Journal of Risk Model Validation

Steve Satchell

University of Cambridge

It is challenging to produce an issue of The Journal of Risk Model Validation in the summer of 2008 that is not going to be dominated by mortgage default issues and other subprime phenomena. It is fair to say that this issue is likewise affected; however, I am pleased to report that a good deal of the material is focused more on Basel II and on long-run, rather than contemporary issues.

Our first paper, “Estimating and validating long-run probability of default with respect to Basel II requirements”, byMiu and Ozdemir, investigates methodologies for estimating and validating long-run probabilities of default in the Basel II context. Noteworthy in this paper is a proposed methodology that allows one to augment internal data with external data. All those who work in the area are aware that internal data deficiencies are the biggest single barrier to effective implementation of any validation procedure.

Our second paper is “Stress-testing retail loan portfolios with dual-time dynamics”, by Breeden et al. This applies an innovative procedure for stress testing to US mortgage markets. The procedure mixes known functions for loan characteristics with extreme macroeconomic events. In the UK, such work was always something sought after by risk managers. However, there was never enough macroeconomic variation to get anything but the nine results. The silver lining to the current cloud is that we are now observing much more macroeconomic variation.

Our third paper is “Modeling and evaluating the credit risk of mortgage loans: a primer”, by Van Order. This paper looks at the optionality in mortgage default and the issues that arise in validating such models. The author identifies a number of particular concerns: moral hazard, the possibility of reverse engineering and the problems that arise from the underlying structure of the model. It is one of the great intellectual marriages, wedding option pricing to the mortgage default decision. This is an area I have always found very stimulating.

Our final paper is “An assessment of the internal rating-based approach in Basel II”, by Varotto. This paper looks at the internal rating-based approach to measuring credit risk in bank portfolios under Basel II. I welcome this paper as there is very little published on the topic and there is now a new industry in having to validate such models. The author looks at some of the assumptions of the internalrating- based approach and shows that the results are very sensitive to them. Whilst I do not find the conclusions of this paper mathematically surprising, I believe that they are of considerable importance from a validation perspective.