When good quants — go bad!

Apologies for not being present here for a while, but it has been a busy couple of weeks at work and I have been fighting to keep my head above water while my colleague lounges about on holiday, leaving me to navigate an absolute pig of a market all by myself. Of course, I have succeeded brilliantly, for now at least, leaving our competitors (including the various Fidelity, Goldmans tech funds) coughing in our dust and further cementing my reputation as the greatest investment mind of my, or possibly any, generation.

This will be my last post on the subject of Quants for a while, but I saw a couple of headlines the past few days which struck me as sort of telling. You may remember that my thesis for some time has been that Quant funds’ dominance ended in the horrendous “style squeeze” of August. Their ecological niche will be filled by fundamental stockpickers, like me.

So I was not overly surprised to see first this, from Yves Smith. Goldmans Quant funds are seeing withdrawals:

“Goldman Sachs Group Inc.’s Global Alpha hedge fund may lose about $6 billion in assets this year, a 60 percent decline, because of trades that went awry and client withdrawals, according to two investors.”

A letter sent by GS to cross investors in Global Alpha (which lost 2-3% in September as well) said the fund “would constrain its borrowing in the future, one of the factors that led to the problems last month, and consider the level of borrowing as a separate risk factor”. As Khatami and Lo point out, assuming mean reversion Quant trading stays as crowded a trade as it has been, reducing leverage is a recipe for radically lower returns. Unless these guys can come up with something new they are beginning to smell like toast.

And then this, from the FT this morning. Goldmans has been reading Baruch! 

Goldman Sachs is to raise $4bn-$6bn for a new hedge fund as the investment bank tries to rebuild its reputation in the hedge fund business, it has told potential investors. The new fund, run by bank partners Raanan Agus, former head of the proprietary trading desk, and Kenneth Eberts, former head of the US prop desk, will be the first from the bank to focus on picking shares, rather than using computerised, or quantitative, approaches.

If scurrilous gossip is true, and it normally is, I do not think things have got much better for the Quants, some of whom apparently decided to replace their moribund mean reversion strategies after August with factor-based, directional strategies which relied excessively on momentum — this is technically known as the “Buy Stuff That Has Gone Up a Lot And Cross Fingers” Strategy. This probably means they switched their computers off to save on overhead and went long Apple, RIMM and Google. This is the staple trade of underperforming managers everywhere at the end of every year, and works more often than not (ask Bill Miller). Just not this year: RIMM is down 20% from its October high. “Doh!” is the technical term we experts use when this sort of thing happens. 


1 thought on “When good quants — go bad!”

  1. I won’t claim to know the details of other Quants’ strategies, but many of the funds have long been doing valuation- and momentum-driven strategies. Those very profitable pair-trade/mean-reversion tactics (profitable, that is, until they are very not) might not have been that large a fraction of the huge buildup of quant strategies, both long/short and long-only, over the past few years.

Comments are closed.