The Yale endowment model 
of investing is not dead

Much of what is discussed here is true, but I would add a few things. First of all, MPT as usually presented does rely on those historical correlations and returns. Adding in the assumptions above (and quite rightly) changes the typical assumptions of MPT a great deal. I also question whether VAR is a very useful metric at all, but we’ll skip that for now.

More interestingly is the claim that Yale changed allocations based on extremes in valuation. That seems to be the crux of the matter. Yale did not. The alternative strategies emphasized had always in the past taken advantage of valuation issues, or hedged them out. However, when in 2006-7 when valuations were so uniformly high across all asset classes the high correlation was not due to a crisis, it was due to them all being over valued. We didn’t see that in the tech bust because assets showed a tremendous spread in valuations across asset classes. The dispersion in returns was predictable. In the last downturn the high correlations were similarly predictable, since assets, whether alternative or not, were all trading at extreme levels. Thus the general decline in them was to be expected, not due to the crisis. Even hedged strategies in fixed income (relative value and convertible arbitrage for example) were dealing with such tiny spreads that to achieve any reasonable return required lots of leverage. Any volatility was bound to lead to crisis. Thus the correlations going to one was not caused by the crisis, the overvaluation and thus extreme correlation caused the crisis.

The answer, Yale should have reduced its exposure to hedged strategies that were in areas that were extremely vulnerable to a general downturn based on valuation. The same for many of their illiquid investments in real assets that were not the undervalued assets that had been so profitable for them for so long. Based on MPT Yale seemed to have assumed there was an illiquidity premium. Unfortunately, by 2006-7 there was none (or little) nor was there an equity premium amongst a host of asset classes (at least over cash.) When that situation exists the asset allocation needs to change. Period. Some strategies are not dependent on dispersion of likely returns amongst asset classes, or risk premiums. Those, very select long short strategies and high quality fixed income and cash were far more responsible as of 2006.

MPT may not be dead, but naive assumptions about efficient markets and risk premiums that are assumed to exist should be.

You can’t really separate MPT from the other aspects of managing an endowment. There are issues of politics, earmarks and fund raising. New donations are as important to the endowment as investment yield is and so is fighting over how to spend the money.

Every time a contract negotiation comes up, the labor unions point to Yale’s big endowment and its market performance to counter the claim that Yale can’t afford to pay employees more or offer even better benefits than they already do. Yale has always responded with the argument that the endowment is full of donations earmarked for some specific purpose. As far as I am aware, management has never made the argument that they are spending as much of the endowment as their risk model will allow. Maybe they should but if they did, it would open the risk model to public debate. Maybe this would be a good thing, maybe not.

As I mentioned earlier, fund raising is very important. It’s the nature of the business that people donate more when you give them a specific project and goal. This gives you more earmarks but that’s okay because, the next time around, you can say that Yale can’t use the general endowment for project xyz because of all the earmarks. “We, therefore, need your donations in order to get project xyz done.” This all means that the endowment is illiquid anyway and you might as well invest it in illiquid assets. If you want to debunk this argument, the unions would love it.

Wonderful article. As a private investor, I used MPT to navigate the crash and was helped enormously by the very classes that the endowments left out. My own story is here:
http://www.scribd.com/full/31174210?access_key=key-bj6f11a6kt0s5xlo6xq