Technical vs Fundamental Analysis for Egyptian Stock Market Prediction using a Unified Ensemble Regressor

There are two competing schools of thought when it comes to making investment decisions in the stock market - technical analysis and fundamental analysis. In this paper, we study the Egyptian stock market to determine which of these two schools of thought can best describe its performance. Toward that end, we come up with different statistical and machine learning techniques that use the core ideas behind these two philosophies to predict the prices of shares. We then design an ensemble regressor that agglomerates the results of these regressors into a final predicted stock value. The meta unit of the ensemble regressor is trained using a novel technique to ensure that it works with any sample it is presented with. The model is designed to extract information about the general market - a single model is trained on several companies rather than having one model per company - this allows the model to generalize about the relationship between past and future share price as well as fundamental data and future share price. Our study indicates that, for the time-scale under consideration, technical analysis outperforms fundamental analysis significantly.