1. Two numbers moving in opposite directions

In April 2026, China's State Administration for Market Regulation released its annual bulletin on the national safety status of special equipment for 2025. In that category-by-category inventory, elevators are the largest single class: as of the end of 2025, the national total of special equipment stood at 24.4169 million units, of which elevators accounted for 12.3159 million, pressure vessels 6.0741 million, lifting machinery 3.1642 million, in-plant motor vehicles 2.5424 million, and boilers 293,300. Put differently: of every two registered pieces of special equipment in China, one is an elevator.

In the same year, output of elevators, escalators and lifts by enterprises above designated size, as measured by the National Bureau of Statistics, was 1.401 million units, down 1.8% year on year. That figure had reached a historic high of 1.557 million units in 2023, fell to 1.492 million in 2024, and continued to slide in 2025.

One number is rising; the other is falling. The installed base is still accumulating at a rate of seven to eight hundred thousand units a year — 11.5324 million at the end of 2024, 12.3159 million a year later — while production has now declined for two consecutive years. Between these two divergent numbers lies the most important transition facing China's elevator industry over the coming decade: from selling equipment to managing it.

For a manufacturing sector accustomed to a million new orders a year, this is not an easy turn. New installation is a one-off business: it can be scheduled against projects, and its gross margin can be calculated cleanly. The installed base is a different animal — dispersed across several million buildings, billed monthly, dependent on technicians making site visits, with unit prices driven to the floor. The organisational capabilities the two require are barely the same.

2. Seventy per cent of the world's elevators are made in China

To understand how large this turn is, one first has to understand how large the preceding thirty years were.

According to the China Elevator Association, roughly 70% of the world's elevators are manufactured in China and 60% to 65% are sold in China. This is one of very few industries in which China simultaneously holds the roles of "world's factory" and "world's market" — unlike smartphones or solar panels, which are export-led, or wide-body aircraft, which are import-led. Elevators are built in China, installed in China, and used in China.

That structure was poured in concrete by urbanisation and commercial housing. Around 2010, newly completed residential floor space in China was being counted in hundreds of millions of square metres a year; residential towers of six storeys and above became the norm, and two to four elevators per building became standard. Output climbed from a few hundred thousand units in 2010 to 1.545 million in 2021, with a double-digit compound annual growth rate. Every major global elevator brand — Otis, Kone, Schindler, thyssenkrupp, Mitsubishi, Hitachi, Toshiba, Fujitec — built factories in China, and for many of them the Chinese plant became the largest production base in their global network.

The starting point of this structure predates commercial housing. The modernisation of China's elevator industry began in the 1980s with joint ventures: the foreign partner brought drawings, processes and quality systems; the Chinese partner brought plants, workers and market access. Over the following two decades almost every international brand followed the same path — first joint venture, then wholly owned subsidiary, then integration of the Chinese plant into the global supply chain. In parallel, a cohort of domestic firms grew up inside the component layer: first guide rails, door systems and sheet metal work, then control cabinets, and finally complete elevators. That China can today design and build the full stack, from traction machines to control systems, is the direct result of those two decades of component accumulation.

The industry logic of that era was simple: follow the developers. Whoever secured a centralised procurement agreement with a large developer secured thousands of units of annual orders; the factory's job was to compress lead times and costs. Sales organisations were built around property developers, product lines around standard residential elevators, and capacity around the assumption that next year would be bigger.

It also created an extraordinarily dense supply chain. Manufacturing of elevators and components in China is heavily concentrated in the Yangtze River Delta, the Pearl River Delta and the Beijing-Tianjin-Hebei region, with the Yangtze River Delta alone accounting for more than 60% of national output; Suzhou in Jiangsu (the Wujiang and Fenhu area), Nanxun in Huzhou, Zhejiang, Shanghai, and Guangzhou in Guangdong are the most representative bases. Disassemble an elevator and you find the traction machine, the control cabinet and inverter, guide rails, the car and door system, the overspeed governor and safety gear, wire ropes and travelling cables, landing door frames, buttons and display boards, plus several hundred stamped and cast metal parts — and behind those parts stand thousands of small and medium factories. The spread of permanent-magnet synchronous traction machines also tied the elevator industry to rare-earth magnetic materials.

In other words, elevators are not an "assembler industry" but a "cluster industry". The assemblers sit in the Yangtze River Delta and their suppliers sit within thirty kilometres of them. When assembler orders for new installations fall, every link in that chain feels it.

There is also a difference that is easy to overlook: China's installed base is structurally unlike that of Europe, the United States or Japan. Europe's stock includes a great many units that have been in service for decades, distributed across low-rise older buildings, with few elevators per building, low duty cycles, and a mature maintenance market where prices hold up. Japan's stock is dominated by commercial and public buildings, with strict seismic and modernisation standards. China, by contrast, laid down 12 million units within twenty years, overwhelmingly in high-rise residential buildings and to broadly similar technical standards — meaning equipment ages are tightly clustered, duty cycles are generally high, and operating environments (renovation debris, overloading, rough use) are harsher.

Tightly clustered installation ages mean a tightly clustered ageing curve. This means China's replacement market will not be released gradually, as in Europe, but will arrive more like a wave of maturities with a clear schedule. For manufacturers that is both a risk and an opportunity: the risk is that safety incidents and public pressure will concentrate in the same period; the opportunity is that replacement demand will concentrate in that period too. Whether service capacity is put in place in advance will determine who catches that wave.

3. New installations peak: an unavoidable real-estate ledger

The 2025 property data is the necessary backdrop for understanding the new-installation market.

According to the National Bureau of Statistics, national real-estate development investment in 2025 was 8.2788 trillion yuan, down 17.2% from the previous year; new housing starts covered 587.7 million square metres, down 20.4%; housing completions covered 603.48 million square metres, down 18.1%; and sales of newly built commercial housing covered 881.01 million square metres, down 8.7%.

These four numbers mean different things to the elevator industry. Sales area affects developers' cash flow and confidence. New starts determine elevator orders two to three years out. Completions determine current-period installation and acceptance work. A drop of more than 20% in new starts in 2025 means demand for residential elevators delivered around 2027 will continue to contract — and the industry's internal debate about "when do we hit bottom" is essentially a bet on when new starts stop falling.

Once orders shrink, price is the first thing to give way. Listed companies describe it in restrained language, but the direction is consistent: demand for new elevators remains under pressure, and both new orders and prices continue to be squeezed. In the market itself, tender prices for standard residential elevators have fallen substantially over recent years, and some winning bids are now close to cost. Competition among assemblers, assemblers squeezing suppliers, and suppliers squeezing each other in turn all transmit down the chain.

Beyond price there is something more troublesome: the ledger. Elevators are a classic project business, with contract payments staged by milestone — deposit, delivery, installation, acceptance, retention. The cash-collection cycle for a single elevator often stretches beyond a year, with retention held for another two. In an upswing this is bearable, because developer credit is backed by land and pre-sales. When developers hit liquidity problems, the elevator maker's receivables are the first casualty: the goods have shipped, the elevator is installed, the crew has been dispatched, but the money has stopped at some milestone in the contract. The marked increase in credit impairment losses on the books of listed elevator companies in recent years is the direct expression of that ledger.

The transmission of this to the supply chain is harsher than price. Once the assembler's collections slow, it passes the payment terms on to its suppliers, who pass them to raw-material vendors. A chain already running on thin margins loses a further slice of profit simply because working capital turns more slowly. More and more firms now treat "customer mix" as a more important metric than "order volume" — better to take fewer projects than to take projects that do not pay on time. This is a management outlook that does not appear during a boom.

