An evaluation of private equity investments
By Emily Ledeboer 10017216 19-01-2014
Abstract
The private equity market has been a rapidly growing market over the past three decades. Some research has been done on its risk and return, but their results differ in many ways. In this paper I used listed private
equity firms as a proxy for the private equity market. By calculating its risk and return profile and by comparing it to the public market, I have tried to find out whether it is possible to outperform the public
1. Introduction
In the last three decades, the private equity market has grown to an important source of funds for start-up firms, private middle-market firms, firms in financial distress and public firms seeking buyout financing. From 1980-2010, over $1,1 trillion has been raised by U.S. buyout funds and roughly $700 billion by venture capital firms (Fenn, Liang and Prowse, 2010)
Despite its increased significance for corporate finance, the private equity market has received relatively little attention in academic literature and there is still a lot to gain from new studies. The lack of knowledge is due to the fact that data is very hard to come by. For when a firm is private its records are not reported in public records. As a
consequence, results from different studies testing more or less the same, vary considerably. Reasons for these different outcomes are that data sets are individually gathered and therefore subject to various selection biases. Another feature that makes the results less accurate is that only book values are available, as market values don’t exist (Zimmermann, Bilo, Christophers and Degosciu, 2004).
However, there is a form of private equity that is obligated to enclose its records publicly. These firms are called Listed Private Equity firms (LPE). It has established itself as an adequate proxy for traditional private equity (TPE), as the underlying business structure is the same. The advantage of this market segment is that since market prices and other figures are publicly available and regulated by authorities, performance
measures are more reliable. To be ranked as a LPE firm, at least 50% of the firm’s equity block should consist of private equity investments, together with some other restrictions (Statman, 1987). The major advantage for investors of LPE to TPE is that its shares are traded on the public stock market. This creates the possibility for small investors to diversify their portfolio to the private equity market, which is not possible with TPE as only large investors have the required amount of funds to invest (Bergmann,
Christophers, Huss and Zimmermann, 2009). However, besides their equal business structures, there are a few drawbacks of LPE to TPE. The underlying exposure of LPE can vary considerably from financing a portfolio with huge parts of borrowed money, to excess cash with no suitable investment options. The risk profile of the portfolio is heavily affected by the risk profile of the money used, which makes the risk profiles of
LPE and TPE less comparable. Nevertheless, LPE firms will be held as a proxy for TPE in this thesis as they share the most similarities.
The central question to be answered in this thesis is the following: Can one, by investing in private equity, while considering its different approach to investing and portfolio setup, outperform the public equity market. First, former studies are evaluated to get some intuition about these two different kinds of investments. The results of the studies on TPE contradict each other in various ways. This is mainly due to their sample selection as they used different approaches to gather data to overcome selection biases. This bias was overcome in the two studies on LPE as data for these firms is available in public records. The first study by Zimmerman et al. (2004) created a benchmark for TPE. The risk and return profiles of LPE and TPE are outlined and compared and the study shows that their business structures are so similar, that they base their study on the assumption that the risk and returns calculated for LPE can be used to make assumptions about the performance of TPE. The second study by Lahr and Herschke (2009) extended this by subdividing the sample in internally and externally managed firms.
The study of this thesis is based on the study by Zimmerman et al. (2004). A new sample is drawn using equal assumptions about liquidity constraints and its performance is measured. Thereafter, the study is taken a step further as the sample will be divided in sub-periods to be able to evaluate the performance more closely. This study will test the hypothesis that investments in private equity outperform the public equity market.
Firstly, in chapter 2, former studies on TPE and LPE will be evaluated to develop some intuition about previous performance of private equity in general. In chapter 3 a new sample will be drawn of LPE from various selected sources. In chapter 4 an
empirical study on the risk and return of LPE firms will be performed. The model used is based on the one used by Zimmermann et al. (2004). After graphing these regressions, the CAPM model (Womack and Zhang, 2003) will be applied to see if alpha is
significantly different from zero. Then in chapter 5, the results are analyzed and finally in chapter 6 a conclusion will be drawn on whether private equity outperforms the public market or not.
2.
Related literature
Traditional private equity versus listed private equity
Traditional Private equity (TPE), also referred to as unlisted private equity, is usually structured as a limited partnership. The two parties involved in this operation are limited partners (LPs) and general partners (GPs). LPs are the main capital providers of equity such as corporate and public pension funds, endowments, insurance companies and wealthy individuals. They commit to an illiquid investment during the lifetime of the investment, which is mostly around a decade. During this period they have no influence on the way the project is managed. If everything goes well, the partners are compensated with a positive return for their investment at the end of the lifetime. The part of GPs is mostly assigned to a PE firm that manages the acquired firm during its lifetime. The firm invests the money of the LPs, with which they restructure the acquired firm and are compensated with performance-based fees. These firms exist of professionals with a lot of field experience and have access to tools that a public firm has not, such as the possibility to increase leverage, reduce agency problems and increased incentives to monitor more closely (Kaplan and Strömberg, 2008).
