Set a Factory Cash Reserve from a Stress Test, Not a Rule of Thumb

A cash reserve should protect a defined promise

A garment factory may show cash in the bank and still struggle to pay for a repair because part of its balance is restricted or already committed to payments issued. A large buyer receipt due next week may also arrive after the next wage payment.

The useful question is therefore quite specific: how much usable cash must the factory hold to meet its chosen payment commitments under a stated stress?

A reserve policy answers that question by linking cash, credit access and agreed actions to a period of risk. It also assigns responsibility for responding when that protection starts to weaken. A fixed number of months may provide an initial discussion point, but it cannot show whether the factory survives its own pattern of receipts and payments.

The worked case below builds that link. All amounts, events, rates and facilities are hypothetical. Values use currency units, or CU. The example explains a method; its reserve level is not an industry benchmark.

Decide what the reserve must cover

Start with the obligations that the business intends to meet on time, including wages, taxes, agreed supplier payments, utility bills and debt service. Their dates and amounts belong in the cash schedule, with any lawful, agreed changes shown separately.

A reserve does not make an obligation optional or give the factory permission to delay a contractual payment. Before treating a delay as a possible response, the finance team needs to establish what the relevant terms allow.

The protection period should reflect both the expected duration of the chosen disruption and the time needed to make a useful response effective. A factory that can obtain approved funding in two days faces a different gap from one that needs several weeks to arrange it. That response time must include the steps needed to get spendable cash, rather than the date on which someone promises to consider a request.

The Association of Corporate Treasurers' liquidity framework uses stress scenarios to examine funding needs. Its guidance also stresses the practical conditions for drawing backup finance. These are useful principles for a factory, even when its cash model is much smaller than a corporate treasury system.

Keep the floor separate from the stress cushion

Three amounts often get called the cash reserve. They serve different purposes:

• The operating floor is the minimum cash balance management wants to preserve.

• The base funding need covers the expected timing gap in normal receipts and payments.

• The stress cushion covers the extra gap caused by the selected disruption.

These amounts must fit together without counting the floor twice. A statement such as “we need 50,000 for operations and another 50,000 for safety” is hard to assess unless both amounts have a clear basis.

For this case, management chooses a 10,000 CU cash floor and a six-week test period, assuming it can review the position and arrange a new plan within that period. These are illustrative policy choices that require operational support before anyone adopts them. Before adopting such a horizon, a real factory would need evidence that its response could take effect in time.

Establish how much cash is usable

The bank balance, the financial statement cash figure and the amount available for this reserve can differ.

Suppose a factory sees 100,000 CU across its bank accounts, of which 20,000 CU must remain available for a separate contractual purpose throughout the six-week test. A further 5,000 CU covers payments already issued but not yet cleared. Its opening cash available to the reserve is therefore:

Usable opening cash = 100,000 − 20,000 − 5,000 = 75,000 CU.

The future payment schedule excludes that 5,000 CU because deducting it from opening cash and including it again as a future payment would count the same outflow twice.

Excluding the 20,000 CU from a management reserve does not, by itself, decide its accounting classification. The IFRS Interpretations Committee's demand-deposit decision addresses a particular fact pattern. A third-party restriction on use does not necessarily stop a demand deposit from being cash under IAS 7. The deposit's nature and the actual terms matter.

For reserve planning, ask the more immediate question: may this money pay these obligations at the time required? A balance can remain visible in accounts while being unsuitable for that purpose.

Access also depends on the account holder, currency and payment route, so money in one group company cannot simply be treated as available to another. A planned transfer may depend on approval, currency conversion or supporting documents. Include it only when those steps and their timing have a sound basis.

Build the base case before adding a shock

The factory's six-week base schedule contains the following receipts and payments. All values in the tables are thousands of CU. Thus, 45 means 45,000 CU.

Receipts total 425,000 CU and payments total 410,000 CU. The period ends with a net inflow of 15,000 CU, yet the factory first needs to fund a 25,000 CU gap.

For this illustration, each week's listed receipts clear before that week's payments, and there are no other cash movements. The opening balance and balances after those payments therefore identify the relevant low points. This assumption makes the weekly table usable for the calculation. It would be unsafe to assume the same order in a real payroll week without checking dates.

The starting cash needed to preserve the floor is:

Required starting cash = cash floor − lowest cumulative net movement.

Include the opening checkpoint, where cumulative movement is zero, when finding that minimum. Here, the minimum is −25,000 CU, so the base need is:

10,000 − (−25,000) = 35,000 CU.

The base opening requirement of 35,000 CU already contains the expected draw of 25,000 CU and the 10,000 CU floor, before any stress cushion is added.

This calculation uses payment timing. It does not use profit, the value of inventory or a sales target as a substitute for receipts. The British Business Bank's reforecasting guide makes the distinction between a cash forecast and a profit forecast. The distinction matters most when a healthy order book ties up cash before buyers pay.

Test the disruption as a dated cash event

The factory now tests two plausible events as planning scenarios chosen to expose specific weaknesses, without assigning a probability to either event.

In the first event, a buyer delays 60,000 CU from week 2 to week 5. The week 2 receipt falls from 70,000 to 10,000 CU, while week 5 rises from 80,000 to 140,000 CU. Total receipts stay at 425,000 CU because this scenario assumes a collection delay without any loss of the sale.

The second event has two parts. A supplier requires 30,000 CU in week 1 instead of week 4. An urgent repair also costs 12,000 CU in week 1. Week 1 payments rise to 112,000 CU, and week 4 payments fall to 30,000 CU. Total payments become 422,000 CU.

Only the repair adds to the six-week cost, while the earlier supplier payment moves cash between weeks and must not be added again to the period total.

Without a new facility, the combined case needs 132,000 CU of starting cash to cover cumulative movements of −67, −122, −117, −72, 3 and 3 thousand CU. Its deepest funding gap occurs in week 2.

Adding the two standalone cash requirements would give 167,000 CU because shared base needs and the floor would be counted more than once. Combining the events within one schedule avoids that duplication.

Events may also be linked. A loss of customer confidence might slow collections just when suppliers reduce credit. PwC's corporate financial resilience guidance discusses these pressures alongside the risk that funding access may weaken. The practical lesson is to test combinations that have a credible cause, rather than treating every shock as independent.

Give a credit facility a date, a limit and conditions

A reserve can consist of cash plus usable backup funding, provided the schedule shows when each resource becomes available for the payments being protected.

Assume the factory has an unused, committed facility of 40,000 CU for which all draw conditions have been checked and met. The cash can reach the account before week 2 payments, but it cannot arrive during week 1. The full principal must be repaid in week 6, with 1,000 CU of assumed interest and fees.

Existing payment commitments are already included in the base schedule, but these new finance movements are excluded so they can be shown separately. No other borrowing is assumed.

The ACT's discussion of corporate debt treats unused committed facilities as part of liquidity planning. Commitment alone, however, does not remove the need to check the actual draw terms, dates and repayment duties.

Work through the facility, including repayment

Adding a 40,000 CU draw in week 2 and a 41,000 CU payment in week 6 changes the combined stress case:

The lowest cumulative movement is now −82,000 CU. Required starting cash is therefore 10,000 − (−82,000), or 92,000 CU.

With 25,000 CU remaining after week 1 payments, the factory does not need the facility before its assumed draw date. The binding cash constraint occurs in week 2, when the balance reaches the 10,000 CU floor. By week 6, the facility and its cost have been paid, leaving 94,000 CU.

The extra cash cushion above the base requirement is 92,000 − 35,000 = 57,000 CU. Without the facility, that cushion was 97,000 CU. This 40,000 CU reduction in required cash depends on the credit being available on the assumed date and under the verified conditions.

Under the stated stress, management can choose between holding 132,000 CU of cash without this facility and holding 92,000 CU while relying on the tested 40,000 CU facility. Calling both “a 92,000 reserve” would hide a material difference in risk.

Compare the policy with today's available cash

The factory currently has 75,000 CU of usable cash. Its conditional requirement is 92,000 CU, leaving a 17,000 CU funding gap.

With only 75,000 CU at the start, the combined scenario leaves 8,000 CU after week 1 payments. That is 2,000 CU below the chosen floor. Even after the week 2 facility draw, cash would fall to −7,000 CU. The shortfall against the floor would then be 17,000 CU.

An extra 17,000 CU that arrives after week 2 payments cannot solve the earlier breach, which is why this case requires any top-up to reach the factory at the start.

If the facility becomes unavailable, the required opening cash rises to 132,000 CU. Against today's 75,000 CU, the gap would be 57,000 CU. Management should know that fallback figure before it approves reliance on the credit line.

Check the payment day when the weekly view looks safe

The order of cash movements deserves a separate check, as another short example shows. In this case, which is separate from the six-week figures, a factory opens a week with 12,000 CU. Payroll of 25,000 CU leaves on Wednesday, while a buyer's 30,000 CU receipt clears on Friday. The week closes at 17,000 CU.

Although the closing balance looks safe against a 10,000 CU floor, cash falls to −13,000 CU on Wednesday. Preserving the floor would therefore require another 23,000 CU before the payroll payment. Friday's receipt cannot pay Wednesday's wages unless some other approved funding bridges the gap.

This is also why a transfer's value date matters more than the date shown on an email. Use the point when funds become available for the required payment. For a critical date, confirm the payment cut-off and the person who can approve the transaction. A holiday or an absent signatory may turn an apparently adequate reserve into a late payment.

The reserve worksheet need not show every small transaction separately. It does need enough detail to expose the largest movements and any clustered due dates. Daily checkpoints around payroll, debt service and large material payments often answer a different question from the weekly management total. Keep the detailed view tied to the same opening cash and totals so that the two reports reconcile.

Ask what would break the plan

A reverse stress test exposes the limits of a chosen reserve by starting with the available resources and asking what would exhaust them.

Keep the 92,000 CU starting cash and the same 40,000 CU facility. Now assume no customer receipts arrive in the first two weeks, while base payments remain due. After week 1, cash is 92,000 − 70,000 = 22,000 CU. In week 2, the draw arrives, but payments then reduce cash to 22,000 + 40,000 − 65,000 = −3,000 CU.

