): 509-27. Ro, B. "Firm Size and the Informational Asymmetry of Annual Earnings Announcements." Working paper, Purdue University, February 1983. SCHROEDER, D. A. "Expectations Data and the Stochastic Properties of Accounting Numbers: A Multiple Time Series Methodology." Ph.D. dissertation, University of Kansas, 1980. SHARPE, W. F. "A Simplified Model for Portfolio Analysis." Management Science (January 1963): 277-93. WATTS, R. L. "The Time Series Behavior of Quarterly Earnings." Working paper, Depart- ment of Commerce, University of Newcastle, 1975. WHITTRED, G. P. "Audit Qualification and the Timeliness of Corporate Annual Reports." The Accounting Review (October 1980): 563-69.