Professional Documents
Culture Documents
kang2014
kang2014
PII: S0927-538X(14)00053-5
DOI: doi: 10.1016/j.pacfin.2014.04.006
Reference: PACFIN 698
Please cite this article as: Kang, Minjung, Lee, Ho-Young, Lee, Myung-Gun,
Park, Jong Chool, The Association between Related-Party Transactions and Control-
Ownership Wedge: Evidence from Korea, Pacific-Basin Finance Journal (2014), doi:
10.1016/j.pacfin.2014.04.006
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 proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
ACCEPTED MANUSCRIPT
T
Title: The Association between Related-Party Transactions and Control-Ownership
P
Wedge: Evidence from Korea
RI
Authors:
(1) Minjung Kang, Ph.D., School of Business, Yonsei University
SC
(2) Ho-Young Lee, Professor, School of Business, Yonsei University
(3) Myung-Gun Lee, Assistant Professor, School of Business, Yeungnam University
(4) Jong Chool Park, Assistant Professor, College of Business & Public Administration,
NU
Old Dominion University
Corresponding author:
ED
Ho-Young Lee
Prof. of Accounting
Yonsei University
PT
hylee@yonsei.ac.kr
82-2-2123-5484
82-10-3398-1847
CE
AC
ABSTRACT
In this study, we examine whether related party transactions (RPT) are used as a
mechanism for tunneling among firms belonging to large business groups in Korea
(chaebols). Using 982 firm-year data of publicly traded firms in Korea, we find that the
control-ownership wedge is positively associated with the magnitude of RPTs. RPTs
increase as voting rights increase, while RPTs decrease as cash flow rights increase. The
control-ownership wedge is more closely related to RPTs among the top 5 chaebol firms
where the agency conflicts between the controlling shareholders and minority shareholders
are more severe than in non-top 5 chaebol firms. While the significant positive association
between the control-ownership wedge and RPTs holds for both operating and non-
operating RPTs, we find that non-top 5 chaebols use only non-operating RPTs whereas the
top 5 firms use both operating and non-operating RPTs. Finally, we find that RPTs of
Korean chaebol firms, on average, reduce firm value, but this value destruction is observed
only when the control-ownership wedge is high and is more pronounced with the top 5
1
ACCEPTED MANUSCRIPT
chaebol firms. Overall, our results together suggest that RPTs occur when the agency
problem is severe and they are used as a means of tunneling, thus destroying firm value.
T
JEL classification: G30, M40
P
RI
1. Introduction
SC
entities such as shareholders, members of the board of directors, and affiliated companies.
NU
Investors and analysts often raise concerns about RPTs, asking whether corporate insiders
are fully focused on the interests of shareholders when conducting such transactions (e.g.,
MA
Wall Street Journal, May 7, 2003, B61). RPTs have also become a central issue in more
advanced capital markets, creating concerns among regulators and other market participants
ED
regarding the appropriate monitoring and auditing of these transactions (Johnson et al.,
PT
2000). Extant academic studies provide inconsistent evidence on the effect of RPTs on firm
value. Some argue that RPTs can be used as efficient contracting mechanisms under
CE
costs and thereby achieving economies of scale (Williamson, 1975; Stein, 1997; Khanna
and Palepu, 1997) (efficient transaction hypothesis), while others argue that RPTs destroy
firm value because they arise from conflicts of interests between controlling shareholders
and minority shareholders and are carried out in the interest of controlling shareholders to
expropriate wealth from minority shareholders (Shin and Park, 1999; Chang and Hong,
1
“These are the kinds of relationships that companies should avoid, in the view of some corporate-
governance experts and investors. Such related-party transactions raise questions about whether corporate
insiders are fully focused on the interests of shareholders, experts say. The deals, no matter how small, can
create the impression that an insider is using company assets for personal benefit, and that the company is
getting the short end of the stick.” (Wall Street Journal, May 7, 2003, “Even Good Insider Deals Raise
Doubts”, B6).
2
ACCEPTED MANUSCRIPT
2000; Johnson et al., 2000) (conflict of interest hypothesis). These opposite and conflicting
predictions and findings in prior literature may be attributed to a failure to examine the role
P T
Concentrated ownership often leads to a divergence of control rights (or voting
RI
rights) and cash flow rights, commonly referred to as the “control-ownership wedge,” or
SC
more simply the “wedge.” A high control-ownership wedge may provide controlling
shareholders, who typically have voting rights in excess of cash flow rights, incentives to
NU
extract private benefits at the expense of minority shareholders. Consistent with this notion,
MA
recent studies document that firm value decreases as corporate insiders control more voting
rights relative to their cash flow rights (Claessens et al., 2002; Lemmon and Lins, 2003;
ED
Lins, 2003; Harvey et al., 2004; Gompers et al., 2010). In this study, by employing the
positive association between the control-ownership wedge and RPTs indicating that related
CE
party transactions are used as a mechanism for tunneling among firms belonging to large
business groups in Korea (chaebols). We further predict the positive association between
AC
the wedge and RPTs is more/less pronounced depending on type of RPTs and size of the
chaebol. Next, we predict a negative association between RPTs and firm value consistent
with the idea that RPTs are potential channels through which this specific ownership
structure leads to firm value destruction. We further predict that the negative effect of RPTs
on firm value is more pronounced when the control-ownership wedge is large and more so
The Korean market has several characteristics that make it particularly suited to our
investigation. Many Korean listed firms belong to business groups (conglomerates) known
3
ACCEPTED MANUSCRIPT
as chaebol which links their affiliated firms via cross and/or circular-shareholdings (Kim
and Yi, 2006). As a result, Korean firms are likely to have a large disparity between voting
rights and cash flow rights, which makes the effect of the control-ownership wedge more
P T
apparent. The divergence between ownership and control rights in the Korean market
RI
implies that the agency conflicts between controlling shareholders and minority
SC
shareholders are severe. This ownership structure suggests that the expropriation of
minority shareholder wealth is a distinct possibility. Korea also has relatively weak
NU
protections for outside minority shareholders (La Porta et al., 1999; Bae et al., 2002). Thus,
MA
the effect of the disparity is likely to be more apparent in Korea than other countries. Lastly,
Korean chaebol firms are likely to exercise more transactions with related parties than those
ED
common. Altogether, the Korean setting provides a more powerful means for examining the
PT
association between the wedge and RPTs than any other country.
CE
establish the relationship between RPTs and the control-ownership wedge and then we test
AC
whether the negative effect of RPTs on firm value is conditional on the degree of the
control-ownership wedge. We find that the magnitude of RPTs is positively associated with
the control-ownership wedge, suggesting that RPTs are carried out more actively when
there is more serious agency problem. Next, we examine whether the positive association
between RPTs and the control-ownership wedge is stronger in firms that belong to large
business groups. Specifically, we focus on the five largest business groups in Korea (the top
5 chaebols, hereafter) and see whether the positive association between RPTs and the
control-ownership wedge is more pronounced in these groups. The top 5 chaebols represent
4
ACCEPTED MANUSCRIPT
Korea’s most prominent chaebols during the past three decades. The economic
concentration in the top 5 chaebols is very large and this phenomenon has further
intensified in recent years.2 Investors and regulators have long criticized the top 5 chaebols
P T
claiming that these firms use RPTs as a means for tunneling in the interest of controlling
RI
shareholders. Consistent with this criticism, we find that among large chaebol firms, the top
SC
5 chaebols show a stronger positive association between RPTs and the control-ownership
wedge.
NU
We then analyze how the control-ownership wedge influences the different types of
MA
RPTs by disaggregating firm-level RPTs into operating (e.g., products/services sales and
property, plant and equipment and investment assets). Generally, transactions associated
with non-operating activities are considered a more preferable means for wealth transfers as
PT
they involve more subjective judgments and discretion than regular operating activities
CE
(Bertrand et al., 2002). However, a firm’s choice between operating and non-operating
RPTs may also depend on the size of the business group. Due to a lack of proper transaction
AC
partners, especially for operating transactions, small chaebols may have to depend more on
non-operating RPTs for tunneling. Our results show that overall, both operating and non-
operating RPTs are positively and significantly associated with the control-ownership
wedge. In particular, we find that the top 5 chaebol firms show significantly positive
association between RPTs and the wedge for both operating and non-operating transactions
2
The proportion of sales by the top 5 chaebols has increased from 49.5 percent in 2001 to 55.7 percent of
gross domestic product (GDP) in 2010 and the total profits (operating profits) of the top 5 account for over 66
(55) percent of those of the largest 500 firms (Hankyre News, 02/13/2012).
5
ACCEPTED MANUSCRIPT
while non-top 5 chaebol firms show a significant association only for non-operating
transactions.
After establishing a direct empirical relation between RPTs and the control-
P T
ownership wedge, we next examine the valuation implication of this relation in the
RI
subsequent analyses. Our results show that the value destroying effect of RPTs is observed
SC
only when the control-ownership wedge is large and stronger in the top 5 chaebol firms as
opposed to non-top 5 chaebol firms, suggesting that the effect of RPTs on firm value is
NU
conditional on the severity of the agency problem arising from the conflict of interest
MA
between controlling and minority shareholders. In the additional tests, we show that both
operating and non-operating RPTs of the top 5 chaebol firms destroy firm value while only
ED
non-operating RPTs destroy firm value in the case of non-top 5 chaebol firms. Furthermore,
this negative valuation effect of operating and non-operating RPTs is observed only when
PT
the control-ownership wedge is large. Overall, our findings suggest that controlling
CE
shareholders for firms with the large control-ownership disparity are more likely to engage
in tunneling through RPTs and RPTs do not destroy firm value for firms with the small
AC
control-ownership disparity. Our results also imply that the severity of the agency problem
plays a key role in determining whether RPTs are more likely to be associated with
tunneling or propping.
This study makes several contributions to the existing literature. First, this study
provides evidence that RPTs are an important channel to exploit minority shareholders’
wealth that can be used when the ownership disparity is large. Second, our study
contributes to the literature on the relationship between RPTs and firm value. Previous
studies examining the effect of RPTs on firm value provide inconclusive evidence. Some
6
ACCEPTED MANUSCRIPT
find that RPTs are more consistent with tunneling while others find they are more
consistent with propping. Such inconclusive evidence is not surprising since RPTs by
P T
interest hypothesis). Our study disentangles the inconsistent evidence by showing that the
RI
effect of RPTs on firm value is conditional on the severity of the agency problem. Due to
SC
the perceived negative consequences of RPTs on firm value, regulators around the world
have imposed strict restrictions on RPTs to make sure transactions are transparent and
NU
objective in order to protect minority shareholders.3 Our results suggest that the regulations
MA
on RPTs need to be made by considering the concerns raised by the investing public.
