Why Banks Need to Run the Numbers
Bank impersonation fraud has has become one of the more pressing challenges facing financial institutions today. Fraudsters pose as the bank, contact customers directly, spoof caller IDs, and walk away with funds before anyone realizes what happened. The losses are real and they tend to compound in ways that don’t show up cleanly in a single report.
That’s why we built the Refine Intelligence Attack Impact Simulator: to give banks a clearer view of the full financial picture before an attack happens, not after.
Why Simulation Matters Now
The volume and sophistication of bank impersonation attacks have grown considerably in recent years. Fraudsters now combine data from prior breaches with AI-generated voice and scripting to create calls that are difficult for customers to distinguish from a genuine fraud alert. Customers act quickly, transferring funds under the belief that they’re protecting their account, which means by the time the bank is aware, recovery options are limited.
What makes this particularly difficult for institutions is that the damage doesn’t end with the stolen funds. There’s the operational burden of investigation, legal and regulatory exposure, mandatory customer communications, and a quieter but equally real cost: customers who experienced the attack, or simply heard about it, gradually reducing their balances or moving deposits elsewhere. That deposit attrition hits the bank’s net interest margin and rarely appears in any fraud report.
For most institutions, these costs have never been modeled together. The fraud team tracks reimbursements. Operations tracks investigation hours. Finance tracks NIM. The full number often goes unseen. That’s the gap this tool is designed to close.
How the Model Works
The simulator takes one input: your total assets under management. From there, it derives everything else using published industry benchmarks, with no historical fraud data required.
- Customer base scales with asset size, accounting for the fact that growth in customer count slows as institutions get larger. A $1B institution typically serves around 35,000 customers; a $10B institution around 244,000; a $100B institution around 1.7 million.
- Victims in a single attack wave also scale sub-linearly, reflecting that larger banks tend to have stronger controls and attackers face more friction at scale. The model assumes roughly 100 victims at a $1B institution, around 500 at $10B, and 1,500 at $60B.
- Fraud losses are blended based on your consumer/commercial account mix. The model uses $3,000 per consumer victim (FTC 2024) and $12,000 per commercial victim (AFP 2024), and both figures are adjustable.
- Recovery and reimbursement reflect current industry norms: a 40% fund recovery rate and a 70% bank reimbursement rate, the latter driven by Reg E obligations and ongoing CFPB pressure.
- Deposit NIM loss captures the 0.4% deposit attrition that FDIC research associates with publicized fraud events, applied against the current community bank NIM rate of 3.2%.
Putting the Numbers in Context
Take a $10B community bank with a 60/40 consumer-to-commercial account split. The model estimates 244,000 customers, with roughly 460 directly victimized in a single wave. The blended average fraud loss comes out to around $6,600 per victim.
Gross stolen: approximately $3M. After a 40% recovery rate and a 70% reimbursement obligation, the bank’s direct payout is around $1.26M. Investigation labor, crisis communications, and regulatory exposure add approximately $275K. Deposit NIM loss from attrition adds another $1.3M. Total economic impact from one wave: roughly $2.8 million.
That’s the realistic scenario. A severe outcome, where response is slower, media coverage follows, and regulators take notice, applies a 1.65x multiplier and pushes the number above $4.6M.
Try It Yourself
The simulator is available now on the Refine Intelligence website. Enter your asset size, adjust the account mix and assumptions as needed, and run the simulation. It takes under a minute and produces a number concrete enough to bring into a planning conversation.
Understanding your exposure is the first step toward addressing it.
