On Jun 28, 2006, at 1:05 PM, Marvin Gandall wrote:
Does anyone know if he is accurate in reporting that "investment strategies that supposedly were well hedged, so that losses in one sector would be offset by gains in another, turned out to be 'correlated', so that they all went down together"? I wasn't able to confirm the information on my own.
That probably means they were doing some version of trading on spreads - e.g., the yield gap between low- and high-grade bonds, to pick a simple example. In normal times, such movements are well- behaved, and when they reach a certain level, you can semi-safely bet they'll revert to the mean. Or some such. But in extraordinary times - those six standard deviation moments - such strategies turn very sour. It's not that surprising, really; in a year with about 250 trading days, even something that works 99% of the time is going to fail 2.5 days. And when they fail badly, they can wipe out the other 247.5 days' gains. Also, like all crowds, hedge funds often do the same thing - short GM bonds, buy Russian stocks, load up on oil futures. So when the hot markets turn, they can turn big. Doug
