Thesis

Bachelor Thesis in Engineering Physics

Stochastic Modeling of Financial Timeseries

Summary Modern finance is dependent on reliable models for stock price processes. We model the time correlation of the square of the logarithm of the increments of the price process using a stochastic volatility process. We also show that the model can be improved further by assuming that data corrected for time correlation is generated by a Lévy process. Based on this assumption we can fit a general model for the distribution of data generated by a Lévy process to empirical, time correlation corrected data, thus resulting in a further improvement of our model.