Cash flow forecasting and runway from real numbers

Most small and mid-market companies do not fail because they were unprofitable on paper. They fail because they ran out of cash while a spreadsheet still said they were fine. Profit is an opinion formed over a period; cash is a fact measured at a moment. A cash forecast is the bridge between the two — and for most teams that bridge is built by hand, in a spreadsheet, from numbers that were already stale when they were typed in.

Cash-flow forecasting and runway are core jobs of FINMOZG's Treasury Agent. Instead of asking you to maintain a model, it computes the forecast continuously from the balances and flows you have already posted — cash position, net flow, projected balance, runway in months, funding gap and a graded liquidity-risk signal — so the number you see is the number your ledger actually supports.

Why most cash forecasts are wrong by the time you read them

A spreadsheet forecast has a structural problem: it is a snapshot of assumptions, disconnected from the ledger that records what really happened. The moment a customer pays late, a supplier invoice lands early, or payroll shifts a day, the model and reality diverge — and nobody re-keys the change until the next review. So the forecast drifts, quietly, between the day it is built and the day a decision depends on it.

Three failure modes recur in hand-built models:

• Stale inputs. Opening balances and recent flows are typed in once and rarely refreshed, so the forecast describes last month, not today.
• Optimistic timing. Inflows are assumed to arrive on the invoice date rather than when customers actually pay, pulling the projected balance higher than reality.
• No alarm. A spreadsheet does not warn you. You only learn the runway is short when you next open the file — which may be after the decision window has closed.

What "runway" really measures

Runway is deceptively simple to say and easy to get wrong. It is the number of months your current cash can sustain operations at your recent net burn rate. Two real quantities define it: your cash position — what is actually in the bank right now — and your net flow, the difference between money coming in and money going out over a recent window.

The traps are in the inputs. If you measure cash position from an invoiced-but-unpaid figure, you overstate what you have. If you measure net flow from a single unusual month — a large one-off receipt, a delayed tax payment — you flatter or scare yourself with noise. Runway is only honest when both numbers come from posted reality and the burn rate reflects a representative window, not a lucky or unlucky month.

How FINMOZG forecasts from real data

FINMOZG's forecast is deterministic, not a guess from a language model. It is computed by an engine that reads the posted ledger and arithmetic that you can trace. That distinction matters: a number a board or a bank relies on cannot come from a black box. The Treasury Agent assembles the forecast from quantities the bookkeeping engine has already posted and evidenced.

The quantities it derives

• Cash position — the consolidated balance across connected accounts, taken from posted bank data rather than an estimate. In the Ukraine pack, ПриватБанк and Монобанк connect via their APIs, so the bank side arrives without manual export.
• Net flow — posted inflows minus posted outflows over the recent window, giving the true burn or build rate.
• Projected balance — the cash position carried forward at that net flow, showing where the balance heads if nothing changes.
• Runway — how many months the projected balance lasts before it crosses zero.

Because every input is posted and evidenced, the forecast updates itself as transactions land. You are not maintaining a model; you are reading one that maintains itself. The same discipline powers the CFO copilot, which answers strategic questions against these same numbers.

Funding gap and liquidity risk

Knowing the runway is short is necessary but not actionable on its own. FINMOZG turns it into two decision objects. The funding gap answers "how much?" — the additional cash you would need to reach a target runway or cover a projected shortfall. It sizes the hole so you can size the raise, the bridge, or the cut.

The liquidity-risk signal answers "how urgent?" It is a graded reading — low, watch, high — derived from how close the projected balance comes to zero and how soon it gets there. A company with twenty months of runway and one with two months both have a forecast; only one has a high-risk grade. Grading the signal stops every forecast from feeling like an emergency and makes the genuine ones impossible to ignore.

Scenario sensitivity

A single forecast hides how fragile it is. The honest question is not "what is the runway?" but "what breaks it?" Because the forecast is computed from real flows, you can test sensitivity against the levers that actually move cash — a slower collection cycle, a delayed round, a step up in payroll, a large supplier payment — and watch the projected balance, runway and risk grade respond. The value is seeing which variable your survival is most exposed to, while there is still time to act on it.

Seeing the wall before you hit it

The point of a live forecast is not a prettier chart. It is the alert that fires before the wall, not at it. Because the Treasury Agent recomputes continuously, it can raise a warning the moment the projected balance or runway crosses a threshold you set — early enough to collect a receivable, delay a discretionary spend, or open a funding conversation while you still have leverage rather than none.

Crucially, the agent forecasts and warns; it does not move money. Outbound payments and drawdowns are a hard human boundary in FINMOZG — they never auto-execute, regardless of confidence. The forecast informs the decision, an immutable audit log records who decided and on what evidence, and a person stays in command of every transaction. You can read how that control model works across the whole platform on the product overview, and the wider treasury picture in treasury and liquidity management.

Cash-flow forecasting in FINMOZG is not a model you babysit. It is a forecast computed from what you have actually posted, graded so the real risks stand out, and wired to warn you in time to do something about them.