QandO
Questions and Observations| « Do you really want to run on the Third New Deal? | Main | Kerry's Record - Navy » |
|
July 30, 2004 News From Iowa The Iowa Electronic Markets have been a very accurate predictor of the results of US elections for the past several years. This market is a real, active futures market, where real people put down real money to buy futures contracts on the outcome of elections. These markets have been shown to be pretty accurate. Ex post evidence suggests that prediction markets can be good at forecasting in the very short run. Berg, Forsythe, Nelson and Rietz (2003) summarize the evidence from 49 IEM election markets run between 1988 and 2000. Election-eve average absolute prediction errors average 1.37% for US Presidential elections, 3.43% for other US elections and 2.12% for non-US elections. They also find that the election-eve market forecasts generally predict better than the latest major national polls. New evidence presented in the current paper shows that markets are generally much better predictors than polls months in advance of the elections. Judging by the IEM, Kerry has done a bit to pull within neck & neck of W over the last month. Based on the current closing price of the presidential election contract, Bush looks to be at about 51% ($0.510) in a race against Kerry. The Kerry contract puts him at 50% ($0.499). Yes, I know this comes out to 101%, but these are two separate contracts, and they are priced separately. What this says is essentially that the race is dead even, with an infinitesimally tiny edge for Bush. Looking at the chart, you can see the "Kerry Bounce" in the IEM by the way the two lines come together. This is not, however, a convention bounce. The prices have been converging since the end of the last month. Interestingly, though, the Kerry price has held rock steady at 0.499 for the past four days. The Bush price, however, has gone from 0.492 on 27 Jul to 0.510 at the close today. For control of Congress , the prices are as follows: Control of the House of Representatives Control of the Senate If you are wise, you'll check these figures often. |
So which options would you rate as a buy Dale? Any you'd rate as a sell? Or are they all holds for you?
I think Kerry is underpriced personally.
Posted by: Elliot Fladen at July 30, 2004 04:13 PM
I think Bush is underpriced. I'd rate him as a buy.
Posted by: Dale Franks at July 30, 2004 04:33 PM
Note that this market is betting on the *popular vote*, not the election outcome! (Read the prospectus).
Posted by: Mike at July 30, 2004 06:42 PM
Yes it is, but that serves, I think, for a decent enough proxy. If the popular vote is within 3% the race is probably uncallable. Outside of that, however, I think you face a vanishingly small chance of getting a large popular vote win without also getting an electoral college win.
Posted by: Dale franks at July 30, 2004 07:38 PM
I have watched the Sportsbooks closely--
Bush used to be a 2-11
(you wager $11 to win $2--i.e. Bush a dead lock)
and both Kerry and Bush have recently moved to 10-11 (pretty much dead heat with the vig, of 10%)
But after the convention, Bush moved to 1-1
So I plunked $100 on him)
I suspect that I at least put the odds slightly in my favor by saving the 10% to the house.
A good move as it has gone back to 10-11 for both.
Fingers crossed for more than one reason :)
Posted by: Ed at July 31, 2004 05:19 AM
Ed,
as a rule of thumb: When the odds are set at even, or the home team(incumbent in this situation) is a slight favorite...the oddsmakers love the visitor or challenger. Given what I know about bookmaking, the oddsmakers set the line so naive individuals would lay their money on the incumbent. Basically what I am saying, is you flushed good money down the toilet when you wagered on Bush to win the election in November. Sorry!
Posted by: JimmyTheGeek at August 13, 2004 05:02 PM
PS
That $100.00 would have bought 5 lap dances at the Gold Club in San Francisco.
Posted by: JimmyTheGeek at August 13, 2004 05:04 PM