A recent poll from August 20th shows McCain actually beating Obama by five points, which is not insubstantial, considering that McCain has been trailing Obama in the polls since his nomination was locked in. Now, this should be taken with a grain of salt, considering that according to the BBC the two are in a virtual dead-heat; most other polls show a statistically insignificant difference between the two. Considering that McCain's fundraising has of late been outpacing Obama's, the polls make some sense. What I find puzzling about all this is that the prediction market intrade.com still shows a substantial difference between the two. As of 6:15PM Pacific Standard Time, Obama was trading at 58.4, and McCain at 36.1. If you do the math, that gives Obama a more-than 20-point lead over McCain, a decidedly different picture from that painted by the polls.
What are the reasons for this discrepancy between what the financial markets and the poll market are saying?
I can think of a couple of reasons offhand, all of which may bear some relevance. Most obvious is that the poll was presumably conducted using a sample made up of US citizens whereas the financial markets are nearly free for anyone in the world to use. If a disproportionate number of foreigners possess an illogical amount of faith that Obama will win the election, this could account for the price-probability discrepancy between the two candidates.
While this seems a nice way of explaining the discrepancy, I find it lacking. Financial markets know more - a whole lot more - than any one, or ten, or one-hundred of us. My intuition is to believe that they know something unique. Now, the poll, (a joint Zogby/Reuters venture), is certainly statistically sound, and the margin of error (+/- 3 percentage points) is less than the difference between the candidates. But does it really make sense that the market knows more about how we will vote than we do?
Not to me, at least. It could be that the financial markets haven't had time to adjust yet, (which is itself a somewhat dubious proposition). I'll try to think of some possible explanations over the next few days...
No comments:
Post a Comment