DSGE Model Restrictions for Structural VAR Identification
by Philip Liua and Konstantinos Theodoridisb
The identification of reduced-form VAR models has been
the subject of numerous debates in the literature. Different
sets of identifying assumptions can lead to very different conclusions
regarding the effects of shocks. This paper proposes
a theoretically consistent identification strategy using restrictions
implied by a DSGE model. Monte Carlo simulations suggest
that both quantitative and qualitative restrictions work
well together, where they act as complements to each other,
in minimizing errors in finding the correct VAR identification.
When using misspecified model restrictions, the data tend to
push the identified VAR responses away from the misspecified
model and closer to the true data-generating process.
JEL Codes: F31, E52.
(PDF, 35 pages 534 kb)
a International Monetary Fund
b Bank of England