Description
Accounting accruals are at the heart of most
accounting systems. A basic premise of accrual
accounting is that it provides a more timely and
relevant performance measure than cash flows
through a better matching of revenues and
expenses. While some prior studies suggest
that managers use individual accrual-related
disclosure items in an opportunistic manner
Accounting Research Journal
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
Tony Kang
Article information:
To cite this document:
Tony Kang, (2005),"Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence", Accounting
Research J ournal, Vol. 18 Iss 1 pp. 6 - 12
Permanent link to this document:http://dx.doi.org/10.1108/10309610580000670
Downloaded on: 24 January 2016, At: 20:55 (PT)
References: this document contains references to 0 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 429 times since 2006*
Users who downloaded this article also downloaded:
Paul Coram, (2005),"Active Learning in Accounting: A Case Study in Preaching to the Unconverted", Accounting Research
J ournal, Vol. 18 Iss 1 pp. 13-20http://dx.doi.org/10.1108/10309610580000671
Paul Mather, Alan Ramsay, (2006),"The Effects of Board Characteristics on Earnings Management around Australian CEO
Changes", Accounting Research J ournal, Vol. 19 Iss 2 pp. 78-93http://dx.doi.org/10.1108/10309610680000680
Prodosh Simlai, (2014),"Firm characteristics, distress risk and average stock returns", Accounting Research J ournal, Vol. 27
Iss 2 pp. 101-123http://dx.doi.org/10.1108/ARJ -06-2012-0046
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
6
Earnings Prediction and the Role of
Accural-Related Disclosure:
International Evidence
Tony Kang
Faculty of Management, McGill University and
School of Accountancy, Singapore Management University
Abstract
Accounting accruals are at the heart of most
accounting systems. A basic premise of accrual
accounting is that it provides a more timely and
relevant performance measure than cash flows
through a better matching of revenues and
expenses. While some prior studies suggest
that managers use individual accrual-related
disclosure items in an opportunistic manner,
hindering market participants’ ability to
predict future firm performance, the market’s
expectation about future firm performance will
become more accurate and consistent under
accrual accounting if the market properly uses
such information to set expectations about
future firm performance. Consistent with this
idea, our evidence shows that the frequency
of accrual-related disclosure is positively
(negatively) associated with analysts’ forecast
accuracy (dispersion). We interpret this finding
as the presence of more detailed accrual-related
disclosure requirements enhancing the market
participants’ ability to predict earnings.
1. Introduction
Accrual accounting, which transforms cash
flows into accruals, is a prominent feature of
any accounting system. A basic premise of
accrual accounting performance measures is
that it reduces uncertainty about future earnings
by guiding investors to better assess firm value
and operating performance than operating cash
flows. Often, this is achieved by providing a
better matching of revenues and expenses than
cash accounting (see for instance, Dechow
1994, Cheng et al. 1996, Basu et al. 1998).
Acknowledgments: We thank Tim Brailsford (the editor)
and an anonymous reviewer for constructive comments.
Under this scenario, accounting accruals can
reduce uncertainty about future earnings and
make investors’ expectation about future
earnings to become more accurate and
consistent by providing useful information
about future earnings.
However, some studies report that this
period’s level of accounting accruals is
negatively related to the size of future stock
returns (see for instance, Sloan 1996) and that
security analysts’ earnings forecasts fail to
incorporate the predictable reversal pattern in
accrual earnings (Bradshaw et al. 2001),
implying that accounting accruals could in fact
increase uncertainty about future earnings,
contrary to the basic premise of an accrual
accounting system. For instance, Bradshaw et al.
(2001) examine the association between firm-
level accruals and the properties of analysts’
forecasts, and find that analysts’ forecast errors
increase with the level of accruals in firm’s
earnings. They claim that even professional
investment intermediaries do not alert investors
to the subsequent earnings problems associated
with high accruals. Consequently, they conclude
that investors do not anticipate fully the negative
implications of unusually high accruals.
Taken together, it is not clear whether the
extensiveness of accrual accounting improves
analysts’ ability to predict future firm
performance more accurately and consistently or
whether it merely provides an avenue for
managers to report the profitability of the
company in such a way that those reports hinder
the market participants’ ability to predict future
earnings.
1
This notion motivates our study.
