Choose Your Own Quarter

Howdy Folks,

I told you the AI issue was next. It is not. ArcBest reported this morning, the last of the seven big public LTL carriers to put out its third quarter, and earnings go stale faster than a fuel table. The AI issue runs next week.

This one works a little differently.

I am not going to tell you what this quarter means.

Partly because the same seven earnings reports make a strong case in both directions. Mostly because you are going to trust your own read of these numbers more than you trust mine, and you should. (Also, the prediction scoreboard I started in I Was Half Right could use a quarter off.)

So here is the deal. You get everything I would want on my desk before making the call: the scoreboard, where the revenue came from, who got more freight and who just got heavier freight, what happened to margins, what the economy underneath is doing, and what every management team said about the fourth quarter. Then you get the bull case and the bear case, each argued as hard as I can argue it.

You pick.

If you grew up on Choose Your Own Adventure books, you know how this works. If you already think diesel did all the work, turn to section 8. If you think the freight is finally coming back, turn to section 7. If you want the numbers first like a responsible adult, keep reading.

The seven are Old Dominion, XPO, Saia, FedEx Freight, ArcBest's asset-based ABF Freight, TFI International's LTL segment, and Knight-Swift's LTL segment, which now runs under the AAA Cooper name. Forward Air, which sat in the last version of this table, reports after the close on Monday, November 9. It gets its seat back in the Q4 issue.

Every number comes from the carriers' own releases, 10-Qs and investor decks unless I say otherwise. Anything said on an earnings call comes from a published transcript, and I flag it. When I did the arithmetic myself, I tell you, because that is where the mistakes live.

1. The scoreboard

Revenue and operating ratio, third quarter 2026 against third quarter 2025.

Four footnotes, and this quarter they matter:

• "Adj." is each company's own adjusted operating ratio, non-GAAP (non-IFRS for TFI). Old Dominion only reports GAAP. FedEx Freight reports an adjusted operating margin, 14.7% against 14.2% a year earlier, so its line is 100 minus that margin. That subtraction is mine.
• TFI and Knight-Swift calculate their adjusted operating ratios on revenue excluding fuel surcharge, and net the surcharge out of expenses too. The other five leave it in revenue. So their ratios are not like-for-like with the rest of the table, which is worth remembering in a quarter when fuel surcharge revenue rose by roughly half or more at every carrier that discloses it.
• The unadjusted numbers tell a different story at five of them. Saia's GAAP operating ratio got 170 bps worse, because Q3 2025 included a $16.4 million gain on a terminal sale, partly offset by a $1.9 million real estate impairment. ArcBest's got 20 bps worse against a $15.7 million gain on two service center sales a year ago. FedEx Freight's got 240 bps worse as spin-off and separation costs rose to $105 million from $30 million a year earlier. Knight-Swift's got 610 bps better, mostly because Q3 2025 carried a $28.8 million write-down of the brand names it retired when it folded its LTL carriers into AAA Cooper. And TFI's unadjusted LTL ratio got about 470 bps worse, nearly double its adjusted decline, on a $14.4 million office building impairment.
• FedEx Freight is a brand-new public company. FedEx spun it off on June 1, and it moved to a calendar year, so its October report covered a four-month transition period, its first as an independent company. The line above is the three months ended September 30, from its earnings release and 10-Q.

Now the fun way to read the same table.

Operating ratio is golf. Lowest score wins, and the number you want next to your name is a negative one.

Old Dominion is still the low round in the group, and it shot 71.5 on 2% fewer shipments a day than a year ago. XPO took 2.6 points off its adjusted OR. Saia shot even par, which in golf is a perfectly respectable round. TFI went 2.5 points the wrong way, and section 4 gets into why.

2. Where the revenue came from

Start with diesel, because it touched every line in that table.

The EIA's U.S. on-highway diesel price averaged about $5.57 a gallon in the third quarter, against $3.76 a year earlier. That is up 48%, my average of its monthly figures. It peaked at $6.529 the week of September 21, and it was still $6.292 on November 2, $2.54 above the same week last year.

The carriers' fuel tables followed it up. As of November 4, Old Dominion's website shows a 54.32% LTL fuel surcharge in effect, and quotes 53.32% for November 11 to 17 against a 28.32% average a year ago. XPO's table shows 57.75% for the week, against 31.25% in the same week of 2025.

The U.S. carriers on this list bill fuel surcharges keyed to that weekly index. On Old Dominion's and XPO's tables it works out to roughly a tenth of a point per penny, as I walked through in Your Carrier Has Two Rate Cards. So every revenue line in section 1 got a lift that had nothing to do with how much freight moved.

How much of a lift? Here is each carrier's revenue growth, split into fuel surcharge and everything else.

XPO, TFI and Knight-Swift report revenue with and without fuel surcharge, so their inputs come straight off the page, and the rest of each row is subtraction and division. The "about" rows are my arithmetic. Saia's uses its disclosed fuel share of revenue, 22.5% against 15.2%. Old Dominion's uses shipments times the gap between its revenue per shipment with and without fuel. FedEx Freight's uses its fuel surcharge per shipment times shipments. ArcBest says higher fuel surcharge revenue drove most of its yield increase but does not give the dollars.

Strip the fuel out and the ranking holds, but the gaps change. Knight-Swift, FedEx Freight and TFI go from 7% to 10% growth to roughly zero or less.

Add up the six that disclose enough to check, and they billed about $636 million more in fuel surcharges than a year earlier, against about $190 million more of everything else. That is my arithmetic on their numbers.

Before anybody calls that free money: fuel is also a cost. Saia said on its call, per the transcript, that its fuel expense rose 55%. But FedEx Freight's 10-Q says that so far in 2026 the higher surcharges more than offset the higher fuel expense, and XPO's release counts higher fuel surcharge revenue, partly offset by higher fuel costs, among the reasons its LTL adjusted EBITDA rose. So at FedEx Freight and XPO, by their own filings, the surcharge added to margins, not just revenue. Whether you count that as earned or borrowed is most of the argument in sections 7 and 8.

3. Who got more freight, and who just got heavier freight

TFI reports quarterly totals rather than daily figures, over the same 64 business days in both years, so the percentages work the same way. ArcBest gives the direction of its ex-fuel yield, not the number.

There are three ways to get more revenue out of freight: more shipments, heavier shipments, or more money per pound. This quarter split the group almost perfectly.

• More shipments. XPO moved 5.8% more shipments a day and TFI 4.6% more. Nobody else grew shipment count by more than 1%.
• Heavier shipments. ArcBest's average shipment got 13% heavier, which is how it grew tonnage 9.6% on 3% fewer shipments. Saia's got 8.1% heavier and Knight-Swift's 4.8%.
• More money per pound. Old Dominion's shipments got 1.9% heavier, but it handled 2% fewer of them, so tons a day slipped 0.1%. Its ex-fuel revenue per hundredweight rose 5.1%, the most in the group.

So where is the heavy freight coming from? The carriers mostly point to truckload.

Per the call transcripts, Old Dominion's CEO said some of its heavier freight is coming from a truckload industry that is short on capacity. XPO's chief strategy officer said more truckload freight converted to LTL as truckload rates climbed sharply, that it runs heavier and more regional, and that XPO repriced a lot of it. XPO's CEO said the shift is mostly about truckload rates, with truckload tender rejections a secondary reason. ArcBest's release credits a continued shift in its freight profile, and its CEO told analysts the company is seeing some truckload spillover. And Knight-Swift's CEO, whose company runs one of the biggest truckload fleets in the country, said in its release that LTL pricing conditions are growing more supportive, partly because truckload is tightening.

Here is what the government and the rate indexes report on truckload capacity and rates:

• Non-domiciled CDLs. FMCSA's final rule narrowing who can hold one took effect March 16, 2026. In the rule's own analysis, FMCSA estimates about 194,000 current holders leave the freight market over roughly five years as their licenses come up for renewal. A court challenge was argued September 15, and I could find no decision as of November 4.
• English proficiency. FMCSA said on August 7 that enforcement had placed more than 26,000 drivers out of service.
• Rates. DAT's national spot van rate, fuel included, averaged $3.01 a mile in September, 96 cents above a year earlier, and contract van averaged $3.24, up 82 cents. DAT puts the fuel part at a record 84 cents a mile. Cass's truckload linehaul index, which leaves out fuel and accessorials, was up 10.1% on a year ago.

Heavy freight from truckload does two things at once.

It fills trailers. It also dilutes yield per pound.

That is how ArcBest can report tonnage up 9.6% and ex-fuel revenue per hundredweight down in the same release.

4. What happened to margins

Carrier by carrier, the short version:

