I ' THE UNIVERSITY OF THE SOUTH PACIFIC
t r LIBRARY Author Statement of Accessibility- Part 2- Permission for Internet Access
Name of Candidate : A 6 L cLqrcc*, b k a
Degree : h 4 e g oC( ~ - c e
Department/School : ~ C L o o l A' ACCWY) & V q
~nstitution/university : ~ln\Ve~sl& E K c 5:~ cc Thesis Title : P~MLSC: CX Eat o w 9 5 c-n+ & a C L 5
Date of completion of requirements for award : 6 13 101
1. I authorise the University to make this thesis available on the Internet for access by USP authorised users.
2. I authorise the University to make this thesis available on the Internet under the International digital theses project
Signed: s Date: \3 !3 lo?
Contact Address Permanent Address
Analyses of Earnings Management Practices
in State Owned Enterprises
Case Studies of the Housing Authority and
the Fiji Electricity Authority
by
Asha Shyreen Lata
A thesis submitted in partial fulfillment of the requirements for the
degree of Master of Commerce in Accounting
School of Accounting and Finance
Faculty of Business and Economics
The University of the South Pacific
Suva, Fiji
March 2007
© Asha Shyreen Lata 2007
ii
DECLARATION OF ORIGINALITY
I, Asha Shyreen Lata, declare that this thesis is my original work and, to the best
of my knowledge, does not contain any material(s) used from elsewhere.
However, where materials have been borrowed, due acknowledgement has been
rendered in an appropriate manner.
----------------------------------
Asha Shyreen Lata
March 6, 2007.
iii
ACKNOWLEDGEMENT
I would like to express my gratitude to the several people who have assisted me
in completing this thesis. I gratefully acknowledge the continuous support and
guidance provided by my supervisors, Associate Professor Arvind Patel and
Professor Michael White, throughout the period of study. Their critical
comments and suggestions on all aspects of the thesis were very useful. I also
acknowledge the two examiners, Professor Howard Davey and an anonymous
examiner, for reviewing this thesis.
I also wish to thank Mr. Parmesh Chand, the Chief Executive Officer of the
Ministry of Public Enterprises and Public Sector Reform and his staff members
for providing relevant information about the organisations on a timely basis. I
am equally thankful to Mr. Narendra Prasad, the former Chief Executive Officer
of the Housing Authority for providing pertinent information about the
Authority.
I am appreciative to my husband, Mr. Rup Singh, for providing editorial
support. His assistance was useful in improving the general structure of the
thesis. I am also grateful to the University of the South Pacific for providing
financial support to complete the study. Finally, I sincerely acknowledge my
family for their moral support, understanding and encouragement during the
completion of my thesis.
iv
DEDICATION
To my parents and the God Almighty
v
ABSTRACT
Earnings management practices lower the quality of financial reporting as they
are often used to obscure the true performance of businesses. This study analyses
the possibilities of earnings management in the two statutory organizations in
Fiji, the Housing Authority and the Fiji Electricity Authority, over the period
1990 to 2004. The results suggest that the entities may have practiced earnings
management by using provisions, amortization and depreciation, revaluation,
capitalisations and misclassification of government grants. These practices were
largely the result of professional judgment and flexibility permitted by the
accounting standards, although, there were a few cases, which occurred due to
non-compliance with generally accepted accounting practices. Further, agency
relationships provided several economic incentives to misreport the financial
outcome, such as reflecting performance of the incoming Board and/or
government, converting debts into equity, satisfying debt covenants and
attracting outside financiers. These findings are consistent with the a priori
expectation derived from studies of earnings management that such practices
arise because of flexibility in accounting standards and agency relationships.
vi
TABLE OF CONTENTS
Contents Page No.
Title Page i
Declaration of Originality ii
Acknowledgement iii
Dedication iv
Abstract v
Table of Contents vi
List of Tables and Figures x
List of Abbreviations xii
Chapter One
Overview of the Thesis
1.1 Introduction 1
1.2 Background 2
1.3 Justifications and contributions of the study 3
1.4 Objectives of the study 5
1.5 Organisation of the thesis 5
Chapter Two
Understanding Earnings Management Behaviour
2.1 Introduction 6
2.2 Why does earnings management occur? 6
vii
2.2.1 Managerial compensation effects 7
2.2.2 Borrowing cost effects 8
2.2.3 Equity offerings 9
2.2.4 Management buyout 10
2.2.5 Meeting analysts’ expectations 11
2.2.6 Reduce(increase) regulatory costs( benefits) 11
2.3 How does earnings management occur? 13
2.4 Conclusion 16
Chapter Three
Earnings Management in State Owned Enterprises
3.1 Introduction 17
3.2 Institutional background 17
3.3 Possibilities of earnings management in SOEs
3.3.1 Economic incentives 21
3.3.2 Accounting regulation 23
3.3.3 Cultural influences 25
3.4 Conclusion 27
Chapter Four
Methodology
4.1 Introduction 28
4.2 Empirical models of earnings management 28
4.3 Data 32
4.4 Research Method 32
4.5 Overlap between the two approaches 39
4.6 Conclusion 41
viii
Chapter Five
Evidence of Earnings Management Practices in the Housing Authority
5.1 Introduction 42
5.2 Overview of the Housing Authority
5.2.1 Background 42
5.2.2 Lending activities 45
5.2.3 Accounting environment and financial performance 47
5.3 Possible instances of earnings management
5.3.1 High provision for write downs of developed lots in 1988 to 1991 48
5.3.2 Capitalisation of administration expenses 54
5.3.3 Deferred interest expense 60
5.3.4 Changes in accounting policies for provision for doubtful debts 67
5.4 Conclusion 80
Chapter Six
Evidence of Earnings Management Practices in the Fiji Electricity Authority
6.1 Introduction 81
6.2 Overview of the Fiji Electricity Authority
6.2.1 Background 81
6.2.2 Electricity generation 84
6.2.3 Accounting environment and financial performance 85
6.3 Possible instances of earnings management
6.3.1 Capitalisation of research and development costs 86
6.3.2 Revaluation of fixed assets 94
6.3.3 Classification of government grants 96
6.3.4 Capitalisation policies 101
6.3.5 Changes in depreciation rates 103
6.4 Conclusion . 112
ix
Chapter Seven
Conclusions and Limitations
7.1 Conclusions 113
7.2 Limitations of the study 116
Bibliography 118
Appendices
1 Checklists 124
2 Current interest rates offered by different financial institutions 130
3A Financial indicators of the Housing Authority 131
3B Financial indicators of the Fiji Electricity Authority 132
x
LIST OF TABLES AND FIGURES
Tables Description Page No.
2.1 The Relationship between Accounting Philosophy and GAAP 14
3.1 List of State Owned Enterprises 18
3.2 Financial Performance of State-Owned Enterprises 20
5.1 Interest Rate Spread of the Housing Authority and the
Commercial Banks 46
5.2 Operating Costs of the Housing Authority and the
Commercial Banks 46
5.3 Auditors Remuneration 47
5.4 Developed Lots Inventory 49
5.5 Abnormal Items 50
5.6 Financial Indicators of the Housing Authority Before and After the
Transfer of Rental Operations (As a percentage of total assets) 52
5.7 Inventory Days 56
5.8 Effects of Capitalising Administrative Costs on Profits 57
5.9 Amortisation of Deferred Interest Expense 63
5.10 Components of Total Expenses in 1993 63
5.11 Interest Expense 64
5.12 Loan Liabilities in 1992, 1993 and 1994 ($000s) 65
5.13 Interest Expense on New Loans Raised During 1992, 1993
and 1994 65
5.14 Discretionary Expenditure 71
5.15 Profits in 1995 and 1996 (as a percentage of assets) 72
5.16 Profitability and Equity Position of the Housing Authority 76
xi
5.17 Effects of Change in Accounting Policy at the Housing
Authority in 1999 78
5.18 Summary of Earnings Management Evidence in the
Housing Authority 80
6.1 Research and Development Cost ($000s) 87
6.2 Capital Works In Progress in 1984 89
6.3 Amortisation of Training Expenses 91
6.4 Effect of Cost Capitalization on Return on Fixed Assets 93
6.5 Effects of Revaluation in 1992 94
6.5 Effects of Misclassification of Deferred Income 99
6.7 Combined Effects of Revaluation and Classification of
Deferred Income 100
6.8 Total Expenses and Profits of the Fiji Electricity Authority 102
6.9 Depreciation Expenses and Profits of the Fiji
Electricity Authority 104
6.10 Summary of Earnings Management Evidence in the Fiji
Electricity Authority 112
xii
LIST OF ABBREVIATIONS
ADB Asian Development Bank
CSA Commercial Statutory Authorities
FASs Fiji Accounting Standards
FEA Fiji Electricity Authority
FIA Fiji Institute of Accountants
FNPF Fiji National Provident Fund
GAAP Generally Accepted Accounting Principles
GCC Government Commercial Companies
HA Housing Authority
IASs International Accounting Standards
NBF National Bank of Fiji
PRB Public Rental Board
RBF Reserve Bank of Fiji
SOEs State Owned Enterprises
Analyses of Earnings Management Practices in SOEs
1
CHAPTER ONE
OVERVIEW OF THE THESIS
“I fear that we are witnessing an erosion in the quality of earnings and therefore the quality of financial reporting.
Managing may be giving way to manipulation; integrity may be losing out to illusion”
Levitt 1998, the former Chairman of Securities and Exchange Commission.
1.1 Introduction
The issue of earnings management has been widely researched and found to be
pervasive in the private sector in many developed economies (DeFond and Jiambalvo,
1994; Teoh et al., 1998; Beneish, 2001; Barton and Simko, 2002 and Krishnan, 2003).
However, little work has been done in the context of developing economies and the
public sector. It will not be clear as to how pervasive the phenomenon of earnings
management really is until the analysis is extended to developing economies and to
public sector entities. This thesis has been undertaken to make a contribution to address
this lacuna. It constitutes an endeavour to identify and analyse possible instances of
earnings management in the public sector of a developing economy, namely Fiji. The
study focuses on two State Owned Enterprises (SOEs) - the Housing Authority and the
Fiji Electricity Authority. At the outset, this thesis can only suggest that there may of
evidence of earnings management in these entities. This introductory chapter begins
with some background information on earnings management and then proceeds to justify
the purpose of the study and its contributions to the literature. Further, the chapter
outlines the research objectives and presents the layout of the thesis.
Chapter 1: Overview of the Thesis
2
1.2 Background
Corporate collapses such as Enron, Qwest, Xerox, WorldCom and HIH Insurance reflect
the failure of conventional accounting and auditing practices. The demise of Arthur
Anderson has further questioned the integrity of financial reporting. Such failures raise
concern about financial reporting, auditing, corporate regulations and the accounting
profession as a whole. Thus, as Fraser and Ormiston (2001:25) state,
The sharper and clearer the picture presented through the financial data and the
closer that picture is to the financial reality, the higher is the quality of financial
statements and reported earnings.
In an effort to improve the quality of financial reporting, the international accounting
profession has taken a number of steps. These include improving the audit committees’
strength and independence, pushing for greater focus on auditor independence and
increasing the review of filings submitted by public companies to detect earnings
management (Behn et al., 2002).
Earnings management is not a new phenomenon. It has received attention from
regulators, practitioners and the media. Several studies strongly suggest that earnings
management is becoming a common practice in firms (Healy, 1985; McNichols and
Wilson, 1988; DeFond and Jiambalvo, 1994; Natarajan, 1999; Nelson et al., 2002;
Beneish and Vargus, 2002 and Palliam and Shalhoub, 2003). As one of their major
findings, the Sponsoring Organisations of the Treadway Commission concluded that
more than 50% of the financial reporting frauds that took place among the USA public
companies over a period from 1987–1997 involved overstatement of revenue (Palliam
and Shalhoub, 2003).
In the literature, earnings management is also referred to as creative accounting,
income/earnings smoothing, financial engineering and cosmetic accounting. It occurs
when managers use judgment in financial reporting and in structuring transactions to
alter financial reports with intentions of either misleading stakeholders about the
Analyses of Earnings Management Practices in SOEs
3
underlying economic performance of the company or to influence the contractual
outcomes that depend on reported accounting numbers (Healy and Wahlen, 1999).
Opportunities for these practices arise because accounting standards allow for flexibility
and judgment in selecting the reporting methods, estimates and disclosures that reflect a
firm’s economic conditions (Vinciguerra and O’Rielly, 2004). Thus, it becomes
important to study earnings management as such practices obscure the firm’s intrinsic
value.
1.3 Justification and Contributions of the Study
Studies on earnings management has mainly focused on the developed economies, such
as the USA, UK, Canada and some continental European countries. In developing
countries, however, such studies are rare and to date, none has been undertaken in the
Pacific Island Economies. As institutional and regulatory differences exist between
developed and developing countries, it is difficult to extrapolate findings/evidences from
one to another. With this gap in the literature, this thesis attempts to examine earnings
management practices in two SOEs in Fiji, which has a relatively weaker financial
regulatory environment than the developed countries. Unlike the USA, Australia and
New Zealand, the accounting standards in Fiji do not have a legal backing.
Consequently, a weaker regulatory environment would permit higher levels of
manipulation.
Furthermore, Fiji is no exception when it comes to failures in financial reporting. An
excellent example is the economic losses incurred when the state owned bank, National
Bank of Fiji (NBF), collapsed. Grynberg et al. (2002) provide a comprehensive insight
into the collapse of the NBF. They describe the NBF’s case as the "largest known
financial scandal in the history of Fiji and the Pacific Islands". According to them, the
bank's failure resulted from corruption, mismanagement and improper accounting
practices. For instance, loans were advanced without adequate security followed by lack
of monitoring of these loans. These peaked in 1996 when bad and doubtful debts
Chapter 1: Overview of the Thesis
4
amounted to $220 million, representing 8% of Fiji's GDP. The NBF failed despite
regulations and monitoring by the Reserve Bank of Fiji (RBF), the Auditor General and
the Ministry of Finance. Its failure emphasised the need for good financial disclosures,
sound corporate governance and robust banking supervision.
The literature on earnings management heavily concentrates on the private sector while
little is known about the prevalence of such practices in different institutional
environments. SOEs, an important sector, particularly in the developing countries, have
been examined in a few studies only (Likerman, 1983 and McInnes, 1990a and 1990b).
Many large organizations in Fiji are state owned, hence substantial investments are made
to these entities. Despite their importance, earnings management in SOEs has been
largely overlooked. Consequently, this study attempts to extend the existing earnings
management literature by exploring two SOEs in Fiji.
Studies on earnings management have mainly used quantitative techniques, the most
common being the variants of the accruals models. However, these models are
constrained by measurement error and their use is limited to the developed countries that
can meet the data requirements (see discussions in Chapter 4). In contrast, this study
examines earnings management using a different approach, viz, the checklists designed
by Mulford and Comiskey (2002). It may be noted that this is a pioneering attempt to
explore earnings management using the qualitative method in the case of a developing
country.
The study is useful to the Fiji Institute of Accountants, the accounting standard setting
body in Fiji, as it identifies
• the specific accounting standards that are used to practice earnings management;
• the extent to which managers misuse accounting discretion and judgment permitted
by the standards and
• the specific accruals that are commonly used to employ such practices in the
aforesaid entities.
Analyses of Earnings Management Practices in SOEs
5
Finally, the study has implications for the auditors of the two entities and the auditing
profession as a whole. The findings would indicate if their audit procedures are adequate
to identify earnings management practices and the areas which require more scrutiny.
1.4 Objectives of the Study
The objectives of this thesis are two-fold. Firstly, it attempts to identify the presence of
earnings management behaviour in the two aforesaid SOEs in Fiji. Secondly, the study
aims to determine the motivations for such practices.
1.5 Organisation of the Thesis
This thesis comprises of seven chapters. The current chapter is an overview of the thesis.
A brief survey of earnings management literature is in Chapter Two. Chapter Three
describes Fiji's SOEs and discusses the possibilities of earnings management in these
entities. Chapter Four outlines the methodology adopted in the study and justifies the
choice of the qualitative approach over the well-known accruals models. The next two
chapters analyse and discuss the results of earnings management in the two entities and
conclusions and limitations are stated in the final chapter.
Chapter 2: Understanding Earnings Management Behaviour
6
CHAPTER TWO
UNDERSTANDING EARNINGS MANAGEMENT BEHAVIOUR
2.1 Introduction
This chapter presents a brief survey of the earnings management literature. It
concentrates on two major issues; why earnings management occurs and how it is
possible in the presence of accounting regulation.
2.2 Why Does Earnings Management Occur?
The agency theory has been used to explain earnings management behaviour. According
to Jensen and Meckling (1976), an agency relationship is a contract under which one or
more persons-the principal(s), engage another person-the agent, to perform some service
on their behalf that involves delegating some decision making authority to the agent. If
both parties are utility maximisers, it is reasonable to believe that the agent will not
always act in the best interest of the principal. Agency problems arise because of
asymmetric information between the two parties, where the management possesses a
more complete set of organizational information relative to the current and potential
stakeholders. Proprietary costs of disclosure, accounting rules and other institutional and
contractual constraints suggest that there is restriction on total communication
(Richardson, 2000). Thus, when management does not communicate all information,
conflict exists between the privileged agents and the remote body of stakeholders.
Richardson provides empirical evidence that information asymmetry is a necessary
condition for earnings management. Using a broad sample of firms, he finds a positive
relationship between the levels of information asymmetry and earnings management.
Analyses of Earnings Management Practices in SOEs
7
This is justified on the basis that when information asymmetry increases, stakeholders
may not have the necessary information to see through the manipulation of earnings.
Alternatively, an environment with a high level of information asymmetry suggests that
stakeholders do not have sufficient resources, incentives or access to relevant
information to monitor management’s actions, which can also give rise to earnings
management. Thus, it is both logical and inescapable that management behaviour will be
self-serving within the agency relationship (Amat and Gowthorpe, 2004). Hence, agency
theory is a solid framework for understanding earnings management as it provides the
agents with economic incentives to manipulate the financial results. The following
section describes the major incentives found in the literature.
2.2.1 Managerial Compensation Effects
One of the most common incentives for earnings management is the manager’s
remuneration package. The employment contract of management often includes
accounting based constraints that determine compensation opportunities, such as annual
salary increases, management bonus, performance evaluation, reaching targets set in the
compensation contracts. It is logical that where management's remuneration is based on
the level of profits, they will manipulate it either to increase or smooth their income.
Healy (1985) is among the first to propose this argument. His study shows how upper
and lower bounds on executive bonus packages encourage managers to make
discretionary accounting accruals in a strategic manner. A common form of
compensation contract specifies the minimum level of profits to grant bonuses. Hence,
such contracts provide incentives for income-increasing or income-decreasing earnings
management depending on the actual level of profits attained in a particular period.
Using accruals and changes in accounting procedures, Healy finds that as long as profits
fall within the minimum and maximum boundaries, managers choose income-increasing
accruals. However, when earnings are above the maximum level, managers tend to move
towards income- decreasing accruals.
Chapter 2: Understanding Earnings Management Behaviour
8
Using a sample of 102 firms for the period 1980-1990, Gaver et al. (1995) extend
Healy’s study by examining the relationship between discretionary accruals and bonus
plan bounds. Contrary to Healy, they find that when earnings before discretionary
accruals are below the lower bound, managers select income-increasing discretionary
accruals (and vice versa). They believe that these results are more consistent with
income smoothing hypothesis than with Healy's bonus hypothesis. However, they do not
rule out the mechanical selection bias in portfolio formation as an alternative
explanation for their results.
Guidry et al. (1999) test and validate Healy’s hypothesis by examining business-unit
managers. They use business unit-level data as this reduces the aggregation problem that
arises from firm-level data. Also, as managers will be paid bonuses based solely on
business-unit earnings, the confounding effects of long-term performance and stock-
based incentive compensation will be absent. Guidry et al. find that business-unit
managers manipulate earnings to maximize their short-term bonuses.
2.2.2 Borrowing Cost Effects
Debt contracts provide another incentive for earnings management. These contain debt
covenants, which are described by Mulford and Comiskey (2002) as stipulations
included in debt agreements designed to monitor corporate performance. For example, a
lender might instruct that a certain value for an accounting ratio be maintained or impose
limits to investing and financing activities. Violating the covenants could result in the
lender increasing the interest rate, requiring additional financial security, or in extreme
cases, calling for immediate repayment. Therefore, debt covenants provide incentives for
earnings management either to reduce the restrictiveness of accounting based constraints
in debt agreements or to avoid the costs of covenant violations (Beneish 2001).
Analyses of Earnings Management Practices in SOEs
9
Studies investigating the debt covenant hypothesis postulate that managers have
incentives to make financial reporting decisions that reduce the likelihood of violating
debt covenants. Using a sample of USA firms that actually violated their lending
covenants, DeFond and Jiambalvo (1994) find that firms accelerate earnings one year
prior to the violation. They interpret this behaviour as evidence of earnings management
by firms that are close to violating their lending covenants. Sweeney (1994) also finds
that covenant violators manipulate earnings, but in the year(s) following rather than prior
to the violation. This behaviour suggests that earnings are not managed specifically to
avoid violating a lending contract but to reduce the likelihood of violating them in the
future. In order to provide large-sample tests of the debt covenant hypothesis, Dichev
and Skinner (2002) use a database of private lending agreements for USA firms. They
find an unusually small number of loans with financial measures just below covenant
thresholds and an unusually large number of loans with financial measures at or just
above covenant threshold. They provide strong evidence that managers take actions to
avoid debt covenant violations.
2.2.3 Equity Offerings
Issuing shares provide a direct incentive to manage earnings since firms that report a
higher earnings power tend to witness a favourable effect on share prices. From the
firm’s perspective, a higher share price implies an increase in its market valuation and a
reduction in cost of capital. As Dechow and Skinner (2000) state, to the extent that
managers can increase reported earnings without being detected, they can improve the
terms of the public offer, providing direct monetary benefits to themselves and their
firms. Studies examining this incentive test whether managers manipulate earnings in
periods prior to initial public offers (IPO) and seasoned equity offers. The IPO process is
susceptible to earnings management since the information is highly asymmetrical
between the investors and IPO issuers. As there is no previous price on the market,
manipulating earnings would increase its introductory price. Using a sample of 1649
IPO firms in the USA, Teoh et al. (1998a) examine the relationship between
Chapter 2: Understanding Earnings Management Behaviour
10
underperformance of IPOs in the subsequent periods and earnings management. They
find evidence that issuers with unusually high accruals in the IPO year experience poor
stock returns in the three years thereafter and conclude that the underperformance is
largely the result of earnings management in the IPO year. Similarly, Rangan (1998) and
Teoh et al. (1998b) provide evidence that earnings are managed at the time of seasoned
equity offerings. Their studies show that at the time of the issue, firms report unusually
high earnings due to unusually high accruals. However, in the subsequent years, they
report poor earning performance. Both studies find strong association between the extent
of earnings management and subsequent stock returns.
2.2.4 Management Buyout
In the previous section, it was noted that upward earnings management is practiced at
the time of equity offerings. However, earnings are managed downwards in the case of a
management buyout. The management is faced with a conflict in such a situation. They
have a fiduciary duty to the shareholders to get the best price for the firm. However, as
buyers, managers would not want to pay a high price. Therefore, they have an incentive
to report reduced earnings prior to the buyout. As earnings information play an
important role in valuation of management buyouts, DeAngelo (1986) claims that
management of buyout firms would understate earnings. She uses a sample of 64 New
York and American Stock Exchange companies over the period 1973-1982, whose
managers proposed to buyout the firm. However, DeAngelo finds little evidence of
earnings management by examining the changes in accruals. In support of her finding,
she asserts that shareholders carefully scrutinize financial statements to detect any
manipulation by the management. A similar study was undertaken by Perry and William
(1994), using a larger sample of 175 management buyouts during 1981-88. By
examining discretionary accruals, they provide convincing evidence of income-
decreasing earnings management prior to a buyout.
Analyses of Earnings Management Practices in SOEs
11
2.2.5 Meeting Analysts’ Expectations
Financial analysts or the management often forecast the entities expected earnings prior
to its financial year-end. Studies have examined whether earnings are managed to meet
these expectations. Among other things, DeGerorge et al. (1999) examine if companies
manage earnings to meet analysts’ earnings forecasts. They use a large sample of USA
firms over the period 1974-1996 and find that firms report abnormally high earnings just
to meet or exceed the analysts’ forecast. Kasznik (1999) investigates whether managers
who issue annual earnings forecasts practice earnings management to meet their
forecasts. Using 499 firm-years (366 firms) with management earnings forecasts issued
between 1987 and 1991, he provides evidence that when earnings are below
managements' forecasts, positive discretionary accruals are used to manage earnings
upwards.
2.2.6 Reduce (Increase) Regulatory Costs (Benefits)
Regulatory considerations can also induce firms to misreport financial performance.
Firms that are vulnerable to anti-trust investigations/other adverse political consequences
or seeking government subsidy/protection have incentives to manage earnings to appear
less profitable. Jones (1991) examine whether regulatory scrutiny increases the
likelihood of earnings management by firms benefiting from import relief assistance,
such as tariff increases and quota reductions. Empirically examining 23 firms from 5
different industries that are investigated by the United States International Trade
Commission, Jones finds convincing evidence that the firms practiced income-
decreasing earnings management during the investigation period in order to qualify for
the import relief assistance. Cahan (1992) investigates the effects of monopoly related
antitrust investigations on firms reported earnings. He uses a sample of 48 firms that
were investigated for monopoly related violations by the USA’s regulatory agencies
between 1970 and 1983 and finds that firms under scrutiny reported income-decreasing
discretionary accruals in the investigation years.
Chapter 2: Understanding Earnings Management Behaviour
12
To conclude the discussion on earnings management incentives, it is appropriate to note
that the presence of earnings management will depend not only on the incentives for
such practices, but also on internal and external control procedures. These could include
entity's internal governance structure, previous accounting decisions made by the firm
that limit future discretionary choices and the costs imposed on the entity when earnings
management is revealed (Becker et al., 1998).
Furthermore, of the incentives discussed above, some are income-increasing incentives
while others are income-decreasing. In reality, entities can face multiple incentives at the
same time. If all motivate the management of earnings in an upward or downward
direction, then the number of incentives present does not matter as they will lead to
either income-increasing or income-decreasing earnings management. However,
incentives of opposing directions are impossible to pursue simultaneously. In such
circumstances, the management would seek to manage other aspects of the financial
report. A similar observation was also made by Mr. Prasad, the former Chief Executive
Officer of the Housing Authority, during an interview, see quotation below:
You will find in the Fiji context that each stakeholder is maximising their relative
utility in the final outcome of the reported results. And more often than not, for
each to keep their position, a consensus is reached and justified before the final
figures are reported.
Such a scenario was noted while analysing the Fiji Electricity Authority (see Chapter 6
for detailed discussion). In 2004, it wanted to attract external financiers and at the same
time increase electricity tariff rates. In order to achieve the former objective, the FEA
had to report increased earnings while the later required evidence of reduced
profitability. Therefore, in such cases, one needs to clearly understand the incentives and
their effects in order to see through the manipulations.
Analyses of Earnings Management Practices in SOEs
13
2.3 How Does Earnings Management Occur?
Annual financial reports contain a wealth of information that could be used by various
stakeholders for different purposes. Despite its importance, the preparers of financial
reports enjoy considerable latitude when reporting transactions. Accounting standards
allow scope for flexibility and judgment in preparing financial reports. Healy and
Wahlen (1999) argue that accounting standards must allow these so that management
can convey information to the stakeholders about the entity's performances. On the other
hand, given that auditing is imperfect, corporate insiders can also abuse this flexibility to
hide the true performance and achieve a desired earnings figure.
