Inflation volatility: an Asian perspective

Abstract

For the quarterly data of 10 Asian economies, ranging from the first quarter of 1991 to last quarter of 2012, we model inflation volatility as a time varying process through different symmetric and asymmetric GARCH specifications. We also propose to model inflation volatility on the basis of cyclic component of inflation obtained from an Hodrick Prescott (HP) filter instead of actual inflation when the latter does not fulfil the criterion of stationarity. Through news impact curves (NICs) we tried to highlight the behaviour of inflation volatility in response to lagged inflation shocks under different GARCH specifications. In our results the leverage parameter shows the expected sign and is significant for almost all countries suggesting strong asymmetry in inflation volatility. The hyperbolic sign integral shape of NICs based on Glosten-Jagannathan-Runkle GARCH (GJR-GARCH) highlights the importance of inflation stabilisation programmes particularly because of the subsequent evidence obtained in favour of bidirectional causality running between inflation and inflation volatility. There is also evidence in favour of the argument that a cyclic component of inflation obtained through an HP filter could be used as a suitable proxy of inflation for volatility estimation.