Trend Extraction From Time Series With Structural Breaks

Schlicht, Ekkehart (May 2007): Trend Extraction From Time Series With Structural Breaks. Discussion Papers in Economics 2007-17 Final version in Journal of the Japan Statistical Society 2009, 38(2), 285-92. http://www.jstage.jst.go.jp/article/jjss/38/2/285/_pdf

Abstract

Trend extraction from time series is often performed by using the filter proposed by Leser (1961), also known as the Hodrick-Prescott filter. A practical problem arises, however, when the time series contains structural breaks (such as produced by German unification for German time series, for instance). This note proposes a method for coping with this problem.

• Item Type: Paper
• Keywords: Trend extraction, structural break, Hodrick-Prescott filter, Leser filter, spline, time-series, smoothing, interpolation.
• Faculties: Economics; Economics > Munich Discussion Papers in Economics; Economics > Munich Discussion Papers in Economics > Statistical Methods; Economics > Chairs > Chair of Institutional Economics (closed)
• Subjects: 300 Social sciences > 300 Social sciences, sociology and anthropology; 300 Social sciences > 330 Economics
• JEL Classification: C22, C32, C63, C14
• Language: English