The Association between Related-Party Transactions and Control-Ownership

Wedge: Evidence from Korea

Minjung Kang, Ho-Young Lee, Myung-Gun Lee, Jong Chool Park

PII: S0927-538X(14)00053-5
DOI: doi: 10.1016/j.pacfin.2014.04.006
Reference: PACFIN 698

To appear in: Pacific-Basin Finance Journal

Received date: 29 January 2013

Revised date: 9 April 2014
Accepted date: 26 April 2014

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

Dear Prof. Jun-Koo Kang, Editor of PBFJ

Attached is the paper submitted to PBFJ.

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

Please, let me know if we can provide any further assistance or clarification.

MA
Sincerely,

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.

Keywords: related-party transactions, control-ownership wedge, control rights, cash flow

rights, chaebol firms

T
JEL classification: G30, M40

P
RI
1. Introduction

Related-party transactions (RPTs, hereafter) are defined as transactions with related

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

incomplete information achieving shareholder value maximization by reducing transaction

AC

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

of the agency problem in relation between RPTs and firm value.

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

control-ownership wedge as a proxy of firm-level agency problem, we first predict a

PT

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

for the top 5 chaebols than for non-top 5 chaebols.

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

in other countries because horizontal diversification across different industries is more

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

Using 982 firm-year observations of large business groups in Korea, we first

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

materials/merchandise purchases) and non-operating RPTs (e.g., sales and purchases of

ED

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

themselves could either be good (efficient transaction hypothesis) or bad (conflict of

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

and relevant information about these transactions. We exploit strict disclosure

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,

and Section 5 concludes this study.

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

2. Related literature and hypotheses development

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

controlling shareholders to pursue their own interests at the expense of minority

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

personal gains (Johnson et al., 2000).

AC

As the control-ownership wedge widens, expropriation of firms’ resources by

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

those where they have high cash flow rights.

Extant literature documents a negative association between the control-ownership

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

another firm where they have large cash flow rights.

2.2. Related party transactions (RPTs)

RPTs are defined as transactions with related entities such as shareholders,

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

interest of controlling shareholders in order to expropriate wealth from minority

AC

shareholders. (e.g., Johnson et al., 2000).

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

agency conflicts between controlling shareholders and minority shareholders, especially in

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

Many Korean firms belong to business groups known as chaebol, a collection of

ED

both public and private companies in a pyramidal and circular ownership structure, that are

typically controlled by members of a founding family. This ownership structure allows a

PT

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

and cross-shareholdings increase the control-ownership wedge of Korean firms leading to

AC

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

Korea may exacerbate the expropriation of minority investors by controlling shareholders

through various self-dealing RPTs. In addition, it is difficult for outside investors to

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

minority shareholders. Based on the above discussion, Hypothesis 1 written in the

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)

suggest that management entrenchment and expropriation of minority shareholders is most

apparent for the top 5 chaebol firms.

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

more pronounced for the top 5 chaebols.

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.

RPTs can be used as a means of tunneling in either operating (e.g., products/services

ED

sales and materials/merchandise purchases) or non-operating transactions (e.g., sales and

PT

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

operating transactions tend to be infrequent and large in amount. Relative to non-operating

AC

RPTs regulators can monitor the operating RPTs without much difficulty due to the

recurring nature of transactions and comparability of these transactions with those of

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

to identify unfair transactions that involve non-operating transactions such as

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

using transactions primarily associated with operating activities as a means of tunneling.

To provide empirical evidence on the effect of the different types of related-party

PT

transactions on the association between the control-ownership wedge and related party
CE

transactions, we test the following hypothesis in the null form:

Hypothesis 1b: The positive association between the control-ownership wedge

AC

and RPTs is not different between operating and non-operating

RPTs.

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.

Hypothesis 1d: The positive association between the control-ownership wedge

NU
and RPTs pertaining to non-top 5 chaebols is more pronounced in
non-operating RPTs. MA
3. Research design and sample selection

3.1. Measurement of control-ownership wedge and related-party transactions

ED

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

chaebol with N affiliated firms we define DirectCFi as a direct ownership portion of

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

affiliated firm j, respectively. Thus, CFj×AffiliateCFji captures an indirect ownership

portion over firm i. Considering all direct and indirect cash flow rights, the total cash flow

15
 ACCEPTED MANUSCRIPT

right of a controlling shareholder over an affiliated firm i is obtained by solving the

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

literature measures the control-ownership wedge (i.e., control-ownership disparity) either as

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

our study defined as follows.

