Accepted Manuscript

Title: How Efficient Ownership Structure Monitors Income

Manipulation? Evidence of Real Earnings Management
among Malaysian Firms

Authors: Mojtaba Shayan Nia, Philip Sinnadurai, Zuraidah

Mohd Sanusi, Ancella Hermawan

PII: S0275-5319(16)30344-0
DOI: http://dx.doi.org/doi:10.1016/j.ribaf.2017.04.013
Reference: RIBAF 629

To appear in: Research in International Business and Finance

Received date: 16-10-2016

Accepted date: 5-4-2017

Please cite this article as: Nia, Mojtaba Shayan, Sinnadurai, Philip, Sanusi,
Zuraidah Mohd, Hermawan, Ancella, How Efficient Ownership Structure
Monitors Income Manipulation? Evidence of Real Earnings Management
among Malaysian Firms.Research in International Business and Finance
http://dx.doi.org/10.1016/j.ribaf.2017.04.013

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
 How Efficient Ownership Structure Monitors Income Manipulation? Evidence of Real Earnings
Management among Malaysian Firms

Mojtaba Shayan Nia

Faculty of Accountancy, Universiti Teknologi Mara, Malaysia
Philip Sinnadurai
Faculty of Business and Economics, Macquarie University, Australia
Zuraidah Mohd Sanusi*
Accounting Research Institute, Universiti Teknologi MARA, Malaysia
Ancella Hermawan
Faculty of Economics and Business, Universitas Indonesia, Indonesia
Abstract
This study investigates income manipulation through real earnings management, by listed companies in
Malaysia, prior to being officially designated as “financially distressed”, by this country’s stock exchange
listed rules. The hypotheses relate to whether the degree of upwards real earnings management, conducted
during the four-year period prior to financial distress, can be explained by ownership structure (measured
with three variables: managerial ownership, institutional ownership and foreign ownership). Using a sample
of 1,180 firm-year observations of financially distressed companies, over the investigation period 2001-
2011, the findings suggest that the degree of real earnings management is not associated with ownership by
management or institutional investors. Conversely, the evidence indicates that foreign shareholders are able
to constrain upwards real earnings management related to discretionary expenditure but not the operating
cycle. This study contributes to the importance of diversity of ownership structures in monitoring income
manipulation among firms.
Keywords: managerial ownership; institutional ownership; foreign ownership; real earnings
management; and financial distress
*Contact author:

Assoc. Prof Dr Zuraidah Mohd Sanusi, Accounting Research Institute, Universiti Teknologi MARA, Shah
Alam 40450, Selangor, Malaysia, Email to: zuraidahms@salam.uitm.edu.my

Abstract
Key words: managerial ownership, institutional ownership, foreign ownership, real earnings
management, and financial distress

1. Introduction
This paper investigates income manipulation through real activities management by listed Malaysian
companies, prior to being officially designated by the stock exchange as being “financially distressed”.
Roychowdhury (2006 p.337) defines real activities manipulation as “departures from normal operational
practices, motivated by managers’ desire to mislead at least some stakeholders into believing certain
financial reporting goals have been met in the normal course of operations”. The paper is grounded in the
assumption that these companies do engage in real activities manipulation. The study examines whether
the extent to which they do so may be explained by ownership structure characteristics, unique to the
Malaysian institutional setting.

1
 There are several motivations studying real activities management. Firstly, a review of the earnings
management literature reveals that managers are shifting away from accrual based earnings management
techniques onto real transaction based earnings management techniques. This is because accrual
manipulation may be restricted in an environment with high quality corporate regulation and national-level
agency mechanisms (Cohen, Dey and Lys, 2008). Hence, managers presumably perceive the risks and costs
of detection to be higher than the benefits. Secondly, Practice Note 17 of the listing requirements of the
Malaysian stock exchange specifies a list of criteria for identifying financially distressed companies. A
company need only meet one of the criteria to be designated as financially distressed. One of the criteria,
that shareholders’ funds are less than one-quarter of paid-up capital, directly uses financial accounting.
Hence, listed companies could potentially use earnings management (real or accruals) to avoid being
caught. Unlike other settings that have investigated potential earnings management to avoid being caught
by GAAP-related thresholds (such as debt covenants), in this setting, the threshold is directly observable to
researchers. Thirdly, investigation of real earnings management (rather than accruals management)
recognizes limitations in the extent to which companies may engage in accruals manipulation, due to the
reversing nature of accruals, constraints of GAAP, auditor and regulatory scrutiny and other agency
mechanisms (Graham, Harvey, and Rajgopal, 2005).

Companies facing pending financial distress are therefore unlikely to rely solely on accruals
manipulation (Rosner, 2003; Roychowdhury, 2006). Furthermore, investigation of real earnings
management acknowledges evidence that discretionary accruals models have limited effectiveness in
isolating discretionary and non-discretionary accruals (Dechow, Hutton, Kim and Sloan, 2012; Ismail and
Sinnadurai, 2012). Similarly, there are motivations for examining companies facing financial distress.
Unsurprisingly, when a company faces the possibility of financial distress, its executives may intentionally
package the financial statements, to avoid being officially classified as “financially distressed”. Hence,
evidence on the relation between real earnings management and ownership structure, obtained from other
settings, and may not generalize to companies facing pending financial distress.

Three hypotheses are tested, regarding the association between the extent to real earnings
management and ownership structure. The first and second hypotheses respectively conjecture that the
degree of real earnings management is positively associated with the level of management ownership and
the level of shareholding by institutional investors. The third hypothesis postulates a negative association
between the degree of real earnings management and the level of ownership by foreign investors.

The methodology is consistent with Roychowdhury (2006). The dependent variables of interest are
abnormal cash flow from operations and abnormal discretionary expenses, calculated via estimation of
Dechow, Kothari and Watts (1998) models on an industry-year basis. The models are estimated by using

2
cross-sectional regression for every year and industry. This requires at least 15 observations for each
industry-year grouping. Deviation from the normal levels is termed abnormal CFO, abnormal discretionary
expenses and abnormal production costs. Imposing all the data availability requirement yields 1,180 firm-
years over period of 2001 to 2011. The independent variables of interest flag companies as being in “high”
or “low” categories, regarding management ownership, institutional ownership and foreign ownership
accordingly.

The results document limited support for the hypotheses. Failure to find evidence of an association
between real earnings management and managerial ownership could be due to lack of separation of
ownership and management in Malaysian family companies (Heng and Sieh, 2004; Ali, Chen and
Radhakrishnan, 2007). Lack of support for the second hypothesis could be due to the heterogeneity of
Malaysian institutional investors with respect to concern for corporate governance quality (including
quality of earnings) (Bushee and Noe, 2000). The evidence partially supports the third hypothesis,
suggesting that foreign investors constrain upward real earnings management related to discretionary
expenditure but not the operating cycle. A possible explanation for the inconsistency is that foreign
investors are more concerned with their companies’ long-term financial health than their daily operations.
The remaining of the paper is organized as follows. Section 2 defines “financial distress”, with
particular regard to the Malaysian setting and discusses the extant literature. Section 3 presents the research
hypotheses. Section 4 explains our measures of real earnings management and discusses our research
design, sample selection and data collection processes. Section 5 presents our empirical findings and
Section 6 concludes the paper.
2. Literature review
2.1. Financial distress and real earnings management
Financial reports specifically earnings figures are important source of information available to investors,
analysts, and board members. They provide information about companies' profitability, financial
performance and financial position. Several figures (e.g. dividends, cash flows, and capital investments)
provided in the financial statements are good performance indicators. These earnings figures provide
significant insights into a firm’s performance on both the medium and long term. It is therefore not
surprising that managers aim to maximise earnings figures through real activities management. This is
particularly true when a firm faces deteriorating financial performance and its results of operations is not
sufficient to meet the analysts forecast for a given period. Moreover, it is hard for a company to survive
when sufficient profit cannot be made as reported by low or even negative profits for several consecutive
years (Ismail and Sinnadurai, 2012). Hence, considering the importance of earnings, managers may feel

3
under pressure and use tools such as real activities management to disguise the actual financial performance
of a company in an attempt to avoid being diagnosed as a “distressed firm”.
In Malaysia, financial distress is often associated with the PN17 status of companies in pursuant to
Paragraph 8.14C (2) of the Bursa Malaysia Listing Requirements. These companies have met one or more
requirements specified under the provision of Practice Note No. 17/2005 of the Bursa Malaysia Listing
Requirements and they are perceived as financially distressed firms (Parker, Peters and Turetsky, 2002).
According to the Bursa Malaysia Listing Requirements, a listed company is financially distressed, when it
triggers any one or more of the PN17 criteria. PN17 firms are required to submit to Bursa Malaysian their
plan on how to regularise or face possible delisting from Bursa Malaysia (2006). Companies that fall within
the definition of PN17 incur some direct and indirect costs, such as more expensive financing opportunity
costs of projects and ultimately delisting from stock exchange. Managers presumably perceive the risks and
costs of detection by the PN17 criteria to be higher than the risks and costs of managing earnings. Therefore,
financially troubled firms are more likely to package their financial statements in an attempt to obscure the
true picture of their performance.