According to Mulford and Comiskey (2002) and Fraser and Ormiston (2001), one of the
ways of practicing earnings management is through a firm’s selection policies that are
employed to present financial reports. They further state that the selection and
application of generally accepted accounting principles (GAAP) is flexible, leaving
room for exercising professional judgment in several cases. Upon reviewing the
corporate financial reporting practices, Loomis (1999, c.f. Brewer et al., 2002:2) asserts
that
…managing earnings is the fundamental motivation for many accounting frauds.
Missing earnings estimates by even a penny is interpreted as an important failure, in
part because managers have so many earnings management tools at their disposal.
The gaps and flexibility in the accounting standards allow a greater variety of accounting
results. Findings of Nelson et al. (2002) point out that earnings management occurs in
several accounting areas such as revenue recognition, business combination, intangibles,
fixed assets, investments and leases and the most frequently identified is reserves. These
practices can include accounting treatments that are intentionally aggressive,
conservative or fraudulent financial reporting. The former two treatments are within the
boundaries of accounting standards while fraudulent financial reporting is outright
violation. Dechow and Skinner (2000) argue that a within-GAAP choice can be
considered earnings management only if it is used to obscure the true economic position
Chapter 2: Understanding Earnings Management Behaviour
14
of the entity. They further argue that to determine whether a particular choice is earnings
management or a legitimate managerial judgment hinges on managerial intent. The
following table illustrates how the accounting philosophy and GAAP interact to create a
range of accounting policy choices.
Table 2. 1 The Relationship between Accounting Philosophy and GAAP
Accounting philosophy
“Conservative” accounting
“Neutral” accounting
“Aggressive” accounting
“Fraudulent” accounting
Relationship with GAAP
Within GAAP Violates GAAP
Impact
Examples Accounting policy changes
-overly aggressive recognition of provisions or reserves -overstatement of restructuring charges and asset write-offs.
-earnings that result from a neutral operation
-understatement of bad debts provision -drawing down provisions or reserves in overly aggressive manner
-recording sales before realisable -recording fictitious sales -backdating sales invoices -recording fictitious inventory
“Real” Cash Flow Choices
-delaying sales -accelerating R&D or advertising expenditures
-postponing R&D or advertising expenditures; -accelerating sales
(Adapted from DeChow and Skinner 2000 and Auditing and Assurance Standards Board 2001.)
Amat and Gowthorpe (2004), Brewer et al. (2002), Healy and Wahlen (1999) and Blake
et al. (1998) discuss several potential areas for earnings management resulting from the
accounting standards. Based on their studies, these areas can be grouped into the
following categories:
1. Management exercises judgment in order to estimate several future economic events.
These events include the expected lives and salvage values of long-term assets,
obligations for pension benefits and other post employment benefits, deferred taxes,
and losses from bad debts and asset impairment. McNichols and Wilson (1988)
examine the discretionary and non-discretionary elements of bad debts provisions
Tendency to understate earnings and overstate liabilities
Tendency to overstate earnings and assets
Analyses of Earnings Management Practices in SOEs
15
and found evidence of earnings management by firms with unusually high or low
earnings.
2. Managers can report the same economic transactions using a range of acceptable
accounting methods. For instance, non-current assets can be depreciated using
straight line or accelerated depreciation method. Likewise, inventory could be valued
using First In First Out or weighted average methods or borrowing costs can be
treated as part of the cost of the asset or an expense. Schipper (1989) points out that
the effects of these changes can easily be identified in the year of change but it is less
readily discernible thereafter.
3. Management exercise judgment when evaluating whether a transaction meets the
criteria for a particular accounting treatment. For example, for an item to be
extraordinary (when this was permitted by GAAP), it has to be both unusual and
infrequent.
4. Management can decide the timing of genuine transactions in order to give the
desired impression. That is, they can make or defer expenditures, such as research
and development, advertising or maintenance.
5. Management can choose how to structure corporate transactions so that the financial
statements can be manipulated. For example, in a sale and leaseback arrangement,
the sale proceeds of the asset could be artificially increased or decreased, with an
equivalent adjustment to related rental payments. Another possibility relates to
equity investments, which can be structured to avoid or require consolidation.
Chapter 2: Understanding Earnings Management Behaviour
16
2.4 Conclusion
In an environment where the users and preparers of financial reports have asymmetric
information, the agency relationship and the flexibility permitted by the accounting
standards can be used to explain the presence of earnings management. The principal-
agent relationship provides the agents with economic incentives to practice earnings
management and flexibility in accounting regulation provides the management avenues
to manipulate the financial outcome. Therefore, it can be concluded that managers face a
wide variety of financial reporting situations in the real world and a correspondingly
wide variety of opportunities and methods to misreport the financial performance of
their organisations.
Analyses of Earnings Management Practices in SOEs
17
CHAPTER THREE
EARNINGS MANAGEMENT IN STATE OWNED ENTERPRISES
3.1 Introduction
The chapter begins with a general introduction of the SOEs in Fiji. It then discusses the
possibilities of earnings management in these types of entities by relating the economic
incentives identified in the previous chapter, reviewing their accounting environment
and evaluating the influence of cultural factors.
3.2 Institutional Background
Every economy has SOEs to control certain economic activities. These entities engage in
the provision of essential utilities and services such as electricity, aviation, housing and
seaports, to name a few. SOEs are an essential part of Fiji's economic environment as
many of the largest organisations in Fiji are state owned. These entities are large
consumers and producers of resources in Fiji's economy hence provide essential
products and services. This study concentrates on state owned enterprises that are fully
owned by the government. These enterprises are divided into two major categories,
Government Commercial Companies (GCC) and Commercial Statutory Authorities
(CSA). The following table lists the SOEs in Fiji and the year when they were
established.
Chapter 3: Earnings Management in SOEs
18
Table 3. 1 List of State Owned Enterprises
Government Commercial Companies Commercial Statutory Authorities
Airports Fiji Limited (1998) Civil Aviation Authority of Fiji Islands (1997)
Fiji Broadcasting Corporation Limited (1997) Fiji Electricity Authority (1966)
Fiji Hardwood Corporation Limited (1998) Housing Authority (1958)
Fiji Shipbuilding Corporation Limited (2001) Maritime and Ports Authority of Fiji (1998)
Food Processors Limited (2003) Public Rental Board (1989)
Ports Terminal Limited (1997)
Post Fiji Limited (1996)
Rewa Rice Limited (1960)
Unit Trust of Fiji (1976)
Viti Corps Corporation Limited (1995)
Yaqara Pastoral Company Limited (1973)
(Adapted from Ministry of Public Enterprises and Public Sector Reform Annual Report 2003)
These entities are administered by the Ministry of Public Enterprises and Public Sector
Reform. The Ministry monitors compliance with the financial reporting and disclosure
requirements of the Public Enterprise Act 1996. The Boards of GCCs are directly
responsible to the Minister for Public Enterprises for their performances while CSAs
report directly to their respective line ministries who appoints their Boards (such as the
FEA reports to the Minister responsible for energy and the HA to the Minister
responsible for housing). These entities are liable to report regularly and be accountable
to their respective authorities. The Public Enterprise Act is intended to improve their
accountability to the government and the public. The Act requires these entities to
submit the following documents to the Minister:
� half-yearly report within two months after the end of the first half of the year or
later after consulting the Minister responsible;
� draft annual report and the unaudited financial statement within three months
after the end of each financial year and
� annual report with audited financial statements within five months after the end
of each financial year.
Analyses of Earnings Management Practices in SOEs
19
The SOEs financial accounts are audited annually in accordance with the provisions of
the Companies Act and any other legal requirements. They prepare a Corporate Plan,
which specifies forecasts relating to the current and the following financial year's profit
and loss account, balance sheet, sources and applications of funds, cash flows and a
statement of the assumption on which the forecasts are based. The respective Boards
ensure that the entities act in accordance with their Corporate Plan. Additionally, each
SOE prepares a Statement of Corporate Intent. This is an annual document summarising
the main elements of the SOE's Corporate Plan and other matters set out in the Public
Enterprise Act. The Statement lists the financial and non-financial performance targets
for the relevant year. Furthermore, all SOEs prepare an Employment and Industrial
Relations Plan, which articulates the major employment and industrial relations issues
for the entity.
The government provides equity and debt finance to SOEs but receives returns only
from profitable ones. As of 2002, the government required the SOEs to achieve a 10%
benchmark required rate of return on shareholders funds, which is used as a financial
performance indicator for successfully operating SOEs. Government also provides
substantial amount of grants to SOEs. Not only this, as the owner, the government is
often explicitly required to guarantee loans for these enterprises. Although SOEs
generate revenue for the government and are a major source of employment for the local
population, many enterprises often incur large losses. According to the Minister for
Commerce, losses from SOEs are around $20 million a year, including forgone revenue
(Ministry for Commerce, 1998). He further states that if CSAs had paid income tax on
profits between 1988 and 1995, government would have received an additional $50
million. The financial performance of the GCCs and the CSAs over the recent years is
reflected in Table 3.2.
Chapter 3: Earnings Management in SOEs
20
Table 3.2 Financial Performance of State Owned Enterprises
2001 2002 20031
Net Profit Before Tax ($m)
GCC 2.36 2.07 1.00
CSA 2.66 2.05 7.55
Total Assets($m)
GCC 330.51 358.17 391.85
CSA 702.91 748.37 758.84
Return on Assets (%)
GCC 2.87 3.88 2.19
CSA 0.67 1.86 2.14
Shareholders Funds($m)
GCC 107.01 294.68 308.13
CSA 359.69 477.38 487.89
Return on Equity (%)
GCC 9.43 6.43 4.44
CSA (2.63) 1.89 2.29
(Adapted from Ministry of Public Enterprises and Public Sector Reform Annual Report 2004)
The major factors contributing to the poor performance of most SOEs are identified by
the Ministry of Public Enterprises and Public Sector Reform as follows:
� multiple and often conflicting objectives such as commercial, social and
regulatory;
� there is lack of commercial orientation;
� continued reliance on government for financial support;
� lack of accountability on performance leading to a lack of incentive for improved
efficiency;
� government intervention in the decision making for public enterprises;
� protection from competition so there is lack of incentive for improving
performance and
� with regard to the FEA, meeting universal service obligations2.
(Ministry of Public Enterprises and Public Sector Reform Annual Report, 2000:6)
1 Figures beyond this are not available. 2 This is an additional item added to the list as FEA used this factor to justify its reduced profitability (see Chapter 6).
Analyses of Earnings Management Practices in SOEs
21
3.3 Possibilities of Earnings Management in SOEs
3.3.1 Economic Incentives
While Chapter 2 discussed several economic incentives for earnings management in the
private sector, the following section aims to relate those incentives to the SOEs in Fiji.
Although, management compensation is one of the common incentives in the private
organisations, it is not highly prevalent in the public entities. However, in order to
improve performance, some SOEs have implemented performance based contracts, for
instance, the Housing Authority in 1993. Therefore, the executive management of these
entities have incentives for earnings management. Further, the extent of the borrowing
cost incentive in the public sector is not very clear. No doubt, SOEs require large
amount of debt finance from the local as well as international lending agencies, who
would impose certain debt covenants. However, in most of the cases, government
guarantees the debt of these enterprises. For instance, contingent liabilities of
government in relation to government guarantees of SOE borrowings were $213 million
in 2004 (Parliamentary Paper, 2005). These entities may not worry if they violate the
debt covenants. Alternatively, the government as the guarantor of debts may exert
pressure on them to meet the covenants. Therefore, it is not clear if the borrowing cost
would provide an incentive to the SOEs for earnings management.
The incentive of equity offerings does not exist in the SOEs; but only in the publicly
listed firms. Although, management buyout is not a common practice in Fiji,
government considers it as one of the methods for privatising SOEs3. For instance, there
was an attempt of a management buyout of Rewa Rice Limited in 1999 when the
government was planning to sell off its shares. However, the management's offer was
declined. Thus, management buyout is an option, but it has never occurred for an SOE in
Fiji.
3 Other methods of privatisation include the sale of shares or assets, leasing and outsourcing.
Chapter 3: Earnings Management in SOEs
22
Of the incentives discussed in the previous chapter, the one that is highly relevant to
Fiji’s SOEs is that of reducing regulatory cost or increasing regulatory benefit. These
enterprises rely heavily on government assistance, such as grants, subsidies and/or
government guarantees on loans. Therefore, it is expected that SOEs would be inclined
to report conservative earnings so that they can qualify for government grants/assistance.
Pathik's study in 1999 supports this argument. One of her findings was that in the year
the government was planning to withdraw their financial assistance, the Housing
Authority misclassified government grant in its financial reports. As a result of the
misclassification, the Housing Authority was able to report lower earnings as it feared
that large profit would provide the needed rationale to the government to withdraw their
assistance. This is an example of a SOE reporting conservative earnings in order to
continue receiving government grants.
Apart from the private sector incentives, there could be other incentives present in the
SOEs, which may drive them towards earnings management. One of these could be for
purposes of job security. These enterprises are generally characterised as poor
performing entities and therefore, the management is often pressured by the key
stakeholders to perform and failing to do so could cost them their jobs. There have been
instances where the management had been sacked because of poor performance, such as
in Airports Fiji Limited and Fiji Electricity Authority. Thus, to retain their jobs, the
management has the incentive to manipulate financial performance of the organization.
A contrasting feature of the SOEs to the private sector is their social obligations in
addition to the financial objectives. Social objectives would provide reasons to report
conservative earnings, as reporting higher than average profits would instigate a public
scrutiny. Moreover, SOEs that intend to increase the tariff rates (unit cost of supplying
the commodity) would not want to show high profits. For instance, the Fiji Electricity
Authority planned to raise their tariff rates but they could not justify the increase with
high profits as this will be objected by the public (see Chapter 6 for details). Finally,
entities would attempt to manage earnings downwards when there are strong trade
unions. Trade unions with strong negotiating powers will demand wage increments
Analyses of Earnings Management Practices in SOEs
23
when the entity continues to report high profits. Apart from these general incentives,
there could be other entity-specific incentives to practice earnings management, which
would be known by investigating the individual organizations.
3.3.2 Accounting Regulation
The Fiji Institute of Accountants (FIA) was established in 1972 under the provisions of
the Fiji Institute of Accountants Act with the objective of registering accountants and
regulating the practice of accountancy in Fiji. The FIA Council appoints various
committees, one of which is Accounting and Auditing Standards Committee, responsible
for issuing accounting and auditing standards4. This Committee promulgates Fiji
Accounting Standards (FASs) that serve as the basis for preparing financial statements.
The institute is a member of the International Accounting Standards Board (IASB) and
the International Federation of Accountants, and therefore, the Institute's standards are
highly influenced by these governing bodies. Over the period 1976 to 1998, the FIA had
issued 28 standards that were largely based on International Accounting Standards
(IASs). However, Australian and New Zealand accounting standards were also adapted
when these were seen to be relevant to the Fijian society5. In early 1999, the Council
decided to adopt the IASs as the basis for a completely new set of FASs. Consequently,
a two-year comparative review began in November 1999, which concluded that FIA
would apply all extant IASs numbered 1-31 with the exception of IAS 12 with effect
from 1 July 2002. IAS 12 and 32-39 were regarded as guidance standards6. Furthermore,
in order to achieve international standards in accounting, the FIA will be adopting
International Accounting and Financial Reporting Standards (IFRS) from 1st January
2007. The adoption was initially set for January 2006 but was deferred when the Council
4 Detail explanations of the standard setting process in Fiji are provided in Chand (2003) and Pathik (1999). 5 These include FASs 2, 7, 8,12,14,17, 101, 102 and 103. Since 2002, only FASs 101, 102 (now withdrawn) and 103 were drawn from Australia and New Zealand. 6 FASs 15, 19 and 26 also had guidance status.
Chapter 3: Earnings Management in SOEs
24
realised that a longer lead time was required to obtain IFRS compliant comparative
figures than was first thought7.
Thus, the accounting standards applied in Fiji are almost entirely based on IASs, which
provides flexibility in financial reporting. Also, it is worth pointing out that IASB
applies principle-based approach in standard setting process. Consequently, the FAS
requires the accountants to exercise professional judgment. One such example is the
accounting standard on leasing that requires the accountant to use qualitative
benchmarks to distinguish between operating and financing lease instead of quantitative
guidelines. Furthermore, IASs are developed mainly on the basis of accounting
standards that prevail in the developed countries of the UK and USA. Therefore, if the
IAS or equivalent in these countries provided opportunities to manage earnings then the
use of the same standards is also likely to enable such behaviour in adopting countries,
like Fiji.
The Surveillance Panel of the FIA is responsible for monitoring compliance with the
accounting standards. However, so far there has been minimal effort to enforce
compliance. This is firstly because the law does not mandate the accounting standards so
on several occasions, the Panel’s observations are ignored. Further, the Panel cannot
always access all the reports they wish to review. The existence of such behaviour is
perhaps a useful insight into ambivalence within the profession as on one hand the
profession enforces compliance and on the other hand it is not helping to achieve the
desired goal. Secondly, the committee mainly has to rely on volunteer members. This
problem is further compounded as the country loses qualified skilled personnel through
emigration. Also, most of the members who possess the necessary expertise are
precluded from the membership into the Panel due to their workplace associations.
Earnings management practices mainly occur within the boundaries of accounting
standards so if the Panel is weak in monitoring compliance, it would be weaker in
detecting manipulations. This provides additional opportunities for earnings
management in Fiji. 7 IFRS 1 First-Time Adoption of International Financial Reporting Standards requires the compilation of an IFRS compliant balance sheet two years prior to formal adoption.
Analyses of Earnings Management Practices in SOEs
25
Barton and Simko (2002) argue that although generally accepted accounting practices
require managers to make numerous judgments and assumptions to report their firm's
performance, managers do not have unlimited discretion to do so. The accounting
regulators such as the Securities and Exchange Commission and the Financial
Accounting Standards Boards, along with professional bodies like the American Institute
of Certified Public Accountants, provide rigorous guidelines and oversight. On the
contrary, Fiji does not have such bodies that would limit managers' use of discretion.
This suggests that there is even higher scope for earnings management in Fiji.
3.3.3 Cultural influences
The environment in which businesses operate consists of four elements: legal, political,
economic and cultural (Miroshnik, 2002). Culture is “the collective programming of the
mind which distinguishes the members of one human group from another” (Hofstede
1980: 25, c.f. Greer and Patel, 2000). Different cultural environments differ in their
norms and behaviour. According to Dressler and Carns (1969), culture allows us to
communicate with others through a common language. It gives us standards for
distinguishing between conditions such as what is right or wrong. It makes it possible to
anticipate how others in our society are likely to respond to our actions. (c.f. Miroshnik,
2002).
Studies have recognised the importance of cultural factors in shaping a country’s
accounting system (see Fechner and Kilgore, 1994). Some have argued that accounting
in fact is determined by culture. Accordingly, Hines (1992) states that accounting
language is a reflection and social production of a particular set of cultural values and
conceptions of reality (c.f. Greer and Patel, 2000). Culture tends to influence the
accounting standard setting process, behaviour of entrepreneurs and managers,
accountability issues, alternate accounting controls and many others.
Chapter 3: Earnings Management in SOEs
26
Hofstede’s (1980) predominant approach has identified four distinct cultural dimensions.
These include Uncertainty Avoidance, Power Distance, Individualism and Masculinity.
Uncertainty Avoidance measures the extent to which people in a society feel threatened
by ambiguous situations, and the extent to which they try to avoid these situations by
providing greater career stability, establishing more formal rules and rejecting deviant
ideas. It reflects the willingness to undertake risks and the preference for security. Power
Distance relates to the degree to which cultures accept a more autocratic structure or the
degree to which participation is favoured. It measures the extent to which the less
powerful members of societies accept the unequal distribution of power. The
Individualism dimension addresses the issue of interdependency between a society and
its individuals. It measures the extent to which people feel they are supposed to take care
or be cared for by themselves, their families or organizations they belong to. It considers
the relative importance placed on the individual as opposed to the group. Masculinity
measures the extent to which a culture is conducive to dominance, assertiveness and
acquisitions of things. It considers the relative importance placed on income recognition
and advancement.
Countries can be differentiated on the basis of the cultural dimensions identified by
Hofstede. Although these dimensions are diminishing, Chand (2003) argues that there
exist cultural differences between Fiji and the developed countries. The dimension that
differentiates Fiji from the developed countries and has implications for earnings
management is power distance. Relative to the countries like the UK and USA, Fiji is
considered to be a large power distance society. This implies that the Fijian society
accepts the hierarchical order in which everybody has a place and they do not require
justification. Thus, the management knows that financial reports prepared by them
would be accepted as given without questioning. Hence, the cultural value of large
power distance reduces the chances of detecting any creative accounting practices.
Similarly, Chand and White (2006) argue that in general Fiji society exhibits large
power distance. However, contrary to their expectation, they found that accountants are
comfortable in making judgments. Their results are surprising as it is not a consistent
behaviour given a range of cultural factors. Nonetheless, they suggest that the
Analyses of Earnings Management Practices in SOEs
27
accounting community is 'accultured'. That is, it has its own professional culture rather
than the culture of the various ethnic groups they are drawn from. There is, therefore, a
possibility that instances of earnings management will not be challenged in Fiji,
irrespective of whether the accountant's judgments has been exercised to facilitate or
inhibit it.
3.4 Conclusion
There are possibilities for earnings management in Fiji’s SOEs. Although, the economic
incentives are not the same as in the previously examined publicly listed companies,
there exist incentives in the SOEs that could drive them towards earnings management.
Further, the weak accounting regulation and the cultural influences produce an
environment that is conducive to earnings management.
Chapter 4: Methodology
28
CHAPTER FOUR
METHODOLOGY
4.1 Introduction
This chapter briefly reviews the empirical models used in the literature to identify
earnings management practices and evaluates their applicability to Fiji. It then justifies
and explains the research method employed in this thesis.
4.2 Empirical Models of Earnings Management
Earnings management has been empirically investigated in many studies using the
variants of the accruals models (see Phillips et al., 2003; Frankel et al., 2002; Beneish
and Vargus, 2002; Heninger, 2001; Natarajan, 1999 and Dechow et al., 1995)8. The
traditional models by Healy (1985) and DeAngelo (1986) are simple since total accruals
are used to measure earnings management. The Healy Model assumes that earnings are
manipulated in a systematic manner over the years. Therefore, it calculates non-
discretionary accruals9 using the average of total accruals scaled by lagged total assets
during the estimation period. The DeAngelo Model calculates non-discretionary accruals
by taking a ratio of lagged total accruals to lagged total assets on the assumption that
non-discretionary accruals are constant. However, Kaplan (1985, c.f. Dechow et al.,
1995) strongly objects to this assumption and suggests that the very nature of accrual
accounting dictates that the level of non-discretionary accruals varies in response to
changes in the economic conditions.
8 Dechow et al. (1995) and Bartov et al. (2000) provide a detailed discussion of these models. 9 Total accruals is the difference between earnings and net operating cash flows. Discretionary accruals is the difference between total accruals and nondiscretionary accruals.
Analyses of Earnings Management Practices in Fiji's SOEs
29
Consequently, the Jones Model (1991) (JM, henceforth) relaxes this assumption by
controlling for the variables that cause changes in non-discretionary accruals. The
controlling variables are revenue and gross property, plant and equipment. Revenue
controls the economic environment of an entity as it is an objective, although not
completely exogenous, measure of the entities operations before managers’
manipulation. Gross property, plant and equipment controls for the portion of total
accruals related to non-discretionary depreciation expense. Instead of using total
accruals as a measure of earnings management, the JM decomposes total accruals into
discretionary and non-discretionary components, where the former are used to explain
earnings management. The calculation of discretionary accruals is a two-step procedure.
The first step is to estimate the non-discretionary component of total accruals using the
following equation:
tt
t
t
t
tt
t
A
PPE
A
REV
AA
TA εααα +
+
∆+
=
−−−− 13
12
11
1
1 (1)
where
tTA = total accruals
1−tA = total assets
tREV∆ = change in revenue
tREC∆ = change in net receivables
tPPE = gross property, plant and equipment
t = time subscript
The second step is to compute the error term from the above equation, the difference
between total and non-discretionary accruals, which represents the discretionary
accruals. Positive (negative) error term means income-increasing (decreasing) earnings
management.
Chapter 4: Methodology
30
Subramanyam’s (1996) study shows that the cross-sectional JM is better specified than
its time-series counterpart (c.f. Krishnan, 2003). He finds that in the cross section model
the accuracy of the estimated coefficients is higher due to a large number of degrees of
freedom. Moreover, cross sectional models do not impose a requirement for long series
of data, hence, it yields a lower risk of survivorship bias relative to a time series model.
The JM is further modified by Dechow et al. (1995) to improve the predictability of
discretionary accruals. The Modified Jones Model (MJM, henceforth) is specified as
follows:
tt
t
t
tt
tt
t
A
PPE
A
RECREV
AA
TA εααα +
+
∆−∆+
=
−−−− 13
12
11
1
1 (2)
where
tREC∆ = change in net receivables
The only difference in this model compared to the JM in (1) is that the change in
revenue is adjusted for the change in receivables in the event period. While the JM
assumes that no discretion is exercised over revenue, MJM assumes that all changes in
credit sales results from earnings management. This is because it is easier to manage
earnings by exercising discretion over credit sales revenue than cash sales. Thus, the
MJM assumes that the change in revenue less the change in accounts receivable is free
from managerial discretion.
Bartov et al. (2000) evaluated the ability of seven accruals models to detect earnings
management and concluded that the cross sectional JM and the cross sectional MJM
outperform their time series counterparts10. However, these observations are naïve in
light of recent developments in time series techniques such as unit root and
10 The models they examined include the DeAngelo Model, Healy Model, JM (time series and cross sectional), MJM (time series and cross sectional) and Industry model.
Analyses of Earnings Management Practices in Fiji's SOEs
31
cointegration, which are on the forefront in applied econometrics. Furthermore, the cross
sectional approach is not entirely free from shortcomings.
The major criticism of the accruals models is the degree of sophistication involved in
separating discretionary and non-discretionary accruals. The models reflect
measurement error, which arises partly because of the misclassification of non-
discretionary accruals as discretionary accruals (see Bernard and Skinner, 1996).
Findings of Dechow et al. (1995) and Guay et al. (1996) suggest that the discretionary
accrual from the JM is imprecise due to the large variation in estimated coefficients.
Nevertheless, recent studies (Krishnan, 2003; Heninger, 2001 and Teoh et al., 1998)
continue to employ the JM and MJM, despite their limitations.
The methodology adopted in this thesis can be distinguished from earlier studies. Instead
of using the empirical models, it employs a qualitative approach, which is discussed in
Section 4.4. The choice of the methodology, however, was not made on the merits and
de-merits of the two methods but on its applicability to Fiji’s environment.
Unfortunately, the lack of quality and consistent data for Fiji does not permit a reliable
statistically based study to be undertaken. It should be noted that the smaller the sample
size, the greater the impact of one particular observation on the coefficient estimates.
Meeting the data requirement is not a problem for developed countries as the size of
such economies, together with the number of entities and age of entities, provide the
model with the required data. Even the cross sectional models, which require fewer data
points than time series models, cannot be applied in Fiji. Cross sectional models requires
a minimum of six observations within each group in each year. In Fiji's case, if the
entities in the sample were grouped, there would hardly be six entities of similar nature
in one group. Further, consistent data on all entities for the same sample period are
hardly available as most of these organizations fail to produce annual reports.
A similar situation was also faced by Liu and Lu (2003), when examining earnings
management in another developing country, China. Instead of discretionary accruals,
they use total accruals as a proxy for earnings management because they argue that there
Chapter 4: Methodology
32
is no reliable model to estimate discretionary accruals in the Chinese listed companies.
Consequently, Liu and Lu conclude that a well-received model in the developed
countries may not be applicable in other economies without some adjustments.
4.3 Data
The required data for this study are primarily collected from the published annual reports
of the entities. Many of the SOEs do not produce annual reports every year but only
prepare financial statements. While this is sufficient for empirical models, the qualitative
approach requires the financial data together with the narrative information in annual
reports. Thus, to be taken in the sample, the chosen entities should publish their annual
reports periodically. Furthermore, the investigation period is 1990 to 2004 as it is
believed that the time frame is large enough to furnish the necessary materials that
would be needed to undertake this study. Therefore, the entities selected should be in
operation over this period. Of the 16 SOEs in Fiji (see list on page 17), only four meet
these criteria - the Fiji Electricity Authority, Housing Authority, Public Rental Board
and the Unit Trust of Fiji.
Given that the qualitative approach employed in this study has not been widely
employed in previous studies, it is important that more than one case is used so that the
applicability of the approach can be tested. Robson (2002) points out that a common
misconception for using more than one case study is for the purpose of gathering a
‘sample’ to make generalization about the ‘population’. However, more than one case is
used not to make a statistical generalization but analytic generalisation. They are chosen
to demonstrate that the methodology can be applied to cases in general, rather than to a
single instance. Thus, it is possible to study all the four entities as they would provide
better results. Nonetheless, knowing that a case study requires an in-depth examination
of the chosen organization, four cases would become unmanageable and beyond the
scope of this thesis. In light of these factors, two cases will be used in this study.
Analyses of Earnings Management Practices in Fiji's SOEs
33
A review of the annual reports of Unit Trust of Fiji shows that there is not much scope to
identify the possibilities of earnings management as there are not many accounts listed
in the financial statements. So this entity is eliminated. The Housing Authority and
Public Rental Board are similar organisations as both are involved in the housing
market. Also Public Rental Board was established by taking over some of the functions
of the Housing Authority. Therefore, it is better to study the larger organization of the
two, which is Housing Authority. After eliminating Unit Trust of Fiji and Public Rental
Board, the two entities chosen are Housing Authority and the Fiji Electricity Authority.
It should also be noted that out of the four, these are the two largest entities, in terms of
total assets and turnover. Hence, they make appropriate cases to be studied.
4.4 Research Method
Each organization's annual report over the review period was thoroughly read in order to
understand their operations, organisation structure, accounting environment and
products and services. Subsequently, the checklists developed by Mulford and Comiskey
(2002) were applied to detect instances of earnings management, which provide a useful
and practical classification scheme to do so without employing empirical modeling.
They state that such a scheme assists the financial report readers to be more focused in
their search for items that may indicate that the reported results may not be true. Mulford
and Comiskey provide the following categories to detect earnings management:
(i) Recognising premature or fictitious revenue
(ii) Aggressive capitalisation and extended amortisation policies
(iii) Misreported assets and liabilities
(iv) Using operating cash flow to detect earnings management
Based on these categories, they prescribe a guideline in the form of checklists. The
checklists, constructed after considering the international collapse of a number of
organisations, provide a list of items that should be considered in detecting earnings
Chapter 4: Methodology
34
management. These checklists are appended in Appendix 1 and are briefly described
below.
Checklist 1 is based on recognising premature or fictitious revenue. Given the
importance of revenue items in the income statement, it cannot be denied that they are
the most obvious items subject to manipulation. For the purpose of detecting earnings
management, it is not important to identify the difference between premature and
fictitious revenue as both lead to misreported results. Checklist 1 is divided into three
parts; revenue recognition policy, physical capacity and accounts receivables.
In order to identify possible earnings management practices related to revenues, one
needs to understand the entity’s revenue recognition policy. Detailed information on this
is provided in the notes section of the annual report and a careful examination of this
would be useful. To assist in identifying premature and/or fictitious revenue, one also
needs to consider the physical capacity of the entity to generate revenue, which is the
focus of the second part to the checklist. Analysing the physical capacity could indicate
whether the entity has the capacity to generate the reported revenue. Possible measures
of revenue per physical capacity could include revenue per employee, revenue per dollar
of property, plant and equipment or revenue per total assets.
Generally, with fictitious or premature revenue, there would be no associated cash
inflow. Accordingly, a balance sheet account is increased and an obvious one is accounts
receivable. Thus, one needs to be mindful of unusual increases in accounts receivable as
these accompany questionable revenue. However, knowing that readers would look for
such relationships, managers may use other balance sheet accounts as a potential storage
for misreported revenue. These could be property, plant and equipment and prepaid
expenses11. Therefore, the final section of the Checklist 1 analyses accounts receivables
and other asset accounts.
11 It should be noted that this evidence would be available only if it had escaped the notice of the auditors. If auditors had already sighted this trend, they could have taken appropriate action(s).
Analyses of Earnings Management Practices in Fiji's SOEs
35
Checklist 2 focuses on an entity’s capitalisation and amortisation/depreciation policies.
The first part of the checklist analyses the capitalisation policies as they have direct
effects on earnings and assets. Some useful analytical tools that could identify
aggressive capitalisation policies include:
1. Review of the entity’s capitalisation policy
2. Carefully consider what the capitalised costs represent
3. Analyse whether the entity has been aggressive in its past capitalisation policy
4. Search for costs capitalised in stealth
In order to find out whether the entity is taking an aggressive approach of capitalising
costs, it would be useful to compare their capitalization policy with that of the
competitor's and the industry's. Although it is useful to compare with competitors, it is
possible that the competitors are also aggressively capitalising costs. Furthermore, if it
was found that the entity had adopted an aggressive capitalisation policy in the past, then
it is reasonable to assume that it will continue to practice this in future. Thus, caution is
required while examining capitalised costs. Needless to say, entities that have been
aggressive in capitalising costs will seek to mask their actions. Therefore, it is important
to go beyond the accounting policy notes and the capitalised expenses to several other
accounts reported in the balance sheet, such as accounts receivable, inventory, prepaid
expenses, property, plant and equipment and other assets. To detect if aggressive
capitalised costs have been transferred to these accounts, it is worthwhile to examine the
relationship of these accounts to revenue and their rate of change relative to that of
revenue.
The other section of checklist 2 deals with extended amortisation and depreciation
policies since these areas have potential scope for manipulation. In order to detect
extended amortisation and/or depreciation policies, the checklists provide two useful
steps. Step 1 calculates average amortisation period for the entity’s depreciable asset
base. The direct approach to determine extended amortisation/depreciation period is to
check the disclosure of the useful lives of non-current assets. This disclosure, however,
Chapter 4: Methodology
36
in many instances is vaguely stated. For example, the Housing Authority disclosed the
following in its Annual Report (2004:38):
NON CURRENT ASSET USEFUL LIVES
Building 28 - 66 years
Plant and equipment 3 - 5 years
This piece of information makes it difficult to identify either what was the
amortisation/depreciation period or whether the disclosed years had been used. In light
of this problem, an alternative would be to compute the average
amortisation/depreciation period of the entity's collective depreciable base. Step 2
searches for extended amortisation periods in prior years as entities found to be behaving
as such in the past are likely to leave themselves open to such problems in future as well.
Checklist 3 identifies any misreported assets and liabilities as these have direct links
with earnings. The first part of the checklist analyses assets that are not subject to annual
amortisation/depreciation, such as accounts receivable and inventory. Assets subject to
periodic amortisation/depreciation have been excluded as they were examined in the
earlier checklists.
As noted in Checklist 1, premature and/or fictitious revenue improperly accumulates
accounts receivable. Consequently, it grows faster than revenues, yielding accounts
receivable days much higher than normal for the entity. However, even when revenues
are properly recognised, earnings could still be temporarily increased. This could be
achieved through an improper valuation of accounts receivable, as these are reported in
the balance sheet after adjusting for provision for doubtful debts. Entities that want to
misreport earnings to suit their objectives could manipulate the provisions.
Another asset account is inventory, the cost of unsold goods reported in the balance
sheet, which upon sales is transferred to the income statement. Misreported inventory
could lead to misreported cost of sales and therefore misstated net income. There are a
Analyses of Earnings Management Practices in Fiji's SOEs
37
number of ways to manipulate inventory. A direct approach is to misreport the physical
quantity of inventory while the second is to misreport its dollar value without altering
the physical count. The third is to postpone a needed or make an early write-down for
inventory. No matter how this is done, the same steps could be used to detect
manipulation. For example, with these anomalies, there would be an unexplained change
in inventory that outweighs the change in revenue. Moreover, inventory turnover days
will reach levels that are different from the competitors and other entities in the industry.
Checklist 3 then proceeds to examine liabilities such as accrued expenses and accounts
payable. To identify misreported accrued expenses, it is important to carefully review
the trends in such accounts. A useful approach is to compare the growth rate in accrued
expenses with that in revenue. For instance, if revenue increases faster than accrued
expenses, it implies that accrued expenses could be undervalued. Another useful test is
to compare selling, general and administrative expenses as a percentage of revenue with
prior periods. It needs to be considered whether an improvement in the expense ratio
reflects true operating efficiencies or a failure to accrue operating expenses properly.
Accounts payable increases mainly due to credit purchase of inventory. Often when
accounts payable are misstated, so is the inventory purchase. Hence, these lead to
misreported cost of goods sold and net profit. To detect unusual or unexpected changes
in accounts payable, it is useful to compute accounts payable days. Unexpected changes
in gross profit margin may also be an indicator that an inappropriate adjustment was
made to accounts payable. However, if the misstatement in accounts payable and
inventory purchases are of an insignificant magnitude, the effect on accounts payable
days and gross margin may not be noticeable. Therefore, a more useful tool would be to
compare the growth rate in accounts payable with that in inventory.
The fourth checklist deals with how the cash flow from the operating activities could be
used to detect earnings management practices. In order to carry out a thorough
examination of such practices, one has to study the entity’s cash generating abilities. It is
generally agreed that operating cash flow is a key indicator of the entity’s ability to
Chapter 4: Methodology
38
generate sustainable cash. Managing reported earnings do not change operating cash
flow therefore examining the relationship between earnings and operating cash flow is
helpful. To use operating cash flow, one needs to eliminate nonrecurring cash items.
Such items include income taxes on items classified as investing or financing activities,
cash effects of purchases and sales of trading securities for non-financial entities,
capitalised expenditure and other nonrecurring cash flows. Cash flow from operating
activities is not useful on its own but when it is used in conjunction with income from
continuing operations adjusted for nonrecurring events. Using these two variables, the
adjusted cash flow-to-income ratio is calculated in the following manner:
Adjusted cash provided by continuing operations
Adjusted cash flow-to-income ratio =
Adjusted income from continuing operations
This ratio is useful in identifying discernible trends over a period of time. A decrease in
the ratio indicates that earnings are growing faster than operating cash flows. Such an
outcome necessitates a closer examination to determine the underlying reasoning, using
the steps outlined in earlier checklists. It could be possible that management may have
employed earnings management practices to temporarily boost earnings. On the other
hand, an increase in this ratio caused by an increase in operating cash flow in excess of
an increase in earnings also needs to be investigated. Such an outcome may be the result
of management employing income-decreasing earnings management practices.
The above-discussed checklists would identify any unexplained, unusual behaviour,
which would then be analysed on a case-by-case basis. Analysis will involve checking
whether the item is consistent with generally accepted accounting practices and if it was
motivated by any incentives.
Analyses of Earnings Management Practices in Fiji's SOEs
39
4.5 Overlap Between the Two Approaches
Through the scrutinisation of the accruals model and the preceding in-depth analysis of
the qualitative approach, it can be seen that the two approaches overlap. Although, the
two methods approach earnings management differently, the variables used in the
models are similar. To illustrate this, the MJM is reproduced here.
tt
t
t
tt
tt
t
A
PPE
A
RECREV
AA
TA εααα +
+
∆−∆+
=
−−−− 13
12
11
1
1 (2')
On the left-hand side of the above equation is total accruals, which according to the
existing literature can be calculated in one of the following ways:
1. Total accruals = net income before _ net cash flows
extraordinary items from operations
2. Total accruals = change in current assets - change in current liabilities - change in
cash + change in short-term debt – depreciation
3. Total accruals = - (decrease in accounts receivable + decrease in inventory +
increase in accounts payable + increase in taxes payable + net
change in other current assets + depreciation)
The variables that are used to calculate total accruals are also used in the checklists. The
first way of calculating total accruals uses net income before extraordinary items and net
cash flows from operations. These two variables are used in the checklist 4 to calculate
cash flow-to-income ratio. Furthermore, items such as accounts receivable, inventory,
accounts payable are used in the third checklist, which analyses misreported assets and
liabilities. Depreciation expenses are also used in the qualitative method, specifically in
the checklist 2.
Chapter 4: Methodology
40
The right-hand side of the equation considers three important variables, revenue,
receivables and property, plant and equipment. The qualitative method also uses the
same variables; revenue figures to identify any premature and fictitious revenue and
property, plant and equipment to detect if the entity had the physical capacity to generate
the reported revenue. Receivables are used on both sides. It is used on the left-hand side
to calculate total accruals and on the right as an independent variable. The qualitative
method analyses receivables to determine its trends over the years and its relationship
with revenue.
Furthermore, the qualitative models not only analyses the same variables but also
considers all other factors that affect a particular variable. This is highly informative and
useful. The empirical models fail to grasp the complexities of accruals actual behaviour.
For instance, they concentrate on the dollar value of revenue only, but the qualitative
approach analyses revenue values, revenue recognition policy, credit policy, related
party transactions and the physical capacity to generate the reported revenue. It could,
thus, be said that the checklists analyse all the aspects of a particular variable.
Additionally, this method assists in identifying the specific accruals used to manage
earnings, which cannot be obtained using the accruals models as it aggregates the effects
of all accruals.
However, the qualitative approach to earnings management has equally, if not more,
significant weaknesses than accruals models. Firstly, this method has not been
extensively used in the literature since it was recently developed. Hence, there is a lack
of practical guidance and means of comparison. Secondly, while accruals models
objectively determine the presence of earnings management, results from the qualitative
approach are subjective to the interpretation of data based on the researcher’s ability.
The qualitative method will not yield a numerical value, on which to base and compare
results. It provides detailed information that should be interpreted cautiously. Thirdly,
this method is only useful in providing indications of earnings management rather than
providing direct documentary evidence. Finally, it could be possible that the qualitative
method may not be able to identify all instances of earnings management in the two
Analyses of Earnings Management Practices in Fiji's SOEs
41
entities, particularly those which were more creatively practiced. Nonetheless, it is
useful, in cases when the accruals models cannot be applied due to data limitations,
particularly in developing countries such as Fiji.
4.6 Conclusion
The accruals models for detecting earnings management cannot be used in Fiji due to
data limitations. As a result, this thesis employs the checklists designed by Mulford and
Comiskey to identify possible instances of earnings management in the two SOEs. It is
worth pointing out that studying earnings management is a difficult task, no matter
which method is applied. Given the qualitative nature of the study, all instances are
analysed on a case-by-case basis. The following two chapters provide details of this
analysis of the two entities.
Chapter 5: Evidence of Earnings Management Practices in the HA
42
CHAPTER FIVE
EVIDENCE OF EARNINGS MANAGEMENT PRACTICES IN
THE HOUSING AUTHORITY
5.1 Introduction
This chapter explores earnings management practices in the first entity, the Housing
Authority (HA). It begins with an overview of the HA, which provides a brief corporate
background and describes its lending activities and the accounting environment within
which it operates. It then proceeds to identify and analyse the possible instances of
earnings management and the economic incentives that may have motivated these
manipulations over the review period.
5.2 Overview of the Housing Authority
5.2.1 Background
In the mid 1950s, the government realised that housing demands by low-income earners,
particularly in the greater Suva area, had far outgrown the available accommodation.
There were concerns that the projected demand for housing in the future could not be
entirely met by the private builders. Therefore, a resolution was passed in the Legislative
Council in September 1955 calling for the establishment of a statutory authority to deal
with these housing problems. Subsequently, the HA was established in 1955 under cap.
267 of the Housing Act and became an operating entity in 1958. The responsibilities of
the Authority were based on recommendations of a report prepared after studying low
cost housing in the West Indies. The report made recommendations on the approach to
Analyses of Earnings Management Practices in Fiji's SOEs
43
public housing that was best suited to Fiji. These recommendations formed the substance
of the policy on which the Authority commenced with its housing programmes12.
The HA launched its first scheme, the Cash Loan Scheme, in October 1958. It provided
loans to the low-income group living in Suva and Lautoka to buy, build, extend or
substantially alter their houses. The loans were advanced to people, who were on regular
employment with average weekly income of less than $24. Towards the end of the year,
the HA diversified into building houses under its Home Purchase Plan. In 1964, the
Rental Flat Scheme was introduced to assist people who could not meet the loan
repayments under the Cash Loan Scheme or the House Purchase Plan. This scheme
operated on rental subsidy, where the tenants contributed 15% of their gross weekly
income towards the rent while the difference between the economical rent and the
tenant’s contribution was subsidised by the government13.
The HA was on a verge of financial collapse in the late 1980s. Consequently, with the
assistance from the Asian Development Bank (ADB) and World Bank, a team of experts
were engaged to study the functions and financial and administrative problems of the
Authority. On the basis of their findings, the HA developed a major Revitalisation Plan,
which involved restructuring of its rental housing program, finances and operations and
taking corrective actions to turn the HA into a viable and dynamic entity. The Plan was
implemented from the beginning of 1989 as the government amended the Housing Act to
allow for the establishment of Public Rental Board (PRB) to take charge of the HA’s
rental operations. A special project team was appointed to implement the transfer of
assets and liabilities to PRB and assist this new entity to take charge of its
responsibilities. Following the transfer of rental operations, the HA developed its own
corporate plans and established and documented its objectives. Its new role was defined
as follows:
12 The government sent Mr. Bain to the West Indies to study their public housing programmes. The details of his recommendations are explicitly described in the HA’s 2001 Annual Report. 13 Economical rent is the level of repayment to meet the loan uplifted to build the unit.
Chapter 5: Evidence of Earnings Management Practices in the HA
44
…to produce and finance through long-term mortgages, shelter for families who
would not otherwise have available to them the kinds of living accommodation they
would both desire and afford (The HA Annual Report, 1989:10).
The Design and Build programme was introduced in 1992 to cater for the increased
housing demands. This was a major change from the previous practices where the
contractors had to build houses as per HA's standard designs. With the new system, the
builders tendered their own designs and were required to complete the houses within
four months from the award of a contract. This system contributed to greater efficiency
in the HA’s construction activities. In 1996, there was a turnaround in the HA's status as
it was declared a CSA, which required it to operate on commercial basis and provide
returns to the government. While pursuing commercial objectives with a major focus on
profit, the HA developed and constructed houses that were affordable but of poor
quality. Defective houses led to difficulties in sales and customer dissatisfaction.
Following complaints from a number of tenants about the quality of houses, the
Authority paid for their repairs and maintenance. The HA began outsourcing the
designing and building activities from 1997 so that it could concentrate only on the
production of land lots and providing home loans. With the initial vision to provide
affordable housing to the middle and low-income earners, the HA has in recent years
expanded its target market by providing mortgage financing to high-income earners as
well.
The HA Board comprises of a Chairman and five other members, all appointed by the
Minister for Housing. The Chief Executive of the HA is an ex-officio member to the
Board. Over the study period, there have been many changes in the Board’s
composition. All executive management appointments in the recent years have been
conducted through a transparent recruitment process and appointees were allocated jobs
within a competency-based framework. A job evaluation exercise was undertaken in
2003 to appropriately remunerate all staff and for the first time in HA's history, the
wages and salaries for its entire staff came on par with the market.
Analyses of Earnings Management Practices in Fiji's SOEs
45
5.2.2 Lending Activities
Providing home loans is one of the core activities of the HA. The financial products of
this package are cash loans, house sale, land sale and village scheme. Cash loans are
provided for purposes of house repairs, fencing, personal or family commitments. Since
1998, new products such as soft loans, home comfort loans and quick loans were also
included in this package. The HA faces stiff competition from other institutions that also
provide home loans. For instance, in 1999 the HA's average lending rate was 9.71%
while the commercial banks charged an average rate of 7.59%. This resulted in a
massive loss of performing customers to its competitors. Most of these customers were
from middle-income level with no arrears and with repayments made directly from
salary deductions, implying that the HA lost customers who were less risky. Also, the
high interest rates become unsustainable to low-income earners who do not qualify for
commercial bank loans. Consequently, loans began to fall into arrears and contributed
towards high level of non-performing loans for the Authority, which averaged around
20% of total mortgage portfolio.
In order to competitively serve its client base, the HA has to be attractive in terms of its
lending rate. This is quite challenging for the Authority as it faces difficulties in
lowering the rate of interest. One of the major contributors of the high interest rate
charged by the HA is its borrowing cost. Table 5.1 compares the interest rate spread of
the HA with commercial banks. Commercial banks have higher interest margin than the
HA despite charging a reasonably lower lending rate. This is achieved due to its lower
average cost of borrowing as the commercial banks are able to source funds at a
significantly lower rate than the HA.
Chapter 5: Evidence of Earnings Management Practices in the HA
46
Table 5.1 Interest Rate Spread of the HA and the Commercial Banks (%)14
Year HA All Commercial Banks
Average Lending Rate
Average Borrowing
Rate Spread
Average Lending Rate
Average Borrowing
Rate
Spread
1999 9.71 6.81 2.90 7.59 1.69 5.90
2000 9.59 7.74 1.85 7.49 1.81 5.67
2001 9.65 6.72 2.94 7.51 1.54 5.96
2002 7.53 5.63 1.90 6.77 1.12 5.66
2003 7.43 5.51 1.91 6.74 0.88 5.86
2004 7.69 5.19 2.50 6.62 0.89 5.73
(Adapted from Jayaraman and Sharma 2003, the HA Annual Reports 1999-2004 and author’s calculations)
Additionally, given the nature of its business, the HA’s operating costs are relatively
higher than its competitors. This is clearly reflected in Table 5.2, which shows that
operating expenses per dollar of income is around 80% for the HA when compared with
50% for commercial banks.
Table 5.2 Operating Costs of the HA and the Commercial Banks
Year HA All Commercial Banks
Income $000
Operating Expenses $000
Operating Expenses Per
Income %
Income $000
Operating Expenses $000
Operating Expenses Per
Income %
1998 8252 6671 80.84 134218 71435 53.22
1999 9517 7081 74.40 153296 85866 56.01
2000 6526 4965 76.08 157671 81948 51.97
2001 9389 8088 86.14 153821 81739 53.14
2002 7886 6478 82.15 152589 81605 53.48
2003 8157 6418 78.68 170496 85016 49.86
2004 9601 6169 64.25 184377 91938 49.86
(Adapted from the HA Annual Reports 1998-2004, RBF Annual Reports 2002 and 2004 and author’s calculations)
Despite high borrowing and operational costs, the HA reduced its lending rate from
11.5% to 6% for all new low-income customers in late 1999 and beginning 2000, this
was extended to all low-income customers (new and existing). The reduction in the rate
was a result of a government directive and an undertaking to the Authority to secure
cheaper funds. The government provided grants to the HA to compensate for the loss in
14 Interest rate spread = Interest earned __ Interest paid
Average interest earning assets Average Borrowings
Analyses of Earnings Management Practices in Fiji's SOEs
47
income resulting from reduced interest rates. Still faced with intense competition, the
HA launched its "Moving on Package" in September 2001, in which interest rates for the
first year were further reduced to 5.99% and 8.35% for all new and existing customers,
respectively. With the introduction of the "Sapphire Package” in 2003, the HA further
lowered the lending rate to 4.45% for the first two years and 7.99% thereafter. Although,
the HA have been able to compete with other lending institutions in the recent years, this
has been possible with the continued financial assistance from the government.
Appendix 2 details the current lending rates of HA and its competitors.
5.2.3 Accounting Environment and Financial Performance
The HA introduced accrual accounting system in 1993 and by the end of 1994 all
transactions and financial reports were prepared on this basis. The financial reports were
audited by the Auditor General until 1989 and since then, KPMG was responsible for its
external audit functions. The KPMG also provides non-audit services since 1995. Table
5.3 illustrates the fees paid to KPMG for auditing and non-audit services.
Table 5.3 Auditors Remuneration
Year Audit fees/ Total fees
(%) Non-audit fees/ Total fees
(%)
1995 62.16 37.84
1996 50.00 50.00
1997 47.73 52.27
1998 30.30 69.70
1999 54.29 45.71
2000 86.36 13.64
2001 86.36 13.64
2002 21.84 78.16
2003 46.43 53.57
2004 86.67 13.33
(Adapted from the HA Annual Reports 1989-2004 and author's calculations)
Chapter 5: Evidence of Earnings Management Practices in the HA
48
When the same auditor provides both audit and non-audit services, auditor independence
is doubtful. According to section 201 and 202 of the Sarbanes Oxley Act 2002, an
auditing firm can only provide non-audit services together with audit services without
obtaining any prior approval, as long as non-audit fees is less than 5% of total fees.
Table 5.3 illustrates that KPMG's income for non-audit services is significantly higher
than 5%. Given that auditors were not rotated from 1990, it is important to check if
there was rotation of audit partners. However, it is hard to determine as the reports were
signed off by the institution rather than the individual partner. These factors, therefore,
suggest that auditor independence is questionable at the HA.
The HA's total assets increased from $85.29 million in 1990 to $158.70 million in 2004,
growing at an annual average rate of 4.96%. Their debt to equity ratios were quite high
in the early 1990s but it worsened in the later part of 1990s when it became negative.
This was due to large accumulated losses included in equity. The conversion of $44
million debts into equity in 2002 returned HA to a net assets position. After reporting
consecutive losses in the 1980s, 1992 was the first year when the HA showed an
operating profit of $43 million. It continued to report marginal profits for few years, but
losses were again reported from 1998. However, the HA reverted to a net profit from
2001. Abnormal items were quite frequently reported in its financial accounts. These
financial data are provided in Appendix 3A.
5.3 Possible Instances of Earnings Management
5.3.1 High Provision for Write Downs of Developed Lots in 1988 to 1991
The inventory of the HA is categorised into developed lots, unsold properties and work-
in-progress. From 1988 to 1991, there were unusual increases in the provision for write-
down of developed lots, as shown in Table 5.4 below.
Analyses of Earnings Management Practices in Fiji's SOEs
49
Table 5.4 Developed Lots Inventory
Year Beginning Balance ($000s)
Provision for write-down ($000s)
Ending Balance ($000s)
Write-down/ developed lots (%)
1985 3726 0 3726 0
1986 2661 0 2661 0
1987 6738 63 6675 0.9
1988 9916 1561 8355 15.7
1989 9727 2063 7664 21.2
1990 13207 3012 10195 22.8
1991 10039 1756 8283 17.5
1992 9155 0 9155 0
1993 11695 0 11695 0
1994 10819 0 10819 0
1995 8170 163 8007 2
1996 7756 506 7250 6.5
1997 9417 362 9055 3.8
(Adapted from the HA Annual Reports 1985 – 1997 and author’s calculations)
The HA values inventory at cost but when the net selling price is lower than the cost, the
cost is written down to its net realisable value. This implies that from 1988 to 1991, the
selling price of developed lots was lower than the cost, hence the large write-downs.
Accordingly, there is a need to ascertain why the selling price was lower than the cost of
developed lots in these periods. However, no information was available in the annual
reports to explain the large write-downs in inventory, apart from the Board Chairman’s
comments. He stated the following about the HA’s operating performance:
…the Housing Authority has progressed quite satisfactorily over the past three
years. From an operational deficit of $1.144 million in 1989 and $2.216 million in
1990, this deficit has been reduced to $479,000 in 1991. However losses due to
abnormal items specifically with relation to on-going re-negotiation of high interest
Fiji National Provident Fund loans and write-offs due to past over expenditure on
land development have put the 1991 loss at $1.898 million compared to $1.141
million for 1990 and $4.088 million for 1989. (The HA Annual Report, 1991:4).
