Why Risk First?
The best strategy is worthless if you're out in round one. Three cases in which mathematical geniuses, stock market prodigies, and trading legends failed — not because of their strategy, but because of risk discipline they treated as a formality.
„Picking up nickels in front of a steamroller."
How two Nobel laureates and a dream team tried to double the world — and lost $4.6 billion in six weeks.
In 1994, John Meriwether, formerly a legendary bond trader at Salomon Brothers, founded a hedge fund unlike any before: Long-Term Capital Management. On board were finance academics Myron Scholes and Robert Merton — three years later both received the Nobel Prize for their options pricing theory. The strategy sounded elegant and safe: Convergence Trades. They searched for bond spreads that statistically had to converge again and pocketed the gap.
To turn tiny margins into real returns, LTCM needed leverage. And massive leverage at that. At times the fund operated at 30:1 on its own capital, with the notional of open positions at $1.25 trillion — with a "t", larger than Germany's GDP at the time. For four years LTCM delivered over 40% return per year. Wall Street puzzled over how. Banks fought to be counterparty on every trade.
In August 1998, Russia unexpectedly defaulted on its ruble bonds. What was a footnote in the normal distribution models of Scholes and Merton was, in reality, a wildfire. Panicking investors fled to safe assets — all convergence spreads widened simultaneously instead of converging. The leverage that had been a return machine for four years became a wrecking ball. Within a few weeks, LTCM burned through $4.6 billion. The Fed had to organize a consortium of 14 banks, each contributing around $300 million, just to prevent a systemic collapse.
The mathematics behind the trades was brilliant — but it was based on the assumption that returns are normally distributed. Fat Tails didn't fit the model. And anyone leveraged 30:1 has no buffer when the real world refuses to follow the bell curve.