基于小波变换的多尺度跳跃识别与波动性估计研究  

Research of multi-scale jump identification and volatility estimation based on wavelet transform

作　　者：王春峰[1] 姚宁[1] 房振明[1] 

机构地区：[1]天津大学管理学院金融工程研究中心,天津300072

出　　处：《管理科学学报》2010年第10期63-68,共6页Journal of Management Sciences in China

基　　金：国家杰出青年科学基金资助项目(70225002);国家自然科学基金资助项目(70771076)

摘　　要：研究了市场微观结构噪音与跳跃同时存在条件下均衡价格波动性的估计问题.应用时间序列变点的小波分析方法,对2006~2007年上证50指数成分股包含市场微观结构噪音和跳跃的高频价格采样数据的跳跃变差和积分波动性进行了有效的估计.研究结果发现小波分析方法能够准确识别我国股市的价格跳跃,并且跳跃变差与积分波动性的比值很大,忽视跳跃的存在将会引起波动性的有偏估计,表明在金融市场中应当使用均衡价格波动性进行风险管理和资产配置.This paper studies how to estimate the volatility of asset equilibrium price with jump and market microstructure noise existing at the same time.Employing the wavelet method of time series change-point analysis,jump variations and integrated volatilities are estimated using high frequency price sampling data which contain market microstructure noise and jumps for the individual firms of Shanghai 50 index during 2006~2007.We find that price jumps of Chinese stock market are accurately identified by wavelet transform method and the ratios of jump variations and integrated volatilities are large.Ignoring the existence of jump will lead to biased estimation of volatility,indicating that the volatility of equilibrium price should be used for risk management and asset allocation in financial market.

关 键 词：市场微观结构噪音 跳跃 波动 离散小波变换 

分 类 号：F830.91[经济管理—金融学]