Three things are sure in life: death, taxes, and betting against the calls of Goldman's Thomas Stolper. Sure enough:
- We lower our EUR/$ forecast path slightly but keep the same upward-sloping trajectory.
- Our new EUR/$ trajectory is 1.40, 1.45 and 1.50 in 3, 6 and 12 months, from 1.45, 1.50, 1.55 previously.
- The recent increase in the Euro area’s fiscal risk premium is likely to persist.
- Very large short EUR/$ positioning is likely to last in the near future.
- But the underlying Dollar downtrend should drive EUR/$ higher over time.
- We discuss the CHF and safe-haven currencies after the SNB’s commitment to intervene.
- Our new EUR/CHF forecasts are 1.21 flat in 3, 6 and 12 months.
Over the past couple of weeks, the fiscal tensions in the Euro area have risen once again. FX markets have reacted via a substantial increase in speculative short EUR positions, which likely reflects the increase in the fiscal risk premium. We expect only slow progress towards improved fiscal policy and coordination in the Euro-zone, and hence the risk premium may remain wide. At the same time, the Dollar is still under much broader depreciation pressures. We therefore continue to expect EUR/$ to rise again, albeit from a slightly lower starting point. We are also adjusting our EUR/CHF forecast to reflect the recent SNB action.
Lastly, we look at alternative safe-haven currencies.
1. What Has Changed Relative to Our Core Assumptions?
In our last Global FX Monthly Analyst before the Summer, we discussed in detail the different transmission channels from fiscal policy to FX, which can be broadly broken down into a risk premium and a cyclical impact. These transmission channels typically work in opposite directions, in particular if debt levels are relatively high. Our conclusions at the time were that the negative cyclical forces in the US will likely dominate, given the high debt levels in the US and the ongoing debate about the debt ceiling. In Europe, we worried about an increase in the fiscal risk premium and the escalation of concerns regarding Italy, but assumed a relatively benign outcome after temporarily increased volatility.
We also assumed that much of the cyclical slowing induced by the spike in oil prices at the beginning of the year and the Japanese earthquake was past, and had already moved to a more pro-cyclical stance. With hindsight, we now note the following differences relative to our assumptions:
- Fiscal risk premia in the Euro area have risen sharply. The negotiation and implementation of the second Greek program, as well as the frictions surrounding the EFSF enhancement agreed on July 21, led to a substantial increase in default and contagion fears. In response, several countries, including Italy and even France, implemented additional fiscal tightening moves. These failed to build market confidence and, as a final backstop, the ECB was obliged to intervene—reluctantly—in Italian and Spanish bond markets and will likely continue to conduct open-ended purchases for the foreseeable future. This, in turn, could lead to a substantial increase in the ECB’s balance sheet. None of these measures and events was totally unexpected, as discussed in our last FX Monthly. However, the extent of the friction and the relative weakness of the policy response was not our core scenario.
- At the same time, the intense US debate about fiscal consolidation led to relatively little additional tightening as the debt ceiling debate was essentially pushed out until after the next presidential election. In addition, last week President Obama presented an ambitious stimulus program to boost job growth. The details of the latter will depend on approval in Congress, but overall it appears that fiscal policy is not heading for very frontloaded consolidation.
- Lastly, activity data remained weaker than expected and survey data in particular weakened a lot more than we would have thought. Stock markets fell sharply in response to the fiscal concerns in Europe and the US, possibly feeding into additional weakness in the survey data. We did not anticipate the observed weakness in cyclical assets. In response to the weaker cyclical data over the summer and rising growth risks, many central banks, including the ECB and the Fed, have shifted to a more dovish bias. Markets now firmly expect additional unconventional easing by the Fed, and front-loaded rate cuts by the ECB.
The result of these events and developments was that, instead of a continuation of the Dollar downtrend, the USD remained broadly range-bound (see above). Moreover, in recent weeks, we have seen EUR weakness, linked to the intensifying Euro-zone tensions. As so often over the past 18 months, these tensions led to a sharp increase in short EUR/$ positions but relatively little evidence of genuine capital outflows from the Euro-zone. EUR/$ risk reversals have shifted substantially towards favouring put options, at the same time as sovereign CDS have widened (see the chart on the previous page). Our EUR/$ Sentiment index shows the largest short positions since the first round of sovereign tensions last year (see the next page).
Overall, fiscal developments have at the margin not been as big a drag on the Dollar as we previously expected, whereas Euro-zone fiscal stress was more intense and cyclical data weaker than we thought.
2. How Persistent Will These Changes Be?
Looking at the three key issues discussed in the previous section, the Euro-zone’s fiscal risk premium may well remain elevated for the next couple of months, for the following reasons:
- Greece continues to be very slow in implementing the admittedly very large reform package. Repeated stand-offs between the Greek authorities and IMF, ECB and EU over the next couple of quarters would constantly raise the spectre of a possible Greek EMU exit and related contagion risks.
- Given the uncertainty about the strength of EMU bank balance sheets in the face of possible sovereign default, Euro-zone banks may remain under pressure as well. As a reflection of this, the cross-currency basis in EUR/$ has widened out to levels similar to September 2008, signifying that USD funding pressures exist despite the ECB-Fed swap agreement.
- At the same time, high debt levels in some EMU countries reduce the ability of these national governments to inject sufficient capital into their banking sector to protect against the risks of a sovereign default.
- The enhanced EFSF, which could act as a circuit breaker for the negative sovereign-bank capital circularity is not yet up and running. National governments in many EMU countries still need to implement the July 21 summit agreement.
- As the only credible backstop for further escalation and contagion, the ECB could be forced to continue with intervention in sovereign bond markets. The related increase in credit exposure could put reputational pressure on the ECB, and increase fears of ultimate monetisation.
Against the backdrop of these near-term concerns, the Euro-zone continues to make gradual progress with the implementation of structural reforms. For example, the enhanced EFSF currently being approved by national parliaments is effectively already the third generation bail-out vehicle in about 18 months, filling an institutional gap that persisted during the first 11 years of the Euro area. Moreover, many peripheral countries have implemented in a very short period of time structural reform packages that have been at the top of the wish list of most economists for many years. We continue to believe that at some stage the markets will begin to acknowledge these reforms. This, in turn, should lead to a permanent reduction of the fiscal risk premium from very high levels currently. It is also worth noting that market pressures tend to have a positive impact on the speed of policy implementation. In that respect, undisputable improvements in Greek compliance with the IMF/ECB/ EU program and the Italian debt trajectory, as well as swift implementation of the July 21 summit agreement, could lead to a faster decline in the risk premium. We also think it is important to recognise just how powerful ECB bond purchases are to cap yield levels in the periphery.
On the US side, the medium- to long-term fiscal trajectory is also still very worrisome and, ultimately, fiscal tightening is likely to become a major negative for the US Dollar. However, this change may still be many months, or even many years, away. The extent of fiscal tightening in the US will in the end be larger than in the Euro area, partly because the US still runs much larger deficits and the debt level is higher.
Overall, we therefore do not expect the fiscal influence on the EUR and the USD to change much in the coming months. It may, therefore, take a while longer before now very sizable EUR/$ short positions (also visible in our Sentiment Index, see above), are unwound.
The expected persistence of a larger fiscal risk premium in the EUR also suggests that the recent shift in EUR/$ may prove longer-lasting. As a result, we have decided to shift the path of our forecasts slightly lower. This reduced starting point has to be thought of as a direct reflection of a slightly larger fiscal risk premium. We now forecast 1.40, 1.45 and 1.50 on a 3-, 6- and 12-month basis, from 1.45, 1.50 and 1.55 previously.
3. Key Factors That Have Not Changed
Despite the reduced starting point for our forecast path, we have not changed the underlying trajectory. The dominating trend in global FX remains broad Dollar depreciation, in our view.
Continued balance of payments pressures are the main reason for this. As we have pointed out frequently elsewhere, despite the post-housing bubble demand weakness in the US, the trade deficit remains worryingly wide. Looking through the month-to-month volatility in the US trade balance, we see no sign of trend improvement, even though the real trade-weighted Dollar remains close to record lows, and much weaker than during last year’s Euro-zone sovereign crisis.
In terms of monetary policy, the Fed moved in August to an even more dovish stance and continues to conduct one of the most accommodating monetary policies globally. Even relative to the ECB, which has also become more dovish, relevant interest rate differentials have not moved in a meaningful way. For example, the 5-year swap rate spread, which at this point may be a better measure of monetary policy expectations than the 2-year differential, has barely moved in most crosses. Any evidence of a global reacceleration in economic activity would likely lead to a sell-off in many fixed income markets outside the US.
Closely related to the points above, the persistent weakness in US activity, the dovish stance by the Fed and the substantial long-term challenges on the fiscal side all suggest that more attractive investment opportunities may exist outside the US. Even US investors have shown a clear preference for foreign stocks relative to domestic stocks (see charts). Together with the widening trade deficit, this exerts ongoing downside pressure on the USD.
Another factor that has not changed is the correlation to other risky assets. In particular, the correlation in daily returns between the trade-weighted Dollar and the SPX has remained virtually unchanged throughout the whole recent period of rising risk aversion. This suggests that weaker asset prices did have a supporting impact on the Dollar, but this was offset by other, Dollar-negative factors. With a substantial amount of cyclical damage now priced into most risky assets, risk correlations could quickly become a Dollar-negative factor in the event risk sentiment and cyclical data improve relative to low expectations.
