Les bêtes noires de Cliff Asness

Clifford Scott “Cliff” Asness est le président-fondateur de AQR (Applied Quantitative Research), l’un des fonds alternatifs/spéculatifs américains les plus sérieux. Je parlerai brièvement dans 5 de cet ancien assistant d’Eugène Fama (prix Nobel d’Economie 2013) à Chicago, parti créé Global Alpha chez Goldman Sachs dans les années 1980, avant de créer AQR en 1998, dans des conditions difficiles (LTCM venait tout juste de s’écrouler). Après avoir méchamment subi la crise de 2007 (le portefeuille d’AQR passant de 39 milliards à 17 milliards en 2008), AQR semble désormais plutôt en bonne forme puisque le fonds gère pas moins de 90 milliards de dollars en 2013 à l’aide de seulement 300 employés.

Comme Jean-Philippe Bouchaud, co-fondateur et co-dirigeant du fonds français Capital Fund Management (CFM), et au contraire d’autres institutions qui gardent jalousement le fruit de leurs recherches, Asness publie régulièrement sur le site d’AQR des articles scientifiques (ici), écrits de sa main ou de ses collaborateurs. Le style de Cliff Asness est souvent mordant, parfois “rentre dedans”, mais Asness est un esprit fin, qui tape souvent là où il faut.

“My Top 10 Peeves” (« Mes 10 bêtes noires ») est un nouvel article d’Asness à paraître dans le Financial Analysts Journal en janvier 2014, qui a fuité hier soir sur Twitter. Très intéressant article, notamment parce que l’une des bêtes noires qui agace Asness est le trading à haute fréquence, non pas en tant que tel mais en raison des polémiques qu’il suscite. Voilà quelques extraits, des arguments judicieux que l’on pourrait résumer ainsi : inutile de s’exciter plus que de nature sur le hft.

This peeve is about the rolling brouhaha over “high-frequency trading” (HFT); I believe it’s massively overwrought. HFT is mostly a good thing, not an evil conspiracy to crush Main Street. I should mention that my firm is quantitative and algorithmic but is not even close to being a high-frequency trader; we mostly have long-term views and tend to hold our positions for quite a while. But being quantitative and algorithmic is frequently, and incorrectly, mistaken for being high frequency.

HFT smashed the old dealer and exchange cartel (providing a much lower barrier to entry for competing market makers), democratized flow information, and replaced very expensive humans trading a handful of securities with very cheap machines trading a great many.

A new and advanced technology always cre- ates critics and predictors of all kinds of doom. I’ve heard HFT blamed for some bizarre things. I’ve heard it blamed for bubbles. How high-frequency traders who go home flat (close to no positions held) every day create bubbles is beyond me. I’ve heard it blamed for why some markets and strategies seem more correlated today than in the past (again, how this can be true is beyond me).

Once we see that market making is the core activity of HFT, we can also see that it is being compared with a mythical gentle giant of the past, the old-school market makers who allegedly often stopped crashes in their tracks by buying securities at prices they knew were way above current market prices. These heroic figures risked their own bankruptcy to save the financial world. Why do I call this a myth? Perhaps because it has happened zero times in financial history. It is not now and has never been the business of a market maker to go broke buying securities at the wrong price in a crashing market.

By the way, of course there have been glitches, and some were quite scary. That will happen with any new technology, and it will happen more when the system is complicated and organically grown in separate places, which it most certainly is now. But these glitches have actually been more about electronic trading than HFT (again, a longer expla- nation is needed, but these are far from the same), and you would have to turn the clock much further back to eliminate electronic trading in addition to HFT.

Finally, two specific elements of HFT raise concerns among some: the frequent use of cancel-and-correct orders and the speed at which transactions occur (it’s the only aspect of finance I know of where the speed of light matters; our field’s physics envy is finally bearing fruit!). But if you view high- frequency traders as mainly (not entirely) market makers, it is easy to see the reason for both of these aspects of HFT.

En fait, il faudrait citer entièrement la section consacrée aux hautes fréquences tant le propos est à la fois serein, détaché et très juste. L’article intégral est ici. A must read !

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s