ISO 20022: it ain’t over till it’s over

The Sibos Insider Blog provides inspiring insights from financial industry experts and influencers. The blogs provide informed thought-leadership on a range of topics. The opinions expressed in these blogs reflect the personal views of the author and not their organisation, Sibos, or SWIFT.

Over a dinner with an international group of bankers during Sibos in London last year I asked how far along they were with their ISO 20022 migration projects. There were three general forms of answer: European banks said they’d already done it at the launch of the Single Euro Payments Area (SEPA), US banks corrected the form of the question and said they were planning their implementation programmes, and a random assortment of them said they didn’t know because they didn’t work in compliance.

The central banker sitting next to me groaned softly and shook his head remorsefully.

A year later, would any of my fellow dinner guests have changed their answers? I doubt it, for the simple reason that the extent of the challenge is only just dawning on many people. It is not just a question of the size of the migration (or implementation) projects, it is the detail.

The push-back of the deadline for SWIFT MT messages to be replaced by ISO 20022 for cross-border payments and cash reporting to the end of 2022 (with a backward-compatibility get-out clause for use of MT after that) will provide breathing space, a few minutes of which I propose to use to consider some aspects of that detail, using the benefit of hindsight.

Let’s start with XML. In 2005, I stood on a stage moderating an audience discussion on standards migration during which one participant became very irate abut being ‘forced’ to adopt XML. In subsequent years there were many times when people pointed out that ISO 20022 doesn’t mandate the use of XML, which is just as well because the venerable Financial Messaging Exchange (FIX) is still with us and has been joined by the newer JSON data-interchange format. My concern here is that XML has been around a long time now, and the tech world is changing rapidly.

Another worry to me is the much-touted benefits of the richer data that is carried by ISO 20022 messages. The suggestion is that banks, corporates and everyone along the financial value chain will be able to benefit from enriched payment data and there will be further opportunities for banks to add new service capabilities and enhance customer service.

It’s true that ISO 20022 can carry more data, but does that mean it needs to? A few years ago, when the development of a real-time payment variant of the standard was under discussion there was concern that users might attach cat videos to payment messages. For one thing this would increase the payload of the message and burden the network, but a more subtle argument against multimedia attachments is that people might attach advertising material that, while legal in the sender’s jurisdiction is not in the recipient’s: gun adverts was the example used.

Within the (largely) structured data world of banking, the ability to have more detailed transaction data that can be parsed automatically, such as Legal Entity Identifiers is clearly a benefit. (Are LEIs sorted out, by the way? Thought not.)

Then there is the issue of making that data usable in different organisations’ systems. Not much of a problem these days, with modern systems using Unicode capable of dealing with non-Roman character sets, but even eight years ago at Sibos we were talking about the problems this causes. It’s not a problem with the standard, obviously, but many of the then-existing systems in banks, clearing houses and exchanges around the world had Roman characters baked in. That probably includes the Burroughs mainframe that was still being used by one US bank this time last year.

With the pace of technology and societal change that is going on, it is often hard to look up from your desk and look around to get the bigger picture. Sibos provides a chance to do that. When we get to Sibos in 2025, the ISO 20022 deadline will (let’s hope) be looming and the SWIFT 2030 strategy will be driving the agenda.

It’ll have taken the best part of a quarter of a century to get there, but that’s hardly surprising: there are a lot of disparate systems, processes, technologies and regulatory changes that have to be put into place, and a lot of that work has been done by the standards geeks, whose work I’ve been lucky enough to follow closely. That makes me a standards groupie, which is why I’m optimistic that the challenges will be overcome and long in the tooth enough to have learnt that this is a long journey. Maybe it would be better to stop thinking about deadlines and start thinking about milestones: we’ve already passed a lot of those.