Greed is good for predictions

If you really want to know who will be the 2012 Republican presidential nominee, don’t ask a Republican …or a Democrat. They’re all biased. But what if you ask folks to predict who would be the Republican nominee, and they had to bet money on the outcome? Would the results be more accurate than traditional preference polls?

Predictive markets, which allow participants to literally put their money where their minds (if not their mouths) are regarding a host of real-world issues, have been around for at least 15 years. Who plays predictive markets? In the absence of hard data, we can only guess. Unlike fat cats who cut checks to the “liberal” or “conservative” candidate or cause that best matches their beliefs, we suspect predictive market players are akin to conventional stock market options traders. They are driven more by the prospect of monetary gain than political philosophy. Of course, one could argue that many big-time contributors to political campaigns hope to cash in on their “investments” one way or another once their candidate wins, but that’s another story for another day.

If, in fact, the participants in predictive markets have the purest of motives – good, old-fashioned greed – then you have the makings of an unbiased device that might be useful in predicting future events. It’s just like a horse race; the bettors don’t really care what the horse thinks about health care, taxes, gay marriage or gun control; they only care if the horse is fast enough to win. The more players in the market and the more informed each player is about the relevant issue (i.e., the speed and stamina of each horse in the race), the better the prediction.

The “wisdom of crowds,” especially crowds of informed people who wager money on a predicted outcome, is a viable social analytic tool for providing insight into specific future events, perhaps matching or even exceeding the predictive capabilities of sophisticated mathematical models in some instances. Ajit Kambil, global research director of Deloitte Services’ CFO program, predicts that, “prediction markets are emerging as a valuable forecasting tool in diverse application areas from sales forecasts to project success.” For more on the story, see page 33.