Journal of Applied Accounting Research

Efficient contracting, earnings smoothing and managerial accounting discretion

Mohamed Khalil Jon Simon
Article information:
To cite this document:
Mohamed Khalil Jon Simon , (2014),"Efficient contracting, earnings smoothing and managerial accounting
discretion", Journal of Applied Accounting Research, Vol. 15 Iss 1 pp. 100 - 123
Permanent link to this document:
http://dx.doi.org/10.1108/JAAR-06-2012-0050
Downloaded on: 23 December 2014, At: 03:11 (PT)
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

References: this document contains references to 65 other documents.

To copy this document: [email protected]
The fulltext of this document has been downloaded 253 times since 2014*
Users who downloaded this article also downloaded:
Prapaporn Kiattikulwattana, (2014),"Earnings management and voluntary disclosure of management's
responsibility for the financial reports", Asian Review of Accounting, Vol. 22 Iss 3 pp. 233-256 http://
dx.doi.org/10.1108/ARA-11-2013-0075
Tesfaye T. Lemma, Minga Negash, (2014),"Determinants of the adjustment speed of capital structure:
Evidence from developing economies", Journal of Applied Accounting Research, Vol. 15 Iss 1 pp. 64-99
http://dx.doi.org/10.1108/JAAR-03-2012-0023
Ahsan Habib, Haiyan Jiang, Donghua Zhou, (2014),"Audit quality and market pricing of earnings
and earnings components in China", Asian Review of Accounting, Vol. 22 Iss 1 pp. 20-34 http://
dx.doi.org/10.1108/ARA-05-2013-0034

Access to this document was granted through an Emerald subscription provided by 380369 []
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.

 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0967-5426.htm

JAAR
15,1
Efficient contracting, earnings
smoothing and managerial
accounting discretion
100 Mohamed Khalil
Accounting and Finance, Hull University Business School,
The University of Hull, Hull, UK and Accounting Department,
Faculty of Commerce, Tanta University, Tanta, Egypt, and
Jon Simon
Accounting and Finance, Hull University Business School,
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

The University of Hull, Hull, UK

Abstract
Purpose – The purpose of this paper is to examine whether the contracting incentives (i.e. bonus
plans, debt covenants, political costs hypotheses), and income smoothing can explain accounting
choices in an emerging country, Egypt.
Design/methodology/approach – The paper uses the ordinary least square regression model to
examine the relationship between earnings management and reporting objectives. A sample of 438
non-financial firms listed on the Egyptian Exchange over the period 2005-2007 is used.
Findings – The paper finds that the contracting objectives explain little of the variations in
accounting choices (i.e. discretionary accruals) in the Egyptian context. However, the paper finds that
mangers are likely to smooth the reported earnings by managing the accrual component in an attempt
to reduce the fluctuation in reported earnings by increasing (decreasing) earnings when earnings are
low (high) in attempt to reduce the variability of the reported earnings.
Research limitations/implications – The empirical results rely on the ability of earnings
management proxies to adequately capture earnings manipulation activities.
Practical implications – The findings of the study should be of substantial interest to regulators
and policy makers. The results implicitly contribute to the ongoing argument in relation to the optimal
flexibility permitted by standard setting and the argument that tightening the accounting standards
and mandating International Financial Reporting Standards are likely to improve reporting quality
and reduce opportunistic earnings management. The results reveal that many of the weaknesses
related to corporate reporting in emerging countries may result from the inadequate enforcement of
the law and the weak legal protection of minority shareholders. The results also highlight the crucial
role of understanding the reporting incentives, which is mainly shaped by institutional and market
forces and the legal environment, in explaining accounting choices.
Originality/value – Unlike previous studies that tested an individual objective, this study examines
the trade-offs among various reporting objectives in an emerging economy.
Keywords Discretionary accruals, Leverage, Earnings smoothing, Egyptian firms,
Managerial compensation, Political costs
Paper type Research paper

1. Introduction
The efficient contracting perspective of accounting choices provides evidence
Journal of Applied Accounting
consistent with the idea that managers exercise accounting discretion to increase their
Research compensation, avoid debt covenants violation, and reduce the chance of exposure to
Vol. 15 No. 1, 2014
pp. 100-122
r Emerald Group Publishing Limited
0967-5426 The authors are grateful to Aydin Ozkan, Khaled Hussainey, Ros Haniffa, Julia Mundy (Editor),
DOI 10.1108/JAAR-06-2012-0050 and two anonymous referees for their useful comments and suggestions.
 political or governmental intrusions in their business’s affairs. Management may also Managerial
tend to smooth the reported earnings in an attempt to meet investors’ expectations accounting
of future cash flows[1].
Accounting choices have been the subject of several studies, the majority of which discretion
were related generally to well-developed capital markets and in particular to the USA
and the UK, in which the ownership of companies is well-dispersed among outside
shareholders and investor protection is strong. However, relatively few studies have 101
directly addressed the trade-offs among accounting choices in emerging countries. In
this study, we extend this area of research by utilizing a unique data set and focusing
on an explanation of the accounting choices for an emerging market, namely Egypt,
which is characterized by highly concentrated ownership and poor investor protection.
Egypt is considered an ideal setting in which to conduct this study for several
reasons. First, whilst the Egyptian privatization programme started in the second half
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

of the 1990s, corporate ownership is still highly concentrated within families, the State,
and banks. A fundamental problem related to such concentrated ownership is
how information asymmetry between controlling and minority shareholders (and other
users, including debt holders, customers, suppliers, and employees) is addressed.
While recognizing the role of timely financial statements in channelling information,
the information asymmetry problem in emerging countries is more likely resolved by
closer personal channels and private communications with dominant shareholders
(Ball et al., 2000). This is likely to lead to the diversion (or abuse) of firm resources by
controlling shareholders. For example, the existing voting rules in Egypt entitle the
controlling owners to elect board members to represent their interests at the expense of
minority shareholders. The majority of Egyptian board members are considered weak
because they are usually chosen from family, close relatives, and friends who lack
adequate financial knowledge (Sourial, 2004). This is, in turn, more likely to encourage
resource expropriation and allow controlling shareholders to more easily manage the
firm’s reported earnings (e.g. Guthrie and Sokolowsky, 2010).
Second, although a considerable improvement has been made in reducing
differences between the Egyptian Accounting Standards and International Financial
Reporting Standards (IFRSs), there are still some concerns about weak enforcement,
lack of implementation guidelines, and inadequate knowledge of IFRSs (including their
Arabic translation), leading to poor quality financial reporting in general and the
inability to limit managers’ scope to manage earnings (Report on the Observance
of Standards and Codes (ROSC), 2009).
Finally, expected litigation cost is another fundamental variable that influences
managers’ disclosure decisions (Kothari et al., 1988). It is expected that the propensity
to manage earnings is more likely to increase when ligation costs are low. In the
Egyptian setting, managers are more likely to engage in earnings management
because civil litigation and securities lawsuits are rare. For example, although the
Capital Market Authority (CMA) has administrative sanctioning powers, including de-
listing, suspension of licences, cancelling transactions, and imposing monetary
penalties, weak enforcement has been a feature of the Egyptian system (ROSC, 2009).
In addition, the regulatory framework contains a significant number of overlapping
and ambiguous laws, which weakens law enforcement (ROSC, 2009). Consequently,
such institutional characteristics are expected to allow managers to opportunistically
exercise discretion over reported earnings.
Prior theoretical and empirical accounting studies have focused on the extent to
which earnings are managed to achieve particular objectives[2]. Despite valuable
 JAAR contributions provided by this stream of research towards understanding the causes
15,1 and consequences of managerial discretion, such empirical studies give only peripheral
attention to the potential trade-offs among several competing reporting objectives
that are likely to explain accounting choices. Notable exceptions are Young (1998),
Darrough et al. (1998), Heflin et al. (2002), and Dey et al. (2008). Young (1998) finds little
evidence to support the efficient contracting explanation for managerial discretion
102 choices in the UK. Darrough et al. (1998) also provide support for leverage incentive
only for the years after the Japanese market crash of 1990 and for the political costs
hypothesis prior to the crash. They also show that Japanese managers choose income-
increasing accounting accruals to increase their bonus and increase the amount of
outside funding. By focusing on a sample of US listed firms, Heflin et al. (2002) find that
managers use accounting latitude to reduce the possibility of debt covenants violation
and to avoid political costs.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

In the Egyptian context, Dey et al. (2008) find evidence consistent with the bonus
plans and debt contracts objectives. However, their study uses a single account
approach, whereby earnings management is measured in specific areas of the financial
statements such as depreciation and inventory. Our study differs in a significant
respect, in that we use abnormal accruals as a proxy for earnings management
throughout the financial statements rather than a more limited single account
approach[3]. In addition, our study is superior to that undertaken by Dey et al. (2008) as
our sample size is larger and more recently collected.
Our findings contribute to the existing literature in two main ways. First, as the
reporting practices are closely linked to their institutional context (Ball et al., 2000),
generalization of findings from studies conducted in developed countries may be
misleading and inappropriate when used to explain accounting choices in emerging
countries. Therefore, using a unique data set that reflects distinct legal and
institutional features helps shed additional light on the role of the institutional
characteristics in explaining accruals choices in an emerging economy. Furthermore,
the institutional environment and payout preferences of managers and large
shareholders in managing reported earnings have potentially greater influence in
explaining accruals choices than other factors, such as adoption of IFRSs.
Second, unlike previous studies that are restricted to testing an individual objective,
the results of this study provide greater insights into understanding the trade-offs
among competing reporting objectives and determinants of accounting choice.
Focusing on a single objective at a time may lead to insufficient evidence about
incentives that explain accounting choices; the same accounting choice may result in
accomplishing several objectives (Fields et al., 2001). For example, income increasing
choices that drive higher managerial compensation to benefit managers at the expense
of other parties may also serve to avoid debt covenants violations, which may harm
creditors and benefit other stakeholders (Fields et al., 2001). Similarly, results of studies
that focus on a single accounting choice at a time are also limited because most
managers are likely to seek to accomplish one or more reporting objectives using
a single choice, or a portfolio of accounting choices (Fields et al., 2001; Watts and
Zimmerman, 1990).
Our analysis yields interesting results. We find that the traditional costly
contracting incentives explain little of the variations in accounting choices
(i.e. discretionary accruals) in the Egyptian context, while earnings smoothing
activity explains much of the cross-sectional variation in managerial choices.
Specifically, managers are likely to use the accrual component in an attempt to reduce
 the fluctuation in reported earnings by increasing (decreasing) earnings when Managerial
earnings are low (high) in an attempt to reduce the variability of the reported earnings accounting
either to gain personal and/or attain the contractual objectives[4].
The remainder of this paper is organized as follows. In Section 2, we develop our discretion
empirical hypotheses, while in Section 3 we provide the methodology. Section 4
provides description of the data and descriptive statistics. Section 5 presents our
empirical findings from univariate and multivariate analyses. Section 6 introduces 103
additional tests. Finally, Section 7 concludes.

2. Literature review and hypotheses development

2.1 Executive bonus plans hypothesis
A number of studies have reported that managers use discretionary accruals in an
attempt to maximize their compensation. For example, Healy (1985) points out that
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

compensation schemes do not always induce managers to select income-increasing

accounting choices. Rather, managers are likely to choose income-decreasing accruals
to save income, in order to increase their future expected bonus when current reported
earnings are beyond the bounds embedded in compensation contracts (i.e. above the
upper limit or below the lower limit). However, managers may choose income-
increasing accruals when the current level of reported earnings is within these bounds.
In a similar vein, Holthausen et al. (1995) find that managers make more income-
decreasing choices when their bonus is at the upper bound than when it is between the
lower and upper bounds. However, they do not find the same result when bonuses are
below the lower bound. Holthausen et al. (1995) find that discretionary accruals are
more negative (i.e. income-decreasing choice), when the CEO bonus is at the upper
bound than when it is between the lower and upper bounds. They also find that
managers manipulate earnings downwards when their bonuses are at their maximum.
Several studies document evidence supporting the view that managers manipulate
earnings when their potential compensation is linked with the value of shares and options.
For example, Cheng and Warfield (2005) demonstrate that the sensitivity of managers’
wealth to the short-term stock price may motivate managers with high stock-based
compensations and stock equity to manage earnings. More specifically, they emphasize
that those managers are more likely to report earnings that meet or beat analysts’ forecasts
and sell more shares in the year after earnings announcement. Bergstresser and Philippon
(2006) also find a significant association between equity incentives and abnormal accruals.
To test the relationship between discretionary accruals and compensation, detailed
bonus plans data should be available. Due to the unavailability of detailed compensation
data in the annual reports or in any other sources, in addition to the secrecy embedded in
the Egyptian disclosure environment (Dahawy and Conover, 2007; Dey et al., 2008), and
the absence of regulation that enforces disclosure of this information, it is not expected
that managers would disclose such information voluntarily. Therefore, following prior
studies (Young, 1998), executive ownership is used as a proxy for the compensation
objective. Executive equity ownership may reduce the underlying agency conflicts that
exist either between managers and outside shareholders or between controlling
shareholders and minority shareholders. According to this view, the more stocks
executives own, the greater their degree of managerial control and the stronger their
motivation to take actions that may lead to a lower earnings management (Warfield et al.,
1995). Hence, our testable hypothesis is formulated as follows:

H1. Earnings management is negatively related to managerial equity ownership.

 JAAR 2.2 Debt covenants hypothesis
15,1 Since debt agreements depend on accounting numbers reported in financial statements,
managers have the opportunity to choose accounting methods that allow them to avoid
violating these agreements. These contracts often include restrictive covenants that limit
potential conflicts of interest between firms’ debt holders and shareholders, as well as
restricting managers’ scope to engage in activities that may adversely affect the debt
104 holders’ wealth. These include limiting the ability of management to issue new debt and
giving the debt holders the right to demand early repayment of the debt when certain
accounting numbers are not maintained (Press and Weintrop, 1990; Duke and Hunt, 1990).
These studies provide evidence supporting the assertion that firms with debt covenants
based on accounting numbers may have greater incentives to conceal the firm’s real
economic performance and inflate reported earnings by, for example, engaging in income-
increasing accrual choices in an attempt to reduce the possibility of default.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

DeFond and Jiambalvo (1994) report that managers manipulate abnormal accruals
upward to increase the reported income in the year prior to violation and, to a lesser
extent, in the year of the covenant violation. Charitou et al. (2007) find a similar result
in one year prior to bankruptcy-filing. Likewise, Sweeney (1994) finds significantly
greater use of income-increasing accounting changes in defaulting firms relative to a
control sample, matched on industry, size, and time period. In addition, she
demonstrates that defaulting firms tend to undertake early adoption of new accounting
standards when these standards increase the reported net income. In a similar vein,
Healy and Palepu (1990) emphasize that firms which are close to default on their
dividend restriction are likely to reduce dividends payment and switch to income-
increasing accounting methods. However, DeAngelo et al. (1994) demonstrate that
managers of financially troubled firms which reduced dividends make income-
decreasing accounting decisions, even though dividends payments are under pressure
due to private debt agreements. Furthermore, they conclude that accounting choices
reflect the firms’ financial difficulties rather than attempts to avoid debt covenant
violation, or inflate reported income to disguise the financial difficulties. Similar
evidence is also found by Peltier-Rivest (1999) who shows that managers of troubled
firms with binding debt covenants do not adopt income-increasing accounting choices.
Thus, it is expected that managers of highly leveraged firms are likely to make
income-increasing accounting choices in an attempt to avoid such violation. This leads
into the following hypothesis:

H2. Earnings management is positively related to leverage.

2.3 Political costs hypothesis

Since large firms are usually more politically visible, abnormally large increases
in reported earnings may be used as an indicator of a monopoly, or used as an
excuse for political or governmental intrusions in their business’s affairs (Watts
and Zimmerman, 1990). Thus, managers of large firms are expected to have greater
incentives to make accounting choices that reduce the likelihood of these political
costs being incurred.
It is found that firms use income-decreasing discretionary accruals in industries
applying to the United States International Trade Commission for import relief
(Jones, 1991), and in firms under investigation for anti-trust dealings during the year of
the investigation (Cahan, 1992). Similar results are found in the cable television
 industry (Key, 1997), and in chemical firms exposed to the Superfund laws (Cahan et al., Managerial
1997; Johnston and Rock, 2005). accounting
In the oil industry, Han and Wang (1998) show that petroleum refining firms tend to
make negative discretionary accruals and report good news late in an attempt to discretion
reduce political costs, as early release of good news would attract additional public
attention which may increase their exposure to political actions. More recently, Byard
et al. (2007) support the political costs hypothesis for a sample of US-based oil 105
companies. Consistent with Han and Wang (1998), they find that large petroleum
refining firms engage in significant abnormal income-decreasing accruals in the 4th
fiscal quarter of 2005 immediately after the impact of hurricanes Katrina and Rita.
Similarly, Hall (1993) demonstrates that the increased scrutiny of oil firms are likely to
motivate managers to make more income-decreasing accounting changes in periods of
sharp oil price increases than in other periods. This leads to the following hypothesis:
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

H3. Earnings management is negatively related to firm size.

