Accuracy ratio (AR)
FRM Part 2 glossary · Risk & Investment Management
Definition
Discriminatory-power summary of a rating model; AR = 2AUC - 1.
Accuracy ratio (AR) in plain English
Imagine a credit-scoring model that rates borrowers, and you want to know: is it actually good at telling future defaulters apart from people who'll pay you back? The Accuracy Ratio is a single grade from 0 to 1 that answers exactly that — 1 means the model perfectly ranks the risky people above the safe ones, and 0 means it's no better than flipping a coin. It's built from AUC (a related score where 0.5 is pure guessing) via AR = 2 × AUC - 1, which just re-centers the scale so that useless = 0 instead of 0.5. Think of it like a metal detector: a high AR means it beeps loudly for real threats and stays quiet for harmless stuff, while a low AR beeps at random.
Also called: AR
Learn these first
Practice Accuracy ratio (AR)
Related FRM Part 2 terms
FRM Part 2 practice questions · FRM Part 2 flashcards · FRM Part 2 study guide