Explanatory Change Detection in Financial Markets by Graph-Based Entropy and Inter-Domain Linkage

Area of Science:

• Financial markets
• Data visualization
• Network analysis

Background:

• The COVID-19 pandemic caused significant structural shifts in financial markets.
• Understanding these changes is crucial for informed investment decisions and future preparedness.
• Traditional analysis methods may not fully capture complex market dynamics.

Purpose of the Study:

• To analyze structural changes in financial markets during the COVID-19 pandemic.
• To develop a novel method for detecting market changes to aid investors.
• To evaluate the effectiveness of the developed change-detection indicator.

Main Methods:

• Financial markets were modeled as a graph structure.
• The graph was analyzed by dividing it into specific domains.
• An original change-detection indicator was designed based on graph topology.
• The indicator's performance was compared against a benchmark method.

Main Results:

• The developed graph-based indicator showed higher effectiveness compared to the benchmark.
• The indicator demonstrated improvements in both response speed and accuracy.
• Domain-specific analysis provided a detailed understanding of market shifts.

Conclusions:

• Graph visualization and domain analysis offer valuable insights into financial market structures.
• The novel change-detection indicator provides a more effective tool for investors.
• This approach supports the development of specific investment strategies in response to market volatility.