Journal of Risk Model Validation

This issue of The Journal of Risk Model Validation brings us four interesting papers, which I discuss below. I will resist the temptation to contextualize Brexit and risk model validation; much has been said already, much more will be said. Very loosely, two of our papers fall into the high-tech bucket while the other two fall into the practical guidance bucket (although one might be described as a hybrid); hopefully this leaves the journal offering something for everyone.

Our first paper, "Rating-transition-probability models and Comprehensive Capital Analysis and Review stress testing: methodologies and implementation" by Bill Huajian Yang and Zunwei Du, provides a new approach to the construction and validation of transition probability models. The paper explicitly demonstrates how Miu and Ozdemir's original methodology on transition probability models can be structured and implemented with rating-specific asset correlation. A key innovation is to include a conditioning variable called the credit index. There are also applications based on a commercial portfolio illustrating estimation and scenario analyses.

The second paper, by Tae Yeon Kwon, is titled "A correlated structural credit risk model with random coefficients and its Bayesian estimation using stock and credit market information". This paper illustrates a validation of a structural correlated default model applied to Black-Cox setups. While the dependence structure is modeled through the imposition of common factors on the asset process, instead of the assumption of homogeneity in the effects of common factors across the firms, a random coefficient representing the heterogeneity effect is considered. The approach taken is Bayesian; this finds fewapplications among practitioners, principally because of the difficulty of specifying priors that capture notions of skill or expertise in forecasting or risk management. Model parameters are estimated using not only equity prices but also credit default swap spreads. This led to improvements based on simulation relative to just using one of them. In order to demonstrate potential practical applications and an out-of-sample model validation check, the posterior distribution of CDX tranche prices is derived.

"Value-at-risk bounds for multivariate heavy tailed distribution: an application to the Glosten-Jagannathan-Runkle generalized autoregressive conditional heteroscedasticity model" by Imed Gammoudi, Mohamed El Ghourabi and Lotfi BelKacem is our third paper. It derives value-at-risk (VaR) bounds for the portfolios of possibly dependent financial assets for heavy tailed Glosten-Jagannathan-Runkle generalized autoregressive conditional heteroscedasticity processes using extreme value theory copulas. Based on earlier work by Mesfioui and Quess and by Gammoudi et al, the authors show that these bounds have the interesting property of location invariance. As with all good theory papers, empirical studies for several market indexes are carried out to illustrate the authors' approach.

The final paper in the issue, "Some options for evaluating significant deterioration under IFRS 9" by Gaurav Chawla, Lawrence R. Forest Jr. and Scott D. Aguais, addresses some issues to do with IFRS 9 (which, according to my rather limited understanding, is in the process of replacing IAS39). The point they focus on is, if I may quote them:

According to International Financial Reporting Standard 9 (IFRS 9), if the credit risk on an instrument has increased "significantly" since the instrument's original recognition, and the resulting credit risk is more than "low credit risk", then the institution would recognize a loss allowance on the instrument in the amount of lifetime expected credit losses (ECLs). Alternatively, at original recognition, and thereafter in the absence of significant deterioration in credit risk, the institution would recognize an allowance in the amount of twelve months of ECLs (or a lifetime, if the instrument matures in less than twelve months). In clarifying this aspect of IFRS 9, the International Accounting Standards Board (IASB) has specified that, in evaluating whether an instrument has suffered significant deterioration, an institution should consider only lifetime default risk, excluding consideration of possible changes in the exposure at default (EAD) and loss given default (LGD) components of ECLs. The authorities have also stated that the triggering of a lifetime allowance would reflect circumstances under which the spread inherent in contractual pricing no longer fully compensates for credit risk. Further, in its analysis of IFRS 9, the Bank for International Settlements has presented a view that any deterioration in credit risk should be considered significant.

Since many of the strictures in regulatory space are rather low on implementation detail, the authors go to some length to explain how to determine if an instrument has suffered serious deterioration in credit risk.

I will end on this note. In a recent editorial I mentioned reader feedback that had asked for guidance on responding to regulatory changes: it is a pleasure to be able to provide such guidance so soon.

Steve Satchell
Trinity College, University of Cambridge

Value-at-risk bounds for multivariate heavy tailed distribution: an application to the Glosten–Jagannathan–Runkle generalized autoregressive conditional heteroscedasticity model

This paper aims to derive VaR bounds for the portfolios of possibly dependent financial assets for heavy tailed Glosten–Jagannathan–Runkle generalized autoregressive conditional heteroscedasticity processes using extreme value theory copulas.

Read Now Download PDF

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here