ANDRITZ: Service Reached 46% of Revenue, but the Record Backlog Is Hydro-Heavy and Margin-Dilutive

The calculation uses ANDRITZ's H1 2026 disclosed backlog and margins.

The weighted implied margin is only about 8.1%, roughly 50 basis points below the H1 group comparable margin. If Hydropower reaches an 8% margin, the weighted figure becomes about 8.5%; at 9%, it reaches roughly 9.0%. Management's 2027 Hydro target is 7–9%, and the 7.1% H1 result already entered that range.

On current evidence the record backlog is margin-dilutive at today's business-area profitability; it becomes neutral around an 8% Hydro margin and clearly accretive near 9%. There is an important second-order offset. Only about EUR 3.30 billion of the EUR 12.60 billion backlog is service, roughly 26%, versus a 46% service share in current revenue. Service orders naturally carry shorter backlogs, so the capital-heavy backlog does not mean future reported revenue will be only 26% service: new service orders will be booked and executed while the capital backlog converts. But it means the backlog itself deserves a lower margin assumption than an investor might infer from today's blended group margin.

The Hydro demand backdrop has both a structural component and a bunching component. In H1 2025, before the latest surge, ANDRITZ's Hydro orders had already risen 72.1%, supported by plant upgrades and pumped-storage projects in Asia, including major Indian projects such as Tarali. In 2025 it commissioned work at Pinnapuram in India. Globally, the International Hydropower Association says 11.7 GW of pumped-storage capacity was commissioned in 2025, a record, taking worldwide pumped storage above 200 GW. China alone has more than 200 GW under construction, and Reuters reported that global pumped-hydro projects under development approach 570 GW.

The secular case is therefore real: variable wind and solar increase the value of long-duration storage, inertia, fast ramping and grid balancing. The 81.8% H1 2026 order increase is nevertheless mathematically unsustainable. Hydro's Q2 order intake fell year on year after an extraordinary Q1, showing exactly how large projects bunch. My normalized assumption is for roughly EUR 3.0–3.6 billion of annual Hydro orders over the medium term, materially above the historical troughs but well below annualizing H1 2026's EUR 2.45 billion. That is a forecast, not company guidance.

Pulp & Paper tells the other half of the story. H1 2026 orders increased 8.8% to EUR 1.89 billion and revenue rose 6.6% to EUR 1.47 billion. Its 10.3% margin and 59% service mix make it a higher-quality earnings source than Hydro today. Customer spending is not collapsing: Packaging Corporation of America expects 2026 capital expenditure equivalent to roughly EUR 0.73–0.75 billion; Smurfit WestRock guides to roughly EUR 2.07–2.16 billion; International Paper spent roughly EUR 0.91 billion in H1 alone, using the 2026-08-18 ECB rate. But Valmet's H1 2026 orders fell 14%, and its biomaterial-services orders fell 9%, showing that the paper-equipment cycle is selective rather than broadly euphoric.

That changes the company's identity. ANDRITZ is no longer best understood simply as a pulp-equipment maker. Yet calling it a Hydro company with paper attached is premature. Hydro owns almost half the backlog but only about 23% of H1 revenue and has a lower margin. Pulp & Paper remains the largest revenue contributor, generates a much higher service mix and earns a double-digit margin. The appropriate description is a two-engine process-equipment group in transition: Pulp & Paper still anchors earnings quality, while Hydro increasingly determines growth and backlog duration.

H1 earnings quality was solid but exposes another important point. Comparable EBITA margin increased from 8.3% to 8.6% and Q2 reached 8.9%, yet net income increased only 4.9%, from EUR 191.6 million to EUR 201.0 million. The main leak was below operating profit: the financial result deteriorated to minus EUR 13.0 million from minus EUR 0.5 million, reflecting lower interest income, financing costs and the absence of a positive Armis valuation effect enjoyed in H1 2025. The tax rate barely changed, at 25.3% versus 25.5%.

The adjustment between reported and comparable profit also deserves scrutiny. H1 reported EBITA was EUR 312.2 million, or 8.1%, versus comparable EBITA around EUR 330 million, or 8.6%. The gap is concentrated particularly in Metals, where reported margin was 4.5% versus 6.0% comparable. Restructuring and capacity-adjustment items have appeared in several periods in this business. I therefore use comparable EBITA for operating trend analysis but reported cash flow and a normalized adjustment burden in valuation; treating every restructuring charge as economically nonexistent would overstate quality.

At EUR 79.50, ANDRITZ is around 16.8 times trailing earnings, roughly 10 times estimated trailing comparable EBITA on enterprise value, about 0.63 times backlog on equity value and approximately 0.58 times backlog on enterprise value. My estimated trailing free-cash-flow yield is around 6%. Those are neither distressed-project-contractor multiples nor the 25–35 times earnings valuations awarded to the best high-aftermarket industrial names. They price a business somewhere between those two identities, which is fundamentally where ANDRITZ sits.

The qualitative portrait is therefore: company in transition. The transformation is not an invented strategy slogan. It is visible in the revenue mix, with service moving from the mid-30s to mid-40s percent, and in the order book, with Hydro displacing Pulp & Paper as the dominant backlog. The question facing shareholders is whether Hydro can acquire the margin characteristics of a better-executed infrastructure franchise before its capital-heavy backlog overwhelms the mix benefit from expanding service.

Company and financial vertical

ANDRITZ's history explains why it can participate in such different investment cycles without looking like a collection of unrelated assets. Josef Körösi established the business in Graz in 1852 as an iron foundry and machine works. Early products included wire nails, chains and iron bars. The company gradually became an industrial-machinery producer, was converted into a stock corporation around 1900 and passed through several ownership regimes, including the Gutmann industrial group, post-war ownership by Creditanstalt and later Germany's AGIV.

