The entrepreneur's choice of a venture capital firm: Empirical evidence from two VC fund portfolios

The entrepreneur's choice of a venture capital firm: Empirical evidence from two VC fund portfolios

Accepted Manuscript The entrepreneur’s choice of a venture capital firm: empirical evidence from two VC fund portfolios Guillaume Andrieu , Raffaele ...

1MB Sizes 1 Downloads 6 Views

Accepted Manuscript

The entrepreneur’s choice of a venture capital firm: empirical evidence from two VC fund portfolios Guillaume Andrieu , Raffaele Stagliano` PII: DOI: Reference:

S1544-6123(16)30020-4 10.1016/j.frl.2016.03.008 FRL 478

To appear in:

Finance Research Letters

Received date: Revised date: Accepted date:

7 January 2016 22 January 2016 5 March 2016

Please cite this article as: Guillaume Andrieu , Raffaele Stagliano` , The entrepreneur’s choice of a venture capital firm: empirical evidence from two VC fund portfolios, Finance Research Letters (2016), doi: 10.1016/j.frl.2016.03.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Highlights We investigate the determinants of entrepreneurs’ choices of a venture capitalist. We use data from two types of French venture capital firms. The independent firm finances more ventures requiring high support quality. The bank-affiliated firm favors ventures with higher liquidation value.

AC

CE

PT

ED

M

AN US

CR IP T

   

1

ACCEPTED MANUSCRIPT

The entrepreneur’s choice of a venture capital firm: empirical evidence from two VC fund portfolios Guillaume Andrieua*, Raffaele Staglianòa a

Abstract

CR IP T

Montpellier Business School and Montpellier Research in Management, 2300 Avenue des Moulins, 34185 Montpellier Cedex 4, France.

AN US

We examine the determinants of entrepreneurs’ choices between an independent and a bankaffiliated venture capital firm operating in the same region of France. We find empirical support for a significant impact of the firm’s support quality and the venture’s liquidation

M

value.

ED

Key words: Venture Capital; Entrepreneurial Finance; Banking

CE

PT

JEL classification: G21; G24

AC

* Corresponding author at: Montpellier Business School, 2300 Avenue des Moulins, 34185 Montpellier Cedex 4, France, Tel.: +33 (0)4 67 10 28 64 (Fax: +33 (0)4 67 45 13 56). Email addresses: G. Andrieu ([email protected]), R. Staglianò ([email protected]).

2

ACCEPTED MANUSCRIPT

1. Introduction To raise funds for a startup, entrepreneurs can turn to several funding sources: “love money” from friends or family, banks, public subsidies, or more sophisticated sources like angel investors or venture capital (VC) firms.

CR IP T

In the last few decades, the VC industry has played a major role in building strong firms, accounting for notable wealth and job creation in developed economies. Venture capitalists are sophisticated investors with high expertise and broad knowledge on how to

AN US

develop innovative startups (Tykvova, 2007). Unsurprisingly, they are quite rigorous when they select among high-risk projects. Once they have chosen a project, they then provide support to the entrepreneur in defining strategies, networking, setting hiring policies, and much more. They invest in shares or convertible securities, providing high-powered

M

incentives to entrepreneurs (Kaplan and Strömberg, 2004, Cumming, 2006). Although the activities of venture capitalists are much the same, they may be funded by diverse types of

ED

investors having different aims and time expectations and they therefore adopt different types

PT

of organization (Andrieu, 2013). Entrepreneurs need the best type of VC firm to fit with their projects. Our aim in this study is to examine the empirical data from two different types of

CE

VC firm to identify the determinants of the VC firm-entrepreneur fit. The typical VC firm is independent (IVC) and organized as a fund under a limited

AC

partnership agreement (Sahlman, 1990). These funds have a limited lifetime, usually 10 years. They are managed by a general partner and financed by limited partners. In the US, the investors are generally pension funds and have a rather passive role. The general partner has an incentive to perform since he is compensated on the basis of the fund performance (carried interest) and risks exiting the industry if he finds no investors when the fund is closed. IVC firms are also found in France, even though pension funds are not their main investors.

3

ACCEPTED MANUSCRIPT Affiliated VC firms also play an important role. Bank-affiliated VC firms (BVC, henceforth) are a good example of the heterogeneity of the VC industry. It was first demonstrated some time ago that banks have specific skills for credit screening (Rajan, 1992). However, Hellmann et al. (2008) showed that in the VC market, BVC firms are less efficient that IVC firms. Bank-affiliated VC firms predominate in continental Europe. They are found

CR IP T

in the US, although their presence is much more modest. Even though BVC firms may not provide the same quality of support to entrepreneurs as independent firms, they benefit from substantial resources within the banking network. In addition, banks tend to enter the VC market as a way of creating synergies with other activities (e.g., ordinary lending). Black and

AN US

Gilson (1998) further pointed out that BVC firms are also a substitute for IVC firms in continental Europe. These economies are bank-centered economies, whereas the US economy is market-centered.

