What were the arguments for rebuilding HMC’s internal capabilities? What sets us apart cannot be outsourced — our triple-A balance sheet, the permanency of our capital. It costs us a third to half as much to manage a dollar internally as to manage it externally, while with internal management we’ll have complete transparency of all our positions — a benefit in a world that’s become incredibly fluid. That setup is so obvious, and yet so difficult to maintain in an academic context.
Was there any talk of simply outsourcing everything? Absolutely. It would have been a simple option … But it would be the wrong thing to do. Combining internal and external management enables you to run much more sophisticated risk management operations, and I think, as the world is proving, there will be times when we’ll need internal capabilities, not to play offense, but to play defense. That’s how we had such a good July, despite having one of our spinoffs (Boston-based hedge fund Sowood Capital Management) fail.
Did past controversies over HMC compensation packages make it tougher to rebuild? There was a real question of whether we’d be able to hire the people we wanted, but in the end, we managed to hire what I regard as the best. … Let me tell you how I started every interview with (prospective) portfolio managers: “We’re going to pay you a fraction of what you’re getting now. We have no equity to give out, which means you’ll never be able to realize the enterprise value of what you’ve done, and in addition, once a year, during the third week of December, we’re going to humiliate you by issuing a list of the five highest paid (managers). … Do you still want to continue this discussion?” People gave up compensation for two reasons. One is they believe in the mission of Harvard, and the other is we try to create pure investment jobs here: no client work, no marketing.
So you’ve managed to square the circle? The compensation model that we have is very powerful. The issue is not with design, the issue is with implementation. What happened earlier is that inadvertently, as Harvard Management experienced one spinoff after another, more and more of the capital was concentrated in one activity, and that activity was so successful, that it entailed very large compensation. We’re now building — for all sorts of sensible reasons — a more diversified return-generating engine, and I think that our compensation system is going to be very powerful in this context.
You’ve suggested the benefits of diversification have been getting diluted, as more institutional investors follow the “endowment” model. It’s getting very crowded, not only in terms of asset allocations, but in terms of finding the right implementation vehicles. There’s a limit to how much superior investment expertise is out there. So the asset allocation is going to be less potent because there are more people doing it. And then the global liquidity situation is changing as well. So our view is that performance in future needs something more — two things more: first, better risk management, because correlated risk has become a big issue, and diversified asset allocation no longer gives you the risk mitigating characteristics it used to. Second, is identifying new secular themes that will play out over the next five years, and trying to be a first mover in those, and that’s what we’re working very hard at doing.
What does better risk management involve? We have three different (tools.) One is beta-adjusted asset allocation, to make sure we’re taking the risk that we want to take; second, we use overlay strategies; and third, and perhaps least intuitive for an endowment, is we started a year ago implementing a fat tail insurance program.
We chart our asset allocation over time, see how our market value exposures are developing. We found the sensitivity of those exposures to certain factors changes dramatically over time, so we’ve modeled our managers to look at what we call the concurrent beta — how exposed are they to the market. We look at our beta-adjusted market exposures, and what you find looking back, typically, ahead of major market selloffs, more and more people get sucked into the trade, and more and more of our external managers, their beta to the markets keeps on going up. So what we do is take offsetting measures at the level of the endowment as a whole. In January, we noticed, when we beta adjusted, that our managers had gotten sucked into the equity rally. So we, as a house, sold S&P, EAFE and (iShares for emerging market equities) to reduce our market value exposure. Little did we know that on Feb. 27 you were going to have this major correction. If we had known, we would have sold a lot more.
We weren’t doing it for market timing, we were doing it for risk management. We had too much risk on, given what we wanted to have, because our managers had become more and more aggressive. Needless to say, the minute the market sold off, many of them started hedging their positions, and we bought back our exposures.
Again, at the beginning of July, we just couldn’t believe how much risk was on the books of various people, so we took down our exposures, and bought credit protection, and because of that, when the selloff came, we were largely unaffected. That helped us tremendously during Sowood. Even though we sustained a fall in our Sowood investment equal to about 1% of the endowment, and even though the S&P 500 fell 3.1% that month, the endowment ended up in positive territory.
Are other big endowments doing likewise? To do it, you need some very talented internal portfolio managers, because this you cannot do with phone calls. There’s one view that says you don’t have to do it, because if you hire smart enough people, they’ll do it for you. We’d rather do it in-house, because we can respond more quickly. …
Has anything in the recent market turmoil surprised you? What caught me by surprise is the fragility of some of the hedge fund structures, and the fact that the banking system is no longer the shock absorber that it was in the past. When clients, particularly leveraged clients, start having issues, banks — with their leveraged buyout commitments, their prime brokerages and proprietary desks — can no longer be the provider of balance sheet, a new reality that people have to factor in. It means when the downturns occur, they’ll be much more violent than before. …
Where does that leave your fellow institutional investors? The question is, do (institutional investors pursuing endowment-like allocation strategies now) have the risk management capabilities to support ... alternatives. Because when an accident happens there, it’s big, and you have to absorb the blow and move on.