Capital modelling often requires modelling using volatility assumptions that are not derived consistently with the modelled exposure. For example: reserve risk coefficients of variation may be derived based on data as at the previous year-end, but are to be applied to the projected reserves at the next year-end sensitivity testing or multi-year projections often involve changes in exposure, which all else being equal would imply some change to volatility The goal of this talk is to outline some simple rules that capital modellers can use to adjust volatility parameters for changes in volumes in a quantitative, mechanical manner, rather than relying on expert judgment. It covers: a brief outline of the problem and the limitations it implies the analysis performed to investigate a solution, based on publicly available market data, results from recent surveys of market participants, and other sources of information analysis of the results and proposed algorithms or formulae for relating volatility with volume There is some technical content, but the aim is to present the issues in an accessible way that does not require any prior capital modelling expertise. 

Speakers: Neil Gedalla, Adam Smylie, Jade Lagrue

Chair: Matthew Facey