Occasioned by the recent Financial Standards Board reports on Correspondent Banking and on Remittances, I am of the opinion that Wolfsberg Group's 2016 “Wolfsberg Guidance on SWIFT Relationship Management Application (RMA) Due Diligence” should be withdrawn.
I have sent a paper to the Wolsfberg Group to this effect.
The guidance has given a clean bill of health to situations where RMA is in place in support of a correspondent banking relationship – what Wolfsberg calls a “customer” RMA.
But it has brought into question the other application of RMA – what is termed a “non-customer” RMA. The outcome of the Wolfsberg guidance on “non-customer” RMA has in my view been more detrimental to the marketplace than the materialisation of the risks
the guidance was meant to eliminate.
The rationale for my challenge is:
- Wolfsberg’s self-allocated remit is Anti-Money Laundering. They have not made a compelling case that the way in which banks handle RMA gives rise to AML risks. The guidance itself states that the risks are primarily operational. The matter is on principle
outside Wolfsberg’s remit;
- The definitions and use cases quoted in the guidance for “non-customer” RMA are flawed, which invalidates the guidance;
- The use cases do not include the one where “non-customer” RMA is most used and is of most value (or used to be): MT103 sent by the Cover method;
- Wolfsberg’s guidance has led to a decline in the ability of banks to operate normal international payment and trade operations with many other SWIFT members;
- SWIFT members – lacking RMA with which to exchange “value messages” directly - access one another over SWIFT through the intermediation of their correspondent banks, adding inefficiency, time and cost, and possibly precluding the transaction taking place
at all or driving it off SWIFT and into non-transparent networks.
All Wolfsberg Group’s 13 members figure in the Financial Stability Boards list of the world’s 30 Global Systemically Important Banks. Wolfsberg’s focus is on the business of correspondent banking. Indeed, all its members except Goldman Sachs are leading
correspondent banks. Wolfsberg guidance has a meaningful impact on market practice, and on the access of other banks and their clients to international banking services.
The SWIFT RMA may seem like an obscure topic, but it is the mechanism that controls the ability of one SWIFT member to send messages across the SWIFT network to other SWIFT members. The Wolfsberg guidance raises particular concern where SWIFT members send
messages to other SWIFT members when they are not correspondent banks of one another: this is called a “non-customer” RMA.
Wolfsberg comments that such arrangements give rise predominantly to an operational risk but that “poorly controlled “non-customer” RMAs may give rise to a risk of Money Laundering in institutions with inadequate AML systems and controls”.
This Delphic statement has been enough to cause banks to eliminate “non-customer” RMAs wholesale. Wolfsberg’s own definition of a “non-customer” RMA is one which I would take considerable issue with, but it is in the five use cases that the basis of Wolfsberg’s
guidance is undermined.
The first one – exchange of information messages in the MT9nn series - does not require RMA at all.
The second one, applying to the exchange of the MT101 message, describes only one very basic variant of the multibank cash management service which MT101 supports. To offer an enhanced version of the service, banks do hold accounts with one another, and
then their RMA is a “customer” RMA.
The third use case describes a situation between a global custodian and their sub-custodian: I do not believe these relationship are set up as "network banks" in the way in which Wolfsberg terms it, without relationship managers, credit lines or cash accounts.
In my experience the relationship between a global custodian and its sub-custodians is extensive and necessitates that they have “customer” RMAs.
The fourth one, on the usage of the MT7nn message series for Trade Finance, is actually two situations, one of which does involve “non-customer” RMA and one of which does not.
The fifth use case – to do with exchanges with Financial Market Infrastructures - is like the third one: contrary to Wolfsberg’s assertion, cash accounts - and usually relationship managers and credit lines as well - are involved so the parties will have
a “customer” RMA in place.
Three of Wolfsberg’s five use cases fall away because they are not exemplars of “non-customer” RMA. Then there is the missing use case and the most important: MT103 sent by the Cover method.
Wolfsberg’s guidance has led directly to the collapse of the Cover method of making international payments. Now they are predominantly made via the Serial method, with a longer payment chain, more parties requiring fees, and less certainty as to what information
on the payment reaches the beneficiary. End-to-end timing, value-dating, and all-in fees cannot be better for the end user under the Serial method.
The three situations where “non-customer” RMA plays a key role correlate to meeting important needs of international banking customers are:
- MT103 by the Cover method: getting a direct payment order from the Account With Institution at the start of the payment chain to the Beneficiary Institution at the end of the chain and very quickly;
- MT101: enabling a large corporate customer to incorporate all their banks into an International Cash Management scheme and not just the banks that the scheme arranger has preferred links with – which are often its own branches and subsidiaries;
- Trade finance: enabling exporters and importers to enjoy risk mitigation (Country and Bank Risk) on cross-border trade transactions because more banks can be involved than just the main bank of the importer and that bank’s cash correspondent in the country
of the exporter.
In my view this Wolfsberg guidance on RMA has had a major and detrimental impact on the market supply of international banking services. Authorities - most recently the Financial Stability Board - see the shrinkage of supply, are very concerned about it
and attribute it to general de-risking. Authorities have been far less adept at identifying its underlying causes.
Wolfsberg’s remit is to combat Money Laundering and to counter the Financing of Terrorism. They have made no case that banks’ methods of managing RMA meaningfully increase or reduce AML/CFT risks. Indeed they state themselves that the risks are primarily
operational. The response to their guidance, however, has been detrimental and, in my view, wholly disproportionate to the cases of AML/CFT risk that may have been mitigated by it. That is why it should be withdrawn.