Oxford Economic Papers Advance Access originally published online on June 12, 2006
Oxford Economic Papers 2006 58(4):596-635; doi:10.1093/oep/gpl007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© Oxford University Press 2006 All rights reserved
The legislative road to Silicon Valley
,
*Faculty of Law and Centre for Business Research, University of Cambridge, Trinity Hall, Cambridge CB2 1TJ, UK
Director, Severino Center for Technological Entrepreneurship, Lally School of Management and Technology, Rensselaer Polytechnic Institute, Troy, NY 12180
Correspondence: e-mail: j.armour{at}law.cam.ac.uk
Correspondence: e-mail: Douglas{at}Cumming.com
 |
Abstract |
|---|
 TOP Abstract 1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
Must policymakers seeking to replicate the success of Silicon Valley's venture capital market first copy other US institutions, such as deep and liquid stock markets? Or can legislative reforms alone make a significant difference? In this paper, we compare the economic and legal determinants of venture capital investment, fundraising, and exits. We introduce a cross-sectional and time series empirical analysis across 15 countries and 14 years of data spanning an entire business cycle. We show that liberal bankruptcy laws stimulate entrepreneurial demand for venture capital; that government programmes more often hinder than help the development of private equity, and that the legal environment matters as much as the strength of stock markets. Our results imply generalizable lessons for legal reform.
Key Words: JEL classification: G24 • G28 • G33 • G38 • H25
|
1. Introduction |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
An important question asked by industrial policymakers around the world is: ‘how do we replicate the success of Silicon Valley's entrepreneurialism and venture capital finance?’ Whilst the story of Silicon Valley is surely multi-faceted (see Kenny, 2000
), one crucial factor appears to be the use of venture capital, a form of financial intermediation that seems particularly well-matched to the development of innovative, high-tech products. Understanding the way in which venture capital operates, and the economic, institutional, and legal factors that help it to flourish, are therefore important questions for research.
The structure of venture capital investment has in recent years received considerable attention (for reviews, see Klausner and Litvak, 2001; Gompers and Lerner, 2001; Armour, 2003). The principal proposition established in the literature is that venture capital flourishes in countries with deep and liquid stock markets. We build upon prior work by considering how legislative enactments make a difference to venture capital finance. We employ a cross-sectional and time series empirical analysis across an entire business cycle of data drawn from 15 Western European and North American countries. Using simultaneous equations methods, we distinguish between those independent variables that affect the supply of investment funds from venture capitalists to entrepreneurial firms, and those that affect the demand for equity finance by such firms.
Like earlier studies (Gompers and Lerner, 1999; Jeng and Wells, 2000; Cumming and MacIntosh, 2006), our empirical results show that economic factors are important determinants of venture capital investment. This paper's contribution, however, lies in the findings relating to the role of legal variables. For the first time, we employ an index of legal and fiscal variables that pertain specifically to factors considered important by a leading trade association, the European Venture Capital Association (EVCA). Our empirical findings show that the EVCA index of the ‘investor friendliness’ of country's legal and fiscal environment is a significant determinant of the supply of venture capital investment to entrepreneurial firms, and also of fundraising and exit activity by venture capitalists (VCs). Secondly, we investigate the impact of publicly funded venture capital programs on aggregate levels of venture capital investment and fundraising. We find that the introduction of significant publicly-sponsored programs is not associated with any overall increase—and is in some cases associated with a decrease—of the overall level of venture capital investment, even accounting for the possible endogeneity of government programs to low national levels of venture capital activity. Thirdly, we focus on the role of bankruptcy law, hitherto ignored in the venture capital literature. We show that countries with less liberal personal bankruptcy laws, measured by reference to the number of years before a bankrupt individual would obtain a ‘fresh start’ (that is, a discharge from pre-bankruptcy indebtedness) and controlling for countries in which no fresh start is available, have significantly lower demand for venture capital and private equity.
The rest of the paper is structured as follows. Section 2 provides a brief description of the structure of venture capital finance, explains why it is of interest to academics and policymakers, and reviews literature on its determinants. We develop three hypotheses in Section 3, relating respectively to the general legal environment (as measured by the EVCA index), the direct investment of government funds, and the liberality of personal bankruptcy laws. Section 4 describes our dataset and empirical methods. Section 5 sets out our results, and Section 6 explains their implications.
|
2. Venture capital and its determinants |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
2.1 What is venture capital?
Venture capital is a subset of private equity investment, distinguished by the fact that funds are advanced to businesses that are starting up or at an early stage in their development—that is, before a profit has been earned. VCs are financial intermediaries, raising their investment funds from end-investors, the most important of which are institutions managing collective investments. The VCs in turn are active investors in their portfolio companies. They ameliorate agency problems between themselves and entrepreneurs by developing specialist industry expertise and using sophisticated contract terms that give the VC a significant role in the governance of portfolio companies (Sahlman, 1990
; Manigart et al., 1996, 2000, 2002; Gompers and Lerner, 1999; Bascha and Walz, 2001, 2002; Kaplan and Strömberg, 2003; Kaplan et al., 2003; Lerner and Schoar, 2005; Mayer et al., 2005; Cumming, 2005, 2006). VCs typically hold their investments for a period of between 3–7 years, after which successful investments are exited either by listing the company through an initial public offering (IPO), or by selling the company to a competitor (a ‘trade sale’). Unsuccessful investments are liquidated. One good investment can earn enough to outweigh several write-offs and thereby generate a healthy overall portfolio return (Schwienbacher, 2002; Hege et al., 2003; Cumming and MacIntosh, 2003; Cumming et al., 2006).
2.2 Venture capital and innovation
Venture capital is thought to be of disproportionate importance in stimulating innovation (Kortum and Lerner, 2000; Tykvová, 2000; Lerner, 2002b; Lerner et al., 2003). ‘Start-up’ firms developing new technologies commonly do not generate steady cash flows that can be used to make interest payments, and lack liquid assets that could be used as collateral. Instead, the value (if any) of a start-up firm will inhere in the ideas and ‘human capital’ of the entrepreneur and opportunities for growth. This makes such firms unsuitable candidates for debt investment (Berger and Udell, 1998; Carpenter and Petersen, 2002). Rather, there is a strong complementarity between ‘soft’ assets and concentrated equity finance, in the form of venture capital. Empirical findings confirm that equity financing in the form of venture capital, and not debt, predominates in privately-held firms in technology-intensive industries (Freear and Wetzel, 1990; Carpenter and Petersen, 2002).
2.3 Economic determinants
Understanding the determinants of venture capital finance is an important research question for both policymakers and academics. The orthodox wisdom suggests that economic and institutional variables—in particular, economic growth, size and liquidity of stock markets, and returns to stock market investments—are probably the most important determinants. Most obviously, levels of venture capital investment vary both across time and countries (Gompers and Lerner, 1999; Jeng and Wells, 2000; Mayer, 2001), closely tracking business cycles in the economy generally. Theory and evidence also indicate a strong link between the size and liquidity of a nation's stock markets and the extent of its venture capital investment market (Gompers, 2001; Gompers and Lerner, 1991, 1999, 2001; Black and Gilson, 1998; Jeng and Wells, 2000; Lerner, 1999, 2002a; Mayer et al., 2005). If these factors are the most significant determinants of venture capital finance, the question of how to stimulate a venture capital market in systems without deep and liquid stock markets becomes one of ‘chicken and egg’. It is necessary to solve a ‘simultaneity’ problem—that is, capital, VCs, and entrepreneurs must all be present simultaneously in order for a thriving market to develop. This presents policymakers with a formidably difficult engineering problem (Gilson, 2003).
2.4 Legal and fiscal determinants
A related theoretical literature suggests that a range of legal and fiscal factors may also be determinants of both supply of, and demand for, venture capital finance (see, for example, Cosh and Wood, 1998; Keuschnigg and Nielsen, 2001, 2003a, 2003b, 2004; Cressy, 2002; Armour, 2003; Kanniainen and Keuschnigg, 2003, 2004; Keuschnigg, 2003, 2004). In Keuschnigg (2003) and Keuschnigg and Nielsen (2003a, 2004), reductions in capital gains tax are a significant stimulant to venture capital activity, whereas the provision of subsidies to venture capital investment are welfare-reducing. This is because both taxes and investment subsidies drive down profits in the venture capital industry, thus impeding the VCs’ ability to provide hands-on governance for their portfolio companies. However, the effects of some areas of law that may be important for entrepreneurship—such as bankruptcy law—have not been considered in relation to venture capital markets.
These theoretical results receive some empirical support from studies, principally of US time-series data, which show levels of venture capital investment to be affected by the regulation of pension funds (Gompers and Lerner, 1999), levels of capital gains tax (Poterba, 1989a,b; Gompers, 1998; Gompers and Lerner, 1998; Da Rin et al., 2005) and the provision of state subsidies to stimulate the development of venture capital markets (Lerner, 1999, 2002b). There is, however, controversy as to whether such subsidies have a positive or negative impact on venture capital markets (Leleux and Surlemount, 2003; Cumming and MacIntosh, 2005; Cumming, 2007). How, if at all, these various factors make a difference across countries is less clearly understood.
|
3. Theories and formulation of hypotheses |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
To investigate differences in demand for and supply of venture capital investment across countries, it is important to identify an appropriate menu of legal variables, and within it, to distinguish those which are supply-based from those which are demand-based.
3.1 Supply-side: does law matter at all?
The existing literature on cross-country comparisons has tended to fail, on the one hand, to distinguish variables affecting supply from those affecting demand, and on the other hand, to specify legal variables in accordance with a clearly-articulated theory of their expected impact. Thus, a number of studies have investigated whether legal rules affect venture capital finance by using as independent variables a range of ‘legal’ indices drawn from the work of La Porta et al. (1997, 1998)—for example, minority shareholder rights, anti-director rights, and creditor rights (Jeng and Wells, 2000; Allen and Song, 2003; Lerner and Schoar, 2005). Such factors are unlikely to have much impact on venture capital investment activity, as the rights of VCs derive largely from their investment contracts, as opposed to general corporate law (Gompers and Lerner, 1999). Unsurprisingly, several of these studies report findings of no correlation (Jeng and Wells, 2000) or even negative correlations (Allen and Song, 2003) between shareholder rights-type variables and venture capital investments.1
The European Venture Capital Association has in recent years been a vocal lobbyist of European governments, arguing for changes in local laws designed to facilitate venture capital and private equity investment. A recent report (EVCA, 2003) establishes an ‘investor friendliness’ index for national fiscal and legal environments, taking into account a multitude of legal and taxation measures which market professionals consider in their experience to be relevant to investors’ willingness to supply venture capital finance. The EVCA index is a composite ranking of many factors, including tax transparency for domestic investors, the ability to avoid permanent establishment for international investors from treaty or non-treaty countries, the ability to incorporate a tax efficient capital investment regarding incentives for fund managers, the ability to avoid paying value-added tax (VAT) on management charges and/or carried interest, the degree of restrictions on investments, the content of merger regulations (in particular, whether there is an obligation to suspend a deal until the responsible authority makes a decision), the regulation of pension funds’ ability to invest in venture capital, the rate of corporate taxation on profits and dividends, the rate of corporate taxation for SMEs, the capital gains tax rate, the tax incentives for individual investors, stock options taxation, fiscal R&D incentives, and time and capital involved in setting up a private limited partnership or company. The EVCA index is structured in a way that a lower number (on a three-point scale) indicates a better legal and tax environment for the venture capital or private equity fund itself. We make use of this index in our empirical study to identify equations for the supply of venture capital and private equity across countries. For reasons of collinearity across variables discussed further below, we do not use each of these variables separately. The composite EVCA index provides a useful starting-point to test the idea that the legal and fiscal environment affects the supply of venture capital investment. If a composite measure of law's impact on venture capital investors, developed intuitively by the investors themselves, is not a determinant of investment, then the hypothesis that law matters will have been quickly falsified.
Hypothesis 1
If the legal environment affects levels of venture capital investment, we would expect that investment, fundraising and dispositions would be inversely correlated with the EVCA index (in which lower numbers indicate regimes more favourable to investors).
3.2 Supply-side: government funds
A more specific question is whether it is possible for governments to stimulate the development of a venture capital industry by setting up publicly-funded schemes. Such funds may act to ‘pump prime’ demand by entrepreneurs, thereby opening up new markets. Whether this will happen depends on the precise legal framework within which the public funds are established (Keushnigg and Nielsen, 2001, 2004; Lerner, 2002b; Gilson, 2003; Keuschnigg, 2003). To see this, consider that if the incentives of the public fund managers are set inappropriately by mandatory regulation (as opposed to the privately-negotiated covenants used to incentivize private venture capital managers: Gompers and Lerner, 1996), such managers may waste public money,2 or, worse still, ‘crowd out’ private money by competing for the same investments as private VCs. Crowding out occurs where, in the expectation of being outbid for attractive investments by public funds with laxer budget constraints, end-investors commit less money to private equity funds—quite the opposite effect to that desired by policymakers. In Canada, Cumming and MacIntosh (2006) provide evidence consistent with such a crowding out effect.3 Of course, it is perfectly plausible that a public scheme could be designed with more effective governance so as to achieve good rates of return and avoid crowding out. This leads us to formulate two alternative hypotheses:
Hypothesis 2a
Government-sponsored funds seed private investment and increase the overall level of venture capital investment and fundraising.
Hypothesis 2b
Government-sponsored funds crowd out private investment and reduce the overall level of venture capital investment and fundraising.
3.3 Demand side: personal bankruptcy laws and the value of a fresh start
Bankruptcy law provides an orderly mechanism for the realization of an insolvent's assets, which are liquidated to pay creditors. Personal bankruptcy law—that is, the regime that governs the insolvency of individuals—also serves to punish or to rehabilitate financially distressed individuals. The nature and extent of the consequences of personal bankruptcy may be expected to impact upon entrepreneurial incentives and thereby on demand for venture capital finance.
Consider first an ex ante effect. The process of raising venture capital finance itself involves transaction costs. A putative entrepreneur must have at least the genesis of a product, develop a credible business plan and assemble a team in order to convince a VC to invest. Start-up enterprises are of course incorporated. Yet the funds to cover this ‘pre-seed’ stage will need to be sourced from personal finances, or those of friends and family. If, having expended these resources, the entrepreneur then fails to raise venture capital finance,4 he will find himself in a position of considerable financial fragility, with depleted resources, no job, and consequently a greatly increased risk of personal bankruptcy. If potential entrepreneurs have heterogeneous risk preferences, then more severe legal penalties for bankruptcy will, ceteris paribus, tend to make marginal would-be entrepreneurs less willing to leave paid employment so as to pursue innovative business ideas (Georgellis and Wall, 2002; Fan and White, 2003). This in turn will reduce demand for venture capital finance.
