The effect of institutional distance on cross-listing performance. A study of Euronext cross-listing portfolios.

Masters’ Thesis IFM

The effect of institutional distance on cross-listing performance. A study of Euronext cross-listing portfolios.

Student number: 2784173 Name: Tjeerd Jeroense Study Programme: MSc IFM Faculty of Economics and Business

University of Groningen January 2017

supervisor: Dr. W. Westerman

Field Key Words: (Cross-listing, Institutional Distance, Capital markets,

Internationalization)

Abstract

The internationalization of firms and the globalization of markets expand the scope of view of firms beyond country borders in their quest for capital alternatives. The focus lies on the effect of institutional distance on cross-listing performance specifically, thereby it is hypothesized that this effect is negative. By observing a sample of cross- listing portfolios and firm performance belonging to Euronext firms collected from the Datastream database, and a disaggregated proxy of institutional distance from existent literature, an effective research design was created which was analyzed by pooled OLS. While controlling for certain firm attributes, the results provide

significant coefficients that suggest a negative and positive effect of some dimensions of institutional distance on cross-listing performance. Therefor evidence for merely an association is found and the hypothesis is partially supported. Since not all

dimensions turned out significant, future research should focus on the further

investigation of the hypothesized effect of institutional distance by expanding the

scope of the sample with more firms, countries and time series.

1. Introduction

The globalization put forward integration of markets. The markets for products and services, but also the markets for capital and that make it easier for firms to acquire capital across borders (Bel et al., 2012).

Firms seeking finance in other countries (institutions) have to make strategic decisions regarding the location and the method for attaining their capital on foreign markets. One of these options is to partake in cross-listing on multiple stock

exchanges to acquire additional equity (Bel et al,. 2012; Pagano, Roell, and Zechner, 2002). Firms have a choice in this, since cross-listing in another stock exchange is not mandatory. Next to the equity market, there is also a debt market and a market for venture capital (VC) (Bel et. al, 2012). Moreover, the amount of cross-listing and location seem to differ as well (Banalieva and Robertson, 2010). Thus there is some diversity in cross-listing activity.

The differences between countries along the institutional dimensions create costs for Multinational Enterprises or (MNEs). Zaheer (1995) calls this the Liability Of Foreignness (LOF). LOF is the particular cost of doing business abroad incurred by foreign MNEs originating from spatial distance, unfamiliarity with the local environment, host country legitimacy costs and home country restrictions (Zaheer, 1995:343).

However, from this logic it can be suggested that firms seeking finance abroad can face the same difficulties in foreign capital markets. Information flows are

important determinants of cross-border capital transactions (Portes and Rey, 2005).

Moreover, whereas transactions in product markets are characterized by a short

relationship between buyer and seller, transactions in capital markets are more

longitudional by nature (Bel et al., 2012). Therefore, like LOF concerns information

assymetries, and legitimacy (Zaheer, 1995), so do costs for firms facing barriers in

foreign capital markets (Bel et al. 2012; Blass and Yafeh, 2001; Portes and Rey,

2005). Nevertheless, some studies argue that cross-listing diversity can have a

decreasing effect on costs (Banalieva and Robertson, 2010; Pagano et al. 2002,

Sirkissian and Schill, 2004). Hence, there seems to be a positive and negative

influence of institutional distance on cross-listing performance.

The existing literature still has not directly linked institutional distance between listings in the firms’ portfolio with post-cross listing firm performance. Banalieva and Robertson (2010) aim to touch upon this relationship by using economic freedom as a proxy for institutional diversity between listing countries. However, the existent literature on institutional distance (Scott, 1995; Zaheer, 1995) suggests that the effect of institutional diversity within cross-listing portfolios on firm value can even be explained with other proxies than by only using the proxy of economic freedom. This research will try to contribute to the literature by filling that gap as well as using a sample of firms cross-listing from Euronext stock exchanges. This led to the following central research question appropriate for this research.

Does institutional diversity within the firm’s cross-listing portfolio affect the post- crosslisting performance of firms cross-listing from Euronext stock exchanges?

This question will be answered by answering the following research questions:

(1) What is institutional distance, and what is its link with firms? (2) What is cross- listing and what is its link with firms? (3) Is there an effect of institutional distance on cross-listing performance and how? (4) Does this effect change over time?

In the next section an extensive literature review will provide a summary of the body

of literature that acts as the fundament on which the hypothesis was formed. The

subsequent section covers the research methods in which the sample, measurements

and data analysis is discussed. This is followed by the presentation of the results and a

discussion in section four. The final section includes contribution, recommendations

and implications.

2. Literature review 2.1 Institutional Distance

2.1.1. Institutions

Institutions are seen as ‘the rules of the game’ and constrains created by humans that shape human interaction, and reduce transaction costs in the market (North, 1992).

Institutions differ along a cognitive, normative and regulatory dimension (Scott, 1995). Cognitive rules refer to societies’ information collection, processing and interpretation; Normative rules determine the things that should be done in a country, like norms and values. Finally, regulative rules state what can be done in the country, the exact law and rules that need to be obeyed in a country.

2.1.2. Institutions & firms

Existing literature suggests that the institutional distance (the differences between countries’ normative, regulative and cognitive dimension can form a resistance for implementing strategies abroad (Moore et al., 2015; Zaheer, 1995). Zaheer (1995) refers to this as Liability Of Foreignness (LOF). LOF is the particular cost of doing business abroad incurred by foreign MNEs originating from spatial distance, unfamiliarity with the local environment, host country legitimacy costs and home country restrictions.

2.2. Cross-listing

2.2.1. The act of cross-listing

With the globalization of product and service markets, financial markets experienced

a similar liberation (Pagano et al. 2002) in that cross-listing became more available

for (multinational) firms seeking equity to finance their (international) operations as

well as for stock exchanges who are seeking foreign listings. The direct costs for

involving a company into cross-listing varies, but it is expensive (Banalieva and

Robertson, 2010) and there is evidence that the cross-listing fees for American

markets can be substantial, not to mention the managerial, legal and accounting costs

of arranging such a listing (Licht, 2003). With these costs it certainly has to be worth

it for the firm, otherwise it would not involve itself in this activity.

