Some Thoughts on Time Series Analysis in Macroeconomics

In this talk I will bring together two strands of research. First, an investigation into the properties of finite lag VAR(n) models fitted to time series that in truth come from an infinite order VAR(∞) data generating mechanism that suggest  that with sample sizes and lag lengths like those commonly employed in practice inferences based on VAR models are likely to be very untrustworthy. Second, the development of a new methodology for identifying the structure of VARMA time series models that proceeds by examining the echelon canonical form and presents a fully automatic data driven approach to model specification using a new technique to determine the Kronecker invariants. A novel feature of the inferential procedures developed is that they facilitate the construction of algorithms which, from the perspective of macroeconomic modeling, are efficacious in that they do not use AR approximations at any stage. The ideas and techniques discussed are illustrated using an example drawn from the real business cycle literature.

The talk is based around:
Vector Autoregressions and Macroeconomic Modeling:  An Error Taxonomy. Journal of Business and Economic Statistics, 2015.
Vector Autoregressive Moving Average Identification for Macroeconomic Modeling: A New Methodology, Journal of Econometrics, 2016.