corporate governance and the information environment: evidence from chinese stock markets
TRANSCRIPT
�������� ����� ��
Corporate governance and the information environment: Evidence fromChinese stock markets
Lars Helge Haß, Skralan Vergauwe, Qiyu Zhang
PII: S1057-5219(14)00047-7DOI: doi: 10.1016/j.irfa.2014.03.010Reference: FINANA 704
To appear in: International Review of Financial Analysis
Received date: 21 May 2013Revised date: 2 February 2014Accepted date: 24 March 2014
Please cite this article as: Haß, L.H., Vergauwe, S. & Zhang, Q., Corporate governanceand the information environment: Evidence from Chinese stock markets, InternationalReview of Financial Analysis (2014), doi: 10.1016/j.irfa.2014.03.010
This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
1
Corporate governance and the information environment: Evidence from Chinese stock
markets
Lars Helge Haß
1 Skrålan Vergauwe
2 Qiyu Zhang
3
This version: February 2, 2014
ABSTRACT
This article explores the relationship between corporate governance and the information
environment in Chinese stock markets. We construct a parsimonious governance measure for
public firms using a 2003 through 2011 sample period. We use four indicators to proxy for
the information environment: analyst following, analyst forecast accuracy, analyst forecast
dispersion, and price timeliness. We find that better governed firms tend to be associated with
larger analyst followings and more informative forecasts. We also find that better governed
firms tend to improve on the timeliness of bad news relative to good news. Our results are
robust for an instrumental variable analysis, which confirms a causal relationship between the
quality of corporate governance and the information environment of a firm.
JEL Classification: G14; G30; M41
Keywords: Corporate governance; information environment; Chinese stock markets
1
Lancaster University Management School, Lancaster University, LA1 4YX, Lancaster, United Kingdom,
Phone: +44 1524 - 593981, Fax: +44 1524 847321, e-mail: [email protected].
2 Lancaster University Management School, Lancaster University, LA1 4YX, Lancaster, United Kingdom,
Phone: +44 1524 – 594738, Fax: +44 1524 847321, e-mail: [email protected]. 3 Lancaster University Management School, Lancaster University, LA1 4YX, Lancaster, United Kingdom,
Phone: +44 1524 – 593625, Fax: +44 1524 847321, e-mail: [email protected].
Acknowledgments: We are grateful to an anonymous referee for many helpful comments,
and the special issue editors Douglas Cumming, Wenxuan Hou, Edward Lee, and Zhenyu
Wu for very useful suggestions. The authors gratefully acknowledge valuable feedback
from Youchao Tan (discussant) and the participants at the Conference on Corporate
Governance and Entrepreneurial Finance in China.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
2
Corporate governance and the information environment: Evidence from Chinese stock
markets
ABSTRACT
This article explores the relationship between corporate governance and the information
environment in Chinese stock markets. We construct a parsimonious governance measure for
public firms using a 2003 through 2011 sample period. We use four indicators to proxy for
the information environment: analyst following, analyst forecast accuracy, analyst forecast
dispersion, and price timeliness. We find that better governed firms tend to be associated with
larger analyst followings and more informative forecasts. We also find that better governed
firms tend to improve on the timeliness of bad news relative to good news. Our results are
robust for an instrumental variable analysis, which confirms a causal relationship between the
quality of corporate governance and the information environment of a firm.
JEL Classification: G14; G30; M41
Keywords: Corporate governance; information environment; Chinese stock markets
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
3
1. Introduction
The Chinese economy has grown dramatically over the last few decades, primarily because of
substantial economic reforms. It is now the second largest economy in the world by GDP;
growth rates over the last thirty years have averaged 10%.4 This growth spurt, along with the
East Asian financial crisis, has led to increased interest in corporate governance in China by
academics and practitioners. Corporate governance (CG) is generally defined as a set of
mechanisms by which outside investors can protect themselves against expropriation by
insiders (La Porta et al., 2000). It also describes the structure of stakeholders’ rights and
responsibilities (Aguilera and Jackson, 2003).
Despite the numerous reforms made by the Chinese government (Cheung et al., 2008),
however, the prevalence of corporate governance in China remains contested. For example,
China continues to suffer from widespread corruption. In 2013, China ranked 80th out of 178
countries on Transparency International’s Corruption Perceptions Index. This is on a
comparable level with Serbia and Trinidad and Tobago, and a more corrupt level than Sri
Lanka and most developed countries.5
Moreover, collusion between individual
businesspeople and agents of the state permeates Chinese culture, which further inhibits the
cultivation of good governance practices. The secret agreements made between these two
groups typically aim to avoid governance provisions and ultimately negate their intended
effects.
Since the 2000s, however, significant progress has been made by the government to
transform the corporate culture, in line with lofty ideals such as building a “Harmonious
Socialist Society.” The Code of Corporate Governance was released in 2001, followed by
several governance reforms such as the non-tradable shares reform in 2005. Ultimately, it
4 Source: http://www.imf.org/external/pubs/ft/weo/2013/01/weodata.
5 Source: http://cpi.transparency.org/cpi2013/results/.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
4
remains unclear how much good corporate governance matters to corporate operations and
capital markets in China, though. This paper attempts to shed light on this issue by
highlighting the relationship between corporate governance and the information environment
of Chinese firms. The latter is a critical component of information transparency and pricing
efficiency.
The Shanghai and Shenzhen stock markets opened in 1990 and 1991, respectively, and
greatly increased investments from domestic and foreign market participants (Sami et al.,
2011). The increased globalization has played an important role in forcing Chinese firms to
adopt international practices and oversight mechanisms, including corporate governance
rules, in order to increase trade and ties with other countries. However, the Chinese market
has some unique features that challenge the implementation and effectiveness of corporate
governance measures.
First, the existence of state control leads to a specific type of agency problem, whereby the
state retains the power to expropriate minority shareholders (Shleifer and Vishny, 1997;
Clarke, 2003; Bai et al., 2004). Second, China uses a two-tiered board structure consisting of
a main board of directors and a board of supervisors. However, ownership and control are not
fully separate, which can potentially lead to governance problems.
Finally, prior to 2005, listed firm shares were divided into non-tradable and tradable shares.
Non-tradable shares were generally state shares and legal entity shares, while tradable shares
were held by individuals, institutions, and private businesses. Consequently, shares held by
the main shareholders could only be transferred through negotiation and auctions. Thus, there
was often a conflict of interest for the majority shareholders over their motivation to improve
company performance (Chen et al., 2011). This split share structure increased agency
problems between majority and minority shareholders (Claessens and Fan, 2002; Jian and
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
5
Wong, 2010; Jiang et al., 2010). Such features are a prime example of the strong need for
good corporate governance mechanisms.
To investigate the relationship between a firm’s corporate governance level and its
information environment, we construct an aggregate firm-level corporate governance index.
We proxy for the information environment by using the number of analysts following a
particular firm, their forecast accuracy, forecast dispersion, and the timeliness of price
discovery. Our sample consists of listed firms in China from the 2003 through 2011 period.
Our results indicate that better governed firms are associated with larger analyst followings
and more informative forecasts. In addition, we find that better governed firms tend to
improve the timeliness of bad news relative to good news. Results are robust for an
instrumental variable analysis, confirming a causal relationship between the quality of
corporate governance and the informativeness of a firm.
The remainder of this paper is organized as follows. Section 2 discusses the research
background and develops our hypotheses. Section 3 describes the data and methods, while
section 4 presents our empirical results. Section 5 concludes.
2. Research background and hypothesis development
2.1. Research background
La Porta et al. (2000) find that firms in emerging economies may be discounted in financial
markets because of perceptions of weak governance. In fact, a survey by McKinsey (2002)
reveals that investors are willing to pay a 25% premium for well-governed firms on average.
This dramatically highlights the importance of good governance to Chinese firms in
increasing investor confidence and access to capital (Ding and Sun, 1997; Rajagopalan and
Zhang, 2008; Buchanan et al., 2012).
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
6
In recent years, the Chinese government has launched programs and reforms driven by
globalization and privatization that have been declared transformational to the landscape of
corporate governance. However, as we mentioned earlier, the Chinese economy possesses
three unique characteristics that challenge the effectiveness of corporate governance.
First, the Chinese market contains an unusually high proportion of state-owned enterprises
(SOE) (Chen et al., 2009). Privatization, or more diversified ownership structures, can lead to
two agency conflicts and consequently to a greater need for better governance: 1) the
traditional principal agency problem, whereby management’s interests are not aligned with
shareholders’ interests, and 2) the principal-principal agency problem, whereby majority
shareholders expropriate minority shareholders’ interests (Shleifer and Vishny, 1997;
Dharwadkar et al., 2000; Naceur et al., 2007; Liu et al., 2012).6
Second, the customary two-tiered board structure in China, which consists of a main board of
directors and a board of supervisors, can be problematic for the goals of good corporate
governance. Chinese law states that the board of supervisors should be independent of the
board of directors in order to better monitor managerial behavior and decision-making.
However, in practice, the board of supervisors is rather limited in their latitude of action,
because supervisors have no voting rights. Also, the board of supervisors is ultimately subject
to oversight by the board of directors because the supervisory board members are firm
employees.
In addition, the government plays an outsize role in the appointment of board and supervisory
board members. One major concern is that board members’ lack of independence means they
actually contribute little to the monitoring of management, and hence the efficiency of
6In most Chinese-listed firms, there is a single dominant shareholder. Chen et al. (2009) find that, for the 1999-2004 period,
the median holding of the largest shareholder is 42.61%, while the median of the second largest shareholder is only 5%, and
that of the third-largest shareholder is 1.89%. The dominant shareholder therefore wields considerable power and influence
over a firm’s operations.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
7
Chinese firms (Fan et al., 2007; Chen et al., 2009; Hu et al., 2010). As a complement to the
board of supervisors, the independent director system is required by the Code of Corporate
Governance issued by the CSRC (CSRC, 2002). Evidence shows that independent directors
in China can serve as effective corporate governance mechanisms (Kato and Long, 2006; Fan
et al., 2007; Conyon and He, 2011). Furthermore, Lo et al. (2010) find that a more
independent board and a separation between the roles of CEO and chairperson suggest that a
firm is less likely to engage in transfer pricing manipulations; Liu and Lu (2007) find that
firms with better corporate governance have lower levels of earnings management.
