CER is a C-level executive at a top healthcare company in China with a career of more than 2 decades spanning healthcare, internet platforms, and AI in China. CER writes about biotech, advanced manufacturing, and AI in his Substack.

We can’t think of anyone else other than CER who is more capable of commenting on the future of healthcare in China and its implications for the entire world.

We helped CER translate and edit this piece, originally written in Chinese.

China may attain healthcare’s Holy Grail: cracking the “impossible trinity”

by

The “healthcare Holy Grail" is a global problem.

Every healthcare system in the world is trying to solve the same impossible trinity: affordability, service quality and accessibility.

Few can deliver all three at once. Cheap care can erode quality. High-quality care can narrow access. Broad access can push costs upward. China matters because it is trying to move this global frontier at population scale.

The United Kingdom gets relatively close on universal coverage. NHS patients face low direct payment pressure and can access a high-quality clinical system. Accessibility is the constraint. Triage, referrals and queues turn “free or low cost” into delayed care. Nuffield Trust data show that in January 2026, 11.5% of CT waits in England exceeded six weeks. NHS England treats the six-week diagnostic target as a key prerequisite for the 18-week referral-to-treatment standard. The British model shows the tradeoff clearly: affordability and clinical quality can coexist while time becomes the bottleneck.

The United States sits at the other end. It has the world’s strongest high-end hospitals, specialty care, innovative drugs and medical devices. Affordability is weak, and accessibility depends heavily on insurance coverage, network status and ability to pay. KFF’s 2026 update shows that 36% of U.S. adults skipped or postponed needed care in the past year because of cost; among uninsured adults under 65, the share was 75%. HCCI’s imaging brief shows that in 2022 the average employer-sponsored insurance price for a CT scan was about $1,116, versus about $456 under Medicare. Those figures are not directly comparable to Chinese public prices, yet the order-of-magnitude gap says enough: American medical capability is strong, and so are the bill and the insurance gate.

U.S. friction starts with price and continues through insurance process. After the murder of UnitedHealthcare’s CEO, AP reported a surge of public anger online around insurance delays, denials and “delay, deny, depose” rhetoric. KFF data show that Medicare Advantage plans fully or partially denied 4.1 million prior-authorization requests in 2024; only 11.5% were appealed, and 80.7% of appealed denials were fully or partially overturned. A 2025 Commonwealth Fund survey similarly found that among working-age privately insured adults who had experienced a prior-authorization denial, roughly 40% reported delayed care and 28% said a health problem worsened.

China sits in a third position. National medical insurance, public hospitals and government subsidies let most people enter the system and keep many basic services affordable. The weakness is uneven quality. Good doctors, advanced equipment and better patient experience still concentrate in top-tier public hospitals. Patients can enter the system, but getting the best doctors and services can still mean queues, referrals, personal networks and cross-city travel. China is strong on affordability and basic access. It is still trying to spread high-quality service beyond the top hospitals.

China’s healthcare reform is aimed at the hardest part of the triangle: after making care broadly affordable and reachable, China is now using institutional reset and technological advances to make good care more widely available.

But before I explain in detail how China is dealing with this monumental task, I want to first explain to you why this matters for you.

Why does China’s healthcare matter for the world

Over the years, the Western healthcare system has built a complete referral system, strict prescription controls and a professional gatekeeping structure.

On the surface, this system is defended as patient safety. In practice, it also works as a collusion between medical groups and the state, with commercial insurers adding another payment gate. The result is a system that protects institutions first and squeezes consumers second.

The alignment is visible in the way the system behaves. Government protects system legitimacy. Physician groups protect the professional gate. Insurers protect the payment gate. Each side can justify its position with the language of safety, compliance or cost control, yet patients encounter the same practical result: higher consultation fees, higher premiums, longer waits and more uncertainty before a simple health problem can be solved.

The contradiction is sharper because this system has also failed to protect patients from real harm. Purdue’s opioid marketing, and McKinsey’s infamous consulting advice to Purdue, showed how commercial interests could purchase influence throughout the medical system under the cover of professional legitimacy. That institutional legitimacy helped feed America’s opioid disaster.

For many common, low-risk conditions, the real need is much simpler. A patient can often describe symptoms, search basic information, use AI for navigation, consult a pharmacist or remote doctor, and receive common medicines, including antibiotics when medically appropriate, without turning the whole episode into a high-cost institutional journey. Prescription responsibility still matters. The practical question is why every minor illness should be forced through the most expensive gate in the system.

China has left a gray tolerance zone that the West often refuses to allow. Free or very low-cost internet consultations, pharmacy-linked remote doctors and fast medicine delivery make it much easier to obtain common drugs that can solve immediate problems. Its operating logic is to keep complex judgment inside hospitals and doctors’ hands, while moving simple, frequent and low-risk demand into cheaper, faster fulfillment chains.

Subscribe now

Institutional innovation: from hospital center to multi-entry system

China’s earlier healthcare structure was hospital-centered, especially around high-level public hospitals. Patient entry, doctor resources, drug flows, diagnostic and treatment pathways, payment interpretation and trust all concentrated around the hospital node. This structure can handle complex care and basic coverage, but it also pushes many simple consultations, follow-ups, prescriptions, tests and explanations into high-cost in-hospital workflows.

Over the past several years, anti-corruption, volume-based procurement, DRG/DIP payment reform, sunshine income reform and prescription outflow have all been reorganizing that old structure. The point is less a moral indictment of doctors or pharmaceutical companies than a compression of non-clinical incentives and in-hospital channel rents. Resource-allocation power spills out of a small number of top hospital nodes, while both doctor income and hospital income are pushed toward more transparent structures.

DRG stands for Diagnosis Related Groups. It bundles inpatient cases with similar clinical processes into “payment groups”, so insurers pay by group, leading hospitals to have stronger incentives to control cost and standardize care.

DIP is diagnosis-intervention packet payment, a method closer to China’s own insurance practice: historical real cases and cost data assign point values to diseases or combinations, and settlement happens within a budget.

China’s National Healthcare Security Administration says that by the end of 2023, more than 90% of coordinating regions had launched DRG/DIP payment reform; the 2.0 grouping was built from 53.71 million DRG cases in 78 cities and 47.87 million DIP cases in 91 cities.

The core transition moves from a hospital-centered initial state to a future structure.

In the initial state, care, patients, drugs, channels and payers rotate around hospitals. “Care” means doctors and hospitals; “patients” are the demand side; “drugs” includes drug and device companies; “channels” include pharmacies and out-of-hospital service nodes; “payers” include basic medical insurance, commercial insurance, employers and other payment sources.

The future structure is a multi-entry, multi-payer, multi-channel service network. Hospitals remain the center for complex disease, acute care and difficult diagnosis, while basic entry points diffuse into community care, online care, pharmacies, home testing, home service, internet follow-up and compliant prescription loops.

In the transition map below, the horizontal axis is distribution inequality, similar to a quasi-Gini concept. It describes whether benefits, resources and service interpretation power are concentrated in a few in-hospital nodes, prescription authorities and payment chains. The vertical axis is ecosystem diversity: the richness of healthcare services, payment layers, drugs, devices, internet hospitals, home service, pharmacies and channel nodes.

Low diversity and high inequality define the initial state, while higher diversity and lower inequality define the future state. The transition map gives the overall direction: entry, payment and fulfillment nodes spread outward, while hospitals return to complex care and difficult diagnosis. Patient entry points multiply, channel nodes become denser, pharmaceutical companies and hospitals are recalibrated, and payers move from a single basic-insurance anchor toward a more layered organization. What China is essentially doing here is to pursue anti-corruption, payment reform, internet healthcare, drug and device industrialization, and innovation-drug/device capability upgrading at the same time.

Follow the money: who pays for China’s healthcare?

The payment structure shows why institutional reform is necessary. My directional estimate of China’s healthcare payment market reorganizes spending by final payer and risk-pool attributes. Currently, government payment remains the anchor, while patients’ out-of-pocket cash payment is still large. On the other hand, employer payment and commercial insurance are still small.

Future change can come from this underbuilt layer. Government policies increasingly frames human capital as an investment, while companies are also rethinking employee health through productivity, ESG and employer-branding lenses.

If commercial insurance, employer benefits and innovative payment can be improved, China’s healthcare system could avoid a narrow choice between basic insurance and household cash payment. Higher-quality services would gain another more stable layer of purchasing power.

Technology: pushing the frontier outward

While institutional reform opens the door, technology and industrialization decide how far the system can move. China’s healthcare story is policy, AI, digital service layers, cheaper equipment, innovative drugs, and dense service networks moving at the same time.

First comes application technology. China is more willing to apply internet healthcare, AI triage, report explanation, risk prompts and follow-up assistance. AI cannot replace physician diagnosis, prescription responsibility, or pharmacist review, but it can take on health navigation, risk stratification, follow-up reminders, report interpretation, and low-risk consultation entry. Many mild consultations, follow-ups, chronic-disease prescription renewals and standardized medication questions can be organized at lower cost.

Second is industrialized cost reduction. China turns many medical devices and testing services into industrial products. CT, MRI, ultrasound, lab tests, home sampling and portable diagnostic equipment can become much cheaper through manufacturing scale and dense services. A 2025 China Daily report on National Healthcare Security Administration pricing guidelines said CT prices in some provinces had fallen to around RMB 200, while MRI prices had fallen to around RMB 500. A 2024 People’s Daily Chengdu case showed CT plain scans falling from RMB 240 to RMB 200 per scan site, with reports available in about two hours. In general, domestic CT equipment costs about one-third of imported equipment, and the lowest ordinary CT scan price was under RMB 50. Lower equipment and service costs are themselves a form of accessibility.

Third is the strengthening of medical-device and new-drug R&D capability. China’s innovation-drug license-out activity has become an important source for the global biopharma industry. Publicly announced cases show that Chinese capability is extending from medical-device manufacturing into innovative drug assets and clinical value itself. For Chinese consumers, the real significance is better drugs, more reliable devices and lower total cost. Overseas revenue is only the external validation of that capability curve.

Fourth is service-network density. China’s population density, logistics and local-service infrastructure make home testing, medicine delivery, internet follow-up, pharmacy O2O and health management easier to build. The low-friction front door is made of homes, pharmacies, online consultation and offline fulfillment. Its value comes from organizing scattered touchpoints into one service network.

Conclusion: the grail is not captured in one move

China is likely to build a hybrid system: government insurance as the base, commercial insurance and employer payment supplementing higher-quality services, out-of-hospital channels handling fulfillment and follow-up, internet healthcare, AI and home service reducing entry friction, and innovative drugs and medical devices lowering cost through industrial capability.

The system has to prove that it can improve all three metrics at the same time: let more people access care, keep basic care affordable, and allow higher-quality services to spill beyond a small number of top public hospitals. Institutional innovation breaks up the old structure. Technology and industrialization lower cost. Dense service networks move capability closer to patients.

The impossible trinity will remain. The more realistic change is that the feasible boundary moves outward. China is trying to break services that used to require high-cost institutions into smaller, cheaper, and denser modules, then reconnect them through compliant payment and service networks. If this route works, overseas readers should pay attention to a healthcare operating system that is cheaper, denser, and easier to scale.

The arithmetic at the point of issuance is almost misleadingly modest. If CXMT raises RMB 29.5 billion by selling roughly 10% of its post-IPO equity, the deal implies a valuation of around RMB 295 billion. Yet the company’s prospectus estimated that its 2026H1 profit would reach between RMB 66 billion and RMB 75 billion. If we take the median, annualize it, and apply a multiple of only 6-8 times, we would place the company’s value somewhere around RMB 800 billion to RMB 1200 billion, which would still be only ~1/7 the size of SK Hynix or Micron.

The A-share market may go much further. CXMT would give domestic investors their only listed large-scale Chinese pure-play DRAM manufacturer, a strategic semiconductor asset, and a scarce way to gain exposure to the memory boom. The Financial Times reported analyst estimates of a post-listing valuation as high as RMB 3 trillion. That would put CXMT among the largest A-share companies, if not the largest.

The timing could hardly be better. AI has driven demand for high-bandwidth memory and server DRAM, while the established manufacturers have shifted more capacity toward those higher-margin products. That has tightened the supply of conventional DRAM used in PCs, phones, and ordinary servers, exactly where CXMT is strongest. Tencent has reportedly signed a multiyear supply agreement worth nearly US$3 billion, giving CXMT something a commodity-market entrant rarely enjoys: a large domestic customer and several years of demand visibility.

In a previous Baiguan article about CXMT, we argued that the company was not yet a black swan for Samsung, SK Hynix, and Micron. It was a gray rhino. Its technological gaps in HBM and advanced server memory remain real, but its approach has been visible for years and is now too large to ignore.

This article is about the city behind that gray rhino.

A government that invests like a venture capitalist

The “Hefei model” is often described as one unique form of China’s industrial policy. Its most distinctive feature was for the government to act as an equity investor.

