No one is surprised, but whose to blame?
That a once successful hedge fund lost US$6 billion over the course of a few days can’t come as a surprise to the limited partnership community. The local Amaranth fellows that I’ve come across are as smart as you’ll find in the business. They knew which risks they wanted to take, and were picky in terms of which deals they’d do from what I could tell.
One lesson the limited partnership community should take from their Amaranth experience is that many risk management tools are just that: tools, and not guarantees to prevent massive losses. If the risk model requires you to be able to exit the security to prevent further losses (a traditional “stop loss”), and you can’t actually exit, the model is utterly useless to manage your risk. In this case the WSJ reported that the fund managers couldn’t exit “economically” — I assume that is code for the market wouldn’t give them a decent bid. Sounds like they just didn’t like the price, and although they wanted to stem their losses it wasn’t going to be at all costs. In hindsight, you have to wonder if that wasn’t just another bet that the price of natural gas would turn in their favour.
The lesson for Limited Partners from all of this isn’t that they should require stricter controls on risk, or more consultants to analyze which hedge funds to commit funds to (there are plenty of those). The lesson is simple: be consistent across all of your investment categories.
In our world (same goes for merchant banks and venture funds), if we made a sufficient return in 2005 to warrant a “sharing of the upside” with our limited partners, but then had a loss in 2006, we’d have to return the upside we’d earned in the prior year…as it hadn’t actually been earned over the life of the fund. Makes sense, doesn’t it? If you don’t make money for your LPs over the life of a fund, there’s no upside for us to share in.
LPs go so far now as to require all individual participants in the General Partner to guarantee their own share of the promote, in case they take a payout this year and then leave the fund during a down year when the subsequent returns would nullify the prior year’s payout.
In the hedge fund world, if you make $1 billion profit this year and meet the hurdle rates, the GP might keep $200 million of that upside for the team. Good for everyone. But if you lost $1 billion the following year, the members of the GP still get to keep that $200 million. Or at least the summer home in France that you bought with your share of the two hundo.
Herein lies the problem. Conservative pension funds, seeking higher alpha, have put over a trillion dollars into the hedge fund asset class. Unintentionally, they payout system incents hedgies to swing for the fences. If you connect one year, as Brian Hunter is said to have done in 2005, you might clear US$80 – US$100 million personally. Yet unlike every other fund type his LPs have invested in (infrastructure, buyout, venture, mezz and sub debt), he keeps that money and never has to work again. Nor do his kids or their kids.
Institutional LPs need to think hard about the cause and effect of the annual non-recourse hedge payout system. It’s not that Amaranth’s risk management system failed. It’s that by condoning the payout structures, their LPs unintentionally encouraged them to take the very risks that cost them two thirds of their capital in six short weeks.