Paper Session 3 Fraudulent Financial Reporting in China ...Fraudulent Financial Reporting in China:...
Transcript of Paper Session 3 Fraudulent Financial Reporting in China ...Fraudulent Financial Reporting in China:...
4th Symposium of China Journal of Accounting Research (CJAR) 中国会计学刊研讨会
17 – 18 December 2010
Sponsored by:
Paper Session 3
Fraudulent Financial Reporting in China: Consideration of Timing Traits and
Corporate Governance Mechanisms
By
Dan Yang University of Aberdeen
& Roger Buckland
University of Aberdeen
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate
Governance Mechanisms
Dan Yang1, Roger Buckland2
Department of Accounting, Business School, University of Aberdeen, UK
ABSTRACT: This paper develops an analysis of the prevalence and determinants of fraudulent
financial reporting as identified in the Chinese listed firms over the period 1996 to 2007, and
highlights the relationship of financial fraud and corporate governance mechanisms. 82 cases of fraud
identified by the China Securities Regulatory Commission are selected as the study sample, and 82
control peers are designed to correspond to the study sample as closely as possible, regarding the
assets size and industries, in order to exclude the influence of these factors towards the dependent
variable. The findings of this paper challenge the conventional arguments which have been testified
based on the data from the western countries, mostly from the U.S.. Conventional arguments show
financial fraud is associated with weakness of governance. However, the finding in this paper reveals
that as for the high discussed corporate governance characteristics (e.g. independent directors, the
supervisory board, and audit committee), the fraud firms and their non-fraud peers are not statistically
distinct. Statistical differences of external auditor traits and regulation from the securities market are
found between the groups. The negative results of this paper contribute by updating our understanding
of the supervision of China’s equity markets, whether it can be considered effective in uncovering
financial fraud; and also the results may contribute that governance protections that work in one
system may fail in the other.
Keywords: fraudulent financial reporting; corporate governance; audit committee; independent
directors; the supervisory board; regulation.
1 Dan Yang: a PhD candidate in the Business School, University of Aberdeen, AB24 3QY, United Kingdom. Email: [email protected] 2 Roger Buckland: full Professor, Chair of Accountancy and Finance, Business School, University of Aberdeen, AB24 3QY, United Kingdom. Email: [email protected]
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
1
1. INTRODUCTION
More and more accounting scandals and fraud are damaging the accounting profession. Unethical
practices by corporations are making financial statements unreliable and can drag the economy into a
recession. Financial statement fraud has received considerable attention from investors, creditors,
regulators and the public, with high profile reported financial fraud at large companies such as Enron,
WorldCom (both U.S.), Parmalat (Italy) and Royal Ahold (the Netherlands). When compared to the
U.S. and Europe, financial statement fraud in the Chinese listed companies has been even worse.
From the establishment of China security market in 1992 to 2005, 45 listed companies had been
punished for fraudulent behaviours in financial reports by the China Security Regulatory Committee
(CSRC) (Hu, 2006). A series of fraudulent financial reports created by so-called blue chip companies,
such as Yin Guangxia, Lan Tian, and Daqing Hongguang, result in unheard-of credit crisis in the
Chinese stock market. The pervasiveness of reported financial frauds and related alleged corporate
governance failures has eroded investor confidence all over the world.
Prior research (e.g. Beasley, 1996; Dechow et al., 1996; 1999 COSO Report; Beasley et al., 2000)
provides insight into financial fraud instances in the U.S. market, and argues that financial fraud is
associated with weakness of governance. In the Chinese market, one of the largest emerging
economies in the world, corporate governance has also been a hot topic in identifying fraudulent
financial reporting in the recent years. The Chinese corporate governance structure has its unique
features, for example, American independent directors system and German supervisory board system,
as the main traits of the Anglo-American and European Continental corporate governance structures
respectively, are both required to establish in the Chinese firms. It inosculates the characteristics of
both U.S. and Europe, and is reasonably assumed to be special comparing to the experience of the U.S.
financial markets and valuable to do research. This research seeks to develop an analysis of the
prevalence and determinants of financial statement fraud as reported in the Chinese listed firms over
the period 1996 to 2007, by matching to a control sample of equities from companies where financial
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
2
fraud has not been detected. This evidence is important because an understanding of the occurrence
of financial fraud and its relationship with corporate governance in the Chinese market are relevant
to investors, creditors and regulators in that system, and also provide more findings in developing
market to supplement the global academic research on this issue, which present research cannot
address.
From data reported by the China Securities Regulatory Commission, 82 cases of fraud, occurred
during the period from 1996 to 2007, are selected as the study sample. While 82 control peers are
also designed to correspond to the assets size and industries of the study sample as closely as possible.
First, in this paper, the nature and evolution of reported fraud during this 12-year period are
investigated and documented, to show how the nature of fraud has evolved over time. Then this paper
examines the argument that companies are more, or less, prone to fraudulent reporting by reason of
their corporate governance characteristics. It is well known (Fama and Jensen, 1983; Loebbecke et al.,
1989; Bell et al., 1991) that governance structures and histories may impact upon the opportunities for,
incentives towards and the control or identification of fraud within organisations.
The findings of this paper challenge the conventional arguments which have been testified based on
the data from the western countries, mostly from the U.S.. Conventional arguments show financial
fraud is associated with weakness of governance in western companies (e.g. Beasley et al., 2000).
However, the finding in this research reveals that as for some high discussed corporate governance
characteristics (e.g. the supervisory board, audit committee, independent directors), the fraud firms
and their non-fraud peers are not statistically distinct, suggesting that some corporate governance
mechanisms that are designed to reduce the probability of financial fraud fail to work in the Chinese
market.
The paper makes contribution to both academic researchers and policy makers. It contributes most
that regulations and governance protections that work in one system may fail in the other, and
different gatekeepers need to be designed into different governance systems to monitor for frauds.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
3
The results also have the advantage of being able to test the hypotheses about the corporate
governance failures in identifying financial fraud in the developing markets.
The paper is organized as follows. In section 2, a brief review of the relevant literature on financial
statement fraud is provided. Section 3 outlines research design and data issues. Section 4 contains the
description of timing traits of the research sample. Section 5 presents the empirical results. Section 6
draws this paper into a conclusion and provides the limitations.
2. LITERATURE REVIEW
Much of prior research of fraudulent financial reporting develops arguments or provides descriptive
information about financial and nonfinancial characteristics of companies experiencing financial fraud.
“Who” engage in financial statement fraud? Rezaee (2002) states that most of, if not all, financial
statement frauds occur with encouragement, participation, approval, and knowledge of top executives,
including CEOs, CFOs, presidents and treasurers. In the 1999 COSO reports, it states that top
executives are involved in most of the fraud cases, that is, CEO and/or CFO are named in 83 percent
of the cases, while Controller and Chief Operating Officer are engaged in 28 percent.
“What” are used in financial statement fraud? The 1999 COSO report finds that nearly 90 percent of
financial statement fraud involved falsification, alteration and manipulation of reported accounting
information, while only about 10 percent involved misappropriation of assets; and overstating profits
by recording fictitiously or prematurely was used in more than half of the fraud companies. Rezaee
(2002) also considers that misstating revenue is the most common way in financial statement fraud. In
China, some other research also focused on aggressive use of choices of accounting policies and
accounting estimate methods (e.g., Wang and Yang, 2006).
“Why” does financial statement fraud occur? Financial statement fraud are motivated by many
aspects and committed for a variety of reasons. Robertson (2000) identifies some reasons as primary
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
4
motives of financial statement fraud, including meeting company goals and objectives, obtaining new
financing or more favourable terms on existing financing. Lys and Watts (1994), and Carcello et al.
(2000) also investigate financial pressures according to poor financial performance and weak financial
condition as an incentive mechanism.
“Where” and “when” is financial statement fraud likely to happen? O’Brien (1988) finds a
significant positive relationship between analyst following, firm size, and institutional ownership.
Latham and Jacobs (2000) conclude that the extent of monitoring, either direct (corporate governance)
or indirect (analysts), can create an environment that permits no error, irregularities, and fraud. In
China, Huang (2005) and Chen (2007) find that avoiding regulation (e.g., ST and PT) is the primary
pressure, and fraud opportunities mainly come from the higher top 10 shareholders’ ownership
concentration, fewer times of directorate meetings, fewer percentage of shares owned by directorate
members.
