Master Thesis Financial Management

Transcription

1 Master Thesis Financial Management Fundraising in Private Equity: What is the influence of Limited Partners characteristics and investment criteria on the probability and intensity of investing in First Time Funds? M.C.J. van Brussel December 16, 2010 Supervisor: Professor Marco Da Rin

2 Faculty School of Economics and Management Programme MSc. Financial Management Name M.C.J. van Brussel ANR Topic Fundraising in Private Equity: What is the influence of Limited Partners characteristics and investment criteria on the probability and intensity of investing in First Time Funds? Date December 16, 2010 Supervisor Professor Marco Da Rin Second reader Assistant Professor Paul Sengmuller 1

3 Preface My final dissertation is a result of a global and unique research study, lead by Professors Marco Da Rin and Ludovic Phalippou, on how Limited Partners perform their due diligence in Private Equity. Between September 2010 and December 2010 I have been given the opportunity to be part of this challenging project for which I am very grateful. I mainly have been working on data collection, which was a great way to develop cold calling skills. I would like to thank Professor Marco Da Rin for having me on board, supporting me in the calling, advising me in writing the thesis and above all giving me the opportunity to graduate before Christmas so that I can have a fresh start with my new job in the beginning of December 2010, M. C.J. van Brussel 2

4 Executive summary Fundraising in Private Equity is a topic not widely investigated. This study aims at explaining what Limited Partners characteristics comprising of Limited Partner s vintage year and size, and investment criteria comprising of variables related to General Partner s experience, Investment Committee and free riding, are important in the decision of investing in First Time Funds. The focus of this study is specifically on First Time Funds, generally it is harder for these kinds of funds to raise capital resulting from the lack of a track record and associated higher risk perception. I answer the research question by using a global and unique dataset of 288 Limited Partner investors in Private Equity. I found that both Limited Partners characteristics and investment criteria influence the probability and investing in First Time Funds. All hypotheses can be partially confirmed, a larger sample might have further increased the explanatory power. Table 1. Summarized results Hypothesis Result 1. Limited Partner characteristics Larger Limited Partners are more likely to invest in First Time Funds. Partially confirmed 2. More experienced Limited Partners are more likely to invest in First Time Funds. Partially confirmed Experience 3. Limited Partners are more likely to invest in First Time Funds when the Partially confirmed General Partner is highly experienced. Investment committee 4. Limited Partners are more likely to invest in First Time Funds when the Partially confirmed power of the Investment Committee in the investment decision is limited. Free riding 5. Limited Partners are more likely to invest in First Time Funds when they have the opportunity to free ride on the advisors' opinions or investment decisions of third parties. Partially confirmed My findings indicate that larger and more experienced Limited Partners are more likely to invest in First Time Funds. This result proves the underlying rationale, that when the amount of assets under management and/or the number of years active in the industry increases, the likelihood of investing in more risky First Time Funds improves. Furthermore, strong evidence can be found on the importance of the General Partner s experience which is mainly driven by the reputation of the Limited Partner. This is in line with the motivation that Limited Partner s tend to look at the General Partner s personal track record in the absence of a track record of previous funds. For both the importance of the composition and functioning of the Investment Committee and opportunities to free ride, some evidence is found though not as convincing as the evidence on Experience. 3

5 Table of contents Executive summary 3 Introduction 6 Part 1: About Private Equity 7 General 7 Types of funds 7 How does it work? 8 Conflicts of interest 9 Types of investors 10 Risk and reward 11 Part 2: Dataset 12 Part 3: Research question and hypotheses 13 Research question 13 Limited Partner characteristics 17 Limited Partner investment criteria 19 Part 4: Empirical analysis 24 Data analysis 24 Models and expectations 25 Statistical output 27 Part 5: Results 32 General overview 32 Probability 32 Intensity 34 Results hypotheses tests 36 4

6 Conclusion 37 Answering the main research question 37 Limitations 38 References 39 Appendix 42 Appendix 1: Sample description 42 Appendix 2: Reasons to invest and not to invest 44 Appendix 3: Log transformations 45 Appendix 4: Survey 47 List of figures and tables Figure 1: Value chain 8 Figure 2: Cash flow streams in Private Equity 9 Figure 3: Dispersion Limited Partner types 19 Figure 4: Log transformation Size Limited Partners (Assets under Management) 45 Figure 5: Log transformation Limited Partner s Experience (LP Vintage Year) 45 Figure 6: Log transformation Fraction (First Time Funds over all Fund investments) 46 Table 1: Summarized results 3 Table 2: Variables overview 15 Table 3: Descriptive statistics 27 Table 4: Pair wise correlation matrix independent variables 28 Table 5: Logit regression on the Probability of investing in First Time Funds (FTF) 29 Table 6: OLS regression on the intensity of investing in First Time Funds (FTF) 30 Table 7: Qualitative reasons to invest in First Time Funds (FTF) 31 Table 8: Summary on probability 34 Table 9: Summary on intensity 35 Table 10: Outcome Hypotheses Tests 36 Table 11: Sample coverage per Country 42 Table 12: Key characteristics sample 43 5

7 Introduction This thesis is part of a global and unique study that aims at gathering a detailed insight in how Limited Partners perform their due diligence in Private Equity. The analysis takes place on the level of the relation between Limited Partners and General Partners and not so much on the widely investigated relation between General Partners and portfolio companies. Research in Private Equity has been largely focused on trends in Private Equity, the relation between General Partners and entrepreneurs but not so much on fund performance (Kaplan and Schoar, 2005). Most of the little available research tends to focus on fund performance and Follow up Funds and not so much on First Time Funds and what factors actually underlie the investment decision. In this thesis I will investigate which Limited Partners characteristics and which investment criteria influence the decision to invest in First Time Funds. The nature of the organization (type) is included as a control variable. A clarification of the terminology and variables used can be found under Part 3, Table 1. I will both look into both the probability of investing in a First Time Fund and the intensity of doing so. For the analysis I will use the project s dataset based on a questionnaire including various topics regarding how Limited Partners perform their due diligence in Private Equity. The topics are general firm information, allocation of assets, access to funds, co-investments, Investment Committee, HR policy, contracting, due diligence in general, due diligence of funds and monitoring. The contribution of my thesis is that it provides an initial overview of what is important when Limited Partners consider investing in a First Time Fund. Whereas the reasons to invest in Follow up Funds are quite straightforward, i.e. one generally looks at the historical performance by checking for example the Internal Rate of Return (IRR) of the previous funds; it is more difficult for First Time Funds as a track record is not available, hence I attempt to find out what then influences the investment decision. 6

8 Part 1: About Private Equity In this Part, an introduction into Private Equity will be provided. An insight will be given to Private Equity in general, the types of funds, how it works, conflicts of interest, types of investors and risk and reward. This part is meant to provide basic knowledge about Private Equity to build upon in the analysis and to answer the research question. General Private Equity is a form of risk capital that is used in a wide range of transaction for the acquisition of everything between start ups and large quoted companies. A fund generally has no leverage, though the investments done by the fund are generally highly levered (Gilligan, Wright, 2010). I.e. Private Equity investments or Management Buyouts are for a small part financed with the funds provided by Limited Partners but the majority is financed with either ordinary debt or mezzanine financing, which is a hybrid form of debt. Types of funds There are two types of equity investment funds, namely Private Equity funds and Quoted Equity funds. We find that by the end of the 1990s many of the captive Private Equity funds have disappeared. Private Equity funds, which in contrast to the Quoted Equity funds, do not invest in public equity, pursue a different strategy. The fund managers of Private Equity funds watch their investments more closely and perform extensive due diligence before they actually invest in a portfolio company. Within the category of Private Equity funds we distinguish between Venture Capital funds which focus on startup companies and Buyout funds which focus on more mature businesses (Gilligan, Wright, 2010). There are two types of Buyouts; to be more precise, Ecbko and Thorburn (2008) define a Leveraged Buy Out (LBO), a transaction undertaken by a Buyout fund, as the acquisition and delisting of an entire company or a division, financed primarily with debt, the buyer is typically a Private Equity fund managed by an LBO sponsor or sometimes a consortium of funds. A Management Buy Out (MBO) is a transaction in which the incumbent management is part of the group that purchases the firm (Gilligan and Wright, 2010). Metrick and Yasuda (2009), state that about $ 1 trillion of capital is invested in Private Equity globally, of which circa one-third is managed by Venture Capital funds. 7

9 How does it work? A Private Equity fund is a kind of investment club to which Limited Partners like pension funds, insurance companies, corporates, banks and family offices allocate their assets. Note that to enhance credibility, the General Partners invest a small percentage of the capital committed as well. The profits that the fund hopes to reap generally come from capital gains from the sale of portfolio companies after several years, i.e. the core focus is not on dividends and other types of short term returns (Gilligan and Wright, 2010). Kaplan and Schoar (2005) add to this that funds have a finite life and the period in which capital is committed to the funds is generally about five years. After 10 to 12 years in total, the Limited Partners will have their capital returned. When a substantial part of the committed capital is invested, General Partners generally strive towards retrieving new capital for a separate and Follow up Fund. In short as introduced by Gilligan and Wright (2010), the four main roles of a General Partner are to: - raise funds from investors (Limited Partners) - source investment opportunities and make investments - actively manage investments - realize capital gains by selling or floating those investments. Figure 1 shows the value chain of Private Equity; Figure 2 adds to this the cash flow streams between the stakeholders. Figure 1: Value chain Source: Kleinschmidt (2006) 8

10 Figure 2: Cash flow streams in Private Equity Source: Gilligan and Wright ( 2010) Conflicts of interest As is the case in any standard Principal-Agent relation, conflicts of interest and information asymmetry might arise between the Limited Partners and the General Partner. Kleinschmidt (2006), quotes several risk reducing actions and instruments which can reduce the risk in the relation between the Limited Partners and the General Partner: Covenants can limit the decision freedom of the General Partner in regard to for example portfolio diversification. Additionally, the compensation of the General Partner can be linked to the actual success of the investments done. To zoom into how compensation is structured, Kleinschmidt (2006) finds that: The fact that returns after fees are generally lower compared to returns in public stock markets feeds a discussion on fees charged implying the need for well defined compensation contracts without discrepancy between actual and expected fees. Furthermore, managers can have an incentive to liquidate good investments immediately, reinvest and increase assets under management and hence also fees since these are based on asset not equity value. Additionally, it is never stated in advance how much fees will be charged, generally the better the fund performance, the higher the fees will be. 9

11 Phalippou (2009) identifies four types of fees: - management fee: annual percent of the capital committed - carried interest: incentive fee on actual returns based on the committed capital plus write-downs of unrealized investments and the expenses on realized investments including management fees - portfolio company fees: transaction and monitoring fees which are paid directly from the portfolio companies - other fees: other costs or fees like for example the cost of keeping up cash proceeds. As it appears that on average half of the committed capital is actually invested, it should be clear whether the management fee will be paid over capital committed or invested. Concerning the carried interest it should be clear how this will be calculated (Phalippou, 2009). Phalippou (2009) identifies three other factors that can further aggravate conflicts of interest: - managers tend to time cash flows such that they will increase incentive fees - certain contracts provide steep incentives in case of short investment periods - transaction fees might alter decisions of the buyout firms concerning among others the level of leverage and the size of the investment. Types of investors Kleinschmidt (2006) distinguishes several categories of Limited Partners: - institutional investors: banks, pension funds and insurance companies. They are the largest suppliers of funds in Private Equity, though for them Private Equity investments are generally not core, but seen as a supplement to their other less risky portfolio investments - public authorities: these governmental institutions invest on the hand in Private Equity, especially in Venture Capital to stimulate economic development, and on the other hand in more mature growth companies - companies: generally supply funds to General Partners who invest in related industries, from a strategic perspective. Kaplan and Strömberg (2009) add to this the finding that 10

12 nowadays Private Equity firms tend to organize themselves around specific industries of interest. - private investors: this group of investors is rather small, generally because the risks involved are so high that individuals are not willing to bear them. In contrast, some wealthy individuals called business angels, invest directly in high growth companies - Fund of Funds: act as a second intermediary between the portfolio companies and the capital providers. I.e. Fund of Funds raise capital themselves, which is consecutively invested into a General Partner. Risk and reward As stated by Ick (2005), how interesting a Private Equity investment opportunity is depends on the performance compared with public markets. Previous literature focused on both fund level and investment level returns measured with several different methods, the results are mixed. Eckbo and Thorburn (2008) report on estimating fund performance measured against the Fama and French (1993) three-factor model, they find fund excess returns (alphas) that are insignificantly different from zero. Kaplan and Schoar (2005) find that on average fund returns are comparable to the return of the S&P500. Ljungqvist and Richardson (2003) find that on average fund returns are 5% higher than the market in the long term. Phalippou (2009) states that returns after fees are generally lower compared to returns in public stock markets. Cochrane (2004) measures the mean, standard deviation, alpha and beta of venture capital investments using a maximum likelihood estimate that corrects for selection bias. The central question is whether Venture Capital investments behave the same as publicly traded securities. He finds that mean returns, alphas and betas all tend to decline steadily from first to fourth round investments, the idiosyncratic variance remains similar. Venture capital investments are like options, they have a rather small chance to gain a large payoff. These insights on the expected return on investments and its volatility are an important factor in the decision whether or not to invest in unpredictive and riskier First Time Funds, but also in Private Equity in general. 11

13 Part 2: Dataset In this Part I will describe the dataset and the survey used to collect the data. In 2009 the project on how Limited Partners perform their due diligence, initiated by Professor Marco Da Rin and Professor Ludovic Phalippou started. Between September and December 2010 I have been given the opportunity to be part of this global and unique research study in the field of Private Equity. The data used in the analysis of this thesis has been retrieved from the database of the project. The objective of the research study is to gain a comprehensive insight in Limited Partners and Private Equity with a focus on how due diligence takes place. A large pool of Limited Partners were invited to participate in the survey and provided us with the necessary data to be able to conduct a decent statistical analysis. The list of Limited Partners that are potential participants was provided by a leading Private Equity information group called Private Equity International (PEI). The directory that forms the base of this study contains 2,429 Private Equity investors, additionally several contacts from Capital IQ were added. To gather substantially filled surveys, several students and myself, first send s to the Limited Partners with a request to participate, followed by many phone calls, voic s and reminder mails. During the process it appeared that several investors in the sample did not actually invest in Private Equity and many were not willing to participate. The reasons for not willing to participate are often related to time constraints, too many requests for surveys, firm policy or fear of disclosing sensitive information. Even though the response rate at this point lies at 11.9%, which is very high taken into account the extensive nature of this study, though not all respondents contribute substantially. The sample I used in this study comprises of 288 Limited Partners. See Appendix 1 for a description of the sample and Appendix 4 for the survey containing all the questions asked. 12

14 Part 3: Research question and hypotheses In this Part the research question will be presented as well as the hypotheses that will be tested. Research question Johnson (2010) reports that during 2009 only $246 billion in funds were raised compared to $636 billion in Despite of the drop in investments, which is largely attributable to the global economic climate, still very large amounts are invested in Private Equity. Then why is investing in Private Equity so popular? This study aims at explaining what drives Limited Partners to invest in the large and growing Private Equity industry, and specifically in First Time Funds. One may have several reasons to invest in Private Equity. To start, an investment in Private Equity is an example of a good way to diversify the risk of an investment portfolio. Although, the perceived benefits of Private Equity investments are broader than just the diversification advantage; Gilligan and Wright (2010) find that generally Private Equity investments outperform market returns (gross of fees) and they benefit from a cash advantage: when investments are made, commitments to the fund are drawn down immediately, however the capital might not be required immediately by the General Partner. To avoid agency problems clearly defined contracts between the Limited and General Partner are crucial. The nature of the Principal-Agent relation is such that agency problems are to some extend avoidable. Corporate governance is about how shareholders keep managers accountable for their decisions. Extensive compensation schemes which are in the interests of both shareholders (the Limited Partner) and managers (the General Partner) are applied, which are an incentive to reach the common goal of gaining high returns. The potential drawback is that Limited Partners forego the liquidity benefit of public equity investments when making a commitment to a Private Equity fund (Gilligan, Wright, 2010). There are several options for Limited Partners when considering an investment in Private Equity; one can either invest in a First Time Fund or a Reinvestment/Seasoned Fund. Kaplan and Schoar (2005) describe various factors which explain why it is more difficult for a First Time Fund to raise funds than it is for Reinvestment/Seasoned Funds. They find that more established General Partners could have the ability to invest in better investments and the capabilities of the General Partners who also act as advisors can influence the returns of the fund as well. 13

15 Hence, this implies that not knowing what the abilities of a General Partner are, as is the case for First Time Funds, might make Limited Partners more skeptical towards these investments. The factors that might influence the decision on whether or not to invest in a First Time Fund, that are of interest in this thesis are categorized in two main groups: Limited Partners characteristics and investment criteria. The most important characteristics we investigate are size and Limited Partner s Vintage year, the nature of the organization (type) is included as a control variable. The investment criteria are divided into factors relating to the experience of the General Partner, the Investment Committee of the Limited Partner and free riding opportunities. All taken together this results in the following main research question: What is the influence of Limited Partners characteristics and investment criteria on the probability and intensity of investing in First Time Funds? Hence, investing in First Time Funds will be measured by both looking at the probability and intensity of investing. A description of the terms and variables used is presented in Table 2 on the next page. 14

16 Table 2. Variables overview Variables Description Intuition Dependent variables FTF Does the Limited Partner invest in First Time Funds? Dummy indicator (1/0) (1 if one does invest in First Time Funds) Used in the probablity analysis, indicates whether or not a Limited Partner invests in First Time Funds FTF/All What is the intensity of investing in First Time Funds? Natural logarithm of the fraction (Number of FTF/Number Total) Used in the intensity analysis, indicates what the fraction of First Time Fund investments is over all investments in General Partners (includes Seasoned and Reinvestment funds) Panel A: Limited Partner characteristics Size Amount of assets under management in 2008 How large the LP is, is measured by the amount of total assets under management. Larger LPs are expected to be more likely to take on risky investments in First Time Funds Limited Partner's Vintage year Type How many years the Limited Partner has been active What is the type of Limited Partner? Pension Fund dummy indicator (1/0) Corporate investor dummy indicator (1/0) Foundation/Endowment dummy indicator (1/0) Financial services dummy indicator (1/0) * Government owned investor dummy indicator Family office dummy indicator (1/0) Fund of Fund dummy indicator (1/0) Asset manager dummy indicator (1/0) Other dummy indicator (1/0) The experience of the LP is measured by taking the the time between LP-Vintage year and now. LPs that have been active longer are expected to be more likely to take on risky investments in First Time Funds The nature of the LP can influence the likelihood of investing in First Time Funds because of differences in e.g. risk adversity, diversification strategy and the target return on investments 15

17 General Partner's Experience Reputation Education Investment committee Independent IC Size Performance Free riding Advisor TopLP Panel B: Limited Partner s investment criteria The experience of the GP is measured by the reputation and the level of education of the GP. More experienced GPs are expected to be more likely to invest in First Time Funds How important is the reputation of the General Partner? Dummy indicator (1/0) (1 if very important/crucial) How important is the quality of education of the General Partner? Dummy indicator (1/0) (1 if very important/crucial) Are the investment decisions made independently by the Investment committee? Dummy indicator (1/0) (1 if made independent) How many members does the Investment committee have? Is the compensation of the members linked to the financial performance? Dummy indicator (1/0) (1 if yes) How important is the opinion of an advisor/gatekeeper? Dummy indicator (1/0) (1 if very important/crucial) How important is the presence of Top Limited Partners in the fund of interest? Dummy indicator (1/0) (1 if very important/crucial) * Category comprises of both banks and insurance companies The reputation of the GP can be an important investment criterion and an indicator of the experience of the GP The education of the GP can be an important investment criterion and an indicator of the experience of the GP The structure and power of the Investment committee can be an important investment criterion. When the power of the IC in the investment decision is limited, LPs are exptected to be more likely to invest in First Time Funds Whether or not investment decions are made independent can be an important investment criterion and an indicator of the power of the Investment committee The size of the Investment committee can be an important investment criterion and an indicator of the power of the Investment committee Whether or not the compensation of the members is linked to the financial performance can be an important investment criterion and an indicator of the power of the Investment committee The opportunity to free ride on the advise of investment decision or third parties can be an important investment criterion. When there are opportunites to free ride, LPs are expected to be more likely to invest in First Time Funds Whether or not the opinion of an advisor is perceived as important can be an important investment criterion and an indicator of opportunities to free ride Whether or not the investment decisions of Top Limited Partners is perceived as important can be an important investment criterion and an indicator of opportunities to free ride 16

18 Limited Partners characteristics Size Hypothesis 1: Larger Limited Partners are more likely to invest in First Time Funds. Hobohm (2008) investigated what the drivers are for specific types of Limited Partners to invest in a particular General Partner and reports on a strong and significant relation between the size of the Limited Partner and the size of the General Partner resulting from an increased ability to screen funds of interest. This indirectly implies that larger Limited Partners are more likely to invest in First Time Funds since they have the ability to thoroughly investigate the General Partner of interest. Furthermore, it could be the case that even the larger risk of investing in First Time Funds is (partially) offset as follows from the following argument from principle agent theory: larger Limited Partners might compensate the General Partner if they will not take on extensive risks (Hobohm, 2008). In the literature that focuses on Follow Up Funds as well, Kaplan and Schoar (2005) find that fund size appears positively and significantly related to the performance of the previous fund. Hence, with some precaution, it can be stated that size is positively related to investing in Private Equity in general and hence also in First Time Funds. Limited Partner s Vintage year Hypothesis 2: More experienced Limited Partners are more likely to invest in First Time Funds. The experience of the Limited Partner is measured by its Vintage year. Kaplan and Schoar (2005) find that the performance of First Time Funds is lower than that of Follow-up Funds. Possibly, established Limited Partners are more likely to take on more risk and therefore are more likely to invest (a larger amount) in First Time Funds. Kaplan, Schoar and Wongsunwai (2007) add to this that generally more experienced and Limited Partners, i.e. with an earlier vintage year, tend to perform better than the newer ones, though while looking per type of Limited Partner it appears that only the interaction of experience and banks is positive. 17

19 Type (control variable) The nature of the organization i.e. the type of Limited Partner, is used as a control variable. In this study the range of types comprises of: pension funds, corporate investors, foundations/endowments, financial services, government agencies, family offices, Fund of Funds and asset managers. The various types of Limited Partners have different characteristics in terms of for example risk adversity, diversification strategy and the target return on investments. These factors might be of influence on the investment decision to invest in First Time Funds and the dispersion among the types. For example: Greenberger (2007) states that Private Equity and especially opportunities to make co-investments gain popularity among the more passive Limited Partners like pension funds and financial services firms. And Hobohm (2008) refers to Mayer, Schoors and Yafeh (2005) who report that whereas banks are more likely to invest in later stage Venture Capital Funds (hence, generally not a First Time Fund) and corporates invest more in early earlier stage Venture Capital Funds. This is confirmed by Lerner, Schoar and Wongsunwai (2007) who report that the fraction of First Time Funds is larger for later stage Venture Capital Funds and Buyout Funds than it is for early stage Venture Capital Funds. Furthermore, Hobohm (2008) states that as part of their diversification strategy, Fund of Funds hold more diversified portfolios which indirectly implies their possible presence in First Time Funds. In Figure 3 an overview of the types of Limited Partners active between is presented as investigated by EVCA/PWC/Thompson Financial. The dispersion is quite similar with the dispersion of the sample used for this study as can be found in Table 12 in Appendix 1 implying that the sample is a reliable reflection of the Limited Partners in the real world. 18

20 Figure 3. Dispersion Limited Partner types Source: EVCA/PWC/Thompson Financial Limited Partner investment criteria General Partner s Experience Hypothesis 3: Limited Partners are more likely to invest in First Time Funds when the General Partner is highly experienced. The following hypotheses are used to test this conjecture. Hypothesis 3.1. Limited Partners are more likely to invest in First Time Funds when the reputation of the General Partner is perceived as an important investment criterion. Hypothesis 3.2. Limited Partners are more likely to invest in First Time Funds when the education of the General Partner is perceived as an important investment criterion. Previous literature on the experience of the General Partner is widespread. To start, Hoell (2009) finds that reputation is a crucial factor in the Principal-Agent relation between Limited 19

21 and General Partners. Kleinschmidt (2006) adds to this that specifically the experience, personality and competence of the General Partner and its managers are important factors. In line with the abovementioned statements concerning General Partner s experience, Metrick and Yusada (2009) find that when the outperformance of a fund is perceived as related to the skills of the General Partner in the eyes of the Limited Partner, than a demand shift for Follow Up funds will occur resulting in a possible increase in size of the fund or a change in revenue earned by the General Partner. Additionally, Kaplan and Schoar (2005) focus more on the influence of experience on future investments and find that General Partners who outperform in one fund are more likely to outperform in a Follow up Fund as well, a positive relation can be found between the track record and raising new funds. Hence, this can have a positive impact on capital commitments. Consequently, it can be stated that fund flows are positively related to the past performance of the General Partners themselves. Weisbach (2010II) zooms further into pay for performance from future fund flows and finds that if a lot is known about the capabilities of the General Partner, current performance should be related more strongly to future fundraising, the effect appears to have a higher impact when it concerns First Time Funds. Hence, with some precaution, it can be stated that in general, reputation and past performance are positively related to investing in Private Equity Funds in general and thus also in First Time Funds. Investment Committee Hypothesis 4: Limited Partners are more likely to invest in First Time Funds when the power of the Investment Committee in the investment decision is limited. The following hypotheses are used to test this conjecture. Hypothesis 4.1. Limited Partners are more likely to invest in First Time Funds when the Investment Committee does not have the ability to make the investment decision completely independent. 20

22 Hypothesis 4.2. Limited Partners are more likely to invest in First Time Funds when the Investment Committee has fewer members. Hypothesis 4.3. Limited Partners are more likely to invest in First Time Funds when the compensation of the Investment Committee is linked directly to the financial performance. During this research study it appeared that not all Limited Partners have an Investment Committee. Of those who do, it is important to know how, among others, decision making takes place and how compensation is arranged for. For example, it is plausible that if flat compensation is used, it is less likely that members would be willing to take more risk to gain a potential higher return and hence invest in First Time Funds. The reason why is that they cannot benefit from any upside potential that might result from the investment. Since I have experienced that little previous literature is available on Investment Committees, the relation between the management and the board is assumed to be to some extend similar to the relation of the Investment Committee and the management of the Limited Partner. To start, Kleinschmidt (2006) mentions that the composition of the board should be arranged in a way that one can supervise the management effectively, meaning that the members of the board should not be managers or have any other relation with them. This is related to whether or not the Investment Committee has the ability to make decisions independently; it is believed that objective members are a prerequisite for proper decision making. Weisbach, (2010I) states that the independency of the board is closely related to the interaction between the board and the CEO who generally like inferior independence of the board which can occur if the CEO has strong bargaining power; in that case the CEO will lobby for more compensation. In Hermalin and Weisbach (Forthcoming) one referred to previous work by Hermalin and Weisbach (1998) in which it appeared that when the CEO is seen as a kind of crown jewel for the firm, the increased bargaining power that comes with that belief results in less board scrutiny and increased compensation. The abovementioned statements have implications for conjectures 4.1 and 4.3. Assuming that boards and Investment Committees have the intention to make responsible and not too risky decisions (as opposed to CEOs), it is likely that in case of independent decision making and the absence of any financial incentives and consequently any upside potential, a Limited Partner will be less likely to invest in a First Time Fund. 21

23 Substantial evidence on the negative relation between board size/involvement and performance can be found. Kleinschmidt (2006) finds that when boards get larger, the effectiveness decreases as a result from a drop in interaction between the members of the board. S. Weisbach reports on this relation in several papers. First, in Weisbach (2010I) it can be found that a reduction of the size of the board would increase the performance of a firm. In Hermalin and Weisbach (Forthcoming), he again refers to the negative relation between the number of directors present in a board and the performance of the firm, though in this paper he distinguishes between outside and inside directors and finds that in addition to size, the level of inside directors is negatively related to performance as well. As mentioned in Hermalin and Weisbach (Forthcoming), also in former times substantial literature was written on board size and effectiveness. For example, Jensen (1993) and Lipton and Lorsch (1992) find that larger boards are not as effective as smaller boards. It is assumed that a smaller size of the Investment Committee is related to higher effectiveness and more investments in First Time Funds; higher effectiveness is assumed to imply better interaction between members and more and better screening, increasing investments in First Time Funds (less screening is needed for investments in Follow up Funds since one can easily check and use the IRR as benchmark). Last I will check for evidence on differences between the different types of Limited Partners. It can be found that endowments generally have active Investment Committees with a superior performance which largely results from strong bonds with successful alumni. This is in contrast to pension funds Investment Committees in which generally more random people are seated (Lerner and Schoar, Wang, 2008). Lerner, Schoar and Wongsunwai (2007), also state that pension funds committees suffer from conflicting objectives, limited compensation and frequent turnover as well which all tend to make the Investment Committees less efficient. Free riding Hypothesis 5: Limited Partners are more likely to invest in First Time Funds when they have the opportunity to free ride on advisor s opinions or investment decisions of third parties. The following hypotheses are used to test this conjecture. 22

24 Hypothesis 5.1. Limited Partners are more likely to invest in First Time Funds when the opinion of an advisor is perceived as important. Hypothesis 5.2.Limited Partners are more likely to invest in First Time Funds when Top Limited Partners are present as investor in the fund of interest Previous literature on Top Limited Partners and Top Funds report that given that most Limited Partners claim that the top funds are all highly oversubscribed, it seems likely that the better funds voluntarily choose to stay smaller and grow less that proportionally with the increase in performance (Kaplan and Schoar, 2005). Borell (2004) states that what makes placing funds hard is the fact that many top funds are not open to new investors. This indirectly implies that even though small Limited Partners would be willing to free ride on the popularity of specific funds, the industry structure makes it sometimes impossible. Related to this are the findings of Lerner, Schoar and Wongsunwai (2007) who report on the fact that the General Partners of popular funds (i.e. above target size) or funds with a short closing time, have the opportunity to benefit from being selective in which Limited Partners are allowed to invest in the fund. 23

25 Part 4: Empirical analysis In this Part a description of how the data will be analyzed will be provided, followed by the introduction of the models used and the statistical output of the analysis. Data analysis For the analysis the statistics program STATA will be used. First, the correlation between the independent variables will be checked to see how strongly they are associated with each other. When correlation is high, multicollinearity arises which affects the calculations of the individual variables. Probability To analyze the effect of the variables on the probability of investing in a First Time Fund, a Logit analysis will be used. The dependent variable indicates whether or not the Limited Partner invests in First Time Funds. The data for the dependent variable are originally based on count data (i.e. the number of investments in First Time Funds). These are transformed into a dummy variable so that a standard Logit regression can be used by means of maximum likelihood estimation. The outcome of the analysis is a prediction of the probability of investing in a First Time Fund. As the fit of the model improves the Log Likelihood will be less negative and the R- squared will go up 1. The R-squared value tells how much of the fraction of variation in the dependent variable is explained by the independent variables. The interpretation of the coefficients is slightly different than in an ordinary regression analysis; the coefficient of a Logit regression shows the change in the predicted logged odds of having the characteristic of interest for a one unit change in the independent variables 2. When the P-value is 0.05 or less the null hypothesis is rejected with a significance of 95%, with P-values of 0.1 or 0.01 the significance levels are respectively 90% and 99%. When the independent variables are highly correlated it might be that the explanatory value of the independent variables is insignificant on the individual level, but significant at the model level (i.e. Prob>chi-squared). In order to be more specific the marginal effects (i.e. effect of a one unit change in the dependent variable) of each independent variable will be estimated by means of the mfx tool in STATA. 1 Brooks, C., Introductory Econometrics for Finance, Second edition (2008), Cambridge University 2 Logistic regression guide, retrieved from 24

26 Intensity To analyze the effect of the variables on the intensity of investing in a First Time Fund, an OLS regression analysis will be used. The dependent variable is the natural log of the fraction of First Time Funds over total investments. Note that in this analysis the focus is on the number of funds and investments, not on the amount of capital allocated. To find out whether each independent variable has a significant effect on the dependent variable, it will be checked whether the coefficients are significantly different from zero and what the direction of the effects are. When the P-value is 0.05 or less the null hypothesis is rejected with a significance of 95%, with P-values of 0.1 or 0.01 the significance levels are respectively 90% and 99%. When the independent variables are highly correlated it might be that the explanatory value of the independent variables is insignificant on the individual level, but significant at the model level (i.e. Prob>F). Finally the R-squared value will be checked to see how much of the fraction of variation in the dependent variable is explained by the independent variables, an increase in R-squared across the models also indicates whether or not improvements in the explanatory value of the model occur. Models and expectations The models to test the probability of investments and the intensity focus on the Limited Partner characteristics and investment criteria as explained above. For both the analysis on the probability and intensity the same models containing the same variables are used. To start, the base model (Model 1) contains only the Limited Partner characteristics as explanatory variables. The variable Size measures the assets under management in 2008 and the variable Limited Partner s Vintage year measures the number of years the Limited Partner has been active. Because of the lack of a track record, Limited Partners might perceive an investment in a First Time Fund as more risky. Consequently, it is plausible that more experienced and larger Limited Partners are more likely to invest (a large fraction) in First Time Funds resulting from benefits from experience and ability to spread investments among different types of fund. Model 2 attempts to explain the influence of the General Partner s experience on investing in First Time Funds. The base model is extended with the variables that measure the experience of 25

27 the General Partner; including reputation resulting from previous (Private Equity) jobs and the education of the General Partner. The reason why these variables are included is because Limited Partners might be sensitive to the perceived ability of the General Partner since in the absence of a fund track record, the track record of the General Partner him/herself is the closest measure that can mimic this. Therefore, I expect that Limited Partners are more likely to invest (a large fraction) in First Time Funds when the General Partner has a good reputation and/or a good education which might be perceived as a risk reducing factor in the perception of Limited Partners. In Model 3 the variables that measure the influence of the Investment Committee are added. The influence of the number of Investment Committee members, the link between compensation and financial performance and the independency of the investment decision are considered. It seems sensible that Limited Partners are more likely to invest (a large fraction) in First Time Funds when the Investment Committee contains fewer members (easier to get to consensus/majority plus higher effectiveness and screening ability), when compensation is linked to performance (might reap more personal profits so willing to take more risk) and when the investment decision is not made independent (influence of e.g. compensation driven CEOs). The last model, Model 4 is extended with variables measuring the influence of free riding opportunities. The variables of interest are the opinion of advisors and the presence of Top Limited Partners in the fund of interest. It is expected that Limited Partners are more likely to invest (a large fraction) in First Time Funds when advisors advise the Limited Partner to do so and when it appears that some well known and established Limited Partners invest in the fund as well. The effect of the presence of Top Limited Partners is expected to be stronger for smaller Limited Partners since they might not have the resources to perform an extensive due diligence themselves. In the Models 3b and 4b, the variable IC Size has been left out. The reason why is that it appears that many participants in the survey have not answered the questions regarding the composition of the Investment Committee. Excluding this variable will raise the number of observations and will consequently increase the predictive and explanatory power of the models. 26

28 Statistical output Table 3. Descriptive statistics Variable # of obs. Mean Standard deviation Minimum Maximum Limited Partner's Vintage year Size (millions) Reputation Education Independent IC Size Performance Advisor TopLP Number of FTF investments Number of Reinvestments Number of Seasoned investments

29 Size Table 4. Pair wise correlation matrix independent variables Size LP's Experience Reputation Education Independent IC Size Performance Advisor TopLP LP's Vintage year (0.423) Reputation (0.568) (0.338) Education (0.916) (0.952) (0.706) Independent * (0.734) (0.517) (0.085) (0.248) IC Size (-0.123)* (0.509) (0.094) (0.796) (0.822) (0.101) Performance * * 0.202* (0.622) (0.094) (0.440) (0.098) (0.0740) (0.165) Advisor *** (0.804) (0.974) (0.306) (0.002) (0.504) TopLP *** (0.398) (0.757) (0.789) (0.000) (0.920) (0.929) (0.337) (0.199) P-values in parentheses, * (significant at 10% level), ** (significant at 5% level), *** (significant at 1% level) 28

30 Dependent variable = FTF Table 5. Logit regression on the Probability of investing in First Time Funds (FTF) Model 1 Model 2 Model 3a Model 3b Model 4a LP characteristics GP's Experience Investment committee Investment committee~ Free riding Model 4b Free riding~ Coef. Mfx Coef. Mfx Coef. Mfx Coef. Mfx Coef. Mfx Coef. Mfx (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.) Size 0.232** 0.055** -0,006-0,001-0,273-0,057-0,046-0, ** -0,086-0,047-0,010 (0.115) (0.113) (0.209) (0.153) (0.408) (0.153) Limited Partner's 4,988 1,119 16,611 3,523 30,996 6,430 12,773 2, *** * 14,445 2,930 Vintage year (11.821) (15.006) (34.277) (15.415) (56.375) (17.089) Reputation 1.968** 0.456*** dropped dropped 2.634* 0.576** dropped dropped 2.607* 0.572** (1.824) (1.416) (1.545) Education -0,446-0,099 0,417 0,082-0,631-0, * 0,215-0,467-0,100 (0.627) (1.245) (0.759) (2.258) (0.832) Independent -3.08** *** * ** *** ** * ** (1.463) (0.981) (2.027) (1.164) IC Size 0,137 0,028 left out left out 0.292* 0.029* left out left out (0.154) (0.173) Performance -1,362-0,269-0,543-0, ** ** -0,600-0,123 (1.667) (0.678) (2.172) (0.729) Advisor *** *** -0,567-0,121 (1.788) (1.071) TopLP -1,561-0,905-0,285-0,060 (1.221) (0.809) Type of LP Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes _cons -39, , ,444-97, , ,666 (89.93) ( ) ( ) ( ) ( ) ( ) No. of observations Wald chi-squared (df) (10) (10) ( ( (10) (10) (12) (12) (12) (12) (14) (14) Prob>chi-squared 0,311 0,311 0,070 0,070 0,031 0,031 0,076 0,076 0,017 0,017 0,081 0,081 Pseudo R-squared 0,096 0,096 0,156 0,156 0,295 0,295 0,295 0,295 0,553 0,553 0,302 0,302 Log pseudolikelihood -78,167-78,167-49,048-49,048-16,606-16,606-31,120-31,120-10,536-10,536-30,811-30,811 Prediction of y 0,602 0,602 0,695 0,695 0,706 0,706 0,707 0,707 0,888 0,888 0,717 0,717 ~ (excluding IC Size) * (significant at 10% level), ** (significant at 5% level), *** (significant at 1% level) Mfx = Marginal effect of respective independent variable evaluated at the means of the other independent variables estimated in respective model 29