Finally, the strict disclosure requirements for RPTs in Korea enable access to more detailed
ED
requirements4 and provide large sample evidence on tunneling through RPTs based on
PT
more reliable and comprehensive related party transaction data than the data used in most
CE
prior studies.
Section 2 reviews prior studies and develops our hypotheses. Section 3 presents our
AC
research methods and sample selection. Section 4 shows the results of empirical analyses,
3
These restrictions are imposed based on ‘Monopoly Regulations and Fair Trade Law (Korean Fair Trade
Commission: KFTC, 2013), ‘Guidelines for Reviewing Unfair Assistance’ (KFTC, 2011), and ‘Regulation on
Resolution of Board of Directors and Disclosures on Large-scale Intra-group Transactions’ (KFTC, 2012).
More details are described in APPENDIX.
4
KFTC requires business groups with total assets over 2 trillion Korean won (years 2003-2008) and 5 trillion
won (years 2009-present) to disclose a detailed ownership structure. The disclosure must include the list of
affiliated firms within the group, voting rights, cash flow rights, and the control-ownership wedge (Fair Trade
Act Decree 17 Article 9, 2008). Ownership data on the wedge is required to be disclosed on the KFTC
website and is verified by the government. Thus, subjective judgment in computing the wedge is less of a
concern in this study.
7
ACCEPTED MANUSCRIPT
P T
2.1. Control-ownership wedge
RI
In an environment of concentrated ownership, the conflict of interest between
SC
minority shareholders and controlling shareholders creates a serious agency problem (e.g.,
La Porta et al., 1999; Claessens et al., 2000; Lemmon and Lins, 2003). Corporate
NU
ownership is concentrated in the hands of controlling shareholders who typically have
MA
voting rights in excess of their cash flow rights.5 The divergence of control rights (or
voting rights; VR, hereafter) and cash flow rights (CR, hereafter), “wedge,” motivates
ED
shareholders (Johnson et al., 2000; Claessens et al., 2002). The control-ownership wedge
PT
also raises concerns about the degree to which the controlling shareholders of a business
CE
group engage in tunneling, that is, transferring resources out of firms to increase their
controlling shareholders for private gain from minority shareholders increases as the benefit
from expropriation becomes larger relative to the associated cost (Shleifer and Vishny, 1997;
Lemmon and Lins, 2003, Joh, 2003; Baek et al., 2004). Consistent with this concern,
Bertrand et al. (2002) find that controlling shareholders in Indian business groups transfer
5
Voting rights are rights that shareholders can exercise to participate in decision making via votes and are
measured as the proportion of voting shares that are controlled by direct and indirect share investment. On the
other hand, cash flow rights are shares directly controlled by controlling shareholders and are interpreted as
claims on profits or dividend. When voting rights are greater than cash flow rights, controlling shareholders
only bare losses up to the proportion of their cash flow rights.
8
ACCEPTED MANUSCRIPT
resources or tunnel profits from affiliated firms where they have low cash flow rights to
P T
wedge and firm value suggesting that the self-serving behavior of controlling shareholders
RI
undermines corporate performance and thus destroy firm value. (e.g., Clasessens et al.,
SC
2000, 2002; Joh, 2003; Lins, 2003; Lemmon and Lins, 2003). Based on data from eight
East Asian countries, Claessens et al. (2002) report that firm value decreases when the
NU
control-ownership wedge increases, while firm value increases as cash flow rights of the
MA
largest shareholder increase. Similarly, Lemmon and Lins (2003) find that the market value
for firms with a large control-ownership wedge was significantly lower than for other firms
ED
during the Asian financial crisis. Lins (2003) also reports a decline in firm value when a
management group’s control rights are greater than its cash flow rights. He finds that the
PT
negative association between the wedge and firm value is significantly more pronounced in
CE
countries with weaker legal shareholder protection measures. The results of these studies
suggest that value decrease stems from controlling shareholders’ transferring resources or
AC
tunneling profits of one firm where controlling shareholders have small cash flow rights to
members of the board of directors, and affiliated companies.6 Extant academic literature
6
In this paper, we define a related party following the International Accounting Standard No. 24 (IAS 24).
According to IAS 24, a related party is a person or entity that is related to the entity that is preparing its
financial statements. Specifically, a person or a close member of that person’s family is related to a reporting
9
ACCEPTED MANUSCRIPT
provides two competing views on RPTs: the efficient transaction hypothesis and the
conflict of interest hypothesis (Chang and Hong, 2000; Gordon et al., 2004; Cheung et al.,
2006; Kohlbeck and Mayhew, 2010). The efficient transaction hypothesis considers RPTs
P T
as sound economic exchanges achieving shareholder value maximization. By reducing
RI
transaction costs, RPTs can be used as efficient contracting mechanisms under incomplete
SC
information (Williamson, 1975; Stein, 1997; Khanna and Palepu, 1997; Shin and Park,
1999). Chang and Hong (2000) document that affiliated firms within a business group share
NU
tangible and intangible resources and by doing so, they enjoy the economies of scale and
MA
scope. Friedman et al. (2003) argue that some RPTs such as cash receipts by the listed
company are used as a means of propping, not tunneling, and these RPTs are likely to
ED
benefit the listed firm’s minority shareholders. In contrast to the efficient transaction
hypothesis, the conflict of interest hypothesis considers RPTs potentially harmful and value
PT
destroying. In the conflict of interest hypothesis, RPTs arise from conflicts of interests
CE
between controlling shareholders and minority shareholders and they are carried out in the
While RPTs may improve economic efficiency by reducing transaction costs (e.g.,
Ryngaert and Thomas, 2007), a few empirical evidence has shown that RPTs arise from the
emerging markets where legal protection of minority shareholders is weak. For example, La
Porta et al. (1997, 1998, 1999, 2000), Johnson et al. (2000), Glaeser et al. (2001), Jian and
entity if that person (i) has control or joint control of the reporting entity; (ii) has significant influence over the
reporting entity; or (iii) is a member of the key management personnel of the reporting entity or of a parent of
the reporting entity.
10
ACCEPTED MANUSCRIPT
Wong (2004), and Claessens et al. (2006) have found that controlling shareholders extract
private benefits from minority shareholders through “tunneling.” Cheung et al. (2009) also
provide evidence that prices are arbitrarily set in asset sales and purchases with related
P T
parties. Using 174 firms that disclosed RPTs in Hong Kong, they find that firms purchase
RI
products at higher prices while selling them at lower prices in RPTs. Similarly, Cheung et al.
SC
(2006) document that firms listed on the Hong Kong stock market experience significant
negative abnormal stock returns upon the disclosure of RPTs; this indicates that such
NU
transactions have a negative effect on firm value.
MA
2.3. Hypotheses
both public and private companies in a pyramidal and circular ownership structure, that are
small number of controlling shareholders enjoy full control over all affiliated firms despite
CE
holding a relatively small portion of cash flow rights.7 The pyramidal ownership structure
more severe agency conflicts between minority shareholders and controlling shareholders
in Korea than in other countries. The lack of strong legal protection of minority investors in
7
For example, according to Monthly Shin-DongA (2007) Mr. Kunhee Lee, the chairperson of Samsung
Electronics Co., had a complete control over affiliated firms within Samsung Business Group via circular
ownership. Mr. Lee had a significant ownership of Everland Co. which had a significant ownership of
Samsung Life Insurance Co. Samsung Life Insurance Co. had a significant ownership of Samsung Electronics
Co. while Samsung Electronics Co. had a significant ownership of Samsung Card Co. Samsung Card Co. had
a significant ownership of Everland Co. This case well summarizes how circular ownership can be used by
controlling shareholders to control the rest of the affiliated firms within the business group (An external
director’s perspective on a difficulty in Samsung, Monthly Shin-DongA, 2007 (Dec. 1), p. 118).
11
ACCEPTED MANUSCRIPT
scrutinize these transactions due to the nature of complex chaebol structures. As a result,
firms have opportunities and means to divert firm resources through RPTs at the expense of
P T
alternative form is as follows:
RI
Hypothesis 1: There is a positive association between the control-ownership
wedge and RPTs.
SC
The top 5 chaebols have come to represent Korea’s most prominent chaebols during
NU
the past three decades. Economic concentration of the top 5 chaebols is very large and this
MA
phenomenon has been more intensified recently. The proportion of sales by the top 5
chaebols has increased from 49.5 percent in 2001 to 55.7 percent of gross domestic product
ED
(GDP) in 2010 and the total profits (operating profits) of the top 5 account for over 66 (55)
percent of those of the largest 500 firms (Hankyore News, 02/13/2012). Campbell and Keys
PT
(2002) document that firms affiliated with the top 5 chaebols exhibit significantly lower
CE
performance relative to other firms. They also report that top executive turnover is
unrelated to performance for the top 5 chaebol firms, indicating a lack of properly
AC
functioning internal corporate governance among the top 5 chaebols. Jung and Kwon (2002)
RPTs of the top 5 chaebols are generally much larger in scale than those of non-top
5 chaebols and they account for over 50 percent of their combined revenues of the top 5
chaebols (Kyunghyang News, 06/23/2013).8 The top 5 chaebols have long been criticized
8
In reality, a significant proportion of RPTs among affiliated firms within a chaebol have been used as a
means for unfair or expediential wealth transfers (or inheritances or gifts) by supporting insolvent affiliated
12
ACCEPTED MANUSCRIPT
for using RPTs as a means for siphoning off corporate wealth to line the pockets of the
siblings of owner families instead of improving the efficiency of transactions. Although the
top 5 chaebols argue that large amount of RPTs is inevitable to maximize firm value due to
P T
horizontal and vertical diversification investors and regulators believe that the top 5
RI
chaebols are more likely to use RPTs as a means of tunneling (Jung and Kwon, 2002). We
SC
thus predict that the positive association between the control-ownership wedge and RPTs is
NU
Hypothesis 1a: The positive association between the control-ownership wedge
and RPTs is stronger for the top 5 chaebols than for non-top 5
MA
chaebols.
purchases of property, plant and equipment and investment assets). Operating transactions
occur throughout the accounting period in a frequent and recurring manner, while non-
CE
RPTs regulators can monitor the operating RPTs without much difficulty due to the
industry peers. For example, they may cast doubt on the fairness of operating RPTs if
prices in operating RPTs are believed to be unfair compared to industry average prices.
firms and generating large transactions with certain affiliated firms with significant ownership held by
controlling shareholders. RPTs of chaebols have certain characteristics. First, the larger the chaebols, the
greater the RPTs. The amount of RPT sales by all chaebols was 132 trillion Won as of December 12, 2013
accounting for 71 percent of all RPT sales. Second, RPTs of chaebols are concentrated on non-listed firms
with large controlling shareholders’ ownership because they are less subject to market monitoring. These
characteristics are closely related to criticisms that RPTs are wrongfully used as means of unfair supports to
firms where controlling shareholders possess significant share ownership (Hankyre News, 04/17/2013).
