Using Deep Learning to Predict Stock Movements Direction in Emerging Markets: The Case of Qatar Stock Exchange

Deep learning approaches have been utilized to predict stocks. In this study, we use convolutional neural network (CNN) to predict stocks direction in Qatar stock exchange (QE) as a case of emerging markets. Prediction in emerging markets is more challenging than in developed ones because they have higher volatility rate. They are influenced by developed markets and by other external factors including oil price. In this study, we aim to use these external factors to improve the accuracy of the prediction in QE. In addition to historical data, we include data of S&P index, Nikkei index, and oil price in the features of our mode. It is found that using these external factors improves the accuracy of the prediction by 10%.