Iraqi election futures

In the weekend edition of the Fin (reproduced here), Justin Wolfers writes about a betting market on the Iraqi election turnout, run by the Irish betting exchange Tradesports. The bet turned on whether turnout would exceed 8 million and was roughly even money before voting began. The price of the contract rose sharply on early reports of turnouts over 70 per cent, then fell back again when to around even money when it became clear these reports had little basis. The final official turnout was about 8.4 million.

Attentive readers will recall that something very similar happened in the US election when early exit polls favored Kerry. Modifying an old aphorism to say that “two striking observations constitute a stylised fact”, I think we can now say pretty safely that political betting markets display the wisdom of crowds who read blogs.

In economic terms, we need to look at the implications for the efficient markets hypothesis, which comes in various levels of strength. The weak form, that you can’t predicted prices on the basis of their own past movements is well confirmed, and not of much interest here. The semi-strong form is that markets make the best possible estimate, given available public information. This, I think, is still open for debate. Obviously markets react to the news, but in these two instances they appear to have over-reacted. So, it seems likely that in a market with new information arriving continuously, we would replicate the stockmarket finding of excess volatility.

Finally, there’s the strong form of the hypothesis which is that markets make the best use of all available information, public or private. This has clearly failed, or else been shown to be irrelevant. Either there was no information anywhere to suggest that the early reports were wrong in these cases, or there is or there was no useful private information which markets failed to incorporate in prices[1].

Now, let’s look back at the most controversial proposal for use of betting markets, the idea of ‘terrorism futures’. Since most of the interesting information here is private (the public gets color-coded alerts and that’s about it), the claim that a market in terrorism futures would provide useful information depends on the strong form of the efficient markets hypothesis. The (stylised) fact is that strong form efficiency doesn’t hold in political markets, and therefore that terrorism futures would provide no useful information.

On the success side for betting markets, they got it right in predicting that Labor would win in WA, but then, so did everyone else.

fn1. In a strict sense, the private information obviously existed. People knew whether and how they had voted, and a perfectly aggregating information market should have been perfectly informed about the votes that had been cast, and well-informed about votes that people intended to cast.

9 thoughts on “Iraqi election futures

  1. Pr Q manages to offend efficient marketeers and bloggers in the same breath:

    political betting markets display the wisdom of crowds who read blogs.

    The betting & futures markets, apart from the few hours of panic surrounding the exit poll fiasco, consistently picked the correct result in the Howard, Bush, WA & Iraqi turnout elections.
    The exit poll stuff-up did throw the markets. However this cannot be counted as a predictive failure since the voting process was already underway but being falsely reported ie it was an indicative failure. A little bit of volatility never hurt anyone – traders meat.
    Pr Q makes a factual error regarding the CW on the WA election:

    On the success side for betting markets, the [markets] got it right in predicting that Labor would win in WA, but then, so did everyone else.

    In fact the polls (“everyone” alright) predicted a close LIB victory a month out:

    Recent opinion polls tip a Liberal win. A Morgan poll for November-December showed two-party-preferred support for Labor in WA had fallen 3.5 points to 49.5 per cent, while the Liberal-Nationals had picked up 3.5 points, rising to 50.5 per cent.

    The markets easily beat the major alternative predictors in the public domain – the polls, polticometric models and pundits – from much further out. The present commenter, as long ago as late 2003, was predicting, on the basis of betting markets, that the leaders of the Coalition of the Willing would win their respective elections. I think this prediction has held up reasonably well.
    Perhaps it is the underwhelming performance of the pundits, rather than the whelming performance of the punters, which is causing the good Pr’s brow to furrow.

  2. I wasn’t paying close attention, Jack. What were the market odds on the WA election a month out?

  3. It was rumoured that someone who knew about the 9/11 attacks went short on equity markets.

    If this did happen, do economists have any methodology to detect these investment decisions ante facto, when in the minds of many they become perceivable post facto?

    A more practical question: Is there some way these investment decisions can be disguised post facto?

  4. Pr Q asks the key question:

    Jack. What were the market odds on the WA election a month out?

    One month out Centrebet was giving even odds. Then the election was called and they shortened towards the LN/P for a week or so….
    WA Election 2005: betting market shortens for the Coalition Friday 28 January 2005 at 6:05 pm

    The betting market is the other good predictor of election outcomes. Up until today the betting market at Centrebet had the odds at $1.85 each way. That is a win probability of 50% for both the Coalition and Labor.
    Today the odds have the Coalition paying $1.60 for a win and Labor paying $2.20. That translates to a win probability of 58% for the Coalition and 42% for Labor….
    This evening Labor is on $2.10 and the Coalition is on $1.65 – so a bit of a bounce for Dr Gallop. The implied win probabilities from these odds are 56% for the Coalition and 44% for Labor.

    Punters got it right after Barnettt made his canal blunder.

    The betting market responded to two events during the 2005 WA election campaign. The campaign began with the Coalition as the punter’s favourite. The Coalition’s announcement of a Kimberley to Perth water-supply canal saw Labor become the favourite.
    Labor’s odds then firmed this week with the Coalition’s $205 million arithmetic error (or more likely proof reading error that did not detect missing lines of savings) when it revealed its budgetary costings.

    he Punters to swing back to ALP in the last three weeks.
    WA Election 2005: Labor firms as favourite
    Tuesday 8 February 2005 at 8:44 pm

    Labor is firming as the favourite in the WA election. This evening, Centrebet is paying $1.57 for a Labor win and $2.25 for a Coalition win. The implicit probabilities are a 41 per cent chance of a Coalition win and a 59 per cent chance of a Labor win.

    Thus the evidence from the WA election is ambiguous in terms of refuting or confirming the “Punters > Polls/Pundits” thesis. The best of the Punters were not as savvy as the best of the Pundits (Mumbles deserves a tremendous pat on the back here, although he stuffed up the US election).
    I have to admit that the WA evidence is consistent with Pr Q’s claim that the Punters look not much better than glorified bloggers, less competent than…competent pundits.
    But the Punters were better than the Polls, some of which still had the LN/P ahead in key margninals with a week to go.
    As against this, I think that the general performance of Punters in the US & AUS elections was better than most polls, pundits, politicometry and bloggers. I reject Pr Q’s claim that the exit poll fiasco shows the limitations of punters as predictors since this was a case of misreporting a current event rather than mispredicting a future event.

  5. Marginal seat polls correlate fairly poorly with election outcomes. I suspect many are junk: small samples, and muddled interpretations. It is misleading to counterpoise ‘polls’ to ‘punters’: polls are a statement of voting intention, to extrapolate from them to an election at a future date involves an assumption that voters’ behaviour will not change. Punters are parasitical on the polls they are guessing how intentions will change from those recorded in a poll. Election betting markets are recent, but newspaper election predictions predate polls and provide a proxy test. However at least for the SMH in the 1930s they under predict the Labor vote.

  6. You make a logical fallacy in concluding from your premises that market prices do not embody all private information, and that private information is desired about terrorism, that prices would therefore be useless for terrorism. Just obtaining some information could be enough to be useful. Requiring all private information is way too high a standard to impose.

Comments are closed.