Published on 24 September 2019

It is surprising to find nothing on the agenda of SIBOS conference or the SWIFT Payments Market Practice Group about the recent court decision that negates vital content of the SWIFT Message Reference Guide and BIS and PMPG documents on the subject of MT202 COV.

The issue is the extent of the MT103 sending bank’s obligation to the MT103 receiving bank to come good on the cover, if the cover method of making the payment has been chosen.

Current wisdom – and the basis of current market practice – is that the MT103 receiving bank pays away at its own risk, if it does so upon receipt of the MT103 without having confirmed that it has received the cover on its nostro account and in irrevocable funds.

A senior UK judge – Jeremy Lionel Cooke – recently delivered his opinion on the matter, which confounds this wisdom, and states instead that it is the MT103 sending bank’s risk.

While the written evidence and the transcript of the hearing remain unpublished, the judgement is in the public domain.

Mr Cooke’s view is that the MT103 sending bank has an irrevocable and unconditional obligation to deliver the cover, out of its own funds if necessary.

This is even if, by way of example, an MT202 COV had been sent through New York correspondents, and it had failed AML/CFT checks at either the sender’s or receiver’s correspondent, and been put into a compliance queue, perhaps never to emerge. This might not be a False Positive, but a Real Positive.

It might also be when the MT103 sending bank had already sent a cancellation message to the MT103 receiving bank that it might reasonably have expected the latter to have recognised and acted upon, given the timing e.g. if both the 103 and the cancellation had reached a bank based in the Asia/Pacific region after 5pm local time and on a Friday afternoon. In that case it would be reasonable to expect that the receiving bank would have held both messages at its SWIFT gateway until Monday’s opening, and then released them both into the processing flow, at which point their payment application would link the two and present the MT103 together with its cancellation, such that both would be queued and put under investigation.

Not so, writes Mr Cooke. In either case where the receiving bank had paid away on the MT103, the MT103 sending bank would need to find a way to reimburse the receiving bank.

The implication of this judgment is that, if the cover was indeed blocked at New York correspondents and the payment was in USD, the MT103 sending bank should not send another cover payment in USD as that would be blocked as well, but that it would need to find either a different currency or a different method of reimbursing the receiving bank so that the cover did arrive.

In other words, the MT103 receiving bank should complete a payment that had been blocked for AML/CFT purposes, possibly justifiably – it would have no way of knowing at that stage whether the blocking was justified, and the compliance department of its correspondent would not inform it for fear of being liable for “tipping off”.

Aside from the fact that the MT103 sending bank would have made the second payment out of its own funds as the customer funds were blocked, the MT103 sending bank would be at risk if the payment then actually turned out to be a case of money laundering or for financing terrorism, because the bank would appear to financial regulators to be insisting on completing payments – with its own money – even where there was an indicator that the payment should not be completed.

We know what the penalties of doing that can be, both corporately and individually.

In consequence banks should eschew the cover method of making customer payments over SWIFT, and use the serial method exclusively, pending further test cases (which may be some years in coming) or successful challenge of this decision (which may not occur).