Imagine your core banking system as a Jenga tower. With each change in product, integration, or compliance, it feels like you’re pulling out another block, hoping the tower doesn’t fall. As the tower grows, each move becomes increasingly governed by fear rather than ambition. For many banks, this is the reality of hitting the “core ceiling.” It’s not just about running out of capacity, but reaching a point where the risk of touching the core starts to limit what the organisation can do for its customers.
Traditionally, financial institutions have had to choose between ease of configuration, scalability, and flexibility. Regardless of the path chosen, they would inevitably hit the core ceiling at some point. A core that’s ready for 2040 and beyond needs to offer all these aspects, without compromise.
Understanding the Core Ceiling
The core ceiling is a two-pronged problem for banks. The first issue arises when an architecture’s inherent scalability limits – such as customer numbers or transactions per second – start to manifest as visible performance issues. These may include card declines at peak times, slow responses in digital channels, or an inability to keep adding new volumes without significant trade‑offs in other areas.
The second issue is modification constraints. Internally, this is seen as friction from code to client. Engineers find it challenging to make changes, adapt or move swiftly because every change carries the possibility of unintended side-effects somewhere deep in the stack. Releases become more about maintaining what already works and less about unlocking new value.
While it’s tempting to blame this solely on legacy mainframes, even newer platforms often end up in a similar situation. Some platforms are built around pure configuration, which is quick to set up and familiar to business teams. However, your ability to differentiate is ultimately bounded by what the vendor has chosen to build and what they decide to put on their roadmap.
The Open-Close Principle: A CTO’s Goldilocks Zone
The challenge is to find a balance – one that avoids both rigid configuration-only models and chaotic custom build scenarios. The most crucial shift is to adopt an open-close principle in core design. First formalised by Bertrand Meyer in his 1988 book Object-Oriented Software Construction, the Open-Closed principle suggests that “software entities should be open for extension, but closed for modification”. In core banking, this would allow banks to close off the core to direct modification, but open it to extension, enabling the business to innovate safely around it.
Rather than encoding every future product variation directly inside the core, you push as much variability as possible into modular, composable components that sit on top or around it. These components can be configured, combined, and iterated quickly without jeopardising the underlying system’s integrity.
This approach creates a Goldilocks zone. You avoid rigid systems that make modification a painful process, and you also avoid the chaos where layers of one-off customisations slowly undermine reliability.
The Price Maintenance Trap
While the open-close approach seems like an obvious choice, some platforms are architected around being the cheapest option for a particular segment, tying design constraints deeply to price maintenance. As these platforms approach one and a half or two million accounts, or as contactless and real-time payments volumes grow, those design choices start to become problematic. This is the architectural ceiling, where your Jenga tower simply isn’t equipped for any more blocks.
On the other hand, architectures that are designed and tested from the outset for tens of millions of accounts change the narrative. The question is no longer whether the tower will stay standing; instead, it’s about what you can safely build next. Capacity and extensibility become part of the same design envelope, rather than trade-offs you constantly have to juggle.
Choosing How You Want to Grow
Ultimately, breaking through the core ceiling is about being clear on two linked questions: how far do you need to scale before the architecture gets in your way, and where do you want your innovation layer to live?
If your answer to both is “we will work it out later”, you are almost certainly building another Jenga tower. Choosing an open-close approach – a closed, stable core with open, governed paths for customisation at the edges, and headroom for higher volumes and transaction rates – gives you a different future.
Your tower can continue to grow, but each new block no longer feels like a gamble.
Miles Wilson, CTO, 10x Banking
Source: Here
