Explaining the Great Moderation Exchange Rate Volatility Puzzle 
This paper seeks to contribute to a better understanding of the drivers of trends in exchange rate volatility. Using data from the last five decades, the authors find that, against the USD, the volatility of financial center currencies (CHF, DEM/EUR, and JPY) has declined, while the volatility of commodity producer currencies (AUD, CAD, and NZD) has risen. Through the prism of the Great Moderation hypothesis—that macroeconomic volatility has decreased around the world since the mid-1970s—they study how macroeconomic factors relate to these trends. The authors do so using both a reduced-form approach and a novel asset-pricing-based decomposition of exchange rate changes that is disciplined with professional forecasts.
Key Findings
- The disparate trends in exchange rate volatility cannot be explained by differences in macroeconomic volatility.
- The main source of the decrease in the volatility of financial center currencies appears to be the declining variance of the currency risk premium term.
- One explanation for the increase in exchange rate volatility of commodity producer currencies involves the expectations of smaller policy rate responses to movements in inflation.
- A second explanation for the increase in the volatility of commodity producer currencies involves the weakening of the Fama puzzle—the idea that a greater realized excess return from being long a particular country’s currency and short the USD is associated with a larger differential between that country’s interest rates and US interest rates.
Implications
There are disparate trends in exchange rate volatility over the past few decades that cannot be explained by patterns in macroeconomic volatility. Indeed, patterns in the volatility of non-macro components, such as risk premia, are a main driver of cross-currency patterns in exchange rate volatility. However, the behavior of macro components of exchange rates still plays a major role in the rising volatility of commodity currencies through their changing co-movement with the non-macro components of exchange rates.
Abstract
In this paper, we study how the volatility of both realized and expected macroeconomic variables relates to the variation in exchange rate volatility through the prism of the Great Moderation hypothesis. We find significant heterogeneity in exchange rate trend volatility across currency pairs despite decreases in the volatility of expected future interest rate differentials and of realized yields themselves. We argue that time variation in the relationship between macroeconomic variables and exchange rates has prevented the Great Moderation in realized yield volatility from translating to a decrease in exchange rate volatility. Considering a Campbell-Shiller-type decomposition of exchange rate changes into forward-looking components linked to inflation, policy rate, and currency risk premia expectations, we find that the Great Moderation in volatility of expected yield differentials cannot explain the patterns in exchange rate volatility we observe. The main drivers of these patterns were trends in the volatility of the currency risk premium component and in the covariance between the components capturing the strength of the Fama puzzle and the expected responsiveness of monetary policy to inflation.