This is a brief note on a recent book I read called The Endowment Model.1 The authors are eminently well-qualified in the endowment/foundation and institutional investment word, with Martin Leibowitz and Anthony Bova at Morgan Stanley and Brett Hammond (and formerly Martin Leibowitz) at TIAA-CREF. The book has won high praise from many esteemed members of the financial community.
The central theme is about improving returns and diversification for endowment-like funds by looking at asset allocation and portfolio risk relative to the US equity asset class, i.e., expressing risk and return relative to US equity beta. Relating the beta of asset classes to the beta of US equities in this way greatly simplifies the analysis of portfolios, provides better understanding of asset class alpha, and is consistent with the premise that the beta sensitivity to US equities “captures about 90% or more of the volatility risk of most allocations seen in the US institutional market.”2
The authors do not believe the endowment model is wrong, rather, its diversity has been overestimated – hence their suggestion that a beta-based approach recognizes this higher volatility. The authors point out: “investors should be leery of accepting the endowment model’s past periods of high returns as a simplistic template for the future. Many of the more notable early successes were achieved by organizations that enjoyed special advantages in staff and analytical resources, highly committed sponsors, flexible funding needs, extensive access networks, and perhaps most important – early entry.”
The size-effect is significant. The 1990-2008 NACUBO/TIAA-CREF Endowment Surveys, which report equally weighted returns of the largest endowments (>$1B in 2008) of 12.1% pa vs. 8.2% earned by the smaller endowments and 9.0% earned by a passive 60/40 asset mix (S&P500/Lehman Aggregate).3
The authors observe that: “US equities continue to act as the overwhelmingly dominant risk factor for most institutional portfolios.” This is the premise for using US equities (whose beta is set at “1”) as the reference point for the betas of other asset classes. The authors call these derived betas “structural betas”. They then calculate the “structural alphas” of asset classes. The structural alphas range from a low of cash to the highest asset classes (as you would expect) venture capital, private equity, and emerging markets.
The book discusses many topics, including:
- “Dragon Risk”: non-quantitative risks in non-standard asset classes,
- “Alpha Cores”: putting the starting point at assessing the maximum acceptable limits of non-traditional assets (the core assets) and then adding the liquid assets to achieve the desired risk level of the overall portfolio,
- “Bonds as the Risk Free Asset”: how this perspective affects structural alpha,
- “Active Alphas”: these are skill-based and zero sum, and can be further divided into 1 /portable, and 2/ bound ( those alphas for which a fund does not have access to efficient hedging vehicles),
- “Stress-betas”: the increasing betas generated during non-normal times when correlation tightening occurs.
Some key takeaways include:
- Using structural alphas and betas reveals that many institutional portfolios are far more alike in risk-return characteristics than they appear on the surface.
- The reliance on non-standard alternative assets appears to be more a return-enhancing than risk-control technique, although, in periods of stress beta, accumulated past excess returns may be the price of severe underperformance during the period of stress betas.
- The endowment model requires a long-term time horizon to be successful – the author’s example of the 1993 through 2007 followed by 2008 bears this out. The earlier period’s excess return compensated for the subsequent underperformance in 2008.
Overall, the value of the beta-based approach is its simplicity and intuitive appeal in understanding the roles of structural alpha and structural beta in the modern endowment portfolio.
1. A full version of this review is forthcoming in the Journal of Investment Management.
2. An early reference to this result is Leibowitz, Martin “The Beta-Plus Measure in Asset Allocation”, Journal of Portfolio Management, Spring 2004.
3. For a survey of literature on private equity performance, see my article“Living Dead” Canadian Investment Review, Fall, 2008. http://www.investmentreview.com/files/2009/12/livingdead1.pdf.