Hence, the prediction market Predictit now has a 81% (19%) implied probability that Clinton (Trump) will win, while the opinion poll aggregator over at Nate Silver’s FiveThirtyEight-side has a implied probability of 88.1% (11.9%) that Clinton (Trump) will win.

If we look at the last couple of months it seems clear that the polls have been somewhat more extreme in both direction than the prediction markets.

This to me is interesting as it seems to indicate there is more “animal spirits” in the opinion polls than in the prediction markets. Or said, in another way this could be an indications that when we act as investors – betting on the outcome of presidential election – we behave more rationally than when we are voting?

Obviously opinion polls and prediction markets are not measuring the same thing. Prediction markets “predict” what the outcome will be of an election at a future date, while opinion polls measures how voters think they would vote today. That said, Nate Silver’s models try to do something in between the two.

Anyway, to me it is interesting that it seems like we are voters are much more inclined to be driven be “mood swings”, while we as investors seem a lot more cool-headed. This is of course also what we could learn from Bryan Caplan’s The Myth of the Rational Voter.

PS note that I am not here discussing whether opinion polls are better or worse than prediction markets at forecasting the outcome elections. I am discussion the different decree of rationality we have as voters and investors (or consumers for that matter).