高频农产品期货波动率和相关性预测——基于Realized Copula-DCC模型的视角  

Volatility and Correlation Prediction of High Frequency Agricultural Futures:Based on Realized Copula-DCC Model

作　　者：黄雯[1] 黄卓[2] 王天一[3] 

机构地区：[1]北京大学国家发展研究院 [2]北京大学国家发展研究院中国经济研究中心,助理教授北京100871 [3]对外经贸大学金融学院,助理教授北京100029

出　　处：《浙江社会科学》2013年第5期40-47,156,共9页Zhejiang Social Sciences

基　　金：教育部人文社会科学青年基金(项目编号71201001);国家自然科学青年基金(项目编号71201001)的资助

摘　　要：本文构建了Realized Copula-DCC模型,整合Realized GARCH模型和Copula-DCC模型对农产品期货的波动率和动态相关性进行研究。农产品期货不仅表现出波动聚类现象、偏斜和尖峰厚尾的特征,还呈现出非正态性。基于Skewed-t分布的Realized GARCH模型比其他模型更好地刻画了农产品期货的波动率特征。农产品期货的相关性呈现出动态变化,tCopula-DCC模型比其他时变Copula模型更好地反映了农产品期货相关性的动态演化过程。The paper constructed the Realized Copula-DCC model, and studied on the volatility and dynamic correlation of agricultural futures by integrating Realized GARCH model and Copula-DCC model. Agricultural futures not only presented the characteristics of volatility clustering, skewness and fat-tail, but also appeared non-normality. The Realized GARCH model based on Skewed-t distribution better depicted the volatility of agricultural futures than other models. The correlation of agricultural futures showed dynamic changes, t Copula-DCC model better reflected the dynamic evolution process of the correlation of agricultural futures than other time-varying Copula models.

关 键 词：Realized Copula-DCC Realized GARCH COPULA 波动率 动态相关性 

分 类 号：F224[经济管理—国民经济] F323.7F724.5