As I mentioned last time I wrote about my adventures on Intrade, I’m sceptical of the claim, a special case of the (semi-strong version of the) Efficient Markets Hypothesis, that the odds in betting markets provide the best estimate of the probability of political outcomes. I managed to double my small stake betting on Newt Gingrich, and might have made more if I had not overestimated the efficiency with which the Republican electorate processes information. I sold on the news of his work for Fannie Mae, and thereby missed the peak of the market when he won South Carolina.

Having made my point and learned a bit about the practical operation of markets, I meant to cash out my winnings, but that turned out to be a complicated process, and I couldn’t resist another flutter. Rick Santorum was trading at 100-1, and while I didn’t think much of his chances, those are pretty good odds in a four-horse race, especially one with no particularly attractive candidates.

He’s now 17.9 per cent (nearly 4-1 in the old language, if I recall it correctly), so I’ve now made a pretty substantial gain. There’s a bit of a cognitive consistency problem here – I didn’t really mean to make money backing Santorum, so now I need some suggestions as to an appropriate use for the money, one which would offset any damage done by backing him when he was down where he belonged. Orientation can be US, Australian or global.