Foreign ownership and firm performance case of vietnam

Purpose – The thesis aims to investigate the relationship between foreign ownership and firm performance on a selective sample of firms listed onHochiminh Stock Exchange for period 2007-

MINISTRY OF EDUCATION AND TRAINING

UNIVERSITY OF ECONOMICS HOCHIMINH CITY

-o0o -NGUYỄN THỊ THANH TÂM

FOREIGN OWNERSHIP AND FIRM PERFORMANCE: CASE OF VIETNAM

MAJOR: BUSINESS ADMINISTRATION

MAJOR CODE: 60.34.05

MASTER THESIS SUPERVISOR: Dr VÕ XUÂN VINH

HO CHI MINH CITY, 2012

I am grateful to the Examination Committee (Professor Nguyễn ĐôngPhong, Dr Nguyễn Trọng Hoài, Dr Nguyễn Đình Thọ, Dr Nguyễn VănNgãi, and Dr Nguyễn Thị Mai Trang) for their valuable advices to make

my thesis improved

I would like to thank my professors at Faculty of Business Administrationand Postgraduate Faculty, University of Economics Ho Chi Minh City fortheir teaching, guidance and support during my MBA course

I owe my sincere thanks to my classmates for their encouragement and aspecial thank of mine goes to my class monitor, Bùi Hồng Thu, for hisgreat assistance in collecting data for this thesis and for his guidance ineconometrics respects

I wish to thank my family for their unconditional support andencouragement during my MBA course

i

Purpose – The thesis aims to investigate the relationship between foreign

ownership and firm performance on a selective sample of firms listed onHochiminh Stock Exchange for period 2007-2010

Methodology – The thesis applies the Ordinary Least Squares method to

run multiple regressions on the whole sample and on each level of foreignownership in order to give a closer view at the relationship betweenforeign ownership and firm performance

Findings – The empirical results show a significant correlation between

foreign ownership and firm performance, measured by Tobin’s Q Theregressions on each level of foreign ownership indicate that foreignownership was found to be significantly positive correlated with Tobin’s

Q when foreigners own between 5% and 20% of shares in firms, while anegative correlation occurs where foreign holdings are more than 20%,specially considerably negative where the level is more than 40%; andthere is no significant relationship between the two variables whereforeigners own less than 5% of shares

Originality/Value – The thesis tries to analyze how foreign ownership

affects firm performance and suggests that Vietnamese business ownersshould take a scrutiny on benefits and costs of foreign investment

Key words – Foreign ownership, firm performance.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMENT i

ABSTRACT ii

LIST OF TABLES iv

LIST OF FIGURES v

ABBREVIATIONS vi

1 INTRODUCTION 1

1.1 Background 1

1.2 Purpose 3

1.3 Scope 4

1.4 Research questions 5

1.5 Structure 5

2 LITERATURE REVIEW 6

3 METHODOLOGY 16

3.1 Data 16

3.2 The model 16

3.3 Statistical Method 22

4 DATA ANALYSIS 23

4.1 Descriptive Statistics 23

4.2 Correlations 25

4.3 Regression Results 26

5 CONCLUSION 37

REFERENCES 39

APPENDICES 43

4

LIST OF TABLES

Table 4.1 Descriptive Statistics 23Table 4.2 Correlation matrix 25Table 4.3 Ordinary Least Squares Regression Results 27

5

LIST OF FIGURES

Figure 1.1 FDI contributions for the period 2006- 2011 2

Figure 1.2 FDI registered and implemented capital for the period

2006-2011 3

6

ABBREVIATIONS

1 INTRODUCTION

1.1 Background

The relationship between ownership structure and firm performance hasbeen examined since decades Some researches shows no effect ofownership structure on firm performance, while others indicate there is acorrelation between these two factors