M

Andrieu and Groh (2012) provided a theoretical model to explain the determinants of entrepreneur’s choices of an IVC as opposed to a BVC firm. Their model predicts that IVC

ED

firms tend to finance more sophisticated, riskier projects requiring strong support and rely less often on syndication. Low physical distance between the venture and the VC firm and high

PT

investment might lead to higher support quality. On the other hand, BVC firms may finance

CE

projects that are less risky or more often syndicated. They should also liquidate ventures more often than IVC firms and thus ex ante favor projects with higher liquidation value. To our

AC

knowledge, no empirical paper has yet investigated their model’s predictions. The remainder of the paper is organized as follows. The following section describes the database that we used and methodology. The empirical results are presented in Section 3, followed by the concluding remarks.

4

ACCEPTED MANUSCRIPT 2. Data and methodology Our analysis is based on a unique dataset with detailed information on the portfolios of two French VC funds, the first being an IVC firm and the second a BVC firm. The data cover all the VC investments made by these two firms during the 1997-2001 period. This period provided us with a stable sample for a pre-crisis period, before the dramatic drop in VC

CR IP T

investments observed worldwide. The sample comprised 25 ventures supported by the BVC firm and 18 ventures supported by the IVC firm. The two firms handle the great majority of the VC investments in the South West region of France. The geographical homogeneity of the sample, with ventures financed in one of the main bank-centered economies of continental

AN US

Europe, seemed relevant to us in order to check the predictions of Andrieu and Groh (2012). Detailed data on the ventures and their portfolios were obtained directly from their internal memorandums. We specifically obtained data about the sectors, the distance between the

M

venture and the VC firm, and the deal structure: syndicated or not and amount invested per round.

ED

Accounting data were obtained from the Point Risk database, which contains detailed governance and accounting information on French firms. In our analysis, we used accounting

PT

data from the year before an investment was made to circumvent the problem of the direct

CE

impact that a VC firm’s behavior (e.g., value creation) would have on the accounting variables.

AC

Our main goal was to determine the factors that guided the entrepreneur’s choice between a BVC and an IVC firm. We first report univariate tests on our variables of interest and follow these results with our probit regression analysis of the entrepreneurs’ choices of a VC firms. All of our specifications were estimated following the general form:

(

)

(

)

( )

5

ACCEPTED MANUSCRIPT where Φ is the cumulative normal distribution. The BVC variable takes on two values. The value 0 denotes the IVC, and 1 denotes the BVC. Support quality is the vector of independent variables that measure a firm’s ability to support entrepreneurs’ activities. Liquidation is a proxy of the liquidation capacity of the VC. Risk is the vector of variable proxies of the risk level of the entrepreneurial activity. X is the vector of the control variables expected to

CR IP T

influence our dependent variable. The estimates represent the regression coefficients. A significant positive (negative) coefficient implies an increase (decrease) in the probability of BVC financing. We also include marginal effects. Last, we clustered the analyses over the

AC

CE

PT

ED

M

AN US

business sector.

6

CR IP T

ACCEPTED MANUSCRIPT

Table 1 Descriptive statistics by VC type. This table shows the mean and median (in brackets) of selected variables for the BVC and IVC firms from the sample. We report the pvalues for the t-test and the Wilcoxon rank-sum test. Description

Support quality VC majority owner Distance

1 if VC has majority of voting shares, 0 if not Physical distance (in km)

Total investment

Bank-affiliated VC

Independent VC

24%

55.5%

0.005

0.006

131.755 (93.5) 528.772 (381.959)

0.000

0.009

AN US

Variables

Cumulative investments of the VC funds (in thousand euros)

Liquidation Tangibility

The ratio of tangible, long-term assets (property, plant, and equipment) to total assets

Risk factors Syndication

1 if syndicate take place, 0 if not

Control variables No. funding rounds ROA

M

Debt-equity ratio

1 for biotechnology, health, chemistry and new materials and electronics/computer ventures, 0 if not The ratio of total debt to stockholders’ equity

ED

Sophisticated ventures

Numbers of funding rounds

Return on asset (net income divided by total assets) Ln (total assets)

AC

CE

PT

Size

211.656 (203.1) 298.749 (304.873)

t-test Pr(|T| > |t|)

Wilcoxon test Prob > |z|

0.001

0.041

0.392 (0.510)

0.319 (0.384 )

