The crimson tide

No more messing around, it's time to look at what looks likely to be the central battleground between Labour and the Conservatives: the Conservative-held marginals and the Labour targets.  But it's worth noting that these are not quite the same things.  In many seats the Conservatives hold, the Lib Dems are the main challengers, while Labour are gunning for quite a few of the Lib Dem seats.

 

Let's look at it from the Labour perspective first.  Here are their target seats in order of swing:

https://drive.google.com/file/d/0Bygi8eZw-4q1aWFXQmdDaF8tOVU/edit?usp=sharing

 

They need to take a minimum of 68 of these seats to form an overall majority.  Seat number 68 by majority is Norwich North, for which they are an 11/10 shot. 

 

Now let's organise these in order of bookies' odds:

 

https://drive.google.com/file/d/0Bygi8eZw-4q1SE5ocHYwcVFaSkU/edit?usp=sharing

 

And on this occasion, there isn't a radical transformation.  The 68th seat in order remains an 11/10 shot.  As compared with the Betfair price of 2.7 (7/4) for Labour securing an overall majority, this represents poor value.  So my first observation is that it's pretty pointless placing many bets on Labour seats as a proxy for them securing an overall majority, because the prices in general are worse.  Just go to Betfair and lump on the general market.

How about the chances of Labour being largest party?  This is harder to judge because it's not at all clear how many seats Labour would actually need to be largest party, particularly given that the Lib Dems look likely to be under pressure from both main parties.  The Conservatives and Labour currently share 565 seats, meaning that 283 would have been enough to be the largest party last time round.  If you take the view that the magic number next time round might be somewhere around the 290 mark, Labour would need to take a further 32 seats.  The 32nd seat in order of odds is Halesowen & Rowley Regis, a 2/5 shot.  The 42nd seat in order of odds (implying a required seat tally of 300 and that the Lib Dems would lose 35 seats) is Gloucester at 4/7.  Even if you think that the Lib Dems will be turning up to the next Parliament in a taxi, Labour's 52nd seat in order of odds is Loughborough, for which the best price on Labour is 8/11.  On Betfair you can back Labour to have most seats at 1.79 (4/5).  So once again, go to Betfair.

With this kind of discrepancy between the general markets and the constituency markets, you would not expect to find much value in the constituency markets on the Labour side of the fence.  Is there any?  Perhaps there is in some of the non-standard cases.

 

First, we can disregard all the standard Labour/Conservative marginals and semi-marginals.  The bookies have a reasonably consistent basis of pricing, and it's less favourable than Betfair. 

What about seats involving the Lib Dems?  Once again, the bookies have a consistent basis of pricing.  Curiously, the prices seem generally more weighted to Lib Dem survival than the equivalent swings required for Conservative challengers, even though it is commonly thought that the Lib Dems are more likely to pick up tactical votes from the left than from the right.  On the basis of yesterday's ICM poll Labour have had a 9.5% swing from the last general election relative to the Lib Dems.  If that was replicated on a uniform national swing basis, that would see Simon Hughes involved in recounts (Labour 9/4), Viscount Thurso in Caithness, Sutherland & Easter Ross returning to his ancestral home (Labour 3/1) and Danny Alexander in Inverness etc etc etc going back to his park and preparing for memoirs (Labour 9/2).  Labour is bafflingly long at 1/2 in East Dunbartonshire considering it needs a puny 2.3% swing against the Lib Dems there.  Given that the Lib Dems seem to have been hit especially hard in Scotland, these odds seem pretty friendly.

Much will depend on the power of incumbency for Lib Dems, but they're going to need to be playing that power for all that it's worth if they're going to hold out against Labour opponents if the general election result turns out anything like what the polls look like at present.

Again, we need to consider what might be the impact of the referendum result in Scotland.  In the probable case of a No vote, the SNP will be demoralised and broke (they've vowed to spend every last penny on the referendum campaign, defying national stereotypes).  This would accentuate the chances of Labour picking up Lib Dem seats, with fewer votes being diverted to the Scots Nats.  It would also cement Labour as the main challengers to the Lib Dems in the three way marginals. 

If you wanted a brave bet, you might consider betting on Labour to retake Dundee East at 11/4.  It only requires a small swing from the SNP, and while current polling doesn't justify it, you might take the view that after the probable referendum result Labour will rise as the SNP fall, in a bout of internal soul-searching.  But I wouldn't make that bet myself, because the impact of a No vote on the SNP is not entirely predictable – and the vote might yet be Yes.

You might also consider the impact of UKIP on Labour targets.  I'll come back to this in a future post.  I'll say right now though that I'm not convinced that the rise of UKIP is going to be all that directly helpful to Labour.
 

Tomorrow I'll look at this from the viewpoint of the Conservatives defending their seats.  And we should find rather more value bets to place then.

Once in a blue moon: first cut on the chances of a Conservative overall majority

I'm working my way up to the central battleground between Labour and the Conservatives, but let's keep skirting round the edges for now. The Conservatives entertain hopes of an overall majority, despite being behind in the polls at present. But what do the bookies make of their chances?

 

If you were to go onto Betfair today, you could back the Conservatives for an overall majority at 4.3 (100/30). But what of the constituencies that would make up that hypothetical majority? Here are the constituencies that the Conservatives would need to make inroads into if they were to achieve an overall majority:

 

https://drive.google.com/file/d/0Bygi8eZw-4q1aXdyMFd2MGwzcnc/edit?usp=sharing

 

The Conservatives would need to take a minimum of 19 to get a majority. Measured by size of majority, the bookies aren't remotely convinced. Number 19 on this list is Newcastle-under-Lyme, and the best price that you can get is 8/1. But let's sort them by odds instead. And here a very different picture emerges:

https://drive.google.com/file/d/0Bygi8eZw-4q1VVVzb3dlblRob3c/edit?usp=sharing

 

It seems that the bookies regard the Lib Dems as much easier pickings than Labour-held seats. As if by magic, the yellow rises to the top. The first 21 seats on the list are all Lib Dem held, and are at prices of 2/1 or less. Seven of these are seats that don't even feature on the first list at all, being outside the Conservatives' first 75 targets as measured by size of swing required. If the Conservatives could hold their existing seats and take these 21 seats, they would have their overall majority. I shall come back in a separate post to the question whether that is a sensible way of judging the respective merits of a 2/1 constituency bet on Carshalton & Wallington against the 100/30 of the Betfair price on an overall majority (spoiler: it isn't). Are these prices sensible? I shall be coming back to this question in later posts.

 

 But I note that on today's ICM poll, where the Conservatives polled 32% and the Lib Dems polled just 12%, there has been a swing from the Lib Dems to the Conservatives of 3% from the last election. On a uniform swing, that would give the Tories just eight of these target seats. The Lib Dems may have lost half their support, but the Conservatives have also lost some vote share. We shall need to work out what will happen in Lib Dem-held constituencies in particular: what will happen to the Lib Dem vote? And what will happen to the Conservative vote? That is going to take further analysis in future posts.  The destiny of these seats at the next election is not obvious on first inspection.  Of the Lib Dem seats, I like the look of the 11/10 on the Conservatives in St Ives and North Cornwall best.  The odds in Berwickshire, Roxburgh & Selkirk and in North Devon both look decidedly mean.

What about those more difficult Labour seats?  Are there any seats where the Conservatives are worth backing?  Well clearly on current polling the Conservatives cannot hope to take many of these.  If you believe that the pendulum will swing in the Conservatives' favour, you need next to consider where the Conservatives will gain support and where Labour will lose it.  By definition, the current state of play will not prevail.  So Walsall South at 8/1 looks like it might be value, requiring only a 2.2% swing and with a relatively small Lib Dem third party vote.  It certainly looks more appealing than the 8/1 you'd get on the Conservatives defeating the Lib Dems in Westmorland & Lonsdale.  The 16/1 in Scunthorpe is worth considering if you think that UKIP will make inroads into the traditional working class Labour support (but note that it was Elliot Morley's seat, which may have affected Labour's performance in that seat).  The 10/1 in Great Grimsby looks like it might be value as well, for the same reason (though note that there is a big Lib Dem contingent here).  And the 5/1 in Wirral South is relatively long for the minuscule majority Labour have in the seat.

 

It has to be recognised, however, that all of these seats are going to be a challenge for the Conservatives without major movements in the polls.  Any bets placed need to be understood to be outside bets.

The SNP's chances in 2015: the early constituency betting

In stark contrast to the Lib Dem constituencies, all of which have been priced up by one or more bookies, the Scottish constituency markets are still being created.  Most of the interest here is with the SNP's chances.  Will they be able to increase their seat count in what might yet be the final UK Parliament that Scotland participates in?  Their polling is currently running far ahead of 2010 levels.  Can they convert this into seats next year?

 

I set out here the 59 Scottish seats in order of the swing required for the SNP to take them (with the six SNP seats at the end of the table): 

https://drive.google.com/file/d/0Bygi8eZw-4q1eUUxTDhQVGJuWFU/edit?usp=sharing

Two things immediately stand out.  First, the SNP is fighting a battle on three fronts.  And secondly, the size of the swings required to make even limited progress is daunting.  In many of the more promising seats, the SNP would need to come from third or fourth.  Not only would they need to dislodge the incumbent, they would need to muscle past existing established challengers.

 

So how do the bookies rate the SNP's chances?  Well, here are the seats so far quoted listed by odds:

 

https://drive.google.com/file/d/0Bygi8eZw-4q1QVhHYkNFZzBsZ2s/edit?usp=sharing

This list is very incomplete.  The SNP's number one target by swing, Ochil & South Perthshire, is not yet listed by any bookie.  But already we can see that the bookies don't particularly fancy the SNP's chances.  They're odds against in every seat listed so far that they don't yet hold.

 

The SNP are, however, thought to have relatively better chances in Lib Dem seats than in Labour seats.  They are rated as having a far better chance (5/1) of achieving a 15% swing to win North East Fife from fourth place than of winning Edinburgh South on a 13.5% swing from third place (33/1).

The swings required for substantial SNP success are heroic. To put their chances in context, a 10% swing to Labour would give them a landslide majority.  A 10% swing for the SNP would give it five seats.

But the SNP do seem to be performing quite a lot better in Westminster polling than they performed in 2010 (be sure to differentiate between Holyrood polling and Westminster polling). It is far from apparent that this swing is yet enough to justify the short odds on SNP seats that are being quoted so far.

Looming over everything else in Scotland is the independence referendum. How might this affect voting intentions in 2015? It seems inconceivable that such an important referendum result will not have a major impact. In particular, if there is a Yes vote, it seems likely that the SNP will get a major boost in the polls. Seats that looked out of reach may suddenly become winnable.

The consequence of all of this is that some of the longshot seats may make good proxies for a referendum victory for Yes at much better odds. The chances of a referendum success for the SNP seem to have been fully priced into the seats where they are more fully in contention. This means, paradoxically, that the SNP longshot bets in constituencies may well offer better value than the short priced bets in seats that they have much better hopes of taking. Whatever your view of the matter, I maintain that the SNP are considerably more likely to take Edinburgh North & Leith, for which they are rated a 50/1 shot, than Charlie Kennedy's seat, for which they are rated a 10/1 shot. 

How far will the Lib Dems slide down the electoral razor blade?

With the Lib Dems polling 7% in two separate polls today, and with only just over a year to go until the general election, it seems like a good time to see how the bookies assess their chances of keeping the seats that they hold.  We have prices on every single Lib Dem seat, so we can do a complete analysis.  Here are the best prices on the Lib Dems listed by size of their majority (the table is based on UK Polling Report's magnificent work):

 

https://drive.google.com/file/d/0Bygi8eZw-4q1NWFPVjdta3lfMGM/edit?usp=sharing

 

The nearest challenger is denoted by the colour of the majority.

 

But we can give a more informative table than that.  Let's organise the seats not by majority but by best price:

 

https://drive.google.com/file/d/0Bygi8eZw-4q1WF9Sd1lrZ2xyLVE/edit?usp=sharing

 

From this table, some things leap out.  First, the bookies are a lot less confident about the Lib Dem chances against serious Labour challengers than they are about their chances of holding on in two horse races against the Conservatives. The red has risen to the top of the table.

Secondly, there are 27 seats where the odds are between 2/1 and 1/2 that the Lib Dems will hold on - nearly half the total.  The bookies aren't very confident that they really know what is going on.

 

Thirdly, there are only 20 seats where the Lib Dems are rated by the bookies as odds against to keep their seats.  Even with the vagaries of First Past The Post, that seems optimistic if the Lib Dems stay in single figures.

 

Fourthly, there are some anomalies.  It is unclear why the Lib Dems are rated as short as 6/4 to hold on in East Dunbartonshire with a slender majority of 4.6% over Labour, while they are rated at 2/1 to hold on in Cardiff Central with a much healthier majority of 12.7% over Labour, especially as the Lib Dems appear to have lost ground particularly steeply in Scotland.  Some of the pricing is driven by sentiment: only punters' animus against Nick Clegg can explain why he is rated only 1/4 to hold onto Sheffield Hallam with the third safest Lib Dem seat in the country while the nominally more vulnerable Vince Cable is rated at 1/10 to hold Twickenham.  Similar observations could be made about the pricing of Danny Alexander's Inverness seat relative to that of Viscount Thurso's more marginal Caithness seat.

 

On current polling, most of the value must be found betting against the Lib Dems.  For those wanting to bet on the Lib Dems, Cardiff Central, Redcar and Berwickshire, Roxburgh & Selkirk may offer value.  But one thing is for sure - betting on these seats is not for the faint-hearted.