Prediction Markets on the Veep Race
Here’s The New York Times shortlist for the vice president. Here’s a similar list from the Wall Street Journal. But there are some bigger differences of opinion with political prediction markets, which I reviewed in my latest Wall Street Journal column. Let me give you the highlights.
First, on the Democratic side:
“… prediction-market traders are moving sharply towards a new favorite. Virginia Governor Tim Kaine, who rated only one-in-ten odds a week ago, is now given a one-in-three chance of winning the number two slot … Beyond Mr. Kaine, Mr. Bayh is the second favorite with a one-in-four chance, followed by Ms. Sebelius with a one-in-six chance, and Mr. Biden with a one-in-eight chance.”
Interestingly there is much less trade on the Republican side. Nonetheless:
Each man is rated around a one-in-three chance to win the nomination … traders believe Tom Ridge, the former Secretary of Homeland Security, and Alaska governor Sarah Palin to be the next most likely contenders.”
How seriously should we take these markets? After all, we are trying to guess the decision of an individual politician, rather than the decisions of the entire electorate.
“This can change the role played by the market. This column has often characterized prediction markets as aggregating widely dispersed local information. Instead, the VP markets are likely parsing publicly-available information, and pundits may be equally as adept at this task.
If there is the possibility that political insiders are betting, this further complicates things, as no trader wants to be on the losing end of a bet with a true insider. This anxiety, in turn, makes traders particularly sensitive to any new activity, as the fear of being beaten by an insider can lead informed speculation to be mistaken for inside information. In such an atmosphere, prices can move sharply even when only small amounts are bet.”
I’m willing to give betting advice, informed by a combination of history and behavioral economics:
“It is worth recalling the historical tendency for campaigns to announce surprising vice-presidential candidates, such as George W. Bush choosing the head of his search team, Dick Cheney, as his nominee, or his father’s selection of Dan Quayle. While John Edwards was a fairly predictable Democratic pick in 2004, Joe Lieberman was a surprise in 2000.
Compare this history of surprises with the fact that markets believe that they have narrowed the Democratic and Republican shortlists to four candidates — and, in each case, there is greater than an 80 percent chance that this shortlist contains the eventual nominee. Given the psychological evidence that traders are often overconfident, I think there is good reason to take these quantitative assessments with a grain of salt.”
Even though I think the markets have the candidates in the right order, I would still be shorting the favorites, as it remains likely that someone pulls a rabbit out of a hat.