We provide a parametrization of Vector Autoregression (VAR) that enables one to examine the parameters associated with unit root dynamics and those associated with stable dynamics separately. The parameterization is based on a factorization of the VAR polynomial that partitions the polynomial spectrum into unit, stable, and zero roots via polynomial factors. The proposed factorization adds to the literature of spectral factorization of matrix polynomials. The main benefit of the parameterization is that actions could be taken to model the dynamics due to a particular class of roots, e.g. unit roots or zero roots, without changing the properties of the dynamics due to other roots. For example, using the parameterization one is able to estimate the co-integrating space with appropriate rank that maintains the root structure of the original VAR processes; or one can estimate a reduced rank causal VAR process maintaining the constraints of causality. In essence, this parameterization provides the practitioner an option to perform estimation of VAR processes with constrained root structure (e.g., co-integrated VAR or reduced rank VAR) such that the estimated model maintains the assumed root structure. A connection to the stability of higher order discrete dynamical systems is also discussed. An application to nowcasting of the Transportation Services Index is provided.
Tucker McElroy is a time series researcher at the U.S. Census Bureau, where he has been working since 2003. He graduated with a B.A. from Columbia in 1996, and a Ph.D. in mathematics from University of California San Diego in 2001. His research interests are multivariate time series modeling, signal extraction, and frequency domain methods. In his free time he practices Muay Thai and writes poetry.
Monday, September 26, 2022
12:00PM – 1:00PM
In-Person at UFII Office
E252 CSE Bldg
Lunch will precede talk at 11:30AM
Luncheon and talk are free, but registration is required. Please register below.