Demsetz (1983) argues that there should be no relationship betweenownership structure and firm performance Pursuing this argumentempirically, Demsetz and Lehn (1985) find no significant correlationbetween profit rates and various measures of ownership concentration in

a sample of 511 United States companies using 1980 data Himmelberg

et al (1999) extend the Demsetz and Lehn (1985) study by adding new

variables to explain the variation in the ownership structure Ownershipstructure is measured by shareholdings of insiders (officers plusdirectors) Firm performance measure is Tobin’s Q They employ thecapital-to sales, R&D-to-sales, advertising-to-sales, and operatingincome-to-sales ratios as instrumental variables Controlling for thesevariables and fixed firm effects, they find that changes in ownershipholdings have no significant impact on performance Demsetz andVillalonga (2001) continue to examine the ownership-performancerelation by treating ownership structure as an endogenous variable and as

an amalgam of shareholdings owned by persons with different interests

By estimating a two-equation model for United States firms, theirevidence shows that performance (defined as Tobin’s Q or the accountingprofit rate) is not found to be influenced by ownership (defined as

8

2006

FDI contribution to GDP

FDI contribution to the total national investments

managerial ownership (Chief Executive Officers, board of directors, topmanagement) or ownership by the five largest shareholders)

Despite the fact that Demsetz and Lehn (1985), Himmelberg et al (1999),

and Demsetz and Villalonga (2001) find no significant correlationbetween ownership structure and firm performance, series of subsequentresearches, which had been done and built on the Demsetz heritage, prove

the converse results For instance, Andersson et al (2004) find that

dispersed ownership is associated with worse performance whenexamining firms listed on Sweden Stock Exchange A significantnegative relationship is also found by Lee and Chuang (2009) wheninvestigating the relation between insiders and corporate performance of

Taiwanese firms In line with this, Fishman et al (2005) find that

managerial ownership impacts negatively on performance This isopposite to the findings reported by Drakos and Bekiris (2010), wheremanagerial ownership is found to be significantly positive correlated withTobin’s Q

Figure 1.1 FDI contributions for the period 2006-2011

25.7 25.8 26 17.96 18.43 18.33 18.72 19 16

2007 2008 2009 2010 2011 Year

(Source: Foreign Investment Department, Ministry of Planning and Investment)

Figure 1.2 FDI registered and implemented capital for the period 2006-2011

(Source: Foreign Investment Department, Ministry of Planning and Investment)

As a part of ownership structure, foreign ownership plays an importantrole In Vietnam, foreign investments contribute considerably to Vietnameconomy

Figure 1.1 and Figure 1.2 indicate that FDI implemented capitalcontinuously increases during the period 2006-2011 FDI implementedcapital in 2011 reaches 11 billion USD, which contributes 26% to thenational investments and 19% to GDP Companies with foreign capitalparticipation are a form of FDI Foreign investors bring to the receivingcompanies some benefits, such as solid financial sources, moderntechnology, and management skills To some extent, foreign ownershipshould have impact on firm performance

1.2 Purpose

The relation between ownership structure and firm performance remainscontroversial in numerous studies in diverse economies For instance,

Demsetz (1983), Demsetz and Lehn (1985), McConnell and Servaes

(1990), Himmelberg et al (1999), and Demsetz and Villalonga (2001)

conduct survey in United States of America; Mudambi and Nicosia

(1998) and Dinga et al (2009) study the case of United Kingdom; the studies of Welch (2003) and Fishman et al (2005) are based on Australia

data; Kuznetsov and Muravyev (2001) examine the performance relationship in Russia; and other empirical analysis are

ownership-reported by Andersson et al (2004) and Bandick (2005) in Sweden, Shiab and Abu-Tapanjeh (2005) in Jordan, by Aydin et al (2007) in

Al-Turkey, by Kapopoulos and Lazaretou (2007) in Greek, by Lee and

Chuang (2009) in Taiwan, by Hu and Zhou (2006) and Hess et al (2010)

in China, by Drakos and Bekiris (2010) in Egypt, by Priya andShanmughan (2011) in India, by Gelubcke (2011) in Germany, and by

1.3 Scope

The present thesis employs a selected sample of the companies listed onHoSE for the period 2007-2010

The thesis is limited to examine the impact of foreign ownership on firmperformance without taking into account the relation between the origin

of foreign investors and firm performance

The thesis does not examine the relationship between firm performanceand foreign investors who hold managerial positions in the companies

1.4 Research questions

To explore the relationship between foreign ownership and firmperformance, the two specific research questions are set as follows

performance of HoSE listed companies?

• Does foreign ownership affect positively firm performance?

1.5 Structure

The thesis does not follow the conventional way where each section isoutlined into chapter but into section The remainder of this thesis isorganized as follows The literature review section summarizes previousstudies on this topic The data and the methodology employed in thethesis are presented in the following section The analysis sectionintroduces the empirical findings The final section concludes the thesis

2 LITERATURE REVIEW

Though the ownership-performance relationship has been the subject ofvoluminous researches, no agreement has been reached

Some studies find no link between ownership structure and firm

performance, namely Demsetz and Lehn (1985), Himmelberg et al

(1999), Demsetz and Villalonga (2001), Welch (2003), Klungland andSunde (2009), and Mihai (2012)

Demsetz and Lehn (1985) apply the OLS method to run statisticalanalysis over 511 firms in United States of America, for the time period1976-1980, and find no significant relationship between ownershipconcentration and accounting profit rates This cross-sectional analysis is

then extended by Himmelberg et al (1999) by adding new variables to

explain the variation in ownership structure

Himmelberg et al (1999) find that managerial ownership and firm

performance measured by Tobin’s Q are endogenously determined byfirm specific factors and key variables in the firm’s contractingenvironment Controlling both for observed firm characteristics and firmfixed effects, they conclude that managerial ownership does not affectfirm performance However, examining the endogeneity of ownershipstructure by using instrumental variables, they find a quadratic

relationship between ownership and performance Himmelberg et al

(1999) conclude that previous works are unable to examine the observable heterogeneity, and hence any relationship detected mightresult from spurious correlations

non-Demsetz and Villalonga (2001) investigate the link of ownershipstructure and corporate performance, but in a new way, where ownership

is made multi-dimensional and treated as an endogenous variable.Conducting both OLS regression analysis and 2SLS test, over a sample of

223 firms quoted in the Fortune 500 list, for the time-period 1976-1980,they affirm that there was no significant relation between ownershipstructure and firm performance

Applying and developing the models by Demsetz and Villalonga (2001),Welch (2003) examine the connection of ownership structure with firmperformance, on a sample of 114 public companies listed on theAustralian Stock Exchange, for the time-period 1999-2000 The OLSresults show that ownership is significant in explaining performance.However, when endogeneity is taken into account, the 2SLS regressionprovides no statistical dependence of ownership on performance.Additionally, the results from a generalized nonlinear model illustratelimited evidence of a nonlinear relationship between managerial shareownership and firm performance

The similar results are found in the study of Klungland and Sunde (2009),

on a large sample of quarterly data from non-financial Norwegiancompanies listed on the Oslo Stock Exchange in the period 2001-2007.Using OLS analysis, Klungland and Sunde (2009) find a significantnegative relation between ownership concentration and firm performance,measured by Tobin’s Q Nevertheless, when controlling for fixed firmeffects, there is no significant relationship Using the method of

instrument variables (2SLS) to account for endogeneity of ownershipstructure, Klungland and Sunde (2009) find that the choice of instrumenthighly affects the significance of the results Since the results obtainedfrom using instrument variables are questionably driven by weak

(econometrically) that ownership concentration influences firmperformance

Mihai (2012) employs linear regression analysis to investigate whetherthe foreign equity is associated with better performance in the case of 63listed Romanian companies on the Bucharest Stock Exchange Theoutcome of the study suggests that there is no significant link betweenfirm performance and the existence of foreign capital

In contrast to the above-mentioned studies, various empirical analysesprove the reverse, for instance, Forsyth and Dwyer (1968), McConnelland Servaes (1990), Cho (1998), Mudambi and Nicosia (1998), Aitkenand Harrison (1999), Wan (1999), Kuznetsov and Muravyev (2001), Park

(2001), Andersson et al (2004), Grant and Kirchmaier (2004), Jiang (2004), Al-Shiab and Abu-Tapanjeh (2005), Fishman et al (2005), Hu

and Zhou (2006), Lisboa and Esperanca (2006), Alonso-Bonis and

Andrés-Alonso (2007), Aydin et al (2007), Farooque et al (2007),

Kapopoulos and Lazaretou (2007), Szép (2007), Yasar and Paul (2007),

Hake (2008), Laurenceson and Qin (2008), Lee (2008), Abidin et al (2009), Bilyk (2009), Burker et al (2009), Cornett et al (2009), Dinga et

al (2009), Ghahroudi (2009), Lee and Chuang (2009), Drakos and Bekiris

(2010), Hess et al (2010), Gelubcke (2011), Priya and Shanmughan (2011), and Pervan et al (2012).

McConnell and Servaes (1990) investigate the relation between Tobin’s

Q and the structure of equity ownership for a sample of 1,173 firms for

1976, and 1,093 firms for 1986 They find a significant curvilinearrelation between Q and the fraction of common stock owned by corporateinsiders They also found a significant positive relation between Q andthe fraction of shares owned by institutional investors

Cho (1998) uses a cross section of 326 manufacturing firms on Fortune

500 in 1991, and finds a significant relation between insider ownershipand corporate value, where corporate is measured by Tobin’s Q He alsofinds a similar non-monotonic relation between insider ownership andinvestment Therefore, based on OLS analysis, ownership structureaffects investment and corporate value However, when he estimates asimultaneous equations system of ownership structure, investment, andcorporate value, the results show that investments affect firmperformance, which in turn affects ownership structure, but not viceversa

The study of Andersson et al (2004) on 87 Sweden companies indicates

that companies with a dispersed ownership structure are associated withworse performance regarding stock return, ROA, and ROE, but are highlyvalued relating to Tobin’s Q The similar findings from Kapopoulos andLazaretou (2007), who follow the model of Demsetz and Villalonga(2001) to conduct a research on a sample of 175 Greek firms By

comparing OLS estimates to 2SLS estimates, empirical findings indicatethat there exists a linear positive relationship between firm performanceand ownership structure Specifically, the study reveals that managerialshareholdings and important shareholdings positively influence Tobin’sQ; and that higher firm profitability required a less diffused ownership.These findings were contradictory to Demsetz (1983)’s arguments: Asthe number of shareholders increases, the wealth of each will depend less

on the success of the firm But this carries no implication of a resultingreduction in the value of the firm Indeed, profit maximization mayrequire a diffuse ownership structure (Demsetz (1983), p.386) Recently,

Pervan et al (2012) support these arguments of Demsetz (1983) when

analyzing the relationship between firm performance (ROA) and firmownership (ownership concentration and type) on a sample of 1430 listedCroatian firms for the period 2003-2010 Their empirical findingsindicate that ownership concentration is negatively related withperformance, for instance, firms with dispersed ownership perform betterthan firms with concentrated ownership In addition, they find thatforeign controlled firms achieve higher level of profitability thandomestically controlled firms

Another view on the impact of ownership structure on firm performance

from Fishman et al (2005), who examine this relationship by using data

of 50 companies listed on Australian Stock Exchange over the period2002-2003, based on the models by Demsetz and Villalonga (2001) andWelch (2003) in Australian environment The OLS results illustrate that

no relationship between ownership structure and Tobin’s Q exists, whilethe 2SLS results indicate that Tobin’s Q has a significantly negative

17

impact on the level of managerial ownership To further investigate thisrelationship, a three-equation model is developed and the findingsindicate that managerial ownership impacts negatively on firmperformance Al-Shiab and Abu-Tapanjeh (2005) have the similarconclusion when examining the impact of ownership structure on firmperformance in 50 of the largest Jordanian Industrial Companies listed onAmman Stock Exchange, over the period 1996-2002 Controlling forcapital structure, firm size, and sales growth, the OLS regression resultsshow a non-linear and significant effect of ownership concentration onmarket-based measure (namely, market-to-book-value of equity), butnegative effect on accounting-based measure, (namely ROA) Thenegative correlation between ownership and corporate performance issimilarly found in the study of Lee and Chuang (2009) Examining onthe ten-year (1994-2003) panel data of 569 Taiwanese listed companies,the empirical findings reveal that the ratio of mortgaged/pledged shares ofdirectors and supervisors and firm performance constitute a significantlynegative correlation, and that government institutional ownership andincorporated companies’ ownership are found to have a significantnegative impact on Tobin’s Q Besides applying three statisticalmethods, such as OLS, fixed effects, and random effects methods, Leeand Chuang (2009) employ the F-test, Lagrange Multiplier test andHausman test to determine the best method amongst the three However,Lee and Chuang (2009) do not discuss whether the variables in theirstudy are endogenously related.

A different approach regarding the impact of managerial ownership on

firm performance is suggested by Abidin et al (2009), who examine the

association between board structure and corporate performance, whereperformance is defined as the value added efficiency of the firm’sphysical and intellectual resources rather than the more commonly usedTobin’s Q or ROA Testing on a randomly selected sample of 75companies listed on Bursa Malaysia, the study finds that boardcomposition and board size have a positive impact on firm performance,while the effects of directors’ ownership and Chief Executive Officersduality on the value added efficiency of firm’s total resource are notestablished The managerial ownership – corporate value relationship isalso discovered by Drakos and Bekiris (2010) Using a panel data of 146firms quoted on the Athens Stock Exchange for the period 2000-2004 andbased on the model developed by Demsetz and Villalonga (2001), Drakosand Bekiris (2010) estimate a system of two regression equations, inwhich managerial ownership and corporate performance (Tobin’s Q) areendogenous variables The main finding of the study indicated that whenmanagerial ownership is treated as endogenous, this leads to a positiveimpact on corporate value.

Government always plays an important and special role in its economy.Accordingly, state-owned holdings have important and special impact onthe firm performance This is the reason why the relationship betweenstate ownership and firm performance is investigated in the research of

Cornett et al (2009) and Hess et al (2010) Cornett et al (2009) examine

the performance differences between privately-owned and state-ownedbanks in 16 Far East countries from 1989 through 2004 to see howgovernment ownership affects bank performance The study uncoversthat state-owned banks operate less profitably, hold less core capital, and

have greater credit risk than privately-owned banks prior to 2001.However, in the post-Asian financial crisis period 2001-2004, the state-

owned banks’ performance is improved Hess et al (2010) investigate the

effects of state ownership structure on firm values on Chinese listed firmsfor the period 2000-2004 The authors use both OLS and 2SLS analysis,which treats ownership concentration as endogenous The study confirmsthe U-shaped of the state ownership-performance relationship Thisimplies that firms dominated by the various state players continue tomaintain a greater respect by the market and outperform those with lowerlevels of state block-holdings However, at lower levels of stateownership, the firm value declines up to a reflection point ofapproximately 35 per cent beyond which the positive affects of statedominance take effect

Regarding the relationship between foreign ownership and firmperformance, Aitken and Harrison (1999) use a panel of more than 4,000Venezuelan plants between 1976 and 1989 and find that increases inforeign equity participation are correlated with increases in productivityfor recipient plants with less than 50 employees Another finding fromtheir study is that foreign investment negatively affects the productivity

of domestically owned plants Aydin et al (2007) apply t-test statistics to

examine over all quoted firms on the Istanbul Stock Exchange for theperiod 2003-2004 The results reveal that firms with foreign ownershipperform better than domestic owned ones in respect to ROAs The similarresults are found in the research of Yasar and Paul (2007), when theyexamine the performance effects of foreign ownership, from theperspective of firms in five transition economies, namely Poland,

