Classical vs wavelet-based filters Comparative study and application to business cycle
In this article, we compare the performance of Hodrickk-Prescott and Baxter-King filters with a method of filtering based on the multi-resolution properties of wavelets. We show that overall the three methods remain comparable if the theoretical cyclical component is defined in the usual waveband, ranging between six and thirty two quarters. However the approach based on wavelets provides information about the business cycle, for example, its stability over time which the other two filters do not provide. Based on Monte Carlo simulation experiments, our method applied to the American GDP using growth rate data shows that the estimate of the business cycle component is richer in information than that deduced from the level of GDP and includes additional information about the post 1980 period of great moderation.
|Date of creation:||Mar 2010|
|Date of revision:|
|Note:||View the original document on HAL open archive server: http://halshs.archives-ouvertes.fr/halshs-00476022|
|Contact details of provider:|| Web page: http://hal.archives-ouvertes.fr/ |
When requesting a correction, please mention this item's handle: RePEc:hal:cesptp:halshs-00476022. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (CCSD)
If references are entirely missing, you can add them using this form.