Stochastic simulation in life office solvency assessment.

Abstract

In this paper an asset/liability model is used to compare the quality of information available from a set of stochastic simulations with a traditional deterministic sensitivity test approach. The traditional approach applied to a range of variants of the basic model office fails to distinguish adequately very risky strategies from relatively secure strategies. The stochastic simulation method succeeds in ranking the various strategies considered into an intuitively satisfactory order of insolvency risk, as well as giving quantitative information on the relative probabilities of insolvency of different strategies and on the timing of potential solvency problems.