Classifying the time series of stocks into different types of stochastic models

Abstract

Different trading strategies have been developed for different types of models. So being able to identify the asset’s models is crucial in the success of the strategies. In this paper, the behavior and the identification of the time series of stocks into Geometric Brownian Motion, Mean Reversion, or Trend Following processes will be studied. In particular, the Hurst Exponent method will be implemented to classify the stochastic processes, and numerical examples are reported to demonstrate the technique.