中国证券市场无限活跃跳跃问题

Infinite Activity Jumps in China Stock Markets

基于任意时间窗口内具有无限到达率的资产价格跳跃行为近期引起学术界的广泛关注。本文对传统的无限活跃跳跃行为辨识方法——阈值p幂次变差(TM PV)方法存在的阈值时变性问题进行了修正,基于蒙特卡洛技术的模拟结果验证了改进之后的模型具有更好的效果。进一步,基于改进的TM PV模型对中国证券市场不同类型个股进行了实证研究,结果发现在中国证券市场无限活跃跳跃是一种常态下的价格行为,这种现象几乎每天都在发生,因此基于无限活跃跳跃的资产价格模型更适合于刻画我国证券市场的价格过程。

This paper focuses on jump behaviors with the infinite arrival rate in any time windows and improves the threshold of time varying in Threshold Multi Power Variation（TMPV） method to identify the infinite activity jump.Monte Carlo simulation further demonstrates the improved model effective theoretically.In addition,we conduct an empirical study on groups of different stocks on asset price jump in China Stock Markets.Results show that the assets price model with infinite activity jump is more suitable to depict asset price process.