Monetary Policy and Exchange Rate Dynamics in a Behavioral Open Economy Model

Abstract

We analyze the implications of adding boundedly rational agents a la Gabaix (2020) to the canonical New Keynesian open economy model. We show that accounting for myopia mitigates several ``puzzling" aspects of the relationship between exchange rates and interest rates and helps explain why some of them only arise in the nested case of rational expectations. Bayesian estimation of the model demonstrates that a high degree of ``cognitive discounting" significantly improves empirical fit. We also show that this form of bounded rationality makes positive international monetary spillovers more likely and exacerbates the unit root problem in small open economy models with incomplete markets. On the normative side, the model with behavioral agents provides arguments against using the exchange rate as a nominal anchor.