2.4 Earnings smoothing hypotheses

Earnings smoothing has been the subject of concern with regulatory and accounting
studies alike[5]. Managers may tend to use accounting discretion afforded by
accounting standards to reduce the fluctuations of earnings in an attempt to report a
less variable earnings stream and to show that the company has less risk (Fudenberg
and Tirole, 1995). Managers can do so by, for example, understating earnings in years
of high performance in order to create reserves for future periods (Leuz et al., 2003).
However, it is found that managers may engage in earnings smoothing even if
managerial compensation is not tied with earnings (Herrmann and Inoue, 1996).
It is found that managers smooth earnings in an attempt to reduce the possibility of
being dismissed (Fudenberg and Tirole, 1995; DeFond and Park, 1997). Trueman and
Titman (1988) provide evidence for use of earnings smoothing as a cost minimizing
device. They indicate that when a firm faces a high level of earnings volatility, the
possibility of bankruptcy will be greater. They argue that earnings smoothing serves to
influence shareholders’ perception of the stability of reported earnings and therefore their
assessment of the likelihood of firm bankruptcy. Earnings smoothing is also thought of
as an equilibrium solution to compensate informed managers for their information
advantages and for taking additional risks (Tucker and Zarowin, 2006). Developing an
analytical model to explain incentives of managers to smooth earnings, Goel and Thakor
(2003) find that the degree of earnings smoothing is higher for firms whose manager’s
compensation contract is tied to long-run performance, firms with higher uncertainty
about the earnings volatility, and firms characterized by diffuse ownership.
Although earnings smoothing behaviour is documented in several contexts, we
expect that earnings smoothing objectives may explain a larger amount of the
variation in accruals choices in Egypt than efficient contracting objectives. This
expectation stems from the premise that, in developed capital markets, suppliers
of capital commonly contract with managers in an attempt to prevent the diversion of
corporate resources to managers’ personal consumption. Contractual relationships
play a crucial role in aligning divergent objectives of various contracting parties,
reducing information asymmetry and encouraging managers to use the reporting
flexibility to improve reporting quality (Badertscher et al., 2012). However, in less
developed capital markets, the demand for accounting income is not expected to be as
 JAAR clear-cut. This is because the demand for accounting income is closely related to the
15,1 payout preferences of controlling shareholders rather than the demand for public
disclosure (Ball et al., 2000). Also, weak legal protection and lax oversight are more
likely to give greater discretion over earnings and allow managers to further reduce
income volatility.
More generally, Leuz et al. (2003) argue that reporting earnings with lower variance
106 creates opacity, which may help insiders to expropriate from outside investors. The
results of a survey conducted by Graham et al. (2005) show that managers believe firms
with smoother earnings are thought by investors to be less risky and associated with
lower cost of equity and expected return. Therefore, managers can smooth reported
earnings to accomplish several market and contracting motivations, including
managerial compensation (e.g. Trueman and Titman, 1988), political costs (e.g. Cahan,
1992; Godfrey and Jones, 2002), and to signal the financial stability of the firm to meet
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

interest commitments and avoid violation of debt covenants.

Prior studies such as Leuz et al. (2003) and Lang et al. (2006) measure earnings
smoothing as the ratio of standard deviation of operating income and the standard
deviation of operating cash flow (both scaled by lagged total assets). Based on the
preceding discussion, we test the following hypothesis:

H4.a Earnings management is negatively related to the ratio of standard deviation

of operating income and the standard deviation of operating cash flow.

The combination of cash flows from operations and accruals constitutes the level of
reported earnings. Kirschenheiter and Melumad (2002) show that the level of reported
earnings allows investors to infer the level of permanent future cash flows. Keeping
fluctuation to a minimum level, therefore, could improve investors’ expectations about
this important future component. Sloan (1996) finds that investors over-estimate the
persistence of accruals (i.e. as firms with relatively low (high) magnitudes of accruals
earn positive (negative) risk-adjusted returns). In response to this situation, firms facing
an increase (decrease) in operating cash flows may engage in income-decreasing
(increasing) accrual manipulation to maintain smoothed earnings. Although accruals
and cash flows are naturally negatively correlated (Dechow, 1994), larger association
may suggest greater earnings smoothing (Lang et al., 2006; Leuz et al., 2003).
Accordingly, the magnitude of discretionary accruals is expected to be greater (smaller)
for poor (good) cash flow firms. Accordingly, the following hypothesis is formulated:

H4.b Earnings management is negatively related to changes in cash flows from

operation.

3. Methodology
3.1 Proxies for earnings management
We employ the cross-sectional approach of the modified Jones model suggested
by Dechow et al. (1995), and the performance-adjusted Jones model suggested by
Kothari et al. (2005) to isolate discretionary accruals, which are used as proxies for
earnings management. This approach allows us to reduce the survivorship bias
problem inherent in time-series models and to overcome the problem of unavailability
of sufficient time-series data needed (at least nine years) to estimate firm-specific
coefficients, as well as relax the assumption that the estimated coefficients are
 stationary (Kothari et al., 2005). We measure discretionary accruals (DAC), when the Managerial
modified Jones model is used, as the residuals from the following industry-year model: accounting

discretion
TACCi;t =TAi;t1 ¼ a þ b1 DREVi;t  DRECi;t =TAi;t1

ð1Þ
þ b2 GPPEi;t =TAi;t1 þ ei;t
107
where TACC is the firm i’s total accruals, calculated as earnings before extraordinary
items and discontinued operations minus cash flow from operating activities, TA is
firm i’s book value of total assets in year t, DREV is firm i’s changes in net revenues
between year t1 and year t, DREC is firm i’s change in receivables between year t1
and year t, GPPE is firm i’s gross property, plant and equipment, b1, b2, and b3 are the
estimated parameters; e is the error term for firm, i is a firm indicator, and t is a time
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

indicator. The procedures of the performance-adjusted model are similar to that

explained above, with the addition of the contemporaneous return on assets (ROA),
measured as net income before extraordinary items to total assets, to Equation (1).
Two modifications to the original models are adopted. Since there is no particular
event to be examined, the first modification involves adjusting firm discretionary
accruals by subtracting the changes in accounts receivable from the changes in revenues
in the estimation period as in the test period (Kasznik, 1999), as ignoring effects of
receivables may reduce the power of the test (McNichols, 2000). Additionally, there is no
reason to think that earnings management is expected to be only in the test period
(McNichols, 2000). The second modification involves the inclusion of an intercept
without scaling by lagged total assets. This is because there is no theoretical reason for
forcing the regression through the origin, or to believe that total accruals will be zero
when changes in cash sales and gross property plant and equipment are zero.

3.2 Research design and empirical model

To test the trade-offs among multiple contracting and income smoothing objectives, we
include a set of explanatory variables related to determinants of discretionary accruals
choice. We model the contracting objectives as a function of bonus plans, debt
covenants, and political costs. Executive ownership, EXECOWN, defined as the
percentage of equity ownership owned by the CEO and executive directors to the total
shares outstanding, is used as a proxy for the compensation objective. It is argued that
leverage is positively related to both the existence of, and closeness to, accounting-
based debt covenants and debt-to-equity ratio captures the existence and tightness of
most common debt covenant restrictions (see Press and Weintrop, 1990; Duke and
Hunt, 1990, among others). We therefore use the ratio of total debts to net book value of
equity, DEBT/EQUITY, as a proxy of closeness to debt covenants violation[6].
In an attempt to test the political cost hypothesis, researchers usually focus on firm
characteristics, such as firm size. However, this proxy has faced much theoretical
criticism. For example, Watts and Zimmerman (1990) and Christie (1990) describe firm
size as a noisy proxy for political costs that may be used as a proxy for many effects
other than political cost. Moreover, while concentrating only on large firms may make
the test stronger and reduce test noise, it may weaken the power of the test as a result
of testing only small samples (Hall, 1993). We use firm size, SIZE (measured as the
natural logarithm end-year book value of total assets of a firm), as a proxy of political
costs and also to control for the correlation between size and accounting choice (Watts
and Zimmerman, 1990).
 JAAR Following Leuz et al. (2003) and Lang et al. (2006), we measure earnings smoothing,
15,1 SMOOTH, as the ratio of standard deviation of operating income and the standard
deviation of operating cash flow (both scaled by lagged total assets)[7]. We interpret a
low value of this measure as indicating that managers are more inclined to exercise
accounting discretion to smooth reported earnings. An earnings smoothness proxy is
calculated using rolling windows of three annual observations. Cash flow from operation
108 is computed directly from the statement of cash flows. Our analysis also incorporates
change in cash flow from operations, DCFO (defined as cash from operating activities in
the current year less cash from operating activities in the previous year), as a proxy for
implicit income smoothing inherent in accrual generation. Although the association
between accruals and cash flow from operation is naturally negative, a larger magnitude
of this association is more likely to indicate smoothing of reported earnings to conceal
the underlying corporate economic performance (Leuz et al., 2003; Lang et al., 2006).
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

