Inflation Dynamics and Forecast: Frequency Matters

50 Pages Posted: 9 Jun 2021 Last revised: 6 Mar 2023

See all articles by Manuel M. F. Martins

Manuel M. F. Martins

University of Porto, cef.up, Faculdade de Economia

Fabio Verona

Bank of Finland - Research

Date Written: June 8, 2021

Abstract

We use a New Keynesian Phillips Curve (NKPC) to study in-sample inflation dynamics and to forecast inflation out-of-sample in the frequency domain. In-sample, inflation expectations dominate medium-to-long-run cycles, energy inflation dominate short cycles and also longer cycles once expectations became anchored. While statistically significant, unemployment is never economically relevant. Out-of-sample, forecasts from a low-frequency NKPC significantly outperform several benchmark models. The low-frequency component of unemployment is key for such remarkable forecasting performance. Hence, while unemployment is of little relevance in-sample, it remains crucial in predicting inflation out-of-sample due to its role at cycles longer than typical business cycles.

JEL Classification: C53, E31, E37

Suggested Citation

Mota Freitas Martins, Manuel and Verona, Fabio, Inflation Dynamics and Forecast: Frequency Matters (June 8, 2021). Bank of Finland Research Discussion Paper No. 8/2021, Available at SSRN: https://ssrn.com/abstract=3863066 or http://dx.doi.org/10.2139/ssrn.3863066

Manuel Mota Freitas Martins (Contact Author)

University of Porto, cef.up, Faculdade de Economia ( email )

4200-464 Porto
Portugal

Fabio Verona

Bank of Finland - Research ( email )

P.O. Box 160
FIN-00101 Helsinki
Finland

HOME PAGE: http://fabioverona.rvsteam.net/

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
71
Abstract Views
357
Rank
589,797
PlumX Metrics