Working Paper: CEPR ID: DP8432
Authors: Hyungsik Roger Moon; Frank Schorfheide; Eleonora Granziera; Mihye Lee
Abstract: There is a fast growing literature that partially identifies structural vector autoregressions (SVARs) by imposing sign restrictions on the responses of a subset of the endogenous variables to a particular structural shock (sign-restricted SVARs). To date, the methods that have been used are only justified from a Bayesian perspective. This paper develops methods of constructing error bands for impulse response functions of sign-restricted SVARs that are valid from a frequentist perspective. We also provide a comparison of frequentist and Bayesian error bands in the context of an empirical application--the former can be twice as wide as the latter.
Keywords: Bayesian inference; frequentist inference; partially identified models; sign restrictions; structural VARs
JEL Codes: C1; C32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| positive demand shock (E00) | prices (P22) |
| positive demand shock (E00) | output (C67) |
| sign restrictions (C29) | bounds on impulse responses (C22) |
| identified set of impulse responses (C22) | not a singleton (C39) |
| frequentist error bands (C46) | valid insights into monetary policy shocks (E49) |
| frequentist confidence intervals (C46) | larger than Bayesian credible sets (C11) |