Money laundering

Purchases concentrated at one merchant

A cardholder or wallet holder spends most of their money, for significant amounts, at one merchant with no apparent reason: a sign of fake sales, a colluding merchant, or a card being turned into cash.

A card or wallet holder spends most of their money, for significant amounts, at one merchant Most of the money, for significant amounts Card or wallet holder Merchant One merchant Merchant

What it catches

A customer whose card or wallet spending goes mostly to one merchant, for amounts that matter, with no reason that explains it. The purchases may not be purchases at all. A merchant in on the scheme can charge the card for goods that never change hands and hand back cash, keeping a cut; a front business can put criminal cash into the financial system as card sales; or the customer can drain a credit line, a stolen balance or a prepaid card through a shop that pays out. Visa’s rules speak to the first two. “A Merchant must not provide cash to a Cardholder unless the Merchant provides Cash-Back or Cash-Out or sells foreign currency” (Visa Core Rules 1.5.5.6, which “does not apply” in South Africa and has an exception in Nepal for merchants licensed to sell foreign currency). And Visa may bar a merchant for “Presenting Transaction Receipts that do not result from an act between a Cardholder and a Merchant or Sponsored Merchant (laundering)” (1.9.1.4).

Argentina’s UIF names the pattern among the circumstances to weigh when setting alerts: “Alta concentración, sin causa aparente, de compras de un cliente en un mismo establecimiento adherido, por montos significativos” (high concentration, with no apparent cause, of a customer’s purchases at the same merchant, for significant amounts). The Latin American FATF-style body, GAFILAT, describes front merchants that run fictitious card sales through their terminal for foreign cardholders, who receive cash in exchange. The Wolfsberg Group warns card issuers about purchases at a merchant “significantly out of pattern with historical spending behavior”, and acquirers about “collusive merchants” and “charging for non-existent goods”.

The rule finds the shape, not the crime. Plenty of people spend most of their money in one place: a supermarket near home, a fuel station on the way to work, a biller they pay every month. An analyst decides which is which.

How the rule works

For each customer, add up purchases over a look-back window, per merchant. Alert when the spend at one merchant is large in money (above a minimum for the customer’s segment), a large share of everything the customer spent, and made of more than one purchase. The share is what separates concentration from a simple big spender; the minimum keeps out customers who made one small purchase, where any merchant is 100% of their spending; the count keeps out one large purchase, which is a large-transaction rule’s case.

Looks at

Every completed card or wallet purchase a customer makes, grouped by the merchant it was paid to.

Measures

  • Spend at the merchant: the customer’s completed purchases at this merchant in the look-back window, less refunds from it.
  • Total spend: all the customer’s completed purchases in the same window, at every merchant the rule looks at, less refunds.

Both totals are converted to one currency at a documented rate, and the merchant is identified by a stable merchant id, not its name.

Leaves out

  • Merchants where paying one place is the product: billers and utilities, rent and school fees, loan repayments, and loads of the customer’s own wallet or account held elsewhere, unless you review those separately.
  • Advertising (merchant category 7311), unless you review it as a segment of its own: a customer who sells online puts most of their card spending into one ad platform.
  • Your own and test merchants.

Alerts when

All three of these are true:

  1. Spend at the merchant is at least your minimum for the segment.
  2. Spend at the merchant is at least your share line of total spend.
  3. The customer made at least your count of purchases at the merchant in the window.

How often

At most one alert per customer and merchant per look-back window.

Highlighted values are settings you choose. Hover over one, tap it or tab to it for a suggested value.

Common variants:

  • Repeated payments to one merchant. Count rather than share: several payments from the same customer to the same merchant in a short window, adding up to a significant total. It catches a card run repeatedly through a terminal when the customer has little other spending to compare with.
  • One card dominates a merchant (the acquirer’s view). The same relation seen from the merchant: a large share of a merchant’s takings comes from one card or a handful of cards, or the same card is used many times. The OCC lists “multiple use of the same cardholder number” among the basic parameters for merchant exception reports. A merchant with few distinct customers and large tickets is the strongest form. Tell cards apart by the Payment Account Reference or the processor’s card fingerprint, not by the full card number, and never by a token that changes per device or per merchant: a phone wallet’s tokens split one card into several and hide its share of the merchant.
  • Concentration at a cash-like merchant. The spend goes to a casino, a gambling site, a money transfer or quasi-cash merchant, or a seller of gift cards or foreign currency, where what was bought can be turned back into money. In-app currencies (game credit, coins a viewer can give to a live streamer) belong here where the receiver can cash them out. GAFILAT’s casino case lists among its red flags “Utilización de varias tarjetas de crédito / débito para comprar fichas de casino” (using several credit or debit cards to buy casino chips) and “Realización de transacciones con tarjetas de débito y/o crédito, hasta su límite máximo” (debit or credit card transactions up to the card’s limit). Its straw man went to “diversos casinos” (several casinos), so watch the merchant category as well as the single merchant.
  • Purchase then refund. Spend at one merchant followed by refunds or credits from it that do not match a returned purchase, or by a refund of the card’s credit balance. Wolfsberg lists “merchant credits without offsetting merchant transactions” and refunds of credit balances among issuers’ red flags.
  • Foreign cards at one merchant. At an acquirer, a merchant whose sales suddenly come mostly from foreign cards, at round amounts. GAFILAT’s fictitious sales case lists “Ventas por fuera del segmento y que son realizadas en su mayoría con tarjetas extranjeras” (sales outside the merchant’s segment, mostly made with foreign cards) and “Operaciones realizadas por compras en cifras cerradas” (purchases in round figures).

Data it needs

ShowHide
FieldUsually comes fromIf it’s missing
Customer id and card or wallet idIssuer or wallet platformRequired. Supplementary cards should roll up to the account holder, so their spending can be compared with the holder’s. Where cards are counted (the acquirer’s view), key them on the Payment Account Reference or the processor’s card fingerprint.
Merchant idAuthorisation and clearing recordsRequired. Key on a stable id: merchant names change and are spelt several ways, and some merchants put a reference in the descriptor that changes with every charge, or the buyer’s account number on their platform. Where a payment facilitator or marketplace pays under one id for many sellers, the rule sees one “merchant” that is really many.
Merchant category codeAuthorisation recordsNeeded to leave out billers and advertising and to run the cash-like-merchant variant.
Amount and currencyAuthorisation and clearing recordsRequired. Convert to one currency before comparing.
Timestamp and statusAuthorisation and clearing recordsRequired. Count completed purchases; declined attempts are a different signal.
Transaction type (purchase, refund, cash advance, wallet load)LedgerNeeded to net refunds and to keep loads of the customer’s own accounts out of the purchases.
Customer segment (individual, business, corporate or fleet card)Onboarding (KYC/KYB)Needed to set the minimum. Without it, corporate cards that buy from one supplier dominate the alerts.
The merchant’s number of distinct customers or cardsAcquirer or your own dataOptional. Needed for the acquirer’s view, and to tell a busy shop from a merchant whose sales come from a few cards.
Card issuing countryAuthorisation recordsOptional. Needed for the foreign-cards variant.

Thresholds

Each row is a setting the rule needs. Each value is a starting point: test it on your own history, then adjust. Where there is no public value, set it from your own data.

ParameterStarting pointWhere it comes from
Share of the customer’s spend at one merchantNo public value; set from your own dataNo regulator or public rule list gives a share. Start above the top-merchant share most ordinary customers in the segment reach, with billers and advertising left out. Where most customers who spend a significant amount already put most of it in one place, the share adds little and the minimum and the leave-outs do the work.
Minimum spend at the merchantSet per segment and currency from your own spread of spendingArgentina’s text says “montos significativos” (significant amounts) without defining them. Required: without a floor, any customer with one purchase is 100% concentrated.
Purchases at the merchant2 or more, as a starting valueThe smallest count that tells concentration from one purchase. Raise it, or also cap the largest purchase’s part of the spend at the merchant, if one large purchase plus a small one keeps alerting.
Repeated-payments variant: payments to one merchant5 or more in 7 daysOne public rule list’s starting point for repeated payments from one customer to one merchant. Raise it for merchants customers use daily (transport, canteens, coffee shops).
Repeated-payments variant: total at one merchantFrom EUR 1,000 or its equivalent in 7 daysThe same public rule list’s floor. Set it per segment and currency: it is far too low for business cards and may be too high where incomes are low.
Acquirer’s view: payments from one card at one merchant3 or more in 24 hours, totalling from USD 1,000 or its equivalentOne public rule list’s starting point. Raise the count for merchants a customer pays several times a day (cafés, transport), and set the total per merchant category.

By country

  • Argentina. Among the circumstances that “deberán valorarse especialmente” (shall be given special weight) in setting alerts: “Alta concentración, sin causa aparente, de compras de un cliente en un mismo establecimiento adherido, por montos significativos” (high concentration, with no apparent cause, of a customer’s purchases at the same adhered merchant, for significant amounts) (UIF Resolution 200/2024, art. 34(b), circumstance 20). The list gives no role for each circumstance, but who counts as the customer depends on the firm’s role: the cardholder for the card issuer, and the payment account holder for the provider of the account, while “No será considerado Cliente del Adquirente, o del Agregador o Subadquirente, al titular de tarjeta” (the cardholder shall not be considered a customer of the acquirer, or of the aggregator or sub-acquirer), whose customer is the merchant (art. 2(f), points 1, 2 and 5). Read with that definition, this circumstance fits card issuers and payment account providers. For acquirers the merchant-side circumstance is the next one: “Saldo acreedor significativo en la cuenta del establecimiento adherido que no guarda relación con los antecedentes de su situación patrimonial” (a significant credit balance in the merchant’s account that bears no relation to its financial background) (circumstance 21). For cards, the automated rules must take account of joint and additional cardholders, “debiendo prestar especial atención al vínculo con el titular de la misma y la estructura de los consumos generados” (paying special attention to their link with the cardholder and the structure of the spending generated) (art. 34(a)). The resolution sets no number for “montos significativos”; the methodology for setting monitoring rules and parameters must be documented (art. 34(c)).

    Reporting deadline

    Money laundering, within 24 hours of concluding that the operation is suspicious, and no later than 90 calendar days from the operation, performed or attempted; terrorist financing and proliferation financing, within 24 hours of the operation (UIF Resolution 200/2024, art. 36(c)).

  • United States. For monitoring merchants (the one-card-dominates-a-merchant variant), not a customer’s spending: the OCC’s Comptroller’s Handbook on merchant processing, written for its bank examiners, says the primary tool for detecting merchant fraud is “an exception report that details variances from parameters established at account setup”, and that basic parameters “may include daily sales volume, average ticket size, multiple purchases of the same dollar amount, multiple use of the same cardholder number, the percentage of transactions keyed versus the percentage swiped, and charge-back activity” (p.28). It describes factoring as “a form of fraud in which a merchant creates false sales transactions, inflates the sales amount, or alters the sales drafts to receive funds from the issuer”, and as a way “to launder money via credit cards” (pp.10 and 81). The handbook is examiner guidance, not a rule.

    Reporting deadline

    No later than 30 calendar days after initial detection (31 CFR 1020.320 for banks, 1022.320 for money services businesses). A bank that has identified no suspect may take 30 more days to identify one, never more than 60 days in all (1020.320(b)(3)); the rule for money services businesses has no such extension.

False positives and tuning

ShowHide

Common innocent explanations: a customer who does most of their shopping at one supermarket or fuel station; a monthly bill, school fee or rent paid by card; a large one-off purchase (a car deposit, furniture, a holiday) that dwarfs the rest of the month; a corporate, fleet or travel card meant for one supplier; a new card that has only been used in one place so far.

Typical causes of misfires:

  • Customers with little spending. With one or two purchases in the window, any merchant is most of the total. A minimum spend at the merchant, and a minimum number of purchases there, remove them.
  • A book dominated by one merchant. Where one biller or one partner takes most of the platform’s volume, nearly every customer is concentrated there by construction. Leave that merchant out, or compare each customer’s share with the merchant’s share of all customers’ spending.
  • Advertising. A customer who sells online spends most of their card money on ads, often on one platform. Where many customers sell online, concentration on an ad platform is the norm, not the signal. Leave advertising out by default and, if you want to watch it, review it as a segment of its own with its own line.
  • A card kept for one use. Many people hold cards at several institutions and keep one for a single purpose: fuel, ads for their online shop, one subscription. The issuer sees that card’s spending, not the customer’s, and the card is concentrated by design. The share line alone cannot tell this from the pattern; the minimum, the leave-outs and what the customer declared do more.
  • Merchant identity. One merchant can appear under several ids (one per terminal or per store) or under several spellings and categories, and slip under the share line; a payment facilitator or marketplace can appear as one merchant for thousands of sellers and look like concentration. Some merchants put a reference in the card descriptor that changes with every charge, or the buyer’s account number on their platform: a rule keyed on the raw descriptor sees each charge as a different merchant and never finds concentration there. Map terminals, stores and descriptors to one merchant before counting, decide which level the rule measures (the platform, or the account on it), and treat marketplace ids as their own segment.
  • Buying for other people or for resale. A customer who buys stock for resale on a marketplace, clothes for friends or tickets for clients concentrates spending at one marketplace or booking site, and is often paid back by transfers from those people. The purchases are real. A business run through a personal account that the institution does not know about, or does not allow, still deserves review, and the third-party money is better seen by Third-party funding and payment account name mismatch and Peer-group outlier. A sole trader the institution allows to trade on a personal account, with the trade recorded at review, belongs in a segment of its own with its own line.
  • Own-account loads. Loading a wallet, brokerage or exchange account held in the customer’s own name looks like spending at one merchant. Exclude those loads by category or counterparty, and review cash-like merchants in their own variant.
  • Mixed currencies. Spend in several currencies summed without conversion distorts both the amount and the share. Convert to one currency at a documented rate.
  • One big ticket. A single large purchase is not concentration of purchases; it belongs to Large single transaction. Requiring more than one purchase at the merchant, or the repeated-payments form, keeps the two apart in part: one large purchase and one small one at the same merchant still pass a count of two. Also cap the largest purchase’s part of the spend at the merchant where that case keeps alerting.

How to tune: segment by customer and card type; leave out billers, advertising and own-account loads; key merchants on a stable id and map terminals, spellings and descriptors to it; set the minimum per segment and currency; measure the spread of top-merchant shares on your own customers before choosing the line; and cap alerts at one per customer and merchant per window.

When it fires

ShowHide

Analyst checklist:

  1. List the customer’s purchases at the merchant in the window and a few weeks around it: dates, amounts, round amounts, card present or keyed, and any refunds or credits from the merchant.
  2. Look the merchant up: what it sells, its category code, how long it has been accepting cards, and whether its sales fit what it says it does. For a wallet or acquirer, check how many other customers pay it and whether a few cards make most of its takings.
  3. Compare with the customer’s declared occupation, income and usual spending. Purchases near the card’s limit, or soon after the limit was raised, are stronger signals.
  4. Check what followed: cash withdrawals or transfers by the merchant soon after settlement, refunds to the card that match no returned goods, or the card’s credit balance refunded to the customer.
  5. Check whether other customers concentrate spending at the same merchant, and whether the customer’s cards are used by someone else (see Shared identifiers across customers).
  6. Ask the customer what they buy there and why so much, as part of ordinary due diligence and never in a way that reveals a suspicion or a report. Escalate when the purchases cannot be explained, the merchant cannot show the goods or services, or several unrelated customers show the same pattern at the same merchant.

An alert is a reason to review, not a suspicion in itself. Under the FATF standard, suspicious transactions are reported regardless of amount, attempted transactions included (Interpretive Note to Recommendation 20, para 3). Each country’s reporting deadline is under By country.

Sources

8 sourcesHide
  1. Unidad de Información Financiera (Argentina), Resolución 200/2024 (RESOL-2024-200-APN-UIF#MJ), 18 December 2024, Boletín Oficial no. 35571, 19 December 2024: art. 2(f) (who the customer is), art. 34(a) (automated rules; cards), art. 34(b), circumstances 20 and 21, art. 34(c), art. 36(c) (report deadlines). Supports the pattern and the reporting duty. https://www.argentina.gob.ar/normativa/nacional/norma-407429/texto (also on InfoLEG: https://servicios.infoleg.gob.ar/infolegInternet/anexos/405000-409999/407429/norma.htm)
  2. GAFILAT, Recopilación de tipologías regionales de GAFILAT 2009–2016, December 2016, typology 56, “Utilización de comercios pantalla para la colocación de capital ilícito en cuentas de ahorro o corrientes a través de venta [sic] ficticias por datafono a clientes con tarjetas internacionales” (first published in the 2011–2012 typologies report), description and red flags, pp.130–131. Supports the pattern and the foreign-cards variant. https://biblioteca.gafilat.org/wp-content/uploads/2024/04/Recopilacion-Tipologias-2010-2016.pdf
  3. GAFILAT, Informe de tipologías regionales GAFILAT 2014–2016, August 2016, typology 11, “Lavado transnacional de activos con tarjetas internacionales a través de casinos”, case p.36, red flags p.37. Supports the cash-like-merchant variant. https://biblioteca.gafilat.org/wp-content/uploads/2024/04/TipologiasRegionales-2016.pdf
  4. The Wolfsberg Group, Wolfsberg AML Guidance on Credit/Charge Card Issuing and Merchant Acquiring Activities, 2009: 5.7.2 (issuing, transaction monitoring red flags), pp.8–9; section 6 and 6.1 (merchant acquiring; collusive merchants), p.10. Supports the pattern and the refund variant. https://db.wolfsberg-group.org/assets/c2914086-4775-4110-860e-adac57371988/Wolfsberg%20Credit%20Card%20Guidance.pdf
  5. Office of the Comptroller of the Currency, Comptroller’s Handbook: Merchant Processing, version 1.0, August 2014 (reputation-risk references removed March 2025): factoring, p.10; “Fraud Monitoring”, p.28; glossary, “Factoring”, p.81. Supports the acquirer’s-view variant. https://www.occ.gov/publications-and-resources/publications/comptrollers-handbook/files/merchant-processing/pub-ch-merchant-processing.pdf
  6. Visa, Visa Core Rules and Visa Product and Service Rules, edition of 24 October 2026: 1.5.5.6 “Merchant Cash Disbursement Prohibition” (with its Nepal and South Africa exceptions), p.127, and 1.9.1.4 “Visa Right to Terminate Merchants…” (laundering), p.141. A card scheme’s own rules. https://usa.visa.com/content/dam/VCOM/download/about-visa/visa-rules-public.pdf
  7. FATF, The FATF Recommendations, October 2025 edition: Interpretive Note to Recommendation 20, para 3. https://www.fatf-gafi.org/en/publications/Fatfrecommendations/Fatf-recommendations.html
  8. 31 CFR 1020.320 (banks) and 1022.320 (money services businesses), reports of suspicious transactions. https://www.ecfr.gov/current/title-31/subtitle-B/chapter-X/part-1020/subpart-C/section-1020.320

Ask us

A question about this rule?

A person at Archer reads every question and answers by email.

The Fraud & AML Library is a practical reference, not legal advice. What binds is the regulator’s own text, which every rule and regulation page links.

Get new rules and regulation changes by email

When we add a rule or a regulation changes, we’ll email you.

You can unsubscribe at any time. Privacy policy