The playbook works roughly like this. The municipal government identifies an industry with strategic value and finds an anchor company at a critical point in its development. City and development-zone entities take a large equity stake. The government then coordinates land, infrastructure, bank financing, and supplier recruitment around the anchor. If the company succeeds, the city gains a new industrial cluster while its investment vehicles gain valuable shares. Financial returns, experienced personnel, and political credibility can then be recycled into the next project.

The three canonical cases are BOE, CXMT, and NIO, although each represents a different version of the model.

BOE already existed when Hefei made its first large bet on the display-panel maker in 2008. The city helped finance and attract a sixth-generation LCD production line at a time when BOE needed huge amounts of capital, and China remained dependent on imported panels. The bet eventually made Hefei a center of China’s display industry.

CXMT was closer to a company that Hefei helped create. When the DRAM venture was launched in 2016, Hefei-backed capital reportedly provided RMB 14.4 billion of the first phase’s RMB 18 billion funding. The city supplied capital and facilities, then helped build a cluster of packaging, testing, gas, materials, and equipment suppliers around the fab. The technology and management did not come from the municipal government. They came from Zhu Yiming and GigaDevice, former Qimonda engineers, acquired intellectual property, and international DRAM expertise. Hefei’s contribution was to assemble these ingredients and give them time.

NIO was a rescue-and-localization investment. In 2020, when the electric-vehicle maker was close to running out of cash, Hefei-linked strategic investors agreed to put RMB 7 billion into NIO China. NIO moved its China headquarters to Hefei and placed its core China assets in the local entity. The deal saved the company at a critical moment and strengthened Hefei’s EV cluster. Whether NIO will become a durable, profitable company remains unsettled, so it is better treated as a successful rescue than as a finished industrial triumph.

SemiAnalysis calls the CXMT story “The Patience of State-Venture Capital” (emphasis our own):

Hefei’s state-venture capital could afford to lose money for a very long time. Unlike a private venture-capital fund answerable to LPs that expect a return on a fixed timetable, Hefei’s state-venture capital, ultimately backed by the city’s municipal and development-zone state entities, faced no such clock. They kept funding a company that, even after turning its first annual profit in 2025, still carried an accumulated deficit of roughly RMB 36.65 billion built up over nearly a decade.

This view, however, is only partially right. In fact, in China, Hefei is the exception, rather than the rule. Understanding this question well will go a long way to explain why the “Hefei Model” has not been able to inspire more followers today.

The balance sheet has no clock, but the bureaucrats have

The idea that the Chinese state capital is naturally patient often mixes up two different levels of government.

Yes, Beijing can maintain a strategic priority for a very long time. But that does not mean every province, city, development zone, or official operates with the same horizon.

The fundamental reason is that while China’s top officials often enjoy long tenures, its local officials do not.

A Chinese city leader normally has only a few years to make an impression in a particular post. Although not a statutory term, three to four years is a useful rule of thumb for the tenure of local leaders. One study of municipal political turnover found that around 72% of mayors and 80% of city party secretaries served for less than four years. Other research has shown that individual local leaders differ significantly in their effect on growth, and that those differences influence their promotion prospects.

Three years is enough time to build a road, launch an industrial park, or hold a factory-opening ceremony, but not enough to know whether a new DRAM manufacturer can become globally competitive.

Wanda, the largest operator of China’s commercial property, understood this political clock unusually well. A government-hosted profile of Wang Jianlin, Wanda’s founder, once explained part of Wanda Plaza’s appeal to local officials in a single sentence: a mall could open in 18 months, allowing local officials to see their “own” political achievement during their tenure. Under Wang, a former military man, construction began quickly. Investment and employment could be counted. A huge physical structure appeared. The opening ceremony usually took place while the officials who approved the project were still in office.

CXMT was the opposite kind of project. Its early progress was buried inside process recipes, yield rates, patents, and customer qualification. It required years of capital expenditure and operating losses. Almost a decade passed between the project’s launch and its first annual profit. By then, an ordinary city leader might have rotated through several different posts.

For simple bureaucratic reasons, successors in local government often do not prioritize their predecessors’ projects, if not actively undermine them (which is also common). For high-stakes, high-risk projects like these mega ventures, such a tendency not to follow the pre-designated roadmap can only be stronger.1

And this uncertainty enters the project before the investment is even made. An official considering a ten-year semiconductor bet must ask whether a successor will continue it, whether banks will believe that government support will survive a leadership change, and whether suppliers will relocate based on promises made by the current administration. The official must also consider whether years of losses will later be interpreted as proof of recklessness, perhaps by an auditor or disciplinary investigator who never shared the original assumptions.

This risk is not abstract. In a previous article about China’s search for “patient capital”, I described a state-owned fund manager who came under investigation because one portfolio company went bankrupt even though the fund as a whole had performed well. Such inquiries are commonplace today under the high-pressure anti-corruption environment.

Founders and financiers anticipate the same risks too. If they doubt that a city’s commitment will survive the next personnel rotation, they discount the promise today. Banks lend more cautiously. Suppliers hesitate. Project managers are pushed to deliver visible results before the technology is ready.

A conventional venture fund may face a fixed ten-year life, but its managers are financially tied to the same portfolio throughout that period. A municipal investment vehicle may exist forever while its political principals change several times.

Under some circumstances, this can make local state capital behave more short-term than a private fund, despite having a much longer-lived balance sheet. “Patient state capital” describes Hefei’s eventual achievement. It does not explain how unusual that achievement was.

What if the bet had failed?

Success makes the original decision look obvious. To understand the political risk, imagine the same story with a different ending.

Suppose CXMT had consumed tens of billions of yuan but never produced competitive memory. Hefei would have been left with a highly specialized fab, expensive equipment, suppliers recruited around an anchor customer that never reached scale, losses inside municipal investment vehicles, and credit exposure at local banks. A successor administration would have had to restructure the project while answering accusations that the original investment was a prestige project or a reckless gamble.

The same decision would acquire a completely different vocabulary. Strategic vision would become political vanity. Patient capital would become a refusal to recognize losses. Supply-chain construction would become stranded industrial assets.

Wuhan Hongxin Semiconductor shows what the failure scenario looks like. It was launched in 2017 with a planned investment of RMB 128 billion and ambitions to manufacture advanced logic chips. The project stalled amid financing and management problems. The local state-assets authority took control, and in 2021 the company dismissed its workforce without ever realizing its stated manufacturing ambitions.

CXMT and Wuhan Hongxin were not equivalent projects. CXMT had a serious founder, acquired technology, experienced engineers, and a commercially grounded product strategy. The comparison is about the decision facing a local official before the outcome is known.

Hefei had good judgment. It also had the good fortune to avoid a spectacular early failure long enough for the model to become self-reinforcing.

How Hefei escaped the local political clock

Read more

Fact vs. Fiction: The Reality of China’s Hukou System in 2026

By David Fishman

Back in May 2026, China’s State Council issued a new set of guidelines dealing with how basic public services are provided to Chinese citizens. Observers immediately noted a key shift: access to basic public services like education or healthcare was now phrased as based on “where people actually live”, instead of the location of their hukou. It is rare to see a Chinese policy choice that is well received both domestically and abroad, but this was one of those moments.

My social media was quickly flooded with comments supporting the policy change and explaining why it was now such a big deal to disconnect public services from the hukou location. Strictly speaking, they were not wrong: it would have been a big deal if this had actually been new. But it wasn’t. In reality, the State Council document was mostly just formally codifying a policy reality that had already existed on the provincial or municipal levels for several years. While this was already known by people who follow this area closely, the excitement from those who thought was a brand-new development helped to reveal who hadn’t been paying very close attention to the last decade of hukou reforms.

Modern Hukou History

In its earliest form, the hukou registration system was an intrinsic feature of labor allocation in China’s planned economy. Up until the 1980s, the Chinese state assigned urban jobs based on hukou location, and a position in an official work unit was only available to registered citizens of that location. Starting in the 90s, hukou registration no longer had a direct connection to employment location. This coincided with the economic transformation that saw hundreds of millions of inland workers migrate to the rapidly industrializing coastal regions. When they arrived, however, they still faced barriers to true integration with those urban economies. The key barrier was their hukou, because their public benefits as Chinese citizens (like healthcare or children’s education) were bound to their registration location, back in their hometowns.

The narrative employed by urban officials at the time was that this was necessary to prevent a surge of domestic migration from overwhelming limited municipal resources. This thought process is straightforward, and not even particularly unique to China. Many countries employ some version of this logic at local levels, where cities prioritize the provision of public resources to their own (taxpaying) citizens. China’s implementation was especially rigid, however, and when there are large gaps in quality between the public services in different regions, such measures can serve as a barrier to social mobility. When it comes to education, for instance, requiring children from less-developed regions to attend the weaker schools in those regions damages their prospects to attend stronger colleges and have better subsequent careers. Also, migrant parents were often forced to leave their children behind in their hometown for schooling while working in more developed regions, driving the well-known “left-behind children” phenomenon prevalent in the 2000s.

It should come as no surprise that this policy was a magnet for criticism. An arrangement that was, at least in theory, focused on the management and distribution of scarce public resources was clearly also a driver of social inequality. In the minds of its critics, both domestic and abroad, this became the defining characteristic of the hukou system.

Converting Rural to Urban Hukou

An additional wrinkle to the hukou issue is that there were originally two kinds of hukou: urban and rural (this is no longer the case today). Previously, converting a rural hukou to an urban hukou was administratively difficult, but many pursued it anyway. Beyond the obvious function of designating the citizen as a resident of an urban or rural administrative unit, their most notable distinction was that rural hukou holders were exclusively eligible to enjoy land usage rights in their village land collective. This included a share of the dividends if such land was expropriated to make way for a national project like a highway. A hukou in an urban unit enjoyed no such land-related privileges but qualified the citizen to participate in urban social programs, like the urban worker pension plan (which is more generous than the equivalent rural plan) and urban public services (which were generally of higher quality than equivalents in rural areas).

In the late 90s, the Chinese government began relaxing the policies around converting hukou from rural to urban areas, with pilots in small urban towns. By 2001, this program was formalised, substantially lowering administrative barriers to transfer a rural village hukou to an urban town hukou. Proof of employment and residence was still required, and implementation was always subject to local discretion with weak/no national enforcement, but reform was underway. However, it was still difficult for a rural hukou holder to convert to an urban hukou in a large city, typically requiring meeting high thresholds, including formal employment, educational qualifications, or property ownership.

Around this time, there were also some early moves to allow migrant workers to access public services based on where they worked, rather than being strictly tied to their hukou location. These initiatives were largely implemented at municipal levels and varied widely, but they took place against a broader central-government push for gradual hukou reform.

In 2001, the topic of hukou reform also caught the attention of the then-governor of Fujian province completing a doctorate at Tsinghua University, a promising young cadre named Xi Jinping. In his dissertation, titled “A Tentative Study on China’s Rural Marketization”, Xi urged an acceleration to hukou reform, noting that the system had “created a large income gap between rural and urban residents that should be corrected in a prudent and proactive manner” with “reform…to level out social services”.1

But at that time, Mr. Xi would not be advancing to the national spotlight for at least another decade, and the kind of reforms he was describing were still far away too. For the most part, Chinese citizens consumed public services at the location of their hukou registration, and migrant workers were broadly excluded from using the public services in the coastal cities they were helping build, a situation that persisted through the 2000s and into the 2010s.

The Dawn of the Modern Reform Era

In 2014, State Council Document No. 25 (2014), Opinions of the State Council on Further Advancing the Reform of the Household Registration System launched a new phase of hukou reform. First, the urban-rural binary designation for hukou type was formally abolished in favor of a unified system for all registration purposes.2 In addition to doing away with the urban/rural designation, Document No. 25 declared small urban areas “fully open” to hukou transfer, introduced a policy of gradual relaxation for hukou transfer in medium-sized cities, and formalized rule-based eligibility criteria for large cities and megacities, where hukou conversion would be based on conditions such as stable employment, housing, and social insurance contributions.

It also began the crucial task of decoupling the provision of public services from hukou location, tying access to public services to place of residence rather than hukou status. The resident permit (居住证) became the key delivery mechanism, which was solidified with State Council Executive Order 663 in 2015. For many cities, if citizens could demonstrate a certain period of continuous residence and contributions to the local social insurance system, they could get a resident permit and access basic services locally. This replaced the old local/non-local binary with a system of threshold-based inclusion.

Hukou Transfer Relaxation

This relaxation of hukou transfer policies arrived roughly around the same time (2010-2015) as the wave of Chinese modernization and development finally began extending to cities in inland regions that had previously been mocked as undesirable backwaters. I do not think this timing was a coincidence. Opening the figurative floodgates to domestic cross-migration runs less of a risk of straining limited urban resources across a small handful of megacities if broad growth has succeeded in creating new, attractive destinations elsewhere across the country. Shanghai faces less pressure from Anhui migration if cities like Hefei or Wuhu already offer viable, lower-cost urban alternatives.3

Following the State Council’s 2014 Opinions document, the old barriers to hukou conversion started coming down through the rest of the decade. Many small and medium-sized cities immediately opened up unrestricted conversion to a local hukou, while larger cities established transparent entry rules or points-based systems for conversion eligibility. By the end of 2014, every city in Sichuan and Guizhou except for their capitals (Chengdu and Guiyang, respectively) had already effectively opened local hukou conversion. Over the subsequent years, many more cities quietly removed remaining requirements, such as social insurance thresholds or residence duration. By the late 2010s, conversion in much of the country’s smaller cities was already largely a procedural affair. The first large city of note to announce an explicitly “zero-threshold” （零门槛）approach was probably Shijiazhuang (capital of Hebei) circa 2019, which made it the earliest provincial capital to remove all meaningful barriers. Other provincial capitals, including Nanchang, Kunming, Jinan, and Fuzhou, followed rapidly in the 2019–2021 timeframe.

By the end of 2024, a study based on data from 332 cities found that the average “hukou conversion threshold” across the country had fallen to just 12.6%, implying that roughly 85-90% of migrant workers qualified for conversion to a local hukou according to the requirements of their host city. The same study showed that more than 90% of the cities surveyed were now classified as “low barrier” or “no barrier” cities, with just nine cities nationwide still maintaining relatively high thresholds. The nine “high barrier” cities saw hukou thresholds >50%, with four of them having very high thresholds of >85%.4

Despite these reduced barriers, there remains a substantial gap between the share of the Chinese population living in an urban location (i.e., the urbanization rate) which in 2024 was ~67% in according to the National Bureau of Statistics, and the share of the population with a hukou registration in an urban location, which was just ~50% according to the Ministry of Public Security. This gap amounts to 17% of the population, or nearly 240 million people. Part of the gap can be explained by migrant workers in those 9 cities that still don’t qualify, or who live in other cities but don’t yet meet the requirements, but they can’t account for all of it.5 So why are there seemingly so many Chinese citizens choosing to work in an urban location for which they appear to qualify for hukou conversion, but not pursuing it?

Subscribe now

Cross-Provincial Education and Healthcare Access

Part of the answer to that question can be found in the policy changes to how public services are accessed in China. In addition to hukou conversion, the last decade has seen major reforms to how citizens access public services like schooling and healthcare. In 2016, State Council Document No. 40 began requiring migrant children’s compulsory education be handled primarily by destination governments, financed through local planning and budgets, and operationalised through residence permits rather than hukou conversion. By 2020, the Ministry of Education said that there were 14.3 million accompanying children of migrant workers in compulsory education in that year, and 86% were enrolled in public schools or government-purchased seats.

Barriers still exist in the largest cities, however, and enrolment access policies are particularly uneven in downtown areas of the largest cities, which typically also have the most competitive schools. Some recent studies have also noted that even when municipal policies are explicit, implementation at the district level may still lag. So, while access to education is now guaranteed in principle for the children of migrant workers, without owning a local property in their target school district, they may be forced to enrol in a less desirable school in the suburbs. At this point though, it’s no longer a uniquely hukou issue, as such scenarios are broadly comparable to school enrolment policies in other countries like the U.S.

Healthcare has followed a trajectory similar to education, but through different institutional channels. China has various public health insurance schemes for both urban and rural residents, but access to treatment, much like education, was originally tied to the place of hukou registration. Using healthcare services in another location would necessitate out-of-pocket payment and a tedious administrative process to secure reimbursement in the home province. Additionally, while inpatient services were almost always covered, certain types of medical treatments (e.g., for chronic diseases) were basically ineligible for cross-province coverage.

However, as early as the major reforms of 2009, China began progressively dismantling the healthcare access barriers faced by migrants and enhancing the portability of medical insurance across different regions. Most critically, this included a cross-provincial direct settlement system, which was piloted through the middle of the 2010s, in parallel with the rollout of the hukou conversion and education reforms. By 2021, this underlying system was largely in place, but usage rates were still weak, with even the most mature treatment category – inpatient treatment – using direct settlement less than 60% of the time, according to the public WeChat account of China Health Insurance Magazine.

2021 also appears to have been a tipping point, after which usage of the direct settlement system increased rapidly. By 2024, official statistics from China’s National Healthcare Security Administration showed cross‑provincial direct settlement systems in that year handled over 238 million healthcare cases, rising almost 85% year-over-year. Their data showed cross-provincial settlements reduced out-of-pocket payments by nearly RMB 200 billion and were now operating across a nationwide network of more than 640,000 medical institutions. Furthermore, because insurance enrolment itself had already shifted in most places from hukou‑based eligibility to residence‑based participation, migrants were generally able to access care and settle costs at the point of treatment at a level never enjoyed before.

If Access to Public Services is Guaranteed, What’s the Point of Converting?

This is a fair question. After all, if your Anhui village hukou is no longer a barrier to receiving direct settlement of healthcare services in Shanghai, and your kids can attend Shanghai schools, what’s the benefit of switching to a Shanghai hukou? Switching into a better pension plan, but at the expense of losing your land? Once you deregister your home village hukou, you’re also at risk for your household to be unenrolled from the village land collective registry, sacrificing your precious (and unique!) land usage rights.

I’ve increasingly heard this in interviews with migrant workers in places like Shanghai, Hangzhou, or Wuxi, finding out they have no interest in converting to a local hukou – prioritizing the retirement optionality and the residual incomes that accrue from holding onto their rural land over the increasingly meager incremental benefits of changing to a local hukou.

A Shanghai citizen might even be tempted to ask whether they’re getting the short end of the stick these days. After all, the migrants who still hold hukou in their hometown while working in the city get to have their cake and eat it too, enjoying the benefits of collective land-usage rights AND access to top-quality urban public services. Urban residents are, of course, welcome to convert their hukou to a rural location somewhere, but this won’t grant them collective land-use rights.

With Hukou Reforming, What Is Left to Criticize?

In the process of these reforms, a familiar line of criticism has lost much of its bite, though not necessarily its audience. The hukou system has long served as a go-to example for commentators seeking to illustrate China’s structural inequities and authoritarian policymaking, often characterized as a rigid mechanism that traps migrants outside the benefits of urban life.

That framing is now increasingly out of date. As barriers to service access have weakened and migration has become far more flexible, the reality on the ground has grown more complex, and less supportive to simple narratives of exclusion and state control. Unfortunately, some commentary continues to lean on this older, caricaturized depiction, or even seeks out rationales for why the reforms themselves are actually signals of enhanced state control.

For instance, this 2022 paper published in China Quarterly suggested the reforms to the hukou “serve as a political contrivance for the survival of the Chinese party-state regime”, and proposed that the resulting system under the new reforms would be “an even stronger barrier”. Such commentary rarely reflects the reality of how the reforms have actually been implemented, and the persistence of the critiques says as much about the incentives of the commentators as it does about the system itself.

Where Do We Go Now?

There’s still a good distance between where we are now, and the “complete abolishment” of the hukou system (and that’s not to say that’s necessarily where Beijing is intending to go). China’s largest cities are still not fully convertible, and it’s not clear that they ever will be. Meanwhile, the readouts from the NHSA stop short of claiming “complete resolution” of access issues for migrant workers, implying there are still areas of uneven implementation. While concerns over some public resources are easing due to shrinking demand (children’s education, for example), other types of public resources are only going to be under more strain into the future (healthcare and eldercare) and cities will surely want to keep those most scarce resources available for local citizens.

In other words, what remains of the hukou system is less about restricting movement, and more about managing fiscal obligations. The most important question has evolved from “can migrants come and live here?” to “who pays for the long-term costs of social services?” That is a much more familiar question internationally. Whether you look at school district zoning in the U.S., residency requirements for social benefits in Europe, or local government financing constraints almost anywhere, the underlying issue is the same: public services are funded locally, and access for non-locals always entails some friction.

China is converging toward that global norm, albeit from a very different starting point and through a more administratively led system. The resident permit policies, points systems in top-tier cities, and the continued existence of local hukou thresholds are all techniques to help smooth that transition instead of eliminating it outright.

The more interesting question going forward is not whether hukou disappears, but whether the remaining distinctions between hukou in different regions, especially when it comes to pensions, fiscal transfers, and collective land use rights, can be resolved into a system understood as a set of trade-offs, instead of a barrier. It’s clear that the marginal benefits of converting continue to shrink in most of the country. As more migrants choose to retain their original registration location, the system will evolve too, into something that looks less like a rigid hierarchy, and more like a menu of different lifestyle choices. Whether external commentary keeps pace with that shift remains to be seen.

Two years later, the bet looks less like Silicon Valley optimism and more like a prophecy already being fulfilled. Matthew Gallagher (and his brother) launched Medvi, a telehealth startup, from his Los Angeles home with $20,000 and no employees, using over a dozen AI tools. In its first full year, Medvi posted $401 million in sales and a 16.2% net profit margin [*]— nearly three times the margin of Hims & Hers, which ran the same business with 2,400 people.

The one-person company (OPC) — a single founder wielding AI as a force multiplier — has gone from thought experiment to a real entrepreneurship path. And now, this concept that first originated in the US has landed with significant cultural and economic force in China.

7.3 million new one-person companies in China

By the end of 2025, nearly 7.32 million OPCs were newly registered in China, surging 42.3% year on year [*]. By mid-2025, China had registered 16 million one-person limited liability companies (一人有限责任公司), representing a quarter of all newly registered market entities nationwide [*].

With numbers like these, local governments began racing to compete for solo founder talent and build dedicated communities.

Beijing made Zhongguancun Science Park the centerpiece of its OPC push, anchoring the initiative around a ¥2 million innovation subsidy for qualifying solo founders, alongside reduced registered capital requirements and priority access to incubator space. Shanghai’s Lingang New Area went further, launching what it calls the “Super Individual 288” policy — a comprehensive package of 288 specific support measures covering tax rebates, co-working allowances, and visa fast-tracks for foreign founders.

Across virtually all cities with OPC-specific policies, AI infrastructure is the most consistently subsidized element. Chongqing, Qingdao, and Shanghai’s Pudong district have all introduced “computing vouchers,“ with Pudong offering to cover computing costs of up to 300,000 yuan ($44,000) for OPCs [*]. Shenzhen released an action plan to build more than 10 OPC communities of over 10,000 square meters each by 2027, pledging to cultivate more than 1,000 high-growth AI startups in the process [*].

This is not a first-tier-city-only phenomenon. In November 2025, Suzhou — a manufacturing and tech hub in the Yangtze River Delta — announced it was reinventing itself as a destination for AI solopreneurs, pledging to build 30 OPC communities and cultivate 1,000 one-person enterprises by 2028 [*].

Hangzhou has perhaps been the most aggressive of all, having announced an OPC-specific subsidy totaling ¥1 billion. Zhejiang province also introduced a personal bankruptcy insurance product specifically for OPC founders — a genuinely novel instrument designed to address one of the structural fears that have historically kept Chinese entrepreneurs cautious.

By April 2026, more than 20 cities had issued OPC-specific policies.

Subscribe now

What are they actually building?

According to the landmark 2026 OPC Insight Report released by Honghub (China’s first dedicated OPC accelerator in Hangzhou), the developer-dominated era is over. 75% of solo founders in China now come from non-technical backgrounds (marketing, product, or traditional industries). AI code generation has completely flipped the script: for every $1 spent on AI tooling, the equivalent human developer labor now costs roughly $72.

From my observation, many solo founders and indie developers are incredibly pragmatic. Most aren’t pitching moonshot ideas to VCs or chasing the “one-person unicorn” narrative (partly because China’s primary venture capital market is far tighter and less speculative than its US counterpart.) Instead, they focus on low-profile, sometimes even “boring” businesses that nevertheless generate good cash flow.

For instance, a massive wave of OPCs has emerged in the AI-generated short drama sector, where solopreneurs use LLMs for scriptwriting and AI video tools to churn out highly commercialized content for global and domestic audiences. Content creation, boutique consulting, online education, and cross-border e-commerce marketing have become the standard playbooks for Chinese OPCs.

What’s unique about China’s OPCs

But there are also more ambitious projects. “AI + Supply Chain” OPCs are emerging as China’s true competitive differentiator. By pairing AI-driven productivity with a physical proximity to manufacturing that cannot be replicated elsewhere, Chinese founders are launching and iterating products at unprecedented speeds—and many are targeting overseas markets from Day 1.

Take Zhang Lei, a recent Shenzhen University graduate based in Shenzhen’s Bao’an district. Working entirely alone, he used AI to design and build an AI smartwatch. He went from initial concept to a physical, working prototype in just 45 days, secured three investment letters of intent within three months, and cut traditional startup costs by 80%. The Bao’an Dagongfang AI Hardware OPC Hub where Zhang is based, has become the poster child for Shenzhen’s new “one-hour AI hardware production circle.“:

Shenzhen’s ‘OPC communities’ have formed a collaborative network where ‘the next floor down is the next step in the supply chain.’ An AI hardware entrepreneur can easily find an e-commerce company within the same building to help sell their product, while an AIGC creator can partner with a resident content IP studio for marketing. This ‘small yet dense’ ecosystem allows one-person companies to instantly plug into resources and rapidly scale from 1 to 100. [sohu]

The new Huaqiangbei: from copycats to 0-1 innovation

For decades, Shenzhen’s famous Huaqiangbei electronics market was known as a massive wholesale hub for cheap components and unbranded copycats. Historically, it was globally notorious for its hyper-efficient shanzhai (copycat) culture, where entrepreneurs could tear down a brand-new iPhone and have a functional, unbranded knockoff selling on the streets within a week. But today, Huaqiangbei is betting on the OPC movement and AI to revive.

In Shenzhen, OPC founders are launching highly innovative products, ranging from automated medical record systems that have been adopted by major hospitals to popular consumer products like ‘AI + pet’ solutions. Armed with the productivity boost of AI, solo founders can now thrive in small, vertical niches that often require highly customized solutions—markets that large enterprises simply don’t have the agility or capacity to drill into.

In a way, it functions much like the classic e-commerce ‘drop-shipping’ model, but with a massive upgrade: instead of just trading existing goods, solopreneurs can now drive genuine 0-1 product innovation with a small team and a laptop.

By leveraging the industrial proximity, integrated design support, and ultra-fast manufacturing networks perfected over the last thirty years, today’s solo innovators aren’t churning out knockoffs. They are using the exact same ecosystem that once built the world’s electronics to turn highly sophisticated, original concepts into mass-produced global products at lightning speed.

Share

The OPCs in rural China

China’s story is, characteristically, more layered. It’s not just builders with laptops and API keys in the tech hub like Shenzhen or Silicon Valley. It’s also recently laid-off engineers from ByteDance and Alibaba who have nowhere else to go. It’s a 28-year-old in Yiwu who figured out that AI can write product listings in 36 languages. It’s a woman in Beijing whose editorial consulting practice runs entirely from her phone. And it’s increasingly happening far outside first-tier cities; it’s showing up in villages too.

Li Xu, a 27-year-old social media worker from Hefei, visited Laotian Village at the foot of Anhui’s Jiuhua Mountain on vacation in 2025 and never left. He set up “Fucun Laotian,” a rural livestream operation filming local culture, food, and scenery for urban audiences seeking authentic countryside experiences. The village has since attracted more than 20 digital nomads like him and incubated over 10 startup projects — many of them are OPCs.

The productivity boost gave rural China an alternative, and young founders a way to afford their passion with significantly lower costs compared to high-pressure hubs like Beijing or Shanghai.

Building a company, or just surviving?

But let’s talk about the elephant in the room. Anyone following China’s economy over the past few years has likely heard about the bruising labor market, especially for young college graduates.

According to official data, more than 200 million people in China are now engaged in what the government calls “flexible employment.” That means roughly one in four workers is part of the gig economy. This raises an uncomfortable question: Is the surge in Chinese OPCs a genuine entrepreneurial revolution, or is it just unemployment in disguise?

Walk through Beijing’s commercial districts on a Tuesday morning, and something will strike you: the cafés are packed. Not with retirees or tourists, but with people in their 20s and 30s—laptops open, headphones on, intensely focused.

For the young Chinese freelancer or solo founder, the café-as-office has become normalized in a way that would have seemed alien even a decade ago. But this is the visible face of China’s new freelance and digital nomad culture, a cohort that doesn’t neatly fit into traditional categories of “employed” or “unemployed.”

Some are true OPC founders. Some are freelancers serving domestic or international clients. Some are building side projects alongside part-time gigs. And some, frankly, are just trying to figure it out. I cannot help but wonder: how many of these people are genuinely building sustainable, independent careers, and how many are simply struggling to stay afloat? Starting a profitable business is inherently risky and not a path that fits everyone’s skill set.

The OPC wave in China feels both new and familiar.

On Chinese social media, netizens often cut through the hype with blunt sarcasm: “Stop using fancy acronyms. OPC founders are just 个体户 Getihu.”

The term Getihu has no perfect English translation. “Individual business owner” is functional, but it misses the historical weight. In China’s cultural memory, the Getihu were the street vendors who first dared to sell merchandise in the late 1980s as Deng Xiaoping began loosening the centrally planned economy. They were the earliest entrepreneurs of the Reform Era, symbols of opening up, selling food, clothing, and household goods from makeshift stalls.

Yet, for a long time, calling someone a Getihu was far from a compliment. It carried a certain distinct social stigma. To be a Getihu implied that you were doing unstable, low-status manual work because you couldn’t land a “real” job—meaning a secure position in a state-owned enterprise, a government office, or a big multinational corporation. Decades later, the freelancers and OPC founders sitting in cafés on workday morning face a remarkably similar societal skepticism.

Is China’s OPC wave a genuine entrepreneurial revolution, or an elaborate coping mechanism for a labor market under stress?

The long shadow of the small business founder

Subscribe now

Read more

The original article was written in Chinese by Elsewhere别处发生, a writer and podcaster who typically reports on China’s venture capital and startups. It’s a partial translation that leaves out some details unrelated to DeepSeek.

The translation below is not authorized by the original author. Emphasis is our own.

The “four-hour meeting”

First, let’s talk about that investor meeting—the one whispered about as the “four-hour meeting.”

In mid-May, DeepSeek hosted this virtual meeting via Tencent Meeting. By then, the list of investors was mostly settled. Each firm was allowed to send two attendees to the meeting. The session began with a presentation by Liang Wenfeng, followed by a Q&A where he addressed the investors’ inquiries.

Legend has it that one CVC (Corporate Venture Capital) investor spent a full ten minutes just asking questions—starting with a lengthy self-introduction followed by three drawn-out questions. The atmosphere in the virtual room turned awkward for a moment. It’s easy to imagine the scene, with some participants whispering to one another, wondering who this person was.

Liang, however, seemed unfazed and continued to answer every question earnestly. At one point, after finishing the second question, he forgot what the third one was and had the investor repeat it.

For many of those present, even those who had already committed capital, this was their actual first interaction with Liang Wenfeng, as most had never met him in person. By the end of the call, many were left with a lasting impression, as if his words were still echoing in their minds.

Liang Wenfeng’s aphorisms and spirit began to circulate throughout the market.

Restraint and AGI

Liang Wenfeng is not a naturally gifted orator, and he speaks at a measured pace. Yet, several quotes from that meeting are frequently repeated (paraphrased, as the exact wording may vary slightly):

• “We have no fame and no influence; we are just a group of very ordinary people.”

• “The narrative I like is a group of ordinary people achieving something extraordinary. For DeepSeek, our core focus is ensuring team stability, which is more important than money or resources. It is the greatest risk and the greatest challenge.”

• “Assuming AGI will account for 20% of GDP, those who want 5% will lose to those who want 1%; those who want 1% will lose to those who want 0.1%.”

• “What we want to do is anything related to improving intelligence; we will not do anything else.”

• “AGI is a big enough endeavor. Everything else is just a process.”

Throughout those four hours, Liang consistently emphasized the same themes: the singular pursuit of AGI, the philosophy that “less is more,” and the importance of restraint.

From 5 Billion to 1.5 Billion

Official fundraising began around April of this year. It is said there was a “whitelisting” of funds—a list consisting of institutions with significant capital capacity and strong brand reputations.

Initially, DeepSeek required a minimum investment of 5 billion RMB, prohibited the syndication of shares, and insisted on a “pure” RMB fund structure. Given this, one prominent investor who was widely rumored to be a guaranteed participant ultimately did not join. Industry rumors suggest this was because they had tried to sell portions of their stake to Middle Eastern LPs.

However, in China, there are very few market-oriented funds capable of committing 3 billion RMB in a single shot. Consequently, many VCs negotiated to see if they could participate through multiple fund entities.

Ultimately, the minimum investment threshold was lowered to 1.5 billion RMB, and the fundraising structure became more flexible.

There was also a claim that participating institutions were preferred if they had not invested in other LLM companies. This was, in all likelihood, an impossibility for a simple reason: finding a top-tier fund that hasn’t invested in any LLM companies is a contradiction in terms.

During the investor meeting, Liang Wenfeng also expressed that DeepSeek is willing to help other companies, including other large-model developers. The caveat, of course, remains: as long as they don’t try to poach his talent.

[Some parts about participating venture capital funds are redacted. You read about them in the original article.]

The most important demand: don’t poach DeepSeek’s people.

I won’t go into DeepSeek’s investment terms here. I’ll add only one point: whether dealing with big tech companies or VC funds, Liang Wenfeng’s most important demand was this: do not poach DeepSeek’s people, or encourage them to leave and start their own companies.

“China’s largest listed company“

If the issue is reduced to the level of a single investment, what kind of company is DeepSeek, in terms of odds of success?

One investor who participated in the deal once worried, perhaps unnecessarily, whether this might be an investment with too much consensus behind it. According to the experience of the investment industry, consensus investments usually do not end well.

But with DeepSeek, it is hard to say that. Or rather, we cannot understand it through a simple investment lens, whether in terms of its significance for China’s AGI industry or on a more spiritual level.

As one investor put it: “They are doing all of this with enormous goodwill. He seems to be telling us: the world was always supposed to be this way.”

So some hesitated, while others believed there was “no reason at all not to invest.” If, as some expect, DeepSeek eventually goes public, could it become the largest listed company in China’s A-share market?

Low profile and ambition

In following DeepSeek’s fundraising progress, one feeling stood out: for most people, DeepSeek was probably an utterly unexpected answer. Even the investors who most love to win had not imagined that one day it would open itself up to financing.

So, as mentioned earlier, many of the well-known investment heavyweights only recently managed to get in touch with the company. Of course, Liang Wenfeng had long avoided meeting investors. After news of the fundraising emerged, many institutions did not even try to compete for an investment opportunity.

But judging from the outcome, some secured a ticket through so-called synergy, some through persuasion, some through branding, and some simply through persistence.

Elsewhere once tried asking several investors about the deal. Most did not want to say much. The sensitivity of the transaction was only one reason. More than that, there seemed to be a kind of self-conscious restraint: because they felt honored to be part of it, they wanted to preserve its low-key nature.

One participating investor said that throughout the transaction process, he kept every document carefully, hoping that one day he might have the chance to frame them.

Another investor repeated twice the 16-character line released with the preview version of DeepSeek V4:

“Unmoved by praise, unafraid of slander, follow the Way, and uprightly correct oneself. 不诱于誉，不恐于诽，率道而行，端然正己。”

Subscribe now

Related Reads

If you also wish to share your writings with the 12,000+ Baiguan subscribers, please send us an email at baiguan@bigonelab.com

Is CXMT’s IPO a Black Swan Moment for the DRAM Giants?

The last few months saw a historic rally across the global memory-chip industry. Each of the three largest DRAM producers — Samsung, Micron Technology, and SK hynix (the “Big Three”), surpassed $1 trillion in market value, one after another.

Since 2025, the AI boom has driven a surge in demand for high-performance computing and storage equipment. As a result, global memory-chip companies have seen explosive growth in both earnings and share prices, becoming major beneficiaries of the AI rally.

It’s within this context that Chinese memory-chip maker CXMT comes under the spotlight of the capital market. On May 27, its STAR Market IPO application was officially approved by the Shanghai Stock Exchange. The company plans to raise around 29.5 billion yuan ($4.2 billion), making it potentially the largest IPO in China since 2022. Once listed, it could be valued at more than 2 trillion yuan.

Founded in 2016 and headquartered in Hefei, Anhui Province, CXMT is currently the only Chinese DRAM company to have achieved large-scale mass production. In 2025, it ranked fourth globally, with a market share of 7.67%.

2024-2025 Output of DRAM Memory Chips (thousand wafers per month) Source: TrendForce

DRAM is one of the most concentrated segments in the semiconductor industry and has long been dominated by the Big Three. For latecomers, huge capital expenditure is only the starting point. They also need years of accumulated expertise in process technology, yield improvement, patents, equipment, customer certification, and supply-chain management. The fact that CXMT has joined the big league is itself a major breakthrough for China’s semiconductor industry.

As of 2026, CXMT has completed the mass-production transition from its first- to fourth-generation process platforms, with products covering DDR4, DDR5, LPDDR4X, and LPDDR5/5X. The company has also built 12-inch DRAM wafer fabs in Hefei and Beijing. Its monthly capacity is expected to continue rising, further expanding CXMT’s presence in the global DRAM market.

Here is an overall timeline for CXMT:

• 2016: The CXMT was launched.

• 2017: Construction of the first-phase fab began.

• 2019: CXMT began production of 8Gb DDR4 products using a 19nm process, marking the first time a Chinese company achieved mass-production capability for mainstream DRAM products.

• 2021: The company made further progress with its 17nm process.

• 2023: CXMT launched China’s first domestically developed LPDDR5 product and entered the smartphone supply chains.

• 2024: CXMT achieved mass production of DDR5.

• After 2025: The company continued to advance DDR5, LPDDR5X, and selected server-side products, while HBM-related R&D also began.

The Hefei Model

Behind CXMT’s rise is a broader story: “Hefei Model.” Named after the city that nurtures display maker BOE, electric vehicle NIO, and CXMT, the Hefei Model is a Chinese-style approach to industrial policy. In effect, it functions as a form of state-guided venture capital: local authorities identify industries with long-term strategic value, take equity stakes to fund promising companies, and later sell their shares for massive profits once the companies scale.

In recent years, Hefei has been dubbed “China’s best venture-capital city.” Its state-owned capital does not simply hand out subsidies. Instead, it participates deeply in the development of strategic emerging industries through industrial funds, equity investment, land resources, industrial coordination, and long-term policy support. This approach has helped a once relatively poor provincial capital climb sharply in China’s urban income rankings.

According to CXMT’s prospectus, the company is backed by a broad base of long-term state and private capital, including the second phase of the National Integrated Circuit Industry Investment Fund (Big Fund II), state-owned capital from Anhui Province and Hefei, Alibaba Cloud, financial asset investment companies under state-owned banks, and insurance funds. These investment capitals are:

• patient. DRAM is a capital-intensive, cyclical industry that often requires years of losses before scale and competitiveness emerge. Without long-term capital support, it would be difficult for any company to survive the long process of R&D, capacity ramp-up, and industry downcycles.

• closely aligned with industrial policy. China wants to improve the self-reliance of its semiconductor supply chain, and DRAM is one of the few critical segments still dominated by global oligopolies. CXMT, therefore, has strategic significance.

• helping create an industrial cluster. Memory-chip production is not just about a single fab. It requires wafer manufacturing, equipment and materials, packaging, and testing. Through local state capital and industrial policy, Hefei has embedded CXMT within a broader ecosystem.

This is what makes CXMT different from many other semiconductor startups. Its growth is not driven purely by venture-capital logic. It is a long-term industrial project jointly supported by local governments, national industrial funds, financial institutions, and strategic clients. From this perspective, CXMT’s IPO is not the finish line, but a new capital base for the next phase of capacity expansion and technological catch-up.

Strong Market Demand

That capital base matters because CXMT is attracting attention not because it is simply another “domestic alternative” in China, but because it addresses real market demand. In the consumer DRAM market, its 1a-node DDR5 and LPDDR5X products are beginning to serve a broad range of PC, smartphone, and consumer electronics customers. While CXMT’s process technology still trails the Big Three, its products are already commercially viable for mainstream consumer applications.

At home, CXMT’s memory chips have already been adopted by domestic customers. Its DDR5 products are being supplied to PC makers such as Lenovo, while its server-side products are undergoing validation with internet firms including Tencent and ByteDance. Xiaomi’s use of CXMT’s LPDDR5X chips in high-end smartphones also suggests that domestically produced mobile DRAM is moving beyond low- and mid-range devices and into premium handset supply chains.

This shift toward domestic substitution is not driven solely by government policy or supply-chain security concerns. Instead, it reflects a genuine supply-demand imbalance. As the Big Three allocate more capacity to high-margin HBM (high bandwidth memory) and server DRAM, the supply of consumer DRAM used in PCs and smartphones has tightened. CXMT is well-positioned to fill that gap.

That also helps explain why CXMT’s appearance in global consumer-electronics supply chains. Recently, Corsair’s Vengeance DDR5 16GB memory modules have used DRAM chips from CXMT rather than from the Big Three. This is not an isolated case: HP reportedly placed a large LPDDR5 order with CXMT in January; Qualcomm began working with the company on customized DRAM in April; and Dell, ASUS, and Acer have also reportedly been in contact with CXMT.

When the Big Three are constrained by tight capacity, elevated prices, and longer delivery times, DDR5 and LPDDR5X chips from Chinese suppliers become a practical alternative. In other words, CXMT’s impact on global DRAM market is likely to be felt first in consumer-grade DDR5, LPDDR5X, and other standardized DRAM products.

Is CXMT ready to challenge the Big Three Memory Makers?

Read more

Mr. Yao is an expert on credit in China, and I can’t possibly think of anyone better qualified to opine on this topic.

If you want to get in touch with Mr. Yao, and want to know more about his company’s services, please send us an email at baiguan@bigonelab.com

Why China’s April Credit Data Looks Terrifying (And Why It Shouldn't)

By YAO Yu

To a foreign observer accustomed to two decades of China’s relentless, credit-fueled economic miracle, the April 2026 financial data released by the central bank looks like a red flag sounding a severe economic contraction.

The headline number that shocked analysts the most was the outright contraction in bank lending: new Renminbi (RMB) loans actually fell into negative territory at -10 billion RMB. This is a jarring anomaly, indicating not only a slowdown in growth but also that more money was repaid to banks than was lent to the real economy.

Across the board, the metrics point downward. New Total Social Financing, China’s broadest measure of credit and liquidity, came in at only 624.5 billion RMB, which is a massive drop of over 535 billion RMB compared to the same period last year. The pain was particularly acute at the consumer level, with household loans plunging by nearly 787 billion RMB as both short-term and long-term borrowing contracted.

For international readers and investors, these data points traditionally lead to a single, terrifying conclusion: a stalling macroeconomic engine. Negative loan growth usually suggests that businesses are refusing to expand, factories are halting capital expenditure, and consumers are hoarding cash out of fear of a looming recession.

Yet, if you look across the broader macroeconomic landscape, a barrage of contradictory signals emerges that makes it difficult to find consistent evidence of a stalled economy. For instance, trade metrics have maintained high growth, with April exports surging by 14.1% year-on-year in US dollar terms and imports jumping by a remarkable 25.3%. On the industrial side, the manufacturing PMI stayed comfortably in expansionary territory with an improving production sub-index, supported by increased electricity generation and solid growth in high-frequency infrastructure production data. Furthermore, price and currency metrics tell a resilient story: CPI, PPI, and real estate prices all showed improvement in April, while the RMB continued to strengthen. These robust indicators suggest we are witnessing a complex divergence rather than a systemic breakdown.

Subscribe now

Beyond the Aggregate: A Signal of Transition, Not Just Contraction

Why does such a glaring contradiction exist?

The decline in credit isn’t merely a cyclical downturn or a collapse in confidence. It is a fundamental restructuring of how the Chinese economy finances itself. To understand what is actually happening, we have to look at the four major shifts occurring right now across personal, government, and corporate balance sheets.

1. The Rational Deleveraging of the Chinese Consumer

The sharpest decline in credit is happening at the household level, driven primarily by the ongoing housing slump. But this isn’t just about consumers hiding money under the mattress in a panic, but about cold, hard math.

Right now, consumers are facing a stark interest rate inversion. The interest rates on outstanding mortgages remain relatively sticky, while the yields on wealth management products, funds, and standard savings have dropped. If your mortgage costs you 3% but your investments are only yielding 2% or less, the most rational financial decision you can make is to pull your liquidity and pay down your debt early. Chinese consumers are actively and rationally deleveraging to optimize their personal balance sheets.

2. The Illusion of Shrinking Government Credit

On the government and corporate side, the data is being heavily distorted by a massive financial cleanup. Local governments are in the midst of a historic effort to substitute high-interest, hidden debt (often held by Local Government Financing Vehicles) with official, lower-cost sovereign bonds.

When a local government issues a new bond strictly to pay off an old, off-balance-sheet loan, the total amount of “new credit” injected into the real economy is effectively zero. In the aggregate data, this looks like a severe slowdown in credit expansion. In reality, it is a risk-mitigation exercise. The debt isn’t fueling new bridges or factories right now, but being restructured to prevent systemic instability.

3. The Corporate Migration to the Bond Market

Even when corporations are actively seeking capital, they are dramatically changing where they get it.

In the current environment of loose macroeconomic liquidity, issuing bonds has become significantly cheaper than taking out traditional bank loans.

High-quality enterprises are acutely aware of this cost advantage and are actively optimizing their debt structures.

The April 2026 data illustrates this substitution effect perfectly: while medium and long-term corporate bank loans suffered a massive contraction of 410 billion RMB (a year-on-year drop of 660 billion RMB), net financing from corporate bonds surged by 452 billion RMB.

Breaking that down further, industrial bonds jumped by 384.1 billion RMB, and LGFV bonds increased by 54.4 billion RMB. With the median coupon rate for newly issued credit bonds falling to a mere 1.9%, corporate liquidity hasn’t dried up, but it has simply migrated to the cheaper bond market, making traditional bank lending figures look artificially anemic.

4. The New Economy Runs on Equity, Not Debt

Perhaps the most crucial blind spot in traditional credit analysis is how the drivers of China’s future growth are actually funded. For the last two decades, economic growth was driven by real estate and heavy infrastructure, industries that rely heavily on massive, traditional bank loans.

Today, the growth engines are emerging industries: artificial intelligence, advanced semiconductors, and green energy. These sectors do not thrive on traditional bank debt. They rely heavily on equity financing, venture capital, and state-backed investment funds. Because traditional credit metrics were designed to measure the old economy, they are structurally blind to the capital flowing into the new economy.

The Takeaway

Using aggregate credit data to diagnose the health of the Chinese economy in 2026 increasingly looks like using a thermometer to measure wind speed. The tool is simply outdated for the current environment.

The shrinking credit figures are certainly a reflection of real estate pains, but they also represent a deliberate consumer deleveraging, a cleanup of local debt risks, a corporate migration to cheaper bonds, and a pivot toward equity-driven tech growth. The Chinese economy is transitioning, and it is time our analytical frameworks transition with it.

Related Readings

After government’s resolve to boost domestic consumption, curb over-production and a stock market bull, it’s safe to say the “whether” part of that question is already resolved. The question now mainly surrounds “when”.

of , an investment memo-style publication that I avidly read, recent made some observation about her recent trip to China (her own emphasis):

Confidence is a big part of the game, and it feels like the average person has made peace with the new normal. Coffee shops are packed. Consumer energy feels back, and people seem confident again… Consumer confidence seems meaningfully improved. Shanghai IFC Mall traffic is strong; regular cafes are packed with middle-class families and afterschool kids….A key driver is the stock market wealth effect: the Shanghai index being up sharply since the Oct-2024 lows has made “hope” feel investable again for a lot of people…The right frame is K-shaped - but the average mood feels less defensive than the last few years.

My own anecdotal observation, as someone who is based in Shanghai and travels regularly inside the country, is also the same as Freda. It’s clear the sentiment in the air feels very different from the vibe 2 years ago.

But subjective experience can be misleading too, bound by the information bubble created by my small social networks and life experience. To have a complete picture, we also need to check what data has to say.

In the following sections, I will use several KPIs review this question.

The Kentucky Fried Chicken index

One of my favorite proxies to understand deflationary pressure in China is to read YUM China’s quarterly results and find the average selling price of Kentucky Fried Chicken.

KFC is a popular fast food dining brand and the largest restaurant chain in China. It’s also one of the cheaper dining options out there. In other words, KFC is easily the least-cost producer in China’s dining sector. The simple rule of thumb is that if deflation stops in China, it will inflect first at KFC. Conversely, if deflation at KFC doesn’t stop, we definitely are not out of the woods.

In the last many quarters, same-store transaction volume has kept growing, but same-store sales grew much more slowly than volume, which means ASP has been declining, and that people are consuming more chicken wings but paying less for each.

The just-announced 25Q4 results, however, marked the first quarter in a long while that the ASP actually stabilized at KFC.

To be more precise, this comes after 7 quarters of ASP decline. The trend seems undeniable.

Subscribe now

The China Resident Risk Appetite Index (CRRAI)

To find data series more representative than chicken wings, I look to the CRRAI which is designed to quantify the financial behavior and confidence of Chinese households. It has been created by , a friend of the channel who has contributed several pieces here before.

Bob constructed CRRAI upon 4 thematic pillars, each capturing a distinct layer of the household financial decision-making process:

• Macro Liquidity & Business Confidence: This dimension serves as the systemic backdrop, gauging the availability of broad market liquidity and the overall health of the business environment through indicators such as the M1-M2 spread, government bond yields, and manufacturing and service PMIs.

• Household Leverage & Credit Dynamics: This reflects the willingness of households to take on debt. It tracks variables such as long-term and short-term loan growth, deposit term-lengthening, and non-bank financial deposits growth, usually a forerunner of household capital rotation into financial markets.

• Fundamental Bedrock: Risk appetite is constrained by asset values and income security. This dimension monitors the “wealth effect” and the stability of the household balance sheet using metrics such as secondary housing prices, property sales volumes, and employment data.

• Capital Market Sentiment: This most reactive component captures “animal spirits” and speculative momentum within secondary markets. It synthesizes signals such as margin trading ratios, valuation levels (P/E ratios), trading intensity, and new brokerage account openings.

This is what CRRAI looks like right now:

It’s clear that from this perspective, the lowest level of household confidence in the past decade was seen during 2022. Since then, we have seen a year-by-year, staircase-like recovery.

A defining feature of the late-2025 landscape is the persistent lead of Capital Market Sentiment. For 2025, this dimension turned positive as early as August and peaked in September. Although it moderated slightly toward the end of the year, it remains the only pillar comfortably above the historical mean. This sentiment front-running was driven by the recovery of the Chinese stock market, a surge in retail participation, and a sharp increase in non-bank financial deposits, which acted as a forerunner for capital rotation into financial assets.

