基于动态相关性的我国银行系统性风险度量研究

Banking Systemic Risk: Evidence from Dynamic Correlation Study

本文以银行股票收益率的动态相关性作为银行业系统性传染风险的市场衡量指标,采用多元GARCH-BEKK模型对我国上市银行自1999-2010年间的动态相关系数进行测算,并应用Markov区制转换模型对系统性风险状态进行识别和判断,研究表明,银行股票收益率的动态相关系数可以作为监测系统性风险的一项有效市场指标,结合Markov区制转换模型可以很好的判断整体系统性风险的变化,识别出高风险状态,为系统性风险的监测提供一定预警和防范作用。

This paper uses dynamic correlation of bank stock return as market index to investigate bank systemic risk. We use multivariate GARCH-BEKK model to analyze the time varying correlation of all listed banks between 1999 and 2010, and also use Markov regime switching model to test different regimes. We find that dynamic correlation of bank stock return can effectively indicate bank systemic risk. According to Markov regime switching model, ＂high risk＂ regime can be identified to help judge systemic risk and pro- vide a role in early-warning and preventing bank risks.