The above statement indicates that the HA had over-spent on land development
activities in the previous years so in 1991, it wrote off these development costs. Table
Chapter 5: Evidence of Earnings Management Practices in the HA
50
5.5 illustrates the items related to developed lots that were reported as abnormal in the
Income Statement.
Table 5.5 Abnormal Items15
Year Item $000s % of Income % of
Inventory
1987 Provision for write-down of development cost and unsold lots
1037 54.44 5.51
1988 Provision for write-down of development cost and unsold lots
2196 115.70 11.05
1989 Provision for write-down of development cost & unsold lots
1641 41.65 8.98
1989 Development cost written off 1303 33.07 7.13
1990 - - - -
1991 Development cost written off 889 21.29 5.94
1991 Provision for unsold land written back (722) (17.29) (4.83)
(Adapted from the HA Annual Reports 1987 – 1991 and author’s calculations)
In 1991, development cost of $0.89 million, representing 21% of income, was written
off. Costs written down and costs written off are two separate items and each has
different accounting treatment. The former is the expected loss in value of inventory
while the latter is the actual loss. When inventory is written down, a provision account is
created and when inventories are written off, there is a direct reduction to the value of
inventory. The corresponding debits in both the cases will still be reflected in the Income
Statement. Furthermore, costs written down and costs written off were only reported as
abnormal items while no such expenses were reported in the normal operating section of
the financial statements. Therefore, the Chairman’s statement only explains the write-
offs of the development cost and not inventory written down.
Based on the reason for the write-offs, it could be argued that from 1987 to 1989 (see
Table 5.5), the HA may have provided for the write down anticipating the loss that
would be incurred due to high expenditure on developed lots. The 1989 and 1991 values
suggest that an additional write-off was needed when these lots were actually sold off.
This means that the provisions made in the earlier years were not sufficient. Under/over
provisioning does not in itself point towards earnings management, but its magnitude
15 Note that the provisions shown in the table includes the provision for developed lots and work-in-progress. This will explain the high write-down in 1987 even though Table 5.4 shows a provision of only $63,000.
Analyses of Earnings Management Practices in Fiji's SOEs
51
may do so. The under-provision was substantial, such that when the additional write offs
of $1.30 million in 1989 and $0.89 million in 1991 were made, they were treated as
abnormal items.
As a result of over-expenditure on developed lots, the HA wrote off these expenses in
1991. In light of this, it is interesting to note that the provision for developed lots was
written back during the same year. Writing back of provisions is not unusual. Prudent
accounting practices should lead to consistent, but not excessive provisioning. Write
backs will arise as uncertainties resolve themselves. Therefore, it is not the write back
but its magnitude that is important. A large write back suggests deliberate over
provisioning and hence an attempt of earnings management. Therefore, it is essential to
consider if the write back related to a specific year or to several years. However, this
was not possible to verify from the available data. Also, there were no further write
backs of provision after 1991.
The property market boomed in the early 1990s so the write back may simply be a
reflection of this optimism. It would certainly explain the zero provisioning in 1992 to
1994 and the higher inventory levels. The property market slumped in 1995 and HA
resumed provisioning at that time. Therefore, the economic factors explain the pattern in
HA's provisioning for developed lots except the high provisioning from 1988 to 1991,
which is hard to justify. As high provisioning reduces earnings, there is a need to
identify if there was any possible reason(s) for the HA to manage earnings downwards
during these periods.
The HA was plagued with financial losses in the 1980s largely due to its subsidised
rental program. Its rental operations imposed significant costs threatening its viability.
There was poor loan management, lending terms did not take into account the cost of
funds and the HA had to meet the debt service costs of the loans that were acquired to
build the rental estates. Immediately after the military coup of 1987, the borrowing rate
in the local market had sky-rocketed. This led the HA to seek cheaper sources of finance
from the ADB and World Bank, who stipulated that loans would only be given if the
Chapter 5: Evidence of Earnings Management Practices in the HA
52
rental element was separated from the HA's operations. Therefore, the deficit laden
rental operations were transferred to the newly created Public Rental Board (PRB) in
May 1989. Negotiations with the ADB were concluded in December and the loan
agreement was signed in early 1990. The agreement with World Bank was delayed due
to issues over the lending terms but was eventually signed in 1991. Separating the rental
operations proved to be advantageous to the HA as it enabled the Authority to meet the
guidelines of the loan requirements. When the PRB was created, the following items
were transferred from the HA:
Assets $11.14 million
Liabilities $19.37 million
Accumulated losses $11.89 million
Revaluation Reserve $ 3.39 million
Table 5.6 shows the performance of the HA in years before and after the transfer of the
aforesaid assets and liabilities to the PRB.
Table 5.6 Financial Indicators of the HA Before and After the Transfer of Rental
Operations (As a percentage of total assets)
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
Total Equity
13.56 9068 1.86 (3.73) 3.04 2.00 0.93 1.91 3.37 3.78
Total Liabilities
86.44 90.32 98.14 103.73 96.96 98.00 99.07 98.09 96.63 96.22
Interest Income
7.94 7.79 7.08 7.00 8.82 8.28 8.54 6.86 6.88 7.09
Total Income
10.52 9.71 9.29 9.55 12.65 11.84 12.40 12.41 11.85 11.78
Interest Expenses
7.67 7.88 7.31 7.65 8.20 8.28 7.58 6.93 6.09 5.80
Total Expenses
12.34 12.27 12.14 12.87 13.94 14.44 12.95 12.37 11.44 11.33
Net Profit (1.75) (2.60) (8.01) (5.52) (4.62) (1.34) (2.19) 0.23 0.40 0.45
(Adapted from the HA Annual Report 1985-1994 and author's calculations)
It is noted that the HA's financial position and performance improved following the
transfer of rental operations. The equity position increased to $3.04 million in 1989
compared to a negative balance of $3.73 million in the previous year. Return on assets
also improved from negative 4.62% in 1989 to 0.23% in 1992.
Analyses of Earnings Management Practices in Fiji's SOEs
53
There were expectations from the government and the offshore lending agencies that
following the transfer, the HA would be able to operate on a profitable basis (Sharma
2000). The Authority was aware that it would have to meet these expectations. One way
to report improved performance in subsequent years is to report major losses in the
current period. It seems that the HA has behaved in this manner. By excessive
provisioning in years prior to and in the year of separation, the HA reported poor
profitability. Given the nature of its rental operations, there is no doubt that the HA was
making losses. Nevertheless, it is argued that by creating unusually large provisions, the
HA exaggerated its unprofitable position. Furthermore, when the assets and liabilities
were transferred to the PRB, $11.89 million of accumulated losses were also transferred.
Although, the HA states that the transfer related to the rental operations only, it may be
possible that this contained losses pertaining to non-rental operations. No matter how
carefully any transfer of business operations is made, it will be arbitrary to some extent.
Therefore, it seems that the HA was able to shed its unprofitable operations to the PRB
and part of its balance of accumulated losses.
There was an additional possible economic incentive to manage earnings downwards in
the early 1990s. The Japanese Grant provided five experts to the HA to strengthen its
institutional capability and train the local staff to undertake the responsibilities. The
grant provided expatriate consulting services in the fields of physical planning, project
management, procurement, corporate planning, accounting and information systems.
The HA was required to have these experts in place before the loans from World Bank
and ADB became effective. The recruitment process began from January 1989 and all
experts were hired by the end of the year. These consulting services were valued at
$183,000 in 1989, $673,000 in 1990 and $332,000 in 1991. The Board Chairman made
the following comments on these services:
Despite the major changes and obvious benefits that have resulted over the 12
months to December 1990 arising out of institutional factors, due to complicated
skills involved and the need for comprehensive training for the authority's
employees, it is desirable, indeed necessary, that in the meantime, the institutional
consultancies continues and support through the funding of these consultancies be
Chapter 5: Evidence of Earnings Management Practices in the HA
54
sustained, to prevent the process from reversing on itself at this delicate and early
stage. (The HA Annual Report, 1990:2).
This shows that HA not only wanted to continue receiving the consultancies but also it’s
funding. High provisioning would assist the Authority in reporting lower profits hence
they could continue to benefit from the assistance. If HA had reported high profits then
perhaps it would still have received the consultancies, but it may have not been in the
form of grant. Rather, the HA would have had to finance them.
To conclude, it was found that unexplained large provisions were created in 1988 - 1991
period. The analysis suggests that this may have been done to exaggerate the poor
performance of the HA so that in the subsequent periods the financial performance
reported are relatively better. This is argued on the basis that there were general
expectations from key stakeholders that the HA would reflect improved performance
following the transfer of rental operations. Hence, it seems that the HA managed
earnings downwards during these periods. Furthermore, the need to continue receiving
funds for consultancies may have provided an additional incentive to practice downward
earnings management.
5.3.2 Capitalisation of Administration Expenses
The HA reports work-in-progress inventory at the lower of cost or net realisable value.
The cost of work-in-progress includes capitalised indirect administrative costs, which
are incurred as a result of land development and house construction, and development
interest costs of the funds used to finance the development up to the time of completion.
Capitalising interest expense is not an uncommon practice as FAS 23 Capitalisation of
Borrowing Costs allows an entity to do so. The concern here is of capitalising
administrative expenses. According to FAS 11 Accounting for Construction Contracts16,
16 Reference is made to FAS 11 because apart from the lending activities, the HA is also involved in construction activities.
Analyses of Earnings Management Practices in Fiji's SOEs
55
costs such as general administration and selling costs, finance costs, research and
development costs and depreciation of idle plant and equipment that is not used on a
particular contract are examples of costs that relate to the general contract activity but
cannot be related to specific contracts. Paragraph 19 of the standard further states that
these costs are usually excluded from accumulated contract costs because they do not
relate to reaching the present stage of completion of a specific contract. However, in
some circumstances, general administrative expenses, development costs and finance
costs are specifically attributable to a particular contract and are sometimes included as
part of accumulated contract costs.
Capitalising costs rather than expensing postpones the recognition of expense. As a
result, current period’s expense will be reduced and earnings will rise. Since the
accounting regulation permits the capitalisation of administrative expenses in certain
circumstances, if an entity is capitalising this expenditure, it is possible that they are
attempting to boost current period earnings. Consequently, it is possible that the HA
interpreted this paragraph of FAS 11 to suit its own convenience rather than to generally
reflect economic reality!
Capitalising administrative expenses to work-in-progress has two effects on profits.
First, when the costs are capitalised, the current period earnings will increase. Secondly,
these capitalised costs will be expensed when the work-in-progress inventory is
completed and sold, upon which, it will be expensed through cost of sales. Thus, in the
subsequent periods, cost of sales will increase leading to a reduction in earnings. In
order to determine when these costs would be expensed, the following table calculates
inventory days, using all three categories of inventories- land, houses and work-in-
progress
Chapter 5: Evidence of Earnings Management Practices in the HA
56
Table 5.7 Inventory Days
1989 1990 1991 1992 1993 1994
Sale of land & houses 3276 3428 4203 6673 13390 21014
Cost of sales of land & houses
2978 2776 3503 5264 10775 17835
Inventory 14004 12679 12266 20558 23914 22899
Inventory days 1716 1667 1278 1425 810 469
(Adapted from the HA Annual Reports 1989–1994 and author’s calculation)
The calculation yields significantly large number of inventory days. For instance, in
1989, it took HA almost five years to sell inventory. This seems unrealistic as when an
entity has such high inventory days it will not continue with annual production activities.
The HA, however, had continued with its land development and construction activities.
Using inventory days to ascertain the timing of expense recognition for the capitalised
costs would therefore be impractical. Thus, the narrative information provided for land
development and construction activities during the years was used in this analysis. As
there is no consistency in the type and extent of information provided in this form, the
study had to rely on whatever information is made available. Based on the narrative
information, it was found that it took HA an average of two years to complete individual
projects and sell them off. Table 5.8 depicts the profits that were reported after
capitalisation and the profits that would have been reported if the HA had not capitalised
these expenses.
Analyses of Earnings Management Practices in Fiji's SOEs
57
Table 5.8 Effects of Capitalising Administrative Costs on Profits
(Adapted from the HA Annual Reports 1989–1994 and author’s calculations)
Table 5.8 firstly shows the actual profit as reported in the financial statements that is
achieved after capitalising administrative expenses. It then reverses the effect of
capitalisation to calculate the profits without capitalized costs. To do that, capitalised
costs are subtracted from profits since capitalisation has increased current period
earnings. Once the work-in-progress inventory is completed and sold, capitalised costs
would be expensed through cost of sales. So the table adds back cost of sales to profit.
As it takes two years time to expense the deferred costs, the table shows that the amount
capitalised in 1989 is expensed in 1991. Capitalised costs in other years are also
expensed in similar time interval.
The amount of administration expenses capitalised was reported in the financial
statements only until 199417. However the accounting policy note continued to state that
work-in-progress included capitalised administrative expenses. This implies that HA had
continued to capitalise administrative expenditure even though the amount capitalised
was not disclosed from 1995 onwards. Consequently, the table could not be extended
beyond 1994. The effect of capitalisation on profits was more observable in periods
17 Capitalised administrative expenses were not explicitly reported prior to 1989 as the format in the presentation of information had changed after the transfer of rental operations.
Profits with capitalisation
Reversal effects - Profits without capitalization
Year Net Profit $000s
NP/Equity (%)
Capitalised administrative
costs ($000s)
Subtract capitalised costs from profits $000s
Add back cost of sales to profits $000s
NP/Equity (%)
1989 (4088) (151.86) 86 (4088) - 86 = (4174)
(4174) (155.05)
1990 (1141) (67.04) 76 (1141) - 76 =
(1217) (1217) (71.50)
1991 (1898) (236.07) 415 (1898) - 415 = (2313)
(2313) + 86 = (2227)
(276.99)
1992 248 12.09 576 248 - 576 =
(328) (328) + 76 = (252)
(12.28)
1993 518 12.02 575 518 - 575 =
(57) (57) + 415 = 358
8.30
1994 687 11.81 676 687 - 676 =
11 11 + 576 =
587 10.09
Chapter 5: Evidence of Earnings Management Practices in the HA
58
1992 to 1994. Capitalisation improved the profitability position of the HA from a
negative return on equity of 12.28% to positive 12.09% in 1992. Further, in 1993 and
1994, it increased the returns from 8.30% to 12.02% and 10.09% to 11.81%,
respectively. Given that capitalisation resulted in increased profits, the following section
examines if there was any possible incentive(s) for reporting higher earnings during
these years.
The Rabuka government, which came into power in 1992, placed a lot of emphasis on
public sector reforms as a means of bringing efficiency in the operations of SOEs. In
line with the government's reform agenda, the HA's focus changed as it prepared for
privatisation. The Authority began to pursue commercial activities with a view of
making profit. Taking into account the changes that took place in the government, it
could be argued that the newly elected government wanted to reflect its performance
through the performance of the HA, which is its public sector housing institution.
Furthermore, there were major changes in the HA's Board membership in 1993 as four
out of the six board members were replaced during the year. Accordingly the Board,
with a majority of new members, may have also wanted to reflect their performance to
the government and other stakeholders. Thus, the incentive to report high profits in 1993
may have also originated from the HA’s Board.
Additionally, the HA introduced performance based contracts at the senior executive
level in June 1993. Management was exposed to a contractual system where appraisal
techniques and indicators were used to facilitate this exercise. All staff in executive
positions was put on three-year contract and the renewal of their contracts was based on
the achievement of the set targets and objectives for each position. The annual bonus of
the senior executives was also based on the achievement of those targets. The annual
report indicates that
the movement of salaries for these positions therefore will be relative to the
productivity of the incumbents and in turn will be reflective of the overall
financial performance of the Authority. (1993:10).
Analyses of Earnings Management Practices in Fiji's SOEs
59
Performance indicators, comprising of both financial and non-financial, were used as a
measure of the achievement of the targets. They enable managers to evaluate in
numerical terms the extent to which their actions have led to improvement in services
offered. The Chief Executive and the General Manager establishes the targets for the
respective departments while the Board Chairman and the Chief Executive sets for the
Chief Executive. Each division has its own targets to achieve, against which the
managers were evaluated. For instance, the Collection Department had a target
collection policy and the Sales Department performance indicators encompassed
cumulative actual and cumulative target sales of land and houses. Similarly, the
performance of the Lending Department was judged through the volume of loans
approved and the average loan processing time relative to the predetermined monthly
targets.
The annual year-end bonus for managers, which was a component of their salary, was
awarded if the set targets were met. Furthermore, renewal of their contract was subject
to the successful achievement of the targets over the current contract period. When the
targets are not achieved, a review is carried out to determine the underlying reasons.
After a thorough investigation, the targets are reviewed or reduced so that they are
attainable. KPMG also checks the targets from time to time to establish its feasibility.
From the time contractual system was introduced, the employment contracts were taken
seriously by the management. Sharma (2000) reports that one of the managers left the
institution in the year in which the contract was actually introduced as he was unhappy
with the new system. It seems that performance based contracts provided incentives to
manage earnings upwards in order to qualify for annual bonuses and continuity of
employment. However, it is difficult to clearly establish the link between the targets and
performance given that in some instances where the targets were not achieved, they were
reduced.
To sum up, it was found that capitalisation resulted in higher profits in the years 1992 to
1994. In fact 1992 was the first year in which HA reported a profit after years of losses.
The analyses suggest that the possible motive to report high earnings may have come
Chapter 5: Evidence of Earnings Management Practices in the HA
60
from three sources. First, there was a newly elected government in 1992 that was in
favor of reforming the public enterprises. Performance of public enterprises is often seen
as a reflection of the performance of government. Second, the majority of the Board
members were replaced in 1993. They would also want to reflect their good
performance. Finally, performance based contracts were introduced in 1993. This system
provides an incentive to the management to report high earnings in order to qualify for
the annual bonuses and to retain their employment. Capitalising administrative costs is
allowed by the accounting regulation so the practice by the HA is not questionable.
Nonetheless, capitalisation significantly improved the HA’s financial performance in
early 1990s and there were several incentives for the HA to report higher earnings. Thus,
it seems that the deferral of administration expenses was used for earnings management
purposes.
5.3.3 Deferred Interest Expense
In the 1990s, the HA negotiated with the Fiji National Provident Fund (FNPF), their
major financier, to reduce the high lending rates. The following contingent liability was
reported in the Annual Report:
Early in 1990 the Authority negotiated with FNPF to reduce high interest rates on
loans taken out during the years 1983 to 1986. The Authority was advised by FNPF
that it had agreed to the reductions and to make repayments based on the reduced
interest rates, which occurred from December 1990. The Authority's solicitors were
in the final stages of amending the loan covenants when advice was received from
FNPF in March 1991 that their Board had not formally approved the renegotiated
interest rates. The Authority had previously taken up interest at the negotiated rates.
If the renegotiated rates are not approved by FNPF, additional interest of $677,000
will be payable each complete year. In respect to 1990 $188,000 will be payable.
(1990:15).
Analyses of Earnings Management Practices in Fiji's SOEs
61
As no formal approval was received for the negotiated rates by 1991, there was an
additional interest payable as at December 31 1991 amounting to $1.18 million to the
FNPF. This was reported as an abnormal item in the Income Statement. This implies that
it was only in 1990 and 1991 that HA recorded interest payable to FNPF at the reduced
rate. The matter was resolved in 1991 by expensing the additional interest payable.
In 1992, the HA recognised Deferred Interest Expense of $7.56 million as an asset. The
same amount was also found under its Loans Liability in 1992 and it was stated that
$7.56 million is the unexpired portion of differential interest on FNPF loan that would
be converted into 10-year bonds in 1993. From 1995, the annual reports began to state
the following:
The deferred interest expense relates to differential interest on rescheduled FNFP
loans. The deferred interest is amortised over a period of 10 years.
During 1992, the FNPF decided to reschedule the loans advanced to the HA. If the
FNPF reduced the interest rate, it would lose the entire amount of $7.56 million.
Accordingly, the FNPF decided that the HA will continue to pay interest at the lower
negotiated rate and the difference in interest payable between the normal and the
reduced rate will be converted into 10-year bonds in 1993. The following disclosures
were found in the FNPF's annual reports.
…Also excluded from the investments are the unpaid portion of nine HA loans
totaling $7,564,354 which were converted to 10-year zero rated bonds during the
1992/199318 financial year (1993:27).
…Also excluded from the investments is the differential interest totaling $7,564,354
for the unexpired portion of nine rescheduled HA loans, converted to 10-year zero
rated bonds during the 1992/1993 financial year (1995:21).
This provided further information that the differential interest was converted into zero-
rated bonds. Thus, the HA will benefit as they will continue to pay interest at the lower
rate and the differential interest will be paid collectively in the tenth year without further
18 The financial year-end for the HA is 31 December while for the FNPF it is 30 June.
Chapter 5: Evidence of Earnings Management Practices in the HA
62
interest accruing over the ten year period. As a result, the HA was able to conserve its
cash flows. Although, the FNPF will still receive its income at the end of the ten-year
period, it will lose as the delayed cash flows would impact, albeit marginally, on its
reinvestment capacities. The differential interest was finally paid in 2002.
Deferred interest expense is an example of a deferred debit, which is reported as asset in
the balance sheet. Henderson and Peirson (2002) and Goodwin and Howieson (1999)
argue that accountants are so anxious to correctly match revenues with expenses that
they end up creating deferred debits and credits. They state that these accounts fail to
meet the definitions of assets and liabilities. However, Sprouse (1966) argues that some
deferred debits and credits are assets and liabilities. Therefore, it is important to find out
if the deferred debit reported by the HA is actually an asset. The Framework for the
Preparation and Presentation of Financial Statements (FIA, 2001) defines assets as
resources controlled by the enterprise as a result of past events and from which future
economic benefits are expected to flow to the enterprise. The future economic benefit
from the deferred interest expense relates to the difference in the interest expense that
the HA will save each year for the next ten years. The deferred debit arose from the past
event when the FNPF actually decided to reschedule the loans. Through the legal
contract of converting the differential interest into bonds, the HA also has control over
these benefits. Consequently, the deferred interest expense reported by the HA does
meet the definition of an asset. After discussing the nature of deferred interest expense,
the following section focuses on its subsequent amortisation.
Table 5.9 shows how the HA expensed $7.56 million over the 10 years. The table shows
that the pattern in which the HA has expensed the deferred debit does not constitute any
systematic method. Further, in 1993 there was an anomaly in the amount amortised as
the amortisation rate, being 31%, was the largest over the entire ten year period. The
amount amortised of $2.38 million was even greater than the abnormal item of $1.18
million reported in 1991 Annual Report, which related to two years of additional interest
expense payable.
Analyses of Earnings Management Practices in Fiji's SOEs
63
Table 5.9 Amortisation of Deferred Interest Expense
Year Deferred Interest
Expense ($000s)
Amortisation Expense ($000s)
Amortisation Expense/ Deferred Interest Expense (%)19
Return on Assets (%)20
1992 7564 0.23
1993 5188 2376 31.41 0.40
1994 4832 356 4.71 0.45
1995 4504 328 4.34 (1.43)
1996 4175 329 4.34 (3.37)
1997 3870 305 4.03 0.07
1998 3239 631 8.34 0.10
1999 2394 845 11.17 (2.15)
2000 1549 845 11.17 (0.27)
2001 704 845 11.17 0.31
2002 0 704 9.31 0.31
(Adapted from the HA Annual Reports 1992 - 2002 and author's calculations)
Over the amortisation period of 1993 to 2002, no separate disclosure was made for the
amortisation expense either in the Income Statements or in the disclosure notes for
expenses. From 1994 to 2002, it was not possible to identify where the amount is
aggregated as it was not large enough to be separately identifiable. It may be possible to
do this for 1993 given the magnitude of the amortisation. In 1993, the following
categories of expenses were shown in the Income Statement and their corresponding
values.
Table 5.10 Components of Total Expenses in 1993
Item $000s
Interest on Loans 7794
Operating Cost 5584
Provisions 965
Bad and Doubtful Debts 300
(Adapted from the HA Annual Report 1993)
Table 5.10 is used to explain the costs with which the amortisation expense is
aggregated. The expenses categorised as Provisions and Bad and Doubtful Debts are
19 This ratio is calculated as the amortisation expense for each year divided by the initial amount recognised as deferred interest expense of $7.56 million. Thus, the ratio gives the amortisation rate. 20 Return on equity could not be used because for several years the HA was faced with negative equity balance, hence the ratio would not give a meaningful explanation.
Chapter 5: Evidence of Earnings Management Practices in the HA
64
eliminated as the amount for these categories of expenses, $965,000 and $300,000,
respectively, is lower than the year's amortisation expense of $2.37 million. The next
category to consider is Operating Costs. All items in this category cost less than $2.37
million except for wages and salaries of $3.46 million. Therefore, this category is also
eliminated. The analysis now suggests that the amortisation expense is almost certainly
included with Interest on Loans. This implies that the amount of interest expense
reported in the Income Statement included the amortisation of deferred interest expense.
Table 5.11 illustrates the interest expense with and without amortised costs.
Table 5.11 Interest Expense
Year Interest Expense
Including Amortisation Interest Expense
Excluding Amortisation
($000s) % of Borrowings ($000s) % of Borrowings
1990 7377 9.47 7377 9.47
1991 7603 9.77 7603 9.77
1992 7941 8.21 7941 8.21
1993 8442 7.37 8442 - 2376 = 6066 5.30
1994 9792 7.03 9792 - 356 = 9436 6.77
1995 11476 7.43 11476 - 328 = 11148 7.22
1996 12547 7.92 12547 - 329 = 12218 7.71
1997 12358 7.95 12358 - 305 = 12053 7.75
1998 12546 8.11 12546 - 631 = 11915 7.70
1999 10336 6.94 10336 - 845 = 9491 6.38
2000 11585 7.70 11585 - 845 = 10740 7.14
2001 10079 6.74 10079 - 845= 9234 6.17
2002 7827 6.09 7827 - 704 = 7123 5.54
2003 7118 5.49 7118 5.49
(Adapted from the HA Annual Reports 1989 - 2003 and author's calculations)
The above table first shows the interest expense on long-term loans as reported in the
financial statements. It then subtracts the amortisation expenses from the interest
expenses to calculate interest expenses excluding the amortised charges. This represents
the interest payable at the negotiated rates. An examination of the interest payable at the
negotiated rates suggests that the interest expense as a percentage of total borrowings is
gradually declining, except for 1993. The interest expense over total borrowings in 1993
was relatively lower at 5.30%, which seems not to be in line with the general trend. It is,
therefore, argued that given the interest payable excluding the amortisation was low in
Analyses of Earnings Management Practices in Fiji's SOEs
65
1993, the HA amortised a larger amount of the deferred interest expense in that year to
smooth its total interest expense payable and consequently its profits.
As it was found that interest expense excluding amortisation was relatively low in 1993,
an attempt was made to identify its possible reason. The following table presents the
Authority’s loan liabilities from 1992 to 1994.
Table 5.12 Loan Liabilities in 1992, 1993 and 1994 ($000s)
1992 1993 1994
Opening balance 75348 91820 109614
New loans 12138 21034 28457
10 year zero rated bond 7564
Principal repaid (3230) (3240) (2701)
Ending balance 91820 109614 135370
(Adapted from the HA Annual Reports 1992 - 1994)
Table 5.12 shows that the outstanding loan balance had increased over the three years.
Accordingly, the interest expense on these loans would also increase. The table shows
that in each year, new loans were undertaken and repayments were made for existing
loans. New loans raised were relatively higher than repayments, implying that interest
expense in 1993 and 1994 should be increasing. To ascertain the interest expense on the
new loans undertaken in each of the three years, the following calculations are made.
Table 5.13 Interest Expense on New Loans Raised During 1992, 1993 and 1994
1992 1993 1994
Loan $000
Interest rate %
Interest expense $000
Loan $000
Interest rate %
Interest expense $000
Loan $000
Interest rate %
Interest expense $000
1437 8.5 122.15 3034 8.5 257.89 6957 8.50 591.35
3580 8.02 287.12 4000 7.875 315.00 10000 7.50 750.00
1500 8.075 121.13 6000 7.45 447.00 7000 7.60 532.00
320 8.12 25.98 1800 7.35 132.30 4500 8.00 360.00
500 8.36 41.80 4000 7.82 312.80
2100 7.8 163.80 2200 8.375 184.25
2701 9.00 243.09
$12138 $1005.06 $21034 $1649.24 $28457 $2233.35
(Adapted from the HA Annual Reports 1992 - 1994 and author's calculations)
Chapter 5: Evidence of Earnings Management Practices in the HA
66
While drawing up Table 5.13, it was assumed that all new loans were undertaken from
the beginning of the year, so the interest expense represents interest for 12 months. For
1992, interest rate on the amount borrowed from the FNPF of $2.70 million was not
provided in the annual report. Consequently, it was assumed that the interest rate was
9% as the annual report had stated that the interest rate on all loans during the year
varied from 6.5% to 9.0%.
Based on the calculations above, it was expected that interest expense (excluding
amortisation) would have increased by approximately $1.65 million and $2.23 million in
1993 and 1994, respectively. Referring back to Table 5.11, it was seen that interest
expense (excluding amortisation) did increase in 1994 from $6.07 million to $9.44
million. However, this does not hold for 1993. Although, Table 5.13 shows that interest
should have increased over the year in 1993, Table 5.11 shows that interest as reported
in the Income Statement has actually reduced. One could argue that perhaps this was the
outcome of the rescheduling of the FNPF's loans as the FNPF had agreed to reduce the
rate. If this was the case then 1994's interest expense should have also been reduced.
But, as it was found, the 1994's amount in Table 5.11 and 5.13 reconciles. Therefore, no
reasonable explanation could be obtained for the low interest expense reported
(excluding amortisation) in Table 5.11.
In conclusion, it is argued that the actual amortisations do not conform to any
identifiable pattern. The amortisation of the deferred expense seems to suggest that it
was used as a smoothing device, particularly in 1993. Interest expense prior to the
inclusion of the amortised charge was relatively low, hence amortising a significant
amount brought the interest expense in line with the general trend. In the earlier section
on capitalisation of administration expenses, it was argued that the HA practiced upward
earnings management in years 1992-1994 while this section implies that a substantial
amortisation made in 1993 reduced earnings. Both the sections discuss items that are
highly dependent on management's discretion as managers can certainly predetermine
the likely effects of amortising a certain amount of deferred assets and/or capitalising a
certain portion of administrative expenses. The year 1993 was an opportune time for the
Analyses of Earnings Management Practices in Fiji's SOEs
67
HA to amortise a significantly large amount as even after amortising deferred debit of
$2.38 million, it was still able to provide 0.40% return on assets compared to 0.23% in
the previous year.
5.3.4 Changes in Accounting Policies for Provision for Doubtful Debts
Over the review period, there were two occasions when the HA revised its accounting
policies pertaining to the provision for doubtful debts. The first revision was made in
1995 and the accounting policy note on doubtful debts was provided as follows:
During the year, the Authority adopted a policy of creating specific provisions for
doubtful debts on non-performing debtors accounts. In addition a general provision
of 1% of the total debtors portfolio not covered by the specific provision is brought
to account. In the past, provision for doubtful debts was calculated on an assessed
mathematical formula. The level of general provision is determined having regard
to economic conditions and general risk factors. The financial effect of this change
in the method of calculating provision for doubtful debts is a charge of $2,702,000
to the profit and loss account. (The HA Annual Report, 1995:25).
Prior to 1995, the HA provided for doubtful debts by using an assessed mathematical
formula21. Following 1995, it began to create specific and general provisions. With the
change in accounting policy, there was a significant increase in the provision for
doubtful debts. This increase was reported as an abnormal item in the Income Statement,
explaining that it was an impact arising from a change in accounting policy.
Consequently, the HA incurred a net loss of $2.39 million in 1995 compared to an
operating profit of $0.69 million in the previous year. On this performance, the Board
Chairman made the following comment:
The Housing Authority has recorded a net loss of $2.393 million for the year ending
31 December 1995. A loss was inevitable after the adoption by the Housing
21 Details of the formula are not provided in the annual reports but may be obtained from the HA.
Chapter 5: Evidence of Earnings Management Practices in the HA
68
Authority Board of new accounting policies in the preparation of the 1995
Accounts. The changes in policy, in particular for the Provisioning for Doubtful
Debts, adopt practices that are acceptable and in line with industry standards. (The
HA Annual Report, 1995: 8).
The HA Board justified the change on the grounds that the new practice was acceptable
and in line with industry standards. Under Section 14(3) of the Banking Act 1995, as part
of the minimum rules for the conduct of banking business by licensed financial
institutions, the RBF provided guidelines for classifying loans and creating provisions
for impaired assets in 199622. The guideline states the following regarding provisions:
(Licensed Financial) Institutions are required to maintain a prudent level of general
provisions against losses not as yet identified on the good part of the portfolio as
well as specific provisions against reasonably anticipated losses on poor quality
assets and all specifically identified losses on impaired assets. Any understatement
of the level of provisions required may result in an institution's capital and profits
being overstated which could lead to a lack of certainty of the institution's on-going
solvency. (RBF Policy Statement No. 3, 1996:7).
The change in the HA's accounting policy for the provision of doubtful debts was in line
with the accepted practice of the licensed financial institutions as these institutions
created general and specific provisions. Thus, the HA's decision to change the
accounting policy is not questionable as it shows that they are moving in line to the
general industry practice. However, the concern here is the timing of the change. Why
did HA decide to adopt an accounting practice similar to the licensed financial
institutions in 1995? It surely cannot be because of the minimum guidelines of the
Banking Act as this came into effect from the following year. Additionally, the guideline
does not apply to the HA since it is not a licensed financial institution so it is not even
obliged to follow these guidelines. It is a non-bank financial institution, regulated by its
individual statute and not by the RBF.
22 Licensed financial institutions include commercial banks, credit institutions and insurance companies.
Analyses of Earnings Management Practices in Fiji's SOEs
69
Interestingly, the policy change was first advised by their external auditors in 1991 but
its discussion and review had been deferred till 1995. The Chairman admitted the
following in the Annual Report:
Deferring the discussion and review of the Housing Authority's accounting
practices since 1991 means that the picture presented in our Annual Accounts for
the past few years have been misleading as a measurement of our performance
(1995:8).
It is not understood why the HA deferred the revision of its accounting policies when it
knew that the existing policy was giving a misleading picture. One should bear in mind
that a change in accounting policy does not automatically improve the economic reality
of the report. Indeed the motive for the change made in 1995 is not undertaken to better
reflect economic reality, but to bring the Authority into line with the rest of the industry.
The following section considers whether there was any possible motive for the change in
policy.
With the existing government’s public sector reform policy, the HA's operations since
1992 has been production driven and it was mainly satisfying the needs of the middle-
income customers. A large number of families in the low-income group were regularly
dispossessed of their houses through mortgagee sales. These families not only lost their
personal savings but also their retirement funds with the FNPF. In light of this, the
Board directed that there be a change in focus from the beginning of 1995 as it felt that
unless this change was made, evictions would continue. The HA commenced the year by
focusing on both the middle and low-income earners. The Authority suspended its
design and build scheme during the year in order to re-align their focus to their target
market as the houses constructed under this scheme proved to be unaffordable to its low-
income clients.
However, the change resulted in conflict within the HA so the Government sanctioned a
Special Inquiry into the affairs of the HA. The committee of the Special Inquiry was
headed by the former secretary to the Public Service Commission, Winston Thompson.
Chapter 5: Evidence of Earnings Management Practices in the HA
70
The recommendations (of the Thompson Report) were aimed at bringing about
stability in management of HA, reducing the costs of its products and making
housing more affordable to low income earners, improving the delivery of services
to the clients and bringing about general efficiency in the execution of its functions
(The Fiji Times, October 1996:4 ).
Cabinet accepted the findings and recommendations of the Thompson Report. Its
acceptance reconfirmed government's commitment to the HA realising that its original
key objective is to serve the middle and low income earners.
If the HA had not changed their accounting policy, they would have reported a profit of
$0.31 million compared to a loss of $2.39 million. A profit would have not looked good
for the HA when so many of its customers were facing hardship by losing their houses
and savings that were used for financing. Accordingly, the net loss lent legitimacy to the
HA's operations. The government's acceptance of the committee of Special Inquiry’s
findings suggests that Government realised the problem and resolved that the HA should
not depart from the social objectives of providing affordable housing to the low and
middle-income earners. Therefore, this license from the government might have
provided further incentive to the HA to change its accounting policy and to report a net
loss in 1995.
Furthermore, in 1996 KPMG were commissioned by the Ministry of Housing and the
Ministry of Finance to undertake a special audit. Engaging KPMG to undertake a special
audit is odd in itself as KPMG was already providing a significant amount of non-audit
services together with their audit work. This special audit would have been more
appropriately and independently undertaken by another firm, although it does not seem a
matter, which the HA has ‘managed’. The purpose of the audit was to ascertain the HA's
true financial status, including the quality of its assets. On this special audit the Board
Chairman made the following comment:
As a result (of the special audit), the Authority introduced a more realistic and
prudent accounting policy. A consequential increased allocation for provisions
Analyses of Earnings Management Practices in Fiji's SOEs
71
gives a much more reliable representation of the Authority's financial position. (The
HA Annual Report, 1996:4).
According to the HA, the provisions were included in the accounts in line with a more
prudent accounting approach, which compares loan balances against current market
value of assets. The Board argues that this approach ensures that the balance sheet
reflects the state of the HA's finances more accurately. The special audit resulted in an
increase in the provision for doubtful debts, repairs and maintenance and stock write-
down. The following table shows the trend in these expenses over the years.
Table 5.14 Discretionary Expenditure
Year Provision for Doubtful
Debts Provision for Repairs and Maintenance
Stock Write down
Net profit
$000s % of
Receivables $000s
% of Stock & PPE23
$000s % of Stock
$000s % of Assets
1992 395 0.55 132 0.49 0 0.00 248 0.23
1993 300 0.34 34 0.11 0 0.00 518 0.40
1994 963 0.97 57 0.18 0 0.00 687 0.45
1995 3522 2.84 0 0.00 163 0.75 (2393) (1.43)
1996 3302 2.49 1017 3.83 977 5.26 (5695) (3.37)
1997 1542 1.18 455 1.76 170 0.95 115 0.07
1998 1421 1.00 135 0.77 130 1.34 160 0.10
1999 5836 4.09 0 0.00 777 12.90 (3400) (2.15)
2000 1993 1.41 120 0.91 0 0.00 (432) (0.27)
2001 783 0.54 2245 17.50 355 9.49 518 0.31
2002 870 0.54 150 1.16 287 7.56 538 0.31
2003 968 0.64 0 0.00 41 0.59 771 0.43
2004 1365 0.96 224 1.05 250 2.18 2067 1.30
(Adapted from the HA Annual Report 1992-2004 and author's calculations)
In December 1996, all new members were appointed to the HA Board, including the
Chairman. At the senior management level, the general manager positions were reduced
from five to only three. The Board Chairman commented as follows, regarding the new
Board’s role:
23 Repairs and maintenance are usually carried out for property, plant and equipment. However, the HA paid for these expenses upon receiving several complaints from the customers about the defective houses.
Chapter 5: Evidence of Earnings Management Practices in the HA
72
1996 at the Housing Authority can best be described as a year of resolving
problems, introducing change and laying the foundation for future reorganisation.
The new board appointed in December 1996, is now firmly embarked on a new
course which will see the Authority emerge as a viable, efficient and innovative
commercial enterprise. It will be profit driven, dedicated to supplying affordable
and acceptable housing and enabling more and more people to own homes. (The
HA Annual Report, 1996:4).
As the HA was declared a CSA from 1996, the new Board was expected to lead the HA
towards a profit-driven institution. The following table compares the Authority’s
profitability position in 1995 and 1996.
Table 5.15 Profits in 1995 and 1996 (as a percentage of assets)
1995 1996
Total Revenue 19.23 17.23
Total Expenses 19.04 17.86
Profit before abnormal items 0.18 (0.63)
Abnormal items (1.61) (2.74)
Profit after abnormal items (1.43) (3.37)
(Adapted from the HA Annual Reports 1995 and 1996)
Table 5.15 shows that in 1995 the HA’s operating profit before abnormal items was
0.18% of total assets while in 1996 it was negative 0.63%. As is evident from the above
table, the operating loss in 1996 was due to reduced revenue. Given the Board and the
senior management’s profit driven objectives, an operating loss of 0.63% would reflect a
poor performance of the new team, especially when a profit was reported by its
predecessor. Hence, the effects of the special audit in 1996 proved to be an advantage to
the Board and management as the operating loss before abnormal items have been
overshadowed by the losses incurred through abnormal items.
There was yet another change in the accounting policy for provision for doubtful debts
in 1999. This time the change was in the definition of non-performing loans. When this
policy was first changed in 1996, a non performing loan was classified as one on which
Analyses of Earnings Management Practices in Fiji's SOEs
73
interest was in arrears for six months or more. In 1999, this was reduced to three months.
Thus, all loans where the interest was in arrears for three months or more was
considered to be a non-performing loan.
An interesting point was discovered when reviewing the explanation for the change in
this accounting policy. The annual report stated that the first change in the accounting
policy for the provisioning of doubtful debts was made in 1996. This implies that in
1995, the HA created specific and general provisions and then in 1996, it classified a
non-performing loan as one in which interest was in arrears from six months or more.
Although, the accounting policy note on provision for doubtful debts in 1995 stated that
HA created specific provisions for non-performing loans, it did not provide any
information in its disclosure notes on the definition of non-performing loans.
This change increased the non-performing loans portfolio by approximately $24.40
million. Once a loan is deemed to be non-performing, interest revenue will cease to be
recognised on them. Thus, the Authority lost about $2.60 million of interest revenue.
Given that non-performing loans have increased, so will its related provision for
doubtful debts. The Chairman attributed the year's net loss to the change in accounting
policy as reflected in the following statement:
For the year the Authority has recorded a loss of $3.4 million as against
$159,899 profit recorded in 1998. This loss was inevitable because of the
introduction of new accounting policies relating to Provisioning for Doubtful
debts and Stock Write-downs. (The HA Annual Report, 1999:3).
Similar to the justification given in 1995, the HA again argued that the revision in 1999
conformed to commercial and international practices. The Board Chairman's comment
on the issue of change in accounting policy was as follows:
…the Authority proceeded to again redefine its accounting policy relating to the
booking of Provision for Doubtful Debts. These policy changes were not only
prudent but also necessary as it conformed to commercial and international
accounting practices. (The HA Annual Report, 1999:3).
Chapter 5: Evidence of Earnings Management Practices in the HA
74
FAS 30 Additional Disclosure by Financial Institutions was first issued and was
effective for accounting periods beginning on or after 1 January 1999. According to
paragraph 55 of the standard,
Loans or similar facilities which have not been operated within key terms for at
least 90 days and for which the recovery of all amounts of principal and interest
is not regarded as probable are classified as non-accrual loans.
…90-day threshold period of non-compliance with key terms is based on industry
practice.
The HA’s decision to change the definition of non-performing loans was in line with the
requirements of FAS 30. The standard does state the 90-day threshold period for
classifying loans as non-accrual (non-performing). Hence, the change made by the HA is
not questionable. However, one cannot overlook the possibility that the HA could have
exaggerated the effects of the policy change. The subsequent discussions will illustrate
that the events that took place in years surrounding the year of change could possibly
have provided the incentive to the HA to reduce earnings.
Excessive provisioning often leads to write-back in subsequent years. Upon reviewing
the financial accounts, no such write-backs were found. Alternatively, there would be
reduced provisioning in the future periods. Table 5.14 shows the provision for doubtful
debts over the years as reported in the Income Statement. Given that from 1999, loans
were classified as non-performing three months sooner, it is expected that the doubtful
debts should be more than what was provided for prior to the change. The values in the
table show otherwise. The provisions after the change is relatively lower than the
previous years’ provisions. One could argue that the HA may have taken steps to reduce
the non-performing loans and hence its provisions. However, there is no evidence of
such measures being implemented by the HA.
Further support of the argument that the HA may have exaggerated the effects of the
change comes from the stock write-downs. The Chairman stated that the operating loss
was inevitable due to changes in accounting policies for both, provision for doubtful
Analyses of Earnings Management Practices in Fiji's SOEs
75
debts and stock-write downs. Although the change for provision for doubtful debts was
explained, no explanation was given for the change in stock write-downs anywhere in
the report. Yet, a perusal of the expenses reflects a major movement in the account
balance from the previous year as shown in Table 5.14. The table shows that apart from
major increase in the provision for doubtful debts in 1999, stock write-downs
represented 13% of stock, being the largest over the review period. In fact, when the
accounting policies were changed in 1996 and 1999, stock write-downs had increased.
Thus, the trend in the provision for doubtful debts and the stock write-down in 1999
supports the argument that the HA may have increased the provisions.
A possible incentive to manage earnings downward could be the need to convert the $44
million ADB and World Bank debts into equity. From as early as 1999, the HA had
commenced discussions and made submissions seeking government approval for the
conversion. A comprehensive review was conducted in 2001 to assess the financial
performance of the Authority and to determine the factors that would contribute to its
continued financial viability. One of the findings of the review is given below.
The review indicated that unless Housing Authority's debt to Government in respect
of ADB/World Bank loans acquired for the low income housing projects 1990-1994
were converted to equity, (the) Housing Authority would remain unviable. The
Board made representation to the Government in this regard. (The HA Annual
Report, 2000:8).
The review clearly pointed out that for the Authority to continue as a viable institution,
these loans had to be converted into equity. The change in accounting policy in 1999 had
worsened the profitability of the HA and consequently deteriorated the equity balance.
The following table shows the profits and equity position of the HA over the years.
Chapter 5: Evidence of Earnings Management Practices in the HA
76
Table 5.16 Profitability and Equity Position of the HA
Year Net Profits Accumulated Profits Total Equity
$000s % of Assets $000s % of Assets $000s % of Assets
1990 (1141) (1.34) (12444) (14.59) 1702 2.00
1991 (1898) (2.19) (14342) (16.57) 804 0.93
1992 248 0.23 (14094) (13.09) 2052 1.91
1993 518 0.40 (13576) (10.61) 4311 3.37
1994 687 0.45 (12889) (8.38) 5819 3.78
1995 (2393) (1.43) (15282) (9.13) 3426 2.05
1996 (5695) (3.37) (20977) (12.42) (2219) (1.31)
1997 115 0.07 (20862) (12.52) (2104) (1.26)
1998 160 0.10 (20702) (12.31) (1944) (1.16)
1999 (3400) (2.15) (24102) (15.25) (5344) (3.38)
2000 (432) (0.27) (24534) (15.17) (4364) (2.70)
2001 518 0.31 (24016) (14.48) (3846) (2.32)
200224 538 0.31 (23478) (13.34) 38464 21.85
2003 771 0.43 (22707) (12.78) 39235 22.09
2004 2067 1.30 (20640) (13.01) 41302 26.03
(Adapted from the HA Annual Reports 1990 - 2004 and author's calculations)
In fact, the HA had already begun to receive financial assistance from the government in
1999 in regards to its negative equity position. Quoting the Board Chairman,
The government provided a grant of $2 million to improve Housing Authority's
negative equity position and to offset reduction in income arising out of the
implementation of new policies (The HA Annual Report, 1999:3).
The above observation implies that government had realised the negative equity position
of the Authority hence provided assistance. Also, the Labor government that came in
power in 1999 had put a halt to the privatisation policy. This means that the HA would
now possibly be able to receive grants from the government, which was ceased by the
previous government.
The HA reduced its lending rate for low income customers in September 1999 so the
government provided assistance to the HA through a grant of $2 million. The purposes
of the grant were to improve the Authority's financial position and assist in reducing
24 The positive balance for capital and reserves from 2002 is reflective of the ABD and World Bank loans converted into government's equity contribution.
Analyses of Earnings Management Practices in Fiji's SOEs
77
interest rates to customers. The entire amount was reported as an abnormal item in the
Income Statement. The grant was reported in accordance with FAS 20 Government
Grants (issued in 1992).
Government grants are rarely gratuitous. The enterprise earns them through
compliance with their conditions and meeting the envisaged obligations. They
should therefore be taken to income and matched with the associated costs which
the grant is intended to compensate (paragraph 8b).
In certain circumstances, a government grant may be awarded for the purpose of
giving immediate financial support to an enterprise rather than as an incentive to
undertake specific expenditures. Such grants may be confined to an individual
enterprise and may not be available to a whole class of beneficiaries. These
circumstances may warrant taking the grant to income in the period in which the
enterprises qualifies to receive it, as an abnormal item if appropriate, with
disclosure to ensure that its effect is clearly understood (paragraph 14).
The above two sections from FAS 20 justifies why the grant was reported in the Income
Statement as an abnormal item. Further, the accounting policy note on Government
Grants stated:
Government grants are recognised in the profit and loss statement over the periods
necessary to match them with the related costs that they are intended to compensate
(The HA Annual Report, 1999:17).
This implies that the entire amount that was received from the government in 1999 was
matched with its related expenses during the same year.
Eventually in October 2002, the Parliament approved to convert the $44 million owed to
the ADB and World Bank into government equity. This moved the HA’s negative equity
position of $3.8 million in 2001 to a positive position of $38 million in 2002. However,
this financial assistance from the government came with certain conditions. The HA and
the Government signed a Memorandum of Understanding, to ensure good corporate
Chapter 5: Evidence of Earnings Management Practices in the HA
78
governance and the financial viability of the HA in the future. As part of these
conditions, the Board is now accountable for good corporate governance through the
Minister for Housing and the Minister for Public Enterprise. In order to ensure this, the
HA created a Corporate Governance Division, headed by the Manager Corporate
Governance. The division's responsibility is to monitor and report on the progress of the
HA's Annual Work Plan on a quarterly basis and ensure that the performance
benchmarks agreed upon, under the respective divisions are met.
Changing the accounting policy had a substantial effect on the HA's net profits and the
equity balance. Table 5.17 demonstrates the effects of changing the accounting policy in
1999. As it is argued that the HA may have exaggerated the effects, the table also takes
into account the large increase in stock write down.
Table 5.17 Effects of Change in Accounting Policy at the HA in 1999
Change in
Accounting policy No Change in
Accounting policy
$000s % of Assets $000s % of Assets
Interest income 13893 8.79 13893 + 2600 = 16493 10.43
Profit before abnormal items
(1416) (0.90) (1416) + 2600 + 777 =
1961 1.24
Abnormal items (1984) (1.26) (1984) - 4291 = 2307 1.46
Profit after abnormal items
(3400) (2.15) 1961 + 2307 = 4268 2.70
Accumulated profits (24102) (15.25) (20702)25 + 4268 =
(16434) (10.40)
Equity (5344) (3.38) (5344) - (24102) + (16434) = 2324
1.47
(Adapted from the HA Annual Report 1999 and author's calculations)
With the change in the definition of non-performing loans, the HA lost about $2.60
million of interest income. Hence, if the change was not made, the HA would have
earned $16.49 million of interest income. Adding back the $0.78 million of stock write
down to income yields an operating profit before abnormal items of $1.96 million. In
order to determine the profit after abnormal items, the provision for doubtful debts of
$2.31 million is added back to yield $4.27 million. Thus, the table shows that presently
the net profit of the HA is negative 2.15% of total assets. If the policy were not revised,
25 This is the beginning balance for accumulated profits
Analyses of Earnings Management Practices in Fiji's SOEs
79
the net income would be much higher at positive 2.70%. Additionally, after the change
in accounting policy the equity position deteriorated to a negative 3.38% from positive
1.47%. The negative equity position was the major concern that the HA was indeed
using when making requests to the government to convert the debts into equity. Had the
accounting policy not changed, it would be less likely that the government would agree
to the conversion.
Table 5.17 presents the calculations with and without the change in accounting policy.
With FAS 30 being effective from 1999, the HA had to make the change. As it is argued
that the HA may have exaggerated the effects of provisioning, this could not be
incorporated in the above table. The extent of exaggeration is impossible to determine.
Nonetheless, the table shows that if the HA had not changed its accounting policy; it
would have made an operating profit of $4.27 million. This profit would have been
achieved despite that HA had reduced its lending rate. As already been noted, a high
possible reported profit is politically unacceptable in view of the Authority's social
objective. Therefore, the HA may not want to report high profits irrespective of
government policy and management incentive schemes. This proves to be another
incentive for exaggerating the effects of accounting policy change and reporting a loss in
1999.
In conclusion, it is argued that the actual change in accounting policy did not reflect an
attempt of earnings management. However, the timing of the change in 1995 is
questionable and it seems that the HA may have exaggerated the effects of the policy
change in 1999. While analysing the change in policies it was found that several possible
incentives existed that could have driven the change. These incentives include the need
for the incoming board and management to reflect improved performance, the non-
financial or social objectives of the HA and the conversion of debts into equity.
Chapter 5: Evidence of Earnings Management Practices in the HA
80
5.4 Conclusion
The in-depth analysis of the information obtained from the HA's annual reports and
other relevant documents suggests that there may be evidence of earnings management
practices in the relevant years. Table 5.18 summarises the evidence in light of the
possible incentives for its practice.
Table 5.18 Summary of Earnings Management Evidence
Year Event Incentive
1988 - 1991 Provision for write-down of inventory
Meeting stakeholders expectations and continue receiving funds for consultancies
1992-1994 Capitalising administration expenses
Reflect performance of the new government and new Board and introduction of performance based contracts
1993 Amortisation of deferred interest Smooth the trend in interest expense
1995 Provision for doubtful debts Social obligations
1999 Provision for doubtful debts Converting debt into equity
Four major possible instances of earnings management practices were identified and
these related to the provisioning for inventory, capitalised administrative expenses,
amortisation of deferred interest expense and changes in accounting policies for
provision for doubtful debts. To support the presence of such practices, analysis
indicates that there were several incentives for these events. These include the need to
meet expectations of the stakeholders, reflect performance of new government and the
incoming board and management, for annual bonuses and continuity of employment, to
satisfy the social objectives of HA and to convert debt into equity.
Analyses of Earnings Management Practices in Fiji’s SOEs
81
CHAPTER SIX
EVIDENCE OF EARNINGS MANAGEMENT PRACTICES IN
THE FIJI ELECTRICITY AUTHORITY
6.1 Introduction
This chapter analyses earnings management practices in the second organization, the Fiji
Electricity Authority (FEA). It begins by providing a brief corporate background of the
Authority, reviews its sources of electricity generation and describes the accounting
environment within which the FEA operates. The chapter then proceeds to identify and
analyse the possible instances of earnings management in light of the economic
incentives.
6.2 Overview of the Fiji Electricity Authority
6.2.1 Background
The FEA was established under the provisions of the Electricity Act 1966, which
constituted it to provide and maintain a power supply that is financially viable,
economically sound and consistent with the required standards of safety, security and
quality. When the Authority commenced operations on 1 August 1966, it was only
responsible for supplying electricity to Nadi, Lautoka and Levuka. Fiji’s biggest
consumer of electricity, Suva, was not under its jurisdiction. The Authority realised that
until it incorporated Suva within its market, its development would not proceed on a
satisfactory basis. In 1978, the FEA eventually acquired the assets of the Suva City
Council's Electricity Supply and became responsible for supplying electricity to Suva.
Subsequent to the acquisition, the Authority’s customers increased from 13,400 to
Chapter 6: Evidence of Earnings Management Practices in the FEA
82
32,000, a massive increase of 140%, and it evolved from being a regional to a national
entity. The rationale for taking over the Suva supply was that the Monasavu Hydro
Electric Scheme (discussed below) would only be financially viable if the FEA had a
larger customer base.
