Implement Gali (2015), chapter 7

Hi everyone

I am relatively new to Dynare and I tried to run the codes you provided on Github for chapter 7 in Gali(2015). Doing this I am confronted with some problems:

  • As you mention this in the preface, first I should set the taylor_rule= 1 (the other equal 0), then optimal_policy=1 (the other equal 0), and then simple_rule_unemployment=1(the other equal), but when I try the first one out it says that there are many variables not assigned in the steady_state_model-block and there are not enough input arguments. When I try the second option it says "Some of the parameters have no value (lambda_p, lambda_w) when using stoch_simul. " but actually they are included in the steady_state_model. The same happens when I try the third option.

So, what am I doing wrong, did I forget sth. ? I would highly appreciate if anyone could help me out of this. I’ve spent two days trying but I actually do not come ahead.

By the way: Many thanks to jpfeifer for providing and sharing open research. Without your research, it would be quite impossible for students to get their hands dirty with DSGE modelling.

Best regards

1 Like

Thanks for the kind words. Those are warnings, not error messages. As you pointed out, the parameters are set within the steady state file, so there is no problem. Dynare simply does not know that is the case, hence the warning. Similarly the warning about not setting variables in the steady_state_model-block. If something is not set, its value is 0 by default. Given that we work with a linear model, that is correct.

Thanks a lot for your answer, it absolutely makes sense (warnings are not errors, sry for the confusions).
But still one question is left: when I set the taylor_rule= 1 (the other equal 0), there comes indeed an error (and no file is saved i.g. IRFs_taylor_rule.mat, which is needed for the figures at the very end). I attach a picture how it looks like, do you have a suggestion where the mistake lies?

Again, thanks a lot for yor patience.

That was a bug on my behalf. I forgot that branch when we phased out global variables in 4.6. A new version is online.

Thanks a lot for your reply. I tried the new version out but it still gives me an error but now a diferent now (error attached).
Might there be something wrong with my Matlab or is sth. missing.
I am very sorry for questioning again! Much appreciation for your work and effort!

I missed something. Please try again.

Now, the results for the Taylor Rule are visible as IRFs_taylor_rule.mat. But (I am very sorry) figure 5 is not depicted (technology shock): It gives the following error:
Thanks a lot!

Hopefully, the issue is now fixed.

Thanks a lot Prof. Pfeifer !
Now, there was no error.

Much appreciation for your work!

Best regards,