You Bought in Dollars, Sold in Dirhams, and Booked a Profit That Included a Bet
· 7 min read · Written by Faceela Research & Editorial Team
Reviewed by Ahmed Hassan Algammal — Founder and Enterprise Systems Consultant
The dirham has been pegged to the US dollar for a generation, at a rate that has not moved in all that time. That one fact is the reason a UAE trading company can buy in dollars, sell in dirhams, and go twenty years without a single conversation about foreign exchange.
It is also the reason nothing changes on the day it starts buying in euros.
The purchase order is raised the same way. The margin is calculated the same way. The salesperson is told the same number. And somewhere between the order and the payment, the business has taken a position it never decided to take, held it for two or three months, and closed it into an account at the bottom of the profit and loss that nobody can read.
That is not a treasury problem. It is a reporting problem wearing a treasury costume.
The peg is a fact about one currency, not about the world
The peg does exactly what people think it does, and only for the currency it names.
| What you buy in | What the rate does | What you are actually holding |
|---|---|---|
| USD | Pegged to the dirham at a fixed rate | No exposure worth the cost of managing |
| SAR, QAR, OMR, BHD | Pegged to the dollar, and therefore to you | Effectively none |
| KWD | Pegged to a weighted basket, dollar-heavy | Small, and not zero |
| EUR, GBP, JPY, INR, TRY | Floating against the dollar, therefore against you | An open position from the day you commit |
| CNY | Managed against a basket, not fixed to the dollar | An open position, quieter than the others |
The trap is in the last two rows, and it is a trap of habit rather than of knowledge. Nobody in the business believes the euro is pegged. It is that the process built around dollar buying — quote from the PO, book the invoice, pay when it falls due, never look again — was correct for USD and is silently wrong for everything else.
Note the middle rows too, because they cut the other way. A Saudi receivable behaves like a dirham receivable, so a trading business with GCC customers usually has far less exposure on the selling side than it fears and far more on the buying side than it has noticed.
Four dates, and only one of them was there when it mattered
Every foreign-currency shipment has four moments at which a rate can be fixed, and they are days or months apart.
The purchase order is where the buyer converts to dirhams to compare suppliers, and where somebody sets a selling price. The supplier invoice is where the payable enters the ledger and the goods are received at a cost. The payment is where money actually leaves, at whatever the rate is that day. And month end is where every unpaid balance is restated at a closing rate, producing a gain or a loss that reverses and is charged again the following month.
Four rates, one shipment. The margin your reports show was fixed at exactly one of them.
One shipment, four rates
Every rate settles a number the decision no longer needs
Rate fixed at · Purchase order raised
1. The rate you quoted from
What this rate settles
The dirham figure in the buyer's comparison, the cost used in the price list, and the margin the salesperson is told he has.
The decision it is already too late for
Nothing yet. This is the only one of the four that arrives before a decision rather than after it — and it is an estimate, not a settlement.
Rate fixed at · Supplier invoice booked
2. The rate the payable is recorded at
What this rate settles
The liability in the ledger and the cost the goods are received at. From here the item's cost stops moving unless somebody moves it.
The decision it is already too late for
The order. It was placed, the goods are made, and the price was agreed against a different number.
Rate fixed at · Supplier paid
3. The rate the money actually left at
What this rate settles
The cash that went out, and the realised difference against the rate the payable was booked at.
The decision it is already too late for
The sale. On normal terms the goods are sold before the supplier is paid, so the margin was reported using the second rate and settled at the third.
Rate fixed at · Month end
4. The rate everything open is restated at
What this rate settles
Every unpaid supplier balance and every foreign-currency bank account, revalued to a closing rate — an unrealised gain or loss that reverses next month and is charged again.
The decision it is already too late for
All of it. Nothing here can be attributed to a product, a supplier or a buyer, because the revaluation is made on balances rather than on shipments.
The margin you reported used one of the four
It used the second, because that is the rate in the item cost on the day the invoice was raised. The third settled later and the fourth was never attributable at all. The difference between them is not an error — it is a position the business held, without anyone deciding to hold it.
The reason this is worth drawing rather than listing is that the same shape repeats at each stop and the repetition is invisible in prose. Each rate settles a number that will be used. Each one arrives after the decision it should have informed. And the one that produces the margin figure everybody steers by — the invoice rate — is the second of four, with two still open behind it.
Realised and unrealised are two different arguments
They get bundled into one line on the P&L, and they are not the same claim about the business.
An unrealised difference is a restatement. The supplier is unpaid, the balance is revalued at a closing rate, and the entry reverses at the start of the next period. It moves the reported profit of a month without moving a single dirham, which is why finance treats it as noise — correctly, for the month, and incorrectly for the year, because over a long payable cycle those restatements are the early warning that a position exists.
A realised difference is cash. It happened. The money that left the bank was more or less than the liability recorded, and the difference is permanent.
The distinction matters for one practical reason: a business that reports only the combined figure cannot tell whether it is looking at a timing artefact or at money. Both look identical on a variance report, and the one that is money is the smaller of the two in most months, which is exactly how it stays unexamined.
The account nobody can read
Here is the structural problem, and it is the same one as landed cost in a different suit.
FX gain and loss lands in one account, at the bottom of the P&L, as a monthly total. It is arithmetically correct and analytically useless, because it cannot be attributed. You cannot say which product line produced it, which supplier's terms caused it, or which buyer's decision it followed from. It has no dimension other than the month.
So three questions that any trading owner would want answered are unanswerable from the accounts as normally kept:
- Which supplier relationships cost us money on currency, independently of price?
- Which product lines carry an FX cost that is not in their gross margin?
- Did the euro supplier's better price survive the currency, or did we buy a discount and pay for it in translation?
None of those require a treasury function. They require the FX difference to be attributable to the shipment it belongs to — the same fix, exactly, as freight and duty. Once the difference lives on the shipment, the gross margin per item is right and the three questions answer themselves.
The forward nobody told the buying team about
The second half of this is organisational, and it is almost universal.
Finance sometimes buys forward cover on a large foreign-currency commitment. It is prudent, it is cheap relative to the exposure, and it fixes the rate for that payable. And in most trading companies, the buyer who created the commitment never learns that it happened.
That produces two failures at once. The buyer keeps quoting from spot, so his cost is wrong in the direction of whatever the market has done since — while a rate has already been contracted. And the cover is booked in treasury against a bank contract rather than against the purchase order it exists to protect, so the shipment's true cost is sitting in two systems that never meet.
The fix is not a policy document. It is that the forward contract has to be a record attached to the purchase order, visible to the person who negotiates the price. A rate that is locked and unknown is worth less than a rate that is floating and known, because at least the second one can be argued about.
What to make a vendor show you
On a live system, on one shipment, not on slides.
- Raise a purchase order in euros and show the dirham cost it produces, and the rate and rate date it used.
- Book the supplier invoice at a different rate, and show what happened to the item's cost and to the margin on any sale already made from it.
- Pay the supplier at a third rate, and show the realised difference attributed to that shipment, not swept to a general account.
- Run the month-end revaluation, and show unrealised separated from realised on the same report.
- Attach a forward contract to a purchase order and show the buyer's screen using the contracted rate rather than spot.
- Produce gross margin per item for a past quarter, including the FX difference, and reconcile its total to the general ledger.
- Show FX gain and loss analysed by supplier and by product line for a period that has already closed.
Item 3 decides it. Every system revalues balances at month end — that is the part accounting standards force and every package therefore has. Carrying a realised difference back to the shipment that caused it is a different mechanism, and it is the one that turns an unreadable account into an attributable cost.
The short version
The peg makes dollar buying genuinely free of currency risk, and it teaches a trading business a habit that is wrong for every other currency it will ever buy in.
There are four dates on every shipment where a rate can be fixed. The margin you reported used the second. The third settled after the goods were sold, and the fourth was never attributable to anything smaller than the month.
None of that is an error. It is a position the company held without deciding to, and reported a profit that quietly included it.
Putting the difference on the shipment rather than in a single account at the foot of the P&L is the whole of the fix, and it is the same fix that landed cost needs. How it runs with the warehouse, the customs file and the bank facility on one record is Logix; the financing side of the same shipment is letters of credit and trade finance; and the general case for why a correct ledger still produces wrong management numbers is why your ERP dashboard is lying to you.