Algorithmic Currency Trading using NEAT - based Evolutionary Computation

This paper introduces NEAT-based Evolutionary Computation as the basis for a fully automated trading system application. The system is designed to trade FX markets by detecting profitable currency cycles in the most widely traded currencies and forecasting future exchange rates. To do this, it relies on a fitness function that measures profitability, as well as a fitness function that measures accuracy of future predictions. The trading system is able to generate consistent gains based on this back-tested and predictive approach.