4. The Distinction between Flow and Level Effects in FX
When thinking about the influences discussed in previous sections, it is worth making a distinction between flow effects and stock effects. Flow effects drive the steepness of a trend, whereas stock effects only lead to one-off changes in the level of the exchange rate. Specifically, balance of payment imbalances are flows, which translate into the slope of a trend, as our research on the BBoP has shown the past. On the other hand, changes in the fiscal risk premium or the monetary policy stance are one-off level effects.
It is relatively easy for markets to ‘price in’ a change in a stock variable, such as the fiscal risk premium. On the other hand, a trend influence from the balance of payments may change the speed of Dollar depreciation for an unknown period, making it impossible to say when a change in flow has been fully ‘priced in’.
Putting this into the context of recent events, our core assumption is that the trend components of EUR/$ have not changed, as they are mainly driven by balance of payments flows. What has changed is the Euro-zone fiscal risk premium, which effectively shifted the underlying uptrend to a lower starting point. The increasing fiscal risk premium has not affected the underlying uptrend in EUR/$, which remains a function of the size of the US BBoP deficit.
5. SNB Acts to Prevent Disruptive CHF strength
Despite our focus on EUR/$ so far, the intensifying Euro-zone fiscal tensions have also clearly affected other currencies—and none more so than the Swiss Franc, which at one stage almost touched parity against the Euro. The SNB reacted with a credible commitment to prevent EUR/CHF from falling below 1.20. This potentially entails CHF weakening intervention of unlimited size.
When the SNB tried a similar policy in the 1.40-1.50 range last year, it failed and the CHF ended up appreciating substantially. Many market participants therefore wonder if the same could happen again.
We think the likelihood of successful intervention is now much higher for the following reasons:
- The political support for intervention to weaken the CHF is now much stronger in Switzerland, as there are clear signs of currency strength affecting the real economy.
- At EUR/CHF 1.20, the Swiss Franc is now clearly overvalued, which was not the case at 1.50. Our GSDEER ‘fair value’ estimate for EUR/CHF is close to 1.45.
- Contrary to the last round of intervention, the SNB is now facing deflationary pressures, as visible in the rapidly falling CPI readings. Quantitative easing via FX intervention appears to be an appropriate policy.
At the same time, we think CHF appreciation pressures persist from the ongoing tensions in the Euro area and related safe-haven flows. Moreover, the potential for further unwinding of legacy carry trades funded out of CHF remains high—for example, Hungarian mortgages or exotic derivatives held by French municipalities. With this offsetting force, EUR/CHF will therefore likely continue to trade fairly closely to the SNB intervention level.
Consequently, we are changing our EUR/CHF forecasts to 1.21 flat for 3, 6 and 12 months.
6. The Search for New Safe-Haven Currencies
Store-of-value considerations: The recent stand-off over the US debt ceiling and the escalation of the Euro-zone crisis have reignited the debate over what exactly constitutes a ‘safe-haven’ currency. The answer to this question needs to be addressed urgently, especially in the wake of the SNB’s decision to counter safe-haven appreciation pressure on the CHF with a peg of EUR/CHF at 1.20.
Whether the CHF has actually lost its safe-haven status, as some claim, is rather questionable, as we will see below.
In a very broad sense, safe-haven currencies are often seen as having certain attributes that tend to reduce the risk of devaluation in the long run, such as:
- External surpluses, making countries net lenders to the world, are often a characteristic of safe-haven currencies.
- A stable institutional and political framework is also considered a pre-requisite, in particular because this typically translates into strong property rights, efficient banking sectors and a free flow of capital.
- Low levels of sovereign indebtedness are also positive, while low inflation rates help to avoid a rapid erosion of FX valuation.
- Lastly, a high degree of FX liquidity is also important.
We looked at the characteristics of different safe-haven currencies from a theoretical perspective. We found that the JPY and CHF combine most of these characteristics.
The JPY is the best illustration of a country where solid external balances, low inflation and a stable political framework have supported real exchange rate appreciation over long periods of time. The JPY has appreciated in real effective terms by 86.7% since 1970 against other major currencies according to BIS data, even if in recent years our GSDEER valuation model suggests it is overvalued. This appreciation record—and, perhaps more importantly, the ability of the economy to absorb appreciation—creates a perception in the market that the currency is a good ‘store of value’.
In Switzerland, the combination of a strong institutional set-up, low inflation, low debt and large positive external surpluses have made the CHF traditionally a safe-haven currency. In terms of the institutional aspect, the large size of the banking system (at around 500% of GDP in recent years) has driven policy makers typically to favour a strong CHF, as efforts to stem CHF appreciation have historically only been taken reluctantly and with delay, for fear that they could spark a deposit run among foreign depositors and hence weaken the banking system. Bordo et. al. (2006) provide a discussion of this dynamic during the 1930s, when Switzerland left the Gold Standard much later than others, incurring substantial declines in output as a result.
