Silos are now a risk, not an inconvenience
AFP's write-up of its risk webinar opens with a flat statement: companies can no longer let FP&A and treasury operate in silos [1]. The reason is practical. Treasury owns liquidity and balance-sheet risk. FP&A owns scenarios, strategic risk and the income impact. An FX shortage or a supply shock hits both at once.
The CFOs on that webinar show what works. One built ready-made scenarios, so the team follows a plan instead of reinventing one under pressure. He also ran a weekly “war team” of treasury, FP&A and legal, which found a way through a two-year FX shortage [1]. Nilly Essaides reaches the same conclusion from the operating-model side: CFOs expect faster, better-informed decision support, and that means dismantling the silos between finance functions [2].
Collaboration fails at the model layer
The weak spot is what the two teams calculate on. Essaides is explicit: treasury and FP&A each produce short-, medium- and long-term cash forecasts, they are often out of sync, and that creates confusion at planning time. Inconsistent data and models, she argues, undermine data-driven decisions. Her remedy is one data pipeline, shared models and one analytics layer for both teams [2].
Now read AFP's scenario planning against that. A ready-made scenario plan is only as reliable as the model behind its numbers [1]. If treasury runs the FX shock in one workbook and FP&A in another, the weekly meeting spends its first hour reconciling figures instead of deciding.
Where those models live in spreadsheets, the problem compounds. A consultant builds the logic, then whoever inherits the file maintains it. Formulas drift, versions multiply, and nobody can say which one is authoritative. The diagnosis is not new: Essaides published hers in 2021, and AFP was still making the case for collaboration in 2024.
What the evidence demands, and what K-Φ does
| The requirement | Source | What K-Φ does |
|---|---|---|
| One data repository and one analytics layer for both teams | Essaides [2] | Treasury and FP&A analysis run on the same ledger data, pulled from ERP and GL sources such as SAP and QuickBooks |
| Cash forecasts that agree | Essaides [2] | Cash forecasts, financial statements and budget variance are produced from the same model |
| Scenarios ready before the crisis, run on numbers both sides trust | AFP [1] | Model logic is compiled once and governed, so every team runs scenarios on the same rules |
| Risk as a continuous, cross-functional conversation | AFP [1] | The same GL analysis, KPIs and forecasts feed both workspaces, so the conversation starts from agreed numbers |
Deterministic, so the numbers hold under pressure. A war team cannot argue over figures that change when you ask twice. Every K-Φ output traces back to accounts, formulas and inputs, which is what auditors, boards and lenders also need.
Built from the models you already trust. In many companies the best treasury and FP&A logic already exists in a spreadsheet, often built by a consultant. K-Φ compiles that logic into governed, audited software. It keeps the expertise and removes the fragility.
Why this is necessary, not optional
AFP framed the two functions as co-pilots [1]. Co-pilots fly from the same instruments. Finance teams that keep treasury and FP&A on separate models will keep holding the same reconciliation meeting, in every crisis and every planning cycle. Teams that share one governed model spend that time deciding. That is the case for K-Φ.