Comparing Performance of Price Ratio-based and Linear Regression-based Pair Trading Strategies Through Backtesting

A strategy of pairs trading comprises of identifying of relationships in given stock data, determining its direction, and then ascertaining how to execute a trade based on that direction. Among numerous approaches taken to build this strategy, the Price Ratio-based trading strategy (PRBTS) and the Linear Regression-based trading strategy (LRBTS) are most commonly used. This paper seeks to compare and analyze these two distinct strategies using backtesting results and some performance metrics. For backtesting, a portfolio was constructed consisting of pairs that were eligible for trade via both approaches. This portfolio was implemented in two unique trading systems which were designed in compliance with both the strategies to generate signals and execute position sized trades by tracking the pairs on a daily basis. Both the systems were trained and optimized in the training period from 1 st January 2022 to 31 st December 2022 to find the maximum profit potential for the subsequent testing period. The period from 1 st January 2023 to 31 st March 2023 was chosen as the testing period as it was a very turbulent phase for the Indian stock market and assessing the performance in this period would provide a greater insight into the proficiency and capability of the two strategies in adverse market conditions. The portfolio results of both the training and testing periods are revealed and compared with the returns of Nifty 50 Index as well. In the training period, at maximum profit potential, results show that the PRBTS produced more trading signals and profits whilst in the testing period, it couldn’t create as many trading opportunities although contributing sufficiently to the performance of the strategy. However, 417 the LRBTS was consistent in generating adequate trading signals that yielded stable returns throughout both the periods.