Chargeback Ratio and Cost Calculator

Visa's own VAMP fact sheet gives the formula in one line: the VAMP ratio is the count of fraud reports (TC40) plus disputes (TC15) divided by the count of settled transactions (TC05), across card-not-present VisaNet transactions, domestic and cross-border. Three things in that sentence catch merchants out. It is count-based, so a $9 dispute weighs as much as a $900 one. Fraud reports sit in the numerator even when the issuer never files a chargeback, so Visa can push you toward a line with events that never touched your balance. And card-present volume is not in the denominator, so a retailer with an online store is measured on the online store alone.

Mastercard defines its ratio differently, and says so in plain language in the Security Rules and Procedures Merchant Edition dated 4 August 2026: basis points are the number of chargebacks received for a merchant in a calendar month, divided by the number of Mastercard transactions in the preceding month, multiplied by 10,000. The denominator is last month. Not this month, not a trailing average, last month.

That single word does real damage to a growing business. Sell 4,000 orders in March and 6,000 in April, take 90 Mastercard chargebacks in April, and your intuitive rate is 90 over 6,000, or 1.50 percent. Mastercard computes 90 over 4,000, or 2.25 percent. You are 75 basis points worse than your dashboard says purely because you grew. A shrinking business gets the opposite gift, which is one reason chargeback trouble surfaces right after a good quarter.

Neither figure is the one your gateway shows you. Most dashboards divide this month's disputes by this month's transactions across every card brand at once. That is a fine internal metric and it is not what either network scores. Run the two formulas separately, on each brand's own volume, or you are managing to a number nobody enforces.

Every one of these programs is an AND, never an OR, and this is where competing pages contradict each other most often. Visa's merchant Excessive line in the United States is 150 basis points, cut from 220 on 1 April 2026, and it applies only where the monthly count of fraud plus disputes reaches 1,500. Visa restated that minimum on its own site in October 2025. The arithmetic lands the opposite way from what small merchants fear: to hold 1,500 fraud and dispute events at exactly the 1.50 percent line you need roughly 100,000 card-not-present Visa transactions that month. Below that scale your ratio can sit at 4 percent and Visa will not name you under VAMP.

Mastercard's counts are much lower and far easier to trip. An Excessive Chargeback Merchant needs at least 100 chargebacks in the month and a ratio of at least 1.50 percent. A High Excessive Chargeback Merchant needs at least 300 and at least 3.00 percent. Both halves of each pair have to be true in the same month, so a merchant with 95 chargebacks at 4 percent is neither, and one with 350 at 2.1 percent stays an ECM because the HECM ratio test fails. Mastercard does not publish those numbers; this page takes them from two acquirer guides that agree. Assessments start in the second month above the line, and at the HECM tier they climb from USD 1,000 in month two to USD 200,000 a month past eighteen months, with issuer recovery adding USD 5 per chargeback above 300 from the fourth month.

So who actually closes an account at 2 percent with 60 chargebacks a month? The acquirer. Every merchant agreement in the United States gives the acquiring bank the right to hold a reserve, reprice, or terminate on its own risk judgment, and most acquirers run an internal limit near 1 percent that they never publish. Termination can be followed by a MATCH listing, which is the thing that genuinely ends a business, and MATCH is an acquirer action rather than a network threshold. Score yourself against the network lines, but understand that the party who moves first is the one reading your statement every month.

Take an $85 order carrying $34 of goods, $8 of shipping and $2.77 of processing fees already paid, plus a $15 chargeback fee and 45 minutes of somebody's time at $32 an hour. The original settlement and the chargeback debit cancel each other out, so the sale amount is not a cash item twice over. What genuinely leaves the business is the goods, the shipping, the processing fee the processor keeps, the chargeback fee and the labour: 34 plus 8 plus 2.77 plus 15 plus 24, which is $83.77. That is 99 cents on every dollar of the order, gone.

You will read almost everywhere that a chargeback costs two to three times the transaction value. That is not wrong, it is a framing choice, and it is worth watching somebody do it in the open. Add the $85 sale you no longer have to the $83.77 that left the bank and you get $168.77, which is 1.99 times the order. Nudge the labour estimate up or add a write-off and you reach 2.5. The multiple is real, but it turns entirely on whether the reversed sale is counted, and nobody who quotes it says which convention they used.

The figure that changes behaviour is what it takes to get back to even. At $85 with $44.77 of variable cost, this merchant keeps $40.23 of gross margin per order, a 47.33 percent margin. Replacing $83.77 of cash at that margin takes $176.99 of new revenue, which is 2.08 more orders. At 55 chargebacks a month that is about $4,607 of cash a month and roughly $55,288 a year, and the chargeback fee line inside that monthly figure is just $825. The fee is the smallest item in the calculation and the only one most merchants shop on.

Fees vary, though not in the direction you would guess. Square charges nothing for disputes and covers the cost of challenging them. Stripe takes a $15 dispute received fee it never returns, plus a further $15 if you counter, which comes back only if you win. PayPal charges $20 on a card chargeback whether or not the buyer wins, and either $15 or $30 on a PayPal checkout dispute depending on your own dispute ratio. Helcim charges $15 only when you lose. The spread on a single event is about $30 against a total cost near $84.

Under 0.50 percent you are clear on both networks and your time is better spent elsewhere. Between 0.50 and 0.90 percent, start instrumenting: split disputes by reason code, by card brand, and by whether they are fraud or service failures, because the fix for friendly fraud and the fix for a delivery problem have nothing in common. Between 0.90 and 1.50 percent you are inside the range most acquirers quietly watch. Above 1.50 percent you are over Mastercard's ECM ratio and over Visa's US merchant line, and whether either program names you comes down entirely to your counts.

Visa's fact sheet carves two things out of the VAMP ratio: disputes resolved through pre-dispute solutions, and TC40 fraud that qualifies for Compelling Evidence 3.0. That exclusion is why Rapid Dispute Resolution and the alert networks get sold so aggressively. Read what you are buying, though. An RDR resolution is a refund: you lose the sale and the goods, you simply keep the event out of the ratio. It buys ratio, not money. Mastercard counts chargebacks received, so a refund issued before the chargeback posts keeps it out of that count too, the same trade in a different wrapper.

Fighting a dispute and winning does not remove it from either ratio. Visa's published exclusions are pre-dispute resolutions and CE3.0-qualified fraud reports, and a dispute you contested and won is on neither list. Mastercard counts chargebacks received, and you received it. Representment gets your money back and leaves your ratio precisely where it was. It is probably the most expensive misunderstanding in this area.

Finally, budget for the lag. Mastercard's denominator is the preceding month, and both networks report a month behind, so a fix you ship today lands in a report your acquirer reads 30 to 60 days from now. If you are near a line, the deadline you are working to is roughly six weeks earlier than the one printed on the letter.