噪声、跳跃与高频价格波动——基于门限预平均实现波动的分析

摘要

高频数据中的噪声和价格跳跃使得波动的估计缺乏一致性，本文提出用门限预平均实现波动的方法估计同时存在市场微观结构噪声和价格跳跃时高频价格波动，该方法是资产价格实际波动的一致估计，并有最优的收敛速度。模拟发现，门限预平均实现波动和常用的高频波动估计方法相比，有更小的均方误差。中国证券市场的实证分析表明，门限预平均实现波动能减少波动预测误差，得到更为精确的风险管理价值。%This paper provides both theoretical and empirical results of threshold pre-averaging realized volatility which is the consistent estimator of the integrated price volatility. Simulations with the constant and stochastic volatility show that TPRV also has superior robustness properties in finite samples. In the empirical application, threshold pre-averaging realized volatility has smaller HMSE in forecasting than the BPV estimator and more precise Value at Risk.