On robust tail index estimation for linear long-memory processes

被引:8
作者
Beran, Jan [1 ]
Das, Bikramjit [2 ]
Schell, Dieter
机构
[1] Univ Konstanz, Dept Math & Stat, D-78457 Constance, Germany
[2] Swiss Fed Inst Technol, Zurich, Switzerland
关键词
Long memory; infinite variance; tail index; M-estimation; RANGE DEPENDENT SEQUENCES; SAMPLING WINDOW METHOD; CENTRAL LIMIT-THEOREMS; EMPIRICAL PROCESS; ASYMPTOTIC NORMALITY; MOVING AVERAGES; HILL ESTIMATOR; CONVERGENCE; PARAMETER; BOOTSTRAP;
D O I
10.1111/j.1467-9892.2011.00774.x
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
We consider robust estimation of the tail index a for linear long-memory processes with i.i.d. innovations ej following a symmetric a-stable law (1 < a < 2) and coefficients aj similar to c.j-beta. Estimates based on the left and right tail respectively are obtained together with a combined statistic with improved efficiency, and a test statistic comparing both tails. Asymptotic results are derived. Simulations illustrate the finite sample performance.
引用
收藏
页码:406 / 423
页数:18
相关论文
共 63 条
[1]  
[Anonymous], 1986, TRANSLATIONS MATH MO
[2]  
Astrauskas A., 1983, LITOVSK MAT SB, V23, P3
[3]  
AVRAM F, 1986, DEPENDENCE PROBABILI, P399
[4]   Semiparametric lower bounds for tail index estimation [J].
Beirlant, J ;
Bouquiaux, C ;
Werker, BJM .
JOURNAL OF STATISTICAL PLANNING AND INFERENCE, 2006, 136 (03) :705-729
[5]  
Beran J, 1997, ANN STAT, V25, P1852
[6]  
Beran J., 2010, COMPUTATION IN PRESS
[7]  
Beran J., 1994, Statistics for Long-Memory Processes
[8]   Reduction principles for quantile and Bahadur-Kiefer processes of long-range dependent linear sequences [J].
Csorgo, Miklos ;
Kulik, Rafal .
PROBABILITY THEORY AND RELATED FIELDS, 2008, 142 (3-4) :339-366
[9]   CENTRAL LIMIT-THEOREMS FOR SUMS OF EXTREME VALUES [J].
CSORGO, S ;
MASON, DM .
MATHEMATICAL PROCEEDINGS OF THE CAMBRIDGE PHILOSOPHICAL SOCIETY, 1985, 98 (NOV) :547-558
[10]   KERNEL ESTIMATES OF THE TAIL INDEX OF A DISTRIBUTION [J].
CSORGO, S ;
DEHEUVELS, P ;
MASON, D .
ANNALS OF STATISTICS, 1985, 13 (03) :1050-1077