Boosting: Why you can use the HP filter
Boosting: Why you can use the HP filter
We propose a procedure of iterating the HP filter to produce a smarter smoothing device, called the boosted HP (bHP) filter, based on L2‐boosting in machine learning. Limit theory shows that the bHP filter asymptotically recovers trend mechanisms that involve integrated processes, deterministic drifts, and structural breaks, covering the most common trends that appear in current modeling methodology. A stopping criterion automates the algorithm, giving a data‐determined method for data‐rich environments. The methodology is illustrated in simulations and with three real data examples that highlight the differences between simple HP filtering, the bHP filter, and an alternative autoregressive approach.
Phillips, Peter Charles Bonest
f67573a4-fc30-484c-ad74-4bbc797d7243
Shi, Zhentao
157ef919-197f-4e66-9be8-8f75716d3430
Phillips, Peter Charles Bonest
f67573a4-fc30-484c-ad74-4bbc797d7243
Shi, Zhentao
157ef919-197f-4e66-9be8-8f75716d3430
Phillips, Peter Charles Bonest and Shi, Zhentao
(2020)
Boosting: Why you can use the HP filter.
International Economic Review.
(doi:10.1111/iere.12495).
Abstract
We propose a procedure of iterating the HP filter to produce a smarter smoothing device, called the boosted HP (bHP) filter, based on L2‐boosting in machine learning. Limit theory shows that the bHP filter asymptotically recovers trend mechanisms that involve integrated processes, deterministic drifts, and structural breaks, covering the most common trends that appear in current modeling methodology. A stopping criterion automates the algorithm, giving a data‐determined method for data‐rich environments. The methodology is illustrated in simulations and with three real data examples that highlight the differences between simple HP filtering, the bHP filter, and an alternative autoregressive approach.
Text
HP_filter_D3_pcb
- Accepted Manuscript
More information
Accepted/In Press date: 7 November 2020
e-pub ahead of print date: 1 December 2020
Additional Information:
Funding Information:
This article is an updated and revised version of an earlier working paper entitled “Boosting the Hodrick Prescott Filter” (Phillips and Shi, 2019). PCBP and ZS contributed equally to this article. We thank the Editor, two referees, Michael Zheng Song, and Brendan Beare for helpful comments and suggestions. We thank Yang Chen for excellent research assistance. Phillips acknowledges research support from the Kelly Foundation at the University of Auckland, the NSF under Grant No. SES 18‐50860, and an LKC Fellowship at Singapore Management University. Shi acknowledges support from the Hong Kong Research Grants Council Early Career Scheme No. 24614817. Please address correspondence to: Peter C.B. Phillips, Cowles Foundation for Research in Economics, Yale University, Box 208281, New Haven, CT 06520‐8281. E‐mail: peter.phillips@yale.edu . 1
Publisher Copyright:
© (2020) by the Economics Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research Association
Identifiers
Local EPrints ID: 445521
URI: http://eprints.soton.ac.uk/id/eprint/445521
ISSN: 0020-6598
PURE UUID: 0ce8f2b9-a65a-438d-9cb7-43a0f0befbd5
Catalogue record
Date deposited: 14 Dec 2020 17:32
Last modified: 17 Mar 2024 06:08
Export record
Altmetrics
Contributors
Author:
Zhentao Shi
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
