Update: 8/3/2012 This article related to an automatic trading algorithm run a-muck is pertinent to this book. http://dealbook.nytimes.com/2012/08/0...
A larger question is whether this kind of trading benefits capitalism, in the sense it helps supply capital for businesses to grow, or whether it serves only the financial industry in its quest for making huge amounts of money without making anything.
The best thrillers and horror stories don’t involve chain saws or mutated snakes. They take something prosaic, something we are all familiar with, something we trust, and then tweak it.
That’s the premise of a wonderful book I read years ago calledAdolescence of P-1
by Thomas Ryan. An algorithm created to seek out knowledge learns and soon desires to protect itself. Read it.
What separates humans from other animals is language. That used to be the case. It’s no longer true. Computers can now assimilate, translate, and communicate. Not to mention that they are all connected and have access to virtually all of human knowledge. So can they learn and adapt?
The other thread that surfaced while reading Fear Index is what happened to Long Term Capital Management in 2000. A hedge fund founded by a brilliant economist, the premise was that if you get enough smart people together you can write some brilliant computer algorithms that will permit outperformance of the market. They collapsed spectacularly in 2000 requiring a bailout.
Combine those two events and ideas and you have Fear Index.
Alex Hoffman is a brilliant physicist who founds a hedge fund which does spectacularly well bring returns of 80% to its investors. Constantly tweaking its algorithms to provide even better returns, they are poised to add even more money to their fund. But some weird things are beginning to happen. Alex gets a rare book in the mail with no indication who might have sent it. Investigation reveals it was purchased using an account in the Caymans he owned but was not aware of. All the transactions happened over the Internet. Then his house is broken into by someone who had all the key codes.
The algorithm Hoffman’s company created uses sophisticated analysis of human fear to determine market positions. All of a sudden the company finds itself taking huge short positions that pay off when companies they were shorting suffered some form of catastrophe. They reap immense profits, but how could the algorithm have foreseen those seemingly random events? And how did the notes of Alex’s therapy get on the web. And who installed the surveillance cameras and built the server farm? In the meantime, Hugo’s algorithm is learning and doing things on its own. Really good story. 4.5 stars. Less than 5 only because it’s so similar in concept to Ryan’s book.
As an aside, Harris has a very nice explanation of what a hedge fund does and how hedges work. Let’s say, as in his example, that you make a bet with someone for $1,000,000 that the girl across the hall is wearing black knickers. Whether she is or not really isn’t relevant to the bet because you’re going to bet someone else $950,000 that her knickers are NOT black, so your maximum exposure is $50,000. You can, of course, hedge your other bet even further as will all the other betters. There will be winners and losers each time the knickers are revealed but the maximum exposure should be relatively small and if you make enough bets over a period of time it doesn’t matter if you are wrong some of the time, because you’ll win enough times to come out ahead. In theory. That’s what hedge funds do and it explains why there are trillions of dollars being bet on trivialities and when things go wrong, as in Corzine’s little escapade, billions can be lost, or in his case, disappear. (Did the accountants make off with it, perhaps?) In the meantime, I’m making a side bet that the fried, fat encrusted ribs and whipped-cream-covered chocolate chip ice cream I just ate won’t kill me before I get this review up.