Trade-based money laundering: what a payment firm can see, and what it must do

Trade-based money laundering is the typology a payment firm is least equipped to see. The payment itself looks ordinary. A business client pays a supplier abroad, an invoice sits behind it, the amount matches, and the beneficiary is the named supplier. Nothing in the instruction is false. The laundering sits in the trade: in a price that is wrong, a quantity that never shipped, or goods that do not exist.

This article sets out what the typology is, which techniques appear in cross-border business payments, what a payment firm can verify and what it cannot, which red flags are worth acting on, and how the control is built so it survives a review. It is general information, not legal advice.

What is trade-based money laundering, and how does it differ from other typologies?

The Financial Action Task Force describes it as the process of disguising the proceeds of crime and moving value through trade transactions, to give those proceeds a legitimate appearance. The reference text is the joint FATF and Egmont Group report Trade-Based Money Laundering: Trends and Developments, published in December 2020.

The difference from other typologies is where the lie sits. In a simple placement case the money is dirty and the paperwork is thin. Here the money often moves through the banking system in plain sight, with an invoice, a contract and a transport document behind it. The documents are the instrument. Value crosses a border because the price on the paper does not match the value of the goods.

That is why the FATF and Egmont report treats it as one of the hardest methods to detect. A bank sees a payment and a document set. It does not see the warehouse, the container, or the market price of the commodity on the day it shipped. The same gap applies to a credit union serving business members and to a money services business processing supplier payments.

Which trade-based money laundering techniques appear in a cross-border business payment?

Four cover most cases, and all four are forms of misinvoicing.

Over-invoicing moves value to the seller. The buyer pays more than the goods are worth, and the surplus lands abroad as a trade receipt that looks earned. Under-invoicing does the reverse. The buyer receives goods worth more than the sum paid for them, so value moves to the importer.

Multiple invoicing uses one shipment to justify several payments. Each payment is defensible on its own. Only a view across the whole relationship shows the same goods paid for twice. Short shipment and phantom shipment change the quantity rather than the price: fewer goods arrive than the invoice describes, or nothing ships at all.

Two structural variants matter as much as the arithmetic. In a black market peso exchange arrangement, drug proceeds in one country are settled against trade payments in another, and FinCEN addressed suspicious activity reporting on that pattern in its February 2010 advisory FIN-2010-A001. The second variant is routing. Goods pass through a third country, so the paperwork no longer matches the economic route, and the payment corridor stops lining up with the trade corridor.

Only multiple invoicing is visible from a single payment. The other three need either a price benchmark or a view of the relationship over time.

What can a payment firm actually verify, and what is out of reach?

Be honest about this boundary in the policy, because a reviewer will test it.

Some things sit in the firm's own records. Who the client is, who owns it, what it says it buys and sells, which counterparties it pays, in which currencies and corridors, and whether that pattern matches the business described at onboarding.

Other things are checkable with effort. Whether a stated unit price is plausible against a published benchmark. Whether a counterparty exists at the address given. Whether the goods described need an export licence. Whether a transport document names a real route and a real carrier.

One thing is out of reach. What was in the container. A payment firm is not a customs authority, it cannot inspect goods, and its policy should not imply otherwise. The control is about plausibility and consistency, not proof of the underlying trade.

Expected activity is the anchor. A client that described itself at onboarding as importing machinery from one country in one currency, and then begins paying textile suppliers in three new jurisdictions, has changed. Whether or not that change is laundering, it is a change the firm should have noticed and asked about.

Which red flags belong in monitoring, and which are noise?

FATF published Trade-Based Money Laundering: Risk Indicators in March 2021. It is a long list, and reading it as a checklist is the common mistake. Most indicators are weak alone and strong in combination.

The combinations worth a hold: an invoice value far outside the normal range for that commodity and corridor; a payment to a jurisdiction unrelated to the buyer, the seller or the goods; round sum invoices repeating at a cadence no real supply relationship produces; a counterparty incorporated weeks before its first invoice; goods described so vaguely that no price can be tested; and a client whose declared sector has no plausible use for the goods on the paperwork.

Weak alone: a higher risk jurisdiction, a single round number, a new supplier. Each is a scoring input. None justifies stopping a payment by itself.

The tuning decision matters more than the list. An engine that fires on any single indicator will bury a small compliance team and teach it to clear alerts quickly, which is the worst outcome available. Alert on the combination, score the rest, and review the thresholds against real outcomes.

The invoice is the only part of a trade transaction a payment firm ever holds. Treat it as a claim to be tested against the client's own declared business, not as evidence of what shipped.

How should a payment business build the control, step by step?

The sequence below produces the evidence file as the work is done, which is the only version that survives an examination.

• Record expected trade activity at onboarding: what the client buys and sells, to and from which countries, in which currencies, at what typical values, and how often. Without this baseline nothing later can be called anomalous.
• Capture the trade documents with the payment instruction, not in a separate folder. The invoice, the purchase order and the transport document where one exists, each linked to the payment reference.
• Test the invoice for plausibility rather than authenticity: unit price against a published benchmark where the commodity has one, quantity against the stated transport, and the goods against the client's declared sector.
• Screen the goods as well as the parties. A counterparty that clears sanctions screening can still be shipping a dual-use or export-controlled item, and that is a separate check with a separate list.
• Monitor across the relationship, not the single payment, so repeated invoices for one shipment and a drift away from expected activity both surface.
• Write the escalation rule before the first alert: who may hold a payment, what evidence releases it, who decides when the evidence is thin, and how long a client waits for an answer.
• Report on suspicion, not on proof. FATF Recommendation 20 requires a report where suspicion exists, and a payment held and then released still needs the reasoning recorded with a date and a name.

One more point on documents. Asking a client for an invoice and filing it unread is worse than not asking, because the file now shows a control that was never run.

Does a stablecoin settlement leg change the trade-based money laundering risk?

It changes the evidence and the timing. It does not change the typology. Misinvoicing is a lie about goods, and moving the settlement leg onto a public blockchain does not make the goods real.

Three things improve. The payment record becomes exact: the asset, the amount, both addresses and a timestamp, visible to both sides at once rather than reconstructed from a correspondent statement days later. The invoice reference travels with the instruction, so one invoice maps to one settled transfer, which is what makes multiple invoicing visible at all. And ISO 20022 structure carries the detail, with pacs.008 for the customer credit transfer, pacs.009 for the interbank leg, structured remittance information for the invoice reference, and a single UETR joining the chain end to end.

Two things get harder. Settlement is fast and final, so a hold has to be placed before release rather than recalled afterwards. And the counterparty wallet becomes a screening object in its own right, alongside the company and its owners.

The reviewer's question does not change. Show me one payment, the invoice behind it, the test you ran, what you concluded, and who decided.

Where StableNet fits

StableNet, built by SpendTheBits, is a cross-border B2B payment and settlement platform for banks, credit unions, licensed money service businesses, exchange houses and remittance fintechs. Settlement runs in regulated stablecoins on public blockchains and completes in minutes with on-chain auditability, so each supplier payment carries its own record rather than a reconstructed one. KYB and KYC onboarding, KYT, sanctions and PEP screening, FATF Travel Rule data and a tamper-evident audit trail sit inside the platform, and every transfer is ISO 20022 native, carrying pacs.008 customer credit transfers and pacs.009 interbank legs tracked by UETR. Judging whether a trade is real stays the institution's own work. Proving what was paid, to whom, against which reference, no longer has to be assembled after the fact.