Secondly, bankruptcy law will have an ex post effect on the ease with which entrepreneurs may be rehabilitated into the economy after a bankruptcy. Bankruptcy is just as likely to occur because of ‘bad luck’ as because of incompetence on the part of the entrepreneur. If bankruptcy legally disables failed entrepreneurs from re-engaging in business, then they only have one chance to fail. Conversely, a readily available fresh start means that failed entrepreneurs can be rapidly rehabilitated (Jackson, 1985; Georgakopoulos, 2002).
Many countries’ bankruptcy laws do allow a debtor to obtain a fresh start, whereby, after a certain period of time, he is permitted to discharge outstanding credit obligations (Hallinan, 1986). Others, however, do not (see Armour, 2004, for details). For those that do, the length of time which must elapse, and the other conditions which must be fulfilled (for example, demonstration of good behaviour), vary considerably.5 Unless or until such a discharge is granted, any assets the debtor has or acquires will be subject to the claims of his creditors, and so his ability to raise fresh finance will be severely hampered.6 Thus, we formulate our third hypothesis in terms of time to discharge:
Hypothesis 3
Personal bankruptcy laws that treat failed entrepreneurs more liberally, in the sense that they offer a fresh start quickly, will stimulate demand for venture capital finance.
|
4. Data and summary statistics |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
These hypotheses are tested using trade association data on venture capital markets in 15 Western European and North American countries,7 over a period of 14 years (1990–2003). The data are pooled (as described in, for example, Judge et al., 1988
) to form a total of 210 observations. Developing countries are excluded because of pronounced institutional differences that give rise to problems associated with combining analyses and data across countries (see, for example, Gompers et al., 2003, and Lerner and Schoar, 2005). Our analysis is based on countries with legal and institutional structures that have significant differences for the purpose of comparative analyses, but not so different as to call for an entirely different empirical approach for subsets of the data. Moreover, we do not consider the period prior to 1990, because before then, venture capital markets in certain of the sample countries were not well developed and associated data are consequently less reliable. The variables and data are summarized in Table 1 Panel A.
|
For comparative purposes, the venture capital data are scaled by the GDP of each country. A snapshot of the investment data is provided in Figure 1, and the raw data are presented in Table 1 Panel B. These show the total amount of early and expansion stage venture capital investment, total private equity (including all types of venture capital and other forms of private equity such as late stage, buyout and turnaround investments), total fundraising (capital flows from institutional investors into private equity funds, for all types of private equity in each country), and total dispositions or exits (the value of all sale transactions through either IPOs, acquisitions, secondary sales, buybacks and write-offs), each expressed as a proportion of GDP in each country. The values are averaged across the period 1990–2003. There are lags between the time that venture capital and private equity funds receive capital for investment from institutional investors and the time that capital is reinvested into entrepreneurial firms (Gompers and Lerner, 1999
, 2001), hence the total fundraising values do not match the total private equity investment values.
|
|
Figure 1 and Table 1 Panel B indicate that the largest market in terms of early and expansion stage venture capital investment relative to GDP is the US. The UK has the largest total private equity market and greatest amount of fundraising relative to GDP, which is attributable to the greater number of large scale buyout transactions. UK private equity investors have also brought about the largest average value of exits relative to GDP, followed closely by the US. Austria has the smallest venture capital and private equity market on each of the metrics.
Table 2 provides an initial insight into the rationales for the relative size of the values presented in Figure 1 and Table 1 Panel B. Table 2 provides a number of comparison of means tests depending on the value of the country-specific Morgan Stanley Capital International (MSCI) index of stock market returns (lagged one year), real GDP growth (lagged one year), a dummy variable for the 1999–2000 ‘bubble’ period, the proportion of self-employment (lagged one year), the number of patents (lagged one year), the number of years to discharge in bankruptcy, the EVCA tax and legal index (a three-point scale, as explained in Section 3.1), and the extent to which government funds participate in the venture capital market. These comparison of means tests strongly indicate that early and expansion stage venture capital investment, as well as total private equity investment (including buyouts; on which, see Wright et al., 2001), fundraising and exits (all relative to GDP) are each higher when the prior year's MSCI returns and real GDP growth are higher, when there has been greater patent activity in the prior year, and in bubble periods such as 1999–2000. The data also show that legal factors are significant: early stage, expansion stage and total private equity investment, as well as fundraising and exits (all relative to GDP) are each higher when discharge in bankruptcy is available (or is so in a shorter time), when the EVCA tax and legal index is lower (a lower value indicating a more favourable tax and legal environment for venture capital and private equity investors), and when capital gains taxes are lower.8 Most of these effects for each variable are statistically significant. Table 2 indicates that the total size of the venture capital and private equity industries, as well as the amounts of fundraising and exits, are smaller where government programs comprise a larger percentage of total industry fundraising. The economic and statistical significance of all of these variables are assessed in the context of multivariate regressions in Section 5.
|
Figures 2 to 4 are presented to complement the presentation of the data in Table 2, with a focus on bankruptcy legislation. Figure 2 indicates that countries in which a fresh start is available in bankruptcy have seen much more early stage venture capital investment relative to GDP levels for all years, with the exception of 2003. The spike in the bubble years (1999–2000) indicates that venture capital was particularly prone to boom and bust period in countries in which a bankruptcy discharge was unavailable. That is, harsh bankruptcy laws appear to mitigate swings in early stage venture capital activity. Similarly, Figs 3 and 4 indicate that countries in which a fresh start is available systematically have greater levels of fundraising and profits.
|
|
|
Table 3 explores the relations across the economic and legal variables, along with country-specific dummy variables. Table 3 also provides guidance for the simultaneous use of different variables in regression models. Below we present a concise set of regression results that are quite robust to various specifications. It is noteworthy that some of the policy instruments (such as government programs, taxation, and time to discharge in bankruptcy) are highly correlated in Table 3. As such, it is worthwhile to use multivariate regressions to study the relevance of each variable's marginal impact on the overall venture capital market. It is also worth noting that although in Table 3 government programs are pronounced in countries that appear to have robust early stage venture capital markets, this univariate analysis of the data may be misleading, because such countries also have liberal bankruptcy laws and generally favourable legal environments (see the correlations in Table 3 between bankruptcy and government programs). Multivariate regressions will allow us to separate these different policy effects.
|
|
5. Multivariate empirical tests |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
This section presents empirical tests as follows. Section 5.1 describes a system of equations which distinguishes factors that affect the supply of venture capital from those that affect demand. Thereafter, results pertinent to early and expansion stage investment (Section 5.2), total private equity investing and fundraising (Section 5.3), and total dispositions (Section 5.4) are presented. Section 5.5 discusses limitations and possible future research.
5.1 Supply and demand estimations
The data are pooled and stacked by country and year to comprise 210 observations in total. We begin by presenting a set of country-fixed-effect regression results that do not distinguish between supply and demand factors in explaining the level of venture capital investment activity per dollar of GDP in each country-year. This is a useful first step for comparative purposes, since it provides preliminary results using a similar methodology to other studies in the venture capital literature.