Listed private equity (LPE) funds are funds that exist for >50% of private equity funding. These private equity blocks provide GPs with a relative big vote in the way the organization is managed, but they have to work alongside the existing management. Private equity blocks in these kinds of funds result in immediate exposure to the risk and return of private equity investments to shareholders that hold shares in portfolio’s that invest martly in such funds. It basically provides shareholders with a more diversified portfolio (Bergmann et al., 2009).
TPE and LPE firms have a lot in common when referring to their performance, which makes them good proxies. Besides that more than half of the organization’s equity is provided by a private source for LPE, its business model is set up in the same way and their investment and financing styles are highly comparable. Therefore, they exhibit the same risk and return potential as their unlisted counterpart (Bergmann et al., 2009). This is an important feature, as in contrast to TPE data on risk and return is available and can be used in several ways to imitate risk and return structures for TPE.
Fleming and Cumming (2010) summarized a number of important studies that differ in sample period, sample size, sample selection and model to determine the performance for traditional and listed PE firms. As can be seen in table 1, there are two methods evaluated for calculating performance with TPE, and one method used for LPE. It may be noted that less research is performed on LPE compared to TPE, and that the results obtained in TPE studies vary substantially. Apart from these studies, one additional study is added which is performed by Jegadeesh, Pollet and Kräussl (2009), who used a different approach to estimate the risk and return of TPE. They used market prices of exchange-traded funds of funds to measure performance, and the abnormal returns are determined by the market expectations from TPE fund investments, based on the net present value of these investments. Besides the two papers mentioned by Fleming and Cumming in table 1, no additional studies on the risk and returns of LPE were found.
Table 1
Summary of empirical results in previous studies by Fleming and Cumming (2010) The table contains sample period and size for each study and relevant performance measures. G.O.F refers to gross-of-fees studies. 0* denotes no significant difference to corresponding benchmark. Beta may vary between single-factor models, Fama-French Three factor models and a Dimson Beta. For PME, the public market return equals 1.00. Standard deviations are presented in parentheses.
TPE studies: sample sample abnormal
Performance of period size return Beta Alpha
GPs investments
Woodward and Hall (2004) ’80-’04 - - - 8,5%
Hwang et al. (2005) ’87-’03 15.583 0* 0,6 1%
Cochrane (2005) ’87-’00 7.765 43% 1,9 32%
TPE studies: sample sample abnormal IRR(%p.a.) . PI
Fund-level studies period size return Beta PME Equally-W Value-W
Jones&Rhodes-Kropf (2003) ’69-’02 1.245 0* 1,05 - 16.38% (46%) 9.18% (39%)
Ljungqvist&Richardson(2003)’81-’93 73 5%-8% 1,09 - 19.8% (22%) 18.1% (22%)
Phalippou&Gottschalg (2009) ’80-’93 852 -3%--6% 1,03 - 12,13% 12,22% 0,96-0,99 Kaplan and Schoar (2005) ’80-’01 746 0* - 0.96-1,05 18% (26%) 17%(31%)
Lossen (2006) (G.O.F) ’79-’98 134 - - 3,075 50.2%(32%) -
LPE studies: sample sample Equally-weighted value-weighted
Period size Beta Alpha SR Beta Alpha SR
Zimmerman et al (2004) ’86-’03 114 0,6 10,18% 0,57 1,2 -1,20% 0*
Lahr and Herschke (2009) ’86-’08 274 1,2 7,50% 0,13 1,7 0* 0,049
Performance studies on traditional private equity
Woodward and Hall (2004), Hwang, Quigley and Woodward (2005) and Cochrane (2005) studied performance based on a measure of GP’s investments where they analyze
the risk and return of venture capital investments. Although the three of them initially measure the same and they all use the CAPM model, their outcomes differ substantially. This is due to a few fundamental differences in the way they approached the data needed for the model and assumptions to overcome selection biases such as the minimum investment required, the length of the initial investment and the moment the data was obtained.
Woodward and Hall (2004) selected company data based on companies that were likely to go from private to public and excluded the ones that were private due to
leveraged buyouts, management buyouts and private placements in public companies. The index is built based on company-level pricing data, not fund-level return data. The purpose of their paper was to create a valid benchmark index to the S&P500 the same way market-wide indices are used for traded securities. This index could in turn be used for asset allocation analysis and portfolio performance evaluation. The result in Table 1 shows an alpha of 8,5%. This means that the return exceeded the expected return for the amount of risk the portfolio was exposed to, which indicates that on average private firms outperformed the public market based on risk for return.
Cochrane (2005)’s main focus when he computed the risk and return of venture capital investments was to overcome selection bias. Selection bias initially existed, as valuations in his study were only observed when a firm went public, received new financing or was acquired. It’s likely that by observing these kinds of valuations mostly firms are selected that have high annual returns. To reduce this selection bias he used maximum-likelihood estimates. By calculating the arithmetic average of the risks and returns, he measured a return of 43% and an alpha of 32%. These seem to be large compared to other studies, but without implementing the arithmetic average, the average return would 698% with an alpha of 462%. These differences result among other things from large standard deviations of individual firms and a small sample.
Woodward and Hall (2004) compared the paper written by Hwang et al. with Cochrane’s method and point out the similar use of log in their risk and return analyses. They argue that the results of these two studies suffer under this choice, because the actual distribution of returns departs significantly from log-normality. Even though Cochrane and Hwang et al. both use log in their methodology, their results differ
substantially. The beta obtained by Hwang et al. is three times smaller than Cochrane’s, and their abnormal return isn’t significantly different from zero, where Cochrane’s equals 43%. Hwang et al.’s alpha is even lower than Woodward and Hall’s. A reason for these different results may lie in that Hwang et al.’s sample is twice as big, which gives a better representation of the market.
Jones and Rhodes-Kropf (2003), Ljungqvist and Richardson (2003), Phalippou and Gottschalg (2009), Kaplan and Schoar (2005) and Lossen (2006) all investigated relative performance of private equity to public equity markets based on fund-level studies. All studies use different approaches to data selection and measurements, which results in different views of performance of private equity.
Jones and Rhodes-Kropf (2003), Ljungqvist and Richardson (2003) and Kaplan and Schoar (2005) all conclude that on average private equity returns net of fees equal the S&P 500 or outperform them. Ljungqvist and Richardson (2003) argue that this is due to the illiquidity of the investment, and that the higher return should be seen as a premium for illiquidity. Among these three papers they seem to have the least bias in their sample as they obtained the most accurate and detailed data. Furthermore, Kaplan and Schoar (2005) show in their results a positive but concave relation between the size of a fund and its return. So it can be assumed that the greater the fund, the less positive returns will increase.
Lossen (2006) noticed this concave relationship and took a different view to fund performance, namely by investigating the effect of different diversification approaches. He found that private equity funds that were formed of specific financing stages, industries or countries, earned higher returns than more diversified funds composed in other fields.
Phalippou and Gottschalg (2009) find contradicting results to the former four papers. According to them a bias exists towards better performing funds in the data selected for which should be corrected and furthermore that a large part of performance is driven by inflated accounting valuation of ongoing investments. Considering these
assumptions they find that returns net of fees underperform the S&P 500 by 3% to 6%. They estimate fees for GPs on 6%, which indicates that GPs on average don’t make a loss on the investment, but that there is no room for profits.
Jegadeesh, Pollet and Kräussl (2009) studied risk and returns of PE based on the market prices of exchange-traded funds of funds. The advantage of this kind of data is that both the market value and the amount of money that the funds of funds raise are observed. The abnormal returns are determined by the market expectations from TPE fund investments, based on the net present value of these investments. Their estimated abnormal returns are approximately 1% per year, which lies between values of former studies discussed. Their estimated beta is nearly one, which coincides as well with the former studies discussed.
Performance studies on listed private equity
Zimmerman et al. (2004) performed a study on listed private equity (LPE) firms. Their goal was to create a valid benchmark to compare to the traditional global stock market in order to measure risk and returns. A major contribution to existing studies was that they calculated these based on three different approaches named ‘A value weighted portfolio,
buy-and-hold (partially rebalanced): VW-BH’, ‘An equally weighted portfolio, fully rebalanced: EW-RB’ and ‘An equally weighted portfolio, buy-and-hold (partially rebalanced): EW-BH’. These three portfolio strategies differ in weights put on
investments and rebalancing choices, which results in different outcomes as the sample size increases over time.
Initially a sample was selected of 274 vehicles. As mentioned before, one of the major differences of LPE with traditional stock markets is its higher illiquidity because at least 50% of its capital is invested in TPE. This means that at least 50% of its investments are invested on the long term and it is not possible to liquidate them at any moment. The sample had to be revised to correct for liquidity in order to be able to draw an accurate conclusion when comparing it to its benchmark. Their selection criteria consisted of the following 5 to correct for liquidity, which resulted in a sample size of only 114 vehicles. Liquidity restrictions by Zimmerman et al. (2004):
1. We require a minimum of 30 weekly price observations in order to ensure accuracy of parameter estimates.
2. The vehicles must have a minimum average market capitalization of $2 million.
3. To assure a minimal trading activity, we impose a minimum average trading volume of 0.1% per week; the (relative) trading volume is defined by the ratio of the trading volume multiplied by the
price, and the market value of the vehicle.
4. Minimum trade continuity must be satisfied. This is measured by the percentage of weeks (here, 15%) within which at least one transaction occurs.
5. Finally, we require a maximum average bid-ask spread of 20%, which is defined relative to the arithmetic average of the bid and ask quotes.
The three different strategies provide different results when compared to the benchmark, which is the MSCI World index. The EW-RB portfolio results in the highest average return of 15,99% compared to an average return of 5,91% in the VW-BH
portfolio and 5,43% in the EW-BH portfolio. Adjusting for risk results in a comparable result. Based on these findings and small R2–coefficients of the underlying regressions, they argue that private equity exhibits a large diversifiable variance component with respect to standard equity investments (Zimmermans et al. 2005).
Lahr and Herschke (2009) based their paper on Zimmermans et al. (2005). They extended their research by subdividing LPE firms based on their organizational structure in internally and externally managed firms. In their results they show that internally managed firms have a higher exposure to market risk than externally managed firms. When looking at their data in table 1, it can be seen that the performance of the value-weighted portfolio has a Beta of 1,7 and an alpha not significantly different from 0. The equally weighted portfolio however, shows a beta of 1,2 and an alpha of 7,5% which indicates a higher return for risk exposure. They argue in their paper that these different risk profiles are induced by a difference in cash flows such as management-fees and carried interest.
3.
Data-description
To provide good estimates of the performance of the private and public equity market, the data that will be used for the calculations have to be well considered. The sample representing the public market is required to be representative for the global public equity market. Therefore, the MSCI world index is chosen. The MSCI global equity indices are used as a benchmark to all sorts of markets and converge over more than 70 countries. They apply a consistent index methodology, and serve as the basis for more than 500 exchanged traded funds over the world (Fletcher, 2000).
The sample for the private equity market is composed as follows. In the literature review we have learned that LPE firms serve as an appropriate proxy for TPE firms, as the business model is similar, and thereby is their exposure to risk and return. LPE vehicles are identified from the following three sources: The LPX index family, S&P’s LPE top 10 and LPEQ Europe. These sources are selected as they apply liquidity constraints, which make them better proxies for TPE. The liquidity characteristics are similar to those set out by Zimmerman et al. (2004), whose model is representative for the tests performed in this study.
LPX group (LPX-Group, 2013)1 computes LPE indexes to measure performances of private equity. These indices are created to fulfil both the needs for the financial market and academic research. LPE are categorised in different sections for index calculation such as LPX 50, LPX major market and LPX composite. For each group, liquidity constraints vary a little but in general they require the following:
• A maximum average bid-ask spread of 4%
• An average minimum market capitalization of €20 million
• An average minimum trading volume per trading day measured relative to the market capitalization of 0,05%
• A minimum trade continuity >75%
Standard and Poor’s Listed Private Equity index2 consist of firms that meet their special requirements to be selected as LPE firms. Stocks must have a three-month average daily trading value above US$ 500,000, total market capitalization should be above US $150 Million, it must be trading on a developed market-exchange, and the stock must trade, on average, at least 10,000 shares a day (S&P, 2013).
LPEQ Europe (LPEQ, 2013) is an association that aims to make private equity accessible, by selecting its listed counterpart. Its geographical focus is on Europe. Their LPE members3 are to be found on their website. Most of them are similar to the ones named by the LPX-Group and S&P’s and therefore it can be assumed that they comply to the liquidity constrains. The others are checked with the liquidity constraints set out by Zimmerman et al. (2004).
1
In the appendix table 4 shows the vehicles selected by LPX-group. 2
In the appendix table 5 shows the vehicles selected by S&P’s. 3 In the appendix table 6 shows the vehicles selected by LPEQ Europe.
Out of these three different sources a sample was selected of 102 vehicles. Out of these 102 vehicles 3 were excluded because data wasn’t available in DataStream resulting in a sample of 99 vehicles. Statman (1987) stated that a sample is a well-diversified portfolio if it consists of at least 30 vehicles. Therefore the time period for which the indices will be computed and compared will start at 2000, because at that time 39 vehicles existed. All data is extracted from DataStream and valued in $US, which takes exchange rates and interest parity into account.
4.
Model set-up
The model used to calculate the return of LPE vehicles is based on a paper written by Zimmerman et al (2004). They designed different strategies to calculate indices to measure the return of a portfolio. Two of them will be implemented here to measure the returns of these portfolios over a twelve-year period to be compared to the MSCI world index. The value-weighted index is the first one, and puts weights on each vehicle based on market capitalization. Basing it on market capitalization induces higher weight on the returns of vehicles with the highest market capitalization and thus more influence of their performance on the return index. It is known in general, that those vehicles that perform well have more access to capital to invest (Kaplan and Schoar, 2005). This will result in the following: firms that outperform the average market of private equity will exceed more impact in the value-weighted index, and so the index of return of this portfolio is expected to be higher than the equally weighted portfolio. Another feature that is taken into account in both measures is that the portfolio is rebalanced for each week that a firm might enter (Zimmerman et al. 2004). As can be seen in figure 1, only 39 vehicles existed in 2000, and over the years the number increased. This approach results in an selection bias towards better performing firms, as the ones that went bankrupt before 2012 aren’t taken into consideration.
