EU referendum: The problem with polls and what it means for markets
As the vote approaches, financial markets are becoming more interested in it, indeed some sterling assets have already started to react to polling data.
A consensus view had been set in financial markets:
• Polling data is unreliable
• Betting exchange data is more insightful
Polling data had proved inaccurate in the past (the Scottish Independence referendum for example), so the small difference between apparent “Bremain” and “Brexit” (henceforth referred to as Stay and Leave) was largely dismissed as irrelevant.
The betting exchanges showing a strong Stay win have been believed to be the better signal to the likely outcome. We have always suspected that this was wishful thinking, after all the polling data seems to suggest that City workers overwhelmingly support Stay, so it might be they found the evidence that supported their own view (known as confirmation bias) choosing the betting exchange view over the polls. Recent data seems to suggest that the polling data might prove to be a better guide this time.
It has been recently observable that polling data is able to influence the odds offered at the betting exchanges. As at June 15 the FT Poll of Polls polling data put the vote at 47% Leave v 44% Remain, which is not what the consensus had thought could happen.
Bloomberg has a Composite EU Referendum Poll Tracker which has the split at 46% Leave and 41.8% remain. The chart is reproduced below and shows the ascent of the grey line (Leave), which appears to be the result of the “don’t knows” realising that they do know after all (i.e. making their minds up).
Bloomberg's EU referendum poll tracker
Source: Bloomberg 16 June 2016
So what have we learned so far?
The Leave party got off to a flying start, sidelining UK Independence Party Leader Nigel Farage and Respect Party Leader George Galloway (of Grassroots Out) and allowing the more moderate Conservative MPs Michael Gove and Chris Grayling to take the centre stage. But the real coup de grâce was to persuade the former Mayor of London Boris Johnson to join.
David Cameron has managed to turn the Stay argument into a scare story of what might happen should we all vote to leave (this has been dubbed “Project Fear”). There has been widespread criticism of the Project Fear approach, but it does seem to be resonating with some sections of the voting population. A similar tactic was used in Scotland to great effect. There seems to be a widely held view that the Prime Minister David Cameron has not lead the campaign very well, to the extent that there are plenty of rumours regarding his immediate future; after all he has already stated that he would not stay the whole term.
Disappointingly for him, the Chancellor George Osborne has fared no better, so post the referendum the political risk in markets will turn to a Conservative Civil War irrespective of the final outcome. This will be enabled as the Opposition are, themselves, divided so there is no one to keep the Conservative Party in check.
Implications for markets
Currency: A consensus emerged in the currency markets that sterling would be weak due to worries about the possibility of the Leave campaign working.
This view was too simplistic for us, particularly as sterling had been $1.58 in June 2015 and stood at $1.38 in March 2016, so substantial weakening had already happened months ahead of the date of the vote. At the time of writing, sterling is at $1.41, not what the consensus would have had us believe.
Furthermore, Bloomberg gives us access to data relating to the volatility of sterling relative to the euro, which has recently been above the level seen just before the Scottish Independence Vote, when the SNP seemed to be winning the argument.
Fixed income: Again, a consensus had been built that a Leave vote would be bad for UK government debt (foreigners less willing to finance UK deficits) and corporate debt (recession fears would undermine confidence in default rates).
It is interesting to observe that as the Leave campaign has had some success (in the polls and more moderately in the betting exchanges) 10-year gilt yields have come down from circa 1.5% to sit at 1.11% today, not what the consensus would have predicted. It could be that domestic investors have reduced their exposure to higher risk domestic assets (UK equities, UK property etc.) to hide in relatively secure UK government debt. This could imply that the move following a Remain vote could be pretty painful for UK gilts, which had stood at just under 2% at the beginning of the year, so yields have a long way to rise to get back there (creating attendant capital losses).
Whatever the outcome of the vote, 1.11% is not a yield that looks remotely sensible given our current views; we would have to subscribe to Project Fear’s doom and gloom predictions to believe that this offers value (in other words disinflationary bust in the UK economy, in which case equities are the bigger risk).
Equities: The consensus believed that UK listed companies with high levels of UK revenue would be under threat. It is hard to disentangle the UK stockmarket from broader global trends, but a case could be built that FTSE 250 and FTSE Small Cap weakness year-to-date has been in part due to their greater exposure to UK domestic companies. Year-to-date though, the FTSE 250 (-5.0%) and Small Cap ex IT (-1.7%) indices are not that far behind FTSE 100 (-1.7%), but the year began with a pretty decent selloff led by the FTSE 250 index (-11.4%) and from February 12th the FTSE 250 (+7.2%) has been in line with the FTSE 100 (+7.1%). The FTSE 100 benefits from having internationally-exposed companies, particularly commodity-related companies that have done well recently, but that factor can be observed globally, not just in the UK, so it would be expected that this area should outperform in the rebound in markets.
The FT did a survey of strategists/investors (May 11th Brexit Trades: Anticipating the markets if Britain votes to leave) which came to the conclusion that banks, retailers and construction would be sectors that would struggle in the event of a Leave vote. These are areas that (year to date) have already weakened with:
FTSE All-share Banks (-17.5%)
FTSE All-share Retailers (-12.8%)
FTSE All-share (-3.4%)
Perhaps both of these sectors now have a good deal of pessimism already priced in, meaning again that the better risk/reward might be to add to this area. Worryingly though, it is just as plausible that this observed weakness is a signal of a more serious UK economic slowdown that is emerging irrespective of the EU vote.
It feels like this has been a subject of discussion for many months now, and like most of the UK we have found it a pretty uninspiring period, with poor arguments presented by both sides. We now find ourselves in the period where it gets interesting, which means financial markets will suddenly react more frequently and violently to data.
A vote to Leave would appear to create the opportunity for the biggest dislocations in financial assets, but our observation is that many of the consensus victims have already been quite weak. Clearly they can weaken further, which needs to remain under consideration.