financial mark

8/13/2019 financial mark

1/44

Financial Markets versus Institutions in European Countries:

Influence of Culture and Other National Characteristics

by

Raj Aggarwal

Sullivan Professor of International Business and Finance

University of Akron, Akron, OH 44325330-972-7442; [email protected]

and

John W. Goodell

College of Business AdministrationUniversity of Akron, Akron, OH 44325

330-972-5361,[email protected]

JEL Classifications: G10, G20, N20, O16Keywords: financial institutions, banks, financial markets, universal banks, comparative

financial systems, legal traditions, culture, power distance, uncertainty avoidance, socialtrust, property rights

December 2009


2/44

Financial Markets versus Institutions in European Countries:

Influence of Culture and other National CharacteristicsAbstract

While it is clear that some countries are bank oriented and others rely more on equity financing, there is

little research on the degree to which the national architecture for financial intermediation is determined

by legal, cultural, and other national characteristics. Using panel analysis of data for a recent eleven-year

period for nineteen major European countries, this paper documents that the architecture of financial

intermediation is influenced by national cultural, political, and economic factors. Specifically, we provide

robust evidence that a greater predilection for market financing over bank financing is associated withhigher levels of power distance, concentration in equity markets, control of corruption, efficiency of debt

enforcement; and the adoption of the euro. Lower predilection for equity financing is associated with an

English legal origin, greater uncertainty avoidance, and greater political legitimacy. Our results should be

of great interest to managers and policy makers and to scholars interested in the relationship of culture,

political economy, and financial intermediation.

I. Introduction

Financial intermediation, the transmission of funds from savers to investors, is a necessary

function in all countries and can be undertaken through financial institutions and/or via financial markets.

Either channel must necessarily resolve the matters of asymmetric information, adverse selection, moral

hazard, and agency costs involved in financing contracts that cover the monitoring and collection of funds

provided by savers to investors. All optimal contracts are incomplete and the costs and efficiency of

establishing contracts depends not only on the legal environment, but also on ethical and other informal

conventions, industrial structure, and social and cultural values.

Particularly, differing forms of intermediation may be better suited for certain societies with

hierarchical or autonomous orientations. This may result from the nature of the institutions of the political

economy or governance structures being synchronous with the hierarchical nature of particular cultures or

inevitably the systemically consistent nature of societies (Licht, Goldschmidt and Schwartz, 2007;


3/44

This paper examines, for European nations, the association of cultural measures and political

factors and predilection for markets over banking while controlling for relevant economic factors.

European financing system choices are assessed based on data from nineteen countries: Austria, Belgium,

Czech Republic, Denmark, Finland, France, Germany, Hungary Ireland, Italy, Netherlands, Norway,

Poland, Portugal, Russia, Spain, Sweden, Switzerland, and the UK. Annual estimates of relative size of

financing provided by markets and by institutions and other national characteristics are collected for the

recent eight year period, 19962006, that includes the introduction of a common currency, the euro.

Our focus on European markets is motivated by the unique role that banks traditionally have had

in European society. White (1998) notes that historically banks in continental Europe have maintained a

much larger share of intermediated claims than in other groups of countries, with corporate and retail

banking traditionally being overwhelmingly done by national entities, with surprisingly little cross-

border activity. The time period of this study straddles adoption of euro. The adoption of the euro has

likely been greatly impacting European banks and markets (see, Dermine and Hillion, 1999).

Panel analysis for our sample is used to assess the national characteristics that are likely to be

associated with a country being bank-based or markets-based. Specifically, using panel analysis on data

for a recent eleven-year period for nineteen major European countries, this paper documents robust

evidence that in Europe, a greater predilection for market financing is associated with higher levels of

power distance; as well as concentration in equity markets, control of corruption, efficiency of debt

enforcement and the adoption of the euro. Lower predilection for equity financing over bank financing is

associated with higher uncertainty avoidance, an English legal origin, and political legitimacy. As we

discuss below, our results point to both cultural-transactions costs and political-influence determinants of

choices for financing channels across Europe These results should be of great interest for those interested


4/44

Contributions

This paper examines for European countries the association of national cultural and institutional

characteristics with the societal choice between markets or banks while controlling for relevant economic

and other factors. This paper extends the work of Ergungor (2004) on the role of regulation in influencing

predilection for financial markets over banking and of Kwok and Tadesse (2006) on the role of culture in

the development of financial markets by presenting a more comprehensive model focusing specifically on

the determinants of national financing channels in European markets using a wider range of independent

variables that reflect our discussion of how political and cultural dimensions can influence national

financial architecture. Although economists have long noted that financing in some countries is market-

based while in others it is predominantly bank-based (Ndikumana, 2005), the cultural and other national

determinants of why financing in European countries are bank-based or market-based has been largely

ignored by the literature even though such knowledge would be important to managers and policy makers.

Particularly, the focus of our study is on the association of cultural factors such as power distance,

uncertainty avoidance, and individuality with European choices of financing channels while controlling

for other relevant determinants. Extending prior models, we uniquely control for concentration of market

capitalization using a Herfindahl index constructed for this paper and we also use a number of new

independent and control variables reflecting relevant political and economic factors which are described

below in the discussion of methodology. Further, this paper uses better methodological procedures and

more recent data than prior studies, and focuses on European countries given the history and uniqueness

of banking in Europe. This paper offers greater in-depth distinction amongst the influences of political

legitimacy, political stability, regulation, concentration of ownership and culture.

This paper contributes first by presenting a new extended model of how various cultural, social,


5/44

Legal, Social, and Cultural Factors in Financial Intermediation

Financial intermediation is a necessary function in all countries and can be undertaken through

financial institutions and/or financial markets. Financial intermediation involves the transmission of funds

from savers to investors with contracts defining the terms of the exchange, i.e., what can savers expect

from investors in return for postponing their consumption and how they may optimally enforce such

contracts. As such financing contracts involve monitoring and collection of funds provided by savers to

investors and result in a separation of responsibilities between the provider and users of funds, contractual

parties must resolve the resulting fiduciary problems of asymmetric information, adverse selection, moral

hazard, and agency costs, especially since (as noted by Hart (2001) and others) all optimal contracts are

incomplete.

Two contrasting mechanisms have evolved to solve the financial intermediation problems

associated with incomplete financing contracts, financial institutions, and financial markets. Both

mechanisms must perform the essential functions involved in financial intermediation, i.e., selection of

fund recipients, the design of financing contracts, monitoring of the fund recipients, and finally collection

of the returns from the financing activity. However, the relative efficiency and efficacy of each of these

functions and of the overall process of financial intermediation under incomplete contracting may be quite

different between the two mechanisms depending on the business, economic, legal, social, and cultural

environment. As noted earlier, Rajan and Zingales (1998) contend that the relative adoption of institutions

versus markets as channels for financial intermediation in a country may depend on the contractibility of

the environment and the relative value of price signals. Thus, the proportion of financing provided in a

given country by banks and other financial institutions versus that provided by financial markets can be

expected to depend on a number of factors including the nature and economic functions of these two


6/44

national economic, legal, social, political, and cultural characteristics determine the relative national

importance of financial institutions and markets.

Institutions versus markets for financial intermediation

Financial intermediation, whether through institutions or through markets, must necessarily

resolve the matters of asymmetric information, adverse selection, moral hazard, and agency costs

involved in financing contracts that cover the monitoring and collection of funds provided by savers to

investors. Further, as noted by Milgrom and Roberts (1990), levels of hierarchy affect costs, including

influence costs. Further, they note that under particular circumstances, more or less hierarchy is more or

less cost efficient.

As the above discussion suggests, the relative efficiencies of differing forms of intermediation is

influenced by the costs in ensuring contract reliability and the dissemination of political influence.

Therefore, the costs of intermediation in particular environments are heavily influenced by the relative

efficacy of hierarchy or markets. The transaction costs of intermediation occur through ensuring contract

reliability. Ensuring contract reliability can be achieved either through hierarchical, relationship-based

practices or through impersonal agents such as market forces, depending on the national characteristics

surrounding the respective market. Similarly, the efficacy of hierarchy or markets to facilitate political

influence depends on influencing costs and so on national characteristics. Further, institutions or markets

primarily have control over decisions that have distributional ramifications, and so the national choice of

intermediation resolves to how well do the hierarchical aspects of societies influence their respective

choices of financial architecture. As contrasted to financial markets, banking and other financial

institutions clearly reflect a hierarchical solution to the problems associated with financial intermediation

h hi hi l i i lik l f i b ki k f fi i l i di i


7/44

Hierarchy versus Market factors in financial intermediation

Contracting Costs:

Of course, as noted by Coase (1960), in a theoretically ideal financial system it would make no

difference whether financing was privately done through banks or publicly through markets. However, in

an imperfect market with non-zero transactions costs and costly information, other factors must also be

considered. Indeed, according to North (1990), the costliness of information needed for measurement and

enforcement of exchanges creates transaction costs. Transaction costs involve costs of defining

property rights and costs of enforcing contractsincluding costs of information. Transformation costs

are the costs associated with using technology and the efficiency of factor and product markets and are

reflected in transactions costs. Whether institutions lower or raise overall transactions costs has to do in

part with the ability of participants to be informed and to understand the nature of the particular

institutional environment. This includes not just understanding the nature of contracts and their

enforceability, but also the temperament and motivations of other participants.

In addition to institutions, transactions costs are also associated with market transactions. As

noted by Williamson (1988) (and others, e.g., Aggarwal and Zhao (2009)), Transaction Cost Economics

(TCE) suggests that when the costs of market exchange are sufficiently high, firms can obtain cheaper

financing through some other means. The alternative to market financing is typically through some sort of

a prescribed arrangement, such as a bank loan or, more broadly, through a prescribed transfer of resources

through a horizontal or vertical network. Williamson (1988) along with Hart (2001) and Hart (1995)

suggest that, from the point of view of the firm, the choice between debt financing versus equity financing

resolves to a choice between respectfully the costs of complying with rules versus the costs of ensuring

li bili H (2001) d H (1995) i h h i i f k


8/44

Transaction Costs and Contract Reliability:

Modigliani and Perotti (1998) theorize that when societies enforcement regimes are not

adequate, bank financing is favored. In this instance the binding of transactions becomes more private

than public but this binding is concomitant with firms having long-term relationships with banks.

Modigliani and Perotti (1998) suggest that when the rights of minority (or outside) investors are not

adequately protected; less equity investment will be available for new enterprises (see Myers, 1977).

According to Modigliani and Perotti (1998) in such societies, there will be more bank lending instead of

financing with public equity. Modigliani and Perotti (1998) also suggest that banks, because of an

emphasis on collateral, are less able to differentiate firms with good future prospects versus those with

poor future prospects. Alternatively, markets with good governance are better able to distinguish between

these types of firms. This view has been supported by recent literature. For instance, Shirai (2004) reports

that, because of improvements in official oversight for the period 19972001, Indian capital markets

improved significantly in being able to differentiate high-quality firms from low-quality firms.

As noted by Modigliani and Perotti (1998), and earlier by Rajan (1992), it is where contract

enforcement is weak that collateral is emphasized more, leading to an advantage for bank financing.

Underlying Rajan and Zingales (1998) is the notion that when reliable information about firms is too

difficult or costly for the general public to obtain, banks provide delegated monitoring (see Diamond,

1984). However, Rajan (1992) notes that the higher informational advantage for banks means that they

can charge excessively high interest rates; which can weaken economic development.

Culture and transaction costs:

It is reasonable to also consider that cultural dimensions have an effect on transaction costs.


9/44

negative association with cultural hierarchy. Licht, Goldschmidt and Schwartz (2005) suggest that

cultural hierarchy is likely to be associated with greater corruption and disregard for the law. Husted

(1999) finds power distance to have a positive impact on the level of corruption. And so, it might be

expected that contracting in more hierarchical societies, and/or in less democratic societies, will have

greater transaction costs and costs of asymmetric information. Similarly, it is also reasonable to expect an

association of uncertainty avoidance with higher transaction costs. It is natural to expect the costs of

asymmetric information to be greater in environments of greater uncertainty avoidance. Additionally,

Husted (1999) finds a positive association of uncertainty avoidance and corruption.

Political Influences on Financial Services

La Porta, Lopez-de-Silvanes and Shleifer (2002) note that governments can gain control of banks

in order to exchange favors for rents. Beck, Demirguc-Kunt and Levine (2006) note that politicians may

channel the flow of credit to politically connected firms, or powerful banks may capture politicians and

induce official supervisors to act in the interests of banks rather the interests of society. In general, it is

clear that banking is often undertaken as part of a broader network, in the sense that goods and services

are exchanged within a private consortium according to some sort of administrative directive. Allen

(2001) sees banking as providing a governance function in societies that have poor corporate governance.

As observed by Modigliani and Perotti (1998), governments have a great deal of power to determine the

rules of the game. Further, Perotti and Von Thadden (2006) note that the political influence on

governance, and therefore potentially on financial structure, is not limited to codified laws and note that

the polity can sway legislation regarding a host of issues that affect banks and shareholders. Corruption,

trade and labor regulations, the design of regulatory institutions, choices about the nature of enforcement,


10/44

to provide economic efficiency, but rather to serve the interests of those with the residual control rights

and if economies realize gains from trade by creating efficient institutions, it is only because

circumstances provide incentives for those with bargaining strength to alter institutions in ways that

coincidently turned out to be economically efficient. Wealthy families may play a significant role in both

firm ownership and in the nations polity, alleviating or substituting for the political role of banks (see

Roe, 2003; Morck, Strangeland and Yeung, 2000. Rajan and Zingales, 2003b) suggest that incumbents

form self-interest groups that influence the polity to impede new competition. The notion that special

interest groups can influence the polity to their advantage (e.g., Olson, 1982) or that incumbent firms can

lobby to impede healthy capitalist competition (e.g., Rajan and Zingales, 2003b) has also been well

discussed in recent literature.

Finance, Culture, and Political Influence

To our knowledge, little prior literature discusses the political influence on banking and markets

within a cultural context. It would be simple to presume, because of their relationship nature, that banks

would have a greater role in disseminating political services through intermediation in environments of

higher power index. Indeed Beck, et al. (2006) associate national corruption with greater governmental

authority over banking. As noted above, it should be considered that equity markets might also serve as

avenues of political service with state involvement extending beyond simply ownership.1

As this brief review suggests, there is much reason to suspect that national hierarchical tendencies

in finance are influenced by national cultural, social, legal, political, and economic characteristics. After

all, national differences in cultural factors have been shown to be very important in many areas of

international business including recently in exporter-distributor relations (Nes, Solberg and Silkoset,


11/44

2001), as the next section indicates, there has been much less focus on the role of such national factors in

determining national financing architecture with regard to relative reliance on banks versus markets for

financial intermediation.

Literature on the determinants of national financing choices

While Levine (1998) provides a connection between financial system design, laws and

regulations, Ergungor (2004) extends this to show that civil-law countries are more likely to be bank-

based. Ergungor (2004) finds that countries with civil-law traditions are more likely to be bank-based

arguing that this is a result of common-law countries being more adaptable in an environment of

incomplete contracts as civil-law courts are less effective than their common-law counterparts in

resolving conflicts because they have less flexibility in interpreting the laws and creating new rules.

Ergungor (2004) argues that in civil-law economies, banks emerge to act as primary contract enforcers

while common-law courts provide more detailed creditor and shareholder protection laws encouraging

common-law countries to have more developed financial markets compared with civil-law countries. It is

intuitively reasonable that legal origins may have an impact on making financial markets more reliable

and legal issues are closely related to mechanisms for undermining financial markets. Johnson, La Porta,

Lopez-de-Silvanes and Shleifer (2002) posit that in civil-law countries, as opposed to common-law

countries, the law favors tunneling by controlling shareholders. While outright expropriation of resources

by controlling shareholders is almost always illegal, other forms of tunneling are legal in many countries

through courts interpretation of incomplete contracts. Johnson, et al. (2002) assert that common law legal

systems are better able to police tunneling because legal authority is given more judicial discretion.

In an extension examining cultural factors that influence financing channels, Kwok and Tadesse

(2006) fi d h i h h hi h l l f i id lik l b b k


12/44

less social capital would favor institutions over markets because of banks use of collateral and because of

the long-term relationships that banks typically have with clients.2In a related extension, Luhmann (1982)

suggests that financial systems can not function without some minimum level of public trust. Fukuyama

(1995) suggests that societies that have less social trust available to bind contracts must do more

contracting with the cost of this additional contracting effectively acting as a tax. Others look to social

capital as enhancing social trust and thereby facilitating financial markets (e.g. Hong, Kubik and Stein,

2004; Guiso, Sapienza and Zingales, 2004). In the absence of adequate social capital or social trust,

relationship financing via banks is a natural substitute for public markets.