13
ACCEPTED MANUSCRIPT
In contrast to operating RPTs, it may be more difficult for regulators and investors
sales/purchases of investment and property, plant, and equipment assets. For example, real
P T
estate assets held for a long period of time may not have relevant market price data.
RI
Therefore, transactions associated with non-operating activities are considered more
SC
preferable means for tunneling as they generally involve more subjective judgments and
discretion than regular operating activities (Bertrand et al., 2002). However, the large scale
NU
of non-operating transactions may also draw a close attention from regulators and
MA
sometimes prompts outside intervention by regulators and investors, leading to a
disciplinary action against controlling shareholders. If this is the case, firms would prefer
ED
transactions on the association between the control-ownership wedge and related party
CE
A firm’s choice between operating and non-operating RPTs may also depend on the
size of the business group. Large chaebols have various firms in different industries while
relatively small chaebols have a small number of firms in a few industries. The difference
in group size may allow larger chaebols to use operating RPTs for tunneling more
effectively than smaller chaebols. Due to the limited number of firms in the same business
group for operating transactions, small chaebols may have to depend more on non-
14
ACCEPTED MANUSCRIPT
operating RPTs for tunneling. Thus, we test, within the top 5 chaebols, whether the positive
association between the wedge and RPTs is more pronounced in operating RPTs. We also
test, within non-top 5 chaebols, whether the positive association between the wedge and
P T
RPTs is more pronounced in non-operating RPTs.
RI
Hypothesis 1c: The positive association between the control-ownership wedge
and RPTs pertaining to the top 5 chaebols is more pronounced in
SC
operating RPTs.
NU
and RPTs pertaining to non-top 5 chaebols is more pronounced in
non-operating RPTs. MA
3. Research design and sample selection
Chaebols are controlled by families that hold equity stakes in group firms either
PT
directly or indirectly through other firms in the group. We thus define cash flow rights by
incorporating both direct holdings and indirect holdings through affiliated firms (La Porta
CE
et al., 2002; Lemon and Lins, 2003; Kim et al., 2007; Almeida et al., 2011; Joh 2003). For a
AC
controlling shareholder and his/her related parties over an affiliated firm i, AffiliateCFji as a
direct ownership portion of an affiliated firm j over another affiliated firm i, and CFj as
total ownership portion of controlling shareholder and his/her related parties over an
portion over firm i. Considering all direct and indirect cash flow rights, the total cash flow
15
ACCEPTED MANUSCRIPT
following equations.9
T
N
CFi DirectCFi CF j AffiliateC F ji , (i 1,2,...., N)
P
j 1
RI
The total voting right is measured as the sum of the voting rights controlled by a
SC
controlling shareholder and his/her related parties, which include relatives, senior managers,
non-profit organizations, and for-profit corporations under his/her de facto control. Cash
NU
flow rights held by a controlling shareholder will be lower than control rights when indirect
stakes with less than full ownership are used in a firm’s ownership structure. For example,
MA
suppose a controlling shareholder holds a 20 percent share in firm A and a 51 percent share
ED
in firm B. Suppose further that firm B holds a 30 percent share in firm A. If we include firm
B in our computation, its control over firm A’s voting rights would be 50 percent
PT
(=20%+30%) and the cash flow rights would be 35.3 percent (=20%+51%×30%). Prior
CE
the difference of control rights and cash flow rights (Joh, 2003; Kim and Yi, 2006) or the
AC
ratio of control rights and cash flow rights (Lins, 2003). We thus employ both measures in
9
The KFTC provides the control-ownership wedge each year. We do not use this measure because cash flow
right measurement is based only on DirectCFi not considering the indirect ownership. Therefore, cash flow
rights are underestimated in KFTC’s measure, which subsequently overestimate the control-ownership wedge.
10
Alternatively, this measure is called as the control-ownership ratio. Note that we add 1 to both numerator
and denominator to reduce the potential small denominator effect.
16
ACCEPTED MANUSCRIPT
Our dependent variable of this study, RPTt, is computed as the total amount of
operating sales and purchases and non-operating transactions with related parties11 divided
P T
RPTt =
RI
SC
Non-operating transactions with related parties include sales/purchases of property,
plants, and equipment (PPE) and investment assets. In order to test hypothesis 1a, 1c, and
NU
1d, we define TOP5 as one if a firm belongs to one of five largest chaebols, and 0 otherwise.
MA
In order to test hypothesis 1b, 1c, and 1d, this study classifies RPTs in terms of operating
sales and purchases (RPT_O) and non-operating transactions (RPT_NO) defined as follows:
ED
RPT_Ot =
PT
RPT_NOt =
CE
We use the following models to test our main hypotheses. Models (1) and (2) are used
to test Hypotheses 1 and 1a. Model (3) and (4) is used to test Hypothesis 1b, 1c, and 1d.
11
International Accounting Standard 24 (Related Party Disclosure) provides detailed definition/discretion of
RPTs. Korean GAAP (Generally Accepted Accounting Principles) follows the definition of RPTs as
described in IAS 24.
17
ACCEPTED MANUSCRIPT
T
RPT_O (or RPT_NO)t = β0 + β1DVC +β2 TOP5 t + β3 DVC×TOP5t + β4 SIZE t
P
+ β5LEVt + β6ROAt + β7AVOIDt + β8MBt + β9RDt + β10CGIt
+ Σ YEARt + Σ INDt + εt
RI
(4)
SC
where,
RPT = [Amount of related-party transactions (operating sales and purchases + non-operating
transactions)]/Market value of equity;
RPT_O = [Amount of related party operating sales and purchases] / Market value of equity;
NU
RPT_NO = [Amount of related party non-operating transactions ] / Market value of equity;
DVC1 = Control-ownership wedge of the controlling shareholders computed as (Voting right-
Cash flow right);
DVC2
MA
= Control-ownership ratio of the controlling shareholders computed as (1+Voting right) /
(1+Cashflow right);
VR = Controlling shareholder’s voting right;
CR = Controlling shareholder’s cash flow right;
ED
The major variable of interest in model (1) is DVC (i.e., the control-ownership
12
Descriptive statistics of the factors used in computing CGI are as follows:
Variable N Mean Median Std. Min. Max.
Proportion of outside director 982 0.386 0.375 0.180 0.0 0.9
1 if all audit committee members are independent, 0 otherwise 982 0.538 1.000 0.499 0.0 1.0
1 if Big 4, 0 otherwise 982 0.930 1.000 0.256 0.0 1.0
13
We set year 2010 as our base year and Service Industry and Hotel, Transportation, Real Estate, Education,
Entertainment, Other Services as a base industry to control for year and industry fixed effect.
18
ACCEPTED MANUSCRIPT
wedge). A significant positive sign of DVC would suggest that RPTs can be used as a
channel which through the control-ownership wedge leads to firm value destruction.
However, there may be no evidence that DVC is positively associated with RPTs, which
P T
would suggest that controlling shareholders do not use RPTs. Model (2) is used to test
RI
Hypothesis 1a in which we examine the incremental effect of the control-ownership wedge
SC
for the top 5 chaebols compared to non-top 5 chaebols. To capture the effect of the top 5
chaebols, we insert TOP5 as well as an interaction term DVC×TOP5 to model (1). If the
NU
positive association is stronger for the top 5 chaebols than for non-top 5 chaebols, both β1
MA
and β3 are expected to be positive. In order to test Hypothesis 1b, we use model (3) by
classifying RPTs into operating (RPT_O) and non-operating transactions (RPT_NO) and
ED
terms of RPT_O and RPT_NO, the coefficient of DVC (β1) would be not different between
PT
RPT_O and RPT_NO. On the other hand, if controlling shareholders prefer non-operating
CE
RPT_NO are expected to be greater than β1 for RPT_O. We add TOP5 and DVC×TOP5 to
AC
model (4) to see if the size of business group affects firms’ choice between operating and
Hypotheses 1c and 1d, we separate RPT_O and RPT_NO for the top 5 and non-top 5
Control variables used in Models (1), (2), (3) and (4) are taken from Jian and Wong
(2010) and Gordon et al. (2004). SIZE is expected to have a positive association with RPTs,
as large firms may have more transactions with related parties. LEV is expected to have a
19
ACCEPTED MANUSCRIPT
positive association since firms with a high debt ratio have incentives to opportunistically
manage earnings to avoid debt covenant violation using RPTs (DeFond and Jiambalvo,
1994, Kim and Yi, 2006). ROA is expected to be negatively associated with RPTs as firms
P T
with poor performance are more likely to have incentives to make opportunistic earnings
RI
management through RPTs (Skinner and Sloan, 2002). Firms who meet or beat the zero
SC
earnings benchmark marginally tend to use RPTs as a means for opportunistic earnings
management (Jian and Wong, 2010). We thus include AVOID which is expected to have a
NU
positive association with RPTs. We also control for the effect of market-to-book ratio (MB)
MA
on RPTs. RPTs are more likely to happen under high market-to-book ratio (MB) because
overvalued firms are more likely to manage earnings via RPTs to avoid decreases in stock
ED
price (Skinner and Sloan, 2002). However, firms having high-growth opportunities (i.e.
high market-to-book ratio) are less likely to use RPTs as a means for propping or tunneling.