Fuel prices increased in the early 1970s due to the world oil crisis, making the
conventional diesel generators a very costly source of power generation. The
Government decided that hydro power was a viable option as it was expected to free Fiji
from some of the price fluctuations and other problems associated with an unstable oil
market. The government implemented the Monasavu Hydro Electric Scheme to generate
electricity. The FEA was given the responsibility for implementing this first hydro
scheme. Two out of the three stages of the scheme were completed in the second half of
1983 so the FEA began generating hydro electricity from October of that year. Since
then, electricity needs for Viti Levu have been met through hydro generated energy.
Diesel power stations were relegated to a standby role and were mainly used to supply
electricity in other areas of Fiji.
The Monasavu Hydro Electric Scheme saves the country over $40 million annually in
foreign currency and guarantees a secure source of supply through utilisation of a local
natural resource. The Chief Executive of the FEA stated in the 1991 Annual Report that
apart from the savings in foreign exchange, the Monasavu Hydro Electric Scheme had
met FEA’s expectations in operational efficiency. It exposed the Authority’s workforce
to state of the art technology that improved the employees' professional skills. Although
this scheme is an economic success, it represents a severe burden for the FEA in terms
of loan repayments as the total cost of financing the project was $240.50 million
(Parliamentary Debates 2003). The FEA had to borrow large sums of finance from
offshore lenders, such as the ADB ($28 million), Commonwealth Development
Corporation ($25 million), European Investment Bank ($34 million) and World Bank
($28 million) to fund the project26. The costs of servicing these foreign debts placed the
FEA into a difficult financial position, as around 40% of its revenue was used for paying 26 $73 million was borrowed from local lending agencies through bond issues while the remainder was financed by the government.
Analyses of Earnings Management Practices in Fiji’s SOEs
83
interest. The 33% devaluation of the Fiji dollar in 1987 (17.75% in June and 15.25% in
October) increased the Authority's debt by $69 million, thus compounding its
difficulties.
In 1993, the FEA management prepared a corporate plan to give the organization its
planned direction and future development strategies. After intensive consultations and
discussions, the plan was adopted in 1994. It set out the philosophies for guiding the
FEA into corporatisation, placing customer satisfaction at the forefront of its corporate
objectives. In 1997, the FEA was declared a Reorganisation Enterprise by the
Government and was to be split into three government commercial companies, each
responsible for Generation, Transmission and Distribution of power. The government
planned to partially privatise the Generation and Distribution companies by allowing a
strategic investor to bring in new technology, ideas and expertise while the third
company was to remain fully owned by the government. However, the restructuring plan
was reversed in 1999 by the new Labour government and the three companies instead
became three business units and two support divisions. They were placed under one
organisation, the FEA, which remained fully owned by the government.
The FEA Board comprises of a Chairman, Deputy Chairman and five other members, all
appointed by the minister responsible for Energy. The Chief Executive is an ex-officio
member and is responsible for the management and execution of the FEA's policies.
There were three occasions over the review period, where a new Board was appointed,
the most recent one being in 2001. Apart from this, the Board turnover was relatively
stable.
Chapter 6: Evidence of Earnings Management Practices in the FEA
84
6.2.2 Electricity Generation
Although, the hydro generation began operating efficiently since its inception, there was
a relative fall in its production capacity due to severe dry conditions and growing
consumer demand. The Monasavu dam reached a critically low level in 1992 due to the
prolonged drought which necessitated the use of supplementary diesel generators at
additional costs. Further, the consumer demand in Viti Levu had exceeded the long-term
average energy production capacity of the Monasavu Hydro Electric Scheme. The FEA
anticipated that if demand increases as expected, it will have to increase the use of diesel
generators to meet the hydro shortfall, which will significantly increase the FEA’s
operating expenditure. In light of this, the FEA commenced developing further hydro
generation as well as other alternatives (discussed below) to replace costly diesel
generators. In addition to supply issues, the FEA introduced Demand Side Management
programme in 1996 to help its customers, particularly industrial, to reduce their
maximum demands and thereby save on charges and improve their loading pattern. The
concept of Demand Side Management supplements the supply side planning activities
and presents new options to the Authority while trying to meet the demand in a cost
effective and efficient manner.
The FEA also explored the possibilities of environment friendly forms of power
generation. It pursued works on solar energy generation and in late 1997 installed a trial
10kW solar panel in Lautoka. As the solar energy production increased over the years, it
began to receive positive attention, particularly in light of ever increasing fuel costs. The
FEA intended to develop solar energy on a large scale so a feasibility study was
undertaken at Yaqara. Further, the FEA is embarking on a wind farm project in
Sigatoka. The Authority has already acquired the land from Native Land Trust Board on
a 50-year lease and expects to commission the project by mid 2006. Studies are also
underway for the development of wind farms in other parts of Viti Levu, Vanua Levu
and Ovalau. On its quest for renewable energy development, the FEA signed a joint
venture agreement with Australia based Pacific Hydro Limited, one of the largest
renewable energy companies in Australia. The FEA is jointly developing two hydro
Analyses of Earnings Management Practices in Fiji’s SOEs
85
projects worth $18 million with the Australian company. Currently, the FEA is
embarking on a plan of becoming a 100% renewable energy utility by 2011.
The FEA was seriously affected by the El Nino weather pattern that caused another
prolonged drought in Fiji in 2003. The Monasavu dam was below the minimum safe
operating level. The condition deteriorated in 2004 as Monasavu experienced the lowest
rainfall over the past 24 years. The FEA carried out a long-term plan, which showed that
the existing power capacity would not be able to meet electricity demand unless a new
power supply is developed by 2006. The FEA signed a 20 year energy conversion
agreement with Telesource (Fiji) to operate and maintain its diesel power stations at
Vuda and Kinoya. Telesource played an important role in the energy crisis of 2003 in
locating and helping to negotiate rental terms that brought the 38 generator sets required
to cover the shortfall in hydro generated electricity. There are three independent power
producers currently operating in Fiji, viz Fiji Sugar Corporation, Tropik Woods and
Emperor Gold Mines. They produce electricity for their own use and sell the surplus to
the FEA. A new Power Purchase Agreement was signed with Tropik Woods in 2004 to
increase their production for the FEA.
6.2.3 Accounting Environment and Financial Performance
The FEA prepares its financial reports on the basis of the FASs and Electricity Act 1966.
The IASs are also used where the Fijian counterpart does not provide adequate guidance.
From the beginning of the review period to 1994, the FEA’s financial accounts were
audited by Coopers & Lybrand while from 1995-1997, PriceWaterhouse took over the
audit responsibilities. With the merger of the two audit firms, the FEA's accounts were
audited by PricewaterhouseCoopers from 1998. The auditors' remuneration averages
around 0.04% of the FEA's income. Although there has been some rotation of the audit
firms over the review period, it is hard to determine if there was any rotation of the audit
partners, as the audit reports are signed off by the audit firm rather than the partner.
Chapter 6: Evidence of Earnings Management Practices in the FEA
86
The FEA's total assets increased from $337.39 million in 1990 to a peak value of
$526.96 million in 1996. Subsequently, assets declined to $452.12 million in 2004. Its
equity position was negative since 1987 but turned positive in 1991. This, however, was
achieved through profits and not capital injections. Equity was still low at $13.41
million, with a debt to equity ratio of 25:1. Following the revaluation of assets in 1992,
equity increased significantly and since then it has averaged around $283.30 million.
The FEA reported an average operating profit of $16.91 million from 1991-1999.
However, a loss was reported in 2000, which was largely due to the political crisis in the
country. The Authority returned to operating profits from 2001 but these were relatively
lower than what was being reported prior to the crisis. The return on equity in the 1990s
averaged around 7% and was reduced to 1% in the early 2000. Apart from the activities
related to its normal operations, abnormal items during the 1990s were quite frequently
reported in the statement of financial performance. These indicators are appended in
Appendix 3B.
6.3 Possible Instances of Earnings Management
6.3.1 Capitalisation of Research and Development Costs
The review of the FEA’s financial reports revealed that it has capitalised research and
development costs (R&D). The accounting policy it adopted is given below followed by
Table 6.1, which depicts the annual movements in the FEA's R&D over the relevant
periods.
Research and development costs are carried forward as an asset where future
benefits are expected. Amortisation will be effected against future income after
commercial production commences. Research and development expenditure which
no longer satisfies this criterion is written off against revenue. (The FEA Annual
Reports, 1985-1992).
Analyses of Earnings Management Practices in Fiji’s SOEs
87
Table 6.1 Research and Development Cost ($000s)
Year Beginning Balance
Increase During The Year
Amounts Written Off
Ending Balance
1985 3418 628 0 4046
1986 4046 498 413 4131
1987 4131 245 1461 2915
1988 2915 119 257 2777
1989 2777 19 24 2772
1990 2772 7 449 2330
1991 2330 7 372 1965
1992 1965 2 1967 0
(Adapted from the FEA Annual Reports 1985 -1992)
With this capitalisation, it is important to review the requirements of the accounting
standard. FAS 9 Accounting for Research and Development Activities (paragraph 3)
differentiates between research and development phases in the following manner:
Research is the original and planned investigation undertaken with the hope of
gaining new scientific or technical knowledge and understanding.
Development is the translation of research findings or other knowledge into a plan
or design for the production of new or substantially improved materials, devices,
products, processes, systems or services prior to the commencement of commercial
production.
Further, according to paragraph 15 of FAS 9, the costs of research and development
include the following:
a) the salaries, wages and other related costs of personnel engaged in research and
development activities
b) the costs of materials and services consumed in research and development
activities
c) the depreciation of equipment and facilities to the extent that they are used for
research and development activities
d) other costs related to research and development activities, such as the
amortisation of patents and licenses.
Chapter 6: Evidence of Earnings Management Practices in the FEA
88
The accounting standard requires R&D to be expensed in the period in which they were
incurred. However, development costs can be capitalised provided all the following
criteria according to paragraph 17 of FAS 9 are met:
a) the product or process is clearly defined and the costs attributable to the product
or process can be separately identified;
b) the technical feasibility of the product or process has been demonstrated;
c) the management of the enterprise has indicated its intention to produce and
market, or use, the product or process;
d) there is a clear indication of a future market of the product or process, or if it is to
be used internally rather than sold, its usefulness to the enterprise can be
demonstrated; and
e) adequate resources exist, or are reasonably expected to be available, to complete
the project and market the project or process.
Even after meeting the above criteria, all development costs still cannot be capitalised.
Only those costs, which can reasonably be expected to be recovered from related future
revenues, are deferred (paragraph 18). So the accounting regulation allows an entity to
capitalise development costs only and not the costs incurred for undertaking research
activities. Therefore, it is important to ascertain whether the capitalisations made by the
FEA relate only to development costs or it includes costs incurred for research activities.
This is discussed in the following section.
FAS 9 requires the following disclosures to be made for R&D:
i. The total of research and development costs, including the amortisation of
deferred development costs, charged as an expense
ii. The movement in and balance of unamortised deferred development costs
iii. The basis, proposed or adopted, for the amortisation of the unamortised balance
Analyses of Earnings Management Practices in Fiji’s SOEs
89
Along with the above disclosure requirements, paragraph 14 of FAS 9 states that:
Further information which might usefully be provided could include a general
description of the project, the stage which the project has reached and the estimate
future costs to complete it.
Although it is not mandatory, FAS 9 considers it useful that entities provide a general
description of the research and development project. However, the FEA’s annual reports
failed to do so. Thus, it is hard to verify for which product (or process) the research and
development project was undertaken. Also, the reports did not disclose the cost
components of the research and development activities, making it impossible to verify if
all the listed criteria in FAS 9 have been met. Further, R&D prior to 1985 was not
separately listed in the statement of financial position. Rather, it was aggregated with
capital works in progress. Capital works in progress represents expenditure in respect of
labour, material, services, an appropriate proportion of overheads and interest on
borrowings for specific capital works which were incomplete at balance date. It is
impossible to disaggregate the figures for R&D from capital works as the way the
information is presented does not permit to do so. Thus, the initial capitalisation of R&D
is unknown. The following table lists the capital works that were in progress in 1984.
Table 6.2 Capital Works In Progress In 1984
Items $000s
Building 481
Investigation 210
Plant, Machinery & Equipment 361
Reticulation 1623
Power Development 46159
Total 48834
(Adapted from the FEA Annual Report 1984)
Total capital works in progress in 1984 were $48.83 million, which includes $3.42
million of R&D (see Table 6.1). Therefore, the research and development activities
relates to one or a combination of assets in Table 6.2. One of the items listed as part of
capital works was ‘Investigation’. This item was first reported in the statement of
financial position in 1983, where $98,687 was shown under fixed assets and $32,713
Chapter 6: Evidence of Earnings Management Practices in the FEA
90
was reported as capital works in progress. In the following year, the amount reported
under fixed assets was transferred to capital works in progress. This item was not
reported in capital work in progress in the subsequent years. An item termed as
‘Investigation’ seems more likely to be the costs undertaken for research activities.
Hence, its inclusion as part of capital works suggests that the FEA has capitalised costs
incurred for research activities.
Moreover, the review of the FEA’s individual capital works revealed that for years up to
198327, they primarily related to the Monasavu Hydro Electric Scheme and following
that, it related to extension of transmission lines and telecommunications systems and
building/relocating new power stations and substations. This clearly shows that all
capital works for the FEA relates to assets that have long economic lives. Regardless of
whether the research and development expenditure relates to the Monasavu Hydro
Electric Scheme or other projects, it should have been amortised using a longer
economic life. The lives of the FEA’s assets range from 20-80 years (excluding motor
vehicle and furniture and fittings). Hence, if the capitalised R&D includes development
costs, it should have been amortised for at least 20 years. Referring back to Table 6.2, it
can be seen that this is not the case. There had been nominal capitalisations of R&D
from 1987 and large write offs. It is also noted that prior to 1987, annual capitalisations
exceeded write-offs. It can be clearly seen from the table that the write offs seem to be
discretionary, rather than based on any policy. Following a review in 1992, the FEA
expensed the remaining balance of R&D. There has been no further research and
development expenditure reported in FEA's financial accounts. The analysis so far
suggests that the capitalised R&D does include costs incurred for research activities.
During the same periods, FEA also capitalised training expenses. These expenses were
initially treated as part of capital works in progress in 1983 while from the following
year, it were separately reported as an intangible asset. The following policy note was
provided in its annual reports:
27 Although, the Monasavu Project was fully completed in 1986, all major works were completed in 1983 and the FEA began to generate hydro energy from that year.
Analyses of Earnings Management Practices in Fiji’s SOEs
91
During the construction of Monasavu Hydro Scheme $3,787,769 of training
expenses was capitalised, giving rise to an intangible asset in the books of the
Authority. This intangible asset is being amortised over 5 years. (The FEA Annual
Reports, 1984 - 1987).
Table 6.3 shows how training expenses of $3.79 million was amortised over the years.
Table 6. 3 Amortisation of Training Expenses
Year Amortisation Expense
($) Amortisation Expense/Initial Amount Capitalised (%)
1984 757554 20.00
1985 757000 19.99
1986 757600 20.00
1987 1515615 40.01
(Adapted from the FEA Annual Reports 1984 -1987 and author’s calculations)
Even though the FEA had adopted a policy of amortising training expenses over five
years, the expenses were fully amortised in the fourth year. In 1987, the FEA wrote off
the unamortised balance for training expense, as stated in the following policy note:
…however, the Authority views this benefit as fully realised and has amortised an
amount of $1,515,615 in 1987 to bring the balance to nil. The effect of this change
in policy is to increase operating loss by $758,015. (The FEA Annual Report,
1987:21).
The generally accepted accounting practices do not allow an entity to capitalise training
expenses. Further, it was noted that the FEA wrote off two years’ amortisation expense
in 1987. The Authority justified this on the grounds that the benefit from capitalising
training expenses was fully realised in the fourth year. This is purely an act of
management judgment, which would be difficult to question. The large write off of
training expenses was accompanied with a first time significant net write off of R&D.
The 1987 annual report was reviewed to find out the reason(s) for the large write-offs of
R&D and training expenditure. Apart from the suspension of capital works (which may
explain the large write off of R&D), there was no other relevant information provided,
which could explain them.
Chapter 6: Evidence of Earnings Management Practices in the FEA
92
Capitalisation of research and training costs are not permitted by the accounting
standards. By overstating its assets, the FEA was able to defer the recognition of its
expenses. There were unusually large write offs of these expenses in 1987, as it was an
opportune year to for the FEA to write off unwanted debits it has treated as assets. As a
result of political coups and devaluations, the readers would have expected substantial
losses during the year and may not be expected to question possible manipulations in
intangibles. Non-compliance with generally accepted accounting practices does not
necessarily constitute earnings management. It could be possible that the accountants
lacked understanding of the requirements of the accounting regulation on this aspect
therefore, making this a genuine case. However, one cannot overlook the possibility of
opportunistic behaviour. Therefore, it is important to determine if there was any
incentive for deferring these costs. This is discussed in the following section.
As part of the terms and conditions of the loans, the lending agencies required the FEA
to provide a return of 8% on currently valued fixed assets (the FEA Annual Report
1984). Table 6.4 depicts the return that the FEA was able to achieve by deferring the
recognition of research and training costs. The return is calculated using operating
profits before financing charges. The second last column shows that in the 1980s, except
for 1984, the FEA was not able to achieve the 8% return required but was close to the
value. It is important to calculate what the returns would have been had these costs not
been capitalised. However, this cannot be calculated with precision due to several
reasons. Firstly, the FEA calculated the above ratios by using separately maintained CPI
adjusted book values so the denominator in the calculation is unknown. Secondly, as
earlier mentioned, it could not be ascertained when R&D was first capitalised. Finally,
although it has been found that R&D capitalized does include research costs, the
component of each is impossible to determine. Keeping these factors in mind, the last
column of Table 6.4 calculates the approximate returns. As expected, there is a
noticeable effect in 1983, as this was the year when training expenses were first
capitalised. The return in fixed assets after capitalising training expenses yields 7.27%.
Had this cost not been deferred, the returns would have been 1.97%. In both cases, the
covenant is violated but the return without capitalisation is substantially lower than the
Analyses of Earnings Management Practices in Fiji’s SOEs
93
required return. Therefore, it can be argued that research and training expenses were
capitalised either to meet or smooth the return on fixed assets.
Table 6.4 Effect of Cost Capitalisations on Return on Fixed Assets
Year
R&D capitalised (amortised)
$000
Training Exp
capitalised (amortised)
$000
Profit with Capitalisation
$000
Profit without
capitalisation $000
Return with capitalisations
%
Return without
capitalisations %
1983 3788 5198 1410 7.27 1.97
1984 3418 (758) 21290 18630 11.98 10.48
1985 628 (757) 21223 21352 5.8 5.84
1986 85 (758) 27774 28447 6.17 6.37
1987 (1216) (1516) 31519 34251 6.53 7.27
1988 (138) 9669 9807 7.74 7.77
1989 (5) 29736 29741 7.35 7.35
1990 (442) 34352 34794 7.21 7.30
1991 (365) 40179 40544 8.01 8.08
1992 (1965) 40050 42015 8.1 8.50
(Adapted from the FEA Annual Reports 1983 -1992)
Studies have clearly shown that debt covenants are one of the strongest incentives for
earnings management in the private sector. Accordingly, one might argue that as FEA's
debts are guaranteed by the government, observing the terms of a debt covenant cannot
be an incentive for manipulation of profits. It cannot be denied that the presence of
government guarantees weakens this incentive in FEA. However, it can also be argued
that government as a guarantor can pressure FEA to satisfy the covenants. Therefore, in
FEA's case, the pressure to report certain numbers in the financial reports comes more
from the government than meeting the lenders terms. There are occasions where the
FEA has been penalized for not meeting the covenants. The penalties were in the form
of government declining to guarantee further debts and lenders not allowing the raising
of further long- term funds. In conclusion, the analysis indicates that the capitalisation of
research and training costs may have been undertaken with earnings management intent,
since it helped the FEA to meet the covenant set by the lenders.
Chapter 6: Evidence of Earnings Management Practices in the FEA
94
6.3.2 Revaluation of Fixed Assets
Fixed assets of the FEA were revalued on 1 January 1992 by independent valuers. The
existing use value was selected for most of its assets as recommended by its valuers,
except for motor vehicles and furniture and fittings. These two classes of assets were not
revalued due to their more frequent renewal and as any difference between the revalued
and written down values were deemed to be immaterial. The revaluation resulted in an
increase in the value of fixed assets by $164.08 million and the corresponding credit was
made to a newly created Asset Revaluation Reserve. There was a consequential increase
in depreciation of $3.39 million during the year, which was reported as an abnormal
item in the Income Statement. Table 6.5 illustrates the key items of the financial
statements that were affected by the revaluation in 1992.
Table 6.5 Effects of Revaluation in 1992
With Revaluation Without Revaluation
Balance Sheet Accounts ($000s)
Capital 74845 74845
Accumulated Losses (58007) (58007) + 3394 = (54613)
Asset Revaluation Reserve 164076 0
Other Reserves 23455 23455
Total Equity 204369 43687
Fixed Assets 483418 483418-164076 = 319342
Total Assets 509885 509885-164076 = 345809
Profit and Loss Accounts ($000s)
Depreciation expenses 12657 12657 - 3394 = 9263
Net profit 15107 15107 + 3394 = 18501
Debt to Equity Ratios
Total debt Total equity
305,516,000 204,369,000
= 1.49:1
305,516,000 43,687,000 = 6.99:1
Long-term debt Total equity
213,287,000 204,369,000
= 1.04:1
213,287,000 43,687,000
= 4.88:1
(Adapted from the FEA Annual Report 1992 and author’s calculations)
Analyses of Earnings Management Practices in Fiji’s SOEs
95
The revaluation exercise significantly improved the FEA's equity position from $43.68
million without revaluation to $204.37 million, representing an increase of almost 370%.
Further, the higher equity balance helped FEA to reduce the debt to equity ratios. Total
debt to equity ratio fell from 6.99:1 to 1.49:1 and long-term debt to equity ratio also
moved in the similar manner from 4.88:1 to 1.04:1. The following comments were made
by the General Manager Finance in relation to the revaluation exercise:
As a result of continuing profitability and the benefits of the asset revaluation
exercise, the Authority's debt to equity ratio as reflected in the audited published
accounts has now reached a commercially acceptable level. This ratio stands at
60/40 and is equivalent to the accepted industry norm. By excluding all short term
funding in accordance with World Bank and other overseas lending criteria, the
ratio is closer to 50/50. (The FEA Annual Report, 1992:27).
Although, the revaluation reduced the FEA's net profit by 18% through additional
depreciation charges, it proved to be beneficial as it aligned the debt to equity ratios of
the FEA to a commercially acceptable level and within the guidelines set by the offshore
financiers. While the FEA borrows from the local as well as overseas financiers, during
these periods offshore loans were quite substantial, around 50% of total borrowings.
With the exception of the loan from Suva City Council, all of the FEA's indebtedness is
guaranteed by the government, which makes the FEA a less risky borrower to its
lenders. However, even when government guarantees the loans, the FEA still has to
meet the covenants set by the lending agencies, as discussed in the earlier section.
As part of the covenants, the offshore lending agencies required the Authority to
maintain a debt to equity ratio of 3:1 (The FEA Annual Report, 1984). The devaluations
of 1987 resulted in $99.17 million of unrealized foreign exchange losses, which
consequently eroded the equity position of the Authority through accumulated losses.
Since then, the equity position of the FEA had been negative and it was unable to meet
the debt to equity covenant. Although, capital contributions were made by the
government, they were insufficient to offset the deteriorated equity balance. The analysis
Chapter 6: Evidence of Earnings Management Practices in the FEA
96
shows that the revaluation exercise not only improved the equity balance but also
assisted the FEA to meet the covenant of 3:1 debt to equity ratio.
Moreover, from 1992 the FEA adopted a policy of revaluing its assets every five years.
The next valuation was to be undertaken in 1997 but was deferred due to the Authority's
anticipated corporatisation. Despite the FEA's policy, it has never undertaken a
revaluation exercise in the past 12 years since 1992. It is worth noting that reporting
entities in Fiji generally face real problems in revaluing assets as it is a costly exercise28.
The deferral of subsequent valuation and the associated costs of revaluation further
support the argument that the revaluation in 1992 was undertaken with an opportunistic
intent. The frequency of the revaluation is dependent on the movements in the fair
values of non-current assets (FAS 16, paragraph 32) and not the need to improve the
equity position of an entity. Therefore, the analysis indicates that the revaluation in 1992
may have been used as an earnings management tool to improve the deteriorating level
of equity position and to meet the debt covenants. This argument is strengthened in the
following section, which looks at the classification of government grants by the FEA.
6.3.3 Classification of Government Grants
Deferred income of the FEA relates to grants-in-aid, representing the fair value of non-
monetary assets received in the form of grants. These were materials received from the
overseas government, particularly the Chinese government29. These were received
through a bilateral procedure, where the overseas government provided the materials to
the FEA through the Fiji government. These grants were not in the form of cash. From
1995, deferred income also includes ‘Government and FEA Contributions for Rural
28 The cost of valuation in Fiji is as follows: valuation of $10,000 = $100 + total expenses valuation of $10,000 to $50,000 = $220 + $2 for every $1000 + total expenses valuation of $50,000 to $250,000 = $620 + $1.50 for every $1000 + total expenses (These figures were obtained from Mr Hassan, who is a registered valuer in Suva.) 29 This information was supplied by the former Chief Executive of the FEA, Mr., Nizam-Ud Dean, who has been with the Authority since the late 1980s.
Analyses of Earnings Management Practices in Fiji’s SOEs
97
Electrification’, which are cash contributions made by each towards rural electrification.
As advised by the Ministry of Finance, the FEA converted a $5.50 million government
loan into equity in February 1990. The Ministry later discovered that this transaction
was not approved in the parliament in 1990 so the House of Representatives passed a
motion in February 1996 to convert the Loan into a Government Grant for Rural
Electrification instead of loan to equity conversion. This represented the Government's
contribution for rural electrification. Since the conversion of debts into equity was not
approved by the Parliament in 1990, the FEA was liable for interest on the loan from
1990 to 1996. On the request of the Government, the FEA contributed the equivalent
amount of interest forgone towards Rural Electrification, which represented FEA's
contribution.
The FEA reported deferred income as part of equity for years prior to 1995 while from
the following year it was reported as a non-current liability. This reflects inconsistency
in reporting government grants. Alternatively, the nature of the grants may have changed
from 1995 hence the differing treatments. Apart from the above described fact about the
FEA's grants, there was no information disclosed in the annual reports suggesting that
the nature of grants had changed. Also, the accounting polices relating to government
grants during the concerned periods remained unchanged as given below:
Grants in aid received in kind have been recorded as deferred income which is
recognised in the Income Statement on a systematic and rational basis over the
useful life of the assets. (The FEA Annual Reports, 1990-91).
…grants in aid and assets acquired at no cost to the Authority are capitalised and
progressively recognised as Other Income on the basis of the average lives of the
assets to which the grants were deemed to relate. (The FEA Annual Reports, 1992-
96).
Given the inconsistency in the classification of grants and the absence of any
explanation to prove otherwise, it is important to ascertain which of the two
Chapter 6: Evidence of Earnings Management Practices in the FEA
98
classifications (equity or liability) is correct. According to FAS (IAS) 2030 Accounting
for Government Grants and the Disclosure of Government Assistance, there are two
accounting treatments of government grants; the capital approach and income approach.
