Huge Carry Trade Blowup at Goldman

Bloomberg is reporting Goldman's Global Alpha Fund Fell 22 Percent in August

Sept. 13 (Bloomberg) -- Goldman Sachs Group Inc.'s Global Alpha hedge fund fell 22.5 percent in August, its biggest monthly decline, on losses from currency and stock trades.

The fund, managed by Mark Carhart and Raymond Iwanowski, lost a third of its value in 2007, according to an update sent to investors. Investors last month notified New York-based Goldman, the most profitable securities firm, that they plan to withdraw $1.6 billion, or almost a fifth of the fund's assets as of July 31.

Carhart and Iwanowski, who use mathematical formulas to select trades, may get more redemption notices as Global Alpha falls further behind quantitative managers such as James Simons's Renaissance Technologies Corp. Global Alpha generated $700 million in fees after an almost 40 percent gain in 2005. The fund has fallen 44 percent since its March 2006 peak.

"With losses this large, typically you have to look at the return potential going forward," said Gregory Dowling, vice president for alternative investments at Cincinnati-based Fund Evaluation Group LLC, which doesn't have money in the Goldman fund. "If there isn't a possibility of a snap-back, you have to examine where else you can put that capital."

Global Alpha's biggest loss in the month stemmed from the managers' decision to sell Japanese yen and buy Australian dollars. The so-called carry trade unraveled when the Australian dollar fell 6 percent against the yen in August. Equities holdings, including stocks in the U.S., Norway and Finland, declined 4.7 percent.

Fundamental Beliefs

"We still hold our fundamental investment beliefs that sound economic investment principles coupled with a disciplined quantitative approach can provide strong uncorrelated returns over time," Goldman said in the unsigned report.

Goldman blamed its losses on too many quantitative funds making the same trades, and said in mid-August it would have to develop new strategies.

Let's dissect the reasoning from Goldman.

1) Global Alpha is a Quant Fund
2) Quant funds are supposed to trade mathematical models that make money in any market.
3) Apparently Quant funds do not make money in any market if other Quant funds are executing the same strategy.
4) Goldman has not identified conditions in which it can determine if too many other Quant funds are executing the same strategy.
5) What's good for Goldman is only good for Goldman as long as too many others are not doing what Goldman is doing.

Perhaps Goldman should start an anti-quant fund. In other words figure out what other quant funds are doing and bet against them. That would make as much sense as what they are saying. Perhaps even more so. Certainly the Alpha Fund would be up enormously if it bet against its own model.

But let's focus on another statement: "We still hold our fundamental investment beliefs that sound economic investment principles coupled with a disciplined quantitative approach can provide strong uncorrelated returns over time."

Sound investment principles?

Hmmm.

"Global Alpha's biggest loss in the month stemmed from the managers' decision to sell Japanese yen and buy Australian dollars. The so-called carry trade unraveled when the Australian dollar fell 6 percent against the yen in August. Equities holdings, including stocks in the U.S., Norway and Finland, declined 4.7 percent."

That is interesting for multiple reasons.

1) Global Alpha is a Quant fund. It trades according to mathematical models. If it trades according to mathematical models exactly what was a manager doing by deciding to sell the Yen and buy Australian dollars? Or was the "manager" in question a computer program?
2) It seems that a 6 percent decline in the Australian dollar led to a 22.5 percent decline in Global Alpha.
3) Is this what Goldman calls "sound economic investment principles"?

Here is my take. There is no real analysis or brilliant programming going on in any of these quant funds. Every damn one of them is trained to but the dip in anything and everything (with enormous leverage). They are all trend followers that will fail spectacularly when the trend changes. Leverage guarantees it. Buying the dip works until it doesn't. When it doesn't it results in a tremendous number of blowups.

And speaking of blowups this is what I see.

As of now there are 156 Mortgage Lender blowups as depicted on the Mortgage Lender Implode-O-Meter. A few days ago it was at 149. Not that long ago it was at 49.
As of now the Hedge Fund Implode-O-Meter is sitting at 18. That number to is set to rocket, because as even Goldman admits "too many quantitative funds are making the same trades".

Given that Global Alpha is down 44 percent since its March 2006, and given that too many other funds are doing what it is doing inquiring minds just might be asking "How can Goldman get out of this mess?"

Here is the answer: They can't.

Since the problem is "too many doing the same thing" (and now every one of them is wrong) as soon as one Quant fund sells to get out, it will further depress prices for the remainder still in. In this situation, the first to liquidate could be the first to reduce losses. But they all fear the snapback rally so no fund wants to take action.

Thus all the Quant funds including Goldman are likely to hold on, hoping and praying for a reversal. In Goldman's case they need a reversal in the Yen. If a 6% decline led to a 22.5% loss it will not take long at that rate for Goldman's Alpha Fund to be worthless.

I have no stake in the matter but I am rooting for a continued rally in the Yen.

Mike Shedlock / Mish
http://globaleconomicanalysis.blogspot.com/