The Effect of Corporate Governance and Ownership Structure on Firm Value in Thailand

8/10/2019 The Effect of Corporate Governance and Ownership Structure on Firm Value in Thailand

1/22

Form versus substance: The effect of ownership structure

and corporate governance on firm value in Thailand

J. Thomas Connelly a, Piman Limpaphayom b,, Nandu J. Nagarajan c

a Department of Banking and Finance, Faculty of Commerce and Accountancy, Phyathai Road, Chulalongkorn University, Bangkok 10330, Thailandb Portland State University, Sasin Graduate Institute of Business Administration of Chulalongkorn University, USAcJoseph M. Katz Graduate School of Business and College of Business Administration, University of Pittsburgh, Mervis Hall, Pittsburgh, PA 15260, USA

a r t i c l e i n f o

Article history:

Received 11 March 2010

Accepted 25 January 2012

Available online 4 February 2012

JEL classification:

G32

G34

Keywords:

Corporate governance

Family ownership

Firm value

Thailand

a b s t r a c t

We examine the relation between the quality of corporate governance practices and firm value for Thai

firms, which often have complex ownership structures. We develop a comprehensive measure of corpo-

rate governance and show that, in contrast to conventional measures of corporate governance, our mea-

surement, on average, is positively associated with Tobins q. Furthermore, we find that q values are lower

for firms that exhibit deviations between cash flow rights and voting rights. We also find that the value

benefits of complying with good corporate governance practices are nullified in the presence of pyra-

midal ownership structures, raising doubts on the effectiveness of governance measures when ownership

structures are not transparent. We conclude that family control of firms through pyramidal ownership

structures can allow firms to seemingly comply with preferred governance practices but also use the con-

trol to their advantage.

2012 Elsevier B.V. All rights reserved.

1. Introduction

In July 1997, the Bank of Thailand discontinued the fixed ex-

change rate regime which had been in place for decades. Thailand,

an emerging nation, with a promising economic future, tumbled

into the worst financial crisis in its history. In the process, Thailand

also dragged many neighboring economies along with it into a re-

gion-wide economic downturn, the likes of which no country in

the region had ever experienced before. Following this crisis, and

its resolution, corporate governance has received considerable

attention from regulators and practitioners in all the AsiaPacificcountries. The main impetus for this heightened attention stems

fromevidence emerging from the financial crisis that the aggressive

financing practices and poor investment decisions, associated with

the financial downturn, were a result of poor corporate governance

practices among large public corporations and financial institutions

in the afflicted economies. Consequently, the central governments

of most AsiaPacific nations, along withinternational organizations

such as the OECD, implemented corporate governance reforms

throughout the region.1 The effectiveness of these governance re-

forms is an empirical question because business environments in this

region are heterogeneous and Asian companies have many unique

features, the most notable among them being the concentrated

ownership and direct and indirect control exercised by the firms

founding families.

In this paper, we provide empirical evidence on the relation be-

tween control, ownership structure and firm value for all industrial

companies that were publicly traded on the Thai stock exchange in

2005. We find that Thai family companies developed pyramidal

ownership structures after the financial crisis of 1997, probablyin response to reductions in their ownership holdings. In particu-

lar, we find that the governance measures mandated in Thailand,

and subsequently adopted by Thai family firms, are not as effective

in mitigating agency conflicts in this new opaque environment as

they have been shown to be in the US. We also develop a corporate

governance index that we show is a more effective measure of

0378-4266/$ - see front matter 2012 Elsevier B.V. All rights reserved.doi:10.1016/j.jbankfin.2012.01.017

Corresponding author. Tel.: +1 503 725 9991; fax: +1 503 725 5850.

E-mail addresses: [email protected] (J.T. Connelly), [email protected] (P.

Limpaphayom),[email protected](N.J. Nagarajan).

1 There is recent empirical evidence of an association between adoption of

internationally accepted corporate governance practices and firm valuation in these

AsiaPacific economies. See, for example Cheung et al. (2010) who examine large

Chinese firms,Black et al. (2006a, 2009)for example Korean firms, andCheung et al.

(2007, 2011)for Hong Kong firms.

Journal of Banking & Finance 36 (2012) 17221743

Contents lists available at SciVerse ScienceDirect

Journal of Banking & Finance

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j b f
http://dx.doi.org/10.1016/j.jbankfin.2012.01.017mailto:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.jbankfin.2012.01.017http://www.sciencedirect.com/science/journal/03784266http://www.elsevier.com/locate/jbfhttp://www.elsevier.com/locate/jbfhttp://www.sciencedirect.com/science/journal/03784266http://dx.doi.org/10.1016/j.jbankfin.2012.01.017mailto:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.jbankfin.2012.01.017


2/22

compliance with good governance practices in the Thai business

context than other, conventional measures of effective governance.

However, even this governance index appears to be associated

with value only when companies do not have a pyramidal owner-

ship structure, suggesting that Thai families are able to manipulate

governance measures when they have high voting control over

their firms.

Recent anecdotal and research evidence document that family

firms in the US outperform non-family firms.2 Founding families

have considerable wealth invested in their firms and, thus, have

an incentive to manage the firm in ways that improve value. On

the other hand, families and founders may use their greater control

over decision rights to expropriate wealth from minority share-

holders, resulting in a decrease in value.3 Family firms in Thailand,

however, operate in a very different environment from that in the

US. Recent empirical research (La Porta et al., 2000b, 2002) pro-

vides evidence that companies with controlling shareholders have

lower valuations in civil law countries, like Thailand, where minor-

ity shareholders are less well protected from expropriation by a

controlling shareholder, compared to valuations in common law

countries, like the US.Mitton (2002)argues that in an environment

where legal protection for outside shareholders may be insufficient,

the firms themselves can pre-commit to not expropriating wealth

from minority shareholders by implementing appropriate corporate

governance measures. For instance, such family firms could in-

crease the number of independent directors on the board. The fore-

going argument suggests that implementing appropriate

governance measures may be particularly important in East Asian

countries, where families commonly tend to own controlling inter-

ests in firms and legal protections for minority shareholder rights

are limited. The issues that emerge are twofold. First, what mea-

sures of governance are appropriate benchmarks in an environment

where implementation of good governance practices is hard to

measure and monitor? Second, is compliance with prescribed mea-

sures of governance effectiveness sufficient to mitigate agency

problems between majority and minority shareholders in these

East Asian family firms or do pyramidal ownership structures andextra-contractual arrangements between family members and

directors allow subversion and manipulation of these governance

measures, rendering them ineffective proxies for value creation?4

In other words, is the form of governance measures adopted by

these firms consistent with the substance of what such measures

are intended to accomplish?

Previous studies on governance arrangements in Thai family

firms have generally focused on the period before the Asian finan-

cial crisis in 1997 (Bertrand et al., 2008; Wiwattanakantang, 2001).

Before 1997, founding families held significant proportions of the

outstanding shares and had full control over their businesses.

Consequently, these families may have had no need to employ

complicated ownership structures, such as pyramids. In corrobora-

tion,Claessens et al. (2000) document that Thai companies rarely

employed a pyramidal ownership structure prior to 1997. How-

ever, recent changes in the characteristics of Thai family firms have

affected the nature and extent of corporate family ownership and

control. Specifically, many firms in Thailand experienced financial

difficulties and some went bankrupt following the Asian financial

crisis (Zhuang et al., 2000). As a result, some of these firms, includ-

ing family firms, had to raise additional capital and restructure

themselves after 1997, which, in turn, led to reductions in family

ownership.5 More importantly, many publicly traded family firms

have also implemented pyramidal ownership structures, probably

because they wished to counter the dilution in control arising from

reductions in their shareholdings. Based on these considerations,

we expect that family firms in Thailand are no longer the homoge-

neous group of firms with similar characteristics that they were

prior to 1997. In particular, the nature and extent of agency prob-

lems, and corresponding value consequences, can vary across family

firms. Because of these new institutional settings, we expect that our

post-1997 analysis of these family firms can add additional insights

and make a significant contribution to the literature.

In our analysis, we classify industrial companies that were

publicly traded on the Thai stock exchange in 2005 into four

groups, based on the extent of family ownership and the presence

of pyramidal ownership structures.6 Of the total sample of 216

firms, we find that firms with high family ownership are associated

with lower values of Tobins q. In particular, these high family own-

ership firms have an averageq value that is lower than the mean q

for low family ownership firms.7 This difference is not only statis-

tically significant, but also economically significant. Past research

conducted on US firms (e.g.,Yermack, 1996) has revealed a positive

relation between conventional governance variables, such as board

size and firm value. However, we find no such relation between

conventional governance variables andq for Thai firms. We conjec-

ture that this lack of association between broad governance mea-

sures and q for our sample firms arises because Thai family firms

only maintain the external trappings of good governance practices,such as having several directors labeled as independent, while

2 Anderson and Reeb (2003) document superior performance of family-owned

firms among S&P 500 firms, across seven years (19921999). The improved

performance is further increased when a family member holds the CEO position

rather than an outside manager. An article in Business Week (November 10, 2003, p.

100) reports that one-third of the S&P 500 firms have founding families involved in

management and these are the best performers. Villalonga and Amit (2006) and

Perez-Gonzalez (2006)find that founder-controlled firms are associated with higher

values which decline when these firms are managed by heirs.3 Demsetz and Lehn (1985)argue that founder or family control may be exercised

because of amenity potential, the non-pecuniary benefits that family members gain

from control. However, because this leads to entrenchment, family control may also

be associated with weaker management than that provided by professional managers.

Johnson et al. (1985) find that positive unexpected returns are associated with the

sudden demise of founder-CEOs and Morck et al. (1988)find that older firms with

founders present among the management have reduced q values.4 For instance, Hwang and Kim (2009) provide evidence that social ties between

managers and directors can weaken managerial pay-performance and turnover-performance sensitivity.

5 As part of restructuring efforts, many families were forced to sell significant

stakes in their firms. Others sought additional capital through strategic partnerships,

often with foreign companies. The shareholdings of two of the largest commercial

banks in Thailand were drastically restructured as the founding families gave up

significant portions of their ownership stakes. Non-financial companies were also

hard-hit. For example, at SHIN, a large stake in the nations largest telecommunication

company, owned by the former Prime Minister of Thailand, was sold to a consortium

of Singaporean investors. BIG C, a large family-owned discount retailer, forged a

business alliance with a foreign retailer by issuing shares to the new partner. NTS, a

family-owned steel maker, merged its steel business into a new company jointly

owned by another conglomerate. The founder of QH, a large real estate development

company, sold a stake to the Government of Singapore Investment Corporation.6 Bertrand et al. (2008)study a sample of public and private firms using 1996 data,

drawn from 93 Thai family groups. They show that family structure is important to

value creation in Thai companies. Firms with more male heirs end up with sons

having greater control and such firms are associated with lower levels of firm

performance, especially if the founder is dead. Families with more sons show a

greater gap between control and ownership rights. Our study differs in important

ways from theirs because we focus only on publicly traded firms drawn from a period

after the financial crisis of 1997 when family firms ownership structures were

drastically different. The study by Bertrand et al. (2008) consists of 528 firms, of

which only 94 are publicly traded companies. More importantly, our study is able to

assess the relative importance of control and management on agency costs while

examining the relation between market value and a quantified and detailed

governance index.7 As anecdotal evidence,Studwell (2007, p. 24)reports, Despite now bullish stock

markets in the region, the billionaires-with their lousy corporate governance and

manipulation of local banks to provide cheap and easy alternative sources of credit-

also have contributed to the worst long-term emerging-market-equity performance

in the world. From 1993-when the first significant international portfolio investments

came into Southeast Asian bourses-to the end of 2006, total dollar returns withdividends reinvested in Thailand and the Philippines were actually negative.

J.T. Connelly et al. / Journal of Banking & Finance 36 (2012) 17221743 1723


3/22

subverting the substance of such practices through the influence of

family ownership and pyramidal control arrangements.8 To test

this conjecture, we design an in-depth assessment of publicly re-

ported details of governance practices and disclosure, which we

term the Corporate Governance Index (CGI).9 In regression analyses

using all firms, we show that, unlike conventional governance vari-

ables, the CGI is significantly associated with q. Further, in regres-

sion analyses with subgroups, we show that the positiveassociation between CGI and q is driven by family firms without

pyramidal ownership structures. Our results remain robust after

controlling for past performance and the potentially endogenous

relationship between CGI and q through the use of a two-stage

least-squares regression approach.

Overall, our study makes several contributions to the literature.

We document that the nature of family ownership in Thai family

firms changed after the 1997 crisis, facilitating more opaque ways

for families to establish control. Furthermore, we construct and

use a detailed governance index which allows us to establish the

relation among stock ownership, governance and value for Thai

firms. We also provide evidence that, contrary to the findings for

US companies, conventional governance variables such as board

size and board independence are not associated with value for Thai

family firms. On the other hand, the Corporate Governance Index

(CGI) we construct, to assess the effectiveness of the implementa-

tion of governance practices, is significantly associated with value

for family firms without pyramidal ownership structures. That is,

our empirical results showthat (1) governance measures that effec-

tively capture the relation with value for Thai family firms have to

be more detailed than conventional measures such as board inde-

pendence and (2) the relation between these more detailed mea-

sures of good corporate governance and firm value is positive and

statistically significant only for family firms without a pyramidal

ownership structure, suggesting that such complex ownership

structures can subvert and render governance measures ineffective.

In support of this argument, we also find that the useof a pyramidal

structure negatively affects firm value after controlling for relevant

factors. Our results have implications for governance regulation inThailand, namely that the adoption of so called good corporate

governance practices alone is not a guarantee of firm performance.

We organize the remainder of the paper as follows. Section2

discusses the theory and institutional background for our study

of governance arrangements. The data and empirical methods are

presented in Section 3. Section 4 provides the empirical results

and Section5 concludes the paper.

2. Theory and institutional background

The Stock Exchange of Thailand (SET) can be characterized as a

retail-driven market, with domestic retail investors conducting thelions share of trading. Institutional investors make up a smaller,

though growing segment. Overall, Thai companies are mostly

owned by insiders while trading activities are mostly initiated by

small minority investors. Moreover, legal protection for investors

is relatively weak and ownership structures are highly concen-

trated. Taken together, the Thai market provides an interesting set-

ting in which to examine the relation among family ownership,

corporate governance and firm performance.

2.1. Family ownership and firm valuation

Financial economists have long recognized that ownership

structure has an important influence on value creation and firm

performance (Shleifer and Vishny, 1997). For example, Morcket al. (1988) find that, at low levels, managerial ownership is va-

lue-enhancing through the alignment of interests between insiders

and outside shareholders, while at high levels, managerial owner-

ship becomes value destroying through the entrenchment effect.

This type of agency conflict is referred to as Type I (between man-

agers and shareholders). For family firms, the agency conflicts be-

tween majority shareholders and outside/minority shareholders

(referred to as Type II agency conflicts) are more prevalent. Recent

empirical evidence for US firms shows that family firms perform

better than non-family firms (Anderson and Reeb, 2003). US family

firm performance is even stronger when founders manage the firm

(Villalonga and Amit, 2006; Perez-Gonzalez, 2006). Demsetz and

Lehn (1985) argue that, because of their concentrated and undiver-

sified holdings, families have more incentives to minimize agencyconflicts and maximize firm value. There is also evidence that fam-

ily ownership leads to long investment horizons and efficient

investment decisions (Stein, 1989; James, 1999).

However, the relation between family ownership and firm per-

formance can be different in a developing economy with relatively

weak legal rights and weak investor protection. This is because, in

such an environment and by virtue of their larger ownership

stakes, families have more opportunities and the power to take

actions that benefit themselves at the expense of minority share-

holders (Fama and Jensen, 1983). An examination of a sample of

companies in East Asia reveals that the majority of top managers

in these companies come from controlling families (Claessens

et al., 2000). Lins (2003)finds that, among East Asian firms, firm

value is lower when management control is excessive. Followingthis evidence, we expect that Type II agency conflicts for many Thai

companies could be quite severe.

2.2. Ownership structure and firm valuation

Recently, corporate finance theorists and corporate governance

researchers have turned their collective attention to the relation

between ownership structure and corporate governance in East

Asian firms. La Porta et al. (2006)argue that securities laws and

enforcement are critical to financial market development. In fact,

high ownership concentration by insiders is a response to the lack

of legal protection for shareholders (La Porta et al., 1998; hereafter

LLSV). Several prior studies have classified Thailand as a country

with weak legal enforcement and shareholder protection. For exam-ple, LLSV (1998) rate Thailand fairly low in terms of shareholder

8 According to Stock Exchange of Thailand regulations, Thai companies are required

to have at least five members on the board of directors, three of whom must be

independent directors. The CEO or top operating executive is permitted to hold the

board chairmanship (Sersansia and Nimmansomboon, 1996). However, Thai boards

often have directors among their members, who are connected, with an affiliation to a

related organization (Khantavit et al., 2004; Nam and Nam, 2004). For example, a

director may be an employee of a related company or a creditor financial institution.

These affiliations mean true director independence can be difficult to achieve.

Contributing to this mix is the fact that personal relationships, often spanning

decades, play a strong role in Thai culture. Directors are quite likely to have had

previous business, social, or even high school connections with the top executive.

Directors may also feel beholden to the executive who appointed them to the board

and thus may be hesitant about objecting openly to his policies, which would violate

cultural norms (Nam and Nam, 2004).9 The composition and construction of the Corporate Governance Index (CGI) are

described on pages 69. The full set of measurements is shown in Appendix. By

corporate governance activities in closely held firms (where Type II agency problems

exist, that is agency conflicts between majority shareholders and outside/minority

shareholders), we mean actions and disclosures by the firm that can affect the welfare

of minority shareholders. Such actions could include more transparency regarding

board activities, such as minutes of board meetings, public disclosure of compensa-

tion arrangements for directors, shareholders being allowed to ask questions during

annual general meetings, disclosure of crossholdings and pyramidal holdings, and

prevention of insider trading by controlling shareholders. Similarly governance

activities in widely held firms, in which Type I agency problems, that is agency

conflicts between managers and shareholders, are found, would be compensation

arrangements for managers that tie their interests to the shareholders, monitoring ofmanagerial actions by the directors, and other practices.

1724 J.T. Connelly et al. / Journal of Banking & Finance 36 (2012) 17221743


4/22

rights and the rule of law. The same authors (LLSV, 2000a) also

classify Thailand as a low-investor protection country in their

33-country survey of dividend policies around the world.Given that low protection for minority shareholders is associ-

ated with higher ownership concentrations, it is not surprising that

LLSV (2000b) and Claessens et al. (2000)find that East Asian firms

have highly concentrated ownership structures, which, in turn,

allows insiders to exercise effective control over their firms.

Although Thai companies have very high ownership concentra-

tions,Claessens et al. (2000) provide evidence that the ratio be-

tween cash flow rights and control rights for Thai firms is close

to one. However, the evidence inClaessens et al. (2000) describes

Thai ownership structures prior to the Asian financial crisis of

1997. In other words, Thai firms seldom used pyramidal structures

to enhance control before 1997 because family ownership levels

were high and there was no need to do so. Subsequent to the crisis,

many Thai family firms had to dilute their ownership stakes in or-der to raise additional outside investment. To illustrate this phe-

nomenon, we examine the transformation of a large real estate

company, Quality House, PLC (QH). QH was founded by Anan

Asawaphokin as a construction company in 1983. In 1990, QH be-

gan to engage in both residential and commercial real estate

development.

In 1996, QH would not have been considered to have a pyrami-

dal ownership structure (seeFig. 1). The founder, Anan Asawapho-

kin, owned 15.77% directly plus 6.99% through Land and House

PLC, another large real estate company controlled by the Asawaph-

okin family. Another individual from another well-known family,

Chai Srivikorn, owned 10.11% of the outstanding shares. Unaffili-

ated foreign financial institutions owned a total of 15.02%. Consis-

tent with Claessens et al. (2000), there is little evidence of anydeviation between cash flow rights and voting rights for QH and

many other Thai companies during this pre-1997 period.

In 1997, QH experienced financial and operating problems.

After the Asian financial crisis, the Government of Singapore

Investment Corporation became a major shareholder in 2001 with

a 20% holding. By 2005, the ownership structure of QH had become

very complicated. The direct shareholdings in QH by members of

the Asawaphokin family had declined to 7.17%, and the Govern-

ment of Singapore Investment Corporation held 13.23% of the out-

standing shares. However, Anant Asawaphokins shareholdings in

QH, via control of Land and House, increased from 6.99% in 1996

to 23.14%. Two other companies, private firms affiliated with the

Asawaphokin family, held a combined 8.86% ownership stake in

QH. The pyramidal structure, now apparent, pushed the ratio

between cash flow rights and voting rights lower than one. The

transition of the control structure of QH is graphically presented

inFig. 1.

Overall, it appears that in order to maintain control, familiesand founders employ more opaque (pyramidal) ownership struc-

tures that allow the owners to exert significant control even with

lower direct ownership holdings. One potential consequence of

excessively high control is that controlling shareholders are able

to expropriate wealth from minority shareholders. Following the

arguments above, we predict that, for our sample of Thai family

firms, the use of more control through a pyramidal ownership

structure should be detrimental to firm value. In other words,

the presence of deviations between control rights and cash flow

rights should be negatively associated with q.

2.3. Corporate governance and firm performance

Recently, corporate governance researchers have begun to usecomposite indexes to assess governance practices, recognizing that

corporate governance mechanismsmayserve as complementsor as

substitutes for one another. Four studies are notable in this regard.

Gompers et al. (2003), Bebchuk et al. (2009), and Brown and Caylor

(2006) create separate indexes for US firms while Drobetz et al.

(2003) do likewise for German companies. These authors demon-

strate the relation between good corporate governance and firm

value.10 There are also a number of studies examining the gover-

nance to performance link for firms in emerging markets. Generally,

authors of these studies also find that better-governed firms are

associated with better performance.11

Mitton (2002) suggests that in an environment where legal pro-

tection for outside shareholders may be insufficient, firms them-

selves can preclude expropriation of minority shareholdersresources through the use of appropriate corporate governance

measures. Therefore, adoption of appropriate or effective corporate

governance practices is particularly important among closely held

firms in East Asian countries. A further implication of this argument

1996 2005

Land and

House PLC

6.99%

Anan

Asawaphokin

15.77%

Financial

Institutions

15.02%

Chai

Srivikorn

10.11%

QH PLC

Land and

House PLC

23.14%

Asawaphokin

Family

7.17%

Private

Company

8.86%

Government

of Singapore

13.23%

QH PLC

Government

of Singapore

13.23%

Fig. 1. An illustration of the ownership structure changes experienced by publicly traded Thai Family Companies After 1997. The figure shows the transformation of the

ownership structure of QH, a large family-owned real estate company, between 1996 and 2005. The thin arrows indicate the direct shareholdings in QH by the various large

shareholders. The wide arrows show the indirect control by the family over other entities.

10 These types of governance indexes may not be relevant for Thai firms because

they are largely structured to assess the extent of takeover defenses. Hostile takeovers

are infrequent in Thailand and in emerging markets throughout the AsiaPacific

region.11 Using an internationally accepted benchmark (OECD, 2004),Cheung et al. (2010)

document a positive relation between market valuation and corporate governance

practices among large Chinese firms. Other studies on the relation between a portfolio

of governance measures or index and value include Black et al. (2006a, 2009) for

Korean firms; Doidge et al. (2007), Mitton (2004), Durnev and Kim (2005), and

Klapper and Love (2003)which use the index created by Credit Lyonnaise Securities

Asia (CLSA, 2002); Cheung et al. (2007, 2011)for Hong Kong firms; and Black et al.(2006b)for Russian firms.



5/22

is that if appropriate governance measures, such as internationally

accepted corporate governance practices, were to be adopted by

Thai family firms, they could potentially mitigate any adverse im-

pact of family ownership.

In addition to the changes in the ownership structures of Thai

family firms after 1997, the institutional environment in which

these firms operate has also changed drastically. In particular,

many corporate governance reforms have been implemented in

the post-crisis years through the concerted efforts of many organi-

zations (Limpaphayom and Connelly, 2004). For example, the Stock

Exchange of Thailand (SET) required all firms to have audit com-

mittees comprised entirely of independent directors by 1999. The

SET also established several other guidelines for improving the

quality of corporate governance practices and for conducting

shareholders meetings. In addition, new and updated rules from

the SET, new and revised laws, and increased regulatory oversight

have been at the forefront of the push for improved corporate gov-

ernance. For example, in 1999, the Thai government passed laws

requiring disclosure of related party transactions, asset divesti-

tures, and disclosure of accounting practices that deviated from

generally accepted principles. In 2002, the government passed a

new law which shortened the financial statement reporting time

from 60 to 45 days after the end of a fiscal year. In addition to reg-

ulatory reforms, other organizations like the Thai Institute of Direc-

tors Association (Thai IOD), the Thai Investors Association (TIA),

and professional associations for accountants, auditors, and inter-

nal auditors have pushed forward initiatives to improve the quality

of corporate governance practices. Consequently, Thai companies

operate in an environment that is very different from the time be-

fore the Asian financial crisis.

