A trade that is going to go wrong almost always warns you before it ships, with four signs: a margin well below what you usually make, freight well above what you pay to that port, a loading date coming up without a single document, and a shipment milestone passing its date with nobody recording anything. The first two can be seen as soon as you enter the deal with its costs, even if loading is weeks away.
None of them says the trade is lost. They say there is a question to ask today, while you can still renegotiate the freight, change vessel or send the missing document, and not at settlement, when all that is left is explaining the loss. Below are the four with their thresholds, an example with dates and how to track them in a spreadsheet.
The four signs at a glance
| Sign | When it can fire | What it is compared with | When it warns |
|---|---|---|---|
| 1. Margin | Once the deal has prices and costs, also in negotiation | The average of your other deals with costs | Negative margin, or more than 40% below the average |
| 2. Freight | Once the deal has freight and quantity, also in negotiation | The per-ton average of your other deals to the same destination port | More than 40% above the route average |
| 3. Documents | Five days before the loading date, once the deal is confirmed | Nothing: it checks whether any document is uploaded | None with five days or less to loading |
| 4. Milestones | Once the deal is confirmed and its containers are added, from the booking (expected fourteen days before loading) | The expected date of each milestone | More than three days past with no actual date |
The first two are not compared with the market but with your own trades, which is what really tells you whether a deal is unusual for you. That is why they need a minimum: the average is worked out from at least three other deals, not counting the one being checked. With fewer, an average means nothing and the sign stays quiet, except for a negative margin, which always warns.
An example: the deal that warned three weeks before loading
A house exporting sesame to Rotterdam is negotiating 100 tons in five 20-foot containers: it buys at USD 1,650/t and sells at USD 1,860/t. Its forwarder quotes freight at USD 2,000 per container and the other costs add up to USD 2,400. Its other four deals with costs closed at margins of 9%, 10%, 11% and 12%, and three of them went to Rotterdam with freight of USD 62, 70 and 68/t. Loading is on Monday 9 November.
| Date | What happens | Sign |
|---|---|---|
| Friday 16 October | The deal is entered in negotiation, with prices, freight and costs | — |
| Saturday 17 October | 4.6% margin against a 10.5% average; freight of USD 100/t, 50% above the route average (USD 66.67/t) | 1 and 2 |
| Thursday 22 October | The contract is signed, the deal moves to confirmed and its five containers are added | — |
| Monday 26 October | Expected booking date; nobody records the actual one | — |
| Friday 30 October | The booking is four days overdue with no actual date | 4 |
| Wednesday 4 November | Five days to loading and the deal has no documents at all | 3 |
| Monday 9 November | Loading date | — |
The numbers: USD 186,000 of sales minus 165,000 of purchase, 10,000 of freight and 2,400 of other costs leave USD 8,600, a 4.6% margin. The threshold sits at 60% of the average, 10.5 × 0.6 = 6.3%, and the deal falls below it. The freight, USD 10,000 over 100 tons, is USD 100/t against 66.67: 50% more, above the 40% that triggers the warning. Both signs arrive twenty-three days before loading and before signing, when asking for another quote still costs just a phone call.
Sign 1: the margin drifts from what you usually make
A low margin is not always a mistake: sometimes it is a strategic trade or a new customer you give something up for. The sign asks for that decision to be a conscious one. It has to be compared with the GP of deals with costs, because a deal with no cost lines has an inflated margin that pushes the average up and makes every other deal look low. And as soon as you reconcile an invoice, the actual amount has to be used, not the budgeted one: that way freight that comes in dearer than planned also moves the sign.
margin < 0 or margin < average of your other deals with costs × 0.6Sign 2: freight drifts from what you pay to that port
Freight is compared per ton and per destination port, because freight to Rotterdam and freight to Callao are not the same number. Using the same logic as the route average that lets you audit your forwarder, freight more than 40% above what you usually pay to that port deserves a second quote before you close.
If it fires, that does not prove you are being overcharged: the market moves too. According to UNCTAD’s Review of Maritime Transport 2024, the container freight index out of Shanghai averaged 2,130 points in January 2024, more than double its December 2023 level, largely because of the Red Sea diversions. If it is the market, what changes is the selling price or when you close; if it is not, the forwarder.
Sign 3: loading is near and there are no documents
Five days before loading, a confirmed deal should already have something: the contract, the pro forma invoice, the certificate of origin, the draft B/L. If there is nothing, either the documents live in one person’s inbox or they have not been requested. The second is paid for in cash: if you are paid against documents, you are paid when they are right. The US International Trade Administration’s guide to the letter of credit points out that those documents are detailed and prone to errors and discrepancies, and that any error has to be amended and resubmitted.
Sign 4: a shipment milestone passes its date
Each container goes through six milestones, from booking to release, with an expected date worked out from loading and the ETA. If one runs more than three days past its expected date with no actual date, something has stalled: an unconfirmed booking, a rolled container (left behind for the next sailing), a B/L not issued. At destination, that delay is what ends up as the demurrage invoice nobody budgets for.
How to track it in your spreadsheet
You can do it without any system if your sheet has one row per deal with its costs and destination port. These are the steps:
- A margin column worked out with actual costs where you already have them, not only the budgeted ones.
- An average margin column for the other deals with costs, not counting the row it sits in, formatted red when the margin falls below 60% of that average or below zero.
- A freight-per-ton column and its average per destination port, also excluding the row itself, red above 140% of the average. Always write the port the same way: “Rotterdam” and “ROTTERDAM, NL” are two routes to a formula.
- A documents alert column with the loading date minus five days, red if that date arrives and the documents column is empty.
- The milestone columns from the shipment checklist, with the delay column red from the fourth day.
- A daily review of the red rows, with a note of what was asked and to whom.
It works as long as someone opens it every day. It fails where every spreadsheet does: averages break when rows are inserted, dates stop being recorded and nobody looks at the red on a deal that is not theirs.
What Granel does with this, and what it does not
On the Growth and Enterprise plans, Granel checks open deals every morning with these four rules and the same thresholds, and warns the organization’s owners and admins with a link to the deal. It uses the actual amount of each cost line as soon as it is reconciled, converts freight to the base currency before comparing, leaves sample deals out of the comparisons and does not repeat the same warning on a deal within three days.
What it does not do is know why a sign fires: it does not check freight indices or know whether the market has moved, the documents check does not look at what is inside them —only whether there are any— and it does not know about a delay until the expected date passes without an actual one. Nor does it decide anything: a person still asks the question.
Frequently asked questions
How do you detect the risk in a trade before it ships?
By comparing each trade with your own: its margin with your average, its freight with what you pay to that port, its documents with the loading date and each milestone with its expected date. The first two comparisons can be made as soon as the deal has costs, weeks before loading.
Why compare with my own deals and not with the market?
Because a margin or a freight rate is only unusual relative to what you normally do on that route and with that product. The market is useful afterwards, to tell whether the gap is explained by a general rise in freight or by a problem with that trade.
How many trades do you need for the average to work?
At least three others, not counting the one you are checking, with costs recorded; for freight, also to the same destination port. With fewer, the average depends too much on a single deal.
What do I do when a sign fires?
Ask the question that same day of whoever can answer it —the trader, the forwarder, the supplier— and note the answer. If the gap is justified, it stays on record; if not, there is still time to renegotiate, ask for another quote or send the missing document.