CNR.TO — Canadian National Railway

TSX: CNR.TO | NYSE: CNI | Generated May 2026

CN Rail operates the only freight railway in North America connecting three coastlines — Pacific, Atlantic, and Gulf of Mexico — a physical network assembled over a century that no regulator would permit today and no competitor could finance. After a difficult 2024 disrupted by labor lockouts, cascading port strikes, and the opening shock of US-Canada trade friction, the business has largely recovered: the 2025 operating ratio improved 120 basis points to 61.7%, earnings reached a record C$7.57 per share, and the dividend was raised for the 29th consecutive year. The stock now trades at roughly 16 times forward earnings — a meaningful discount to its 10-year average multiple — but the discount exists for a reason: tariff headwinds are real, ongoing, and quantified at C$350 million in revenue impact for 2025 alone; resolution of US-Canada trade relations is the specific catalyst that separates an interesting situation from a compelling one.

---

North American freight has spent the better part of two years working through a hangover. The pandemic-era boom in goods demand — and the supply chain chaos that made every mile of rail and truck capacity precious — gave way in 2023 and 2024 to a protracted freight recession. Volumes across every mode fell back toward trend. Intermodal units declined. Carload revenues weakened. Railroads that had guided to double-digit earnings growth in 2022 were quietly withdrawing their long-range outlooks by 2024. Into that normalization came a second disruption: the United States decision to impose broad tariffs on Canadian goods, eventually reaching 35% on various categories and 50% on steel and aluminum, created genuine trade uncertainty across the cross-border corridors that Class I railroads depend on. For CN, which crosses the US-Canada border more extensively than any competitor, the combination of freight-cycle softness and policy-driven volume loss produced a 2024 where earnings fell nearly 20% from 2023's levels, the operating ratio deteriorated by 2.6 percentage points, and management withdrew its multi-year financial outlook entirely.

The industrial logic of freight rail as a business has not changed. A freight railroad is a toll road on which the toll is set by the economics of the alternative, and the alternative — diesel trucking — has been structurally disadvantaged by fuel costs, driver shortages, and the physics of moving large quantities of bulk commodity over long distances. Rail consumes roughly one-third the fuel of trucking per ton-mile, a gap that widens on corridors above 500 miles and that no improvement in truck technology has materially narrowed in three decades. The structural argument for shifting freight from truck to rail — particularly in intermodal, where standard containers move by train between terminals and by truck for the first and last mile — remains intact. Rail currently handles approximately 28% of North American freight by ton-miles against trucking's 40%, and in the medium-haul corridors of 750 to 1,500 miles, rail captures only 27% of addressable intermodal freight. Every percentage point of conversion in that band represents meaningful incremental volume for the industry.

The Class I railroad industry is a textbook oligopoly. Six carriers — Union Pacific, BNSF, Canadian National, Canadian Pacific Kansas City, CSX, and Norfolk Southern — operate essentially every mile of long-haul freight rail in North America. Collectively they have invested more than $840 billion in their own infrastructure since 1980, an amount that renders new competitive entry impossible under any plausible economic or regulatory scenario. The Surface Transportation Board in the US and the Canadian Transportation Agency have not approved the construction of a competing main line since the late nineteenth century. Rail networks are not disrupted by faster-moving entrants because the capital requirement is not millions of dollars or even billions — it is hundreds of billions, deployed over decades, across lands that would require eminent domain proceedings in two countries. The correct analogy is not a technology business; it is a river. Someone built the river a century ago, and no one is going to build another one.

Within this oligopoly, Canadian National occupies a structurally unique position. Its network forms a Y-shape connecting three coastlines: the Pacific at Vancouver and Prince Rupert in British Columbia; the Atlantic at Halifax, Nova Scotia; and the Gulf of Mexico at New Orleans, Louisiana. This configuration — completed through the 1998 acquisition of Illinois Central Railroad, which added the north-south corridor from Chicago to New Orleans, and the 2001 acquisition of Wisconsin Central, which filled in the Great Lakes linkage — is the only truly transcontinental routing available on a single railway. A shipper moving goods from the Canadian prairies to a Gulf refinery, or an importer receiving Asian container cargo at Prince Rupert destined for Chicago, cannot access these movements on any other single carrier. The network's geographic reach means CN serves all three of North America's major port clusters — the Pacific gateway, the Atlantic gateway, and the Gulf petrochemical complex — without requiring interline handoffs to competitors. This is not brand loyalty or switching cost in the conventional sense. It is physical geography.

The competitive moat claim requires more than geographic description. The test is whether CN's position translates into measurable operating superiority over time — and here the operating ratio is the clearest instrument. A railroad's operating ratio measures operating expenses as a percentage of revenue; at 60%, the railroad keeps forty cents of every revenue dollar as operating income. CN pioneered Precision Scheduled Railroading at Illinois Central under Hunter Harrison in the 1990s, brought PSR to CN when it acquired that road in 1998, and has operated for nearly three decades under a discipline that competitors spent years attempting to replicate. The comparison across the Class I peer group as of 2025 shows CN's structural operational advantage:

Railroad	2025 Operating Ratio	2024 ROIC	Network Position
Union Pacific	~59%	16.3%	US West, 32,000 route miles
Canadian National	61.7%	12.9%	Three coasts, 20,000 route miles
CPKC	~64%	~10%	Three countries (Canada-US-Mexico)
CSX	~64%	~10%	US East, 21,000 route miles
Norfolk Southern	~66%	~11%	US Appalachian/East, 19,000 route miles

CN is the second-most efficient Class I railroad in North America by operating ratio, trailing only Union Pacific, and it has held that position through multiple freight cycles. The gap between CN and the median Class I competitor represents roughly 4 to 5 points of operating ratio — on C$17 billion in revenue, that differential is worth approximately C$700 to C$850 million in additional annual operating income compared to a hypothetical average-efficiency competitor running CN's network. The PSR methodology — scheduling trains like airlines rather than accumulating cars until a train is full, minimizing terminal dwell time, maximizing asset velocity — is now practiced by all Class I carriers, but CN's three-decade head start shows up in operational metrics that remain measurably better than peers: car velocity of approximately 212 miles per day in 2025, average train dwell times that industry analyses consistently show at the low end of the Class I distribution, and a fuel efficiency record that in Q1 2026 hit its best quarterly level ever at 0.892 gallons per 1,000 gross ton-miles.

The one competitive development that does alter CN's long-term landscape in a meaningful way is the 2023 completion of the CPKC merger. Canadian Pacific's acquisition of Kansas City Southern created the first single-line railway connecting Canada, the United States, and Mexico. CN had aggressively pursued KCS itself in 2021, offering US$33.7 billion — outbidding CP by US$5 billion — before the Surface Transportation Board rejected CN's voting trust structure on competition grounds. The failure of that bid was a capital allocation setback that briefly consumed management attention; the subsequent completion of CP-KCS under different regulatory treatment created a competitor with a genuinely differentiated routing option for cross-border Mexico-Canada traffic that CN cannot currently match with single-line service. CN's response has been to partner with Union Pacific and Ferromex to launch the Falcon Premium intermodal service connecting Canada through the US interior to Mexico, effectively assembling a three-country option through interline agreements rather than owned infrastructure. Whether that partnership proves as commercially effective as CPKC's owned network remains to be demonstrated over time.

The financial profile of CN is the record of a business that earns well but does not compound at exceptional rates. Revenue has been essentially flat for four years: C$17.1 billion in 2022, C$16.8 billion in 2023, C$17.0 billion in 2024, and C$17.3 billion in 2025. The flat top line reflects two overlapping dynamics — post-pandemic volume normalization and deliberate tariff-driven diversion of Canadian freight to US ports by shippers uncertain about cross-border service continuity. Against that revenue backdrop, the earnings swings have been driven almost entirely by cost events rather than business quality deterioration. The 2024 earnings decline was anatomized by CN's own disclosure: the August 2024 lockout of 9,300 CN and CPKC employees, resolved after two days by federal arbitration order, had a pre-tax impact of approximately C$0.20 per diluted share in the third quarter. Cascading port labor disputes at Prince Rupert, Vancouver, and Montreal throughout 2024 diverted international container volumes to US Pacific ports before the labor disputes even began. Weather in the fourth quarter added operating leverage headwinds. Strip out those events and the underlying operating margin was modestly better than the headline 63.4% operating ratio suggested.

The 2025 recovery confirms that 2024 was an event-driven trough rather than structural deterioration. Operating ratio improved to 61.7%, approaching the 60–61% range that represents CN's normalized operating efficiency. GAAP net income reached C$4.7 billion, GAAP diluted EPS of C$7.57 was a record, and free cash flow recovered to approximately C$3.3 billion from 2024's C$2.3 billion. The Q1 2026 results introduced new caution: revenue fell 1% year-over-year, adjusted EPS declined 3% against the prior-year Q1, and the stock fell 6% on the report — its largest single-day decline since December 2021. Management quantified the 2025 tariff impact at C$350 million in lost revenue, attributed primarily to weakness in forest products, automotive parts, and cross-border intermodal. For 2026, CN cut its capital expenditure budget by C$500 million to C$2.8 billion and announced workforce reductions to align costs with reduced volume expectations. The dividend increase of 3% — the 29th consecutive annual increase — was announced alongside those cuts: a deliberate signal that management views the revenue headwinds as cyclical rather than structural. Whether that view is correct is the central question for the investment.