Finally, prior to 2005, Chinese firms had both tradable and non-tradable shares. In 2005,
China removed all trading restrictions from non-tradable shares. This reform was one of the
starting points in the transition from primarily stateownership to public ownership.7 The split
share structure, in which both types of shares have the same voting rights but different prices,
greatly impacted corporate governance. First, large shareholders holding non-tradable shares
had little incentive to improve firm performance, because they could only trade their shares at
book value (Chen and Yuan, 2006). Second, the split share structure induced conflicts of
interest between the different types of shareholders (Chen et al., 2011). Larger shareholders
were prone to using their control rights for, e.g., tunneling, whereby they attempt to
artificially increase dividends as a means to transfer more cash to their own pockets (Lee and
Xiao, 2004), and/or transfer resources to benefit controlling shareholders at the expense of
minority shareholders (Aharony et al., 2010; Jiang et al., 2010).
7 Another effort by the government to diversify ownership was to increase the presence of institutional investors in the local
stock market. It is widely acknowledged that institutional shareholders can help improve corporate governance quality while
reducing information asymmetries (Smith, 1996; Woidtke, 2002; Aggarwal et al., 2011). Mutual funds were introduced by
the China Securities Regulatory Commission (CSRC) in 1998. By the end of June 2007, there were 343 open-ended mutual
funds with a total net value of over 1.7 trillion Chinese yuan. In addition, the Qualified Foreign Institutional Investor (QFII)
program, launched in 2002, allows licensed foreign investors to buy and sell yuan-denominated “A” shares on China’s
mainland stock exchanges. According to Lane and Milesi-Ferretti (2007), portfolio equity inflows had grown to U.S. $450
billion by 2007, from U.S. $13 billion in 2001.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
8
Thus, good corporate governance in China is especially important because of the severity of
agency problems. Despite the fact that the third issue became much less important after 2005,
the other two issues remain problematic. Research has found that good corporate governance
is associated with increased market valuations (Bai et al., 2004), a reduced propensity to
commit fraud (Chen et al., 2006), greater influence over capital structure decisions (Wen et
al., 2002; Li et al., 2009), improvements in operating performance (Sami et al., 2011), and
increased firm liquidity (Tang and Wang, 2011).
2.2. Hypothesis development
In theory, corporate governance mechanisms are expected to reduce the agency costs
associated with the separation of ownership and control (Jensen and Meckling, 1976; Fama,
1980; Fama and Jensen, 1983). Such mechanisms are both internal and external, and they
serve to align shareholders’ and management’s interests. Internal mechanisms are related to
the structure of the board, such as duality and the proportion of non-executive directors, debt
financing, and executive director shareholdings. External mechanisms include the effective
takeover market, legal infrastructure, and product market competition (Liu, 2006).
Agency problems can be mitigated by increasing shareholder monitoring and increasing
controlling activities (Shleifer and Vishny, 1986; Huddart, 1993; Noe, 2002) by, i.e., analysts
covering the firm. Analysts seek to forecast earnings accurately (Mikhail et al., 1999) by
using firm-provided disclosures (Stickel, 1989; Lang and Lundholm, 1993; Ashbaugh and
Pincus, 2001; Barron et al., 2002; Byard et al., 2006). Prior research has extensively
investigated the impact of corporate governance on financial reporting quality. Better quality
governance is associated with a lower likelihood of financial statement fraud (e.g., Beasley,
1996), a lower incidence of earnings management (e.g., Dechow et al., 1996; Peasnell et al.,
2000; Klein, 2002), higher voluntary disclosure levels (e.g., Eng and Mak, 2003), and more
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
9
precise and more frequent managerial earnings forecasts (Ajinkya et al., 2005; Karamanou
and Vafeas, 2005).
Overall, firms with better corporate governance levels will have more reliable information
(Yu, 2010), which should lower analysts’ costs of providing forecasts, as well as reduce
forecast errors. Therefore, forecast revisions should be smaller, forecasts should be more
accurate, analyst followings should increase, and there should be fewer disagreements among
analysts (Lang and Lundholm, 1996; Brown et al., 2011).
In our setting, agency costs are relatively high because of the features of Chinese corporate
governance discussed earlier. We thus expect a significantly positive correlation between a
firm’s information environment, proxied for by analyst forecast properties, and corporate
governance levels. This is confirmed by Byard et al. (2006), who find evidence that the
information available to analysts improves with better corporate governance, proxied for by
board independence, audit committee independence, board size, and CEO duality in the U.S.
Beekes and Brown (2006) find that better corporate governance leads to a higher level of
analyst following and greater forecast accuracy among Australian firms; Beekes et al.
(2012a) find that better corporate governance leads to a higher analyst following, lower
dispersion rates, and greater accuracy in Canada.8 Moreover, in a cross-sectional setting, Bhat
et al. (2006) show that analyst forecast accuracy is positively influenced by corporate
governance disclosures. Xu and Tang (2008) find lower analyst forecast accuracy when
internal controls are weaker. Finally, Nowland (2008) reports that the introduction of
voluntary corporate governance codes leads to lower analyst forecast errors in Asia.
8 Beekes and Brown (2006) find greater disagreement among forecasts when corporate governance is higher.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
10
In conclusion, corporate governance generally has a positive impact on the properties of
analysts’ forecasts. Given specific features of the Chinese setting, we believe it is an
especially vital component there. Hence, we pose the following hypotheses:
H1: Firms with better CG are associated with higher levels of analyst following.
H2: Firms with better CG are associated with more accurate forecasts.
H3: Firms with better CG are associated with lower levels of forecast dispersion.
Besides analysts’ forecast properties, we also investigate the relationship between governance
and the timeliness of price discovery. Information disclosed to investors must not only be
credible, but also disclosed in a timely manner. Management has an incentive to provide
investors with timely information, but nevertheless tend to delay delivering bad news
(Kothari et al., 2009). However, as Skinner (1994, 1997) notes, bad news must be disclosed
eventually because of litigation risk and reputation costs.
Information related to earnings announcements gets incorporated into the share price before
the actual earnings release dates (Ball and Brown, 1968). Price discovery is the process
whereby value-relevant, inside information gets incorporated into a stock’s publicly
observable market price (Beekes and Brown, 2006). Prior research has found a correlation
between timely disclosure and corporate governance, for several reasons. First, outside
directors bear reputation costs, and therefore encourage timely disclosure (Ajinkya et al.,
2005; Abdelsalam and Street, 2007). Independent directors are likely to monitor management
more closely in order to align shareholder and management interests (Fama and Jensen, 1983;
Weisbach, 1988; Borokhovich et al., 1996). Second, CEO duality has a negative impact on
reporting quality and timeliness (Blackburn, 1994; Argenti, 1976). Third, firms with more
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
11
shareholders tend to provide more timely information (Bushee and Noe, 2000; Bushee et al.,
2003).
Overall, we expect corporate governance mechanisms to improve the timely release of
information. We also expect the information to be perceived as more credible, and that it will
have a favourable impact on the timeliness of price discovery. Consistent with this prediction,
Beekes and Brown (2006) find evidence of timelier price discovery for better governed
Australian firms, although Beekes et al. (2012a) find weaker evidence for this relationship
among Canadian firms. However, using a cross-country sample, Beekes et al. (2012b) find
that firms with better governance quality substitute governance for greater transparency,
which is proxied for by a more timely release of information to the market. In other words,
price discovery is slower, consistent with prior U.S. evidence (Bushman et al., 2004).
We posit that the relationship between corporate governance and the timeliness of price
discovery will vary with the legal environment and ownership structure. Given the specific
characteristics of our Chinese setting, we predict a positive correlation. We thus test the
following hypothesis:
H4: Price discovery is timelier (faster) for firms with better CG.
3. Data and methods
3.1. Data and sample construction
Our data come from the China Stock Market and Accounting Research (CSMAR) platform,
which is a collection of databases of Chinese-listed firms. Specifically, our primary source
for corporate governance data is the China Listed Firms Corporate Governance Research
Database; data on analyst coverage and forecasts come from the China Securities Market
Analyst Forecasts Research Database; financial data come from the China Stock Market
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
12
Financial Statements Database; stock prices and trading data are from the China Stock
Market Trading Database; and information on auditors is from the China Stock Market
Financial Database – Audit Opinion.
Our sample selection begins with firms that have available data on corporate governance. Our
initial sample consists of 2,324 firms over the 2003-2011 period. We begin our sample in
2003 because it is the first year when data on information environment, corporate governance
attributes, and control variables are all available. We exclude firms with missing information
environment variables and control variables. We then exclude financial firms because of their
unique accounting and financial characteristics. We also exclude B-shares and H-shares that
are open to international investors, because these stocks are subject to different reporting
requirements. We are left with a final sample of 2,152 firms for the analyst following model,
2,176 firms for the forecast accuracy model, 1,995 firms for the forecast dispersion model,
and 2,134 firms for the price timeliness model.9
3.2. Measures of firm-level corporate governance
We construct a parsimonious firm-level score of corporate governance that captures seven
characteristics. We base our selection of governance characteristics and our score
construction on prior literature, the potential changes in governance brought by relevant
reforms, and data availability. For example, prior studies have found that governance
attributes such as board size, percentage of independent directors, separation of the chairman
and CEO positions, and percentage of managers’ and directors’ stock ownership can impact
the valuation and operation of listed firms in China (e.g., Bai et al., 2004; Chen et al., 2006;
Sami et al., 2011).
9 Our sample is comparable to other recent research on Chinese firms (e.g. Xu et al., 2013).
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
13
Fan and Wong (2005) find that external auditors perform a corporate governance-type role
for firms with agency problems in East Asian emerging markets. Note that we are particularly
interested in governance items that might be affected by recent reforms. During the 2000s,
several reforms were implemented, such as the independent director provision in 2002, the
introduction of Qualified Foreign Institutional Investors (QFII) in 2002, and the non-tradable
shares reform in 2005. Following the above standards, we select the following seven
attributes: independent outside directors as a percentage of total number of board members
(INDIV); total number of directors (including board chairman) on the company’s board
(BOARDSIZE); whether one person shares both the board chairman and general manager
positions (DUAL); whether there are any relationships among the top ten shareholders
(TOP10RELATION); shares held by directors, supervisors, and executives as the proportion
of total number of shares (MANAGEMENT); shares held by foreign investors as the
proportion of total number of shares (FOREIGN); and whether the auditor is a member of one
of the joint ventures of the Big Four international audit firms and domestic audit firms
(BIG4).
In light of results from previous studies (e.g., Brown and Caylor, 2006, 2009; Chung et al.,
2010; Aggarwal et al., 2011), we impose criteria for each attribute.10
We construct a dummy
variable that is set equal to 1 if the governance attribute meets certain criteria, and 0
otherwise in a given year. We then combine the seven categories into a governance score that
measures overall governance quality by summing all the dummy variables, denoted as CG. A
higher CG value suggests a better governance mechanism.
Panel A of Table 1 gives the selected governance features, as well as some basic statistics.
