Application of the Method of Information Decomposition for Revealing the Latent Periodicity in Financial Time Series

Abstract

In the given work results of search of the latent periodicity in financial time series are presented. For these purposes we used the method of information decomposition (ID) developed earlier for revealing periodicity in symbolical sequences. Application of this method to numerical rows became possible after conversion of numbers in symbols. It is shown, that the method of ID is more sensitive, than the classical approach applied to search of periodicity - Fourier analysis.