The Dynamic Correlation between News Sentiment and Housing Price Volatility: A Heterogeneity Perspective

• Published in: Journal of Applied Science and Engineering Vol. 29; no. 2; pp. 315 - 327
• Main Authors: Yuxing Liu, Wei Chong Choo, Keng Yap Ng, Feifei Li
• Format: Journal Article
• Language: English
• Published: Tamkang University Press 2026
• Subjects: dcc-garch model; heterogeneity analysis; housing price volatility; news sentiment; spillover index model; tvp-sv-var model
• ISSN: 2708-9967; 2708-9975
• DOI: 10.6180/jase.202602_29(2).0007

Based on the differences between traditional news and digital news, this paper takes a heterogeneous perspective to systematically analyze the dynamic correlation degree, time-varying impulse response, and spillover effects between traditional news sentiment, digital news sentiment, and housing price volatility using DCC-GARCH model, TVP-SV-VAR model, and Spillover Index model. Through the application of machine learning techniques, the research results are obtained: (a) The correlation between the three variables exhibits significant temporal variability, especially in the later stage of the sample period. (b) There is a positive dynamic correlation between news sentiment from two different sources, and housing price volatility is significantly affected by spillover effects from both types of news sentiment. (c) Traditional news sentiment dominates in the early stage of the sample period, but its influence gradually fades in the later stage. In contrast, the positive impact of digital news sentiment on housing price volatility is more sustained. We believe that in the current context of diversified development in the media industry, policy makers should pay attention to the complex dynamic correlation between different news sentiment and housing price volatility. While focusing on regulating digital media, the influence of traditional media cannot be ignored.