How to read a 90-day cash forecast

The first time most people open a cash forecast, their eye goes straight to the number on the far right. That is the least interesting thing on the chart. It is the end of a ninety-link chain of assumptions, and every link between here and there is a place where reality gets a vote.

The forecast earns its keep earlier in the line. Here is the order worth reading it in.

1. Start at the low point, not the end

The single most useful figure on a 90-day chart is the lowest balance the line touches, and the date it touches it. That is the moment the business is thinnest. Everything after it is recovery; everything before it is a countdown.

A company can end the quarter comfortably and still have a Tuesday in November where payroll clears three days before the biggest receipt lands. The ending balance hides that. The low point is the whole point.

2. Read the shape before the level

Cash lines have a vocabulary, and once you can read it you can diagnose a business from across the room.

What you see	What it usually is
Regular sawtooth, even teeth	Payroll on a fixed cycle
Steps up on the same days each month	Retainers or subscriptions landing
One tall cliff	Tax, insurance, or an annual renewal
A long flat drift downward	Day-to-day burn outrunning the retainer base
A jump that never repeats	A milestone or one-off invoice — do not plan on it twice

A line that drifts down between the teeth is telling you something different from a line that drops in one cliff. The first is a pricing or headcount conversation. The second is a timing conversation.

3. Treat the band as the honest part

A single line implies a confidence nobody has. The shaded range around the projection is the same arithmetic with every receipt arriving ten days late — which is not pessimism, it is the median experience of anyone who has ever sent an invoice.

If the low point stays comfortable inside that band, you have slack. If the band dips somewhere the line does not, you do not have a cash problem yet; you have a collections problem that will become one.

4. Ask what would have to be true

The forecast is most valuable as an argument starter. Take the low point and work backwards:

• What lands before it? If one receipt is carrying the week, name it, and find out where that invoice actually is.
• What could move? A lease payment can usually shift by a week. Payroll cannot. Sort your outflows into those two piles.
• What is missing? Feeds see what has already happened. The contractor invoice sitting in someone's inbox is not on the chart until you put it there.

Add the missing items and watch the line redraw. That gap — between the forecast as the feed sees it and the forecast as you know it — is usually where the real answer lives.

5. Know what it cannot do

A 90-day projection is arithmetic: balances, dated items, and a recent run rate carried forward. It does not know a client is unhappy, that a contract is up for renewal, or that your best account manager just resigned. It will happily extrapolate a good quarter into a great one.

Which is fine, as long as you read it as a floor plan rather than a photograph. The number is only ever as good as the items behind it.