Panel B presents the criteria used for governance attributes and the proportions of qualified
observations. We observe some notable variations among attributes. For example, for 10 See Table 2 for more details.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
14
BOARDSIZE and DUAL, more than 80% of firm-year observations meet the criteria, while
only about 2% do for INDIV. Furthermore, qualified firm-year observations are at
intermediate levels for TOP10RELATION, MANAGEMENT, and FOREIGN.
[Insert Table 1 about here]
Table 2 presents the summary statistics and the Pearson correlation coefficients of the
corporate governance attributes. Panel A gives the summary statistics for the CG score and
the governance attributes in our sample. The mean (median) CG is 2.014 (2.000). Panel B
shows the Pearson correlations among the composite CG score and its seven components.
Note that CG is significantly correlated with each component, but the components are not
significantly correlated with each another. This latter result indicates that the seven
components capture different aspects of the corporate governance mechanism.
[Insert Table 2 about here]
3.3. Proxies for the information environment and estimation models
We create four measures of the information environment within which firms operate: analyst
following, analyst forecast accuracy, analyst forecast dispersion, and timeliness of price
discovery. We then examine the relationships among these measures and the corporate
governance measure described previously.
3.3.1. Analyst following and forecast accuracy
To measure analyst following (COVERAGE), we calculate the natural logarithm of the
number of unique analysts covering a firm in a particular year. Forecast accuracy
(ACCURACY) is the absolute value of the forecast EPS minus the actual EPS and deflated by
the stock price at the beginning of the year, where forecast EPS is the most recent figure prior
to the announcement date of the year but subsequent to the previous EPS announcement.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
15
Forecast dispersion (DISPERSION) is the standard deviation of the forecast EPS of the year,
deflated by the stock price at the beginning of the year. We calculate all of these variables on
an annual basis for each firm.
3.3.2. Timeliness of price discovery
The measure of price timeliness is derived from Beekes and Brown’s (2006) approach. The
measure has its origins in the seminal work of Ball and Brown (1968), who stated that annual
income reports are not a timely enough format for disclosing price-sensitive information,
because most of their content (85%-90%) has already been captured by more timely media.
Hence, it seems more suitable to assess how accurately a firm’s share price (Pt), observed at
daily intervals throughout the year, approximates the market’s valuation two weeks (14 days)
after the annual earnings (P0) release date. Specifically, we calculate the timeliness of price
discovery (T) as:
(1)
where Pt is the market-adjusted share price, which is observed at daily intervals from day
-365 until day -1, and P0 is the price 14 days after the release date.11
-0.5/365 is an adjustment
made to recognize the flow of information, which is reflected in returns over the day.
The idea behind this measure is simple. The longer it takes a firm’s share price to capture
information and converge to its “final” price P0 (which reflects all value-relevant information
discovered during the year), the larger the value of TIMELINESS. A high value for
TIMELINESS thus indicates low intra-year timeliness. In contrast, if all the information that
affects the final price was incorporated at day -365, TIMELINESS would be at its minimum,
and the speed of adjustment at its maximum. We can interpret TIMELINESS as a measure of
11 Prices are backfilled on days the market was closed (e.g., weekends), or when there was no trading in the stock. We set the
ending day to be fourteen days after the earnings release date, because the market may need time to absorb information.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
16
how much information regarding the final earnings number is already known from other
sources prior to its release. In this sense, the TIMELINESS variable is inextricably linked to
the timeliness of other disclosures related to earnings that are more timely than an annual
report.
3.3.3. Estimation models
To investigate how corporate governance is related to analyst following, forecast accuracy,
forecast dispersion, and price timeliness, we use the following models, which include year
and industry fixed-effects.
(2)
(3)
, (4)
(5)
We use four information environment variables as dependent variables: COVERAGE,
ACCURACY, DISPERSION, and TIMELINESS. The main independent variable, corporate
governance, is again denoted by CG. As the literature suggests, we include a large number of
control variables for each model (e.g., Bhushan, 1989; Brennan and Hughes, 1991; Lang and
Lundholm, 1996; Beekes and Brown, 2006; Landsman et al., 2012; Horton et al., 2013).
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
17
We control for firm size (SIZE), defined as the natural logarithm of market capitalization at
year-end, because larger firms are more likely to attract higher levels of analyst coverage. We
control for profitability by including the return on equity ratio (ROE), and a dummy variable
(LOSS) that is equal to 1 if the firm reports a negative net income, and 0 otherwise. We
control for financial leverage (LEV) and growth opportunities (MB) by including the ratio of
total liabilities to total assets and the market-to-book ratio, respectively. We further control
for stock return volatility (RETVOL) and trading volume (VOLUME), where RETVOL is
measured as the standard deviation of daily stock returns over the 360 days prior to the end of
the year, and VOLUME is defined as the natural logarithm of trading volume for the year. We
also consider the effects of IFRS adoption in China on the information environment by
including a dummy variable (IFRS) that is equal to 1 for the year 2007 and afterward, and 0
otherwise. Also, higher brokerage commissions are likely to suggest a higher level of analyst
following. To account for the rate of brokerage commissions, we include the inverse of the
mean stock price for the year (INVPRICE).
These variables denote factors that affect analysts’ incentives to collect information, and are
thus likely to affect the properties of their forecasts. In addition, for model (3), we control for
the length of the forecast horizon (HORIZON), defined as the natural logarithm of the number
of days from the most recent forecast date until the EPS announcement date of the year. This
is because forecasts tend to improve as the date of the earnings release draws closer, due to
the progressive release of information throughout the year.
Finally, as a robustness check, we consider three ownership variables to control for the
governance characteristics that are not included in the aggregate CG score, but that will
influence the information environment. These variables are state-owned shares as the
proportion of total number of shares (STATE), non-tradable shares as the proportion of total
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
18
number of shares (NONTRADE), and shares held by the largest shareholder as the proportion
of total number of shares (TOP1).
Table 3 shows the summary statistics for information environment and the control variables.
The values for average analyst following (NUMEST), forecast accuracy (ACCURACY),
forecast dispersion (DISPERSION), and price timeliness (TIMELINESS) are 1.723, 0.009,
0.008, and 0.177, respectively. The average firm size value is 21.747, measured by the
natural logarithm of market value. The sample firms are profitable, as shown by the average
ROE ratio of 5.6%. The debt ratio averages 51.2%, while the market-to-book ratio averages
3.412. The averages are 0.024, 27.229, 0.128, and 4.208 for return volatility (RETVOL), the
natural logarithm of trading volume (VOLUME), the mean inverse of the stock price
(INVPRICE), and forecast horizon (HORIZON), respectively. The averages of ownership
control variables (STATE, NONTRADE, and TOP1) are 20.8%, 43.6%, and 37.6%,
respectively.
[Insert Table 3 about here]
4. Empirical results
4.1. Univariate analysis
Table 3 also compares firms with higher and lower CG scores. The mean equality test reveals
that firms with a higher CG score tend to have a larger analyst following, more accurate
forecasts, less volatile forecasts, and faster price discovery. These characteristics suggest that
better governed firms mitigate agency problems, and thus improve corporate transparency
and the information environment, a result that is consistent with our predictions.
Table 4 shows the Pearson correlation coefficients between variables. The relationships
between CG and the information environment variables confirm some of our findings in
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
19
Table 3. Corporate governance is positively correlated with analyst following, and negatively
correlated with forecast accuracy and price timeliness. The univariate analyses provide prima
facie evidence that stronger corporate governance is associated with an increase in analyst
coverage, forecast accuracy, speed of price discovery, and a decrease in forecast volatility. In
the next section, we will further explore the relationship between the information
environment and corporate governance using multivariate regression analyses.
[Insert Table 4 about here]
4.2. Regression analysis
Tables 5-8 report the regression results for the models proposed in Section 3.3.3. In each
table, we arrange regression specifications as follows. For models 1-4, we use the pooled
ordinary least squares (OLS) method, with standard errors clustered by firm; for models 5-8,
we use fixed-effects (FE) regressions, with standard errors clustered by firm. The fixed-effect
approach controls for unobservable firm characteristics that remain constant through time,
which could result in spurious relationships among the variables we examine. Specifically, in
models 1 and 5, we estimate the effects of the aggregate CG score. In models 3 and 7, we
replace the aggregate CG score with individual components. In the rest of the models, we
include the three additional ownership variables.
4.2.1. Analyst Models
Table 5 gives the regression results for analyst following. Firms with higher CG scores are
associated with higher analyst followings in both the OLS and fixed-effects models (i.e., the
CG score has a positive and significant coefficient in models 1, 2, 5, and 6). This finding
confirms H1. As expected, better governed firms attract more analyst coverage: A 1-standard
deviation increase in CG is associated with a 0.027 increase in the natural logarithm of the
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
20
number of unique analysts following a stock, which is approximately equivalent to one more
analyst.12
For SIZE, which is one of the major control variables, a 1-standard deviation increase is
associated with an estimated increase of two more analysts. We do not, however, find a
strong and robust relationship between analyst following and individual governance
attributes. The impression is thus that governance attributes work collectively to affect
analyst following.
[Insert Table 5 about here]
The results of forecast accuracy are in Table 6. The coefficient on CG is -0.001, which is
significant at the 1% level in the OLS and fixed-effects models. This is consistent with the
prediction in H2 that better corporate governance improves forecast accuracy. A 1-standard
deviation increase in CG is associated with a 0.0005 increase in forecast accuracy, deflated
by the stock price. The magnitude for SIZE is 0.001, suggesting that the effect of CG on
forecast accuracy is economically significant.
Our results largely mirror those in Beekes and Brown (2006), who use Australian data and
document that better governed firms have larger analyst followings, and that analysts’
consensus forecasts are more accurate. Good corporate governance seems to offer the same
beneficial effects under two different institutional environments. Among the seven attributes
of corporate governance, foreign ownership has a negative coefficient across different
specifications. This echoes the finding of Bae et al. (2006) that openness to foreign investors
has a favourable effect on the information environment of emerging stock markets.
[Insert Table 6 about here]
12
The economic magnitude of the effect is based on model 1.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
21
Under H3, we would expect to find that a better CG score is associated with lower forecast
volatility. As shown in Table 7, however, a higher CG score is negatively correlated with
DISPERSION in the OLS models, but no significant relationship is found for the fixed-effects
models. It is worth noting that the proportion of independent directors and management
shareholding variables have negative coefficients, suggesting lower dispersion if a firm has
more independent directors on the board and if management holds a moderate amount of
shares.
[Insert Table 7 about here]
4.2.2. Timeliness model
The results for the timeliness model are reported in Table 8. We detect only modest effects of
CG on timeliness, at the 10% and 5% levels, respectively, in models 1 and 3. This suggests
that value-relevant information is priced more rapidly when the firm has better corporate
governance mechanisms.
Some earlier research has documented how investors’ short-term mindsets can result in
increased volatility of share prices when firms increase their levels of disclosure. This could
partially explain our results. For example, Bushee and Noe (2000) report that high levels of
disclosure attract more transient institutional investors who trade aggressively on short-term
earnings news, thus increasing return volatility. Regarding the effects of governance
attributes, there is evidence that a higher proportion of independent directors and the
involvement of the Big Four auditors contribute to faster price timeliness. This is consistent
with the argument that outside directors contribute to more timely disclosures (e.g., Ajinkya
et al., 2005; Abdelsalam and Street, 2007).
[Insert Table 8 about here]
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
22
4.3. Alternative estimates
4.3.1. Instrumental variables estimation
We argue that better corporate governance leads to a better information environment.
However, we note that the directions of causality could be reversed. For example, a higher
level of analyst coverage may put external pressure on firms, causing them to improve their
corporate governance. To identify the causal effects of CG on the information environment,
we estimate an instrumental variables (IV) regression using prior year CG scores and average
industry level of CG to stand in for CG. We use two-stage least squares (2SLS) estimation
methods, with standard errors clustered by firm. The results are reported in Table 9.
[Insert Table 9 about here]
The effects of corporate governance on information environment variables are somewhat
weaker than those found earlier using the OLS method. For analyst following and timeliness,
the coefficients of CG are not significant, which echoes our earlier concern. Although less
significant than before, the effects of CG remain significant for accuracy models at the 10%
level. For forecast dispersion, estimates are greater in magnitude than the estimates from the
OLS models, and are significant at the 5% level.
To test whether the instruments can satisfy the irrelevance condition and thus be considered
valid, we report the p-values of Hansen’s J-statistics, where a rejection of the null hypothesis
casts doubt on validity. The p-values suggest that the null hypothesis cannot be rejected, thus
the instruments are valid. In other words, they are uncorrelated with the error term, and are
correctly excluded from the estimated equations. We also report the p-values of the
underidentification test, i.e., whether our instruments are correlated with the endogenous CG
variable, where a rejection of the null indicates that the model is identified. The p-values
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
23
suggest that the null hypothesis that the instruments are correlated with CG and the models
are identified in all cases is rejected.
4.3.2. Timeliness of good news versus bad news
The imbalance in corporate disclosures has long been a familiar topic in the literature. For
example, managers are more likely to withhold bad news due to their private incentives (e.g.,
Mendenhall and Nichols, 1988; Kothari et al., 2009). If better CG can mitigate the agency
problems between managers and shareholders, we posit that bad news will be disclosed
earlier for better governed firms. To test this notion, we use two alternative measures of
timeliness: timeliness of good news, and timeliness of bad news.
To measure the timeliness of news, we use the approach discussed in Beekes and Brown
(2007). For the timeliness of good news, we first construct a time series of good news returns,
which includes positive market-adjusted daily log returns. Previous positive returns
are carried forward for negative return days. We then create cumulative log return series, ,
by setting and combining the good news return series as
from
day -364 to day 0:13
(6)
We estimate the timeliness of bad news in the same way. Table 10 gives the regression
results for the timeliness measures. The coefficient on CG is not significant for the timeliness
of good news, implying there is no relationship between corporate governance and the speed
of recognition of good news. The negative and significant coefficient on CG for the
timeliness of bad news suggests that price discovery of bad news is faster for better governed
firms than for poorly governed firms. Note that releasing bad news earlier could significantly
13 We thank Philip Brown for suggesting this timeliness estimation method. Note that we use cumulative returns and a
discounting factor, , in the equation, rather than prices. We believe this will help mitigate any volatility-induced bias in
the Beekes and Brown (2006) measure.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
24
benefit firms by reducing litigation risk, and by eliminating greater stock price impacts at the
end of the year (Cornell and Landsman, 1989; Skinner, 1994, 1997; Baginski et al., 2002). To
this end, better governed firms seem to be more forthcoming with bad news.
[Insert Table 10 about here]
5. Conclusion
Our Chinese setting has several specific characteristics that are likely to cause severe agency
conflicts. Better corporate governance mechanisms can help overcome these conflicts. We
investigate the relationship between higher levels of corporate governance and a firm’s
information environment, measured as analyst coverage, analyst forecast accuracy, analyst
forecast dispersion, and timelier price discovery. The underlying reasoning is that better
corporate governance can enhance the reliability of the information available to analysts (Yu,
2010). Better corporate governance can also enhance disclosure, reduce fraud and earnings
management, and increase the frequency of management forecasts (Beasley, 1996; Dechow
et al., 1996; Peasnell et al., 2000; Klein, 2002; Eng and Mak, 2003; Ajinkya et al., 2005;
Karamanou and Vafaes, 2005). Moreover, better corporate governance can enhance the
timeliness of price discovery, because better governed firms tend to release information in a
more timely fashion, and that information is perceived as more credible (e.g., Bushee and
Noe, 2000; Bushee et al., 2003; Beekes and Brown, 2006).
By using a variety of proxies for the information environment, we are able to examine the
relationship between several aspects of the information environment and the quality of a
firm’s corporate governance. Using a sample of Chinese-listed firms over the 2003-2011
period, we find that firms with higher levels of corporate governance are associated with
larger analyst followings and more informative forecasts. In addition, our results indicate that
better governed firms improve the timeliness of bad news relative to good news.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
25
The results of an instrumental variables analysis suggest a causal relationship between the
quality of corporate governance and the informativeness of a firm, thus posing the benefits in
terms of information transparency. In this regard, our findings have useful implications for
corporate managers, market participants, and policymakers.
There are several pertinent directions for future research. For example, would good corporate
governance criteria work the same way in China? We consulted previous works and used
their criteria when building our own CG score for Chinese firms. Those “good governance”
criteria, however, were designed primarily for developed countries. Many cross-country
studies use corporate governance data from Institutional Shareholder Services (ISS), which is
a database covering OECD countries. Given the differences in institutions, business
environment, and culture, it is certainly possible that the governance provisions may work
differently in China. Hence, further exploration of China-specific governance attributes and
criteria would be useful.
One could explore which channels of corporate governance affect the information
transparency of Chinese firms. Do better governed firms have higher disclosure levels in
terms of frequency and amount? This is likely to result in better informational efficiency.
Examining possible working paths would provide an even clearer picture of how corporate
governance works in China.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
26
References
Abdelsalam, O., Street, D., 2007. Corporate governance and the timeliness of corporate
internet reporting by UK listed companies. Journal of International Accounting,
Auditing and Taxation 16, 111-130.
Aggarwal, R., Erel, I., Ferreira, M., Matos, P., 2011. Does governance travel around the
world? Evidence from institutional investors. Journal of Financial Economics 100,
154-181.
Aguilera, R., Jackson, G., 2003. The cross-national diversity of corporate governance:
Dimensions and determinants. Academy of Management Review 28, 447-465.
Aharony, J., Wang, J., Yuan, H., 2010. Tunneling as an incentive for earnings management
during the IPO process in China. Journal of Accounting and Public Policy 29, 1-26.
Ajinkya, B., Bhojraj, S., Sengupta, P., 2005. The association between outside directors,
institutional investors and the properties of management earnings forecasts. Journal of
Accounting Research 43, 343-376.
Argenti, J., 1976. Corporation collapse: The causes and symptoms. John Wiley & Sons.
Ashbaugh, H., Pincus, M., 2001. Domestic accounting standards, international accounting
standards, and the predictability of earnings. Journal of Accounting Research 39,
417-434.
Bae, K., Bailey, W., Mao, C., 2006. Stock market liberalization and the information
environment. Journal of International Money and Finance 25, 404-428.
Baginski, S., Hassell, J., Kimbrough, M., 2002. The effect of legal environment on
voluntary disclosure: Evidence from management earnings forecasts issued in U.S.
and Canadian markets. Accounting Review 77, 25-50.
Bai, C., Liu, Q., Lu, J., Song, F., Zhang, J., 2004. Corporate governance and market valuation
in China. Journal of Comparative Economics 32, 599-616.
Ball, R., Brown, P., 1968. An empirical evaluation of accounting numbers. Journal of
Accounting Research 6, 159-178.
Barron, O., Byard, D., Kim, O., 2002. Changes in analysts’ information around earnings
announcements. Accounting Review 77, 821-846.
Beasley, M., 1996. An empirical analysis of the relation between the board of director
composition and financial statement fraud. Accounting Review 71, 443-465.
Beekes, W., Brown, P., 2006. Do better-governed Australian firms make more informative
disclosures? Journal of Business Finance and Accounting 33, 422-450.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
27
Beekes, W., Brown, P., 2007. On the timeliness of price discovery. Available at SSRN:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=938982.
Beekes, W., Brown, P., Chin, G., Zhang, Q., 2012a. The effects of corporate governance on
information disclosure, timeliness and market participants’ expectations. Available at
SSRN: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2122300.
Beekes, W., Brown, P., Zhan, W., Zhang, Q., 2012b. The relationship between disclosure,
information timeliness and corporate governance: A cross country study. Available at
SSRN: http://papers.ssrn.com/sol3/papers.cfm?abstract_id =2139641.
Bhat, G., Hope, O., Kang, T., 2006. Does corporate governance transparency affect the
accuracy of analyst forecasts? Accounting and Finance 46, 715-732.
Bhushan, R., 1989. Firm characteristics and analyst following. Journal of Accounting and
Economics 11, 255-274.
Blackburn, V., 1994. The effectiveness of corporate control in US corporations.
Corporate Governance: An International Review 2, 196-202.
Borokhovich, K., Parrino, R., Trapani, T., 1996. Outside directors and CEO selection.
Journal of Financial and Quantitative Analysis 31, 337-355.
Brennan, M., Hughes, P., 1991. Stock prices and the supply of information. Journal of
Finance 46, 1665-1691.
Brown, P., Beekes, W., Verhoeven, P., 2011. Corporate governance, accounting and finance:
A review. Accounting and Finance 51, 96-172.
Brown, L., Caylor, M., 2006. Corporate governance and firm valuation. Journal of
Accounting and Public Policy 25, 409-434.
Brown, L., Caylor, M., 2009. Corporate governance and firm operating performance. Review
of Quantitative Finance and Accounting 32, 129-144.
Buchanan, B., Le, Q., Rishi, M., 2012. Foreign direct investment and institutional quality:
Some empirical evidence. International Review of Financial Analysis 21, 81-89.
Bushee, B., Matsumoto, D., Miller, G., 2003. Open versus closed conference calls: The
determinants and effects of broadening access to disclosure. Journal of Accounting
and Economics 34, 149-180.
Bushee, B., Noe, C., 2000. Corporate disclosure practices, institutional investors, and
stock return volatility. Journal of Accounting Research 38, 171-202.