The modern ANDRITZ was effectively forged around the turn of the millennium. A financial-investor group together with Wolfgang Leitner's Custos interests acquired control in 1999. The acquisition of Ahlström Machinery in 2000 transformed the scale of the pulp-and-paper franchise, and the company listed in Vienna on 25 June 2001, after an earlier postponement during the market turmoil following the technology bubble. Company share-price histories published later show a split-adjusted IPO reference of EUR 5.25 after the 2007 four-for-one split; the 2012 two-for-one split reduces the fully adjusted historical comparison to about EUR 2.63. This implies an original pre-split IPO price of EUR 21.00.

The sources reviewed do not give me a sufficiently reliable primary-source gross-proceeds figure for the 2001 offer, so I do not invent one. That is a small historical blind spot rather than a valuation issue.

The first post-IPO phase was built around consolidation. VA TECH Hydro in 2006 gave ANDRITZ the global hydropower platform that now drives the order book. Schuler, acquired in 2013, moved the company deeper into metal forming and automotive press systems. Xerium in 2018 expanded paper-machine fabrics and rolls, a particularly important transaction because consumables and replacement products increase recurring installed-base revenue. Sovema in 2022 extended the battery-production equipment portfolio. The acquisition history is therefore less random than it appears: management repeatedly bought either a new industrial platform or products that sit downstream of an installed machine and can be sold for decades after the initial capital order.

A useful way to divide the modern history is into four stages.

The 2000–2006 period created the current industrial skeleton. Ahlström Machinery established the scale in pulp, the IPO provided public-market currency, and VA TECH Hydro established the second major long-cycle platform. By the end of this phase, ANDRITZ was no longer an Austrian machinery company expanding abroad; it was an international system supplier exposed to global capital expenditure.

The 2007–2018 period was the expansion era. The company grew across paper, metal forming and consumables through Schuler and Xerium. That brought revenue scale and installed base, but also introduced heavier restructuring risk. The financial-crisis period showed how sharply project demand could turn. Later, Metals and Hydro both suffered periods where execution and cost overruns overwhelmed revenue growth. Historical company risk disclosures explicitly acknowledge “considerable losses” on certain projects and warn that long-term fixed-price contracts can produce margins materially different from initial estimates as material, labor and outsourced-component costs change.

The 2019–2021 phase was a repair period rather than a growth story. Management focused on structural cost reductions, particularly in Metals and Hydro. The pandemic then disrupted customer investment and execution. This matters in hindsight because today's 7.1% Hydro margin comes after years of work on project selection and cost discipline; it should not be treated as a natural margin that existed throughout the cycle. The 2020 management report explicitly identified optimization of Metals and Hydro cost structures as a central task for 2021.

The current phase began around 2022. Service intensity rose, profitability stabilized at higher levels, and the Hydro market began to inflect as renewable penetration increased the value of storage and grid-stabilization assets. ANDRITZ simultaneously kept making bolt-on acquisitions, including LDX Solutions and A.Celli-related paper assets in 2025. The result is a business with a record order book and more recurring sales, but with the capital component of that order book concentrated in precisely the division that has historically required the most margin repair.

The long-run service development shows how much the earnings engine has changed.

The history and H1 2026 service figures come from ANDRITZ's results presentation; H1 service revenue is my calculation from the disclosed 46% share.

The progression is not a straight line. Capital projects drove the 2022–23 revenue surge, temporarily pulling service's percentage downward even though service revenue itself kept growing. That distinction matters. The numerator increased from EUR 2.08 billion in 2018 to EUR 3.43 billion in 2025; the temporary declines in percentage mix were caused mainly by unusually strong capital sales, not erosion of the service franchise.

A five-year business-area service history and, more importantly, service EBITA are not disclosed as a clean comparable series. ANDRITZ provides current area service shares and has periodically disclosed them in interim presentations, but it does not give a five-year service-versus-capital profit split. I therefore refuse to assign a “known” service margin in the valuation. The SOTP later uses explicit assumed service and project margins, making that uncertainty visible rather than hiding it in a blended multiple.

Cash conversion is also better than one bad year can suggest. Company presentations emphasize that working-capital movements tied to project milestones make annual operating cash flow volatile, while the three-year rolling operating-cash-flow average rose from EUR 207 million in 2018 to EUR 574 million in 2024. Actual operating cash flow was EUR 529.6 million in 2021, EUR 710.8 million in 2022 and about EUR 375 million in 2023; 2025 generated EUR 652.7 million. Over a full five-year window the cumulative operating-cash-flow/net-income relationship is roughly 1.3 times, depending on treatment of reconstructed 2024 rolling data. I use the approximate ratio because it is more honest than presenting a falsely precise five-year number from differently formatted historical presentations.

The balance sheet needs a similar adjustment in interpretation. H1 2026 net liquidity was EUR 593 million. That is clearly preferable to leveraged project contracting, but part of the cash is linked economically to customer financing. ANDRITZ says trade working capital averages roughly 16% of revenue, while contract working capital can move between roughly 3% and 10% depending on customer prepayments and project progress. In H1 2026 alone, contract liabilities added EUR 316.7 million to working capital as strong orders generated advance payments.

The company's reported ROIC illustrates the effect. ANDRITZ calculated H1 2026 ROIC at 18.5% versus a 9.1% WACC, using its stated methodology and only 5% of revenue as operating cash. That is a healthy economic spread. Yet customer advances also reduce invested capital. A simple sensitivity makes the issue visible: annualizing H1 after-tax operating profit implies roughly EUR 470 million of NOPAT; an 18.5% return corresponds to around EUR 2.5 billion of invested capital. Adding even EUR 1 billion of customer financing back to that denominator would lower the illustrative return to around 13–14%. That is still respectable, but much less spectacular. The calculation is illustrative because ANDRITZ does not publish a “ROIC before customer advances” metric.

This is why the quality-screen interpretation needs care. High ROE, net cash and strong apparent capital efficiency contain a real operating advantage, particularly service economics, but also a project-business financing advantage. Customer prepayments are useful financing; they are not free equity. As backlog converts, contract liabilities can reverse into cash outflow.

