When Genius Failed

Book review:

$ $ $ $ $ $

In light of the current turbulence in the markets, I downloaded and listened to a book that has been on my list for a while: When Genius Failed, The Rise and Fall of Long Term Capital Management by Roger Lowenstein. The book was read by the author which I always enjoy -- it gives it a little personal touch. If you don’t instantly fall asleep at the mention of derivatives and equity swaps I would recommend it. I gave it 5 out of 6 dollar signs.

Quick summary:

As the name implies, the book is about the astounding rise and fall of Long Term Capital Management from 1994-1998 (not that long ago!). Long Term Capital was one of the first firms to really hire and rely on phd’s and academics to drive their decisions when making trades. They were also one of the most successful investment groups ever, until very quickly in a matter of one month everything started to fall apart. In the end the company is brought down by overconfidence in it’s own models. It turns out the economy of all of Russia including its people, its culture and its politics cannot be easily rolled up into a formula of volatility and expected returns.

Personal touch:

Part of what makes the book enjoyable is that it is largely a story of the people involved. In between the talk of stocks and bonds, it’s full of little windows into the lives of the Wall Street CEOs and the hedge fund managers. It talks about their friendships, their loyalties, their arrogance, and their tears when things start to go wrong. There are even dramatic scenes where the heads of the major banks in wall street get together like heads of competing mafias to make decisions that would affect and maybe save the whole market, but at the same time they are secretly working in their own self interest. More than anything, it paint the people involved as very human and ironically far removed from the robotic beings that LTCM’s models assume traders to be.

A lesson:

Economics models just aren’t that great. If you were to use Long Term Capital’s models to predict the chance of their own failure then you would have concluded that failure as it occurred was for all practical purposes impossible. According to their models if the markets had been open everyday from the beginning of the universe until now it would still have been unlikely for some of events that occurred to have happened… but of course they did. In the late 90s LTCM had the best models available… that wasn’t that long ago and as is now obvious current models aren’t much better.