Outlook 2017: Breaking up is hard
The UK referendum result raises the possibility of other countries leaving the European Union. Christoph Schon and William Morokoff examine the impact of such a scenario
At a glance
•A break-up of the EU cannot be ruled out.
•Possible plebiscites on EU membership will key themes in next year’s Dutch and French elections.
•An EU break-up would hit the region’s banks which would have a damaging effect on the broader economy.
•Stress tests show that an EU break-up would probably lead to steep losses in global equities and a weakening of the euro.
After the initial turmoil following the Brexit vote in June, markets appear to have calmed and even recovered to pre-referendum levels. Yet the success of the British Leave camp and the realisation that Britain is not doing as badly as many politicians and market analysts predicted seems to have fuelled EU-critical movements in other nations. Some are now calling for plebiscites on EU membership in their own countries – a groundswell that could ultimately dismantle the 28-country union.
In this article, we examine the effects of a potential EU break-up on a European multi-asset class portfolio by stressing sovereign yields in some member countries to levels seen at the height of the euro-zone debt crisis. As might be expected, our transitive stress test, in which we increased sovereign yield spreads while other factors moved according to historical correlations calibrated from 2011-12, shows steep declines in global equity markets, in particular in the financial sectors (losing between 25% and 55% in some countries), and a drop of the euro versus all leading currencies, ranging from 3% versus the pound and more than 10% against the dollar.
The historic decision of British voters to leave the European Union in the referendum in June created substantial political and market turmoil, with both the main political parties in the UK plunging into leadership crises, and the pound falling to its lowest level against the dollar in 30 years. Equity markets around the world dropped significantly in the days immediately following the vote, with one of the hardest hit market segments being the British financial sector, where shares in Barclays, for example, declined by 30% from Thursday afternoon to Monday afternoon.
Since then, however, the country has welcomed a new prime minister and equity indices have recovered to or even surpassed their pre-Brexit levels. In particular, the UK market profited from a weaker currency, which helped exports and made the prices of British stocks more attractive to foreign buyers. Several banks have already reconsidered their post-Brexit recession forecasts and now predict modest GDP growth over the next two years (just under 2% for 2016 and about 0.5% in 2017), pointing to more resilient economic data than expected and less political uncertainty.
However, these expectations assume that the UK and the EU continue as currently configured, with Britain simply no longer operating within the European framework. On the other hand, speculation has grown substantially since the vote about the validity of this assumption.
Various leaders of far right and eurosceptic parties across the EU, most prominently in the Netherlands and France, are calling for referendums on whether their countries should remain in. The Dutch Party for Freedom (PVV) has been leading opinion polls for almost a year now and its leader, Geert Wilders, has pledged that he will do everything he can to lead the Netherlands out of the 28-nation community. French presidential hopeful and leader of the far-right National Front, Marine Le Pen, has also called for a popular vote on French EU membership.
While the appetite for such plebiscites appears to be quite low within mainstream political parties, the issue will certainly be an important theme in the upcoming elections in both countries next spring. An ensuing groundswell of popular support for a referendum could ultimately prevail, with a potential eventual outcome of a significantly reconfigured EU. Given the current climate, a complete break-up of the EU is within the realm of possibility.
France and the Netherlands are among the biggest net contributors to the EU. With the UK on the way out of the EU, this could mean that the contributions of up to three of the five top net payers could be in jeopardy. In the UK, voices were raised after the referendum, calling for the government to make up for lost EU support and subsidies in, for example, the agricultural sector and academia. While this would probably not be a problem for the financially stronger potential leavers (after all, all they need to do is divert the money they no longer have to send to Brussels to those sectors), similar calls are likely to erupt in the countries that receive a lot more in subsidies than they pay into the EU budget.
The governments of the latter may then feel pressure to give in to these demands. Such a move would probably lead to increased borrowing needs, potentially at yields seen at the height of the European sovereign debt crisis.
Most European banks have exposure to euro-zone government debt, in particular that of their own home country. They do this for various reasons. One is the preferential treatment of sovereign debt under Basel rules on banks’ capital adequacy, both in terms of a zero-percent risk weighting and the liquidity coverage ratio.
Sovereign bonds are also heavily used as collateral in the interbank funding market and the long-term refinancing operation (LTRO) of the European Central Bank (ECB). The latter can be particularly profitable in countries where government bond coupons are higher, as banks can finance those purchases cheaply through the LTRO. The concern is that this might lead to a similar ‘doom loop’ as observed in 2011-12. Sovereign yields could go up, which could lead to banks which hold these bonds on their balance sheets or use them as collateral for their funding starting to struggle. Such banks might even need to be bailed out by their government, which, in turn, could lead to even higher borrowing and increasing yields for the government, and so on. The graphs in figure 1 show the spreads of 10-year government bond yield spreads for Spain, Italy, Portugal and Greece, respectively, over German Bunds, together with the five-year CDS spreads of the two largest banks in each country.
