Working Paper: CEPR ID: DP6373
Authors: Giovanni Amisano; Oreste Tristani
Abstract: We estimate the approximate nonlinear solution of a small DSGE model on euro area data, using the conditional particle filter to compute the model likelihood. Our results are consistent with previous findings, based on simulated data, suggesting that this approach delivers sharper inference compared to the estimation of the linearised model. We also show that the nonlinear model can account for richer economic dynamics: the impulse responses to structural shocks vary depending on initial conditions selected within our estimation sample.
Keywords: Bayesian estimation; DSGE models; inflation persistence; second order approximations; sequential Monte Carlo
JEL Codes: C11; C15; E31; E32; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Positive inflation target shock (E31) | Output (Y10) |
| Positive inflation target shock (when inflation is low) (E31) | Output (Y10) |
| Positive inflation target shock (when inflation is high) (E31) | Output (Y10) |
| Initial conditions of inflation (E31) | Impulse responses to structural shocks (C22) |
| Nonlinear model performance (C52) | Predictive log density (C59) |
| Linear model performance (C52) | Marginal likelihood (C11) |