8 February 2008
In an earlier post I discussed how intrade.com on election nights anticipates electoral outcomes a few hours before they are publicly known. Well, what happened on Super/Duper/Tsunami Tuesday? Here is a chart that does not do justice to the wild vacillations (a finer granularity focused on just 3pm to 3am super tuesday would have been ideal; unfortunately, I could not produce a chart with exactly the right time period.):
Late afternoon/early evening Obama shot up from mid 40s to about 60, and then collapsed to about 40, and steadily climbed to a level a bit above 55.
What happened? The early movement in Obama shares certainly did not anticipate the outcomes of the evening. Well, apparently, the private information flowing through the network "in the know" was that early exit polls were very good for Obama-- potentially in a good position to win California, for example. This would have been a huge victory for him. Of course, as the returns came in (and more data to the exit polls, which generally did pretty well), it became increasingly clear that while Obama would do fine for the night, he would not win California, resulting in a lot of selling of Obama stock (note the volumes).
And the information since Tuesday, especially on the money front, has been very good for Obama, presumably fueling an increase in value Obama shares (I have not parsed the timing of the Clinton announcement and the movement of Obama/Clinton shares).
In any case, sometimes where there's smoke, there is a fire. Just, perhaps, hidden in the network....
Posted by David Lazer at February 8, 2008 3:53 PM