Large Bayesian VARs: A Flexible Kronecker Error Covariance Structure

Joshua Chan (2020)
Journal of Business and Economic Statistics, 38(1), 68-79
[ Journal Version | Working Paper | Code ]

This code estimates eight large Bayesian VARs with a variety of flexible covariance structures, allowing for non-Gaussian, heteroscedastic and serially dependent innovations.

The application is illustrated using a US dataset of 20 macroeconomic variables at quarterly frequency. This code also computes the marginal likelihood for each model.

The eight large Bayesian VARs are:

1. BVAR: standard BVAR with iid Gaussian innovations
2. BVAR-t: BVAR with t innovations
3. BVAR-CSV: BVAR with a common stochastic volatility
4. BVAR-MA: BVAR with MA(1) Gaussian innovations
5. BVAR-t-CSV: BVAR with a common stochastic volatility and t innovations
6. BVAR-t-MA: BVAR with MA(1) t innovations
7. BVAR-CSV-MA: BVAR with a common stochastic volatility and MA(1) innovations
8. BVAR-CSV-t-MA: BVAR with a common stochastic volatility and MA(1) t innovations