Sandvik AB: A Record 22.6% Margin With SEK 550 Million of Tungsten Inside It, and No Margin of Safety at SEK 368.70

There is an important information limit here. Sandvik does not publicly disclose a clean aftermarket EBITA margin versus original-equipment EBITA margin. It describes aftermarket as higher-margin and more resilient, and the installed-base logic supports that claim, but deriving separate margins from the published segment data would be underdetermined. Any exact figure would be fabricated. Economically, original equipment is the more capex-sensitive component: miner project approvals and fleet expansion can be deferred quickly when commodity economics deteriorate. Parts, rock tools and service are driven more by the installed fleet and operating hours, so they normally weaken later and less sharply, although a severe mining downturn eventually reduces utilization and aftermarket consumption as well.

The second distinction is between mining and short-cycle manufacturing. Mining entered the summer of 2026 with high customer activity. Sandvik Mining’s Q2 organic order growth was 11%; aftermarket orders grew 17% organically while equipment orders rose only 2%, showing that the quarter was not simply a fleet-capex blowout. Epiroc, the closest listed pure-play mining peer, reported a different mix in the same quarter: organic orders rose 13% and equipment orders 30%, with management pointing to historically high copper and gold prices. That gives some cross-company confirmation that mining demand is genuinely strong, while Sandvik’s particular Q2 composition was unusually aftermarket-heavy.

Machining operates on a shorter clock. Cutting tools are consumables, which gives the business recurring characteristics. But consumption is tied closely to factory output in aerospace, automotive and general engineering, and distributor inventories can amplify short-cycle movements. Q2 organic Machining order growth of 30% and revenue growth of 36% look spectacular, but cutting-tool orders rose a lower 20% and powder orders more than doubled amid high tungsten prices. Sandvik itself described early-July cutting-tool order conditions as stable but unusually uncertain because of tungsten dynamics. The quarter cannot be projected as a 30%-plus organic growth run rate.

Intelligent Manufacturing is a smaller but strategically important counterweight. Q2 revenue was only SEK 880 million, roughly 2.4% of group revenue, yet its 22.5% EBITA margin and subscription conversion point toward a less capital-intensive revenue model. Organic orders rose 7% and revenue 9%; acquisitions lifted reported growth further, while the ongoing transition toward subscriptions reduced recognized order and revenue growth by around two percentage points. Hexagon’s Q2 2026 results provide a useful benchmark for the broader industrial-software/metrology environment: Hexagon reported 12% organic growth, a 24.3% EBITAC margin and 4% organic recurring-revenue growth. Sandvik’s software operation is not yet large enough to determine group valuation, but it is large enough to alter the group’s mix over time.

The capital market is trading two narratives at once. The first is straightforward cyclical strength: copper and gold economics are supporting mining activity, Sandvik has posted double-digit organic order growth quarter after quarter, and manufacturing orders rebounded. The second is a quality re-rating: recurring revenue is higher, digital revenue is larger, cost flexibility is better, the group targets 20–22% EBITA through a cycle, and balance-sheet leverage remains modest. Sandvik itself says it has divested businesses representing roughly SEK 30 billion of annual revenue and acquired businesses representing more than SEK 22 billion, while recurring revenue and digital exposure have moved higher.

Q2’s market reaction revealed where the disagreement lies. Despite the record result, the shares fell roughly 9% after the report. The obvious interpretation is that investors already expected excellent numbers and focused instead on how much of the margin was repeatable, particularly the tungsten benefit in Machining. That reaction is consistent with the fundamental decomposition: Sandvik produced a genuinely strong underlying quarter, but the headline margin overstated the sustainable earnings run rate.

That debate is occurring at a demanding valuation. At SEK 368.70, Sandvik’s market capitalization is about SEK 462.4 billion. Using 2025 adjusted EPS of SEK 12.17, subtracting H1 2025 adjusted EPS of about SEK 5.97 and adding H1 2026 adjusted EPS of SEK 7.86 gives roughly SEK 14.06 of trailing adjusted EPS, implying about 26.2 times adjusted earnings. The equivalent trailing reported EPS calculation is roughly SEK 13.42, or about 27.5 times earnings. Those are above Sandvik’s reported year-end P/E levels for 2021–2025, which ranged from the high teens to 25.7 times at the end of 2025.

The balance sheet does not require a discount. At the end of 2025 Sandvik had SEK 93.2 billion of equity, financial net debt of SEK 26.5 billion and financial net debt/EBITDA of 0.9 times; Q2 2026 leverage was about 1.0 times, safely inside the company’s below-1.5-times target. The more relevant balance-sheet issue is composition: intangible assets were SEK 62.6 billion at year-end 2025, reflecting years of acquisitions, which raises the amount of capital whose ultimate return depends on successful integration and durable acquired earnings.

Share-price history explains why the current multiple is plausible but dangerous. Sandvik entered the 2020s as a company whose old conglomerate structure was being dismantled. It later separated Alleima, bought software and automation businesses, increased recurring revenue and defended profitability through a weak manufacturing cycle. In 2024, by contrast, weaker short-cycle demand contributed to a cost program involving around 1,100 jobs after Q4 profits disappointed. In 2025, strong mining orders, supported by high gold and copper prices, helped push the share sharply higher; Sandvik’s own share statistics show a 51.6% share-price increase in 2025. The market is now rewarding the idea that the old Sandvik’s cyclicality has been permanently dampened.

My qualitative-portrait label is company in transition, specifically a high-quality cyclical industrial transitioning toward a more recurring and software-rich earnings base. The transition is far enough along to be visible in margins, recurring sales and portfolio composition, but mining capex and manufacturing production still have enough weight that “high-quality compounding growth” would overstate the defensiveness.

Vertical history, financial review, and price history

Origins and listing path.

Sandvik’s origin is a metallurgy story before it is an equipment story. Göran Fredrik Göransson acquired rights to the Bessemer process in 1857 and in 1858 succeeded in producing high-quality steel at industrial scale. Högbo Stål & Jernwerks AB was founded in Sandviken on January 31, 1862. Financial stress forced the company into receivership in 1866, after which the assets were reconstructed in 1868 as Sandvikens Jernverks AB. The early problem was not “mining productivity” in the modern sense. It was reliable mass production of high-grade steel at a cost and consistency that could serve industrial customers.

Internationalization began unusually early. The company used foreign agents in the nineteenth century and sold products such as railway material, drill steel and wire outside Sweden. That export orientation became structural: in 2025 the United States, Australia, China and Canada were among Sandvik’s largest country markets, while Sweden represented only a small portion of group sales. The modern FX sensitivity is an extension of a commercial model established more than a century ago, not a recent strategic choice.

Sandvik’s shares were introduced on the Stockholm Stock Exchange in 1901. I could not locate a reliable primary record of a modern-style IPO price, proceeds figure or flotation valuation from the 1901 event, and it would be misleading to retrofit today’s IPO concepts onto that listing without archival evidence. The company did not reach the market through a SPAC, reverse merger or contemporary carve-out; it has a long-standing Swedish primary listing.

Nor does the old archival record provide a defensible named peer set for Sandvik’s first decades. Competition was primarily among European and British steel and engineering producers adopting new metallurgy processes. The enduring lesson is more useful than a speculative list of nineteenth-century competitors: Sandvik’s earliest advantage came from materials science and industrial process know-how, the same capability that later enabled carbide tools and rock-drilling products.

The first strategic turn: from steel producer to engineered consumables.

Cemented carbide produced the decisive twentieth-century shift. Sandvik established the Coromant brand in 1942, introduced carbide-tipped drill steel around the same period, entered cemented-carbide metal-cutting tools in 1943 and developed its own carbide production. By 1954, rock tools generated essentially all group profit even though steel still represented the majority of sales. That mismatch forced the company toward the businesses where metallurgy could be sold as productivity rather than as commodity tonnage.

Indexable inserts and later surface-coated carbide widened that gap. T-Max tooling in the 1950s and coated carbide in 1969 gave customers faster cutting, longer tool life and lower downtime. By around 1970, carbide products represented roughly 40% of sales, rising further later in the decade. The modern Machining business carries one of Sandvik’s oldest economic advantages. The product is physically small relative to the customer’s machine tool and final component value, yet a failed or inefficient cutting tool can destroy far more value than its purchase price. That asymmetry supports pricing power.

The crisis that created the modern operating model.

The 1982–83 downturn was the negative turning point. Sandvik suffered a sharp sales decline, cut more than 2,200 positions and recorded its first loss in 62 years in 1983. The response included decentralization, divestments and the sale of hydropower assets to strengthen cash flow. The long-term consequence was larger than the recession itself: management learned that a technically strong industrial portfolio can still destroy value if central costs and capital allocation are allowed to expand during good times.

That lesson reappears in the 2020s strategy. Sandvik now describes itself as a decentralized portfolio of 23 businesses, with divisional leaders accountable for profit and balance sheets. The structure is not simply management fashion; it has roots in the 1980s crisis, and is designed to shorten the distance between demand changes and cost action.

Global mining and tooling consolidation.

The 1990s and early 2000s created the recognizable mining/tooling group. Industrivärden became a major shareholder in the late 1990s. Sandvik acquired Tamrock in 1997, reorganized around three major businesses, divested Saws and Tools in 1999, bought Svedala’s crushing and screening operations in 2001 and acquired Walter in 2002. The economic logic was concentration: exit businesses where Sandvik lacked a durable technological edge and deepen areas in which installed equipment, materials science, tooling know-how and distribution could reinforce one another.

Mining automation emerged alongside the hardware. Sandvik’s AutoMine concept dates to the early 2000s, giving the company more than two decades of automation experience by 2026. That matters because autonomy in underground mining is not a standalone software sale: the control system must understand loaders, trucks, drilling, mine plans, safety systems and the operating environment. A supplier with both hardware and software can capture more workflow value than a component specialist, although Epiroc offers a similarly integrated automation proposition in drilling and underground equipment.

Portfolio reconstruction in the last decade.

The most recent stage has been a deliberate retreat from conglomerate breadth. Sandvik says it has divested businesses representing around SEK 30 billion of annual revenue and acquired businesses representing more than SEK 22 billion, moving toward automation, digital manufacturing, higher aftermarket content and faster-growth niches. The 2022 separation of Alleima, the former materials-technology operation, was central to this process. Under the distribution terms, every five Sandvik shares entitled the holder to one Alleima share; Alleima began trading in Stockholm in August 2022. Historical Sandvik price charts around that distribution need total-return adjustment before conclusions are drawn about operating performance.

This period also put software inside the core portfolio rather than beside it. Acquisitions in CAM, mine planning and automation helped lift digital revenue from below SEK 1 billion in 2019 to SEK 5.5 billion in 2025, and Sandvik targets SEK 13 billion by 2030. The company says most of the incremental digital growth should be organic, supplemented by acquisitions.

The 2026 reporting split is the organizational recognition of that change. Intelligent Manufacturing is now visible as a separate business rather than disappearing within Machining. That makes investor analysis better because a 22%-plus-margin software-heavy operation with subscription conversion has different capital requirements and cyclicality from cutting tools, even though both serve component manufacturers.

Current portfolio moves.

The announced acquisition of Diemme Filtration extends Rock Processing downstream into filtration and dewatering. Sandvik said Diemme should have roughly SEK 1.1 billion of 2026 revenue, operates in an addressable market exceeding SEK 20 billion, carries a sizeable aftermarket component and is expected to have an EBITA margin accretive to Rock Processing. The transaction was expected to close in Q3 2026 and Sandvik stated that returns should reach its cost of capital within three years and exceed it thereafter. The acquisition price was not disclosed in the primary announcement I found.

As of the August 21 research base date, I found no subsequent primary completion notice, so I treat Diemme as agreed but not yet verified as closed. That is different from assuming it failed to close. The latest primary disclosure available in the research set still described Q3 2026 as the expected closing period.

The Additive Manufacturing divestment runs in the opposite direction. Sandvik agreed in May to sell the unit to Mimir, with closing expected in Q3 subject to customary regulatory approvals. The Q2 accounts included impairment charges associated with the signed Additive Manufacturing divestment and the Zero Touch disposal; group items affecting comparability in Q2 totaled SEK 582 million, including SEK 320 million of impairment losses related to those signed divestments. No transaction consideration was stated in the primary Additive Manufacturing announcement reviewed for this report, and I found no verified completion notice by the base date.

These two transactions show the portfolio test management is applying. Additive Manufacturing consumed capital in a business where Sandvik had struggled to produce acceptable returns; filtration adds a process technology with an installed-base aftermarket into an area Sandvik already understands. Whether Diemme actually earns above the cost of capital after year three will be a useful test of management discipline rather than the acquisition announcement itself.

Financial vertical review.

Five-year comparisons require care because Alleima became discontinued operations and acquisitions changed the perimeter. On the continuing basis used in later annual reports, revenue expanded from about SEK 85.7 billion in 2021 to SEK 120.7 billion in 2025. The large step in 2022–23 came from acquisitions, price and organic growth, not from a smooth same-company CAGR. Operating cash flow was much more volatile around the supply-chain period because working capital absorbed cash in 2021–22 and released it later.

Sources are Sandvik annual-report financial statements; 2021–22 comparisons were restated around the Alleima separation, so the table is best read as a direction-of-travel view rather than an untouched historical perimeter.

The revenue table shows why Sandvik should not be valued on sales CAGR alone. Acquisitions expanded the denominator, Alleima reduced it, FX moves can push reported growth far away from fixed-currency growth, and mining versus manufacturing cycles move independently. In 2025 alone, reported revenue fell to SEK 120.7 billion from SEK 122.9 billion, while FX reduced adjusted EBITA by approximately SEK 2.48 billion. Operational quality improved more than the reported top line suggested.

Profitability has been substantially more stable than in older industrial cycles. Sandvik generated a 20.0% adjusted EBITA margin in 2023, 19.2% in 2024 and 19.3% in 2025 despite weak short-cycle manufacturing conditions during parts of that period. Q1 2026 moved to 20.0%, followed by Q2’s 22.6%. The evidence supports a higher margin floor than the old Sandvik, although Q2’s temporary tungsten benefit argues strongly against extrapolating the record.

Cash conversion has also been credible over a full cycle rather than every quarter. Cash flow from operating activities was SEK 19.2 billion in 2025 against net profit of SEK 14.7 billion. Across 2021–25, the annual statements imply aggregate operating cash flow of roughly SEK 82 billion against aggregate net income of roughly the high-SEK-60-billion range, or around 1.2 times. That is healthy long-run earnings conversion even though working-capital movements can produce weak individual quarters.

Q2 2026 is a good warning against using a single cash-conversion print. Free operating cash flow was SEK 3.59 billion and cash conversion only 46% despite record profit, reflecting working-capital demands as revenue accelerated. Sandvik defines free operating cash flow as EBITDA adjusted for noncash and certain acquisition items, plus the working-capital change, less rental-equipment and tangible/intangible investment; importantly, that company metric is before financial items and paid taxes, so it is not identical to equity free cash flow.

Balance-sheet quality is strong enough to support bolt-on acquisitions. At December 2025, cash was SEK 5.0 billion, equity SEK 93.2 billion, financial net debt SEK 26.5 billion and net debt/EBITDA 0.9 times; Sandvik had an S&P BBB+ rating with stable outlook and roughly SEK 11 billion of unused committed facilities. Q2 leverage moved to around 1.0 times.

Goodwill and other acquired intangibles are the balance-sheet item worth watching. Intangible assets were SEK 62.6 billion at year-end 2025, approximately two-thirds of book equity. That does not imply impairment is imminent; it means future returns on acquired software, tooling and mining assets matter materially to book-value quality. The Q2 impairment linked to portfolio exits is a reminder that unsuccessful acquisitions eventually reappear as accounting charges even when management excludes them from adjusted EBITA.

Capital spending is modest relative to sales. Sandvik spent about SEK 2.84 billion on tangible assets and SEK 0.97 billion on intangible assets in 2025, and it guides to SEK 4.0–4.5 billion of capex in 2026. R&D is more important strategically: Sandvik says it spends around 4% of annual revenue on R&D and that products launched in the prior five years represented 25% of 2025 sales. The business is knowledge-intensive without being exceptionally fixed-capital-intensive.

Price and valuation history.

The last decade can be divided into four market narratives. The first was restructuring and focus: investors rewarded a more decentralized, less sprawling industrial group. The second was the COVID shock and recovery, when cash generation proved stronger than a conventional capital-goods model might have implied. The third, across 2022–24, combined the Alleima distribution, higher rates, supply-chain normalization and weak manufacturing demand. The fourth began in 2025 as mining orders accelerated and margins remained resilient.

Sandvik’s official year-end share data make the re-rating visible. The shares ended 2021 at about SEK 246.3, 2022 at SEK 188.4, 2023 at SEK 218.1, 2024 at SEK 198.3 and 2025 at SEK 300.6. The corresponding reported P/E multiples were approximately 21.9, 18.4, 17.9, 20.3 and 25.7 times. The 2025 share-price gain was 51.6%. At SEK 368.70 on August 20, 2026, the share price and trailing multiple are above all of those five year-end levels.

Sandvik has had several historical stock splits, including 5-for-1 splits in 1993 and 2006; those are outside a modern 10-year window but matter for very long-run per-share histories. The more relevant modern adjustment is the 2022 Alleima distribution. A raw Sandvik price chart that ignores the value transferred to shareholders in Alleima will understate total shareholder return around the spin-off date.

The valuation center has shifted for a business reason and a market-preference reason. The business reason is the move toward recurring aftermarket, software and higher through-cycle margins. The preference reason is that investors have been willing to pay premiums for industrial companies combining high returns, automation exposure and resilient aftermarket cash flows. Both can be true. The current question is whether the business-quality upgrade justifies moving from the high teens/low 20s P/E range toward the mid-to-high 20s permanently.

Business model, moat, industry, and cycle

Current revenue and profit structure.

The Q2 2026 divisional data show how different the four Sandviks are. Mining accounts for roughly half of revenue and slightly less than half of adjusted divisional EBITA; Machining accounts for about 40% of sales but, in this exceptional quarter, generated more adjusted EBITA than Mining. Rock Processing is much smaller and less profitable. Intelligent Manufacturing is still tiny in group revenue terms but already operates at software-like margins.

The table uses Sandvik’s divisional Q2 disclosures. “n.d.” means Sandvik does not disclose the same aftermarket measure for those businesses.

Behind the table, the business reason matters more than the quarter’s ranking. Mining combines equipment orders with a vast installed base. Rock Processing is trying to raise aftermarket content and expand downstream. Machining is a consumables business whose products turn over much faster but whose demand follows manufacturing output. Intelligent Manufacturing sells workflow and software economics. Combining those cash-flow patterns makes Sandvik less cyclical than a pure equipment manufacturer, while leaving it more cyclical than a software or service company.

The 2025 pro-forma structure gives a more normal margin baseline than Q2:

Mining and Rock Processing are taken from Sandvik’s 2025 reporting; Machining and Intelligent Manufacturing are from the company’s pro-forma split prepared for the January 2026 organizational change. Group revenue was SEK 120.680 billion and group adjusted EBITA SEK 23.309 billion after group activities and reporting differences.

The 2025-versus-Q2 comparison is the cleanest evidence against treating Q2 Machining profitability as normal. Mining and Rock Processing margins barely moved; Intelligent Manufacturing improved modestly; Machining moved by nearly nine percentage points, with 3.8 points explicitly attributed to the tungsten effect and the remainder reflecting volume, pricing and absorption.

Equipment versus aftermarket.

Mining is built around a “seed and harvest” model, although Sandvik itself does not use that accounting label. A fleet sale places drilling rigs, loaders, trucks and rock tools into a mine. The mine then needs parts, maintenance, service, ground-support consumables and other inputs over years of operation. The installed base creates repeated customer contact, while safety and uptime make procurement less price-sensitive than the initial machine purchase in many applications.

Q2 showed the asymmetry clearly: Mining aftermarket organic orders rose 17%, while equipment grew only 2%. Rock Processing aftermarket orders rose 8% versus 5% for equipment. The group’s record earnings did not require an equally exceptional original-equipment cycle.

In a mining-capex downturn, new equipment would normally react first. Expansion projects can be postponed, replacement intervals extended and large orders cancelled or delayed. Aftermarket demand is tied more closely to ore production and machine utilization, so it should be more resilient until miners actually cut production. This is why aftermarket creates margin stability rather than true non-cyclicality. Epiroc’s high recurring content and Metso’s large service installed base are built on the same industrial logic.

Sandvik does not disclose exact aftermarket and OE margins, and that omission matters. A segment can show rising aftermarket mix and rising overall margin, but price, volumes, factory absorption and geography move at the same time. There is insufficient information to solve the separate margins reliably. For valuation, the defensible treatment is to credit aftermarket with lower revenue volatility and probably higher incremental profitability, while refusing to assign a fabricated “aftermarket margin” such as 30% or 40%.

Machining is economically recurring in a different way. A carbide insert is consumed by production rather than serviced after an equipment sale. Sandvik benefits every time a customer cuts metal, but production shutdowns hit consumption rapidly. This makes Machining less project-lumpy than mining equipment but more sensitive to short-cycle industrial output and distributor inventories.

Cost structure and operating leverage.

Sandvik combines variable material and production costs with a significant fixed base in factories, application engineering, sales, R&D and software development. That creates operating leverage when volumes rise. The effect was unusually visible in Q2: organic operating leverage was reported at 25% in Mining, 25% in Rock Processing, 55% in Machining and 41% in Intelligent Manufacturing.

Machining’s 55% Q2 operating leverage also shows the danger. Higher volumes and favorable powder absorption can push margin up quickly, while a production slowdown can reverse absorption with similar speed. Sandvik has reduced that downside through cost programs and a more variable cost base, but it cannot eliminate plant utilization economics. The company says its 2022 and 2024 programs generated around SEK 2 billion of annualized savings and a 2025 initiative is targeting another SEK 1 billion.

R&D is a recurring competitive cost rather than optional “growth spending.” Sandvik spends about 4% of revenue on R&D because cutting performance, battery-electric mining equipment, autonomy, mine-planning software and manufacturing simulation continuously improve. A company that cut R&D aggressively to defend one year’s margin would eventually lose pricing power.

Moat.

The first real moat is installed-base stickiness in mining and rock processing. Uptime, safety certification, operator familiarity, maintenance systems and parts availability matter more underground than they do in many ordinary machinery categories. Once Sandvik has equipment running at a mine, aftermarket economics can persist through multiple replacement cycles. The 66% Mining aftermarket share in Q2 and 68% in 2025 are evidence that this installed-base relationship has monetized in practice rather than existing only in sales presentations.

Second comes materials science and application know-how in Machining. Cutting tools are small relative to the cost of a machine tool, aerospace component or scrapped workpiece, so customers optimize for tool life, speed, surface quality and process reliability rather than the cheapest insert. Sandvik Coromant’s multi-decade development of carbide grades, coatings and cutting geometries has survived repeated technological cycles, which is stronger evidence of a moat than current market share alone.

Third is the combination of hardware and digital workflow. Sandvik can connect mine planning, drilling, loading, hauling, automation and parts/service; in manufacturing it is assembling CAD/CAM, simulation, tool selection and process optimization. Software by itself does not create a network effect comparable with a consumer platform, but embedding digital tools into expensive industrial workflows raises switching costs and improves the amount of productivity value Sandvik can price against.

Fourth is global technical service and distribution. A remote mine will pay for reliability only if technicians, parts and logistics exist where the machine operates. This infrastructure is expensive for a new entrant to replicate and becomes more economic as the installed fleet expands. The same applies to global machining customers that need local application engineering across plants.

Sandvik’s moat is real in Mining and Machining, moderate in Rock Processing and still being built in Intelligent Manufacturing. The group moat comes from combining recurring consumption, installed-base service and application know-how rather than from monopoly technology. Epiroc, Metso, Weir, Kennametal and large software vendors prove that customers have credible alternatives.

Management and governance.

Stefan Widing has led Sandvik since 2020, a period spanning the Alleima separation, substantial M&A, digital expansion and the increase in recurring revenue. The strongest evidence for management credibility is operational rather than rhetorical: Sandvik reports 6% revenue CAGR since 2019 despite difficult end markets, recurring revenue rising to 40%, digital revenue reaching SEK 5.5 billion, and a balance sheet still around 1.0 times net debt/EBITDA after acquisitions.

Capital allocation priorities are stated as organic investment first, dividend commitments second and value-creating M&A thereafter. Sandvik targets a dividend payout around 50% of adjusted EPS through the cycle; the 2025 payout was 49%, with a SEK 6.00 dividend. Buybacks exist as an authorization rather than a central capital-return program: only SEK 6 million of own-share repurchases appeared in the 2025 cash-flow statement.

Governance follows Swedish corporate-law and Nasdaq Stockholm standards. Sandvik’s nomination committee is built around representatives of the four largest voting shareholders plus the board chair; Industrivärden has long been influential and is represented in that process. This is a widely held institutional ownership structure rather than founder control.

I found no primary disclosure suggesting a material accounting-fraud investigation or a governance crisis as of the base date. The more relevant governance risk is ordinary acquisitive-industrial-company risk: management can overpay, exclude recurring restructuring charges from adjusted profit, or tolerate underperforming acquired assets for too long. The Additive Manufacturing impairment is a small example of why acquisition returns should be followed independently of adjusted EBITA.

Industry structure and cycles.

Mining equipment is mature in physical penetration but structurally growing in automation, electrification, productivity software and more demanding mineral extraction. Sandvik and Epiroc are especially strong in underground drilling, loading and automation; Caterpillar and Komatsu are formidable in large surface equipment; Metso, Weir and Sandvik compete more directly around comminution, crushing, screening, pumps and mineral processing. The profit pools sit disproportionately in consumables, service, parts and high-value process technology rather than in low-differentiation fabrication.

Current mining conditions are favorable. Epiroc said Q2 2026 mining customer activity remained high, supported by historically high prices in important copper and gold exposures; its organic equipment orders rose 30%. Sandvik’s own mining orders remained in double-digit organic growth, with particularly strong aftermarket demand. This is an upcycle, even though Sandvik’s particular Q2 mix was not dominated by equipment.

Machining belongs to the macro, industrial-production and inventory cycle. Aerospace can remain strong while automotive or general engineering slows; distributors can destock even when underlying production is merely flat. That is why a single blended “Sandvik cycle” is analytically wrong. Mining could remain strong through 2027 while cutting-tool demand weakens, or the reverse could happen.

Intelligent Manufacturing adds a technology-iteration and subscription cycle. Growth depends on CAM penetration, software conversion and cross-selling rather than simply industrial tonnage. Q2’s 7% organic order growth and 9% organic revenue growth were slower than Machining but less exposed to tungsten and factory absorption.

Regulation is not Sandvik’s dominant risk in the way it would be for a bank or pharmaceutical company. Trade barriers, tariffs, sanctions, local-content requirements and export controls can nevertheless move costs and production footprints. Currency is even more immediate. Sandvik’s globally distributed revenue base caused a SEK 2.48 billion negative FX effect on adjusted EBITA in 2025, making the distinction between reported and fixed-currency growth essential.

Horizontal competitors and current fundamentals

A single peer table would obscure more than it clarifies because no public company matches all four Sandvik businesses. The correct peer changes with the question.

For Mining, Epiroc is the cleanest listed comparison. Both companies sell drilling equipment, rock tools, automation and aftermarket service into mines; Epiroc is the purer mining exposure, while Sandvik adds loaders, trucks, mine-planning software and a broader group portfolio. In Q2 2026 Epiroc recorded 13% organic order growth and 20.1% adjusted operating margin, compared with Sandvik Mining at 11% organic orders and 20.5% adjusted EBITA margin. Epiroc equipment orders rose 30% organically, while Sandvik Mining equipment orders rose just 2%; Sandvik’s 17% aftermarket growth was the stronger driver.

Customers choose Epiroc when its drilling, exploration, automation and service ecosystem fits the mine architecture best; they choose Sandvik when Sandvik’s drilling, loading/hauling, rock tools and digital workflow offer superior whole-mine productivity. Neither has a monopoly, and large miners often use more than one supplier. The cross-sectional conclusion is that both are high-quality mining franchises, with Epiroc offering a cleaner pure-play and Sandvik a broader earnings mix.

The market prices that purity. Epiroc A traded around the mid-SEK-200s in August 2026 at roughly 35 times trailing earnings, materially above Sandvik’s approximately 27.5 times reported trailing P/E and 26.2 times trailing adjusted P/E. A valuation discount for Sandvik is understandable because its short-cycle Machining exposure and portfolio complexity dilute pure mining scarcity; the size of the discount is less obviously justified now that Sandvik’s recurring mix and through-cycle margins have improved.

Caterpillar and Komatsu matter where scale and large mobile mining equipment are decisive. Caterpillar’s Q2 2026 group revenue jumped 24% and adjusted operating margin reached 21.9%, showing how profitable the broader heavy-equipment cycle can become at high utilization. These companies have stronger surface-haulage scale and dealer networks, while Sandvik has deeper specialization in underground hard-rock mining, rock tools and mine automation. They are competitive references, not clean valuation twins.

Rock Processing has a different peer map. Metso is the most relevant process-equipment competitor across aggregates and minerals, with crushing, grinding and mineral-processing equipment plus a large service base. Weir is particularly strong in slurry handling, pumps and wear-intensive mining applications. Sandvik historically had greater strength upstream in crushing and screening; Diemme is strategically important because filtration moves it further downstream into a process step where specialist know-how, installed base and recurring service can support higher margins.

Metso’s own history illustrates why service mix matters. In earlier quarters its Minerals business has reported service shares around two-thirds of sales, providing resilience when equipment orders fell sharply. In H1 2026 Metso said sales and profitability improved, with parts of the portfolio reaching mid-teens adjusted EBITA margins. Sandvik Rock Processing’s 14.6% Q2 margin is credible but still below Sandvik’s 17–19% 2025–30 ambition for the business area; Diemme and greater aftermarket penetration are part of the bridge management is trying to build.

Weir’s strategic trajectory offers another benchmark. Its mining franchise has emphasized aftermarket growth, cost improvement and a move toward 20%-plus operating margins. This matters because Sandvik Rock Processing cannot claim that a larger aftermarket mix is proprietary strategy; competitors are pursuing the same profit pool. Execution, product performance and installed-base wins will determine who captures it.

Machining’s most useful public peer is Kennametal, though the quality gap is substantial. Kennametal’s fiscal 2025 Metal Cutting adjusted operating margin was only 7.9% in the fourth quarter, while Sandvik Machining generated 19.8% adjusted EBITA for full-year 2025 before exploding to 28.7% in Q2 2026. Even after mechanically removing Sandvik’s 380-basis-point tungsten benefit, the Q2 margin would have been about 24.9%.

That gap says more than a product-parameter table could. Sandvik became the premium productivity supplier whose economics depend on high-value inserts, round tools, materials expertise and global application engineering. Kennametal remains a credible competitor across metal cutting and wear-resistant materials, but recent results have required restructuring and capacity actions to offset weaker volumes and inflation. Customers will still switch when performance, price or technical support favors Kennametal or Japanese competitors, yet Sandvik’s sustained margin premium indicates that its product mix and pricing power have generated real economic differentiation.

Kennametal also demonstrates why current peer P/Es need caution. Data services showed an unusually low trailing P/E after a sharp change in fiscal-2026 earnings, while forward measures were much higher because analysts expected normalization. Using that distorted trailing figure to call Sandvik “expensive” or Kennametal “cheap” would confuse temporary earnings with business quality.

Intelligent Manufacturing belongs beside Hexagon and, more loosely, industrial-software companies rather than mining-equipment peers. Hexagon’s measurement, positioning, digital-twin and autonomy technologies touch many of the same manufacturing customers. Hexagon produced 12% organic Q2 2026 growth, a 24.3% EBITAC margin and 4% organic recurring-revenue growth; Sandvik Intelligent Manufacturing produced 9% organic revenue growth and a 22.5% adjusted EBITA margin. Sandvik’s operation is much smaller, but the margin profile is already comparable with specialist industrial technology.

That leaves Sandvik in a distinctive ecological niche. It is not the largest heavy-equipment company, the purest mining company, the largest mineral-processing supplier or the largest industrial-software vendor. It owns several premium niches where a small product or software layer controls a large amount of customer productivity, then uses installed bases and consumable demand to turn those niches into recurring revenue. The threat is equally clear: Epiroc can take mining automation and drilling economics, Metso or Weir can take downstream process economics, Kennametal and private/Japanese tooling groups can attack machining share, and Hexagon or broader software ecosystems can capture the digital workflow.

The last four reported quarters.

Sandvik’s acceleration began before the record Q2, which is why the current thesis cannot be reduced to one anomalous quarter.

Reported growth calculations use the disclosed SEK values; fixed-currency and organic figures are Sandvik’s own disclosures. For Q3 and Q4 2025, the releases supplied fixed-currency growth but did not give a separate organic percentage in the headline data, so none is invented.

The underlying turn is visible in orders. Fixed-currency order growth moved from 16% in Q3 to 15% in Q4, 23% in Q1 and 17% in Q2. Reported figures were much weaker in late 2025 because currency moved against Sandvik. Mixing the reported and fixed-currency series would make the improvement look much later than it actually occurred.

Q3 2025 was still a 19% margin quarter despite reported revenue contraction, illustrating improved cost flexibility. Q4 moved to 19.6%, Q1 to 20.0%, and Q2 then benefited from both broad volume growth and the Machining raw-material effect. The sequence supports some structural margin improvement before tungsten enters the analysis.

Cash flow moved the opposite way as growth accelerated. Q3 and Q4 2025 generated SEK 5.6 billion and SEK 6.7 billion of free operating cash flow, while Q1 and Q2 2026 each generated around SEK 3.6 billion. That does not invalidate the earnings acceleration, but it raises the burden of proof for H2: inventories and receivables should ultimately convert as the order book delivers.

What the market is trading now.

Real fundamentals include double-digit organic group orders, double-digit Mining aftermarket growth, resilient Mining margins, a strong Machining volume recovery, rising subscription activity and modest leverage. These are observable. The narrative layer is that Sandvik has become so structurally resilient that 20–22% margins deserve to be capitalized at a premium industrial multiple even near a strong point in the mining cycle.

The Q2 share-price selloff indicates that the narrative had run ahead of the headline result. A company can beat earnings expectations and fall because the market had already paid for those earnings. The tungsten disclosure gave investors a concrete reason to question the marginal quality of the surprise.

Publicly available sources do not provide a sufficiently clean time series of post-Q2 sell-side EPS revisions to claim that “consensus was raised X%.” Sandvik lists more than a dozen covering analysts, and public data pages show 2026–27 estimates, but I did not find a primary or LSEG series documenting the before-and-after revision magnitude. One external research provider raised its valuation after Q2, but that is not a market-wide consensus revision. I therefore leave the consensus-revision question unquantified.

Bull/bear divergence.

The bulls’ strongest evidence is the margin floor, not the record margin. Sandvik generated about 19–20% adjusted EBITA through weaker manufacturing conditions, then moved above 20% when volumes returned. Recurring revenue is higher than in 2019, Mining aftermarket demand is strong and the balance sheet can fund further bolt-ons without approaching the 1.5-times leverage limit.

Their second argument is that mining demand is more than a Sandvik-specific order bulge. Epiroc simultaneously reported 13% organic order growth and 30% organic equipment-order growth, citing high copper and gold activity. If miners sustain production and fleet investment, Sandvik can keep harvesting its installed base even if new-equipment growth moderates.

The bears’ strongest evidence is the quality of Q2 Machining profit. A SEK 550 million tungsten effect was worth 380 basis points to the area margin. When one temporary pricing/raw-material mechanism produces about 1.5 percentage points of group margin, a headline “record 22.6%” becomes a poor starting point for normalized earnings.

Their second argument is valuation. Sandvik trades at roughly 26 times trailing adjusted earnings and 27.5 times reported earnings, above all of its 2021–25 year-end P/E observations. That multiple leaves limited room for a conventional mining or manufacturing downturn.

The final disagreement concerns the nature of the 40% recurring mix. Bulls see it as a permanent reduction in cyclicality. Bears can correctly point out that “recurring” does not mean “independent of customer production”: mines consume fewer parts if machines run fewer hours, and factories consume fewer inserts if they machine fewer components. The portfolio has become more resilient; the cycle has not disappeared.

Valuation, risk, catalysts, and tracking

Historical valuation.

Using Sandvik’s own year-end data, the stock traded around 17.9–21.9 times reported earnings from 2021 through 2024 before rising to 25.7 times at the end of 2025. At SEK 368.70, the current roughly 27.5-times reported trailing P/E sits above all five 2021–25 year-end observations. Calling that an exact “historical percentile” would imply precision the sparse year-end sample cannot support, but it is plainly the expensive end of recent history.

A separate market-data calculation put Sandvik’s EV/EBITDA at about 15.5 times on August 20, 2026. That cross-check points in the same direction: the market is not valuing Sandvik as an ordinary mid-cycle machinery company.

The premium partly reflects real change. Recurring revenue rose by nine percentage points of group sales between 2019 and 2025, digital revenue increased by more than fivefold, and management raised the target architecture to 20–22% adjusted EBITA through a cycle. A permanent valuation shift from the old low-teens industrial range is defensible. Moving into the upper 20s requires confidence that those margins will survive the next synchronized mining/manufacturing downturn.

Peer valuation.

Epiroc’s roughly 35-times trailing P/E is higher than Sandvik’s. That premium reflects mining purity, high recurring content and a similarly strong margin profile. Sandvik’s discount is justified to some extent because Machining adds short-cycle exposure and group complexity, but Sandvik also owns a high-quality cutting-tools franchise and a growing software option that Epiroc lacks. The valuation gap should narrow if Sandvik proves 20%-plus margins in a downturn; it should widen if Machining mean-reverts sharply.

Kennametal is cheaper on normalized market metrics but also produces materially lower Metal Cutting margins, so that discount does not establish that Sandvik is cheap. Metso and Weir provide more useful operating benchmarks for Rock Processing than whole-group multiple anchors. Hexagon’s software-heavy economics deserve a higher valuation framework than machinery, but Intelligent Manufacturing is only a few percent of Sandvik revenue, so applying a Hexagon-like multiple to the whole group would be inappropriate.

Cash-flow passthrough.

Operating cash flow converted earnings well over five years. The 2021–25 annual cash-flow statements sum to roughly SEK 82 billion of operating cash flow, versus roughly SEK 68 billion of net profit on a comparable broad basis, giving an operating-cash-flow/net-income ratio around 1.2 times. The weak years were primarily working-capital years rather than evidence that reported earnings were structurally fictitious.

