Contributions for modeling characterization of heavy-tail time series

Abstract

The occurrence of extreme phenomena and their devastating impact have been on the agenda, especially in areas of environmental and economic-financial sciences, extending to insurance activity. The theory of extreme values allows an adequate approach in the statistical study of data associated with this type of phenomena. Heavy tail models thus play an important role and are increasingly a resource. In this work we will revisit some max/min-autoregressive and maximum-moving models and contribute to their characterization by deriving their autocorrelation structure based on the Spearman and Kendall coefficients, both useful tools in the identification of models in real data applications.