Using Short-Term Memory Model (ARIMA) to Predict Value-at-Risk for Damascus Stock Exchange Indices

Abstract

     Investors and decision makers suffer from the volatility and turmoil witnessed in financial markets, which has led to a decline in investor confidence regarding these markets and financial transactions. Therefore, it is necessary to adopt analytical models to model these turbulences, including the Autoregressive Integrated Moving Average Model ARIMA). Therefore, this study aimed to predict the value at risk (VAR) of the Damascus Securities Exchange indices (DWX index, number of transactions index, and trading value index). The researcher calculated the largest possible loss for these indices using the historical method, relying on Excel. She then predicted the VaR for these indices using the ARIMA model, based on Eviews13. The most important results reached by the researcher were that the value at risk of the studied indicators took a decreasing trend, which indicates that the value at risk is constantly increasing. Also, the time series of the VaR of indicators is unstable because the Damascus market is considered an frontier market), and the best model for predicting the value at risk of the DWX index is ARIMA(2,1,2), while the best model for predicting the number of transactions and trading value indicators is ARIMA(0,1,1).