Sandvik does not split maintenance and growth capex. So for owner-earnings purposes I use an explicit assumption rather than presenting false precision: of 2025’s approximately SEK 3.81 billion tangible and intangible capex, I treat 70–80%, or roughly SEK 2.7–3.0 billion, as maintenance and 20–30%, or roughly SEK 0.8–1.1 billion, as growth/strategic investment. The range is consistent with a mature industrial asset base, modest total capex intensity and ongoing capacity/digital investments, but it is an analyst assumption, not company disclosure.

On that basis, 2025 owner earnings approximate SEK 16.1–16.5 billion: SEK 19.19 billion of operating cash flow less SEK 2.7–3.0 billion of estimated maintenance capex. That is about SEK 12.8–13.2 per share. At SEK 368.70, the owner-earnings P/E is approximately 28–29 times, or a 3.5–3.6% owner-earnings yield. The difference from the roughly 27.5-times headline reported trailing P/E is far below 30%, so the framework does not trigger the instruction to discard accounting earnings entirely.

Trailing adjusted EPS can be estimated at about SEK 14.06 by taking 2025 adjusted EPS of SEK 12.17, subtracting approximately SEK 5.97 for H1 2025 and adding SEK 7.86 for H1 2026. That gives the approximately 26.2-times adjusted P/E noted above. Trailing group revenue is roughly SEK 129.1 billion and trailing adjusted EBITA roughly SEK 26.35 billion, or about a 20.4% margin.

Sandvik’s own free-operating-cash-flow definition gives roughly SEK 19.5 billion on the same trailing construction, equivalent to about 4.2% of market capitalization. That ratio is only a cross-check because Sandvik’s FOCF is calculated before financial items and paid taxes; it should not be presented as a conventional equity FCF yield.

Absolute valuation.

Normalized owner-earnings/P/E, EV/EBITDA and cash-yield cross-checks are the most suitable methods here. A full DCF would give an appearance of precision while the key uncertainty is simply how much of a record-margin cycle is sustainable. The scenarios normalize margins and earnings first, then apply multiples consistent with different business-quality outcomes.

This is valuation-scenario analysis within a research framework, not investment advice. Inputs are intentionally normalized below Q2’s headline Machining margin; Sandvik’s 7% through-cycle growth target and 20–22% group margin target inform, but do not dictate, the base and optimistic cases.

The conservative case is not a recession apocalypse. Revenue remains roughly around today’s trailing level and margin settles at 18.5%, still respectable for an industrial company. A 21-times multiple remains generous compared with Sandvik’s weaker historical valuation points. Yet that combination produces only about SEK 273 per share, showing how much current value depends on the structural-margin thesis.

The base case gives Sandvik credit for most of management’s transformation. Revenue compounds toward SEK 150 billion by 2029, recurring/digital mix rises further, and adjusted EBITA settles around 20.5%. With normalized owner earnings of about SEK 15.8 a share and a 23-times multiple, fair value is around SEK 363, almost exactly where the stock trades. The current price behaves like a fair price for successful execution, not a price that protects against disappointing execution.

The optimistic case assumes Sandvik reaches the upper end of its structural margin framework while maintaining growth close to its 7% cycle target. Even there, a SEK 450 central value implies only a low-twenties percentage of aggregate price upside from the current quote before dividends. This is the core valuation problem: excellent execution is already required to generate an attractive result.

Expectation gap.

The market appears to price a through-cycle EBITA margin around 20% or better, continued mining strength and a durable premium multiple. It does not appear to price 22.6% forever; the post-Q2 selloff makes that clear. The expectation gap will be created by whether normalized Machining margin remains in the mid-20s after tungsten unwinds and whether Mining aftermarket can continue double-digit growth if new-equipment orders slow.

Sandvik’s next earnings report is scheduled for October 22, 2026. Investors should care less about whether group revenue beats consensus by a few hundred million SEK than about three quality indicators: the Machining margin bridge after tungsten normalization, Mining aftermarket versus equipment orders, and cash conversion/working capital.

A fourth signal is Intelligent Manufacturing subscription growth. The business is too small to rescue a mining downturn today, but persistent high-single-digit or better organic growth with 20%-plus margins would support a gradual conglomerate-multiple re-rating as software becomes more material.

Margin-of-safety recheck.

The current SEK 368.70 price trades about 35% above the conservative scenario’s SEK 273 central fair value. On the discipline specified in this framework, the margin of safety against the conservative outcome is zero.

The most fragile base-case assumption is normalized owner earnings around SEK 15.8 per share, which embeds a roughly 20.5% group EBITA margin and continued mix improvement. Mechanically reducing that earnings assumption to 70%, while leaving the 23-times multiple unchanged, gives owner earnings of about SEK 11.1 per share and a valuation near SEK 254. The exercise is deliberately severe, but it makes clear that the current price is highly sensitive to the new margin floor.

If earnings remain flat for three years and the multiple does not expand, the investor mainly receives the dividend. The SEK 6.00 2025 dividend represents only about a 1.6% yield at SEK 368.70. Sweden’s 10-year government bond yielded about 3.03% on August 20, 2026. Under the framework’s required comparison, there is no margin of safety at this buy price.

This is close to a “good company, demanding price” case. Waiting carries opportunity cost because Sandvik may continue growing earnings and dividends and may never revisit a deep conservative value. That cost is real; it does not change the arithmetic of paying more than the conservative and base central values.

Margin-of-safety sufficiency verdict: none.

Permanent-loss risks.

The highest-impact business risk is a mining-capex reversal. Probability is medium; impact is high. The earliest observable indicator would be Mining equipment organic orders turning negative by double digits for more than one quarter, followed later by slowing aftermarket orders. The transmission path runs from fleet orders to plant absorption and then, if mine production weakens, into parts and service. Lower earnings would coincide with a market reclassification from “structurally improved compounder” back toward “premium cyclical.” Epiroc’s current 30% equipment-order growth shows how far the industry is presently on the favorable side of that cycle.

The most immediate earnings-normalization risk is tungsten. Probability is high because the Q2 effect was explicitly temporary; impact is medium to high. The observable variables are tungsten prices, cutting-tool pricing and Machining margin. SEK 550 million contributed around 380 basis points to Machining margin in Q2. If customer pricing adjusts faster than input costs, the benefit reverses; if underlying manufacturing demand simultaneously weakens, absorption becomes a second hit.

A synchronized short-cycle manufacturing downturn has medium probability and high impact because Machining is currently producing a disproportionate share of profit. Cutting tools are consumable, but consumption follows production. The alert would be cutting-tool organic orders below zero for two quarters together with a Machining margin below roughly 21–22% after raw-material normalization. The share-price impact could exceed the earnings decline because the market is currently paying a premium for the belief that portfolio changes reduced cyclicality.

Acquisition and intangible-asset risk has medium probability and medium-to-high long-run impact. With SEK 62.6 billion of intangibles against SEK 93.2 billion of equity, repeated acquisitions that fail to beat the cost of capital can erode true owner returns even while adjusted EBITA rises. The observable indicators are acquired-business growth, ROCE, impairment charges and leverage. Diemme’s stated three-year cost-of-capital hurdle is a tracking metric, not merely transaction boilerplate.

FX risk has high probability but generally medium permanent-loss impact. Sandvik lost about SEK 2.48 billion of adjusted EBITA to currency in 2025, yet currency translation does not normally destroy the underlying local-market franchise. It becomes more serious when FX coincides with local-cost mismatches or forces price increases that reduce competitiveness. Reported growth should never be used as a substitute for organic/fixed-FX growth.

Valuation compression is currently the most direct capital-market risk. Probability is medium to high, impact high. A move from roughly 26–28 times normalized earnings back toward 18–20 times would cause substantial price loss even if earnings did not decline. The observable indicators are Sandvik’s P/E versus its own history, Swedish bond yields and the premium to industrial peers.

