Loop over parameters

[htastan]

Suppose that we want to estimate the parameters using GMM. In particular, we would like to match the theoretical moments implied by the linearized model to the sample moments based on the data. Also assume that we have a separate toolbox for GMM estimation in which we just need to code the moment conditions in a separate m file. In this case, we need to pass the information obtained from first run of the Dynare to parse the code to the moment function in GMM and the objective function. Can we do this as follows:

1. In a Matlab script file define the parameters and all other necessary stuff along with

global var_list_ oo_ M_ ex0_ options_ estimation_info info
save parameters gamma1 alpha1 delta1 rho1 beta1 sigma1
dynare stochgrowth_dyna noclearall

2. In the mod file define

parameters gamma alpha delta rho beta sigma;

load parameters;

set_param_value('gamma',gamma1) % utility parameter
set_param_value('alpha',alpha1) % production elasticity
set_param_value('delta',delta1) % depreciation rate
set_param_value('rho',rho1) % production shock autocorrelation
set_param_value('beta',beta1 ) % discount factor
set_param_value('sigma',sigma1) % production shock volatility

at the beginning, the rest of the mode file is standard.

3. In the GMM moment function matlab m file:

function mom = GMMmoments(theta,X,avg)

global var_list_ oo_ M_ ex0_ options_ estimation_info info

gamma1 = theta(1);
alpha1 = theta(2);
delta1 = theta(3);
rho1 = theta(4);
beta1 = theta(5);
sigma1 = theta(6);

yobs = X(:,1);
cobs = X(:,2);
kobs = X(:,3);
iobs = X(:,4);

set_param_value('gamma',gamma1)
set_param_value('alpha',alpha1)
set_param_value('delta',delta1)
set_param_value('rho',rho1)
set_param_value('beta',beta1 )
set_param_value('sigma',sigma1)

info = stoch_simul(var_list_);
if info
disp('Stochastic Simulation failed');
end

xmean = oo_.mean;
xvar = diag(oo_.var);

Basically, what I did here is to define Dynare solution variables as GLOBAL variables in the main file and the moment file. I am not a big fan of using global variables but I could not use local variables. Is this procedure OK? Is there any other ways?

(By the way I did not post the full codes but they seem to work OK)

[jpfeifer]

Looks correct to me. There are ways to get around using global variables in the main file, but they are typically not worth the hassle. The only global variables you should need are

Code: Select all

oo_ M_  options_

One important comment: if you are using simulated moments, make sure you use the same random numbers for every parameter draw.

[htastan]

Exactly. The same set of shocks has to be used in the SMM or indirect inference. One can easily code this using Dynare with second or third order linearization.
Better way is of course to solve the nonlinear eq'm conditions using residual based methods, e.g. collocation, and then to embed it into the indirect inference framework. That would be a very nice extension to the Dynare.

By the way I think we also need

var_list_

as a global variable in addition to the ones you mentioned.
Thanks for your reply,

Cheers

[jpfeifer]

If you did not specify variables for var_list_, you can just use

Code: Select all

var_list_=[];
info = stoch_simul(var_list_);


If you did, you can either make var_list_ global or use something along the lines of

Code: Select all

var_list_=[];
var_list_ = 'yobs ';
var_list_ = char(var_list_, 'cobs');
var_list_ = char(var_list_, 'kobs');
var_list_ = char(var_list_, 'iobs');
info = stoch_simul(var_list_);


[bpyemba]

Hi all,

I am trying to evaluate the conditional welfare and I am getting the following error

??? Reference to non-existent field 'eigval'.

Error in ==> check at 80
eigenvalues_ = dr.eigval;

Error in ==> openeconomy_loop_det at 626
oo_.dr.eigval = check(M_,options_,oo_);

Error in ==> dynare at 120
evalin('base',fname) ;

Error in ==> max_cond_welfare2 at 31
dynare openeconomy_loop_det noclearall nolog

How can I solve the problem?

Thanks Boniface

[jpfeifer]

Your openeconomy_loop_det.mod is unable to compute the steady state.

[bpyemba]

The steady state is reached using dynare 4.3.3. It doesn't work using the version 4.4.1.

[bpyemba]

Hi,

Could do you please run with 4.3.3?

Boniface

[jpfeifer]

No, also with Dynare 4.3.3 it does not find the steady state. This is the fundamental source of your problem.

[bpyemba]

Thanks, I don't understand how it works perfectly with 4.3.3 in my computer, but not with 4.4.1. I am working on.

[jpfeifer]

You could try

Code: Select all

steady(solve_algo=4);

in Dynare 4.4.2

[dphm]

Hi everyone

If I may follow up from this post, I am actually doing the same as kipfilet. But I gotta get the impulse response functions from the newly resolved steady state.

Previously, when I executed the "dynare" command, I simply could use oo_.irfs in my m file. But now, I don't know where the IRFs are stored after I run the command "[dr_,~,M_,~,oo_] = resol(0,M1_,options1_,oo1_);".

May I know how I can plot the impulse response functions thereafter?

Thanks.

[jpfeifer]

in that case you actually need to loop over

Code: Select all

info = stoch_simul(var_list_);


instead of resol. The irfs will then be stored in their usual place.

[HouseC]

Dears，
I'm following discussion here and what I'm also doing with grid search is to generate expected welfare for different monetary rules setting.

Here is my example code for loop.

Code: Select all

aaa=[
    0.1 0.3
    0.1 0.3
    0.1 0.3
    0.1 0.3];
first_time = 1;
welfare=zeros(1,2);
for i1=1:2
    e=aaa(1,i1);
    f=aaa(2,i1);
    g=aaa(3,i1);
    h=aaa(4,i1);

    if first_time
save parameterfile1 e f g h;
dynare LcpTax.mod noclearall;
       first_time = 0;
    else
set_param_value('sigma1',e);
set_param_value('sigma2',f);
set_param_value('sigmas1',g);
set_param_value('sigmas2',h);

nv=101;
eu_v=zeros(1,nv);
eus_v=zeros(1,nv);
eu_w=zeros(1,nv);

for jj=1:nv;
v=(jj-1)/(nv-1);
info=stoch_simul(var_list_);
eu_v(jj)=oo_.mean(26);
cbar=oo_.steady_state(1);
lbar=oo_.steady_state(3);
eu_v(jj)=(((eu_v(jj)*(1-beta)+eta*lbar)*(1-rho))^(1/(1-rho))/cbar-1)*100;

eus_v(jj)=oo_.mean(27);
csbar=oo_.steady_state(13);
lsbar=oo_.steady_state(15);
eus_v(jj)=(((eus_v(jj)*(1-beta)+eta*lsbar)*(1-rho))^(1/(1-rho))/csbar-1)*100;
eu_w(jj)=n*eu_v(jj)+(1-n)*eus_v(jj);

end;


My questions is:
based on the discussion, we only need to call dynare once and then change different parameter and redo the simulation. I double check the result in this example code, and find that the estimated welfare are different between calling dynare each time and the method mentioned in the previous code.
I believe that calling dynare for each loop will return the correct result, then anyone could suggest what's the problem with my code listed above?

[monsoon]

Just clarifying a doubt with reference to the above topic. Is looping important when we deal with endogenous collateral constraints?