Monday, August 10, 2009

Bounded Rationality piquing my interest

I was reading an article this weekend talking about the microfoundations of macroeconomic theories having series structural issues. The author maintained that uber-rational homo economics does not exist, and that people do not have perfect foresight, or even optimization skills. The author invoked bounded rationality. This led me to think about a book I haven't thought of in years, Max Bazerman's "Judgement in Managerial Decision Making." I recommend the book; it is very accessible. There are plenty of surveys and games that you can play with others to see just how limited our cognitive functions are, and it is, dare I say, fun.

Basically Bazerman theorizes that humans have two thinking states: the 1st is based on simple rules of thumb and allow you to make quick decisions. These rules are created by you either through your own experience or training. The second type of thinking is the rational version of deliberate debate, weighing the pros and cons of an action on several measures, such as: defining the problem, creating alternatives, weighing trade-offs. The reason humans have these two systems is that if we had to base every decision with the superior method of system two we would take a disproportionate amount of time choosing between different brands of Frosted Flakes.

Here is a quick example of what Bazerman terms an availability heuristic. You receive a newsletter in the mail and it states that the stock market will go up over the next 6 months. You ignore it. 6 months hence the market has gone up and you receive another letter that states that market will go down over the next half year. 6 months hence the market is down and you receive another letter stating that the market will go up. This time you consider the letter for awhile, maybe it sits on the coffee table for a week, but you still throw it away. 6 months go by and the market is up and you receive another letter stating that the market will go up again. This time you thoroughly read the letter and send your money into the broker who has been sending you the letter for two years. You never hear another word and lose 10,000 dollars. What happened?

Well imagine instead you are this deceptive broker. You have a list of 100,000 people. Every 6 months you send out the letter to everyone, except that only half get the market goes up letter. The other half get the market goes down letter. So in the first half only 50,000 are correct. Then 25,000. Then 12,500. Finally, the last letter 6,250. The broker gets 10,000 from each he walks away with 62,500,000 dollars. Not bad for two years work. This is also what people call survivorship bias, which happens in the mutual fund industry. Not that I am calling them crooks, but when you read stats like 90% of our funds have been up every year, well that is with the caveat that the poor performing funds are dead. Thus, no longer in the reported data.

No comments:

Post a Comment