Very excited to share new work with Tao Zha! Identification has long been disciplined in structural VARs; in the big data era, the information set deserves the same. We provide it: a new Bayesian method, applied to household credit & monetary shocks under recursive & proxy identification.
Every structural vector autoregression result depends on identifying restrictions and variables. Economists discipline the first but often choose the second by hand. This paper ends this asymmetry, from @yuchengyang1993.bsky.social and Tao Zha www.nber.org/papers/w35604