Linde plc Ordinary Shares · Moat analysis

Report previewLIN.USReport date Last updated
← Back to reports
Report contents
01

Industry landscape

Where the company competes: density inside industrial clusters, not global share

Linde makes and delivers industrial gases. It separates air into oxygen, nitrogen and argon, produces hydrogen, carbon monoxide, carbon dioxide and helium, and sells those molecules through three delivery models: a plant built on the customer's site and connected by pipeline, cryogenic tanker loads of liquid, and cylinders or packs for smaller users. The Engineering segment designs and builds air separation and hydrogen plants, both for Linde and for third parties. Upstream, the business is a large buyer of electricity, natural gas, tanks, trailers and compressors. Downstream, it sells into steel, refining, chemicals, electronics, healthcare, glass, food and beverage, and general manufacturing. Linde reports along regional lines — the Americas, EMEA and APAC, plus Engineering — which follows where the customers sit. The company is incorporated in Ireland, with principal offices in Danbury, Connecticut and Woking, England, per its 2025 Form 10-K.

The industry is usually described as one global market. Operationally it is four overlapping markets, each with a different geographic boundary, and that split determines how much pricing power Linde actually has at any given site.

Four arenas, four different borders

On-site supply is the tightest. A plant sits inside a refinery, steel mill, float-glass line or electronics fab, connected by a short pipeline under a contract that commonly runs 15 to 20 years on take-or-pay terms. The competitive radius here is a few hundred meters. Once the unit is built, no rival can serve that customer; the contest happened years earlier, at the bidding stage.

Merchant liquid is regional. A tanker can economically reach roughly 200 to 300 miles, so the relevant market is a metro area or an industrial corridor, not a country. Packaged and specialty gases are more local still and more fragmented, with independent distributors and welding suppliers in the chain. Large projects — ammonia, low-carbon hydrogen, carbon capture — are bid globally by a short list of firms and priced off long-term offtake.

The practical consequence is that Linde's worldwide revenue share says less than its share of installed plants inside a given cluster. In the US Gulf Coast, northern Europe and coastal China, its pipelines and plants function as local infrastructure, and a competitor would have to duplicate that network to take a customer.

Market structure: four global names, two or three bidders per site

No company publishes global market share, so the figures below are rough estimates built from each firm's reported gas revenue. Treat the ranges as indicative; the ranking is the point.

CompanyHQEst. share of global industrial gas revenue (%)Footprint
LindeIreland / US-UK20–25Global, strongest in the Americas
Air LiquideFrance18–22Global, strong in Europe and Asia
Air ProductsUS9–12Global, skew to large projects
Taiyo Nippon SansoJapan6–8Asia, US, Europe
MesserGermany3–5Europe, China, Americas
Regional and state-owned (Baowu, Air Water, Gulf Cryo and others)—remainderNational

Globally that is concentrated enough to matter: four firms account for more than half of industry revenue, and they meet each other in most large tenders. Locally it is tighter still, because a customer choosing an on-site supplier usually has two or three credible bidders, not ten.

The barrier is capital plus contract, not chemistry. A world-scale air separation unit costs hundreds of millions of dollars, needs power, permits and pipeline rights of way, and cannot be justified without an anchor customer signed up before construction. The incumbent already holds that customer. Add density economics — every new plant in a region spreads fixed distribution costs across more molecules — and the entrant faces a cost disadvantage before it produces anything. So entry barriers are high for on-site and pipeline business, medium for merchant liquid, and low in packaged gas, where trucks and cylinders are the only real assets.

The weakest parts of the perimeter are the ones growing fastest in headline value. Low-carbon hydrogen and carbon capture projects attract subsidy-backed entrants and pull Air Products and Air Liquide into the same tenders, and the returns depend on offtake contracts that have not yet been signed. Whether Linde's site-level advantages carry into those projects, or whether they become a capital-intensive commodity business with more bidders, is the open question this section leaves for the moat assessment.

02

Competitive advantage

Does the company have a moat? Yes, and it sits inside customer fence lines

Overall moat rating: Fairly Strong. Linde's protection is local, not global. The relevant unit is a plant built inside a customer's site and connected by a short pipeline, plus the depot-and-truck network serving the industrial corridor around it. Judged at that level, the framework's two tests come out positive, though the return test is the weaker of the two.