PT
We thus have no directional expectation for MB. Gordon et al. (2004) report that shared
CE
R&D arrangements are often associated with RPTs. We thus expect positive association
between RD and RPTs. Gordon et al. (2004) and Cheung et al. (2009) report the importance
AC
of corporate governance mechanisms such as the board of directors, audit committee, and
external auditors in RPTs. This study, therefore, includes CGI representing the
indicators representing independences of board of directors, audit committee, and the Big 4
auditors. Firms with good governance are expected to reduce RPTs in order to protect
minority shareholders. Thus, CGI is expected to have a negative association with RPTs.
This study covers publicly listed chaebol firms in Korea from 2003 through 2010.
20
ACCEPTED MANUSCRIPT
Financial firms and non-December year-end firms are excluded to make the sample
comparable. This study also excludes deficit firms and delisted firms. Financial and stock
market data used in this study are extracted from the Data Guide Pro database provided by
P T
FnGuide Co., and related-party transaction data are from the Kis-value database provided
RI
by NICE Information Service Co., Ltd. Data Guide Pro and Kis-value is equivalent to
SC
CRSP and Compustat in US, respectively covering all publicly listed firms in Korea. These
procedures leave the final sample of 982 firm-years representing 181 firms from 57
NU
business groups on average across years.14 Table 1 provides the summary of sample
MA
selection procedures.
********************
Insert Table 1 here
ED
********************
PT
dependent variable of this study (RPT) represents the proportion of RPTs to the market
value of equity. The magnitude of RPTs is quite large for our sample firms with a mean
14
The KFTC (2011) reported that there are 172 publicly traded firms that belong to large business groups as
of the end of 2010 among 770 firms listed on the Korea Stock Exchange. See http://groupopni.ftc.go.kr
for detailed information of Korean business groups.
Year 2003 2004 2005 2006 2007 2008 2009 2010 2011
Business groups with total assets:
Over 2 trillion Korean won Over 5 trillion Korean won
N 49 51 55 59 62 79 48 53 55
15
In order to mitigate the effect of extreme values on the results, all continuous variables are winsorized at
the top and bottom one percentile.
21
ACCEPTED MANUSCRIPT
value of 1.041. RPT_O and RPT_NO have mean values of 0.413 and 0.585, respectively. It
is noteworthy that the magnitude of non-operating RPTs is greater than the magnitude of
operating RPTs. The control-ownership wedge (DVC1) has a mean value of 0.214. Voting
P T
rights (VR), one of the components of the control-ownership wedge, has a mean value of
RI
0.411; cash flow rights (CR), the other component, has a mean value of 0.197. From the
SC
perspective of ownership percentage, controlling shareholders have about 41 percent of
voting rights, while they have about 20 percent cash flow rights, showing that the control-
NU
ownership wedge is very large among Korean chaebol firms. Among control variables, the
MA 16
mean value of the natural log of sales volume (SIZE) is 20.955 representing 1,260.8
billion Korean won; the mean value of ROA is 0.055, indicating that firms make about 5.5
ED
percent of total assets as net income during the sample period. Sample firms spend on
average about 1.6 percent of total sales on research and development (RD).
PT
Table 2, Panel B, compares the top 5 chaebols with non-top 5 chaebols. It shows
CE
that the top 5 chaebols perform significantly more RPTs, and have significantly lower
voting and cash flow rights than those of non-top 5 chaebols. In particular, the top 5
AC
chaebols tend to have cash flow rights significantly lower than voting rights making the
********************
Insert Table 2 here
********************
coefficient between DVC1 and RPT is 0.126 (p-value < 0.001), and the coefficient between
16
SIZE is defined as the natural logarithm of total sales in thousand Korean Won.
22
ACCEPTED MANUSCRIPT
CR and RPT is -0.104 (p-value = 0.001) indicating that firms with a larger wedge have
more RPTs while firms with higher cash flow rights (CR) have less RPTs. The correlation
coefficient between DVC1 and RPT_O is 0.081 (p-value = 0.011), and the coefficient
P T
between DVC1 and RPT_NO is 0.099 (p-value = 0.002). The correlation coefficients
RI
between dependent variables and control variables are generally consistent with our
SC
expectations. Table 3 also shows that RPT is positively correlated with SIZE, LEV, and
AVOID, while it is negatively correlated with ROA, MB, and CGI. The highest correlation
NU
among independent variables is between SIZE and CGI (0.627, p-value = 0.000). 17
MA
Variance Inflation Factor (VIF) values in all empirical models do not exceed 10, suggesting
********************
Insert Table 3 here
********************
PT
Table 4, Column (1) and (2) show the results of testing whether there is a positive
RPTs in our study include operating sales, purchases, and non-operating transaction with
wedge (DVC1) and control-ownership ratio (DVC2). The coefficients of DVC1 and DVC2
in Column (1) and (2) are 1.232 (t-value=4.60) and 1.365 (t-value=4.76), indicating that
17
Dropping CGI from the model does not change our conclusion.
18
The largest variance inflation factor (VIF) values are 2.36 in hypothesis 1, 4.93 in hypothesis 1a, 4.93 in
hypothesis 1b, and 3.11 in other analyses. The VIFs of the additional analysis 1a and 1b are larger probably
because of the interaction variable.
23
ACCEPTED MANUSCRIPT
firms with a larger control-ownership wedge tend to perform more RPTs. The results
support the notion that RPTs arise from the agency conflicts between controlling
shareholders and minority shareholders (Morck et al., 1988; Fan and Wong, 2005).
P T
For our control variables, both the magnitude and statistical significance of the
RI
parameter estimates are fairly stable across the three model specifications shown in Table 4.
SC
Most of the parameter estimates for the control variables are consistent with our prediction.
Specifically, we observe that SIZE has a significantly positive effect on RPTs. LEV also has
NU
a significantly positive effect on RPTs, suggesting that firms are more likely to rely on
MA
RPTs as their financial risk increases. ROA has a positive, but insignificant effect on RPTs.
AVOID has a significant positive effect on RPTs, suggesting that firms that marginally meet
ED
or beat zero earnings target are more likely to manage earnings by using RPTs (Jian and
Wong, 2010). MB has a significant negative effect on RPTs, suggesting that growing firms
PT
are less likely to use RPTs possibly because of weak incentives for opportunistic earnings
CE
management. This result is consistent with the finding of Jian and Wong (2010) that firms
with high growth rates decrease propping or tunneling through RPTs. Consistent with
AC
Gordon et al. (2004), RD has a significant positive effect on RPTs. Finally, CGI, a proxy
for the effectiveness of corporate governance, has a significant negative effect on RPTs,
suggesting that in firms with strong governance RPTs are more closely monitored and thus
We also analyze how voting rights and cash flow rights affect RPTs separately by
replacing DVC with VR and CR in regression model.19 In Table 4, Column (3), we report
19
Additionally, we test the effect of VR and CR in separate regressions. The estimation results are that the
effect of VR on RPTs is positive though not significant (t-value=1.45) and the effect of CR on RPTs is
24
ACCEPTED MANUSCRIPT
the estimation results. The coefficient estimate of VR is 1.088, significant at the 1 percent
level, suggesting that controlling shareholders who have higher voting rights are more
likely to use RPTs to pursue their own benefits. On the other hand, the coefficient estimate
P T
of CR is -1.377, significant at the 1 percent level, suggesting that RPTs decrease as cash
RI
flow rights increase because RPTs lower the value of controlling shareholders’ equity stock
SC
(i.e., firm value), which is consistent with the alignment of interests effect (Jensen and
Meckling, 1976; La Porta et al., 1999). The results in Column (3) suggest that as a result of
NU
large difference between CR and VR controlling shareholders are more likely to use RPTs
MA
for their own entrenchment supporting our argument that RPTs are an important channel
benefits.
********************
PT
depending on the size of business groups. A large business group generally has more
affiliated firms, which provides more opportunities for RPTs. For example, Hyundai
Motors Inc. has a steel manufacturing company as an affiliated firm, while General Motors
Korea Inc. does not have steel manufacturing company as an affiliated firm. In this case,
relative to General Motors Korea Inc., Hyundai Motor Inc. has more opportunities to
conduct RPTs.
significantly negative, which is generally consistent with the estimation results including both CR and VR in
the same regression model.
25
ACCEPTED MANUSCRIPT
In order to test Hypothesis 1a, we add an indicator variable, TOP5, which equals
one for the largest five business groups and zero for all other groups. We then construct an
P T
of DVC on RPTs for the top 5 chaebols. Hypothesis 1a predicts that the interaction term has
RI
a positive effect on RPTs. Estimation results in Table 5, Panel A, show that TOP5 has
SC
significant positive effect on RPTs in the first two specifications, indicating that largest five
business groups tend to perform more RPTs for the same level of the control-ownership
NU
disparity. Furthermore, a significant positive coefficient (1.067) of DVC×TOP5 indicates
MA
that for the same degree of the control-ownership wedge the top 5 chaebols tend to perform
more RPTs than non-top 5 chaebols.20 To check the robustness of the results in Panel A,
ED
we perform a sub-sample test and provide the results in Panel B of Table 5. Column (1)
reports estimation results for the top 5 group. The coefficient estimate of DVC is 2.078 for
PT
the top 5 chaebols and it is significantly greater than the coefficient estimate of DVC (0.742)
CE
for non-top 5 chaebols (parallelism test: t-value=1.81, significant at the 10 percent level)
corroborating the results in Panel A.21 Overall, our evidence in Table 5 reveal that the
AC
positive association between RPTs and the control-ownership wedge is more pronounced in
********************
Insert Table 5 here
********************
20
We confirm that the results with DVC2 are qualitatively the same as those with DVC1. In Panel A (3), the
coefficient of DVC2×TOP5 is 1.088 (t-value=1.67) while in Panel B (1), (2), the coefficients of DVC2 are
2.233 (t-value=3.18) and 0.826 (t-value=2.50), respectively. For convenience’ sake, DVC1 is referred as DVC
from Table 5.
21
Alternatively, we perform additional tests by replacing a top 5 dummy to top 3, top 4, top 6, and top 7, etc.
Untabulated results show that the incremental effect of DVC×TOP_N increases as N decreases and is
significant only in the top 3 and top 4 groups. This result implies that the incremental effect of top business
groups is significant only for the top 5 chaebols and larger chaebols.