One means of packaging financial statements is by managing accruals. It is well known that
managers use accrual based earnings management techniques to provide flexibility within the accounting
rules and mask true economic performance (García Lara, Osma and Neophytou, 2009). Some examples
include under-provisioning bad debts and delaying writing off assets. However, accrual management
techniques are not the only tools available to managers for manipulating earnings. Managers have a greater
incentive to manipulate actual transactions (Roychowdhury, 2006). This is because transactions
management does not tail the cost and risk of Generally Accepted Accounting Principle (GAAP) violation
or scrutiny by auditors, especially in a heightened regulatory environment (Li, Rider and Moore, 2009).
This phenomenon is often referred to as “real activities management” (this study uses the terms real
activities management, real earnings management, real activities manipulation and real transactions based
earnings management interchangeably).

In contrast to accrual based earnings management techniques, in which managers adjust

assumptions and estimates within the accounting system, real activities based earnings management
involves the timing and structuring of actual business activities in order to achieve a desired financial
reporting result (Li et al., 2009). These activities include managers’ attempts to temporarily increase sales,
reduction of discretionary expenditures such as research and development expenses (R&D) and
overproduction (Roychowdhury, 2006).

However, real activities manipulation is expected to reduce future cash flows, and, consequently,
firm value (Roychowdhury, 2006). As pointed out by Peasnell et al. (2000, p. 420–421) “the use of sub-

4
optimal operating strategies is more costly than the reversals from accruals, and, consequently a more
aggressive form of earnings management and thus, a last resource for management”. In the case of
financially troubled companies, even if real activities management is costly, being designated as
“financially distressed firms” is certainly even more so. Hence, they may engage in the management of real
activities to avoid financial distress.

2.2. Ownership structure and financial distress

When firms face financial distress, a divergence of interests between shareholders and management can
lead to suboptimal management decisions because the goals of the managers and their shareholders are not
necessarily aligned (Alves, 2012). Pressure caused by poor results or extreme financial situations motivate
managers to manipulate accounts in order to alter the firms’ financial performances (Campa and Camacho-
Miñano 2014). Managers of distressed firms are posited to opportunistically structure real transactions to
avoid management turnover during the distressed period (Gilson, 1998), or temporarily inflate the market
price to increase their compensation or gain from cashing stock-based compensation holdings.
However, a variety of factors that may restrict real activities management among distressed firms do
exists. Some studies have indicated that certain corporate governance factors such as ownership structures
have an important impact on corporate accounting behaviour, including earnings management (Ajona,
Dallo and Alegria, 2008; Ali et al., 2010; Alves, 2012). In this study, to analyse whether a firm’s ownership
structure provides effective monitoring of real activities management among distressed firms, three types
of ownerships are considered: managerial ownership, institutional ownership and foreign ownership. The
individual variables for each of these proxies and their possible association with earnings management are
discussed below.
3. Hypothesis development
3.1. Managerial ownership and real earnings management
The hypotheses relating management ownership to real earnings management ownership in financially
distressed companies is argued from the perspectives of two different types of agency relationships of
equity. The first type of agency relationship is between managers and shareholders and referred to as the
“Type 1 agency relationship”. The second type of agency relationship is between controlling shareholders
(who are typically also managers) and non-controlling shareholders. This type of relationship is referred to
as the “Type 2 agency relationship” (Ali et al., 2007).
Consideration of Type 1 agency relationships yields conflicting theoretical expectations regarding
the direction of association between managerial ownership and earnings management by financially
distressed companies. The traditional position is that higher managerial shareholdings merge the identity
between management and shareholders, providing disincentive for managerial opportunism (Fama and

5
Jensen, 1983; Koh, 2003; Yang, Lai and Tan, 2008). Based on this premise, Warfield, Wild, and Wild
(1995) hypothesized that low managerial ownership provides deeper incentives for managers’ to manipulate
earnings. Consistent with this hypothesis, the study documents a negative association between the absolute
value of discretionary accruals and insider ownership in the United States.
A contrary position is that, to the extent that managers and shareholders’ interests are not fully
aligned, higher stock ownership exacerbates managerial entrenchment. Once entrenched, managers can
extract private benefits of control, to the detriment of shareholder wealth. Entrenched managers would be
relatively insulated from the disciplinary agency mechanisms of the markets for managerial labour and
corporate control (Denis and McConnell, 2003; Fama and Jensen, 1983). Managerial entrenchment is
consistent with a positive association between the levels of managerial ownership and real earnings
management.
The second position is likely to be stronger than the first, for companies facing financial distress.
Managers of distressed companies are likely to be more vulnerable that managers of healthy companies.
Hence, managers of distressed companies would be fearful of losing their wealth and employment, if the
company does not emerge from the financial distress.
H1. Distressed firms with high managerial ownership engage in real earnings management activities
more than the distressed firms with less managerial ownership.
3.2 Institutional ownership and real earnings management
The direction of association between earnings management and institutional ownership is likely to differ
according for different types of institutional investors. One category, long-term institutional investors, have
high shareholdings and the intention of holding their stake over a long period. This type of institutions has
greater incentives to collect information, monitor management actions and urge better performance. Long-
term institutions may be concerned about the sustained underlying profitability of the companies. These
investors would want corporate managers to focus on long-term profitability rather than being pre-occupied
with managing earnings on a year-by-year basis. This type of institutional investor would function as a
monitoring mechanism, constraining earnings management by their investee companies. (Bushee, 2001;
Velury and Jenkins, 2006).
Conversely, institutional shareholders with a short-term orientation (also referred to as “myopic”
or “transient” institutional investors) are active portfolio managers, seeking to make short-term risk-
adjusted gains. This group of investors is likely to be attracted to companies with forthcoming disclosure
policies. Evidence suggests that their trading activity induces increased share price volatility of their
investee companies (Bushee and Noe, 2000). When myopic investors are dissatisfied with the performance
of their portfolio companies, they would prefer to liquidate their holdings, rather than monitoring
management (Bushee, 1998; Coffee, 1991).

6
 Empirical evidence suggests that in general, long-term institutional investors, acting as active
monitors, are more pervasive than myopic investors. However, evidence also indicates that the extent to
which the active monitoring hypothesis dominates depends of the size of the institutional holdings.
Naturally, investors with large shareholders are more likely to have a long-term orientation than investors
with small shareholdings (Bushee, 1998; Bushee and Noe, 2000; Chung et al., 2002).
There is a dearth of empirical literature the effectiveness of institutional investors, as monitors of
management, in the presence financial distress. Arguably, both types of institutional investors, in the face
of deteriorating performance, would want to take actions to minimize their losses. One such action may be
pressuring managers to engage in real earnings management. Proponents of the private benefits hypothesis
argue that large institutional investors (regardless of their orientation as long-term versus myopic) are more
likely to enjoy certain benefits such as access to private information, which can be exploited for trading
purposes. Hence, in the event of their shareholdings pending financial distress, these investors would seek
sell their shares prior to the forthcoming share price decline and postpone delisting. The secondary research
hypothesis follows:
H2: Distressed firms with high institutional ownership engage more in real earnings management
activities than distressed firms with less institutional ownership.
3.3 Foreign ownership and real earnings management
Foreign investors (institutional, retail and individual) play a crucial role in improving firm-level
governance. In particular, there is evidence that foreign investors from countries with high quality
governance countries, can improving monitors in “poor governance” countries (Aggarwal, Erel, Ferreira
and Matos, 2011; KHO, Stulz and Warnock, 2009). Foreign investors can influence the managerial
decisions of local companies via direct and indirectly. Foreign investors act as direct monitors via a
“knowledge spillover effect”. Foreign investor import new knowledge and management styles and improve
oversight of accounting processes and operational efficiency (Guo, Huang, Zhang and Zhou, 2014).
Foreign investors may also implement advanced technologies and more efficient organizational structures.
Their presence adds depth to otherwise constrained local markets for managerial labour. Shareholding by
foreign investors facilitates sharing of risk between domestic and foreign shareholders (Henry, 2000;
Berger, Deyoung, Genay and Udell, 2000). Foreign investors indirectly monitor management, due to the
threat of being able to divest their shareholdings in local companies (Gillan and Starks, 2003). These
arguments collectively suggest that foreign ownership would constrain real earnings management. The
arguments are supported by empirical evidence of a negative association between foreign ownership and
real earnings management (using the Roychowdhury (2006) and Cohen (2008) models), using data from
Japan (Guo et al.)

7
 The strength of these “knowledge spillover” mechanisms may be constrained by exacerbated
information asymmetry between managers and foreign shareholders. There is evidence that distance and
lack of knowledge of the local institutional environment impedes the effectiveness of monitoring by foreign
shareholders. For example, the findings of Dvořák (2005) showed that domestic investors earn higher
returns than foreign investors in the Indonesian market. The information asymmetry affect would impair
the ability of foreign investors to constrain real earnings management.
In the case of financially distressed companies, factors causing foreign investors to constrain real
earnings management may be stronger than counter factors. When the level of shareholding by foreign
investors is high and in jeopardy (due to the pending financial distress), the foreign investors may cease to
regard distance as an obstacle. Furthermore, they may take steps to overcome lack of familiarity with the
local environment, say by appointing local managers who are not from the controlling block of shareholders
(Yeung, 2006). The third research hypothesis follows.
H3: Distressed firms with high foreign ownership engage less in real earnings management activities
than the distressed firms with low foreign ownership.

4. Methodology
4.1 Measuring ownership structure
The ownership structure of firms in this study is represented by managerial ownership, institutional
ownership and foreign ownership. This study adopts the definition of Lasfer (2006) about managerial
ownership. It is defined as ownership by members of the board of directors and measured by the percentage
of ordinary shares held by the directors and shares in which the directors are deemed to have an interest. In
this respect, any interest held by family members or related firms are regarded as part of directors’ interests.
Share options were excluded in this study because as suggested by Ofek and Yermack (2000), increase in
managerial ownership via share options exercise is unlikely to be permanent.
In this study, institutional ownership is measured as the percentage of shares held by private
pension funds, public pension funds such as Employee Provident Fund (EPF), mutual funds, insurance
firms, investment funds, and funds managed by banks or foundations and private firms. Following Alves
(2012) data institutional ownership were collected based on the top twenty largest shareholders. Similarly,
foreign ownership is measured as the percentage of shares held by foreign individual and foreign
institutions. Foreign ownership was also collected based on the top twenty largest shareholders.
4.2 Measuring real activities management
Following Roychowdhury (2006), this study investigates patterns in cash flow and discretionary expenses
to detect the real activities manipulation. Deviations from the normal levels are termed abnormal. The
Roychowdhury (2006) models are estimated cross-sectional Ordinary Least Squares (OLS) regressions, on
an industry-year basis. Equation (1) models abnormal cash flow from operations.

8
𝐶𝐹𝑂𝑖𝑡 1 𝑆𝑖𝑡 ∆𝑠𝑡
𝐴𝑖𝑡−1
= 𝛼0 + 𝛼1 (𝐴 ) + 𝛽1 (𝐴 ) + 𝛽2 (𝐴 ) + 𝜀𝑖𝑡 , (1)
𝑖𝑡−1 𝑖𝑡−1 𝑖𝑡−1
where:
CFOi,t = Cash flow from operations of company i, during year t.
Si,t = Sales revenue of company i, for year t.
ΔSi,t = Si,t - Si,t-1
Ai,t-1 = Total assets of company i, as at the end of year t-1.
α0, α1, β1 and β2 are regression parameters.
εi,t is a stochastic disturbance term.
Abnormal cash flow from operations (ABCFO) is the residual from Equation (1). Smaller ABCFO
indicates more real earnings management. In addition, Equation (2) models abnormal discretionary
expenses. Abnormal discretionary expenses (ABDISEXP) are the residual from Equation (2). Smaller
ABDISEXP means more real earnings management.
𝐷𝐼𝑆𝐸𝑋𝑃𝑖𝑡 1 𝑆𝑖𝑡
𝐴𝑖𝑡−1
= 𝛼0∗ + 𝛼1∗ (𝐴 ) + 𝛽1∗ (𝐴 ) + 𝜀∗𝑖𝑡 (2)
𝑖𝑡−1 𝑖𝑡−1

where:
DISEXPi,t = Disretionary expenses of company i for year t, estimated as the sum of research and
development expense, advertising expense and selling, general and administrative expenses, provided
selling, general and administrative expense is available. Advertising expense and research and
development expense are set to zero if they are missing.
α0*, α1* and β1* are regression parameters.
ε*i,t is a stochastic disturbance term.
All other terms are as defined previously.
Models used to test research hypotheses
Equation (3) is estimated, via OLS regression, on a pooled basis, to test the research hypotheses.

𝑌𝑖𝑡 = 𝛼 + 𝛽′1 (𝑂𝑆)𝑖𝑡 + 𝛽′2 (𝑆𝐼𝑍𝐸)𝑖𝑡−1 + 𝛽′3 (𝑀𝑇𝐵)𝑖𝑡−1 + 𝛽′4 (𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒)𝑖𝑡 + 𝜀′𝑖𝑡 (3)

where:
Yi,t = Abnormal CFO or abnormal discretionary expenses (alternatively) for company i in year t.
OSi,t is indicator variable capturing ownership structure. OSi,t is alternatively set to one for firm-years
belongs with high managerial ownership, institutional ownership and foreign ownership, and zero
otherwise.
SIZEi,t-1 = natural logarithm of the market value of equity of company i as at the end of year t-1.
MTBi,t-1 = Market-to-book value of equity of company i as at the end of year t-1.
Net Incomei,t = Net income of company i for year t, scaled by opening total assets.
α, β’1, β’2, β’3 and β’4 are regression parameters.
ε’i,t is a stochastic disturbance term.
The ownership structure variables are the independent variables of interest are managerial
ownership, institutional ownership and foreign ownership. Since Yi,t is an inverse measure of the extent of
real earnings management, the first two hypotheses, conjecturing positive associations with the level of
block holder ownership, are consistent with negative coefficients of the ownership structure variables.
Similarly, the third hypothesis, postulating a negative association, is consistent with a positive coefficient.

9
 Following Roychowdhury (2006), Equation (3) to controls for company size, growth opportunities
and net income as potential determinants of real earnings management. Size is proxied by the natural
logarithm of beginning-of-year market value of equity. We anticipate a positive coefficient attaching to
company size. Larger companies may engage in more earnings management because they stand to suffer
more from the consequences of being caught by PN17 status. The present value of growth options is
proxied by MTBi,t-1, the ratio of beginning-of-year market value of equity to book value of equity. A positive
coefficient of MTBi,t-1 is anticipated. From a shareholder’s point of view, growth options are riskier than
fixed assets. Hence, companies whose value comprises growth options, to a greater extent than assets-in-
place, would have higher required rates of return on equity (Penman, Richardson and Tuna, 2007). They
would experience greater share price decline as a result of being caught by PN17 status. Previous empirical
evidence suggests that discretionary accruals models confound extreme performance (positive or negative)
with poor quality accruals (Dechow et al., 2012). Controlling for scaled, contemporaneous net income
recognizes that the Dechow et al. (1998) models of abnormal cash flow from operations and abnormal
discretionary expenditure may suffer from the same problems. A positive coefficient for net income is
expected.
4.3 Data collection and sample characteristics
The initial sample comprises all companies listed on the Malaysian Stock Exchange that were caught by
PN17 between 2001 and 2011. Data for the companies were hand-collected from the annual reports, for the
four years prior to being caught by PN17. After excluding banks and financial institutions, the final sample
comprises 295 PN17 firms and 1,180 firm-year observations. Table 1 presents the industry distribution of
the final sample and the average of each ownership structure variable within each industry.
Table 1
Industry distribution of final sample and average of ownership structure variables within each industry stratum

INDUSTRY SECTOR Number (%) Average Average Average

of distressed MO IO FO
firms % % %

Industrial Construction & Materials 45 (15) 0.20 0.36 0.06

Industrial General Industrials 26 (9) 0.15 0.41 0.01
Industrial Industrial Transportation 16 (5) 0.31 0.21 0.13
Industrial Industrial Engineering 14 (5) 0.19 0.28 0.00
Industrial Electronic & Electrical Equipment 21 (7) 0.48 0.14 0.00
Consumer Goods Household Goods & Home Construction 15 (5) 0.29 0.42 0.22
Consumer Goods Automobiles & Parts 15 (5) 0.27 0.47 0.09
Consumer Goods Food Producers 28 (9) 0.43 0.16 0.03
Consumer Goods Personal Goods 9 (3) 0.51 0.21 0.00
Consumer Services Leisure Goods 11 (4) 0.35 0.15 0.00
Consumer Services General Retailers 16 (5) 0.23 0.34 0.11
Consumer Services Real Estate Investment & Services 19 (6) 0.29 0.19 0.03
Basic Materials Forestry & Paper 17 (6) 0.41 0.24 0.10
Basic Materials Chemicals 11 (4) 0.09 0.41 0.21

10
 Technology Technology Hardware & Equipment 14 (5) 0.29 0.21 0.05
Technology Software & Computer Services 18 (6) 0.42 0.09 0.00
TOTAL 295 (100) 0.30 0.26 0.06

The above table shows the percentage of shares held by managers, institutions and foreign investors. We hand-collect ownership
data from Bursa Malaysia. The sample period spans 2001-2011. Data on ownership structure is available for a sub-sample of 1,179
firm-years. Industry classifications are based on the Bursa Malaysia industry classification, excluding utility and financial firms.
In order to estimate measures of real earnings management for every industry-year, industry-years with fewer than 9 observations
are eliminated from the sample. Frequency refers to the number of firm-year observations in one industry sector.

Table 1 documents two industry classifications by the Malaysian stock exchange: “industry”
grouping and “sector” grouping. The industry groups reflect a higher level of aggregation that the sector
groupings. The Dechow et al. (1998) models are estimated on a sector-year basis, rather than an industry-
year basis. Naturally, there is more homogeneity of business activities within the finer grouping. Hence,
this research design feature amplifies the power of the tests. Sector-years with fewer than nine observations
were deleted from the final sample, following Roychowdhury (2006).

Table 1 reveals that the construction and materials industry has the largest representation in the
final sample. This statistic could reflect the vulnerability of this sector to macroeconomic business cycles.
Furthermore, the prevalence of the constructions industry likely reflects Government policies to redress the
situation that post-independence, the primary and tertiary sectors comprised much larger percentages of the
Malaysian Gross Domestic Product than the second sector.

The table also shows trends in ownership structure reflective of the Malaysian institutional
environment. The average management ownership is high. The industry with the lowest average
management ownership is chemicals (nine percent) and the industry with the highest management
ownership is personal goods (51%). The average level of institutional ownership is also high, ranging from
nine percent (software and computer services) to 47% (automobiles and parts).

A low level of foreign ownership in Malaysia is also apparent. In five sectors out of 16, there is no
foreign ownership at all. In the sector with the most foreign ownership, household good and home
construction, the average level is only 22%. Foreign investors typically don’t invest in Malaysia with the
principal objective of maximizing portfolio returns. Rather, they invest to enhance their control, for
diversification and exposure to another environment. Hence, their level of ownership tends to be low.

5. Empirical results

Table 2 reports descriptive statistics relating to the estimates of the Dechow et al. (1998) models of
abnormal cash flow from operations and abnormal discretionary expenditure. The table reports the mean

11
coefficients across all 16 industry-years and t-statistics calculated using the standard error of the mean
across industry-years.

Table 2
Summary statistics for abnormal cash flow from operations and abnormal discretionary expenditure

𝐶𝐹𝑂𝑡 /𝐴𝑡−1 𝐷𝐼𝑆𝐸𝑋𝑃𝑡 /𝐴𝑡−1

Intercept 0.0251 0.064

t-statistics (2.35)** (7.68)**
1/𝐴𝑡−1 -3.2141 1.961
t-statistics (-4.16)** (10.52)**
𝑆𝑡 /𝐴𝑡−1 .0412 .0467
t-statistics (5.46)** (7.086)**
𝛥𝑆𝑡 /𝐴𝑡−1 0.0124
t-statistics (2.45)**
Adjusted 𝑅2 0.42** 0.62**

*Significant at the 10% level. **Significant at the 5% level.

This table reports the estimated parameters using Dechow, Kothari and Watts (1998) in the following regressions :

𝐶𝐹𝑂𝑖𝑡 1 𝑆𝑖𝑡 ∆𝑠𝑡

1) = 𝛼0 + 𝛼1 ( ) + 𝛽1 ( ) + 𝛽2 ( ) + 𝜀𝑖𝑡
𝐴𝑖𝑡−1 𝐴𝑖𝑡−1 𝐴𝑖𝑡−1 𝐴𝑖𝑡−1

𝐷𝐼𝑆𝐸𝑋𝑃𝑖𝑡 1 𝑆𝑖𝑡
2) = 𝛼0∗ + 𝛼1∗ ( ) + 𝛽1∗ ( ) + 𝜀∗𝑖𝑡
𝐴𝑖𝑡−1 𝐴𝑖𝑡−1 𝐴𝑖𝑡−1

where:

CFOi,t = Cash flow from operations of company i, during year t.

Si,t = Sales revenue of company i, for year t.
ΔSi,t = Si,t - Si,t-1
Ai,t-1 = Total assets of company i, as at the end of year t-1.
DISEXPi,t = Disretionary expenses of company i for year t, estimated as the sum of research and development expense, advertising
expense and selling, general and administrative expenses, provided selling, general and administrative expense is available.
Advertising expense and research and development expense are set to zero if they are missing.

α0, α1, β1, β2, α0*, α1* and β1* are regression parameters.
εi,t and ε*i,t are stochastic disturbance terms.

The regressions are estimated for every industry-year, using the Bursa Malaysia classification. Industry-years with fewer than 9
firms are eliminated from the sample. The table reports the mean coefficient across all industry-years and t-statistics calculated
using the standard error of the mean across industry-years. The table also reports the mean R2s (across industry-years) for each of
these regressions. Variable definitions are reported in the Appendix. *, **, and *** denote two-tailed significance at the 10%,
5%, and 1% levels, respectively.

The mean adjusted R2 is 42% across industry-years for the cash flow from operations
model and 62% for the discretionary expense model. The signs and significance of the independent
variables of interest conform reasonably with a priori expectations. The results document
reasonably high explanatory power of the Dechow et al. (1998) models.

12
Table 3
Univariate statistics (N = 1,179 firm-years)

Variables Mean Median Std Dev

MVE (millions) 275.32 61.12 4.37

BE ( millions) 250.74 162.48 11.45
T-Assets( millions) 173.12 43.2 6.41
Sales (millions) 371.26 44.26 9.42
CFO (millions) 22.57 10.45 6.28
Sales/A (%) 3.16 2.14 7.55
IBEI/A (%) -1.09 -1.04 3.32
CFO/A (%) 1.59 0.5 3.47
DISEXP/A (%) 17.34 4.1 3.41
ABCFO -0.14 -0.013 0.09
ABDISEXP -0.17 -0.147 0.06
SIZE 7.87 5.00 0.74
MTB 0.95 0.68 0.14
INCOME -3.12 -1.24 0.32
MO (%) 0.30 11.00 6.45
IO (%) 26.35 0.25 11.51
FO (%) 5.91 1.07 6.21

Where:
MVE = Beginning-of-year market value of equity
BE = Beginning-of-year book value of equity
T-Assets = Beginning of year book value of assets
Sales = Sales revenue
CFO = Cash flow from operations
Sales / A = Sales deflated by beginning-of-period total assets
IBEI / A = Income before Extraordinary Items deflated by beginning-of-period total assets
DISEXP/A = Discretionary expenses deflated by beginning-of-period total assets
ABCFO = Abnormal cash flow from operations
ABDISEXP = Abnormal discretionary expenses
SIZE = Natural logarithm of beginning of year market value of equity, expressed as deviation from the corresponding
industry-year mean
MTB = The ratio of beginning-of-year market to book value of equity, expressed as deviation from th corresponding
industry-
year mean
INCOME = Income before extraordinary items (IBEI) scaled by lagged total assets (A), expressed as deviation from the
corresponding industry-year mean.
MO (%) = Percentage of shares held by management
IO (%) = Percentage of shares held by institutional investors
FO (%) = Percentage of shares held by foreign investors

This table provides summary statistics for our full sample of 15,212 firm-year observations over the period from 2001 to 2011.

13
 Table 3 reveals that the mean book value of distressed firm-years, approximately RM 250.74
million, is almost equal of the mean for the market value, RM 275.32 million. These statistics suggest that
distressed firms experience a considerable devaluation of share price. The mean and median values of
IBEI/A (-1.09 and -1.04 respectively) elucidate the financial condition of the full sample. Average losses
were larger in magnitude than total assets. Moreover, the mean values of abnormal cash flows from
operations (ABCFO), abnormal discretionary expenses (ABDISEXP) are -0.14 and -0.17 respectively (both
negative). These statistics support the maintained assumption that PN17 companies do engage in upwards
real earnings management.
To test whether real earnings management associate with the structure of distressed firms’
ownership, companies are placed into “high” and “low” groups, according to the three ownership structure
variables. Companies are assigned to the low managerial ownership group, if their managerial ownership
is less than 15% and to the high managerial ownership group, if its managerial ownership is greater than or
equal to 15%1. Similarly, firms with institutional ownership less than 20% are allocated to the low
institutional ownership group; firms with institutional ownership greater than or equal to 20% are assigned
to the high institutional ownership category. Companies are respectively designated as having low and
high foreign ownership if the level of foreign ownership is less than three percent and at least equal to three
percent.

Table 4 reports univariate tests of equality of the independent variables of interest, according to
ownership structure category. Panel (a) reports the tests of mean/median difference between firms with high
managerial ownership and firms with low managerial ownership. From the table, it is evident that the
mean/median values of ABDISEXP are significantly higher for the low managerial ownership firms than
for the high managerial ownership firms (p<0.01), preliminarily suggesting that high managerial ownership
firms engage more in real earnings management in the presence of financial distress. However, there is no
significant difference in ABCFO between the strata. These statistics suggest that PN17 companies prefer
to conduct real earnings management via curtailing discretionary expenditure, rather than activities related
to their operating cycles.

Table 4
Tests of mean and median differences between ownership structure categories

Panel (a) – Low versus high managerial ownership

1 We assume managers own more than 10% holdings will have more impact on firms’ financial performance. However,
if we use the sample median of foreign ownership 3.6% to divide the sample into two groups, we have similar results.

14
 Distressed firms Distressed firms with Difference in
with low managerial high managerial
ownership ownership

Mean Median Mean Median Means (t-test) Median (Z-test)

MVE ($ millions) 162.32 42.12 216.24 132.24 -53.92 -12.34** -90.12 -21.43**
BVE ($ millions) 84.52 63.18 194.12 86.52 -109.6 -15.37** -23.34 -6.78**
T-Assets($ millions) 94.11 32.1 124.21 34.21 -30.1 -11.72** -2.11 -8.35**
Sales ($millions) 264.25 65.24 465.32 64.21 -201.07 -34.12* 1.03 2.57*
CFO ($millions) 12.52 9.04 23.24 11.21 -10.72 -7.21 -2.17 -1.14
Sales/A (%) 3.16 2.14 7.55 4.33 -4.39 -1.24 -2.19 -1.22
IBEI/A (%) -1.09 -1.04 3.32 2.65 -4.41 -1.32 -3.69 -0.45
CFO/A (%) 1.59 0.5 3.47 4.64 -1.88 -0.11 -4.14 -1.24**
DISEXP/A (%) 7.24 2.1 9.12 6.13 -1.88 -1.72** -4.03 -2.47*
ABCFO -0.012 -0.001 -0.125 -0.091 0.113 0.45 0.09 0.71
ABDISEXP 0.042 0.024 -0.178 -0.082 0.22 1.75* 0.106 12.26*
SIZE 6.25 3.45 9.15 5.12 -2.9 -1.94* -1.67 -0.46
MTB 0.65 0.24 2.14 1.86 -1.49 -0.51 -1.62 -1.12
INCOME -1.05 -0.04 -1.42 -0.75 0.37 1.76* 0.71 1.41
MO (%) 0.06 0.03 0.26 0.12 -0.2 -0.91 -0.09 -0.14
IO (%) 0.32 0.16 0.21 0.14 0.11 13.06** 0.02 11.84**
FO (%) 0.03 0.01 0.09 0.04 -0.06 -6.48** -0.03 -2.81**

*Significant at the of 10% level. ** Significant at the 5% level.

Panel (b) – Low versus high institutional ownership

Distressed firms Distressed firms with Difference in

with low high institutional
institutional ownership
ownership

Mean Median Mean Median Means (t-test) Median (Z-test)

MVE ($ millions) 231.23 65.15 198.45 82.46 16.23* -15.46**

32.78 * -17.31
BVE ($ millions) 98.46 41.21 77.41 41.76 14.11* -7.17**
21.05 * -0.55
T-Assets($ millions) 132.86 71.41 166.15 63.17 - 11.08**
27.18*
-33.29 * 8.24
Sales ($millions) 362.78 124.84 461.14 102.73 -98.36 -3.42** 22.11 18.94**
CFO ($millions) 42.14 15.11 22.74 9.28 19.4 -4.25** 5.83 7.41**
Sales/A (%) 2.37 1.46 2.27 1.75 0.1 -.98 -0.29 -4.09**
IBEI/A (%) -0.92 -0.41 -1.47 -0.76 0.55 -1.32 0.35 1.41
CFO/A (%) 0.42 0.17 0.13 0.06 0.29 5.64** 0.11 6.71**
DISEXP/A (%) 6.37 2.91 7.24 3.71 -0.87 -1.47 -0.8 -2.75**
ABCFO -0.14 -0.09 -0.21 -0.112 0.07 2.41** 0.022 1.96*

15
 ABDISEXP -0.21 -0.09 -0.13 -0.06 -3.87** -
-0.08 -0.03 12.53**
SIZE 4.26 1.43 6.24 4.48 -1.98 -6.14** -3.05 -7.18**
MTB 2.78 0.94 3.76 1.46 -0.98 -4.84** -0.52 -11.09**
INCOME -2.17 -0.87 -3.14 -1.24 0.97 6.41** 0.37 4.71**
MO 0.15 0.09 0.11 0.08 10.15* 8.18**
0.04 * 0.01
IO 0.18 0.11 0.33 0.12 -0.15 -4.78** -0.01 -6.27**
FO 0.08 0.03 0.06 0.02 0.02 5.49** 0.01 4.78**

*Significant at the of 10% level. ** Significant at the 5% level.

