The side hustle that lost $4.8 billion in a single day

The Smartest Guys in Every Room

Long-Term Capital Management launched in 1994 with the kind of pedigree that made investors write checks before asking questions. John Meriwether, formerly of Salomon Brothers, assembled a partnership that included Myron Scholes and Robert Merton—who would win the Nobel Prize in Economics in 1997, mid-fund. The pitch was elegant: use complex mathematical models to identify tiny pricing inefficiencies in bond markets, then leverage those tiny edges into massive returns. Minimum investment was $10 million. The fund returned 21% in 1995, 43% in 1996, and 17% in 1997. By early 1998, LTCM managed $4.8 billion in capital and had borrowed against it so aggressively that its total positions exceeded $1.25 trillion—more than the GDP of most countries. The models said it was safe. The Nobel laureates said it was safe. Every risk metric said losses of this magnitude were statistically impossible.

When Russia Broke the Math

The unraveling began in August 1998 when Russia defaulted on its government debt—an event LTCM's models rated as essentially impossible. Suddenly every "safe" convergence trade moved in the wrong direction simultaneously. The fund had bet that similar bonds would move toward the same price; instead, panicked investors fled to the safest assets, making spreads widen dramatically. LTCM lost $1.9 billion in August alone. September was worse. The fund's leverage meant tiny market moves translated into catastrophic losses. By late September, LTCM had lost $4.6 billion, erasing 90% of its capital. The problem wasn't just that LTCM was dying—it was dying so large that its collapse would force liquidation of $1 trillion in positions, potentially triggering a cascade of bank failures. The fund had 60,000 derivatives contracts with most major Wall Street firms. Every counterparty would take simultaneous hits. The Federal Reserve Bank of New York called an emergency meeting.

The Bailout That Wasn't Quite a Bailout

On September 23, 1998, fourteen banks gathered in the Fed's offices overlooking Wall Street. William McDonough, president of the New York Fed, made clear this wasn't a government rescue—but if the banks didn't act, some of them would fail along with LTCM. The solution was a consortium bailout: fourteen firms contributed a total of $3.625 billion in exchange for 90% of the fund. Meriwether and his partners were wiped out but kept 10% and stayed on to unwind positions. The banks weren't being charitable. They were protecting their own derivative books. If LTCM collapsed suddenly, it would trigger margin calls across Wall Street, forcing everyone to sell identical positions into a panicked market. Over the following year, the consortium slowly unwound LTCM's positions, ultimately recovering the full $3.6 billion investment. The partners got nothing. By 2000, the fund was formally dissolved. Total time from "statistical impossibility" to total collapse: four months.

The Legacy Nobody Wanted to Learn

Long-Term Capital Management became business school shorthand for the danger of models that can't imagine their own failure. The firm's core mistake wasn't mathematical—it was philosophical. The models assumed markets would always provide liquidity, that you could always exit a trade at model-predicted prices. In August 1998, liquidity vanished. Nobody wanted the other side of LTCM's trades at any price. Leverage turned a bad month into a terminal event. Congress held hearings. Regulations were proposed. Then everyone moved on. The same banks that funded the LTCM bailout would, a decade later, use similar leverage and similar confidence in quantitative models to trigger the 2008 financial crisis. The lesson was clear. The industry didn't learn it. Today, Meriwether's name is a Wall Street punchline, and Scholes and Merton are still Nobel laureates. LTCM's models are still taught in finance courses—as examples of what happens when mathematical elegance meets market reality, and reality doesn't blink first.