|
In this paper we propose a new jump robust measure of ex-post return variation that can be
computed using potentially noisy data. The estimator exploits the link between return quantiles
and volatility and is consistent for the integrated (diffusive) variance under weak conditions on
the price process. We present various central limit theorems which show that the estimator
converges at the best attainable rate and has excellent efficiency. Asymptotically, the estimator
is immune to finite activity jumps and simulations show that also in finite sample it has superior
robustness properties. In modified form, the estimator is applicable with market microstructure
noise and therefore operational on high frequency data. As such, it constitutes an appealing
alternative to the existing jump-robust or noise-corrected realised variance measures. An
empirical application using low and high frequency data is included to further illustrate the
properties of the estimator.
Joint with Kim Christensen and Mark Podolskij
|