Contract accounting adds another layer. ANDRITZ has historically disclosed contract liabilities associated with sales recognized over time and explicitly warns that many projects involve long-term fixed-price contracts. Under IFRS 15, over-time recognition means revenue and estimated profit are recognized as performance obligations are satisfied rather than waiting for final delivery. This gives a more economically representative revenue pattern for multi-year projects, but reported period margins depend on current estimates of total project cost. Cost revisions can therefore move earnings before the underlying project is completed.

The contract protections work both ways. ANDRITZ commits to performance guarantees and deadlines; failure can result in remedial work or damages, and serious non-performance can entitle a customer to terminate, return the plant and seek damages. The risk disclosure also states that outsourced parts may be quoted to customers at fixed prices before exact procurement costs are known. The company has experienced losses from these mechanisms before. The backlog is therefore high-quality evidence of future workload, but it is not equivalent to cash or guaranteed profit.

Capital allocation has been broadly rational. The company has used acquisitions to deepen installed bases, pays a meaningful dividend and has maintained net liquidity. The 2025 dividend of EUR 2.70 represents roughly 58% of the EUR 4.67 EPS. Net liquidity fell from EUR 713 million at the end of 2025 to EUR 593 million at mid-year, after the EUR 265 million dividend paid in the second quarter, EUR 131 million of capital expenditure and bolt-on M&A, with EUR 291 million of operating cash flow covering most of that outflow.

Governance is unusually owner-influenced for a European industrial company. Custos Privatstiftung and Wolfgang Leitner interests together control roughly 31.5%; Leitner chairs the supervisory board. Joachim Schönbeck is CEO, while Vanessa Hellwing became CFO in March 2025. The ownership creates meaningful long-term alignment, although a 30%-plus block also means minority shareholders have less influence than in a fully dispersed company. ANDRITZ states that it complies with the Austrian Corporate Governance Code.

The share-price history reflects the changing identity. ANDRITZ's own 2010 report showed a roughly 1,240% gain from the split-adjusted 2001 IPO price to year-end 2010, far ahead of the ATX. Thereafter the stock periodically de-rated when Hydro, Metals or paper capital spending disappointed and re-rated as margins and cash generation recovered. The latest phase has rewarded the service shift and Hydro backlog: the 52-week trading range through August 2026 was roughly EUR 57.95 to EUR 84.40, with the current EUR 79.50 only modestly below the recent high.

I cannot derive a defensible exact “historical valuation percentile” from the company's public filings because they do not provide a continuous 10-year P/E or EV/EBITA time series, and I will not reverse-engineer one from sparse observations. The evidence supports a narrower statement: today's roughly 17 times trailing earnings is above the valuation normally associated with a distressed project contractor, but remains far below the multiples of premium aftermarket industrials. That shift is rational to the extent that 44–46% service is durable.

Business model, moat, industry and peers

ANDRITZ's four divisions sit on different economic clocks.

Pulp & Paper sells complete pulp mills, recovery and chemical systems, fiber lines, paper and tissue machinery, automation, consumables, upgrades and service. Large new-build plants create order spikes, while the installed base produces a recurring stream of parts, fabrics, rolls, rebuilds and process optimization. H1 2026 revenue was EUR 1.47 billion, backlog EUR 3.28 billion, service share 59% and comparable EBITA margin 10.3%. This is currently the highest-quality major division because recurring revenue and project capability reinforce each other.

Hydropower sells turbines, generators, pumped-storage equipment, turbo-generators, control systems, rehabilitation and service. H1 revenue was EUR 873 million, backlog EUR 6.12 billion, service share 36% and comparable EBITA margin 7.1%. The value proposition depends heavily on engineering references, reliability and the ability to manage projects that remain in operation for decades. The installed base creates service opportunities, but the current book is still dominated by large capital work.

Metals combines processing lines with Schuler's forming systems. It is the weakest margin business: H1 comparable EBITA margin was 6.0% and reported margin 4.5%, against a 29% service share. Customer decisions are tied to steel, automotive and industrial investment cycles, and cost adjustment has been recurring enough that I give this division little valuation premium.

Environment & Energy is more heterogeneous. It includes clean-air technologies, separation, pumps, feed and biofuel systems and newer decarbonization applications such as carbon capture and hydrogen engineering. The business can earn a good margin, 10.0% comparable in H1 2026, but orders fell 13% and revenue 2% as customers postponed investment decisions. ANDRITZ's own 2025 CEO commentary pointed to legal uncertainty and weak economics around parts of Europe's hydrogen and power-to-X investment pipeline.

The current operating picture is compactly shown below.

ANDRITZ H1 2026 segment disclosures are the source.

The moat is strongest where three conditions overlap: a large installed base, process knowledge that materially affects a customer's output, and equipment whose replacement cycle is measured in decades. A pulp mill owner does not casually replace the process supplier around a critical recovery boiler, fiber line or automation system to save a small percentage on parts. A hydro owner makes an even more conservative choice when a turbine-generator set must operate reliably for decades. Those are genuine switching costs, although they are strongest in service and modernization rather than in greenfield tenders.

The second moat is global project reference density. ANDRITZ operates through more than 280 locations in over 80 countries. For highly engineered equipment, a prospective customer can see operating references using similar wood species, mill scale, metal process or turbine conditions. References reduce perceived technical risk, which matters when downtime can cost far more than the equipment price.

The third is the installed-base flywheel, although “flywheel” should not be confused with a network effect. Each capital installation creates service revenue, spare-parts demand and modernization possibilities for decades. Xerium and subsequent paper acquisitions increased precisely this opportunity. The evidence that the mechanism works is the rise in service revenue in absolute terms and as a percentage of group sales.

The moat is weaker in new-build pricing. Customers routinely put large projects out for competitive tender, and fixed-price execution risk can eliminate the economic benefit of a strong order book. No brand name prevents steel, engineering hours or subcontractor costs from exceeding the bid assumption. ANDRITZ's own historical loss disclosures confirm this.

