Long-Term Capital Management (LTCM) Collapse: Causes and U.S. Intervention

Search Fintrev

Search our guides, reviews and glossary.

Long-Term Capital Management (LTCM) was a Greenwich, Connecticut hedge fund that collapsed in 1998. It had used enormous leverage to bet that small price gaps between related bonds would close. When Russia defaulted on its debt in August 1998, investors worldwide fled to the safest assets. The gaps LTCM expected to narrow widened instead, and the fund lost most of its capital within weeks. The U.S. response was not a taxpayer bailout. The Federal Reserve Bank of New York brought LTCM’s major creditors together, and those banks put in $3.625 billion of their own money to take over the fund and wind it down in an orderly way. Officials feared that a sudden default would force fire sales across markets that were already under stress.

What Was Long-Term Capital Management?

LTCM was a private investment partnership open only to institutions and wealthy investors. Meriwether had run Salomon Brothers’ bond arbitrage group, which made a fortune in the 1980s and early 1990s. He left Salomon after a 1991 Treasury auction scandal involving one of the firm’s traders. In 1994 he rebuilt much of that team at his own firm.

The roster was what made LTCM famous. Alongside Meriwether’s traders sat Scholes, co-author of the Black-Scholes option pricing model, and Merton, who extended that work; the two shared the 1997 Nobel Prize in economics. David Mullins, a former vice chairman of the Federal Reserve, was also a partner. The minimum investment and fees were steep.

The fund launched with more than $1 billion in capital. By 1998 it had drawn about $3.5 billion from outside investors.

How LTCM Made Money

Convergence trades

LTCM’s main trades relied on securities that should be priced almost identically but temporarily were not. The classic example involved U.S. Treasury bonds. A newly issued (“on-the-run”) 30-year Treasury trades more actively than a nearly identical bond issued a few months earlier (“off-the-run”). Investors pay a small premium for that liquidity, so the older bond yields slightly more. LTCM would buy the cheaper off-the-run bond and short the pricier on-the-run bond. As the new bond aged and lost its liquidity premium, the two prices would converge and LTCM would collect the difference.

The same logic applied across many markets:

• Government bond spreads between countries expected to join the euro, whose yields LTCM bet would converge before the 1999 launch.
• Interest rate swaps, trading the spread between swap rates and government bond yields.
• Mortgage-backed securities and corporate bonds against Treasuries, betting that credit and liquidity premiums were too wide.
• Equity volatility, selling long-dated options when implied volatility looked high against LTCM’s estimates. The positions grew large enough that some traders called the fund the “central bank of volatility.”
• Merger arbitrage and relative-value equity pairs, such as the dual-listed shares of Royal Dutch and Shell.

Each position was designed to be roughly market-neutral. The fund was not betting on whether rates or stocks rose or fell, only on whether one price moved back toward another.

Why leverage was essential

Convergence spreads were often measured in hundredths of a percentage point. Earning a meaningful return on such small differences required very large positions, and that required borrowing.

A hypothetical example shows the effect. A fund with $25 of assets for every $1 of capital, close to LTCM’s roughly $5 billion of capital against more than $100 billion of positions, earns about 12.5% on capital if its assets gain 0.5%, before financing costs. If they lose 4%, it loses about 100% of capital. With off-balance-sheet derivatives above $1 trillion in notional value, LTCM’s true exposure was larger than that ratio suggested.

Banks were willing to lend on generous terms. LTCM’s reputation meant many lenders accepted thin collateral (low “haircuts”) and gave the fund little information about its total book. By 1998 it had borrowed more than $155 billion. Its positions were estimated at about 5% of the global fixed-income market.

Early Success

At first, the approach worked spectacularly. After fees, investors earned roughly 21% in 1994, more than 40% in both 1995 and 1996, and about 17% in 1997.

That success created two problems. First, rivals copied the trades, so spreads narrowed and opportunities thinned. Second, LTCM had more capital than good trades. At the end of 1997 it returned about $2.7 billion to outside investors but kept its positions roughly the same size. Leverage rose just as the margin for error was shrinking.

