Chapter 15: Further Reading — Credit Risk Modelling and Model Risk Management

Regulatory Primary Sources

These documents carry the force of law or supervisory expectation in their respective jurisdictions. Practitioners working in credit risk modelling or model risk management should be familiar with the relevant sources for their jurisdiction.

Essential Books

Credit Risk Scorecards

Naeem Siddiqi, Credit Risk Scorecards: Developing and Implementing Intelligent Credit Scoring (2nd ed.) John Wiley & Sons, 2012

The definitive practitioner reference for credit scorecard development. Siddiqi covers the full development lifecycle from data preparation through WoE transformation, logistic regression, scorecard scaling, and validation. The book is written for practitioners, not academics: every technique is motivated by a real modelling problem. Essential reading before building any retail or SME scorecard. The WoE and Information Value treatment in this chapter draws directly on Siddiqi's exposition.

Naeem Siddiqi, Intelligent Credit Scoring: Building and Implementing Better Credit Risk Scorecards (3rd ed.) John Wiley & Sons, 2017

A more recent successor that extends the treatment to include machine learning considerations, IFRS 9 implications, and the evolving regulatory landscape. Complements the 2nd edition rather than replacing it.

Logistic Regression and Statistical Foundations

David W. Hosmer Jr., Stanley Lemeshow, and Rodney X. Sturdivant, Applied Logistic Regression (3rd ed.) John Wiley & Sons, 2013

The authoritative statistical reference for logistic regression. Chapter 5 (model building strategies) and Chapter 8 (assessment of fit) are most directly relevant to credit scorecard practitioners. The Hosmer-Lemeshow goodness-of-fit test — used for calibration assessment in credit models — is explained in full here. More statistical than Siddiqi, but essential for understanding the foundations of what logistic regression scorecards are actually doing.

Credit Risk — Wholesale and Enterprise-Level

Darrell Duffie and Kenneth J. Singleton, Credit Risk: Pricing, Measurement, and Management Princeton University Press, 2003

A graduate-level treatment of credit risk pricing and measurement, covering structural models (Merton model, KMV), reduced-form intensity models, and credit derivatives. Essential background for practitioners working on wholesale credit risk or credit portfolio models. More mathematically demanding than Siddiqi, but provides the theoretical grounding for understanding why PD, LGD, and EAD are the key parameters.

Michael Bluhm, Ludger Overbeck, and Christoph Wagner, Introduction to Credit Risk Modeling (2nd ed.) Chapman & Hall/CRC Financial Mathematics Series, 2010

A comprehensive reference covering credit portfolio models, correlation, and the mathematics of the IRB formula. Chapter 3 on the Vasicek model and its relationship to the Basel IRB formula is particularly valuable for understanding why the regulatory capital formula takes the form it does.

IFRS 9 and Expected Credit Loss

KPMG, IFRS 9 — Expected Credit Losses KPMG International, updated annually

Not a book but a freely available guidance document from KPMG covering the IFRS 9 ECL requirements in practical detail, with worked examples and jurisdictional commentary. The most accessible starting point for IFRS 9 ECL methodology. Available at kpmg.com.

Moody's Analytics, IFRS 9 Impairment Implementation Guide Moody's Analytics, 2018

A practitioner guide to IFRS 9 model architecture, PD term structure estimation, and macroeconomic scenario integration. Reflects Moody's practical experience implementing IFRS 9 models at large banks. Available from Moody's Analytics website (registration required).

Model Risk Management

Eduardo Epperlein, "Model Risk: Illuminating the Black Boxes" Published in Risk Magazine, various editions 2010–2015

A series of practitioner articles on model risk that preceded and informed the SR 11-7 framework. Accessible through Risk.net (subscription required).

Office of the Comptroller of the Currency (OCC), Model Risk Management Handbook OCC, 2021

The OCC's companion handbook to SR 11-7, providing supplementary guidance on model risk management in banks. More prescriptive than SR 11-7 in several areas; includes case examples and self-assessment checklists. Available free at occ.gov.

