The Performance of Secondary Buyouts

The following post comes to us from François Degeorge, Professor of Finance, Swiss Finance Institute, University of Lugano (USI); Jens Martin, Assistant Professor of Finance at the University of Amsterdam; and Ludovic Phalippou, Associate Professor of Finance at the University of Oxford Saïd Business School.

In the past two decades, private equity buyout transactions have grown from a niche phenomenon to a ubiquitous form of corporate ownership (e.g., Strömberg, 2008). Traditionally buyouts have involved private equity funds buying companies or divisions from families or conglomerates: such transactions are known as primary buyouts (PBOs). A major trend accompanying the growth of private equity has been the rise of secondary buyouts (SBOs): transactions in which a private equity fund buys a company from another private equity fund. In our paper, The Performance of Secondary Buyouts, which was recently made publicly available on SSRN, we compare buyer returns in SBOs and PBOs.

We find that SBOs generate much lower average returns for buyers than comparable PBOs. The average SBO distributes $0.4 less cash per $1 of cash invested and has an approximately 15% lower Internal Rate of Return (IRR). These results hold irrespective of the control variables we use, including time, country, industry, and private equity firm fixed effects.

The underperformance of such a large and fast-growing class of private equity investments is a puzzle. We analyze three potential explanations, which are not mutually exclusive: (1) agency problems between private equity funds’ managers and investors generate an incentive to overpay for SBOs; (2) SBOs are less risky; and (3) significant value has already been extracted by the selling fund. Our evidence is most consistent with explanation (1). Specifically, we find evidence consistent with the “go for broke” hypothesis of Axelson, Strömberg, and Weisbach (2009).

In their model, Axelson et al. (2009) show that fund managers sometimes have a strong incentive to overpay for transactions occurring towards the end of their investment period. This is because the investment period is fixed (usually five years) and investors have committed capital ex-ante. Funds have an incentive to spend all of the capital committed before “losing it” at the end of the investment period. We argue that SBOs are natural investments for funds who want to spend money fast because SBO buyers suffer from less asymmetric information and bear lower search costs than PBO buyers.

We find that SBOs made late in the buying fund’s investment period underperform relative to similar PBOs, while SBOs made in the first half of the fund’s investment period have the same performance as similar PBOs. Furthermore, the underperformance of “late SBOs” is exacerbated when the fund has more capital left to be spent, i.e. when it is under buying pressure. Late SBOs also underperform more when the transaction is large relative to the fund’s size. The sign and large economic magnitude of these two cross effects are consistent with the “go for broke” hypothesis.

To evaluate the risk hypothesis, explanation (2) above, we run Probit regressions to model the likelihood of the return falling below certain thresholds (e.g. Cash Multiple falling below one or below Mean). After accounting for the lower average return of late SBOs, we find that late SBOs are equally likely to have returns below these thresholds as similar PBOs. We interpret these results as showing that downside risk is similar for late SBOs and other buyout transactions. We also find no evidence that late SBOs are more liquid, in that they do not return cash faster to investors. These results also hold irrespective of the control variables we use.

When considering whether value gains are too limited for the buyer of an SBO, as in explanation (3) above, we find evidence that, although a large number of SBOs are driven by agency costs, certain SBOs are value-enhancing transactions. Case-based evidence on SBOs indicates that funds may have different specialties and that a company may need a different set of skills, hence a different owner, as it evolves over time (e.g. Strömberg, 2013). Consistent with this idea, we find that SBOs acquired by a specialized fund perform better.

We find evidence of performance persistence at the target company level: The returns of successive buyout transactions are positively correlated. We view this result as consistent with the idea that some companies are better off under private equity fund ownership (e.g., Jensen, 1989; Kaplan, 1991; Strömberg, 2008), implying that some SBOs are motivated by the necessity to stay in private equity hands. We also find evidence that some SBO buyers perform better when sellers are under pressure to exit their investments: SBOs purchased from private equity firms that are fund raising at the time of the transaction outperform other SBOs.

Finally, we investigate the Net Present Value (NPV) achieved by the end-clients of the buyout funds (i.e., pension funds, endowment etc.). We use the stock-market index as a discount rate and find that only late SBOs, taken together, have a negative NPV. Only 88% of what is invested is returned to investors, in present value terms. Other transactions (early SBOs, early and late PBOs) have positive NPVs, in line with what has been documented in the literature (Robinson and Sensoy, 2011, Harris et al., 2013). We also find that these end-clients seem to vote with their feet: the follow-on fund is significantly smaller when the focal fund has made some late SBOs, even controlling for the performance of the focal fund. Hence end-clients penalize the fund managers beyond the penalty for lower overall performance.

A few contemporaneous studies examine secondary buyouts empirically and present results that are consistent with and complementary to ours. Unlike this paper, most focus on the corporate finance side of SBOs. Wang (2013), Jenkinson and Sousa (2012), Zhou, Jelic and Wright (2013) and Bonini (2012) find that, on average, SBOs exhibit smaller operating performance gains than PBOs. Arcot, Hege, Gaspar, and Fluck (2013) find that price multiples (ratio of purchase price to EBITDA) are higher for SBOs, and in particular for SBOs transacted in the second part of the investment period. This finding is consistent with the “go for broke” hypothesis. Our access to return data allows us to show that these higher prices translate into lower returns and do not reflect lower risk. Finally, Achleitner and Figge (2013) study SBO returns and find lower average performance for SBOs; their results are not statistically significant, however, perhaps due to a smaller sample size.

The full paper is available for download here.

Both comments and trackbacks are currently closed.