Market action since June 7 suggests equity investors have found some resolve regarding the upcoming U.K. referendum on whether to leave the European Union. As Exhibit 1 below illustrates, stocks of the United Kingdom, as proxied by the iShares MSCI United Kingdom ETF ($EWU), recently kicked off a strong advance versus the iShares S&P Europe 350 Index ($IEV). In fact, the slope of this month’s relative rally already suggests the trend of underperformance, in effect since the summer of 2012, is no more.
Our momentum work corroborates the idea that recent strength indeed represents the early phase of more enduring trend. Specifically, our medium and long-term momentum oscillators both support continuing outperformance. Weekly momentum calls for the U.K. to lead European equities into late August, while monthly momentum looks set to support excess returns at least into Spring 2017. Exhibit 2 below illustrates graphically the weekly and monthly momentum setup for the U.K. stock market versus that of broader Europe.
Not pictured, quarterly momentum likewise appears to be in the process of forging a bottom (we anticipate this very long-term momentum gauge will inflect no later than 1Q17). Thus, the present might mark an opportune point for investors to express an overweight to the U.K. (vs. European stocks) over a multi-year horizon as well.
Consistent with the indications from our weekly momentum oscillator, Oddschecker reports that William Hill is currently offering 3:1 odds to backers of the leave campaign. This translates to a mere 25% implied probability that Thursday’s proceedings will culminate in the decision for the U.K. to sever ties with the 28-nation bloc.
On the other hand, as Exhibit 3 demonstrates, per the most recent polling data, the Telegraph makes the decision a toss-up.
Our experience leads us to discount polling data, as betting markets have demonstrated stunning historical accuracy in forecasting political outcomes in years past. It makes intuitive sense to us that bettors represent the smarter money, as this cohort avoids negative financial repercussions only by correctly handicapping outcomes; conversely, poll respondents face no direct undesirable consequence if they, for instance, change their minds.
Consistent with this rationale, after noting the accuracy of prediction markets in calling both election outcome and various individual electorates in Australia in 2001, Wolfers and Leigh (2002) remarked that the “the press may have better served its readers by reporting betting odds than by conducting polls.”
Similarly, Snowberg, Wolfers and Zeitzewitz (2012) in a NBER working paper found that prediction markets feature several attributes suitable to the task of generating accurate forecasts: they quickly incorporate new information, are largely efficient and are impervious to manipulation. This examination also found prediction markets to demonstrate lower forecasting error than either polls or professional forecasters.
Moreover, Rothchilds (2009) examined the 2008 U.S. elections and, after adjusting the data for inherent biases, concluded that prediction market-based forecasts trumped poll-based forecasts in accuracy as well as information content.
Bottom line: We liken the U.K.’s June 23rd referendum to the 2012 U.S. Presidential Election. In both instances polls called for a close contest, while betting markets were fairly certain of the outcome. On the basis of our momentum work and betting markets historical superiority to polling data, we expect bettors will find vindication again in this instance.
However, if our read is incorrect, and the U.K. secedes from the European Union, we see scope for a material hiccup in the prices of British and European equities, as well as for the Euro and the Pound Sterling. But, as bettors and investors alike are offering strong signals that BRemain is several times more likely than BRexit, we will position accordingly.