A comparative performance analysis for the commonly used time series filters in economics : Hodrick-Prescott versus Baxter-King

Abstract

This thesis compares the performance of the Hodrick-Prescott filter commonly employed in economic analysis to separate the trend of a given non-stationary time series from its cyclical components, to that of the Band-Pass filter developed by Baxter and King. The performances of detrending techniques under consideration are evaluated by constructing special time series that mimic the pattern of actually observed series of interest using synthesized cyclical and trend components. As an illustration of the use of this approach, the behavior of the ISE-100 index of Istanbul Stock Exchange and the Jasdaq index of Japanese Stock Market are analyzed.