Translate

Monday, March 03, 2008

UNP-D (Part 2A): Rasmussen Markets

Going by the Offer on the contracts (more reliable than the last strike price, from my observation), Clinton is at 57% to win in Ohio and 27% to win in Texas. If we assume these are accurate and independent, that means she has about 15-16% chance of winning both.

We can compare these with the 12% she's at to win the nomination in the trading, the 21% chance she's now given in Pennsylvania, and the 11% for her to win the "Lifeline" parley of Ohio, Texas, and Pennsylvania. Essentially, this means that if she were to win both, her chances in PA would shift from 20-80 against to 80-20 in favor.

Given my predictions (that she can win both OH and TX and lose the nomination; surely PA's outcome is most likely to ratify the nomination winner), I have to bet against this combination of expectations.

I'm betting against Clinton on the parley, though I really don't know which state(s) she will lose. I have ridden Obama's odds on TX upward to a potential fat profit (though I'd have to sell before Tuesday night to secure it with certainty).

Less proudly, I am also showing a considerable loss on HRC's nomination chances, which I put some fake money behind when she was around a 60-40 favorite (that calculation persisted right through the UNP until the Obama sweep in the Potomac Primary outcomes became clear to forecasters). So I'm also trying to sell her contracts and limit my loss on that one, asking about 12%.

1 comment:

Chin Shih Tang said...

There was a significant shift in sentiment favoring Hillary after this posting--to about 40% in Texas--which held through the evening until it finally became clear Obama would not close the gap there.

There are still some big gaps in the logic: somehow Clinton is 74% to win PA, but Obama 74% to win the nomination.

The big swing thus hasn't helped HRC nomination chances much (and closed my loss) yet. I didn't close that position, nor did I close the one on Obama in Texas at 48% until it dropped to about 3.

The parley is also very interesting; I'm looking at it to fix at about 60%, which is the true probability in PA for HRC.