It’s been an interesting couple of weeks in New Zealand politics which means an active time on iPredict. Much has been made of a supposed challenge to Mr Shearer’s leadership by Mr Cunliffe. He appears safe for now, but what actually happened over the last two weeks? Can iPredict trading give us answers?
To give some context, here is a media timeline of the early days of the ‘coup’. There was a large cluster of both blog and media posting on Mr Shearer’s leadership leading up to the Labour Party conference, starting from November 10th:
Mr Cunliffe Listener story published
Eddie posts at The Standard calling for Mr Shearer to go - 271 comments
Vernon Small suggests Mr Shearer will depart unless the Labour conference goes well
QoT posts on The Standard calling for Mr Shearer to go - 88 comments
Andrea Vance suggests that time running out for Mr Shearer
Irish Bill posts on The Standard calling for Mr Shearer to go – 170 comments
Lprent posts on The Standard calling for Mr Shearer to go – 89 comments
Brian Edwards calls for Mr Shearer to go on this blog
Tapu Misa calls for Mr Shearer to go in NZ Herald
By Tuesday November 13th leadership speculation had built significant momentum and was dominating New Zealand news media.
The first observation to make is this: key stocks concerning Shearer’s leadership didn’t start reacting to media coverage on November 10th, but a day earlier on November 9th. The first stock to move were ‘David Shearer to depart as Leader of the Labour Party in 2013’ which rose from 32% to a high of 57% before settling at around 46%. ‘Shearer to be Prime Minister before 2015’ experienced a similar trading pattern, falling from 37% to 26% before settling on 28%. These are very sharp movements.
The most likely explanation for this is insider trading by sources within media circles or the Labour Party. But common media narratives explaining what really happened run into problems. Trading suggests that its unlikely Mr Cunliffe had been planning a coup for some time to be launched the week before the Labour conference, or that bad press for Mr Shearer in the lead up to the party conference was inevitable given the criticisms of his leadership (a theory suggested by danylmc at dim post here). The movements would have occurred much earlier if that was the case as they could have been anticipated by either insiders or the public.
But nor is it likely that the timing of the anti-Shearer press was coincidental. If that was the case the November 9th trades would have most likely come from an isolated member of the media or blogosphere, perhaps someone who knew of one of the anti-Shearer articles about to be published (but not others), and was simply lucky that an anti-Shearer narrative subsequently developed. This would seem awfully coincidental given how large the November 9 movements were.
The reality probably occurs somewhere in the middle. No specific coup was planned, but in the week before the conference opponents of Mr Shearer decided to launch a media offensive against him with some degree of coordination, perhaps with plans to launch a challenge in February if all went well. This is perhaps closest to the theory proposed by Mr Farrar in his November 12 “The war against Shearer” where he wrote that ‘someone has decided to push a button’.
I don’t purport to believe this explanation is rock solid, or somehow proven by iPredict stock movements. There are many ways to interpret stock movements and it’s possible I’m doing little more than reading tea leaves. iPredict is just one tool useful in understanding New Zealand politics, but given the mechanism prediction markets provide and the incentives at play we have reason to believe they accurately predict political outcomes. iPredict was for instance more accurate than 15 of 19 polls published in the run up to the 2008 general election.
I’ll be doing another 2 parts to this post. I’ll analyse stock movements in the lead up to and during the Labour Party conference, and then look at the winners and losers and what traders expect to happen within the Labour Party over the next 12 months.