Concentration first. Four suppliers account for more than half of global industrial gas revenue, and Linde is one of them. Locally the count is smaller: a refinery, a float-glass line or a large electronics fab choosing an on-site supplier in 2025 had two or three credible bidders. That structure has not changed shape in five years, and it is what keeps a barrier standing.

Market position: Strong in on-site supply, Moderate in merchant and packaged gas. No producer publishes a global share, so the figure is an estimate built from reported gas revenue: Linde holds roughly 20-25% of world industrial gas revenue, level with Air Liquide and ahead of Air Products and Taiyo Nippon Sanso. It ranks first or second depending on how engineering and equipment revenue is counted. The more telling number is the bid list at a specific site, where Linde's installed base in the US Gulf Coast, northern Europe and coastal China makes it the incumbent in most large tenders.

Stability: Fairly Strong. Across 2021-2025 the estimated global share held in the low-to-mid 20s and the four-firm structure did not shift. That falls inside the rough boundaries of the framework: a five-year move of more than five percentage points would suggest no barrier, and one under two points would suggest a strong one. The honest limit is disclosure. Annual company-by-company share data are not published, so the reading rests on a stable ranking and on revenue that grew from $30.8 billion in 2021 to $34.0 billion in 2025 at roughly the pace of industrial production.

Fiscal yearRevenue (USD bn)Gross margin (%)Net margin (%)ROE (%)
202130.828.012.48.4
202233.429.112.49.9
202332.935.118.915.5
202433.036.619.916.9
202534.043.320.318.0

Revenue sat almost flat from 2022 to 2025 while gross margin gained about 14 points and net margin eight. Because volume barely moved, the improvement came from price and cost, and the reported figures do not split the two. That matters for the moat question: a margin gain achieved without volume growth is consistent with pricing power on a fixed asset base, but it is also consistent with a one-time cost program, and five years of data cannot separate those explanations.

On total invested capital the return is more modest. After-tax ROIC, computed as EBIT taxed at the reported effective rate over equity plus debt less cash, comes to roughly 12-13% in each of 2023, 2024 and 2025. That sits below the 15-25% band the framework treats as clear evidence of advantage, and well above the 6-8% band that signals none. The gap has a specific cause: $27.9 billion of an $86.8 billion balance sheet is goodwill, largely from the 2018 all-share merger of Linde AG and Praxair that created the present company. A new plant carries no such goodwill, so the return on the capital an entrant would actually spend is far higher than 12%. Both figures answer real questions. The goodwill-inclusive number is what shareholders earned on the price paid for the business; the tangible number is what makes the incumbent hard to displace.

Profitability: Fairly Strong. Returns clear the cost of capital with room to spare, margins have widened for three consecutive years, and the pattern is not a single good year. They fall short of the top band on a total-capital basis, which is what keeps this from a Strong rating. Averaging profitability and market position, with the extra weight on profitability the framework calls for, the overall rating is Fairly Strong.

Two developments would change that view. A full-cycle ROIC that settled below 10% would undercut the return test, and evidence that on-site contracts are renewing at materially thinner spreads would undercut the local concentration that produces it. Neither is visible in the numbers through 2025.

Where the advantage comes from and how long it can last: contracts signed before the steel goes in

Three sources do the work here, and they are not equal. Customer captivity is the strongest, local scale reinforces it, and a cost and technology edge sits underneath at a lower order of strength. Together, captivity plus scale is the combination the framework treats as most durable, but both parts are bounded by geography and by contract length.

Customer captivity: Fairly Strong

The sequence starts before the plant exists. Oxygen and nitrogen are difficult to store and expensive to move in volume, so a large customer takes them through a dedicated pipeline from a unit built on its own land, under a contract commonly running 15 to 20 years on take-or-pay terms. Commitment precedes construction, and that ordering is where the advantage begins.

Once the unit is running, the customer's alternative is not a cheaper supplier. It is a new plant costing hundreds of millions of dollars, a multi-year permit and construction cycle, and a shutdown of the production line the gas feeds. Take-or-pay terms mean the customer pays for reserved capacity whether or not it runs, which insulates Linde's revenue from the customer's production cycle in a way a spot supplier's revenue is not insulated. The gas itself is usually a small share of the customer's total cost, whether oxygen for a glass furnace or a steel converter, while an interruption is very expensive. Low price sensitivity combined with high cost of interruption is what converts a supply relationship into captivity; brand preference on its own would not.

The limit is the calendar. Contracts expire, and at renewal the customer has years of operating data on the plant and two or three credible bidders to play against the incumbent. Linde's leverage is highest mid-contract and lowest at the end. Captivity protects the installed base and says nothing about the next plant. Merchant liquid and packaged gas customers sign nothing long-term and can switch on price within weeks, which is why the same company earns very different returns across its own portfolio. That boundary keeps the rating at Fairly Strong rather than Strong.

Scale advantage: Fairly Strong inside a corridor

The fixed cost base is large and lumpy: the plant, depots, tankers, drivers, maintenance engineers and safety systems. Every additional customer in a corridor where Linde already delivers spreads those costs over more molecules, and the truck is often going that way already. The advantage comes from relative share in that corridor, not from national or global size.

Entry economics follow directly. A rival without an anchor customer cannot justify a plant, and without a plant it cannot win the anchor customer. If it wins one anyway, it still builds its depot and fleet before it has the volume to fill them, so its unit cost is higher at every volume short of scale. That is a genuine supply-side barrier. It depends, however, on captivity doing half the work: scale without a locked-in anchor is contestable, because an entrant can discount to buy volume and grow into its fixed costs. The long-term contract is what denies it that volume for a decade and a half.

The limits are as important as the strength. Packaged gas carries little fixed cost beyond trucks and cylinders, so scale there is thin and the segment is fragmented among independent distributors. Merchant liquid is regional, and a single-plant independent can undercut on spot volume when its plant is running. The scale advantage is real in on-site and pipeline corridors and weak almost everywhere else.

Cost and technology: Moderate

Linde's Engineering segment designs and builds air separation and hydrogen plants, for itself and for third parties, and repeated designs, operating data and maintenance experience lower capex and improve on-stream time at new units. That is a real operating edge, and it compounds slowly.

It is not a patent wall. Air separation is a mature process that Air Liquide, Air Products and Taiyo Nippon Sanso also build, and licensed technology is available. Energy is the largest single variable cost, and Linde buys power at scale on long-term contracts, but it does not control privileged generation; in regions with cheap hydro or nuclear, every producer competes for the same electricity. The edge supports returns and is hard to copy quickly, yet it is imitable, and on its own it would not protect the business for long.

Durability: the risk sits in the customer's process

Substitution threatens Linde from the demand side, not from inside the gas industry. Blast-furnace steel is the most oxygen-intensive common process; a shift toward electric arc furnaces uses less oxygen per ton of steel, and refining, the other large outlet, is exposed to the energy transition. If those customers shrink, no contract term preserves the volumes. That is a slow substitution and it is visible in advance, which makes it manageable, but it is not something the moat can stop.

The nearer question is the next wave of projects. Low-carbon hydrogen, ammonia and carbon capture draw a wider bid list, including engineering firms and subsidy-backed entrants, and returns depend on offtake agreements and tax credits that can be revised. If Linde commits capital to those projects at returns below its legacy air separation base, the moat still protects the old business while the new one dilutes group returns. So far the public evidence does not establish which way that goes; the offtake contracts that would settle it are not yet signed.

One structural feature cuts both ways. With 15- to 20-year contracts, any erosion of the on-site base will show up over a decade, not a quarter. That makes the moat durable in appearance in any single year and slow to falsify. It also means the installed base should keep earning for the rest of this decade regardless of how competitive the next round of bidding turns out. Scale of capital is sometimes offered as a moat. It is not: Linde's investment-grade balance sheet lets it fund projects smaller rivals cannot, but capital is available to competitors too, and access to it does not keep anyone out.

Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.

Continue the comparison

Continue reading the full report

  1. Comparison and summaryHow Linde plc Ordinary Shares compares with its closest rivals, and our moat rating

Free to sign up · 20 research points · unlocking this report uses 10

20 points cover 2 full reports. No card needed.

Research reflects its generation date. Preview the first two modules.

More on Linde plc Ordinary Shares