Smets and wouters

Steady state
1

Greetings … I am in the process of building the Smets and Wouters 2007 model, and I have reached the practical side on Dynare 4.6.4 , and I faced two problems that I am trying to find solutions to:
1- Warning in the ‘steady_state _model’ block, variable …
2- Not knowing the extraction of prioer and posterior data in the output
data.xlsx (16.3 KB)
iraq.mod (20.8 KB)
many thanks

2
  1. The warnings are normal. Most variables are 0 in steady state, so there is no point in setting them.
  2. What exactly is your problem?
1 Like
3

Thank you very much, Professor. My main problem is that the backward estimates do not appear in the final results. I want to compare the results of the parameter estimated using the MH method and the method of likehood
Many thanks professor

4

What are

?

1 Like
5

Sorry, typing error, I mean piror estimate parameter

6

And what are

?

7

Thanks professor @jpfeifer : the Esimate parameters by bayesian approach, the priors of parameters are determined and presented, for exampel the inter temporal elasticity of substitution … coefficients for inflation, such as table
Parameters Pdf type Prior distribution mean 1st, prior distribution st. Error and prior distribution mean 2nd, for to all parameters coefficient of relative risk aversion , output elasticity of capital
Ar for lagged interst rate
Inflation cofficient ……. I want comper between calibrated parameters and estimated, exactly estimated parameters by baysian Aproach and Maximum Likelihood

8

Your data treatment seems inconsistent with the model variables. The means of your observables do not match. Please read Pfeifer(2013): “A Guide to Specifying Observation Equations for the Estimation of DSGE Models”

1 Like
9

Thank you very much Professor @jpfeifer , the paper has been modified and is working correctly, I have a question about how to add induced impulse response and Univariate Dignostic tes or Multivariate Diagnostic by use MCMC to the file mod in the model smets and woutres 2007,
Best

10

The IRFs are triggered by the bayesian_irfs-option (see the manual). The convergence diagnostics are provided by default if mh_replic>2000.