A model for stock price forecasting based on ARMA systems

Abstract

The Prediction of the future values of a stock market signal on the basis of its past and present data series, is one of the most necessities of all financial applications. In this study, one special stock market signal is considered and analyzed using “ARMA” model with different number of poles and zeros, in order to estimate the values for the next days` prices. The estimated and the actual data for the next day is compared and the amount of error for each system is calculated, resulting into selection of most efficient model.