On industry structure, the closest public peer is Valmet rather than a generic capital-goods name. Valmet sells pulp, paper and energy technology alongside automation, flow control and a large service operation. H1 2026 orders were EUR 2.47 billion, down 14%; sales were EUR 2.56 billion, up 6%; comparable EBITA was EUR 266 million at a 10.4% margin; backlog was EUR 4.26 billion. Its Process Performance Solutions segment generated an 18.6% comparable EBITA margin, illustrating how automation and recurring process products can earn much higher returns than large mill projects.

Valmet has become the cleaner pulp-and-paper process-control story. Its strength is a tighter portfolio around process industries and higher-margin automation. ANDRITZ is more diversified and has the much larger current backlog because Hydro has exploded. A paper customer may choose either for broad process capability; an investor should choose between them based partly on whether they want Hydro exposure. At the moment Valmet's capital orders are softer and ANDRITZ's Hydro orders much stronger, so headline group growth makes ANDRITZ look better than the pulp-equipment market actually is.

Metso is less direct in product overlap but is the best public benchmark for what ANDRITZ could become if installed-base economics dominate. H1 2026 aftermarket sales represented 56% of Metso revenue; adjusted EBITA margin was 16.4%. Minerals orders rose strongly, and net debt/EBITDA was 1.3 times. Metso demonstrates the valuation and margin benefits that can follow when customers consume replacement and wear products continuously rather than only during periodic plant rebuilds.

Kadant is an even purer benchmark. Its paper and industrial processing products are smaller-ticket and much more aftermarket-heavy. Parts and consumables were 68% of Q2 2026 revenue, down from 71% a year earlier as capital shipments grew. Q2 adjusted EBITDA margin reached 21.8%. Capital-project timing remains sluggish, but management said installed-base aftermarket demand drove record revenue. Kadant therefore shows the destination of a recurring industrial model rather than being a direct scale competitor for a complete pulp mill.

Voith is strategically more important than either Metso or Kadant because it competes directly in both paper and hydro. It is private, removing the possibility of a clean market-multiple comparison. Fiscal 2024/25 sales were EUR 4.85 billion and orders EUR 5.45 billion excluding discontinued Commercial Vehicles. Voith's historical Hydro results are also a warning: in fiscal 2023/24 it recorded additional provisions for expected cost increases on individual Hydro customer projects. In December 2025 Voith said it was examining a reduction of up to 2,500 positions to strengthen competitiveness.

That Voith experience is highly relevant. Pumped-storage demand may be structurally excellent while equipment suppliers still earn poor returns if contracts are priced too aggressively. Industry growth and supplier profitability are separate questions.

The public-peer numbers make the quality ladder visible.

† Kadant's U.S.-dollar market capitalization, revenue and backlog are converted at EUR 1 = USD 1.1576 on 2026-08-18. Its mix and profitability figures are second-quarter 2026 rather than first-half. ‡ Valmet figure is biomaterial-services revenue as a percentage of total group H1 sales and excludes Process Performance Solutions; it is therefore not directly comparable with ANDRITZ's service measure. § Kadant reports the cited profitability figure as adjusted EBITDA margin, not EBITA.

ANDRITZ and peer operating data come from their latest company releases; market multiples use market data around 2026-08-18.

The table explains the valuation better than a simple “ANDRITZ is cheaper” statement. Metso and Kadant earn materially higher margins and have more recurring revenue. Valmet is much closer to ANDRITZ's valuation and operating profile. ANDRITZ deserves some premium for net cash and Hydro growth, but a Kadant-like multiple would require a service mix and margin structure that ANDRITZ has not yet reached.

The pulp industry is currently mixed rather than depressed. International Paper, Smurfit WestRock and Packaging Corporation of America are still committing substantial capital. Converted at the August 18 ECB rate, Smurfit WestRock's 2026 capex plan is roughly EUR 2.1 billion; PCA's is around EUR 0.74 billion; International Paper spent around EUR 0.91 billion in H1. Those budgets support modernization and efficiency projects, although only a portion is addressable by ANDRITZ.

Valmet's softer biomaterial orders show why those budgets should not be translated one-for-one into equipment orders. Packaging producers spend on maintenance, conversions, environmental compliance, logistics, internal construction and assets where neither Valmet nor ANDRITZ participates. What matters for ANDRITZ is the subset involving fiber lines, paper machines, recovery systems, automation and mill modernization.

Hydro is at a different point in its cycle. IHA reports 28 GW of global new hydro capacity commissioned in 2025, of which a record 11.7 GW was pumped storage. Europe alone had a pumped-storage development pipeline around 52.9 GW in 2025, while China's construction pipeline exceeds 200 GW. Renewables make the need for dispatchable storage and grid services more structural than the old model of building hydro simply to add generation.

Yet pumped storage faces permitting, geology, cost overruns, long construction schedules and competition from rapidly falling battery-storage costs. Reuters and the IEA have both documented the storage-cost decline and the long-duration role that hydro still fills. The likely future is coexistence: batteries dominate shorter-duration, faster-deployment applications while pumped hydro remains attractive for very large, long-duration systems where geography and regulation permit it.

Environment & Energy's weakness alongside record Hydro orders therefore makes economic sense. Utilities and governments can approve grid-stability infrastructure under long-lived regulated or contracted frameworks even while industrial firms delay hydrogen, filtration, separation or emissions-related investments because commodity prices, policy support and project economics are uncertain. These are different investment cycles sitting inside one group.

The ecological niche is unusual. ANDRITZ is a broad process-plant supplier with two unusually strong installed-base franchises, paper and hydro. It takes profit pools from specialized OEMs and engineering contractors by offering whole process islands or complete plants, and then keeps part of the downstream service economics. Its greatest strategic danger comes from competitors that can separate those layers: Valmet can take high-value paper/automation work, Voith can take paper and hydro, and specialist aftermarket firms can take the recurring profit pool without accepting the risk of a complete fixed-price plant.

