Estimation and comparison of signed symmetric covariation coefficient and generalized association parameter for alpha-stable dependence modeling

Bernédy Kodia (1) , Bernard Garel (1)

1 IMT - Institut de Mathématiques de Toulouse UMR5219

In this paper we study the estimators of two measures of dependence: the signed symmetric covariation coefficient proposed by Garel and Kodia and the generalized association parameter put forward by Paulauskas. In the sub-Gaussian case, the signed symmetric covariation coefficient and the generalized association parameter coincide. The estimator of the signed symmetric covariation coefficient proposed here is based on fractional lower-order moments. The estimator of the generalized association parameter is based on estimation of a stable spectral measure. We investigate the relative performance of these estimators by comparing results from simulations.

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Bernédy Kodia, Bernard Garel. Estimation and comparison of signed symmetric covariation coefficient and generalized association parameter for alpha-stable dependence modeling. Communications in Statistics - Theory and Methods, 2014, vol. 43 (24), p. 5156-5174. ⟨10.1080/03610926.2012.730167⟩. ⟨hal-00951885⟩