Cummins Inc.(CMI) · Diversified Industrials

Cummins Long-Term Value Investment Analysis

Other languages
Quick ReadPlain-language overview · read this first

Cummins is a platform industrial company built around powertrains, critical components, aftermarket service, and power generation systems, operating across five segments — Engine, Components, Distribution, Power Systems, and Accelera — with 2025 revenue of $33.67 billion. Its roughly 80-year partnership with PACCAR (a single customer at 13% of sales), together with 640 distribution nodes and more than 13,000 certified dealer points, forms a real but moderate service moat. Rating: Watcha good company, but the current price already pays up-front for multiple tailwinds.

The core tension lies in the valuation premise, not the business itself. At the current price of $639.55, the P/FCF is about 37x and conservative owner earnings about 32–36x, with a starting earnings yield of only just over 3% — actually below the 10-year Treasury's 4.56%. Discounting owner earnings across three scenarios gives $220–280 (conservative), $350–430 (base), and $500–590 (optimistic) per share, and the current price has already reached or even crossed the upper edge of the optimistic scenario. The market is not treating it as a cheap industrial stock in a down cycle, but paying at once for three things: the data-center backup-power boom, the North American truck recovery, and the narrowing of new-energy losses.

The three most fragile assumptions are exactly these three things — a $458 million electrolyzer-business charge in 2025 and a further $199 million loss from selling the low-pressure fuel-cell business in the first quarter of 2026 make the new-energy trial-and-error cost clear; add the receivables sale being classified into CFO, and the cash-flow measure is less pure than it appears. If data centers cool, the truck recovery is weak, Accelera impairs again, and the valuation reverts to the industrial-stock average — all four combined — a return of the share price to $190–270 would imply a permanent capital loss of about 58%–70%. With an ideal buy range of $280–350, the current price is better suited to a watch list.

Lead

A global platform spanning heavy-duty engines, components, and power generation systems, with 2025 revenue of $33.67 billion and robust demand for data-center backup power; at the current price of $639.55 and roughly 37x P/FCF, the stock has already priced in multiple tailwinds and offers an insufficient margin of safety. Rating: Watch.

Full report

Prices in the article are as of publication; see the valuation band above for the live price.

Bottom Line Up Front

Preliminary Conclusion

If I evaluate Cummins by the standard of "acquiring a business to own for the long term," my conclusion is this: this is an industrial company you can broadly understand and that does carry genuine quality, but at today's price it is not cheap and the margin of safety is insufficient. My rating is: Watch. Cummins's core strengths are real: it has a deep installed base, a global distribution network, and strong engineering capability across heavy-duty engines, critical components, aftermarket service, and power generation systems. When the North American truck cycle turns down, distribution and power systems can still support earnings, and demand for data-center backup power in particular has been very strong over the past several quarters. The problems are just as real: the business is inherently strongly cyclical, and profits and cash flow in recent years have been heavily distorted by large one-off items such as the emissions settlement, the Atmus spin-off gain, and Accelera's restructuring and impairments. Yet the price the capital markets now assign it looks more like paying up-front for "a durable data-center tailwind + a truck-cycle recovery + a narrowing of new-energy losses" all at once. For an investor with a "balanced-to-conservative" risk appetite and a holding horizon of more than ten years, I would rather treat it as a good company whose current price simply isn't friendly enough.

Summary of Conclusions

To unpack that judgment: the investment rating is "Watch." The core judgment is that the business is understandable, quality is above average, the moat exists but is not "unassailable," management is broadly rational though its capital allocation is not top-tier, and the current valuation lacks appeal for conservative long-term capital. Is there a margin of safety at the current price? No, or at least not an obvious one. The type of investor it suits is the long-term value investor who can tolerate the industrial cycle and is willing to wait through sizable drawdowns; it is less suited to anyone buying it as a "steady compounding utility." The biggest uncertainties are three: how long the data-center backup-power boom can last; the strength of the North American medium- and heavy-duty truck-cycle recovery; and whether Accelera's transition losses can genuinely narrow.

The core basis for the above judgment: as of May 22, 2026, CMI traded at about $639.55, with a market cap of roughly $88.77 billion; while in 2025 the company posted revenue of $33.67 billion, net income attributable to the parent of $2.843 billion, operating cash flow of $3.621 billion, capital expenditure of $1.235 billion, and GAAP EBITDA of $5.385 billion. Estimating from 2025 actual free cash flow, the current price corresponds to a P/FCF of around 37x; even under a more lenient "conservative owner earnings" measure, the current price sits at roughly 32 to 36 times owner earnings, which is not generous for a cyclical heavy-industrial company.

Understanding the Business and Industry Landscape

How Does This Company Actually Make Money

【Fact】 Cummins currently has five reporting segments: Engine, Components, Distribution, Power Systems, and Accelera. The company's disclosed 2025 external revenue for each was roughly: Engine $8.104 billion, Components $8.643 billion, Distribution $12.386 billion, Power Systems $4.114 billion, and Accelera $423 million. Together these businesses cover diesel and natural-gas engines, aftertreatment systems, turbochargers, fuel systems, axles and brakes, transmission systems, generator sets, spare-parts service, and zero-emission technologies such as batteries, electric drives, and fuel cells.

【Inference】 From a long-term owner's perspective, Cummins is not a "single-engine company" but is closer to a platform industrial company built around powertrains + critical components + aftermarket + power generation systems. What truly supports its resilience through the cycle is not just selling new engines, but a three-layer revenue structure: the first layer is OEM-fitted new units and components; the second is the globally distributed parts, service, maintenance, and remanufacturing; the third is power generation systems, which have relatively low correlation to the diesel cycle and have been extremely strong over the past two years, especially data-center backup power. For this reason, Cummins's "understandability" is higher than that of the average industrial conglomerate, but lower than a consumer-goods or pure-software business.

【Fact】 Looking at the revenue mix of the distribution business, within the 2025 Distribution segment: Parts $4.073 billion, Service $1.798 billion, Power generation $4.928 billion, Engines $1.587 billion. In other words, within Distribution alone, parts and service together already reach $5.871 billion, about 47% of that segment's revenue; and if the distribution segment's power generation is included, the share that leans "more toward aftermarket or service capability" within Distribution is even higher. Within the Power Systems segment's 2025 external revenue, Power generation was $2.229 billion, Industrial $1.238 billion, and Generator technologies $647 million.

【Opinion】 This means that if you see Cummins only as a "heavy-truck engine cyclical," you will underestimate its quality; but if you see it as a "stable, predictable compounding machine almost immune to the cycle," you will overestimate its stability. What it most resembles is: a heavy-industrial platform with a fairly deep service moat, with profit quality above the typical OEM and volatility above that of a true consumer monopoly or asset-light platform.

Customers, Pricing Model, Recurrence, and Dependence

【Fact】 Cummins's customers include truck and commercial-vehicle OEMs, construction-machinery OEMs, mining, rail, marine, power generation and data-center customers, dealers, and end-repair customers. The company explicitly discloses that PACCAR is its largest customer, accounting for 13% of consolidated net sales in 2025; in 2024, PACCAR's share was 16%, and it has long-term supply agreements with customers such as PACCAR, Traton, Daimler, and Stellantis. Cummins also discloses that in 2024 its four largest customers, including these four, together accounted for 36% of revenue; on this same basis the 2021 figure was 33%.

【Inference】 This shows that the customer structure is not diversified to the point of "no concentration risk at all." A single customer like PACCAR at 13% is already worth taking seriously for an industrial company; but it is also not a single-dependency company that would be "crippled by losing one customer." More importantly, the supply relationships are built on decades of product validation, regulatory certification, after-sales service, and platform fit, so switching is not easy.

【Fact】 At the end of 2025 the company served roughly 190 countries and regions, with about 640 company-owned, joint-venture, and independent distribution locations and more than 13,000 Cummins-certified dealer points.

【Opinion】 The pricing model is very traditional and very clear: sell engines, sell components, sell power generation systems, sell service, sell maintenance and support. This is an old-school, unglamorous, but verifiable industrial business. You don't need to believe a fuzzy TAM story, and you don't need to understand a complex advertising algorithm; you only need to understand two things: whether the installed base will keep expanding, and whether the aftermarket value of already-installed equipment can continue to be realized. For Cummins, the answers to both remain modestly positive for now.

Industry Stage, Competitive Landscape, and Whether It's Worth Holding Long-Term

【Fact】 This industry as a whole belongs to the cyclical sub-industries within a mature industry. Demand for heavy-duty commercial vehicles, construction equipment, industrial power, and distributed power exists over the long run, but is affected by the macro economy, freight rates, capital expenditure, emissions regulations, and shifts in technology pathways. In 2025, Cummins was under pressure in the North American medium- and heavy-duty truck market, but Distribution and Power Systems achieved record sales and profit, driven mainly by strong power-generation demand, especially data-center backup power; by the first quarter of 2026, the company further raised its full-year revenue growth guidance to up 8% to 11% year over year and its EBITDA guidance to 17.75% to 18.50%, one core reason being the improving North American on-highway market and strong power demand.

【Fact】 At the same time, the industry is not static. In 2025 the company conducted a strategic review of the electrolyzer business under Accelera and recognized $458 million in charges related to that business for the full year, of which $415 million was non-cash; the company also said it planned to stop new commercial activity in the electrolyzer business while continuing to honor existing customer commitments. By the first quarter of 2026, the company had completed the sale of its low-pressure fuel-cell business, recognizing $199 million in net charges as a result.

【Inference】 This industry is therefore not one that "will see no technological disruption over the next decade." Diesel will not disappear tomorrow, but it will not be the only answer forever either. Cummins's advantage is that it has not locked itself into a single technology path — the HELMTM fuel-agnostic platform, natural-gas engines, hybrid mining conversions, batteries, and electric drives are all part of its plans; its difficulty is that every transition path costs money, and not every road can turn a profit. The pullback in the electrolyzer and part of the fuel-cell business already shows that "positioning for the transition" does not equal "succeeding at the transition."

【Opinion】 If the stock market were closed for five years, I would be willing to hold the Cummins business, but only if the purchase price is reasonable. Because it is not the kind of absolute top-quality asset where "even if you overpay a bit today, compounding over the next decade will most likely bail you out"; it is more of a "good business + cyclical swings + occasional special items + a need for price discipline" industrial leader. Business understandability score: 4/5. Industry attractiveness score: 3/5.

Moat and Management

Does the Moat Actually Exist

Cummins's moat is not an internet-style network effect but a more traditional combination of industrial moats.

Brand and reliability. In heavy-duty power systems, brand is not advertising but the accumulated perception among grassroots fleets, equipment owners, and OEMs of "reliability, fuel economy, ease of maintenance, and residual value." Cummins's relationship with PACCAR has lasted about 80 years, and it has long-term supply arrangements with Traton, Daimler, and Stellantis — something an ordinary new entrant cannot replicate through marketing alone.

Scale and channel. About 640 distribution nodes + more than 13,000 certified dealer points, layered on top of a global installed base, give Cummins real advantages in service capability, parts supply, regional coverage, and fault response. This kind of network is not a classic "network effect," but it can significantly lower customers' downtime costs and raise aftermarket stickiness.

Switching costs. For end customers, the engine, aftertreatment, control system, service network, repair technicians, parts supply, and regulatory certification form one integrated system; for OEMs, switching product platforms involves engineering validation, calibration, emissions certification, and rebuilding the after-sales system. Switching costs are real, but not insurmountable, so I define this as a moderate moat, not an impregnable one.

Operating capability and culture. In 2025, with the North American truck cycle turning down, the company still delivered strong profit through Distribution and Power Systems, and in the first quarter of 2026 Power Systems again drove a segment EBITDA margin of 29.5%, showing that operational execution is genuinely not weak.

Moats it lacks. Cummins has almost no network effect or data-monopoly advantage worth singling out, and it is not a company that monopolizes its industry through regulatory licenses. Patents and emissions certifications do form a barrier, but they are mainly an entry barrier rather than a source of permanent monopoly.

My judgment: Cummins's moat is "a stable-to-narrowing traditional internal-combustion-engine moat + an expanding service/power-generation moat." In other words, the moat around the old business has not vanished, but it faces technological substitution and regulatory pressure; at the same time, the competitive position of the aftermarket and power generation systems is becoming more important. Moat strength score: 3.5/5.

Is Management Trustworthy, and Is Capital Allocation Rational

【Fact】 Jennifer Rumsey has served as CEO since August 2022 and as chair since August 2023. The 2026 proxy statement shows that management's pay system still emphasizes performance: the annual bonus uses EBITDA at 70% + operating cash flow at 30% as its core metrics; long-term incentives are 100% performance-based, and within the domestic plan they consist of 70% performance shares + 30% performance cash, with ROIC weighted 80% and EBITDA weighted 20% over the three-year assessment. The 2025 say-on-pay vote received about 91% support.

【Fact】 The 2026 proxy statement also shows that Rumsey's beneficial ownership was about 119,419 shares, and all directors and current executive officers together held about 415,453 shares, in each case less than 1% of the total. This shows that management is not "almost entirely unaligned" with shareholders, but it is by no means a founder-style, very-high-ownership structure either.

【Fact】 On capital returns, in 2025 the company raised its annual dividend to $7.64 per share, marking 16 consecutive years of dividend increases; over full-year 2025 it returned about $1.055 billion to shareholders through dividends alone. In the first quarter of 2026, the company said it returned a total of $519 million to shareholders, and continued small-scale buybacks under the $2 billion repurchase authorization approved in 2021; in 2024, the Atmus split-off exchange offer further reduced Cummins's shares outstanding by about 5.6 million shares.

【Opinion】 My assessment of management is: honesty and operational rationality above average, but capital allocation not yet excellent. There are three things I would give it "credit" for. First, management does not shy away from bad news: whether the large electrolyzer-business impairment in 2025 or the sale of the low-pressure fuel-cell business in the first quarter of 2026, both were clearly disclosed. Second, the pay metrics put operating cash flow and ROIC at the core rather than simply chasing EPS. Third, the dividend policy is consistent and steady, without overdrawing the balance sheet for short-term returns.

There are also three things I would "dock" it for. First, between 2023 and 2025, Accelera went through repeated restructurings, impairments, and shifts in direction, indicating there were exploration costs in its new-energy capital deployment — at the very least, resource allocation was not brilliant all the way through. Second, the emissions settlement disclosed in 2023 and effective in 2024 is a major governance and compliance stain: the company accrued about $2 billion for the matter in the fourth quarter of 2023 and paid $1.9 billion in settlement-related cash in the second quarter of 2024. Third, in 2024 the company established a receivables-sale arrangement and included the related collections in operating cash flow — this does not amount to accounting manipulation, but it does make the cash-flow measure somewhat less "clean."

Management and capital-allocation score: 3.5/5. I will not dismiss Cummins over these issues, but I also will not place it in the "management is nearly Buffett-like" category.

Financial Quality and Owner Earnings

How Has Financial Quality Been in Recent Years

The table below organizes the core metrics from 2020 to 2025 on the basis a "long-term owner cares most about." The operating margin and free cash flow in the table are calculated by me from the company's disclosed statements; among them, the 2023 and 2024 GAAP profits were significantly distorted by items such as the emissions settlement, the Atmus spin-off, and Accelera's restructuring, so they must be viewed together with cash flow.

Year Revenue Net Income (to parent) Operating Cash Flow CapEx Free Cash Flow Diluted EPS Gross Margin Operating Margin
2025 33.67 2.84 3.62 1.24 2.39 20.50 25.3% 12.0%
2024 34.10 3.95 1.49 1.21 0.28 28.37 24.7% 11.0%
2023 34.07 0.74 3.97 1.21 2.75 5.15 24.2% 5.2%
2022 28.07 2.15 1.96 0.92 1.05 15.12 23.9% 10.4%
2021 24.02 2.13 2.26 0.73 1.52 14.61 23.7% 11.3%
2020 19.81 1.79 2.72 0.53 2.19 12.01 24.7% 11.5%

Three conclusions after reading the table.

First, revenue growth is real, but it is not linear, high-quality compounding. From 2020 to 2025 revenue rose from $19.81 billion to $33.67 billion, decent growth in absolute scale; but revenue was nearly flat from 2023 to 2025, indicating the company expanded more through business-mix optimization and cyclical offsetting than through a straight-line rise in unit volume. The quality of growth depends on product mix, not pure volume.

Second, the margin center is not bad, but it gets heavily contaminated by special items. In 2025 gross margin and operating margin were about 25.3% and 12.0%; operating margins in 2021, 2022, and 2020 were also all in the 10% to just over 11% range. The real large swings appeared in 2023 and 2024: 2023 GAAP profit was significantly depressed by the emissions-settlement accrual, while 2024 was significantly lifted by the Atmus spin-off gain. So if you look only at the 2024 P/E, you will overestimate earnings power; if you look only at the 2023 P/E, you will underestimate it.

Third, cash flow is broadly real, but distorted in some years, with a measurement caveat. In 2025 operating cash flow was $3.621 billion and CapEx $1.235 billion; in 2023 operating cash flow of $3.966 billion was actually far higher than that low-base year's profit, indicating that year's low profit contained a large amount of non-cash factors. Conversely, 2024 operating cash flow was only $1.487 billion, mainly affected by the settlement payment and working capital. The company also introduced a receivables-sale arrangement in 2024 and classified the related cash flow as operating cash flow, so when looking at Cummins's CFO one should keep a stance "a bit more conservative than the surface suggests."

Balance Sheet, Leverage, and Cash-Flow Quality

【Fact】 At the end of 2025 the company had cash and cash equivalents of about $2.845 billion, long-term debt of about $6.792 billion, and total equity of about $13.408 billion; 2025 interest expense was about $329 million and operating profit about $4.025 billion, so on this simple basis operating profit / interest expense was about 12.2x. This multiple was also above 10x in 2024. This is not a fragile balance sheet.

【Inference】 On a rough 2025 basis, Cummins's net debt / EBITDA is roughly below 1x, clearly within an acceptable range. It is not a highly leveraged company betting on the cycle. Even after paying the large emissions-settlement cash in 2024, liquidity showed no danger signals. The key restriction disclosed in the company's debt covenants is that net debt / total capital not exceed 0.65x, and it was in compliance at the end of 2022; in the materials retrieved, there is no sign of recent default or liquidity strain.

【Fact】 On the share-count side, the diluted weighted-average share count fell from about 149 million shares in 2020 to about 138.7 million shares in 2025, a decline of about 7%. This shows per-share value has not been continuously diluted but has instead slowly improved; but the share decline does not come entirely from secondary-market buybacks — it also includes the share reduction from the 2024 Atmus split-off exchange.

【Opinion】 I have not seen evidence sufficient to support "financial fraud" or "aggressive accounting manipulation of profit," but I would put the following three items on the ongoing-watch list: first, the impact of the receivables-sale program on the comparability of operating cash flow; second, whether new-energy-related impairments and exits will continue; third, whether major compliance fines or litigation cash outflows will occur again in the future. In other words, I do not think it is a company with financial risk, but I also do not think it is a completely noise-free, clean set of statements.

Owner Earnings Analysis

For this kind of manufacturing company, I lean toward a conservative version of owner earnings rather than mechanically equating free cash flow with owner earnings.

【Fact】 In 2025, Cummins's net income attributable to the parent was about $2.843 billion; depreciation and amortization about $1.093 billion; operating cash flow about $3.621 billion; and total CapEx about $1.235 billion.

【Assumption】 Maintenance CapEx is hard to read directly from the statements. Given that Cummins is a heavy manufacturer while simultaneously working on new platforms, plants, and product transitions, I adopt a conservative assumption: treat 85% to 90% of 2025 total CapEx as maintenance CapEx, i.e., roughly $1.05 billion to $1.11 billion. The implication of this assumption is that I do not treat most of the investment as "high-return growth investment," but instead default to a substantial portion being merely what is necessary to maintain existing competitiveness. This assumption is not company-disclosed data but a prudent assumption within the valuation.

【Inference】 Under this assumption, 2025 conservative owner earnings can be approximated as: Operating cash flow $3.621 billion − maintenance CapEx $1.05 billion to $1.11 billion = $2.51 billion to $2.57 billion. Further considering that the 2024 settlement payment distorts the three-year average cash flow, while 2025 also carried some Accelera non-cash charges, a more reasonable conservative-to-neutral owner-earnings range can be placed at roughly $2.5 billion to $2.8 billion. This implies the current price corresponds to an equity-basis owner-earnings multiple of about 32 to 36 times. For an industrial stock, this multiple is not low.

【Opinion】 Can this company generate "real, distributable cash flow" over the long run? The answer is: yes, but not smoothly. I would rather see it as a company that "produces meaningful cash across the full cycle" than one that "steadily disgorges cash every year." Investing in this kind of company requires valuing it on cyclically adjusted, normalized earnings, rather than taking peak-cycle profit and simply multiplying it by a high multiple.

Intrinsic Value, Margin of Safety, and Valuation Comparison

Where Does the Current Valuation Stand

As of May 22, 2026, CMI traded at about $639.55, with a market cap of about $88.77 billion. Market-data pages show its trailing-twelve-month P/E at about 33.22x; Yahoo's public summary statistics also show its enterprise value / EBITDA at about 17.85x. Combined with year-end 2025 cash of $2.845 billion, long-term debt of $6.792 billion, and equity of $13.408 billion, it is clear that the price the market currently pays for this company is plainly not an "ordinary industrial-stock valuation" but carries a fairly strong quality premium and boom-cycle expectation.

By my estimate based on 2025 actual free cash flow of $2.386 billion, the current price corresponds to a P/FCF of about 37x; on shareholders' equity of $13.408 billion, the current P/B is about 6.6x. All of this shows: the market is not pricing Cummins as a "cheap manufacturing stock in a down cycle," but as an "industrial platform with higher certainty." If you agree this premium is entirely reasonable, then you can buy; if you require a sufficient margin of safety, then this price is not friendly.

Intrinsic Value Through Three Methods

Owner-Earnings Discount Method

Below is my three-scenario valuation based on conservative owner earnings. To be clear: this is not "forecasting the stock price," but "testing whether buying today is worthwhile."

Scenario Initial Owner Earnings First-Decade Growth Discount Rate Terminal Growth Implied Intrinsic Value
Conservative $2.5 billion 2% 10% 2% about $220–280/share
Base $2.7 billion 4.5% 9.5% 2.5% about $350–430/share
Optimistic $3.0 billion 7% 8.5% 3% about $500–590/share

【Assumption Notes】 The conservative scenario assumes: data-center power-generation demand returns to normal, the North American medium- and heavy-duty truck market sees only a mild recovery, and Accelera's losses narrow but keep consuming some cash. The base scenario assumes: Distribution and Power Systems maintain fairly strong performance, Engine/Components recover with the cycle, and new-energy losses fall markedly. The optimistic scenario assumes: the data-center boom lasts for years, the truck-cycle improvement continues, capital-expenditure returns improve, and Accelera gradually stops the bleeding. The starting points of these scenarios are all based on my estimate of normalized owner earnings for 2025 to 2026, rather than simply using GAAP net income. The inputs come from the company's 2025 cash flow, capital expenditure, and earnings, along with the first-quarter-2026 guidance raise.

【Conclusion】 Viewed through long-term owner-earnings discounting, the current stock price is broadly already near or even above the upper edge of my optimistic scenario. This means that if you buy today, you are not leaving room for "mean reversion" but paying a price for "good things continuing to happen." For a conservative investor, this is not an ideal starting point.

Relative Valuation Method

To avoid the illusion that "peers are all expensive, so it must be cheap," I use relative valuation only as a supplement, not as the main judgment.

Company Current Price / Market-Cap Basis P/E EV/EBITDA P/B Notes
Cummins Current 33.2x 17.9x about 6.6x Not cheap, especially relative to its cyclicality
Caterpillar Current about 35–44x about 29.6x about 18.9–21.7x Clearly more expensive, but the market grants it a "stronger capital-goods leader" premium
PACCAR Current 23.3x 17.8x 2.9x EV/EBITDA close to Cummins, but P/B clearly lower
Wabtec Current 36.3x 21.2x 4.0x Also not cheap, but its rail-equipment/service model differs substantially

The most important thing here is not the multiples down to the decimal, but three conclusions: First, Cummins is indeed not as expensive as CAT, but CAT's expensiveness comes with its "extremely strong capital-goods leader + high capital returns + stronger pricing power" backdrop. Second, Cummins's EV/EBITDA is already close to PACCAR's, which means it does not get a clear discount for its "engine/components being more cyclical." Third, on 2025 free cash flow, Cummins itself is at about 37x P/FCF, which for conservative value investing is clearly nowhere near cheap.

Asset or Liquidation Value Method

【Fact】 At the end of 2025, Cummins had total assets of about $33.992 billion, total equity of about $13.408 billion, cash and cash equivalents of about $2.845 billion, and net PP&E of about $6.958 billion.

【Opinion】 For a company like Cummins, the asset method is more of a "floor mindset" than the primary valuation method. The reason is simple: what is truly valuable about it — the installed base, global service network, customer relationships, brand, engineering capability, and OEM embeddedness — is not fully reflected in book net assets. Seen from another angle, this also means book assets can provide almost no protection for the current price. What the market grants it is an earnings-power premium, not an asset-revaluation premium. Comparing the current market cap of about $88.77 billion with equity of $13.41 billion, the market is clearly buying "future cash flows," not "liquidatable assets." The asset method does not support the case for buying today.

Is the Margin of Safety Enough

My judgment is very clear: not enough.

The three most fragile valuation assumptions are these. First, how long the high boom in data-center backup power lasts. Over the past several quarters it has greatly lifted the profit quality of Distribution and Power Systems, one of the core reasons the current price is willing to pay a premium. Second, the speed and magnitude of the North American medium- and heavy-duty truck market's rebound. If the recovery falls short of expectations, the profit recovery of Engine and Components will be slower than the market imagines. Third, whether Accelera's loss-narrowing is truly sustainable; if a new round of large impairments or exit costs appears in the future, normalized free cash flow will keep being eroded.

To put it more bluntly: buying Cummins today is more like accepting a "good company but bad price" payoff structure. If future growth comes in below expectations, margins pull back slightly, and the valuation multiple reverts from a high level to normal, investors could still end up with very mediocre or even negative long-term returns. For investors who emphasize control of permanent capital loss, this is not a comfortable starting point.

Based on the discounting and relative valuation above, I offer the following range judgments: Conservative intrinsic value range: $220–320/share. Fair intrinsic value range: $350–450/share. Optimistic intrinsic value range: $500–600/share. Ideal buy range: $280–350/share. Acceptable holding range: $350–500/share. Clearly overvalued range: above $550. The current price of about $639.55, in my view, is at least in the "on the expensive side" zone rather than the "has a margin" zone.

Risks, the Bear Case, and Alternative Opportunities

The Most Important Risks and the Strongest Bear Case

The most important risk is not short-term volatility but the following scenarios that could lead to overpaying + a reversal of the thesis.

Competition and technological-substitution risk. Cummins's traditional moat is mainly built around the internal-combustion engine and related components. If truck electrification, hydrogen, alternative powertrains, or OEM in-house development advance faster than expected, the durability of Cummins's existing high-margin business will be compressed. The company has already made strategic retreats in the electrolyzer and low-pressure fuel-cell businesses, showing that betting on new technology is not easy.

Cyclical risk. Engine and Components remain highly exposed to the North American medium- and heavy-duty truck market and parts of the industrial market. 2025 and the first quarter of 2026 clearly demonstrated: Power Systems being good does not mean the entire business is good. If data-center demand cools while the truck recovery is weaker than expected, the market's current imagination of the profit center will be revised down.

Customer-concentration and OEM-bargaining risk. PACCAR as a single customer accounted for 13% in 2025, and it has close relationships with large customers such as Traton, Daimler, and Stellantis. Long-term agreements do enhance visibility, but they also mean large OEMs retain strong bargaining power over price, quality, delivery, and platform switching. Cummins is not a consumer brand with full pricing power.

Governance and compliance risk. The emissions settlement effective in 2024 is a historical stain that cannot be glossed over. It brought not only a large cash outflow but also showed that even a century-old industrial company can incur enormous costs from a compliance failure. If a similar regulatory or recall event occurs in the future, it will do substantive damage to the long-term valuation.

Accounting and cash-flow-measure risk. The receivables-sale program being classified as operating cash flow lowers the "natural purity" of CFO. This does not mean the company has a problem, but that investors should be more conservative when judging "truly distributable cash flow."

Overvaluation risk. This is what I consider the most realistic risk right now. Today's investor is not buying a cheap cyclical stock but a significantly re-rated industrial platform. If the future is merely "no better" rather than "clearly worse," it could still produce poor returns over a long stretch.

The strongest bear case can be summarized in one sentence: "You have not bought, at a low, an industrial leader wrongly punished by the market; you have bought, at a high, three things at once — the data-center boom, the truck recovery, and the narrowing of transition losses." If this bear case is right, then the worst outcome is not necessarily the company's fundamentals collapsing, but that you bought at a high valuation point a business that is still cyclical in nature, and ultimately go through profit mean-reversion + multiple compression, resulting in years of no returns or even a permanent drawdown of more than 50%. This scenario does not require the company to hit disaster; it only requires expectations to return to reality.

Comparison With Other Opportunities

Against its strongest peer competitors, Cummins is neither the worst nor the best. Relative to CAT, it is somewhat cheaper, but its CapEx, cyclicality, customer structure, and business quality are also not as "irreplaceable" as CAT's; relative to PCAR, it gets no clear discount on EV/EBITDA; relative to WAB, it is not cheap either. In other words, it is not now "the cheapest good company among peers," but more like "a company of decent quality among peers whose price already reflects a lot of good news."

Against broad indices, I am not confident that buying Cummins today will clearly beat simply buying the S&P 500. The reason is not that the company is not good enough, but that at the current price, the "starting free-cash-flow yield / owner-earnings yield" you buy for this company is not high; and the S&P 500 at least offers stronger asset diversification. The company's own shareholder-return performance in 2025 was very impressive — full-year total shareholder return of 49%, higher than its self-defined peer group's 24% and the S&P 500's 18% — but that describes past price performance and does not automatically translate into excess returns over the next decade.

Against the risk-free rate, the issue is even sharper. The Reuters market-data page shows the U.S. 10-year Treasury yield at about 4.56%. Yet by my conservative owner-earnings estimate, the current price corresponds to an owner-earnings yield of only about 3%. In other words, if you buy today, the starting "business owner-earnings yield" it offers is even below the risk-free rate; the investment holds up only by relying on years of future growth and sustained high margins to compensate. For a "balanced-to-conservative" investor, this is not an ideal payoff.

If I could hold only five assets, at the current price I would not put Cummins among them; but if in the future it falls, due to the cycle, market panic, or a cooling boom, to the ideal buy range I gave, it would re-enter the candidate list.

Investment Checklist and Final Conclusion

Investment Checklist

Checklist Question Conclusion
Can I understand this business? Pass
Does it have long-term stable demand? Pass
Does it have a durable moat? Pass, but not extremely wide
Does it have pricing power? Uncertain / partial pass
Can it generate stable free cash flow? Pass, but cyclically volatile
Is its return on capital excellent? Pass, but volatile and disturbed by special items
Is management trustworthy? Pass, but with a reserved discount
Is capital allocation rational? Pass, but not excellent
Is the balance sheet sound? Pass
Is the valuation below intrinsic value? Fail
Is the margin of safety sufficient? Fail
Does long-term holding leave me at ease? Uncertain; the price determines the degree of ease
Which key facts would make me sell? If data-center demand slows, Accelera keeps posting large losses, key OEM relationships deteriorate, or another major compliance event recurs, I should reassess or even sell
Am I only wanting to buy because the stock has risen or because of market sentiment? Right now, that is easily the case

Open Questions and Limitations

The main conclusions of this report are built on the latest official annual report, quarterly materials, proxy statement, company IR pages, and authoritative market data; but three limitations still need to be made clear: First, maintenance CapEx cannot be read directly from public statements, so owner earnings must rely on a conservative estimate; Second, some peer ROIC and P/FCF metrics are not fully consistent in the public summaries retrieved, so I treat relative valuation only as a supplement; Third, the sustainability of data-center-related orders and profit after 2025 still needs to be verified by tracking subsequent quarters.

Final Investment Conclusion

【Final Rating】 Watch

【One-Sentence Investment Thesis】 Cummins is a long-established power-systems leader with a real industrial moat and fairly strong cash-generating ability, but the current price has already priced in a good deal of good news and lacks the margin of safety that would reassure conservative long-term investors.

【Core Bull Case】

  • Its global power-systems, components, distribution, and service network is deep, and its installed base and aftermarket value are real.

  • In 2025, with North American trucks on the weak side, Distribution and Power Systems still set high-boom records, showing a business mix superior to a pure engine company.

  • Operating cash flow and full-cycle cash generation are decent, the balance sheet is sound, and dividend discipline is fairly strong.

  • Management's pay system puts ROIC, EBITDA, and operating cash flow at the core, and the governance framework is broadly rational.

  • If the price falls significantly in the future, it is qualified to be a candidate for a long-term value position.

【Core Bear Case】

  • The current valuation is on the high side, at about 32–36x on conservative owner earnings, lacking a margin of safety.

  • Data-center power demand and the North American truck recovery are already highly anticipated by the market, an unfriendly payoff.

  • Accelera's repeated restructurings, impairments, and business exits show new-energy capital deployment still carries clear trial-and-error costs.

  • The historical emissions settlement is a serious compliance stain that cannot be easily erased by "it's already been paid."

  • The receivables-sale program makes the operating-cash-flow measure less pure than it appears.

【Key Assumptions】

  • Data-center backup power-generation demand stays at a fairly high level for the next several years.

  • The North American medium- and heavy-duty truck market recovers from the bottom of the cycle without falling into a deeper recession.

  • Accelera's losses keep narrowing, with no larger strategic failure recurring.

  • Major OEM relationships stay stable, with no significant loss of share among large customers such as PACCAR.

  • No major compliance cash blow on the scale of 2023–2024 recurs.

【Fair Buy Price】 The buy range I more readily endorse is $280–350/share. The basis is not "the historical low," but that within this range Cummins comes closer to my conservative-to-neutral owner-earnings discounted value and begins to offer a margin of safety sufficient to hedge the industrial cycle, transition risk, and valuation-contraction risk. If it falls to $350–450, it can enter the "researchable, buy-in-tranches" range; above $550, I consider it clearly on the expensive side.

【Target Holding Period】 If the purchase price is good enough, it suits a hold of 5–10 years or more; if bought at today's price, it is more like betting on a medium-term boom being maintained than locking in a high-conviction long-term return.

【Expected Annualized Return】 This is my conservative estimate based on the current price, not a market forecast:

  • Conservative scenario: -2% to +1%.

  • Base scenario: +3% to +6%.

  • Optimistic scenario: +7% to +10%. The reason the center is not high is not that the company is poor, but that the starting valuation is too high, with a large part of future returns already reflected in advance.

【Maximum Loss Risk】 If, over the next 3 to 5 years, a combination of "data-center demand cooling + a weak truck-cycle recovery + Accelera posting losses/impairments again + the valuation multiple reverting to the industrial-stock average" occurs, I think a return of the share price to $190–270 is not unimaginable. For someone buying near $639 today, this implies a potential permanent capital loss scenario of about 58% to 70%. This is an inference, not a forecast, but it is enough to show that the risk at the current price is not a small fluctuation but a deterioration of the payoff.

【Tracking Metrics】 Going forward I will keep tracking the following metrics:

  • Power Systems and Distribution revenue growth and EBITDA margins.

  • Data-center-related power-generation orders, deliveries, and shifts in management's commentary.

  • Engine and Components volume and profit elasticity in the North American medium- and heavy-duty truck recovery.

  • Accelera revenue, losses, impairments, and CapEx.

  • The gap between operating cash flow and CapEx, rather than only net income.

  • Receivables, inventory, payables, and use of the AR-sale program.

  • Sales, partnership, and share changes at PACCAR and other major OEM customers.

  • New regulatory, recall, environmental, or litigation matters.

  • Whether dividends and buybacks continue to maintain discipline.

  • Whether leverage levels and interest coverage deteriorate noticeably.

【Signals That Would Trigger Reassessment】

  • The data-center-driven high power-generation boom cools noticeably.

  • Power Systems or Distribution margins pull back significantly and persist for more than two quarters.

  • Accelera again shows large impairments, exit costs, or widening cash losses.

  • The partnership with PACCAR or other key OEMs turns clearly unfavorable.

  • A major regulatory settlement, recall, or compliance event recurs.

  • Capital allocation shifts to high-priced buybacks or high-premium M&A.

  • Operating cash flow keeps relying on working-capital release or off-balance-sheet arrangements for support.

【Final Recommendation】 Put plainly: Cummins deserves respect, but its current price does not deserve excitement. If you already hold it at a reasonable cost, I lean toward viewing it as an industrial leader you "can keep holding, while closely tracking the boom and the valuation"; if you are about to build a new position today, from a long-term value perspective I would rather suggest putting it on a high-quality watch list than rushing to buy. The truly favorable returns often come from "buying a good company when it meets bad sentiment," not from "chasing in when a good narrative, a good boom, and a good stock price all appear at once."

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.

CumminsEnginesPower SystemsData Center Backup PowerDiversified IndustrialsValue Investing
Reader Q&A10

Baillie Framework · Ten Questions for Growth Investing

10

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

Baillie Framework · Ten Questions for Growth Investing — score profile: 42/100 total Ceiling 5/10 · Revenue 2x 3/10 · Next engine 4/10 · Moat 5/10 · Reinvention 5/10 · Management 4/10 · Customer need 6/10 · Unit economics 5/10 · 5x path 2/10 · Blind spot 3/10 0510 How high is its market ceiling? Is it enlarging an existing pie, or creating an entirely new market? — 5/10 Ceiling 5 Can its revenue at least double over the next five years? Is growth driven mainly by volume, price, or new businesses? — 3/10 Revenue 2x 3 Five years from now, what will take the baton as the next growth engine? Does this "second curve" exist today? — 4/10 Next engine 4 What is its core competitive advantage? Will this moat widen or narrow over the next three to five years? — 5/10 Moat 5 If its core business is disrupted, does it have the genes to reinvent itself? How does it treat mistakes and bad news? — 5/10 Reinvention 5 Does management (especially the founder) have a long-term vision, with interests deeply tied to the company? Is it willing to sacrifice current profit for five to ten years out? — 4/10 Management 4 If it disappeared tomorrow, how much would customers miss it? Is its way of growing sustainable, without relying on harming society or regulators? — 6/10 Customer need 6 How is this business's unit economics (gross margin, incremental returns)? Does it get better or worse at scale? Where is the money it earns spent? — 5/10 Unit economics 5 For it to rise five-fold in ten years, which conditions must simultaneously hold? Are these conditions realistic? What expectations does today's stock price imply? — 2/10 5x path 2 Why has the market not yet realized all this? Is it that it can't understand it, looks down on it, or can't see far enough? What will become the "narrative inflection point"? — 3/10 Blind spot 3
  • How high is its market ceiling? Is it enlarging an existing pie, or creating an entirely new market?5/10

    The ceiling is limited; in essence it is "enlarging and defending a slice of an existing, mature pie" rather than creating an entirely new market. Measured against Baillie Gifford's "five-fold in ten years" yardstick, Cummins is not the kind of target whose TAM can open up exponentially — its main battlegrounds (heavy-duty diesel/natural-gas engines, components, aftermarket service, generator sets) are all mature cyclical industries that have existed for decades. The report is clear on this characterization: "this industry as a whole belongs to the cyclical sub-industries within a mature industry," where demand exists over the long run but is at the mercy of the macro economy, freight rates, capital expenditure, and emissions regulations. So the sources of growth are mainly share, mix, and unit price, not opening up a demand pool that did not previously exist.

    The only thing with a whiff of a "new pie" is data-center backup power, but it looks more like "an old pie being enlarged again by AI" than something created from nothing. Backup generators are themselves a mature category; what has truly changed is that AI compute is pushing per-rack power higher and blowing out the backup-power architecture needs of hyperscale data centers. A third-party estimate puts the data-center generator market at about $7.88 billion in 2026 and about $9.84 billion by 2031, a compound annual growth rate of only about 4.55% — the absolute size is not large and the growth rate is hardly explosive. Cummins's positioning advantage is that it and Caterpillar together hold more than 35% share of data-center backup generators, and in early 2026 it added about $150 million of capacity to expand its high-power QSK95 units. This line is indeed ramping quickly (the report discloses first-quarter-2026 Power Systems revenue of $2 billion, up 19% year over year), but it is "capturing the AI dividend within an existing category," not creating a new market.

    Conclusion: the ceiling is the "enlarge the existing pie + defend share" type, not the "open up without limit" type. The report scores industry attractiveness at 3/5 and business understandability at 4/5, which is exactly the footnote to this positioning: the business is understandable and demand exists over the long run, but the upper bound of its growth imagination is boxed in by the mature-industry attribute and the cyclicality. For the Baillie Gifford paradigm, this is an inherently disadvantaged dimension — there is no entirely new demand curve capable of supporting a multi-fold revenue increase over ten years, and forcibly applying a "blue-sky imagination" would distort things.

    Jun 11, 2026
  • Can its revenue at least double over the next five years? Is growth driven mainly by volume, price, or new businesses?3/10

    Doubling in five years is almost impossible. Doubling revenue in five years implies a compound annual growth rate of about 15%, yet Cummins is a mature cyclical giant with annual revenue already reaching $33.67 billion and 2025 net income of $2.843 billion, where the sheer size is itself a drag on growth. The most direct evidence the report offers: from 2020 to 2025 revenue rose from $19.81 billion to $33.67 billion, but across the three years from 2023 to 2025 it was "essentially flat" (fluctuating around $34.07 billion → $34.10 billion → $33.67 billion). A company that has been treading water for nearly three years lacks a realistic basis for doubling its scale over the next five.