The plan breaches its 10,000 CU floor at the week 2 payment checkpoint, although identifying the exact day of failure requires a dated payment schedule.

The original reserve calculation remains valid for its stated assumptions, but the reverse stress test shows that it does not cover a total stop in receipts over this interval. Management must decide whether that exposure is acceptable, whether it needs more funding, or whether a credible action can reduce it.

Because these scenarios contain chosen assumptions rather than a measured distribution of future events, they do not establish a precise probability of survival.

Use evidence that can change a decision

A stress amount should have a traceable basis. For the buyer delay, examine actual payment terms, recent disputes, overdue balances and the share of receipts due from that buyer. A large invoice should not become a certain receipt simply because its due date appears in a contract.

Forecast errors can help show where the plan is fragile. ApparelDiary's guide to MAE, WAPE and signed bias explains why total error and directional bias answer different questions. For cash planning, keep late receipts distinct from receipts that are no longer expected. A delay changes dates; a loss changes both dates and totals.

The repair scenario needs a different evidence trail, using maintenance records, supplier quotations and spare availability to support its cost and likely timing. The guide to critical spare-parts decisions helps connect stock choices with the effect of a failure. A reserve should not quietly assume a spare is available when the factory has neither stock nor a confirmed delivery route.

Each proposed response needs evidence, so a supplier extension belongs in the plan only after there is a credible basis for the relevant agreement. Selling inventory may take time and may return less cash than its book value. Canceling a project may create a fee. Show those effects in the same schedule rather than treating every action as immediate, free cash.

Challenge the stress without pretending to predict it

A scenario can be useful without being the most likely outcome: the buyer delay tests reliance on one receipt, while the supplier change tests the value of existing payment terms. The repair adds a separate demand that is not recovered later in the schedule.

Review these assumptions with the people who hold the evidence. The sales team may know that an invoice is disputed rather than merely late. Maintenance may know that the quotation covers parts but excludes installation. Purchasing may know that a supplier's new terms apply to every open order, not just the next delivery. Each finding changes a different part of the model.

Recording the reason for accepting or rejecting an assumption prevents a later reviewer from mistaking a convenient estimate for a confirmed fact. It also lets the team focus its next review on changed evidence rather than rebuild the whole case from memory.

The hardest useful question is often what has been left outside the test. A six-week horizon may miss a large payment in week seven. Extend the view when a known commitment or recovery action falls beyond the chosen end date. A reserve that passes only because the worksheet stops early provides weak evidence of protection.

Weigh carrying cost without losing the purpose

Holding more cash has a cost, but a cheap reserve that fails at the first payment gap is poor protection.

Consider a separate annual cost illustration. Suppose an extra 40,000 CU of cash is funded by debt costing 8% a year and earns 2% a year. These are assumed rates, not current market quotes. On a simple annual basis, debt interest is 3,200 CU, cash income is 800 CU and the net carrying cost is 2,400 CU, before fees and taxes.

This annual comparison is separate from the six-week facility example and its assumed 1,000 CU cost. Combining the two without matching their periods would be misleading. If the cash comes from existing equity rather than new debt, the same borrowing-cost calculation does not automatically apply.

The decision should compare this cost with the protection actually gained. It should also examine the cost and reliability of other available funding. Chasing a higher cash yield can change access or risk, so the reserve policy must keep the required payment access clear. The ACT's guidance on board treasury policies links these choices to an agreed appetite for risk.

Turn the calculation into a working policy

The approved policy should name the cash floor, stress horizon, tested events and allowed funding sources, with a clear distinction between cash and facility-dependent protection. Recording that basis gives the owner a way to explain why the requirement has changed.

A useful trigger comes before a missed payment. If a buyer disputes the week 2 invoice, the team should rerun the affected dates at once. If a draw condition changes, the facility should no longer appear as certain cash until the issue is resolved. A fixed review date is helpful, but it cannot replace a review after a material event.

After a reserve is used, a strong receipt may make the balance look comfortable again, as the worked case shows. Week 5 cash reaches 135,000 CU, but 41,000 CU is still committed to the week 6 facility settlement. Releasing that cash before checking the remaining schedule would defeat the policy.

Replenishment should use an updated forward view. Check new orders, wage dates, material commitments and the status of backup funding. Set who may approve a temporary shortfall, what action they must record and when the revised plan returns for review. Keep a second authorised person able to act if the usual approver is absent.

The final reserve is a documented choice about protection, linking usable money and payment commitments to credible shocks and timely action. With that basis in place, a factory can explain what its buffer covers and where the cover ends.

References

• Association of Corporate Treasurers: Building a robust liquidity risk framework.

• IFRS Interpretations Committee: Demand deposits with restrictions on use arising from a contract with a third party, April 2022 decision, updated January 2026.

• British Business Bank: Reforecasting in uncertain times.

• PwC: Corporate financial resilience—recent financial market events, April 2023.

• Association of Corporate Treasurers: Treasurers in the driving seat for corporate debt.

• Association of Corporate Treasurers: How treasurers and NEDs can inform the board on treasury policy.