This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2022
  4. Preliminary Program
  5. JBES Session: Structural Macro – Narrative Restrictions and Proxies
« Back to Results

JBES Session: Structural Macro – Narrative Restrictions and Proxies

Paper Session

Friday, Jan. 7, 2022 12:15 PM - 2:15 PM (EST)

Hosted By: Econometric Society
  • Chair: Christian Hansen, University of Chicago

Narrative Restrictions and Proxies

Raffaella Giacomini
,
University College London
Toru Kitagawa
,
University College London
Matthew Read
,
University College London

Abstract

We compare two approaches to inference in structural vector autoregressions given information about the signs of structural shocks at specific dates: a likelihood-based approach that imposes `narrative restrictions' (NR) on the shock signs and a `narrative-proxy' (NP) approach that casts the information about the shock signs as proxies for the shocks. When the number of NR is fixed, the robust Bayesian approach to inference under NR described in Giacomini, Kitagawa and Read (2021) delivers robust credible intervals with valid asymptotic frequentist coverage of the true impulse responses. In contrast, under the NP approach, the assumptions for validity of the weak-proxy robust confidence intervals in Montiel Olea, Stock and Watson (2021) are violated. A Monte Carlo exercise suggests that the weak-proxy robust confidence intervals can have incorrect coverage unless the sign of the shock is known in a large number of periods, whereas the robust credible intervals under the NR approach always display coverage exceeding the nominal level.

Discussant(s)
Lutz Kilian
,
Federal Reserve Bank of Dallas
Mikkel Plagborg-Møller
,
Princeton University
Juan Rubio-Ramírez
,
Emory University
JEL Classifications
  • C1 - Econometric and Statistical Methods and Methodology: General
  • C3 - Multiple or Simultaneous Equation Models; Multiple Variables