26
ACCEPTED MANUSCRIPT
To test Hypothesis 1b, we divide RPTs into two types: (1) sales and purchases
P T
transactions as parts of operating transaction and (2) non-operating transactions. Then we
RI
replace our main dependent variable, RPTs, in model (1) with RPT_O and RPT_NO,
SC
respectively. Table 6 presents the estimation results from the regression model (3) & (4),
examining whether the effect of control-ownership wedge on RPTs varies depending on the
NU
nature of RPTs. MA
In Table 6, Panel A, for both operating and non-operating RPTs, the coefficient
estimate of DVC22 is positive and significant, indicating that the control-ownership wedge
ED
is positively associated with both operating and non-operating RPTs. The coefficient
estimate of DVC for non-operating RPTs is 0.783, which is greater than the coefficient
PT
estimate of DVC for operating RPTs (0.447), but the difference is not statistically
CE
This result suggests that on average firms use both operating and non-operating RPTs
AC
Firms’ choice between operating and non-operating RPTs depends also on the
availability of the proper transaction partner in the group. Large chaebols have various
firms in different industries while relatively small chaebols have a small number of firms in
22
Table 6 reports the results with DVC1. We also find that the results with DVC2 are qualitatively the same
as those with DVC1. In Panel A (1), (2), the coefficients of DVC2 are 0.479 (t-value=2.25) and 0.880 (t-
value=4.16), respectively while, in Panel B (1) and (2), the coefficients of DVC2*TOP5 are 1.170 (t-
value=2.42) and -0.325 (t-value=-0.67), respectively.
23
Our p-value is obtained from the bivariate normal regression analysis in which we include both
specification (1) & (2) of Panel A, Table 6 in the same regression model (see Timm, 1975; Bock, 1975;
Morrison, 1976; and Mardia et al., 1979 for details of multivariate tests).
27
ACCEPTED MANUSCRIPT
a few industries. The difference in group size may allow large chaebols with high DVC to
use operating RPTs more effectively for tunneling than others. We find DVC×TOP5 in
Column (1) in Panel B, Table 6 is significant while DVC×TOP5 in Column (2) in Panel B
P T
is insignificant. This result suggests that the top 5 chaebols use operating RPTs more
RI
extensively when the wedge is high supporting Hypothesis 1c.
SC
In addition, in Table 6, Panel C, we separate the sample into the top 5 and non-top 5
chaebol sub-samples. While DVC is significant with both RPT_O and RPT_NO in the top 5
NU
chaebols, DVC is significant only with RPT_NO in non-top 5 chaebols suggesting that non-
MA
top 5 firms use non-operating RPTs. This result supports Hypothesis 1d.
********************
ED
So far we provide evidence that there is a positive association between the control-
AC
ownership wedge and RPTs arguing that RPTs are used as an important channel through
which the high control-ownership wedge leads to firm value destruction. However, the
positive association alone may not be direct evidence of tunneling without showing that
increased RPTs affect firm value negatively. In this section, we test whether RPTs of
To assess the relation between RPTs and firm value, we use the following OLS
regressions in which Tobin’s Q is the dependent variable and RPTs is the main independent
variables.
28
ACCEPTED MANUSCRIPT
Variable Definitions:
T
Tobin Q = (Market value of equityt+1 + Book value of debtt+1) / Total assetst+1;
P
GROW = (Salest – Salest-1) / Salest-1;
RI
AGE = Natural logarithm of (Current year – Establishment year);
The rest of variable definitions are discussed in Models (1), (2), (3), and (4).
SC
Tobin’s Q is widely used as a measure of firm value in the existing literature. To
NU
facilitate the inference on the causal relationship that runs from RPTs to Tobin’s Q we use a
lead value as of (t+1) for Tobin’s Q in the model. As suggested by the conflict of interest
MA
hypothesis, if RPTs are used as means of tunneling firm value will be negatively affected
ED
by them (Kohlbeck and Mayhew, 2010; Cheung et al., 2006). Therefore, if RPTs of Korean
chaebols destroy firm value, RPT will have a negative coefficient (i.e., β1<0). Our
PT
regression includes a variety of control variables to ensure that the effects we attribute to
CE
RPTs are not due to other correlated factors. Specifically, we control for firm size defined
as the natural logarithm of assets (ASSET). We add leverage (LEV) to control for its effect
AC
on cost of capital (Modigliani and Miller, 1963). Firm profitability and growth are added
because they are positively associated with firm value (Smith and Watts, 1992; Yermack,
1996; Myers, 1997). Research and development (RD) is expected to increase firm value
(Chan et al., 1990) and age (AGE) is negatively associated with firm value (Drobetz et al.,
2004; Black et al., 2006). We thus add them as additional control variables. Lastly, we
control for governance quality (CGI). Table 7 presents the estimation results from the
regression model (5). Table 7 shows that RPTs have a significant negative effect on Tobin Q
29
ACCEPTED MANUSCRIPT
We next test whether value destruction by RPTs is more severe when a firm has
P T
where DVCD equals one if DVC is greater than or equal to the sample median and zero
RI
otherwise.
SC
Tobin Qt+1 = β0 + β1RPTt + β2RPTt×DVCDt + β3DVCD + β4ASSETt + β5LEVt
+ β6 ROA t + β7GROWt + β8RDt+ β9AGEt + β10CGIt
+ Σ YEARt + Σ INDt + εt (6)
NU
Estimation results of Column (2) of Panel A, Table 7, show that the coefficient of
MA
RPT is negative but insignificant (-0.021). The coefficient of interaction term (RPT×DVCD)
is negative and significant (-0.041). Subsample tests in Panel B, Table 7 show similar
ED
results. For both high and low wedge groups, the coefficient of RPTs is negative, but is
PT
significant only for the high control-ownership wedge group. These results suggest that
value destruction by RPTs occurs only when the control-ownership wedge is high, that is,
CE
the agency conflicts between controlling shareholders and minority shareholders are severe.
AC
Prior studies provide inconsistent evidence on whether RPTs are more likely to be used as a
means of tunneling or propping (Shin and Park, 1999; Chang and Hong, 2000; Johnson et
al., 2000 etc.). The results in Table 7 suggest that this inconsistency may be at least partially
attributed to the lack of controlling the effect of agency problems in prior studies. When the
agency conflicts are properly controlled like the estimations in Table 7 RPTs do not always
destroy firm value. In sum, the results suggest that RPTs are used as an important channel
through which the high control-ownership wedge leads to firm value destruction.
********************
Insert Table 7 here
30
ACCEPTED MANUSCRIPT
********************
Consistent with the notion that the top 5 chaebols are more likely to use RPTs as a
T
means of tunneling (Jung and Kwon, 2002), we predict that the negative association
P
RI
between RPTs and firm value is more pronounced for the top 5 chaebols.
SC
Estimation results in Table 8 show that the negative association between RPTs and
NU
firm value is more pronounced in the top 5 firms than non-top 5 firms consistent with the
argument that the top 5 chaebols are more likely to use RPTs as a means for wealth transfer
MA
from minority shareholders to controlling shareholders. However, Table 8, Panel B, shows
that the effect of RPTs is significant only for high-wedge subsample firms, consistent with
ED
the results in Table 7. We also examine the possibility of differential effect of RPTs on firm
PT
value between operating and non-operating RPTs. In Table 8, Panel C, we report that the
effect of operating and non-operating RPTs is insignificant for the low-wedge subsample.
CE
For high-wedge group, both operating and non-operating RPTs destroy the top 5 chaebol
AC
********************
Insert Table 8 here
********************
5. Conclusion
Using firm-level RPTs data from large Korean firms in which the ownership
disparity is also large, this study examines the association between the control-ownership
wedge and RPTs and the effect of RPTs on firm value. Prior studies report that the agency
31
ACCEPTED MANUSCRIPT
problem between controlling shareholders and minority shareholders amplifies as the level
affects firm value. However, prior studies provide contrasting views on the effect of RPTs
P T
on firm value, e.g., efficient transaction hypothesis vs. conflict of interest hypothesis. In
RI
this study, we aim to disentangle the inconsistent effect of RPTs on firm value reported in
SC
prior studies by investigating the relationship between RPTs and firm value conditional on
the extent of the agency conflict between controlling shareholders and minority
NU
shareholders. MA
In this paper, we show that RPTs increase as the control-ownership wedge increases,
suggesting that controlling shareholders who have a greater wedge are more likely to use
ED
RPTs opportunistically. We also show that RPTs increase as voting rights increase, while
RPTs decrease as cash flow rights increase, suggesting that controlling shareholders with a
PT
strong influence over a firm’s decision-making tend to prefer RPTs. We find that the
CE
positive association between the control-ownership wedge and RPTs is more pronounced in
the top 5 chaebols of Korea wherein the agency conflicts between controlling shareholders
AC
and minority shareholders are more severe. Analyses that separately examine operating and
non-operating transactions show that a significant positive association between the control-
ownership wedge and RPTs holds for both types of RPTs in the top 5 chaebols. However,
non-top 5 firms typically use non-operating RPTs as a primary means of tunneling. Finally,
we show that RPTs of Korean chaebol firms reduce firm value only when the control-
ownership wedge is high and this value destruction becomes more severe in the top 5
chaebols. Overall, our results together suggest that RPTs occur when the agency problem is
severe and are used as a means of tunneling and thus destroying firm value. Our results also
32
ACCEPTED MANUSCRIPT
suggest that RPTs not arising from the agency conflict do not affect firm value negatively
indicating they are not related to tunneling. The implications of our study can be extended
P T
RI
SC
NU
MA
ED
PT
CE
AC
33
ACCEPTED MANUSCRIPT
Reference
Almeida, H., S. Y. Park, M. G. Subrahmanyam, and D. Wolfenzon. 2011. The structure and
formation of business groups: evidence from Korean chaebols. Journal of Financial
Economics 99, 447-475.
T
Bae, K., J. Kang, and J. Kim. 2002. Tunneling or value added? evidence from mergers by
P
Korean business groups. The Journal of Finance 6, 2695-2740.
RI
Baek, J., J. Kang, and K. Park. 2004. Corporate governance and firm value: evidence from
the Korean financial crisis. Journal of Financial Economics 71, 265-313.
