Hierarchical Compartmental Reserving Models

Bayesian Mixer London
Bayesian Mixer London
Public group
Location image of event venue


Markus Gesmann: Hierarchical Compartmental Reserving Models

Abstract: Hierarchical compartmental reserving models provide a parametric framework for describing the high-level business processes driving claims development in insurance using differential equations.

We will discuss how those models can be presented in a fully Bayesian modelling framework for the aggregated claims settlement process to capture trends observed in paid and outstanding claims development data reflecting the random nature of claims and latent underlying process parameters.

We show how the experienced modeller can utilise her expertise to describe the volatility of the underlying risk exposure profile and uncertainty on prior parameter assumptions and highlight in particular the subtle, but important difference between modelling incremental and cumulative claims payments.

The talk is based on joint work with Jake Morris.

Bio: Markus is an analyst/data scientist with over 15 years’ experience in the London Insurance and Capital Markets. He is the maintainer of the ChainLadder and googleVis R packages. Markus is the co-founder of the Insurance Data Science conference series and the Bayesian Mixer Meetups in London.

Website: https://magesblog.com
Twitter: @MarkusGesmann


6:30 - 7:00pm: Arrival & welcome to Lockton Re
7:00 pm: Presentation