It's Monday, 8:30 AM. The CEO calls: "the bank wants to know if we draw the credit line before Wednesday. What do I tell them?"
There are two versions of this call. In the first, you open your 13-week cash forecast — last refreshed Friday — look at week 4, and answer in 90 seconds: "Yes, $3M, before Thursday. Reason is the large Sodimac payment plus the 15th payroll." In the second, you say "let me check and call you back this afternoon."
The difference between those two versions isn't talent. It's a 45-minute weekly ritual that separates the treasurer who sees problems 6 weeks ahead from the one who finds out the day they happen.
By the end of this module you'll have it down. Let's go.
Opening cash $4M. Safety floor $2M. This week: $1.4M in (collections), $2.4M out (payroll $480K, debt $250K, suppliers $1.65M). Do you close above or below the floor?
In plain language
Before the mechanics, the four basic questions.
Why do this at all?
To not go bust by surprise. The P&L tells you if the business is making money. The cash forecast tells you if you're going to be able to pay people on Friday the 15th. Two different questions. Profitable companies go bankrupt every year because they have no 13-week cash visibility. This is the radar.
Who uses it?
Treasury builds and maintains it weekly (or the CFO in smaller companies). It's used by: the CFO to take to the board, the CEO to talk to the bank, the VP Finance to decide whether to delay a supplier payment, the controller to sync with AP. In a mid-market company, 5-7 different people open this file in a single week.
When does it show up?
Every Monday (or Friday) in a 45-60 minute session. Additionally: every time a cash-moving decision shows up (large payment, investment, capex), before any bank conversation about credit lines or covenants, and before each month-end close to validate the projected position.
What if we skip it?
You live in reactive mode. You find out about problems the day they happen, not 6 weeks before. The credit line you ask for when you no longer have leverage to negotiate. Supplier payments slip without warning (which destroys your payment reputation). And a bad operational month becomes a liquidity crisis because nobody saw it coming.
Meet Andina S.A. (from the treasury seat)
Same Andina from Pillar 2 — LATAM coffee company, $200M revenue, based in Chile. But today we sit in the treasurer's chair, not FP&A's. The question changes: not "how is plan vs actual" but "is there cash in the account the day I need it?"
Andina operates with $4M opening cash, has a $2M operational floor (below that the business gets fragile), and a committed $5M credit line they'd rather not touch because it costs and reports to the bank.
The big outflows are clear: biweekly payroll ($480K), monthly debt service ($250K), and the large raw-material supplier payment in week 4 ($1.4M, vs ~$820K other weeks). Inflows are B2B customer collections with typical 60-day terms.
In the base case, the business breathes. Under stress (AR aging slipping by 10 days, which happens all the time in LATAM), things change. The visual below shows exactly how much.
13 weeks of cash, live
Thirteen weeks, four flow groups (collections, payroll, suppliers, debt), a cumulative cash line and a safety floor. It's the cleanest version of the file your treasurer opens every Monday.
Hit Stress: AR +10 days and watch what happens: a typical collections slip pushes the cash curve under the floor for three weeks. This isn't theoretical. It's the most common scenario that destroys liquidity in LATAM companies.
What matters isn't the numbers — it's the highlighted weeks. If a week lights up red in your file, that's the week the board needs to hear about this Monday, not next.
Interactive visual
Andina S.A. — 13-week cash forecast
Opening cash
$4.0M
Ending cash
$7.0M
Weekly low
$4.0M
Weeks below floor
—
What you see in the base case
The business clears 13 weeks above the floor, but week 4 is tight: large supplier payment lands together with payroll and debt service. Narrow margin. This is exactly when the CFO should be looking at this table every Monday.
What does a treasurer do when they see this?
Accelerate week 3–4 collections. Negotiate the large supplier payment in week 4 (can it be split?). Pre-arrange a $1M credit line before week 4 just in case.
The difference between "clears above floor" and "punches floor for 3 weeks" is 10 days of collection delay. Not 30, not 60. Ten. That's why the weekly ritual matters: you spot the slip in the first week it diverges, not the fourth.
The treasurer who looks at this every Monday draws the credit line in week 1, not week 4. Accelerates collections on the top 3 overdue accounts. Renegotiates the large supplier payment to split into two. All of that happens with time. Without the file, it happens under pressure.
The mechanics: how to build it well
- 13 weeks, not 12, not 14. The standard is 13 because it covers a full quarter + one buffer week and aligns with monthly payment cycles. Shorter and you lose visibility. Longer and projection quality drops fast.
- Weekly granularity, not monthly. Payroll lands on the 15th and last day. Debt lands on the 5th. Large suppliers cluster in specific weeks. If your view is monthly, you miss the dips inside the month — and the dips are the problem.
- Collections = top-down + bottom-up. Top-down: monthly revenue / 4.3 weeks as sanity check. Bottom-up: actual AR aging, top-10 customer outreach, adjustment by real days-to-pay. The gap between the two methods is your error margin.
- Outflows have 3 tiers of certainty. Tier 1 (must pay): payroll, debt, taxes. Tier 2 (committed): suppliers with signed POs. Tier 3 (discretionary): new contracts, capex. When stress shows up, you cut Tier 3 first.
- Refresh every Monday. You compare the closed week (actual) vs what you projected 7 days ago. That gap tells you if your model is calibrated. If you see systematic >5% errors, the model needs adjusting, not the result.
- One row up top: end-of-week cash position. That's the line the CEO will read. If it's above the floor, all good. If below, we talk.
- Annual budget = financial commitment for the year, monthly granularity, used to set targets and allocate resources. Doesn't work for liquidity (weekly dips don't show up).
- Cash flow statement (accounting) = accounting view of how cash was generated in a closed period, based on activities (operating / investing / financing). Retrospective, not operational.
- 13-week cash forecast = operational view, weekly, prospective, granular about who pays what when. The liquidity tool. Treasury builds it with AR/AP data, not accounting with GL data.
- Practical rule: if your CFO shows you the "quarterly cash flow" when you ask about liquidity, they're showing you the statement, not the forecast. Different tools.
Adversarial check
Adversarial check
1.The CEO asks you Monday 8AM: "can we pay the annual employee bonus this Friday?" The bonus is $1.8M. Your 13-week shows end-of-week cash at $2.4M (floor: $2M). What do you answer?
2.Your team shows you the refreshed 13-week on Monday. The collections line says $1.4M for the week, same as the last 4. Last week's actual was $1.1M. What do you do?
3.What's the main difference between the 13-week cash forecast and the accounting cash flow statement?
Exit checklist
Suggested re-review: 90 days, or when a forecast week lights up red for the first time in the year.
Optional
Go deeper
Sources and books to dig into the original material