It is worth noting that output has fallen far more gently than property: down just 1.8% in 2025 against a more than 20% drop in new starts. The gap comes from three directions: replacement of ageing elevators is generating real orders; exports are growing; and non-residential segments (industrial plants, hospitals, schools, metro systems, airports, commercial retrofits) are more resilient than housing. Those three directions are precisely the three legs the industry is now betting on.

4. The other side of twelve million units

Now back to that figure of 12.3159 million.

The design life of an elevator is typically around fifteen years; how long it actually lasts depends on duty cycle and maintenance quality. On industry estimates, as of the end of 2024 roughly 900,000 elevators nationwide had been in service for more than fifteen years; by the end of 2025 that number had passed 1.1 million. A separate count for the first quarter of 2025 put elevators in service for more than twenty years at 296,200 units.

These numbers grow quickly every year, for a plain reason: the period around 2010 was China's first installation boom, and that cohort is now crossing the fifteen-year line en masse. The much larger installation peak from 2015 to 2021 will come due around 2030. In other words, today's "ageing elevator problem" is only the prelude; the real peak is still ahead.

Accidents and hazards: an industry under constant examination

According to the 2025 bulletin, China reported 188 special-equipment accidents, related accidents and related safety events during the year, with 156 deaths. The elevator sub-totals were: 11 accidents with 4 deaths; 26 related accidents with 35 deaths; and 14 related safety events with 5 deaths.

These three categories need explaining. "Accidents" refers to events caused by failure of the elevator itself — ascending car overspeed, hard landing, unintended movement. "Related accidents" refers to injuries arising in use: passengers caught in doors, falls on escalators, casualties among maintenance workers operating in the shaft. "Related safety events" sit between the two. Taken together, elevators accounted for 51 events and 44 deaths in 2025. Against a denominator of 12.3159 million units, that is a very low rate — China's overall elevator safety performance is not behind international levels. But to any individual family, rates mean nothing.

In 2025 regulators did two things that had not been done before. The first was a foundation-strengthening campaign for elevator safety, under which 132,000 elevators were inspected for hazards. The second was the release of the Rules for the Recall of Defective Special Equipment, under which five elevator and component manufacturers were, for the first time, required to recall 53,461 elevators.

Recalling more than fifty thousand elevators is a first in the history of Chinese special-equipment regulation. The signal it sends matters more than the number: elevators are now regulated as a product, not merely as a facility. A defective product must be recalled, and responsibility traced back to the assembler and the component maker rather than pushed entirely onto the operating entity and the maintenance contractor. For manufacturers, this significantly increases the long-tail liability attached to design and component selection.

A quiet rewriting of the inspection regime

A change that came earlier and is more structural than recalls is the reform of the inspection regime.

In 2023 the State Administration for Market Regulation issued a new edition of the Rules for Supervisory Inspection and Periodic Inspection of Elevators (TSG T7001—2023) together with the companion Rules for Self-Testing of Elevators (TSG T7008—2023); after a one-year transition, they were fully implemented from April 2024. The core of the reform is to split what used to be a single periodic inspection performed by inspection bodies into two parallel tracks: statutory supervisory and periodic inspection by accredited inspection bodies, and self-testing organised by the operating entity (or its appointed maintenance contractor). Self-testing must produce a memorandum and a report, which must be posted for at least fifteen days where passengers can read them.

The significance of this change only becomes visible against a denominator of twelve million units. Under the old regime every in-service elevator required a full annual inspection by an inspection body, meaning national inspection resources had to grow linearly with the installed base — and in 2025 there were 5,341 inspection and testing bodies nationwide holding 5,787 licences. There is a ceiling on human capacity. Delegating part of the technical checking to operating entities and concentrating inspection bodies on higher-risk links is a reallocation of resources made unavoidable by a swelling installed base.

At the same time it pushes responsibility forward: whoever operates it is responsible, whoever maintains it signs for it, and the results are posted at the elevator door for every passenger to see. For maintenance firms this converts service quality — previously something that could be handled vaguely — into a written commitment that must be signed, published and recorded. In a market competing on price, that record-keeping is itself a filter.

