Realized range-based threshold estimation for jump-diffusion models

We develop a framework for estimating the quadratic variation of discontinuous semi-martingales with intra-day high-low statistics. Restricting the realized rangebased variance smaller than a suitably defined threshold, we propose an integrated volatility estimator and consider its consistency and asymptotic normality under a set of weak conditions. We find that the precision of our statistics is about five times greater than that of realized variance purely restricted by threshold. Simulation results illustrate the good finite sample properties of our estimator.