The Futility of the Endowment Model
In the past two decades, the so-called endowment model has been adopted by hundreds of endowments, foundations and advisors – particularly those serving ultra-high-net-worth clients. By aggressively allocating to illiquid alternative asset classes, those investors hoped to duplicate the results of Yale and other top-tier institutions.
New research exposes the futility of those efforts.
The average university endowment does not earn any risk-adjusted return (alpha) and would have done just as well investing in a 60/40 stock/bond portfolio, according to a new paper, “Do (Some) University Endowments Earn Alpha?”, by Brad Barber and Guojun Wang, professors at the University of California at Davis.
Some elite institutions – like Yale – may appear to earn alpha, according to Barber and Wang, but it does not come from selecting skillful managers or from dynamically changing asset allocations. Instead, the apparent alpha comes from the mere fact that they hold assets like private equity and hedge funds.
But those private equity and hedge fund assets may not really be earning alpha. Barber and Wang cite evidence that private equity doesn’t deliver alpha, and only meager evidence that hedge funds deliver alpha.
Another pair of studies, “How Institutional Investors Form and Ignore Their Own Expectations” and “Picking winners? Investment Consultants' Recommendations of Fund Managers,” by a group of researchers from the University of Oxford, further illustrate the imperfections of the manner in which many endowments are managed.
The first Oxford study shows that institutional investors largely ignore their own expectations about the expected performance of the managers they employ and instead rely heavily on the recommendations of consultants. Those funds recommended by consultants, however, do not significantly outperform other products, according to the second study.
In other words, the consultants upon whom endowments, foundations and many advisors rely add no value.
I’ll look at those studies and conclude with the implications for financial advisors.
Universities’ fruitless search for alpha
Barber and Wang studied annual return data from 279 endowments over the 21-year period beginning in 1991. They found that 94% of the volatility of those returns was explained by a two-factor stock-bond benchmark model (i.e., the returns had an r-squared of 0.94). When they added an international equity benchmark to the model, it explained 99% of the volatility. In both cases, the alpha of the average endowment was statistically indistinguishable from zero.
“We find no evidence that the average endowment is able to deliver alpha relative to public stock/bond benchmarks,” they wrote.
Naturally, there were some endowments that performed better and worse than average. Barber and Wang found persistence among those groups: strong performers in one year tended to do better in the next, and poor performers were more likely to underperform in subsequent years.
They also studied two groups of elite institutions: the eight Ivy League schools and the 30 non-Ivy schools with the top SAT scores (based on the assumption that SAT scores are an objective measure of a school’s “status”). Those two groups delivered positive alphas (3.15% and 3.82%; 1.74% and 2.28%) relative to the two- and three-factor model portfolios.
Next, they added two more benchmarks to their model, for hedge funds and private equity. Using this five-factor model, the alpha disappeared from the Ivy and top-30 groups. Barber and Wang concluded that the performance of those institutions comes from their allocations to hedge funds and private equity, not from manager selection or tactical asset allocation.
Unquestionably, the strong performance of elite institutions was attributable to outperformance among hedge fund and private equity funds over the time period they studied.
“The intriguing evidence of superior returns among the top performing and elite institutions is completely explained by their strategic asset allocation decisions,” Barber and Wang wrote.
That finding, the authors admit, rests on the assumption that neither the hedge fund nor private equity benchmarks generate alpha. Barber and Wang did not assert that that assumption is valid, but they cited strong evidence that private equity benchmarks did not generate alpha, and nearly-as-strong evidence for hedge funds.
But their key finding – that elite institutions did not earn alpha – ultimate rests on whether those two asset classes truly delivered risk-adjusted return. That question has been the subject of debate in academic circles and remains unanswered.
Do consultants add value?
Many of the endowments in Barber and Wang’s sample rely on consultants to guide their asset-allocation and manager-selection decisions, as do many foundations, institutions, family offices and advisory practices. Let’s turn to the question of whether those consultants helped to overcome the performance challenges they identified.
The two Oxford studies looked at the role consultants play in helping institutions select managers. Their data reflected assets managed by endowments, public and private pension funds and foundations; 94% of plan sponsors managing those assets use a consulting firm. Among the biggest names, according to a New York Timesarticle, are Mercer Investment Consulting, Russell Investments, Towers Watson Investment, Cambridge Associates, Hewitt EnnisKnupp, R. V. Kuhns & Associates, Callan Associates, Pension Consulting Alliance, Strategic Investment Solutions and Wilshire Associates.
The authors looked at surveys of consultants conducted by Greenwich Associates between 1999 and 2011 and at the investment performance data of the funds recommended by consultants, as reported by eVestment.
