Financial Engineering Workshop: Colin Turfus (Deutsche Bank)

Scenario Generation for Risk Management

Colin Turfus, Deutsche Bank

Risk management increasingly requires consideration of exposures (to counterparty or market risk) under multiple future scenarios. Various strategies can be adopted. Credit and rates markets raise non-trivial challenges around the evolution of curves (level and shape). While low-dimensional pricing models may give accurate fair value as the first moment of a distribution, exposure calculations require realistic representation of the overall distribution of prices (quantiles), not just its mean. So the use of higher-dimensional rates and credit models is desirable. We show how recently published analytic pricing kernels for one-factor Hull-White and Black-Karasinski models can be extended to an arbitrary number of dimensions to allow multi-dimensional rates and credit curves to be evolved simultaneously.”