CMStatistics 2022: Start Registration
View Submission - CFE
Title: Dealing with the statistical representation of DSGE models Authors:  Alessia Paccagnini - University College Dublin (Ireland) [presenting]
Abstract: Novel empirical evidence about model validation in DSGE modeling is provided. First, we use several small - and medium-scale models as data-generation processes to create artificial pseudo-data. Second, using this pseudo-data, we identify shocks by using VAR and Local Projections to evaluate the effects of macroeconomic shocks such as monetary policy and fiscal policy shocks. As main findings, we document how both VAR and Local Projection help researchers recover the dynamics of a DSGE, even if they cannot always recover the dynamics of the true data generation processes. These results provide a guideline to investigate the misspecification in empirical DSGE modeling due to the statistical representation.