Weighted empirical processes in the nonparametric inference for Lévy processes

Given observations of a Lévy process, we provide nonparametric estimators of its Lévy tail and study the asymptotic properties of the corresponding weighted empirical processes. Within a special class of weight functions, we give necessary and sufficient conditions that ensure strong consistency and asymptotic normality of the weighted empirical processes, provided that complete information on the jumps is available. To cope with infinite activity processes, we depart from this assumption and analyze the weighted empirical processes of a sampling scheme where small jumps are neglected. We establish a bootstrap principle and provide a simulation study for some prominent Lévy processes.