A simple lesson could be the key to understanding the failure of Long-Term Capital Management LP. Give a money manager the license to use leverage -- via debt, derivatives or both -- and you give him (or her) an unusual power and temptation.
The power is the ability to sculpt return patterns in ways that make it totally dissimilar to the "bell curve" distribution that we generally accept as applying to a more traditional portfolio. This leads to the temptation to create a portfolio in which, on an ongoing basis, a small possibility of a huge loss is traded for the much greater likelihood of a series of small gains.
This ultimately can lead to disaster for the investor, but for the money manager it may seem a risk worth taking. With a series of small gains, the money manager can earn some tidy bonuses and establish his reputation. An optimistic money manager will not necessarily assume that things will never go wrong, but he will hope that if things do go wrong, he'll somehow be able to trade his way out of it.
There are two basic strategies money managers can use with leverage and derivatives to sculpt return patterns in such a manner.
The first is a strategy well known in Las Vegas. Most of the time, it is easy money. Bet $100 on black. If you lose, bet $200 on black. If you lose again, bet $400. Then $800. And so on. Running out of money? No problem -- max out your credit cards, hit up your friends, take out a second mortgage. If 10 times is your limit, about 99.9% of the time you'll win $100. Of course, the problem is, when you lose, you lose more than $100,000.
My point is, you can make some nice money betting against three standard deviation events -- until the three standard deviation event occurs. Could a lesson this simple be the key to understanding the Long-Term Capital failure?
If a traditional money manager tried this, it would be obvious pretty quickly what he was doing. But macro hedge fund managers have such a mystifying bag of tricks (leverage, exotic options, multiple markets) they might get away with it, at least for a while.
The other strategy is for the manager to use securities or derivatives that have the "rare occurrence of big loss/usual occurrence of small gain" embedded in them. Just one example is a swap with an above-market yield that will incur a huge liability if some unlikely major event occurs.
Until the unlikely event occurs, the money manager can demonstrate a great track record. Returns can be above market, with market (or lower) volatility. But volatility as a measure loses its usefulness when applied to potentially abnormal return patterns.
Is the money manager consciously duping the investor? Not necessarily. The manager may believe (and it may be true) that his ability to predict the likelihood of an unlikely event is greater than that of his counterparty. If his strategy does blow up, he undoubtedly will blame the unusual event: "Who could ever expect spreads to widen the way they did?" Or "If it hadn't been for that drop in the price of oil . . ."
Investors therefore should have a thorough understanding of the theoretical source of the money manager's profits. The trading activity should be researched and (if possible) constrained to ensure it is consistent with that theoretical profit source. Examining past trades can help as well.
For example, investors might examine how the money manager has handled periods of unexpected volatility. When the volatility has hurt the manager, has he minimized his losses, or taken on bigger bets?
If possible, investors should be precise in limiting leverage. While the leverage inherent in overlaying S&P 500 futures on a fully invested stock portfolio may be obvious, the leverage created by selling an outperformance option on the French vs. the Spanish market, and offsetting it by buying a knockout butterfly collar, may be far less plain to see. A money manager may plan to offset a leveraged position dynamically, but if his volatility assumptions are flawed, or if he is not true to the strategy, the offset may fail.
Clearly, leverage and derivatives have many appropriate uses, both as tools for hedging risk and for legitimate speculative purposes.
But as has become increasingly obvious, they must be approached with great caution -- not because they are risky, but because of the manner in which they can be risky. They should be given a wide berth by anyone unwilling to do a thorough analysis that goes far beyond the basic assumptions generally applied to the underlying instruments and markets they are used to purchase, or from which they are derived.
Long-Term Capital's investors and lenders clearly failed to make this kind of thorough analysis. Some events are so unlikely (a meteor strike on New York City, for example) that even the most prudent investor may ignore the chance that they'll occur.
But the recent simultaneous run on emerging markets (and associated widening of U.S. spreads) does not qualify as such an unlikely event. What's worse, the managers of Long-Term Capital didn't just ignore the chance of its occurrence: rather, they placed huge, highly leveraged bets against the possibility that it would. These bets resulted not only in huge losses to the investors and lenders, they may well serve to exacerbate the effects of the emerging markets' decline.