Monday, 6 September 2010

Market manipulation? Or is Coulson just a jolly good bet?


Bookies give Coulson a 75% chance of going before Conservative Party conference and only a 25% chance of him staying. Guido’s money is on him staying.


Sometime this afternoon, Paul Staines posted that on the Guido Fawkes blog, in an update appended later on to a 2pm post.

In truth, it was a very tiny market to begin with. But newspapers being newspapers, that could have been reportable.

Until, that is, the odds suddenly changed dramatically - now giving Coulson a 90% survival chance.

Unless I am misreading what the screens mean, that was because a brave punter - or somebody or other - has decided to plonk £990 on Coulson staying, with the effect of shifting the odds this afternoon from 1-3 on (25% of survival) by offering 9-1 against (90% survival chance). (I am not a smarkets expert; will correct this post if that is not accurate).

Just like that.

No doubt, the most plausible explanation is that Guido's betting tips are very influential indeed.

Though one or two others may have got a voicemail saying it was a bit of a steal.

Others will know more about whether chucking £990 about for a bit of liquidity (and to freeze the odds) - when there is about £72 on the other side of the ledger - is just what happens on online betting when the City boys are bored. So perhaps we can all be grateful at this significant indicator that the economic downturn has well and truly gone.

***

Newspapers should be much more cautious in reporting betting odds around political odds. Given that these are a key "momentum" indicator, it is not difficult to see why it could be in the interests of a public figure, campaign or their supporters to take an interest in what the odds are (and perhaps make a few quid out of it too). There was much speculation about what was going on when ex-banker Chris Huhne was became the bookies' favourite to win a leadership contest he narrowly lost to Ming Campbell.

As Andrew Grice of The Independent has written:


One Huhne aide said: “It is true that the changing odds were useful to us and we did everything to publicise them. Yes. Did we put money on? No.” The Huhne camp claims there was one occasion on which supporters of Sir Menzies might have put money on him to stop his odds lengthening.

Mike Smithson, a former BBC producer at Westminster who set up the politicalbetting.com
website, says that there was some “artificial, odd betting taking place to position Huhne as favourite”.

Although Huhne’s rivals claim relatively small amounts could move the market, Smithson believes that the sums involved could have been as high as £50,000-£60,000.

8 comments:

Guido Fawkes said...

The great thing about gambling and free markets is that if you disagree you can put your money where your mouth is and profit. If you are right.

So if you think the market is wrong, put your money on it and end up quids in.

(Technical point - £990 wasn't staked, if I have done my sums right £110 @ 90% required £990 to cover it. Less exciting now?)

Incidentally 90% is the best offer, the best bid is at 65% currently - you should look at what people are prepared to pay rather than sell at in general.

Sunder Katwala said...

Thanks for clarification. So somebody has offered to risk up to £990, by offering something like 9/1.

On the question of whether the primary motivation was to win cash or shift the public odds, that suggests it was the latter doesn't it?

I think the natural reading of that is less that they thought "ooh, Guido's right, there's lots of value here" (you don't need to stake @90% to do that do you?) but rather that they could not be confident that punters would themselves put the odds strongly in the "staying" direction quickly enough for comfort, hence the intervention.

Which strengthens a hypothesis that this is a friend or supporter of Coulson, or News International, or the party, or an interested party - and suggests there is a fair amount rattled-ness, despite the business as usual public stance.

Guido Fawkes said...

Somebody was offering £110 @ 90% requiring £990 from bidder(s) to cover the risk.

What is always more firm an indication is the best bid price.

I'm more than willing to offer my house for £5m and happy to buy another for £50,000. Realistically the true price is somewhere in between.

Robert said...

Similar odds fixing during the 2008 US Presidential elections:
http://www.fivethirtyeight.com/2008/09/intrade-betting-is-suspcious.html

Jason Trost said...

Market manipulation in general and price manipulation of prediction markets specifically is quite a large discussion that requires more than a post from Nate Silver to do the topic justice. I've been following markets and have been trading for quite sometime, and I've seen a lot of different tricks to push a price in a certain direction. It's something that happens in all markets but I rarely see it happen in prediction markets and if it does, the market corrects itself quickly. As Guido said, the markets on Smarkets are free for anyone to participate.

Over the coming weeks we will expose more transparency in our markets so that integrity can be ensured and trading activity can be better disseminated.

@Robert Odds aren't "fixed" on a prediction market. In a continuous double auction, any market participant can enter a order on either side at any price. Intrade had a moderately low volume market which I suspect was subject to big price swings with orders of a certain size.

Please email me if you guys have feedback for ways to make our markets more transparent.

Jason / founder of Smarkets

Sunder Katwala said...

Guido - Appreciate the tutorial! what then seems odd to me is somebody saying 'OK, I'll put £110 on him staying' but you'll in effect need to give me 8-1 against has the effect of "less likely to resign" in impact on headline %. The punter's bet seems to say the opposite. But changing the headline worth doing.

Jason - Thanks for taking the time to comment.

Perhaps the point is less about the integrity of your market (you can hardly identify potential weight of £/other motives in staking money, and clearly small events like LibDem leaderships cd be susceptible) but the reportability of these as political evidence in themselves, without a major health warning.

Jason Trost said...

Health warnings should accompany all forms of political evidence including articles by journalists, opinion polls and party literature. It's not only prediction markets that are open to manipulation.

Prediction markets are a tool like everything else with advantages and disadvantages.

Don't believe everything you read :-)

. said...

I'm not a betting man but I think I'd plump for him going sooner rather than later.