As this brief review of the relevant literature indicates, national financial architectures regarding

banking or markets as the preferred channels for financial intermediation are likely to depend not only on

the previously examined economic factors, but also on factors such as national cultural characteristics and

measures of political influence on the nature of financial intermediation. Prior literature has not examined

such a broad set of determinants for the national tendency to favor banking versus financial markets. The

research design described next reflects an attempt to overcome this gap in the prior literature.

III. Research Design

Based on Demirguc-Kunt and Levine (2001) we suggest that the assessment of national financing

predilection for markets versus institutions can and should be measured based on the relative sizes of

financial institutions and financial markets in a country. Size measures are more likely to be stable and

more representative of inter-national variations. Our dependent variable, therefore, is the domestic stock

market capitalization relative to domestic assets of deposit money banks. This relative measure is

constructed by dividing the ratio of domestic stock market capitalization relative to GDP and domestic


13/44

be seen in Table 1, Finland and Sweden are the most market oriented nations in our sample while

Germany, Portugal and Austria are the least market oriented.

(please insert Table 1 about here)

Kwok and Tadesse (2006) discuss at great length how and why culture may influence financial

systems. As in Kwok and Tadesse (2006), to account for cultural determinants of financing choices we

include as independent variables three measures of cultural characteristics from Hofstede (2001) that may

influence attitudes towards hierarchy in a society: the Power Distance Index (PDI), The Uncertainty

Avoidance Index (UAI) and the Index of individualism (IDV). Power distance describes the extent of

inequality in the social system of a nation. Hofstede (2001) argues that people exhibit a pattern of

dominance in the social order across cultures. Hofstede (2001) notes that power distance is related to the

extent to which the less powerful members of organizations and institutions accept and expect that

power is distributed unequally and that power distance is related to other forms of inequality such as

ethnic fractionalization and economic inequality. This represents an inequality that is endorsed by the

followers as much as by the leaders. Power Distance reflects faith in existing rules and so we hypothesize

that it favors market financing over bank financing. Hofstede (2001) also suggests that societies with

greater power distance will naturally have less political legitimacy, which means that such societies will

be less able to favor banks and would rely more on markets for financing.

According to Hofstede (2001), uncertainty avoidance reflects a society's tolerance for uncertainty

and ambiguity. Uncertainty avoiding cultures try to minimize the possibility of unstructured situations by

strict laws and rules, and safety and security measures. Higher uncertainty avoidance in a society clearly

indicates a desire to avoid markets compared to banks as markets present much more risk to participants.

Also as noted above Kwok and Tadesse (2006) find that uncertainty avoidance is a significant


14/44

after him/herself and his/her immediate family. Licht, et al. (2007) suggest that egalitarianism is likely

associated with greater rule of law and less corruption. Therefore, one might expect nations that are less

individualistic and more hierarchical or more collectivist would be more corrupt and thereby incur greater

transactions costs through market participation. So, we can expect higher individualism to favor markets

rather than institutions like banks.3

Hypotheses and MethodologyHypotheses related to culture as a partial determinant of financial architecture

We put forward the following hypotheses related to culture:

H1: Culture has a significant influence on financial architecture.

H1.1: Higher power distance in a society will favor financing via markets compared to banks.

H1.2: Higher uncertainty avoidance in a society will favor financing via banks compared to markets.

H1.3: Higher individualism in a society will favor financing via markets rather than banks.

Kwok and Tadesse (2006) document significant support for H1 and H1.2.To our knowledge, the

associations of individualism and power distance with financial architecture have not been documented in

prior literature.

In this study we control for the concentration of equity ownership. Kwok and Tadesse (2006) note

that market-based intermediation is better suited to countries populated with larger firms while bank-

based intermediation is more appropriate for countries characterized by a lot of smaller firms. According

to Tadesse (2002), markets are important for public pricing. We theorize that as market concentration

becomes more diffuse it becomes less efficient for the marketplace to disseminate information and banks

consequently assume this function.


15/44

difficult. Leland and Pyle (1977) also argue that informational asymmetries can be significantly assuaged

by the signaling effect of investment by known insiders or entities expected to have inside information. In

this sense then banks are information brokers (Ramakrishnan and Thakor, 1984). It is reasonable to

suppose that both these functions or information collection and verification of banks are greatly enhanced

by an environment of many smaller firms as opposed to one characterized by a few large firms. It is

simply more efficient for the marketplace to disseminate information about firms when market

capitalization is concentrated in fewer firms.

We also consider the possibility that market concentration is also important in controlling for the

potential for existing owners to influence respective nations polity. Perhaps governments are more

efficiently able to exercise political control over a small number of firms directly while a larger number of

smaller firms are more efficiently influenced through banks aligned to the polity. Given the nature of

political economy, it is appropriate to consider the influence of equity owners on nations polities

especially if such ownership is concentrated. It is natural to think that higher ownership concentration of

equity will allow a smaller set of owners to exert greater influence on the polity. Krueger (1974) notes

that government intervention in economic activities creates large political rents and, as noted earlier, La

Porta, et al. (2002) claim that governments can gain control of banks in order to exchange favors for rents

supporting the Beck, et al. (2006) contention that powerful banks may act as agents of the polity. In

addition, Fogel (2006) documents that family control of large businesses in a country is generally

associated with these families having great influence over the polity.

Hypotheses regarding market concentration as a determinant of financial architecture

To account for equity market concentration we establish a Herfindahl index for each country for

each year A Herfindahl index close to 1 would suggest that most market capitalization for a particular


16/44

firms favor banks. Market capitalization data is obtained from the Institutional Brokers Estimate System

(I/B/E/S) of Thomson Financial. Thus, we have the following hypotheses related to market concentration:

H2: Market concentration has a significant influence on financial architecture.

H2.1: Higher market concentration will favor financing via markets compared to banks.

Tadesse (2002) finds that countries dominated by large firms grow faster in market-based systems. Under

the broad assumption that societies migrate toward a financial architecture that provides better growth, it

could be argued therefore that Tadesse (2002) provides tangential support for H2 and H2.1.

Hypotheses for legal origin, politics, and governance as determinants of financial architecture

As discussed above, transaction costs of market participation are reduced if the contracts inherent

in market participation are made more secure. Consequently, we also include a measure of legal origin,

English or other (ENGLISH_LEGAL). This variable is a dummy variable that is assigned a value of 1 if

the country has an English legal origin (La Porta, Lopez-de-Silvanes and Shleifer (2006)). This variable is

included because some theorize that common-law systems offer better investor protection (e.g. Johnson,

et al. (2002)). Additionally, prior research, in particular Ergungor (2004), finds legal origin to be

significant in partially determining whether nations are either market- or bank-based. English legal origin

can strengthen the rights of minority owners reducing the benefits of control by the larger shareholders

and, thus, the advantages of financial markets. Our hypotheses are Ergungor (2004) and Kwok and

Tadesse (2006) document support forH3andH3.1):

H3: Legal origin has a significant influence on financial architecture.

H3.1: English legal origin will favor financing via markets compared to banks.

As a measure of governance quality, we include in our models independent variables regulatory

quality (REG QUAL) from the indicators of governance of Kaufmann Kraay and Mastruzzi (2008) It is


17/44

barriers to self-dealing will be more bank-based because banks are good monitors. Thus, our hypotheses

regarding regulatory quality are:

H4: Regulatory quality has a significant influence on financial architecture.

H4.1: Higher regulatory quality will favor financing via markets compared to banks.

Kwok and Tadesse (2006) include as independent variables an Institutions Index, which is an

amalgamation of measures from Kaufmann, Kraay and Zoido-Lobaton (1999), as well as a measure of

accounting standards from the Center for International Financial Analysis and Research (CIFAR). Kwok

and Tadesse (2006) find in some of their models association of greater efficiency of these variables with a

predilection for markets.

We also include a more pointed measure of regulatory efficiency, a factor for the efficiency of

debt enforcement from Djankov, Hart, McLiesh and Shleifer (2008) (DEBT_EFFICIENCY). This

measure is constructed from surveys conducted in a wide sample of countries and assesses such aspects of

debt enforcement as the amount of time and cost in the disposition of assets; as well as the structure of the

appeals process, floating charges, whether seniority can be abrogated etc. This variable captures not only

the efficiency of creditor rights enforcement in the case of a business in default, but as noted by Djankov,

et al. (2008), is also a strong predictor of private credit. And so the inclusion of this measure amongst the

independent variables helps to control for governance as well as general prevalence of private credit. To

our knowledge, the association of debt efficiency with financial architecture has not been studied before.

Our hypotheses regarding this variable are:

H5: Debt efficiency has a significant influence on financial architecture.

H5.1: Higher debt efficiency will favor financing via markets compared to banks.

We also control for political stability using estimates from Kaufmann et al (2008) As noted


18/44

welfare gains. Controlling for political legitimacy (POLITICAL_LEGITIMACY) is motivated by a desire

to distinguish between political stability, the ability of the polity to simply endure, and aspects of political

governance that are generally considered more evolved in the sense of democratic accountability and

social welfare. According to Fligstein (2001), governance, property rights, control rules of exchange

develop out of the same societal processes that evolve political institutions. It is expected that greater

political legitimacy is associated with an enhanced ability of governments to influence banks and so

predilection for banks will be enhanced in such countries. Our hypotheses regarding these variables are:

H6: Political stability has a significant influence on financial architecture.

H6.1: Higher political stability will favor financing via markets compared to banks.

H7: Political legitimacy has a significant influence on financial architecture.

H7.1: Higher political legitimacy will not favor financing via markets compared to banks.

Kwok and Tadesse (2006) and Ergungor (2004) have investigated issues related to H6 and H6.1.

Both of these papers include measures of number of revolutions and/or number of assassinations as

measures of political stability. Kwok and Tadesse (2006) do not find these measures significant.

However, Ergungor (2004) does find support for H6 and H6.1, documenting evidence of an association of

more revolutions with a predilection for banking.

To our knowledge, H7 and H7.1 have not previously been directly investigated in the literature.

Our measure of political legitimacy from Gilley (2006)) incorporates measures of good governance,

democratic rights, and welfare gains. While Kwok and Tadesse (2006) include control of corruption as an

independent variable, and Ergungor (2004) measures of bureaucratic efficiency (control of corruption, red

tape, and efficiency of judiciary system), these variables do not incorporate relative estimates of

democratic rights and social welfare along with control of corruption into relative measures of political


19/44

these cultural qualities may increase social fractionalization and so lower public trust (Bjornskov

(2008)). A lowering of trust would increase the transaction costs of market involvement.Further, Perotti

and Von Thadden (2006) offer a Median-Voter Theorem wherein if wealth is concentrated in a rich

minority (wealth skewed) the median voter has relatively more at stake in the form of firm-specific

human capital and therefore supports dominance by banks. We measure national economic inequality

(GINI) with the Gini coefficient from the 2005 CIA World Fact Book. Our measure of ethnic

fractionalization (ETHNIC_FRACTION) comes from Alesina, Devleeschauwer, Easterly, Kurlat and

Wacziarg (2003). To our knowledge, these hypotheses have not been previously investigated. Our

hypotheses regarding these two variables are:

H8: Economic inequality has a significant influence on financial architecture.

H8.1: Higher economic inequality will not favor financing via markets compared to banks.

H9: Ethnic fractionalization has a significant influence on financial architecture.

H9.1: Higher ethnic fractionalization will not favor financing via markets compared to banks.

Hypotheses regarding wealth and region as partial determinants of financial architecture

In order to control for wealth affects on cultural measures and levels of governance efficiency as

well as levels of market development we include a measure of GDP per capita (GDP). This variable is

purchasing power parity (PPP) GDP per capita from World Development Indicators. Relative levels of

wealth is important to control for: it is reasonable to consider that measures of market development as

well as regulatory quality are associated levels of wealth, Further, Hofstede (2001) suggests an

association of wealth with a lower power distance. In this case, our hypotheses are:

H10: Wealth has a significant influence on financial architecture.

H10 1: Higher wealth will favor financing via markets compared to banks


20/44

We also include two dummy variables: a dummy variable (EURO) that receives 1 if the euro is

adopted as the national currency and 0 otherwise for particular country/years as well as a dummy

variable (FORMER_SOCIALIST) that receives a 1 if the nation is formerly a socialist country and a

0 otherwise. As suggested by Dermine and Hillion (1999), the development of capital markets is likely

influence by the adoption of a common currency. Licht, et al. (2005) suggest that the former communist

countries of Eastern Europe have stronger cultural hierarchy. Controlling for socialist origins also helps

assure that our results for power distance and other variables are not driven by the presence of the four

former socialist countries in our sample, Czech Republic, Hungary, Poland, and Russia. To our

knowledge the impact of these variables on financial architecture has not been studied before and our

hypotheses are:

H11: Euro introduction has a significant influence on financial architecture.

H11.1: Euro introduction will favor financing via markets compared to banks.

H12: Former Socialist status has a significant influence on financial architecture.

H12.1: Former Socialist status will not favor financing via markets compared to banks.

Table 2 presents summary statistics of these variables. Together these independent and control

variables reflect the factors we described earlier that may affect nations financing predilections: cultural

aspects, market concentration, governance, legal origin, wealth, economic inequality, and ethnic

fractionalization.4We see examining the coefficients of variation in Table 2, that of those variables that

are not dummy variables, that market concentration, and ethnic fractionalization vary the most. Both these

variables are related to concentration and hierarchy. Further, of the cultural dimensions, power distance is

more variable than uncertainty avoidance or individualism. Table 2 suggests that hierarchy is more

variable in European countries than variation in regulatory quality In examining Table 2 we also note that


21/44

partially determined by additional factors besides these. The last column of Table 2 summarizes our

hypotheses and indicates the expected relationship of each variable with our dependent variable.

(Please insert Table 2 about here)

V. Results

We first estimate models that focus on cultural variables and then add structural variables and

regulatory variables. Generally our models have variance inflation factors (VIF) of less than 10 for all

regressors. Exceptions are Table 3, Model 6, where POLITICAL_LEGITIMACY has a VIF of 10.65, and

Table 5, Model R4, where GDP has a VIF of 10.34. Overall, our generally low VIFs indicate that

multicollinearity is unlikely to be a problem. We also further reduce multicollinearity in additional tests

with a residual-based procedure which we describe below. We also do Hausman tests for each model to

determine whether to report random-effects or fixed-effects estimates. For all of our models, the Hausman

test is insignificant. Consequently we report random-effects estimates.

Determinants of Markets versus Institutions Relative Financing Roles

Table 3 shows the results of regressions for the determinants of the relative national size of

financial markets versus financial institutions (the dependent variable). Model 1 just uses the cultural

variables PDI, IDV and UAI along with our Herfindahl for concentration of market ownership

(MKT_CONC). We also control for GDP per capital and for the adoption of the euro (EURO). The

results of Model 1 are that PDI, MKT_CONC, and EURO are significantly positive at the 1% level, while

UAI is negatively significant at the 1% level. This suggests that societies with greater power distance, less

uncertainty avoidance and concentration of equity markets have a higher predilection for markets. GDP

and IDV are not significant. With regard to Model 1, which has only a small number of independent


22/44


Model 2 adds a dummy variable for English legal origin to the independent variables of Model 1.

As discussed above, prior research, namely Ergungor (2004) and Kwok and Tadesse (2006), finds legal

origin to be significant in partially determining whether nations are either market- or bank-based.

However, results for Model 2 report ENGLISH_LEGAL as not significant. PDI, MKT_CONC, and

EURO are again significantly positive at the 1% level, while UAI is again negatively significant at the 1%

level. GDP and IDVare again not significant.

Model 3 adds to the variables of Model 2 the regulatory and governance variables REG_QUAL,

POLITICAL_STABILITY and POLITICAL_LEGITIMACY. These variables are not significant. PDI

MKT_CONC, are again significantly positive at the 1% level, while UAI is again negatively significant at

the 1% level. EURO is positively significant at the 1% level. GDP, IDV and ENGLISH_LEGAL are

again not significant. The results of Model 3 suggest that cultural variables are more important than

governance variables in partially determining whether European nations are either market- or bank-based.

Model 4 adds to the independent variables of Model 3 a variable that assesses in a composite

manner the efficiency of debt enforcement, DEBT_EFFICIENCY. DEBT_EFFICIENCY is positively

significant, suggesting that greater ease of private credit is associated with a predilection for markets. PDI

and MKT_CONC are again significantly positive at the 1% level, while UAI is again negatively

significant at the 1% level. EURO is again positively significant at the 1% level. ENGLISH_LEGAL is

now significantly negative at the 5% level. REG_QUAL, POLITICAL_STABILITY, GDP are again not

significant. POLITICAL_LEGITIMACYis now negatively significant at 10%.

Model 5 adds to the independent variables of Model 4 variables for economic inequality (GINI)

and ethnic fractionalization (ETHNIC FRACTION). GINI is significant at 1%, suggesting that our


23/44

at 1% level. DEBT_EFFICIENCYis now positively significant at 1%. IDV is now significantly positive

at 10%. POLITICAL_LEGITIMACY is now not significant. GINI is positively significant at 5%.

Models 6 adds to the independent variables of Model 5 a dummy variable that receives 1 if the

country is formerly socialist (FORMER_SOCIALIST). FORMER_SOCIALIST is not significant,

suggesting that the significance of many of the other variables is not driven by the Eastern European

countries in our sample.PDI is still positively significant, now at 5%. MKT_CONC is again significantly

positive at the 1% level, while UAI is again negatively significant at the 1% level. EURO is again

positively significant at the 5% level. REG_QUAL, POLITICAL_STABILITY,

POLITICAL_LEGITIMACY, ETHNIC_FRACTION, and GDP are not significant. GINI is now not

significant. ENGLISH_LEGAL is again significantly negative now at the 5% level.

DEBT_EFFICIENCY is again significantly positive at the 5% level. IDV is not significant.

In examining these results based on the relative size of bank versus market financing presented in

Table 3, PDI, EURO, and MKT_CONC are positively significant at 1% or 5% in all models. Further,

UAI is negatively significant at the 1% level in all models. Measures of wealth, ethnic fractionalization,

regulatory quality and political stability are never significant. English legal origin is negatively significant

in the more comprehensive models. Efficiency of debt enforcement is positively significant in all models

in which it is present. Individualism and political legitimacy are generally not significant.

Regarding the results of Table 3 with respect to the hypotheses presented above, we find support

for H1, that culture has a significant influence on financial architecture. We also find support for H1.1,

that higher power distance is associated with markets, and H1.2, that higher uncertainty avoidance is

associated with banks. But we find no support for H1.3 that higher individualism favor financing via

markets We also find support for H2/H2 1: market concentration has a significant influence on financial


24/44

insignificant. We do however find support for H5 and H5.1 as efficiency of debt enforcement is

associated with financing via markets compared to banks. We find no support for H6/H6.1 or H7/H7.1:

political stability and political legitimacy are not significant in Table 3. We find some support for H8, that

economic inequality has a significant influence on financial architecture; however no support for H8.1,

that higher economic inequality will not favor financing via markets compared to banks. In fact we find

some support for an association of economic inequality and markets. We find no support for H9 and H9.1

or H10 and H10.1 as ethnic fractionalization and wealth are not significant. We find support for

H11/H11.1 as euro introduction has a significant association with markets. But we find no support for

H12/H12.1 as former-Socialist status is not significant.

Additional robustness tests

Orthogonal variables

As noted above, as governance factors, we have included in our models in Table 3 independent

variables measuring governance. However, we report in Table 4 results designed to test whether the

significance we have reported for these variables in Table 3 is independent of effects of cross-national

variation in wealth. In order to do this we employ a two-stage model. We first regress the respective

independent variable on our measure of GDP per capita. Specifically we regress each measure on GDP

and then use the residuals from these regressions as our independent variables.

1governance variable

it i it it GDP e = + + (1)

We include the residuals from Equation 1 as substitute independent variables for the respective

governance variables. We refer to these variables as orthogonal regulatory quality (ORTH_REG_QUAL),

orthogonal political stability (ORTH_POL_STABILITY), orthogonal political legitimacy


25/44

variable (including this variable in the models of Table 3 would have resulted in excessive

multicollinearity). It is expected that control of corruption will be associated with a greater predilection

for markets as trust will be enhanced. In Table 4 we include the measure of control of corruption from

Kaufmann, et al. (2008), made orthogonal to GDP (ORTH_CNTRL_CORRUPTION).

Table 4 Model R1, includes the independent variables PDI, UAI, IDV, MKT_CONC, GDP,

EURO, ENGLISH_LEGAL, ORTH_REG_QUAL, ORTH_POLITICAL_STABILITY,

ORTH_POLITICAL_LEGITIMACY, and ORTH_CNTRL_CORRUPTION. Similar to the results of

Table 3, PDI and MKT_CONC are significantly positive at the 1% level. UAI is negatively significant at

the 1% level. EURO is positively significant at 5%. Additionally ORTH_CNTRL_CORRUPTION is

positively significant at the 1% level. This suggests that control of corruption is more important than other

regulatory measures in partially determining a preference for markets. Further, in this model, political

legitimacy is now negatively significant at the 5% level. This suggests that, when controlling for other

factors, more democratic legitimacy is associated with a preference for relationship financing over market

financing for European markets.

Table 4 Model R2 adds to the independent variables of Model R1, ORTH_DEBT_EFFICIENCY.

This variable is positively significant at the 10% level. PDI and MKT_CONC are again positively

significant at 1%. UAI is again negatively significant at 1%. Euro is now positively significant at the 1%

level. English legal origin is negatively significant now at the 5% level.

ORTH_POLITICAL_LEGITIMACY is again negatively significant. ORTH_CNTRL_CORRUPTION is

again positively significant at 1%.

Table 4 Model R3 adds to the independent variables of Model R2 the independent variables GINI

and ETHNIC FRACTION ETHNIC FRACTION is not significant GINI is significantly positive at the


26/44

ORTH_POLITICAL_LEGITIMACY is again negatively significant. ORTH_DEBT_EFFICIENCY and

ORTH_CNTRL_CORRUPTION are again positively significant.

Table 4 Model R4 adds to the independent variables of Model R3 the dummy variable

FORMER_SOCIALIST. This variable is not significant. GINI is now not significant.

ORTH_CNTRL_CORRUPTION is again positively significant, but now at 10%. PDI and MKT_CONC

are again positively significant at 1%. UAI is again negatively significant at 1%. Euro is positively

significant at the 1% level. English legal origin is negatively significant at the 1% level.

ORTH_POLITICAL_LEGITIMACY is again negatively significant. ORTH_DEBT_EFFICIENCY is

again positively significant.

Regarding the results of Table 4 with respect to the hypotheses presented above, we again find

support for H1/H1.1/ H1.2: higher power distance is associated with markets, while uncertainty avoidance

is associated with banks. We again, with the exception of one model, find no support for H1.3 that higher

individualism favor financing via markets. We again find support for H2/H2.1: market favors financing

via markets compared to banks. We also find support for H3, that legal origin has a significant influence

on financial architecture; however not for H3.1, that English legal origin will favor financing via markets

compared to banks. In fact, we again find support for a negative association of common law and markets.

We find no support for H4/H4.1: regulatory quality is insignificant. We do however again find support for

H5/H5.1 as efficiency of debt enforcement is associated with financing via markets compared to banks.

We again find no support for H6/H6.1: political stability is not significant. However we do find support

for or H7 as political legitimacy is significant in Table 4 but not H7.1 as it is negatively significant. We

find some support for H8, that economic inequality has a significant influence on financial architecture;

however no support for H8 1 that higher economic inequality will not favor financing via markets


27/44

Logit modeling

As a further test of robustness we conduct logit regressions on the same sets of independent

variables as the models of Table 3. Table 5 shows the results of logit regressions for the determinants of

the relative national size of financial markets versus financial institutions (the dependent variable). Model

L1 just uses the cultural variables PDI, IDV and UAI along with our Herfindahl index for concentration

of market ownership (MKT_CONC). We also control for GDP per capital and for the adoption of the euro

(EURO). The result of Model L1 is that PDI and MKT_CONC are significantly positive at the 1% level,

while UAI is negatively significant at the 1% level. This suggests that societies with greater power

distance, less uncertainty avoidance and concentration of equity markets have a higher predilection for

financial markets over banks. GDP, EURO and IDV are not significant. With regard to Model L1, which

has only a small number of independent variables, the significance, at the 1% level of two measures from

Hofstede (2001), as well as the significant of our measure for market concentration reaffirms our

inference from the results of Table 3, Model 1 that cultural and political economy variables alone are of

note in explaining financing predilection in European markets.


Model L2 adds a dummy variable for English legal origin to the independent variables of Model

L1. PDI and MKT_CONC are again significantly positive at the 1% level, while UAI is again negatively

significant at the 1% level. GDP, EURO and IDV are again not significant. As with Table 3, Model 2,

Model L2 reports ENGLISH_LEGAL as not significant.

Model L3 adds to the variables of Model L2 the regulatory and governance variables

REG_QUAL, POLITICAL_STABILITY and POLITICAL_LEGITIMACY. PDI MKT_CONC, are again

i ifi l i i h 1% l l hil UAI i i i l i ifi h 1% l l GDP


28/44


29/44

Overall, in examining the results based on the logit regressions presented in Table 5, PDI,

DEBT_EFFICIENCY, and MKT_CONC are positively significant at 1% or 5% in all models. Further,

UAI is negatively significant at the 1% level in all models, while English legal origin is again negatively

significant in the more comprehensive models. This confirms our results of Table 3 for these independent

variables. However, unlike the results of Table 3, EURO is generally not significant in Table 5. Further,

our independent variables for Individualism, regulatory quality, and ethnic fractionalization, which are

generally not significant in Table 3, are all positively significant in Table 5. Additionally, while not

generally significant in Table 3, political legitimacy is negatively significant in all models of Table 5. It

seems that these variables being significant in Table 5 but not in Table 3 is likely due to logit modeling

exaggerating variation in our dependent variable by imposing a 1 or 0 classification instead of

smaller continuous variations in the dependent variable.

Regarding the results of Table 5 with respect to the hypotheses presented above, we again find

support for H1/H1.1/H1.2: higher power distance is associated with markets, while uncertainty avoidance

is associated with banks. We also find support for H1.3 that higher individualism favor financing via

markets. We again find support for H2/H2.1: market favors financing via markets compared to banks. We

also find support for H3, that legal origin has a significant influence on financial architecture; however

not for H3.1, that English legal origin will favor financing via markets compared to banks. In fact, we

again find support for a negative association of common law and markets. Unlike Tables 3 and 4, Table 5

documents support for H4/H4.1: regulatory quality is positively significant. We also again find support

for H5/H5.1 as efficiency of debt enforcement is associated with financing via markets compared to

banks. We again find no support for H6/H6.1: political stability is not significant. However we do find

support for or H7 as political legitimacy is significant in Table 4 but not H7 1 as it is negatively


30/44

in Tables 3 and 4, we again find no support for H10/H10.1 or H12/H12.1 as wealth and former-Socialist

status are not significant. Unlike Tables 3 and 4 we find no support for H11/H11.1 as the introduction of

the Euro is not significant.

Discussion of results

Examination of results

The results presented in Table 3 indicate reliable model estimates and no significant missing

variables based on the fact that the intercepts are never significant for any of the models and on the

summary statistics presented for the model estimates. PDI, MKT_CONC, and EURO are all significantly

positive at a better than 1% significance level in all models. Similarly, UAI is negatively significant at

better than the 1% significance level in all models. These are very strong, consistent, and reliable results.

ENGLISH_LEGAL is negatively significant and DEBT_EFFICIENCY is positively significant in some

models and these results are, thus, less reliable. These Table 3 results and conclusions are confirmed

strongly in Tables 4 and 5. In addition, the robustness tests in Tables 4 and 5 also indicate that in Europe,

political legitimacy may be associated with a predilection for banks, while control of corruption, ethnic

fractionalization, and individualism may be associated with a preference for markets.

In summary, based on the results presented in Tables 3, 4, and 5, we document that a greater

predilection for market financing is consistently associated with higher levels of power distance,

concentration in equity markets, and the introduction of the euro.5These results make sense as higher

power distance reflects enhanced faith in the rules of exchange favoring the development of financial

markets. Because of more efficient dissemination of information and more efficient political control,

nations with higher concentration in equity markets tend to favor market financing. Additionally, the

i d i f h f d fi i l k b f i d


31/44

equity financing over bank financing is strongly associated with higher uncertainty avoidance. This

finding supports, for European countries, the Kwok and Tadesse (2006) findings that greater predilection

for market financing is associated with less uncertainty avoidance. This is an intuitively expected result as

uncertainty avoidance would lessen the confidence in the contracts of market exchange.

In addition, there is some (weaker and less consistent) evidence that control of corruption, ethnic

fractionalization, efficiency of debt enforcement and individualism may be positively, and an English

legal origin and political legitimacy negatively, related to a predilection for financial markets. Control of

corruption can favor financial markets over banks, as noted earlier, because corruption is closely linked to

government power and control of banks and their regulation. Similarly, efficient debt enforcement reflects

reliability of contracts between un-related parties and so it favors financial markets. English legal origin

can strengthen the rights of minority owners reducing the benefits of control by the larger shareholders

and, thus, the advantages of financial markets. However, these results should be viewed with caution as

only two countries (the UK and Ireland) in our sample have an English legal origin.6Lastly, political

legitimacy is negatively related to a predilection for financial markets as more legitimate governments can

exert stronger influences on the banking system and so favor banks for providing financial intermediation.

The few significant results we obtain for economic inequality and ethnic fractionalization are inconsistent

with our hypotheses; but these variables are not always consistently significant and likely distorted by the

high negative correlations of Gini with regulatory quality, political stability, and control of corruption.

Summary of findings in regard to hypotheses and previous literature.

We document robust support for several hypotheses which have not been previously investigated.

Our robust findings that a predilection for markets is associated with higher levels of power distance,


32/44

concentration in equity markets, and the introduction of the euro, affirm our hypotheses H1.1, H2.1, and

H11/H11.1, which have not previously been investigated. In addition, we document some less consistent

evidence that ethnic fractionalization, efficiency of debt enforcement and individualism may be positively

associated with a preference for markets, supporting hypotheses H5/H5.1 (debt efficiency) and H1.3

(individualism), while affirming H9 (that ethnic fractionalization is significant) while refuting hypothesis

H9.1 (that greater ethnic fractionalization is associated with a predilection for banking). These hypotheses

have not been previously investigated.

We also document some association of political legitimacy with a preference for banking,

supporting our hypotheses H7/H7.1. These hypotheses have not been previously investigated to the extent

of this paper. While Kwok and Tadesse (2006) and Ergungor (2004)control for measures against

corruption (as we also do in our robustness tests), and Kwok and Tadesse (2006)) include as an

independent variable an Institutions Index which incorporates amongst the six measures of Kaufmann,

et al. (1999) a measure of Voice and Accountability, this paper also investigates the relationship

between financial architecture and democratic rights and social welfare. While we find for European

countries that political stability is not a significant determinant of architecture (rejecting H6/H6.1);

whereas Ergungor (2004) finds political stability (measured as number of revolutions) as a significant

determinant, we suggest our results are likely due to examining the association of political legitimacy as a

political factor beyond simply political stability.

We also document evidence that both support and contradict hypotheses that have been

previously investigated. We document strong support for hypotheses H1.2, that greater uncertainty

avoidance is associated with a preference for banking. This result affirms, for European countries, the

result of Kwok and Tadesse (2006) We also find that for European countries English legal origin is


33/44

choices between the two major channels of financial intermediation, banking and markets, are influenced

significantly by measures of culture and political structure even after controlling for other relevant

economic variables (many of our results have not been presented in prior literature).

International business (IB) scholarship focuses on the issues in doing business across national

boundaries and, thus, in diverse institutional and cultural settings (Aggarwal, 2008). Thus, the business

impact of cross-border differences in cultural and social environments is a central focus of IB research

and scholarship. Indeed, it has been noted that geography and location continue to be important foci for

IB scholarship (Buckley, 2002). Consequently, it is important that IB scholars develop a systematic

understanding of how international differences in social, cultural, and political environments impact

multinational firms and their choices regarding an important resource, finance.

This paper contributes to the international business literature by providing an empirically

supported, extended new model of how the cultural, social, and economic environments interact in

influencing national financial architecture and the availability of financial resources to an MNC. Thus, it

extends the role of cultural and social factors in understanding business and finance in cross-national

settings. For example, finance is an important resource for firms and feedback and signals from providers

of finance are an important factor in the management of firms. Multinational firms must generate

financing and parse and decipher managerially useful information from multiple countries. It is important

for such cross-national companies to understand the nature of the financial environment to optimize and

manage global resources available to them. Managers of multinational firms seeking to optimize local

financing strategy will also benefit from understanding that national financial strategies for a

multinational optimally may not all be the same and may need to differ from a uniform global financial

strategy These differing national financial strategies may have to reflect varying financial architectures


34/44


35/44

References

Aggarwal, Raj, (2008), "International Business and Finance Scholarship", in Jean J. Boddewyn,ed.:International Business Scholarship: AIB Fellows on the First 50 Years and Beyond:

Research in Global Strategic Management(London: Emerald Group Publishing).

Aggarwal, Raj, and Shelly Zhao, (2009), "The diversification discount puzzle: Empirical

evidence for a transactions cost resolution", Financial Review44 (No. 1, February) pp.

113-135.

Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat, and Romain

Wacziarg, (2003), "Fractionalization",Journal of Economic Growth8 (No. 2, June) pp.

155-194.

Allen, Franklin, (2001), "Do financial institutions matter?"Journal of Finance56 (No. 4,

August) pp. 1165-1175.

