An application of Rolling chaos 0-1 test on Stock Market

Abstract

In this paper we apply a rolling 0-1 test for chaos on different stock market indices returns in the world, considering different time period windows to capture the effects of adding new information. A rolling sample is defined for each index and at the same time, wavelet denoising has been employed since approximately 1995 to the end of 2012. Empirical evidence of continuous chaotic behavior for all indices is found.