October 20, 2004 - 2:49 PM CDT
the other Shu
I also spotted the Bainbridge article today, and I have some mixed feelings about it. On the one hand, he's absolutely right that this debate matters a lot. As he, and the original WSJ article point out, assumptions about whether actors in markets are rational have large effects on things like public policy - take for example the major election issue of privatizing social security. Support for privatization should be sensitive to whether you think investors are rational about where they put that money. (Thaler has nice evidence that they are not.)
Bainbridge also nicely points out, again in concert with the WSJ piece, the issue that even if we can find the behavioral anomalies, it's not so easy to take advantage of them as an investor. Holding the index still makes the best long-run sense. Fuller-Thaler has figured out how to profit from individual biases, but their strategy is not easily copied by the individual investor. I have serious doubts about other firms or financial advisors that claim to use "behavioral finance" strategies for investing. Some try to chase the previously documented anomolies, which is a losing game since the anomolies often go away once they're identified (as Bainbridge notes). Others do "behavioral profiles" of the individual investors and call it behavioral finance. That's not the behavioral finance that I know.
Where I disagree the most with Bainbridge is at the end when he states that since regulators, legislators, and bureaucrates also suffer from bias (which they surely do), it makes no sense to adapt legislation to address individual biases. This makes an assumption of a central planner who completely supercedes the free market. Being well trained in the history of Chicago economics, I'm not much of a fan of central plannners. However, I do think that biases need to be taken into account when options are defined. For example, which option should be the default? How should the options be described (framed)? How many options should there be, and what is the effect of too many versus too few? Sunstein and Thaler have a wonderful theory called libertarian paternalism... it suggests that we can still allow individuals to make all their own choices (the libertarian aspect) but we can frame those choices in a way that helps them make sensible decisions (the paternalistic part). So Bainbridge's (and Arlen's) argument that the behavioral approach "loops back upon itself" is incorrect. Many behavioralists still argue for free markets - they just encourage the market organizers to frame the choices in a way sensitive to what we know about bias.