B0513
Title: Prediction outstanding insurance claims via in-sample forecasting
Authors: Munir Hiabu - Cass Business School (United Kingdom) [presenting]
Enno Mammen - Heidelberg University (Germany)
M. Dolores Martinez-Miranda - Universidad de Granada (Spain)
Jens Perch Nielsen - City, University of London (United Kingdom)
Abstract: Non-life insurance companies traditionally use the so called chain ladder method to reserve for outstanding liabilities. We will show how to translate the chain ladder method into a continuous framework while keeping the basic structure. The problem will hereby be translated into a survival analysis setting. As it turns out, chain ladder, and thus also our continuous analogue, is a in-sample technique where no extrapolation is needed to forecast the reserve. The in-sample area is defined as one triangle and the forecasting area as the second triangle that added to the first triangle produces a square. We call our approach in-sample forecasting. It is defined as forecasting a structured density to sets where the density is not observed. The in-sample forecasting will be done with nonparametric kernel methods. Firstly we will focus on the multiplicative density structure which also is the underlying assumption of chain ladder. Secondly, we will show how to go beyond this multiplicativity assumption in order to get more accurate forecasts.
