资产价格跳跃行为是连续时间金融学研究的重要问题之一。在Ait-Sahalia成功将跳跃研究视野由狭义事件驱动有限活跃跳跃扩展到广义订单驱动无限活跃跳跃层次的基础上,对阈值多幂次变差模型(TMPV)广义跳跃系列问题研究中存在的阈值时变性问题进行了改进,对其尚未讨论的广义跳跃行为属性存在的采样频率依赖性问题进行了实证研究,在此基础上对不同超高频率下的广义跳跃行为的属性特征进行了刻画,进一步分析不同类型股票的跳跃行为属性特征,结合实证结果对其原因进行了初步分析。
Asset price jump is one of the most important and basic issue in Continuous-Time Finance. On the basis of Sahalia, expending the research perspectives from finite activity jump in the issue of information to infinite activity jump in the issue of order successfully, this article corrects the threshold time-varying problem in Sahalia's new lines of investigation into generalized jump with TMPV method, and discusses the behavioral genotype's sampling frequency dependency in generalized jump which hasn't been studied in Sahalia's research. Further more, this article characterizes the property of jumps under different frequencies, and studies the features of behavioral genotypes of different stocks, and carries out preliminary analysis combined with empirical results.