Global Elevators and Escalators Market Size, Share and Forecast 2025-2033

Global Elevators & Escalators Market Overview

The elevator market is not a construction market with a service tail. It is a service business that manufactures its own future installed base. That sentence is the whole of this report, and it is the reverse of how the category is normally described. Otis — the largest company in the industry — reported 2025 net sales of USD 14,431 million, of which new equipment contributed 35 per cent and service 65 per cent, and of segment operating profit new equipment contributed 9 per cent and service 91 per cent. A market usually described through its cranes and towers is, on the evidence of the most complete set of globally consolidated accounts in the industry, a market described through its maintenance routes.

On the model set out in this report, the global market for elevators, escalators and moving walkways was worth USD 102.14 billion in 2025 and reaches USD 154.10 billion in 2033, a compound annual rate of 5.28 per cent and USD 51.96 billion of absolute growth. New equipment is 38.0 per cent of the 2025 figure, maintenance and repair 43.0 per cent, modernization 19.1 per cent. By 2033 those shares are 32.2, 41.6 and 26.1 per cent — modernization takes a quarter of the market, up from under a fifth, and new equipment falls to under a third.

What this report counts

The perimeter is revenue paid to the vertical transportation industry: new elevators, escalators and moving walkways supplied and installed; the maintenance and repair contracts that keep them running; and the modernization of units already in service. Value is at contract price to the building owner or developer, not at factory gate, because in this industry the installation is not separable from the product — a lift does not exist until it is built into a shaft.

Three perimeter decisions move the number by tens of billions, so they are stated rather than implied.

First, installation is inside the figure. A new-equipment contract is equipment plus field labour plus commissioning, and in most markets the labour is 40 to 60 per cent of the contract. Excluding it would cut the new-equipment line roughly in half and create a market that no participant recognises.

Second, maintenance is inside the figure and it is measured on the installed base, not on the units shipped. This is the only reason the market is larger than the equipment cycle suggests: 26.05 million units in service worldwide each generate a service contract. Otis maintains approximately 2.5 million units, KONE approximately 1.8 million, and the combined KONE and TK Elevator group would hold approximately 3.2 million.

Third, modernization is inside the figure and is the least well defined line in the industry. It runs from a full car and machine replacement to a controller and drive upgrade to a package of door operators, fixtures and safety gear booked inside a maintenance contract. The European Lift Association records EUR 2.0 billion of lift modernization across 32 countries in 2025 against EUR 6.2 billion of new lifts; the world's largest OEMs report modernization at roughly 20 to 21 per cent of group sales. Both can be true only if most component-level renewal is booked as repair rather than as modernization. This report counts code-driven and performance-driven renewal of an existing unit as modernization wherever the work is contracted as a defined project, and states the resulting base explicitly.

Outside the figure: stairlifts and consumer mobility aids sold direct to households; component and sub-assembly sales between manufacturers (door systems, ropes and belts, drives, controllers); building automation and access control beyond the lift itself; and pure inspection and certification revenue. Where those adjacencies matter to the argument they are named, and not counted.

The volume anchors the model is built on

New equipment. 1,581,400 units were installed worldwide in 2025 at an average contract value of USD 24,520. China accounted for approximately 1,000,000 of those units — 63.2 per cent — at an average contract value of USD 15,500, which is the USD 15.50 billion of Chinese new-equipment value in this model. Chinese factory output was larger: 1,401,000 elevators, escalators and lifts produced in 2025, down 1.8 per cent on 2024, on National Bureau of Statistics data — a production figure that includes exports and units not installed in the year, and that is 88.6 per cent of the world's installations rather than a share of a market. The two numbers are routinely quoted as one. They are not the same quantity, and treating them as one implies a Chinese contract at USD 11,064 a unit against the USD 15,500 this model actually uses. Europe's 151,400 new lifts are worth EUR 6.2 billion on European Lift Association figures — an average of EUR 40,950 installed. India installs 80,000 to 85,000 units a year, a figure given by Otis India in November 2025.

The production row is not a fifth row of the same series. Chinese factory output of 1,401,000 units is 88.6 per cent of the world's 1,581,400 installations, and no country installs 88.6 per cent of the world's lifts: Europe alone takes 151,400 and India 80,000 to 85,000, which would leave under 30,000 units for North America, Japan, Korea, the Middle East, Latin America and the rest of Asia Pacific combined. The roughly 400,000-unit gap between Chinese output and Chinese installation is export, inventory and definitional scope, and this report does not attempt to allocate it. The residual row is arithmetic on the model's own total and its two identified blocks, not an estimate of any market.

Installed base. 26.05 million units in service in 2025, of which China holds 12.0 million (SAMR counted 11,532,400 at end-2024) and the 32 countries covered by the European Lift Association hold 7.15 million (6,980,000 at end-2024, plus 175,000 escalators). The base grows at 2.63 per cent a year to 32.07 million in 2033 — far slower than it grew in the 2010s, and that slowdown is the single most important fact in the forecast.

Modernization. 474,600 units were modernized in 2025, 1.82 per cent of the installed base, of which China accounted for 53.1 per cent of the units. SAMR recorded 900,000 Chinese elevators over fifteen years old at end-2024 — a cohort that grows by the construction volume of 2005 to 2015 and that is now the target of direct subsidy.

Market snapshot

Regional components are individually rounded and may not sum exactly to the column totals.

Five findings the model produces and the conventional account does not

1. Unit volume is almost flat and value is not. New-equipment installations rise from 1.58 million to 1.70 million units over eight years — 0.89 per cent a year — while the value of those installations rises at 3.15 per cent. The difference is mix and price, not volume. Any forecast of this market that leads with units installed is describing a stagnant industry that is not stagnant.

2. China stops being a new-equipment market and becomes a service and replacement market. Chinese new-equipment value grows from USD 15.50 billion to USD 17.06 billion, 1.21 per cent a year, while Chinese maintenance grows from USD 9.60 billion to USD 14.04 billion and Chinese modernization from USD 8.06 billion to USD 16.29 billion. The country that installs 63.2 per cent of the world's new lifts contributes 27.4 per cent of the market's absolute growth by 2033, and only 11.0 per cent of that contribution comes from new equipment.

3. Modernization is the growth line and the least certain one. It grows at 9.52 per cent and supplies 40.0 per cent of absolute growth from 19.1 per cent of the base. If it grows at 6.12 per cent instead — the conservative case — the 2033 market is USD 135.30 billion, not USD 154.10 billion. No other single assumption in this report is worth USD 19 billion.

4. Service is where the money is, and service is where the industry is consolidating. Maintenance and repair supplies 39.0 per cent of absolute growth and, at Otis, 91 per cent of segment operating profit. It is also the only line where acquisition buys anything: KONE's combination with TK Elevator, agreed on 29 April 2026, would produce a group with EUR 20.5 billion of sales, 65 per cent of it service and modernization, and 3.2 million units under maintenance, for EUR 700 million of annual run-rate synergies. Those synergies come from service density, not from factories.

5. The buyer of a new lift and the buyer of its maintenance contract are different people. New equipment is sold off-plan to a developer twelve months before handover; the service contract is sold to a building owner, a facility manager or a housing association, often a decade later, and it is worth more over the life of the asset than the equipment margin ever was. Every commercial strategy in this industry is an attempt to keep those two buyers attached to each other, and the fact that independent service providers hold approximately half of the world's service units — Otis's own estimate — is the measure of how often the industry fails.

The year-by-year path, 2025–2033

The snapshot above is two endpoints. This is the path, and in this market the path is the argument: a service business growing at 5.3 per cent a year does not grow evenly, because two of its three lines sit on stocks rather than flows.

Three drivers produce the series. The installed base rises from 26.05 million units to 32.07 million at 2.63 per cent a year, and maintenance revenue per serviced unit rises from USD 1,685 to USD 1,999.7 — mix and escalation rather than a price increase. New-equipment units rise from 1,581,400 to 1,697,500 while the average contract value rises from USD 24,520 to USD 29,272, so almost all of that line’s growth is mix and price rather than volume. Modernization projects rise from 474,600 to 800,000 — 1.82 per cent of the installed base to 2.49 per cent — at a value per project rising from USD 41,010 to USD 50,338.

Line values are rounded independently, so a row may differ from the sum of its rounded components by up to USD 0.01 billion. The terminal-year lines sum to USD 154.09 billion against the USD 154.10 billion carried in the headline — USD 10 million, six thousandths of one per cent, arising from rounding the 2033 drivers before the lines were built. The snapshot table has always carried the same USD 0.01 billion gap; this table makes it visible rather than leaving it to be found.

1. The growth rate tapers and never accelerates. Year on year the market runs 5.77, 5.62, 5.47, 5.33, 5.19, 5.06, 4.94 and 4.83 per cent. That is what a stock-anchored market looks like: the installed base compounds at 2.63 per cent, the maintenance line sitting on it compounds at 4.85, and the only line that can move faster is modernization — which is large enough to hold the total at 5.3 and not large enough to lift it.

2. Modernization is the growth line and the only accelerating one. USD 19.46 billion to 40.27 at 9.52 per cent, against 3.15 for new equipment and 4.85 for maintenance, and 40.0 per cent of the market’s absolute growth from 19.1 per cent of the base. In unit terms 474,600 defined projects become 800,000 — 1.82 per cent of the installed base rising to 2.49 per cent — which is the ageing cohort being converted into contracted work: 900,000 Chinese lifts were already past fifteen years old at the end of 2024.

3. Only 35.0 per cent of the growth lands by 2028, the reverse of an equipment-cycle market. Two thirds of the increase arrives in the last five years, so this forecast is far more exposed to one assumption — how fast an ageing installed base is actually renewed — than to anything already contracted. That is why the conservative case, modernization at 6.12 per cent rather than 9.52, costs USD 19 billion of 2033 value.

How the report is organised

Chapter two sets out the competitive landscape and explains why a market this fragmented in equipment is consolidating in service. Chapter three follows the value chain from a gearless machine to a maintenance route, and locates the point where the profit changes hands. Chapter four covers investment activity and argues that almost no capital in this industry goes into capacity. Chapter five covers technology and treats connectivity as a service-cost technology rather than a passenger-experience one. Chapter six ranks six risks, led by China's property reset and by modernization timing. Chapter seven maps the regulatory regimes that decide when a lift must be replaced rather than repaired.

Method note. The base year is 2025 and all values are nominal US dollars. Volumes are units of equipment, not currency. The model is built from units multiplied by realised contract value, region by region, line by line, and cross-checked against three independent anchors: the reported 2025 segment mix of Otis and KONE, the European Lift Association's unit and value statistics for 32 countries, and the Chinese installed base and production series. Where those disagree, the report says so in the chapter where it matters rather than smoothing the difference. No figure in this report is taken from a market research publication.

Companion analysis

This report is the global parent of a regional series on the same perimeter. The Asia Pacific elevators and escalators market covers that region and the Chinese core inside it — the country that installs 63.2 per cent of the world's new equipment units and accounted for 53.1 per cent of its modernization volume; its figures should be read against the Asia Pacific and China lines above, not against the global total.

Two other construction-material reports sit on the same installation cycle and the same contractor base: building and construction glass, which shares the façade and curtain-wall specification chain on tall buildings, and sanitary ware, which shares the fit-out stage and the same distribution to developers. None of the three should be added to the others; they are separate trades on one building programme.

Table of Contents

• Market overview, definition and perimeter
• Year-by-year path and 2033 composition
• Competitive landscape
• Value chain
• Investment activity
• Technology innovation
• Market risk
• Regulatory landscape

Competitive Landscape

The elevator industry looks like an oligopoly and behaves like a bazaar. Four Western companies and a cluster of Japanese and Korean majors sit at the top of every ranking, and between them they account for a little over half of the world’s market value. But that aggregate conceals two markets with almost opposite structures — equipment, where nothing concentrates, and service, where everything is trying to — and the difference explains every strategic decision the industry has taken since 2020.

Where the profit sits, and it is not where the equipment is

Otis’s 2025 accounts settle the question in a single disclosure. Net sales were USD 14,431 million. New equipment was 35 per cent of sales and 9 per cent of segment operating profit. Service was 65 per cent of sales and 91 per cent of segment operating profit. The world’s largest elevator company is, financially, a maintenance company with a manufacturing division attached to it.

