Voluntary disclosures and corporate governance characteristics evidence from vietnam

Voluntary Disclosures and Corporate Governance
Characteristics: Evidence from Vietnam
Phuong Le
Linh
Nguyen
International University, Vietnam National University HCMC, Vietnam
Abstract
Voluntary disclosure practice has become more important especially in developing markets
where there is a lack of studies on voluntary disclosure and its determinants. This research is
conducted with the aim to provide a better understanding of voluntary disclosure and its
relationship with corporate governance characteristics in Vietnam. Our study is based on a
sample of 100 largest market capitalization companies during a three- year period starting
from 2014 to 2016. Analysis indicates a moderate extent of voluntary disclosure compared
to other countries in different studies. Foreign ownership is found to have a significantly
positive link with the volume of voluntary disclosure. The study also reveals significant
relations between board independence, CEO separation, audit committee existence, audit
quality and the

extent of sub- categories of voluntary disclosure. Interestingly, there are

significantly positive links between audit committee presence, audit quality, foreign ownership
and voluntary disclosure in the following year. Our findings are expected to contribute to
literature on voluntary disclosure in Vietnam. They can be helpful for policy makers

to

gradually improve the voluntary disclosure practice as well as information asymmetry in
Vietnam market.
Keywords: Voluntary Disclosure, Corporate Governance, Vietnam, VN100.
1.Introduction
After some worldwide accounting and financial reporting scandals like Enron, WorldCom, and

Paramalat in 2012, scholars have raised ideas about the concept “corporate governance”. The
governance of a corporation is said to be the sum of actions forming its regulations which are
complied with legislation, ownership and control (Cannon, 1994). According to Apostolou &
Nanopoulos (2009), poor corporate governance was blame for most world crises. In this thesis,
it focuses on transparency and disclosure aspect, one of five pillars of good corporate
governance. Companies prove their transparency by disclosing as much information as
possible through mandatory and voluntary disclosure. Mandatory disclosure is to comply with
laws and regulations, therefore all companies have to disclose all information required.
However, voluntary disclosure is based on the willingness of the managers, so it is needed to
be studied more in accounting field.
Companies are encouraged to voluntarily supply additional information because it
generates many benefits. Corporations can lower their cost of capital when raising capital
outside (Francis, Nanda, & Olsson, 2008). Firms can make their value increased by disclosing
more information to the public (Healy & Palepu, 1993). It gives a good effect on companies’

1


reputation, attracts more investors, and lowers cost of capital when companies give out more
information as doing this will distinguish companies form other competitors in the industry
(Hawashe & Ruddock, 2014).
It is not companies who benefit from voluntary disclosure, but investors and other stake
holders gain benefits as well. Disclosures increase market transparency which is considered as
basic mechanism to reduce

2


the information asymmetry among the participants in the market (Bleck & Liu, 2007). Investors
always have uncertainty about the quality of firm as well as their securities, so they need

adequate information value the firms, judge the opportunities and riskiness of their
investments to come up to best investing decisions (Al- Janadi, Rahman, & Omar, 2013; Meek,
Roberts, & Gray, 1995). Moreover, voluntary disclosure is believed to provide a clearer view
about business’ sustainability, reduce agency conflicts between managers and investors
(Boesso & Kumar, 2007; Healy & Palepu, 2001).
The association between corporate governance and optional disclosures has become
headlines in many journals in recent years. There are many studies conducted in developed
countries and most of them give mixed results. S. S. Ho & Wong (2001) found that the
existence of audit committee is significantly and positively related to the level of voluntary
disclosure of listed firms in Hong Kong. A significant association between some corporate
governance variables and voluntary disclosure of Swedish companies was found by (Cooke,
1989). It is found that board composition and the extent of voluntary disclosure of
information in annual reports of 181 Australian companies had positive relation (Lim, Matolcsy,
& Chow, 2007). There have been other scholars who found positive results, such as (Chau &
Gray, 2002; Chen & Jaggi, 2001; Forker, 1992,; Haniffa & Cooke, 2002; Klein, 2002), etc.
However, there are many studies which found negative relation like (Barako, Hancock, & Izan,
2006; Eng & Mak, 2003; Gul & Leung, 2004; Haniffa & Cooke, 2002).
Although this topic has been considered as an increasingly important topic in the accounting
field in recent years, there are a few studies on this topic that are conducted by Vietnamese
scholars. Vu (2012) through her thesis findings suggested that corporate governance can be
considered as an effective monitoring method to increase the voluntary disclosure practice.
Hieu & Lan (2015) found that board size had a positive link with the extent of voluntary
disclosure, while board independence, role duality, and type of auditors were found to have no
significant associations.
In recent years, corporate governance practice has attracted more and more attention.
Vietnam has jointed ASEAN corporate governance scorecard since 2011. Moreover, decree
71/2017 NĐ-BTC about corporate disclosure took effect in 2017.

In summary, the shortage of research on the relationship between corporate governance attributes
and optional disclosure together with the importance of corporate governance in today business

motivate for conducting this research.
This study is conducted to mainly examine the association between corporate governance
mechanisms and voluntary disclosure practice in the case of VN100 companies listed in Ho Chi
Minh stock exchange from 2014 to 2016. In the process, the author also observes the
voluntary disclosure index of these companies. Research questions are formed based on
mentioned objectives:
 To what extent do listed firms in Vietnam voluntarily disclose?
 What is the relationship between corporate governance characteristics and the
extent of voluntary disclosure of listed firms in Vietnam?
2.Literature Review
2.1.

Definitions of Voluntary Disclosure

Voluntary disclosure is defined as an enlargement and supplement of compulsory disclosure
(Tian & Chen, 2009), which is not affected by any authority body (Scaltrito, 2015).
Financial Accounting Standards Board (FASB) defines that the information that is disclosed by
listed firms rather than to be obligated to disclose by laws, standards and regulations of
authority committees.


2.2.

Relevant Theories


2.2.1.

Vietnam Corporate Governance Regulations

“Listed company” is defined as company whose shares are listed on one stock exchange.
Listed companies are regulated by Law on Enterprise 2005 (with the revision on 2014), and
Corporate Governance Code 2007 (with the amendments on 2012) issued by Ministry of
Finance.
The Law of Securities 2016 applied main standards for listed companies to practice
better corporate governance. These standards are:
 Internal governance structure.
 Rights of shareholders.
 Conflicts of interests and related parties’ transactions.
 Information disclosure and transparency.
Circular No 52/2012/TT-BTC took effect on 1st June 2012. This circular provides each type
of entity in Vietnamese market with requirements and guidelines in disclosing information,
such as: what information must be publicized, who is authorized to do, and which forms are
appropriate to use.
Circular No 71/2017/TT-BTC was issued in July, 2017 to amend the circular 52 in public
companies’ section only. The new requirements are related to general shareholders, board of
directors, board of advisory, related parties’ transactions, and information disclosure.
Moreover, all listed companies have to follow listing rules of stock exchanges that they are
listed on. In Vietnam, there are two largest stock exchanges which are Ho Chi Minh Stock
Exchange (HOSE) and Ha Noi Stock Exchange (HNX).
2.2.2.

Agency Theory

Agency theory, which is widely used to study on accounting and finance literature, was
developed by (Jensen & Meckling, 1976). The key concept of this theory is the agency
relationship that is said as a contract between the principal and the agent. In the contract,
principal gives the agent an authority to perform some tasks on behalf of them. In the
organizational context, the principal is shareholders and the agent is managers. The
authorization leads to the separation between ownership and control, which is the main

cause of information asymmetry. Information asymmetry is the situation that the managers
possess more information about company than the shareholders. Having information
advantages, agents may act as their own interests instead of principals’ interests if principals
cannot effectively monitor the managers’ behaviors. Agency problems can be categorized as
moral hazard and adverse selection.
 Moral hazard or hidden cost: It is a situation that the principal cannot access the
agent’s performance directly but only based on the outcome. In this case, the
agents can have abilities to give themselves more benefits.(Folkare & Andersson,
2015)
 Adverse selection: It is the situation that the principal is able to observe the agent’s
behaviors, but unable to determine whether these behaviors are the most appropriate ones.
(Subramaniam, 2006).
Agency problems cause some costs for corporations to mitigate them and these costs are
generally named agency costs. Agency costs can be classified into three kinds:
 Monitoring costs: These costs are incurred with the aim to monitor the agent’s
behaviors. For example: audit costs.
 Bonding costs: These costs are related to incentives that given to agents to align
their interests to principals’ interests. For example: stock options.
 Residual costs: These costs are arisen when there are conflicts of interests between
principal and agent despite monitoring and bonding processes.


In general, agency problem can be minimized by two main and common strategies:
monitoring-related and incentive-focused strategies. In monitoring-related strategies,
voluntary disclosure can be considered as a


mechanism to reduce information asymmetry. This mechanism suggests that the agent
voluntarily discloses a significant amount of information to the principal, mostly under many
kinds of repots.

2.2.3.

Signaling Theory

Signaling theory was first suggested in solving information asymmetry in labor market by
Spence (1973). However, the theory is used in corporate financial reporting by Ross (1977). In
his study, Ross uses signaling theory to explain for voluntary disclosure. Following him, there
have been more and more researchers that use signaling theory as a literature review when
studying on voluntary disclosure.
Disclosing more information to outside on voluntary basis is considered as a signaling
mechanism as a firm can give a signal that it is performing better than other competitors in
the industry (Campbell, Shrives, & Bohmbach-Saager, 2001).
Signaling theory suggests that voluntary disclosure can be considered as a signal to improve
the company’s image or reputation, and relevant relationship with many other stakeholders
(Hawashe & Ruddock, 2014).
2.2.4.

