Marouane LOUNIS
1084 1261Which variables drive performance in
Private Equity transactions:
A deal-level study of Mid-Market buyouts in Europe
Master Thesis – MSc Business Economics
Supervisor: Jens Martin
Abstract
In this study we use a novel dataset of 337 mid-market European private equity transactions exited from 2005 to 2015 in order to analyze the effect of some selected deal-level and macroeconomic variables on their overall performance measured as the Money Multiple.
We confirm several previous findings such as the strong impact on performance of operating improvements but we find the Ebitda multiple to have a negligible impact. We also find that IPOs are the most favorable exit way and that general macroeconomic conditions strongly impact returns, specifically the GDP growth rate. Analysis further shows that private equity is not strongly influenced by public market benchmarks.
Statement of Originality
This document is written by Marouane Lounis who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in
creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
1) Introduction
Private Equity is an alternative asset class consisting of equity in companies that are not publicly traded. It has strongly increased in popularity these last two decades and is usually decomposed in two categories: venture capital and buyouts. This research project will focus exclusively on buyouts, which are defined as a strategy of making equity investments as part of a transaction in which a company is acquired from its current shareholders, usually with the use of financial leverage.
Private Equity firms typically raise a fund every 3 to 5 years that has an average lifetime of 5 to 10 years. During this time the firm, also called General Partner (GP) manages the fund,
investing in companies to ultimately sell them and realize a profit at exit. The Limited Partner (LP) commits money to the PE Firm in order to realize some capital gains. The LP pays fees to the GP that are generally composed of a Management Fee (2% of the amount committed every year) and a Carried Interest (usually 20% of the gains made by the GP). Some other fees may apply (Transaction fees, etc.). The LP does not participate in deals selection and has to carefully select the GP it commits capital to, putting selection skills as the foremost conditions for
success.
This thesis is the result of collaboration with a leading LP that kindly let us access its databases and funds’ financials.
Source: British Private Equity & Venture Capital Association (BVCA)
Despite a healthy literature on buyouts, little is known about the details of private equity transactions, as most studies rely on publicly available data or confidential data from a single buyout firm. Still previous similar studies have been done focusing on deal-level variables and several factors have been found to strongly relate to performance. Research on macroeconomic factors is also quite developed and findings often classify private equity as a pro-cyclical asset class, i.e. strongly influenced by the general state of the economy, especially the GDP growth rate as stated by Kaplan and Schoar (2005) and Phalippou and Zollo (2005).
This thesis analyzes the specific impact of a selected list of variables on deals’ gross performance to the GP measured as the Money Multiple.
Hypotheses are tested on a sample containing 337 buyout transactions in Europe belonging to the Mid-Market segment and exited between 2005 and 2015, capturing the recent 2008 Financial Crisis. OLS regression analysis is the empirical strategy used to test the hypotheses. The results are interpreted among some limitations, notably the relative limited size of the sample, which is a recurrent issue in PE research.
Main findings are that IPOs are the Exit way that brings the most return to investors, while buying a company from a corporate seller relates to highest performance at exit.
We also confirm that Private Equity is strongly subject to general macroeconomic conditions, most notably the GDP growth rate but does not seem to relate to public market benchmarks. The thesis will be organized as follows. In Section II we will start by the theoretical framework in which we will expose the most relevant literature on the subject and formulate hypotheses. Section III will describe each variable, introduce the dataset and detail the empirical strategy. Section IV will show and discuss the results from the regressions. Finally, Section V will conclude the research and summarize the main findings.
2) Theoretical Framework A) Literature Review
According to previous research, many variables have a strong impact on the performance of private equity deals. In this study, we will focus on deal-level variables and Macroeconomic variables.
We will measure the performance of a given deal by its Money Multiple, that is to say the Total Proceeds including any previous distribution divided by the Total Investment cost. We will not use the IRR as it has been attributed a number of pitfalls in the recent literature, notably
because of its time component, it favors firms that minimize their holding periods by buying and selling quick even if they create relatively less value.
The research will be focused on fully exited deals for which the Money Multiple is only computed with money that has effectively been paid back to the fund nullifying the risk for potential errors in fair valuations common for unrealized or partially realized deals.
The literature about Private Equity is now quite developed following the very high growth of this sector in the last decades. The most relevant issue is finding the variable that relate to higher returns. Several papers have been written on the subject, each analyzing a certain set of variables and drawing conclusions about their explanatory power and significance. We will expose the findings that are most relevant to this study.
Deal-level
Achleitner et al. (2011) find that sales growth over the holding period has a positive impact on performance, but that EBITDA margin improvements have no significant influence on LBO pricing at exit while Acharya et al. (2013) associate both revenues and profits growth to performance. Kaserer (2011) confirms the latter view.
According to Guo (2009) nearly 20% of the realized performance comes from the change in the EBITDA multiple, and this is supported by Acharya, Gottschalg, Hahn and Kehoe (2013) who relate multiple expansion strongly to performance. The opposite view is held by Kaserer
(2011) who states that the return contribution stemming from EBITDA multiple enhancements is negligible.
Pindur (2009) points leverage at entry as a cause of value creation, stating that debt has a disciplinary effect on managers therefore the more debt the least ways for managers to “waste” investors’ money. The latter conclusion is quite frequent in corporate finance with the theory of the disciplining power of debt. This is further confirmed by Achleitner et al. (2010) who relate more leverage at entry to more experienced GPs, since they are more trusted by financial institutions. Further more experienced GPs are related to more returns as stated by Kaplan and Schoar (2005). But an excessive amount of leverage can lead a company to bankruptcy inducing a write-off, the worse outcome for investors.
