The weekly Intrade analysis continues to show the Democrats improving both their win probabilities and EV numbers in the Intrade markets – yet over the last few days a holding pattern has developed suggesting that this might well be the end of any major movement we’ll see between now and Election Day.
If we look at how the weekly probabilities and the weekly Electoral College estimates come out we get:
Now compare these to the daily Intrade tracker:
A clear holding pattern has developed where the probabilities in the simulations have hit the high 90’s, the headline Intrade market is stuck in the mid 80’s and the Electoral College votes are stuck on 364. This holding pattern has a lot to do with the way Indiana, West Virginia and Georgia pulled back towards the Republicans over the last week which we can see by looking at the weekly change in probabilities for all States.
Even though North Dakota moved 20 points to the Democrats, it’s still only sitting on a 31% chance of being taken by Obama. Worth mentioning is how Colorado and Virginia are now considered safe Democrat gains with both having 80%+ Democrat win probabilities. If we chart all States on a fixed probabilities axis, it’s not hard to see why the numbers have settled, with very few states remaining contestable:
North Carolina is the only State sitting between 40 and 60% Democrat win probabilities, and only barely on 59.5% . Unless some major late swings happen somewhere, this 364 Electoral College Vote level could well be the final result (well, maybe 365 if Obama can win the second Congressional District of Nebraska).
This increasingly polarised positioning of the States has also dramatically tightened up our simulation results.
With less states remaining contestable, our simulations end up with more certain results with less variance.
As always, our US Election page has all these results and much more including descriptions of the simulation methodology itself and other goodies.
You may remember us talking here before numerous times about how the Intrade contender markets – the 2008.PRES.OBAMA and 2008.PRES.MCCAIN markets – are a little rough around the edges and are filled with people doing silly things, or rather a particular group doing silly things, like trying to create artificial media narratives by pissing hundreds of thousands of dollars up the wall to generate headlines of “McCain’s New Momentum” etc. It’s why I don’t use the contender markets for analysis – preferring the party markets and the individual State markets. Intrade released a media statement confirming that a single institutional investor was responsible for these plays in the market.
We can see what they look like – here’s a couple of them. This one is during the Presidential debate:
Large volumes of trades designed specifically to artificially push up the McCain price and the McCain win probability. As you can see, it fell flat (as all of these idiotic trades do – you simply can’t outspend reality). Here’s another example of a fool being separated from its money a few days before the debate where they tried 3 times to move the price.
Yet within a few hours it had all washed out of the system. It’s a pathetic spectacle when an organised force, probably with party links, tries to manipulate a big prediction market like the one Intrade runs on the Presidential election –it shows an extraordinary level of naivety.
I’m sure plenty of traders went “Thanks for the free money – tool”.
6 Comments
Nice analysis.
One point is that the manipulation did have an impact on the prevailing price for the candidates. For the last month Intrade has consistently been 8-10% more in favor of McCain than the other exchanges (Betfair, IEM, Hubdub etc). The really surprising thing was that this ‘free money’ wasn’t being arb’ed away.
The differences in the exchange prices are surprising.Even within single exchanges.
The party market President.Dem2008 has always been 2-5 points higher than the Obama contender market which I’ve always found a little strange, to the point where Clinton legacy contracts playing in the contender markets were the only reason I could think of that would cause such a thing.The gap has been closing a bit lately, but still isnt parity.
The $500 betting limit on IEM could be a reason for part of the discrepancy with Intrade, but it doesn’t explain Betfair or you guys.
There hasn’t been a McCain money push in the contender markets at Intrade for over a week, yet the price discrepancies between exchanges are still there. Considering the volume that’s been going on, one would have expected it to wash out by now if it was just that silly McCain money holding up the price discrepancy.
One possibility is that maybe Intrade itself is institutionally predisposed to less certainty because of its clientele, but the Intrade State markets have extraordinary certainty, so it kind of kills that off too.
I’m still of the belief that there’s a lot of inside information in the State markets (which explains their comparatively high levels of certainty) and the contender markets have too many newstraders that make bets on the basis of media headlines driven by an awful lot of second-rate polling that’s floating around this election.
Or it could be like it was here in Australia last election – half the country couldn’t bring themselves to believe what was happening was actually true!
It’s clear that the “large institution” is building a big position long on McCain and short on Obama, but the evidence doesn’t support the charge that the intention behind this is to manipulate the prices. It could just be, as Intrade claim, a risk-management strategy – it could even be a fundraising arm of the Democrats putting in a hedge against an election loss.
Caf – the timing of the events and the volume of trades as a function of time are what convinces me.
With those two examples up there, one McCain price drive happened on the debate and when the McCain campaign were building hype over “something big” happening that day – which turned out to be the Ayers/Obama ad
http://www.youtube.com/watch?v=ONfJ7YSXE5w
Very sus.
The other one happened when Obama hit the 70% probability threshold after a 17 day upward run in the markets and where his polling margins doubled from around 4 to around eight in the space of 3 days.
Again, the timing is very sus.
If it was a Democrat front hedge play, they wouldn’t compress their trades over a small window, but spread them out over days so the Obama price remained high (and/or the McCain price remained low) to maximise the return on the hedge, and minimise it’s cost.
By running the trades over a small window, especially the one on the day of the debate and the Ayers ad, the costs of some of those trades for any hypothetical hedge would have increased by up to 50%.
All very good points. If the idea is to push up the prices, why do you think they’re not getting the message that it won’t work?
Never underestimate the politics of desperation? Republican fronts filled with white collar 20 somethings might not have the wisdom needed to match their intentions?
Maybe it was the Bear Stearns Traders for McCain foundation?
Maybe it was a campaigning experiment?
They have stopped though since they were partially outed – so that’s sort of interesting in itself (not that they really needed outing for people to realise that someone was playing silly buggers).
The silliest thing about it though is that let’s say in a hypothetical universe they pulled it off and put a temporary 40% implied probability floor under McCains price in the Contender Markets. If that occurred, the results would have been such that McCain might may well have had a 40% probability of winning, but the State markets would have been showing Obama on 364 Electoral College votes!
The amount of money required to then sustain a trading defence of that numerical dislocation over the following weeks would have been enormous. We’d be talking about sums of money that could run television ad campaigns in mid-sized states!
Intrade just isn’t that important in the broader scheme of things – if this was done as a manipulation attempt at McCains price for political reasons, the people behind it had extraordinary naivety.