Although conventional wisdom would suggest that firms con-

forming to mandated governance requirements (e.g., board inde-

pendence) are less likely to experience impairments in value,

there are often extra-contractual relationships between the top

managers and directors that can complicate the situation (Romano,

2005). The number and extent of these relationships in Thai com-

panies can be considerable, stemming from long associations andpersonal or family contacts. These extra-contractual relationships

can result in boards, with ostensibly independent directors, being

unable to effectively monitor and control management. One impli-

cation of the foregoing observation is that traditional measures of

board effectiveness, such as board size or board independence,

while effective as measures of corporate governance in the US,

may be inadequate to capture the actual quality of corporate gov-

ernance practices among Thai family firms. Therefore, we predict

that these conventional proxies for effective corporate governance

should exhibit no significant relation with firm value for Thai firms.

In addition, we note that other governance measures commonly

used to evaluate the level of entrenchment of insiders, such as the

GIM index (Gompers et al., 2003) or the E-Index (Bebchuk et al.,

2009) are not applicable to Thai family firms because takeoversare not common in Thailand and restrictions in shareholder rights

such as staggered boards or supermajority amendments to the

charter are not permitted (Limpaphayom and Connelly, 2004).

However, we expect that more comprehensive and detailed mea-

sures of the quality of corporate governance practices, that go be-

yond superficial conformity to prescribed standards, by capturing

how governance practices are actually implemented, should exhi-

bit a positive relation with value. Furthermore, we predict that the

use of a complex ownership structure, such as a pyramidal struc-

ture, can moderate the relation between the quality of corporate

governance practices and firm value. In particular, we expect that

implementation of prescribed governance standards will have little

effect on corporate value if owners maintain excessive control over

the firm through the use of a pyramidal structure, which allows

them to effectively subvert these governance measures through

indirect means.

Overall, we address the following research questions: First, how

does the quality of corporate governance practices relate to firm

value in Thailand? Second, what is the relation between family

influence, in the form of ownership or control structures, and firm

value in Thailand? Finally, does family ownership or an enhanced

control structure moderate the relation between the quality of cor-

porate governance practices and firm value? The next section pre-

sents the empirical approach we use to address these questions.

3. Data and methodology

The sample used in this study comprises all industrial compa-

nies traded on the Stock Exchange of Thailand (SET) in 2005. Firms

in the financial services industry (banks, insurance companies, fi-

nance and securities companies, listed mutual fund companies,

and property investment funds) are excluded from the sample.

Firms that do not have complete financial data available for fiscal

year 2005, newly listed firms, delisted firms, inactive firms, or

firms undergoing financial rehabilitation or restructuring are also

excluded from the sample. We next eliminate firms where themajority shareholder is the government or a widely-held domestic

or foreign financial institution. Lastly, we eliminate firms that are

subsidiaries of foreign corporations.12 This procedure results in a

sample size of 216 firms. Financial data are obtained from Data-

stream, published by Thomson Financial, with additional financial

information supplied by the Stock Exchange of Thailand through

the SETSMART data service.

3.1. Identification of family-owned firms

We determine whether or not a sample firm is considered to be

family-owned based on the identity of the shareholders. We use the

SETSMART database, provided by the Stock Exchange of Thailand,

which lists the top ten shareholders of each firm, to identify shareownership. In addition to shareholding records, we use annual re-

ports and other outside sources to trace share ownership and, thus,

identify family ownership. For direct shareholdings, we count

shareholders with the same surname as the family as well as

shareholders with known familial relationships (relatives, spouse,

children, etc.), even if the last names are different. Many firms in

the sample exhibit indirect shareholdings; that is, shares owned

by investors in a sample firm through another public or pri-

vately-held company. For each firm in the sample, we trace any

indirect ownership of shares upwards through networks of public

and/or private companies. We also add together shares owned by

individual family members or owned by family-affiliated firms un-

der family control to find total family ownership. We classify any

company in the sample, which is part of a family-controlled net-

work, as a family firm.

We begin by classifying firms as high family ownership if the

total family ownership exceeds the median ownership for all firms

(41.2%).13 The objective of this initial classification of all sample

firms into two groups based on the level of family ownership is to

12 We are interested in the effects of family versus non-family ownership. Thus

government-owned firms are eliminated because a government may pursue public

policy-related objectives through state-owned firms rather than policies designed to

maximize shareholder wealth. For example, Gugler (2003) finds that state-owned

Austrian firms smooth dividends and are much less likely to cut payouts when

reductions are warranted compared with firms owned by families. In a similar vein,

for a local (Thai) subsidiary of a foreign corporation, countervailing issues of tax

management, profit repatriation, or transfer pricing may shift management priorities

away from shareholder wealth maximization.

13 We also confirm that a more restrictive definition of high ownership (50%) yieldsqualitatively similar results.



6/22

compare the characteristics of firms that have high family ownership

to firms with low family ownership. This initial analysis provides us

with some insights about family firm behavior that we use to guide

our subsequent empirical analyses. For the subsequent regression

analyses, we maintain the division of high and low family owner-

ship.14 This classification scheme results in two distinct groups: (i)

high family ownership firms (family ownership above the median

for all family firms); and (ii) low family ownership firms (familyownership below the median for all family firms).

3.2. Identification of pyramidal structures (ratio of cash flow rights to

voting rights)

Though Thai law requires one share, one vote, it is possible to

have differences between voting or control rights and cash flow

rights through use of indirect shareholdings or pyramidal owner-

ship. The deviation between voting rights and cash flow rights for

firms in our sample indicates the extent of control-enhancing mech-

anisms in place, such as pyramidal ownership. We follow prior re-

search in the approach we use to identify cash flow rights and

voting rights (La Porta et al., 1999).15 The ratio of cash flow rights

to voting rights, or the wedge, is the key measure we use to deter-

mine the ultimate owners of a firm through the chain of control.

For companies that only have direct shareholdings, (i.e., no chain

of control or holdings by affiliated companies), the cash flow rights

are exactly equal to the voting rights and thus the ratio of the cash

flow rights to the voting rights or the wedge is equal to one. A

wedge value lower thanone indicates that,throughindirect or pyra-

midal ownership, firms have voting rights that exceed cash flow

rights. For example, a sample firm may be majority-owned (50%)

by a second public company which is, in turn, majority-owned

(60%) by a group of family members. This firm is not widely held

and clearly has a chain of control. Examining the chain of control,

the family is the ultimate owner of the sample firm via its indirect

holdingsthrough the second firm. Thus, the sample company is con-

sidered a family firm. As in other studies, we determine the voting

rights as corresponding to the smallest shareholding or weakest

link in a chain of control. In this example, the voting rights would

be 50%. However, the cash flow rights are 50% 60% or 30%, deter-

mined by multiplying the successive shareholdings along the chain

of control. Thus, the deviation from cash flow rights and voting

rights or wedge would be equal to the cash flow rights of 30% di-

vided by thevoting rights of 50%, yielding a wedge value of 0.60. We

use the procedure described above to calculate the cash flow rights,

voting rights, and wedge values for each firm in our sample.

3.3. Corporate governance measurement

Beyond measurement of the usual variables, such as board size

and independence, we construct a corporate governance index

(CGI) in order to assess the quality of corporate governance prac-

tices among listed Thai firms. Calculated from a total of 117 sepa-

rate criteria, the CGI quantifies the overall quality of corporate

governance practices. We develop the criteria from the OECDs

(Organization for Economic Cooperation and Development) five

corporate governance principles (OECD, 1999; 2004) and then

adjust the criteria to take into account the subtleties of Thai laws

and regulations.16 We construct this corporate governance index,

rather than using an existing index (e.g., Gompers et al., 2003; Beb-

chuk et al., 2009; Brown and Caylor, 2006) because these other mea-

sures of governance may not be fully applicable to Asian markets, in

general, and Thai firms, in particular. The other indexes are built

primarily from provisions relating to takeover defenses and other

shareholder rights. Hostile takeovers are rare in Asian marketslargely because of concentrated and complex ownership structures,

and unique institutional settings.

The strengths of the corporate governance measure (CGI) used

in this study are twofold. First, the CGI index measures the actual

quality of corporate governance, i.e., the corporate governance-re-

lated activities or firm disclosures which, thus, reveal the practices

actually implemented by the firms. The CGI also acts as a gauge of

quality, showing whether the observed practices are missing

(poor), match the level required by law (good), or reach the highest

level of quality equivalent to international best practices (best).

The second strength of this measure concerns the foundations of

the survey measures themselves. The measures are rooted in eco-

nomic and financial research findings and theories, aggregating

many of the conclusions that have received empirical support from

prior research. These theories and findings form the basis for the

OECDs corporate governance principles. The complete scorecard

is shown inAppendix A. The scorecard criteria span five sections,

matching the five OECD corporate governance principles: the

rights of shareholders, equitable treatment of (minority) share-

holders, role of stakeholders, disclosure and transparency, and

board responsibilities.

3.3.1. Shareholders rights

Shareholders rights should be protected by the corporate gov-

ernance structure.17 In addition, the corporate governance structure

should also make it easy for shareholders to exercise their rights. The

first section of the measurement assesses the provision of basic

shareholder rights, including the right to: (i) secure methods of own-

ership registration; (ii) convey or transfer shares; (iii) obtain relevant

and material information on the corporation on a timely and regular

basis; (iv) participate and vote in general shareholder meetings; (v)

elect and remove members of the board; and (vi) share in the profits

of the corporation. Twenty-two measures capture the actual rights of

shareholders. To evaluate the quality of protection of shareholders

rights, we examine in detail the relevant documents provided by

14 We identify 17 firms that have no family ownership. These firms are included in

our low family ownership group. Our results remain qualitatively unaffected if we

repeat the subsequent analyses after dropping these firms.15 La Porta et al. (1999) use six classifications of ownership: widely-held (no

dominant owner); family (members of the same family with the same last name);

state (government ownership); widely-held financial institutions (financial institu-

tions that do not have a single controlling large shareholder); widely-held corpora-

tions (corporations that do not have a single controlling large shareholder; and other.

Essentially the same classification scheme is used in subsequent studies such asClaessens et al. (2000, 2002).

16 The measurement itself was initially developed as part of an annual project by the

Thai Institute of Directors Association (Thai IOD) called Corporate Governance

Baselining in Thailand, which commenced in 2000. The late Mr. Charnchai

Charuvastr, the former President of the Thai IOD, was instrumental in working with

the Thai government to include this annual project as a part of the Thai National

Governance Committee. To ensure adherence to international standards and compa-

rability, the original questionnaire was designed with technical assistance from

McKinsey and Company and the World Bank. The steering committee supervising the

development and scoring process consists of representatives from the Stock Exchange

of Thailand, the Securities and Exchange Commission, and related parties (e.g., the

Institute of Certified Accountants and Auditors of Thailand, the Thai Investors

Association, and the Government Pension Fund). The Steering Committee also

determines the composition and weighting scheme of the CGI. In the first year, the

World Bank generously provided financial support in the early stage of the project in

Thailand.17 Protection of shareholders rights is critical to economic and capital market

development (LLSV, 1997). Previous studies have shown evidence that firm value

declines when the control rights of the largest shareholders exceed their ownership

rights (Claessens et al., 2002). An example of a shareholders right that should be

protected is the ability to review the compensation for board members (Bushman

et al., 2004; Murphy, 1999). Shareholders need to receive company information that

is relevant and material (Gillian and Starks, 2000; Karpoff et al., 1996). Owners should

also be able to participate and vote in general shareholder meetings (Bhagat and

Brickley, 1984; Gordon and Pound, 1993; Ferris et al., 2003; Fich and Shivdasani,

2005), and elect and remove members of the board ( LLSV, 1998; Fama and Jensen,1983).



7/22

the companies and the regulators. The resultant score for this sub-

section is scaled to comprise 25% of CGI, the governance index.18

3.3.2. Equitable treatment of shareholders

All shareholders should be treated similarly whether they are

dominant owners, minority owners, or foreign shareholders. In East

Asia, the treatment of shareholders is critical because the compa-

nies operate in an environment in which inside, majority share-holders have advantages over outside, minority shareholders

(Claessens et al., 2002). For example, the second section of the

CGI measurement assesses whether processes and procedures for

general shareholder meetings allow for equitable treatment of all

shareholders and whether company procedures make it unduly dif-

ficult or expensive for shareholders to cast votes.19 There are 13

items to evaluate the equitable treatment of shareholders reported

in the official documents provided by the firms and the regulators.

The score for this sub-section makes up approximately 15% of CGI.

3.3.3. Role of stakeholders

The corporate governance framework should recognize the

rights of stakeholders established by law or through mutual agree-

ments and encourage active cooperation between corporations and

stakeholders in creating financially sound enterprises. The third

part of the scorecard covers the role of stakeholders, specifically

the interactions of the firm with stakeholder groups, such as

employees, creditors, suppliers, shareholders, and the environ-

ment.20 This section of the CGI measurement assesses whether

stakeholders can participate in the corporate governance process

or have access to relevant and reliable information on a timely and

regular basis. There are nine survey items in this sub-section, making

up approximately 10% of CGI.

3.3.4. Disclosure and transparency

Corporate disclosure and transparency are two cornerstones of

good governance.21 Therefore, the corporate governance framework

should ensure timely and accurate disclosures are made on all mate-

rial matters regarding the corporation, including the financial situa-

tion, performance, ownership, and governance of the company. This

section of the CGI measurement assesses whether the information

was prepared and disclosed in accordance with acceptable corporate

standards for accounting and financial and non-financial disclosures.

Further, it assesses whether the channels for disseminating informa-

tion provide for all users equal, timely, and cost-efficient access tothe relevant information. There are 32 items in the survey that

examine disclosure practices. These measures account for approxi-

mately 25% of the CGI score.

3.3.5. Board responsibilities

The final section of the scorecard addresses board responsibili-

ties, an area that has generated an extensive amount of research

interest. The OECD principles reaffirm that the corporate gover-

nance framework should ensure the strategic guidance of the com-

pany, the effective monitoring of management by the board, and

the boards accountability to the company and the shareholders.22

Much of the prior research focuses on governance variables, identi-

fying broad board characteristics such as independence and board

size. Often, however, in Thailand, there are deep personal connec-tions between the CEO and board members. Extra-contractual mech-

anisms are usually available to the CEO or family for disciplining and

controlling directors, who are often family members or otherwise

connected to the family through business or personal dealings. We

believe our in-depth assessment of board practices and policies pro-

vides a more accurate gauge of a boards ability to implement delib-

erate and meaningful practices designed to enhance corporate

governance. In total, there are 41 survey items to assess board

responsibilities. The items in this section make up approximately

30% of CGI.

We draw the data used to evaluate the governance practices for

each firm from a wide variety of publicly available information

sources, such as annual reports, Securities and Exchange Commis-

sion and Stock Exchange of Thailand filings, minutes from annualshareholders meetings, articles of association, company by-laws,

and company websites. To reinforce the emphasis on minority

shareholders, this study assumes the viewpoint of an outside

investor. Specifically,only publicly available official documents serve

as source documents since this information would be readily avail-

able to outside investors. We score each company on all applicable

areas of CGI.

As assessing the level of corporate governance for an individual

company can be subjective, we design the scoring scheme to min-

imize this problem. In addition to cross-checking and auditing by

different raters, we quantify nearly every governance measure in

our study. This is also a unique feature of this study, as previous

research has only checked for the presence or absence of a specific

corporate governance measure. With our assessment procedure,

companies that omit or do not comply with a specific scoring

18 The Steering Committee (See Footnote 16) assigned weights to each component

of CGI. However, our results are robust to alternative weighting schemes including

assigning equal weights to each component.19 There are several aspects of treatment of shareholders included in the measure-

ment. For example, Thai law prescribes one share, one vote, which has been shown to

benefit shareholders (Grossman and Hart, 1988; Harris and Raviv, 1988). Additionally,

shareholders should have the opportunity to obtain effective redress should their

rights be violated. One example would be the abuse of minority shareholders at the

hands of controlling shareholders. Abusive related-party transactions (tunneling or

propping) have been documented as a serious violation of shareholder rights in

emerging markets (Cheung et al., 2006; Friedman et al., 2003; Johnson et al., 2000). In

addition, internal control systems need to be established to prevent the use of

material inside information (Givoly and Palmon, 1985), for example to prevent insider

trading.20 Jensen (2002) contends that a firm cannot maximize value if it ignores the

interest of its stakeholders. This view is supported byAllen et al. (2009)who note that

firms may voluntarily choose to be stakeholder-oriented because this increases their

value in certain circumstances. Connelly and Limpaphayom (2004) find that an

optimally designed environmental policy can maximize market valuation in an

emerging market.21 Prior research (e.g., Ball, 2001; Bushman et al., 2004; Khanna et al., 2004)

demonstrates that both the quantity and the quality of corporate disclosure are

integral parts of effective governance practices. First, the ownership stakes of a firm

should be transparent (LLSV, 1998; Bushman et al., 2004; Claessens et al., 2002;

Mallette and Fowler, 1992; Himmelberg et al., 1999). The quality of information

provided to investors is another important dimension(Meek et al., 1995; Singhvi and

Desai, 1971). The quality of disclosure can also be a useful indicator of the levels of

agency conflicts and information asymmetries within the firm ( Healy and Palepu,

2001; Meek et al., 1995). Further, information such as related-party transactions,

external audit results, and insider transactions would be important to investors

(Cheung et al., 2006; Johnson et al., 2000; Fan and Wong, 2005 ). Also, prior research

has identified the benefits of using multiple channels to communicate with investors

(Ashbaugh et al., 1999; Lang and Lundholm, 1993, 1996; Farragher et al., 1994).

Anderson et al. (2009) find that family firms in the US lack transparency and that

founder- and heir-controlled firms are associated with weaker performance whenthey are opaque.

22 An effective board can reduce agency conflicts by exercising its power to monitor

and control management (Fama and Jensen, 1983). Frequent and well-attended board

meetings can have a positive effect on firm performance (Vafeas, 1999; Ferris et al.,

2003; Fich and Shivdasani, 2005). The auditing role of the board is one area where

substance, in this case concrete actions that enhance governance effectiveness, can be

separated from form (Adams, 1994; Scarbrough et al., 1998; Carcello et al., 2002;

Turpin and DeZoort, 1998). Directors must also exercise their role in oversight,

including evaluation of the CEO (Boyd, 1994) and themselves (Ingley and van der

Walt, 2002). One area where substance would clearly supersede form is the important

area of board committees (Klein, 1998). A number of researchers have investigated

the roles of board committees in propagating principles of good governance, e.g., on

the importance of audit committees (Carcello and Neal, 2000; Klein, 2002; Krishnan,

2005). Well-functioning audit, compensation, and director nomination committees,

each with a clear mission and each effectively discharging its duties, play a role in

protecting shareholder interests and increasing firm value (Brick et al., 2006; Dailyet al., 1998).



8/22

criterion receive a poor score. Meeting the legal compliance stan-

dard earns a firm a score of good, while firms which exceed the

regulatory requirements and/or meet international standards re-

ceive the highest score. Once the assessment is complete, we scale

the CGI score, aggregated across the five sub-sections, from zero to

100%.

The original version of the survey was designed and piloted on a

select group of 133 companies in 2000. The Thai Institute of Direc-

tors Association conducted the second and third studies in 2002

and 2004. After the initial survey and during the intervening years,

a number of questions in the original survey were revised while

new questions were added. The original version contained 57questions whereas the latest version contains 117 questions (see

Appendix A). While the number of questions and the format of

the survey have changed over the years, the structure of the survey

retains the same mapping of the OECD Principles of Corporate Gov-

ernance (1999). The survey was then executed again in 2005 with

the most comprehensive coverage: nearly all firms listed on the

Stock Exchange of Thailand were included in the study.

The time-series results of the survey from 2002 to 2005 are

shown in Table 1 and, graphically, inFig. 2. The results from the

overall CGI scores show that the quality of corporate governance

practices of Thai companies has gradually improved. Specifically,

the average CGIscore improved from 57.61 (out of 100) in 2002

to 61.79 in 2005. This is attributable to the reform efforts initiated

by the government and related organizations. The results clearlyshow that Thai companies are going through a major transformation

and, therefore, no longer have the same characteristics as Thai

companies examined in previous studies using samples before

the Asian financial crisis (e.g., Claessens et al., 2000; Wiwattana-

kantang, 2001).

From the descriptive statistics in Table 1, the increase in the

overall quality of corporate governance practices comes largely

from two major areas, disclosure and transparency (Section D)

and role of stakeholders (Section C). The improvement in the area

of disclosure and transparency is shown by the increase in the

average score from55.54 in 2002 to 73.09 in 2005. This rise is most

likely a result of regulatory reforms and the implementation of

new rules and regulations by the Stock Exchange of Thailand(SET) and the Securities and Exchange Commission. During this

period, the SET and the Thai Institute of Directors Association also

launched programs to increase the awareness of the role of stake-

holders among listed firms which, in turn, led to a sizable increase

in the score (43.06 in 200264.04 in 2005) in Section C of the sur-

vey. The equitable treatment of shareholders (Section B) also

shows some improvement (69.32 in 200274.85 in 2005) whereas

the scores for the board responsibilities segment (Section E) show

virtually no improvement.

The most striking result is a substantial declinein the scores for

the rights of shareholders (Section A), going from 72.45 in 2002 to

64.40 in 2005. This is largely due to an increase in cross-holdings

andthe use of pyramidalstructures by Thai family firms. In a pre-cri-

sis study byClaessens et al. (2000), the average ratio between cashflow rights and voting rights among Thai firms is close to 1.0 (i.e.,

Table 1

Descriptive statistics of corporate governance index by survey year.

Year CGI Section A Section B Section C Section D Section E Number of firms in survey

2002 57.61 72.45 69.32 43.06 55.54 49.26 294

2004 61.30 63.68 72.52 61.80 74.05 50.09 327

2005 61.79 64.40 74.85 64.04 73.09 50.84 364

This table presents theaverage of theCorporateGovernance Index (CGI) andthe fiveCGI sub-indexes based on theOECD corporategovernance principles (1999) for three CGI

surveys. The surveys were completed during 2002, 2004, and 2005. The sample is drawn from publicly-traded firms in Thailand. The CGI and each sub-index ranges areexpressed as percentages and range from 0 to 100. The five subsections are: rights of shareholders (Section A); equitable treatment of shareholders (Section B); role of

stakeholders (Section C); disclosure and transparency (Section D); and board responsibilities (Section E). Survey questions are shown in Appendix A.

02

02

02

02

02

02

0404

04

04

04

04

0505

05

05

05

05

0

10

20

30

40

50

60

70

80

CGI Section A Section B Section C Section D Section E

CGIScores by Section, based on the OECD Principles of Corporate Governance (1999)

CorporateGovernanceIndex(CGI)

2002,

2004,and2

005

Surveys

Fig. 2. Average corporate governance survey scores (CGI). The figure shows the average scores of the Corporate Governance Index (CGI) in Thailand for three survey years:

2002, 2004, and2005. Thefive subsections are: rights of shareholders (Section A); equitabletreatment of shareholders (Section B); role of stakeholders (Section C); disclosure

and transparency (Section D); and board responsibilities (Section E).



9/22

no pyramidalstructures areobserved). In 2005, the average ratio be-

tween cash flow rightsand voting or ownership rightsis 0.833,indi-

cating that Thai firms now have substantially increased the use of

this type of control-enhancing structure. Overall, it appears that

while Thai firms have improved some governance measures, these

firms have also experienced an effective decline in shareholder

rights.

3.4. Empirical methods

In our initial analyses, we examine the influence of family con-

trol and family ownership on firm performance using OLS regres-

sion analyses.23 Tobins q, a measure of firm market valuation, is

the dependent variable.24 We include variables for family ownership

and various board characteristics in the regressions, as well as other

firm characteristics, which serve as control variables.

