Assessing monetary policy surprises in Japan by high frequency identification

Abstract

The use of changes in short-term interest rates (STIRs) within 30 min before and after monetary policy announcements, or so-called high-frequency identification (HFI), has been attracting attention as a method of extracting monetary policy surprises. In this paper, we use the Japanese data during the 2000s and 2010s, which includes periods when interest rates hovered around the ELB, to construct an indicator of monetary policy surprises using HFI and document its properties. We find that the STIR futures variations within 30 min around monetary policy announcements are more closely correlated with key financial variables than those outside that window. We also find that the impulse responses of macroeconomic variables to the identified shocks are overall in line with what conventional theory predicts.