Thereafter, to investigate the demand for venture capital funds by entrepreneurial firms and the supply of funds by VCs (and their institutional investors), we use three-stage least squares to estimate the following two equations. As discussed herein, the results are quite robust to alternative specifications.9
- Demand for Capital =
1 + ß11 MSCI Public Market Return (contemporary/lagged 1 year) + ß12 Real GDP Growth (lagged 1 year) + ß13 Dummy Variable for 1999 and 2000 + ß14 Trend + ß15 Self Employment/Working Population (lagged 1 year) + ß16 Patents (lagged 1 year) + ß17 Time to Discharge in Bankruptcy + ß18 Capital Gains Tax + 
1
- Supply of Capital = 2 + ß21 MSCI Public Market Return (contemporary/lagged 1 year) + ß22 Real GDP Growth (lagged 1 year) + ß23 Dummy Variable for 1999 and 2000 + ß24 Trend + ß25 Capital Gains Tax + ß26 EVCA Tax and Legal Index + ß27 Government Programs + 2
All variables are defined in Table 1 Panel A. Several affect both the demand for and supply of venture capital, and therefore appear in both equations. Following Gompers and Lerner (1998), Black and Gilson (1998), and Jeng and Wells (2000), we associate stock market returns with both supply of, and demand for, venture capital. Stock returns also act as the equilibriating mechanism in the system of simultaneous equations. We use the MSCI index of national stock market returns. Because both venture capital data and the MSCI index are annualized, there is a risk that contemporary stock market returns might take into account returns that resulted subsequent to many of the venture capital transactions. To check for this, we construct alternate specifications using the MSCI index lagged by one year. We also include a dummy variable for the bubble period of 1999–2000. Similarly, there is a connection between real economic activity and the demand for venture finance, for which we use a real GDP growth variable. The GDP data are lagged one year to avoid timing and/or endogeneity problems. A trend term is included to de-trend the data so that spurious correlations are not picked up in the regressions from two or more positively trending time series (see, e.g., Johnson et al., 1984).
We use the EVCA index (discussed in Section 3.1), for which a lower value indicates a more favourable environment for venture capital and private equity funds, in the supply equation because it is specifically designed to reflect the attractiveness of different countries for supply-side activity—that is, setting up a venture capital fund. The index is a composite of both fiscal and legal factors. If the taxation elements dominate, then any observed correlation would simply reflect favourable fiscal, as opposed to legal, environments. In order to check for this possibility, we also include a variable based on capital gains tax rates in each country-year.
The association of other variables with supply or demand is intuitive. The supply equation also includes a variable for the amount of funds made available by government to publicly-funded venture capital investment vehicles. Owing to the incentives for a government to set up and give capital to a private equity fund, the potential for endogeneity in this variable for government funds is considered in the system of equations to check for robustness.
Following from the discussion of bankruptcy law in Section 3.3, we include in the demand-side specification a variable denoting the number of years until a discharge from pre-bankruptcy debt is available. Where no discharge is available, a measure of average life expectancy is substituted, capturing the fact that bankrupts in such jurisdictions face ‘social death’.10 Where discharge is available only at the discretion of the court, we substitute half the life expectancy, in order to discount for uncertainty about how the court may exercise its discretion. Where a change in bankruptcy laws occurred during our sample period, an average value is used, calculated on a monthly basis across the year in which the change came into force.
It is of course possible that any observed correlation between liberal bankruptcy laws and demand for venture capital finance may be the result of other variables that are correlated with both. For example, it may be that some countries simply have a culture that is more conducive to entrepreneurship, which in turn might lead their legislatures to enact more liberal bankruptcy laws and also generate greater demand for venture capital. To check for this possibility, we also include two control variables in the demand equation as proxies for national propensity for entrepreneurship: self-employment and patent applications. We further consider the possible endogeneity of the bankruptcy variable—namely, that policymakers may have changed their bankruptcy laws in response to a desire to increase the amount of venture capital activity in the country. Various specifications are considered to check fully for robustness.
5.2 Estimates of early and expansion stage venture capital investment relative to GDP
This subsection reports the estimates based on the empirical methods outlined immediately above in Section 5.1. Before presenting estimates that distinguish between demand and supply, however, we first present standard single equation estimates in Table 4. These preliminary results provide a comparison with results in other studies in the venture capital literature using a similar methodology (for example, Jeng and Wells, 2000; Leleux and Surlemount, 2003). The regressions in Table 4 use country-fixed-effects to account for unobserved heterogeneity across countries in the panel dataset. Owing to perfect collinearity, it is not possible to include time-invariant independent variables when a complete set of country dummy variables is used for the fixed effects regressions, so the EVCA index, time trend and bubble dummy variables are not included in Table 4. The dependent variable, early stage venture capital, is expressed relative to the GDP in the year in the country.
|
Five models reported in Table 4 to show robustness to the inclusion or exclusion of different right-hand-side variables. Models 1 to 4 treat the direct government programs and the time to discharge in bankruptcy as exogenous variables. Model 5 uses instruments for those two policy variables, to consider the possibility that changes in direct government spending and/or bankruptcy laws were at least partly a response to prior levels of venture capital activity. The instruments selected included self employment per capita of population, patents, and MSCI returns (each lagged by one year). These instruments are not ideal, since in theory they might also directly influence the level of venture capital activity, as we in fact describe above and test in the results. We nevertheless selected these instruments for two reasons. First, the data did not indicate a statistically significant relation between these variables and early stage venture capital investment activity (see, for example, models 2 to 4 in Table 4). Secondly, there were no obviously better instruments that might influence the right-hand-side policy variables but not the left-hand-side variable. Regardless, we did consider numerous alternative specifications, and the results discussed below were quite invariant to the specification.
The results in Table 4 indicate the following. First, as regards direct government investment programs, the data provide more support for Hypothesis 2b than Hypothesis 2a (see Section 3.2): government funds appear to crowd out private investment in venture capital. Support for Hypothesis 2a would require a positive and significant coefficient for government programs, but the coefficient is not statistically significant in any model.11 This implies that the investment of government funds does not increase the overall amount of venture capital investment in a particular country: rather, public funds crowd out private funds such that there is no overall change in the total amount invested. If the purpose of government programs is to expand the overall size of the pool of investment activity, this evidence suggests that government programs frustrate this objective rather than fulfil it.
Secondly, with respect to bankruptcy laws, the data indicate that more liberal personal bankruptcy laws stimulate venture capital investment, thus supporting Hypothesis 3 in Section 3.3. The coefficient is consistently negative and statistically significant (to the 1% level in models 1, 3 and 4, and to the 10% level in Model 5). This effect is also economically significant. For example, Germany introduced a fresh start in personal bankruptcy during the late 1990s. The results in Table 4 (using Model 5) indicate that early stage venture capital, per dollar of GDP, increased by 1.229E-03 as a result. Put differently, the changes in the bankruptcy legislation increased the size of the German venture capital market by 6.4%, based on the average size of its early stage venture capital market reported in Table 1 Panel B.
Thirdly, note that in Table 4, levels of capital gains taxes are negatively related to venture capital activity. Our panel estimates—such as Model 4 in Table 4—indicate, for example, that the reduction in capital gains taxes by 8% in the US in the 1990s gave rise to an increase in early stage venture capital investment, per dollar of GDP, of 7.75E-02, or 1.3% based on the average size of the US early stage venture capital market, reported in Table 1 Panel B. This finding is consistent with prior work (see Poterba, 1989a, 1989b; Gompers and Lerner, 1998; Jeng and Wells, 2000).
Finally, note that in Table 4, lagged GDP growth is the principal economic variable which is both robust and significant. A 5% increase in real GDP growth from the prior year causes an increase in early stage venture capital investment, per dollar of GDP, of 3.021E-04, or 1.0% based on the size of the average early stage venture capital market for the countries in the data, reported in Table 1 Panel B. The other economic control variables, including patents and self-employment, are not significant in Table 4, nor are MSCI returns lagged 1 year. Contrary to our expectations—although consistent with the correlations reported in Table 3—contemporary MSCI returns are negative and significant. This latter result is perhaps related to the fact that early stage venture capital investments can take from 3–7 years to come to fruition, so that exit requirements might be based on expectations of changes in market conditions.