The analysis also considers several other control variables. Two dummy variables are
used. First, CFOH, defined as a dummy variable that takes the value of one when the
cash flow from operations is included in the highest decile of cash flow from operations
and zero otherwise. Second, CFOL, defined as a dummy variable that takes the value
of one when the cash flow from operations is included in the lowest decile of cash flow
from operations and zero otherwise. Both dummy variables are used to control for
discretionary accruals measurement error, which is found to be negatively associated
with cash flow performance (Dechow et al., 1995; Young, 1999). In addition, two dummy
variables are used to control for abnormal reported earnings. First, EARNH, defined as a
dummy variable that takes the value of one when the reported earnings is included in the
highest decile of the reported earnings and zero otherwise. Second, EARNL, defined as a
dummy variable that takes the value of one when the reported earnings is included in the
lowest decile of the reported earnings and zero otherwise.
Additionally, the ratio of long-term assets to total assets is used as a proxy for assets
intensity, ASSINT, to control for the possible impact of the depreciation charge on
estimations of discretionary accruals (Young, 1998). Market-to-book ratio, MTBOOK
(measured as the ratio of book value of total assets minus the book value of equity plus the
market value of equity to book value of assets), is used as a proxy for growth opportunities
(Krishnan, 2003). Additionally, firms that constitute the Egyptian Exchange Index 30 may
have larger abnormal accruals because they possibly have the ability and resources to
boost the reported earnings through using, for example, discretionary accruals. EGX30 is
a dummy variable introduced in the analyses to control for this possibility. Finally, we
control for industry and time effects (not reported) using IndustryDum and TimeDum
as indicator variables. The following regression is used to test the hypotheses:
DAC i;t ¼ a þ b1 EXECOWN i;t þ b2 LEV i;t þ b3 SIZE i;t þ b4 DCFOi;t
þ b5 SMOOTH i;t þ b6 ASSINT i;t þ b7 CFOLi;t þ b8 CFOHi;t
þ b9 EARNLi;t þ b10 EARNHi;t þ b11 MTBOOKi;t þ b12 EGX 30i;t þ ei;t
where i and t are firm and time subscripts, respectively, e is an error term, and all other
variables are as defined in Table I.

4. Data and descriptive statistics

4.1 Data
For our empirical analysis, we use a sample of listed non-financial Egyptian firms over the
period (2005-2007). Accounting data are obtained from the Egyptian Exchange (EGX). Data
 Variable Definition
Managerial
accounting
Discretionary accruals variables
TACC The total accruals calculated as earnings before extraordinary
discretion
items and discontinued operations minus cash flow
from operating activities
CACC The current accruals calculated as the sum of changes in inventory,
accounts receivable, and other current assets less changes in
109
accounts payable, income taxes payable and other current liabilities
TA The book value of total assets
DREV The change in net revenues between the current year and prior year
DREC The change in receivables between the current year and prior year
GPPE The gross property, plant and equipment
ROA The return on total assets
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

Dependent variables
MJTDA The signed total discretionary accruals scaled by lagged total assets
as measured by the cross-sectional modified Jones model
PATDA The signed total discretionary accruals scaled by lagged total
assets as measured by the cross-sectional performance-adjusted
Jones model
MJCDA The signed current discretionary accruals scaled by lagged total
assets as measured by the cross-sectional modified Jones model
PACDA The signed current discretionary accruals scaled by lagged total
assets as measured by the cross-sectional performance-adjusted
Jones model
Independent variables
EXECOWN The percentage of equity ownership owned by CEO and executive
directors to the total shares outstanding
DET/EQUITY The ratio of total debts debt to book value of equity
SIZE The natural logarithm end-year book value of total assets of firm in
million (Egyptian) pounds
DCFO Change in cash from operations as measured by cash from operating
activities in the current year less cash from operating activities
in prior year
SMOOTH The ratio of standard deviation of operating income and the
standard deviation of operating cash flow (both scaled by lagged
total assets)
Control variables
ASSINT The ratio of long-term assets to total assets
CFOL A dummy variable that takes the value of one when the cash flow
from operations is included in the lowest decile (the extreme low
CFO) of cash flow from operations and zero otherwise
CFOH A dummy variable that takes the value of one when the cash flow
from operations is included in the highest decile (the extreme high
CFO) of cash flow from operations and zero otherwise
EARNL A dummy variable that takes the value of one when the reported
earnings are included in the lowest decile (the extreme low reported
earnings) of reported earnings and zero otherwise
EARNH A dummy variable that takes the value of one when the reported
earnings are included in the highest decile (the extreme high
reported earnings) of reported earnings and zero otherwise
MTBOOK The ratio of book value of total assets minus the book value of
equity plus the market value of equity to book value of assets
EGX30 A dummy variable that takes the value of one when the firm is one Table I.
of the EGX30 companies and zero otherwise Variables definitions
 JAAR on ownership structure are collected from the Egypt for Information Dissemination (EGID)
15,1 and the CMA. The market values of equity are extracted from monthly and annually
bulletins issued by EGID (various issues). Several screening criteria were applied to the data
before carrying out the empirical analysis. First, for the purpose of discretionary accruals’
estimations, we chose firms with no missing data over the period (2004-2007)[8]. Second,
firms should not be involved in merger or acquisition events as these firms tend to be larger
110 for reasons other than earnings management behaviour. Third, firms should not belong to
the financial or regulated sectors as their disclosure requirements and accruals generation
are different from those of other firms. In addition, regulation of these firms makes their
accounting information incomparable to that in other industries and earnings management
incentives differ from those of unregulated industries. Finally, we cleared outliers and
potential data error in the data set by excluding the values of each variable that lie outside
the first and the 99th percentiles. This process yields a final sample of 438 observations.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

4.2 Descriptive statistics

Table II reports summary descriptive statistics for the main variables used in the
analysis. The table indicates that the average (median) abnormal accrual as a
percentage of beginning total assets is 0.0000 (0.007) using the modified Jones model
and that they are qualitatively similar when the performance-adjusted model is used.
On average, the executive directors own 10.6 per cent of firm shares. In addition, the
leverage ratio, on average, is 41.1 per cent.
The Pearson correlation matrix in Table III shows that correlation coefficients seem
reasonable. For example, larger companies are more likely to have a large ratio of debt-
to-equity and their motivations to smooth earnings are higher than in small companies.
Also, most of the firms that constitute EGX30 are large firms[9].
5. Empirical results
5.1 Univariate analysis
The univariate analysis includes a mean (median) comparison test using t-test
(Wilcoxon-Mann-Whitney). The samples of discretionary accruals are designed on the

Variable Mean SD Q1 Median Q3 Min. Max.

TACC/lagTA 0.007 0.115 0.050 0.001 0.062 0.389 0.551

CACC/lagTA 0.023 0.122 0.081 0.024 0.037 0.437 0.551
D(REV-REC)/lagTA 0.068 0.437 0.026 0.044 0.125 1.076 7.896
GPPE/lagTA 0.678 0.438 0.365 0.678 0.979 0.001 3.101
lagROA 0.084 0.133 0.029 0.072 0.128 0.569 1.808
ROA 0.092 0.110 0.033 0.087 0.143 0.569 0.471
MJTDA 0.000 0.105 0.058 0.007 0.05 0.412 0.476
PATDA 0.000 0.099 0.058 0.006 0.052 0.395 0.434
MJCDA 0.000 0.107 0.060 0.006 0.053 0.429 0.485
PACDA 0.000 0.102 0.062 0.002 0.059 0.413 0.441
EXECOWN 0.106 0.237 0.000 0.000 0.060 0.000 1.000
DEB/EQUITY 0.411 1.262 0.000 0.131 0.520 7.197 19.603
SIZE 12.735 1.562 11.712 12.632 13.782 9.057 17.965
DCFO 0.016 0.142 0.040 0.017 0.080 0.953 1.015
SMOOTH 0.949 0.763 0.432 0.758 1.250 2.506 3.939
Table II. Notes: This table shows the descriptive statistics for 438 observations used in the analyses over the
Descriptive statistics period 2005-2007. Definitions for all variables are provided in Table I
 Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

