摘要
本文介绍了对ARCH/GARCH模型的两种估计方法:准极大似然估计和极小绝对偏差估计,并提出了一种基于自助法(Bootstrap)对估计方法的选择。在厚尾程度不同的情况下进行了模拟分析,表明对于一个具体的数据,该选择法能够自动选择较优的估计方法。并用该方法对上海证券交易所A股和B股的股价指数进行了分析,印证了上海股市B股收益率的尾部厚于A股收益率尾部。
In this paper,we introduce two estimations on GARCH model:one is Quasi-Maximum Likelihood Estimation(QMLE),the other is the Least Absolute Deviation Estimator(LADE).We also suggest an estimation selection method based on bootstrap strategy.We present the effect of this selection method by models with tails of different heaviness and show the method's power in selecting a better estimation.We also implement our methods on the real data of Shanghai Stock Price Index,and confirm that the yield's volatility of B share is heavier than its counterpart of A share.
出处
《数理统计与管理》
CSSCI
北大核心
2010年第3期544-549,共6页
Journal of Applied Statistics and Management
基金
国家自然科学基金资助
项目批准编号10771006