Social Learning and Higher Order Beliefs: A Structural Model of Exchange Rates Dynamics

This paper proposes a structural model of exchange rates where agents formulate their one-step ahead predictions based on social learning process and higher order beliefs. Individual choices are then aggregated and plugged into a rather standard macroeconomic model to derive the dynamics of exchange rates. Bayesian estimation of the structural parameters is implemented exploiting Foreign exchange Consensus Survey data of heterogeneous forecasts and fundamentals. Results show that higher order beliefs accounts for a large part of the total value, while public information play the most important role in determining individual expectations. Although the precision of the private signal is larger than the public one, information publicly revealed does exert a disproportionate influence, and differences in the estimated signals determine the equilibrium strategy of each agent as a combination between personal beliefs and higher order expectations.