The Long-Term Capital Management Collapse

Editor's Note

A hedge fund staffed by Nobel laureates, a former Federal Reserve vice chairman, and the sharpest minds on Wall Street lost nearly its entire $4.7 billion in capital in under five months. The 1998 collapse of Long-Term Capital Management remains one of the most instructive episodes in the history of financial risk; a case study in how leverage, model overconfidence, and interconnectedness can threaten the stability of the entire global financial system.

The Dream Team of Finance

Long-Term Capital Management was founded in 1994 by John Meriwether, a legendary bond trader who had built the enormously profitable Arbitrage Group at Salomon Brothers during the 1980s. Meriwether's reputation had survived a Treasury bond-bidding scandal that forced his departure from Salomon in 1991, and he attracted an extraordinary roster of talent to his new venture. The founding partners included Myron Scholes and Robert C. Merton, who would share the 1997 Nobel Memorial Prize in Economic Sciences for their work on options pricing theory. David Mullins, formerly the vice chairman of the Federal Reserve Board, also joined as a principal. The academic and institutional pedigree was without parallel in the hedge fund world.

LTCM launched with $1.25 billion in capital; at the time, the largest initial fund-raising in hedge fund history. Meriwether charged investors a 2 percent management fee and 25 percent of profits, above the standard hedge fund terms of the era. The minimum investment was $10 million, with a three-year lock-up period during which investors could not withdraw their money. Major investors included Merrill Lynch, the University of Pittsburgh endowment, the Bank of Italy, and numerous wealthy individuals. The fund's mystique was such that investors competed for the privilege of placing their capital with the firm.

The Strategy: Convergence and Relative Value

LTCM's core strategy was relative value arbitrage; identifying small pricing discrepancies between related securities and betting that these discrepancies would narrow over time. The fund's traders focused primarily on fixed-income markets, exploiting the spread between on-the-run Treasury bonds (the most recently issued and therefore most liquid) and off-the-run bonds (older issues that were slightly less liquid but otherwise nearly identical). These spreads were tiny; often just a few basis points; but LTCM amplified them through massive leverage.

The firm also traded mortgage-backed securities spreads, European government bond convergence plays driven by the approaching launch of the euro, interest rate swap spreads, and equity volatility positions. The underlying logic was consistent: markets contained small inefficiencies that would correct over time, and a sufficiently patient, well-capitalized investor could harvest these corrections reliably.

The mathematical models that guided LTCM's trading were sophisticated extensions of the Black-Scholes framework. They relied on historical correlations between asset classes, estimates of the probability distribution of price movements, and the assumption that markets would, in the medium term, revert to equilibrium relationships. The models were elegant and, for the first several years, spectacularly successful.

The Golden Years

LTCM's early returns were extraordinary. In 1994, the fund returned 20 percent net of fees. In 1995, it returned 43 percent. In 1996, it returned 41 percent. These were remarkable numbers for a strategy that was supposed to be low-risk; the returns reflected the enormous leverage the fund employed to magnify small pricing discrepancies into outsized profits.

By the end of 1997, LTCM's capital had grown to $7 billion and the fund controlled a portfolio of approximately $125 billion in assets; a leverage ratio of roughly 25 to 1. But this figure understated the true risk. The fund also held off-balance-sheet derivative positions with a notional value exceeding $1.25 trillion. The total risk exposure dwarfed the capital base by a factor that made even experienced Wall Street risk managers uneasy Lowenstein (2000).

Ironically, the fund's very success planted the seeds of its destruction. By late 1997, LTCM's partners concluded that the fund had grown too large relative to the opportunities available. They returned $2.7 billion to outside investors, reducing capital to $4.7 billion while maintaining the same portfolio size. The leverage ratio rose accordingly; from roughly 25:1 to approximately 28:1 on the balance sheet alone. The partners were, in effect, concentrating risk at precisely the moment when markets were about to turn against them.

The Russian Crisis and the Flight to Quality

The catalyst for LTCM's collapse came from an unexpected direction. On August 17, 1998, the Russian government defaulted on its domestic ruble-denominated debt and devalued the ruble. The default itself was not a major direct exposure for LTCM, but its secondary effects were devastating.

The Russian crisis triggered a global flight to quality of extraordinary intensity. Investors worldwide panicked and rushed to sell risky assets; emerging market bonds, corporate debt, mortgage-backed securities; and buy the safest instruments available: US Treasury bonds. The spreads that LTCM had bet would narrow instead blew out to levels that the fund's models considered virtually impossible. The spread between on-the-run and off-the-run Treasuries, which had historically averaged about 10 basis points, widened to more than 30 basis points. Swap spreads, corporate bond spreads, and mortgage spreads all moved violently against the fund's positions simultaneously.

The timing was catastrophic. The Asian financial crisis had already been destabilizing global markets since mid-1997, and the Russian default transformed what had been regional stress into a worldwide panic. Correlations between asset classes; which LTCM's models had assumed would remain low or negative; spiked toward one. Diversification, the foundation of modern portfolio theory, ceased to function. As one market participant later observed, in a crisis, the only thing that goes up is correlation.

The Spiral

In August 1998 alone, LTCM lost $1.85 billion; roughly 45 percent of its remaining capital. The fund lost $553 million on August 21 in a single day. September brought further losses. By mid-September, LTCM's capital had fallen to approximately $600 million, while its balance sheet assets still exceeded $100 billion; an effective leverage ratio of roughly 167 to 1.

The losses were self-reinforcing. As LTCM's capital shrank, the fund was forced to reduce positions to meet margin calls from its counterparties. But selling into a panicked market drove prices further against the fund's remaining positions, triggering more margin calls and more forced selling; a vicious cycle that risk managers call a "doom loop." Other funds that held similar positions; and many did, having been inspired by LTCM's early success; were also liquidating, amplifying the downward pressure.

The fundamental problem was not that LTCM's trades were wrong in the long run; most of the convergence bets did eventually prove correct. The problem was that the fund could not survive long enough for the trades to work. As Keynes famously cautioned, the market can remain irrational longer than you can remain solvent. LTCM's extreme leverage left no margin for error, and the magnitude of the short-term moves vastly exceeded what the Value-at-Risk models predicted was possible.

The Bailout

By late September 1998, LTCM's impending collapse posed a threat to the entire financial system. The fund had derivative contracts with virtually every major financial institution on Wall Street. A disorderly liquidation of its $100 billion-plus portfolio would have forced fire sales across global markets, potentially triggering a chain reaction of failures among LTCM's counterparties; the largest banks and investment houses in the world.

William McDonough, president of the Federal Reserve Bank of New York, convened an emergency meeting on September 22 at the New York Fed's offices on Liberty Street. Representatives of the major Wall Street firms; including Goldman Sachs, Merrill Lynch, JPMorgan, Morgan Stanley, Salomon Smith Barney, Lehman Brothers, Bear Stearns, Chase Manhattan, Bankers Trust, Credit Suisse First Boston, Deutsche Bank, UBS, Barclays, and Societe Generale; gathered to assess the situation and negotiate a rescue Edwards (1999).

After intense negotiations that lasted through the night, a consortium of 14 financial institutions agreed to inject $3.65 billion into LTCM in exchange for 90 percent ownership of the fund. The existing partners, including Meriwether, Scholes, and Merton, saw their stakes diluted to just 10 percent. The consortium would manage the orderly liquidation of the portfolio over time, allowing positions to mature rather than being dumped into a panicked market.

Notably, the Federal Reserve did not commit any public funds to the rescue. McDonough's role was that of a convener and facilitator; using the moral authority of the central bank to bring the parties together and prevent a collective action problem in which each firm might have preferred to let others bear the cost of stabilization. Some critics argued that even this degree of Fed involvement constituted a dangerous precedent, signaling that firms deemed systemically important would be protected from the consequences of their risk-taking; a concern that would prove prescient when the 2008 financial crisis unfolded a decade later.

Systemic Risk and the Lessons Unlearned

The LTCM debacle exposed fundamental weaknesses in the architecture of global financial markets. The fund had borrowed from and entered into derivative contracts with dozens of counterparties, none of whom had a complete picture of LTCM's total exposure. Each bank knew its own relationship with the fund but not the aggregate risk that LTCM posed to the system. This opacity; the inability of any single regulator or market participant to see the full web of interconnections; was a defining feature of pre-crisis financial markets and one that the crisis did remarkably little to correct.

The President's Working Group on Financial Markets, led by Treasury Secretary Robert Rubin and including Federal Reserve Chairman Alan Greenspan, published a report on the LTCM episode in April 1999. The report recommended improved risk management practices, greater transparency from hedge funds to their creditors, and better regulatory oversight of leverage. But few of these recommendations were translated into binding regulation. The prevailing ideology of the late 1990s held that financial markets were self-correcting and that sophisticated market participants did not require the protection of government regulation Rubin and Greenspan (1999).

The lessons that LTCM should have taught; about the dangers of excessive leverage, the fragility of historical correlations in crisis conditions, the limits of mathematical models, and the systemic consequences of interconnected financial institutions; went largely unheeded. Within a decade, the same dynamics; leverage, model overconfidence, opacity, and interconnectedness; would re-emerge on a vastly larger scale in the mortgage-backed securities market, producing a crisis that dwarfed LTCM's collapse by orders of magnitude.

John Meriwether went on to found JWM Associates in 1999 and later JM Advisors, both of which eventually closed after suffering significant losses. Myron Scholes co-founded Platinum Grove Asset Management. The appetite for leveraged, model-driven strategies proved irrepressible; a reminder that in financial markets, the lessons of history are learned slowly, if they are learned at all.