This is not a quirk of one balance sheet. KONE’s 2025 mix reports the same gravity from the other direction: new building solutions fell from 41 to 36 per cent of sales in one year, service held at 42, and modernization rose from 19 to 21 per cent. The combined KONE and TK Elevator group that was agreed in April 2026 would take that to roughly 65 per cent of sales from service and modernization together. When two competitors of that size describe the same destination, it is not a strategy; it is the structure of the industry asserting itself.

The reason is arithmetic. A new-equipment contract is a one-off sale to a professional purchaser who runs a tender and buys on price and delivery. The maintenance contract attached to that same unit is renewed every one to five years for the life of the building, is priced locally against the incumbent’s own cost base, and requires no factory. Across the industry, service revenue per unit is small and service margin per unit is large; equipment revenue per unit is large and equipment margin per unit is thin. A firm that installs a lift and loses the service contract has paid the cost of acquiring a customer for a competitor.

The equipment market is fragmented, and structurally so

New-equipment value is not concentrated to the degree the brand names imply. Otis’s new-equipment revenue was USD 5,051 million in 2025, or 13.0 per cent of the modelled USD 38.78 billion world equipment market. KONE’s new building solutions were EUR 4,048 million, or 12.0 per cent at EUR/USD 1.15. No company holds more than about one unit in seven of the world’s new-equipment value, and the two largest hold a quarter between them.

The reason is China. Chinese factories produced 1,401,000 elevators, escalators and lifts in 2025 on National Bureau of Statistics data, and Chinese installations are a different and smaller quantity: approximately 1,000,000 units, 63.2 per cent of the 1,581,400 installed worldwide, at an average contract value of USD 15,500 against USD 47,100 in Europe. Conflating the two makes the Chinese price gap look like a rounding error, because dividing output into Chinese equipment value implies USD 11,064 a unit, which is not a price this model builds any Chinese contract at. A market where two thirds of the units clear at a third of the price cannot concentrate on brand, because a domestic producer with a lower cost base and a local service network is a credible bidder in almost every Chinese tender. Otis’s own structure records the history of that: it operated in China through joint ventures — Otis China from 1998 with Tianjin Tai Kang, and Otis Electric from 1997 with Xizi Elevator Group — and in October 2025 it bought out the noncontrolling shareholder of Otis Electric. Localising was how you competed; consolidating the local entity is how you now capture the service profit.

China’s 2025 split: a shrinking new-build market with a growing total

Chinese demand disaggregated sharply in 2025, and the pattern matters more than the headline. Traditional new installations fell 12.9 per cent. Freight elevators were broadly flat. Home and self-built housing elevators rose 12.5 per cent. The two "old" businesses — renovating old elevators and installing lifts into old residential compounds — rose 52 per cent. Exports rose 20.87 per cent, and total industry shipments still rose by roughly 5 per cent. Those are the figures of an industry review of the Chinese market published in February 2026.

They do not reconcile neatly with the National Bureau of Statistics series, which records production down 1.8 per cent for 2025. The two measures are counting different things — factory output including exports, against domestic deliveries including renovation — and this report states the divergence rather than choosing the more convenient one. What both agree on is that the replacement and renovation channel is now large enough to offset a double-digit decline in new construction, which is the single most important transition in the global market this decade.

The same review records the competitive consequence: the top ten producers averaged growth above 6 per cent and the fastest grew more than 15 per cent, while small and medium producers lost share. Treasury-bond-funded renovation has high project access requirements, tight execution schedules and strict fund supervision. That is a channel deliberately designed to favour firms that can bond, schedule and document at scale, and it is the mechanism by which a fragmented industry consolidates without anyone buying anyone.

The consolidation that is happening is in service, and it is happening now

On 29 April 2026 KONE and TK Elevator agreed to combine, in a cash and share transaction under which KONE acquires the entire share capital of the entity holding TK Elevator. The consideration is EUR 5 billion in cash plus up to 270 million new KONE class B shares, valued at EUR 15.2 billion at the 28 April close, giving an enterprise value of EUR 29.4 billion including TK Elevator’s approximately EUR 9.2 billion of interest-bearing net debt, most of which is refinanced. On illustrative last-year figures the combined group would have EUR 20.5 billion of sales, approximately 65 per cent of it service and modernization, more than EUR 2.7 billion of adjusted EBIT before synergies, and 3.2 million units under maintenance.

Three readings of that number matter here.

First, it is a service-base purchase. On KONE’s standalone sales of EUR 11,245 million, the implied TK Elevator contribution is EUR 9,255 million, so the enterprise value is roughly 3.2 times the acquired revenue and about EUR 9,188 per unit of maintenance portfolio. No one pays three times revenue for commodity equipment installation. The listed synergy drivers begin with "higher density of service networks" and run through R&D, platform optimisation, procurement and SG&A. Density is the product being bought.

Second, it changes the league table. 3.2 million units under maintenance is 12.3 per cent of the modelled 26.05 million global base, against Otis’s approximately 2.5 million and KONE’s approximately 1.8 million. On the combined sales figure at EUR/USD 1.15, the group would be 23.1 per cent of the total market including service and modernization — against 14.1 per cent for Otis and 12.7 per cent for KONE as standalone companies.

Third, it is not done. Completion is expected earliest in the second quarter of 2027, subject to regulatory approvals and to KONE’s extraordinary general meeting, which was expected to be convened in June 2026 and at which shareholders holding approximately 40.3 per cent of shares and 74.3 per cent of votes had irrevocably undertaken to vote in favour. The UK Competition and Markets Authority opened an invitation to comment on 22 September 2026 and had not formally commenced a phase 1 investigation at the time of writing; reviews were under way in the United States, India and Singapore. The purchase price is adjustable for divestments required to obtain clearance, which tells you where the regulator’s attention is: not in equipment, where share is diffuse, but in service, where a combined route in a dense city is a local monopoly.

The independent half is the real competitor

Otis states that independent service providers have an aggregate portfolio of about 50 per cent of service units worldwide, while accounting for a smaller percentage of the service business by value because of the types of units and level of maintenance they cover. Every strategic question in this industry sits inside that sentence. The OEMs have installed most of the base; they maintain slightly less than half of it; and every unit they lose at first renewal is lost permanently, because the independent that wins it prices against a lower cost structure and never gives it back.

That is what connectivity is for (see the technology chapter), and it is why the industry’s reported order books are a poor guide to future service revenue. KONE’s order book fell 2.8 per cent at reported rates in 2025 while orders received rose 3.8 per cent. Otis’s new-equipment backlog ended 2025 up 2 per cent and its modernization backlog up 30 per cent. The line that is growing is the line that attaches to units already installed.

How to read a share number in this market

This report does not publish a single global market-share table, because a whole-market share in this industry is an artefact of currency, perimeter and whether installation labour is counted. What it publishes instead is value share by line, which is comparable: 13.0 per cent of new equipment for Otis, 12.0 for KONE; and, after completion, 12.3 per cent of the world’s serviced units for the combined KONE and TK Elevator group. Those are the numbers that decide who has pricing power, and they point in opposite directions.

The structure is therefore fragmented at the level of the market, consolidating at the level of the service route, with eight tier-one players and high competitive intensity in equipment and rising intensity in service as the majors buy density and the independents defend price. Distribution is direct — Otis operates 1,400 branches and offices and KONE runs its own service organisation — because a lift cannot be serviced through a distributor once the warranty has expired. The operational model is hybrid: manufacturing for a minority of revenue, field service for the majority of profit.

Value Chain

A lift is sold before it exists, built into a building somebody else designed, handed over to an owner who did not buy it, and then maintained for twenty-five years by whoever can physically reach it. Seven links separate a steel coil from a maintenance route, and each is governed by a different constraint. Only one of them is a machine, and the machine is not where the money changes hands.

Link one: components, where two parts decide everything

A modern traction lift is a gearless permanent-magnet machine, a variable-voltage variable-frequency drive, ropes or a coated belt, a car, a counterweight, safety gear, a controller and a door operator. Two of those dominate the engineering economics. The door operator is the most frequently cycling assembly on the equipment and the source of most call-outs, so it drives the cost of the maintenance contract long after the warranty has expired. The controller is where lock-in lives: a proprietary controller with proprietary diagnostics is the difference between a service contract that is competed and one that is not.

Component supply is more concentrated than elevator supply. Door systems, drives and ropes come from a smaller group of specialists than the OEM badges suggest, and KONE is in the process of separating its door business into its own legal and operating structure — a disclosure that says something about where the company thinks the value in that link sits.

Link two: the factory, which is real but not decisive

Otis runs 16 factories and 11 research centres with 1,300 engineers; KONE runs 10 manufacturing units and 7 research units and works with more than 40,000 suppliers. That is substantial industrial scale. It is also not a moat. A Chinese producer with a lower labour cost, a shorter supply chain and a domestic service network can put a compliant lift into a Chinese shaft at USD 15,500 when a European contract clears at USD 47,100. Where the factory does matter is in the tall-building segment and in the standardised modular platforms the majors now sell — Schindler lists a standardised modular platform for US mid-rise and standardised modernization packages as its 2025 product achievements. Standardisation is a factory strategy for a service problem.

Link three: the contract, signed before the building exists

New equipment is sold off-plan, typically to a developer or a general contractor, at an early stage of the construction cycle. Delivery is generally within twelve months of booking, shorter in China. Customers pay in advance to cover design and contract engineering, then at milestones — delivery to site, completion of installation, commissioning — and revenue is recognised on a percentage-of-completion basis. Two consequences follow. The new-equipment order book is a twelve-to-twenty-four-month leading indicator of the service base, because every unit installed today is a maintenance contract tomorrow. And the new-equipment margin is structurally thin, because the buyer is a professional purchaser with a tender, three credible bidders and a construction budget.

Link four: installation, the half of the contract nobody models

Field labour and commissioning are 40 to 60 per cent of a new-equipment contract in most markets. Installation is carried out by the OEM’s own technicians or by subcontractors, with the OEM completing final inspection and commissioning. This is the least scalable and least automatable part of the chain, and it is where a lift’s first five years of reliability are decided: alignment, door set-up and controller configuration at commissioning determine the call-out rate, and the call-out rate determines whether the service contract is profitable. It is also why the installer and the maintainer are usually the same firm, and why that link is worth defending.

Link five: handover, where the profit changes hands

The unit is commissioned, warranted for a limited period, and then the warranty expires. At that moment a decision is made by somebody who was not party to the construction contract: a building owner, a facility manager, a housing association or a government agency. If the incumbent converts the unit, it holds a revenue stream for two decades. If it does not, the unit joins the independent pool permanently, because an independent service provider prices against a lower cost base and rarely loses the contract back.

This is the hinge of the entire value chain, and it is where the industry’s economics are decided — not in the factory, not in the showroom. Otis’s estimate that independent providers hold approximately half of the world’s service units, against a smaller share of service value, is the aggregate measure of how often that handover is lost.

Link six: the route, which is a density business pretending to be a service business

The 26.05 million units in service worldwide are maintained by technicians whose cost is fixed per day and whose output is not. Across the 32 countries covered by the European Lift Association, 6,980,000 lifts and 175,000 escalators are served by 174,800 employees — roughly one employee per 40 units. Otis employs 37,000 service mechanics out of more than 1,400 branches and offices, which on an implied maintenance and repair revenue of USD 6,364 million is about USD 172,000 of revenue per mechanic. KONE’s 1.8 million units and Otis’s 2.5 million are, in operating terms, route maps.

Which is why the price of a contract varies by a factor of nearly six across the world on the same asset: USD 800 a unit a year in China, USD 2,300 in Europe, USD 4,600 in North America. The physical task is identical. The difference is wage levels, code-mandated visit frequency, the scope of coverage, and whether the market will bear a price that funds a modern fleet of technicians.

Link seven: the second sale

Eventually the unit is worth more renewed than repaired. Modernization ranges from finishes and fixtures to a full car, machine, controller and safety-gear replacement, and at USD 41,010 average contract value it is the largest single transaction available on an existing asset — larger in cash terms than the original installation at today’s prices, and bought by an owner who already knows the building. Modernization is 1.82 per cent of the installed base a year and grows at 9.52 per cent in this report’s model, which makes it the second sale that finally justifies the first.

Where China’s chain differs, and why it is being unwound

The Chinese value chain was deliberately localised. Foreign OEMs entered through joint ventures — Otis China in 1998 with Tianjin Tai Kang, Otis Electric in 1997 with Xizi Elevator Group — because local manufacture and local partners were the condition of access. That structure optimised for winning new equipment in a market growing at a million units a year. It is the wrong structure for a market where the growth has moved to service and replacement, which is why Otis bought out the noncontrolling shareholder of Otis Electric in October 2025 and now owns it outright. The chain is being consolidated at precisely the moment its economics change.

What the chain says about the market

The margin in this industry is created at handover, defended on the route, and harvested at modernization. Equipment manufacture is the price of admission, not the business. A supplier that sells equipment and loses the route has donated a twenty-five-year annuity to a competitor; a supplier that holds the route can afford to sell equipment at a thin margin, which is exactly what the majors do. Supply chain complexity is moderate — the inputs are steel, electronics and field labour, none of them scarce — and the value chain model is hybrid: internal manufacturing, internal service, subcontract installation. That combination is the reason capital allocation in this industry looks the way it does, which is the subject of the next chapter.

Investment Activity

Almost no capital in this industry goes into making more elevators. It goes into owning more of the ones already installed. That distinction explains the shape of every transaction, every capacity announcement and every research budget in the sector, and it is the reason the market can grow 5.28 per cent a year while the four largest Western manufacturers report organic growth between flat and 4 per cent.

The research budget of the world’s largest elevator company is USD 152 million

Otis spent USD 152 million on research and development in 2025, or 1.1 per cent of net sales. It spent a further USD 45 million on digital and strategic initiatives, taking the combined figure to 1.4 per cent of sales. It owns approximately 4,600 issued patents and 1,300 pending applications, and filed roughly 900 applications across the last three years. It employs 1,300 engineers worldwide and runs 11 research centres, with 16 factories.

Those numbers are small for a USD 14.4 billion industrial company, and they are the correct size for this one. The physics of lifting a mass in a shaft has not changed; the last platform to reach genuine scale is Otis’s Gen2, launched in 2000 and past one million units sold. What the industry buys with its research budget is not a new machine. It is the sensor, the connection and the dispatch algorithm that reduce the cost of the route — which is why the digital line is reported separately by management even though it is a seventh the size of the R&D line.

The capital that did move went into a maintenance portfolio

The KONE agreement to acquire TK Elevator, announced on 29 April 2026, is the largest capital allocation in the industry’s recent history and it is not a capacity investment in any conventional sense. The terms: EUR 5 billion in cash plus up to 270 million newly issued KONE class B shares, worth EUR 15.2 billion at the 28 April close, for an enterprise value of EUR 29.4 billion including approximately EUR 9.2 billion of TK Elevator’s interest-bearing net debt, most of which is refinanced. The combined group would have EUR 20.5 billion of sales, more than EUR 2.7 billion of adjusted EBIT before synergies, and 3.2 million units under maintenance.

On KONE’s standalone sales of EUR 11,245 million, the implied acquired revenue is EUR 9,255 million, so the multiple is approximately 3.2 times revenue. That is not a price for equipment manufacturing; a plant is replaceable at book value. It is a price for 3.2 million service contracts, or about EUR 9,188 per unit of portfolio. Management identifies approximately EUR 700 million of annual run-rate pre-tax cost savings, with full P&L effect by the end of the third year and one-off costs of one to 1.2 times the run-rate, and lists the sources in order: higher density of service networks, combined R&D, platform optimisation, procurement and SG&A. Four of those five are portfolio economics, not factory economics.

It is also, explicitly, a re-rating exercise. KONE states that the transaction resets its long-term targets and supports an ambition to push adjusted EBIT margin substantially beyond its current standalone target of 16 per cent. Consolidating routes in overlapping cities is how a service business raises margin without raising price.

The deal is not closed. Completion is expected earliest in the second quarter of 2027, subject to regulatory approvals and to KONE’s extraordinary general meeting, and the consideration is adjustable for any divestments required to obtain clearance. The UK Competition and Markets Authority issued an invitation to comment on 22 September 2026 without having formally commenced a phase 1 investigation; reviews were under way in the United States, India and Singapore. Any remedy that requires divesting a national service business would hand a route network to a competitor and change the arithmetic of the synergies, which is the principal execution risk in the transaction.

India is where capacity is genuinely being added

The one region where announced investment is capacity investment is India, and the disclosure is unusually precise. Otis India doubled new equipment orders over three years, expanded its Bengaluru plant by 20 per cent in 2024 after an earlier doubling in 2012, sources more than 90 per cent of local demand from that plant, and treats India as one of only two global escalator manufacturing hubs alongside China. India installs 80,000 to 85,000 units a year, and the country is the reason Asia Pacific orders grew more than 20 per cent in the first quarter of 2025.

The logic is straightforward. A market growing from a low base with a rising share of mid-rise residential and metro infrastructure needs local assembly and a local service organisation, and freight costs on a lift are material relative to its price in a low-price market. India is the only large market where the equipment business is still worth building plants for.

Chinese capacity is being rationalised, not expanded

Chinese production of elevators, escalators and lifts was 1,401,000 units in 2025, down 1.8 per cent, on National Bureau of Statistics data, after a year in which traditional new installations fell 12.9 per cent and the renovation channel rose 52 per cent. The recovery that began in 2026 — production up 3.2 per cent in the first half, with June up 9 per cent — is a recovery in replacement demand, not in new construction. Capacity is being worked harder on the same buildings.

This is the least capital-intensive kind of growth available to a manufacturer, and it is why Chinese producers have been able to gain share while their market shrank: the marginal unit of demand requires no new plant, and a renovation contract is shorter, more local and more service-weighted than a new-build contract.

Why modernization needs almost no capital at all

The fastest-growing line in this report, at 9.52 per cent, is also the one that consumes the least capital. Modernization uses the existing factory for cars, doors and controllers, the existing branch network, and the existing technician — the unit is already on the route. It is, in capital terms, closer to a service product than a manufacturing one, which is why the majors can grow it at double digits while their capital expenditure stays flat, and why Schindler’s 2025 disclosure that modernization orders grew 19 per cent and modernization revenue grew 12 per cent in local currencies sits alongside an operating cash flow that fell 7 per cent rather than a capex programme.

