This business-focused course provides a high-level introduction to credit risk management; detailed end-to-end methodology training for risk scorecard development for retail and SME portfolios; and discussions about scorecard implementation and risk strategy development and about scorecard and portfolio management reporting. The focus of the course is the development of application scorecards, but issues relating to behavior scorecard development are also explored. Detailed mechanics of scorecard development are covered, but more importantly, students learn to think critically about each phase and use a combination of business and statistical criteria to make better decisions. These objectives are reflected in three sections.

Section 1, Introduction to Credit Risk: This section provides a high-level overview of the credit risk industry, risk management tools, and strategies. Students will understand the different uses of credit risk scorecards and learn industry terminology, as well as understand the main personas involved in successful credit scoring projects.

Section 2, Risk Scorecard Development: Students learn how grouped-variable, points-based credit risk scorecards are developed, from the planning stages to delivery. The main focus is on business issues, but statistical aspects of scorecard development are also explored. Students learn practical tips and tricks for tasks such as variable binning, creating intelligent variables and scorecards, and using a wide variety of data sources.

Section 3, Implementation and Maintenance: This section covers post-development activities, including setting cutoffs, developing strategies, and creating scorecard maintenance reports. Students will understand how scorecards are used to make better decisions, enabling them to build more relevant scorecards.

Learn how to

• Recognize the different uses for credit risk scorecards.
• Recognize the main personas involved in successful credit scoring projects.
• Plan a successful scorecard development project.
• Develop grouped-variable, points-based credit risk scorecards, from start to finish.
• Identify and fix common data issues such as sample biases.
• Create intelligent derived variables from a variety of sources, including telco, online, and transactional data.
• Perform validation, use multiple scorecards, and set cutoffs.
• Develop strategies using scorecards.
• Create reports for scorecard validation and maintenance. TEST

Who should attend

Credit risk/scoring managers and data miners; those involved in model vetting/validation and auditing; risk strategy developers; and credit risk executives

Formats available	Standard Duration
Classroom:	2 days

No SAS experience or programming experience is required. Students should be familiar with logistic regression. Business experience is preferred, particularly in the following sectors:

• Banking/finance.
• Retail.
• Insurance.
• Telecommunications.

This course addresses SAS Risk Management software.

Although the solution is used to illustrate the examples, the course is software-neutral. There are no hands-on programming practices in this course.

Introduction to Credit Risk

• Risk management areas.
• Terminology.
• Applications of credit scoring in various industries.

Preliminaries and Planning

• Create business plan: identify organizational objectives and identify internal versus external development and scorecard type.
• Create project plan: identify project risks and Identify the project team.

Data Review and Project Parameters

• Data availability and quality review.
• Data gathering for definition of project parameters.
• Definition of project parameters: performance window and sample window, performance categories definition, and exclusions.
• Segmentation.

Development Database Creation

• Development sample specifications.
• Development data collection and construction.
• Adjusting for prior probabilities/factoring.

Model Development

• Missing values and outliers.
• Initial characteristic analysis (binning): statistical measures and logical trend.
• Building a preliminary scorecard using logistic regression.
• Reject inference.
• Final scorecard production: scaling, misclassification, and scorecard strength measures.
• Validation.

Scorecard Management Reports

• Gains tables.
• Characteristic reports.

Pre-implementation

• Pre-implementation validation: scoring accuracy and front end validation.

Strategy Development

• Strategy development considerations.
• Scoring strategies.
• Setting cutoffs.
• Strategy actions.
• Policy rules.
• Overrides.

Post-implementation

• Review/postmortem.
• Reporting: system stability, characteristic, override, final score, delinquency, and vintage.