US Intermodal Gains Ground as Diesel and Truck Capacity Push Costs Up

The Analysis

Every corridor has a pothole, and on the Los Angeles to Chicago lane the pothole sits at the Chicago end, not the California end. I have run boxes along this line often enough to know its shape: the ocean leg gets all the attention, the inland leg gets the bill. What changed this quarter is not the ocean leg. Diesel moved up, truck capacity got tighter, and the third-quarter results from the big railroads showed domestic intermodal volumes growing. Put those three together and you get a familiar migration, importers pulling containers off the long-haul inland leg and putting them on rail.

The pitch writes itself. Rail is cheaper over distance, trucks are expensive when fuel and drivers both cost more, and the trade press reported the trend and stopped there. I am not going to argue with the pitch. I am going to argue with the way it gets quoted, because the number your broker hands you and the number that lands on your cost sheet are two different animals, and the gap between them is built out of drayage, dwell and fixed cut-offs. What never reaches the headline is this: on this lane, the length of the rail haul is not the variable that decides whether intermodal pays. The distance from each door to the nearest ramp is.

So my assumptions go on the table first and stay there. The lane behind every number in this piece is a plain one: a forty-foot container, loaded, door at the Los Angeles and Long Beach port complex, destination a distribution centre in the Chicago metro area, roughly two thousand miles of inland haul. Assume cargo worth 120,000 dollars in the box. Assume a truckload dry van quoted at 2 dollars 15 per mile, so about 4,300 dollars door to door. Assume an intermodal quote of 2,750 dollars ramp to ramp plus drayage at both ends. Everything that follows is arithmetic on those assumptions, and all of them move. That is not a flaw in the analysis. It is the analysis.

Start with what the update actually says, because the hard news is thin and I am going to build on it. Southern Star put three things in one paragraph: diesel prices up, truck capacity tighter, and third-quarter results from the big railroads showing domestic intermodal volumes growing. No diesel number, no capacity index, no volume percentage. So before I put anything on top of it, my assumptions go on the table and stay there for the rest of this piece. Assume diesel moved from about 3 dollars 80 a gallon to about 4 dollars 35 over the period in question.

Assume the truckload rate on the Los Angeles to Chicago lane sits at 2 dollars 15 per mile. Assume an intermodal quote of 2,750 dollars ramp to ramp for the same box, plus drayage at both ends. Assume 2,000 miles of inland haul and cargo worth 120,000 dollars in the container. Every figure that follows is arithmetic on those assumptions, and every one of them can move. That is not a weakness in the analysis. It is the analysis.

Why this quarter and not last spring. Diesel is the only input in trucking that carries a published index and a contractual pass-through, so it moves fast and it moves visibly. A fuel surcharge table keyed to a weekly or monthly diesel index absorbs a move like this inside two to six weeks, and the shipper sees it as a line item on the invoice. Capacity has no index. When capacity tightens, what you feel first is not a price, it is a lead time: your broker stops saying yes and starts saying let me check and come back to you. Assume cover time on a spot truckload on this lane goes from one day to three. Three days of cover on a lane that takes four days to drive is a lane that no longer behaves like a truck lane. That is the moment importers start pricing rail, and it is where the intermodal volume growth in the third-quarter numbers comes from.

This lane has six nodes and I want you able to name them, because the money is never lost on the long middle part. One: the container yard at the port complex and the chassis pool underneath it. Two: the Los Angeles end dray to the ramp, and there are several ramps to choose from. Three: the ramp itself, where the box gets lifted onto a well car. Four: the line haul. Five: the Chicago end ramp, where it comes off and sits. Six: the Chicago end dray plus the delivery appointment at the distribution centre.

Ask which node blocks and the answer is five and six, not four. Chicago is the funnel of the American rail network, east-west and north-south traffic both change hands there, and when the funnel backs up, dwell at the Chicago ramps is where it shows first. It shows at the appointment desk too, because the same congestion holding your box in the ramp is holding the dray driver who was going to pull it out, and the receiving dock has only so many doors and so many hours.

What does it cost to switch. Switching mode is the obvious one and the cheapest to price: you take the 650 dollar spread and buy four days with it. Switching ramp costs almost nothing in money and a lot in attention, since moving from an inner Chicago ramp to an outer one adds maybe 30 to 40 miles of dray and buys you out of the worst queue. Switching port is the expensive one and the one people underprice. Route through Houston or Savannah instead and move the box north from there: assume the ocean leg costs another 600 dollars and adds eight to twelve days, while the inland leg saves around 900.

Net 300 saved, twelve days added. That trade works for a replenishment programme carrying a 30 day buffer and kills a promotion with a fixed shelf date. Switching container is the one nobody prices. Transloading into a domestic 53 foot box costs around 500 dollars and a day, but a 53 gives you roughly 20 per cent more cube, so a shipper whose 40 foot boxes cube out before they weigh out can sometimes merge two into one and delete an entire inland move. I have watched that pay for a quarter of inland budget, and I have watched it fall apart because nobody checked whether the destination dock could receive a 53.

The impact is not evenly spread and the split is not by company size alone. A shipper moving 100 boxes a month can sign a rail contract with allocated slots and negotiated drayage: 650 dollars a box across 1,200 boxes a year is 780,000 dollars. A shipper moving 10 boxes a month buys intermodal on the spot market with no allocation, pays roughly 15 per cent over the contract rate so the spread drops to about 240 a box, and in peak season gets rolled off the train while the contract holders load. Same mode, opposite outcome. Weight splits it too. A box grossing above roughly 24 tonnes either takes a heavy-weight surcharge or does not get a slot, because the railroad sells space on a car and wants to know what the car weighs.

Light bulky freight is the reverse, it cubes out long before it weighs out, and rail prices per slot, so rail loves it. Temperature control runs on its own rules: reefer slots are scarce, they need power at the ramp and someone watching them, and the surcharges are quoted separately. Distance from the ramp splits it hardest of all, which is exactly where the headline misses the story.

None of this counts until it lands on a cost or a stockout, so here is the timing. Diesel into freight rates: two to six weeks through the fuel surcharge, and watch the ratchet, because it goes up on the index schedule and comes down when somebody renegotiates, so the up leg is always faster than the down leg. Truck capacity into spot rates: four to eight weeks before peak, topping out about two weeks before the volume does. Rail cut-offs into inventory is the one that hurts, because a missed cut is not a late arrival, it is a missing week. Assume two fixed cut-offs a week in peak on this lane.

Miss one and you wait three to four days for the next, then add the dwell degradation that arrives with the same congestion, and the realistic worst case is six days. Against a distribution centre holding 10 days of safety stock, six days is not a delay, it is a stockout. The early signals cost nothing to watch: how many days your broker needs to cover a load, terminal dwell as the railroads disclose it in their quarterly reporting, and whether the dray appointment at the Chicago end is next day or third day.

The third-quarter numbers do give you one figure worth carrying into a negotiation, if you read them properly: terminal dwell. The railroads disclose it, the regulator publishes it, and it is the only number in this whole story that is somebody else's audited fact rather than a quote. Assume this lane behaves at one to two days of dwell in a normal month and degrades past three in peak. If disclosed dwell at your destination ramp is already above two days in October, then every assumption above is optimistic and your break-even drayage distance shrinks: add roughly 210 dollars of storage and chassis to the rail side and set the two modes equal again, and the break-even D falls from 484 miles to about 410. That is a real figure to take into a rate conversation, and it is not the sort of thing a broker volunteers.

The chassis deserves its own paragraph, because on this lane it behaves less like a transport asset and more like a queue. In most intermodal moves here the chassis is not the railroad's property. It is pooled, it goes scarce in exactly the weeks when volume is highest, and it bills by the day whether the box is moving or sitting. Assume 60 dollars a day and assume you cannot get one for two days in the second week of October. Two days is 120 dollars, and it lands on the same invoice as the storage. It matters far more than its size suggests, because a missing chassis breaks the dray appointment, a broken appointment keeps the box occupying a slot at the ramp, and the occupied slot is where the storage charge comes from. One small missing asset turns into three charges on three lines. That is what a pothole looks like out here: not a crater, a chain.

Here is what the headline does not carry. The intermodal quote you are handed is ramp to ramp. It does not include the delivery dray, it does not include the chassis, it does not include storage once free time runs out, it does not include a heavy-weight surcharge, and if it settles collect there is a fee for that too. Assume those add 250 to 350 dollars at the far end, so a 2,750 dollar quote lands closer to 3,050 and the spread you thought was 650 is closer to 350.

Now add the variable nobody models, which is combined drayage distance. Written out: truck is 2.15 a mile for the whole 2,000, so 4,300. Rail is 1.35 a mile on the rail portion plus about 220 dollars of terminal handling plus 4.20 a mile on the drayage, and if D is total drayage miles across both ends that comes to 2,920 plus 2.85 D. Set them equal and D lands near 484 miles. Past roughly 480 miles of combined drayage, call it 240 at each end, the rail line-haul advantage is gone and anything you still show as a saving is coming from somewhere other than mode. The length of the rail haul does not decide this. The distance from each door to the nearest ramp does. I have never seen a broker quote sheet carry that number, and it is the number that decides the answer.

Fixed cut-offs are the second thing nobody prices. A truck leaves when you say go. A train leaves when the timetable says it leaves. In peak, assume two departures a week and a ramp that fills: your box rolls to the next train and that is three or four hard days, not a soft delay. Free time is next. Assume 48 hours free at the Chicago ramp, then 150 dollars a day of storage and 60 a day on the chassis, so if dwell stretches from one day to three in peak that is 210 dollars a box you did not budget, on top of the 250 to 350 at the end.

Add the four days of transit and the carrying cost on 120,000 dollars of cargo at 10 per cent a year, which is 33 dollars a day and 132 for the trip, and the honest net is about 218 dollars a box. Across 480 boxes a year that is 104,640 dollars. Put the peak risk on the same page. Assume four peak months carrying 200 boxes on rail, a missed-cut rate of 8 per cent, so 16 boxes, and assume half of those turn into a real stockout at 5,800 dollars each between expedited trucking and customer penalties.

That is 46,400 dollars against you and the year ends at 58,240 dollars net, or 121 dollars a box. Not 650. Get the missed-cut rate down to 3 per cent by locking slots 48 hours ahead and keeping a truck fallback and the year ends near 87,240 dollars, or 182 a box, and the 61 dollar gap is the measured value of one operational habit.

The conditions that flip it, then. Diesel back to 3 dollars 50 and the fuel surcharge gives back around 12 cents a mile, so the truck quote lands near 4,060 and the rail spread drops to 240, which after end charges and carrying cost is close to nothing, and at that point you stay on trucks for reasons that have nothing to do with the rate. Dwell at the Chicago end out to five days and door to door becomes twelve, which breaks any receiver running 10 days of safety stock no matter how good the number looks. Cargo worth 400,000 dollars a box carries 110 dollars a day, so four extra days costs 440 and wipes a 350 dollar saving: the higher the value, the worse the rail case, which is the opposite of what most people assume.

Capacity tightening further pushes it the other way. If spot truckload on this lane reaches 3 dollars a mile in peak, the truck quote is 6,000 and the rail spread is over 2,000, and at that point the constraint is not money but allocation, so whoever locked slots in September wins December. Pin these four numbers to the wall, 480 miles of combined drayage, 3 days of dwell, 10 days of safety stock, 400,000 dollars of cargo value, and check them before you sign, because crossing any one reverses the answer.

So does this corridor still work. It does, but you assign it lane by lane rather than converting everything. My split runs three ways. Lanes with more than 25 days of safety stock and under 60 miles of drayage go to rail whole, no hesitation. Lanes with 10 to 25 days run rail as base load and truck for the peak increment, because that is exactly what the truck premium is for. Lanes under 10 days of safety stock, or more than 150 miles from the nearest ramp, stay on trucks, and what you are buying there is not freight, it is the right to leave whenever you want.

Price everything door to door before you compare it to anything, and keep a truck option open on the lanes where your buffer is thin. The pitfalls are dull and that is where the money leaks: quoting your boss the ramp-to-ramp number and then explaining the end charges; signing a rail contract without an allocation and discovering in October that a good rate is not a seat on the train; rerouting through another port because the inland leg is shorter and forgetting the ocean leg is longer, thinner and harder to book in peak; and letting a box sit outside the fence because the receiver never booked an unloading appointment, which is your bill and not theirs. Nadia would put the same point on a clock instead of a map: the four extra days on the train are survivable, the eight hours lost at the Chicago appointment desk are what break the promise you made to your customer.

↑ Back to the key points

— By Ömer Kaya