Under the capital approach, the grant is taken directly to the equity section of the
balance sheet while the income approach requires the grant to be taken to income on a
systematic basis. As grants-in-aid relates to assets, FAS 20 further states the following
about grants relating to assets:
Government grants, including non-monetary grants at fair value, should not be
recognised until there is reasonable assurance that (i) the enterprise will comply
with the conditions attaching to them, and (ii) the grants will be received. They
should not be credited to shareholder's interests. (Paragraph 37)
Government grants should be recognised in the income statement over the periods
necessary to match them with the related costs which they are intended to
compensate, on a systematic basis. (Paragraph 38)
FAS 20 requires non-monetary grants to be accounted for using the income approach. It
is further stated in paragraph 39 that grants related to assets should either be presented as
deferred income (liability) or deducted from the carrying amount of the asset. The nature
of FEA’s grants requires it to apply the income approach. Thus, its treatment from 1995
is correct while in the earlier years it resulted in non-compliance with FAS 20. It should
be noted that accounting standards were revised in 1996, but there was no change in
FAS 20. Hence, the differing treatment could not be the result of change in accounting
practice. As indicated earlier, non-compliance with accounting standards does not
always constitute earnings management. It could be possible that the FEA made a
genuine error in classifying government grants. It is worth pointing out that the
misclassification did not occur in a single but a series of years. With due respect, the
auditors should have been able to spot this error. Their failure to do so suggest either the
audit team was lax or they "condoned" the accounting treatment. Also, during the years
of misclassification, the same firm audited their accounts and the correct approach was
applied when there was a change their auditor. In light of this, the FEA's 30 FAS 20 was issued in 1992 and was identical to IAS 20. In the absence of any national accounting standards on government grants prior to 1992, the FEA relied on IAS 20.
Analyses of Earnings Management Practices in Fiji’s SOEs
99
misclassification seems more to be opportunistic rather than a general error in practice.
This undoubtedly seems to be a case where the FEA has stretched the limits of generally
accepted accounting practices to the extent that it actually led to non-compliance. It
should, however, be noted that the misclassifications did not affect profits. In both the
treatments, grants were amortised to the Income Statement. The following table
illustrates the effects of reporting government grants as equity in years prior to 1995.
Table 6.6 Effects of Misclassification of Deferred Income
Year Deferred Income Taken To Equity Deferred Income Taken To Liability
Equity $000s
Total Debt To Equity Ratio
Long-Term Debt To
Equity Ratio
Adjusted Equity $000s
Total Debt To Equity Ratio
Long-Term Debt To
Equity Ratio
1990 (3929) (86.86) (61.24) (3929) – 1911 =
(5840) (58.77) (41.53)
1991 13407 24.52 18.32 13407 - 4891 =
8516 39.18 29.42
1992 204369 1.49 1.08 204369 – 12077
= 192292 1.65 1.21
1993 224677 1.29 0.94 224677 – 12781
= 211896 1.43 1.06
199431 241995 1.11 0.99 241995 – 13125
= 228870 1.23 1.10
(Adapted from the FEA Annual Report 1990-1994 and author’s calculations)
Table 6.6 shows that the misclassification of grants improved the equity balance of the
FEA, which consequently affected its debt to equity ratios. Applying the capital
approach improved the total debt to equity ratios and long-term debt to equity ratios.
Hence, the motive behind the misclassification was to improve the equity position of the
FEA. In the earlier section, it was found that the FEA revalued its assets in 1992 and this
section revealed that it misreported government grants as equity instead of liabilities in
years prior to 1995. It seems that both these events were undertaken with a view to
improve the FEA's equity position and consequently the debt to equity ratios. Table 6.7
shows the combined effects of revaluation and misclassification of government grants.
31 The table is not extended beyond 1994 as grants were correctly reported from 1995.
Chapter 6: Evidence of Earnings Management Practices in the FEA
100
Table 6.7 Combined Effects of Revaluation and Classification of Deferred Income
Scenario 1 No Revaluation and Deferred Income is Reported as Liability
Year Equity ($000s)
Non-Current Liabilities ($000s)
Total Liabilities ($000s)
Total Debts to Equity
Long-Term Debts to Equity
1989 (7756) 258943 342294 (44.13) (33.39)
1990 (5840) 242545 343234 (58.77) (41.53)
1991 8516 250528 333669 39.18 29.42
1992 31610 233425 317593 10.05 7.38
Scenario 2 No Revaluation and Deferred Income is Reported as Equity
Year Equity ($000s)
Non-Current Liabilities ($000s)
Total Liabilities ($000s)
Total Debts to Equity
Long-Term Debts to Equity
1989 (5832) 257019 340370 (58.36) (44.07)
1990 (3929) 240634 341323 (86.87) (61.25)
1991 13407 245637 328778 24.52 18.32
1992 43687 221348 305516 6.99 5.07
Scenario 3 Revaluation and Deferred Income is Reported as Equity
Year Equity ($000s)
Non-Current Liabilities ($000s)
Total Liabilities ($000s)
Total Debts to Equity
Long-Term Debts to Equity
1989 (5832) 257019 340370 (58.36) (44.07)
1990 (3929) 240634 341323 (86.87) (61.25)
1991 13407 245637 328778 24.52 18.32
1992 204369 221348 305516 1.49 1.08
(Adapted from the FEA Annual Report 1990-1992 and author’s calculations)
The table presents three scenarios. Scenario 1 analyses the effects if there were no
revaluation and deferred income was reported as a liability. In the second scenario, again
there is no revaluation but deferred income is now classified as equity. The third
scenario describes what actually happened at the FEA and the values for this scenario
are directly taken from its financial statement. It presents a case of a revaluation and
deferred income being reported as equity. It was impossible to extend the table beyond
1992 since the full effects of a revaluation cannot be determined precisely in the future
years. Given that revaluation results in increased depreciation expense in the current
year as well as in the subsequent years, the extent of the increase in depreciation expense
in subsequent years cannot be determined.
Analyses of Earnings Management Practices in Fiji’s SOEs
101
In 1989 and 1990, the FEA had a negative equity position hence the debt to equity ratios
in these years are not really meaningful. The difference in the ratios is more observable
in the remaining years. In scenario 1 with no revaluation and deferred income correctly
reported as a liability, the total debt to equity ratio and long-term debt to equity ratio was
the highest of all the three scenarios. By reporting deferred income as equity, the FEA
was able to reduce the ratios in scenario 2. Although, the ratio was reduced, it was still
high when compared with industry standards and those set by the overseas lenders.
Hence, the revaluation exercise in the third scenario produced the lowest debt to equity
ratios, which were now within the industry and overseas lenders' standards. The
combined analysis illustrates that the FEA had intended to reflect an improved debt to
equity ratios. This was initially achieved by classifying deferred income as equity and
the subsequent revaluation activity further strengthened the financial position. Hence,
the revaluation and the misclassification of government grants might be a possible case
of earnings management at the FEA.
6.3.4 Capitalisation Policies
The cost of the FEA's fixed assets includes the cost of all materials, direct labour and
associated overheads. Over the review period, there were two occasions where the FEA
changed the policy of capitalising overheads to property, plant and equipment. The first
change was made in 1994 where certain overheads, which could not reasonably be
identified as forming part of the cost of purchase or acquisition of an asset or of bringing
it to its working condition were expensed from 1 January. These overheads were
previously capitalised to property, plant and equipment. Two years later, the FEA ceased
to capitalise administrative salaries and motor vehicle operating costs to the acquisition
cost of property, plant and equipment. The FEA stated that by expensing these costs
from 1996, they were able to reflect actual operating expenditure. It should be noted that
neither of the two changes was done retrospectively. The changes increased the FEA's
operating expenses by $0.90 million and $2.50 million in 1994 and 1996, respectively.
These are reflected in Table 6.8. In 1994, the operating expenses represented 74% of
Chapter 6: Evidence of Earnings Management Practices in the FEA
102
income and in 1996 it increased to 78%. Higher operating expenses reduced FEA’s
profits during the concerned periods.
Table 6.8 Total Expenses and Profits of the FEA
Year Total Expenses Profit
$000s % of Income $000s % of Assets
1992 36907 68.29 17294 3.39
1993 39636 68.75 18189 3.53
1994 45260 73.80 16379 3.21
1995 48343 70.97 20470 3.92
1996 56116 78.20 17320 3.29
1997 55605 76.60 18408 3.68
1998 63472 74.91 21256 4.27
1999 83378 90.23 9028 1.87
2000 93694 107.64 (6647) (1.40)
2001 97971 99.97 31 0.01
2002 110886 96.70 3787 0.83
2003 125204 97.96 2609 0.57
2004 119449 92.77 9306 2.06
(Adapted from the FEA Annual Report 1992-2004 and author’s calculations)
Given the changes in the capitalisation policies, it is important to find out the
requirements of the accounting standard on this aspect. FAS 16 Accounting for Property,
Plant and Equipment (paragraph 11) states that
the cost of an item of property, plant and equipment comprises its purchase price,
including import duties and non-refundable purchase taxes, and any directly
attributable costs of bringing the asset to working condition for its intended use; any
trade discounts and rebates are deducted in arriving at the purchase price.
Examples of directly attributable costs are:
a) site preparations;
b) initial delivery and handling costs;
c) installation costs, such as special foundations for plant;
d) professional fees
Analyses of Earnings Management Practices in Fiji’s SOEs
103
The accounting standard further states that
administration and other general overhead expenses are not a component of the
cost of property, plant and equipment, unless they can be specifically related to the
acquisition of the asset or bringing it to its working condition ( Paragraph 13).
Paragraph 13 illustrates an example of accounting regulation that allows management to
apply discretion and flexibility in financial reporting. According to the accounting
standard, the capitalisation or current expensing of administration and other general
overhead expenses is dependent on whether these expenses were related to the assets
acquisition or bringing it to its working condition. This criterion is available to the
preparers of financial reports so that the economic reality of the transactions can be
reported. However, in the presence of economic incentives, the same condition can be
used to misreport the reality. The FEA used this condition to justify their policy change
by stating that these costs are expensed as these are not specifically related to the assets
acquisition. On the surface of it, it seems that the policy change was made to reflect
economic reality. Knowing that the determination of whether an expenditure is an asset
or an expense can have a significant effect on the entity’s operating results, it could also
be possible that the policy change was an attempt of earnings management. This
possibility is supported by the fact that there were incentives for the FEA to report
reduced earning in these periods. These incentives will be discussed in the next section
together with the changes in depreciation rates as these changes would appear to have
been made for same reasons.
6.3.5 Changes in Depreciation Rates
During the review period, there were two occasions where the FEA changed the
depreciation rates for fixed assets. The first revision was in 1999 when the FEA began to
apply tax depreciation rates, which were higher than the previously applied rates.
Although the rate had varied, the method of depreciation remained unchanged as straight
line. The depreciation rates were changed for the new assets as well as assets acquired
Chapter 6: Evidence of Earnings Management Practices in the FEA
104
earlier than 1 January 1999. As a result, depreciation expense increased by $21.16
million during the year. As changing depreciation rates have on-going effects, the FEA
has managed to increase its depreciation expenses in the future periods as well. The
following table illustrates the Authority's depreciation expenses and its impact on
profits.
Table 6.9 Depreciation Expenses and Profits of the FEA
Year Depreciation Profit If the rates were not changed
$000s % of PPE $000s % of Assets Depreciation $000s
Profit $000s
1992 9263 1.87 17294 3.39
1993 13671 2.64 18189 3.53
1994 14118 2.68 16379 3.21
1995 14309 2.66 20470 3.92
1996 14484 2.67 17320 3.29
1997 14538 2.64 18408 3.68
1998 14391 2.56 21256 4.27
1999 36095 6.31 9028 1.87 14467 30656
2000 35733 6.22 (6647) (1.40) 14535 14551
2001 35902 6.03 31 0.01 15056 20877
2002 41030 6.39 3787 0.83 16257 28560
2003 43301 6.52 2609 0.57 16791 29119
2004 20260 3.02 9306 2.06 16970 12596
(Adapted from the FEA Annual Reports 1992-2004)
The profits reduced to 1.87% of assets in 1999 when compared to 4.27% in the
preceding year and the FEA attributed this to the change in depreciation rates. The
following statement was made by the Authority on its financial performance:
Net profit before abnormal items and income tax decreased from $21.3 million to $9
million in the current year. This was due mainly to a change in accounting policy
when FEA adopted tax depreciation rates from 1 January 1999. The Authority
considered it prudent to adopt tax depreciation rates because it reflects a more
realistic life of the fixed assets and FEA became a taxable entity from 1 January
1998. (The FEA Annual Report, 1999:22).
Analyses of Earnings Management Practices in Fiji’s SOEs
105
Table 6.9 also illustrates the depreciation expenses and profits that would have been
reported if the rates were not changed. Depreciation rates are calculated using an
average rate of 2.53% of gross property, plant and equipment.32 It is evident from the
table that the change in rates has significantly affected FEA’s profits. Depreciation rates
were changed for two reasons; prudence and the FEA’s taxable status. According to
FAS 1 Disclosure of Accounting Policies, prudence is one of the factors that govern the
selection and application of accounting policies. Paragraph 8(a) of the standard states
that:
Uncertainties inevitably surround many transactions. This should be recognised by
exercising prudence in preparing financial statements. Prudence does not, however,
justify the creation of secret or hidden reserves.
Upon selecting and applying accounting policies, entities need to exercise prudence. The
FEA asserts that by employing tax depreciation rates, it would be able to reflect a more
realistic life of its fixed assets. This reason seems to be justifiable as long as the FEA
had not used it for the purpose of creating hidden reserves. Nonetheless, it should be
noted that tax depreciation rates are arbitrary and apply to certain types of assets rather
than to industries. It is impossible for tax depreciation rates to fairly reflect the
economically useful life of an asset. As they are often structured to provide investment
incentives, tax depreciation rates may not be specifically related to the assets economic
lives. Further, being a taxable entity does not oblige it to adopt depreciation rates used
by the tax authorities; it can still employ the accounting rates. Hence the FEA’s reason
to adopt tax depreciation rates due to its taxable status seems to be neither necessary nor
justifiable. The increase in depreciation rates and changes in capitalisation polices in
1994 and 1996 discussed earlier seem to be related. As a result of all the three changes,
the FEA’s expenses (depreciation and operating) increased resulting in reduced
profitability. There were possible motives for the FEA to report reduced earnings during
these periods.
32 This average is computed using 1992 to 1998 values in Table 6.9 since the FEA changed the format of presenting its fixed assets and their corresponding depreciation charges. Unlike in the previous years, from 1999, the Authority ceased to separately disclose its individual assets.
Chapter 6: Evidence of Earnings Management Practices in the FEA
106
The FEA was involved in a long running debate with the government over their tax
status. Since inception it has been exempted from tax in pursuant to section 17(25) of the
Income Tax Act. However, the Income Tax (Amendment) Decree repealed the FEA’s
tax exempt status from 1 July 1992 given that the FEA was expected to be corporatised
in the near future. The FEA made submissions to the appropriate Ministries for an
extension of the tax exempt status, back dating from 1 July 1992 to 31 December 1996.
In light of this, the FEA did not adopt tax effect accounting and had not provided for any
taxes payable on its profits during these periods. If the FEA's submissions had not been
accepted, its tax payable from 1 July 1992 to 31 December 1996 would have been
expected to be $23 million (Parliamentary Debates, 2005). As a result of this, the
auditors had raised a matter of emphasis when signing off the 1996 financial statements.
The FEA argued that its operational costs made it difficult to cope with the universal
service obligations that they are undertaking33. The cost of these services is an average
of $25 million per year. Although the Public Enterprise Act requires the government to
reimburse the universal service obligation costs to the FEA, such costs have not been
refunded. The Authority argues that it was able to afford the universal service obligation
in the past when it was exempted from paying tax and dividends to the government. This
has changed as it is now required to pay taxes and provide dividends. Further, the FEA
as a commercial statutory authority is required to earn the government a target return on
equity of 10% from 2002.
The FEA’s Board Chairman asserted that as long as the government reimburses these
costs, the Authority could earn the benchmark return:
The government as its right as a shareholder, expects a 10% return on investment
(ROI) from all CSA, like FEA. This is not an unreasonable demand. In terms of
profit, it means that FEA should be making over $30m a year. In other words, if the
Government were to reimburse for its social obligation cost as intended under the
Public Enterprise Act, then its financial performance is right on track with the
current reforms. The government will need to seriously consider reimbursing FEA 33 As a universal service obligation, the FEA subsidises power to rural areas in Viti Levu as well as the entire islands of Vanua Levu and Ovalau.
Analyses of Earnings Management Practices in Fiji’s SOEs
107
for these social obligations subsidies if it insists on this return benchmark. If not, it
will have to allow an increase in tariffs. (The FEA Annual Report, 2002:9)
In September 2002 the Cabinet finally decided to extend the FEA’s tax exempt status
from 1 July 1992 to 31 December 1996. It was also decided that the FEA’s universal
service obligation be recognised as its annual 'social' dividend to the government.
Although, the recognition of the universal service obligation cost as a social dividend is
a relief to the FEA to some extent, the Authority still argues that:
…if it (universal service obligation costs) were part of the profit, it would mean that
we would have to pay well in excess of 90% of our profit as a dividend. For such a
capital intensive industry, this, as a dividend policy is financially unsustainable and
untenable. (The FEA Annual Report, 2003:10)
The above discussion highlights that the FEA sought to extend its tax exempt status and
reimbursement of universal service obligations costs. The decision would be made by
the government so the financial performance of the FEA has to be reported such that it
can make a justifiable case to the government. An entity with marginal or reduced
profitability has a high chance of getting a favorable response in contrast to an entity
with a series of high reported profits. The changes made to the capitalisation policies in
1994 and 1996 have led to a continuous reduction in earnings. The increase in the
depreciation rates from 1999 further reduced profitability (see Table 6.9). Therefore, the
analysis indicates that the changes to the accounting polices may have been used as an
earnings management device since the changes have undoubtedly assisted the FEA to
make a permissible case to the government.
There was yet another change in the depreciation rates in 2004. Following a
comprehensive review of the remaining useful lives of its asset base, the FEA adopted
new depreciation rates from 1 January 2004. These rates were lower than the previously
used tax depreciation rates and for most of the assets, were the same as the ones that
were applicable prior to 1999. By applying lower depreciation rates, the FEA was able
to reduce its depreciation expense from 6.52% of gross property, plant and equipment in
Chapter 6: Evidence of Earnings Management Practices in the FEA
108
2003 to 3.02% in 2004 and return on assets increased from 0.57% to 2.06% (Table 6.9).
The FEA again attributed the year's profit to the changes in depreciation rates. The
Chairman of the Board indicates:
Whilst it (the FEA) recorded a headline operating profit before tax of $9.3m, this
has really been the result of a significant reduction in depreciation charges due to
lower accounting depreciation rates used for 2004 (The FEA Annual Report,
2004:7).
When the depreciation rates were increased in 1999, the FEA argued that this was
undertaken to reflect a more realistic life of its fixed assets. It then believed that the
fixed assets had a shorter useful life. When reducing the depreciation rates in 2004, the
Chief Executive of the FEA argued the following:
Operating profit before tax is $9.3m. This would have been a loss of $13.7m if the
depreciation rates used by FEA for accounting purposes were not revised
downwards. Lower accounting depreciation rates were adopted by FEA in 2004
following a comprehensive review of the remaining useful lives of its existing asset
portfolio. Previously higher depreciation rates allowed by the Income Tax Act for
tax depreciation purposes have been used for accounting purposes also. Given the
longer useful lives of FEA’s assets such as hydro dams, generators and plant and
equipment, the Board considered that the lower accounting depreciation rates will
reflect fair and more reasonable financial statements as required under the current
financial reporting standards. The effect on the income statement is a reduction in
the depreciation charge of $23m. (The FEA Annual Report, 2004:17).
The reduced depreciation rates were justified by the FEA on the grounds that lower rates
reflected a fair and reasonable financial statement. It now believed that its fixed assets
had longer lives. It was also noted that the new rates resulted in $9.30 million of
operating profit before tax. If FEA had continued to use the higher tax depreciation
rates, it would have ended up with a major loss of $13.70 million. On the basis of the
Chief Executive’s statement, one can argue that the FEA had not wanted to report a
Analyses of Earnings Management Practices in Fiji’s SOEs
109
material operating loss. This argument is further supported by the fact there were
possible incentives for the FEA to report high earnings in 2004.
The FEA has been operating at near full capacity since the mid 1990s and the demand
for electricity has also been escalating. In order to meet electricity demand, the FEA
argues that it needs to undertake capital investment of $500 million. Its surpluses have
clearly been inadequate to finance the investment needed, hence it requires external
funding. In order to attract outside financiers, the FEA needs a healthy financial
performance and position. By reducing the depreciation rates in 2004, the FEA was able
to avoid reporting significant losses as a loss making entity would fail to attract potential
investors.
To further improve its operating results, the FEA intended to increase the electricity
tariff rates. Since the 10% increase in tariff in 1991, there has been no other increase in
the rate and the FEA contends that its current tariff structure would not attract investors.
The Board Chairman argued that
...the tariff structure that prevailed at the beginning of 2004 is inadequate to
provide the necessary cash flows for FEA to fund building new power capacity.
Also, the tariff rates were considered too low to attract private investors and
independent power producers as there were not enough economic returns to such
investors. Therefore one of the main objectives of FEA was to achieve an
appropriate increase in tariff rates which have not increased since 1991 and in fact
have reduced three times, in 1996, 1997 and 1998. (The FEA Annual Report, 2004:
9-10).
However, the increase in tariffs would be decided by the government on
recommendations of the Commerce Commission so in the second half of 2004, the FEA
made a submission for a substantial increase in electricity tariff to the Commission34.
The Commission reported that
34 The Commerce Commission is an independent statutory body, established under the provision of the Commerce
Act 1998. The Commission promotes effective competition and informed markets, encourages fair-trading, protects
Chapter 6: Evidence of Earnings Management Practices in the FEA
110
In its submission, FEA requires up to $500 million to meet new capital
investment to cater for the increased power capacity and to achieve 10% Return
on Shareholders Fund (ROSF) in line with the Public Enterprise Act. (The
Commerce Commission Report, 2004:3).
According to the Commerce Commission’s Report, the FEA proposed three major
reasons, amongst others, for the increase in tariff. Firstly, the FEA argues that the higher
prices would provide new capacity to meet new investment. It states that unless it
installs new power capacity, there would be power shortages after 2006. Based on past
trends, it anticipates that growth in future electricity demand would not be met by
current capacities and its diesel generators are already 30 years old and are becoming
unreliable. Secondly, it argues that the current tariff is inadequate to meet new funding
and ROE of 10%. In order to increase its power capacity, it would need to undertake
substantial borrowing, which would increase its debt level and interest rates. These
finance charges would not enable the FEA to meet the 10% bench mark return. The FEA
argues that the tariff increase would bring it closer to meeting government’s ROE,
maintaining acceptable debt levels as well as meeting new investment requirements.
Finally, of the $500 million capital investment needed, the FEA expected that
Independent Power Producers (IPPs) would contribute $150 million. The FEA argues
that the current tariff structure would not attract IPPs as it would fail to provide the
return that they would require to invest. Thus, higher prices would encourage IPPs.
When making a case to the Commission, the FEA reviewed its accounting policies and
decided, inter alia, to revalue its non-current assets.
The FEA points out in its submission that it has revalued its non-current assets.
In so doing, it increases the denominator in the ROSF calculation, thus driving
up the tariff required to secure the 10% required by the government. (The
Commerce Commission Report, 2004:14)
consumers and businesses from restrictive trade practices and control prices in regulated industries and other markets where competition is lessened or limited.
Analyses of Earnings Management Practices in Fiji’s SOEs
111
A revaluation requires the recognition of a Revaluation Reserve as part of equity. Given
that the valuation of non-current assets will increase, depreciation expenses will also
increase, which would then reduce the profits of the FEA. So revaluation will increase
the equity and reduce profit, thus reducing the ROE. Hence, the profit figure will have to
rise proportionately in order to achieve the 10% target return. Revaluation increases the
denominator in the ROE calculation, which requires the tariff increase to earn the 10%
return. Thus, the proposed revaluation may have helped to influence the Commission's
decision in FEA's favour.
The Commission subsequently approved the increase in electricity tariff rates in
November 200435and was implemented from 1 January 2005. The reduced depreciation
rates and the approved increase in electricity tariff rates helped to strengthen the
financial performance and position of the FEA. As matters transpired the FEA secured
concessionary funding from the Fiji National Provident Fund of $230 million supported
by a government guarantee. The proposal to revalue non-current assets was
consequently shelved.
Jones (1991) study shows that when organizations expect to receive assistance from the
government, they are likely to report reduced earnings. Similarly, when the FEA makes
its case for an increase in its price, it is expected that they would have also behaved in
the similar manner. Therefore, one would argue that reducing depreciation rates does not
help but contradicts the normal expectation. In order to reconcile this, one needs to
clearly understand FEA’s situation. Since 2004, the FEA’s main objective was to obtain
external funding, for which it has to report higher profits. Higher prices assist in
reporting more profits but government would not approve the increase when high profits
are reported. Thus, the FEA used other arguments (as discussed above) to get the
approval from the government.
35 The approved increases in electricity tariff rates were 3.30% in 2005, 3.20% in 2006 and 3.10% in 2007 for domestic customers. The Commission also approved increases in electricity tariff rates of 6.90% in 2005, 6.45% in 2006 and 5.70% in 2007 for commercial and industrial customers and the Maximum Demand tariff customers.
Chapter 6: Evidence of Earnings Management Practices in the FEA
112
Therefore, it seems that the reduction in depreciation rates may have been used to
manage earnings upwards as the FEA wanted to attract external financiers. The FEA’s
submission to the Commerce Commission further supports the argument as higher tariff
rates would increase its profitability. Not only this, the FEA used the revaluation
proposal to receive favourable response from the Commission.
6.4 Conclusion
From an in-depth analysis of the information obtained from the FEA's annual reports
and other relevant documents, it seems that there may be evidence of earnings
management practices in the FEA. Five possible instances were identified and these
related to the capitalisation of research and training expenses, revaluation of non-current
assets, classification of government grants, changes in capitalisation policies and
changes in depreciation rates. Table 6.10 summarises the evidence and the incentives for
its practice.
Table 6.10 Summary of Earnings Management Evidence
Year Event Incentive
1985 to 1992
Capitalisation of research and training costs
Avoid reporting large expense in a single year and to meet the debt covenant
1992 Revaluation of non-current assets Improve the equity position and meet the debt covenant set by the offshore lenders
Prior to 1995
Deferred income was reported as equity instead of liability
Improve the equity position and the debt to equity ratios
1994 and 1996
Ceased to capitalise certain overheads to property, plant and equipment
1999 Adopted higher tax depreciation rates
To make a case to the government to extend the tax exempt status and reimburse the cost of universal service obligations.
2004 Adopted lower tax depreciation rates To attract external financiers
Several incentives seem to be present during the times of these earnings management
practices. These include the need to improve the equity position of the Authority, to
meet the debt covenants, extend its tax exempt status, re-imbursement for the universal
service obligation costs and to attract potential investors.
Analyses of Earnings Management Practices in Fiji’s SOEs
113
CHAPTER SEVEN
CONCLUSIONS AND LIMITATIONS
7.1 Conclusions
The unexpected collapses of large corporations reflect erosion in the quality of financial
reporting. Earnings management is one of the contributing factors as it is used to
manipulate the true and fair view of the entity. The review of the related literature shows
that earnings management arises because of two broad reasons. Firstly, accounting
standards provide considerable discretion and flexibility to the preparers of financial
reports. Secondly, agency relationships provide economic incentives to the agents.
Considerable attention has been devoted to earnings management issues in the
developed countries, largely focusing on the publicly listed companies. Conversely,
studies in developing countries are rare. Therefore, the primary objective of this thesis is
to find evidence of earnings management in commercial statutory bodies of Fiji. Due to
the inapplicability of the widely used accruals models, the study applied the checklists
designed by Mulford and Comiskey (2002) to identify and analyse the possibilities of
earnings management practices in two SOEs over the period 1990 to 2004.
