Evaluating a stock market pairs trading strategy using a dependent functional data cointegration method
Danni Wanga, Zhifang Sua
Keywords:
Dependent functional data, Co-integration test, Pair trading
Abstract:
Pairs trading, recognized as an effective high-frequency investment strategy to mitigate risk, has been widely employed in statistical arbitrage. Traditional pairs trading approaches often impose strict constraints on asset selection, and in view of the functional and dependent characteristics of financial market data, this study proposes an innovative pairs trading strategy based on a dependent functional cointegration model. Specifically, we first utilize a dependent functional cointegration test to identify contract pairs exhibiting cointegration relationships, then establish a signal mechanism and threshold setting for the trading phase, and finally carry out arbitrage tests in stock markets. The results show that, compared with traditional methods, this approach not only respects the stochastic nature of data but also operates under a nonparametric framework, enabling accurate recognition of dependence and dynamic changes in high-frequency data. It thereby precisely captures short-term mean reversion trends, achieves higher trading frequency, and yields significantly improved returns. Consequently, this method offers trading institutions a practical tool for constructing a universal quantitative trading system in fiercely competitive markets, contributing to the overall competitiveness and efficiency of the financial sector.
Fulltext download:
 Evaluating a stock market pairs trading strategy using a dependent functional data cointegration method [PDF file] [Filesize: 993.73 KB]
 10.7441/joc.2025.04.08
Wanga, D., & Sua, Z. (2025). Evaluating a stock market pairs trading strategy using a dependent functional data cointegration method. Journal of Competitiveness, 17(4), 207-229. https://doi.org/10.7441/joc.2025.04.08
|
