基于动态分位点回归模型的金融传染分析

金融危机传染检验一直是国际金融研究中的重要问题,大多数传染效应存在性的检验采用相关性方法.本文应用动态平滑数系数分位点回归模型研究不同国家股票市场之间的分位点相关关系,通过系数函数的变化趋势对危机传染问题进行检验和预测.其中在对参数函数进行非参数估计时,应用局部线性回归方法.为了分析亚洲金融危机期间的危机传染问题,对亚洲几个相关国家的指数数据进行了实证分析,实证结果发现,通过分析常数项函数以及系数函数的变化趋势,不仅可以对危机传染问题进行检验,还可以对金融危机传染的发生时刻以及金融危机的缓解时刻进行相应的分析.

Analysis of financial contagion based on dynamic quantile regression model

The analysis of financial contagion is always an important problem in international finance field. When testing financial contagion,the dependence method is usually adopted.In this paper,a new method called dynamic smooth coefficient quantile regression model is used to analyze the correlation among stock markets from different countries,and the financial contagion is tested and forecasted from the varying tendency of coefficient function.The local linear method is used to estimate the regression coefficient.In order to analyze the financial contagion of Asian Crisis,an empirical analysis is presented for the stock indices from five correlated countries.The empirical result shows that from the varying tendency of the constant and slope coefficient function,not only the contagion can be tested,but also the time when financial contagion starts and eases can be determined.