CLARK ABRAHAMS is the Director for Fair Banking at SAS, where he leads business and product development. He has over thirty years of experience in the financial services industry, at corporations including Bank of America and Fair Isaac Corporation. MINGYUAN ZHANG is Solutions Architect for SAS Financial Services. Over the last 10 years with SAS Institute, he has successfully developed and implemented many economic forecasting, data mining, and financial risk management solutions for various industries. Prior to joining SAS, he served as an economic and financial analyst for a leading telecommunications consulting firm.

• Foreword ixPreface xiiiAcknowledgments xvii1 Credit Access and Credit Risk 1Enterprise Risk Management 2Laws and Regulations 4Changing Markets 6Prepare for the Challenges 8Return on Compliance 14Appendix 1A: Taxonomy of Enterprise Risks 17Appendix 1B: Making the Business Case 182 Methodology and Elements of Risk and Compliance Intelligence 23Role of Data in Fair Lending Compliance Intelligence 23Sampling 29Types of Statistical Analysis 35Compliance Self-Testing Strategy Matrix 36Credit Risk Management Self-Testing Strategy Matrix 38Matching Appropriate Statistical Methods to Regulatory Examination Factors 42Case for a Systematic Approach 43Summary 44Appendix 2A: FFIEC Fair Lending Examination Factors within Seven Broad Categories 463 Analytic Process Initiation 51Universal Performance Indicator 51Overall Framework 53Define Disparity 53Derive Indices 58Generate Universal Performance Indicator 65Performance Monitoring 75Summary 80Appendix 3A: UPI Application Example: Liquidity Risk Management 834 Loan Pricing Analysis 85Understanding Loan Pricing Models 87Systematic Pricing Analysis Process 91Overage/Underage Analysis 112Overage/Underage Monitoring Overview 123Summary 125Appendix 4A: Pricing Analysis for HMDA Data 126Appendix 4B: Pricing and Loan Terms Adjustments 133Appendix 4C: Overage/Underage Data Model (Restricted to Input Fields, by Category) 137Appendix 4D: Detailed Overage/Underage Reporting 139Appendix 4E: Sample Size Determination 1425 Regression Analysis for Compliance Testing 147Traditional Main-Effects Regression Model Approach 148Dynamic Conditional Process 151DCP Modeling Framework 154DCP Application: A Simulation 168Summary 180Appendix 5A: Illustration of Bootstrap Estimation 1816 Alternative Credit Risk Models 183Credit Underwriting and Pricing 184Overview of Credit Risk Models 185Hybrid System Construction 201Hybrid System Maintenance 216Hybrid Underwriting Models with Traditional Credit Information 222Hybrid Underwriting Models with Nontraditional Credit Information 234Hybrid Models and Override Analysis 237Summary 248Appendix 6A: Loan Underwriting with Credit Scoring 250Appendix 6B: Log-Linear and Logistic Regression Models 254Appendix 6C: Additional Examples of Hybrid Models with Traditional Credit Information 255Appendix 6D: General Override Monitoring Process 2657 Multilayered Segmentation 267Segmentation Schemes Supporting Integrated Views 267Proposed Segmentation Approach 269Applications 275Summary 297Appendix 7A: Mathematical Underpinnings of BSM 298Appendix 7B: Data Element Examples for Dynamic Relationship Pricing Example 3018 Model Validation 305Model Validation for Risk and Compliance Intelligence 305Typical Model Validation Process, Methods, Metrics, and Components 307An Integrated Model Validation Approach 317Summary 344Closing Observations 344Index 347