摘要：This paper tests whether information derived from 144 economic variables (represented by only a few constructed factors) can be used for the forecasting of consumer prices in Croatia. The results obtained show that the use of one factor enhances the precision of the benchmark model’s ability to forecast inflation. The methodology used is sufficiently general to be able to be applied directly for the forecasting of other economic variables.
关键词：factor models; time series analysis; inflation; forecasting