    On growth composition, each of the three engines — volume, price, and new businesses — has its own constraints.

    • Volume: the core Engine and Components are highly exposed to the North American medium- and heavy-duty truck cycle, which was in a downturn in 2025; the report explicitly says "the North American medium- and heavy-duty truck market is under pressure." A cyclical recovery can contribute elasticity, but that is a return to the mid-point, not structural expansion.
    • Price: pricing power is localized — the report lists the checklist conclusion for "does it have pricing power?" as "uncertain / partial pass," and notes that large OEMs (per the report, PACCAR accounted for 13% of consolidated net sales in 2025) retain strong bargaining power on price and delivery. Driving a doubling through price increases is unrealistic.
    • New business / boom: the real increment is in Power Systems (data-center backup power); in the first quarter of 2026 that segment's revenue was $2 billion, up 19% year over year, and the company accordingly raised full-year revenue guidance to up 8% to 11% year over year. But that is high growth in a single segment, and even if the company-wide 8–11% growth were sustained for five years it would cumulate to only about 47%–68%, nowhere near a doubling.

    Conclusion: over the next five years revenue is most likely to be moderate growth — "low-to-mid single-digit compounding, the cycle layered with the data-center dividend" — rather than a doubling. The drivers are primarily "the Power Systems ramp pulled by data centers + volume from the truck-cycle recovery," with price as a secondary factor, and there is no new-business curve capable of pushing company-wide revenue to a double. Cummins clearly fails Baillie Gifford's "double in five years" gate.

    Jun 11, 2026
  • Five years from now, what will take the baton as the next growth engine? Does this "second curve" exist today?4/10

    The "second curve" visible today is Power Systems / data-center backup power, but it looks more like the main engine being realized right now than the next-stage rocket that takes the baton five years out; the "zero-emission transition" line that should really take the baton has not yet been proven out. Baillie Gifford's question here is "what relays five years from now," and Cummins's answer is a bit awkward: its strongest growth story is burning right now, while the next-stage curve is still in trial and error.

    The curve being realized (data-center power generation): it exists, but belongs to "now" rather than "five years out." In the first quarter of 2026 Power Systems revenue was $2 billion, up 19% year over year, with a segment EBITDA margin as high as 29.5%; on this basis the company raised full-year revenue guidance to up 8% to 11% year over year, with EBITDA guidance of 17.75% to 18.50%. AI compute has blown out the backup-power demand of hyperscale data centers, and in early 2026 Cummins added about $150 million of capacity to expand its high-power QSK95 units. The catch is that the report lists "how long the data-center backup-power boom can last" as the first of its three biggest uncertainties — it is a strong current dividend, but whether it can become a stable second curve supporting "years 3 through 10" is in doubt.

    The curve that ought to take the baton (Accelera's zero emissions: batteries, electric drives, fuel cells, electrolyzers): it exists today, but is contracting rather than expanding. This is where the most caution is warranted. The report discloses that in 2025 the company conducted a strategic review of the electrolyzer business under Accelera, recognized $458 million in related charges for the full year (of which $415 million was non-cash), and planned to stop new electrolyzer commercial activity; in the first quarter of 2026 it completed the sale of its low-pressure fuel-cell business, recognizing $199 million in net charges. Accelera's 2025 external revenue was only $423 million, the smallest of the five segments. In other words, the new-energy platform that ought to shoulder "the next decade" is, for now, cutting losses while shrinking rather than investing while growing.

    The potential true long curve (the HELM fuel-agnostic platform + natural-gas engines): it extends the core business rather than being an independent new growth pole. The report notes the company has an HELM fuel-agnostic platform, natural-gas engines, hybrid mining conversions, and other positioning, and has not locked itself into a single technology path. But these look more like defensive extensions that "let the engine core business live longer" than an incremental engine capable of recreating another Cummins-sized business.

    Conclusion: the second curve (data-center power generation) does exist and is ramping, which is a plus; but its timing is mismatched with "taking the baton five years out," while the zero-emission curve genuinely aimed at the 2030s is still in trial and error, even in retreat. The report's line "positioning for the transition does not equal succeeding at the transition" is the most honest footnote to this dimension — the baton is visible but not yet firmly grasped.

    Jun 11, 2026
  • What is its core competitive advantage? Will this moat widen or narrow over the next three to five years?5/10

    The core moat is a traditional industrial moat made of three pieces — "installed base + global distribution-service network + OEM-embedded switching costs" — of moderate strength (the report rates it 3.5/5); over the next three to five years it is most likely a trade-off in which "the old-business moat narrows slowly while the aftermarket and power-generation moat widens," rather than the whole widening significantly. This is not an internet-style network effect but a real barrier built up over decades of a reliability reputation and a physical network.

    Breaking down the moat's sources (all evidence-based):

    • Brand and reliability: in heavy-duty power, brand equals the long-term perception among fleets, equipment owners, and OEMs of "reliability, fuel economy, maintenance, and residual value." The report notes Cummins's relationship with PACCAR has lasted about 80 years, with long-term supply arrangements with Traton, Daimler, and Stellantis — something a new entrant cannot replicate through marketing.
    • Scale and channel: about 640 distribution nodes + more than 13,000 certified dealer points + service in about 190 countries, significantly lowering customers' downtime costs and raising aftermarket stickiness.
    • Switching costs: for end customers it is a whole system of "engine + aftertreatment + control + service + technicians + parts + regulatory certification"; for OEMs, switching platforms means redoing engineering validation, calibration, and emissions certification. It is real, but the report explicitly characterizes it as "not insurmountable," so it is classed as a moderate moat.

    Why the old-business moat is narrowing: the traditional barrier is mainly built around the internal-combustion engine. The report flags the risk — if truck electrification, hydrogen, alternative powertrains, or OEM in-house development advance faster than expected, the durability of the existing high-margin business will be compressed; the company's own strategic retreats in electrolyzers and low-pressure fuel cells precisely show that the long-term dominance of the internal-combustion engine carries a tail risk of technological substitution.

    Why the aftermarket/power-generation moat is widening: the data-center backup-power boom has raised the strategic standing of distribution and power generation systems. In the first quarter of 2026 the Power Systems segment's EBITDA margin reached 29.5%, and Cummins and Caterpillar together hold more than 35% share of data-center backup generators; the delivery and after-sales barrier formed by high-power units (such as the QSK95) plus a global service network is being reinforced by this wave of demand.

    Conclusion: the moat is real but not extremely wide, and the direction is a structural migration in which "the internal-combustion-engine end narrows slowly while the service/power-generation end widens." The report's checklist judges "does it have a durable moat?" as "pass, but not extremely wide." For Baillie Gifford, this is a "qualified but unremarkable" dimension — there is a real barrier to defend, but not the kind of top-tier moat that keeps self-reinforcing and widening over time.

    Jun 11, 2026
  • If its core business is disrupted, does it have the genes to reinvent itself? How does it treat mistakes and bad news?5/10

    The gene for self-reinvention "exists, but it is the cautious trial-and-error type, not the aggressive self-disruption type"; its treatment of mistakes and bad news is quite mature — it owns up, cuts losses, and discloses honestly. The implicit premise of this question is "can it save itself when the core business is disrupted," and for a century-old company that has bet its life on the internal-combustion engine, this is exactly where it should be interrogated.

    Reinvention gene: right sense of direction, conservative execution. Cummins has not locked itself into a single technology path — the report notes it has multiple lines of positioning such as the HELM fuel-agnostic platform, natural-gas engines, hybrid mining conversions, and batteries/electric drives/fuel cells, and it set up the Accelera segment specifically to carry the zero-emission transition. This shows it recognizes the long-term substitution risk to the internal-combustion engine and has placed real-money bets. But "betting" does not equal "proving out": in 2025 the company conducted a strategic review of the electrolyzer business, recognized $458 million in related charges (of which $415 million was non-cash) and planned to stop new commercial activity; in the first quarter of 2026 it sold the low-pressure fuel-cell business, recognizing $199 million in net charges. The report's line "positioning for the transition does not equal succeeding at the transition" pinpoints the limit of this gene — it can recognize change and is willing to invest, but it lacks the disruptive edge to "revolutionize itself and voluntarily cannibalize its cash cow."