On the other hand, the most encouraging signal in December was the rapid recovery of the Fundamental Bedrock pillar, which rose from -0.47 to -0.21. This improvement reflects a stabilizing job market and a marginal recovery in secondary housing transactions. Historically, a sustainable “risk-on” environment requires the composite index to be supported by these “real economy” pillars.

Despite these recoveries, Macroeconomic Liquidity and Household Credit Dynamics continue to weigh on the overall index. The “water level” in the system remains low, with the M1-M2 spread still indicating sluggish capital activation. Furthermore, while the pace of deleveraging has slowed, the willingness of households to expand their long-term balance sheets (mortgages) has not yet returned to growth.

But if they can continue their upward trajectory into Q1 2026, the index is well-positioned to break above the neutral threshold, marking a definitive end to the multi-year “risk-off” cycle.

Ratingdog China PMI

Ratingdog China PMI, formerly known as Caixin PMI, is the unofficial macro-economic indicator that the market is watching closely. The one sub-index that I pay the most attention to is the service PMI because the service sector is more directly related to domestic consumption rather than the manufacturing sector.

This data series suggests that China’s services sector has entered 2026 on a firmer footing, with this reading rising to 52.3, the strongest reading in three months.

Yao Yu, founder of Ratingdog commented:

Service sector business activity accelerated in January, primarily driven by faster growth in total new business.

Employment returned to expansion, with a slight increase in backlogs. Supported by sustained inflows of new business, service sector firms saw improved demand for labor, with the employment index rising into expansionary territory for the first time in six months.

On the price front, cost pressures eased somewhat. Input costs rose for the eleventh consecutive month, driven mainly by higher fuel and purchased item prices, but the reading fell to a five-month low. The decline in selling prices eased, with the reading essentially returning to the breakeven level, indicating some relief in terminal pricing pressure.

Overall, the service sector started the year on a stable footing. Looking ahead to February, consumption-oriented services such as culture and tourism, catering, and instant retail may see growth driven by the extended nine-day.

The Services PMI is expected to remain in expansionary territory. From a medium- to longterm perspective, the slight dip in business confidence suggests that external uncertainties remain a constraining factor. The sustainability of the service sector’s recovery will still depend on the strength of domestic demand improvement.”

(Ratingdog is the creator of YY Rating, which is perhaps the most objective credit rating system in the PRC at the moment. If you are interested in getting in contact with them, please shoot us a message at baiguan@bigonelab.com)

Official CPI…

So far, we haven’t even looked at official statistics yet. But we should.

Many observers have this unconscious bias that the numbers made by Beijing are not trustworthy. We may write a separate post about this topic in the future. But in a nutshell, we believe that most of the official data is quite dependable. Its credibility gap compared with other major economies has also visibly shrunk.

So what do official statistics tell? Let’s look at the consumer price index (CPI).

Clearly, CPI has hit some kind of inflection point in 2025Q4, with YoY change decidedly entering positive territory.

Core CPI, which excludes food and energy components that are susceptible to seasonal and non-economic factors, did even better. Throughout 2025, the core CPI steadily recovered, gradually rising from 0.5% yoy in March to above 1% yoy in the fourth quarter.

Researchers at CF40, our partner think-tank, did some further digging by segmenting core CPI into several main factors.

CF40 showed that there are three components that made significant contributions to core CPI growth in 2025: gold jewelry, trade-in goods, and medical services. Although these three components account for approximately 22% of core CPI, their average contribution to core CPI throughout 2025 was around 40%. The increases in these components were all driven by exogenous (international gold price hike) or policy-related factors (trade-in subsidies and continuous advancement of healthcare payment reforms).

To better look at the “core” core CPI without the influence of these exogenous factors, CF40 made an adjusted core CPI by excluding these three components. The final, adjusted core CPI yoy growth remained essentially flat at around 0.6% across all months in 2025. However, this flat trend, which came after significant growth moderation in the previous years, suggested that China’s core CPI has indeed bottomed out and stabilized.

Investment implications (for paid subscribers)

Read more

Looking back today, four of those tailwinds are clearly intact.

Geopolitically, the external pressure has continued to ease materially. Every day, Trump’s antics and the new Monroe Doctrine have even made China a more viable trading partner, and China’s bargaining position has strengthened as its industrial base continues to move up the value chain.

On policy and expectation management, the learning curve is real. Regulators are no longer operating with the blunt, surprise-driven approach that characterized the 2021–2022 period. Policy signals are now layered, iterative, and increasingly conscious of second-order market effects.

On the economic engine, the transition away from real estate is still painful, but it is no longer directionless. High value-added manufacturing, export-oriented Chinese brands, and frontier technologies are now doing the incremental work that property once did.

And most importantly, China’s pivot toward a capital-market-centric development model is no longer theoretical.

In just the past few months, a growing list of strategically important AI-related companies — from chips to large models to hard tech — have either entered the capital market or are lining up to do so. Equity financing is becoming the default channel for frontier industries. Compared with the old land-based and banking-based financing system, the capital market’s role is not only larger — it is structurally indispensable.

The remaining piece, of whether Chinese residents’ excess savings will move into the equities market, however, is far more complex.

Juggling of Objectives

Today, Chinese policymakers are juggling three objectives that are sometimes in tension:

First, they want the stock market to rise. That requires all investors to believe equities are a legitimate long-term asset class.

Second, they are counting on the market to fund strategically critical industries — AI, semiconductors, commercial space — where balance-sheet lending simply does not work.

Third, they do not want a speculative frenzy that ends badly. The explicit preference is a slow bull, not a crazy bull.

But if you want to achieve Objective 2 too dearly, you risk hurting investors’ interests, making you unable to achieve Objective 1. On the other hand, if you want Objective 1 too much, you risk attracting retail money too quickly, thus making Objective 3 of a slow bull elusive.

The difficulty of balancing these three goals became unusually clear over the past two weeks, as we are only fresh into 2026.

How the Frenzy Took Shape

From mid-2025 onward, a sector bull quietly formed in A-share technology stocks, centered on AI infrastructure.

For instance, optical module leaders such as Innolight (300308.SH) and Eoptolink (300502) rose close to tenfold in roughly half a year, reflecting their bottleneck position in AI data centers. Cambricon (688256.SH), one of China’s main challengers to Nvidia, also saw its market capitalization balloon as restrictions on advanced Nvidia chips pushed capital toward domestic alternatives.

Then came commercial space.

A string of supportive policies — most notably the easing of STAR Market IPO rules for space companies (as Amber just documented in our last piece) — coincided with renewed global attention on SpaceX and its potential IPO. Capital rapidly flooded into what began to be labeled the “commercial space sector.”

But in reality, it was barely a sector at all, as most true leaders were still private. And none of the Chinese contenders have achieved the same kind of success that SpaceX has achieved.

Yet, that didn’t stop speculation. Listed companies with only tangential exposure to satellites, launch systems, or aerospace materials were aggressively re-rated. Prices doubled, tripled, and quadrupled within weeks. As investors returned from the New Year holiday, early 2026 saw the frenzy intensify further.

Soon after, the successful listings of LLM companies such as MiniMax and Zhipu AI ignited yet another narrative: “AI applications.”

In a country that has not yet fully cracked advanced GPUs or state-of-the-art foundation models, speculative capital had already leapt ahead to applications. Once again, stocks with only remote relevance were swept up.

This could be the long-awaited moment many had hoped for: retail deposits finally beginning to migrate into the capital market — Except the migration was at the risk of happening too fast.

Last Monday, on Jan 12, total A-share turnover hit 3.6 trillion yuan, an all-time high. I also started hearing tales of massageurs chatting about the opportunity to buy commercial space stocks. An unlicensed influencer reportedly attracted tens of billions of RMB from retail investors into a newly formed fund focused on “AI applications and commercial space” in one live-streaming session.

It was also obvious how this would end if left unchecked. If anything, it would not become a “slow bull”, but would very likely end up in tears again.

The Intervention, Managed Upside, Managed Downside

And soon enough, regulators stepped in.

Several of the most speculative stocks — particularly those with the most extreme short-term gains and thin fundamentals — were ordered to suspend trading pending disclosures. Margin financing rules were tightened. That unlicensed KOL was also reportedly under investigation.

Yet, it was also clear that this is not an all-out campaign against the market. There was no sweeping crackdown, no moralized rhetoric, no attempt to crush sentiment across the board. The actions were selective, technical, and deliberately contained. They were designed to slow momentum, not reverse direction.

Beijing is attempting to short-circuit a feedback loop of leverage, narrative, and retail FOMO that can lead to collapse, while it’s still in its early stages.

This episode also needs to be seen in symmetry.

In mid-December, the market was threatening to break down. At that point, the “national team” (a catch-all phrase for market stabilization mechanism) stepped in with visible buying support, initiating a rare 13-day winning streak — one of the longest in A-share history.

Now, with sentiment tipping toward euphoria, the same apparatus appears to be operating in reverse. Market chatters also suggest that the stabilization fund” may be selling into strength this time around — a rumor that would not be surprising if true.

That is, after all, what a stabilization fund is meant to do: buy when fear dominates, sell when euphoria takes over.

If this interpretation is correct, it suggests that Chinese regulators have become far more sophisticated at managing the market — 操盘 in the literal sense.

And the market is taking cues. Many of the most red-hot speculative stocks in commercial space and so-called “AI applications” have started to tumble, and so far, there is no end in sight for where that slide will land.

For now, this crackdown on over-speculation seems to be working, without hurting the broader sentiment.

The Real Risk

And yet, one concern remains unresolved.

Read more

For context, Elon Musk’s Starlink—the undisputed leader in low-Earth orbit (LEO) today—has a long-term target of around 42,000 satellites [*].

The scale alone signals that something fundamental has changed.

Historically, China’s space industry was a closed system dominated by state-owned “national champions,” primarily China Aerospace Science and Technology Corporation (CASC) and China Aerospace Science and Industry Corporation (CASIC). This model delivered strategic reliability, but it struggled to compete with SpaceX’s Starlink in the areas that now matter most: cost efficiency, iteration speed, and control over low-Earth orbit (LEO) resources, including military-grade communications.

Faced with Starlink’s overwhelming first-mover advantage, Chinese policymakers concluded that a purely state-led model was no longer sufficient. As a result, China’s policy direction has undergone a fundamental shift—from merely allowing private companies to participate, to actively supporting private firms as the main drivers of the commercial space sector.

In late 2025, the China Securities Regulatory Commission (CSRC), together with the Shanghai Stock Exchange (SSE), revised the STAR Market (科创板, or China’s Nasdaq) listing rules. A dedicated IPO “green channel” was introduced for high-barrier, long-cycle hard-tech companies, particularly those focused on reusable rockets and space infrastructure [*].

The most important change: loss-making companies are now allowed to go public. As long as a firm demonstrates breakthroughs in core technologies—such as liquid rocket engine recovery—and aligns with national strategic priorities like China’s state-backed satellite constellations, traditional profitability requirements can be waived.

The impact was immediate.

On December 31, 2025—just days after the new rules took effect—LandSpace, one of China’s leading private rocket companies, had its STAR Market IPO application formally accepted [*]. The company is seeking to raise RMB 7.5 billion, at a reported valuation of RMB 75 billion. 2026 is shaping up to be a peak year for Pre-IPO monetization across China’s commercial space industry.

Secondary markets have already started to price this in. Beginning in December 2025, speculative enthusiasm surged, pushing several satellite and space-themed ETFs to nearly double from end-2025 levels (perhaps too much speculative betting at this point). Investor attention has shifted decisively from policy vision to execution—and to who will become China’s answer to SpaceX.

Subscribe now

Why the Rush? The Race Against the ITU and the “Iron Curtain”

The sense of urgency within China’s orbital strategy is driven by a “red zone” countdown imposed by the International Telecommunication Union (ITU). Under the ITU’s “Use it or Lose it” rules, satellite constellations must hit strict deployment milestones to retain their claimed frequencies. Once a frequency is filed, the operator has exactly seven years to launch its first satellite and begin operations (Bringing Into Use, or BIU). But the clock doesn’t stop there: by year nine, 10% of the total constellation must be in orbit; by year twelve, 50%; and the entire fleet must be operational by year fourteen [*].

For China’s largest constellation, “GW” (Guowang), which filed for approximately 13,000 satellites in September 2020, the BIU deadline is September 2027. To avoid losing these precious spectrum rights, China must transition from ambition on paper to mass industrial delivery by 2026.

The Battle for Orbital Real Estate: Why 550km is the “Park Avenue” of Space

This regulatory pressure is compounded by Starlink’s hegemony. As of early 2026, Starlink has already launched over 9,422 satellites, and an additional 7,500 satellites were approved a few days ago, with 50% to be launched by 2028. Starlink has effectively occupied the prime orbital shells at 530 km and 550 km, and is now moving to compress its orbits down to 480 km.

Low Earth Orbit (LEO) is generally categorized into three distinct layers: Ultra-Low Orbit (VLEO) below 400km, Core LEO between 400km and 600km, and High LEO above 1,000km.

The narrow bands at 530km and 550km are considered the “prime real estate” of the sky. At these specific altitudes, the physics of communication and safety reach a perfect equilibrium. These orbits are high enough to minimize atmospheric drag, allowing satellites to stay aloft for years without excessive fuel consumption, yet low enough to keep signal latency at a crisp 20–40 milliseconds—ideal for high-speed trading and gaming. Critically, these layers also benefit from “passive safety”: if a satellite fails here, the thin atmosphere will naturally pull it down to incinerate within a few years, preventing a permanent buildup of space junk[*].

Starlink has effectively “colonized” these core shells, forcing competitors to either climb higher (thus more latency) or dive deeper into the more challenging VLEO zones (difficult because intense atmospheric drag requires constant orbital maneuvers and specialized aerodynamic designs to prevent satellites from prematurely re-entering the atmosphere).

But if China does not scale up mass production and launch by 2026, it will be orbitally blocked by Starlink’s sheer physical presence.

Share

Is This Space Race 2.0?

At first glance, the echoes of the 1960s are unmistakable. Once again, two superpowers are vying for the “ultimate high ground” to secure not just national pride, but undisputed strategic dominance. In the past, whoever reached space first demonstrated the superiority of their system and their intercontinental missile capabilities. Today, whoever controls low-Earth orbit satellite networks controls a “god’s-eye view” of global communications and modern military command.

But this time, there are three key differences.

First: orbit and spectrum as non-substitutable strategic resources. Orbital slots and frequencies are physically scarce. If China does not file and launch now, once Musk’s 42,000 satellites are fully deployed, there may simply be no room left for latecomers.

Second: the primary driver is also commercial, not just political. The Apollo program was never profitable, but satellite connectivity today is a trillion-dollar market—powering 6G, autonomous driving, and global broadband. For example, China Mobile (0941.HK) has also applied for 2,520 satellite launches, not because a telecom operator suddenly developed ambitions for Mars, but because it wants to place its base stations in orbit—selling 24/7 connectivity and controlling the pricing and data flows of direct-to-device communications.

Third: the players have fundamentally changed. Space Race 1.0 was a pure state-to-state competition—NASA versus the Soviet space program. This time, it is a hybrid model: SpaceX as a private-sector champion versus China’s state-backed giants operating alongside a growing ecosystem of private companies.

How China Advances Its Orbital Strategy

To compete with Starlink’s established dominance, China has moved away from a pure “National Team” model. Instead, it has adopted a more distributed approach, spreading the burden of deployment across central authorities, regional governments, and private companies.

You can also get free access by sharing us. If you find our content helpful, please consider getting a paid subscription to access the exclusive benefits for Baiguan paying subscribers.

Subscribe now

Read more

Baiguan is a data-driven newsletter dedicated to scouting and synthesizing diverse data sources to provide deep market insights—this commitment to alternative data is the foundation of our collaboration with Skysight Technology, a leading provider of satellite remote sensing and alternative data services.

Established in 2017, Skysight brings together an elite background in aerospace research, macroeconomics, and financial analysis, providing institutional-grade solutions including standardized financial indices, supply chain insights, sector monitoring, and tailored analytical models across macro trends, commodities, and ESG.

On November 12th, 2025, the U.S. government concluded a record-breaking 43-day shutdown. While the machinery of the state was paralyzed, the global market was obsessing over a singular data vacuum: Was the American soybean harvest a bust, and had Beijing finally blinked?

The 2025 Timeline of the “Soybean Gambit”

• Feb - March: A 10% U.S. tariff on China takes effect; China retaliates by weaponizing soybeans.

• April - August: A brutal “tug-of-war.” China effectively halts procurement; U.S. exports to China drop to near zero.

• August - October: Harvest season. China maintains its “zero-buying” strategy, delivering a massive blow to U.S. farmer confidence.

• October 30: The Busan Summit offers a “de-escalation” signal, with soybean purchase commitments as a centerpiece.

• December: U.S. officials report that China is actively fulfilling a 12-million-ton commitment, with over 7 million tons already ordered.

With the USDA officially blind due to the shutdown, satellite remote sensing by Skysight Technology revealed a startling divergence -- while total U.S. soybean plant area is contracting, productivity is surging. Yields in primary producing states hit a projected 3.72 tons per hectare.

In an era where AI, semiconductors, and humanoid robotics dominate the headlines, it seems counterintuitive that a “low-tech” primary agricultural product could remain the most volatile chip on the US-China geopolitical chessboard. The question then is, how much weight does a single grain of soy actually carry?

Subscribe now

1. China as a Customer

Importing Soybeans = Importing Virtual Lands

The logic of the U.S.-China soybean trade is often reduced to a simple supply-and-demand curve, but viewed through the lens of macro-strategy, it is a resource arbitrage.

The U.S. possesses the world’s most efficient farmland, and China commands the world’s most massive demand for protein. For Beijing, importing soybeans is not a failure of domestic farming, by importing “land and water-intensive” crops, China effectively outsources the environmental and resource costs to foreign soil. This allows it to hold the “18亿亩耕地红线”(120 mn hectare Farmland Redline Policy) and prioritize the absolute self-sufficiency of primary staples like rice and wheat.

Food or Feed Divide

Despite record-breaking domestic harvests in 2024, China’s import dependency remains locked above 80%. This reveals a structural “hard floor” in the market:

• Domestic Soybeans: Primarily reserved for human consumption—the edible market for tofu and soy milk.

• Imported Soybeans: The industrial backbone used for large-scale crushing and livestock feed that’s largely driven by the rise of the middle class as the Chinese diet shifts toward meat

Unless there is a massive regression in Chinese dietary habits, this “80/20” dependency is a permanent fixture of the geopolitical landscape.

De-risking the “Soybean Circle”

Before 2018, the U.S. was one of the primary providers, commanding 40% of the market. Today, American beans have been relegated to a marginal buffer role. This is a proactive de-risking designed to move the U.S. from a core supplier to a secondary swing producer.

State-owned buyers now prioritize South American crops by default, turning to the U.S. only during the South American off-season or as a diplomatic currency to satisfy specific “de-escalation” agreements.

This transition has paved the way for a new Brazilian hegemon. By 2024, Brazil dominated nearly 70% of the Chinese market, exporting 3.5 times more than the U.S. This dominance was not an accident of nature but the result of a decade-long, aggressive “picks and shovels” strategy orchestrated by China’s state-owned giant, COFCO International.

Through the strategic acquisitions of Nidera and Noble Agri, COFCO secured the critical logistics nodes that bridge the gap between Brazilian farms and Chinese ports. The expansion of the STS11 terminal at the Port of Santos has effectively created a logistical moat. By slashing transport costs and streamlining the supply chain, China has ensured that Brazilian beans remain price-competitive even during the traditional U.S. export window from October to January.

Subscribe now

2. U.S. as a Supplier

Why the U.S. Midwest is Nature’s Factory

Viewed through the lens of natural endowment, the American “Corn-Soybean Belt” is an unrivaled agricultural fortress. Sitting on the world’s most fertile soft black soil, U.S. farmers enjoy deep organic layers and high water retention that their counterparts—who must battle acidic soils with massive infusions of lime and fertilizer—simply cannot replicate. This is a “natural cost advantage” that begins beneath the topsoil.

The true industrial cheat code isn’t just the soil; it’s the hydrological geography. The Mississippi River system and its tributaries (the Illinois, Ohio, and Missouri) act as high-velocity arteries pumping commodities from the heartland to the global market.

• The Efficiency Gap: A single 15-barge tow moves the equivalent of 1,000 large trucks, with many farms located within a stone’s throw of a river terminal.

• The Price Wall: In 2016, this geographic advantage translated to a staggering cost differential: shipping a ton of soybeans from Illinois to Shanghai cost $75, compared to $120 from Brazil’s Mato Grosso. This logistics premium is the primary reason the U.S. remains the world’s “preferred” high-efficiency supplier.

The Heartland’s Fragile Crown

Agriculture is the rare bright spot in the U.S. trade balance—a sector that consistently generates a surplus. At the center of this success are soybeans, the “crown jewel” of American exports.

In 2024, the soybean complex (including meal and oil) accounted for nearly 18% of all U.S. agricultural export revenue.

While this sounds like a triumph, it reveals a dangerous single-crop risk. By specializing so heavily in a crop grown for export, the U.S. agricultural sector has developed a structural fragility. Striking at soybeans is a direct hit on one-fifth of America’s total agricultural export engine. This concentration makes the Heartland uniquely vulnerable to targeted geopolitical tariffs.

A Dependency That Won’t Quit

If 2018 was a “black swan” event for trade, it also exposed a painful reality: the U.S. soybean industry is suffering from a “China addiction” it cannot kick.

Before the trade war, the dependency was stable, with over 60% of all exported U.S. soybeans destined for China.

In 2018, that figure cratered to 17.3%, leading to a domestic glut and a total price collapse.

Despite a partial recovery fueled by “de-escalation” agreements, exports have never returned to the 2016 peak of 36 million tons. By 2024, export dependency hovered around 44%, a sign that the industry is still breathing through an oxygen mask.

So, why can’t the U.S. find a new “China”? Washington has attempted to diversify its “protein basket,” but the results have been underwhelming:

• The Scale Problem: Combined, the growth in exports to Mexico, the EU, Japan, and Egypt has only recovered 6 to 8 million tons—failing to fill the 11-million-ton crater left by the shrinking Chinese market.

• The Crisis Dividend: The 2018 spike in EU imports was a mirage. Europe didn’t suddenly develop a taste for more soy; they simply bought U.S. beans because they were piled high and sold cheap after being blocked from China. This is bargain hunting, not structural demand.

The macro-reality is stark: China still consumes over half of all U.S. soybean exports. If Beijing were to fully sever this artery, the U.S. would face an immediate export loss of over $12 billion.

No other market on Earth—not the EU, not India, not Mexico—possesses the crushing capacity or the livestock scale to absorb what China takes. This is the asymmetric leverage Beijing holds. While the U.S. has the best soil and the best rivers, it is trapped in a buyer’s market where its biggest customer is also its biggest rival.

Subscribe now

3. The Great Insulation: Hedging Against Strategic Insecurity

The American Pivot: From “Land Grab” to “Fortress Efficiency”

The U.S. soybean landscape is undergoing a fundamental mutation. This year, the total hectares dropped by 7% —not a coincidence, but a long-term trend accelerated by the 2018 trade war.

On the ground, regional performance is diverging. The ‘powerhouse’ states, Illinois and Iowa, remain anchored at 17.44% and 16.56% of national output, respectively. Minnesota is seeing robust momentum as yields continue to climb. Kansas, however, has hit a turbulent patch; a yield slump in November dragged total production down by over 8% year-over-year.

Even during the USDA shutdown, eyes above the sky remained sharp. Through continuous tracking, the satellite captured an interesting shift: yield potential is climbing steadily.

U.S. farmers are no longer chasing expansion; they are chasing margins.

The American Heartland is moving toward an intensive production model:

• The Yield Explosion: Despite less land, total output remains resilient. Satellite data confirmed that improved climate conditions and high-tech management have pushed yields to a projected 3.72 tons per hectare—a massive jump from last year’s 3.41 tons per hectare.

• The Domestic Floor: The U.S. is aggressively building an industrial safety net via the Biodiesel Boom. Domestic crushing is projected to hit a record 68 million tons for the 2025–2026 cycle.

• Strategic Infrastructure: A wave of new processing facilities is hitting the soil in North Dakota, Nebraska, and Iowa. By processing beans at home into fuel, the U.S. is reducing its exposure to Chinese “buy-or-die” cycles.

China’s “Strategic Grain Fortress”: The Weapon of Patience

In August 2025, China imported a record 12 million tons of soybeans in a single month without purchasing a single American bean.

If the U.S. is hedging via energy, China is hedging via absolute volume and diversification. This was made possible by a two-pronged strategy of Extreme Tariffs and Aggressive Diversification:

• The Tarriff Wall: Under tarriff, American soybeans have effectively been priced out of the Chinese commercial market, allowing Brazil to capture nearly 90% of imports between May and August.

• The Argentina Hedge: In 2025, the Argentine government’s removal of grain export taxes significantly boosted its competitiveness. While Argentina’s current exports are a fraction of Brazil’s, its growth trajectory is a key part of China’s dispersed supply roadmap.

• The Strategic Stockpile: By the end of 2025, China’s soybean reserves are expected to reach 43.86 million tons—accounting for 36% of total global stocks. This volume has quadrupled since 2015.

By buying heavily when U.S. prices hit a five-year low in late 2024 (anticipating political shifts), Beijing converted high-quality, low-moisture U.S. crops into a strategic reserve. And because China now sits on enough grain to ignore the U.S. market for months at a time, it no longer has to buy out of panic. The U.S. may have the efficiency engine, but China has built a “Grain Fortress” that allows it to dictate the timing of trade de-escalation.

Subscribe now

4. Geopolitical Decoding: Why Soybean?

Economic data shows the intensity of a trade war, but only political geography explains why China chose to “strike here.” In 2018, when the U.S. launched its Section 301 investigation, Beijing needed a target that would cause maximum pain in the White House.