DEBT/
EXECOWN QUITY SIZE DCFO SMOOTH ASSINT CFOL CFOH EARNL EARNH MTBOOK EGX30

EXECOWN 1.000
DEBT/QUITY 0.010 1.000
SIZE 0.007 0.115*** 1.000
DCFO 0.023 0.006 0.041 1.000
SMOOTH 0.009 0.059 0.044** 0.029 1.000
ASSINT 0.086* 0.034 0.040 0.062 0.062 1.000
CFOL 0.023 0.078 0.131*** 0.370*** 0.024 0.042 1.000
CFOH 0.101** 0.034 0.156*** 0.356*** 0.116** 0.128*** 0.110** 1.000
EARNL 0.014 0.030 0.235*** 0.020 0.034 0.112***** 0.131*** 0.110** 1.000
EARNH 0.039 0.059 0.614*** 0.090* 0.018 0.172*** 0.081* 0.242*** 0.108** 1.000
MTBOOK 0.062 0.044 0.156*** 0.001 0.023 0.055 0.053 0.131*** 0.032 0.126*** 1.000
EGX30 0.057 0.005 0.418*** 0.029 0.009 0.097* 0.015 0.065 0.106** 0.349*** 0.244*** 1.000
Notes: Definitions for all variables are provided in Table I. *,**,***Indicate that correlation is significant at the 10, 5 and 1 per cent level, respectively
accounting
Managerial

Pearson correlation matrix

discretion

Table III.
111
 JAAR basis of the median of each explanatory variable in the case of scale variables or using
15,1 the two categories in the case of dichotomous variables. These tests aim to show
whether the mean of discretionary accruals measures differs across the two categories
of each explanatory variable. For example, it is hypothesized that there is a significant
difference in terms of executive directors’ ownership, debt-to-equity ratio, firm size and
earnings smoothing between firms in the above and below median subsamples.
112 The results in Table IV show that firms with negative cash flow changes make
significantly higher discretionary accruals compared to those with positive cash flow
changes. In contrast, the results do not support the managerial ownership, leverage
and political costs hypotheses for any earnings management proxy. In summary, the
univariate analysis provides evidence that accruals choices in Egypt are likely to be
driven by income smoothing objectives and contracting objectives explain little of the
variations in abnormal accruals.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

5.2 Regression results

The results of univariate analysis reveal a weak association between earnings
management and all efficient contracting objectives. The univariate analysis, however,
does not control for the effects of other variables that may be related to discretionary
accruals or to other efficient contracting objectives, which may result in potential effects,
confounding the earnings management-efficient contracting objectives relationship.
Therefore, ordinary least square (OLS) regression with robust standard errors to
correct for heteroskedasticity is employed to test the trade-offs between efficient
contracting and earning smoothing incentives. In Table V, we use total discretionary
accruals as measured by the modified Jones model (MJTDA) and the performance-
adjusted Jones model (PATDA) as our dependent variables. We start the multivariate
analysis with Model 1, in which only the control variables are included. In order to judge
the marginal predictive power of explanatory variables in determining discretionary
accruals choices, all independent and control variables are included in Model 2. In
addition, a vector of industry dummies IndustryDum, and time dummies TimeDum are
incorporated to control for industry-fixed effects and year-fixed effects, respectively.
In general, the coefficients of most variables are in line with their predicted signs.
However, in contrast to our expectations, the results in Table V reveal that the estimated
coefficients of EXWCOWN, SIZE, and DEBT/EQUITY are not significant. These results
provide no support for the executive bonus plans, political costs, or debt covenants
hypotheses. This implies that large firms are less likely to make income-decreasing choices
in an attempt to increase their compensation or to reduce the likelihood of governmental
intrusions in firm affairs. This result is inconsistent with the debt covenants hypothesis,
suggesting that managers of highly leveraged firms are unlikely to make income-
increasing accounting choices in an attempt to prevent debt covenants violation.
However, both income smoothing variables are highly significant for both
alternative discretionary accruals estimation models. More specifically, the coefficient
of CFO is negative and statistically significant at the 1 per cent level for both
discretionary accruals proxies. This result implies that firms with positive cash flow
changes are more likely to manipulate earnings downward to adversely affect the
reporting earnings level in order to smooth the reporting earnings and reduce their
fluctuations at minimum levels. This negative association is consistent with prior
studies (e.g. DeFond and Jiambalvo, 1994). Similarly, the coefficient of SMOOTH is
significantly negative at the 1 per cent level for the two earnings management proxies.
Managers, therefore, are more likely to exercise accounting discretion using the accrual
 Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

The modified Jones total discretionary accruals model The performance-adjusted Jones total discretionary accruals model
MJTDA mean MJTDA mean MJTDA mean MJTDA mean
above variable below variable Mann- above variable below variable Mann-
median median t-value Whitney median median t-value Whitney

EXECOWN 0.0041 0.0036 1.203 1.077 0.0070 0.0025 0.8076 0.645

DEBT/QUITY 0.0039 0.0035 0.750 0.887 0.0040 0.0035 0.7431 0.609
SIZE 0.0065 0.0069 1.345 0.504 0.0068 0.0072 1.3712 0.561
DCFO 0.038 0.039 10.155*** 10.030*** 0.037 0.037 10.165*** 10.002***
SMOOTH 0.005 0.004 4.231*** 5.160*** 0.0071 0.0067 3.3553*** 4.141***
Notes: Definitions for all variables are provided in Table I. *,**,***Indicate statistical significance at the 10, 5 and 1 per cent levels, respectively
accounting
Managerial

Univariate analysis
Table IV.
discretion

113
 JAAR MJTDA PATDA
15,1 Pred. Sign Model 1 Model 2 Model 3 Model 4 VIF

Constant þ / 0.001 0.004 0.000 0.020

(0.05) (0.09) (0.02) (0.40)
EXECOWN  0.024 0.014 1.11
114 (1.49) (0.90)
DEBT/QUITY þ 0.001 0.001 1.11
(0.98) (1.59)
SIZE  0.001 0.002 2.12
(0.19) (0.52)
DCFO  0.353*** 0.353*** 1.41
(9.35) (9.02)
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

SMOOTH  0.003*** 0.003*** 1.06

(2.69) (2.60)
ASSINT þ 0.001 0.001 0.002 0.002 1.38
(0.05) (0.06) (0.15) (0.15)
CFOL þ 0.097*** 0.037** 0.095*** 0.036*** 1.31
(4.84) (2.04) (4.69) (2.93)
CFOH  0.089*** 0.035** 0.099*** 0.045*** 1.33
(4.70) (2.29) (4.98) (2.73)
EARNL  0.114*** 0.104*** 0.109*** 0.098*** 1.14
(6.90) (7.27) (6.31) (6.36)
EARNH þ 0.037 0.033 0.033 0.025 2.03
(1.63) (1.33) (1.41) (0.98)
MTBOOK þ 0.002 0.001 0.002 0.001 1.16
(0.49) (0.28) (0.52) (0.26)
EGX30 þ 0.017 0.024 0.019 0.024 1.52
(1.01) (1.39) (1.08) (1.34)
Industry dummies Included Included Included Included
Time dummies Included Included Included Included
Table V. No of observations 438 438 438 438
Regressions of total Adj. R2 0.1900 0.3629 0.1830 0.3476
discretionary accruals on
the contracting incentives, Notes: Definitions for all variables are provided in Table I. t-statistics in parentheses. For the
earnings smoothing and estimation the consistent to heteroskedasticity standard errors has been used. *,**,***Indicate
control variables statistical significance at the 10, 5 and 1 per cent levels, respectively

component in order to smooth reported earnings in an attempt to reduce the variability

of earnings by altering discretionary accruals. Therefore, the results reveal that none of
the contracting hypotheses are confirmed. In contrast, income smoothing hypotheses
are accepted irrespective of the measure of abnormal accruals. This result ties in
closely with findings in, Ball et al. (2000), Leuz et al. (2003), and Lang et al. (2006),
suggesting that pervasiveness of earnings management is more apparent in countries
with concentrated ownership, weak investor rights and legal enforcement.
The weak evidence for effects of contracting objectives in the Egyptian context may
be captured by the strong effects of income smoothing. This means that income
smoothing may be seen as a method by which managers can reduce the variability of
the reported earnings, either to gain personal advantage or attain some contractual
objectives. Doing so is likely to increase the likelihood of keeping their jobs (Fudenberg
and Tirole, 1995), decrease the probability of political and governmental intervention
as well as increase managerial compensation, which in turn can help to signal their
ability to the capital market and build their reputation. Furthermore, management may
 tend to use income smoothing as a signal to convey private information about the Managerial
firm’s cash flow and future profitability. accounting
Another possible explanation for earnings smoothing is based on the results of Goel
and Thakor (2003) who demonstrate that earnings smoothing may not be determined discretion
by self-interest or leverage concerns, but is driven by managers’ efforts to increase the
firm’s stock price by reducing the losses shareholders may bear when they are forced to
trade for liquidity reasons. Goel and Thakor (2003) argue that earnings smoothing may 115
be desirable by uninformed shareholders who trade for liquidity reasons because any
increase in the volatility of reported earnings are likely to increase their trading losses.
Since investors are less likely to pay for firms with high earnings volatility, managers
may respond and prohibit speculators from acquiring private information that could
be used to trade against uninformed shareholders.
Collectively, the results of the regression analysis lend credence to the idea that the
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

traditional costly contracting incentives provide little explanation for discretionary

accruals choices in Egypt, while income smoothing activity explains much of the cross-
sectional variation in managerial choices.

