Banks are spending real money building internal workarounds to imitate instant payments, inside systems that were never designed to produce them. The workarounds cost time and money to develop and to maintain, and what they deliver is a transaction that looks instant to the customer of one bank, while the books behind it settle on the next opening banking day. Between two banks, or between a bank and a payment provider, the appearance is harder to hold together.
That gap is what the architecture was built to produce. In a legacy core, the ledger and the processes that act on it live inside the same platform. Every process reads and writes against the same tightly held data, and keeping that ledger correct under those conditions means lining the work up at a fixed point and running it in sequence, with nothing else touching the data while it runs. The nightly batch follows from the design.
Fintech platforms hold data as its own layer, reachable by any process through an API and updated by events as they happen. Nothing in that design waits for a closing window, because the firms that built it started later and had no ledger already sitting in a box that had to keep working through the migration.
There is now a date on this. Albania signed the letter of intent for the Western Balkans instant-payments platform in January 2025, with Kosovo, Montenegro and Bosnia and Herzegovina; North Macedonia joined that October, bringing in five of the six. The platform is a clone of the Eurosystem's TIPS, operated by Banca d'Italia, settling around the clock every day of the year, and its cross-currency model closes both legs of a payment atomically in about twenty seconds. Go-live was planned for July 2026 and has slipped to autumn. Banks connect as participants, build their own applications against it, and run their own testing. Twenty seconds of settlement then arrives into a ledger that closes its books once a night, and the payment waits there for the batch.
Most modernisation programmes read digital-first as a channel problem: rebuild the app and put a better interface in front of the same core. People are not so clear what it means, and they do not distinguish quite well between digitalisation and digital-first. Digitalisation puts an existing process on a screen. Digital-first separates the data from the processes that act on it, so that any process can reach the ledger through an API or an event stream without waiting for a nightly cycle to finish. The order runs one direction. Decoupling data from process comes first, and real-time and 24/7 availability are what a bank gets once that separation exists. Running the batch more often, or wiring a fast rail onto the front end, leaves the data and the processes sharing one box.
Investing in the architectural modernisation of the core banking is something that will come seriously in front of banks, and the question is at what cost, and the time that it will come. The time is the more sensitive part, because it is tied to the market share that fintechs can take from banks. Technically, and on user friendliness, fintechs are obviously more advanced. The key and only advantage of banks is customer trust. Trust is something that fintechs have started to build faster, and when they have that improved level of it, banks will face a real threat from other financial institutions offering innovative solutions.