The exit polls were wayoff. The pre-election opinion polls weren't much better. But how did the markets do at predicting the outcome Tuesday night?
The past few years have witnessed the rise of political futures markets—online trading operations that allow you to bet on (or, if you prefer, "invest in") candidates. The theory behind the markets is that they ought to do a better job of forecasting elections than polls. After all, if stock investors efficiently factor all available information into the price of a stock, why wouldn't they do the same for a presidential candidate?
There are two main venues for trading political futures: the Iowa Electronic Markets, at the University of Iowa, which has been around since 1988; and TradeSports, a 3-year-old Dublin, Ireland-based exchange. IEM offers winner-takes-all contracts, which pay off $1 if your candidate wins and nothing if he loses, as well as contracts on the percentage of the national two-party vote share. (IEM limits investors to $500 in wagers.) TradeSports offers contracts on the candidates' election prospects, as well as on the outcome of individual states. Thomas Rietz, a University of Iowa professor and an IEM director, told me this summer that a comparison of 596 opinion polls with Iowa's presidential futures prices at the time the polls were taken shows that the futures prices were closer to the actual result 76 percent of the time. For its part, TradeSports says its investors successfully predicted the winner of every Democratic primary and tagged John Edwards as the vice presidential nominee in May. (There are criticisms of the political markets. Read stock-market strategist Barry Ritholtz's comprehensive take-down here.)
So, how'd they do yesterday? The political markets performed creditably during this campaign, though by no means perfectly, especially on Election Day itself. Throughout the campaign, the IEM and TradeSports investors seemed quite effective at and dispassionate about processing information. Both markets generally had Bush easing to re-election. This graph of the IEM's winner-takes-all contracts shows Bush leading since early September. IEM's vote-share graph shows Bush pulling ahead in August and Kerry narrowing his lead but not catching up. On IEM, as of midnight on Nov. 1—just before Election Day started—the Bush winner-take-all contract traded at .512, meaning traders gave Bush a slight edge. They also thought the vote share would break down 50.45 for Bush and 49.55 for Kerry. In other words, IEM investors forecast the ultimate result (Bush victory) and were close on the vote split. (According to CNN's latest tally, Bush won the two-party total 51.6 to 48.6.) As Election Day dawned, TradeSports also had the right call, a Bush victory with 53 percent of the popular vote.
But in the afternoon of Election Day, when exit polls started to leak showing surprising strength for Kerry, the political futures traders freaked out and rushed to dump Bush and buy Kerry. By 4:30 p.m. ET, IEM's Kerry winner-takes-all contract had risen from below 50 in the morning to more than 70. The vote-share contracts on IEM also shifted in Kerry's favor. TradeSports' traders, who had shown Bush winning for the entire campaign, suddenly bid the Kerry election contract up to 67—meaning they thought he had a two-thirds chance at winning. By 5:44 p.m. ET, TradeSports gave Bush just a 39 percent chance of winning Ohio and a 43 percent chance at winning Florida. All on the basis of a few stray data points. Surely, as voracious consumers of all information relevant to political campaigns, these financially motivated investors should have known that exit polls can be unreliable. But instead they panicked.
As results started to flow in, and John Kerry's "seven-hour presidency" began to ebb, the traders—like the rest of us watching on television and the Internet—realized the exit polls were wrong. By 10:26 p.m., the Bush re-election contract on TradeSports was up to 68.9. The rest is history.
So, what are we to conclude? Over the long-term of the campaign, IEM was quite accurate in projecting the vote distribution. And over the long term, both markets projected the winner. But so, by and large, did the polls. The furious and seemingly irrational Election Day market action stands as evidence that the traders are more poll-followers than poll-beaters.