In the talk we discuss the results of several recent papers of V. Paulauskas, M. Vaiciulis, and Y. Qi on the tail index estimation. In these papers generalization of an estimator, introduced by Davydov, Paulauskas and Rackauskas and which is based not on rank statistics, are proposed. Performance of these generalized estimators is compared with the well-known estimators, based on rank statistics. As one of advantages of these new estimators is the fact that they are well adapted to work with the high frequency financial data on real time.
Keywords: Tail index; Heavy-tailed distributions; Estimation
Biography: Vygantas Paulauskas - professor, habilitated doctor, head of the chair of Mathematical analysis at Vilnius university. Elected member of Lithuanian Academy of Sciences and International Statistical Institute, member of the boards of Lithuanian Mathematical Society and Lithuanian Statistical Society, president of Lithuanian Statistical Society (2003-2009). The holder of two National Awards in Science (1979, 1998), the author of more than 100 research papers and one monograph.