Comparative Study of Stock Price Forecasting Using Arima and Arimax Models
This thesis evaluates the in-sample forecasting accuracy of two forecasting models namely ARIMA and ARIMAX. . Data used was monthly adjusted close price of four stocks in the Oil and Gas Industry in the London Stock Exchange from 2005 - 2010 with a total of 72 observations. The Mean Square Error (MSE), Mean absolute Error (MAE) and Root Mean Square Error (RMSE) serve as the error matrices in evaluating the forecastability of the models. The effect of Akaike Information Criterion (AIC) and the linear correlation on candidate models among the considered stocks were tested.The ARIMAXmodels performed well with lower error metrics as compared to the ARIMA models in all time regimes.The Linear Correlation on the other hand had little or no influence at all on the in – sample forecastability as compared to the AIC which had significant influence on the error metric. The results support that themarket is efficient and hence no investor has undue advantage of gaining from it.
A thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirements for the degree of Master Philosophy in Applied Mathematics