In the first study, they found that plan sponsors used the past performance of funds to project future performance and that perceived past performance is a driver of flows into funds. But consultants’ recommendations were also responsible for flows into funds. Plan sponsors’ expectations of future performance were also influence by so-called “soft factors,” such as their perception of managers’ styles and processes.
But none of those factors – past performance, expected future performance or soft factors – were predictive of future performance.
The second study looked directly at fund recommendations provided by the consultants for actively managed U.S. equity funds.
It showed that consultants do not chase performance when selecting funds; instead, they rely on a number of soft factors in forming their recommendations. Consultants’ recommendations have a sizeable influence over flows of funds, according to the study. Asset flow is correlated with the number of consultants recommending a particular fund.
Most importantly, they found that funds recommended by consultants, on average, did not exhibit any significant outperformance on a risk-adjusted basis, relative to those funds that were not recommended.
The authors found a negative impact of fund size on performance. Much of the reason why funds recommended by consultants underperformed was explained by the fact that those were the larger funds, according to the authors. They hypothesized that consultants may be forced to recommend larger funds because of capacity constraints of smaller ones, even if the larger funds are not as high-performing.
“The analysis finds no evidence that the recommendations of the investment consultant for these U.S. equity products enabled investors to outperform their benchmarks or generate alpha,” they wrote.
Indeed, it could be that the Barber and Wang’s finding that manager selection did not generate alpha for the average university was due to the failures of consultants to recommended good managers.
The study was based on gross data, and the authors wrote that results would be even worse if they considered fees, which average 50 basis points on the products they studied – plus whatever the consultants are charging.
“In light of the evidence presented, what is striking is that fund sponsors follow such recommendations to the extent and at the expense that they do,” they wrote.
The authors offered three explanations for this. First, investors may value the “hand-holding” service offered by consultants. Second, consultants may provide a “shield” in case their selections are challenged in court or elsewhere.
Most disturbing, however, is the third possible explanation: investors are not aware of the inaccuracy of consultants’ recommendations. Unlike managers, who provide fully transparent data for their fund’s performance, consultants do not report the results of their recommendations. The authors used aggregate data for their study, so it is possible that some consultants are better than others.
Implications for advisors
An obvious implication of the Oxford studies is that anyone using – or contemplating using – a consultant should insist on a fully transparent record of their recommendations and an analysis of whether those recommendations resulted in risk-adjusted outperformance.
But the larger question is whether advisors embracing the endowment model are acting in their clients’ best interests, or whether they will fail to produce alpha by allocating to alternative asset classes, like hedge funds and private equity.
I spoke last week with Barber, who has been in the position of advising universities on how to manage their endowments. His recommendation, typical of academics, is to rely on a passive, index-based approach. Reducing the cost of investments leaves more for the investor.
“The vast majority of endowments choose to play the loser’s game, with mixed results,” Barber and Wang wrote in their paper. “The average endowment allocates 73% of its domestic public equity portfolio and 66% of fixed income assets to active management – markets in which it is notoriously difficult to beat public indexes.”
For advisors, Barber recommended starting with a simple asset allocation using index funds with low fees and avoiding private equity, hedge funds and active management. He said advisors should focus on estate planning, insurance and “making sure clients meet their life goals.”
“I see so many people trying to do the investment-picking thing in the advisor arena,” he said, “and that’s not where the value is added. It is hard to argue for advisors using actively managed public-equity funds.”
Barber said that an argument could be made for endowments to have a small allocation to private equity, because of their long investment horizons and ability to bear liquidity risk. But that should be a “small slice,” he said, for two reasons. First, investors should try to replicate the broad market portfolio, and private equity is relatively small compared to public stocks. Second, substantial capital has flowed into private equity over the last two decades, and there is little reason to believe the strong returns over that period – whether or not they represent alpha – will continue in the future.
He said the case is weaker for hedge funds, which represent a “side bet on alpha for which house takes a big cut.”
Barber’s recommendations and the findings of these studies will not be surprising to readers of this publication, who have been warned before, by academics and others, about the inapplicability of the endowment model to client portfolios. But the key findings – that the average endowment fails to deliver alpha and that the average consultant does not add value – reinforce the message that the odds against successfully adopting the endowment model are very steep.
These findings also have public-policy implications. Many of the investors in these studies are public institutions, such as the endowments of state schools. Those institutions are placing risky bets on private equity and similar funds, to some degree at the taxpayer’s expense. They are relying on consultants who, on average, add no value and no prospect of delivering alpha. These studies document the futility of those efforts. Taxpayers would be better served if public institutions adopted a low-cost index-based approach to endowment management.