中国股市波动的异方差模型及其SPA检验  

Heteroskedastic Volatility Models in Chinese Stock Market and Their SPA Test

作　　者：魏宇[1] 

机构地区：[1]西南交通大学经济管理学院,成都610031

出　　处：《系统工程理论与实践》2007年第6期27-35,共9页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70501025);国家杰出青年科学基金(70229001)

摘　　要：以中国股票市场最具代表性的股价指数-上证综指的高频(High-frequency)数据样本为例,实证计算了以GARCH族模型和随机波动(Stochastic volatility)模型为代表的不同异方差模型对中国股市波动率的预测,并进一步运用SPA(Superior predictive ability)检验法,实证检验了不同异方差模型对中国股市波动的刻画能力和预测精度问题.实证结果显示,就中国股市而言,随机波动(Stochastic volatility)模型是预测精度最高的异方差模型,但在某些损失函数标准下,EGARCH模型也具有良好的波动预测表现.One high-frequency dataset of the most important stock index in Chinese stock market is used to calculate the volatility forecasts based on different heteroskedastic volatility models, such as GARCH models and stochastic volatility model. We compare the forecasting performance of different kinds of volatility models using SPA test. The empirical results show that, stochastic volatility model is the best models for volatility forecasts in Chinese stock market. However, under some kinds of loss functions, EGARCH model also performance quite well.

关 键 词：异方差 实现波动率 GARCH模型 随机波动模型 SPA检验 

分 类 号：F224[经济管理—国民经济]