Nowcasting Inflation

RePEc Short ID: pkn16
ORCID iD: 0000-0001-7378-916X

Knotek, Edward S.; Zaman, Saeed

doi:10.26509/frbc-wp-202406

Working Paper

Nowcasting Inflation

Abstract: This chapter summarizes the mixed-frequency methods commonly used for nowcasting inflation. It discusses the importance of key high-frequency data in producing timely and accurate inflation nowcasts. In the US, consensus surveys of professional forecasters have historically provided an accurate benchmark for inflation nowcasts because they incorporate professional judgment to capture idiosyncratic factors driving inflation. Using real-time data, we show that a relatively parsimonious mixed-frequency model produces superior point and density nowcasting accuracy for headline inflation and competitive nowcasting accuracy for core inflation compared with surveys of professional forecasters over a long sample spanning 1999–2022 and over a short sample focusing on the period since the start of the pandemic.

Keywords: inflation; nowcasting; mixed-frequency models; survey nowcasts; real-time data;

JEL Classification: C53; E3; E37;

https://doi.org/10.26509/frbc-wp-202406