Whole Earth: When genius failed Failed: The Rise and Fall of Long-Term Capital Management.

When Genius
Failed
The Rise and Fall
of Long-Term Capital
Management
Roger Lowenstein
2000; 264 pp.
$26.95
Random House

This compelling book is the story of some guys who really needed to do scenario planning. One trillion dollars almost went down with Long-Term Capital Management (LTCM), a hedge fund that was created by the best minds on Wall Street (not to mention a pair of future Nobel Laureates). The principals were extremely successful and respected--they knew what they were doing. Yet their overdependence on financial models, exacerbated by overconfidence, blinded them to a cascade of loan defaults that came close to collapsing the US financial system.

If the financial establishment--coerced at the eleventh hour by the New York Federal Reserve--had not acted together to rescue the company when its elaborate "sure-fire methods" failed, they would likely have unleashed an unprecedented national and global financial crisis.

LTCM could have survived, except that greed got in the way Flush by their early successes, the principals returned their investors' money so that they could keep all the gains themselves. With those [investor] funds, LTCM might have weathered the crisis. --Peter Schwartz (courtesy Global Business Network)

"This one obscure arbitrage fund had amassed an amazing $100 billion in assets, virtually all of it borrowed--borrowed, that is, from the bankers at [New York Fed President] McDonough's table. As monstrous as this indebtedness was, it was by no means the worst of Long-Term's problems. The fund had entered into thousands of derivative contracts, which had endlessly entwined it with every bank on Wall Street. These contracts, essentially side bets on market prices, covered an astronomical sum--more than $1 trillion worth of exposure.

"One can be big (and therefore illiquid); one can (within prudent levels) be leveraged. But the investor who is highly leveraged and illiquid is playing Russian roulette, for he must be right about the market not only at the end, but every single day. (One wrong day, and he is out of business.) Long-Term was so self-certain as to believe that the markets would never--not even for a wild swing some August or September--stray so far from its predictions.

... The professors' conceit was to think that models could forecast the limits of behavior. In fact, the models could tell them what was reasonable or what was predictable based on the past. The professors overlooked the fact that people, traders included, are not always reasonable. This is the true lesson of Long-Term's demise. No matter what the models say, traders are not machines guided by silicon; they are impressionable and imitative; they run in flocks and retreat in hordes.