Twilight of the Quants (or yet another reason for the current financial turmoil)

Over the last few months, I've tried to offer some analysis about the recent market turmoil that I feel has not been fully understood by the financial press and so-called expert commentators (e.g., the role of credit derivatives in the current crisis). It turns out that there is another contributing factor to the recent turmoil in the markets: quantitative hedge funds.

The most recent Buttonwood column in The Economist (25 Oct. 2007), sheds light on how 'quants' have been adding risk to the market because of their strategies. These quant funds are staffed by mathematicians, physicists, etc., that are suppose to come up with ultra-sophisticated models that should, they promise, lead to higher returns. But the problem is that other funds also hire quants that use similar models. This leads to a kind of feeding frenzy (or a vicious cycle) where an increasing number of very smart people are chasing ever shrinking opportunities for genuine arbitrage profits, which causes them to either ratchet up risks and/or try ever more esoteric strategies (which are copied almost as soon as there is any profit to be made), which, in turn, lead them further down the slippery slope.

For that reason, as well as others, these funds -- just as recently foreclosed home owners and their subprime mortgage lenders (who, presumably, didn't do PhDs in maths, physics, or engineering) -- loaded up on leverage (either directly or indirectly via derivatives, etc.). When the markets turned against the quants, they tried to offload their now toxic investments (in credit derivatives, etc.). Of course they couldn't find buyers, so they had to unload more liquid and valuable investments in order to meet margin calls and/or stay in business. But all this did was to cause a stampede among quant funds to get rid of assets that they really shouldn't have gotten rid of but HAD TO because of margin calls, the need to delta hedge their positions, plain desperation, etc.

This is the story of so many financial bubbles bursting -- except on a high tech, post-modern level. This, according to Buttonwood, is what happened during August. It seems that no one on Wall Street or in the City has a memory beyond their last bonus. Didn't we go through this with LTCM? Program trading in October 1987? Yet we are told that ever more sophisticated models or ever faster execution of orders will lead to greater returns. But what these 'geniuses' don't realize is that the world doesn't always fit their 'sophisticated' models. Common sense should have told them that strategies and assets that weren't correlated before will become correlated once everyone else with a PhD in Computational Chemistry decides to sell similar things at the same time! The rules of the game are the same whether you are a down-on-your-luck homeowner or an over-educated quant.

But to be fair, quants shouldn't shoulder all the blame. Yes, it's true that, as the article points out, "instead of providing liquidity in a crisis, the quants added to the instability." But the same could be said for those who participated in the dotcom bubble, tulip-mania, the Roaring 20s, and a whole host of financial fads and fashions that went awry. The problem is not that we are too sophisticated or that we are not sophisticated enough; the problem is arrogance. The problem is being willfully blind to 'Black Swans.' I doubt that any amount of rigorous modelling, algorithmic trading, statistical arbitrage, or improvements in order executions will solve that problem.

Why some people won't stop teaching others how to lose money (or Still Getting Fooled by Randomness)

Apologies for not updating more frequently. I have been trying to adjust to life in the U.K., so I haven't had much free time.

I wanted to put up a quick blog post about an interesting thing I experienced last week. I went to a presentation by a prominent finance academic (he holds posts in some top programs ... programmes in British English ... in finance, and has written a well known book about quantitative investing). He presented the results of a theoretical paper he wrote extending some of the standard neoclassical financial models in what I hoped would be interesting directions. Instead, this chap seemed to be making some of the same mistakes -- and, in some cases, introducing new errors into already flawed models -- that had been made before by some of the (unfortunately) leading figures in quantitative finance.

I politely asked him how his model related to the types of difficulties we have just faced (i.e., the credit crisis, real estate bubble bursting, Northern Rock, etc.) -- and his model, as advertised, should have had something of substance to say about it. Sadly, he and his cohorts (mildly) dismissed my musings. Their rationale was that their simplistic linear model couldn't cope with the non-linearity of the way real financial markets work (aside: actually, even based on linear type maths -- vector spaces, independence, etc. -- I and a few others doubted that the presenter's model was valid). Rhetorical questions: Isn't that the point? Why claim that you have a model of how financial markets should work when it doesn't and probably won't work that way?

I wanted to respond by saying something like: "Sorry for bringing in a dose of reality and spoiling the fun of building models of fantasy worlds." But, prudently, I didn't say anything of the sort.

But what this incident goes to show is that -- despite a cold slap to the face by the true nature of uncertainty (not continuous, but extremely discrete; not linear, but non-linear) by events that are barely a month old -- there are still people teaching finance to impressionable minds that refuse to incorporate some degree of humility and reality into a subject that can use quite a lot of both.