QZ decomposition error in Dynare

Stochastic simulations
1

Hi.
I run a New Keynesian DSGE model with 23 equations and 23 endogenous variables.

When I run the model I have QZ decomposotion error in solving the model. I used model_diagonistic command to recognize the problem. This command indicates that:

There are not enough input arguments.

But I have 23 endogenous variables and also 23 equations in the model.

What can be the cause of this problem ?
NEWNKDSGE.mod (4.1 KB)

2

Without the file it is impossible to tell.

3

NEWNKDSGE.mod (4.1 KB)

I undestood the problem.by this command:

model_diagnostics(M_ , options_ , oo_)

Variable RN is not in present time in the model.

4

Hi, Professor.
my model has a problem. Variables of T and B
are collinar and model by QZ decomposition is not solveable in Dynare.

T is tax and B is bonds.

T is in the government budget equation.

How we can solve collinearity problem between two variables in a DSGE model ?
NEWNKDSGE.mod (4.2 KB)

What is the main reason of collinearity in DSGE models ?

Can we solve a New Keynesian model with not equal of number of equations and endogenous variables ?

for example number of equations higher than the number of endogenous variables or vice versa.

5

Does Ricardian equivalence hold in your model? If yes,

would apply.

6

Hi.
I changed the model structure. Many problems solved completely but impulse response functions have explosive form and are not normal. When I use model_diagnostics(M_,options_,oo_) command this command indicates that there is no obvious problem in the model and resids are zero for all equations.

I attach you my DSGE file for better investigation.
ANEWMODEL.mod (4.6 KB)

7

See

8

No, I did not test different values for parameters of the model.