Bushman, R., Chen, Q., Engel, E., Smith, A., 2004. Financial accounting information,
organizational complexity and corporate governance systems. Journal of Accounting
and Economics 37, 167-201.
Byard, D., Li, Y., Weintrop, J., 2006. Corporate governance and the quality of financial
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
28
analysts’ information. Journal of Accounting and Public Policy 25, 609-625.
Chen, C., Jin, Q., Yuan, H., 2011. Agency problems and liquidity premium: Evidence from
China’s stock ownership reform. International Review of Financial Analysis 20, 76-
87.
Chen, G., Firth, M., Gao, D., Rui, O., 2006. Ownership structure, corporate governance, and
fraud: Evidence from China. Journal of Corporate Finance 12, 424-448.
Chen, G., Firth, M., Xu, L., 2009. Does the type of ownership control matter? Evidence from
China’s listed companies. Journal of Banking and Finance 33, 171-181.
Chen, K., Yuan, H., 2006. Market segmentation and the pricing of earnings. Hong Kong
University of Science and Technology, Working paper.
Cheung, Y., Jiang, P., Limpaphayom, P., Lu, T., 2008. Does corporate governance matter
in China? China Economic Review 19, 460-479.
Chung, K., Elder, J., Kim, J., 2010. Corporate governance and liquidity. Journal of Financial
and Quantitative Analysis 45, 265-291.
Claessens, S., Fan, J., 2002. Corporate governance in Asia: A survey. International Review
of Finance 3, 71-103.
Clarke, D., 2003. Corporate governance in China: An overview. China Economic Review
14, 494-507.
Conyon, M., He, L., 2011. Executive compensation and corporate governance in China.
Journal of Corporate Finance 17, 1158-1175.
Cornell, B., Landsman, W., 1989. Security price response to quarterly earnings
announcements and analysts’ forecast revisions. Accounting Review 64, 680-692.
CSRC, 2002. Code of corporate governance for listed companies in China. China Securities
Regulatory Commission.
Dechow, P., Sloan, R., Sweeney, A., 1996. Causes and consequences of earnings
manipulation: An analysis of firms subject to enforcement actions by the SEC.
Contemporary Accounting Research 13, 1-36.
Dharwadkar, B., George, G., Brandes, P., 2000. Privatization in emerging economies: An
agency theory perspective. Academy of Management Review 25, 650-669.
Ding, D., Sun, Q., 1997. The information content of FDI announcements: Evidence from an
emerging market. International Review of Financial Analysis 6, 63-76.
Eng, L., Mak, Y., 2003. Corporate governance and voluntary disclosure. Journal of
Accounting and Public Policy 22, 325-345.
Fama, E., 1980. Agency problems and the theory of the firm. Journal of Political
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
29
Economy 88, 288-307.
Fama, E., Jensen, M., 1983. Separation of ownership and control. Journal of Law and
Economics 26, 301-325.
Fan, J., Wong, T., 2005. Do external auditors perform a corporate governance role in
emerging markets? Evidence from East Asia. Journal of Accounting Research 43, 35-
72.
Fan, J., Wong, T., Zhang, T., 2007. Politically connected CEOs, corporate governance,
and Post-IPO performance of China’s newly partially privatized firms. Journal of
Financial Economics 84, 330-357.
Horton, J., Serafeim, G., Serafeim, I., 2013. Does mandatory IFRS adoption improve the
information environment? Contemporary Accounting Research 30, 388-423.
Hu, H., Tam, O, Tan, M., 2010. Internal governance mechanisms and firm
performance in China. Asia Pacific Journal of Management 27, 727-749.
Huddart, S., 1993. The effect of a large shareholder on corporate value. Management Science
39, 1407-1421.
Jensen, M., Meckling, W., 1976. Theory of the firm: Managerial behavior, agency costs
and ownership structure. Journal of Financial Economics 3, 305-360.
Jian, M., Wong, T., 2010. Propping through related party transactions. Review of
Accounting Studies 15, 70-105.
Jiang, G., Lee, C., Yue, H., 2010. Tunneling through inter-corporate loans: The China
experience. Journal of Financial Economics 98, 1–20.
Karamanou, I., Vafeas, N., 2005. The association between corporate boards, audit
committees, and management earnings forecasts: An empirical analysis. Journal of
Accounting Research 43, 453-486.
Kato, T., Long, C., 2006. CEO turnover, firm performance, and enterprise reform in China:
Evidence from micro data. Journal of Comparative Economics 34, 796-817.
Klein, A., 2002. Audit committee, board of director characteristics, and earnings
management. Journal of Accounting and Economics 33, 375-400.
Kothari, S., Shu, S., Wysocki, P., 2009. Do managers withhold bad news? Journal of
Accounting Research 47, 241-276.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000. Investor protection and
corporate governance. Journal of Financial Economics 58, 3-27.
Landsman, W., Maydew, E., Thornock, J., 2012. The information content of annual earnings
announcements and mandatory adoption of IFRS. Journal of Accounting and
Economics 53, 34-54.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
30
Lane, P., Milesi-Ferretti, G., 2007. The external wealth of nations mark II: Revised and
extended estimates of foreign assets and liabilities, 1970-2004. Journal of
International Economics 73, 223-250.
Lang, M., Lundholm, R., 1993. Cross-sectional determinants of analyst ratings of
corporate disclosures. Journal of Accounting Research 41, 317-345.
Lang, M., Lundholm, R., 1996. Corporate disclosure policy and analyst behavior.
Accounting Review 71, 467-492.
Lee, C., Xiao, X., 2004. Tunneling dividends. Available at SSRN:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=693361.
Li, K., Yue, H., Zhao, L., 2009. Ownership, institutions, and capital structure: Evidence from
China. Journal of Comparative Economics 37, 471-490.
Liu, Q., 2006. Corporate governance in China: Current practices, economic effects and
institutional determinants. CESifo Economic Studies 52, 415-453.
Liu, Q., Lu, Z., 2007. Corporate governance and earnings management in Chinese
listed companies: A tunneling perspective. Journal of Corporate Finance 13,
881-906.
Liu, C., Uchida, K., Yang, Y., 2012. Corporate governance and firm value during the global
financial crisis: Evidence from China. International Review of Financial Analysis 21,
70-80.
Lo, A., Wong, R., Firth, M., 2010. Can corporate governance deter management from
manipulating earnings? Evidence from related-party sales transactions in China.
Journal of Corporate Finance 16, 225-235.
McKinsey & Company, 2002. Corporate governance developments in emerging markets. McKinsey on Finance 3, 15-18.
Mendenhall, R., Nichols, W., 1988. Bad news and differential market reactions to
announcements of earlier-quarter versus fourth-quarter earnings. Journal of
Accounting Research 26, 63-86.
Mikhail, M., Walther, B., Willis, R., 1999. Does forecast accuracy matter to security
analysts? Accounting Review 74, 185-200.
Naceur, S., Ghazouani, S., Omran, M., 2007. The performance of newly privatized firms in
selected MENA countries: The role of ownership structure, governance and
liberalization policies. International Review of Financial Analysis 16, 332-353.
Noe, T., 2002. Investor activism and financial market structure. Review of Financial
Studies 15, 289-318.
Nowland, J., 2008. The effect of national governance codes on firm disclosure practices:
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
31
Evidence from analyst earnings forecasts. Corporate Governance: An International
Review 16, 475-491.
Peasnell, K., Pope, P., Young, S., 2000. Accrual management to meet earnings targets:
UK evidence pre-and post-Cadbury. British Accounting Review 32, 415-445.
Rajagopalan, N., Zhang, Y., 2008. Corporate governance reforms in China and India:
Challenges and opportunities. Business Horizons 51, 55-64.
Sami, H., Wang, J., Zhou, H., 2011. Corporate governance and operating performance of
Chinese listed firms. Journal of International Accounting, Auditing and Taxation 20,
106-114.
Shleifer, A., Vishny, R.W., 1986. Large shareholders and corporate control. Journal of
Political Economy 94, 461-488.
Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of
Finance 52, 737-783.
Skinner, D., 1994. Why firms voluntarily disclose bad news. Journal of Accounting
Research 32, 38-60.
Skinner, D., 1997. Earnings disclosures and stockholder lawsuits. Journal of Accounting and
Economics 23, 249-282.
Smith, M., 1996. Shareholder activism by institutional investors: Evidence from CalPERS.
Journal of Finance 51, 227-252.
Stickel, S., 1989. The timing of and incentives for annual earnings forecasts near interim
earnings announcements. Journal of Accounting and Economics 11, 275-292.
Tang, K., Wang, C., 2011. Corporate governance and firm liquidity: Evidence from the
Chinese stock market. Emerging Markets Finance and Trade 47, 47-60.
Weisbach, M., 1988. Outside directors and CEO turnover. Journal of Financial Economics
20, 431-460.
Wen, Y., Rwegasira, K., Bilderbeek, J., 2002. Corporate governance and capital structure
decisions of the Chinese listed firms. Corporate Governance: An International
Review 10, 75-83.
Woidtke, T., 2002. Agents watching agents? Evidence from pension fund ownership and firm
value. Journal of Financial Economics 63, 99-131.
Xu, N., Chan, K., Jiang, X., Yi, Z., 2013. Do star analysts know more firm-specific
information? Evidence from China. Journal of Banking and Finance 37, 89-102.