2.2.2. Reasons for cross-listing

Not necessarily stated as a reason for cross-listing but, at least a side effect of cross- listing is that it provides firms with more experience and learning regarding global transactions and contracts, because managing a cross-listing is a difficult task (Licht, 2003). It is needless to say however that following that logic other global business transactions also provide experience and induce learning. This is however dependent on each firms’ management absorptive capacity. More to the point of strategy, Durant, Gunawan and Tarca, (2006) suggest that firms are either going abroad to raise capital or are in the game to attain debt. Other reasons are lower costs of capital (Banalieva and Robertson, 2010), new cheaper sources of capital (Biddle and

Saudogaran, 1991: Pagano et al., 2002), better explicit visibility to foreign investors (Banalieva and Robertson, 2010; Coffee, 2002; Licht 2003), increase in stock liquidity (Kyle, 1985: Admati and Pfleiderer, 1988), increase of shareholder base (Coffee 1999; Foerster and Karolyi 1993, Karolyi, 1996;) , “bonding” for foreign investors’ trust (Banalieva and Robertson, 2010), and better valuation. As to the latter, Miller (1999) finds that investors respond positively on cross-listings. These benefits of better valuation and lower cost of capital are further enforced when multiple cross- listings are realized. The exposure to risk of foreign markets is therefore spread over multiple geographic areas (Banalieva and Robertson, 2010).

2.2.3. Bonding

“Bonding” and its anti-hypothesis “Avoiding” received tremendous attention in the cross-listing literature (Coffee, 1999, 2002; Doidge, 2004; Doidge et al., 2004; Stulz, 1999); they are effects of cross-listing, and are some of the reasons for firms to engage in cross-listing. These studies have focused mainly on the effects from cross- listing on American stock exchanges, markets characterized by having very high disclosure standards. Licht (2003) argues that bonding is used by firms to improve their corporate governance system. By means of cross-listing, the firm then adheres voluntarily to stricter regulation regarding minority protection in the host country. A better corporate governance system then results in extra benefits for minority

shareholders above the other conventional benefits of cross-listing. Adhering to

stricter corporate governance systems like (advanced investors, institutions and

technology) guarantees for effective monitoring of performance and management and

thus increases the investor confidence in that firm’s stock, which is better for firm

performance (Coffee, 2002; Stulz, 1999). Moreover, it is suggested that these benefits of bonding regarding investor confidence are also present if the listing was set-up without the intention of raising capital (Licht, 2003). However, the difficulty of using the bonding hypothesis in predicting cross-listing performance is that bonding effects and thereby the benefits are not even for everyone. Cross-listing involving “bonding”

requires firms to seek higher disclosure regimes. However, this also entails that firms already listed in a country with high disclosure standards like the United States (Doidge et al., 2004) cannot improve much in terms of better corporate governance.

In fact, for such firms it means that cross-listing in another country means adhering to equal or lower disclosure standards, which could give different signals to investors about firm quality. The evidence for these mixed signals are already shown by Coffee (1999), for instance where he argues that firms that cross-list and abide to the higher disclosure standards present in the United States will indeed increase firm value, however the effects are also the other way around. His study argues that negative returns ought to be expected when cross-listing from the United States to another country and that it may imply foggy (undetectable) insider behavior or other

misconduct according to investors, which is undesirable. Indeed, Durant et al. (2006) also find that Australian firms receive negative performance after cross-listing in the United States. They argue that these firms appear to use cross-listing as a final means of getting capital, because they cannot find it otherwise. This is a bad signal for investors. Therefore, investors judge this as a lower quality, which is bad for firm performance. The negative judgment of investors stems from agency theory and the corporate governance system. Since managers have a fiduciary duty they have to act in all shareholders’ interests. However, managers also have their own interests.

Adhering to lower disclosure standards is seen by investors as an act by managers to avoid regulation, neglecting their fiduciary duty and conduct managerial opportunism.

This link between cross-listing in lower disclosure standards and the resulting investor distrust and negative firm performance is called the “Avoiding” hypothesis.

2.2.4. Avoiding

Licht, (2003) brings forward the agency problem that resides between management

and the investor of the firm. Cross-listing is a managerial decision, an activity that

either benefits the growth of the company or the expropriation power of the

controlling shareholders. It is expected that in the case of bonding, controlling

shareholders or managers of the firm will cross-list in countries with regulations for high disclosure standards (high protection of minority shareholders). These particular countries are chosen by them to benefit from the financial benefits or business

opportunities that listing there could generate for the firm, and by doing this they give up their own private benefits for the sake of firm growth. However, it may also imply that if the controlling shareholders get the chance to expropriate and follow their private benefits, they will and use cross-listing as a piggyback ride towards avoiding regulation. Coffee, (2002) acknowledges this avoiding for expropriation. The

governance structure and power of controlling shareholders in a firm will affect the decision to opt for cross-listing in a particular country.

2.2.5. Stock Exchange heterogeneity

Coffee (1999) suggests in his study that while the standards of protecting minority shareholders is based on legal rules, the systems underlying the differences in corporate governance (dispersed versus centralized ownership) is a normative one, where investors depend on relationships in centralized ownership rather than

regulative rules (law) in the dispersed system. It is relevant to note that both systems differ among each other, and that both systems have preferred countries to list in.

Firms from Common Law countries prefer cross-listing in countries with high- disclosure regimes and Code Law firms rather cross-list in low-disclosure countries.

Coffee (2002) and Licht (2003) suggest that both type of firms want liquidity benefits, but the Common Law firms are not necessarily looking for minority shareholders. It is therefore important to realize that these different corporate governance structures can have a huge impact on cross-listing decisions. Furthermore, Reese and Weisbach, (2002) find that firms from low-disclosure markets cross-list in high disclosure markets, but subsequent equity issues are done outside the United States. They find that firms from Common Law countries cross-list in the United States for other reasons than firms from Civil Law countries do. Moreover, Licht (2003) found that some firms get certain exceptions in adhering to the rules by the SEC providing stronger evidence for the existence of an “avoiding” hypothesis as well. Based on several firm characteristics and country characteristics firms might “Bond” or

“Avoid” but the literature provides mixed results in predicting cross-listing

performance. Where bonding/avoiding is related to firm performance through the

level of firm quality perceived by investors, another study suggests (Pagano et al.,

2002) that these attention-seeking efforts towards investors ripple far beyond the investor itself, as it also has a legitimacy increasing effect for other potential

stakeholders of the firm. The various stakeholders of a firm in the host market follow the judgments of foreign investors on the particular firm. This puts more weight on the investors’ perception of firm quality and can hence have both a positive and negative effect on the firms legitimacy and therefore firm performance.

2.2.6. Cross-listing destination

In deciding whether to cross-list or not the influence of location specific factors are weighted in as well. Licht, (2003) mentions that reasons for choosing specific cross- list locations include the access to cheaper financing and better visibility. Better visibility for foreign investors, but also for potential customers as it can be used as a platform to communicate product and service information producing more demand for the goods/services.

Pagano et al. (2002), focus on differences among cross-listings between Europe and the United States and find that there are many cross-listings from Europe to the United States, but not so much the other way around. They also observe that the United States have a more liquid market, higher accounting standards and a stricter minority shareholder rights’ protection as well as lower trading costs including direct costs, listing charges, fees for professional advice, but also the cost of legislation and adhering to foreign accounting standards. Furthermore the same study reveals that locations where analysts in the industry have superior knowledge are attracting cross- listing since it lowers information asymmetries thereby increasing investor attention and thus improves firm value.

