We introduce a general procedure for macroeconomic models’ calibration and validation. Configurations of parameters are selected on the basis of a loss function involving a distance between model-derived structural coefficients and their empirical counterparts. These, in both cases, are locally identified by exploiting non-Gaussianity in a structural vector autoregressive framework under a data-driven approach. We use model confidence set to account for the uncertainty in the selection procedure. We provide a measure of validation by comparing (model’s and empirical) shocks-variables structure. We apply our procedure to a complex macroeconomic simulation model that studies the link between climate change and economic growth.
Calibration and Validation of Macroeconomic Simulation Models by Statistical Causal Search
Mario Martinoli;Alessio Moneta
;Gianluca Pallante
2024-01-01
Abstract
We introduce a general procedure for macroeconomic models’ calibration and validation. Configurations of parameters are selected on the basis of a loss function involving a distance between model-derived structural coefficients and their empirical counterparts. These, in both cases, are locally identified by exploiting non-Gaussianity in a structural vector autoregressive framework under a data-driven approach. We use model confidence set to account for the uncertainty in the selection procedure. We provide a measure of validation by comparing (model’s and empirical) shocks-variables structure. We apply our procedure to a complex macroeconomic simulation model that studies the link between climate change and economic growth.File | Dimensione | Formato | |
---|---|---|---|
Calibration_large_scale_ABM (4).pdf
accesso aperto
Tipologia:
Documento in Pre-print/Submitted manuscript
Licenza:
Dominio pubblico
Dimensione
2.52 MB
Formato
Adobe PDF
|
2.52 MB | Adobe PDF | Visualizza/Apri |
Martinoli_et_al_JEBO.pdf
accesso aperto
Tipologia:
PDF Editoriale
Licenza:
Copyright dell'editore
Dimensione
1.45 MB
Formato
Adobe PDF
|
1.45 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.