Multi-peak Algorithmic Trading Strategies Using Grey Wolf Optimizer

摘要

In this paper, we propose a new method of algorithmic trading for short term investors in the financial markets, by applying swarm intelligence. We apply a well known meta-heuristic, known as Grey Wolf Optimizer (GWO), and find multi-peak optimisation solutions having different expected risk and return ratios, to propose 3 automated trading strategies. The novelty of our method is how we leverage three best swarm agents to construct multi-peak solutions that are best suited for the stochastic nature of financial markets. We utilise the variance between the positions of swarm agents in GWO to construct different algorithmic approaches to day trading, with an aim to diversify expected portfolio volatility. Our research showcases how the three best swarms of GWO are best suited to predict stochastic time series problems, as we typically find in the field of finance. Our experiments demonstrate the capability of our model compared to industry benchmark indices and evaluates the effectiveness of the proposed strategies.