Issues with prior distribution

Estimation ML/Bayesian estimation
1

Dear all,

I’m trying to estimate a DSGE model for my thesis, i use Rossana Merola’s work and try to estimate the model with germany’s data. Until past week the code worked but today the code send me this error:
Error using inverse_gamma_specification
Too many input arguments.

Error in set_prior (line 221)
[bayestopt_.p6(k(i)),bayestopt_.p7(k(i))] = inverse_gamma_specification(bayestopt_.p1(k(i)), bayestopt_.p2(k(i))^2, bayestopt_.p3(k(i)), 1, false, bayestopt_.name{k(i)});

Error in dynare_estimation_init (line 145)
[xparam1,estim_params_,bayestopt_,lb,ub,M_] = set_prior(estim_params_,M_,options_);

Error in dynare_estimation_1 (line 120)
dynare_estimation_init(var_list_, dname, , M_, options_, oo_, estim_params_, bayestopt_);

Error in dynare_estimation (line 105)
dynare_estimation_1(var_list,dname);

Error in SW_FA_Dynare.driver (line 1434)
oo_recursive_=dynare_estimation(var_list_);

Error in dynare (line 310)
evalin(‘base’,[fname ‘.driver’]);

I try run the code without the prior distribution with ‘invgamma’ and the code work, i also try to modify in various ways the shock’s prior but the code keep giving me the same error.

germanydata.xlsx (77.9 KB)
SW_FA_Dynare.log (975 Bytes)
SW_FA_Dynare.mod (9.8 KB)

Thanks!

2

now the error is: Unrecognized function or variable ‘estim_params_’.

Error in SW_FA_Dynare.driver (line 1258)
if isempty(estim_params_)

Error in dynare (line 310)
evalin(‘base’,[fname ‘.driver’]);

3

You are loading an m-file for that data-file that you did not provide.

4

sorry,
germanydata.m (563 Bytes)

5

Sorry, but I cannot replicate the issue.

6

I tried reinstalling matlab and dynare and it worked thank you very much

7

However, I get some strange IRFs (especially for the model without financial frictions)
germanydata.xlsx (77.9 KB)
germanyloading.m (545 Bytes)
SW_FA_Dynare.mod (9.2 KB)
SW_NOFA_Dynare.mod (7.7 KB)

8

What is strange about them? And did you check that the estimated parameters make sense?

9

because some don’t converge to 0, irf must converge to 0 or am I wrong? some seem to move further and further away from 0 over time
i use the prior of an another dsge model for euro area. Thank you very much
This is the IRFs of the model with and without financial frictions and in each folder there are the irfs both without proceeding with the estimate and with the estimate. I change something with the previous code.

IRFFA.zip (1.1 MB)
IRFNOFA.zip (934.9 KB)
germaniaFA.mod (9.3 KB)
germanydata.xlsx (77.9 KB)
germaniaNOFA.mod (7.8 KB)
germanyloading.m (543 Bytes)

10

Your model just features a lot of persistence. The standard IRF length of 20 periods is insufficient to see the mean reversion.