Las series de tiempo fractales y un método de pronóstico

Abstract

The authors present here a forecast method for fractal time series, and a variation of the formula of Feder, which simulates fractional brownian motion. We applied this method to the daily returns of the Indice de Precios y Cotizaciones (IPC) of the Mexican Stock Market, from November 22. 2002 to February 16.2004. Finally we compare the results with other two methods, including and econometric AR(1) prediction method.