Intrade is an interesting site that the political junkies all love these days. Basically it lets you trade shares of real-world events based on whether or not you think they will happen. For example, right now there’s a 25% chance that any country using the Euro will drop it before the end of the year. If you think there’s a bigger chance, you can buy shares for $2.50, and if it actually happens, you get four times the money when the contract closes. If it doesn’t happen, you lose it all.
It’s like a cross between stock trading and gambling, but it’s interesting because you can “put your money where your mouth is” based on your bias about what’s going to happen in the world, and as lots of people do this we can get an interesting snapshot in the overall “conventional wisdom” by where the percentage is holding at any given time.
Anyway, last night before Super Tuesday election results rolled started rolling in, Romney was an 85% favorite on Intrade to win Ohio. Poll junkie @fivethirtyeight thought this was way too high considering the close polling: “Continue not to understand why Intrade has Romney at 85% to win OH. A ~2 point lead in the polls simply isn’t 85% safe.” Of course, this prompted the usual discussion about how Intrade is not a very reliable prediction market because it’s not used by enough people and markets are unreliable anyway, etc, etc.
The contract even approached 90%. I toyed with the idea of finally creating an account and buying a few Santorum shares. It looked like it was gonna be a close race, and if Santorum won, you could put in $10 and make $100. But I didn’t.
Then the results started coming in. Santorum broke into an early, close lead. By the time half the results were in he was leading by several thousand votes. People started selling Romney shares and he dropped to 70%, then 50%. I believe there was even a point, with almost 2/3 of the results in, where Santorum’s lead had consistently grown to 15,000 votes and his Intrade shares for Ohio were up to 60-70%. I almost regretted not putting money down on such an obvious misalignment.
It was looking like Intrade traders had been hopelessly wrong about Romney’s 90%. This prompted some liberal mocking of conservative beliefs in efficient markets along with some pro-market mocking of their mocking: “Intrade fails to perfectly predict highly uncertain events. Market failure!” Left-leaning economist Justin Wolfers offered consolation: “I’ll concede prediction markets aren’t perfect. But the question is what alternatives are less imperfect. So far, I’m yet to find ’em.”
But everyone tweeted too soon. Apparently most of the results were in from rural areas, where Santorum had done best, and there were still a lot of results coming in from large urban areas where Romney had done best. Santorum’s lead dropped to around 10,000 and before long Romney’s contract had surged back to 90%. As the urban precincts finished reporting, Romney broke into a lead and won by over 12,000 votes.
Now I was glad I hadn’t done anything so financially rash. In hindsight it would have been best to buy Santorum at 10% and sell at 50% or better for at least a five-fold gain, but at that point it looked like he was going to end with a win and I’m sure I wouldn’t have been clever enough to sell, holding on for the hope of 100% that actually ended at 0.
But what does all of this mean for last night’s handwringing about Intrade? Did the people who originally traded Romney up to 90% know something that made them more confident than the close polling, or did they just get lucky? I don’t know. But I do know that the pundits who were ready to wax prolific last night about why Intrade was wrong now need to wax about why it was right.
The results also don’t hurt the notion that Intrade is a useful view of the conventional wisdom. These markets still give Obama a 60% chance of being re-elected and Republicans a 60% chance of controlling the House and Senate – that’s optimistic for hopes of divided government.
I’ll probably continue to hold off putting any money into Intrade. But last night’s outcomes make Intrade look at least as useful as the opinions of arrogant pundits, if not more. I will continue to watch it with a cautious eye.