A0675
Title: Asymptotic expansions for blocks estimators of cluster indices
Authors: Rafal Kulik - University of Ottawa (Canada) [presenting]
Abstract: Cluster indices describe extremal behaviour of stationary time series. We consider their disjoint and sliding blocks estimators. Using a modern theory of multivariate, regularly varying time series, we obtain a sharp asymptotic expansion on the difference between these two types of estimators. As a consequence, we show that in the Peaks-Over-Threshold framework, sliding and disjoint blocks estimators have the same limiting variance.