A semi-parametric l-moments estimation method for copulas under right-censoring

Abstract

We present a new semi-parametric estimator for Archimedean copula models in the present research. Based on the classical copula L-moment estimation in complete data, we improve this approach for data that is subject to right-censorship. The estimation relies on conditional moment methods, accounts for various censoring patterns, and is based on the empirical survival copula. Moreover, the suggested estimator's asymptotic normality is proven, offering theoretical support for the proposed estimator. A simulation study is conducted to illustrate the performance and robustness of the estimator, highlighting its advantages in modeling the dependence structure in censored bivariate survival settings.