Speak to a bookmaker!

There is a question that I get asked every time I travel outside the UK. In Shanghai, in Berlin, in Innsbruck, they all want to know: “So tell me Nina, how will the referendum go? Will they vote to leave? Or will they remain?” Interestingly enough, all the Germans that I speak to assume that there will be a clear Yes for Remain – simply because they cannot imagine the UK leaving the EU and secondly because they don’t understand why there is a referendum in the first place. Unfortunately, it is not that easy.

Taking into consideration that I am a foreigner in the UK who is not allowed to vote and who does not cover politics but business and finance, this is a pretty difficult question for me. For sure, I know that I am not the only one struggling with this question. I guess not only Prime Minister Cameron and Finance Minister Osborne but also thousands of bankers, entrepreneuers and other decision-makers will be racking their brains these days because of the upcoming EU-Referendum.

The reason for that is not just the magnitude of what is at stake – Finance Minister Osborne recently called the referendum a “once in a generation-decision” but also the fact that a lot of people do take the polls with more than a pinch of salt. Given that the pollsters were so bad at forecasting the outcome of the UK general election in 2015 (virtually no one predicted a sound Tory-majority), there is a lot of scepticism about the accuracy of their Brexit-forecasting.

That’s not only because the models or the methodology could be wrong. Many survey institutes and pollsters took a very close look at how they conducted their surveys after the 2015 election and vowed to improve the quality of their forecasts. But there is a second issue which makes the British EU-Referendum so hard to forecast. This is because many voters allegedly don’t tell the truth when they are asked whether they would vote for Brexit or not. Some of them state they will vote remain although they are actually in favour of leaving and plan to vote acccordingly.

The third issue with polling the referendum comes down to the methodology. Do you rely on landline calls? Do you focus on online surveys? Or do you do a mixture of both? Provided that many younger people in this country (like me) don’t have a landline anymore, this of course does affect the outcome. According to a recent Guardian/ICM survey, phone polls lead to significantly different results than online polls. Whereas the phone poll found a majority of eight percent for a Remain-vote (43 percent Remain, 39 percent Leave, 13 percent undecided), the online poll showed a four percent lead for Leave (47 percent Leave, 43 percent Remain, 10 percent undecided).

In addition to that, the survey discovered an overrepresentation of pro-Remain Labour-voters in phone polls and an overrepresentation of pro-Leave supporters of the UK Independence Party (UKIP). “There is no definitive way of adjudicating between the two polls, but as good a guess as any is that the right answer lies somewhere in between”, the Guardian quoted Martin Boon of ICM Unlimited.

So where does that leave me? What do I say to my friends next time they ask me about my predictions for the outcome of the referendum? Fortunately, I did not exhaust all of my options when I looked at the polls. There is second means for when you want to get a better understanding of probabilities and likelihoods. Speak to a bookmaker!

That’s what I did some days ago. I met with Matthew Shaddick, Head of Political Odds at Ladbrokes, one of the big bookies here in the UK. According to Shaddick, more than 90 percent of the money is placed on Remain. “There is more money on the Remain-side, but more bets on Leave”, Shaddick said. On Saturday, Ladbrokes put the probability of a Remain-vote even higher, at 79 percent and cut the odds for staying in the EU at 1/6. William Hill, one of their competitors, even went further with stating that the likelihood of a Remain-vote based on their bets was 85 percent – a significantly higher number than what the polls predict.

“You need to look at where people put their money”, Matthew Shaddick of Ladbrokes explained. Covering around ten percent of the British betting market, Ladbrokes is expecting most bets during the last two weeks before the referendum. “That’s when people will make up their mind”, said Shaddick. He expects some bigger bets closer to time. “We will take in some for 100.000 pounds and more.” Shaddick grinned.

According to him, the current closeness of the polls is actually benefitting Remain. “It will lead to a higher turnout which could be good for Remain”, he said. In contrast, a lower turnout is seen to be favouring the Leave-side. The assumption is that older voters are more in favour of leaving and have a higher likelihood of voting, whereas younger voters are more in favour of remaing but might not show up at the ballot box. So far, so good.

Up until that point, I had the feeling I knew a bit more about the potential outcome on June 23rd. But Shaddick destroyed all my confidence with a single sentence. “You know, we lost a million pounds last year after the general election”, he admitted. My heart sunk. Another deadend? Shaddick saw my dissappointment. “We don’t have as much data for our political analysis”, he said. “In Tennis or Football, you have matches every weekend, you have milions of datapoints.” As a result of which the accuracy of sport bets tends to be much higher than that of political bets. “That’s part of our business. This is not a superpredictor”, Shaddick quipped. I smiled. Maybe I should look into the betting industry instead of trying to understand the polls. And place a bet myself.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s