Domestic and Foreign Institutional Investor

Behavior in China

Ningyue LiuUniversity College Dublin ∗

Don BredinUniversity College Dublin†

Liming WangUniversity College Dublin‡

Zhihong YiRenmin University of China§

Abstract

This paper compares the investment characteristics between foreign funds operatingunder Qualified Foreign Institutional Investor (QFII) in China and domestic Chinesefunds and analyzes the firm-level drivers that influence their allocation choices. Theanalysis reveals that foreign funds have a preference for a range of sectors such astransportation, metals & non-metals, and machinery, as opposed to industries with arequirement for local knowledge. The portfolios of domestic Chinese funds are dis-tributed more evenly, compared to foreign funds. The comparative analysis indicatesthat foreign funds invest in firms that are significantly different from those favoredby domestic funds in terms of size, profit and compensation of management. Finally,we find that when making investment decisions, foreign funds tend to rely on somecorporate governance indicators which is not consistent with the results from previousstudies examining developed markets. In particular, foreign funds have a preferencefor firms with a high percentage of state-owned shares, while the reverse is the casefor domestic funds. These empirical findings highlight the differences between QFIIsand domestics funds investment preferences, and will be of value to policy makers inemerging markets, and China in particular, in gauging the important drivers of foreigninvestment. At the emerging market level our results highlight the sensitivity of foreigninstitutional investment to market cycles. For China in particular, our results indicatethat the Chinese ownership structure is an important attraction for foreign investors.

Keywords: Qualified Foreign Institutional Investor, Corporate Governance, ChinaStock Market.JEL Classification: G11; G34.

∗Graduate School of Business, University College Dublin, Blackrock, Dublin, Ireland. E-mail:ningyue.liu@ucd.ie.

†Graduate School of Business, University College Dublin, Blackrock, Dublin, Ireland. E-mail:don.bredin@ucd.ie.

‡Irish Institute for Chinese Studies, University College Dublin, Belfield, Dublin 4, Ireland. E-mail:liming.wang@ucd.ie.

§School of Business, Renmin University of China, Beijing, China. E-mail: zhyi7620@163.com.

1 Introduction

Although foreign direct investment (FDI) has had a dramatic impact on economic devel-

opment in China, it is only since 2003 that the Chinese government has permitted foreign

institutional investors (FIIs) to directly invest in Chinese securities market (Cheng and

Kwan, 2000; Zhang, 2001). Since China became a member of the World Trade Organi-

zation (WTO) in December 2001, it has implemented numerous measures to liberalize its

economy and improve its investment environment. One of the most significant measures

has been the Qualified Foreign Institutional Investor (QFII) scheme, which is designed to

allow the largest overseas institutions access to China’s debt and equities markets. The

QFII scheme represents a significant departure from China’s traditional approach of strict

capital controls. As of 30th September 2010, 93 QFIIs had been approved by the China Se-

curities Regulatory Commission (CSRC).1 The total investment quota of QFIIs has grown

from 425 million US dollars at the beginning of the scheme in 2003 to 19 billion US dollars

by the end of September 2010 (see figure 1). In spite of the exceptional growth, there is

limited research on how QFIIs determine the allocations across different listed companies

and the factors influencing their investment behavior.

In this paper, we provide an introduction to the QFII scheme in China, examine the

firm-level characteristics of stocks that fund managers invest in, and also investigate whether

such stock preferences vary across foreign and domestic fund managers. Specifically, we aim

to address three questions. Firstly, what are foreign funds’ industrial preferences? Secondly,

how do firm-level characteristics compare to domestic funds? Finally, we examine what firm

attributes impact the security holdings of foreign and domestic funds.

All previous research to date examining the preference of foreign investors has focused

on either macro level analysis or firm level analysis for developed markets. For exam-

ple, La Porta et al. (1997, 1998, 2000) find that stronger investor protection laws, high

enforcement and high quality accounting disclosures have a positive impact on market de-

velopment. Chan, Covrig and Ng (2005) conclude that economic development, capital

controls, and withholding tax variables have significant effects on foreign investors’ invest-

ment allocation. While, Falkenstein (1996), Kang and Stulz (1997) and Dahlquist and

Robertsson (2001) investigate the preference of foreign institutional investors in firm level

1Source is the Investment Quota Approval Form of QFIIs issued by State Administration of Foreign

Exchange (SAFE), 10th November 2010.

1

analysis using US, Japan and Swedish market data respectively. Given the short history of

institutional investors in emerging stock markets and the difficulty of obtaining data, there

is a considerable dearth of empirical evidence with regard to emerging markets. Moreover,

Sercu and Vanpee (2007) find that the equity home bias is lower in developed markets and

higher in emerging markets. Since developed markets have higher standards of information

disclosure, we would expect that the information asymmetry between foreign and domestic

investors in these markets would be less severe. However, foreign investors may face more

severe information asymmetry in developing markets due to imperfect market, cultural, and

political differences. It is therefore important to examine whether foreign funds exhibit a

different preference compared to domestic investors in emerging markets. Our study sheds

lights on the stock preference of foreign investors in emerging markets with information

asymmetry and imperfect market.

The rate of economic growth in China in recent years has been dramatic, with an average

annual growth rate of around 10% from 1978 to 2010.2 China’s gross domestic product

(GDP) reached 5.7 trillion US dollars in 2009. Thus, China has become the world’s third

largest economy, only second to the Eurozone and the US.3 The size and growth of Chinese

equity markets is equally dramatic, growing from 104 billion Chinese yuan (CNY) in 1992

to more than 26 trillion CNY by the end of 2009.4 The number of listed companies has

grown from 14 in 1991 to 2,063 by the end of 2010, while the number of trading accounts

stands at 133 million.5

Given the major differences between China and other developed markets, e.g. in terms

of culture and language, we would expect significant information asymmetries to exist.

Grinblatt and Keloharju (2000) cite these type of asymmetries as having implications for

portfolio decisions. Turning to the specific case of foreign investment behavior in China,

Naughton (2007) concludes that insider control and manipulation are particular charac-

teristics of the Chinese stock market. In addition to insider control, weak disclosure and

regulation, along with continued policy-driven government investment are further character-

istics of the Chinese stock market (Naughton, 2007). These aspects of the Chinese market

stand in stark contrast to the case of developed markets. However, these considerations

2Source is the National Bureau of Statistics of China.3International Monetary Fund, World Economic Outlook Database, October 2010.4China’s two domestic stock exchanges, Shanghai Stock Exchange and Shenzhen Stock Exchange, were

established in December 1990 and July 1991, respectively.5Source is the CSRC from 26th January 2011.

2

appear to count for very little when we view the dramatic growth in foreign institutional

investment in China, as highlighted in figure 1.6 Clearly investors continue to see potential

for further economic growth in China.

Given the dramatic growth and the specific features of the Chinese market outlined

in the previous paragraph, we examine the impact of a comprehensive range of firm-level

characteristics, financial indicators and corporate governance indicators, on the investment

decisions of foreign funds, with a comparative analysis to domestic funds. Although financial

characteristics have been investigated by previous studies examining developed markets, our

comprehensive range of indicators offers a more robust analysis of the Chinese market. In

particular, a number of additional financial aspects are included, such as operating ability,

which is an important indicator used to evaluate the financial condition of a company.

The role of corporate governance on the investment decisions of foreign (and domestic)

funds is also examined. McKinsey and Company (2002) have highlighted the importance of

governance to the investment decisions in the global market. This study also highlights that

corporate governance considerations dominate any other issues when it comes to investment

decisions in East Asia.

Our results will be of value to policy makers in both developed and developing markets

in gauging the important drivers of foreign investment. Given the increasing significance

of foreign financing and the fact that access to foreign capital is uneven across firms, it

is important to fully understand the factors that influence investors’ behavior. Moreover,

Huang and Shiu (2006) find that stocks with high foreign ownership outperform stocks with

low foreign ownership. Our findings will be particularly relevant to policy makers and firms

in creating an environment conductive to foreign investment.

The paper adopts annual Chinese stock market data for the period 2003 to 2009 for both

foreign and domestic funds. Our results indicate that foreign institutional investors prefer

sectors like transportation, metals & non-metals, and machinery as opposed to industries

requiring local knowledge, e.g. real estate, construction, media & culture. The portfolios

of domestic funds are distributed more evenly across sectors compared to foreign funds.

This paper examines the impact of a comprehensive range of firm-level characteristics,

6Stulz (2005) indicates that home bias can be caused by the agency problem of state-ruler discretion,

meaning that state rulers can expropriate investors by regulations and taxes that are used for the state rulers

own benefit. The high percentage of state-owned shares in Chinese listed firms could be another important

factor in portfolio choices of foreign investors in China.

3

financial indicators and corporate governance indicators, on the investment of foreign and

domestic funds. Although some characteristics have been investigated by previous studies,

our comprehensive range of indicators offer a more robust analysis. In particular, a number

of new financial measures are included, such as operating ability and compensation of

management, which are important indicators used to evaluate the financial condition of

a company. We find some characteristics of the companies foreign investors invest in,

such as size, profit and compensation of management are significantly different from those

companies preferred by domestic funds.

Finally, we find when making investment decisions, foreign funds tend to have a prefer-

ence for certain corporate governance indicators, e.g. ownership structure and concentra-

tion, which is inconsistent with prior research in developed markets (Kang and Stulz, 1997;

Dahlquist and Robertsson, 2001; Aggarwal, Klapper and Wysocki, 2005). In particular,

foreign funds prefer firms with a high percentage of state-owned shares, which is one of the

characteristics of Chinese companies. In contrast, however, domestic funds place consider-

able emphasize on financial characteristics, e.g. size and return on shareholders as well as

corporate governance when they make investment decisions. Although ownership structure

is one of the strongest determinants of the investment decisions of both foreign and domes-

tic funds, their preferences represent investment strategies that represent two very different

requirements. While foreign investors want to access priority industries like manufactur-

ing, with the implications of high state ownership, domestic investors have a preference for

investing in industries that have lower state involvement and greater potential for external

investment derived growth. These empirical findings highlight the differences between for-

eign and domestic funds investment preferences. In particular our results have important

implications for policy makers aiming to attract foreign investors to invest in emerging

markets.

The outline of the rest of the paper is as follows. In section 2, a brief review of the

literature on foreign institutional investors’ investment behavior, before introducing the case

of China and the related research and research hypotheses development. In section 3 we

discuss the data and indicators used in this paper. In section 4, we present the empirical

results including industrial distribution, comparative analysis of financial indicators and

corporate governance indicators as well as the impact factors on the investment decisions

of foreign and domestic funds. Finally section 5 provides some concluding remarks.

4

2 Literature Review and Research Hypotheses

2.1 Preference of Foreign Investors

Frenkel and Poterba (1991), Cooper and Kaplanis (1994) and Tesar and Werner (1995) doc-

ument that although the barriers to international investment have declined dramatically,

foreign ownership of shares is still extremely limited and much smaller than one would ex-

pect. A large number of papers have examined the home bias issue and several explanations

have been suggested. Typically, researchers compare the aggregated holdings of investors

in foreign markets with their domestic holdings (Lewis, 1999). Kang and Stulz (1997) an-

alyze this issue from a different perspective and study the shareholdings of foreigners in

individual firms in a specific market. They adopt data for Japan from 1975 to 1991 and

find that foreign investors hold disproportionately more shares of firms in manufacturing

industries, large firms, and firms with good accounting performance, low unsystematic risk,

and low leverage.

Dahlquist and Robertsson (2001) analyze the determinants of foreign ownership in

Swedish firms and find that foreigners have a preference for large firms, firms paying low

dividends, and firms with large cash positions on their balance sheets. Foreign investors

also tend to underweight firms with a dominant owner.7 Covirg, Lau and Ng (2006) investi-

gate a range of stock preferences of domestic and foreign fund managers from 11 developed

countries and conclude that foreign fund managers have less information about the domes-

tic stocks than their domestic counterparts. They find that ownership by foreign funds

is related to the size of foreign sales, index membership and stocks with foreign listing.

Jiang and Kim V (2004) find that foreign investors in Japan tend to avoid stocks with high

cross-corporate holdings. The authors suggest that foreign institutional investors are likely

to be efficient processors of public information and are attracted to Japanese firms with

low information asymmetry.

Fast growing emerging markets are attractive to foreign institutional investors and port-

folio investments by foreign institutional investors were the most important source of capital

for emerging markets, in particular during 1990s (Frenkel and Menkhoff, 2004). Aggarwal,

Klapper and Wysocki (2005) look at the portfolio holdings of 576 US mutual funds invested

7Dahlquist and Robertsson (2001) find there is a negative relationship between foreign institutional

ownership and ownership concentration, implying that foreigners avoid companies with a dominant owner.

5

in emerging markets as of February 2002 and analyze both country- and firm-level disclo-

sure and institutional policies that influence mutual funds’ allocation choices. At the firm

level, US funds have a preference for firms that adopt discretionary policies such as greater

accounting transparency and the issuance of an American Depositary Receipt (ADR). Vo

(2010) employs the data from Vietnam stock markets and finds that foreign institutional

investors avoid firms with dominant shareholders and favor to invest in firms where they

have less informational asymmetry.

2.2 QFII: A Formal Classification and Recent Developments

QFII is defined as “overseas fund management institutions, insurance companies, securities

companies and other assets management institutions which have been approved by CSRC

to invest in China’s securities market and granted investment quota by SAFE”. 8 Term 18

of the Provisional Measure states that QFIIs can invest in A Shares, treasuries, convertible

bonds and corporate bonds listed in China’s stock exchanges and other financial instruments

as approved by CSRC.9 The requirement for QFII’s qualification states that fund managers

must hold assets in excess of 10 billion US dollars during the latest accounting year and have

operated for over 5 years.10 Hence only large foreign institutional investors are qualified to

apply as QFIIs. Two important conditions of the regulations influencing the investment of

QFIIs arise from Term 20 (1), which states that “shares held by each QFII in any one listed

company should not exceed 10% of the total outstanding shares of the company” and Term

20 (2), which further states that “total shares held by all QFIIs in one listed company should

not exceed 20% of the total outstanding shares of the company”.11 The first three QFIIs

approved in June 2003 were UBS, Nomura Securities, and Citigroup Global Markets. They

8This definition can be found in “Provisional Measures on Administration of Domestic Securities Invest-

ments of Qualified Foreign Institutional Investors” issued by CSRC on 5th November 2002, page 1.9A Shares in Shanghai and Shenzhen stock exchanges refers to those that are traded in CNY. Currently

only mainlanders and QFIIs are allowed to trade A Shares. Some shares in the two mainland Chinese stock

exchanges, known as B Shares, are traded in foreign currencies. In the past, only non Chinese were allowed

to trade B Shares. Starting from March 2001, mainlanders can trade B Shares as well. However, they must

trade with legal foreign currency accounts. QFIIs are not permitted to invest in B Shares.10Insurance companies must have assets in excess of 10 billion US dollars during the latest accounting

year and have operated for over 30 years. Commercial banks must be ranked in the world top 100 banks in

terms of assets and have assets in excess of 10 billion US dollars during the latest accounting year as well.11Two conditions can be found in “Provisional Measures on Administration of Domestic Securities Invest-

ments of Qualified Foreign Institutional Investors” issued by CSRC on 5th November 2002, page 5.

6

were soon followed by Morgan Stanley International and Goldman Sachs. QFIIs are all

large international institutions from major developed countries, e.g. USA, UK, Germany,

France, Japan, Canada, Netherlands and Switzerland. The details of QFIIs’ approved date

and investment quota are reported in table 1.

CSRC, People’s Bank of China (PBC) and SAFE jointly issued new regulations, entitled

“Measures for the Administration of Investment in Domestic Securities by Qualified Foreign

Institutional Investors” (the “New QFII Rules”) on 24th August 2006. The new QFII rules

supersede the original ones that were in place since 2002 and took effect from the 1st

September 2006. The qualifying criteria in terms of assets under management for QFII

applicants that are fund management institutions has been reduced from 10 billion US

dollars to 5 billion US dollars during the latest accounting year. This will enable more

fund management companies to apply for QFII approval.12 In addition, while the old

rules did not specifically cater for other categories of institutional investor, it is now stated

that they are subject to the qualifying criteria of having been established for at least

five years and having assets under management of at least 5 billion US dollars in the

most recent accounting year.13 Furthermore, on 29th September 2009, SAFE released the

“the Provisions on Foreign Exchange Administration of Domestic Securities Investment by

Qualified Foreign Institutional Investors”. The maximum accumulated investment quota of

one single QFII has been increased to 1 billion US dollars from 800 million US dollars. The

New QFII Rules contain significant improvements from the original rules, particularly for

fund management companies wishing to invest in the Chinese domestic securities market

through the QFII scheme. The positive changes under the New QFII Rules open the way

for the further development of investment fund products in China.

2.3 Empirical Evidence on QFII

The QFII system was introduced in Taiwan in the late 1980s and begun in 1991, when Tai-

wan’s stock market was particularly popular with foreign investors (Dean, 2003). Research

on QFIIs in Taiwan is divided into two main categories. One area of research has examined

the impact of QFIIs on Taiwan’s stock market and local companies’ performance (Lin and

12The qualifying criteria for insurance companies is changed from at least 10 billion US dollars to 5 billion

US dollars in assets held in the most recent accounting year.13Other categories of institutional investors include pension funds, charitable funds, donation funds, trust

companies and government investment companies.

7

Chen, 2006; Huang and Shiu, 2006). The other investigates the behavior of QFIIs, e.g.

the extent of momentum and herd behavior, in Taiwan (Lin and Swanson, 2003; Lu, Fang,

Wong and Wang, 2009; Lai, Lou and Shiu, 2008). Lin and Shiu (2003) investigate foreign

ownership in Taiwan stock market form 1996 to 2000 and find foreign investors appear

to favor large firms and low book-to-market stocks. Analytical results show that foreign

investors have a preference for firms with high export ratios, which may be as a result of

brand recognition given their foreign sales. Furthermore, Korea announced guidelines for

the limited opening up of its equities market to foreign investment in 1991. Using Korean

data, Choe, Kho and Stulz (2005) show that foreign institutional investors pay more than

their domestic counterpart when they buy and receive less when they sell for medium and

large trades which indicate that domestic individual investors have an edge over foreign

investors.

The extent of research on QFII in China is limited given the difficulties in accessing

data and the short time period of the QFII scheme. Most of the research to date provides

an introductory analysis, e.g. Yeo (2003) introduces QFII scheme in China and compares

it with other Asian QFII markets. Ting, Yen and Chiu (2008) examine the relationships

between audit opinions and the default probability within the Chinese stock market, and

find audit opinions begin providing signals of potential default risk only after QFIIs entered

the market. Chen and Yu (2003) investigate market reactions around the announcement of

the QFII scheme and find no significant abnormal returns in market indices in the short-

term leading up to the announcement, negative abnormal returns in short-term following

the announcement, and no significant abnormal returns in long-term. Most recently, Huang

and Zhu (2011) use the data on QFII in China around the Split Share Structure Reform

to investigate the different roles of QFIIs and domestic funds in the firms’ voting process.

They conclude that foreign institutional investors are subject to less political pressure than

domestic counterparts. They also find that foreign and domestic funds differ in portfolio

selection.

2.4 Research Hypotheses

Drawing on the literature to date and the research questions, we develop a number of

research hypotheses in this paper.

Hypothesis 1: Foreign institutional investors have a preference for manufacturing industry

8

in China.

Firstly, what are foreign funds’ industrial preferences? Merton (1987) argues that in-

vestors prefer securities they know about. In other words, foreign investors invest less in

the industry requiring local knowledge, e.g. real estate. Furthermore, is it the case that

foreign funds have a preference for manufacturing industries which are consistent with the

preference of foreign institutional investors in developed markets (see Kang and Stulz, 1997)

or are there special features associated with the Chinese market? Accordingly, we propose

Hypothesis 1.

Hypothesis 2: Foreign and domestic funds’ differences exist in relation to financial and

corporate governance characteristics.

Secondly, how do firm-level characteristics of foreign firms compare to domestic funds?

The question of whether foreign and domestic institutional investors have the same in-

vestment behavior is increasingly controversial. One argument is that foreign institutional

investors are more sophisticated investors than their domestic counterparts (Grinblatt and

Keloharju, 2000). An alternative is that they are regarded as equally sophisticated but

not as well informed (Covirg, Lau and Ng, 2006). Therefore, they will invest in companies

with different characteristics compared to those that domestic funds invest in. Conversely,

Chang (2010) highlights that foreign investors in emerging markets might be expected to

suffer from an informational disadvantage given a lack of local knowledge and contacts.

For example, foreign investors need to take account of China’s unique characteristics, such

as insider control and the influential role of government policy (Naughton, 2007).14 This

information asymmetry, combined with China’s unique characteristics may motivate for-

eign institutional investors to choose the “free rider strategy” and follow the investment

decisions of domestic funds. Formal empirical evidence on this issue is also mixed. Grin-

blatt and Keloharju (2000) using Finnish data, Choe, Kho and Stulz (2005) using Korean

data, and Covirg, Lau and Ng (2006) using a data set of 11 developed countries find sig-

nificant differences in the investment behavior of foreign and domestic investors. While,

Kang and Stulz (1997) using Japanese data find no difference in the performance of foreign

and domestic investors. Chang (2010) using data from Taiwan Stock Exchange find that

14Naughton (2007) indicates that Chinese stock market is heavily influenced by changes in government

policy, along with limitations on individual firm information and shareholder control. Chinese market

fluctuations are better explained as reactions to government policy changes rather than as reaction to

changes in underlying fundamentals of individual companies.

9

foreign and local institutional traders herd around expatriates for information reasons. Is

it the case that foreign funds follow the investment decisions of domestic funds because of

information asymmetry and China’s unique characteristics or do they invest in companies

with different characteristics compared to those that domestic funds invest in because of

more sophisticated investment skills? Thus we propose Hypothesis 2.

In order to compare the different firm-level characteristics of firms QFIIs invest in and

those favored by domestic funds, we divide the whole sample into two groups: one group

includes the firms that have at least one QFII holding their equity, and another group

includes the firms without QFII presence. Then, we compare the firm-level characteristics

between these two groups. Additionally, the firm-level characteristics are divided into two

groups: financial indicators and corporate governance indicators which will be discussed

detailedly in the section 3. According to the above analysis, we could expect that there

are significant differences on both financial indicators and corporate governance indicators

between the two groups if foreign funds don’t follow the investment decisions of domestic

funds even though they have to face information asymmetry and are not familiar with the

unique features of Chinese stock markets.

Hypothesis 3: Corporate governance (financial characteristics) has a significant impact

on the investment of foreign (domestic) institutional investors.

Finally, we examine what firm attributes impact the security holdings of foreign and

domestic funds. Most of previous studies emphasize the impact of financial characteristics,

e.g. size, on the share holdings of institutional investors (Dahlquist and Robertsson, 2001).

However, Leuz, Lins and Warnock (2009) believe that corporate governance, e.g. owner-

ship structure, is another group of factors that are considered when investment decisions

are made. We examine whether this is also the case for foreign institutional investors in

China. Given the potential long-term investment benefits, we would expect to find that

foreign investors (representing sophisticated investors) are more likely to consider the cor-

porate governance characteristics of firms. Given the short history of the local markets, the

trading experiences and investment sophistication of the Chinese investors are unlikely to

be comparable with those investors from developed markets (Ng and Wu, 2007). We could

expect that Chinese domestic institutional investors (representing relatively less sophisti-

cated investors) might have a preference for financial characteristics, given the short-term

capital gain. Specifically, domestic fund managers may put more emphasis on firm’s finan-

10

cial characteristics, e.g. return on shareholders which measured by earning per shares and

price to earning ratio in this paper. For these reasons, we propose Hypothesis 3.

Hypothesis 3 predicts the differences in the firm characteristics that explain the “abnor-

mal” holdings between foreign and domestic investors. “Abnormal” holdings are measured

by the percentage relative spread which are the over and under-investment of funds by cal-

culating the difference between the firms’ allocated weight and the Chinese market weight

for each firm in this paper. According to the above analysis, we could expect that corporate

governance (financial) indicators significantly explain the “abnormal” holdings of foreign

(domestic) funds in Chinese stock markets.

3 Data Issues and Key Characteristics

3.1 Data Description

The shareholder data of foreign funds and domestic funds are all sourced from the Wind

database while financial and corporate governance data are from the China Stock Market

Accounting Research (CSMAR) at GTA Research Service Center.15 The data is annual

and covers the period 2003 to 2009.

In order to investigate the behavior of foreign and domestic funds, our sample includes

only those companies whose stocks are held by either foreign funds, domestic funds or both.

Table 2 presents the breakdown of both foreign and domestic funds for individual years.

Over this period, the number of foreign and domestic funds grew dramatically. The average

number of firms foreign and domestic funds invest in each year are 123 and 970 respectively.

3.2 Industry Characteristics

Firms are classified in 13 industrial sectors using codes provided by CSRC. Given foreign

investors traditionally hold more shares of firms in manufacturing industries (Kang and

Stulz, 1997) and the large number of manufacturing companies in China, we subdivide the

manufacturing industry with CSRC industrial code C into 10 sublevel industries. There

15Wind Info is a financial data provider in Mainland China. It provides large-sized financial database to

academic researchers as well as financial organizations, including Merrill Lynch. Poon and Chan (2008) and

Wu, Xu and Yuan (2009) are example of recent studies that applied data from the Wind database.

11

are 22 sectors in total. Dummy variables are employed to capture the industrial effect. For

instance, the dummy variable for the agriculture, forestry, animal husbandry & fishery (A)

industry will be assigned a value of one if a company belongs to that specific industry and

a zero otherwise.

3.3 Financial Characteristics

The firm-level characteristics are divided into two groups: financial indicator group and cor-

porate governance indicator. In each group, we choose a range of indicators to measure the

firm-level characteristics according to previous literature and the characteristics of Chinese

capital market. We explain the detailed financial indicators’ definition in table 3. The first

column is the classification of financial characteristics. In column 2, we choose indicators

to measure the characteristics. For example, total assets and capitalization are adopted to

indicate company size. In the last column, the formula is stated for each indicator.

Drawing on existing studies, we use total assets and capitalization to measure size, cur-

rent ratio and quick ratio to measure short-term liabilities paying capability, assets-liabilities

ratio and long-term liabilities ratio to measure long-term liabilities paying capability, earn-

ings per share and price to earning ratio to measure return on shareholders, and return on

assets and return on equity to measure company profit. In addition to these five firm-level

characteristics, our analysis also includes four additional characteristics, namely financial

risk, cash flow, growth and operating capability. In order to investigate a comprehensive

range of firm-level characteristics, we also add cash flow and operating capability which are

important characteristics for a company but are omitted by previous studies.

3.4 Corporate Governance Characteristics

Table 4 reports the definition of the corporate governance indicators. Since the late 1980’s,

shareholder activism has played a predominate role in improving corporate governance

structures and several studies have investigated this issue (Gillan and Starks, 2000; Kar-

poff, Malatesta and Walkling, 1996). However, there is an endogeneity problem, suggest-

ing that institutions are good at investing in the firms with better corporate governance

structure, leading to the observed relationship between institutional presence and better-

governed firms without any active participation (Chen, Harford and Li, 2007). They focus

12

on independent long-term institutions and find the corporate governance index (G-score)

(Gompers, Ishii and Metrick, 2003) has no effect on institutions’ shareholding decisions.

On the other side, Leuz, Lins and Warnock (2009) study 4,409 firms from 29 countries and

find foreigners invest less in firms due to governance problems.

For ease of comparison, we first choose the same attributes, ownership concentration

as Dahlquist and Robertsson (2001) and management structure as Bushee, Carter and

Gerakos (2009). In addition to these two classifications, our analysis also includes two

more measures, namely ownership structure and compensation of management. Naughton

(2007) indicates that circulating and noncirculating shares is a notable characteristic of the

Chinese equity market. Therefore, we employ the percentage of state-owned shares and the

percentage of circulating shares in the ownership structure subgroup. Finally, compensation

of management is also an important characteristic of a company and is also included here.

4 Empirical Results

In this section, we present evidence of foreign and domestic ownership in China. We begin

by showing the industrial distribution of foreign and domestic funds. Next, we compare

the firm-level investment preferences between foreign and domestic funds in relation to the

financial characteristics and corporate governance. We then extend our examination of the

determinants of foreign ownership by analyzing the firm-level drivers of foreign institutional

investors. Finally, we consider the different drivers of foreign and domestic funds.

4.1 Industrial Distribution

Table 5 reports the relative importance of foreign over domestic fund ownership and the

industry weights in the Chinese market for a total of 22 industries. The numbers are in

percentage terms, 1% means that foreign institutional investors invest 1% more of their

Chinese A-share portfolio in an industry than they would if their investment weights were

those of the Chinese market portfolio. Column one lists the industry sector. In column two

and three (Panel A), we report the number of firms (in %) broken down by industry. In

parentheses, the numbers of firm-year observations for each industry are reported. Column

four and five (Panel B) present the number of shares (in %) and column six and seven

13

(Panel C) present the market value (in %) for each industry from 2003 to 2009.16

Columns two, four and six report foreign institutional investors’ ownership and the do-

mestic institutional investors’ ownership are presented in columns three, five and seven.

Foreign institutional investors hold disproportionately more of the transportation sector

with 4.82%, 4.17% and 5.11% overweighted in these portfolios in Panel A, B and C re-

spectively. The ownership of domestic institutional investors in the transportation sector

is similar with the Chinese market portfolio (0.99% and 0.38% weight in their portfolio in

Panel A and B respectively). Foreign institutional investors are also overweighted in met-

als & non-metals (3.47%, 14.02% and 9.72% overweight in their portfolio in three panels

respectively) which are part of the manufacturing industry. This finding is consistent with

Kang and Stulz (1997). It is also clear that foreign institutional investors hold dispropor-

tionately more investments in the machinery industry. Therefore, our research Hypothesis

1, foreign institutional investors have a preference for manufacturing industry in China, is

supported. In Panel A, foreign institutional investors’ allocations in the finance & insur-

ance sector (0.39%) are generally consistent with Chinese market portfolio which means

that foreign institutional investors invest in most firms in this industry. However, foreign

institutional investors hold disproportionately less in the finance & insurance sector and

the deviation for it is extremely large in Panel B and Panel C.17 Between 2003 and 2009,

foreign institutional investors invested 19.44% and 35.63% less of their Chinese portfolio in

finance & insurance sector than in the market portfolio in Panel B and Panel C respectively.

One possible reason for this phenomenon could be some QFIIs are also strategic investors

in Chinese banks.18 Strategic investor is an alternative facility to invest in the finance &

16In some cases, investors invest in large number of firms in one industry but hold small number and

market value of shares in each firm. Shareholding numbers and market value represent the preference of

investors in each industry. Therefore, we also report the number of shares (in %) and the number of market

value (in %) for each industry.17The companies in finance & insurance sector will not be examined in later sections. This is due to data

limitations. A number of indicators used in this paper do not exist in this industry, e.g. current ratio and

quick ratio, as there are no current assets and current liabilities terms in banks’ balance sheets. Kang and

Stulz (1997) also employ the data from nonfinancial Japanese firms.18Strategic investor is another form of foreign institutional investors with access to China’s market. They

invest A shares through negotiations and initial public offerings rather than trading under the QFII scheme.

Most strategic investors acquire interests in Chinese banks. For example, Temasek, through its wholly-

owned subsidiary Asia Financial Holdings (AFH), acquired a 10% interest in Bank of China in August 2005.

Meanwhile, Temasek is also a member of QFIIs.

14

insurance sector for QFIIs.19 Therefore, foreign investors are more likely to invest more in

finance & insurance sector as a strategic investor in order to play a dominate role as a big

shareholder rather than invest in this sector as a QFII.

Our results indicate that foreign investors have little preference for real estate, in partic-

ular relative to domestic investors. The lower levels of investment in real estate by foreign

institutional investors are consistent with empirical evidence from developed markets that

finds a considerably larger role played by domestic investors given the requirement of local

knowledge (Dahlquist and Robertsson, 2001). Foreign institutional investors also hold fewer

allocations of construction, media & culture, other manufacturing and conglomerates. The

portfolio of domestic funds are distributed more evenly across sectors compared to foreign

funds. Domestic funds invest 0.56%, 6.11% and 6.56% more in metals & non-metals in the

three panels respectively which is consistent with the industrial preferences of foreign insti-

tutional investors. However, they hold disproportionately more of wholesale & retail and

real estate sectors which is inconsistent with foreign institutional investors. The require-

ment for local knowledge is again a likely explanation for such a finding. Finally, domestic

funds also hold disproportionately lower levels of the finance & insurance with a portfolio

under weighting of 17.81% and 44.79% in Panel B and C respectively.

4.2 Foreign and Domestic Funds: An Empirical Comparison

In this section, we present evidence of different investment characteristics between foreign

and domestic funds. In order to take account of outliers, we winsorize the data before

computing the statistics.20 We use the t test and the median test to compare the mean

and median between the two groups of nonfinancial firms. Although the extent of foreign

investment has grown dramatically, the portfolio holding remains relatively small compared

to the domestic counterparts. Foreign and domestic funds’ portfolio holdings consist of 858

and 6,793 firm-year observations respectively. The sample size of domestic funds’ portfolio

19While QFII restricts investments to be 10% or less of the total outstanding shares of the company, the

strategic investor facility requires that investments be at a minimum 10% of the total outstanding shares.20Winsorising or Winsorization is the transformation of statistics by transforming extreme values in

the statistical data, and is named after the engineer-turned-biostatistician Charles P. Winsor (1895-1951).

Winsorised estimators are usually more robust to outliers than their unwinsorised counterparts. All data

below the 1st percentile are set to the 1st percentile, and data above the 99th percentile are set to the 99th

percentile.

15

holdings is almost eight times of that of foreign funds.

The comparative results of financial characteristics are reported in table 6. The two

groups are statistically different in terms of both size and profit. Consistent with prior

research (see Dahlquist and Robertsson, 2001; Aggarwal, Klapper and Wysocki, 2005), we

find that the size of the firms in foreign funds’ portfolios are larger than those in domestic

funds’ portfolios. We use current ratio and quick ratio as a proxy for short-term liabilities

paying capability. The mean of both the current ratio and quick ratio in the two groups are

not significantly different, which is consistent with the findings of Kang and Stulz (1997).21

22 A higher P/E ratio, indicates that investors are paying more for each unit of net income,

with the ratio in China several times higher than in many countries.23 Therefore, it is not a

surprise that the P/E ratio of the firms in foreign funds’ portfolios are higher than those in

domestic funds’ portfolios according to the median test. Another notable difference is that

both ROA and ROE in the foreign funds’ group are larger than those in the domestic funds’

group which is consistent with the result in Kang and Stulz (1997) using ROA, although

inconsistent with the findings in Dahlquist and Robertsson (2001) and Aggarwal, Klapper

and Wysocki (2005) using ROE.

The comparative results of corporate governance characteristics are reported in table

7. The two groups are statistically different in terms of both the percentage of circulating

shares and compensation of management. We find that the percentage of circulating shares

of the firms in foreign funds’ portfolios are larger than those in domestic funds’ portfolios.

This is a significant issue in China. Policy makers aversion to the loss of state owned

assets in China means that there are a large percentage of non-tradable shares in Chinese

stock markets. The implications of greater levels of circulating shares, is that foreign

institutional investors have more freedom to trade the shares. In contrast, however, there

21The current ratio and the quick ratio are calculated as current assets divided by current liabilities and

current assets minus inventory divided by current liabilities respectively, and measure the ability of the

company to meet its short-term payment requirements.22In Kang and Stulz (1997), the coefficient on current ratio in the regression of foreign ownership on

several explanatory variables is significantly positive in 1 year out of 16 years. Therefore, we believe it is

almost nonsignificant.23Source from P/E Ratio Global Stock Markets Analysis and Technical Outlook, November 07 issued on

18th November 2007 by The Market Oracle http://www.marketoracle.co.uk/. A high P/E ratio signifies

high expectations on the growth. Many developed countries have low P/E ratios, but they also have low

GDP growth, while developing countries may have higher market valuations as well as stronger GDP growth,

in particular the high P/E ratio and strong GDP growth in China.

16

are no significant differences between the two groups with regard to most indicators in

the management structure aspect except for the number of directors. Another notable

difference is the compensation of the management group. Both indicators are significantly

different between the two groups (at the 1% level of significance).

Our findings have highlighted a number of distinct issues between foreign and domestic

funds in particular in relation to financial and corporate governance characteristics. Al-

though commonalities exist, foreign and domestic institutional investors differ significantly

in portfolio selection in terms of size and profit of firms. In addition, there is also a con-

siderable difference between the companies in the two groups in relation to the aspect of

corporate governance, e.g. management compensation. Our results indicate that foreign

institutional investors don’t follow the “free rider strategy” even though they are not famil-

iar with the Chinese market. Therefore, our research Hypothesis 2, foreign and domestic

funds’ differences exist in relation to financial and corporate governance characteristics, is

supported.

4.3 Preferences of Foreign and Domestic Funds

In this section, we focus on analyzing the firm-level drivers that influence foreign and do-

mestic investors’ allocation choices.24 Both fixed effects models (FEM) and random effects

models (REM) are used in this paper to examine the relationships (Greene, 2002).25 Due

to the high correlation between the indicators in several classifications, we adopt a modified

step-wise regression approach (see Ghosh, Harding and Phani (2008) for a recent empirical

example) to indicate the appropriate proxy for each case. We regress the ownership of

foreign funds in each company against the industry and each of the individual indicators

to identify the appropriate determining variable, using the R2 of each regression.26 Fi-

nally, we estimate a combined model by selecting the most statistically significant measure

24Consistent with a number of studies examining developed markets (Dahlquist and Robertsson, 2001;

Aggarwal, Klapper and Wysocki, 2005), our approach focuses primarily on contemporaneous relationships.

Given the emerging nature of the market, the low level of foreign ownership and the limited channels of

corporate participation, we would expect there to be no potential for endogeneity in our study.25Hausman (1978) designs a test to indicate the appropriateness of the FEM versus the REM and decide

which is a better one. The null hypothesis states REM as the correct form while the alternative is FEM.26For one of the groups, we find there is no statistically significant indicator. In this case, we choose

the indicator with the higher overall R2 to join the final model because we hope to test the impact of a

comprehensive range of firm-level characteristics on the investment decisions of foreign institutional investors.

17

for each characteristic to examine the drivers of institutional investors. Consistent with

previous studies in the literature, we track funds’ portfolio holdings using the percentage

relative spread (see Dahlquist and Robertsson (2001) and Kang and Stulz (1997)).27 The

percentage relative spread captures funds’ deviations from investment allocations predicted

by Chinese market portfolio.28

Table 8 presents the selection process. Model 1 reports the results of the regression

of foreign funds’ percentage relative spread on size indicators and industrial dummy vari-

ables.29 30 For each explanatory variable, the table reports the parameter coefficient and

the overall R2 in parenthesis below. The superscripts of coefficients denote the significance

of the individual coefficient and the superscripts of the overall R2 denote the model used

to estimate. Overall R2 of the natural logarithm of total assets (0.0527) is smaller than

that of the natural logarithm of capitalization (0.0547), thus we choose the latter indicator

in the final model. The selecting processes of other financial subgroups are reported from

Model 2 to Model 13 which are similar to the format of Model 1. Duality of chairman and

CEO and percentage of independent directors are not included in M12 because they are

not highly correlated with the other two variables and they are included in the final model

directly.31

Table 9 reports the results for the overall model where the percentage relative spread of

foreign funds is regressed on the representative financial indicators and the representative

corporate governance indicators. We also control for industry and year effect by adding

industry and yearly dummy variables into the final model. In order to take account of the

27The percentage relative spread measures the over and under-investment of funds by calculating the

difference between the firms’ allocated weight and the Chinese market weight for each firm.28Dahlquist and Robertsson (2001) compare the portfolio of foreign investors with the Swedish market

portfolio. Kang and Stulz (1997) also adopt Japanese market portfolio as a benchmark to analyze the

investment of foreign investors.29Given the fact that the natural logarithm of total assets and the natural logarithm of capitalization are

so highly correlated (Pearson correlation=0.912), we do not include them in a single regression. To resolve

the problem of multicollinearity, we regress the two variables in two separate regressions.30As a sensitivity test, we employ the investment level of institutional investors (the percentage of market

capitalization of each firm in the total market capitalization of all the firms held by funds) as an alternative

method to track funds’ portfolio holdings. The results are consistent with those using the relative spread as

the dependent variable. Results are available from the authors upon request31Pearson correlations between duality of chairman and CEO, percentage of independent directors and

the number of directors, number of supervisors are all below 0.10.

18

introduction of new regulations on QFII’s in 2006, a dummy variable is included in the final

empirical model. The dummy variable, year 2006, is assigned a value of one if the year is

after 2006 and a zero otherwise. Table 9 reveals several findings on foreign institutional

investors’ preferences. Inconsistent with prior research in developed markets (see Kang and

Stulz, 1997; Dahlquist and Robertsson, 2001; Aggarwal, Klapper and Wysocki, 2005), we

find that firm size as measured by the natural logarithm of capitalization is not one of

the strongest determinants of foreign institutional investors’ investment decisions in China.

Furthermore, the insignificant relation between foreign institutional investors’ ownership

and firms’ profit is inconsistent with the finding in Kang and Stulz (1997) employing the

data of Japanese market. There is no significant relation between foreign institutional

investors’ ownership and firms’ liabilities paying capability no matter short-term or long-

term. Finally, free cash flow reports a negative relationship and statistically significant at

the 1% level. This result indicates that foreign institutional investors prefer the firms with

less free cash flow which is not consistent with the findings of Dahlquist and Robertsson

(2001).

Regarding corporate governance, the percentage of state-owned shares and Z index are

statistically significant with the ownership of foreign institutional investors. The results

indicate that foreign institutional investors have a preference for the firms with larger

percentage of state-owned shares. The influential role of government policy is a feature

of Chinese market. Naughton (2007) believes that Chinese market fluctuations are better

explained as reactions to government policy changes rather than as reaction to changes in

underlying fundamentals of individual companies. This China’s unique characteristic drives

foreign funds to invest more in firms with larger percentage of state-owned shares because

the higher percentage of state-owned shares in one company indicates the closer relation

with Chinese government. Consistent with Dahlquist and Robertsson (2001) and Huang

and Zhu (2011), there is a negative relationship between foreign institutional ownership

and firms’ ownership concentration measured by Z index, implying that foreigners avoid

companies with a dominant owner.32 Another notable significant finding is the year 2006

32The opposite signs for ownership of state-owned shares and the Z index are not conflict. Z index equals

to the number of shares held by the largest circulating shareholder divided by the number of shares held by

the second largest circulating shareholder. It stresses the control power of the largest shareholders. Even

the firm with high percentage of state-owned shares as the largest shareholder, Z index will be still small as

long as the number of shares held by the second largest shareholder is also large.

19

dummy variable. The statistically significant coefficient with a negative sign for the 2006

dummy variable indicates that the investment of QFIIs declined after 2006, although CSRC

issued the new regulation to increase the investment quota of QFII in China. However, it

is more likely that this dummy variable is taking account of the substantial volatility in

Chinese stock markets post 2006. Following the global economic downturn, the Shanghai

Stock Exchange Composite Index fell from a peak of 6,092 points on 16th October 2007 to

a trough of 1,719 on 3rd November 2008. In particular, Kaminsky, Lyons and Schmukler

(2001) conclude that although international mutual funds are one of the main sources of

capital flows to emerging economies, withdrawals from emerging markets during crisis are

large. Our results provide further indications of the sensitivity of institutional investment

to downturns in the market.

Table 10 presents the impact factors of domestic funds’ investment behavior. In order

to compare with the preference of foreign funds, we employ the same indicators as table 9 to

analyze the preference of domestic funds. Several notable observations emerge from table

10. Firm size has significant explanatory power, indicating that domestic funds prefer large

firms which is consistent with the finding of Huang and Zhu (2011). Although not a critical

indicator for foreign funds, earnings per share is a significant financial factor influencing

domestic funds’ investment behavior. For both foreign and domestic funds, ownership

structure is a dominate factor influencing their investment behavior but for different reasons.

While foreign investors want to access priority industries like manufacturing, with the

implications of high state ownership, domestic investors have a preference for investing in

industries that have lower state involvement and greater potential for external investment

derived growth. Clearly the investment preferences of foreign and domestic funds differ

substantially.

The conventional view is that government intervention has a negative impact on firm

performance. However, Cao, Lemmon, Pan, Qian and Tian (2011) highlight the potential

of political promotion providing incentives for managers in Chinese state-owned enterprises

to maximize firm value, which has a positive impact on corporate performance. Our find-

ings in relation to the behavior of foreign investors are consistent with this literature on

political connections. On the other hand, Firth, Lin and Zou (2010) find the government

exerts political pressure on domestic mutual funds to accept the compensation package to

complete the split share structure reform quickly. Specifically, the negative and significant

relationship between mutual fund ownership and the final compensation ratio indicates that

20

mutual funds help SOEs firms complete the reform at a lower cost which actually damages

their own benefits. Research also indicates that foreign institutional investors tend to gain

superior bargaining power in China, while local funds yield to political pressure from gov-

ernments (Huang and Zhu, 2011). Inferior bargaining power coupled with adverse political

pressure from governments, compared to foreign institutional investors, has the implication

that domestic funds in China invest less in firms with a high percentage of state-owned

shares. Finally, foreign institutional investors require approval by stock market authori-

ties, e.g. CSRC and SAFE, under the QFIIs scheme which further underlines the political

influence.

We conclude that the drivers of foreign funds in emerging markets are not consistent

with those in developed markets, e.g. size, liabilities paying capability and profit (see Kang

and Stulz, 1997; Dahlquist and Robertsson, 2001). Foreign investors in China place more

emphasis on the corporate governance characteristics. In particular, they have a preference

for the firms with a high percentage of state-owned shares which indicates more political

connections. Consistent with the finding in Leuz, Lins and Warnock (2009), corporate

governance is indeed an important factor influencing the investment decisions of foreign

institutional investors. In contrast, however, domestic funds place greater emphasize on

financial characteristics, e.g. size and return on shareholders as well as corporate gov-

ernance structure. Therefore, our research Hypothesis 3, corporate governance (financial

characteristics) has a significant impact on the investment of foreign (domestic) institu-

tional investors, is supported. We conclude that the investment preferences of foreign and

domestic funds differ substantially.

5 Conclusions

This paper employs a unique data set to analyze the key firm-level drivers of foreign insti-

tutional investors, including financial indicators and corporate governance indicators, and

identifies the similarities and differences between foreign and domestic funds. In the first

part of the study we analyze the industrial distribution of the companies foreign and do-

mestic funds invest in. Furthermore, we use the t test and median test to compare the

mean and median of financial and corporate governance indicators of the companies held

by foreign and domestic funds respectively. Finally, a modified step-wise regression, con-

trolling for the effects of different industry types and years, is employed to reveal firm-level

21

factors influencing the investment by foreign and domestic funds.

Results of our analysis indicate that foreign institutional investors have a preference

for industries like transportation, metals & non-metals, and machinery. However, there

is very little foreign institutional investment in real estate, construction, and media &

culture, where there is a requirement for local knowledge. This result is consistent with

evidence from developed countries (Kang and Stulz, 1997; Dahlquist and Robertsson, 2001).

The portfolios of domestic Chinese funds are distributed more evenly compared to foreign

funds in different sectors. The characteristics of firms that foreign and domestic funds

invest in are significantly different in terms of size and profit which are consistent with

the findings of developed markets (Kang and Stulz, 1997). The comparative analysis of

firm-level characteristics between firms held by foreign and domestic funds partly indicates

that foreign institutional investors do not follow the investment decisions of domestic funds,

even taking account of the limited information associated with the Chinese market.

Finally, corporate governance characteristics, e.g. ownership structure and concentra-

tion play a significantly strong role in determining foreign funds holdings, whereas some

basic financial indicators, e.g. size and profit of a firm do not have a substantial impact

which is inconsistent with the findings in developed markets. Our results indicate that

corporate governance has an impact on the investment decisions of foreign institutional

investors.

Our empirical findings will be of value to policy makers in emerging markets, and China

in particular, in gauging the important drivers of foreign institutional investors. For the

emerging markets, our results highlight the sensitivity of foreign institutional investment to

market cycles. While Choe, Kho and Stulz (1999) find no evidence that trades by foreign

investors had a destabilizing effect on Korea’s equity market over the crisis in 1997, our

results do indicate the sensitive nature of foreign institutional investment in China. For

China in particular, our findings indicate that the ownership structure of Chinese listed

firms is a critical attraction for foreign institutional investors. Finally, given the increasing

significance of foreign financing and the fact that access to foreign capital is uneven across

firms, it is important for firms to fully understand the factors that influence investors’

behavior. In particularly, stocks with high foreign ownership outperform stocks with low

foreign ownership (Huang and Shiu, 2006).

22

References

Aggarwal, R., Klapper L., and Wysocki, P. 2005. Portfolio Preferences of Foreign Institu-

tional Investors. Journal of Banking and Finance 29, no. 12: 2919-2946.

Bushee, B., M. E. Carter, and J. Gerakos. 2009. Institutional Investor Preferences for

Corporate Governance Mechanisms. SSRN Working Paper Series.

Cao, J., Lemmon, M., Pan, X., Qian, M, and Tian, G. 2011. Political Promotion, CEO

Incentives, and the Relationship between Pay and Performance. Working paper.

Chan, K., Covrig, V., and Ng, L. 2005. What Determines the Domestic Bias and Foreign

Bias? Evidence from Mutual Fund Equity Allocations Worldwide. Journal of Finance

60, no. 3: 1495-1534.

Chang, C. 2010. Information Footholds: Isolating Local Presence as a Factor in Analyst

Performance and Trading. Journal of International Money and Finance 29, no. 6: 1094-

1107.

Chen, M., and Yu, W. 2003. Stock Market Liberalization and Market Returns in China:

Evidence from QFII Announcement. SSRN Working Paper.

Chen, X., Harford, J., and Li, K. 2007. Monitoring: Which Institutions Matter? Journal

of Financial Economics 86, no.2 :279-305.

Cheng, L. K., and Kwan, Y. K. 2000. What Are the Determinants of the Location of Foreign

Direct Investment? The Chinese Experience. Journal of International Economics 51, no.

2: 379-400.

Choe, H., Kho, B., and Stulz, R. M. 1999. Do Foreign Investors Destabilize Stock Markets?

The Korean Experience in 1997. Journal of Financial Economics 54, no. 2: 227-264.

Choe, H., Kho, B., and Stulz, R. M. 2005. Do Domestic Investors Have an Edge? The

Trading Experience of Foreign Investors in Korea. The Review of Financial Studies 18,

no. 3: 795-829.

Cooper, I. A., and Kaplanis, E. 1994. What Explains the Home Bias in Portfolio Investment.

Review of Financial Studies 7, no. 1: 45-60.

23

Covirg, V., Lau, S. T., and Ng, L. 2006. Do Domestic and Foreign Fund Managers have

Similar Preferences for Stock Characteristics? A Cross Country Analysis. Journal of

International Business Studies 37, no. 3: 407-429.

Dahlquist, M., and Robertsson, G. 2001. Direct Foreign Ownership, Institutional Investors,

and Firm Characteristics. Journal of Financial Economics 59, no. 3: 413-440.

Dean, J. 2003. Taiwan’s stock market could get a lift. Wall Street Journal (Eastern Edition),

C 14.

Falkenstein, E. G. 1996. Preferences for Stock Characteristics as Revealed by Mutual Fund

Portfolio Holdings. Journal of Finance 51, no. 1: 111-135.

Firth, M., Lin, C., and Zou, H. 2010. Friend or Foe? The Role of State and Mutual

Fund Ownership in the Split Share Structure Reform in China. Journal of Financial and

Quantitative Analysis 45, no. 3: 685-706.

Frenkel, K. R., and Poterba, J. M. 1991. Investor Diversification and International Equity

Markets. American Economic Review 81, no. 2: 222-226.

Frenkel, M., and Menkhoff, L. 2004. Are Foreign Institutional Investors Good for Emerging

Markets? The World Economy 27, no. 8: 1275-1293.

Ghosh, C., Harding, J., and Phani, B. 2008. Does Liberalization Reduce Agency Costs?

Evidence from the Indian Banking Sector. Journal of Banking and Finance 32, no. 7:

405-419.

Gillan, S. L., and Starks, L. T. 2000. Relationship Investing and Shareholder Activism by

Institutional Investors. Journal of Financial Economics 57, no. 2: 275-305.

Gompers, P., Ishii, J., and Metrick, A. 2003. Corporate Governance and Equity Prices.

Quarterly Journal of Economics 118, no. 1: 107-155.

Greene, W. H. 2002. Econometric analysis (5th edition). Prentice Hall.

Grinblatt, M., and Keloharju, M. 2000. The Investment Behavior and Performance of Var-

ious Investor Types: A Study of Finland’s Unique Data Set. Journal of Financial Eco-

nomics 55, no. 1: 43-67.

24

Hausman, J. A. 1978. Specification Tests in Econometrics. Econometrica 46, no. 6: 1251-

1271.

Huang, R. D., and Shiu, C. 2006. Local Effects of Foreign Ownership in an Emerging

Financial Market: Evidence from Qualified Foreign Institutional Investors in Taiwan.

Financial Management 38, no. 3: 567-602.

Huang, W., and Zhu, T. 2011. Foreign Investment vs. State Ownership: Is There a Political

Pressure? The Illustrative Case of Qualified Foreign Institutional Investors in China.

Working Paper.

Jiang, L., and Kim V, J. 2004. Foreign Equity Ownership and Information Asymmetry:

Evidence from Japan. Journal of International Financial Management and Accounting

15, no. 3: 185-211.

Kaminsky, G. L., Lyons, R. K., and Schmukler, S. L. 2001. Mutual Fund Investment in

Emerging Markets: An Overview. The World Bank Economic Review 15, no. 2: 315-340.

Kang, J., and Stulz, R. M. 1997. Why is There a Home Bias? An Analysis of Foreign

Portfolio Equity Ownership in Japan. Journal of Financial Economics 46, no. 1: 3-28.

Karpoff, J. M., Malatesta, P. H., and Walkling, R. A. 1996. Corporate Goverance and

Shareholder Initiatives: Empirical Evidence. Journal of Financial Economics 42, no. 3:

365-395.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. 1997. Legal Determinants

of External Finance. Journal of Finance 52, no. 3: 1131-1150.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. 1998. Law and Finance.

Journal of Political Economy 106, no. 6: 1113-1155.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. 2000. Investor Protection

and Corporate Governance. Journal of Financial Economics 58, no. 5: 3-28.

Lai, C., Lou, K., and Shiu, C. 2008. Does Foreign Investors Trade Predict Stock Return?

Evidence from Taiwan. Information and Management Sciences 19, no. 3: 477-494.

Leuz, C., Lins, K. V., and Warnock, F. E. 2009. Do Foreigners Invest Less in Poorly

Governed Firms? Review of Financial Studies 22, no. 8: 3245-3285.

25

Lewis, K. K. 1999. Trying to Explain Home Bias in Equities and Consumption. Journal of

Economic Literature 37, no. 2: 571-608.

Lin, A. Y., and Swanson, P. E. 2003. The Behavior and Performance of Foreign Investors

in Emerging Equity Markets: Evidence from Taiwan. International Review of Finance 4,

no. 3-4: 189-210.

Lin, C. H., and Shiu, C. 2003. Foreign Ownership in the Taiwan Stock Market: An Empirical

Analysis. Journal of Multinational Financial Management 13, no. 1: 19-41.

Lin, A., and Chen, C. 2006. The Impact of Qualified Foreign Institutional Investors on

Taiwan’s Stock Market. Journal of Chinese Management Review 9, no. 2: 1-27.

Lu, Y., Fang, H., Wong, J., and Wang, Y. 2009. Information Contents of QFIIs’ Cascades

in the Taiwan Stock Market. SSRN Working Paper.

McKinsey and Company. 2002. Global Investor Opinion Survey. July.

Measures for the Administration of Investment in Domestic Securities by Qualified Foreign

Institutional Investors, China Securities Regulatory Commissionon, People’s Bank of

China and State Administration of Foreign Exchange, 24th August 2006.

Merton, R. C. 1987. A Simple Model of Capital Market Equilibrium with Incomplete In-

formation. Journal of Finance 42, no. 3: 483-510.

Naughton, B. 2007. The Chinese Economy Transitions and Growth. The MIT Press.

Ng, Lilian and Fei Wu. 2007. The Trading Behavior of Institutions and Individuals in

Chinese Equity Markets. Journal of Banking and Finance 31, no. 9: 2695-2710.

Poon, W. P. H., and Chan, K. C. 2008. An Empirical Examination of the Informational

Content of Credit Ratings in China. Journal of Business Research 61, no. 7: 790-797.

Provisional Measures on Administration of Domestic Securities Investments of Qualified

Foreign Institutional Investors, China Securities Regulatory Commissionon, 5th Novem-

ber 2002.

Sercu, P., and Vanpee, R. 2007. Home Bias in International Equity Portfolios: A Review.

Working paper, Katholieke University Leuven.

26

Stulz, R. M. 2005. The Limits of Financial Globalization. Journal of Finance 60, no. 4:

1595-1638.

Tesar, L., and Werner, I. M. 1995. Home Bias and High Turnover. Journal of International

Money and Finance 14, no. 4: 467-493.

Ting, W., Yen, S., and Chiu, C. 2008. Risk and Audit Opinions: Evidence form the Chinese

Stock Market. Corporate Governance: An International Review 16, no. 5: 400-415.

Vo, X. V. 2010. Foreign Ownership in Vietnam Stock Markets - An Empirical Analysis.

MPRA Working Paper.

Wu, S., Xu, N., and Yuan, Q. 2009. State Control, Legal Investor Protection, and Ownership

Concentration: Evidence from China. Corporate Governance: An International Review

17, no. 2: 176-192.

Yeo, S. 2003. The PRC Qualified Foreign Institutional Investors Market. China Economic

Review 14, no. 4: 443-450.

Zhang, K. H. 2001. China’s Inward FDI Boom and the Greater Chinese Economy. Chinese

Economy 34, no. 1: 74-88.

27

Jun 03 Jan 04 Jun 04 Jan 05 Jun 05 Jan 06 Jun 06 Jan 07 Jun 07 Jan 08 Jun 08 Jan 09 Jun 09 Jan 10 Sep 100

20

40

60

80

100

120

140

160

180

200

month

tota

l inv

estm

ent q

uota

of Q

FIIs

(10

0 m

illio

n U

S d

olla

rs)

Figure 1: Investment Quota of QFIIs in China

28

Table 1: QFII list

QFII’s Name Approved Date Investment Quota Investment Quota(Approved Date) (September 2010)

(100 million US dollars) (100 million US dollars)1 UBS AG 04/06/2003 3.00 8.002 Nomura Securities Co.,Ltd. 04/06/2003 0.50 3.503 Citigroup Global Markets Limited 18/06/2003 0.75 5.504 Morgan Stanley & Co. International Limited 01/07/2003 3.00 4.005 Goldman, Sachs & Co. 24/07/2003 0.50 3.006 The Hongkong and Shanghai Banking Corporation Limited 26/08/2003 0.50 4.007 Deutsche Bank AG or Deutsche Bank Aktiengesellschaft 26/08/2003 0.50 4.008 ING Bank N. V. 16/10/2003 1.00 4.009 Jpmorgan Chase Bank 04/11/2003 0.50 1.5010 Credit Suisse (HongKong) Limited 28/11/2003 0.50 5.0011 Nikko Asset Management Co.,Ltd. 09/02/2004 0.50 4.5012 Standard Chartered Bank (HongKong) Limited 19/05/2004 0.75 0.7513 Hangseng Bank 22/06/2004 0.50 1.0014 Daiwa Securities SMBC Co.,Ltd. 05/07/2004 0.50 0.5015 Merrill Lynch International 16/07/2004 0.75 3.0016 Lehman Brothers International (Europe) 16/08/2004 0.75 2.0017 Bill & Melinda Gates Foundation 28/08/2004 1.00 3.0018 ABN AMRO Bank N.V. 17/09/2004 0.75 1.7519 Socit Gnrale 17/09/2004 0.50 0.5020 Barclays Bank PLC 15/10/2004 0.75 4.0021 BNP Paribas 27/10/2004 0.75 2.0022 Dresdner Bank Aktiengesellschaft 08/11/2004 0.75 0.7523 Fortis Bank SA/NV 21/11/2004 1.00 5.0024 Power Corporation of Canada 21/11/2004 0.50 0.5025 Calyon S.A. 10/01/2005 0.75 0.7526 INVESCO Asset Management Limited 08/03/2005 0.50 2.5027 Government of Singapore Investment Corporation Pte Ltd 16/11/2005 1.00 3.0028 Goldman Sachs Asset Management International 16/11/2005 2.00 5.0029 Martin Currie Investment Management Ltd 24/11/2005 1.20 1.2030 Temasek Fullerton Alpha Investments Pte Ltd 12/12/2005 1.00 3.0031 AIG Global Investment Corp 12/12/2005 0.50 0.5032 The Dai-ichi Mutual Life Insurance Company 22/02/2006 1.00 2.0033 DBS Bank Ltd. 12/04/2006 1.00 1.0034 JF Asset Management Limited 12/04/2006 1.50 2.7535 KBC Financial Products UK Limited 09/06/2006 1.00 1.0036 Scotia Bank or The Bank of Nova Scotia 09/06/2006 1.50 1.5037 La Compagnie Financierr Edmond de Rothschild Banque 19/07/2006 1.00 1.0038 Yale University 01/08/2006 0.50 1.5039 AMP Capital Investors Ltd. 01/08/2006 2.00 3.00

Continued on next page

29

Table 1 – continued from previous page

QFII’s Name Approved Date Investment Quota Investment Quota(Approved Date) (September 2010)

(100 million US dollars) (100 million US dollars)40 Morgan Stanley Investment Management Inc. 05/09/2006 2.00 4.5041 Prudential Asset Management (Hongkong) Limited 12/10/2006 2.00 3.0042 Stanford University 07/11/2006 0.50 1.0043 United Overseas Bank Limited 07/11/2006 0.50 0.5044 Schroder Investment Mangement Limited 11/12/2006 2.00 2.0045 GE Asset Management Incorporated 11/01/2007 2.00 3.5046 UBS Global Asset Management (Singapore) Ltd 11/01/2007 2.00 2.0047 Shinko Securities Co.Ltd 13/02/2007 0.50 0.5048 HSBC Investments (Hongkong) Limited 13/02/2007 2.00 3.5049 Sumitomo Mitsui Asset Management Company, Limited 13/02/2007 2.00 3.5050 Norges Bank 24/01/2008 2.00 7.0051 Pictet Asset Management Limited 01/04/2008 1.00 1.0052 The Trustees of Columbia University in the City of New York 07/04/2008 1.00 1.0053 Prudential Asset Management Co.,Ltd. 04/05/2008 0.75 0.7554 Robeco Institutional Asset management B.V. 20/06/2008 1.50 1.5055 KBC Asset Management N.V. 31/07/2008 1.50 1.5056 Mirae Asset Investment Management Co., Ltd 02/09/2008 1.50 2.5057 Platinum Investment Company Limited 10/09/2008 1.50 1.5058 State Street Global Advisors Asia Limited 03/11/2008 0.50 0.5059 Caisse de d pot et placement du Qu bec 03/11/2008 2.00 2.0060 Samsung Investment Trust Management Co., Ltd. 07/11/2008 1.50 3.0061 Oversea-Chinese Banking Corporation Limited 12/11/2008 1.50 1.5062 AllianceBernstein Limited 12/11/2008 0.50 1.5063 ACE INA International Holdings,Ltd. 13/11/2008 1.50 1.5064 President and Fellows of Harvard College 14/11/2008 2.00 2.0065 T. Rowe Price International, Inc. 03/12/2008 1.10 1.1066 DAIWA Asset Management Co. 26/12/2008 1.00 1.0067 ABU Dhabi Investment Authority 17/01/2009 2.00 2.0068 Allianz Global Investors Luxembourg S.A. 04/03/2009 1.00 1.0069 Mitsubishi UFJ Securities Co., Ltd. 25/03/2009 1.00 1.0070 Capital International,inc. 31/03/2009 1.00 1.0071 Credit Suisse 22/05/2009 2.00 2.0072 Emerging Markets Management, L.L.C. 03/06/2009 0.50 0.5073 First State Investment Management (UK) Limited 16/06/2009 1.20 1.2074 Hanwha Investment Trust Management Co., Ltd 10/08/2009 0.70 0.7075 UOB Asset Management LTD 25/08/2009 0.50 0.5076 Bank Negara Malaysia 04/09/2009 2.00 2.0077 DWS Investment S.A. 09/09/2009 2.00 2.0078 Lloyd George Management (Hong Kong) Limited 06/11/2009 0.50 0.5079 The Korea Development Bank 09/11/2009 1.00 1.00

Continued on next page

30

Table 1 – continued from previous page

QFII’s Name Approved Date Investment Quota Investment Quota(Approved Date) (September 2010)

(100 million US dollars) (100 million US dollars)80 Templeton Investment Counsel, LLC 08/12/2009 2.00 2.0081 Shell Asset Management Company B.V. 08/12/2009 1.00 1.0082 BEA Union Investment Management Limited 08/12/2009 1.00 1.0083 Woori Bank Co., Ltd 30/12/2009 0.50 0.5084 Korea Investment Trust Management Co.,Ltd 30/12/2009 1.00 1.0085 The Sumitomo Trust & Banking Co. Ltd 31/12/2009 0.50 0.5086 Baring Asset Management Ltd 10/02/2010 2.00 2.0087 Ashmore Investment Management Ltd 10/02/2010 2.00 2.0088 Nomura Asset Management CO., LTD 04/05/2010 2.00 2.0089 Manulife Asset Management (Hong Kong) Limited 31/05/2010 2.00 2.0090 Tongyang Investment Trust Management Co., Ltd 22/07/2010 1.00 1.0091 Royal Bank of Canada 19/08/2010 1.00 1.0092 Ivy Investment management Company 01/09/2010 1.00 1.0093 DIAM Co., Ltd. 01/09/2010 1.00 1.00

Total 108.2 190.0

Note: In this table, we report the list of QFIIs from June 2003 to September 2010. The source is from Investment Quota Approval Form of QFIIs issued bySAFE, 10th November 2010. The last two columns are the investment quota of each QFII at approved date and investment quota by the end of September2010 respectively. The units are 100 million US dollars in each column. Up to 30th September 2010, there were 93 QFIIs approved by CSRC and SAFE andthe investment quotas were 19 billion US dollars.

31

Table 2: Distribution of funds by year

QFII Domestic Fund

Year No. of QFIIs No. of firms QFIIs invest in No. of domestic funds No. of firms domestic funds invest in

2003 10 17 110 5162004 24 35 161 10492005 31 122 218 10622006 44 196 301 11132007 49 154 346 8932008 66 124 439 8872009 85 210 557 1273

mean 44 123 305 970

Note: This table shows the distribution of QFIIs and domestic funds in terms of year. The information on amountsof QFIIs and domestic funds is obtained from Investment Quota Approval Form of QFIIs issued by SAFE, 14th July2010 and monthly report issued by CSRC respectively. Following the global economic downturn, the number of firmswhose share are partially held by foreign institutional investors fell in 2007 and 2008. Shanghai Stock ExchangeComposite Index fell from the peak of 6,092 points on 16th October 2007 to the bottom of 1,719 on 3rd November2008.

32

Table 3: Financial Indicator Definition

Classification Indicator Symbol Formula

Size Total assets LNTA ln(total assets)Capitalization LNC ln (capitalization)

Short-term liabilities Current ratio CR current assets/current liabilitiespaying capability Quick ratio QR (current assets-inventory)/current liabilities

Long-term liabilities Assets-liabilities ratio ALR total liabilities/total assetspaying capability Long-term liabilities ratio LDR long-term liabilities/total liabilities

Financial risk Degree of financial leverage DFL earnings before interest and taxes (EBIT)/(EBIT-financial cost)

Degree of operating leverage DOL % change in operating income/% change in sales

Return on shareholders Earnings per share EPS net profit/total shares outstandingPrice to earning ratio P/E price per share/earnings per share

Profit Return on assets ROA netprofit/average total assetsReturn on equity ROE net profit/average total equity

Cash flow Free cash flow FCF net profit+amortization/depreciation-changes in working capital-capital expenditures

Net cash flow per share NCFPS net increase in cash and cash equivalents/total shares outstanding

Growth Growth rate of fixed assets GRFA (ending fixed assets-beginning fixed assets)/beginning fixed assets

Growth rate of total assets GRTA (ending total assets-beginning total assets)/beginning total assets

Operating capability Turnover rate of receivables TRR operating income/average receivablesTurnover rate of inventory TRI operating cost/average inventory

Note: In this table, we explain the financial indicators employed in the paper. We divide the comprehensive range of financial characteristics into 9 subgroups,namely size, short-term liabilities paying capability, long-term liabilities paying capability, financial risk, return on shareholders, profit, cash flow, growth, andoperating capability. In each subgroup, two indicators are used to measure this classification for robust. In the last column, we also introduce the way tocalculate the indicator.

33

Table 4: Corporate Governance Definition

Classification Indicator Symbol Formula

Ownership structure Percentage of state-owned shares PSOS number of state-owned shares/number of total sharesPercentage of circulating shares PCS number of circulating shares/number of total shares

Ownership concentration Ownership percentage of the largestcirculating shareholder

OPLCS number of shares held by the largest circulating share-holder/number of total shares

Z index Z number of shares held by the largest circulating share-holder/number of shares held by the second largest cir-culating shareholder

Management structure Number of directors ND number of directorsNumber of supervisors NS number of supervisorsDuality of chairman and CEO DCEO a dummy variable equals to 1 if chairman and CEO is

one person and 2 if otherwisePercentage of independent directors PID number of independent directors/number of directors

Compensation of ln of sum of top three compensation LNSD3 ln of Sum to top three compensation of directorsmanagement of directors

ln of sum of top three compensationof senior executives

LNSSE3 ln of sum of top three compensation of senior executives

Note: In this table, we explain the corporate governance indicators employed in the paper. We divide the comprehensive range of corporate governancecharacteristics into 4 subgroups, including ownership structure, ownership concentration, management structure, and compensation of management. In eachsubgroup, some indicators are used to measure this classification for robust. In the last column, we also introduce the way to calculate the indicator.

34

Table 5: Industry Allocations of Foreign and Domestic Funds in China

Firm Number Shareholding Number Market Value(Panel A) (Panel B) (Panel C)

Industry Foreign Domestic Foreign Domestic Foreign Domestic

Agriculture(A) 0.29 -0.08 -0.01 -0.26 0.65 0.19(21) (142)

Mining(B) 1.80 0.18 -4.26 -0.85 -3.44 0.38(36) (172)

Food & Beverage(C0) 1.46 0.39 2.38 1.23 3.13 5.18(48) (304)

Textiles & Apparel(C1) -1.22 -0.44 -0.61 -0.74 0.02 -0.01(25) (252)

Timber & Furnishings(C2) -0.01 -0.07 -0.01 0.04 0.07 0.07(3) (20)

Paper & Printing(C3) -0.30 -0.20 0.30 0.13 0.56 0.33(15) (125)

Petrochemicals(C4) -3.04 -0.16 -0.95 0.57 1.15 3.41(64) (705)

Electronics(C5) -1.52 -0.37 -0.51 -0.52 0.44 0.34(25) (277)

Metals & non-Metals(C6) 3.47 0.56 14.02 6.11 9.72 6.56(105) (628)

Machinery(C7) 0.02 -0.93 5.10 2.71 10.85 6.81(140) (1040)

Pharmaceutical(C8) 0.35 0.27 -0.10 1.46 1.38 3.48(58) (452)

Other Manufacturing(C9) -0.15 -0.17 -0.04 -0.05 0.07 0.07(7) (54)

Utilities(D) 1.86 0.71 2.24 0.41 1.07 0.81(50) (315)

Construction (E) -0.37 -0.04 -1.39 0.12 -0.93 -0.52(17) (157)

Transportation (F) 4.82 0.99 4.17 0.38 5.11 3.41(78) (351)

Information Technology(G) -2.75 -1.05 -0.85 2.15 2.15 3.07(37) (410)

Wholesale & Retail(H) -0.19 0.68 0.35 2.36 1.58 4.71(47) (431)

Finance & Insurance(I) 0.39 0.14 -19.44 -17.81 -35.63 -44.79(19) (132)

Real Estate(J) -2.76 0.19 -0.23 2.94 0.86 4.41(25) (402)

Social Services(K) 1.37 -0.20 1.03 0.55 1.02 1.23(40) (207)

Media & Culture(L) -0.84 -0.03 -0.23 0.40 -0.11 0.61(0) (55)

Conglomerates(M) -2.68 -0.37 -0.96 -1.33 0.28 0.25(17) (293)

Total (877) (6925)

Note: This table shows statistics of industry allocations by foreign and domestic funds. Column 1 lists the industriesin China according to the classification of CSRC. Column 2 and 3 (Panel A) report the deviation in percent of eachindustry’s weight in the portfolio held by foreign and domestic funds from its weight in the Chinese market portfolio interm of the number of firms. For example, the first row shows that 0.29% means that foreign funds invest 0.29% moreof their portfolio in Agriculture industry than they would if their investment weights were those of the Chinese marketportfolio. It indicates the fund’s over/under investment in an industry relative to the Chinese market portfolio. Inparentheses, the numbers of firms for each industry are reported. Column 4 and 5 (Panel B) report the deviationin term of the number of shares. Column 6 and 7 (Panel C) present it in term of the number of market value. Thepercentage of each industry’s weight in the Chinese market portfolio is the value by the end of 2009. The full nameof industry A is agriculture, forestry, animal husbandry & fishery and F is communication, transportation & storage.

