Brexit Is A Toss-Up Now

| About: iShares MSCI (EWU)
This article is now exclusive for PRO subscribers.


The odds of a Brexit have been rising in recent days, as polls show the Leave side increasing its lead.

Past independence/sovereignty referendums indicate that most undecided voters will vote to Remain.

However, after allocating most undecided voters to Remain, the Leave vote is projected to be around 49-51%.

I think the chance of a Brexit is now around 50/50, and the markets have priced in a lower probability of this happening.

A speculative long gold and short British/European currencies and equities strategy may have solid expected returns.

Recent EU referendum polls have shown the Leave side with a solid lead and has led to fears that a Brexit will actually occur. Past independence/sovereignty referendums indicate that most undecided voters will likely chose to Remain. However, even after accounting for this, the results are very close to even. I think that the probability of a Brexit is being discounted too much still. Even though a Brexit is still far from certain, the probability appears to be close to 50/50 right now, while the markets may only be pricing in a 30-40% chance or less.

Scottish And Quebec Referendums

It may be instructive to look at polling and results for the 2014 Scottish independence referendum and the 1995 Quebec sovereignty referendum. Both referendums had their equivalent of the Leave side gain momentum in the final weeks before voting, but both fell short (with the Quebec vote being quite close).

Polling for the 2014 Scottish independence referendum indicated that the "No" vote had a slight lead during the last week before the referendum. Polls conducted during this period indicated an average of 48.3% for the "No" vote, with 44.2% for the "Yes" vote and 7.5% undecided. The actual vote ended up as 44.7% "Yes" and 55.3% "No," so it appears that nearly all of the undecided voters ended up voting "No" in the end.

Although the "Yes" campaign gained some ground in the final weeks, only two polls in 2014 ever showed the "Yes" side with a lead, and in both of those polls, the "Yes" vote was still less than 50%. Therefore, a victory for Scottish independence seemed unlikely.

The 1995 Quebec sovereignty referendum was closer, and polls taken the week before the referendum showed the "Yes" side with a modest lead of around 4%. The "Yes" vote averaged 46.3% in the polls, compared to 42% for the "No" vote. There was a large amount of undecided voters, at 11.7% of those surveyed.

The final vote was a narrow win for the "No" side, which took 50.6% of the vote, compared to 49.4% for the "Yes" side. In this case, around 75% of the undecideds ended up voting "No." It is notable that none of the polls showed 50%-plus support for the "Yes" side in the Quebec referendum either.

Current Polling

The most recent polls indicate that the Leave side has around a 5% lead, with an average of 48.6% compared to 43.5% preferring Remain. Around 8% of voters are still undecided. None of the polls have the Leave side at above 50% yet, although the ICM phone poll has Leave very close to 50%, at 49.8%.

This means if the polling averages are accurate, the Leave side will get 50% of the vote if it can capture 17.7% of the undecided voters. This is in between the results from the two other referendums, with under 10% of undecided voters choosing independence in the Scottish referendum, and just over 25% of undecided voters choosing sovereignty in the Quebec referendum. This, of course, assumes that the polling averages were an accurate reflection of voter intentions going into the referendum, and that only undecided voters changed their position from the last polls.

Thus, I would currently put the chances of a Brexit to be roughly 50/50 currently. This may change with the results of future polling.

Effect On Various Items

Morgan Stanley recently estimated the effect on Remain and Leave votes on various items, such as the pound (NYSEARCA:FXB) and euro (NYSEARCA:FXE) as well European (NYSEARCA:VGK) and British (NYSEARCA:EWU) equities. It used a 30% probability that Britain will leave the EU in its calculations, which I feel is still an underestimate given the most recently available information. A higher Brexit probability would appear to support a short pound/euro and short European/British equity strategy as well as a long gold (NYSEARCA:GLD) and safe haven strategy.

Source: Morgan Stanley and Business Insider

With an estimated 50% probability of a Brexit, though, playing for either result is quite speculative. However, given the potential underestimation of Brexit chances, it appears that betting on a Brexit would result in higher expected returns.


Despite the lead for the Leave side in referendum polling, I'd consider the current polls as pointing toward a toss-up result. The Leave vote has come very close to 50% in the polls, but hasn't topped 50% in any poll so far, while past independence/sovereignty referendums indicate that most undecided voters will end up voting for the status quo. Thus, based on current polls, I'd estimate that Leave has the support of around 49-51% of voters (with no undecideds). The referendum result is truly up in the air.

The betting markets still indicate a higher probability of a Remain vote (around 60% probability), although this has decreased as the Leave side has increased its polling lead. Personally, I think both the betting markets and financial markets have somewhat underestimated the probability of a Brexit at this point. This makes betting on a Brexit potentially more lucrative than betting against it, even though the actual outcome is a toss-up at this point.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in GLD over the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: May short European/British equities and currencies