the impact of internal governance mechanisms on audit ... · 1 the impact of internal governance...
TRANSCRIPT
1
The impact of internal governance mechanisms on audit quality: a study of large listed
companies in China
Yuan George Shan
Business School, University of Adelaide, Adelaide
10 Pulteney Street, Adelaide, South Australia 5005, Australia
Email: [email protected]
Abstract
Purpose – This paper examines whether audit quality of Chinese listed companies is affected
by internal governance mechanisms (IGMs).
Design/methodology/approach – The sample contains a panel data set of 117 companies
with 540 firm-year observations during 2001–2005. The logistic regression model is used to
investigate the three sets of hypotheses regarded IGMs.
Findings – The results show that foreign ownership and the number of professional
supervisors are positively related to audit quality, but the size of supervisory board shows a
negative correlation. Other IGMs including state ownership, the size of board of directors, the
number of independent directors, the frequency of board meetings, and the frequency of
supervisory board meetings are found to have no impact on audit quality.
Practical implications – This paper offers insights to policy-makers and regulators interested
in enhancing corporate governance in China. It also provides recommendations to prescribe
the legal responsibilities and obligations for two-tier boards in the Chinese context, allowing
them to undertake their duties diligently.
Originality/value – The contribution of this paper is twofold. First, prior literature
incompletely uncovers the impact of concentrated ownership and two-tier board system on
2
audit quality. This paper provides a comprehensive evaluation of independent and
interdependent effects of IGMs of Chinese characters and investigates how these IGMs
jointly affect audit quality in China. Second, this paper considers the improvement of legal
environment as one of external governance mechanisms (EGMs) proposed by Denis and
McConnell (2003) and examine the relationship between company‟s IGMs and it‟s the
decision for audit quality choice.
Keywords Corporate Governance, Audit Quality, China
Paper type Research paper
1. Introduction
From early 2001, corporate financial scandals involving listed companies in China (e.g.,
Guanxia Industry Co. Ltd, Macat Optics and Electronics Co. Ltd, Sanjiu Pharmaceutical Co.
Ltd and Lantian Co. Ltd) prompted policy-makers, regulators and officials at the China
Securities Regulatory Commission (CSRC) and the State Economic and Trade Commission
(SETC) to make corporate governance a priority issue for their agenda in 2002. The
government came under increasing public pressure to improve its policy towards corporate
governance structures and grant permission to a select set of accounting firms to audit public
companies (DeFond et al., 2000). Additionally and importantly, according to the WTO entry
agreement, China should be committed to capital market liberalisation and corporate reform,
in particular with respect to state-owned enterprises (SOEs). An effective and efficient China
corporate governance system must coincide with international standards, allowing both
domestic companies and foreign companies to enjoy free market competition within a fair
economic and policy environment.
3
The purpose of this paper is to examine in a more comprehensive way than has been
the case previously the key internal governance mechanisms (IGMs) of listed companies in
China in their association with audit quality.
According to the Cadbury Report (1992, p. 36) the annual audit is “one of the
cornerstones of corporate governance…The audit provides an external and objective check on
the way in which the financial statements have been prepared and presented.” But the
effectiveness and efficiency of external auditing is subject to the actuality and the
development of the corporate governance environment (Holm and Laursen, 2007). A sound
corporate governance mechanism plays an important role in enhancing the audit function
because it assists in ensuring that directors and management of listed companies appoint high-
quality auditors who will exercise independent and effective monitoring over the financial
reporting process (Lin and Liu, 2009a).
Under agency theory, principal-agent problems stem from the complex set of
contracts engaged in by the firm. Such contracts may be either explicit or implicit in the
separation of ownership from control of the firm. Perhaps best known principal-agent
problem is where a divergence of interests between owners and managers causes the agent
(management) to fail to maximize the welfare of the principal (shareholders) (Jensen and
Meckling, 1976; Shleifer and Vishny, 1997; Denis and McConnell, 2003). From this
perspective, the audit function represents a vital corporate governance mechanism that helps
shareholders in their monitoring and control of company management. The audit of a
company‟s financial statements makes disclosures more credible, thereby instilling
confidence in the company‟s transparency. Indeed, auditing has been considered to play a
role in contract monitoring, as a company‟s auditors contract with debtholders to report any
observed breaches of restrictive covenants and audited earnings numbers are used in bonus
plans (Watts and Zimmerman, 1986). This argument is commonly accepted in prior auditing
4
research that examines the quality of auditing (e.g., DeAngelo, 1981; DeFond, 1992;
O'Sullivan, 2000; Kane and Velury, 2004; Chan et al., 2007; Boo and Sharma, 2008; Al-Ajmi,
2009; Lin and Liu, 2009a, 2009b; Lin et al., 2009; Gul and Goodwin, 2010; Zerni et al., 2010;
Zaman et al., 2011; Liu et al., 2011). Some of these studies introduce corporate governance
mechanism in terms of principal-agent conflicts to analyse audit quality through proxies such
as audit firm size (Kane and Velury, 2004; Lin and Liu, 2009b), audit fees (O'Sullivan, 2000;
Boo and Sharma, 2008; Zaman et al., 2011), auditor switching (Lin and Liu, 2009a; Lin et al.,
2009) and modified audit opinions (Liu et al., 2011). For example, Kane and Velury (2004)
hypothesise that American institutional owners prefer audits conducted by large audit firms
due to their belief that large audit firms provide relatively higher audit quality. Their findings
confirmed the positive association between institutional ownership and audit firm size (Big 6).
O'Sullivan (2000) uses audit fees as a surrogate for audit quality to examine the
impact of board composition and ownership structure on audit quality in the UK prior to the
adoption of the recommendations of the Cadbury Committee (1992). His findings suggest
that audit fees have a positive correlation with the proportion of non-executive directors and a
negative association with the proportion of equity owned by executive directors, but have no
impact with ownership by large institutional blockholders or CEO/chairman duality. Boo and
Sharma (2008) investigate the nature of the relationship between three corporate governance
mechanisms (board/audit committee independence, external auditor size and regulators) for a
sample of industrial firms and financial/utility firms subject to industry-specific regulation.
They find weaker relationships between audit fees and board/audit committee independence
and size for regulated firms. This finding suggests that regulatory oversight partially
substitutes the external audit as a monitoring mechanism. Zaman et al. (2011) examine the
correlation between governance quality and auditor remuneration in the UK based on a
composite measure of four dimensions of audit committee effectiveness (independence and
5
financial expertise of audit committee members, frequency of meetings and size of the audit
committee. They find a significant positive association between audit committee
effectiveness and audit fees for large firms after controlling for board of director
characteristics, which indicates that effective audit committees undertake more monitoring,
resulting in higher audit fees.
Rather than the traditional principal-agent problem, a principal-principal problem is
introduced as a major concern in relation to corporate governance in emerging economies
(Dharwadkar et al., 2000). These economies are characterized by high ownership
concentration, extensive family ownership and control, and weak legal protection of minority
shareholders (Dharwadkar et al., 2000; Young et al., 2008). To moderate the principal-
principal conflict of interests between controlling shareholders and minority shareholders
Denis and McConnell (2003) propose improving IGMs (ownership structure, board of
directors and supervisory board) and external governance mechanisms (EGMs) (corporate
control, legal environment and market development). However, in the context of China,
EGMs such as corporate control and market development are underdeveloped and state
ownership remains for most Chinese listed firms. Even reform of EGMs is an important issue
but arguably a prerequisite for establishment of EGMs is to establish a series of effective
IGMs because they play a substantial role at the juncture of corporate governance
development in China (Hu et al., 2010).
Using a panel data set of 117 companies with 540 firm-year observations during
2001–2005, this paper examines whether audit quality for Chinese listed companies is
associated with IGMs. The results reveal that foreign ownership and the number of
professional supervisors are positively related to audit quality, but the size of the supervisory
board shows a negative correlation. Other IGMs including state ownership, the size of the
6
board of directors, the number of independent directors and the frequency of each of board
and supervisory meetings are found not to be associated with audit quality.
In the context of studying the principal-principal agency conflict of interest between
controlling and minority shareholders the contribution of this paper is twofold. First, prior
literature incompletely uncovers the impact of concentrated ownership and two-tier board
system on audit quality. This paper provides a comprehensive evaluation of independent and
interdependent effects of IGMs of Chinese characters (i.e., highly concentrated state
ownership, board of directors and supervisory board) and investigates how these IGMs
jointly affect audit quality in China. Second, this paper considers the improvement of legal
environment as one of EGMs proposed by Denis and McConnell (2003). Thereby data was
selected during an important regulatory reform period between 2001 and 2005. Within this
period, The Guidelines for Introducing Independent Directors to the Board of Directors of
Listed Companies (The Guidelines) was introduced in August 2001 and mandated by 30 June
2003, and The Code of Corporate Governance for Listed Companies in China (The Code)
was promulgated in January 2002. The improvement on legal environment affords a good
opportunity to examine the relationship between company‟s IGMs and it‟s the decision for
audit quality choice.
The remainder of this paper is organized as follows. The following section provides
an overview of the literature on audit quality in China and develops three sets of hypotheses
in accordance with the IGMs to be tested. Section 3 outlines the research method and
describes the data, while Section 4 presents the results of the analysis and discusses the
findings. Section 5 discusses the implications of the findings and notes the paper‟s
limitations and future research ideas.
7
2. Literature and hypotheses
Study of corporate governance and audit quality has been drawing attention in China recently.
However, none of these studies includes IGMs in a comprehensive way. For example, Chan
et al. (2007) investigate whether the demand for quality-differentiated audits by listed
Chinese firms is related to changes in state or institutional ownership structure. Their results
suggest that a decrease in state ownership and a corresponding increase in institutional
ownership results in demand for higher quality auditing by the firms. They conclude that
managers of listed firms have incentives to supply credible accounting information through
quality audits when the institutional features became absent. Thus, the introduction of large
institutional shareholders represents a sound development in terms of economic reform in
China. Lin and Liu (2009a) examine the association between Chinese listed firms‟ internal
corporate governance mechanisms and their auditor switching decisions. Their results show
that firms with higher ownership concentration and firms with CEO duality are more likely to
switch to a smaller auditor, and this downward switching reduces audit quality and allows the
controlling shareholder to sustain opaque gains. In a similar vein, Lin and Liu (2009b) use
audit firm size as a proxy for audit quality and investigate the determinants of auditor choice
in China in respect of their corporate governance mechanisms. They find that firms with
higher ownership concentration and with smaller supervisory board or CEO duality are less
likely to hire a higher-quality audit firm. Lin et al. (2009) study how investors respond to
audit quality and auditor switches in the Chinese context. Their findings suggest that firms
with positive abnormal earnings, good audit quality and upward auditor switching are
positively correlated with firms‟ earnings response coefficients. Liu et al. (2011) study the
effects of the two types of connection (firm-level connections derived from state ownership
and personal connections developed through management affiliations with the external
8
auditor) and their joint effect on audit quality. Their results suggest that state ownership and
management affiliations with the external auditor impair auditor independence.
However, as alluded to earlier, these studies lack reflection of the full range of IGMs
within the Chinese context. For instance, proxies for the proportions of state and foreign
ownership, board independence, board of directors and supervisory board activity, and the
proportion of professional supervisors are absent from Lin and Liu (2009a) and Lin and Liu
(2009b); proxies for the proportion of foreign ownership, board size and independence, board
of directors and supervisory board activity and the proportion of professional supervisors are
absent from Liu et al. (2011); proxies for the proportion of foreign ownership, board
independence, board of directors and supervisory board activity and the proportion of
professional supervisors are absent from Chan et al. (2007). In order to provide a
comprehensive empirical evaluation of the independent and interdependent effects of
different IGMs, this paper investigates how they jointly affect audit quality in the Chinese
context.
2.1 Ownership structure
A dominant feature of concentrated ownership by the state is the non-tradable nature of the
equity ownership, which is held either through direct investment or indirectly through
holdings by domestic institutions. These institutions are entirely or partially owned by either
China‟s central government or its provincial governments. Thus, China‟s privatisation
program is different from that of many other transition economies. In China, state
shareholders are recognised as the controlling shareholders and often seek objectives other
than efficiency or profitability. For example, they may place a high priority on maintaining
social order and affecting wealth redistribution that may favour increasing employment,
rather than considerations of efficiency or profitability for public shareholders (Xu and Wang,
9
1999). Lin and Liu (2009a) argue that the role of auditing in minimising principal-principal
agency costs does not exist in the context of China because the controlling shareholders are
also seized with the power to appoint the auditor. Thus there is motivation to select low-
quality auditors in order to transfer benefits among related parties maintaining their
operations and expropriating the interests of minority shareholders (Felo et al., 2003).
The objective of foreign investors is to maximize profits and their shareholder wealth.
Most of these foreign investors are financial institutions based in developed economies such
as those in Europe, North America, Japan and Hong Kong. They thereby have resources to
analyse firm performance and have experience as well as capability to effect operational and
management changes when profitability and efficiency are poor (Chen et al., 2006b).
Moreover, foreign investors can play active and positive roles in bringing about
improvements in corporate governance, such as by appointing a high-quality auditor. In light
of the above discussion, the first set of hypotheses is formed as follows:
H1a. Ceteris paribus, state ownership is associated with lower audit quality.
H1b. Ceteris paribus, foreign ownership is associated with higher audit quality.
2.2 Board of directors
China has adopted a two-tier board system as a means to promote better governance. This
choice was made in the early 1990s, partly because many directors and their enterprises were
perceived to be engaged in opaque related-party transactions (Jian and Wong, 2010). The
Code, issued in January 2002 by the CSRC and the SETC, reinforces the role that the two
boards (i.e., board of directors and supervisory board) are supposed to play in corporate
governance. It gives particular attention to aspects of the boards. Since 2003 at least one-third
of the directors on the board of directors has been required to be independent. Independence
is required from both the listed company that appoints them and its major shareholders. The
10
Code also requires that for these directors, their role in the listed company is limited to that of
an independent director. Independence is argued to be important due to the behavioural
motivations that flow from it, or rather from the lack of it. From this perspective, independent
directors work in the best interests of the minority shareholders in order to maintain their own
good reputation in society (Fama and Jensen, 1983). This suggests that both larger boards and
those with a higher proportion of independent directors will have more individuals possessing
these incentives, improving the effectiveness of corporate governance and transparency of
disclosures in a sensitive area such as related party transactions.
Prior literature suggests that independent directors are effective in monitoring the
performance of management because they do not have a financial interest in the company in
the form of shares or psychological ties to management. As such, they are expected to
challenge management objectively and support the auditor (Beasley, 1996; Carcello et al.,
2002; Klein, 2002; Abbott et al., 2003; Abbott et al., 2004; Bedard and Johnstone, 2004; Boo
and Sharma, 2008). For example, Carcello et al. (2002) and Abbott et al. (2003) find a
positive association between the proportion of independent directors and audit fees and infer
that higher audit quality results. Bedard and Johnstone (2004) argue that higher proportions
of independent directors on the board and audit committee are more likely to provide vigilant
oversight of the financial reporting process, and their result also supports this argument.
To extend this argument, board meetings can indicate the level of diligence exercised
by directors (Zaman et al., 2011) and is higher activity expected to work as a good means to
discuss and solve the most widely shared problem that directors face. Hence the number of
board meetings represents an important resource in improving the effectiveness of a board
(Conger et al., 1998). Carcello et al.‟s (2002) results support higher board meeting frequency
indicating a higher level of control in the company, leading to higher audit fees. Based on a
consideration of above arguments, the second set of hypotheses is formed as follows:
11
H2a. Ceteris paribus, the size of the board of directors is positively related to audit
quality.
H2b. Ceteris paribus, the number of independent directors on the board of directors is
positively related to audit quality.
H2c. Ceteris paribus, the frequency of meetings of the board of directors is positively
related to audit quality.
2.3 Supervisory board
Dahya et al. (2003) outline the range of competencies required for a supervisory board to
effectively fulfil its stated roles. The Code identifies four distinct types of roles that
supervisory boards may undertake, depending on the independence and competencies of the
board‟s members; namely honoured guest, friendly advisor, censored watchdog, or
independent watchdog. The role of supervisory board independent watchdog requires that
members on the supervisory board have the necessary competencies in terms of knowledge
and experience to act with expertise and sufficient independence. Logically, this type of
supervisory board has a larger number of members with appropriate professional knowledge
or work experience and hence should be in a better position to improve corporate governance.
Chen (2005) argues that larger supervisory boards enhance their monitoring role. Her finding
confirms a positive correlation between the size of the supervisory board and the level of
corporate governance in Chinese companies. Lin and Liu (2009b) use the size of the
supervisory board as a proxy to monitor the strength of this board. Their finding indicates that
if there is a large supervisory board that is relatively independent from the board of directors
and the management, the firm is motivated to engage a higher-quality auditor to enhance the
supervision or monitoring role. Furthermore, frequent meetings of the supervisory board are
expected to improve the effectiveness and efficiency of the board (Conger et al., 1998), and
12
are more likely to reduce opaque related-party transactions. Accordingly, the third set of
hypotheses is formed as follows:
H3a. Ceteris paribus, the size of the supervisory board is positively related to audit
quality.
H3b. Ceteris paribus, the number of supervisors with professional knowledge or work
experience on the supervisory board is positively related to audit quality.
H3c. Ceteris paribus, the frequency of board meetings of supervisory boards is
positively related to audit quality.
3. Research method
3.1 Sample and data
The data are derived from two main sources. First, the CSRC requires that all listed
companies in China publish information regarding their stock issues, half-year reports, an
annual report and reports in respect of important events. Annual reports are chosen for this
paper as both financial and non-financial data (i.e., ownership structure and board
composition) can be extracted from them. Second, the remaining financial data are sourced
from the China Stock Market Finance Database (CSMAR-A) and the Trading Database
(CSMAR-T) produced by the Shenzhen Guotaian Information Technology Co., Ltd.
This paper focuses on the non-financial sector A-share companies listed on either the
SHSE or the SZSE. In order to test the effects of various types of ownership (i.e., high levels
of state and foreign investor ownership), the sample of companies is divided into three groups:
A-share, AB-share, and AH-share companies. A-share companies are companies that have
issued A-shares1 only, and are listed on the domestic stock exchanges. AB-share companies
are those that have issued both A-shares and B-shares,2 with an initial A-share offering. They
are also listed on the domestic stock exchanges in China. However, AH-share companies are
13
those that have issued both A-shares and H-shares3
and have floated their shares
simultaneously on the Hong Kong Stock Exchange and one of China‟s two mainland stock
exchanges.
INSERT FIGURE I ABOUT HERE
According to the rule-of-thumb of Green (1991),4
a sample size of 117 listed
companies was selected from the listed companies included in China‟s Shanghai SSE1805
and Shenzhen SSE1006 for the period from 2001 to 2005. This was achieved through use of a
stratified sampling method. As shown in Figure I, 45 companies7 were randomly selected
from the A-share group, and 42 companies were randomly selected from the AB-share group.
There were only 30 eligible companies listed on both the Hong Kong Stock Exchange and
one of the two mainland Chinese stock exchanges in 2005. For this reason, all of these
companies were selected for the sample.
INSERT TABLE I ABOUT HERE
As shown in Table I Panel A, the sample consists of 117 companies listed on the
SHSE and the SZSE in 2005 (45 A-share, 42 AB-share and 30 AH-share companies),
representing weights of 38.46%, 35.9% and 25.64%, respectively, in the sample. Table II
Panel B shows that the panel dataset is unbalanced because some companies were listed after
2001. Over the five-year period from 2001 to 2005, there were 97, 102, 109, 115 and 117
listed companies within the sample frame. This results in the sample data set having a total
of 540 firm-year observations.
3.2 Dependent variable
DeAngelo (1981) proposes that the independence of an auditor has positive correlation with
the probability of the auditor finding and reporting misstatements or irregularities in the
financial statements (i.e., providing a high-quality audit). Independence is particularly
14
relevant to larger auditors, such as members of Big 4 which want to protect their reputation
and avoid costly litigation (Francis and Krishnan, 1999; Francis and Yu, 2009). Prior studies
have substituted audit firm size as a common surrogate for audit quality (e.g., DeAngelo,
1981; Moore and Scott, 1989; Becker et al., 1998; Lennox, 1999, 2005; Kane and Velury,
2004; Lee et al., 2004; Mansi et al., 2004; Lin and Liu, 2009b; Lennox and Pittman, 2010)
because large auditors possess a higher degree of independence and expertise. The Big 4
audit firms is used as the proxy for audit quality (AQ) in this paper.8 The dichotomous
dependent variable–AQ is coded as 1 if the firm is audited by Big 4 audit firm, otherwise
coded as zero.
3.3 Independent variables
The independent variables used to examine the factors that are expected to be associated with
audit quality are divided into three main categories. The first category, ownership structure,
consists of state ownership (STATE), which represents the proportion of shares held by the
state and foreign ownership (FOREIGN), which measures the proportion of shares held by
foreign investors. The second category, board of directors, comprises BSIZE, which
represents the number of directors on the board of directors; INPD, which represents the
number of independent directors on the board of directors and BDMEET, which represents
the number of meetings held by the board of directors in the fiscal year. The third category,
supervisory board characteristics, consists of SBSIZE, which represents the number of
supervisors on the supervisory board, PROFSB which represents the number of supervisors
with professional knowledge or work experience and SBMEET, which represents the number
of meetings held by the supervisory board in the fiscal year.
15
3.4 Control variables
Control variables include return on assets (ROA), Tobin‟s Q (TOBINSQ), firm size (FSIZE)
and firm age (AGE). ROA is equal to a fiscal year's net income divided by total assets, which
provides an indication of how efficient management is at using its assets to generate earnings.
Tobin‟s Q (TOBINSQ) is equal to the market value of stock and the book value of debt
divided by the book value of total assets, which reveals the value investors assign to a firm‟s
tangible and intangible assets based on predicted future revenue and cost streams. Firm size
(FSIZE) is measured by the natural logarithm of total assets, which is often found to have a
significant impact on IGMs (e.g., Chan et al., 2007; Boo and Sharma, 2008). Firm age (AGE)
measures the number of years since initial listing.
In addition, dummy variables representing year effects (YEARDUMMY) and
controlling for listing status (SHAREDUMMY) are included in alternative analyses.
3.5 Model specification
Based on the three sets of hypotheses presented in this paper and the dichotomous nature of
the dependent variable, the logistic regression model to be empirically investigated is as
follows, where the variables are as defined the previous section:
it
4
1l
2
1kkl
4
1jitjit8it7
it6it5it4it3it2it1it
SHAREDUMMYYEARDUMMYCONTROLSSBMEETPROFSB
SBSIZEBDMEETINDPBSIZEFOREIGNSTATEAQ
This model is useful when examining the correlation between the probability of
employing an auditor assumed to provide a greater likelihood of high audit quality and the
factors of IGMs as follows:
1)1()Pr('
itjjx
it ex
16
where, )Pr( itx is the probability that the observation firm i will fail at time t; itjx is a 1j vector
of predictor observations for the ith observation at time t; 'j is a j1 vector of coefficient
estimates.
As shown in Table II Panel A, there are no Spearman correlations between
independent variables that reach .8. However, multicollinearity may still be a concern, even
where the bivariate correlation coefficients are low (Gujarati, 2003). Hence, the degree of
multicollinearity is examined through estimation of Variance Inflation Factors (VIF).9 The
results, reported in Table II Panel B, highlight that the largest VIF is 1.97 while the remainder
are below 1.68. Thus, there is no evidence of a serious multicollinearity problem being
present in the regression model.
INSERT TABLE II ABOUT HERE
4. Results and discussion
4.1 Descriptive statistics
The descriptive statistics presented in Table III Panel A provide a longitudinal profile of the
key variables. With respect to ownership structure, state ownership (STATE) displays a
declining pattern from 68%, 67%, 66%, 62% and 61% during 2001 to 2005 respectively. The
mean ratio of STATE is 65%, which indicates that state ownership continues to maintain a
dominant role. In contrast, the mean foreign ownership (FOREIGN) is only 14%. With
respect to the board of directors, the means of board size (BSIZE) and number of independent
directors (INDP) are 10.41 and 2.93 respectively, which indicates that the mean proportion of
independent directors is around 28.15%. In more detailed, this proportion can be calculated
each year as 9.87% in 2001, 24.55% in 2002, 33.21% in 2003, 34.58% in 2004 and 34.79%
in 2005. These proportions suggest that appointing independent directors was relatively
uncommon before 2002. However, a rapid growth ratio of independent directors occurred
17
between 2001 and 2003, but this growth virtually ceased between 2003 and 2005. Hence the
sample companies only just satisfy the minimum one-third ratio nominated by The Guidelines.
Examining the supervisory board, the means for supervisory board size and professional
supervisors are 4.59 and 1.86 respectively, and these numbers are well below the average size
of board of directors and number of independent directors. Regarding the control variables–
ROA and TOBINSQ, the means are .08 and 1.16 respectively. The mean TOBINSQ shows
that the market value of equity of the sampled companies is high due to surging stock prices.
I suggest that these listed companies utilize the stock market as a means to benefit the
controlling shareholders and expropriate the interests of minority shareholders, which is
indicated by the poor ROA.
INSERT TABLE III ABOUT HERE
4.2 Regression analyses
Table IV provides the regression results of logistic models that examine the three categories
of IGMs concurrently. Results are provided for analyses that exclude year and share-type
effects [model (1)], include year effect but exclude share-type effect [model (2)], include
share-type effect but exclude year effect [model (3)] and include all variables [model (4)].
The model fit is good for each model, reporting pseudo R2s of 16.17%, 16.6%, 26.83% and
26.98% respectively. The Wald 2 statistics for all models indicate that statistically
significant components of the variation in the chosen measure of AQ are explained by
variation in the set of independent variables.
INSERT TABLE IV ABOUT HERE
Regarding the first set of hypotheses, the coefficients for state ownership (β1) are
insignificant at the 5% level for all models. Thus, Hypothesis 1a is not supported. In the
context of China, state shares are still retained by state asset management bureaus (AMBs) or
their agencies, and are not usually allowed to be publicly traded in the market. This highly
18
concentrated ownership (65% on average) creates principal-principal agency problems. The
principal owner is the state, representing all Chinese people (Chan et al., 2007). The
management or agents represents multi-level AMBs (i.e., state AMBs, provincial AMBs, and
their agents). According to agency theory, management or agents will serve the best interests
of the owners. Due to lack of a clear party of ownership, no management or agent has an
adequate motivation to achieve profit maximization for the principal owner – all Chinese
people (Chan et al., 2007). From the results I infer that the controlling party for state shares
has no incentive to engage a high-quality auditor for the company.
The coefficients for foreign ownership (β2) are positively significant at the .1% level
for models (1) and (2), and at the 10% level for models (3) and (4). These findings are
consistent with Hypothesis 1b and suggest that the higher the proportion of equity owned by
foreign investors, the more likely the engagement of a high-quality auditor. I infer a
motivation to secure benefits and not have them expropriated by other parties.
In respect of the second set of hypotheses, the coefficients for board size (β3) are
insignificant at the 5% level for all models. Thus, Hypothesis 2a is not supported. The state/
provincial AMB or their agency relies on control over the board of directors to preserve their
property rights. But arguably most directors actually represent the interests of the state
because they are appointed and remunerated by the various levels of government according to
their administrative rankings rather than their ability (Xu and Wang, 1999; Zhou and Wang,
2000). Hence they may have insufficient managerial ability to monitor management‟s
behaviour. Furthermore, the state, through its control over the board of directors, may seek
objectives other than profit maximization and place a priority on maintaining social order or
reducing unemployment (Xu and Wang, 1999).
The coefficients for the number of independent directors (β4) are insignificant at the
5% level for all models. Thus, Hypothesis 2b is not supported. In the context of China, the
19
key problem of board independence is related to the directors and managers. Most directors
are insiders or executive directors, and listed companies on average only just meet the one-
third ratio of independent directors required by The Guidelines since 2003. Most independent
directors are appointed by the board members or controlling shareholders, hence they owe
their loyalty and patronage to their appointers (Tan and Wang, 2007). As a consequence, the
independent directors of Chinese listed companies may demonstrate limited independence
and easily become „window dressing‟ or „symbolic‟ as board members (Luan and Tang,
2007).
The coefficients for board meeting activity (β5) are insignificant at the 5% level on all
models. Thus, Hypothesis 2c is not supported. This finding is not consistent with Conger et al.
(1998) and Carcello et al. (2002). The frequency of board meetings may be attributable to
many reasons, such as financial distress or controversial decisions on illegal or questionable
activities (Vafeas, 1999; Chen et al., 2006a) and does not necessarily work as a means to
enhance corporate governance.
Regarding the third set of hypotheses, the coefficients for supervisory board size (β6)
are negative and hence opposite to the expected sign as I hypothesized and are inconsistent
with the findings of Lin and Liu (2009b). Hence even though they are significant at the 10%
level in model (1), at the 5% level in model (2) the significances cannot be used to support
Hypothesis 3a. I speculate that the supervisory boards of Chinese listed companies have
tended to become „censored watchdogs‟ in the words of Dahya et al. (2003) during a period
when rapid corporate expansion and the dominance of the board of directors has occurred.
Xiao et al. (2004, p. 47) concluded that supervisory board members describe their behaviour
as where “You sometimes have to open one eye and close the other”. If the „censored
watchdog‟ is the role of the supervisory board, then perhaps what is occurring is that the
more qualified the supervisory board members are, the greater the internal effort made to
20
censor or review works about earnings management and tunneling, which could include
information deemed sensitive to the board of directors and controlling shareholders. In reality,
these supervisory board members could be regarded as „rubber stamp‟ supervisors.
The coefficients for professional supervisors on the supervisory board (β7) are
positively significant at the 1% level in models (1), (2) and (4), and at the 5% level in model
(3). Thus, Hypothesis 3b is supported. This finding suggests that professional supervisors are
relatively independent from management and the board of directors (Lin and Liu, 2009b), and
advocate for high-quality auditors as a means to improve the corporate governance of listed
companies.
The coefficients for supervisory board meetings (β8) are insignificant at the 5% level
on all models. Thus, Hypothesis 3c is not supported. Once more, as with the frequency of the
meetings of board of directors, these results confirm the inferences of Vafeas (1999) and
Chen et al. (2006a) reveal the reason for companies having more board meetings during
periods of financial distress or during times of controversial decisions.
In terms of control variables, the coefficients for ROA are positively significant at the
10% in model (3) and at the 5% level in model (4), but insignificant in models (1) and (2).
The coefficients for TOBINSQ are positively significant at the 1% level in model (1) and at
the 5% level in model (2), but insignificant in models (3) and (4). These findings suggest that
listed companies with better profitability and market performance are more likely to hire a
high-quality auditor. The coefficients for FSIZE are positively significant at least the 5%
level in all models. The coefficients for AGE are positively significant at the 10% level in
model (1) and at the 5% in model (2), but insignificant in models (3) and (4). These findings
suggest that listed companies of larger size or longer listing status are more likely to hire a
high-quality auditor.
21
Concerning year and share-type effects, the results indicate that there are no year
effects, and positively significant share-type effect in models (3) and (4). This suggests that
listed companies that issue both A- and B-shares or A- and H-shares have better are more
likely to choose a high-quality auditor than listed companies that issues A-shares only.
4.3 Sensitivity test
A sensitivity test is used as an alternative technique to examine the regression model and the
empirical results. According to composition of the sample presented in Panel A Table I, I
perform the regression model used in Table IV to investigate the association of the IGMs on
each of the three share-types separately. These findings may differ from the main results
because the proportions of ownership and monitoring power of the board of directors and
supervisory board of each listing status are diversified. The smaller sample sizes need to be
considered in interpreting the results of this secondary analysis.
INSERT TABLE V ABOUT HERE
For the companies that issue A-shares only, as shown in Table V, the results are
consistent with the findings in Table IV, except for insignificant coefficients for foreign
ownership on models (1) and (2). This finding suggests that foreign ownership in listed
companies with A-shares is only 2% in comparison with 75% of state ownership (as shown in
Panel B of Table III), and this small stake reduces the efficiency of foreign investors to
actively enhance corporate governance (Chen et al., 2006b).
Regarding the companies that issue A- and B-shares, as shown in Table V, the results
are consistent with the findings in Table IV, except for a positively significant coefficient for
independent directors in model (4), a negatively significant coefficient for board meetings in
model (4), and negatively significant coefficients for supervisory board size in models (3) and
(4). The positively significant coefficient for independent directors adds weight to the agency
22
argument of (Fama and Jensen, 1983) that independent directors of these companies are
motivated to work in the best interests of shareholders in order to maintain their good
personal reputation, and also supports Chen and Jaggi‟s (2000) findings that independent
directors will improve corporate governance even in countries that have fewer incentives for
transparency, such as China. The negatively significant coefficients for board meetings and
supervisory board size suggest that the IGMs of board of directors and supervisory board are
ineffective in enhancing corporate governance, and that the effectiveness of corporate
governance can be improved by increasing the proportion of foreign equity.
In respect of the companies that issue A- and H-shares, as shown in Table V, the
results are consistent with the findings in Table IV, except for negatively significant
coefficients for state ownership in models (5) and (6), positively significant coefficients for
independent directors and supervisory board size in model (5) and insignificant coefficients
for foreign ownership in models (5) and (6). The negative coefficients for state ownership are
consistent with the findings of Lin and Liu (2009a) suggesting controlling shareholders are
less motivated to hire high-quality auditors in order to transfer benefits among related parties.
The discussion in the above two paragraphs in terms of the positive coefficient for
independent directors and the insignificant coefficient for foreign ownership can be applied to
the findings of AH-share companies.
5. Conclusion
The objective of this paper is to investigate whether the audit quality of Chinese listed
companies is associated with IGMs. I used a panel data set comprising 117 companies with a
total of 540 firm-year observations between 2001 and 2005. The results suggest that foreign
ownership and the number of professional supervisors are positively associated with audit
quality, however the size of the supervisory board revealed an unexpected negative
23
association. Other IGMs including state ownership, the size of board of directors, the number
of independent directors, the frequency of board meetings, and the frequency of supervisory
board meetings were found to have no association with audit quality.
At least two implications can be drawn from this paper. First, I believe that the
corporate governance reforms regarding the board of directors and the supervisory board of
Chinese listed companies have not yet been sufficient to ensure that they properly fulfil their
roles of oversight. Rather, they are perhaps playing more a „window dressing‟ or „rubber
stamp‟ or „symbolic‟ role within the current two-tier corporate governance system. Second,
despite recent reforms, such as mandating the introduction of independent directors and
minority foreign ownership after China‟s entrance into the WTO in 2001, the dominance and
influence of state ownership has not changed. Highly concentrated state ownership appears to
have resulted in ineffective IGMs among Chinese listed companies. For example, the
supervisory boards consist of a high proportion of insiders who may be political officers of a
certain level of a government organ, leaders of a non-functional trade union, or close friends
and allies of senior executives. These supervisors, appointed by the senior executives who
also determine their compensation, have virtually no independence from management. In
order to overcome this problem, Tam (2000) suggests that the proportion of inside
supervisors be kept below 20%, and a set of functional committees, such as nomination,
remuneration and audit, should be established to remove some of the grounds on which
insider control has been built.
Other than these innovations, changes could be made to The Company Law and The
Code. The supervisory board could be given greater power to hire and remove directors. This
authority would improve the supervisor‟s status within and outside the company. Moreover,
The Company Law could prescribe legal responsibilities for the supervisory board so that
supervisors would be statutorily obliged to undertake their duties diligently. On the other
24
hand, The Company Law could clarify an obligation of the board of directors to provide
supervisors with timely information regarding the firm so that the supervisory board can
effectively exercise its monitoring function.
There are several limitations of this paper. The first limitation is that I only employed
a dichotomous dependent variable (Big 4 audit firm) as the proxy for audit quality due to
limited data. However, audit quality can also be surrogated by other measures, such as audit
fees (O'Sullivan, 2000; Boo and Sharma, 2008; Zaman et al., 2011), auditor switching (Lin
and Liu, 2009a; Lin et al., 2009) and modified audit opinions (Liu et al., 2011). The second
limitation regards the data set used for this paper. Although the sample was selected from the
Shanghai SSE 180 and the Shenzhen SSE 100, it may not be generalizable to all listed
companies in China. Additionally the proxies for IGMs may not differentiate well the
characteristics of sound corporate. Due to these limitations, the findings of this paper should
be interpreted with care. Future research can be undertaken that improves the degree of
confidence by extending sample sizes in the coming years and by introducing better and
additional proxies for audit quality such as audit fees and audit switching.
Despite these limitations, this is the first study to provide a comprehensive empirical
analysis of the key parameters that underlie China‟s IGMs impact on audit quality during the
important period of regulatory change and organizational reform between 2001 and 2005.
This paper also fills a gap in the literature concerning the impact on audit quality choice of
high state ownership concentration and the two-tier board corporate governance system.
Notes 1A-shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the
mainland stock exchanges, and are reserved for trading by Chinese citizens. The A-share market was launched
in 1990 in Shanghai. 2 B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock
exchanges. The B-share market was launched in 1992 and was restricted to foreign investors before 19 February
2001.
25
3
H-shares are securities of companies incorporated in mainland China and nominated by the Chinese
government for listing and trading on the Hong Kong Stock Exchange, being quoted and traded in HKD. There
are no restrictions on holdings by international investors. 4 Green (1991) suggests that the power for a test of a multiple correlation with a medium effect size is
approximately 0.80 if N ≥ 50 + 8m, where N is number of subjects, and m is the number of predictors. Thus, the
sample size should be at least 114 (N = 50 + 8 x 8). 5 The Shanghai SSE180 Index was created by restructuring and renaming the SSE30 Index. Through scientific
and objective methods it selects constituents that best represent the market. The SSE is a benchmark index
reflecting the Shanghai market and serves as a performance benchmark for investment and a basis for financial
innovation. 6 The Shenzhen SSE100 is a benchmark index reflecting performance in the Shenzhen market and serves as a
performance benchmark for investment and as a basis for development of financial innovations. 7 These samples are randomly selected from A-share companies included in the Shanghai SSE180 and the
Shenzhen SSE100 after removing dual listed companies (these being either AB-share companies or AH-share
companies). 8According to Chinese government regulation, the international Big 4 accounting firms, i.e., PwC, KPMG,
Deloitte and Ernst & Young, were not permitted to provide auditing services before 2005 in China unless they
became joint-ventures or partnerships with large domestic Chinese CPA firms. Thereby, the Big 4 audit firms in
this paper comprise PwC ZhongTian, KPGM Huazhen, Deloitte Huayong and Ernst & Young Hua Ming. 9 The critical value of the VIF to test for multicollinearity is 10. Gujarati (2003) suggests that there is no
evidence of multicollinearity unless the VIF of a variable exceeds 10. All values used in this paper are well
below this critical level.
26
References
Abbott, L.J., Parker, S. and Peters, G.F. (2004), “Audit committee characteristics and
restatements”, Auditing: A Journal of Practice and Theory, Vol. 23 No.1, pp. 69–87.
Abbott, L.J., Parker, S., Peters, G.F and Raghunandan, K. (2003), “The association between
audit committee characteristics and audit fees”, Auditing: A Journal of Practice and
Theory, Vol. 22 No.2, pp. 17–32.
Al-Ajmi, J. (2009), “Audit firm, corporate governance, and audit quality: evidence from
Bahrain”, Advances in Accounting, Vol. 25 No. 1, pp. 64–74.
Beasley, M.S. (1996), “An empirical analysis of the relation between the board of director
composition and financial statement fraud”, The Accounting Review, Vol. 71 No. 4,
pp. 443–465.
Becker, C.L., DeFond, M.L., Jiambalvo, J. and Subramanyam, K.R. (1998), “The effect of
audit quality on earnings management”, Contemporary Accounting Research, Vol. 15
No. 1, pp. 1–24.
Bedard, J.C. and Johnstone, K.M. (2004), “Earnings manipulation risk, corporate governance
risk, and auditors' planning and pricing decisions”, The Accounting Review, Vol. 79
No. 2, pp. 227–304.
Boo, E. and Sharma, D. (2008), “Effect of regulatory oversight on the association between
internal governance characteristics and audit fees”, Accounting and Finance, Vol. 48
No. 1, pp.51–71.
Cadbury Report. (1992), Report of the Committee on the Financial Aspects of Corporate
Governance: The Code of Best Practice, Gee Professional Publishing, London.
Carcello, J.V., Hermanson, D.R., Neal, T.L. and Riley Jr., R.A. (2002), “Board characteristics
and audit fees”, Contemporary Accounting Research, Vol. 19 No. 3, pp. 365–384.
Chan, K.H., Lin, K.Z. and Zhang, F. (2007), “On the association between changes in
corporate ownership and changes in auditor quality in a transitional economy”,
Journal of International Accounting Research, Vol. 6 No. 1, pp. 19–36.
Chen, C.J.P. and Jaggi, B. (2000), “Association between independent non-executive directors,
family control and financial disclosures in Hong Kong”, Journal of Accounting and
Public Policy, Vol. 19 No. 4/5, pp. 285–310.
Chen, G., Firth, M., Gao, D.N. and Rui, O.M. (2006a), “Ownership structure, corporate
governance, and fraud: evidence from China”, Journal of Corporate Finance, Vol. 12
No. 3, pp.424–448.
Chen, G., Firth, M. and Rui, O.M. (2006b), “Have China's enterprise reforms led to improved
efficiency and profitability?”, Emerging Markets Review, Vol. 7 No.1, pp. 82–109.
27
Chen, J.J. (2005), “China's institutional environment and corporate governance”, Corporate
Governance: A Global Perspective Advances in Financial Economics, Vol. 11, pp.
75–93.
Conger, J.A., Finegold, D. and Lawler III, E.E. (1998), “Appraising boardroom performance”,
Harvard Business Review, Vol. 76 No. 1, pp. 136–148.
Dahya, J., Karbhari, Y., Xiao, J.Z. and Yang, M. (2003), “The usefulness of the supervisory
board report in China”, Corporate Governance: An International Review, Vo. 11 No.
4, pp. 308–321.
DeAngelo, L.E. (1981), “Auditor size and audit quality”, Journal of Accounting and
Economics, Vol. 3 No.3, pp. 183–199.
DeFond, M.L. (1992), “The association between changes in client firm agency costs and
auditor switching”, Auditing: A Journal of Practice and Theory, Vo. 11 No. 1, pp.16–
31.
DeFond, M.L., Wong, T.J. and Li, S. (2000), “The impact of improved auditor independence
on audit market concentration in China”, Journal of Accounting and Economics, Vol.
