A comparison of two business cycle dating methods

Don Harding; Adrian Pagan

doi:10.1016/S0165-1889(02)00076-3

A comparison of two business cycle dating methods

Don Harding^*, Adrian Pagan

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

145 Citations (Scopus)

Abstract

We study the suggestion that Markov switching (MS) models should be used to determine cyclical turning points. A Kalman filter approximation is used to derive the dating rules implicit in such models. We compare these with dating rules in an algorithm that provides a good approximation to the chronology determined by the NBER. We find that there is very little that is attractive in the MS approach when compared with this algorithm. The most important difference relates to robustness. The MS approach depends on the validity of that statistical model. Our approach is valid in a wider range of circumstances.

Original language	English
Pages (from-to)	1681-1690
Number of pages	10
Journal	Journal of Economic Dynamics and Control
Volume	27
Issue number	9
DOIs	https://doi.org/10.1016/S0165-1889(02)00076-3
Publication status	Published - Jul 2003

Access to Document

10.1016/S0165-1889(02)00076-3

Cite this

@article{3288956a71754b749dfa4edbfef5f629,

title = "A comparison of two business cycle dating methods",

abstract = "We study the suggestion that Markov switching (MS) models should be used to determine cyclical turning points. A Kalman filter approximation is used to derive the dating rules implicit in such models. We compare these with dating rules in an algorithm that provides a good approximation to the chronology determined by the NBER. We find that there is very little that is attractive in the MS approach when compared with this algorithm. The most important difference relates to robustness. The MS approach depends on the validity of that statistical model. Our approach is valid in a wider range of circumstances.",

keywords = "Business cycle, Dating rules, Markov switching, NBER dates, Turning points",

author = "Don Harding and Adrian Pagan",

year = "2003",

month = jul,

doi = "10.1016/S0165-1889(02)00076-3",

language = "English",

volume = "27",

pages = "1681--1690",

journal = "Journal of Economic Dynamics and Control",

issn = "0165-1889",

publisher = "Elsevier B.V.",

number = "9",

}

TY - JOUR

T1 - A comparison of two business cycle dating methods

AU - Harding, Don

AU - Pagan, Adrian

PY - 2003/7

Y1 - 2003/7

N2 - We study the suggestion that Markov switching (MS) models should be used to determine cyclical turning points. A Kalman filter approximation is used to derive the dating rules implicit in such models. We compare these with dating rules in an algorithm that provides a good approximation to the chronology determined by the NBER. We find that there is very little that is attractive in the MS approach when compared with this algorithm. The most important difference relates to robustness. The MS approach depends on the validity of that statistical model. Our approach is valid in a wider range of circumstances.

AB - We study the suggestion that Markov switching (MS) models should be used to determine cyclical turning points. A Kalman filter approximation is used to derive the dating rules implicit in such models. We compare these with dating rules in an algorithm that provides a good approximation to the chronology determined by the NBER. We find that there is very little that is attractive in the MS approach when compared with this algorithm. The most important difference relates to robustness. The MS approach depends on the validity of that statistical model. Our approach is valid in a wider range of circumstances.

KW - Business cycle

KW - Dating rules

KW - Markov switching

KW - NBER dates

KW - Turning points

UR - http://www.scopus.com/inward/record.url?scp=0037410926&partnerID=8YFLogxK

U2 - 10.1016/S0165-1889(02)00076-3

DO - 10.1016/S0165-1889(02)00076-3

M3 - Article

SN - 0165-1889

VL - 27

SP - 1681

EP - 1690

JO - Journal of Economic Dynamics and Control

JF - Journal of Economic Dynamics and Control

IS - 9

ER -

A comparison of two business cycle dating methods

Abstract

Access to Document

Other files and links

Fingerprint

Cite this