    Treatment of mistakes and bad news: this is Cummins's genuine plus. The report gives management an assessment of "honesty and operational rationality above average," with three pieces of evidence:

    • Does not avoid bad news: whether the large electrolyzer impairment or the low-pressure fuel-cell sale, both were clearly disclosed.
    • Cuts losses proactively: for loss-making new-energy businesses it cuts when it should cut and exits when it should exit, without saving face by hanging on.
    • Owns up to historical stains too: for the emissions settlement disclosed in 2023 and effective in 2024, the company accrued about $2 billion in the fourth quarter of 2023 and paid $1.9 billion in settlement cash in the second quarter of 2024 — a serious compliance stain, but the disclosure and handling were faced head-on.

    Conclusion: the reinvention gene leans "steady trial and error," not "self-disruption"; the error-correction culture is clearly mature and trustworthy. But note that "daring to cut losses" itself also means its new-business bets have a low hit rate (Accelera's repeated restructurings and impairments). For the Baillie Gifford paradigm, the ideal target is a species that "can grow an entirely new self when its core is disrupted"; Cummins is more of a mature industrial body that "can sustain its core for a long time, cautiously positions its follow-up moves, and candidly cleans up the mess when it errs" — reliable, but lacking the imagination for disruptive regeneration.

    Jun 11, 2026
  • Does management (especially the founder) have a long-term vision, with interests deeply tied to the company? Is it willing to sacrifice current profit for five to ten years out?4/10

    Management has a long-term vision and healthy pay-metric orientation, but the item "interests deeply tied to the company" clearly falls short of Baillie Gifford's preferred standard — this is a century-old company run by professional managers, without founder-style high-ownership alignment. The report scores management and capital allocation 3.5/5, characterizing it as "honesty and operational rationality above average, but capital allocation not yet excellent."

    Interest alignment: on the weak side, the main point deducted on this question. Jennifer Rumsey has been CEO since August 2022 and chair since August 2023. Per the 2026 proxy statement, Rumsey's beneficial ownership was about 119,419 shares, and all directors and current executive officers together held about 415,453 shares, each less than 1% of total shares outstanding. This shows executives and shareholders are "not almost entirely unaligned," but it is by no means a founder-style, very-high-ownership structure. Baillie Gifford most values "a founder/helmsman putting their net worth on the line with the company, in the same boat as minority shareholders," and Cummins is inherently disadvantaged on this dimension — it is an institutionalized mature company, not a founder-driven growth machine.

    Long-term vision and "sacrificing the present for five to ten years out": there is real evidence.

    • Pay metrics oriented to the long term, not the EPS game: the annual bonus centers on EBITDA at 70% + operating cash flow at 30%; long-term incentives are 100% performance-based, and in the domestic plan's three-year assessment ROIC is weighted 80% and EBITDA 20%. Putting ROIC and cash flow at the core is far better than simply chasing earnings per share.
    • Willing to swallow short-term losses for the long term: continuing to invest in Accelera's zero emissions and positioning even while knowingly losing money in the short term is itself an expression of "ceding current profit for five to ten years out" (even if the hit rate is not high).
    • Governance recognized: the 2025 say-on-pay vote received about 91% support.

    Capital allocation: rational but not excellent, with both pluses and minuses. Pluses — the dividend has been raised for 16 consecutive years, and in 2025 the company raised its annual dividend to $7.64 per share without overdrawing the balance sheet for short-term returns. Minuses — Accelera's repeated restructurings and impairments from 2023 to 2025 show new-energy deployment carried exploration costs; the emissions settlement disclosed in 2023 (with about $2 billion accrued) is a major compliance and governance stain; and in 2024 the company established a receivables-sale arrangement and included the collections in operating cash flow, making the cash-flow measure less "clean."

    Conclusion: management is trustworthy (honest, long-term, disciplined), but "deep interest alignment" can only be rated moderate-to-below by Baillie Gifford's standard. The report's checklist judges "is management trustworthy?" as "pass, but with a reserved discount" and "is capital allocation rational?" as "pass, but not excellent" — this is a professional, reliable team, but not a "nearly Buffett-like / founder-like" one.

    Jun 11, 2026
  • If it disappeared tomorrow, how much would customers miss it? Is its way of growing sustainable, without relying on harming society or regulators?6/10

    If Cummins disappeared tomorrow, customers would "miss it quite a lot" — especially fleets, mining, and data-center operators with already-installed equipment, who would immediately face parts-supply cutoffs, service interruptions, and soaring downtime costs; at this layer it is "indispensable but not the only one." Its way of growing is broadly sustainable and does not rely on harming society, but internal-combustion-engine emissions compliance is a structural constraint that must be watched continuously. This question's implicit dual premise — "indispensability" and "social/regulatory sustainability" — sees Cummins strong on one and moderate on the other.

    Indispensability: high, but not irreplaceable. What truly makes customers unable to do without it is not selling new engines, but the physical network the report describes: about 640 distribution nodes + more than 13,000 certified dealer points, service in about 190 countries, layered on a global installed base. For a piece of equipment running on the road, working in a mine, or on standby in a data center, parts supply and fault response directly determine downtime losses — this layer of service capability cannot be replaced at equal scale in the short term, so "it would be missed a lot." But at the "selling new units" layer, OEMs have choices other than Cummins (in-house or competitors), and large customers such as PACCAR retain bargaining and switching leverage (report: pricing power "uncertain / partial pass"), so it is "high dependence, not deadlocked dependence."

    Social/regulatory sustainability: growth itself is legitimate, but emissions are a real regulatory constraint.

    • Does not make money by harming society: Cummins's business is selling power, components, power generation, and service — old-school, clear, verifiable, not reliant on advertising manipulation or regulatory arbitrage. Data-center backup power and truck power are rigid demands of the real economy, and the growth logic is clean.
    • Emissions compliance is a constraint that must be faced: internal-combustion engines are inherently subject to tightening emissions regulations. The emissions settlement the report discloses — about $2 billion accrued in the fourth quarter of 2023 and $1.9 billion in settlement cash paid in the second quarter of 2024 — is exactly the proof that "a compliance failure can bring an enormous cost." This is not a "society-harming business model," but "a core business subject to strict social/regulatory standards that has stumbled historically." The company's positioning in the HELM fuel-agnostic platform, natural gas, and zero-emission technologies is in part a response to this long-term regulatory trend.

    Conclusion: indispensability is on the high side (the service network is a real moat), social sustainability is acceptable (the business is legitimate), but emissions regulation is a long-term constraint hanging over the internal-combustion-engine core business, and it has carried a hard cash cost. To Baillie Gifford's question of "how much would customers miss it if it disappeared, and can its growth withstand social scrutiny," Cummins's answer is "it would be missed a lot, growth is legitimate, but it must be run long-term with the sword of emissions compliance hanging over it" — steady, but not the kind of target with zero regulatory tail risk.

    Jun 11, 2026
  • How is this business's unit economics (gross margin, incremental returns)? Does it get better or worse at scale? Where is the money it earns spent?5/10

    The unit economics are "upper-mid industrial-manufacturing level, but not a high-leverage asset-light model": gross margin about 25%, operating margin about 12%; growing scale brings structural improvement (a rising share of aftermarket/power generation) rather than an exponential jump in marginal margins; and the money earned is spent mainly on dividends, maintenance CapEx, and the new-energy transition. Placed in the Baillie Gifford paradigm, this is a "solid but not sexy" dimension — real profit, real cash generation, but without software-style elasticity in incremental returns.

    Gross margin and margins: stable and upper-mid, but heavily contaminated by special items. In 2025 gross margin was about 25.3% and operating margin about 12.0%, and operating margins in 2020–2022 were also all just over 10%–11%, a stable center. But the report cautions: 2023 and 2024 GAAP profits were heavily distorted by the emissions settlement (depressing 2023) and the Atmus spin-off gain (lifting 2024) — looking at a single year's margin would mislead, and one must look across the cycle. This is a heavy-asset manufacturing business that structurally cannot reach the gross-margin level of a consumer monopoly or pure software.

    Incremental returns and scale effects: they come from "business mix" rather than a "unit-cost curve." This is the most crucial feature of Cummins's unit economics. The report repeatedly stresses that what is truly cycle-resistant and higher in profit quality is the aftermarket and power generation systems — in 2025 parts and service within the Distribution segment totaled $5.871 billion, about 47% of that segment; and Power Systems achieved a segment EBITDA margin of 29.5% in the first quarter of 2026. In other words, if growing scale is accompanied by a rising share of aftermarket and high-end power generation, profit quality will improve; but if it merely piles up cyclical OEM new-unit volume, the marginal-margin improvement is limited. Put differently, its "bigger is better" depends on product-mix optimization, not natural operating leverage.

