13-Week Cash Flow Forecast: How Singapore SMEs Fund Q4 Without a Crunch

If your business sells into the year-end peak, your cash position between now and Christmas is not a smooth curve. It is a trough followed by a spike, and most Singapore SMEs hit the bottom of that trough in the second half of October — after the stock is paid for, before the sales are paid out. A 13-week cash flow forecast started this week covers 21 September to 20 December and answers one question precisely: can you fund the peak you have already planned? It works by listing every committed outflow week by week against realistic inflow dates, which are almost never the dates you invoice. Built properly it takes a day to set up and twenty minutes a week to maintain, and it is the cheapest financial control available to an SME this quarter.

Why does a profitable Q4 still run out of cash?

Because profit and cash separate under seasonal load. In a normal month you buy stock, sell it, and collect within a fairly stable cycle. In Q4 that cycle stretches at both ends simultaneously.

On the outflow side, suppliers want deposits earlier and larger. Factory orders placed ahead of China's Golden Week and the Chinese New Year 2027 shutdown typically carry 30% deposits paid at order and the balance before shipment — money out in September and October for goods that generate revenue in November or February. Marketplace and search advertising is prepaid or charged on short cycles, and it peaks exactly when your bank balance is thinnest. Freight rates rise into the peak, and air freight, if you end up needing it, is a cash event with no notice.

On the inflow side, payment lags widen. Marketplace payouts are released after the buyer protection window closes, not when the order ships — realistically 7 to 16 days after delivery, which for an 11 November sale means cash in late November. B2B customers slow down in December as their own finance teams wind down for the year. Returns from the peak land in December and January and reverse revenue you have already spent against.

Layer statutory payments over that and the picture sharpens. The result is an SME that ends the year with a strong P&L and a genuinely uncomfortable November.

What belongs in a 13-week forecast?

Keep it to one tab and three blocks. Complexity is what kills these forecasts by week four.

• Opening bank balance for each of the 13 weeks, carried forward from the prior week's closing position — actual balances, not accounting balances, and across every account you genuinely draw on.
• Inflows by expected receipt date: marketplace payouts by platform, customer collections from your aged receivables, and any grant disbursements or tax refunds you are reasonably confident of. Date these by when the money clears, not when you raised the invoice.
• Outflows by committed payment date: supplier deposits and balances, freight and duty, payroll, CPF, rent, advertising, software subscriptions, loan repayments, GST, corporate tax and year-end bonuses.

Two disciplines make it useful. First, separate committed from discretionary outflows in different colours — the committed rows are your real floor, and the discretionary rows are the levers you pull when a week goes negative. Second, show the closing balance as a running line and mark your minimum acceptable buffer on it. The forecast's job is not to be accurate to the dollar; it is to show you which week goes below the line and how many weeks of warning you have.

Which Singapore dates should already be on it?

These are fixed, they are knowable today, and they are the ones most commonly missed:

• 14 October, 14 November, 14 December — CPF contributions for the preceding month. December's is larger if you pay bonuses in November.
• 31 October — GST return and payment for the quarter ending 30 September, if you file quarterly on the Mar/Jun/Sep/Dec cycle. GIRO deductions fall on the 15th of the following month instead.
• 30 November — Form C-S or Form C filing deadline. The filing itself costs nothing, but the assessment that follows does, and tax payable is due within one month of the Notice of Assessment unless you are on a GIRO instalment plan agreed earlier in the year.
• Late November to December — AWS, 13th month and performance bonuses. Remember the CPF Additional Wage ceiling: bonuses can push employer CPF higher than the monthly run rate suggests.
• December to early January — Chinese New Year 2027 supplier deposits. CNY falls in mid-February 2027, factories stop earlier than that, and deposits are typically demanded four to six weeks ahead of production.
• 31 January 2027 — GST for the quarter ending 31 December, sitting just outside the 13 weeks but worth carrying as a note so it does not ambush the next forecast.

How do you populate it without re-keying everything?

Manual forecasts die because maintaining them is someone's Friday afternoon. Three sources do most of the work automatically.

Your accounting system already holds aged receivables and unpaid supplier bills with due dates — export both weekly rather than retyping them. Your marketplace seller centres expose settlement and payout reports that show pending amounts and release dates; these can be pulled on a schedule. Your bank feed, or a CSV export, gives you the actual opening balance that anchors the whole model.

Start in a spreadsheet with those three feeds landing in a raw data tab and a single formula layer summarising them into weeks. That is enough for most SMEs, and it is honest about where the numbers came from. Only move to a dedicated cash flow tool once you have run the spreadsheet for a full quarter and know which assumptions you actually argue about — buying software to answer a question you have not yet framed is how SMEs end up with a subscription nobody opens.

What decisions should the forecast drive?

A forecast that does not change a decision is a report. This one should force four, and force them this month while the options are still cheap.

How much peak stock can you actually fund? If week 6 goes below your buffer, the order you place in early October is the problem, not the sales you make in November. Second, do you need a facility? Arranging a trade finance line or an overdraft takes weeks; asking in November, with a negative forecast in hand, gets a worse answer than asking in September with a plan. Third, can you reshape terms? A supplier deposit moved from 30% to 20%, or a payment date pushed two weeks, often costs nothing but a conversation — and you only know which conversation to have once the weeks are laid out. Fourth, which discretionary spend is genuinely deferrable? Better to consciously delay a software rollout to January than to discover in week 9 that you are choosing between ad spend and payroll.

Frequently asked questions

How accurate does a 13-week forecast need to be?

Directionally accurate is sufficient. Aim for inflow timing within a week and outflow amounts within 10%. The forecast exists to identify which week breaches your buffer and by roughly how much, not to predict a closing balance to the cent. Over-engineering the accuracy is the most common reason these models get abandoned.

Should I use my accounting software's built-in cash flow report instead?

Those reports are usually backward-looking or project forward from invoice due dates, which overstates inflows because customers and marketplaces do not pay on the due date. Use the software as the data source and keep the forecast layer separate, where you control the timing assumptions.

We are a services business with no stock — is this still relevant?

Yes, though the shape differs. Your Q4 pressure comes from bonuses, CPF on additional wages, the corporate tax assessment following the 30 November filing, and December collections slowing while payroll does not. The trough is shallower and later, but it is still there, and the same 13-week view finds it.

If you want a second pair of eyes on your Q4 cash position — or help wiring your accounting and marketplace data into a forecast that updates itself — that is a conversation worth having before October, not during it.