Algorithmic Forex Trading using Combination of Numeric Time Series and News Analysis

Abstract

Foreign Currency Exchange market (Forex) is a highly volatile complex time series. Predicting the trends in Foreign Exchange prices is a very challenging task due to the many uncertainties involved and many variables that influence the market value in a particular day such as economic factors, political events, government debt etc which cannot be anticipated just by considering numeric time series data. This paper proposes a prediction model that combines artificial neural network (Long Short term memory model) for historical analysis and support vector machine for news analysis which will consider the impact of above mentioned factors. The performance measure for LSTM is quantified in terms of mean absolute error, mean square error and root mean square error. SVM is evaluated using confusion matrix. Next, the predictions from both these models were integrated to design trading rules which eventually maximizes the profit and reduces the risk.