Description
The purpose of this paper is to investigate, using data on US manufacturing firms, how
and when corporate governance affects managers’ decisions to use discretionary accruals and thereby
artificially influence company financial reports
Accounting Research Journal
Financial status, corporate governance quality, and the likelihood of managers using
discretionary accruals
Sebahattin Demirkan Harlan Platt
Article information:
To cite this document:
Sebahattin Demirkan Harlan Platt, (2009),"Financial status, corporate governance quality, and the likelihood
of managers using discretionary accruals", Accounting Research J ournal, Vol. 22 Iss 2 pp. 93 - 117
Permanent link to this document:
http://dx.doi.org/10.1108/10309610910987475
Downloaded on: 24 January 2016, At: 21:08 (PT)
References: this document contains references to 60 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1957 times since 2009*
Users who downloaded this article also downloaded:
Marion Hutchinson, (2009),"Governance issues in accounting", Accounting Research J ournal, Vol. 22 Iss 2
pp. 89-92 http://dx.doi.org/10.1108/10309610910987466
Raghavan J . Iyengar, J udy Land, Ernest M. Zampelli, (2010),"Does board governance improve
the quality of accounting earnings?", Accounting Research J ournal, Vol. 23 Iss 1 pp. 49-68 http://
dx.doi.org/10.1108/10309611011060524
Ahsan Habib, Istiaq Azim, (2008),"Corporate governance and the value-relevance of accounting
information: Evidence from Australia", Accounting Research J ournal, Vol. 21 Iss 2 pp. 167-194 http://
dx.doi.org/10.1108/10309610810905944
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Financial status, corporate
governance quality, and the
likelihood of managers using
discretionary accruals
Sebahattin Demirkan
School of Management, SUNY Binghamton University,
Vestal, New York, USA, and
Harlan Platt
College of Business Administration, Northeastern University,
Boston, Massachusetts, USA
Abstract
Purpose – The purpose of this paper is to investigate, using data on US manufacturing ?rms, how
and when corporate governance affects managers’ decisions to use discretionary accruals and thereby
arti?cially in?uence company ?nancial reports.
Design/methodology/approach – Three-stage least squares is employed to study the relationship
between ?nancial status, corporate governance and ?nancial reporting discretion. The sample spans
the years 2001-2003 during a severe downturn in the US stock market. Financial status is measured
with the Altman Z-score.
Findings – A signi?cant difference is found between ?rms not classi?ed as healthy or failed (i.e. the
mid-range group) and the two extreme categories when examining governance quotient using a
well-known index. A positive relationship is found between discretionary accruals and the governance
index. Strong governance appears to reduce the incidence of mid-range ?rms engaging in accruals
management. The least healthy and the most distressed companies have the weakest relationship
with discretionary accruals. By contrast, mid-range ?rms are more likely to resort to discretionary
accruals.
Practical implications – Non-executive members of boards of directors are warned to be
particularly vigilant about discretionary accruals with ?rms transitioning between healthy and
high-failure risk.
Originality/value – The relationship between ?rms’ ?nancial health and discretionary accruals
reveals an agency problem in credit markets with ?nancially stressed ?rms. More attention is required
on ?rms whose ?nancial condition is uncertain. Also, it is documented that signi?cant ?ndings of
importance to the earnings quality and corporate governance literature by documenting the role of
corporate governance on discretionary accruals and ?nancial status.
Keywords Corporate governance, Financial reporting, Financial performance, UnitedStates of America,
Manufacturing industries
Paper type Research paper
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
The authors are grateful to the reviewers for their keen insights and extraordinary diligence.
The authors also thank Arnie Wright for his valuable comments. The authors would like to
acknowledge ?nancial support from College of Business Administration, Northeastern
University and School of Management, SUNY Binghamton University.
Discretionary
accruals
93
Accounting Research Journal
Vol. 22 No. 2, 2009
pp. 93-117
qEmerald Group Publishing Limited
1030-9616
DOI 10.1108/10309610910987475
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
Accounting literature uses discretionary accruals to measure earnings management and
market ef?ciency since earnings are composed of cash ?ow from operations (CFO) and
accruals (DeFond and Jiambalvo, 1994; Rees et al., 1996). Bernstein (1993) argues that:
Cash ?owfromoperations, as a measure of performance, is less subject to distortion than is the
net income ?gure. This is so because the accrual system, which produces the income number,
relies on accruals, deferrals, allocations and valuation, all of which involve higher degrees of
subjectivity than what enters the determination of CFO.
His explanation highlights why academic researchers focus on discretionary accruals to
detect all kinds of earnings management and opportunistic behaviors of managers.
Market ef?ciency is also measured by how the information in discretionary accruals is
contained in investors’ decision-making process (Sloan, 1996).
Managers may be motivated to manage their ?rm’s earnings as a consequence of
self interest (e.g. compensation, stock awards or pension contributions) especially if the
remunerative event depends on observable performance measures such as earnings or
pro?tability (Fields et al., 2001). Earnings manipulations affect the quality of earnings
by distorting the portion of earnings that are a result of corporate success and the
portion resulting from managerial discretion. Not surprisingly, investors respond
appropriately if they believe earnings numbers are distorted (Dechow and Schrand,
2004). On the other hand, successful detection of earnings management is problematic
due to the surfeit of variables that in?uence the managers’ decision set about the ?rm.
Greater recent attention on discretionary current accruals (DCA) is a consequence of
major accounting frauds (such as Enron) and in response to mounting evidence of
corporate earnings management (Clarke, 2005). It is doubtful that all companies use
DCA to in?uence their earnings level. We speculate that the greatest incidence of DCA
occurs when companies’ ?nancials are under the closest scrutiny[1]. Such times include
when credit lines are being reviewed by lenders, unexpected shocks hit either the
company or its industry, or the company’s operating and ?nancial results are under
stress. This research focuses on this later case. That is, we study how a company’s
?nancial condition (as measured by the Altman Z-score) in?uences its use of DCA by
taking into consideration their corporate governance quality[2].
Financial status of the ?rm is important for both creditors and investors. The
literature also shows that corporate investment decisions are related to ?nancial
factors (Cleary, 1999). Firm’s investment strategies are closely related to their accruals
(Ohlson, 1995; Fair?eld et al., 2003). Basically, investment strategies of ?rms are
sensitive to the availability of both internal and external funds and as a consequence
accruals play a signi?cant role in their decisions.
External sources of ?nancing may come with agency costs. Therefore, ?rms may
prefer to ?nance their investments with internal sources of ?nancing to avoid this
additional agency cost. Following the above argument, Bernanke and Gertler (1990),
Gertler (1992) and Fazzari et al. (1988) classify ?rms according to their ?nancial
constraints by using dividend payout, size, age, group memberships or debt ratings,
and show that investment decisions of ?rms that are ?nancially constrained are more
sensitive to ?rm liquidity than otherwise. Kaplan and Zingales (1997) found the
opposite result that investment decisions of the least ?nancially constrained ?rms
are less sensitive to the availability of cash ?ows. Cleary (1999), like Kaplan and
Zingales’ (1997), classi?es ?rms according to their ?nancial variables which is similar
ARJ
22,2
94
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to how the Altman Z predicts bankruptcy[3]. Cleary (1999) notes that corporate
investment (real activity decision) is sensitive to ?rm liquidity and that the relationship
is more severe among less creditworthy ?rms. We document a strong relationship
between a ?rms’ ?nancial classi?cation (according to the Altman Z-score which
includes liquidity ratios) and its use of performance adjusted DCA (PADCA)[4] which
is ?nancial accounting reporting choice of managers, see the Appendix for PADCA’s
calculation. This relationship is most sensitive for ?rms that are in the mid-range
between healthy and distressed classi?cations.
This study contributes to both the accounting and the ?nance literature in various
ways. First of all, we document the relationship between ?rms’ ?nancial health and
their discretionary accruals. This result is important because it shows the existence of
agency problems in credit markets if ?nancial statements signal that the ?rm is
?nancially stressed. Second, we document the importance of paying more attention to
?rms whose ?nancial condition is uncertain since they appear to be more prone to
accrual manipulations. This is especially important for the non-executive members of
boards of directors of the ?rm concerned with governance issues in the Sarbanes-Oxley
constrained environment. Finally, we document signi?cant ?ndings of importance to
the earnings quality and corporate governance literature by documenting the role of
corporate governance on PADCAs and ?nancial status.
The remainder of the paper is organized as follows. Section 2 discusses the
motivation behind the study, Section 3 reviews existing literature, Section 4 develops
hypotheses, Section 5 presents research design, Section 6 documents empirical results
and the conclusions are offered in Section 7.
2. Motivation
A number of motivations for managers to use discretion in accounting have been
proposed in the literature including hitting targets to affect employee bonuses and debt
covenants as well as motivations affecting stakeholders and stock price by achieving
targeted earnings goals (Healy and Wahlen, 1999; Dechow and Skinner, 2000;
Fields et al., 2001). In addition, managers may want to increase their reputation with
stakeholders, such as customers, suppliers, and creditors, to gain bargaining power
and thereby in?uence their terms of trade (Bowen et al., 1995; Burgstahler and Dichev,
1997). Companies reporting better results, even if they have been managed, are likely to
be rewarded with advantageous credit terms relative to similar companies not
managing their accounts. Executive pay may be related to some preset earnings
number pushing managers to use accounting discretion to realize earnings goals and
thereby increase their personal wealth and continue their tenure at the ?rm (Healy,
1985; Matsunaga and Park, 2001; Farrell and Whidbee, 2003; Francis et al., 2008).
Managers may also manage earnings to avoid violating debt covenants that would
increase the cost of capital for the ?rm (Watts and Zimmerman, 1990).
Our study contributes to the stream of research that asks “why do managers use
accounting discretion?” We incorporate two factors that previously were separately
tested as determinants of discretionary accruals: the company’s ?nancial status and
corporate governance. A company’s ?nancial health is related to the other factors
discussed above that explain discretionary accruals including impacts on executives’
careers, corporate reputation, stock price, and debt covenants. Our study also makes
a contribution to the literature regarding classi?cation of ?rms according to their
Discretionary
accruals
95
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
?nancial health by considering the role that discretionary accruals play in that
matter[5].
Graham et al. (2005) report that companies are more likely to take real actions such
as asset sales or reducing investments to facilitate the meeting of earnings’ targets.
They note that “accounting actions to meet earnings benchmarks get notably little
support” (p. 36), from their survey respondents, 401 ?nancial executives. It is precisely
this negation of the admission that discretionary accruals are used by ?rms that leads
us to investigate their use by ?nancially stressed ?rms. That is, we agree that it is
likely that normal or healthy companies are averse to using discretionary accruals;
however, it is companies that either are falling from a state of health or which are in the
process of working their way out of distress that we believe are likely candidates to be
users of discretionary accruals.
3. Literature review
Yeh and Lee (2004) ?nd that weak corporate governance increases the probability of a
?rm being classi?ed as ?nancially distressed, but not vice versa. Observant and
involved members of a board of directors provide a forumthat promotes better managed
?rms. They also note that healthy ?rms are more likely to have strong corporate
governance than weaker ?rms.
Cook et al. (2008) examine the effect of tax planning on the relationship between
effective tax rates (ETRs) and earnings management that is documented byDhaliwal et al.
(2004). Following that argument they also look at the effect of the Sarbanes-Oxley Act
(SOX) of 2002 onthe relationship between earnings management and ETR. They ?nd that
tax planning effects a ?rm’s ETRs and its earnings management. On the other hand, they
?nd no signi?cant result between SOX and earnings management.
Bowen et al. (2008) investigate whether managers’ use of accounting discretion for
opportunistic purposes or for shareholder wealth maximization, as ef?cient contracting
theory anticipates. They ?nd some evidence that managers use accounting discretion to
increase shareholder wealth. Unlike the conclusions of the prior literature, they also
document that poor governance is positively associated with future CFO and return on
assets (ROA). Therefore, they assert that earnings management may bene?t
shareholders by signaling future performance of the company. On the other hand,
their ?nding does not support the idea that managers exploit lax governance structures
and engage in accountingdiscretion at the shareholder’ expense. Similar to Bowen’s et al.
(2008) argument in our case, managers may use discretionary accruals to signal that a
?rm deserves a healthy classi?cation if it would be classi?ed otherwise. On the other
hand, opportunistic usage of discretionary accruals may increase the manager’s own
wealth and entrench their position at their company by avoiding disclosure of ?nancial
distress or bankruptcy. Corporate governance quality plays a signi?cant role in the
decision to exercise discretionary accrual in ?nancial reports.
Jones et al. (2008) investigate the relationship between restatements and several
models of discretionary accruals. They document that accrual estimation errors are
correlated with restatements. Their ?nding suggests the need for research on whether
?rms use discretionary accruals to avoid being classi?ed as distressed ?rms. Both
restatements and distressed classi?cations lead investors to discount the value of ?rms
signi?cantly. Later after a ?rm’s health has organically improved, they may correct
?nancial statement manipulation through restatements.
ARJ
22,2
96
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Holthausen et al. (1995) document that managers manipulate earnings to maximize
their bonus plan payments. This ?nding supports Healy’s (1985) idea but did not
support his other idea that managers manipulate earnings downward (income
decreasing earnings manipulation) when earnings fall belowthe minimumnecessary to
receive bonuses. We support the income increasing argument and think that managers
may forsake accruals management when their ?rm is stressed because manipulation
then has no affect on their compensation.
Dhaliwal et al. (2005) examine the relationship between institutional ownership,
?nancial health and market valuation weights on earnings and book value of equity. They
document that ?rms that have high-institutional ownership are ?nancially healthier. Their
?nding is supportive of prior studies that show institutional investors playing a ?duciary
and positive governance role (Shleifer and Vishny, 1997; Bushee, 1998). Prior studies also
suggest that institutional investors prefer to invest in ?nancially healthy ?rms (Hessel and
Norman, 1992; Del Guercio, 1996). Managers may seek to keep institutional investors as
owners in order to bene?t fromtheir ?nancial power but also to achieve stock appreciation
whichmayincrease the value of managers’ bonuses. Our studycontributes tothis streamof
research by showing that managers use accounting discretion to make mid-range
companies appear healthier and thereby become the targets of institutional investors in the
future, which may affect their corporate governance quality as well.
Ali et al. (2007), Mather and Ramsay (2006) and Hsu and Koh (2005) examined the
impact of corporate governance on discretionary accruals. Ali et al. (2007) ?nd better
quality earnings reports among family ?rms who they ?nd are also more reticent about
their corporate governance practices. Mather and Ramsay (2006) consider how
characteristics of strong corporate boards are associated with the degree of earnings
management at the time of chief executive of?cer (CEO) transition. Notably, they uncover
evidence in cases where the CEOresigned of negative unexpected accruals. This is limited
by larger boards and boards with a higher proportion of independent directors. When the
CEOretires they ?nd that discretionary accruals are somewhat controlled when there is a
higher proportion of executive and af?liated director shareholdings. Hsu and Koh (2005)
consider the corporate governance role of institutional investors. They ?nd upward
accruals management at companies whose shares are owned by institutional investors
that sell out as opposed to take long-term positions. In contrast, institutions with a
long-term hold strategy constrain upward accruals management even among companies
that have strong incentives for such behavior.
4. Hypothesis development
4.1 Altman Z-score
There are many studies in the ?nance and accounting literature that investigate how to
classify and to measure ?rms according to their ?nancial health. Altman’s (1968) initial
study developed a measure to estimate bankruptcy risk based on accounting ratios.
Follow up literature to improve the ef?ciency of Altman’s approach include Merton
(1974), Shumway (2001), Zmijewski (1984) and Platt and Platt (1991). These studies
generally conclude that bankruptcy is determined by:
.
?rm liquidity, pro?tability and leverage;
.
market variables; and
.
macroeconomic or business cycle variables.
Discretionary
accruals
97
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In this study, we use the Altman Z-score to classify ?rms according to their ?nancial
health.
4.2 Discretionary accruals
Investors respond negatively to information indicating that a ?rmis distressed (Chi and
Tang, 2007). This is called the distressed ?rm discount in the academic literature. In
order to avoid this discount executives of a ?rm may use discretionary accruals to
manipulate their disclosures, e.g. earnings management (Kasznik, 1999). Accrual
management is one of the ways managers obtain desired earnings numbers (Dechow
and Schrand, 2004). Accruals are open to manipulation because they require the
manager to forecast and to make judgments. Consequently, accruals are subject to all
kinds of management discretion[6]. Of course, accrual management changes only the
timing of the recognition of earnings. That is, managers shift earnings between
quarters but do not create real earnings.
The earnings management literature generally reports that high-accrual companies
have the potential to manage their earnings (Dechowet al., 1996; Richardson et al., 2002).
These studies have bifurcated accruals into those that are nondiscretionary and those
that are discretionary in order to detect earnings management by managers. Jones (1991)
employed what is now a commonly used method to determine discretionary accruals.
The main criticism of her model is that it does not control for performance and growth.
Subsequent developments in the literature have attempted to improve the power of this
model to estimate the amount of nondiscretionary accruals (Dechow et al., 1996;
Kothari et al., 2005). In this study, we follow the method developed by Kothari et al.
(2005). Their discretionary accrual estimation method is an extension of Dechow et al.
(1996) but they go one step further and adjust discretionary accruals with the
performance of the ?rm which is measured by lagged ROA.
Firms wishing to appear healthier than their true condition use discretionary
accruals in order to avoid being identi?ed as in distress and thereby avoid the ?nancial
market discount. Corporate governance of the ?rms in?uences their ?nancial reporting
decisions (Bowen et al., 2008). By following above argument, our main hypotheses
concern the timing of when managers engage in discretionary accruals. Speci?cally, we
test the following hypotheses in alternative format:
H1. Discretionary accruals of ?rms are related to their ?nancial health and
corporate governance quality.
Firm’s ?nancial classi?cations are calculated using the Altman Z-score which depends
upon numbers that come from ?nancial statements. Therefore, the Z-score itself may
be affected by those accounts. Moreover, corporate governance quality may affect
corporate ?nancial reporting decisions by inducing strategic management decisions.
This leads to our second hypothesis becomes:
H2. Financial classi?cations of ?rms are related to discretionary accruals and
corporate governance quality.
A ?rm’s governance mechanism is mainly affected by its economic environment,
shareholders and creditors. Therefore, a ?rm’s ?nancial status may in?uence
management’s strategic decisions. While the effect of discretionary accruals on
corporate governance is less clear, we include that possibility in our H3:
ARJ
22,2
98
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H3. Corporate governance quality is related to ?nancial health and discretionary
accruals.
5. Research design, measurements and sample
5.1 Altman Z-score
Altman (1968) used multiple discriminate analyses to determine the ?nancial conditions
of ?rms. From a list of 22 ?nancial ratios, he put ratios in ?ve categories including
liquidity, pro?tability, leverage, solvency and activity to discriminate between healthy
and to-be-bankrupt ?rms. Altman selected the ratios according to their popularity in the
literature rather than a theoretical basis resulting in a linear combination of variables
that discriminated between bankrupt and non-bankrupt ?rms. The linear equation is as
follows:
Z ¼ 0:012X
1
þ 0:014X
2
þ 0:033X
3
þ 0:006X
4
þ 0:999X
5
ð1Þ
where X
1
– working capital/total assets; X
2
– retained earnings/total assets; X
3
–
earnings before interest and taxes/total assets; X
4
– market value of equity/total
liabilities; X
5
– sales/total assets; Z – overall index, the lower a company’s Z-score the
higher its probability to bankrupt.
We use the Altman Z-score to measure the ?nancial health of companies. Later
research con?rms that the Altman Z-score works better for manufacturing ?rms (Grice
and Ingram, 2001). Therefore, we con?ned our study to manufacturing ?rms[7].
Following Altman’s (1968) study we classi?ed ?rms into three groups according to
their Z-score level. If Z-score exceeds 2.67 then the ?rm is called ?nancially healthy,
if the Z-score is between 1.81 and 2.67, ?rms are called gray-area companies, and ?nally
if the Z-score is smaller than 1.81, ?rms are classi?ed as ?nancially distress.
5.2 Discretionary accruals
Examples of items that may be subject to discretionary include accounts payable,
account receivable, future tax liability, depreciation, and future interest expense among
others. SFAS No. 95 requires companies to disclose information necessary to compute
accrual components of earnings in the operating section of cash ?ows. Accrual
components of earnings are computed by following prior literature (Dechowet al., 1995):
Total current accruals ¼ ðDCA 2DCashÞ 2ðDCL 2DSTDÞ ð2Þ
where DCA – change in current assets (Compustat item no. 4); DCash – change in
cash/cash equivalents (Compustat item no. 1); DCL – change in current liabilities
(Compustat itemno. 5); DSTD – change in debt includedin current liabilities (Compustat
item no. 34).
As mentioned above, managers may in?uence accruals to get their desired earnings
numbers. According to this hypothesis, companies do not change their activities rather
they report opportunistic (manipulated) earning numbers. For example, companies
may be avoiding write-offs, or capitalizing some expenses. The idea of accruals being
discretionary means that they require managers to forecast, estimate and make
judgments (Dechow and Schrand, 2004). The discretionary accrual measure is obtained
by subtraction of normal accruals from estimated total accruals thereby providing an
estimate of discretionary accruals that a ?rm might take in a year (Kothari et al., 2005).
Discretionary
accruals
99
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A higher estimate of discretionary accruals may signal a greater level of earnings
management (Dechow and Schrand, 2004)[8]. Lower discretionary accruals estimate
may signal higher earnings quality.
To estimate abnormal discretionary accruals, we follow Kothari et al.’s (2005) and
calculate PADCA. First, we estimate the modi?ed Jones model separately for each year
for each two-digit Standard Industrial Classi?cation (SIC) code (Dechow et al., 1995).
Then, PADCA is computed as the difference between the abnormal accrual and the
closest matched ?rm’s abnormal accrual, where the closest matched ?rm is the ?rm in
the same two-digit SIC code with the closest ROA in the prior year[9].
To test whether a ?rm’s health in?uences its use of PADCA, we examine whether
the amount of PADCA is related to its Altman’s Z-score. There is a possibility that the
Altman Z-score itself, which relies on ?nancial data that may be altered by the ?rm’s
use of PADCA, may be affected by PADCA. That is, there is a possibility that the
estimated level of Altman’s Z itself is related to the ?rm’s use of PADCA. Healthy ?rms
may have stronger governance because of their ability to hire better skilled executives
than distressed ?rms. Therefore, the corporate governance mechanism may in?uence
?nancial status classi?cation. This relationship may hold for discretionary accruals as
well. Moreover, corporate governance quality may affect the discretionary accruals,
and vice versa. This may lead to three-way endogeneity in our model between
corporate governance, ?nancial health and discretionary accruals. We thus consider a
simultaneous relationship model with PADCA, Altman’s Z and corporate governance
quality score (GINDEX). To test this question, we use a three-stage least squares
(3SLS) model similar to Al-Tuwaijri et al. (2004)[10].
The basic three equation model is shown below:
PADCA ¼ f ðY
1
; Altman Z 2score; GINDEXÞ ð3Þ
Altman Z ¼ gðY
2
; PADCA; GINDEXÞ ð4Þ
GINDEX ¼ gðY
3
; PADCA; Altman Z 2scoreÞ ð5Þ
where Y
1
, Y
2
and Y
3
are vectors of additional variables that in?uence PADCA, Altman’s
Z-score, and GINDEX, respectively, where Altman Z is a proxy for corporate health and
GINDEX is an indicator for the corporate governance quality. The incidence of PADCA
is assumed to be related to Y
1
(a vector of determinants), Altman Z, and GINDEX. We
believe that companies engage in PADCA as their health deteriorates and no
pre-assumption for corporate governance. Moreover, we believe that a company’s health
(measured by Altman’s Z) is affected by Y
2
(a vector of determinants), by PADCA, and
by GINDEX. That is, a company affects its measured ?nancial health (via Altman’s
Z-score) when it engages in the use of PADCA, and having good corporate governance.
Finally, corporate governance quality (measured by GINDEX) is in?uenced by Y
3
(a vector of determinants), PADCA and Altman Z.
The simultaneous relationship between PADCA, Altman Z and GINDEX creates an
identi?cation problem which leads to biased and inconsistent parameter estimates if
the model is estimated using ordinary least squares. To avoid this problem, a 3SLS
regression is performed[11]. Using some of the control variables similar to Ali et al.
(2007) and including more, a 3SLS model for the years 2001-2003 is run to examine the
relationship between discretionary accruals and the ?rms’ ?nancial health[12]. We also
include dummy variables for each year and two-digit industry classi?cations[13]:
ARJ
22,2
100
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
PADCA ¼ a þb
1
Altman 2Z þb
2
GINDEX þb
3
HEALTH þb
4
AltmanZ
£ HEALTH þb
5
DISTRESS þb
6
Altman 2Z £ DISTRESS
þb
7
LagPADCA þb
8
LEVERAGE þb
9
SIZE þb
10
LOSS þb
11
ROA
þb
12
CFO þ error
ð6Þ
Altman 2Z ¼ a þb
1
PADCA þb
2
GINDEX þb
3
SIZE þb
4
LOSS þb
5
ROA
þb
6
CFA þb
7
LagAltman 2Z þb
8
EPS þb
9
MB
þb
10
LITIGATION þb
11
R&D þb
12
DEPRECIATION
þb
13
PP&E þb
14
INTANGIBLE þb
15
SPECIAL
þb
16
INTEREST þ error
ð7Þ
GINDEX ¼ a þb
1
Altman 2Z þb
2
PADCA þb
3
SIZE þb
4
LITIGATION
þb
5
MERGER þb
6
PREFERRED þb
7
DIVIDEND þ error
ð8Þ
where PADCA (see the Appendix for the calculation) is performance adjusted
discretionary accruals that is calculated by following Kothari et al. (2005). PADCA is
scaled by total assets[14]. Altman Z is the Z-score developed by Altman (1968) to
estimate the ?nancial condition of ?rms. The variables in equation (7) that explain the
level of Altman Z beyond PADCA and GINDEX re?ect hypotheses about ?rm
characteristics and Altman Z. SIZE, R&D, MB, ROA and EPS are expected to control
for the performance and information environment of the ?rms therefore we expect they
have positive coef?cients as higher values for these variables contribute to healthier
companies. In contrast, LOSS, LITIGATION, CFO, DEPRECIATION, PP&E,
INTANGIBLE, SPECIAL and INTEREST, capture the operational and ?nancial
riskiness of the ?rm, are expected to have negative coef?cients since higher values of
these variables reduce the health of ?rms. MERGER and LEVERAGE are used
speci?cally to capture agency problem which are interpreted as the higher the values
for these variables the higher the agency costs. All of the variables have expected signs
except for ROA.
GINDEX is a corporate governance index that is developed by Gompers et al. (2003)
to measure the governance quality of the ?rm where GINDEX score is extracted from
the IRRC database. HEALTH is an indicator variable that takes the value of one if the
Z-score exceeds the value of 2.67, and zero otherwise. DISTRESS is also an indicator
variable that takes the value of one if Z-score is smaller than 1.81, and zero
otherwise[15]. LagPADCA is equivalent to last year’s PADCA scaled by total
assets[16]. LEVERAGE is the ratio of total debt to total assets at the beginning of the
?scal period (total assets minus its book value divided by its total assets) scaled by
total assets. SIZE is the natural logarithm of total assets. MERGER (MA) is a dummy
variable that takes the value of one if the ?rm has engaged in a merger and/or
acquisition activity during the current ?scal year, zero otherwise. LOSS is an indicator
Discretionary
accruals
101
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
variable that takes the value of one if the ?rm reports a net loss for the ?scal period,
and zero otherwise. ROA is the current year’s return on asset calculated as net income
before extraordinary items divided by total assets scaled by total assets. CFO is scaled
by the beginning of the year’s total assets scaled by total assets. EPS is earnings
per share which is calculated by dividing net income before interest and taxes by
numbers of shares outstanding scaled by total assets. MB is the market-to-book ratio
that is calculated by using the market value of equity divided by the book value of
equity scaled by total assets. LITIGATION is a dummy variable that equals one if the
?rm operates in the high litigation industries with the SIC codes of 2833-2836,
3570-3577, 3600-3674, 5200-5961 and 7370-730, zero otherwise. R&D is research and
development expense scaled by total assets. DEPRECIATION is depreciation expense
in the current year scaled by total assets. PP&E is total net plant, property and
equipment that is reported on the ?rm’s balance sheet scaled by total assets.
INTANGIBLE is total intangible assets scaled by total assets. SPECIAL is total special
items scaled by total assets. INTEREST is total interest expense within a year reported
on a ?rm’s income statement scaled by total assets. PREFERRED is total preferred
stock holders equity that is reported under the owners’ equity section of the balance
sheet in the ?rm’s annual report scaled by total assets. DIVIDEND is total dividend
issued and distributed by the ?rm in the current year scaled by total assets[17].
5.3 Sample
We use US ?rm data, over the period 2001-2003. This period is chosen because it
includes a large downturn in the stock market which may be an environment more
susceptible to PADCA activity[18]. Only, manufacturing ?rms are analyzed to conform
to the Altman Z-score original design and due to its greater ef?cacy for these ?rms
(Altman, 1968; Grice and Ingram, 2001). Specially, we collected all the data from the
Compustat database between the years 2001 and 2003 for manufacturing ?rms.
Manufacturing ?rms have the SIC code between 2000 and 4000. From 67,899 ?rm-year
observations originally gathered, the sample decreases to 8,655 ?rm-year observations
after calculating PADCA, Altman Z-score, and other variables in the two equation
model. We used Gompers et al. (2003) governance index (GINDEX) to measure the
governance strength of the ?rms in our sample.
6. Empirical results and analysis
Table I, Panel A provides descriptive statistics of the dependent variables. The mean
(median) Altman Z for sample ?rms is 2.8374 (2.6980). The mean (median) performance
adjusted discretionary accruals (PADCA) 0.0006 (20.0005) indicates on average
sample ?rms have income increasing accruals. GINDEX 9.1836 (9.000) suggests that on
average sample ?rms have good corporate governance quality. Panel B of Table I
presents the descriptive statistics for independent variables. Leverage with 0.6659
(0.3996) is positively skewed that indicates that sample ?rms have on average
66.8 per cent total liabilities on their balance sheets. Average size which is measured
with market value of equity in our sample is approximately $126 million. The merger
variable shows that, on average, 27 per cent of companies engage in M&A. In our
sample, 49 per cent of companies report losses which may be because our sample
consist of only manufacturing ?rms. ROA average is 217.22 per cent and its
median is 0.2 per cent in part because our sample period is between 2001 and 2003.
ARJ
22,2
102
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
CFO is negative (a value of 8.19 per cent) with a positive median value which indicates
skewness to the left. EPS share is also negative again because our sample period spans
an economic downturn. Market to book ratios show that on average sample ?rms are
established, and conservative in their reporting decisions. About 35 per cent of ?rms
are in litigated industries as de?ned earlier. On average ?rms spend 10 per cent of their
Variables Mean SD 25th percentile Median 75th percentile
Panel A: dependent variables
Altman Z 2.8374 14.6280 1.0901 2.6980 5.1304
PADCA 0.0006 0.1779 20.0832 20.0005 0.0806
GINDEX 9.1836 2.6813 7.0000 9.0000 11.000
Panel B: predetermined variables
LEVERAGE 0.6659 6.4737 0.2128 0.3996 0.5877
SIZE 4.8391 2.4434 3.1286 4.7694 6.5128
MERGER 0.2659 0.4419 0.0000 0.0000 1.0000
LOSS 0.4934 0.5000 0.0000 0.0000 1.0000
ROA 20.1722 0.5052 20.2079 0.0020 0.0638
CFO 20.0819 0.4257 20.0893 0.0454 0.1105
EPS 20.0189 0.0655 20.0130 0.0000 0.0022
MB 20.0480 1.8147 0.0010 0.0075 0.0443
LITIGATION 0.3546 0.4784 0.0000 0.0000 1.0000
R&D 0.1059 0.5052 0.0000 0.0322 0.1171
DEPRECIATION 0.0527 0.0445 0.0274 0.0429 0.0626
PP&E 0.4903 0.3601 0.2130 0.4111 0.6871
INTANGIBLE 0.1100 0.1537 0.0000 0.0344 0.1671
SPECIAL 20.0486 0.1604 20.0249 0.0000 0.0000
INTEREST 0.0294 0.0643 0.0017 0.0130 0.0297
PREFERRED 0.0447 0.2191 0.0000 0.0000 0.0000
DIVIDEND 0.0003 0.0011 0.0000 0.0000 0.0000
Notes: All variables are winsorized at 1 per cent level; reported values of variables are for a ?rm at current
year if it is not stated otherwise; variable de?nitions: Altman Z is ?nancial indicator score that was developed
by Altman (1968). Detailed calculation for Altman Z is provided in the text; PADCA is performance adjusted
discretionary accruals that is calculated by following Kothari et al. (2005) which is scaled by total assets;
GINDEX is governance index that is developed by Gompers et al. (2003). This variable is extracted from
Thomson IRRC governance database; LEVERAGE is the ratio of total debt to total assets at the beginning of
the ?scal period scaled by total assets; SIZE is the natural logarithm of the total assets; MERGER (MA) is a
dummy variable that takes the value of 1 if the ?rm has engaged in a merger and/or acquisition activity, 0
otherwise. LOSS is an indicator variable that takes the value of 1 if the ?rm reports a net loss for the ?scal
period and 0 otherwise, i.e. negative net income before extraordinary items; ROA is the current year’s return
on assets calculated as net income before extraordinary items divided by total assets scaled by total assets;
CFO is cash ?ow from operations scaled by total assets; EPS is earnings per share which is calculated as
dividing net income before interest and taxes by numbers of shares outstanding corresponding year scaled
by total assets; MB is the market-to-book ratio that is calculated by using the market value of equity divided
by the book value of equity scaled by total assets; LITIGATION is a dummy variable that equals one if the
?rm operates in the high litigation industries with the SIC codes of 2833-2836, 3570-3577, 3600-3674, 5200-
5961, and 7370-730, 0 otherwise; R&D is research and development expense scaled by total assets;
DEPRECIATION is depreciation expense scaled by total assets; PP&E is total net plant, property and
equipment that is reported at ?rm’s balance sheet scaled by total assets; INTANGIBLE is total intangible
assets scaled by total assets; SPECIAL is total special items reported in income statement scaled by total
assets; INTEREST is total interest expense within a year that is reported at ?rm’s income statement scaled
by total assets; PREFERRED is total preferred stock holders equity that is reported under the owners’ equity
section of balance sheet at ?rm’s annual report scaled by total assets; DIVIDEND is total dividend issued and
distributed scaled by total assets
Table I.
Descriptive statistics
of 8,655 numbers
of observations except
GINDEX variable has
1,340 observations
Discretionary
accruals
103
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
total assets on research and development investments. Median value for this variable is
3.22 per cent. Depreciation expense is 5.27 per cent of total assets. About 49.03 per cent
of assets are PP&Es which is not surprising given the heavy investment of
manufacturing ?rms in machineries, buildings, etc. On the other hand, intangible
assets are 11 per cent of total assets and could be generated by research and
development investments or patents. Special items which are usually one-off items
have mean and median values of 24.86 and 0.00 per cent, respectively. Total interest
expense is on average 2.94 per cent with an average rate of 5 per cent (2.94/66.59).
Preferred stock ownership percentage is 4.47 per cent on average with a median value
of 0.00 per cent. Finally, sample ?rms distribute few dividends with the average size of
0.3 per cent. This may be because half of our sample ?rms report loss.
Table II presents descriptive statistics for healthy, distressed and gray area ?rms.
The average Altman Z-scores for healthy ?rms is 9.4361, for distressed ?rm is
26.4160, and for gray area ?rms is 2.2336. Differences between the means are
signi?cant at the 1 per cent level according to the mean difference t-test. The PADCA
score reports a different story than the Altman Z-score using the mean comparison test.
The mean PADCA score for healthy ?rms is 0.0057, for distress ?rms is 20.0088, and
for gray area ?rms is 0.0057. Both healthy and gray area ?rms have statistically
signi?cant mean PADCA differences compared with distressed ?rms. On the other
hand, healthy ?rms and gray ?rms’ average PADCA are not statistically different.
After determining the Altman Z-score and PADCA for ?rms in our sample, we had
three sets of companies: those in the healthy area (787 ?rms), those in the mid-range or
gray area (262 ?rms), and those in the distress region (291 ?rms). We calculated the
value of the governance index (GINDEX) for the healthy, gray, and distressed ?rm
groups. We found mean values of 9.531, 9.137 and 8.997, respectively, as seen in
Table II. The gray area has a mean according to t-tests that is signi?cantly different at
the 5 per cent level from the mean values of the healthy ?rms but is not different from
distressed area ?rms at conventional levels. Distressed ?rms have smaller GINDEX
values than healthy ?rms and their mean GINDEX value is signi?cantly different at
the 5 per cent level from that of healthy ?rms. In other words, there appears to be less
corporate governance quality when ?rms depart from being healthy and are not clearly
identi?ed as distressed. That is, gray area companies have less governance quality
applied to them than their counterparts’ healthy ?rms. Mean difference tests document
that healthy ?rms’ corporate governance quality is better than both gray and
distressed ?rms[19]. Later, we introduce the governance index into our regression
model to see how well it explains the level of PADCA.
There are fewer discretionary accruals among distressed companies; healthy and
gray area companies have more discretionary accruals than distressed companies but
about the same level of discretionary accruals between them. Managers of these ?rms
may feel that it is in their best interest to produce better results this quarter/year than
the ?rm justi?ably deserves. Some healthy and gray area companies appear to
creatively use accruals to bolster their ?nancial results. Some healthy ?rms may be
using PADCA to achieve a healthy classi?cation. Similarly, some gray area companies
may use PADCA to avoid being classi?ed as distressed. In fact, the lower proportion of
PADCA among distressed companies may result from those ?rms having fewer
remaining PADCA opportunities. Another reason for higher discretionary accruals
among gray area ?rms may be noise in their accounting information, i.e. information
ARJ
22,2
104
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
M
e
a
n
M
e
a
n
d
i
f
f
e
r
e
n
c
e
ª¯
V
a
r
i
a
b
l
e
s
H
e
a
l
t
h
y
D
i
s
t
r
e
s
s
G
r
a
y
H
e
a
l
t
h
y
-
d
i
s
t
r
e
s
s
H
e
a
l
t
h
y
-
g
r
a
y
G
r
a
y
-
d
i
s
t
r
e
s
s
A
l
t
m
a
n
Z
9
.
4
3
6
1
2
6
.
4
1
6
0
2
.
2
3
3
6
1
5
.
8
5
2
1
*
*
*
7
.
2
0
2
5
*
*
*
8
.
6
4
9
6
*
*
*
P
A
D
C
A
(
%
)
0
.
0
0
5
7
2
0
.
0
0
8
8
0
.
0
0
5
7
0
.
0
1
4
5
*
*
*
0
.
0
0
0
0
0
.
0
1
4
5
*
*
G
I
N
D
E
X
9
.
5
3
1
8
.
9
9
7
9
.
1
3
7
0
.
5
3
4
*
*
0
.
3
9
4
*
*
0
.
1
4
0
L
a
g
P
A
D
C
A
(
%
)
0
.
0
0
5
4
2
0
.
0
0
5
8
0
.
0
1
1
0
0
.
0
0
1
2
*
*
2
0
.
0
0
5
6
0
.
0
1
6
8
*
*
*
L
E
V
E
R
A
G
E
0
.
2
9
1
0
1
.
2
7
5
9
0
.
4
9
5
0
2
0
.
9
8
4
9
*
*
*
2
0
.
2
0
4
0
*
*
*
2
0
.
7
8
0
9
*
*
S
I
Z
E
5
.
1
2
2
5
3
.
9
5
7
2
5
.
9
9
7
5
1
.
1
6
5
3
*
*
*
2
0
.
8
7
5
0
*
*
*
2
.
0
4
0
3
*
*
*
M
E
R
G
E
R
0
.
2
6
7
3
0
.
2
4
7
5
0
.
3
0
6
5
0
.
0
1
9
8
*
2
0
.
0
3
9
2
*
*
*
0
.
0
5
9
0
*
*
*
L
O
S
S
0
.
3
2
6
7
0
.
7
9
3
6
0
.
3
4
5
9
2
0
.
4
6
6
9
*
*
*
2
0
.
0
1
9
2
2
0
.
4
4
7
7
*
*
*
R
O
A
2
0
.
0
5
6
8
2
0
.
4
0
0
0
2
0
.
0
2
1
3
0
.
3
4
3
2
*
*
*
2
0
.
0
0
3
6
*
*
*
0
.
3
7
8
7
*
*
*
C
F
O
0
.
0
2
4
5
2
0
.
2
8
6
5
0
.
0
4
3
5
0
.
3
1
1
1
0
*
*
*
2
0
.
0
1
9
0
*
*
0
.
3
3
0
0
*
*
*
L
a
g
A
l
t
m
a
n
Z
1
3
.
1
2
3
9
2
1
.
6
0
9
0
3
.
0
8
8
9
1
4
.
7
3
2
9
*
*
*
1
0
.
0
3
5
0
*
*
*
4
.
6
9
7
9
*
*
*
E
P
S
2
0
.
0
0
2
1
2
0
.
0
4
9
1
2
0
.
0
0
5
0
0
.
0
4
7
0
*
*
*
0
.
0
0
2
9
*
*
*
0
.
0
4
4
1
*
*
*
M
B
0
.
1
6
3
9
2
0
.
3
8
1
1
0
.
0
2
0
0
0
.
5
4
5
0
*
*
*
0
.
1
4
3
9
*
*
*
0
.
4
0
1
1
*
*
*
L
I
T
I
G
A
T
I
O
N
0
.
3
8
9
4
0
.
3
7
4
3
0
.
1
8
4
2
0
.
0
1
5
1
0
.
2
0
5
2
*
*
*
2
0
.
1
9
0
1
*
*
*
R
&
D
0
.
0
8
6
1
0
.
1
6
3
5
0
.
0
3
5
6
2
0
.
0
7
7
4
*
*
*
0
.
0
5
0
5
*
*
*
2
0
.
1
2
7
9
*
*
*
D
E
P
R
E
C
I
A
T
I
O
N
0
.
0
4
2
5
0
.
0
6
9
8
0
.
0
4
6
9
2
0
.
0
2
7
3
*
*
*
2
0
.
0
0
4
4
*
*
*
2
0
.
0
2
2
9
*
*
*
P
P
&
E
0
.
4
1
7
6
0
.
5
6
8
9
0
.
5
5
4
1
2
0
.
1
5
1
3
*
*
*
2
0
.
1
3
6
5
*
*
*
2
0
.
0
1
4
8
I
N
T
A
N
G
I
B
L
E
0
.
0
9
3
3
0
.
1
1
6
8
0
.
1
4
9
2
2
0
.
0
2
3
5
*
*
*
2
0
.
0
5
5
9
*
*
*
0
.
0
3
2
4
*
*
*
S
P
E
C
I
A
L
2
0
.
0
2
7
4
2
0
.
0
8
5
2
2
0
.
0
3
3
7
0
.
0
5
7
8
*
*
*
0
.
0
0
6
3
*
0
.
0
5
1
5
*
*
*
I
N
T
E
R
E
S
T
0
.
0
1
0
5
0
.
0
5
8
3
0
.
0
2
5
6
2
0
.
0
4
7
8
*
*
*
2
0
.
0
1
5
1
*
*
*
2
0
.
0
3
2
7
*
*
*
P
R
E
F
E
R
R
E
D
0
.
0
1
0
0
0
.
1
0
9
3
0
.
0
0
8
9
2
0
.
0
9
9
3
*
*
*
0
.
0
0
1
1
2
0
.
1
0
0
4
*
*
*
D
I
V
I
D
E
N
D
0
.
0
0
0
5
0
.
0
0
0
0
0
.
0
0
0
2
0
.
0
0
0
5
*
*
*
0
.
0
0
0
3
*
*
*
0
.
0
0
0
2
*
*
*
N
o
.
o
f
o
b
s
e
r
v
a
t
i
o
n
s
4
,
3
7
4
3
,
0
3
8
1
,
2
4
3
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
0
.
1
0
,
*
*
0
.
0
5
a
n
d
*
*
*
0
.
0
1
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
ª¯
m
e
a
n
t
e
s
t
b
a
s
e
d
o
n
t
-
t
e
s
t
;
H
E
A
L
T
H
Y
?
r
m
s
a
r
e
t
h
e
o
n
e
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
e
x
c
e
e
d
s
t
h
e
v
a
l
u
e
o
f
2
.
6
7
;
D
I
S
T
R
E
S
S
?
r
m
s
a
r
e
t
h
e
o
n
e
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
i
s
s
m
a
l
l
e
r
t
h
a
n
1
.
8
1
;
a
l
l
o
t
h
e
r
?
r
m
s
a
r
e
c
l
a
s
s
i
?
e
d
a
s
G
R
A
Y
a
r
e
a
?
r
m
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
s
m
a
l
l
e
r
t
h
a
n
2
.
6
7
a
n
d
l
a
r
g
e
r
t
h
a
n
1
.
8
1
;
n
u
m
b
e
r
o
f
o
b
s
e
r
v
a
t
i
o
n
s
f
o
r
G
I
N
D
E
X
f
o
r
h
e
a
l
t
h
y
,
d
i
s
t
r
e
s
s
a
n
d
g
r
a
y
a
r
e
7
8
7
,
2
6
2
a
n
d
2
9
1
,
r
e
s
p
e
c
t
i
v
e
l
y
;
v
a
r
i
a
b
l
e
s
a
r
e
d
e
?
n
e
d
i
n
T
a
b
l
e
I
;
L
a
g
P
A
D
C
A
(
A
l
t
m
a
n
Z
)
i
s
p
r
e
v
i
o
u
s
y
e
a
r
’
s
P
A
D
C
A
(
A
l
t
m
a
n
Z
)
Table II.
Comparative descriptive
statistics
Discretionary
accruals
105
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
environment. On the other hand, these ?rms are probably under the close scrutiny of
?nancial market participants such as auditors, banks, investors or banks and therefore
they may not have a tendency to manage their earnings (Bushee, 1998). In that case, the
larger discretionary accrual result may be the outcome of a noisy information
environment. There may be incentives for managers to use discretionary accruals
to retain their job and to get high compensation until regulators catch up with them.
