How Fast is Too Fast?

Unlike last year where much of the discussion was around ‘why’, this year’s debate focused on ‘how’ and the routes to solving some of the issues.

Now the ‘why’ is now pretty much a given, as mobile and the Internet have raised consumer expectations. When consumers buy items on the web, they no longer will put up with the fact that the seemingly straightforward process of debiting one bank account and crediting another can take longer than the physical movement of goods.

As one panellist put it, “if the banks can’t satisfy the customers, then the upstart new entrants will. The genie was out of the bottle with Internet and mobile - I don’t hear banks discussing business cases anymore - there’s definitely no turning back now for real-time payment systems.”

But how fast do we really need to go? Well it depends. “Arguably, a commercial delivery-versus-payment should be completed within a few minutes. Paying back a friend in a restaurant might be ok within in a minute, but paying a taxi driver whilst standing in the rain needs to be done in a few seconds”, said one analyst. “In reality ‘real-time’ is short-hand for ‘very fast’ - no more than a few seconds.”

But it’s more complicated than that. “People think that this is all about speed. It’s not. When I’m awake I want (near) instant gratification when it comes to my money,” said a session moderator. “Yes, people also sleep, but not everybody sleeps at the same time”, he added, “this means that customers now expect us to operate 24x7x365, which is far from easy.”

On both of these counts, market expectations have outpaced existing legacy batch-based infrastructure, and many countries in all corners of the world are already implementing domestic real-time payments systems. In SWIFT’s recent White Paper “RT-RPS Global Adoption Landscape”, they noted that 18 countries are currently ‘live’ with real-time systems, 12 countries are ‘exploring / planning / building’, and an additional block of 17 countries are ‘exploring’ through a pan-Eurozone initiative.

And most seemed to be represented at this years’ Market Infrastructures Forum, as it was packed with workshops where customers described their implementation approaches and upcoming plans.

“It’s interesting that all of these different markets want the same thing”, said one real-time practitioner, “but each are approaching it from different angles – you have centralised and decentralised schemes, different methods for the provision of liquidity, different models for clearing and settlement, some start from scratch and some re-use existing infrastructure”, said one panellist. “There’s more than one way to make all the plumbing work, but from the customer’s point of view, it should all be transparent and seamless.”

But despite this diversity, everybody in this new market seemed to agree on one thing. “You have to use ISO 20022 as the base standard for these messages. The market knows that if they pick something else, then they’ll be left outside the tent - banks need to connect to all these systems, so conforming to one protocol just makes good technical and commercial sense”, he continued.

One analyst went even further, “today it’s just the banks that need to connect, but tomorrow it’ll all change. Regulation will seek open up these markets, and I can see that corporate firms, payment service providers and other new entrants will want to connect to these infrastructures and create new services. Selecting one standard just makes it so much easier for interoperability.”

But is it only about the retail, end-customer experience? Interestingly, many participants reported that they are looking at this technology to solve some old institutional problems.

As one practitioner put it, “in the U.S., there are perhaps 15,000 banks and countless merchants that still rely on physical cheques. Imaging for corporate cheques is very expensive. One holy-grail of this technology would be to automate the remittance process – the seamless reconciliation of order and invoice for a particular payment transaction. Reconciling one invoice with one order is not too tricky, but reconciling one-to-many, many-to-one or partial payments are really time consuming and operations-heavy. Real-time and ISO 20022 now give us the tools to solve these long-standing and ingrained issues.”

So, domestically it all seems within our grasp - “now we can get real-time systems that can do 1000 TPS” said one session moderator, “it’s not a question of ‘how fast is too fast’. For some countries, the right question is ‘how slow is too slow’. Some of the big markets have realised that they were about to miss the boat, and now they are playing catch-up. But once you start on this journey there’s no looking back. It’s a really exciting time and we don’t know how the market will evolve.”

“Banks have already become coffee shops, and, who knows, coffee shops might even join the party and become banks, offering real-time services that we can’t yet imagine”, he added.

However, there are still challenges when we step out of purely domestic schemes. “Going cross-border we get into issues such as the sovereignty of payment data and addressing data, AML and sanctions screening to keep the bad guys out, and ensuring inter-system interoperability. Sibos is the perfect venue to debate and solve these issues with our community.”

This year’s conference certainly furthered the real-time discussion – everybody agreed we need it and everybody agreed it needs to be fast, ubiquitous and ‘on’ all of the time. While domestic schemes are charging ahead, the debate is just starting for pan-regional schemes where payments will have to move cross-border and the level of complexity increases.

“I can’t wait to see how this story ends,” said one practitioner, “roll on Sibos in Geneva next year!”