Xu, L., Tang, A., 2008. Internal control, material weakness, analysts’ accuracy and bias, and
brokerage reputation. Working paper, Morgan State University, Baltimore, MD.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
32
Yu, M., 2010. Analyst forecast properties, analyst following and governance disclosures: A
global perspective. Journal of International Accounting, Auditing and Taxation 19, 1-
15.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
33
Table 1. Corporate governance attributes and standards used to construct the aggregate governance score
Panel A: Corporate governance attributes
Mean
N = 12,649
Percentile
1% 5% 25% 50% 75% 95% 99%
1. Independent outside directors as the proportion of total number of board members 0.356 0.222 0.300 0.333 0.333 0.375 0.444 0.556
2. Total number of directors (including board chairman) on the company’s board of directors 9.309 5 6 9 9 10 13 15
3. Whether the same person holds both the board chairman and general manager roles
(1 = same person; 2 = different persons)
1.845 1 1 2 2 2 2 2
4. Whether there are any relationships among the top ten shareholders
(1 = no relationship; 2 = presence of a relationship; 3 = unknown)
2.428 1 1 2 3 3 3 3
5. Shares held by directors, supervisors, and executives as the proportion of total number of shares 0.040 0.000 0.000 0.000 0.00004 0.0004 0.371 0.640
6. Shares held by foreign investors as the proportion of total number of shares 0.012 0.000 0.000 0.000 0.000 0.000 0.054 0.333
7. Whether the auditor is part of a joint venture of the Big Four international audit firms and domestic audit firms
(1 = Big Four; 0 = Non-Big Four)
0.063 0.000 0.000 0.000 0.000 0.000 1.000 1.000
Panel B: Criteria used to construct the CG score
Analyst following
sample
Forecast accuracy
sample
Forecast dispersion
sample
Price timeliness
sample
1. Board is controlled by more than 50% independent directors 2% 2% 2% 1%
2. Board size is greater than 6 but fewer than 13 89% 89% 89% 88%
3. The chairman and general manager are not the same person 82% 82% 81% 85%
4. There are no relationships among the top ten shareholders 6% 5% 5% 8%
5. Management ownership (directors, supervisors, and executives) is greater than 1% but less than 30% 12% 12% 13% 7%
6. Foreign investor ownership is greater than zero 8% 7% 8% 6%
7. Firm is audited by one of the joint ventures of the Big Four international audit firms and domestic audit firms 9% 8% 9% 6%
Notes: This table reports the summary statistics of selected governance attributes and the scoring standards of the CG index. Panel A presents the summary statistics of governance attributes; panel B presents the proportion
of observations that meet the criterion for each governance attribute. The statistics in Panel A come from the sample of the price timeliness model because that sample has the most observations.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
34
Table 2. Summary statistics of the aggregate governance score and correlations of governance attributes
Panel A: Summary statistics of the corporate governance score (CG)
Variable
Mean
N = 12,649
Std. Dev. Percentile
1% 25% 50% 75% 99%
CG 2.014 0.680 0 2 2 2 4
INDIV 0.013 0.113 0 0 0 0 1
BOARDSIZE 0.882 0.323 0 1 1 1 1
DUAL 0.845 0.362 0 1 1 1 1
TOP10RELATION 0.080 0.271 0 0 0 0 1
MANAGEMENT 0.073 0.260 0 0 0 0 1
FOREIGN 0.058 0.234 0 0 0 0 1
BIG4 0.063 0.243 0 0 0 0 1
Panel B: Correlations of corporate governance attributes
[2] [3] [4] [5] [6] [7] [8]
CG [1] 0.154 0.448 0.493 0.351 0.307 0.326 0.364
INDIV [2] -0.016 -0.007 -0.021 -0.011 -0.011 0.042
BOARDSIZE [3] 0.0002 -0.029 0.033 -0.016 -0.056
DUAL [4] -0.007 -0.099 -0.044 0.052
TOP10RELATION [5] -0.045 -0.012 -0.014
MANAGEMENT [6] -0.009 -0.045
FOREIGN [7] 0.064
BIG4 [8]
Notes: This table provides the summary statistics and correlation coefficients of the corporate governance attributes. Panel A
presents the summary statistics, and panel B presents the Pearson correlation coefficients. The statistics in Panel A come from the
sample of the price timeliness model because that sample has the most observations. Characters in boldface indicate significance
at the 1% level.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
35
Table 3. Descriptive statistics of the information environment and control variables
Mean Median Std. Dev. Min Max Mean: Mean: Mean equality test
High CG score Low CG score t-stat
NUMEST 1.723 1.792 1.115 0.000 3.738 1.910 1.667 -8.347***
ACCURACY 0.009 0.003 0.019 0.000 0.132 0.007 0.010 5.973***
DISPERSION 0.008 0.005 0.010 0.000 0.066 0.007 0.008 3.613***
TIMELINESS 0.177 0.143 0.121 0.038 0.652 0.163 0.180 6.248***
SIZE 21.747 21.637 1.067 19.710 24.966 21.992 21.686 -13.501***
ROE 0.056 0.070 0.231 -1.379 0.956 0.073 0.052 -4.241***
LOSS 0.115 0.000 0.320 0.000 1.000 0.083 0.123 5.854***
LEV 0.512 0.496 0.310 0.047 2.401 0.471 0.522 7.668***
MB 3.412 2.556 3.386 -6.122 21.022 3.280 3.445 2.278**
RETVOL 0.024 0.023 0.007 0.011 0.043 0.023 0.025 11.461***
VOLUME 27.229 27.256 1.209 24.446 29.934 27.041 27.276 9.120***
INVPRICE 0.128 0.111 0.085 0.017 0.441 0.112 0.133 11.651***
HORIZON 4.208 4.500 1.236 0.000 5.938 4.113 4.235 3.753***
STATE 0.208 0.056 0.244 0.000 0.75 0.205 0.209 0.907
NONTRADE 0.436 0.496 0.256 0.000 0.832 0.501 0.419 -15.052***
TOP1 0.376 0.356 0.158 0.091 0.750 0.390 0.373 -5.056***
Notes: This table shows the summary statistics of the variables from 2003 through 2011. NUMEST is the natural logarithm of the number of unique analysts
covering a firm in a year; ACCURACY is the absolute value of the most recent forecast EPS minus the actual EPS of the year, deflated by the stock price at the
beginning of the year; DISPERSION is the standard deviation of the forecast EPS of the year, deflated by the stock price at the beginning of the year; SIZE is
the natural logarithm of market capitalization at the end of the year; ROE is the ratio of net income to the book value of total shareholder equity at the end of
the year; LOSS is a dummy variable equal to 1 if the net income of the year is negative, and 0 otherwise; LEV is the ratio of total liabilities to total assets at the
end of the year; MB is the ratio of market capitalization to the book value of total shareholder equity at the end of the year; RETVOL is the standard deviation
of daily stock returns over the 360 days prior to the end of the year; VOLUME is the natural logarithm of the trading volume of the year; INVPRICE is the
inverse of the mean stock price of the year; HORIZON is the natural logarithm of the number of days from the most recent forecast date until the EPS
announcement date of the year. STATE is state-owned shares as the proportion of total number of shares; NONTRADE is non-tradable shares as the proportion
of total number of shares; TOP1 is the shares held by the largest shareholder as the proportion of total number of shares. High (low) CG score refers to firms
with an aggregate CG score greater (lower) than the sample average. All time-varying variables are winsorized at the 1% and 99% levels. ***, **, and *
denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
36
Table 4. Correlation matrix
[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
COVERAGE [1] -0.253 -0.038 -0.048 0.054 -0.154 -0.030 0.034 0.515 0.232 -0.155 -0.144 0.133 0.022 0.125 0.392 -0.442 -0.482
ACCURACY [2] 0.326 0.073 -0.038 0.033 -0.084 -0.040 -0.117 -0.303 0.493 0.246 -0.019 0.054 0.105 -0.092 0.277 0.226
DISPERSION [3] 0.101 -0.013 0.051 -0.155 0.015 0.116 -0.048 0.231 0.256 -0.033 0.043 0.289 -0.054 0.228 -0.001
TIMELINESS [4] -0.046 0.013 0.053 -0.055 -0.053 -0.087 0.167 0.155 0.104 0.213 0.028 -0.041 0.096 -0.064
CG [5] 0.040 0.083 0.053 0.103 0.023 -0.045 -0.051 -0.034 -0.074 -0.032 -0.053 -0.066 -0.014
STATE [6] 0.439 0.468 0.018 -0.009 -0.008 -0.001 -0.101 -0.080 -0.105 -0.379 0.176 0.032
NONTRADE [7] 0.345 -0.160 0.017 -0.036 -0.124 -0.065 -0.076 -0.552 -0.425 0.037 -0.003
TOP1 [8] 0.209 0.075 -0.111 -0.096 -0.050 -0.105 -0.090 -0.120 -0.073 -0.050
SIZE [9] 0.222 -0.256 -0.159 0.219 0.051 0.562 0.435 -0.466 -0.352
ROE [10] -0.410 -0.035 -0.170 -0.024 0.025 0.076 -0.206 -0.181
LOSS [11] 0.320 0.024 0.059 -0.010 -0.084 0.297 0.102
LEV [12] -0.123 0.013 0.096 -0.074 0.344 0.056
MB [13] 0.174 0.059 0.252 -0.283 -0.122
RETVOL [14] 0.346 0.450 -0.169 -0.028
VOLUME [15] 0.477 0.012 -0.067
IFRS [16] -0.529 -0.219
INVPRICE [17] 0.266
HORIZON [18]
Notes: This table reports the simple correlations among the information environment variables and explanatory variables. See Table 3 for variable definitions. All time-varying variables are winsorized at the 1% and 99% levels.