Bertrand, M., P. Mehta, and S. Mullainathan. 2002. Ferreting out tunneling: an application
SC
to Indian business groups. Quarterly Journal of Economics 118, 121-148.
Black, B., H. Jang, and W. Kim. 2006. Does corporate governance affect firms’ market
value? evidence from Korea. Journal of Law, Economism and Organization. 22,
NU
366-413.
Bock, R. D. 1975. Multivariate Statistical Methods in Behavioral Research. New York:
McGraw-Hill.
MA
Campbell II, T.L., and P.Y. Keys. 2002. Corporate governance in South Korea: the chaebol
experience. Journal of Corporate Finance 8(4), 373-391.
Chan, S. H., J. D. Martin, and J. W. Kensinger. 1990. Corporate research and development
expenditures and share value. Journal of Financial Economics 26, 255-276.
ED
Cheung, Y., P. Rau., and A. Stouraitis. 2006. Tunneling, propping and expropriation:
evidence from connected party transactions in Hong Kong. Journal of Financial
Economics 82, 343-386.
CE
Cheung, Y., Y. Qi, P. Rau, and A. Stouraitis. 2009. Buy high, sell low: how listed firms
price asset transfers in related party transactions. Journal of Banking & Finance 33,
914-924.
AC
Claessens, S., S. Djankov, and L.H.P. Lang. 2000. The separation of ownership and control
in East Asian corporations. Journal of Financial Economics 58, 81–112.
Claessens, S., S. Djankov, J.P.H. Fan, and L.H.P. Lang. 2002. Disentangling the incentive
and entrenchment effects of large shareholdings. The Journal of Finance 57, 2741–
2771.
Claessens, S., J.P.H. Fan, and L.H.P. Lang. 2006. The benefits and costs of group affiliation:
Evidence from East Asia. Emerging Markets Review 7, 1-26.
DeFond. M., and Jiambalvo J. 1994. Debt covenant violation and manipulation of accruals.
Journal of Accounting and Economics 17, 145-176
Drobetz, W., A. Schillofer, and H. Zimmermann. 2004. Corporate governance and expected
stock returns: evidence from Germany. European Financial Management 10, 267-
293.
Fan, J., and T. J. Wong. 2005. Do external auditors perform a corporate governance role in
emerging markets? evidence from East Asia. Journal of Accounting Research 43
(1), 35-72.
34
ACCEPTED MANUSCRIPT
Friedman, E., S. Johnson and T. Mitton. 2003. Propping and tunneling. Journal of
Comparative Economics 31, 732–750.
Glaeser, E., S. Johnson and A. Shleifer. 2001. Coase versus the Coasians. Quarterly Journal
of Economics 116, 853–899.
Gompers, P. A., J. Ishii, and A. Metrick. 2010. Extreme governance: an analysis of dual
T
class firms in the United States. Review of Financial Studies 23(3), 1051-1088.
P
Gordon, E., E. Henry, and D. Palia. 2004. Determinants of related party transactions and
RI
their impact on firm value. Working paper.
Hankyre News, April 17, 2013. “17 percent of total sales in the top 5 chaebols are related
party transactions”, Available at
SC
http://www.hani.co.kr/arti/economy/economy_general/583356.html. [printed in
Korean].
Hankyore News, February 13, 2012. “Lives of ordinary people are besieged by 0.1%
NU
chaebols”, Available at
http://www.hani.co.kr/arti/economy/economy_general/518598.html. [printed in
Korean].
MA
Harvey, C.R., K.V. Lins, and A.H. Roper. 2004. The effect of capital structure when
expected agency costs are extreme. Journal of Financial Economics 74, 3-30.
Jensen, M., and W. Meckling. 1976. Theory of the firm: managerial behavior, agency costs
and ownership structure. Journal of Financial Economics 3, 305-360.
ED
Jian, M. and T. J. Wong. 2004. Earnings management and tunneling through related party
transactions: Evidence from Chinese corporate groups. American Accounting
Association, 2004 Annual Conference Paper.
PT
Jian, M., and T. Wong. 2010. Propping through related party transactions. Review of
Accounting Studies 15, 70-105.
Joh, S. W. 2003. Corporate governance and firm profitability: evidence from Korea before
CE
Jung, K., and S.Y. Kwon. 2002. Ownership structure and earnings informativeness
Evidence from Korea. The International Journal of Accounting 37(3), 301-325.
Khanna, T., and K. Palepu. 1997. Why focused strategy may be wrong in emerging markets.
Harvard Business Review 75(4), 41-51.
Kim, J., and C. Yi. 2006. Ownership structure, business group affiliation, listing status and
earnings management: evidence from Korea. Contemporary Accounting Research
23 (2), 427-464.
Kim, W., Y. Lim, and T. Sung. 2007. Group control motive as a determinant of owernship
structure in business conglomerates evidence from Korea’ chaebols. Pacific-Basin
Finance Journal 15: 213-252.
Kohlbeck, M., and B.W. Mayhew. 2010. Valuation of firms that disclose related party
transactions. Journal of Accounting and Public Policy 29(2), 115-137.
Korea Accounting Standards Board. 2004. K-GAAP No. 20, Related Party Transactions
Disclosure.
Korea Fair Trade Commission (KFTC). 2011. Guidelines for Reviewing Unfair Assistance.
Available at: http:// www.ftc.go.kr/laws/laws/laws.jsp?lawDivCd=01.
35
ACCEPTED MANUSCRIPT
T
Korean Institute of Certified Public Accountants. 2005. Auditing Standards No. 550
P
Related Party.
Kyunghyang News, June 23. 2013. “More than a half of total sales of the top 5 chaebols are
RI
related party transactions”, Available at
http://news.khan.co.kr/kh_news/khan_art_view.html?artid=201306232228075&cod
SC
e=920401. [printed in Korean].
La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1997. Legal determinants of
external finance. Journal of Finance 52, 1131–1150.
NU
La Porta, R., F. Lopez-de-Silanes, A. Shleifer and R. Vishny. 1998. Law and finance.
Journal of Political Economy 106, 1113–1155.
La Porta, R., F. Lopez-de-Silanes, and A. Shleifer. 1999. Corporate ownership around the
MA
world. The Journal of Finance 54(2), 471–518.
La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R.W. Vishny. 2000. Investor protection
and corporate governance. The Journal of Financial Economics 58, 3–27.
La Porta, R., F. Lopez-de-Silanes, A. Shleifer, and R.W. Vishny. 2002. Investor protection
ED
58(4), 1445–1468.
Lins, K. V. 2003. Equity ownership and firm value in emerging markets. Journal of
Financial and Quantitative Analysis 38(1), 159–184.
CE
Mardia, K. V., Kent, J. T., and Bibby, J. M. 1979. Multivariate Analysis. London:
Academic Press.
Ministry of Strategy and Finance. 2010. Corporation Tax Act, Article 52, Rejection of
AC
36
ACCEPTED MANUSCRIPT
Ryngaert, M. and S. Thamas. 2007. Related party transactions: their origins and wealth
effects. Working paper.
Shin, H., and Y. S. Park. 1999. Financing constraints and internal capital markets: evidence
from Korean chaebols, Journal of Corporate Finance 5, 169-194.
Shleifer, A., and R. Vishny. 1997. A survey of corporate governance. The Journal of
T
Finance 52 (2): 737-783.
P
Skinner, D., and R. Sloan. 2002. Earnings surprises, growth expectations, and stock returns
or don’t let an earnings torpedo sink your portfolio. Review of Accounting Studies
RI
7, 289-312.
Smith, C. and R. Watts. 1992. The investment opportunity set and corporate financing,
SC
dividend, and compensation policies. Journal of Financial Economics 32, 263-292.
Stein, J. 1997. Internal capital markets and the competition for corporate resources. Journal
of Finance 52, 111 – 133.
NU
Timm, N. H. 1975. Multivariate Analysis with Applications in Education and Psychology.
Monterey, CA: Brooks-Cole.
Wall Street Journal, May 7, 2003.“Even good insider deals raise doubts”, B6.
MA
Williamson, O. 1975. Market and Hierarchies: Analysis and Antitrust Implications. New
York: Free Press.
Yermack, D. 1996. Higher market valuation of companies with a small board of directors.
Journal of Financial Economics 40, 185-111.
ED
PT
CE
AC
37
ACCEPTED MANUSCRIPT
APPENDIX
T
Classification Content
P
Designated The Fair Trade Commission determines a business group as a group of
RI
Business firms in which “the same person virtually dominates the business
Groups decisions of a group” and specifies, among business groups, a large scale
SC
business group (chaebol) as having total assets of 5 trillion Won (2 trillion
Won before 2009). [Fair Trade Act Decree 17 Article 9, amended 2008]
Regulations by Through “Monopoly Regulations and Fair Trade Law”, business groups
NU
the Fair Trade are under strict discipline as for internal transactions.
Commission (1) Chaebol firms must undergo a board decision and disclose relevant
information when large internal transactions with related parties (i.e.,
more than 10 percent of company capital stock or more than a 10 billion
MA
Won deal) are made. [KFTC, Regulation on Resolution of Board of
Directors and Disclosures on Large-Scale Intra-Group Transactions,
2012].
(2) Unfair internal transactions among chaebol affiliates are audited and
ED
are subject to penalties. Here, unfair internal transactions are those with
related parties or other companies who provide goods, services, capital,
assets, and personnel free of charge or on favorable terms which might
PT
groups, shares, transactions with related parties, etc. [Fair Trade Act
Article 11, amended 2007 and Decree 11 Article 17, amended 2009]
AC
38
ACCEPTED MANUSCRIPT
cannot obtain sufficient and appropriate audit evidence about the related
parties and transactions and/or the disclosure on related party transactions
are inappropriate, auditors are mandated to modify audit reports. [KICPA,
Auditing Standards No. 550 Related Party, 2005]
T
Tax Law Through the provision of rejection of unfair acts and calculations, in the
case wherein special transactions are conducted with significantly lower
P
or higher prices than fair market price, all persons who received personal
RI
benefits are subject to income tax or inheritance or gift tax. Here, related
party transactions include transactions of goods and services as well as
direct transactions or indirect transactions through stocks etc. [MSF,
SC
Corporation Tax Act, Article 52, Rejection of Unfair Act and Calculation,
amended 2010],[MFS, Income Tax Act, Article 41, Unfair Act and
Calculation, amended 2009], [MFS, Inheritance Tax and Gift Tax Act,
NU
Article 4, Obligation to Pay Gift Tax, amended 2013]
MA
ED
PT
CE
AC
39
ACCEPTED MANUSCRIPT
TABLE 1
Panel A of TABLE 1 represents the procedures of sample selection. Panel B represents sample distribution
and the list of 5 largest Business Groups by year. Panel C reports the sample distribution by industry.