The Collapse in 1998

Early strain

Trouble began before Russia. The 1997 Asian financial crisis had rattled emerging markets. In mid-1998 Salomon Brothers, by then part of Travelers, closed its own bond arbitrage desk. Many of its trades resembled LTCM’s, and unwinding them pushed spreads against LTCM. The fund lost money in May and June.

Russia’s default

On August 17, 1998, Russia devalued the ruble and defaulted on its domestic government debt, including short-term GKO bonds in which LTCM held a position. The direct losses mattered less than the reaction that followed. Investors around the world dumped anything risky or hard to sell and piled into the most liquid government bonds.

That flight to quality hit almost every LTCM trade at once. On-the-run Treasuries became more expensive compared with off-the-run issues. Credit spreads widened. Swap spreads blew out. Equity volatility jumped.

LTCM’s risk models were built on historical relationships and assumed that losses across its trades would partly offset each other. In this environment, they did not. Even as the fund lost hundreds of millions of dollars a day, its models signaled that the spreads were now even more attractive. In August alone it lost nearly half its capital.

Why LTCM could not get out

Cutting losses was nearly impossible. Many positions were so large that selling them would have moved prices sharply against the fund. Other market participants knew or guessed what LTCM held, which made the trades harder to exit. Lenders demanded more collateral as positions lost value. Each margin call forced the fund either to raise cash or to sell into a falling market.

Meriwether tried to raise new capital, including from Warren Buffett, George Soros and others. By September those efforts had failed. A proposal from a group led by Buffett was rejected because of its terms. As losses neared $4 billion, LTCM’s capital had shrunk to a fraction of its starting level while its positions remained enormous.

The U.S. Intervention

Why the Fed got involved

The concern was not LTCM’s investors. It was LTCM’s counterparties. Almost every major Wall Street and European bank had lent to the fund or held the other side of its derivatives. If LTCM defaulted, those banks would rush to seize collateral and close out positions at the same moment, in markets that were already illiquid. Officials at the New York Fed, led by President William McDonough, judged that such a scramble could drive prices down sharply, inflict large losses on the banks themselves, and spread stress through the financial system.

How the rescue worked

On September 23, 1998, the New York Fed hosted a meeting of LTCM’s major creditors. The outcome was a private recapitalization:

1. Fourteen banks and securities firms contributed a combined $3.625 billion. Bear Stearns declined to take part.
2. In exchange, the consortium took about 90% of the fund’s equity and oversight of its management.
3. LTCM’s partners and original investors were left with about 10%, and most of their earlier gains were wiped out.
4. The consortium funded the fund’s positions while they were unwound gradually instead of dumped.

No public money went into LTCM, and the Fed did not guarantee the loans. Its role was to organize and broker the deal. The Federal Reserve also cut interest rates three times in the fall of 1998, which helped calm markets more broadly.

Liquidation and aftermath

With the new capital in place, LTCM survived the rest of the 1998 turmoil and repaid the consortium over the following year. It was liquidated in an orderly way in early 2000.

Critics argued that by arranging the deal, the Fed had signaled that firms with enough counterparties might be protected from a disorderly failure, encouraging lenders to keep extending credit to highly leveraged borrowers. In 1999 the President’s Working Group on Financial Markets issued a report on hedge funds and leverage that called for better risk management by lenders and more disclosure.

What LTCM’s Collapse Shows

Leverage converts small errors into total losses. A strategy can be right in the long run and still fail if the fund cannot hold its positions through a short-term move against it.

Correlations change in a crisis. Trades that look diversified in calm markets move together when everyone sells at once, and models calibrated on normal periods understated that risk.

Liquidity is a risk, not just a cost. By collecting liquidity premiums, LTCM was effectively short liquidity exactly when liquidity disappeared.

Interconnection drives intervention. LTCM was not a bank, yet its links to banks made its failure a public concern. The problem returned on a larger scale a decade later with the collapse of Lehman Brothers, where no private rescue was arranged.

FAQs

The Bottom Line

LTCM's collapse matters to investors, lenders and regulators who rely on leverage, liquidity and historical models. A strategy can be right in the long run and still fail when borrowing leaves no room to hold through a short-term move against it. Trades that look diversified in calm markets can move together in a crisis, and a heavily connected firm can threaten the banks that lend to it even when it is not a bank itself.