A major challenge for proxy vector autoregressive analysis is the construction of a suitable instrument variable for identifying a shock of interest. We propose a simple proxy that can be constructed whenever the dating and sign of particular shocks are known. It is shown that the proxy can lead...

Studies of the crude oil market based on structural vector autoregressive (VAR) models typically assume a time-invariant model and transmission of shocks or they consider a time-varying model and shock transmission. We assume a heteroskedastic reduced-form VAR model with time-invariant slope...