Capital Need Theory

Listed companies commonly finance their capital by borrowings or equity. Capital need
theory says that voluntary disclosure can help companies to finance capital at a low cost
(Choi, 1973). This idea can be explained based on information asymmetry. The cost of
capital of a firm includes a premium for investors’ uncertainty about information given by the
firms. The more uncertain the investors feel, the higher the cost of capital. If managers who
possess more reliable information are willing to share a significant amount of this information
to investors, investors do not face with information asymmetry and they are confident when
investing to companies. As a result, the cost of capital is reduced (Financial Accounting
Standards Board, 2001).
According to the theory, the increase in voluntary disclosure can help companies to
attract more new investors, which helps maintain a healthy demand for companies’ shares.

Moreover, companies with higher degree of disclosure tend to gain higher stock price in the
long run (Cooke, 1989).
2.2.5.

Legitimacy Theory

Magness (2006) suggests that legitimacy theory is a contract between a company and the
society in which it is operating. Coebergh (2011) gives out an opinion that low legitimacy may
cause the firms to discontinue operations if firms do business under the expectations of the
society in which firms are operating. In his study, he suggests that investors should be well
informed about on-going situation of firms. Therefore, managers should disclose more
information to outside investors to protect companies’ legitimacy.
The annual report is said as a main source of legitimation (Cadiz Dyball, 1998).
Legitimation can comes from both mandatory disclosure and voluntary disclosure (Magness,
2006).
2.3. Corporate
Governance
Characteristics
(Board
Independence,
CEO
Separation, Audit Committee Existence, Audit Quality, and Foreign Ownership) and
Voluntary Disclosure.
First, the study will review how board independence is related to level of voluntary
disclosure. Board independent is characterized as the percentage of outside members to total
members of the board. A board of directors consists of two main kinds of members: inside
members and outside members. Outside directors, independent directors, are defined in
circular 121/2012/TT-BTC as following:






Not hold any position in management board (non-executive).
Not a member of board, CEO, managers of subsidiaries, associated companies.
Not a large shareholders or related persons of the companies’ major shareholders.
Not working for organizations providing legal advisory services, audit organization for
company in the most recent two years.


Not a relevant partner or partners who have an annual value of transactions with
companies accounted for 30% or more of total revenue or total value of products
purchased by company the most recent two years.
If dependent members have specialized skills, valuable experiences in industry in which
the firm is operated, and deep knowledge about firm’s activities, independent members play a
controlling role to monitor actions of executive members (Jensen & Meckling, 1976). Another
study Fama & Jensen (1983) suggests that independent directors act as check and balance
mechanism in enhancing the effectiveness of monitoring system. Moreover, Adams &
Hossain, 1998; Klein (2002) both say that high rate of independent members can reduce
accounting fraud and earnings management.
H1: There is a positive relation between board independence and the extent of voluntary
disclosure.
CEO duality is a situation that one person holds two positions CEO and Chairman
simultaneously. There is a high likelihood that conflicts of interests between these two
positions may cause agency problem to increase. The chairman is expected to represent the
shareholders to oversee the CEO’s performance. In the case of leadership duality, the CEO
may put his interests on first priority, which may violate the shareholders’ interests. Therefore,
Gul & Leung (2004) suggest that one person cannot hold two positions at the same time, and

by separating roles, chairman can have more power to have CEO disclose more information on
the firm to reduce information asymmetry between shareholders, the principle, and
managers, the agent. Dalton & Kesner (1987); Davidson III, Worrell, & Cheng (1990) are in
favor of this idea in their studies.
H2: There is a positive relation between CEO separation and the extent of voluntary
disclosure.
Internal audit has been raised more concerns since the accounting scandals of WorldCom
and Enron. Accounting fraud is believed to be less likely to happen if there is a collaboration
between internal and external audit. Because of this reason, the Sarbanes Oxley Act of (2002)
has required that companies must establish and maintain an audit committee that have
independent directors as members and at least one member among them has financial
expertise.
Circular 71 of Vietnam Ministry of Finance has recommended that each company should
have an audit committee having minimum of three members and maximum of five members.
Members in audit committee are required not to be employees working financial departments
of company or employees of an auditing firm that have conducted financial audit for
company within three consecutive years. There are two main responsibilities for audit
committee: overseeing the financial performance of firm and assisting board of directors in
making financial statements.
The presence of audit committee improves internal control system and increases the
quality of reports thereby increases transparency of corporation (Collier, 1993; Forker, 1992;
S. S. Ho & Wong, 2001).
H3: There is a positive relation between audit committee existence and the extent of
voluntary disclosure.
“Corporate disclosure is affected by differences in type of auditing firms.”(Vu, 2012).
Similarly, Surendra
S. Singhvi & Desai (1971) explain that big audit firms put their reputation at first, so they
always want to co- operate with client companies who provide adequate information to be
sure about good audit quality. Therefore, most of big auditing firms encourage their clients to
disclose information as much as possible.

