Measuring the Value of LP Fund-Selection Skill: A Fair Comparison Framework

6. Apr. 2020

15 mins read

We recently introduced a framework to compare risky assets, both public and private, on a fair basis (Jeet, 2019). Our findings were conditional on various settings: 1) the horizon period over which terminal wealth accumulates; 2) the number of fund commitments per vintage; 3) the selection of the private market strategy such as buyout or venture capital; 4) the choice of the default investment such as S&P 500 or 3m LIBOR; and 5) a commitment pacing strategy that specifies the percentage of uncommitted capital committed to each new vintage. We now examine the sensitivity of our findings to these settings.

Our earlier findings were also based on an assumption that the LP had no fund-selection skill – in other words, we assumed the LP simply picks funds at random. In practice, however, it is likely that an LP’s fund-selection skill has a meaningful impact on the risk and return profile of a private capital portfolio. How much does LP fund-selection skill matter to portfolio performance? Is fund-selection skill more important for some private market strategies than others?

Lerner, et al. (2007) found that experience, sophistication, and access are among the top factors that cause a wide variation in the returns that institutional investors realize from private equity. The three factors together constitute fund-selection skill. Consequently, we analyze the sensitivity of the risk and return of a private asset allocation to the LP’s fund-selection skill.

We present five representative examples of skill types along with their associated selection probability distributions (Figure 1).

Figure 2 provides a comparison of terminal wealth resulting from investing in US buyout funds with different skill types. In terms of mean returns, Poor Skill and Some Skill are clearly different. However, mean returns for Hit or Miss, No Skill, and OK Skill are indistinguishable. In terms of volatility, the skill types are quite different as Hit or Miss is much worse than No Skill and OK Skill. Consequently, risk-adjusted returns (i.e., returns per unit of risk) are better for a No Skill investor than a Poor Skill or a Hit-or-Miss investor. Some Skill or OK Skill are clearly the best skill types in terms of the risk-adjusted returns. Having skill helps improve mean returns while keeping volatility low.

In the full version of the paper, we explore the interaction of skill and other portfolio construction choices for several private market strategies including buyout, venture capital, real estate, and mezzanine.

Figure 3 compares the risk-adjusted performance of various private market strategies with the public market in three different settings: 1) As reported, 2) No Skill, and 3) With Skill (both in public and private markets).

While the ordering of various public and private markets investments more or less does not change from As Reported to No Skill, (real estate took the second position pushing buyouts to third), the magnitudes of risk-adjusted returns did change significantly. This is because the fair comparison framework improves the risk estimates as it does not suffer from return smoothing, which is the case of As Reported numbers. The With Skill comparison is also interesting; the ordering is again more or less the same as before, but the magnitude has gone up again because of skill. Mezzanine remained the best investment. Early and late-stage venture funds came in last, even with skill.

Among the private market strategies, we found that mezzanine performed best in terms of risk-adjusted returns, irrespective of fund selection skill. However, mezzanine is a small market compared to buyout and real estate, hence there may be less opportunity for larger investments and cross-sectional diversification. The next best strategy is real estate followed by buyout. Venture investments have the lowest risk-adjusted returns because of their high underlying risk, but they still offer a better investment opportunity than public markets with or without skill.

Download white paper

Erfahren Sie mehr