Advanced IRB Modelling

Agenda Program
divider graphic
Prague, NH Hotel Prague
divider graphic
Dynamics of Internal Rating Systems
Building, Validating, and Using PD Models
Measuring Market, Workout and Implied LGD
Modelling PD and LGD in Low-Default Portfolios (LDP)
Measuring EAD of Loans, Credit Lines and OTC Derivatives
Backtesting and Validation of PD, LGD and EAD Models
Practical Applications of Internal Rating Models
A two-day course on how to build, implement and validate models for calculating PD, LGD, EAD and RWA

The purpose of this workshop is to give you a practical understanding of methods for estimating and using the risk parameters (PD, LGD and EAD) in the Basel III IRB Risk Weight Function.

The course is primarily targeted towards persons who work "hands-on" with credit risk modelling in banks. The course is, however, also very relevant to senior staff members and to regulators, who need a thorough understanding of the "workings" of the IRB-A approach and of the challenges of estimating, implementing and validating models for quantifying credit risk.

We start with a brief introduction to the IRB approach to measuring credit risk. We explain and demonstrate how estimates for the risk parameters are translated into risk weights using the Supervisory Risk Function, and we discuss the underlying assumptions.

We then turn to look in more detail at various approaches to modelling PD, LGD and EAD. We will focus particularly on challenging issues such as the modelling of PD and LGD for low-default portfolios and the estimation of recoveries under the "workout" method. We also thoroughly explain methods for estimating the EADs for credit lines and OTC derivatives.

Finally, we explain how PD and LGD models can be validated and how the outputs can be used for regulatory and internal management purposes.

The course is designed as mix of theoretical presentations/discussions, practical examples and small exercises.

Program of the seminar: Advanced IRB Modelling

The seminar timetable follows Central European Time (CET).

09.00 - 09.15 Welcome and Introduction

09.15 - 12.00 Understanding Internal Rating Systems and the IRB Risk Weight Function

  • Background and Economic Foundations
  • Model Specification
  • Worked Examples: Calculating RWA for Different Exposure Types
  • Main Problems in Building, Validating, and Using Internal Rating Systems
    • Capturing financial statement information accurately
    • Measuring default rates
    • Using the appropriate model

Modelling PD

  • Approaches to Modelling PD
    • Heuristic approaches
    • Statistical/causal approaches
    • Hybrid solutions
  • Dynamics of Internal Rating Systems
    • Obligor-specific default probabilities
    • Classification of rating systems
    • Quantifying pooled PDs
  • An Illustrative Model of Rating System Dynamics and PD Quantification
  • Small Exercises

12.00 - 13.00 Lunch

13.00 - 16.30 Modelling PD (Continued)

  • Modelling PD in Low-Default Portfolios (LDP)
    • From the perspective of Basel
    • Augmenting the data set with "quasi-defaults"
    • Using expert opinions
  • LDP Verification
    • Single factor analysis
    • Weight optimization
    • Mapping optimization
    • Model reliability test
  • Dynamic scoring: Survival Analysis
  • Backtesting Pooled PDs
  • Validation of Ratings Systems and PD Models
    • Key components of validation
    • Discrimatory power
    • Calibration
  • Worked Examples and Small Exercises

Day Two

09.00 - 09.15 Recap

09.15 - 12.00 Modelling LGD

  • Definitions of Default and Loss
  • What Does Basel Demand on LGD Key Criteria
  • LGD Drivers
  • An Illustrative LGD Decision Tree Framework
  • Three Broad Ways of Measuring LGD
    • Market LGD
    • Workout LGD
    • Implied LGD
  • A Closer Look at Workout LGD
    • Recoveries
    • Costs
    • Discount rates
  • Collateral Allocation and Collateral Valuation
  • Measuring LGD for Non-Performing Portfolios
  • Measuring LGD Low-Default Portfolios
  • Validation of LGD
    • Proportional decomposition method
    • Marginal decomposition method
  • Worked Examples and Small Exercises

12.00 - 13.00 Lunch

13.00 - 16.30 Modelling EAD

  • Credit Characteristics
    • Fixed vs. floating, revolving or non-revolving
    • Covenants, restructuring
    • Obligor-specific characteristics
  • Three Important Concepts in EAD Modelling: LEQ, CCF and EADF
  • CCF Risk Drives - Choice of Method
  • EAD - Design of Model and Back-Testing
  • Measuring Downturn EAD
  • Measuring EAD of OTC Derivatives Portfolios
    • The add-on method
    • The internal modelling method

Putting It All together

  • Complete Worked Example of the Estimation and Applications of PD, LGD and EAD Models

Evaluation and Termination of the Seminar

Training catalogue in PDF
arrow-up icon