Because a supplier that does not own its production can be replaced, and a supplier that does can set terms.
Automotive supply has always contained two very different kinds of business. One designs a component, contracts its manufacture to somebody else's plant and takes a margin on the difference. The other owns the building, the tooling, the process knowledge and the people who run them. In a stable market both can be profitable. In a market being restructured around electrification, only the second survives a programme cancellation.
Bosch operates roughly 100 dedicated automotive production sites inside a network spanning more than 60 countries, and it fabricates its own automotive-grade silicon carbide wafers — the base layer of the inverters every electric vehicle needs. CATL controls the chain from lithium refining through cell production to a recycling operation that processed 210,000 tonnes of spent batteries in 2025, and holds 772 GWh of capacity with 321 GWh under construction. Aisin melts aluminium ingots, die-casts them under high pressure and assembles the finished solenoid valve bodies on its own lines. None of those positions can be replicated by a competitor writing a purchase order.
Revenue alone cannot distinguish between them. A design house with a large contract and a vertically integrated manufacturer can report similar sales while carrying completely different risk. Factory ownership is what determines whether a company can hold price when its customer demands a reduction, and whether it can absorb the cost of a cancelled platform — which is precisely the event that produced ZF's EUR 1.6 billion write-down and forced Continental to give its electronics division away.
That is why production scale and capacity carries 40% of the weight here — more than revenue, more than brand, more than any single technology claim.
Disclaimer: Rankings are compiled from publicly available filings and independent research. VerityRank does not accept payment for inclusion or position.
Three things at once: the defect rate, the qualification time, and the fact that the customer can end the programme.
Start with the defect rate. CATL describes its manufacturing as reaching PPB-level defect control — parts per billion — because a single contaminated cell in a battery pack can cause a thermal event that destroys a vehicle and a brand. A consumer electronics factory operating at parts per million would be considered excellent. Automotive safety components are held to a standard two or three orders of magnitude tighter, and the cost of achieving it is borne entirely by the manufacturer.
Then there is qualification time. A new braking system or steering component typically takes two to four years from design freeze to production release, during which the supplier builds test fleets, validates in extreme climates, and pays for tooling that generates no revenue. That investment is only recovered if the vehicle sells in volume for years afterwards, and the automaker is under no obligation to make it do so.
That is the third difficulty, and the one that has reshaped the industry. Vehicle programmes are cancelled. ZF took roughly EUR 1.6 billion of charges in 2025 after negotiating the early termination of electric powertrain contracts that would never have reached profitable volume. Magna impaired European capacity built for programmes that were delayed, and absorbed losses from the collapse of EV start-up Fisker. Bosch booked EUR 2.7 billion in restructuring provisions and plans up to 22,000 job losses by 2030.
A manufacturer in most other industries can redirect an idle production line toward a different customer or a different product. An automotive supplier frequently cannot: the tooling is designed for one component, the plant is contracted to one programme, and the equipment has no second-hand market worth mentioning.
Because they bought electronics and interiors to escape low-margin hardware, and discovered that software development costs more than it earns.
The strategic logic of the past decade was convergence. A supplier that made seats should also make the electronics that control them; a company that made lighting should also make the radar that sees by it. Forvia was built exactly that way, when Faurecia absorbed Germany's Hella in 2022 on the theory that interiors plus electronics would produce a higher-value supplier than either alone. Continental pursued the same idea internally, assembling automotive electronics, braking and tyres under one roof.
By 2025 both had reversed course. Forvia agreed to sell its interiors business to Apollo Funds for EUR 1.82 billion, a transaction expected to cut net debt by more than EUR 1 billion, and cancelled its dividend to accelerate deleveraging. Continental separated its automotive electronics division entirely, listing it in Frankfurt as Aumovio in September 2025 and keeping the tyres and braking hardware — the businesses with 13.6% and double-digit margins respectively.
The reason is arithmetic. Software-defined vehicle development requires engineers who must be paid every quarter, against revenue that arrives only when a programme reaches production years later. In a market where vehicle production is flat and price reductions are contractual, that mismatch is unsustainable inside a group whose traditional hardware businesses are also under margin pressure. Both companies concluded that the electronics operations would be worth more — and cost less to fund — as separate businesses.
The result is an industry splitting along cash-flow lines: manufacturing businesses that generate money, and software businesses that consume it, no longer housed in the same corporate structure.
The difference is not engineering capability or capital — it is the time between deciding to build a plant and producing parts in it.
CATL reached 772 GWh of battery capacity in 2025 with a further 321 GWh under construction, spread across fifteen plants in five countries, and reported capacity utilisation of about 96.9%. Building and ramping that volume in roughly a decade represents a construction and commissioning speed that European and North American suppliers have not matched for any comparable product.
HASCO illustrates the same capacity at component level. The company operates more than 300 research, manufacturing and service bases, mostly in China, and grew revenue 8.49% to RMB 183.99 billion in 2025 while Chinese automakers were forcing annual price reductions through their supply chains. Growing revenue and profit simultaneously during a price war is only possible if the underlying manufacturing cost falls faster than price.
European suppliers face constraints Chinese ones do not. ZF plans to remove 11,000 to 14,000 German positions by 2028, and Bosch up to 22,000 globally by 2030 — reductions that are negotiated with works councils and unions, and take years to implement. Closing a European plant is a political process; commissioning a Chinese one is a construction schedule.
The consequence for global sourcing is already visible. Chinese component makers now supply vehicle programmes in Europe, Southeast Asia and Latin America, and Western automakers have begun structuring joint ventures specifically to access that manufacturing pace — as CATL did with Stellantis, committing about EUR 4.1 billion to a 50 GWh lithium-iron-phosphate plant in Zaragoza, Spain.
The tooling is stranded, the workforce is protected by agreements, and the write-down lands in the supplier's accounts — not the automaker's.
An automotive production line is not a general-purpose asset. The presses, moulds, welding cells and assembly fixtures are built for one component of one vehicle programme, and when that programme stops, they have almost no alternative use and virtually no resale value. ZF demonstrated the arithmetic in 2025: after agreeing with customers to end electric powertrain programmes early, the company took roughly EUR 1.6 billion in one-time charges, which turned an operationally improving year into a reported net loss of EUR 2.1 billion. Adjusted EBIT margin had actually risen to 4.5%, above its own guidance.
Magna faced a sharper version of the same problem. Its exposure to EV start-up Fisker produced impairment charges and losses in its assembly operations when the company failed, and it separately booked USD 591 million of goodwill and intangible write-downs against its electronics reporting unit. Bosch set aside EUR 2.7 billion for restructuring as it reduced its automotive workforce.
Employment protection makes the adjustment slower and more expensive in Europe than anywhere else. German suppliers cannot simply close a site; they negotiate social plans, phased reductions and transfer arrangements, which converts a sudden loss of volume into a multi-year cost. ZF's plan to remove 11,000 to 14,000 German positions by 2028 and Bosch's target of up to 22,000 global reductions by 2030 are both consequences of programmes that no longer exist.
The strategic response is now visible in how suppliers contract. Manufacturers increasingly require minimum volume commitments, shorter tooling amortisation periods and cancellation compensation before committing capacity — terms that shift some of the programme risk back to the automaker that cancelled it.
Disclaimer: Rankings are compiled from publicly available corporate and financial disclosures. VerityRank does not accept payment for inclusion or position.