Characters in boldface indicate significance at the 1% level.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
37
Table 5. Regression results for analyst following and corporate governance
Dependent variable COVERAGE
(1)
OLS
(2)
OLS
(3)
OLS
(4)
OLS
(5)
Fixed-effects
(6)
Fixed-effects
(7)
Fixed-effects
(8)
Fixed-effects
CG 0.038** 0.028* 0.057*** 0.053**
(2.45) (1.84) (2.69) (2.53)
INDIV -0.029 -0.020 0.061 0.045
(-0.35) (-0.25) (0.56) (0.40)
BOARDSIZE 0.034 0.036 0.086* 0.089*
(0.91) (0.99) (1.76) (1.84)
DUAL -0.088*** -0.057* -0.003 -0.006
(-2.99) (-1.92) (-0.08) (-0.14)
TOP10RELATION 0.050 0.034 0.020 0.020
(1.24) (0.84) (0.41) (0.40)
MANAGEMENT 0.184*** 0.144*** 0.091 0.099
(5.41) (4.31) (1.39) (1.49)
FOREIGN 0.108** 0.044 0.086 0.079
(2.54) (1.02) (1.61) (1.42)
BIG4 0.029 0.020 0.118* 0.090
(0.55) (0.39) (1.75) (1.35)
STATE -0.355*** -0.311*** -0.326*** -0.327***
(-5.55) (-4.79) (-3.86) (-3.77)
NONTRADE 0.308*** 0.264*** 0.072 0.079
(4.93) (4.16) (0.88) (0.92)
TOP1 -0.412*** -0.388*** -0.212 -0.196
(-4.51) (-4.26) (-1.04) (-0.97)
SIZE 0.584*** 0.605*** 0.586*** 0.609*** 0.562*** 0.612*** 0.559*** 0.607***
(29.77) (29.23) (29.09) (28.80) (16.58) (18.29) (16.56) (18.11)
ROE 0.252*** 0.270*** 0.256*** 0.265*** 0.236*** 0.239*** 0.235*** 0.244***
(2.98) (3.17) (3.00) (3.10) (2.64) (2.66) (2.62) (2.70)
LOSS -0.168*** -0.168*** -0.166*** -0.168*** -0.103** -0.100* -0.102** -0.098*
(-3.39) (-3.38) (-3.36) (-3.37) (-1.98) (-1.92) (-1.97) (-1.89)
LEV -0.263*** -0.172** -0.221*** -0.156** -0.107 -0.120 -0.099 -0.125
(-4.01) (-2.53) (-3.31) (-2.26) (-0.91) (-1.03) (-0.84) (-1.07)
MB -0.035*** -0.035*** -0.036*** -0.036*** -0.024*** -0.020*** -0.024*** -0.019***
(-6.98) (-6.86) (-7.22) (-7.12) (-4.10) (-3.49) (-4.09) (-3.38)
RETVOL -14.021*** -16.936*** -14.416*** -16.676*** -8.288** -6.658** -8.154** -4.655
(-5.43) (-6.45) (-5.62) (-6.36) (-2.54) (-2.41) (-2.49) (-1.49)
TRADE -0.122*** -0.092*** -0.107*** -0.088*** -0.077*** -0.070*** -0.077*** -0.072***
(-6.97) (-4.63) (-6.08) (-4.48) (-3.35) (-2.93) (-3.34) (-3.01)
IFRS 0.624*** 0.558*** 0.601*** 0.546*** 1.177*** 1.092*** 1.162*** 1.091***
(9.80) (8.67) (9.08) (8.19) (16.33) (13.83) (15.56) (13.14)
INVPRICE -4.656*** -4.575*** -4.555*** -4.493*** -2.483*** -2.293*** -2.464*** -2.208***
(-16.43) (-16.16) (-16.26) (-16.03) (-7.09) (-6.66) (-7.03) (-6.31)
Constant -7.962*** -9.053*** -8.299*** -9.167*** -9.237*** -10.398*** -9.152*** -10.264***
(-23.57) (-23.36) (-22.62) (-22.15) (-11.17) (-12.47) (-11.06) (-12.32)
N 8,265 8,265 8,265 8,265 8,265 8,265 8,265 8,265
Adjusted R-squared 0.52 0.53 0.52 0.53 0.47 0.47 0.47 0.47
F-test 372.38 345.80 300.11 281.83 225.03 222.40 169.67 162.24
Year dummy Yes Yes Yes Yes Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes No No No No
Notes: The table reports the regression results for analyst following. In models 1-4, we use pooled ordinary least squares (OLS) with standard errors clustered by firm, and
in models 5-8, we use fixed-effects (FE) regressions with standard errors clustered by firm. See Table 3 for variable definitions. All time-varying variables are winsorized at
the 1% and 99% levels. The values of t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
38
Table 6. Regression results for forecast accuracy and corporate governance
Dependent variable ACCURACY
(1)
OLS
(2)
OLS
(3)
OLS
(4)
OLS
(5)
Fixed-effects
(6)
Fixed-effects
(7)
Fixed-effects
(8)
Fixed-effects
CG -0.001*** -0.001*** -0.001*** -0.001***
(-2.87) (-2.78) (-2.99) (-2.94)
INDIV -0.001 -0.001 -0.004 -0.005
(-1.33) (-1.32) (-1.42) (-1.51)
BOARDSIZE -0.0004 -0.0004 -0.002** -0.002**
(-0.68) (-0.69) (-2.31) (-2.31)
DUAL -0.0004 -0.0003 -0.001 -0.001
(-0.81) (-0.78) (-0.66) (-0.70)
TOP10RELATION 0.001 0.001 0.0002 0.0003
(1.00) (1.05) (0.21) (0.30)
MANAGEMENT -0.001*** -0.001*** -0.0001 -0.001
(-2.94) (-2.86) (-0.09) (-0.72)
FOREIGN -0.001** -0.001** -0.003*** -0.002*
(-2.24) (-2.08) (-2.63) (-1.72)
BIG4 -0.001* -0.001* -0.001 -0.002
(-1.67) (-1.70) (-0.69) (-0.97)
STATE -0.0003 -0.001 0.0004 0.0002
(-0.31) (-0.59) (0.23) (0.11)
NONTRADE -0.001 -0.0005 -0.004*** -0.004**
(-0.76) (-0.49) (-2.75) (-2.45)
TOP1 0.001 0.001 -0.012*** -0.012***
(0.78) (0.73) (-2.58) (-2.60)
SIZE -0.001*** -0.001*** -0.001*** -0.001** -0.001 0.001 -0.001 0.001
(-3.10) (-2.85) (-2.69) (-2.46) (-1.00) (0.92) (-0.96) (0.94)
ROE -0.009** -0.009** -0.009** -0.009** -0.015*** -0.016*** -0.015*** -0.015***
(-2.06) (-2.07) (-2.07) (-2.08) (-3.16) (-3.33) (-3.15) (-3.32)
LOSS 0.038*** 0.038*** 0.038*** 0.038*** 0.030*** 0.030*** 0.030*** 0.030***
(14.32) (14.31) (14.31) (14.31) (11.38) (11.31) (11.40) (11.32)
LEV 0.010*** 0.010*** 0.010*** 0.010*** 0.008** 0.008** 0.008** 0.008**
(8.87) (8.26) (8.75) (8.26) (2.07) (2.24) (2.05) (2.23)
MB 0.0001 0.0001 0.0001 0.0001 -0.0001 0.00004 -0.0001 0.00004
(0.49) (0.43) (0.45) (0.40) (-0.34) (0.17) (-0.37) (0.16)
RETVOL 0.086* 0.093** 0.084* 0.089* 0.149** 0.099** 0.152** 0.098**
(1.92) (2.00) (1.88) (1.91) (2.22) (2.29) (2.27) (2.26)
TRADE 0.001*** 0.001*** 0.001*** 0.001*** 0.001** 0.001*** 0.001** 0.001***
(3.77) (2.86) (3.47) (2.82) (2.21) (3.59) (2.05) (3.57)
IFRS -0.0001 -0.0002 -0.00003 -0.0001 -0.0004 -0.005*** 0.0003 -0.004***
(-0.08) (-0.13) (-0.03) (-0.10) (-0.29) (-3.24) (0.23) (-2.73)
HORIZON 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002***
(13.55) (13.52) (13.56) (13.54) (10.89) (10.96) (10.93) (10.98)
Constant -0.008 -0.007 -0.009 -0008 -0.007 -0.046*** -0.007 -0.047***
(-1.38) (-0.96) (-1.39) (-1.13) (-0.37) (-2.70) (-0.35) (-2.71)
N 8,402 8,402 8,402 8,402 8,402 8,402 8,402 8,402
Adjusted R-squared 0.32 0.32 0.32 0.32 0.23 0.22 0.23 0.22
F-test 50.23 45.12 39.52 36.36 29.52 26.73 22.08 19.96
Year dummy Yes Yes Yes Yes Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes No No No No
Notes: This table reports the regression results for forecast accuracy. In models 1-4, we use pooled ordinary least squares (OLS) with standard errors clustered by firm; for models 5-8, we use
fixed-effects (FE) regressions with standard errors clustered by firm. See Table 3 for variable definitions. All time-varying variables are winsorized at the 1% and 99% levels. The values of t-
statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
39
Table 7. Regression results for forecast dispersion and corporate governance
Dependent variable DISPERSION
(1)
OLS
(2)
OLS
(3)
OLS
(4)
OLS
(5)
Fixed-effects
(6)
Fixed-effects
(7)
Fixed-effects
(8)
Fixed-effects
CG -0.0004*** -0.0004** -0.0002 -0.0003
(-2.58) (-2.56) (0.83) (-0.90)
INDIV -0.002** -0.002** -0.003** -0.003**
(-2.36) (-2.37) (-2.15) (-2.27)
BOARDSIZE -0.001 -0.001 -0.001 -0.001
(-1.53) (-1.55) (-1.64) (-1.56)
DUAL -0.0003 -0.0003 -0.001 -0.001
(-1.18) (-1.03) (-0.84) (-0.88)
TOP10RELATION 0.0003 0.0003 0.0001 0.0001
(0.62) (0.73) (0.22) (0.14)
MANAGEMENT -0.001** -0.001** -0.001** -0.001*
(-2.28) (-2.31) (-2.05) (-1.67)
FOREIGN 0.00003 0.00003 0.002** 0.002**
(0.07) (0.07) (2.26) (2.14)
BIG4 -0.0002 -0.0003 0.001 0.001
(-0.40) (-0.46) (0.95) (0.87)
STATE -0.001* -0.001* -0.003** -0.002*
(-1.76) (-1.69) (-2.25) (-1.70)*
NONTRADE -0.0003 -0.0003 0.001 -0.0001
(-0.39) (-0.48) (0.56) (-0.06)
TOP1 0.001 0.001 0.006* 0.006*
(1.35) (1.46) (1.77) (1.73)
SIZE 0.0001 0.0001 0.0001 0.0001 -0.0004 -0.0004 -0.0005 -0.0004
(0.49) (0.61) (0.30) (0.43) (0.77) (-0.67) (-0.93) (-0.78)
ROE 0.005 0.004 0.005* 0.004 0.003 0.003 0.003 0.003
(1.62) (1.57) (1.65) (1.60) (0.86) (0.84) (0.94) (0.91)
LOSS 0.012*** 0.012*** 0.012*** 0.012*** 0.011*** 0.011*** 0.011*** 0.011***
(8.41) (8.44) (8.40) (8.43) (6.62) (6.66) (6.60) (6.64)
LEV 0.005*** 0.005*** 0.005*** 0.005*** 0.003* 0.003 0.004** 0.003*
(7.00) (6.66) (6.95) (6.63) (1.84) (1.55) (1.97) (1.67)
MB -0.0001 -0.0001 -0.0001 -0.0001 0.0002 0.0002 0.0002 0.0002
(-0.94) (-1.02) (-0.82) (-0.91) (1.42) (1.34) (1.49) (1.37)
RETVOL 0.070** 0.074** 0.068** 0.073** 0.083** 0.081** 0.082** 0.083**
(2.48) (2.48) (2.43) (2.46) (2.10) (2.04) (2.08) (2.10)
TRADE 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002***
(12.65) (10.25) (12.46) (10.32) (6.28) (6.48) (6.38) (6.36)
IFRS -0.004*** -0.004*** -0.003*** -0.004*** -0.0003 -0.0003 -0.0001 -0.0001
