Abstract:
Central banks have usually employed short-term rates as the main instrument of monetary policy. In the last decades, however, forward guidance has also become a central tool for monetary policy. In an innovative way this paper combines two sources of extraneous information - high frequency surprises and narrative evidence - with sign restrictions in a structural vector autoregressive (VAR) model to fully disentangle the effects of forward guidance shocks from the effects of conventional monetary policy shocks. Results show that conventional monetary policy has the expected effects even in a recent US sample, in contrast with the evidence reported by Barakchian and Crowe (2013) and Ramey (2016), and that forward guidance is an effective policy tool. In fact, it is at least as strong as conventional monetary policy