FedEx Express fuel surcharge holds at 38.00% for 31 Aug-6 Sep as jet fuel stays elevated

The Analysis

It is twenty to seven at the collection point and the van is already idling. The driver has nine minutes before he has to pull, and there are eleven parcels on the table that are not scanned yet. Nine minutes. That is the unit this business actually runs on, and it is the unit nobody prices.

FedEx published its Express fuel surcharge at 38.00 percent for the week of 31 August to 6 September 2026. The week before it was 37.00. Early in July it was 30.75. One percentage point does not sound like much when you say it out loud. Seven and a quarter points against July does. Both of them are sitting in the same week's invoice, and which one you actually feel depends entirely on whether you have been re-pricing every week or sleeping through the resets.

The part I care about is buried in the mechanism rather than in the number. This is a weekly index tied to US Gulf Coast kerosene jet fuel prices with a two-week lag. Two weeks is three hundred and thirty-six hours. That means the rate on this week's invoice was already decided before anybody on your team saw it, and it also means you could have seen it coming three hundred and thirty-six hours earlier if somebody had been looking. Most operations do not look. They find out on the invoice, three weeks after the fuel moved.

Take one parcel. Three kilos, base express freight of USD 42.00, and I want to be clear that both of those are my assumptions on a lane I invented for the arithmetic, not a rate I am quoting you. At 30.75 percent the fuel line on that parcel was twelve dollars and ninety-two cents. At 37.00 percent it was fifteen fifty-four. At 38.00 percent it is fifteen ninety-six. Against July that is three dollars and five cents more per parcel. Against last week it is forty-two cents.

Now scale it to a week. Assume eight hundred parcels a week on that lane. Against July, three point zero four five times eight hundred is two thousand four hundred and thirty-six dollars a week. Against last week it is three hundred and thirty-six dollars.

Divide by one hundred and sixty-eight, because a week is how the invoice is cut and hours are how the floor runs. Two thousand four hundred and thirty-six over one hundred and sixty-eight is fourteen dollars and fifty cents an hour of pure surcharge against the July level, and the week-on-week move is two dollars an hour. Fourteen fifty an hour does not sound like a crisis either, until you notice it is running twenty-four hours a day with no overtime and no weekend off.

Absorbing it or passing it through is worth doing on paper once, because deciding by instinct is exactly how margin leaks. Assume a contribution margin of eight dollars a parcel on that lane before fuel, so six thousand four hundred dollars a week across eight hundred parcels. Absorb the two thousand four hundred and thirty-six and you are left with three thousand nine hundred and sixty-four. That is thirty-eight percent of the contribution, gone, on a move that happened while nobody was looking. If your selling price has not moved since early July, this is where that money went. I am not telling you to pass it through. I am telling you to look at the number before you decide, because a decision nobody writes down becomes a decision nobody revisits.

A percentage is only as honest as the number it is applied to, so let me be explicit about the base. Assume the surcharge is assessed on the base transportation charge, which is my assumption and the single line you must confirm on your own rate card. If it is also assessed on the accessorials stacked on top of that base, the picture changes. Assume a parcel carrying six dollars and fifty cents of accessorials. At 38 percent that is another two dollars and forty-seven cents. At 30.75 percent it was two dollars even. Forty-seven cents a parcel, and on eight hundred parcels a week that is three hundred and seventy-six dollars nobody budgeted for.

The lag is what turns this from a bill into something you can plan against. Assume the index tracks US Gulf Coast kerosene with a two-week lag. Two weeks is three hundred and thirty-six hours. So the rate published for the week of 31 August was locked when fuel moved around the middle of August. You cannot argue with it, and you cannot negotiate it. What you can do is watch the thing it is derived from and buy yourself three hundred and thirty-six hours of warning.

Concretely, give one person thirty minutes a week to read the fuel price and write down which way it is going. Thirty minutes a week is twenty-six hours a year. Against a surcharge that has moved seven and a quarter points in about eight weeks, twenty-six hours a year is the cheapest insurance on the floor, and it is cheaper than finding the move on an invoice.

Then build a trigger, because a number without a threshold is just trivia. Pick the rate at which you act. Mine would be forty percent. Below that, absorb it and keep watching. At or above it, re-quote inside forty-eight hours and tell the customer team the same morning. The trigger matters more than the watch, and most operations have neither.

While you are rebuilding that habit, look hard at the validity window on your quotes. Assume your agreement holds a rate for thirty days. Against an index that resets every one hundred and sixty-eight hours, thirty days is roughly four resets, and somebody is carrying every one of them. That risk does not disappear just because it is not on the page; it gets priced into the base you were quoted. If you can move the fuel line to a pass-through instead of a fixed percentage, you hand the volatility back and should get a leaner base in exchange. Model that trade before peak rather than during it, because during it you will take whatever you are handed.

And give the whole thing a weekly slot. Thirty minutes, same morning every week: read the published rate, read the fuel price behind it, update the sheet, tell the dock and the customer team what changed. A standing half hour beats a monthly post-mortem every time, and it is the only defence I know against finding a move like this on an invoice three weeks late.

Where the field can take money back on a percentage surcharge is not the rate itself, it is the base, because the surcharge is a percentage of that base. Cut the base and the fuel line comes down with it.

Assume a second lane priced at six dollars per chargeable kilo. A carton measuring thirty-five by twenty-five by twenty centimetres is seventeen thousand five hundred cubic centimetres. Assume a divisor of five thousand, and go check that rather than taking my word for it. That divides to three point five kilos volumetric against an actual weight of one point nine. You are paying on three point five.

At six dollars a kilo the base on that carton is twenty-one dollars, and the fuel line at 38 percent is seven dollars and ninety-eight cents. Right-size the box so it bills at one point nine kilos and the base is eleven forty, with a fuel line of four dollars and thirty-three cents. Three dollars and sixty-five cents falls off the fuel line alone, and nine dollars and sixty cents falls off the base underneath it.

Run four hundred cartons a day through that change. Three point six five times four hundred is one thousand four hundred and sixty dollars a day off the fuel line, and across a seven-day week that is about ten thousand two hundred and fourteen dollars. The base saving is nearly three times that again on top. One afternoon with a tape measure, and no system project, no steering committee, no vendor.

The loud number this week is the percentage. The quiet one is the retry hour, and when peak demand builds the quiet one is the thing that closes the door on you.

Assume a residential lane doing six hundred deliveries a day. Assume first-attempt success sits at ninety percent, so sixty parcels a day do not land. Assume sixty-five percent of those land on the second attempt: thirty-nine resolved, twenty-one sinking into the exception pile. Assume twenty-two minutes of driver route time per failed attempt, counting the stop, the scan, the card and the walk back to the vehicle. Sixty times twenty-two is one thousand three hundred and twenty minutes, which is twenty-two hours of driver capacity a day spent moving nothing that was not already supposed to move.

Twenty-two hours is close to three shifts. In the back half of the year you cannot buy that capacity back at any price, and every hour you burn on a retry is an hour you will not have for the volume that is about to land on you.

Then the desk. Assume the twenty-one parcels that went deeper each need seven minutes: a look-up, a message, a rebook instruction. Twenty-one times seven is one hundred and forty-seven minutes a day, call it two and a half hours. Assume a third of them wait on a customer reply before anything can happen, and assume that reply takes five hours. Seven parcels a day sitting five hours is thirty-five parcel-hours of pure waiting, every day, and it compounds straight through the weekend when nobody is watching it.

Peak does not arrive as a rate. It arrives as a van that pulls nine minutes early because the driver is behind, and eleven parcels left on the table.

Assume your pickup cutoff is eighteen forty-five. Assume the manifest closes at eighteen fifty-two on a Thursday in November. Those seven minutes are not the cost. The cost is that everything not scanned waits for the next collection, and on most lanes the next collection is twenty-four hours later. If thirty parcels miss, that is thirty times twenty-four: seven hundred and twenty parcel-hours, and all thirty are now competing for room on a van that was already tighter than it was the week before.

That is the arithmetic I want on the board, because it reframes the argument completely. The surcharge raises the cost of the parcel. The missed cutoff raises the cost of the hour, and the hour is what the customer actually counts. You cannot move 38.00 percent. You can absolutely decide whether the last scan happens at eighteen thirty or eighteen fifty-two.

There is also an hour-scale lever here that costs nothing in rate. Assume the collection that misses is the last one of the day, because it always is. Shifting a fifth of that volume into the morning collection costs you nothing on the invoice and buys back the twenty-four hours you keep losing on thirty parcels. That is a schedule change a supervisor can make on a Tuesday, not a procurement project, and it is the kind of move that never shows up in a rate negotiation and always shows up in the delivery numbers.

The decision in front of you is not whether to pay 38 percent. It is which parcels need the hours you still control and which ones can give those hours up. That is a sorting problem on a table, not a pricing problem in a spreadsheet.

Split the table into three piles. The first is anything with a promise that expires inside this week, plus anything feeding a line that stops if it does not arrive. That goes, and the surcharge is simply the cost of a promise you already sold. The second is anything with slack in the promise. That is where the money is, because every parcel you can move forty-eight hours later without breaking anything is a parcel you can consolidate, re-weigh or re-price before it goes. The third is anything going to an address you do not trust. That one does not get scanned until somebody fixes the address, because a parcel that fails twice and comes back has eaten forty-four minutes of driver time and fourteen minutes of desk time and generated two customer contacts, and you still have not delivered it.

I want to be straight about the second pile, because holding is not free. You are trading a known rate for an unknown one, and the rate for the week of 7 September has not been published as I write this. What you do have, which you do not get with most surcharges, is a direction to look at, because the index tracks a published fuel price with a two-week lag. Use the direction. Do not pretend it is certainty.

Consolidation earns more here than people expect, and for a specific reason: the surcharge is a percentage of a base that itself falls when you consolidate. Fewer pieces, lower base, and a lower fuel line sitting on the lower base. That double effect is the one structural lever you have against a percentage, and it is worth more than arguing about a point of surcharge.

Mode shift is the slower lever. Moving off express onto anything deferred changes the promise by days rather than hours, and you cannot do it inside one week for freight that has already been sold. Where it earns its keep is replenishment rather than orders. If you are paying express rates to refill a forward warehouse because the slower pipeline underneath is thin, seven and a quarter points of surcharge since early July is the argument you needed to go and fix that pipeline. That is a quarter-scale job, and this week's rate is a reason to start it rather than a reason to panic.

On capacity, I am not going to invent figures I have not seen. What I know from the field is that as peak demand builds, collection and linehaul slots tend to go to whoever committed earliest, and the lead time that worked in June stops working later in the year. If your booking lead time is still measured in hours, move it to forty-eight to seventy-two hours now, while there is still flexibility in the slots. Lead time is nearly free this week and expensive in ten weeks.

The last piece before the traps is what you say to the customer. A surcharge you swallow silently is a margin decision, and you should make it deliberately rather than by default. A promise you miss loudly is a service decision, and it costs you retry hours, contact minutes and a slice of a relationship. If attention is scarce this week, spend it on the promise and let the surcharge flow through.

Now the traps, and the first one is the base. Confirm what the 38 percent is applied to before you build any savings case. A percentage of the base transportation charge and a percentage of base plus accessorials are different numbers, and I priced the gap between them at three hundred and seventy-six dollars a week on eight hundred parcels.

The reset day is the next one. This is a weekly index, so find out which date or timestamp assigns a shipment to a week. It is printed in your agreement and enforced by your shipping system, and it is worth real money on a weekend boundary.

Stacking is the one that bites in peak. This fuel percentage sits alongside whatever peak-season, residential and remote-area charges your lane carries. They do not replace one another. A parcel that catches all of them has very little base left underneath.

The lag itself is a trap in both directions. It means the invoice always tells you about fuel from roughly three hundred and thirty-six hours ago, so an invoice that looks calm may be hiding a move that has already happened, and an invoice that looks alarming may already be stale by the time you read it. Work off the fuel price and the published weekly rate, not off last month's invoice.

Anchoring is the last trap, and it is the one I see most. The week-on-week move is one point, from 37.00 to 38.00, and one point is easy to wave through. The move from 30.75 in early July to 38.00 now is seven and a quarter points. If you anchor on last week you will do nothing. If you anchor on July you will re-price, and re-pricing is the right answer.

If you only do one thing this week, re-price. Not renegotiate, re-price. Get the 38.00 percent onto your own quote sheet with the week printed next to it, and find out what base it sits on. Everything else I have written here is worth more money over a quarter, but it takes longer to show up, and the re-price is the one you can finish before the van pulls at ten to seven.

Three numbers on the board by the manifest printer and the job mostly runs itself: chargeable kilos per carton, first-attempt success rate, and cutoff misses per week. Measured the same way every day. I have just priced those three on unremarkable lanes at ten thousand two hundred and fourteen dollars a week, twenty-two hours a day and seven hundred and twenty parcel-hours per missed collection. The surcharge belongs to somebody else. Those three are yours, and this week is a good week to start measuring them.

↑ Back to the key points

— By Nadia Pryce