Empirical exchange-rate models fit the U.S. dollar very well in the 21st century. A “standard” model that includes real interest rates, a measure of expected inflation for the U.S. and the foreign country, and the U.S. comprehensive trade balance—augmented with measures of global risk and liquidity demand—is well supported by the data. From the 1970s to the 1990s, the fit of the model was poor, but it has improved for both traditional and risk variables almost monotonically to the present day. We provide evidence that better monetary policy has led to this improvement.

The global secular decline in long-term interest rates since the 1990s has been largely driven by cumulative yield changes during U.S. monetary policy announcement windows (FOMC windows). We document this FOMC-window effect for a group of ten major advanced economies. Based on a term structure model, we show that the effect operates through the expected path of future short-term interest rates rather than term premia. We further find that the global secular decline in interest rates has been concentrated in FOMC windows preceded by macroeconomic news. We interpret this pattern as a ``Fed response to news'' channel and provide empirical evidence to support this interpretation. 

The surge in post-COVID-19 inflation has raised critical questions about its persistence and underlying drivers. We employ a high-dimensional Factor-Augmented Vector Autoregression (FAVAR) model, utilizing Lasso techniques to simultaneously capture unobserved common factors, sectoral heterogeneity, and latent intersectoral spillovers. We find that spillovers play a dominant role in sustaining aggregate inflation, surpassing the effects of sectoral heterogeneity and common factors. During the post-COVID period, sector-specific shocks significantly contributed to the variance and persistence of aggregate inflation. Counterfactual analysis reveals that eliminating spillovers would have reduced average post-COVID inflation by 22%, with inflation reverting to the mean three quarters earlier. 

This paper studies the interaction between foreign exchange reserves and the currency composition of sovereign debt in emerging countries. Focusing on inflation targeting countries, we find that holdings of foreign reserves are associated with higher local currency sovereign debt, an exchange rate which is less sensitive to global shocks, and a lower exchange rate risk premium in local currency sovereign spreads. We rationalize these findings within a financially constrained model of a small open economy. The Sovereign values local currency debt as a hedge against endowment risk, but since the exchange rate tends to depreciate in times of global downturns, risk averse international investors charge an additional currency risk premium on this debt. When a country optimally uses foreign reserves to lean against the wind in response to global shocks, this dampens the response of the exchange rate, providing insurance for the global investor. By reducing the risk premium on local currency debt, foreign exchange reserves therefore facilitate a higher share of local currency debt in the sovereign portfolio. Quantitatively, we find the welfare benefits for the sovereign from optimal foreign reserves management can be very large.

Corporate external debt in emerging countries is very dollarized. We show this could create an externality to the sovereign and is reflected in sovereign spreads. Empirically, decomposing sovereign spreads into their credit default premium (default probability) and credit risk premium components, an increase in foreign-currency corporate debt is associated with a significant increase in the sovereign risk premium but does not change the sovereign default premium. We reconcile both findings in a quantitative model with risk-averse international investors, foreign-currency corporate debt makes the sovereign more likely to default in investors’ bad times when foreign-currency appreciates, thus increases the risk premium.