‘Time is money’: An heterogeneous agent model for the FX

Abstract

This paper deals with a Bayesian extension of a behavioral finance framework ‘à la’ De Grauwe and Grimaldi (The Exchange Rate in a Behavioural Finance Framework, Princeton University Press, Princeton, 2006) in which agents operating in the FX market differ in their forecasting time horizon for the exchange rate. In the short run, if we believe in the world described by Meese and Rogoff (J. Int. Econ., 14(1–2):3–24, 1983), this leads to a chartist rule, whereas in the long run, the PPP condition appears as a natural anchor. In between, i.e. in the medium run, we implement an APEER model using Bayesian tools, as an alternative to the FEER-BEER nexus. Our results show that the stabilizing impact of the intermediate rule depends on agents’ good perception of the fundamentals.