6. Additional tests
6.1 Alternative discretionary accruals proxies: current discretionary accruals
It is widely believed that the scope for manipulating non-current accruals (i.e. non-
working capital accruals) is relatively limited for management because they can exercise
more discretion over the choice of regular revenue and expense items. Therefore, it is
expected that the (current) working capital discretionary accruals component is an
effective device for managers to manipulate earnings without being easily detected
(DeFond and Jiambalvo, 1994; Teoh et al., 1998a, b). To assess whether previous results
are sensitive to the measure of earnings management, the statistical procedures were
repeated using only the modified Jones current discretionary accruals (MJCDA) and the
performance-adjusted current discretionary accruals (PACDA) models. The current
accruals were calculated as the sum of changes in inventory, accounts receivable, and
other current assets less changes in accounts payable, income taxes payable and other
current liabilities. The results in Table VI, show Models 2 and 4 are qualitatively similar
to those reported using total discretionary accruals models at a relatively higher R2. The
results again confirm the highly negative association between discretionary accruals and
income smoothing hypotheses and reveal no evidence to support the bonus plans, the
political costs, or the debt covenants hypotheses.

6.2 Managerial ownership: piecewise and non-linearity tests

Two alternative additional tests were performed to investigate the possibility of a non-
linear relationship between executive ownership and discretionary accruals similar to
that documented by Teshima and Shuto (2008). First, we used piecewise linear models
as EXECOWN is decomposed as follows:

EXECOWNL ¼ EXECOWN if EXECOWN o5%; and ¼ 5% if

EXECOWNX5%; EXECOWNM ¼ 0 if EXECOWN o5%;
and ¼ EXECOWN  5% if 5%pEXECOWN o25%;
EXECOWNH ¼ 0 if EXECOWN o25%; and ¼ EXECOWN  25%
if EXECOWN X25%:
 JAAR MJCDA PACDA
15,1 Pred. Sign Model 1 Model 2 Model 3 Model 4

Constant þ / 0.005 0.006 0.005 0.021

(0.51) (0.13) (0.48) (0.44)
EXECOWN  0.020 0.010
116 (1.38) (0.75)
DEBT/QUITY þ 0.001 0.001
(1.21) (1.56)
SIZE  0.000 0.002
(0.07) (0.44)
DCFO  0.300*** 0.304***
(9.40) (9.05)
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

SMOOTH  0.003*** 0.003***

(2.82) (2.63)
ASSINT þ 0.012 0.013 0.013 0.013
(0.97) (1.06) (0.94) (0.98)
CFOL þ 0.115*** 0.065*** 0.111*** 0.060***
(6.35) (3.90) (5.94) (3.50)
CFOH  0.112*** 0.067*** 0.119*** 0.073***
(6.53) (4.67) (6.58) (4.73)
EARNL  0.074*** 0.067*** 0.072*** 0.063***
(4.72) (4.53) (4.30) (3.97)
EARNH þ 0.020 0.017 0.017 0.011
(0.91) (0.75) (0.76) (0.44)
MTBOOK þ 0.001 0.001 0.001 0.001
(0.11) (0.19) (0.18) (0.15)
EGX30 þ 0.022 0.028* 0.023 0.027*
(1.49) (1.87) (1.47) (1.71)
Industry dummies Included Included Included Included
Time dummies Included Included Included Included
Table VI. No of observations 438 438 438 438
Regressions of current Adj. R2 0.2339 0.3755 0.2227 0.3566
discretionary accruals on
the contracting incentives, Notes: Definitions for all variables are provided in Table I. t-statistics in parentheses. For the
earnings smoothing and estimation the consistent to heteroskedasticity standard errors has been used. *,**,***Indicates
control variables statistical significance at the 10, 5 and 1 per cent levels, respectively

In unreported tests[10], the results confirm those previously documented. More

specifically, the coefficient of the intermediate ownership level is negative and none of
the ownership terms are significant. Second, regression models are re-estimated after
including a squared term of ownership. The results also show that the coefficients of
both ownership variables are not significant for both proxies, suggesting no evidence
of a non-linear relationship between managerial ownership and discretionary accruals.
Furthermore, the results do not support any of the traditional contracting hypotheses.
However, both measures of earnings smoothing are negative and highly significant at
the 1 per cent level.

6.3 Prior period discretionary accruals

Since discretionary accruals revert over the firm’s lifetime (Dechow et al., 2012; Dechow,
1994), the discretionary accruals in any period consist of the initial discretionary accrual
in that period plus portions of prior periods (McNichols, 2000). However, the ability of
 managers to inflate the current period’s reported earnings will undoubtedly shrink, Managerial
as the level of lagged total accruals rises (Koh, 2007). Hence, the failure to control accounting
for reversal of prior years’ accruals may lead to seriously invalid conclusions
(Kasznik, 1999). As a result, the relationship between current period discretionary discretion
accruals and lagged accruals is expected to be significantly negative. To control for
the effects of accruals reversal, we include in the empirical analysis lagged total
discretionary LAGDA. 117
We find that discretionary accruals are subject to short term reversal, although
the results confirm previous findings concerning the income smoothing hypotheses.
To examine the possibility that the results reported earlier are not driven by accruals
reversal, the regressions are re-estimated after including the interactions between LAGDA
and efficient contracting and income smoothing explanatory variables. One would observe
significant coefficients on interacting variables if the accruals reversal had significant
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

effects on the prior results. Our findings reveal that income smoothing variables remain
highly significant and all coefficients of interactions are not significant. However, the debt
covenant hypothesis is only confirmed at the 5 per cent level. These results provide
evidence that the initial findings are robust, even after taking the effects of accruals
reversal into consideration.

6.4 Size effect

The results documented above provide little support for the political costs hypothesis.
Sloan (1996) finds a non-linear relationship between total accruals and firm size. To allow
for this possibility, the statistical analysis was repeated after adding a squared term for
firm size. Our analysis shows no indication of such a relationship. Another point to
note is that the smoothing variables are still statistically significant at the 1 per cent level.
It appears from the correlation matrix that SIZE is highly correlated with other
explanatory variables, suggesting a need to investigate whether the earlier results will
be affected exclusive of SIZE or using a different proxy for size, (defined as the natural
logarithm of sales). The findings are similar to those reported above and the results reveal
no evidence of a non-linear relationship between earnings management and firm size.
Despite the careful treatment of the variables used in the analysis and the
methodology adopted, the results of this study are subject to some caveats. First,
as in any accruals-based earnings management study, a key concern regarding the
explanation of results relies on the ability of earnings management proxies to
adequately capture earnings manipulation activities. It is well-known that measurement
errors related to abnormal accruals measurement are a concern. Although alternative
discretionary accruals models and different measurement error-related variables are
used, the findings are still not totally free of this concern. Thus, we cannot rule out
the possibility that our results are influenced by omitted variables. Second, our efficient
contracting objectives and income smoothing objectives may exhibit considerable
measurement error, which may bias the magnitude of the estimated effects. For
example, if leverage measures closeness to covenants violation with error, the
parameter estimates using these surrogates may be biased and inconsistent. Third,
corporate governance mechanisms such as board composition and ownership audit
quality (which monitor managerial opportunism), are not included in this study. This is
because the main objective of this study is to document only the trade-offs between
contacting and income smoothing objectives. Thus, the effectiveness of corporate
governance mechanisms in constraining opportunistic earnings management is left
for future investigation.
 JAAR 7. Concluding remarks
15,1 This study examines whether discretionary accruals choices can be explained by the
costly contracting incentives, as well as income smoothing. Accounting discretion has been
modelled as a function of two competing accounting choices incentives: efficient contracting
(i.e. bonus plans, debt covenants, and political costs), and income smoothing. The modified
Jones and performance-adjusted Jones models are used to isolate the discretionary accruals
118 component. Based on 438 non-financial Egyptian observations over the period (2005-2007),
the results indicate that the associations between the measures of earnings management
and contracting variables are not significant. Overall, the results of regression analysis
lend support to the notion that the traditional costly contracting incentives provide little
explanation for discretionary accruals choices in Egypt, while income smoothing activity
explains much of the cross-sectional variation in managerial choices.
These findings are in contrast with studies that test only one reporting objective
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

at a time. More specifically, managers tend to reduce the fluctuations in reported