T
Criteria Number of Observations
4,986
P
Total related party transactions between 2003 and 2010
Less: Financial institutions 352
RI
Less: Non-December fiscal year firms 149
Less: Delisted firms 96
SC
Less: Deficit firms 9
Less: Firms without DVC data 3,398
Final sample 982
NU
Panel B. Sample Distribution by Year and the Description of 5 Largest Business Groups
year N 5 largest business groups
103 Samsung, LG, SK, Hyundai Motors, Hanjin
2003
MA
2004 110 Samsung, LG, Hyundai Motors, SK, Hanjin
2005 121 Samsung, Hyundai Motors, LG, SK, Lotte
2006 127 Samsung, Hyundai Motors, SK, LG, Lotte
2007 135 Samsung, Hyundai Motors, SK, LG, Lotte
ED
Total 982
Industry N
Farming, Fishery, Mining, Forestry 2
Food, Beverage, Tobacco 54
AC
40
ACCEPTED MANUSCRIPT
TABLE 2
Panel A reports the descriptive statistics of the variables of total sample.
Variable definition are as follows: (1)RPT = [RPTs (operating sales and purchases + non-operating
transactions)]/Market value of equity, (2)RPT_O = [RPTs (operating sales and purchases)]/Market value of
equity, (3)RPT_NO = [RPTs (non-operating transactions)]/Market value of equity, (4)DVC1 = Ownership
T
wedge 1 of the controlling shareholders (Voting right - Cash flow right), (5)DVC2 = Ownership wedge 2 of
the controlling shareholders (1+Voting right)/(1+Cash flow right), (6)VR = Controlling shareholders’ voting
P
right (%), (7)CR = Controlling shareholders’ cash flow right (%), (8)SIZE = Natural logarithm of total sales
in thousands of Korean won, (9)LEV = Total liabilities / Total equity, (10)ROA = Net income / Total assets,
RI
(11)AVOID =If 0< ROA<2 %, then AVOID is “1.” Else AVOID is “0”, (12) MB = Market value of equity /
Book value of equity, (13) RD = R&D Expenses / Total sales, (14) CGI = Corporate governance
SC
measurement using outside director(s), independence of audit committee, and audit firms.
Panel B represents mean, standard deviation (std.) and number of observations (N) of Top 5 and Non-Top 5
sample, respectively. Top 5 means the firms belongs to one of five largest chaebols. T-values refer to the t-test
statistics of the means between Top 5 sample and Non-Top 5 sample.
NU
Panel A : N=982
Variables Mean Std.
MA Median Min Max
RPT 1.041 1.429 0.478 0.000 8.085
RPT_O 0.413 1.006 0.000 0.000 6.149
RPT_NO 0.585 0.945 0.151 0.000 4.446
DVC1 0.214 0.163 0.211 0.000 0.899
ED
41
ACCEPTED MANUSCRIPT
T
CGI 0.744 0.307 0.614 0.334 6.10 ***
P
RI
SC
NU
MA
ED
PT
CE
AC
42
ACCEPTED MANUSCRIPT
TABLE 3
T
TABLE 3 represents Pearson correlations among the variables used in the regression analysis. P-values are in parentheses.
IP
The definitions of variables are presented in TABLE 2.
N=982
CR
RPT RPT_O RPT_NO DVC1 DVC2 VR CR TOP5 SIZE LEV ROA AVOID MB RD CGI
RPT 1.000 0.616 0.645 0.126 0.129 0.015 -0.104 0.105 0.107 0.157 -0.067 0.125 -0.143 -0.026 -0.059
(0.000) (0.000) (0.000) (0.000) (0.645) (0.001) (0.001) (0.001) (0.000) (0.036) (0.000) (0.000) (0.423) (0.065)
S
RPT_O 1.000 -0.163 0.081 0.079 0.028 -0.047 0.004 -0.012 0.086 -0.043 0.037 -0.129 -0.025 -0.146
(0.000) (0.011) (0.013) (0.379) (0.139) (0.902) (0.709) (0.007) (0.177) (0.251) (0.000) (0.438) (0.000)
NU
RPT_NO 1.000 0.099 0.104 -0.001 -0.095 0.131 0.159 0.110 -0.052 0.123 -0.074 0.000 0.079
(0.002) (0.001) (0.970) (0.003) (0.000) (0.000) (0.001) (0.103) (0.000) (0.020) (0.998) (0.013)
DVC1 1.000 0.990 0.510 -0.405 0.202 -0.138 -0.002 -0.021 0.003 -0.051 -0.087 -0.206
MA
(0.000) (0.000) (0.000) (0.000) (0.000) (0.951) (0.509) (0.933) (0.110) (0.007) (0.000)
DVC2 1.000 0.427 -0.484 0.256 -0.101 0.003 -0.012 -0.009 -0.028 -0.071 -0.174
(0.000) (0.000) (0.000) (0.002) (0.916) (0.709) (0.767) (0.380) (0.026) (0.000)
ED
VR 1.000 0.580 -0.234 -0.448 -0.081 -0.037 0.029 -0.196 -0.162 -0.345
(0.000) (0.000) (0.000) (0.011) (0.252) (0.364) (0.000) (0.000) (0.000)
CR 1.000 -0.440 -0.345 -0.084 -0.019 0.028 -0.160 -0.090 -0.172
PT
(0.000) (0.000) (0.008) (0.554) (0.376) (0.000) (0.005) (0.000)
TOP5 1.000 0.319 -0.132 0.136 -0.093 0.169 0.162 0.186
CE
(0.000) (0.000) (0.000) (0.004) (0.000) (0.000) (0.000)
SIZE 1.000 0.179 0.154 -0.022 0.152 0.009 0.627
(0.000) (0.000) (0.491) (0.000) (0.787) (0.000)
AC
43
ACCEPTED MANUSCRIPT
T
controlling shareholders (1+Voting right)/(1+Cash flow right), (3)VR = Controlling shareholders’ voting right
(%), (4)CR = Controlling shareholders’ cash flow right (%), (5)SIZE = Natural logarithm of total sales in
P
thousands of Korean won, (6)LEV = Total liabilities / Total equity, (7)ROA = Net income / Total assets;
(8)AVOID =If 0< ROA<2 %, then AVOID is “1.” Else AVOID is “0”, (9) MB = Market value of equity /
RI
Book value of equity, (10) RD = R&D Expenses / Total sales, (11) CGI = Corporate governance measurement
using outside director(s), independence of audit committee, and audit firms, (12) IND = Industry indicators,
SC
(13) YEAR = Year indicators.
NU
+ β6MBt+ β7RDt + β8CGIt + Σ IND t + Σ YEAR t + ε t
Dependent = RPT
Exp.
Variables (1) (2) (3)
sign
MA
Coefficients t-value Coefficients t-value Coefficients t-value
Intercept +/- -3.547 -4.96*** -4.816 -5.97 *** -3.155 -3.82***
DVC1 + 1.232 4.60***
DVC2 + 1.365 4.76 ***
ED
VR + 1.088 3.53***
CR - -1.377 -4.46***
SIZE + 0.216 5.83*** 0.212 5.73 *** 0.202 5.04***
PT
Industry
Included Included Included
Dummies
Year
Included Included Included
Dummies
F-value 10.85*** 10.93*** 10.48***
Adjusted-R² 0.207 0.208 0.207
Sample Size 982 982 982
***, **, and * represent a significance at the 1, 5, and 10 percent level, respectively.
44
ACCEPTED MANUSCRIPT
TABLE 5
This table reports the results of association between Wedge and RPT: Top 5 vs. Non-Top 5 chaebols. Panel A
represents the results using Top 5 interaction term, and Panel B shows the results of Top 5 and Non-Top 5
sub-sample. The dependent variable of the regression equation is RPT, which is [RPTs (operating sales and
purchases + non-operating transactions)]/Market value of equity. The independent variable include: (1) DVC
T
= Ownership wedge 1 of the controlling shareholders (Voting right - Cash flow right), (2) TOP5= 1 if a firm
belongs to one of five largest chaebols, and 0 otherwise, (3) SIZE = Natural logarithm of total sales in
P
thousands of Korean won, (4)LEV = Total liabilities / Total equity, (5)ROA = Net income / Total assets;
(6)AVOID =If 0< ROA<2 %, then AVOID is “1.” Else AVOID is “0”, (7) MB = Market value of equity / Book
RI
value of equity, (8) RD = R&D Expenses / Total sales, (9) CGI = Corporate governance measurement using
outside director(s), independence of audit committee, and audit firms, (10) IND = Industry indicators, (11)
SC
YEAR = Year indicators.
NU
+ β7AVOIDt + β8MBt + β9RDt + β10CGIt + Σ IND t + Σ YEAR t + ε t
Dependent = RPT
Exp.
Variables (1) (2) (3)
sign
MA
Coefficients t-value Coefficients t-value Coefficients t-value
Intercept +/- -2.047 -2.77*** -2.828 -3.69*** -2.826 -3.69***
DVC + 0.999 3.54*** 0.743 2.34**
*** **
TOP5 + 0.387 3.88 0.267 2.54 0.003 0.02
ED
45
ACCEPTED MANUSCRIPT
T
Adjusted-R² 0.297 0.207
Sample Size 336 646
P
Notes: ***, **, and * represent a significance at 1, 5, and 10 percent level, respectively.
RI
SC
NU
MA
ED
PT
CE
AC
46
ACCEPTED MANUSCRIPT
TABLE 6
This table reports the results of association between Wedge and RPT: Operating vs. Non-operating RPTs.
Panel A represents the results using RPT_O and RPT_NO as the dependent variable in the column (1) and (2),
respectively. Panel B shows the results using TOP 5 interaction term, and Panel C shows the each result of
Top 5 and Non-Top 5 sub-sample.
