I cannot know why fulfill Blanchard & Kahn conditions are not satisfied: no stable equilibrium

jpmodel5.mod (9.5 KB)
jpdat.mod (19.9 KB)
equation.pdf (243.1 KB)
Kotera_Sakai(2018)PPR.pdf (2.9 MB)
Hi all.
I’m confused by this error which equation doesn’t have a unique solution. This is because the formula has once been estimated by another paper through a regular Kalman filter and should meet Blanchard & Kahn conditions.
I’m currently looking for the cause, but it’s not clear. If anyone knows, I would be grateful if you could point out what went wrong.

I attached equation of dynare at equation.pdf. If some wrong in the equation, please tell me.
My research is focusing Forecasting of Tax in Japan economy under DSGE-VAR model. So that, Prior research is mainly using Kotera_Sakai(2018)PPR.pdf to fulfill my paper in undergraduate school. Therefore, any mistakes is on Dynare model program but no mistake in my model in paper because that model is discussed on many papers in Japan.


I know where is the mistake because I checked into author’s laboratory and return following answers.

“The mistake is in parameter’s initval values. when I chenged the values of esitimated_param’s initval values, the model is worked.”

And recommended not using Octave but Matlab is best for solving that problem.Then we recommend to you using gensys program".

Therefore, I am starting debugging of the model by using gensys in matlab by using 30 day’s free program.

Thank you(:slight_smile:)


I don’t see why using Matlab with Gensys would help. After all, Dynare employs gensys in the background as well. It’s a C++ routine, so Octave or Matlab does not make a difference.

I am also I do not know how to provide values on gensys_check. But it needs to provide steady_state_model and making matrix for fiting Blanchard_Kahn Theory’s calculation…

“Octave and Matlab does not make difference” is noticed by using Matlab…(:

For others meeting Blanchard-Kahn condition problem,
the error of mine is concerned as parameters of reaction of interst rates to inflation(phi_r_pi) must be <1. therefore need to be nomal distribution and <1.
This is strange but system imply this…()

See Why Inflationary reaction to interest rates is under 1 in Taylor rule?