Dynamic trade execution: a grammatical evolution approach

Abstract

Trade execution is concerned with the actual mechanicsof buying or selling the desired amount of a financialinstrument. Investors wishing to execute large ordersface a tradeoff between market impact and opportunitycost. Trade execution strategies are designed tobalance out these costs, thereby minimising totaltrading cost. Despite the importance of optimising thetrade execution process, this is difficult to do inpractice due to the dynamic nature of markets and dueto our imperfect understanding of them. In this paper,we adopt a novel approach, combining an evolutionarymethodology whereby we evolve high-quality tradeexecution strategies, with an agent-based artificialstock market, wherein the evolved strategies aretested. The evolved strategies are found to outperforma series of benchmark strategies and several avenuesare suggested for future work.