Catalysts.

Positive catalysts over the next year include Q3 evidence that Machining can retain a margin above its 2025 level after the tungsten effect fades, continued double-digit Mining aftermarket orders, and completion and clean integration of Diemme. Add to that a return of cash conversion toward normal as working capital unwinds, and Intelligent Manufacturing sustaining high-single-digit or better organic growth while subscription revenue rises. Each would support the structural-margin thesis with evidence independent of Q2’s temporary benefit.

Negative catalysts are the mirror image: Mining equipment orders turning sharply negative, aftermarket falling after a lag, Machining margin dropping below its pre-Q2 run rate, Q3 cash conversion remaining weak, or a large acquisition pushing leverage toward the 1.5-times target without a clear return path.

Tracking dashboard.

The “normal” and “alert” values below are research thresholds, not management guidance unless explicitly tied to Sandvik’s disclosed target.

Sandvik’s formal group targets are 7% growth through a cycle excluding currency, a 20–22% adjusted EBITA margin and financial net debt/EBITDA below 1.5 times. The dashboard deliberately sets some alerts before the formal leverage covenant-like ceiling; an investor needs an early warning, not confirmation after balance-sheet flexibility has already disappeared.

Mining aftermarket and cutting-tool orders should be followed separately because they answer different questions. Aftermarket diagnoses mine utilization and installed-base health. Cutting-tool demand is one of Sandvik’s fastest indicators of global factory activity. Group order intake alone can hide deterioration in one while strength in the other dominates the total.

Cash conversion and net working capital identify whether reported growth is consuming disproportionate capital. Q2’s 46% cash conversion is not alarming by itself because sales were expanding rapidly, but another several quarters of record profit without cash realization would weaken the quality thesis.

The October 22 report is the first major scheduled test after Q2. The most important bridge will be from the 22.6% headline margin toward normalized profitability after the tungsten windfall.

Cross-synthesis, key data, uncertainties, and sources

Vertically, Sandvik has proven one capability more consistently than any other: it can turn materials and engineering know-how into high-value industrial productivity products, then migrate the portfolio toward the product categories where that know-how earns the best return. The company began by monetizing a better steelmaking process, shifted into drill steel and carbide, then built high-margin cutting tools and rock equipment. It survived the 1980s by decentralizing, expanded in mining through Tamrock and Svedala, then spent the last decade disposing of lower-quality businesses and acquiring software, automation and specialized technology. The individual products changed repeatedly; the economic pattern did not.

Past success came from a combination of technological competence, cycle exposure and management adaptation. The mining boom has periodically made Sandvik look better than it was; industrial recessions have periodically made it look worse. Yet pure cycle does not explain why carbide tools became profit engines while old steel operations were de-emphasized, why Sandvik survived the 1983 loss and emerged decentralized, or why recurring revenue rose from 31% to 40% while digital revenue exceeded SEK 5 billion. Management and portfolio choices matter.

Those success factors are still present. Sandvik remains willing to divest underperforming operations, as Additive Manufacturing shows, while acquiring adjacent installed-base businesses such as Diemme. It funds R&D at around 4% of revenue, maintains modest leverage and gives divisions P&L accountability. These attributes raise the probability that the current portfolio will keep improving. They do not guarantee that each acquisition earns its cost of capital.

Horizontally, Sandvik’s real advantage is breadth within carefully chosen high-value niches. Epiroc is a purer mining company and can be stronger in particular drilling and automation applications. Caterpillar and Komatsu dwarf Sandvik in large mobile surface equipment. Metso has deeper process-plant positions in parts of mineral processing; Weir dominates selected pump and wear niches. Kennametal competes directly in tools, and Hexagon has much greater scale in measurement software. Sandvik’s strength is that several of these businesses sit inside one decentralized capital-allocation platform while maintaining strong specialist brands and technical sales organizations.

That breadth also creates the weakness. Investors can buy Epiroc for cleaner mining exposure or Hexagon for a purer industrial-digital thesis. Sandvik earns a conglomerate discount when one business is booming and another is obscuring the earnings quality. The 2026 reporting split helps because Intelligent Manufacturing can now prove its economics separately, but at SEK 880 million of quarterly revenue it cannot yet dominate the group multiple.

The most likely market misjudgment is subtle. I do not think the market is wrong to assign Sandvik a higher through-cycle margin than five or ten years ago. The recurring mix, digital growth and cost structure support that. The more questionable assumption is how much further the valuation multiple should rise simply because a structurally better business is also enjoying a favorable mining cycle and a temporary Machining raw-material benefit. Business quality and cycle strength have arrived together, making them difficult to separate in headline EPS.

The Q2 decomposition solves part of that problem. Strip only the disclosed SEK 550 million tungsten contribution and the group’s 22.6% margin falls mechanically to about 21.1%. That is still excellent and sits within management’s 20–22% through-cycle range. The structural improvement survives the adjustment. What does not survive is the idea that 22.6% itself should be capitalized as normalized earnings.

The next 12 months revolve around normalization. Mining aftermarket must stay healthy, Machining must reveal its post-tungsten margin, and working capital must turn record accounting profit into cash. The downside is asymmetric because the multiple already assumes a high-quality outcome. A merely “good” Q3 or Q4 can disappoint at 26–28 times earnings even if the long-term franchise remains intact.

Over three years, the question changes from quarterly margin normalization to portfolio returns. Intelligent Manufacturing needs to become large enough to matter, Rock Processing needs to move toward its 17–19% margin ambition, Diemme should reach the promised cost-of-capital return threshold, and Mining must preserve its installed-base economics through whatever commodity cycle arrives. Success on those variables could make the current premium multiple structurally defensible.

Over five years, the decisive variable is whether Sandvik can compound without repeatedly paying away the value in acquisitions. A group that grows roughly 7% through the cycle, holds 20–22% margins, converts earnings into cash and maintains leverage around one turn deserves to be viewed as a high-quality industrial compounder. A group that reaches the same revenue growth through expensive acquisitions, goodwill accumulation and peak-cycle margins does not. ROCE and owner earnings matter more than the SEK 13 billion digital-revenue target in isolation.

Bull and bear reasons

Bull reasons:

• Mining aftermarket represented 66% of Mining Q2 revenue and organic aftermarket orders rose 17%, giving Sandvik a recurring earnings buffer that does not depend solely on new fleet capex.
• Aftermarket and other recurring revenue has risen from 31% of group sales in 2019 to about 40% in 2025, while digital revenue rose from below SEK 1 billion to SEK 5.5 billion.
• After removing the disclosed tungsten benefit mechanically, Q2 group EBITA margin is still roughly 21.1%, supporting management’s 20–22% through-cycle target.
• Financial net debt/EBITDA was only about 1.0 times in Q2 versus a below-1.5-times target, leaving substantial balance-sheet flexibility.
• Epiroc’s simultaneous 13% organic order growth and 30% equipment-order growth confirm that current mining strength is industry-wide rather than uniquely Sandvik backlog noise.

Bear reasons:

• Q2 Machining’s 28.7% margin contained a disclosed SEK 550 million tungsten effect worth 380 basis points, so headline earnings materially overstate the quarter’s normalized profitability.
• At SEK 368.70, Sandvik trades around 26.2 times trailing adjusted EPS and 27.5 times reported EPS, above every 2021–25 year-end P/E in Sandvik’s own share statistics.
• Mining is in a favorable commodity-driven cycle, and recurring aftermarket will eventually weaken if mine utilization falls; 40% recurring revenue reduces cyclicality but does not remove it.
• Q2 cash conversion fell to 46% as working capital absorbed cash, so record accounting earnings still need to prove full cash realization.
• Intangible assets of SEK 62.6 billion are large relative to SEK 93.2 billion of equity, making acquisition returns and future impairments material to owner economics.

Pre-mortem

The first concrete three-year failure script is a 2027–28 mining reversal. Copper and gold investment slows, Epiroc and Sandvik compete aggressively for a smaller pool of automation and replacement projects, Sandvik Mining equipment organic orders fall roughly 20% and aftermarket orders turn negative after a lag. Machining is simultaneously back in a normal industrial downturn. Group adjusted EBITA margin falls from a normalized 20–21% to about 17.5%, normalized EPS drops toward SEK 10, and the market moves from roughly 26 times earnings to 17 times. A price around SEK 170 would represent a decline of more than 50% from SEK 368.70. The numbers are a stress script, not a forecast; the transmission mechanism is historically plausible for a premium cyclical.

The second script starts in Machining. Tungsten prices normalize in 2027, customers demand price concessions just as general engineering and automotive production weaken, while Kennametal and other tooling suppliers compete harder on price. Machining margin falls from the Q2 28.7% headline, past the roughly 24.9% tungsten-adjusted level, toward 20%. Group margin falls below 18.5%, normalized EPS settles around SEK 11 and the stock is valued at 18 times as investors reject the “structurally non-cyclical” narrative. That produces a price around SEK 200, roughly 45% below the base-date close before dividends.

Final research conclusion

Sandvik is a substantially better business than the old cyclical-engineering label implies. Installed-base aftermarket, cutting-tool consumables and software now create more recurring revenue; management has removed large lower-quality operations; margins remained around 19–20% even during weak manufacturing conditions; and balance-sheet leverage is modest. Q2 2026 strengthened that thesis because the underlying margin remains around 21% even after mechanically removing the disclosed tungsten boost.

The problem is price, not franchise quality. At SEK 368.70, the shares trade around 26 times trailing adjusted earnings, while my base normalized value is approximately SEK 363 and conservative value about SEK 273. The current quote pays in advance for a large part of the portfolio transformation and leaves little protection against a normal industrial downturn.

For an existing long-horizon holder, the recurring mix, strong Mining franchise and capital structure are sufficient reasons not to exit a high-quality company solely because the valuation is full. For new capital, the expected return is weak unless Sandvik reaches something close to the optimistic case. The evidence supports holding a strong franchise rather than chasing a record-margin quarter.

【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: cyclical

【Investment rating】

• Rating: Hold
• One-line thesis: Recurring mining aftermarket and digital mix support a higher through-cycle margin, but SEK 368.70 already discounts most of that improvement.
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes. For new capital, the framework requires SEK 205–218 or evidence materially stronger than the base case, such as sustainable post-tungsten margins above 21%, healthy Mining aftermarket and normalized cash conversion. The opportunity cost is foregoing dividends and further earnings growth if the share never enters that range.
• Target holding horizon: 3–5 years
• Expected annualized return: conservative approximately -7.6%; base approximately +1.3%; optimistic approximately +8.6%, each over three years including the scenario dividend assumptions.
• Max-loss risk: roughly 50–55% in the pre-mortem case where Mining orders contract sharply, Machining normalizes below 20–21%, group EBITA margin falls toward 17–18% and the P/E compresses to roughly 17–18 times.

【Ideal Buy Price】205–218 SEK

Basis: the conservative central valuation is approximately SEK 273; the upper end of SEK 218 is about 20% below that value, while SEK 205 provides a larger buffer against cyclical and multiple risk.

Acceptable hold price: SEK 330–400. This range surrounds the approximately SEK 363 base valuation and stays inside the framework’s ±15% hold discipline.

Clearly overvalued price: SEK 495 and above. This begins 10% above the SEK 450 optimistic central valuation.

Reassessment-trigger signals:

• Reassess downward if group adjusted EBITA margin falls below 19% for two consecutive quarters without an identifiable temporary factor.
• Reassess downward if Mining aftermarket organic orders turn negative while equipment orders are below -10% for two quarters.
• Reassess downward if financial net debt/EBITDA rises above 1.3 times and management announces another material acquisition before leverage normalizes.
• Reassess upward if post-tungsten Machining maintains at least a mid-20s adjusted EBITA margin while cutting-tool organic orders remain positive.
• Reassess upward if Intelligent Manufacturing sustains high-single-digit-plus organic growth and Sandvik reaches its 20–22% group margin range with R12 cash conversion above roughly 80%.

【Valuation Range】

• current: 368.70 SEK (close as of 2026-08-20)
• bear (conservative · ideal buy zone): [205, 218]
• base (fair · acceptable hold zone): [330, 400]
• bull (optimistic · above the clearly-overvalued line): [495, 530]

The range is derived directly from the valuation framework: a roughly SEK 273 conservative central value with at least a 20% purchase margin of safety, a SEK 363 base value with a narrower hold corridor, and a SEK 450 optimistic value whose clearly-overvalued threshold begins 10% higher.

Key data

Calculations combine Sandvik’s 2025 annual data, H1 2025 comparatives and H1 2026 interim data with the August 20 share-price close.

The snapshot captures the central tension. A 20.4% trailing adjusted EBITA margin and 1.0-times leverage are high-quality industrial numbers. A 26-times-plus earnings multiple and dividend yield materially below the Swedish 10-year government yield provide little standalone valuation support.

Research uncertainties

First, Sandvik does not disclose an exact EBITA margin for aftermarket versus original equipment. The report quantifies revenue mix but refuses to manufacture a margin split from insufficient data.

Second, Diemme Filtration and Additive Manufacturing were announced as Q3 2026 closings, but I did not find a primary completion notice by the August 21 research base date. Consideration was also not stated in the primary announcements reviewed. Their final cash/debt effects remain uncertain.

Third, Sandvik does not disclose maintenance versus growth capex. The owner-earnings calculation uses a stated 70–80% maintenance-capex assumption rather than presenting the estimate as reported data.

Fourth, a clean market-wide before/after series of post-Q2 analyst EPS revisions was not publicly available in the sources reviewed. Individual analyst estimates exist, but claiming a quantified consensus revision would overstate the evidence.

Fifth, the exact IPO price, capital raised and listing valuation for Sandvik’s 1901 Stockholm introduction were not located in reliable primary material. The event itself is documented; modern-style deal metrics are not.

Sources

The primary research base is Sandvik’s Q2 2026 interim report and divisional pages, which provide the group results, margin bridge, aftermarket mix, operating leverage and tungsten effect.

Historical and balance-sheet work relies on Sandvik’s 2025 Annual Report, earlier annual-report cash-flow statements, official share statistics and the company’s history archive.

The structural transformation, financial targets, recurring-revenue mix, digital revenue, capital-allocation principles, R&D spending and 2030 ambitions come from Sandvik’s current investor materials.

Competitive evidence is drawn principally from Epiroc’s Q2 2026 report, Metso’s 2026 half-year reporting, Weir’s 2026 reporting, Caterpillar’s Q2 2026 results, Komatsu’s investor disclosures, Kennametal’s earnings materials and Hexagon’s Q2 2026 interim report.

Portfolio-transaction evidence comes from Sandvik’s Diemme Filtration and Additive Manufacturing announcements, while market-price and government-yield checks use dated August 2026 market sources and Swedish bond-market data.

Other tickers mentioned

EPI-A.ST: Epiroc is the closest listed pure-play comparison for Sandvik’s mining equipment, drilling, automation and aftermarket franchise.

METSO.HE: Metso is the most relevant listed competitor in mineral processing, crushing, aggregates and downstream mining services.

WEIR.LSE: Weir is a mining-process and aftermarket benchmark, particularly in slurry handling, pumps and wear-intensive applications.

CAT.US: Caterpillar provides the global scale benchmark for large surface-mining and heavy mobile equipment.

6301.TSE: Komatsu competes in large mining and construction machinery and provides another surface-equipment reference point.

KMT.US: Kennametal is the most useful listed public comparator for Sandvik’s metal-cutting consumables business.

HEXA-B.ST: Hexagon is a reference for industrial measurement, manufacturing software, autonomy and recurring digital economics.

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