Practical 'modern' Bayesian statistics in actuarial science [Submitted for the Brian Hey Prize]

Document description

The aim of this paper is to convince actuaries that Bayesian statistics could be useful for solving practical problems. This affirmation is due to two main characteristics of Bayesian modelling not yet fully explored by practitioner actuaries: first, the possibility of absorbing subjective information and second, the wider range of models available in the Bayesian framework. In order to help in this convincing process this paper includes an overview of Bayesian statistics in actuarial science and cites many published papers based on this theory with insurance applications. An approach with as few formulae as possible will be used to make it easier to follow for all actuaries, independently of their involvement in statistics.