Atlas Copco AB: Q2 2026 Group Orders Grew 26% Organically at a 1.13x Book-to-Bill, and SEK 208.80 Sits 34% Above the SEK 156 Conservative Value

The Q2 book-to-bill ratio was 1.13x, calculated as MSEK 50,951 orders divided by MSEK 44,974 revenue. Orders exceeded revenue by MSEK 5,977. H1 book-to-bill was similarly strong at 1.13x, with MSEK 96,346 of orders against MSEK 85,514 of revenue. That establishes a direction of backlog accumulation, but not a backlog balance: Atlas Copco's current Q2 report and 2025 annual material do not disclose a group order-backlog amount, its business-area composition or a percentage contractually scheduled to convert within 12 months. The MSEK 5,977 quarterly gap must therefore not be relabeled “backlog”; opening backlog, cancellations, delivery timing, acquisitions and currency all prevent that shortcut.

Vacuum Technique supplied the acceleration. Atlas's Q2 presentation says the business represented 25% of group orders on a rolling-12-month basis and delivered 59% organic order growth in the quarter. Compressor Technique represented 44% of rolling-12-month group orders and grew Q2 organic orders 19%; Industrial Technique was 14% and +11%; Power Technique 17% and +14%. Those percentages mix two deliberately different bases: the order shares are rolling 12 months, while the growth rates are Q2 year-on-year organic growth. They should not be read as Q2-quarter revenue weights.

Vacuum’s Q2 orders were MSEK 14,409, up 59% organically; revenue was MSEK 10,531, up 19% organically; operating profit reached MSEK 2,216 and margin rose to 21.0% from 18.9%. Semiconductor equipment demand was the principal driver, alongside healthy industrial, scientific and service activity. Compressor, still the larger profit pool, generated MSEK 4,745 of Q2 operating profit at a 24.1% margin. Thus the market's most dramatic growth story is Vacuum, while the earnings ballast remains Compressor.

The semiconductor demand evidence outside Atlas supports a real cycle turn rather than an isolated company-specific spike. SEMI's 2Q26 update projected global fab-equipment spending to rise 24% in 2026 and another 9% in 2027; its June update projected 300mm memory-equipment investment up 29% in 2026, supported by AI infrastructure, HBM and DDR5. TSMC, one of the most important end-demand signals in leading-edge semiconductor manufacturing, said in April that its 2026 capital budget was moving toward the high end of its range and in July raised its 2026 revenue-growth expectation to slightly above 40% in US-dollar terms.

That evidence also shows why 59% cannot be treated as a run rate. Atlas Vacuum's own 2025 Capital Markets Day material shows a much smoother long-term picture: from 2015 to the period ending September 2025, service revenue had compounded at roughly 17% and equipment at about 14%, including structural and currency effects. Atlas also described a shift from reactive toward planned and predictive service and further local capacity expansion. Those are durable drivers. A 59% quarterly order jump is a cyclical overshoot around them.

What would need to remain true for the exceptional order pace to persist is demanding: leading-edge fab construction would need to stay above prior plans; Atlas would need to retain or gain sub-fab vacuum content per wafer; customer utilization would need to translate into further equipment orders rather than merely service; and the current geographic capacity build would need to avoid a digestion phase. A move back toward low-single-digit or negative Vacuum order growth, book-to-bill below 1.0, or material downward revisions to semiconductor-fab spending would falsify the extrapolation. This is an inference from Atlas's order pattern and the external capex data, rather than company guidance.

The broader group is less dependent on one cycle than the Q2 headline suggests. Compressor serves general manufacturing and process industries. Industrial Technique is exposed to automotive assembly, general industry, machine vision and joining. Power Technique reaches construction, infrastructure, specialty rental and mobile-energy applications. Vacuum is tied to semiconductors, scientific applications and industrial vacuum. Q2 delivered double-digit organic order growth in all four, which is unusual because their end markets normally do not turn at exactly the same time.

The market is therefore trading a convergence narrative: a semiconductor upturn powerful enough to reaccelerate group orders, while Compressor service and the installed base prevent that upturn from becoming a fragile single-cycle bet. The A share's 2026 year-to-date appreciation of roughly 26% through September 8 and approximately 34% one-year appreciation have already reflected a material part of that improvement.

History explains why investors pay such a premium. Atlas was founded in Stockholm in 1873 to supply equipment for Sweden's railway build-out, nearly failed in the 1880s, then pivoted toward steam machinery and eventually pneumatic equipment. It abandoned diesel engines after World War II, globalized around compressed-air technology and service, and repeatedly remade its portfolio through acquisitions. Edwards and Leybold helped create today's Vacuum franchise; Vacuum became a separate business area in 2017. The mining and rock-excavation operation was then separated into Epiroc, leaving Atlas with Compressor, Vacuum, Industrial and Power Technique.

Epiroc's 2018 separation is the most important accounting boundary in the historical record. Atlas shareholders received one Epiroc A share for each Atlas A share and one Epiroc B for each Atlas B; Epiroc began trading on Nasdaq Stockholm on June 18, 2018. The information brochure showed that the continuing Atlas businesses generated 2017 revenue of MSEK 85,522 and operating profit of MSEK 18,251, while the Epiroc operations generated MSEK 31,440 and MSEK 5,949 respectively. Pre-separation consolidated history is consequently not directly comparable with today's company.

Post-separation financial quality is real. Revenue rose from MSEK 99,787 in 2020 to MSEK 176,771 in 2024 before declining to MSEK 168,343 in 2025, while return on capital employed stayed in the 20s or around 30%: 23% in 2020, 27% in 2021, 29% in 2022, 30% in 2023 and 28% in 2024. The pandemic year cut 2020 revenue by 4% but group operating margin still held at 19.2%; 2021 and 2022 margins recovered to 21.2% and 21.4%. That is stronger evidence for the installed-base quality thesis than a generic ten-year quality score.

Cash conversion is good over a full cycle but not mechanically perfect. Aggregate Atlas-reported operating cash flow for 2021–2025 was MSEK 117,446 against aggregate net income of MSEK 125,887, a 0.93x ratio. The two weak years were 2022 and 2023, when cash conversion was about 0.73x and 0.83x; 2024 recovered to roughly 1.04x and 2025 to 1.01x. The 2022 annual report explicitly linked the deterioration to a MSEK 7,415 working-capital increase associated with inventory and supply constraints.

The serial-acquisition model is material but has not overwhelmed organic economics. Atlas completed 18 acquisitions in 2019 and 30 in 2022, and the company maintains an extensive acquisition ledger. In H1 2026, acquisitions added roughly 4 percentage points to reported order growth and 3 points to revenue growth, while currency subtracted about 8 and 7 points respectively. Organic growth therefore remained the dominant explanation for the current H1 acceleration.

