Economics Webinar - Exchange Rate Models are Better than You Think, and Why They Didn't Work in the Old Days

Contrary to popular belief, exchange-rate models fit very well for the U.S. dollar over the past quarter century. We find that a “standard” single-equation model that includes real interest rates for the U.S. and the ‘foreign’ country, a measure of expected inflation for each country, the U.S.comprehensive trade balance, measures of global risk and liquidity demand, and the lagged real exchange rate level is well-supported in the data for the U.S. against other G10 currencies. This explanatory power does not come just from the global risk and liquidity variables, and the fit of the model is not driven by the episode of the global financial crisis. The monetary variables are highly significant and important in explaining exchange rate movements. Importantly, the model is fit on monthly changes in exchange rates, which have proven in the past to be difficult to explain. We show that in the 1970s – early 1990s, the fit of the model was poor but the fit (as measured by t- and F-statistics, and R2s) has increased almost monotonically to the present day. We make the case that it is improved monetary policy (inflation targeting) that has led to the better fit, as the scope for “sunspots” to affect real and nominal exchange rates has disappeared.