Updating the Yale Model

The so-called “endowment model” of investing – often called the “Yale Model” because it was pioneered by David Swensen at Yale – remains the most prominent institutional investment approach out there. That approach may still be working for Yale (although some would disagree), but in the aggregate it is not.

Data from 831 U.S. college and university endowments and affiliated foundations, representing over $400 billion in endowment assets, shows that the average return for U.S. college endowments was minus-0.3 percent in the 12 months ended in June 2012 (the most recent period for which data is available).   Longer-term returns are not a lot better.  College and university endowments returned an average of only 1.1 percent annually over the past five fiscal years and 6.2 percent over the past decade, net of fees. There is no way to sugarcoat those numbers.  They aren’t very good.  Moreover, Swensen himself concedes that the “average endowment” runs a 28 percent likelihood of losing fully half of its assets (in real terms) over the next 50 years based on its asset allocations and a 35 percent risk of a “spending disruption” over the next five years.

It should not be surprising, then, that The Portfolio Whiteboard Project brought together a group of next generation institutional investment leaders and asset managers to update the Yale Model by considering afresh what an asset allocation model for the future might look like. The end result (memorialized via a noteworthy “white paper“) was significant and insightful in unanticipated ways, suggesting that institutions do need to go back to the drawing board, but not just for the reasons initially thought.

From the outset, the group set out to answer one intriguing question: “How would you allocate your portfolio if you could start from just a blank whiteboard?”  The folks who posed the question thought they were positing an asset allocation challenge.  But the group quickly decided that they needed to back up a bit before tackling asset allocation.  They suggested a full re-examination of the very framework of their investment approaches first, from governance to execution. They wanted to consider their systems and their processes. Significantly, they realized that institutions which had copied Yale’s asset allocation often failed to learn from Yale’s investment process.

As a starting point, the group valued and desired the freedom and flexibility to do what was needed to accomplish their investment goals. What works today may not work tomorrow. Past performance is not indicative of future results. “Governance establishes the process and puts the policies and procedures in place to ensure investment officers can execute.” In other words, the investment team must be empowered to find what is likely to work and then to go out and do it. Without clear support from “above,” there is little chance that that sort of freedom to step “out of the pack” will be exercised. Being wrong with everyone else has astonishingly little cost. Being wrong alone, even temporarily, often means career suicide.

With respect to asset allocation, the group decided that the key is to “choose objectives first and [only] then select assets based on their underlying risk premia.” Thus asset classes that have traditionally been viewed as distinct and separate – such as high yield bonds and equities – might be grouped together due to similar performance characteristics. Moreover, there should be specific budgets and allocations for liquidity needs and to deal with inflation as well as for capital growth.

It’s a good idea to invest based upon objectives, of course, but executing that idea isn’t easy.  “Draw down tolerance must be established, stress testing must be instituted and cross-correlations must be studied. This translates into a bigger investment in systems and technology and more carefully defined processes and procedures, probably adding more personnel.” Even so, investment realities don’t always cooperate with intentions. As one project participant put it, “‘The more of us that buy illiquid investments, the smaller the premium becomes.’ If you’re there first, you benefit from the piling in from peers.” Moreover, “[h]edge funds have been largely disappointing in recent years; they are pricey and even when they do well, they might not do well enough to justify the time spent vetting them” (much less to justify their cost).

The group’s fundamental conclusion is that investment process is paramount. As one participant summed up: “So it is the journey, not the destination.” This idea is a great one, but the analysis didn’t go nearly far enough.  Fixing the investment process alone won’t do the trick. The problems are more foundational than that.

Research on the performance of institutional portfolios (not just endowment portfolios) shows that after risk adjustment, 24 percent of funds fall significantly short of their chosen market benchmark and have negative alpha, 75 percent of funds roughly match the market and have zero alpha, and well under 1 percent achieve superior results after costs—a number not statistically significantly different from zero. Mutual funds aren’t any better. In a random 12-month period, about 60 percent of mutual fund managers underperform. Lengthen the period to 10 and 20 years and the proportions of managers who underperform rises to about 70 percent and 80 percent, respectively.  Perhaps even more importantly, money managers who underperform do so by roughly twice as much as the “outperforming” funds beat their chosen benchmarks.

This lack of performance is not the only reason to question the foundations of the institutional investment process. Data from over 35 years of research on individual managers at institutional funds show that new accounts tend to go to managers who have produced superior recent results—mostly after their best-performance years—and away from underperforming managers after their worst-performance years. Pick your favored metaphor – they are fighting the last war and investing while looking in the rear-view mirror. Moreover, another oft-repeated problem is moving into asset classes after prices have risen and out of asset classes after prices have fallen—moving assets in the wrong direction at the wrong time, which also burdens institutional investors with billions of dollars in costs.

Clearly, the investment process is broken at the institutional level. The evidence is remarkably clear that institutions would have been better off doing nothing – partly on account of slightly better returns, but mostly on account of costs, particularly the costs of active management itself (which is far more expensive than passive management) and the costs of the consultants that recommended those managers in the first place. These consultants concede (as they must) that past performance is not indicative of future results. But they still make recommendations almost entirely due to such past performance. Why should anyone be surprised, then, when this approach fails (at least from the perspective of the institution, not the consultant)?

This failure isn’t a matter of process so much as a failure of investment belief and philosophy. As consistently (and ironically) reiterated by Charley Ellis (long-time head of Yale’s Investment Committee) and David Swensen (long-time Yale Chief Investment Officer), unless you’re Yale – with its resources and advantages – you shouldn’t expect performance like Yale’s. What really matters are the investment beliefs and the overarching philosophy instituted so as to give investors a reasonable chance to succeed.

To pick the two most obvious examples: (a) passive management should be the default setting with active management used only to supplement it with very good reasons; and (b) low costs are crucial. While the data supporting these concepts is wide, deep and perhaps incontrovertible (and some institutional investors now seem to “get it”), they are not commonplace among the institutional investment world and received barely a mention in the otherwise outstanding white paper. 

Of course, since following these beliefs would likely require smaller staffs and many fewer consultants, it should hardly surprise us that the seemingly obvious gets neglected.  As newly minted Nobel Prize winner Robert Schiller sagely notes, “[m]oney management has been a profession involving a lot of fakery — people saying they can beat the market and they really can’t.” So here’s to better investment process, but a process supported by investment beliefs and an undergirding philosophy that is reality-based and actually supported by the data. Fakery not required.