GM's world still exists, but there aren't enough new cars.

Author｜Liu Xinyi Editor｜Mao Shiyang

Auto Pixel (ID:autopix)Original

SAIC-GM's farewell and rebirth arrived in almost the same month.

On August 5, SAIC and GM extended SAIC-GM's joint venture term by 20 years, until 2047.

Four words stood on site: Together to a New Realm. Along with the renewal, the new task for this joint venture was also confirmed.

Three weeks later, on August 28, the Buick Envision 2026 model will be discontinued in the US.

The two events are independent but coincided in the same month.

The Envision story goes back eleven years. At the end of 2015, the first batch of Envisions destined for North America departed from Yantai Port, entering Buick dealerships in the US, Canada, and Mexico the following year. Over the next ten years, cumulative exports reached 420,000 units. After 2017, it became GM's only model imported from China to the US.

Starting in 2018, the Envision bore a 25% tariff, and GM's application for exemption was unsuccessful. On January 22, 2026, GM announced stopping production of Envisions for the US market in China; the next generation will move to the Fairfax, Kansas factory in 2028.

By May this year, the inventory cycle for this model in the US had exceeded 300 days.

▍Old photo of Buick Envision at Yantai Port

One door is closing, while another is opening. According to the arrangement after the renewal, the Buick Zhijing series independently developed by SAIC-GM will be exported overseas starting from October this year, with target markets including the Middle East, Africa, South America, Mexico, and Asia-Pacific.

Tariffs shut out the US car market, and GM hands the world outside to Shanghai.

01

Outside the US, GM is Short of Cars

This renewal did not change the 50:50 equity ratio, nor disclose new investment amounts. What changed is the division of labor.

At the signing ceremony, GM Global Senior Vice President and China President Luo Xu said, SAIC-GM has established local capabilities in engineering, manufacturing, and quality, "We can bring these advantages to the international markets of the Middle East, Africa, South America, Mexico, and the Asia-Pacific region", relying on GM's existing sales and service layouts in these regions.

In organizational language, what he revealed is that SAIC-GM and Pan Asia define the products, while GM provides global channels outside the US market.

The specific implementation method is quite light. SAIC-GM will not build its own sales network overseas, but will fully reuse GM's existing system.

The first exported car is the Buick Zhijing E7, with the first stop in South Korea. GM Korea is responsible for pricing, marketing, and after-sales, while SAIC-GM is only responsible for production, supply, and technical support. Later in other markets, a similar model is expected to be used.

The flow of products and resources has changed. In the past, GM brought global products to China; next, SAIC-GM will send products developed in China back to the world.

This is not common in "Joint Venture 2.0". When most foreign parties allow Chinese joint ventures to go global, they first demarcate their core markets. GM did not give SAIC-GM just an edge market to test the waters; it opened the Middle East, Africa, South America, Mexico, and Asia-Pacific all at once, almost including everything GM owns outside the US. And it allows Chinese products to directly connect to its sales and service system built over many years.

Why now, why these markets, the answers are written on GM's two reports respectively.

▍Group photo when GM Brazil was founded in 1925

The first is North America. In Q2 2026, GM North America's adjusted EBIT margin reached 8.6%, full-size pickup share 43%, No. 1 for the seventh consecutive year. The company raised its performance guidance for the second time this year, raising the adjusted EBIT expectation to $14-16 billion.

This is a self-sustaining machine, expensive, large, high-spec, with brand premium, plus the cash flow of subscription services. It does not need external funding and less external products.

The second report is outside the US.

In 2025, the US market contributed nearly 70% of GM's global sales. Following the US were China, Uzbekistan, Brazil, Kazakhstan, Colombia, Chile, Ecuador, Peru, Kuwait. Apart from the US and China, the rest are all emerging markets.

The cars these markets want have almost no overlap with the cars produced by that machine in North America.

And GM's position in these places is very uneven.

Mexico remains GM's stronghold, market share long second only to Nissan. South America slowed down in defense: In 2025, GM's share in Brazil dropped from 12.0% to 10.3%, Volkswagen rose to 17.1%, BYD rose to 4.4%.

The absence in the Middle East is even more obvious. In Saudi Arabia, GM has fallen out of the top mainstream brands, while SAIC MG has entered the top sales list. Africa has left a damaged network; Egypt still has manufacturing and operational systems, but GM has long exited the major business of South Africa and East Africa.

For Mexico and South America, SAIC-GM is defending the ring; for the Middle East, it's supplementing products; for Africa, there is a hint of returning to old markets.

02

Ultium Failed to Stitch GM Together

In 2014, Mary Barra became GM's first female CEO.

The GM she took over had just emerged from bankruptcy restructuring, still retaining a huge global business map. Opel and Vauxhall in Europe, Chevrolet laying networks in India, South Africa, and Southeast Asia, Holden stayed in Australia, Korea Daewoo system responsible for small cars, Brazil responsible for South American products.

These businesses allowed GM to maintain the image of a global automaker, but also kept it burdened with the costs of a global automaker for a long time.

Mary Barra adopted a clear financial standard. If a business cannot generate sufficient returns, it is no longer retained due to scale, history, or global status.

From 2015 to 2020, GM began to intensively take action against markets that could not generate sufficient shareholder returns, withdrawing from Europe, Russia, India, Southeast Asia, and Australia, selling Opel, Vauxhall, and the Thailand factory.

Profit discipline achieved results, capital gradually flowed to large cars North America was best at; the cost was that GM simultaneously dismantled the product system serving global needs.

Opel once provided front-wheel-drive sedans and European engineering capabilities, Holden mastered rear-wheel-drive platforms, Korea was responsible for low-cost small cars. After exiting these markets, GM also lost most of the capabilities to serve global needs.

By 2020, GM had basically completed global contraction. North America kept pickups, large SUVs, and high-profit fuel cars; China still had a huge joint venture system; markets like South America, Korea, Middle East were compressed into regional businesses maintaining their own profitability.

They were like isolated islands scattered on the world map. With the competitive landscape unchanged, they could continue to exist relying on inertia, contributing limited but stable profits.

But now, competition has intensified.

In 2025, the Brazilian car market continued to grow, but GM's share dropped from 12.0% to 10.3%; Volkswagen rose to 17.1%, BYD rose to 4.4%. In the Mexican market, GM's sales dropped 3.4%, while Nissan, Toyota, Kia, and Mazda all grew.

The results are written on GM's International Operations (GMI) report outside North America. In 2025, GMI wholesale sales dropped 8.1%, revenue dropped 3.3%; excluding Chinese joint venture business, adjusted EBIT dropped 32.8%, leaving only $426 million.

What these regional businesses lacked was not brand and channels, but a set of new products that could be continuously delivered to dealerships.

GM certainly knew the side effects of contraction. Once contraction was basically completed, Mary Barra had to find a new technical base, without re-establishing those regional R&D centers, to reconnect North America, China, and the remaining international markets.

Electric vehicles seemed to provide this opportunity exactly.

By 2019, Tesla had completed the difficult climb of Model 3, delivering 367,500 units for the year; the Shanghai factory from groundbreaking to delivery took less than a year, proving that a highly centralized model, battery, and software system can be quickly replicated between the US and China.

The signal from the capital market was more direct. In January 2020, Tesla's market value surpassed GM and Ford combined for the first time, and exceeded Toyota six months later. For GM which had just completed global contraction, electrification not only represents the future of the auto industry but also provides a shortcut to cover the global again without rebuilding the old system.

Starting in 2018, GM turned the funds saved after cutting losses to electrification and autonomous driving; by March 2020, GM officially released Ultium and promised to invest over $20 billion before 2025. China subsequently held the corresponding technology release.

Ultium's engineering ambition is huge. GM hopes it will simultaneously cover compact crossover vehicles, SUVs, Hummers, and large pickups, and allow North America and China to share scale, supply chain, and R&D costs.

But the product that first proved Ultium was the Hummer EV equipped with about 205 kWh batteries. This number is double the common battery capacity of China's mainstream large pure electric vehicles. From the beginning, Ultium was pulled towards American-style large cars by the goals of ultra-long range, thousand-horsepower power, and large pickups; it can continuously add modules to cover upwards, but it is difficult to use the same system to build down cheap models.

More fatally, GM bet on the wrong direction of market convergence. Ultium bet on highly modular battery cells, modules, and e-drive to拼出 as many models as possible; the place where markets truly form barriers has shifted to software, electrical/electronic architecture, and continuous iteration capabilities, while model diversity is actually converging.

In April 2026, GM notified suppliers to indefinitely postpone the development of the next generation Silverado EV and Sierra EV models; major platform updates are not expected until after 2030.

