Estimation of large dimensional time varying VARs using copulas

Tsionas, Mike G.; Izzeldin, Marwan; Trapani, Lorenzo

Abstract

This paper provides a simple, yet reliable, alternative to the (Bayesian) estimation of large multivariate VARs with time variation in the conditional mean equations and/or in the covariance structure. The original multivariate, n-dimensional model is treated as a set of n univariate estimation problems, and cross-dependence is handled through the use of a copula. This makes it possible to run the estimation of each univariate equation in parallel. Thus, only univariate distribution functions are needed when estimating the individual equations, which are often available in closed form, and easy to handle with MCMC (or other techniques). Thereafter, the individual posteriors are combined with the copula, so obtaining a joint posterior which can be easily resampled. We illustrate our approach using various examples of large time-varying parameter VARs with 129 and even 215 macroeconomic variables.

Citation

Tsionas, M. G., Izzeldin, M., & Trapani, L. (2022). Estimation of large dimensional time varying VARs using copulas. European Economic Review, 141, Article 103952. https://doi.org/10.1016/j.euroecorev.2021.103952

• Journal Article Type: Article
• Acceptance Date: Oct 18, 2021
• Online Publication Date: Nov 6, 2021
• Publication Date: Jan 1, 2022
• Deposit Date: Oct 20, 2021
• Publicly Available Date: Nov 7, 2023
• Journal: European Economic Review
• Print ISSN: 0014-2921
• Electronic ISSN: 1873-572X
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 141
• Article Number: 103952
• DOI: https://doi.org/10.1016/j.euroecorev.2021.103952
• Keywords: Economics and Econometrics; Finance
• Public URL: https://nottingham-repository.worktribe.com/output/6505084
• Publisher URL: https://www.sciencedirect.com/science/article/abs/pii/S0014292121002439?via%3Dihub