Algorithmic Trading Strategies in the Cryptocurrency Market

Details

Supervisors

Iania, Leonardo

Faculty

Louvain School of Management

Degree label

Master [120] en Ingénieur de gestion

Abstract

Cryptocurrencies have recently received increasing attention in the fields of finance and computer science. Likewise, the application of machine learning techniques have shown successful in cryptocurrency price forecasting. In our paper, we investigate the prediction performance of Support Vector Machine (SVM), the Multilayer Perceptron (MLP) and the Long Short Term Memory (LSTM) for directional price movement on Bitcoin and Ethereum, based on technical, fundamental and social indicators. Subsequently, we build an hourly trading algorithmic strategy based on the best-in class prediction performances for both Bitcoin and Ethereum separately. Obtained results suggest the MLP outperforms both SVM and the LSTM for predictions directional cryptocurrency price movements. Highest prediction performances are achieved through the combination of the trading indicators. Finally, our trading algorithm outperforms the classic buy-andhold benchmark strategy on absolute and risk-adjusted performance metrics.