On the other hand, if the market expects discretionary accruals, managers may be
compelled to manage earnings in order to reach equilibrium[20].
Almost all other ?rm characteristics for healthy ?rms are signi?cantly different
from distressed ?rms at the 1 per cent level (5 per cent for L1PADCA and 10 per cent
for MERGER dummy variables) with the exception of LITIGATION. Similar results
are noted between gray area and distressed ?rms. Between healthy and gray area ?rms
the other ?rm characteristic variables are signi?cantly different at 1 per cent level
(CFO at 5 per cent and SPECIAL at 10 per cent levels) except for the LagPADCA,
PREFERRED and LOSS. Overall, healthy, distressed and gray area ?rms have
different characteristics which need to be controlled for in our multivariate analysis.
Table III presents the correlation matrix of our dependent and independent
variables. As depicted in Table III PADCA and Altman Z-score are positively
correlated. On the other hand, Altman Z-score and GINDEX are negatively correlated
with p , 0.001. PADCA and governance score are not correlated signi?cantly at
conventional levels with the association being positive[21].
Table IV presents two-stage least squares (2SLS) regression results[22]. We used
2SLS regressions to resolve the identi?cation problem possibly caused by the Altman
Z-score and PADCA simultaneously in?uencing each other. As shown in Table III,
Altman Z-score (0.0035), and HEALTH (20.0187) of the ?rm (a health indicator
variable) in?uence PADCA signi?cantly at the 1 per cent level. These results indicate
that PADCA increases with an increase in the Altman Z-score, and decreases with ?rm
health. An interaction variable between Altman Z and HEALTH is statistically
signi?cant at the 1 per cent level (t ¼ 3.24) with a coef?cient value of 20.0018. This
indicates that being healthy decreases the effect of Altman Z-score on PADCA, almost
50 per cent from 0.0035 to 0.0017 (0.0035-0.0018). This result indicates that ?nancially
healthy ?rms use less PADCA type accounting discretion than other ?rms. The
coef?cient estimates for DISTRESS and the interaction variable Altman Z
£ DISTRESS are not signi?cant at conventional levels indicating that ?nancially
distressed ?rms are not engaged in manipulating their earnings. Perhaps, they fail to
perceive a bene?t from such manipulation[23].
The second regression reported in Table IV gives the results for the relationship
going from the level of PADCA to the level of Altman Z. That is, it examines whether a
?rm’s Altman Z is related to its use of PADCA. We are particularly concerned with the
coef?cient estimated on the PADCA variable. Notably, that coef?cient (144.944) is
signi?cant at the 0.01 level and suggests that ?rms have better (higher valued) Altman
Z-scores when they make more use of PADCA to in?uence their ?nancial data.
The implication of this ?nding is that analysts who believe that a company is not
distressed as a result of a score from the Altman Z model may be misled if the ?rm’s
use of PADCA has contributed to its perceived Altman Z-score value.
We report 3SLS regression results in Table V. This technique accounts for the
three way endogeneity problem identi?ed in equations (6)-(8). We discuss the three
ARJ
22,2
106
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
A
D
C
A
A
L
T
Z
G
I
N
D
E
X
L
E
V
S
I
Z
E
M
A
L
O
S
S
R
O
A
C
F
O
E
P
S
P
A
D
C
A
0
.
0
6
*
0
.
0
3
2
0
.
0
4
*
2
0
.
0
0
*
0
.
0
0
*
2
0
.
0
7
*
0
.
0
7
*
2
0
.
0
6
*
0
.
0
8
*
A
L
T
Z
0
.
0
4
*
2
0
.
0
9
*
2
0
.
1
9
*
0
.
2
9
*
0
.
0
1
2
0
.
1
7
*
0
.
3
5
*
0
.
5
2
*
0
.
4
6
*
G
I
N
D
E
X
0
.
0
2
2
0
.
0
7
*
0
.
0
9
*
0
.
2
1
*
0
.
1
1
*
2
0
.
2
0
*
0
.
0
7
*
0
.
0
7
*
0
.
0
4
L
E
V
2
0
.
0
6
*
2
0
.
6
5
*
0
.
1
6
*
2
0
.
0
8
*
2
0
.
0
0
*
0
.
0
4
*
2
0
.
0
9
*
2
0
.
1
7
*
2
0
.
1
8
*
S
I
Z
E
0
.
0
0
0
.
2
2
*
0
.
2
4
*
0
.
2
2
*
0
.
2
2
*
2
0
.
3
7
*
0
.
4
3
*
0
.
4
8
*
0
.
4
0
*
M
A
0
.
0
1
0
.
0
1
0
.
1
1
*
0
.
0
6
*
0
.
2
3
*
2
0
.
0
8
*
0
.
0
5
*
0
.
0
8
*
0
.
0
7
*
L
O
S
S
2
0
.
0
8
*
2
0
.
3
6
*
2
0
.
1
5
*
0
.
0
6
*
2
0
.
3
6
*
2
0
.
0
8
*
2
0
.
5
1
*
2
0
.
4
2
*
2
0
.
4
0
*
R
O
A
0
.
0
9
*
0
.
4
6
*
0
.
0
7
*
2
0
.
1
1
*
0
.
4
3
*
0
.
0
9
*
2
0
.
8
7
*
0
.
7
0
*
0
.
5
8
*
C
F
O
2
0
.
1
4
*
0
.
3
8
*
0
.
0
9
*
2
0
.
0
6
*
0
.
4
5
*
0
.
0
7
*
2
0
.
6
3
*
0
.
7
5
*
0
.
6
5
*
E
P
S
0
.
1
0
*
0
.
4
4
*
0
.
0
4
2
0
.
1
4
*
0
.
3
3
*
0
.
0
7
*
2
0
.
8
7
*
0
.
8
9
*
0
.
6
7
*
M
B
0
.
0
6
*
0
.
2
9
*
2
0
.
1
6
*
2
0
.
3
9
*
2
0
.
5
6
*
2
0
.
1
4
*
0
.
1
0
*
2
0
.
0
7
*
2
0
.
1
5
*
0
.
0
0
*
L
I
T
G
2
0
.
0
2
0
.
0
7
*
2
0
.
1
8
*
2
0
.
2
0
*
2
0
.
1
2
*
2
0
.
0
2
0
.
2
4
*
0
.
2
3
*
2
0
.
2
3
*
2
0
.
2
2
*
R
D
2
0
.
0
4
*
0
.
0
5
*
2
0
.
0
7
*
2
0
.
2
5
*
2
0
.
2
5
*
2
0
.
0
5
*
0
.
3
5
*
2
0
.
3
8
*
2
0
.
4
0
*
2
0
.
3
6
*
D
E
P
2
0
.
1
5
*
2
0
.
3
1
*
0
.
0
0
0
.
2
5
*
2
0
.
0
4
*
2
0
.
0
3
*
0
.
1
6
*
0
.
1
0
*
0
.
0
1
2
0
.
2
0
*
P
P
&
E
2
0
.
0
2
*
*
*
2
0
.
2
5
*
0
.
1
6
*
0
.
3
0
*
0
.
1
2
*
2
0
.
0
7
*
2
0
.
1
2
*
0
.
1
0
*
0
.
2
3
*
0
.
0
8
*
I
N
G
2
0
.
0
5
*
2
0
.
0
4
*
0
.
0
8
*
0
.
1
3
*
0
.
2
9
*
0
.
2
4
*
2
0
.
1
2
*
0
.
1
1
*
0
.
1
6
*
0
.
1
0
*
S
P
E
C
0
.
0
8
*
0
.
1
3
*
0
.
0
1
2
0
.
0
6
*
2
0
.
1
0
*
2
0
.
1
1
*
2
0
.
2
9
*
0
.
2
9
*
0
.
0
9
*
0
.
2
9
*
I
N
T
2
0
.
0
3
*
2
0
.
5
1
*
0
.
1
3
*
0
.
7
6
*
0
.
0
7
*
0
.
0
3
*
0
.
0
4
*
2
0
.
0
8
*
0
.
0
1
2
0
.
0
9
*
P
R
E
F
2
0
.
0
2
2
0
.
2
5
*
0
.
0
3
0
.
1
7
*
2
0
.
1
1
*
2
0
.
0
0
0
.
1
5
*
2
0
.
1
7
*
2
0
.
1
5
*
2
0
.
2
0
*
D
I
V
2
0
.
0
1
0
.
1
3
*
0
.
2
7
*
0
.
0
4
*
0
.
4
6
*
0
.
0
8
*
2
0
.
3
8
*
0
.
3
7
*
0
.
3
7
*
0
.
3
2
*
(
c
o
n
t
i
n
u
e
d
)
Table III.
Correlation coef?cients
for all variables included
in simultaneous equation
model
Discretionary
accruals
107
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
M
B
L
I
T
G
R
D
D
E
P
P
P
&
E
I
N
G
S
P
E
C
I
N
T
P
R
E
F
D
I
V
P
A
D
C
A
0
.
0
3
*
2
0
.
0
1
2
0
.
0
5
*
2
0
.
1
6
*
2
0
.
0
4
*
2
0
.
5
0
*
0
.
0
7
*
2
0
.
0
6
*
2
0
.
0
3
*
2
0
.
0
0
*
A
L
T
Z
0
.
3
8
*
0
.
0
7
*
2
0
.
2
6
*
2
0
.
4
1
*
2
0
.
2
7
*
2
0
.
0
3
*
0
.
1
9
*
2
0
.
5
1
*
2
0
.
3
0
*
0
.
0
4
*
G
I
N
D
E
X
0
.
0
1
2
0
.
1
7
*
2
0
.
1
2
*
2
0
.
0
5
*
*
*
0
.
1
4
*
0
.
0
5
*
*
*
0
.
0
1
0
.
0
1
2
0
.
0
6
*
*
2
0
.
0
3
L
E
V
2
0
.
1
2
*
2
0
.
0
1
0
.
0
3
*
0
.
0
7
*
0
.
0
3
*
0
.
0
1
2
0
.
0
2
*
*
*
0
.
1
9
*
0
.
1
6
*
2
0
.
0
1
S
I
Z
E
0
.
0
8
*
2
0
.
1
1
*
2
0
.
3
4
*
2
0
.
1
9
*
0
.
0
6
*
0
.
2
0
*
0
.
0
6
*
2
0
.
2
6
*
2
0
.
0
2
*
2
0
.
0
2
*
*
*
M
A
0
.
0
2
2
0
.
0
2
*
*
2
0
.
0
8
*
2
0
.
0
3
*
2
0
.
0
9
*
0
.
2
2
*
2
0
.
0
4
*
2
0
.
0
3
*
2
0
.
0
2
*
*
2
0
.
0
3
*
L
O
S
S
2
0
.
0
4
*
0
.
2
4
*
0
.
3
4
*
0
.
2
3
*
2
0
.
0
7
*
2
0
.
0
6
*
2
0
.
2
1
*
0
.
1
9
*
0
.
1
7
*
2
0
.
1
9
*
R
O
A
0
.
2
1
*
2
0
.
1
8
*
2
0
.
5
3
*
2
0
.
2
9
*
0
.
0
4
*
0
.
0
2
*
*
*
0
.
2
5
*
2
0
.
4
4
*
2
0
.
2
8
*
0
.
1
1
*
C
F
O
0
.
2
5
*
2
0
.
1
6
*
2
0
.
6
3
*
2
0
.
3
3
*
0
.
0
1
0
.
0
8
*
0
.
1
8
*
2
0
.
4
3
*
2
0
.
3
1
*
0
.
1
1
*
E
P
S
0
.
2
2
*
2
0
.
0
7
*
2
0
.
3
8
*
2
0
.
3
4
*
2
0
.
0
6
*
0
.
0
6
*
0
.
2
7
*
2
0
.
4
5
*
2
0
.
3
3
*
0
.
1
2
*
M
B
0
.
0
2
*
*
*
2
0
.
1
1
*
2
0
.
1
3
*
2
0
.
1
0
*
0
.
0
2
*
*
0
.
0
4
*
2
0
.
2
5
*
2
0
.
1
6
*
0
.
0
1
L
I
T
G
0
.
1
6
*
0
.
3
3
*
0
.
0
4
*
2
0
.
2
2
*
2
0
.
0
9
*
2
0
.
0
5
*
2
0
.
0
5
*
0
.
0
4
*
2
0
.
1
0
*
R
D
0
.
2
6
*
0
.
4
8
*
0
.
2
7
*
2
0
.
0
5
*
2
0
.
0
9
*
2
0
.
1
1
*
0
.
1
5
*
0
.
2
3
*
2
0
.
1
0
*
D
E
P
2
0
.
1
4
*
2
0
.
0
3
*
0
.
0
4
*
0
.
4
1
*
0
.
0
4
*
2
0
.
3
0
*
0
.
3
1
*
0
.
2
1
*
2
0
.
0
5
*
P
P
&
E
2
0
.
2
4
*
2
0
.
1
2
*
2
0
.
2
8
*
0
.
5
6
*
2
0
.
2
3
*
2
0
.
0
2
0
.
1
7
*
0
.
0
3
*
0
.
1
1
I
N
G
2
0
.
1
7
*
2
0
.
1
2
*
2
0
.
1
3
*
2
0
.
1
8
*
2
0
.
1
1
*
2
0
.
0
4
*
0
.
0
3
*
*
2
0
.
0
1
2
0
.
0
7
*
S
P
E
C
0
.
1
3
*
2
0
.
0
6
*
2
0
.
0
8
*
2
0
.
1
8
*
0
.
0
2
*
*
2
0
.
1
2
*
2
0
.
1
7
*
2
0
.
0
8
*
0
.
0
5
*
I
N
T
2
0
.
3
7
*
2
0
.
2
4
*
2
0
.
3
4
*
0
.
2
2
*
0
.
3
1
*
0
.
1
7
*
2
0
.
0
5
*
0
.
2
4
*
2
0
.
0
7
*
P
R
E
F
2
0
.
1
2
*
2
0
.
0
0
0
.
0
1
0
.
1
0
*
0
.
0
3
0
.
0
2
2
0
.
0
3
4
0
.
1
5
*
2
0
.
0
5
*
D
I
V
2
0
.
2
9
*
2
0
.
2
6
*
2
0
.
2
9
*
0
.
0
4
*
0
.
3
0
*
0
.
1
3
*
0
.
0
5
*
0
.
0
5
*
2
0
.
0
7
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
,
*
*
5
a
n
d
*
*
*
1
0
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
a
l
l
v
a
r
i
a
b
l
e
d
e
?
n
i
t
i
o
n
s
a
r
e
p
r
o
v
i
d
e
d
i
n
T
a
b
l
e
I
;
n
u
m
b
e
r
o
f
o
b
s
e
r
v
a
t
i
o
n
f
o
r
G
I
N
D
E
X
i
s
1
,
3
4
0
a
n
d
f
o
r
o
t
h
e
r
v
a
r
i
a
b
l
e
s
8
,
6
5
5
;
P
e
a
r
s
o
n
c
o
e
f
?
c
i
e
n
t
s
a
r
e
i
n
u
p
p
e
r
r
i
g
h
t
a
n
d
S
p
e
a
r
m
a
n
c
o
e
f
?
c
i
e
n
t
s
a
r
e
i
n
l
o
w
e
r
l
e
f
t
Table III.
ARJ
22,2
108
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
regressions separately starting with the PADCA regressions. Unlike in the 2SLS
regression where the coef?cient estimated for the Altman Z was positive, with 3SLS the
coef?cient is negative (t ¼ 22.55) and signi?cant at 5 per cent level. The other results
are similar between the 2SLS and 3SLS. The healthy dummy coef?cient is 20.0241
(t ¼ 21.94) which indicates that healthy ?rms have lower discretionary accruals.
Distressed ?rms also have lower discretionary accruals. We suspect that this later case
arises because these ?rms have fewer discretionary accrual opportunities.
The GINDEX coef?cient is 0.0221 and signi?cant at 1 per cent level. This result
indicates that companies with good corporate governance use income increasing
discretionary accruals. This positive association may be attributable to governance
quality and the subsequent performance outcome may bene?t the shareholders
(Subramanyam, 1996). Governance affects the use of PADCA by ?rms possibly
Dependant variables
PADCA Altman Z
Independent variables Coef?cient t-statistics Coef?cient t-statistics
Intercept 0.0381 4.22
* * *
20.2380 20.19
Altman Z 0.0035 4.96
* * *
PADCA 144.944 6.52
* * *
HEALTH 20.0187 22.69
* * *
Altman Z £ HEALTH 20.0018 23.24
* * *
DISTRESS 0.0061 0.80
Altman Z £ DISTRESS 20.0000 20.61
LagPADCA 0.0101 0.83
LEVERAGE 20.0028 20.96
SIZE 20.0029 22.56
* *
0.5450 4.01
* * *
MERGER 0.0074 1.44 21.8104 2.84
* * *
LOSS 20.0335 25.64
* * *
1.5252 2.30
* *
ROA 0.0744 8.54
* * *
29.2826 25.80
* * *
CFO 20.1510 212.79
* * *
28.8656 10.11
* * *
LagAltman 0.1801 16.55
* * *
EPS 213.4442 21.69
*
MB 0.9028 5.38
* * *
LITIGATION 0.2234 0.36
R&D 8.6875 3.49
* * *
DEPRECIATION 48.3391 3.87
* * *
PP&E 26.1869 26.45
* * *
INTANGIBLE 22.2409 21.12
SPECIAL 6.5919 3.61
* * *
INTEREST 236.3476 26.86
* * *
PREFERRED 26.34052 24.67
* * *
DIVIDEND 299.9099 1.17
Adjusted R
2
(per cent) 12.95 4.74
No. of observations 8,655 8,655
Notes: Signi?cant at:
*
0.10,
* *
0.05 and
* * *
0.01 per cent levels, respectively; all variables are
winsorized at 1st and 99th percentiles; HEALTH is an indicator variable that takes the value of 1 if the
Z-score exceeds the value of 2.67, and zero otherwise; DISTRESS is also an indicator variable that
takes the value 1 if Z-score is smaller than 1.81 and 0 otherwise; other variable de?nitions are provided
in Table I
Table IV.
2SLS regression
estimates
Discretionary
accruals
109
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
D
e
p
e
n
d
a
n
t
v
a
r
i
a
b
l
e
s
P
A
D
C
A
A
l
t
m
a
n
Z
G
I
N
D
E
X
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
I
n
t
e
r
c
e
p
t
2
0
.
1
4
7
0
2
2
.
2
0
*
*
1
8
.
3
8
0
4
5
.
6
0
*
*
*
7
.
6
7
6
4
1
4
.
2
0
*
*
*
A
l
t
m
a
n
Z
2
0
.
0
0
7
7
2
2
.
5
5
*
*
2
0
.
1
1
7
6
2
3
.
8
6
*
*
*
P
A
D
C
A
4
1
.
9
1
5
5
2
.
5
5
*
*
2
.
6
2
8
4
1
.
1
0
G
I
N
D
E
X
0
.
0
2
2
1
2
.
9
3
*
*
*
2
1
.
9
7
2
8
2
4
.
9
5
*
*
*
H
E
A
L
T
H
2
0
.
0
2
4
1
2
1
.
9
4
*
A
l
t
m
a
n
Z
£
H
E
A
L
T
H
0
.
0
1
1
7
1
3
.
6
4
*
*
*
D
I
S
T
R
E
S
S
2
0
.
0
3
6
4
2
3
.
1
2
*
*
*
A
l
t
m
a
n
Z
£
D
I
S
T
R
E
S
S
0
.
0
1
8
0
8
.
3
8
*
*
*
L
a
g
P
A
D
C
A
2
0
.
0
1
4
2
2
0
.
4
9
L
E
V
E
R
A
G
E
2
0
.
0
0
8
5
2
0
.
3
3
S
I
Z
E
2
0
.
0
0
3
1
2
0
.
8
3
0
.
3
3
3
2
1
.
9
4
*
0
.
2
9
3
6
4
.
1
9
*
*
*
L
O
S
S
0
.
0
0
2
5
0
.
2
9
2
0
.
5
6
3
2
2
1
.
3
6
R
O
A
0
.
1
5
3
8
3
.
8
8
*
*
*
2
0
.
9
0
3
9
2
0
.
2
0
C
F
O
2
0
.
3
5
8
8
2
8
.
3
2
*
*
*
1
9
.
4
3
2
4
2
.
8
3
*
*
*
L
a
g
A
l
t
m
a
n
0
.
3
3
1
8
1
8
.
2
1
*
*
*
E
P
S
1
9
.
4
2
7
1
0
.
8
5
M
B
0
.
0
9
3
0
0
.
3
8
L
I
T
I
G
A
T
I
O
N
2
1
.
1
0
8
3
2
2
.
3
9
*
*
2
0
.
5
7
7
5
2
2
.
6
4
*
*
*
R
&
D
3
.
6
8
9
7
0
.
8
4
D
E
P
R
E
C
I
A
T
I
O
N
2
1
3
.
6
7
0
1
2
1
.
4
4
P
P
&
E
2
0
.
2
3
3
1
2
0
.
1
9
I
N
T
A
N
G
I
B
L
E
2
2
.
2
2
4
5
2
2
.
5
6
*
*
S
P
E
C
I
A
L
2
0
.
5
1
9
5
2
0
.
2
7
I
N
T
E
R
E
S
T
2
1
4
.
9
2
5
2
2
1
.
6
2
M
E
R
G
E
R
0
.
2
7
1
0
1
.
6
3
P
R
E
F
E
R
R
E
D
2
1
.
3
2
7
3
2
0
.
9
2
D
I
V
I
D
E
N
D
1
3
1
.
2
1
6
1
1
.
0
3
S
y
s
t
e
m
w
e
i
g
h
t
e
d
R
2
(
p
e
r
c
e
n
t
)
4
3
.
5
1
D
u
r
b
i
n
-
W
a
t
s
o
n
1
.
8
9
1
.
8
9
2
.
0
3
F
-
v
a
l
u
e
3
.
7
9
*
*
*
3
.
7
4
*
*
*
2
.
0
3
*
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
0
.
1
0
,
*
*
0
.
0
5
a
n
d
*
*
*
0
.
0
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
v
a
r
i
a
b
l
e
d
e
?
n
i
t
i
o
n
s
a
r
e
g
i
v
e
n
i
n
T
a
b
l
e
I
;
n
¼
7
6
7
Table V.
3SLS regression
estimates with
governance index
ARJ
22,2
110
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
because managers may use their discretion to achieve some reporting objective to
bene?t shareholders[24]. Managers apparently are not grilled on the details of ?nancial
reports by non-executive members of their board of directors. This may be due to the
extreme level of detail that would be required for the board of directors to uncover
cases of PADCA[25].
The second regression equation has Altman Z-score as the dependent variable.
Here, GINDEX is negatively associated with Altman Z-score. This indicates that
as ?rms strengthen their corporate governance that their Altman Z-score declines.
Since this result is found with the 3SLS regression there is no confusion as to the
direction of the relationship. That is, our result suggest that as corporate governance
strengthens that ?rms become less healthy which may occur because they may lose
markets and pro?ts to ?rms with less governance. The positive and signi?cant
coef?cient on PADCA suggests that the more ?rms use discretionary accruals the
better their ?nancial health appears, as measured by Altman Z.
In equation (3), the GINDEX is the dependent variable. The signi?cant coef?cient on
the Altman Z-variable suggests that the higher is the Altman Z-score (the better the
?rm’s ?nancial health) the lower is the GINDEX. This result serves as a counterpoint to
the results in the second equation; now the argument is that members of the board of
directors of healthier ?rms are less vigilant and have lower governance scores. The
coef?cient estimated on PADCA is not signi?cant, as we expected a priori. The amount
of discretionary accruals does not affect the governance index. We also (results not
shown) tested lagged PADCA as a determinant of the GINDEX thinking that the board
might increase its vigilance after managers engaged in discretionary accruals. The
coef?cient estimated on lagged PADCA was not signi?cant.
Perhaps, the most interesting ?nding is that a higher GINDEX lowers a ?rm’s
Altman Z while a higher Altman Z lowers the ?rms GINDEX. The interpretation is that
better corporate governance is not associated with improved corporate health and that
improved corporate health is not associated with better corporate governance. Firms
may reach an equilibrium level of GINDEX and Altman Z after which increases in
either governance or health leads to undesired results.
7. Conclusion
We investigate the relationship between discretionary accrual level, ?rms’ ?nancial
distress as measured by the Altman Z, and corporate governance quality as extracted
from the IRRC database applied to US manufacturing ?rms between the years 2001
and 2003. We ?nd that ?rms’ managers appear to exercise discretionary accruals when
their ?rms are in the mid-range or gray area of the Altman Z in order to avoid being
classi?ed as distressed or if they are successful in doing it to be classi?ed as a healthy
?rm. We did not observe the same behavior for ?rms in either the distressed or healthy
classi?cation areas based on the Altman Z-score. Gray area companies have larger
PADCAs. There are two possible explanation for this; they are either opportunistic or
they face noise in their accounting information. We do not ?nd that heightened
governance reduces the incidence of discretionary accruals. Our ?nding suggests that
non-executive members of boards of directors need to pay closer attention to ?rms in
the gray region.
This study contributes to both the accounting and ?nance literature in three ways.
First of all, we document the relationship between a ?rms’ ?nancial health and its use of
Discretionary
accruals
111
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
discretionary accruals by taking into consideration corporate governance quality. This
result is important because it shows the existence of agency problems in credit markets
when ?nancials signal stressful results especially when there is good corporate
governance. Second, we document the importance of paying attention to ?rms at the
boundaries of healthy and gray areas as they appear to be more open to accrual
manipulation. This result is interesting in the sense that both creditor and equity
investors must be careful about the ?nancial classi?cations of ?rms when they evaluate
potential investments. They should also pay close attention to accruals, if the ?rm is on
the boundary or in the gray area that corporate governance may play intervening role
in this.
When endogeneity is tested between the three variables, GINDEX, Altman Z and
PADCA, the 3SLS results yield an interesting conclusion that better corporate
governance is not associated with improved corporate health and that improved
corporate health is not associated with better corporate governance. We suspect that
?rms that deviate from an equilibrium or best level for their GINDEX and Altman Z
suffer unexpected changes in the other variable.
Notes
1. One may argue that a ?rm under close scrutiny may not be able to manage its earnings since
detection of earnings management would not be good for managers. On the other hand,
managers know that ?nancial reports are value relevant because they are closely followed
therefore they are motivated to fudge the numbers to in?uence the investors’ expectations.
2. Some might argue that a 40-year-old model, the Altman Z, is not the best choice since other
models exist (e.g. the Shumway, the Merton, the KMV, or the Ohlson, 1980 models). Our
choice of Altman’s model was based on its widespread acceptance and reuse throughout the
world. Some authors, Begley, Ming and Watts, for example argue that his cutoff points are
outdated. We continue with the original cutoff points since our objective is not to predict
?nancially distressed ?rms but rather to document the relationship between ?nancial status
classi?cation, corporate governance quality and the likelihood managers’ using
discretionary accruals.
3. The advantage of using an Altman Z-approach to classify ?rms is that it “considers an entire
pro?le of characteristics shared by a speci?c ?rm and transforms that into a univariate
statistic” (Cleary, 1999).
4. Reasoning behind the use of PADCA speci?cally to measure discretionary accruals will be
explained in subsequent sections.
5. The sample period is chosen speci?cally to select ?rms with a higher probability of having
lower Altman Z-scores than during periods of economic health. We also choose the
manufacturing industry since it has been established that the Altman Z-methodology works
well for these ?rms. We also control for leverage and other ?rm characteristics as a result of
the signaling hypothesis.
6. The more discretion exercised by management, the greater the opportunity for them to
manipulate reported earnings.
7. The Altman Z-score is a widely used classi?cation and bankruptcy estimation measurement
tool in the ?eld of credit risk analysis, distressed investing; mergers and acquisitions (M&A)
target analysis, and turnaround management by both practitioners and academics
(Calandro, 2007).
ARJ
22,2
112
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
8. Accruals may also be used opportunistically by executives, because they need to make
judgments while reporting. Therefore, the greater a company’s management discretion
concerning accruals the greater the opportunity to manipulate reported earnings.
9. When calculating PADCA, previous year ROA is used. That may create limitation in our
research design since we do not match ?rms according to their ?nancial status, i.e. Altman
Z-score. This is also limitation of PADCA measures that assume ?rms are identical at the
previous year just by looking at their ROA. But this method is the most ef?cient one
available compared with other methods (Kothari et al., 2005). In our design to decrease this
measurement issue, we used ROA at the current period as a control variable that is
correlated signi?cantly with Altman Z, PADCA and G-score.
10. Regular ordinary least squares (OLS) regression provided similar results with the same
coef?cient signs. Firms were divided into distressed, gray and healthy classi?cations and
OLS was run for each subsample. We tested for multicollinearity in both our 2SLS and 3SLS
speci?cation and rejected it.
11. We employed 2SLS to solve the endogeneity problem between PADCA and Altman Z
without taking into consideration corporate governance quality where sample size decreases
from 8,655 to 767 with 3SLS regression.
12. Ali et al. (2007) investigates the corporate disclosure practices of family ?rms, and they
document that family ?rms report better disclosure quality earnings and are more likely to
warn for a given magnitude of bad news in comparison with non-family ?rms. We also
reported the 2SLS regression model results because with 3SLS our sample size decreases
from 8,655 to 767. Some inferences are also affected because of this sample size difference,
but overall our conclusions are similar.
13. The t-statistics are corrected using the Huber-White procedure by following Petersen (2009).
14. Discretionary accruals reverse over time (Guay et al., 1996). They state that “the
discretionary accrual is expected to reverse in the future periods and nondiscretionary
earnings are also expected to decline”. They do not say accruals reverse at the next period,
but in the future. For example, a manager may think to boost earnings by choosing a lower
depreciation expense in the current period. Of course, a higher expense is required in
subsequent periods to make up for the difference but it may take many years for the two
depreciation methods to be equal. The authors recognize this problem and control for
depreciation and other kinds of discretionary accruals. The correlation and time series
properties of discretionary accruals and earning process are shown in Beaver et al. (1980) and
Ashton (2005). They document that the serial correlation of discretionary accruals from one
to subsequent periods is negative but not perfect which indicates that they do not reverse
next period but takes time. On the other hand, one discretionary accrual may be replaced
with another one over time that provides a large room for managers to maneuver their
?nancial reporting strategies. The time horizon of our study may not be long enough for
discretionary accruals to reverse.
15. Altman Z-score is a widely used classi?cation and bankruptcy estimation measurement tool
in the ?eld of credit risk analysis, distressed investing; M&A target analysis and turnaround
management by both practitioners and academics (Calandro, 2007).
16. The LagPADCA variable is included in the regression model to test for the reversal of
accruals over time.
17. Firms in the Compustat database for which the required data is available are included in the
estimation process. We also use the two-digit industry de?nition for these tests.
18. Obviously, there are other periods of market downturn before 2001. Moreover, the economic
downturns may have different effects on the business environment. The sample period may
Discretionary
accruals
113
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
be extended to other periods when the US economy hit bottom in November 1982 and March
1991 according to National Bureau of Economic Research.
19. Mean tests would not be conclusive since other ?rms’ characteristics are diverse for different
classes of ?rms, i.e. healthy, gray and distress ?rms as reported at Table II.
20. Since markets expect managers to manage their earnings, investors discount the value of the
earnings such as for $100 reported earnings they discount it to $95. Therefore, managers
manage the earnings to $105 to reach equilibrium at $100.
21. We look at the Pearson and Spearman correlations between lagged PADCA and GINDEX
and could not ?nd association as well.
22. We include 2SLS regression results, because our sample size decreases drastically when we
include GINDEX in our regression analysis which is reported as the 3SLS model.
23. Variance in?ation factor (VIF) are found to range between 1 and 4 which is less than the
critical value of 20. Therefore, we conclude that there is not a multicollinearity problem.
24. We do not test the market response, therefore, we may not know whether using discretionary
accruals bene?ts shareholders.
25. VIF scores for the variables that are used in 3SLS have the values that are less than 20,
therefore, we do not have tolerable multicollinearity problem.
References
Ali, A., Chen, T. and Radhakrishnan, S. (2007), “Corporate disclosures by family ?rms”, Journal
of Accounting and Economics, Vol. 44, pp. 238-86.
Altman, E. (1968), “Financial ratios, dicriminant analysis and prediction of corporate
bankruptcy”, Journal of Finance, Vol. 23, pp. 589-609.
Al-Tuwaijri, S.A., Christensenand, T.E. and Hughes, K.E. II (2004), “The relation among
environmental disclosure, environmental performance, and economic performance:
a simultaneous equations approach”, Accounting, Organizations and Society, Vol. 29,
pp. 447-71.
Ashton, D. (2005), “Discussion of cost of capital, strategic disclosures and accounting choice”,
Journal of Business Finance & Accounting, Vol. 32, pp. 635-41.
Beaver, W.H., Lamber, R. and Morse, D. (1980), “The information content of security prices”,
Journal of Accounting and Economics, Vol. 2 No. 1, pp. 3-28.
Bernanke, B. and Gertler, M. (1990), “Financial fragility and economic performance”,
The Quarterly Journal of Economics, Vol. 105, pp. 87-114.
Bernstein, L. (1993), Financial Statement Analysis, 5th ed., Irwin, Homewoood, IL.
Bowen, R., Ducharmeand, L. and Shores, D. (1995), “Stakeholders’ implicit claims and accounting
method choice”, Journal of Accounting and Economics, Vol. 20, pp. 255-95.
Bowen, R., Rajgopal, S. and Venkatachalam, M. (2008), “Accounting discretion, corporate
governance, and ?rm performance”, Contemporary Accounting Research, Vol. 25,
pp. 310-405.
Burgstahler, D. and Dichev, I. (1997), “Earnings management to avoid earnings decreases and
losses”, Journal of Accounting and Economics, Vol. 24, pp. 99-126.
Bushee, B.J. (1998), “The in?uence of institutional investors on myopic R&D investment
behavior”, The Accounting Review, Vol. 73, pp. 305-34.
Calandro, J. (2007), “Considering the utility of Altman’s Z-score as a strategic assessment and
performance management tool”, Strategy & Leadership, Vol. 35, pp. 37-43.
ARJ
22,2
114
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Chi, L. and Tang, T. (2007), “Impact of reorganizational announcements on distressed-stock
return”, Economic Modeling, Vol. 24, pp. 749-67.
Clarke, T. (2005), “Accounting for Enron: shareholder value and stakeholder interests”,
Corporate Governance: An International Review, Vol. 13, pp. 598-612.
Cleary, S. (1999), “The relationship between ?rm investment and ?nancial status”, Journal
of Finance, Vol. 54, pp. 673-92.
Cook, K.A., Huston, G.R. and Omer, T.C. (2008), “Earnings management through effective tax
rates: the effect of tax planning investment and Sarbanes-Oxley Act of 2002”,
Contemporary Accounting Research, Vol. 25 No. 2, pp. 447-71.
Dechow, P.M. and Schrand, C.M. (2004), Earnings Quality, The Research Foundation of CFA
Institute, New York, NY.
Dechow, P.M. and Skinner, D. (2000), “Earnings management: reconciling the views of
accounting academics, practitioners and regulators”, Accounting Horizons, Vol. 14,
pp. 235-50.
Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1995), “Detecting earnings management”,
The Accounting Review, Vol. 70, pp. 3-42.
Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1996), “Causes and consequences of earnings
manipulation: an analysis of ?rms subject to enforcement actions by the SEC”,
Contemporary Accounting Research, Vol. 13, pp. 1-36.
DeFond, M. and Jiambalvo, J. (1994), “Debt covenant violation and manipulation of accruals”,
Journal of Accounting and Economics, Vol. 17, pp. 145-76.
Del Guercio, D. (1996), “The distorting effect of the prudent-man laws on institutional equity
investments”, Journal of Financial Economics, Vol. 40, pp. 31-62.
Dhaliwal, D.S., Gleason, C. and Mills, L. (2004), “Last-chance earnings management: using the tax
expense to meet analysts’ forecasts”, Contemporary Accounting Research, Vol. 21,
pp. 431-59.
Dhaliwal, D.S., Li, O.Z. and Xie, H. (2005), “Institutional investors, ?nancial health equity
valuation”, available at: http://ssrn.com/abstract¼493046 (accessed 3 June 2009).
Fair?eld, P., Whisenant, J. and Yohn, T. (2003), “Accrued earnings and growth: implications for
future pro?tability and market mispricing”, The Accounting Review, Vol. 78, pp. 353-71.
Farrell, K. and Whidbee, D. (2003), “Impact of ?rms performance expectations on CEO turnover
and replacement decisions”, Journal of Accounting and Economics, Vol. 36, pp. 165-96.
Fazzari, S., Hubbard, G. and Petersen, B. (1988), “Financing constraints and corporate
investment”, Brooking Papers on Economic Activity, Vol. 1, pp. 141-206.
Fields, T.D., Lys, T.Z. and Vincent, L. (2001), “Empirical research of accounting choice”, Journal
of Accounting and Economics, Vol. 31, pp. 255-307.
Francis, J., Huang, A.H., Rajgopal, S. and Zang, A.Y. (2008), “CEO reputation and earnings
quality”, Contemporary Accounting Research, Vol. 25, pp. 109-47.
Gertler, M. (1992), “Financial capacity and output ?uctuations in an economy with multi-period
?nancial relationships”, Review of Economic Studies, Vol. 59 No. 3, pp. 455-72.
Gompers, P.A., Ishii, J.L. and Metrick, A. (2003), “Corporate governance and equity prices”,
The Quarterly Journal of Economics, Vol. 118, pp. 107-55.
Graham, J.R., Harvey, C.R. and Rajgopal, S. (2005), “The economic implications of corporate
?nancial reporting”, Journal of Accounting and Economics, Vol. 40, pp. 3-73.
Grice, J.S. and Ingram, R.W. (2001), “Tests of the generalizability of Altman’s bankruptcy
prediction model”, Journal of Business Research, Vol. 54, pp. 53-61.
Discretionary
accruals
115
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Guay, W.R., Kothari, S.P. and Watts, R.L. (1996), “A market-based evaluation of discretionary
accrual models”, Journal of Accounting Research, Vol. 34, pp. 83-105.
Healy, P. (1985), “The impact of bonus schemes on the selection of accounting principles”, Journal
of Accounting and Economics, Vol. 7, pp. 85-107.
Healy, P. and Wahlen, J. (1999), “A review of the earnings management literature and its
implications for standard setting”, Accounting Horizons, Vol. 13, pp. 365-83.
Hessel, C.A. and Norman, M. (1992), “Financial characteristics of neglected and institutionally
held stocks”, Journal of Accounting, Auditing and Finance, Vol. 7 No. 3, pp. 313-30.
Holthausen, R., Larcker, D. and Sloan, R.G. (1995), “Annual bonus schemes and manipulation of
earnings: additional evidence on bonus plans and income management”, Journal
of Accounting and Economics, Vol. 19 No. 1, pp. 29-74.
Hsu, G.C.-M. and Koh, P.-S. (2005), “Does the presence of institutional investors in?uence
accruals management? Evidence from Australia”, Corporate Governance: An International
Review, Vol. 13, pp. 809-23.
Jones, J. (1991), “Earnings management during import relief investigations”, Journal
of Accounting Research, Vol. 29, pp. 193-228.
Jones, K.L., Krishnan, G.V. and Melendrez, D. (2008), “Do models of discretionary accruals detect
actual cases of fraudulent and restated earnings? An empirical analysis”, Contemporary
Accounting Research, Vol. 25, pp. 499-531.
Kaplan, S. and Zingales, L. (1997), “Do investment cash-?ow sensitivities provide useful
measures of ?nancing constraints?”, Quarterly Journal of Economics, Vol. 112, pp. 169-216.
Kasznik, R. (1999), “On the association between voluntary disclosure and earnings
management”, Journal of Accounting Research, Vol. 37, pp. 57-81.
Kothari, S.P., Leone, A. and Wasley, C. (2005), “Performance matched discretionary accrual
measures”, Journal of Accounting and Economics, Vol. 39, pp. 163-97.
Mather, P. and Ramsay, A. (2006), “The effects of board characteristics on earnings management
around Australian CEO changes”, Accounting Research Journal, Vol. 19, pp. 78-93.
Matsunaga, S. and Park, C. (2001), “The effect of missing quarterly earnings benchmark on the
CEO’s annual bonus”, The Accounting Review, Vol. 76, pp. 313-32.
Merton, R.C. (1974), “On the pricing of corporate debt: the risk structure of interest rates”, Journal
of Finance, Vol. 29, pp. 449-70.
Ohlson, J. (1980), “Financial ratios and the probabilistic prediction of bankruptcy”, Journal
of Accounting Research, Vol. 18 No. 1, pp. 109-31.
Ohlson, J. (1995), “Earnings, book values, and dividends in equity valuation”, Contemporary
Accounting Research, Vol. 18 No. 1, pp. 661-87.
Petersen, M. (2009), “Estimating standard errors in ?nance panel data sets: comparing
approaches”, Review of Financial Studies, Vol. 22, pp. 435-80.
Platt, H.D. and Platt, M.B. (1991), “A note on the use of industry-relative ratios in bankruptcy
prediction”, Journal of Banking & Finance, Vol. 15, pp. 183-1194.
Rees, L., Gill, S. and Gore, R. (1996), “An investigation of asset write-downs and concurrent
abnormal accruals”, Journal of Accounting Research, Vol. 34, pp. 157-69.
Richardson, S.A., Tuna, A.I. and Wu, M. (2002), “Predicting earnings management: the case of
earnings restatements”, available at: http://ssrn.com/abstract¼338681 (accessed 3 June
2008).
Shleifer, A. and Vishny, R.W. (1997), “A survey of corporate governance”, Journal of Finance,
Vol. 52, pp. 737-83.
ARJ
22,2
116
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Shumway, T. (2001), “Forecasting bankruptcy more accurately: a simple hazard model”,
The Journal of Business, Vol. 74, pp. 101-24.
Sloan, R.G. (1996), “Do stock prices fully re?ect information in accruals and cash ?ows about
future earnings?”, The Accounting Review, Vol. 71, pp. 289-316.
Subramanyam, K.R. (1996), “The pricing of discretionary accruals”, Journal of Accounting and
Economics, Vol. 22, pp. 249-81.
Watts, R.L. and Zimmerman, J. (1990), “Positive accounting theory: a ten year perspective”,
The Accounting Review, Vol. 65, pp. 131-56.
Yeh, Y.H. and Lee, T.-S. (2004), “Corporate governance and ?nancial distress: evidence from
Taiwan”, Corporate Governance: An International Review, Vol. 12, pp. 378-88.
Zmijewski, M. (1984), “Methodological issues related to the estimation of ?nancial distress
prediction models”, Journal of Accounting Research, Vol. 22, pp. 59-82.
Appendix. PADCA calculation
Following Kothari et al. (2005), we estimate the cross-sectional Jones model for each two-digit SIC
industry group with at least ten ?rms in year t. Speci?cally, we estimate the following modi?ed
Jones model to determine coef?cient values and the value of the error term:
TCA
j;t
¼ b
0
þ b
1
ðSCASSET
j;t
Þ þ b
1
ðDREV
j;t
2DAR
j;t
Þ þ error
j;t
ðA1Þ
where TCA – change in non-cash current assets minus the change in current liabilities excluding
the current portion of long-term debt so that TCA ¼ DCurrent assets-DCash and short-term
investments-DCurrent liabilities þ DDebt in current liabilties; SCASSET – one scaled by lagged
total assets (1/lagASSET); DREV
j,t
– change in total revenue (sales) for ?rm j at year t scaled by
lagged total asset; DAR
j,t
¼ change in account receivable for ?rm j at year t scaled by lagged
total assets; error
j,t
– error term for equation (A1) for ?rm j at year t.
The two-digit industry and year-speci?c estimates obtained from estimating equation (A1)
are used to compute normal accruals (NA
j,t
). The difference between TCA
j,t
and NA
j,t
is the
abnormal (discretionary) accruals for ?rm j in year t. Following Kothari et al. (2005) ?rm j in year
t is matched with a companion ?rm with the closest lagged-return on assets in the same two-digit
SIC code in year t. We obtain PADCA by subtracting matched ?rm’s abnormal accruals from
?rm j’s abormal accrual at year t.
Corresponding author
Harlan Platt can be contacted at: [email protected]
Discretionary
accruals
117
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Nuraddeen Usman Miko, Hasnah Kamardin. 2015. Impact of Audit Committee and Audit Quality on
Preventing Earnings Management in the Pre- and Post- Nigerian Corporate Governance Code 2011.
Procedia - Social and Behavioral Sciences 172, 651-657. [CrossRef]
2. Aziatul Waznah Ghazali, Nur Aima Shafie, Zuraidah Mohd Sanusi. 2015. Earnings Management: An
Analysis of Opportunistic Behaviour, Monitoring Mechanism and Financial Distress. Procedia Economics
and Finance 28, 190-201. [CrossRef]
3. Nor Farhana Selahudin, Nor Balkish Zakaria, Zuraidah Mohd. Sanusi. 2014. Remodelling the Earnings
Management with the Appearance of Leverage, Financial Distress and Free Cash Flow: Malaysia and
Thailand Evidences. Journal of Applied Sciences 14, 2644-2661. [CrossRef]
4. Nor Farhana Selahudin, Nor Balkish Zakaria, Zuraidah Mohd Sanusi, Pornanong Budsaratragoon. 2014.
Monitoring Financial Risk Ratios and Earnings Management: Evidence from Malaysia and Thailand.
Procedia - Social and Behavioral Sciences 145, 51-60. [CrossRef]
5. Panagiotis Dimitropoulos. 2011. Corporate Governance and Earnings Management in the European
Football Industry. European Sport Management Quarterly 11, 495-523. [CrossRef]
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_425876692.pdf
The purpose of this paper is to investigate, using data on US manufacturing firms, how
and when corporate governance affects managers’ decisions to use discretionary accruals and thereby
artificially influence company financial reports
Accounting Research Journal
Financial status, corporate governance quality, and the likelihood of managers using
discretionary accruals
Sebahattin Demirkan Harlan Platt
Article information:
To cite this document:
Sebahattin Demirkan Harlan Platt, (2009),"Financial status, corporate governance quality, and the likelihood
of managers using discretionary accruals", Accounting Research J ournal, Vol. 22 Iss 2 pp. 93 - 117
Permanent link to this document:
http://dx.doi.org/10.1108/10309610910987475
Downloaded on: 24 January 2016, At: 21:08 (PT)
References: this document contains references to 60 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1957 times since 2009*
Users who downloaded this article also downloaded:
Marion Hutchinson, (2009),"Governance issues in accounting", Accounting Research J ournal, Vol. 22 Iss 2
pp. 89-92 http://dx.doi.org/10.1108/10309610910987466
Raghavan J . Iyengar, J udy Land, Ernest M. Zampelli, (2010),"Does board governance improve
the quality of accounting earnings?", Accounting Research J ournal, Vol. 23 Iss 1 pp. 49-68 http://
dx.doi.org/10.1108/10309611011060524
Ahsan Habib, Istiaq Azim, (2008),"Corporate governance and the value-relevance of accounting
information: Evidence from Australia", Accounting Research J ournal, Vol. 21 Iss 2 pp. 167-194 http://
dx.doi.org/10.1108/10309610810905944
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Financial status, corporate
governance quality, and the
likelihood of managers using
discretionary accruals
Sebahattin Demirkan
School of Management, SUNY Binghamton University,
Vestal, New York, USA, and
Harlan Platt
College of Business Administration, Northeastern University,
Boston, Massachusetts, USA
Abstract
Purpose – The purpose of this paper is to investigate, using data on US manufacturing ?rms, how
and when corporate governance affects managers’ decisions to use discretionary accruals and thereby
arti?cially in?uence company ?nancial reports.
Design/methodology/approach – Three-stage least squares is employed to study the relationship
between ?nancial status, corporate governance and ?nancial reporting discretion. The sample spans
the years 2001-2003 during a severe downturn in the US stock market. Financial status is measured
with the Altman Z-score.
Findings – A signi?cant difference is found between ?rms not classi?ed as healthy or failed (i.e. the
mid-range group) and the two extreme categories when examining governance quotient using a
well-known index. A positive relationship is found between discretionary accruals and the governance
index. Strong governance appears to reduce the incidence of mid-range ?rms engaging in accruals
management. The least healthy and the most distressed companies have the weakest relationship
with discretionary accruals. By contrast, mid-range ?rms are more likely to resort to discretionary
accruals.
Practical implications – Non-executive members of boards of directors are warned to be
particularly vigilant about discretionary accruals with ?rms transitioning between healthy and
high-failure risk.
Originality/value – The relationship between ?rms’ ?nancial health and discretionary accruals
reveals an agency problem in credit markets with ?nancially stressed ?rms. More attention is required
on ?rms whose ?nancial condition is uncertain. Also, it is documented that signi?cant ?ndings of
importance to the earnings quality and corporate governance literature by documenting the role of
corporate governance on discretionary accruals and ?nancial status.
Keywords Corporate governance, Financial reporting, Financial performance, UnitedStates of America,
Manufacturing industries
Paper type Research paper
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
The authors are grateful to the reviewers for their keen insights and extraordinary diligence.
The authors also thank Arnie Wright for his valuable comments. The authors would like to
acknowledge ?nancial support from College of Business Administration, Northeastern
University and School of Management, SUNY Binghamton University.
Discretionary
accruals
93
Accounting Research Journal
Vol. 22 No. 2, 2009
pp. 93-117
qEmerald Group Publishing Limited
1030-9616
DOI 10.1108/10309610910987475
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
1. Introduction
Accounting literature uses discretionary accruals to measure earnings management and
market ef?ciency since earnings are composed of cash ?ow from operations (CFO) and
accruals (DeFond and Jiambalvo, 1994; Rees et al., 1996). Bernstein (1993) argues that:
Cash ?owfromoperations, as a measure of performance, is less subject to distortion than is the
net income ?gure. This is so because the accrual system, which produces the income number,
relies on accruals, deferrals, allocations and valuation, all of which involve higher degrees of
subjectivity than what enters the determination of CFO.
His explanation highlights why academic researchers focus on discretionary accruals to
detect all kinds of earnings management and opportunistic behaviors of managers.
Market ef?ciency is also measured by how the information in discretionary accruals is
contained in investors’ decision-making process (Sloan, 1996).
Managers may be motivated to manage their ?rm’s earnings as a consequence of
self interest (e.g. compensation, stock awards or pension contributions) especially if the
remunerative event depends on observable performance measures such as earnings or
pro?tability (Fields et al., 2001). Earnings manipulations affect the quality of earnings
by distorting the portion of earnings that are a result of corporate success and the
portion resulting from managerial discretion. Not surprisingly, investors respond
appropriately if they believe earnings numbers are distorted (Dechow and Schrand,
2004). On the other hand, successful detection of earnings management is problematic
due to the surfeit of variables that in?uence the managers’ decision set about the ?rm.
Greater recent attention on discretionary current accruals (DCA) is a consequence of
major accounting frauds (such as Enron) and in response to mounting evidence of
corporate earnings management (Clarke, 2005). It is doubtful that all companies use
DCA to in?uence their earnings level. We speculate that the greatest incidence of DCA
occurs when companies’ ?nancials are under the closest scrutiny[1]. Such times include
when credit lines are being reviewed by lenders, unexpected shocks hit either the
company or its industry, or the company’s operating and ?nancial results are under
stress. This research focuses on this later case. That is, we study how a company’s
?nancial condition (as measured by the Altman Z-score) in?uences its use of DCA by
taking into consideration their corporate governance quality[2].
Financial status of the ?rm is important for both creditors and investors. The
literature also shows that corporate investment decisions are related to ?nancial
factors (Cleary, 1999). Firm’s investment strategies are closely related to their accruals
(Ohlson, 1995; Fair?eld et al., 2003). Basically, investment strategies of ?rms are
sensitive to the availability of both internal and external funds and as a consequence
accruals play a signi?cant role in their decisions.
External sources of ?nancing may come with agency costs. Therefore, ?rms may
prefer to ?nance their investments with internal sources of ?nancing to avoid this
additional agency cost. Following the above argument, Bernanke and Gertler (1990),
Gertler (1992) and Fazzari et al. (1988) classify ?rms according to their ?nancial
constraints by using dividend payout, size, age, group memberships or debt ratings,
and show that investment decisions of ?rms that are ?nancially constrained are more
sensitive to ?rm liquidity than otherwise. Kaplan and Zingales (1997) found the
opposite result that investment decisions of the least ?nancially constrained ?rms
are less sensitive to the availability of cash ?ows. Cleary (1999), like Kaplan and
Zingales’ (1997), classi?es ?rms according to their ?nancial variables which is similar
ARJ
22,2
94
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to how the Altman Z predicts bankruptcy[3]. Cleary (1999) notes that corporate
investment (real activity decision) is sensitive to ?rm liquidity and that the relationship
is more severe among less creditworthy ?rms. We document a strong relationship
between a ?rms’ ?nancial classi?cation (according to the Altman Z-score which
includes liquidity ratios) and its use of performance adjusted DCA (PADCA)[4] which
is ?nancial accounting reporting choice of managers, see the Appendix for PADCA’s
calculation. This relationship is most sensitive for ?rms that are in the mid-range
between healthy and distressed classi?cations.
This study contributes to both the accounting and the ?nance literature in various
ways. First of all, we document the relationship between ?rms’ ?nancial health and
their discretionary accruals. This result is important because it shows the existence of
agency problems in credit markets if ?nancial statements signal that the ?rm is
?nancially stressed. Second, we document the importance of paying more attention to
?rms whose ?nancial condition is uncertain since they appear to be more prone to
accrual manipulations. This is especially important for the non-executive members of
boards of directors of the ?rm concerned with governance issues in the Sarbanes-Oxley
constrained environment. Finally, we document signi?cant ?ndings of importance to
the earnings quality and corporate governance literature by documenting the role of
corporate governance on PADCAs and ?nancial status.
The remainder of the paper is organized as follows. Section 2 discusses the
motivation behind the study, Section 3 reviews existing literature, Section 4 develops
hypotheses, Section 5 presents research design, Section 6 documents empirical results
and the conclusions are offered in Section 7.
2. Motivation
A number of motivations for managers to use discretion in accounting have been
proposed in the literature including hitting targets to affect employee bonuses and debt
covenants as well as motivations affecting stakeholders and stock price by achieving
targeted earnings goals (Healy and Wahlen, 1999; Dechow and Skinner, 2000;
Fields et al., 2001). In addition, managers may want to increase their reputation with
stakeholders, such as customers, suppliers, and creditors, to gain bargaining power
and thereby in?uence their terms of trade (Bowen et al., 1995; Burgstahler and Dichev,
1997). Companies reporting better results, even if they have been managed, are likely to
be rewarded with advantageous credit terms relative to similar companies not
managing their accounts. Executive pay may be related to some preset earnings
number pushing managers to use accounting discretion to realize earnings goals and
thereby increase their personal wealth and continue their tenure at the ?rm (Healy,
1985; Matsunaga and Park, 2001; Farrell and Whidbee, 2003; Francis et al., 2008).
Managers may also manage earnings to avoid violating debt covenants that would
increase the cost of capital for the ?rm (Watts and Zimmerman, 1990).
Our study contributes to the stream of research that asks “why do managers use
accounting discretion?” We incorporate two factors that previously were separately
tested as determinants of discretionary accruals: the company’s ?nancial status and
corporate governance. A company’s ?nancial health is related to the other factors
discussed above that explain discretionary accruals including impacts on executives’
careers, corporate reputation, stock price, and debt covenants. Our study also makes
a contribution to the literature regarding classi?cation of ?rms according to their
Discretionary
accruals
95
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
?nancial health by considering the role that discretionary accruals play in that
matter[5].
Graham et al. (2005) report that companies are more likely to take real actions such
as asset sales or reducing investments to facilitate the meeting of earnings’ targets.
They note that “accounting actions to meet earnings benchmarks get notably little
support” (p. 36), from their survey respondents, 401 ?nancial executives. It is precisely
this negation of the admission that discretionary accruals are used by ?rms that leads
us to investigate their use by ?nancially stressed ?rms. That is, we agree that it is
likely that normal or healthy companies are averse to using discretionary accruals;
however, it is companies that either are falling from a state of health or which are in the
process of working their way out of distress that we believe are likely candidates to be
users of discretionary accruals.
3. Literature review
Yeh and Lee (2004) ?nd that weak corporate governance increases the probability of a
?rm being classi?ed as ?nancially distressed, but not vice versa. Observant and
involved members of a board of directors provide a forumthat promotes better managed
?rms. They also note that healthy ?rms are more likely to have strong corporate
governance than weaker ?rms.
Cook et al. (2008) examine the effect of tax planning on the relationship between
effective tax rates (ETRs) and earnings management that is documented byDhaliwal et al.
(2004). Following that argument they also look at the effect of the Sarbanes-Oxley Act
(SOX) of 2002 onthe relationship between earnings management and ETR. They ?nd that
tax planning effects a ?rm’s ETRs and its earnings management. On the other hand, they
?nd no signi?cant result between SOX and earnings management.
Bowen et al. (2008) investigate whether managers’ use of accounting discretion for
opportunistic purposes or for shareholder wealth maximization, as ef?cient contracting
theory anticipates. They ?nd some evidence that managers use accounting discretion to
increase shareholder wealth. Unlike the conclusions of the prior literature, they also
document that poor governance is positively associated with future CFO and return on
assets (ROA). Therefore, they assert that earnings management may bene?t
shareholders by signaling future performance of the company. On the other hand,
their ?nding does not support the idea that managers exploit lax governance structures
and engage in accountingdiscretion at the shareholder’ expense. Similar to Bowen’s et al.
(2008) argument in our case, managers may use discretionary accruals to signal that a
?rm deserves a healthy classi?cation if it would be classi?ed otherwise. On the other
hand, opportunistic usage of discretionary accruals may increase the manager’s own
wealth and entrench their position at their company by avoiding disclosure of ?nancial
distress or bankruptcy. Corporate governance quality plays a signi?cant role in the
decision to exercise discretionary accrual in ?nancial reports.
Jones et al. (2008) investigate the relationship between restatements and several
models of discretionary accruals. They document that accrual estimation errors are
correlated with restatements. Their ?nding suggests the need for research on whether
?rms use discretionary accruals to avoid being classi?ed as distressed ?rms. Both
restatements and distressed classi?cations lead investors to discount the value of ?rms
signi?cantly. Later after a ?rm’s health has organically improved, they may correct
?nancial statement manipulation through restatements.
ARJ
22,2
96
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Holthausen et al. (1995) document that managers manipulate earnings to maximize
their bonus plan payments. This ?nding supports Healy’s (1985) idea but did not
support his other idea that managers manipulate earnings downward (income
decreasing earnings manipulation) when earnings fall belowthe minimumnecessary to
receive bonuses. We support the income increasing argument and think that managers
may forsake accruals management when their ?rm is stressed because manipulation
then has no affect on their compensation.
Dhaliwal et al. (2005) examine the relationship between institutional ownership,
?nancial health and market valuation weights on earnings and book value of equity. They
document that ?rms that have high-institutional ownership are ?nancially healthier. Their
?nding is supportive of prior studies that show institutional investors playing a ?duciary
and positive governance role (Shleifer and Vishny, 1997; Bushee, 1998). Prior studies also
suggest that institutional investors prefer to invest in ?nancially healthy ?rms (Hessel and
Norman, 1992; Del Guercio, 1996). Managers may seek to keep institutional investors as
owners in order to bene?t fromtheir ?nancial power but also to achieve stock appreciation
whichmayincrease the value of managers’ bonuses. Our studycontributes tothis streamof
research by showing that managers use accounting discretion to make mid-range
companies appear healthier and thereby become the targets of institutional investors in the
future, which may affect their corporate governance quality as well.
Ali et al. (2007), Mather and Ramsay (2006) and Hsu and Koh (2005) examined the
impact of corporate governance on discretionary accruals. Ali et al. (2007) ?nd better
quality earnings reports among family ?rms who they ?nd are also more reticent about
their corporate governance practices. Mather and Ramsay (2006) consider how
characteristics of strong corporate boards are associated with the degree of earnings
management at the time of chief executive of?cer (CEO) transition. Notably, they uncover
evidence in cases where the CEOresigned of negative unexpected accruals. This is limited
by larger boards and boards with a higher proportion of independent directors. When the
CEOretires they ?nd that discretionary accruals are somewhat controlled when there is a
higher proportion of executive and af?liated director shareholdings. Hsu and Koh (2005)
consider the corporate governance role of institutional investors. They ?nd upward
accruals management at companies whose shares are owned by institutional investors
that sell out as opposed to take long-term positions. In contrast, institutions with a
long-term hold strategy constrain upward accruals management even among companies
that have strong incentives for such behavior.
4. Hypothesis development
4.1 Altman Z-score
There are many studies in the ?nance and accounting literature that investigate how to
classify and to measure ?rms according to their ?nancial health. Altman’s (1968) initial
study developed a measure to estimate bankruptcy risk based on accounting ratios.
Follow up literature to improve the ef?ciency of Altman’s approach include Merton
(1974), Shumway (2001), Zmijewski (1984) and Platt and Platt (1991). These studies
generally conclude that bankruptcy is determined by:
.
?rm liquidity, pro?tability and leverage;
.
market variables; and
.
macroeconomic or business cycle variables.
Discretionary
accruals
97
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
In this study, we use the Altman Z-score to classify ?rms according to their ?nancial
health.
4.2 Discretionary accruals
Investors respond negatively to information indicating that a ?rmis distressed (Chi and
Tang, 2007). This is called the distressed ?rm discount in the academic literature. In
order to avoid this discount executives of a ?rm may use discretionary accruals to
manipulate their disclosures, e.g. earnings management (Kasznik, 1999). Accrual
management is one of the ways managers obtain desired earnings numbers (Dechow
and Schrand, 2004). Accruals are open to manipulation because they require the
manager to forecast and to make judgments. Consequently, accruals are subject to all
kinds of management discretion[6]. Of course, accrual management changes only the
timing of the recognition of earnings. That is, managers shift earnings between
quarters but do not create real earnings.
The earnings management literature generally reports that high-accrual companies
have the potential to manage their earnings (Dechowet al., 1996; Richardson et al., 2002).
These studies have bifurcated accruals into those that are nondiscretionary and those
that are discretionary in order to detect earnings management by managers. Jones (1991)
employed what is now a commonly used method to determine discretionary accruals.
The main criticism of her model is that it does not control for performance and growth.
Subsequent developments in the literature have attempted to improve the power of this
model to estimate the amount of nondiscretionary accruals (Dechow et al., 1996;
Kothari et al., 2005). In this study, we follow the method developed by Kothari et al.
(2005). Their discretionary accrual estimation method is an extension of Dechow et al.
(1996) but they go one step further and adjust discretionary accruals with the
performance of the ?rm which is measured by lagged ROA.
Firms wishing to appear healthier than their true condition use discretionary
accruals in order to avoid being identi?ed as in distress and thereby avoid the ?nancial
market discount. Corporate governance of the ?rms in?uences their ?nancial reporting
decisions (Bowen et al., 2008). By following above argument, our main hypotheses
concern the timing of when managers engage in discretionary accruals. Speci?cally, we
test the following hypotheses in alternative format:
H1. Discretionary accruals of ?rms are related to their ?nancial health and
corporate governance quality.
Firm’s ?nancial classi?cations are calculated using the Altman Z-score which depends
upon numbers that come from ?nancial statements. Therefore, the Z-score itself may
be affected by those accounts. Moreover, corporate governance quality may affect
corporate ?nancial reporting decisions by inducing strategic management decisions.
This leads to our second hypothesis becomes:
H2. Financial classi?cations of ?rms are related to discretionary accruals and
corporate governance quality.
A ?rm’s governance mechanism is mainly affected by its economic environment,
shareholders and creditors. Therefore, a ?rm’s ?nancial status may in?uence
management’s strategic decisions. While the effect of discretionary accruals on
corporate governance is less clear, we include that possibility in our H3:
ARJ
22,2
98
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
H3. Corporate governance quality is related to ?nancial health and discretionary
accruals.
5. Research design, measurements and sample
5.1 Altman Z-score
Altman (1968) used multiple discriminate analyses to determine the ?nancial conditions
of ?rms. From a list of 22 ?nancial ratios, he put ratios in ?ve categories including
liquidity, pro?tability, leverage, solvency and activity to discriminate between healthy
and to-be-bankrupt ?rms. Altman selected the ratios according to their popularity in the
literature rather than a theoretical basis resulting in a linear combination of variables
that discriminated between bankrupt and non-bankrupt ?rms. The linear equation is as
follows:
Z ¼ 0:012X
1
þ 0:014X
2
þ 0:033X
3
þ 0:006X
4
þ 0:999X
5
ð1Þ
where X
1
– working capital/total assets; X
2
– retained earnings/total assets; X
3
–
earnings before interest and taxes/total assets; X
4
– market value of equity/total
liabilities; X
5
– sales/total assets; Z – overall index, the lower a company’s Z-score the
higher its probability to bankrupt.
We use the Altman Z-score to measure the ?nancial health of companies. Later
research con?rms that the Altman Z-score works better for manufacturing ?rms (Grice
and Ingram, 2001). Therefore, we con?ned our study to manufacturing ?rms[7].
Following Altman’s (1968) study we classi?ed ?rms into three groups according to
their Z-score level. If Z-score exceeds 2.67 then the ?rm is called ?nancially healthy,
if the Z-score is between 1.81 and 2.67, ?rms are called gray-area companies, and ?nally
if the Z-score is smaller than 1.81, ?rms are classi?ed as ?nancially distress.
5.2 Discretionary accruals
Examples of items that may be subject to discretionary include accounts payable,
account receivable, future tax liability, depreciation, and future interest expense among
others. SFAS No. 95 requires companies to disclose information necessary to compute
accrual components of earnings in the operating section of cash ?ows. Accrual
components of earnings are computed by following prior literature (Dechowet al., 1995):
Total current accruals ¼ ðDCA 2DCashÞ 2ðDCL 2DSTDÞ ð2Þ
where DCA – change in current assets (Compustat item no. 4); DCash – change in
cash/cash equivalents (Compustat item no. 1); DCL – change in current liabilities
(Compustat itemno. 5); DSTD – change in debt includedin current liabilities (Compustat
item no. 34).
As mentioned above, managers may in?uence accruals to get their desired earnings
numbers. According to this hypothesis, companies do not change their activities rather
they report opportunistic (manipulated) earning numbers. For example, companies
may be avoiding write-offs, or capitalizing some expenses. The idea of accruals being
discretionary means that they require managers to forecast, estimate and make
judgments (Dechow and Schrand, 2004). The discretionary accrual measure is obtained
by subtraction of normal accruals from estimated total accruals thereby providing an
estimate of discretionary accruals that a ?rm might take in a year (Kothari et al., 2005).
Discretionary
accruals
99
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A higher estimate of discretionary accruals may signal a greater level of earnings
management (Dechow and Schrand, 2004)[8]. Lower discretionary accruals estimate
may signal higher earnings quality.
To estimate abnormal discretionary accruals, we follow Kothari et al.’s (2005) and
calculate PADCA. First, we estimate the modi?ed Jones model separately for each year
for each two-digit Standard Industrial Classi?cation (SIC) code (Dechow et al., 1995).
Then, PADCA is computed as the difference between the abnormal accrual and the
closest matched ?rm’s abnormal accrual, where the closest matched ?rm is the ?rm in
the same two-digit SIC code with the closest ROA in the prior year[9].
To test whether a ?rm’s health in?uences its use of PADCA, we examine whether
the amount of PADCA is related to its Altman’s Z-score. There is a possibility that the
Altman Z-score itself, which relies on ?nancial data that may be altered by the ?rm’s
use of PADCA, may be affected by PADCA. That is, there is a possibility that the
estimated level of Altman’s Z itself is related to the ?rm’s use of PADCA. Healthy ?rms
may have stronger governance because of their ability to hire better skilled executives
than distressed ?rms. Therefore, the corporate governance mechanism may in?uence
?nancial status classi?cation. This relationship may hold for discretionary accruals as
well. Moreover, corporate governance quality may affect the discretionary accruals,
and vice versa. This may lead to three-way endogeneity in our model between
corporate governance, ?nancial health and discretionary accruals. We thus consider a
simultaneous relationship model with PADCA, Altman’s Z and corporate governance
quality score (GINDEX). To test this question, we use a three-stage least squares
(3SLS) model similar to Al-Tuwaijri et al. (2004)[10].
The basic three equation model is shown below:
PADCA ¼ f ðY
1
; Altman Z 2score; GINDEXÞ ð3Þ
Altman Z ¼ gðY
2
; PADCA; GINDEXÞ ð4Þ
GINDEX ¼ gðY
3
; PADCA; Altman Z 2scoreÞ ð5Þ
where Y
1
, Y
2
and Y
3
are vectors of additional variables that in?uence PADCA, Altman’s
Z-score, and GINDEX, respectively, where Altman Z is a proxy for corporate health and
GINDEX is an indicator for the corporate governance quality. The incidence of PADCA
is assumed to be related to Y
1
(a vector of determinants), Altman Z, and GINDEX. We
believe that companies engage in PADCA as their health deteriorates and no
pre-assumption for corporate governance. Moreover, we believe that a company’s health
(measured by Altman’s Z) is affected by Y
2
(a vector of determinants), by PADCA, and
by GINDEX. That is, a company affects its measured ?nancial health (via Altman’s
Z-score) when it engages in the use of PADCA, and having good corporate governance.
Finally, corporate governance quality (measured by GINDEX) is in?uenced by Y
3
(a vector of determinants), PADCA and Altman Z.
The simultaneous relationship between PADCA, Altman Z and GINDEX creates an
identi?cation problem which leads to biased and inconsistent parameter estimates if
the model is estimated using ordinary least squares. To avoid this problem, a 3SLS
regression is performed[11]. Using some of the control variables similar to Ali et al.
(2007) and including more, a 3SLS model for the years 2001-2003 is run to examine the
relationship between discretionary accruals and the ?rms’ ?nancial health[12]. We also
include dummy variables for each year and two-digit industry classi?cations[13]:
ARJ
22,2
100
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
PADCA ¼ a þb
1
Altman 2Z þb
2
GINDEX þb
3
HEALTH þb
4
AltmanZ
£ HEALTH þb
5
DISTRESS þb
6
Altman 2Z £ DISTRESS
þb
7
LagPADCA þb
8
LEVERAGE þb
9
SIZE þb
10
LOSS þb
11
ROA
þb
12
CFO þ error
ð6Þ
Altman 2Z ¼ a þb
1
PADCA þb
2
GINDEX þb
3
SIZE þb
4
LOSS þb
5
ROA
þb
6
CFA þb
7
LagAltman 2Z þb
8
EPS þb
9
MB
þb
10
LITIGATION þb
11
R&D þb
12
DEPRECIATION
þb
13
PP&E þb
14
INTANGIBLE þb
15
SPECIAL
þb
16
INTEREST þ error
ð7Þ
GINDEX ¼ a þb
1
Altman 2Z þb
2
PADCA þb
3
SIZE þb
4
LITIGATION
þb
5
MERGER þb
6
PREFERRED þb
7
DIVIDEND þ error
ð8Þ
where PADCA (see the Appendix for the calculation) is performance adjusted
discretionary accruals that is calculated by following Kothari et al. (2005). PADCA is
scaled by total assets[14]. Altman Z is the Z-score developed by Altman (1968) to
estimate the ?nancial condition of ?rms. The variables in equation (7) that explain the
level of Altman Z beyond PADCA and GINDEX re?ect hypotheses about ?rm
characteristics and Altman Z. SIZE, R&D, MB, ROA and EPS are expected to control
for the performance and information environment of the ?rms therefore we expect they
have positive coef?cients as higher values for these variables contribute to healthier
companies. In contrast, LOSS, LITIGATION, CFO, DEPRECIATION, PP&E,
INTANGIBLE, SPECIAL and INTEREST, capture the operational and ?nancial
riskiness of the ?rm, are expected to have negative coef?cients since higher values of
these variables reduce the health of ?rms. MERGER and LEVERAGE are used
speci?cally to capture agency problem which are interpreted as the higher the values
for these variables the higher the agency costs. All of the variables have expected signs
except for ROA.
GINDEX is a corporate governance index that is developed by Gompers et al. (2003)
to measure the governance quality of the ?rm where GINDEX score is extracted from
the IRRC database. HEALTH is an indicator variable that takes the value of one if the
Z-score exceeds the value of 2.67, and zero otherwise. DISTRESS is also an indicator
variable that takes the value of one if Z-score is smaller than 1.81, and zero
otherwise[15]. LagPADCA is equivalent to last year’s PADCA scaled by total
assets[16]. LEVERAGE is the ratio of total debt to total assets at the beginning of the
?scal period (total assets minus its book value divided by its total assets) scaled by
total assets. SIZE is the natural logarithm of total assets. MERGER (MA) is a dummy
variable that takes the value of one if the ?rm has engaged in a merger and/or
acquisition activity during the current ?scal year, zero otherwise. LOSS is an indicator
Discretionary
accruals
101
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
variable that takes the value of one if the ?rm reports a net loss for the ?scal period,
and zero otherwise. ROA is the current year’s return on asset calculated as net income
before extraordinary items divided by total assets scaled by total assets. CFO is scaled
by the beginning of the year’s total assets scaled by total assets. EPS is earnings
per share which is calculated by dividing net income before interest and taxes by
numbers of shares outstanding scaled by total assets. MB is the market-to-book ratio
that is calculated by using the market value of equity divided by the book value of
equity scaled by total assets. LITIGATION is a dummy variable that equals one if the
?rm operates in the high litigation industries with the SIC codes of 2833-2836,
3570-3577, 3600-3674, 5200-5961 and 7370-730, zero otherwise. R&D is research and
development expense scaled by total assets. DEPRECIATION is depreciation expense
in the current year scaled by total assets. PP&E is total net plant, property and
equipment that is reported on the ?rm’s balance sheet scaled by total assets.
INTANGIBLE is total intangible assets scaled by total assets. SPECIAL is total special
items scaled by total assets. INTEREST is total interest expense within a year reported
on a ?rm’s income statement scaled by total assets. PREFERRED is total preferred
stock holders equity that is reported under the owners’ equity section of the balance
sheet in the ?rm’s annual report scaled by total assets. DIVIDEND is total dividend
issued and distributed by the ?rm in the current year scaled by total assets[17].
5.3 Sample
We use US ?rm data, over the period 2001-2003. This period is chosen because it
includes a large downturn in the stock market which may be an environment more
susceptible to PADCA activity[18]. Only, manufacturing ?rms are analyzed to conform
to the Altman Z-score original design and due to its greater ef?cacy for these ?rms
(Altman, 1968; Grice and Ingram, 2001). Specially, we collected all the data from the
Compustat database between the years 2001 and 2003 for manufacturing ?rms.
Manufacturing ?rms have the SIC code between 2000 and 4000. From 67,899 ?rm-year
observations originally gathered, the sample decreases to 8,655 ?rm-year observations
after calculating PADCA, Altman Z-score, and other variables in the two equation
model. We used Gompers et al. (2003) governance index (GINDEX) to measure the
governance strength of the ?rms in our sample.
6. Empirical results and analysis
Table I, Panel A provides descriptive statistics of the dependent variables. The mean
(median) Altman Z for sample ?rms is 2.8374 (2.6980). The mean (median) performance
adjusted discretionary accruals (PADCA) 0.0006 (20.0005) indicates on average
sample ?rms have income increasing accruals. GINDEX 9.1836 (9.000) suggests that on
average sample ?rms have good corporate governance quality. Panel B of Table I
presents the descriptive statistics for independent variables. Leverage with 0.6659
(0.3996) is positively skewed that indicates that sample ?rms have on average
66.8 per cent total liabilities on their balance sheets. Average size which is measured
with market value of equity in our sample is approximately $126 million. The merger
variable shows that, on average, 27 per cent of companies engage in M&A. In our
sample, 49 per cent of companies report losses which may be because our sample
consist of only manufacturing ?rms. ROA average is 217.22 per cent and its
median is 0.2 per cent in part because our sample period is between 2001 and 2003.
ARJ
22,2
102
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
CFO is negative (a value of 8.19 per cent) with a positive median value which indicates
skewness to the left. EPS share is also negative again because our sample period spans
an economic downturn. Market to book ratios show that on average sample ?rms are
established, and conservative in their reporting decisions. About 35 per cent of ?rms
are in litigated industries as de?ned earlier. On average ?rms spend 10 per cent of their
Variables Mean SD 25th percentile Median 75th percentile
Panel A: dependent variables
Altman Z 2.8374 14.6280 1.0901 2.6980 5.1304
PADCA 0.0006 0.1779 20.0832 20.0005 0.0806
GINDEX 9.1836 2.6813 7.0000 9.0000 11.000
Panel B: predetermined variables
LEVERAGE 0.6659 6.4737 0.2128 0.3996 0.5877
SIZE 4.8391 2.4434 3.1286 4.7694 6.5128
MERGER 0.2659 0.4419 0.0000 0.0000 1.0000
LOSS 0.4934 0.5000 0.0000 0.0000 1.0000
ROA 20.1722 0.5052 20.2079 0.0020 0.0638
CFO 20.0819 0.4257 20.0893 0.0454 0.1105
EPS 20.0189 0.0655 20.0130 0.0000 0.0022
MB 20.0480 1.8147 0.0010 0.0075 0.0443
LITIGATION 0.3546 0.4784 0.0000 0.0000 1.0000
R&D 0.1059 0.5052 0.0000 0.0322 0.1171
DEPRECIATION 0.0527 0.0445 0.0274 0.0429 0.0626
PP&E 0.4903 0.3601 0.2130 0.4111 0.6871
INTANGIBLE 0.1100 0.1537 0.0000 0.0344 0.1671
SPECIAL 20.0486 0.1604 20.0249 0.0000 0.0000
INTEREST 0.0294 0.0643 0.0017 0.0130 0.0297
PREFERRED 0.0447 0.2191 0.0000 0.0000 0.0000
DIVIDEND 0.0003 0.0011 0.0000 0.0000 0.0000
Notes: All variables are winsorized at 1 per cent level; reported values of variables are for a ?rm at current
year if it is not stated otherwise; variable de?nitions: Altman Z is ?nancial indicator score that was developed
by Altman (1968). Detailed calculation for Altman Z is provided in the text; PADCA is performance adjusted
discretionary accruals that is calculated by following Kothari et al. (2005) which is scaled by total assets;
GINDEX is governance index that is developed by Gompers et al. (2003). This variable is extracted from
Thomson IRRC governance database; LEVERAGE is the ratio of total debt to total assets at the beginning of
the ?scal period scaled by total assets; SIZE is the natural logarithm of the total assets; MERGER (MA) is a
dummy variable that takes the value of 1 if the ?rm has engaged in a merger and/or acquisition activity, 0
otherwise. LOSS is an indicator variable that takes the value of 1 if the ?rm reports a net loss for the ?scal
period and 0 otherwise, i.e. negative net income before extraordinary items; ROA is the current year’s return
on assets calculated as net income before extraordinary items divided by total assets scaled by total assets;
CFO is cash ?ow from operations scaled by total assets; EPS is earnings per share which is calculated as
dividing net income before interest and taxes by numbers of shares outstanding corresponding year scaled
by total assets; MB is the market-to-book ratio that is calculated by using the market value of equity divided
by the book value of equity scaled by total assets; LITIGATION is a dummy variable that equals one if the
?rm operates in the high litigation industries with the SIC codes of 2833-2836, 3570-3577, 3600-3674, 5200-
5961, and 7370-730, 0 otherwise; R&D is research and development expense scaled by total assets;
DEPRECIATION is depreciation expense scaled by total assets; PP&E is total net plant, property and
equipment that is reported at ?rm’s balance sheet scaled by total assets; INTANGIBLE is total intangible
assets scaled by total assets; SPECIAL is total special items reported in income statement scaled by total
assets; INTEREST is total interest expense within a year that is reported at ?rm’s income statement scaled
by total assets; PREFERRED is total preferred stock holders equity that is reported under the owners’ equity
section of balance sheet at ?rm’s annual report scaled by total assets; DIVIDEND is total dividend issued and
distributed scaled by total assets
Table I.
Descriptive statistics
of 8,655 numbers
of observations except
GINDEX variable has
1,340 observations
Discretionary
accruals
103
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
total assets on research and development investments. Median value for this variable is
3.22 per cent. Depreciation expense is 5.27 per cent of total assets. About 49.03 per cent
of assets are PP&Es which is not surprising given the heavy investment of
manufacturing ?rms in machineries, buildings, etc. On the other hand, intangible
assets are 11 per cent of total assets and could be generated by research and
development investments or patents. Special items which are usually one-off items
have mean and median values of 24.86 and 0.00 per cent, respectively. Total interest
expense is on average 2.94 per cent with an average rate of 5 per cent (2.94/66.59).
Preferred stock ownership percentage is 4.47 per cent on average with a median value
of 0.00 per cent. Finally, sample ?rms distribute few dividends with the average size of
0.3 per cent. This may be because half of our sample ?rms report loss.
Table II presents descriptive statistics for healthy, distressed and gray area ?rms.
The average Altman Z-scores for healthy ?rms is 9.4361, for distressed ?rm is
26.4160, and for gray area ?rms is 2.2336. Differences between the means are
signi?cant at the 1 per cent level according to the mean difference t-test. The PADCA
score reports a different story than the Altman Z-score using the mean comparison test.
The mean PADCA score for healthy ?rms is 0.0057, for distress ?rms is 20.0088, and
for gray area ?rms is 0.0057. Both healthy and gray area ?rms have statistically
signi?cant mean PADCA differences compared with distressed ?rms. On the other
hand, healthy ?rms and gray ?rms’ average PADCA are not statistically different.
After determining the Altman Z-score and PADCA for ?rms in our sample, we had
three sets of companies: those in the healthy area (787 ?rms), those in the mid-range or
gray area (262 ?rms), and those in the distress region (291 ?rms). We calculated the
value of the governance index (GINDEX) for the healthy, gray, and distressed ?rm
groups. We found mean values of 9.531, 9.137 and 8.997, respectively, as seen in
Table II. The gray area has a mean according to t-tests that is signi?cantly different at
the 5 per cent level from the mean values of the healthy ?rms but is not different from
distressed area ?rms at conventional levels. Distressed ?rms have smaller GINDEX
values than healthy ?rms and their mean GINDEX value is signi?cantly different at
the 5 per cent level from that of healthy ?rms. In other words, there appears to be less
corporate governance quality when ?rms depart from being healthy and are not clearly
identi?ed as distressed. That is, gray area companies have less governance quality
applied to them than their counterparts’ healthy ?rms. Mean difference tests document
that healthy ?rms’ corporate governance quality is better than both gray and
distressed ?rms[19]. Later, we introduce the governance index into our regression
model to see how well it explains the level of PADCA.
There are fewer discretionary accruals among distressed companies; healthy and
gray area companies have more discretionary accruals than distressed companies but
about the same level of discretionary accruals between them. Managers of these ?rms
may feel that it is in their best interest to produce better results this quarter/year than
the ?rm justi?ably deserves. Some healthy and gray area companies appear to
creatively use accruals to bolster their ?nancial results. Some healthy ?rms may be
using PADCA to achieve a healthy classi?cation. Similarly, some gray area companies
may use PADCA to avoid being classi?ed as distressed. In fact, the lower proportion of
PADCA among distressed companies may result from those ?rms having fewer
remaining PADCA opportunities. Another reason for higher discretionary accruals
among gray area ?rms may be noise in their accounting information, i.e. information
ARJ
22,2
104
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
M
e
a
n
M
e
a
n
d
i
f
f
e
r
e
n
c
e
ª¯
V
a
r
i
a
b
l
e
s
H
e
a
l
t
h
y
D
i
s
t
r
e
s
s
G
r
a
y
H
e
a
l
t
h
y
-
d
i
s
t
r
e
s
s
H
e
a
l
t
h
y
-
g
r
a
y
G
r
a
y
-
d
i
s
t
r
e
s
s
A
l
t
m
a
n
Z
9
.
4
3
6
1
2
6
.
4
1
6
0
2
.
2
3
3
6
1
5
.
8
5
2
1
*
*
*
7
.
2
0
2
5
*
*
*
8
.
6
4
9
6
*
*
*
P
A
D
C
A
(
%
)
0
.
0
0
5
7
2
0
.
0
0
8
8
0
.
0
0
5
7
0
.
0
1
4
5
*
*
*
0
.
0
0
0
0
0
.
0
1
4
5
*
*
G
I
N
D
E
X
9
.
5
3
1
8
.
9
9
7
9
.
1
3
7
0
.
5
3
4
*
*
0
.
3
9
4
*
*
0
.
1
4
0
L
a
g
P
A
D
C
A
(
%
)
0
.
0
0
5
4
2
0
.
0
0
5
8
0
.
0
1
1
0
0
.
0
0
1
2
*
*
2
0
.
0
0
5
6
0
.
0
1
6
8
*
*
*
L
E
V
E
R
A
G
E
0
.
2
9
1
0
1
.
2
7
5
9
0
.
4
9
5
0
2
0
.
9
8
4
9
*
*
*
2
0
.
2
0
4
0
*
*
*
2
0
.
7
8
0
9
*
*
S
I
Z
E
5
.
1
2
2
5
3
.
9
5
7
2
5
.
9
9
7
5
1
.
1
6
5
3
*
*
*
2
0
.
8
7
5
0
*
*
*
2
.
0
4
0
3
*
*
*
M
E
R
G
E
R
0
.
2
6
7
3
0
.
2
4
7
5
0
.
3
0
6
5
0
.
0
1
9
8
*
2
0
.
0
3
9
2
*
*
*
0
.
0
5
9
0
*
*
*
L
O
S
S
0
.
3
2
6
7
0
.
7
9
3
6
0
.
3
4
5
9
2
0
.
4
6
6
9
*
*
*
2
0
.
0
1
9
2
2
0
.
4
4
7
7
*
*
*
R
O
A
2
0
.
0
5
6
8
2
0
.
4
0
0
0
2
0
.
0
2
1
3
0
.
3
4
3
2
*
*
*
2
0
.
0
0
3
6
*
*
*
0
.
3
7
8
7
*
*
*
C
F
O
0
.
0
2
4
5
2
0
.
2
8
6
5
0
.
0
4
3
5
0
.
3
1
1
1
0
*
*
*
2
0
.
0
1
9
0
*
*
0
.
3
3
0
0
*
*
*
L
a
g
A
l
t
m
a
n
Z
1
3
.
1
2
3
9
2
1
.
6
0
9
0
3
.
0
8
8
9
1
4
.
7
3
2
9
*
*
*
1
0
.
0
3
5
0
*
*
*
4
.
6
9
7
9
*
*
*
E
P
S
2
0
.
0
0
2
1
2
0
.
0
4
9
1
2
0
.
0
0
5
0
0
.
0
4
7
0
*
*
*
0
.
0
0
2
9
*
*
*
0
.
0
4
4
1
*
*
*
M
B
0
.
1
6
3
9
2
0
.
3
8
1
1
0
.
0
2
0
0
0
.
5
4
5
0
*
*
*
0
.
1
4
3
9
*
*
*
0
.
4
0
1
1
*
*
*
L
I
T
I
G
A
T
I
O
N
0
.
3
8
9
4
0
.
3
7
4
3
0
.
1
8
4
2
0
.
0
1
5
1
0
.
2
0
5
2
*
*
*
2
0
.
1
9
0
1
*
*
*
R
&
D
0
.
0
8
6
1
0
.
1
6
3
5
0
.
0
3
5
6
2
0
.
0
7
7
4
*
*
*
0
.
0
5
0
5
*
*
*
2
0
.
1
2
7
9
*
*
*
D
E
P
R
E
C
I
A
T
I
O
N
0
.
0
4
2
5
0
.
0
6
9
8
0
.
0
4
6
9
2
0
.
0
2
7
3
*
*
*
2
0
.
0
0
4
4
*
*
*
2
0
.
0
2
2
9
*
*
*
P
P
&
E
0
.
4
1
7
6
0
.
5
6
8
9
0
.
5
5
4
1
2
0
.
1
5
1
3
*
*
*
2
0
.
1
3
6
5
*
*
*
2
0
.
0
1
4
8
I
N
T
A
N
G
I
B
L
E
0
.
0
9
3
3
0
.
1
1
6
8
0
.
1
4
9
2
2
0
.
0
2
3
5
*
*
*
2
0
.
0
5
5
9
*
*
*
0
.
0
3
2
4
*
*
*
S
P
E
C
I
A
L
2
0
.
0
2
7
4
2
0
.
0
8
5
2
2
0
.
0
3
3
7
0
.
0
5
7
8
*
*
*
0
.
0
0
6
3
*
0
.
0
5
1
5
*
*
*
I
N
T
E
R
E
S
T
0
.
0
1
0
5
0
.
0
5
8
3
0
.
0
2
5
6
2
0
.
0
4
7
8
*
*
*
2
0
.
0
1
5
1
*
*
*
2
0
.
0
3
2
7
*
*
*
P
R
E
F
E
R
R
E
D
0
.
0
1
0
0
0
.
1
0
9
3
0
.
0
0
8
9
2
0
.
0
9
9
3
*
*
*
0
.
0
0
1
1
2
0
.
1
0
0
4
*
*
*
D
I
V
I
D
E
N
D
0
.
0
0
0
5
0
.
0
0
0
0
0
.
0
0
0
2
0
.
0
0
0
5
*
*
*
0
.
0
0
0
3
*
*
*
0
.
0
0
0
2
*
*
*
N
o
.
o
f
o
b
s
e
r
v
a
t
i
o
n
s
4
,
3
7
4
3
,
0
3
8
1
,
2
4
3
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
0
.
1
0
,
*
*
0
.
0
5
a
n
d
*
*
*
0
.
0
1
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
ª¯
m
e
a
n
t
e
s
t
b
a
s
e
d
o
n
t
-
t
e
s
t
;
H
E
A
L
T
H
Y
?
r
m
s
a
r
e
t
h
e
o
n
e
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
e
x
c
e
e
d
s
t
h
e
v
a
l
u
e
o
f
2
.
6
7
;
D
I
S
T
R
E
S
S
?
r
m
s
a
r
e
t
h
e
o
n
e
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
i
s
s
m
a
l
l
e
r
t
h
a
n
1
.
8
1
;
a
l
l
o
t
h
e
r
?
r
m
s
a
r
e
c
l
a
s
s
i
?
e
d
a
s
G
R
A
Y
a
r
e
a
?
r
m
s
w
i
t
h
A
l
t
m
a
n
Z
-
s
c
o
r
e
s
m
a
l
l
e
r
t
h
a
n
2
.
6
7
a
n
d
l
a
r
g
e
r
t
h
a
n
1
.
8
1
;
n
u
m
b
e
r
o
f
o
b
s
e
r
v
a
t
i
o
n
s
f
o
r
G
I
N
D
E
X
f
o
r
h
e
a
l
t
h
y
,
d
i
s
t
r
e
s
s
a
n
d
g
r
a
y
a
r
e
7
8
7
,
2
6
2
a
n
d
2
9
1
,
r
e
s
p
e
c
t
i
v
e
l
y
;
v
a
r
i
a
b
l
e
s
a
r
e
d
e
?
n
e
d
i
n
T
a
b
l
e
I
;
L
a
g
P
A
D
C
A
(
A
l
t
m
a
n
Z
)
i
s
p
r
e
v
i
o
u
s
y
e
a
r
’
s
P
A
D
C
A
(
A
l
t
m
a
n
Z
)
Table II.
Comparative descriptive
statistics
Discretionary
accruals
105
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
environment. On the other hand, these ?rms are probably under the close scrutiny of
?nancial market participants such as auditors, banks, investors or banks and therefore
they may not have a tendency to manage their earnings (Bushee, 1998). In that case, the
larger discretionary accrual result may be the outcome of a noisy information
environment. There may be incentives for managers to use discretionary accruals
to retain their job and to get high compensation until regulators catch up with them.