Wedge (DVC1) t = Voting rights t - Cash flow rights t

Wedge (DVC2) t = (1+Voting rights t) / (1+Cash flow rights t)10

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

by the market value of equity.

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

3.2. Research model

AC

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.

RPTt = β0 + β1DVC t (or VR t, CR t) + β2 SIZE t + β3LEVt + β4ROAt + β5AVOIDt

+ β6MBt + β7RDt + β8CGIt + Σ YEAR t + Σ INDt + εt (1)

RPTt = β0 + β1DVCt + β2TOP5 t + β3 DVC×TOP5t + β4 SIZE t + β5LEVt

+ β6ROAt + β7AVOIDt + β8MBt + β9RDt + β10CGIt
+ Σ YEARt + Σ INDt + εt (2)

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

RPT_O (or RPT_NO)t = β0 + β1DVCt + β2 SIZE t + β3LEVt

+ β4ROAt + β5AVOIDt + β6MBt + β7RDt + β8CGIt
+ Σ YEARt + Σ INDt + εt (3)

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

TOP5 = 1 if a firm belongs to one of five largest chaebols, and 0 otherwise

SIZE = Natural logarithm of total sales in thousands of Korean won;
LEV = Total liabilities / total equity;
PT

ROA = Net income / total assets;

AVOID = 1 if 0< ROA<2%, and 0 otherwise;
MB = Market value of equity / Book value of equity;
CE

RD = Research and development cost / Total sales;

CGI = Corporate governance index (GCI) computed as the sum of proportion of outside
directors, an indicator of independent audit committee (i.e., 1 if all audit committee
members are independent and 0 otherwise), and an indicator of Big 4 external auditors
AC

(i.e., 1 if Big 4 and 0 otherwise) divided by 3;12

YEAR = Year indicators; and
IND = Industry indicators.13

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

using them as a dependent variable. If controlling shareholders do not differentiate RPTs in

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

transactions to operating transactions, either only RPT_NO would be significant or β1 for

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

non-operating transactions in association with the ownership structure. In order to test

Hypotheses 1c and 1d, we separate RPT_O and RPT_NO for the top 5 and non-top 5

chaebols, respectively. This separation allows us to see a differential effect of DVC on

RPT_O and RPT_NO between the top 5 and non-top 5 chaebols.

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

effectiveness of corporate governance. CGI is measured as the mean value of three

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.

3.3. Sample selection

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

4. Empirical analyses and results

4.1. Descriptive statistics

CE

Table 2 presents descriptive statistics of variables used in this study. 15 The

AC

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

wedge (DVC) significantly greater than the wedge of non-top 5 firms.

********************
Insert Table 2 here
********************

Table 3 presents Pearson correlations among study variables. The correlation

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

the multicollinearity is not severe.18

ED

********************
Insert Table 3 here
********************
PT

4.2. The results of empirical analyses

CE

4.2.1. The association between the control-ownership wedge and RPTs

AC

Table 4, Column (1) and (2) show the results of testing whether there is a positive

association between the control-ownership wedge and RPTs supporting Hypothesis 1.

RPTs in our study include operating sales, purchases, and non-operating transaction with

related parties. Control-ownership disparity is measured in terms of control-ownership

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

they are less likely to be used.

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

for controlling shareholders to expropriate firm resource to maximize their personal

ED

benefits.

********************
PT

Insert Table 4 here

********************
CE

4.2.2. Top 5 vs. Non-Top 5

We examine whether the influence of the ownership structure on RPTs is different

AC

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

interaction term, DVC×TOP5, as presented in equation 2 to capture the incremental effect

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

the top 5 chaebols supporting Hypothesis 1a.

********************
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

4.2.3. Operating vs. Non-operating RPTs

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

significant (multivariate test: p-value=0.28),23which is not able to reject Hypothesis 1b.

This result suggests that on average firms use both operating and non-operating RPTs
AC

equally as a means of tunneling.