On GAAP versus adjusted figures: the primary reconciling item between CN's reported GAAP results and adjusted results is depreciation on acquired assets, charges related to labor disruption costs, and certain restructuring items. None of the adjustments reflect business model problems or obscure structural deterioration — they are genuine non-recurring items in a business with a long history of transparent reporting. The more important accounting note is what ROIC actually measures for a railroad. CN's return on invested capital of 12.9% in 2024 and approximately 13% in 2025 is calculated against approximately C$39–40 billion in invested capital, most of which is physical track, locomotives, and rolling stock carried at depreciated historical cost. If CN's network were to be valued at replacement cost — the amount it would cost to build it today — that cost would be several multiples of the book value, possibly C$200 billion or more. The 12–13% ROIC on historical cost becomes a much lower return on replacement cost. This is the correct way to think about why new entry does not occur: the required capital to replicate CN's network would earn inadequate returns on investment even at CN's efficiency levels, because the replacement cost is so enormous that the economics do not work. CN earns a reasonable but not spectacular return on its cost base; it earns a very high return relative to what it would cost a competitor to displace it. The moat is real but the returns do not compound like a software business — they compound like the infrastructure they are.

Tracy Robinson became CN's President and CEO in February 2022, the first female CEO of a Canadian Class I railroad, following 27 years at Canadian Pacific and seven years leading TC Energy's Canadian natural gas pipeline business. Her capital allocation record through three and a half years in the role reflects consistent priorities: approximately C$2 billion annually in share repurchases, 29 consecutive annual dividend increases, and disciplined maintenance capital investment in the C$2.8 to C$3.4 billion range. In November 2024, at the depths of the CN earnings anxiety, Robinson personally purchased 3,300 shares at C$148 per share — a symbolic but real signal. The KCS episode predates Robinson's tenure; the capital she has deployed has been returned to shareholders rather than allocated toward transformative acquisitions. The 2026 capex cut, while a meaningful reduction in near-term infrastructure investment, reflects the right instinct for a business navigating trade policy uncertainty: preserve financial flexibility, maintain the dividend, buy back shares at a discount to normalized value.

The share count reduction since the PSR era accelerated materially: CN has repurchased approximately 15 million shares in 2025 (approximately C$2 billion) and another 6 million in Q1 2026 alone (C$869 million), leaving the share count at approximately 610 to 616 million versus over 640 million three years ago. The dividend record — 29 consecutive years of increases, with the 2026 increase of 3% coming even as management cut capex and reduced the workforce — reflects a company that has internalized the idea that shareholder returns are not optional in good years and discretionary in bad ones.

The growth runway table for CN requires selecting variables that are leading indicators for this specific business rather than generic financial metrics. The right questions to ask of a mature freight railroad are: Is actual freight volume moving in the right direction? Is operational efficiency holding or improving? Is the cash generation reliable? And is the intermodal segment — the primary secular growth driver — recovering after the 2024 disruptions?

Year	Revenue (C$B)	Operating Ratio	Free Cash Flow (C$B)	RTMs (Billions)
2021	14.5	61.9%	~3.0	233
2022	17.1	60.0%	3.0	235
2023	16.8	60.8%	2.9	233
2024	17.0	63.4%	2.3	236
2025	17.3	61.7%	~3.3	238

The table tells a story of fundamental stability rather than growth. Revenue ton-miles — the actual measure of freight moved — grew from 233 billion in 2021 to 238 billion in 2025, a 2% increase over four years. The revenue line jumped in 2022 when freight rates spiked post-pandemic, then held approximately flat despite the absence of rate tailwinds. Operating ratio was at 60.0% in 2022, the best in CN's recorded history, before widening to 63.4% in 2024 on the disruption events and recovering to 61.7% in 2025. Free cash flow compressed to C$2.3 billion in 2024 and recovered to approximately C$3.3 billion in 2025 — demonstrating that the 2024 decline was event-driven, not structural. This is not a business growing at 8–10% per year; it is a business growing at 1–2% volume per year with pricing that tracks inflation and occasional commodity mix shifts.

The case for a more compelling growth runway rests on two specific pillars that are both real and time-delayed. The first is the Prince Rupert gateway expansion. Prince Rupert, British Columbia sits closer to Asian ports than any other major North American harbor — approximately two days closer to Shanghai than Vancouver — making it the fastest Asia-Pacific container gateway on the continent. CN holds the exclusive railway franchise serving the port. Prince Rupert moved 23.1 million tonnes in 2024 and 26.3 million tonnes in 2025, a 14% year-over-year gain. The port is capacity-constrained. A critical infrastructure bottleneck — the Zanardi Rapids rail bridge — currently limits train movements to approximately 24 per day. CN is constructing a new 1,600-foot replacement bridge, with completion expected in 2027, that will remove this constraint and enable materially higher daily train counts. When that infrastructure comes online, Prince Rupert could handle 30 million-plus tonnes annually, with CN capturing the economics of every container moving through it. This is a defined capital investment with a defined completion date and a defined volume expansion, not a speculative market share argument.

The second pillar is intermodal conversion. CN's intermodal segment — containers moving by rail between terminals — declined in 2024 amid labor uncertainty and then recovered 6% in 2025. Rail currently captures 27% of medium-haul freight (750 to 1,500 miles) that could move in intermodal. Moving that capture rate to 30% across the industry would represent approximately C$1.7 billion in additional annual revenue industry-wide, with CN's share proportional to its network access. CN has captured approximately 14% of Class I rail ton-miles, a position that has held essentially flat across the measurement period. The untapped intermodal opportunity in medium-haul corridors — where rail economics work but shipper relationships favor established truck carriers — is the structural growth case. It is real but materializes over years, not quarters. Against that medium-term backdrop, the near-term tariff-driven volume loss in forest products, automotive, and cross-border intermodal is the dominant consideration for 2026 earnings.

At a current price of C$149.87 on the TSX and USD $109.38 on the NYSE, CN trades at a market capitalization of approximately C$91 billion. Adding net debt of approximately C$20.5 billion (at 2.6 times adjusted EBITDA, well within the company's target range) produces an enterprise value of roughly C$112 billion, or approximately 11.6 times trailing EBITDA. The trailing twelve-month P/E ratio of approximately 20 times is exactly at CN's 10-year median multiple of 20.04 times. On a forward basis, with 2026 guidance implying earnings that are roughly flat to modestly below 2025's record due to tariff headwinds, the forward P/E contracts to approximately 16 times — a meaningful discount to the 20-times historical norm. The stock's 6% decline on Q1 2026 results reflected the market absorbing the news that 2026 will not be a repeat of 2025's recovery year.

This is not an exceptional compounder in the sense of a business earning 25% or more on unleveraged net tangible assets. CN earns approximately 13% return on its C$39 to C$40 billion invested capital base — a return that falls in the "Good" rather than "Great" category. For context, Union Pacific earns 16% ROIC on a network half again as large. The difference is partly structural: CN's three-coast routing gives it geographic breadth but involves long corridors through lower-density territory, while Union Pacific's Western US franchise connects dense industrial markets. CN's ROIC has ranged from approximately 13% to 17% over the past five years, with 2024's 12.9% representing a trough driven by identifiable event-based disruptions rather than competitive displacement or secular decline. The honest assessment is that a business earning 13% on its invested capital, with 1–2% annual volume growth, priced at 20 times trailing earnings, offers a reasonable but not exceptional prospective return. The question is whether the tariff headwinds are temporary — in which case 2025's C$7.57 GAAP earnings are closer to normalized than 2026's likely lower figure — or persistent, in which case the forward discount is not a discount at all but a fair assessment of structurally reduced earnings power.

The intelligent bear argues that CN's tariff problem is not temporary: US-Canada trade relations have deteriorated structurally, forest products and automotive supply chains are being reorganized away from cross-border flows, and a business that earns 13% ROIC at peak efficiency does not deserve a premium to fair value even at 16 times earnings. This argument has real force. The tariff regime imposed in 2025 was not a brief bargaining chip; it has persisted and deepened, and CN's own management has responded by cutting capital investment and reducing headcount — hardly the posture of a leadership team expecting near-term reversal. The counter-argument is that Canadian-US trade has been disrupted multiple times historically — through NAFTA renegotiation, lumber disputes, softwood tariff cycles — and has repeatedly normalized because the economic logic of integrated North American production chains is compelling regardless of political posture. The Falcon Premium service CN assembled with UP and Ferromex demonstrates that CN has not surrendered the Mexico corridor but is actively building alternative commercial pathways. The question is not whether CN has a permanent moat — it demonstrably does — but whether the catalyst for earnings recovery is six months away or three years away.

The conclusion that follows from the evidence is this: CN is a high-quality infrastructure business trading at approximately fair value on trailing earnings and at a modest discount on a forward basis, with near-term earnings held below normalized levels by trade policy headwinds that are political rather than structural in character. The Zanardi Rapids bridge is coming. Intermodal will recover when shipper confidence stabilizes. The three-coast network is not going anywhere. But the investor buying at C$150 today is not getting the network for free — they are paying roughly the historical average multiple for a business that will likely earn somewhat less in 2026 than it earned in 2025. The specific catalyst that makes this more than an acceptable-return infrastructure holding is evidence that US-Canada trade tension is reducing rather than compounding — that tariff rates are declining, that forest products and automotive volumes are recovering, that CN's 2026 intermodal guidance can be revised upward. Without that evidence, the investment is interesting. With it, it becomes compelling.

At 16 times normalized earnings, the tollbooth is available at a discount. The question is how long one is willing to wait for political conditions to normalize around a physical asset that has no expiration date.