Authors: Clément Marsilli
Addresses: Banque de France, International Macroeconomics Division, 31 rue Croix des Petits Champs, 75001 Paris, France
Abstract: We present a mixed-frequency model for real-time monitoring of US inflation. Our approach relies on a mixed-data sampling (MIDAS) augmented Phillips curve with daily oil prices for nowcasting inflation. In line with the literature we find that performances of inflation nowcasting models rely on two key elements: the inclusion of high-frequency oil prices and a rolling-window framework. Our approach succeeds in providing a policy-oriented tool for monitoring inflation in real-time.
Keywords: inflation nowcasting; MIDAS; mixed-data sampling; oil prices; US inflation; USA; United States; Phillips curve; inflation monitoring; rolling window; real-time monitoring.
International Journal of Computational Economics and Econometrics, 2017 Vol.7 No.1/2, pp.64 - 77
Received: 21 Mar 2015
Accepted: 23 Feb 2016
Published online: 21 Nov 2016 *