The strength of its safe-haven status is also mirrored in the high ranking that the CHF achieves in our score card above. And these factors are unlikely to change because of the SNB’s decision to counter deflationary pressures via FX invention. Moreover, the CHF has not been a safe haven primarily because people expect appreciation. For most people, the CHF appreciation to current levels was an unexpected ‘bonus’, while the other characteristics dominate. Even if EUR/CHF continues hovering close to 1.20 for a long time, and assuming no change to the inflation profile, the Swiss Franc should remain a pretty safe place.
The Deutsche Mark has also worked as a safe haven in the past, mostly due to the conservative monetary policy of the Bundesbank and the ECB, influenced by memories of the German hyperinflation experience in the 1930s. At certain points in time, the EUR has also behaved like the DEM. However, the recent crisis has significantly increased the risk premium on the single currency, as we discussed above.
All told, however, none of the attributes described and exemplified above are either necessary or sufficient, with the USD being perhaps the most prominent outlier. The US runs a large current account and broad balance of payments deficit, is a net debtor to the rest of the world, had double-digit inflation in the 1980s and is a large commodity importer. Yet, the USD tends to strengthen in risk-off episodes, most visibly during the global financial crisis in 2008/09. That said, the USD benefited from many of these factors when it cemented its status as primary reserve currency in the first half of the last century. And the reserve currency status of the USD has been weakening in recent years, as reflected in ongoing FX reserve allocation shifts.
Risk correlations. At an important level, above and beyond the theoretical characteristics discussed in the previous section, a safe-haven currency also needs to perform strongly at times of risk aversion. As mentioned above, the USD has been the key currency to display this countercyclical behaviour, particularly since the 2008 crisis. The JPY and the CHF also possess this feature the chart below shows weekly return correlations with the SPX over the past 12 months. Most currencies, except for the CHF and JPY, have depreciated against the Dollar (even marginally) at times of SPX decline, with the AUD, PLN, NZD, ZAR and HUF most correlated to risk.
This raises a few issues from a strategy perspective. With the SNB capping the upside of the CHF in a determined fashion and the risk of intervention looming in abrupt JPY appreciation moves, should investors count on these correlations remaining in place—and thus continue to use the JPY and CHF as safe-haven solutions? And with the Fed possibly moving into an easier policy stance, will long USD positions offer a good enough offset to risk aversion episodes? For now, it is reasonable to assume that at least the CHF, USD and JPY will not depreciate vs other currencies in the near term as risk aversion picks up.
Interestingly, the CNY is starting to look interesting both in terms of a strong underlying fundamental backdrop (external surpluses, low debt) and from a correlation perspective; its tight management against the Dollar makes it follow the greenback at times of risk aversion. If the institutional set-up and market liquidity were to improve over the years, the CNY could become a safe haven. However, these factors are not yet in place, in particular as the CNY is not yet fully convertible, a key pre-requisite for a safe-haven currency.
Beyond the outright correlation between currencies and risk sentiment, it is important to understand how shifts in risk aversion affect currencies on a relative basis, while accounting for other macro factors, including Dollar direction. To do this, we use our FX betas framework, which we revamped in our latest annual Foreign Exchange Market publication. The framework is based on regressions of weekly returns on different currencies against a number of macro factors, including local 2yr rates, US 2yr rates, risk reversals (as proxies for positioning), Dollar direction (proxied by USD TWI returns), oil prices and SPX returns. Looking at the betas to the SPX, we can gain a sense of which currencies benefit most and which currencies are hurt most from shifts in risk sentiment, all other things considered.
We found that a basket of short AUD, MXN and ZAR vs long CHF, EUR and GBP would offer a good currency hedge to pure risk aversion spikes. Interestingly, this exercise suggests that currencies with sound fundamentals (such as the CAD, SEK, NOK and SGD) would depreciate vs both the USD and the EUR at times of intense risk aversion and hence should not be considered safe havens for USD-based or EUR-based investors. Of course, this finding is conditional on historical correlations remaining in place. For example, if volatility in the Euro-zone escalates from here, or if the UK re-engages in QE, these results would be subject to change.
7. Other Forecast Changes
Besides the EUR/$ and the EUR/CHF revisions mentioned above, we have made several forecast changes reflecting the increased financial stress in Europe and the gloomier US outlook. These are most reflected in the weaker forecasts of the MXN and CLP. Most other changes mainly reflect the recent moves, which we expect to be relatively short-lived, before the fundamentals once again become the main driver. The revisions in the ARS and PEN are due to local factors, as we expect the Peruvian central bank to turn more hawkish and forecast accelerated portfolio dollarisation in Argentina.