Jeffrey Larson's Sowood Capital: A Lesson In Risk Mis-Management

The weekend Wall Street Journal carried an amazing story concerning Jeffrey Larson, once a member of the Harvard University endowment's management team, now ex-partner of his failed hedge fund group, Sowood Capital Management LP.

As an asset manager with experience managing both my own and other people's money, and having clear memories of market shocks all the way back to 1987, I am a firm believer in the weakness of so-called 'dynamic' hedges, or cross-instrument hedges. Next, I learned first-hand, watching a hedge fund, for which I sub-advised some money with my own equity strategy, come a cropper on 3:1 leveraged money in 1990. The crux of the Sowood article proves my personal contention, as Larson's big losses came via complex, cross-instrument hedges, the risks of which were accentuated with highly leveraged money.

Thus, Larson's Sowood did two things my partner and I won't do. That could be why my large-cap equity strategy currently has a 33% YTD return, and our related options strategy has a return in the 60% range after only roughly six months of operation. Both are comparatively simple, single-instrument strategies using no leverage. We would prefer that customers, if and when we choose to open our strategies beyond their current proprietary nature, bring their own borrowed money, if any is to be used.

Perhaps the fundamental moral of the Sowood, and other, similar stories, comes from this paragraph in the Journal article,

"The $35 billion Harvard endowment would have been able to absorb Mr. Larson's losses. But like many traders who have recently left large organizations to start their own funds, Mr. Larson found he didn't have any leeway in times of crisis."

Another lesson I observed from my one-time colleagues who ran a stat-arb hedge fund in the late 1990s was that often, a strategy will work in the long term, but fail to sustain itself amidst losses and redemptions in the short term. In Sowood's case, Larson sold his portfolio to another hedge fund, Citadel Investments. The Journal piece notes,

"...Mr. Larson has told investors he'd like to resume his trading career. He said he wasn't reckless, and that his investments eventually would have led to profits. On this point, few would dispute: Citadel has made big gains on the investments, according to people familiar with the situation."

This is quite believable, as highly-leveraged trading/investing strategies often take losses which wipe out the sliver of equity supporting the total asset position, for a time, before rebounding after whatever financial crisis caused the unexpected reversal of the hedged positions' values.

It seems that some managers are incapable of understanding lessons which they have not personally experienced. For instance, it's fairly obvious now that the correlations upon which many cross-instrument hedges are based can rapidly and easily evaporate in times of unexpected financial crises. And the crises are typically things which nobody foresaw, or for which low probabilities were generally assigned- the Russian sovereign debt default, LTCM's massive hedging losses affecting even equities, or the recent large-scale seizure of 'markets' for structured finance instruments.

Years of respectable, if not out sized portfolio returns can be dwarfed by losses in a crisis lasting only weeks or months. Leverage and the need to provide margin funding can quickly consume the relatively small amount of equity in a leveraged portfolio. And, last of all, the worst time to try to trade one's way out of trouble is typically in the midst of a crisis in which prices, if they exist at all, may be fleeting or ultimately unknown with certainty for some time.

It's a fine line between confidence in a successful asset management strategy, and hubris that nothing can go wrong with it- ever. It seems to me that avoiding crippling, lethal losses on the downside is more important than attaining the maximal possible returns to a strategy on the upside.