Allen, Franklin, and Anthony M. Santomero, (2001), "What do financial intermediaries do?"

Journal of Banking & Finance25 (No. 2) pp. 271-294.

Beck, Thorsten, A. Demirguc-Kunt, and Ross Levine, (2006), "Bank supervision and corruption

in lending",Journal of Monetary Economics53 (No. 8, November) pp. 2131-2163.

Beck, Thorsten, Asli Demirguc-Kunt, and Ross Levine, (2000), "A new database on financialdevelopment and structure", World Bank Economic Review14 (No. 3, September) pp.

597-605.

Bjornskov, Christian, (2008), "Social trust and fractionalization: A possible reinterpretation",

European Sociological Review24 (No. 3, February) pp. 271-283.

Boot, Arnoud W. A., and Anjan V. Thakor, (1997), "Financial system architecture",Review of

Financial Studies10 (No. 3) pp. 693-733.

Buckley, Peter J., (2002), "Is the international business research agenda running out of steam?"Journal of International Business Studies33 (No. 2) pp. 365-373.

Coase, Ronald H., (1960), "The problem of social cost",Journal of Law and Economics3 (No.

1, October) pp. 1-44.

Dam, Kenneth W., (2006), The Law-Growth Nexus: The Rule of Law and Economic

Development(Brookings Institution Press, Washington, D.C.).

Delerue, Helene, and Eric Simon, (2009), "National cultural values and the perceived relationalrisks in biotechnology alliance relationships",International Business Review18 (No. 1,

February) pp. 14-25.

Demirguc-Kunt, Asli, and Ross Levine, (2001), "Financial Structure and Economic Growth", in

(Cambridge, Massachusetts: MIT Press).

Dermine, Jean, and Pierre Hillion, (1999), "European Capital Markets with a Single Currency",

in (Oxford: Oxford University Press)


36/44

Ergungor, Ozgur E., (2004), "Market- vs. bank-based financial systems: do rights and regulations

really matter?"Journal of Banking and Finance28 (No. 12, December) pp. 2869-2887.

Fligstein, Neil, (2001), The Architecture of Markets: An Economic Sociology of Twenty-First-

Century Capitalist Societies(Princeton University Press, Princeton, N.J.).

Fogel, Kathy, (2006), "Oligarchic family control, social economic outcomes, and the quality of

government",Journal of International Business Studies37 (No. 5) pp. 603-622.

Fukuyama, Francis, (1995), Trust: The Social Virtues and the Creation of Prosperity(Free Press,

New York).

Gilley, Bruce, (2006), "The determinants of state legitimacy: results for 72 countries",

International Political Science Review27 (No. 1) pp. 47-71.

Guiso, Luigi, Paola Sapienza, and Luigi Zingales, (2004), "The Role of Social Capital inFinancial Development",American Economic Review94 (No. 3, June) pp. 526-556.

Hart, Oliver, (1995), Firms, contracts, and financial structure(Oxford University Press, New

York).

Hart, Oliver, (2001), "Financial contracting",Journal of Economic Literature39 (No. 4,

December) pp. 1079-1100.

Hofstede, Geert, (2001), Culture's Consequences(Sage Publications, London).

Hong, Harrison, Jeffrey D. Kubik, and Jeremy Stein, (2004), "Social interaction and stock-market participation",Journal of Finance59 (No. 1, February) pp. 137-163.

Husted, Bryan W., (1999), "Wealth, culture, and corruption",Journal of International Business

Studies30 (No. 2) pp. 339-359.

Johnson, Simon, Rafael La Porta, Florencio Lopez-de-Silvanes, and Andrei Shleifer, (2002),"Tunneling", in Klaus J. Hopt, ed.: Capital Markets and Company Law(Oxford: Oxford

University Press).

Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi, 2008, Governance matters VII:Aggregate and individual governance indicators, 1996-2007.

Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton, 1999, Governance Matters, World

Bank Policy Research Department Working Paper No. 2196.

Krueger, Anne O., (1974), "The political economy of rent-seeking society",American Economic

Review64 (No. 3, June) pp. 291-303.

Kwok, Chuck CY, and Solomon Tadesse, (2006), "National culture and financial systems",Journal of International Business Studies37 (No. 2, March) pp. 227-247.

La Porta, Rafael, Florencio Lopez-de-Silvanes, and Andrei Shleifer, (2002), "Government

ownership of banks",Journal of Finance57 (No. 1, February) pp. 265-301.

La Porta, Rafael, Florencio Lopez-de-Silvanes, and Andrei Shleifer, (2006), "What works in

securities laws?"Journal of Finance61 (No. 1, February) pp. 1-32.

Leland Hayne E and David H Pyle (1977) "Informational asymmetries financial structure


37/44

Licht, Amir N., Chanan Goldschmidt, and Shalom H. Schwartz, (2007), "Culture rules: The

foundations of the rule of law other norms of governance",Journal of Comparative

Economics35 (No. 4) pp. 659-688.

Luhmann, Niklas, (1982), Trust and Power(John Wiley & Sons, New York).

Magnusson, Peter, Daniel W. Baack, Srdan Zdravkovic, Karin M. Staub, and Lyn S. Amine,

(2008), "Meta-analysis of cultural differences: Another slice at the apple",International

Business Review17 (No. 5, October) pp. 520-532.

Milgrom, Paul, and John Roberts, (1990), "Bargaining costs, influence costs, and organization of

economic activity", in James E. Alt, and Kenneth A. Shepsle, (eds), Perspectives on

Positive Political Economy(Cambridge: Cambridge University Press).

Modigliani, Franco, and Enrico Perotti, 1998, Security versus bank finance: the importance of aproper enforcement of legal rules, Tinbergen Institute Discussion Paper 04-012/2.

Morck, Randall K., David A. Strangeland, and Bernard Yeung, (2000), "Inherited Wealth,

Corporate Control, and Economic Growth: The Canadian Disease?" in Randall Morck,

ed.: Concentrated Corporate Ownership(Chicago: University of Chicago Press).

Myers, Stewart, (1977), "The determinants of corporate borrowing",Journal of Financial

Economics5 (No. 2) pp. 147-175.

Ndikumana, Leonce, (2005), "Financial development, financial structure, and domesticinvestment: International evidence",Journal of International Money and Finance24 (No.

4, June) pp. 651-673.

Nes, Erik B., Carl Arthur Solberg, and Ragnild Silkoset, (2007), "The impacts of national cultureand communication on exporter-distributor relations and on export performance",

International Business Review16 (No. 4, August) pp. 405-424.

North, Douglas C., (1990),Institutions, Institutional Change, and Economic Performance

(Cambridge University Press, Cambridge, UK).

Olson, Mancur, (1982), The Rise and Decline of Nations: Economic Growth, Stagflation, and

Social Rigidities(Yale University Press, New Haven, CT).

Pagano, Marco, and Paolo Volpin, (2005), "The political economy of corporate governance",

American Economic Review95 (No. 4, September) pp. 1005-1030.

Perotti, Enrico C., and Ernst Ludwig Von Thadden, (2006), "The political economy of dominantinvestors",Journal of Institutional and Theoretical Economics162 (No. 1, March) pp.

222-224.

Rajan, Raghuram, G., (1992), "Insiders and outsiders: The choice between informed and arm'slength debt",Journal of Finance47 (No. 4, September) pp. 1367-1400.

Rajan, Raghuram, G., and Luigi Zingales, (1998), "Which capitalism? Lessons from the East

Asian crisis",Journal of Applied Corporate Finance11 (No. 3, Fall) pp. 40-48.

Rajan, Raghuram, G., and Luigi Zingales, (2003a), "Great reversals: The politics of financial


38/44

Roe, Mark J., (2003), Political Determinants of Corporate Governance(Oxford University

Press, New York).

Roland, Gerard, (2004), "Understanding institutional change: Fast-moving and slow-moving

institutions", Studies in Comparative International Development38 (No.) pp. 109-132.

Shirai, Sayuri, (2004), "Assessing the impact of financial and capital market reforms on firms'

corporate financing patterns in India", South Asia Economic Journal5 (No. 2) pp. 189-

208.

Tadesse, Solomon, (2002), "Financial architecture and economic performance: International

evidence",Journal of Financial Intermediation11 (No. 4, October) pp. 429-454.

White, William R., 1998, The coming transformation of continental European banking?, BIS

Working Paper.Williamson, Oliver E., (1988), "Corporate Finance and Corporate Governance",Journal of

Finance43 (No. 3, July) pp. 567-592.


39/44

- - 38

Table 1: Summary Statistics for Size of Stock Market Development versus Banking Development

For Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and UK for the period 19962003, this

table displays mean, median and standard deviations for 1) size of stock market activity as measured by domestic stock market capitalization to GDP (Beck, et al. (2000), et al.(2000)); size of banking activity as measured by domestic assets of deposit money banks to GDP (Beck, et al. (2000), et al. (2000)); and 3) a relative measure of stock market

develop to banking development based on size as measured by 1) divided by 2). Nations are rank ordered by relative market development to banking development.

Domestic Stock Market Capitalization to

GDP

Domestic Assets of Deposit Money

Banks to GDP

Domestic Stock Market Capitalization to

Domestic Assets of Deposit MoneyBanks

Country Mean Standard Deviation Mean Standard Deviation Mean Standard Deviation

Finland 1.26 0.68 0.64 0.07 2.03 1.25

Sweden 1.09 0.23 0.79 0.34 1.69 0.87

Russia 0.35 0.24 0.22 0.04 1.44 0.72Switzerland 2.34 0.49 1.74 0.04 1.34 0.28

UK 1.47 0.24 1.33 0.17 1.13 0.27

France 0.75 0.23 1.04 0.04 0.73 0.23

Denmark 0.56 0.11 1.12 0.62 0.72 0.62

Belgium 0.80 0.34 1.20 0.17 0.71 0.40

Netherlands 1.17 0.35 1.80 0.44 0.66 0.28

Spain 0.70 0.17 1.20 0.20 0.58 0.10

Norway 0.43 0.11 0.73 0.06 0.58 0.12

Ireland 0.60 0.12 1.10 0.28 0.57 0.15

Hungary 0.24 0.08 0.44 0.11 0.57 0.23

Italy 0.43 0.14 0.90 0.11 0.48 0.15

Poland 0.17 0.09 0.34 0.04 0.47 0.21

Czech Republic 0.23 0.05 0.56 0.10 0.41 0.10

Germany 0.47 0.13 1.41 0.05 0.33 0.09

Portugal 0.40 0.11 1.28 0.21 0.31 0.10

Austria 0.21 0.11 1.23 0.03 0.18 0.10

Mean Overall 0.72 1.00 0.79

T bl 2 S St ti ti


40/44

- - 39

Table 2: Summary Statistics

This table reports summary statistics for country variables used in Tables 3, 4 and 5 for Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Hungary, Ireland, Italy, Netherlands,Norway, Poland, Portugal, Spain, Russia, Sweden, Switzerland and UK for the period 19962006. PDI, IDV and UAI are Power Distance, Individualism and Uncertainty Avoidance of Hofstede

(2001). MKT_CONC is a Herfindahl Index of market concentration created for this paper. GDP is PPP GDP per capita from World Development Indicators. ENGLISH_LEGAL is a dummy that isassigned 1 if nation has English legal origin POLITICAL_STABILITY, REG_QUAL and CNTRL_CORRUPTION are from governance indicators of Kaufmann, et al. (2008); GINI is from 2005 CIA

World Fact Book, DEBT_EFFICIENCY is the efficiency of debt enforcement from Djankov, et al. (2008); POLITICAL_LEGITIMACY is the Index of Political Legitimacy from Gilley (2006) .FORMER_SOCIALIST is a dummy that receives 1 is country has a socialist legal origin. EURO is a dummy that receives 1 if Euro is currency for country/year.ETHNIC_FRACTION is the

Index of Ethnic Fractionalization from Alesina, et al. (2003). Overall values are reported. The last column indicates the expected relationship of each independent variable with our dependent variable.

Variable MeanStandard

DeviationCoefficient of Variation Minimum Maximum N

Expected

Sign*

Domestic Stock Market

Capitalization to Domestic Assetsof Deposit Money Banks

0.79 0.66 0.83 0.11 4.69 201

PDI 44.26 20.08 0.45 11.00 95.00 209 +

IDV 66.00 13.56 0.21 27.00 89.00 209 +

UAI 66.37 23.48 0.35 23.00 104.00 209

MKT_CONC 0.15 0.12 0.80 0.02 0.74 205 +

GDP 24,473.05 8490.84 0.35 6271.20 50077.99 209 +

EURO 0.38 0.49 1.29 0.00 1.00 209 +

ENGLISH_LEGAL 0.11 0.31 2.82 0.00 1.00 209 +

REG_QUAL 1.21 0.53 0.44 0.78 2.01 209 +

POLITICAL_STABILITY 0.89 0.53 0.60 1.04 1.65 209 +

CNTRL_CORRUPTION 1.47 0.85 0.58 0.98 2.58 209 +

GINI 31.07 4.95 0.16 23.20 40.50 209

DEBT_EFFICIENCY 67.01 24.31 0.36 0.07 92.40 209 +

POLITICAL_LEGITIMACY 6.15 1.19 0.19 2.27 7.62 209 +

FORMER_SOCIALIST 0.21 0.41 1.95 0.00 1.00 209

ETHNIC_FRACTION 0.19 0.15 0.79 0.05 0.56 209

*


41/44


42/44

Table 4: Cross-National Determinants of Market versus Institutional Financing for European Countries: Results of Robustness Tests

This table reports results of panel data regressions for Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Hungary, Ireland,Italy, Netherlands, Norway, Poland, Portugal, Spain, Russia, Sweden, Switzerland and UK for the period 19962006. PDI, UAI and IDV are

Power Distance, Uncertainty Avoidance, and Individualismism of Hofstede (2001). MKT_CONC is a Herfindahl Index of market concentrationcreated for this paper. GDP is PPP gdp per capita from World Development Indicators. ENGLISH_LEGAL is a dummy that is assigned 1 if nationhas English legal origin. ORTH_POLITICAL_STABILITY, ORTH_CNTRL_CORRUPTION, and ORTH_REG_QUAL are derived fromgovernance indicators of Kaufmann, et al. (2008); GINI is from 2005 CIA World Fact Book., ORTH_DEBT_EFFICIENCY is derived from the

efficiency of debt enforcement from Djankov, et al. (2008), ORTH_POLITICAL_LEGITIMACY is derived from the Index of PoliticalLegitimacy from Gilley (2006). FORMER_SOCIALIST is a dummy that receives 1 is country has a socialist legal origin. EURO is a dummythat receives 1 if Euro is currency for country/year.ETHNIC_FRACTION is the Index of Ethnic Fractionalization from Alesina, et al. (2003)).Variance inflation factors are less than 10 for all variables and models except Model R4 where GDP has a VIF of 10.34. Random-effects estimates

are reported. P values are in parentheses.

MODELDependent Variable: domestic stock market capitalization

relative to domestic assets of deposit money banksR1 R2 R3 R4

INTERCEPT0.38

(0.653)0.35

(0.653)1.72(0.149)

1.35(0.433)

PDI0.03***(0.001)

0.02***(0.002)

0.02***(0.002)

0.02***(0.006)

UAI0.02***

(0.000)

0.02***

(0.000)

0.03***

(0.000)

0.03***

(0.000)

IDV0.00

(0.956)

0.00

(0.501)

0.01*

(0.073)

0.01*

(0.111)

MKT_CONC3.14***

(0.000)

3.02***

(0.000)

3.15***

(0.000)

3.16***

(0.000)

GDP0.00

(0.284)

0.00

(0.565)

0.00

(0.309)

0.00

(0.761)

EURO0.25**(0.017)

0.28***(0.008)

0.28***(0.006)

0.27***(0.010)

ENGLISH_LEGAL0.38(0.194)

0.66**(0.034)

1.31***(0.001)

1.26***(0.005)

ORTH_REG_QUAL

0.17

(0.382)

0.20

(0.294)

0.24

(0.188)

0.24

(0.195)

ORTH_POLITICAL_STABILITY

0.05

(0.773)

0.06

(0.736)

0.04

(0.787)

0.04

(0.806)

ORTH_POLITICAL_LEGITIMACY

0.30**

(0.023)

0.39***

(0.003)

0.31**

(0.014)

0.33**

(0.037)

ORTH_CNTRL_CORRUPTION

0.51***(0.005)

0.48***(0.006)

0.41**(0.013)

0.37*(0.063)

ORTH_DEBT_EFFICIENCY

0.01*(0.063)

0.01***(0.005)

0.01**(0.012)

GINI0.05**

(0.032)

0.05

(0.137)

ETHNIC_FRACTION0.42

(0.362)

0.39

(0.438)

FORMER_SOCIALIST0.12

(0.747)

Hausman Test3.08

(0.688)

3.84

(0.699)

3.55

(0.737)

4.12

(0.533)


43/44


44/44

Documents

financial mark