Page 1 of 1

24 structural shocks and 24 observed variables, ridiculous?

PostPosted: Wed Apr 19, 2017 4:57 am
by Jesse
Dear Johannes,
I recently formulate a DSGE model with around 100 parameters, and I use 24 structural shocks and 24 observed variables, I am able to run dynare program without problems, however, when i presented my paper, some macroeconomists laughed at my idea of including 24 structural shocks. My problem is that I have a large scale macroeconometric DSGE model and I also use the dsge model to formulate priors for var model, I need a large data set-24 observed variables, to validate my DSGE-VAR model, I need to have 24 structural shocks. Do you think my inclusion of 24 structural shocks is ridiculous?
Thank you and look forward to hearing from you.
Best regards,
Jesse

Re: 24 structural shocks and 24 observed variables, ridiculo

PostPosted: Wed Apr 19, 2017 3:55 pm
by jpfeifer
If it's a one-country model, you will be hard-pressed to come up with 24 structural shocks. However, in open economy papers, it is not unheard of. The paper "What Drives the German Current Account? And How Does it Affect Other EU Member States?" by Robert Kollmann, Marco Ratto, Werner Roeger, Jan in’t Veld and Lukas Vogel, Economic Policy, 2015 at http://www.robertkollmann.com/KOLLMANN_PUBL_RATTO_ROEGER_intVELD_VOGEL_ECONOMIC_POLICY_2015.pdf has 46 shocks.