Journal of Risk Model Validation

This issue of The Journal of Risk Model Validation comprises four papers: the first two deal with practitioner credit risk models in an intensive fashion, while the remaining two could be deemed to be theoretical reviews of topics that are central to risk model validation. The issue's first paper is "Comparative analysis of credit risk models for loan portfolios" by Chulwoo Han. The paper deals with a range of models and conducts both theoretical and simulation analysis on them. Practitioners should find this of value as the models considered are those available to, and widely used by, industry. The author also provides some interesting insights into the applicability of particular models and the relative importance of different risk components.

Our second paper, "An analytic approach to quantify the sensitivity of CreditRisk+ with respect to its underlying assumptions" by Matthias Fischer and Florian Kaufmann, deals with extensions to the well-known CreditRisk+ model in a framework that the authors call the analytic Bernoulli CreditRisk+, or just ABCreditRisk+, model, which allows them to calculate the sensitivity of the basic CreditRisk+ model within an analytical framework and can also be used for risk calculation in its own right. In statistical parlance, this encompasses or nests the CreditRisk+ model and allows model users to gauge the sensitivity of the model to its rather restrictive distributional assumptions. This gives a very good framework for sensitivity analysis, which is a central aspect of model validation.

The third paper in the issue is "Review, theory and implementation of convertible bonds for commercial investment" by Raj Kunwar, Zhihui Yang, Jonathan Lai and Jerrold Cline. Quoting from the abstract:

"This technical paper provides a comprehensive literature review of the convertible bond model, as well as discussing the theory behind it and its validation. We analyze several different deal terms of a chosen convertible bond in order to value the embedded optionality of the convertible bond and systematically carry out the model validation of a convertible bond model for commercial investment purposes."

It is hoped that readers will find such a detailed and comprehensive paper a valuable resource when researching this particular area.

In a similar way, our final paper, "Backtesting value-at-risk tail losses on a dynamic portfolio" by Alasdair Graham and János Pál, gives a comprehensive overview of backtesting, with a special focus on tail risk. The authors present a number of technically challenging procedures, such as the saddle-point approximation, but before we reach for our abaci, I should tell readers that this paper is very well-written and will hopefully be accessible to most of the journal's readership.

As always, I welcome all feedback and particularly feedback regarding whether papers that investigate a topic in depth are of value to readers. In my own experience, as both a user and provider of papers, I have found such papers to be very helpful to my research.

Steve Satchell
Trinity College, University of Cambridge