We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.AP

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Applications

Title: Lapse risk modelling in insurance: a Bayesian mixture approach

Abstract: This paper focuses on modelling surrender time for policyholders in the context of life insurance. In this setup, a large lapse rate at the first months of a contract is often observed, with a decrease in this rate after some months. The modelling of the time to cancellation must account for this specific behaviour. Another stylised fact is that policies which are not cancelled in the study period are considered censored. To account for both censuring and heterogeneous lapse rates, this work assumes a Bayesian survival model with a mixture of regressions. The inference is based on data augmentation allowing for fast computations even for data sets of over a million clients. Moreover, scalable point estimation based on EM algorithm is also presented. An illustrative example emulates a typical behaviour for life insurance contracts and a simulated study investigates the properties of the proposed model. In particular, the observed censuring in the insurance context might be up to 50% of the data, which is very unusual for survival models in other fields such as epidemiology. This aspect is exploited in our simulated study.
Subjects: Applications (stat.AP); Computation (stat.CO)
Cite as: arXiv:2207.07203 [stat.AP]
  (or arXiv:2207.07203v1 [stat.AP] for this version)

Submission history

From: Viviana Lobo [view email]
[v1] Thu, 14 Jul 2022 21:03:02 GMT (2746kb,D)

Link back to: arXiv, form interface, contact.