Problem using Identification command

[Azh]

Hello,

I was trying to implement identification analysis for Christiano, Motto, Rostagno paper, Risk shock. But there is a mistake using svd matlab function.
I consider only simplified version of the paper, without financial frictions. There are no measurement equations in this case. I tried to use different data set, but I still have the same error:

Testing prior mean
Evaluating simulated moment uncertainty ... please wait
Doing 300 replicas of length 300 periods.
Simulated moment uncertainty ... done!
??? Error using ==> svd
Input to SVD must not contain NaN or Inf.
Error in ==> rank at 15
s = svd(A);
Error in ==> identification_analysis at 200
if iflag || rank(MIM)>rank(tildaM),
Error in ==> dynare_identification at 280
[idehess_point, idemoments_point, idemodel_point, idelre_point,
derivatives_info_point, info] = ...
Error in ==> cmr at 677
dynare_identification(options_ident);
Error in ==> dynare at 180
evalin('base',fname) ;

What could possibly trigger this problem?
I would be very grateful if you could help me to solve this mystery.

[jpfeifer]

Could you please provide the files to replicate this issue?

[Azh]

Sure, Pls find the code in the attached file.

I use dynare\4.4.2.

File to run : cmr.mod

[jpfeifer]

Identification cannot deal with a singular Jacobian that is not due to the presence of a unit root.
Model_diagnostics points to the following issue:

Code: Select all

MODEL_DIAGNOSTICS:  The Jacobian of the static model is singular
MODEL_DIAGNOSTICS:  there is 1 colinear relationships between the variables and the equations
Colinear variables:
c             
Fp             
Fw             
g             
h             
i             
kbar           
lambdaz       
phi           
Colinear equations
    35

MODEL_DIAGNOSTICS:  The presence of a singularity problem typically indicates that there is one
MODEL_DIAGNOSTICS:  redundant equation entered in the model block, while another non-redundant equation
MODEL_DIAGNOSTICS:  is missing. The problem often derives from Walras Law.

[Azh]

Thank you so much for you quick answer.

Does it mean that the modal of Christiano, Motto Rostagno is false?

But then, why dynare finds the solution of the model without complaining for the redundant equation?

[jpfeifer]

Is this the original code or did you modify it?

Even in the presence of collinearity, a model sometimes runs. But identification can currently not deal with this.

The model itself is really big and I am not that familiar with it. Given the size of the model, I would not be surprised if there were issues with the code. But the authors are known to work extremely diligently and know what they are doing. Hence, I would rather believe that there is a good explanation for this. Maybe you could ask them why there is a collinearity issue.