Simulation-based models used within general insurance for calculating capital requirements typically have an unsophisticated view of reserve risk over time. Limitations of the widely used ‘emergence factor’ approach include:

• one-year and ultimate results are (nonsensically) correlated 100%
• one-year correlations between tranches necessarily match ultimate correlations
• there is no sensible view of future reserve risk conditional on the one-year result

This session will demonstrate a new approach which resolves these limitations. Given an arbitrary ultimate loss distribution, we construct a martingale stochastic process which begins at the opening reserve, develops randomly over time and terminates at the ultimate loss amount. Deterministic or stochastic emergence patterns may be used.

This framework enables calculation of reserve distributions conditional on future development.

Furthermore, in the simplest form of the model, we can calculate the expected value of future capital requirements at future points in time, enabling a proper calculation of the Solvency II Risk Margin.

The session will be a presentation on the model and its potential applications, with space for questions throughout.

The target audience would be interested in and broadly familiar with general insurance capital modelling methods. Technical details such as proofs and equations will be largely glossed over, but provided separately.

Speakers: Ramsay Nashef, Direct Line Group