Forecasting In a Non-Linear DSGE Model

A medium-scale nonlinear dynamic stochastic general equilibrium (DSGE) model is estimated (54 variables, 29 state variables, 7 observed variables). The model includes a observed variable for stock market returns. The root-mean square error (RMSE) of the in-sample and out-of-sample forecasts is calculated. The nonlinear DSGE model with measurement errors outperforms AR (1), VAR (1) and the linearized DSGE in terms of the quality of the out-of-sample forecasts. The nonlinear DSGE model without measurement errors is actually of a quality equal to that of the linearized DSGE model.