A day late and about a million dollars short

by Daniel on November 2, 2004

Thank you from the bottom of my heart to the traders on the Iowa Electronic Markets, who, on the last day of the campaign, have bid Kerry up to 51% chance of winning. Thus ensuring that, whoever wins, I will have ample material to spend the next four years teasing James Surowiecki about the “Wisdom of Panicky Crowds”.

(the really interesting thing is that the single most probable IEM outcome is still Bush to win with less than 52% of the popular vote. The big move of the bids has been from Bush>52 to Kerry >52!)



FMguru 11.02.04 at 10:15 am

So what, exactly, is the predictive power of these online markets? This is like the Iowa Primary results, which showed a strong Dean lead up until the day before the election, when it became clear that his poll numbers were collapsing.

Looks like these electronic markets simply follow the news and the polls and regurgitate the conventional wisdom. Where’s the value in that?


Andrew Boucher 11.02.04 at 12:16 pm

fmguru: If the electronic markets did that, it would have value, since they would be summarizing all that disparate information in one handy number. On the other hand, I doubt that the electronic markets actually do that (although Iowa should be right about Kerry).


James Surowiecki 11.02.04 at 1:36 pm

Panicky, or suddenly realizing that there’s no more time left to find a greater fool? It’s probably a mix of both.

In any case, that Kerry number is a little fishy, because the bid-ask spread is absurdly large — I don’t think there was any real liquidity there at the end, and there was probably some attempt to manipulate the market in a pro-Kerry direction. (The last trades in all four contracts add up to a total of 1.02, which is a mathematical impossibility.)

I’ve always thought the average price was a more reasonable reflection of the market’s verdict. It has Kerry’s chances of winning at 50.2% and Bush’s chances at 50%. Really going out on a limb there.

Anyway, I think the more interesting IEM market is the vote-share market, since even though it’s traditionally a lot smaller and less liquid than the WTA market, there at least we know definitively how accurate (or not) the market was. (Its historical performance has also been very good.) Last trade has Kerry getting 51% of the two-party popular vote, Bush 49%. (Average price has Kerry at 50.6%, Bush at 48.7%.)


dsquared 11.02.04 at 1:59 pm

I’d agree with James on this one; I think that the WTA market has regularly been all over the place, particularly with respect to the VS market.


dsquared 11.02.04 at 2:10 pm

It’s important that we record this, btw, as neither the closing graph nor the historical data file for IEM will record the astonishing intraday volatility I’m seeing (other than through the High and Low of today’s trade). The Kerry contract has oscillated between 46 and 51 in front of my eyes, while the spread has been all over the place. Meanwhile, the Bush contracts have stayed fairly flat. I really think it would be an improvement to IEM if they didn’t dividen up the liquidity pool and just had an “INCUMBENT” contract.


James Surowiecki 11.02.04 at 2:25 pm

I agree with Daniel about recording this — the WTA prices are jumping all over the place. (Even the Bush prices have been moving recently — I think at 8:45 EDT the two Bush contracts added up to something like 54%, even as Kerry was at 52%.)

Absolutely right, too, that dividing up the liquidity pool into four contracts was not a good idea, and that there should probably be only be one INCUMBENT contract. If you’re trying to predict a single-event outcome, it probably works best when you only have a contract in that outcome.


Ken Houghton 11.02.04 at 2:42 pm

Not absurd. If you’re expecting record turnout (which I’m not, though ca. 57% would be unsurprising), then one of two things is happening:

(1) The Democratic base is out in force, abetted by all the people who ask themselves “are you better off than four years ago” and those who have friends/family/relatives attempting their second or third “retirement” from the Armed Forces, and Kerry will dominate, or

(2) Karl Rove really is a genius, James Cavaziel and Curt Schilling are major boosts, “the four million” who failed to vote because of the DUI in 2000 will be out in force, and the voter “challenges” will go well, so Bush wins the popular vote in a landslide.

Turnout well over 50% of the electorate implies the people are making a clear choice–not a time for middling bets.


dsquared 11.02.04 at 3:27 pm

But that would imply that the two >52% contracts would be doing well, and they’re not; the action is just switching around from a high probability of a narrow Kerry win to a low probability.


Alex 11.02.04 at 3:36 pm

IEM’s platform suffers from the inability to see the ‘stack’ of pending bids and offers. Often the best bid/last print represent a single contract. FWIW, the much deeper Tradesports voteshare/WTA market has had Bush ahead consistently with less volatility than the IEM WTA equivalent. The intermarket arb has at times been as wide as 7 or 8 cents (even greater in the Senate contracts.)

Don’t know if anyone caught wind of this when it happened, but a couple of weeks ago a single individual sold $100,000 worth of Bush contracts down to 10 cents on Tradesports (whereupon it very quickly recovered to 55 cents.) Very odd.


Andrew Boucher 11.02.04 at 4:15 pm

By the way tradesports has Bush victory bid at 55.


dsquared 11.02.04 at 8:45 pm

Slate’s just put out a bunch of early exit polls with Kerry leading Ohio and Florida 50-49. Given that Dems usually do better as the day goes on (it sez ‘ere), that ought to put a few cats among the pigeons … Kerr currently at 52, with the central <52 quotes at 33.4 for Kerry and 34.2 for Bush ... this isn't a market, it's a fucking dartboard.


dsquared 11.02.04 at 9:09 pm

Coo, CT readers could have made a bit of money there … Kerry now quoted @ 57-60, and KERR<52 is 36 to BUSH<52 at 27.


dsquared 11.02.04 at 9:14 pm

Consumer alert, by the way; tradesports is just simply not responding, which I would imagine would be a bit of an annoyance if you had bought Bush @ 56.


dsquared 11.02.04 at 9:21 pm

pfffft … and there goes IEM too.

There’s probably a microstructure paper in this; you have to make allowances for the fact that you won’t be able to rely on trading out of your position as witching hour comes round.


abb1 11.02.04 at 10:21 pm

Here’s the current IEM:

Contract BestBid
DEM04_G52 0.234
DEM04_L52 0.395
REP04_L52 0.186
REP04_G52 0.071

Comments on this entry are closed.