Pre-Deal Clearance vs KYC in Oil Trading: Which Control Fires First on an Inbound Crude Offer?

Pre-deal clearance and KYC are sequential controls, not substitutes. A pre-deal clearance read is a triage pre-screen that fires at inception, the moment an unsolicited ICPO, LOI or EN590 offer lands in an originator's inbox, and it answers exactly one question: is this counterparty and this structure worth another hour of desk time? A full seven-step KYC dossier is a different instrument entirely. It is the regulator-grade evidentiary record that supports onboarding and transaction approval under FATF Recommendation 10, OFAC SDN list screening, US Bank Secrecy Act obligations, the UK Money Laundering Regulations 2017 and the EU AML framework. Collapse the two and you get one of two failures: a desk that runs full customer due diligence on inbound noise, or an MLRO holding a fast verdict where a defensible file should be.

The two controls answer two different questions

Most desks do not have a diligence problem. They have a sequencing problem.

The pre-deal read exists to protect originator attention. Its output is a directional verdict on whether an approach is worth advancing, produced fast enough that the originator has not already invested a morning in a mandate chain that dissolves on contact. It is a filter, and filters are judged on throughput and false-negative discipline, not on evidentiary completeness.

The KYC dossier exists to protect the institution. Its output is a file: identification and verification of the counterparty, beneficial ownership to natural persons, sanctions and PEP screening including OFAC SDN and consolidated EU and UK lists, source of funds and source of wealth, documentary authentication, vessel and routing exposure, and a written risk rating with the rationale that produced it. That file is what a supervisor, a correspondent bank or a trade finance credit committee reads two years later when they ask why you transacted.

One answers should we look? The other answers can we transact, and can we evidence why? Neither substitutes for the other, and the order is not negotiable.

What a sub-30-second pre-deal verdict is built to produce

OilFlow's pre-deal clearance read is designed to return a verdict in under thirty seconds. That constraint is deliberate and it is a workflow constraint, not a compliance claim. The value is that the verdict lands before the originator has replied, before the NCNDA is signed, before the desk has entered into an implied conversation it now feels social pressure to conclude.

What the read produces is a first-pass position on structural coherence. Does the claimed allocation route make commercial sense against the stated grade and delivery basis? Is the mandate chain plausible or is it a layer cake of introducers, each claiming authority they cannot document, sitting between the approach and a refinery that has never heard of any of them? Are the named entities, vessels and jurisdictions carrying obvious sanctions nexus? Does the paperwork sequence match how the transaction would actually finance, or does it invert the order, demanding a DLC MT700 or proof of funds before any verifiable seller identity is established?

That is triage. It is not customer due diligence, and it should never be filed as such.

What the seven-step dossier is built to produce

The seven-step KYC dossier is the opposite instrument. It is slow on purpose, because the thing it produces is durability under scrutiny.

Its steps run across the evidentiary domains that FATF Recommendation 10 sets out as the core of customer due diligence: identifying the customer and verifying identity from reliable, independent source documents; identifying beneficial owners and taking reasonable measures to verify them; understanding and, where appropriate, obtaining information on the purpose and intended nature of the business relationship; and conducting ongoing due diligence across the relationship. Layered on top of that are the sanctions obligations that sit outside the CDD framework and operate on strict liability logic, principally OFAC SDN and sectoral screening for US-nexus exposure, alongside UK and EU designations.

For physical crude and products, the dossier extends into domains a generic KYC template does not reach. The mandate chain must be evidenced, not asserted, with authority traced document by document from the approaching party back to a title holder. Vessel exposure has to be assessed, including AIS behaviour, ownership and management history, flag changes and any pattern consistent with dark fleet operation. Document authentication has to cover the artefacts that fraudulent approaches lean on hardest, from tank storage receipts to SGS reports to bank instruments.

The dossier's defining property is that it is legible to someone who was not in the room. That is what makes it an audit record.

The failure mode runs in both directions

Desks get burned two ways, and both are workflow errors rather than diligence errors.

Failure one: KYC as triage. The desk routes every inbound approach into full due diligence. Compliance becomes the bottleneck, queues build, and originators start informally pre-filtering by instinct to keep their pipeline moving. The unlogged instinct filter is now the real first control, and it produces no record at all. Worse, the genuine counterparties in the queue experience the same delay as the noise, so commercial pressure builds to expedite selectively, which is precisely the discretion a well-constructed approach is engineered to exploit.

Failure two: the fast verdict as the file. A pre-screen returns a clean directional read. The deal advances. Nobody goes back and builds the dossier, because the clean read created a felt sense of clearance. Six months later the MLRO is asked to evidence the CDD that supported onboarding, and what exists is a triage output that was never designed to carry that weight. The gap is not that the screening was wrong. The gap is that the institution cannot demonstrate what it did or why, which is itself the regulatory finding.

The second failure is the more dangerous one because it feels like efficiency right up until the moment it does not.

Why the current market raises the cost of a slow first filter

Sequencing matters more when the market is moving. Brent at $92.25 against WTI at $85.40 leaves a Brent-WTI arb near $6.85, and Dubai at $90.25 puts the Brent-Dubai EFS around $2.00. Spreads at those levels generate genuine arbitrage traffic, and genuine traffic is exactly the cover that fraudulent approaches need. When real cargoes are moving on real economics, an implausible offer looks less implausible.

At the same time, renewed US-Iran escalation and continued Houthi activity in the southern Red Sea mean sanctions nexus and vessel routing risk are live variables, not background conditions. Rerouting, ship-to-ship transfer activity and dark fleet participation all increase in exactly the conditions the market is presenting. A desk without a fast, consistent first filter will either slow down and miss legitimate flow, or speed up and admit approaches it has not structurally examined.

The sequencing model, stated plainly

1. Inbound arrives. Unsolicited offer, ICPO, LOI, introducer email, broker forward.
2. Pre-deal clearance fires. Sub-thirty-second directional verdict. Outcome is advance, decline or escalate. Log the verdict and the reason. It is a triage record, not a CDD record.
3. If advance, KYC opens. The seven-step dossier is built against FATF Recommendation 10, applicable OFAC SDN and UK/EU sanctions obligations, and the domestic regime, whether UK MLR 2017, the EU AML framework or BSA requirements.
4. Risk rating and MLRO sign-off. Enhanced due diligence where the risk assessment requires it. Written rationale attached.
5. Transaction approval, then ongoing monitoring. The relationship file stays live, not archived.

If your workflow diagram cannot show which of those steps produced which artefact, the sequencing is not implemented, it is assumed.

What compliance teams should do

• Write the two controls into policy as separate objects with separate owners, separate outputs and separate retention treatment. Ambiguity in the policy becomes ambiguity in the file.
• Instrument the triage layer. Every pre-deal verdict should generate a logged reason, even a decline. Declines are intelligence, and an unlogged filter is an invisible control.
• Never permit a triage output to satisfy a CDD requirement. Make the system enforce this. If the dossier is not built, the counterparty is not onboarded, regardless of how clean the first read was.
• Test the mandate chain as a discrete control. Require documentary authority at each link. Treat an unevidenced layer cake as a decline trigger, not a follow-up item.
• Refresh vessel and sanctions screening at transaction time, not just at onboarding. Designations change. Dark fleet ownership structures change faster.
• Run a file audit. Pull ten recently onboarded counterparties and ask whether a supervisor reading the file cold could reconstruct the decision. If not, you have a gap regardless of outcome.

If you want to see what each control produces side by side, the pre-deal verdict and the full seven-step dossier, book a walkthrough with OilFlow Intelligence. For ongoing typology teardowns aimed at MLROs and trade finance compliance teams, subscribe to the OilFlow Intelligence briefing.