Furthermore, several studies (Grinblatt and Keloharju, 2001; Sarkissian and Schill,

2004) find specific strategies for firm choice of cross-listing destinations. The

preference goes towards geographically and culturally (language, colonial ties),

economically (heavy bilateral trade), and industrial base close countries. This is

because their potential investors prefer shares whom with they are familiar with and

the firms would like to adhere to those investor preferences and close countries have

lower familiarity costs to overcome. This is in line with the investor recognition

hypothesis of Merton (1987: 506) “each investor knows some securities and does not

invest in securities he knows nothing about”. When firms cross-list they get into the

crosshairs of investors because they become familiar, and then get invested in.

This suggests that specific country factors positively affect the cross-listing performance for the firm. However, other studies argue that Cross-listers seem to be disadvantaged in the foreign capital market (Bel et al. 2012; Blass and Yafeh, 2001;

Portes and Rey, 2005). Cross-listing firms face a higher cost of capital, lower liquidity, less capital coverage and more unfamiliarity among foreign investors than host country firms. To reduce this unfamiliarity and to reduce this legitimacy deficit from the information asymmetries the MNE has to incur more costs in spite of performance. Therefore a set of multiple strong institutional diverging listings will increase these costs with negative effects on firm performance. Others also find that cross-listing may not give a clear positive signal about the firm (Coffee, 1999; Durant et al., 2006) depending on various country factors (Karolyi, 1996). They find that firms cross-listing from the United States and Australia result in negative cross-listing performance. Banalieva and Robertson (2010) find a negative relation between

institutional diversity in cross-listing portfolios and firm performance.

2.3. The link between Institutional Distance and cross-listing performance

In search for evidence for and relationship between institutional distance and cross- listing performance several studies give partial support for this connection. Sundaran and Black (1992) suggest that the organizational form of the firm is influenced by the aspects of the environment in which it undertakes transactions, from this logic it follows that the institutional environment (cognitive, normative and or regulative) (Scott, 1995) for each country its own will influence the transactions of the firm.

Furthermore, although Stulz (1999) does not suggest that the degree of institutional differences between the countries’ legal regimes effect the ability to trade shares, the difference itself is considered a trade barrier effecting firm performance after cross- listing, which relates to an effect between regulative dimensions or explicit laws (Scott, 1995) between countries affecting cross-listing firm performance. There are however more studies suggesting evidence for a more specific direction of the relation.

2.3.1. A negative relation

In the process of acquiring capital, firms face a “home bias” from investors in foreign

countries (French and Poterba ,1991; Tesar and Werner, 1995), which raises the cost

of capital for them. For example, Ferreira and Matos (2008) found that institutional investors prefer stocks from countries with strong disclosure standards. Foreign firms face information asymmetries with the local investors. This is often referred to as the reason for the “home bias” (Portes and Rey, 2005). Others call it institutional barriers that cross-listing firms face resulting in problems they cannot oversee or mediate (Sundaran and Black, 1992), or political differences between listing destinations that impede trade of shares (Stulz, 1999). This also raises the cost of capital for the firm.

Across all dimensions, there are examples of problems in getting correct information.

Furthermore, firms face problems in acquiring legitimacy in acquiring financing in foreign markets, what Schmidt and Sofka (2009) referred to as "legitimacy deficit".

The lack of legitimacy makes it necessary for the firm to put more effort to acquire this legitimacy, increasing the cost of capital. Chan et al. (2005) and Tesar and Werner (1995) acknowledge this and argue that information asymmetry, institutional distance, unfamiliarity and cultural distance create costs for firms acquiring finance in foreign capital markets. This is also referred to in (Bel et al., 2012) as capital market liability of foreignness (CMLOF). It can be than be said that cross-listing firms face certain institutional barriers, in every country they cross-list in, barriers that local companies do not incur. This distance will influence their total cost of capital.

However, institutional distance is not merely a part of these predictors, but it is a broader concept capturing and covering all four of these antecedents of foreign costs.

Firms listed on foreign stock exchanges therefore face a larger cost of capital than in home markets, and these costs vary depending on which country is chosen as a candidate for cross-listing (Scott, 1995; Zaheer, 1995).

2.3.2. A positive relation

However, regulation and the availability of (correct) information is vital for

transactions in the capital market (Bel et al., 2012), and cross-listing will, by allowing firms to tap into investment pools from various global markets, decrease investor risks (Banalieva and Robertson, 2010). Moreover, Cross-listing firms from low investor- protection countries can improve corporate governance systems by “bonding” by means of issuing shares in better investor protection countries. (Coffee, 1999). This lowers the cost of capital and therefore this shows that institutional distance,

differences between countries can have positive effects on cross-listing performance

as well. Furthermore, some studies suggest that cross-listing reduces the regulative

barrier for foreign investors, but also provides timely transparent information to local investors signaling that gaps in normative or cognitive dimensions are filled by cross- listing in that market (Pagano et al, 2002; Saudagaran, 1988).

2.3.3. Hypothesis

Although institutional distance seems to effect the decision for cross-listing both positively and negatively, a negative relationship is expected in the end. Previous empirical studies have shown a significant overall negative effect of institutional diversity on cross-listing performance (Banalieva and Robertson, 2010; Grinblatt and Keloharju, 2001; Sarkissian and Schill, 2004,). Since both bonding and avoiding give inconclusive effects on firm value, it is expected that Institutional distance negatively affects cross-listing performance. Hence, the hypothesis is as follows

H1: Institutional distance has a negative effect on cross-listing performance

3. Research design

This section covers the research design. It includes the sample, variables, measurements and the method of data analysis.

3.1. Databases

The research is of quantitative nature and made use of the database Datastream to gather the data for the dependent variable “Cross-listing performance” and the three control variables. The independent variable cultural distance is measured with the Hofstede values and the other nine independent variables of institutional distance are measured with the data gathered from the (Berry et al., 2010) study.

3.2. Sample

3.2.1. Firms

Following Saudagaran and Biddle (1992), it is opted to choose internationally traded firms listed in other stock exchanges. However since this research is also interested in countries other than the biggest capital markets the sample will include more than the eight countries from their study. Banalieva and Robertson (2010) used thirteen Triad countries to represent the population of firms from developed countries, since these firms are considered to be able to capitalize successfully on their resources to raise foreign capital. This research focuses on the non-financial cross-listing firms home- listed on two Euronext stock exchange countries as recorded by Datastream. The Euronext countries in the sample are the Netherlands and Belgium. At the best of the researchers’ knowledge no prior research on cross-listing diversity has used the Euronext stock exchange as focus, therefore this setting will contribute to existent literature even further. Moreover, geographical nearness to the Euronext firms allows for effective data collection in situations where direct contact with firms is required.

Furthermore, by using Euronext exchanges there is a significant pool of firms with one single currency, the Euro, minimizing data misinterpretation.

3.2.2 Sample size

The procedure used to arrive at the final sample size is described below.

3.2.2.1. Filtering. A list containing all the firms listed on the Euronext

countries (Netherlands, Belgium and France) was obtained from Datastream. From all of these firms, only the firms were selected that were indeed cross-listing to another country, here “Other” and “Over The Counter” were not considered as cross-listing locations and were neglected. Moreover, only those firms that were cross-listing from the Netherlands, Belgium or France were selected. Finally, like in (Banalieva and Robertson, 2010) the sample included only those firms that were operating in the non- financial industries, because it is argued that financial firms can have a different motivation for cross-listing than non-financial firms could have. To make this distinction the firms’ industries were identified along six industry coded provided by Datastream (code WC06010); 1. Industry, 2. Utilities, 3. Transport, 4. Banks/savings - loan, 5. Insurance and 6. Other financial. Firms belonging to classes 4, 5 and 6 where excluded from the sample. After applying these filters 102 firms remained. 55 from the Netherlands, 43 from Belgium and four from France, and these were operating in either 1. Industry, 2. Utilities or 3. Transport.

3.2.2.2. Missing values. Any firm that missed data along one of the

(in)dependent or control variables has been omitted from the analysis. Therefore, the sample is further reduced to a common sample of 54 firms, of which 29 firms are from the Netherlands, 22 from Belgium and three from France operating in either 1.

Industry or 2. Utilities.

3.2.2.3. Time frame. The set included three years of data between 2009 and 2011. This specific scope was chosen because the data from Berry et al. (2010) only covered the years up until 2011. The specific three year scope was taken from both the Banalieva and Robertson, 2010) study as well as the (Durant et al., 2006) study.

Given this underrepresentation of firms from France in the distribution, the firms from

France were excluded from the dataset. The final dataset therefore included 29 firms

from the Netherlands and 22 firms from Belgium, measured over three years resulting

in a common sample of 144 observations.

3.3. Variables

3.3.1. The dependent variable

The dependent variable of this research is the firm performance measured after cross- listing measured from 2009. The dependent variable, cross-listing performance, is measured using the yearly Return on Assets (ROA) score of the firm observed in the years 2009 until 2011. This allows for measuring the ability of the firm to transform its means (assets) into profits (Banalieva and Robertson, 2010).

3.3.2. The independent variables

The independent variables are the nine dimensions of institutional distance as proposed by (Berry et al., 2010). It comprises out of nine dimensions and the dataset is available for use in other studies. The nine dimensions ultimately capture the different dimensions of institutional distance.

The measure of Berry et al., (2010) is used because it tries to measure institutional distance along the dimensions: economic, financial, political, administrative, cultural, demographic, knowledge, global connectedness and geography. They developed nine dimensions, which is more comprehensive then other measures developed in this field (Ando and Paik, 2013; Elango et al., 2013; Moore et al., 2015). Moreover, the data constituting this measure is readily available. The datasets are comprised out of distances between countries and are all time varying except for administrative distance, cultural distance and geographic distance.

3.3.3.1. Economic distance. This dimension correlates with the macroeconomic stability of a country, the consumer purchasing power and

preferences and the openness of the country to external influences. This dimensions’

score is built up from a country’s inflation, income, imports and exports. The data

3.3.3.2. Financial distance. Berry (2010) links this dimension with corporate governance, foreign investment and corporate acquisitions. This dimension is created from scores on private credit, stock market capitalization and number of listed

companies.

3.3.3.3. Political distance. This distances correlates with choice of foreign location, entry mode and Foreign Direct Investment (FDI) flows. This distance is made up from policymaking uncertainty, democratic character, size of the state, WTO membership and regional trade agreements.

3.3.3.4. Administrative distance. Measuring bureaucratic patterns, this dimension is constructed from the colonizer-colonized link, common language, common legal system and common religion.

3.3.3.5. Cultural distance. The cultural distance within the cross-listing

portfolios was measured using the original values from Hofstede, and not the measure created by Berry et al. (2010). It was computed with the Kogut and Singh (1988:422) formula (see Figure 1). This formula is used, because of its widespread use in

calculating cultural distance (Ando and Paik, 2013; Elango et al., 2013; Moore et al., 2015). This would improve the generalizability of the research. Index Scores along four dimensions of cultural distance were adopted from Hofstede (Hofstede, 2001).

The dimensions are power distance, individualism, masculinity and uncertainty avoidance. Per dimension i, the difference between the index score I of the firm’s home country j and the score of the firm’s cross-listing (host) country u was calculated and squared. Since the formula is designed to calculate the distance between two countries, this step had to be repeated for each country the firm was cross-listed in. These scores were summed and divided over the total variance V of the ith dimension of all the firms in the sample. After this, the average distance score of one dimension of a firm’s portfolio is known. This procedure is then repeated for all four dimensions after which all firms will have four dimension scores. For each firm the average of these scores is the cultural distance within their cross-listing portfolio.

Figure 1.

The formula for cultural distance.

This is the formula for calculating cultural distance as developed by Kogut

and Singh (1988).

3.3.3.6. Demographic distance. This distance would portray patterns of international corporate expansion and share prices. This measure consists of life expectancy, birth rate, population under the age of 14 years, and the population under the age of 65 years.

3.3.3.7. Knowledge distance. This distance is the indicator of the capacity to create knowledge and to innovate. It is built up from the amount of patents in the country and the amount of scientific articles per capita.

3.3.3.8. Global connectedness distance. This indicates the ability of the country’s residents and companies to communicate with the rest of the world, collect information from- and diffuse to this network. It consists of tourism expenditure and receipts over GDP, and internet use per 1000 persons.

3.3.3.9. Geographic distance. Geographic distance affects the costs of

communication and transportation. This distance is computed out of the difference in the great circle distance between each countries’ geographic center.

The institutional distance scores between the home-country and each of its cross- listings in the host-countries where summed. When several different listings in one country occurred, this distance was multiplied with the amount of listings in the particular country. This summation of distances between home and host are chosen, as countries have their own set of institutional distance scores. This is expected not to interfere with institutional distance felt between the home country and another host- country, which would be the case if it was to be averaged over the total amount of host country listings.

3.3.3. Controls

There are certain factors driving cross-listing found in earlier research that have to be controlled for. Multinationality and firm size are adopted from Banalieva and

Robertson (2010). According to the study both variables explain why a firm is

successful in capitalizing on a cross-listing and increase their performance. This is

complemented with a third control, operating industry, and is also partially derived

from (Banalieva and Robertson, 2010).

3.3.3.1. Firm size. Large firms are argued to be more visible for investors than smaller firms. Since there are firms of all sizes in the sample this control is included.

Following Banalieva and Robertson, (2010) it is measured by taking the natural logarithm of the total sales (in euro) for each firm for each year.

3.3.3.2. Multinationality. The higher the internationalization of its market penetration, the higher the chance of cross-listing and better firm performance. Given the various operating industries and size of the firms, this control seemed appropriate as well. The level of internationalization of each firm was measured by its foreign sales to total sales ratio.

3.3.3.3. Industry. Banalieva and Robertson (2010) use industry diversification as a control. However, since this paper is focused on the cross-listing portfolio of firms and not so much the organizational structure of the firm, the general industry classification was chosen as control. Hence, industry specific variation due to trends like technological advancement or fashion are controlled for. This is a categorical variable and as such it will be included by using dummy coding for correct interpretation in the data analysis. The classification code of each firm’s general industry class was collected from Datastream. As previously discussed in the paragraph on the sample size, the financial firms where excluded. Moreover, the transport industry is excluded due to missing data. The remaining industries in the sample are “Industry” and “Utilities”. The base industry is “Industry” and this is compared to “Utilities” (Field, 2009: 254).

3.4. Data analysis

Since the nature of the data is Panel, meaning cross-sectional over multiple time series, the statistical inference must be done either through Pooled OLS, Fixed effects or Random effects. This research design is very similar to the design of Banalieva and Robertson (2010). They also have panel data, but used three equations with an 3SLS method focusing on three dependent variables, whereas this research solely focuses on one dependent variable in one regression analysis. They used a (within estimator) , fixed effects, whereas this study used a Pooled OLS estimation. According to

(Brooks, 2014) the first step of handling panel regression is to attain whether Fixed

effects could be favored over Pooled OLS, this before taking into account any random

effects estimators. By using the redundant fixed estimator from the statistical package

Eviews and the statistical test provided by the statistical package STATA it was

concluded that Pooled OLS should be favored in spite of fixed effects. Henceforth Pooled OLS regression was used to estimate the models, therefore according to the procedure (Brooks, 2014) determining fixed effects over random effects with a Hausman test, was then deemed unnecessary.

3.4.1. Robustness checks

The assumptions were checked and tests for serial correlation and heteroskedasticity were conducted. The option “White period” was used a covariance estimator provided by the statistical package EViews. It is a robust estimator that controls for serial correlation and heteroskedasticity, two violations of assumptions that are common in corporate finance models.

3.4.2. Backward enter method

The input of the variables in the regression was realized by using the Backward enter method. This is done because it was not hypothesized that all the institutional distance variables would affect the dependent variable, where the “enter” method would have been more appropriate. It was hypothesized that the total combined effect of all nine dimensions of institutional distance would have an effect. Therefor in this research it was opted to use a Backward enter method to stepwise eliminate the insignificant variables until the overall fit of the model (proxied with the adjusted R

) would not drop anymore.

3.4.3. Assumptions

In this research three assumptions for classical linear regression models are checked before interpretation of the data. First, the assumption of normality is checked. The Central Limit Theorem states that when a sample surpasses a size of 30 it can be considered normally distributed. Since the sample size (N) in this research is larger than 30, it is suggested there is enough evidence to conclude that the sample is normally distributed. The second and third assumption, heteroskedasticity and serial/

auto-correlation are violated, therefor robust standard errors are used to account for

both of these violations.

4. Results

This section includes the descriptive statistics of the sample and the results from the regression analysis followed by a discussion.

4.1. Descriptive statistics

The descriptive statistics between the variables are displayed in table 1. Table 1 shows that the average cross-listing firm in the sample has a positive return on assets of 4.58%. Furthermore, the average firm derives most of its sales from foreign sales, which indicates that the average firm in the sample is a multinational company (Banalieva and Robertson, 2010). The average firm is an “Industry” firm, and the average geographic distance between listing destinations in a firm’s portfolio is at least 1997 kilometers. For more information regarding the interaction between the variables please see the correlation table (Table 3) in the appendix.

Table 1.

Descriptive statistics.

This table displays the descriptive statistics of the sample. The panel consists of 51 firms. 22 firms are from Belgium and 29 firms are from the Netherlands. The time period is three years. This resulted in a total set of 144 (N) observations. (M) is the mean, and (SD) is the standard deviation.

Variables Common sample N= 144

Variables (M) (SD)

1. Performance (ROA) 4.580 7.157 2. Economic Distance 10.658 6.166 3. Financial Distance 11.561 10.815 4. Political Distance 215.731 210.552 5. Administrative Distance 25.736 18.681 6. Cultural Distance 6.692 5.180 7. Demographic Distance 7.756 13.214 8. Knowledge Distance 11.046 9.685 9 Global Connectedness Distance 1.766 1.310 10 Geographic Distance 1997.296 1801.945 11 Multinationality 70.001 24.482

12 Firm Size 16.385 2.331

13 Industry Utility dummy 0.181 0.386

4.2. Regression results

Now the outcome of the regression analysis will be discussed. Table 2 shows the results displayed in two models. Model 1 is a model specified regressing cross-listing performance with only the controls. Model 2 is specified with the controls and the independent variables altogether. The signs and significance will be discussed below.

Table 2.

The final regression results.

This table shows the results of the regression. The sample size (N) is 144. (B) is the regression coefficient. ((SE)) is the standard deviation. (T) is the t-value. (*) is significant at the 10% significance level. (**) is significant at the 5% significance level. (***) is significant at the 1% significance level.

As explained in the research design, the backward elimination model is used since it was hypothesized that the institutional distance construct in total consists of nine independent variables. However, the proxy for institutional distance discussed in Berry et al., (2010) was created to investigate different internationalization strategies.

Therefor the purpose/ relevance of the measure was not directly translatable to this research design. It is hypothesized that Institutional distance has an influence on

Independent variables Cross-listing performance 2009-2011 N = 144

Main effects B(SE) T B(SE) T

Cultural Distance -.404(.156) -2.598***

Demographic Distance .175(.047) 3.725***

Economic Distance -.387(.088) -4.378***

Geographic Distance -.003(.001) -2.922***

Knowledge Distance .452(.159) 2.850***

Controls

Firm Size .343(.483) .710 .833(.462) 1.800*

Multinationality .119(.041) 2.915*** .139(.039) 3.533***

Utility industry dummy 2.518(1.679) 1.500 3.884(1.570) 2.474***

Adjusted R

.086 .236

F-test (probability F=0) 0.001 0.000

cross-listing performance (ROA), however literature has yet to find evidence which underlying aspects of institutional distance affect ROA. Since there is no specific background literature, here instead of using an Enter method of variables clearly indicating expecting all to be fit for the model, this paper will use the backward model, thereby eliminating the overfit variables to proceed towards the variables creating the highest possible fitted model. This resulted in the elimination of the variables Political Distance, Financial Distance, Global Connectedness Distance and Administrative Distance and an overall R

of .236.

