Typically, when conducting econometric forecasting, estimation is carried out on a forecasting model that is built upon some assumed economic structure. However, such techniques cannot avoid running into the possibility of misspecification, which will occur should there be some error in the assumptions underlying this economic structure. In this paper, in which we concentrate upon inflation forecasting, we present a method of hitting every vector autoregression (VAR) and forecasting under model uncertainty (HEVAR/FMU) that stresses statistical relationships among time-series data, and that makes no structural assumptions, other than to set up the underlying variables. Use of this HEVAR/FMU, in addition to establishing a more objective setting and enabling us to produce forecasts that take uncertainty into account, gives better results when forecasting qualitative movements in inflation. Therefore, we can state that the HEVAR/FMU can also play a valuable role in providing a cross-check for forecasts produced using such structural-type models.
Keywords: Inflation; Forecast; Reduced rank VAR; Kernel smoothing; Mixture distribution; Nonparametric test
Views expressed in the paper are those of the authors and do not necessarily reflect those of the Bank of Japan or Institute for Monetary and Economic Studies.