Indeterminacy problem 1

Stochastic simulations
1

Dear all,

I encounter this problem: “Blanchard Kahn conditions are not satisfied: indeterminacy”. I have checked the timing of the model and it looks correct. Appreciate if anyone can give me advice on this.

Felice.mod (1.3 KB)

Thanks,
Vannak

2

Check the timing of your equations. Are you sure that e.g. is that forward-looking?

r=sigmata*(yn(+1)-yn)+(sigma-sigmata)*(cf(+1)-cf);
3

Hello,
Perhaps it is a time specification issue as Johannes suggests. Without knowing your model, I see that sigma is equal to sigmata, which is 1. Some equations are reduced or even not significant, so as some parameters become zero. Maybe a check of your parameters sections might help.
Best,
FgB

4

Hi Johannes,

Thanks for your response. They are exactly the same as the paper I wish to replicate.

5

Hi Fernando,

Thanks for your suggestion. I tried to change the value of my theta to ensure sigma is not equal to sigmata. Still, the problem exists.

Regards

6

What is the paper you are trying to replicate?

7

it is Felices and Tuesta (2013) “Monetary Policy in Dual Currency Environment”.

8

I have checked the timing again. They are correct as per original paper.
When I use model_diagnostic, it mentioned about singularity problem. I am not sure if this is the root cause.

Appreciate if you can advise me how to go about this problem. 1. Monetary policy in a dual currency environment.pdf (449.1 KB)

9

Hello, maybe I am wrong, but you are using equations from a subsection and mix them with another one (section III). The issue that I obtain from the model_diagnostics is that you may have a redundant equation.
Perhaps you can start with only the equations expressed in the subsection “The log-linear version in a limiting case: (n → 0)”. Then you can move to the replicate the model in section III, which is loglinear representation in terms of output gap and domestic inflation.

10

Hi Fernando,

Thanks for your suggestion. I have tried your suggestion, but still the same problem arises. sub.mod (2.1 KB)

1 Like
11

I see that it is not running and you follow the equations. Maybe there is one variable that is predetermined. I can’t follow which one (I am a bit restricted in time for now).
If you find which one(s), include after the "varexo …; " section,

predetermined_variables …;

and write instead the … the variable(s) you find it is predetermined. Let me know if it helps.

12

Hi Fernando,

If i choose s as the predetermined_variables, the model runs. However, I am not sure if I choose the right variable to be predetermined variable.

13

I don’t think that the use of exchange rate as predetermined variable is correct. I am not sure whether the paper mentions something about predetermined variables, have a look. At the moment I cannot devote lots of time for checking the code, I am sorry. Perhaps you can ask for directions from the authors of the paper you try to replicate.

14

Apart from thoroughly rechecking your equations and the ones in the paper, I also can only recommend asking the authors.

15

Thanks for your response. exo.mod (1.2 KB). After checking all the equation and simplify the model, I now encounter another problem “Blanchard Kahn conditions are not satisfied:
indeterminacy”. Appreciate if anyone is able to give me advise me on how to go about it.