Bloomberg reports that Goldman’s big hedge fund, Global Alpha, which took a beating along with other quantitatively oriented traders, was down 22.5% in August. Even among quant funds, this was lackluster performance. James Simons’ Renaissance Technologies recouped the 8.7% loss it suffered at the beginning of the month.
But here comes the juicy bit: of that 22.5% loss, the biggest contributor was an 8.9% fall in currencies due to being hit by an unwinding of the carry trade. The fund had sold yen and bought Aussie dollars, and the Australian dollar fell 6% while the yen rose.
Let’s get this picture straight. Goldman is charging hedge fund fees and claiming to deserve them by virtue of having highly sophisticated models, and one of its biggest positions in a yen/Aussie dollar bet? You don’t need Goldman for that. Japanese housewives are doing far more complicated currency trades on their own. They can try to hide behind their models, but this is plain simple point of view punting on currencies.
But Goldman would have you believe otherwise:
“Longer term, successful quant managers will have to rely more on unique factors,” the firm’s fund-management division said in a report to clients. “While we have developed a number of these factors over the last several years, in hindsight we did not put sufficient weight on these relative to more popular quant factors.”
The claim that too many other quants were making the same trade is utter baloney. Huge numbers of Japanese speculators were into that trade. Everyone knew that. Every time the yen stated to appreciate, retail traders in Japan would sell it down, until even they lost their nerve.
But Goldman is fond of coming up with justifications for Global Alpha’s performance that presume that their investors are financially illiterate. At the time of the meltdown, they offered this explanation:
“We were seeing things that were 25-standard deviation moves, several days in a row,” said David Viniar, Goldman’s chief financial officer. “There have been issues in some of the other quantitative spaces. But nothing like what we saw last week.”
Brad DeLong had pointed out a mere day before that the world wasn’t old enough to have have a mere 16 standard deviation event.
If you are going to make excuses, you need to come up with something better than the investor relations version of “the dog ate my homework.”