Thesis (M.Sc.) - Cairo University - Faculty of Commerce - Department of Business Administration

Predicting stock prices and index has been an essential issue for all participants in the financial market nowadays. Through accurate prediction investors can hedge against risks and achieve more profits, governmental institutions can monitor market fluctuations and managers can improve their investment decisions. Different methods and techniques had been already used to make accurate prediction but selecting the most suitable method to improve forecasting performance is the corner stone of prediction process.This study aims at predicting the Egyptian stock market index EGX30 and 14 shares listed on the EGX30 using Artificial Neural Network (ANN) model and comparing the results of this model with ones that were provided by another forecasting techniques.This study employs ARCH/GARCH model to control the forecasting results of the ANN model and to compare between forecasts and errors of both techniques using seven input independent variables to forecast closing prices of shares, these variables are (trade volume, closing price, moving average of five days, moving average of twenty days, stochastic oscillator %k, and stochastic oscillator % D and EGX30 index)

Issued also as CD