Bias-correcting the realized range-based variance in the presence of market microstructure noise (Q964674)

The authors analyze the impact of noisy high-frequency data on the realized range-based variance (RRV). They propose a new robust range-based estimator, which is consistent for the integrated variance and asymptotically mixed Gaussian in the presence of simple forms of microstructure noise. It is shown how to optimally divide high-frequency data such that the conditional variance of the asymptotic distribution is minimized. The finite sample properties of the estimator are studied by Monte-Carlo simulations. It is established that a bias-corrected range-statistic often leads to much smaller confidence intervals for the integrated variance, relative to the realized variance.