Forecasting US Commercial Property Price Indexes Using Dynamic Factor Models

The general purpose of a dynamic factor model (DFM) is to summarize large number time series into few common factors. In this paper we explore several DFMs on 80 granular, non-overlapping commercial property price indexes in the US, quarterly from 2001Q1 2017Q2. We examine nature and structure factors index forecasts that can be produced DFMs. consider specifications one four As major motivation for use their ability improve out-of-sample forecasting systems numerous related series, apply DFM estimated an Autoregressive Distributed Lag (ARDL) forecast individual market returns. compare markets those benchmark univariate autoregression. results show & ARDL predicts crisis subsequent recovery really well, whereas typically extrapolates past trend.