Bivariate Archimedean copula modelling for loss-ALAE data in non-life insurance
(2004)
Files
dp0423.pdf
Open Access - Adobe PDF
- 706.16 KB
Details
- Authors
- Abstract
- This paper considers the bivariate loss-ALAE modelling problem in actuarial science, taking into account the particular form of the censorship present in the data. Specifically, a graphical selection procedure for the generator of the underlying archimedean copula is proposed, based on a nonparametric estimation of the generator. The approach is in line with Genest & Rivest (1993) who considered complete data and Wang & Wells (2000b) who treated doubly censored data. A loss-ALAE data set provided by the US Insurance Services Office is used for the numerical illustrations, and comparisons with previous results appeared in the actuarial literature are performed.
- Affiliations
Citations
Denuit, M., Purcaru, O., & Van Keilegom, I. (2004). Bivariate Archimedean copula modelling for loss-ALAE data in non-life insurance (STAT Discussion Paper 0423). https://hdl.handle.net/2078.5/34201