The Yale Myth: Analyzing the Poor Performance of Endowments

The investment performance of Yale’s and Harvard’s endowments has drawn interest from other endowments and investors alike, but little research has been done on non-profit endowment performance overall, despite the $700 billion in assets they collectively manage.

One recent study,  “Investment Returns and Distribution Policies of Non-Profit Endowment Funds,” released by Georgetown University and the Stern School of Business, shed some light on endowment performance.  This paper, by professors Sandeep Dahiya and David Yermack, used IRS filing from 2009-2016 to look at the investment performance of more than 28,000 organizations, and the authors made some surprising conclusions.

• In general, they found, endowments averaged a return of only 3.75% over the period, badly underperforming a 60/40 equity/bond split portfolio by 5.53%.  They were even handily beaten by Treasury bonds, which returned an average of 4.89% over that period.
• Using the Fama-French-Carhart four factor model (risk, size, value and momentum), the authors found a statistically significant (to below 1%) alpha of -1.01%.
• Despite the reputations of Yale’s and Harvard’s endowments, higher education endowments performed significantly worse than other endowments, with alphas of -1.89% and -.93% respectively.
• Smaller endowments perform better than larger endowments, contrary to expectation.  It is often assumed that larger endowments have access to better managers and can negotiate lower fees.
• The top 20 national universities, as ranked by U.S. News and World Report, performed better, with abnormal returns of zero percent.  While this is much better than the negative alphas of the endowment group as a whole, a zero percent alpha is the abnormal return expected simply by chance. 

The authors surmise that while both stocks and bonds experienced a bull market for most of the study period, endowments may have been sitting on the sidelines, with large amounts of their assets in cash or equivalents.  They also found an interesting performance pattern.  Large endowments, which make up the majority of investment assets, had poor performance if they were located closer to a major financial center, while smaller endowments tended to perform better the closer they were to a major financial center.  It is not clear why this would be the case, but one possibility is that large funds are likely to attract, and possibly over-invest in, money managers pitching expensive, illiquid products.

Some of these finding are backed up by an earlier study from Vanguard.  Their September 2014 study, “Assessing Endowment Performance: The Enduring Role of Low-Cost Investing” noted that in the 25 years prior to 2014, the investment strategies of many endowments changed dramatically.  A balanced 60%/40% stock/bond split was the traditional investment norm, but over the 25 year period, investments in alternatives such as hedge funds and private equity increased dramatically, possibly in response to Yale’s success.  By 2013, the largest endowment portfolios had 60% of their assets tied up in such investments.  Vanguard’s study looked at whether or not this strategy shift had paid off.

The answer, unsurprisingly, was no.  The table below, from their study, shows the 5 – 25 year returns of all endowments in the study, compared to a 60% stock / 40% bond benchmark.

	5 years	10 years	15 years	20 years	25 years
Yale University	3.3%	11.0%	11.8%	13.5%	13.2%
Harvard University	1.7	9.4	9.6	11.9	11.5
All endowments	3.8	6.8	5.6	7.7	8.4
All active balanced mutual funds	5.1	6.0	4.9	7.0	7.9
60% stock/40% bond benchmark	5.9	7.4	5.7	7.6	8.3

This table also shows the long term performance of Yale’s and Harvard’s endowment compared to the rest of the group.  Their past outperformance is clear and is what led to the “Yale Model” that other funds and investors have tried to replicate, however, in the most recent 5 year period, both of those endowments have performed poorly, returning less, on average, than endowments as a group, and coming in far below the returns of a simple 60% stock / 40% bond portfolio.

Vanguard found that endowments had, over time, significantly increased the portfolio percentages allocated to alternative investments.  In the 10 years up until June 2013, large endowments increased their alternative allocation from, on average, 31% to 59%.  Medium sized endowments went from 16% to 36%, while small endowments went from 5% to 18%.  But there is no evidence that this pivot towards alternative investments has increased returns.

There are a number of possible reasons for this.  One explanation is that the performance of hedge funds overall has decreased over the last couple of decades.  Vanguard found that the excess returns realized by large endowments mostly occurred in the early to mid 2000s, before many smaller endowments began moving into the alternatives space.  But aside from timing issues, there is the problem of active management in general.  While there are clearly successful outliers, alternative investments as a group do not outperform public market benchmarks.  Generating outsized returns from alternatives, then, requires finding and accessing top managers.  And, of course, it requires those top managers to continue their out-performance over extended periods of time, something that most studies show simply does not happen.   It is also possible that expenses dragged down performance for small endowments especially.  The thinking here is that large funds should be able to negotiate better pricing based on the amount of money they are able to invest.  Smaller funds with less to invest could be paying considerably more.

There is also the question of all around deteriorating performance over the last 5-15 years ending in 2013.  10 years prior to 2013, the performance gap between a balanced portfolio and endowment returns was less than 1%.   In the 5 years prior to 2013, during a period of steady increases in the percentage of portfolios allocated to alternatives, that gap has increased to more than 3%.  Since the study periods here overlap with Dahiya and Yermack, the possibility of endowments holding large cash positions needs to be considered as a reason for this underperformance, but it is also possible that the increase in alternatives was at least partly responsible.

Whatever the reason, it seems clear that endowments have investing in an unsuccessful strategy.  With so many low cost, liquid investment vehicles available, endowments would be better off focusing on factors and asset classes they want exposure to, and utilizing those vehicles to build inexpensive, transparent portfolios to protect and grow their funds.