The workforce behind ten thousand elevators

The truly scarce resource in the age of the installed base is not equipment; it is people.

The 96333 elevator emergency response platforms operated by cities offer one window onto that workforce. In Nanjing, for instance, the 96333 centre in the single month of September 2025 answered 3,041 calls, handled 672 entrapment incidents, and rescued 1,433 trapped people, with rescuers arriving on site in an average of 10.7 minutes. Comparable figures repeat every month in every large and mid-sized city. In breakdowns of causes, door-system faults and external factors (renovation debris jamming doors, for example) usually rank near the top — which is both an equipment issue and a usage issue.

What sustains this response system are maintenance crews distributed across every urban district. Their working rhythm is set by two things: the routine servicing required by contract, and the entrapment call that may come at any moment. A crew typically covers dozens to more than a hundred elevators across multiple residential compounds, and travel time is often longer than working time.

The economics of the job are not attractive. When maintenance is priced at around four thousand yuan per elevator per year, what remains for the technician after vehicles, tools, spare parts, insurance and overhead is very limited — while the work itself requires dual certification for electrical and special-equipment operations, entry into high-risk environments such as shafts and pits, and availability around the clock. The result is predictable: fewer young entrants, skilled technicians leaving for other industries, and a rising average age. This is one of the practical reasons that condition-based maintenance and remote monitoring keep being raised — they are not merely an efficiency question but a necessary means of keeping twelve million units safe with a finite workforce.

A knock-on effect that is often ignored: the stability of the maintenance workforce directly determines whether an assembler's aftermarket strategy is feasible at all. A company can decide in a meeting room what share of revenue maintenance should represent, but what actually determines it is whether it can retain tens of thousands of workers willing to climb into shafts, in three hundred cities.

Who pays for replacement

The hardest part of replacing an ageing elevator has never been technical. It is money, and people.

Full replacement of a residential elevator typically costs on the order of two to three hundred thousand yuan, depending on floors served and brand. Institutionally that money should come from the residential special maintenance fund, but the obstacles in practice are multiple: many older compounds never collected such a fund, or the balance has long been exhausted; drawing on the fund requires a vote of owners, and residents on high and low floors often disagree about the elevator; and in older reform-era or work-unit housing, ownership is so fragmented that no clear responsible party can be found.

This is precisely the case for fiscal money to step in.

5. Subsidies arrive: from "equipment renewal" to "retrofitting"

In 2024, replacement of ageing residential elevators was brought within the scope of ultra-long special treasury bond funding, becoming one of the priority areas of the "two new" policies (large-scale equipment renewal and consumer-goods trade-in).

In 2025 the subsidy took the form of a flat grant: 150,000 yuan per qualifying elevator replaced. In 2026 the method was significantly revised — from a flat grant to a tiered one based on the number of floors (or stops) served: 100,000 yuan per unit for nine floors and below, 150,000 yuan for ten to eighteen, and 200,000 yuan for nineteen and above.

The change looks technical but matters a great deal. Floors served directly determine guide-rail length, travelling-cable length, shaft work volume and total cost; a 24-storey elevator can cost more than twice a 6-storey one. A flat grant under-subsidises high-rise buildings and over-subsidises low-rise ones, misallocating funds. Tiering improves the match between subsidy intensity and actual cost.

The more substantive change in 2026 is something else: for the first time, retrofitting elevators into existing residential buildings receives national-level subsidy — 250,000 yuan per unit for buildings of four to six storeys and 300,000 yuan for seven storeys and above.

Retrofitting had until then been a field where local governments improvised: standards varied widely by province and city, mostly in the range of a hundred-odd to a couple of hundred thousand yuan, with uneven disbursement. Central funding turns retrofitting from a local livelihood project into a nationwide policy product — which, for contractors and equipment suppliers, means a market that can be planned against a national baseline.

Execution data already shows an effect. On "two new" policy figures, 17,000 elevators have been retrofitted and 150,000 ageing elevators replaced nationwide since the start of 2025. In Beijing, 911 elevator replacements received treasury bond support in 2024; in 2025 that rose to more than 10,000, of which 9,008 had started work and 1,692 were complete. Beijing has retrofitted 3,391 elevators cumulatively since the start of the 14th Five-Year Plan period. Chongqing completed 6,073 units of ageing-housing renovation and retrofitting combined in 2025.

Set 150,000 replacements against annual output of 1.401 million units and the ratio already exceeds ten per cent. This is the most certain increment the elevator industry can obtain during a downcycle in new installation. And its character differs sharply from new-build orders: the decision-makers in a replacement project are owners and neighbourhood offices, not developers; the payers are fiscal funds and maintenance funds, not property companies; and tenders are judged on construction management capability and the ability to fit an existing shaft, rather than on centralised procurement pricing. This effectively requires assemblers to rebuild a sales and engineering organisation aimed at consumers and grassroots government.

Retrofitting is hard in a different way from replacement. Replacement means swapping equipment in place — familiar work. Retrofitting means hanging an elevator onto a building that never had a shaft, which touches structural safety, daylight and natural lighting, ventilation and fire-escape routes, and utility relocation — and then the hardest part: getting the neighbours to agree. Residents on lower floors barely use the elevator yet bear the loss of daylight and the noise of construction, and many projects stall exactly there. Central subsidies change the arithmetic of that negotiation: when 250,000 to 300,000 yuan per unit comes from the public purse, each household's share falls sharply and the probability of agreement rises. This is why the industry calls 2026 the turning point for the retrofit market — what has been unlocked is not the technology but the consent rate.

The limits should be seen too. Fiscal subsidy covers the portion that "should be replaced but cannot afford to be". It accelerates replacement; it does not create new buildings. The more than 1.1 million elevators over fifteen years old form a pool that will release over a decade, not a one-off flood. And predictable, smooth demand that does not depend on the property cycle is exactly what the industry needs.

There is a further effect that is easy to miss: replacement and retrofitting swap the industry's customers — from a few dozen large developers to hundreds of thousands of owners' assemblies, neighbourhood offices and property management firms. The purchasing logic of the former is centralised price pressure; that of the latter is "who can actually get this done". Sales organisation, settlement methods, project management and after-sales commitments all have to be rebuilt. Whoever completes that organisational rebuild first will establish position first in the installed-base market.

6. Exports: a leg that has genuinely grown

If modernisation demand is an increment pushed by policy, exports are one the industry won for itself.

According to customs data compiled by industry associations, China exported 126,833 passenger elevators in 2025, worth 16.011 billion yuan, along with 13,222 escalators and moving walks worth about 3.079 billion yuan. In the first half of 2025, passenger elevators and escalators together totalled 61,881 units and 8.596 billion yuan, of which passenger elevators accounted for 55,957 units, up 22.5% year on year. By market, demand was strongest in the Middle East, Central Asia and Southeast Asia; the United Arab Emirates was the largest single market with 9,299 units and 1.3019 billion yuan, followed by Saudi Arabia and Russia.

Exports of a bit over 130,000 units represent less than a tenth of output of 1.4 million — but their significance exceeds that share.

First, what is exported are complete elevators, not parts. China's elevator internationalisation is not travelling the old road of contract manufacturing for global brands — that road was completed twenty years ago, since global brands' Chinese plants already supply the world. What is growing now is Chinese brands going abroad with their own trademarks, their own installation crews and their own spare-parts systems.

Second, the structure of export markets differs from the domestic one. High-rise housing, hotels and commercial complexes in the Middle East and Central Asia are in a construction boom that closely resembles China's own in the 2010s; these markets were previously dominated by European and Japanese brands, at price levels well above those in China. For manufacturers used to a domestic price war, the gross-margin headroom abroad is a novelty.

Third, exporting forces Chinese elevator makers into a stricter rulebook. Elevators are a heavily regulated product: entering the EU means passing EN 81-20/50 and the Lifts Directive; entering the Middle East and Southeast Asia means local certification and acceptance regimes, plus building local installation and maintenance networks — because selling an elevator is only the beginning, and a brand that cannot install or service well is finished. This process steadily filters out those looking for a one-off sale and leaves those willing to build service organisations abroad.

Fourth, the payment terms of export business differ from domestic ones. Overseas projects are typically settled by letter of credit or advance payment, with far greater collection certainty than domestic property projects; meanwhile overseas spare-part prices and service rates are markedly higher than in China, so the service revenue generated over an elevator's life is considerable. For firms struggling in the domestic maintenance price war, this is structurally healthier revenue.

The cost is heavy investment. Building an overseas service network means warehouses, stocked parts, trained technicians and a crew able to reach a site within contractual time. Below a certain shipment volume that investment is pure loss; only when the installed base reaches a certain density does it amortise. Overseas markets therefore have their own "lose first, earn later" threshold, and getting through that phase is the real barrier to internationalisation for Chinese elevator brands — and the biggest difference between this industry and sectors such as solar or home appliances that can go abroad on pure trade.

From a cluster perspective, the pull of exports is selective: it lifts assemblers and high-value components (traction machines, control systems, door operators) but does less for low-value hardware, since overseas projects often require a share of local sourcing. This is why, among elevator suppliers in the Yangtze River Delta, those making controls, drives and door systems have had an easier few years than those doing sheet metal and stamping.

7. Competitive landscape: foreign strength in the aftermarket, domestic openings in replacement

The competitive structure of China's elevator market has long been "foreign brands in the premium segment and the installed base, domestic brands in the mid-market and the increment".

On publicly comparable market-share measures, the largest domestic listed player, Shanghai Mechanical & Electrical Industry (whose subsidiary Shanghai Mitsubishi Elevator is the operating business), sits at roughly ten per cent measured by principal business revenue, with Canny Elevator in the low single digits, and others such as Guangri and Jiangnan Jiajie holding regional strengths. Foreign and joint-venture brands including Otis, Kone, Schindler, thyssenkrupp, Hitachi, Toshiba and Fujitec together still hold a substantial share, particularly in office towers, commercial complexes, airports and metro systems, where requirements for speed, capacity and reliability are higher.

Entering the age of the installed base, this structure shifts subtly: the strengths of foreign brands lie precisely in the aftermarket.

There are two reasons. One is installed base: the highest-quality commercial projects of the past two decades were largely won by foreign brands, and maintenance and modernisation of those elevators naturally flow back to the original manufacturer. The other is organisational capability: multinational elevator groups have long derived the majority of their global revenue from service, with maintenance and modernisation accounting for a large share of turnover, and the China business is simply replicating a mature model. Public information indicates Otis's maintenance portfolio in China grew about 15% year on year in the first half of 2025, with modernisation revenue up more than 20%; Kone has likewise identified maintenance and modernisation as the two priorities for its China business.

The window for domestic firms lies in residential replacement and retrofitting. These markets are fragmented, low in unit value, heavy in cross-department coordination and cost-sensitive — precisely where multinationals' organisational cost structure is at a disadvantage, and precisely the battlefield Chinese firms have spent twenty years mastering. Shanghai Mechanical & Electrical has explicitly identified the aftermarket — servicing, spare parts, repair, modernisation and replacement — as a strategic direction, listing elevator replacement, retrofitting in older buildings, home elevators and exports as new growth drivers; Canny Elevator, in its annual report, characterises the industry as being in "an era of dramatic change in operating environment and demand structure", pointing the same way.