We evaluate the first model, shown below as Eq.(1), using the

full sample.

qi bob1BOD SIZEib2BOD INDib3FAM OWNi

b4FAM OWN2i b5SIZEib6FIRM AGEib7ROAi

b8CAPITAL EXPENDITURESib9LEVERie1i 1

For the second regression, we add a dummy variable (Wedge dum-

my), indicating the presence of a control-enhancing pyramidal

ownership structure, as an additional explanatory variable to the

first model.

qi bob1WEDGE DUMMYib2BOD SIZEib3BOD INDi

b4FAM OWNib5FAM OWN2i b6SIZEi

b7FIRM AGEib8ROAib9CAPITAL EXPENDITURESi

b10LEVERie1i 2

For the third regression, we add CGI, the corporate governance

score, as an additional explanatory variable to Model 1.

qi bob1CGIib2BOD SIZEib3BOD INDi

b4FAM OWNib5FAM OWN2i b6SIZEi

b7FIRM AGEib8ROAib9CAPITAL EXPENDITURESi

b10LEVERie1i 3

Lastly, we include both the Wedge dummy variable and CGI as addi-

tional explanatory variables.

qi bob1CGIib2WEDGEDUMMYib3BOD SIZEi

b4BOD INDi b5FAM OWNib6FAM OWN2i

b7SIZEib8FIRM AGEib9ROAi

b10CAPITAL EXPENDITURESi b11LEVERie1i 4Tobins q, as a market-based measure of firm performance, is de-

fined as the sum of the book value of long-termdebt and the market

value of equity divided by the book value of total assets. CGI is the

percentage score from the Corporate Governance Index based on

the OECD Principles of Corporate Governance (2004). The CGI score

is comprised of scores on five sub-components of CGI: rights of

shareholders, treatment of shareholders, role of stakeholders, dis-

closure and transparency, and board responsibilities. Both CGI and

the sub-component scores are expressed as percentages. Family

ownership is the proportion of outstanding shares held by the

founding family and affiliated members.

We include several board-related measures that have been used

in other studies as control variables. These elements, i.e., board size

and board composition, and their relation to firm performance,

have been investigated in the literature but show mixed results.

In this study, board size is the number of directors on a firms board

while board independence (IND) is the proportion of directors who

are independent/outside directors.

For robustness, we also include several other control variables.

We calculate profitability by taking the ratio of net income after

taxes divided by total assets. Firm size is the natural log of total as-

sets. Firm age is the natural logarithm of the years since the firms

founding. We define financial leverage as the ratio of long-term

debt divided by total assets.

4. Results

Table 2contains the descriptive statistics for the 216 firms in

the sample. We classify companies into High Family Ownership

and Low Family Ownership groups depending on whether the

proportion of outstanding shares held by family or related mem-bers is above or below the median for all firms (41.2%). Table 2

shows three sets of statistics: (i) for the full sample of 216 firms,

(ii) for 108 firms with high family ownership, and (iii) 108 firms

with low family ownership (including firms with zero family own-

ership). When comparing the high family ownership firms to low

family ownership firms, the descriptive statistics show some strik-

ing differences. The average share ownership of family and affili-

ated members is 39.29% for the full sample. Low family

ownership firms have a mean family ownership percentage of

20.21% whereas the mean percentage is 58.37% for high family

ownership firms. The difference is statistically significant at con-

ventional levels (t = 20.76). The difference in firm performance,

as measured by Tobins q, is also statistically significant and lower

for high family ownership firms. The average Tobins q for the full

sample is 0.82. Low family ownership firms have an average q of

0.88, which is higher than the averageq for high family ownership

firms (0.76). The difference is statistically significant at the 10% le-

vel (t= 1.77). The average return on assets for high family owner-

ship firms is also lower than that for low family ownership firms

but the difference is not statistically significant (t= 1.18). Firms

with high family ownership are, on average, slightly smaller (mean

firm size of 14.74, measured by the natural logarithm of total as-

sets) than low family ownership firms (15.10; t= 2.07) and have

lower levels of financial leverage, measured by the ratio of long-

term debt to total assets (0.09 versus 0.14; t= 2.48), as judged

by the ratio of long-term debt to total assets, suggesting that high

family ownership is associated with more conservative financing

strategies.25

With respect to general governance characteristics, the two

types of firms have similar percentages for the average proportion

of independent directors on the board. The mean is 35% (low own-

ership) versus 33% (high ownership) for the two groups of firms

(t= 1.37). Similarly, the two types of firms have average board

sizes (11.1 directors versus 11.3) that are nearly identical. While

the overall CGI scores for the two groups are similar, with a mean

score of 69.17 for high family ownership firms versus a mean of

70.25 for low family ownership firms (t= 0.84), some CGI sub-

section scores are lower for high family ownership firms than low

family ownership companies. Rights of Shareholders and Disclosure

23 In subsequent analyses, we use a simultaneous equations approach to control for

a potentially endogenous relation between CGI and q.24 To make the results comparable toBebchuk et al. (2009), the empirical analyses

are repeated using the natural logarithm ofq as the dependent variable. The overallresults are qualitatively similar to the reported results.

25 Family-controlled firms also appear to be associated with lower levels of debt in

the US. Agrawal and Nagarajan (1990) report that all-equity firms have higher cash

holdings and significantly higher family control compared to a matched controlsample of levered firms.



10/22

and Transparency are both lower for high family ownership com-

panies. However, the difference is statistically significant only for

Rights of Shareholders. The scores for the Treatment of Sharehold-

ers, Roles of Stakeholders, and Board Responsibilities sections arestatistically indistinguishable between the two types of firms.

We classify our sample companies into two groups based on the

ratio between cash flow rights and voting rights. Companies with a

wedge or pyramidalstructure aredefinedas those thathave a ratio

of cash flow rights to voting rights lower than one. The comparison

betweensamplefirmsthat employa pyramidalownershipstructure

and other firms, shown in columns 4 and 3, respectively ofTable 2,

reveals some interesting findings. From the descriptive statistics in

Table 2, both groups of companies are quite similar with respect to

financial characteristics. There is no statistical difference in firm

age, firm size, profitability, capital expenditures and leverage be-

tween the two groups. However, companies witha pyramidal struc-

ture have higher average family ownership than those with no

wedge (t=

1.80). This is consistent with the notion that foundingfamilies attempt to use the pyramidal structureto maintain control

of their firms. Interestingly, firms with pyramidal structures have a

highermean valuationthan those that do not employ thepyramidal

structure(t= 2.03). It also appears that companies withpyramidal

structures exhibit better corporate governance, as measured by CGI.On the surface, it appears that pyramidal structures are value-

enhancing. However, these univariate comparisons could be mis-

leadingnot onlybecause of the complexand potentially endogenous

nature of the interrelationshipamongownership, control and value,

but also because greater control may allow family firms to manipu-

latethe corporate governance measures to showbetterscores,with-

out actually implementing value enhancing strategies. We examine

these issues further in subsequent empirical analyses.

Table 3 contains a correlation matrix forthe variables used in the

study. Panel A contains the correlations between the overall CGI

score and firm characteristics. The correlation between Tobins q

andboardindependence is positive andstatistically significant while

the correlation with board size is not statistically significant. Most

importantly, there appears to be a positive and statistically signifi-cant relation between q and the quality of corporate governance

Table 2

Descriptive statistics.

All

firms

High family

ownership

Low family

ownership

t-

Statistic

No pyramidal structure

observed

Pyramidal structure

observed

t-

Statistic

(1) (2) (12) (3) (4) (34)

Family ownership (%) 39.29 58.37 20.21 20.76*** 37.56 44.00 1.80*

(23.39) (11.46) (15.28) (23.47) (22.72)

Wedge 0.83 0.82 0.85 0.87 1.00 0.38 35.13***

(0.30) (0.31) (0.29) (0.00) (0.22)Tobinsq 0.82 0.76 0.88 1.77* 0.78 0.93 2.03**

(0.50) (0.47) (0.52) (0.49) (0.49)

Profitability (ROA) 0.05 0.05 0.06 1.18 0.05 0.06 0.95

(0.07) (0.07) (0.08) (0.08) (0.05)

Firm size 14.92 14.74 15.10 2.07** 14.88 15.03 0.74

(1.31) (1.21) (1.37) (1.31) (1.30)

Firm age 3.18 3.24 3.13 1.75* 3.19 3.17 0.28

(0.47) (0.41) (0.51) (0.44) (0.53)

Capital expenditures 0.08 0.08 0.08 0.20 0.08 0.09 0.94

(0.08) (0.07) (0.08) (0.07) (0.09)

Financial leverage 0.11 0.09 0.14 2.48** 0.11 0.11 0.01

(0.13) (0.12) (0.14) (0.12) (0.16)

Board size 11.20 11.26 11.14 0.29 10.91 12.00 2.38**

(3.02) (3.24) (2.81) (2.98) (3.04)

Board independence 0.34 0.33 0.35 1.37 0.34 0.32 1.21

(0.10) (0.10) (0.11) (0.10) (0.11)

Corporate governance

index

69.71 69.17 70.25 0.84 68.40 73.28 3.44***

(9.47) (9.40) (9.55) (9.49) (8.51)