(-4.80) (-4.84) (-4.20) (-4.26) (0.29) (-0.28) (-0.07) (-0.09)
Constant -0.047*** -0.047*** -0.047*** -0.047*** -0.031** -0.038*** -0.029** -0.035***
(-11.60) (-10.18) (-10.89) (-9.67) (-2.57) (-3.07) (-2.44) (-2.82)
N 6,868 6,868 6,868 6,868 6,868 6,868 6,868 6,868
Adjusted R-squared 0.19 0.19 0.19 0.19 0.12 0.12 0.12 0.12
F-test 47.49 43.04 37.23 34.60 26.58 22.50 19.68 17.42
Year dummy Yes Yes Yes Yes Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes No No No No
Notes: This table reports the regression results for forecast dispersion. In models 1-4, we use pooled ordinary least squares (OLS) with standard errors clustered by firm; for models 5-8, we use
fixed-effects (FE) regressions with standard errors clustered by firm. See Table 3 for variable definitions. All time-varying variables are winsorized at the 1% and 99% levels. The values of t-
statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
40
Table 8. Regression results for price timeliness and corporate governance
Dependent variable TIMELINESS
(1)
OLS
(2)
OLS
(3)
OLS
(4)
OLS
(5)
Fixed-effects
(6)
Fixed-effects
(7)
Fixed-effects
(8)
Fixed-effects
CG -0.003* -0.003** -0.0003 -0.0003
(-1.82) (-2.08) (0.15) (-0.14)
INDIV -0.014** -0.013** -0.023** -0.021*
(-2.29) (-2.26) (-2.10) (-1.96)
BOARDSIZE 0.0001 0.0003 0.007 0.007
(0.03) (0.09) (1.34) (1.39)
DUAL -0.003 -0.001 -0.0004 -0.0003
(-0.92) (-0.34) (-0.08) (-0.07)
TOP10RELATION -0.001 -0.002 0.005 0.005
(-0.31) (0.49) (1.06) (1.11)
MANAGEMENT 0.002 -0.001 -0.008 -0.009
(0.42) (-0.39) (-0.91) (1.12)
FOREIGN -0.001 -0.005 -0.007 -0.008
(-0.13) (-1.03) (-0.92) (1.03)
BIG4 -0.014*** -0.014*** -0.009 -0.009
(-3.39) (-3.40) (-1.23) (-1.22)
STATE -0.015** -0.016** 0.019** 0.018*
(-2.40) (-2.46) (1.97) (1.82)
NONTRADE 0.014** 0.014** 0.002 0.004
(2.32) (2.29) (0.22) (0.42)
TOP1 -0.024*** -0.024*** -0.040** -0.042**
(-2.98) (-2.93) (-2.15) (-2.27)
SIZE 0.008*** 0.009*** 0.009*** 0.011*** 0.010*** 0.010*** 0.010*** 0.010***
(5.23) (5.73) (5.67) (6.11) (2.96) (3.08) (2.96) (3.06)
ROE -0.008 -0.009 -0.009 -0.009 -0.018** -0.018** -0.018** -0.018**
(-1.09) (-1.17) (-1.18) (-1.26) (-2.38) (-2.38) (-2.38) (-2.38)
LOSS 0.042*** 0.042*** 0.041*** 0.041*** 0.022*** 0.022*** 0.022*** 0.022***
(9.75) (9.80) (9.63) (9.69) (4.71) (4.71) (4.65) (4.65)
LEV 0.052*** 0.053*** 0.053*** 0.053*** 0.044*** 0.044*** 0.044*** 0.043***
(11.17) (11.24) (11.23) (11.24) (4.80) (4.74) (4.74) (4.68)
MB 0.003*** 0.003*** 0.003*** 0.003*** 0.002*** 0.002*** 0.002*** 0.002***
(6.63) (6.48) (6.44) (6.28) (4.66) (4.61) (4.60) (4.54)
RETVOL 4.036*** 3.967*** 3.972*** 3.912*** 3.985*** 3.994*** 3.841*** 3.830***
(12.85) (12.66) (12.38) (12.19) (20.60) (20.45) (15.54) (15.47)
TRADE -0.009*** -0.008*** -0.008*** -0.008*** -0.007*** -0.008*** -0.008*** -0.008***
(-5.67) (-5.28) (-4.76) (-4.61) (-3.07) (-3.09) (-2.93) (-2.91)
IFRS -0.077*** -0.076*** -0.076*** -0.075*** -0.067*** -0.068*** -0.064*** -0.064***
(-11.96) (-11.88) (-11.65) (-11.53) (-14.36) (-14.34) (-11.92) (-11.86)
Constant 0.160*** 0.126*** 0.120*** 0.089** 0.097 0.086 0.103 0.090
(4.68) (3.44) (3.30) (2.32) (1.32) (1.17) (1.40) (1.23)
N 12,649 12,649 12,649 12,649 12,649 12,649 12,649 12,649
Adjusted R-squared 0.13 0.13 0.13 0.13 0.09 0.09 0.09 0.09
F-test 88.65 68.95 78.84 63.06 85.80 58.04 71.09 51.67
Year dummy Yes Yes Yes Yes Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes No No No No
Notes: This table reports the regression results for price timeliness. See Table 3 for variable definitions. All time-varying variables are winsorized at the 1% and 99% levels. The values of t-
statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
41
Table 9. Effects of corporate governance on information environment, instrumental variable (IV) estimates
Dependent variable COVERAGE ACCURACY DISPERSION TIMELINESS
(1) (2) (3) (4) (5) (6) (7) (8)
CG 0.028 0.020 -0.001* -0.001* -0.001** -0.001** -0.003 -0.004
(1.16) (0.84) (-1.95) (-1.88) (-2.12) (-2.19) (-1.37) (-1.57)
STATE -0.248*** 0.0003 -0.002* -0.016**
(-3.44) (0.02) (-1.92) (-2.30)
NONTRADE -0.038 -0.002 0.0002 0.013**
(-0.55) (-1.61) (0.32) (2.00)
TOP1 -0.416*** 0.001 0.001 -0.028***
(-3.95) (0.75) (1.16) (-3.16)
SIZE 0.556*** 0.619*** -0.001*** -0.001*** 0.0003 0.0003 0.007*** 0.009***
(23.34) (24.30) (-3.11) (-2.74) (1.47) (1.38) (4.50) (5.16)
ROE 0.362*** 0.334*** -0.008* -0.008* 0.004 0.004 -0.009 -0.010
(3.89) (3.61) (-1.76) (-1.78) (1.52) (1.50) (-1.13) (-1.24)
LOSS -0.122 -0.130** 0.037*** 0.037*** 0.012*** 0.012*** 0.041*** 0.041***
(-2.33) (-2.47) (13.59) (13.59) (8.08) (8.14) (8.75) (8.64)
LEV -0.065 -0.051 0.012*** 0.011*** 0.006*** 0.006*** 0.053*** 0.053***
(-0.88) (-0.68) (8.50) (7.99) (6.26) (6.14) (10.30) (10.33)
MB -0.022*** -0.027*** 0.0001 0.0001 -0.0001 -0.0001 0.003*** 0.003***
(-3.88) (-4.76) (0.82) (0.74) (-1.32) (-1.33) (5.65) (5.42)
RETVOL -31.528*** -28.118*** 0.084 0.100 0.155*** 0.153*** 4.002*** 3.966***
(-8.65) (-7.63) (1.30) (1.49) (3.63) (3.41) (11.89) (11.56)
TRADE -0.008 -0.049* 0.001*** 0.001** 0.002*** 0.002*** -0.009*** -0.009***
(-0.31) (-1.87) (2.81) (1.98) (7.30) (6.73) (-5.08) (-4.52)
IFRS 0.007 -0.048 -0.006*** -0.006*** -0.003*** -0.004*** -0.076*** -0.077**
(0.15) (-0.98) (-6.63) (-6.60) (-5.31) (-5.64) (-17.11) (-16.64)
INVPRICE -5.296*** -4.968***
(-17.28) (-16.12)
HORIZON 0.002*** 0.002***
(12.93) (12.88)
Constant -9.028*** -9.125*** -0.005 -0.001 -0.044*** -0.046*** 0.175*** 0.139***
(-21.25) (-19.64) (-0.64) (-0.14) (-8.95) (-8.38) (4.65) (3.50)
Hansen J-statistic (p-value) 0.19 0.18 0.21 0.21 0.65 0.71 0.19 0.16
Underidentification test (p-value) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
N 6,751 6,751 6,907 6,907 5,614 5,614 11,010 11,010
Adjusted R-squared 0.52 0.52 0.32 0.32 0.17 0.17 0.12 0.12
F-test 302.14 278.95 46.40 41.23 35.05 30.88 76.00 67.33
Year dummy Yes Yes Yes Yes Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes Yes Yes Yes Yes
Notes: This table reports the instrumental variables (IV) results for analyst following, forecast accuracy, forecast dispersion, and price timeliness. See Table 3 for variable definitions. All time-varying variables are winsorized
at the 1% and 99% levels. The values of t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
42
Table 10. Effects of corporate governance on timeliness measures (good news versus bad news)
Dependent variable TIMELINESS
(Good news)
TIMELINESS
(Bad news)
(1) (2) (3) (4)
CG -0.00001 0.0001 -0.001*** -0.001***
(-0.02) (0.19) (-2.87) (-2.72)
STATE 0.008*** 0.002
(3.74) (1.19)
NONTRADE -0.004* -0.007***
(-1.91) (-4.22)
TOP1 0.001 -0.010***
(0.36) (-4.27)
SIZE -0.008*** -0.008*** 0.006*** 0.007***
(-14.07) (-13.17) (14.49) (15.85)
ROE -0.009*** -0.008*** -0.004** -0.004**
(-4.20) (-4.10) (-2.10) (-2.27)
LOSS 0.002 0.002 0.009*** 0.009***
(1.10) (1.13) (8.49) (8.44)
LEV -0.003* -0.003** 0.004*** 0.004***
(-1.96) (-2.08) (3.72) (3.41)
MB -0.001*** -0.001*** 0.001*** 0.001***
(-5.62) (-5.49) (6.94) (6.24)
RETVOL -2.666*** -2.632*** -1.177*** -1.087***
(-22.43) (-21.82) (-13.01) (-11.78)
TRADE 0.002*** 0.002** -0.006*** -0.007***
(3.50) (2.47) (-13.90) (-14.61)
IFRS 0.001 0.001 -0.038*** -0.039***
(0.56) (0.61) (-25.09) (-25.81)
Constant 0.693*** 0.708*** 0.558*** 0.574***
(52.25) (49.27) (61.11) (55.43)
N 12,649 12,649 12,649 12,649
Adjusted R-square 0.20 0.20 0.32 0.32
F test 163.38 142.33 366.76 326.00
Year dummy Yes Yes Yes Yes
Industry dummy Yes Yes Yes Yes
Notes: This table reports the regression results for the timeliness measures (good news versus bad news). See Table 3 for
the variable definitions. All time-varying variables are winsorized at the 1% and 99% levels. The estimation method is
pooled ordinary least squares (OLS) with standard errors clustered by firm. The values of t-statistics are reported in
parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively.
ACC
EPTE
D M
ANU
SCR
IPT
ACCEPTED MANUSCRIPT
43
Highlights:
Corporate governance is important for the information environment of Chinese firms
Good corporate governance increases analyst following and analyst forecast accuracy and
decreases analyst forecast dispersion
Good corporate governance improves timeliness of bad news relative to good news