What should a pre-deal check on an oil trade counterparty actually test?

A pre-deal check is not a name screen. Screening a name against the OFAC SDN List answers exactly one question, whether that name is designated, and it answers nothing about whether the entity holds the position it claims to hold. The four inputs that carry the diagnostic weight, counterparty, product, corridor and payment structure, each interrogate a structurally different failure mode, and together they operationalise what FATF Recommendation 10 actually asks for: not just identification, but an understanding of the purpose and intended nature of the relationship. An originator who knows which field tests what can read an inbound LOI, ICPO or draft MT700 as a set of testable assertions rather than a document to be forwarded.

This is the ten minute read for the front office, before the reply goes out.

Field one: counterparty. It tests standing, not existence

Companies house registrations, a trade licence, a website with a Dubai or Houston address: none of these are evidence of standing. They are evidence of existence, which is cheap. The counterparty field is testing something harder, whether the entity is transacting in its own name and whether it occupies the commercial position it asserts.

The dominant failure mode here is first-party misrepresentation, not impersonation. Sanctions screening and adverse media are built to catch someone pretending to be another party. They are poorly suited to catching an entity that is entirely who it says it is and is simply not the seller. That distinction matters because the mandate chain is where the ambiguity lives. An inbound offer arrives with an NCNDA and an IMFPA attached, five intermediaries deep, each described as a "mandate" or "direct seller representative". None of them will appear on a bill of lading. None of them holds title. The commission layer cake is fully documented while the underlying title position is not documented at all.

The test questions follow directly. Who is the title holder at the moment of transfer. What is the documented relationship between the named seller and the producer, term supplier or storage operator. Is there an allocation or contract reference that can be verified upstream. Under the OFAC 50 percent rule, ownership by designated persons is a blocking issue whether or not the operating entity is listed, so beneficial ownership resolution belongs in this field too, not in a later stage.

Field two: product. It tests whether the named seller can plausibly lift the grade

The product field is a capability test on the counterparty, not a quality test on the cargo. The question is not whether the grade exists. It is whether anyone in that seller's stated position could realistically get access to it.

This is where EN590 10ppm ex Rotterdam does the most work in the fraud economy. The grade is real, widely traded, and instantly recognisable to a non specialist, which is precisely why it appears in offers from entities with no tank lease, no terminal relationship, no nominated tank number and no history of notice of readiness anywhere. Crude offers show the same pattern in a different register. Term barrels of a specific grade are allocated by producers to a known and finite list of term lifters. An unknown intermediary offering rolling monthly allocations of a term grade, on a twelve month evergreen with rollovers and extensions, is asserting an access position that the market structure does not support.

FATF trade-based money laundering typologies catalogue falsely described goods as a core mechanism, and the same descriptive slippage shows up here in a pre-trade form. The tell is rarely a spec that is technically impossible. It is a spec that is entirely ordinary, paired with a seller who has no route to it.

Field three: corridor. It tests routing logic and sanctions exposure

Sanctions exposure in refined products and crude sits in the route far more often than in the name. Load port, transshipment, discharge and the vessel's behaviour between them are the exposure surface. OFAC's 2020 advisory for the maritime industry set out the indicators that have since become standard supervisory expectations: AIS transmission gaps, ship-to-ship transfers outside recognised transfer areas, falsified cargo and vessel documents. Dark fleet activity is defined by these behaviours, not by a designation that may or may not exist yet. Under the G7, EU and UK price cap regime, the attestation and recordkeeping model puts the burden of route knowledge directly on service providers, which means a corridor that cannot be explained is a corridor that cannot be attested.

The second half of the corridor test is plain commercial coherence. Does the stated origin to destination movement make economic sense given freight, quality and the prevailing arb. Unfamiliar corridors are not inherently suspicious, but they need a reason.

That is where current structure matters, lightly. With Brent at $86.77, down $0.30, WTI at $80.86, down $0.39, and Dubai at $84.77, a Brent-Dubai EFS near $2.00 per barrel keeps Atlantic Basin grades, WAF, CPC and Forties, competitive into Asia, while Brent-WTI near $5.91 sustains USGC exports into Northwest Europe and the Mediterranean. When arbs move, genuinely unfamiliar corridors appear in the inbox, and the volume of legitimate anomalies rises. That is exactly when the corridor field earns its keep, because it forces the originator to state the economic logic rather than assume it.

Field four: payment structure. It tests whether the instrument matches the trade

Documentary fraud lives in structure mismatch. UCP 600 article 5 is unambiguous that banks deal with documents and not with the goods, services or performance to which those documents relate, and article 34 disclaims bank liability for the genuineness or legal effect of any document presented. Those provisions are the operating logic of documentary credits and they are also, read the other way, the exact aperture that documentary fraud exploits.

The payment structure field asks whether the instrument named is the instrument the trade would actually require. A SWIFT MT700 is the issuance of a documentary credit. An MT760 covers guarantees and standbys. An MT799 is a free format message and is not an instrument at all, which is why "we will send an MT799 to activate" belongs on the anomaly list rather than in the workflow. Other mismatches are equally legible once you look for them: a documentary credit for a prompt cargo written with long usance and no bill of lading requirement, a non-operative credit whose operative clause depends on a condition the buyer cannot control, a performance bond demanded from the buyer before any instrument is issued, or an IMFPA commission layer that settles ahead of the underlying trade.

Advance fee patterns cluster here too. Tank inspection fees, dip and pay arrangements requiring buyer funds at the terminal before verification, and proof of funds demands escalating in sequence are all structure level tells rather than name level ones. ICC guidance including ISBP 821 gives examiners the reference for what a compliant presentation looks like, which is also the reference for what a non-compliant proposal is quietly avoiding.

Why a thin output usually means a thin input

If the front office submits a name and nothing else, no system, ours included, can return more than a screen. The pre-deal check that OilFlow Intelligence is built around takes these four inputs specifically because each one opens a different line of inquiry, and the output resolution is bounded by the input resolution. A read that comes back sparse is usually not a clean counterparty. It is an under-specified submission.

The corollary is the useful part. The output should be read as an operational instruction, not a traffic light. Each field failure maps to a document request: a standing failure asks for the title chain or allocation reference, a product failure asks for the tank nomination or lifting agreement, a corridor failure asks for vessel particulars and voyage history, a structure failure asks for the draft instrument text before anyone discusses price.

What compliance teams should do

• Mandate all four fields at intake. Counterparty, product, corridor, payment structure. Reject single field submissions back to the originator rather than screening them, so the front office learns that a name alone produces nothing usable.
• Separate standing from identity in your CDD file. FATF Recommendation 10 requires understanding the purpose and nature of the relationship. Record the counterparty's asserted commercial position and the evidence for it, not just the registry extract and the SDN screening result.
• Treat grade plausibility as a counterparty control. Ask what access route supports the offered grade. Where EN590 or a term crude allocation appears from an unknown intermediary, require the terminal or supplier relationship before proceeding.
• Route every corridor through sanctions and coherence checks. Apply the OFAC maritime advisory indicators, AIS gaps, irregular STS, document irregularities, and require the originator to state the arb or commercial rationale for unfamiliar routings.
• Escalate instrument mismatch before price discussion. Any reference to an MT799 as an activating instrument, a non-operative credit, or a buyer side bond preceding issuance should trigger MLRO review, not a negotiation.
• Log the mandate chain and the commission structure. Layer cake IMFPA arrangements that settle ahead of the underlying trade are a documented pattern, and they belong in the file whether or not the trade proceeds.
• Keep the record even when the deal dies. Abandoned approaches are intelligence. The same structures recirculate under new entity names.

If you want to see how the four field read is constructed and what each field returns, request a walkthrough. For weekly typology briefings written for MLROs and trade finance compliance teams, subscribe to the Forensics Friday newsletter.