Syndication of Venture Capital Investments

Rieder, Finn

Syndication of Venture Capital Investments

Theory and Practice in Germany

von Finn Rieder (Autor:in)

BWL - Investition und Finanzierung

Zusammenfassung

Inhaltsangabe:Abstract:
Cooperation among financial institutions is a persistent feature of the equity issuance process. Also, the syndication of venture capital investments is common practice among venture capitalists. Despite the importance of syndication, surprisingly little is known on the motives and structure of syndication. Further-more, there is hardly any empirical evidence for Europe or particularly Germany on the syndication behaviour of VC organisations and the factors influencing their overall propensity to co-invest.
On the one hand, the purpose of this paper is to compile and summarise all aspects of the existing theory on VC syndication. Therefore, the evidence from the few empirical studies that were so far carried out on this topic is discussed. Thus, the general theory on syndication, which is thought to be independent from peculiar VC industries, serves as the framework throughout this paper which is complemented by the results of the empirical analyses done so far by different authors. On the other hand, this text has also an explorative component where it is analysed if and to what degree findings of the existing studies also hold for the German VC market. This analysis is done by the use of a data base listing a total of 3,230 VC investments in German portfolio companies. The nature of the data base at hand is not designed to allow for conclusions on all relevant aspects of syndication. But the claim of this paper is to use and analyse it whenever it is possible and the data base can be linked to aspects on syndication discussed throughout the different chapters.
Firstly, this paper gives a theoretical overview on general motives behind VC syndication which leads to a discussion which rational is more important in explaining VC syndicates and if there are differences between the North American and European VC market with regards to this. Then, as the principal part of this paper, various factors influencing the propensity of VC investors to syndicate are analysed and, whenever possible and appropriate, also reviewed based on the data base. The next part examines the process of forming and managing a syndicate, after the decision to co-invest a deal was made built upon the motives and factors of influence, and to what extent networking activities play a role in it. After this, it is discussed if syndicates enhance the firm value
If the portfolio company and whether there is a difference between the value added of syndicates […]

Leseprobe

Inhaltsverzeichnis

ID 7149

Rieder, Finn: Syndication of Venture Capital Investments - Theory and Practice in

Germany

Hamburg: Diplomica GmbH, 2003

Zugl.: European Business School Schloß Reichartshausen, Oestrich-Winkel,

Internationale Wirtschaftshochschule, Diplomarbeit, 2003

Dieses Werk ist urheberrechtlich geschützt. Die dadurch begründeten Rechte,

insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von

Abbildungen und Tabellen, der Funksendung, der Mikroverfilmung oder der

Vervielfältigung auf anderen Wegen und der Speicherung in Datenverarbeitungsanlagen,

bleiben, auch bei nur auszugsweiser Verwertung, vorbehalten. Eine Vervielfältigung

dieses Werkes oder von Teilen dieses Werkes ist auch im Einzelfall nur in den Grenzen

der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundesrepublik

Deutschland in der jeweils geltenden Fassung zulässig. Sie ist grundsätzlich

vergütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des

Urheberrechtes.

Die Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in

diesem Werk berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme,

dass solche Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als frei

zu betrachten wären und daher von jedermann benutzt werden dürften.

Die Informationen in diesem Werk wurden mit Sorgfalt erarbeitet. Dennoch können

Fehler nicht vollständig ausgeschlossen werden, und die Diplomarbeiten Agentur, die

Autoren oder Übersetzer übernehmen keine juristische Verantwortung oder irgendeine

Haftung für evtl. verbliebene fehlerhafte Angaben und deren Folgen.

Diplomica GmbH

http://www.diplom.de, Hamburg 2003

Printed in Germany

-II-

Table of Contents

List of Addenda ...V

List of Tables... VI

List of Figures...X

List of Abbreviations ... XI

1 Introduction... 1

1.1. Problem set and course of investigation... 1

1.2. Definition of venture capital ... 2

1.3. Definition of venture capital syndication... 2

1.4. Syndication practice of venture capitalists in Europe ... 3

2 Exploration of the data set... 5

2.1. Methodology ... 5

2.2. Description of the sample... 6

2.2.1.

Number of investors... 6

2.2.2.

Syndication Ratio of the investors ... 8

2.2.3.

Number of portfolio companies ... 9

2.2.4.

Geographical characteristics of the sample ... 10

2.2.4.1.

Distribution of the venture capital investors in Germany ...10

2.2.4.2.

Distribution of the portfolio companies in Germany ...11

3 Rationales for venture capital syndication... 13

3.1. The finance-based motive ... 13

3.2. The resource-based motive ... 14

3.2.1.

The Superior Investment Selection Hypothesis ... 15

3.2.2.

The Value Added Hypothesis ... 16

3.2.3.

Evaluation of the Value Added Hypothesis and the Selection Hypothesis .. 17

3.3. The deal flow motive ... 18

3.4. Other motives ... 19

3.4.1.

Window dressing... 19

3.4.2.

Collusion ... 20

3.5. Empirical relevance of the major motives... 20

-III-

4 Factors influencing the propensity to syndicate... 24

4.1. Degree of uncertainty... 25

4.2. Industry of the investee company... 26

4.3. Type of the venture capital investor ... 30

4.4. Characteristics of the venture capital organisation... 32

4.4.1.

Size ... 32

4.4.2.

Maturity... 35

4.4.3.

Specialisation... 36

4.4.4.

Investment amount ... 37

4.4.5.

Workload of the investment manager ... 37

4.5. Location of the venture capital firm and the portfolio company... 37

4.6. Investment stage ... 40

4.7. Competition in the venture capital market ... 41

5 Syndicate formation and networking of venture capital organisations ... 42

5.1. Partner selection in syndicates ... 42

5.2. The structure of venture capital syndicates... 44

5.2.1.

The role of the lead investor... 45

5.3. Syndication under a networking perspective ... 47

5.3.1.

Networking activities of venture capitalists... 47

5.3.2.

Implications on venture capitalists ... 48

5.3.3.

Implications on entrepreneurs ... 48

6 Syndication and firm value of the portfolio company ... 49

6.1. Syndicates of independent venture capitalists ... 49

6.2. Syndicates of corporate venture capitalists ... 51

7 Potential problems with syndication ... 54

7.1. Informational asymmetry ... 55

7.2. Coordination problems ... 55

8 Conclusion... 57

8.1. Prospects... 57

8.2. Avenues for further research... 58

8.3. Closing remarks... 59

-IV-

Appendix ... 61

Addenda ... 61

Tables

... 120

Figures

... 216

References... 222

Endnotes... 227

-V-

List of Addenda

Addenda 1: Variables listed in the primary Excel file ...61

Addenda 2: Data processing and consolidation of the original Excel file ...63

Addenda 3: Consolidation of the industries of the portfolio companies...71

Addenda 4: Data aggregation of the venture capital investors and portfolio companies...80

Addenda 5: Creation of the Syndication Ratio of the venture capital investors ...86

Addenda 6: Frequency distribution of the variable "IC Industry" ...91

Addenda 7: Aggregation of the Syndication Index and the variable "IC Industry" ...91

Addenda 8: Handling of Chi-squared tests with under-represented cell counts ...97

Addenda 9: Description and geographical distribution of the types of venture capital

investors ...102

Addenda 10: Aggregation of the variable "VC Type" and the Syndication Ratio ...104

Addenda 11: Aggregation of the variable "VC Size Categorised" and the Syndication

Ratio ...106

Addenda 12: Calculation of the co- investment ratios of the most frequented clusters...107

Addenda 13: Calculation of the Syndication Index of the foundation types of the

portfolio companies...113

Addenda 14: Aggregation of the variable "IC Failure" and the Syndication Index ...115

Addenda 15: Frequency Distribution of the variable "IC Failure" ...118

-VI-

List of Tables

Table 1: Total investment amount and number of venture capital investments syndicated in

Europe in 2001 ...120

Table 2: Syndication rates of German venture capitalists 2001 (all stages) ...120

Table 3: Syndication rates of German venture capitalists 2001 (early stage) ...121

Table 4: Syndication rates of German venture capitalists 2001 (later stage)...121

Table 5: Average number of investments per investor

(total sample)...121

Table 6: Average number of investments per investor (single-time investors excluded) ...122

Table 7: Frequency distribution of the number of investments per investor ...122

Table 8: Names of the in vestors with at least five investments ...123

Table 9: Frequency distribution of the categorised number of investments per investor ...128

Table 10: Statistics on the particular size classes

of the investors

...128

Table 11: Frequency distribution of the particular size classes of the investors...130

Table 12: Frequency distribution of the Syndication Ratio of the investors...131

Table 13: Mean of the Syndication Ratio (total sample) ...133

Table 14: Mean of the Syndication Ratio (single-time investors excluded)...133

Table 15: Average number of investors per portfolio company ...134

Table 16: Frequency distribution of the number of investors per company ...134

Table 17: Name of the portfolio companies with at least three investors...135

Table 18: Average Number of investors per portfolio company (companies with less

than two investors excluded) ...143

Table 19: Portfolio companies categorised according to their number of investors ...143

Table 20: Statistics on the particular syndication classes of the portfolio companies ...143

Table 21: Frequency distribution of the particular syndication classes of the portfolio

companies ...145

Table 22: Frequency distribution of the cities of the investor s ...145

Table 23: Frequency distribution of the clusters of the investors ...148

Table 24: Frequency distribution of the states of the in vestors...150

Table 25: Frequency distribution of the cities of the portfolio companies (Cities with

less than two portfolio companies excluded) ...151

-VII-

Table 26: Frequency distribution of the clusters of the portfolio companies ...156