That research of financial fraud provides much of the foundation for the inclusion of many of the
fraud risk factors in regulation. Then several studies (e.g. Beasley, 1996; Dechow et al., 1996;
McMullen, 1996; Beasley et al., 2000) move research forward to “Does corporate governance
matter?” An association between weaknesses in certain corporate governance mechanisms and
fraudulent financial reporting have been documented. For example, Beasley et al. (2000) provides
insight into financial statement fraud instances investigated during the late 1980s through the 1990s
within three volatile industries, and highlights important corporate governance differences between
fraud companies and no-fraud benchmarks. However those studies focus on only one or a few internal
corporate governance mechanisms. Also many studies in this area in China (e.g., Huang, 2005; Hu,
2006) lie in investigating the influence of ownership structure and board governance mechanisms on
financial fraud. Most of other relevant research (e.g., Liu and Du, 2003; Wang and Li, 2004) in fraud
in China has attributed financial fraud to auditing failures and theoretically argues that external
auditors act the first role in deterring financial statement fraud. The empirical research of financial
fraud and its relationship with corporate governance mechanisms, especially some high discussed
mechanisms, is rare. Thus, this paper is motivated and warranted.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
5
3. RESEARCH DESIGN AND DATA ISSUES
The Chinese corporate governance structure has unique features, for example, American independent
directors system and German supervisory board system, are both required to establish in the Chinese
firms. In this paper, we focus on some high discussed corporate governance mechanisms, including
independent directors, the supervisory board, audit committee, external auditors and regulation rules
in the securities market. It aims to examine whether these mechanisms work as well in identifying
financial fraud in the Chinese market.
3.1 Research Hypotheses
3.1.1 The Relationship of Board of Directors and Financial Statement Fraud
The situation that the board is controlled by top management team is the combination of chairman of
the board and CEO. When the chairman also serves as CEO, the board may be absolutely dominated
by insiders. It leads to weakness of the board’s supervisory independence, and increases the
probability for managers to engage in earnings manipulation and misappropriation of external
investors’ benefits (Jensen 1993). Thereby, it can be expected that the firms with the combination of
chairman of the board and CEO are more likely to perpetrate financial fraud or financial statements
prettification, and may be more inclined to conceal bad accounting information.
H1.1: The situation that the board is controlled by top management team and the occurrence of
financial statement fraud are positively related
Independent director is a member of the board of directors who is differentiated from inside director,
related director and gray director (Daniel, 2002), and they do not form part of the executive
management team or are not involved in the day-to-day running of business. In 23rd August 2001, the
Chinese Securities Regulatory Commission issued Introduction on Development of Independent
Director Mechanism in the Chinese Listed Companies, stating that until 30th June 2002, the board of
directors in the listed companies should include at least two independent directors; until 30th June
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
6
2003, one third of the directorate members should be independent directors, and one of them should
be accounting professionals. Prior research (e.g. Leftwich et al. 1981, Fama and Jesen 1983) argue
that high proportion of independent directors in the board could availably monitor the opportunism
existing in top management. Many Chinese scholars, such as Lu and Wang (2004), also argue that the
independence of the board is improved through independent director mechanism in preventing the
majority shareholders and managers illimitably expand their self-awareness. And we expect that the
listed companies with greater independent directors are less likely to engage in financial statement
fraud.
H1.2: The percentage of independent directors in the board is lower in fraud firms compared to
matched non-fraud firms
3.1.2 The Relationship of the Supervisory Board and Financial Statement Fraud
As for the members of the supervisory board, an effective shareholding contract is one important way
to harmonize their interest with that of shareholders, which can inspirit them to behave towards the
shareholders’ profit maximization at low cost. The greater shares they hold, the stronger motivations
they are given to monitor the actions of top management and also the higher efficiency the board
operates. Therefore, according to the general governance theory, we can expect that high proportion
of shares held by the supervisors improves the disclosure of accounting information and its quality in
the listed companies.
H2: The firms committing financial statement fraud are more likely to have a smaller proportion of
shares owned by supervisors
3.1.3 The Relationship of Audit Committee and Financial Statement Fraud
The western accounting literature, especially empirical studies, cites evidences proving the potential
for a vigilant audit committee as one important mechanism of corporate governance. Previous
research (Beasley 1996, Dechow et al. 1996, McMullen 1996) indicates that firms engaging in
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
7
financial statement fraud are more apt to have ineffective audit committees which meet infrequently
or have no audit committee. DeFond and Jiambalvo (1994) find that non-fraud firms have more active
and effective audit committee than fraud firms. Several reports, such as the Treadway Commission,
the NYSE and NASD, have addressed the role of audit committee in the areas of corporate
governance and Blue Ribbon Committee (BRC) issued ten recommendations and five guiding
principles aimed at enhancing audit committee members’ independence and qualification to ensure a
reliable financial reporting process. The role and function of audit committee have evolved over the
years in the western countries and now its future is expected to be the ultimate guardians of investors’
interest and accountability. While in China, audit committee is still a new concept both in literature
and in the practical operations of firms. Although the Chinese Securities Regulatory Commission
(CSRC) announced that the Chinese listed companies should set up audit committee and issued its
functions in 2002, not all the Chinese listed companies have established audit committee until now.
Most of the Chinese research (e.g. Zhang 2007, Wen 2007, Ye and Ban 2008) focus on theoretically
arguing the functions, significance and problems of audit committee. Based on analysis of the role and
function of audit committee, we expect that in the Chinese stock market, firms with audit committee
are less likely to perpetrate financial statement fraud.
H3: The firms experiencing financial statement fraud are less likely to have an audit committee
than non-fraud firms
3.1.4 The Relationship of External Auditors and Financial Statement Fraud
Users of disclosed accounting information have held external auditors responsible for assuring
financial reports free of material misstatements caused by error or fraud and detecting financial
statement fraud. Auditor’s independence is the cornerstone of auditing quality. The literature has long
been concerned that change and tenure of accounting firms potentially affect audit quality, but
contains conflicting arguments. DeAngelo (1981) and Watts and Zimmerman (1990) consider that as
the auditor-client relationship lengthens, auditors may easily develop personal friendships and
economic dependency on audit clients which impairs audit independence and auditing quality (Mautz
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
8
and Sharaf, 1961; AICPA, 1997; SEC, 2002); and Davis et al. (2003) find a positive relationship
between audit firm tenure and discretionary accruals. While other researchers hold that information
asymmetry between auditors and clients reduces over time as client-specific knowledge is needed in
auditing. For the reason that client-specific knowledge can help detecting material misstatements in
financial statements, the lack of the knowledge in the beginning of audit engagement may lead to
lower auditing quality (Beck et al. 1988; Knapp, 1991; Solomon et al. 1999; Geiger and
Raghunandan, 2002). In China, few scholars study if the duration of the relationship between auditors
and clients is related with the occurrence of financial statement fraud, and most of them focus on
emphasizing the importance of auditor’s independence (e.g. Yu and Li, 2003). In this paper, we expect
a positive relationship between change of accounting firms and the occurrence of financial statement
fraud, and a negative relation between tenure of accounting firms and the likelihood of financial
statement fraud.
H4.1: Change of accounting firms and the occurrence of financial statement fraud are positively
related
H4.2: Tenure of accounting firms and the occurrence of financial statement fraud are negatively
related
As for the influence of accounting firm size on auditing quality, no conclusive and supportive
evidence indicates that the larger accounting firms (e.g., Big Four now, previous Big Eight, Big Six,
or Big Five) can provide more effective audit service and do better in detecting financial statement
fraud. For example, Dechow et al. (1996) find no significant relationship between a non-Big Four
audit firm and financial fraud reporting firms, whereas McMullen (1996) indicates a positive
relationship. In China, research on this issue contains different arguments as well. Hu (2002) shows
that Big Four accounting firms have reputation premium in the Chinese capital market from the aspect
of IPO. While Liu et al. (2003) and Chen and Xia (2006) do not get those supportive conclusions but
even opposite results. As expected the Big Four accounting firms have more professional personnel
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
9
and circumstance, we assume here that larger accounting firms can result in higher auditing quality
and lower possibility of financial statement fraud.
H4.3: Auditing quality and the occurrence of financial statement fraud are negatively related
3.1.5 The Relationship of Regulation Rules and Financial Statement Fraud
One motivation for perpetration of financial statement fraud is associated with the need to hide loss to
prevent ST (Special treatment) or PT (Particular transfer) in the capital market. In terms of Company
Law and Qualifications of Stock being Listed, it can be found that a corporation is much apt to be
treated specially, including ST, suspending being listed or being de-listed, by the Securities
Regulation Sector when it reports continuous loss. When the company is specially treated, on one
hand, the creditors may deem its solvency decreasing, and in order to protect their own benefits, they
may press on with dunning the debt or requiring the company to pay back in advance. On the other
hand, it may increase difficulty for the company in getting loans from banks. Therefore, both debt
financing and equity financing are influenced or even become impossible, and the company may get
into serious financial difficulties. We expect that the listed companies may attempt to utilize
diversified earnings management ways, even fraudulent methods, to make up deficits or prevent
reporting loss.
H5.1: Avoiding regulation in securities market (e.g. ST) and the occurrence of financial statement
fraud are positively related
H5.2: Time length of being listed and the occurrence of financial statement fraud are negatively
related
3.2 Research Variables
The explained variable is whether the listed company engages in financial statement fraud, and the
value can be set to 1 for the fraud firms, while 0 for the non-fraud firms. The explanatory variables are
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
10
as indicated in Table 1. Eight control variables are also involved which relate to motivational and
conditional factors identified in Loebbecke et al. (1989) and Beasley (1996), and most of them are
known to affect the possibility of the occurrence of financial statement fraud in the literature (Simpson,
1986; Baucus and Near, 1991; Bell et al., 1991). The control variables are indicated in Table 2.
Table 1 The explanatory variables of corporate governance
Item Variables Description and measurement of variables
BD1 A dummy variable with a value of one when the chairman of the board
serves as CEO or president and a value of zero otherwise Board of
directors BD2 The proportion of independent directors in the board
SB1 Size of the supervisory board Supervisory
board SB2 The percentage of shares owned by the supervisory board members
Audit
committee AC
A dummy variable with a value of one if the firm has an audit committee
in the year of financial statement fraud and a value of zero otherwise
EA1 A dummy variable with a value of one when accounting firms are
changed in the fraud year and a value of zero otherwise
EA2 The tenure of accounting firms External
auditors
EA3 The authority of accounting firms with a value of one when the
accounting firm is the Big Four and a value of zero otherwise
RG1 Special treated in fraud-occurring year, with a value of one when a firm
is signed ST and a value of zero otherwise
RG2 Special treated in fraud-reporting year, with a value of one when a firm
is signed ST and a value of zero otherwise
Regulators
RG3 Time length of a firm being listed in public markets
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
11
Table 2 The control variables
Variables Description and measurement of variables
IC Industry culture, with a value of one for industry with high competition and
scare resources, such as manufacturing, and a value of zero otherwise
CS Corporate size which is the natural logarithm of total assets of a firm
H Fraud history of a firm, with a value of one when financial statement fraud
occurs in the previous year and a value of zero otherwise
AO Auditing opinion, with a value of one when it is perfect opinion and a value of
zero otherwise
LEVER Financial leverage ratio which is the ratio of liabilities and assets
ROE Return on equity of a firm
ST Special treated, with a value of one when a firm is signed ST and a value of zero
otherwise
RSG Rate of sales growth
3.3 Research Methods
3.3.1 Variance Analysis
Variance analysis is a collection of statistical sample and their associated procedures, in which the
observed variance is partitioned into components due to different explanatory variables. Variance
analysis is employed in this study to provide insight into corporate governance differences between
study sample and control sample benchmarks on an industry basis, and the formula can be seen as
followed:
CGij=ui+eij (F 1)
Where
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
12
CGij= the corporate governance variable of firm j in group i;
ui= the average value of corporate governance variable in group i;
eij= the residual of firm j in group i.
The eij above describes the differences between the corporate governance variable of individual firms
and the average level of that group, and it is assumed normally distributed. Further, Formula 4.2 is
usually transformed to Formula 4.3 for statistical purpose:
CGij=u+ai+eij (F 2)
Where
u= the general average value of corporate governance variable without considering fraud or not;
ai= additional effects when the sample are fraud firms or non-fraud firms.
When corporate governance mechanisms has no relationship with financial statement fraud, there are
no additional effects of the two samples, and a1=a2=0. Therefore, in order to test the role of corporate
governance mechanisms in the commission of fraud, the following assumptions should be tested:
H0: a1=a2=0; H1: at least a1 or a2 is not 0
3.3.2 Factor Analysis
Factor analysis is a statistical method used to describe variability among the observed variables in
terms of a potentially lower number of unobserved variables called factors. One advantage of Factor
Analysis is reduction of number of variables, by combining two or more variables into a single factor.
Factor Analysis is performed in this study to avoid the probable multicollinearity phenomenon
resulting from highly correlated relationship among the research variables.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
13
3.3.3 Logistic Regression
Logistic regression is involved in the research design of this paper, because financial statement fraud,
as the dependent variable, is dichotomous. The study sample and one-to-one matched control sample
are analyzed in the estimation, although we cannot assume that the rate of companies that are alleged
to have experienced financial statement fraud within the total population of the listed companies is 50
percent, and this approach in the study is different from a pure random sampling approach. Logistic
regression is employed in empirical analysis of each corporate governance factors to test the
hypothesized relationship between the occurrence of financial statement fraud and corporate
governance mechanisms. The logistic regression model is developed as followed:
FSFi= α+ β1CGi + β2ICi+ β3CSi+ β4Hi+ β5AOi+ β6LEVERi+ β7ROEi+ β8STi+ β9RSGi+ εi
(F 3)
Where
FSF- a dummy variable with a value of one when the firm is disclosed to have engaged in
financial statement fraud and a value of zero otherwise;
CG- governance characteristics of corporate governance parties;
IC- industry culture, with a value of one for industry with high competition and scare
resources, such as manufacturing, and a value of zero otherwise;
CS- corporate size which is the natural logarithm of total assets of a firm;
H- fraud history of a firm, with a value of one when financial statement fraud occurs in the
previous year and a value of zero otherwise;
AO- auditing opinion, with a value of one when it is perfect opinion and a value of zero
otherwise;
LEVER- financial leverage ratio which is the ratio of liabilities and assets;
ROE- return on equity of a firm;
ST- special treated, with a value of one when a firm is signed ST and a value of zero
otherwise, which is used to measure if the firm is experiencing financial trouble;
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
14
RSG- rate of sales growth;
i- firm 1 through 164;
ε- the residual.
3.4 Data Collection
Study sample and control sample are mainly chosen from the China Securities Regulatory
Commission (http://www.csrc.gov.cn), Shanghai Stock Exchange website (http://www.sse.com.cn)
and Shenzhen Stock Exchange website (http://www.szse.cn), and other helpful websites that can
provide relevant financial information of the listed companies are http://www.homeway.com.cn,
http://www.cnlist.com and http://www.cninfo.com.cn.
Through carefully reading the punitive notices disclosed by the China Securities Regulatory
Commission from 2002 to 2009, 82 instances of financial statement fraud alleged by the CSRC and
occurred during the 12-year period between 1996 and 2007 are chosen as the study sample1. The
fraudulent financial reports without influence on profits and assets are not involved for lack of fraud
signals on financial data. The periodic punitive notices by the CSRC, which contain summaries of
violation actions of the listed companies, represent one of the most important sources of alleged cases
of fraudulent financial reporting in China and provide recent relevant information. The financial
statement fraud involves violations of Article 177 of Securities Law which is “disobeying the relevant
rules to disclose financial information, or disclosing sham, misguided or highly incomplete
information”; violations of Article 74(2) of Provisional Regulations on the Administration of Share
Issuance and Trading that is “providing fallacious or misleading statements, or omitting momentous
financial information during the process of share issuance and trading”; and violations of Article 12 of
Provisional Regulations on Prohibiting Securities Fraud.
1 The listed companies in Financials industry are not included in the study sample because of different accounting policy and regulation.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
15
The control sample is also chosen those are designed to correspond to the assets size and industries of
the study sample as closely as possible. Each fraud firm is matched with a non-fraud firm according to
four requirements, which are Stock Exchange, industry, firm size and time period. (Beasley, 1996)
The benchmark industry category is according to industry regulations of Shanghai Stock Exchange
and Shenzhen Stock Exchange. Prior studies on financial fraud consider the absence of non-fraud
control sample as a limitation and have provided empirical evidence of differences in governance
characteristics between fraud firms and non-fraud firms (Beasley, 1996; McMullen, 1996; Beasley et
al., 2000). However, potential possibility of bias exists in choosing the control sample, which is
potential cases of financial statement fraud might be included in the control sample. This represents a
limitation of this paper and we expect that the listed companies that are not in the punitive notices of
the CSRC are less likely to engage in fraudulent activities. In addition, in order to minimize the
likelihood of some misclassification when a company identified as control sample has an occurrence
of financial statement fraud that has not yet been detected, its financial indices are checked. Those
with abnormal indices (such as unusually high level of cash, abnormal assets/debt ratio) are not
involved.
4. TIMING TRAITS OF THE RESEARCH SAMPLES
Table 3 shows that financial statement fraud occurs in each year during the 12-year period between
1996 and 2007, therein, most of the fraudulent financial reports are produced from 2000 to 2004,
which represents 68.13% of the total. The engagements of companies in financial statement fraud are
mostly between 2000 and 2004 as well, which represents 80% of the total. With the development of
the Chinese capital market, since 2000, the securities market has come into an active period. Many
companies consider stock market as “capital resource without cost” and the listing qualification as a
rare resource. Rapid development of internet makes the listed companies pay more attention to their
financial information disclosure (Wang, 2001). Thus, because of the existence of agency problem and
information asymmetry, and imperfect regulation system of the Chinese securities market, many listed
companies are inclined to meet their needs through financial statement fraud. From Figure 1, it can be
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
16
found that before 2004, the occurrence of financial statement fraud has a steady increasing trend, but
decreases suddenly after 2004. One probable reason is that, with increasing disclosure of fraudulent
financial reporting, more pressure from the stakeholders of the listed companies, such as shareholders,
investors and creditors, have been put on the companies and external auditors to get transparent and
accurate accounting information. The other probable reason is, it takes a long time from investigating
financial fraud to disclosing the punitive notices, not all financial statement frauds in the recent three
years have been detected by the China Securities Regulatory Commission yet, and many listed
companies have been put on record to investigate fraudulent activities by the commission but haven’t
got the results.
Based on above analysis, we have a position that we argue that: 1) financial statement fraud occurs
and varies according to pressure and opportunities for misrepresentation, fuelled by return to
successful fraud, and this proportion varies over time; 2) some proportion of fraud has been
uncovered, reported and penalised, and this proportion varies over time as well; 3) the expectation of
detection changes the opportunities and the expected return to fraud, then affecting (1) recursively.
As indicated in Table 4, from the occurrence of financial statement fraud to that the fraud has been
detected and disclosed in the punitive notices by the China Securities Regulatory Commission, time
delay is usually between 2 and 6 years, with 3 years and 4 years in the majority. This indicates that
time delay is one of the main reasons that instances of financial statement fraud suddenly decrease
after 2004, and not all cases of financial statement fraud in recent years have been identified and
reported. Figure 2 describes the relationship line of occurring year and reporting year of financial
statement fraud, and we focus on the fraud cases that occur between 1999 and 2004, because the
delayed reporting years of the fraud cases in this period are mainly from 2002 to 2009. Figure 2
clearly shows that in reporting year t+1, the instances of financial statement fraud identified in a
certain occurring year are more than that in reporting year t, then we can reasonably expect that, a
kind of function relation exists between time delay of the fraud occurring year and reporting year Δt
and the increased instances of financial statement fraud indentified over time ΔFSF, which is assumed
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
17
as ΔFSF= f (Δt). Based on Table 4, we can find out the increased instances of financial statement
fraud indentified in each occurring year when Δt is from 2 to 6 years, as shown in Table 5. Therefore,
it can be reasonably estimated that although in the study sample, the cases of financial statement fraud
identified in 2005, 2006 and 2007 are respectively 2, 1 and 1, some proportion of fraud has not been
detected yet, and the estimated cases of financial statement fraud in these three years are probably
around 14, 13 and 13.
From Table 6, the average of financial statement fraud’s duration in the Chinese listed companies is
2.28 years which is more than a single year. Previous studies (Baucus and Near, 1991; Mulford and
Comiskey, 2002) state that illegal activities usually do not occur in one single accounting period, and
companies disclosed to commit financial fraud often have criminal records before. The 1999 COSO
Report also concludes that most fraudulent activities are not isolated to a single fiscal period.
5. EMPIRICAL ANALYSIS RESULTS
5.1 The Relationship of Board of Directors and Financial Statement Fraud
Based on the statistical description (Table 7) and analysis of variance (Table 8), it is found that the
situation that the board is controlled by the top management team (BD1) is more common in the fraud
firms than that in non-fraud firms, and the chairperson of the boards in more than a half of fraud
companies also serves as CEO or president. The percentage of independent directors in the board
(BD2) for both the fraud group and non-fraud group are on a low level, for the reason that
independent directors system in the Chinese securities market was set up in 2001, and the Chinese
Securities Regulatory Commission required that at the end of 2003, one third of directorate members
in the Chinese listed companies should be independent directors. Young independent directors system,
without actual supervision efficacy, is easily to be pro forma in some of the listed companies, which is
one of the probable reasons for the analysis results.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
18
Logistic regression is further deployed to test the hypothesized relationship of the board traits and
financial statement fraud. When univariate differences exist, the logistic regression offers advantages
over the comparison of Variance analysis because some firm specific factors which have been known
to have an effect on occurrences of financial statement fraud can be controlled. The Logistic
regression model here is:
FSFi= α+ β1BD1i+ β2BD2i+ β3ICi+ β4CSi+ β5Hi+ β6AOi+ β7LEVERi+ β8ROEi+ β9STi+ β10RSGi+ εi
(F 4)
Factor Analysis is also employed in advance of Logistic regression, so as to avoid the probable
multicollinearity problem because of highly correlated relationship among the variables. Table 13
reveals the output results of Logistic analysis. As regards variable BD1, its coefficient is positive and
statistically significant, suggesting that firms with a board that is controlled by top executive team are
more likely to engage in fraudulent behaviors. As for the board traits variable BD2, its significant
value is greater than 0.05 that fails in the significance test, suggesting that the proportion of
independent directors in the board does not affect the occurrence of financial statement fraud. The
results are consistent with what we get from Variance analysis. The results thus succeed in supporting
the research hypothesis.
H1.1 The situation that the board is controlled by top management team and the occurrence of
financial statement fraud are positively related: confirmed.
While the results fail to support the research hypothesis,
H1.2 The percentage of independent directors in the board is lower in fraud firms compared to
matched non-fraud firms: not confirmed.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
19
5.2 The Relationship of Supervisory Board and Financial Statement Fraud
From Table 7 and Table 8, the sizes of the supervisory board (SB1) for the fraud firms and non-fraud
firms are both on a low level2, with the averages of approximate 4 persons. The proportions of shares
owned by the board members (SB2) for the groups are also relatively low, with the average values of
0.0056 percent and 0.0049 percent respectively. Despite the results of variance analysis indicating that
there are no significant differences of the proportion of shares held by the supervisory board members
between the two groups, based on statistical description, it can be seen that the supervisory board
members of the fraud firms generally hold more shares than that of non-fraud firms. In China, the
board of directors and the supervisory board are parallel and independent, and the members of the
supervisory board are mostly shareholders and employees. In this binary corporate governance
structure, the power of the supervisory board is not like the German style and has no advantages of
Anglo-American Independent Directors. When the efficacy of the supervisory board is weak and
ineffective, and together with the situation that the directorate is controlled by top management, the
likelihood that managers engage in fraudulent financial reporting is enlarged a lot.
In the Logistic analysis, the regression model is developed as follows.
FSFi= α+ β1SB1i+ β2SB2i+ β3ICi+ β4CSi+ β5Hi+ β6AOi+ β7LEVERi+ β8ROEi+ β9STi+ β10RSGi+ εi
(F 5)
Table 13 reveals the output results. The results presented show that the size of the supervisory board
and the percentage of shares owned by the supervisory board members do not influence the likelihood
that top management produce fraudulent financial reports, as evidenced by the insignificant
coefficients on SB1 and SB2 (Sig. value=0.893, 0.840). The result is consistent with what is got from
Variance analysis that the presence of the supervisory board does not play a role in reducing and
2 According to the Chinese Company Law, the size of the supervisory board in the listed companies should be at least three persons. Available: http://www.gov.cn/ziliao/flfg/2005-10/28/content_85478.htm (in Chinese)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
20
preventing the occurrence of fraudulent financial reporting in practice. The empirical results fail to
support the research hypothesis.
H2 The firms committing financial statement fraud are more likely to have a smaller proportion of
shares owned by supervisors: not confirmed.
5.3 The Relationship of Audit Committee and Financial Statement Fraud
As indicated in Table 7 and Table 8, audit committees in the fraud group and non-fraud group are
both not common. The results of Variance analysis reveal that the fraud firms have significantly more
audit committees than non-fraud firms, suggesting probably 1) that although the oversight functions
of audit committee have been recognized in theory, it does not play a role in the practical operations;
2) that with reverse causation: firms with Audit committees may be more likely to have fraudulent
behaviours uncovered. For the two probable explanations, this paper prefers the latter one. Audit
committee system is still not universal in the Chinese corporate governance structure. It can be
expected that some fraud firms set up audit committee only in form but not in substance, in order to
show a good corporate governance structure, or to hide unfavourable performance.
In Logistic analysis, the regression model is developed as follows. Beasley (1996) argues that because
of many the board composition reform proposals, e.g., AICPA National Commission on Fraudulent
Financial Reporting 1987; AICPA Public Oversight Board 1993, 1994, the presence of an audit
committee indirectly influences board composition when outside directors are involved in the board to
serve on the audit committee. Therefore, some board traits variables that are relevant to the audit
committee are added to the Logistic model.
FSFi= α+ β1ACi+ β2OUTi+ β3INDi+β4ICi+ β5CSi+ β6Hi+ β7AOi+ β8LEVERi+ β9ROEi+ β10STi+
β11RSGi+ εi
(F 6)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
21
Where
AC- a dummy variable with a value of one if the firm has an audit committee in the year of
financial statement fraud and a value of zero otherwise;
OUT- the percentage of outside directors in the board;
IND- the proportion of independent directors in the board;
Table 13 reveals the output results of Logistic analysis. The results presented indicate that the
presence of an audit committee has no significant effect on the occurrence of financial statement fraud,
as evidenced by the insignificant coefficient on AC (Sig. value=0.093). Young audit committee
system in the Chinese capital market is still one of the main reasons that the presence of an audit
committee is less likely to reduce the possibility of financial statement fraud. However, this finding is
consistent with many western scholars’ reports (Sommer 1991, Beasley 1996) noting that considerable
anecdotal evidence is found that many, if not most, audit committees fall short of doing what they are
expected to do as their duties. The results gained fail to support the research hypothesis.
H3 The firms experiencing financial statement fraud are less likely to have an audit committee than
non-fraud firms: not confirmed.
5.4 The Relationship of External Auditor and Financial Statement Fraud
Based on the statistical description and analysis of variance (Table 7 and Table 8), it is found that both
the fraud firms and non-fraud firms have some instances of accounting firms changes (EA1), and
comparing to non-fraud firms, more changes of accounting firms occur in the fraud firms. The tenures
of accounting firms (EA2) for the matched control sample are longer, with the average of 4.48 years,
while for the study sample, the average tenure of accounting firms is 3.70 years. The probable reasons
for relatively more accounting firm changes in the fraud companies are as followed. First, because of
the lack of the knowledge in the beginning of audit engagement, it is impossible for external auditors
to fully get hold of the client’s financial situation and the probability that fraudulent financial
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
22
activities cannot be detected is high; as the auditor-client relationship lengthens, information
asymmetry between auditors and clients reduces over time, and the fraud companies, in order to make
their illegal acts not be uncovered, may choose to change the accounting firms to hide their financial
fraud engagement. Second, during the audit engagement, when auditors find out some clients’
fraudulent activities and require relevant revisions which are inconsistent with the clients’ interests,
the clients may not be willing to accept the revision requirements, and as a result, poor coordination
between auditors and the clients leads to termination of the cooperative contract and accounting firm
changes. Therefore, frequent changes of accounting firms can be deemed as one of indicators of
financial abnormality in the listed companies. As regards the tenure of accounting firms, the analysis
results indicate that the average for the fraud group is relatively smaller and the tenure is shorter,
suggesting again that more accounting firm changes happen in the fraud group. Comparing to the
western highly concentrative auditing market 3 , much dispersive auditing market in China and
numerous small-sized accounting firms in the market that engage in financial audits for the listed
companies, result in extremely serious low-cost competition in auditing sector. In such operating
environment, when independent auditors have irresolvable contradiction with the clients, they are
more likely to be fired. As to the effect of authority of accounting firms on financial statement fraud
(EA3), we find that the cases that employ the Big Four accounting firms as their external auditing
body are quite few in both the study sample and control sample, therein, 4 percent of fraud firms
employ the Big Four for financial audits while only 1 percent of non-fraud firms surveyed hire the Big
Four as external auditors, which reflects again the operating environment in the Chinese auditing
market. Thereby, there is no conclusive and supportive evidence to show that the larger accounting
firms provide more effective audit service and do better in detecting financial statement fraud.
Logistic regression is then involved to test the hypothesized relationship of external auditor traits and
financial statement fraud. The Logistic regression model is developed as followed, and eight control
3 The U.S. Government Accountability Office (GAO) study found the Big Four audit firms collected 94 percent of all audit fees paid by public companies in 2006 (96 percent in 2002); 82 percent of the large public companies surveyed see their auditor choice as limited to the Big Four in 2008.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
23
variables that have also been discussed to influence fraudulent financial reporting in the literature (e.g.,
Loebbecke et al., 1989; Beasley, 1996) are included.
FSFi= α+ β1EA1i+ β2EA2i+ β3EA3i+β4ICi+ β5CSi+ β6Hi+ β7AOi+ β8LEVERi+ β9ROEi+ β10STi+
β11RSGi+ εi
(F 7)
In advance of Logistic regression, Factor Analysis of the external auditor variables is performed in
order to avert the probable multicollinearity problem resulting from highly correlated relationship
among the variables. In the situation that research variables contain highly similar information, the
coefficient estimates may change erratically in response to small changes in the model. The
correlation matrix of Factor analysis can be seen in Table 9.
The variable EA1 (whether accounting firms are changed in the fraud year) and EA2 (the tenure of
accounting firms) indicate highly correlated relations, with the correlation coefficient of -0.459 and
the significant value of 0.000. According to Table 10, in matrix terms, we have
F1= 0.862*EA1- 0.844*EA2+ 0.049*EA3
F2= -0.034*EA1- 0.123*EA2+ 0.995*EA3 (F 8)
Factor F1 corresponds to the turnover of accounting firms that engage in financial audits for the listed
companies, while factor F2 corresponds to the authority of accounting firms. We substitute the two
factors F1 and F2 for the three external auditor variables EA1, EA2 and EA3 into original Logistic
regression model and new Logistic model can be seen as followed.
FSFi= α’+ β’1F1i+ β’2F2i+ β’3ICi+ β’4CSi+ β’5Hi+ β’6AOi+ β’7LEVERi+ β’8ROEi+ β’9STi+ β’10RSGi+
εi
(F 9)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
24
In Table 13, as regards the turnover of accounting firms factor F1, the results exhibited show that
financial fraud is more likely to be reported in the Chinese firms with more turnovers of accounting
firms, as evidenced by its significant coefficient (Sig. value= 0.039); while in the light of factor F2
representing the authority of accounting firms, no supportive evidence shows that accounting firms
with high authority (e.g. the Big Four) provide more effective audit service and perform better in
detecting financial statement fraud, as indicated by its insignificant coefficient (Sig. value=0.403).
The result is consistent with that of Variance analysis that fraud firms have significantly more
instances of accounting firms changes and shorter tenures of audit service, and no significant
variability regarding the authority of accounting firms is found between the fraud sample and the
matched control sample. The results gained succeed in supporting the research hypotheses.
H4.1 Change of accounting firms and the occurrence of financial statement fraud are positively
related: confirmed;
H4.2 Tenure of accounting firms and the occurrence of financial statement fraud are negatively
related: confirmed.
But the results are inconsistent with the research hypothesis,
H4.3 Auditing quality and the occurrence of financial statement fraud are negatively related: not
confirmed.
5.5 The Relationship of Regulation and Financial Statement Fraud
As indicated in the statistical description and analysis of variance (Table 7 and Table 8), in the fraud-
occurring years, quite a few companies that engage in financial statement fraud are signed ST, and
most of the fraud companies indicate profitable operating performance; while in the matched non-
fraud group, the companies that have been signed ST are much more than that in fraud group.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
25
Seemingly, the occurrence of financial statement fraud has no significant association with regulation
pressure, and companies that are not signed ST and not pressured by regulation requirements commit
more fraudulent financial activities. However, with consideration of variable RG2, the true story is
presented. Variable RG2 measures in fraud-reporting year when financial fraud is disclosed by the
China Securities Regulatory Commission, if the firm is signed ST or not. As regards variable RG2, no
significant difference is found between fraud companies and non-fraud companies, with respectively
30 percent and 21 percent of the companies that are special treated because of poor financial
performance. Based on the Timing description of the fraud sample in Part 4, it is known that from the
occurrence of financial statement fraud to that the fraud has been detected and disclosed in punitive
notices by the China Securities Regulatory Commission, time delay is usually between 2 and 6 years,
with 3 years and 4 years in the majority. According to variable RG1 and RG2, it is reasonably
interpreted that when the listed companies are running into financial difficulties, for fear that both
debt financing and equity financing are influenced resulting from being ST, they may perpetrate
fraudulent behaviours to falsely present good and stable operating performance. However, financial
fraud only works to temporarily cover the financial difficulties of the companies, and if no effective
solutions are performed, the financial difficulties will finally be disclosed over a period to come,
which leads that the companies are signed ST with a few years delayed. Thereby, when the companies
are possible and pressured to be signed ST by regulation commission, they have greater motivations
to engage in financial statement fraud. As for time length of a firm being listed in public markets, we
find that non-fraud group’s listing time length is significantly longer than that of fraud group, which is
consistent with the results of The National Commission on Fraudulent Financial Reporting (AICPA,
1987) studies, and companies that are newly listed in public markets have a proportionately greater
risk of financial statement fraud for the probable purpose of meeting earnings expectations and
increase financing degree in the stock market.
In Logistic analysis, the regression model is:
FSFi= α+ β1RG1i+ β2RG2i+ β3RG3i+β4ICi+ β5CSi+ β6Hi+ β7AOi+ β8LEVERi+ β9ROEi+ β10RSGi+ εi
(F 10)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
26
Factor Analysis is also performed in advance of Logistic regression. The correlation matrix of factor
analysis is shown in Table 11. The variable RG1 (whether the firm is signed ST in the fraud-occurring
year) and RG2 (whether the firm is signed ST in the fraud-reporting year) indicate highly correlated
relations, with the correlation coefficient of 0.457 and the significant value of 0.000. On the basis of
Table 12, in matrix terms, we have
F1’= 0.850*RG1+ 0.848*RG2+ 0.108*RG3
F2’= 0.088*RG1+ 0.096*RG2+ 0.994*RG3 (F 11)
Factor F1’ corresponds to the consistency of financial performance of the listed companies, while
factor F2’ corresponds to time length of the listed companies being listed in public markets. We
substitute F1’ and F2’ for the regulation variables RG1, RG2 and RG3 into original Logistic
regression model. The new Logistic model is seen as followed.
FSFi= α’+ β’1F1’i+ β’2F2’i+ β’3ICi+ β’4CSi+ β’5Hi+ β’6AOi+ β’7LEVERi+ β’8ROEi+ β’9RSGi+ εi
(F 12)
Table 13 reveals that as to the consistency of financial performance factor F1’, the results exhibited
demonstrate that financial fraud is more likely to be reported in the Chinese firms with inconsistent
financial performance in fraud-occurring years and fraud-reporting years, as evidenced by its
significant coefficient (Sig. value= 0.040); while in the light of factor F2’ representing time length of
the listed companies being listed in public markets, supportive evidence is found that companies that
are newly listed in public markets have proportionately greater likelihood of financial statement fraud,
as indicated by its significant coefficient (Sig. value=0.034). The results are consistent with that of
Variance analysis that when the listed companies are running into financial difficulties, for fear of the
negative influences resulted from being ST, they may perpetrate fraudulent behaviours to temporarily
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
27
hide poor finance performance, and regarding time length of the companies being listed in capital
market, non-fraud group’s listing time length is significantly longer than that of fraud group. The
results gained succeed in supporting the research hypotheses.
H5.1 Avoiding regulation in securities market (e.g. ST) and the occurrence of financial statement
fraud are positively related: confirmed;
H5.2 Time length of being listed and the occurrence of financial statement fraud are negatively
related: confirmed.
6. CONCLUSIONS
The findings of this research challenge the conventional arguments which have been testified based
on the data from the western countries. Conventional arguments show financial fraud is associated
with weakness of governance in western companies (e.g. Beasley et al., 2000). However, the finding
in this research reveals that as for some high discussed corporate governance characteristics (e.g. the
supervisory board, audit committee, independent directors), the fraud firms and their non-fraud peers
are not statistically distinct, suggesting that corporate governance mechanisms that are designed to
reduce the probability of financial fraud fail to work in the Chinese market. Statistical differences of
external auditor traits and regulation from the securities market are found between the groups.
The negative results bring deep thinking to both academic researchers and policymakers. Why are
contradictions gained in the Chinese market? It is difficult to answer the question at the moment.
One probable explanation provided here relates to the system of share ownership. Two basic
ownership systems are: disperse ownership system (e.g., the U.S. market) and concentrated ownership
system (e.g., the Continental European, and the Chinese markets) (Wang, 2003; Coffee, 2005; Zhang,
2007). In the dispersed ownership system, corporate performance is mostly controlled by managers
and governance issues arise from agency relationship between shareholders and managers (Jensen
and Meckling, 1976; Jensen, 1986); while in the concentrated ownership system, there is usually a
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
28
controlling shareholder or shareholder group in the corporations who has absolute control on
managers and corporate performance, and governance issues refer to if the controlling shareholders
overreach minority shareholders (Gordon, 1999; Kirchmaier and Grant, 2004). The closest argument
is supported by Coffee (2005), who argues that gatekeepers, that is professional agents serving
shareholders but selected by the corporation, are not believed to work as well in concentrated
ownership regimes as in dispersed ownership regimes. Whether different ownership system accounts
for the inconsistent findings in different economies, further research is warranted. However, it is
worth noting that this research develops the analysis based on the data from the concentrated
ownership system, and all the research findings are valid only in the concentrated ownership system.
The research contributes most that regulations and governance protections that work in one system
may fail in the other, and different gatekeepers need to be designed into different governance systems
to monitor for different frauds.
In appraising the findings of this paper, it is important to consider the following limitations. The
studied cases in this paper are mainly chosen from the punitive notices disclosed by the Chinese
Securities Regulatory Commission. For the reason that the CSRC is most inclined to take formal
investigations with strong evidences and high probability of success, focusing on the punitive notices
of the CSRC might offer the possibility of bias. There are likely to be potential instances of fraudulent
financial reporting not included in the study sample. Also, potential possibility of bias may exist in
choosing the control sample. Potential cases of financial statement fraud might be included in the
control group. Although due to this limitation, the matched non-fraudulent sample is carefully chosen
and also checked if they have some abnormal financial indices that may suggest a greater possibility
of engaging in fraudulent activities. The bias of choosing control sample is only minimized but cannot
be eliminated. Furthermore, because of the timing traits of the research sample, the fraud sample
would have over-representation of past fraud, and again, there would also be probability of recent
fraudulent cases involved in the control sample.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
29
REFERENCE
American Institute of Certified Public Accountants (AICPA). (1997). Serving the Public Interest: A New Conceptual Framework for Auditor Independence, The Independence Standards Board (November). Exposure Draft. New York: AICPA.
Baucus, M.S. and Near, J.P., (1991). Can Illegal Corporate Behaviour Be Predicted? An Event History Analysis. Academy of Management Journal. (34), pp. 9-36.
Beasley, M.S., (1996). An Empirical Analysis of the Relation between the Board of Director Composition and Financial Statement Fraud. Accounting Review, 71(4), pp. 443-465.
Beasley, M.S., Carcello, J.V. and Hermanson, D.R., (1999). Fraudulent Financial Reporting: 1987-1997, An Analysis of U.S. Public Companies. New York, NY: COSO.
Beasley, M.S., Carcello, J.V., Hermanson, D.R. and Lapides, P.D., (2000). Fraudulent Financial Reporting: Consideration of Industry Traits and Corporate Governance Mechanisms. Accounting Horizons, 14(4), pp. 441-454.
Beck, P., Frecka, T. and Solomon, I., (1988). A Model for the Market for Mas and Audit Service: Knowledge Spillovers and Auditor-Auditee Bonding. Journal of Accounting Literature. 7(1), pp. 50-64.
Bell, T.B., Szykowny, S. and Willingham, J.J., (1991). Assessing the Likelihood of Fraudulent Financial Reporting: A Cascaded Logit Approach. Working Paper, KPMG Peat Marwick, Montvale, NJ.
Carcello, J.V., Hermanson, D.R. and Huss, H.F., (2000). Going-Concern Opinions: the Effects of Partner Compensation Plans and Client Size. Auditing, 19(1), pp. 67-77.
Chen, G., (2007). Positive Research on the Financial Statement Fraud Factors of Listed Companies in China. Auditing Research. (5), pp. 91-96. (In Chinese)
Chen, X. and Xia, L., (2006). Audit Tenure and Audit Quality—Empirical Evidence from the Chinese Securities Market. Accounting Research. (1), pp. 44-54. (In Chinese)
Coffee Jr., J.C., (2005). A Theory of Corporate Scandals: Why the USA and Europe Differ. Oxford Review of Economic Policy, 21(2), pp. 198-211.
Daniel, A., (2002). The Determinants of the Amount of Information Disclosed about Corporate Restructurings. Journal of Accounting Research. 40(1), pp. 1-19.
Davis, L.R., Soo, B. and Trompeter, G., (2003). Auditor Tenure, Auditor Independence and Earnings Management. Working Paper, Boston College.
DeAngelo, L.E., (1981). Auditor Independence, 'Low Balling', and Disclosure Regulation. Journal of Accounting and Economics, 3(2), pp. 113-127.
Dechow, P.M., Sloan, R.G. and Sweeney, A.P., (1996). Causes and Consequences of Earnings Manipulation: An Analysis of Firms Subject to Enforcement Actions by the SEC. Contemporary Accounting Research, 13(1), pp. 1-36.
DeFond, M.L. and Jiambalvo, J., (1994). Debt Covenant Violation and Manipulation of Accruals. Journal of Accounting and Economics, 17(1-2), pp. 145-176.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
30
Fama, E.F. and Jensen, M.C., (1983). Separation of Ownership and Control. Journal of Law and Economics, 26(6), pp. 301-325.
Geiger, M.A. and Raghunandan, K., (2002). Auditor Tenure and Audit Reporting Failure. Auditing: A Journal of Practice and Theory. 21(1), pp. 67-78.
Gordon, J. N., (1999). Pathways to Corporate Convergence?: Two Steps On the Road to Shareholder Capitalism in Germany. Columbia Journal of European Law. 5(219).
Hu, H., (2006). A Theoretical and Empirical Research on Fraudulent Financial Report. Working Paper, Wuhan University of Technology, China. (In Chinese)
Hu, X., (2002). Does the Big Five’s Reputation Has Significant Information Value? Securities Market Herald, (3), pp. 43-46. (In Chinese)
Huang, Q., (2005). A Study on the Structure of Board of Directors and Financial Reports Fraud. Journal of Finance and Accounting. (3), pp. 49-52. (In Chinese)
Jensen, M. and Meckling, W., (1976). Theory of Firm:Managerial Behaviour,Agency Costs,and Ownership Structure. Journal of Financial Economics. 3, pp. 305-360.
Jensen, M., (1986). Agency Costs of Free Cash Flow, Corporate Finance and Takeovers. American Economic Review. (76), pp. 323-329.
Jensen, M., (1993). The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems. Journal of Finance. 48, pp. 831-880.
Kirchmaier, T., and Grant, J., (2004). Financial Tunnelling and the Revenge of the Insider System: How to Circumvent the New European Corporate Governance Legislation. Available at http://ssrn.com/abstract=613945.
Knapp, M., (1991). Factors that Audit Committees Use as a Surrogates for Audit Quality. Auditing: A Journal of Practice and Theory. 10 (1), pp. 35-52.
Latham, C.K. and Jacobs, F.A., (2000). Monitoring and Incentive Factors Influencing Misleading Disclosures. Journal of Managerial Issues, 7(2), pp. 169-187.
Leftwich, R.W., Wattts, R.L. and Zimmerman, J.L., (1981). Voluntary Corporate Disclosure: The Case of Interim Reporting. Journal of Accounting Research. 19, pp. 246-271.
Liu, L. and Du, Y., (2003). An Empirical Research on the Relationship between Corporate Governance and the Quality of Accounting Information. Accounting Research. (2), pp. 28-37. (In Chinese)
Liu, M., Li, L. and Zhang, Y., (2003). Empirical Analysis on Concentration of the Chinese Audit Market and Audit Quality. Accounting Research. (7), pp. 37-42. (In Chinese)
Loebbecke, J.K., Eining, M.M. and Willingham, J.J., (1989). Auditors’ Experience with Material Irregularities: Frequency, Nature, and Detectability. Auditing: A Journal of Practice & Theory, (Fall), pp. 1-28.
Lu, C. and Wang, L., (2004). An Empirical Study on Boards Design of Public Companies and Related Performance --Some Facts from Northeast China. Research on Financial and Economic Issues. (10), pp. 58-67. (In Chinese)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
31
Lys, T. and Watts, R., (1994). Lawsuits against Auditors. Journal of Accountancy Research. 32(Supplement), pp. 65-93.
Mautz, R.K. and Sharaf, H.A., (1961). The Philosophy of Auditing, American Accounting Association, Sarasota, FA. pp. 245-248.
McMullen, D.A., (1996). Audit Committee Performance: An Investigation of the Consequences Associated with Audit Committees. Auditing, 15(1), pp. 87-103.
Mulford, C.W. and Comiskey, E.E., (2002). The Financial Numbers Game, John Wiley & Sons, Inc.
O'brien, P.C., (1988). Analysts' Forecasts as Earnings Expectations. Journal of Accounting and Economics, 10(1), pp. 53-83.
Rezaee, Z., (2002). Financial Statement Fraud: Prevention and Detection, John Wiley & Sons, Inc.
Robertson, J.C., (2000). Fraud Examination for Managers and Auditors. Austin, TX: Viesca Books.
Securities and Exchange Commission (SEC). Speech by SEC Commissioner Hunt Jr Isaac C Remarks before the ABA Committee on Federal Regulation of Securities; 2002a April 6. Available at http://www.sec.gov/new/speech/spch550.htm.
Simpson, S.S., (1986). The Decomposition of Antitrust: Testing a Multi-Level Longitudinal Model of Profit-Squeeze. American Sociological Review. (51), pp. 859-875.
Solomon, I., Shields, M. and Whittington, R., (1999). What Do Industry Specialist Auditors Know? Journal of Accounting Research. 37(1), pp. 191-208.
Sommer, A.A., (1991). Auditing Audit Committees: An Educational Opportunity for Auditors. Accounting Horizons. (6), pp. 91-93.
The U.S. Government Accountability Office (GAO), 2009. Improvements Needed in Corrective Action Plans to Remediate Financial Reporting Material Weaknesses. Available: http://www.gao.gov/new.items/d1065.pdf.
Wang, F. and Yang, D., (2006). Empirical Research on Accounting Estimate Application’s Impact of Earnings Management in Chinese Public Companies, Proceedings of International Conference on Construction and Real Estate Management 2006, Edited By Yaowu Wang. Orlando, FL.: China Architecture & Building Press, pp.1094-1099.
Wang, G. and Li, X., (2004). The Auditing Responsibility for the Irregularities of Financial Statement. Commercial Research. (03), pp. 142-144. (In Chinese)
Wang, Y., (2001). Three Hot Points in the Chinese Stock Market in 2001. Available: http://finance.sina.com.cn/t/34687.html. (In Chinese)
Wang, Y., (2003). Empirical Study on Index of Voluntary Corporate Disclosure. Securities Market Herald. (9), pp. 10-19.
Watts, R.L. and Zimmerman, J.L., (1990). Positive Accounting Theory: A Ten Year Perspective. The Accounting Review, pp. 65.
