In certain cases partial sums of i.i.d. random variables with finite variance are better approximated by a sequence of stable distributions with indices \alpha_n \to 2 than by a normal distribution. We discuss when this happens and how much the convergence rate can be improved by using...

Estimators of the extreme-value index are based on a set of upper order statistics. We present an adaptivemethod to choose the number of order statistics involved in an optimal way, balancing variance and biascomponents. Recently this has been achieved for the similar but somewhat less involved...

In certain cases partial sums of i.i.d. random variables with finite variance are better approximated by asequence of stable distributions with indices alpha n - 2 than by a normal distribution. We discusswhen this happens and how much the convergence rate can be improved by using penultimate...

In certain cases partial sums of i.i.d. random variables with finite variance are better approximated by asequence of stable distributions with indices alpha n - 2 than by a normal distribution. We discusswhen this happens and how much the convergence rate can be improved by using penultimate...

An abundance of high quality data sets requiring heavy tailed models necessitates reliablemethods of estimating the shape parameter governing the degree of tail heaviness.The Hill estimator is a popular method for doing this but its practical use isencumbered by several difficulties. We show...

In certain cases the distribution of the normalized maximumof a sample can be better approximated by a sequence ofdifferent extreme value distributions than by the final one. Weshow that these cases are rather restricted and that the possibleimprovement is not spectacular.

The theory of stable probability distributions and their domains of attraction is derived in a direct way(avoiding the usual route via infinitely divisible distributions) using Fourier transforms. Regularly varyingfunctions play an important role in the exposition.

Estimators of the extreme-value index are based on a set of upper order statistics. We present an adaptive method to choose the number of order statistics involved in an optimal way, balancing variance and bias components. Recently this has been achieved for the similar but somewhat less...

The paper characterizes first and second order tail behavior ofconvolutions of i.i.d. heavy tailed random variables with supporton the real line. The result is applied to the problem of riskdiversification in portfolio analysis and to the estimation of theparameter in a MA(1) model.

Estimators of the extreme-value index are based on a set of upper order statistics. We present an adaptivemethod to choose the number of order statistics involved in an optimal way, balancing variance and biascomponents. Recently this has been achieved for the similar but somewhat less involved...