This week in the markets saw a move towards the Coalition of around 2% points of implied probability of victory, with Centrebet and Betfair leading the push. As a result, our all agency aggregate came in on Labor having a 73.6% implied probability of victory and the Coalition sitting on 26.4%.
Before we get to the usual charts, Sportingbet and Sportsbet are both offering markets on the size of the swing to the ALP. If we aggregate the data coming out of those markets, this is what the implied probability is for a given swing to each party at the moment – where the red bars are a swing to the ALP and the blue bars a swing towards the Coalition:
The markets aren’t expecting much movement at the election, with over a 63% implied probability of the swing falling between 1% toward the ALP and 2% toward the Coalition – meaning an ALP two party preferred of 50.7 to 53.7%. To win the election, the Coalition effectively need a swing of over 2% – yet the market is pricing that possibility in at around 17% – much less than the headline market probability of a Coalition victory that is sitting on 26.4%.
Meanwhile, the usual charts come in like this:
8 Comments
I took a $100 punt some weeks back just because the odds seemed so ridiculous.
If Tone’Marge are Lodge-bound, I’ll get $440.
We’ll all lose quite a deal more than $100 each if the ALP win an undeserved second term.
In what we’ve lost so far….how does $180 billion get divvied into 22 million in ALP-land?
Just thought I’d get the ball rollin’ Poss……..
Just checked Sportsbet and they currently have ALP at 1.30 & Coalition at 3.40
Let it be said James that all threads should start with a bit of cheek!
Paddy – that change would have occurred sometime early today. I take the prices at around 5-6am on a Friday morning.
Poss- would love to see a comparitive timeseries of the bookies odds on the last couple of elections along with where we are at now 3 weeks out, and the same for the polls. do you have that anywhere?
Cheers
Rich
Possum writes: “To win the election, the Coalition effectively need a swing of over 2% – yet the market is pricing that possibility in at around 17% – much less than the headline market probability of a Coalition victory that is sitting on 26.4%.”
That sort of makes sense, at least that the 26.4% is greater than the 17% makes sense. Victory in the election is approximately a combination of the overall swing and how it’s arranged. If those don’t correlate too much (OK, that might be a big if) then you’d expect the variation in election-effective swing to be bigger than the variation in raw swing. So you’d expect the coalition’s chance of victory to be closer to 50% than the coalition’s chance of getting the swing. Pretending we live in a Gaussian world 17% is 0.95 standard deviations and 26.4% is 0.63 standard deviations: I don’t know if that’s a usual shift but it’s not obviously ridiculous.