Median-Unbiased Optimal Smoothing and Trend Extraction
25 Pages Posted: 4 Dec 2008
Date Written: December 4, 2008
Consider the problem of smoothing a time series for extracting its low frequency characteristics, collectively called its trend. This paper proposes a competitive, to existing methods, solution in choosing the optimal degree of smoothing based on the distribution of the residuals from the smooth trend. The methodology is illustrated with simulations and with an application to the U.S. real GDP series, where a comparison of the proposed methodology with the Hodrick-Prescott filter is made.
Keywords: Hodrick-Prescott filter, local linear, moving average, singular spectrum analysis, smoothing, splines, time series, trend extraction, U.S. real GDP
JEL Classification: C14, C32, E32
Suggested Citation: Suggested Citation