噪音环境下跳特征函数的非参数估计  

Nonparametric Estimation of Jump Characteristics under Market Microstructure Noise

作　　者：付金玉 林金官 FU Jinyu;LIN Jinguan(School of Mathematics,Southeast University,Nanjing,210096,China;School of Statistics and Data Science,Nanjing Audit University,Nanjing,211815,China)

机构地区：[1]东南大学数学学院,南京210096 [2]南京审计大学统计与数据科学学院,南京211815

出　　处：《应用概率统计》2022年第6期887-903,共17页Chinese Journal of Applied Probability and Statistics

基　　金：国家自然科学基金项目(批准号:11971235、11831008);全国统计科学研究项目(批准号:2021LY007)资助。

摘　　要：本文研究了在噪音环境下,跳特征函数的非参数估计问题.结合极差增量和门限技术提出新的统计量,并给出相应的大样本性质.实验表明提出的估计量对噪音是稳健的,尤其是在极高频数据的时候.By exploiting financial high frequency data,we nonparametrically estimate the jump characteristic in the presence of market microstructure noise.Our estimator is based on the realized range increments and threshold technique.Besides,the bias caused by microstructure noise can be estimated and removed,if it is modeled as ask-bid spread,which is a used frequently assumption.We further present the asymptotic properties of the proposed estimator.Simulation studies show the estimator works well under microstructure noise.Finally,the estimator is also applied to the real data.

关 键 词：跳特征函数 噪音 极差 买卖差价 门限技术 

分 类 号：O212.1[理学—概率论与数理统计]