The weakening of the US dollar against the euro since the beginning of 2017 has been a perplexing issue for market observers, central banks, and financial institutions alike. Following Donald Trump’s presidential election, the general expectation was that his administration’s proposed economic policies – including infrastructure spending, tax reforms, and deregulation – would bolster the US dollar’s value. However, the opposite occurred, raising concerns given the US dollar’s pivotal role in the global economy (Boz et al. 2018).
What factors can explain this unexpected behavior in the dollar-euro exchange rate? This article argues that shifts in two key interest rate spreads largely account for these fluctuations. This analysis is particularly relevant considering prior research, such as Cheung et al. (2017), which highlights the inherent difficulty in accurately forecasting the dollar-euro exchange rate outside of sample data.
It’s logical to begin by examining the exchange rate in relation to the interest rate differential between the United States and the Eurozone.
Figure 1 US dollar–euro rate and German–US 10-year spread
Figure 1 illustrates that after President Trump’s election in November 2017, the dollar initially strengthened, moving from $1.10 per euro in October to $1.08 in November and further to $1.05 in December. The election also spurred a rise in US interest rates as market participants anticipated that President Trump’s economic agenda would stimulate growth and inflation. Consequently, the spread between German and US 10-year bond yields narrowed from -1.76% in October to -1.95% in November, and then to -2.24%. Thus, the initial impact of President Trump’s election was higher US interest rates and a stronger dollar.
However, from January to September, this trend reversed. The interest rate differential favoring the US diminished to 1.85%, and the dollar depreciated to $1.19 per euro. Two primary factors likely contributed to this shift. Firstly, it became apparent during the spring that President Trump’s economic program was facing delays and might not be implemented as swiftly or comprehensively as initially anticipated. Secondly, the Eurozone economy experienced stronger growth, and investor confidence in the euro improved following a series of European elections. These included the Dutch parliamentary elections in March, the French presidential elections in late April and early May, and the French parliamentary election in June, all of which saw losses for euro-skeptic political parties.
Yet, starting in October, the correlation between the exchange rate and the German-US interest rate spread weakened significantly.
It’s also important to consider that yields on government bonds from financially weaker Eurozone countries might also influence the exchange rate. For example, concerns regarding public debt or political instability in these countries could lead to increased long-term yields and a simultaneous euro depreciation.
To investigate this further, Figure 2 displays the spread between bond yields in Germany and Spain, serving as an indicator of yield trends in the Eurozone periphery (note the inverted scale). The figure reveals some correlation between these variables. Notably, the euro’s strengthening at the end of the period coincided with a decrease in the spread between Spanish and German yields, from 130 basis points to around 85 basis points. This development would typically be expected to contribute to euro appreciation.
Figure 2 US dollar–euro rate and Spanish–German 10-year spread
To proceed with a more quantitative approach, we calculated the monthly changes in both interest rate spreads and the monthly percentage change in the exchange rate. Intriguingly, the correlation between changes in the German-US spread and changes in the exchange rate is 0.79, which is statistically significant. Similarly, the correlation between changes in the Spanish-German spread and changes in the exchange rate is -0.53, also statistically significant.
With these two potentially explanatory variables, we conducted a regression analysis, examining the change in the exchange rate against changes in the German-US spread and the Spanish-German spread. These two interest rate spreads together explained a substantial 78% of the exchange rate’s variation. Moreover, their impacts were both statistically significant and economically plausible. Specifically, a 10 basis point increase in German yields relative to US yields leads to a 1.1% appreciation of the euro (t = 8.5), contradicting the uncovered interest parity theory. Conversely, a 10 basis point decrease in Spanish yields relative to German yields results in a 0.6% euro appreciation (t = 4.1).
Figure 3 visually decomposes the changes in the exchange rate into components attributed to changes in the German-US yield spread and the Spanish-German spread. [2] The fluctuations in the German-US spread appear to have been the primary driver of the exchange rate from late 2016 to October 2017. Changes in the Spanish-German yield spread also appear to have played a significant role, particularly in early 2018.
Figure 3 Monthly changes in the US dollar–euro rate and the parts due to changes in the German–US and Spanish–German 10-year spreads
This analysis suggests that the dollar’s depreciation in 2017 was largely driven by an increase in long-term German bond yields relative to US yields, and a decrease in long-term Spanish yields (representing the Eurozone periphery) relative to German yields. However, it’s important to acknowledge that with only 17 data points in this specific sample, these initial findings might warrant cautious interpretation.
For this analysis to be more broadly applicable, it’s crucial to assess the model’s usefulness beyond this particular timeframe. Re-estimating the model using data extending back to the euro’s inception reveals that incorporating a lagged dependent variable is necessary. [3] Furthermore, with a larger dataset of 228 observations, the explanatory power does decrease to 22%. Nevertheless, the estimated parameters for changes in the German-US spread and the Spanish-German spread remain highly statistically significant. Importantly, a test for structural breaks at unknown points in time does not indicate any parameter instability.
The central takeaway from this analysis is that a significant portion of the fluctuations in the US dollar compared to the euro since President Trump’s election can be attributed to shifts in the relative attractiveness of holding US dollars versus the euro, as reflected in interest rate differentials.
References
Boz, E, G Gopinath and M Plagborg-Moller (2018) “Global Trade and the dollar,” VoxEU.org, February 11.
Cheung, Y-W, M Chinn, A Garcia Pascual and Y Zhang (2017), “Exchange rate prediction redux,” VoxEU.org, February 11.
Marcellino, M and A Abbate (2017), “Reducing the uncertainty around exchange rate forecasts: A new model,” VoxEU.org, February 4.
Rossi, B (2013), “Are exchange rates predictable?” VoxEU.org, November 14.
Endnotes
[1] See Rossi (2013) and Marcelliono and Abbate (2017) for discussion of difficulties forecasting exchange rates out-of-sample.
[2] The constant in the regression (0.92) has been subtracted from the change in the exchange rate.
[3] The parameter is 0.27 (t = 5.25).
