Extremal clustering under moderate long range dependence and moderately heavy tails (Q2074983)

The paper highlights an unusual clustering of extreme values that occurs in stationary stochastic sequences with subexponential tails. Some basic facts on the subexponential distributions in the Gumbel maximum domain of attraction are reviewed and completed. Random closed sets, null-recurrent Markov chains and limiting random sup-measure are discussed. The authors introduce a family of stationary subexponential infinitely divisible processes. In this frame, some extremal limit theorems in the space of sup-measures and in the space of cadlag functions are proved. Two suitable examples are presented.