Table 27: Frequency distribution of the states of the portfolio companies ...158

Table 28: Frequency distribution of the distance between the investor and its portfolio

company ...160

Table 29: Frequency distribution of the industries of the portfolio companies ...160

Table 30: Syndication Index of the factor "IC Industry" (all portfolio companies) ...161

Table 31: Significance of the factor "IC Industry" on the Syndiction Index (all portfolio

companies)...162

Table 32: Comparison of the categories of the factor "IC Industry" (all portfolio

companies)...162

Table 33: Syndication Index of the factor "IC Industry" (only co-invested portfolio

companies)...170

Table 34: Significance of the factor "IC Industry" on the Syndiction Index (only co-

invested portfolio companies) ...170

Table 35: Comparison of the categories of the factor "IC Industry" (only co-invested

portfolio companies)...171

Table 36: Chi-squared test for the variables "VC Type" and "IC Industry" (all company

industries and investor types) ...178

Table 37: Chi-squared test for the variables "VC Type" and "IC Industry" (company

industries and investor types with a too low cell count excluded) ...178

Table 38: Contingency table for the variables "VC Type" and "IC Industry" (portfolio

company industries and investor types with a too low cell count excluded) ...179

Table 39: Frequency distribution of the investor types...180

Table 40: Chi-squared test for the variables "VC Type" and " VC Cluster"...181

Table 41: Contingency table for the variables "VC Type" and " VC Cluster"...182

Table 42: Syndication Ratio of the investor types ...183

Table 43: Significance of the factor "VC Type" on the Syndiction Ratio ...183

Table 44: Comparison of the categories of the factor "VC Type" (on the Syndication

Ratio) ...184

Table 45: Significance of the factor "VC Type" on the Syndication Index...190

Table 46: Average number of co-investors of the investor types ...191

-VIII-

Table 47: Comparison of the categories of the factor "VC Type" (on the Syndication

Index)...191

Table 48: Syndication Ratio of the investors according to their size...198

Table 49: Significance of the factor "VC Size" on the Syndiction Ratio ...198

Table 50: Comparison of the categories of the factor "VC Size" (on the Syndication

Ratio) ...199

Table 51: Frequency distribution of the single-time investors according to their type...201

Table 52: Average number of co-investors of the investors according to their size ...202

Table 53: Significance of the factor "VC Size" on the Syndiction Index...202

Table 54: Comparison of the categories of the factor "VC Size" (on the Syndication

Index)...203

Table 55: t-test on the Syndication Ratio of Young- and Established Specialist...205

Table 56: t-test on the Syndication Ratio of Specialised- and Non-Specialised investors .205

Table 57: Significance of the factor "VC Cluster" on the Syndication Ratio ...206

Table 58: Significance of the factor "Distance" on the Syndication Index...207

Table 59: Average number of investors of the portfolio companies according to their

mean distance to their investors ...207

Table 60: Comparison of the characteristics of the factor "Distance"...208

Table 61: Syndication Index of the portfolio companies according to their foundation

type ...209

Table 62: Significance of the factor "IC Foundation Type" on the Syndiction Index...209

Table 63: Comparison of the categories of the factor "IC Type"...210

Table 64: Chi-squared test for the variables "VC Type" and " VC Lead yes/no"...211

Table 65: Contingency table for the variables "VC Type" and " VC Lead yes/no"...211

Table 66: Significance of the factor "IC Failure" on the Syndication Index ...212

Table 67: Comparison of the categories of the factor "IC Failure"...212

Table 68: Syndication Index of the portfolio companies according to their failure-

characteristics ...213

Table 69: Chi-squared test for the variables "VC Type CVC yes/no" and "IC Failure"...213

Table 70: Contingency table for the variables "VC Type CVC yes/no" and "IC Failure".214

Table 71: Chi-squared test for the variables "VC Type" and "IC Failure"...214

-IX-

Table 72: Contingency table for the variables "VC Type" and "IC Failure" ...215

-X-

List of Figures

Figure 1: Number of investments distributed according to the name of the investors...216

Figure 2: Frequency distribution of the number of investments per investor ...216

Figure 3: Number of investments per investor according to their size class...217

Figure 4: Frequency distribution of the portfolio companies according to their total

number of investors ...217

Figure 5: Histogram of the number of investors per portfolio company ...218

Figure 6: Frequency distribution of the size classes of the portfolio companies ...218

Figure 7: Frequency distribution of the clusters of the investors...219

Figure 8: Frequency distribution of the clusters of the portfolio companies ...219

Figure 9: Frequency distribution of the distance between the investor and its portfolio

company...220

Figure 10: Frequency distribution of the industries of the portfolio companies ...220

Figure 11: Frequency distribution of the investor types ...221

-XI-

List of Abbreviations

Approx.

Approximately

BVK

Bundesverband Deutscher Kapitalbeteiligungs-

gesellschaften

CVC

Corporate venture capital

B2B

Business to Business

B2C

Business to Consumer

BVCA

British Venture Capital Association

e.g.

For example

EVCA

European Venture Capital Association

GF500ICTs

Global Fortune 500 InfoCom Corporations

HIVC

High Innovative Technology Companies

Investee company

IPO

Initial Public Offering

LIVC

Low Innovative Technology Companies

Resp.

Respectively

Venture Capital

-1-

Introduction

1.1.

Problem set and course of investigation

Cooperation among financial institutions is a persistent feature of the equity issu-

ance process. Also, the syndication of venture capital investments is common practice

among venture capitalists. Despite the importance of syndication, surprisingly little is

known on the motives and structure of syndication (Manigart et al. (2002)). Further-

more, there is hardly any empirical evidence for Europe or particularly Germany on the

syndication behavior of VC organisations and the factors influencing their overall pro-

pensity to co-invest.

On the one hand, the purpose of this paper is to compile and summarise all aspects

of the existing theory on VC syndication. Therefore, the evidence from the few empiri-

cal studies that were so far carried out on this topic is discussed. Thus, the general the-

ory on syndication, which is thought to be independent from peculiar VC industries,

serves as the framework throughout this paper which is complemented by the results of

the empirical analyses done so far by different authors. On the other hand, this text has

also an explorative component where it is analysed if and to what degree findings of the

existing studies also hold for the German VC market. This analysis is done by the use of

a data base listing a total of 3,230 VC investments in German portfolio companies. The

nature of the data base at hand is not designed to allow for conclusions on all relevant

aspects of syndication. But the claim of this paper is to use and analyse it whenever it is

possible and the data base can be linked to aspects on syndication discussed throughout

the different chapters.

Firstly, this paper gives a theoretical overview on general motives behind VC syn-

dication which leads to a discussion which rational is more important in explaining VC

syndicates and if there are differences between the North American and European VC

market with regards to this. Then, as the principal part of this paper, various factors in-

fluencing the propensity of VC investors to syndicate are analysed and whenever pos-

sible and appropriate also reviewed based on the data base. The next part examines the

process of forming and managing a syndicate, after the decision to co-invest a deal was

made built upon the motives and factors of influence, and to what extent networking

activities play a role in it. After this, it is discussed if syndicates enhance the firm value

-2-

of the portfolio company and whether there is a difference between the value added of

syndicates formed by independent VC organisations and corporate venture capital

(CVC) firms. Then, the major potential drawbacks of syndicates are reviewed to the ex-

tent they were researched so far. Finally, the paper concludes with a short prospect on

the development of the importance of VC syndication and names some avenues for fur-

ther research.

1.2.

Definition of venture capital

The VC industry, unlike most other parts of the financial sector, lacks a precise le-

gal or regulatory definition. Looking at the broad variety of common definitions of VC,

they usually focus on four characteristics. The first characteristic sees VC as a source of

financing for privately held companies. Secondly, this sort of financing happens usually

in form of either equity or long term convertible debt. Thirdly, the venture capitalist is

seen as an intermediary between investor and entrepreneur (Schilit (1991)). Lastly, the

combination of equity investments and active involvement in the development of the

company is a unique characteristic of the VC industry (Kunze (1990)).

The principle of venture capital is to provide high potential growth companies with

the required funds and market expertise they need to make their business model a suc-

cess. Venture capitalists strive for substantia l capital gains and returns in the medium or

sometimes long term, compensating them for the high uncertainty and risk they accept

with their investments (Sahlmann (1990)). The ability to select investment opportunities

from a wide range of expected returns is vital to any venture capital organisation. Dif-

ferent to other institutional investors, venture capitalists face an informational disadva n-

tage as they do not invest in public quoted companies (Fama (1991)).

With regard to deal selection and monitoring, venture capital firms have developed

different strategies to reduce uncertainty in their high risk environment. One of these

strategies is to syndicate investments with other venture capitalists (Manigart et al.

(2002), Wright and Robbie (1998)).

1.3.

Definition of venture capital syndication

Syndication is one form of collaboration among venture capitalists. As Lerner

(1994) points out, cooperation among financial institutions is an enduring feature of the

-3-

equity issuance process. The first syndicated underwriting to be recorded dates back to

1870 and was an offering by Pennsylvania Railroad (Lerner (1994)).

An equity syndicate involves several venture capitalists taking an equity stake in an

investment (Lockett and Wright (1991)). It involves "[...] a group of individuals who

must make a common decision under uncertainty that will result in a payoff to be shared

jointly among them" (Wilson (1968), p. 119).

Thereby it is possible to make a distinction between a narrow and broad definition

of VC syndication. The narrow definition takes only venture capitalists that make the

investment decision simultaneously in the first financing round into consideration,

whereas the broader definition allows for investments to be syndicated, none important

whether the investment decision was made at the same or different time (Brander

(2002)). The analysis of the data set throughout this paper is always based on the

broader definition of syndication considering an investment as syndicated when the

portfolio company is funded by mo re than one VC investor.

1.4.

Syndication practice of venture capitalists in Europe

The concept of syndicating investment deals is very common among US- and

European venture capitalists. In August 2002, Tornado Insider, a European trade journal

focusing on start-ups, identified 100 promising European based technology companies.

The study showed that more than 90% of these start-ups were given venture capital,

with at least 85% of them being funded by a syndicate of venture capital firms (Tornado

Insider (2002)).

According to the European Venture Capital Association (EVCA) almost 30% of the

amount invested by European venture capitalists and number of VC deals were synd i-

cated in 2001 (see table 1). This appears relatively low compared to the fact that over

60% of VC investments in the US were syndicated in 2000. However, a 30% share

equaling a total amount syndicated of 7 billion still emphasizes the importance synd i-

cation plays in the European VC market and that it is well worthwhile to analyse the

backgrounds of syndication practice (EVCA (2002)).

As Chiplin et al. (1997) point out, the UK has the most established and mature VC

industry in Europe and syndication has been common practice in its VC market for

many years. Thus, with a total investment volume of almost 7 billion, the UK is the

-4-

biggest VC market in Europe and ranks world-wide second behind the US. Germany

(4.4 billion investment volume) and France (3.3 billion) are the next biggest Euro-

pean VC markets (see table 1).

Although being the largest European VC market, Germany (1.5 billion) and

France (1.4 billion) are ahead of the UK (1.3 billion) in terms of investment amount

syndicated by venture capitalists. Additionally, the UK shows the lowest number of in-

vestments syndicated (13.7%) and syndicated invested amount (19.3%) in Europe.

This

disagrees with Ooghe et al. (1991) who find that increasing maturity of markets is often

associated with increased syndication. (see table 1)

According to the EVCA (2002), among the six countries Belgium, France, Ger-

many, The Netherlands, Sweden and the UK, syndication in terms of amounts invested

is most common in Belgium: In 2001, 53.9% of the total VC investment amount was

invested through syndicates (while only 24% of all VC deals were syndicated). Contrary

to this, in Sweden only 9% of the total investment amount is syndicated, whereas the

number of deals syndicated (40%) is significantly higher (see table 1).

For the German VC industry, the "Bundesverband Deutscher Kapitalbeteiligungs-

gesellschaften" (BVK (2002)) reports that 32.9% of the total investment amount was

syndicated in 2001. This represents an amount syndicated of 1.5 billion in Germany.

Moreover, 37.5% of all German VC investments were syndicated (see table 1). Of the

total amount syndicated by German venture capitalists 66.9% were invested on a na-

tional, 28.7% on a European and only 4.4% on a trans-European level (see table 2).

Nearly 65% of the total amount syndicated by the German VC industry was spent

on the funding of late stage ventures and 35% on early stage ventures respectively

(EVCA (2002)). There is a distinct difference between the pattern of syndication of

early- and late stage investments: In Germany 92.1% of the amount syndicated in early

stage ventures is invested by German venture capitalists (see table 3). The amount syn-

dicated for late stage ventures, however, is partly financed by German syndicates

(52.1%), European syndicates (43.5%) and even trans-European syndicates (4.4%) (see

table 4).

-5-

Exploration of the data set

2.1.

Methodology

The empirical analys es in this paper are carried out on a data base from the Chair of

Entrepreneurship lead by Prof. Dr. Klandt at the European Business School in Oestrich-

Winkel, Germany. In an Excel file, this data base lists a total of 3,272 valid transactions

where a VC investor (German or non-German) funds a German investee company.

Originally, the database was built upon approx. 1,500 VC transactions from the

year 1998 on that were researched for the project "e-startup.org"

. In early 2002, Prof.

Dr. Klandt's chair adopted these 1,500 transactions and extended them through continu-

ous research to a total of 3,272 until December 2002. The information on these VC

transactions in Germany are exclusively based on secondary sources, such as the inter-

net, business reports of VC firms, print media, Genios-, MARKUS- and other data

bases.

For the purpose of this analysis, 42 transactions were excluded in which a one-time

investor (all companies from various industries which did a single occasional VC in-

vestment) took an equity stake in a company which has no investors else wise. These

investments were of no interest in terms of VC syndication as they involved a non-VC

orientated single-time investing company and a portfolio company that was solely

funded by this one investor and not through a syndicate.

The 3,230 remaining transactions were originally described by 86 variables, giving

further information on different aspects of each transaction (see Addenda 1). Most of

these variables in the original Excel file were either of little interest with regards to the

study on syndication or could not be used for proper statistical analysis due to data qual-

ity reasons. Thus, the analyses in this paper are mostly based on twelve of the original

variables which will be described later in the text and also serve to derive a whole set of

new variables from them.

The sample was described and analysed by importing the original Excel file into the

statistic software program SPSS. Due to the heterogeneous nature of the secondary

sources and the way the information on the different variables of each transaction were

entered in the original Excel data base, it was necessary to edit, modify and complete

the whole data set in SPSS before it was used for statistical purposes. The major meas-

-6-

ures and steps undertaken to improve the overall quality of the sample are described in

addenda 2 and 3

The data base with its entries on all the transactions as well as the SPSS-edition at

hand are in German. The names of the variables as well as the newly created variables

with their entries are translated into English, but it would have been to elaborate to

change all original data entries of the sample and the automated SPSS outputs to Eng-

lish.

2.2.

Description of the sample

The sample consists of a total number of 3,230 transactions where a VC investor

takes in some way an equity stake in a portfolio company. For each of these 3,230

transactions, there exist no missing values on the names of the VC investor and investee

company, thus the bottom line information of each transaction is that the participating

investor and portfolio company are known. The 3,230 transactions are distributed on

812 different VC investors and 1,962 portfolio companies. 1,377 transactions (42.6%)

are investments in companies that have only one VC provider and the other 1,853 trans-

actions (57.4%) involve co-invested companies with more than one investor. In the fol-

lowing, the distribution of the VC investors and portfolio companies is described (see

addenda 4).

2.2.1.

Number of investors

All transactions in the sample are completed by N=812 different VC investors. The

number of investments per single VC provider ranges from 1-120 recorded transactions

(range=119). As can be seen from table 5, the 812 investors of the sample have on aver-

age an equity stake in four different portfolio companies, whereat VC investors with

only one investment form the most represented group (mode=1). If only considered

those investors with two or more investments (N=301), the mean number of investments

per VC provider significantly increases to nine (see table 6), with two being the most

frequent value (mode).

A high standard deviation of almost 9.36, thus nearly 2.5 times the arithmetic mean,

indicates a strong dispersion of the distribution of the 812 investors. As demonstrated

by the extreme sinistral asymmetry ("Linksschiefe") of seven, the sample does not show

a symmetric distribution around the mean of number of transactions. The asymmetric

-7-

distribution form becomes apparent in figure 1. Obviously, the graph is rather complex

due to the high number of different VC firm names but reflects well the trend of the ac-

cumulation of FC investors in Germany: Firstly, some very few VC firms are involved

in remarkably many investments. Secondly, the number of different VC investors dis-

tinctly grows with decreasing number of investments. Thus, the sample of 812 VC pro-

viders consists of very few large investors (with tbg being the biggest one with 120

transactions) and a strong majority of small ones (see figure 2).

511 investors (62.9%) were only invested in a single company. Moreover, the num-

ber of VC providers with less than ten investments accounts for over 90% of the total

sample (see table 7).

Table 8 lists all investors with more than five investments with their particular

number of investments. This table points up that the 31 VC firms with 20 or more in-

vestments only make up 3.8% of the total number of VC investors, but made almost

40% of all 3,230 transactions of the sample. The four largest VC firms (0.5% of all in-

vestors) tbg Technologie-Beteiligungsgesellschaft (N=120 investments), TFG Venture

(N=100), 3i Deutschland (N=94), Knorr Capital Partners (N=86) combine 400 transac-

tions, 12.4% of the whole sample. In contrast to this, the 511 investors with only a sin-

gle transaction account for 62.9% of all investors in the data set, but cover only 15.8%

of the transactions in the sample. This emphasises again on the one hand the high con-

centration of VC investments on a few very large VC companies in Germany and on the

other hand the vast majority of small VC investors having a portfolio of only one or

very few VC investments.

On the basis of table 7 and 8 it seems appropriate to classify the investors according

to their number of transactions in groups of "one time investor" (1 investment), "very

small VC" (2-3 inv.), "small VC" (4-6 inv.), "lower middle field VC" (7-10 inv.), "up-

per middle field VC" (11-20 inv.), "large VC" (21-50 inv.) and "very large VC" (>50

inv.) (see table 9). This classification concisely displays the character of the distribu-

tion:

Almost 2/3 (N=511) of all investors in the sample take only once a VC equity stake

in a company, 14.7% (N=119) are very small VC investors with 2-3 transactions, and

9.6% (N=78) are small with still less than seven companies. 41 VC firms (5%) in the

lower middle field have up to ten investments and 32 venture capitalists (3.9%) form the

upper middle field with in-between 11-20 investments. There are 25 large VC firms

-8-

with up to 50 investments (3.1%) and only six organisations (0.7%) completed more

than 50 transactions each and thereby constitute the top flight of the 812 VC investors.

The allocation of the investors in size classes is visualised in figure 3. Table 10

contains detailed statistics on each particular size class and table 11 lists their individual

frequency distribution.

2.2.2.

Syndication Ratio of the investors

It is interesting to describe the propensity to syndicate of the VC investors in the

sample. The propensity of an investor to co-invest is expressed in this paper by its ratio

of syndicated investments to those that he did on his own. The Syndication Ratio di-

vides the number of co-invested deals by the total number of deals the investor com-

pleted in the sample (see addenda 5). The higher the Syndication Ratio of an investor,

the more he tends to invest in portfolio companies that are funded through a co-

investment.

Table 12 shows the frequency distribution of the Syndication Ratio of the investors.

The value "0" in the table indicates that the specific investor invested exclusively on his

own and was not involved in any co-investment. This is the case for 126 investors,

equaling a share of 15.5%. All other 686 investors of the sample (84.5%) did at least

participate in one syndicate during their investment activities. Only 185 investors

(22.8%) have a syndication ratio other than "0" or "1". By far the most VC providers

(N=501; 61.7%) have a syndication ratio of "1", meaning that they invested solely in

companies that were funded through a co-investment.

Based on all 812 VC investors of the sample the mean Syndication Ratio is striking

73.5%, saying that the investors of the sample tend by far more to invest in portfolio

companies that have more than one investor than to invest in companies that have no

other investors (see table 13).

However, it is important to note that this number is biased in a way. In the sample,

there are 511 investors who did only one single VC investment. If this particular com-

pany was funded through a syndicate, the single time investor gets the maximum synd i-

cation ratio value of 100% (1.0). Still it would be misleading to reason that this single

time investor has a higher propensity to syndicate than for example 3i Deutschland

which invested in 94 companies and has a Syndication Ratio of 70%.

Therefore, these

511 one-time investors which can only have a syndication ratio of either "1.0" or "0.0"

-9-

were excluded from the calculation of the next mean. In doing so, the new mean of the

overall Syndication Ratio of the sample is 0.6 and applies to the remaining 301 VC in-

vestors of the sample which have at least two investments (see table 14).

2.2.3.

Number of portfolio companies

The 3,230 transactions are distributed on a total of 1,962 different capital seeking

companies. The range with regards to the number of investors involved in the funding

of a single company is between 1-17, with one being the mode (see table 15). More than

70% (N=1,377) of all investee companies in the sample have only one investor and 92%

of them are funded by not more than three different VC investors (see table 16). Com-

panies with more than five investors on board account for 3% of the sample. Finally,

there are only nine portfolio companies being funded by ten or more VC providers. A

list with the names of the portfolio companies and their number of investors is given in

table 17.

The distribution of the companies according to their total number of investors

is shown in figure 4.

Throughout this paper, each portfolio company is allocated a Syndication Index

which is created as an important input factor for the statistical analysis throughout the

text. The Syndication Index simply calculates how many different investors a company

has. A Syndication Index of four means that the company is funded by a co-investment

of four different VC providers, for example.

From all transactions results an average of

1.65 investors per portfolio company (Syndication Index=1.65). However, if only calcu-

lated on the basis of the companies that have at least two funding investors (N=585) the

mean almost doubles to 3.17 (see table 18). Thus, if it comes to a co-investment of a

portfolio company in the sample, the average number of investors is 3.2.

The standard deviation of the number of investors (1.45) for the sample of the

1,962 companies is lower than the one of the number of transactions of the VC investors

(9.36), however also here, the distribution shows a sinistral asymmetry of four, indicat-

ing that it is non-symmetric distributed around the arithmetic mean (see figure 4).

With regards to further statistical analyses, the histogram for the portfolio compa-

nies suggests rather awry values as borders for a classification according to their num-

ber of investors (see figure 5). When looking at the frequency distribution in table 16, it

appears to be more reasonable to classify each of the 1,962 companies in groups of be-

ing funded by a "one single investor", "small syndicate" (2-3 investors); "middle synd i-

-10-

cate" (4-5 investors), "large syndicate" (6-9 investors), and "very large syndicate" (>9

investors). The overall frequency distribution of this classification is summarised in ta-

ble 19.

As already stated, only 30% of the portfolio companies have more than one inves-

tor. The aggregated 30% share of those companies with at least two investors (N=585)

is portioned out on 435 investee firms funded by a small syndicate (22.2%), 94 funded

by middle syndicates (4.8%), 47 funded by large syndicates (2.4%) and nine portfolio

companies funded by very large syndicates (0.5%) (see table 19). The overall distribu-

tion of the size classes is pictured in figure 6.

Further detailed statistics on the classification of the portfolio companies with re-

gards to by what kind of syndicate they are funded is given in table 20. Finally, table 21

contains the detailed frequency distribution within each syndication class.

2.2.4.

Geographical characteristics of the sample

In the following, the distribution characteristics and statistics of the variables de-

scribing geographical aspect of the 3,230 transactions are analysed. In doing so, the base

entity always is the transaction and not, as the names of the variables used to describe

the transactions might suggest, the VC investors or portfolio companies themselves.

2.2.4.1.

Distribution of the venture capital investors in Germany

The city of the VC investor (variable "VC City") is known for 2,679 of the 3,230

transactions of the sample. These 2,679 transactions are distributed on 90 cities of the

VC investors in Germany and 20 cities abroad (see table 22). As can be seen from this

table, Munich is the explicit leading VC city of Germany with more than ¼ of all trans-

actions. The metropolis Munich, Berlin and Frankfurt combine a share of 50.6%

(N=1,355) of all transactions emphasising their dominant role in the German VC indus-

try. The strong position of Bonn (5

place with 5.4% of all transactions) is due to one

dominant VC firm, tbg. The geographical concentration of VC investors in few large

cities is also emphasised by the fact that 68.2% (N=75 cities) of the 110 cities are less

than ten times the basis for transactions of VC investors.

On a next level, smaller cities in the catchment area of a metropolis are allocated to

clusters. Based on the variable "VC Cluster", 2,535 transactions of VC investors are

spread on 47 clusters in Germany and show for the large cities a highly similar distribu-

-11-

tion as the variable "VC City" (see table 23 and figure 7). Also here, the cluster Munich

clearly is at the top with 711 transactions (28%). It is followed by Berlin (N=409,

16.1%) and Frankfurt (N=336, 13.3%). The three most frequent clusters Munich, Berlin,

and Frankfurt combine a share of 57.7% (N=1,465) of the 2,535 transactions, which is

7% more than for the variable "VC City". This is explained by the fact, that these clus-

ters include transactions from smaller cities nearby. Surely, the high concentration of

the clusters is even higher than that of the VC firm cities: ¾ of all 2,535 recorded trans-

actions are allocated to VC investors coming from one of the six largest out of 47 total

clusters.

The state the VC investor comes from (variable "VC State") is known for 2,644 of

all 3,230 transactions in the sample. As expected from the distribution of the cities and

clusters, by far the greatest share of these transactions (29%; N=767) are allocated to

VC firms from Bavaria (see table 24). The next top three states Nordrhein-Westfalen

(N=451; 17.1%), Berlin (N=409; 15.5%) and Hessen (N=371; 14.1%) all have a similar

share of the transactions. Combined with Bavaria, they bundle more than ¾ of all trans-

actions in German portfolio companies. Hessen clearly ranks before Hamburg (N=184;

7%) as there are a number of large VC firms based in Frankfurt. VC investors from for-

eign countries account for 3.1% of the transactions which still places them before

Thüringen (1.9%), Brandenburg (1.5%), Niedersachsen (1%), and Rheinland-Pfalz,

Saarland, Schleswig-Hostein, Mecklenburg-Vorpommern and Bremen, the latter ones

being home to VC investors which did less than 1% of the 2,644 transactions.

2.2.4.2.

Distribution of the portfolio companies in Germany

The city of the portfolio company (variable "Investee Company (IC) City") is

known for 2,980 of all 3,230 transactions of the sample (see table 25). These transac-

tions are distributed on 478 different cities, which meets the expectation that the inves-

tee companies are much stronger spread across Germany than the VC firms. Similar to

the VC firms, Munich (N=473) and Berlin (N=366) are the most dominant cities in the

sample and account for 28.2% of the 2,980 transactions. Frankfurt (3.1%), the 3

most

popular VC city (9.6% on the variable "VC City"), is home to less capital seeking firms

than Hamburg (7.3%), Cologne (5.2%) and Martinsried (3.7%). Generally, the distribu-

tion of the transactions across the group of the most frequent investee company cities is

more balanced than the one for the VC firm cities. But the degree of concentration of

-12-

the distribution still is very high with the top eight of the 478 portfolio company cities

combining 50.6% (N=1,509) of all 2,980 transactions.

The information from which of the 47 clusters the portfolio company comes (vari-

able "IC Cluster") from is given for 2,517 of the 3,230 transactions (see table 26). The

dominant position of Munich (N=679; 27%) is at least partly explained by the allocation

of Martinsried (N=111 for the variable "IC City") to this cluster. As for the variable "IC

City", Berlin forms the second largest cluster (14.8%) followed by Hamburg (9.1%),

Cologne (7.7%) and Frankfurt (5.8%). The five largest clusters make up 64.4% of all

2,517 transactions. The distribution in form of a bar chart is shown in figure 8.

The variable "IC State" provides on 2,740 transactions the information in which

state the investee company is based. Bavaria with almost 1/3 of all transactions (N=805)

is by far the most represented state of the portfolio companies of the sample (see table

27). The second most frequent state is Nordrhein-Westfalen (N=430; 15.7%) followed

by Berlin where the companies in 370 transactions came from (13.5%). These three

states already make up 58.6% of the 2,740 transactions. If complemented by Hamburg

(N=222; 8.1%), Hessen (N=210; 7.7%), and Baden-Württemberg this number increases

to 81.6%. The remaining 504 transactions (18.4%) are distributed on ten German states

and 21 foreign countries.

Overall the distribution shows that the VC investors, as well as the portfolio com-

panies, are highly concentrated on few regions in Germany. Clearly the cluster Munich,

where by far the largest number of VC firms and investee companies are located, is the

most popular German metropolitan area for VC. The clusters Munich, Berlin, Frankfurt,

Cologne and Hamburg play the absolute prominent role in the German VC industry.

Other cities and regions are only of secondary importance. The distribution also clearly

reflects that there are by far more portfolio companies in West - than in East Germany

and that those VC investors based in East Germany mostly operate from Leipzig, Jena

and Dresden.

The variable "Distance" measures for 1,390 transactions the number of kilometres

(km) between the VC firm and its portfolio company. In most cases (N=604; 43.5%),

the VC investor and the investee company are located in the same city (see table 28).

There is also a considerable number of transactions (N=125; 9%) where both parties are

very close to each other (up to 50km). In contrast to this, 500 transactions (36%) show a

distance of more than 200km between the two parties. Within this group, 212 transac-

-13-

tions (15.3%) involve a company and an investor which are more than 500km afar and

in 288 cases (20.7%) 200-500km lie in-between the two. The classes "up to 200km"

(N=97; 7%) and "up to 100km" (N=64; 4.6%) are less frequent with the latter being the

scarcest one. Thus, one can conclude that the investors and their portfolio companies

either tend to be very close together with a distance of not more than 50km (N=729;

52.5%), or rather far away from each other with a distance of more than 200km or even

500km (N=500; 36%). This distribution is visualised in figure 9.

Rationales for venture capital syndication

There exist two dominant competing views as to why venture capitalist syndicate,

which are the traditional finance-related perspective and the resource-based perspective.

A third perspective acknowledges the impact of reciprocated deal flow which some-

times is also related to the resource-based perspective. In the following, these three per-

spectives will be primarily investigated. Other rationales behind syndication, such as

window dressing and the collusion of VC firms, are also presented. Then, the empirical

relevance of the finance-related rational and the resource-based rational in Europe and

the US are analysed.

All rationales as follows are described from a perspective to syndicate out an in-

vestment. Locket and Wright (1999) find that the motivation to join a syndicate is ex-

plained by the same factors to syndicate out an investment: the traditional finance per-

spective, the resource-based perspective or deal flow.

3.1.

The finance-based motive

The finance-related rational for syndication originates from finance theory and is

therefore also seen as the traditional approach. The finance perspective is to see synd i-

cation as a mean for venture capitalists to build up a well-diversified portfolio and re-

duce risk without reducing return. The relevant risk-consideration for a VC investor is

the contribution of an investment to the overall risk of his portfolio. This depends on the

covariance of the portfolio and the investment opportunity. There are two subdivisions

of risk involved in an equity investment. While the market component is systematic and

cannot be eliminated, the firm specific risk component is non-systematic and can there-

fore be reduced by holding a well-diversified portfolio. In a well-balanced VC portfolio

-14-

there exists a minimum level of co-variance between the different investments (Mani-

gart at al. (2002), Lo cket and Wright (2001), Markowitz (1959)).

Syndication allows a venture capitalist to participate in more deals and to spread his

financial risk on a larger number of investments. Therefore, syndication can serve as a

mean to considerably reduce the variation of returns without lowering the overall ex-

pected return of the portfolio (Manigart et al. (2002)). The constraints on investment

activities are based on the Modern Portfolio Theory. Its main principle is the efficient

diversification of investments (Elton and Gruber (1995)).

Firstly, venture capitalists encounter the difficulty to obtain a well-diversified port-

folio, since they do not invest in listed stocks as institutional investors. The difficulty

arises on the one hand from ex-ante asymmetric information and also from the size of

the fund required (capital restraints).

This demonstrates that through syndication smaller venture capitalist can actually

invest in deals with a high amount of required funds. Thus, syndication enables venture

capitalists to invest in more companies and thereby increases the balance of their portfo-

lio. The result is a reduction of the overall risk of the fund (Manigart et al. (2002)).

Another characteristic of the VC industry promoting syndication is the illiquidity of

investments. This means that venture capitalists have their capital bound to the investee

company for a longer period. The minimum investment period makes equity illiquid in

the short- to medium-term and therefore equity cannot be continuously traded. Further-

more, due to ex-ante information asymmetry, the actual risk of a project may turn out to

be higher than expected. Under these circumstances it may be difficult to adjust the

portfolio by divesting because the capital is bound to the investee company. As a result,

venture capitalists should strive to access a diversity of projects at the initial investment

stage because these require a lower investment amount (Lockett and Wright (1999)).

To summarise, syndication enables the sharing of risk on a deal-by-deal basis that

could lead to a reduction of the overall portfolio risk without giving up return (Manigart

et al. (2002)).

3.2.

The resource-based motive

Frequently syndication occurs in the VC industry even though the amount of fund-

ing required for the investment opportunity is relatively modest compared the financial

-15-

resources of the venture capitalists. The resource-based perspective raises the question

why venture capitalists still syndicate in such situations. (Brander (2002)). The re-

source-based approach sees the VC market as a pool of productive resources in which a

VC organisation can access resources of another venture capitalist through syndication

(Manigart et al. (2002), Bygrave (1987).

3.2.1.

The Superior Investment Selection Hypothesis

At the pre-investment stage, Lerner (1994) suggests the Selection Hypothesis as a

rational for VC syndication. Under this hypothesis the evaluation process before the se-

lection of an investment opportunity is undertaken by more than one venture capitalist.

The evaluation of the same venture proposal by different VC companies operating in a

syndicate reduces therefore the potential danger of adverse selection (Lerner (1994,

Houben (2002)). The combined effort to assess the quality of a venture helps VC inves-

tors to overcome informational asymmetries as the entrepreneurs typically know more

about the investment opportunity they seek funding for and might overstate the attrac-

tiveness of his business proposal (Sorenson and Stuart (1999)).

Sah and Stiglitz (1986) compare the decisio n-making process under different sce-

narios: In the first scenario the project is already accepted when a single party thinks

that it is worth undertaking. In the second scenario, however, two or more separate par-

ties must be convinced by the investment opportunity before the project is undertaken.

Sah and Stiglitz (1986) conclude that the decision making process is more efficient and

leads to better results if the project is only undertaken when approved by two or more

parties.

Transferred to the situation of a VC syndicate this means that the involvement of

two or more venture capitalists leads to better decisions whether to invest in a venture

not. The same investment opportunity is screened and evaluated by different VC firms

under different aspects. If all potential syndicate members believe that it is worthwhile

to invest in the venture this is a good indication for the success of the investment.

The Superior Selection model expects syndicated deals to be of higher quality than

deals selected and completed by a single VC investor. However, Brander (2002) takes a

different view on the common assessment of an investment opportunity. He points out

that the reputation of the investing parties plays an important role since the more estab-

lished and renowned VC firms will firstly receive projects for evaluation. Therefore, at

-16-

first they can assess and decide solely over the project. If they find the project not to be

of sufficient quality and potential they will reject it without further consideration. On

the other extreme, if the project is very promising and offers a high expected return they

will accept the project without the consultation of a second party. The more established

VC firms then also have mostly the financial resources to fund the project without the

help of a syndicate and try as a single investor to get all the financial benefits from the

lucrative project.

These two extreme scenarios of a very bad and a very promising project lead to the

conclusion that syndication occurs in the intermediate range of the assessed project

quality. If the evaluating venture capitalist is not able to clearly identify a project as

profitable it will pass it on to other VC firms for their review and evaluation. In case

different other VC providers evaluate the venture and believe that it is profitable enough

then an investment syndicate may be formed for the project of intermediate quality

(Brander (2002).

This results in the conclusion that stand-alone investment projects should promise

higher average returns than syndicated investments. Therefore, by assuming that the

Superior Selection Hypothesis is a central motive for syndication, it has to be acknowl-

edged that syndicated investments possibly lead to lower average returns than stand-

alone projects (Brander (2002)).

Finally, it has to be stated that the Superior Selection Hypothesis does not hold in

the case of a lead investor undertaking the whole deal at first and then syndicating down

the investment to other investors. In this case the risk-sharing perspective is more ap-

propriate as the decision to invest has already been made (Lockett and Wright (1999)).

3.2.2.

The Value Added Hypothesis

The Superior Selection Hypothesis applies to the pre-investment stage whereas the

Value Added Hypothesis in terms of managerial activities is a resource-based motive

for syndication which holds for the post-investment stage.

Under the Value Added Hypothesis venture capitalists are considered to add value

to the performance of the venture after they invested in it. This contrasts with the selec-

tion hypothesis, where syndication helps investors to select the best projects, but does

not influence the performance of the investee company (Brander (2002)).

-17-

A lead investor acts according to the Value Added Hypothesis when he believes

that the involvement of other venture capitalists would add some value to the venture.

The benefit of involving co-investors is derived from heterogeneous skills and informa-

tion different venture capitalists can contribute to the management of the venture com-

pany. The need for such additional resources is anticipated to be greater in earlier stages

of an investment, than in later-stage investments. This is mainly due to the fact that

more mature investee-companies already have an established management structure and

market position and have already built relationships with suppliers and customers

(Locket and Wright (1999), Brander (2002)).

3.2.3.

Evaluation of the Value Added Hypothesis and the Selection Hypothesis

When evaluating the Value Added Hypothesis it is important to compare the ob-

served financial returns of investments that were syndicated to those that were funded

by a single investor. The Value Added Hypothesis predicts that the financial returns on

syndicated investments should exceed the return of a stand-alone investment due to the

sharing of resources in a syndicate. Also the Superior Selection Hypothesis expects

syndicated investments to be of higher quality as they were reviewed and approved by

different VC investors.

Contrary to this, Brander (2002) also sees The Value Added Hypothesis as a rele-

vant motive to syndicate leading to higher returns of co-invested deals but expects the

Selection Hypothesis to lead to lower returns of syndicated investments. Thereby, he

finds the two hypotheses to be mutually exclusive. In order to test if syndicated VC in-

vestments have higher or lower returns he uses the Macdonald & Associates database

for returns on Canadian VC investments. He finds that syndicated investment projects

have significantly higher returns than investments of a stand-alone status. Thus, accord-

ing to Brander (2002) the Value Added Hypothesis should be favoured over the Selec-

tion Hypothesis in order to explain the resource-based rational for syndicating VC in-

vestments.

Brander (2002) does not deny the value of a second opinion for the investment se-

lection process, but emphasizes that venture capitalists do much more than merely se-

lecting profitable projects. He sees the post-management and development of the ve n-

-18-

ture as the core competency of venture capitalist activities rather than the investment

selection.

Considering the possibility that both motives and hypothesis could operate at the

same time it could be assumed that "Syndication is a response to the need to share in-

formational resources in the ex ante selection and ex post management of invest-

ments."(Lockett and Wright (1999), p. 307).

3.3.

The deal flow motive

Due to syndication, venture capitalists can expand their geographical and industrial

reach. This is mainly based on two factors trust and reciprocity. Prior co-investing ex-

perience can create a relationship based on trust and competence and can be a strong

influence in the selection of future potential partners. Trust is an important factor since

VC firms do not only invest in local ventures. Therefore, they often rely on the evalua-

tions of another investor since it could be closer to the venture in a physical or industry-

related way. Reciprocity is reflected in the fact that different investments are often made

by the same group of investors which already shared a number of investments before.

This demonstrates that venture capitalists build up ties with one another and then invite

each other to participate in new deals (Sorenson and Stuart (1999)).

A lead investor hopes that in the future he will be invited by his co-investors who

then put together a syndicate under their lead. By syndicating deals to others, venture

capitalists anticipate to receive the same gesture in the future, thereby increasing deal

flow (Lockett and Wright (1999)).

As a result, deal flow can be seen as an important intangible resource. VC organisa-

tions compete for as many investment opportunities as possible because it is favourable

to be able to select from a wide range of ventures. Thus, new VC firms have to establish

themselves in the local market before being invited by geographically distant venture

capitalists. These networks are important to venture capitalists in order to establish deal

flow, especially in times of high competition where reciprocation can be a mean to

maintain deal flow. In this context it is not important any more whether the firm was

originator of the deal or not. A steady deal flow is vital for venture capitalists, especially

for those in highly competitive markets with only few good deals (Boviard (1990),

Sorenson and Stuart (1999)).

-19-

Although investors strive to expand their radii of investments, this expansion oc-

curs mainly by joining syndicates with geographically distant lead venture capitalists. In

this sense, syndication can serve as a mechanism to enter trans-national networks in or-

der to expand spatial reach for potential investments. Syndication can thereby reduce

effective spatial limitations on the information flow and thus increases the possibility of

getting access to co-investments and local expertise beyond market boundaries

(Sorenson and Stuart (1999)). Therefore, it is possible to venture capitalists who have

already built up a good syndication network to "[...] transcend the natural boundaries

on the dissemination of information and expand the radius of their investment activity."

(Sorenson and Stuart (1999), p.5)

3.4.

Other motives

3.4.1.

Window dressing

In this case, venture capitalists use the same strategy as money managers who ad-

just their portfolios through investing in successful ventures in order to receive good

publicity. In the VC industry, such kinds of investments are usually made shortly before

the target company goes public with its lead investor. Investments in already successful

projects at a later stage offer to the VC investor to "window dress" his investment per-

formance. For publicity reasons, the venture capitalist can demonstrate his investment

know-how by emphasising his participation in successful ventures. He will not mention

that he joined the syndicate in the last financing round and was not really involved in

the development of the venture. The expected return under these circumstances is only a

secondary factor. Also early-stage investors receive the opportunity to invest in projects

at later stages thus receiving the same benefits (Gompers and Lerner (1999)).

The window dressing hypothesis suggests that the lead investor also benefits from

offering to other venture capitalists to co-invest in his successful deals. The lead inves-

tor should identify those co-investors that are well-established. These are the most inter-

esting since they are the most capable to reciprocate. The idea behind reciprocation is

that the invited venture capitalist will himself offer to join in its successful later stage

investments. The concept of window dressing can thereby create deal flow and increase

spatial reach (Lerner (1994)).

-20-

Brander (2002) used his research to evaluate the relevance of window dressing in

the VC industry. Since it involves two investments at two different points in time, he

uses the broader definition of syndication (see chapter 1.3). Yet, as he compared the av-

erage returns for narrow and broader defined syndicated investments, he found little dif-

ference. He also states that most syndicates in his sample were of narrow definition and

therefore it seems that only few venture capitalists are joining late. Overall, (Brander

(2002) sees little if no evidence for window dressing as a motive for syndication.

3.4.2.

Collusion

One further potential explanation for syndication is collusion. Hereby, venture capi-

talists would strive to syndicate in order to band together rather than competing against

each other. Under this scenario the main aim of the VC investors would be the attempt

to improve their own bargaining position with entrepreneurs.

Syndication of the deal could also have an important impact on control rights

within the venture. Especially in VC financing control of a venture plays a key role

(Brander (2002).

3.5.

Empirical relevance of the major motives

Different empirical studies over the last years come to varying conclusions as to

whether the desire to share risk and increase portfolio diversification is a more impor-

tant rationale for syndication than the desire to access additional informational resources

or deal flow. As this is the prevailing controversy and issue in the literature on VC syn-

dication, it is appropriate to draw a comparison to what conclusions the existing empir i-

cal studies come.

Bygrave (1987) found that there is more co-investing when there is a higher level

of uncertainty. His comparison of the more conservative consumer and the more risky

computer industry in the USA showed a clear tendency of co-investing in the high in-

novative computer sector.

There was also more syndication in early-stage investments

than later-stage investments, even though the investment amount required was on aver-

age 40% lower for early-stage investments. These findings seem to question the fi-

nance-based rational as it assumes higher syndication for deals requiring a la rger

amount of funding. Thus, Bygrave (1987) concluded that the main motive for syndica-

tion was rather the sharing of experience and other intangible resources than capital re-

-21-

straints and the spreading of financial risk. In his findings he also refers to Pfeffer and

Salancik (1978) who found similar evidence in their studies on joint ventures. In an-

other publication, Bygrave and Timmons (1992) again emphasise the great role uncer-

tainty plays in the decision to syndicate which can be reduced by the sharing of infor-

mation and the access to resources from the syndicate members.

In 1997, Chiplin et al. (1997) also drew conclusions in favour of the resource-based

motive in their study. They found the greater support for syndication as a mean to im-

prove deal selection through joint decision making. Chiplin et al. (1997) acknowledge

the importance of costs in the VC market, but can only find weak support for the risk

sharing perspective as a motive to syndicate.

Contrary to this, Lockett and Wright (1999) are with their findings on the UK mar-

ket clearly in favour of the financial-based rational as primary explanation for syndica-

tion.

They find that the large size of a deal compared to the funds that are available to

a single venture capitalist is significantly more important than all other factors. The

need for additional information before making a decision turned out to be the least rele-

vant explaining factor. This leads to the conclusion that that the ex-ante Superior Selec-

tion Hypothesis is less important than the ex-post Value Added Hypothesis both being

allocated to the resource-based perspective. According to Locket and Wright (1999), the

best explanation for syndication, however, is based on the traditional finance perspec-

tive.

Lockett and Wright (1999) split the UK venture capitalists into two groups. On the

one hand those with a maximum investment preference of £5 million and on the other

hand those with a minimum investment preference of £5 million. The reasoning behind

this is to separate those VC firms that exclusively invest in MBO/MBI later stage in-

vestments from the rest of the sample being characterised by a higher degree of uncer-

tainty and variability in outcomes. Locket and Wright (1999) find that the two groups

with their different minimum investment preference ha ve different attitudes towards

syndication in a way that they have different financing requirements. The level of risk

associated with smaller early stage investments is considerably higher than for late stage

investments. Thus, venture capitalists that have a preference or limitation to smaller

minimum investment amounts need to diversify away the greater risk through the crea-

tion of a portfolio of syndicated investments. Moreover, the findings suggest that the

exchange of informational resources in the ex-ante selection and ex-post management of

-22-

investments is greater for VC firms with a focus on earlier stage investments. The

asymmetry of information is generally higher for these deals and a higher degree of spe-

cialisation is required.

To summarise, Locket and Wright (1999) come to the conclusion that both the fi-

nance-based rationale and the resource-based rational are more important to VC firms

with a lower minimum investment preference with the finance motive being generally

significantly more relevant. The deal flow rational does not differ across the two groups

with their differing investment stage preferences.

In 2000, Hoje Jo (2000) again focuses on the resource-based rational to conduct his

research. However, he emphasises that it is unclear to him to what extent risk sharing

aspects influence venture capitalists to syndicate. Maula and Murray (2000) support this

view in their study. They identify the need for complementary resources, including in-

tangible assets like industry experience or

tangible assets like warehousing. They offer

no explicit findings regarding to what extent the financial perspective is involved as a

motive for syndication.

In 2002, Brander (2002) concentrates like Maula and Murray (2000) on the re-

source-based rational. In his conclusion, he clearly favours the Value Added Hypothe-

sis. This is underlined by his findings that syndicated investments have higher rates of

return than stand-alone investments. He acknowledges the value of a second opinion in

the investment selection process, but states that his empirical analysis identifies the

value added effect as the driving force behind VC syndication. He concludes that risk-

sharing might play a role, but emphasises at the same time that he sees capital con-

straints only as an issue in come cases and rather not for large VC firms which do most

of VC investing.

The most up to date research on this discussion was just recently finished by Mani-

gart et al. (2002) who pick up the study by Locket and Wright (1999) on the UK and

extend it to a European context.

In their study of six European countries, Manigart et al. (2002) come to the conclu-

sion that the traditional finance perspective is significantly more important than all other

perspectives and motives.

It is remarkably that this finding is consistent across all

countries of the study. For the UK and France, they even find significant evidence for

the deal flow motive being more important than the resource-based perspective. The

-23-

same is found for Germany, however only at slightly more significant level. The impor-

tance of the deal flow and resource-based perspective is equal in Sweden, Belgium and

The Netherlands. Overall, the results of this comprehensive study suggest that the fi-

nancial motive is the only important motive European VC firms consider when synd i-

cating.

A new aspect discussed in this study is the differentiation between the syndication

motives of lead- and non lead investors. Lead investors usually are the driving force in

the constitution of a deal while non-lead investors more passively react to the invitation

to join a syndicate. Thus, they are expected to have a different decision-making process

with regards to syndication. A lead investor is more likely to syndicate a forthcoming

investment if it is a high risk project or if the required funds exceed his financial limits.

This suggests that the financial motive is more important to a lead venture capitalist

than to a non-lead venture capitalist because it allows him to balance its portfolio at a

given investment strategy. However, this expectation could not be verified by the results

of the study. The finance-based rational for syndication seems to be equally important

to lead investors than to non-lead investors. The same holds for the deal flow motive

(Manigart et al. 2002).

Manigart et al. (2002) find strong support for the hypothesis that the resource-based

motive is more important to non-lead investors than to lead investors. Being invited by a

respected lead investor to join a syndicate enables a non-lead investor to increase his

resource base in terms of legitimacy and reputation. It is a signal of its quality to the VC

community and moreover allows the co-investor to learn from other syndicate members.

To summarize, both studies on the European VC industry lend strong credibility to

the traditional finance rational which is also found to be the dominant motive for Ger-

man VC firms. Opposing to this, the findings of studies on North American syndicates

come to the conclusion that the motive for VC syndication is derived from additional

resources and added value that are brought in by syndicate partners although they also

acknowledge the relevance of financial considerations.

There is a potential explanation for the different findings on the rationales behind

syndication in the US and Europe. Manigart et al. (2002) and Locket and Wright (1999)

use perceptual data originating from a survey. The North American studies use outcome

data and thereby derive their findings on syndication patterns from actual investment

decisions. From this, Manigart et al. (2002) conjecture: ,,Maybe European VC managers

-24-

unconsciously underestimate the importance of access to resources and deal flow, but

their actions might nevertheless reflect the fact that these are important. Ho wever, if this

is true, then European VCs should be made more conscious of the potential benefits

syndication may yield, beyond risk reduction considerations. In this way, they might

more consciously develop a strategy relating to this important dimension of ma naging a

VC firm." (Manigart et al. (2002), p. 18).

And Locket and Wright (1999) write in this context with regards to the UK market:

,,The lack of support for the skills accessing and deal flow rationales for syndication

casts some doubt on the efficacy of attempts to add value in new types of investments

such as investor-led buy-outs and leveraged build-ups. VC firms may be giving inade-

quate emphasis to the need to access other skills to enable them to make superior deal

selection decisions and more importantly in the light of the present findings, in terms of

the subsequent management of investments"(Locket and Wright (1999), p. 322).

Thus, as opposed to North American VC firms, European VC managers may not

sufficiently appreciate resource-based benefits such as getting access to additional re-

sources, establishing strong and trustworthy networks and increasing deal flow (Mani-

gart et al. (2002)).

Factors influencing the propensity to syndicate

The herein before discussed motives play an important role in the propensity to

syndicate. This is due to the finding as postulated by Manigart "[...] that a firm's pro-

pensity to syndicate will be positively associated with finance-based, resource-based,

and deal flow motives for syndication, controlling for firm-level characteristics." (Ma-

nigart et al. (2002), p. 8).

The stronger VC firms believe that financial, resource-based

and deal flow rationales are important the higher their propensity to syndicate their

deals.

In the following, factors that influence a venture capitalist's propensity to syndicate

are further investigated.

-25-

4.1.

Degree of uncertainty

Syndication can be seen as a powerful mean to reduce the overall high degree of

uncertainty in which VC organizations operate. Therein, Bygrave (1988) found that the

degree of uncertainty a firm faces determines its amount of co-investing. Generally, a

higher level of uncertainty implies a higher amount of syndication.

This relation is theoretically based on the Resource Exchange Model by Pfeffer and

Salincik (1978) postulating that the interconnectedness of firms increases with the

amount of uncertainty in their environment. As pointed out by Bygrave (1988), there

also exist competing models, such as the Population Ecology Model by Wholey and

Brittain (1986), but the Resource Exchange Model turns out to be most suitable to ex-

plain the networks formed by VC syndicates. The model consists of the three structural

variables "concentration", "munificence and "uncertainty" which determine the inter-

connectedness of a firm. Interconnectedness describes the stability of linkages between

firms within an industry. Among others (e.g. interlocking directorates), joint ventures

are one common form of interconnectedness. The model is based on the assumption that

the interconnectedness of companies increases with the degree of uncertainty in the en-

vironment.

Concentration as the first variable is the number of competing firms in an indus-

try.

Back in 1987, Bygrave (1987) classified the VC industry as having intermediate

concentration at most. It is expected that in times of prosperity the number of invest-

ment opportunities increases and concentration is low and vice versa in times of hard-

ship. Locket and Wright (1999) come to similar conclusions.

Under the variable munificence, Pfeffer and Salincik (1978) understand the scarcity

or abundance of the main resources a firm derives from its environment. Transferred to

the VC industry these would be resources such as capital, deal flow and people to ma n-

age the investments. The manipulation of these factors influences the behaviour of a VC

organization towards syndication. According to Bygrave (1987), low amounts of spend

able cash or few investment opportunities tend to lead to more syndication.

The third and most important variable is uncertainty measuring the degree to which

the outcome of an event cannot be predicted (Knight (1921)). Each industry has a cer-

tain degree of uncertainty and especially the VC industry with its focus on high risk

high return investments has a high degree of uncertainty. Apparently, the level of uncer-

-26-

tainty and risk widely depends on in what kinds of portfolio companies a venture capi-

talist primarily invests. Bygrave (1987) splits his sample in two kinds of VC firms: On

the on hand venture capitalists that mainly invest in high innovative technology compa-

nies (HIVCs) operating in very uncertain environments. The top 21 US HIVCs in his

sample comprised a tightly coupled network (connectedness: 37%)

. On the other hand

Bygrave (1987) separates the venture capital firms primarily funding low innovative

technology companies (LIVCs) which are less risky on average. The connectedness of

the top 21 US LIVCs was clearly lower with 22%. Thus, these findings demonstrate that

a greater degree of uncertainty (HIVCs) implies a greater degree of connectedness in

form of syndication, whereas the LIVCs are more loosely bound and show less syndica-

tion. This is in accordance with the predictions by the Resource Exchange Model.

To centralise, the Resource Exchange Model provides an appropriate theoretical

framework for analysing the VC industry and for explaining the relationship between

uncertainty and syndication. It predicts that the interconnectedness of firms is greater in

more uncertain and less munificent industries. The interconnectedness is supposed to be

highest in industries with intermediate levels of concentration which applies for the VC

industry (Bygrave (1987), Bygrave (1988)).

4.2.

Industry of the investee company

It is examined if the industry of a portfolio company has an impact on the degree of

its co-investing. Bygrave (1987) picked the consumer and the computer industry for a

comparison because they largely differ in their market uncertainty and risk. He found

the ratio of pairs to single investments to be five times greater in computer- than in con-

sumer companies indicating that investments in the riskier computer industry have a

higher propensity to be syndicated. Keeping in mind that there was hardly any differ-

ence in the average total investment amount per company in the two industries, Bygrave

(1987) concludes that this lends support to the resource-based motive rather than the

financial motive.

To see if Bygrave's (1987) findings that there is higher connectedness and more co-

investing in riskier industries can be confirmed with the data base, a univarite linear

model is run with the Syndication Index of the investee companies as dependent vari-

able and their industry as independent factor (Stapleton (1995), Brase and Brase

(1999)).

-27-

For a detailed description of the frequency distribution of the variable "IC Industry"

in the data set refer to addenda 6. The processing of the data set that was needed to set

up the model is described in addenda 7. As a result of the model, table 30 calculates the

average number of investors per company for all different industries of the portfolio

companies in the data set. Internet Service Providers have on average 2.5 VC investors

followed by Biotechnology firms with a mean of 2.2 investors. Both types of E-

Commerce firms (B2C and B2B) have on average two VC investors. These four indus-

tries with the largest average number of investors are all of high market risk.

Portfolio companies from industries with a lower market uncertainty such as Indus-

try & Construction (1.2 VC investors) or Electronics (1.2) ha ve on average one investor

less than Biotechnology firms or Internet Service Providers. However, there are also

industries of relatively higher market risk such as the Internet sector (1.5) or Techno l-

ogy (1.4) having less VC investors than the more conservative Infrastructure & Logis-

tics sector (1.6), for example.

Overall, there is a high significance of the factor "IC Industry" on the Syndication

Index explaining 4.2% of the variance of the number of investors per portfolio company

(see table 31). This low overall impact on the variance is also reflected in table 32. It

calculates that there is not a single significant difference in the direct comparisons of the

means of the Syndication Index of the different industries. Even though portfolio com-

panies from some industries have on average one investor less than other companies,

these differences are not found to be statistically significant.

Based on this finding it is in a next step analysed if the model comes to the same

conclusions if only those portfolio companies are analysed that have at least two VC

investors. Thus, it is analysed how many investors the companies from the different in-

dustries on average have if it actually comes to a co-investment with a number of inves-

tors. Table 33 shows for the different industries the average number of investors if a

company actually is co-invested. This time, Biotechnology firms (4.2 VC investors)

rank before Internet Service Providers (3.9). If it comes to an co-investment, the number

of VC investors involved is approx. the same for companies from the Internet- (3.4),

Life Science & Pharmacy- (3.4), B2C- (3.2) and B2B industry (3.3). Portfolio compa-

nies from the Hardware & IT- (2.1), Services- (2.7) and Media & Communication (2.8)

sectors are at the other end of the distribution having on average the lowest number of

investors.

-28-

Again, there is an overall high significance of the portfolio company industry on

the variance of the number of investors accounting for 7.8% of that variance (see table

34). But also this time, the means of the Syndication Index of the different industries are

not found to be significantly different in their direct comparison (see table 35).

Overall, there is no evidence that portfolio companies from certain industries have

statistically significantly more or less investors than companies from other industries.

However, it is at least noticeable that companies in high market risk industries such as

Biotechnology, E-Commerce and Internet Service Providing have a clearly higher Syn-

dication Index than firms from less risky industries such as Finance & Banks, Industry

& Construction and Infrastructure & Logistics. The average number of investors in a

syndicate in the different industries varies by up to two investors

In a next step, it is interesting to find what types of VC firms tend to syndicate in

the different industries of the portfolio companies. This would clarify if there is a cer-

tain pattern whereupon different types of VC investors preferably syndicate investments

in certain industries. The proper statistical method to approach this question is the use of

a Chi-squared test (Greenwood and Nikulin (1996), Dowdy and Wearden (1991)).

For an N=1,689 transactions the Chi-squared test of the variables "VC Type" and

"IC Industry" yields a significance level of 0.00 (see table 36). But 172 cells (67%) of

the contingency table have an expected cell count of less than five and the minimum

expected cell count is below one (0.07) which both violates Cochran's rule (Greenwood

and Nikulin (1996)).

The high number of under-represented cells is caused by the

many different characteristics of the variables "VC Type" (16 different VC types) and

the variable "IC Industry" (17 different industries). The contingency table (which is a

16x17 matrix) must be cut on both dimensions (axes) for those variable characteristics

whose cell count is too little (see addenda 8).

The significance of the adjusted contingency table still is 0.00 and based on N=

1,229 transactions (see table 37). Only 14.3% of the cells in the matrix have an expected

cell count of less than five and there is no cell with an expected cell count of less than

one now allowing for an interpretation.

The contingency table reveals the following interesting syndication patterns (see

table 38): Foreign- (+36%)

and Public VC investors (+63%) tend significantly more to

syndicate in the Biotechnology sector than would have bee expected from their distribu-

-29-

tion in the sample. But also Established Specialists (+20%) disproportionately prefer

this high risk industry, whereas Corporate/Industry- (-95%), Non-Professional VC in-

vestors (-60%) and Young Specialists (-50%) distinctly avoid the Biotechnology sector.

The discussion earlier in the text suggested that next to Biotechnology companies

also Interne t Service Providers have on average clearly more investors than companies

from other industries. The contingency table suggests that it is mostly Foreign investors

(+133%) and Bank VC firms (+122%) who have a high propensity for co-investments

in this industry.

On the other hand, Public venture capitalists and Young- and Estab-

lished VC firms clearly refrain from co-investments in this sector.

The two industries B2C- and B2B E-Commerce cannot be as clearly attributed to

certain types of VC investors as it is the case for Biotechnology. Only Young Specia l-

ists and partly also Foreign- and Non-Professional VC investors reveal a higher propen-

sity to co-invest in these industries, while most other VC types show no noticeable

deviation from their expected frequencies.

Those industries with the lowest average Syndication Indexes and a relatively low

market risk such as Hardware & IT, Finance & Banks and Industry & Construction are

not listed in the contingency table as their expected cell counts are too little and there-

fore cannot be attributed to certain VC types.

The Service industry being of moderate risk seems to be disproportionately attrac-

tive to Non-Professional VC investors (+79%) and Young Specialists (+17%) whereas

Established Specialists are less represented (-22%). Corporate/Industry investors

(+75%) seem to prefer co-investments in the Infrastructure & Navigation sector what

this industry is surprisingly equally distributed on all other types of VC investors.

These findings also widely apply if not only co-invested portfolio companies are

included in the analysis but rather the whole sample of transactions is analysed. A de-

tailed description and report on the general investment preference (co-invested as well

as non-co-invested deals) of the VC types is therefore omitted.

The exemplary interpretations of the contingency table demonstrate that some types

of VC investors seem to have a significant preference for or against some portfolio

company industries. However, these characteristics are particularly different for each

VC type and there is no overall pattern whereupon certain groups of VC investors have

the same or distinct different investment behaviour in terms of their focal industries.

-30-

4.3.

Type of the venture capital investor

Addenda 9 describes in detail the different types of VC investors and their distrib u-

tion in the sample. It is analysed if these different types of the 812 VC investors in the

data base generally have a differing propensity to syndicate. This is done through a

comparison of the means the different VC types score on the Syndication Ratio and

Syndication Index (see addenda 10).

Table 42 lists the average Syndication Ratio for each type of VC investor in the

sample. Most noticeable in this table is that Non-Professional VC providers have a syn-

dication ratio of 1.0, meaning that none of the 89 one-time-occasion VC investors

funded their portfolio company on their own. But also practically all of the 123 different

Foreign VC investors, which are mostly professional venture capitalists, invested in

their German portfolio companies together with a number of co-investors (Syndication

Ratio=0.88).

Private VC Investors are at ¾ invested in portfolio companies with more than one

investor, which is similar to Corporate/Media- (0.79) and Bank VC firms (0.78). The

eleven Specialists in the sample have the lowest Syndication Ratio (0.37) followed by

Incubators (0.41). Young Specialists (0.56) and Established Specialists (0.58) have al-

most the same ratio of co-invested to non-co-invested deals. Also Public venture capi-

talists (0.65) and Business Angels (0.65) do not differ in their propensity to syndicate.

The analysis of variance rejects the hypothesis of no effects at the maximum sig-

nificance level of 0.00. 22% of the total variance of the VC inve stors' propensity to co-

invest can be explained by its type (see table 43). The contingency table reveals that

Non-Professional- and Foreign VC investors syndicate significantly more deals than

Business Angels, Incubators, Established Specialists, Young Specialists, Specialists,

Public- and Sparkassen VC investors (see table 44). All other direct comparisons be-

tween the average Syndication Ratios of the VC types show no significant differences.

Thus, although there are noticeable differences in the Syndication Ratio of the VC

types, they are not found to be significant except for Non-Professional- and Foreign VC

investors who co-invest up to 60% more than VC types with the lowest Syndication Ra-

tio.

The test of another univariate linear model serves the purpose to reveal which type

of VC investor has on average the most co-investors throughout his investment activi-

Details

Seiten
Erscheinungsform: Originalausgabe
Jahr: 2003
ISBN (eBook): 9783832471491
ISBN (Paperback): 9783838671499
DOI: 10.3239/9783832471491
Dateigröße: 1.5 MB
Sprache: Englisch
Institution / Hochschule: European Business School - Internationale Universität Schloß Reichartshausen Oestrich-Winkel – unbekannt
Erscheinungsdatum: 2003 (September)
Note: 1,0
Schlagworte: syndicat co-investments

Autor

Finn Rieder (Autor:in)