The efficient market hypothesis and identification in structural VARs

Structural vector autoregression (SVAR) models are commonly used to investigate the effect of structural shocks on economic variables. The identifying restrictions imposed in many of these exercises have been criticized in the literature. This paper extends this literature by showing that, if the SVAR includes one or more variables that are efficient in the strong form of the efficient market hypothesis, the identifying restrictions frequently imposed in SVARs cannot be satisfied. The authors argue that this analysis will likely apply to VARs that include variables that are consistent with weaker forms of the efficient market hypothesis, especially when the data are measured at the monthly or quarterly frequencies, as is frequently the case.