摘要
在较弱的假设下导出合格波动率代理变量的必要条件,并借助重抽样技术构建了检验该必要条件的非参数方法。用其进行中外股指高频数据的实证研究表明:对于中国股指的数据,采用"已实现波动率"作为代理变量违背了上述必要条件而有不适合作为合格代理变量之虞;但欧美股市的数据却未检测出上述问题。最后,根据实证结果对"已实现波动率"进行适当改造,使其成为能避免上述偏差的改进代理变量。
In view of the importance of volatility proxy for model evaluation,we propose a necessary condition to be qualified proxies of volatility.A robust approach is given to check the condition in order to investigate the justification of realized volatilities as proxies of volatility.This approach works through comparing the performance of proxies in each sub-sample and whole sample respectively.Applying the approach to stock index,we found that there are some evidences which strongly suggest that realized volatilities are not qualified proxies of volatility for Shanghai stock market data,while the evidences that realized volatilities fail to meet the necessary condition were not detected for Europe and America stock market data.Finally,based on the empirical results,this paper further propose a simple method to revise those proxies and a revised volatility proxy is given as well.
作者
施雅丰
应婷婷
史彦龙
范奎奎
SHI Ya-feng;YING Ting-ting;SHI Yan-long;FAN Kui-kui(School of Science,Ningbo University of Technology,Ningbo 315211,China;Faculty of Business,University of Nottingham,Ningbo 315100,China;School of Basic Courses,Zhejiang Pharmaceutical College,Ningbo 315100,China;School of Statistics,Capital University of Economics and Business,Beijing 100070,China)
出处
《统计与信息论坛》
CSSCI
北大核心
2018年第10期43-48,共6页
Journal of Statistics and Information
基金
浙江省教育厅一般科研项目<股市技术分析交易与市场波动率的相关性研究>(Y201635394)
宁波市软科学项目<进口食品安全问题对消费者购买行为的影响因素分析及对策研究--基于风险认知和信任视角>(2017A10113)
国家自然科学青年基金项目<新视觉模型下非完整移动机器人同时镇定和跟踪控制研究>(61503205)