Fraud Decision Net Economic Yield (FDNEY)

GMI-PAY-001Payments & Transaction ProcessingPilot-Ready

Definition

Net economic value of fraud model decisions per period, calculated as fraud losses prevented minus revenue destroyed by false declines.

The Measurement Gap

The Measurement Gap

No metric in the standard payments measurement stack tells you the net economic value of fraud model decisions. The fraud model makes two types of consequential decisions every time it scores a transaction: it either correctly prevents fraud (saving the organization money) or it incorrectly declines a legitimate transaction (destroying revenue the organization would have earned). Every fraud metric in common use measures one side of this equation while ignoring the other.

Authorization approval rate tracks the volume of approvals but says nothing about the economic quality of the declines it excludes. Fraud basis points measures aggregate fraud losses as a share of transaction volume but cannot tell you how much revenue was destroyed to achieve that loss rate. False positive rate tracks the error rate of the model but assigns no dollar value to those errors. A fraud model can report excellent fraud basis points, an acceptable false positive rate, and a healthy approval rate while simultaneously destroying more revenue through false declines than it saves through fraud prevention, making it a net-negative investment that every dashboard says is performing well.

The gap is particularly dangerous because false decline costs are estimated at 10-15x actual fraud losses for card-not-present merchants. The unmeasured side of the equation is an order of magnitude larger than the measured side.

Decision Supported

Is the fraud model a net-positive or net-negative economic investment when both fraud prevention value and false decline revenue destruction are accounted for?

Formula

Plain Language

Sum the dollar value of fraud correctly prevented, subtract the revenue margin lost on legitimate transactions incorrectly declined, then subtract the operating cost of the fraud model infrastructure. The result is the net economic contribution of your fraud model.

Notation

FDNEY = Σ(TxnAmount_TP × FraudRecoveryRate) − Σ(TxnAmount_FP × MarginRate) − ModelOperatingCost

Variables

Variable Definition Unit
TxnAmount_TP Dollar amount of correctly declined fraud transactions (true positives) $
FraudRecoveryRate Estimated unrecoverable fraud fraction; typically 0.85–0.95 for card-not-present transactions Ratio (0–1)
TxnAmount_FP Dollar amount of legitimate transactions incorrectly declined (false positives) $
MarginRate Revenue margin per transaction (interchange + processing fee) Ratio (0–1)
ModelOperatingCost Total fraud model infrastructure cost per measurement period $

Proposed formula, pending validation pilots.

At a Glance

Unit
Currency
Trend Direction
Higher
Reporting Frequency
Monthly
Category
Fraud Model Economic Integrity
← Authorization Threshold Staleness Index (ATSI)Fraud Score Calibration Index (FSCI) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment