!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> Streamline Training & Documentation: Prediction Market Hiccups

Wednesday, February 13, 2008

Prediction Market Hiccups

Thanks to an article by David Leonhardt in today's New York Times, we have an update on how one popular prediction market — Intrade — is doing in predicting the outcomes of the current round of primary elections.

The predictions have been less than accurate, notably, in the case of New Hampshire, which raises the obvious question, Why?

Leonhardt cites the analysis of Barry Ritholtz on his well-trafficked blog, The Big Picture. In a January 11 post, Ritholtz points to three problems with the political prediction markets: their thinness, their small trading volumes, and the low number of dollars at risk.

The small scale of the Intrade market means that
  • if a significant portion of the participants are biased, they can skew the prices. An example is Ron Paul's non-zero Intrade odds of winning any primary in which he is running.

  • sluggish response to new information, with eventual overreaction all too common. Leohhardt suspects that this is due to a shortage of smart money in the market.
Leonhardt concludes his article by noting that as more traders recognize the opportunity to make money off the inefficiency of Intrade's political markets, the level of inefficiency will drop, i.e., more smart money will make Intrade's prices more accurately reflective of true probabilities.