Investors await the next leg of Class I rail earnings growth as the PSR story moves toward the rear-view mirror. This report explores the factors at play that can drive the future success of the railroads. We believe the Class I’s need change; if they cannot show volume growth, the group may be at risk to de-rate lower or be forced to look at merging.

Our Thesis

While PSR gains are not completely in the rear-view mirror, rails are likely to need another story to maintain the investor interest the group has enjoyed over the last two decades. Management teams steer investors towards a return to carloadings growth, despite LT trends of declines and loss of share to truck. This focus arises because the Class I railroads have underpenetrated their TAM for more than a decade, with carload growth lagging GDP. At the root of share loss has been an inconsistent service offering, rectifying which is intended to drive growth.

Our survey data validates that the shipper experience with the railroads is deficient. A persistently tenuous correlation between metrics and the shipper experience (lamented for decades), does not bode well for the conversion thesis if not addressed. We believe increasing 1) freight visibility, 2) ease of business, 3) operational flexibility and 4) service quality will be paramount to the Class I’s ability to stay relevant in public markets. We view the levers of growth as largely in control of the railroads.

Since 2000, U.S. Class I rail multiples have increased ~5 turns despite declining volumes (versus ~3 turns for S&P), due primarily to:

• Pricing higher than inflation.
• Implementing Precision Scheduled Railroading (PSR) and other productivity initiatives that led operating margins to increase +20 points. If Class I’s cannot show growth, the group may be at risk of de-rating lower.

Mergers have entered the conversation again recently as sources of growth are questioned. Establishing a transcontinental railroad would reduce interchange complexities and offer cost synergies. However, achieving this is a tall order from a regulatory perspective. Current regulatory precedent for the STB to approve a merger is prohibitive and would require a change in current rules and/or the railroads to agree to forms of increased competition such as full reciprocal switching. While investors should keep an eye on any rule changes/shifting precedents, we believe it is too early to lean on mergers as a growth lever for rail earnings. Focusing on organic growth is both a higher probability and more durable path to growth.

We analyzed 20+ years of rail history, conducted a proprietary TD Cowen rail shipper survey, and spoke with industry participants. We also examined industrial end markets, technological initiatives, and OTR trends to better understand the positioning of the Class I carriers.

Financial and Industry Model Implications

We believe there is a risk to carloading growth over the long term if service and growth are not prioritized by the Class I rails and their managements. While the near/medium term offers more idiosyncratic opportunities that may enable another investment cycle to play out, we see potential downside risks to longer-term carloadings growth.

What To Watch

• Service metrics: Rail service deteriorated sharply during COVID when layoffs were met with an inflection in freight. Each of the U.S. Class I’s saw high-double-digit deteriorations in avg speed peak to trough, with the Eastern rails seeing the worst declines. Rail service metrics have shown improvements, but shippers need service consistency to make long-term changes to their supply chains. We continue to monitor service metrics the rails publish and believe consistent steady improvements will be imperative.
• Shipper perception (TD Cowen proprietary surveys): Despite service showing improvements, there is a gap between published railroad service metrics and shipper sentiment on rail service. We caution investors against relying solely on railroad service metrics data. Rail shippers have consistently (since we began digitally tracking our survey in ‘16) wanted to put more freight onto the railroads, though poor service metrics have dissuaded them. We will continue to monitor our quarterly survey data which captures shipper sentiment on rail service and modal shifts.
• Regulatory changes: Rail mergers have entered the conversation again recently as sources of growth are examined. If the rails fail to prove they can grow with a concerted effort made by the industry, they may be forced to press the merger issue or else face multiple contraction. We acknowledge that this is a tall order given regulatory obstacles. There is still a remaining seat left on the STB and we expect it to be filled shortly, which will give Republican nominees a 3-2 board advantage.
• Technology adoption: Lack of shipment visibility is a key pain point for shippers. The RailPulse initiative spearheaded by a broad coalition of rail industry participants, including Class I’s, railcar lessors and equipment manufacturers, aims to plug these gaps. We look for continued adoption of RailPulse (currently 1% of the North American fleet), among other technology initiatives. Regulatory changes at the FRA could also enable the industry to fully take advantage of some overlap between labor and technology. We expect AI will be a key theme for railroads as the technology evolves; areas for focus will likely be in 1) quoting bids, 2) optimized scheduling, 3) safety improvements, predictive maintenance, and 4) corporate back-office costs. While the Class I’s have not explicitly outlined the opportunities from AI, we expect this to be an emerging theme in 2026.
• Gains made by autonomous trucks: Autonomous trucking has always been on the proverbial radar screen for rail operators but never taken that seriously as it was viewed as being “7-10 year away” 7-10 years ago. While it has not arrived in the ubiquitous fashion envisioned, we are starting to see the beginnings of a driver out model for some operators. We do not see autonomy being a headwind to short- to near-term growth, but note it bears watching for the longer term.

Stock Conclusion: One More Cycle in the Tank.

We believe the U.S. Class I’s have idiosyncratic tailwinds that should support another cycle before investors look for another leg of growth. Longer term the railroads face a serious threat of earnings deceleration and multiple de-rating if Class I’s cannot improve their product offering and ultimately win back share from the highway. While the 6-year average is at ~17.5x, the group was trading in the low teens in the early 2000s. Was the 21x + multiple the group traded at during 2022 the high watermark for the group? We favor Union Pacific in the U.S. Class I group because of 1) positive pricing momentum in contract renewals, particularly in multi-year contracts, 2) strong operational gains and productivity initiatives, and 3) with a forward PE at 17.5x it is the only U.S. Class I trading below its forward five-year average.

Executive Summary

Investors continue to await the next leg of Class I rail earnings growth. While margin gains from current levels are expected, PSR is no longer the investor story it once was, and we believe the rails must prove they can grow to maintain their premium multiples. In this report we look to explore potential drivers and factors that can drive future success including 1) service-driven truck conversions, 2) regulatory relief allowing technological advancement, 3) industrial development benefits, and 4) potential merger producing a transcontinental network (something we view as a last option for the industry). While the track record for industry volume growth has been questionable, we view the levers of growth as largely in control of the railroads.

Growth initiatives are one of the more recent strategic focus areas of the rail industry. This focus was borne out of the fact that the Class I railroads have underpenetrated their TAM for over a decade with carload growth lagging GDP. At the same time, trucks’ share in NA freight has expanded considerably, reaching ~70%-80% of the total market in 2024 (up from the low-to-mid 60s in the early 2000s) despite rail being the cheaper mode. At the root of share loss has been an inconsistent service offering, brought starkly into attention during COVID but improved since, which the Class I’s believe will drive growth.

Our proprietary survey finds that although rail service metrics reported by the Class I’s have improved considerably, shipper experience remains lacking: only 66% of shippers receive >80% on-time performance from rail partners compared to truck and only 4% receive a service guarantee compared to 37% for truck. A persistently tenuous correlation between metrics and the shipper experience, one that has been lamented for decades, does not bode well for the conversion thesis if not addressed. Freight visibility initiatives in the form of RailPulse are encouraging but are in relatively early stages. It is our view that the industry needs to improve connectivity with its customer base. A key driver for volume growth will also be driven by the Class I’s’ ability to increase the ease of business, which has been one of the leading reasons for loss of share to the highway. Since we began our proprietary rail survey in 2016, shippers have had a ~similar answer, in that they believe 1) rail service needs to improve, and 2) they would move more freight via rail if the service did improve. Rail service metrics have improved over time, though some industry participants disagree (and as the old saying goes, perception is everything).

While we have seen service quality generally improve over the past year, consistency remains key for shippers and overall rail reliability has not proved resilient. PSR has been a leading thesis as to why service has not been consistent, though our analysis suggests no correlation between operating margins and deteriorating service. The decline in rail employees pre-PSR has not been overly concentrated in one part of the business.

Total U.S. rail carloadings are down 11% between 2014 and 2024, and down 1% when excluding coal. A secular decline in coal is the primary driver for overall volume declines, with coal declining 52% over the same period. If we exclude coal carloadings, volumes are down just LSD with the industry being buoyed by automotive and intermodal growth. We estimate U.S. intermodal volume growth of 3.5% in 2026 and expect LSD volume growth over the long term. International intermodal growth will be dictated by global trade and expanded market reach (service levels and the ability to use transloading options). Automotive has been the largest growing carload segment; the primary driver is growth in North American auto production – Mexico has been a key partner that has enabled growth. We estimate 2.3% automotive volume growth for the U.S. Class I group in 2026, which could prove conservative given TD Cowen’s forecast of 4% for the overall auto industry.

Since 2000, U.S. Class I rail multiples have increased ~5 turns compared to ~3 turns for the S&P (when comparing 2000-2007 averages (ex. ‘00 outliers) vs 5-year-present avg ex. COVID). This was largely due to two factors. The first was pricing as railroads were able to reprice long-term legacy contracts that were at severely below market levels starting in the back half of 2003. Pricing above rail cost inflation enabled the industry to quickly improve margins. After that, the industry saw the implementation of precision scheduled railroading (PSR) beyond the first railroad (Illinois Central/CN) that took on the operating model. Operating margins have increased 20+ points over the same period and rails saw significant stock multiple expansion. This result, however, came despite a decline in volumes.

While PSR gains are not completely in the rear-view mirror (we note that the last railroads to announce PSR were Norfolk Southern and KCS in 2019), rails are likely to need another story to maintain the investor interest the group has enjoyed over the last two decades. In our view, it is growth that will help keep that interest. However, if Class I’s cannot show growth, the group may be at risk to re-rate lower. The 6-year Class I rail multiple average is at ~17.5x, but investors should note that the group was trading in the low teens in the early 2000s. This leads us to question if the 21x + multiple the group traded at during 2022 was the high watermark.

The growth outlook for the Canadian rails is quite similar vs. that of their U.S. counterparts, in that it is heavily predicated on intermodal, including truck-to-rail conversion, as well as certain franchise-specific initiatives. One key differentiator, however, is that the Canadian rails are under far less pressure to backfill a secular decline in utility coal vs. the U.S. rails, which is important, because coal is much higher margin vs. intermodal, particularly for the eastern U.S. rails.

Given the more service-sensitive nature of intermodal, and the negative regulatory/ political attention that the industry attracted during the post-COVID service breakdowns, the Canadian rails, along with their U.S. peers, are taking a different approach to resources, particularly labor, going forward. The industry has effectively committed to carrying surge capacity during low points in the cycle to enable better service responsiveness when demand inflects positively. The result should be better earnings over the cycle, including better peak earnings, but also less downside cost variability than has been the case historically.

CN’s multi-year growth plan includes contributions from intermodal, energy/chemicals/plastics, metals/minerals and bulk (grain/potash/renewables). To date, the company has seen the most traction in refined fuels, NGLs and frac sand. Intermodal growth has been hampered by rail/port labor disruption in 2023/2024 and a weak trucking market.

CPKC is three years into a transformative merger, which provided unique single-line connectivity between Canada, the U.S., and Mexico. Expected merger synergies are relatively balanced between intermodal/automotive, bulk, and merchandise, but the company has seen the most traction to-date in intermodal, automotive, energy/chemicals/plastics, and grain.

For the industry, an improving U.S. regulatory environment is evidently materializing with “pro-carrier” nominations to the STB and FRA though we believe these are likely to have marginal benefits on network fluidity rather than support growth in a major way. Automated inspections in particular hold promise per multiple conversations we have had with the Class I’s and their short line partners.

Finally, rail mergers have entered the conversation again recently as sources of growth are examined. The establishment of a transcontinental railroad would likely improve efficiency by reducing interchange complexities and drive cost synergies. However, we acknowledge achieving this is a tall order. Established regulatory precedent for the STB to clear a merger remains prohibitive in our view following the scrapped

BNSF/CN combination and subsequent large mergers are required to prove that they increase competition (as opposed to simply preserving it – something that the CPKC merger has clearly accomplished). In a change of control scenario, we calculate that the CSX CEO has a materially higher payout compared to the Norfolk Southern CEO – up to ~7x.

That is not to say the path to such a venture would be insurmountable. To make the path viable for the industry, one of two things must happen in our view.

• The U.S. rails could offer up a much more expanded version of reciprocal switching, something closer to what exists in Canada vs. the much watered-down version the U.S. currently possesses.
• A change in current STB rules, an unlikely enough proposition that CPKC’s CEO Keith Creel emphasized a deal would face “mountains of regulatory risk.” While investors should keep an eye on any rule changes or shifting precedents, we believe it is too early to lean on mergers as a growth lever for rail earnings. If, however, the rails fail to prove they can grow over the medium term with a concerted effort, the industry may be forced to press the merger issue or else face multiple contraction from current levels.

Stock Takeaways

We believe rail earnings remain partially at the mercy of the marketplace and macro in the near term as shipper testimony does not suggest a step change in modal choices favoring rail is imminent. Regulatory relief and industrial development offer benefits but do not address the core service bottleneck or ease of business issues. This is not to say that the current outlook is uniformly bleak. Indeed, Union Pacific is operating very well and Norfolk Southern is in the midst of turning around operations after a year the railroad would like to forget. CSX remains an underperformer in the U.S. group as the network copes with a major intermodal project on one of its most important lanes and the rebuilding of an entire subdivision that was devastated by historical flooding conditions in April (although recent numbers are starting to show improvement). We favor Union Pacific in the U.S. Class I group because 1) positive pricing momentum in contract renewals, particularly in multi-year contracts 2) strong operational gains and productivity initiatives and 3) forward PE at 17.5x is the only U.S. Class I trading below its forward five-year average.

In Canada, CPKC continues to benefit from its ~$30 billion acquisition of KCS, while CN expects earnings growth to accelerate in 2H25, as it laps easy prior year comparables. We prefer CN over CPKC because 1) CPKC has greater exposure to cross-border revenue vs. CN (41% of total revenue in 2024 vs. 32% at CN), and cross-border Mexico volume growth is an important component of CPKC’s expected multi-year merger synergies; and 2) CN is currently the cheapest stock in the group based on an average of P/2025E and P/2026E consensus EPS, a discount which is highly unusual by historical standards, and which should reverse as the company’s earnings growth reaccelerates higher in H2/25, partly due to the benefit of easy prior year comparables.

To get more confident in the longer-term rail secular growth thesis we would need to see 1) shippers acknowledge structural improvements in on-time performance, 2) rails continue to press for increased supply chain visibility and the improvement in the ease of doing business (AI could eventually help with this), 3) an upside surprise in reshoring activity/conversions or 4) major shifts in legal precedents governing large rail mergers. Cyclical catalysts mainly in the form of a recovery in competing TL pricing following a historic downturn could support the rails near-term. However, the broader investor appeal they enjoy requires durable growth drivers over the longer term in our view.

Historical Rail Volume Drivers

The Class I rails possess a wide TAM as they transport the physical goods share of North American real GDP as well as import volumes across the full range of commercial end markets, both industrial and consumer. Rail carloads are therefore naturally sensitive to broad economy-wide macro indicators.

Even so, Class I carloads in aggregate have recorded meager growth since the turn of the millennium posting a 0.4% CAGR between 2000 and 2024. Excluding coal (which remains in a secular decline unrelated to macro with a -3% CAGR), carloads fared only slightly better with a 0.9% CAGR. At the same time real GDP grew at a 1.8% CAGR and industrial production grew at a 1.7% CAGR during the same period.

In the 2000-2024 timeframe, carload growth has been driven by intermodal growth of ~53% or a 1.8% CAGR. Ex intermodal, carloads would have clocked a -0.6% CAGR or ended down 14% from the starting point, speaking to the significant influence this mode has had on rail volumes. That said, intermodal growth has lagged not only real GDP but also real import growth which FRED data places at a significantly robust 3.1% CAGR. This validates a well understood phenomenon namely that the rail network has ceded intermodal-eligible freight share as well, strong growth notwithstanding. In fact, railroads appear to have given up market share in lanes where they have had a historical advantage in due to a longer length of haul, according to several conversations we have had with industry stakeholders.

Rail Service

Underlying this volume share loss is a wide gap in service quality offered by the rail network compared to over-the-road (OTR) trucking alternatives that has been well understood for a long time but was exacerbated and therefore pushed to the forefront of the rail industry narrative following the extended spell of COVID congestion. Rail service quality deteriorated sharply during this period as the Class I’s found themselves woefully under resourced for the freight deluge that took place in response to stimulus spending. Train speed and dwell displayed sharp deteriorations during this period. Each of the U.S. Class I’s saw high double-digit deteriorations in avg speed peak to trough, with the Eastern rails seeing the worst declines as seen in the chart below. Correspondingly, the network also saw sharp increases in terminal dwell during the same timeframe. Declines in intermodal were tempered relative to merchandise but still, because of the quick decline in service metrics, the OTR mode was able to resume freight share gains that have been underway since the 1990s. Indeed, truck share of total U.S. freight transport (by value) has increased ~7-14 pts in the past three decades depending on the chosen source (Bureau of Transportation Statistics data lands in the 7-8 point range from 2000 to 2023). Other sources have truck gaining another 8 pts from 2022-2044.

PSR and Its Impact

The decline in rail employees pre-PSR has not been overly concentrated in one part of the business. The largest decline (when comparing Jan ‘17 to Jan ‘25) is in the maintenance of equipment and stores segment that saw a 33% decline. Professional and administrative employment has declined 27%, and Executive rolls have declined 24%, according to the STB.

In 2017, CSX began implementation of PSR (precision scheduled railroading) which worked to improve operational efficiencies on moving trains, not cars. Previously, Class I carriers focused on long, fast trains; carriers would wait to build long trains to move freight. CSX began reporting trip plan compliance in 2018; in short, there is no obvious correlation between operating margins and deteriorating service when looking at the numbers. Operating margins worsened by ~300bps from the 2018 period to 2024, and trip plan compliance has improved by over 10 points over the same time frame. Part of the reason is likely due to the pressure on intermodal rates in 2024 due to the ongoing weakness in the OTR market. CSX has 7% fewer employees in ‘24 vs 2017; total carloadings have decreased by 2% over the same time frame. Overall, Class I employees have declined meaningfully since ‘15, down 30%. In our 2017 rail shipper survey, we asked participants who utilize CSX to see if issues came out of the PSR implementation. 84% of rail shippers stated they were having service issues with CSX when PSR was being rolled out; 42% of shippers stated they switched less than 25% of their freight to the highway.

Total U.S. rail carloadings are down 11% from 2014-2024 levels, and down 1% when excluding fuel, suggesting share loss to truck mode predates PSR given the long-term trend of declines. When focusing in on U.S. carloadings excluding fuel, carloadings are still down 3%. However, PSR evidently added to prevailing pressures. The charts below plot intermodal carloadings against relevant TAM benchmarks (indexed to 2000) and the post PSR/COVID divergence in addressable freight and carloadings is stark.

Shipper Perceptions of Prevailing Rail Service

Despite well-documented improvement in rail service in the past 2 years, this notable service quality gap continues to prevail per our proprietary rail shipper survey as we obtained the following findings:

• 66% of shippers receive >80% on-time performance from truck vs only 16% from rail.
• Six times as many shippers receive service guarantees from their truck carriers compared to rail. 94% of rail shippers do not receive any guarantees.
• 54% of shippers wait for more than a week just to receive a quote from a rail provider.

The other issue in getting quotes out in a timelier manner for the rail industry has to do with the fact that the majority of freight does not originate and terminate on the same line. Hence, most quotes involve two or more railroads and could also include a drayage move. When we were conducting a fireside chat with a chemical shipper, he revealed that one of his quotes from an unnamed Class I took 64 days. While rails remain more constrained in giving quotes given most railroad moves include more than one carrier (and sometimes more than one mode), most would agree that more than 9 weeks is not acceptable when shippers are attempting to improve their own supply chain.

Ease Of Doing Business

With optionality between rail and OTR transport, it is essential for shippers to have a seamless, long-term business partnership with carriers. This section will analyze the ease of business with rail and trucking companies, looking at the following factors: 1) accessibility and reach 2) operating flexibility and 3) service quality that shippers get from trucking vs what they get from rails.

Accessibility and reach. OTR transport offers the easiest accessibility given its door-to door access across North America. Unlike railroads that operate on a fixed rail network, OTR carriers benefit from comprehensive highway coverage that create benefits like speed. Long haul lanes (which is where rail vs. truck becomes a more prevalent question) can take ~twice the length to cross the country via rail when compared to truck. Terminal dwell, which measures the average hours at a terminal for a rail car, is running at ~23 hours YTD in ‘25. While losing a full day of travel time when compared to truck, the Class I network has improved dwell by 21% when compared to pre-COVID levels and is running 13% below (better than) its long-term average. Decreasing dwell (driven by productivity measures) shortens overall transit times and improves ease of business. Train speed is up marginally vs its LT trend of 25.5 MPH.

Operational Flexibility

Historically, trucking offers shippers more operational flexibility because of its ability to choose schedules (vs rails that have fixed routes and scheduling), deliveries, and more seamless/timely quotes. We asked rail carriers how long it takes them, on average, to get a quote from a rail provider. Less than 50% can get a quote in less than a week, with 42% of respondents stated that it takes 1-2 weeks to get a quote. We asked this same question but for an OTR carrier/broker, and 94% of participants can get a trucking quote within a week (in some cases just hours). Automation, which has been a key theme for asset-light providers in recent years, can offer quotes in seconds. Additionally, 94% of rail shippers stated that their contracts do not have service guarantees, further impacting shipper flexibility and ease of business. We believe, however, that this presents an opportunity for rail carriers; creating more flexibility for shippers is a sticking point in modal choice.

Service Quality

Service quality ultimately depends on what a shipper is moving, as the product/volume depend greatly on whether a shipment is better fit for a rail or OTR haul. Trip plan compliance is the key indicator for rail service quality. Union Pacific, who calls their trip plan compliance intermodal service performance index (% of IM boxes on time), has made steady improvements on its on-time index, which improved 15% in 1Q when compared to 1Q23 levels. Service has held up well for the whole group despite strong volumes in 1Q as pull forward took place.

In our TD Cowen proprietary rail survey, we asked shippers if rail service has impacted modal transportation decisions; 35% answered yes, 50% answered No, and the remaining did not have an opinion.

We also asked shippers if rail service did improve, how much more intermodal freight they would push towards the railroads; 71% said they would only move between 0%-5% of freight, and 19% would move between 5%-10%.

Since we began our proprietary rail survey in 2005 (we have been using Surveyplanet to amalgamate the data since 2016), we ask shippers the state of rail service and how they would change their shipments if rail service improved. For nearly 10 years, shippers have had a ~similar answer, in that they believe rail service 1) needs to improve and 2) they would move more freight via rail if the service did improve. Rail service metrics have improved over time, though some industry participants disagree. One panelist we recently had on an industry call pointed to 50K of merchandise shipments he tracked, dwell time worsened at 75% of the terminals and transit times worsened by 10% over the last 13 months. Acknowledging severe weather this winter, panelists were skeptical on the prioritization of rail service and volume growth at the Class I’s.

“Modifying data is an industry sport” – Railroad stakeholder in a statement that suggests the railroads posted service metrics do not align with service rail shippers receive.

Shipment Visibility and the RailPulse Initiative: A Long Runway

Another notable element in the service gap between rail and truck transportation is the lack of shipment visibility in the former. The deficiencies are hardly on the shipper’s side who employ TMS software and have advanced ERP systems to aide in their inventory management. Rails intimate shippers once a shipment is created but on-time performance is relatively poor (~60% of shippers attested to <70% OTP) and in our conversations with shippers, not knowing where your railcars are at any given time is described as a natural state. Additionally, shippers typically own and maintain their railcars and therefore are interested in the condition and health of the equipment from a financial standpoint. Knowing how your rail partners are handling assets on your balance sheet and the cargo within it is therefore valuable to shippers.

The RailPulse initiative spearheaded by a broad coalition of rail industry participants including Class I’s, railcar lessors and equipment manufacturers aims to plug these gaps. Shippers have attempted to invest in telematics in the past, but efforts were naturally fragmented and occurred at a smaller scale than necessary given the size and complexity of the rail network. This model involves shippers fitting their owned or leased railcars with sensors and fittings from approved vendors and entering a subscription agreement to receive access to real-time data offered through an API. Features include comprehensive railcar location data, loaded/unloaded status, transit status as well as railcar condition information such as temperature, handbrakes and G forces.

These initiatives will likely produce a relatively inexpensive but significant improvement in visibility off a low base for the rails. Commentary from the Class I’s over the years has indicated that sensors report to the RailPulse database every 15 minutes which only costs $800-$1,000 per car (as of 2024) and is orders of magnitude better than Automatic Equipment Identification technology currently employed which utilizes RFID tags.

RailPulse is an encouraging initiative, and its promise lies in the broad coalition of support it receives. The interconnectedness of the rail network makes this type of coordinated effort more effective to the overall customer experience. Indeed, CSX recently highlighted that ~40% of their traffic touches another railroad. To be sure however, we are still in relatively early innings of the RailPulse rollout. Tech instrumenting on cars began in early 2023 and 5 telemetry vendors are likely approved at this stage (3 were certified as of 2024 and two more were scheduled for 1Q25). At the time of writing ~16,000 railcars are operating with RailPulse enabled, i.e. ~1% of the total North American fleet. RailPulse coalition members and partners account for 35% of the ~1.6 million-strong North American fleet and therefore expansion of the offering is likely to continue in the coming years in our view and is likely to be a significantly protracted process. Norfolk Southern disclosed at a conference last year that universal application of telematics in the rail industry has a 2035 target. Additionally, while improved visibility should support shippers’ inventory management processes, it is not a direct cure for inconsistent service, which relates more to core operating philosophy.

Rail Regulation’s Role in the Growth Algorithm

With a change in the Administration, we believe we will see a shift at the main U.S. rail regulatory authorities; the Surface Transportation Board (STB) and the Federal Railroad Administration (FRA). Both changes are likely to be positives for the rail group in our view but for different reasons. Neither are likely to have an outsized impact on growth, but one change has the potential to help in our view. In January of this year, Patrick Fuchs became Chairman of the STB, succeeding acting Chairman Robert Primus who took over for Marty Oberman upon his retirement last May. We view Fuchs as more favorable to the railroads space versus the more adversarial role former Chairman Primus tended to take. There is still a remaining seat left on the STB and we expect it to be filled very shortly (see note here). This will give Republican nominees a 3-2 board advantage. While we do not feel the board will change how it looks at cases brought before it, we do believe it will alter its approach to being more reactive rather than the more pro-active regulatory style it exhibited in recent years.

While those changes at the STB are viewed by us in a positive light, we believe the nomination of David Fink to head the FRA will be more impactful for the rail industry despite his expression of support for two-person crews at his congressional hearing. Investors may recall that the FRA is the agency that oversees the creation and enforcement of rail safety regulations, administers railroad funding and researches technologies associated with rail operational improvement. He was nominated by POTUS 47 back in January and recently cleared the U.S. Senate Committee on Commerce, Science and Transportation. The nomination must now go through the Senate. Mr. Fink is a lifelong railroader, having most recently been the President of the Pan Am Railways (a Class II New England rail network that was sold to CSX). We view Mr. Fink as likely to have a very “pro-carrier” stance. The FRA has been long accused by industry insiders of limiting the exclusive use of automated inspection devices to favor labor friendly visual inspections. Getting clearance to utilize more automated inspection devices in lieu of mandated visual inspections, however, could offer a minor boost to network fluidity in our view. Several Class I executives have mentioned that this move would enable the railroads to turn their workforce from “finders” to “fixers.” In a recent investor dinner we hosted, CSX expressed optimism on FRA’s likely direction under Fink despite commentary on the two-person crew rule. Significant savings in the inspections process was highlighted.

Industrial Development Growth Hinges on Reshoring

We believe industrial development represents a growth driver, but its sustainability relies on reshoring trends which can be driven by regulations/laws out of Washington, DC. In particular, the current “Big, Beautiful Bill” that is being taken up by the Senate includes a new provision that allows bonus depreciation on structures. The Eastern railroads benefit from this trend to a greater degree.

The rails’ deep infrastructure and construction expertise affords a unique degree of integration particularly with industrial shippers which is expressed in the form of industrial development partnerships. Under these arrangements, the rails offer industrial shippers value added services in the form of real estate selection and some architectural design while integrating rail solutions and capital investment to service these shippers’ transportation needs. All the Class I’s have highlighted industrial development projects as a growth driver, particularly as manufacturing reshoring emerges as a policy priority in the U.S. Indeed, the Southeast has been a focal point for the industrial development pipeline at the Class I’s.

Our view is based on reported carload projections from industrial development. CSX and Norfolk Southern have recently projected incremental carloads from industrial development business at ~150K-300K and >150K carloads over a multi-year period respectively which represent volume additions of ~1%-2% annually (net attrition). Tailwinds to overall revenue are likely to be slightly more favorable given industrial volumes resulting from these projects have higher than average yields.

Nonetheless, we view the industrial development story as somewhat linked with the service and truck conversion narrative discussed in detail above. Almost all industrial development customers for the rails will also construct truck docks or terminals on their sites as shippers value the flexibility of multimodal transport. At the recent NEARS City Series event that we hosted, the head of Business Development at CSX attested to site customers’ heavy utilization of the OTR mode. While this represents a conversion opportunity for the railroads, it also means industrial shippers are not necessarily captive. Minimum volume commitments likely place a floor on volumes and support project IRR but upside and growth is dependent on the service proposition in our view. That said, if the service product can improve, industrial development represents an increased base of business to go after.

If All Else Fails, Attempt to Merge

Rail mergers have entered the conversation again recently as sources of growth are examined internally and external forces question if such a move is needed. If the rails fail to prove they can grow with a concerted effort given by the industry, they may be forced to press the merger issue or else face multiple contraction. We acknowledge that this is a tall order given regulatory obstacles.

Rationale for a rail merger from an efficiency and profitability standpoint are well understood across the industry and investor base. In particular, a single transcontinental network would significantly reduce interchange complexities which add a significant friction to freight flows. Another supporting rationale is the dwindling pipeline of executive talent in the railroad industry which also bolsters the case for a consolidated network established under the leadership of an able operator.

However, we acknowledge achieving this is a tall order as the regulatory paradigm based on prevailing precedents raises serious obstacles that even a move toward reduced oversight is unlikely to overcome. Key of these include:

• Enhancing competition: Following the scrapped BNSF/CN combination, large mergers are required to prove that they increase competition as opposed to simply preserving it. With only 6 Class I rails remaining this remains a high bar. Merger candidates would need to demonstrate continued access to open gateways and would likely have to justify price hikes beyond a certain level from the STB, as in the case of CPKC.
• Minimal overlap: Critical and related to competition is network overlap. CPKC deliberations benefited from the two networks having relatively minimal overlap (“end-to-end”) which placed the emphasis on truck conversions. CPKC CEO Creel recently noted that a prospective merger would not meet this standard for overlap.
• Shipper support: The CPKC deal received considerable backing from the shipper community (450 support letters) as a single-line service offering was indeed service and competition enhancing. The STB will be intensely attuned to shipper feedback and a prospective merger would require robust outreach and communication with shippers.

Some have speculated that a would-be rail merger could offer to accept full reciprocal switching in order to maintain the competitiveness of a transcontinental railroad. Recall that existing rules fall well short of a blanket reciprocal switching paradigm with eligibility established on case-by-case basis depending on service metrics that a captive shipper is facing, a much higher hurdle that gives room for tightening of that rule. That said, full switching will likely face opposition from the broader industry (i.e. the two U.S. Class I’s not merging and shortlines), notwithstanding the preferences of merger parties. Commentary from the Class I’s and short lines during the hotly contested 2016 litigation of a full reciprocal switching proposal illustrates the broader opposition clearly.

• CP argued switching in the U.S. would significantly disrupt service and that Canada’s interswitching rules were designed for dense urban areas.
• CN emphasized that voluntary switching agreements were sufficient to service captive shippers and echoed CP’s comments about the limited context in which Canadian interswitching operated. 3) Norfolk Southern and Union Pacific noted that full switching would disincentivize rails from making necessary capital investments in the network and ultimately would result in poorer infrastructure and service. A Union Pacific executive noted that rules would add 48 to 96 hours to transit times.
• BNSF pointed to limited capacity at the STB to process switching requests and reminded the Board about a failed switching experiment run in Louisiana in the prior decade.
• Short lines like Watco, Patriot Rail and Genessee & Wyoming were sympathetic to the spirit of reciprocal switching in terms of supporting competition but highlighted that execution would raise challenges.
• Due to these reasons, we believe switching might not be the silver bullet for a transcontinental deal and is unlikely to even be the only concession that would need to be made. We also see opposition from the unions as a hurdle and the current administration appears to be more pro-union than the prior Trump administration. While investors should keep an eye on any rule changes or shifting precedents, we believe it is too early to lean on mergers as a main investment thesis.

End Market Outlook, Rail-Specific Growth Initiatives

Intermodal

Since 2011, rail carloadings have increased 33% and is the leading growth story for rail volumes. The selling point for IM is that it can offer truck-like service for a lower price and help shippers with ESG initiatives, given it is 3-4x more fuel efficient than truck (which we have shown the math for in prior notes a few years ago).

We estimate U.S. intermodal volume growth of 3.5% in 2026, and expect LSD volume growth over the long-term, largely driven by the assumption of the development pipeline the rails have spoken to. CSX stated that 95% of its development projects, which are expected to add 150K-300K carloads by 2027, are in its merchandise segment. This new capacity is part of the company’s growth strategy and plays into the re-shore theme.

Despite near-term trade uncertainties, Mexico remains the U.S.’s second largest trading partner and is a key area for cross border intermodal growth. International intermodal growth will likely be dictated by global trade and expanded market reach. Sales teams must win share, though longer-term trends of re-shoring may create a more uncertain picture for international trade lanes.

Motor Vehicles and Equipment

Automotive has been the largest growing segment for carloading growth over the past 10+ years, at +51% since 2011. The primary driver is growth in North American auto production – Mexico has been a key partner that has enabled growth. EV sales growth is also a driving factor, which was also discussed by CSX at its investor day in 2024. Automotive production has had a slow start to 2025, due to lower production, largely driven by tariff uncertainty. Longer-term tariff impacts could be a positive for domestic auto production. This would likely benefit the Eastern carriers more, given most auto plants are located on the East Coast.

We model 2.3% volume growth for the U.S. Class I group in 2026. This is more conservative than TD Cowen’s auto analyst Itay Michaeli’s motor vehicle production forecast of 4% for 2026. U.S. carloads for autos are down LSD to begin ‘25, and we don’t expect carloads to improve in the near term, largely due to tariff uncertainty and higher interest rates.

Coal

Coal volumes have steadily declined, decreasing by over 50% compared to the 2011 period. In the medium term, we would expect this trend to remain ~flat (CSX called for flat volumes for coal at its investor day) in the medium term, but longer term expects this segment to be a secular decliner (albeit with an acknowledgement that the current Administration could help slow the rate).

We model a 2.3% decline in coal volumes for U.S. rail carriers in ‘26. This is despite coal carloadings starting off the year strong (after a decade + secular decline) at +6% YTD. This is largely due to natural gas pricing and increased energy demand. Over the medium-term, there is potential for coal to fill the energy gap given lower grid capacity as data centers/AI demand stresses the grid. The current administration is less sensitive to clean energy projects, which may keep more coal plants online. Coal RPU is also significantly higher than, for example, intermodal RPU; stronger coal demand (medium-term growth unlikely but stabilization may be more realistic) is accretive to margins.

AI’s Future Impact

We expect AI will be a key theme over the longer term for railroads as the technology evolves; areas for focus will likely be in 1) quoting bids 2) optimized scheduling 3) safety improvements, predictive maintenance and 4) corporate back-office costs. While the Class I’s have not explicitly outlined the opportunities from AI, we expect this to be an emerging theme as we move into 2026.

Currently, the Class I’s are in slightly different, though very early stages of AI deployment. Union Pacific expects AI driven productivity savings of $100MM over the next three years, focusing on areas of transport planning, optimizing resources, decision support tools, and automation.

Norfolk Southern is developing AI to help with digital inspection portals, track inspections, and route/yard planning. AI can analyze data sensors that should create a more efficient, safer railroad by eliminating human error with higher volumes of decisions being made. Internal “ChatGPT-like” tools are being implemented across organizations that should improve decision making for employees with internal policies and procedures. This may enable a do more with less approach to administrative work.

While management teams we speak with acknowledge the opportunity AI has for the quoting process, it is not yet in practice. Given the ease-of-use gap between rail and truck when getting quotes, we believe AI will offer significant service quality benefits over the longer-term. Railroads that embrace and adopt AI strategically will likely gain a competitive edge over competitors, in our view, and has the potential for another leg of margin improvement for the group.