Blanchard and Kahn condition failed: no stable equilibrium!

Hi everyone,

I am working on a code of my macro network model. I can find out the steady-state but the simulation does not work at all! I upload the code below and does anyone have an idea?

All the best,
Andre

Benchmark.mod (3.6 KB)

You did not provide the modelparm_37sec.mat to run the model.

Hi jpfeifer!

Thanks for replying and sorry for forgetting to upload the file. Here it is!
modelparm_37sec.mat (21.6 KB)

Thanks for any help!!

All the best,
Andre

It’s very hard to see what is going on in your model. Have you tried a simplified version with fewer states? I also fail to see the purpose of

    exp(Kd_3)=exp(Ks_3(-1));

That seems like causing a timing error.

Thanks for the respondence! I have tried exp(Kd_3)=exp(Ks_3) and the results do not change. I might need to try it again with fewer states. Thanks again!

And the reason for this exp(Kd_3)=exp(Ks_3(-1)); is that Ks is determined in last period and Kd is determined in this period in the model.

That cannot be correct. exp(Kd_3)=exp(Ks_3(-1)) implies that Kd_3 is known one period in advance.

Thanks for reminding me of this! I revised my model and attached it below here:
Model (263.9 KB)

And this is the updated code:
Only_K.mod (3.4 KB)
modelparm_37sec.mat (21.6 KB)

Any idea? Thanks for any help!

All the best,
Andre

If you add check;, you will see

There are 3 eigenvalue(s) larger than 1 in modulus
for 0 forward-looking variable(s)

This suggests a systematic timing error in one variable that affects all three sectors. Why is your capital FOC backward-looking and not forward-looking? For example

beta*(exp(mu)*(1-delta_3)+exp(lambda_3)*alpha_3*theta_3*exp(Y_3)*(1/exp(K_3(-1))))=exp(mu(-1));

I eventually figured it out! Thanks a lot for all the help! This forum is defintely 5 points out of 5!