October 05, 2004

Telling stories with pictures

Posted by Daniel

Deja vu.

All over again?

Posted on October 5, 2004 07:49 PM UTC
Comments

Nice to know that the roles are reversed this time. Though I don’t quite understand why Bush stock crashed so drastically on November 6 - didn’t they say he had basically won by then?

Posted by Kenny Easwaran · October 5, 2004 10:40 PM

It looks like the contracts pay off on the popular vote, not the electoral college. So did the Gore contracts pay off in 2000?

Posted by Robert · October 5, 2004 11:17 PM

Seems plausible that share value would get more extreme as the time till the election shrinks and the prevailing polls are therefore thought to be better predictors, subject to perhaps one major reversal of opinion. But you’d think traders would gain a better grasp of that dynamic more quickly — ie, that the values would be largely flat at 50 just about until election eve.

Posted by George · October 5, 2004 11:50 PM

Robert: you are correct. Gore paid off in 2000. I don’t see the similarities, myself. If anything that 2000 graph (which had Gore losing the popular vote in late Oct.) shows what a seriously bad predictor the IEM can be. Perhaps that’s Daniel’s point?

Tradesports odds are more even (54/56 ish for the popular vote contract.)

Posted by Alex · October 6, 2004 01:44 PM

But wouldn’t an ideal predictor not change at all, it’d be right from day one?

Posted by Simstim · October 6, 2004 02:59 PM

via Marginal Rev, the value of a good economist.

Posted by nnyhav · October 6, 2004 04:57 PM

Simstim — a good predictor could still be better close to the event than it was further out — that doesn’t really seem to be true of the IEM for 2000 as shown in this graphic though.

Posted by Jeremy Osner · October 6, 2004 06:23 PM

In other news, the IEM is set to open a market on when Rodney Dangerfield will die.

Posted by nick · October 7, 2004 09:26 AM
Followups

This discussion has been closed. Thanks to everyone who contributed.