There seems to be some misunderstanding of prediction markets and some readers think they are no better than polling so I've added a quote from a WSJ article on the Iowa Electronic Market's' website which explains why these markets are, in fact, a very good tool and better than polls as far as monitoring the odds of a candidate getting elected. Hope this ads value to this post and useful information about these markets to the readers.
Begin quote:
"The premise behind the IEM, which was first used in the 1988 presidential election, is that voters don't have any incentive to tell pollsters the truth. But when their own money is on the line, people will act on information that affects the chances the candidates have of winning or losing an election.
"There is a strong incentive for players in these political markets to process information about the candidates and their conclusions are revealed to us through the markets' prices," says Steven Feinstein, an assistant professor of finance at Boston University's School of Management. The IEM works on the same principle as any futures market, he says."
End Quote
I thought it would be interesting to look at the latest in the prediction markets to see how those who are actually gambling their money on who will win the democratic nomination and ultimately the presidential election are betting.
There's research to show that prediction markets have a respectable track record for calling the outcomes - better than pundits that's for sure.
Below the fold I've included some links to graphs from two established prediction markets, the Iowa Electronic Markets and Intrade Prediction Markets.
First, who do the market players think has the highest odds of winning the democratic nomination? The way to think of the prices is to think of them as probabilities that that candidate will win the nomination. So, if the last trade for Obama is $0.78, the markets putting a 78% probability that he will win at the time that trade was made.
Iowa Electronic Markets:
The last trade price shows Obama leads by a wide margin $0.78 to HRC's $0.199. To see an interesting graph which puts it all into perspective copy paste the following link:
http://iemweb.biz.uiowa.edu/...
Intrade:
Intrade shows almost the same result with Obama's last trade price at $0.795 with HRC's last trade at $0.20 but she's being offered at $0.199 (below last price). Go to the link below and you can see a graph of each candidate's price performance since the race began:
http://www.intrade.com/
So, the bettors put Obama way out in front with about a 79% probability of winning.
Presidential Race:
The last trade prices show the democrats well in the lead:
Iowa has dems at $0.559 and repugs at $0.451, i.e., 55.9% vs 45.1% probability that dems will win.
And Intrade shows:
On the odds of it being an Obama-McCain or Clinton-McCain race in November, Obama-McCain given highest chance of occuring:
An Obama-McCain race is at 79.5% vs a Clinton-McCain at
20%.
Now which candidate do the players think will actually win the whole enchilada?
Obama 48%, McCain 39.5%, and Clinton 13.5%
The bettors have basically written Clinton off as the democratic nominee barring nasty surprises. It will be very interesting to follow these numbers once Barack is freed up to concentrate solely on John McCain. Like Barack once said, "I look forward to that debate", and so do I.