At the component level the shift has been more thorough than at the assembler level. Permanent-magnet synchronous traction machines, variable-frequency drives and elevator-specific control systems — all once imported — are now supplied mainly by domestic firms, with a cohort of companies from industrial automation entering the elevator supply chain on the strength of drive and control technology. The path mirrors hydraulics in construction machinery and inverter compressors in appliances: first validated in the market with the largest demand, then working back into the global supply chain. Elevators produced today in foreign brands' Chinese plants already use large volumes of Chinese components, and use China as a base for supplying overseas markets.

For that reason, framing this as "foreign versus domestic" is no longer very accurate. What distinguishes competitiveness is no longer shareholder background but three things: whether a firm can win maintenance contracts on the installed base, whether it can execute modernisation work in existing buildings, and whether it can roll out a service network overseas.

One emerging segment deserves separate mention: home elevators. Villas, self-built houses and age-friendly renovation have created demand for small-capacity, small-shaft, low-speed products, and the sales logic here is closer to consumer appliances — individual buyers, brand premium, fragmented channels, and installation service as word of mouth. Volumes are still modest, but this is one of the few parts of the elevator industry whose demand is not determined by property investment, and one of the few where domestic firms and foreign brands start from roughly the same line.

The decisive factor may not be market share but service network density. Installed-base business is fundamentally about response time: when someone is trapped, rescue must arrive within a set time; when a part fails, the replacement must arrive the same day. That means maintaining outlets, inventory and skilled technicians across hundreds of cities. Whoever has the denser network and the lower cost per node survives the maintenance price war. This is a heavier fight than manufacturing, requiring quite different capital expenditure and management reach.

8. What happens to the supply chain

For a supply chain organised around new-build assembly, a change in the structure of demand transmits unevenly.

As new-installation orders fall, the first to be hit are standard and low-value machined parts — car sheet metal for standard residential elevators, landing door frames, generic hardware. These products are highly homogeneous with weak bargaining power, so when orders thin the price war is especially brutal.

Demand generated by modernisation is different in kind. Replacing an old elevator can be done several ways: full removal and reinstallation; replacing only the machine and control system; or upgrading only safety components (adding brake monitoring, governor testing, unintended car movement protection). The latter two impose exacting requirements for fitting an existing shaft — old shafts are non-standard in dimension, requiring custom traction-machine mounts, rail brackets and door-system conversion parts. That is exactly the strength of small and mid-sized suppliers: small batches, high variety, fast turnaround, work that is uneconomic for large plants.

Retrofitting is almost a separate sub-industry. It requires shaft steelwork, machine-room-less or compact machine-room configurations, compact traction systems suited to old buildings, and a great deal of on-site construction and structural reinforcement. Firms doing retrofitting are frequently not traditional assemblers at all but a new type of company sitting between building construction and equipment installation.

Exports pull a different set: control systems, drives, permanent-magnet synchronous traction machines, door operators and safety components able to pass overseas certification. Certification is the most tangible barrier here — a CE or local certificate often takes one to two years and a non-trivial outlay, but once obtained it is a moat.

Another link being rewritten is spare parts. In the age of the installed base, the spare-parts business scales steadily with equipment age: twelve million units, each requiring replacement of wire ropes, guide shoes, door sliders, contactors, brake friction pads and door-operator belts. This business has historically been dispersed among original manufacturers, agents and a mass of small traders, with chaotic pricing and uneven quality. Once regulators begin tracing defect liability, and once self-testing requires records to be kept and posted, the origin and quality of spare parts become questions that must be answered. This will push the parts supply chain toward two poles: an original-equipment certification system at one end, and qualified independent suppliers able to provide traceability documentation at the other — while the space for untraceable cheap parts is compressed. For component factories, "can we get onto a given assembler's approved parts list" is becoming worth more than "can we win one more new-build order".

Layer these three demand types together and the stratification of the supply chain becomes sharper: some factories move toward service — smaller batches, more customisation, faster response; others move upmarket — higher precision, stronger certification, more expensive components; and those left making only standard parts, surviving on volume and payment terms, will be cleared out in this cycle.