The value-weighted index (VW)
This index puts weights on each vehicle based on market capitalization
It=It1 x
•
(Mit/Mt) x [(P it+ D it) /P it-1] (1) As we have learned in the literature review, Lahr & Herschke (2009) argue in their paper that the last term in this equation [….] equals a ratio of the two return indices[(P it+ D it) /P it-1] = R it/ R it-1 (2)
Applying formula 2 to formula 1 results in the following simplification:
It=It1 x
•
(Mit/Mt) x R it/ R it-1 (3)D it = Dividend of a single vehicle It = Value-weighted index at time t
It1 = Index of the previous period, at t=0 I=100
Mit = Market value of a single vehicle
Mt = Total market value of the sample portfolio P it = Price of a single vehicle
P it-1 = Price of a single vehicle of the previous period, at t=0 R=100 R it = Return index of a single vehicle
R it-1 = Return index of a single vehicle of the previous period, at t=0 R=100
The equally weighted index (EW)
This index puts an equal weight on each vehicle
It=It1⋅ 1/n ⋅
•
[(P it +D it)/ P it-1] (4) Applying formula 2 to formula 4 results in the following simplification:It=It1⋅ 1/n ⋅
•
R it/ R it-1 (5)D it = Dividend of a single vehicle It = Equally weighted index at time t
It1 = Index of the previous period, at t=0 I=100
n = Number of vehicles in the sample portfolio, rebalanced over time
P it = Price of a single vehicle
P it-1 = Price of a single vehicle of the previous period, at t=0 R=100 R it = Return index of a single vehicle
The values needed for these two strategies can be found in DataStream. Market value is indicated as MV and return index as RI. The value ‘n’ indicates the number of firms that exist in the sample at a certain time. The total return provided in DataStream assumes that cash-dividends are reinvested in the underlying stocks, so they are already accounted for (Campbell, Huisman and Koedijk 2001).
As a benchmark to represent investments in the public market the MSCI world price index is chosen. This is an index that is globally used as a benchmark to investigate risk and returns. It gathers data from over 70 countries and measures various indices categorized by country or region, sector, size and style. These two return indices, VW and EW and the price index of the MSCI, are shown in the graph below from 2000-2012. The horizontal axis is divided in 12 periods of one year starting in June. The vertical axis shows the index values. All three indices are calculated based on the same beginning date, 12-06-2012.
Graph 1: indices calculated weekly over 2000-2012, indicating the MSCI world index, VW-index and the EW-index.
The second analysis is based on a regression output. In contrast to the graph where only the performance is analyzed, the regression adjusts the returns for their exposure to risk. The papers studied in the literature review calculated Jensen’s alpha or abnormal returns
0 50 100 150 200 250 VALUE WEIGHTED INDEX EQUALLY WEIGHTED INDEX MSCI WORLD INDEX
to adjust them for risk. Here, the CAPM model is used for a similar purpose. The CAPM model provides the expected return of a vehicle compared to the market (Womack and Zhang, 2003).
( Ri,t – Rf,t ) = estimated alpha + estimated beta ( Rm,t – Rf,t )
The returns of the market sample, the MSCI return index, and the LPE sample are adjusted for risk, the risk free rate. The firms are situated among various countries, but as they are converted to $US, the three month treasury bill of the US serves as a reliable risk free rate and is extracted from DataStream. A positive alpha, significant at the 5%
significance level, will indicate that the LPE sample outperforms the market sample. A negative significant alpha will indicate underperformance to the market sample.
At first, the estimated alpha and beta of the LPE portfolio are computed over the whole period from 2000-2012. These outcomes are shown in table 2. Thereafter the sample period is split up in 4 equal periods of 157 weeks, to analyze the estimated alphas and betas for these sub-periods. These outcomes are shown in table 3 to 6. In the
equation, Y = (Ri,t – Rf,t) and X = Rm,t – Rf,t). SPSS is used to regress Y on X.
Table 2: Regression results of the CAPM model from 12-01-2000 to 18-01-2012 to calculate the expected return of a vehicle compared to the market.
Period Coefficient SD t P > | t |
1/628 Estimated β .710 .031 22.675 0.000
Estimated α .007 .001 8.072 0.000
Table 3: Regression results of the CAPM model divided in four equal sub-periods from 12-01-2000 to 18-01-2012 to calculate the expected return of a vehicle compared to the market.
Period Coefficient SD t P > | t | 1/157 Estimated β .554 .068 8.181 0.000 Estimated α .007 .002 3.794 0.000 158/314 Estimated β .629 .062 10.076 0.000 Estimated α .005 .001 4.557 0.000 315/471 Estimated β .770 .061 12.708 0.000 Estimated α .009 .002 4.045 0.000 472/627 Estimated β .787 .061 12.919 0.000 Estimated α .007 .002 3.705 0.000
5.
Results/analysis
Graph 1 shows that the private equity market, indicated by the VW- and EW indices, outperforms the public market, indicated by the MSCI index, at almost every date of the past 12 years. The indices based on the LPE sample fluctuate together with a high volatility. The value-weighted index measures a steady better performance
compared to the equally weighted performance. This result was already predicted, as it puts relative higher weights on vehicles with a higher market capitalization, which mostly are better performers than smaller vehicles. The MSCI index follows the trend of the curves with a smaller volatility than the EV index and the VW index. The reason for this more steady movement is that the market portfolio includes all firms and is therefore more diversified than the sample portfolio that consists only of LPE firms. Besides outperformance and similar movement, there is another aspect that deserves additional attention. After 2000, the graph reaches two peeks and two down lows and both the market and the private samples are affected as they follow the same movement. It is likely to assume that it is the result of an extern effect that can be explained by economic theory.
The CAPM model takes a closer look to the data, which results in more precise estimations of the movements of the two markets. The CAPM model is composed in a way that the estimated alpha indicates under- or outperformance of the LPE sample to the public market sample. First, in table 2, the estimation is performed over the whole sample period 2000-2012. With an estimated beta of 0.710 it explains what we already saw in graph 1, which is that the private and public markets move together. The alpha is estimated at .007, which indicates that over the whole sample period, the private market outperformed the public market. Both estimations are significant at the 5% level with a p-value of 0.000.
The graph shows two highs and two lows over the sample period. Therefore, the sample period was split up in four equal periods of 157 weeks to see if calculated over a smaller period, the private market still outperforms the public market. Table 3 shows the output results of the regression. The four estimated alpha’s are 0.007 0.005 0.009 and 0.007, all significant at the 5% level, so it can be concluded that based on these sample estimations, the private market outperforms the public market.
Comparing these results to our findings in the related literature shows whether the study is consistent with former estimations of the alphas and betas of TPE and LPE. First look at the whole sample period 2000-2012, with an estimated alpha of 0.007 and an estimated beta of 0.710. The estimated betas of former studies on TPE range between
0,6-1,9 and studies on LPE show beta’s between 0,6-1,7. The estimated beta lies closest to the lower bound of both two equity groups, and as it lies between the former estimated values, it can be said that our study is consistent. Betas of the four sub-periods are
respectively 0,554 0,629 0,770 and 0,787, where only the first beta is lower than the lower bound of former studies and I conclude that the study is still consistent. The alpha’s of former studies are very variable and range from -3% to 43%. It is not possible to draw a straightforward conclusion of whether or not our estimated alpha is consistent with former studies as they are calculated in different ways.
Now, some possible biases are considered that may influence the sample
estimations. The sample is randomly drawn from three different sources with a minimum of 30 vehicles, composed of firms located all over the world and with constraints for liquidity taken into account. The first bias that isn’t corrected for is a form of selection bias, as the sample doesn’t include firms that stopped existing before the sample was drawn. Firms were selected that were alive by 2012, based on sources that only track existing firms as for those sources it’s not relevant to indicate the faulted ones. The fact that these are not represented in the sample data results in a somewhat upward bias of the performance results. Another aspect referred to by Zimmerman et al. (2004) is that although a correction for liquidity constraints is applied, the illiquid features of limited private equity still result in a bias that exposes the sample portfolio to higher volatility. Another possible bias is the rebalancing of the portfolio. As is regarded in the literary review, Lossen (2006) examined the effect of diversification in portfolios and showed that the field of the private equity investment matters for its performance. In this paper there is only rebalanced for a value-weighted and an equally weighted portfolio, but no attention is given to possible diversifications of the sample based on industries. This could have resulted in different risk and return profiles.
Overall, the sample can be considered a good representation of the traditional private equity market with some biases that should be taken in consideration while evaluating the estimates.
6.
Conclusion-discussion
The private equity sector has developed and increased its investments during the last three decades. The question to be answered was therefore: can one, by investing in private equity, outperform the public equity market.
First, former studies were evaluated. The studies on traditional private equity showed contradicting results induced by selection biases, different methods of measure and the fact that no public records were available. In order to develop a more unbiased sample it is assumed that Listed private equity could serve as a valid proxy for traditional private equity, as its business models are highly comparable. The hypothesis that
investments in private equity will outperform the public equity market based on risk and return is based on these studies.
The return of our listed private equity sample was tested by computing the value-weighted-index and the equally value-weighted-index, based on a former study by Zimmerman et al. (2004). Graph 1 shows the VW-index, EW-index and the index of the market. From this graph, it can be concluded that the private market has a higher volatility than the public market, outperforms the public market based on performance indices, and that they are affected by similar economic factors as they move together. The regression results displayed in table 2 and 3 confirm these findings. The estimated alpha’s are positive even when adjusted for risk in the regression and the estimated beta’s lie between 0.554 and 0.787 which indicate that both markets move together.
Based on these findings I conclude that the private market outperforms the public market, though with a small percentage, and that my hypotheses based on former studies is true. There is still a lot of research to be done on this subject. For example, this paper could provide as a basis for a new study where the focus is shifted to other variables such as cultural, social and economic growth effects. Also, the equation used for the CAPM model could be extended with control variables to reduce biases resulting in a more conclusive view on this subject.
7.
Literature
Bergmann, B., Christophers, H., Huss, M., & Zimmermann, H. (2009). Listed private equity. Private
Equity: Fund Types, Risk, Return and Regulation, 53-70.
Brown, C., & Kraeussl, R. (2010). Risk and return characteristics of listed private equity, 558. Campbell, R., Huisman, R., & Koedijk, K. (2001). Optimal portfolio selection in a Value-at-Risk framework. Journal of Banking & Finance, 25(9), 1789-1804.
Cochrane, J. H. (2005). The risk and return of venture capital. Journal of financial economics, 75(1), 3-52. Fleming, G., & Cumming, D. (2010). Institutional Investment in Private Equity: Motivations, Strategies, and Performance. Private equity: fund types, risks and returns, and regulation, 7-29.
Doukas, J., Gonenc, H., & Plantinga, A. (2013). Private acquisition gains: A contingent claims explanation.
The European Journal of Finance, (ahead-of-print), 1-24.
Fenn, G.W., Liang, N. and Prowse, S. (2010). The Economics of the Private Equity Market.
Fletcher, J. (2000). On the conditional relationship between beta and return in international stock returns.
International Review of Financial Analysis, 9(3), 235-245.
Hwang, M., Quigley, J. M., & Woodward, S. E. (2005). An index for venture capital, 1987-2003. Contributions in Economic Analysis & Policy, 4(1).
Jegadeesh, N., Kräussl, R., & Pollet, J. (2009). Risk and expected returns of private equity investments: evidence based on market prices (No. w15335). National Bureau of Economic Research.
Jones, C., & Rhodes-Kropf, M. (2003). The price of diversifiable risk in venture capital and private equity. Kaplan, S. N., & Schoar, A. (2005). Private equity performance: Returns, persistence, and capital flows.
The Journal of Finance, 60(4), 1791-1823.
Kaplan, S. N., & Strömberg, P. (2008). Leveraged buyouts and private equity (No. w14207). National Bureau of Economic Research.
Lahr, H., & Herschke, F. T. (2009). Organizational forms and risk of listed private equity. The Journal of
Private Equity, 13(1), 89-99.
Ljungqvist, A., & Richardson, M. (2003). The cash flow, return and risk characteristics of private equity (No. w9454). National Bureau of Economic Research.
Lossen, Ulrich. 2006. The Performance of Private Equity Funds: Does Diversification Matter? Working Paper 2006-14, Munich School of Management.
LPEQ (2013).
http://www.lpeq.com/AboutLPEQ/Memberoverview.aspx LPX group (2013a). Guide to the LPX equity indices. P. 6 LPX group (2013b).
Phalippou, L., & Gottschalg, O. (2009). The performance of private equity funds. Review of Financial
Studies, 22(4), 1747-1776.
S&P (2013). http://us.spindices.com/indices/equity/sp-listed-private-equity-index
Statman, M. (1987). How Many Stocks Make a Diversified Portfolio? Journal of Financial and
Quantitative Analysis, 22(03), 350-360.
Woodward, S. E., & Hall, R. E. (2004). Benchmarking the returns to venture (No. w10202). National Bureau of Economic Research.
Zimmermann, H., Bilo, S., Christophers, H., & Degosciu, M. (2004). The risk and return of publicly traded private equity. University of Basel, Working Paper.
Womack, K., & Zhang, Y. (2003). Understanding risk and return, the CAPM, and the Fama-French three-factor model. Tuck Case, (03-111).
8.
Appendix
Table 4 : LPE vehicles selected by LPX-Group Date as of: 30-09-2013
Name 3I Group plc
Aberdeen Development Capital PLC Altamir Amboise S.A.
Amphion Innovations plc ARC Capital Holdings Arques Industries AG Aurelius AG
Aurora Russia Limited Better Capital Limited BMP AG
Brait S.A. Bure Equity AB
Candover Investments plc
China Growth Opportunities Limited
China Merchants China Direct Investments Limited Cleantech Invest AG
Compass Diversified Holdings DeA Capital S.p.A.
Deutsche Beteiligungs AG
Dinamia Capital Privado, S.C.R., S.A. Dunedin Enterprise Investment Trust PLC Electra Private Equity PLC
Eurazeo S.A.
HBM BioVentures AG
Heliad Equity Partners GmbH & Co. KGaA Helikos SE
HgCapital Trust plc
Imperial Innovations Group plc Ingenious Media Active Capital Limited Internet Capital Group, Inc.
IP Group plc Jafco Co., Ltd.
Japan Asia Investment Co., Ltd. k1 Ventures Limited
Kubera Cross Border Fund Limited Ledstiernan AB
LMS Capital plc Management & Capitali MVC Capital, Inc. New Value AG New Venturetec AG
Northern Investors Company PLC Novestra AB
OFI Private Equity Capital SCA Onex Corporation SV
Origo Partners PLC Promethean PLC Ratos AB
Safeguard Scientifics, Inc. Spark Ventures plc
Symphony International Holdings Limited TVC Holdings
Unternehmens Invest AG Wendel S.A.
Apollo Investment Corporation Ares Capital Corporation
BlackRock Kelso Capital Corporation Equus Total Return, Inc.
Fifth Street Finance Corp. Gladstone Capital Corporation Gladstone Investment Corporation Greenwich Loan Income Fund Ltd GSC Investment Corp.
Henderson Diversified Income Limited Hercules Technology Growth Capital, Inc. Intermediate Capital Group PLC
JZ Capital Partners Limited
Kayne Anderson Energy Development Company Kohlberg Capital Corporation
Main Street Capital Corporation MCG Capital Corporation NGP Capital Resources Company PennantPark Investment Corporation Prospect Capital Corporation
Quorum Oil and Gas Technology Fund Limited TICC Capital Corporation
Triangle Capital Corporation CapMan Plc
KKR & Co. (Guernsey) L.P. Partners Group Holding AG The Blackstone Group L.P. Absolute Private Equity AG Amanda Capital Plc AP Alternative Assets, L.P. APEN AG
Castle Private Equity AG Conversus Capital, L.P. East Capital Explorer AB EIH plc
F&C Private Equity Trust plc B Graphite Enterprise Trust PLC Harbourvest Global Private Equity
Henderson Private Equity Investment Trust plc ING Private Equity Access Limited
J.P. Morgan Private Equity Limited Mithras Investment Trust plc
NAXS Nordic Access Buyout Fund AB NB Private Equity Partners Limited Oakley Capital Investments Ltd.
Pantheon International Participations PLC Princess Private Equity Holding Limited Private Equity Holding AG
Private Equity Investor PLC Scandinavian Private Equity AS shaPE Capital AG
Standard Life European Private Equity Trust PLC SVG Capital plc
Table 5 : S&P’s listed private equity index, full consistuents list Date as of: 30-09-2013
NAME
3I Group
Alaris Royalty Corp. American Capital Ltd
Apollo Global Management LLC Apollo Inv Corp
Ares Capital Corp Asiasons Capital Ltd Aurelius AG
Blackrock Kelso Capital
Brookfield Asset Management Inc Capital Southwest Corp
Carlyle Group LP/The Compass Diversified Trust Electra Investment Trust Eurazeo
Fidus Investment Corp Fifth Street Finance Corp
Firsthand Technology Value Fund Inc Fortress Investment Group LLC Garrison Capital Inc
GIMV NV
Gladstone Capital Corp Gladstone Investment Golub Capital BDC Inc GSV Capital Corp
Hercules Technology Growth Capital Inc ICG Group Inc
Intermediate Capital Group Jafco Co
KCAP Financial, Inc. KKR & Co
Main Street Capital Corp
Marfin Investment Group Holdings SA MCG Capital Corp
Medley Capital Corporation MVC Capital
New Mountain Finance Corp Oakley Capital Investments Ltd Onex Corp Subvtg
Partners Group Hldg
PennantPark Floating Rate Capital Ltd PennantPark Investment Corp Prospect Capital Corp Ratos AB B
Safeguard Scientifics Solar Capital Ltd. Solar Senior Capital Ltd. SVG Capital PLC TCP Capital Corp THL Credit Inc TICC Capital Corp Triangle Capital Corp Wendel
Dated as of 30-09-2013 Name
Aberdeen Private Equity Altamira SCA
Deutsche Beteiligungs AG Dinamia Capital Privado Dunedin Enterprise Electra Private Equity F&C Private Equity Gimv NV
Graphite Enterprise HarbourVest Global PE
HBM Healthcare investments AG HgCapital
JP Morgan Private Equity JZ Capital Partners Ltd NB Private Equity Partners Pantheon Int’l Participations Private Equity Holdings Standard Life European PE
