Skip to content

FRM Part II · FRM Exam Part II · Introduction to Credit Risk Modeling and Assessment

A validation team evaluates a scorecard using the ROC curve. Model A has an AUC of 0.82 and a Gini (accuracy ratio) coefficient the team must report. Model B has a Gini of 0.70. Assuming Gini = 2*AUC - 1, which statement is correct?

Model A's Gini is 2×0.82 − 1 = 0.64, while Model B's Gini of 0.70 implies an AUC of 0.85. Model B therefore discriminates better. Calibration does not affect this comparison, which depends only on ranking power.

  1. AModel A has Gini 0.64 and is less discriminating than Model B
  2. BModel A has Gini 0.82 and is more discriminating than Model B
  3. CModel A has Gini 0.64 and is less discriminating than Model B only if calibrated
  4. DModel A has Gini 0.64, and Model B has an AUC of 0.85, so Model B discriminates betterCorrect

Explanation

Model A: Gini = 2*0.82 - 1 = 0.64. Model B: AUC = (0.70+1)/2 = 0.85. Since 0.85 exceeds 0.82, Model B discriminates better. Reporting Gini equal to AUC is a conversion error, and calibration is irrelevant to ranking power.

Did you get it right without looking?

One question tells you little. A timed set on Introduction to Credit Risk Modeling and Assessment shows your real accuracy, how long you take and where you lose marks.

Practise Introduction to Credit Risk Modeling and AssessmentFree diagnostic mock

More Introduction to Credit Risk Modeling and Assessment questions