A new approach on tail index estimation is proposed based on studying the in-sample evolution of appropriately chosen diverging statistics. The resulting estimators are simple to construct, and they can be generalized to address other rate estimation problems as well.
Une nouvelle méthode est proposée pour l'estimation de l'index d'une queue de distribution. Elle est basée sur l'étude de statistiques divergentes. Les estimateurs résultants sont simples à construire et peuvent être utilisés pour résoudre d'autres problèmes d'estimation.
Accepted:
Published online:
Dimitris N. Politis 1
@article{CRMATH_2002__335_3_279_0, author = {Dimitris N. Politis}, title = {A new approach on estimation of the tail index}, journal = {Comptes Rendus. Math\'ematique}, pages = {279--282}, publisher = {Elsevier}, volume = {335}, number = {3}, year = {2002}, doi = {10.1016/S1631-073X(02)02450-0}, language = {en}, }
Dimitris N. Politis. A new approach on estimation of the tail index. Comptes Rendus. Mathématique, Volume 335 (2002) no. 3, pp. 279-282. doi : 10.1016/S1631-073X(02)02450-0. https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/S1631-073X(02)02450-0/
[1] Kernel estimates of the tail index of a distribution, Ann. Statist., Volume 13 (1985), pp. 1050-1077
[2] Estimating the tail index (B. Szyszkowicz, ed.), Asymptotic Methods in Probability and Statistics, North-Holland, Amsterdam, 1998, pp. 833-881
[3] Modelling Extremal Events, Springer, Berlin, 1997
[4] Necessary conditions for the bootstrap of the mean, Ann. Statist., Volume 17 (1990), pp. 684-691
[5] T. McElroy, D.N. Politis, Robust inference for the mean in the presence of serial correlation and heavy tailed distributions, Econometric Theory, 2002, forthcoming
Cited by Sources:
Comments - Policy