BBE FAVAR replication code does not run with RATS Version 8

[hawkey]

I am really new to the RATS software. My aim was to replicate the FAVAR model by Bernanke, Boivin & Eliasz (2005), "Measuring the Effects of Monetary Policy: A Factor-augmented Vector Autoregressive (FAVAR) Approach," The Quarterly Journal of Economics, vol. 120(1), pages 387-422 with the codes provided by Tom Doan and posted on ideas: https://ideas.repec.org/c/boc/bocode/rtz00012.html.

Unfortunately, the code does not run smoothly with my RATS 8 version. Could you please provide me with some information on the version needed for this programming codes? Or whether there is a way to make the code run even in more recent versions of RATS.

Thank you.

[TomDoan]

I posted information about a revision of that at https://estima.com/forum/viewtopic.php?f=8&t=1029. Apparently, the Gibbs sampler needs a weak prior on the VAR to prevent the sampler from occasionally going unstable. (On looking at the comments in the original program, the authors started without a prior and added one later, so I assume they ran into this as well).

[hawkey]

Dear Tom,

Many thanks for the link to the updated version. It runs smoothly now. I was wonderring whether you could provide me with the information where the structural shocks from the BBE VAR estimation are stored. I would be particularly interested in the (identified) monetary policy shock time series.