Анотация: One of the most important questions in Extreme Value Theory is the task for estimation of cumulative distribution function (c.d.f.) and quantiles of an observed random variable (r.v.) outside the range of the data. In particular, its solution helps one to answer the questions about Value at Risk estimation, description of the annual maximum temperatures within a region, maximum level of water in a river, sea or ocean, maximum monthly precipitations, the size of the claim causing ruin, among others. It is closely related to estimation of the return period of a fixed level, which is important to know for predicting natural disasters, financial crises, as well as any kind of extreme behaviour. (авторът)