Codes used to estimate a Dynamic Stochastic General Equilibrium (DSGE) model using Bayesian Estimation techniques.

These codes are available online at:
https://github.com/vcurdia/VC-BayesianEstimation

The codes were tested using Matlab (R) R2018a with the following toolboxes

• Symbolic Toolbox
• Statistical Toolbox
• Optimization Toolbox

LaTeX is used by some tools to compile certain documents.

epstopdf, included in most LaTeX releases, is used by some tools.

Codes from Vasco Cúrdia:

• VC-Tools, version v3.1.0

Codes from Chris Sims:

• gensys
• optimize
• KF

The script SetDSGE.m is an example of how to setup the model and estimate it using this package.

The main structure for setup

1. Set file names for the data input and the output
2. Set parameters list and priors
3. Set list of observation variables
4. Set list of State space variables
5. Set list of iid shocks
6. Generate symbolic variables
7. Construct any necessary auxiliary definitions (optional)
8. Set observation equations
9. Set state Equations

The above 9 steps will set up the model. After setup the Bayesian estimation proceeds by finding the mode of the posterior using MaxPost.m and then generating MCMC samples, using MCMC.m. Analysis of estimation results is done with MCMCAnalysis.m.

See the example SetDSGE.m for basic options and how to call the sequence of steps in more detail.

Reports in pdf format are generated along the way.

Each of the functions and scripts contains help at the beginning of the codes, including options and flags. This help can be accessed in an interactive Matlab (R) session using the help or doc commands:

help SetDSGE

or

doc SetDSGE