Switching Dynamic Factor Models for High-Dimensional Time Series

• Activity Number: 576
• Type: Invited
• Date/Time: Wednesday, August 12, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Rong Chen*
• Companies: Rutgers University

Abstract

We consider a factor model for high-dimensional time series with a regime switching dynamics. The switching is assumed to be driven by an unobserved Markov chain; the mean, factor loading matrix and covariance matrix of the error process are different among the regimes. The model is an extension of the traditional factor models for time series and provides flexibility in dealing with real applications in which underlying states may be changing over time. We propose an iterative approach to estimate the loading spaces of each regime as well as the states of the hidden Markov chain at each time, by combining eigenanalysis and Viterbi algorithm. The theoretical properties of the procedure are investigated. Simulation results and analysis of a real example are presented.