Agency theory says that the choice of auditors is a mechanism to reduce the conflicts of
interest between managers and shareholders (Jensen & Meckling, 1976). Bushman, Piotroski, &



Smith (2004) agree that external audit indicator is a confirmation measure of how reliable
disclosures are.
Moreover, big auditing firms have many clients, so they are not sensitive to the economic
loss if they miss some clients; they care for the audit quality instead. It is the reason why there
is a difference in disclosure level between choosing a big auditing firm and a smaller one.
Many studies have provided positive relation between audit quality and the level of
voluntary disclosure, such as: (Bonson & Escobar, 2006; OGWE, 2014; Patton & Zelenka, 1997;
Qu, 2011; Raffournier, 1995; Wallace
& Naser, 1995).
H4: There is a positive relation between audit quality and the extent of voluntary disclosure.


Haniffa & Cooke (2002) find out the positive link between foreign ownership and the volume
of voluntary disclosure in Malaysia. They reason that foreign owners want to have more
information to gain more control over the local management. Similarly, Surendra Singh Singhvi
(1968) suggests that companies whose majority of shares are owned by foreigner present a
higher disclosure quality than locally owned by Indian companies. There are some other
scholars who have the same research results with mentioned scholars, and some names can
be mentioned:(Barako, 2007; Xiao et al., 2004).
H5: There is a positive relation between foreign ownership and the extent of voluntary
disclosure.
3.Research Method
3.1.

Sample

Secondary data is collected through consolidated, audited firms’ annual reports which are
gathered from companies’ website and other public, reliable websites (vietstock, vndirect, or
cophieu68) for three financial years 2014-2016. Our sample includes firms in the list of VN100
which are 100 largest companies in terms of market capitalization. Following Vu (2012), we
exclude:
 Bank, finance, and insurance listed firms because of differences in reporting regime.
 Firms whose annual reports are unavailable for any of three years 2014-2016.
 Firms that miss any information about independent or control variables.
 Firms that have financial reporting period longer or shorter than 12 months.
 Firms that listed after January 01, 2015 as listing duration variable cannot be
measured.
3.2.

Measurements

In the Vietnamese literature, Vu (2012) has been considered to be one of the first scholars
who study about voluntary disclosure. She constructed her checklist that consists of 84 items
with 5 categories. Following her and some other scholars, Hieu & Lan (2015) constructed their
own checklist. This checklist was updated with changes of Circular 52/2012 TT-BTC about how
to disclose information on Stock Exchange. The final checklist has 42 items with 4 categories
of information: strategy, finance, future, social activities. In this study, the authors employ
the checklist from the study of Hieu & Lan (2015). Following Hieu and Lan (2015), VDI is
calculated as follow:
 Each firm is given “1” score for 1 item disclosed within the checklist, and “0”
otherwise.
 The total disclosure score that a firm achieves by disclosing items within the checklist
is computed by adding all item-scores.
 The total disclosure score is divided by total voluntary disclosure score.
The formula to calculate the VDI of one firm is adapted from many other studies: (Hieu &

Lan, 2015; Soliman, 2013; Vu, 2012), etc.

∑�
���� =

Where
:

83


�=1
��

��

VDI� = Voluntary disclosure index for ��ℎ firm;
�� = Voluntary disclosure item applicable to ��ℎ firm (n ≤ 42 items);
�� = “1” for disclosed item and “0” otherwise; so that: 0 = VDI� ≤ 1.
This study utilizes the unweighting approach when calculate the VDI, which means that it
assumes all
information is equally important to the users. (Chow & Wong-Boren, 1987; Vu, 2012) proved
that it made no significant differences between two approaches.

83


The authors also employ the same formula when investigating sub-categories
separately. Independent and control variables are measured as presented in table
1.

Table 1: Summary of Variables Measurements.
Variables
Dependent Variables:

Measurements

References

∑��=1
��
��� =

Voluntary Disclosure Index

�

VDI-CSI, VDI-FCMI, VDI-FLI,
VDI-SIR

��

∑�

(Hieu & Lan, 2015;
Soliman, 2013; Vu, 2012)
��

���(��� − ��������)� =

�=1

(Hieu & Lan, 2015;
Soliman, 2013; Vu, 2012)

��

Independent Variables:
The proportion of independent
members on board of directors.
Dummy variable: code “1” if there
CEO Separation (CEOSEPAR) is a separation between these two
positions, otherwise code “0”.
Audit Committee
Dummy variable: code “1” if the
Existence (ACE)
company has audit committee,
otherwise
code “0”.coding “1” if the
Dummy variable:
company
is
audited by one of “Big
Audit Quality (B4)
4” auditing firm, otherwise coding
“0”.
The total the percentage of equity
Foreign Ownership (FOROW)
shares owned by foreign investors.
Control Variables:
Listing Duration (LIST)

Natural logarithm of years listed.
Firm Size (FSIZE)
Natural logarithm of total assets.
��� �� ����
Profitability (PROFIT)
����� ������
Board Independence (BI)

3.3.

(Al-Janadi et al., 2013; S.
S. Ho & Wong, 2001)
(Al-Janadi et al., 2013)

(Hieu & Lan, 2015)
(Al-Janadi et al., 2013)

(Al-Janadi et al., 2013)

(Kim et al., 2005)
(Vu, 2012)
Vu (2012)

Model Specifications

Model 1 exams the relationship between voluntary disclosure index and corporate
governance variables. In most of previous studies, multiple regression was employed with a
large sample size (data was collected in one year). This study cannot follow them as data is not
large enough if data is examined in one year. Instead, we apply panel regression with a 3-year
period. This falls into the same situation with study from Barako, Hancock, & Izan (2006).

Model 1: Corporate governance characteristics and voluntary disclosure
����� = �0 + �1���� + �2 ���������� + �3 ����� + �4 �4�� + �5 ������� +
�6������ + �7������� +
�8�������� + ��.
The relations between 4 sub-categories of voluntary disclosure and corporate governance
variables are
investigated from model 2 to model 5.
Model

2

��� − ����� = �0 + �1���� + �2���������� + �3����� + �4�4��

+ �5 ������� + �6 ������ +
�7������� + �8�������� + ��.
3
VDI − FCMIit = β0 + β1BIit + β2CEOSEPARit + β3ACEit + β4B4it + β5FOROWit
+ βModel
6LISTit + β7FSIZEit + β8PROFITit + it.
Model
4
��� − ����� = �0 + �1���� + �2���������� + �3����� + �4�4��
+ �5 ������� + �6 ������ +
�7������� + �8�������� + ��.


Model
5
– ��� − ����� = �0 + �1���� + �2���������� + �3����� + �4�4��
+ �5 ������� + �6 ������ +

�7������� + �8�������� + ��.


Model 6 is conducted to see whether the effects of corporate governance variables on
voluntary disclosure with one-year lag.
Model 6 ����� = �0 + �1����−1 + �2����������−1 + �3�����−1 + �4�4��−1 +
�5�������−1 + �6 ������−1 +
�7 �������−1 + �8 ��������−1 + ��.
Where:
i, t: Indices for companies and time, respectively.
�����: Voluntary disclosure index – calculated as total disclosure items score over
possible maximum disclosure items in year t.
��� − ����� , VDI − FCMIit, ��� − ����� , ��� − ����� :Voluntary

disclosure

index
(sub-categories)
–
calculated as total disclosure items score over possible maximum disclosure items in sub-group
in year t.
���� , ����−1: Board independence – calculated by the percentage of independent
members on board of
directors in the year t and t-1, respectively.
���������� , ����������−1 : CEO separation – calculated by dummy variable,
respectively in the year t
and t-1.
����� , �����−1: Audit committee existence - calculated by dummy variable,
1. respectively in the year t and t�4��, �4��−1 : Audit quality (Big4 or not) - calculated by dummy variable, respectively in
the year t and t-1.

������� , �������−1 : Foreign ownership – calculated by the total the percentage of
owned
by equity
foreignshares
investors,
respectively in the year t and t-1.
������ , ������−1 : Listing duration – calculated by natural logarithm of numbers of
years listed in the year t and t-1.
������� , �������−1 : Firm size – calculated by natural logarithm of total assets in the
year t and t-1.
�������� , ��������−1 : Profitability – calculated as Net income divided by total
assets in the year t and t-1.
4.Results
4.1.

Voluntary Disclosure

The figure 1 presents the level of voluntary disclosure index and its sub-categories from
2014 to 2016. The average voluntary disclosure index of companies in VN100 is 0.47 with a
large range from 0.23 to 0.71. The index is higher than those calculated by (Hieu & Lan, 2015;
Vu, 2012). The mean VDI of 0.47 is lower than in the studies by M. Akhtaruddin & Haron (2010)
in Malaysia (0.58); Rouf & Harun (2011) in Bangladesh (0.48) and is higher than studies by S.
S. Ho & Wong (2001) in Hong Kong (0.29); Eng & Mak (2003) in Singapore (0.22). It is shown
from above analysis that voluntary disclosure score of Vietnamese listed companies is in the
middle level compared to other countries.


In analyzing the sub-categories disclosure, it is found that companies disclose
information on social reporting activities by far the most (the score goes form 0.62 in 2014 to
0.70 in 2016), which is consistent with the study by Hieu & Lan (2015) and inconsistent with

the study by Vu (2012). In her study, Vu (2012) found that companies disclosed least
information on this category in 2012. However, in 2015, Hieu & Lan (2015) found the
opposite result and it is the same with the finding in this thesis in three years from 2014 to
2016. This case indicates that firms more and more pay attentions to reporting social activities
with the purpose of showing their sustainable growths which are an increasingly focus of the
society. By doing that, companies can increase the attractions from much people in society,
which may benefit their productions.


VDI
VDI-CSI
VDI-FCMI
VDI-FLI
VDI-SRI

Voluntary Disclosure Level.

Table 2: Voluntary Disclosure and Its Sub-Categories.
Minimum
0.23
0.18
0.07
0.13
0.11

Maximum
0.71
0.64
0.86
0.86

0.89

Mean
0.47
0.38
0.40
0.48
0.67

Std. Deviation
0.09
0.10
0.14
0.16
0.16

Mean Voluntary Disclosure Level And Its Categories.
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
2014
2015
2016

VDI
0.46
0.47
0.48

VDI-CSI
0.37
0.38
0.38

VDI-FCMI
0.41
0.40
0.41

VDI-FLI
0.49
0.49
0.48

VDI-SRI
0.62
0.68
0.70

Figure 9 Mean Voluntary Disclosure and Its Categories.
(Note: VDI stands for voluntary disclosure index; CSI stands for corporate strategy information; FCMI
stands for financial capital and market information; FLI stands for forward looking information; SRI
stands for social reporting information).

Table 3 summarizes the descriptive analysis results and correlation matrix of independent
and control variables. The correlation values range from r=-0.15, p<0.1 to r=0.29, p<0.1,
which means that the correlation problem cannot cause biased results as all values are less
than 0.9.
Table 3: Mean, Standard Deviation, and Pearson’s Correlations
for the Variables.

BI
CEOSEPAR
ACE
BIG4
FOROW
LIST
FSIZE
PROFIT

Mea
n
(St.d
0.32
(0.18)
0.72
(0.45)
0.13
(0.34)
0.59
(0.49)
0.21
(0.17)
1.77

(0.62)
28.80
(1.09)
0.08
(0.06)

Min
(Max
) 0.00
(0.80)
0.00
(1.00)
0.00
(1.00)
0.00
(1.00)
0.00
(0.59)
-2.48
(2.77)
25.48
(31.92
) -0.36
(0.28)

BI

CEOSEPAR

ACE

BIG4

FOROW

LIST

FSIZE

PROFIT

1.00
0.08

1.00

-0.05

-

1.00

0.06

0.07

-0.01

1.00

-0.10*

0.09

0.09

0.26***

1.00

-0.05

-

0.10*

0.17***

0.15***

-0.15***

0.04

0.05

0.28***

0.11**

-0.10*

0.11*

0.0009

-0.01

1.00

0.001
6
0.29*** -0.02

1.00
0.21**

1.00


*,**,*** Correlation is significant at 10%, 5%, and 1%.


In this study, the authors employ Variance Inflation Factors (VIF) to test the
multicollinearity between independent variables. All the VIF values are less than 10 (table 4),
which can come to the conclusion of no presence of multicollinearity.
Table 4: Multicollinearity Test.
Variables
BI
CEOSEPAR

ACE
BIG4
FOROW
LIST
FSIZE
PROFIT
Mean VIF

VIF
1.08
1.04
1.03
1.20
1.24
1.07
1.21
1.20
1.13

4.2.

Regression Results

4.2.1.

Corporate governance variables and voluntary disclosure index (model 1)

Table 5 presents detailed results on model 1. Fixed-effect model (FEM) is employed after
F-test and Hausman test. Two more additional tests are conducted, which reveals that data
suffers autocorrelation, but not heteroscedasticity. In conclusion, the final regression is run

again by FEM with cluster standard errors to remove autocorrelation.
Table 5: Corporate Governance Characteristics and Voluntary Disclosure
(Model 1)
FEM with Cluster robust
Independent variables
BI
CEOSEPAR
ACE
BIG4
FOROW
Control variables
LIST
FSIZE
PROFIT
Constant
Observations
R²
F
Prob > F

Coefficient

Robust SE

Sig.

0.03
0.03
-0.01
-0.02

0.05

0.02
0.02
0.01
0.02
0.03

0.22
0.12
0.30
0.36
0.08*

0.03
0.01
0.15
-0.03
282
0.15

0.01
0.01
0.08
0.26

0.04**
0.13
0.06*
0.91

2.41
0.02

*,**,*** Correlation is significant at 10%, 5%, and 1%. Source: the author.

It can be seen from table 5 that the model 1 is highly significant (F=2.41 and p=0.02) and
the model explains 15% of the variations in the voluntary disclosure (R-sq within=0.15).
H5 is supported by the result from panel 1. In the panel, a positive and significant relation is
found between foreign ownership and the level of voluntary disclosure at 10% level with
(β=0.05, p=0.08). This finding is similar to (Bonson & Escobar, 2006; OGWE, 2014; Patton &
Zelenka, 1997; Qu, 2011; Raffournier, 1995; Wallace
& Naser, 1995).


The finding can be explained based on the high average proportion of foreign owners. In the
sample, more firms have 21% of their shares owned by foreigners, which means that these
foreign shareholders are powerful


enough to get more additional information disclosed. It is consistent with what is discussed in
hypotheses developments as foreign owners want to have control on local management team.
Other than that, the relationship between voluntary disclosure and corporate governance is
also confirmed by control variables listing duration and profitability. Findings for control
variables show that listing duration is positively and significantly related to the volume of
voluntary disclosure at 5% level of significance (β=0.03, p=0.04). This result is consistent with
those from (Alsaeed, 2006) and (Md Akhtaruddin, 2005). It is reasoned that old firms have
more experiences and resources to outweigh the younger ones in the number of information
disclosed voluntarily. It is also a reason that elderly listed firms want to show more about them
to gain more reputation. This is also the reason for a positive relation between listing duration

and the volume of disclosure in social reporting information sub-category. Finally, firms with
many years listed have to face with public scrutiny, so they want to give out more information
to reduce stress from society. Empirical data provides a result that profitability has positive
effects on the level of information that firms are willing to report in their annual reports at
10% level of significance (β=0.15, p=0.06). The finding supports for other previous finding
from many authors (Haniffa & Cooke, 2002; Vu, 2012). This outcome can be explained that
firms with high profit are usually scrutinized by public, so they have to give out more evidence
to prove for their transparency. It can be another reason that profitable firms want to
distinguish themselves from others in the industry to appeal to more outside capital.
Besides, board independence (β=0.03, p=0.22) is found to have no significant relation
with voluntary disclosure, which means that H1 is not supported. However, this finding is
consistent with those from (Haniffa
& Cooke, 2002; S. S. Ho & Wong, 2001), and contradict to study from (Chen & Jaggi, 2001).
There are some potential reasons for this insignificant result. First, as discussed in the previous
part, the rate of independent director to total board member (which is 0.32) is lower than the
required rate, so it is difficult for the outsiders’ votes to outweigh those of insiders. Therefore,
they are not powerful enough to make the decision to disclose more information become
reality. Second, the sample has a relatively high average voluntary disclosure index (0.47) and
low proportion of independent board members, which means that the insider board
members voluntary can disclose more information as well as the outsiders do. In this case,
the insiders can act as the roles of outsiders and the agency problem has expectedly less
effects.
It is revealed from the result that CEO separation (β=0.03, p=0.12) is in no significant
relation with volume of voluntary disclosure, which means that H2 is not supported. This
finding is opposite to those from (Collier, 1993; Dalton & Kesner, 1987; Gul & Leung, 2004),
and the same with study from (Barako, Hancock, & Izan, 2006). The insignificant relation can
be explained by one following reason. There may be no much difference between a voluntary
disclosure index generated by a CEO-separating firm and a non-separating firm. Therefore,
this relation is so weak in statistics.
Firm size (β=0.01, p=0.13) is found to have no relation to the amount of information

disclosed voluntarily, which is in contradiction with findings from many other researchers
(Camfferman & Cooke, 2002; Cooke, 1989; Raffournier, 1995). There is one possible reason for
this case. There are more big firms as the mean of the firm size variable is relatively high
which is 28.8. It can be inferred that there are not great differences between sizes of firms in
the sample, so it cannot generate an expected relation.
Empirical data generates an insignificant relation between having audit committee (β=
-0.01, p=0.30) and the extent of voluntary disclosure. It means that H3 is not supported. The
finding is different from studies from (Collier, 1993; S. S. Ho & Wong, 2001). The insignificant
association can be well explained for one reason that based on an evidence. The main model
tests how independent variables affect dependent variable in the same year, so result
indicates that the change in audit committee will not have immediate effects on level of


disclosure. However, in the model 6 with lagged time effects, it can be seen that there is a
significantly positive association between those two variables. Meaningfully, the effect of
having an audit committee is delayed to one-year period. This result matches with what is
happening in Vietnamese context. There is always a gap from the time a person is elected to
the date when his signature is powerful. The replaced person has a time


to complete unfinished tasks and transfer them to the new one. Therefore, any changes in
personnel in one year are going to affect the quality annual report made in the year after.
Audit quality (β= -0.02, p=0.36) both have negative but insignificant relation with the extent
of voluntary disclosure. The panel 1 generates result that level of disclosure is in no relation
with the audit quality, which is not in favor of H4. This finding is similar to those from
(Mohammad Hossain, Perera, & Rahman, 1995a; Malone, Fries, & Jones, 1993; Wallace, Naser,
& Mora, 1994), and contradict to study from (Qu, 2011). Actually, the effects of changing
auditing firm on the quality of annual report happens after one year. Usually, firms will
change from less famous auditing firms to four big ones. At the year of changing, the effects
may be little because a firm need time to adapt to the changes and from one year later on the

huge changes are recognized.
4.2.2.
5)

Corporate governance and voluntary disclosure sub-categories (model 2 to model

In the model 2 and model 5, the results from tests show that FEM should be applied. Data in
two models both suffered autocorrelation, but not heteroscedasticity. Therefore, FEM model will
be run with cluster robust to remove autocorrelation.
In the model 3 and 4, the results imply that REM should be the most appropriate among 3
models. Data in two models both suffer autocorrelation. However, only model 5 also has
heteroscedasticity. Therefore, the final regression is REM model with cluster robust to remove
autocorrelation and heteroscedasticity at the same time.
Table 6: Corporate Governance Characteristics and Voluntary Disclosure Subcategories (Model 2,3,4,5)
VDI-CSI (model 2) VDI-FCMI (model 3)
276
276
0.0
0.12
2.0
7
0.0
5

VDI-FLI (model 4)
276
0.09

11.5
6

0.17

4.9
1
0.0
0

0.002
5

Sig.

