By LARRY SWEDROE
Assets managed by private equity managers grew from $130 billion in 1998 to $2 trillion at the end of 2018. But has the dramatic increase in their allocation been justified by performance?
My September 12, 2019, article for Advisor Perspectives provided a summary of the research on the performance of private equity. Unfortunately, it was not encouraging — in general, private equity has underperformed similarly risky public equities, even without considering their use of leverage and adjusting for their lack of liquidity. However, the authors of the 2005 study Private Equity Performance: Returns, Persistence, and Capital Flows offered some hope. They concluded that the evidence suggests that private equity partnerships are learning—older, more experienced funds tend to have better performance — and there’s some persistence in performance. Thus, they recommended that investors choose a firm with a long track record of superior performance.
The most common interpretation of this persistence has been either skill in distinguishing better investments or in the ability to add value post-investment (e.g., providing strategic advice to their portfolio companies or helping recruit talented executives). The research does offer another plausible explanation for persistence: Successful firms are able to charge a premium for their capital.
Robert Harris, Tim Jenkinson, Steven Kaplan and Ruediger Stucke confirmed the prior research findings of persistence of outperformance in their November 2020 study, Has Persistence Persisted in Private Equity? Evidence from Buyout and Venture Capital Funds. However, it was only true for venture capital (not the full spectrum of private equity), as they found that there was no persistence of outperformance in buyout firms.
An interesting question is: Do institutional private equity investors follow the research findings and choose funds run by managers that have performed well in the past, particularly so-called top-quartile funds, while avoiding first-time funds? Amit Goyal, Sunil Wahal and Deniz Yavuz sought the answer to that question in their August 2021 paper, Picking Partners: Manager Selection in Private Equity, in which they analysed the selection of investment managers — general partners (GPs) — to fulfil allocations to private equity. They assembled a database from Pregin of 100,506 capital commitments originating from 8,801 public and corporate pension systems, endowments, foundations and sovereign wealth funds from 61 countries over the period 1990-2019. Capital commitments were designated to a diverse set of sub-asset classes often referred to as alternatives (buyouts, direct lending, distressed equity, growth, infrastructure, mezzanine financing, natural resources, real estate and venture capital), all of which used the same closed-end delivery vehicle.
Their analysis began by first estimating selection equations that sought to explain GP choice relative to an opportunity set. For each commitment, they generated the opportunity set from GPs raising capital in the same sub-asset class, within one year of the actual decision, and comparable in size to the fund that won the capital allocation. They then asked which criteria influenced the choices of limited partners (LPs) and used the counterfactual opportunity set to assess post-selection excess performance. This allowed them to understand whether the criteria used by LPs provided information that was helpful in selection. In each step, they organised their tests around two selection criteria: (1) prior performance of the GP and (2) channels through which LPs could acquire information about GPs. These channels included LPs’ prior investments, LPs’ peers’ investments, and geographic proximity between the LPs and the GPs. Following is a summary of their findings:
Almost 50 percent of capital commitments were made to GPs without any performance history — performance history may not have existed if the GP was raising capital for the first time or, for younger GPs, where sufficient time had not passed since their first funds to observe distributions.
GPs without a performance history are 30 percent more likely to be selected than those with observable performance. Public information about first-time funds is likely to be limited, but it is possible that LPs choosing these funds have access to private information.
The excess internal rates of returns (IRRs) of selected first-time funds were more negative than those of non-selected first-time funds.
Similar selection effects were found across all types of funds, including buyout and venture capital.
When prior performance was observable, a GP in the 4th quartile (top performance) of the distribution of pre-decision IRRs was 33 percent more likely to be selected than a 1st quartile GP. However, the average difference in excess IRRs between selected and non-selected funds sponsored by 4th quartile GPs was -1.10 percent (t-statistic = -0.58).
Performance chasing by U.S. LPs, though not by international LPs, was observed.
Prior hiring experiences by peers did not convey information about the future performance of selected funds.
There was a 16.5 percent probability of a repeat investment, a 634 percent increase over the unconditional probability—an LP’s prior experience with a GP was the most influential determinant of choice. However, selected GPs did not outperform non-selected GPs.
While an LP was almost twice as likely to choose a local GP (home bias) over a nonlocal GP from the opportunity set, the preference for local GPs did not provide information, as the average excess IRRs of selected funds was statistically indistinguishable from non-selected funds in the same group.
While home bias existed both in the U.S. and internationally, it was dramatically higher (more than four times as great) outside the U.S. In both cases, home bias did not deliver larger post-selection IRRs.
U.S. university endowment selections are highly influenced by prior investments made by their peers.
There was no evidence that combinations of selection criteria generated positive differences in excess IRRs — all excess IRR differences were either negative or statistically indistinguishable from zero.
Goyal, Wahal and Yavuz also examined whether the choice of first-time GPs could be explained by the fact that they were not truly rookies — their founding partners may have been veterans of well-established firms. They found: “The excess IRRs of selected rookie first-time funds are lower than those of non-selected rookie first-time funds by 3.06% (t-statistic = 2.09), and the equivalent difference for veteran first-time funds is 2.33% (t-statistic = 1.70).”
Goyal, Wahal and Yavuz noted: “The tendency to invest in GPs without prior performance is surprising, particularly since three-to five-year track records are standard screening devices in public equity and fixed income.” This is particularly surprising because the tendency was more apparent in the latter time period of the sample — the LPs not only ignored the academic research findings but also their own experiences!
The researchers also noted that one explanation for the poor performance generated by home bias is that political representation on investment committees in public pension plans deleteriously affects performance either because of suboptimal decision-making due to control or corruption issues. They found that for public pension plans reinvesting with the same GP, the difference in excess IRRs between selected and non-selected funds for public pension systems was -5.92% (t-statistic = -2.84). Such stark performance differences were not present for other types of LPs.
Summary of the findings
Summarising their findings, Goyal, Wahal and Yavuz found:
the lack of a track record is not a deterrent to capital commitments; there is evidence of performance chasing;
prior investment by a peer institution, or by the LP itself, matters for GP choice; and
U.S. LPs display home bias, as they are more likely to select local GPs.
Unfortunately, these behaviours subtract from performance, begging the question: Why do LPs ignore the evidence suggesting that they invest only with GPs with a long record of outperformance? Their findings led the authors to conclude: “There is no clear-cut evidence that LPs can pick private equity funds that deliver future excess IRRs relative to the opportunity set.”
Investor takeaway
The late David Swensen, legendary chief investment officer of Yale’s endowment fund, offered this caution on private equity investing: “Understanding the difficulty of identifying superior hedge fund, venture capital and leveraged buyout investments leads to the conclusion that hurdles for casual investors stand insurmountably high. Even many well-equipped investors fail to clear the hurdles necessary to achieve consistent success in producing market-beating, active management results. When operating in arenas that depend fundamentally on active management for success, ill-informed manager selection poses grave risks to portfolio assets.”
Those who choose to ignore Swensen’s warnings still need to understand that, due to the extreme volatility and skewness of returns, it is important to diversify the risks of private equity. This is best achieved by investing indirectly through a private equity fund rather than through direct investments in individual companies. Because most such funds typically limit their investments to a relatively small number, it is also prudent to diversify by investing in more than one fund. And finally, top-notch funds are likely closed to individual investors.
They get all the capital they need from the Yales of this world.
© The Evidence-Based Investor MMXXIV. All rights reserved. Unauthorised use and/ or duplication of this material without express and written permission is strictly prohibited.