Fragility Distributions and their approximations

Abstract

Given a sequence of $n$ identically distributed random variables with common distribution $F$, the \emph{fragility distribution of order $m$}, represented by $\FD$, is the limit conditional distribution of the number of exceedances given there are at least $m$ exceedances, as the threshold tends to the right end point of $F$. In this paper we are concerned with the existence of $\FD$ and its asymptotic behaviour when $n$ becomes large. For a stationary sequence with its exceedance process converging to a compound Poisson process, we derive an explicit formula for calculating $\lim_{n \to \infty} \FD$. We also establish Stein's method for estimating the errors involved in fragility distribution approximations.