COMPARISON OF OBJECT-ORIENTED PROGRAMMING AND DATA-ORIENTED DESIGN FOR IMPLEMENTING TRADING STRATEGIES BACKTESTER

Abstract

This research proposes a way to accelerate backtesting of trading strategies using data-oriented design (DOD). The research discusses the differences between DOD and object-oriented approach (OOP), which is the most popular at the current moment. Then, the paper proposes efficient way to parallelize a backtesting using DOD. Finally, this research provides a performance comparison between DOD and OOP backtester implementations on the example of typical technical indicators. The comparison shows that use of DOD can speed up the process of quantitative features calculation up to 33% and allows for parallelization scheme that better utilizes resources in multiprocessor systems.