Wednesday, April 28, 2010

Blankfein's Apha Deception


Loyd Blankfein was grilled by the Senate yesterday, and he highlighted one of the oldest tricks in the alpha deception handbook. Don't admit to being a merely a middleman, because that's too transparent. That's a skill to be sure, but something many people can do, and most don't make $10MM a year doing it. Instead, say, you are 'transferring risk', 'taking risk', 'selling risk'. If you buy and sell investments, this is technically true.

But it's hugely misleading. Senator Claire McCaskill better characterized Goldman as a bookie whose main job is setting a line so they aren't taking a position on the outcome, their customers are, just in offsetting ways. Making markets is first and foremost about pricing (setting the line), secondarily about hedging, and finally about how the residual risk agregates up. If you price correctly, the other two are de minimus.

But as I explain in my Finding Alpha book, one reason why 'risk' remains prominent in academic finance though it has never been identified precisely is that it has the ability to rationalize a lot of useful deception. Risk is presumably the most important thing in finance, its essence. But what is 'risk'? That depends: it could be beta, a regression coefficient with the aggregate market; it could be volatility, or the correlation with the wealth-to-income ratio. Bill Sharpe, one of the founders of the Capital Asset Pricing Model, now prefers a 12 factor model of risk that totally obviates his Nobel Prize winning insight, though no one seems to note the inconsistency.

Risk is that which, in combination with an expected return, defines how much of some asset one wants to buy. Now, because an expected return is a straightforward concept, alpha deceptors don't like to go there. Instead, they talk about the risk that could rationalize any desire for any expected return. And since 'risk' has both an insanely rigorous definition (covariance with the Stochastic Discount Factor!), and also has a subjective, intuitive definition, it allows one to say nonsense and get away with it.

It's sort of like post-modern gibberish where one can talk about 'Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity', and get smart people to think you, too, are smart, because it uses profound but amorphous words in a plausible sequence.

So, when alpha deceptors hide behind the 'we are risk managers' defense, remember, a real risk manager has a prosaic job, doing things that one can understand: verifying income on loan applications, measuring CAPM betas, calculating VaRs even. They are straightforward, and you can argue about key assumptions. The whole 'risk manager' spiel is because if you are getting paid $1MM+ a year, you know that there's probably someone just as smart as you making half that who wants your job, so better make it sound like you are doing financial string theory. As 'risk' is essential, important, and undefinable, Blankfein and Sparks could talk about 'selling risk' and keep senators at bay.

A financial market middleman's job is difficult, and they have an essential and salutary function for society, but all the 'selling risk' stuff was obfuscatory: they sold Paulson short CDO exposure because he was bearish, so saying they were 'selling risk' was simply a way to avoid talking about it.

Goldman didn't do a good job defending themselves. You have to be a 'bright young thing' to get a job there, but I think their brand name and massive connections, make them appear much smarter than they really are. They could have mentioned that for most things they make markets in, it would be impossible to keep a consistent position (eg, long housing), because you have hedges, people bet not just on the direction but the volatility, you might believe a certain subset of a broad sector is bad, or you might change your mind every week. Also, in a large institution, you probably have certain groups with opposite views, and let the magic of their pnl determine who is right: they are net zero on the 'beta' bet, but one might demonstrate 'alpha', so your organization wins. Further, every asset, even 'shitty' ones, is attractive at the right price. A regulation on position direction is antithetical to the essence of being a market maker, who occasionally ends up with inventory long or short, which then needs to be exited.

Blankfein is a crony capitalist, begging for more 'regulation' because he knows that a 1300 page bill basically only helps those with connections and extant massive legal infrastructure, and hurts potential competitors who merely have good ideas. I'd miss them as much as I'd miss Drexel, a giant which died with much schadenfreude.

2 comments:

Anonymous said...

Eric
You have been updating your blog far more than you had indicated to keep up with the events presumably.

Absolutely enjoy your blog and the thoughtful comments

J said...

Complex regulation generally favors established firms and increases the barriers to new would-be entrants. It also engourages completely new ways of doing business that are not covered by the regulation. I think that new irregulated financial instruments will be invented.