Publication details
What drives the estimation results of DSGE models? Effect of the input data on parameter estimates
| Authors | |
|---|---|
| Year of publication | 2018 |
| Type | Article in Proceedings |
| Conference | Proceedings of 36th International Conference Mathematical Methods in Economics |
| MU Faculty or unit | |
| Citation | |
| web | Conference proceedings |
| Keywords | Bayesian estimation; DSGE model; Matching moments; Model simulation; Parameter identification |
| Description | In this contribution, I compare three different Bayesian dynamic stochastic general equilibrium (DSGE) models in a simulation-estimation exercise. This exercise is aimed at revealing the capabilities of these models to re-estimate, during the estimation phase, values of parameters previously set in the simulation phase. The first model is the renowned work of Smets and Wouters. The second one is the rather small DSGE model of a closed economy with search and matching frictions on labour market proposed by Lubik. The third one is based on the paper written by Sheen and Wang, where they introduce a model of a small open economy with various labour market frictions. The aim of this contribution is to examine how the complexity of the model and the amount of information needed, represented by the number of observations in the observables, affect the results when the parameters are estimated. At first, I shortly introduce all presented models. Based on the given calibration, trajectories of main endogenous variables are simulated. These simulated trajectories with a various number of observations are then used as observables for estimation of the model parameters to reveal how rich information is needed for each model to properly identify its parameters. |
| Related projects: |