JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

1 of 51

�

Misconduct Risks,

Enforcement & VC Networks

Douglas Cumming

Florida Atlantic University

Simona Zambelli

University of Florence

Ali Mohammadi

Copenhagen Business School

Quality of Law

Investor

Behavior

Enforcement & VC Networks

VC Networks

Culture

Syndication

Networks in VC

Dec. 2021,

Accepted,

European

Financial

Management

2- 1

2 of 51

This paper investigates the impact of law and culture on VC Network. It contributes to the literature on VC financing, by empirically investigating whether the quality of law & legal systems is associated with the density of

VC syndication networks and whether Culture can moderate this relationship.

Introduction

Motivation, Context & Literature
Prior studies & new contribution
Research questions
Hypothesis

Data and Methodology

Results & Discussion

Descriptive Statistics & Multivariate Analysis

Conclusions

Outline

VC Networks

Enforcement & VC Networks

2- 2

3 of 51

This study was inspired by a recent debate on the impact of the quality of law especially in terms of enforcement, and on the way to enforce the law

Motivation (1)

VC Networks

Enforcement & VC Networks

Law in Finance Literature for public markets found a Differential Impact of LAWS & legal institutions depending on:

- the ways in which securities laws are enforced 🡪 emphasizing the role of private enforc. (La Porta et al. 2006)

resources spent by public actors in enforcement emphasizing the role of public enforc. (Roe & Jackson 2009)
unequal costs/benefits associated to enforcement handled by different types of firms, e.g., bigger firms appears to better handle higher disclosure costs associated with private enforcement

(Cumming et al. 2015)

This debate on enforcement for

public market contexts made us

wonder whether and how the quality

of law, type of enforcement of securities laws, matters for PE inv. as well, and whether it may play a differential role /impact for private capital markets as well, such as VC finance.

Q: Does enforcement of securities

laws matter for VC markets and VC

contracting behavior?

SYNDICATION NETWORKS

debate

2- 3

4 of 51

Various comparative law studies (on the quality of law around the world), have been criticized because they do not consider the so called

“Law in Context”. 🡪 Legal systems are in fact subsets of societies and can be fully understood only if the cultural background is considered.

VC Networks

Enforcement & VC Networks

Urgent Call for more empirical research (from the economics and

business field) integrating the analysis of

FORMAL Institutions

INFORMAL Institutions

Motivation (2)

Legal rules

Political rules

Behavioral & Cultural norms

Some evidence shows Culture moderates the impact of formal institutions on:

entrepreneurial opportunities (Fuentelsaz et al. (2019)
development of VC activity (Li and Zahra 2012)

We intend to fill this GAP 🡪 by considering the interplay of Formal - Informal Institutions and focusing on

(This focus is not trivial 🡪 the current evidence on formal & informal

institutions on VC behavior & syndication is inconsistent and inconclusive)

VC BHEAVIOR

SYNDICATION NETWORKS

2- 4

5 of 51

VC & PE

The major source of Finance for innovative firms (not quoted) and new venture. PE can be defined as the provision of capital to innovative firms, through an equity participation (VC is a subset, focused on new ventures)

Enforcement & VC Networks

First, VCs raise money from external investors to create a closed-end fund
Second, they manage the pooled fund (PE fund), selecting the firms to finance, through a detailed due diligence (Cumming and Zambelli, 2017).

INV. CYCLE

EXIT (in 3-7 Y)

PE investors

2- 5

6 of 51

From a legal point of view, a common misconception in economics and finance is to think that the securities laws apply only to public firms .This misconception results in many startup issuers breaking the law unintentionally.

Common Misconception

VC Networks

Enforcement & VC Networks

1) The more relevant and most dreamed divestment routs (Exit) for VCs is the IPO. In the vast majority of cases the IPO is pre-planned since the beginning (the moment of the investment).

2) There are a various number of specific rules and aspects in public and private enforcement that applies directly to public and private firms, regardless of the fact that the portfolio companies will public or not (see Appendix)

Securities laws do apply to publicly traded

and privately held (non-listed) firms

2- 6

7 of 51

Chief of the SEC Enforcement Division

“In 2010, the Commission created 5 specialized units within the Division of Enforcement to address specific areas of the financial markets where specialization would enhance the Commission’s ability to carry out its investor protection mission.

“The Asset Management Unit, is the largest of those units and focuses on investment advisers and investment companies, which of course includes managers of private equity funds.

“It’s worth mentioning that we are national in scope — we have approximately 75 staff across 11 offices covering large metropolitan areas with high concentrations of private equity managers.” Source: Karpati (2013).

Enforcement & VC Networks

2- 7

8 of 51

Forms of PE Misconduct

Type of Fraud	Detailed Description	Examples	Related Research
Inflate the value of illiquid assets	Private equity fund investments range normally between 2-7 years prior to an exit (IPO, acquisition, secondary sale, buyback, write-off). Each year, and sometimes 2-4 times per year, funds report valuations of unexited deals to institutional investors. Those valuations may be exacerbated in order to attract new investment.	Advisors SEC v. Yorkville Advisors, No. 12 Civ. 7728 (S.D.N.Y. filed Oct. 17, 2012). This case involved hedge funds, but the issues involved were similar to those in private equity cases as reported by the SEC https://www.sec.gov/news/speech/2013-spch012313bkhtm. Additional examples include: In re Advanced Equities, Inc., Adm. Proc. File No. 3-15031 (instituted Sept. 18, 2012); In re KCAP Financial, Inc., Adm. Proc. File No. 3-15109 (instituted Nov. 28, 2012).	Cumming and Walz (2010 JIBS); Johan and Zhang (2021 JBE).

2- 8

9 of 51

Forms of PE Misconduct

Type of Fraud	Detailed Description	Examples		Related Research
Conflicts of Interest	Conflicts of interest can arise between fund managers and different investors, products, and investees. Example: fee shirting of organization expenses for co-mingled vehicles, or broken deal expenses, to lower expenses of preferred clients. Example: complementary products used to generate fund commitments to support co-investment vehicles. Example: usurping investment opportunities or entering into related party transactions at the expense of some investors.	In re Pinkas, Adm. Proc. File No. 3-14759 (instituted Feb. 15, 2012); SEC v. Onyx Capital Advisors, LLC, No. 10-cv-11633 (E.D. Mich. filed April 22, 2010); In re Crisp, Adm. Proc. File No. 3-14520 (instituted Aug. 30, 2012).	Atanasov et al. (2012 JF); Cumming and Dai (2010 EFM) in the context of hedge funds.

2- 9

10 of 51

Forms of PE Misconduct

Type of Fraud	Detailed Description	Examples	Related Research
Theft; insider trading	Theft of confidential information by PE investors to illegally profit.	SEC v. Gowrish, No. 09-cv-5883 (N.D. Cal. filed Dec. 16, 2009).	Archarya and Johnson (2010).

2- 10

11 of 51

VC Syndication

Introduction Data Results Conclusions

VC Networks

SYNDICATION

Reasons for syndication

Recent evidence shows that VC syndication seems to be more common in countries with stronger and efficient legal settings. 🡪 But this evidence does not take into account that in weaker legal environment there are less VCs due to lower exit opportunities. Therefore, this evidence is INCONCLUSIVE

It occurs when more VCs jointly invest in the same target company (with binding contracts that bind VCs through financial securities)

Sharing capital requested and risk ;

Sharing knowledge and experience with other VCs, with different skills, and this has positive impact in the entire VC cycle, from screening & due diligence to value adding activities (Dai and Nahata, 2016; Huang et al., 2015; Manigart and Wright, 2013; Trapido 2013)

Firm

Crucial role for VC success (Nahata

et al. 2014; Dai et al., 2016; Brander et al., 2012)

Common characteristics in VC finance

(Dai and Nahata 2015). In particular:

2- 11

12 of 51

Syndication & Culture

Introduction Data Results Conclusions

VC Networks

CULTURE

Also syndication seems more valuable in some cultural settings depending on the level of trust (for example, for low-trust cultural environments; Bottazzi et al., 2016)

Cultural characteristics influences collaboration and learning (Ado et al., 2016). In a more general context of entrepreneurial finance, culture plays a crucial role in promoting new businesses (Nahata et al. 2016; Perry et al. 2015; Perry et al. 2015)

Recent evidence in cross boarder VC investments shows that cultural differences play a crucial role in influencing VC activity. In particular, higher cultural differences are associated with smaller collaborations and lower VC syndication (Dai et al. 2016)

Recent studies show:

Recent evidence: VC syndication activity is influenced by culture (Mauleman et al 2017)

VCs prefer to syndicate with other with the same cultural background (Bird-of-a-feather-flock-together effect; Dai et al.2016) and this, in turn, leads to a reduction in the success of the venture, probably due to the consequent poor decision making driven by high affinity syndicates (Gompers et al. 2016)

2- 12

13 of 51

Syndication & Networks

Introduction Data Results Conclusions

VC Networks

The syndication of VC investments

leads to the establishment of webs of relationships or networks among VCs

NETWORKS

among VC investors, with positive benefits for future deal flows, and transferring of knowledge\

With the adoption of the Network Analysis tools, a VC syndication network can be represented as a network of investors, where:

each node represents a VCst

each link (ties) between 2 nodes represents the coinvestment in the same investee firm.

Therefore, a tie exists between 2 VCs if they have invested in the same company (Hocheberg 2007; Bringman et al. 2016)

vc 1

vc2

target firm

2- 13

14 of 51

Structure of VC Networks

Introduction Data Results Conclusions

VC Networks

In the contexts of organization behavior, network analysis has been applied extensively: networks play a crucial role in learning, sharing, and transferring of knowledge (Ado et al. 2016)- Network Analysis has been applied to different contexts in finance, as well.

For example, in the context of VC, recent literature has focused on the way in which networks are structured, in order to investigate which structural and spatial patterns of syndication networks are most beneficial for portfolio firms (Bringmann 2015).

It has been shown that the structure of a network matters and can affect performance. For example, the particular position inside a network matters for performance, in such a way that more central VC firms obtain better performance in terms of exit routes via IPOs or TS to another company. Also, more centrally located VCs are associated with higher investee performance (Hochberg 2007; Bringmann, 2015; Chemmanur 2016)

Also, from a VC context, literature has shown that the density of networks matters and and highly dense networks create entry barrier (they seem to limit the entry into a local market (Hocheberg et al., 2010).

2- 14

15 of 51

Gap & Contribution

Is it possible that legal institutions play a differing role depending on the different benefits and costs associated with the different type of law enforcement and different cultural settings?

While the impact of legal environment on finance and growth is very well documented in the literature, no previous study has investigated the link between VC networks & institutional environments.

We contribute to filling this gap in the context of VC, by empirically investigating, on a worldwide scale, the link between:

Introduction Data Results Conclusions

VC Networks

VC syndicate networks

Legal settings

Prior studies are inconclusive on the impact of legal instit. on VC syndication, and have never investigated whether or not VC syndicate networks are positively or negatively related to a country’s legal condition (especially with enforcement).

and how VC networks changes

in response to changes:

- in legal environments

- and cultural settings;

Our paper is the first step

toward this direction

no previous study

2- 15

16 of 51

What we investigate

Introduction Data Results Conclusions

VC Networks

Different types of Enforcement

networks

different cultural settings

culture

can affect or moderate this link

between law and networks

VC networks

density

Impact of Quality of legal system & Enforcement of law

Public

Private

considering if and how

In particular, for the first time, in our study we consider the differential role played by law enforcement, in the context of VC markets

Formal Inst.

Informal Inst.

2- 16

17 of 51

Types of Enforcement

Introduction Data Results Conclusions

VC Networks

PUBLIC ENFORCEMENT

External supervisors or Public Agent (e.g., public regulators
or public Authorities) with the power to:
Impose rules
Investigate (investigative power)
Inflict sanctions (criminal sanctions in case of rule breaches)
Impose Orders (e.g. “stop and do” or “refrain” from a range of activities

Private parties (through private litigation aimed at protecting the rights of a contractual party 🡪final aim: receiving compensation or a contract rescission). Strong effective Enforcement, involves
Disclosure requirements standards (regarding for example the ownership structure, directors compensation, equity structure…)
Liability rules
Burden of proofs in private lawsuits

PRIVATE ENFORCEMENT

2- 17

18 of 51

Previous EVIDENCE on �Enforcement & its impact

The impact of law enforcement has been addressed for public firms. For example, the evidence on public markets shows that:

No studies has investigated its impact in the context of private firms. It is not clear how the quality of enforcement (public and private enforcement) affect private markets, such as VC, PE markets

Introduction Data Results Conclusions

VC Networks

PUBLIC ENFORCEMENT & IPO

PRIVATE ENFORCEMENT & SIZE

Strong public enforcement of securities

laws discourages or diminish the probability of IPOs IPOs (La Porta et al. 2006)

Private enforcement is much more beneficial to large firms (rather than small firms) 🡪 Larger firms are better equipped to handle the costs of higher disclosure, burden of proofs, and liability standards related to private enforcement (Jackson&Roe, 2009; Cumming et al.2015)

…No studies…

2- 18

19 of 51

Research Questions

Is there a differential effect of Public vs Private enforcement on VC behavior and on the ways they carry on their activity, especially with reference to syndication and VC network formation? Is it possible that private vs public enforcement facilitates, in a differential way, the formation of networks among VCs

Private enforcement (disclosure standards and burden of proof rules)?
Public enforcement (investigative powers & sanctions by public regulators)?
or None?

in direct influencing VC behavior and VC network formation?
in moderating the relationship between law and VC networks?

Introduction Data Results Conclusions

VC Networks

Q: Which matters more for VC markets?

Q: Is there a differential effect of Enforcement on VC networks?

Q: Does culture matter or moderate the impact of law on VC networks?

2- 19

20 of 51

HYPOTHESES

Introduction Data Results Conclusions

VC Networks

2- 20

21 of 51

Hypotheses: 1^st Set

Guided by the law &finance literature we developed two sets of Hypotheses

We expect a differential role of public and private enforcement on Network formation

Introduction Data Results Conclusions

VC Networks

1^st Set

ENFORCEMENT IMPACT

HIGH PUBLIC ENFORCEMENT with strong investigative powers against any member of a syndicate, and strong group sanctions, increases the cost of VC syndication net. 🡪 it will discourage the formation of VC networks due to the fact that the high investigative power and sanctions can be easily extended to any member in the syndicate network

PUBLIC ENFORCEMENT

PRIVATE ENFORCEMENT

on VC syndic. networks

HIGH PRIVATE ENFORCEMENT with strong disclosure requirements and liability standards will encourage the formation of VC syndicate networks via: clarification of liability rules, standardizing contracts, and cost sharing amongst VCs belonging in the syndicate. High disclosure requirements are better handled by larger firms and by a network

2- 21

22 of 51

The level of acceptance of risky and uncertain situations. High scores on this dimension represent more risk-adverse systems. Low scores reflect more risk lover systems.

Uncertainty Avoidance

(Risk Aversion)

Min

Max

Risk- Adverse

Risk- Lover

The level of acceptance of inequalities among people. High scores represent more hierarchical systems and inequalities. It also measures the willingness of people to accept hierarchical differences

Power Distance

Min

Max

High Inequality

Equality

Cultural Impact – Hofstede

Min

Max

High Masculinity

Low Masc.

Min

Max

High

Individualism

Collectivism

Masculinity reflects the level of competitiveness, and the focus on achievement of materialistic goals among people. High scores reflects highly competitive aggressive systems (🡪 low feminine society). The opposite, or Low scores, reflects high Femininity: the tendency to focus on cooperation, relationships, quality of life, caring of others (High scores of this index reflects high differences in gender discrimination).

Masculinity

Individualism

Individualism reflects the importance that a system gives to individual achievements and results. High scores reflects highly individualistic and selfish systems. In contrast, low scores reflect collectivism systems promoting collective actions.

Introduction Data Results Conclusions

Low

Femininity

2- 22

23 of 51

Introduction Data Results Conclusions

VC Networks

CULTURE EFFECT

on VC syndicate

networks

encourage the formation of VC syndicate networks. In more Risk Adverse contests, we expect a higher propensity to syndicate and, as such, more dense VC networks, as they act as a form of insurance of the syndicate members

discourage the formation of syndicates and as such of VC networks. We expect the same

in higher power distance (with higher inequalities) societies, and in

higher masculinity contexts (which are more competitive - and less prone to cooperation). In these contexts, we expect to see less intense VC network formation

2^nd Set

HIGH RISK AVERSE contexts

HIGH INDIVIDUALISM contexts

Hypotheses: 2^nd Set

HIGH MASCULINITY contexts

2- 23

24 of 51

Hypotheses Summary

Public Enforcement

Private enforcement

discourages network density

facilitates network formation

High Risk contexts

High Individualism / High Masculinity

facilitates network formation

discourages network formation

2- 24

25 of 51

Introduction

Motivation
Prior studies & new contribution
Research questions
Hypothesis

Data and Methodology

Results & Discussion

Descriptive Statistics & Multivariate Analysis

Conclusions

Outline

VC Networks

Enforcement & VC Networks

2- 25

26 of 51

Data and Key Variable

From Thompson SDC we collected a worldwide dataset covering VC transactions carried out in 31 countries from: 1995 2014

The dataset includes 100,018 worldwide VC investments in 55,621 investee firms and involves 201,552 relationships, or VC-investee firm deals.

To form VC network observations we need 5 years of data (Hochberg, et al., 2007). So, from a network perspective, our time period is

2000 2014,

Unit of observation: [country – year] 🡪 in our data, each [country – year] represents an observation 🡪 15 years (for 31 countries) = 465 observations (-) 2 (country-year) where we do not have observations = 463 observations

Our main dependent variable is VC’s network density, calculated as the ratio between the number of existing connections divided by the maximum possible number of connections (see, e.g., Friedkin, 1981; Hochberg, et al., 2010). This measure indicates how well connected VCs are in terms of nodes (or agents) in their real network relative to the theoretical number of all possible connections.

Introduction Data Results Conclusions

VC Networks

2- 26

27 of 51

Key variables

Introduction Data Results Conclusions

VC Networks

Enforcement

VC Network density

cultural settings

across countries & over time

How network density

Public

Private

varies

regulatory environments

culture

contexts

with different

Key dependent variable

+ Other Control var.: other legal indexes; size of networks in different countries; size of economies; presence of foreign investors; VC industry characteristics; country’s investment profile safety

Key independent variables

To investigate

This measure indicates how well connected VCs are

in their real network relative to the theoretical number

of all possible connections.

2- 27

28 of 51

Network Density Example

Introduction Data Results Conclusions

VC Networks

In line with Hochberg, et al., 2010

it is calculated as the % ratio between

Network Density is the most commonly reported property of a network. It is a standard measure generally adopted when comparing networks against each other (especially when comparing networks in different countries). It captures both syndication and the diversity of syndicate partners (Friedkin 1981)

It shows how nearly a network is “complete” (context in which each member is connected directly with every other member and reflects the strengths of group cohesion)

Hochberg et al. 2010; Bygrave 1988

2- 28

29 of 51

Low - High Density

Figure - Low density network

(with 8 VCs), characterized by less connected members

Figure - High density network

(with 8 VCs) characterized by a

higher number of connected members

Introduction Data Results Conclusions

VC Networks

The maximum density of a network is reached when all members are connected to

each other (theoretical situation).

2- 29

30 of 51

Introduction

Motivation
Prior studies & new contribution
Research questions
Hypothesis

Data and Methodology

Results & Discussion

Descriptive Statistics & Multivariate Analysis

Conclusions

Outline

VC Networks

Enforcement & VC Networks

2- 30

31 of 51

Impact of Enforcement

Difference test statistics -0.31***

Difference test statistics -0.14***

Introduction Data Results Conclusions

VC Networks

Impact of Public Enforcement

Settings characterized by

HIGHER Private Enforcement

🡪 HIGHER network density

For Settings characterized by HIGHER Public Enforcement

🡪 LOWER network formation

Impact of Private Enforcement

When we split our sample 2 sub-groups (according to their level of LOW vs HIGH public - private enforc. for being above or below the median), both public & private enforc. are negatively associated with syndicate network

(🡪 partially supporting our Hypotheses)

√

2- 31

32 of 51

Trend Impact of Public Enforcement on Network Density

Full sample

The relationship between public enforcement & network density remains strongly negative, with or without a few outliers

Introduction Data Results Conclusions

√

2- 32

33 of 51

Introduction Data Results Conclusions

The Puzzling Trend :

Impact of Private Enforcement on Network Density

Full sample (with outliers)

weakly negative relationship

BUT …. Without a few outliers

the relationship changes

cultural settings may

explain this puzzling pattern

maybe, it is plausible that cultural differences play a moderating role and it is plausible that legal standards are “dominant” when culture is not so dominant and viceversa

√

2- 33

34 of 51

Role of Cultural Differences

Difference test statistics 0.196***

Introduction Data Results Conclusions

VC Networks

The same pattern is observed for higher masculinity (or low femininity), but the

diff is not stat significant

Impact of Risk aversion

Observations with HIGHER RISK AVERSION

🡪 facilitate VC networks density

Impact of individualism

Observations with HIGHER INDIVIDUALISM

lower sense of cooperation
lower VC networks density

2- 34

35 of 51

Introduction

Motivation
Prior studies & new contribution
Research questions
Hypothesis

Data and Methodology

Results & Discussion

Descriptive Statistics & Multivariate Analysis

Conclusions

Outline

VC Networks

Enforcement & VC Networks

2- 35

36 of 51

Dependent Variable:
Network Density (%)	8	9	10	11	12	13
Public enforcement	-1.023***	-0.927**	-1.004***	-1.042***	-1.045**
	-0.331	-0.385	-0.292	-0.276	-0.463
Private enforcement	0.172	0.028	0.277	0.144		-0.987*
	-0.365	-0.457	-0.408	-0.347		-0.595
Anti-self-dealing			-0.206
			-0.282
Anti-director-rights			0.048	0.042
			-0.093	-0.096
Efficiency of judiciary			-0.099	-0.1
			-0.077	-0.078
Investment profile		0.055	0.083	0.081
		-0.041	-0.056	-0.056
Cluster size	0.657	0.572*	0.609*	0.613*	0.630***	0.731***
	-0.405	-0.345	-0.359	-0.359	-0.199	-0.256
Share of foreign VCs	0.301	0.221	-0.027	-0.058	-1.071	-2.038
	-0.897	-0.836	-0.911	-0.93	-1.439	-1.601
Geographic diversity of ventures	-0.089	-0.061	-0.113	-0.123*	-0.081**	-0.074
	-0.062	-0.049	-0.074	-0.075	-0.04	-0.062
Share of CVC deals	0.19	0.396	0.922	0.953	0.385	1.494
	-1.03	-0.956	-0.841	-0.822	-0.953	-1.709
GDP (log)	-0.223***	-0.220***	-0.301***	-0.289***	-0.241***	-0.204*
	-0.069	-0.072	-0.102	-0.101	-0.087	-0.109
No. VC firms	0.011	0.01	0.018**	0.017*	0.012	0.007
	-0.007	-0.007	-0.009	-0.009	-0.01	-0.011
Financial crisis	-0.011	-0.043	-0.087	-0.09	-0.045	-0.032
	-0.071	-0.081	-0.095	-0.096	-0.09	-0.096
Constant	1.966***	1.546***	2.266***	2.206***	2.404***	2.235***
	-0.432	-0.508	-0.833	-0.82	-0.792	-0.815
Year Fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Country Fixed Effect	No	No	No	No	No	No
N	463	463	463	463	463	463
No. countries	31	31	31	31	31	31
R2 within	0.223	0.237	0.24	0.239
R2 overall	0.311	0.302	0.345	0.344
R2 between	0.475	0.424	0.541	0.539
chi2	1018.304	1156.811	2361.493	1490.72	1435.897	715.499
p	0	0	0	0	0	0
First stage F-statistics					8.0672	23.1595
First stage r-square					0.3941	0.5676

Negative and significant association

between Public enforcement and

Network Density

2- 36

37 of 51

Economic Significance

The economic significance is such that a 1-standard deviation increase in public enforcement is associated with a 25% reduction in density relative to the average density in the data.

Also, the data show a substantial improvement in goodness of fit for overall (from 23.7% to 30.8%) and between (from 26.8% to 46.1%) by adding these variables (although not within, due to the lack of time series variation).

Robust to IV estimates (legal origin instrument)

Introduction Data Results Conclusions

VC Networks

2- 37

38 of 51

Dependent Variable:
Network Density (%)	8	9	10	11	12	13
Public enforcement	-1.023***	-0.927**	-1.004***	-1.042***	-1.045**
	-0.331	-0.385	-0.292	-0.276	-0.463
Private enforcement	0.172	0.028	0.277	0.144		-0.987*
	-0.365	-0.457	-0.408	-0.347		-0.595
Anti-self-dealing			-0.206
			-0.282
Anti-director-rights			0.048	0.042
			-0.093	-0.096
Efficiency of judiciary			-0.099	-0.1
			-0.077	-0.078
Investment profile		0.055	0.083	0.081
		-0.041	-0.056	-0.056
Cluster size	0.657	0.572*	0.609*	0.613*	0.630***	0.731***
	-0.405	-0.345	-0.359	-0.359	-0.199	-0.256
Share of foreign VCs	0.301	0.221	-0.027	-0.058	-1.071	-2.038
	-0.897	-0.836	-0.911	-0.93	-1.439	-1.601
Geographic diversity of ventures	-0.089	-0.061	-0.113	-0.123*	-0.081**	-0.074
	-0.062	-0.049	-0.074	-0.075	-0.04	-0.062
Share of CVC deals	0.19	0.396	0.922	0.953	0.385	1.494
	-1.03	-0.956	-0.841	-0.822	-0.953	-1.709
GDP (log)	-0.223***	-0.220***	-0.301***	-0.289***	-0.241***	-0.204*
	-0.069	-0.072	-0.102	-0.101	-0.087	-0.109
No. VC firms	0.011	0.01	0.018**	0.017*	0.012	0.007
	-0.007	-0.007	-0.009	-0.009	-0.01	-0.011
Financial crisis	-0.011	-0.043	-0.087	-0.09	-0.045	-0.032
	-0.071	-0.081	-0.095	-0.096	-0.09	-0.096
Constant	1.966***	1.546***	2.266***	2.206***	2.404***	2.235***
	-0.432	-0.508	-0.833	-0.82	-0.792	-0.815
Year Fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Country Fixed Effect	No	No	No	No	No	No
N	463	463	463	463	463	463
No. countries	31	31	31	31	31	31
R2 within	0.223	0.237	0.24	0.239
R2 overall	0.311	0.302	0.345	0.344
R2 between	0.475	0.424	0.541	0.539
chi2	1018.304	1156.811	2361.493	1490.72	1435.897	715.499
p	0	0	0	0	0	0
First stage F-statistics					8.0672	23.1595
First stage r-square					0.3941	0.5676

No impact of private enforcement

on Network Density, as well as for

other legal indexes

2- 38

39 of 51

TAB 7 IMPACT ON DENSITY - subsamples	Excluding US				Excluding US and UK
excluding US & UK	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Public enforcement		-0.956***		-1.023***		-0.898***		-0.977***
		(0.273)		(0.341)		(0.276)		(0.343)
Private enforcement			-0.329	0.207			-0.239	0.252
			(0.318)	(0.353)			(0.310)	(0.343)
Cluster size	0.670*	0.661*	0.676*	0.656*	0.660*	0.656*	0.666*	0.650*
	(0.390)	(0.385)	(0.393)	(0.389)	(0.384)	(0.379)	(0.388)	(0.384)
Share of foreign VCs	0.409	0.625	0.423	0.657	0.430	0.630	0.445	0.669
	(0.605)	(0.587)	(0.606)	(0.581)	(0.609)	(0.591)	(0.608)	(0.585)
Geographic diversity of ventures	-0.113	-0.090	-0.105	-0.092	-0.128	-0.102	-0.121	-0.104
	(0.083)	(0.083)	(0.081)	(0.081)	(0.090)	(0.093)	(0.088)	(0.092)
Share of CVC deals	0.417	0.302	0.526	0.204	0.448	0.320	0.524	0.198
	(1.180)	(1.130)	(1.198)	(1.142)	(1.191)	(1.147)	(1.213)	(1.168)
GDP (log)	-0.140	-0.215***	-0.151	-0.213***	-0.120	-0.197**	-0.128	-0.194***
	(0.106)	(0.079)	(0.116)	(0.076)	(0.103)	(0.078)	(0.111)	(0.074)
No. VC firms	-0.004	0.012	0.002	0.011	0.010	0.009	0.011	0.010
	(0.010)	(0.009)	(0.011)	(0.008)	(0.009)	(0.011)	(0.009)	(0.008)
Financial crisis	-0.023	-0.017	-0.017	-0.020	-0.030	-0.024	-0.025	-0.027
	(0.081)	(0.079)	(0.079)	(0.078)	(0.084)	(0.083)	(0.082)	(0.082)
Constant	1.038*	1.931***	1.279*	1.839***	0.916	1.793***	1.094	1.674***
	(0.624)	(0.558)	(0.716)	(0.497)	(0.609)	(0.566)	(0.683)	(0.485)
Year Fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	448	448	448	448	433	433	433	433
No. countries	30	30	30	30	29	29	29	29
R2 within	0.224	0.228	0.226	0.228	0.228	0.231	0.229	0.231
R2 overall	0.239	0.297	0.240	0.298	0.235	0.287	0.235	0.289
R2 between	0.282	0.430	0.277	0.436	0.262	0.405	0.259	0.411

When we exclude US (and US + UK), results remain confirmed:

Negative and significant association between Public enforcement and

Network Density; but NO significant IMPACT of private eqnforcement

2- 39

40 of 51

Difference-in-Differences

We also implemented difference-in-differences tests of countries that had regulatory changes (Brazil, India and South Korea, as identified in Cumming and Knill, 2012) in regards to improvement in public governance standards (improvements in public enforcement) and compared these countries (Treat Group) with other countries which did not experienced regulatory changes (Control Group). We compared the VC network density of treated countries pre and post treatment with VC network density of match control group (Japan, Argentina, Indonesia).
Countries with improved laws show substantial reductions in network density in the post-regulatory change period (from 1.23 on average to 0.69), and the difference is statistically significant at the 1% level. The control sample shows no material change in network density.

Introduction Data Results Conclusions

VC Networks

Control group:

Argentina, Indonesia, Japan

* we use 3 specifications for control gr.

a) Argentina, Indonesia, Japan

b) All the other countries except US

c) All the other countries

Treat group:

Brazil, India, South Korea

🡪 Countries that experienced an Improvement in public enforcement (Cumming and Knill, 2012)

2- 40

41 of 51

Countries with improved laws show substantial

reductions in network density in the post-regulatory change

period and the difference is statistically significant at the 1% level. The control sample shows no material change in network density.

2- 41

42 of 51

CULTURE

Introduction Data Results Conclusions

VC Networks

Split our data

as explanatory variables

Direct Impact of cultural differences

to see which one matters the most, or has the most pronounced impact on VC networks density

(models 1-4, Tab 9)

dividing our data according to the different cultural characteristics of countries

(models 5-12, Tab 9)

Sub-samples

Moderating Impact of cultural differences

We test 2 sets or regressions

2- 42

43 of 51

Data remain confirmed with reference to public enforcement impact 🡪 Negative and signif. association between public enforcement and Network Density

Data indicate a positive and signif. association between

Risk aversion and Net.Density -- investors in countries with high uncertainty avoidance are more likely to form stronger networks to mitigate the risk of operating alone,

consistent with Hypothesis 3.

2- 43

44 of 51

Economic Significance of Culture�TAB 9 models 1-4

The economic significance is such that a 1-standard deviation increase in uncertainty avoidance is associated with a 25.5% and 20.4% increase in density in Model 1 and Model 4, respectively, for the least and most conservative estimates.

……. The impact is as strong as the impact of enforcement

Introduction Data Results Conclusions

2- 44

45 of 51

Sub-samples culture

When we distinguish our sample into sub-groups of countries according to the cultural characteristics, we do find that

BOTH

Introduction Data Results Conclusions

PUBLIC ENFORCEMENT

PRIVATE ENFORCEMENT

matter

…But….only for some cultural

settings

in influencing network density and we find support for our two main hypotheses

2- 45

46 of 51

The impact of public enforcement on network density remain confirmed with a negative and significant association, for

LOW uncertainty avoidance (Low risk aversion)
High masculinity (LOW femininity)
LOW power distance

2- 46

47 of 51

Private enforcement becomes significant 🡪 with a positive and significant association, contributing to facilitate syndicate network formation (in line with our H2) but only in certain cultural settings:

- LOW risk aversion

- high masculinity (LOW femininity)

- LOW level of power distance

2- 47

48 of 51

Overall, both effects of BOTH private and public enforcement are

in line with our main hypotheses, but only for certain cultural settings:

LOW risk aversion societies,
and higher masculine countries (LOW femininity)
LOW power distance societies

in these cultural settings, LAW MATTERS MORE (LEGAL STANDARDS appear to matter more in driving Network Density)

2- 48

49 of 51

While in other contexts with STRONG CULTURAL TRAITS:

HIGH risk aversion
HIGH femininity
HIGH power distance
LEGAL STANDARDS are never significant and
Law appears to matter less in driving network density, probably because CULTURE PLAYS a more dominant impact… and this calls for more research on the field

2- 49

50 of 51

Conclusions 1

Introduction Data Results Conclusions

VC Networks

Differential role in the impact of Public and Private Enforcement

culture

discourages network formation

(in full sample and sub-samples)

VC behavior in their propensity to form denser networks

Law & Enforcement of law

Public Enforcement

Private enforcement

also considering

if and how

discourages network formation

facilitates network formation

Public Enforcement

…But…

only for some cultural settings

Results:

√

moderates

the relationship

In this paper

√

2- 50

51 of 51

Conclusions 2

Introduction Data Results Conclusions

VC Networks

Differential role in the impact of public and private Enforcement

networks

Lower risk aversion

Lower power distance

only for some cultural settings

When we take into account subgroups of countries according to cultural characteristics we do find support for both our main hypotheses, as we find a differential role in the impact on network density played by public and private enforcement, in line with our expectations ….. but… this occurs only for certain cultural settings

Lawer Femininity

The Impact of both public and private

enforcement is

significant only for settings with:

For these cultural settings, legal standards appears to matter more

and where there are strong cultural traits, law appears to play a weaker impact

🡪 our findings call for more research on the impact of cultural differences

2- 51