Algorithmic Trading Strategies: Automating and Back-testing the Perfect Order Strategy

Abstract

The evolution of technology alongside the development of new techniques of algorithmic trading over the past 30 decades (Narang, 2009) allowed global financial markets to achieve higher transaction volume and execution efficiency (Kissell, 2006). In this context, those who fail to adapt to this reality may not survive in financial markets in the future (Chan, 2009). For that, as an attempt to participate in the ongoing automated trading evolution, the present study aims to back-test the Perfect Order Strategy (Lien, 2015) in some selected FX pairs through a fully automated trading system. As a part of the methodology process, the author developed the referred automated trading system through the use of different algorithmic techniques, trading, and risk management models available in the literature, see (Basso, 2019; Leshik & Cralle, 2011; Narang, 2009; Neely et al., 2014; Wilder Jr., 1978). Although the strategy had a positive return at the end of the tests, it performed below the S&P500 index over the same period. Moreover, the results from the back-test showed that the strategy was able to identify trends in its early stages reasonably. In turn, the automated trading system and the advantages that an algorithm execution-based system brought to the strategy played an important role in controlling losses and, therefore, protecting the investment capital. However, the procedures for establishing the stop-loss limit order and the take-profit target showed a flaw and were responsible, in part, for the poor performance of the strategy. Indeed, we are confident that further research in general, particularly in the stop-loss and take-profit target procedures, could improve the strategy's overall performance.