The global economic meltdown as a failure of modeling

March 2, 2009

Like most people, I’ve recently developed an intense interest in the workings of the interconnected systems known as the global economy. One thing I’ve learned about it: it’s really big. And as this 2008 New Yorker interview with Ben Bernanke makes clear, no one, not even the chair of the Fed, exactly knows how it works. Bernanke has as much information as anyone, possibly more, and even so has to fall back on best guesses. His models are incomplete.

Wired has a fascinating though completely depressing article describing the current global economic crisis in terms of bad modeling. Is a mathematical formula to blame for the current collapse? That’s the article’s premise, and it makes a good case. The formula in question is the Gaussian copula model, the logic behind credit default swaps. It models default correlation, that is, the probability that a mortgage will be defaulted on, given a default on a related mortgage. This is a hard modeling problem, and the formula cleverly (or so it seemed at the time) begged that problem by relying on the market’s beliefs about risk rather than predicting risk based on past history. In short, it reduced complex and poorly understood interactions to a single, unchanging number. So of course, everyone loved it:

The effect on the securitization market was electric. Armed with Li’s formula, Wall Street’s quants saw a new world of possibilities. And the first thing they did was start creating a huge number of brand-new triple-A securities. Using Li’s copula approach meant that ratings agencies like Moody’s—or anybody wanting to model the risk of a tranche—no longer needed to puzzle over the underlying securities. All they needed was that correlation number, and out would come a rating telling them how safe or risky the tranche was.

As a result, just about anything could be bundled and turned into a triple-A bond—corporate bonds, bank loans, mortgage-backed securities, whatever you liked. The consequent pools were often known as collateralized debt obligations, or CDOs. You could tranche that pool and create a triple-A security even if none of the components were themselves triple-A. You could even take lower-rated tranches of other CDOs, put them in a pool, and tranche them—an instrument known as a CDO-squared, which at that point was so far removed from any actual underlying bond or loan or mortgage that no one really had a clue what it included. But it didn’t matter. All you needed was Li’s copula function.

Of course, we can’t really blame the formula, or even its creator David X. Li. Li’s assumptions were faulty, but he’s not responsible for the enthusiastic and widespread misuse of his model. That came from pure greed, which is no doubt missing from most economic models.

I’m an uninformed armchair economist. The niece of two accomplished economists, I’ve somehow managed to avoid any formal study of the field. So my critique is no doubt ill-informed, and maybe someone (uncles?) will be moved to correct it in a comment. That said, I’ll plunge boldly ahead. Several of the fundamental assumptions of economics strike me as naive (the rational actor and perfect markets to name two). Economic models seem to consistently draw strange lines around themselves, notoriously failing to perceive value where price does not exist. Like Li’s copula function, economic models tend to stub in simplistic stand-ins for complex interactions. This may work well enough some of the time, but clearly has its limits.

Some people are working on ways to “complex up” economic models. Behavioral economics and cognitive economics reject the notion of the rational actor and investigate cognitive and neurological bases of financial decision making.  On the squishier side of things, triple bottom line accounting attempts to incorporate social and environmental costs into the evaluation of corporate or national economic success. Better models can’t prevent greed, bubbles, or the periodic downturns of market capitalism, but we can hope they’ll lead to better understanding of the overall system.

Advertisements

One Response to “The global economic meltdown as a failure of modeling”


  1. […] the comments to find some good points about modeling and the history of analogue computing. Given my feelings on the the limits of economic models, I especially appreciated this one on the limits of modeling. In a less negative vein, this […]


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

%d bloggers like this: