Journal of Credit Risk

With this issue of The Journal of Credit Risk, we enter the journal's tenth year. Over this time the journal has flourished in terms of both readership and quality of papers, while the world has seen a credit crisis and a broader economic recession. Structured credit products and credit underwriting of nonprime mortgages have been at the center of it all. Lessons learned from the credit crisis have been discussed in various papers that we have published in the last three years, and it is clear that the importance of credit risk has only increased post-crisis. With the advent of the Comprehensive Capital Analysis and Review (CCAR), the stress testing of credit portfolios has taken center stage in many financial institutions. An issue of paramount importance is how to reconcile CCAR results with economic capital results. For both CCAR and economic capital, credit risk is by far the most important risk stripe.

In this issue we present two research papers and two technical reports.

The first research paper in this issue is "Dynamic affordability assessment: predicting an applicant's ability to repay over the life of the loan" by Katarzyna Bijak, Lyn C. Thomas and Christophe Mues. In this paper the authors outline a method of identifying and measuring an applicant's ability to pay (ie, an affordability assessment measure). Insofar as a borrower's probability of default is dependent on their existing debt, assets, income and expenditures, an affordability measure should be conditioned on the interaction of those factors over time and the impact the interaction has on the borrower's ability to carry the amount of new debt offered/requested. The authors develop a conceptual framework that incorporates information known about the borrower at time of application (ie, borrower characteristics, cohort effects) with key assumptions about income and consumption/savings growth and future payment behavior. Under the dynamic modeling approach, income and expenditures (consumption) are allowed to vary over the life of the loan. Because that information is not known, they outline a conceptual approach for estimating future income and consumption using methods developed in the economics literature, and then simulate the impact of additional amounts of debt on the likelihood of default conditioned on repeated random draws of the variation in individual effects and idiosyncratic components (under a random effects model design). The simulation is performed over "all possible" or "feasible" installment amounts and it is therefore able to identify the loan amount under which the borrower's likelihood of default is less than some preset limit.

This issue's second research paper, "Modeling the credit contagion channel and its consequences via the standard portfolio credit risk model" by Yongwoong Lee and Ser-Huang Poon, divides its sample into four data sets based on size and implied credit ratings and investigates default dependence between the four groups, using a single-factor modified Merton-KMV model. The authors use actual default events for all the listed firms in thirty economies over the period from the first quarter of 2000 to the second quarter of 2011. Utilizing a technique of particle filtering and smoothing with Monte Carlo Markov chain, the authors find strong evidence that defaults of small high-yield firms infect large high-yield firms, and this in turn generates a contagious feedback effect back onto small high-yield firms. All investment-grade firms are unaffected by this channel of credit contagion.

A technical report describes a particular practical technique and enumerates situations in which it works well and others in which it does not. Such reports provide extremely useful information to practitioners in terms of saved time and minimizing duplication of effort. The contents of technical reports complement rigorous conceptual and model developments presented in the research papers and provide a lot of value to practitioners.

Our first technical report is "Usage and exposures at default of corporate credit lines: an empirical study" by Janet Yinqing Zhao, Douglas W. Dwyer and Jing Zhang. This paper empirically examines the use of lines of credit around default. Using Moody's Credit Research Database, the study covers information on the credit line usage of private firms in the United States during the period 2000-2010. This data is augmented with manually collected lines of credit information of public firms during this period. The study finds that defaulted borrowers draw more on lines of credit than nondefaulted borrowers. The study also finds that bank monitoring reduces the draw-down activity of high-default risk borrowers. The results suggest that credit line usage is a function of both borrowers' characteristics and banks' monitoring and control of these lines.

The second technical report in this issue is "Backtesting counterparty risk: how good is your model?" by Ignacio Ruiz. Starting from a description of the Basel backtesting framework for market risk value-at-risk, the author develops a backtesting methodology for counterparty credit risk that exhibits a red-yellow-green classification to reflect model quality. The paper focuses on single-risk-factor evolution models, which are the most critical part of counterparty credit risk models, but it also shows how to extend the proposed backtesting methodology to incorporate dependency structures between risk factors. The item to be backtested is the whole distribution, which for expected positive exposure models is more appropriate than quantile-based backtesting. Using different metrics and weights the method allows for more emphasis on certain parts of the distribution. The paper does not elaborate on the backtesting of representative counterparty portfolios or the relationship between different tenors of the same risk factor.

Ashish Dev
JPMorgan Chase, New York