What does the empirical evidence suggest about the drivers of
dividend recaps and a PE’s subsequent decision on the exit route?
Juan Ruik Beyhaut Master in International Finance
Amsterdam Business School University of Amsterdam
juanruik@student.uva.nl
Thesis supervisor Jens Martin
Abstract
Using a self-collected sample of 436 portfolio companies of Private Equity sponsors that have used dividend recapitalizations as a partial exit, this paper concludes that IPO activity and the LIBOR are key drivers for the level of debt a company is willing to take with the purpose of paying a special dividend to their private equity investor. After observing the exit route of 133 portfolio companies, we conclude that the lower the ratio of recap to initial investment in the company or the longer the holding period of the investment at the moment of the recap, the more likely to be exited through an IPO relative to a Secondary Buyout. Last, we conclude that after doing a recap, a company is less likely to exit through a Trade Sale relative to a Secondary Buyout
Table of Contents
Introduction ... 4
Literature review ... 6
Data sources and sample sources ... 8
Sample ... 8
Summary statistics ... 10
The model and empirical results ... 11
Regression analysis ... 11
Multinomial logistic analysis ... 13
Conclusion ... 15
References ... 16
Introduction
Dividend recapitalizations, leverage recapitalizations, or simply “recaps,” have recently been in the spotlight because banks and bond investors do not like them since General Partners can secure profits before the actual sale through a traditional exit, leaving the company in a vulnerable condition (Financial Times, 14th April 2013). However, Harford and Kolasinski, (2012) suggest that special dividends are not correlated with future company distress.
The market for dividend recaps have seen strong growth and represents approximately 25 per cent of the leverage loan and high yield bond market in the United States (Financial Times, 23rd October 2012). Figure 1, shows the dividend recaps number, and figure 2, displays the volume, both relative to other traditional exit types. It is possible, based on figures 1 and 2, to observe the increased relevance of recaps relative to the overall exit routes. Indeed, recaps represent 14.9 per cent of all the number of deals and 33.2 per cent of the total proceeds of the exits in the period between January 2013 and June 2013.
To explain the rationale of recaps as the topic of this paper, it is necessary to define some relevant concepts. First, a Private Equity (PE) firm is in an asset management company that through a PE fund directly invests in private companies. For them to invest, it is necessary to go through different phases: raise capital from investors to a PE fund, invest in private companies, improve the value of the portfolio company, and exit from the investment. Each of these steps is crucial to the bottom line of a PE fund. A good performance is essential for the carried interest fee gains and the reputational issues that impact the ability to raise funds in the future (Lerner & Hardymon, 2002). While each of the four steps listed above are equally important, this research project will focus on the exit from an investment.
There are different exit routes for PE fund investments which are: Initial Public Offering (IPO), Trade Sale, Secondary buy-out, Buy-back, Leverage Recapitalization (“recaps”), and Liquidation (“write-offs”). Recently, as it was mentioned before, the use of leverage dividend recapitalizations in PE investments has achieved an important role in PE exit strategies. Despite the recent attention, there still is not academic literature strictly focusing on recaps.
A recap is a process where the General Partner (GP) and the Limited Partners (LP) obtain proceeds from the portfolio companies through the re-leverage of the firm and the payment of
a special dividend to the PE fund. Based on B. Espen Eckbo and Karin S. Thorburn (2008) definition, “a leveraged recapitalization is a significant payment to shareholders financed by new debt borrowed against the firm’s future cash flow”. Although this partial exit route has been an important source for the recent success of PE firms, criticism about the economic effects follows. For instance, this kind of exit in the portfolio company could generate a higher likelihood of bankruptcy due to the pressure on a firm with increased leverage. However, it is expected that these concerns should be outweighed by the reputational loss of the PE firm and the fact that leverage multiples tend to be at reasonable levels, indicating a conservative use of debt. All in all, as Axelson, Jenkinson, Strömberg & Weisbach (2010) suggest, these effects are not clear, and therefore, are subject to future research.
As a result, this research is relevant firstly, because the PE industry has become an important player in the global economy. Based on the prediction of Jensen (1989), the Private Equity industry will surge and eclipse public firms in the corporate world due to its superior organizational structure. Secondly, this partial exit strategy has gained importance and it, therefore, becomes an attractive topic. Thirdly, recaps are in the spotlight because of the expected vulnerability of the portfolio companies after a re-leverage. Lastly, there is a lack of academic research regarding this type of partial exit.
This paper aims to find empirical evidence that low interest rates and the lack of activity of traditional exit strategies are the drivers of dividend recaps in the private equity industry. Specifically, the question that this thesis attempt to answer is: What does the empirical evidence suggest about the drivers of dividend recaps and a PE’s subsequent decision on the exit route? As a result, a detailed analysis can be found within this paper which lists companies that were involved in at least one recap between January 2000 and June 2013 and their subsequent traditional exit route.
Literature review
As has been mentioned in the introduction, there is a lack of academic research of dividend recapitalizations as an exit route. However, articles in the press suggest some factors as possible the drivers of the recaps as an exit. For instance, Financial Times (FT) stated the 20th of December of 2012 that the “availability of cheap debt”, “dividend tax changes”, “market was incredibly liquid”, “delivered balance sheets” and the depressed IPO market.
Miguel Sousa (2010) in his paper states that the easy access and the low cost of debt have made this practice very popular. Kushner (2004) states in his paper that recaps are seen as an alternative to a public offering in very specific market conditions: “public offerings are rare, strategic buyers are scarce and acquisition financing is available only episodically”.
The lack of traditional exit routes means that there are fewer options for the PE sponsors to sell their investments and to return the money to the investors. This lack of exit opportunities meant less carried interest earned by private equity firms and, more importantly, that these PE firms faced a real challenge in returning money to their investors within the initial period of the fund of 10 years with the expected return. By completing successful and fast exits, the private equity firms are aware that they will be rewarded in the form of carried interest and, perhaps, more importantly, in form of reputation gain, which is vital toward raising the next fund. Therefore, a good exit is important in order to ensure attractive returns for investors and so their own profit, but based on Lerner and Hardymon (2002), they are also important to facilitate the raising of additional funds.
“You can’t exit, you can’t do a public offering, it’s a way to take the pressure off” (Financial Times, 23rd October 2012). Ideally, a PE firm should keep their investment period short due to IRR and reputation considerations. Based on Schwienbacher (2009), it is crucial to try to avoid long holding periods and having to request an extension on the life of the fund because that is not in the limited partner interest.
Based on the quotation of a comment from the founder of the private equity firm Better Capital, he states that the “failures of companies post-dividend recaps will definitely bring the private equity industry back into the political focus” (Financial Times, 14th April 2013). However, Harford & Kolasinski (2012) and Hotchkiss, Smith & Strömberg (2012) conclude
in their papers that there is no evidence that dividend recapitalizations increase the likelihood of a future portfolio company financial distress. Despite the focus of this thesis is not to look for evidence of financial distress after a recap, in table 3 we can observe that only four out of 436 companies were written off were done during the period of study. Moreover, Cumming and MacIntosh (2003) suggest that partial exits are used as a signal of a portfolio company quality.
Data sources and sample sources
Sample
A multi-step approach was used to collect data from Capital IQ Leverage Commentary and Data (LCD), Thomson ONE, Data Stream, Zephyr, Bloomberg and General Partner’s, which was then analysed for the purposes of this thesis.
It is extremely important to mention the constraints faced when conducting a study on the Private Equity industry due to the limited availability of data, or in some cases, contradicting data from various data sources versus what was actually reported. Indeed, some General Partner’s websites stated the realization of an investment; however, the data sources did not support any activity linked with that exit. Nevertheless, for the purpose of this study, we opted to use the data sources information.
In short, we can say that the entire process of data collection can be summarized in the following steps. First, we obtained from Capital IQ LCD the data of all the dividend recapitalizations processed. Second, we matched this data with data gathered from Thomson ONE regarding all investments and exits of portfolio companies. Third, we utilized Bloomberg to obtain the different names attributed to a single portfolio companies. Lastly, when the information was not available in Thomson ONE, the Zephyr database was used to provide the missing information. Nonetheless, we will discuss in more detail the steps taken to obtain our final data set.
Initially, it was necessary to obtain all the Dividend Recaps done in the US debt market from firms based in the United States, Canada or Puerto Rico, regardless of the origin of their Private Equity sponsor. All other deals of portfolio companies from outside that region, that did their recaps in the US market, were excluded. From data gather from Capital IQ LCD, we obtained the information concerning all the announcement of the deals that took place during between January 2000 and June 2013, including the name of the company, the sponsor, the amount of the recap, the industry where the company operates, and more. It is important to clarify that the announcement date of the recap was used due to the lack of accessibility to data that displays competition dates.
From this data set, we excluded all the deals of public companies and companies that were not sponsored by a PE firm at the moment of the recap. The reason for this is to maintain our focus on the Private Equity industry and since our main data source does not follow the
ownership of PE sponsors in public companies. Moreover, it assumes that all private equity deals exited through an IPO are no longer part of the portfolio of the sponsors; however, it does not consider those cases where general partners exit from their investment at a later date than the IPO date. This is normally the case due to lock-up period agreements. In fact, public companies have easier access to the debt capital market, an issue that may be a source of distortion in our analysis.
In many cases, various private equity firms co-invest or invest at different times in a company. To identify the main PE investor, Capital IQ was used to obtain the name of the sponsor responsible for the recap announcement. It is assumed that the name provided in that data source is the investor-leader.
Secondly, from Thomson ONE we obtained a list of all exits and investments still under the management of a Private Equity sponsors over the same time period stated above. From this data source, we collected data regarding the investment date, initial invested amount, investment status, date completed or exited, holding period, and more. Next, each of the companies that were included in the list provided by Capital IQ was matched with this historical list of private equity investments. Many of the deals were easily matched since the data sources identify a company with the same name while many other were not matched. Thirdly, given that many companies were missing after the second step, Bloomberg was used to find different names, aliases, former names, names used to operate, names of the parent company, or names of a subsidiary that were attributed to the company. This allowed us to increase the number of matches done in the second step.
Fourthly, since many of the companies that did the recaps were not matched with an exited investment or current portfolio company, we searched for the information in Zephyr to create a new deal that was not registered in Thomson ONE. Moreover, in the cases where information over the investment amount was missing from Thomson ONE, Zephyr was also used to provide this information. However, this data source was only relevant for investments done after the year 2000, and rarely before.
Lastly, each of the companies was verified on the Private Equity sponsor’s website.
After the collection process, certain companies were excluded from the sample. The reasons for this are either:
the sponsor is not a Private Equity firm;
the firm did the recap when it was publicly traded;
the PE firm did many rounds of funding in the company;
the recap was announced before the acquisition of the company was completed; or
the portfolio company is from outside the region of study.
Finally, data about the “North American” IPO market was obtained from Thomson ONE; and data about the M&A market, LIBOR (6 months), US treasury bills rate (3 months), and Barclays US Corporate Investment Grade Yield were obtained from Data Stream. Also, LIBOR was used since the debt of the dividend recaps were structured based on this rate. Furthermore, US Treasury bill rates and Barclays US Corporate Investment Grade Yield were used to calculate a proxy of the risk aversion, or risk appetite, that is observed in the market. Summary statistics
From the data collection phase, it was possible to identify information about 436 companies, where 263 of them were exited investments and 173 cases where the companies were still in the portfolio of the General Partners at the time of this study. Unfortunately, the availability of information regarding the initial investment of PE sponsors was limited. Therefore, from those companies identified, 236 had the initial investment data available, 137 of whom were exited and 99 that were still in the portfolio of the General Partner.
Table 2 displays the descriptive statistics about the recap amount, the holding period before the recap, the holding period of the investment (only for those exited investments), and the ratio of recap to initial investment. All this information is organized by the round of recap done in the firm.
Table 3 shows the same variables as table 2, but it organizes the information based on type of exit used. Note, that it also includes the descriptive statistics of the current portfolio companies that have done a recap.
Finally, table 1 displays the quantity of recaps done by different Private Equity firms between January 2000 and June 2013. It includes those General Partners that did more than 7 recaps. It can be observed, that the great majority have executed more than one recap in their portfolio companies. Furthermore, the status of those investments can be observed.
The model and empirical results
Two multivariate analyses will be applied to answer our thesis question. First, a regression analysis was used to verify the drivers of the dividend recaps amounts. Second, a multinomial logistic model was used with the purpose of finding explanatory variables of the exit route of those companies that did a dividend recap.
Regression analysis
The objective of the first analysis is to observe the correlation between the recaps amount and the independent variables stated below. Therefore, the amount of the recaps is the dependent variable of the regression analysis. For this, a regression analysis will be applied, using the following variables in the model:
In this model, 3 macroeconomic independent variables are considered: IPO activity, M&A activity and LIBOR. For the purposes of this regression, the IPO and M&A activity are considered on a rolling 12 months basis. The values used are of the same month of the announcement of the recap as well as 3 and 6 months prior. The units used are in thousands of US dollars. LIBOR 6 months is the last macroeconomic explanatory variable.
Specific information about the portfolio companies is included in the model: holding period of the investment until recap, initial investment in the company, and industry. The first is used as an explanatory variable; however, the last two are considered control variables. The reason an initial investment in the company is used as a control variable, is that the higher the initial investment, the higher recap is expected. As a result, this variable is used as a proxy for the size of the company. Additionally, the industry of the companies is the last firm specific control variable incorporated in the model, which is controlled by the use of dummies variables.
Risk aversion in the market is the last control variable considered. It is calculated as the difference between Barclays US Corporate Investment Grade Yield and US Treasury bill
rates. The rationale for using this variable is that the higher the risk aversion, the higher the level of the recap.
Table 4 displays the results of 6 models of the regression analysis by using different combinations of the variables.
Model 1 includes all the variables stated in the model equation. This model illustrates what we can expect about the variables that may be drivers for the value of a recap. Although, the IPO market activity at the moment of the recaps, and 3 months before, are not significant variables, they have a negative influence on the value of a recap. This is expected based on the common belief that the more depressed the IPO market, the more likely the private equity sponsor will partially exit from its investment with a higher amount of dividend recap. Nevertheless, contrary to what can be anticipated, IPO activity 6 months before the recap has a significant positive coefficient. Moreover, LIBOR has a significant negative influence on the value of the recap. This is in line with our belief that the cheaper the debt, the more will it be used. Finally, the control variables coefficients for size and aversion in the market show the expected sign. The bigger the company is or the lower aversion in the market, the larger the value of the recap.
Model 2 through 6 suggest similar conclusions for the macroeconomic variables as the ones illustrated in model 1. In fact, the IPO level 3 months and 6 months before the recap are significant with a 90% confidence level. As expected, the sign of the coefficient for the 3 months before the recap is negative. The reason for its significance implies that the announcement of a dividend recap is a planned process; however, the coefficient for the 6 months prior is positive and economically significant (0.004), an issue that contradicts our belief. Moreover, M&A activity in the market seems to not have any economically significant influence in the amount a company is willing to pay out as a dividend. Finally, LIBOR is a key driver, since leverage loans are structured based on this rate. In fact, the pipeline of the dividend recaps done within the study period of this thesis, shows that all the deals are designed based on this rate. The negative sign of the coefficient is straightforward: the higher the cost of the debt, the lower the debt the company is willing to use.
Regarding the holding period until the first recap, the sign of its coefficient is neither significant nor consistent over the different models. The pattern observed is that when we
control for industry by using dummy variables, it displays a positive sign. When we do not control for the industry, the output has a negative sign; therefore, we cannot say anything about this variable.
Concerning the control variables, we can observe that the initial investment and the risk aversion in the market are significant variables. They show a positive and negative sign in the coefficient, respectively, which is in line with what is expected from these variables.