0.018

0.025

80%

66.6%

0.334

0.328

28%

28%

0.987

0.987

2.692 (2.102)

5.119 (3.248)

0.076

0.054

1.52 (1) 0.074 (0.051) 7.773 (7.896)

1.333 (1) 0.061 (0.035) 8.042 (8.218)

0.071

0.087

0.064

0.030

0.247

0.579

7

ACCEPTED MANUSCRIPT 3. Results Table 1 presents the complete descriptions of the variables and the median and mean values for all continuous variables of the sample. It also shows the percentages of cases for the dummy variables. We report t-statistics for the differences in the means and the nonparametric Wilcoxon rank-sum test for the differences in the medians. The mean values

CR IP T

suggest that the IVC firm became the majority shareholder in 55.5% of cases (24% for BVC). According to both the median and mean values, the IVC firm tended to be physically closer to the funded ventures. The mean (median) distance in km was 131.755 (93.5) for the IVC firm compared with 211.656 (203.1) for the BVC firm. The IVC firm tended to invest more per

AN US

venture than the BVC firm. The mean (median) total investment was 528.772 (381.959) for the IVC firm compared with 298.749 (304.873) for the BVC firm. The mean (median) tangibility value was 0.392 (0.510) for the ventures in the BVC portfolio, which was significantly higher than the value obtained for the ventures in the IVC portfolio. The mean

M

(median) debt-equity ratio for the ventures in the IVC portfolio was 5.119 (3.248), much

ED

higher than in the BVC portfolio. We observed no significant difference for syndication or the propensity to finance sophisticated projects. The number of funding rounds and the ROA

PT

variables were significantly higher for the BVC-financed ventures, whereas size was not

CE

significantly different.

Table 2 presents the correlations between all the variables used in the model. The possibility

AC

of multicollinearity among the independent variables was also tested using variance inflation factors (VIFs). The maximum VIF that resulted from any of the models was 1.74, which is far below the generally employed cut-off of 10 (or, more prudently, 5) for regression models. The results show that the absence of multicollinearity can be accepted. Table 3 presents the results for the probit models across four specifications. Column 1 shows the results from the headline specification, which identifies the VC entrepreneur choice as the effect of support quality variables. We can see that the VC majority owner, distance and total 8

ACCEPTED MANUSCRIPT investment variables were as expected and statistically significant, with marginal effects respectively of 20.8%, 0.2% and 10.4%. The results suggest that the IVC firm generally provided better support quality. Column 2 includes the risk factor variables, with only the debt-equity ratio showing a negative and significant coefficient, with a marginal effect of 0.3%, as expected. Column 3 includes tangibility and, as expected, the estimated coefficient

CR IP T

was positive and significant, with a marginal effect of 21.3%. Last, we re-estimated the models including all variables and column 4 shows the results were similar. These results suggest that IVC firms are more willing to finance riskier projects requiring strong support, whereas BVC portfolio consists primarily of safer projects that are less difficult to support

AC

CE

PT

ED

M

AN US

and offer higher liquidating values.

9

CR IP T

ACCEPTED MANUSCRIPT

1 1. BVC 1

2

3

1

2. VC 2 majority owner

-0.322

1

3. Distance 3

0.387

-0.044

1

4. Total 4 investment

-0.299

0.125

0.151

5. Tangibility 5

0.243

-0.192

-0.054

6. Syndications 6

0.151

-0.230

7. Sophisticated 7 ventures

0.025

-0.049

8. Debt-equity 8 ratio

- 0185

9. No. funding rounds

0.148

10. ROA

0.114

-0.106

AC

CE

VIF

4

5

6

7

9

10

11

1

1

-.0.030

0.007

0.075

1

0.099

0.188

0.063

0.246

1

M

0.223

ED

8

0.156

-0.047

0.076

0.217

0.096

-.0.080

1

0.072

0.219

-0.115

-0.136

0.245

-0.025

0.087

1

0.007

0.068

0.279

0.1663

0.141

0.210

0.020

0.001

1

0.046

0.210

0227

0.174

0.146

0.189

0.143

-0.012

0.169

1

1.74

1.22

1.58

1.48

1.62

1.26

1.21

1.29

1.32

1.40

PT

11. Size

AN US

Table 2 Correlation matrix and variance inflation factors. Pearson’s correlations between variables and variance inflation factors (VIFs) to test the absence of multicollinearity.

10

ACCEPTED MANUSCRIPT

Table 3 Factors influencing the entrepreneur choice in a BVC and an IVC firm. In this table, the dependent variable is a dummy equal to one if the observation concerns a BVC firm and equal to zero if it concerns an IVC firm. We measured all independent accounting variables in year –1. (4) Probit BVC

-0.849** (0.356) -0.208 0.008** (0.004) 0.002 -0.423*** (0.132) -0.104

-0.856** (0.363) -0.206 0.008* (0.004) 0.002 -0.406*** (0.154) -0.098

-0.608** (0.227) -0.141 0.001** (0.004) 0 .002 -0.528*** (0.181) -0.122

-0.527** (0.247) -0.118 0.009** (0.004) 0.002 -0.500** (0.199) -0.113

Distance

+

Total investment

-

Liquidation Tangibility

+

0.922** (0.459) 0.213

1.031* (0.565) 0.232

Risk factors Syndications

+

-

ED

Sophisticated ventures

AN US

-

-

PT

Debt-equity ratio

-0.037 (0.265) -0.009 0.250

0.075 (0.265) 0.048 0.215

(0.261) 0.060 -0.0143* (0.008) -0.00344

(0.267) 0.017 -0.019*** (0.007) -0.0042

0.752* (0.426) 0.184 2.917** (1.143) 0.714 -0.014 (0.078) -0.003 -2.173 (1.873)

0.751* (0.440) 0.181 2.609*** (0.812) 0.631 -0.025 (0.067) -0.006 -2.130 (1.847)

0.883** (0.400) 0.204 2.405*** (0.798) 0.557 0.001 (0.080) 0.000 -2.297 (1.795)

0.877** (0.417) 0.198 1.854** (0.785) 0.418 -0.004 (0.077) -0.000 -2.285 (1.800)

Observations

43

43

43

43

Pseudo-R2

0.410

0.416

0.432

0.442

CE

No. funding rounds

CR IP T

(3) Probit BVC

Exp. sign

Support quality VC majority owner

Control

(2) Probit BVC

M

VARIABLES

(1) Probit BVC

AC

ROA

Size

Constant

Note: z-values are given in italics, marginal effects in bold. Robust standard errors (over business sector) in parentheses *** p<0.01, ** p<0.05, * p<0.1

11

ACCEPTED MANUSCRIPT

4. Concluding remarks This paper contributes to entrepreneurial finance research by empirically confirming the predictions of Andrieu and Groh’s (2012) model of the behavior of IVC and BVC firms. The

CR IP T

estimated results show that support quality, liquidation propensity and risk factors are relevant factors in determining the best fit between entrepreneurs and the VC firm. Entrepreneurs make decisions depending on the venture’s sector, the liquidation value, and the importance of the investor’s support. Our findings should therefore help entrepreneurs who want to choose

AN US

the venture capitalist most in line with their project’s characteristics. Our results are also important for policymakers because they corroborate the hypothesis that IVC and BVC firms cannot substitute for each other, thus suggesting that bank-centered economies in which BVC firms predominate cannot replicate the US model. In addition to legal aspects and cultural

M

differences, the dominance of VC firms in continental Europe is also a key factor explaining

PT

may deepen this analysis.

ED

the relatively low creation of innovative and disruptive firms. New research in other countries

Acknowledgments

CE

We are grateful for the comments of Catherine Casamatta, Antoine Renucci, Philippe

AC

Desbrières and Ulrich Hege. Special thanks to Laurent Germain.

12

ACCEPTED MANUSCRIPT References Andrieu, G., Groh, A. P., 2012. Entrepreneurs' financing choice between independent and bank-affiliated venture capital firms. J. Corp. Financ. 18(5), 1143-1167. Andrieu, G., 2013. The impact of the affiliation of venture capital firms: A survey. J. Econ. Surv. 27(2), 234246. Black, B. S., Gilson, R. J., 1998. Venture capital and the structure of capital markets: banks versus stock markets. J. Financ. Econ. 47(3), 243-277.

CR IP T

Cumming, D., 2006. Adverse selection and capital structure: evidence from venture capital. Entrep. Theory Pract. 30(2), 155-183. Hellmann, T., Lindsey, L., Puri, M, 2008. Building relationships early: Banks in venture capital. Rev. Financial Stud. 21(2), 513-541. Kaplan, S. N., Strömberg, P. E., 2004. Characteristics, contracts, and actions: Evidence from venture capitalist analyses. J. Finance 59(5), 2177-2210.

AN US

Rajan, R. G., 1992. Insiders and outsiders: The choice between informed and arm's‐length debt. J. Finance 47(4), 1367-1400. Sahlman, W. A., 1990. The structure and governance of venture-capital organizations. J. Financ. Econ. 27(2), 473-521.

AC

CE

PT

ED

M

Tykvová, T., 2007. What do economists tell us about venture capital contracts? J.Econ. Surv. 21(1), 65-89.

13