Counterparty Credit Risk and Credit Value Adjustment: A Continuing Challenge for Global Financial Markets, 2nd Edition

9.4 Models for Credit Exposure

In this section, we provide details on models used for exposure simulation considering some basic points on the modelling in different asset classes (see also Appendix 9A for more mathematical detail). We consider the calibration issues, the balance between complex and simple models, and give some specific examples for various asset classes. We will provide only an overview and the reader is directed to Cesari et al. (2009) for a more mathematical treatment of the modelling of credit exposure.

A key point to bear in mind is that, sometimes, simple and parsimonious approaches with fewer parameters can justifiably be favoured over ones that are more complex. Modelling many risk factors to define a curve evolution is unnecessary and may lead to excessive complexity and unrealistic future scenarios. This is especially true given the complex co-dependencies that may need to be modelled and the fact that problems such as wrong-way risk (Chapter 15) are still to be addressed. Whilst the total number of risk factors needing to be modelled is likely to be in the hundreds (depending on asset class coverage), it is important to keep this number to a minimum.

Whilst more advanced models for exposure are important, the possibility of false accuracy must also be considered. If there are other components of CVA with a large degree of uncertainty then a more sophisticated modelling of exposure may be futile. An obvious example of such a component would be the assessment ...