CREDIT RISK ANALYTICS

MEASUREMENT TECHNIQUES,
APPLICATIONS, and EXAMPLES

  CREDIT RISK ANALYTICS
  • HOME
  • TABLE OF CONTENT
  • DATASETS
  • TRAINING
    • ONLINE
    • ZOOM MASTERCLASS
    • IN HOUSE
  • AUTHORS
  • PAPERS
  • UPDATES
  • CONTACT

We have noticed the following errata in the SAS book:

  • Introduction
  • Introduction to SAS software
  • Exploratory Data Analysis
  • Data Preprocessing
  • Credit Scoring
    • Page 107: All headings need to move on cell to the right;
  • Probabilities of default (PD): discrete time hazard models
    • Page 146: Change from PROC GPLOT DATA = test; to PROC GPLOT DATA = graph;
    • Page 155: Add between PROC LOGISTIC and the data step: PROC SORT DATA=probabilities2; BY time; RUN; PROC MEANS DATA=probabilities2; BY time; OUTPUT OUT=means MEAN(default_time PD_TTC_time PD_PIT_time)=default_time PD_TTC_time PD_PIT_time; RUN;
    • Page 168: Move code "OUTALL" into one line
    • Page 170: Exhibit 6.23: change heading to "Selection Indicator = 0", SAS may output Exhibit 6.23 before Exhibit 6.22
  • Probabilities of default: continuous time hazard models
    • Page 194 - data creation code needs to come before the survival code chunk above
    • Page 198 - change  LTV_time to LTV_orig_time on page 198
      Page 199: Change: MODEL (time1,time2)*default_time(0)=FICO_orig_time LTV_time gdp_time TIES=EFRON; to MODEL (time1,time2)*default_time(0)=FICO_orig_time LTV_orig_time gdp_time /TIES=EFRON;
    • Page 199 - coverates_time data creation code chunk needs to come before chunk on page 198
      Page 208: Change: xbeta = 2.0998+ 0.0044* FICO_orig_time-0.0159*LTV_orig_time  to
      xbeta = 2.1137+ 0.0043* FICO_orig_time-0.0155*LTV_orig_time
  • Low Default Portfolios
  • Default Correlations and Credit Portfolio Risk
  • Loss Given Default (LGD) and Recovery Rates
    • Page 293 - add MODEL lgd_time ~ GENERAL(ll); before RUN;
    • Page 302 - need to add ODS GRAPHICS ON/OFF to display all graphics shown in book for this specific analysis
  • Exposure at default (EAD) and adverse selection
  • Bayesian Methods for Credit Risk Modeling
  • Model validation
    • Page 402/403: MODEL statement may be removed for all models with INMODEL statement in PROC LOGISTIC
  • Stress testing
  • Concluding remarks
GO TO DATASETS

Copyright © 2023

Privacy Policy

Terms & Conditions

  • HOME
  • TABLE OF CONTENT
  • DATASETS
  • TRAINING
    • ONLINE
    • ZOOM MASTERCLASS
    • IN HOUSE
  • AUTHORS
  • PAPERS
  • UPDATES
  • CONTACT