Monday, November 05, 2012

Theory: “impact on the crowds”

Intrade allows people to bet on the election. Right now, Intrade gives Obama a 66% chance of winning vs. 33% for Romney. This is called the “wisdom of the crowds”. The idea is that if somebody has some private information, they’ll profit it from that information by going to Intrade and buying/selling shares, moving the price. Thus, the final price encapsulates everything known about the election. Thus, according to this theory, the best anybody knows about Obama winning the election is 66%. I use this theory in my previous post.
But that idea isn’t entirely correct.

If you are among the 1% of top earners, an Obama victory almost certainly means your taxes are going up even more. To compensate, you can do what’s known as a “hedge”. You buy shares on Intrade betting on an Obama victory (even if you think it unlikely), so that if Obama wins, you can use the winnings to offset the extra amount you’ll have to pay on taxes.

The same is true of other Intrade bets on things like the weather. Instead of betting, you could be hedging on the outcome. If bad weather will cause a crop loss, then buy Intrade shares betting on bad weather.

Thus, Intrade doesn’t necessarily reflect the “wisdom of the crowds”, but the “impact on the crowd”.

I suspect Intrade still reflects the “wisdom of the crowds” to some extent, because it’s volume is too low to hedge. I actually looked into this on the last election. I wanted to hedge against Obama’s rise in taxes (which turned out to be 8.9% for me), but I found the Intrade volume to be too low. Any realistic attempt at hedging would swing the price to the point it wasn’t worth it. Thus, the only people betting on Intrade are those betting for fun, not profit.

2 comments:

Anonymous said...

Paddy Power pays out £400,000 on Obama victory in U.S. Presidential election - http://blog.paddypower.com/2012/11/04/paddy-power-pays-out-400000-on-obama-victory-in-u-s-presidential-election/

Gary Myers said...

It also doesn't reflect exactly on the quality of the information, only the confidence and betting capacity of the information holder.
As a rough guide, someone with four Queens will bet high (and maybe all they can afford), and so will someone with Kings. But if the person with the Kings has less money or is more risk averse, the odds would point to the Queens