Review of empirical research on the performance of the US private equity market

(1)

Review of empirical research on the

performance of the US private equity market

What Dutch investors and regulators can learn

Augustus 2007

Rijksuniversiteit Groningen Doctoraalscriptie Economie

Afstudeerrichting Financiering en Belegging Begeleider: Professor van der Meer

Tweede begeleider: Professor Plantinga

(2)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 2

Introduction

One of the most famous anomalies in the field of finance is centered around the equity premium. Although no precise definition exists the equity premium is usually defined as the difference between the return on the market portfolio of common stock and the risk free interest rate. The equity premium is central to corporate finance and investment. It is important in portfolio allocation decisions and it is a major determinant in cost of capital estimates. Despite this importance there exists no consensus among finance practicers and researchers on how to compute and apply the equity premium. One finding seems to be generally shared. Historical stock returns seem to be too high compared to the risk free rate. Large differences are found that cannot be justified with standard finance models like CAPM (see for example Mehra and Prescott 1985 and Siegel and Thaler 1997). This anomaly has come to be known as the equity premium puzzle.

Mehra and Prescott (1985) 1 were the first to speak of an equity premium puzzle. They found that stock returns were too high given the observed volatility of consumption. Ever since attempts have been made to solve this puzzle. These attempts have come in two forms. First, research has been done to come up with new ways to compute the premium or to find new data. New numbers might result into equity premia that are more in line with the theory. Second, new theories have been developed in which the high equity premium is no longer considered a puzzle. Despite these efforts the debate about the equity premium is still ongoing.

The same kind of debate is currently going on about the performance of private equity investments. At the end of the last century the market for this asset class was booming. After a few bad years during the recession of 2001 and 2002 its star is rising again. In a 2004 cover story British magazine The Economist crowned private equity to be ‘capitalism’s new king’, describing how private equity has changed the business world.2 Newspapers are filled with stories about private equity funds taking over or investing in

(5)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 5 companies. Examples from the Dutch economy are the takeover of VNU in 2006 and the sale of the Philips chip division to a consortium of private equity investors in the same year. Enormous amounts of money are at stake in these deals. The Economist reported in February 2007 that in the United States and Europe 440 billion dollar was spent on buy outs by private equity funds. In the same year the National Venture Capital Association (NVCA) states that over 30 billion dollar of funds was raised by venture capital funds in the United States alone.3 However, it is not just the vast amount of money flowing in and out private equity funds that makes private equity an investment class to care about. Private equity funds have become an important player at the economic stage. Their performance affects the whole economy through the way they finance their deals and the composition of their investment base.

Despite these numbers and the importance of private equity as an investment class it is still not fully understood in terms of risk, returns and cash flows. One of the main reasons is the lack of data. Private equity is called private for a reason; the private equity funds are to a great extent free from public disclosure requirements. Information being only limited available makes analyzing developments in the private equity market difficult. Because of this data problem earlier research focused mainly on either aggregate trends in private equity or on the relationships between entrepreneur, private equity fund and the investor.

At the end of the nineties the attention shifted to the performance of the private equity funds and their investments. Quantity and quality of available data increased and more and more investors became interested in private equity as an asset class that could offer diversification benefits or lead to higher returns. This led to an increase in research on the performance of private equity. The key topic here is the risk-return relationship of private equity, its performance relative to public markets and overall the attractiveness of its investment opportunities.4

3_{http://www.nvca.org/}

4_{Ick, Matthias M., 2005, Performance measurement and appraisal of private equity investments relative to public equity markets,}

(6)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 6 The central question in the debate about the performance of private equity is whether its returns imply a private equity premium. That is, are returns to private equity higher than returns to publicly traded equity? Of course private equity has some specific characteristics that public equity doesn’t have. Therefore it is logical to assume that private equity investing differs from public equity in terms of unsystematic risk. According to standard finance theory a difference in unsystematic risk is not priced in a perfect capital market. However, the private equity market is far from perfect. This might lead to unsystematic risk being priced and cause private equity to trade at a (systematic) risk adjusted premium or discount

As with the public equity premium puzzle the understanding of the risk and return of private equity investing is of great importance. It is a major class of financial assets and its main investors (mostly institutional investors like banks or pension funds) are among the most important economic actors. However, in the existing literature so far there is no consensus on whether private equity outperforms public equity on a risk adjusted basis.

The US private equity market developed earlier than the private equity market in the Netherlands. Therefore, most of the empirical research done on the performance of private equity investing is based on the US private equity market. Based on this research I will answer the following question. What are the lessons that Dutch investors can learn from the historical performance of the US private equity market? After answering this question I will look at the role regulation can play in avoiding problems that might arise in buy out transactions performed by private equity funds.

(7)

(8)

1 Historical developments in the US private equity market

1.1 Definition of private equity

Private equity is the name for a very broad asset category, literally meaning all investments that are private. The focus of this thesis will be on the organized private equity market. The organized private equity market is defined by Fenn, Liang and Prowse (1997) as professionally managed equity investments in the unregistered securities of private and public companies. An equity investment is any form of security that has an equity participation feature.5

Three players can be identified in the organized private equity market; the investors in private equity, the issuers of private equity and the intermediary who professionally manages the investment. I will come back to the role of the intermediary in chapter two. First I will discuss the different types of investors in private equity and the characteristics of firms that issue private equity.

A number of institutional investors invest in private equity. Public and private pension funds are the largest group. They are followed by endowments, foundations, insurance companies and bank holding companies. Wealthy families and individuals, investments banks and non financial corporations also invest in this asset class. Within these groups, the larger institutions are typically the ones that invest most in private equity.

Companies that issue private equity in general have one common characteristic. They need private equity funding because they cannot get financing from the public equity or debt market. Based on firm specific characteristics the issuers can be divided into two categories. Firms that seek venture capital funding and firms financed by buy out funds.

5_{Fenn, George W., Nellie Liang and Stephen Prowse, 1997, Economics of the private equity market, Financial Markets, Institutions}

(9)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 9 1.1.2 Venture Capital

There exists no widely agreed definition of firms that seek venture capital financing. See for example Ick (2005) page nine for a classification in three stages. For simplicity I follow Fenn (1997). in making a distinction between early stage new ventures and later stage new ventures.

Early stage new ventures are firms with unproven production technologies or large uncertainty about the existence of a market for their product. That is, their reason for existence still has to be proved. Some early stage new ventures only have a business plan and won’t be making a profit for years to come. It can be difficult for these young start-up firms to acquire funds needed to operate and expand their business. In this case venture capital can provide an attractive way of funding. A venture capital fund will invest in a young promising firm to ensure it can stay in business and fulfill its potential. Financing needs can range from investments in research and development, marketing issues or to set up initial operating facilities to start selling their product on a commercial scale. These ventures are surrounded by high risks of failure because they haven’t proved yet that their business is actually feasible.

Later stage new ventures have a more proven production technology and market for their product. They are already in business and making a profit. Their risk comes more from the threats and uncertainties that affect a young small business. They need venture capital financing to increase capacity, add production facilities or update equipment. In short, they need money to grow. The goal of both early stage new ventures and later stage new ventures is the same however. The aim is to grow fast enough to ultimately be able to get access to outside financing. Finally this should result in going public through an IPO or a sale of the company to another party.

1.1.3 Buy Outs

(10)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 10 Newspapers today are still filled with publications about large deals where a private equity fund takes a public company private. Typical features of these companies are slow growth rates, high levels of free cash flows and management under fire of shareholders for not pursuing the right strategy. Related to these buy outs are the management buy outs where the management of a public firm buys the company using private equity financing. Another area where buy out funds are active is the divesture of business units by public firms. When for some reasons a company wants to divest one his units, doing this through a private equity fund is one of the options. For example, the chips division of Philips was sold to a private equity fund in 2006.

Firms in financial distress, being overleveraged for example, are often not able to acquire financing from the public market. At the same time, they need financing in order to survive, stay in business and manage a turnaround. Private equity investing might be an outcome. In exchange, the private equity fund often gains some control over the future direction of the firm.

Table 1.1 shows the most important differences between the typical venture capital investment andthe typical buy out investment. Table 1.2 summarizes the characteristics of the different issuers of private equity funding. From now one, when I am talking about venture capital funds and buy out funds I follow the distinctions made in these tables. In the remainder of this chapter the historical developments in the US organized private equity market are discussed.

Table 1.1 Differences between the typical venture capital investment and the typical buy out investment. Based on Woodward (2004)

Venture capital investment Buy out investment

Company stage Early Mature

Company revenues None/small Sustaining

Funds investing in same company Many One

Screening and selection Heavy Moderate

Management involvement Moderate Heavy

Number of financing rounds 4-7 1

(11)

Table 1.2 This table shows the different issuers of private equity. The table lists their characteristics, why they need private equity financing and which type of private equity funds they rely on. Based on Fenn (1997)

Characteristics Early stage

new ventures Later stage new ventures Corporate divestures Public and private firms in financial distress

Public buy outs Other public

firms Size Revenues between zero and $15 million Revenues between $15 and $50 million Established with stable cash

flows between $25 and $500

million

Any size Any size Any size

Financial attributes High growth potential High growth potential Growth prospects vary widely Overleveraged or operating problems Underperforming

High levels of free cash flow Depends on reasons for seeking private equity Reasons for seeking private equity Start operations Expand production facilities and operations Cash out early stage investors Finance a required change in ownership or capital structure (repay debt) Focus on core business Effect a turnaround Finance a change in management or a change in management incentives Variety of reasons. See text for a summary Major sources of private equity Venture capital funds Venture

(12)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 12 1.2 The early period: 1946-19706

In the post World War 2 period the organized private equity market was devoted solely to providing venture capital financing. The first institution to professionally manage venture capital investments was the American Research and Development Corporation (ARD), founded in 1946. Three reasons were at the heart of this foundation. First of all, the founders were worried about the lack of long term financing possibilities for new ventures. They also wanted to create an institution that not only provided these new ventures with the necessary capital but also with valuable managerial advice. Managerial skills and experience was seen as just as critical as capital in the success of the business. Finally they hoped that the corporation would attract large institutional investors. More and more the money was shifting from the wealthy individuals and families into the hands of these investors. Unfortunately, the ARD was no big success since it attracted much less capital than it aimed for.

In this period a number of private venture capital companies where formed. They managed the venture capital investments of wealthy families on an ad hoc deal by deal basis.

The first step towards professionally managed venture capital investing was made with the establishment of Small Business Investment Companies (SBIC). These companies were aimed to provide capital to new, small and risky companies. To encourage this they had access to supplementary loans and were eligible for certain tax benefits. Despite a number of flaws in the SBIC system (i.e. abuse of loan option, lack of institutional investors, and poor quality of investment managers) a record amount of capital flowed to small fast growing companies and the SBIC’s proved a fruitful training ground for future venture capitalists.

1.3 Development period 1970-1980

This period saw the start of the limited partnership as a way of organizing the venture capital investments. However, this did not result in growing capital committed to these

(13)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 13 partnerships. Very poor exit conditions (no IPO market and economic recession) made investors reluctant to invest.

This did create attention among venture capitalists to develop strategies for non venture private equity investing. In this period, the first leveraged buy outs occurred. Despite these developments, capital under management did not grow by much.

1.4 The boom: 1980-2000

(14)

2 Private equity as an asset class

2.1 Reason of existence of private equity

The first question to answer is why private equity investing exists in the first place. After all, it is one of the most expensive forms of finance. And despite of its riskiness, in a perfect public market it should be possible to get financing for profitable investments. Smit (2002) therefore concludes that private equity financing exists because of imperfections in the public market. He labels these imperfections information asymmetry and agency problems.7

Unique risks are involved in private equity investing. Therefore it is important for an investor to know as many as possible of the firm he invests in and perform thorough due diligence. This requires time, effort and specific knowledge normally not available to an investor.

In a private equity investment the investor has a clear focus on the future of the company. To make sure the management of the company acts on behalf of the investor, tries to achieve the same goals and agrees on the strategy for the future, private equity investors generally want to exert significant influence in exchange for their money. Again, this is a time consuming task requiring specific expertise.

These reasons bring us to the way private equity investments are organized and managed; the limited partnership.

7_{Smit, J.T.J. and van den Berg , W., 2007, De private-equity golf, Maandblad voor accountancy en Bedrijfseconomie, jaargang 81,}

(15)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 15 2.2 The limited partnership8

When defining private equity earlier I talked about the three players active on the private equity market. Investors and issuers are discussed. It is now time to look at the role of the third player, the intermediary between investor and issuer and the general manager of the private equity investment.

Why can’t an institutional investor invest directly in private equity? The answer to this question lays in the reasons for the existence of the private equity market; information asymmetry and agency problems. Fenn (1997) sees the same problems but name them sorting issues and incentive issues. It requires a lot of information when selecting an investment. Incentive issues arise after the selection process when managers have the opportunity to act at in a way that might benefit themselves at the expense of the investor. As noted by Smit and van den Berg (2007), private equity financing is used when these issues are most severe.

Extensive pre-investment screening and intensive post-investment monitoring are therefore essential in the private equity market. This is where the intermediary comes in. According to Fenn (1997) it is most efficient when these screening and monitoring activities are delegated to one single intermediary. Single investors simply lack the time, resources and expertise to perform these tasks.9 The way in which the institutional investor and the intermediary interact in the private equity market is called the limited partnership. In this partnership institutional investors are the limited partners and the specialized investment manager, the intermediary, serves as the general partner. Examples of specialized private equity managers are Kohlberg, Kravis and Roberts (KKR) a well known buy out group and Kleiner, Perkins, Caufield and Byers, specialized in venture capital.

(16)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 16 2.2.1 Activities of the investment manager

The private equity managers are responsible for investing the funds provided by their limited partners and for generating a return as high as possible on these investments. In doing this, four stages of manager activities can be identified.10

2.2.1.1 Selection and screening

The process starts with the selection of the investments. A private equity manager uses his network of relationships in the financial world to get access to the necessary information on the available deals in the private equity market. This information is then sorted and thoroughly evaluated before the decision to invest is made. This requires specific knowledge and expertise from the part of the investment manager. Because of this private equity managers often tend to specialize their activities based on region or industry.

2.2.1.2 Structuring the deal

When the decision to invest in a company is made the next step is the structuring of the deal. Negotiations with the company are started to set the terms. Two issues are at stake here. The first is the valuation issue. How much is being paid and how big is the ownership stake the private equity investor acquires? The second issue relates to governance. How does the private equity investor make sure the incentives of the management are in line with his goals and to what extent can he exert influence over the company? The answers to these questions decide for a great deal the final success of the investment.

2.2.1.3 Managing the investment

Successful negotiations will lead to the closing of the deal. Not only will the investment manager monitor the company but when necessary he will also participate actively in management. In fact, the private equity industry claims that their ability to add value to the company through management assistance is what distinguishes them from other

(17)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 17 providers of finance.11 This management assistance can take several forms, ranging from providing help with finding additional financing to developing a whole new strategy and deciding on operational issues. Of course the degree of management interference differs per investment. Early stage ventures require more care than later stage ventures. Buy outs where bad management was the underlying reason for the takeover also lead to heavy managerial involvement of the private equity general partner. In general, a limited partnership is rarely a completely passive investor. As with the selection process being active in management requires knowledge and time from the general partner, leading to specialization.

2.2.1.4 Exit the investment

The fourth and final stage is the exit of the investment. Without exit there will be no return on the investment for the limited partners so clearly this is a very important phase. There are four different ways to exit the investment, each with their own consequences for the company and the limited partnership.12

The first option is to take the company public through an initial public offering (IPO). In general this leads to the highest valuation of the company and gives the management of the company access to capital for the future. However, there are drawbacks to this method. An IPO can be costly. An underwriter needs to be paid to market the offering and legal and accounting costs are incurred. Besides that, restrictions might be placed on the number of shares the partnership is allowed to sell in the market. If not all shares can be sold the limited partners are left with limited liquidity. Moreover, active involvement of the general partner with the company usually only ends when all the shares are sold, resulting in time consuming activity.

A private sale to a strategic buyer might be another option. All shares are sold and transaction costs are lower than in the case of an IPO. However, for the company a private sale almost always means a loss of independence because he is now part of another company.

11_{Fenn, George W., Nellie Liang and Stephen Prowse, 1997, page 54}

12_{Dronkers, Wouter, 2005, De Secondary Buy-Out Onder de Loep; Een Onderzoek naar de Kenmerken, Drijfveren en Toekomst van}

(18)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 18 When no buyer can be found a share buy-back by the company can offer an outcome. The company purchases his own shares back from the private equity fund. This option offers the limited partner the advantage of liquidity but the price is generally lower than with the first two options. It is therefore considered the second least favorite exit option and happens at investments that are labeled unsuccessful.

With completely failed investments a liquidation of the company is the only exit option left to the investors. Any remaining assets are sold and the investment is written off. Of course this is the least favorite option. Not only because of losing the invested money, but also because of the effects a write off has on the reputation of the investment manager. I will come back to this effect later in my thesis. The exit options are summarized in table 2.1

Table 2.1 Possible exit options for a private equity investment and their main advantages and disadvantages. Based on Dronkers (2005)

Exit option Advantage Disadvantage

IPO High valuation

Access to capital

High transaction costs Limited liquidity

Private (strategic) sale Complete exit Loss of independence company

Share buy back Complete exit Low price

Liquidation/write off - Loss on investment

Reputation effects

2.2.2 Relationship between limited and general partner

(19)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 19 The most effective way to limit potential conflicts of interest are two broad mechanisms used in the contractual agreement between limited and general partner.13 These broad mechanisms are the compensation scheme and direct means of control. A typical compensation scheme for a limited partnership consists of two parts. First of all, the limited partners pay a management fee to the investment manager. This fee is usually around 2 or 3 percent of capital committed by the limited partner. Besides this fee the general partner also shares in the capital gains made on the investment. As a general rule, the general partner gets 20% of the profits made by the partnership. It is this agreement that serves to align the interest of all parties.

Reputation effects also play a role. When a general partner wants to raise a new fund, limited partners will look at his track record to decide whether or not to trust him with their money. In order to be able to come back to the institutional investor a private equity manager cannot afford a bad reputation. Therefore, reputation will act as an incentive to act in the limited partners interests.

The agreement between limited and general partner can also contain aspects of direct control. Covenants can be included that forbid the general manager to invest in specific industries or companies to avoid excessive risk taking. In some cases, the limited partners (at least the larger ones) are actively involved on advisory boards and thus can exert some influence over the general manager.

2.2.3 Value, cash flows and return

In the end of course the goal of the investors in private equity is to earn a return on their investment that compensates them for the risk they have taken on. How does this process work?

It starts with the valuation of the target company. During the selection and structuring stage the investment manager decides how much he thinks the company is worth and how much value he thinks the private equity manager is able to add. Next step is to decide on the discount rate to apply to this value, the required rate of return the investment manager wishes to earn. Fenn (1997) has interviewed private equity investment managers. Based

(20)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 20 on these interviews they estimate that required rates of return for investing in venture capital are between twenty five percent for later stage deals and fifty percent for more risky later stage deals. Buy out funds generally require a return between fifteen percent and twenty five percent.14 These returns seem extremely large. One explanation is that the required return is a conditional return. That is, the return the investment manager requires when the investment is a success, considering the failure rate, which is high especially amongst early stage ventures. The estimated company value, combined with the required rate of return, determine the price the private equity manager is willing to pay for his stake in the company.

When it is time to make the investment the general partner will call money from its limited partners to invest. This is called a draw down of capital. Now the general partner monitors and manages the investment in order to make sure the estimated value is being obtained. Finally, an exit option is chosen and the capital gains are being redistributed to the limited partners, after deduction of fees and profit shares for the general partner. Figure 2.1 shows how cash flows between the institutional investor, the limited partnership and the company in a typical private equity investment.

Figure 2.1 This figure shows the cash flows between the investors in private equity, the limited partnership led by the general manager and the firm receiving the private equity funding. The returns accruing to the investors are net of fees paid to the general partner and net of the general partner’s share of the profit. Based on Gompers and Lerner (2000).

14_{Fenn, George W., Nellie Liang and Stephen Prowse, 1997, page 51}

(21)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 21 This is however a simplified picture because it doesn’t take into account the timing of these cash flows. Ljungqvist and Richardson (2002) document on the timing of the cash flows in typical limited partnership. They find that the average private equity fund is a ten year limited partnership. At the beginning of a funds life the limited partners commit to an amount of capital they are willing to invest. However, the actual draw down of this capital only takes place when the general partner finds a good investment opportunity. On average, a fund draws down fifty six percent of the committed capital in the first three years of its existence. After that it takes another three years before ninety percent of the capital is invested.15 However, the speed at which capital is invested varies across funds. When calculating the return on a private equity investment it is important to take into account this timing of the cash flows. Assuming that all committed capital is invested immediately will ignore the time value of money.

2.3 Financial characteristics of private equity investing

In the previous chapter we have seen how private equity deals are structured, who the players are on the private equity market and how they interact. The next step is to look at the financial characteristics of investing in private equity. From the viewpoint of the institutional investor, next to the return on their investments these are the most important characteristics. That is, these characteristics determine the (unsystematic) risk of investing in private equity and with that determine whether the return earned is high enough to compensate for this risk. The three main characteristics typical for private equity investments are illiquidity, lack of homogeneity and information asymmetry

2.3.1 Illiquidity

Probably the most important characteristic of a private equity investment is the illiquidity that comes from holding it.16 This illiquidity stems from two reasons.

15_{Ljungqvist, Alexander and Matt Richardson, 2002, The cash flow, return and risk characteristics of private equity, Working paper,}

New York University, page 11

16_{Ick (2005) for example does not discuss the existence of the private equity premium puzzle but questions the existence of a private}

(22)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 22 First, the structure of the limited partnership leads to illiquidity. An investor has to commit capital to a fund with a typical lifespan of ten years. During this time, he can not exert influence on how his capital is being invested. So illiquidity comes from a forced long holding period and a lack of influence.

The second factor leading to illiquidity is the lack of an active and liquid secondary market for private equity. Once the investment is made, it is not possible for the investor to trade it in a public market.

2.3.2 Heterogeneity

In developing models to value financial assets one of the assumptions often made is that investors form homogeneous expectations about the risk and return of these assets. As described earlier in this chapter, private equity investment managers actively manage their portfolios to earn a sufficient return. However, active portfolio management implies heterogeneous expectations.17 The assumption of homogeneous beliefs does not hold for private equity investments made in the organized private equity market.

2.3.3 Information asymmetry

In the beginning of this chapter a lot has been said about information asymmetry in private equity investing. Information asymmetry exists in the relationship between firm and general partner and also in the relationship between the general partner and his limited partners. This stems from two simple facts. Of course, managers of the firm know their own company better than anyone. Second, the general partner devotes more time and energy to the screening and selection process to get to know the companies than his limited partners do.

As discussed before, the structure of private equity deals and the interaction in the limited partnership take away a great deal of this information asymmetry and serve to align goals

17_{Meer, van der Robert, Auke Plantinga and Roelof Salomons, 2007, Performance Measurement and Evaluation as Part of the}

(23)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 23 and interests. However, not all the uncertainty and risk surrounding the existence of information asymmetry can be taken away.

2.3.4 Other characteristics

2.3.4.1 Leverage

Private equity deals that attract the most attention from the financial press are without a doubt the large leveraged buy outs of big public companies. Not only because these deals are big in terms of valuation but also because of the way these deals are financed. More specific, these investments tend to be highly levered, with leverage ratios of 75% being no exception.

This leverage has consequences for the required rate of return for the equity providers in these deals, the limited partners. With higher leverage, the equity stake becomes more risky resulting in a higher expected return. When comparing returns of private equity investing with returns on public market investments this leverage effect has to be bourn in mind when making risk adjustments because leverage results in a higher equity beta.

2.3.4.2 High failure risk

Early venture capital partnerships invest in companies that in some cases are not more than a business plan. No certainty about success of this plan exists. This makes these investments extremely risky, and it is no surprise that failure rates among these companies are high, sometimes estimated to be around fifty percent.18

Venture capital managers acknowledge these risks, reflected by the high required rates of return they demand for these investments. However, the extreme high failure rate remains an area of concern.

2.4 Asset pricing models

The important question is whether these characteristics of investments in private equity have an influence on return. That means, are these characteristics priced? The answer to

(24)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 24 that question is the topic of chapter 3 in which research on risk and return of private equity is discussed. Here I will first look at the capital asset pricing model (CAPM) as a way of pricing financial assets.

The CAPM is a model that specifies the relation between risk and expected return on a financial asset. The main insight of the CAPM is that the variance of a financial asset by itself is not an important determinant of its expected return. What is important is the covariance of the return with the return on the market portfolio. This figure is called beta and is a measure of the systematic risk of the asset.19

According to the CAPM the expected return of a financial asset is mathematically determined by the following equation;

(2.1) E(r)=rf +β(E(rm)−rf)

So according to the CAPM the expected return of a security is solely determined by its systematic risk, its covariance with the market. That would mean that the above mentioned characteristics of private equity, which affect variance and therefore unsystematic risk, are not priced. In the CAPM framework unsystematic risk can be diversified away, leading to the prediction that private equity will not earn higher (systematic) risk adjusted returns. In other words, no private equity premium or liquidity premium exists.

However, the CAPM is based on a number of simplifying assumptions. Brown, Elton, Goetzmann and Gruber (2003) identify no less than ten underlying assumptions for the CAPM to work.20 Summarized, it comes down to the following three assumptions;

(1) Investors care only about the mean and variance of their portfolio returns (2) Markets are frictionless

(3) Investors have homogeneous expectations

19_{Grinblatt, Mark and Sheridan Titman, 2002, Financial markets and corporate strategy, 2}nd_{edition, McGraw-Hill, New York, page}

151

(25)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 25 From all the above it is immediately clear that these assumptions are violated in the organized private equity market. Especially the assumption about frictionless markets doesn’t apply. Illiquidity, management fees and information asymmetry all stand in the way of the private equity market to be frictionless. So if CAPM doesn’t hold, the nonsystematic risk might be priced and private equity might trade at a premium or a discount. This is confirmed by a study performed by Jones and Rhodes-Kropf (2003). They find that principal-agent problems in the private equity market, based on information asymmetry, lead to returns that are higher when unsystematic risk increases. Their results hold, even when the investor can fully diversify and private equity markets are competitive.21

To summarize, in a frictionless and completely competitive private equity market the theory predicts that private equity funds returns are only determined by systematic risk. However, the private equity market it characterized by a number of market inefficiencies. Therefore, the market does not fulfill the assumptions underlying the CAPM. Because investors seek compensation for the high illiquidity and unsystematic risk of the organized private equity market, one expects a (systematic) risk adjusted premium on the return to private equity investments.

2.5 Reasons to invest in private equity

Before we turn to chapter three were we will see whether or not private equity earns a risk adjusted premium I will conclude this chapter with a short summary of reasons for institutional investors to invest in private equity. Despite the market imperfections discussed here private equity is popular among institutional investors. The reasons can be divided into financial reasons and strategic reasons.

21_{Jones, Charles and Matthew Rhodes-Kropf, 2003, The price of diversifiable risk in venture capital and private equity, Working}

(26)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 26 2.5.1 Financial reasons

Of course institutional investors are attracted by high reported returns on private equity in the past. Several buy outs and venture investments have shown enormous returns. Institutional investor’s portfolio managers simply might want to grab a share of these profits.

According to standard portfolio theory an investor is interested in the risk and the return of his total portfolio of assets. For an institutional investor private equity is part of a broader portfolio of financial assets. When the correlation between private equity and the other assets in the portfolio is low enough adding private equity can result in diversification benefits. Fenn (1997) states that in many cases private equity investments are part of a portfolio of so called alternative assets, held for diversification purposes. These alternative assets include among others emerging market investments, commodities like oil and gas and investments in real estate.22

A final financial argument is derived from the balance sheet of namely pension funds. The liabilities on their balance sheet typically have a long time to maturity. Private equity investments, as we have seen, in general are investments with a maturity of ten years plus. This quality ensures a good match for the pension fund between the duration of its assets and liabilities.

2.5.2 Strategic reasons

Next to purely financial reasons aimed at improving the performance of the investment portfolio there are also non-financial strategic reasons for an institutional investor to invest in private equity. These reasons apply mostly to the case where the institutional investor is a bank or commercial institution.

First of all, economies of scope might exist between investing in private equity and the other activities employed by the institution. Participating in the private equity market can also serve as a means to open doors to other markets. Ljungqvist and Richardson (2003) interviewed a limited partner who said his objective was not only to obtain a high risk

(27)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 27 adjusted return but also “to increase the likelihood that the funds partners will purchase the services our limited partner’s corporate parent has to offer”.23

In line with this is a study documented by Phalippou and Zollo (2005a) which shows that banks invest in venture capital in order build relationships that can be used for lending activities in the future. The same study shows that some limited partners invest in private equity in order to stimulate the local economy, again with the objective to reap the benefits in the future.24

23_{Ljungqvist, Alexander and Matt Richardson, 2003, The investment behavior of private equity fund managers, Working paper, New}

York University, page

(28)

3 Performance of the US private equity market

In the previous chapters we have seen that private equity investing has experienced massive growth over the recent decades. Its proponents argue that investing in this asset class offers the investor a superior risk return trade off compared to public equity. As we have seen there is a theoretical foundation to support this claim.

In this chapter I will present the outcomes of the different studies that test this claim. These studies performed research into the performance of US private equity investments. I will start with a discussion on methodologies used by the different authors. Estimating private equity returns is difficult and several potential biases must be avoided. Next I will look at the different ways in which the return outcomes are evaluated, that is, which performance measures are used.

Finally, the main results of the studies are summarized and compared. Hopefully this will answer the question whether private equity offered superior risk adjusted performance in the US. In the end this should lead to an advice to Dutch investors what to do with private equity in their portfolios.

3.1 Methodologies and biases

3.1.1 Potential biases

Virtually all studies start with pointing out the difficulties the authors had to overcome to perform their research and arrive at reliable outcomes. These difficulties are inherent in the characteristics of private equity as an asset class and the workings of the private equity market.

(29)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 29 be successfully exited, leading to an upward biased estimate of the underlying mean return of all (including failed) projects and funds.

A second potential sample bias problem arises when a fund has not yet liquidated investments or investments that haven’t shown any activity for a number of years, so called living dead investments. The problem is to decide whether to include these funds in the sample and if so, how to value the ongoing and living dead investments.

Normally a return is measured with a time series of historical observed returns. There are two reasons why this approach does not work for private equity. First, the lack of transparency of investments in private equity results in a far from complete data set on which research can be based. Second, the private equity market is characterized by a highly imperfect and sometimes even missing secondary market. This means that for a private equity fund no time series of historically observed market returns exits. For buy out funds, only when the investment is made and when the investment is exited transaction prices, and thus market values, can be observed. In the case of venture capital funds intermediate financing rounds also reveal a market value but this still leaves an irregular and small sample of values.25 One way to deal with this problem is to estimate the Net Asset Value (NAV) of a fund and use this value to calculate a return. The NAV is the value of all investments in the portfolio of the private equity fund that did not exit yet. The problem here is that the NAV is based on accounting numbers and not on market valuations. These accounting numbers can be manipulated and be subject to a valuation bias, making the returns they produce subject to the same bias.

3.1.2 Methodologies and performance measurement

The most common way in the private equity literature to calculate return is the use of the Internal Rate of Return (IRR). The IRR is the discount rate that would result in a net present value (NPV) of zero for a series of cash in- and outflows. Mathematically this results in the following equation;

25_{Cochrane (2005) argues that these intermediate valuations should not be used for constructing returns. A new round determines the}

(30)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 30 (3.1) 0 ) 1 ( 0 = + ∑T₌ t _t t IRR CF where,

CFt= Cash flow to or from investor at time t

IRR = Internal rate of return

Although the internal rate of return represents the standard performance measure for private equity investments and is even recommended by the EVCA, it is associated with some shortcomings with respect to evaluating the investment return. When calculating the IRR the assumption is made that distributed capital during the life of the investment is reinvested at the IRR. This is not a very realistic reinvestment hypothesis and will lead to distorted return figures. On the one hand, extremely well performing investments expel very high IRR's owing to a reinvestment assumption with high project rates. On the other hand, an implicated low reinvestment rate decreases overall IRR of badly performing investments. 26

To use the IRR as a performance measure the excess IRR can be calculated. This is the IRR of the portfolio relative to the IRR of a chosen benchmark. A positive excess IRR indicates over performance of the private equity investment relative to the chosen benchmark.

(3.2) ER_pe =IRR_pe−IRR_bpwhere,

ERpe= Excess IRR of the private equity investment IRRpe= Internal rate of return private equity investment IRR_bp= Internal rate of return benchmark portfolio

The second method used to evaluate performance of private equity investments is the public market equivalent (PME). Basically the PME determines how many dollars one would need to invest in the chosen benchmark to generate a return equal to that of a one dollar investment in private equity on a present value basis. It assumes that intermediate cash flows are

26_{Schmidt, Daniel, 2003, Private equity, stocks- and mixed asset portfolios: A bootstrap approach to determining performance}

(31)

reinvested in the public benchmark and determines the value of the cash flows accordingly.27 This is presented by the following equation;

(3.3) ) ) 1 ( ( ) ) 1 ( ( 1 , 0 , 0 1 , 0 , 0 − = = − = = + ∏ ∑ + ∏ ∑ − = n b t n pe t T t n b t n pe t T t pe r CFI r CFO PME where,

PMEpe= Public market equivalent private equiyt investment

CFOtpe= Cash outflow for private equity investor at time t

CFItpe= Cash inflow to private equity investor at time t

r_bn= return on benchmark portfolio over period n

A PME of more than one is a sign of over performance relative to the chosen benchmark.

As Ick puts forward, discussion is possible about what discount rates to apply to the different cash flows.28 The discount rate applied to cash inflows and outflows should resemble the risk of that particular cash flow.

The Sharpe ratio is one of the most frequently used performance measures. It is defined as the ratio of the excess return above the risk free rate and the standard deviation of the portfolio. Simply stated the Sharpe ratio measures the compensation in return for taking on one unit of risk. The measure can be written as follows;

(3.4) pe f pe R R E S σ − = ( )

Specialized investment managers like private equity investors are likely to construct a portfolio containing a high degree of diversifiable risk Ultimate risk reduction is not the objective of these portfolio managers. According to van der Meer et al. (2007) Jensen’s alpha is a suitable performance measure for these situations. This alpha is the difference

(32)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 32 between the expected return based on systematic risk (beta) and the realized return29 Jensen’s alpha is measured with the following equation;

(3.5) Rpe−Rf =αpe +βpe(Rbn −Rf)where,

=

pe

R Return on private equity

=

pe

α Alpha for private equity

=

pe

β Beta for private equity

=

bn

R Return on benchmark portfolio 3.2 Results of empirical studies

There are a number of possible ways to classify studies on risk and return of private equity performance. I follow the distinction Phalippou and Zollo (2005a) make. They make a distinction between studies that focus on risk and return of individual private equity investments made by general partners on the one hand and studies that look at the performance of private equity funds on the other hand.30 The first type is the most pure form of studying the performance of private equity as an asset class since returns are measured gross of fees. It tests the quality of the available private equity investments universe. The second type judges private equity from the viewpoint of an investor. Since the average investor can only invest in private equity through a fund or a fund of funds, the performance of private equity funds is his primary concern.31

29_{Meer, van der Robert, Auke Plantinga and Roelof Salomons, 2007, page 9 and 10} 30_{Phalippou, P. and M. Zollo, 2005a, page 4}

(33)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 33 3.2.1 Studies of individual private equity investments performance

3.2.1.1 Venture Capital Indices

The S&P 500, the NASDAQ and the Dutch AEX index are examples of indices of publicly traded securities. These indices can be used as a benchmark to evaluate the performance of a portfolio of traded securities or to asses the historical performance of the asset class captured by the index. Venture capital, being a non-publicly traded security, does not have such an index or benchmark. Quigley and Woodward (2003) and Peng (2003) have tried to overcome this problem by constructing an index of venture capital. According to Quigley and Woodward (2003), this index can be useful in evaluating the performance of a particular venture capital fund or limited partnership, in comparing the performance of venture capital to that of other asset classes and in deciding on the appropriate portfolio allocation for an investor, that is whether or not to include venture capital.32 Both indices show the returns of direct investments in venture capital, gross of any fees.

Peng (2003) uses data on 16.720 financing rounds by venture capitalists in the US in the period 1987-2000. He eliminates the financing rounds in which information about either the date or amount of fundraising or, in the case of an exit, date and return on exit is missing. This leaves him with a database of 12.946 observations from which to construct the venture capital index. The resulting index is a portfolio of the venture capital investments in the database that meet the information criteria.

Peng (2003) starts by creating two sub indices. One for investments that have been successfully exited through an IPO or an acquisition and another one for all other investments that have been liquidated. The next step is to assign weights to both indices. Peng (2003) makes assumptions about observable characteristics of the investments that influence the probability of success. He shows that age of the firm, the total number of financing rounds in one firm and the relative size of the last financing round are

(34)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 34 significant predictors of success of an investment.33 He uses these factors to assign weights to the good and the bad sub indices. The final step is to make assumptions about the return of liquidated investments. Peng (2003) sets this return at -80%.34 Now all the information is available to construct a venture capital index.

The sample set of Quigley and Woodward (2003) consists of 12.553 pricing events in the US in the years between 1987 and 1990, the same period as investigated by Peng (2003). The main problem identified by Quigley and Woodward (2003) is that pricing events happen on an irregular basis, and more importantly, not every pricing event necessarily reveals a value. Because of the private nature disclosing the value at which the pricing event took place is not obligatory. This results in a biased sample of events that do show value.

To estimate the degree of bias Quigley and Woodward (2003) use all pricing events that do reveal value to estimate the probability that a pricing event reveals value. This probability is used in a repeat sales method to overcome the problem of irregular pricing events. The way they correct for selection bias makes their research different from that of Peng. In fact, Quigley and Woodward (2003) criticize Peng (2003) for the crude way in which he corrects his sample for selection bias.35

Quigley and Woodward (2003) analyze their results in two ways. First they compare the venture capital index to the S&P 500 and the NASDAQ index on returns and standard deviations. Second, they measure the correlation between their venture capital index on the one hand and these public market indices and the bonds and T-bill markets on the other. Table 3.1 summarizes the results of both Peng (2003) and Quigley and Woodward (2003).

33_{Peng, L., 2003, Building a venture capital index, Working paper, Yale University, page 19}

34_{Peng, L., 2003, page 26, of course one can argue about this number. Setting the loss rate lower would reduce the performance of the}

venture capital index.

(35)

Table 3.1 This table shows the average annual returns and standard deviations of the venture capital indices composed by Quigley and Woodward (2003) and Peng (2003). It also shows the correlation of both indices with two public market indices in the period 1987-2000

Peng (2003) finds an average annual return of 55,18% for his index of venture capital. This return seems extremely large. However, if you take a closer look at his results it shows that the high average return is for a great deal explained by a return of 168% in 1996 and a mind blowing 681% in 1999.36 This also explains the high volatility. Apparently, the high return is driven by a few major successes. This shows the importance of selection skills from the part of the general manager. Peng (2003) also finds a high correlation of the NASDAQ index return with his venture capital index. Not shown in the table 3.1 is that also the correlation between the standard deviations of both indices is large.37 Peng (2003) therefore concludes that his venture capital index shows resemblance to the behavior of small NASDAQ stocks.

Quigley and Woodward (2003) find an annual return for venture capital of only 8,28%. This is less than the return of the S&P 500 and the NASDAQ, with a higher standard deviation. Venture capital seems unattractive in this perspective. However the authors find very low correlation between the returns on venture capital and the returns on the other asset classes. This makes venture capital a potentially interesting candidate to include in the investment portfolio.

36_{Peng, L., 2003, page 40} 37_{Peng, L., 2003, page 28} Quigley and Woodward (2003)

Peng (2003) S&P 500 NASDAQ

Return 8,28 % 55,18 % 11,2% 13,4%

Standard deviation 14,56 % 28,11% 9,11% 14,45%

Correlation with S&P

500 0,044 0,038 _ 0,84

Correlation with

(36)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 36 Using mean variance analysis Quigley and Woodward (2003) try to decide whether or not venture capital should be included in the optimal portfolio when these three indices and bonds and T-bills are the available investment opportunities.

They arrive at the following conclusions. When short selling is allowed and risk aversion is low a large fraction of the portfolio is allocated to venture capital. The average institutional investor however cannot sell short unlimited. In the more realistic case of limited or no short selling the optimal portfolio has 10 to 15% of its assets invested in the venture capital index at moderate risk levels. When the risk level rises, the focus shifts to the NASDAQ index in stead of venture capital. In the best scenario, inclusion of venture capital in the investment portfolio increases its annual return by a half percent for moderate risk levels.38

Related to the work of Peng (2003) and Quigley and Woodward (2003) is the study of John Cochrane (2005). Cochrane (2005) doesn’t build an index but measures risk and return of venture capital investments and its correlation with the market directly. The central question he asks is whether venture capital investments behave the same way as publicly traded securities. Do venture capital investments show better risk adjusted performance, do they resemble any publicly traded security and how large are their betas and residual risk? Cochrane (2005) identifies selection bias as the main problem in evaluating venture capital investing. Overcoming this bias is therefore the main focus of his work. Cochrane (2005) uses a dataset comprising of 16.613 financing rounds in the US in the period between 1987 and 2000. For this set he finds a mean arithmetic return of 59% and a standard deviation of 107%.

Cochrane (2005) performs a regression on the S&P 500. He finds a beta of 1,7; venture capital has higher systematic risk than the S&P 500. Jensen’s alpha is 32%, indicating that the diversifiable risk is priced. Although he recognizes some reasons as why risk and return of venture capital might differ from publicly traded securities, to Cochrane these returns and alphas are still surprisingly large.39

38_{Quigley, J. and S.E. Woodward, 2003, page 19}

(37)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 37 When comparing to publicly traded securities Cochrane (2005) finds that a sample of small NASDAQ stocks shows the same behavior as venture capital. High mean arithmetic returns and very high volatility for individual stocks. The remaining puzzles are not exclusive for venture capital. When performing a regression however Cochrane (2005) finds only a small beta of the venture capital on the small stock portfolio. This leads him to conclude that NASDAQ small stocks performance does not explain the high returns of venture capital. The two events are similar, but they are not the same event.40 A more methodologically based study has been performed by Woodward (2004) Woodward (2004) proposes an expansion of the Jensen’s alpha performance measure which will give a more accurate estimate of the beta, alpha and the correlation. She identifies two problems that are related to the way in which returns for venture capital are calculated. The first one is stale pricing problem. This occurs when an asset is infrequently priced, like a venture capital investment. When regressing a venture capital index on a public benchmark index this will understate beta and overstate alpha because the underlying time intervals are not synchronous.41

The second problem comes from the structure of venture capital investments. A financing round can take several months to be completed. A financing round is considered a pricing moment so the investment will show the same value during these months. This leads to serial correlation over time in the measured returns on venture capital. To adjust for these problems Woodward (2004) suggests including several lags of the benchmark index as independent variables in the regression to better capture the before mentioned problems. Woodward (2004) performs a regression using a venture capital index constructed by the Sand Hill Company. This index is constructed out of individual venture capital

investment in the period 1989-2004; returns are gross of any fees. As independent variable the return on the Wilkshire 5000 index is used. The Wilkshire index is broader than the S&P 500 and Woodward wants her investigation of private equity to capture as much of the variation in value of small public companies as possible.

40_{Cochrane, John C., 2005 page 50-51}

(38)

Doctoraalscriptie - Matthijs van der Hoeven - Rijksuniversiteit Groningen 38 The results of the regression show that at least 18 monthly lags (one and a half year) show significant influence on the returns of the index. Woodward finds a corrected beta of 2, an annualized alpha of 0,36% and a correlation with the index of almost 0,7.42 These results question the quality of venture capital as an investment opportunity. After fees the alpha will turn negative, indicating that the return does not even cover systematic risk. Moreover, the high correlation suggests that it will be hard to sufficiently reduce non systematic risk through diversification. The low returns on venture capital are in line with earlier research performed by Quigley and Woodward (2003). The high correlation is a new fact, probably caused by the inclusion of several lags of the public market index.

3.2.1.2 Performance of individual private equity investments

In this section I will discuss papers that investigate the performance of individual private equity investments, with the sample including both investments in venture capital and investments in buy outs. As Ick (2005) puts it, the focus is on “pure” investment returns prior to any pooling activity by private equity funds. The authors are interested in project returns and the quality of the private equity investment universe. They strive to assess whether private equity projects generate adequate return premiums over public market investments.43

Schmidt (2003) analyzes the risk and return characteristics of private equity investments. He also investigates potential diversification benefits when private equity is included in an investment portfolio. Schmidt (2003) decides to only include completed investments in his sample, investments which are successfully exited or liquidated. This allows him to focus solely on investments for which the complete cash flow pattern from draw down to exit is known. This leaves him with a final sample of 642 completed investments made in the US in the period between 1980 and 2002.

The cash flow pattern of these completed investments is used to calculate the average IRR of the private equity investments. This IRR is then compared to the IRR of a

Review of empirical research on the performance of the US private equity market