摘要
本文采用半鞅过程的双重次方变差和多重次方变差构造的跳检验统计量,对沪深300指数进行日内跳检验。检验结果表明,沪深300指数多于1/3的交易日有跳发生,跳发生的概率在不同交易时段并不均匀,并表现出一定的持续性。计算结果表明,跳二次变差在二次变差中占的比例高达32.48%,是二次变差的重要组成部分。
With the statistics constructed from bipower variation and multipower variation of semimartingale, the paper tests for intraday jumps in CSI300.The tests showed that jumps in CSI300 occurred more than one third trading days.The probability of jump varies over trading time and the persistence of jump is moderate.By calculation,the quadratic variation from jump accounts for 32.48 percent of total quadratic variation and therefor is important component.
出处
《数理统计与管理》
CSSCI
北大核心
2010年第4期713-718,共6页
Journal of Applied Statistics and Management
