Their finest hour

Far from failing in New Hampshire, prediction markets performed at their best, providing traders with a much needed reality check.

Hillary Clinton’s upset victory in New Hampshire has encouraged venomous criticism of prediction markets for having overwhelmingly predicted an Obama victory instead. Proof, they say, that all this talk about the superior ability of markets to peer into the future is just so much hype.

This isn’t the first time that an election cycle provides the fuel needed to burn markets at the stake for the heresy of pretending to make useful predictions by harnessing the wisdom of the crowds. We’ve been here before. Remember the Democrats retaking the Senate in 2006, against all odds, or Howard Dean’s unforeseen collapse in 2004 in Iowa?

The classic first line of defense in these cases is to remind people that market “predictions” are really just probabilities, so any one outcome cannot invalidate the approach. The argument is sound and backed up by loads of data. But it would of course be much more convincing if we, as an industry, would remember to show at least as much humility when our market “predictions” appear correct instead. If you’re going to spread the idea that your market called all 50 states in the last U.S. presidential election because each correct outcome was predicted with over 50% chance, then you can’t hide behind probabilities when an 80% prediction comes to naught, as in Obama’s NH collapse.

The second typical line of defense is to admit fallibility, while, paraphrasing Churchill, claiming that “prediction markets are the worst form of forecasting, except for all those other forms that have been tried from time to time”. It is indeed easy to argue that pundits and polls erred even more than markets last Tuesday.

But over here at NewsFutures we don’t think markets have anything to apologize for. We think, rather, that they performed perfectly in New Hampshire, doing exactly what they are designed to do: capturing the consensus opinion in a much finer and dynamic way than all the amorphous media buzz, and, remarkably, giving Clinton a significant chance to win — 20%, give or take a few — even while the pundits and the polls, and some in her own campaign, had already left her for dead.

In fact, it is possible that the people who, on election day, had the best notion that Clinton could still pull it off were those staring at the price of that Obama stock, having to make a decision to buy, hold or sell. Why, they wondered, is it stuck at 80% if it’s such a sure thing? Why are there still people on the other side willing to put down good money on Clinton? What do they know that I don’t? What reports are they reading, what polls are they looking at? Prediction traders were surely among the very few people who, on election day, were confronted with the reality that Obama could still lose, despite what they heard from the pundits and the polls. By virtue of playing the market they were thus better informed than anyone else. For that reason, the NH primary should be chalked up as one of the prediction markets’ finest hours.

Advertisements

1 Comment

Filed under Political Markets

One response to “Their finest hour

  1. Furthermore, the NH results only concerned NH and not the whole US of A. I wouldn’t think Prediction markets were restricted to NH voters. Prediction markets betters from NH were not dominant in the global US prediction market US users. Different population => different results.
    On the other hand polls were constructed upon NH voters not the whole US population. So the same reason for prediction inaccuracy does not work for polls.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s