We consider kernel-based estimators of integrated variances in the presence of independent market microstructure effects. We derive the bias and variance properties for all regular kernel-based estimators and derive a lower bound for their asymptotic variance. Further we show that the subsample-based estimator is closely related to a Bartlett-type kernel… (More)
A number of authors have reported empirically observed scaling laws of the absolute values of log returns of stocks and exchange rates, with a scaling coefficient in the order of 0.58–0.59. It is suggested here that this phenomenon is largely due to the semi-heavy tailedness of the distributions concerned rather than to real scaling.