1 An underlying assumption is that at least some accrual
items are not manipulated by managers in an
opportunistic manner.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
7
An alternative approach to examine this
issue, yet to be taken in prior studies, is to look
at this issue in a cross-country setting. Our
cross-country approach departs from the
conventional approach that uses ex-post firm-
level data in a particular country (i.e., mostly
the U.S.) in the sense that it focuses on the
extensiveness of accrual accounting rules in a
given jurisdiction, rather than realized firm-
level accrual numbers, which could be driven
by a number of managerial factors such as
contracting motives, agency issues, etc. Along
these lines, we claim that our approach is worth
an investigation and that it could provide
evidence which will complement the findings
of prior studies.
Using the accrual index developed by Hung
(2000) as a measure for the effectiveness of
accrual accounting system, we find, in a sample
of non-U.S. firms cross-listed in the U.S., that
the market participants predict future earnings
in a more consistent manner (i.e., the analysts’
forecast dispersion decreases) as the accrual
accounting system becomes more effective.
Thus, consistent with the premise of accrual
accounting, our finding suggests that earnings
predictability increases with the effectiveness
of an accrual accounting system.
The rest of this paper is organized as
follows. In the next section, we develop the
hypothesis. In Section III, we describe research
design and the sample. Subsequently in Section
IV, we discuss the results. We then conclude in
Section V.
2. Hypothesis Development
Accrual accounting systems are expected to
generate more decision-useful performance
measures (i.e., earnings) than just cash flows
because the accrual accounting systems are
better at matching revenues and expenses (see
for instance, Dechow 1994, Cheng et al. 1996).
Dechow (1994) hypothesizes and finds that the
importance of accruals decreases with the
performance measurement interval, that the
importance of accruals increases with the
volatility of the firm’s working capital
requirements and investment and financing
activities, and that the importance of accruals
increases with the length of the firm’s operating
cycle. Under each of these circumstances, she
finds that cash flows suffer more severely from
timing and matching problems that reduce their
ability to reflect firm performance. Cheng et al.
(1996) find that the incremental information
content of cash flows increases when earnings
are transitory. Specifically, they find that the
incremental information content of accounting
earnings and accruals decreases with a decrease
in the permanence of earnings.
Hung (2000) finds that the use of accrual
accounting (versus cash accounting) negatively
affects the value relevance of financial
statements in countries with weak shareholder
protection. Bradshaw et al. (2002) find that the
analysts’ earnings forecasts do not incorporate
the predictable future earnings declines
associated with high accruals. However, they
find no evidence that auditors signal the higher
likelihood of GAAP violations associated with
high accruals through either their audit
opinions or through auditor changes. They
conclude that analysts and auditors do not alert
investors to the future earnings problems
associated with high accruals.
Fairfield et al. (1996) find that earnings sub-
components (mostly accrual-related) have
predictive power for future earnings beyond
current earnings (suggesting that detailed
accrual-related disclosure requirements convey
valuable information about future firm
performance). Thus, to the extent that analysts
understand the implications of the accrual-
related disclosures for future firm performance
and use those earnings components to predict
future earnings, their forecasts for future
earnings will become more accurate as the
frequency of accrual-related accounting
standards increases. More formally, our first
hypothesis (in an alternative form) is:
H
1
: Analysts’ forecast accuracy will increase
with the effectiveness of an accrual
accounting system in an economy.
It is also possible that a more effective
accrual accounting system not only increases
the market participants’ ability to predict future
earnings more accurately, but also reduce
heterogeneous beliefs about future firm
prospects among them, decreasing uncertainty
about future earnings and enabling the market
to predict earnings in a more consistent
manner. Under this scenario, the magnitude of
forecast dispersion will also be related to the
effectiveness of an accrual accounting system.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
8
Thus, our second hypothesis (in an alternative
form) is:
2
H2: Analysts’ forecast dispersion will decrease
with the effectiveness of an accrual
accounting system in an economy.
3. Research Design and Sample
In this study, we define the use of accrual
accounting as the extent that the accounting
system deviates from a cash method of
accounting as in Hung (2000), and use an
accrual index developed by her to measure
the frequency of accrual-related accounting
standards. She constructs an accrual index for
each country by equally weighting 11 accrual-
related accounting standards that are directly
related to the timing differences between cash
receipt/disbursement and revenue/expense
recognition.
3
The index represents the degree to
which the accounting system moves away from
a cash method measure of performance. A
higher index value indicates higher use of
accrual accounting. These scores, along with
the number of firm observations for each
sample country, are reported in table 1. The
accrual index score ranges from 0.32 for Brazil
and Switzerland to 0.82 for Australia, Ireland,
Norway, and U.K. U.K. has the largest number
of firms included in the sample with 63 firms.
2 Taken together, if one posits that an accrual-based
performance measure is a better indicator of the
underlying firm value at a given point in time, this also
means that such a measure will also be a better predictor
of future firm performance (which is reflected in stock
price) than a strictly cash-based measure. Along these
lines, one may observe a stronger statistical association
between an accrual-based performance measure and
future firm performance than between a cash-based
measure and future profitability. Prior evidence on this
issue, however, is not conclusive.
3 The eleven items are based on whether: (1) goodwill is
capitalized; (2) equity method is used; (3) accelerated
depreciation is allowed; (4) purchased intangible is
capitalized; (5) developed intangible is capitalized; (6)
research and development expenditure is capitalized; (7)
interest is capitalized; (8) finance lease is capitalized;
(9) percentage of completion method is allowed; (10)
pension costs are accrued; (11) other post-retirement
benefits are accrued. Hung notes that this index is
created from the data in the 1993 International
Accounting Summaries by Coopers and Lybrand
(Coopers & Lybrand 1993). She assigns a weight of one
if an accounting standard applies specific accrual
methods. Refer to Hung (2001) for further details on
how this index is constructed.
We estimate the following set of regression
models to test our hypotheses. Since it is likely
that forecast dispersion is simultaneously
determined with forecast accuracy, we estimate
the following set of equations both separately
using OLS and simultaneously using the two-
stage least squares method. Forecast accuracy
(ACRCY) is defined as the negative of
the absolute difference between actual
earnings and the I/B/E/S median consensus
forecast at the fiscal year end month, deflated
by stock price (Lang and Lundholm
1996). Specifically, we compute forecast
accuracy at time t (ACRCY
t
) as in the
following:
1 t
t
FORECAST
?
is the mean
I/B/E/S consensus forecast of period t earnings
made at time t-1,
t
EPS is the actual earnings
per share before extraordinary items at time t,
taken from I/B/E/S, and
t
PRICE is the stock
price at the end of period t.
( )
1
1
1
?
?
?
? =
t
t
t
t
t
PRICE
EPS FORECAST
ACRCY
Forecast dispersion (i.e., DISP) is the
standard deviation of forecasts issued as of
fiscal year end month, as available in the
I/B/E/S Summary File. The accrual index score
(ACCR) is taken from Hung (2001). Firm size
(SIZE) is the log of market value of equity at
fiscal year end. Loss firm year (LOSS) is an
indicator variable that takes the value of one
(zero) if the firm reported a loss (non-loss)
during the year, as in Hwang et al. (1996).
Earnings surprise (SURPRISE) is this year’s
earnings per share less last year’s, deflated by
the stock price. Earnings volatility (EVOL) is
measured as the standard deviation in earnings
per share over the past five years.
ACRCY = !
0
+ !
1
ACCR + !
2
SIZE
+ !
3
LOSS + !
4
SURPRISE
+ !
5
DISP + " (1)
DISP = #
0
+ #
1
ACCR + #
2
SIZE
+ #
3
LOSS + #
4
SURPRISE
+ #
5
EVOL + " (2)
Our sample covers a subset of non-U.S.
firms cross-listed in the U.S. between 1992 and
1999. To form the sample, we start from the
Bank of New York American Depository
Receipt (ADR hereafter) list and identify firms
that have earnings forecast data and the control
variables available in 2001 I/B/E/S and
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
9
Compustat. We then eliminate firm
observations that do not have an accrual index
score in Hung. This procedure results in 1,074
firm-year observations from 362 distinct firms.
Using this sample has an advantage of being
able to control for differences in the reporting
environment across countries (i.e., all of the
sample firms reconcile to U.S. GAAP) as well
as for the enforcement environment (subject to
the Securities Act of 1934) that are known to
associate with the properties of analysts’
forecasts (see for instance, Ashbaugh and
Pincus 2001, Hope 2003).
Sample characteristics are reported in table
2. The mean and median forecast accuracies in
the entire sample are -0.031 and -0.012 and the
mean and median forecast dispersion are 0.211
and 0.100. The mean and median percentage
forecast errors (e.g., forecast error as a
percentage of the actual EPS) is 0.304 and
0.129 respectively. The mean and median
accrual index scores in the sample are around
0.627 and 0.640. Found in the panel B of the
same table is a correlation matrix among the
regression variables. Not surprisingly, most of
the variables are highly correlated with each
another, and especially with the forecast
variables. An interesting observation is that
while ACCR has a strong partial correlation
with the forecast variables, it is not strongly
correlated with the other determinants of the
forecast characteristics, suggesting that it may
contain information about forecast accuracy
that is not captured in the control variables.
Table 1
Home Country Distribution of Sample Firms
Home Country Number of Firms Accrual
Index Score
Argentina 11 0.64
Australia 20 0.82
Belgium 3 0.68
Brazil 22 0.32
Denmark 2 0.55
Finland 2 0.55
France 23 0.64
Germany 19 0.41
Greece 1 0.55
Hong Kong 28 0.64
Ireland 6 0.82
Italy 14 0.45
Japan 32 0.55
Korea 5 0.55
Mexico 43 0.55
Netherlands 24 0.73
New Zealand 5 0.73
Norway 4 0.82
Portugal 2 0.55
Singapore 7 0.64
South Africa 2 0.68
Spain 7 0.77
Sweden 10 0.59
Switzerland 7 0.32
United Kingdom 63 0.82
Total 362
A list of currently cross-listed securities (as of February 2000) was obtained from the Bank of New York Global
Investors Guide.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
10
Table 2
Descriptive Statistics and Correlation Matrix
Panel A: Descriptive Statistics
Variable Mean Median Std Dev Min Max
ACRCY -0.031 -0.012 0.053 -0.496 0.000
%FORCERROR 0.304 0.129 1.279 -4.158 11.333
DISP 0.211 0.100 0.350 0.000 3.320
ACCR 0.627 0.640 0.147 0.320 0.820
SIZE 7.888 8.181 1.855 -2.827 12.143
LOSS 0.083 0.000 0.276 0.000 1.000
SURPRISE -0.002 -0.005 0.078 -0.411 0.600
EVOL 0.615 0.358 0.997 0.000 8.904
Panel B: Correlation Matrix
ACRCY DISP ACCR SIZE LOSS SURPRISE EVOL
ACRCY
DISP -0.204
***
ACCR 0.116
***
-0.143
***
SIZE 0.375
***
-0.027
***
0.043
LOSS -0.502
***
0.101
***
-0.014 -0.168
***
SURPRISE -0.426
***
0.117
***
-0.001
-0.090
***
0.400
***
EVOL -0.081
***
0.490
***
-0.054
*
0.074
***
0.063
**
0.116
***
ACRCY is the accuracy in analysts’ earnings forecasts, measured as the forecast available as of the
beginning month of the fiscal year and the actual earnings per share for the year. DISP refers to dispersion
in analysts’ forecasts, and it is taken from 2001 I/B/E/S Summary File. %FORCERROR is percentage
forecast error, computed as ([Forecast-Actual EPS]/Actual EPS). ACCR refers to the accrual index score
taken from Hung (2001). SIZE is the log of market value of equity at fiscal year end month. LOSS is an
indicator variable that takes the value of one (zero) if the firm reported a loss (gain) during the year.
SURPRISE is this year’s earnings minus last years’ earnings, deflated by stock price. EVOL is the standard
deviation of earnings over the previous five years.
4. Results and Discussions
Because analysts’ forecast dispersion is
modeled endogeneously in equations (1) and
(2), we first estimate the two equations
separately using the OLS method, and then we
also estimate them jointly using the two-stage
least squares regression methodology. These
results are reported in table 3.
Consistent with the view that accounting
accruals provide useful information about
future earnings, our evidence shows that the
forecast accuracy (dispersion) of I/B/E/S/
analysts increases (decreases) with the degree
of effectiveness of accrual accounting system
of a firm’s country of domicile.
4
This
association is robust to the inclusion of various
firm- and country-level determinants of
analysts’ forecasts identified by prior studies.
In both models, ACCR is significant at
p
Accounting accruals are at the heart of most
accounting systems. A basic premise of accrual
accounting is that it provides a more timely and
relevant performance measure than cash flows
through a better matching of revenues and
expenses. While some prior studies suggest
that managers use individual accrual-related
disclosure items in an opportunistic manner
Accounting Research Journal
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
Tony Kang
Article information:
To cite this document:
Tony Kang, (2005),"Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence", Accounting
Research J ournal, Vol. 18 Iss 1 pp. 6 - 12
Permanent link to this document:http://dx.doi.org/10.1108/10309610580000670
Downloaded on: 24 January 2016, At: 20:55 (PT)
References: this document contains references to 0 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 429 times since 2006*
Users who downloaded this article also downloaded:
Paul Coram, (2005),"Active Learning in Accounting: A Case Study in Preaching to the Unconverted", Accounting Research
J ournal, Vol. 18 Iss 1 pp. 13-20http://dx.doi.org/10.1108/10309610580000671
Paul Mather, Alan Ramsay, (2006),"The Effects of Board Characteristics on Earnings Management around Australian CEO
Changes", Accounting Research J ournal, Vol. 19 Iss 2 pp. 78-93http://dx.doi.org/10.1108/10309610680000680
Prodosh Simlai, (2014),"Firm characteristics, distress risk and average stock returns", Accounting Research J ournal, Vol. 27
Iss 2 pp. 101-123http://dx.doi.org/10.1108/ARJ -06-2012-0046
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
6
Earnings Prediction and the Role of
Accural-Related Disclosure:
International Evidence
Tony Kang
Faculty of Management, McGill University and
School of Accountancy, Singapore Management University
Abstract
Accounting accruals are at the heart of most
accounting systems. A basic premise of accrual
accounting is that it provides a more timely and
relevant performance measure than cash flows
through a better matching of revenues and
expenses. While some prior studies suggest
that managers use individual accrual-related
disclosure items in an opportunistic manner,
hindering market participants’ ability to
predict future firm performance, the market’s
expectation about future firm performance will
become more accurate and consistent under
accrual accounting if the market properly uses
such information to set expectations about
future firm performance. Consistent with this
idea, our evidence shows that the frequency
of accrual-related disclosure is positively
(negatively) associated with analysts’ forecast
accuracy (dispersion). We interpret this finding
as the presence of more detailed accrual-related
disclosure requirements enhancing the market
participants’ ability to predict earnings.
1. Introduction
Accrual accounting, which transforms cash
flows into accruals, is a prominent feature of
any accounting system. A basic premise of
accrual accounting performance measures is
that it reduces uncertainty about future earnings
by guiding investors to better assess firm value
and operating performance than operating cash
flows. Often, this is achieved by providing a
better matching of revenues and expenses than
cash accounting (see for instance, Dechow
1994, Cheng et al. 1996, Basu et al. 1998).
Acknowledgments: We thank Tim Brailsford (the editor)
and an anonymous reviewer for constructive comments.
Under this scenario, accounting accruals can
reduce uncertainty about future earnings and
make investors’ expectation about future
earnings to become more accurate and
consistent by providing useful information
about future earnings.
However, some studies report that this
period’s level of accounting accruals is
negatively related to the size of future stock
returns (see for instance, Sloan 1996) and that
security analysts’ earnings forecasts fail to
incorporate the predictable reversal pattern in
accrual earnings (Bradshaw et al. 2001),
implying that accounting accruals could in fact
increase uncertainty about future earnings,
contrary to the basic premise of an accrual
accounting system. For instance, Bradshaw et al.
(2001) examine the association between firm-
level accruals and the properties of analysts’
forecasts, and find that analysts’ forecast errors
increase with the level of accruals in firm’s
earnings. They claim that even professional
investment intermediaries do not alert investors
to the subsequent earnings problems associated
with high accruals. Consequently, they conclude
that investors do not anticipate fully the negative
implications of unusually high accruals.
Taken together, it is not clear whether the
extensiveness of accrual accounting improves
analysts’ ability to predict future firm
performance more accurately and consistently or
whether it merely provides an avenue for
managers to report the profitability of the
company in such a way that those reports hinder
the market participants’ ability to predict future
earnings.
1
This notion motivates our study.
1 An underlying assumption is that at least some accrual
items are not manipulated by managers in an
opportunistic manner.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
7
An alternative approach to examine this
issue, yet to be taken in prior studies, is to look
at this issue in a cross-country setting. Our
cross-country approach departs from the
conventional approach that uses ex-post firm-
level data in a particular country (i.e., mostly
the U.S.) in the sense that it focuses on the
extensiveness of accrual accounting rules in a
given jurisdiction, rather than realized firm-
level accrual numbers, which could be driven
by a number of managerial factors such as
contracting motives, agency issues, etc. Along
these lines, we claim that our approach is worth
an investigation and that it could provide
evidence which will complement the findings
of prior studies.
Using the accrual index developed by Hung
(2000) as a measure for the effectiveness of
accrual accounting system, we find, in a sample
of non-U.S. firms cross-listed in the U.S., that
the market participants predict future earnings
in a more consistent manner (i.e., the analysts’
forecast dispersion decreases) as the accrual
accounting system becomes more effective.
Thus, consistent with the premise of accrual
accounting, our finding suggests that earnings
predictability increases with the effectiveness
of an accrual accounting system.
The rest of this paper is organized as
follows. In the next section, we develop the
hypothesis. In Section III, we describe research
design and the sample. Subsequently in Section
IV, we discuss the results. We then conclude in
Section V.
2. Hypothesis Development
Accrual accounting systems are expected to
generate more decision-useful performance
measures (i.e., earnings) than just cash flows
because the accrual accounting systems are
better at matching revenues and expenses (see
for instance, Dechow 1994, Cheng et al. 1996).
Dechow (1994) hypothesizes and finds that the
importance of accruals decreases with the
performance measurement interval, that the
importance of accruals increases with the
volatility of the firm’s working capital
requirements and investment and financing
activities, and that the importance of accruals
increases with the length of the firm’s operating
cycle. Under each of these circumstances, she
finds that cash flows suffer more severely from
timing and matching problems that reduce their
ability to reflect firm performance. Cheng et al.
(1996) find that the incremental information
content of cash flows increases when earnings
are transitory. Specifically, they find that the
incremental information content of accounting
earnings and accruals decreases with a decrease
in the permanence of earnings.
Hung (2000) finds that the use of accrual
accounting (versus cash accounting) negatively
affects the value relevance of financial
statements in countries with weak shareholder
protection. Bradshaw et al. (2002) find that the
analysts’ earnings forecasts do not incorporate
the predictable future earnings declines
associated with high accruals. However, they
find no evidence that auditors signal the higher
likelihood of GAAP violations associated with
high accruals through either their audit
opinions or through auditor changes. They
conclude that analysts and auditors do not alert
investors to the future earnings problems
associated with high accruals.
Fairfield et al. (1996) find that earnings sub-
components (mostly accrual-related) have
predictive power for future earnings beyond
current earnings (suggesting that detailed
accrual-related disclosure requirements convey
valuable information about future firm
performance). Thus, to the extent that analysts
understand the implications of the accrual-
related disclosures for future firm performance
and use those earnings components to predict
future earnings, their forecasts for future
earnings will become more accurate as the
frequency of accrual-related accounting
standards increases. More formally, our first
hypothesis (in an alternative form) is:
H
1
: Analysts’ forecast accuracy will increase
with the effectiveness of an accrual
accounting system in an economy.
It is also possible that a more effective
accrual accounting system not only increases
the market participants’ ability to predict future
earnings more accurately, but also reduce
heterogeneous beliefs about future firm
prospects among them, decreasing uncertainty
about future earnings and enabling the market
to predict earnings in a more consistent
manner. Under this scenario, the magnitude of
forecast dispersion will also be related to the
effectiveness of an accrual accounting system.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
8
Thus, our second hypothesis (in an alternative
form) is:
2
H2: Analysts’ forecast dispersion will decrease
with the effectiveness of an accrual
accounting system in an economy.
3. Research Design and Sample
In this study, we define the use of accrual
accounting as the extent that the accounting
system deviates from a cash method of
accounting as in Hung (2000), and use an
accrual index developed by her to measure
the frequency of accrual-related accounting
standards. She constructs an accrual index for
each country by equally weighting 11 accrual-
related accounting standards that are directly
related to the timing differences between cash
receipt/disbursement and revenue/expense
recognition.
3
The index represents the degree to
which the accounting system moves away from
a cash method measure of performance. A
higher index value indicates higher use of
accrual accounting. These scores, along with
the number of firm observations for each
sample country, are reported in table 1. The
accrual index score ranges from 0.32 for Brazil
and Switzerland to 0.82 for Australia, Ireland,
Norway, and U.K. U.K. has the largest number
of firms included in the sample with 63 firms.
2 Taken together, if one posits that an accrual-based
performance measure is a better indicator of the
underlying firm value at a given point in time, this also
means that such a measure will also be a better predictor
of future firm performance (which is reflected in stock
price) than a strictly cash-based measure. Along these
lines, one may observe a stronger statistical association
between an accrual-based performance measure and
future firm performance than between a cash-based
measure and future profitability. Prior evidence on this
issue, however, is not conclusive.
3 The eleven items are based on whether: (1) goodwill is
capitalized; (2) equity method is used; (3) accelerated
depreciation is allowed; (4) purchased intangible is
capitalized; (5) developed intangible is capitalized; (6)
research and development expenditure is capitalized; (7)
interest is capitalized; (8) finance lease is capitalized;
(9) percentage of completion method is allowed; (10)
pension costs are accrued; (11) other post-retirement
benefits are accrued. Hung notes that this index is
created from the data in the 1993 International
Accounting Summaries by Coopers and Lybrand
(Coopers & Lybrand 1993). She assigns a weight of one
if an accounting standard applies specific accrual
methods. Refer to Hung (2001) for further details on
how this index is constructed.
We estimate the following set of regression
models to test our hypotheses. Since it is likely
that forecast dispersion is simultaneously
determined with forecast accuracy, we estimate
the following set of equations both separately
using OLS and simultaneously using the two-
stage least squares method. Forecast accuracy
(ACRCY) is defined as the negative of
the absolute difference between actual
earnings and the I/B/E/S median consensus
forecast at the fiscal year end month, deflated
by stock price (Lang and Lundholm
1996). Specifically, we compute forecast
accuracy at time t (ACRCY
t
) as in the
following:
1 t
t
FORECAST
?
is the mean
I/B/E/S consensus forecast of period t earnings
made at time t-1,
t
EPS is the actual earnings
per share before extraordinary items at time t,
taken from I/B/E/S, and
t
PRICE is the stock
price at the end of period t.
( )
1
1
1
?
?
?
? =
t
t
t
t
t
PRICE
EPS FORECAST
ACRCY
Forecast dispersion (i.e., DISP) is the
standard deviation of forecasts issued as of
fiscal year end month, as available in the
I/B/E/S Summary File. The accrual index score
(ACCR) is taken from Hung (2001). Firm size
(SIZE) is the log of market value of equity at
fiscal year end. Loss firm year (LOSS) is an
indicator variable that takes the value of one
(zero) if the firm reported a loss (non-loss)
during the year, as in Hwang et al. (1996).
Earnings surprise (SURPRISE) is this year’s
earnings per share less last year’s, deflated by
the stock price. Earnings volatility (EVOL) is
measured as the standard deviation in earnings
per share over the past five years.
ACRCY = !
0
+ !
1
ACCR + !
2
SIZE
+ !
3
LOSS + !
4
SURPRISE
+ !
5
DISP + " (1)
DISP = #
0
+ #
1
ACCR + #
2
SIZE
+ #
3
LOSS + #
4
SURPRISE
+ #
5
EVOL + " (2)
Our sample covers a subset of non-U.S.
firms cross-listed in the U.S. between 1992 and
1999. To form the sample, we start from the
Bank of New York American Depository
Receipt (ADR hereafter) list and identify firms
that have earnings forecast data and the control
variables available in 2001 I/B/E/S and
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Earnings Prediction and the Role of Accural-Related Disclosure: International Evidence
9
Compustat. We then eliminate firm
observations that do not have an accrual index
score in Hung. This procedure results in 1,074
firm-year observations from 362 distinct firms.
Using this sample has an advantage of being
able to control for differences in the reporting
environment across countries (i.e., all of the
sample firms reconcile to U.S. GAAP) as well
as for the enforcement environment (subject to
the Securities Act of 1934) that are known to
associate with the properties of analysts’
forecasts (see for instance, Ashbaugh and
Pincus 2001, Hope 2003).
Sample characteristics are reported in table
2. The mean and median forecast accuracies in
the entire sample are -0.031 and -0.012 and the
mean and median forecast dispersion are 0.211
and 0.100. The mean and median percentage
forecast errors (e.g., forecast error as a
percentage of the actual EPS) is 0.304 and
0.129 respectively. The mean and median
accrual index scores in the sample are around
0.627 and 0.640. Found in the panel B of the
same table is a correlation matrix among the
regression variables. Not surprisingly, most of
the variables are highly correlated with each
another, and especially with the forecast
variables. An interesting observation is that
while ACCR has a strong partial correlation
with the forecast variables, it is not strongly
correlated with the other determinants of the
forecast characteristics, suggesting that it may
contain information about forecast accuracy
that is not captured in the control variables.
Table 1
Home Country Distribution of Sample Firms
Home Country Number of Firms Accrual
Index Score
Argentina 11 0.64
Australia 20 0.82
Belgium 3 0.68
Brazil 22 0.32
Denmark 2 0.55
Finland 2 0.55
France 23 0.64
Germany 19 0.41
Greece 1 0.55
Hong Kong 28 0.64
Ireland 6 0.82
Italy 14 0.45
Japan 32 0.55
Korea 5 0.55
Mexico 43 0.55
Netherlands 24 0.73
New Zealand 5 0.73
Norway 4 0.82
Portugal 2 0.55
Singapore 7 0.64
South Africa 2 0.68
Spain 7 0.77
Sweden 10 0.59
Switzerland 7 0.32
United Kingdom 63 0.82
Total 362
A list of currently cross-listed securities (as of February 2000) was obtained from the Bank of New York Global
Investors Guide.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
0
:
5
5
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
ACCOUNTING RESEARCH JOURNAL VOLUME 18 NO 1 (2005)
10
Table 2
Descriptive Statistics and Correlation Matrix
Panel A: Descriptive Statistics
Variable Mean Median Std Dev Min Max
ACRCY -0.031 -0.012 0.053 -0.496 0.000
%FORCERROR 0.304 0.129 1.279 -4.158 11.333
DISP 0.211 0.100 0.350 0.000 3.320
ACCR 0.627 0.640 0.147 0.320 0.820
SIZE 7.888 8.181 1.855 -2.827 12.143
LOSS 0.083 0.000 0.276 0.000 1.000
SURPRISE -0.002 -0.005 0.078 -0.411 0.600
EVOL 0.615 0.358 0.997 0.000 8.904
Panel B: Correlation Matrix
ACRCY DISP ACCR SIZE LOSS SURPRISE EVOL
ACRCY
DISP -0.204
***
ACCR 0.116
***
-0.143
***
SIZE 0.375
***
-0.027
***
0.043
LOSS -0.502
***
0.101
***
-0.014 -0.168
***
SURPRISE -0.426
***
0.117
***
-0.001
-0.090
***
0.400
***
EVOL -0.081
***
0.490
***
-0.054
*
0.074
***
0.063
**
0.116
***
ACRCY is the accuracy in analysts’ earnings forecasts, measured as the forecast available as of the
beginning month of the fiscal year and the actual earnings per share for the year. DISP refers to dispersion
in analysts’ forecasts, and it is taken from 2001 I/B/E/S Summary File. %FORCERROR is percentage
forecast error, computed as ([Forecast-Actual EPS]/Actual EPS). ACCR refers to the accrual index score
taken from Hung (2001). SIZE is the log of market value of equity at fiscal year end month. LOSS is an
indicator variable that takes the value of one (zero) if the firm reported a loss (gain) during the year.
SURPRISE is this year’s earnings minus last years’ earnings, deflated by stock price. EVOL is the standard
deviation of earnings over the previous five years.
4. Results and Discussions
Because analysts’ forecast dispersion is
modeled endogeneously in equations (1) and
(2), we first estimate the two equations
separately using the OLS method, and then we
also estimate them jointly using the two-stage
least squares regression methodology. These
results are reported in table 3.
Consistent with the view that accounting
accruals provide useful information about
future earnings, our evidence shows that the
forecast accuracy (dispersion) of I/B/E/S/
analysts increases (decreases) with the degree
of effectiveness of accrual accounting system
of a firm’s country of domicile.
4
This
association is robust to the inclusion of various
firm- and country-level determinants of
analysts’ forecasts identified by prior studies.
In both models, ACCR is significant at
p