Making Sure Bets On Online Prediction Markets
from the psst-i-have-a-tip dept
Online prediction markets like Intrade have started to make a name for themselves in this year's exciting political arena. By trading futures on political outcomes like the Democratic and Republican presidential nominations, the idea is that the efficiency of the market will ultimately serve as a better predictor of outcome than traditional methods like exit polls. However, as some traders are starting to find, prediction markets still have many inefficiencies in their current state, allowing shrewed traders to make tidy profits as a result. In the Intrade market, a political future is worth $10 if the political outcome occurs, and $0 if it does not occur. Therefore, a $5 market price on a particular future is supposed to correlate with a 50% prediction of that future to occur. However, in practice, certain factors push these prices out of the range of their realistic probability. For example, contracts for Ron Paul's predicted as high as a 9% chance of him being selected as the Republican nominee, when in reality, his chances were probably closer to nil. Perhaps driven by a small cadre of Paul supporters, the Intrade market was able to be swayed by a small number of trades. Even today, Intrade shows Paul at a 1.2% -- which is a great opportunity for someone to make money on taking the short side of that contract. On the Democratic side, Al Gore supporters have put a 1% chance on his head, who never even appeared on the ballot -- wishful thinking indeed. That said, the limited amount of volume on these contracts precludes anyone from actually making a crazy amount of money on them, but it does remind us of an important fact about markets -- while they do tend to come up with the right answer in the long term, in the short term, they are incredibly susceptible to very human factors like optimism and group think.
Reader Comments (rss)
(Flattened / Threaded)
by toby scammell on Feb 13th, 2008 @ 7:49pm
There are definitely inefficiencies that can be exploited but many of them are mirages. As you would expect, the contracts that seem to be most out of line with conventional thought are those with the least volume. Plenty of contracts are so illiquid that you can't actually trade into them at anywhere near the price of the last trade, which in some cases might have happened days or weeks ago. Most interesting is that there is a pretty consistent lag factor--the intrade market is slow to arrive at conclusions that normal people arrive at. I think part of that has to do with how difficult it is to get money into the system. If there's a contract you really want to buy, it might take you 4 days to transfer the money so you can buy it.
(reply to this comment) (link to this comment)
Manipulators Don't Have an Effect by russ on Feb 14th, 2008 @ 12:14am
There are published articles out there that demonstrate how insignificant 'shrewd traders,' (manipulators) are in prediction markets. I say that after reading http://people.ucsc.edu/~roprea/manipEX1.pdf (there are other publishing out there as well on this subject). I would recommend browsing the conclusion, if nothing else. In addition, there are other published experiments whose data shows that auctions injected with manipulators actually produce lower ending prices compared to regular auctions with no presence of manipulating betters(but still not low enough to be significant). Yes, quick traders my be able to capture minimal profits in the VERY short run, but the data shows that they do NOT have a significant effect on the overall price or validity of the prediction market. Not to mention, there is an inherent risk of losing money while betting far above current prices. Overall, the claims of the inefficiencies in these prediction markets remain to be stated; at least by this author.
(reply to this comment) (link to this comment)
Love your circular reasoning by Lawrence D'Oliveiro on Feb 14th, 2008 @ 12:56am
Q: How long is "long term"?
A: Long enough to come up with the right answer.
(reply to this comment) (link to this comment)
by Anonymous Coward on Feb 14th, 2008 @ 6:40am
The University of Iowa has been doing this for years (at least 40) and it really does work. The one thing it does is force people to put their money where their mouth is.
(reply to this comment) (link to this comment)
Prediction Markets by Patrick L on Feb 15th, 2008 @ 7:51am
Great article! Clearly, liquidity and volume play an important role in the accuracy of the outcome. "many are smarter than the few"
I have included a Wikipedia link to "Wisdom of the Crowds". Remember, the government wanted to trade a futures contract on terrorism.
(reply to this comment) (link to this comment)
PopSci Predictions Exchange by Zatch on Feb 23rd, 2008 @ 9:50pm
Anyone checked out the Popular Science's PopSci Predictions Exchange? popsci.com/ppx
(reply to this comment) (link to this comment)
Add Your Comment