Trading in political futures on the leading InTrade market gives Obama an 84% chance of the White House, against 16% of McCain.
Check out the graphs - and notice how late the Obama surge broke.
And now we know why. InTrade has revealed in a statement that a single institutional investor spent hundreds of thousands to make McCain look more competitive on the market than he was. This reduced the Obama probability of winning down by around 10 points over a month. (This is not simply counter-cyclical betting by somebody who thought the herd were getting it wrong: they were deliberately betting at much worse odds on InTrade than were readily available elsewhere).
CQ has all the details of how it was done. I heard about this on Paul Krugman's blog, but even the Nobel winner has to doff his hat at Nate Silver, who spotted something suspicious and worked out from the betting patterns a few weeks ago. (Sliver's www.fivethirtyeight.com blog takes political numbers to a new art form).
But there wasn't enough of a future in it. You can buck the market for a while. But you can't buck the election.
'I can live with defeat says McCain' probably isn't the best headline in the momentum stakes either.