Long-Term Capital Management (LTCM) – 1998

Given LTCM’s pedigree, it was often referred to as the Rolls Royce of hedge funds. It was headed by an erstwhile
Salomon Brothers bond market wizard and boasted of two Nobel Laureates amongst its partners as well as a
former Vice Chairman of the Federal Reserve. Dozens of PhDs in Mathematics and Physics were employed in
LTCM to undertake quantitative analysis and build mathematical models in support of its arbitrage trading
strategy in bond markets.

The dominance of “quants” (Quantitative analysis) was to be accepted given that the two Nobel Prize winners on
its rolls were Myron Scholes, the co-inventor of the famous Black Scholes model for pricing and hedging options,
and Robert Metton, an equally famous financial economist. The credibility and performance of the fund in its
early years had been such that it was able to leverage its trading activity on a scale far beyond what most other
hedge funds could manage. Further, many of the top names in commercial and investment banking world had
put their money in the fund in preference to their own proprietary trading activities. This apart, LTCM charged
higher fees than other fund managers (25% of profits) and did not allow investors to withdraw money in less than
three years. LTCM was also extremely secretive about its trading strategy. In fact, Arrogant and Secretive were
the two adjectives often used to describe LTCM.

LTCM was getting so much money in the fund that just about a year earlier, when the amount reached USD 7
billion, LTCM returned USD 2.7 billion to its investors. Its record spoke for itself – returns in excess of 40% in 1995
in 1995 and 1996 and a lower but still a strong return of 17% in 1997. With the sharp rise in the volatility in
financial markets in late summer of 1998, following Russia’s default on its domestic debt and the collapse of the
rouble, LTCM lost almost 90% of its capital in barely one month.

What had gone wrong? In many ways, and unlike others ( for example, George Soros’ Quantum Funds), LTCM was
operating more like a traditional “hedge” fund, whose returns, in theory at least, are independent of ,market
movements. Indeed, the name “hedge Funds” came into being because such funds shorted certain securities or
commodities, and went long in others of a similar class. If this were done, for example, in equity markets, returns
would, in theory, not be dependent on whether the stock market as a whole went up or down. The key, of
course, was in selecting the short and long positions in relation to historical price patterns. To give a simple
example, consider that the historical yield difference between AAA and AA bonds is, say, 0.5%. If, at a given point
of time, the yield difference widens to, say, 1%, it can be expected that the price of the AA bond would rise
relative to that of the AAA bond. If one shorts the AAA and goes long on the AA, a profit would be realised with a
reversion to the historical yield difference (the age old “reversion to mean” principle). Such a strategy obviously
required an enormous amount of quantitative analysis in different markets to pinpoint minute differences
between current prices and the historical pattern, in order to spot arbitrage opportunities. And this is what LTCM
was doing. Again, if, in efficient markets, minute price disparities are to lead to large profits, leveraging will have
to be very high, which too part of LTCM’s strategy. But leveraging is a double edged weapon. If the view on
“reversion to mean” turns out to be correct, profits will be high. If it goes wrong, losses too will be equally high.
And, the resultant margin calls from lenders would lead to liquidation of positions in unfavorable market
conditions, and this is exactly what happened to LTCM.
Ironically, while LTCM’s strategy became unstuck after the collapse of the rouble, it had very little exposure to the
Russian market. Most of its investments and positions were in European, Japanese and US bond markets. What
all the quants, rocket scientists and trading strategies could not forecast, however, was the scale of the “flight to
quality” that the rouble collapse would lead to. In principle, this consisted of investors preferring German and US
government bonds and running away from all other investments. The prices of these two bonds, therefore, rose
and the yield differences on which LTCM’s strategy was based, widened instead of narrowing. For example, LTCM
had shorted German bonds and went long on higher yielding Italian bonds in the hope that as the date for the
introduction of the Euro came near, so would the yields. Instead, they widened. Given the scale of LTCM’s
leverage, it was faced with margin calls, and could not hold the positions. It was forced to liquidate at huge losses
– so large that from the beginning of August to the third week of September 1998, LTCM lost 90% of its equity
and still had huge positions in the market. Its exposure as a borrower or a counterparty to transactions with the
banking system totaled an estimated USD 200 billion, and posed a huge systemic risk.

In a move that is bound to be debated for quite some time in hindsight, The Federal Reserve Bank of New York
took the lead in getting 15 leading commercial and investment banks to put in USD 3.7 billion in LTCM, which
gave them 90% management control for an orderly liquidation of the portfolio.

The two lessons of LTCM –

There are two major lessons to be learnt from the LTCM case. The first of course is that the quantitative models
based on historical patterns, however sophisticated they may be, cannot always predict the future in any reliable
way. As Lord Keynes said in a different context, “The inevitable never happens and the unexpected always
happens”. There is no way to structure al eventualities which may affect markets, in such models; the Russian
default on domestic debt, and its impact on bond yields worldwide, was one such. The prudent risk manager,
instead of relying totally on his models, should also limit his risks to what he can afford if his expectations go
wrong. In other words, there must be prudent limits on leveraging by the risk manager. Again, he must “stress
test” the models that he is using in order to estimate the impact on the portfolio value of unexpected events.

The second lesson is for banks dealing with highly leveraged counterparties. In such cases, banks should
understand precisely the strategy that is being adopted by the counterparty and not allow leveraging beyond
prudent limits. As LTCM’s bankers did not do so, they were forced to rescue the fund by putting in a huge amount
of their own money.

The Basel Committee on banking Supervision has published in 1999 a report on exposures of banks to highly
leveraged entities. The study was prompted by the LTCM debacle and the systemic risk that it entailed.