Best Quadratic Unbiased Estimators of Integrated Variance in the Presence of Market Microstructure Noise

Yixiao Sun (University of California, San Diego)


We consider the best quadratic unbiased estimators of the integrated variance in the presence of independent market microstructure noise. We establish the asymptotic normality of a feasible best quadratic unbiased estimator under the assumption of constant volatility and show that it is asymptotically efficient when the market microstructure noise is normal. Since the class of quadratic estimators includes all the existing estimators of the integrated variance as special cases, the best quadratic unbiased estimator outperforms the existing estimators in terms of root mean squared error, often by a large margin.