Tuesday, October 26, 2010

Putting Intrade to the test

Intrade is a kind of stock market for predictions. You can bet real money on all kinds of things, including election results. It's much more objective than a poll or a pundits analysis. When people are asked their opinion or even for their objective analysis, they have a tendency to provide answers and insights that favor what they hope will happen, probably in hopes that they might influence someone in the audience.

With Intrade, your betting real money, anonymously. There's no audience to influence. You're not stating what you'd like to happen. You're making a statement about what you believe will happen. That doesn't make it perfect, but it does make it much more objective. How accurate is it? Let's find out. Note: if you visit the website, keep in mind the market prices don't reflect the predicted margin of victory, just the likelihood of victory. If 90% of investors believe a candidate will win by 4%, you're going to see a 90 - 10 (roughly) difference in price between the candidates.

Here's how Intrade sees some key, relatively close races: (I'll check again on Nov 1, then see how predictions jive with results after the elections.)

Note: I'm posting what I believe is the better known candidate in each of the races below and whether they win or lose.

Patty Murray wins (Washington Senate)
Russ Fiengold loses (Wisconsin Senate)
Brown wins (CA Gov)
Boxer wins (CA Senate)
Buck wins (CO Senate)
Rubio wins (FL Senate)
Miller wins (AK Senate)
Paul wins (KY Senate)
Harry Ried loses (Nevada Senate)
Toomey wins (PA Senate)
Hickenlooper wins (CO Gov)

That's a pretty good sampling I think. I should point out that personally, I don't like this as an investment vehicle because the spreads are quite wide on most contracts. Still, it could be a good forecasting tool.

No comments: