Ruin problems in the generalized Erlang(n) risk model
Author(s) -
Agnieszka I. Bergel,
Alfredo D. Egı́dio dos Reis
Publication year - 2016
Publication title -
european actuarial journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.661
H-Index - 13
eISSN - 2190-9741
pISSN - 2190-9733
DOI - 10.1007/s13385-016-0130-2
Subject(s) - erlang (programming language) , mathematics , integro differential equation , erlang distribution , risk theory , differential equation , computation , mathematical finance , statistics , mathematical analysis , gamma distribution , first order partial differential equation , computer science , actuarial science , algorithm , economics , financial economics , functional programming , theoretical computer science , business
For actuarial applications we consider the Sparre–Andersen risk model when the interclaim times are Generalized Erlang(n) distributed. Unlike the standard Erlang(n) case, the roots of the generalized Lundberg’s equation with positive real parts can be multiple. This has a significant impact in the formulae for ruin probabilities that have to be found. We start by addressing the problem of solving an integro–differential equation that is satisfied by the survival probability, as well as other probabilities related, and present a method to solve such equation. This is done by considering the roots with positive real parts of the generalized Lundberg’s equation, and then establishing a one-one relation between them and the solutions of the integro–differential equation mentioned above. We first study the cases when all the roots are single and when there are roots with higher multiplicity. Secondly, we show that it is possible to have double roots but no higher multiplicity. Also, we show that the number of double roots depend on the choice of the parameters of the generalized Erlang(n) distribution, with a maximum number depending on n being even or odd. Afterwards, we extend our findings above for the computation of the distribution of the maximum severity of ruin as well as, considering an interest force, to the study the expected discounted future dividends, prior to ruin. Our findings show an alternative and more general method to the one provided by Albrecher et al. (Insur Math Econ 37:324–334, 2005), by considering a general claim amount distribution.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom