Wednesday, December 4, 2013

The Poverty of Behavioral Economics

With the awards of the recent Nobel Prizes in Economics to Eugene Fama, Lars Peter Hansen and Robert Shiller, the whole discussion about market efficiency and behavioral economics surfaced anew in the press.  Professor Fama's efficient market theories are said to be discredited, according to the financial press.  The evidence?  The last financial crisis and its aftermath.  Never mind that the supposed evidence is weakly related, if at all, to the theory.

Professor Shiller, on the other hand, is suddenly lionized as a foil to Professor Fama for being a 'behavioral economist.'  Having viewed Shiller's "Financial Economics" class that he teaches at Yale online, I'm very puzzled, as most of that course is a nice exposition of modern portfolio theory, of which Fama is a father along with several others.

In the investments class I taught to upper level finance MBAs at the University of St. Thomas, there was a small discussion from Shiller about how markets can deviate from efficient equilibria due to something he describes as "noise trading."  Since this was a counterargument to the mainline theory from an economist of quality and stature, I dutifully taught it in our discussions.

The problem?  It sounds good, and it intuitively matches our ex-post experience of market runs and crashes.  However, there's no good model for how fundamental traders and noise traders behave, and therefore there is no model to test and no data.  It's an interesting idea, but empty.  It's a metaphor, but little else.

Recently, I received two different kinds of material from Chicago Booth School of Business.  One was an informative and thoughtful piece by Professor John Cochrane, one of my favorite researchers, on why Gene Fama was awarded the Nobel Prize. If you'd like to understand the theory apart from its facile characterization in the press, have a read, linked above.

By contrast, I also received a link to a video in which three 'behavioral' researchers from different disciplines at Chicago Booth expound findings from their latest research.  The research results seemed either blindingly obvious, somewhat puzzling, or downright counterexperiential.  These are all very smart people, but the discussion betrays the real poverty of behavioral economics.

Going into their relatively small experiments, there is no theoretical model from which they can measure their actual results compared to the expected results from their theories.  They find some correlation or trend, give it a name, and say that the result is "quite surprising."  Why?  What exactly did you expect?  I was really interested to know about the relative effectiveness of intrinsic versus extrinsic rewards in the performance of marathoners. I don't know anything interesting, thought provoking or useful after this discussion.

No comments: