By Greg Brown, Executive Director of the Kenan Institute of Private Enterprise, and Research Director of the Institute for Private Capital
There’s been quite a bit of interest in a recent white paper by Antii Ilmanen, Swati Chandra, and Nicholas McQuinn of AQR, which examines expected returns for Private Equity (PE).
The paper’s authors claim that, when properly risk-adjusted, the returns on private equity are not attractive relative to public market stocks. The conclusion on lower returns is surprising to many because it’s at odds with what is now the well-documented outperformance of PE over the last few decades.
So what gives? How do we reconcile this disparity?
Let’s start with the AQR research. To reach their conclusion, the authors considered three methods for examining returns. I’ll focus on just one of them, because it’s where I think the controversy lies. Specifically, I’ll examine the historical evidence on past average returns.
The authors make the case that PE has a higher hurdle rate than large cap stocks because more leverage implies a higher market beta and because PE companies tend to be smaller and have a value tilt.
First, consider the fact that after a buyout, leverage is much higher than for a typical public firm. This is true—it’s common practice for buyout firms to add leverage as part of a transaction. A caveat is that firms targeted for buyouts tend to have lower risk in their underlying business, or what financial economists call “asset betas” in the typical model of firm capital structure. Consequently, the overall riskiness of a typical buyout with increased leverage is not as high as we might naively assume, given the amount of debt. The AQR analysis assumes leverage (beta) of 1.2 times the index. This seems about right, given what various academic studies have found (see Korteweg, 2018, for a summary). So, leverage doesn’t appear to resolve the discrepancy.
Second, buyouts tend to be smaller in size than the typical public company, mostly in the small-cap or even micro-cap range. We know that small-cap stocks are considered riskier than large cap stocks, yet most studies compare performance of PE to large-cap stock performance and the S&P 500 in particular. Thus, this seems like an important point.
Likewise, buyouts are biased toward value stocks versus growth stocks. In “factor parlance,” buyouts tend to have high exposure to the high-minus-low-book-to-market, or HML, factor, which also has a significant premium associated with it historically. So this also seems like an important point.
Given this direct analogy to public markets, the authors of the AQR paper then calculate what historical returns would have been for a levered version of the Russel 2000 Value Index, and compare this to the historical return on the Cambridge Private Equity (U.S. Buyout) Index.
The authors show (in their Exhibit 1; see below) that there was outperformance of PE based on IRRs compared to annualized returns for small-cap value stocks, so no real controversy there. In fact, even after desmoothing the PE returns, the Sharpe ratios for PE (not reported by AQR) are much better than for public equity.
The real discrepancy seems to be the results presented on Public Market Equivalents (PMEs) in Exhibit 2 of the AQR paper. The idea behind PMEs is fairly straightforward. The measure provides a multiple of what an investor would have earned in private equity versus their earnings from investing the same amount in a public market index. Thus, values greater than 1.0 suggest PE outperformance versus the public benchmark; values less than 1.0 suggest the opposite. The AQR Exhibit 2 shows that, over the last decade, PE returns have moderated, and were worse than a public benchmark with appropriate leverage, size and sector adjustments.
This is where the discrepancy begins to creep in. The AQR PME figure relies on data published about three years ago in a paper published in the Financial Analysts Journal (L’Her et al, 2016) that made adjustments for leverage, size, value, and sector. Let’s see what happens when we expand the dataset and bring it up to date. Through the Private Equity Research Consortium, I have access to data for a larger set of funds (also collected by Burgiss). The dataset covers 1,159 U.S. buyout funds with about $1.4 trillion in committed capital, from 1987-2018. These data, which include almost all institutional-quality buyout funds, reveal a somewhat different story–one more consistent with the results presented in Exhibit 1 of the AQR paper.
With the help of Wendy Hu at Burgiss, I calculated PMEs using the Russell 2000 Value Index levered 1.2 times, using data through the third quarter of 2018 (the most recent data currently available). The results, displayed in the table below, show that a pro rata investment in all U.S. buyout funds over the last 30 years would have yielded a PME of 1.09–which indicates better performance than the benchmark, and substantially better than the values reported in Exhibit 2 of the AQR paper.
Another important part of the argument put forth by AQR revolves around a declining trend in returns to buyouts. Once again, though, the trend is inferred from less data. The graph below plots a longer (and more up-to-date) history that reveals that the period of very strong performance was limited to vintages in the early 2000s. If we look back to 1987, then the recent experience seems much more typical. In fact, the average PME of vintage years from 2005 through 2014 is 1.10—almost exactly the same as for the full sample. Consequently, it appears that the AQR conclusions about low and declining PMEs when referenced against levered small-cap value stocks do not hold up against a larger, up-to-date dataset.
The AQR paper makes some very good, interesting points about the worrisome nature of valuation levels in the past few years. In the light of a more relevant dataset, however, the inference based on historical average returns should be reconsidered.
Korteweg, Arthur, 2018, Risk Adjustment in Private Equity Returns, University of Southern California working paper.
L’Her , Jean-François, Rossitsa Stoyanova, Kathryn Shaw, William Scott, & Charissa Lai, 2016, A Bottom-Up Approach to the Risk-Adjusted Performance of the Buyout Fund Market, Financial Analysts Journal 72 (4), 36-48.