调整"已实现"波动率与GARCH及SV模型对波动的预测能力的比较研究  

The Comparative Research on Volatility Prediction Ability of Adjusted Realized Volatility,GARCH Model and SV Model

作　　者：徐正国 张世英[1] 

机构地区：[1]天津大学管理学院,天津300072

出　　处：《系统工程》2004年第8期60-63,共4页Systems Engineering

基　　金：国家自然科学基金资助项目(70171001)

摘　　要：高频金融时间序列的分析与建模是金融计量学的一个全新的研究领域,"已实现"波动率是针对高频金融时间序列的一种全新的波动率的度量方法。为了降低"已实现"波动的测量误差,提出更有效的调整"已实现"波动。针对调整"已实现"波动的长记忆性和"杠杠"效应建立ARFIMAX模型。通过设定一系列标准,全面比较基于调整"已实现"波动的ARFIMAX模型、GARCH模型以及SV模型的预测能力。High-frequency financial time series analysis and modeling is a new research field in financial econometrics, and (realized) volatility is a new measure approach of volatility in high-frequency data field. The paper puts forward a more (efficient) approach which is adjusted realized volatility based on realized volatility. This paper constructs ARFIMAX model (aimed) at the adjusted realized volatility's long memory characteristics and leverage effect. At last, through a variety of (criterions) the paper studies prediction ability of adjusted realized volatility, GARCH model and SV model.

关 键 词：高频金融时间序列 调整“已实现”波动率 二次变差 长记忆性 Mincer—Zarnowitz回归 

分 类 号：F830[经济管理—金融学]