Forecasting the Futuristic Directions of DWX Returns Using ARMA-GARCH Models

الملخص

This research aimed to evaluate the effectiveness of using the Generalized Autoregressive Conditional Heteroskedasticity models to study the volatility of the Damascus Stock Exchange index during the period from 14/9/2015 to 1/12/2016 and also to forecast the last month futuristic values. To achieve this holding period return was used under the assumption that the cash distributions are zero, and by following the Box-Jenkins method, which is used in time series analysis, a group of results were obtained. The time series of the returns of the index does not follow a random walk status during the studied period due to inability to detect a unit root, also the DWX returns time series follow Autoregressive process of the second grade, in addition to following a process of moving averages of the first grade ARMA (2,1), that is, the current value of the return of the index is affected by its value in the previous two days as well as affected by a set of random variables of the current day and the previous day. And the variance of the series follows a GARCH (1, 1). Based on the estimated model, Risk-Returns of the index were forecast for the period from 1/11/2016 to 1/12/2016. When compared with the actual values, the convergence between them was observed, demonstrating the ability and effectiveness of the proposed model to describe the behavior of the index's return and volatility during the period In addition to its ability to deliver relatively small error predictions, which declared through the low values of MSE and MAE.