Multiresolution analysis of stock market price fluctuations

Abstract. – Recently, we have developed a method based on discrete wavelets to characterize the correlation and scaling properties of non-stationary time series. This approach is local in nature and it makes use of wavelets from the Daubechies family for detrending purpose. The natural built-in variable windows in wavelet transform makes this procedure well suited for the non-stationary data. We analyze daily price of NASDAQ composite index for a period of 20 years, starting from 11-Oct-1984 to 24-Nov-2004, and BSE sensex index, over a period of 15 years, starting from 2-Jan-1991 to 12-May-2005. We find that the present wavelet based analysis clearly reveals long-range correlation as well as multifractal behavior for both the stock index values, which differ from each other significantly.