Back to Glossary

Fraud & AML

What is a fraud score calibration curve?

A fraud score calibration curve measures how well predicted fraud probabilities align with actual fraud rates across score ranges, ensuring model outputs represent true likelihood of fraudulent transactions.

Why It Matters

Proper calibration prevents cost overruns from false positives that can reject 15-25% of legitimate transactions worth millions in revenue. Well-calibrated models reduce manual review queues by 40-60% while maintaining detection rates above 85%. Financial institutions save $2-4 million annually per billion in transaction volume through improved precision in fraud scoring.

How It Works in Practice

  1. 1Segment historical transactions into score deciles based on fraud model predictions
  2. 2Calculate actual fraud rates within each score bucket over 30-90 day periods
  3. 3Plot predicted probability against observed fraud rate to visualize calibration gaps
  4. 4Apply statistical tests like Hosmer-Lemeshow to quantify calibration quality
  5. 5Adjust model outputs using isotonic regression or Platt scaling techniques
  6. 6Monitor calibration drift weekly and retrain models when deviation exceeds 5% threshold

Common Pitfalls

Seasonal fraud patterns can create temporary calibration drift that appears as model degradation

Regulatory stress testing requirements may demand specific calibration standards that conflict with operational performance

Sample bias from declined transactions creates incomplete fraud labels that skew calibration assessment

Over-calibration on recent data can reduce model sensitivity to emerging fraud vectors

Key Metrics

MetricTargetFormula
Calibration Error<0.05Mean absolute difference between predicted and observed fraud rates across deciles
Brier Score<0.15Mean squared difference between predicted probabilities and binary fraud outcomes
AUC-ROC>0.85Area under receiver operating characteristic curve measuring discrimination ability

Related Terms