the attractiveness of central eastern european countries ...€¦ · country rankings that vary...

The Attractiveness of Central Eastern European Countries for

Venture Capital and Private Equity Investors

Alexander Peter Groh*, Heinrich von Liechtenstein**, and Karsten Lieser***

Abstract

We address the attractiveness of Central Eastern European countries for Venture Capital and

Private Equity investors by the construction of a composite index. For the index’s

composition we refer to the results of numerous prior research papers that investigate

relevant parameters determining entrepreneurial activity and/or the engagements of

institutional investors. We aggregate the index via five different methods and receive

country rankings that vary only slightly, signaling a robust index calculation. We

benchmark the Central Eastern European countries with the EU 15 countries, Norway and

Switzerland and identify six tier groups of attractiveness for all of our sample countries. We

compare our index with the actual Venture Capital and Private Equity activities in the

individual countries and reveal a reasonable correlation of both figures. The results highlight

the strengths and weaknesses of the particular economies and provide guidelines for policy

improvements to attract Venture Capital and Private Equity and hence, to spur innovation,

entrepreneurship, employment, and growth.

JEL codes: G23, G24, M13, O16, P34, P52

Keywords: Central Eastern Europe, Economic Transition, Venture Capital, Private Equity

*corresponding author, GSCM Montpellier Business School, 2300 Avenue des Moulins,

34185 Montpellier Cedex 4, France, [email protected] and IESE Business

School – University of Navarra, Finance Department, Av. Pearson, 21, 08034 Barcelona,

Spain

**IESE Business School – University of Navarra, Finance Department, Av. Pearson, 21,

08034 Barcelona, Spain, [email protected]

*** IESE Business School – University of Navarra, Finance Department, Av. Pearson, 21,

08034 Barcelona, Spain, [email protected]

mailto:[email protected]:[email protected]:[email protected]

The Attractiveness of Central Eastern European Countries for

Venture Capital and Private Equity Investors

Abstract

We address the attractiveness of Central Eastern European countries for Venture Capital and

Private Equity investors by the construction of a composite index. For the index’s

composition we refer to the results of numerous prior research papers that investigate

relevant parameters determining entrepreneurial activity and/or the engagements of

institutional investors. We aggregate the index via five different methods and receive

country rankings that vary only slightly, signaling a robust index calculation. We

benchmark the Central Eastern European countries with the EU 15 countries, Norway and

Switzerland and identify six tier groups of attractiveness for all of our sample countries. We

compare our index with the actual Venture Capital and Private Equity activities in the

individual countries and reveal a reasonable correlation of both figures. The results highlight

the strengths and weaknesses of the particular economies and provide guidelines for policy

improvements to attract Venture Capital and Private Equity and hence, to spur innovation,

entrepreneurship, employment, and growth.

JEL codes: G23, G24, M13, O16, P34, P52

Keywords: Central Eastern Europe, Economic Transition, Venture Capital, Private Equity

1

1. Introduction

The Central Eastern European (CEE)1 countries are still in a transitional stage. EBRD

(2005) emphasizes that improvements in governance, enterprise restructuring, and the

financial sector have been the main features of the transition process in the last years. The

CEE countries lessened the burden of business regulation, such as licensing and tax

administration, and they progressed in reducing corruption and organized crime. EBRD

(2006) highlights that the speed of the transition process varies in each country; some of

them show strong attempts to reform while others have decreased the pace of reform, partly

influenced by lately elected new governments.

Kolodko (2000) and Wagner and Hlouskova (2005) argue that the CEE countries are in a

period of catch-up that might last for several decades. This view is typically based on the

observation that per-capita GDP are still below the level of the current EU member states,

while education in CEE countries is at a high level, and institutional structures have been

converging for some time, as Süppel (2003) highlights. Growth estimates above the

European average, and the policy will to promote innovative enterprises should lead to a

strong demand for risk capital in the CEE countries and hence, to a high attractiveness for

Venture Capital and Private Equity (VC/PE) investors.

Hellmann and Puri (2000), and Kortum and Lerner (2000) show that VC/PE-backed

companies are more efficient innovators, and Belke et al. (2003), and Fehn and Fuchs

(2003) prove that they create more employment and growth than their peers. Levine (1997)

documents well the role of VC/PE funds in fostering innovative firms, and indeed, there

now exists a broad consensus that a strong VC/PE culture is a cornerstone for

commercialization and innovation in modern economies. Hence, policymakers should focus

on the creation of an adequate setting for a prospering VC/PE market to support

entrepreneurial activities and growth, especially in transition countries. However, the risk

capital supply is rather small compared to other European economies and relative to the

1 We define CEE countries as those Central Eastern European countries that lately (i.e. 2004, and 2007 respectively) accessed the European Union, namely Bulgaria, Czech Republic, Hungary, Poland, Romania, Slovakia, Slovenia, and the Baltic States, including Estonia, Latvia, and Lithuania.

2

expected growth opportunities in the CEE countries, even if institutional investors are

increasingly looking internationally for new investment opportunities. The first funds were

raised shortly after the fall of communism. According to EVCA (2004, 2005 and 2006),

since then only a little more than € 9bn have been raised by VC/PE funds dedicated to CEE

countries. This raises questions about the reasons that constrain the development of the

VC/PE market in that region.

In this paper, we address the attractiveness of the CEE countries for VC/PE investors by

means of the construction of a composite index. For the index’ composition we refer to the

results of numerous prior research papers that investigate relevant parameters determining

entrepreneurial activity and/or the engagements of institutional investors. We aggregate the

information to the index via five different methods. As benchmarks, we also calculate the

index scores for the 15 countries that belonged to the European Union before May 1st 2004

(the EU-15 countries), and for the non-EU countries Switzerland and Norway. As a result,

we obtain a ranking of all the individual economies based on their attractiveness for VC/PE

allocation by institutional investors. The rankings vary only slightly with the five different

index calculation methods, signaling a robust index definition. We clearly identify six tier

groups of attractiveness for all of our sample countries and three tier groups for the CEE

countries, that all rank below the EU-15 average. We compare our index with the

fundraising activities in the individual countries and reveal a reasonable correlation of both

figures.

Policy makers will benefit from our results by realizing the weaknesses of their countries to

attract international VC and PE. Improvements of the revealed weaknesses shall lead to

more supply of risk capital and will hence spur innovation, entrepreneurship, employment,

and growth.

The paper is structured as follows: After a brief introduction to our assumptions about

supply and demand in the VC/PE market, we review the most important related literature,

and discuss the relevant parameters for our model. Next, we explain the data and the

technical background for our index calculations. We verify the appropriateness of the sub-

indicators included, and discuss different normalization techniques, weighting, and

aggregation methods. Then, we present the index’s results, the strengths, and weaknesses of

the individual countries, perform robustness checks and determine tier groups regarding the

3

attractiveness of the various economies for VC/PE investors. Finally, we summarize this

paper.

2. Supply and Demand in the VC/PE Market

Models and conditions for achieving equilibrium in supply and demand in a VC/PE market

are comprehensively discussed in Gompers and Lerner (1998), Balboa and Martí (2003),

and Jeng and Wells (2000). We assume that the institutional investors supplying VC/PE

analyze several economies and choose among them for their international asset allocation.

The usual fundraising process implies, and statistics on supply and demand for VC/PE, such

as EVCA (2006) confirm this, that usually there is no lack of supply of funds. On the

contrary, the amount of funds raised in a particular year is generally higher than the amount

invested in the same period, and the funds raised are invested progressively in subsequent

years. This leads us to conclude that the suppliers of capital estimate the demand for VC/PE

with a one-to two-year horizon and make their allocations accordingly. Consequently, they

judge the individual countries’ attractiveness, which is determined primarily by expectations

about the ability of local VC/PE funds to perform a sufficient number of transactions with

satisfactory risk and return ratios. Hence, the predominant issue regarding the attractiveness

of a particular region for an institutional investor is probably the availability of adequate

investment opportunities. These opportunities are probably depending on local

entrepreneurial activities and are associated, among other factors, with innovations,

restructurings, the size of the economy, growth expectations, and the entrepreneurial spirit

of people. However, it is not clear to what extent these and related factors influence the

attractiveness of individual economies for investors in VC/PE funds. Therefore, in the

following section, we provide an overview of the literature dealing with success factors for

entrepreneurial activities, and the volume of VC/PE investments.

3. Literature Review

Intuitively, the state of a particular country’s economy affects VC/PE activities. Gompers

and Lerner (1998) point out that there are more attractive opportunities for entrepreneurs if

the economy is growing quickly. Wilken (1979) argues that economic development

facilitates entrepreneurship as it provides a greater accumulation of capital for investments.

The ease of start-ups is expected to be related to societal wealth, not only due to the

4

availability of start-up financing, but also to higher income among potential customers in

the domestic market. Romain and van Pottelsberghe de la Potterie (2004) find that VC/PE

activity is cyclical and significantly related to GDP growth.

Likewise, Jeng and Wells (2000) stress that the main force behind the cyclical swings is the

IPO activity because it reflects the potential return to the VC/PE funds. Kaplan and Schoar

(2005) confirm this. Black and Gilson (1998), and Gompers and Lerner (2000) point out

that risk capital flourishes in countries with deep and liquid stock markets. Schertler (2003)

uses either the capitalization of stock markets or the number of listed firms as a measure for

the liquidity of stock markets. He finds that the liquidity of stock markets has a significant

positive impact on VC investments at early stages.

The availability of debt financing is another entry key for start-ups, and, emphasized by

Greene (1998), in many countries it is the most important entrepreneurial obstacle.

Entrepreneurs need to find backers who are willing to bear this risk, like banks or VC/PE

funds. Hellmann et al. (2004) argue that banks represent the dominant financial institutions

in most of the countries. They examine the role of banks for the VC/PE industry and stress

that banks invest in VC/PE mainly for strategic reasons. They try to build early relationships

for future lending activities. Cetorelli and Gambera (2001) provide evidence that bank

concentration promotes the growth of those industrial sectors that have a higher need of

external finance by facilitating credit access to younger firms.

Additionally, the VC/PE activity in a particular country relates to the status of the VC/PE

market’s maturity level. Sapienza et al. (1996) mention that acceptance in a country’s

society and the historical evolvement of its VC/PE market determine investor confidence.

Balboa and Martí (2003) find that annual fundraising volume is dependent on the previous

year’s market liquidity. Chemla (2005) argues that the management of VC/PE funds is

costly. Particular regions become attractive to investors when the transaction volumes and

expected payoffs exceed a certain amount to cover the management fees.

Legal structures and the protection of property rights also appear to influence the

attractiveness of a VC/PE market. La Porta et al. (1997 and 1998) confirm that the legal

environment strongly determines the size and extent of a country’s capital market and local

firms’ ability to receive outside financing. They emphasize the difference between law on

books and the quality of law enforcement in some countries. Glaeser et al. (2001), Djankov

5

et al. (2003 and 2005) suggest that parties in common-law countries have greater ease in

enforcing their rights from commercial contracts. However, Cumming et al. (2006a) find

that the quality of a country’s legal system is stronger connected to facilitating VC/PE

backed exits than the size of a country’s stock market. Cumming et al (2006b) extend this

finding and show that cross-country differences in legality, including legal origin and

accounting standards, have a significant impact on the governance of investments in the

VC/PE industry. Desai et al. (2006) discuss that fairness and property rights protection

largely determine the growth and emergence of new enterprises. Cumming and Johan

(2007) highlight that the perceived importance of regulatory harmonization increases

institutional investors’ allocations to the asset class. La Porta et al. (2002) find lower cost of

capital for companies in countries with better investor protection. Lerner and Schoar (2005)

confirm these findings. Johnson et al. (1999) show that weak property rights limit the

reinvestment of profits in start-up firms. Even so, Knack and Keefer (1995), Mauro (1995),

and Svensson (1998) demonstrate that property rights significantly affect investments and

economic growth.

Gompers and Lerner (1998) stress that the capital gains tax rate influences VC/PE activity.

In fact, they confirm Poterba’s finding (1989), who builds a decision-model to become

entrepreneur. Bruce (2000 and 2002), and Cullen and Gordon (2002) prove that taxes matter

for businesses entry and exit. Bruce and Gurley (2005) explain that increases in the tax rates

on wages raise the probability of becoming an entrepreneur. Hence, the difference between

personal income tax rates and corporate tax rates tends to be an incentive to create self-

employment.

Rigid labor market policies negatively affect the evolvement of a VC/PE market. Lazear

(1990), and Blanchard (1997) discuss how protection of workers can reduce employment

and growth. Black and Gilson (1998) show that variations in labor market restrictions

correlate with VC/PE activity.

Djankov et al. (2002) investigate the role of administrative and bureaucratic burdens for

start-ups in different countries. They conclude that the highest barriers and costs are

associated with corruption, a larger unofficial economy, and bureaucratic delay. Baughn and

Neupert (2003) argue that bureaucracy in form of excessive rules and procedural

requirements, multiple institutions from which approvals are needed, and numerous

documentation requirements may severely constrain entrepreneurial activity. Lee and

6

Peterson (2000) stress that the time and money required to meet such administrative burdens

may discourage new venture creations.

Access to viable investments is probably the most important factor for the attractiveness of a

regional VC/PE market. In order to foster a growing risk capital industry, Megginson (2004)

argues that R&D culture, especially in universities or national laboratories, plays an

important role. Gompers and Lerner (1998) show that both industrial and academic R&D

expenditure is significantly correlated with VC/PE activity. Kortum and Lerner (2000)

highlight that the growth in VC/PE fundraising in the mid-90s may be due to a surge of

patents in the late 1980s and 1990s. Schertler (2003) emphasizes that the number of

employees in the field of R&D, and the number of patents, as an approximation of the

human capital endowment, has a positive and highly significant influence on VC/PE

activity. Furthermore, Romain and von Pottelsberghe de la Potterie (2004) find that the level

of entrepreneurship interacts with the R&D capital stock, with technological opportunities,

and the number of patents. Lee and Peterson (2000), and Baughn and Neupert (2003) argue

that national cultures shape both individual orientation and environmental conditions, which

lead to different levels of entrepreneurial activity in particular countries.

Summarizing this literature overview, we identify six key drivers determining the

attractiveness of an individual country for VC/PE investors: economic activity, size and

liquidity of capital markets, taxation, investor protection and corporate governance, human

and social environment, and entrepreneurial opportunities. We also refer to these key drivers

as level 1 indexes. To proxy the desired characteristics of these latent variables we find 42

different sub-indicators, which we can group into lower index levels (level 2 and 3). These

indicators are available for every individual sample country and ultimately determine its

attractiveness for VC/PE investors. We will describe the data and the framework for the

index aggregation in the following chapter.

4. Data and Aggregation Methodology

4.1 Data Sample

In general, composite indicators are used to summarize a number of underlying individual

indicators or variables. An indicator is a quantitative or qualitative measure derived from a

series of observed facts that can reveal or proxy characteristics. To ensure that our cross-

7

country aggregations are comparable, we deflate variables by the sizes of the

economies/countries and use either GDP or population as deflators.

We use several databases with yearly data ranging from 2000 to 2005 and usually refer to

the last data record. Some of the data-points are averages over a certain time-period to

smooth fluctuations. GDP figures, PE activity or M&A transaction volume among others

are such averages considering the period from 2000 to 2005. For large fluctuations and large

differences between the countries we also use logs of the averages (please refer to the

legend of table 1 for detailed information). In less than one percent of all cases, data was not

available for a certain year. If data-points are missing, we apply the three methods suggested

by Nardo et al. (2005a) in the following order: a) We try to find missing data in other

databases or via the Internet, b) we interpolate between the adjacent data records, and c) we

use the latest available data before 2005.

However, we do not always use raw data but sometimes refer to ready-made indexes like

the “doing business indexes” from the World Bank.2 For instance, our indicator for investor

protection and corporate governance is a ready-made index. The following table 1 presents

all the raw data and ready-made indexes and their sources (resp. alternative databases if

data-points are missing), we use to determine the “Venture Capital and Private Equity

Attractiveness Index” (VC/PEAI). For descriptions of the individual index items, we refer to

the sources, where comprehensive definitions and descriptions of the data series are

available.

Table 1: VC/PEAI – List of raw data and ready-made indexes and their sources

1 Economic activity 1.1 Gross Domestic Product 1.1.1 Total GDP [€/capita]* Global Market Inform. Database 1.1.2 Total GDP y-o-y growth [%]** Global Market Inform. Database 1.2 General Price Level [Index=1995]*** Global Market Inform. Database 1.3 Working force (unemployment rate) [%]* Global Market Inform. Database 1.4 Foreign direct investment, net inflows [% of GDP]*** Global Market Inform. Database 2 Capital market 2.1 IPO [IPO volume in % of GDP]**** Thomson Financial Data 2.2 Stock market 2.2.1 Stock market capitalization [% of GDP]* Worldbank Data

2 See http://www.doingbusiness.org.

8

2.2.2 Stock Market Total Value Traded / GDP [% of GDP]* Worldbank Data2.3 M&A market [sales % of GDP]* Global Market Inform. Database 2.4 Debt & Credit market 2.4.1 Central bank discount rate [%]* IMF2.4.2 Private Credit by Deposit Money Banks and Other

Financial Institutions [% of GDP]* Worldbank Data

2.4.3 Number of Banks [per Capita] EBRD, EUROSTAT Database2.5 Private equity activity [funds invested in % of GDP]**** Thomson Financial Data3 Taxation 3.1 Highest marginal tax rate, corporate rate (%) Worldbank Data 3.2 Difference between income and corporate tax rate [%] The Heritage Foundation 4 Investor protection and corporate governance 4.1 Extent of disclosure index Worldbank Data 4.2 Extent of director liability index Worldbank Data 4.3 Ease of shareholder suits index Worldbank Data5 Human & social environment 5.1 Education 5.1.1 Government expenditure on education, total [% of GDP]* Global Market Inform. Database 5.1.2 Amount employes as researcher in the university sector

[per capita]EUROSTAT

5.1.3 Amount university students [per capita]* Global Market Inform. Database 5.1.4 Amount university establishements [per capita] Global Market Inform. Database 5.2 Labor regulations 5.2.1 Rigidity of employment 5.2.1.1 Difficulty of hiring index Worldbank Data5.2.1.2 Rigidity of hours index Worldbank Data5.2.1.3 Difficulty of firing index Worldbank Data5.2.2 Hiring cost [% of salary] Worldbank Data5.2.3 Firing costs [weeks of wages] Worldbank Data5.3 Bribing & corruption index Transparency5.4 Crime 5.4.1 Juvenile offenders [per capita]* Global Market Inform. Database 5.4.2 Offences [per 100,000 habitants]* Global Market Inform. Database 6 Entrepreneurial opportunities 6.1 General Innovativeness Index TrendChart.Cordis6.2 R&D expenditure 6.2.1 Public R&D expenditures [% of GDP] EUROSTAT, OECD6.2.2 Business R&D expenditures [% of GDP] EUROSTAT, OECD6.3 Enterprise restructuring 6.3.1 Small-scale privatisation index EBRD 6.3.2 Large-scale privatisation index EBRD 6.3.3 Governance and enterprice restructuring index EBRD 6.4 Enterprise stock activity 6.4.1 Number of enterprises [per capita] World Bank, EUROSTAT, OECD 6.4.2 Enterprise foundation rate [%]* World Bank, EUROSTAT, OECD 6.5 Burden: Starting a Business 6.5.1 Procedures [numbers] Worldbank Data 6.5.2 Time [days] Worldbank Data 6.5.3 Cost of business start-up procedures [% GNI per capita] Worldbank Data 6.5.4 Min. capital [% of income per capita] Worldbank Data

* = arithmetic average of annual data from 2000 to 2005, ** = geometric average of annual data from 2000 to 2005, *** = log of arithmetic average of annual data from 2000 to 2005, **** = arithmetic average of annual data since coverage in the database for CEE countries, arithmetic

average of annual data from 2003 to 2005 for the other countries,

otherwise: 2005 data record.

http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.doingbusiness.org/ExploreTopics/HiringFiringWorkers/http://www.transparency.org/policy_research/surveys_indices/http://www.transparency.org/policy_research/surveys_indices/http://www.transparency.org/policy_research/surveys_indices/http://trendchart.cordis.lu/http://trendchart.cordis.lu/http://trendchart.cordis.lu/http://trendchart.cordis.lu/http://trendchart.cordis.lu/http://trendchart.cordis.lu/

We choose the 42 sub-indicators according to the referenced literature and according to our

comprehensive search of adequate data-series to proxy the latent variables. However, the

selection of the index-items is arguable. Additional data-series might be included some

might be discarded, or exchanged against others. We propose our selection and emphasize

that changes of particular data-series do not meaningfully affect the overall results. The

results are stable due to the magnitude of index-items and due to the index consistency, as

proofed by the reported and unreported robustness checks.

Due to the large number of index items (42), and data-points (105) per country (including

the data records over a certain period to calculate the averages), we follow the method

proposed by Nicoletti et al. (2000) and determine a pyramidal structure of three sub-index

levels for the index aggregation (see Figure A1 in the Appendix). We group the items that

we expect to correlate with each other. The main advantage of this pyramidal structure is

that we can trace back indicator values to increasing levels of detail. This will help in

interpreting the strengths and weaknesses of the individual countries and in drawing up the

conclusions.

Using this composition technique, we have to prove that the raw data and the ready-made

indexes are consistent for their aggregation. Thus, we perform a reliability analysis of all the

individual items, using Cronbach’s Alpha to ascertain the consistency of the chosen data.

This procedure is described in the subsequent section.

4.2 Analysis of Index Consistency

Cronbach’s Alpha3 is a measure of internal consistency of items in a model or survey.4 It

assesses how well a set of items measures a single one-dimensional object

(unidimensionality). Here, we use it to approve the consistency of our index and all the sub-

indexes we aggregate. Cronbach’s Alpha is defined as:

( )RnRn

11 −+=α (1).

3 Cf. Cronbach (1951). 4 Cf. Raykov (1998), Cortina (1993), Feldt et al. (1987), Green et al. (1977), Hattie (1985), and Miller (1995).

9

Thereby, n is the number of the components of a (sub-) index and R is the mean correlation

of the items (e.g. the mean of the non-diagonal terms of the correlation matrix). The

coefficient increases with the number of sub-indicators and with the correlation of each

tuple. Cronbach’s Alpha is equal to zero if no correlation exists and the sub-indicators are

independent. The coefficient is equal to one if sub-indicators are perfectly correlated.

Hence, a high alpha indicates that the underlying items proxy well the desired variable.

Nunnally (1978) suggests a value of 0.7 as acceptable threshold.

The following Table 2 presents the consistency of the six main key drivers measured by the

Cronbach’s Alphas of the underlying level 2 sub-indexes. We do not consider Taxation and

Investor protection in this calculation, because they consist of too few underlying items.

Table 2: Consistency analysis of the underlying items on the level of the six key drivers

Sub-indicator

Cronbach's Alpha

Economic activity 0.553Capital market 0.729Taxation -Investor protection -Human and Social environment 0.750Entrepreneurial opportunities 0.785

Cronbach’s Alpha for economic activity lies below Nunnally’s (1978) cut-off value of 0.7.

This could lead us to exchange or to drop some items that proxy economic activity.

However, the aggregation of the six key drivers to the overall VC/PE Attractiveness Index

yields a Cronbach’s Alpha of 0.784. Thus, we propose that our selection of items represents

well a country’s attractiveness for VC/PE investors and we do not discard sub-indexes from

our sample.

4.3 Normalization and Standardization

All data-points need to be normalized for their index aggregation. There exist various

techniques, each one with particular advantages and disadvantages as discussed by

Freudenberg (2003), Jacobs et al. (2004), and Nardo et al. (2005a). We use two different

methods - standardization and rescaling - in our calculations. Lastly, we analyze the

differences resulting from using both methods in a robustness check.

10

Standardization (or z-scores) converts the underlying data to a common scale of the normal

distribution with a mean of zero and a standard deviation of one. Hence, variables with

extreme values have a greater effect on the indicator. The z-score is defined as:

sxxz −= (2).

The rescaling method is used to normalize indicators to an identical range by linear

transformation. This method is vulnerable for extreme values or outliers that can distort the

transformation. However, rescaling can widen the range of indicators lying within small

intervals more than using the z-scores transformation. The rescaling method is defined as:

)min()max()min(

xxxxy

−−

= (3).

Ebert and Welsch (2004) discuss that the selection of a suitable normalization method is not

trivial and requires special attention. The method shall consider the properties of the

underlying data, as well as the objective of the summarized indicator. The z-scores and the

rescaling approach are the most commonly used because they have desirable characteristics

when it comes to aggregation.

Considering our data, where the values of the variables are rather close to each other, the

rescaling method seems most appropriate because it widens the countries’ spread, and

hence, allows better interpretations. Accordingly, we use the rescaling method and convert

all variables of the particular sub-indexes to a common scale from 0 to 100 points. Thereby,

100 represents the best score, while 0 is the worst. For every individual variable, we define

if high values positively or negatively influence the attractiveness for VC/PE investors,

regarding the above-cited literature findings. In our robustness checks, we investigate the

difference resulting from using z-scores for standardization. The next step deals with the

weighting of the individual factors and sub-indexes, and the aggregation of all items to the

VC/PE Attractiveness Index.

4.4 Weighting of the Index Items

If there are no statistical or empirical grounds for choosing a different scheme, we could use

equal weights for the index items to calculate the index. This implies an equal contribution

of all sub-indicators to the VC/PE attractiveness, which is arguable. Equal weighting, as

11

discussed by Nardo et al. (2005a), can be the result of insufficient knowledge about causal

relationships, ignorance about the correct model to apply or even stem from the lack of

consensus on alternative solutions. There are a number of weighting techniques derived

from statistical models. Manly (1994) discusses principal component analysis. Nardo et al.

(2005a) propose factor analysis, and data development analysis. Kaufmann et al. (1999 and

2003) use an unobserved component model. Other weighting techniques are derived from

analytic hierarchy processes, as described in Forman (1983), or Saaty (1987), or from

conjoint analysis, as in Green and Srinivasan (1978), Hair et al. (1998), and McDaniel and

Gates (1998).

We use both, one approach with equal weights among all the sub-index items and one

approach based on factor analysis where we differentiate between the index levels. Level 3

sub-indexes are still equally weighted, but for the level 2 and level 1 sub-indexes we follow

Berlage and Terweduwe (1988). In this weighting method, each component of the level 1

and level 2 sub-indexes is weighted according to its contribution to the total variance in the

data. This is an attractive feature, because it ensures that the resulting summary indicators

account for a large part of the cross-country variance of the underlying items. That makes

this method independent of prior views on their relative economic importance, which is an

arguable issue, but, as highlighted in Nicoletti et al. (2000), these properties are particularly

desirable for cross-country comparisons. However, using the two weighting approaches, we

can investigate the results of both in our robustness check.

A detailed discussion of factor analyses is carried out e.g. in Hair et al. (1998). The general

linear factor model for p observed variables and q factors or latent variables takes the form:

iqiqiii eFFFx ++++= ααα ...2211 (i = 1,…,p) (4).

Where xi represent standardized variables, and αi1,…,αiq are factor loadings related to the

factors Fi,…,Fq, while ei are residuals. We assume that the factors are uncorrelated with each

other, and with the residuals. Further, they have zero means, and unit variance. Additionally,

the residuals are uncorrelated with each other, have zero means, but not necessarily equal

variances.

Now, the most common method used to extract the first m components is principal

component analysis. The decision of when to stop extracting factors depends on the point

when only little “random” variability remains. Various stopping rules have been developed

12

as described in Dunteman (1989): Kaiser’s Criterion, Scree Plot, variance explained criteria,

Joliffe Criterion, Comprehensibility, Bootstrapped Eigenvalues and Eigenvectors. However,

Kaiser’s Criterion is one of the most widely used stopping rules and recommends to drop all

factors with an Eigenvalue below one. Due to Kaiser (1958), most of the total variance is

determined by components beyond the Eigenvalue of one. However, regarding the

Eigenvalues in our sample, there is another large decrease of explained variance below

Kaiser’s mark. As demonstrated in Table 3, we obtain three components that represent

83.8% of the total variance given by the underlying data.

Table 3: Total Variance explained by Components

Total Variance Explained

2,747 45,776 45,776 2,747 45,776 45,776 2,578 42,961 42,9611,306 21,766 67,541 1,306 21,766 67,541 1,272 21,195 64,156,976 16,268 83,810 ,976 16,268 83,810 1,179 19,654 83,810,446 7,431 91,241,300 5,005 96,246,225 3,754 100,000

Component123456

Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance Cumulative %Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings

Extraction Method: Principal Component Analysis.

The next step deals with the rotation of factors (see Table 4). According to Hair et al. (1998)

the usual rotation method is the Varimax Rotation. Rotation is used to minimize the number

of sub-indicators that have a high loading on the same factor. Ideally, each indicator is

loaded exclusively on one of the factors. Kline (1998) points out, that the rotation changes

the factor loadings, and hence, the factors’ interpretation, but leaves the analytical solutions

ex-ante and ex-post rotation unchanged.

Table 4: Rotated Component Matrix

Rotated Component Matrixa

,849 ,077 ,249,700 -,457 ,391,018 ,924 ,154,089 ,133 ,966

,796 ,376 -,088

,851 -,211 ,004

Economic activityCapital marketTaxationInvestor protectionHuman & SocialenvironmentEntrepreneurialopportunity

1 2 3Component

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.

Rotation converged in 5 iterations.a.

Table 4 presents the component matrix after Varimax Rotation and allows an interesting

interpretation of the resulting factors: Economic activity, capital market, human & social

13

environment and entrepreneurial opportunities have high loadings on the first factor. Hence,

it represents the general socio-economic conditions. Taxation and investor protection have

each high loadings on the remaining two factors. Consequently, we can name the two other

factors correspondingly. As a result, the socio-economic environment, the tax regime and

property rights protection determine the index.

The last step (see Table 5) of the weighting procedure deals with the construction of the

weights from the matrix of factor loadings after rotation. The square of a factor loading

represents the proportion of the variance of the indicator explained by the factors. Now, the

three intermediate components are aggregated by weighting each composite using the

proportion of the explained variance in the dataset: 0.513 for the first

(0.513 = 2.578/(2.578+1.272+1.179)), 0.254 for the second component and 0.235 for the

third one. We rescale the final weights to sum up to one to preserve the comparability.

Finally, we can calculate the overall weights as a linear combination of the different

components.

Table 5: Calculation of the weights Private Equity AttractivenessRotated Component Matrix(a) Overall weights

1 2 3 1 2 3Economic activity 0.849 0.077 0.249 0.280 0.005 0.053 0.157Capital market 0.700 -0.457 0.391 0.190 0.164 0.130 0.169Taxation 0.018 0.924 0.154 0.000 0.671 0.020 0.174Investor protection 0.089 0.133 0.966 0.003 0.014 0.791 0.191Human & Social environment 0.796 0.376 -0.088 0.246 0.111 0.007 0.156Entrepreneurial opportunity 0.851 -0.211 0.004 0.281 0.035 0.000 0.153Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a Rotation converged in 5 iterations.Expl. Var 2.578 1.272 1.179 1 1 1 1Expl. /Tot 0.513 0.253 0.235 Sum Sum Sum Sum

Component loadings Component weights

Table 5 presents the weights of the six key drivers (or level 1 sub-indexes). It becomes

obvious that the difference between the weights of the key drivers is not very large,

probably allowing for equal weightings of the sub-indexes as well. However, we will

address this issue in the robustness check.

Anyway, tables 3, 4, and 5 so far present the final procedure for the weights of the already

aggregated level 1 sub-indexes (refer to Figure A1 in the appendix). To determine those

sub-indexes we had to perform the same procedure for all the six key drivers one-step

before. The calculations of the factor analyses for each key driver are described in the

appendix (Tables A1 – A12).

14

Nardo at al. (2005b) discuss the advantages and disadvantages of factor analysis. Factor

analysis can summarize a set of sub-indicators while preserving the maximum possible

proportion of the total variation in the original set. This is a very desirable feature for cross-

country comparisons. Contrarily, the determined factor loadings might not represent the real

influence of sub-indicators. Furthermore, factor analysis is highly sensitive towards

modification of the sample due to data revisions or updates of new countries. Factor analysis

is also very sensitive to the presence of outliers, which may introduce a spurious variability

in the data, and to the sample size.

4.5 Aggregation

There also exist various procedures for the index aggregation. Nardo et al. (2005a and

2005b) distinguish additive methods, geometric aggregation and non-compensatory multi-

criteria analysis. We focus on linear and geometric aggregation because they are in common

use.

Linear aggregation is an additive method and defined as:

∑ ∑ =≤≤=i i

iii wandwwherexwx 1,10, (9).

Geometric aggregation is defined as:

∑∏ =≤≤=i

ii

ii wandwwherexwx 1,10, (10).

Ebert and Welsch (2004) recommend that the linear aggregation method is useful when all

sub-indicators have the same measurement unit, and geometric aggregation is better suited,

if non-comparable and strictly positive sub-indicators are expressed in different ratio scales.

Nardo et al. (2005a) highlight that linear aggregation assigns base indicators proportionally

to the weights, while geometric aggregation rewards those countries or those sub-indicators

with higher scores. Overall, a shortcoming in the value of one variable or sub-index can be

compensated by a surplus in another. Compensability is constant in linear aggregation,

while it is smaller in geometric aggregation for the sub-indicators with low values.

Therefore, countries with low scores in some sub-indexes benefit from linear aggregation.

Due to the properties of the rescaling method from 1 to 100 index points, we prefer linear

aggregation. However, in the robustness analyses we will prove if the geometric aggregation

method yields different index results.

15

16

5. Results

5.1 Base Findings

We calculate the base case of our VC/PE Attractiveness Index according to the procedures

described using rescaling, factor analysis, and linear aggregation. Figure 1 presents the

index rankings for the EU-15 countries, for the CEE countries, for Switzerland and Norway,

and for the GDP-weighted averages of the EU-15 and the CEE. The EU-15 states are chosen

as the benchmark. Their score is indexed to 100 points to simplify country comparisons.

That means that the CEE region, with 85 points, is 15% less attractive compared to EU-15

benchmark.

The top performers are Ireland, Luxembourg, United Kingdom, Sweden and Denmark. The

Central Eastern European Countries lag behind the EU-15 states. The best CEE performer

with 96 index points is Hungary, followed by Slovenia with 95 points, which even rank

before France. The CEE average still ranks before Italy and Spain. However, Bulgaria,

Romania and Slovakia constitute less attractive economies for VC/PE investors while

Greece qualifies as least attractive. The decisive factors for the individual countries shall not

be discussed here, but are presented in the appendix in Tables A12 to A15.

Focusing on the CEE region, and disaggregating the index result to the six key drivers, we

can present the region’s strengths and weaknesses in Figure 2. The chart shows the six key

drivers (and their index weights according to our factor analysis) of attractiveness for the

CEE region, and the EU-15 states as the benchmark. Taxation is the strongest component of

the CEE countries’ attractiveness for VC/PE investors. However, this criterion is highly

dependent on the local legislations, and relatively quickly and arbitrarily adaptable by

politicians. United Nations (2004) reports that CEE governments try to attract investors with

low corporate tax rates and tax incentives within the accession process.

Investor protection & Corporate Governance is another criterion where local legislatives

copied Western European standards to catch up quickly in the accession process. Generally

speaking, investors are as well protected by law on books and by enforcements in the CEE

countries as they are in the average EU-15. Both, the character of the legal rules, and the

quality of law enforcement is covered in the selected sub-indexes. The human & social

environment is also on a par with the EU-15 level. However, the other key drivers cannot

reach the EU-15 average. Economic activity, entrepreneurial opportunities, and particularly

capital markets lag (far) behind the EU-15 countries.

Figure 1: Country Ranking According to the VC/PE Attractiveness Index – Method: Rescaling, Factor Analysis, and Linear Aggregation

69

76

77

79

81

82

85

85

89

100

102

102

104

110

111

121

122

124

127

130

99

93

95

95

96

96

0 100

Greece

Slovakia

Romania

Bulgaria

Italy

Spain

CEE

Czech Republic

Poland

Baltic States

France

Slovenia

Hungary

Portugal

Germany

Austria

Belgium

Netherlands

Finland

Switzerland

Norway

Denmark

Sweden

UK

Luxembourg

Ireland

VC/PE-Attractiveness Index

EU15=100 Index Points

17

Figure 2: Central Eastern European Strengths & Weaknesses (1)

0

50

100

150

200Economic activity ; (0.156)

Capital market; (0.169)

Taxation; (0.174)

Investor protection; (0.190)

Human & Social environment;(0.155)

Entrepreneurial opportunity;(0.153)

CEEEU15

Figure 3 breaks down the index aggregation and presents the level 2 sub-indexes (and their

weights calculated by our factor analysis) for the average of the CEE countries with the EU-

15 states as the benchmark. It reveals that relatively small economies, high unemployment

rates, and small and illiquid capital markets characterize the CEE countries. The capital

markets in particular constitute a strong deficit in every sub-criterion compared to the EU-

15 benchmark.

The Human & social environment of the CEE countries is judged to be equal to the EU-15

average. High educational standards, good labor regulations, and low crime rates constitute

the strengths of the CEE culture. However, bribery and corruption remain higher in the CEE

countries than in the West European benchmarks.

While privatization and large enterprise restructuring processes are nearly completed,

entrepreneurial opportunities are rather weak in CEE. In particular, the burden for starting a

business is much higher than the EU-15 average. Additionally, the innovativeness of the

CEE countries is ranked very poorly. The small number of patents and low public and

private R&D expenditure contribute to that deficit.

18

Figures A54 and A55 in the appendix present the scores of the six key drivers for all of our

sample countries, and for the CEE and the EU-15 averages. Further, spider charts and bar

charts, as in the appendix Figures A2 to A53, allow comprehensive comparisons of all our

sample countries.

Figure 3: Central Eastern European Strengths & Weaknesses (2)

Central Eastern Europe

69.71

58.80

53.48

102.49

24.76

10.80

70.77

31.53

19.03

320.31

100.37

72.34

105.33

126.54

88.83

110.13

52.87

120.79

26.84

39.39

83.65

165.78

67.29

0 50 100 150 200 250 300 350

1.1 Gross Domest ic Product; (0.27)

1.2 General Price Level; (0.25)

1.3 Working force (unemployment rate); (0.25)

1.4 Foreign direct investment, net inf lows; (0.21)

2.1 IPO Volume; (0.23)

2.2 Stock market ; (0.17)

2.3 M &A market; (0.14)

2.4 Credit and Debt market ; (0.20)

2.5 Private equity act ivity; (0.24)

3.1 Highest marginal tax rate, corporate rate; (0.5)

3.2 Dif ference between income and corporate tax rate; (0.5)

4.1 Extent of disclosure index; (0.33)

4.2 Extent of director liability index; (0.33)

4.3 Ease of shareholder suits index; (0.32)

5.1 Education; (0.27)

5.2 Labor regulat ions; (0.22)

5.3 Bribing & corruption; (0.25)

5.4 Criminality; (0.24)

6.1 General Innovativeness; (0.22)

6.2 R&D expenditure; (0.20)

6.3 Enterprise restructuring; (0.19)

6.4 Enterprise stock; (0.20)

6.5 Burden: Start ing a Business; (0.17)

EU 15 =100

5.2 Attractiveness Regarding Actual VC/PE Activity

Klonowski (2005) argues that the activity of VC/PE fundraising indicates the attractiveness

of a particular country to foreign and domestic investors. Hence, there should be a strong

correlation between our VC/PE attractiveness ranking and the fundraising activities in the

various countries. In Figure 4, we present this relationship, adding the size of the individual

19

20

economies measured by GDP per capita into the chart. The expected VC/PE activity is

measured as averages of the ratios between raised funds and GDP for several years to

smooth fluctuations. The data is taken from EVCA (2003, 2004, 2005, and 2006 and an

additional EVCA database on the CEE market). Especially for the CEE countries, the raised

funds fluctuate strongly among different periods. Therefore, we average the raised

funds/GDP ratios for the CEE countries for all the periods since coverage in the EVCA

database.5 For the other countries, we use the average of the ratios from 2003 to 2005. The

chart shows a strong correlation between the actual VC/PE activity in the individual

countries and our index results. This indicates that our index is a good proxy for the

countries’ attractiveness for VC/PE investors. The Pearson Correlation coefficient is 0.571

at a 0.002 significance level (two tailed). If we drop outliers beyond the percentile marked

by two times the standard deviation of one of the variables (namely Ireland because of its

high ranking, and the United Kingdom because of its high activity), the correlation becomes

0.676 at a 0.000 significance level.

Regarding Figure 4, it should be emphasized that there is a large volume of cross border

transactions in some of the countries with large raised fund figures. Especially in the United

Kingdom, a certain amount of the funds raised is allocated to other European economies.

For the CEE market, Poland plays a similar role as a hub for fundraising activities.

Furthermore, a high level of fund raising activities in the CEE region is, in some cases,

attributable linked to a small number of (larger and regional) funds.

5 The coverage in the EVCA database differs for the particular countries. Some countries such as the Czech Republic, Hungary, or Poland are covered since 1998. The Baltic States are least covered, since 2003.

Figure 4: VC/PE-Attractiveness vs. VC/PE-Activity

Slo

vaki

a

Rom

ania B

ulga

riaCze

ch R

epub

lic

Pol

and

Balti

c St

ates

-0.2

5

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

6575

8595

105

115

125

135

VC/P

E-At

trac

tiven

ess

Inde

x 200

6 [EU

15 =

100

]

Expected VC/PE-Activity [raised funds in % of GDP]

Gre

ece

Slo

vaki

aR

oman

iaB

ulga

riaIta

lyS

pain

Cze

ch R

epub

licP

olan

dBa

ltic

Stat

esFr

ance

Slov

enia

Hun

gary

Por

tuga

lG

erm

any

Aus

tria

Bel

gium

Net

herla

nds

Finl

and

Sw

itzer

land

Nor

way

Den

mar

kSw

eden

UK

Luxe

mbo

urg

Irela

nd

Bub

ble

size

= T

otal

GD

P pe

r cap

ita

21

5.3 Robustness Check

As discussed before, the index is affected by the normalization, the weighting, and the

aggregation technique. For the results commented above we used rescaling, factor analysis,

and linear aggregation for the index calculation. Now, we investigate how other

normalization, weighting, and aggregation techniques impact the results. Table 6 presents

the ranks of the individual countries with respect to the approach chosen. The first two

columns show the results for rescaling and linear aggregation, with altered weighting

schemes. In column three and four, we use rescaling and geometrical aggregation, and again

alter the weighting schemes. In column five, we use z-scores as standardization with equal

weights and linear aggregation. The last column presents the volatility of the rankings

measured by the standard deviation of the ranks for each country. The volatility ranges from

0.4 to 2.5 with an average of 1.09. This means that ranks change very little across the

different approaches.

Table 6: Robustness of the VCPEAI

Score Ranking Score Ranking Score Ranking Score Ranking Score RankingAustria 100.10 11 99.94 11 82.80 11 83.85 11 99.42 12 0.4Baltic States 89.63 17 92.61 17 68.65 19 73.16 18 88 17 0.9Belgium 100.01 12 102.01 10 79.61 12 82.17 12 100 10 1.1Bulgaria 74.49 22 79.00 22 40.06 23 44.12 22 75 23 0.5Czech Republic 82.39 19 85.01 19 70.86 17 73.51 17 80 19 1.1Denmark 121.02 5 120.60 5 112.46 3 110.50 3 120 4 1.0Finland 109.16 8 104.22 8 96.35 9 91.21 8 108 8 0.4France 94.94 13 94.96 16 78.61 13 78.37 15 95 13 1.4Germany 100.62 10 99.49 12 89.86 10 87.90 9 99 11 1.1Greece 67.20 25 69.38 25 43.63 22 42.10 23 63 25 1.4Hungary 93.29 15 95.87 14 77.07 14 79.83 14 92 15 0.5Ireland 127.85 1 130.03 1 114.56 2 111.77 2 127 1 0.5Italy 80.99 21 81.22 21 69.44 18 68.69 19 77 21 1.4Luxembourg 124.30 3 127.21 2 96.38 8 98.59 6 121 3 2.5Netherlands 105.00 9 102.34 9 97.24 7 94.05 7 106 9 1.1Norway 110.36 7 110.63 6 105.70 5 105.97 5 110 7 1.0Poland 83.00 18 89.23 18 55.43 21 64.23 20 85 18 1.4Portugal 92.69 16 95.89 13 76.79 15 80.34 13 90 16 1.5Romania 73.34 24 76.69 23 31.15 25 36.85 25 74 24 0.8Slovakia 73.81 23 76.27 24 38.80 24 41.41 24 77 22 0.9Slovenia 94.14 14 95.34 15 75.87 16 77.72 16 93 14 1.0Spain 81.97 20 82.41 20 59.10 20 56.97 21 79 20 0.4Sweden 124.96 2 122.26 4 111.77 4 109.31 4 122 2 1.1Switzerland 117.90 6 109.83 7 97.72 6 87.57 10 114 6 1.7UK 121.37 4 124.26 3 115.54 1 117.61 1 116 5 1.8

Z-scoreEqual weights - LIN

Volatility

Re-scalingEqual weights - GME

Re-scalingFactor analysis - GME

Re-scalingEqual weights - LIN

Re-scalingFactor analysis - LIN

Summarizing, the index is robust against different normalization, weighting, and

aggregation approaches. Figure 5 shows the rankings for the different approaches. The lines

present the span of the ranks determined by the different approaches, while the dots indicate

22

23

the average rank among the five scenarios from Table 6.6 We can clearly identify six tier

groups of attractiveness as marked by the dashed lines. Within those tier groups, the ranks

might change slightly due to the different approaches. However, there is hardly any

transition across the tier groups.

6 Note that the order of the individual countries in Figure 5 is according to their average rank across the different calculation approaches from Table 6. Hence, their order is not the same as in Figure 1.

Figure 5: Box plot of Rankings According to Different Index Calculation Approaches

0 5 10 15 20 25

Irelan

d

UKSw

eden D

enma

rkLu

xemb

ourg

Norw

ay Switz

erlan

dFin

land

Nethe

rland

s Germ

any

Austr

iaBe

lgium

Fran

ceHu

ngary

Portu

gal Sl

oven

iaBa

ltic S

tates

Czec

h Rep

ublic

Polan

d

Italy

Spain

Bulga

riaSlo

vakia

Gree

ce Rom

ania

Average Ranking Position

24

25

6. Conclusions and Outlook

In this paper, we present a composite index to measure the attractiveness of Central Eastern

European countries for VC/PE investors. The index relies on several sub-indexes and raw

data series. We review the literature on attractiveness determinants and refer to the most

important findings for choosing the relevant data series and sub-indexes. For the index

construction, we follow a pyramidal structure of three index levels and show that the

selected data lead to a consistent result. We aggregate the data as suggested in literature and

propose different methods for calculating the index, namely for normalization, weightings,

and aggregation. The different methods give only slightly different results regarding the

sample countries’ attractiveness, indicating the calculations’ robustness. We find six tier

groups regarding the sample countries’ attractiveness rankings. The CEE region lags behind

the average of the EU-15 states. Some of the CEE countries are more attractive for VC/PE

investors than certain EU-15 states. A detailed analysis of the strengths and weaknesses of

the individual countries, as presented in numerous figures in the appendix, provides

conclusions for policy attempts to attract risk capital investors, and hence, to spur

innovation, entrepreneurship, employment, and growth.

The calculation of the index is based on publicly available data and ready-made sub-

indexes. All of the data used are only proxies for the characteristics of the latent drivers. We

have chosen the data series carefully, and we believe to present the best available selection

of proxies for the desired parameters. However, some data series might not adequately

indicate the latent variables in particular countries. Additionally, there are differences

regarding definitions of the data series and regarding the methods to measure and aggregate

the data among the individual countries. This is even valid for the EU-15 states. On the

other hand, it has to be emphasized that whenever the ranking of a country might

unwarrantedly benefit from particularities of an individual index item, the situation might be

reversed at another stage. This is similar to the question about adding, discarding or

exchanging particular data series. Adding, discarding, or exchanging individual index items

will not influence the overall results meaningfully as proved in many unreported robustness

checks.

The availability of the necessary data series limits the scope of our index. The full data set

with the required quality, and based on single sources, is not (yet) available to any useful

26

extent for other regions of the world (beside North America). Hence, further research should

tackle how the quality of our findings changes with a reduced number of index items.

Additionally, other worldwide available data series could qualify as proxies for creating a

worldwide VC/PE Attractiveness Index, and including other emerging regions.

Our economic approach cannot cover special situations or special opportunities in particular

countries. This is notably the case for tax considerations. It is impossible to cover and

compare individual countries’ tax regimes on a general level, especially considering taxes

on dividends, and capital gains taxes, which might be of particular importance for the asset

class in question. Moreover, our approach relies on least available and historic (averaged)

data, and cannot consider the latest changes of individual items. Anyway, we attempt to

contribute to the transparency of the VC/PE fund allocation process, and to discover

strengths and weaknesses of the CEE economies to spur innovation, entrepreneurship,

employment, and growth.

27

References

Balboa, M.; Martí, J. (2003): An integrative approach to the determinants of private equity

fundraising. SSRN working paper 493344.

Baughn, C. C.; Neupert, K. E. (2003): Culture and National Conditions Facilitating

Entrepreneurial Start-ups. In: Journal of International Entrepreneurship, vol. 1, pp. 313-

330.

Belke, A.; Fehn, R.; Foster, N. (2003): Does venture capital investment spur employment

growth? CESIFO working paper 930.

Berlage, L.; Terweduwe, D. (1988): The classification of countries by cluster and by factor

analysis. In: World Development, vol. 16, no. 12, pp. 1527-1545.

Black, B.; Gilson, R. (1998): Venture Capital and the structure of capital markets: Banks

versus stock markets. In: Journal of Financial Economics, vol. 47, no.3, pp. 243-277.

Blanchard, O. J. (1997): The medium run. In: Brookings Papers on Economic Activity,

1997, no. 2, pp. 89-158.

Bruce, D. (2000): Effects of the United States’ tax system on transition into self-

employment. In: Labour Economics, vol. 7, no. 5, pp. 545–574.

Bruce, D. (2002): Taxes and entrepreneurial endurance: Evidence from the self-employed.

In: National Tax Journal, Vol. 55, no. 1, pp. 5-24.

Bruce, D.; Gurley, T. (2005): Taxes and entrepreneurial activity: An empirical

investigation using longitudinal tax return data. Small Business Research Summary, no.

252.

Cetorelli, N.; Gamera, M. (2001): Banking market structure, financial dependence and

growth: International evidence from industry data. In: The Journal of Finance, vol. 56, no. 2,

pp. 617-648.

Chemla, G. (2005): The determinants of investment in private equity and venture capital:

Evidence from American and Canadian pension funds. SSRN working paper 556421.

Cortina, J. M. (1993): What is coefficient alpha? An examination of theory and

applications. In: Journal of Applied Psychology, vol. 78, no. 1, pp. 98-104.

28

Cronbach, L. J. (1951): Coefficient alpha and the international structure of tests. In:

Psychometrika, vol. 16, pp. 297-334.

Cullen, J. B.; Gordon, R. H. (2002): Taxes and entrepreneurial activity: Theory and

evidence for the U.S. NBER working paper 9015.

Cumming, D.; Fleming, G.; Schwienbacher, A. (2006a): Legality and venture capital

exits. In: Journal of Corporate Finance, Vol. 12, pp. 214 – 245

Cumming, D.; Schmidt, D.; Walz, U. (2006b): Legality and venture capital governance

around the world. SSRN working paper 537243

Cumming, D.; Johan, S. (2007): Regulatory harmonization and the development of private

equity markets. In: Journal of Banking and Finance (forthcoming)

Desai, M.; Gompers, P.; Lerner, J. (2006): Institutions and entrepreneurial firm

dynamics: Evidence from Europe. Havard NOM Research Paper 03-59.

Djankov, S.; La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A. (2002): The regulation of

entry. In: Quarterly Journal of Economics, vol. 117, no. 1, pp. 1-37.

Djankov, S.; La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A. (2003): Courts. In:

Quarterly Journal of Economics, vol. 118, no. 2, pp. 453-517.

Djankov, S.; La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A. (2005): The Law and

Economics of Self-Dealing. NBER working paper 11883.

Dunteman, G. H. (1989): Principal components analysis. Newbury Park

Ebert, U.; Welsch, H. (2004): Meaningful environmental indices: a social choice

approach. In: Journal of Environmental Economics and Management, vol. 47, pp. 270-283.

EBRD (2005): Transition Report 2005: Business in Transition. London.

EBRD (2006): Transition Report 2006: Finance in Transition. London.

EVCA (2003): EVCA yearbook 2003 – Annual survey of Pan-European private equity &

venture capital activity. EVCA European Venture Capital Association, Brussels.



29





Fehn, R.; Fuchs, T. (2003): Capital market institutions and venture capital: Do they affect

unemployment and labour demand? CESIFO working paper no. 898.

Feldt, L. S.; Woodruffe, D. J.; Salih, F. A. (1987): Statistical Inference for Coefficient

Alpha. In: Applied Psychological Measurement, vol. 11, no. 1, pp. 93-103.

Forman, E. H. (1983): The analytical hierarchy process as a decision support system. In:

Proceedings of the IEEE Computer society.

Freudenberg, M. (2003): Composite indicators of country performance: a critical

assessment. OECD Economics Department working paper JT00139910.

Glaeser, E. L.; Johnson, S.; Shleifer, A. (2001): Coase vs. the Coasians. In: Quarterly

Journal of Economics, vol. 116, pp. 853-899.

Gompers, P.; Lerner, J. (1998): What Drives Venture Fundraising? In: Brooking Papers

on Economic Activity, Microeconomics, pp. 149-192.

Gompers, P.; Lerner, J. (2000): Money chasing deals? The impact of funds inflows on the

valuation of private equity investments. In: Journal of Financial Economics, vol. 55, no.2,

pp. 281-325.

Green, S. B.; Lissitz, R. W.; Mulaik, S. A. (1977): Limitations of coefficient alpha as an

index of test unidimensionality. In: Educational and Psychological Measurement, vol. 37,

pp. 827-838.

Green, P. E.; Srinivasan, V. (1978): Conjoint analysis in consumer research: issues and

outlook. In: Journal of Consumer Research, vol. 5, pp. 103-123.

Green, P.G. (1998): Dimensions of Perceived Entrepreneurial Obstacles. In: P. Reynolds

(ed.) Frontiers of Entrepreneurship Research. Babson Park: Center for Entrepreneurial

Studies, Babson College, pp. 48-49.

Hair, J. F.; Anderson, R. E.; Tatham, R. L.; Black, W. C. (1998): Multivariate Data

Analysis. fifth ed., Englewood Cliffs.

30

Hattie, J. (1985): Methodology Review: Assessing unidimensionality of test and items. In:

Applied Psychological Measurement, vol. 9, no. 2, pp. 139-164.

Hellmann, T.; Puri, M. (2000): The interaction between product market and financing

strategy: The role of venture capital. In: Review of Financial Studies, vol. 13, no. 4, pp.

959-984.

Hellmann, T.; Lindsey, L.; Puri, M. (2004): Building relationship early: Banks in venture

capital. NBER working paper 10535.

Jacobs, R.; Smith, P.; Goddard, M. (2004): Measuring performance: an examination of

composite performance indicators. Centre for Health Economics, technical paper series 29.

Jeng, L. A.; Wells, Ph. C. (2000): The deteminants of Venture Capital funding: evidence

across countries. In: Journal of Corporate Finance, vol. 6, no. 3, pp. 241-289.

Johnson, S. H.; McMillan, J.;Woodruff, C. M. (1999): Property Rights, Finance and

Entrepreneurship. SSRN working paper 198409.

Kaiser, H.F. (1958): The varimax criterion for analytic rotation in factor analysis. In:

Psychometrika, vol. 23, pp. 187-200.

Kaufmann, D.; Kraay, A.; Zoido-Lobatón, P. (1999): Aggregating Governance

Indicators. World Bank policy research working papers.

Kaufmann, D.; Kraay, A.; Zoido-Lobatón, P. (2003): Governance matters III:

governance Indicators for 1996-2002. World Bank policy research working papers.

Kaplan, S. N.; Schoar, A. (2005): Private equity performance: Returns, persistence, and

capital flows. In: Journal of Finance, vol. 60, no. 4, pp. 1791-1823.

Kline, R. B. (1998): Principles and practice of structural equation modeling. New York

Klonowski, D. (2005): The evolution of the venture capital industry in transition

economies: The case of Poland. In: Post-Communist Economies, Vol. 17, No. 3, September

2005, pp. 331-348.

Knack, S.; Keefer, P. (1995): Institutions and economic performance: Cross-country tests

using alternative institutional measures. In: Economics and Politics, vol. 7, no. 3, pp. 207-

228.

31

Kolodko, G. W. (2000): Globalisation and catching-up: From recession to growth in

transition economies. IMF working paper 00/100.

Kortum, S.; Lerner, J. (2000): Assessing the contribution of venture capital to innovation.

In: Rand Journal Economics, vol. 31, no. 4, pp. 674-692.

La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A.; Vishny, R. (1997): Legal Determinants

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

La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A.; Vishny, R. (1998): Law and finance. In:

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

La Porta, R.; Lopez-de-Silanes, F.; Shleifer, A.; Vishny, R. (2002): Investor Protection

and Corporate Valuation. In: Journal of Finance, vol. 57, no. 3, pp. 1147-1170.

Lazear, E. P. (1990): Job security provisions and employment. In: Quarterly Journal of

Economics, vol. 105, pp. 699-726.

Lee, S. M.; Peterson, S. J. (2000): Culture, Entrepreneurial Orientation and Global

Competitiveness. In: Journal of World Business, vol. 35, no. 4, pp. 401-416.

Lerner, J.; Schoar, A. (2005): Does legal enforcement affect financial transactions? The

contractual channel in private equity. In: Quarterly Journal of Economics, vol. 120, no. 1,

pp. 223-246.

Levine, R. (1997): Financial development and economic growth: Views and agenda. In:

Journal of Economic Literature, vol. 35, pp. 688-726.

Manly, B. (1994): Multivariate statistical methods. London.

Mauro, P. (1995): Corruption and growth. In: Quarterly Journal of Economics, vol. 110,

pp. 681–712.

McDaniel, C.; Gates, R. (1998): Contemporary Marketing Research. Cincinnati.

Megginson, W. (2004): Toward a global model of venture capital? In: Journal of Applied

Corporate Finance, vol. 16, no. 1, pp. 89-107.

Miller, M. B. (1995): Coefficient Alpha: A basic introduction from the perspectives of

classical test theory and structural equation modeling. In: Structural Equation Modelling,

vol. 2, no. 3, pp. 255-273.

32

Nardo, M.; Saisana, M.; Saltelli, A.; Tarantola, S., Hoffman, A.; Giovannini, E.

(2005a): Handbook on constructing composite indicators: Methodology and user guide.

OECD statistics working paper STD/DOC(2005)3.

Nardo, M.; Saisana, M.; Saltelli, A.; Tarantola, S. (2005b): Tools for Composite

Indicators Building. European Commission, Joint Research Centre working paper EUR

21682 EN.

Nicoletti, G.; Scarpetta, S.; Boylaud, O. (2000): Summary indicators of product market

regulation with an extension to employment protection legislation. OECD, Economics

Department working paper 226, ECO/WKP(99)18.

Nunnaly J. (1978): Psychometric Theory, New York.

Poterba, J. (1998): Venture Capital and capital gains taxation. In: Lawrence Summers

(ed.): Tax Policy and the Economy. Cambridge.

Raykov, T. (1998): Cronbach’s Alpha and reliability of composite with interrelated non-

homogenous items. In: Applied Psychological Measurement, vol. 22, pp. 375-385.

Romain, A.; van Pottelsberghe de la Potterie, B. (2004): The Determinants of Venture

Capital: A panel analysis of 16 OECD countries. Université Libre de Bruxelles working

paper WP-CEB 04/015.

Saaty, R. W. (1987): The analytical hierarch process: what it is and how it is used. In:

Mathematical Modelling, vol. 9, pp. 161-176.

Sapienza, H; Manigart, S.; Vermeir, W. (1996): Venture capitalist governance and value

added in four countries. In: Journal of Business Venturing, vol. 11, no. 6, pp. 439-469.

Schertler, A. (2003): Driving Forces of Venture Capital Investments in Europe: a Dynamic

Panel Data Analysis. European Integration, Financial Systems and Corporate Performance

(EIFC) working paper No. 03-27, United Nations University.

Süppel, R. (2003): Comparing economic dynamics in the EU and CEE accession countries.

European Central Bank working paper 267.

Svensson, J. (1998): Investment, property rights and political instability: Theory and

evidence. In: European Economic Review, vol. 42, no. 7, pp. 1317-1341.

33

United Nations (2004): Economic survey of Europe 2004 no.1. United Nations publication

E.04.II.E.7.

Wagner, M.; Hlouskova, J. (2005): CEEC growth projections: Certainly necessary and

necessarily uncertain. In: Economics of Transition, vol. 13, no. 3, pp. 341-372.

Wilken, P. H. (1979): Entrepreneurship: A Comparative and Historical Study, Norwood.

Appendix

Table(s) A1: Factor Analysis - Economic activity

Communalities

1,000 ,8231,000 ,7911,000 ,789

1,000 ,640

Gross Domestic ProductGeneral Price LevelWorking forceForeign direct investment,net inflows [% of GDP]

Initial Extraction



1,892 47,289 47,289 1,892 47,289 47,289 1,715 42,887 42,8871,152 28,810 76,099 1,152 28,810 76,099 1,328 33,212 76,099,697 17,414 93,513,259 6,487 100,000

Component1234




,881 ,216-,141 ,878,556 ,693

,781 -,175

Gross Domestic ProductGeneral Price LevelWorking forceForeign direct investment,net inflows [% of GDP]

1 2Component



Table A2: Weights - Economic activity Economic activityRotated Component Matrix(a) Overall weights

1 2 1 2 WeightsGross Domestic Product 0.881 0.216 0.453 0.035 0.270General Price Level -0.141 0.878 0.012 0.580 0.260Working force 0.556 0.693 0.180 0.362 0.259Foreign direct investment, net inflows [% of GDP] 0.781 -0.175 0.355 0.023 0.210Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a Rotation converged in 3 iterations.Expl. Var 1.715 1.328 1 1 1Expl. /Tot 0.564 0.436 Sum


Sum

34

Table(s) A3: Factor Analysis - Capital market

Communalities

1,000 ,8991,000 ,6401,000 ,5421,000 ,7701,000 ,907

IPOStock marketM&A market activityCredit and Debt marketPrivate equity activity

Initial Extraction



2,606 52,117 52,117 2,606 52,117 52,117 2,011 40,226 40,2261,153 23,064 75,181 1,153 23,064 75,181 1,748 34,955 75,181,740 14,805 89,986,357 7,150 97,136,143 2,864 100,000

Component12345




,948 -,009,522 ,607,068 ,733,108 ,871,908 ,289

IPOStock marketM&A market activityCredit and Debt marketPrivate equity activity

1 2Component



Table A4: Weights - Capital market Capital marketRotated Component Matrix(a) Overall weights

1 2 1 2 WeightsIPO 0.948 -0.009 0.447 0.000 0.239Stock market 0.522 0.607 0.135 0.211 0.170M&A market activity 0.068 0.733 0.002 0.308 0.144Credit and Debt market 0.108 0.871 0.006 0.434 0.205Private equity activity 0.908 0.289 0.410 0.048 0.241Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a 1 components extracted.Expl. Var 2.011 1.748 1 1 1Expl. /Tot 0.535 0.465 SumSum


35

Table(s) A6: Factor Analysis - Investor protection

Communalities

1,000 ,767

1,000 ,760

1,000 ,729

Extent of disclosure indexExtent of director liabilityindexEase of shareholdersuits index

Initial Extraction



1,147 38,248 38,248 1,147 38,248 38,248 1,133 37,765 37,7651,109 36,983 75,231 1,109 36,983 75,231 1,124 37,466 75,231,743 24,769 100,000

Component123




-,167 ,860

,839 -,239

,634 ,572

Extent of disclosure indexExtent of director liabilityindexEase of shareholdersuits index

1 2Component



Table A7: Weights - Investor protection Investor protectionRotated Component Matrix(a) Overall weights

1 2 1 2Extent of disclosure index -0.167 0.860 0.025 0.658 0.340Extent of director liability index 0.839 -0.239 0.621 0.051 0.337Ease of shareholder suits index 0.634 0.572 0.354 0.292 0.323Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a Rotation converged in 3 iterations.Expl. Var 1.133 1.124 1 1 1Expl. /Tot 0.502 0.498 SumSum


36

Table(s) A8: Factor Analysis - Human & Social environment

Communalities

1,000 ,8721,000 ,7101,000 ,8121,000 ,794

EducationLabor regulationsBribing & corruptionCrime

Initial Extraction



2,045 51,131 51,131 2,045 51,131 51,131 2,045 51,128 51,1281,143 28,586 79,718 1,143 28,586 79,718 1,144 28,590 79,718,525 13,118 92,835,287 7,165 100,000

Component1234




,145 ,923,666 -,517,893 ,117

-,885 -,107

EducationLabor regulationsBribing & corruptionCrime

1 2Component



Table A9: Weights - Human & Social environment Human & Social environmentRotated Component Matrix(a) Overall weights

1 2 1 2Education 0.145 0.923 0.010 0.745 0.274Labor regulations 0.666 -0.517 0.217 0.233 0.223Bribing & corruption 0.893 0.117 0.390 0.012 0.255Crime -0.885 -0.107 0.383 0.010 0.249Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a Rotation converged in 3 iterations.Expl. Var 2.045 1.144 1 1 1Expl. /Tot 0.641 0.359 SumSum


37

Table(s) A10: Factor Analysis - Entrepreneurial opportunity

Communalities

1,000 ,9291,000 ,8571,000 ,7821,000 ,823

1,000 ,696

General InnovativenessR&D expenditureEnterprise restructuringEnterprise stock activityBurden: Starting aBusiness

Initial Extraction



2,988 59,754 59,754 2,988 59,754 59,754 2,421 48,412 48,4121,100 22,009 81,763 1,100 22,009 81,763 1,668 33,351 81,763,486 9,727 91,490,372 7,431 98,921,054 1,079 100,000

Component12345




,854 ,447,855 ,355,875 -,132,019 -,907

,441 ,709

General InnovativenessR&D expenditureEnterprise restructuringEnterprise stock activityBurden: Starting aBusiness

1 2Component



Table A11: Weights - Entrepreneurial opportunity Entrepreneurial opportunityRotated Component Matrix(a) Overall weights

1 2 1 2General Innovativeness 0.854 0.447 0.301 0.120 0.227R&D expenditure 0.855 0.355 0.302 0.076 0.210Enterprise restructuring 0.875 -0.132 0.316 0.010 0.191Enterprise stock activity 0.019 -0.907 0.000 0.493 0.201Burden: Starting a Business 0.441 0.709 0.080 0.301 0.170Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a Rotation converged in 3 iterations.Expl. Var 2.421 1.668 1 1 1Expl. /Tot 0.592 0.408 SumSum


38

Table A12: Level I Performance – EU (EU15=100)7

Economic activity

Capital market Taxation

Investor protection

Human & Social environment

Entrepreneurial opportunity

Weights 0.16 0.17 0.17 0.19 0.16 0.15Ireland 125.96 62.39 281.63 135.61 115.17 88.84Luxembourg 155.34 120.63 152.76 130.12 112.33 98.86UK 109.84 214.36 79.08 142.08 100.19 110.04Sweden 111.96 124.99 192.45 82.38 110.29 143.39Denmark 115.25 70.06 187.88 112.29 120.89 125.35Norway 114.87 68.37 135.41 118.21 112.78 107.62Switzerland 119.23 113.03 135.91 70.72 119.44 118.28Finland 97.97 87.81 93.63 100.38 99.85 136.70Netherlands 116.29 115.57 117.26 76.81 99.99 102.58Belgium 107.52 73.11 80.50 130.12 96.47 102.18Austria 110.19 46.52 179.97 65.33 109.59 104.99Germany 92.94 70.57 138.49 94.64 91.83 112.94Portugal 99.00 55.74 134.34 106.60 110.96 67.54Hungary 95.14 22.34 224.69 82.09 91.71 82.35Slovenia 96.46 21.76 179.97 100.46 104.23 73.82France 93.75 85.09 74.04 95.15 96.79 114.18Baltic States 82.23 27.55 178.48 100.50 99.85 74.15Poland 53.06 23.15 200.66 112.04 90.45 68.25Czech Republic 95.27 43.78 86.79 88.07 98.37 84.75CEE 63.36 27.91 169.03 98.72 91.10 66.79Spain 89.48 66.92 46.63 83.22 108.60 82.61Italy 87.40 54.33 59.61 83.14 94.90 91.53Bulgaria 35.40 35.21 171.86 94.56 95.85 55.72Romania 74.94 26.07 124.30 101.21 69.36 59.63Slovakia 66.26 39.69 104.84 70.77 103.00 72.33Greece 79.84 38.65 56.97 59.11 94.40 75.46

7 Single observation and comparison of each item is made possible in this table.

39

Table A13: Level II Performance – EU (1) (EU15=100)8

Gross Domestic Product

General Price Level

Working force (unemployment rate)

Foreign direct investment, net inflows IPO Volume Stock market M&A market

Credit and Debt market

Private equity activity

Code 1.1 1.2 1.3 1.4 2.1 2.2 2.3 2.4 2.5Weights 0.27 0.26 0.26 0.21 0.24 0.17 0.14 0.20 0.24Ireland 174.00 94.90 135.74 169.20 37.39 49.11 115.08 101.47 11.01Luxembourg 233.45 100.77 141.52 325.67 1.00 78.70 397.05 157.81 80.64UK 115.74 98.92 129.31 91.22 327.94 179.80 187.53 96.77 330.37Sweden 117.68 101.57 129.89 95.86 69.15 144.83 112.54 94.27 218.13Denmark 138.62 99.94 128.10 114.17 10.54 57.93 74.43 113.49 69.29Norway 188.00 95.21 137.77 49.88 45.09 44.07 80.40 99.49 57.75Switzerland 135.74 103.46 142.42 97.94 18.82 266.65 118.93 122.95 54.02Finland 106.38 101.07 90.02 99.50 28.79 178.47 117.74 87.93 52.35Netherlands 118.42 97.67 136.33 130.19 38.62 169.65 166.68 113.35 123.20Belgium 104.79 101.34 102.95 147.48 114.84 54.50 61.43 87.12 31.39Austria 109.07 102.23 135.17 71.78 15.26 10.41 31.56 109.43 17.34Germany 89.35 102.70 85.95 78.29 38.47 66.91 152.81 97.99 21.62Portugal 55.27 95.62 123.19 89.84 12.51 39.76 24.07 113.64 37.13Hungary 83.95 79.60 117.07 105.67 21.66 19.17 45.29 16.48 23.64Slovenia 92.32 88.69 116.48 74.58 1.00 9.42 58.04 27.49 26.19France 100.68 101.84 85.40 78.45 97.34 94.55 42.66 91.50 75.51Baltic States 85.66 91.27 62.11 100.17 2.29 8.76 15.33 56.76 29.15Poland 40.11 87.13 1.00 80.74 41.20 10.77 55.58 8.75 22.62Czech Republic 61.67 93.24 99.61 126.38 13.73 15.23 110.78 70.29 23.60CEE 69.71 58.80 53.48 102.49 24.76 10.80 70.77 31.53 19.03Spain 107.67 97.26 67.46 102.20 3.77 137.55 38.52 97.08 34.63Italy 89.65 98.33 90.81 32.61 36.17 61.56 19.62 88.94 28.71Bulgaria 67.28 1.00 36.00 134.68 1.00 1.30 147.11 62.07 1.00Romania 199.08 14.71 104.67 95.63 25.09 1.54 28.04 47.27 14.04Slovakia 72.90 92.60 5.23 128.92 1.00 2.58 219.40 47.64 8.91Greece 93.51 93.56 83.64 1.00 1.09 66.82 1.00 80.55 3.98

8 Single observation and comparison of each item is made possible in this table.

40

Table A14: Level II Performance - EU (2) (EU15=100)9

Highest marginal tax rate, corporate rate

Difference between income and corporate tax rate

Extent of disclosure index

Extent of director liability index

Ease of shareholder suits index Education

Labor regulations

Bribing & corruption Criminality

Code 3.1 3.2 4.1 4.2 4.3 5.1 5.2 5.3 5.4Weights 0.50 0.50 0.34 0.34 0.32 0.27 0.22 0.25 0.25Ireland 468.77 196.70 131.19 117.15 155.36 102.86 121.96 97.51 135.

the attractiveness of central eastern european countries ...€¦ · country rankings that vary...

Documents