Share price changes using MST-based sequential change point detection

• Published in: Sequential analysis Vol. 44; no. 3; pp. 351 - 365
• Main Authors: Mahawar, Jayant, Chattopadhyay, Bhargab
• Format: Journal Article
• Language: English
• Published: Taylor & Francis 03.07.2025
• Subjects: Graph theory; minimum spanning tree; scan statistics; sequential change point detection
• ISSN: 0747-4946; 1532-4176
• DOI: 10.1080/07474946.2025.2506420

Critical events, when they occur, pose a significant impact on the stocks of one or more sectors of stock markets. Because of that effect, the prices of stocks belonging to the same stock market sector change comparably over time. Small variations in stock prices are natural, but if the change is significant, it results in a change in the statistical distribution of the stock price. Early detection of this distributional change in prices of stocks belonging to a particular sector often helps decide on investment decisions like selling, buying, or holding stocks. Thus, the accurate and fast detection of these change points in stock prices is critical in share trading. In this article, we consider the problem of early detection of a change point, if it occurs. In the process, we developed a sequential change point detection algorithm with no distributional assumptions and compared it with existing algorithms. Application to a portfolio of stocks from the banking sector associated with the National Stock Exchange of India and simulation study shows that our detection algorithm outperforms an existing change point detection method to detect change earlier.