Steps in Vector Autoregressive (VAR) and Vector Error Correction (VEC) Models

Steps in Vector Autoregressive (VAR) and Vector Error Correction (VEC) Models

1. Check co-integration using Johansen co-integration test
2. If you find any co-integration equation, go for VEC model. However, VAR is fine too for simplicity. Estimation parameters do not differ drasticially.
3. Decide on lag orders.
4. Perform diagnostic check.
5. If all good, perform out-sample forecast.
6. Check accuracy of both in-sample and out-sample forecast.

Video on how to perform VAR and VEC models in R will be posted in ResearchHUB YouTube channel soon.