1. EURUSD long-term drivers
With US interest rates testing their multi-year highs on the long-end of the curve (i.e. 10Y trading at 3.22%), some market participants have argued that the US Dollar should strengthen against all other currencies. Exchange rates have had many fundamental drivers through history, however FX strategists have struggled to come up with a FX Fair Value model that has consistently performed well during a long period of time. For instance, a popular driver widely used in the market is the interest rate (IR) differential. Figure 1 (left frame) shows that the EURUSD exchange rate tends to co-move strongly with the 10-year German-US IR differential over time, and therefore some analysts have concluded that the rise in the 10Y US yield should benefit to the US Dollar and that the current IR differential is pricing in a much lower exchange rate (around parity).
However, we know from empirical studies that currencies’ drivers vary all the time depending on the macroeconomic situation, hence the divergence between the two times series can widen for a significant amount of time before closing out. Another interesting long-driver of the EURUSD exchange rate is the natural rate of interest differential (Holston, Laubach and Williams (2016)) estimated by a version of the Laubach-Williams model (2003). The natural or ‘equilibrium’ real interest rate, firstly introduced by Knut Wicksell in 1898 and sometimes referred as r*, provides a benchmark for measuring the stance of monetary policy in most of the economies. Figure 1 (right frame) plots the long-term relationship between the r* differential between the Euro area and the US and the EURUSD exchange rate since 1975 (using a proxy of the EURUSD prior 1999). We can notice that after a long decline from 50bps in 2011 to -100bps in early 2015, the r* differential has been rising over the past couple of years as the natural real interest rate in the Euro area moved from -45bps in the middle of 2014 to roughly 10bps in Q1 2018. Hence, if we use that proxy as a LT driver of the exchange rate, EURUSD should continue to rise as the r* differential increases.
Source Eikon Reuters, Holston et al. (2016)
Another important LT driver that some practitioners like to use is the current account differential. We saw previously that significant successive ‘twin’ deficits (current account and budget deficits) tend to weight on the country’s currency in the long run. For instance, as we expect the twin deficit to widen in the US on the back of a large fiscal stimulus (1tr USD deficits for the fiscal year 2019 and 2020 according to the CBO), the US Dollar should underperform against other strong currencies (i.e. Euro). In Figure 2 (left frame), we can notice some co-movements between the USD index and the US twin deficits since the early 1970s.
In the Euro area, the relationship is less visible when we plot the two times series starting 1997, however the exchange rate and the current account differential (EU-US) tend to not deviate too far away from each other in the long run. Over the past few years, the current account differential has constantly been increasing in favour of the Euro area (up from 4% in early 2013 to over 5.5% in the middle of 2018) while the EURUSD exchange rate has been depreciating on the back of a monetary policy divergence. As a reminder, the Fed stepped out of the bond market in the end of October 2014 while the ECB started its QE bazooka in March 2015. Now that the ECB is ‘exiting’ the market at the end of the year, we will see if fundamentals such as the current account will start to matter once again.
Source: Eikon Reuters, CBO
A popular model to ‘value’ exchange rates in the medium is the Behavioural Equilibrium Exchange Model (BEER), which estimates the FX fair value according to a set of macroeconomic and price data. For instance, one of our BEER models uses the interest rate (10Y), the inflation rate and the terms-of-trade differentials to value currencies. Figure 3 (left frame) shows the fluctuations of the EURUSD spot rate with its BEER value since 1991. The nominal exchange rate is much more volatile than its macroeconomic neutral rate, however we can notice that spot rates mean revert if they diverge too far from their equilibrium values (statistically significant when testing for mean reverting process using an ECM model). Currently, our BEER model is pricing a fair value at around 1.13, which is slightly lower than the current level of EURUSD (1.15). However, results can change drastically depending on the variable we use for the analysis. As we included the interest rate differential, we knew originally that we were not going to get a very elevated EURUSD fair value as IR differentials have been trending down over the past few years (as shown in Figure 1, left frame).
Another important chart that we like to look at is the Fed-to-ECB balance sheet total assets ratio. When a central bank is intervening in the market through its open market operations, the dynamics of its balance sheet changes and may impact the currency. For instance, between September 2012 and May 2014, the balance sheet of the ECB was contracting on the back of the early LTROs reimbursements by banks, while the Fed was running QE3 purchasing USD 85bn of Treasuries and MBS through the entire year 2013. As a result, the Fed-ECB ratio increased from 0.72 to 1.5 during that period, which played in favour of the single currency (EURUSD went up from 1.20 in August 2012 to 1.40 in May 2014). After reaching a high of 1.86 in March 2015, the ratio inverted as the ECB started its QE program that month. The Euro started to depreciate after the May 2014 meeting as market participants were already anticipating that move after Draghi’s ‘ready to act’. In short, monetary policy divergence should be automatically reflected in the exchange rate.
Source: Eikon Reuters, Rothko Research
2. Euro outlook
Since the beginning of the year, two important drivers of the EURUSD exchange rate have been the US Dollar strength that kicked off in mid-April, shorty followed by a sudden rise in political uncertainty in Italy. As we can see in figure 4 (left frame), the 10Y yield in Italian bonds suddenly jumped following the March elections, soaring from 1.7% to over 3%, and has been weighing on the Euro since the beginning of the summer. We usually see the Peripheral-Core yield spreads as a fundamental short-term driver of EURUSD. As long as the political instability persists in Italy, LT yields will remain elevated, depriving the Euro from appreciating.
For short term traders, it is also interesting to watch the behaviour of the EURUSD vis-à-vis the VIX index. If we plot a 30-min chart (see appendix 1), we can notice that the two times series have co-moved significantly since June. EURUSD tends to act as a risk-on asset, which means that the Euro appreciates (resp. depreciates) when the implied volatility is decreasing (spiking). There were some periods when the Euro was positively correlated to the VIX, which means that the single currency was behaving like a safe-haven asset (see what the EURUSD did during the August 2015 sell-off in Appendix 2).
Overall, there are a lot of challenges in the short-term for the Euro – elevated political uncertainty in Italy, slowing fundamentals, negative reactions to a sudden risk-off environment – however the single currency remains significantly undervalued looking at a variety of ‘fair’ value metrics. For instance, the Purchasing Power Parity (PPP) index is pricing an exchange rate at 1.33 according to the yearly Eurostat-OECD calculations. We could see EURUSD hitting new lows in the short run (1.11 is the next support in case of a sudden sell-off), however we think that the Euro has more upside potential in the medium / long term against the US Dollar, and especially if the situation improves in Italy.
Source: Eikon Reuters, OECD
Appendix. EURUSD (yellow line, rhs) vs. VIX (inv, lhs)
Source: Eikon Reuters