Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Credit Risk Models


Published on

Presentation Slide from WMSL's seminar

Published in: Economy & Finance
  • Be the first to comment

  • Be the first to like this

Credit Risk Models

  1. 1. Credit Risk Models August 24 – 26, 2010
  2. 2.  1st Case Study : Credit Rating Model Borrowers and Factoring (Accounts Receivable Financing) pages 3 – 10  2nd Case Study : Credit Scoring Model Automobile Leasing pages 11 – 20  3rd Case Study : The Validation of Internal Rating Systems pages 21 – 28  Credit Risk Model : What is Credit Risk Model? page 29 What Properties to be expected! page 29 Application page 30 AGENDA
  3. 3. 1st Case Study Credit Rating Model Borrowers and Factoring (Accounts Receivable Financing) 3
  4. 4. Credit Rating Model ….. (1) Model Development Portfolio Analysis Performance Test Data Analysis Data Preparation Implementation Recommendation Calibration and Mapping Data Cleansing Data Partition Univariate Correlation Multivariate Accuracy Blind Test 4
  5. 5. Credit Rating Model ….. (2) Portfolio Analysis Distribution of Financial Statements and Default Data -Year; Business Type / Group / Concentration; Size Performance Window Bad Definition Coverage Number of Accounts 5
  6. 6. Credit Rating Model ….. (3) Data Preparation Data Cleansing Data Partition Sample Grouping Dimension of Sampling Existing and Completeness of Data Variables -FinancialRatio -Qualitative Data Correctionof Data Logic Loan Status comparedto FinancialPerformance In Sample / Out-of-Sample Out-of-Universe Out-of-Time Product Type Business Type Asset Size 6
  7. 7. Credit Rating Model ….. (4) Data Analysis Univariate Correlation Multivariate Logistic Regression Screening Criteria Multiple Logistic Regression Quantitative Qualitative Make Business Sense Powerful Understandable / Intuitive Enough observation to develop and validate model 7
  8. 8. Credit Rating Model ….. (5) Performance Test Accuracy Blind Test 8
  9. 9. Credit Rating Model ….. (6) Calibration and Mapping Calibration Mapping 9
  10. 10. Credit Rating Model ….. (7) 10
  11. 11. 2nd Case Study Credit Scoring Model Automobile Leasing 11
  12. 12. Credit Scoring Model Scorecard Development Portfolio Analysis Model Development Portfolio Distribution Vintage Analysis Test on Similarity and/or Differentiation of Bad Rates Data Gathering Model Building Data Cleansing Variable Classing Preliminary Variable Selection Purposes of analyzing portfolio: -To understand the overall picture of portfolio -To know the portfolio’s default rates in terms of Marginal and Cumulative Default Rates -To help set the sample group to be collected for model development 12
  13. 13. Model Development ….. (1) Data Gathering Performance Window VariablesSample Size Bad Definition Dependent Variables Independent Variables Borrower’s Characteristics Loan Status Facility’s Characteristics Collateral’s Characteristics 13
  14. 14. Model Development ….. (2) Data Cleansing Data Exploration Problem and Solution Number of Fields Number of Complete Cases Missing Value Error from Data Transformation Extreme Value Vague / Unclear Data Reliability of Data Missing Data Discrepancy of Data 14
  15. 15. Model Development ….. (3) Variable Classing Manage the number of attributes per characteristics Make the Weight of Evidence – and thereby the number of points in the scorecard – vary smoothly or even linearly across the attributes Select predictive characteristics Improve the predictive power of the characteristics Classing is process of automatically and / or interactively binning and grouping interval, nominal, or ordinal input variables in order to 15
  16. 16. Model Development ….. (4) Variable Selection Preliminary Statistical Missing Value Consistency Completeness and relationship with other variables Univariate data analysis Multivariate data analysis 16
  17. 17. Model Development ….. (5) Model Building Why Logistic Regression! -It can handle discrete variable or qualitative variable. -The dependent variable need not to be normally distributed. -The dependent variable need not to be homoscedastic for each level of the independents. -Normally distributed error terms are not assumed. -It does not require that the independents be interval. 17
  18. 18. Model Development ….. (6) Model Building Sample Accuracy Training Sample Out-of-time Sample Type I & II errors K-S Statistics Testing Sample Discriminatory Power of the model : The maximum difference between the cumulative percent good distribution and the cumulative percent bad distribution 18
  19. 19. Model Development ….. (7) Criteria for Model Selection Accuracy : Type I & II errors; K-S statistics Including or excluding Policy Indicators in the model Number of Variables in the model Consistency of testing results among Training, Testing and Out-of-time Samples ExampleofGainTable 19
  20. 20. Model Development ….. (8) Scorecard Implementation and Application 20
  21. 21. 3rd Case Study The Validation of Internal Rating Systems 21
  22. 22. Scope of Work and Validation Aspects Scope of Work Analyze the discriminatory power of rating models ValidationAspects Analyze the stability of rating models Analyze the connection between PD and Grade Analyze the models design Analyze the rating process 22
  23. 23. Validation Method Study and Analyze Data from Documents Interview and Site VisitDefault Probability Model Validation Statistical and Mathematical Tests 23
  24. 24. Validation Results Quantitative : Discriminatory Power Type I & II errors in theory RelativeFrequency Rating Class Density Function for Bad Cases Density Function for Good Cases Cut-Off Point Type I error Type II error Project Financing <15M; Hire Purchase; Leasing Project Financing <15M Hire Purchase and Leasing Project Financing >= 15M Factoring Type I error of each model 24
  25. 25. Validation Results Quantitative : Discriminatory Power (ROC Curve) Project Financing < 15M; Hire Purchase; Leasing Project Financing >= 15M Factoring 25
  26. 26. Validation Results Quantitative : Stability Project Financing < 15M; Hire Purchase; Leasing Project Financing >= 15M Factoring % PD from Actual Default Rate : ADR (%) 2006 2007 2008 Overall 26
  27. 27. Validation Results Quantitative : Calibration Actual Default Rate (%) of each model compared to Implied PD (%) by CQC Grade PF < 15M PF >= 15M HP: LS PF < 15M; HP; Leasing Implied PD Actual Default Rate (%) of each model compared to S&P’s Default Rate by Rating PF >= 15M HP: LS S&P’s Default Rate PF < 15M PF < 15M; HP; Leasing 27
  28. 28. Executive Summary 28
  29. 29. Credit Risk Model ….. (1) What is Credit Risk Model?  A tool used to evaluate the level of risk associated with applicants or borrowers.  It consists of a group of characteristics, statistically determined to be predictive in separating “good” and “bad” accounts.  It provides statistically odds or probability that an applicant or borrower with any given rating or score will be “good” or “bad”. What Properties to be expected!  Understandable  Powerful  Calibrated  Empirically validated 29
  30. 30. Credit Risk Model ….. (2) Application  Origination Decisions  Given the risk and a fixed price, is the asset worth taking?  Given the risk, what price is required to make the asset worth buying?  Portfolio Optimization  To reduce the portfolio’s risk, concentrations of risk and how the risk can be diversified must be known.  Capital Management  To set capital, the loss level is needed.  Credit Process Management  To gain the efficiencies of application processing that comes through automation.  To gain control and consistency in lending practices for the entire credit portfolio.  To identify the variables which are important in the credit evaluation process  To improve delinquency statistics while maintaining desired approval rates 30
  31. 31. Credit Risk Model ….. (3) Credit Risk Model only classifies and predicts risk; It does not tell the lender how to manage it. 31