Goldman Sachs Group Inc. blamed its $3 billion in hedge-fund losses this month on too many quantitative managers making the same trades and said it needed to develop new investing strategies.
[MiSh comment: Too many Quant managers making the same trades? Uh... Wasn't that OK when momentum was flowing the other way? By the way, how is it that they were all making the same trades anyway, given that a quant fund is completely driven by computer programs". Is every quant fund using the same program?]
"Longer term, successful quant managers will have to rely more on unique factors," New York-based Goldman's fund- management division said this week 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."
[MiSh comment: In essence Goldman is saying they need to develop an "anti-quant" program. Why bother? Why not take the existing program you have and bet against it? Hmm . What happens if all the quant managers do the same thing? ]
The world's most profitable securities firm and second- largest hedge-fund manager lost $1.4 billion, or 28 percent, in its Global Equity Opportunities fund this month as computer- driven models selected losing trades. Its largest hedge fund, Global Alpha, fell 27 percent this year through last week. Goldman waived fees to attract clients to the smaller fund and said Aug. 13 it's investing $2 billion of its own money as well.
Last week, "quantitative portfolios fell (and rose) by unprecedented amounts -- far more than at any time in the history of our data," according to the Goldman report.
Goldman's quantitative funds use six major investment "themes" that it identified as momentum, earnings quality, valuation, profitability, analyst sentiment and management impact. All suffered "extreme negative returns" from July 30 through Aug. 10 in the U.S., Japan and Europe, according to the report.
[MiSh comment. Goldman is programming in "analyst sentiment"? Earnings quality? Valuation? There is not much earnings quality anywhere, especially in financials. I would be curious to see exactly what Goldman's model says about "earnings quality" in the financial sector. I bet it differs vastly from what I think and what the market has done recently. ]
When adjusted for volatility, last week's losses at Goldman's funds were five times greater than any another prior period, the report says.
"The magnitude of this move alone argues that fundamental economic news could not be responsible," according to Goldman. "Our conclusion is that there was a major supply/demand imbalance caused by many quant managers unwinding simultaneously, which caused outsized negative" returns.
[MiSh comment: The fundamental economic news could not be responsible eh? GDP has fallen like a rock (with one minor and short lived rebound that will be revised lower). 120 mortgage lenders have blown up. Consumers are tapped out and cutting back on purchases. Walmart is struggling and cut prices on 16,000 items smack in the face of higher energy costs. (Please see Price Wars for further discussion). Capex is slowing. Weekly jobless claims are ticking up and recent jobs reports from the BLS look like they are from Mars (See Fed questions the BLS jobs model and Employment on Pluto Rises for more discussion about jobs). Foreclosures are soaring, and a huge number of ARMs need to be reset but can't because of rising default risk. I would be curious to see what fundamentals Goldman is looking at. Whatever it is, it's surely wrong. Poor assumptions will eventually produce poor results. It's no wonder the losses were big, especially if a dozen other quant funds were all doing the similar things and all had to bail.]
Goldman's quant funds, overseen by Mark Carhart and Raymond Iwanowski, have "good investment opportunities in the near term," according to the report. "However, substantial risks still accompany these opportunities."
Current securities prices don't reflect what they're worth, Goldman said in the report. "If so, the current environment may represent a significant investment opportunity."
Of all the things Goldman said, that last paragraph is arguably the most dangerous. Goldman has been reduced to "arguing with the market". We saw the same thing with the dotcom bust in 2000-2001. Stocks looked cheap simply because they were even more insanely overpriced before.
Clearly Goldman does not see the risks, and if it does not see the risks how the heck are they going to get their model correct? And what about that $250 trillion in derivatives floating around? Is there not massive risk because of that? Is Goldman completely ignoring the possibility of a huge consumer led recession? Is Goldman listening to Paulson chirp about how "strong" the global economy is? What about stocks reverting to the mean? It happened in housing, it's now starting to happen with stocks. How long did Goldman's model think we could go without a 10% market correction?
Momentum suggests it's now time to sell the rallies not buy the dips. Missing the turn is very expensive when done with leverage. And not only did Goldman miss the turn in momentum big time, it missed an even bigger turn in market psychology. Genius Fails Again. Goldman's answer to all of this is to create a new model based on what everyone else's model is not doing. This suggests Goldman does not really have a clue as to what happened, what to do about it, or even what the risks are.