PURPOSE

To provide modern techniques for evaluation of credit risk and prediction of bankruptcy with emphasis on applications to the Brazilian market.

 MAIN TOPICS

. Univariate and multivariate discriminant analysis
. Approval, price and collaterals
. Behavior score
. Sorting of variables into categories and dummies generation
. Comparison of models
. Correlations and portfolio risk
. Criteria for selecting models
. Definition of good and bad payer
. Performance of a scoring model
. Population stability
. Statistics of Kolmogorov and Smirnov
. Identification of potential variables
. Logit, Probit and discriminant analysis
. Regression methods
. Nature of the credit risk
. Cut-off points and score ranges
. Statistical Prediction of Bankruptcy
. Defaulting probability
. Revenues and recovery rate
. Regression with discreet dependent variables
. Revision of cut-off points
. Selection of sample and type I and type II errors
. Statistical tests and selection of variables and models
. Validation of a scoring model