My intention is to retrieve forecast error variance decomposition from a MF-VAR with no latent processes following Ghysels (2012)
I found the following site:
that has MatLab codes for this, but would be more comfortable with using R instead.
As far as I understand the mfbvar-package in R does not bend to this. I also looked at the midasr-package, but am not convinced it can do the trick. Are there any relevant packages I am missing or could the aforementioned packages be modified to suit a general MF-VAR?