Listen to this article
Ask investors whether they think Donald Trump can still win the US presidency and they invariably say something like: “Well, we all thought Brexit wouldn’t happen, didn’t we?”.
Britain’s vote to leave the EU has made the art of prediction that much more perilous, even in the US. The Democrat nominee, Hillary Clinton may enjoy a lead over her Republican rival across a number of polls, yet the market remains unwilling or unable to price in a Clinton victory. After Brexit, can anyone really trust pollsters?
We should not discount them. True, the reputation of UK pollsters was hammered on account of their dismal performance during the UK’s 2015 general election campaign. By and large, they had forecast a hung parliament. None forecast that the Conservatives would end up with a parliamentary majority, prompting the industry to mount an inquiry. But they were on the money predicting a No vote in the 2014 Scottish referendum, as well as the Labour party candidate Sadiq Khan’s victory in this year’s London mayoral campaign.
However, the performance of pollsters over Brexit was erratic. The polls swung several times between favouring Leave and favouring Remain, and as Mike Smithson of politicalbetting.com says, from the start of postal voting to referendum day, “there were more polls showing Leave leads than Remain ones”.
Even so, only one of the leading seven polling companies went into referendum day predicting a Leave victory.
Alan Wilde, head of fixed income at Barings, the asset management company, has some sympathy for pollsters. “Brexit was, in truth, a tricky exercise for pollsters as there was only two options and no historic data to stratify by constituency, region or even country, by gender, age cohort, average earnings or by category of work inter alia,” he says. “Raw polls varied by sampling technique with face-to-face, telephone or online showing inconsistencies.”
Brexit looked more damaging to the reputation of prediction markets that create prices for trading the outcome of events, which all but discounted the chances of a Leave vote. “Don’t Believe the Brexit Hype: Remain are still strong favourites,” was the headline on a blog posted by Almatis, a company that deals in prediction markets, ahead of the referendum.
At face value, bookmakers looked to have got the outcome spectacularly wrong. They priced Remain as heavy favourites, although with most of the money going on the Remain outcome, bookies shortened the odds.
Had the political classes probed a little further, they would have discovered that the number of punters betting on Leave far outweighed those betting on Remain. The bookies were protecting themselves, as they always do. And guess who ultimately profited out of the £120m gambled on the Referendum? The bookies.
The real money was won or lost in the currency markets. The pound became a political weathervane, swinging with every pronouncement, headline, economist’s warning and opinion poll in the run-up to the referendum.
A year before the referendum, the pound was worth close to $1.60. It only started turning down with the dawning realisation that David Cameron was going to make good on his 2015 election promise to hold a referendum.
The first real sign of market concern followed the declaration in support of the Leave campaign by the popular Conservative former London mayor and Brexit campaigner (now foreign secretary) Boris Johnson. But the market was more inclined to believe the polls favouring Remain than those showing Leave in the lead, and as voting closed on the night of June 23 the pound was ready to reclaim the $1.50 mark.
Many investors are inclined to cover their bets, and the extent of nervousness was captured in the futures market, where the cost of insuring against a sharp drop in the pound in the event of Brexit was soaring to levels not seen since the financial crisis.
But if politicians are looking to markets for pointers to political outcomes, they should think again. Markets tend to get round to considering political risk pretty late in the day, probably too late to have a bearing on the event.
Foreign exchange volatility came late in the Brexit campaign. Traders paid attention to the US election campaign only when Mr Trump cut Mrs Clinton’s polling lead to negligible levels after the party conventions in July, and when the polls narrowed ahead of the first presidential debate last month.
Even then, their view has been expressed mainly via the Mexican peso. It was sold off in response to the Republican candidate’s protectionist rhetoric in the summer, but has recovered as Mr Trump’s poll numbers have suffered.
Investors are no better than prediction markets, bookmakers or polling companies at judging the likely outcome of a political event. At best, they can hedge their bets, if they can afford to do so. At worst, they risk heavy losses if they try to predict the shape and direction of politics in an increasingly unpredictable world.
The best course of action is to remain cold and stay aloof. “Financial markets and election pundits alike need to always assimilate all relevant data including polls and bookmakers odds to decide what the most likely outcome will be,” says Mr Wilde.
“The secret to making money from any election outcome, is surely to ignore one’s personal emotions and act dispassionately and objectively.”
Get alerts on UK politics opinion polls when a new story is published