A0429
Title: Corrected inference about the extreme expected shortfall in the general max-domain of attraction (part II)
Authors: Abdelaati Daouia - Toulouse School of Economics (France)
Gilles Stupfler - University of Angers (France)
Antoine Usseglio-Carleve - Avignon Université (France) [presenting]
Abstract: The finite-sample applicability of asymptotic theory for expected shortfall estimation above extreme quantiles in the class of distributions with finite first tail moment, regardless of whether the underlying extreme value index is positive, negative, or zero, is discussed and found to often fail to yield confidence intervals whose empirical coverage probability matches nominal coverage. By relying on the moment estimators of the scale and shape extreme value parameters, as well as on a fine understanding of the dependence structure between these estimators and intermediate expected shortfall estimators, corrected asymptotic confidence intervals are constructed whose finite-sample coverage is found to be close to the nominal level on simulated data. The usefulness of the construction is illustrated on two sets of financial loss returns and flood insurance claims data.
