The Endowment Effect at MIT
William J. Hartnett
This article is the first of two parts, the second of which will run in the June 5th Commencement issue.
What is the purpose of the endowment? Most people answer: to smooth out fluctuations in the budget and be a source of funding. However, investment swings in the MIT endowment over the past five years have dwarfed distributions (see Figure 1). Furthermore, these investment swings have been used to justify painful cuts in budget and staff, so that the endowment has been a cause of fluctuations in the budget instead of smoothing them out.
During the current process of renewal at the Institute, we need transparency while reformulating the endowment strategy -- defined as the development, investment and distribution strategies taken together. Transparency promotes support from current members of the Institute community and future donors as well.
In that spirit, I suggest that we have been overexposed to the volatile stock market (see Figure 2). Coupled with a distribution strategy which transmits some of those fluctuations of the market to the budget, this has led to problems despite significantly outperforming the market over time. In other words, the problem has been excess volatility, rather than cumulative returns.
If combinations of more return and less volatility “dominate”, then over five years MIT endowment performance has been ex-post dominated by a simple bond portfolio (see Figure 3). While it has also been dominated by select peers, they are not as useful as a reference.
Might we benefit from reconceptualizing our endowment strategy as aggressively as other aspects of the Institute have been scrutinized? Can we really afford to lose in the stock market amounts that we can’t afford to distribute to our budget? I suggest we establish a target for portfolio volatility which is only modestly above low risk bonds, coupled with a distribution strategy which prioritizes smoothing out the budget.
Big gains in the stock market year-to-date provide a golden opportunity to take the money off the table. “Speculation” could be used to refer to investments with so much volatility as to affect consumption standards in the relatively near term. From my point of view, many donors could be disinclined to pay for (or see their contributions exposed to) speculative losses. “Gambling” might be used to describe speculation with a negative expected return, which might well apply to the post-election stock market.
In finance, the “endowment effect” means having a reference, and that relative to that reference losses are felt more strongly than gains. Our endowment reference should be a conservative investment strategy, coupled with a distribution strategy whose first priority is to avoid cutbacks in programmed budget outlays. I believe this could be an important first component of making the endowment better support MIT’s mission of socially beneficial research and education.
William J. Hartnett PhD ’96 is a former Visiting Scholar at the Sloan School Of Management’s Fin-ance Group.