/* ---- Google Analytics Code Below */

Wednesday, May 13, 2009

A Bonfire of the Economic Models?

Why did most academic economists fail to see the sharp downturn coming? In Knowledge@Wharton: Why Economists Failed to Predict the Financial Crisis?. This presents an overview of what modelers considered when trying to understand the economy.

From my perspective ... either a model is incorrect in describing the first principle operations of the economy. Or there are aspects of the economy that were not included in the model to begin with. Note that the second case is just a form of the first. Again, check out game developer/modeler Ken Karakotsios economic modeling example.

' ... A sense that they failed to see the financial crisis brewing has led to soul searching among many economists. While some did warn that home prices were forming a bubble, others confess to a widespread failure to predict the damage the bubble would cause when it burst. Some economists are harsher, arguing that a free-market bias in the profession, coupled with outmoded and simplistic analytical tools, blinded many of their colleagues to the danger. A recent paper from a conference of economists calls for changes in the way they are trained....

1 comment:

sking said...

I don't like to admit it in public:), but I started my career as an economist doing economic forecasts using econometric models.

The firm I worked for, DRI, had some of the world's best and smartest economists working for it. We also had massive computing power and huge economic databases.

And our forecasts were consistently wrong.

The reasons were: (1) the models were too simplistic; and (2) the models required inputs to drive the forecasts.

These inputs were based on human judgement. And we proved over and over again that humans are not good at predicting the future - even with the help of very sophisticated models.

I left the forecasting field for many years and interestingly enough I have returned (although not as an economist).

And guess what - our models are still too simplistic and humans still can't predict the future.

Oh well....