A Series of Uncertain Events

In “A Series of Uncertain Events”, Harris + David break down the “Telegraph Market Consist”  for 1H 2026 + what the next three to six months may hold for freight rail. The outlook is…murky, with carload volumes projected to dip 3–5% + intermodal 5–6%. The bright side? This all potentially ripens the conditions for what looks like a shipper’s market.

They unpack the forces behind those lower volumes - from international trade headwinds + tariff turbulence to weaker domestic indicators. They also dive into lanes where intermodal is gaining ground versus truckload. For rail shippers the message is pretty simple: in a cycle like this, leverage matters + this may be just the moment to use it.

Talk Track, hosted by Harris Ligon + David Correll of Telegraph™, is a spin-off series dedicated to timely rail industry news. From service shakeups to technology breakthroughs, each episode delivers a behind-the-scenes perspective on all the happenings shaping the future of freight rail. 

Harris + David will bring their decades of rail experience to help them parse through the latest industry headlines, evolving regulations, + the long-term forecasts for how railroads move freight across North America. Find us at telegraph.io/insights, Apple Podcasts, Spotify, or wherever you listen to your favorite podcasts. 

About Our Hosts

Harris Ligon is the co-founder + CEO of Telegraph. Prior to launching Telegraph, he spent nearly 15 years in surface transportation at Uber Freight, Norfolk Southern, + BNSF Railway. During this time, he led teams in operations, strategy, business development, + product development. 

David Correll is the Director of Freight Market Intelligence at Telegraph. He has spent two decades in transportation and logistics with the US Department of Transportation, the US Department of Energy, the Massachusetts Institute of Technology, and Clark University.

About Telegraph

Telegraph is a leader in delivering digital solutions to railroads, shippers, logistics service providers, terminals, + railcar leasing companies. With an integrated platform that prov...

Transcript

SPEAKER_01 0:01

I'm Cindy Trick.

SPEAKER_00 0:02

And I'm Aaron Sligan. You're listening to Don't Dwell On, the podcast where we pause just long enough to figure out what's really moving freight rail forward. Let's get into it. Hey folks, welcome back to TalkTrack, a telegraph podcast, where we talk about everything that's going on in today's events in the rail industry, really for anybody that's involved in the overall freight rail space in North America. So that could be railroads, shippers, equipment owners, and even third-party logistics companies. And so today we're joined by our very own David Carell, who uh leads our freight market intelligence function. And uh we're gonna be diving in to the second rendition of our Telegraph Market Consest. David, what welcome back? How are you doing?

SPEAKER_01 0:50

I'm very excited to be back with a second report to share and to talk about with you and to share with this audience.

SPEAKER_00 0:57

Yeah, look, David, it's uh, you know, the the initial cut was uh was fairly broad and we were wading into some territory um that I think other folks gave us feedback on, that there wasn't quite a bit there. But in this second rendition, I think that you found a few more levers and trends that are really interesting, especially in the commodity markets. And so I'm candidly really excited to dive into that as well. And so, look, um, what we're trying to do with the market consist is give the readers some actionable insights around what we perceive to be trends and changes happening in the market over the next three, six, nine, and even twelve months. And so the question that I think you and I were asking ourselves as we were kind of rounding the corner on bringing this to town was what is gonna happen in rail in the next three to six months, considering everything else that's going on? What do you what say you?

SPEAKER_01 1:57

Gosh, yeah, well, well, I will I think first it's best if I speak to the output of the quantitative modeling. And then I'd love to get your take on you know what that means on the ground, you know, drawing on the many experiences both here at Telegraph and you've had with customers and and with carriers, and and you know, happy to have that conversation because I think there's more to unpack than we could write in the text. And so conversations like this, you know, give us an opportunity to think more deeply about it. So I think the the bottom line, and this speaks to when you were mentioning the goals of the project, we want to say what's gonna happen and not sort of hand walk sort of hand washing way or hand-waving way, excuse me, any sort of vague way. We we want to be really rather quantitative and specific about what readers can take away as expectations for the next three to six months. And I gotta say, volume-wise, you know, depending on who you ask, it's it's not great news from what we see. We're projecting carload volumes to be relatively steady Q4, but down 3% in Q1 2026, down 5% in Q2, and intermodal volumes down 5-6% in the near term, flattening off towards Q2 2026. And so I think the first thing to think about there is we expect relative underperformance in volume at a national level, car load and intermodal. And in the way our modeling works, and you know, as is helpful, we'll get into that, is that that then feeds into projections of actually faster service times at the Chicago Gateway Hub as there's just less volume pulling on those resources. So that's a long way of saying we see a shipper's market going into the new year.

SPEAKER_00 3:39

And when when you kind of think about you know the the broader view of things, and when I say things, I mean volume, being being depressed, when you kind of looked at some of the factors that are really that are really driving that, where where is the pressure coming from that is going to basically keep volumes lower than I think they have been historically when we think about maybe the the prior five years?

SPEAKER_01 4:06

Sure. I I think there's two ways to kind of approach that. One is what do we feel pretty confident is going to underperform last year? And then the second is when we look for green shoots, what do we see? And so when we think about what we feel pretty confident is going to underperform, I think the story of 2025 was international trade and tariff regimes, and and our models take that into account, particularly trade with Canada on the carload side and uh trade with China and Mexico on the intermodal side. And you know, I think we in the world have reason to believe that those trade relationships are not what they once were. Uh, brass tax, what does that mean in our models? It means that we've taken our seasonal forecast for trade with those trading partners down a few pegs going into the new year, and we think that that will give us a more accurate model. Uh, and and so you know, the first point I guess is internationally there's a lot of headwind. I think one thing that you could look at and say is, well, yeah, but isn't some of that intended to create tailwinds domestically? Aren't we hoping for domestic growth to replace that? And you know, maybe that certainly is the intention, but I just don't think from a quantitative point of view, we have reason to expect that in the near term. Some of the things that we were looking at to get a sense of what our manufacturers are doing were the production managers index. Um, this time around, we also look closely at the logistics managers index. Those are both, you know, for people who don't know, uh, diffusion indices of what those managers expect going forward. So a production managers index is basically a survey of production managers' sentiment. What are we purchasing for inputs to be ready for our plans in the next several months? Right. They're trending downward. So, you know, I say that in that, you know, as we considered our modeling approach, we had to consider both the headwinds on the international market and the general sort of malaise we observe through the survey data in the domestic market. Those are the things that I think came together to give us these underperforming forecasts.

SPEAKER_00 6:19

Right. What what's what's interesting is obviously you you set kind of the foundation with the first section. Now with with this, we're we're revitalizing that. And I think you did a lot of a lot of additive effort. What's really interesting is we were recording this on the day after CSX released its um fourth quarter and full year 2025 kind of uh financial discussion. And the clearest thing that I took away from that, yes, they pulled guidance, yeah. A lot of it was a very clear statement. We do not anticipate any meaningful improvement in the macroeconomic conditions in the near term. So that I thought I think that's that's a very interesting piece, and we we can definitely talk about that later. Um, look, one of the things that I've always struggled with as an individual, when I consume macroeconomic forecasts, and oftentimes there is some layering of opinions that to your word tends to be very hand-wavy. Um how do how do we how are we giving something to readers that they can trust? Like, like how are we kind of like validating the data and the application of that? Give me more there, because I I think that's a big question that that many people probably have.

SPEAKER_01 7:26

No, that that there's two broad ways to do it, and I appreciate the question so much because I think we all should be critical readers. You know, if if you're reading things that we write or you're reading things that anyone writes, you know, continuously interrogating that text for should I believe you is 100% what we all should be doing. And and so two ways I, you know, I hope everyone reads our report. I'm very glad that this is something that we can share with everyone for free. And I hope people take that same interrogation to to our work here. And so the the two ways I would suggest you know, we all do that. One is just, you know, plainly doesn't make sense. Do the the facts of this story fit together in a way that aligns with the way I think the world works? Um, and then the second one, and we did the legwork for you there, is well, if you're making a point forecast, you know, we're quantitatively saying this is how much we think will move car load in the US uh in February and March and April 2026. You can look back and compare our previous forecast to the data that came out after we made the forecast. And we did that for you, and and you know, I'm I'm very proud and excited to share that our results were pretty good. 98% accuracy in our car load forecast, 97% accuracy in our intermodal forecasts. But you know, as as you and I have digested those results, we want to be very upfront with ways that people should think about those high accuracy numbers. So, yes, it's good. Um we've plotted them in the report so you can see where we were up and down. But you know, really, we forecasted um we had four months of forecast data come out since we released our forecast. So four points of comparison on the intermodal and the car load side. And those are mostly Q3 and one Q4 data point, which is kind of a steady period. So, yeah, it's great we got this over 90% accuracy, but to kind of couch that it's really only four points in the data series, and it's during relatively calm waters. I think the real test for our method going, you know, will be this report going forward, because we see a big dip in January. That's a seasonal dip, um, but that's that's real movement on the line chart, so to speak. So if we got that right, I think we'll have even more confidence, and hopefully our readers will have even more confidence going forward.

SPEAKER_00 9:46

Yeah, yeah. Look, I love the results, but I also love the explanation of how we got to the results. I think one of the really interesting things is that there are always pockets of um, I would say, disagreement with how the kind of the freight market is gonna go. And some of that is based on fundamentals, some of that is based on desiring people to click certain headlines. But I think one of the things that we've seen is that the truck market has had some undulating thing. There's an idea that this the storm coming is gonna drastically reshape freight demand for a long, long period of time. Um, and there's also been a thought that exports um from China heading uh kind of building up towards Chinese New Year have actually been depressed. And I think there's some initial signs the bookings may be higher than anticipated there. So it'll be interesting to see how all of that play plays out. Going back to our methods and approach, one of the things that I always get a little uh animated about is how how is kind of the how are the things made in the kitchen? And is this a situation of, hey bro, did you just like do AI to like get the things? So tell us a little bit more about how uh in non-technical terms, how this all kind of comes together.

SPEAKER_01 10:56

No, I I I'm glad you asked and and I agree with that sentiment so much because we are in a unique um point in human history, really, where anyone can go to a large language model and ask any question and get any result. So I think kind of to your earlier point about how do we know what we should trust, that question is even more pressing in our current environment, perhaps than it's than it's ever been. So when when we build our forecast, one of the things that you know was important, I think, philosophically to the team and to you, as you and I discussed it, is like we we don't just want a number to come out and then we say, look, it's 40 and walk away. Like we want to know how and why and which ingredients went into that dish and and what justifies including those ingredients, if you will. So the way we did that is we said we need two broad buckets of um inputs or ingredients to continue the analogy. One is the time series history. So, what is the history of carload movements in the United States? What is the history of intermodal movements in the United States? And then also we we spent a lot of time and we put a lot of sort of lab resources to what external data produces good forecasts of those values. So, you know, to contrast that, the hey bro, is it AI approach would be here's all the data I can download, model, please look at it, and then give me a forecast. In that case, you don't, you're not really curating the inputs, and you don't understand the relationships to the output. We chose not to go that way. We chose to say, let's curate our inputs so we have different, what I would call exogenous factors that feed into our forecasts. Um, you know, for example, for the car load model, that includes things like trade with Canada, U.S. coal production, U.S. steel production, U.S. paper production, things that we know and have rigorously tested how they correlate to uh carload volumes.

unknown 12:57

Right.

SPEAKER_01 12:57

And then we use those known ingredients as the inputs to our model. And you know, for any technical people watching, I don't want to get too technical, but uh, the resulting model is something called a Ceramax formulation with exogenous variables. So it allows us to use the history and the outside information to get a forecast. One more point on that, just because I know and know everyone is understandably excited about machine learning and AI tools. You know, why are we using what is essentially a statistical probabilistic measure here when all those other things are out there for free? Because the results are so good. Like right now, we have really good accuracy and we can explain every step. To my mind, if we were not getting good accuracy, then we would start to weigh the benefits of giving up some of the explainability by using some sort of large language model with unknown inputs. But for now, our models have been performing very well and they've allowed us to understand the results and to also massage them. You know, we were able to say, what does the data tell us to expect for trade with Canada, for example, over the next six months? And then we can manually ratchet that down and say, but what if the relationship gets worse and see the results? You know, and we're able to do that because we've built the models from scratch, if you will.

SPEAKER_00 14:19

Things are really made. So laying that foundation actually added a ton of new perspective into this second edition. And so, what's new in this installment that we actually didn't include in the first one?

SPEAKER_01 14:31

Sure, there's a brand new visual and analysis that I'm really excited about, and I hope readers will take in and let us know what they think of it. And it's it's um using the Telegraph platform data for the first time. So our last installment didn't use Telegraph's own data, and here we're using it to one very specific end. We're looking at across thousands of observations in our platform, how long did it take to go from metro area to metro area by rail? So, Chicago to Los Angeles, for example, over thousands of observations. We're looking at a confidence interval around that time. So, you know, how many hours did it take, give or take? And then we're comparing that against two benchmarks. One, how long it would take to make that same trip via truck, and two, where does that confidence interval fit compared to the Q4 2025 average? That second piece is to give a sense of what was trending faster or slower over the course of the year. So I'm really excited for readers to digest that, you know, maybe help us improve it if they see things that we could find ways to make it even more helpful. But what we are able to do there is identify origin destination pairs that are getting increasingly competitive intermodal versus truckload. And Chicago to Los Angeles is one of those lanes highlighted for that reason in the report.

SPEAKER_00 15:55

Yeah, look, I think one of the things that I've always appreciated about um railroads coming out and publicly talking about their dwell numbers and train speeds being being where they are is that that that's great. Let's actually make it super applicable. Um, and and so that's the piece that's always been missing. And so I'm glad we've kind of stepped up to the plate and are sticking our neck out and actually putting something useful in front of readers and saying rail in these lanes is getting better or more competitive relative to truck. And so when we come out and say it's a shipper's market, hey, you should evaluate over-the-road conversion, especially in a winter storm, right? There's actually some some teeth behind that, but we're actually doing the hard work of measuring that. Um, you know, and it would be great if the railroads did that and provided that, but candidly, like they don't see across their uh outside of their own network. And so that's what I think one of the benefits that we can really bring to the table. Um, and thinking kind of like a little bit more about us continuing to drive to why should shippers care and how does this influence the shipper perspective? What are some of some of those big takeaways that you think readers will have as they think about planning their procurement decisions over the next nine to 12 months?

SPEAKER_01 17:07

Sure. You know, two come to mind, and I've always enjoyed our conversations on this topic as we prepare the report. So so I hope you'll you'll you know build upon these comments because I know I get so much out of it when you do. But I I think the first one is is fairly obvious in that a shipper's market means that there's a certain um amount of power that confers to the shipper, the carriers want your business, is another way to say that. Now, railroading is a little different than trucking, it's not, is you know, the power dynamics are different. But I think we we've seen even with some of the customers who we've had these one-on-one conversations with, you know, there are lanes where the railroads are really seeking out volume, and shippers who have consistent volumes, who have high volumes on these lanes, are moving commodities that are not sensitive to some of these up and downs that we've been talking about. I think ought to seize that opportunity to flex that power and get competitive rates, particularly if the rates were negotiated several years ago, as we've seen, a lot of those are higher than I think they need to be. So we hope we can you know help people, help shippers make that case with this kind of analysis. I think the second thing is, you know, oftentimes we're talking about intermodal conversion, converting from truck to rail, but people have vast networks, so you know they need help identifying where the the most opportune lanes to do that are, and we have some of that going here. And in in a shipper's market, I think you're moving on rates and you're moving on where can I reposition my logistics decisions, you know, perhaps moving from truck to an intermodal solution most effectively. I think we're helping people find that in this market.

SPEAKER_00 18:51

Yeah, I think those are actually really two good points. It's funny, one of one of the one of the things that I think now being no longer working as a railroad or and being very much on the on the other side, I've it it lends a certain creativity that that um I feel like it's based in reality, but you know, who who really knows? I've oftentimes thought about if you're an intermotor marketing company or you're a railroad and you were desiring and you've you've said publicly, and you say it every quarter in the in the earnings call that over-the-road, truck-to-rail conversion is really the good the growth lever. And you've said that so many times for the past many years that I think hopefully people really believe it now. But one of the best ways to do that from a shipper's standpoint, and how do you kind of like you know, really incentivize, is that railroads really benefit when there is large amounts of consistent volume. Sometimes that is just simply moving on a stack train, sometimes that is filling a shuttle train, whatever it is, volume really benefits the railroads. But oftentimes the undulating forecast, which is something we're talking about today, can have a real big impact when you're running effectively kind of a fixed network. And so one of the things that I've often thought about is if there are shippers or IMCs or steamship companies that can commit certain amounts of volume to the railroad over a specific period, can they, in return for that, gain some value in rates or gain some value on a guarantee of consistency of service? And guaranteeing service is something that is very antithetical to the railroads, but candidly, that's something that you and I as consumers, and even if we were running a business and shipping packages, that some providers would offer in that space. And so, going back to your point around it being a shipper's market, I think there are things that you can do, and that if you're sitting across the table from some entrepreneurial under the gun, we've got to drive up revenue railroads, they they should be willing to kind of entertain those opportunities, if nothing more in the short term, to get the numbers back to where I think people want them to be.

SPEAKER_01 20:56

Yeah. Oh, totally agree. And and I think some of the national level analysis we do is really just backdrop or context to those one-on-one conversations between shippers and carriers. You know, if if we see this going on nationally, we're probably part of that trend. How do we get ahead of it? How do we make our smartest decisions given that context?

SPEAKER_00 21:15

Right, right. And that I think that is going back to why we even started the market contest was to really equip the broader industry with similar information and inferences and insights that they can then leverage to hopefully continue the conversion, leverage rail more, and then generally the the industry should benefit from that. Um, David, it has been an absolute blast catching up with you again. We we don't do this enough. Um it's been really fun to talk through this kind of second edition of the market consist. I I'm really excited about getting this up and out there. I think it's a it's a healthy, helpful publication, and looking forward to getting some feedback from uh from the broader environment. Thanks so much for taking the time to chat today. And uh for the audience, thanks for taking the time to listen. Thank you. And thanks for watching. Don't dwell on it. Stay tuned for more. Follow the show wherever you listen to podcasts and find us on LinkedIn to keep up with what we're building.