Machine Learning in Credit Risk

Christoph Molnar, Interpretable Machine Learning: A Guide for Making Black Box Models Explainable Lulu.com, 2019 (also freely available at christophm.github.io/interpretable-ml-book/)

The standard practitioner reference for SHAP, LIME, partial dependence plots, and other interpretability techniques. Directly applicable to the challenge of explaining ML credit models to regulators, auditors, and customers. The SHAP chapter is the most immediately useful for credit risk practitioners.

Moritz Baur, Corinna Ewelt-Knauer, Axel Kind, and Peter Löw, "Machine Learning Applications in Credit Risk" Journal of Risk and Financial Management, 2021 (open access)

A well-structured academic survey of ML applications in credit risk modelling, covering random forest, gradient boosting, and neural network approaches, with performance benchmarks against traditional logistic regression. Provides a balanced view of where ML adds value and where it does not.

Technology References

Vendor Platforms for Credit Risk Modelling

Open-Source Python Libraries for Credit Risk

Academic Papers — Key References

Thomas E. Keenan and Jorge A. Sobehart, "Performance Measures for Credit Risk Models" Moody's Risk Management Services, 1999

One of the foundational papers establishing the Gini coefficient and Cumulative Accuracy Profile (CAP) as credit model performance metrics. Historical but essential for understanding the theoretical basis of the discrimination metrics covered in this chapter.

Basel Committee on Banking Supervision, "Studies on Validation of Internal Rating Systems" (Working Paper No. 14) BIS, 2005

A detailed treatment of IRB model validation from the Basel Committee's perspective: backtesting, benchmarking, stability testing. Includes a comprehensive discussion of statistical tests for PD validation. Available free at bis.org.

Tobias Berg, "Got Rejected? Real Effects of Not Getting a Loan" Review of Financial Studies, 2018

An empirical study using German bank data examining the real-economy effects of credit scoring decisions, including the role of model overrides. Relevant to understanding what is at stake when credit models produce wrong outputs — the consequences extend to borrowers and the economy, not just bank capital.

Andreas Behr and Frank Güttler, "The Information Content of Unsolicited Ratings" Journal of Banking & Finance, 2008

Relevant to the wholesale credit model discussion: examines how much information agency ratings add beyond observable financials, with implications for shadow-rating model design.

Further Online Resources

Bank for International Settlements (BIS) — Credit Risk Papers bis.org/list/bcbspapers/index.htm

The complete catalogue of Basel Committee working papers, including historical research on IRB model validation, credit risk data requirements, and through-the-cycle versus point-in-time PD estimation. All freely available.

European Banking Authority — Publications eba.europa.eu/regulation-and-policy/credit-risk

All EBA guidelines, discussion papers, and technical standards on credit risk, including regular updates to IRB requirements and IFRS 9 implementation guidance.

UK Finance — Credit Risk and Modelling Briefings ukfinance.org.uk

Industry body publications on credit risk practice in the UK, including guidance on IFRS 9 implementation, model governance, and engagement with the PRA on IRB.

IFRS Foundation ifrs.org/issued-standards/list-of-standards/ifrs-9-financial-instruments/

Access to IFRS 9 and accompanying materials. The Educational Material on Expected Credit Losses (2021) provides accessible worked examples of Stage 1/2/3 classification and ECL calculation.

Practitioner Certifications Relevant to This Chapter

Suggested Reading Sequence

For a practitioner new to credit risk modelling:

1. Start with Siddiqi (2012) chapters 1–8 for scorecard methodology fundamentals
2. Read SR 11-7 in full — it is only 20 pages and is the most important single document for model governance
3. Work through the EBA/GL/2017/16 overview sections for IRB data requirements
4. Read IFRS 9 Section 5.5 with the KPMG guidance document alongside it
5. Explore Molnar's Interpretable Machine Learning (free online) for the ML context
6. Return to Hosmer, Lemeshow & Sturdivant for statistical foundations when needed

For a practitioner deepening existing knowledge:

1. Basel 3.1 Final Standard for the output floor and IRB scope restrictions
2. ECB TRIM Guide for the most granular regulatory expectations on IRB models
3. EBA/DP/2021/04 on ML in IRB models for the frontier regulatory perspective
4. Duffie and Singleton for wholesale credit risk theoretical foundations

Chapter 15 of Regulatory Technology (RegTech): A Practitioner's Guide