IFRS 9 and CECL Credit Risk Modelling and Validation covers a hot topic in risk management. Both IFRS 9 and CECL accounting standards require Banks to adopt a new perspective in assessing Expected Credit Losses. The book explores a wide range of models and corresponding validation procedures. The most traditional regression analyses pave the way to more innovative methods like machine learning, survival analysis, and competing risk modelling. Special attention is then devoted to scarce data and low default portfolios. A practical approach inspires the learning journey. In each section the theoretical dissertation is accompanied by Examples and Case Studies worked in R and SAS, the most widely used software packages used by practitioners in Credit Risk Management.
Key Features
- Offers a broad survey that explains which models work best for mortgage, small business, cards, commercial real estate, commercial loans and other credit products
- Concentrates on specific aspects of the modelling process by focusing on lifetime estimates
- Provides an hands-on approach to enable readers to perform model development, validation and audit of credit risk models
1. Introduction to Expected Credit Loss Modelling and Validation1.1 Introduction 1.2 IFRS 9 1.21 Staging Allocation 1.22 ECL Ingredients1.23 Scenario Analysis and ECL1.3 CECL 1.31 Loss-Rate Methods 1.32 Vintage Methods1.33 Discounted Cash Flow Methods1.34 Probability of Default Method (PD, LGD, EAD)1.35 IFRS 9 vs CECL 1.4 ECL and Capital Requirements1.41 Internal Rating-Based Credit Risk-Weighted Assets1.42 How ECL Affects Regulatory Capital and Ratios1.5 Book Structure at a Glance1.6 Summary
2. One-Year PDs2.1 Introduction 2.2 Default Definition and Data Preparation 2.21 Default Definition 2.22 Data Preparation 2.3 Generalized Linear Models (GLMs) 2.31 GLM (Scorecard) Development2.32 GLM Calibration2.33 GLM Validation2.4 Machine Learning (ML) Modelling2.41 Classification and Regression Trees (CART)2.42 Bagging, Random Forest, and Boosting2.43 ML Model Calibration2.44 ML Model Validation2.5 Low Default Portfolio, Market-Based, and Scarce Data Modelling2.51 Low Default Portfolio Modelling2.52 Market Based Modelling2.53 Scarce Data Modelling2.54 Hints on Low Default Portfolio, Market-Based, and Scarce Data Model Validation 2.6 SAS Laboratory 2.7 Summary 2.8 Appendix A From Linear Regression to GLMs2.9 Appendix B Discriminatory Power Assessment
3. Lifetime PDs 13.1 Introduction3.2 Data Preparation 3.21 Default Flag Creation 3.22 Account-Level (Panel) Database Structure3.3 Lifetime GLM Framework3.31 Portfolio-level GLM Analysis3.32 Account-Level GLM Analysis3.33 Lifetime GLM Validation3.4 Survival Modelling 3.41 Kaplan Meier (KM) Survival Analysis3.42 Cox Proportional Hazard (CPH) Survival Analysis3.43 Accelerated Failure Time (AFT) Survival Analysis3.44 Survival Model Validation3.5 Lifetime Machine Learning (ML) Modelling3.51 Bagging, Random Forest, and Boosting Lifetime PD3.52 Random Survival Forest Lifetime PD3.53 Lifetime ML Validation3.6 Transition Matrix Modelling3.61 Na_ve Markov Chain Modelling3.62 Merton-Like Transition Modelling3.63 Multi State Modelling3.64 Transition Matrix Model Validation3.7 SAS Laboratory 3.8 Summary
4. LGD Modelling4.1 Introduction4.2 LGD Data Preparation4.21 LGD Data Conceptual Characteristics 4.22 LGD Database Elements4.3 LGD Micro-Structure Approach4.31 Probability of Cure 4.32 Severity4.33 Defaulted Asset LGD4.34 Forward-Looking Micro-Structure LGD Modelling4.35 Micro-Structure Real Estate LGD Modelling4.36 Micro-Structure LGD Validation4.4 LGD Regression Methods441 Tobit Regression4.42 Beta Regression4.43 Mixture Models and forward-looking Regression4.44 Regression LGD Validation4.5 LGD Machine Learning (ML) Modelling4.51 Regression Tree LGD4.52 Bagging, Random Forest, and Boosting LGD4.53 Forward-Looking Machine Learning LGD4.54 Machine Learning LGD Validation4.6 Hints on LGD Survival Analysis4.7 Scarce Data and Low Default Portfolio LGD Modelling4.71 Expert Judgement LGD Process4.72 Low Default Portfolio LGD4.73 Hints on How to Validate Scarce Data and Low Default Portfolio LGDs4.8 SAS Laboratory4.9 Summary
5. Prepayments, Competing Risks and EAD Modelling5.1 Introduction5.2 Data Preparation5.21 How to Organize Data5.3 Full Prepayment Modelling5.31 Full Prepayment via GLMs5.32 Machine Learning (ML) Full Prepayment Modelling5.33 Hints on Survival Analysis5.34 Full Prepayment Model Validation5.4 Competing Risk Modelling5.41 Multinomial Regression Competing Risks Modelling5.42 Full Evaluation Procedure5.43 Competing Risk Model Validation5.5 EAD Modelling5.51 A Competing-Risk-Like EAD Framework5.52 Hints on EAD Estimation via Machine Learning (ML)5.53 EAD Model Validation5.6 SAS Laboratory 5.7 Summary
6. Scenario Analysis and Expected Credit Losses6.1 Introduction6.2 Scenario Analysis6.21 Vector Auto-Regression (VAR) and Vector Error-Correction (VEC) Modelling6.22 VAR and VEC Forecast6.23 Hints on GVAR Modelling6.3 ECL Computation in Practice 6.31 Scenario Design and Satellite Models6.32 Lifetime ECL6.33 IFRS 9 Staging Allocation6.4 ECL Validation6.41 Historical and Forward-Looking Validation6.42 Credit Portfolio Modelling and ECL Estimation6.5 SAS Laboratory6.6 Summary
- Christoffersen, Elements of Financial Risk Management, 2e, 9780123744487, 2011, 344pp, $69.95
- Gilli, Numerical Methods and Optimization in Finance, 9780123756626, 2011, 600pp, $105.00
- Stowell, Investment Banks, Hedge Funds, and Private Equity, 2e, 2012, 9780124158207, 672pp, $79.95
Upper-division undergraduates, graduate students, and professionals working in economic modelling and statistics.
