From Treasury Jigsaw to Operating Model: Why Integration is Harder Than it Looks
Published: May 21, 2026
Excel persists for a reason, but is there an alternative? Emmanuel de Rességuier, CEO, Fennech Financial, explores the answer to this rather tricky question.
Three minutes before the usual CFO call at 9am, the question lands, inevitably and without ceremony: “What’s our cash position?” Three systems produce three different answers. A fourth sits in Excel, carefully adjusted, faintly trusted, yet entirely indispensable. Everyone knows which number will be used. Nobody is entirely sure why that is still so.
If that sounds familiar, it’s because treasury fragmentation rarely looks like failure but instead appears as something that has worked well enough, often for many years.
A question nobody likes to answer
Before considering technology, data, or architecture, there must be governance. Its frequent absence is a clue to the problem. Who owns liquidity? Who defines cash? Who decides when IT pushes for ERP standardisation while treasury insists on functionality that reflects market reality?
In many organisations, these questions are never formally settled. They are revisited repeatedly, project by project. As one seasoned treasurer put it: “The system is usually the least of our problems, but alignment is a major issue.”
Indeed, most transformations do not fail in testing. They fail much earlier, in steering committees where no one quite has the mandate to say no.
What you mean, ‘cash’?
Ask 10 treasurers for their cash position and you will receive 10 defensible answers. Available cash, trapped cash, forecast liquidity, regulatory buffers: each response has its own logic. The difficulty is not conceptual, it is organisational: integration requires that these definitions are agreed and consistently applied across entities, banks, and systems.
Christian Mnich, Vice President, Global Go-To Market and Partner Lead for Office of the CFO Solutions, SAP,1 framed it bluntly when he said data integration enables timely reporting and crisis modelling. The implication is simple. Without a shared data model, integration accelerates confusion rather than clarity. This is where many projects struggle: not in implementing tools, but in agreeing on what those tools are meant to represent.
No one mapped the operating model
Treasury operating models are rarely designed from scratch. They evolve. A regional structure here, a payment factory there, an IHB introduced after a liquidity scare, a local exception granted for regulatory reasons. Each decision makes sense. Collectively, they form a system that is coherent mainly only to those who built it.
The challenge becomes visible after major events, particularly acquisitions. When HeidelbergCement integrated Italcementi2, treasury was not simply merging systems, it was reconciling two distinct views of liquidity, risk, and control.
Integration, in these cases, is less about technology than about aligning financial reality. And alignment requires choices. But by the time organisations debate systems, much has already been decided implicitly. Some opt for ERP-centric models.
Sanofi’s move towards a global SAP S/4HANA backbone3 illustrates the approach: reduce complexity through a single financial core. Others favour modular ecosystems, prioritising flexibility and depth. In practice, most land somewhere in between.
Industry mainstay and CEO, Simply Treasury, François Masquelier, has argued4 that treasury architecture should be reconsidered as a whole, not extended piece by piece. Yet many organisations do the latter, adding interfaces and layering tools, building what he once described as a “pile of solutions”.
The result is familiar. Systems connect technically, but do not integrate operationally. Architecture, in effect, reflects prior compromises.
Why Excel survives everything
For all the investment in systems, Excel remains treasury’s constant. This is not nostalgia. It is pragmatism. When systems fail to capture operational nuance, users recreate it manually. Over time, these spreadsheets become sophisticated, combining data extraction, adjustments, and reconciliation logic. They work, and they are trusted, but they are also fragile.
Karen Van den Driessche, former Vice President and Group Treasurer, Ontex (now Vice President, Group Treasurer, Head of Tax, Lipton), noted5 that automation delivers value only when it simplifies workflows enough to shift focus towards strategy. When it does not, people revert.
Excel survives not because treasury resists change, but because it fills the gaps.
Integration is not a project, it is a reset
Some organisations address fragmentation incrementally. Others choose more decisive moments.
RTL Group’s treasury transformation6, including a co-ordinated overhaul of FX platforms, TMS and payment factory, took what is often known as a ‘big bang’ approach. Doing so delivered measurable results but required a level of alignment that is uncommon.
At the other end of the spectrum, greenfield environments such as GE Healthcare’s treasury build7 show a different reality. Integration is far simpler when it is designed from day one.
Most organisations sit somewhere in between, managing inherited complexity while trying to impose coherence.
A quieter, yet more demanding ambition
An integrated treasury is not defined by elegant systems or the latest technology. It is defined by something more practical: the ability to answer, consistently and without negotiation, a small set of questions. Where is the cash? What is available? What is at risk? What can move, and who decides?
Masquelier has suggested8 that treasury’s natural destination is a centre of excellence for liquidity and risk. That ambition is not achieved through tools alone. It is achieved when systems, data, processes, and people align sufficiently often to support confident decisions.
The 9am question will, regardless, continue to arrive.