0.06 0.37 0.10
0.02 0.06* 0.04
0.02 0.02* 0.03 0.65 0.08
0.07 0.04* -0.03
0.03 0.73 0.15
0.01 0.12 0.03
0.14 0.37 0.16
0.35 0.84 -0.45

Robust
SE

-0.05
-0.05
0.05
0.01
0.14
0.01

0.02
0.13
-0.07

Coeff

0.27
0.78
0.73
0.84
0.04
0.36
0.32
0.12
0.74

Sig.

Coeff

0.05
0.02
0.03
0.03
0.05
0.02
0.01
0.11
0.26

Robust
SE

Sig.

0.04 0.81 -0.06
0.02 0.14 -0.01
0.03 0.11 0.01
0.02 0.02* 0.01
0.09 0.82 0.11
0.04 0.67 0.02
0.02 0.95 0.01
0.13 0.09* 0.18
0.59 0.48 0.09

VDI-SRI (model 5)
273
0.2

23.76

Robust
SE

Coeff

Sig.

BI
0.01

CEOSEPAR 0.03
ACE
-0.05
BIG4
-0.05
FOROW
0.02
LIST
0.02
FSIZE
PROFIT
-0.22
Constant
0.41

Robust
SE

Coeff

Observatio
R-sq
F/
Wald
Chi2
Prob>F/
Prob>Chi
2

0.05 0.05**

0.03
0.13
0.03
0.91
0.06
0.20
0.05
0.58
0.05 0.002*
0.02
0.19
0.14
0.28
0.50
0.37

*,**,*** Correlation is significant at 10%, 5%, and 1%. Source: the author. (Note: VDI stands for voluntary
disclosure index; CSI stands for corporate strategy information; FCMI stands for financial capital and
market information; FLI stands for forward looking information; SRI stands for social reporting
information).

In the model 2, both audit quality and profitability have significantly negative relations with
voluntary disclosure in corporate strategy information sub-category, which is proved by β=0.05 with p=0.02 for BIG4 variable and β=-0.22 with p=0.09 for PROFIT variable.
The model 3 with Chi2=11.56 and p=0.17 is not statistically significant. For that reason, the
author decides to put the model out of concerns.


In the model 4, audit committee presence (β=0.05, p=0.02) and foreign ownership (β=0.14,
p=0.04) are both significantly related to voluntary disclosure in future looking information
sub-group. Differently, CEO separation (β=-0.05, p=0.06) is in a significantly negative relation

with VDI-FLI sub-case.
Finally, In the model 5, both board independence (β=0.10, p=0.05) and listing duration
(β=0.15, p=0.002) have significantly positive relations with social reporting information subcategory of voluntary disclosure.
4.2.3 Corporate governance and voluntary disclosure with time lag
Besides the main model, the authors wonder whether the effects of corporate governance
variables on the extent of voluntary disclosure may be delayed a year. Therefore, the author
decides to conduct an additional test to check for the mentioned possibility.
It is indicated from tests that REM is more appropriate for this model 6. Data used in this
model suffered autocorrelation, but no heteroscedasticity. Therefore, the final regression is run
by REM with cluster standard errors to remove autocorrelation.
Table 7: Corporate Governance Variables and Voluntary Disclosure with Time Lag
(Model 6)
REM with Cluster robust
Independent variables
BI1
CEOSEPAR1
ACE1
BIG41
FOROW1
Control variables
LIST1
FSIZE1
PROFIT1
Constant
Observations
R²
Wald Chi2 (8)
Prob > Chi2

Coefficient

Robust SE

Sig.

-0.01
0.02
0.06
0.02
0.06

0.02
0.01
0.02
0.01
0.04

0.70
0.12
0.004***
0.08*
0.10*

0.02
0.0044
0.01
0.27
240
0.09

0.01
0.005
0.02
0.15

0.21
0.38
0.49
0.07

23.17
0.0032

*,**,*** Correlation is significant at 10%, 5%, and 1%. Source: the author.

Final regression of model 6 generates the results summarized in table 7. There are 3
variables having significantly positive relationship with the voluntary disclosure: audit
committee presence (β=0.06, p=0.004), audit quality (β=0.02, p=0.08), and foreign
ownership (β=0.06, p=0.10) with the significant level of 1%, 10%, and 10% respectively.
Besides, board independence (β=-0.01, p=0.70) shows a negative but insignificant
association with voluntary disclosure.
Moreover, CEO separation (β=0.02, p=0.12), listing duration (β=0.02, p=0.21), firm size
(β=0.00, p=0.38), and profitability (β=0.01, p=0.49) all affected positively the volume of
voluntary disclosure. However, these links are not strong enough to be considered as
statistically significant links.
5.Conclusion
The authors study the relation between corporate governance mechanisms and the extent
of voluntary disclosure with the sample of 101 companies through a 3-year period starting
from 2014 to 2016. The researchers adapt a checklist from (Hieu & Lan, 2015) to figure out

the value of dependent variables. As for independent and control variables, this study uses
many proxies mentioned in many previous papers to