« previous | TPM CAFÉ READER POSTS HOME | next »
Intrade Predictions -- being manipulated??
Can anyone explain why Hillary's numbers keep inching up in the Intrade predictions? I know this isn't a scientific survey, and frankly, it isn't important, but it's bugging me.
Are Clinton folks deliberately going and buying to edge her up? Is there any other rational explanation?
Surely one of you numbers wonks out there can explain this to me.
And while you are at it, also explain why the numbers add up to more that 100 (off by more than a rounding error, which was my first thought).
Thanks.
Advertisement
Comments (7)
Look at how high Al Gore's numbers are, or Giuliani's for that matter.
I'd definitely recommend a good dose of Hanlon's razor about now:
May 28, 2008 10:51 AM | Reply | Permalink
I have no easy answers for your 100% question, however. For the Dem side, perhaps they've introduced another off-by-a-factor-of 10 thing again with Gore (it wouldn't be the first time). Changing his 2.6% to 0.26% gives numbers that come close to 100%. I can't figure out how to do that with the Republican side, however.
May 28, 2008 10:58 AM | Reply | Permalink
This is because you can't arbitrage away the theoretical inefficiency due to transactions costs. Intrade's commissions are 0.5 per contract, or 0.3 when the price is very low, plus there's a 1.0 point per contract expiration fee of any open position. So in a 2-contract market, you'd need to get prices above 103 before you could actually profit from arbitrage after commissions.
For estimation purposes, the Iowa Electronic Markets don't have this problem, since the exchange is always willing to buy or sell a "bundle" (one of each contract in each market) at a price of 1, so there you can arbitrage away any theoretical inefficiencies.
The Iowa market is less liquid, though, and they don't show market depth. Intrade is a for-profit company, while the Iowa market is a non-profit experiment run by the university's business school.
May 28, 2008 11:03 AM | Reply | Permalink
At the risk of being macabre, perhaps Hillary's RFK comments moved the market. The comments don't need to raise the probability of the event itself to move the market, they just need to raise the awareness (or fear) in market participants that it could occur to cause some price shift. Markets don't care about being PC, they care about turning a profit.
Also, the moves really aren't that big, so it's hard for me to distinguish them from normal fluctuations. The most accurate answer is that, for whatever reason, more market participants want to buy Hillary than before. But you knew that already. If you're convinced Obama will be the nominee, consider this an opportunity to buy more (or short Hillary) at a better price.
May 28, 2008 11:10 AM | Reply | Permalink
It's an open market. I doubt very much there's any manipulation going on, although I suppose some wealthy Hillary supporter could buy up shares in order to boost her numbers. It would be a pretty good way to throw good money after bad.
IEM is better:
http://www.biz.uiowa.edu/iem/markets/nomination08.html
May 28, 2008 11:19 AM | Reply | Permalink
Here's IEM's graph:
http://iemweb.biz.uiowa.edu/graphs/graph_DConv08.cfm
May 28, 2008 11:20 AM | Reply | Permalink
I've been watching the Intrade numbers for a few months now. My unscientific analysis is that there is a primary or some actual event and Hillary's numbers take a hit. Then the spin cycle starts, Hillary is in full bullshit mode, and her numbers start inching up. Then another event with actual significance takes place and her numbers take another hit, usually going lower than last time.
For all I know, this is exactly what will happen after the next round of primaries.
May 28, 2008 11:50 AM | Reply | Permalink
Post a Comment