Machine Tool Orders to 2027, Margins at 3.7%

Machine Tool Talent Research · Orders, Margins and Delivery Headcount

China's machine tool industry booked RMB 19.6 billion in first-half profit, up 110.8% year on year, on an average margin of just 3.7%. Orders stretch into 2027; the profit surge is a repair off a low base. The headcount being added sits in assembly, commissioning and application.

Published by SunTzu China | Data as of 28 October 2026

Source tiers: industry association operating data / customs statistics / listed-company disclosures / live job postings

Research question: machine tool orders in China are now booked into the second quarter of 2027, and first-half industry profit doubled. Does that make 2026 a boom year, and what should anyone hiring machining talent in China read into it? The two signals point in different directions. Profit rose 110.8% from a depressed base, yet the average margin is 3.7%, still below the 4.5% average across the machinery sector. Volume has returned; pricing has not. The headcount implications follow the orders, not the profit line.

Key readings: metal-cutting machine tool new orders up 20.5% in the first half | machining centre export volumes down 25.6% under tightened export controls | five-axis application engineer postings at RMB 15,000-25,000 a month in Shenzhen and RMB 10,000-15,000 in Changsha.

Three numbers are usable straight away. Metal-cutting machine tools booked a 20.5% rise in new orders, and that figure is the main valve for every downstream hiring plan. A five-axis application engineer with five years and a bachelor degree is quoted at RMB 15,000-25,000 a month by a listed builder in Shenzhen. And the machine tool fitter-and-commissioner occupation has entered provincial shortage lists, carrying an apprenticeship subsidy of RMB 4,000-8,000 per trainee per year, with a 20-30% uplift where the occupation is designated scarce.

This report reads four things out of the data. What the profit figure actually measures, and why hiring budgets should track the order book instead. How one industry now runs two opposite order cycles, metal-cutting up and metal-forming down, which forces a budget split. Where capacity goes as export controls tighten, and why that lands on domestic delivery quality. And what the five-axis role structure costs, where the talent comes from, and when to move.

On data provenance: industry-level figures come from the China Machine Tool & Tool Builders' Association half-year operating report, with revenue and profit on the association's surveyed-enterprise estimation basis; trade figures come from customs; company figures from periodic disclosures of Neway, Kede and Haitian Precision; pay anchors from live job postings. No institutional talent-gap estimate is used anywhere in this report.

I. Orders to 2027, Margins Stuck at 3.7%

Set the two figures side by side. In the first half of 2026 the China machine tool industry recorded revenue of RMB 530.9 billion, up 7.6%, and total profit of RMB 19.6 billion, up 110.8%, on an average margin of 3.7% against the machinery-sector average of 4.5%. Doubling profit and a bottom-shelf margin can both be true at once, and the reason sits in the base: industry profit has hovered at low levels for years, and last year's first half was a trough. This year's surge is a climb out of that trough. The association says it plainly: the margin "remains below the 4.5% average of the machinery sector, and profitability is still at a low level".

The split by segment matters more than the total. Metal-cutting machine tools turned RMB 102.6 billion of revenue, up 15.6%, into RMB 8.1 billion of profit, up 43.0%, at a 7.9% margin, the best in years. Cutting tools grew revenue 26.4% and profit 127.9% at a 13.5% margin, the most profitable niche in the industry. Function components grew 20.1%. At the other end, metal-forming machine tools saw revenue fall 3.0% and profit 3.2%, still declining, while abrasives lost RMB 5.5 billion, narrowed from a bigger loss but not out of it, with about 70% of that segment tied to construction materials, steel and cement demand that keeps shrinking.

Basis note: revenue and profit are the association's surveyed-enterprise estimation data; margins are segment averages. Source: CMTBA, Half-Year Operating Report 2026.

The hiring signal sits in the structure, not the headline. Money is concentrating in cutting machines and cutting tools, so headcount follows those two segments. The association attributes the cutting-machine surge to a specific downstream list: cooling-plate and server precision-structure parts for AI infrastructure, core components of humanoid robots entering volume production, commercial-scale low-altitude aircraft, and energy-storage structural parts and inverters. That list decides which roles a builder adds: people who can build, commission and hand over high-precision, intelligent, line-level equipment. The two cold columns, forming machines and abrasives, send the opposite signal: postings there backfill attrition, and net headcount does not grow.

The abrasive loss matters for hiring in one way only, as a restraint. A segment that loses RMB 5.5 billion while the rest of the industry repairs does not add headcount; at best it holds a service skeleton. Buyers consolidating abrasives suppliers should expect service capability to thin before prices rise, which is a sourcing risk rather than a hiring opportunity.

Company results already show it. Neway Numerical Control turned RMB 1.032 billion of third-quarter revenue, its first quarter above one billion, up 30.35%, with essentially every machine line at full capacity by end-August; contract liabilities and construction in progress rose together, which is expansion running alongside full utilisation. Haitian Precision's interim report describes its Ningbo intelligent machine tool base entering phased production with monthly output and utilisation still climbing, plus capacity additions in South China and overseas. Both companies are in the same posture: capacity runs ahead, people chase it.

Can the 110.8% figure be converted directly into a hiring budget? No, and this is the most common misreading. The number measures the speed of a repair, not earning power. A 3.7% margin means the industry keeps RMB 3.7 of every hundred; hiring budgets will not loosen because the growth rate looks large. Expansion lands precisely on roles that convert orders into cash: assembly, commissioning, application engineering and after-sales delivery. There is no matching evidence of a research hiring wave; Haitian's own language is "optimise the organisation, raise management efficiency and innovation capability, cut manufacturing costs" — cost discipline and delivery expansion run in parallel.

II. One Industry, Two Order Books

The association's order data is the most consequential set in this report for anyone planning headcount. Industry-wide, new orders for metal-working machine tools rose 9.2% in the first half and order backlogs rose 12.1%. Split apart, metal-cutting machine tools posted new orders up 20.5% and backlogs up 19.1%, both accelerating from the first quarter. Metal-forming machine tools posted new orders down 12.3% and backlogs down 0.3%. One industry, one statistical table, two order curves pointing in opposite directions.

Why is the forming side falling? The association names its customers: after the earlier expansion wave, new-energy-vehicle makers have entered a capacity digestion period, so stamping equipment demand is saturating, and falling combustion-vehicle output compounds it. Appliance stimulus has faded and property-linked demand stays depressed. In hiring terms, the teams expanded for stamping lines over the past three years now sit on a business that is digesting capacity; headcount there attrits naturally and is not replaced. The one growth pocket in forming is battery structural parts, aerospace precision forming and semiconductor equipment enclosures — and even there the hires are high-precision process and commissioning people, not general operators.

For budgets the consequence is direct. A group with both a cutting line and a forming line must split its recruitment budget. The cutting line staffs its delivery function against a 19.1% backlog rise, starting assembly, commissioning and application hiring one recruiting cycle early. The forming line backfills only, freezing net additions unless a battery-structure or aerospace order actually lands. Pool the two into one budget and both sides misfire: cutting hires arrive too late for deliveries, forming hires sit idle.

The export split mirrors the same divide. Metal-forming machine tool exports held at USD 1.63 billion, up 1.4%, while metal-cutting exports slipped 0.3% — overseas demand is what keeps the forming side of the industry alive while its domestic book shrinks. A builder reading only its own domestic order screen will underestimate how much of the forming business now depends on buyers in Southeast Asia and India.

There is also a timing lag to respect. The association expects second-half growth to continue "at a possibly slightly slower pace", and export growth already peaked in the first quarter. Orders signed now deliver three to nine months out; the backlog stretching into 2027 maps to a hiring peak between late 2026 and mid-2027. By the time a slower profit number reaches the news, half of the delivery-side hiring window has closed. In machine tools, hiring timing follows the order book, not the income statement.

The forming side has a timeline too. Battery structural parts and inverters are scaling now, and equipment selection for those lines is happening, but the headcount only lands when the orders show up in reports. The practical move is to hold the forming-line core team while watching for those orders, then hire with one recruiting cycle of lead time.

III. Export Controls Push Capacity Home

Customs data carries the variable most commentary misses. As dual-use export controls tightened, metal-cutting machine tool exports flipped from first-quarter growth to decline, and machining centre exports fell 25.6% in volume and 21.4% in value. Domestic orders are rising while the export channel narrows; squeezed between the two, capacity has one place to go — the domestic high-end market.

The export book itself is large. Machine tool exports ran USD 11.84 billion in the first half, up 5.8%, led by India at USD 1.05 billion, with Russia at 1.02 and Vietnam at 1.01, Southeast and South Asia the core growth markets. The controls bite in one slice, high-precision machining centres, which is exactly the category that demonstrates five-axis capability. The export capacity blocked there, added to substitution demand, lands on domestic customers: aerospace, machine-tool remanufacturing and precision tooling. Kede Numerical Control is the cleanest illustration: new orders up nearly 25% in the first half, backlog of RMB 266 million up 64.53%, five-axis machine localization above 90%, its in-house GNC62 control system above 95%, and volume deliveries already running for aero-engine integral blisk machining.

Imports carry a warning for anyone assuming substitution is easy. Machine tool imports rose 6.3% to USD 5.35 billion, with Japan first at USD 1.84 billion, up 11.6%. The association adds the key line: "the yen's depreciation has further strengthened the competitiveness of Japanese products". The substitution rival has not slowed; the exchange rate is cutting its prices. For domestic builders the pressure in the mid-to-high end is real, and delivery quality with response speed are among the few remaining differentiators — both of which depend on commissioning and application teams.

Import composition adds one more read. Machining centres took USD 860 million of the import bill, 32.9% of the total, and special-processing machines grew 21.0% — the segments where domestic five-axis capability is still being proven. Every imported machine placed in a Chinese plant creates a service and application tail: installation, post-processor work, operator training, spare-parts planning. Distributors and agents staff that tail, and it draws on the same small pond of five-axis talent as the builders themselves.

The net headcount effect of this redirection is additive. Domestic customers demand more than export orders did: sample cutting on site, process validation, operator training, each requiring people at the customer's floor. Dayang Numerical's investor disclosures describe its CCD six-axis drilling machine, carrying 3D back-drilling and drill-test integration, as certified for next-generation AI server PCB production and in volume use at several leading makers. The product name marks a new delivery scenario: after the machine lands, the builder's application engineers spend months tracking the customer's yield metrics.

IV. What a Five-Axis Application Engineer Costs

The five-era role structure splits into three layers. Commissioning sits in the factory: final assembly, geometric accuracy, control-system parameters and acceptance cutting. Application sits at the customer: machining plans, post-processor work, on-site debugging and training. Programming and process engineering sit between, turning customer part drawings into toolpaths and cycle times. The three layers demand different skills, work in different places, and price differently. Live postings show the spread. A Changsha equipment builder lists a five-axis application engineer at RMB 10,000-15,000 for three-plus years and a bachelor degree. Genesis Industrial Equipment in Shenzhen lists the same role at RMB 15,000-25,000 for five-plus years, requiring familiarity with "machining processes in any of: drones, robots, AI liquid cooling, new energy vehicles, aerospace, medical". Principle's Jinling Machine Tool lists RMB 10-15k in Changsha. All three postings converge on one control-system requirement: fluency in at least one of Siemens, Heidenhain, SynTEC or Huazhong CNC, with independent post-processor debugging. Across the three, the application engineer salary band runs RMB 10,000-25,000, and five years of five-axis machining experience is the common entry line.

Basis note: pay anchors from live postings, October 2026, not normalised across employers; subsidy figures are government standards, subject to local implementation rules.

Why does an application engineer price two to four notches above a commissioning technician of the same seniority? The answer sits in the delivery chain. The commissioning technician brings the machine to factory spec inside the plant, with standard procedures as a safety net. The application engineer faces the customer's parts, fixtures, cutting parameters and yield targets, none of which the factory standard covers. Genesis's posting lists the job plainly: trial-machine follow-up at the customer, machining-trouble technical support, customer staff training, efficiency optimisation and process improvement. Each item happens on the customer's production line; failure refunds not one machine but an entire line order. The role is part of the selling price, so its pay follows industry process coverage: add drone structural parts and the price steps up; add aerospace blisks and it steps up again.

The pricing mechanism here matches the silicon-carbide process-integration role: the most expensive person is not the one with the highest title but the one who covers the most process segments. It also contrasts with the four-band structure of general manufacturing roles, where pay follows the radius of responsibility. In one machine tool builder both logics run at once — a buyer benchmarking against an average "mechanical engineer" salary will overprice management roles and underprice application roles simultaneously.

One price band is missing from public markets: post-processing. All three postings fold it into the application role, and standalone post-processor engineer openings almost never appear on job boards. The skill grows inside factories; neither supply nor demand runs through open channels. If an employer needs that capability, keyword searches will not produce candidates. Two routes work: identify application engineers willing to specialise and develop them in place, or approach the post-processing teams inside peer factories directly.

V. Where the Talent Comes From

The supply-side change this cycle is that policy has written the fitter-commissioner occupation into shortage lists. Yangzhou's human resources bureau published its 2026 scarce-occupation catalogue in September, listing machine tool fitter-commissioners, multi-process CNC machine operators and metal heat-treatment workers in its advanced-manufacturing tier, with training subsidies up to RMB 4,300. Heilongjiang's document is more granular: enterprise apprenticeship training is subsidised by grade, RMB 4,000 per trainee per year at entry grade, 5,000 at intermediate, 6,000 at advanced, 7,000 at technician and 8,000 at senior technician, with up to 20% uplift for scarce occupations — RMB 9,600 a year at the ceiling. Guangdong's standard runs RMB 5,000-8,500 with a 30% uplift for scarce occupations at advanced grade or above.

The practical meaning: an employer that trains its own commissioning technicians recovers three to four tenths of the cost from government. An advanced-grade trainee at RMB 6,000, uplifted 20%, draws RMB 7,200 a year against apprenticeship wages — a material share. Yangzhou's catalogue carries a second signal: the multi-process CNC operator sits on the same list as the fitter-commissioner, so the province treats the operating tier and the commissioning tier as joint gaps and channels training resources to both. The procedure matters: employers must file the apprenticeship programme with the local HR bureau before training starts, may draw up to 50% in advance within three months of opening the class, and collect the balance after apprentices pass assessment.

The apprenticeship model is worth understanding before budgeting for it. Recruitment doubles as enrolment: the trainee joins the payroll on day one and splits time between the shop floor and a partner vocational school, with a dual-mentor structure pairing a shop-floor master with a school teacher. Trainees are typically 16 to 40 and remain on the company's payroll throughout. For an employer the cost maths follows: the subsidy covers a share of wages already being paid, so the marginal cost of a training cohort is close to the administrative overhead of the filing itself.

Where do trained technicians come from today? The development path does not compress. From apprentice to independent geometric-accuracy commissioning and acceptance cutting takes three to five years in the normal course. Technicians who can handle thermal-error compensation and RTCP calibration are produced one project at a time. China's five-axis commissioning talent concentrates in the head builders and core component factories, in a circle small enough that job changes travel by referral. Two consequences follow: sourcing runs against a named target-company list by industrial belt, not keyword trawling; and internal training should start inside the subsidy window, because catalogues and standards adjust annually and a missed filing year is a missed year.

The hiring radius is drawable. Five-axis builders cluster in four belts: Dalian (Kede), Ningbo (Haitian Precision, Neway), Shenzhen and Dongguan (Genesis), and Changsha (Jinling and peers). Commissioning technicians rarely relocate across regions; mobility runs inside a belt, so same-belt approaches succeed far more often than cross-province ones. Campus hiring is expanding in parallel: the capacity additions at Ningbo, at Neway's construction-in-progress and at Genesis's Huzhou base all lift entry-level demand, concentrated in assembly process, commissioning technology and technical service tracks at vocational-to-bachelor level — the same structure humanoid-robot actuator suppliers are hiring against, which makes their published pay curves a usable cross-reference.