This script is not unique to elevators. It closely resembles what construction machinery, home appliances and building materials each went through after their own peaks: the moment output stops growing, the industry's organising principle has to be rewritten.

9. Variables worth tracking

First, the turning point in new housing starts. This is the most reliable leading indicator of new-installation demand, with a lead of roughly two to three years. Until it stops falling, price pressure in the new-installation market will not ease.

Second, annual replacement volumes for ageing elevators, and the continuity of subsidies. The 150,000 replacements in 2025 were achieved with treasury bond support; the subsidy method changed to tiering in 2026, and policy intensity for 2027 and beyond is not yet defined. This demand is highly policy-dependent.

Third, signs of maintenance prices bottoming out. As long as average maintenance prices keep falling, the industry cannot genuinely build service capability. Whether condition-based maintenance moves from pilot to general practice is the key to this variable.

Fourth, whether export growth is sustained. Construction cycles in the Middle East and Central Asia will not last forever, and the pace at which local production and local service networks are built will determine whether Chinese elevator brands can stay in those markets.

Fifth, the intensity of enforcement of the defect recall system. The first recall of 53,461 units in 2025 set a precedent; whether recalls become routine will directly shape assemblers' attitudes toward component selection and design margins, and therefore the quality standard of the entire supply chain.

10. The two lives of an elevator

The life of an elevator runs roughly like this: a design life of fifteen years, an actual run of twenty or more, several hundred trips a day, well over a hundred thousand cycles a year, serving dozens to hundreds of households, punctuated by several hundred maintenance visits, a number of major repairs, and one or more modernisations.

For the past thirty years, China's elevator industry has mainly been in the business of day one of that life — building it and installing it. The scale of that business was set by real estate, its rhythm by developers, its price by centralised procurement.

From 2025 onward, the industry has no choice but to run the remaining seven thousand-odd days as well. Each of those 12.3159 million units is a twenty-year service contract, a replacement order that will arrive sooner or later, and an emergency call that must be answered. The scale of this second business is set by the installed base, its rhythm by the ageing curve, and its price by whether service quality can be told apart at all.

The first business built the largest elevator manufacturing industry in the world. Whether the second can be built as successfully turns on a harder question: can a manufacturing industry accustomed to delivering by the unit learn to serve by the year.

Data sources and principal references

All facts and figures in this article are drawn from public sources and cross-checked across multiple sources; where a figure is an industry estimate rather than a statutory statistic (such as the ageing-elevator installed base or average maintenance prices), the basis and nature of the source are stated in the text, and no subjective extrapolation is made.

  • Tianxia Gongchang Industrial Platform — China factory database and industrial-chain structural data
  • State Administration for Market Regulation: Bulletin on the National Safety Status of Special Equipment 2025; Rules for the Recall of Defective Special Equipment
  • National Bureau of Statistics: National real-estate market data for 2025; output statistics for elevators, escalators and lifts among enterprises above designated size
  • China Elevator Association: China's share of global elevator manufacturing and sales; industry operating bulletins
  • General Administration of Customs: import and export statistics for elevators, escalators and moving walks (as compiled by industry associations)
  • Ministry of Finance and National Development and Reform Commission: policy documents on ultra-long special treasury bond support for the "two new" programme and the 2026 revision of subsidy standards
  • Beijing Municipal Administration for Market Regulation and relevant Chongqing authorities: public information on progress in ageing-elevator replacement and retrofitting of existing residential buildings
  • Shanghai Mechanical & Electrical Industry Co., Ltd. and Canny Elevator Co., Ltd.: industry assessments and business-structure disclosures in periodic reports
  • Otis, Kone and other multinational elevator groups: publicly disclosed information on their China businesses
  • China Newsweek, Xinhua News Agency, Securities Times and other media: field investigations into elevator maintenance and ageing-elevator replacement