We now turn to the system of equations in Table 5, as described above in Section 5.1. The specifications reported in Table 5 are for the value of early stage transactions only (models 6 and 7) and expansion stage transactions only (models 8 and 9). These systems each employ 11 country dummy variables. Country fixed effects cannot be used because of the inclusion of variables that are invariant across countries, including the EVCA index, the bubble dummy variable, and the time trend. As many country dummy variables as possible were used. The reported specifications exclude dummy variables for Canada, Denmark, Germany, and the US. Those countries were excluded due to higher correlations with those country dummy variables and the included policy variables (such as the EVCA index, government programs, and capital gains taxation); see Table 3.12 Importantly, we point out the fact that the results in Table 4 with the exclusion of country-invariant variables and country-fixed-effect dummy variables are consistent with the results in Table 5 where 11 of 15 country dummy variables can be used with some of the country-invariant variables. Finally, it should also be noted that in Table 5, models 6 and 8 treat government venture capital programs and capital gains taxes as exogenous and time to discharge in bankruptcy as endogenous; Models 7 and 9 treat all three of these variables as endogenous.13 Each dependent variable is expressed relative to the GDP in the year in the country.
|
Models 6 and 8 in Table 5 indicate strong support for Hypothesis 1 (Section 3.1) pertaining to the EVCA tax and legal index for early stage venture capital relative to GDP. The evidence is statistically significant at the 1% level in both models. Moreover, it is also economically significant. A one-point improvement (decrease) in the index (that is, the approximate difference between moving from Spain to the UK) gives rise to an increase in early stage venture capital, per dollar of GDP, of 8.856E-04 (based on Model 7, and similar for Model 6). This is about 3% of the size of the average early stage venture capital market, as reported in Table 1 Panel B. The EVCA index is similarly significant for all of the venture capital and private equity stages and for fundraising (Models 6–13 in Tables 5 and 6).
|
As expected, capital gains tax rates are negatively correlated with investment, statistically significant at the 1% level on both the supply and demand sides in Models 6, 8, and 9. However, note that the economic significance indicated for capital gains taxes is smaller in Table 5 than in Table 4. For example, the estimates in Table 5 (Model 6) indicate the reduction in capital gains taxes by 8% in the US in the 1990s gave rise to an increase in early stage venture capital activity, per dollar of GDP, of 2.02E-04, or 0.3% based on the average size of the US early stage venture capital market, reported in Table 1 Panel B. Much of this difference between Tables 4 and 5 is owing to the inclusion of the EVCA index in Table 5, which itself in part accounts for tax differences across countries, along with the other legal differences discussed above in Section 3.1.
Regarding government programs (considered in Section 3.2), the results again—consistently with Table 4—support Hypothesis 2b rather than Hypothesis 2a. That is, government funds appear to crowd out private investment. In Models 6, 8 and 9, the coefficient for government programs is statistically insignificant, implying that government investment has neither increased nor decreased the total amount of early stage investment. This indicates that the investment of public funds has crowded out an equal amount of private funds. In Model 7, the presence of significant government programs is, quite extraordinarily, associated with a reduced overall level of early stage investment, implying that for each dollar of government money invested, more than a dollar of private money that would otherwise be invested is held back. In terms of economic significance, the result in Model 7 implies that a 1% increase in government venture capital funds (relative to the size of the venture capital market in the country) reduces early stage investment per dollar of GDP by 1% of the size of the average early stage venture capital markets in the data reported in Table 1 Panel B. However, this reduction effect–implying more than 100% crowding out–in Model 7 is not robust to the specification: the other three specifications in Table 5 and the specifications in Table 4 only suggest 100% crowding out.14
Regarding Hypothesis 3 (discussed in Section 3.3) pertaining to bankruptcy legislation, the data indicate that more liberal bankruptcy laws—in terms of a shorter time to discharge—increase the demand for venture capital. The economic significance found in Table 5 is approximately one-half that reported above for Table 4. For instance, the changes to the bankruptcy laws in Germany discussed above increased the size of its early stage venture capital markets by 3.0% relative to the average size of German venture capital markets reported in Table 1 Panel B. It is also worth noting that the statistical significance of the bankruptcy effect is quite robust. Finally, the results are statistically significant—both for early and expansion stage investment—where the time to discharge variable is treated as exogenous (although, for conciseness, not explicitly reported).
Many of the control variables for economic effects are also significant in Table 5, as is the trend term. The results pertaining to the MSCI, real GDP growth, and the bubble effect depend on which of these variables are included or excluded. That is, when the bubble variable is included, the others are generally insignificant. Conversely, if the bubble variable is excluded, the other economic effects tend to be positive and significant. Either way, the particular specification in terms of economic control variables does not materially impact the legal variables indicated above. It is worth noting that the self-employment and patent variables are negative and significant in some of the specifications in Table 5, counter to our expectations. These later results appear to be driven by the Mediterranean countries in our data (including Greece, Italy, Spain), which appear to be outliers in these respects. That is, self-employment levels are particularly high in these countries, whilst venture capital activity is particularly low.15 Regardless, these results were not central to, and did not affect results related to, the primary three hypotheses of interest outlined in Section 3.
5.3 Estimates of total private equity investing and fundraising relative to GDP
In order to complement the analyses in Tables 4 and 5 above, we now assess the determinants of the total amount of private equity investment: including early and late stage venture capital, buyouts and turnaround investments (Models 10 and 11 in Table 6). We further assess the determinants of the flow of funds from institutional investors to venture capital funds (fundraising) (Models 12 and 13 in Table 6), as distinct from the flow of capital vis-à-vis entrepreneurial firms and venture capital funds (investing).
The evidence on fundraising in Table 6 is generally consistent with, and supportive of, the evidence on investing in Tables 4 and 5 pertaining to both the legal variables (Hypotheses 1 to 3) and the economic control variables. With the detrended data—that is, specifications that include a trend term to avoid spurious correlations between the left- and right-hand-side variables—the main economic factor driving fundraising across countries during 1990–2003 was the existence or otherwise of the investment bubble in 1999–2000. There is, again unexpectedly, some evidence of a negative relation between fundraising and patent activity and employment; however, this evidence is not robust to the particular specification. Similarly, there is unexpected evidence of a negative relation between self-employment and fundraising. This latter finding is probably attributable to a comparative dearth of pension fund capital—the largest contributor to venture capital funds in Europe and North America—amongst countries with higher levels of self-employment. It also appears to be driven by the outlier Mediterranean countries in our data (including Greece, Italy, Spain), as discussed in the prior subsection.
The legal and institutional variables have much more robust and stronger economic effects on venture fundraising than do the economic variables. Total private equity investing and total fundraising is much greater among countries and time periods with shorter times to discharge in bankruptcy. For instance, the introduction of a fresh start in bankruptcy law in Germany (as discussed above in Section 5.2) increased the size of total private equity investing (fundraising) by 2.6% (2.9%), based on the average size of the markets reported in Table 1 Panel B.
Similarly, a one-point decrease (improvement) in the EVCA index (that is, the approximate difference between moving from Spain to the UK) gives rise to an increase in total private equity investing (fundraising)/GDP by 1.73E-03 (2.03E-03), which is about 1.03% (0.98%) of the size of the average total private equity investing across all of the country years reported in Table 1 Panel B. Overall, therefore, the EVCA index appears to have a less pronounced impact on total private equity than on early stage investing (as reported above in Section 5.2).
In all specifications, government support programs do not increase the total amount of venture funds raised, which again implies crowding out—that is, the substitution for, or deterrence of, private investment by public investment. As indicated, a finding that government programs have their intended effect of stimulating venture capital would require the government variable to be positive and significant.
5.4 Estimates of dispositions (exits) relative to GDP
To complement the foregoing evidence, we provide estimates of the legal variables pertaining to Hypotheses 1 to 3 on the total value of exits (dispositions or sale transactions) per dollar of GDP in Table 7. These models employ OLS and IV panel regressions. A simultaneous equation specification would be inappropriate, because distinguishing between demand and supply for dispositions would require modelling of factors that affect the respective demands for IPOs, trade sales, and other forms of disposition, which are beyond the scope of the data. Country fixed effects are used in these specifications and so once again time-invariant variables are not included.
|
Once more, the results show strong support for Hypothesis 3, regarding time to discharge in bankruptcy, and Hypothesis 2b, regarding crowding out. Government investment in no case increases overall levels of venture capital exits. Unfavourable bankruptcy legislation further reduces industry dispositions. For instance, the changes to the bankruptcy laws in Germany above in Section 5.2 increased the annual value of dispositions per dollar of GDP in that venture capital market by 2.4%, again based on the average size of German dispositions reported in Table 1 Panel B. One explanation for this result is that marginal entrepreneurs with risky but potentially very valuable projects do not want to take the risk of bringing them to market because of unfavourable bankruptcy laws.
Note that the effect from bankruptcy in Table 7 is much stronger and more robust relative to the taxation and other institutional and economic control variables. The only other statistically robust effect on exits is real GDP growth. The estimates indicate that a 5% increase in real GDP growth causes an increase in the value of exits, per dollar of GDP, of 2.31E-04, or 0.3% based on the average value of dispositions across all markets reported in Table 1 Panel B. It is also worth noting that the variable for government venture capital programs is consistently insignificant, indicating such programs have not on average enhanced the returns to venture capital markets.
5.5 Limitations and future research
Whilst our indices for bankruptcy laws and public funds are true time series, the EVCA index of legal and fiscal measures was first constructed in 2003. Our preliminary consideration of a modified EVCA index with changes over time did not yield material changes to our econometric estimates. The main reason is that, as described above in Section 3.1, the EVCA index is a weighted average of 10 legal factors; therefore, legislative changes to a subset of the variables are relatively immaterial to the overall index value for a country year index value.16 By contrast, our public venture capital funds, tax data, and bankruptcy law index do change over time for each of the countries considered in our sample.
Our conclusions in this paper are of course confined to the countries considered. We did consider segregating the sample by dropping countries, but this did not materially affect our presented results. As datasets develop over a longer period from developing nations, it would also be worthwhile to assess the role of legal systems in facilitating the development of nascent venture capital markets.
There are of course limitations with the use of aggregate data; in particular, it does not capture the details of specific micro-level effects on transactions. Transaction-specific data could be used to explore certain issues in more detail, and such an analysis would usefully extend and complement the results presented in this paper. Yet to assess overall industry effects resulting from the legal environment, it is necessary to employ a comprehensive set of data, and these currently exist only on an aggregate basis.
|
6. Implications and conclusion |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
Based on aggregate industry venture capital and private equity data spanning the period 1990–2003 from Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands, Portugal, Spain, Sweden, the UK, and the US, we show that the legal environment is of paramount importance in measuring the supply of and demand for venture capital. Favourable fiscal and legal environments facilitate the establishment of venture capital and private equity funds and increase the supply of capital. Similarly, liberal bankruptcy laws stimulate entrepreneurialism and increase the demand for venture capital. Government programs, by contrast, crowd out private equity investment. These effects are both statistically and economically significant, and more pronounced than the effects from control variables such as stock market returns, real GDP growth, and patent activity—even including controls for the endogeneity of government programs. Moreover, we showed that these results are supported by alternative estimation methods and an analysis of fundraising across countries.
In the course of our complementary analyses, we demonstrate that industry dispositions per dollar of GDP are enhanced by favourable tax and legal environments for funds and by liberal bankruptcy laws for entrepreneurs. Government programs, by contrast, significantly reduce overall industry dispositions per dollar of GDP. These results are robust to the endogeneity of the establishment of and support for government programs.
These results have significant policy implications. The prevailing wisdom has been that deep and liquid stock markets are the most important determinant of venture capital investment. It was therefore thought that policymakers wishing to foster venture capital markets could only do so indirectly, by implementing legal measures that are conducive to the development of liquid stock markets, such as disclosure laws, minority shareholder protection, and anti-director rights (La Porta et al., 1997, 1998; Black, 2001). An alternative route would be for governments to supply capital themselves, through publicly-funded schemes that would seek to stimulate the growth of private equity markets, with the success of such schemes being highly contingent on the appropriate design of incentives (Gilson, 2003).
Our results cast doubt on both aspects of this wisdom. They imply first that a range of legal factors may affect venture capital investment directly; secondly that publicly-funded venture capital funds frequently fail to achieve their objective; and thirdly that a liberal personal bankruptcy law increases demand for venture capital finance, even controlling for other demand-side factors such as patent activity and levels of self-employment. Interestingly, but consistently with theoretical work by Keuschnigg (2003) and Keuschnigg and Nielsen (2003a, 2004), these results imply that legislators may successfully stimulate venture capital markets by reducing direct taxation, but not by providing investment subsidies. This is because increasing the supply of investment without a corresponding increase in projects to be funded results in greater competition and diminished returns, driving private investors from the market. Generally, the results indicate the road to establishing a Silicon Valley-like private equity market outside the US is paved with favourable tax laws and legal structures that accommodate the establishment of private equity funds, liberal bankruptcy laws that provide little or no time to discharge for entrepreneurs, and at most only a very small scope for direct government investment programs.
|
Acknowledgements |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
Armour gratefully acknowledges ESRC funding and Cumming gratefully acknowledges University of Alberta SAS funding. For helpful comments and discussions on earlier versions (some of which were under different titles), we thank Robert Ahdieh, Jöchen Bigus, Heather Gibson, Henry Hansmann, Yair Listokin, Kate Litvak, Joe McCahery, Richard Nolan, Ed Rock, Roberta Romano, Erik Vermeulen, Michael Wachter, Wolfgang Weigel, and Ralph Winter. We are also grateful to seminar participants at Athens University (EALE 19th), Tilburg University, the University of Pennsylvania Law School, the Cass Business School, the ECB/Bank of Greece Workshop on European Financial Integration, the Yale Law School LEO Workshop, Queen's University Belfast, the Babson Entrepreneurship Conference and ETH Zürich (CLEF 11th).
|
Notes |
|---|
1 Other studies suggest that differences in corporate and tax law may result in differences in transaction structure, but not necessarily affect overall investment levels (Gilson and Schizer, 2003
; Lerner and Schoar, 2005). 
2 Crucial to the success of private venture funds is the use of appropriate contractual technology. Contracts ensure that the VCs themselves are appropriately incentivized to select good investments and to keep up their monitoring efforts so as to ensure that the maximum return is achieved on them in due course (Sahlman, 1990; Gompers and Lerner, 1999; Gilson, 2003).
3 On the other hand, Leleux and Surlemount (2003) do not find crowding out in a study of investment data from 15 European countries in the early 1990s. Leleux and Surlemount's dataset, however, comprises the period 1990–96, and it is therefore worthwhile to explore this issue further over at least one full business cycle.
4 Or if venture capital is raised and the business subsequently fails: simply because venture capital is invested in the business does not mean the entrepreneur's personal finances will have been restored to their previous position.
5 In almost all jurisdictions, a debtor may emerge from bankruptcy by entering into a ‘composition’ with his creditors, whereby he agrees to repay a proportion of the face value of his debts and the rest is treated as discharged. The difference between this and the fresh start discussed in the text is, however, that a composition requires the agreement of a majority of the debtor's creditors. A fresh start regime on the other hand entitles the debtor to be discharged against the wishes of creditors.
6 Furthermore, many jurisdictions specifically prohibit undischarged bankrupts from trading or participating in the management of a company.
7 Namely: Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, The Netherlands, Portugal, Spain, Sweden, the UK, and the US.
8 Capital gains taxes are just one of many aspects of a tax system, and it is extremely difficult to identify a country-year with a single number. The capital gains tax rates are often graduated so that they depend on the level of capital gains, and the inclusion rates (the amounts and type of capital gains subject to tax) can vary. Each country typically has special exclusions for different industries, including high-tech industries in which VCs often invest. As such, our tax figures are at best proxies for the general state of the tax environment with regard to venture capital activity. Limited degrees of freedom prevent inclusion of additional tax variables in our estimates.
9 A number of robustness checks are provided in the tables; alternative specifications not reported are available on request.
10 The figure used is national life expectancy minus 30 years, to simulate the impact on a relatively young entrepreneur. The results are robust to alternative specifications.
11 The specification that we explicitly report (government VC/total VC) is that also used by Leleux and Surlemount (2003). As a robustness check, we considered the government VC variable as expressed in total amounts of government investment. Similarly to the results reported, the government variable was insignificant in all the models in Table 4, suggesting 100% crowding out. We also considered different specifications of the government VC variable in Tabels 5–7 and did not find material differences. It should be noted that it is inappropriate to subtract government expenditures from the total venture capital amounts because (1) the stage of development of government expenditures is not known in all countries, and (2) private and public VCs may syndicate with one another, and hence the subtraction may yield an incorrect measure of the total amount of venture capital even if the stages of government expenditures were known.
12 It is appropriate to exclude the variables that are the most highly correlated to avoid estimation bias. Inclusion of variables that were more highly correlated in some cases distorted our estimates due to statistical bias.
13 The instruments are analogous to those used in Model 5 in Table 4. As discussed immediately above, ideal instruments do not exist; however, the results reported are quite robust to alternative specifications.
14 As discussed above (see note 11), the inferences from using alternative specifications of the government VC variable are extremely similar.
15 One possible solution to this issue in the data would be to include interaction terms (dummy variables for these countries multiplied by these employment variables); however, limited degrees of freedom and data mining concerns suggested this would not be warranted.
16 An absence in time variation in a legal index is of course shared with all studies that employ La Porta et al. (1997, 1998) legality variables in relation to time series financial data.
|
References |
|---|
|
TOPAbstract1. Introduction2. Venture capital and its...3. Theories and formulation...4. Data and summary...5. Multivariate empirical...6. Implications and...AcknowledgementsReferences |
|---|
-
Allen F and Song W-L. (2003) Venture capital and corporate governance. In Cornelius PK and Kogut B (Eds.). Corporate Governance and Capital Flows in a Global Economy(Oxford University Press, New York) pp. 133–56.
Armour J. (2003) Law, finance and innovation. In McCahery JA and Renneboog L (Eds.). Venture Capital Contracting and the Valuation of Hi-Tech Firms(Oxford University Press, Oxford) pp. 133–61.
Armour J. (2004) Personal insolvency law and the demand for venture capital finance. European Business Organization Law Review 5:87–118.
Bascha A and Walz U. (2001) Convertible securities and optimal exit decisions in venture capital finance. Journal of Corporate Finance 7:285–306.[CrossRef]
Bascha A and Walz U. (2002) Financing practices in the German venture capital industry: an empirical assessment. Working Paper No. 2002/08Center for Financial Studies, University of Frankfurt.
Berger AN and Udell GF. (1998) The economics of small business finance: the roles of private equity and debt markets in the financial growth cycles. Journal of Banking and Finance 22:613–73.[CrossRef]
Black BS. (2001) The legal and institutional preconditions for strong securities markets. UCLA Law Review 48:781–858.
Black BS and Gilson RJ. (1998) Venture capital and the structure of capital markets: banks versus stock markets. Journal of Financial Economics 47:243–277.[CrossRef][Web of Science]
Carpenter RE and Petersen BC. (2002) Capital market imperfections, high-tech investment, and new equity financing. Economic Journal 112:F54–72.[CrossRef][Web of Science]
Cosh A and Wood E. (1998) Innovation: scale, objectives and constraints. In Cosh A and Hughes A (Eds.). Enterprise Britain: Growth, Innovation and Public Policy in the Small and Medium Sized Enterprise Sector 1994–1997(ESRC Centre for Business Research, Cambridge) pp. 38–48.
Cressy R. (2002) Funding gaps: a symposium. Economic Journal 112:F1–16.[CrossRef][Web of Science]
Cumming DJ. (2005) Capital structure in venture finance. Journal of Corporate Finance 11:550–85.[CrossRef]
Cumming DJ. (2006) The determinants of venture capital portfolio size: empirical evidence. Journal of Business 79: forthcoming.
Cumming DJ. (2007) Government policy towards entrepreneurial finance innovation investment funds. Journal of Business Venturing 22: forthcoming.
Cumming DJ, Fleming G, Schwienbacher A. (2006) Legality and venture capital exits. Journal of Corporate Finance 12:214–45.[CrossRef]
Cumming DJ and MacIntosh JG. (2003) A cross-country comparison of full and partial venture capital exits. Journal of Banking and Finance 27:511–48.[CrossRef]
Cumming DJ and MacIntosh JG. (2005) Crowding out private equity: Canadian evidence. Journal of Business Venturing 21: forthcoming.
Da Rin M, Nicodano G, Sembellin A. (2005) Public policy and the creation of active venture capital markets. Working Paper No. 430(European Central Bank, Frankfurt).
EVCA (European Venture Capital Association). (2006) Benchmarking European Tax and Legal Environments(EVCA, Zaventum).
Fan W and White MJ. (2003) Personal bankruptcy and the level of entrepreneurial activity. Journal of Law and Economics 46:543–68.[CrossRef]
Freear J and Wetzel WE Jr. (1990) Who bankrolls high-tech entrepreneurs? Journal of Business Venturing 5:77–89.
Georgakopoulos NL. (2002) Bankruptcy law for productivity. Wake Forest Law Review 37:51–95.
Georgellis Y and Wall HJ. (2002) Entrepreneurship and the policy environment. Working Paper No. 2002–019BFederal Reserve Bank of St. Louis.
Gilson RJ. (2003) Engineering a venture capital market: lessons from the American experience. Stanford Law Review 55:1067–103.
Gilson RJ and Schizer DM. (2003) Understanding venture capital structure: a tax explanation for convertible preferred stock. Harvard Law Review 116:874–916.
Gompers PA. (1998) Venture capital growing pains: should the market diet? Journal of Banking and Finance 22:1089–102.[CrossRef]
Gompers PA and Lerner J. (1996) The use of covenants: an empirical analysis of venture capital partnership agreements. Journal of Law and Economics 39:463–98.[CrossRef][Web of Science]
Gompers PA and Lerner J. (1998) What drives venture fundraising? Brookings Proceedings on Economic Activity—Microeconomics pp. 149–92.
Gompers PA and Lerner J. (1999) The Venture Capital Cycle(MIT Press, Cambridge, MA).
Gompers PA and Lerner J. (2001) The venture capital revolution. Journal of Economic Perspectives 15:145–68.
Gompers P, Lerner J, Desai M. (2003) The two faces of entrepreneurship: evidence from Eastern and Western Europe. Working Paper(Harvard Business School, Cambridge, MA).
Hallinan CG. (1986) The ‘fresh start’ policy in consumer bankruptcy: a historical inventory and an interpretative theory. University of Richmond Law Review 21:49–160.
Hege U, Palomino F, Schwienbacher A. (2003) Determinants of venture capital performance: Europe and the United States. Working Paper(HEC School of Management, Paris).
Jackson TH. (1985) The fresh-start policy in bankruptcy law. Harvard Law Review 98:1393–448.
Jeng LA and Wells PC. (2000) The determinants of venture capital funding: evidence across countries. Journal of Corporate Finance 6:241–89.
Johnson SR, Hassaon ZA, Green RD. (1984) Demand System Estimation(Iowa State University Press, Ames, IA).
Judge GG, Griffiths WE, Hill RC, Lütkepohl H, Lee T-C. (1988) Introduction to the Theory and Practice of Econometrics 2nd (John Wiley & Sons, Hoboken, NJ).
Kanniainen V and Keuschnigg C. (2003) The optimal portfolio of start-up firms in venture capital finance. Journal of Corporate Finance 9:521–34.[CrossRef][Web of Science]
Kanniainen V and Keuschnigg C. (2004) Start-up investment with scarce venture capital support. Journal of Banking and Finance 28:1935–59.[CrossRef]
Kaplan SN and Strömberg P. (2003) Financial contracting theory meets the real world: an empirical analysis of venture capital contracts. Review of Economic Studies 70:281–315.[CrossRef]
Kaplan SN, Martel F, Strömberg P. (2003) How do legal differences and learning affect financial contracts? NBER Working Paper w10097, Cambridge, MA.
Kenny M. (2000) Understanding Silicon Valley: The Anatomy of an Entrepreneurial Region(Stanford University Press, Stanford, CA).
Keuschnigg C. (2003) Optimal public policy for venture capital backed innovation. CEPR Working Paper No. 3850(Centre for Economic Policy Research, London).
Keuschnigg C. (2004) Taxation of a venture capitalist with a portfolio of firms. Oxford Economic Papers 56:285–306.
Keuschnigg C and Nielsen SB. (2001) Public policy for venture capital. International Tax and Public Finance 8:557–72.[CrossRef][Web of Science]
Keuschnigg C and Nielsen SB. (2003a) Tax policy, venture capital and entrepreneurship. Journal of Public Economics 87:175–203.[CrossRef]
Keuschnigg C and Nielsen SB. (2003b) Public policy for start-up entrepreneurship with venture capital and bank finance. Department of Economics Discussion Paper No. 2003–02 CESifo Working Paper Series No. 850 University of St. Gallen.
Keuschnigg C and Nielsen SB. (2004) Start-ups, venture capitalists and the capital gains tax. Journal of Public Economics 88:1011–42.[CrossRef]
Klausner M and Litvak K. (2001) What economists have taught us about venture capital contracting. In Whincop MJ (Ed.). Bridging the Entrepreneurial Finance Gap: Linking Governance with Regulatory Policy(Ashgate, Aldershot) pp. 54–74.
Kortum S and Lerner J. (2000) Assessing the contribution of venture capital to innovation. RAND Journal of Economics 31:647–92.
La Porta R, Lopes-de-Silanes F, Shleifer A, Vishny RW. (1997) Legal determinants of external finance. Journal of Finance 52:1131–50.[CrossRef][Web of Science]
La Porta R, Lopes-de-Silanes F, Shleifer A, Vishny RW. (1998) Law and finance. Journal of Political Economy 106:1113–55.[CrossRef][Web of Science]
Leleux B and Surlemont B. (2003) Public versus private venture capital: seeding or crowding out? A pan-European analysis. Journal of Business Venturing 18:81–104.[CrossRef]
Lerner J. (1999) The government as venture capitalist: the long-run effects of the SBIR Program. Journal of Business 72:285–318.[CrossRef][Web of Science]
Lerner J. (2002a) Boom and bust in the venture capital industry and the impact on innovation. Federal Reserve Bank of Atlanta Economic Review 2002:425–39.
Lerner J. (2002b) When bureaucrats meet entrepreneurs: the design of effective ‘public venture capital’ programmes. Economic Journal 112:F73–84.[CrossRef][Web of Science]
Lerner J, Shane H, Tsai A. (2003) Do equity financing cycles matter? Evidence from biotechnology alliances. Journal of Financial Economics 67:411–46.[CrossRef]
Lerner J and Schoar A. (2005) Does legal enforcement affect financial transactions? The contractual channel in private equity. Quarterly Journal of Economics 120:223–46.[CrossRef]
Manigart S, Sapienza H, Vermeir W. (1996) Venture capital governance and value-added in four countries. Journal of Business Venturing 11:439–69.[CrossRef][Web of Science]
Maginart S, De Waele K, Wright M, Robbie K, Desbrières P, Sapienza H, Beekman A. (2000) Venture capital, investment appraisal, and accounting information: a comparative study of the US, UK, France, Belgium and Holland. European Financial Management 6:380–404.
Maginart S, De Waele K, Wright M, Robbie K, Desbrières P, Sapienza H, Beekman A. (2002) The determinants of the required returns in venture capital investments: a five-country study. Journal of Business Venturing 17:291–312.[CrossRef][Web of Science]
Mayer C. (2001) Institutional investment and private equity in the UK. OFRC Working Paper No. 2001fe10Oxford Financial Research Centre, Oxford University.
Mayer C, Schoors K, Yafeh Y. (2005) Sources of funds and investment activities of venture capital funds: evidence from Germany, Israel, Japan and the UK. Journal of Corporate Finance 11:586–608.[CrossRef]
Poterba J. (1989a) Capital gains tax policy towards entrepreneurship. National Tax Journal 42:375–89.
Poterba J. (1989b) Venture capital and capital gains taxation. In Summers LH (Ed.). Tax Policy and the Economy(MIT Press, Cambridge, MA) 3: pp. 47–67.
Sahlman WA. (1990) The structure and governance of venture capital organizations. Journal of Financial Economics 27:473–521.[CrossRef][Web of Science]
Schwienbacher A. (2002) An empirical analysis of venture capital exits in Europe and the United States. EFA 2002 Berlin Meetings Discussion PaperUniversity of Amsterdam.
Tykvová T. (2000) Venture capital in Germany and its impact on innovation. Working PaperCentre for European Economic Research (ZEW), University of Mannheim.
Wright M, Hoskisson RE, Busenitz LW, Dial J. (2001) Finance and management buyouts: agency versus entrepreneurship perspectives. Venture Capital: An International Journal of Entrepreneurial Finance 3:239–62.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Armour and D. Cumming Bankruptcy Law and Entrepreneurship Am. Law Econ. Rev., September 1, 2008; 10(2): 303 - 350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Cumming Contracts and Exits in Venture Capital Finance Rev. Financ. Stud., September 1, 2008; 21(5): 1947 - 1982. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Cumming and J. G. MacIntosh Mutual funds that invest in private equity? An analysis of labour-sponsored investment funds Camb. J. Econ., May 1, 2007; 31(3): 445 - 487. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||