Sibos 2019, held in London, was the biggest ever – 13,000 attendees – and entertaining as always. Payments and cyber-security were top of mind and talk of DLT is receding in favour of solutions. And as always, surprises abound.

Resilience from the shark pool

Wandering around Innotribe in a quiet moment, I came across Paul de Gelder explaining how he got over losing half his limbs in a shark attack in Sydney Harbour – not only recovering and thriving but also applying his personal resilience to campaigning for planetary resilience. He definitely put into perspective the normal issues we treasurers think of when discussing overcoming challenges.

Payments

With faster payments progressing at different speeds around the globe, payments were naturally high on the agenda. The Bank of England will be rolling out its next generation payment platform over the next few years. They have left DLT for the next upgrade cycle. Their RTGS vision includes:

• Increased resilience.
• Strengthened end-to-end security.
• Improved user functionality.
• Wider interoperability.
• Greater access.

It is interesting that two big themes surface in the Bank of England vision. First: resilience, which is clearly a prime responsibility of central banks and regulators everywhere. Second: access, which echoes the open banking being promoted across the Eurozone and elsewhere. At some level there is a tension between the two, because superficially, at least, more open systems can be more vulnerable.

This tension was a major theme at Sibos, engendering sessions on a wide array of topics, from securing APIs to new forms of biometrics, harder encryption, and training. We even had a session on secure coding for non-coders that was very enlightening.

It seems that the technology is reasonably robust and the bug bounty programmes keep the code security abreast with hacker’s creativity, so the big risk is the wetware. As one panellist put it: “Every organisation has their ‘Dave’” – the person who uses their birthday as a password and clicks on phishing links.

It’s not just errant Luddites who pose a threat. One panellist cited statistics showing that only 3% of banks have activated multi-factor authentication. That is truly shocking, if correct.

Cacophony of payments

Solving these issues will require plenty of experimentation. Corporates lament the complexity and diversity we already face when faster payments co-exist with legacy batch systems, not to mention wallets and various mobile solutions like Zelle (a bank created counter to Venmo in the USA). Each one of these has different APIs and file formats as well as different cut-offs (or none) and operating parameters.

For most, this diversity of payment rails represents a challenge impeding centralised straight through processing of large payable volumes. To some, diversity is a business opportunity in so far as solution providers who can mask the complexity can present valuable services. Yet others accept diversity as a reality that must be handled to ensure the widest possible access for their products and services.

One panellist opined that there will be two (cross-border) payment rails in the (unspecified) future – one for China and one for the rest of the world. But most agreed that the near-term promises much more diversity than harmonisation. Even within the relatively coherent confines of the Eurozone, it seems that every bank is creating its own API formats and processes. Meanwhile in the USA, a bank consortium called The Clearing House (THC) is slugging it out with the Fed to determine the future of faster payments; THC is already live while the Fed has deep research into market needs.

Ultimately payments need to become free as email has become, and indeed as are retail payments in many markets. An interesting anecdote from the early years of TransferWise was that the best price they could get from the GBP clearers was £1.30 per payment; they became the first clearing member of the Bank of England and dropped their cost of GBP clearing to £0.03 (three pennies) thereby enabling even cheaper customer transfers.

Cross border payments

Ripple was back (and on stage) at Sibos – they ran their own competing event called Swell for a couple of years – indicating that SWIFT now feels sufficiently confident of GPI’s robustness. One panellist described cross border payments as two domestic payments with FX in between – which is basically how Ripple treats it (although it is now trying to squeeze RPX into the middle).

There were many cross-border ACH service providers at Sibos this year – entities that put their own domestic bank accounts at the service of their customers to enable them to offer cheaper cross-border payments to their customers in turn, thereby avoiding the cost and complexity of the four corner correspondent banking model.

Cross-border ACH addresses the price issue with cross border payments – that they are too expensive. Cross-border ACH costs between zero and US$1 per payment and generally arrive same day and in many cases within minutes, depending on the local ACH used. Cross-border ACH normally uses domestic low value systems and many of these cap the value of payments, which precludes their use for high value treasury payments. Nonetheless, the technology is finding broad adoption within its niche.

SWIFT’s gpi addresses several problems with cross border payments – notably speed and obfuscation – by enforcing a rule book for cross-border payments which includes:

• Fast (within 24 hours).
• Trackable (GPI tracker).
• Transparent (all fees advised upfront).
• Complete (no more data truncation).

GPI will be mandatory for all 10,000 or so SWIFT member banks in 2020, and this will certainly improve treasury operations. SWIFT intends to continue to address issues that have come from its corporate advisory group including:

• Instant international payments and 24/7.
• Pre-validation.
• Pay and trace.
• Inbound tracking.
• Request to pay.

A strength of GPI is that it builds on the existing and well-worn four corner correspondent model so that it accesses the extensive coverage of SWIFT’s 10,000 or so member banks. Correspondent banking also brings weaknesses. The correspondent banking model consumes large amounts of precious bank liquidity – a common bank heuristic is that the liquidity required for cross-border payments is 60% of the daily volume. Another issue is that tightening AML and KYC enforcement has made big holes in correspondent relationships.

Ripple is more efficient from a liquidity perspective, and this benefit may well offset its relatively limited coverage compared to GPI.

Conclusion

Payments are progressing rapidly. With diverse use cases searching for solutions and a variety of providers providing solutions, payments will become more heterogenous in the near future. The landscape may only simplify when government institutions like central banks decide that free, ‘anywhere anytime’ payments are a public good – after all, payments are a more obvious common good than many things governments currently provide to their citizens.