A1097
Title: Nonparametric range-based estimation of integrated variance with episodic extreme return persistence
Authors: Yifan Li - The University of Manchester (United Kingdom)
Ingmar Nolte - Lancaster University (United Kingdom)
Sandra Nolte - Lancaster University (United Kingdom)
Shifan Yu - Lancaster University (United Kingdom) [presenting]
Abstract: A new nonparametric estimator of integrated variance is developed based on intraday candlestick information (high, low, open, and close prices in short time intervals). This range-return difference volatility (RRDV) estimator is robust to short-lived extreme return persistence hardly attributable to the diffusion component, such as gradual jumps and flash crashes. By modelling such sharp but continuous price movements following two recent influential works, it is shown that RRDV can provide consistent estimates with relatively small variances. Simulation results demonstrate the reliability of the proposed estimator in practice with some finite-sample refinements. An empirical illustration of volatility forecasting shows the RRDV-based Heterogeneous Autoregressive (HAR) model performs well relative to existing procedures according to standard out-of-sample loss functions.