Rights of shareholders 70.93 69.27 72.59 1.76* 69.91 73.70 1.79*

(13.89) (12.78) (14.78) (14.11) (12.96)

Treatment of

shareholders

74.06 73.72 74.40 0.70 73.81 74.74 0.85

(7.15) (6.59) (7.68) (7.44) (6.29)

Role of stakeholders 70.49 71.05 69.94 0.38 68.64 75.54 2.14**

(21.18) (20.09) (22.30) (22.24) (17.15)

Disclosure and

transparency

81.74 80.99 82.48 1.22 80.62 84.77 3.07***

(8.99) (8.96) (8.98) (9.45) (6.77)

Board responsibilities 53.93 53.70 54.15 0.23 51.82 59.67 3.74***

(14.08) (14.50) (13.72) (13.60) (13.88)

N 216 108 108 158 58

This table presents summary statistics of variablesused in thestudy. Thesample consists of firms listedon theStockExchange of Thailandin 2005. Thesample is split based onthemedian value of familyownership. Familyownership is thenumber of outstanding shares held by thefounding familyand affiliatedmembers divided by thetotalnumber

of shares outstanding. Wedge is the ratio of the cash flow rights divided by the voting rights. Tobins q is the book value of long-term debt plus the market value of equity

divided bythe book value of total assets. Profitability(ROA) is theratio of netincome after taxesdivided bytotalassets. Firm size is thenatural logof total assets. Firm ageis the

natural log of the years since the firms founding. Capital expenditures are the ratio of capital expenditures divided by total assets. Financial leverage is the ratio of long-term

debt divided by total assets. Board size indicatesthe numberof directorson theboardof directors. Board independenceis thenumber of directorswho areindependent/outside

directors divided by board size. CGI is the percentage score from the Corporate Governance Index based on the OECD Principles of Corporate Governance. The five sub-

components of CGI are: rights of shareholders, equitable treatment of shareholders, role of stakeholders, disclosure and transparency, and board responsibilities, expressed as

percentages. Standard deviations are shown in parentheses. t-Statistics are calculated for the differences between family-owned firms and other firms.* statistical significant differences at the 10% level (two-tailed).** statistical significant differences at the 5% level (two-tailed).*** statistical significant differences at the 1% level (two-tailed).



11/22

practices, as measured by the CGI.Panel B shows the correlationsbe-

tween CGIandthe fivesub-components. Thecorrelations amongthe

sub-components are positive and statistically significant, showing

that these five aspects of corporate governance practices are

interrelated.

Table 4 presents regression results using Tobins qas the depen-

dent variable. In this table, we run regressions using all firms in the

sample, adding, in turn, additional variables of interest such as

dummy variables for the presence of pyramidal ownership and

the CGI. To recap, the purpose of this series of regression analyses

is twofold. First, we want to examine the relation between firm va-

lue, as measured by q, and (1) the proportion of family ownership

and (2) conventional governance variables. We also wish to

compare the predictive ability of the CGI variable to that of

conventional governance variables. Secondly, we want to look at

the elation between the use of pyramidal ownership structures

and firm value, after controlling for other factors.

In the first model, the coefficient for family ownershipis not statis-

tically significant at conventional levels. This indicates that after con-

trolling for other factors, there is no significant relation between

family ownership and firm value, consistent with ownership being

endogenously determined (Demsetz and Lehn, 1985). In Model 2, we

include a dummy variable to indicate the presence of a pyramidal

structure (Wedge dummy). The coefficient for the presence of a pyra-

midal structure is positive and statistically significant. However, the

statistical significance disappears after adding additional control vari-

ables, as shown in the subsequent models. Looking at the results from

models 1 and2, a striking finding is that there is no significant relation

betweenq and conventional corporate governance variables, such as

boardcharacteristics.It appears thatthesegovernancevariables,which

have traditionally been found to be associated with firm value (q)in

developed economies following common law,26 apparently do not

exert any influence on firm value for Thai family firms. This finding

also provides preliminary support for our contention that many gov-

ernance variables represent the form but not the substance of

effective corporate governance practices in countries where the fam-

ily exerts substantial control over voting rights and control mecha-

nisms are not transparent.

In order to shed further light on the relation between the qual-

ity of governance practices and value, we go beyond the standard

governance variables and turn to the Corporate Governance Index

(CGI) described earlier. Recall that the CGI index is a composite

score, aggregating many of the mandated and voluntary provisions

and disclosures made by firms. These provisions and disclosures

are relevant to the quality and visibility of corporate governance

practices. Thus, we expect the CGI index to exhibit an association

with market valuation. Models 3 and 4 present regression analyses

with the addition of CGI. We find that as before, the coefficients for

conventional governance variables (e.g., board size, board indepen-

dence and family ownership) are still not statistically significant.

Most interestingly, the regression shows a positive and statistically

significant relation between CGI and Tobins q. The coefficient for

CGI is positive in model 3 and in model 4, which includes the dum-

my variable for the presence of a pyramidal ownership structure.

Moreover, in model 4, the regression coefficient for the presence

of a pyramidal structure is no longer statistically significant. We

find that of the control variables, profitability shows a positive

and statistically significant relation to Tobins q in all models.

The coefficients for capital expenditures are also positive and sig-

nificant in all four regression models. Financial leverage does not

have a statistically significant coefficient in any model.

It is possible that our results are affected by an endogenous

relation between CGI and q. Lehn et al. (2007)find that after con-

trolling for past performance, there is no contemporaneous associ-

ation between the quality of corporate governance, as measured by

the GIM Index,27 and firm valuation. As a robustness check, we

Table 3

Correlation matrix.

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

Panel A: Correlations between CGI and firm characteristics

(1) Family 1.0

(2) Wedge 0.09 1.0

(3) Tobinsq 0.09 0.14 1.0

(4) ROA 0.05 0.08 0.49 1.0

(5) Firm size 0.18 0.01 0.20 0.12 1.0(6) Firm age 0.11 0.01 0.06 0.01 0.08 1.0

(7) Cap ex 0.02 0.05 0.24 0.18 0.05 0.14 1.0

(8) Leverage 0.19 0.02 0.14 0.01 0.47 0.14 0.31 1.0

(9) Board size 0.08 0.19 0.05 0.02 0.14 0.26 0.06 0.04 1.0

(10) Independence 0.13 0.14 0.12 0.09 0.06 0.14 0.02 0.13 0.62 1.0

(11) CGI 0.12 0.19 0.30 0.25 0.47 0.02 0.02 0.27 0.05 0.07 1.0

(1) (2) (3) (4) (5) (6)

Panel B: Correlations among CGI and sub-components of CGI

(1) CGI 1.0

(2) Rights of shareholders 0.80 1.0

(3) Treatment of shareholders 0.24 0.13 1.0

(4) Role of stakeholders 0.74 0.44 0.07 1.0

(5) Disclosure and transparency 0.75 0.52 0.18 0.37 1.0

(6) Board responsibilites 0.87 0.63 0.04 0.54 0.60 1.0

This table presents correlation coefficients among variables used in thestudy. Thesample consists of firms listedon the Stock Exchange of Thailand in 2005. In Panel A, family

ownership is the number of outstanding shares held by the founding family and affiliated members divided by the total number of shares outstanding. Wedge is the ratio of

the cash flow rights divided by the voting rights. Tobins q is the book value of long-term debt plus the market value of equity divided by the book value of total assets.

Profitability(ROA) is theratioof netincome after taxes divided by total assets. Firm size is thenatural logof total assets. Firm ageis thenatural logof theyearssince thefirms

founding. Capital expenditures is the ratio of capital expenditures divided by total assets. Financial leverage is the ratio of long-term debt divided by total assets. Board size

indicates the number of directors on the board of directors. Board independence is the number of directors who are independent/outside directors divided by board size. CGI

is the percentage score from the Corporate Governance Index survey based on the OECD Principles of Corporate Governance. Panel B presents the correlation coefficients

between CGI and the five sub-components of CGI: rights of shareholders, equitable treatment of shareholders, role of stakeholders, disclosure and transparency, and board

responsibilities. Correlations that are statistically significant at the 10 percent level are shown in bold.

26 Thus, our comparison group is primarily firms in the US and UK. For instance, see

Yermack (1996) on the significance of board size and Morck et al. (1988) on therelat

The Effect of Corporate Governance and Ownership Structure on Firm Value in Thailand

Documents

Transcript of The Effect of Corporate Governance and Ownership Structure on Firm Value in Thailand