|
| 文章编号：1672-3104(2013)06-0086-05 |
| |
|
我国创业板市场与中小板市场动态相关性实证研究
——基于方法比较视角 |
| |
耿庆峰 |
| |
|
（闽江学院公共经济学与金融学系，福建福州，350108） |
| |
| 摘 要： DCC-GARCH模型能较好地刻画金融市场间的线性动态相关性，而时变相关的正态Copula模型主要用于考察金融市场间的非线性动态相关性，选取创业板指数(399006)和中小板指数(399005)，通过对两大指数收益率序列进行建模，基于DCC-GARCH模型和Copula模型分别计算两市场间的动态相关系数，以考察两市场间的动态相关性。研究结果表明：① 创业板市场与中小板市场存在正相关关系，且相关性很强；② 时变Copula模型因捕捉了市场收益率随时间变化的特性，在刻画金融市场间的相关性方面，效果要优于常相关Copula模型；③ 除个别时点外，DCC-GARCH模型计算的动态相关系数处于稳定的区间，而Copula模型计算的动态相关系数会表现出异常值的出现，这是由于其考虑了市场间的非线性因素所致，效果优于前者。 |
| |
| 关键词： 创业板市场；中小板市场；动态相关性；DCC-GARCH模型；Copula模型 |
| |
|
| |
The Empirical Research of the Dynamic Correlation Between the Second Board Market And the Small and Medium-sized Board Market in China ——Based on Comparative Perspective of Different Methods |
| |
GENG Qingfeng |
| |
（Department of Public Economics and Finance, Minjiang University, Fuzhou 350108, China） |
| |
| Abstract: The DCC-GARCH model can describe the linear dynamic correlation between financial markets, while the time-varying normal Copula model is mainly used for nonlinear dynamic correlation study between financial markets. This paper aims to study the dynamic correlation between the second board market and SME board market by building models to the return series of the two boards’ indexes and calculating dynamic correlation coefficient of the two markets on the basis of DCC-GARCH model and Copula model. The results show as as follows: ① there is positive correlation between the second board market and SME board market and the correlation is very strong; ② time-varying Copula model is better than constant correlation Copula model in describing the correlations among financial markets as it captures market return’s feature of time-varying; ③ except for a little time-points, dynamic correlation coefficient calculated on the basis of DCC-GARCH model is in a stable interval. Whereas, there are abnormal values in the dynamic correlation coefficients calculated on the basis of Copula model as a result of taking nonlinear factors into consideration. Hence the latter is better than the former. |
| |
Key words: Growth Enterprise Market; Small and medium-sized market; Dynamic Correlation; DCC-GARCH Model; Copula model |
| |
|
