Glossary term
Hit Rate
Subject: Financial Crime, Sanctions
Also known as: Screening hit rate
Hit rate is the share of screening checks that produce an alert, or the share of those alerts that turn out to be real matches. It measures how well a screening system is tuned.
Explanation
The term is used in two ways, so teams state which one they mean. The first is the alert rate: of all customers or payments checked, how many throw up an alert for review. The second is the true hit rate: of all alerts reviewed, how many prove to be genuine matches with a listed party. A very high alert rate with almost no genuine matches means analysts spend their days clearing noise. A rate near zero can be good news, or a sign that matching is set so tightly that real risks slip past. Hit rate is tracked by list, by screening type and by threshold setting, and compared before and after any tuning change. Model validators and regulators ask for it when they test whether a screening tool is effective. Read alongside alert volumes, it shows whether review effort is landing on real risk.
Example
A mid-sized bank screens 200,000 payments a month and its system raises 4,000 alerts, an alert rate of 2%. Analysts confirm only 8 of them as genuine matches. The compliance team treats the confirmed share, 0.2%, as the true hit rate. It tests a stricter fuzzy-matching threshold on historical data, finds that all 8 true matches would still be caught, and cuts the alert volume by about a third.
Not to be confused with
- False Positive rate
- The share of alerts that prove wrong. Hit rate can describe all alerts or only the genuine ones, so the two are not interchangeable.