Multinomial logistic analysis
The objective of this second analysis is to observe the behaviour of those firms that have done a dividend recap and to verify if there is evidence of a subsequent traditional exit type. This will be done by using a multinomial logistic regression. For this purpose, the equation of the model is as follows:
Since in only a few cases the company was exited through a Buyback or a Reverse Takeover, or it was Written Off, we decided to exclude those cases in our analysis. Therefore, we will only take into account those exits through an IPO, Trade Sale, and Secondary Sale.
In this second study, company specific information is used as independent variables. These exogenous variables are the ratio of first recap to initial investment and the holding period until recap.
Since the larger the size of a company the more likely to go public because of less direct and indirect costs in the IPO process, Ibbotson and Jaffe (1975), we use the initial investment variable so as to control for size. In addition, based on Sudarsanam (2005), firm size and lenghts of holding period are significant determinants of the exit strategy. Furthermore, as we expect that a higher recap is positively correlated with the company size, we are also going to use the recap value as a proxy for the magnitude of the company.
This model was controlled for the holding period of the investment, the IPO market activity three months before the exit, and the industry of the company. The rationale for using the holding period as a control variable is that an exit through an IPO after a certain period of time is not possible. Based on Sousa (2010), an IPO is the preferred exit route, but when it is not possible, a secondary sale is the alternative exit in order to avoid long holding periods. The reason for considering the IPO market activity at the moment of the exit is that the IPO market “hotness” is directly correlated with a more likely exit through an IPO. Indeed, we used the three months before the exit date because firms are more likely to exit their investments through a public offering when the market returns are higher in the period between six to three months before the public offering. In addition, the industry of the companies is the last control variable included in the model, which is controlled by the use of dummies variables.
Table 5 displays the results of three models of the logistic model regression by using different combinations of the variables. The secondary buyout is the base case; therefore, the first equation compares the probability of an exit through an IPO relative to a Secondary Sale, and the second equation compares the probability of a Trade Sale exit relative to a Secondary Buyout.
Models 1 through 3 consistently suggest similar conclusions about the ratio of the recap to initial investment and holding period to recap. First, based on the results of the models, we can say that the higher the ratio recap initial investment, the less likely an exit through an IPO relative to a secondary buyout. Although, not statistically significant, we can also say that the higher this ratio, the lower the probability of exiting through a Trade Sale. Second, we can conclude that the longer the holding period until the recap, the more likely an exit through an IPO. This can be indicative of a pattern observed in the market about the use of dividend recaps prior to an IPO. Finally, we can observe that the longer the holding period of an investment after a recap, the less likely that the portfolio company will be exited through a Trade Sale. That said, this is not in line with Sousa (2010), who concludes that PE firms tend to exit their investment through a trade sale as a last resource route.
Conclusion
A regression and a logistic multivariate analysis were used for the purpose of answer the thesis question: What does the empirical evidence suggest about the drivers of dividend recaps and a PE’s subsequent decision on the exit route? For the purpose of this study we analysed a sample of 436 companies that have done a dividend recap during the period between January 2000 and June 2013.
From the first study, the regression model, we can conclude that the IPO activity and the LIBOR are key drivers for the level of debt a company is willing to take with the purpose of pay a special dividend to their private equity investor.
After the second analysis we arrived to three conclusions. First, we concluded that the higher the ratio of recap to initial investment in the company, the less likely that the PE sponsor will exit through an IPO relative to a Secondary Sale. Second, we conclude that the longer the holding period of the investment at the moment of the recap, the more likely to be exited through an IPO relative to a Secondary Buyout. Last, we conclude that after a recap, a company is less likely to exit through a Trade Sale relative to a Secondary Buyout.
References
Axelon, U., Jenkinson, T., Strömberg, P., and Weisbach, M., 2010. Borrow cheap, buy high? The Determinants of Leverage and Pricing in Buyouts. Working paper.
Cumming, D., and MacIntosh, J., 2003. A cross-country comparison of full and partial venture capital exits. Journal of Banking & Finance 27 (2003) 511–548.
Chapman, J., and Klein, P., 2010. Value Creation in Middle-Market Buyouts: A Transaction-Level Analysis. Private Equity Fund Types, Risks and Returns, and Regulation, Wiley and Sons, New York.
Eckbo, B.E., Thorburn, K.S., 2008. Corporate Restructuring: Breakups and LBOs. Tuck School of Business at Dartmouth. Working paper No. 2008-49.
Financial Times, 2012. US companies return to ‘dividend recaps’. Vivianne Rodrigues, October 17, 2012.
Financial Times, 2012. Private equity eyes dividend ‘recaps’. Michael Stothard and Dan McCrum, October 23, 2012.
Financial Times, 2012. US dividend recaps rise. Stephen Foley, December 20, 2012.
Financial Times, 2013. Dividend recaps make return in Europe. Anne-Sylvaine Chassany, April 14, 2013.
Harford, J., and Kolasinski, A., 2010. Evidence on how private equity sponsors add value from a comprehensive sample of buyouts and exit outcomes.
Harris, R., Jenkinson, T., and Kaplan, S., 2012. Private Equity Performance: What Do We Know? Working paper.
Hotchkiss, E., Smith, D., and Strömberg, P., 2012. Private Equity and the Resolution of Financial Distress. Working paper.
Jensen, M.C., 1997. Eclipse of the Public Corporation. Harvard Business School; Social Science Electronic Publishing (SSEP), Inc.; National Bureau of Economic Research (NBER); European Corporate Governance Institute (ECGI).
Kaplan, S., and Schoar, A., 2005. Private equity performance: Returns, persistence, and capital flows. Journal of Finance 60, 1791-1823.
Kaplan, S. and Stromberg, P., 2009. Leveraged Buyouts and Private Equity, Journal of Economic Perspectives 23, 121-146.
Kushner, M., 2004. Searching for exits in a challenging market. International Financial Law Review 2004 supplement, 95-98.
Lerner, J., Hardymon, F., 2002. Venture Capital and Private Equity – A Casebook: Volume II. New York: John Wiley & Sons, Inc.
Lerner, J., Sorensen, M., and Stromberg, P., 2010. Private equity and long-run investment: the case of innovation, Journal of Finance 66, 445-477.
Reuters, 2013. Moody’s sees dividend recaps slowing in 2013. Joy Ferguson, February 27, 2013.
Reuters, 2013. Dividend recap loans back on European agenda. Claire Ruckin, April 8, 2013.
Schwienbacher, A., 2009. Venture Capital Exits. Forthcoming as book chapter “Venture Capital Exits” in Companion to Venture Capital (Robert W. Kolb Companion to Finance Series): Wiley/Blackwell (Ed. D. Cumming).
Sousa, M., 2010. Why Do Private Equity Firms Sell to Each Other? Working paper.
Sudarsanam, P., 2005. Exit Strategy for UK Leveraged Buyouts: Empirical Evidence on Determinants. Cranfield School of Management. Working paper.
Wall Street Journal, 2013. Risky Business? PE’s Dividend Recap Gold Rush. Shasha Dai, January 10, 2013.
Annexes
Figure 1: Number of deals
This figure shows the number of deals based on the type of exit route. It only considers those portfolio companies of Private Equity sponsors that have their headquarters in Canada, Puerto Rico or the United States. Furthermore, it includes only the dividend recap, excluding recapitalization through stock repurchases. Finally, this figure displays the number of recaps relative to the total number of traditional exits.
0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 200 400 600 800 1,000 1,200 1,400 1,600 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Trade Sale Secondary Sale IPO Reverse Takeover
Figure 2: Volume of deals
This figure shows the total proceeds of the deals based on the type of exit route. As in the previous graph, it only considers those portfolio companies of Private Equity sponsors that have their headquarters in Canada, Puerto Rico or the United State, and it only includes the dividend recap, excluding recapitalization through stock repurchases. Finally, this figure displays the total proceeds of recaps relative to the total proceeds of traditional exits.
0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 50,000 100,000 150,000 200,000 250,000 300,000 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Trade Sale Secondary Sale IPO Reverse Takeover
Table 1: Recaps by private equity firm
Table 1 reports the quantity of dividend recaps done by the top 21 General Partners between January 2000 and June 2013. If more than one private equity firm is involved in the recap, only the leader is considered.
Number of Recaps Only one Recap More than one Recap Exited
Companies Still Held
ABRY Partners 13 9 4 6 3 Berkshire Partners 12 7 5 4 3 Bain Capital 11 7 4 5 2 Warburg Pincus 11 8 3 4 4 Carlyle Group 10 10 0 8 2 GTCR Golder Rauner 10 9 1 6 3
Madison Dearborn Partners 10 8 2 8 0
Welsh Carson Anderson & Stowe 10 9 1 6 3
American Securities Capital Partners 9 7 2 3 4
Texas Pacific Group 9 7 2 3 4
Thoma Bravo LLC 9 7 2 1 6
Hellman & Friedman 8 7 1 5 2
Apollo Partners 7 6 1 5 1
Code, Hennessy & Simmons 7 7 0 5 2
Court Square Capital Partners 7 6 1 2 4
Joseph Littlejohn Levy 7 3 4 1 2
Leonard Green 7 6 1 1 5
New Mountain Capital 7 5 2 0 5
Oak Hill Capital Partners 7 4 3 2 2
TA Associates 7 7 0 5 2
Thomas H. Lee 7 6 1 6 0
Status Recap Numbers
Table 2: Descriptive stats by different rounds of recaps
Table 2 compares the recap amount, holding period before the recap, holding period of the investment (only for those exited investments), and the ratio of recap to initial investment. All this information is organized by the round of recap done in the firm. Panel A contains
information about both exited and companies under PE management. Panel B observes only the exited investments and panel C considers the investments under PE management
Panel A: Exited investments and companies under PE management
Panel B: Exited investments
Panel C: Investments under PE management
Round of Recap N Mean Median SD Mean Median SD N Mean Median SD
1st Round 436 369.60 280.00 339.31 3.65 2.99 3.25 236 1.09 0.83 1.20 2nd Round 81 401.93 255.00 415.95 3.96 3.65 2.97 44 0.88 0.74 0.83 3rd Round 13 330.24 300.00 223.18 5.33 4.71 1.94 7 1.23 0.56 2.71 4th Round 5 686.00 360.00 865.72 6.29 5.21 2.73 2 0.68 0.68 0.08
Recap/ Initial Investment Holding Period before
Recap Recap
Round of Recap N Mean Median SD Mean Median SD Mean Median SD N Mean Median SD
1st Round 263 334.86 235.00 314.38 4.14 3.23 3.72 6.29 5.67 3.87 137 1.18 1.00 1.05 2nd Round 44 362.84 287.50 295.82 3.75 2.67 3.58 7.19 6.45 3.96 22 1.01 0.77 0.93 3rd Round 7 276.87 230.00 184.89 4.32 3.98 1.43 6.70 6.50 1.41 2 0.29 0.29 0.38 4th Round 3 958.33 360.00 1,097.52 3.96 3.37 1.09 6.60 6.50 1.35 1 0.74 0.74
-Recap Holding Period before Recap
Holding Period of Investment
Recap/ Initial Investment of Exited Companies
Round of Recap N Mean Median SD Mean Median SD N Mean Median SD
1st Round 173 422.40 330.00 368.75 3.16 2.56 2.26 99 0.99 0.60 1.37 2nd Round 37 448.42 250.00 525.23 4.16 4.36 2.05 22 0.75 0.69 0.70 3rd Round 6 392.50 360.00 264.36 6.33 6.55 2.00 5 2.18 1.49 3.12 4th Round 2 277.50 277.50 180.31 8.63 8.63 1.15 1 0.63 0.63
-Recap Holding Period before
Recap
Recap/ Initial Investment of Still Held Companies
Table 3: Descriptive stats of different type of exits
Table 3 compares the recap amount, holding period before the recap, holding period of the investment, and the ratio of recap to initial investment. All this information is organized by the type of exit on the investment. Note, the current portfolio companies are also considered.
Holding Period before 1st Recap Holding Period of Investment
Exit N Mean Mean Mean N Mean
Trade Sale 93 323,12 3,93 6,15 54 1,22 Secondary Sale 111 284,15 4,19 6,52 51 1,30 IPO 53 470,26 4,42 5,89 28 0,80 Reverse Takeover 1 400,00 6,26 7,05 1 2,99 Buyback 1 165,00 7,94 8,80 1 1,10 Write Off 4 247,38 2,55 7,82 2 1,61 Currently PE/VC Backed 173 422,40 3,16 99 0,99
Total 436 236
Recap Value
1st Recap/ Initial Investment
Table 4: Regression analysis
The dependent variable is the value of a recap. The exogenous variables are the IPO volume, the M&A volume, the LIBOR and the holding period before the recap date. Standard errors are between parentheses. The significance level of 10%, 5% and 1% are indicated with a *, ** and ***. The variables ipo, ipo3m, and ipo6m, refer to the level of the level of IPO
activity at the moment of the recap, 3 months before and 6 months prior. The variables M&A, M&A-3M, and M&A-6M, are the M&A activity at the moment of the recap, 3 months earlier and 6 months before. Libor refers to the interest rate at the moment of the recap. The
variables hprecap and initialinv are the holding period of the investment until the first recap and the initial investment done by the sponsor in the company. The aversion variable is the difference between the US Treasury bill rates and Barclays US Corporate Investment Grade Yield at the moment of the recap. Finally, consumer, healthcare, industrial and servinfomedia (Service, Information and Media) are dummies variables that control for the different kind of industries.
Table 5: Logistic multinomial regression
The dependent variable is the type of exit and the base case is the Secondary Buyout. The exogenous variables are the ratio of recap to initial investment (recapinitialinv) and the holding period of the investment until the first recap (hprecap). Standard errors are between parentheses. The significance level of 10%, 5% and 1% are indicated with a *, ** and ***. The variables recap, initialinv, and hp, refer to the level of the value of the first recap, the initial investment in the company and the holding period of the investment. The variable ipo3m is the IPO activity 3 months before the exit date. Finally, consumer, healthcare, industrial and servinfomedia (Service, Information and Media) are dummies variables that control for the different kind of industries.
Variables IPO TS IPO TS IPO TS
recapinitialinv - 0.007 - 0.305 0.001 - 0.141 - 0.714 * - 0.091 (.33) (.23) (.32) (.20) (.39) (.18) hprecap 0.414 * 0.037 0.410 * 0.046 0.361 * 0.046 (.21) (.14) (.21) (.14) (.21) (.14) recap 0.000 0.001 0.002 ** 0.001 (.00) (.00) (.00) (.00) initialinv 0.001 ** - 0.001 0.001 *** - 0.000 (.00) (.00) (.00) (.00) hp - 0.667 *** - 0.129 - 0.658 *** - 0.111 *** - 0.608 *** - 0.121 (.21) (.14) (.21) (.14) (.20) (.14) ipo3m - 0.000 * - 0.000 - 0.000 * - 0.000 - 0.000 - 0.000 (.00) (.00) (.00) (.00) (.00) (.00) consumer 2.409 ** 0.673 2.351 ** 0.625 1.868 ** 0.617 (1.02) (.62) (1.00) (.61) (.86) (.61) healtcare 1.862 0.409 1.835 0.482 1.004 0.505 (1.21) (.74) (1.19) (.73) (1.09) (.73) industrial 2.234 *** - 0.048 2.218 *** 0.075 1.307 ** 0.058 (.85) (.49) (.82) (.48) (.66) (.48) servinfomedia - - - -(.) (.) (.) (.) (.) (.) cons 0.024 0.975 0.048 0.794 1.341 0.543 (1.33) (.00) (1.32) (.92) (1.16) (.89) Observations LR Chi 2 Pseudo R2 133 30.17 0.1068
Model 1 Model 2 Model 3
133 46.15 0.1634 133 43.03 0.1524