What this means for the forecast

Capital intensity in this industry is moderate and falling in real terms per unit of revenue. The binding constraint on growth is not factory capacity, working capital or engineering; it is the number of qualified technicians and the density of the route they can cover. That is why this report projects the service line growing faster than the equipment line, why it expects capital to keep flowing into portfolios rather than plants, and why the investment trend is rising — rising in absolute money, directed at ownership of the installed base rather than at the means of building it.

The one capital question the industry has not answered is the one China poses. A market that installs a million units a year for fifteen years creates a cohort that must be replaced in the following fifteen, and someone has to fund the factories, the inventory and the technicians to do it. At present that funding is coming from the state — through subsidised replacement — rather than from the industry’s own balance sheet, and a forecast that assumes the subsidy continues is making a policy assumption, not an industrial one. Chapter six treats that as the second-largest risk in the report.

Technology & Innovation

The most important technology in this industry is not the machine that lifts the car. It is the connection that tells somebody in a van whether the car needs lifting today. That is not a rhetorical flourish: it is the allocation of the industry’s research money, its patent activity and its margin, and it follows directly from the fact that 91 per cent of Otis’s segment operating profit comes from service.

The drive train stopped being the story a generation ago

The last elevator platform to reach genuine scale was Otis’s Gen2, launched in 2000 and past one million units sold. Machine-room-less gearless traction, regenerative drives and coated belts are now standard across the majors and are no longer differentiators; they are the entry ticket. A Chinese producer sells all three at USD 15,500 a unit. When a technology has diffused to the low-cost bidders in the largest market in the world, it has stopped being a technology advantage and become a manufacturing cost.

What has changed since is the electronic architecture around the mechanical one. Otis’s Gen360 replaces a set of mechanical components with electronic ones, which does two things that matter commercially. It reduces the potential for entrapments, and it lets a technician perform maintenance safely from inside the car rather than on top of it — which, depending on local regulation, removes the need for a refuge space above the car and the roof protrusion that goes with it. That last clause is the most commercially interesting sentence in the product literature. It is a building-cost innovation sold as an elevator feature: a flat roof instead of a parapet and a machine-room-like box changes the developer’s structural cost, and the developer is the person who signs the new-equipment contract.

Connectivity is a service-cost technology, and it is now mainstream

At 31 December 2025 approximately 1.1 million of Otis’s 2.5 million units were connected — about 44 per cent of the global portfolio, including units still under warranty. Schindler reports more than 40 per cent of its maintenance portfolio cloud-connected. KONE bundles 24/7 connected service into its maintenance offering across its markets. Within five years the connected unit has gone from a premium option to the default condition of a competitive service contract.

The value is routinely described as predictive maintenance. That is the marketing frame, and it oversells what the sensors do: the failure modes that matter in a lift — door faults, levelling, drive faults, entrapments — are largely known failure modes, and a connected unit is better at triaging them than at predicting them. The four things connectivity actually buys are more prosaic and more valuable:

Remote reset. A large share of call-outs are nuisance faults that clear on a controller reset. Doing that from a desk converts a two-hour van journey into a two-minute action, and the saving lands directly on the service margin.

Fault triage before dispatch. Knowing which part to bring to which unit converts a diagnostic visit into a repair visit. On a route where revenue per mechanic is approximately USD 172,000 a year, the number of productive visits per day is the whole of the operating leverage.

Route density. Connected units generate a live fault stream that can be scheduled geographically rather than chronologically. This is the mechanism by which the service networks that the majors are buying become worth more owned together than apart, and it is why KONE lists density of service networks first among its synergy drivers.

The renewal argument. A building owner choosing between an OEM with remote monitoring and an independent with a lower price is being asked to value uptime. Connectivity is the only durable answer the OEM has to a 30 per cent price gap, which is precisely the situation created by the independent half of the service market.

Destination dispatch is the one passenger-facing change that pays

Otis’s Compass 360 destination management system groups passengers by destination and assigns them to a specific car, with algorithms that anticipate traffic demand within a building. The engineering claim — shorter waiting and ride times — matters less than the capital claim: in a tall building, better dispatch can reduce the number of cars or the size of the core required for a given handling capacity. As with Gen360, the buyer of the technology is the developer’s cost model, not the passenger’s experience.

Patent counts are a poor guide to innovation here

Otis holds approximately 4,600 issued patents and 1,300 pending applications, and filed roughly 900 applications in the last three years. For a company of its size and history that is a modest, steady portfolio concentrated in Europe, the United States and Asia, and it is consistent with an industry that innovates incrementally around a mature mechanical principle rather than in bursts. Any claim that this is a high-patent-activity industry is measuring the wrong thing: the defensible assets are the service route, the code approval and the installed base, none of which is patentable.

The technology that has not arrived

Ropeless and multi-car systems remain a tall-building niche, constrained by the shaft and by code rather than by cost. Hydraulic drives are declining. Escalators and moving walkways have seen their innovation in energy management — sensor-equipped units that run only when a passenger approaches, or at reduced speed when idle — which is a running-cost feature for the building, not a product revolution.

The genuinely unsolved problem is labour. Installation and maintenance are 40 to 60 per cent of a new-equipment contract and the majority of a service contract, and both are delivered by people climbing into shafts. No technology in this industry has yet reduced the technician-hours per unit per year in a way that shows up in the accounts. Until one does, the constraint on growth in the service line is the number of qualified people, not the number of lifts.

Where that leaves the technology rating

Innovation intensity is moderate: real, continuous, and concentrated in electronics, software and service process rather than in the machine. Patent activity is moderate. The technology is mature and the industry knows it — which is why its research budget is 1.4 per cent of sales and its acquisition budget is EUR 29.4 billion. In a mature technology market, the way to buy growth is to buy the installed base of the technology you already make.

Market Risk

Six risks sit under this forecast, ranked by what they would cost. One is a policy and property risk in a single country that supplies two thirds of the world’s units. One is a timing assumption inside this report’s own model. The remaining four are structural and slower. The ranking matters more than the list: the first two together account for most of the distance between the conservative and the optimistic case.

1. China’s new-equipment reset — dominant

China produced 1,401,000 elevators, escalators and lifts in 2025, down 1.8 per cent, while traditional new installations fell 12.9 per cent. That single market installs 63.2 per cent of the world’s new units and takes 32.5 per cent of the world’s total market value at USD 33.16 billion, which is 40.0 per cent of world new-equipment value: the like-for-like comparison, and the gap between the two is the price of a Chinese contract at USD 15,500 against USD 47,100 in Europe. It is also 40 per cent of KONE’s sales as of 2025, up from 38 per cent the year before even as the market contracted. No other country has that weight in any industrial forecast in this report.

The base case assumes Chinese new-equipment value grows from USD 15.50 billion in 2025 to USD 17.06 billion in 2033, which is 1.21 per cent a year — a floor, a modest price recovery, and a mix shift towards larger and taller projects. If instead the Chinese equipment market declines at 2.95 per cent a year, 2033 lands at USD 12.20 billion, which removes USD 4.86 billion from the forecast, or 3.2 per cent of it. That is the mechanical arithmetic. The unmechanical risk is worse: a Chinese property market that resets further would take the maintenance base with it, because a building that is never completed never generates a service contract.

2. Modernization timing — the model’s largest single assumption

Modernization grows at 9.52 per cent in the base case and supplies 40.0 per cent of the market’s absolute growth from 19.1 per cent of its 2025 value. It is also the line with the widest definitional disagreement in the industry: the European Lift Association recorded EUR 2.0 billion of lift modernization across 32 countries in 2025 against EUR 6.2 billion of new lifts, while the majors report modernization at 20 to 21 per cent of group sales. Those two statements are only compatible if most component-level renewal is booked as repair.

The consequence is that this line can be moved without anyone lying. At a 6.12 per cent rate — the conservative case — 2033 is USD 31.30 billion instead of USD 40.27 billion, and the total market is USD 135.30 billion at 3.58 per cent instead of USD 154.10 billion at 5.28 per cent. No other assumption in this report is worth USD 19 billion. A reader who distrusts the modernization build should use the conservative path and disregard the rest of this chapter.

3. Consolidation and remedy risk

The KONE and TK Elevator combination would create a group with 3.2 million units under maintenance, 12.3 per cent of the world base, and 23.1 per cent of the total market by value. The consideration is adjustable for divestments required to obtain clearance, and the reviews under way in the United States, India and Singapore, alongside the UK Competition and Markets Authority’s invitation to comment of 22 September 2026, will determine how much of a national service network has to be sold. Two outcomes matter for this forecast, and they cut in opposite directions: remedies that hand routes to a competitor raise service prices and margin across the industry; remedies that delay or block completion strand EUR 700 million of expected synergies and leave a distracted competitor in the market for two years.

The wider risk is not the transaction but the precedent. If density of service routes is worth EUR 9,188 a unit, every mid-sized independent portfolio becomes an acquisition target, and the half of the world’s service units now held by independents becomes contested ground rather than a stable feature of the market.

4. Service margin compression

Service supplies 39.0 per cent of absolute growth in this report and 91 per cent of Otis’s segment operating profit. Its economics are set locally, against an independent competitor with a lower cost base, on a contract that is renewed every one to five years. The Chinese maintenance price of USD 800 a unit a year against USD 4,600 in North America is not a discount to be closed; it is what the market will bear in a country with a young, price-sensitive base and a large supply of low-cost technicians. As China’s share of the global base rises, the blended price of service falls, and this report’s 4.85 per cent service growth already assumes that drag.

Wage inflation is the other side. Technician pay rises with local labour markets and cannot be passed on in a competitive tender. Connectivity — 44 per cent of Otis’s portfolio connected, more than 40 per cent at Schindler — is the industry’s principal hedge, and it is a hedge that only works at scale.

5. Input cost and trade measures

A lift is steel, copper, electronics and labour. Steel and electronic components are globally traded inputs, and in North America the assembled unit is exposed to the tariff structure of 2025 and 2026, which is why this report assumes North American new-equipment prices rise 2.75 per cent a year against 2.5 per cent elsewhere. That is a small wedge, and it is deliberately small: tariffs raise the price of a lift to a developer, which reduces units, so the revenue effect is closer to neutral than the price effect suggests.

6. Safety, liability and the ageing fleet

China’s market regulator recorded 41 elevator accidents and 27 deaths in 2024, alongside 900,000 units over fifteen years old and no national mandatory scrappage standard for a complete lift — only component disposal standards. Every accident that reaches public attention tightens inspection, shortens the interval at which an old unit must be renewed, and increases the demand for documented maintenance. In this industry safety regulation is both a cost and a demand driver, and it is the only one a government can switch on at will.

The three paths

The conservative path is not a recession scenario. It is what this model produces if Chinese equipment stabilises rather than recovers, if modernization runs at the lower of the two rates the industry reports, and if service prices stay flat in real terms. The optimistic path is what it produces if the KONE and TK Elevator combination reprices service, if India sustains 7 to 8 per cent unit growth, and if code deadlines in Europe and China compress the modernization cycle. The baseline is the middle, and the honest summary of the range is this: the industry will grow, and the argument is about which line grows.

How the risks score

Overall risk is moderate. Geopolitical risk is high, not because of conflict exposure but because two thirds of world unit volume and 40 per cent of the largest manufacturer’s revenue sit in one regulatory jurisdiction whose property market is in a multi-year reset. Substitution risk is low: below roughly eight storeys stairs are a substitute for a lift and above it nothing is, and no technology in the forecast window changes that. Regulatory complexity is high — because the rules that decide when a unit must be renewed differ by country and, in the United States, by state — and that complexity is the subject of the final chapter.

Regulatory Landscape

In most capital equipment markets, regulation sets the cost of compliance. In this one it sets the replacement date. A lift is a safety device installed in a public space, its condition is inspected by someone with statutory authority, and the point at which repair stops being acceptable is a legal judgement rather than an engineering one. That is why the same asset has a different economic life in Shanghai, Munich and Ohio, and why the regulatory map is the closest thing this market has to a demand schedule.

Europe: the Directive does the work, and it is written around installation

Lifts in the European Union are governed by Directive 2014/33/EU, the recast of the 1995 Lifts Directive, adopted on 26 February 2014 and in force, with a consolidated version current as of 30 May 2026. Its structure is unusual and it explains the European market.

Recital 4 of the Directive states the governing logic: lifts "only come into existence as finished products once they have been permanently installed in buildings or constructions", and consequently they cannot be imported into the Union and are only placed on the market. There are, in the Directive’s own words, no importers or distributors of lifts. The conformity assessment therefore attaches to the installed unit and to its installer, not to a box shipped from a factory. Every lift placed on the market must carry CE marking, be accompanied by an EU declaration of conformity, and be designed and installed against the harmonised standards — principally EN 81-20 for the lift itself and EN 81-50 for the testing and examination of components.

Two consequences follow for the numbers in this report.

A substantial modification pulls the unit back into conformity assessment. Because the obligation attaches to the unit as placed on the market, a change of sufficient scope — a new controller, a new drive, a new safety gear, a changed rated load or speed — can require the lift to be treated as newly placed on the market, with a fresh declaration and a fresh assessment. That is the legal engine behind European modernization: it is not a discretionary upgrade cycle, it is a compliance cycle, and it is why Europe modernizes at a higher rate of its installed base than any other region in this report’s model.

Periodic inspection is national, and it is strict. Member states operate their own inspection and licensing regimes on top of the Directive, which is why the European service contract carries a price of roughly USD 2,300 a unit a year against USD 800 in China: a mandated visit frequency, a licensed inspector and a documented trail are embedded in the price. The Directive harmonises what a lift must be; the member states decide how often someone must certify that it still is.

The scale of the resulting regime is visible in the industry’s own statistics. The European Lift Association counts 6,980,000 existing lifts and 175,000 escalators across 32 countries, served by 174,800 employees, with 151,400 new lifts placed in 2025 at EUR 6.2 billion and EUR 2.0 billion of modernization work.

China: the state is the demand driver

China combines the largest installed base in the world with the weakest mandatory replacement standard, and has chosen to close that gap with money rather than with regulation.

The regulatory baseline: the State Administration for Market Regulation counted 11,532,400 elevators at end-2024, of which 900,000 were more than fifteen years old, and recorded 41 elevator accidents with 27 deaths in 2024. There is no national mandatory scrappage standard for a complete lift, only standards for the disposal of individual components. "Old" is an industry convention of more than fifteen years of service, not a legal status.

On 8 January 2026 the National Development and Reform Commission and the Ministry of Finance issued notice 发改环资〔2025〕1745号 setting out the 2026 large-scale equipment renewal and consumer trade-in policy. Four features of it bear directly on this market:

Residential old elevators remain a supported category of equipment renewal, alongside industrial, energy and medical equipment, funded through ultra-long special treasury bonds.

Installation of lifts into old residential compounds was added to the supported scope for 2026 — a category distinct from replacing an existing lift, because these buildings have never had one. That is new demand rather than replacement demand, and it is aimed at the low-rise stock that China built before 2000.

The conditions tightened rather than loosened. The notice requires authorities to set quantitative minimum service lives and technical criteria by sector, and makes equipment age and depreciation hard conditions for an application, while lowering the investment threshold and extending support to small and medium enterprises and to commercial complexes, shopping centres, department stores and supermarkets.

The funding is conditional and time-limited. The mechanism is a central allocation against approved projects, with local implementation, and unused allocations are recovered. A subsidised replacement is therefore not a market price; it is a programme, and a forecast that assumes it continues is making a policy assumption.

The earlier version of the programme set the price signal: from 2024, replacement of a lift more than fifteen years old attracted a subsidy of RMB 150,000 a unit. Against a Chinese modernization contract value of roughly USD 32,000, the subsidy covers a substantial share of the transaction, which is why the renovation channel grew 52 per cent in 2025 while new installations fell 12.9 per cent.

United States: fifty regimes and no national deadline

The United States regulates lifts through ASME A17.1/CSA B44 for new installations and ASME A17.3 for existing ones, adopted state by state and sometimes city by city, with inspection and licensing at municipal level. The consequence is the opposite of Europe’s: there is no single national instrument that pulls an ageing unit into a modernization cycle, so renewal is driven by local ordinance, by insurance requirements and by the owner’s own judgement about downtime.

That is visible in the model, where North American modernization runs at 1.4 per cent of the installed base rising to 2.0 per cent, against Europe’s 1.7 per cent rising to 2.4 per cent, even though the North American base is older on average and the contract value per modernization, at USD 95,000, is roughly twice the European figure. The United States modernizes fewer units and spends more on each, because the trigger is discretionary and the building is usually occupied.

India and the enforcement gap

India works to IS 14665 with state-level lift acts, licensing of erectors and periodic inspection. The binding constraint is not the standard but its enforcement: a market installing 80,000 to 85,000 units a year, growing at double digits, with a large share of low-rise residential buildings and a young installed base, generates new equipment demand far faster than it generates renewal demand. This report models Indian maintenance at a blended price well below the European level for that reason, and expects the renewal cycle to become material only after 2030.

What the regulatory map does to the forecast

Three things, and they are the reason this chapter is not an appendix.

It sets the modernization rate, which is the largest single assumption in the model. Europe’s compliance-driven cycle and China’s subsidy-driven cycle produce the two highest modernization penetrations in the report; the United States’ discretionary cycle produces the lowest relative to the age of its stock.

It sets the service price. A mandated inspection interval is a contract that cannot be competed away on scope, only on price, which is why European service revenue per unit is nearly three times the Chinese figure on identical equipment.

It is the only demand driver a government can switch on. Capacity, steel and technology are private decisions. A code revision, an inspection interval or a replacement subsidy is a public one, and in 2026 the most consequential of those decisions was taken in Beijing on 8 January, in a document about equipment renewal that does not use the word elevator in its title.

Regulatory complexity in this market is therefore high, and the approval pathway is standards and conformity assessment rather than product authorisation: the supplier’s obligation is to certify an installation against a standard, in a jurisdiction whose inspector has the authority to take the unit out of service.