摘要
探索金融市场极端风险传导机理一直是政府管理当局、投资者关注的焦点。本文针对股市中存在的典型事实及股市损失分布复杂性特点,运用ARMA-GJR对股市指数条件损失进行建模分析,进而运用EVT对标准残差的极值尾部建模估计出股市极端风险ES,然后运用Granger-Causality检验技术,分别考察两个市场间极端风险ES的传导关系。实证结果表明:在整个样本期间,中国大陆沪深股市极端风险具有双向传导关系,香港市场向深市传导风险,而深市不能向香港传导风险,东京市场与香港市场、香港与台湾市场具有双向传导关系;而在熊市期间,中国大陆与周边市场极端风险ES传导关系变得更为复杂。
Both the government and investors are keen to find contagion mechanism of extreme risk in financial markets. This paper applies ARMA-GJR to model conditional loss of different stock index for Chinese stock and other circumjacent markets in Asia based on some stylized facts, then applies EVT to model their standard residuals series and gets dynamic extreme risk ES time series for every stock market index. At last, the paper applies Granger-Causality technology to analyze contagion effect for these extreme risk ES. The results show that there exist two-way contagion relationship between SSEC and SZEC, between SZEC and STI, and extreme risk contagion from HSIN to SZEC risk throughout the entire sampling period; the contagion effect become more complex during bear period.
出处
《管理评论》
CSSCI
北大核心
2012年第9期64-74,共11页
Management Review
基金
国家自然科学基金项目(71071131
71171025)
国家社会科学基金项目(12BGL024)
教育部人文社会科学研究青年基金(10YJCZH086)
成都理工大学中青年骨干教师培养计划(2011-013)
关键词
金融市场
典型事实
极端风险
熊市期间
传导效应
financial market, stylized facts, extreme risk, bear period, contagion effect