Sensitivity analysis of a DSGE model with time-varying parameters identified by a nonlinear particle filter
| Authors | |
|---|---|
| Year of publication | 2015 |
| Type | Article in Proceedings |
| Conference | 33rd International Conference Mathematical Methods in Economics Conference Proceedings |
| MU Faculty or unit | |
| Citation | |
| Web | conference proceedings |
| Field | Economy |
| Keywords | DSGE model; time-varying parameters; particle filter; sensitivity analysis |
| Attached files | |
| Description | In this paper, we assess the robustness of the time-varying parameter estimates of a dynamic stochastic model of a small open economy general equilibrium with financial frictions that were obtained with the use of a nonlinear particle filter. First, we assess the sensitivity to the selection of the subset of model parameters that are assumed to be time-varying. For that purpose, the model is estimated in three different configurations of the time-varying and constant parameters. Second, we examine the sensitivity of the time-varying estimates with respect to the calibration of the initial value of the adhesion parameter. The value of a time-varying parameter is given by the model law of motion as a weighted average of the last known value of the parameter and its initial value plus stochastic innovation. The weights are determined by the adhesion parameter that is common for all the time-varying parameters. Thus, the adhesion parameter determines the general tendency of the time-varying parameters to return to their respective initial values. Three different calibrations of the adhesion parameter are compared in the paper. The model is estimated using the data of the Czech economy. |
| Related projects: |