Statistical Empowerment Based on the GJR-GARCH Three-Stage Model: An Empirical Investigation
ITM Web of Conferences · 2026 · 93 · 01034
DOI: 10.1051/itmconf/20269301034 ↗Kurz gefasst
Fachartikel in der Zeitschrift ITM Web of Conferences, online erschienen am 13.11.2026. Verfasst von Xuanming Cui und 3 weiteren Autor:innen. Die Arbeit nennt 11 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study converts qualitative event descriptions into a mathematical feature system for volatility forecasting and strategy construction. The first stage follows the event-study framework and builds an Event Feature Matrix from five dimensions: occurrence probability, impact intensity, impact breadth, decay, and duration. Expected returns are estimated with the Fama–French three-factor model; cumulative abnormal return (CAR) and cumulative average abnormal return (CAAR) measure intensity and breadth; Poisson and exponential-decay specifications describe event occurrence and persistence. The second stage embeds the resulting Event Impact Score (EIS) sequences in a GJR-GARCH-X model. Maximum likelihood estimation is used to distinguish event-driven volatility from volatility clustering and leverage effects. In the reported out-of-sample tests, the GJR-GARCH-X model produces lower root mean square error (RMSE) values than the benchmark GJR-GARCH model, indicating that event information improves volatility forecasts for the examined stocks. The third stage links these forecasts to a dynamic arbitrage framework based on mean-variance optimization. Transaction costs and position constraints are included, while a pre-trade CVaR constraint and a post-trade dynamic stop-loss rule control tail risk. Backtesting against the CSI 300 Index and an event-free specification shows improvements in the reported Sharpe ratio, maximum drawdown, and alpha. Overall, the framework provides a structured route from event quantification to volatility forecasting and portfolio decisions.
Zitieren
APA
Cui, X., Liu, E., Chen, S., & Li, Z. (2026). Statistical Empowerment Based on the GJR-GARCH Three-Stage Model: An Empirical Investigation. ITM Web of Conferences, 93, 01034. https://doi.org/10.1051/itmconf/20269301034
Vancouver
Cui X, Liu E, Chen S, Li Z. Statistical Empowerment Based on the GJR-GARCH Three-Stage Model: An Empirical Investigation. ITM Web of Conferences. 2026;93:01034. doi:10.1051/itmconf/20269301034
BibTeX
@article{Cui2026,
author = {Cui, Xuanming and Liu, Enxu and Chen, Shunzhe and Li, Zhengqi},
title = {Statistical Empowerment Based on the GJR-GARCH Three-Stage Model: An Empirical Investigation},
journal = {ITM Web of Conferences},
year = {2026},
volume = {93},
pages = {01034},
doi = {10.1051/itmconf/20269301034}
}
EUVIMED