Long-Term Capital Management: When the Smartest People in the Room Used Too Much Leverage
A hedge fund run by Nobel laureates and star traders lost over $4 billion in a matter of weeks, threatening the wider financial system and forcing a Wall Street-funded bailout.
- Peak leverage
- Reported around 25:1 on capital
- Losses
- $4.6 billion in under 4 months
- Resolution
- $3.6B bailout by 14 banks, coordinated by the Fed
What happened
Long-Term Capital Management was a hedge fund founded in 1994, staffed with elite traders and two Nobel Prize-winning economists. Its strategy relied on sophisticated statistical models identifying small, reliable pricing gaps between related securities — and using very high leverage to turn those small gaps into large returns.
The models assumed that historical relationships between markets would hold, or at worst, revert gradually. In 1998, Russia defaulted on its debt, and global markets reacted in ways the models had not anticipated — volatility spiked and the statistical relationships LTCM relied on broke down simultaneously, across nearly every position the fund held at once.
Because the fund was leveraged so heavily, losses that would have been manageable at a normal position size instead threatened to wipe out the fund entirely — and because LTCM's positions were so large relative to the markets it traded, its potential collapse threatened to disrupt those markets directly. The Federal Reserve organized a consortium of banks to inject capital and unwind the fund in an orderly way rather than risk a disorderly one.
Why it still matters
LTCM is the definitive case study for why leverage turns a survivable loss into an existential one. The fund's underlying trades were not obviously wrong — many of the pricing gaps it identified did eventually close as predicted. It failed because it was sized so aggressively that it could not survive the time it took to be right.
It is also a lesson about model risk: a strategy validated on historical data can fail exactly when it matters most, because a genuine crisis is, by definition, a period where historical relationships stop holding. Sizing a leveraged position assuming your model is correct is different from sizing it assuming your model could be wrong at the worst possible time.