The qualitative portrait is high-quality compounding growth with a cyclical semiconductor accelerator. The label is deliberately different from “pure high-quality growth”: Atlas's service installed base, decentralization, capital allocation and returns on capital look structural, while Vacuum's current order impulse remains cyclical. The disagreement in the stock is therefore less about whether Atlas is a strong business than about how much of the semiconductor recovery, future acquisition compounding and margin resilience an A-share buyer should pay for in advance.

Vertical history and financial record

Atlas Copco began as AB Atlas in Stockholm in 1873. Sweden was building railways and needed rolling stock and industrial equipment; banker André Oscar Wallenberg was a driving force behind the venture, while engineer Eduard Fränckel led the technical side. The initial model bore little resemblance to the current company: heavy railway machinery was capital intensive, domestically focused and exposed to infrastructure investment. A severe Swedish downturn in the 1880s pushed Atlas close to failure. New Wallenberg capital and a reorganization as Nya Atlas forced a practical strategic pivot into steam engines, boilers and other machinery with wider industrial demand.

The decisive product discovery came almost by accident. Atlas used pneumatic tools inside its own plants, learned to maintain and improve them, then started selling them externally around 1901. That path contains an early version of the later Atlas formula: a technically demanding product first proves itself under operating conditions, then customer support and application knowledge become as important as manufacturing the unit.

The 1917 combination with Diesels Motorer created Atlas Diesel. Diesel engines initially appeared to offer a larger industrial future, but pneumatic equipment produced better economics and stronger competitive positioning. Atlas stopped manufacturing diesel engines in 1948. That decision matters more than the name change: management abandoned a large legacy line rather than protect historical identity, concentrated capital where it had differentiated engineering, and set the company on the compressed-air path that still generates its largest earnings pool.

The next phase was internationalization rather than product reinvention. Atlas bought Belgian compressor maker Arpic Engineering in 1956 and adopted the Atlas Copco name. The company built direct sales, distribution and service organizations abroad, then used acquisitions to enter adjacent compressor, tool and assembly niches. Its post-war expansion occurred while manufacturing globalized and compressed air became basic industrial infrastructure. This created a valuable characteristic: compressor demand follows industrial investment, but servicing existing compressors follows the installed base.

Atlas has traded in Stockholm since 1920, making the modern concept of an IPO prospectus, offer price and first-day valuation poorly applicable. The company's contemporary primary-history material confirms the 1920 listing but does not provide a reliable modern-format IPO price, capital-raised figure or listing valuation. I therefore leave those data points blank rather than reconstructing them from potentially inconsistent historical currency and capitalization records.

A third stage came as Atlas used its balance sheet and decentralized operating model to build specialist franchises. Industrial tools broadened the company's exposure to automotive and assembly. Vacuum was transformed through acquisitions including Edwards and Leybold. By 2016 Atlas was preparing Vacuum Technique as a standalone business area, operational from January 1, 2017. That move separated a semiconductor-sensitive, high-technology capital-equipment franchise from the slower and larger compressor operation, making the group's internal cycle more visible to investors.

The fourth stage was portfolio surgery. Atlas announced the split that became Epiroc because mining equipment had developed into a sufficiently distinct business with its own customers, cycle and capital-market peer group. From January 1, 2018 Mining and Rock Excavation was classified as discontinued operations. Atlas shareholders received Epiroc shares class for class, and Epiroc listed on June 18, 2018.

The separation changed the quality investors were buying. The 2017 information prepared for the transaction showed continuing Atlas revenue of MSEK 85,522 and operating profit of MSEK 18,251, an operating margin above 21%, while Epiroc revenue was MSEK 31,440 and operating profit MSEK 5,949. Atlas excluding Epiroc had generated an average operating margin of roughly 20.8% and average ROCE of about 28% over the stated 2010–2017 historical period.

Historical comparisons before that boundary need care. Older Atlas figures include a mining-equipment business that now belongs to Epiroc, while 2017 figures were later restated for continuing operations and IFRS 15. The cleanest long-run operating comparison for today's Atlas begins in 2018.

From 2018 onward Atlas became more visibly an acquisition-and-installed-base compounder. The four business areas were internally decentralized; acquisitions were normally small enough to fit into existing divisions rather than require group-scale integration. Atlas recorded 18 acquisitions in 2019 and 30 in 2022. The number itself does not prove value creation, but sustained ROCE in the 20s despite repeated purchases argues that the portfolio has not simply been grown by overleveraging.

The pandemic provided the most useful recent stress test of operating resilience. Revenue declined 4% in 2020 to MSEK 99,787, operating profit fell to MSEK 19,146, and the margin compressed from 21.1% in 2019 to 19.2%. ROCE fell from 30% to 23%. That was a genuine downturn, yet the company remained solidly profitable and lightly leveraged. Compressor Technique alone still produced a 22.5% operating margin and 79% ROCE in 2020.

The recovery that followed was unusually strong. Revenue increased from MSEK 110,912 in 2021 to MSEK 141,325 in 2022, then to approximately MSEK 172,664 in 2023. Supply constraints caused working capital to absorb cash in 2022 even while accounting earnings expanded. By 2024 revenue reached MSEK 176,771 and operating cash flow a record MSEK 30,981.

Sources: Atlas Copco annual and quarterly primary reporting; 2025 EPS is consistent with MSEK 26,425 net income and the prevailing weighted share count.

The table shows two different stories. Earnings quality is high over the cycle, but cash flow can lag during fast growth. The five-year cumulative OCF/net-income ratio is about 0.93x, rather than comfortably above one. Most of that gap comes from 2022–2023; 2024–2025 returned to approximately full cash conversion. The quality thesis therefore survives, but a valuation model that assumes every krona of accounting profit becomes immediately distributable cash would be too generous.

2025 was the first recent warning against treating the compounder as linear. Revenue fell to roughly MSEK 168,343 and net profit to MSEK 26,425. Q2 2025 disappointed: organic orders fell 1%, revenue fell 2% organically, and the A share dropped about 7.6% after results as adjusted operating profit missed expectations. That episode is useful because the business did not structurally deteriorate; the market simply removed part of the premium when the cycle and earnings failed to validate it.

The market's long-run repricing is visible through share-price anchors. The A share closed 2016 at SEK 277.50 before the later four-for-one split, equivalent to roughly SEK 69.38 on today's share-count basis, though that price still embeds the mining business distributed as Epiroc in 2018 and is therefore not comparable with the later anchors on a like-for-like basis. It closed 2019 at SEK 373.60 pre-split, roughly SEK 93.40 adjusted. By the end of 2022 the post-split A share was SEK 123.10. It was SEK 208.80 on September 8, 2026. The 2022 four-for-one split affects nominal historical prices but created no economic value by itself.

Using earnings available at those dates illustrates the multiple shift. The 2022 close against SEK 4.82 of EPS was about 25.5x earnings. The 2019 split-adjusted price against split-adjusted earnings of roughly SEK 3.40 was around 27.5x. The current A share is roughly 38x trailing earnings using the latest approximately SEK 5.46 trailing EPS. Exact daily historical-percentile data are not available from the primary source set used here, so I would describe today's valuation as roughly the upper quintile of recent post-Epiroc valuation experience rather than assign a false-precision “87th percentile.”

Some re-rating is deserved. Atlas today has a larger Vacuum franchise, a bigger service mix, a cleaner portfolio after Epiroc and a demonstrated acquisition architecture. The question at SEK 208.80 is whether the premium rewards those changes or capitalizes too many future years of them.

Management succession has so far preserved continuity rather than reset the strategy. Vagner Rego became CEO on May 1, 2024 after joining Atlas in Brazil as a trainee engineer in 1996. His career included Compressor Technique service roles, leadership of the Compressor Technique Service division and ultimately presidency of the Compressor Technique business area. That background is unusually aligned with the economics investors value most: installed-base service, decentralized operating responsibility and compressor profitability.

Ownership reinforces that continuity. Investor AB's 22.4% voting interest gives the group a large shareholder with a time horizon much longer than a quarterly earnings cycle. The advantage is strategic patience: management can build local production, service capacity and bolt-on acquisitions without optimizing every quarter for consensus. The constraint is equally clear. A minority A-share holder has substantially less ability to alter governance or capital allocation when Investor and other long-duration Swedish institutions support the incumbent direction.

Capital returns coexist with reinvestment. Atlas historically targets ordinary dividends around half of earnings and has periodically used extra distributions and redemption mechanisms when the balance sheet carried surplus capital. The 2022 transaction combined an ordinary dividend, an extra redemption distribution and the four-for-one share split. This history suggests that low leverage is deliberate financial policy rather than a temporary absence of investment opportunities.

At Q2 2026 net indebtedness was MSEK 22,305, net debt/EBITDA about 0.5x and net debt/equity 21%, with average interest-bearing-debt maturity of about 4.1 years. That balance sheet gives Atlas room to continue bolt-ons without making an economic downturn into a refinancing event.

The acquisition machine nevertheless deserves a stricter test under Rego because higher purchase prices can erode the return on capital even when reported EPS rises. In H1 2026 structural change from acquisitions contributed 4 percentage points to orders and 3 to revenues. Q1 disclosed purchases including Cypress Equipment Rental, Zind Verfahrenstechnik and Centroar; Q2 added businesses including LACO Technologies and the Stainless Design Concepts operation in Vacuum. LACO alone brought approximately MSEK 247 of prior-year revenue.

The acquisition list is too extensive for each tiny distributor to deserve equal weight in an investment argument. The better scorecard is post-deal organic growth, margins, cash conversion and ROCE. On that test Atlas remains strong, but ROCE's move from 30% in 2023 toward the mid-20s is worth monitoring.

Business model, industry and peers

The economic machine has four different front ends and one common back end. Equipment is designed for applications where downtime, energy consumption, contamination, torque accuracy or process stability cost customers much more than the purchase price of the equipment. Once installed, Atlas gains recurring access through spare parts, maintenance, optimization and replacement. That installed-base relationship is strongest where the process is mission critical: compressed air in a factory, vacuum in semiconductor fabrication, tightening quality in an automotive line, or portable power on a construction or infrastructure site.

Compressor Technique remains the core franchise. In Q2 it booked MSEK 21,704 of orders and MSEK 19,653 of revenue; organic orders increased 19% while organic revenue increased 3%. Operating profit was MSEK 4,745 and margin 24.1%, compared with 25.0% a year earlier. Industrial compressors and service were solid; gas and process demand was particularly strong. Acquisitions diluted the margin.

Vacuum Technique is the growth engine. Its products create and control vacuum environments in semiconductor manufacturing, scientific applications and industrial processes. Q2 orders rose to MSEK 14,409 and revenues to MSEK 10,531, with 59% and 19% organic growth respectively. The huge gap is precisely why order conversion rather than order growth alone should drive a forecast.

Industrial Technique booked MSEK 6,953 of orders and MSEK 6,643 of revenue in Q2, with both organically up 11%. Operating profit was MSEK 1,328 and margin 20.0%, up from 17.1%. Its automation, assembly and joining exposure makes automotive manufacturing important, but machine vision and general industrial applications broaden the base.

Power Technique recorded MSEK 8,097 of orders and MSEK 8,336 of revenue; organic growth was 14% for orders and 9% for revenue. Operating profit was MSEK 1,459 and margin 17.5%. Specialty rental acquisitions materially increased reported growth, showing why this business area's organic and structural components need to be separated.

The business-area sum differs slightly from group order/revenue totals because of group eliminations. Sources: Atlas Copco Q2 2026 report and business-area disclosures.

The profit hierarchy is more stable than the growth hierarchy. Before central costs and eliminations, Compressor supplied roughly half of the four business areas' Q2 operating profit, Vacuum around 23%, Industrial around 14% and Power around 15%. A semiconductor boom can therefore move group growth materially without making the entire group a semiconductor-equipment company.

The 38% service share is the first genuine moat. A compressor or vacuum system is an operating asset rather than a discretionary gadget. Customers value uptime, energy efficiency, contamination control and a service technician who understands the installed configuration. In Vacuum, Atlas has explicitly described the strategic move from reactive repairs toward planned and predictive service and recorded faster historical service growth than equipment growth in its long-run business-area presentation.

The moat is strongest where Atlas combines a large installed base with application knowledge and local service, because the customer is purchasing process reliability over the equipment's life rather than a one-off machine.

This is also why the precise service margin is less important than the observed consolidated downside behavior. Atlas does not reveal a service margin, but the group stayed near a 19% operating margin through the 2020 industrial shock and returned above 21% afterward. Service cannot be given sole credit; pricing, decentralized cost control and product mix contributed too. Yet the evidence is consistent with the resilience thesis.

The second moat is decentralization combined with a global local-service footprint. Atlas operates through divisions with operational responsibility rather than forcing all businesses through a centralized product platform. Its current corporate material describes a presence in about 180 countries and operations in roughly 70, with a “local for local” strategy of producing where it sells and sourcing where it produces. That architecture reduces lead times and tariff exposure while leaving division managers accountable for customer economics.

The third is engineering specialization. Semiconductor vacuum is especially demanding because contamination, particle control, process chemistry and uptime affect fab yield. Compressor efficiency matters because lifetime electricity cost can dwarf the upfront purchase price. Tightening systems in automotive assembly need traceability and repeatability. These are markets where the technically cheapest product does not automatically minimize customers' total cost. Atlas's margins and service retention are indirect evidence that it captures part of that customer value.

The fourth is the acquisition process. A decentralized group can buy a small local distributor, specialist pump maker or technology company and leave customer accountability close to the acquired operation. H1 2026 structural growth of 3–4% shows that the machine remains meaningful. The moat would weaken if acquired growth began to require group-level restructuring, if goodwill rose faster than earnings, or if ROCE kept falling despite accelerating EPS.

Brand by itself is the least convincing moat. Atlas owns numerous brands, but an industrial customer does not stay with a vacuum platform because a logo is famous. The economic defense is the installed system, application engineering, service network and switching cost. Semiconductor demand itself is likewise no moat: every competent vacuum supplier benefits from rising fab capex.

Cost behavior follows the same logic. Equipment manufacturing has raw-material, component and assembly costs that vary with volume; engineering, sales, service infrastructure, R&D and factories contain fixed costs. Acquisitions add amortization and integration expenses. A downturn therefore creates operating leverage in equipment, but service revenue and the decentralized ability to adjust local cost bases reduce the group-level hit. The 2020 result provides the empirical boundary: a 4% revenue decline caused operating profit to fall more sharply, but margin remained 19.2%.

Atlas is not an extraordinarily capex-heavy manufacturer. Its balance-sheet risk is more likely to come from working capital during fast growth and from acquisition consideration than from building gigantic owned production plants. The 2022 cash-flow shortfall illustrates the first risk; the steady acquisition program illustrates the second.

The industrial cycle is a composite rather than a single index. Compressor responds to broad manufacturing capex and utilization. Vacuum carries a semiconductor and scientific-capex cycle. Industrial Technique carries automotive and general assembly cycles. Power has construction, infrastructure and rental exposure. Service follows installed-base utilization and tends to move less violently than new equipment.

In semiconductor equipment, the cycle is currently favorable. SEMI's latest 2Q26 World Fab Forecast projected global equipment spending up 24% in 2026 to a new high, followed by another 9% in 2027. Earlier SEMI forecasts were already calling for double-digit 300mm spending increases, and the later forecast became more bullish. That revision is itself useful: the demand environment strengthened through the forecast cycle.

TSMC provides a customer-side confirmation. In April 2026 it said its capital budget would be toward the high end of its planned range and described additional N3 capacity in Taiwan, Arizona and Japan; by Q2 it reported 33.7% US-dollar revenue growth and expected 2026 revenue growth slightly above 40%. That does not translate one-for-one into Atlas orders, but it supports continued demand for the vacuum infrastructure surrounding advanced semiconductor production.

Geopolitics cuts both ways. Semiconductor self-sufficiency policies are leading to more geographically dispersed fabs, creating more equipment-installation opportunities and fitting Atlas's local-for-local strategy. Export controls can simultaneously alter which fabs receive advanced equipment and introduce compliance friction. For Atlas the likely long-run effect is more regionalized supply chains rather than a simple reduction in vacuum demand. That is an inference from semiconductor-capacity regionalization and Atlas's own manufacturing strategy.

Tariffs and energy costs are broader industrial risks, but Atlas's direct geographic manufacturing footprint reduces some border exposure. The company's low financial leverage means higher rates matter more through its equity multiple and customer capex than through its own interest bill.

The closest horizontal comparison is not one company. Atlas straddles several niches, so a useful peer portrait needs different references.

Ingersoll Rand is the cleanest compressor and aftermarket analogue. It also operates a decentralized industrial portfolio, uses acquisitions extensively and builds recurring aftermarket sales around installed flow equipment. Q2 2026 aftermarket was about 36% of revenue, remarkably close to Atlas's 38% group service mix; Ingersoll Rand reported roughly 2% organic order growth and 4% organic revenue growth, with adjusted EBITDA margin around 25.4%. Compared with Atlas, it offers purer compressor/flow exposure but lacks a Vacuum franchise of comparable semiconductor sensitivity.

That explains why customers pick the two differently. Ingersoll Rand competes directly in industrial compression and flow creation and has built its own service installed base. Atlas's advantage is a broader mix of premium industrial niches and a deeper semiconductor-vacuum growth option. Ingersoll's relative advantage is that its portfolio is easier to read as a flow-equipment compounder and currently carries less of the semiconductor-cycle expectation embedded in Atlas.

Ebara is the most informative listed Vacuum comparator. Its Precision Machinery business sells dry vacuum pumps and semiconductor systems. In Q1 2026 Precision Machinery orders more than doubled year-on-year and revenue rose roughly one-third, driven by AI-related demand, higher customer-factory utilization and production-expansion capex. By H1 Ebara said group orders, revenue and operating profit had reached record first-half levels, with Precision Machinery a major contributor, and it was expanding vacuum-pump overhaul service in Taiwan and South Korea as the installed base grew.

Ebara's acceleration is important evidence against the idea that Atlas's +59% Vacuum orders came only from company-specific share gains. The semiconductor equipment tide is lifting multiple vacuum suppliers. Atlas can still gain share, but the burden of proof is higher: a genuine share-gain thesis would need Atlas Vacuum orders to continue outperforming relevant peer equipment orders after the industry normalizes.

Sandvik is a capital-market and organizational comparator more than a product twin. It is another decentralized Swedish engineering group with premium tools, mining exposure, service and acquisitions. Q2 2026 organic orders rose 17%, organic revenue 23% and adjusted EBITA margin reached 22.6%. Sandvik therefore shows that Atlas is not the only high-margin Nordic industrial entering the second half of 2026 with strong order momentum.

The differences matter. Sandvik has far more mining and machining exposure, while Atlas has compressors and semiconductor vacuum. Sandvik's Q2 revenue growth exceeded its order growth; Atlas's order growth exceeded revenue growth by a wide margin. The two stocks therefore represent different points in order conversion even while both are enjoying strong end demand.

Alfa Laval provides another useful Swedish quality-industrial cross-check. Q2 organic orders rose 29%, while organic sales rose 6% and adjusted EBITA margin was 17.0%. The order-versus-revenue pattern resembles Atlas much more than Sandvik's and shows how strong project/capex order intake can precede recognized revenue. Atlas earns a higher operating margin, reflecting a different product and service mix rather than simply superior execution.

Selected trailing market multiples reinforce the premium. Atlas A trades at roughly 38x trailing earnings. Same-period market data put Ingersoll Rand around 31–32x, Alfa Laval around 30x and Ebara around 26x. These are not perfectly homogeneous accounting bases, and the Atlas A figure includes the high-vote premium, but the ranking is clear enough: the market assigns Atlas a substantial quality-and-growth premium to relevant industrial references.

These values are market-data snapshots rather than standardized consensus forward estimates. Atlas multiple is explicitly for ATCO A at SEK 208.80.

The premium has a business basis: higher long-run ROCE, a 38% service component, low leverage and exposure to semiconductor vacuum layered on top of the Compressor cash engine. It is nevertheless large enough that execution must stay superior. Peer valuation cannot make Atlas cheap simply because other quality industrials are expensive.

Atlas's niche can therefore be defined precisely. It is the premium installed-base operator across several relatively narrow industrial technologies, with Compressor harvesting the largest mature profit pool and Vacuum taking a growing share of semiconductor sub-fab economics. A new generic machinery entrant would struggle because distribution and installed service take decades to build. The credible threats come from established specialists such as Ingersoll Rand in compression and Ebara in semiconductor vacuum, where customers already trust the competing technical stack.

Current fundamentals and capital-market narrative

The last four quarters show the turn more clearly than a single Q2 comparison.

Sources: quarterly Atlas Copco reporting. Organic order growth and organic revenue growth are deliberately separate columns.

The sequence says the business reached an order trough around the second half of 2025 and began reaccelerating before revenue did. Organic orders went from 0% to +4%, +5% and +26%; organic revenue went from +1% to 0%, +3% and +8%. Q2 therefore contains both a real inflection and a risk of overinterpretation.

Reported growth tells another useful part of the story. Q1 2026 orders were down 3% reported even though they rose 5% organically, and revenue was down 5% despite 3% organic growth. H1 currency effects were roughly -8 percentage points on orders and -7 on revenue; acquisitions added +4 and +3 points. By Q2 the organic impulse was so large that reported orders still rose 27% and revenue 9% despite currency.

This is why an investor forecasting from reported revenue alone would have missed the operational turn, while an investor annualizing Q2 orders would overshoot it. The correct bridge begins with business-area organic orders, then applies conversion timing, acquisitions and currency separately.

Q3 2025 was still mixed. Orders were MSEK 40,517, down 4% reported but flat organically; revenue was MSEK 41,621, down 3% reported and up 1% organically. Operating cash flow was MSEK 7,330. Service demand remained constructive.

Q4 improved on orders but not on revenue. Organic orders increased 4%, revenue was flat organically, reported operating margin was 19.8% and operating cash flow MSEK 6,777. Management's near-term message remained measured rather than forecasting a surge.

Q1 2026 was the first clearer semiconductor signal. Organic orders rose 5% and organic revenue 3%; Vacuum demand strengthened while Compressor industrial demand was roughly unchanged and automotive remained weak. Services performed well across business areas and regions. Operating cash flow was unusually soft at MSEK 4,355, against MSEK 6,575 a year earlier.

Q2 broadened the recovery. All four areas posted double-digit organic order growth; regional order growth in local currencies was particularly strong in North America and Asia/Oceania. Management nevertheless guided only that near-term customer activity should “remain at the current level.” That wording matters: management did not tell investors to annualize the 26% organic order growth.

The market is trading the order reacceleration first and the eventual revenue conversion second. That interpretation is supported by the 1.13x Q2 book-to-bill, Vacuum's 59% organic order increase, the roughly 26% A-share year-to-date rise and the improving semiconductor-capex environment.

Q2 also beat the market more decisively on orders than on earnings. Secondary consensus data showed MSEK 50,951 of orders versus about MSEK 44,961 expected, a roughly 13% beat; revenue beat by around 2%. Reported operating profit of MSEK 9,249 was about 0.5% below that consensus measure, while adjusted operating profit was around 2% ahead. The implication is straightforward: the surprise was the future workload, not a sudden change in current-quarter profitability.

That is the present bull/bear divide.

The bull case starts with breadth. Vacuum is not carrying a shrinking rest of group: Compressor organic orders rose 19%, Industrial 11% and Power 14% in Q2. A broad order recovery raises the odds that the group can grow through 2027 even if Vacuum's 59% normalizes rapidly.

The second bull point is margin resilience. Reported Q2 group operating margin held at 20.6% and the adjusted margin improved to 21.0% from 20.4%, which Atlas attributes mainly to volume, price and mix, with a smaller favorable currency effect, against dilution from recent acquisitions and the spending to ramp Vacuum capacity. The currency help is worth naming, because the same translation effect cut about 3 percentage points from reported revenue growth: it flatters the margin line while it depresses the top line. Vacuum's own margin rose to 21.0% as revenue caught up. If order conversion lifts utilization, group earnings can grow faster than revenue without requiring an unprecedented margin.

The third is external semiconductor evidence. SEMI has revised capex expectations upward and TSMC is expanding advanced-node capacity. Ebara's own Precision Machinery orders have surged. Vacuum's growth is therefore supported by an industry upcycle with multiple independent signals.

The fourth bull argument is the installed base. Service is 38% of group revenue and has continued to grow even when equipment orders were mixed. This makes a semiconductor-led revenue upswing additive to a recurring base rather than the only source of earnings.

The bear case begins at almost the same facts. A 59% Vacuum organic-order increase invites extrapolation exactly when semicap orders are most cyclical. Ebara's equally dramatic acceleration says the current boom is industry-wide, so Atlas may be experiencing strong beta to semiconductor capex rather than a permanent step change in share.

The second bear point is the conversion gap. Q2 orders exceeded revenue by almost MSEK 6,000, yet Atlas gives investors no current group backlog amount or 12-month conversion schedule. That prevents a precise bridge from the order surprise to 2027 revenue. Higher orders are directionally favorable; the amount and timing of earnings remain uncertain.

Third, cash conversion over five years is 0.93x rather than greater than one. The recent normalization is reassuring, but another combination of rapid growth, inventory buildup and acquisition spending could repeat 2022–2023.

Fourth, the A share already trades at roughly 38x trailing earnings and a meaningful premium to other high-quality industrial names. A company can meet its earnings forecast and still generate a poor shareholder return if the valuation multiple normalizes.

There is also a share-class issue that is easy to overlook. Because A and B have the same economic claim, the SEK 208.80 A price versus SEK 180.70 B price implies that an A investor accepts a materially lower earnings and cash-flow yield in exchange for the additional vote and market attributes. Using the same group per-share earnings, the B-share earnings multiple would be roughly 13% lower. Investors indifferent to voting rights should consciously price that difference rather than assume the two Atlas tickers are interchangeable.

The next scheduled earnings event is Q3 2026 on October 22, 2026, with the report planned around 12:00 CEST and the conference call at 14:00 CEST. The next print will matter disproportionately because it provides the first test of whether Q2 Vacuum orders were a concentrated release of customer projects or the start of several strong quarters.

Valuation, risk and tracking

A valuation of Atlas needs to begin with cash pass-through because a premium industrial should not receive a premium multiple simply for reporting noncash earnings.

Atlas-reported operating cash flow totaled MSEK 117,446 over 2021–2025, against MSEK 125,887 of net income, producing a cumulative OCF/net-income ratio of about 0.93x. The annual sequence was approximately 1.07x, 0.73x, 0.83x, 1.04x and 1.01x. The weak middle years coincide with the large 2022 working-capital absorption and expansion investments; the last two years show normalization.

Atlas does not publish a reliable audited split of physical capex into “maintenance” and “growth.” Forcing a precise split would therefore create false accuracy. The company's reported operating-cash-flow KPI already captures investment cash flows more conservatively than a simple net-income measure, so I use reported cash flow as the principal owner-earnings cross-check rather than subtract an invented maintenance-capex number a second time.

On 2025 operating cash flow of MSEK 26,796 and the Q2 2026 net share count, cash flow was approximately SEK 5.49 per economic share. At the A-share price of SEK 208.80 that corresponds to a price/cash-owner-earnings proxy near 38.0x and a yield around 2.6%. The FY2025 EPS multiple is about 38.5x; using latest trailing EPS of roughly SEK 5.46 gives about 38.2x. The cash and accounting valuations therefore differ by roughly one percent, far too little to justify abandoning earnings-based valuation.

That is reassuring on earnings quality and demanding on price.

The rate comparison makes the price tension visible. Sweden's September 2, 2026 auction of a government bond maturing in October 2036 cleared at an average yield of 3.1495%. Atlas A's current owner-earnings yield of about 2.6% is below that risk-free nominal yield before assigning any equity-risk premium. An investor therefore needs meaningful future growth and/or sustained premium valuation merely to justify today's price.

A reverse owner-earnings calculation illustrates the burden. Starting from approximately SEK 5.49 per share, an 8% required equity return and 2.5% terminal growth would require owner earnings to compound at roughly 11.6% annually for ten years to reproduce SEK 208.80. At a more forgiving 7% required return, the implied decade growth falls to roughly 8.8%. These are model outputs rather than consensus forecasts. They show that the A price embeds more than a normal mature-industrial trajectory.

Historical anchors point in the same direction. Atlas A traded around 25.5x FY2022 EPS at the 2022 year-end and around the high-20s on a split-adjusted basis around 2019. The current approximately 38x multiple incorporates the stronger Vacuum mix and post-Epiroc quality, but also makes the investment materially more dependent on long-duration execution.

Peer valuation does not rescue the price. Ingersoll Rand trades around 31–32x trailing earnings, Alfa Laval near 30x and Ebara in the mid-20s on the same broad market-data window. Atlas deserves some premium; a 20–50% premium over those references requires persistent superior returns, growth and capital allocation.

For absolute valuation I use a 2027 earnings framework. The revenue assumptions deliberately do not annualize Q2 2026 organic orders. Conservative revenue assumes only modest progression from the 2024–2025 base; the optimistic case requires strong Vacuum conversion plus renewed growth across the other three areas. Net financial expense is normalized at roughly MSEK 1,000 and the model uses a 23% effective tax rate and about 4.879bn shares.

The scenario inputs are research assumptions; historical anchor data come from Atlas's filings and current price from the 2026-09-08 A-share close.

The conservative case assumes the Q2 order surge converts only partially before semiconductor demand normalizes, general-industrial growth stays modest and the group margin slips below its recent 20%–21% range. A 30x multiple still grants Atlas a premium to an ordinary machinery company. Its SEK 156 central value is therefore conservative without assuming structural damage.

The base case assumes the existing order excess converts over subsequent quarters, Vacuum growth normalizes sharply from +59% orders but remains positive, Compressor continues to earn around its historical premium margin, and acquisitions add low-single-digit structural growth without materially harming returns. The resulting SEK 200 value is close to the current A price.

The optimistic case requires a durable semicap expansion, broad industrial demand and group operating margin around 22%, near the upper end of recent experience. Even then, a 37x multiple is required to reach roughly SEK 249. That scenario is achievable only if Atlas keeps its quality premium while delivering the growth that currently justifies it.

Using a conservative recurring cash distribution of roughly SEK 3 per share over the next 12 months as a return placeholder rather than assuming extraordinary distributions recur, those central cases imply approximately -24% one-year total return in the conservative case, -3% in the base case and +21% in the optimistic case. These are valuation-scenario outputs, not investment advice. Atlas's actual distributions can differ.

For a 3–5-year holder, the arithmetic improves only if earnings continue compounding. A base path in which EPS grows about 7% annually for five years but the terminal multiple normalizes into the low-to-mid 30s produces roughly mid-single-digit annualized returns after ordinary dividends. The company can therefore remain excellent while the stock delivers merely acceptable returns.

A reverse multiple test is harsher. The most fragile base-case assumption is the 34x terminal A-share P/E. Cutting that assumption to 70% leaves 23.8x; applied to SEK 5.89 base EPS, the valuation falls to about SEK 140 per A share. This is why valuation risk is a permanent-loss mechanism rather than short-term volatility.

The independent margin-of-safety test produces an unambiguous answer. Current A price of SEK 208.80 is about 34% above the conservative central value of SEK 156. The discount to conservative value is therefore negative.

If earnings stayed flat for three years at roughly the current SEK 5.46 trailing level and the market eventually priced the company at 30x, the terminal share price would be about SEK 164. Adding roughly SEK 9 of cumulative ordinary dividends gives a three-year annualized return near -6%. That is below the approximately 3.15% Swedish ten-year government yield used above. There is no margin of safety at this buy price.

Margin-of-safety sufficiency verdict: none.

This is a good-company/bad-initiation-price case under a strict margin-of-safety discipline. That does not imply the A share must fall. It means the investor presently depends on successful growth and continued multiple support instead of buying those outcomes at a discount.

The first permanent-loss risk is a semiconductor-capex reversal. Probability is medium and impact high. The observable signal would be a reversal in SEMI fab-spending forecasts, cuts in major foundry/memory capacity plans, Vacuum book-to-bill below 1 and negative Vacuum organic orders. The transmission path is direct: fewer fab projects cut Vacuum orders first, revenue follows with a lag, factory utilization and mix pressure Vacuum margin, and the equity multiple contracts when the market stops treating Vacuum as a structural AI beneficiary. Current SEMI and customer evidence points upward, which is precisely why the downside from a reversal could be large.

The second is multiple compression. Probability is medium-high at an A-share trailing P/E around 38x; impact is high. The observable indicators are the Swedish long yield, the A-share earnings yield and the valuation gap versus peers. A move from 38x toward 28–30x with unchanged earnings would cut the share price by roughly one-quarter. No operational failure is required.

The third is acquisition-return erosion. Probability is medium and impact medium-to-high. Atlas has the balance sheet to keep purchasing companies, but success should be judged by ROCE and cash conversion rather than deal count. A sustained ROCE drop below 20%, acquisition-driven growth accompanied by falling organic growth, or two years of OCF/net income below 0.7x would suggest capital is being deployed at weaker incremental returns.

The fourth is broad industrial recession. Probability is medium and impact medium. Compressor's service base would cushion earnings, but new compressors, tools and power equipment are still capex products. The 2020 experience suggests a deep shock can push the group margin below 20%, though the balance sheet makes a solvency event unlikely.

The fifth is competitive share loss in high-value vacuum. Probability is low-to-medium, impact high if sustained. Ebara is already seeing strong AI-related Precision Machinery demand and expanding its service infrastructure. A period in which Ebara and other vacuum specialists continue growing orders while Atlas Vacuum turns negative would challenge the assumption that Atlas has maintained its competitive position through the latest process transition.

Governance is a lower-probability permanent-loss risk. Investor AB's 22.4% voting interest stabilizes ownership but reduces the practical influence of dispersed holders. No evidence in the primary sources reviewed points to a current accounting scandal, material related-party deterioration or balance-sheet governance problem; Q2 states there had been no significant change in related-party relationships versus the 2025 annual report.

The positive catalysts for the next 12 months are consequently measurable rather than thematic. Continued Q3/Q4 book-to-bill above one, Vacuum revenue catching up without a drop in its 21% margin, Compressor maintaining its mid-20s margin, service growth across all four areas, and continued upgrades in semiconductor fab-spending forecasts would turn more of today's order narrative into realized earnings.

Negative catalysts are the mirror image: Vacuum orders reverting to negative growth before revenue catches up, group book-to-bill below one, a broad industrial-order slowdown, operating margin below 19%, or cash conversion deteriorating during the growth ramp.

Current/reference data come from Atlas Q2 2026, historical annual reports, current A-share market data and Atlas's financial calendar.

The dashboard should be read in order. Vacuum orders lead. Group book-to-bill tells whether they are translating into a larger future workload. Organic revenue shows conversion. Margin reveals the profitability of that conversion. Cash flow and ROCE then test whether growth is worth the capital consumed. The P/E decides how much of the result the shareholder already paid for.

The principal research uncertainties are concentrated rather than broad. Atlas does not disclose a current consolidated backlog balance or its 12-month conversion schedule; it does not publish a separate group service margin; it does not separate maintenance and growth capex in the precise owner-earnings format requested; current primary disclosure does not supply a single cross-class average-liquidity ratio; and an exact ten-year daily valuation-percentile series was not available from the primary materials used here. None of those gaps justifies filling the report with modeled “facts.” They are the variables where the investment case necessarily carries a wider confidence interval.

Cross-synthesis and conclusion

Vertically, the capability Atlas Copco has proven over more than a century is portfolio adaptation without losing operating discipline. The company started in railway equipment, nearly failed, pivoted into machinery, abandoned diesel engines, globalized around compressed air, used service to deepen installed-base economics, built industrial tools and vacuum through acquisitions, and eventually separated mining into Epiroc. The common thread is not loyalty to a particular machine. It is the ability to identify industrial niches where application engineering and post-sale service earn attractive returns, then leave operating accountability close to the customer.

That makes management capability a larger explanation of historical success than any single era tailwind. Atlas certainly benefited from post-war industrialization, global manufacturing investment, automotive automation and more recently semiconductor capex. Yet the product set changed repeatedly while returns on capital stayed unusually high. The 2018–2024 post-Epiroc period is particularly useful because it removes mining from the comparison: ROCE stayed mostly in the high-20s or around 30% through a pandemic, supply constraints and several acquisition waves.

The success factors remain present. Service is a larger share of revenue than several years ago. Compressor still earns a mid-20s operating margin. The balance sheet is lightly geared. Vacuum has a larger installed base and a service strategy built around planned and predictive maintenance. Atlas continues to acquire small specialist companies rather than betting the franchise on giant transformative deals.

Horizontally, Atlas's advantage is that no major listed peer reproduces the same portfolio. Ingersoll Rand comes closest in compression and aftermarket but lacks a Vacuum franchise of the same importance. Ebara competes directly in dry vacuum and semiconductor equipment but does not have Atlas's Compressor profit engine. Sandvik matches the decentralized Nordic industrial culture but has mining and machining cycles. Alfa Laval resembles Atlas in premium process equipment and recent order/revenue divergence but earns lower group margins.

This portfolio structure makes Atlas more resilient than a pure semiconductor supplier while giving it more semiconductor upside than a pure compressor manufacturer. The same structure complicates valuation because the market can assign a “quality compounder” multiple to Compressor and simultaneously capitalize the fastest part of a semicap upcycle in Vacuum.

The current valuation is doing both.

At SEK 208.80, roughly 38x trailing A-share earnings, the stock is well above the earnings multiples observable at Ingersoll Rand, Alfa Laval and Ebara. Its cash-owner-earnings yield is around 2.6%, below Sweden's recent roughly 3.15% ten-year government-bond yield. That spread can make economic sense only if future earnings grow materially and the market continues to treat Atlas as a superior long-duration asset.

The market is rewarding proven quality while pre-spending part of the semiconductor recovery.

What the market may still be underestimating is the breadth of the order turn. Q2 organic orders did not rise 26% because Vacuum alone offset collapsing businesses; all four areas grew double digits. If Compressor's +19% order growth and Industrial/Power's double-digit growth persist even as Vacuum normalizes, 2027 revenue could be stronger than a semiconductor-only reading implies.

What the market may be overestimating is the duration of the exceptional order rate. Ebara's surge and SEMI's booming fab forecasts establish a cycle. They do not establish a permanent 50%-plus Atlas Vacuum growth rate. Atlas's own long-run Vacuum data point toward mid-teens equipment/service growth over a decade, far below Q2's 59% order number.

The 12-month variable that matters most is order conversion. Q2 generated almost SEK 6bn more orders than revenue. Q3 and Q4 need to show that revenue growth catches up without Vacuum margin giving back its recent improvement. An order slowdown is acceptable if it reflects normalization from an exceptional comparison; a simultaneous slowdown in orders and failure of revenue to accelerate would be far more damaging.

Over three years the crucial variable becomes capital allocation. Atlas can create value with 3–4% annual structural growth from acquisitions only if acquired companies earn attractive incremental returns. ROCE staying above the low-to-mid 20s and cash conversion near one would confirm the model. A persistent move below 20% ROCE would change the interpretation from “compounding through acquisitions” to “buying growth at progressively lower returns.”

Over five years the most important variable is the economic durability of service. The 38% group service share should rise or at least remain stable as the installed equipment base expands. If service begins growing below equipment for several years, customer stickiness or competitive intensity may be changing. Vacuum's service-led strategy makes that test especially important as today's semiconductor equipment eventually moves into the installed fleet.

Bull reasons

Atlas has five specific reasons the long-run compounder thesis can continue.

First, Q2 2026 organic orders grew double digits in every business area, with group book-to-bill at 1.13x, so the growth inflection is broader than semiconductor vacuum alone.

Second, service accounted for 38% of 2025 revenue and continued growing across business areas and regions in Q2, providing a recurring installed-base layer beneath equipment cycles.

Third, Compressor generated a 24.1% Q2 operating margin while Vacuum improved to 21.0%, showing that the group is combining its highest current growth with strong, rather than collapsing, profitability.

Fourth, external semicap indicators remain favorable: SEMI forecasts strong 2026–2027 fab-equipment spending and Ebara is seeing its own AI-related semiconductor equipment surge, validating the industry demand signal.

Fifth, net debt/EBITDA of 0.5x gives Atlas the financial capacity to keep investing and acquiring through a downturn rather than issue equity under stress.

Bear reasons

First, Atlas A trades near 38x trailing earnings and at a meaningful premium to relevant industrial peers, leaving the shareholder exposed to multiple compression even if earnings remain intact.

Second, Vacuum's 59% Q2 organic order growth is far above its long-run growth record and coincides with an industry-wide semiconductor-capex surge, making cyclicality a more plausible explanation than a permanent new run rate.

Third, the almost SEK 6bn quarterly order-over-revenue gap cannot be translated precisely into future sales because Atlas does not disclose a current consolidated backlog balance or 12-month conversion schedule.

Fourth, five-year aggregate OCF/net-income conversion is only 0.93x, with 2022–2023 showing that fast growth and working capital can temporarily consume a large part of reported earnings.

Fifth, the A share costs 15.6% more than B despite identical economic rights, so an A investor is paying an additional voting/liquidity premium on top of an already high corporate valuation.

Pre-mortem

The first three-year failure script begins in semiconductor equipment during 2027. Global fab plans are cut after the 2026 AI-capacity rush, Ebara and other vacuum competitors defend share aggressively, and Atlas Vacuum organic orders swing from +59% to roughly -25%. Revenue follows with a two- or three-quarter lag, Vacuum margin falls from about 21% toward 16–17% as capacity utilization drops, and group operating margin falls below 18%. If group EPS then settles near SEK 4.5 and investors stop paying 38x for a cyclical earnings profile, a 22x multiple would imply an A price near SEK 99, approximately 53% below the 2026-09-08 close. The numbers are scenario assumptions, but the transmission path follows the present order concentration and valuation.

A second failure script is slower and less dramatic operationally. From 2027–2029 Atlas keeps buying businesses, reported revenue grows, but organic growth fades to low single digits, ROCE falls below 20% and OCF/net-income stays below 0.7x for two years. Ingersoll Rand holds compressor pricing while Ebara captures more leading-edge vacuum placements. Atlas still earns roughly SEK 5 per share, but the market reclassifies it from an elite compounder to an ordinary high-quality industrial at 25x earnings. The A share would be around SEK 125, about 40% below today's price.

Neither script requires bankruptcy. The principal permanent-loss mechanism is paying a premium multiple for quality and then discovering that incremental growth no longer deserves that multiple.

Final research conclusion

Atlas Copco is one of the rarer industrial companies for which the quality thesis survives close inspection. The evidence is concrete: service has reached 38% of revenue; Compressor still earns around a mid-20s operating margin; the group crossed the pandemic with a 19.2% margin; post-Epiroc ROCE has generally stayed in the 20s or around 30%; net leverage is low; and Q2 2026 delivered broad rather than single-segment order growth. The acquisition model is also producing meaningful structural growth without yet dominating organic growth.

The A share is a different proposition from the company. At SEK 208.80 the investor pays roughly 38x trailing earnings, receives an owner-earnings yield around 2.6% and pays a 15.6% share-class premium over B for the higher vote. My base valuation of roughly SEK 200 already assumes successful conversion of today's orders, a return toward revenue growth and continued premium margins. The conservative value is closer to SEK 156. The current A price therefore offers a reasonable hold price for an existing long-term owner who accepts multiple risk, but little protection for a new buyer demanding a conventional margin of safety.

The evidence that would change that assessment upward is straightforward: several more quarters of broad book-to-bill above one, Vacuum revenue conversion while its margin stays around 20% or better, ROCE stabilizing above the mid-20s and cash conversion around one. A substantial price decline without a deterioration in those variables would be even more powerful because it would separate business quality from valuation. The judgment would move downward if the semiconductor order surge reverses before converting, group margin falls below 19%, or ROCE moves structurally below 20%.

【Company-profile scores】

• Fundamental quality: high
• Growth: medium
• Moat: strong
• Financial soundness: strong
• Management credibility: high
• Valuation attractiveness: low
• Risk level: medium
• Suitable investor type: long-term growth

【Investment rating】

• Rating: Hold
• One-line thesis: Strong installed-base economics and broad order recovery are offset by an A-share valuation near 38x trailing earnings.
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes. For a new A-share position at a balanced risk tolerance, I would wait for SEK 117–124 with fundamentals intact; the opportunity cost is missing further semicap-driven earnings and multiple expansion if the cycle remains unusually strong.
• Target holding horizon: 3–5 years, with the next 12 months focused on order conversion.
• Expected annualized return: conservative approximately -24%, base approximately -3%, optimistic approximately +21% on the 12-month scenario framework, including a conservative SEK 3 ordinary-cash-distribution placeholder.
• Max-loss risk: approximately 50–55% in the semiconductor-cycle/multiple-compression pre-mortem, with Vacuum order contraction, group margin below roughly 18% and a P/E reset toward the low 20s.
• Reassessment triggers: Vacuum organic orders below 0% for two consecutive quarters; group operating margin below 19% for two consecutive quarters; ROCE below 20%; OCF/net income below 0.7x for two annual periods; net debt/EBITDA above 1.5x without a clearly accretive acquisition rationale.

【Ideal Buy Price】117–124 SEK

Basis: 20%–25% below the SEK 156 value implied by the conservative 2027 scenario. This is deliberately a margin-of-safety entry range, rather than my estimate of the most likely near-term trading price.

【Valuation Range】

• current: 208.80 SEK (ATCO A close as of 2026-09-08)
• bear (conservative · ideal buy zone): [117, 124] SEK
• base (fair · acceptable hold zone): [185, 215] SEK
• bull (optimistic · above the clearly-overvalued line): [274, 300] SEK

The bear band is a 20%–25% margin-of-safety discount to the conservative SEK 156 central value. The base band surrounds the SEK 200 base value. The clearly-overvalued threshold starts at approximately 10% above the SEK 249 optimistic value, or about SEK 274. The bands therefore arise from the scenario inputs rather than from the current quotation.

The 12-month and 3–5-year views are different. Over 12 months, order conversion and the semiconductor cycle can still push earnings estimates higher enough for the stock to outperform despite today's valuation. Over 3–5 years, starting valuation becomes increasingly important: if Atlas compounds EPS around mid-to-high single digits but the P/E migrates from 38x into the low 30s, shareholder returns are likely to be only mid-single-digit annualized despite a very good underlying business.

The final judgment is therefore narrower than either the bull or bear caricature. Atlas is neither a valuation bubble unsupported by earnings nor an obviously cheap compounder. It is a proven industrial compounding system entering a favorable cyclical phase at a price that already assumes much of the system keeps working. For an existing A shareholder, that supports patience. For a fresh buyer demanding protection against permanent loss, the valuation asks for patience of another kind.

Other tickers mentioned

• EPI-A.ST: Epiroc is the former Atlas Copco mining and rock-excavation business distributed to shareholders and separately listed in 2018.
• IR.US: Ingersoll Rand is the closest listed compressor, flow-equipment and aftermarket comparator.
• 6361.TSE: Ebara is a direct semiconductor dry-vacuum and precision-machinery reference for Vacuum Technique.
• SAND.ST: Sandvik is a Swedish decentralized industrial-quality and cycle comparator.
• ALFA.ST: Alfa Laval is a Swedish premium process-equipment peer useful for order, margin and valuation comparisons.
• 2330.TW: TSMC's fab investment and advanced-node expansion are important external demand signals for Atlas Copco's semiconductor Vacuum business.

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