Monetary reaction functions in a small open economy

Monetary reaction functions are constructed in an intertemporal optimization framework. The time-varying coefficients to be put on the variations of the exchange rate and the asset stock are then derived by solving a matrix Riccati equation. As the planning horizon extends to infinity, a related algebraic Riccati equation coincides with those derived from minimization of stationary state variances only under certain conditions. Under the assumed full information by the government, reaction functions using realized values of exogenous shocks result in dynamics that do not enjoy the same stability property as the ones obtained by explicit intertemporal optimization.