Imagine running a hedge fund so successful you must return billions because you cannot find enough trades to deploy. That was Long-Term Capital Management at the end of 1997, and the decision the founders made next, with two Nobel laureates on the masthead, nearly took the global financial system down. The paradox matters because the same instinct shows up in retail accounts today.
The setup that looked like genius
LTCM was the smartest room on Wall Street. Nobel Prize winners Myron Scholes and Robert Merton were partners. Founder John Meriwether had built the legendary bond arbitrage desk at Salomon Brothers. The fund’s core trade was convergence arbitrage, betting that unusually wide spreads between similar bonds would narrow back to normal. In quiet markets, this works beautifully. Small edges, repeated thousands of times, financed with enormous borrowing, produce enormous returns.
By late 1997 the fund had made so much money that Meriwether wrote to investors that “the fund has excess capital” because of “a substantial increase in the capital base from the larger than expected past realized rates of return.” In plain English, they were too good. There are not enough mispriced spreads to soak up all this money at the returns you expect.
The fatal decision to hand back $2.7 billion
So LTCM did something almost unheard of. It forced investors to take back $2.7 billion at the end of 1997. Many investors resisted, because a spot in LTCM was a status object. Crucially, the partners kept their own money in. A larger share of the remaining fund now belonged to them.
If the story stopped there, it would be a cute anecdote about a disciplined firm returning capital it could not deploy. Instead, LTCM kept the positions the same size and made up the difference by borrowing more. Leverage went up. The bets did not shrink to fit the smaller equity base. The cash pile shrank to fit the same bets, and debt filled the gap.
If a trade is levered ten to one, a 1% move against you costs 10% of equity. If you push it to twenty five to one, that same 1% move costs 25%. If the market goes against you, instead of losing a little, you lose a lot. The upside got concentrated in the partners’ pockets. So did the downside, which nobody was modeling correctly.
When the convergence trade stopped converging
In August 1998 Russia defaulted on its ruble debt. Investors ran for the safest, most liquid securities. Spreads that LTCM had bet would narrow blew out instead. Liquidity vanished at exactly the moment LTCM needed to reduce positions. The Federal Reserve Bank of New York eventually organized a consortium of banks to recapitalize the fund and unwind it in an orderly way, a moment Alan Greenspan later defended before Congress as necessary to prevent a broader chain reaction.
What still makes this story live is that spreads move violently. The 10-year minus 2-year Treasury spread compressed from 0.74% in early February 2026 to 0.27% by late June, a fast repricing that would have punished anyone levered against it. The VIX still spikes on cue, hitting 31.05 in late March 2026 before settling back to 16.45 by June 30. Forced sellers show up when volatility jumps.
The lesson for your own account
Leverage is a crowbar.
It pries open returns you could not otherwise reach, and it pries open holes in your account when the trade turns. LTCM’s mistake was believing its own track record so completely that it added borrowed money precisely as its edge was shrinking. If you use margin, options, or leveraged ETFs, size the position around the loss you can survive. Past success, even the Nobel kind, does not repeal arithmetic.
Contact [email protected] for any questions or corrections.