35

Table 6: Financial Indicator Comparison

Classification Indicator Test QFII Domestic Fund p-valueSize ln of total assets mean 22.05 21.65 .000∗∗∗

median 21.87 21.51 .000∗∗∗

ln of capitalization mean 22.52 22.04 .000∗∗∗

median 22.44 21.90 .000∗∗∗

Short-term current ratio mean 1.69 1.82 .161liabilities paying capability median 1.20 1.27 .014∗∗

quick ratio mean 1.27 1.36 .294median 0.78 0.84 .009∗∗∗

Long-term assets-liabilities ratio mean 0.47 0.48 .305liabilities paying capability median 0.48 0.49 .060∗

long-term liabilities ratio mean 0.19 0.16 .000∗∗∗

median 0.12 0.09 .002∗∗∗

Financial risk degree of financial leverage mean 1.20 1.31 .033∗∗

median 1.13 1.14 .129degree of operating leverage mean 1.76 1.98 .008∗∗∗

median 1.64 1.78 .001∗∗∗

Return on shareholders earnings per share mean 0.49 0.33 .000∗∗∗

median 0.39 0.26 .000∗∗∗

price to earning ratio mean 65.76 68.24 .689median 29.53 32.27 .002∗∗∗

Profit return on assets mean 0.07 0.05 .000∗∗∗

median 0.06 0.04 .000∗∗∗

return on equity mean 0.14 0.09 .000∗∗∗

median 0.12 0.09 .000∗∗∗

Cash flow free cash flow mean 8.94 4.86 .005∗∗∗

median 2.09 1.09 .000∗∗∗

net cash flow per share mean 0.34 0.29 .245median 0.09 0.07 .081∗

Growth growth rate of fixed assets mean 0.29 0.31 .756median 0.08 0.06 .020∗∗

growth rate of total assets mean 0.24 0.23 .557median 0.15 0.12 .002∗∗∗

Operating capability turnover rate of receivables mean 125.79 86.79 .303median 9.20 7.74 .000∗∗∗

turnover rate of inventory mean 31.89 17.75 .089∗

median 4.85 4.22 .001∗∗∗

Note:In order to compare the different firm-level characteristics of firms QFIIs invest in and those favored by domestic funds, we divide thewhole sample into two groups: one group includes the firms that have at least one QFII holding their equity, and another group includes thefirms without QFII presence. In this table, we compare a range of financial indicators between companies in two groups. The first column is theclassification of financial characteristics. In the second column, we choose some indicators to measure this characteristic. T test and median testare employed to compare the difference between these two groups. The mean and median of each indicator of these two groups are in the fourthand fifth column respectively. P-value of test is in the last column. ∗∗∗,∗∗,∗ represent significance at 1%,5%, and 10% levels respectively. Theunit of total assets, capitalization, earnings per share, and net cash flow per share is CNY while the unit of free cash flow is 100 million CNY.

36

Table 7: Corporate Governance Comparison

Classification Indicator Test QFII Domestic Fund p-valueOwnership structure percentage of state-owned shares mean 0.29 0.28 .645

median 0.30 0.28 .308percentage of circulating shares mean 0.54 0.51 .000∗∗∗

median 0.50 0.46 .000∗∗∗

Ownership concentration ownership percentage of the largest circulating shareholder mean 8.72 6.23 .000∗∗∗

median 3.11 1.95 .000∗∗∗

Z index mean 9.25 6.18 .014∗∗

median 1.60 1.57 .554Management structure number of directors mean 9.74 9.48 .002∗∗∗

median 9.00 9.00 .011∗∗

number of supervisors mean 4.16 4.11 .469median 3.00 3.00 .493

duality of chairman and CEO mean 1.85 1.85 .903median 2.00 2.00 .900

percentage of independent directors mean 0.35 0.35 .565median 0.33 0.33 .450

Compensation of management ln of sum to top three compensation of directors mean 13.49 13.26 .000∗∗∗

median 13.52 13.30 .000∗∗∗

ln of sum of top three compensation of senior executives mean 13.71 13.43 .000∗∗∗

median 13.73 13.46 .000∗∗∗

Note: In order to compare the different firm-level characteristics of firms QFIIs invest in and those favored by domestic funds, we divide the whole sampleinto two groups: one group includes the firms that have at least one QFII holding their equity, and another group includes the firms without QFII presence.In this table, we compare some corporate governance indicators between the companies in two groups. The first column is the classification of corporategovernance characteristics. In the second column, we choose some indicators to measure this characteristic. T test and median test are employed to comparethe difference between these two groups. The mean and median of each indicator of these two groups are in the fourth and fifth column respectively. P-valueof test is in the last column. ∗∗∗,∗∗,∗ represent significance at 1%,5%, and 10% levels respectively. The unit of compensation is CNY.

37

Table 8: Test for firm-level characteristics in explaining investment allocation of

foreign funds

Model Classification Indicator

Model 1 Size LN of total assets -0.5801(0.0527)F

LN of capitalization -0.1247(0.0547)F

Model 2 Short-term liabilities current ratio -0.0242paying capability (0.0226)R

quick ratio -0.0283(0.0225)R

Model 3 Long-term liabilities assets-liabilities ratio 0.0798paying capability (0.0224)R

long-term liabilities ratio 0.7683(0.0249)R

Model 4 Risk degree of financial leverage -0.0586(0.0227)R

degree of operating leverage -0.1227(0.0040)F

Model 5 Return on shareholders earnings per share 0.8455∗∗∗

(0.0429)R

price to earning ratio -0.0008(0.0238)R

Model 6 Profit ROA 4.6722∗∗∗

(0.0308)R

ROE 2.4670∗∗∗

(0.0345)R

Model 7 Cash flow free cash flow -0.0169∗∗

(0.0371)F

net cash flow per share 0.4592∗∗∗

(0.0025)F

Model 8 Growth growth rate of fixed assets -0.0126(0.0223)R

growth rate of total assets 0.1894(0.0229)R

Model 9 Operating capability turnover rate of receivables -0.0001(0.0224)R

turnover rate of inventory -0.0001(0.0223)R

Model 10 Ownership structure percentage of state-owned 2.7856∗∗∗

shares (0.0043)F

percentage of circulating -2.1622∗∗

shares (0.0013)F

Model 11 Ownership concentration Ownership percentage of -0.0260∗∗

the largest circulating shareholder (0.0007)F

Z index -0.0323∗∗∗

(0.0036)F

Model 12 Structure of management number of directors 0.1222∗∗∗

(0.0330)R

number of supervisors 1.0175∗∗∗

(0.0149)F

Model 13 Compensation of management ln of sum of top three 0.5083∗∗∗

compensation of directors (0.0522)R

ln of sum of top three 0.6159∗∗∗

compensation of senior executives (0.0544)R

Note: Panel regression estimates the % relative spread of foreign institutional investors in each company on financialand corporate governance firm-level characteristics. Both Fixed Effects Model and Random Effects Model are usedto explain the impacted factors of the investment of foreign institutional investors in China’s Stock Market. Wealso include the industrial dummy variables in each regression to control the industrial effect. For each explanatoryvariable, the table reports the coefficient above and the overall R2 in parenthesis below. The superscripts of coefficientdenote the significance of the coefficient(∗∗∗ ,∗∗,∗ represent significance at 1%,5%, and 10% levels respectively) andthe superscripts of the overall R2 suggest the model used to estimate (F and R represent Fixed Effects Model andRandom Effects Model, respectively). In the last column, we report the indicators which are selected into final modelin bold. The specification test devised by Hausman (1978) tests the null hypothesis that the coefficients estimatedby the efficient random effects estimator are the same as the ones estimated by the consistent fixed effects estimator.If they are insignificant, it is safe to use random effects. If it is a significant P-value, however, you should use fixedeffects.

38

Table 9: Preference of QFIIs

Classification Indicator

Size LN of capitalization 0.5155(1.11)

Short-term liabilities paying capability current ratio -0.2216(-0.92)

Long-term liabilities paying capability long-term liabilities ratio 1.2435(0.99)

Risk degree of financial leverage 0.0250(0.11)

Return on shareholders earnings per share -0.9645(-1.31)

Profit ROE 4.6125∗

(1.85)Cash flow free cash flow -0.0194∗∗∗

(-2.56)Growth growth rate of total assets -1.0335∗

(-0.92)Operating capability turnover rate of receivables -0.0001

(-0.01)Ownership structure percentage of state-owned shares 2.1179∗∗

(2.27)Ownership concentration Z index -0.0296∗∗∗

(-3.27)Structure of management number of directors 0.0124

(0.10)duality of chairman and CEO -0.0312

(-0.04)percentage of independent directors -1.8448

(-0.40)Compensation of management ln of sum of top three compensation of senior executives 0.8696∗

(1.91)Dummy variable year 2006 -5.3861∗∗∗

(-4.42)

industry controlyear control

intercept -17.2309(-1.64)

model Fixedhausman 50.37

p-value of hausman 0.00overall R2 (%) 11.01%

F test 4.18p-value of F test 0.0000

N 858

Note: In this table, fixed effects model is used to explain the impacted factors to the investment of foreign institutionalinvestors in China’s Stock Market. The dependent variable is the % relative spread of foreign institutional investorsin each company. The independent variables are financial and corporate governance firm-level characteristics selectedfrom table 8. We also include the industrial and year dummy variables to control the industrial and year effectrespectively. For each independent variable, the table reports the coefficient above and the t-statistic in parenthesisbelow. The superscripts of coefficient denote the significance of the coefficient(∗∗∗ ,∗∗,∗ represent significance at1%,5%, and 10% levels respectively). We report the intercept, t-statistics of the intercept in parenthesis below, themodel type in this table, the hausman test value, p-value of hausman test and number of observations at the end oftable 9. The specification test devised by Hausman (1978) tests the null hypothesis that the coefficients estimatedby the efficient random effects estimator are the same as the ones estimated by the consistent fixed effects estimator.If they are insignificant, then it is safe to use random effects. If you get a significant P-value, however, you shoulduse fixed effects.

39

Table 10: Preference of Domestic Funds

Classification Indicator

Size LN of capitalization 0.0396∗∗∗

(3.57)Short-term liabilities paying capability current ratio 0.0077∗

(1.73)Long-term liabilities paying capability long-term liabilities ratio -0.0212

(-0.63)Risk degree of financial leverage -0.0039

(-1.01)Return on shareholders earnings per share 0.0740∗∗∗

(3.23)Profit ROE -0.0248

(-0.38)Cash flow free cash flow 0.0002

(0.51)Growth growth rate of total assets -0.0058

(-0.48)Operating capability turnover rate of receivables 0.0001

(0.78)Ownership structure percentage of state-owned shares -0.0769∗∗∗

(-2.80)Ownership concentration Z index -0.0024∗∗∗

(-6.93)Structure of management number of directors 0.0068∗∗

(2.04)duality of chairman and CEO 0.0065

(0.50)percentage of independent directors 0.1969∗∗

(2.11)Compensation of management ln of sum of top three compensation of senior executives -0.0043

(-0.45)

industry controlyear control

intercept -0.9145∗∗∗

(-3.33)model Fixed

hausman 116.96p-value of hausman 0.0000

overall R2 (%) 5.32%F test 7.81

p-value of F test 0.0000N 6,793

Note: In this table, fixed effects model is used to explain the impacted factors to the investment of domestic funds inChina’s Stock Market. The dependent variable is the % relative spread of domestic funds in each company. In orderto compare with the previous results, the independent variables are financial and corporate governance firm-levelcharacteristics as same as those used in table 9. We also include the industrial and year dummy variables to controlthe industrial and year effect respectively. For each independent variable, the table reports the coefficient above andthe t-statistic in parenthesis below. The superscripts of coefficient denote the significance of the coefficient(∗∗∗ ,∗∗,∗

represent significance at 1%,5%, and 10% levels respectively). We report the intercept, t-statistics of the interceptin parenthesis below, the model type in this table, the hausman test value, p-value of hausman test and numberof observations at the end of table 10. The specification test devised by Hausman (1978) tests the null hypothesisthat the coefficients estimated by the efficient random effects estimator are the same as the ones estimated by theconsistent fixed effects estimator. If they are insignificant, then it is safe to use random effects. If you get a significantP-value, however, you should use fixed effects.

40