China, however, quickly shifted to lower-cost lithium iron phosphate, faster charging speeds, plug-in hybrids and range extender, and smart cockpits and assisted driving driven by local suppliers.

The base originally used to unify China and the US ultimately failed to connect either end.

Ultium did not stitch GM back together after it was dismantled. It left behind battery, e-drive, manufacturing, and safety verification capabilities, and made an inescapable conclusion: North America, China, and outside the US can no longer be defined by the same base.

03

SAIC-GM, Established Twice

At the end of 2024, SAIC-GM was not facing how to go global, but whether this company was worth continuing to exist.

That year, SAIC-GM sales dropped to 435,000 units, down 56.54% year-on-year, the largest decline among SAIC Motor's whole vehicle enterprises. In December, GM accrued 2.6 to 2.9 billion USD in impairment for its Chinese joint venture business, and confirmed about 2.7 billion USD in restructuring costs, totaling over 5 billion USD.

This money was both a settlement for the past and drew the survival line for SAIC-GM.

Ultium did not bring the expected transformation to the China business, and Detroit had no other mature global plan to send over. SAIC-GM must reduce scale, restore profitability, and rely on the China team to re-answer the most basic question: How exactly should the next generation of products be built.

In April 2025, the Xiaoyao Architecture and Buick's high-end new energy sequence Zhijing were released together.

Xiaoyao retained Ultium's e-drive, manufacturing, and safety accumulation, but the technical route has shifted to China: batteries switched to 6C Lithium Iron Phosphate, power covers pure electric, plug-in hybrid and range extender, electronic/electrical architecture also shifted to central computing.

More important than parameter changes is that product definition rights remain in China.

In the past, Pan Asia mainly adapted Detroit platforms for China. With Xiaoyao and Zhijing, technical routes, product forms, prices, and supply chains started to be defined by the China team.

The results came faster than expected.

Zhijing L7 first stabilized in the mid-to-large new energy sedan market; on April 22, 2026, Zhijing E7 went on sale, with the lowest trade-in equity price of 154,900 yuan, breaking 10,000 major orders within 90 minutes of launch, exceeding 10,000 deliveries in the first month, becoming the first joint venture new energy model to achieve this.

The reversal at the operational level started earlier. In 2025, SAIC-GM sales rebounded to 535,000 units, up 22.99% year-on-year; GM's business in China achieved profitability for consecutive quarters.

It first proved it no longer needed continuous blood transfusion from headquarters, and then proved that new energy vehicles defined by the China team can obtain the market.

Thus, the direction of renewal discussion changed. From whether SAIC-GM needed to continue existing to how to maximize its value.

This capability did not appear suddenly. In 1997, Pan Asia and SAIC-GM were established on the same day, going from localization, model modification all the way to whole vehicle forward development. In 2010, the new S-Orio developed throughout Pan Asia went on sale and went overseas; in over twenty years, it developed over 30 series and over 150 models for Buick, Chevrolet, and Cadillac.

In the past, this capability was mainly used to change GM's global products into what China needs. In the Zhijing era, the direction reversed; products defined by the China team began to fill the gaps in GM's global product line.

Chinese joint ventures outputting products in reverse is no longer an isolated case. Volkswagen, Nissan, Mazda, and Kia are all sending models developed and produced in China overseas.

The real difference is the scale of openness; the more important overseas profits and capacity are, the clearer the boundaries drawn by the foreign side are; the more the overseas system needs replenishment, the larger the space gained by the China team.

GM's speciality lies in that its most protected North American profit core has almost no conflict with products SAIC-GM can provide. What GM needs most to protect is large pickups and SUVs in North America, which happen to fill the product gaps in GM's overseas network.

South America, Korea, Middle East, and Africa still have GM's brands, factories, and channels, but new product supply has become thin. Incorporating Chinese models is faster and cheaper than rebuilding R&D systems for each region.

Eleven years ago, the Envision sailed from Yantai to the US, China's role in GM's global system was still a manufacturing base. Eleven years later, SAIC-GM went global again with products it defined itself.

What GM handed over was not a world still growing, but a network where channels are still valuable but products are increasingly cut off. What SAIC-GM obtained was not a ready-made market, but the power to repair it.

This 20-year contract is not just a continuation of the previous joint venture relationship. It is more like SAIC-GM's second establishment.

This article is original content by Auto Pixel (autopix)

Unauthorized, do not reproduce