How to measure inFLAtion volatility. A note

Abstract

This paper proposes a statistical model and a conceptual framework to estimate inflation volatility assuming rational inattention, where the decay in the level of attention reflects the arrival of news in the market. We estimate trend inflation and the conditional inflation volatility for Germany, Spain, the euro area and the United States using monthly data from January 2002 to March 2022 and test whether inflation was equal to or below 2% in this period in these regions. We decompose inflation volatility into positive and negative surprise components and characterise different inflation volatility scenarios during the Great Financial Crisis, the Sovereign Debt Crisis, and the post-COVID period. Our volatility measure outperforms the GARCH(1,1) model and the rolling standard deviation in one-step ahead volatility forecasts both in-sample and out-of-sample. The methodology proposed in this article is appropriate for estimating the conditional volatility of macro-financial variables. We recommend the inclusion of this measure in inflation dynamics monitoring and forecasting exercises.