Volatility measurement with pockets of extreme return persistence
针对日内出现极端收益持续性的时段，提出一种新的波动率估计量（差分收益波动率估计量），能减少常用估计量的偏差，并在S&P 500指数期货和个股的预测应用中表现更优。
Increasing evidence points towards the episodic emergence of pockets with extreme return persistence. This notion refers to intraday periods of non-trivial duration, for which stock returns are \nhighly positively autocorrelated. Such episodes include, but are not limited to, gradual jumps and \nprolonged bursts in the drift component. In this paper, we develop a family of integrated volatility estimators, labeled differenced-return volatility (DV) estimators, which provide robustness \nto these types of Itˆo semimartingale violations. Specifically, we show that, by using differences \nin consecutive high-frequency returns, our DV estimators can reduce the non-trivial bias that \nall commonly-used estimators exhibit during such periods of apparent short-term intraday return predictability. A Monte Carlo study demonstrates the reliability of the newly developed \nvolatility estimators in finite samples. In our empirical volatility forecasting application to S&P \n500 index futures and individual equities, our DV-based Heterogeneous Autoregressive (HAR) \nmodel performs well relative to existing procedures according to standard out-of-sample MSE \nand QLIKE criteria. \n