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

Insert Table 6 here

********************
PT

4.3. Further analyses

4.3.1. Effect of RPTs on firm value

CE

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

Korean chaebols destroy firm value.

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

Tobin Qt+1 = β0 + β1RPTt + β2ASSETt + β3LEVt + β4 ROAt + β5GROWt + β6RDt

+ β7AGEt + β8CGIt + Σ INDt + Σ YEARt + εt (5)

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

consistent with the conflict of interest hypothesis.

29
 ACCEPTED MANUSCRIPT

4.3.2. RPTs, firm value, and the control-ownership wedge

We next test whether value destruction by RPTs is more severe when a firm has

higher degree of control-ownership wedge by employing an interaction term RPT×DVCD

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.

4.3.3. RPTs, firm value, Top 5 vs. Non-Top 5

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

firms, while only non-operating RPTs destroy non-top 5 chaebol firms.

********************
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

of ownership wedge increases by showing that the control-ownership wedge negatively

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

to companies in other East Asian countries in which ownership is highly concentrated.

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

Chang, S., and J. Hong. 2000. Economic performance of group-affiliated companies in

Korea: intergroup resource sharing and internal business transactions. Academy of
Management Journal 43(3), 429-448.
PT

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

the economic crisis. Journal of Financial Economics 68: 287-322.

Johnson, S., R. La Porta, F. Lopes-de-Silanes, and A. Shleifer. 2000. Tunneling. American
Economic Review 90(2), 22-27.
AC

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

Korea Fair Trade Commission (KFTC). 2012. Regulation on Resolution of Board of

Directors and Disclosures on Large-scale Intra-group Transactions. Available at:
http://www.ftc.go.kr/laws/laws/laws.jsp?lawDivCd=01.
Korea Fair Trade Commission (KFTC). 2013. Monopoly Regulations and Fair Trade Law.
Available at: http:// www.ftc.go.kr/laws/laws/laws.jsp?lawDivCd=01.

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

and corporate valuation. The Journal of Finance 57, 1147–1170.

Lemmon, M. L., and K. V. Lins. 2003. Ownership structure, corporate governance, and
firm value: evidence from the East Asian financial crisis. The Journal of Finance
PT

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

Unfair Act and Calculation.

Ministry of Strategy and Finance. 2009. Income Tax Act, Article 41, Unfair Act and
Calculation.
Ministry of Strategy and Finance. 2013. Inheritance Tax and Gift Tax Act, Article 4,
Obligation to Pay Gift Tax.
Modigliani, F. and M. B. Miller. 1963. Corporate income taxes and the cost of capital: a
correction. American Economic Review 53(3), 433-443.
Monthly Shin-DongA. Dec 1, 2007. “An external director’s perspective on a difficulty in
Samsung,” p. 118.
Morck, R., A. Shleifer, and R. Vishny. 1988. Management and ownership and market
valuation: an empirical analysis. Journal of Financial Economics 20, 293-315.
Morrison, D. F. 1976. Multivariate Statistical Methods, Second Edition. New York:
McGraw-Hill.
Myers, S. C. 1977. Determinants of corporate borrowing. Journal of Financial Economics 5,
147-175.

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

Regulations on Large Business Groups and Related Party Transactions

(1) Regulations on Large Business Group

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

substantially interfere with fair deals. [KFTC, Guidelines for Reviewing

Unfair Assistance, 2011].
(3) Chaebol firms must disclose, quarterly, the general status of business
CE

groups, shares, transactions with related parties, etc. [Fair Trade Act
Article 11, amended 2007 and Decree 11 Article 17, amended 2009]
AC

(2) Regulations on Related Party Transactions

Classification Content
Financial The K-GAAP requires firms to disclose related parties and related party
Reporting transactions in footnotes. Specifically, regardless of the presence or
Standards absence of transactions, a parent company must disclose its special
relationship and transactions with its subsidiaries. In addition, if the
transaction is with a related party within the accounting period, the
company needs to disclose the transactions and potential impact of
transactions on financial transactions, including contract bonds, receivable
and payable balances, etc. Information about the nature of the special
relationship should also be disclosed. [KFASB, Korean Generally
Accepted Accounting Principles (K-GAAP) No. 20, Related Party
Transactions Disclosure, 2004]
Auditing The auditor should carefully review data provided by management in
Standards regards to related parties and related party transactions. If the auditor

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.

