Analysis of stock market data by using Dynamic Fourier and Wavelets techniques

This work deals with the analysis of daily and minute sampled financial stock market data. We propose a Dynamic Fourier Transform (DFT) and a Wavelet Transform to estimate the power spectrum of returns. In order to estimate the power spectrum, we used the tapering process with the DFT technique and the scaling function with the wavelets methodology to avoid the spectral leakage or discontinuity in the sequence. Our result suggest that the power spectrum are effective in characterizing the minute and daily based data corresponding to different frequencies. This type of modeling techniques help to characterize some key variables of stationary time series that are very useful for making informed decisions in the stock market such as assessing financial risk in the market.