28 No. 3, pp. 269–305.
Denis, D.K. and McConnell, J.J. (2003), “International corporate governance”, Journal of
Financial and Quantitative, Vol. 38 No. 1, pp. 1–36.
Dharwadkar, B., George, G. and Brandes, P. (2000), “Privatization in emerging economies:
an agency theory perspective”, Academy of Management Review, Vol. 25 No. 3, pp.
650–669.
Fama, E.F. and Jensen, M.C. (1983), “The separation of ownership and control”, Journal of
Law and Economics, Vol. 26 No. 2, pp. 301–325.
Felo, A.J., Krishnamurthy, S. and Solieri, S.A. (2003), “Audit committee characteristics and
the perceived quality of financial reporting: an empirical analysis”, Working paper,
Penn State Great Valley.
Francis, J.R. and Krishnan, J. (1999), “Accounting accruals and auditor reporting
conservatism”, Contemporary Accounting Research, Vol. 16 No. 1, pp. 135–165.
Francis, J.R. and Yu, M.D. (2009), “Big 4 office size and audit quality”, The Accounting
Review, Vol. 48 No. 5, pp. 1521–1552.
Green, S.B (1991), “How many subjects does it take to do a regression analysis?”,
Multivariate Behavioral Research, Vol. 26 No. 3, pp. 499–510.
Gujarati, D.N. (2003), Basic Econometrics, 4th edn. McGraw-Hill, New York .
Gul, F.A. and Goodwin, J. (2010), “Short-term debt maturity structures, credit ratings, and
the pricing of audit services”, The Accounting Review, Vol. 85 No. 3, pp. 877–909.
28
Holm, C. and Laursen, P.B. (2007), “Risk and control developments in corporate governance:
Changing the role of the external auditor?”, Corporate Governance: An International
Review, Vol. 15 No. 2, pp. 322–333.
Hu, H.W., Tam, O.K. and Tan, M.G-S. (2010), “Internal governance mechanisms and firm
performance in China”, Asia Pacific Journal of Management, Vol. 27 No. 4, pp. 727–
749.
Jensen, M.C. and Meckling, W.H. (1976), “Theory of the firm: managerial behavior, agency
costs, and ownership structure”, Journal of Financial Economics, Vol. 3 No. 4, pp.
305–360.
Jian, M. and Wong, T.J. (2010), “Propping through related party transaction”, Review of
Accounting Studies, Vol. 15 No. 1, pp. 70–105.
Kane, G.D. and Velury, U. (2004), “The role of institutional ownership in the market for
auditing services: an empirical investigation”, Journal of Business Research, Vol. 57
No. 9, pp. 976–983.
Klein, A. (2002), “Audit committee, board of director characteristics, and earnings
management”, Journal of Accounting and Economics, Vol. 33 No. 3, pp. 375–400.
Lee, H.Y., Mande, V. and Ortman, R. (2004), “The effect of audit committee and board of
director independence on auditor resignation”, Auditing: A Journal of Practice and
Theory, Vol. 23 No. 2, pp. 131–146.
Lennox, C. (1999), “Non-audit fees, disclosure and audit quality”, European Accounting
Review, Vol. 8 No. 2, pp. 239–252.
Lennox, C. (2005), “Management ownership and audit firm size”, Contemporary Accounting
Research, Vol. 22 No. 1, pp. 205–227.
Lennox, C. and Pittman, J.A. (2010), “Big five audits and accounting fraud”, Contemporary
Accounting Research, Vol. 27 No. 11, pp. 209–247.
Lin, Z.J. and Liu, M. (2009a), “The determinants of auditor switching from the perspective of
corporate governance in China”, Corporate Governance: An International Review,
Vol. 17 No. 4, pp. 476–491.
Lin, Z.J. and Liu, M. (2009b), “The impact of corporate governance on auditor choice:
evidence from China”, Journal of International Accounting, Auditing and Taxation,
Vol. 18 No. 1, pp. 44–59.
Lin, Z.J., Liu, M. and Wang, Z. (2009), “Market implications of the audit quality and auditor
switches: evidence from China”, Journal of International Financial Management and
Accounting, Vol. 20 No. 1, pp: 35–78.
Liu, J., Wang, Y. and Wu, L. (2011), “The effect of Guanxi on audit quality in China”,
Journal of Business Ethics, Vol. 103 No. 4, pp. 621–638.
29
Luan, C-J. and Tang, M-J. (2007), “Where is independent director efficacy?” Corporate
Governance: An International Review, Vol. 15 No. 4, pp. 636–643.
Mansi, S.A., Maxwell, W.F. and Miller, D.P. (2004), “Does auditor quality and tenure matter
to investors? evidence from the bond market”, Journal of Accounting Research, Vol.
42 No. 4, pp.755–793.
Moore, G. and Scott, W.R. (1989), “Auditors' legal liability, collusion with management, and
investors' loss”, Contemporary Accounting Research, Vol. 5 No. 2, pp. 754–774.
O'Sullivan, N. (2000), “The impact of board composition and ownership on audit quality:
Evidence from large UK companies”, British Accounting Review, Vol. 32 No. 4, pp.
397–414.
Shleifer, A. and Vishny, R.W. (1997), “A survey of corporate governance”, Journal of
Finance, Vol. 52 No. 2, pp. 737–782.
Tam, O.K. (2000), “Models of corporate governance for Chinese companies”, Corporate
Governance: An International Review, Vol. 8 No. 1, pp. 52–64.
Tan, L.H. and Wang, J. (2007), “Modelling an effective corporate governance system for
China's listed state-owned enterprises: Issues and challenges in a transitional economy”,
Journal of Corporate Law Studies, Vol. 7 No. 1, pp. 143–183.
Vafeas, N. (1999), “Board meeting frequency and firm performance”, Journal of Financial
Economics, Vol. 53 No. 1, pp. 113–142.
Watts, R.L. and Zimmerman, J.L. (1986), Positive Accounting Theory, Prentice-Hall:
Englewood Cliffs, NJ.
Xiao, J.Z., Dahya, J. and Lin, Z. (2004), “A grounded theory exposition of the role of the
supervisory board in China”, British Journal of Management, Vol. 15 No. 1, pp. 39–
55.
Xu, X. and Wang, Y. (1999), “Ownership structure and corporate governance in Chinese
stock companies”, China Economic Review, Vol. 10 No. 1, pp. 75–98.
Young, M.N., Peng, M.W., Ahlstrom, D., Bruton, G.D. and Jiang, Y. (2008), “Corporate
governance in emerging economies: a review of the principal-principal perspective”,
Journal of Management Studies, Vol. 45 No. 1, pp. 196–220.
Zaman, M., Hudaib, M. and Haniffa, R. (2011), “Corporate governance quality, audit fees
and non-audit services fees”, Journal of Business Finance and Accounting, Vol. 38
No. 1/2, pp. 165–197.
Zerni, M., Kallunki, J-P. and Nilsson, H. (2010), “The entrenchment problem, corporate
governance mechanisms, and firm value”, Contemporary Accounting Research, Vol.
27 No. 4, pp. 1169–1206.
30
Zhou, M. and Wang, X. (2000), “Agency cost and the crisis of China's SOE”, China
Economic Review, Vol. 11 No. 3, pp. 297–317.
31
FIGURE I. Sampling frame a, b, c, d
H-Share firms
B-Share firms
AH-Share firms 30
selected for sample
AB-Share firms 42
selected for sample
A-Share firms 45 selected for sample
from Shanghai
SSE180 and Shenzhen SSE100
Notes: a A-shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the
mainland stock exchanges, and are reserved for trading by Chinese citizens. The A-share market was launched
in 1990 in Shanghai. b B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock
exchanges. The B-share market was launched in 1992 and was restricted to foreign investors before 19 February
2001. AB-share companies in this paper indicate those that have issued both A-shares and B-shares, with an
initial A-share offering. They are also listed on the domestic stock exchanges in China. c H-shares are securities of companies incorporated in mainland China and nominated by the Chinese
Government for listing and trading on the Hong Kong Stock Exchange, being quoted and traded in Hong Kong
Dollar. There are no restrictions on holdings by international investors. AH-share companies in this study
indicate those that have issued both A-shares and H-shares, and have floated their shares simultaneously on the
Hong Kong Stock Exchange and one of China‟s two mainland stock exchanges. d
The Shanghai SSE180 Index was created by restructuring and renaming the SSE30 Index. Through scientific
and objective methods it selects constituents that best represent the market. The SSE is a benchmark index
reflecting the Shanghai market and serves as a performance benchmark for investment and a basis for financial
innovation. The Shenzhen SSE100 is a benchmark index reflecting performance in the Shenzhen market and
serves as a performance benchmark for investment and as a basis for development of financial innovations.
32
Table I. Sample and observation a, b, c
Panel A: Composition of sample
Share Type No. in Sample Percentage of
Sample
Number of firm-
year observations
Percentage of firm-year
observations
A
45 38.46 191 35.37
AB
42 35.90 210 38.89
AH
30 25.64 139 25.74
Total 117 100.00 540 100.00
Panel B: Distribution of observations by share type and year
Share Type 2001 2002 2003 2004 2005 Total
A
31 33 38 44 45 191
AB
42 42 42 42 42 210
AH
24 27 29 29 30 139
Total 97 102 109 115 117 540
Notes: a A-Shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the
mainland stock exchanges, and are reserved for trading by Chinese citizens. The A-share market was launched
in 1990 in Shanghai. b B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock
exchanges. The B-share market was launched in 1992 and was restricted to foreign investors before 19 February
2001. AB-share companies in this paper indicate those that have issued both A-shares and B-shares, with an
initial A-share offering. They are also listed on the domestic stock exchanges in China. c H-Shares are securities of companies incorporated in mainland China and nominated by the Chinese
Government for listing and trading on the Hong Kong Stock Exchange, being quoted and traded in Hong Kong
Dollar. There are no restrictions on holdings by international investors. AH-share companies in this study
indicate those that have issued both A-shares and H-shares, and have floated their shares simultaneously on the
Hong Kong Stock Exchange and one of China‟s two mainland stock exchanges.
33
Table II. Collinearity diagnostics a
Panel A: Spearman matrix
Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
(1) STATE –
(2) FOREIGN –.52***
–
(3) BSIZE –.07† .21***
–
(4) INDP –.10*
.20***
.49***
–
(5) BDMEET –.06 .13**
.03 .17***
–
(6) SBSIZE .13**
.02 .43***
.28***
.02 –
(7) PROFSB .06 .09*
.23***
.23***
.06 .62***
–
(8) SBMEET –.04 .09* –.02 .01 .29
*** .02 .01 –
(9) ROA –.05 –.04 .15***
.08† .04 .14**
.09*
–.07 –
(10) TOBINSQ –.28***
.20***
–.15***
–.30***
–.01 –.08†
–.11**
.13**
–.01 –
(11) FSIZE .05 .25***
.28***
.33***
.06 .17***
.17***
–.06 .10*
–.42***
–
(12) AGE –.17***
.14**
–.15***
.08† .05 –.06 .01 .16***
–.27***
.21***
–.14***
–
Panel B: VIF diagnostic b
STATE FOREIGN BSIZE INDP BDMEET SBSIZE PROFSB SBMEET ROA TOBINSQ FSIZE AGE
VIF 1.37 1.41 1.55 1.59 1.14 1.97 1.68 1.13 1.02 1.51 1.58 1.13
Notes: a STATE = state ownership concentration, proportion shares held by the state; FOREIGN = foreign ownership concentration, proportion of shares held by foreign investors;
BSIZE = board size, the number of directors on the board; INDP = the number of independent directors on the board; BDMEET = the number of meetings of the board of
directors in the fiscal year; SBSIZE = the number of members on the supervisory board; PROFSB = professionalism of the supervisory board, number of supervisory board
members with professional knowledge or work experience; SBMEET = the number of meetings of the supervisory board in the fiscal year; ROA = return on total assets;
TOBINSQ = Tobin‟s Q, market value of stock and book value of debt divided by book value of total assets; FSIZE: firm size, natural logarithm of value of total assets at the
end of fiscal year; AGE: firm age, years since initial listing b The critical value of the VIF to test for multicollinearity is 10. Gujarati (2003) suggests that there is no evidence of multicollinearity unless the VIF of a variable exceeds 10.
All values used in this paper were well below this critical level.
†if p<.10; * if p <.05;
** if p<.01;
*** if p<.001 (two-tailed p-values are used in determining significance)
34
Table III. Descriptive statistics a
Panel A: Statistics by year
2001 2002 2003 2004 2005 All years
Variable Mean Med SD Mean Med SD Mean Med SD Mean Med SD Mean Med SD Mean Med SD
STATE .68 .78 .29 .67 .74 .29 .66 .70 .29 .62 .68 .32 .61 .67 .32 .65 .71 .31
FOREIGN .14 .04 .16 .14 .04 .17 .14 .04 .17 .14 .04 .17 .14 .05 .17 .14 .04 .17
BSIZE 10.13 10 2.78 10.55 11 2.31 10.63 10 2.32 10.44 9 2.4 10.29 9 2.33 10.41 10 2.42
INDP 1 0 1.34 2.59 2 .94 3.53 3 .85 3.61 3 .83 3.58 3 .87 2.93 3 1.38
BDMEET 6.04 6 2.29 8.21 8 3.08 7.39 7 3.31 7.32 7 2.78 7.95 7 3.13 7.41 7 3.03
SBSIZE 4.59 5 1.56 4.6 5 1.57 4.67 5 1.8 4.6 5 1.72 4.49 5 1.69 4.59 5 1.67
PROFSB 1.67 2 .84 1.83 2 .92 1.92 2 1.07 1.91 2 1.1 1.93 2 1.17 1.86 2 1.04
SBMEET 3.46 3 1.51 4.35 4 2.02 3.83 3 2.09 3.33 3 1.74 3.42 3 1.99 3.67 3 1.92
ROA .23 .05 1.75 .03 .04 .17 .05 .05 .04 .06 .06 .07 .05 .05 .08 .08 .05 .75
TOBINSQ 1.47 1.36 .57 1.24 1.18 .41 1.14 1.09 .31 1.03 1.00 .26 .96 .93 .25 1.16 1.07 .41
FSIZE 21.74 21.64 1.04 21.88 21.79 1.02 22 21.89 1.05 22.09 22.05 1.11 22.17 22.15 1.18 21.99 21.88 1.09
AGE 5.37 6 2.73 6.06 6 2.99 6.61 7 3.38 7.21 8 3.7 8.07 9 3.82 6.73 7 3.49
Panel B: Statistics by share type
A-share b
AB-share c
AH-share d
Variable Mean Med SD Mean Med SD Mean Med SD
STATE .75 .89 .30 .59 .71 .35 .60 .60 .19
FOREIGN .02 .00 .07 .12 .06 .14 .33 .36 .12
BSIZE 10.09 9 2.51 10.08 9 2.39 11.35 11 2.1
INDP 2.76 3 1.42 2.68 3 1.44 3.53 4 1.01
BDMEET 6.81 6 2.63 7.57 7 3.28 7.99 8 3.03
SBSIZE 4.52 5 1.44 4.38 3 1.63 4.99 5 1.94
PROFSB 1.74 2 .93 1.69 2 1.03 2.27 2 1.08
SBMEET 3.3 3 1.8 3.92 3.5 2.1 3.79 4 1.7
ROA .07 .06 .05 .03 .04 .13 .17 .05 1.47
TOBINSQ 1.00 .93 .36 1.35 1.26 .43 1.08 1.06 .30
FSIZE 21.74 21.85 .77 21.61 21.63 .81 22.9 22.92 1.31
AGE 4.99 5 3.3 8.98 9 2.32 5.71 6 3.39
Notes: a STATE = state ownership concentration, proportion shares held by the state; FOREIGN = foreign ownership concentration, proportion of shares held by foreign investors;
BSIZE = board size, the number of directors on the board; INDP = the number of independent directors on the board; BDMEET = the number of meetings of the board of
directors in the fiscal year; SBSIZE = the number of members on the supervisory board; PROFSB = professionalism of the supervisory board, number of supervisory board
members with professional knowledge or work experience; SBMEET = the number of meetings of the supervisory board in the fiscal year; ROA = return on total assets;
35
TOBINSQ = Tobin‟s Q, market value of stock and book value of debt divided by book value of total assets; FSIZE: firm size, natural logarithm of value of total assets at the
end of fiscal year; AGE: firm age, years since initial listing bA-shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the mainland stock exchanges, and are reserved for trading by Chinese
citizens. The A-share market was launched in 1990 in Shanghai. c B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock exchanges. The B-share market was launched in 1992 and was
restricted to foreign investors before 19 February 2001. AB-share companies in this paper indicate those that have issued both A-shares and B-shares, with an initial A-share
offering. They are also listed on the domestic stock exchanges in China. d H-shares are securities of companies incorporated in mainland China and nominated by the Chinese Government for listing and trading on the Hong Kong Stock Exchange,
being quoted and traded in Hong Kong Dollar. There are no restrictions on holdings by international investors. AH-share companies in this study indicate those that have
issued both A-shares and H-shares, and have floated their shares simultaneously on the Hong Kong Stock Exchange and one of China‟s two mainland stock exchanges.
36
Table IV. Logistic regression results a, b
Model (1) Model (2) Model (3) Model (4)
Expected Main Year Effect Share-Type Effect Year & Share-Type Effect
sign β z β z β z β z
Constant –20.10 –6.71***
–19.55 –6.41***
–11.56 –3.47***
–11.37 –3.37***
STATE – [H1a] –.16 –.40 –.24 –.61 –.54 –1.28 –.61 –1.41
FOREIGN + [H1b] 2.38 3.53***
2.26 3.32***
1.82 1.78† 1.87 1.83†
BSIZE + [H2a] .09 1.63 .06 1.08 .05 .92 .04 .62
INDP + [H2b] –.08 –.90 .04 .29 –.09 –.85 –.04 –.27
BDMEET + [H2c] –.01 –.32 –.01 –.28 –.04 –1.17 –.05 –1.23
SBSIZE + [H3a] –.16 –1.85† –.18 –2.03*
–.14 –1.52 –.15 –1.57
PROFSB + [H3b] .39 3.07**
.40 3.12**
.36 2.57*
.36 2.58**
SBMEET + [H3c] .04 .82 .03 .45 .01 .20 –.01 –.03
ROA ? .97 .73 1.29 .92 3.35 1.91† 3.58 2.00*
TOBINSQ ? .97 3.21**
.79 2.39*
.01 .04 –.08 –.21
FSIZE ? .77 6.03***
.77 5.99***
.42 2.98**
.43 3.00**
AGE ? .06 1.99† .09 2.44*
–.03 –.79 –.02 –.42
Y2002 - - –.02 –.06 - - .20 .48
Y2003 - - –.38 –.80 - - –.10 –.19
Y2004 - - –.57 –1.14 - - –.25 –.46
Y2005 - - –.70 –1.34 - - –.24 –.43
ABSHARE - - - - 2.75 6.85***
2.72 6.77***
AHSHARE - - - - 3.57 6.97***
3.56 6.90***
Pseudo R2
16.17% 16.6% 26.83% 26.98%
Wald 2 114.4
*** 117.31
*** 189.83
*** 190.68
***
Cross-section 117 117 117 117
Observations 540 540 540 540
The table reports the results of logistic regression models:
it
4
1l
2
1kkl
4
1jitj
it8it7it6it5it4it3it2it1it
SHAREDUMMYYEARDUMMYCONTROLS
SBMEETPROFSBSBSIZEBDMEETINDPBSIZEFOREIGNSTATEAQ
where, AQ = audit quality, the dichotomous dependent variable is coded as 1 if the firm is audited by a Big 4 audit firm, otherwise coded as zero; STATE = state ownership
concentration, proportion of shares held by the state; FOREIGN = foreign ownership concentration, proportion of shares held by foreign investors; BSIZE = board size, the
number of directors on the board; INDP = the number of independent directors on the board; BDMEET = the number of meetings of the board of directors in the fiscal year;
SBSIZE = the number of members on the supervisory board; PROFSB = professionalism of the supervisory board, number of supervisory board members with professional
knowledge or work experience; SBMEET = the number of meetings of the supervisory board in the fiscal year; ROA = return on total assets; TOBINSQ = Tobin‟s Q, market
37
value of stock and book value of debt divided by book value of total assets; FSIZE: firm size, natural logarithm of value of total assets at the end of fiscal year; AGE: firm
age, years since initial listing
Notes: a In logistic regression model we coded as 1 if the firm is audited by a Big 4 auditor, otherwise coded as 0.
b Two-tailed p-values are used in determining significance: †if p<.10;
* if p <.05;
** if p<.01;
*** if p<.001
38
Table V. Logistic regression by share type a, b
A-share c AB-share
d AH-share
e
Model (1) Model (2) Model (3) Model (4) Model (5) Model (6)
Expected Main Year Effect Main Year Effect Main Year Effect
sign β z β z β z β z β z β z
Constant –31.49 –2.16*
–35.86 –2.28*
–5.55 –.99 –2.76 –.47 –5.41 –.69 –5.51 –.70
STATE – [H1a] 1.47 .63 1.69 .67 –.24 –.45 –.56 –.99 –5.77 –1.93† –6.08 –1.93†
FOREIGN + [H1b] .63 .11 1.09 .18 2.62 1.99*
2.77 2.07*
–2.73 –.73 –2.78 –.73
BSIZE + [H2a] .10 .49 .21 .86 .04 .56 –.01 –.16 .08 .57 .05 .35
INDP + [H2b] –.36 –1.23 –1.01 –1.44 .15 1.00 .40 1.69† .47 1.89† .38 1.32
BDMEET + [H2c] –.11 –.61 –.09 –.51 –.09 –1.58 –.10 –1.68† –.04 –.57 –.04 –.52
SBSIZE + [H3a] –.54 –1.12 –.61 –1.22 –.47 –2.87**
–.49 –2.93**
.29 1.66† .27 1.45
PROFSB + [H3b] 1.79 2.81**
1.83 2.81**
.76 3.14**
.83 3.33***
.44 1.61 .46 1.65†
SBMEET + [H3c] .40 1.67 .50 1.61 –.07 –.87 –.09 –1.15 .19 1.34 .20 1.36
ROA ? .18 .02 –2.66 –.27 5.73 1.92† 5.19 1.72† 8.24 2.36*
8.77 2.43*
TOBINSQ ? –.45 –.27 .09 .05 .68 1.48 .14 .25 –1.99 –2.08*
–2.29 –2.17*
FSIZE ? 1.22 1.84† 1.36 1.94† .29 1.21 .21 .86 .54 2.01*
.58 2.07*
AGE ? –.23 –1.45 –.34 –1.77† –.11 –1.37 –.02 –.22 .04 .45 .07 .68
Y2002 - - .75 .46 - - –.22 –.34 - - –.41 –.52
Y2003 - - 2.12 .90 - - –1.13 –1.25 - - –.29 –.35
Y2004 - - 2.09 .85 - - –1.42 –1.46 - - –.71 –.76
Y2005 - - 2.97 1.15 - - –1.71 –1.63 - - –.64 –.62
Pseudo R2
31.36% 33.31% 11.47% 13.14% 21.72% 22.12%
Wald 2 29.78
** 31.63
** 32.75
** 37.52
** 37.7
*** 38.39
***
Cross-section 45 (38.46%) f
42 (35.9%) f
30 (25.64%) f
Observations 191 (35.37%) g
210 (38.89%) g
139 (25.74% ) g
The table reports the results of logistic regression models:
it
4
1l
2
1kkl
4
1jitj
it8it7it6it5it4it3it2it1it
SHAREDUMMYYEARDUMMYCONTROLS
SBMEETPROFSBSBSIZEBDMEETINDPBSIZEFOREIGNSTATEAQ
where, AQ = audit quality, the dichotomous dependent variable is coded as 1 if the firm is audited by a Big 4 audit firm, otherwise coded as zero; STATE = state ownership
concentration, proportion shares held by the state; FOREIGN = foreign ownership concentration, proportion of shares held by foreign investors; BSIZE = board size, the
number of directors on the board; INDP = the number of independent directors on the board; BDMEET = the number of meetings of the board of directors in the fiscal year;
SBSIZE = the number of members on the supervisory board; PROFSB = professionalism of the supervisory board, number of supervisory board members with professional
knowledge or work experience; SBMEET = the number of meetings of the supervisory board in the fiscal year; ROA = return on total assets; TOBINSQ = Tobin‟s Q, market
39
value of stock and book value of debt divided by book value of total assets; FSIZE: firm size, natural logarithm of value of total assets at the end of fiscal year; AGE: firm
age, years since initial listing
Notes: a In logistic regression model we coded as 1 if the firm is audited by a Big 4 auditor, otherwise coded as 0.
b Two-tailed p-values are used in determining significance: †if p<.10;
* if p <.05;
** if p<.01;
*** if p<.001
c A-shares are common stock issued by mainland China firms, subscribed and traded in RMB, listed on the mainland stock exchanges, and are reserved for trading by Chinese
citizens. The A-share market was launched in 1990 in Shanghai. d B-shares are issued by mainland China firms, traded in foreign currencies, and listed on the mainland stock exchanges. The B-share market was launched in 1992 and was
restricted to foreign investors before 19 February 2001. AB-share companies in this paper indicate those that have issued both A-shares and B-shares, with an initial A-share
offering. They are also listed on the domestic stock exchanges in China. e H-shares are securities of companies incorporated in mainland China and nominated by the Chinese Government for listing and trading on the Hong Kong Stock Exchange,
being quoted and traded in Hong Kong Dollar. There are no restrictions on holdings by international investors. AH-share companies in this paper indicate those that have
issued both A-shares and H-shares, and have floated their shares simultaneously on the Hong Kong Stock Exchange and one of China‟s two mainland stock exchanges. f Percentage of sample.
g Percentage of observations.