Writing our book Asset Allocation and Private Markets has led its authors to revisit an old adage, venture on an unusual path, and confront an academic theory.
Do not try to time the market
There is an old practitioner’s saying: investors should not try to time the market. Investing in funds at a specific point does not guarantee an immediate deployment of capital. Managers have significant latitude to deploy funds with three- to five-year investment periods.
Moreover, even if investors identify a future macro event, they are unable to decide when a fund should be raised. As managers raise funds on average every two to three years, there might not be any fund on offer at the time. To compound the issue, private market funds are raised over 6-18 months.
Even if a fund is closed before a major macroeconomic event, the manager might take advantage of (or suffer from) it by adapting the strategy. These actions cannot be foreseen by investors. Adding more spice to the mix, it is difficult to identify which vintage year of a strategy will benefit (or suffer) the most. Even if the performances of individual vintages tend to be correlated, the absolute performance varies. An additional twist is that private markets cycles are appearing, but most strategies are too recent to have documented patterns.
The best way to eliminate these risks is to invest regularly in private markets and avoid timing, as confirmed by academic literature. The logical consequence is that investors should give up tactical asset allocation. This is a big ask. This is also where we found surprising conclusions.
We ran portfolios of private market funds using historical cash flow data from Cambridge Associates. To account for the cost of unused capital, we apply a 0% risk-free rate. We then compute a complete annualised return (CAR) to evaluate private market investments entirely, including the performance drag (or opportunity costs) of the unused capital. We used a ‘market neutral’ portfolio of private market funds, replicating the overall regional and strategic breakdown of capital raised by managers, which had to be simplified due to data limitations. We implemented it in three-, five- and seven-year investment programmes, with individual starting dates ranging from 1993 to 2003.
To our surprise, the best option was to deploy over seven years. The CAR ranges from 4.14% to 5.13%. Running a programme over five years lead to a range of 3.80% to 6.74%, and over three years of 3.31% to 8.08%. Indeed, time-diversification is a ‘free lunch’ and advisable. We made a second surprise discovery, leading us down an unusual path.
The wild side of overcommitting
We discovered investors would effectively deploy at best 22.19% to 44.79% of their capital over a seven-year cycle. The figures are 31.07% to 50.94% for a five-year programme, and 43.04% to 61.32% for three-years. We decided to explore the path of overcommitting: how much would investors have to commit to reach an effective 90% net maximum exposure?
The results are not for the faint-hearted. With a seven-year programme, overcommitments reach 100% to 305%, and an average CAR of 9.56%. With a three-year programme, the overcommitment would be 46% to 109% and an average CAR of 8.57%.
In practice, it is impossible to forecast the next three to seven years. Logically, investors would apply the lowest level of overcommitment historically validated, that is, 46%, 76%, and 100% respectively. With these scenarios, investors would generate an average CAR of 7.22% (three-year) to 7.83% (seven-year). The variation is low with a seven year programme as returns range from 7.14% to 8.59%. Overcommitting reconciles predictable and significant returns with prudent deployment of capital.
Pension funds might face regulatory constraints preventing overcommitments. The logical option is to use a separate investment vehicle enjoying more operational freedom. This goes against the grain and led us to confront academic theory.
The fund is not a disposable artefact
The fund structure is often seen as an annoyance in the analysis of private markets. Some authors decided to force private markets into the listed markets framework by ‘de-smoothing’ fund returns. Focusing on leveraged buyouts (LBOs), proponents of this approach argue it does not create value and is essentially a leveraged index of listed small- and mid-sized companies. That is an error.
Scholars recognise that listed markets are not efficient. Nevertheless, private markets are seen as a ‘sticky’ version of their listed counterparts, flawed by stale pricing. Regardless of these opinions, the facts are stubborn. First, market caps are at best a proxy of the value of listed companies. Upon sale, the price differs from its market cap. Logically, the fair value of a private asset differs from its transaction price. The fair value method used to assess private market funds is the result of regular appraisals of stakes by managers with different views and shareholder rights. The framework of the transaction is key. Valuations are, therefore, a subjective exercise. ‘De-smoothing’ the returns by computing more frequent proxies only raises noise.
The proponents of private markets as “listed market returns plus leverage” argue that the performance of LBO funds can be replicated by leveraging an index of listed stocks. This is misguided. First, they replicate ex post the exposure of LBO funds with a bespoke index. In practice, managers are opportunistic and do not know in advance where they will invest. Moreover, some sectors of LBO investment are not represented on stock exchanges. Applying the replication approach ex ante is impossible. Second, the risk associated with the “listed-markets-plus leverage” approach is higher than with LBO investing, where default rates are lower, and overall the performance is stable.
Third, the performance of LBO managers is well documented. The financial leverage accounts for about a third of total performance. Operational value creation within the assets acquired represents over half. A buy-and-hold strategy on listed stocks does not achieve this. The acid test is provided thanks to the Public Market Equivalent method. If LBOs did not create specific value, the PME would show a limited and normative outperformance versus listed indexes. In practice, LBO funds outperform by 500 to 800 basis points, far beyond the contribution of financial leverage.
Cyril Demaria is affiliate professor at EDHEC and an independent consultant. He is the co-author of Asset Allocation and Private Markets (Wiley, 2021) and author of Introduction to Private Equity, Debt and Real Assets (Wiley, 3rd ed., 2020)