Selling delusion short
A few weeks ago, Intrade had Obama at 60%. Believers in the efficiency of prediction markets see the Intrade price as the best estimate of the probability that Obama will be elected. At 60%, I think that’s reasonable. Now Obama is at 85%, McCain at 15%. Is this still a reasonable prediction? Is there really a 15% chance that McCain will do something mavericky in the next two weeks and turn the campaign around?
We know that Intrade is inefficient at the extremes. The reasons are partly structural and partly psychological. For example, the “Hillary for President” contract held at 5% for a few months after Obama locked up enough delegates to win the nomination. Even after the convention, I was able to sell Hillary short at 3%.
I’m not interested in flipping a fair coin. I’m interested in the free lunches, the inefficiencies. If Obama-McCain is anywhere near even odds, I don’t want either side of the bet. However, this looks like a one-sided election, and at some point, McCain will be trading at X percent, where X mostly represents delusion. The closer the election gets, the less time for mavericky surprises, and the more delusional the long odds. The problem is recognizing X when I see it. Is X 15%? How delusional are Republicans? Ron Paul got close to 10% delusion. It’s hard to say, but somewhere between 5% and 10% I think I will short McCain. This stuff is always more obvious in retrospect.
