Specifying and estimating vector autoregressions using their Eigensystem representation

• Main Author: KRIPPNER, Leo
• Format: text
• Language: English
• Published: Institutional Knowledge at Singapore Management University 2024
• Subjects: Vector autoregression (VAR); companion matrix; eigenvalues; eigenvectors; Econometrics; Finance and Financial Management
• Online Access: https://ink.library.smu.edu.sg/skbi/42; https://ink.library.smu.edu.sg/context/skbi/article/1041/viewcontent/Specifying_and_estimating_vector_autoregressions..pdf
• Institution: Singapore Management University

This article introduces the principles and mechanics of the eigensystem vector autoregression (EVAR) framework, where a VAR may be specified and estimated directly via its eigenvalue and eigenvector parameters. Using explicit constraints on the eigensystem permits control of a VAR ís allowable dynamics, which is illustrated empirically with standard and time-varying VAR estimations specified to be always non-explosive.