Guidelines on Credit Risk Management
Rating Models and Validation
I INTRODUCTION
7II ESTIMATING AND VALIDATING PROBABILITY
OF DEFAULT (PD)
81 Defining Segments for Credit Assessment
82 Best-Practice Data Requirements for Credit Assessment
112.1 Governments and the Public Sector 12
2.2 Financial Service Providers 15
2.3 Corporate Customers — Enterprises/Business Owners 17
2.4 Corporate Customers — Specialized Lending 22
2.4.1 Project Finance
242.4.2 Object Finance
252.4.3 Commodities Finance
262.4.4 Income-Producing Real Estate Financing
262.5 Retail Customers 28
2.5.1 Mass-Market Banking
282.5.2 Private Banking
313 Commonly Used Credit Assessment Models
323.1 Heuristic Models 33
3.1.1 Classic Rating Questionnaires
333.1.2 Qualitative Systems
343.1.3 Expert Systems
363.1.4 Fuzzy Logic Systems
383.2 Statistical Models 40
3.2.1 Multivariate Discriminant Analysis
413.2.2 Regression Models
433.2.3 Artificial Neural Networks
453.3 Causal Models 48
3.3.1 Option Pricing Models
483.3.2 Cash Flow (Simulation) Models
493.4 Hybrid Forms 50
3.4.1 Horizontal Linking of Model Types
513.4.2 Vertical Linking of Model Types Using Overrides
523.4.3 Upstream Inclusion of Heuristic Knock-Out Criteria
534 Assessing the Models Suitability for Various Rating
Segments
544.1 Fulfillment of Essential Requirements 54
4.1.1 PD as Target Value
544.1.2 Completeness
554.1.3 Objectivity
554.1.4 Acceptance
564.1.5 Consistency
574.2 Suitability of Individual Model Types 57
4.2.1 Heuristic Models
574.2.2 Statistical Models
584.2.3 Causal Models
605 Developing a Rating Model
605.1 Generating the Data Set 62
5.1.1 Data Requirements and Sources
625.1.2 Data Collection and Cleansing
645.1.3 Definition of the Sample
725.2 Developing the Scoring Function 82
5.2.1 Univariate Analyses
755.2.2 Multivariate Analysis
805.2.3 Overall Scoring Function
825.3 Calibrating the Rating Model 84
5.3.1 Calibration for Logistic Regression
855.3.2 Calibration in Standard Cases
865.4 Transition Matrices 88
5.4.1 The One-Year Transition Matrix
885.4.2 Multi-Year Transition Matrices
916 Validating Rating Models
946.1 Qualitative Validation 96
6.2 Quantitative Validation 98
6.2.1 Discriminatory Power
986.2.2 Back-Testing the Calibration
1156.2.3 Back-Testing Transition Matrices
1326.2.4 Stability
1346.3 Benchmarking 128
6.4 Stress Tests 130
6.4.1 Definition and Necessity of Stress Tests
1306.4.2 Essential Factors in Stress Tests
1316.4.3 Developing Stress Tests
1336.4.4 Performing and Evaluating Stress Tests
137III ESTIMATING AND VALIDATING LGD/EAD
AS RISK COMPONENTS
1397 Estimating Loss Given Default (LGD)
1397.1 Definition of Loss 140
7.2 Parameters for LGD Calculation 140
7.2.1 LGD-Specific Loss Components in Non-Retail Transactions
1407.2.2 LGD-Specific Loss Components in Retail Transactions
1437.3 Identifying Information Carriers for Loss Parameters 144
7.3.1 Information Carriers for Specific Loss Parameters
1447.3.2 Customer Types
1467.3.3 Types of Collateral
1487.3.4 Types of Transaction
1497.3.5 Linking of Collateral Types and Customer Types
1507.4 Methods of Estimating LGD Parameters 151
7.4.1 Top-Down Approaches
1517.4.2 Bottom-Up Approaches
1537.5 Developing an LGD Estimation Model 157
8 Estimating Exposure at Default (EAD)
1628.1 Transaction Types 162
8.2 Customer Types 163
8.3 EAD Estimation Methods 165
IV REFERENCES
167V FURTHER READING 170