The findings of the study indicate a number of possible earnings management practices
in the organizations. In the HA, these were related to:
i. provisioning of inventory,
ii. capitalisation of administrative expenses,
iii. amortisation of deferred interest expense and
iv. changes in accounting policies for provision for doubtful debts.
The HA increased the provision for write-downs of inventory in the late 1980s and
transferred part of the resultant losses to the PRB. Higher provisions assisted HA to meet
the stakeholders’ expectations of reporting improved performance after the separation of
Chapter 7: Conclusions and Limitations
114
rental operations. Furthermore, by capitalising administrative expenses, the HA was able
to report increased profitability in the early 1990s. Higher reported profits reflected
better performance of the new government and the Board. Performance based contracts
were also implemented during this period, which are argued to have provided added
incentives to the management to report improved performance.
The HA recognised deferred interest expense in 1992, which was amortised over a ten-
year period. Its recognition is not questionable but its subsequent amortization seems to
be opportunistic as the trend in the amortisation expense suggests that it was used as a
smoothing device. This was particularly evident in 1993 when a large amount was
amortised to smooth the HA's interest expense. The final evidence relates to the changes
in accounting policy for provision for doubtful debts. There were two occasions over the
investigation period, when the policy was changed and in both cases the change was
justified on the grounds of prudence and conforming to industry standards. However, the
timing of the change in 1995 was questionable while in 1999 it was argued that the HA
may have exaggerated the effects of the policy change. The change in policies increased
the doubtful debts expenses, resulting in reduced earnings. It seems that the events that
took place in years surrounding the policy change, such as achieving its social objectives
and the need to convert debts into equity, may have led the HA to practice downwards
earnings management.
The analysis of the second entity, the FEA, revealed the following possible earnings
management practices:
i. capitalisation of research and training expenses,
ii. revaluation of non-current assets,
iii. classification of government grants,
iv. changes in capitalisation policies and
v. changes in depreciation rates.
The FEA capitalised research and development expenditure from 1985 to 1992, which
included costs incurred during the research phase. During the same period, it also
Analyses of Earnings Management Practices in Fiji’s SOEs
115
deferred training expenses. Capitalisation of these costs is not permitted by generally
accepted accounting practices, hence, it can be argued that FEA has possibly practiced
earnings management with the incentive to meet debt covenant set by its lenders. The
majority of these costs were expensed in 1987, as it proved to be an opportune time.
Furthermore, the FEA misclassified government grants as equity instead of liabilities in
years prior to 1995 and revalued its non-current assets in 1992. It seems that these events
were undertaken with similar intentions. As a result of the two devaluations, the FEA’s
net worth has been negative since 1987, which prevented it from meeting the 3:1 debt to
equity ratios imposed by the offshore financiers. Although, misclassification of grants
improved the equity position, it had insignificant influence on debt to equity ratios. The
subsequent revaluation further improved the equity balance and the ratios. Hence, the
two events improved the deteriorated equity balance and aligned debt to equity ratios to
a commercially acceptable level and those required by the lending bodies.
From 1994 and 1996, the FEA began to expense certain overheads that were previously
capitalised, resulting in higher operating expenses. Additionally, the FEA increased the
depreciation rates from 1999 by adopting the higher tax depreciation rates. These policy
changes are related and seem to be undertaken for similar reasons as they resulted in
reduced profitability. Lower reported earnings assisted the FEA to justify its case to the
government when requesting for an extension of their tax exempt status and
reimbursement of universal service obligation costs. The government subsequently
extended the FEA’s tax exempt status and universal service obligations costs were
acknowledged as the 'social' dividend. Finally, it was found that the FEA reduced the
depreciation rates from 2004 so that it could attract external finance for its capital
investment. This argument is further strengthened when the FEA made a submission to
the Commerce Commission for an increase in electricity tariff rates and used the
proposed revaluation to influence the Commission’s decision in their favour. The
increase in the tariff rates was subsequently approved by the Commission. Lower
depreciation rates and higher tariff rates strengthened the financial performance of the
FEA, enabling it to receive concessionary external funding.
Chapter 7: Conclusions and Limitations
116
The possible evidence of earnings management was more clearly observable in the FEA
than the HA. The results indicate that multiple earnings management practices were
undertaken for similar reasons at the FEA while the HA used a single event at a
particular time. The results illustrate that the key accounting areas used by two entities
for these practices were provisions, amortization and depreciation rates, capitalisation
policies, revaluations and classification of government grants. All these practices were
within the boundaries of generally accepted accounting practices except for the
classification of grants and capitalisations of research and training expenses. It should be
noted that the analysis only suggests that earnings management may be practiced in
these organizations. Overall, the findings are consistent with a priori expectation that in
the presence of agency relationships and flexibility in accounting standards, earnings
management practices will exist. These observations suggest that apart from the
presence of earnings management in publicly listed companies, it can also be practiced
in statutory organizations. However, these observations should be interpreted in light of
the following limitations.
7.2 Limitations of the Study
Earnings management practices are normally examined using variants of accruals
models, which require large samples of data. Accordingly, these models are confined to
the developed countries, where data is readily available. In small developing countries,
such as Fiji, the limited number of reporting entities does not permit the application of
the widely used statistical models. Therefore, the study had to pursue with a qualitative
case study method of investigation. This is the major drawback of this study. Moreover,
the entities examined are government owned, where the public is privy to certain
documents/data. If these were made available, it would have added more validity to the
findings of this study. Further, given the scope of the thesis, only two entities were
reviewed. Had more entities been included, the confidence in the obtained results would
be higher.
Analyses of Earnings Management Practices in Fiji’s SOEs
117
Even though, the qualitative approach is less powerful than the accruals models, it
certainly provides indications of earnings management. It should be noted that this is a
pioneering attempt and can be used as a guideline by other researchers who may explore
earnings management practices in countries where the accruals models are inapplicable.
If data permits, a promising avenue for future research would be to compare the results
obtained using the accruals models with the qualitative approach to ascertain the degree
of similarities in the results obtained by the two methods.
Bibliography
118
BIBLIOGRAPHY
Amat, O. and Gowthorpe, C. 2004. Creative accounting: nature, incidence and ethical issues.
Working paper, Spain.
Australian Accounting Research Foundation. 2001. Earnings management: audit considerations
& issues. Auditing & Assurance Alert 10, Auditing & Assurance Standards Board.
Barton, J. and Simko, P. The balance sheet as an earnings management constraint. The
Accounting Review. Vol. 77(supplement), pp. 1-27.
Bartov, E., Gul, F. and Tsui, J. 2000. Discretionary accruals models and audit qualifications.
Journal of Accounting and Economics. Vol. 30(3), pp. 421-452.
Becker, C., DeFond, M., Jiambalvo, J., and Subramanyam, K. 1998. The effect of audit quality
on earnings management. Contemporary Accounting Research. Vol. 15 (1), pp. 1- 24.
Behn, B., Nagy, A. and Riley, R. 2002. The association between stock/compensation mix and
earnings usefulness. Working paper, Tennessee.
Bernard, V. and Skinner, D. 1996. What motivates managers’ choice of discretionary accruals?
Journal of Accounting and Economics. Vol. 22(1-3), pp. 313-325.
Beneish, M. 2001. Earnings management: a perspective. Managerial Finance. Vol. 27(12), pp.
3-17.
Beneish, M. and Vargus, M. 2002. Insider trading, earnings quality and accrual mispricing. The
Accounting Review. Vol. 77(4), pp. 755-791.
Blake, J., Clarke J. and Gowthorpe C. 1998. The ethics of accounting regulation. In Gowthorpe
C., and Blake J. (eds.). Ethical Issues in Accounting. London.
Brewer, B., Angell, R. and Mautz, R. 2002. The proliferation of “special” accounting items: a
threat to corporate credibility. Southern Business Review. Vol. 28(1), pp.1-10.
Cahan, S. 1992. The effect of antitrust investigations on discretionary accruals: a refined test of
the political cost hypothesis. The Accounting Review. Vol. 67(1), pp. 77-95.
Analyses of Earnings Management Practices in Fiji’s SOEs
119
Chand, P. 2003. Relevance of International Accounting Standards to developing countries: the
case for Fiji. Unpublished MA thesis, The University of the South Pacific.
Chand, P. and White, M. 2006. The influence of culture on judgments of accountants in Fiji.
Australian Accounting Review. Forthcoming.
Commerce Commission. About us. www.commcomm.gov.fj/
Commerce Commission. 2004. Price Determination on the Fiji Electricity Authority's proposal
for tariff increase.
DeAngelo, L. 1986. Accounting numbers as market valuation substitutes: A study of
management buyouts of public stockholders. The Accounting Review. Vol. 61(3), pp.
400-420.
DeChow, P. and Skinner, D. 2000. Earnings management: reconciling the views of accounting
academics, practitioners, and regulators. Accounting Horizons. Vol. 14(2), pp. 235-250.
DeChow, P., Sloan, R. and Sweeney, A. 1995. Detecting earnings management. The Accounting
Review. Vol. 70(2), pp. 193-225.
DeFond, M. and Jiambalvo, J. 1994. Debt covenant violation and manipulation of accruals.
Journal of Accounting and Economics. Vol. 17(1-2), pp. 145-176.
Degeorge, F., Patel, J. and Zeckhauser, R. 1999. Earnings management to exceed thresholds.
Journal of Business. Vol. 72(1), pp. 1-33.
Dichev, I. and Skinner, D. 2002. Large-sample evidence on the debt covenant hypothesis.
Journal of Accounting Research. Vol. 40(4), pp. 1091-1123.
Fechner, H. and Kilgore, A. 1994. The influence of cultural factors on accounting practice. The
International Journal of Accounting. Vol. 29, pp. 265-277.
Fiji Electricity Authority. 1983-2004. Annual Reports.
Fiji Institute of Accountants. 1995. Fiji Accounting Standards. Suva, Fiji.
Fiji Institute of Accountants. 2001. Fiji Accounting Standards. Suva, Fiji.
Fiji Institute of Accountants. 2001. Framework for the preparation and presentation of financial
statements. Suva, Fiji.
Fiji National Provident Fund. 1993 and 1995. Annual Reports.
Bibliography
120
Frankel, R., Johnson, M. and Nelson, K. 2002. The relation between auditors fees for nonaudit
services and earnings management. The Accounting Review. Vol. 77 (supplement), pp.
71-105.
Fraser, L. and Ormiston, A. 2001. Understanding financial statements. Prentice Hall, Inc.; New
Jersey.
Gaver, J., Gaver, K. and Austin, J. 1995. Additional evidence on bonus plans and income
management. Journal of Accounting and Economics. Vol. 19(1), pp. 3-28.
Goodwin, J. and Howieson, B. 1999. In Leo, K. and Radford, J. (eds.) Financial Accounting
Issues. John Wiley & Sons, Australia.
Greer, S. and Patel, C. 2000. The issue of Australian indigenous world-views and accounting.
Accounting, Auditing and Accountability Journal. Vol. 13(3), pp. 307-329.
Grynberg, R., Munro, D. and White, M. 2002. Crisis: the collapse of the National Bank of Fiji.
University of the South Pacific Book Centre, Suva, Fiji.
Guay, W., Kothari, S. and Watts, R. 1996. A market-based evaluation of discretionary- accrual
models. Journal of Accounting Research. Vol. 34(supplement), pp. 83-105.
Guidry, F., Leone, A. and Rock, S., 1999. Earnings-based bonus plans and earnings management
by business-unit managers. Journal of Accounting and Economics. Vol. 26(1-3), pp.
113-143.
Healy, P. 1985. The effect of bonus schemes on accounting decisions. Journal of Accounting
and Economics. Vol. 7(1), pp. 85-107.
Healy, P. and Wahlen, J. 1999. A review of the earnings management literature and its
implications for standard setting. Accounting Horizons. Vol. 13(4), pp. 365-383.
Henderson, S. and Peirson, G. 2002. Issues in financial accounting. Pearson Education
Australia.
Heninger, W. 2001. The association between auditor litigation and abnormal accruals. The
Accounting Review. Vol. 76(1) pp. 111-126.
Housing Authority. 1985– 2004. Annual Reports
Analyses of Earnings Management Practices in Fiji’s SOEs
121
Jayaraman, T. and Sharma, R. 2003. Why is interest rate spread high in fiji? results from a
preliminary study. Fijian Studies: A Journal of Contemporary Fiji. Vol.1 (1), pp.75-104.
Jensen, M. and Meckling, W. 1976. Theory of the firm: managerial behaviour, agency costs and
ownership structure. Journal of Financial Economics. Vol. 3(4), pp. 305-360.
Jones, J. 1991. Earnings management during import relief investigations. Journal of Accounting
Research. Vol. 29(2), pp. 193-228.
Kasznik, R. 1999. On the association between voluntary disclosure and earnings management.
Journal of Accounting Research. Vol. 37(1), pp. 57-81.
Krishnan, G. 2003. Audit quality and the pricing of discretionary accruals. Auditing: A Journal
of Practice and Theory. Vol. 22(1), pp. 109-126.
Levitt, A. 1998. The numbers game. Unpublished remarks. Available at
http://www.SEC.gov/news/speech/spch220.txt.
Likerman, A. 1983. Evidence on accusations of manipulating profitability: adjustments for
inflation by nationalised industries 1976-81. Accounting and Business Research. Vol.
14(53), pp. 29-34.
Liu, Q. and Lu, Z. 2003. Earnings management to tunnel: evidence from China’s listed
companies. University of Hong Kong, Hong Kong.
McInnes, W. 1990a. A longitudinal study of accounting changes: The UK Gas Industry 1969-
1974. Accounting and Business Research. Vol. 20(80), pp. 315-327.
McInnes, W. 1990b. Further evidence on accounting choices: The South of Scotland Electricity
Board, 1978-1988. Accounting and Business Research. Vol. 20(81), pp. 57-66.
McNichols, M. and Wilson, G. 1988. Evidence of earnings management from the provision for
bad debts. Journal of Accounting Research. Vol. 26(supplement), pp. 1-31.
Ministry of Public Enterprises and Public Sector Reform. 2000-2004. Annual Reports.
Ministry for Commerce, Industry, Cooperatives and Public Enterprises. 1998. Towards a Better
Fiji: The Public Enterprise Reform Programme.
Miroshnik, V. 2002. Culture and international management: a review. The Journal of
Management Development. Vol. 21(7), pp. 521-544.
Bibliography
122
Mulford, C and Comiskey, E. 2002. The financial numbers game: detecting creative accounting
practices. John Wiley & Sons. Inc., USA.
Natarajan, R.1999. Earnings based compensation schemes and discretionary accrual and
expenditure decisions over CEO tenure. Managerial Finance. Vol. 25(9), pp. 89-118.
Nelson, M., Elliot, J. and Tarpley, R. 2002. Evidence from auditors about managers and auditors
earnings management decisions. The Accounting Review. Vol. 77(supplement), pp. 175-
202.
Palliam, R. and Shalhoub Z. 2003. The phenomenology of earnings management within the
confines of agency theory. International Journal of Value Based Management. Vol.
16(1), pp.75-88.
Parliament of Fiji. 2003 and 2005. Parliamentary Debates (Hansard).
Parliament of Fiji. 2005. Accounts and Finance for the year ended 31 December 2004.
Parliamentary Paper No. 80.
Pathik, K. 1999. The standard setting process in Fiji and the level of compliance. Unpublished
MA thesis, The University of the South Pacific, Fiji.
Perry, S. and William, T. 1994. Earnings management preceding management buyout offers.
Journal of Accounting and Economics. Vol. 18(2), pp. 157-179.
Phillips, J., Pincus, M. and Rego, S. 2003. Earnings management: new evidence based on
deferred tax expense. The Accounting Review. Vol. 78(2), pp. 491-521.
Rangan, S. 1998. Earnings management and the performance of seasoned equity offerings.
Journal of Financial Economics. Vol. 50(1), pp. 101-122.
Reserve Bank of Fiji. 1996. Guidelines for loan classification and provisioning for impaired
assets. Banking supervision policy statement number 3.
Reserve Bank of Fiji. 2001. Guidelines for loan classification and provisioning for impaired
assets. Banking supervision policy statement number 3(revised).
Reserve Bank of Fiji. 2002 and 2004. Annual Reports.
Analyses of Earnings Management Practices in Fiji’s SOEs
123
Robson, C. 2002. Real world research: a resource for social scientists and practitioner
researchers. Blackwell Publishers, UK.
Richardson, V. 2000. Information asymmetry and earnings management: some evidence. Review
of Quantitative Finance and Accounting. Vol. 15(4), pp. 325-347.
Schipper, K. 1989. Commentary on earnings management. Accounting Horizons. Vol. 3(4), pp.
91-102.
Sharma, U. 2000. Management accounting and control system changes: a case study of the
Housing Authority of Fiji. Unpublished MA thesis, The University of the South Pacific,
Fiji.
Sprouse, R. 1966. Accounting for what-you-may-call its. The Journal of Accountancy. Vol.
122(4), pp. 45-53. In Keller, T. and Zeff, S. (eds.) Financial Accounting Theory Part II:
Issues and Controversies. 1969. pg 367-379.
Stolowy, H. and Breton, G. 2000. A framework for the classification of accounts manipulations.
Working paper, HEC School of Management.
Sweeney, A. 1994. Debt covenant violations and managers’ accounting responses. Journal of
Accounting and Economics. Vol. 17(3), pp. 281-308.
Teoh, S., Welch, I. and Wong, T. 1998a. Earnings management and the long-run market
performance of initial public offerings. The Journal of Finance. Vol. 53(6), pp. 1935-
1974.
Teoh, S., Welch, I. and Wong, T., 1998b. Earnings management and the long run
underperformance of seasoned equity offerings. Journal of Financial Economics. Vol.
50(1), pp. 53–100.
The Government of Fiji. Public Enterprise Act, 1996.
Wise, M. 1996. HA close to objective. The Fiji Times. 127th year (288) Oct 16, pp.4.
United States of America. 2002. Sarbanes-Oxley Act.
Vinciguerra, B. and O’Rielly, M. 2004. An examination of factors influencing managers and
auditors assessments of the appropriateness of an accounting treatment and earnings
management intentions. American Business Review. Vol. 22(1), pp. 78-87.
Appendix
124
APPENDIX 1 CHECKLISTS
Checklist 1 Premature or Fictitious Revenue
What is the entity’s revenue recognition policy?
1. Before delivery or performance
a. Is it really earned?
2. At delivery or performance
a. Is there a right of return or price protection?
i. Has the entity provided adequately for returns or price adjustments?
3. After delivery or performance and full customer acceptance
Was there a change in the revenue recognition policy?
1. Did the change result in earlier revenue recognition?
Are there any unusual changes in revenue reported over the years?
1. What is the revenue for each year over the investigation period?
2. Does any one year show unusual activity not explained by economic factors?
Analyses of Earnings Management Practices in Fiji’s SOEs
125
Checklist 1 Continued
Does the entity have the physical capacity to generate the revenue reported?
1. What is revenue per appropriate measure of physical capacity for each of the last ten years?
a. Possible measures of revenue per physical capacity:
i. Revenue per employee
ii. Revenue per dollar of property, plant and equipment
iii. Revenue per dollar of total assets
Are there signs of overstated accounts receivable or other accounts that might be used to
offset premature or fictitious revenue?
1. Compare the percentage rate of change in accounts receivable, property, plant and equipment
and other assets with the percentage rate of change in revenue for each year.
a. What are the implications of differences in the rates of change in these accounts and revenue?
2. Consider whether unexplained changes in other asset or liability accounts might be explained
by premature or fictitious revenue.
3. Compute accounts receivable days for each year.
a. What are the implications of changes noted in accounts receivable days over the investigation
period?
Appendix
126
Checklist 2 Capitalization and Amortization Policies
PART 1 Detecting Aggressive Capitalization Policies
For cost capitalization generally:
1. What are the entity’s policies with respect to cost capitalization?
a. Is the entity capitalizing costs that should be expensed?
b. Does the entity expense more, taking a more conservative approach?
c. Are capitalized costs increasing faster then revenue over lengthy periods?
Has the entity shown evidence in the past of being aggressive in its capitalization policies?
Is there an example of prior year write down of capitalized costs that should not have been
capitalized?
Has a regulator forced a change in accounting policies in the past?
Has the entity capitalized costs in stealth?
Examine unusual changes in and relationships with revenue of the following: accounts
receivable, inventory, property, plant and equipment and other assets.
PART 2 Amortization Policies
Has the entity selected extended amortization and depreciation periods for capitalized
costs?
1. How does the calculated average amortization period for long-lived assets compare with
competitors or other entities in the industry?
Consider the extended amortization periods in the following situations:
1. Entity’s industry is experiencing price deflation
2. Entity is in an industry that is experiencing rapid technological change
3. Entity has shown evidence in the past of employing extended amortization periods
a. Is there an example of a prior year write down of assets that became value impaired?
Analyses of Earnings Management Practices in Fiji’s SOEs
127
Checklist 3 Misreported Assets and Liabilities
Detecting overvalued/undervalued assets
Accounts receivable
1. Compare the percentage rate of change in accounts receivable with the percentage rate of
change in revenue for each year.
a. What are the implications of differences in the rates of change?
2. Is the allowance for doubtful accounts sufficient to cover future collection problems?
a. Compute accounts receivable days for each year.
i. Is the trend steady, improving or worsening?
3. Have economic conditions for the entity’s customers worsened recently?
a. Are entity’s sales declining?
b. Are there other general economic reasons to expect that customers are or may be having
difficulties?
4. Are sales growing rapidly?
a. Has the entity changed its credit policy?
i. Is credit being granted to less creditworthy customers?
b. Have payment terms been extended?
Inventory
1. Are inventories overstated due to inclusion of nonexistent inventories or by the reporting of
true quantities on hand at amounts that exceed replacement cost?
a. Compute gross margin and inventory days for the last ten years
i. Is the trend steady, worsening or improving?
b. Do ongoing entity events and fortunes suggest problems with slackening demand for the
entity’s products?
i. Are sales declining?
ii. Have raw material inventories declined markedly as a percentage of total inventories?
Appendix
128
Checklist 3 continued
c. Are prices falling, suggesting general industry weakness and an increased chance that
inventory cost may not be recoverable?
d. Is the entity in an industry that is experiencing rapid technological change, increasing the risk
of inventory obsolescence?
e. Has the entity shown evidence in the past of inventory overvaluation/undervaluation?
i. Is there an example of a prior year write-down of inventory that became value impaired?
Detecting overvalued/undervalued liabilities
Accrued expenses payable
1. What is the trend in accrued expenses payable?
2. Compare the percentage rate of change in accrued expenses payable with the percentage rate
of change in revenue for each year.
a. What are the implications of differences in the rates of change?
3. Does an improvement in selling, general and administrative expense as a percentage of
revenue reflect true operating efficiencies?
Accounts payable
1. Compute accounts payable days for each year.
a. Is the trend steady, worsening or improving?
2. Was there an unexpected improvement/decline in gross profit margin?
3. How does the percent change in accounts payable compare with the percent change in
inventory?
Analyses of Earnings Management Practices in Fiji’s SOEs
129
Checklist 4 Using Cash Flows To Detect Earnings Management Practices
Isolate nonrecurring cash inflows and outflows, including:
1. Income taxes paid or recovered on transactions classified as investing or financing activities,
including:
a. gain or loss on sale of assets, investments or businesses
b. extraordinary items
c. changes in accounting principle, if any
2. Cash flow from the purchase and sale of trading securities
3. Capitalized expenditures that should be expensed as incurred
4. Nonrecurring cash income and expense
5. Significant isolated events leading to changes in operations-related assets and liabilities,
including:
a. special inventory reduction sale outside normal channels
Compute adjusted cash flow provided by continuing operations
1. Adjust reported cash flow provided by operating activities for identified nonrecurring cash
flow items.
Compute adjusted income from continuing operations
1. Adjust reported income from continuing operations for nonrecurring items of income and
expense
Compute the adjusted cash flow-to-income ratio
1. Adjusted cash flow provided by operating activities divided by adjusted income from
continuing operations.
a. Compute for several years
b. Examine results for discernible trend.
Appendix
130
APPENDIX 2
Current Interest Rates Offered by Different Financial Institutions
Institution Rate36
Home Finance Company 8.65% fixed for first 6 months, 10% variable thereafter
Westpac Banking Corporation 8.4% fixed for 1 year, 10.25% variable thereafter
Colonial National Bank 9.40% fixed for 1 year, 10.25% variable thereafter
ANZ 8.25% fixed for 1 year, 10.25% variable thereafter
Fiji Development Bank Variable rate 8.95%
Housing Authority 3.99% fixed for 18 months, 7.99% variable thereafter
(Adapted from the financial institutions)
36 These rates are effective August 2006 after the RBF increased its policy indicator rate in June from 3.25% to 4.25%. However, HA's rates remained unchanged.
Analyses of Earnings Management Practices in Fiji’s SOEs
131
APPENDIX 3A
Financial Indicators of the HA
Year Assets
$m
Equity
$m
Liabilities
$m
Operating
Income37
$m
Debt to
equity
Current
ratio
ROE
%
1990 85.29 1.70 83.59 (2.22) 49.11 1.93 (130.20)
1991 86.55 0.80 85.75 (0.48) 106.65 1.44 (59.58)
1992 107.70 2.05 105.64 0.04 51.48 1.73 2.10
1993 127.98 4.31 123.67 0.52 28.69 2.07 12.02
1994 153.74 5.82 147.92 0.69 25.42 2.15 11.81
1995 167.32 3.43 163.90 0.31 47.84 3.35 9.02
1996 168.92 (2.22) 171.14 (1.07) (77.13) 1.59 47.99
1997 166.55 (2.10) 168.65 0.12 (80.16) 1.08 (5.47)
1998 168.13 (1.94) 170.08 (0.77) (87.49) 0.61 39.40
1999 158.06 (5.34) 163.40 (1.42) (30.58) 1.02 26.50
2000 161.77 (4.36) 166.14 (1.96) (38.07) 0.83 44.84
2001 165.87 (3.85) 169.71 0.45 (44.13) 0.65 (11.80)
2002 176.05 38.46 137.58 0.54 3.58 1.08 1.40
2003 177.62 39.24 138.39 0.77 3.53 1.07 1.97
2004 158.70 41.30 117.39 2.07 2.84 1.23 5.00
(Adapted from the HA Annual Reports 1989-2004 and author's calculations)
37 This represents income before abnormal items.
Appendix
132
APPENDIX 3B
Financial Indicators of the FEA
Year Assets
$m
Equity
$m
Liabilities
$m
Operating
Income38
$m
Debt to
Equity
Ratio
Current
Ratio
ROE
%
1990 337.39 (3.93) 341.32 (3.38) (86.92) 0.23 85.97
1991 342.19 13.41 328.78 13.84 24.52 0.32 103.25
1992 509.89 204.37 305.52 17.29 1.49 0.31 8.46
1993 515.23 224.68 290.56 18.19 1.29 0.34 8.10
1994 509.53 242.00 267.53 16.38 1.11 0.85 6.77
1995 522.51 240.87 281.64 20.47 1.17 1.56 8.50
1996 525.96 259.61 266.35 17.32 1.03 1.33 6.67
1997 500.47 277.61 222.85 18.41 0.80 0.45 6.63
1998 497.61 299.78 197.83 21.26 0.66 0.55 7.09
1999 482.08 318.17 163.91 9.03 0.52 1.61 2.84
2000 473.28 315.91 157.37 (6.65) 0.50 0.33 (2.10)
2001 470.07 319.46 150.61 0.03 0.47 0.99 0.01
2002 456.70 324.91 131.79 3.79 0.41 2.21 1.17
2003 459.92 330.14 129.78 2.61 0.39 3.47 0.79
2004 452.12 325.38 126.74 9.31 0.39 1.89 2.86
(Adapted from the FEA Annual Reports 1990-2004 and author's calculation)
38 This represents income before abnormal items and income tax.