Credit Scoring & Rating Systems Flashcards
7 cards from real CCM practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 7 Credit Scoring & Rating Systems flashcards as text
What is the primary advantage of an internally developed credit scoring model over relying solely on external credit scores?
Answer: Internal models can incorporate company-specific payment history, industry data, and tailored risk appetite
Internal models can be calibrated to the company's specific customer base, products, and risk tolerance, using proprietary payment data that external bureaus do not have.
What does 'scorecard validation' refer to in credit scoring model development?
Answer: Testing the scoring model against holdout or out-of-sample data to assess its predictive accuracy
Scorecard validation involves applying the model to data not used in its development to confirm it predicts credit risk reliably and has not been overfitted.
What is 'behavioral scoring' in the context of credit management?
Answer: Using an established customer's ongoing payment behavior and account activity to dynamically reassess creditworthiness
Behavioral scoring continuously monitors existing customers' actual payment patterns and account usage to update credit risk assessments and adjust credit limits over time.
Which statistical technique is most commonly used to develop credit scoring models?
Answer: Logistic regression
Logistic regression is the industry standard for credit scoring because it models the probability of a binary outcome — default versus non-default — as a function of predictor variables.
In credit scoring, the Gini coefficient is used to measure:
Answer: The discriminatory power of a scoring model to separate good accounts from bad accounts
In credit risk, the Gini coefficient (derived from the ROC curve) quantifies how well a model distinguishes between creditworthy and non-creditworthy obligors; higher values indicate better separation.
When a credit manager applies a 'judgmental override' to a credit scoring model result, what are they doing?
Answer: Manually adjusting a credit decision away from the model's recommendation based on qualitative information
A judgmental override allows a credit professional to deviate from the model's output when additional qualitative factors or specific circumstances warrant a different decision.
What is the Population Stability Index (PSI) used for in credit scoring?
Answer: Detecting significant shifts in the distribution of a scoring model's input population over time
PSI measures whether the distribution of a key input variable has changed materially from the development population, signaling that the model may need recalibration or redevelopment.