Current fundamentals and valuation

The last four quarters show a business moving from revenue digestion back toward backlog expansion.

Q3 2025 orders increased 14.5% while revenue remained below the prior year because earlier order weakness was still moving through the income statement. Full-year 2025 order intake reached EUR 8.91 billion, backlog EUR 10.46 billion and revenue EUR 7.88 billion. Comparable EBITA margin held at 8.9%, despite lower sales. Q1 2026 then produced a record EUR 3.6 billion of orders, and H1 reached EUR 5.92 billion. The leading indicator has therefore been strengthening for more than a single quarter.

H1 2026 confirms that revenue is beginning to follow, but gradually. Revenue rose 5.2%, comparable EBITA margin increased 30 basis points to 8.6%, and net income grew 4.9%. Management guides to 2026 revenue of EUR 8.0–8.3 billion and comparable EBITA margin of 8.7–9.1%. That means the company itself is not extrapolating the 25% order surge into equivalent near-term sales growth.

My backlog conversion framework is deliberately more conservative than an order-growth extrapolation.

Backlog/revenue ratios are calculated from H1 disclosures. Conversion percentages and horizons are my analytical assumptions, anchored to ANDRITZ's statement that historical aggregate conversion was around two-thirds within a year but fell below 60% as Hydro grew.

This creates the 2027–29 revenue runway, but also accounting risk. Multi-year fixed-price work is sensitive to wage inflation, steel and equipment procurement, subcontractor prices, geology and on-site delays. Because a meaningful portion is recognized over time, updated total-cost estimates can alter the margin recognized before completion. That is why I place more weight on multi-year segment-margin development and cash conversion than on any single quarterly EBITA print.

The cash flow in H1 2026 was sound but not spectacular. Operating cash flow was about EUR 291 million and free cash flow EUR 160 million after EUR 131 million of capex. Inventory and payables consumed cash, while contract liabilities added EUR 317 million. This is exactly what a project-cycle cash-flow statement looks like: customer advances and milestone timing dominate quarter-to-quarter movement.

At the current price, a simple valuation snapshot is:

Price and market data are as of 2026-08-18. Enterprise-value, backlog and FCF multiples are my calculations using H1 net liquidity, the EUR 12.60 billion backlog and reported cash flow.

Backlog multiples are descriptive rather than intrinsically meaningful. A euro of Pulp service backlog is economically worth much more than a euro of low-margin fixed-price Hydro EPC backlog. This is precisely why a blended EV/backlog number cannot establish cheapness.

The owner-earnings exercise gives a more useful cross-check. ANDRITZ does not disclose maintenance capex separately from growth capex. Total 2025 capex was EUR 269.5 million. I assume 65–75% is maintenance, with 70% as the midpoint; acquisitions are excluded. Adding depreciation and amortization back to net income and deducting the assumed maintenance component gives normalized owner earnings around EUR 0.50–0.53 billion. The resulting owner-earnings multiple is about 15–16 times, or a 6.3–6.7% yield.

That is close enough to the headline 16.8 times P/E that accounting earnings are not overstating normalized owner earnings by the 30% threshold that would force me to abandon P/E altogether. The bigger valuation uncertainty is the service/project split, not maintenance capex.

The SOTP therefore separates the two economic streams. No service EBITA split is reported, so the margins below are assumptions, openly shown. The “service” multiple reflects recurring aftermarket economics; “project” covers capital equipment and long-cycle execution.

This is valuation-scenario analysis within a research framework, not investment advice.

The conservative case roughly describes a backlog that converts but fails to improve economically: Hydro remains near the bottom of its target range, service stops gaining mix and capital projects receive only contractor-like multiples. The base case requires Hydro to progress toward 8%, continued service growth and no material fixed-price losses. The optimistic case effectively assumes that today's Hydro boom becomes a multi-year structural cycle, while service reaches high-40s mix and project execution approaches the upper end of management's targeted profitability.

The current price is therefore below my base SOTP but above the conservative value. That is a very different proposition from “cheap at 17 times earnings.”

Current market capitalization also lets us reverse the SOTP. Under my base operating assumptions, if the market gives the project business 7 times EBITA, the current enterprise value implies only about 10.5–11 times service EBITA rather than the 13 times I assign in the base case. Alternatively, awarding the service operation 13 times would imply an extremely low multiple on projects. The gap says the market is either skeptical that the assumed service margin is achievable, skeptical about Hydro execution, or both.

That skepticism is reasonable. The service backlog is only 26% of total backlog. A large portion of the visible future revenue is capital-heavy. The bull case therefore depends on service orders replenishing continuously outside the long-term backlog and Hydro's own margin catching up.

Peer valuation reinforces that conclusion. Valmet's trailing P/E is around 16–17 times, very close to ANDRITZ. Metso trades around the high-20s and Kadant above 30 times because their recurring mix and operating margins are substantially higher. ANDRITZ should close part of that valuation gap only after the earnings mix has actually improved; using Kadant's multiple now would pre-spend the transformation.

The market's main expectation gap in the next two earnings prints is likely to be Hydro margin rather than order intake. A fall in order intake from H1's extraordinary level should not by itself invalidate the thesis; large Hydro orders are inherently lumpy. A Hydro margin that falls back toward 6%, by contrast, would challenge the economic value of the EUR 6.1 billion backlog. Conversely, an 8% margin with continued service growth would make the current mix dilution largely disappear.

The second expectation gap is cash. Investors may treat EUR 593 million of net liquidity as excess balance-sheet value. A rapid increase in contract assets and decline in customer advances as Hydro work converts could consume part of it. I therefore include only EUR 0.4–0.6 billion of normalized cash in the scenarios rather than extrapolating the largest historical cash balance.

The independent margin-of-safety check is less flattering than the base upside.

Current EUR 79.50 is about 18% above the EUR 67.6 conservative SOTP. On that test, the margin of safety is zero.

