My introduction to futures markets came in a 1960s TV show called Dobie Gillis, most notable for the ‘beatnik’ character of Maynard G. Krebs, played by Bob Denver, later the eponymous Gilligan of Gilligan’s Island*. In one episode, Dobie’s high school class was assigned to do a fantasy investment exercise, and Dobie along with a cute & smart fellow student chose the egg futures markets. The smart student’s moves always paid off, and Dobie decided to copy them all in real life, building up a huge profit. Finally, the smart student announced she was selling, but Dobie intoxicated with success, decided to keep playing. This went on for a few days of rising prices, but finally when Dobie announced he had held on yet again, the fellow student said, with some horror, “But, Dobie, today is spot day”. The final scene was of the back room of Dobie’s dad’s store, filled to the rafters with the eggs on which Dobie had been forced to take physical delivery.**
Today’s Fin, with a piece by Andrew Leigh*** praising betting markets on events like unemployment rates and so on, brought this episode to mind. It struck me that the case for weak-form market efficiency (the market price incorporates all publicly available information) is much better in a market with a spot date in the near future, as is typical of real betting markets. If you are betting on a race to be run this afternoon at 3, there’s not much point in trying to track movements in the market: you’re better off backing whichever horse is offering the best odds relative to your best estimate of winning probability. More generally, markets with a spot date in the near future ought to be relatively immune to bubbles.
While I’m on the topic, I thought I would repost a piece I wrote before the 2007 elections which, I think, puts a bit of a dent in claims that betting markets really represent the superior wisdom of crowds: in this case, the polls were right long before the pundits and the punters came in last. Of course, one data point doesn’t constitute proof, but supporters of betting markets have made strong claims on the basis of fairly limited data.