Estimating Structural Shocks with the GVAR-DSGE Model: Pre- and Post-Pandemic

Kwok, C. 2022. Estimating Structural Shocks with the GVAR-DSGE Model: Pre- and Post-Pandemic. Mathematics. 10 (10) e1773.

TitleEstimating Structural Shocks with the GVAR-DSGE Model: Pre- and Post-Pandemic
TypeJournal article
AuthorsKwok, C.
AbstractThis paper investigates the possibility of using the global VAR (GVAR) model to estimate a simple New Keynesian DSGE-type multi-country model. The long-run forecasts from an estimated GVAR model were used to calculate the steady-states of macro variables as differences. The deviations from the long-run forecasts were taken as the deviation from the steady-states and were used to estimate a simple NK open economy model with an IS curve, Philips curve, Taylor rule, and an exchange rate equation. The shocks to these equations were taken as the demand shock, supply shock, monetary shock, and exchange rate shock, respectively. An alternative model was constructed to compare the results from GVAR long-run forecasts. The alternative model used a Hodrick−Prescott (HP) filter to derive deviations from the steady-states. The impulsive response functions from the shocks were then compared to results from other DSGE models in the literature. Both GVAR and HP estimates produced dissimilar results, although the GVAR managed to capture more from the data, given the explicit co-integration relationships. For the IRFs, both GVAR and HP estimated DSGE models appeared to be as expected before the pandemic; however, if we include the pandemic data, i.e., 2020, the IRFs are very different, due to the nature of the policy actions. In general, DSGE−GVAR models appear to be much more versatile, and are able to capture dynamics that HP filters are not.
Keywordsglobal VAR
New Keynesian model
structural model
pandemic data
Article numbere1773
Journal citation10 (10)
Publisher's version
CC BY 4.0
File Access Level
Open (open metadata and files)
Digital Object Identifier (DOI)
Publication dates
Published23 May 2022
LicenseCC BY 4.0

Permalink -

Share this

Usage statistics

10 total views
1 total downloads
These values cover views and downloads from WestminsterResearch and are for the period from September 2nd 2018, when this repository was created.