Wen, C., (2007). Effect Analysis on Audit Committee. China Economist. (7), pp. 87-89. (In Chinese)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
32
Ye, C. And Ban, W., (2008). Analysis on Establishing Audit Committee Mechanism in the Chinese Listed Companies. Financial and Accounting Monthly. (6), pp. 74-80. (In Chinese)
Yu, Y. and Li, L., (2003). Theoretical Analysis on Audit Tenure and Audit Quality. Economic Review. (5), pp. 125-129. (In Chinese)
Zhang, B., (2007). Evolution and Revelation of Audit Committee. China Economist. (5), pp. 243-244. (In Chinese)
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
33
Table 3 Time distribution of the studied sample
Occurring
year Fraudulent financial
statements (occurred) Fraud in this year/
the total (%) Companies with FSF
finished Companies in this year/ the total (%)
1996 2 2.20 0 0
1997 5 5.49 1 2.50
1998 9 9.89 2 5.00
1999 9 9.89 2 5.00
2000 14 15.38 6 15.00
2001 13 14.29 8 20.00
2002 11 12.09 3 7.50
2003 14 15.38 7 17.50
2004 10 10.99 8 20.00
2005 2 2.20 2 5.00
2006 1 1.10 0 0
2007 1 1.10 1 2.50
Total 914 100 40 100
Table 4 The relation of financial statement fraud (FSF) occurring year and reporting year
Number of fraudulent financial statements identified in the reporting year Occurring year 2002 2003 2004 2005 2006 2007 2008 2009
1996 1 0 0 0 1 0 0 0
1997 2 1 1 0 1 0 0 0
1998 3 2 3 0 1 0 0 0
1999 2 3 3 0 1 0 0 0
2000 1 5 5 2 1 0 0 0
2001 1 2 7 2 1 0 0 0
2002 ---- 0 2 4 3 2 0 0
2003 ---- ---- 0 4 4 5 0 1
2004 ---- ---- ---- 1 3 3 2 1
2005 ---- ---- ---- ---- 0 1 0 1
2006 ---- ---- ---- ---- ---- 0 1 0
2007 ---- ---- ---- ---- ---- ---- 1 0
Total 10 13 21 13 16 11 4 3
4 The fraud sample here is 91 financial statements in 40 firms. Due to lack of the relevant financial information in 9 cases, the research sample in variance analysis and Logistic regression is 82 fraudulent cases.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
34
Table 5 The relation of ΔFSF identified and time delay Δt
Δt, time delay of FSF occurring year and reporting year Occurring year Δt=2 Δt=3 Δt=4 Δt=5 Δt=6
1999 ---- 2 5 8 9
2000 1 6 11 13 14
2001 3 10 12 13 13
2002 2 6 9 11 11
2003 4 8 13 13 14
2004 4 7 9 10 ----
Average 2.8 6.5 9.8 11.3 12.2
Table 6 Fraud duration of the studied sample
Fraud duration 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years Total
Companies 13 15 6 4 0 1 0 1 40
Average 2.28
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
35
Figure 1 Time distribution line of the studied sample
Figure 2 The relation of financial statement fraud occurring year and reporting year
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
36
Table 7 Statistical description of corporate governance characteristics
Fraud firms Non-fraud firms
Variables N Mean Std. D Variance Variables N Mean Std. D Variance
BD1 82 0.51 0.503 0.253 BD1 82 0.20 0.399 0.159
BD2 82 14.856 14.638 214.291 BD2 82 15.696 15.122 228.679
SB1 82 4.02 1.100 1.209 SB1 82 4.11 1.305 1.704
SB2 82 0.0056 0.0077 0.000 SB2 82 0.0049 0.0090 0.000
AC 82 0.29 0.458 0.210 AC 82 0.16 0.367 0.135
EA1 82 0.17 0.379 0.143 EA1 82 0.11 0.315 0.099
EA2 82 3.70 2.147 4.610 EA2 82 4.48 3.159 9.981
EA3 82 0.04 0.189 0.036 EA3 82 0.01 0.110 0.012
RG1 82 0.02 0.155 0.024 RG1 82 0.11 0.315 0.099
RG2 82 0.30 0.463 0.215 RG2 82 0.21 0.408 0.166
RG3 82 5.68 2.238 5.009 RG3 82 6.79 3.332 11.105
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
37
Table 8(a) Summary of output results of test of variance analysis
Variables df Mean Square F Sig.
Between Groups 1 28.963 0.131 0.718BD2
Within Groups 162 221.485
Between Groups 1 0.299 0.205 0.651SB1
Within Groups 162 1.457
Between Groups 1 0.000 0.251 0.617SB2
Within Groups 162 0.000
Between Groups 1 0.115 5.184 0.028LgEA25
Within Groups 162 0.097
Significance level: 0.05
Table 8(b) Summary of output results of test of variance analysis6
Variables Statistic* df1 df2 Sig.
Welch 20.013 1 153.990 0.000BD1
Brown-Forsythe 20.013 1 153.990 0.000
Welch 4.282 1 154.763 0.040AC
Brown-Forsythe 4.282 1 154.763 0.040
Welch 5.259 1 156.731 0.026EA1
Brown-Forsythe 5.259 1 156.731 0.026
Welch 1.019 1 130.576 0.315EA3
Brown-Forsythe 1.019 1 130.576 0.315
Welch 4.858 1 118.243 0.029RG1
Brown-Forsythe 4.858 1 118.243 0.029
Welch 2.049 1 159.449 0.154RG2
Brown-Forsythe 2.049 1 159.449 0.154
Welch 6.267 1 141.722 0.013RG3
Brown-Forsythe 6.267 1 141.722 0.013
*Asymptotically F distributed; Significance level: 0.05 5 In variance analysis, variable EA2 fails in the homogeneity testing. By dealing with original EA2 values logarithmically, variance analysis of LgEA2 passes the homogeneity test. 6 The variables in this form all fail in homogeneity testing of variance analysis. When their original values cannot be dealt with logarithmically, Statistic Brown-Forsythe and Statistic Welch are involved, which are better than Statistic F, when homogeneity of variances for some variables cannot be assumed.
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
38
Table 9 Correlation matrix of Factor analysis for external auditor variables
EA1 EA2 EA3
EA1 1.000 -0.459 0.050
EA2 -0.459 1.000 -0.122Correlation
EA3 0.050 -0.122 1.000
EA1 0.000 0.263
EA2 0.000 0.061Sig.
(1-tailed) EA3 0.263 0.061
KMO and Bartlett’s Test: 40.605 (Sig.: 0.000)
Significance level: 0.05;
Table 10 Rotated component matrix for external auditor variables
Component
1 2
EA1 0.862 -0.034
EA2 -0.844 -0.123
EA3 0.049 0.995
Rotation Method: Varimax with Kaiser Normalization Where EA1- A dummy variable with a value of one when accounting firms are changed in the fraud year and a value of zero otherwise EA2- The tenure of accounting firms EA3- The authority of accounting firms with a value of one when the accounting firm is the Big Four and a value of zero otherwise
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
39
Table 11 Correlation matrix of Factor analysis for regulation variables
RG1 RG2 RG3
RG1 1.000 0.457 0.181
RG2 0.457 1.000 0.185
Correlation
RG3 0.181 0.185 1.000
RG1 0.000 0.010
RG2 0.000 0.009
Sig.
(1-tailed)
RG3 0.010 0.009
KMO and Bartlett’s Test: 45.340 (Sig.: 0.000)
Significance level: 0.05;
Table 12 Rotated component matrix for regulation variables
Component
1 2
RG1 0.850 0.088
RG2 0.848 0.096
RG3 0.108 0.994
Rotation Method: Varimax with Kaiser Normalization Where RG1- Special treated in fraud-occurring year, with a value of one when a firm is signed ST and a value of zero otherwise RG2- Special treated in fraud-reporting year, with a value of one when a firm is signed ST and a value of zero otherwise RG3- Time length of a firm being listed in public markets
Fraudulent Financial Reporting in China: Consideration of Timing Traits and Corporate Governance Mechanisms
40
Table 13 Summary of output results of Logistic regression
Independent variable
Predicated relation
Estimated coefficients
Standard Errors Wald value Sig.
The board traits variables:
BD1 + 1.473 0.531 7.682 0.006
BD2 none -0.026 0.020 1.604 0.205
The supervisory board variables:
SB1 none 0.023 0.169 0.018 0.893
SB2 none 4.907 24.372 0.041 0.840
The audit committee variables:
AC none 1.029 0.613 2.820 0.093
The external auditor variables: turnover of accounting firms factor F1
F1 + 0.367 0.222 4.424 0.039
The authority of accounting firms factor F2
F2 none 0.219 0.261 0.700 0.403
The regulation traits variables: the consistency of financial performance factor F1’
F1’ - -0.047 0.245 4.419 0.040
Time length of the companies being listed factor F2’
F2’ - -0.493 0.233 4.475 0.034
Chi-Square Test of Model’s Fit Sig.: 0.000
Total cases: 164
Fraud cases: 82
Non-fraud cases: 82
Significance level: 0.05 Where BD1- A dummy variable with a value of one when the chairman of the board serves as CEO or president and a value of zero otherwise BD2- The proportion of independent directors in the board SB1- Size of the supervisory board SB2- The percentage of shares owned by the supervisory board members AC- A dummy variable with a value of one if the firm has an audit committee in the year of financial statement fraud and a value of zero otherwise F1- The factor corresponding to turnover of accounting firms that engage in financial audits, with a value obtained from Factor analysis F2- The factor corresponding to the authority of accounting firms that engage in financial audits, with a value obtained from Factor analysis
F1’- The factor corresponding to the consistency of financial performance of the listed companies, with a value obtained from Factor analysis F2’- The factor corresponding to time length of the listed companies being listed in public markets, with a value obtained from Factor analysis