Financial forecasting using the Kolmogorov-Feller equation

Abstract

An approach to analysing a financial time series using the Kolmogorov-Feller Equation is considered, in particular, the Generalised Kolmogorov-Feller Equation (GKFE), subject to variations in the Stochastic Volatility. Using the Mittag-Leffler memory function, we derive an expression for the Impulse Response Function associated with a short time window of data which is then used to derive an algorithm for computing a new index using a standard moving window process. It is shown that application of this index to financial time series, subject to a low volatility condition, correlates with the start, direction and end of a trend depending on the sampling rate of the time series and the look-back window or 'period' that is used. An example of this is provided in the chapter using MetaTrader4. © 2013 Springer Science+Business Media Dordrecht.