When insight isn't enough: the liquidity architecture problem

Most banks understand liquidity behaviour better than ever, so what’s holding them back from acting on these insights? This article explores why architecture is now the real constraint on action, and what it takes to build a liquidity capability that works across the entire enterprise.

Banks have spent years sharpening their understanding of liquidity behaviour. Stress testing now uses increasingly granular data, richer behavioural assumptions, and broader scenarios. Treasury and risk functions have also gotten much better at monitoring and forecasting liquidity.

But even though firms can spot liquidity risks and opportunities much more accurately than they could five years ago, many still can’t turn that understanding into coordinated action fast enough to change outcomes. The challenge has moved from analysis to execution, which matters because liquidity itself has changed shape, and the cost of holding too much has risen significantly.

This article is part of a wider Baringa series exploring how banks can rebalance risk and growth across the balance sheet, from the buffers and models that shape liquidity to the behavioural insight that increasingly drives it. Here, we look at why even strong insight often stalls before it turns into action, and what that reveals about the liquidity architecture behind it.

Moving beyond cash

Liquidity has never been just about cash. It has always rested on a broader set of assets, rights, and funding sources. Regardless, most liquidity management frameworks were built around cash anyway, because it was the easiest component to observe, mobilise, and govern. Operational complexity, settlement constraints, and questions around collateral usability made it difficult to gain a broader view, so cash became the default lens even though it never provided the full picture of liquidity.

That approach is getting harder to sustain. Today, liquidity sits across interconnected pools of cash, securities, and collateral, each governed by different timings, constraints, and operational processes. Even when the overall position seems strong, effective liquidity depends less on ownership than usability: where assets sit, how quickly they can be mobilised, and what restrictions apply along the way.

Liquidity, in other words, is best understood as a position across interacting assets rather than a balance held in a specific place.

The challenges of fragmentation

Once liquidity is viewed through this broader lens, another problem becomes clear: architectural fragmentation.

Most banks already have the data they need, but accessing it rapidly and reliably can be difficult. Information is scattered across a complex web of payment platforms, nostro account and reconciliation data, trading and collateral engines, custodians, central securities depositories (CSDs), and regulatory reporting infrastructures. None of these were designed to operate as part of a single liquidity architecture. Each system may provide a valid view of its own domain, but when put together they rarely add up to a coherent firm-wide picture of liquidity.

These limitations tend to show up intraday, which is increasingly where liquidity stress first manifests through settlement concentrations, margin volatility, and rapidly changing client flows. Cash teams watch payment queues, collateral teams track margin calls, and market desks see funding requirements. Treasury gets a consolidated picture, but it often arrives late and without the context needed to understand what’s driving it. This means that liquidity events are often managed locally but understood only partially.

The lack of a shared, timely view becomes a constraint in its own right. As liquidity grows more dynamic, uncertainty can emerge because institutions cannot see, mobilise, or redeploy it with enough confidence.

Moving towards an integrated liquidity architecture

Most existing liquidity architectures in use today were built for a slower world, where settlement cycles were longer, collateral mobility was lower, and reporting cycles ran in days rather than hours. Many of those assumptions no longer hold, yet balances, collateral positions, securities holdings, and regulatory views often remain spread across different platforms.

As a result, even well-governed institutions can struggle to answer relatively simple questions in real time:

• What liquidity is genuinely available?
• Where is it trapped?
• How fast can it be mobilised?
• What does a change in one position mean elsewhere on the balance sheet?

Closing this gap does not necessarily mean ripping everything up and starting from scratch, but it does mean bringing more consistency to data, systems, and decision-making.

In practical terms, this means creating a shared view of liquidity across cash, securities, and collateral that’s based on usability rather than pure ownership. It means seeing beyond end-of-day balances, so firms can track movements, constraints, and encumbrances as they develop throughout the day. It also means closing the gap between internal liquidity management and regulatory reporting, so both draw from the same, trusted data.

Some institutions call this a liquidity control tower. Whatever you call it, what matters most is the capability it describes: a timely, trusted view of liquidity that lets decision-makers work from one shared version of the truth.

Regulation is adding momentum to this shift. Supervisors increasingly expect greater granularity, transaction-level visibility, and demonstrable liquidity usability rather than simple aggregate balances. When internal management and regulatory reporting draw on the same underlying data, the benefits go beyond compliance, with institutions typically gaining greater transparency, lower reconciliation effort, and more confidence in their decision-making.

Liquidity is now everyone’s responsibility

Taken together, these pressures point to something bigger than architecture alone: the need for an enterprise liquidity capability. While liquidity management used to be treated largely as treasury’s responsibility, this is now getting harder to sustain. Treasury, risk, finance, markets, and operations each hold a piece of the liquidity picture, but no single function can see the whole of how liquidity is created, consumed, or mobilised across the institution.

As behavioural insight grows more sophisticated and liquidity management becomes more intraday, coordination across these functions starts to matter just as much as the analysis itself. Shared visibility, faster escalation, and clearer accountability are now must-haves if institutions want to manage liquidity as a real-time portfolio rather than a collection of individual positions and processes.

Closing thought

The question is no longer whether institutions understand how liquidity behaves. What matters now is whether that understanding can be shared, trusted, and acted upon consistently across the organisation. As liquidity becomes more dynamic, intraday, and cross-asset, these organisational capabilities may end up mattering as much as the quality of the insight itself.