Working Paper: NBER ID: w24059
Authors: Charles Engel; Dohyeon Lee; Chang Liu; Chenxin Liu; Steve Pak Yeung Wu
Abstract: Recent research has found that the Taylor-rule fundamentals have power to forecast changes in U.S. dollar exchange rates out of sample. Our work casts some doubt on that claim. However, we find strong evidence of a related in-sample anomaly. When we include U.S. inflation in the well-known uncovered interest parity regression of the change in the exchange rate on the interest-rate differential, we find that the inflation variable is highly significant and the interest-rate differential is not. Specifically, high U.S. inflation in one month forecasts dollar appreciation in the subsequent month. We introduce a model in which a Taylor rule determines monetary policy, but in which not only monetary shocks but also liquidity shocks drive nominal interest rates. This model can potentially account for the empirical findings.
Keywords: exchange rates; Taylor rule; uncovered interest parity
JEL Codes: F3; F31; F41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| US inflation (E31) | exchange rate changes (F31) |
| interest rate differentials (E43) | exchange rate changes (F31) |