Cash Forecast Accuracy

A cash forecast predicts the dates and amounts of incoming and outgoing money, and accuracy asks how closely that forecast matched later cash movements over the same period. The comparison needs a stated boundary, timing convention and treatment of extraordinary flows.

CashAnalytics explains why an aggregate net-cash comparison can conceal offsetting receipts and payments errors, and Ripple Treasury describes absolute percentage error and directional measures, but there is no single universal score that answers every treasury question. Decide what is being forecast, because a closing bank balance differs from gross customer receipts, supplier payments or net movement and each answers a different operating question.

Pick a horizon too, since tomorrow's forecast can draw on scheduled bank transactions while a three-month forecast contains more estimates, and compare forecasts only with actuals for the same horizon. Freeze the forecast by keeping the version issued before the outcome was known, because editing last week's estimate after seeing actual receipts makes the result meaningless.

Align accounts, entities and timing. A group forecast may net subsidiary surpluses against deficits even when cash cannot be transferred freely, so measure the legal entities that actually need funding.

Weekly aggregation can hide a Tuesday shortfall offset by Friday's inflow, so daily assessment matters when payments are due before receipts arrive, and an incorrect opening balance carries through to the closing balance even if the period's receipts and payments were predicted well. Separate inflows and outflows, because accurate net cash could arise from both gross streams missing by the same amount, which is not a strong forecast.

Classify payments as well, since payroll and rent may be more predictable than tax, inventory purchases or project milestones, and segmenting helps assign owners to errors. Define the error measure: absolute error prevents positive and negative misses from cancelling, while a signed error shows whether the team systematically forecasts too much cash.

Use percentages carefully, because dividing by a near-zero actual cash movement produces a huge or undefined percentage, so present the amount error and a sensible scale alongside it. Link forecasts to evidence, since an expected customer receipt should have a due date, payment history and collection status and a manager's hope is not a committed bank deposit.

Compare like periods, because seasonal peaks, holidays and short weeks change the cash pattern, and backtest by horizon so that four-week-ahead forecasts are compared only with other four-week-ahead versions. Investigate large misses to find whether the cause was a late customer, an omitted payment, a timing shift or an assumption change, since the corrective action depends on the cause.

Watch the liquidity consequence, because a forecast miss near a credit limit can be urgent even if the percentage score seems small. For an owner, cash forecast accuracy is a test of planning reliability, and the useful result is not merely a high number but earlier warning of shortfalls and clear learning from misses.