Private equity (PE) and venture capital (VC) firms generally view the initial public offering (IPO) as the most desirable exit scenario by which to realize their returns, sparking keen interest in the risk-return profile of these stocks as compared to the overall market.
In his seminal work on IPOs, Ritter (1991) first drew attention to the significant underperformance of newly issued stocks during their first three years in the aftermarket. Interestingly, subsequent literature has found evidence that private-equity-backed (PE-backed) IPOs defy this trend. For example, studies by Degeorge and Zeckhouser (1993) and Holthausen and Larcker (1996) show that leveraged buyouts (LBOs) returning to the public markets outperform other newly listed stocks. In the case of venture capital, a study by Brav and Gompers (1997) establishes that venture-capital-backed (VC-backed) IPOs significantly outperform non-VC-backed IPOs.
Previous studies have sought to identify risk and return characteristics that might explain the differential performance of PE-backed IPOs. Most relevant to this paper is a study authored by Levis (2011) that uses Fama and French’s three-factor model to analyze a sample of PE-backed, VC-backed and other non-backed IPOs. Levis finds that PE-backed IPOs have larger market capitalizations and are more profitable than their non-sponsored counterparts. In examining the aftermarket performance of these IPOs, he also finds that the three-factor alphas of VC-backed IPOs are poorer than those of PE-backed IPOs, but that both of these groups generate higher returns than non-backed IPOs.
Although the three-factor profiles of private equity IPOs has been well established in the financial literature, such studies have yet to incorporate the factors from Fama and French’s most recent five-factor model (2014). In addition to the traditional factors for market risk (MKTRF for “market return over risk free rate”), size (SMB for “small minus big”) and value (HML for “high minus low”), this model includes two new factors: operating profitability (RMW for “robust minus weak”) and investment (CMA for “conservative minus aggressive”).
RMW is particularly relevant to the private equity industry in two respects. First, operational improvement is one of the primary means by which sponsoring funds seek to create value in their portfolio companies, with operating multiples such as EV/EBITDA serving as key metrics for evaluating investment performance (Axelson et. al, 2010). Additionally, expectations for operational profitability differ significantly by fund strategy. For example, a paper on private equity performance by Ljungvist and Richardson (2003) finds that venture capital investments tend to be significantly less profitable than private equity investments.
The investment factor was first introduced by Zhang et. al (2008) in response to the apparent underperformance of equity and debt issuers as compared to matched non-issuers. The authors propose that the reduced cost of capital from issuing securities incentivizes firms to increase investment, which in turn lowers average returns. They find that the new factor does indeed explain a significant part of the new issues puzzle, reducing 80 percent of IPO underperformance.
Fama and French (2014) find that positive exposures to RMW and CMA capture the high average returns associated with low market betas and conservative investing, whereas negative exposures explain the low average returns characteristic of unprofitable firms that invest aggressively. Given the differing return characteristics of PE and VC investments, it is likely that the loadings on these two factors would differ based on the stage of an investment’s sponsoring fund, although no research has been conducted on this topic to date.
In addition to five-factor analyses, there is also a notable lack of research on the specific risk and return characteristics of distressed stage funds. This gap in the literature is surprising, as distressed deals have become more popular over time, accounting for 3.2 percent of all LBO transactions from 2001-2007 versus 1.2 percent from 1970-2000 (Stromberg 2007); moreover, the most recent financial crisis in 2008 sparked further opportunities for distressed strategies. Research into the characteristics of distressed investing has potentially interesting implications, given the highly differentiated platform of such funds. For example, Stromberg (2007) establishes that distressed investments are the most risky form of LBO deals, with bankruptcy rates that are twice as high as other deal types.
The purpose of my analysis is two fold. First, I apply Fama and French’s most recently published five-factor model to establish the risk and return characteristics of IPOs backed by buyout, venture and distressed stage funds. I then proceed to parse out significant differences in the alphas and betas of IPOs based on each of the three types of sponsor in my dataset. My analysis is one of the first to apply the new five-factor terms for operating profitability and investment to the private equity industry, as well as one of few papers to give specific attention to distressed stage investments.
The main dataset on private equity IPOs is from Thompson One, which covers all private equity exits since 1970. I filtered this data to focus on companies that were backed by either venture-, buyout- or distressed-stage funds and exited via IPO. The sample I obtained from Thompson One only includes IPOs that took place from 1970 to 2014 on the New York Stock Exchange, NASDAQ Stock Market or American Stock Exchange. Of the 4,131 total IPOs in the dataset, venture-backed firms comprise over half of the sample with 2,729 IPOs; the 1,329 buyout- and 73 distressed-backed firms account for the remaining observations. Table 1 provides summary statistics for the price, shares issued and proceeds of IPOs based on sponsor stage.
Table 2 shows the output from the five-factor regressions. Results are broken down by each of the five time intervals following a firm’s IPO date: 1 month, 6 months, 12 months, 24 months and 36 months. These panels are further subdivided according to the stage of the firm’s sponsoring fund—venture, buyout or distressed.
The findings indicate that the five-factor model has significant descriptive power for sponsored IPOs. Venture-, buyout- and distressed-backed firms are similarly characterized by greater volatility, smaller market capitalizations, lower book-to-market values, lower operational profitability, higher levels of investment and some degree of abnormal returns in the first month following an IPO.
While these initial results hint at differences in alphas and risk factor loadings between sponsor stages, they do not establish whether such differences are statistically significant. To follow up on this point, I conducted a second analysis adding interaction terms to the five-factor model to test for IPO performance differentials based on sponsor stage.
Thus, in addition to the original five factors, the regressions shown in Table 3 include dummy variables for buyout and distressed sponsorship, plus interaction terms between these dummies and each of the five factors. Because the base case is venture, the constant and all five uninteracted factors in this table are identical to the coefficients from the venture regressions for the corresponding periods in Table 2.
Although there were no consistently significant differential effects for distressed terms, the analysis does reveal a few points of significant difference in the performance attribution of venture- and buyout-backed IPOs. Buyout sponsored firms are less volatile than venture sponsored firms during the first six to twelve months following an IPO. Consistent with previous studies, buyout-backed firms also tend to be larger and more value-oriented than venture-backed firms.
While there are significant differential effects for the original three Fama-French factors, no such differences appear to exist for RMW or CMA. Consistent differences in abnormal returns are also lacking, although the end-period alpha differential for buyout-backed IPOs is significant and negative with respect to venture-backed IPOs.
My results generally confirm findings from previous studies based on the original three Fama-French factors—market risk, SMB and HML. Regardless of sponsor stage, I show that all firms backed by private equity funds tend to be smaller and more growth-oriented than other public companies, even generating some degree of excess returns in the first month following their IPO date. Furthermore, venture-backed IPOs are more volatile than the market, as to be expected from the riskier strategies that characterize venture capital industry. Differential analyses for the three factors are also consistent with preceding financial literature; buyout-backed IPOs are less volatile, larger and more value-oriented than their venture-backed counterparts.
The new RMW and CMA factors shed light on previously unexplored aspects of private equity performance attribution. My results show that sponsored IPOs as a whole exhibit lower operating profitability and higher levels of investment during their first 36 months on the public markets. These results are consistent with Fama and French’s five-factor correlations (2014), which show that RMA and CMA vary positively with book-to-market value. Unlike the first three Fama-French factors, however, I find no evidence of differential effects for RMW or CMA by sponsor stage.
This paper set out to profile the risk and return characteristics of private-equity-backed IPOs based on the stage of their sponsoring fund. In addition to confirming previously established findings, the analysis also contributes new insights to the performance attribution of private-backed-IPOs via the five-factor model. These findings lay the groundwork for future research on the value-add of private equity funds in managing their portfolio companies.
This article is an excerpt from Ms. Eubank’s paper for her finance class at Dartmouth.