• Old Dominion got to 71.5% on yield and cost, not volume. Salaries, wages and benefits fell as a share of revenue, and average full-time headcount was down 4.1% to 20,202. It still ran 99% on time with a 0.1% cargo claims ratio.
• XPO grew adjusted operating income 35%. Its investor deck shows it now sends 5.6% of its linehaul miles to outside carriers, down from 25.2% in 2020.
• FedEx Freight improved its adjusted margin by half a point while GAAP operating income fell 13% under the separation costs. Its claims ratio was 0.43% over the four months, which it calls a record low.
• Saia held 87.6% adjusted while hauling much heavier freight. Management said on the call that the terminals it opened in 2023 and 2024 grew operating income 22%. It now has 219 terminals, after opening one on Long Island in October.
• ArcBest took 210 bps off its non-GAAP OR, with help from the July restructuring, which is expected to cut about 2% of positions and close ten ABF service centers. ArcBest says the plan saved about $6 million across the whole company in the quarter.
• TFI went the other way. Ex-fuel revenue per shipment fell 4.4%, and LTL operating income fell 43%, including a $14.4 million impairment on an office building. Per the transcript, CEO Alain Bédard called U.S. LTL the one disappointment in an otherwise strong quarter, put the U.S. business at roughly a 97% operating ratio against below 75% in Canada, and said TFI will bring in executives from its peers. TFI also disclosed that its Ground Freight Pricing contract, which Bédard described as its UPS reseller arrangement, ends January 31, 2027. It brought in $30.4 million of revenue and $4.2 million of operating income in Q3.
• Knight-Swift's adjusted OR got 80 bps worse, mostly because gains on sales fell $4.6 million. Ex-fuel revenue per shipment rose 4.0%, and renewals kept coming in at mid-single-digit increases.

Then Wall Street graded the homework.

Those are closing-price changes from Yahoo Finance, Nasdaq.com and StockAnalysis daily data. ArcBest was still trading when I wrote this.

The carrier with the lowest operating ratio in the group fell 7%. Two carriers whose revenue growth was essentially all fuel surcharge went in opposite directions: FedEx Freight jumped 15% after raising its revenue and adjusted profit outlook, and TFI fell 17%. If you can build one model that explains all three, send it to me, then quit your job and go start a hedge fund.

5. The economy underneath

The carriers' numbers sit on top of these. Each is the latest release as of November 4.

Read it twice. Manufacturing has been expanding for ten months, and production is at its strongest reading in nearly five years. Truck tonnage is flat on a year ago. Cass shipments have grown year over year two months running, after what Cass calls the longest downturn on record, three and a half years. Freight spend is up 20%, and most of that is price and fuel, not boxes: Cass's two indexes imply spend per shipment up about 18%, my arithmetic.

ATA's chief economist, Bob Costello, said in ATA's October 20 release that tonnage is still soft, and that weakness in July and August left the third quarter down both sequentially and year over year. He also said capacity is still contracting, which is making for a better market for motor carriers.

On price, the BLS producer price index for long-distance LTL, which includes fuel surcharges, averaged 14.9% above a year earlier in the third quarter, and the truckload index rose almost exactly as much. Both are my averages of the monthly values, and the summer months are still preliminary.

Some general rate increases are arriving earlier. Old Dominion's 4.9% took effect October 5, four weeks ahead of last year's November 3. AAA Cooper's website lists a 5.9% increase on its base rates effective the same day. FedEx Freight's rate-change page lists a 6.9% average increase effective January 8, 2027, against 5.9% on January 5, 2026.

6. What they said about Q4

Two patterns worth noticing. First, XPO, Saia and ArcBest guided their Q4 margins to hold up better than their normal seasonal slide. Old Dominion guided to the low end of its usual 200 to 250 bps seasonal increase, and TFI guided its LTL margin flat to worse than a year ago. Second, October tonnage was up at all four carriers that gave a number, by about 15% at ArcBest and about 10% at Saia, mostly on weight.

Old Dominion's management also said October tonnage ran below normal seasonality compared with September, and that it expects November and December to come in above it. So even the carrier with the lowest operating ratio in the group is giving you both stories in one call.

7. The bull case

Here is the strongest version I can build of the case that LTL is turning up.

• The industrial economy is growing. ISM's PMI has been above 50 for ten straight months, its production index is the highest since 2021, and industrial production is up 2.2% on a year ago. LTL is an industrial business.
• Truckload capacity is leaving, and the rules say it keeps leaving. By FMCSA's own estimate, about 194,000 non-domiciled CDL holders exit over five years, and more than 26,000 drivers have been put out of service on English proficiency. The freight that cannot find a truck is already showing up in LTL: five of these seven carriers grew tonnage.
• Price is holding where you can measure it cleanly. Old Dominion's ex-fuel yield rose 5.1% and XPO's 3.6%, and XPO said on its call that October ran about 5%. Saia said on its call that contract renewals averaged a 10.8% increase, ArcBest reported 5.1%, and ArcBest's release says "LTL industry pricing remains rational." Old Dominion pulled its 4.9% general rate increase forward by a month.
• The networks have room. Old Dominion said on its call that it has more than 35% spare real estate capacity, and XPO's 10-Q says it has more than 30% excess door capacity. New freight lands in doors that are already built and paid for. That is what operating leverage looks like.
• The margins are already moving. Four of seven improved their operating ratio year over year, adjusted where the company reports one, and three guided Q4 to beat their normal seasonal slide.
• The trend is pointing up. FedEx Freight's daily shipments went from down 7.6% in January to up 0.9% in September. Old Dominion's tons per day turned positive year over year in September. October tonnage is up at every carrier that gave a number.
• Weak carriers are leaving. Mountain Valley Express, a 13-terminal regional carrier, shut down in July, and TP Freight Lines, which listed seven terminals, suspended operations in August, per FreightWaves. Every exit is freight the survivors get to price.
• Claims are at record lows. Saia and FedEx Freight both report record-low claims ratios, 0.3% and 0.43%. XPO is below 0.2% and Old Dominion is at 0.1%. Saia said on its call that its service and claims record supported its 10.8% renewal rate.

8. The bear case

And here is the strongest version I can build of the case that this was a fuel quarter dressed up as a recovery.

• Fuel did most of the growth. At the six carriers where you can check, fuel surcharge was 59% to more than 100% of revenue growth. Ex-fuel revenue grew less than 1%, or shrank, at FedEx Freight, Knight-Swift and TFI.
• Fuel reverses. The EIA's October Short-Term Energy Outlook has retail diesel averaging $5.81 this quarter, then falling every quarter of 2027, to $4.18 by the fourth. On November 2, NYMEX diesel futures for December 2027 settled about 94 cents under December 2026, per Barchart. XPO's strategy chief said on its call that he expects diesel to come down meaningfully through the rest of the quarter, though he guided XPO's Q4 fuel surcharge revenue to roughly flat with Q3. Whenever the index does fall, the surcharge that padded Q3 revenue and margin shrinks with it.
• Price per pound is weak where the freight is heavy. Ex-fuel revenue per hundredweight fell at Saia, down 2.2%, at TFI, down 0.9%, at Knight-Swift, down 0.7%, and at ArcBest, down in the low single digits. It rose just 0.4% at FedEx Freight.
• The growth is weight, not shipments. Shipments per day fell at four of the seven: Old Dominion down 2.0%, ArcBest down 3.0%, Knight-Swift down 1.4% and FedEx Freight down 0.7%. Tonnage that showed up because truckload got tight can leave when truckload loosens.
• The freight economy is flat. ATA's truck tonnage index was unchanged on a year ago in September. FedEx Freight's own 10-Q expects softness in the industrial economy to keep pressuring LTL demand for the rest of the year. In April, SJ Consulting sized the 2025 U.S. LTL market at $51.8 billion, down 1.9% from 2024.
• The seasonal slide is real. Old Dominion said on its call that October tonnage came in below normal seasonality. ArcBest guided its Q4 operating ratio as much as 150 bps worse than Q3. TFI guided its LTL margins flat to worse than a year ago.
• The market sold the leader. Old Dominion fell 7.3% the day it posted the best operating ratio in the group. TFI lost 17.1% in a session.
• There is a new competitor in the building. Amazon, which touts 80,000 trailers and 24,000 intermodal containers, opened its LTL network to all businesses in June, per FreightWaves.

9. What would settle it

Neither side has to wait long. Here is what each one needs to see, and when it shows up.

1. Ex-fuel yield in the Q4 prints, which start in late January. Bulls need it to accelerate. Bears expect it to stall while heavier truckload freight keeps diluting yield per pound.
2. Shipments per day versus weight per shipment, same prints. Bulls want shipment counts growing, because that is more demand, not just heavier loads. Bears expect weight to keep doing the work.
3. Diesel. The EIA's next Short-Term Energy Outlook comes out November 10, and its weekly diesel price posts every Tuesday. On these tables, every 10 cents of diesel is about a point of surcharge.
4. The truckload rules. As of November 4, I could find no D.C. Circuit ruling on the non-domiciled CDL rule, and FMCSA's English-proficiency rule is still a proposal after comments closed October 9. A ruling or a final rule in either direction changes the capacity math.
5. January. General rate increases and contract renewals for 2027. Watch whether shippers sign the multi-year base-rate deals Old Dominion says customers are asking for.
6. ISM. The next manufacturing report is December 1. Bulls need it to stay above 50.
7. The mid-quarter updates. Old Dominion, XPO, Saia and ArcBest all published mid-quarter numbers between September 3 and September 8 this year. If they do it again in early December, you get November before you get the Q4 prints.

What's next

Next week is the AI issue I keep bumping: the full teardown of the plugin that quotes and books freight from inside our inbox, including what broke.

The Q4 version of this issue lands once the last carrier reports in early 2027, with Forward Air back in the table. Same tables, same order, so you can hold this quarter up against the next one.

See ya next week.

Gabe

Here is what I want to know: bull or bear? And which single number in this issue decided it for you? Reply and tell me. I will tally the answers, and when the Q4 numbers land, we will find out which half of this list was right.