Panel A. Sample Selection

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

2008 113 Samsung, Hyundai Motors, SK, LG, Lotte

2009 130 Samsung, Hyundai Motors, SK, LG, Posco
2010 143 Samsung, Hyundai Motors, SK, LG, Lotte
PT

Total 982

Panel C. Sample Distribution by Industry

CE

Industry N
Farming, Fishery, Mining, Forestry 2
Food, Beverage, Tobacco 54
AC

Fabrics, Clothes, Shoes, Bags 8

Woods, Pulps, Paper, Publishing 11
Chemicals, Lubber, Plastics, Oil Refinery 137
Nonmetallic Mineral, Base Metal 23
Assembled Metal, Primary Metal 90
Machine, Equipment, Computer, Transportation Equipment, Auto Manufacturing 208
Electricity, Gas, Water 32
Construction 88
Whole Sale, Retail 110
Hotel, Transportation, Real Estate, Education, Entertainment, Other Services 219
Total 982

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

DVC2 1.190 0.150 1.187 1.000 1.817

VR 0.411 0.182 0.403 0.000 1.000
CR 0.197 0.172 0.142 0.000 0.859
PT

SIZE 20.955 1.554 20.962 16.198 23.728

LEV 1.360 1.377 1.047 0.061 10.931
ROA 0.055 0.071 0.054 -0.367 0.317
CE

AVOID 0.154 0.361 0.000 0.000 1.000

MB 1.429 1.158 1.091 0.138 6.413
RD 0.016 0.028 0.005 0.000 0.159
AC

CGI 0.659 0.331 0.670 0.000 1.000

Panel B
Top 5 Non-Top 5
T-test
Variables (N=336) (N=646)
Mean Std. Mean Std. t-value
RPT 1.248 1.632 0.933 1.299 3.07 ***
RPT_O 0.418 1.080 0.410 0.966 0.12
RPT_NO 0.757 1.094 0.495 0.844 3.83 ***
DVC1 0.260 0.137 0.191 0.170 6.92 ***
DVC2 1.243 0.131 1.162 0.152 8.69 ***
VR 0.352 0.146 0.442 0.192 -8.17 ***
CR 0.092 0.104 0.251 0.175 -17.86 ***
SIZE 21.642 1.439 20.598 1.492 10.53 ***
LEV 1.108 1.106 1.491 1.483 -4.56 ***

41
 ACCEPTED MANUSCRIPT

ROA 0.068 0.062 0.048 0.074 4.52 ***

AVOID 0.107 0.310 0.178 0.383 -3.13 ***
MB 1.700 1.299 1.288 1.051 5.02 ***
RD 0.022 0.033 0.013 0.024 4.68 ***

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

LEV 1.000 -0.341 0.129 0.159 -0.067 0.106

(0.000) (0.000) (0.000) (0.035) (0.001)
ROA 1.000 -0.276 0.255 -0.065 0.079
(0.000) (0.000) (0.043) (0.013)
AVOID 1.000 -0.161 0.042 -0.004
(0.000) (0.187) (0.902)
MB 1.000 0.116 0.167
(0.000) (0.000)
RD 1.000 0.233
(0.000)
CGI 1.000

43
 ACCEPTED MANUSCRIPT

TABLE 4. Association Between Wedge and RPT

This table reports the results of association between Wedge and RPT. 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)DVC1 = Ownership wedge 1 of
the controlling shareholders (Voting right - Cash flow right), (2)DVC2 = Ownership wedge 2 of the

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.

RPTt = β0 + β1DVC(or VR, CR)t + β2 SIZE t + β3LEVt + β4ROAt + β5AVOIDt

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

LEV + 0.171 4.79*** 0.172 4.81 *** 0.173 4.82***

ROA - 0.581 0.83 0.580 0.83 0.626 0.89
AVOID + 0.340 2.84*** 0.345 2.89 *** 0.343 2.87***
CE

MB +/- -0.192 -4.49*** -0.194 -4.54 *** -0.198 -4.58***

RD + 3.714 2.13** 3.669 2.11 ** 3.541 2.02**
CGI - -0.682 -3.90*** -0.683 -3.91 *** -0.676 -3.86***
AC

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.

Panel A. Top 5 Interaction

RPTt = β0 + β1DVCt + β2TOP5t + β3DVC×TOP5t + β4 SIZE t + β5LEVt + β6ROAt

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

DVC×TOP5 + 1.067 1.75*

*** ***
SIZE + 0.153 3.90 0.180 4.54 0.183 4.61***
***
LEV + 0.202 5.55 0.189 5.20*** 0.187 5.15***
PT

ROA - 0.741 1.06 0.651 0.93 0.711 1.02

AVOID + 0.351 2.93*** 0.348 2.92*** 0.354 2.97***
*** ***
MB +/- -0.222 -5.12 -0.210 -4.85 -0.208 -4.82***
CE

RD + 1.572 0.88 2.634 1.47 2.942 1.64

*** ***
CGI - -0.740 -4.24 -0.666 -3.81 -0.675 -3.87***
Ind.Dummies Included Included Included
AC

Yr.Dummies Included Included Included

*** ***
F-value 10.55 10.75 10.50***
Adjusted-R² 0.202 0.212 0.213
Sample Size 982 982 982
Panel B. Top 5 vs. Non-Top 5
RPTt = β0 + β1DVCt + β2 SIZE t + β3LEVt + β4ROAt + β5AVOIDt
+ β6MBt+ β7RDt + β8CGIt + Σ IND t + Σ YEAR t + ε t
Dependent = RPT
Exp.
Variables (1) TOP 5 (2) Non-TOP 5
Sign
Coefficients t-value Coefficients t-value
***
Intercept +/- -7.163 -4.36 0.023 0.03
DVC + 2.078 3.09*** 0.742 2.49**
***
SIZE + 0.382 4.76 0.042 0.92
LEV + 0.341 4.03*** 0.116 3.01***
ROA - -0.158 -0.13 0.009 0.01
AVOID + 0.673 2.55** 0.231 1.84*
***
MB +/- -0.244 -3.10 -0.096 -1.75*

45
 ACCEPTED MANUSCRIPT

RD + 3.953 1.10 2.779 1.31

CGI - -0.737 -1.95* -0.487 -2.54**
Industry Dummies Included Included
Year Dummies Included Included
F-value 7.16*** 7.48***

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

Industry Dummies Included Included

Year Dummies Included Included
F-value 5.67*** 6.02***
Adjusted-R² 0.110 0.117
AC

Sample Size 982 982

Panel B: Top 5 Interaction
RPT_O(or RPT_NO)t = β0 + β1DVCt + β2TOP5 t + β3DVC*TOP5t + β4SIZEt + β5LEVt
+ β6ROAt + β7AVOIDt + β8MBt+ β9RDt + β10CGIt + Σ INDt + Σ YEARt + εt
Exp. (1) RPT_O (2) RPT_NO
Variables
sign Coefficients t-value Coefficients t-value
Intercept +/- -1.379 -2.42** -2.259 -3.99***
DVC + 0.167 0.71 0.654 2.79***
*
TOP5 + -0.237 -1.74 0.273 2.01**
**
DVC*TOP5 + 1.069 2.36 -0.238 -0.53
SIZE + 0.072 2.45** 0.098 3.35***
***
LEV + 0.076 2.82 0.084 3.13***
ROA - 0.316 0.61 0.115 0.22
AVOID + -0.002 -0.02 0.296 3.37***
***
MB +/- -0.101 -3.13 -0.108 -3.38***
RD + 1.524 1.14 1.001 0.75
CGI - -0.526 -4.06*** -0.052 -0.40
Industry Dummies Included Included

47
 ACCEPTED MANUSCRIPT

Year Dummies Included Included

F-value 5.49*** 5.90***
Adjusted-R² 0.114 0.123
Sample Size 982 982
Panel C. Top 5 vs. Non-Top 5

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

Adjusted-R² 0.1739 0.2675 0.1057 0.1026

Sample Size 336 336 646 646
Notes: ***, **, and * represent a significance at the 1, 5, and 10 percent level, respectively.
CE
AC