A close relationship is evident between sovereign yield spreads and financial bond risk premia. This does not only apply for those countries that see a significant upward move of their government curve, but for financial institutions in general. Even spreads for German, French and Dutch banks widened by about 200bps at the height of the crisis, although this was still some way below the levels observed in countries such as Spain or Italy, as can be seen in figure 2, which shows average five-year bank credit default swap (CDS) spreads by country.
Once banks encounter financial problems, there is likely to be an effect on the wider economy, too, as lending dries up. There is also a view that country-specific risk (expressed as a sovereign yield or CDS spread) sets an approximate floor for most corporate borrowers in that country, under the assumption that if the country defaults, there will likely be substantial credit deterioration for all local corporations. They will certainly have difficulty getting access to international capital markets, and banks and other industries cannot rely on sovereign bailouts when the sovereign has defaulted.
The analysis above suggests that a reasonable stress test would be to raise government bond yields to the levels seen at the height of the sovereign debt crisis in 2011-12. We therefore increased 10-year yields in Spain, Italy, Ireland, Portugal and Greece by 400bps to 800bps. The stress test was set up as transitive, meaning that the other risk and pricing factors affecting the portfolio would move according to historic correlations with the stressed factors. We used weekly returns and factor changes over the 2011-12 period to calibrate those correlations.
To demonstrate the effect of these stress tests on a European investor, we constructed a hypothetical euro-denominated multi-asset class portfolio composed of 40% equities, 30% corporate bonds, 20% euro-zone government bonds, and 5% each of foreign currency cash (sterling, dollar and yen) and commodities (oil and gold). The currency distribution was 58% euros, 21% dollars and 8% sterling, with the remainder being a mix of other currencies.
The simulated overall loss for our multi-asset class portfolio was 5.5%. The biggest contribution, with a drop of almost 3% came from the global equities portion which, on a standalone basis, declined by 7.4%. The biggest category loss could be observed for government bonds, which gave up almost 12% of their value, translating into a contribution of a negative 2.3%. A further 0.5% was taken away by a drop of 1.5% in corporate bonds, while a strong performance of more than 5% in foreign currency cash added back 0.3% to the overall portfolio value. Commodities, on the other hand, showed a mixed picture, with oil increasing and the gold price going down, thus having no return impact on aggregate.
In terms of the time frame of these returns, currency markets and stocks, and financials in particular, can be expected to react almost instantly to a vote to leave in either France or the Netherlands, or even just the announcement of a referendum. The mechanics described above following increasing sovereign yields, on the other hand, are likely to take much longer, so the losses shown here should be viewed as cumulative over a 12-to-18 month period.
Stock and bond impact
Not too surprisingly, the biggest losses occurred in the countries where we had stressed government bond yields. But it was not only the government bonds that declined, with decreases ranging from 20% in Spain to 40% in Portugal, but also the corresponding equity and corporate bond markets. As expected, these were largely driven by steep falls in financial stock and bond prices. The largest simulated negative financial stock return of 55% was recorded in Ireland, but even in the US, the UK and Germany banks could lose as much as 15-30% on average.
Losses were less pronounced in the corporate bond market (1.5% on average), but were again led by financial issuers dropping between one and five per cent, with some individual ones losing 10-15%. A small number of sectors, such as consumer staples, healthcare and information technology, proved to be more resilient in some countries. However, all industries were down in Italy, Spain, Ireland and Portugal.
Even Asia-Pacific stock markets could offer only limited relief for our European investor, as they were barely lifted out of the red by gains of 6% in the yen and 10% in the dollar (both relative to the euro). UK Gilts, however, showed a healthy return of 11%, helped by a 3% strengthening of the pound versus the single currency. And German Bunds also affirmed their safe-haven status by gaining more than 6% on average.
The stress test simulates the potential market impact of a significant increase in European government yield spreads, should other net payers also decide to leave the EU. The test was calibrated based on market movements during the height of the euro sovereign debt crisis in 2011-12 and results are likely to be different when using other time periods or correlations. It also does not capture any potential long-term structural changes caused by such an event.
The analysis shows that, apart from the government bonds of weaker member states, financial institutions (stocks and bonds) are the most likely sector to be hit by further EU membership referendums, not only in the countries where sovereign yields are going up, but also in ‘stronger’ countries.
Diversifying internationally can help dampen the effect of falling stock prices, although most of the positive returns (from the perspective of a euro-based investor) are likely to come from FX rate gains of other currencies versus a weakening euro. However, there is the possibility of a euro rebound, if Germany or the ECB decides to support their currency, which could then lead to further losses, if investors decided to shift funds into, for example, US Treasuries.
We also saw that some sectors, such as consumer staples, healthcare and technology, were less prone to rising sovereign yields, although not in countries where the government is in distress, too. A better alternative proved to be traditional safe-haven assets, such as Bunds and Gilts, although, again, the latter carry some currency risk in case of a euro recovery. There is also some risk around commodity returns. In our analysis, oil contributed positively to the overall portfolio return, but it has been positively correlated with stock prices recently, so that there is a danger that it might go the other way if equity markets were to fall significantly.
Christoph Schon is director of applied research and William Morokoff is head of research at Axioma