Using Forecast-Augmented VAR Evidence to Dampen the Forward Guidance Puzzle

We estimate the effects of interest rate forward guidance (FG) using a parsimonious VAR, augmented with survey forecast data. The identification strategy of FG shocks via sign and zero restrictions is successfully tested by the recovery of true IRFs from simulated data. The identified shocks from the VAR suggest that FG has a stronger effect on macro variables and deviations are more instantaneous compared to the hump-shaped response following unanticipated changes in monetary policy. We apply this evidence to calibrate free parameters of an otherwise estimated DSGE model in order to dampen the FG Puzzle