The 1998 failure of Long-Term Capital Management (LTCM) is said to have nearly blown up the world's financial system.1 Indeed, the fund's woes threatened to create major losses for its Wall Street lenders. LTCM was so big that the Federal Reserve Bank of New York took the unprecedented step of facilitating a bailout of the private hedge fund, out of fear that a forced liquidation might ravage world markets.

For such a near-catastrophic event, the finance profession has precious little information to draw lessons from this failure. No doubt this is due to the secrecy of the hedge fund, which never revealed information about its positions, even to its own investors. As LTCM is now in the process of rehabilitating itself, it is slowly disclosing information about its risk management practices. By piecing together publicly available information, we can learn why LTCM failed.

This issue is vitally important, as LTCM's failure has been widely ascribed to its use of Value at Risk (VaR), the disturbing implication that the method currently used to set capital adequacy requirements for the banking sector is woefully inadequate. VaR itself was not the culprit, however. Rather it was the way this risk management tool was employed. LTCM used parameters for the model suitable for a commercial bank, but entirely unsuitable for a hedge fund. As an example, one of the parameters used to set the amount of equity capital was a 10-day horizon. The horizon must be related to the liquidity of the assets, or the time necessary for an orderly liquidation. Alternatively, the horizon should cover the time necessary to raise additional funds or for corrective action. Ten days may be sufficient for a commercial bank, which is closely supervised by a regulator who can step in at the first sign of trouble. For a hedge fund, however, the horizon should correspond to the period required to raise additional funds. This may be no easy matter, as additional capital will be needed precisely after the fund suffers a large loss. The 10-day horizon was clearly insufficient for LTCM. This, and the other unsuitable parameters produced a flawed model, which led LTCM to drastically underestimate the amount of capital they should set aside against market risks. At the same time, many other factors conspired to bring LTCM down as well.

How LTCM Lost its Capital

The seeds of LTCM can be traced to a highly profitable bond-arbitrage group at Salomon Brothers run by John Meriwether. LTCM was founded in 1994 by Meriwether, who left after the 1991 Salomon bond scandal. Meriwether took with him a group of traders and academics, who had been part of Salomon's bond-arbitrage group that had racked up billions of dollars in profits. Together, they set up a "hedge" fund using similar principles to those they had been using at Salomon.

A hedge fund is a private investment partnership fund that can take long and short positions in various markets and is accessible only to large investors. As such, hedge funds are not regulated by the SEC. Of course, the term "hedge" is somewhat of a misnomer, if not misleading, since these investment vehicles are leveraged and can be quite risky.

Initially, the new venture was eminently profitable. Capital grew from $1 billion to more than $7 billion by 1997. The firm was charging sky-high fees consisting of an annual charge of 2% of capital plus 25% of profits. By comparison, other hedge funds charge a 1% fixed fee and 20% of profits; the typical mutual fund fee is about 1.41%. By 1997, total fees had grown to about $1.5 billion. LTCM's 16 partners had invested roughly $1.9 billion of their own money in the fund.

Much has been said about LTCM's positions in the press. LTCM was supposed to have wagered $125 billion. This represents the total assets of the fund, most of it borrowed.2 Compared to equity of about $5 billion only, the size of assets appears ludicrous.

Even more scary was the off balance sheet position, including swaps, options, repurchase agreements and other derivatives, that added up to a notional principal amount of over $1 trillion. Many of these trades, however, were offsetting each other, so that this notional amount is practically meaningless. What mattered was the total risk of the fund.

LTCM was able to leverage its balance sheet through sale-repurchase agreements (repos) with commercial and investment banks. Under "repo" agreements, the fund sold some of its assets in exchange for cash and a promise to repurchase them back at a fixed price on some future date. Normally, brokers require collateral that is worth slightly more than the cash loaned, by an amount known as a haircut, designed to provide a buffer against decreases in the collateral value. In the case of LTCM, however, the fund was able to obtain next-to-zero haircuts, as it was widely viewed as "safe" by its lenders. This must have been due to the fact that no counterparty had a complete picture of the extent of LTCM's operations.

The core strategy of LTCM can be described as "convergence-arbitrage" trades, trying to take advantage of small differences in prices among near-identical bonds. Compare, for instance, a corporate bond yielding 10% and an otherwise identical Treasury bond with a yield of 7%. The yield spread of 3% represents some compensation for credit risk. If the corporate borrower does not default, a trade that is long the corporate bond and short the Treasury bond would be expected to return 3% for every dollar in the first bond. Short-term, the position will be even more profitable if the yield spread narrows further. The key is that eventually the two bonds must converge to the same value. Most of the time, this will happen - barring default or market disruption.

This strategy worked excellently for LTCM in 1995 and 1996, with after-fees returns above 40%. The fund placed large bets on convergence of European interest rates within the European Monetary System that paid off handsomely.

By 1997, however, convergence had occurred in Europe, as the common currency, the Euro, came into being on January 1999. Credit spreads were almost as narrow as they had ever been since 1986 and considerably lower than the average over the period 1986-93. Convergence trades had generally become less profitable. In 1997, the fund's return was down to only 17%. This performance, unfortunately, was trounced by U.S. stocks, which gained 33%. This was embarrassing, as LTCM touted itself as having the same risk as equities. If it had lower returns, why would anybody invest in the fund? LTCM had to look for other opportunities.

To achieve the 40% returns it had become accustomed to, the firm had to assume greater leverage. So, LTCM returned $2.7 billion of capital to investors in 1997 while keeping total assets at $125 billion. By shrinking the capital base to $4.7 billion, the leverage ratio went up, amplifying returns to investors that remained in the fund. Unfortunately, this also increased the risks. Troubles began in May and June of 1998. A downturn in the mortgage-backed securities market led to a 16% loss in the value of equity. LTCM's capital had just dropped from $4.7 to $4.0 billion.

Then came August 17. Russia announced that it was "restructuring" its bond payments - de facto defaulting on its debt. This bombshell led to a reassessment of credit and sovereign risks across all financial markets. Credit spreads jumped up sharply. Stock markets dived. LTCM lost $550 million on August 21 alone on its two main bets, long interest rate swap spreads and short stock market volatility.

By August, the fund had lost 52% of its December 31 value. Its equity had dropped faster than its assets, which still stood around $110 billion. The resulting leverage ratio had increased from 27 to 50-to-1. LTCM badly needed new capital. In his September 2 letter to investors, Meriwether revealed the extent of the losses and wrote that "Since it is prudent to raise additional capital, the Fund is offering you the opportunity to invest on special terms related to LTCM fees. If you have an interest in investing, please contact . . ." There were no takers, though.

The portfolio's losses accelerated. On September 21, the fund lost another $500 million, mostly due to increased volatility in equity markets. Bear Stearns, LTCM's prime broker, faced a large margin call from a losing LTCM T-bond futures position. It then required increased collateral, which depleted the fund's liquid resources. Counterparties feared that LTCM could not meet further margin calls, in which case they would have to liquidate their repo collateral.

A liquidation of the fund would have forced dealers to sell off tens of billions of dollars of securities and to cover their numerous derivatives trades with LTCM. Because lenders had required very low haircuts, there was a potential for losses to accrue while the collateral was being liquidated. In addition, as the fund was organized in the Cayman Islands, there was uncertainty as to whether the lenders could have liquidated their collateral. In contrast, such liquidation is explicitly allowed under the U.S. Bankruptcy Code. As it was believed that the fund could have sought bankruptcy protection under Cayman law, LTCM's lenders could have been exposed to major losses on their collateral.

The potential effects on financial markets was such that the New York Federal Reserve felt compelled to act. On September 23, it organized a bailout of LTCM, encouraging 14 banks to invest $3.6 billion in return for a 90% stake in the firm.

These fresh funds came just in time to avoid meltdown. By September 28, the fund's value had dropped to only $400 million. If August was bad, September was even worse, with a loss of 83%. Investors had lost a whopping 92% of their year-to-date investment. Of the $4.4 billion lost, $1.9 billion belonged to the partners, $700 million to Union Bank of Switzerland, and $1.8 billion to other investors.

LTCM is now operating under the control of a 14-member consortium, formally known as Oversight Partners I LLC. Helped by recovering financial markets, the portfolio gained 13% to December 1998. Since the bailout, the risk profile has been halved. As of this writing, though, the future of the fund is still in doubt.

Endnotes

1. Chairman Alan Greenspan (1998) testified that "Had the failure of LTCM triggered the seizing up of markets, substantial damage could have been inflicted on many market participants, including some not directly involved with the firm, and could have potentially impaired the economies of many nations, including our own."

2. According to the President's Working Group report (1999), the fund had 60,000 trades on its books. The balance sheet consisted of over $50 billion of long positions and short positions of an equivalent magnitude.

Philippe Jorion is a Professor of Finance at the Graduate School of Management at the University of California at Irvine.