VI. Limits, Indicators, and What This Means for Hiring

Five limits bound this report. Industry-level: revenue and profit are surveyed-enterprise estimates, and segment margins show direction without revealing company-level distribution. Order-level: new-order and backlog growth rates come from key-enterprise statistics without absolute values, so they cannot be converted into headcount per role. Pay-level: anchors come from live postings at different employers without normalisation between monthly and annual packages; they indicate magnitude only. Policy-level: apprenticeship subsidies differ by province and change annually; the Yangzhou, Heilongjiang and Guangdong figures cited here do not transfer to other provinces, and current local documents must be checked before use. Sample-level: Neway, Kede and Haitian are verification cases, not a statistical sample; their full order books do not represent the industry, and the forming-side decline carries three variables — combustion-vehicle decline, fading appliance stimulus, property depression — not quantified down to headcount.

Suggested cadence: association data quarterly, customs and filings on reporting schedules, posting pay quarterly, HR documents annually. None of the five indicators depends on third-party estimates; each traces to a checkable original.

For anyone running manufacturing recruitment in China, the first move is switching the budget basis from the profit line to the order book. A 110.8% profit rise invites hiring on "boom" logic; a 3.7% margin argues for restraint. When the two signals conflict, the only reliable anchor is a backlog growing 19.1% and deliveries scheduled into 2027. Budgeting on orders means: delivery-side roles in cutting lines — assembly, commissioning, application — staffed first, development and overheads held; forming lines backfill only; abrasives-linked businesses add nobody.

For five-axis application hiring, screen on three layers. Systems: fluency in at least one of Siemens, Heidenhain, SynTEC or Huazhong CNC with independent post-processor debugging is the entry gate. Process: the industry segments covered — drone structures, robot components, cooling plates, blisks — set the pay band. Fit: before opening the role, decide which industry your machines actually sell into and screen against that segment's process requirements. Screening on "five years of five-axis experience" alone excludes qualified candidates while admitting operators who have only ever run standard parts.

On development, use the apprenticeship window. The fitter-commissioner occupation is in Yangzhou's 2026 catalogue; Heilongjiang's ceiling reaches RMB 9,600 per trainee per year and Guangdong's standard tops at 8,500 with a 30% uplift, together covering three to four tenths of training cost. The sequence: confirm the occupation is listed in your province's current-year catalogue, file the apprenticeship programme with the HR bureau, draw the 50% advance within three months of opening. Unclaimed, the money is simply forfeited — and catalogues adjust annually, so there is no guarantee the occupation stays listed next year.

On timing, treat the window as 12 to 18 months. Backlogs scheduled into the first half of 2027 put the delivery peak between late 2026 and mid-2027, which sets the latest hiring start at the fourth quarter of 2026. The association has already flagged a possible second-half slowdown and export growth peaked in the first quarter; waiting for the slowdown to reach the news means starting a cycle late. Low-altitude aircraft and energy storage keep adding downstream demand, and an application team's size decides how much of it a builder can capture — the most practical answer to why now. Where outside support is useful, an executive search in China manufacturing engagement can compress coverage of the target-company list, with the cost structure fixed before signing.

This report uses no third-party talent-gap estimate and no purchased salary survey. Conflicting versions of both circulate for the machine tool industry, none traceable to an original sample. The association's key-enterprise statistics and customs data are the only industry-level basis used here.

Sources: China Machine Tool & Tool Builders' Association, Half-Year Operating Report 2026; China customs trade statistics; CCTV Finance reporting, 28 October 2026; periodic disclosures and investor materials of Neway Numerical Control, Kede Numerical Control and Haitian Precision; human resources and social security documents of Yangzhou, Heilongjiang and Guangdong; live job postings and regional job boards, October 2026.