Katz (2009) finds that a majority stake in the private company has usually a positive effect on the performance, as opposed to Loos et al. (2007) who find no significant relation. A majority stake enables investors to efficiently leverage their operational knowledge to bring changes in companies they buy and improve governance through a better incentivization of the
management.
Regarding the way a Private Equity firm exits its investments Loos et al. (2007) find that IPOs bring the most return, closely followed by Trade Sales. This view is shared by Guo, Hotchkiss and Song (2011) putting IPO as the preferred exit route.
But Degeorge, Martin and Phalippou (2014) claim that the best performance is generated in secondary buyout exits when the selling and the buying financial sponsors have a
Macroeconomic
Phalippou and Zollo (2005) argue that PE returns are highly cyclical and covariate positively with public stock markets (SP500, etc.), corporate bond yields, Price/Earnings ratios
(especially for specific sectors, for example tech companies) and the state of the IPO market at the exit of the investment (i.e. more IPOs happening will correlate with a higher average performance of buyout funds). They also find a negative effect from interest rates and a
positive effect from a higher GDP growth on returns. They claim that PE systematic risk is not a driver of performance but rather macroeconomic context is more significant. Kaplan and Schoar (2005) confirm this view stating that PE returns are subject to a boom and bust cycle. Further, Achleitner et al. (2011) claim that higher valuations at exit are realized when public benchmarks are also higher and that buyout multiples are strongly influenced by general economic conditions, notably the GDP growth.
The opposite view is held by Gompers and Lerner (2000) who find that during periods of economic expansion, higher valuations of investment opportunities increase the competition between Private Equity investors and may have a negative effect on performance therefore attributing a negative effect to GDP growth on returns. Kaserer and Diller (2005) state that PE returns are not influenced by public market returns and are negatively correlated to the overall economic development as measured by the GDP growth rate, qualifying Private Equity as an asset class with low or even no market risk.
How much money in the economy is allocated to PE is also a determining factor of average funds performance. Gompers and Lerner (2000) claim that “money chasing deals” impact returns negatively. Kaplan and Schoar (2005) note that funds raised during periods of intense competition for deals tend to underperform funds raised during other periods.
Harris, Jenkinson and Kaplan (2013) suggest that an influx of capital into buyout funds is associated with lower subsequent performance, consistently with Kaplan and Stromberg (2008).
Jegadeesh Kräussl and Pollet (2014) claim that returns are negatively correlated to the credit spread, as a consequence of Private Equity’s high exposure to credit risk. Widening spreads are usually associated with worsening general business conditions according to Fama (1990) thus making potential exits harder and also increase the cost of raising new risky debt for the leveraged buyouts.
B) Hypotheses
In this paper, we are aiming at finding to what extent specific variables have an effect on the performance of a given buyout transaction. We will focus the research on 8 main hypotheses.
H1: Sales growth over the holding period is positively related to performance H2: EBITDA margin improvements have no effect on performance
H3: IPOs are the exit mode bringing the higher level of performance H4: Leverage at Entry has a positive effect on performance
H5: Leverage at Exit has a negative effect on performance
H6: Change in the EBITDA multiple has a negligible impact on performance H7: GDP growth rate is positively related to performance
C) Variables Description Fund-Level:
Firm name: name of the Private Equity firm Fund name: name and sequence of the fund
Vintage:year in which the first influx of investment capital is delivered Fund size: total amount of money committed to the fund
Fund country: country where the Private Equity firm is legally incorporated
Deal-Level:
Seller: it categorizes the 5 types of sellers the company was bought from; it can be corporate (the firm was bought from an operational corporation, spin offs for example), financial sponsor (secondary/ tertiary buyouts), family/entrepreneur (the company was acquired directly from its founders), public (the company was acquired as a publicly-traded company) or government. Sector: it classifies the sector of operation of the private company in 26 categories
Country: it is the country where the private company’s headquarters are located
Invested (Total Cost): it is the total amount of money invested in the specific deal by the PE firm during the company’s holding time in the portfolio
Realized (Realized Amount- DVPI): it is the amount of money effectively received by the PE firm from the company during its holding period (i.e. proceeds)
Unrealized (Unrealized Amount - RVPI): it is the fair valuation of the fund’s remaining share in the company. This thesis focuses exclusively on fully realized deals, therefore the vast majority of them has no unrealized amount, the only exception is when the exit is an IPO and the lock-up period prevents the fund from selling its shares immediately. In this case we value the shares at the March 2015 stock price.
Total (Total Amount - TVPI): it is the sum of the realized and unrealized amounts
Multiple: it is the Total Amount divided by the Total Cost (Total Amount/Total Cost). This is the exclusive variable we use to measure deals’ performance. It is also called Money Multiple or Realization Multiple.
Acquisition date: date when the first investment was made in the company Exit date: date when the company was entirely divested by the fund
Investment Status: 1 for fully exited deals and 2 otherwise. This thesis focuses exclusively on fully exited deals therefore the value will always be 1 in the database.
Duration: it is the amount of time (in years) the company was in the fund’s portfolio; we compute it as the exit date minus the entry date
Exit: this categorizes the 7 different ways companies are exited by the fund. It can be sold to a financial buyer (secondary/ tertiary buyouts), an IPO (the fund takes the company public), a management buyback (management buys the company back from the fund), public (the fund sells shares back on the public market, implying it also bought them on the public market in the first place without taking the company private in between, i.e. PIPE), a trade sale (the company is bought by a strategic buyer), a write-off (the investment is cancelled) or a government exit. Ownership: it is the percentage of equity owned by the fund at entry (first investment) EV Entry/ EV Exit: it is the Enterprise Value of the private company
Debt at Entry/ Debt Exit: it is the amount of debt the company has right after the first investment/ right before being sold. This includes the debt contracted by the fund to finance the LBO.
Leverage at Entry/ Exit: it is the Financial Leverage at the entry/ exit of the private company; it is defined as the Debt divided by the Enterprise Value
Equity at Entry/ Equity Exit: we compute the Equity as (EV- Debt)
Margin at Entry/ Margin Exit: it is the company’s operating margin. We compute it as (EBITDA/ Sales)
Ebitda Multiple: it is the EV divided by the Ebitda (EV/ EBITDA)
Margin Variation: it is the percentage of variation of the operating margin over the holding period.
Enterprise Value Variation: it is the Enterprise Value Variation over the holding period. Sales Growth: it is the percentage growth of the sales over the holding period
EBITDA Multiple Variation: it is the EBITDA Multiple Variation over the holding period. Leverage Variation: it is the Leverage Variation over the holding period.
Macroeconomic:
Market Return: it is the annual growth rate of the private company’s country main index (for example CAC40 in France or DAX in Germany) at the exit year
PE Ratio: it is the average PE ratio of all the companies included in the previously mentioned index during the exit year
3) Methodology A) Dataset
The dataset is a hand-collected sample of 337 deals stemming from private equity funds that have been invested by the LP and that were exited between 2005 and 2015. Private Equity firms in the sample span from Global firms with Pan-European funds (Permira, BC Partners, CVC Capital, etc.) to local firms solely investing in their country of origin (Gresham, Bencis, Procuritas, etc.). Table 4 displays the funds origination.
The data gathering has been the largest part of the research project. For obvious reasons private equity deal-level data is not available on regular financial data outlets and the industry is characterized with very low transparency.
Data has been collected by hand from Quarterly Reports and to a lesser extent from PPMs (Private Placement Memorandums are documents sent by PE firms as part of their fundraising efforts where they give details about deals from prior funds). Note that PPMs are fundraising material and therefore tend to focus on the deals that performed best rather than on the less performing ones. Access to Quarterly Reports enables us to mitigate this survivorship bias. Indeed when a LP invests a certain amount of money in a Private Equity fund, the GP keeps the LP up-to-date about the fund’s performance through Quarterly Financial Statements and Quarterly Reports. The latter usually contain a page with the general news regarding the fund (new investments, exits, etc.) and a “Summary of Investments” table. This table lists each deal (exited or current) with information such as “Total Amount Invested”, “Proceeds” and “Fair Valuation” of the remaining fund’s share in the private company. Deal’s performance was collected from March 2015 Quarterly Reports (except Sovereign Capital that only provide Annual Reports) and Financial data are updated to December 2014.
The macroeconomic data (Market return, PE ratios and GDP growth rates) was obtained from the LP’s Macroeconomic research department.
B) Empirical Strategy
The main objective of this paper is to evaluate the impact that a selected list of variables have on the overall deal performance measured as the Money Multiple.
We will use OLS regressions to estimate these coefficients. For each variable we will set the Multiple as the dependent variable and diverse variables as the independent one.
The general form of the equation is:
Multiple = β0 + β1*Variable1 + β2*Variable2+ … +βN*VariableN +ε
We will further use fixed effects to minimize the omitted variable bias. These are PC Country, PC Industry, Acquisition Year and Exit Year.
4) Results
Now that the data and the methodology have been outlined, we will proceed to present the results of the regression analysis.
A) Financials Variation
Table 5 displays OLS regressions in which the dependent variable is the Multiple and the independent variables are variation percentages of numerous PC Financials over the holding periods. They are the EBITDA Multiple Variation followed by the Sales growth, the Operating Margin variation and the Enterprise Value Variation.
R2 is equal to 4% when analyzing the EBITDA Multiple effect on performance only while it is 12% when adding the Sales Growth impact and 19% when adding the Operating Margin’s effect. These variables consequently have a relatively low explanatory power over
performance. R2 jumps to 56% when adding Enterprise Value variation to the regression giving the latter variable the highest explanatory power over performance by far. This finding confirms the view held by Kaserer (2011) that EBITDA Multiple change has a negligible effect on performance while increasing the intrinsic value of the company through operational improvements is the most important driver of performance.
Sales growth has a strong and significant positive impact on performance but it becomes negligible and slightly negative when controlling for Enterprise Value variation. Similarly Operating Margin variation’s impact on performance becomes almost negligible and lowly significant when controlling for Enterprise Value variation. This is because the latter variable contains the impact of the former. These three variables together account for the Operational Improvements part of the performance while the EBITDA Multiple accounts for the Market Timing component and has a negligible impact on returns.
B) Seller Types
Table 8 displays the regression analysis of the performance on the category of the seller the firm was acquired from.
R2 is around 8% giving the Seller type a relatively weak explanatory power on performance. The analysis shows that despite some previous research’s findings, buying a company from a Financial Sponsor (i.e. SBO) brings the least amount of performance. Indeed the previous owner probably already made a large part of the operational improvements a private equity firm can possibly bring. The cases were we could expect better performance from secondary buyouts, according to Degeorge, Martin and Phalippou (2015) are if the private equity firms have complimentary skillsets or in the particular case were the previous owner had to sell quickly because Limited Partners demanded exits, making the deals quite attractive to potential buyers. Also the higher interest a Financial Sponsor has in obtaining the best price possible when selling its investment, coupled with better negotiating skills from experience can explain why buying from a Financial sponsor might not be the best strategy. Oppositely a founding family is rather preoccupied by cashing out and retiring and a corporate seller wants to get rid of a costly business unit through a spinoff. Similarly a government is not supposed to have any explicit profit motive when privatizing a company.
The analysis further shows that the highest return by far is obtained when the private company was acquired from a corporate seller. In most cases the cause of the sale is a spin-off and as discussed above the standalone company sold is usually less valuable than it was when still a part of the parent company, mostly because of the synergies abandoned. This makes it an attractive deal for a buyer with a good strategy and the right operational skillset. Furthermore even if a strategic buyer has a vested interest in selling the company for the highest price possible, this is not the core of its operations, as opposed to a financial sponsor whose main agenda is to buy low, improve and sell high.
Buying a company from a Family/ Entrepreneur tends to generate a relatively low amount of return at exit, which is mostly because of the transition time to new management. This is a very
familiar problem with companies at the lower end of the Mid-market where the founder has some insider knowledge that is hard to be replaced. A common solution to this issue is to keep the retiring founder as a consultant to the new (MBI) or already existing (MBO) management for a few months or years. Therefore we believe that a company could be acquired at a relatively good price from a Family/ Entrepreneur but maintaining/ increasing the operating income can be more challenging than average.
Public sellers are related to a slightly above average performance although the significance is zero, which is explained by the large heterogeneity of this type of operations making it hard to single out any particular trend. Indeed private equity firms usually invest in public companies opportunistically.
With the government being the seller in one case only we cannot conclude anything significant about this category.
C) Exit Routes
Table 7 displays the fixed-effect regression analysis of the Multiple on the different ways a private company has been exited from the portfolio, namely through a Financial Buyer, an IPO, a Strategic Buyer (Trade Sale), a Government, a public sale or a Write-Off.
R2 is equal to 0.18 giving the Exit Route an explanatory power close to 20% over the deals’ performance.
IPOs have the largest positive relation with performance. A deal exited this way will have an average Money Multiple of 3.1, highly above the sample average at 2.4. IPOs have been
considered as the Holy Grail by investors for a long time. Despite bringing the largest amount of return, they also carry a certain prestige and recognition. The drawback though is the lock-up period usually lasting 180 days that is a risk for investors as the price of the shares can still go down during this time.
The next best way to exit for investors is Management Buybacks, very closely followed by Trade Sales and Financial Buyers (SBOs). The latter category brings a return slightly above the sample average but not as high as some previous research seems to show. As discussed earlier for Seller types, the main motivation of a financial buyer will be to profit by selling the company 3 to 5 years after acquiring it therefore their main goal is to buy low. Strategic buyers tend to pay a higher price since the company will be worth more only by realizing synergies and they usually acquire companies to obtain more market share among other strategic reasons. The best example for this phenomenon is the Tech companies that spend tremendously high amounts to acquire targets that are not even profitable yet (Facebook acquisition of Whatsapp).
Management buybacks are also motivated by long-term visions rather than “flipping” the company for a profit.
Public exits come next and bring a lower performance than the average sample. This type of Exit follows an Entry route called a PIPE (Private Investment in Public Company) that is often an opportunistic investment for PE firms to invest in the stock market when private deals become scarce. PIPEs do not usually allow for ownerships of large amount of stock, thus control.
The lowest coefficient is unsurprisingly associated with Write-Offs closely followed by Government. The only observation where the company was exited through a government buyout was very close to a Write-Off with the state bailing out the company through its acquisition.
C) Debt and Leverage
Table 9 displays OLS regression analysis of the Multiple on several debt variables, namely the Leverage at Entry, the Leverage at Exit, the Leverage Variation over the holding period and the Debt/ Ebitda ratio at Entry. Fixed-effects are used to minimize the omitted variable bias, they are namely the Acquisition year, the Exit year, Private company country and industry.
R2 is equal to 28% when analyzing the effect of Leverage at Entry only and increases to 32% when analyzing the cumulated impact of leverage at entry and exit. It jumps to 42% when we control for the variation of the leverage over the holding period.
The regression analysis shows that leverage at entry and Debt/Ebitda at entry are negatively correlated to performance. Even though debt has a disciplining power over managers it also increases bankruptcy risks leading to Write-Offs. But the impact of these variables are relatively weaker compared to the effect of Leverage at exit.
Leverage at exit has a strongly significant negative impact on performance. It becomes slightly lower when controlling for leverage variation over the holding period. Indeed a high amount of leverage at exit implies that the private company did not manage to pay off enough debt, which is often due to insufficient cash flows over the holding period. The amount of leverage at exit is a negative function of the operating margin improvements discussed earlier because when negotiating LBOS financing with banks, GPs have to demonstrate that the acquired company cash flows will cover the loan.
More debt at exit also mechanically lowers the Enterprise Value. We further find a negative impact of leverage variation on the performance, confirming our views.
Therefore it seems that what matters most is having a relatively low leverage at exit rather than at entry and the disciplining power of debt at entry on management seems to be outweighed by the extra amount of risk it brings.
D) Duration
Table 1 shows an average deal duration of 4.6 years with a minimum of 0.83 years and a maximum of 11 years. It falls right into the commonly accepted 3 to 5 years holding period. Table 9 displays the regression of the Multiple on the Duration of the holding period with country and industry fixed-effects. R2 is equal to 16%.
The table shows a significant negative relation of duration with performance even if Multiple does not have a time dimension (as opposed to IRR). This is strongly influenced by the several write-offs that have a longer average duration than other Exit routes since GPs tend to keep them longer in their portfolio in a hope of improvements (the human tendency to delay losses is a major insight from Behavioral Economics).
F) Macroeconomic Variables
Table 6 displays regression analysis where the Multiple is the dependent variable on some macroeconomic variables, namely the Market Return, the PE Ratio and the GDP Growth Rate at exit year.
R2 oscillates between 1% and 2% when analyzing public benchmarks (PE ratio and market returns) and it jumps to 5% when we control for GDP growth. These variables thus have only a slight explanatory power on the performance.
The regressions show that the GDP growth is strongly and significantly positively related to overall deal’s performance. Surprisingly the PE Ratio and the Market Return (public
benchmarks) are negatively related to returns.
Because of its intrinsic high illiquidity and lower volume, private equity is relatively
disconnected from public benchmarks. Indeed as opposed to public shares, if a GP wanted to liquidate its portfolio in a period of crisis it might not be possible to find buyers unless the prices offered were especially low. Therefore they will tend to wait longer before selling
portfolio companies. The cases where PE firms still sell at a high loss during a downturn are when there are under LP pressure. As we will see later this is a particularly bad move for the selling GPs but can create tremendous opportunities for some buyers. Also the private equity market is composed of what has lately been called “Smart Money” meaning that exclusively sophisticated investors have access to this asset class while in public equity it is easier to find a “less sophisticated” counterparty to trade with even in bad economic times.
GDP Growth’s coefficient almost doubles in value when using the fixed effects (Country, Sector, Acquisition and Exit years) which confirms its preponderant impact on performance. Graph 1 paints a clear picture of the strong correlation between GDP growth rate and PE returns. Therefore we accept H7 and reject H8.
G) Acquisition and Exit years
Table 10 displays the fixed-effect regression of the Multiple on the private company acquisition year while Table 11 displays the same analysis on the exit year.
R2 is about 10% for acquisition year and around 9% for exit year brining the explanatory power of a right timing of the investments to close to 20% of the performance.
Private companies acquired in 2007 and 2008 bring the lowest performance at Exit because they have been acquired at overvalued prices, similarly companies acquired in 2009 are associated to the highest returns of the sample. The picture is inverted for exit years where 2007 and 2008 are correlated with the highest performance while 2009 was the worst year to sell for GPs on average.
These finding confirm the boom-and-bust nature of Private Equity returns claimed by Kaplan and Schoar (2005). A GP would have optimized its aggregate returns by selling its portfolio companies in 2007 and 2008 (right before the financial crisis at a time of “irrational
exuberance”) and by acquiring new private companies in 2009 (right after the crisis when the markets where rather “bearish”). Therefore Private Equity is associated with a relatively high market risk.
5) Conclusion and Discussion
In this paper, we investigate the performance of private equity using a novel hand-collected dataset of 337 buyout deals exited between 2005 and 2015. We focus on transactions involving European companies in the Mid-Market segment.
Private Equity is a relatively recent alternative asset class that has seen its market share strongly grow during the last two decades among asset managers.
From a Limited Partner standpoint the key factor of success is selection skills regarding the PE firms to invest in. Therefore to be able to predict and reproduce performance we need to single out specific factors that drive returns from those that do not. This study aims at reaching this goal through OLS regressions of the transactions’ Multiple on a selected list of deal-specific and macroeconomic variables.
Main findings are a strong positive correlation between deal returns and general
macroeconomic conditions defined as GDP growth, qualifying Private Equity, just like its public counterpart, as an asset class that is subject to a boom and bust cycle. But we find the asset class to be disconnected from public markets’ benchmarks mainly because of private equity’s lower trading volume and highly illiquid nature.
Furthermore it seems that Leverage at Entry or at Exit has a negative impact on performance and that the role played by changes in the EBITDA Multiple is negligible compared to concrete operational improvements reflected in the Enterprise Value variation over the holding period. Among exit routes IPOs bring the largest amount of returns to GPs, followed by management buybacks. Regarding sellers, buying a firm from a financial sponsor is related to the lowest amount of return while the best performance comes from acquiring a firm from a corporate seller followed by buying out a public firm.
Some limitations that were encountered in this study could be turned into further research topics. Firstly the dataset has been collected among funds invested by the LP which could be the source of some omitted variable bias and a selection bias.
Secondly when comparing private equity performance to public benchmarks, we used indexes and PE ratio of large-cap stocks since they have the greatest availability.
Therefore some further research could aggregate several databases to mitigate the selection bias and use Mid-Cap benchmarks when comparing private equity to public markets.
Bibliography
Braun, R., Jenkinson, T., & Stoff, I. (2013). How Persistent is Private Equity Performance? Evidence from Deal-Level Data. Working Paper, Said Business School, University of Oxford
Gompers, P., & Lerner, J. (2000). Money chasing deals? The impact of fund inflows on private equity valuation. Journal of Financial Economics
Harris, R. S., Jenkinson, T., & Kaplan, S. N. (2012). Private Equity Performance: What Do We Know? Working Paper
Harris, R., Jenkinson T., Kaplan S., Stucke R. (2013). Has Persistence Persisted in Private Equity? Evidence From Buyout and Venture Capital Funds. Working Paper. University of Virginia (UVA) -
Darden School of Business, University of Oxford - Said Business School, University of Chicago - Booth School of Business and University of Oxford - Said Business School
Higson C. and Stucke R. (2012). The Performance of Private Equity. Working Paper, London
Business School
Jegadeesh, N., Kräussl, R., & Pollet, J. (2009). Risk and Expected Returns of Private Equity Investments : Evidence Based on Market Prices. NBER Working Paper
Kaplan, S. N., & Schoar, A. (2005). Private equity performance: Returns, persistence, and capital flows. Journal of Finance
Kaserer, C., & Diller, C. (2004). European Private Equity Funds – A Cash Flow Based Performance
Analysis ∗. CEFS Working paper.
Korteweg and Sorensen. (2014). Skill and Luck in Private Equity Performance. Rock Center for Corporate Governance at Stanford University. Working Paper No. 179
Lopez-de-Silanes, F., Phalippou, L., & Gottschalg, O. (2013). Giants at the Gate: On the Cross-section of Private Equity Investment Returns. Journal of Financial & Quantitative Analysis, forthcoming
Paper, W., Kaplan, S., & Schoar, A. (2003). Returns, Persistence and Capital Flows. Social Science
Research, MIT Sloan School of Management
Phalippou, L. (2014). Performance of buyout funds revisited? Review of Finance, 18(1), 189–218.
Phalippou, L. & Gottschalg, O. (2007). The Performance of Private Equity Funds. Review of
Financial Studies
Phalippou, L., & Zollo, M. (2006). What Drives Private Equity Fund Performance? Working Papers
Robinson, D. T., & Sensoy, B. A. (2011). Cyclicality, Performance Measurement, and Cash Flow Liquidity in Private Equity. Working Paper
Steven N. Kaplan and Per Stromberg (2008). Leveraged Buyouts and Private Equity, Working paper 14207, National Bureau of Economic Research
Braun, Engel, Achleitner (2011). Value creation and Pricing in Buyouts, Review of Financial
Economics
Acharya, V. V., Gottschalg, O. F., Hahn, M., & Kehoe, C. (2013). Corporate governance and value creation: Evidence from private equity. Review of Financial Studies, 26(2), 368-402.
Pindur, D. (2009). Value creation in succesful LBOs. Springer Science & Business Media.
Loos, N., Hilb, M., Bernet, B., Gottschlag, O., & Zollo, M. (2007). Value creation in leveraged
buyouts: analysis of factors driving private equity investment performance. Springer Science &
Business Media
Katz, S. P. (2009). Earnings quality and ownership structure: The role of private equity sponsors.
The Accounting Review, 84(3), 623-658.
Degeorge F., Martin J., Phalippou L., (2015). On Secondary Buyouts, Journal of Financial
Economics, forthcoming
Chapman, J. L., & Klein, P. G. (2010). Value creation in Middle-Market Buyouts: A Transaction-level analysis. In D. J. Cumming, Private Equity: Fund Types, Risks and Returns, and Regulation. Wiley, Hoboken, NJ.
Kaserer, C. (2011). Return attribution in mid-market buy-out transactions- New evidence from Europe. Available at SSRN 1946110.
Guo, S., Hotchkiss, E., & Song, W. (2011). Do buyouts (still) create value? The Journal of Finance, 66(2), 479-517.
Achleitner, A. K., Braun, R., Engel, N., Figge, C., & Tappeiner, F. (2010). Value creation drivers in private equity buyouts: Emperical evidence from Europe. The Journal of Private Equity, 13(2), 17-27.
Phalippou, L. (2009) The Hazards of Using IRR to Measure Performance: The Case of Private Equity
Table 1 - Sample Description
Table 1 displays the Minimum, Maximum and Average values of several variables in the sample.
Minimum
Maximum
Average
Entry year 2001 2012 2006.1 Exit year 2004 2015 2010.7 Money Multiple 0 13.7 2.4 Duration 0.83 11 4.6 Ownership at entry 0% 100% 55%
EBITDA Multiple at entry 0 21.7 8.14
Table 2 – Private Companies Country
Table 2 displays the origination of the portfolio companies: the number of companies in each country and the percentage each country represents in the whole sample. More than half of them are whether from the UK or France.
PC Country # % Austria 1 0.3 Belgium 8 2.37 Czech Republic 3 0.89 Denmark 11 3.26 Finland 4 1.19 France 60 17.8 Germany 37 10.98 Hungary 1 0.3 Italy 18 5.34 Luxembourg 1 0.3 Netherlands 20 5.93 Norway 5 1.48 Spain 8 2.37 Sweden 36 10.68 Switzerland 4 1.19 Turkey 3 0.89 UK 117 34.72 Total 337 100
Table 3 – Sector of Operation
Table 3 decomposes the portfolio companies by their sectors of operation indicating the number of companies belonging to each sector and the percentage represented by companies in a given sector across the whole sample.
Sector # % IT Services 7 2.1 Aero 2 0.59 Apparel 6 1.8
Automobiles and Trucks 9 2.7
Beer and Liquor 1 0.3
Business Services 53 15.7 Chemicals 9 2.66 Construction 9 2.7 Consumer 33 9.8 Education 2 0.6 Entertainment 4 1.18 Fabricated products 48 14.2 Financial services 21 6.2 Food products 21 6.2 Healthcare 30 8.9 Machinery 3 0.9 Manufacturing 3 0.9 Media 7 2.07 Personal Services 3 0.9 Real estate 1 0.3 Recreation 5 1.5 Retail 18 5.33 Shipping 3 0.89 Steel 3 0.89 Telecom 22 6.51 Transport 5 1.48 Utilities 9 2.7 Total 337 100
Table 4 – Private Equity Firm repartition
Table 4 displays the repartition of deals in private equity firms indicating the number of deals made by any specific PE firm and the percentage they represent across the whole sample.
Private Equity Firm # %
21 CP 8 2.37
21 Investimenti 5 1.48
3i 62 18.4
Accent Equity Partners 13 3.86
Advent International 2 0.59
BC Partners 10 2.97
Bencis 8 2.37
CVC Capial Partners 13 3.86
Charterhouse Capital Partners 19 5.64
Cinven 20 5.93 EQT 12 3.56 Equistone Partners Europe 82 24.33 Exponent 8 2.37 Gresham 10 2.97 Investindustrial 2 0.59 Lyceum Capital 3 0.89 Mid Europa 3 0.89 Nordic Capital 13 3.86 PAI Partners 7 2.07 Permira 17 5.05 Polaris 2 0.59 Procuritas 3 0.89 Sovereign 12 3.56 Terra Firma 1 0.3
Turkish Private Equity 2 0.59
Total 337 100
Table 5 – Financials Variation
Table 5 displays OLS regressions in which the dependent variable is the Multiple and the
independent variables are variation percentages of numerous financials over the portfolio companies’ holding periods. They are namely the EBITDA Multiple Variation, the Sales growth, the Operating Margin variation and the Enterprise Value Variation.
Multiple
EBITDA Multiple Variation 0.037 0.032 0.030 0.001 (3.69)** (3.23)** (3.11)** (0.14) Sales Growth 0.808 0.924 -0.062 -0.062 (5.30)** (5.92)** (0.47) (0.47) Margin Variation 0.632 0.126 0.125 0.132 (4.91)** (1.25) (1.24) (1.33) EV Variation 1.006 1.008 0.994 0.924 (16.02)** (16.46)** (18.52)** (18.20)** Constant 2.471 2.172 2.007 1.662 1.660 1.645 1.694 (21.82)** (17.38)** (15.65)** (16.51)** (16.60)** (17.41)** (17.78)**
R2
0.04 0.12 0.19 0.56 0.56 0.56 0.52N
327 319 310 295 295 295 309 * p<0.05; ** p<0.01Table 6 – Macroeconomic Variables
Table 6 displays regression analysis where the Multiple is the dependent variable on some macroeconomic variables, namely the Market Return, the PE Ratio and the GDP Growth Rate. Fixed-effects are used to minimize the omitted variable bias, they are namely the Acquisition year, the Exit year, Private company country and industry T-statistics are reported between brackets under each coefficient. Multiple Market Return -1.29 -1.00 -0.76 -0.87 -0.89 (2.10)* (1.44) (1.09) (1.41) (-0.59) PE Ratio -0.02 -0.01 -0.02 -0.04 (0.86) (0.36) (0.96) (-1.27) GDP Growth Rate 19.75 20.06 20.42 21.09 33.40 (3.41)** (3.5)** (3.54)** (3.79)** (3.17)** Constant 2.53 2.81 2.28 2.16 2.37 2.05 1.09 (19.9)** (8.03)** (6.03)** (13.14)** (6.45)** (14.11)** (-0.33) Fixed effects:
Acquisition Year No No No No No No Yes
Exit Year No No No No No No Yes
PC Country No No No No No No Yes PC Industry No No No No No No Yes R2 0.01 0.02 0.05 0.05 0.05 0.04 0.31 N 336 336 336 336 336 337 336 * p<0.05; ** p<0.01
Table 7 – Exit routes
Table 7 displays the fixed-effect regression analysis of the Multiple on the different ways a private company has been exited from the portfolio. T-statistics are reported between brackets under each coefficient. The table further displays the number of deals exited through any specific exit route and their percentage among the whole sample.
Multiple # % Financial Buyer 0.23 117 34.6 (0.12) IPO 0.7 37 10.9 (0.36) Trade Sale 0.36 110 32.5 (0.19) Government -2.13 1 0.3 (-0.79) Public -0.97 2 0.6 (-0.41) Management Buyback 0.38 27 8 (0.19) Write-Off -2.22 43 12.7 (-1.15) Constant 2.4 (1.25) R2 0.18 N 337 Total 337 100 * p<0.05; ** p<0.01
Table 8 – Seller Types
Table 8 displays a fixed-effect regression analysis of the Multiple on the different types of sellers a private company has been bought from. T-statistics are reported between brackets under each coefficient. The table further displays the number of deals acquired through any specific seller type and their percentage among the whole sample.
Multiple # % Corporate 1.08 75 28.2 (0.5) Family/ Entrepreneur -0.23 72 27.07 (-0.11) Financial Sponsor -0.43 93 34.96 (-0.2) Public 0.13 24 9.02 (0) Government -0.005 1 0.38 (0.06) Constant 2.4 (1.11) R2 0.08 N 265 Total 265 100 * p<0.05; ** p<0.01
Table 9 – Debt
Table 9 displays OLS regression analysis of the Multiple on several debt variables, namely the Leverage at Entry, the Leverage at Exit, the Leverage Variation over the holding period and the Debt/ Ebitda ratio at Entry. Fixed-effects are used to minimize the omitted variable bias, they are namely the Acquisition year, the Exit year, Private company country and industry.
T-statistics are reported between brackets under each coefficient.
Multiple Leverage at Entry -1.08 -0.64 0.19 (-2.25)* (1.46) (0.31) Leverage at Exit -0.92 -0.78 (-5.37)** (-0.78)** Leverage Variation -0.06 (-2.32)* Debt/ Ebitda at Entry
-0.15 (-1.83) Duration -0.13 (-2.33)* Constant 1.36 0.7 0.6 1.88 (0.53) (0.32) (0.25) (0.83) Fixed effects:
Acquisition year Yes Yes Yes No
Exit year Yes Yes Yes No
PC country Yes Yes Yes Yes
PC industry Yes No Yes Yes
R2 0.28 0.32 0.42 0.16
N 329 288 279 337
Table 10 – Acquisition year
Table 10 displays the fixed-effect regression of the Multiple on the private companies acquisition year. T-statistics are reported between brackets under each coefficient.
Multiple 2002 0.92 (0.44) 2003 0.46 (0.22) 2004 1.09 (0.54) 2005 0.69 (0.34) 2006 -0.02 (-0.01) 2007 -0.73 (-0.36) 2008 -0.33 (-0.16) 2009 1.02 (0.49) 2010 0.44 (0.21) 2011 -0.21 (-0.1) 2012 -0.28 (-0.11) Constant 2.17 (1.07) R2 0.1 N 337 * p<0.05; ** p<0.01
Table 11 – Exit year
Table 11 displays the fixed-effect regression of the multiple on the private companies exit years. T-statistics are reported between brackets under each coefficient.
Multiple 2005 0.16 (0.17) 2006 0.41 (0.47) 2007 1.11 (1.29) 2008 0.77 (0.8) 2009 -1.7 (1.77) 2010 -0.52 (-0.61) 2011 -0.22 (-0.27) 2012 -0.14 (-0.17) 2013 -0.73 (-0.88) 2014 -0.6 (-0.75) 2015 -0.44 (-0.47) Constant 2.6 (3.43)** R2 0.086 N 337 * p<0.05; ** p<0.01
Table 12 – Literature Review
Table 12 displays the dataset used and most relevant findings of selected papers from the literature review.
Paper Sample Findings
Acharya et al. (2013)
395 buyout deals completed between 1991 and 2007 (Western Europe)
-EBITDA multiple strongly and positively related to performance -Sales growth and EBITDA margin improvements have a positive impact on performance
Kaserer (2011)
332 mid-market buyout deals completed between 1990 and 2011 (Europe)
-EBITDA multiple has a negligible impact on performance
-Sales growth and EBITDA margin improvements have a positive impact on performance
Pindur (2009)
42 buyout deals completed between 1993 and 2004 (Europe)
-Leverage at Entry is positively related to performance
Achleitner et al. (2010) 206 buyout deals (Europe)
-Leverage at Entry is positively related to performance
-Sales growth has a positive impact on performance but not EBITDA margin improvement
- Performance is positively impacted by GDP growth rate and public benchmarks
Loos et al. (2007)
3000 buyout deals completed between 1973 and 2003 (US and Europe)
-IPOs is the exit route bringing the most performance
Hotchkiss, Guo and Song (2011)
192 buyout deals completed between 1990 and 2006 (US)
-EBITDA multiple strongly and positively related to performance -IPOs is the exit route bringing the most performance
Gompers and Lerner (2000)
Sample of 7235 venture capital rounds between 1987 and 1995 (US)
- GDP growth has a negative impact on performance
Kaserer and Diller (2005)
200 private equity funds closed between 1980 and 2003 (Europe)
- GDP growth and public benchmarks have a negative impact on performance: private equity has no market risk
Graph 1 – Average Multiple and Average GDP Growth at Exit
Graph 1 displays the average Multiple per exit year and the average annual GDP growth rate for the countries represented in the sample from 2005 to 2015.
Graph 2 – Average Multiple and Average Market Return at Exit
Graph 2 displays the average Multiple per exit year and the average annual Market Return rate for the countries represented in the sample from 2005 to 2015.
Graph 3 – Average Multiple and Average GDP growth at Entry
Graph 3 displays the average Multiple per entry year and the average annual GDP growth rate for the countries represented in the sample from 2004 to 2012.
Graph 4 – Average Multiple and Average Market Return at Entry Graph 4 displays the average Multiple per entry year and the average annual Market Return rate for the countries represented in the sample from 2004 to 2012.
