The direction of credit research has changed significantly during the past decade — from a classic accounting and finance framework to an emphasis on quantitative analysis of default and loss probabilities. Although the old ways of firm-specific analysis remain helpful, they are no longer foundational for successful measurement, integration, and aggregation of pooled credit risks in a portfolio. The growth in structured products and derivatives requires greater understanding of the mechanics of pricing credit risk. Whether through the bundled credit risks found within mortgage pools and asset-backed securities or through analyzing credit default swaps and credit products, measuring conditional probabilities based on scoring and segmentation is essential to credit research. Basel III and other regulatory constructs centered on risk categorizations make quantitative credit risk measurement an even more critical part of the financial landscape.
The educational tools available to understand credit risk measurement have not kept pace with structural advancements. A knowledge gap exists between the traditional credit analyst and the credit quant who uses advanced econometrics to measure and price risk. For professionals who want to learn the pertinent quantitative techniques, the various strains of econometrics associated with credit risk analysis are not often available in one book.
Princeton University Press’s paperback reissue of The Econometrics of Individual Risk helps alleviate this gap, especially as we enter the late stage of the current credit cycle. Originally published in 2007, this book may be even more pertinent today, given the intervening advances in credit structuring dynamics, the continued growth of consumer lending, and the absence of any competing short text that explores these issues together.
Pricing risk based on measurable characteristics or covariates is fundamental to all banking, mortgage, and consumer loan scoring, as well as to the bundling of any credit portfolio. The authors explore credit risk from four distinct quantitative perspectives, each requiring its own econometric tool: the occurrence, frequency, timing, and severity of a loss. The Econometrics of Individual Risk focuses on the core econometric techniques for measuring each aspect.
The four risks can be quantified and priced by examining a credit pool’s measured statistics. Relying on scores and statistics without knowing how they were generated, however, may have contributed to the gross risk management failures during the 2008 financial crisis. Would key senior executives or investors have taken on the same mortgage exposures if they were truly familiar with how risks were measured for subprime pools?
The sophistication necessary to apply these techniques may be beyond the statistics expertise of the typical CFA charterholder, but a working knowledge of quantitative credit analysis has become critical for success in the credit asset class. Evaluating quantitative assessments of credit risk may be too important to be left to the quants. Only through familiarity with these econometric intricacies can analysts appreciate both the value and limits of credit scoring and pricing.
Christian Gourieroux of the University of Toronto and Joann Jasiak of York University, Toronto, have written a cogent monograph on the econometrics of risk that presents the theory along with examples of its practical application. It is an econometrics book and not a how-to manual on credit scoring, however, so the important foundational concepts are handled succinctly, allowing the authors to focus on practicality. Each chapter presents a different credit problem along with descriptions of how to use the relevant econometric methods. The examples would benefit, however, from additional detail on exactly how a credit analyst can apply the models. Although theory and foundations are important, practitioners must understand how to use these techniques. Using tools without experience is always dangerous.
The econometric techniques the authors describe focus on scoring, or measures of risk, based on defined characteristics and segmentation under uncertainty. With pools of credits, there is no certainty regarding behavior. Statistics on conditional probabilities derived from large datasets can, however, provide insights into an event’s likelihood.
Gourieroux and Jasiak cover three key problems and the associated quantitative credit techniques. First, the workhorse model for finding the characteristics of, or covariates with, defaults is a linear discriminant or logit model that measures a variable’s effect on the likelihood of failure. This question is binary (i.e., whether a default occurs). The second major topic explores count methods of predicting events occurring during a specific period. For instance, how many accidents will occur during the next year? The third topic is timing or hazard models, often used in pricing corporate bonds to determine when a default will occur. For example, how long will it take before a mortgage goes bad?
In the final chapters, the authors explain the econometrics of more difficult problems, such as the endogeneity of variables; transition models that look, for example, at the probability of being in a particular credit-rating state; and the problem of multiple scoring of risks. The authors conclude with a review of value at risk modeling, looking at how it relates to the previously discussed econometric techniques.
It may be a bold assertion, but some of the failures of subprime lending, asset securitization, collateralized loan obligations, and credit derivative swaps might have been mitigated had more analysts understood the mechanics of the quantitative risk tools presented in this book. This knowledge is all the more relevant considering that the current risk-weighting systems used by banks and insurance companies are founded on scoring principles. For financial professionals engaged in credit work, this book provides a good reference for basic understanding of quantitative credit techniques. Although some background in econometrics is required, the authors of this short monograph ensure that readers will become better informed about the key quantitative credit risk concepts and measurement tools.
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.