Realized GARCH模型在黄金期货市场的应用  

Application of Realized GARCH Model in Gold Future Market

作　　者：王玲[1] 

机构地区：[1]上海理工大学

出　　处：《财务与金融》2015年第3期89-95,共7页Accounting and Finance

摘　　要：采用结合高频数据实现计量的Realized GARCH模型对黄金期货市场的杠杆效应、波动集群性等波动性特征进行研究。选取对价格稳健的实现计量作为Realized GARCH模型的解释变量,并对我国黄金期货的高频样本数据进行检验。极大似然方法估计结果表明:对价格跳稳健的实现双幂变差波动性(RBV)所得结果更优,黄金期货存在明显的杠杆效应和波动集群性,且波动具有长持续性,而实现方差并不能有效的刻画数据实际特征。This paper studies the leverage effect and volatility clustering of gold future in China by using realized GARCH model combined with realized measures calculated from high-frequency data. The realized measure robust to price jumps is selected as the explanatory variable of realized GARCH model to test the features of high-frequency data of gold future in China. The results got from maximum likelihood estimation show： the Realized Bipower Variation （RBV） which robust to price jumps performs good, gold future presents leverage effect and volatility clustering, and volatility lasts a long time; but Realized Variance （RV） can＇t describe the real characters of the data effectively.

关 键 词：黄金期货 非对称性 Realized GARCH 极大似然估计 

分 类 号：F831[经济管理—金融学]