ISO20022: The big shift to structured addresses

The global financial system is changing. Soon, ISO 20022, a messaging standard designed to improve how payments are processed and communicated, will become the global norm. ISO 20022 was introduced in 2004 and has since been part of a long and complex journey to improve international payment messaging. 

A critical shift will happen this year on 22 November, when the coexistence period for MT (Message Type) and MX (Message Exchange) messages will end. From then on, banks need to be able to process MX messages, which contain much richer and more structured data. 

A few banks have already transitioned, but many more have yet to make the switch. Non-compliance may result in failed payments, operational inefficiencies, regulatory scrutiny, and reputational damage. And unfortunately, more work is ahead for everyone.

The next major shift

After November 2025, banks will need to prepare for yet another major shift: structured addresses. 

Starting November 2026, unstructured postal addresses will no longer be supported in CBPR+ messages. From this point on, Swift, and payment schemes, such as SEPA and CHAPS, will begin rejecting transactions with unstructured data. 

Banks need to adopt either a fully structured or hybrid address format to become compliant. Neglecting this may lead to payment failures, processing delays, higher costs, false compliance hits, and increased scrutiny from regulators and correspondent banks.

The benefits are far-reaching. Structure will bring clarity and precision to payments, improving financial crime compliance, sanctions screening, and processing speed.

But there are big, expensive hurdles to overcome first. Namely, updating libraries of messy customer data.

The case for structured addresses

To illustrate, let’s start with the problem: unstructured addresses. Free-text fields in MT messages have no predefined structure. The entire message is lumped into one box, without a clear or specific space for street name, city or postcode. 

Hard to read and automate, unstructured addresses lead to inconsistencies, errors, and difficulty processing payments.

Now consider the ISO 20022 Structured Address element, which includes fields for specific address data. Unlike unstructured addresses, these are clear, quick and machine-readable. 

Now, it looks good on paper. But address formats differ across regions. In the West, for example, we prefer a bottom-up approach, which starts with a house number or name and ends with a postcode (House>Street > City > State > Country). In the East, a top-down approach is the norm (Country > State > City > House>Street). 

In some countries, such as Hong Kong and Somalia, there are no postcodes. And in others, like India and Kenya, addresses often include landmarks to help identify a location (e.g. "near City Mall, opposite the bus stop"). Unstructured, free-text fields could accommodate these differences. Structured Addresses cannot.

So, this is where hybrid addresses come in. A hybrid address combines both structured and unstructured elements to allow for differences in address formats. This makes them more flexible than structured addresses.

Updating legacy systems

The trouble is that many corporate Enterprise Resource Planning (ERP) and Treasury Management Systems (TMS) use free-text or semi-structured fields for address information. 

That means corporates and their payment service providers have big databases filled with customer and counterparty addresses in an unstructured format. All of which need updating to meet compliance.

But updating these legacy systems and databases is a huge task. One that’s particularly painful for banks and large multinationals with extensive payment networks. Converting years of historical data into a structured or hybrid format could cost millions in IT infrastructure upgrades and operational expenses. 

What’s more, banks will need to invest in training staff to ensure the migration is handled efficiently. Most of the heavy lifting needs to be done in 2025 and 2026. The cost of inaction will be a far bigger burden.

How to plan for the future

Banks must begin planning now if they want to be prepared for the November deadlines. As a first step, they should already be reviewing and migrating their existing address data to a more structured format. They can also use AI-based address validation software to help automate the process. 

Banks need to future-proof their systems and this will only be possible if they prioritise a transition to fully structured addresses. Hybrid addresses will remain valid until 2026, but their status may evolve as the migration progresses. 

As systems become fully compliant with ISO 20022, the preference may shift toward fully structured addresses, so banks need to take proactive measures to avoid having to redo this work in the future. They will also need to look at upgrading their ERP and TMS systems to support ISO 20022.

By capturing address data in a structured format from the start, banks improve their integration with modern ERP systems, data accuracy, and compliance. ISO 20022 is just around the corner, and only proactive banks will stay ahead and reap the rewards of smarter payments and a stronger competitive edge.