Nov. 3, 2008 at 7:09 PM ET
Kim Carney / msnbc.com / IEM
This chart shows rising levels for Barack Obama's stock (and falling levels for John
McCain's stock) on the Iowa Electronic Markets' winner-take-all market.
Even though GOP presidential candidate John McCain insists he's catching up to Democratic rival Barack Obama, the gap on the political prediction markets is widening on the final day of the campaign.
McCain's shares fell to about 10 cents on the Iowa Electronic Markets, and to even lower levels on other markets. That means buying his stock now would reap a tenfold payoff ... if he wins, that is.
Over the past 20 years, studies have shown that prediction markets do a better job than polls when it comes to anticipating the winners of presidential contests.
Such markets let "investors" put down play money or, in some cases, real money on the prospects for parties or individual candidates. For example, today you can buy Obama's stock at 90 cents on the IEM's winner-take-all market. If he wins, you get $1 per share. But if McCain pulls off an upset, you get nothing. That upset would make those 10-cent McCain shares worth $1 each - which provides a choice buying opportunity for the contrarian investor.
"McCain still has a 10 to 12 percent probability of winning - that is better than zero," said Tom Snee, a spokesman for the University of Iowa's Tippie College of Business, which operates the IEM as a research project. "If you buy some McCain now, you could make some serious change."
Throughout the final weeks of the campaign, we'vebeentracking five winner-take-all prediction markets dealing with real money (Betfair, Intrade and the IEM) or play money (Inkling and NewsFutures). Here are today's quotes, as of 4 p.m. ET:
Snee said there appears to be a GOP selloff aimed at settling accounts on Election Eve: "People holding McCain are looking to clear," he said.
Comparing markets, models and polls
Another type of market at the IEM predicts what the eventual popular vote total will be, and those shares trade in a much narrower range. The current verdict is for a 53-47 split that favors the Democrats.
How does that stack up against the political polls? It's hard to judge, because even on Election Eve, a good number of voters are telling pollsters they're undecided. The breakdown in the final NBC News/Wall Street Journal poll is 51 percent for Obama to 43 percent for McCain, leaving 6 percent in the undecided-or-other category.
If we assume that all these figures correctly reflect the final result, the implication is that as much as two-thirds of the undecided vote will break McCain's way, still leaving Obama on top. But that's a big if.
The NBC/WSJ poll, conducted Nov. 1-2 among 1,011 likely voters, has a margin of error of plus or minus 3.1 percentage points. And there are plenty of intangible factors, ranging from the turnout levels for young voters to the potential for a race-driven "Bradley effect."
Another way to predict the popular vote is to use Yale economist Ray Fair's model, which is based purely on economic statistics. The final prediction, issued last week, comes up with a Democratic-leaning 52-48 split on the presidential vote, and a 56-44 percentage split in the House. (The current House split is 54-46, favoring the Democrats.)
Fair has set a plus-or-minus margin of 2.5 percent for his model, and there are additional intangible factors to consider here as well:
Even if Obama ends up winning the popular vote, there's still a chance that McCain will prevail in the state-by-state electoral vote. That's what happened in 2000: George W. Bush may have ultimately won the presidential election, but Al Gore piled up a bigger total of votes.
Eight years ago, Gore's investors won the big payoff from the IEM, while Bush's backers went away empty-handed. The same thing could conceivably happen this year: Obama's investors could win even if their man loses. Nate Silver, founder of the FiveThirtyEight.com election prediction Web site, said there's even a chance that McCain could win the popular vote while losing the electoral vote.
Neither the markets nor the polls are perfect predictors, as Hillary Clinton's surprise win in the New Hampshire Democratic primary illustrated 10 months ago. Clinton's investors cleaned up back then, and McCain's investors could pull off a similar coup this time around.
It all goes to show that there's no such thing as a sure thing, in the markets as well as in politics. Or do you disagree? Feel free to add your comments below.