Statistical analysis of financial time series is studied. We use wavelet analysis to study signal to noise ratios along with auto-correlation function to study correlation length for time series data of daily stock prices for specific sectors of the market. We study the "high beta" stocks versus the "low beta" stocks. We sample ten companies from both of these sectors. We find that the signal to noise ratio is not uniformly high for the "high beta" classified stocks nor is the correlation length large for the "high beta" classified stocks. We explain reasons for this and possible further applications.

Author Bio

This work was done in the Summer 2001 REU program at Indiana University Mathematics Department. At that time (and currently), I was an undergraduate at Princeton University. Currently, I am a rising senior (in the fall 2002) at Princeton University. I am mainly interested in working on problems in applied mathematics. During the latter parts of my high school and initial parts of my undergraduate years, I worked on a problem in mathematical biology. I worked on studying the information content calculations for different types of DNA based sequences across all three domains of life. I have published a paper on this work in Physica D: Nonlinear Phenomenon Journal(). I intend to go to graduate school in applied mathematics. I plan to pursue a Ph.D. in a problem in financial mathematics. I believe that there are a lot of possibilities of using methods in classical physics and mathematics to attack challenging problems in quantitative finance.