Quantifying the Linguistic Complexity of Pan-Homophonic Events in Stock Market Volatility Dynamics

Area of Science:

• Behavioral Finance
• Computational Linguistics
• Financial Econometrics

Background:

• Pan-homophonic events represent a novel phenomenon where phonetic similarities influence stock market behavior.
• Understanding these linguistic-driven market dynamics is crucial for financial analysis.

Purpose of the Study:

• To investigate the impact of pan-homophonic events on stock prices, using Chuandazhisheng as a case study.
• To quantify the dynamic and time-varying effects of these events on market volatility.

Main Methods:

• A linguistic amplitude segmentation method based on information entropy was used to classify event intensity.
• The Relational Stock Ranking (RSR) model and Dynamic Conditional Correlation-Generalized Autoregressive Conditional Heteroskedasticity (DCC-GARCH) framework were employed to isolate event-specific price movements.

Main Results:

• Stock prices exhibit a sequential response, with low-amplitude events causing moderate fluctuations and high-amplitude events leading to significant volatility.
• Price surges, typically occurring within 1-2 days, tend to revert within three weeks.
• Repeated exposure to pan-homophonic stimuli results in an attenuated market response, indicating a decaying spillover effect.

Conclusions:

• Linguistic cues significantly impact financial markets, influencing stock price dynamics.
• The study provides insights into investor behavior, information processing, and the need for regulatory oversight in response to such events.