Page 1 of 1
Dynare++ and Likelihood
Posted:
Sun Aug 02, 2015 11:15 pm
by mjs19
Hi,
Can Dynare++ calculate likelihoods for models solved above third order? If the answer is yes, I am assuming its a solution obtained via pruning?
Thanks.
Re: Dynare++ and Likelihood
Posted:
Mon Aug 03, 2015 5:38 am
by jpfeifer
There is no closed form solution for likelihoods at order>1. Thus, you need to use e.g. the particle filter. It is not implemented in Dynare++. As the particle filter is simulation-based, people typically also use pruning.
Re: Dynare++ and Likelihood
Posted:
Mon Aug 03, 2015 5:57 pm
by mjs19
The particle filter is implemented in Dynare up to second order?
Re: Dynare++ and Likelihood
Posted:
Tue Aug 04, 2015 5:28 am
by jpfeifer
Yes, at third order, you would need to do e.g. the simulated method of moments.