Page 1 of 1

DSGE model for other country

PostPosted: Mon Jul 14, 2014 12:58 pm
by cobzac
Hello, everybody!

I'm interested in the way of applying a given DSGE model to a different country. What I mean is: suppose I have a DSGE model that describes US economy, say US_SW07 (.mod file, from MMB 2.0) and I want to apply it to Estonia. Which lines and parameters I should change in original model? Or may be some equations should be changed also?

I don't have an economic education and also I'm new to Dynare

Re: DSGE model for other country

PostPosted: Mon Jul 14, 2014 2:06 pm
by jpfeifer
You need to adjust all parameters and priors. It should usually not be necessary to change any model equations unless you want to modify the model structure itself. But given your example, Estonia, using a small open economy version of the Smets/Wouters model might be more appropriate. Given your statement that you don't have an economics education, I would strongly discourage you from attempting this. It will be way over your head, because this is something you might want to attempt in the second year of your economics PhD.

Re: DSGE model for other country

PostPosted: Mon Jul 14, 2014 2:27 pm
by cobzac
jpfeifer wrote:You need to adjust all parameters and priors. It should usually not be necessary to change any model equations unless you want to modify the model structure itself. But given your example, Estonia, using a small open economy version of the Smets/Wouters model might be more appropriate. Given your statement that you don't have an economics education, I would strongly discourage you from attempting this. It will be way over your head, because this is something you might want to attempt in the second year of your economics PhD.

Thanks for reply. How exactly I should adjust the parameters? What are priors, initial conditions? Am I supposed to use real data and compare dynare simulations with it? Or I can get the parameters from estonaian central bank, for example? And mainly what are these parameters? Something like discount factor, capitals share, depreciation rate, persistence of technology shocks?
And which model would you recommend for small economy?
To make it clearer I have a problem to estimate economical state of the country, so I'm not so bothered with how the model works, I just need it to work correctly and show quality results.