On the other hand, if the market expects discretionary accruals, managers may be
compelled to manage earnings in order to reach equilibrium[20].
Almost all other ?rm characteristics for healthy ?rms are signi?cantly different
from distressed ?rms at the 1 per cent level (5 per cent for L1PADCA and 10 per cent
for MERGER dummy variables) with the exception of LITIGATION. Similar results
are noted between gray area and distressed ?rms. Between healthy and gray area ?rms
the other ?rm characteristic variables are signi?cantly different at 1 per cent level
(CFO at 5 per cent and SPECIAL at 10 per cent levels) except for the LagPADCA,
PREFERRED and LOSS. Overall, healthy, distressed and gray area ?rms have
different characteristics which need to be controlled for in our multivariate analysis.
Table III presents the correlation matrix of our dependent and independent
variables. As depicted in Table III PADCA and Altman Z-score are positively
correlated. On the other hand, Altman Z-score and GINDEX are negatively correlated
with p , 0.001. PADCA and governance score are not correlated signi?cantly at
conventional levels with the association being positive[21].
Table IV presents two-stage least squares (2SLS) regression results[22]. We used
2SLS regressions to resolve the identi?cation problem possibly caused by the Altman
Z-score and PADCA simultaneously in?uencing each other. As shown in Table III,
Altman Z-score (0.0035), and HEALTH (20.0187) of the ?rm (a health indicator
variable) in?uence PADCA signi?cantly at the 1 per cent level. These results indicate
that PADCA increases with an increase in the Altman Z-score, and decreases with ?rm
health. An interaction variable between Altman Z and HEALTH is statistically
signi?cant at the 1 per cent level (t ¼ 3.24) with a coef?cient value of 20.0018. This
indicates that being healthy decreases the effect of Altman Z-score on PADCA, almost
50 per cent from 0.0035 to 0.0017 (0.0035-0.0018). This result indicates that ?nancially
healthy ?rms use less PADCA type accounting discretion than other ?rms. The
coef?cient estimates for DISTRESS and the interaction variable Altman Z
£ DISTRESS are not signi?cant at conventional levels indicating that ?nancially
distressed ?rms are not engaged in manipulating their earnings. Perhaps, they fail to
perceive a bene?t from such manipulation[23].
The second regression reported in Table IV gives the results for the relationship
going from the level of PADCA to the level of Altman Z. That is, it examines whether a
?rm’s Altman Z is related to its use of PADCA. We are particularly concerned with the
coef?cient estimated on the PADCA variable. Notably, that coef?cient (144.944) is
signi?cant at the 0.01 level and suggests that ?rms have better (higher valued) Altman
Z-scores when they make more use of PADCA to in?uence their ?nancial data.
The implication of this ?nding is that analysts who believe that a company is not
distressed as a result of a score from the Altman Z model may be misled if the ?rm’s
use of PADCA has contributed to its perceived Altman Z-score value.
We report 3SLS regression results in Table V. This technique accounts for the
three way endogeneity problem identi?ed in equations (6)-(8). We discuss the three
ARJ
22,2
106
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
P
A
D
C
A
A
L
T
Z
G
I
N
D
E
X
L
E
V
S
I
Z
E
M
A
L
O
S
S
R
O
A
C
F
O
E
P
S
P
A
D
C
A
0
.
0
6
*
0
.
0
3
2
0
.
0
4
*
2
0
.
0
0
*
0
.
0
0
*
2
0
.
0
7
*
0
.
0
7
*
2
0
.
0
6
*
0
.
0
8
*
A
L
T
Z
0
.
0
4
*
2
0
.
0
9
*
2
0
.
1
9
*
0
.
2
9
*
0
.
0
1
2
0
.
1
7
*
0
.
3
5
*
0
.
5
2
*
0
.
4
6
*
G
I
N
D
E
X
0
.
0
2
2
0
.
0
7
*
0
.
0
9
*
0
.
2
1
*
0
.
1
1
*
2
0
.
2
0
*
0
.
0
7
*
0
.
0
7
*
0
.
0
4
L
E
V
2
0
.
0
6
*
2
0
.
6
5
*
0
.
1
6
*
2
0
.
0
8
*
2
0
.
0
0
*
0
.
0
4
*
2
0
.
0
9
*
2
0
.
1
7
*
2
0
.
1
8
*
S
I
Z
E
0
.
0
0
0
.
2
2
*
0
.
2
4
*
0
.
2
2
*
0
.
2
2
*
2
0
.
3
7
*
0
.
4
3
*
0
.
4
8
*
0
.
4
0
*
M
A
0
.
0
1
0
.
0
1
0
.
1
1
*
0
.
0
6
*
0
.
2
3
*
2
0
.
0
8
*
0
.
0
5
*
0
.
0
8
*
0
.
0
7
*
L
O
S
S
2
0
.
0
8
*
2
0
.
3
6
*
2
0
.
1
5
*
0
.
0
6
*
2
0
.
3
6
*
2
0
.
0
8
*
2
0
.
5
1
*
2
0
.
4
2
*
2
0
.
4
0
*
R
O
A
0
.
0
9
*
0
.
4
6
*
0
.
0
7
*
2
0
.
1
1
*
0
.
4
3
*
0
.
0
9
*
2
0
.
8
7
*
0
.
7
0
*
0
.
5
8
*
C
F
O
2
0
.
1
4
*
0
.
3
8
*
0
.
0
9
*
2
0
.
0
6
*
0
.
4
5
*
0
.
0
7
*
2
0
.
6
3
*
0
.
7
5
*
0
.
6
5
*
E
P
S
0
.
1
0
*
0
.
4
4
*
0
.
0
4
2
0
.
1
4
*
0
.
3
3
*
0
.
0
7
*
2
0
.
8
7
*
0
.
8
9
*
0
.
6
7
*
M
B
0
.
0
6
*
0
.
2
9
*
2
0
.
1
6
*
2
0
.
3
9
*
2
0
.
5
6
*
2
0
.
1
4
*
0
.
1
0
*
2
0
.
0
7
*
2
0
.
1
5
*
0
.
0
0
*
L
I
T
G
2
0
.
0
2
0
.
0
7
*
2
0
.
1
8
*
2
0
.
2
0
*
2
0
.
1
2
*
2
0
.
0
2
0
.
2
4
*
0
.
2
3
*
2
0
.
2
3
*
2
0
.
2
2
*
R
D
2
0
.
0
4
*
0
.
0
5
*
2
0
.
0
7
*
2
0
.
2
5
*
2
0
.
2
5
*
2
0
.
0
5
*
0
.
3
5
*
2
0
.
3
8
*
2
0
.
4
0
*
2
0
.
3
6
*
D
E
P
2
0
.
1
5
*
2
0
.
3
1
*
0
.
0
0
0
.
2
5
*
2
0
.
0
4
*
2
0
.
0
3
*
0
.
1
6
*
0
.
1
0
*
0
.
0
1
2
0
.
2
0
*
P
P
&
E
2
0
.
0
2
*
*
*
2
0
.
2
5
*
0
.
1
6
*
0
.
3
0
*
0
.
1
2
*
2
0
.
0
7
*
2
0
.
1
2
*
0
.
1
0
*
0
.
2
3
*
0
.
0
8
*
I
N
G
2
0
.
0
5
*
2
0
.
0
4
*
0
.
0
8
*
0
.
1
3
*
0
.
2
9
*
0
.
2
4
*
2
0
.
1
2
*
0
.
1
1
*
0
.
1
6
*
0
.
1
0
*
S
P
E
C
0
.
0
8
*
0
.
1
3
*
0
.
0
1
2
0
.
0
6
*
2
0
.
1
0
*
2
0
.
1
1
*
2
0
.
2
9
*
0
.
2
9
*
0
.
0
9
*
0
.
2
9
*
I
N
T
2
0
.
0
3
*
2
0
.
5
1
*
0
.
1
3
*
0
.
7
6
*
0
.
0
7
*
0
.
0
3
*
0
.
0
4
*
2
0
.
0
8
*
0
.
0
1
2
0
.
0
9
*
P
R
E
F
2
0
.
0
2
2
0
.
2
5
*
0
.
0
3
0
.
1
7
*
2
0
.
1
1
*
2
0
.
0
0
0
.
1
5
*
2
0
.
1
7
*
2
0
.
1
5
*
2
0
.
2
0
*
D
I
V
2
0
.
0
1
0
.
1
3
*
0
.
2
7
*
0
.
0
4
*
0
.
4
6
*
0
.
0
8
*
2
0
.
3
8
*
0
.
3
7
*
0
.
3
7
*
0
.
3
2
*
(
c
o
n
t
i
n
u
e
d
)
Table III.
Correlation coef?cients
for all variables included
in simultaneous equation
model
Discretionary
accruals
107
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
M
B
L
I
T
G
R
D
D
E
P
P
P
&
E
I
N
G
S
P
E
C
I
N
T
P
R
E
F
D
I
V
P
A
D
C
A
0
.
0
3
*
2
0
.
0
1
2
0
.
0
5
*
2
0
.
1
6
*
2
0
.
0
4
*
2
0
.
5
0
*
0
.
0
7
*
2
0
.
0
6
*
2
0
.
0
3
*
2
0
.
0
0
*
A
L
T
Z
0
.
3
8
*
0
.
0
7
*
2
0
.
2
6
*
2
0
.
4
1
*
2
0
.
2
7
*
2
0
.
0
3
*
0
.
1
9
*
2
0
.
5
1
*
2
0
.
3
0
*
0
.
0
4
*
G
I
N
D
E
X
0
.
0
1
2
0
.
1
7
*
2
0
.
1
2
*
2
0
.
0
5
*
*
*
0
.
1
4
*
0
.
0
5
*
*
*
0
.
0
1
0
.
0
1
2
0
.
0
6
*
*
2
0
.
0
3
L
E
V
2
0
.
1
2
*
2
0
.
0
1
0
.
0
3
*
0
.
0
7
*
0
.
0
3
*
0
.
0
1
2
0
.
0
2
*
*
*
0
.
1
9
*
0
.
1
6
*
2
0
.
0
1
S
I
Z
E
0
.
0
8
*
2
0
.
1
1
*
2
0
.
3
4
*
2
0
.
1
9
*
0
.
0
6
*
0
.
2
0
*
0
.
0
6
*
2
0
.
2
6
*
2
0
.
0
2
*
2
0
.
0
2
*
*
*
M
A
0
.
0
2
2
0
.
0
2
*
*
2
0
.
0
8
*
2
0
.
0
3
*
2
0
.
0
9
*
0
.
2
2
*
2
0
.
0
4
*
2
0
.
0
3
*
2
0
.
0
2
*
*
2
0
.
0
3
*
L
O
S
S
2
0
.
0
4
*
0
.
2
4
*
0
.
3
4
*
0
.
2
3
*
2
0
.
0
7
*
2
0
.
0
6
*
2
0
.
2
1
*
0
.
1
9
*
0
.
1
7
*
2
0
.
1
9
*
R
O
A
0
.
2
1
*
2
0
.
1
8
*
2
0
.
5
3
*
2
0
.
2
9
*
0
.
0
4
*
0
.
0
2
*
*
*
0
.
2
5
*
2
0
.
4
4
*
2
0
.
2
8
*
0
.
1
1
*
C
F
O
0
.
2
5
*
2
0
.
1
6
*
2
0
.
6
3
*
2
0
.
3
3
*
0
.
0
1
0
.
0
8
*
0
.
1
8
*
2
0
.
4
3
*
2
0
.
3
1
*
0
.
1
1
*
E
P
S
0
.
2
2
*
2
0
.
0
7
*
2
0
.
3
8
*
2
0
.
3
4
*
2
0
.
0
6
*
0
.
0
6
*
0
.
2
7
*
2
0
.
4
5
*
2
0
.
3
3
*
0
.
1
2
*
M
B
0
.
0
2
*
*
*
2
0
.
1
1
*
2
0
.
1
3
*
2
0
.
1
0
*
0
.
0
2
*
*
0
.
0
4
*
2
0
.
2
5
*
2
0
.
1
6
*
0
.
0
1
L
I
T
G
0
.
1
6
*
0
.
3
3
*
0
.
0
4
*
2
0
.
2
2
*
2
0
.
0
9
*
2
0
.
0
5
*
2
0
.
0
5
*
0
.
0
4
*
2
0
.
1
0
*
R
D
0
.
2
6
*
0
.
4
8
*
0
.
2
7
*
2
0
.
0
5
*
2
0
.
0
9
*
2
0
.
1
1
*
0
.
1
5
*
0
.
2
3
*
2
0
.
1
0
*
D
E
P
2
0
.
1
4
*
2
0
.
0
3
*
0
.
0
4
*
0
.
4
1
*
0
.
0
4
*
2
0
.
3
0
*
0
.
3
1
*
0
.
2
1
*
2
0
.
0
5
*
P
P
&
E
2
0
.
2
4
*
2
0
.
1
2
*
2
0
.
2
8
*
0
.
5
6
*
2
0
.
2
3
*
2
0
.
0
2
0
.
1
7
*
0
.
0
3
*
0
.
1
1
I
N
G
2
0
.
1
7
*
2
0
.
1
2
*
2
0
.
1
3
*
2
0
.
1
8
*
2
0
.
1
1
*
2
0
.
0
4
*
0
.
0
3
*
*
2
0
.
0
1
2
0
.
0
7
*
S
P
E
C
0
.
1
3
*
2
0
.
0
6
*
2
0
.
0
8
*
2
0
.
1
8
*
0
.
0
2
*
*
2
0
.
1
2
*
2
0
.
1
7
*
2
0
.
0
8
*
0
.
0
5
*
I
N
T
2
0
.
3
7
*
2
0
.
2
4
*
2
0
.
3
4
*
0
.
2
2
*
0
.
3
1
*
0
.
1
7
*
2
0
.
0
5
*
0
.
2
4
*
2
0
.
0
7
*
P
R
E
F
2
0
.
1
2
*
2
0
.
0
0
0
.
0
1
0
.
1
0
*
0
.
0
3
0
.
0
2
2
0
.
0
3
4
0
.
1
5
*
2
0
.
0
5
*
D
I
V
2
0
.
2
9
*
2
0
.
2
6
*
2
0
.
2
9
*
0
.
0
4
*
0
.
3
0
*
0
.
1
3
*
0
.
0
5
*
0
.
0
5
*
2
0
.
0
7
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
1
,
*
*
5
a
n
d
*
*
*
1
0
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
a
l
l
v
a
r
i
a
b
l
e
d
e
?
n
i
t
i
o
n
s
a
r
e
p
r
o
v
i
d
e
d
i
n
T
a
b
l
e
I
;
n
u
m
b
e
r
o
f
o
b
s
e
r
v
a
t
i
o
n
f
o
r
G
I
N
D
E
X
i
s
1
,
3
4
0
a
n
d
f
o
r
o
t
h
e
r
v
a
r
i
a
b
l
e
s
8
,
6
5
5
;
P
e
a
r
s
o
n
c
o
e
f
?
c
i
e
n
t
s
a
r
e
i
n
u
p
p
e
r
r
i
g
h
t
a
n
d
S
p
e
a
r
m
a
n
c
o
e
f
?
c
i
e
n
t
s
a
r
e
i
n
l
o
w
e
r
l
e
f
t
Table III.
ARJ
22,2
108
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
regressions separately starting with the PADCA regressions. Unlike in the 2SLS
regression where the coef?cient estimated for the Altman Z was positive, with 3SLS the
coef?cient is negative (t ¼ 22.55) and signi?cant at 5 per cent level. The other results
are similar between the 2SLS and 3SLS. The healthy dummy coef?cient is 20.0241
(t ¼ 21.94) which indicates that healthy ?rms have lower discretionary accruals.
Distressed ?rms also have lower discretionary accruals. We suspect that this later case
arises because these ?rms have fewer discretionary accrual opportunities.
The GINDEX coef?cient is 0.0221 and signi?cant at 1 per cent level. This result
indicates that companies with good corporate governance use income increasing
discretionary accruals. This positive association may be attributable to governance
quality and the subsequent performance outcome may bene?t the shareholders
(Subramanyam, 1996). Governance affects the use of PADCA by ?rms possibly
Dependant variables
PADCA Altman Z
Independent variables Coef?cient t-statistics Coef?cient t-statistics
Intercept 0.0381 4.22
* * *
20.2380 20.19
Altman Z 0.0035 4.96
* * *
PADCA 144.944 6.52
* * *
HEALTH 20.0187 22.69
* * *
Altman Z £ HEALTH 20.0018 23.24
* * *
DISTRESS 0.0061 0.80
Altman Z £ DISTRESS 20.0000 20.61
LagPADCA 0.0101 0.83
LEVERAGE 20.0028 20.96
SIZE 20.0029 22.56
* *
0.5450 4.01
* * *
MERGER 0.0074 1.44 21.8104 2.84
* * *
LOSS 20.0335 25.64
* * *
1.5252 2.30
* *
ROA 0.0744 8.54
* * *
29.2826 25.80
* * *
CFO 20.1510 212.79
* * *
28.8656 10.11
* * *
LagAltman 0.1801 16.55
* * *
EPS 213.4442 21.69
*
MB 0.9028 5.38
* * *
LITIGATION 0.2234 0.36
R&D 8.6875 3.49
* * *
DEPRECIATION 48.3391 3.87
* * *
PP&E 26.1869 26.45
* * *
INTANGIBLE 22.2409 21.12
SPECIAL 6.5919 3.61
* * *
INTEREST 236.3476 26.86
* * *
PREFERRED 26.34052 24.67
* * *
DIVIDEND 299.9099 1.17
Adjusted R
2
(per cent) 12.95 4.74
No. of observations 8,655 8,655
Notes: Signi?cant at:
*
0.10,
* *
0.05 and
* * *
0.01 per cent levels, respectively; all variables are
winsorized at 1st and 99th percentiles; HEALTH is an indicator variable that takes the value of 1 if the
Z-score exceeds the value of 2.67, and zero otherwise; DISTRESS is also an indicator variable that
takes the value 1 if Z-score is smaller than 1.81 and 0 otherwise; other variable de?nitions are provided
in Table I
Table IV.
2SLS regression
estimates
Discretionary
accruals
109
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
D
e
p
e
n
d
a
n
t
v
a
r
i
a
b
l
e
s
P
A
D
C
A
A
l
t
m
a
n
Z
G
I
N
D
E
X
I
n
d
e
p
e
n
d
e
n
t
v
a
r
i
a
b
l
e
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
C
o
e
f
?
c
i
e
n
t
t
-
s
t
a
t
i
s
t
i
c
s
I
n
t
e
r
c
e
p
t
2
0
.
1
4
7
0
2
2
.
2
0
*
*
1
8
.
3
8
0
4
5
.
6
0
*
*
*
7
.
6
7
6
4
1
4
.
2
0
*
*
*
A
l
t
m
a
n
Z
2
0
.
0
0
7
7
2
2
.
5
5
*
*
2
0
.
1
1
7
6
2
3
.
8
6
*
*
*
P
A
D
C
A
4
1
.
9
1
5
5
2
.
5
5
*
*
2
.
6
2
8
4
1
.
1
0
G
I
N
D
E
X
0
.
0
2
2
1
2
.
9
3
*
*
*
2
1
.
9
7
2
8
2
4
.
9
5
*
*
*
H
E
A
L
T
H
2
0
.
0
2
4
1
2
1
.
9
4
*
A
l
t
m
a
n
Z
£
H
E
A
L
T
H
0
.
0
1
1
7
1
3
.
6
4
*
*
*
D
I
S
T
R
E
S
S
2
0
.
0
3
6
4
2
3
.
1
2
*
*
*
A
l
t
m
a
n
Z
£
D
I
S
T
R
E
S
S
0
.
0
1
8
0
8
.
3
8
*
*
*
L
a
g
P
A
D
C
A
2
0
.
0
1
4
2
2
0
.
4
9
L
E
V
E
R
A
G
E
2
0
.
0
0
8
5
2
0
.
3
3
S
I
Z
E
2
0
.
0
0
3
1
2
0
.
8
3
0
.
3
3
3
2
1
.
9
4
*
0
.
2
9
3
6
4
.
1
9
*
*
*
L
O
S
S
0
.
0
0
2
5
0
.
2
9
2
0
.
5
6
3
2
2
1
.
3
6
R
O
A
0
.
1
5
3
8
3
.
8
8
*
*
*
2
0
.
9
0
3
9
2
0
.
2
0
C
F
O
2
0
.
3
5
8
8
2
8
.
3
2
*
*
*
1
9
.
4
3
2
4
2
.
8
3
*
*
*
L
a
g
A
l
t
m
a
n
0
.
3
3
1
8
1
8
.
2
1
*
*
*
E
P
S
1
9
.
4
2
7
1
0
.
8
5
M
B
0
.
0
9
3
0
0
.
3
8
L
I
T
I
G
A
T
I
O
N
2
1
.
1
0
8
3
2
2
.
3
9
*
*
2
0
.
5
7
7
5
2
2
.
6
4
*
*
*
R
&
D
3
.
6
8
9
7
0
.
8
4
D
E
P
R
E
C
I
A
T
I
O
N
2
1
3
.
6
7
0
1
2
1
.
4
4
P
P
&
E
2
0
.
2
3
3
1
2
0
.
1
9
I
N
T
A
N
G
I
B
L
E
2
2
.
2
2
4
5
2
2
.
5
6
*
*
S
P
E
C
I
A
L
2
0
.
5
1
9
5
2
0
.
2
7
I
N
T
E
R
E
S
T
2
1
4
.
9
2
5
2
2
1
.
6
2
M
E
R
G
E
R
0
.
2
7
1
0
1
.
6
3
P
R
E
F
E
R
R
E
D
2
1
.
3
2
7
3
2
0
.
9
2
D
I
V
I
D
E
N
D
1
3
1
.
2
1
6
1
1
.
0
3
S
y
s
t
e
m
w
e
i
g
h
t
e
d
R
2
(
p
e
r
c
e
n
t
)
4
3
.
5
1
D
u
r
b
i
n
-
W
a
t
s
o
n
1
.
8
9
1
.
8
9
2
.
0
3
F
-
v
a
l
u
e
3
.
7
9
*
*
*
3
.
7
4
*
*
*
2
.
0
3
*
*
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
0
.
1
0
,
*
*
0
.
0
5
a
n
d
*
*
*
0
.
0
1
p
e
r
c
e
n
t
l
e
v
e
l
s
,
r
e
s
p
e
c
t
i
v
e
l
y
;
v
a
r
i
a
b
l
e
d
e
?
n
i
t
i
o
n
s
a
r
e
g
i
v
e
n
i
n
T
a
b
l
e
I
;
n
¼
7
6
7
Table V.
3SLS regression
estimates with
governance index
ARJ
22,2
110
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
because managers may use their discretion to achieve some reporting objective to
bene?t shareholders[24]. Managers apparently are not grilled on the details of ?nancial
reports by non-executive members of their board of directors. This may be due to the
extreme level of detail that would be required for the board of directors to uncover
cases of PADCA[25].
The second regression equation has Altman Z-score as the dependent variable.
Here, GINDEX is negatively associated with Altman Z-score. This indicates that
as ?rms strengthen their corporate governance that their Altman Z-score declines.
Since this result is found with the 3SLS regression there is no confusion as to the
direction of the relationship. That is, our result suggest that as corporate governance
strengthens that ?rms become less healthy which may occur because they may lose
markets and pro?ts to ?rms with less governance. The positive and signi?cant
coef?cient on PADCA suggests that the more ?rms use discretionary accruals the
better their ?nancial health appears, as measured by Altman Z.
In equation (3), the GINDEX is the dependent variable. The signi?cant coef?cient on
the Altman Z-variable suggests that the higher is the Altman Z-score (the better the
?rm’s ?nancial health) the lower is the GINDEX. This result serves as a counterpoint to
the results in the second equation; now the argument is that members of the board of
directors of healthier ?rms are less vigilant and have lower governance scores. The
coef?cient estimated on PADCA is not signi?cant, as we expected a priori. The amount
of discretionary accruals does not affect the governance index. We also (results not
shown) tested lagged PADCA as a determinant of the GINDEX thinking that the board
might increase its vigilance after managers engaged in discretionary accruals. The
coef?cient estimated on lagged PADCA was not signi?cant.
Perhaps, the most interesting ?nding is that a higher GINDEX lowers a ?rm’s
Altman Z while a higher Altman Z lowers the ?rms GINDEX. The interpretation is that
better corporate governance is not associated with improved corporate health and that
improved corporate health is not associated with better corporate governance. Firms
may reach an equilibrium level of GINDEX and Altman Z after which increases in
either governance or health leads to undesired results.
7. Conclusion
We investigate the relationship between discretionary accrual level, ?rms’ ?nancial
distress as measured by the Altman Z, and corporate governance quality as extracted
from the IRRC database applied to US manufacturing ?rms between the years 2001
and 2003. We ?nd that ?rms’ managers appear to exercise discretionary accruals when
their ?rms are in the mid-range or gray area of the Altman Z in order to avoid being
classi?ed as distressed or if they are successful in doing it to be classi?ed as a healthy
?rm. We did not observe the same behavior for ?rms in either the distressed or healthy
classi?cation areas based on the Altman Z-score. Gray area companies have larger
PADCAs. There are two possible explanation for this; they are either opportunistic or
they face noise in their accounting information. We do not ?nd that heightened
governance reduces the incidence of discretionary accruals. Our ?nding suggests that
non-executive members of boards of directors need to pay closer attention to ?rms in
the gray region.
This study contributes to both the accounting and ?nance literature in three ways.
First of all, we document the relationship between a ?rms’ ?nancial health and its use of
Discretionary
accruals
111
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
discretionary accruals by taking into consideration corporate governance quality. This
result is important because it shows the existence of agency problems in credit markets
when ?nancials signal stressful results especially when there is good corporate
governance. Second, we document the importance of paying attention to ?rms at the
boundaries of healthy and gray areas as they appear to be more open to accrual
manipulation. This result is interesting in the sense that both creditor and equity
investors must be careful about the ?nancial classi?cations of ?rms when they evaluate
potential investments. They should also pay close attention to accruals, if the ?rm is on
the boundary or in the gray area that corporate governance may play intervening role
in this.
When endogeneity is tested between the three variables, GINDEX, Altman Z and
PADCA, the 3SLS results yield an interesting conclusion that better corporate
governance is not associated with improved corporate health and that improved
corporate health is not associated with better corporate governance. We suspect that
?rms that deviate from an equilibrium or best level for their GINDEX and Altman Z
suffer unexpected changes in the other variable.
Notes
1. One may argue that a ?rm under close scrutiny may not be able to manage its earnings since
detection of earnings management would not be good for managers. On the other hand,
managers know that ?nancial reports are value relevant because they are closely followed
therefore they are motivated to fudge the numbers to in?uence the investors’ expectations.
2. Some might argue that a 40-year-old model, the Altman Z, is not the best choice since other
models exist (e.g. the Shumway, the Merton, the KMV, or the Ohlson, 1980 models). Our
choice of Altman’s model was based on its widespread acceptance and reuse throughout the
world. Some authors, Begley, Ming and Watts, for example argue that his cutoff points are
outdated. We continue with the original cutoff points since our objective is not to predict
?nancially distressed ?rms but rather to document the relationship between ?nancial status
classi?cation, corporate governance quality and the likelihood managers’ using
discretionary accruals.
3. The advantage of using an Altman Z-approach to classify ?rms is that it “considers an entire
pro?le of characteristics shared by a speci?c ?rm and transforms that into a univariate
statistic” (Cleary, 1999).
4. Reasoning behind the use of PADCA speci?cally to measure discretionary accruals will be
explained in subsequent sections.
5. The sample period is chosen speci?cally to select ?rms with a higher probability of having
lower Altman Z-scores than during periods of economic health. We also choose the
manufacturing industry since it has been established that the Altman Z-methodology works
well for these ?rms. We also control for leverage and other ?rm characteristics as a result of
the signaling hypothesis.
6. The more discretion exercised by management, the greater the opportunity for them to
manipulate reported earnings.
7. The Altman Z-score is a widely used classi?cation and bankruptcy estimation measurement
tool in the ?eld of credit risk analysis, distressed investing; mergers and acquisitions (M&A)
target analysis, and turnaround management by both practitioners and academics
(Calandro, 2007).
ARJ
22,2
112
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
8. Accruals may also be used opportunistically by executives, because they need to make
judgments while reporting. Therefore, the greater a company’s management discretion
concerning accruals the greater the opportunity to manipulate reported earnings.
9. When calculating PADCA, previous year ROA is used. That may create limitation in our
research design since we do not match ?rms according to their ?nancial status, i.e. Altman
Z-score. This is also limitation of PADCA measures that assume ?rms are identical at the
previous year just by looking at their ROA. But this method is the most ef?cient one
available compared with other methods (Kothari et al., 2005). In our design to decrease this
measurement issue, we used ROA at the current period as a control variable that is
correlated signi?cantly with Altman Z, PADCA and G-score.
10. Regular ordinary least squares (OLS) regression provided similar results with the same
coef?cient signs. Firms were divided into distressed, gray and healthy classi?cations and
OLS was run for each subsample. We tested for multicollinearity in both our 2SLS and 3SLS
speci?cation and rejected it.
11. We employed 2SLS to solve the endogeneity problem between PADCA and Altman Z
without taking into consideration corporate governance quality where sample size decreases
from 8,655 to 767 with 3SLS regression.
12. Ali et al. (2007) investigates the corporate disclosure practices of family ?rms, and they
document that family ?rms report better disclosure quality earnings and are more likely to
warn for a given magnitude of bad news in comparison with non-family ?rms. We also
reported the 2SLS regression model results because with 3SLS our sample size decreases
from 8,655 to 767. Some inferences are also affected because of this sample size difference,
but overall our conclusions are similar.
13. The t-statistics are corrected using the Huber-White procedure by following Petersen (2009).
14. Discretionary accruals reverse over time (Guay et al., 1996). They state that “the
discretionary accrual is expected to reverse in the future periods and nondiscretionary
earnings are also expected to decline”. They do not say accruals reverse at the next period,
but in the future. For example, a manager may think to boost earnings by choosing a lower
depreciation expense in the current period. Of course, a higher expense is required in
subsequent periods to make up for the difference but it may take many years for the two
depreciation methods to be equal. The authors recognize this problem and control for
depreciation and other kinds of discretionary accruals. The correlation and time series
properties of discretionary accruals and earning process are shown in Beaver et al. (1980) and
Ashton (2005). They document that the serial correlation of discretionary accruals from one
to subsequent periods is negative but not perfect which indicates that they do not reverse
next period but takes time. On the other hand, one discretionary accrual may be replaced
with another one over time that provides a large room for managers to maneuver their
?nancial reporting strategies. The time horizon of our study may not be long enough for
discretionary accruals to reverse.
15. Altman Z-score is a widely used classi?cation and bankruptcy estimation measurement tool
in the ?eld of credit risk analysis, distressed investing; M&A target analysis and turnaround
management by both practitioners and academics (Calandro, 2007).
16. The LagPADCA variable is included in the regression model to test for the reversal of
accruals over time.
17. Firms in the Compustat database for which the required data is available are included in the
estimation process. We also use the two-digit industry de?nition for these tests.
18. Obviously, there are other periods of market downturn before 2001. Moreover, the economic
downturns may have different effects on the business environment. The sample period may
Discretionary
accruals
113
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
be extended to other periods when the US economy hit bottom in November 1982 and March
1991 according to National Bureau of Economic Research.
19. Mean tests would not be conclusive since other ?rms’ characteristics are diverse for different
classes of ?rms, i.e. healthy, gray and distress ?rms as reported at Table II.
20. Since markets expect managers to manage their earnings, investors discount the value of the
earnings such as for $100 reported earnings they discount it to $95. Therefore, managers
manage the earnings to $105 to reach equilibrium at $100.
21. We look at the Pearson and Spearman correlations between lagged PADCA and GINDEX
and could not ?nd association as well.
22. We include 2SLS regression results, because our sample size decreases drastically when we
include GINDEX in our regression analysis which is reported as the 3SLS model.
23. Variance in?ation factor (VIF) are found to range between 1 and 4 which is less than the
critical value of 20. Therefore, we conclude that there is not a multicollinearity problem.
24. We do not test the market response, therefore, we may not know whether using discretionary
accruals bene?ts shareholders.
25. VIF scores for the variables that are used in 3SLS have the values that are less than 20,
therefore, we do not have tolerable multicollinearity problem.
References
Ali, A., Chen, T. and Radhakrishnan, S. (2007), “Corporate disclosures by family ?rms”, Journal
of Accounting and Economics, Vol. 44, pp. 238-86.
Altman, E. (1968), “Financial ratios, dicriminant analysis and prediction of corporate
bankruptcy”, Journal of Finance, Vol. 23, pp. 589-609.
Al-Tuwaijri, S.A., Christensenand, T.E. and Hughes, K.E. II (2004), “The relation among
environmental disclosure, environmental performance, and economic performance:
a simultaneous equations approach”, Accounting, Organizations and Society, Vol. 29,
pp. 447-71.
Ashton, D. (2005), “Discussion of cost of capital, strategic disclosures and accounting choice”,
Journal of Business Finance & Accounting, Vol. 32, pp. 635-41.
Beaver, W.H., Lamber, R. and Morse, D. (1980), “The information content of security prices”,
Journal of Accounting and Economics, Vol. 2 No. 1, pp. 3-28.
Bernanke, B. and Gertler, M. (1990), “Financial fragility and economic performance”,
The Quarterly Journal of Economics, Vol. 105, pp. 87-114.
Bernstein, L. (1993), Financial Statement Analysis, 5th ed., Irwin, Homewoood, IL.
Bowen, R., Ducharmeand, L. and Shores, D. (1995), “Stakeholders’ implicit claims and accounting
method choice”, Journal of Accounting and Economics, Vol. 20, pp. 255-95.
Bowen, R., Rajgopal, S. and Venkatachalam, M. (2008), “Accounting discretion, corporate
governance, and ?rm performance”, Contemporary Accounting Research, Vol. 25,
pp. 310-405.
Burgstahler, D. and Dichev, I. (1997), “Earnings management to avoid earnings decreases and
losses”, Journal of Accounting and Economics, Vol. 24, pp. 99-126.
Bushee, B.J. (1998), “The in?uence of institutional investors on myopic R&D investment
behavior”, The Accounting Review, Vol. 73, pp. 305-34.
Calandro, J. (2007), “Considering the utility of Altman’s Z-score as a strategic assessment and
performance management tool”, Strategy & Leadership, Vol. 35, pp. 37-43.
ARJ
22,2
114
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Chi, L. and Tang, T. (2007), “Impact of reorganizational announcements on distressed-stock
return”, Economic Modeling, Vol. 24, pp. 749-67.
Clarke, T. (2005), “Accounting for Enron: shareholder value and stakeholder interests”,
Corporate Governance: An International Review, Vol. 13, pp. 598-612.
Cleary, S. (1999), “The relationship between ?rm investment and ?nancial status”, Journal
of Finance, Vol. 54, pp. 673-92.
Cook, K.A., Huston, G.R. and Omer, T.C. (2008), “Earnings management through effective tax
rates: the effect of tax planning investment and Sarbanes-Oxley Act of 2002”,
Contemporary Accounting Research, Vol. 25 No. 2, pp. 447-71.
Dechow, P.M. and Schrand, C.M. (2004), Earnings Quality, The Research Foundation of CFA
Institute, New York, NY.
Dechow, P.M. and Skinner, D. (2000), “Earnings management: reconciling the views of
accounting academics, practitioners and regulators”, Accounting Horizons, Vol. 14,
pp. 235-50.
Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1995), “Detecting earnings management”,
The Accounting Review, Vol. 70, pp. 3-42.
Dechow, P.M., Sloan, R.G. and Sweeney, A.P. (1996), “Causes and consequences of earnings
manipulation: an analysis of ?rms subject to enforcement actions by the SEC”,
Contemporary Accounting Research, Vol. 13, pp. 1-36.
DeFond, M. and Jiambalvo, J. (1994), “Debt covenant violation and manipulation of accruals”,
Journal of Accounting and Economics, Vol. 17, pp. 145-76.
Del Guercio, D. (1996), “The distorting effect of the prudent-man laws on institutional equity
investments”, Journal of Financial Economics, Vol. 40, pp. 31-62.
Dhaliwal, D.S., Gleason, C. and Mills, L. (2004), “Last-chance earnings management: using the tax
expense to meet analysts’ forecasts”, Contemporary Accounting Research, Vol. 21,
pp. 431-59.
Dhaliwal, D.S., Li, O.Z. and Xie, H. (2005), “Institutional investors, ?nancial health equity
valuation”, available at: http://ssrn.com/abstract¼493046 (accessed 3 June 2009).
Fair?eld, P., Whisenant, J. and Yohn, T. (2003), “Accrued earnings and growth: implications for
future pro?tability and market mispricing”, The Accounting Review, Vol. 78, pp. 353-71.
Farrell, K. and Whidbee, D. (2003), “Impact of ?rms performance expectations on CEO turnover
and replacement decisions”, Journal of Accounting and Economics, Vol. 36, pp. 165-96.
Fazzari, S., Hubbard, G. and Petersen, B. (1988), “Financing constraints and corporate
investment”, Brooking Papers on Economic Activity, Vol. 1, pp. 141-206.
Fields, T.D., Lys, T.Z. and Vincent, L. (2001), “Empirical research of accounting choice”, Journal
of Accounting and Economics, Vol. 31, pp. 255-307.
Francis, J., Huang, A.H., Rajgopal, S. and Zang, A.Y. (2008), “CEO reputation and earnings
quality”, Contemporary Accounting Research, Vol. 25, pp. 109-47.
Gertler, M. (1992), “Financial capacity and output ?uctuations in an economy with multi-period
?nancial relationships”, Review of Economic Studies, Vol. 59 No. 3, pp. 455-72.
Gompers, P.A., Ishii, J.L. and Metrick, A. (2003), “Corporate governance and equity prices”,
The Quarterly Journal of Economics, Vol. 118, pp. 107-55.
Graham, J.R., Harvey, C.R. and Rajgopal, S. (2005), “The economic implications of corporate
?nancial reporting”, Journal of Accounting and Economics, Vol. 40, pp. 3-73.
Grice, J.S. and Ingram, R.W. (2001), “Tests of the generalizability of Altman’s bankruptcy
prediction model”, Journal of Business Research, Vol. 54, pp. 53-61.
Discretionary
accruals
115
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Guay, W.R., Kothari, S.P. and Watts, R.L. (1996), “A market-based evaluation of discretionary
accrual models”, Journal of Accounting Research, Vol. 34, pp. 83-105.
Healy, P. (1985), “The impact of bonus schemes on the selection of accounting principles”, Journal
of Accounting and Economics, Vol. 7, pp. 85-107.
Healy, P. and Wahlen, J. (1999), “A review of the earnings management literature and its
implications for standard setting”, Accounting Horizons, Vol. 13, pp. 365-83.
Hessel, C.A. and Norman, M. (1992), “Financial characteristics of neglected and institutionally
held stocks”, Journal of Accounting, Auditing and Finance, Vol. 7 No. 3, pp. 313-30.
Holthausen, R., Larcker, D. and Sloan, R.G. (1995), “Annual bonus schemes and manipulation of
earnings: additional evidence on bonus plans and income management”, Journal
of Accounting and Economics, Vol. 19 No. 1, pp. 29-74.
Hsu, G.C.-M. and Koh, P.-S. (2005), “Does the presence of institutional investors in?uence
accruals management? Evidence from Australia”, Corporate Governance: An International
Review, Vol. 13, pp. 809-23.
Jones, J. (1991), “Earnings management during import relief investigations”, Journal
of Accounting Research, Vol. 29, pp. 193-228.
Jones, K.L., Krishnan, G.V. and Melendrez, D. (2008), “Do models of discretionary accruals detect
actual cases of fraudulent and restated earnings? An empirical analysis”, Contemporary
Accounting Research, Vol. 25, pp. 499-531.
Kaplan, S. and Zingales, L. (1997), “Do investment cash-?ow sensitivities provide useful
measures of ?nancing constraints?”, Quarterly Journal of Economics, Vol. 112, pp. 169-216.
Kasznik, R. (1999), “On the association between voluntary disclosure and earnings
management”, Journal of Accounting Research, Vol. 37, pp. 57-81.
Kothari, S.P., Leone, A. and Wasley, C. (2005), “Performance matched discretionary accrual
measures”, Journal of Accounting and Economics, Vol. 39, pp. 163-97.
Mather, P. and Ramsay, A. (2006), “The effects of board characteristics on earnings management
around Australian CEO changes”, Accounting Research Journal, Vol. 19, pp. 78-93.
Matsunaga, S. and Park, C. (2001), “The effect of missing quarterly earnings benchmark on the
CEO’s annual bonus”, The Accounting Review, Vol. 76, pp. 313-32.
Merton, R.C. (1974), “On the pricing of corporate debt: the risk structure of interest rates”, Journal
of Finance, Vol. 29, pp. 449-70.
Ohlson, J. (1980), “Financial ratios and the probabilistic prediction of bankruptcy”, Journal
of Accounting Research, Vol. 18 No. 1, pp. 109-31.
Ohlson, J. (1995), “Earnings, book values, and dividends in equity valuation”, Contemporary
Accounting Research, Vol. 18 No. 1, pp. 661-87.
Petersen, M. (2009), “Estimating standard errors in ?nance panel data sets: comparing
approaches”, Review of Financial Studies, Vol. 22, pp. 435-80.
Platt, H.D. and Platt, M.B. (1991), “A note on the use of industry-relative ratios in bankruptcy
prediction”, Journal of Banking & Finance, Vol. 15, pp. 183-1194.
Rees, L., Gill, S. and Gore, R. (1996), “An investigation of asset write-downs and concurrent
abnormal accruals”, Journal of Accounting Research, Vol. 34, pp. 157-69.
Richardson, S.A., Tuna, A.I. and Wu, M. (2002), “Predicting earnings management: the case of
earnings restatements”, available at: http://ssrn.com/abstract¼338681 (accessed 3 June
2008).
Shleifer, A. and Vishny, R.W. (1997), “A survey of corporate governance”, Journal of Finance,
Vol. 52, pp. 737-83.
ARJ
22,2
116
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Shumway, T. (2001), “Forecasting bankruptcy more accurately: a simple hazard model”,
The Journal of Business, Vol. 74, pp. 101-24.
Sloan, R.G. (1996), “Do stock prices fully re?ect information in accruals and cash ?ows about
future earnings?”, The Accounting Review, Vol. 71, pp. 289-316.
Subramanyam, K.R. (1996), “The pricing of discretionary accruals”, Journal of Accounting and
Economics, Vol. 22, pp. 249-81.
Watts, R.L. and Zimmerman, J. (1990), “Positive accounting theory: a ten year perspective”,
The Accounting Review, Vol. 65, pp. 131-56.
Yeh, Y.H. and Lee, T.-S. (2004), “Corporate governance and ?nancial distress: evidence from
Taiwan”, Corporate Governance: An International Review, Vol. 12, pp. 378-88.
Zmijewski, M. (1984), “Methodological issues related to the estimation of ?nancial distress
prediction models”, Journal of Accounting Research, Vol. 22, pp. 59-82.
Appendix. PADCA calculation
Following Kothari et al. (2005), we estimate the cross-sectional Jones model for each two-digit SIC
industry group with at least ten ?rms in year t. Speci?cally, we estimate the following modi?ed
Jones model to determine coef?cient values and the value of the error term:
TCA
j;t
¼ b
0
þ b
1
ðSCASSET
j;t
Þ þ b
1
ðDREV
j;t
2DAR
j;t
Þ þ error
j;t
ðA1Þ
where TCA – change in non-cash current assets minus the change in current liabilities excluding
the current portion of long-term debt so that TCA ¼ DCurrent assets-DCash and short-term
investments-DCurrent liabilities þ DDebt in current liabilties; SCASSET – one scaled by lagged
total assets (1/lagASSET); DREV
j,t
– change in total revenue (sales) for ?rm j at year t scaled by
lagged total asset; DAR
j,t
¼ change in account receivable for ?rm j at year t scaled by lagged
total assets; error
j,t
– error term for equation (A1) for ?rm j at year t.
The two-digit industry and year-speci?c estimates obtained from estimating equation (A1)
are used to compute normal accruals (NA
j,t
). The difference between TCA
j,t
and NA
j,t
is the
abnormal (discretionary) accruals for ?rm j in year t. Following Kothari et al. (2005) ?rm j in year
t is matched with a companion ?rm with the closest lagged-return on assets in the same two-digit
SIC code in year t. We obtain PADCA by subtracting matched ?rm’s abnormal accruals from
?rm j’s abormal accrual at year t.
Corresponding author
Harlan Platt can be contacted at: [email protected]
Discretionary
accruals
117
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Nuraddeen Usman Miko, Hasnah Kamardin. 2015. Impact of Audit Committee and Audit Quality on
Preventing Earnings Management in the Pre- and Post- Nigerian Corporate Governance Code 2011.
Procedia - Social and Behavioral Sciences 172, 651-657. [CrossRef]
2. Aziatul Waznah Ghazali, Nur Aima Shafie, Zuraidah Mohd Sanusi. 2015. Earnings Management: An
Analysis of Opportunistic Behaviour, Monitoring Mechanism and Financial Distress. Procedia Economics
and Finance 28, 190-201. [CrossRef]
3. Nor Farhana Selahudin, Nor Balkish Zakaria, Zuraidah Mohd. Sanusi. 2014. Remodelling the Earnings
Management with the Appearance of Leverage, Financial Distress and Free Cash Flow: Malaysia and
Thailand Evidences. Journal of Applied Sciences 14, 2644-2661. [CrossRef]
4. Nor Farhana Selahudin, Nor Balkish Zakaria, Zuraidah Mohd Sanusi, Pornanong Budsaratragoon. 2014.
Monitoring Financial Risk Ratios and Earnings Management: Evidence from Malaysia and Thailand.
Procedia - Social and Behavioral Sciences 145, 51-60. [CrossRef]
5. Panagiotis Dimitropoulos. 2011. Corporate Governance and Earnings Management in the European
Football Industry. European Sport Management Quarterly 11, 495-523. [CrossRef]
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
1
:
0
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_425876692.pdf