    Return on capital: excellent but highly volatile. The report's checklist judges "is return on capital excellent?" as "pass, but volatile and disturbed by special items," and notes the pay system puts ROIC (weighted 80% in the three-year assessment) at the core — management does watch capital efficiency. But strong cyclicality plus occasional large special items means the ROIC at any single point in time needs to be viewed on a normalized basis.

    Where the money earned goes (direction of capital allocation):

    • Maintaining existing competitiveness: the report uses a conservative assumption, treating 85%–90% of 2025 total CapEx ($1.235 billion) as maintenance, i.e., about $1.05 billion to $1.11 billion — meaning a substantial portion of investment merely "holds the line on competitiveness" rather than high-return expansion.
    • Returning capital to shareholders: over full-year 2025 it returned about $1.055 billion through dividends ($7.64 per share, raised for 16 consecutive years), with small-scale buybacks under the $2 billion authorization approved in 2021; in the first quarter of 2026 it returned a total of $519 million to shareholders.
    • Transition investment: continuing to invest in Accelera's zero emissions, but with a low hit rate and multiple rounds of impairment.

    Conclusion: the unit economics are solid and can keep generating cash (the report says "pass, but cyclically volatile"), but there is no software-style incremental return, and the scale dividend hinges on mix optimization. To Baillie Gifford's triple question of "unit economics, incremental returns, where the money goes," Cummins gives an honest answer of "upper-mid industrial standard + structural aftermarket improvement + disciplined shareholder returns" — a good business, but not the top-tier model that automatically earns more and more with scale.

    Jun 11, 2026
  • For it to rise five-fold in ten years, which conditions must simultaneously hold? Are these conditions realistic? What expectations does today's stock price imply?2/10

    A five-fold rise in ten years (per the question's framing, corresponding to about 17.5% annualized) requires several strong assumptions to hold at once, and these conditions are not realistic for a mature cyclical industrial giant; more crucially, today's stock price of about $630 has already priced in plenty of good news, leaving little room for "mean reversion." This is the question on which Cummins fares worst in the Baillie Gifford framework — it was never a "blue-sky five-bagger" species, and the current price has overdrawn the optimistic scenario.

    The conditions for a five-fold rise in ten years that must all hold (none can be missing):

    1. The data-center backup-power boom lasts for many years: yet the report lists "how long the data-center backup-power boom can last" as the first of its three biggest uncertainties; a third-party estimate puts the data-center generator market at a 2026–2031 compound growth rate of only about 4.55%, with an absolute size of only around ten billion dollars, hard to single-handedly support a company-wide five-fold rise.
    2. The North American medium- and heavy-duty truck cycle recovers strongly and durably: but this is a cyclical return to the mid-point, not structural expansion.
    3. Accelera's losses genuinely narrow, or even turn into a new growth pole: yet the reality is a 2025 electrolyzer impairment of $458 million and a first-quarter-2026 low-pressure fuel-cell sale recognizing $199 million in net charges — contracting rather than ramping.
    4. The margin center shifts up systematically + the valuation multiple does not contract: but the current valuation is already at a high level (see below), and the multiple is more likely to compress than expand. All four conditions need to be realized "simultaneously" to approach a five-fold rise, and the report's judgment is that the current stock price is "already near or even above the upper edge of my optimistic scenario" — which itself shows the five-fold narrative lacks payoff support.

    What today's stock price implies (the core): As of the close on June 10, 2026, CMI was about $630.52, with a market cap of about $87 billion, up about 93.6% over the past 52 weeks. On valuation: TTM P/E about 32.75x, EV/EBITDA about 18.4x, P/B about 7.0x, P/FCF about 32.6x. Based on 2025 actual free cash flow of $2.386 billion, the report estimates P/FCF at about 37x and, on a conservative owner-earnings basis, about 32 to 36 times. This set of multiples is not cheap for a strongly cyclical heavy-industrial company — the market is not pricing "a cheap manufacturing stock in a down cycle," but paying at once for three things: "a durable data-center dividend + a truck-cycle recovery + a narrowing of new-energy losses."

    A hedging signal worth pointing out: the current forward P/E of about 20.7x is notably below the TTM 32.75x, implying the market already expects a sharp earnings rebound in 2026 (pulled by data centers). This slightly softens the "absolutely on the expensive side" conclusion — if forward earnings come through, the valuation is not outrageously high; but even on a forward basis, it is still far from the growth needed to support "five-fold in ten years."

    Conclusion: the combined probability of the conditions needed for a five-fold rise in ten years is low, and the current price has already overdrawn the optimistic scenario, with an insufficient margin of safety (the report's checklist judges both "is the valuation below intrinsic value?" and "is the margin of safety sufficient?" as fail). The report gives an ideal buy range of $280–350/share, while the current price of about $630 has already been classed by it as "on the expensive side." To Baillie Gifford's question of "what a five-fold rise requires and what the current price implies," Cummins's answer is: a five-fold rise requires several strong assumptions to hold at once, with low realism, and today's price looks more like paying for "good things continuing to happen" than "leaving room for mean reversion."

    Jun 11, 2026
  • Why has the market not yet realized all this? Is it that it can't understand it, looks down on it, or can't see far enough? What will become the "narrative inflection point"?3/10

    Baillie Gifford's question here assumes there is "a perception gap the market has not yet realized," but for today's Cummins the conclusion is exactly the opposite: the market is not "unable to understand it, looking down on it, or unable to see far," but has already fully — even somewhat optimistically — realized its merits. The direction of the perception gap is negative (the market has priced it in), not positive (wrongly punished and waiting to be discovered). This must be honestly acknowledged; one cannot force-fit "undiscovered value" to serve a growth narrative.

    Why the market has long "understood it, taken to it, and seen far":

    • Understood its quality: over the past 52 weeks the stock rose about 93.6%, with a TTM P/E of about 32.75x and P/B of about 7.0x — the market is clearly pricing it as "an industrial platform with higher certainty" rather than "a cheap manufacturing stock in a down cycle."
    • Taken to the data-center story: UBS has upgraded it to buy with a target price of $850, precisely on the grounds of an improving truck market and power demand — the sell side is far from ignoring the core bull logic; it is enthusiastic about it.
    • Seen far to forward earnings: the current forward P/E of about 20.7x is notably below the TTM, showing the market has already priced in the 2026 earnings rebound pulled by data centers.

    The report's judgment on this is incisive: "You have not bought, at a low, an industrial leader wrongly punished by the market; you have bought, at a high, three things at once — the data-center boom, the truck recovery, and the narrowing of transition losses." On the checklist, the item "am I only wanting to buy because the stock has risen or because of market sentiment?" concludes "right now, that is easily the case."

    So what will the true "narrative inflection point" be? For Cummins, the inflection is more likely to be "falsification to the downside" than "discovery to the upside." Because the good news is already priced in, the marginal information that can change the narrative is mainly on the negative side:

    • The data-center boom cools: backup power-generation orders, deliveries, or management's commentary weaken — this is the core pillar of the current premium, and once it loosens, the valuation will quickly revert toward the industrial-stock average.
    • The North American medium- and heavy-duty truck recovery is weaker than expected: Engine and Components profit recovery is slower than the market imagines.
    • Accelera again posts large impairments/exit costs: normalized free cash flow is further eroded (the 2025 electrolyzer impairment of $458 million and the first-quarter-2026 low-pressure fuel-cell sale recognizing $199 million in net charges are already prior offenses).
    • A major compliance/regulatory event recurs: a replay of an emissions-settlement shock on the 2023–2024 scale. Conversely, upside inflections (the data-center boom lasting for years beyond expectations + a strong truck recovery + new energy turning profitable) of course also exist, but those would be "realizing already-full expectations," an asymmetric payoff — the good news has been paid for, and only the bad news is unexpected.

    Conclusion: the honest answer to this question is that "the market has not misjudged, and the perception gap leans negative," while the narrative inflection is most likely a downside falsification of the boom/compliance rather than an upside value discovery. This is the opposite of the Baillie Gifford ideal of "undervalued, awaiting re-rating" — Cummins is a good company already fully priced by the market, and the truly favorable entry point will appear only when "a good company meets bad sentiment," not now.

    Jun 11, 2026
Ask about this report

Members can ask about this report; once answered it appears under "Reader Q&A" on this page. You can also highlight a passage in the text to ask about it directly.