An advance is not free money

The request everybody says yes to

11 June 2026. Nabaruh's finance manager has looked at the same table you built in lesson 2 and reached the obvious conclusion. The factory is carrying USD 299,000.00 for months, and the cheapest money on the ladder is money that arrives before the goods do.

So Nabaruh asks Thurlemont for 30% with the order, balance under the sight credit as agreed. The proforma goes out on 8 July. Thurlemont's finance director replies on 14 July.

We can do the advance. Payment goes out on the 15th against your bank's advance payment guarantee for the same amount. Given that we are funding your production from July rather than from October, the price for this order needs to be USD 21.60.

Everybody at Nabaruh reads that as a win. Fifteen cents is not much, the cash arrives more than three months early, and the merchandiser records the concession in the order file as advance secured.

It is a win. It is a much smaller one than anybody in the room thinks, and by the following March it has turned into a loss. This lesson is the arithmetic that would have shown that in July, in about ten minutes.

The factory's side

The advance is 30% of USD 400,200.00, so USD 120,060.00, arriving on 15 July instead of 26 October. That is 103 days earlier on that slice of the money.

• Saved: 120,060.00 at 13.5% for 103 days is USD 4,573.79
• Cost: Rasheed Commercial Bank issues the advance payment guarantee at 1.4% a year on the guaranteed amount, running from 15 July to shipment on 14 October, which is 91 days: USD 419.06. Issuance fee USD 120.00
• Cost: the price reduction, USD 0.15 on 18,400 jackets: USD 2,760.00

A guarantee is the bank's own promise to hand the money back to the buyer if the goods never ship.

USD 1,274.73. Positive, so the decision was right. But it is worth 3.6% of the order's margin, not the windfall the room believed it had won. Nearly two thirds of the saving went straight back across the table.

That is the general rule, and it is why the first band of the ladder is not the free lunch it looks like:

An advance is a loan from the buyer to the factory, and buyers who are good at their jobs price loans.

The buyer's side, which is why the price moved

The number Thurlemont came back with was not a bargaining position. It was a calculation. It is worth being able to reproduce it, because a supplier who can reproduce the buyer's arithmetic argues from a completely different position.

Thurlemont borrows at 5.9%. On USD 120,060.00 for 103 days that is USD 1,998.92, which is less than the USD 2,760.00 they asked for. So they were not merely recovering their cost of funds. Their treasury policy prices a supplier prepayment at cost of funds plus two points, and the reason is not greed.

A receivable and a prepayment are different animals on a buyer's books. A receivable is a right to receive cash. It can be insured, it can be sold, and it sits inside machinery a finance department already has. A prepayment to a supplier is a right to receive goods. It is not a financial asset at all. It cannot be sold on, ordinary trade credit insurance does not reach it, and if the factory fails in September, what Thurlemont holds is a claim in an Egyptian insolvency for jackets that do not exist.

At 7.9% for 103 days the cost is USD 2,676.52, and Thurlemont asked for USD 2,760.00. They rounded the price to a sensible number and took USD 83.48 of margin on the arrangement. It is a fair price, worked out properly. It is also exactly why the advance guarantee was a condition and not a courtesy: the guarantee keeps that extra risk inside a bank rather than on Thurlemont's own books. Course 13.7 teaches guarantees and bonds in full.

The mistake nobody made

Now the part that cost real money. Notice that no single person in it did anything wrong.

July 2026. Nabaruh wins an advance and gives up USD 0.15 a jacket. Net gain USD 1,274.73. Correct decision, correctly taken.

October 2026. The order ships and is paid. The file is closed.

February 2027. Thurlemont's buying office builds next season's costing. It opens the supplier price file. The last agreed price for style CJ-118 at Nabaruh is USD 21.60, so that is the price the new range is costed against. Nobody is being sharp. A price file holds prices.

March 2027. The repeat order is confirmed at USD 21.60, on a sight credit, with no advance. The advance was a one-off arrangement for a season when Thurlemont's own cash was easy.

Nabaruh has now given up USD 2,760.00 on the repeat order and received nothing at all for it. The concession outlived its reason by one season, and it will outlive it again next year unless somebody argues the price back up. That is a much harder conversation than the one in July.

A price concession is permanent and a cash gap is temporary. Trading one for the other gets worse every season you do not reverse it.

The fix is not to refuse advances. It is one sentence in the July reply:

Agreed at USD 21.60 for this order only, on account of the advance. The list price for CJ-118 remains USD 21.75 and applies to any order not carrying an advance.

Seven seconds of typing. On the repeat order alone it is worth USD 2,760.00, which is more than twice what the advance itself earned.

Act as a treasury analyst who has worked on both the buying side and the supplier side, and has no interest in flattering either. I want an advance payment priced from BOTH sides, so I can see what it is really worth and reproduce the buyer's own arithmetic. My facts: order value [AMOUNT], quantity [QTY] pieces, agreed price [PRICE] per piece, on board [DATE], expected date the money arrives without an advance [DATE]. Advance requested [PERCENT] of order value, payable [DATE]. My working capital facility costs [RATE] a year. My bank will issue an advance payment guarantee at [RATE] a year on the guaranteed amount, plus an issuance fee of [AMOUNT], running from the advance date to [DATE]. The price the buyer is asking in exchange: [NEW PRICE] per piece. What I know about the buyer, if anything: their borrowing cost [RATE], their size, whether they are listed, and whether they have done this before. Do the following. First, price my side: the finance saved on the advanced slice, less the guarantee commission, less the issuance fee, less the price reduction across the WHOLE order. Give me the net in money and as a share of the order's margin. Second, rebuild the buyer's side: what the advance costs them at their own cost of funds, and what a sensible premium above that would be for holding a claim on goods rather than on cash, given that a prepayment cannot be insured or sold the way a receivable can. Tell me whether the price they asked is fair, greedy or generous against that rebuild. Third, find the price at which the advance becomes worthless to me, and the percentage advance at which it becomes worthwhile again at their asking price. Fourth — and do not skip this — tell me what happens to the reduced price NEXT season if I do not attach it to its reason in writing, and draft the one sentence that ties the concession to this order only. Fifth, list what could go wrong with the guarantee: what triggers a demand under it, and what I would owe and to whom. Show every calculation as amount times rate times days over 365, so I can check it.

What you should be able to do now

1. Price an advance net, after the guarantee and after whatever the buyer takes off the price. The gross saving is not the answer.
2. Reproduce the buyer's arithmetic. Their cost of funds, plus a premium for holding a claim on goods rather than on cash. That is where their number comes from.
3. Attach every concession to its reason, in writing, in the same message. For this order only, on account of the advance.
4. Go and look at your own price file for a price still carrying a reason that expired two seasons ago.