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

LEV + 0.055 4.28 0.056

***
ROA + 2.098 8.53 2.147 8.71***
GROW + -0.012 -0.21 -0.014 -0.26
RD + 2.061 3.22*** 2.060 3.21***
PT

***
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

F-value 12.78*** 12.09***

Adjusted-R² 0.238 0.240
Sample Size 982 982
AC

Panel B. Divided by DVC

Exp. Dependent = Tobin Q
Variables Sign (3) DVC >= Median (4) DVC < Median
Coefficients t-value Coefficients t-value
Intercept +/- 3.110 5.57*** 1.711 2.93***
RPT - -0.066 -4.08*** -0.022 -1.21
ASSET - -0.073 -3.36*** -0.024 -1.08
LEV + 0.064 3.32*** 0.053 2.98***
***
ROA + 2.250 6.33 2.099 5.93***
GROW + -0.057 -0.58 0.035 0.53
RD + 3.729 3.45*** 1.422 1.72*
**
AGE - -0.082 -2.04 -0.012 -0.38
CGI + 0.328 3.01*** 0.039 0.40
Industry Dummies Included Included
Year Dummies Included Included
F-value 7.74*** 6.50***
Adjusted-R² 0.263 0.226
Sample Size 491 491
Notes: ***, **, and * represent a significance at the 1, 5, and 10 percent level, respectively.

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

ASSET - -0.062 -4.06*** -0.198 -6.86*** -0.008 -0.53

***
LEV + 0.067 5.38 0.076 2.85*** 0.045 3.57***
*** ***
ROA + 2.019 8.45 3.999 8.76 1.090 4.32***
CE

GROW + -0.003 -0.05 -0.083 -0.72 0.018 0.34

RD + 1.249 1.99** 3.281 2.83*** 0.185 0.26
AGE - -0.057 -2.63*** -0.140 -3.36*** -0.021 -0.92
2.84*** 3.07***
AC

CGI + 0.189 0.098 0.72 0.210

Ind.Dummies Included Included Included
Yr.Dummies Included Included Included
F-value 14.85*** 13.87*** 9.26***
Adjusted-R² 0.283 0.469 0.249
Sample Size 982 336 646
Panel B. Top 5 vs. Non-Top 5
Tobin Qt+1 = β0 + β1RPTt + β2ASSETt + β3LEVt + β4 ROA t + β5GROWt + β6RDt + β7AGEt + β8CGIt
+ Σ IND t + Σ YEAR t + ε t
Top 5 Non-Top 5
Exp. Model (4) Model (5) Model (6) Model (7)
Variables sign DVC >= Median DVC < Median DVC >= Median DVC < Median
Coef. t-value Coef. t-value Coef. t-value Coef. t-value
Intercept +/- 6.321 5.92*** 9.114 8.06*** 1.970 3.28*** 0.371 0.57
RPT - -0.070 -2.71*** 0.002 0.10 -0.039 -2.22** -0.024 -0.92
ASSET - -0.184 -4.64*** -0.252 -5.93*** -0.037 -1.56 0.011 0.44
LEV + 0.110 3.49*** 0.089 1.74* 0.049 2.26** 0.043 2.64***
ROA + 1.937 3.10*** 5.970 8.42*** 0.981 2.78*** 1.216 3.23***

50
 ACCEPTED MANUSCRIPT

GROW + -0.218 -1.14 -0.105 -0.80 0.082 0.98 -0.009 -0.12

RD + 11.060 4.01*** 1.484 0.97 0.673 0.58 0.038 0.04
AGE - -0.048 -0.70 -0.187 -3.64*** -0.060 -1.65 0.017 0.50
CGI + 0.461 2.46** -0.104 -0.56 0.227 2.21** 0.142 1.40
Ind.Dummies Included Included Included Included

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

Panel C : RPT_O and RPT_NO divided by Top 5

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

Ind.Dummies included included included included

Yr. Dummies included included included included
F-value 5.28*** 14.19*** 4.18*** 5.69***
CE

Adjusted-R² 0.361 0.635 0.210 0.282

Sample Size 168 168 323 323
Notes: ***, **, and * represent a significance at the 1, 5, and 10 percent level, respectively.
AC

51
 ACCEPTED MANUSCRIPT

Highlights for the Reviewer

"The Association between Related-Party Transactions and Control-Ownership Wedge

: Evidence from Korea"

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