Credit Scoring and Internal Rating
This course provides practical knowledge in the creation of credit scoring and internal rating systems with a focus on the relevant statistical techniques and tools.
Attendees should have a background in predictive modelling or credit risk management and want to learn more about building and validating
predictive models in the credit risk area, especially in the context of Basel 2.
Dr Hendrik Wagner
 For 9 years Dr Wagner was Product Manager Data Mining Solutions at the SAS Institute covering Europe, Middle East and Africa.
He introduced scorecard development functionality into SAS' flagship data mining solution Enterprise Miner and made it the market leading solution for inhouse scorecard development. He also led the creation of an end-to-end model development, deployment and monitoring solution and
defined specific functionality for building internal rating systems for Basel 2 -PD and LGD modeling, pooling and backtesting.
After writing SAS' first Risk Weighted Assets calculation code, he helped launch SAS' market leading Credit Risk Management solution.
He became a consultant in 2006 providing credit risk and internal audit departments with advisory and implementation services, such as readiness assessment, model development and rating system auditing.
Clients include inter alia GHB bank, Thailand, (Housing Loan Application Scorecard), Samlink, Finland, (Behavioral PD Model),Maybank Malaysia(Corporate PD Model Validation), National Australia Group UK, (Retail PD, LGD and EAD Model Validation for Basel2 IRB Approval), Deutsche Telekom Germany (PD model validation and development, early warning system, profit scoring) and BHW Bausparkasse ( PD and LGD model validation of a home loans portfolio).
Dr Wagner holds a doctorate in Computer Sciences.
Duration: 2-3 days
Summary
Part 1: Introduction to Credit Scoring and Internal Rating
Part 2: Scoring Model Development
- Classic Scorecard Development
- Sampling and Data Partitioning
- Multivariate Variable Selection
- Dealing with Interactions
- Trees and Neural Networks
Part 3: Portfolio Specifics- Application Scoring
- Behavioural Scoring
- Corporate and SME Rating
Part 4: Internal Rating Systems
- PD Calibration
- Rating System Validation
- LGD and EAD Modelling
Part 1: Introduction to Credit Scoring and Internal Rating
Goal: Get to know the reasons for credit scoring and internal rating. Understand the environment in which scoring models get created, live and die.
- Credit Scoring and Internal Rating
- Decision Process Optimization
- Model Lifecycle Management
- Model Development
- Model Deployment
- Model Monitoring
Part 2: Scoring model development
Goal: Learn to build Probability of Default models. Learn to build a basic scorecard and then to optimize it. Get to know alternative modelling methods. Learn to explain model predictions.
- Getting started
- Principles of predictive modelling (Generalization)
- Building a first scorecard
- Grouped Variable Regression
- Weight of Evidence coding versus dummy coding
- Automatic and Interactive Variable Grouping and Selection
- Splitting criteria and options
- Optimizing the Scorecard
- Sampling, biased sampling and adjustments
- Data partitioning
- Correlation, stepwise regression and variable clustering
- Interactions and segmentation
- More flexible modelling methods
- Tree models
- Missing values and outliers
- Neural network models
Part 3: Portfolio Specifics
Goal: Get to know the specific challenges of building PD models for scoring new applicants for retail products, existing retail clients and Small and Medium Enterprise applicants and clients. Appreciate how differences in data availability and model purpose influence the model development
process.
Retail Application Scoring
- Variables for application scoring
- Co-applicants
- Credit bureau
- Bad definition
- Reject Inference
- Setting cut-off
Retail Behavioural Scoring
- Account level or customer level scoring
- Variables behavioural scoring
- Account behaviour
Corporate and SME Rating
- Corporate rating models
- Balance sheet analysis
- Segments and sectors
- Quantitative and qualitative rating
- Shadow rating
Part 4: Internal Rating Systems
Goals:
Learn how to calibrate a graded PD rating system.
Learn how to assess and monitor model quality in terms of concentration, stability, discriminatory power and calibration.
Understand the data preparation for defining development tables for LGD prediction in retail.
Get to know modelling methods that are especially suited for LGD prediction.
Learn how to validate LGD and EAD models. Calibrating PD Grades for Basel 2
- Calibrating scoring models
- Grade definition
- Economic cycle adjustments
Rating System Validation
- Data preparation for validation and monitoring
- Defining default rate
- Measuring concentration
- Measuring stability
- Measuring discriminatory Power
- Measuring calibration
- Confidence intervals
LGD and EAD Modelling and Validation
- Averaging and cohort approaches
- Development sample definition
- Regression trees
- Linear regression
- Beta regression
- Two-stage scenario modelling
- Generalized Additive Neural Network
- LGD scorecard
- EAD modelling
- Validating LGD and EAD models
| 
