Skip to Main content Skip to Navigation
Master Thesis

Prise en compte de l'hétérogénéité des individus dans le modèle de Lee-Carter

Abstract : Looking at the mortality rates in time, one can check that people use to live older. This phenomenon leads to important matters on financial institutions such as pension fund system or life annuities business. This is why, actuaries are getting used to model and forecast mortality in order to assess the longevity risk. The most famous model is the Lee-Carter one, which has been applied to several studies and american institutions, and most of times, gave consistent and attractive results. However, this model gives rise to several inconsistencies, like an homoskedascticity of errors terms and narrow confidence bound. The model we developed here is an extension of that one, based on the assumption that the number of deaths can be model with a Binomiale Negative distribution. According to that hypothesis, we remove the homoskedasticity assumption and take into account the degree of heterogeneity that arise from the population. The perks of this model lead to wider confidence intervals for the projection of the mortality rates and a better goodness-of-fit.
Document type :
Master Thesis
Complete list of metadata

Cited literature [81 references]  Display  Hide  Download
Contributor : Ufr de Mathématique-Informatique / Irma <>
Submitted on : Friday, October 10, 2014 - 9:46:32 AM
Last modification on : Friday, October 23, 2020 - 4:56:28 PM
Long-term archiving on: : Sunday, January 11, 2015 - 10:20:12 AM


  • HAL Id : dumas-01073547, version 1



Cédric Boureau. Prise en compte de l'hétérogénéité des individus dans le modèle de Lee-Carter. Gestion des risques [q-fin.RM]. 2014. ⟨dumas-01073547⟩



Record views


Files downloads