Friday, May 13, 2011

Wagering on Climate Change

Wagering on climate change seems to be growing. At least offering wagers, I don't know how many takers are out there. Bishop Hill mentioned on his blog a new offered wager of one thousand pounds based on the GISS global data being below the 2008 average of 0.5 in the year 2015.

I modified my stack regressions for the wager using GISS monthly data from 2001 to present which is March 2011. I have a thick purple line which is the regression for the whole period and a thick greenish colored line which is the regression for 2008 to present. Of course I have all my stacked regression lines. Since the object of wagering is to predict an outcome, I have modified the stack by including the most recent regressions as thicker dashed lines. The wager is if the GISS 2015 average will be above or below 0.50, 50 on the chart. This looks to be a fair wager.

It is fair because the possible paths average just above the wager temperature but the most recent paths are below the wager temperature. Based on the temperature track record from 2001 it is close to a coin toss. This type of wagering is a good opportunity to sucker people in to being over confident in their predictive skills. I would not take this wager without negotiating the odds. 2:1 odds may be pushing it, but after a few courage inducing cocktails, may be worth the bar tab. So a 500 payout above 0.5 with 1000 and the push going to guys betting on below 0.5 sounds like fun, 5:4 is doable.

Remember, predictions are hard, especially predicting the future. The stacked regression method does give you a little more to work with, I think.

No comments: