On the ideality of filtering techniques in the business cycle analysis under conditions of European economy

Abstract

The paper deals with the applications and comparison of various filtering techniques, which is an integral part of the business cycles identification and their further analysis. In particular, the paper examines four high-pass and band-pass filters and compares them to an ideal filter. The gain function is estimated and shown by the periodogram to assess the filters' efficiency in the extraction of undesired periods in the time series. The monthly data of industrial production indices of chosen EU countries are used to analyse the filtering techniques. The results show the Butterworth filter as a high pass and the Christiano-Fitzgerald filter as a band-pass to be closer to an ideal filter than the Hodrick- Prescott and Baxter-King filters.