Community

Starting 2008, it took around 7 years for SIPS, real-time payments in the UK Faster Payments system, to reach ~1bn annual txns, or just over 15 real-time payments per capita per year, and just 7 months for the Nets real-time payments system in Denmark to reach the same level after it launched in late 2014.

It is over 3 years since the first SCT Inst payment in the Eurozone, and looking at EPC stats, the current run rate of SCT Inst is about 1.8bn txns per year, or 5 real-time payments per capita per year in the Eurozone. This is more like the steady adoption of the UK experience than the fast adoption of the Denmark experience, and a little surprising given we are now much more advanced in a digital age of always on, instant services.

Given that real-time payments are core to a modern digital economy, and that even targets for adherence to SCT Inst (under Article 4 of the SEPA Regulation2) appear still to be met in Eurozone countries, I expect that the EC may mandate adoption targets for SCT Inst, and soon.

Meanwhile, UK SIPS are currently running at about 32 real-time payments per capita per year, or 2bn payments, growing at over 20% annually (and on a system that is 13 years old needing replacement), illustrating that Eurozone banks and other PSPs have a massive upside from SCT Inst, and should get motoring, with or without a regulatory push.

I wonder how much regulators focus on intraday liquidity for cross-border payments?

It is one thing to measure and manage intraday liquidity in settlement accounts at the central bank and intraday positions in closely regulated high value payments systems like CHAPS, Target, Fedwire, it is a completely different challenge managing liquidity in nostro accounts (which can run into 100s for some FIs) at correspondents banks abroad - end-of-day liquidity management is difficult enough, intraday is much more complex.

I guess that regulators are less interested, as the liquidity in a nostro in a correspondent outside their jurisdiction is outside their control, while availability of funds in nostros in their jurisdiction to meet payment obligations made abroad is outside of their concern.

SWIFT GPI makes a great play on enabling very fast cross-border payments - the messaging component has always been fast (compliance hits aside), but it is the liquidity component that is the greatest challenge. To enable GPI, funding liquidity in overseas nostros at end of day to release payments the following day needs to be replaced by funding liqudity in advance if GPI targets for near-real-time payments are to be met - pre-funding nostros the previous day, or intraday, requires accurate forecasting, including incoming payments to net, and requires continuous, accurate management.

Liquidity has been fairly abundant over the past decade, and pre-funding nostros by over-funding has been an easy, if inefficient solution. However, if liquidity were to become tighter, settlement and credit risks in the global cross-border payments system will escalete rapidly.

It will be interesting to see how this is addressed in the G20/FSB project to enhance cross-border payments.

useful report, three observations:

1. the report talks about "aspects of fintech regulation may require adjustment" - much, much more ambition is needed, i.e. full regulatory reinvention. Take advantage of emerging technologies to do away with huge frictions in e.g. KYC and AML by approaching them very differently with better outcomes, using zero knowledge proofs, homomorphic encryption etc to shift financial crime compliance from an individual entity to collective responsibility, and with explicit safeguards to protect individual privacy and prevent government and big tech surveillance. The Alliance for Innovative Regulation (AIR) has some great ideas and principles for regulatory reivention to consider. As an example, cross-border micropayments are potentially transformational but are impractical when burdened with the same compliance requirements as £1m payments (as low value remittance payments have long suffered) 

2. the CBDC piece needs to move beyond the inward looking central bank lens into how digital sterling can create a new ecosystem of open finance innovation, with programmable money and new digital financial services outside of the traditional banking system - this is where the true benefit of digital sterling lies 

3. the interoperability of a retail CBDC is important, but more important is the holistic co-existence of an open finance system based on digital sterling, the DeFi open finance system already emerging in non-sterling digital assets, and the existing financial system (where most of Fintech outside of crypto currently operates) 

just two hours ago I wrote a comment below the Finextra news on the European EPI payments initiative saying it is probably a bad idea to go for a centralised infrastructure - this outage at the Fed confirms why (imagine it happening in, say in 10 years time, even for a few minutes when volumes are 10 to 100 times where they are today, all real-time, and connected to 100s of millions of end points - mobiles, meters, sensors etc)

#distributedarchitecture

There are very little details about EPI released publicly, but as the RFI states, EPI is based on SCT Inst, so at its core it is an account-to-account credit transfer payment system, rather than a legacy-style card system, which makes it very different and very interesting compared to previous attempts for a European payments solution (Monnet, Berlin Group, Payfair).

A great model for EPI is India's payments infrastructure run by NPCI  which has a switch (old style ATM) for clearing (NFS), supporting an instant mobile payment service (IMPS), which supports the universal payments interface (UPI) giving API access to Fintechs and banks, and innovations such as digital wallets. UPI volumes are enormous and have grown very quickly, so opening up the infrastructure clearly works. Hopefully, NPCI's international arm will be bidding for this, they have delivered a fantastic system for India.

Building something similar (but on more advanced technology in the clearing and settlement layer) would give European payments a huge boost and set it up for decades.

I am unsure though that a centralised infrastructure is a good idea. A distributed architecture (using blockchain type technology) would be a cheaper, more resilient system, faste to implement and easier to innovate on. Especially with the digital euro on its way... 

this looks like yet another data collection point for Big Tech surveillance to feed their advertising machine picture - email, search requests, location, travel, social media, payments... the list goes on.

correction 2nd para - BTC reached $50k on 17 Feb 21 not 17 Feb 20

"BIS Innovation Hub has centres in Switzerland, Hong Kong and Singapore, with locations in Toronto, London, Frankfurt and Paris, and Stockholm on the horizon"..... sounds more like a network than a hub. 

It reveals the innate bias among central bankers to centralise, as is natural - I expect this tension between the instinct to centralise and logic to decentralise will be a key feature of these projects, let's hope it leads to some creativity and innovation.

For CBDCs to be successful in public use they need to be indistinguishable from other forms of electronic money for the public to adopt them. Otherwise, the public will have to understand the difference between emoney, bank deposits and CBDC, risking mass confusion.

This could happen with the digital euro if it has maximum holding limits, which would flag it as different to bank deposits (emoney in the form of prepaid cards can have holding limits too but this restricts the use and appeal of the products emoney supports).

It is rare to see the public lens considered in CBDC analysis, which in my opinion risks CBDCs being designed with flawed assumptions on public acceptance and adoption.

@s_ketharaman

Your memory serves you well.

In the UK individuals can pay their personal tax electronically by direct debit (one off), online/telephone bank transfer, debit card or corporate credit card (and non-electronically by cheque in the post or in person at a branch).

The payment process is separate from the process to submit a tax return. If paying by online bank transfer (an option which has been around for some time, probably 10 years, even 20 years), you are given a choice of two HMRC bank accounts to pay into, one being a default if you are uncertain (which I suspect many people are as it is unclear how you find out the one to use). Some banks will have these HMRC details set up already, other banks require you set them up yourself as a beneficiary on your account - when submitting the payment, you need to include a payment reference which you construct as your 10 digit tax number (UTR) followed by the letter K.

So plenty of scope for payers to make errors - mistype your UTR, forget to add a K and send to the wrong HMRC account. These HMRC accounts are used for all types of tax payments (VAT, corporation tax etc), so I expect HMRC must have a huge reconciliation process to match payments in error. Since they receive millions of payments, it means 10,000 errors per error rate % per million payments, so quite a headache.

With OB, there is much scope to improve the customer experience as well as reduce errors. For example on submitting a tax return, connecting to your bank from the HMRC website (either in the same session or later), authorising the payment and that's it, without entering any information (with the amount and reference are formatted automatically).