Forecasting with a noncausal VAR model

Abstract

We propose simulation-based forecasting methods for the noncausal vector autoregressive model proposed by Lanne and Saikkonen (2012). Simulation or numerical methods are required because the prediction problem is generally nonlinear and, therefore, its analytical solution is not available. It turns out that different special cases of the model call for different simulation procedures. Simulation experiments demonstrate that gains in forecasting accuracy are achieved by using the correct noncausal VAR model instead of its conventional causal counterpart. In an empirical application, a noncausal VAR model comprised of U.S. inflation and marginal cost turns out superior to the bestfitting conventional causal VAR model in forecasting inflation. Keywords: Noncausal vector autoregression, forecasting, simulation, importance sampling, inflation. JEL codes: C32, C53, E3l.AC