Improving the Predictability of Damascus Securities Exchange index's Movement Using a Proposed Method to Build a Hybrid Model between Artificial Neural Networks and ARCH-ARIMA Models

الملخص

This study aimed at presenting a proposed method for constructing a hybrid model between artificial neural networks and ARIMA-ARCH models. In order to improve the ability to predict the direction of the movement of the financial market index, by applying it to DSE index, and to achieve the research goal, a set of common hybridization methods have been applied and compared with the proposed method, the forecast periods were divided into two periods, from (19/08/2019 to 19/09/2019) and (from 19/09/2019 to 21/10/2019) and calculating the trend-fit ratio and RMSE for the two periods. The most important results reached are; the proposed method was the most appropriate to predict the movement of the DSE index movement. As this method does not depend on calculating the merging weights on model errors in the estimation stage, but on the ability of the models to simulate the direction of movement of the DSE Index series, which gives a greater ability to improve the predictability of the DSE index directions' during the two forecasted periods when compared to other hybridization methods.