T
The dependent variable of the regression equation is RPT_O and RPT_NO. (1) RPT_O = [RPTs (operating
sales and purchases)]/Market value of equity (2) RPT_NO = [RPTs (non-operating transactions)]/Market
P
value of equity. The independent variable include: (1) DVC = Ownership wedge 1 of the controlling
RI
shareholders (Voting right - Cash flow right), (2) TOP5= 1 if a firm belongs to one of five largest chaebols,
and 0 otherwise, (3) SIZE = Natural logarithm of total sales in thousands of Korean won, (4)LEV = Total
liabilities / Total equity, (5)ROA = Net income / Total assets; (6)AVOID =If 0< ROA<2 %, then AVOID is
SC
“1.” Else AVOID is “0”, (7) MB = Market value of equity / Book value of equity, (8) RD = R&D Expenses /
Total sales, (9) CGI = Corporate governance measurement using outside director(s), independence of audit
committee, and audit firms, (10) IND = Industry indicators, (11) YEAR = Year indicators.
NU
Panel A: PRT_O vs. RPT_NO
RPT_O(or RPT_NO)t = β0 + β1DVCt + β2 SIZEt + β3LEVt + β4ROAt + β5AVOIDt
+ β6MBt+ β7RDt + β8CGIt + Σ IND t + Σ YEAR t + ε t
Exp.
MA (1) RPT_O (2) RPT_NO
Variables
sign Coefficients t-value Coefficients t-value
Intercept +/- -1.455 -2.74*** -2.835 -5.37***
DVC + 0.447 2.25** 0.783 3.96***
***
SIZE + 0.073 2.66 0.128 4.68***
ED
***
LEV + 0.076 2.87 0.070 2.63***
ROA - 0.248 0.48 0.070 0.14
AVOID + -0.009 -0.10 0.291 3.30***
PT
***
MB +/- -0.100 -3.15 -0.094 -2.97***
RD + 1.325 1.02 1.935 1.50
CGI - -0.518 -4.00*** -0.068 -0.52
CE
47
ACCEPTED MANUSCRIPT
T
RPT_O(or RPT_NO)t = β0 + β1DVCt + β3SIZEt + β4LEVt + β5ROAt + β6AVOIDt + β7MBt
+ β8RDt + β9CGIt + Σ INDt + Σ YEARt + εt
P
Top 5 Non-Top 5
Exp.
RI
(1) (2) (3) (4)
variables sign RPT_O RPT_NO RPT_O RPT_NO
Coef. t-value Coef. t-value Coef. t-value Coef. t-value
SC
** ***
Intercept +/- -2.344 -2.00 -5.056 -4.24 -1.035 -1.51 -0.010 -0.02
DVC + 1.027 2.14** 0.826 1.70* 0.247 1.06 0.556 2.55**
** ***
SIZE + 0.139 2.42 0.202 3.47 0.047 1.32 -0.003 -0.08
NU
** ** *
LEV + 0.137 2.30 0.147 2.43 0.058 1.90 0.047 1.68*
ROA - 0.026 0.02 0.032 0.03 0.327 0.55 -0.386 -0.69
AVOID + 0.483 2.56** 0.155
MA 0.81 -0.144 -1.46 0.320 3.48***
***
MB +/- -0.152 -2.76 -0.082 -1.45 -0.061 -1.42 -0.052 -1.30
RD + -0.172 -0.07 2.399 0.92 2.942 1.77* -0.533 -0.34
*** ***
CGI - -1.043 -3.89 0.447 1.64 -0.430 -2.86 -0.014 -0.10
Industry
Included included included included
Dummies
ED
Year
Included included included included
Dummies
*** *** ***
F-value 4.07 6.32 3.93 3.84***
PT
48
ACCEPTED MANUSCRIPT
TABLE 7
This table reports the regression results on Valuation Effect of RPTs and the Control-Ownership Wedge.
Panel A represents the results using interaction term of RPT and DVCD, and Panel B shows the results of sub-
sample divided by DVC. The dependent variable of the regression equation is Tobin Q, which equal to (Market
value of equity t+1 + Book value of debt t+1) / Total assets t+1. The independent variable include: (1) RPT = [RPTs
T
(operating sales and purchases + non-operating transactions)]/Market value of equity, (2) DVCD = One if
DVC is greater than or equal to the sample median and zero otherwise. (3) ASSET = Natural logarithm of total
P
asset in thousands of Korean won, (4) LEV = Total liabilities / Total equity, (5) ROA = Net income / Total
assets; (6) GROW = (Sales t – Sales t-1) / Sales t-1, (7) RD = R&D Expenses / Total sales, (8) AGE = Natural
RI
logarithm of (Current year – Establishment year), (9) CGI = Corporate governance measurement using outside
director(s), independence of audit committee, and audit firms, (10) IND = Industry indicators, (11) YEAR =
Year indicators.
SC
Panel A. DVC Interaction
Tobin Qt+1 = β0 + β1RPTt + β2RPT×DVCDt + β3ASSETt + β4LEVt + β5 ROA t + β6GROWt + β7RDt
+ β8AGEt + β9CGIt + Σ IND t + Σ YEAR t + ε t
NU
Exp. Dependent = Tobin Q
Variables Sign (1) (2)
Coefficients t-value Coefficients t-value
Intercept +/- 2.156 5.50*** 2.259 5.67***
MA
RPT - -0.046 -3.99*** -0.021 -1.25
RPT*DVCD - -0.041 -1.85*
DVCD - 0.000 -0.01
ASSET - -0.036 -2.40** -0.040 -2.60***
***
4.37***
ED
***
AGE - -0.063 -2.81 -0.068 -3.01***
**
CGI + 0.165 2.40 0.168 2.45**
Industry Dummies Included Included
Year Dummies Included Included
CE
49
ACCEPTED MANUSCRIPT
TABLE 8
This table reports the regression results on Association between RPT and Tobin Q by the Top 5 Chaebols.
Panel A represents the results using interaction term of RPT and TOP5, and Panel B shows the results of sub-
sample divided by DVC. Panel C reports the results using RPT_O and RPT_NO as dependent variables
divided by DVC and TOP5. The dependent variable of the regression equation is Tobin Q, which equal to
T
(Market value of equity t+1 + Book value of debt t+1) / Total assets t+1. The independent variable include: (1) RPT =
[RPTs (operating sales and purchases + non-operating transactions)]/Market value of equity, (2) RPT_O =
P
[RPTs (operating sales and purchases)]/Market value of equity (3) RPT_NO = [RPTs (non-operating
RI
transactions)]/Market value of equity (4) TOP5 = 1 if a firm belongs to one of five largest chaebols, and 0
otherwise, (5) ASSET = Natural logarithm of total asset in thousands of Korean won, (4) LEV = Total
liabilities / Total equity, (6) ROA = Net income / Total assets; (7) GROW = (Sales t – Sales t-1) / Sales t-1, (8) RD
SC
= R&D Expenses / Total sales, (9) AGE = Natural logarithm of (Current year – Establishment year), (10) CGI =
Corporate governance measurement using outside director(s), independence of audit committee, and audit
firms, (11) IND = Industry indicators, (12) YEAR = Year indicators.
NU
Panel A. Top 5 Chaebols as an Interaction
Tobin Qt+1 = β0 + β1RPTt + β2RPT×TOP5t + β3ASSETt + β4LEVt + β5 ROA t + β6GROWt + β7RDt
+ β8AGEt + β9CGIt + Σ IND t + Σ YEAR t + ε t
MA Dependent = Tobin Q
Exp. Model (1) Model (2) Model(3)
Variables sign
Total Top5 Non-Top5
Coef. t-value Coef. t-value Coef. t-value
Intercept +/- 2.733 6.97*** 7.434 9.78*** 1.043 2.51**
ED
*** **
RPT - -0.042 -2.73 -0.045 -2.47 -0.029 -2.11**
*
RPT×TOP5 - -0.037 -1.70
TOP5 +/- 0.310 7.42***
PT
50
ACCEPTED MANUSCRIPT
T
Yr. Dummies Included Included Included Included
5.54*** 14.84*** 4.30*** 5.88***
P
F-value
Adjusted-R² 0.364 0.635 0.210 0.283
RI
Sample Size 168 168 323 323
SC
Top5 Non-Top5
Exp. Model (1) Model (2) Model (3) Model (4)
variables sign DVC >= Median DVC < Median DVC >= Median DVC < Median
NU
Coef. t-value Coef. t-value Coef. t-value Coef. t-value
Intercept +/- 6.243 5.82*** 9.122 8.05*** 1.995 3.32*** 0.465 0.70
RPT_O - -0.081 -2.11** 0.017 0.47 -0.030 -1.38 -0.019 -0.55
RPT_NO - -0.091 -2.30** -0.028 -0.68 -0.056 -2.26** -0.059 -1.30
MA
ASSET - -0.180 -4.50*** -0.253 -5.94*** -0.038 -1.60 0.008 0.31
LEV + 0.108 3.42*** 0.090 1.74* 0.050 2.29** 0.043 2.60***
***
ROA + 1.863 2.97 6.076 8.44*** 0.964 2.72*** 1.191 3.16***
GROW + -0.203 -1.06 -0.126 -0.95 0.067 0.79 -0.005 -0.06
ED
***
RD + 10.779 3.87 1.430 0.94 0.644 0.56 -0.046 -0.05
AGE - -0.058 -0.84 -0.189 -3.67*** -0.058 -1.60 0.018 0.53
CGI + 0.448 2.33** -0.067 -0.36 0.237 2.29** 0.147 1.45
PT
51
ACCEPTED MANUSCRIPT
T
Pacific-Basin Finance Journal
P
RI
We thank you very much for your insightful and constructive comments. We greatly value
the opportunity to revise our paper. The paper has benefited from your insightful
SC
suggestions and we are grateful for that. We have done our best to address your comments
and hope that we have responded in a satisfactory fashion. Major changes made in this
revision are listed below.
NU
First, we removed footnote 1 and a sentence as suggested. Second, we corrected errors in
writing AffiliateCFij, unit of measurement in a footnote, and the expression on
MA
simultaneous equations. Third, we clarified the top 5 chaebol list in Table 1 by changing
Hyundai to Hyundai Motors. Fourth, following the editorial policy, we included detailed
self-contained table legends in all of the tables.
ED
PT
CE
AC
52