The most fragile base-case assumption is the multiple awarded to the recurring service stream, rather than the revenue forecast. Cutting my 13 times service EBITA multiple to 70%, or 9.1 times, while leaving the other base assumptions unchanged reduces fair value from about EUR 91.9 to roughly EUR 72.1. That is below the current price.

If earnings remain flat for three years and the valuation multiple also remains unchanged, the primary return is the EUR 2.70 dividend, currently about 3.4% a year. Austria's 10-year government-bond yield was about 3.48% on 2026-08-18. On that deliberately static test, there is no margin of safety at this buy price.

Margin-of-safety sufficiency verdict: none.

That does not say intrinsic value equals EUR 67.6. It says the current share price requires at least some of the service and Hydro improvement to occur. The market is not charging an obviously excessive amount for that possibility, but neither is it giving the investor the downside case for free.

Risks, catalysts and tracking

The largest permanent-loss risk is fixed-price Hydro execution. I assign medium probability and high impact. The transmission path is direct: material, subcontracting or civil-engineering cost overruns reduce Hydro margin; because almost half the backlog is Hydro, the hit moves group EBITA; recurring restructuring or project provisions then undermine the market's belief that Hydro deserves a higher multiple. ANDRITZ's own historical risk disclosure and Voith's recent Hydro provisions show that this is an industry reality rather than a hypothetical accounting concern.

The observable indicator is the Hydro comparable EBITA margin. A sustained move above 8% would greatly reduce the risk; two quarters below 6.5% while the backlog remains above EUR 5 billion would be a major warning.

The second risk is confusing project bunching with structural growth. Probability is high that order growth slows sharply; impact is medium unless the absolute level collapses. H1 2026's 81.8% Hydro increase cannot repeat indefinitely, and Q2 already showed the effect of the prior-year comparison. What matters is whether normalized annual Hydro intake settles around EUR 3 billion-plus or falls back toward historical troughs after the current group of pumped-storage awards.

The structural demand evidence makes a complete reversal less likely: record pumped-storage commissioning, large Chinese construction pipelines and growing European development activity all support multi-year demand. The risk is timing and price discipline, not the disappearance of grid-storage need.

The third risk is service-mix disappointment. I assign medium probability and high valuation impact. Today's 46% group service share is the strongest evidence for a quality re-rating. If that figure falls sustainably below 42–43% because capital expansion outruns aftermarket growth or customers use alternative service providers, the multiple should migrate back toward a conventional project contractor. Kadant and Metso show how large the valuation gap can be between an aftermarket-heavy business and a capital-heavy one.

The fourth risk is paper-capex weakness. Probability is medium, impact medium. Current U.S. customer budgets remain substantial, but Valmet's H1 orders and biomaterial-service orders were down. A weaker containerboard/pulp price environment can delay major mills for years. Pulp & Paper's unusually high service mix cushions the decline, but a prolonged capital pause would remove the group's highest-margin large project business just as Hydro increases its share.

The fifth is working-capital reversal. Probability is high; impact on intrinsic value is low-to-medium unless combined with project problems. Strong orders generate customer prepayments and contract liabilities; project execution unwinds them. A declining net-cash balance during rapid backlog conversion would therefore be normal. It becomes problematic when cash falls at the same time that EBITA margins fall, contract assets rise disproportionately and provisions increase. That combination would suggest cost overruns rather than harmless milestone timing.

Environment & Energy represents a smaller but useful macro warning. Customer decisions are currently delayed across several industries, while European policy uncertainty has particularly affected hydrogen and related investment economics. The probability of continuing softness into 2027 is medium, but group impact is moderate because E&E is only around 11% of backlog. A rebound would add attractive margin because the division currently earns around 10%.

Positive catalysts over the next year are straightforward: Hydro margin reaching 8% without loss of order discipline, sustained group service share around 46–48%, stronger Pulp & Paper capital awards, an E&E order recovery and free cash flow that rises as revenue catches up with the order book. The most important negative catalysts are the mirror image: Hydro project provisions, a drop in service share, order cancellations or prolonged deferrals, and a cash outflow that cannot be explained by healthy project conversion.

The next scheduled results publication is 2026-10-29, when ANDRITZ will release Q3/Q1–Q3 2026 results.

The dashboard thresholds are analytical monitoring levels rather than company targets, except Hydro's disclosed 7–9% 2027 margin range and the earnings date. Current operating values come from H1 2026 disclosures.

The reason to watch service share and Hydro margin together is that either can offset the other. A 50% Hydro backlog is not necessarily bad if that division earns 8.5–9% and throws off a growing service stream. A 50% Hydro backlog at 6% creates a structurally weaker group even if revenue grows.

Similarly, the order/revenue ratio is useful only over multiple quarters. H1's 1.54 times is too high to normalize. A drop toward one would be expected as large orders become harder comparisons. A ratio below 0.9 for two or more quarters while backlog conversion accelerates would suggest the peak has passed.

Research uncertainties remain material in four places.

First, ANDRITZ does not disclose EBITA separately for service and capital. That is the largest valuation blind spot because the entire SOTP rests on assumed rather than reported stream margins.

Second, the company does not publish a consistently comparable five-year service-margin series by business area. Current revenue shares are clear; historical segment profitability by service versus project is not.

Third, detailed contractual cancellation-for-convenience clauses vary by project and are not publicly disclosed. The company does disclose termination rights following serious performance failures, but the exact economic firmness of every backlog euro cannot be independently audited.

Fourth, no public primary-source dataset gives a continuous 10-year valuation percentile, so I do not claim false precision about whether 16.8 times earnings is, for example, the 63rd or 71st percentile.

The source hierarchy for this report was ANDRITZ's H1 2026 results presentation and release, the 2025 annual report and historical financial reports; Vienna Stock Exchange for the share; Valmet, Metso, Kadant and Voith corporate disclosures for peers; IHA and IEA-related material for hydropower; customer filings/results for paper capex; and ECB data for currency conversion.

Cross-synthesis and final research conclusion

Looking vertically, the capability ANDRITZ has genuinely proven over decades is not simply manufacturing machinery. It has repeatedly learned to absorb complex engineering platforms, build a global installed base and then convert that base into service revenue. The acquisitions of Ahlström Machinery, VA TECH Hydro, Schuler and Xerium created most of today's industrial shape. Some integrations produced painful execution periods, particularly where large fixed-price projects were involved, but the group survived those cycles without permanent balance-sheet impairment and emerged with a larger recurring revenue stream.

The service data provide the cleanest evidence that this capability is durable. Revenue from service rose from roughly EUR 2.1 billion in 2018 to EUR 3.4 billion in 2025 even though total revenue moved through several industrial cycles. The percentage increased to 46% in H1 2026. Pulp & Paper has already crossed into a majority-service model at 59%. Those figures represent a real change in business quality.

Past success nevertheless depended partly on era tailwinds and acquisition timing. The early-2000s consolidation created scale at a favorable point in globalization. Emerging-market pulp capacity, Chinese industrialization and later environmental investment enlarged the addressable market. The current Hydro cycle is another large tailwind, driven by renewables and grid storage. Management deserves credit for owning the capability when the cycle arrived; it did not create the need for grid-scale storage.

The next three years will tell us how much of the return is management skill rather than cycle. A record Hydro backlog only creates value if it is priced correctly and executed within cost. Voith's Hydro provisions prove that a competitor can participate in the same excellent end market and still suffer project economics. ANDRITZ's own historic project-loss disclosures make the same point.

Horizontally, ANDRITZ sits between two industrial archetypes. Valmet is the closest operational peer and trades at almost the same earnings multiple. Metso and Kadant represent the higher-quality recurring model: more aftermarket, much higher margins and correspondingly higher multiples. Voith is the most direct strategic competitor but lacks a public valuation.

ANDRITZ's real advantage versus Valmet is diversification into a Hydro market that is currently much stronger than pulp-equipment capital ordering. Its disadvantage is that Hydro is lower margin and much longer duration. Against Metso and Kadant, ANDRITZ has a larger order book and more whole-plant capability but meaningfully weaker recurring economics. The appropriate valuation therefore lies between a contractor multiple and a premium aftermarket multiple.

The market appears to understand much of this. At EUR 79.50, investors are not paying 25–30 times earnings for a flawless transformation. The stock trades around 17 times trailing earnings and roughly 10 times comparable EBITA on enterprise value. That leaves room for upside if service keeps compounding and Hydro reaches 8–9%. Yet the price is already above my conservative SOTP, so some success is embedded.

The misjudgment I think is most likely is a tendency to focus on the absolute EUR 12.6 billion backlog without distinguishing its composition. There are two opposite errors.

The bullish error is to treat every backlog euro as if it carried today's 8.6% group margin. Current segment arithmetic says closer to 8.1% before accounting for the capital-heavy nature of the backlog. A service-heavy Pulp euro and a multi-year Hydro project euro are economically different.

The bearish error is to conclude from the 8.1% arithmetic that group margins must fall. New service orders do not need to sit in the long-term backlog before becoming revenue. Group service revenue already represents 46% of sales while service is only around 26% of backlog. If that short-cycle stream keeps growing and Hydro reaches 8%, the mix dilution largely disappears.

This tension means the Hydro margin is more informative than headline order growth. At 7.1%, the backlog is a volume story. At 8%, it becomes approximately margin-neutral. At 9%, the largest order book in company history can become a genuine earnings-quality upgrade.

For the next 12 months, the critical variables are Hydro execution, service mix and working capital. Management has already given a fairly modest EUR 8.0–8.3 billion revenue guide relative to the size of the order book, so a simple revenue beat would matter less than a clean margin and cash conversion.

For the next three years, the critical variable is whether pumped storage converts from an order boom into a durable installed-base franchise. Every project commissioned today should generate decades of modernization and service opportunities. The strongest long-term bull case requires the current projects to produce an aftermarket similar to the one Pulp & Paper already enjoys, rather than Hydro capital orders continuing to rise 80%.

For the five-year horizon, the identity question dominates. If service reaches or exceeds half of group revenue while Hydro sustainably earns 8–9%, ANDRITZ deserves to be valued much closer to higher-quality industrial aftermarket peers. If service stalls in the low-40s and Hydro remains a 6–7% project business, the current transition will have produced more revenue visibility without much more economic quality.

The company becomes a substantially better investment under one of two conditions: either the operating proof improves enough that the conservative value rises, or the share price falls far enough to supply a true margin of safety. Today the first process is underway, while the second condition is absent.

Core bull reasons:

• Group service share has risen from 35% in 2018 to 46% in H1 2026, while Pulp & Paper has reached 59%, providing a tangible recurring-revenue base beneath the project cycle.
• The EUR 12.6 billion backlog is a record and provides multiple years of visibility, with Hydro alone holding EUR 6.1 billion.
• Pumped-storage demand has structural support: record global installations in 2025 and hundreds of gigawatts under development make the Hydro opportunity broader than a single group of projects.
• Hydro margin has already reached 7.1%, the bottom of management's 7–9% 2027 target, so only another roughly 90 basis points are needed for the current backlog mix to become approximately margin-neutral.
• The balance sheet retains roughly EUR 0.6 billion net liquidity despite dividends, capex and acquisitions, giving management room to absorb project volatility.

Core bear reasons:

• Applying current business-area margins to backlog gives only about an 8.1% implied EBITA margin, below the current 8.6% group level, because nearly half the backlog sits in lower-margin Hydro.
• The backlog itself is only about 26% service against 46% service in current revenue, so the most visible future revenue is disproportionately capital-heavy.
• Hydro's 81.8% H1 order growth is highly lumpy and cannot be annualized; a post-award air pocket could emerge even if the long-term pumped-storage market remains attractive.
• ANDRITZ retains long-term fixed-price exposure, and both its own history and Voith's recent Hydro provisions show that cost overruns can destroy the apparent value of a strong order book.
• At EUR 79.50 the stock trades around 18% above my conservative SOTP, leaving no conventional margin of safety despite apparent base-case upside.

The first pre-mortem script is a Hydro execution failure. Assume that during 2027–28 several Indian and European pumped-storage contracts encounter material, civil-engineering and subcontracting inflation. Voith and other suppliers keep tender pricing aggressive, preventing repricing of new work. ANDRITZ Hydro margin falls from 7.1% to 5%, group EBITA margin slips toward 7.2%, restructuring and project provisions reappear and EPS falls toward EUR 3.7. If the market then reclassifies ANDRITZ as a conventional cyclical contractor at 10 times earnings, the shares could trade around EUR 37–40. That is roughly a 50% loss from today's price despite a still-large order book.

The second script is less dramatic operationally but equally damaging to the valuation. Global pulp and containerboard customers enter a 2027 capex pause, Valmet and specialist service firms compete more aggressively for aftermarket work, and ANDRITZ's service share falls from 46% to around 42%. Hydro orders normalize sharply after the current pumped-storage awards, while its margin never exceeds 7%. Group EPS stagnates near EUR 4.0–4.2. A 12 times P/E would then imply roughly EUR 48–50 a share, a decline near 40% before dividends.

Those are stress cases rather than forecasts. Their purpose is to identify the combination that can create permanent loss: lower margin plus lower recurring quality plus multiple compression. Any one variable by itself is less dangerous.

The final research judgment is that ANDRITZ has become a better company faster than its conventional “European machinery” label suggests. A 46% service share, double-digit Pulp & Paper profitability, net liquidity and a structural pumped-storage opportunity distinguish it from a low-quality project contractor. The diversification also works: weakness in E&E and softer parts of pulp capital spending are being offset by Hydro rather than hitting the whole portfolio simultaneously.

The price, however, no longer offers the asymmetry I would require for a Buy. My EUR 91.9 base value provides upside, but it relies on service receiving a quality multiple and capital-project margins improving. The EUR 67.6 conservative value sits below the market price. A three-year flat-earnings case yields little more than the dividend, approximately the Austrian 10-year government-bond yield. The stock is therefore fairly investable for an existing long-term holder but lacks a clean downside cushion for a new position at EUR 79.50.

The right classification is a quality-improving cyclical company whose backlog is stronger than its current margin mix. The evidence that would change my judgment upward is straightforward: Hydro sustains at least 8%, service stays above 46% and free cash flow converts without a large erosion of net cash. A move to 9% Hydro margins would materially raise my valuation because it would resolve the central mix problem. The evidence that would make me more negative is Hydro below 6.5%, service below 42% or a working-capital unwind accompanied by project provisions.

【Company-profile scores】

• Fundamental quality: high
• Growth: medium
• Moat: medium
• Financial soundness: strong
• Management credibility: high
• Valuation attractiveness: medium
• Risk level: medium
• Suitable investor type: cyclical

【Investment rating】

• Rating: Hold
• One-line thesis: Record backlog gives multi-year visibility, but its Hydro-heavy mix dilutes current margins and EUR 79.50 offers no discount to the conservative SOTP.
• Ideal buy price: see the dedicated line below.
• Acceptable hold price: EUR 78–105, corresponding to roughly ±15% around the EUR 91.9 base SOTP.
• Clearly overvalued price: EUR 133–145, beginning more than 10% above the EUR 120.5 optimistic SOTP.
• Current-price classification: acceptable hold.
• Whether to wait for a better price: yes. For a full Buy-level margin of safety I would require the EUR 50–54 range, provided Hydro margin remains at least 7%, group service share remains above 44% and no material backlog loss emerges. Waiting sacrifices roughly a 3.4% dividend yield and risks missing a re-rating if Hydro reaches 8–9%.
• Target holding horizon: 3–5 years.
• Expected annualized return: conservative about -1.4%; base about 7.6%; optimistic about 17.2% over three years including an assumed EUR 2.70 annual dividend.
• Max-loss risk: roughly 45–55% in the combined pre-mortem of Hydro margin falling toward 5%, EPS falling toward EUR 3.7–4.0 and the market compressing the P/E to 10–12 times.
• Reassessment-trigger signals: Hydro comparable EBITA below 6.5% for two consecutive quarters; group service share below 42%; order intake/revenue below 0.9 for two quarters after backlog begins shrinking; rolling FCF/net income below 0.8 with rising contract assets; or net liquidity turning negative without a major acquisition.

【Ideal Buy Price】50–54 EUR

Basis: the EUR 50–54 range is at least 20% below the EUR 67.6 conservative SOTP and provides a material cushion against Hydro execution risk rather than relying on the base-case service re-rating.

【Valuation Range】

• current: 79.50 EUR (close as of 2026-08-18)
• bear (conservative · ideal buy zone): [50, 54]
• base (fair · acceptable hold zone): [78, 105]
• bull (optimistic · above the clearly-overvalued line): [133, 145]

Other tickers mentioned

• VALMT.HE: Valmet is the closest listed peer in pulp-and-paper equipment, automation and lifecycle service.
• METSO.HE: Metso is the key benchmark for a higher-aftermarket industrial model and the margins such a mix can support.
• KAI.US: Kadant is the premium recurring-parts benchmark, with roughly 68% of Q2 2026 revenue from parts and consumables.
• IP.US: International Paper's capital spending is a direct demand signal for pulp, packaging and mill-modernization equipment.
• SW.US: Smurfit WestRock's large 2026 capex budget tests the health of downstream packaging investment.
• PKG.US: Packaging Corporation of America provides another direct read on North American containerboard capital expenditure.
• GEV.US: GE Vernova is a relevant power-equipment reference for the broader grid-stability investment cycle.
• ENR.DE: Siemens Energy is a broader power-equipment comparator for generation and grid infrastructure economics.
• VWS.CO: Vestas is relevant to the renewable build-out whose intermittency increases demand for storage and grid balancing.

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