earnings by increasing (decreasing) earnings when earnings are low (high) in an
attempt to reduce the variability of the reported earnings. Such smoothing behaviour is
likely to help managers retain their position, decrease the probability of political and
governmental intervention, and increase their compensation, which in turn can help to
signal their ability to the capital market and build their reputation. Furthermore,
management may tend to use income smoothing as a signal to convey private
information about the firm’s cash flow and future profitability.
The findings of our study should be of substantial interest to regulators and policy
makers. The results implicitly contribute to the ongoing arguments in relation to the
optimal flexibility permitted by standard setters and the reduction in permissible
accounting treatments in order to improve reporting quality and reduce opportunistic
earnings management. Many of the weaknesses related to corporate reporting in
emerging countries may result from the inadequate enforcement of company law, as
well as the weak legal protection of minority shareholders. Our results highlight the
crucial role of understanding the reporting incentives in such an environment. There is
a need to put greater emphasis on proper enforcement and protecting minority
shareholders’ rights, e.g. by adopting cumulative voting to give minority shareholders
the chance to elect their own representatives.
Several avenues for future research exist. First, whereas the focus of the current study
is restricted to the contracting and income smoothing objectives, a more comprehensive
approach is needed to include and test more multiple (even conflicting) motivations,
especially those related to the equity market. Second, despite the complexities and
difficulties to develop such a model, progress towards providing a comprehensive model
that explains accounting choices would be valuable. Finally, there is a need to refine the
discretionary accruals models and develop a generally accepted model to appropriately
isolate the discretionary accruals component and/or focus on alternative measures of
earnings management, such as those related to real earnings management activities.

Notes
1. Section 2 provides more discussion for studies related to these objectives.
2. See Fields et al. (2001) and Dechow et al. (2010) for a survey of research.
3. In essence, using the single account approach is problematic for three reasons (McNichols, 2000).
First, the explanatory power is expected to be low as it is not clear which accrual managers use
to manipulate earnings. In addition, the validity of this approach tends to be reduced when the
 aim is to identify the magnitude of manipulation rather than testing factors associated Managerial
with a specific accrual. Thus, the feasibility of employing such an approach is questionable,
since an individual model is required for each accrual used to manipulate earnings. Second, accounting
the generalizability of findings of this approach might be limited due to the small number discretion
of firms for which a specific accrual is manipulated. Finally, earnings management is
associated with aggregate accounting adjustments rather than the choice of specific accruals
(DeAngelo, 1988).
4. The terms discretionary accruals and abnormal accruals are used interchangeably.
119
5. Earnings smoothing is a special case of earnings management involving intertemporal
smoothing of reported earnings relative to economic earnings to reduce the variability of earnings
over time (Goel and Thakor, 2003). It is important to note that earnings smoothing can be
achieved through real activities, real smoothing, or the reporting flexibility provided by GAAP
through accruals, artificial smoothing. While the former reduces volatility by directly affecting the
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

distribution of underlying cash flows, the latter directly affects only earnings volatility. Because
real smoothing has obvious costs and artificial smoothing costs are unobservable, it is less costly
for management to smooth earnings through accruals (Pincus and Rajgopal, 2002; Goel and
Thakor, 2003).Therefore, earnings smoothing in this study is related to artificial smoothing.
6. Due to the high costs of accessing actual debt covenant information, previous accounting
studies use leverage as a surrogate for the possibility of violating accounting based debt
covenants. The variables commonly used in prior studies as a proxy for existence and
tightness of covenant restrictions (i.e. leverage) are the debt/equity ratio, total debts to total
assets, long-term debt to total assets, and total liabilities to total assets. In the presence of
secrecy imbedded in the Egyptian environment and the lax oversight, obtaining such data is
very difficult. In addition, with no legal obligation to disclose such data, it is not expected
that managers voluntarily disclose such sensitive data.
7. Using this proxy assumes that cash flow is free of manipulation, although real activities
manipulation affects cash ows (e.g. Roychowdhury, 2006).
8. It is worth noting that dropping firms with missing data might induce a size bias in the sample.
Against this concern, the size of firms included in the final sample is compared with that of
firms that have missing data. The results of t-test comparison reveal no statistical significance
between the two groups of firms, which mitigates the concern of selection bias in the sample.
9. It is evident that relatively high correlations among some explanatory variables raise
econometric concern about the possible impact of collinearity on the drawn inferences.
Variance Inflation Factor (VIF) scores and condition indices are calculated to ensure that the
sample did not suffer from possible harmful collinearity. Belsley et al. (1980) suggest that a
condition index 415 signifies a possible problem and in excess of 30 suggests potentially
severe collinearity among the explanatory variables. Since the highest VIF score (2.12) is less
than ten, multicollinearity is not a problem in this study.
10. These results are not reported for the sake of brevity, but are available from the authors
upon request.

References
Badertscher, B., Collins, D. and Lys, T. (2012), “Discretionary accounting choices and the
predictive ability of accruals with respect to future cash flows”, Journal of Accounting and
Economics, Vol. 53 Nos 1-2, pp. 330-352.
Ball, R., Kothari, S. and Robin, A. (2000), “The effect of international institutional factors on properties
of accounting earnings”, Journal of Accounting and Economics, Vol. 29 No. 1, pp. 1-51.
Belsley, D., Kuh, E. and Welsch, R. (1980), Regression Diagnostics, Wiley, New York, NY.
 JAAR Bergstresser, D. and Philippon, T. (2006), “CEO incentives and earnings management”, Journal of
Financial Economics, Vol. 80 No. 3, pp. 511-529.
15,1
Byard, D., Hossain, M. and Mitra, S. (2007), “US oil companies’ earnings management in response
to hurricanes Katrina and Rita”, Journal of Accounting and Public Policy, Vol. 26 No. 6,
pp. 733-748.
Cahan, S. (1992), “The effect of antitrust investigations on discretionary accruals: a refined test of
120 the political-cost hypothesis”, The Accounting Review, Vol. 67 No. 1, pp. 77-95.
Cahan, S., Chavis, B. and Elmendorf, R. (1997), “Earnings management of chemical firms in
response to political costs from environmental legislation”, Journal of Accounting,
Auditing and Finance, Vol. 12 No. 1, pp. 37-65.
Charitou, A., Lambertides, N. and Trigeorgis, L. (2007), “Managerial discretion in distressed
firms”, The British Accounting Review, Vol. 39 No. 4, pp. 323-346.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

Cheng, Q. and Warfield, T.D. (2005), “Equity incentives and earnings management”, The
Accounting Review, Vol. 80 No. 2, pp. 441-476.
Christie, A. (1990), “Aggregation of test statistics: an evaluation on the evidence on contracting
and size hypotheses”, Journal of Accounting and Economics, Vol. 12 Nos 1-3, pp. 15-36.
Dahawy, K. and Conover, T. (2007), “Accounting disclosure in companies listed on the Egyptian
Stock Exchange”, Middle Eastern Finance and Economics, No. 1, pp. 5-20.
Darrough, M., Pourjalali, H. and Saudagaran, S. (1998), “Earnings management in Japanese
companies”, International Journal of Accounting, Vol. 33 No. 3, pp. 313-334.
DeAngelo, L. (1988), “Discussion of evidence of earnings management from the provision for bad
debts”, Journal of Accounting Research, Vol. 26, Supplement, pp. 32-40.
DeAngelo, H., DeAngelo, L. and Skinner, D. (1994), “Accounting choice in troubled companies”,
Journal of Accounting and Economics, Vol. 17 Nos 1-2, pp. 113-143.
Dechow, P. (1994), “Accounting earnings and cash flows as measures of firm performance: the
role of accounting accruals”, Journal of Accounting and Economics, Vol. 18 No. 1, pp. 3-42.
Dechow, P., Sloan, R. and Sweeney, A. (1995), “Detecting earnings management”, The Accounting
Review, Vol. 70 No. 2, pp. 193-225.
Dechow, P., Ge, W. and Schrand, G. (2010), “Understanding earnings quality: a review of the
proxies, their determinants and their consequences”, Journal of Accounting and
Economics, Vol. 50 Nos 2-3, pp. 344-401.
Dechow, P., Hutton, A., Kim, J. and Sloan, R. (2012), “Detecting earnings management: a new
approach”, Journal of Accounting Research, Vol. 50 No. 2, pp. 275-334.
DeFond, M. and Jiambalvo, J. (1994), “Debt covenant violation and manipulation of accruals”,
Journal of Accounting and Economics, Vol. 17 Nos 1-2, pp. 145-176.
DeFond, M. and Park, C. (1997), “Smoothing income in anticipation of future earnings”, Journal of
Accounting and Economics, Vol. 23 No. 2, pp. 115-139.
Dey, C., Grinyer, J., Sinclair, C. and El-Habashy, H. (2008), “Determinants of accounting choices in
Egypt”, Journal of Applied Accounting Research, Vol. 8 No. 3, pp. 48-92.
Duke, J. and Hunt, H. (1990), “An empirical examination of debt covenant restrictions and accounting-
related debt proxies”, Journal of Accounting and Economics, Vol. 12 Nos 1-3, pp. 45-63.
Fields, T., Lys, T. and Vincent, L. (2001), “Empirical research on accounting choice”, Journal of
Accounting and Economics, Vol. 31 Nos 1-3, pp. 255-307.
Fudenberg, D. and Tirole, J. (1995), “A theory of income and dividend smoothing based on
incumbency rents”, Journal of Political Economy, Vol. 103 No. 1, pp. 75-93.
Godfrey, J. and Jones, K. (2002), “Political cost influences on income smoothing via extraordinary
item classification”, Accounting & Finance, Vol. 39 No. 3, pp. 229-253.
 Goel, A. and Thakor, A. (2003), “Why do firms smooth earnings?”, The Journal of Business, Managerial
Vol. 76 No. 1, pp. 151-192.
accounting
Graham, J., Harvey, C. and Rajgopal, S. (2005), “The economic implications of corporate financial
reporting”, Journal of Accounting and Economics, Vol. 40 No. 1, pp. 3-73. discretion
Guthrie, K. and Sokolowsky, J. (2010), “Large shareholders and the pressure to manage
earnings”, Journal of Corporate Finance, Vol. 16 No. 3, pp. 302-319.
Hall, S. (1993), “Political scrutiny and earnings management in the oil refining industry”, Journal 121
of Accounting and Public Policy, Vol. 12 No. 4, pp. 325-351.
Han, J. and Wang, S. (1998), “Political costs and earnings management of oil companies during
the 1990 persian Gulf crisis”, The Accounting Review, Vol. 73 No. 1, pp. 103-117.
Healy, P. (1985), “The effect of bonus schemes on accounting decisions”, Journal of Accounting
and Economics, Vol. 7 Nos 1-3, pp. 85-107.
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

Healy, P. and Palepu, K. (1990), “Effectiveness of accounting-based dividend covenants”, Journal

of Accounting and Economics, Vol. 12 Nos 1-3, pp. 97-124.
Heflin, F., Kwon, S., Wild, J. and Building, R. (2002), “Accounting choices: variation in managerial
opportunism”, Journal of Business Finance & Accounting, Vol. 29 Nos 7-8, pp. 1047-1078.
Herrmann, D. and Inoue, T. (1996), “Income smoothing and incentives by operating condition: an
empirical test using depreciation changes in Japan”, Journal of International Accounting,
Auditing and Taxation, Vol. 5 No. 2, pp. 161-177.
Holthausen, R., Larcker, D. and Sloan, R. (1995), “Annual bonus schemes and the manipulation of
earnings”, Journal of Accounting and Economics, Vol. 19 No. 1, pp. 29-74.
Johnston, D. and Rock, S. (2005), “Earnings management to minimize superfund clean-up and
transaction costs”, Contemporary Accounting Research, Vol. 22 No. 3, pp. 617-642.
Jones, J. (1991), “Earnings management during import relief investigations”, Journal of
Accounting Research, Vol. 29 No. 2, pp. 193-228.
Kasznik, R. (1999), “On the association between voluntary disclosure and earnings
management”, Journal of Accounting Research, Vol. 37 No. 1, pp. 57-81.
Key, K. (1997), “Political cost incentives for earnings management in the cable television
industry”, Journal of Accounting and Economics, Vol. 23 No. 3, pp. 309-337.
Kirschenheiter, M. and Melumad, N. (2002), “Can ‘big bath’ and earnings smoothing co exist as
equilibrium financial reporting strategies?”, Journal of Accounting Research, Vol. 40 No. 3,
pp. 761-796.
Koh, P. (2007), “Institutional investor type, earnings management and benchmark beaters”,
Journal of Accounting and Public Policy, Vol. 26 No. 3, pp. 267-299.
Kothari, S., Leone, A. and Wasley, C. (2005), “Performance matched discretionary accrual
measures”, Journal of Accounting and Economics, Vol. 39 No. 1, pp. 163-197.
Kothari, S., Lys, T., Smith, C. and Watts, R. (1988), “Auditor liability and information disclosure”,
Journal of Accounting, Auditing & Finance, Vol. 3 No. 4, pp. 307-339.
Krishnan, G. (2003), “Does Big 6 auditor industry expertise constrain earnings management?”,
Accounting Horizons, Vol. 17, Supplement, pp. 1-17.
Lang, M., Smith Raedy, J. and Wilson, W. (2006), “Earnings management and cross listing: are
reconciled earnings comparable to US earnings?”, Journal of Accounting and Economics,
Vol. 42 No. 1, pp. 255-283.
Leuz, C., Nanda, D. and Wysocki, P. (2003), “Earnings management and investor protection:
an international comparison”, Journal of Financial Economics, Vol. 69 No. 3, pp. 505-527.
McNichols, M. (2000), “Research design issues in earnings management studies”, Journal of
Accounting and Public Policy, Vol. 19 Nos 4-5, pp. 313-345.
 JAAR Peltier-Rivest, D. (1999), “The determinants of accounting choices in troubled companies”,
Quarterly Journal of Business and Economics, Vol. 38 No. 4, pp. 28-44.
15,1
Pincus, M. and Rajgopal, S. (2002), “The interaction between accrual management and hedging:
evidence from oil and gas firms”, The Accounting Review, Vol. 77 No. 1, pp. 127-160.
Press, E. and Weintrop, J. (1990), “Accounting-based constraints in public and private debt
agreements: their association with leverage and impact on accounting choice”, Journal of
122 Accounting and Economics, Vol. 12 No. 1, pp. 65-95.
Report on the Observance of Standards and Codes (ROSC) (2009), “A corporate governance
country assessment for the Arab Republic of Egypt”, available at: www.worldbank.org/ifa/
rosc_cg_egy.pdf (accessed 28 February 2013).
Roychowdhury, S. (2006), “Earnings management through real activities manipulation”, Journal
of Accounting and Economics, Vol. 42 No. 3, pp. 335-370.
Sloan, R. (1996), “Do stock prices fully reflect information in accruals and cash flows about future
Downloaded by BASKENT UNIVERSITY At 03:11 23 December 2014 (PT)

earnings?”, The Accounting Review, Vol. 71 No. 3, pp. 289-315.

Sourial, M. (2004), “Corporate governance in the Middle East and North Africa: an overview”,
SSRN, available at: http://ssrn.com/abstract ¼ 508883 (accessed 15 February 2012).
Sweeney, A. (1994), “Debt-covenant violations and managers’ accounting responses”, Journal of
Accounting and Economics, Vol. 17 No. 3, pp. 281-308.
Teoh, S., Welch, I. and Wong., T. (1998a), “Earnings management and the long-run market
performance of initial public offerings”, Journal of Finance, Vol. 53 No. 6, pp. 1935-1974.
Teoh, S., Welch, I. and Wong., T. (1998b), “Earnings management and the underperformance of
seasoned equity offerings”, Journal of Financial Economics, Vol. 50 No. 1, pp. 63-99.
Teshima, N. and Shuto, A. (2008), “Managerial ownership and earnings management: theory and
empirical evidence from Japan”, Journal of International Financial Management &
Accounting, Vol. 19 No. 2, pp. 107-132.
Trueman, B. and Titman, S. (1988), “An explanation for accounting income smoothing”, Journal
of Accounting Research, Vol. 26, Supplement, pp. 127-139.
Tucker, J. and Zarowin, P. (2006), “Does income smoothing improve earnings informativeness?”,
The Accounting Review, Vol. 81 No. 1, pp. 251-270.
Warfield, T., Wild, J. and Wild, K. (1995), “Managerial ownership, accounting choices, and
informativeness of earnings”, Journal of Accounting and Economics, Vol. 20 No. 1, pp. 61-91.
Watts, R. and Zimmerman, J. (1990), “Positive accounting theory: a ten year perspective”,
The Accounting Review, Vol. 65 No. 1, pp. 131-156.
Young, S. (1998), “The determinants of managerial accounting policy choice: further evidence for
the UK”, Accounting and Business Research, Vol. 28 No. 2, pp. 133-143.
Young, S. (1999), “Systematic measurement error in the estimation of discretionary accruals:
an evaluation of alternative modelling procedures”, Journal of Business Finance &
Accounting, Vol. 26 Nos 7-8, pp. 833-862.

Further reading
Hall, S. and Stammerjohan, W. (1997), “Damage awards and earnings management in the oil
industry”, The Accounting Review, Vol. 72 No. 1, pp. 47-65.

Corresponding author
Dr Mohamed Khalil can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints