A Flexible Framework for Stochastic Reserving Models

Maximum likelihood estimators provide a powerful statistical tool. In this paper we directly deal with non-linear reserving models, without the need to transform those models to make them tractable for linear or generalized linear methods. We also show how the same general approach can be easily adapted to provide estimates for a very wide range of reserving methods and models, making use of the same framework, and even much of the same computer code. We focus on the triangle of incremental average costs, and show how five common methods can be set in a stochastic framework.