摘要
本文在对沪深300和S&P500股指期货的当月连续合约进行展期处理的基础上,基于条件收益分别服从正态、学生t、GED和skewed-t分布的假设,运用GJR GARCH模型对波动非对称性建模,并对模型设定偏误进行严格诊断检验。研究发现:GJR GARCH模型能很好地捕捉股指期货市场波动的非对称性;基于skewed-t分布的波动模型的准确性明显优于其他分布下的相同模型;与S&P500股指期货市场相比,我国股指期货市场波动的非对称性较弱。
In this paper, two continuous price series of CSI300 and S&PS00 index future are constructed with rollover. Under the assumption of the conditional return obeying normal, student t, GED and skewed - t distributions, four GJR GARCH models are used to model asymmetry volatility, and Sign Bias Tests are used to test the misspecification of the GJR model. The results show that the GJR models can effectively capture the asymmetry of index futures ; the GJR mod- els with skewed- t distribution is superior to the models with other distributions; contrasted with obvious asymmetry of S&PS00 index future, CSI300 index future is less asymmetric.
出处
《商业研究》
CSSCI
北大核心
2015年第5期73-78,共6页
Commercial Research
基金
国家自然科学基金项目
项目编号:71301095
上海市自然科学基金项目
项目编号:11ZR1411800
上海财经大学博士研究生创新基金项目
项目编号:CXJJ-2013-407