DSGE Dynare Code

General DSGE Modeling
1

Dear community,
I am looking for the DSGE dynare code of Smets-Wouters (2007).
Moreover, does the use of the R code give the same results as the Dynare code?
Sincerely

2

Hi E.Ahmed,

there is plenty of SW code in the forum if you look for it. Also Johannes Pfeifer has amazing Dynare code on many models on his Github, including SW.
I do not understand your second question though. What do you mean by R code? In general it should be the same but it, of course, depends on what your goal is and what you are trying to do.

Best

1 Like
3

Hi DoubleBass,

Thank you about Dynare code.
The second question, there is a package “gEcon” for DSGE estimates on r software. But, because a lot of people use just “Dynare” and not “R”, I would like know if there is a difference in the software output.
Best Regards

4

Hi E.Ahmed,

I am sorry but I am not familiar with this software-package. Why don’t you just try it with a super simple three equation model and find out for yourself.

Best

1 Like
5

There is no trouble, now I found a Dynare code that is important for now. I thank you for your time and I thank also Johannes Pfeifer for sharing the code.

Thank you :blush:.

1 Like
6

I have used gecon a couple of times, results are similar to dynare…just slight differences in values, sometimes. Maybe, correlation is 1.233565 for one and 1.234589 for the other. Of course, that should be expected due to differences in numerical approximation techniques.

2 Likes
7

Thank you Dear @kofiemma, Your comment is very important to me. Thanks

8

Hi @kofiemma
I am interested in this gEcon package. Do you have any tip or reference to share with a beginner in this tool?
Thank you

9

That maybe useful for you

1 Like
10

Yeah. It is pretty straightforward when you read the documentation. In the package, you just have to enter the model as it appears on say an exam. Thus, just enter the objective functions and constraints of the household, firm, whatever agent. And then it will do everything, basically. Thus, calculate FOC, compute steady-state, and the other results like IRFs, etc. They have a lot of models to practice with from simple RBC to Smets and Wouters. Heterogenous agents are also available. I practiced with it for a while, but Dynare is better. Maybe because I already know how to find my way in Dynare when the code does not compute. gEcon has a forum too, but actually, nothing there. And it takes a long time to get help when the code does not work. Many times, it can not find the steady-state unless the model is really simple or you slightly adapt one of the models they provide (or maybe I didn’t become an expert before I gave up. lol). But at least, it tries to calculate FOC and then get stuck with computing steady-state. So if you are bored with calculating FOCs with pen and paper, maybe you can use it, and then compute steady-state yourself.

Dynare is still way better because it has a lot of functionality and the forum is great - a lot of solved problems when you get stuck. Anyway, gEcon is fun to practice with, so yeah you can try.