Scholarlink

<p>The selection of stock pairs in pairs trading strategy (PTS) is critical to the strategy’s success, as it determines the ability to exploit price movements and manage risk effectively. Existing methods like correlation-based, cointegration-based, and machine learning often overlook deeper relationships, and short-term fluctuations and struggle with data complexity, performing poorly in volatile periods like COVID-19. This paper focuses on a novel pair selection method that enhances performance by leveraging volatility and residuals for improved long-term stability and profitability. This approach improves pair selection precision and enhances trading performance by integrating volatility metrics with regression-derived residuals. Analyzing historical data, the proposed strategy reveals significantly superior returns, showcasing its peak performance in a volatile market environment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing Pairs Trading Through a Volatility-Based Residual Selection Approach

  • 期刊论文

  • 发表日期：

    2026年5月12日

查看全文
  • Pranjala G. Kolapwar,
  • Uday V. Kulkarni,
  • Jaishri M. Waghmare

摘要

The selection of stock pairs in pairs trading strategy (PTS) is critical to the strategy’s success, as it determines the ability to exploit price movements and manage risk effectively. Existing methods like correlation-based, cointegration-based, and machine learning often overlook deeper relationships, and short-term fluctuations and struggle with data complexity, performing poorly in volatile periods like COVID-19. This paper focuses on a novel pair selection method that enhances performance by leveraging volatility and residuals for improved long-term stability and profitability. This approach improves pair selection precision and enhances trading performance by integrating volatility metrics with regression-derived residuals. Analyzing historical data, the proposed strategy reveals significantly superior returns, showcasing its peak performance in a volatile market environment.