Moldova, Tajikistan, Uzbekistan, and Kyrgyz Republic in 2002 Theempirical findings indicate that foreign owned firms perform better thandomestic counterparts, namely higher productivity, capital intensity,export and import shares, employment, and wages are found in firms withforeign ownership Bilyk (2009) investigates the effects of foreignownership on performance of 264 Ukrainian manufacturing companies in2002-2006 Bilyk (2009) divides foreign ownership into two types,namely foreign offshore and foreign non-offshore firms When fixed-effects regression is estimated, no consistent relationship between foreignownership and performance is found Nonetheless, when more detailedclassifications of foreign ownership are used, empirical results providethe evidence that there is a positive effect of investment coming fromforeign investors on firms’ profitability in Ukraine Recently, Priya andShanmughan (2011) use ownership structure of 425 Indian firms over theperiod 2003-2009 to examine whether differences in ownership structurecan explain their differences in performance The empirical findingssuggest foreign ownership relates positively to firm performancemeasured by Tobin’s Q and ROA These above-mentioned findings areopposite to those reported by Mihai (2012), who finds that there is nosignificant link between firm performance and the existence of foreigncapital when investigating the foreign ownership-performance relation on

a sample of 63 Romanian listed companies for the period 2000-2010,where ROA and ROE are used as measures of financial and economicperformance This indicates that not all cases prove the theory of foreigninvestment by Hill (2009) FDI recipient economy can benefit capitalsupplies, modern technologies, and management resources, which boostthe host country’s economic growth (Hill (2009))

Trying to explain about discrepancies amongst researches regarding theownership-performance relation, Hu and Izumida (2008) believe that theimportant reason was the realities of corporate governance environments

in which firms were embedded The disagreement was also owing tomodel specification and estimating technique applied Moreover, variablemeasurements and data issues could explain partially the differences

3 METHODOLOGY

3.1 Data

The data used for this thesis was collected from the website

www.cophieu68.com1 It consisted of a representative sample selectedfrom the total number of firms listed on the Hochiminh Stock Exchangefor the period 2007-2010 The data was collected during a period of fouryears to avoid the bias when running regression on data collected for asingle year The firms employed in this study must meet the followingcriteria: The firms must be quoted on Hochiminh Stock Exchange at leastone year before the year of analysis; the proportion of foreign ownership

in the firms must be more than 0%; and the firms must be in operation inthe course of the thesis With these criteria, the dataset included 83 firmsfor the year of 2007, 125 for 2008, 151 for 2009, and 208 for 2010,obtaining a total sample of 567 firm-years

3.2 The model

The past literature reveals that most of previous researches develop theirown models based on the model put forward by Demsetz and Villalonga(2001) Similarly, the thesis applies the model built by Drakos andBekiris (2010), which is the most recently modified version of Demsetzand Villalonga’s model The model is presented as follows:

Q = β0 + β1foreign_own + β2ln_assets + β3debt_asset + ε

1 This website was selected to collect data for the thesis based on its sufficient and reliable information for the model in this thesis.

for the period of 2007-2010

Annual Tobin’s Q is calculated as the sum of totalyear-end book value of debt and total year-end marketvalue of equity, divided by total year-end book value

of assets

foreign_own is Foreign Ownership, measured by percentage of

shares owned by foreigner in the firms

ln_assets is Firm Size, calculated by logarithm of total assetsdebt_asset is Leverage, followed the formula of Total year-end

debt over total year-end assets

Firm Performance

There are two common measures of firm performance One is accountingmeasure and the other is market-value measure Hirschey and Wichern(1984) examine the relationship between accounting and market-valuemeasures on a 386-firm sample taken from the 1977 Fortune 500; andthey suggest both accounting and market-value measures can be used asunique but imperfect indicators of profitability Sauaia and Castro (2002)use the Multinational Management Game to test whether Tobin’s Q is agood indicator of a company’s performance Comparing to the otherseven past performance indicators in the same (namely market-share;return on sales; asset turnover; inventory turnover; return on assets; debt

to total assets; and return on equity), they find that companies thatachieved a better performance, are associated with a higher value of theTobin’s Q.

Existing researches commonly employs two measures of firmperformance, namely Tobin’s Q and accounting profit rates These twomeasures differ in some respects Tobin’s Q is a forward-lookingperformance measure, which takes investor psychology into account Incontrast, accounting profit rate is backward-looking and unaffected bymarket psychology While accounting profit rates are affected byaccounting practices, such as the different methods applied to valuations

of tangible and intangible capital; Tobin’s Q distorts performancecomparisons of firms, because the numerator of Tobin’s Q (market value)partly reflects a firm’s intangible assets, yet the denominator of Tobin’s Q(replacement cost2) includes the firm’s tangible assets only (Demsetz andVillalonga (2001); Hu and Izumida (2008))

Since most economists have a better understanding of market constraintsthan of accounting constraints, they are in favor of Tobin’s Q (Demsetzand Villalonga (2001)) Moreover, accounting data is not thought aseligible variables in measuring a firm’s performance in countries wherethe accounting standards are imperfect, especially in developing countries(Hu and Izumida (2008)) Accordingly, like most of previous researcheswhere the simple Tobin’s Q formula is commonly used, which iscalculated by summing up market value of equity and book value of total

2 Replacement cost of assets is the book value of a firm’s assets with inflation adjustments (Cho (1998), p.107).

debt, and divided it by book value of total assets, Tobin’s Q ratio ischosen in the thesis as measure of firm performance.

Foreign Ownership

The measures of ownership structure have been defined in diverse ways

in previous empirical studies in accordance with individual researchpurpose, but in general, they all express the fraction of shares held by afirm’s most significant shareholders

In order to determine whether exist the interrelation between firm

performance and ownership structure, Himmelberg et al (1999) employ

the fraction of common equity holdings of top-level managers to measureownership concentration Cho (1998) uses the percentage of insiderownership as ownership variable in his model

The proportions of shares held by the State and held by private block

holders are chosen by Cornett et al (2009) to explore the impact of state

ownership on performance by comparing the performance of owned banks with that of state-owned banks in 16 Far East countries

privately-Hess et al (2010) use these ownership measures to examine the effect of

the dominance of state and private block-holders on firm performance ofChinese listed firm during 2000-2004

Demsetz and Lehn (1985) measure concentration of ownership by thefraction owned by the five and twenty largest shareholders Demsetz andVillalonga (2001) continue to use percentage of equity owned by the five

largest shareholders as the measure of ownership structure This variable

is employed again in the researches by Welch (2003) and Fishman et al

(2005) when applying the model by Demsetz and Villalonga (2001) toexamine the ownership-firm performance relationship on Australian

Stock Exchange; and then is used in the study by Hess et al (2010) to

explore the link between state-dominant and non-state-dominantownership and firm performance of Chinese listed companies Still based

on the model by Demsetz and Villalonga (2001), yet Drakos and Bekiris(2010) replace the ownership measure with the percentage of sharesowned by directors of the Board, in order to correspond to theinvestigation of the relationship between managerial ownership andcorporate performance on Athens Stock Exchange

The current thesis employs the percentage of shares owned by foreigner

in a firm as measure of foreign ownership This is consistent with theproxy of percentage of equity held by foreign investors used the previousstudies (Yasar and Paul (2007); Bilyk (2009); and Mihai (2012))

Firm Size

Most of past researches include firm size in their models as control

variable Himmelberg et al (1999) use logarithm of firm sales to measure

firm size, and argue that firm size has an ambiguous effect Himmelberg

et al (1999) argue that on the one hand, larger firms have more

difficulties in controlling all their activities, which can lead to agencyproblems and consequently to a decline in the performance; on the other

27