Panel (c) – Low versus high foreign ownership

Distressed firms Distressed firms with Difference in

with low foreign high foreign ownership
ownership

Mean Median Mean Median Means (t-test) Median (Z-test)

MVE ($ millions) 234.15 76.41 201.41 65.74 16.84* 15.61**

32.74 * 10.67
BVE ($ millions) 117.18 63.51 82.16 33.45 11.21* 7.94**
35.02 * 30.06
T-Assets($ millions) 99.07 45.18 117.41 55.71 - -7.16**
11.06*
-18.34 * -10.53
Sales ($millions) 441.58 185.67 336.45 115.98 105.13 8.46** 69.69 12.35**
CFO ($millions) 68.56 24.18 66.15 31.15 6.45** -
2.41 -6.97 17.16**
Sales/A (%) 4.45 2.23 2.87 1.02 1.58 1.74 1.21 4.72**
IBEI/A (%) -1.23 -0.91 -1.33 -0.84 12.1 -
0.1 -0.07 14.54**
CFO/A (%) 0.67 0.13 0.23 0.07 0.44 1.46 0.06 6.79**
DISEXP/A (%) 5.65 2.17 6.15 3.45 -5.06** -
-0.5 -1.28 11.27**
ABCFO -0.26 -0.09 -0.06 -0.03 -0.2 -4.16** -0.06 -8.46**
ABDISEXP -0.14 -0.04 -0.13 -0.06 - 12.28**
21.81*
-0.01 * 0.02
SIZE 6.12 2.15 4.86 2.18 1.26 5.64** -0.03 -6.26**
MTB 1.94 0.64 2.04 0.76 - -13.71**
13.24*
-0.1 * -0.12
INCOME -3.14 -1.07 -4.04 -1.13 0.9 2.41** 0.06 0.16
MO 0.16 0.13 0.04 0.01 11.84* 3.78**
0.12 * 0.12
IO 0.23 0.10 0.21 0.13 6.41** -
0.02 -0.03 14.76**

16
 FO 0.01 0.00 0.03 0.00 - 0
14.76*
-0.02 * 0

*Significant at the of 10% level. ** Significant at the 5% level.

where all variables are as defined previously.

Panel (b) of Table 4 reports the tests of differences in mean and median abnormal cash flow from
operation and abnormal discretionary expenditure, between firms with high versus low institutional
ownership. Both strata have negative means and medians for both variables, suggesting that both groups of
companies engaged in upwards real earnings management, prior to being caught by PN17. The t- and Z-
statistics comparing the mean and median levels of abnormal cash flow are positive (p<0.10); while the
counterpart statistics comparing levels of abnormal discretionary expenditure between the strata are
negative (p<0.05). Consistent with expectations, these statistics suggest that companies with high
institutional ownership appear to engage in more upwards real earnings management related to
discretionary expenditure than companies with low institutional ownership. However, contrary to
expectations, the opposite appears to be true regarding real earnings management affecting cash flow from
operations.

Panel (c) of Table 4 reports the tests of mean/median difference between firms with high foreign
ownership and firms with low foreign ownership. It is evidence that the mean/median values of ABCFO
and ABDISEXP are significantly higher for the high foreign ownership firms than for the low foreign
ownership firms with p-values less than 0.01, suggesting that high foreign ownership firms engage in less
real earnings management.

Table 5
Regression analysis of managerial ownership and real earnings management

Abnormal CFO Abnormal discretionary

expenses
Coefficient t-statistics Coefficient t-statistics
Intercept 0.164 2.51** -0.045 -1.73*

MO -0.14 -1.35 -0.04 -1.06

SIZE -0.018 -1.37 0.032 -2.04**

MTB 0.14 0.36 0.04 1.85*

NET INCOME 0.232 2.36** 0.05 0.45

17
 R2 0.12 0.24

*Significant at the 10% level. **Significant at the 5% level.

This table reports the result of the Ordinary Least Squares estimates of (3), where MO is the independent variable of interest. The
models are estimated using data over the five years before distressed firms fall under the PN17 list. The total sample includes
2,700 observations. The regressions being estimated are of the form;

𝑌𝑖𝑡 = 𝛼 + 𝛽1 (𝑀𝑂)𝑖𝑡 + 𝛽2 (𝑆𝐼𝑍𝐸)𝑖𝑡−1 + 𝛽3 (𝑀𝑇𝐵)𝑖𝑡−1 + 𝛽4 (𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒)𝑖𝑡 + 𝜀𝑖𝑡 (3)

Where:
Y = Abnormal cash flow from operations or discretionary expenses, alternatively.
MO = 1 if the percentage of shares held by management exceeds 15. Otherwise, MO = 0.
All other variables are as defined previously.

t-statistics are calculated using standard errors corrected for autocorrelation using the Newey-West procedure.

α, β1, β2, β3 and β4 are regression parameters.

εi,t is a stochastic disturbance term.

Table 5 provides evidence on H1 (managerial ownership). Contrary to H1, the coefficients of MO,
managerial ownership, are uniformly insignificant. A possible explanation for the null result is that in
companies with high managerial ownership, such as family companies, it is not necessary to manage
earnings to achieve real wealth transfers from the company to management, owing to relative lack of
separation of management and ownership (Ali et al., 2007). Neither of the R2 in Table 5 is significant,
indicating model misspecification.
The coefficients of the control variables in Table 5 display mixed conformity with expectations.
Contrary to expectations, the coefficient of SIZE is not significant in the model with abnormal cash flow
from operations as the dependent variable. In the model with abnormal discretionary expenditure as the
dependent variable, this coefficient is significant (p<0.05), in the opposite direction from anticipated. The
coefficient of MTB, the ratio of market-to-book value of equity, is not significant in the model with
abnormal cash flow from operations as the dependent variable. However, as anticipated, this coefficient is
positive and significant in the model with abnormal discretionary expenditure as the dependent variable.
The coefficient of NETINCOME is positive and significant (p<0.05) in the model within abnormal cash
flow from operations as the dependent variable but not significant in the other model. This result could be
due to the model for abnormal cash flow from operations (but not the model for abnormal discretionary
expenditure) confounding extreme performance with real earnings management (Dechow et al., 2012). In
totality, these statistics suggest that in the context of financially distressed Malaysian companies, the

18
Dechow et al. (1998) models of abnormal discretionary expenditure are better specified than the counterpart
models for abnormal cash flow from operations.

Table 6
Regression analysis of institutional ownership and real earnings management

Abnormal CFO Abnormal discretionary

expenses
Coefficient t-statistics Coefficient t-statistics
intercept -0.272 -2.36** -0.045 -1.24
IO 0.13 0.85 0.03 0.242
SIZE 0.068 6.37** -0.041 -8.431**
MTB -0.07 -3.134** 0.02* 5.551**
INCOME -0.16 -0.321 -0.022 -0.38
R2 0.26 0.11

*Significant at the 10% level. **Significant at the 5% level.

This table reports the result of the Ordinary Least Squares estimates of (3), where IO is the independent variable of interest. The
models are estimated using data over the five years before distressed firms fall under the PN17 list. The total sample includes
2,700 observations. The regressions being estimated are of the form;

𝑌𝑖𝑡 = 𝛼 + 𝛽1 (𝐼𝑂)𝑖𝑡 + 𝛽2 (𝑆𝐼𝑍𝐸)𝑖𝑡−1 + 𝛽3 (𝑀𝑇𝐵)𝑖𝑡−1 + 𝛽4 (𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒)𝑖𝑡 + 𝜀𝑖𝑡 (3)

where:
Y = Abnormal cash flow from operations or discretionary expenses, alternatively.
IO = 1 if the percentage of shares held by institutional investors exceeds 20. Otherwise, IO = 0.
All other variables are as defined previously.

t-statistics are calculated using standard errors corrected for autocorrelation using the Newey-West procedure.

α, β1, β2, β3 and β4 are regression parameters.

εi,t is a stochastic disturbance term.

Table 6 provides evidence on H2 (institutional ownership). Both coefficients of IO (institutional

ownership) are insignificant, failing to support H2. Lack of evidence of an association between real
earnings management and institutional ownership is consistent with the results of Roychowdhury (2006).
However, the result in Table 6 may be (at least partially) due to unique attributes of the Malaysian setting.
Some institutional investors in Malaysia may be unconcerned about the earnings quality of their investee
companies or may even benefit from the companies being officially classified as financially distressed. A
small percentage of institutional investors in Malaysia are from the private sector. Some of these investors

19
would actively manage their portfolios and hence regard the quality of corporate governance (including
quality of earnings) of their investee companies as being less important to maximization of portfolio returns
(Bushee and Noe, 2000). Other institutional investors in Malaysia may use their companies’ official status
as financially distressed (and subsequent share price decline) to purchase shares at low prices, to facility
intra-industry consolidation. A third possible reason for benefiting from investee companies being
classified as financially distressed is that some institutional investors may wish to simultaneously own
shares in industrial rivals, to diversify portfolio risk. The literature of information transfers suggests that
when one company announces earnings, or issues a management earnings forecast, there may be a share
price impact for other companies in the same industry (Foster, 1981). The share price impact may be due
to a shift in the competitive balance among industry incumbents (Pyo and Lustgarten, 1990). These
institutional investors’ lack of unconcern about real earnings management would offset the concern of the
other institutions, trying to avoid having their companies being caught by PN17.

The coefficients of the control variables in Table 6 display limited conformity with a priori
expectations. While the coefficients of SIZE are both significant (p<0.05), they have different signs in the
two models. (The coefficient is positive in the model with abnormal cash flow from operations as the
independent variable and negative in the model with abnormal discretionary expenditure as the dependent
variable.) Similarly, the coefficients of MTB, while both significant (p<0.05), they have opposite signs in
the two models. (The coefficient is negative in the model with abnormal cash flow from operations as the
independent variable and positive in the model with abnormal discretionary expenditure as the dependent
variable.) The coefficients of NET INCOME are both insignificant. Neither of the R2 in Table 6 is
significant, indicating that these problems could be due to model misspecification.

Table 7
Regression analysis of foreign ownership and real earnings management

20
 Abnormal CFO Abnormal discretionary
expenses
Coefficient t-statistics Coefficient t-statistics
Intercept -0.212 -1.45 -0.045 -3.45*
FO 0.034 1.153 0.36 2.349**
SIZE 0.057 8.935** -0.042 -3.541**
MTB -0.018 -3.235** 0.019 5.106**
INCOME -0.14 -1.37 -0.011 -0.366
R2 0.31 0.16

*Significant at the 10% level. **Significant at the 5% level.

This table reports the result of the Ordinary Least Squares estimates of (3), where FO is the independent variable of interest. The
models are estimated using data over the five years before distressed firms fall under the PN17 list. The total sample includes
2,700 observations. The regressions being estimated are of the form;

𝑌𝑖𝑡 = 𝛼 + 𝛽1 (𝐹𝑂)𝑖𝑡 + 𝛽2 (𝑆𝐼𝑍𝐸)𝑖𝑡−1 + 𝛽3 (𝑀𝑇𝐵)𝑖𝑡−1 + 𝛽4 (𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒)𝑖𝑡 + 𝜀𝑖𝑡

where:
Y = Abnormal cash flow from operations or discretionary expenses, alternatively.
FO = 1 if the percentage of shares held by institutional investors exceeds 3. Otherwise, FO = 0.
All other variables are as defined previously.

t-statistics are calculated using standard errors corrected for autocorrelation using the Newey-West procedure.

α, β1, β2, β3 and β4 are regression parameters.

εi,t is a stochastic disturbance term.

t-statistics are calculated using standard errors corrected for autocorrelation using the Newey-West procedure.

Table 7 reports results of tests of H3 (foreign ownership). The results partially support H3. The
coefficient of foreign ownership, FO, is positive in both models and significant (p<0.05) in the model with
abnormal discretionary expenses as the dependent variable. These results suggest that foreign investors are
able to constrain curtailment of discretionary expenditure as a model of upwards real earnings management
but are unable to constrain upwards real earnings management related to the operating cycle. A possible
explanation for these results is that foreign investors may regard curtailment of long-term discretionary
expenditure, such as expenditure on research and development as more critical for a company’s long-term
solvency, than management of its working capital.

The coefficients of the control variables in Table 7 display limited consistency with expectations
and differ according to the dependent variable. While the coefficients of SIZE are both significant (p<0.05),
they have different signs in the two models. (The coefficient is positive in the model with abnormal cash
flow from operations as the independent variable and negative in the model with abnormal discretionary

21
expenditure as the dependent variable.) Similarly, the coefficients of MTB, while both significant (p<0.05),
they have opposite signs in the two models. (The coefficient is negative in the model with abnormal cash
flow from operations as the independent variable and positive in the model with abnormal discretionary
expenditure as the dependent variable.) The coefficients of NET INCOME are both insignificant. These
results are similar to those reported in Table 6. Neither of the R2 in Table 7 is significant, indicating that
these problems could be due to model misspecification.

6. Conclusion

This paper investigates the extent of income manipulation, measured as real earnings management,
conducted by Malaysian companies officially designated by the securities regulator as financial distressed.
The research question is whether the degree of real earnings management, conducted prior to being caught
by the regulator’s criteria for financially distressed, may be explained by features of the nation’s
institutional features. The study is predicated on the assumption that these companies did engage inincome
manipulation. The descriptive statistics suggest that the maintained assumption is upheld. Use of the
Malaysian setting is motivated by the fact that the securities regulator uses a financial accounting-based
threshold to identify financially distressed companies. Three hypotheses are tested. The first, second and
third hypotheses respectively conjecture that the degree of real earnings management is positively
associated with the level of managerial ownership, institutional ownership and foreign ownership.

The study uses the methodology developed in Dechow, Kothari and Watts (1998) and Roychowdhury
(2006) to investigate real earnings management. The research hypotheses are tested via pooled Ordinary
Least Squares regressions. The dependent variables measure abnormal cash flows from operations and
discretionary expenditure, alternatively. The independent variables of interest measure the extent of
ownership by management, institutional investors and foreign investors. Dechow, Kothari and Watts
models of abnormal cash flow from operations and discretionary expenditure are estimated on an industry-
year basis. Forty-nine industry-years are represented. The industry boundaries are narrower than the
principal categories used by the Malaysian stock exchange, resulting in enhanced homogeneity of business
models within each industry-year group. Tests of the research hypotheses control for company size and
growth options, as potential determinants of real earnings management. They also control for net income,
in recognition of the possibility that models of abnormal cash flow from operations and abnormal
discretionary expenditure, like discretionary accruals models, may misclassify extreme performance as real
earnings management (Dechow, Hutton, Kim and Sloan, 2012).

22
 The final sample comprises 1,180 company-years that were officially designated as financially
distressed (by the Malaysian regulators) during the investigation period 2001 – 2011. Tests of the research
hypotheses model the degree of real earnings management in the four years prior to receiving this official
designation. The parameters of the Dechow, Kothari and Watts (1996) models conform reasonably with
theoretical expectations.

The empirical results suggest that foreign owners constrain upwards earnings management in the
form of reduction of discretionary expenditure. This evidence, consistent with prior findings, indicates that
in developing countries, foreign investors function as an agency mechanism. Foreign investors are able to
import knowledge spillovers and high standards of corporate governance from their countries of origin
(Guo, Huang, Zhang and Zhou, 2014). However, the results suggest that foreign shareholders don’t
constrain upwards real earnings management related to the operating cycle.

The results do not support the other two hypotheses, related to management ownership and
institutional ownership. Failure to support the hypothesis about managerial ownership could reflect lack of
separation of ownership and management in family companies. In this situation, there would be reduced
need to resort to (any) earnings management to effect wealth transfers from shareholders to management
(Ali, Chen and Radhakrishnan, 2007; Radha, 2012).

Lack of support for the hypothesis about institutional investors may be due to the existence of
different type of institutional investors in Malaysia. As in Anglo-American countries, some of these
Malaysian institutional investors may be unconcerned about the earnings quality of their investee
companies (Bushee and Noe, 2000). A limitation of this study is the absence of theory to facilitate
categorisation of Malaysian institutional investors according to this dimension. Further research could
develop theory of this nature and compare earnings and disclosure quality of the investment companies of
the two categories of institutional investor (Bushee and Noe, 2000). Another suggestion for further research
is to compare the extent of real earnings management between Malaysian companies designated as
financially distressed and their counterparts who are not caught.

This study has limitations in the measurement errors of proxies for the variables of interest. The
Dechow, Kothari and Watts (1998) models of abnormal cash flows from operations and abnormal
discretionary expenditure suffer from generic limitations. One of these problems is the assumption that the
processes of generating normal cash flows from operations and discretionary expenditure are homogeneous
across the estimation sample. Hence, further research could investigate real earnings management by
companies classified as financially distressed, via examination of specific management actions (such as

23
curtailing expenditure on research and development and timing of non-current asset sales), rather than
relying on the Dechow, Kothari and Watts model.

Another variable measured with error is the construct of “financial distress”. Notwithstanding
evidence of high quality regulation in Malaysia (Frost, Gordon and Hayes, 2006), it would be extremely
difficult for the Malaysian stock exchange to operationalize this concept. Companies in Malaysia have
been subject to liquidation, having by-passed the Practice Note 17 criteria for “financial distress”. A
suggestion for further research would be to investigate the research question of interest using data from
other institutional settings that use different approaches to defining “financial distress” (Cevik, Dibooglu
and Kenc, 2016). Estimates of the models used to test the research hypotheses are mis-specified, possibly
due to correlated omitted variables. Hence, further research could develop theory, conjecturing other
possible determinants of upwards real earnings management, by financially distressed companies.

References
Aggarwal, R., Erel, I., Ferreira, M., Matos, P. 2011. Does governance travel around the world?
Evidence from institutional investors. Journal of financial economics, 100(1), 154-181.
Ali, A., Chen, T., Radhakrishnan, S. 2007. Corporate disclosures by family firms. Journal of
Accounting and Economics 44 (1), 238-286.
Bursa Malaysia Securities Berhad, 2006. Practice Note 17, Criteria and Obligations of PN17
Issuers, accessed at
http://www.bursamalaysia.com/misc/system/assets/4913/regulation_rules_bm
_main_pn17.pdf, 10 October, 2014 at 8:43am.
Bushee, B. J. 1998. The influence of institutional investors on myopic R&D investment behavior.
Accounting Review, 305-333.
Bushee, B. J. 2001. Do institutional investors prefer near‐term earnings over long‐run value?.
Contemporary Accounting Research, 18(2), 207-246.
Bushee, B. J., Noe, C. F. 2000. Corporate disclosure practices, institutional investors, and stock
return volatility. Journal of Accounting Research, 171-202.
Cevik, E.I., Dibooglu, S., Kenc, T. 2016. Financial stress and economic activity in some emerging
Asian economies. Research in International Business and Finance, 36, 127-139.
Chung, R., Firth, M., Kim, J.B. 2002. Institutional monitoring and opportunistic earnings
management. Journal of Corporate Finance, 8(1), 29-48.
Claessens, S., Djankov, S. Lang, L. 2000. The separation of ownership and control in east Asian
corporations. Journal of Financial Economics 58 (1-2), 81-112.
Claessens, S., Djankov, S., Fan, J. and Lang, L. 2002. Disentangling the incentive and
entrenchment effects of large shareholdings. Journal of Finance 57 (6), 2,741-2,771.
Coffee, J. C. 1991. Liquidity versus control: The institutional investor as corporate monitor.
Columbia Law Review, 1277-1368.
Cohen, D., A. Dey, T. Lys. 2008. Real and accruals-based earnings management in the Pre- and
Post-Sarbanes-Oxley Periods. The Accounting Review, 757-787.

24
Dechow, P., Ge, W., Schrand, C. 2010. Understanding earnings quality: A review of the proxies,
their determinants and their consequences. Journal of Accounting and Economics, 50 (2),
344-401.
Dechow, P. M., Hutton, A. P., Kim, J. H., Sloan, R. G. 2012. Detecting earnings management: A
new approach. Journal of accounting research, 50(2), 275-334.
Dechow, P. M., Kothari, S. P., Watts, R. 1998. The relation between earnings and cash flows.
Journal of Accounting and Economics, 25(2), 133-168.
Dechow, P. M., Sloan, R. G., Sweeney, A. P. 1996. Causes and consequences of earnings
manipulation: An analysis of firms subject to enforcement actions by the sec*.
Contemporary accounting research, 13(1), 1-36.
Denis, D. K., McConnell, J. J. 2003. International corporate governance. Journal of financial and
quantitative analysis, 38(01), 1-36.
Dvořák, T. 2005. Do domestic investors have an information advantage? Evidence from Indonesia.
The Journal of Finance, 60(2), 817-839.
Fama, E. F., Jensen, M. C. 1983. Agency problems and residual claims. Journal of law and
Economics, 327-349.
Foster, G. 1981. Intra-industry information transfers associated with earnings releases. Journal of
Accounting and Economics, 3(3), 201-232.
Frost, C., Gordon, E., Hayes, A. 2006. Stock exchange disclosure and market development: An
analysis of 50 international exchanges. Journal of Accounting Research 44 (3), 437-483.
Gillan, S. L., Starks, L. T. 2003. Corporate governance, corporate ownership, and the role of
institutional investors: A global perspective. Journal of Applied Finance, 13(2), 4-22.
Gopalan, R., Jayaraman, S. 2012. Private control benefits and earnings management: evidence
from insider-controlled firms. Journal of Accounting Research 50 (1), 117-157.
Graham, J. R., Harvey, C. R., Rajgopal, S. 2005. The economic implications of corporate financial
reporting. Journal of Accounting and Economics, 40(1-3), 3-73.
Guo, J., Huang, P., Zhang, Y., Zhou, N. 2014. Foreign Ownership and Real Earnings Management:
Evidence from Japan.
Heng, P., Sieh. L. 2004. The Chinese business community in Peninsular Malaysia – 1957-1999, Chapter 5
in Lee, K. and Tan, C., The Chinese in Malaysia, published by Penerbit Fajar Bakti Sendirian
Berhad, Shah Alam, Malaysia, under licence from Oxford University Press, second impression of
first edition, pp. 123-168.
Ismail, N., Sinnadurai, P. 2012. Does ownership concentration type affect the mapping of earnings
quality into value? Malaysian evidence. Journal of Business and Policy Research, 7 (3),
24-47.
Kho, B. C., Stulz, R. M., Warnock, F. E. 2009. Financial globalization, governance, and the
evolution of the home bias. Journal of accounting research, 47(2), 597-635.
Koh, P.S. 2003. On the association between institutional ownership and aggressive corporate
earnings management in Australia. The British Accounting Review, 35(2), 105-128.
Okposin, S., Abdul Hamid, A. and Ong, H. (2005), The Changing Phases of the Malaysian
Economy, Pelanduk Publications (Malaysia) Sendirian Berhad.
Parker, S., Peters, G. F., Turetsky, H. F. 2002. Corporate governance and corporate failure: a
survival analysis. Corporate Governance, 2(2), 4-12.
Penman, S., Richardson, S., Tuna, I. 2007. The book-to-price effect in stock returns:
Accounting for leverage. Journal of Accounting Research 45 (2), 427-67.
Pyo, Y., Lustgarten, S. 1990. Differential intra-industry information transfer associated with
management earnings forecasts. Journal of Accounting and Economics, 13(4), 365-379.

25
Radha, A. S. 2012. Ownership Structure and Earnings Management: Evidence from Portugal.
Australasian Accounting Business and Finance Journal, 6(1), 57-74.
Rosner, R. L. 2003. Earnings manipulation in failing firms. Contemporary Accounting Research,
20(2), 361-408.
Roychowdhury, S. 2006. Earnings management through real activities manipulation. Journal of
Accounting and Economics, 42(3), 335-370.
Velury, U., Jenkins, D. S. 2006. Institutional ownership and the quality of earnings. Journal of
business research, 59(9), 1043-1051.
Warfield, T. D., Wild, J. J., Wild, K. L. 1995. Managerial ownership, accounting choices, and
informativeness of earnings. Journal of Accounting and Economics, 20(1), 61-91.
Yang, C.-Y., Lai, H.N., Tan, B. L. 2008. Managerial ownership structure and earnings
management. Journal of Financial Reporting and Accounting, 6(1), 35-53.
Yeung, H. 2006. Change and continuity in Southeast Asian ethnic Chinese business. Asia-Pacific
Journal of Management, 23 (3), 229-254.

26