Figuratively speaking, large foundations and pension funds can be compared to tankers on the high seas: they have a clear objective, are on a direct, previously calculated course and are usually heavily loaded - in other words, almost fully invested. This means that in the event of turbulence, the course taken cannot be changed quickly. Instead, they rely on their investment process, due diligence and the diversification effect. This allows them to weather even severe storms on the markets, as previous experience has shown, before returning to calm waters.
However, the forces of nature on the markets are quite different from those on the high seas - and far more difficult to calculate. Just how challenging the environment really is was illustrated by a paper recently published by Richard M. Ennis entitled "Endowment Performance", which documents recent underperformance by foundations. 
The investment style of foundations
According to the study, there are recurring elements that are common to various foundations in their investment style. In addition to a general tendency towards value investing, there are the following three points in particular: 
The investment style of foundations:
Active management: The (assumed) ability of investment managers of foundations to identify and exploit special expertise in target funds. According to the paper, large foundations usually rely on a large number of different target investments, of which passive products account for only about 6 percent on average. The willingness to pay for the perceived skills of external managers is particularly evident in the considerable proportion of cost-intensive investments, for example in hedge funds.
Focus on equities: The view that equities in particular should determine the allocation is widespread among investment managers of foundations. However, the proportion of shares varies depending on the size of the overall portfolio: large foundations have an average effective equity exposure of 72 percent, while small foundations have 63 percent.
Private Equity / Debt / Real Estate: Private markets are another focus of many investment managers of foundations and their advisors. It is often assumed that these markets offer good diversification and that experienced managers can benefit more from them. In addition to hedge funds, this is the alternative component of the portfolio.
This roughly defined investment style of foundations has achieved very good results in the 1990s and at the beginning of this century. Market participants' perception of it was high, so that this style has met - and in some cases still meets with - sustained interest among investors worldwide.
Let us return to the underperformance mentioned at the beginning. In his study, Richard Ennis looks at 43 large foundations in the period from June 2008 to June 2019, which manage more than USD 1 billion. He examines these foundations in comparison to individually defined benchmarks. As a result of his regressions, he obtains alphas that range between -3.56 and + 2.07 percent. While none of the positive alphas are statistically a significant positive, about one in four negative alphas are significant. The author concludes from this that large foundations clearly underperformed their benchmarks in the period under review. 
This result is also confirmed in the long-term study by Dennis Hammond, which analyzes the entire available data period on the performance of university foundations of 58 years.  According to the study, foundations achieved an average annual return of 8.5 percent, which is worse than the traditional 60/40 benchmark of stocks and bonds, which achieved 9.3 percent per year. The large foundations still performed best overall. This could be due to the fact that - unlike small foundations - they have better or even exclusive access to the really profitable target investments and/or can enforce lower fees.
Other points of view
However, there are also other voices. This is mainly due to the fact that the assessment of under- and outperformance can depend on both the benchmark and the time horizon. Hossein Kazemi of the University of Massachusetts argues that an appropriate benchmark is not always used. Instead of a US benchmark, which has been difficult to beat anyway due to the high returns over the past 20 years, a benchmark must be used that also includes international equities - after all, these are usually an essential component of portfolios. In addition, cash holdings are often not taken into account. Although this is small in most cases, it does have a systematic influence. Taking these aspects into account, the average returns of foundations since 2000 - including the eight good years until 2008 - have been above the benchmark, according to Kazemi. He also points out that classic alternative investments, which are increasingly used by foundations, have performed well overall over the past 20 years. 
As an explanation for the observed underperformance, however, the study by Richard Ennis uses precisely these alternative investments - i.e. hedge funds and investments in the private markets already mentioned. This is initially in contradiction to the advantage often described by representatives of alternative investments, that they would offer better risk-adjusted returns and diversification potential.
But the argument carries weight. Another study by the author examines which asset classes actually contribute to diversification. 4] For this purpose, these are added step by step as explanatory variables to a regression model and the corresponding statistical variables - specifically the coefficient of determination and the standard error of the regression - are calculated in parallel. This is used to check how well the returns are explained in the model.
The results illustrate the paradigm shift. In the period from July 1999 to June 2008, alternative investments had a significant diversification effect: by adding them to a portfolio of stocks and bonds, for example, the coefficient of determination increased from 0.66 to 0.97. But since July 2008, the diversification effect has been almost negligible. Here, the coefficient of determination was 0.99 if global equities and bonds are included, which means that the returns of the foundations could be almost completely explained by these two asset classes alone. According to the author, the same pattern is also found in the returns of public pension funds during this period. 
These are amazing results. Although the portfolios of many foundations have a significant proportion of alternative investments, returns can be explained by only two variables, global stocks and bonds. Specifically, the selection of large university foundations examined by the author during the period under study can best be described by a benchmark of three indices with the following weightings: 53 percent Russell 3000, 17 percent MSCI ACWI ex US and 30 percent Bloomberg Barclays Global Aggregate Bond.