The regression was performed with a Pooled OLS estimator. However, using Pooled OLS assumes there is no heterogeneity (Brooks, 2014), whereas this is almost never the case in finance, therefore interpretation should be done with care.

4.2.1. Model 1

Firm size, multinationality and the utility industry dummy are regressed standalone in this model to check their respective explanatory power apart from the main

independent variables. The controls firm size and multinationality have a positive coefficient consistent with the literature (Banalieva and Robertson, 2010), however firm size was insignificant in the model. Interpretation of these coefficients is difficult, because they are not standardized. The utility industry dummy showed a positive beta, indicating a higher ROA for utility firms than industry firms, but

unfortunately the result was insignificant. Furthermore, the overall F-test is significant providing evidence that the inclusion of the controls significantly improved the

overall fit of the model compared to a model with only the constant.

4.2.2. Model 2

After the inclusion of the independent variables, the adjusted R

became .236. The controls firm size and multinationality were significant, and positive. However, multinationality was significant at the 99% confidence level, unlike firm size, which was only significant at the 90% confidence level. The utility industry dummy was significant at the 99% confidence level in this model. This suggests that the utility firms in the sample on average had a higher cross-listing performance than the industry firms.

Cultural distance, geographic distance and economic distance carry a negative

coefficient and are significant. This means that all others held constant, an increase in

cultural distance, geographic distance or economic distance will have a negative effect on cross-listing performance. Furthermore, both knowledge distance and demographic distance are positively correlated with cross-listing performance and are both significant. All others variables held constant, an increase in knowledge distance or demographic distance will increase the cross-listing performance. This means that not all dimensions of institutional distance in this model follow a coherent direction.

Finally, the significance of the overall F-test of the model indicates that there is enough evidence to suggest that also in this model the indicators included in the model improved the fit of the model in favor of a model with only the constant.

4.3. Hypothesis testing

With the data in this research five out of the nine original dimensions of Institutional Distance were significant. Therefor there is partial support for the hypothesis. Even when controlling for certain firm characteristics, there is evidence for a relationship between institutional distance and cross-listing performance, however the sign of the effect cannot be verified conclusively since some dimensions are insignificant and the remaining significant variables carry opposite coefficients.

4.4. Discussion

Now follows the discussion on the results. Together with background literature

regarding cross-listing and the measures of institutional distance possible explanations for the results are formed.

4.4.1. Insignificance of the other variables

The main hypothesis was partially supported as some dimensions of institutional distance are significantly affecting cross-listing performance, however not all variables constituting institutional distance in Berry et al. (2010) are significant.

Moreover, not all the predictor variables hold the same coefficient sign. This provides insufficient evidence to support a clear negative effecting relationship between

institutional distance and cross-listing distance. A possible explanation could be that

the explanatory power of the insignificant variables are captured in the prediction of

the significant variables. More statistical analysis outside of the scope of this research is necessary to verify this.

4.4.2. Negative effects

It is suggested that cultural distance, economic distance and geographic distance are negatively affecting cross-listing performance. This is in line with the information overflow hypothesis (Banalieva and Robertson, 2010), and the problems arising from institutional barriers that cannot be overseen and mediated (Sundaran and Black, 1992). It is hard to maintain a diversified portfolio in a set of countries that have different norms and values. Moreover, information overflows and managerial complexity increases as transaction in multiple institutionally different and geographically distant environments have to be managed at the same time.

Furthermore, when economic distance increases, it can be suggested that the trade flows between countries decreases, and that this decreased interaction between countries feeds the “home bias” of foreign investors in spite of the firms cross-listing from the more alienating country. Hence, this is also in line with the “home bias”

hypothesis (French and Poterba, 1991; Portes and Rey, 2005, Tesar and Werner, 1995), institutional investors prefer investing in high-disclosure markets (Ferreira and Matos, 2008) and there is evidence for CMLOF being present in financial markets (Bel et al., 2012). Especially cultural distance attributes to these costs since

information asymmetries in this dimension are rather abstract given the underlying cognitive and normative dimensions (Scott, 1995). Both the resulting imperfect information between firm and investors as well as the legitimacy deficit (Zaheer, 1995) are not easy to quantify in costs and firms may not even know how to surpass this hurdle (Sundaram and Black, 1992). This in turn decreases their ability to get explicit investors’ attention (Banalieva and Robertson, 2010; Coffee, 2002; Licht 2003), but also the expanded shareholder base from it (Coffee 1999; Foerster and Karolyi 1993; Karolyi, 1996;), which leads to lower cross-listing performance.

4.4.3. Positive effects

There is evidence found for a positive effect of both demographic distance and

knowledge distance on cross-listing performance. An explanation could be found in

the risk-diversifying benefits (multiple investment pools) that a cross-listing portfolio

could bear (Banalieva and Robertson, 2010) When knowledge distance or

demographic distance increases, arguably the differences in labor work force or technology also increase thereby enlarging the differential of information access between countries. This implies that having a very broad cross-listing portfolio

composed out of distant cross-list destinations would diversify the risk as arguably the chance of risk spill-overs are getting smaller that way. This also entails that investor pools are not alike and that certain cross-list locations would diversify the risks more than others. More research is necessary to investigate the existence of this

heterogeneity in investor pools and its effect on the institutional distance- cross-listing performance relationship.

Furthermore, the differences in information access increases the complexity of

managing the cross-listing portfolio, and the ability of a firm to coop with this

complexity provides a positive signal to foreign investors about a firm’s quality,

hence it could improve cross-listing performance.

5. Conclusion

This section presents the contributions, recommendations and implications.

5.1. Contribution

In conclusion, the results of the quantitative analysis provided evidence for partially supporting the main hypothesis. Some of the institutional dimensions are affecting cross-listing performance, therewith providing support for the existence of an effect of institutional distance on cross-listing performance even after controlling for certain firm specific factors. However, there was insufficient evidence found for the expected negative effect of this effect since some of dimensions in the model have contrasting coefficients. The suggestion of a partial effect indicates that the relationship between institutional distance and cross-listing performance is more complex and deserves more attention in future research.

5.1.1. Academic research.

This research contributes to future research by not using just major stock exchanges (Banalieva and Robertson, 2010), with panel data over three years measured.

Furthermore, to the reseacher’s knowledge, the sample consisting of Euronext

companies has never been used before in previous literature. Moreover, the concept of institutional distance has been disaggregated in its effect to cross-listing performance further explaining the dynamics of cross-country differences and their interplay in predicting cross-listing performance.

5.2 Recommendations

Since the sample captured Euronext firms on a majority of European cross-listings, one can posit that the overall characteristic of the sample was a “centralized

ownership” set of firms (Doidge et al., 2004). It is suggested that firms from code law

countries (Europe & Asia) prefer stock exchanges in code law countries. (Coffee,

2002). This suggestion is supported in this research since no firm in this sample cross-

listed in the United States, the country with very high disclosure standards (Doidge et

al., 2004) and only a few on the other common law country in Europe (the United

Kingdom). This clustering of Euronext listings cross-listing in European countries then also could give more evidence for the hypothesis discussed in (Sarkissian and Schill, 2004), where it is suggested that firms list in nearby countries (low

geographic-, economic-, cultural- and industry distances) as to overcome the small familiarity costs for attaining foreign investors’ attention.

Therefore, where most literature based on “bonding” is referring to the United States (Coffee, 1999, 2002; Doidge et al., 2004; Licht, 2003; Reese and Weisbach, 2002) or other Common Law countries (Coffee, 1999) one could argue that there is little evidence that bonding occurred during the cross-listings in this research. This is more in line with the reasoning discussed in (Coffee, 1999), low-disclosure firms rather cross-list on low-disclosure stock exchanges. This also suggests that avoiding was perhaps more prevailing in the sample (Licht, 2003), were “bonding” effects are non-existent, foreign investors may think that firms are avoiding certain country characteristics. Perhaps because they are desperate and have not other means to raise capital, which is a bad sign for investors and firm performance (Durant et al., 2006).

This reasoning could have more power in this research and could bias the negative results reported here, however further statistical analysis beyond the scope of this research has to verify this. Hence, future research is invited to investigate a larger sample over a larger scope of firms and countries and over more than three years.

Moreover, a different sample could also contribute to a further understanding of the research topic at hand. There was not enough evidence for a negative

relationship, but this could perhaps be observed by a set of different firms over different time periods.

Furthermore, this research could also not account for the absorptive capacity for each firm (Licht, 2003) resulting in different learning benefits and thus

management effectiveness and firm performance. It is recommended that future research takes this effect into account.

Another uncontrolled factor is the effectiveness of the marketing visibility and thus performance of marketing efforts (Licht, 2003), thereby influencing the overall performance of the firm. These benefits are arguably not the same for each firm. In future research it is better to control for this effect.

Also, the fact that there is no information regarding the stakeholder network of

each firm in the sample indicates that there is not yet a view on the magnitude of the

effect of negative or positive quality perceived by investors (Pagano et al., 2002).

This is a suggestion for future research.

The total costs for listing on each of the different stock exchanges can vary as well with the “race to the top” and “race to the bottom” of stock exchanges (Coffee, 1999, 2002). Different costs can attribute to different firm cross-listing performance, therefor uncertainty regarding these costs could results in biased interpretation. This has to be controlled for in future research.

Unfortunately, Datastream only has current data on cross-listing portfolios. It has no information regarding the portfolio between 2009 and 2011. For the purpose of this research, the portfolio’s are assumed to be constant in between 2009 and 2011, but future research should account for potential time differences in cross-listing portfolios.

There were some companies who have Over The Counter (OTC) and/or Other (OTH) listings in their portfolio. The author neglected these listings in the calculation of the institutional distance. It was assumed that Datastream had no sufficient

information on these listings. Banalieva and Robertson (2010) argue that making inferences on cross-listing among Triad countries is more practical. Future research should try to identify exactly what the other listings are and what the effect is of over the counter listings on cross-listing performance.

This research only implemented one measure of cross-listing performance (ROA). It is recommended to use multiples measures of cross-listing performance, like the return on sales or Tobin’s Q as cross-reference in future research.

Some institutional distance scores from Berry et al. (2010) were not available for all the years. Administrative distance, cultural distance and geographic distance were fixed over time. Perhaps for future research it is better to use measures of administrative distance and cultural distance that are not fixed over time, in order to investigate any time effects more appropriately.

In this research it was found that the various dimensions that constitute

institutional distance have contrasting relationships with cross-listing performance

and make the overall relationship more complex than previously hypothesized. Future

research could contribute by further investigating the overall sign of the relationship

to identify what other dimensions of institutional distance are affecting cross-listing

performance.

5.3. Implications

5.3.1. Practical implications for issuers

This result suggests that the positive effect of geographic, economic and cultural proximity on investor attention (Sarkissian and Schill, 2004) also exists between European countries. Firms that wish to increase their performance by cross-listing, should bear in mind that an increased geographic, cultural and economic proximity with the target countries could reap benefits for the firm, unlike the demographic and knowledge proximity, which decrease the firm performance in this research. This shows that the firm should carefully take into account in their cost-benefit decisions the various negative and positive aspects of proximity to the target country in their cross-listing portfolio.

5.3.2. Practical implications for investors

While firms have to take into account different costs and benefits so do the investors willing to acquire a stake in the firm. There are various indicators for firm value that indicate that valuating a firm is more complex than looking at just one dimension alone. Judging a firm’s attractiveness on its suggested acquired benefits of knowledge or demographic distance could mislead an investor in thinking that the barriers have been overcome. Other barriers such as different norms and values could convey a different message. This implies the importance of looking and valuating a firm from different perspectives.

Moreover, by taking a sample of more than 50 Euronext firms observed over three years this research finds that the overall destination of the Euronext firms was to other European countries and not to the US. Hence, there is more evidence for the clustering effect described in (Sarkissian and Schill, 2004), which conveys

information about a firm’s capital raising strategy for investors.

5.3.3. Practical implications for policy makers

One of the reasons that firms seek listing abroad is because certain institutional

barriers make raising capital or facilitating debt cheaper that way (Banalieva and

Robertson, 2010; Biddle and Saudogaran, 1991; Durant et al., 2006; Pagano et al.,

2002). If policy makers in the home country want to keep the firms listed solely in the

home country than it could try to make a more flexible credit regulation, or try to

develop policies that would let the firm bear similar risk to other countries, thereby

decreasing the incentives to go abroad. Policy makers in the foreign country could

initiate cross-listing by facilitating the mitigation of costs resulting from cultural

distance, economic distance or geographic distance since there is evidence that a

bigger distance between norms and values, trade flows and locations will decrease

cross-listing performance of the firm.

References

Admati, A.R., Pfleiderer, P. 1988. A theory of intraday trading patterns: volume and price variability. Review of Financial Studies 1, 3-40

Ando, N., Paik, Y., 2013. Institutional distance, host country and international business experience, and the use of parent country nationals. Human resource Management Journal 23(1), 52-71

Banalieva, E.R, Robertson, C.J., 2010. Performance, diversity, and multiplicity of foreign cross-listing portfolios. International Business Review 19, 531-547

Berry, H., Guillén, M.F., Zhou, N., 2010. An institutional approach to cross-national distance. Journal of International Business Studies 41(9), 1460-1480

Bell, R. G., Filatotchev, I., Rasheed, A.A., 2012. The liability of foreignness in capital markets: Sources and remedies. Journal of International Business Studies 43(2), 107- 122

Biddle, G.C., Saudagaran, S.M., 1991. Foreign stock listings: Benefits, costs, and the accounting policy dilemma. Accounting Horizons 5(3), 69-81

Blass, A., Yafeh, Y., 2001. Vagabond shoes longing to stray: Why foreign firms list in the United States. Journal of Banking & Finance 25(3), 555-572.

