Two Nobel Laureates and a Fund That Nearly Broke Wall Street
Long Term Capital Management collapsed in 1998 running strategies its founders had helped invent. The failure was not a modeling error so much as a leverage error.
The Firm
Long Term Capital Management launched in 1994, founded by a former Salomon Brothers trader and staffed with an extraordinary roster including Myron Scholes and Robert Merton, who shared the Nobel Prize in economics in 1997 for work on option pricing.
The fund produced exceptional returns in its early years and attracted enormous capital. Then in 1998 it lost most of its value in a matter of weeks and required an intervention organized by the Federal Reserve Bank of New York, in which a consortium of major banks recapitalized it to permit an orderly wind down.
The Strategy
The core approach was relative value arbitrage. The fund identified pairs of closely related securities whose prices had diverged slightly from their historical relationship, bought the cheaper and sold the more expensive, and waited for convergence.
A classic example involved on the run and off the run Treasury bonds. The most recently issued Treasury of a given maturity trades slightly richer than an otherwise nearly identical older bond, because it is more liquid. The spread is small and has historically converged as the newer bond ages.
These trades were genuinely low risk in the sense that the relationships were stable and well documented. The problem is that a small spread produces a small return.
When the edge on each trade is tiny, the only way to earn a large return is enormous leverage, and leverage converts a small adverse move into a fatal one.
The Leverage
To convert small spreads into large returns, the fund borrowed heavily, operating at leverage ratios that left a very thin equity cushion. Additional exposure through derivatives made the effective position larger still.
At that leverage, a move of a few percent against the portfolio exhausts the capital. The strategy did not need to be wrong. It only needed to be temporarily wrong by a small amount.
What Went Wrong
In August 1998 Russia defaulted on domestic debt and devalued. The event itself was not enormous in global terms. Its effect on markets was.
Investors fled to the safest and most liquid assets available, which meant exactly the on the run instruments the fund was short, while selling the less liquid instruments it was long. Every convergence trade diverged simultaneously.
The positions were not diversified in the way the models assumed. They were many expressions of one underlying bet: that liquidity premiums would normalize. When the market demanded liquidity above all else, that single bet lost everywhere at once.
The Part That Made It Systemic
The fund's positions were so large that liquidating them would have moved markets against the counterparties holding the other side, many of whom were the same banks that had lent to it. Its trades were also widely imitated, so other firms held similar positions and faced the same losses.
That is why the intervention happened. Not to rescue the partners, who lost most of their money, but because a disorderly liquidation of that size threatened the institutions on the other side.
The Bottom Line
LTCM's trades were reasonable and its leverage was not. Positions that look diversified because they involve different securities can be one position if they depend on the same condition, and leverage removes the ability to wait for it to return.