Risk, Returns, and Hot Breakfast

The Harvard Management Company has come under increasing scrutiny in the wake of its investment performance in fiscal year 2015. Newspapers around the country and here at Harvard have noted HMC’s returns this past year lagged significantly behind its peers, citing the superior performance of endowments at Bowdoin, the University of Virginia, MIT, Stanford, Princeton, and Yale. Most recently, it has even been argued in The Crimson that the returns on Harvard’s endowment scandalously demonstrate that HMC’s professed prerogative to invest in any securities it sees fit — including fossil fuel companies — is illegitimate, and its continued refusal to divest from fossil fuels is baseless.

While it is true that Harvard’s returns have been outpaced by some of its competitors, calls for pay cuts and accusations of scandal on the basis of a single return metric demonstrate a fundamental misunderstanding of the ways in which investment managers should be evaluated. Cogent versions of such arguments would require substantially more data than is currently available to the public.

Before proceeding, remember that Harvard’s endowment is thirty times larger than Bowdoin’s and six times larger than the University of Virginia’s. Attractive investments of a given size, equally elusive for everyone, will do far more to boost the return of a small endowment than a large one. Comparisons of Harvard to these particular institutions are strained at best.

More broadly, a key fact to recognize is that maximizing returns is never the only objective for professional investors. Minimizing risk is often at least as important. Endowments are particularly concerned about limiting risk, since the university budgets they support, which fund financial aid, salaries, and hot breakfast, cannot countenance large and unpredictable swings.

Therefore, return metrics can only offer part of the story—they must be evaluated in the context of their accompanying risk. Simply put, there is not enough public data to make meaningful claims about Harvard’s performance vis-à-vis the schools commonly cited by the media.

In particular, the endowments of MIT, Stanford, and Princeton publish little more than a summary of their returns. Even attempts to make comparisons on a risk-adjusted basis are beyond our reach.

This leaves Yale. Harvard’s performance was nearly doubled by its perennial competitor, which returned a stunning 11.5 percent in fiscal year 2015. However, this comparison has its own problems. Though data is sparse, one can find that Yale’s target allocation for 2015 of 31 percent in private equity and venture capital, asset classes that have done extraordinarily well this past year, dwarfed Harvard’s 18 percent. Notwithstanding the fact that the valuation of these highly illiquid assets is difficult and subjective, such investments do offer attractive long-term risk-adjusted returns. That said, their illiquidity means that attempts to sell them in short time-frames result in transactions at precipitous discounts to fair value. Both Yale and Harvard learned this the hard way as they attempted to do just that in 2008-2009; operating budgets at both schools were cut in large part due to poor performance in their private equity holdings.

While Yale’s return this year looks attractive next to Harvard’s, we must remember this number exists in the context of divergent risk appetites. Perhaps Yale is just willing to take the chance that in the next bad year, it may be forced to cut the university’s operating budget significantly, in exchange for more attractive long-term returns; its allocation to private equity might make perfect sense in this case. Harvard, on the other hand, may be content with a lower average return, knowing that during crises, it might avoid similarly deep cuts to university spending. Though private equity allocations are merely a single dimension of incredibly complex investment programs, it should be clear that in general, without knowing far more about the risk-taking of the investment managers in question, it is impossible to make meaningful comparisons.

To be clear, I do not intend to make a directional claim on HMC’s performance here, nor do I intend to pass judgment on the various calls for change at HMC. However, it is certain that the return metrics widely cited in the media are incommensurable without a wealth of additional context, and that one ought not to read too deeply into the numbers that have thus far been disseminated.

Marshall Zhang ’16, a joint concentrator in Statistics and Mathematics, lives in Mather House.