Working Paper: NBER ID: w4298
Authors: Graciela L. Kaminsky; Karen K. Lewis
Abstract: A frequently cited explanation for why sterilized interventions may affect exchange rates is that these interventions signal central banks' future monetary policy intentions. This explanation presumes that central banks in fact back up interventions with subsequent changes in monetary policy. We empirically examine this hypothesis using data on market observations of U.S. intervention together with monetary policy variables, and exchange rates. We strongly reject the hypothesis that interventions convey no signal. However, we also find that in some episodes, intervention signaled changes in monetary policy in the opposite direction of the conventional signaling story. This finding can explain why in some periods exchange rates moved in the opposite direction of that suggested by intervention.
Keywords: Foreign Exchange Intervention; Monetary Policy; Signaling Hypothesis
JEL Codes: F31; F33; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Foreign exchange interventions (F31) | future monetary policy changes (E52) |
| Federal Reserve sells domestic currency (F31) | future expansionary monetary policies (E49) |
| Interventions perceived as correct signals (C92) | significant exchange rate movements in intended direction (F31) |
| Interventions perceived as incorrect signals (D91) | opposite movements in exchange rates (F31) |