Odds for a record year hit 2-1 on Intrade

Since I penned this post, Intrade's odds of 2010 being the hottest year in the instrument record have risen from 50% to 66%.

It's a little hard to know why this is happening. Low volumes (meaning few traders making few trades) are the Achilles' heel of Intrade, and they may be a factor here. On the one hand, we are certainly on pace to break the record. Intrade uses the GISS surface data, and while both January and February were second-hottest overall (at +0.88C and +0.91C, respectively), this was only because of two crazy hot months in two different years (1998 and 2007). What do I mean when I say "crazy hot"? January 2007 was the hottest month in the instrument record with an anomaly of 1.08C. Not the hottest January -- the hottest month, period, in 131 years of records. The third hottest month, ever, was February 1998 (+1.01C). The second hottest -- as luck would have it -- was a March (March 2002, +1.02C). So March 2010 may not break the March record either, but dollar to donuts (looking at the daily satellite temperatures here) the first quarter is going to set a record.

But even so, it's early days yet; we only have data for three months of the year, and the third month still has a week to run. Why are the numbers so high?

It may have to do with the fact that the current El Nino is just hanging in there, and is now expected to continue through Spring. Or it may be the fact that we are emerging from the solar minimum, as I discussed in this post. Or perhaps people are looking at local forecasts, like those predicting the hottest summer ever for the UK.

It's been five years since the last "hottest ever" year, so we can expect the the underlying baseline temperature is about 0.1C greater than it was then (half the decadal trend of 0.2C.) By itself, that's not enough to give us a two-thirds chance to break the record: the short-term fluctuations in the climate are greater in magnitude.

Maybe all these factors together give the savvy Intrader a confidence in a record-breaking 2010 that I, for one, can't entirely parse. Or maybe the contract is a little overpriced. But for whatever it's worth, the money in the market says the odds of a record year are 86 times greater than random chance without a warming trend.