Filtering short term fluctuations in inflation analysis

Many economic time series, specifically inflation, are subject to seasonal fluctuations. Seasonal adjustment is a powerful tool for removing such fluctuations. Meanwhile, seasonal adjustment may provide highly volatile series, as it deals with certain type of movements completed on seasonal frequencies. However, short term fluctuations occurring at non-seasonal frequencies are also possible. In the context of inflation, we propose a two-step methodology combining wavelet and band-pass filters and aiming at removing all short term fluctuations within a year. This method yields much smoother series than seasonal adjustment provides. Moreover, filtered series capture the dynamics of inflation in subgroups well.