Modeling the dependence structure of coupled lives in the Philippines using some Archimedean copulas

• Main Authors: Cagalpin, Anne Katherine A., Micaller, Dhianna Monica B.
• Format: text
• Language: English
• Published: Animo Repository 2009
• Subjects: Copulas (Mathematical statistics); Dependence (Statistics); Statistics and Probability
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/5223
• Institution: De La Salle University

Dependence can be modeled by the use of copulas. This research presents the application of copulas in modeling the dependence structure using a dataset from the Philippine Social Security System. Empirical marginal survival distribution were first computed from the data using the Kaplan-Meier Estimator that led to the computation of Gompertz distribution as the parametric marginal distribution. The bivariate survival function was estimated by the use of a simple non-parametric estimator presented in the study of Lin et al. Through the use of generators for each copula, the difference between the empirical estimates and the theoretical ones were compared and analyzed. Graphs and errors were the basis in choosing which among the candidate copulas best describes the Philippine dataset. Clayton copula best fits the joint survival function in describing the dependence structure for the Philippine dataset.