Brooks, C. 2014. Introductory Econometrics for Finance. Third edition. Cambridge University Press.

Chan, K., Covrig, V., Ng, L., 2005. What determines the domestic bias and foreign

bias? Evidence from mutual fund equity allocations worldwide. Journal of Finance

60(3), 1495-1534

Coffee, J.C., 1999. The future as history: The prospects for global convergence in corporate governance and its implications. Northwestern University Law Review 93(3), 641-708.

Coffee, J.C., 2002. Racing towards the Top? The Impact of Cross-Listings and Stock Market Competition on International Corporate Governance. Columbia Law Review 102(7), 1757-1831

Doidge, C., 2004. U.S. Cross-listings and the benefits of control: evidence from dual- class firms. Journal of Financial Economics 72, 519-553

Doidge, C., Karolyi, A., Stulz, R.M., 2004. Why are foreign firms listed in the U.S.

worth more? Journal of Financial Economics 71, 205-238

Durant, R.B., Gunawan, F., Tarca, A., 2006. Does cross-listing signal quality? Journal of Contemporary Accounting and Economics 2(2), 170-189

Elango, B., Lahiri, S., Kundu, S.K., 2013. How does firm experience and institutional distance impact ownership choice in high-technology acquisitions? RD Management 43(5), 501-516

Ferreira, M.A., Matos, P., 2008. The colors of investors' money: The role of

institutional investors around the world. Journal of Financial Economics 88(3), 499- 533.

Field, A. 2009. Discovering Statistics using SPSS. Sage publications LTD.

Foerster, S.R., Karolyi, G.A., 1993. The long-run performance of global equity

offerings. Journal of Financial and Quantitative Analysis 35, 499-528

French, K.R., Poterba, J.M., 1991. Investor diversification and international equity markets. The American Economic Review 81(2), 222-226.

Grinblatt, M., Keloharju, M., 2001. How Distance, Language, and Culture Influence Stockholdings and Trades. Journal of Finance 56(3), 1053-1073

Hofstede, G., 2001. Culture’s Consequences: Comparing Values, Behaviors,

Institutions, and Organizations Across Nations. Second Edition, Thousand Oaks CA:

Sage Publications

Karolyi, G.A., 1996. What happens to stocks that list shares abroad? A survey of the evidence and its managerial implications. Unpublished working paper, University of Western- Ontario, 1-46

Kogut, B., Singh, H., 1988. The Effect of National Culture on the Choice of Entry Mode. Journal of International Business Studies 19(3), 411-432

Kyle, A., 1985. Continuous auctions and insider trading. Econometrica 53, 1315-1335

Licht, A.N., 2003. Cross-listing and Corporate Governance: Bonding or Avoiding?

Chicago Journal of International Law 4(1), 141-163

Merton, R., 1987. A simple model of capital market equilibrium with incomplete information. Journal of Finance 42, 483-510

Miller, D.P., 1999. The market reaction to international cross-listings: Evidence from depositary receipts. Journal of Financial Economics 51(1), 103-123

Moore, C.B., Payne, G.T, Bell, R.G., Lewis, J.L., 2015. Institutional Distance and Cross-Border Venture Capital Investment Flows. Journal of Small Business Management 53(2), 482-500

North, D.C., 1992. Institutions and economic theory. The American Economist

(spring), 3-6

Pagano, M., Roell, A.A., Zehcner, J., 2002. The geography of equity listings: Why do European companies list abroad? Journal of Finance 57, 2651–2694

Portes, R., Rey, H., 2005. The determinants of cross-border equity flows. Journal of International Economics, 65(2), 269-296

Reese, W.A., Weisbach, M.S., 2002. Protection of minority shareholder interests, cross-listings in the United States, and subsequent equity offerings. Journal of financial economics 66, 65-104

Saudagaran, S.M., 1988. An empirical study of selected factors influencing the decision to list on foreign stock exchanges. Journal of International Business Studies 19(1), 101–127

Saudagaran, S.M., Biddle, G.C., 1992. Financial disclosure levels and stock foreign stock exchange listing decisions. Journal of International Financial

Management and Accounting 4, 106–148.

Saudagaran, S.M., Biddle, G.C., 1995. Foreign listing location: A study of MNEs and stock exchanges in eight countries, Journal of International Business Studies 26(2), 319–341.

Schmidt, T., Sofka, W., 2009. Liability of foreignness as a barrier to knowledge spillovers: Lost in translation? Journal of International Management 15(4), 460-474.

Scott, W.R., 1995. Institutions and Organizations. Ideas, Interests and Identities Thousand oaks CA: Sage.

Sirkissian, S., Schill, M.J., 2004. The Overseas Listing Decision: New Evidence Of Proximity Preference. The Review of Financial Studies 17(3), 769-809

Stulz, R.M., 1999. Globalization, Corporate Finance, and the Cost of Capital. Journal

of Applied Corporate Finance 12(3), 8-25

Tesar, L., Werner, I., 1995. Home bias and high turnover. Journal of International Money and Finance 14(4), 467-492

Zaheer, S., 1995. Overcoming the liability of foreignness. Academy of Management Journal 38(2), 341-363.

Appendix

Table 2

This is the correlation matrix of all the variables in the research. Bold indicates significance at 5%. The variables 10,11, 12 and 13 are the variables later removed in the actual regression.

Bold indicates significance at 5%. Variables 10, 11, 12 and 13 are the variables later removed in the actual regression model

variable 1 2 3 4 5 6 7 8 9 10 11 12 13

1 Performance (ROA) 1.000

2 Economic Distance -.019 1.000

3 Cultural Distance .017 .345 1.000

4 Demographic Distance .207 .643 .547 1.000

5 Knowledge Distance .052 .296 .641 .476 1.000

6 Geographic Distance -.091 .084 .513 .232 .880 1.000

7 Firm Size .008 .068 .243 .104 .292 .407 1.000

8 Multinationality .289 .111 .362 .231 .226 .064 -.311 1.000

9 Industry Utility dummy -.101 .050 -.207 -.087 -.149 -.025 .173 -.627 1.000

10 Financial Distance .051 .438 .887 .572 .437 .261 .095 .385 -.193 1.000

11 Political Distance .070 .307 .937 .482 .577 .424 .206 .401 -.195 .867 1.000

12 Administrative Distance .028 .357 .937 .561 .523 .373 .161 .281 -.056 .877 .921 1.000

13 Global Connect. Distance .089 .265 .861 .523 .385 .236 .192 .334 -.217 .880 .859 .895 1.000