R Package Vignette for Boosted HP Filter

Abstract

This vignette introduces the HP filter, the boosted HP filter, and the usage of the R package bHP. The Hodrick-Prescott filter (HP filter; Hodrick and Prescott (1997)) is one of the fundamental statistical tools in macroeconomic data analysis. Thanks to its simplicity, it has been widely used in empirical macroeconomics studies. As an operational algorithm, its pros and cons have been debated over decades, and academic interest has been renewed in recent years to investigate its properties and extensions. While Hamilton (2018) argues against the usage of the HP filter, P. C. Phillips and Shi (2021) propose a machine learning enhancement of the original HP filter, called the boosted HP filter (bHP), and establish consistency when it is applied to a large class of trended time series in macroeconomic applications.

Keywords: HP filter, time series

JEL Classification: C32