Cash visibility and cash forecasting are frequently discussed as if they're the same capability or as if improving one automatically improves the other. They're not the same, and they don't automatically combine. Each solves a different decision problem, operates on different data, and has different accuracy characteristics. Understanding the distinction is important for treasury teams evaluating tools — both for choosing the right product and for setting accurate expectations about what any given solution will and won't deliver.
Cash Visibility: Where You Are
Cash visibility refers to the ability to see current bank account balances, positions, and transaction activity across all entities and banking relationships — ideally in real time or near-real time, aggregated into a single view. Good cash visibility answers the question: What is our cash position right now, by entity and in total?
The inputs to cash visibility are bank data feeds — BAI2 files, MT940/MT942 messages, or direct bank API connections. The data is historical in the sense that it reflects completed transactions; the "real-time" aspect is simply about how quickly the system pulls and displays that data. A system with intraday bank API connectivity and a 15-minute refresh interval provides near-real-time visibility. A system with prior-day BAI2 feeds provides visibility into yesterday's closing position.
Cash visibility is a necessary foundation for treasury operations. Without it, treasury teams spend hours each morning manually collecting position data — the portal-logging ritual that every mid-market treasurer knows well. But visibility alone is not forecasting. It tells you where cash is; it tells you nothing about where cash will be in three weeks.
Cash Forecasting: Where You're Going
Cash forecasting projects future cash flows — inflows from AR collections, outflows from AP payments and payroll, debt service, capital expenditures — over a forward horizon, typically 13 weeks for operational treasury purposes. It answers: Will each entity have sufficient cash to meet its obligations over the next 13 weeks, and where are the gaps?
The inputs to cash forecasting are not primarily bank data. They are AR aging (invoice-level, with payment timing models), AP aging (with approval status and timing assumptions), payroll schedules, debt service schedules, and any other committed or probabilistic future cash flows. Bank data feeds provide the actual opening balance that the forecast begins from, but the forecast itself is built on ERP data and operational information — not on bank transaction history.
This distinction matters for system design. A cash visibility tool that pulls bank data very well is not necessarily a good forecasting tool. A forecasting tool that has deep ERP integration and sophisticated AR/AP timing models is not necessarily providing real-time bank position data. Platforms that genuinely combine both capabilities have done two different integration and modeling jobs, not one.
How They Work Together in Practice
The interaction between visibility and forecasting is where the operational value compounds. Consider how a treasury team at a mid-size distribution company — eight entities, 14 bank accounts, Tuesday morning cash position meeting — uses both capabilities together.
At 8:00 AM, visibility provides the current opening position across all entities. The treasurer sees that the Wisconsin subsidiary is holding $3.2M and the Texas entity is at $1.1M, up from last week. That's a visibility output. At 8:15 AM, the forecast updates automatically with those opening positions as the starting balance for each entity's 13-week projection. The updated forecast shows that the Texas entity's $1.1M will drop below the $500K minimum operating floor in week 6 due to a $680K quarterly equipment lease payment and a seasonal AR slowdown. That's a forecasting output.
The actionable insight — that a $400K interco sweep from Wisconsin to Texas should be executed by week 4 to pre-fund the week-6 trough — could not have been generated by either capability alone. Visibility without forecasting shows today's healthy positions and generates no action. Forecasting without current positions either uses yesterday's stale opening balances (introducing a modeling error) or requires manual updates. Together, they drive a proactive funding decision that prevents a reactive crisis.
Why Most Platforms Optimize for One
The treasury technology market has historically split along the visibility-forecasting axis for a structural reason: the data engineering problems are different. Bank connectivity (visibility) requires relationships with financial institutions, SFTP infrastructure, format normalization across BAI2/MT940/MT942/API variants, and real-time or near-real-time data pipeline architecture. Forecasting requires ERP connectivity, AR/AP data modeling, time-series analysis, and forecast accuracy measurement infrastructure.
Legacy TMS platforms (the enterprise category: Kyriba, GTreasury, ION) generally have deeper forecasting and analytics capabilities, with bank connectivity as a complement. Newer API-first connectivity tools sometimes provide excellent real-time visibility but deliver forecasting as a simpler, less accurate add-on. The gap between what each category provides in its secondary capability versus its primary is often larger than marketing materials suggest.
We're not saying that any platform offering both capabilities should be dismissed as a jack-of-all-trades compromise. The argument is to evaluate each capability independently when assessing a platform: how does the bank data connect, at what latency, for which banking relationships? And separately: what is the AR/AP timing model, how is it calibrated to our customer payment patterns, and what accuracy benchmarks support the forecast claims?
The Decision Horizon Difference
Cash visibility and cash forecasting also serve different decision horizons, which is worth making explicit when treasury teams are mapping tools to workflows.
Visibility is the foundation for same-day and next-day decisions: does the operating account have sufficient funds for today's ACH file? Is there an unexpected large debit that needs explanation? Does the overnight position require an early-morning investment or funding action?
Forecasting supports medium-term decisions (1-13 weeks): when should the interco sweep be executed? Should we draw on the revolver now or wait three weeks? Is there a projected covenant breach in week 10 that the CFO needs to know about today? Should we accelerate a capex payment to capture a vendor discount, and does the cash position support it?
These are different conversations, with different audiences, happening at different points in the treasury day. A treasury infrastructure that delivers both — current positions in the morning, entity-level forward projections for the weekly CFO review — is what the decision cadence of a well-run mid-market treasury function actually requires. The combination isn't a luxury feature; it's the complete picture.