The soybean was selected because it possesses the perfect “weapon attributes”:

• Massive Volume: It is the single most valuable agricultural product the U.S. sells to China.

• Substitutability: Unlike semiconductors, soybeans can be sourced from Brazil or Argentina, providing China with a strategic off-ramp.

• Political Sensitivity: The “Soybean Belt” is the heart of Trump’s electoral base.

This asymmetry turned soybeans into the ultimate bargaining chip. By hitting the income of U.S. farmers, China aimed to force agricultural interest groups to pressure the White House into concessions.

The Electoral Heartlands: A “Red” Infrastructure

Of the top ten soybean-producing states, eight are steadfastly Republican. While states like Illinois and Minnesota often lean Democratic, this is largely driven by urban centers like Chicago and the Twin Cities. Peel back the city layers, and the rural “Soybean Belt” remains a fortress of GOP support.

For the Trump administration, China’s “zero-purchase” strategy was a brutal political stress test. By weaponizing the economic anxiety of the American heartland, China is attempting to force a retreat at the negotiating table by holding the 2026 Midterm elections hostage.

If the GOP loses control of the House or Senate in 2026, Trump’s second-term legislative agenda faces total paralysis. And if China can manufacture enough panic in the “Soybean Belt,” the resulting electoral hemorrhage will force the White House to trade away its “America First” tariffs for farm-belt security.

But, is China’s soybean tactic working?

The soybean weapon may be causing pain, but it isn’t flipping votes.

• The Iowa Expansion: In Plymouth County, a titan of soybean production, Trump’s margin of victory actually widened from 9.4% in 2016 to 13.2% in 2024, despite years of bearing the brunt of Chinese retaliation.

• The Wisconsin Wall: In Rock and Grant counties—critical bellwethers in the swing state of Wisconsin—GOP support did not collapse under the weight of the trade war.

On the surface, the conclusion seems obvious: now that Trump has successfully secured a second term and the farming states remain “deep red,” then China’s “soybean revenge” has fundamentally failed.

But is that the whole story?

Did the soybean strategy fail?

“Failure” is the wrong framing.

This is not a win-loss binary; it is a high-stakes endurance test where both superpowers are locked in a short-term game of brinkmanship and a long-term state of Strategic Stalemate.

First, the hard fundamentals of global supply dictate that a total severance is mathematically impossible.

On the U.S. side, the stability was not a result of market resilience, but of a massive Fiscal Shield.

• The Subsidy Injection: The White House successfully blunted the trade war’s trauma by injecting tens of billions of dollars into the rural economy via “Market Facilitation Programs”. These were effectively “fiscal painkillers” that masked the true wound caused by the loss of the Chinese market.

• The Sustainability Trap: This is a hollow victory. The U.S. cannot indefinitely sustain a multi-billion-dollar “governance tax” to keep its farming base afloat. With a ballooning deficit and inflationary pressures, this fiscal safety net is fundamentally unsustainable.

• No Alternative to Scale: Even with the “Biodiesel Boom” absorbing more domestic supply, the U.S. has no external market capable of digesting its massive surplus. Abandoning China would mean a painful structural contraction of the entire U.S. agricultural engine.

On the Chinese Side, the structural pivot has introduced its own set of geopolitical and environmental vulnerabilities.

• The Two-Stroke Engine: Brazil harvests from February to May, and the U.S. harvests from September to November. This seasonal alternation is the world’s only true guarantee of year-round protein security.

• The Brazilian Exposure: By tethering itself almost exclusively to Brazil, Beijing has traded “U.S. Political Risk” for “South American Climate Risk”. Any major drought or political upheaval in Brazil now poses an existential threat to China’s livestock industry, as there are no other suppliers with the requisite scale to fill a Brazilian-sized gap.

Second, and perhaps most crucially, the essence of deterrence here is the systematic escalation of the “governance tax.”

Beijing’s soybean strategy was never a simplistic attempt to flip the ideological switch of the American farmer. In the rural heartland, cultural identity and patriotic narratives frequently override pure economic rationality; a trade war rarely turns a “Red” county “Blue” simply through market pain.

If the soybean weapon didn’t change the ideology of the American farmer, what did it achieve?

To raise the “Cost of Governance” for the opponent.

China’s strategy was never about winning hearts and minds. Instead, the strategy is designed to force the opponent into a state of fiscal and political exhaustion. By making the trade war as expensive as possible for the White House to manage, Beijing turned a “low-tech” commodity into a permanent drag on the U.S. Treasury.

Entering the Era of High-Friction Coexistence

For all the ideological noise suggesting a total decoupling, the recurring tariff salvos and purchase freezes of the 2018–2025 cycle were never truly intended to kill trade. Instead, they represent a paradigm of leveraged interdependence.

Both superpowers are now weaponizing election cycles and short-term policy tools—tariffs, subsidies, and tactical pauses—to extract maximum gain while hedging against the risks of mutual uncertainty.

The soybean has ultimately become the defining microcosm of this High-Friction Coexistence. We must accept that while trade volumes may never return to their 2017 peak, the “Free Trade” honeymoon of decades ago has officially ended.

Yet, the two economies remain more deeply fused than the decoupling narrative suggests. While both nations are aggressively building domestic “fortresses”, they remain tethered by the objective gravity of supply and demand. Beijing cannot forever sever its ties to American protein, and Washington cannot survive the loss of its most significant agricultural customer.

Both sides are desperately seeking “security” through localization, yet they are forced to maintain a minimum viable connection to keep their respective economies functioning.

This state of constant collision and collision-management is not a temporary crisis but the new structural reality. In the coming years, the relationship will be defined by this permanent tension—an era where friction is the constant, but the inescapable pull of mutual dependence continues to foster new, albeit difficult, opportunities.

Related Reads

Hainan is China’s southernmost province, an island roughly the size of Taiwan. For decades, it has been best known domestically as a tropical vacation destination — a place of beaches, resorts, and winter getaways for Chinese families escaping the cold north.

To the casual observer, Hainan is often described as “China’s Hawaii.”

But for global investors and policymakers, December 18 marked something far more consequential. It was the moment Hainan formally transitioned from a leisure-oriented island economy into what Chinese policymakers call a “within the border but outside the customs” special economic zone — the operational launch of the Hainan Free Trade Port (FTP).

Subscribe now

How Does “Customs Closure” Work?

The term “customs closure” — or 封关运作 in Chinese — is easy to misunderstand.

“Closure” sounds restrictive, even isolationist. In reality, it represents the opposite. It is a deliberate step in China’s opening-up strategy, designed to expand zero-tariff access and ease cross-border trade — but within a carefully controlled geographic boundary.

Under the new system:

• Hainan and overseas markets effectively operate under zero-tariff treatment for eligible products

• Hainan and the Chinese mainland are separated by customs controls

This framework is often summarized by a phrase Chinese policymakers use repeatedly:
“First line open, second line controlled.”

In other words, Hainan becomes more open to the world, while its economic interface with the mainland remains regulated.

A Long-Anticipated Move

This shift did not happen overnight.

As early as 1988, Hainan was established simultaneously as a province and a special economic zone — the fifth in China, following Shenzhen, Zhuhai, Shantou, and Xiamen. Yet unlike Shenzhen, Hainan never became an economic legend. For most of the past three decades, its economic scale remained modest.

In 2010, Beijing elevated Hainan’s development to a national strategy, branding it an International Tourism Island and launching pilot programs such as offshore duty-free shopping. These policies helped tourism and consumption, but they stopped short of transforming Hainan’s economic structure. In 2024, Hainan’s GDP only ranked 28th out of 31 provincial-level regions in China.

And this time, the scope of opening is far broader: Duty-free product categories expanded from roughly 1,900 to over 6,600; cross-border foreign exchange controls were further relaxed to facilitate trade, reducing approval requirements and administrative friction; corporate and personal income taxes were also lowered to attract businesses and high-end talent.

Share

So Is China Building a New Hong Kong?

For foreign observers, this raises a natural question:

Is China trying to build a new Hong Kong — this time in Hainan?

After all, the similarities are hard to ignore.

Both are islands that sit at critical maritime crossroads. Both are designed as international trade hubs. And now, Hainan is introducing policies that sound very familiar: zero tariffs, simplified customs, looser capital controls, and preferential tax treatment for companies and talent.

So what’s really going on?

Subscribe now

Read more

That essay did not age particularly well.

By the end of 2025, few people would seriously argue that domestic consumption had become China’s top priority. Yes, consumption mattered. Yes, there were incremental pro-demand policies that would have been unthinkable a few years earlier—childbirth subsidy program, being the most visible example.

But taken together, these measures felt too small and too cautious. Even the birth subsidy program felt too modest. The familiar hesitation was still there. Consumption was encouraged, but never fully embraced.

And yet, over the past weeks, something genuinely different has begun to emerge.

First came Qiushi, the theoretical journal of the CPC Central Committee, publishing a curated series of Xi Jinping’s speeches from the past decade on domestic consumption. This was not a casual archival exercise. Several things stood out immediately. Consumption—more precisely, domestic demand—was explicitly framed as a strategic issue, not merely a desirable outcome. More importantly, a 2023 passage was highlighted in which Xi stated plainly that the key to boosting consumption is boosting income.

This matters because it eliminates any ambiguity about diagnosis. The leadership understands the problem. It understands the causal chain, so there is no intellectual confusion here. If execution fails, it is not because Beijing does not know what needs to be done.

Soon after, a senior official from the Central Office for Economic and Financial Affairs went further, stating outright that:

扩大内需是明年排在首位的重点任务。

“Expanding domestic demand will rank as the foremost priority next year.”

At the same time, People’s Daily, CPC’s main mouthpiece, just issued an explainer on why domestic demand ranks at the No.1 spot of all eight main priorities of next year:

Looking ahead, it is essential to genuinely enhance households’ consumption capacity and improve the social security system.

The investment structure should be optimized, with greater emphasis on consumption-related sectors, to promote a virtuous cycle between investment and consumption.

The official signaling can’t be clearer. So why is the market still not paying enough attention?

Because for years, the story of “boosting China’s domestic demand” has been a textbook case of that boy who cried wolf. More than a decade of speeches and urgings about consumption, demand, and rebalancing (including the Xi articles from Qiushi) can just as easily be read as proof that recognition has never been the constraint. It’s all about execution. If the problem has been so well understood for so long, why hasn’t it been decisively solved?

From this perspective, skepticism is entirely understandable.

Still, I think this time may be different—and that this possibility is being under-discussed. Let me explain.

First, I need to introduce a mental framework I use when thinking about China’s policymaking system. I sometimes call it the “Zhongnanhai Priority Framework.”

China is an all-encompassing state. Party documents touch everything: the economy, national security, education, technology, healthcare, the environment, ethnic policy, ideology. On paper, everything is always important.

In reality, nothing ever works that way.

A system this large cannot pursue a hundred objectives simultaneously. What matters is not what appears in documents, but what sits at the very top of the priority list on the leadership’s work desk—and that list is reordered constantly. When an issue genuinely rises to the top, the system can be astonishingly effective. Environmental enforcement in 2015–2016 is one example. Poverty alleviation was another. Anti-corruption at its peak was a third.

China does priorities extremely well. But the flip side of that is that it does non-priorities poorly.

For something as fundamental and sticky as shifting the fundamental growth engine from investment to consumption, partial prioritization is not enough. This is not a small technocratic tweak. It requires consumption to sit at the absolute top of that priority list, because there is immense institutional inertia against it.

Put yourself in the shoes of a mayor or county chief. You have a roughly three-year term. You want GDP growth, fiscal revenue, and employment stability. Do you focus on attracting a factory line that shows up cleanly in investment and output statistics? Or do you distribute cash or income support to households, hoping it turns into consumption rather than the more likely destination of even more savings?

The rational choice is obvious. Investment is fast, measurable, and politically safe. Consumption is uncertain, indirect, and often fiscally unrewarding for local governments. Even when households spend more, China’s tax structure means local governments capture little of the upside. Not to mention, China’s local governments have been severely underfunded for the past few years due to the real estate slump, making fiscal resources even more precious.

The local pushback is thus structural, and it is why exhortations alone do nothing. Overcoming it requires overwhelming top-down force, coordination across ministries, and meaningful changes to fiscal and tax incentives. Anything less simply dissipates.

Seen through this lens, it becomes obvious why 2025 did not see real action on demand. The core priority of 2025 was unmistakable. Everyone knows: it’s none other than the second Trump administration and the renewed trade war. The entire government sector braced itself for the impact of this event. Everything else, including consumption, was secondary. That was not a close call.

Here is where the setup for 2026 begins to look different.

Read more

Read more

At the Wuwei Thorium-based Molten Salt Reactor (TMSR-LF1), a 2MWt (megawatt-thermal) experimental reactor, Chinese scientists successfully completed the facility’s first cycle of converting Thorium-232 into fissile Uranium-233 within a liquid fuel environment [*].

This breakthrough, achieved by the Shanghai Institute of Applied Physics (SINAP), marks the world’s first successful use of thorium in a molten salt reactor, which could be pivotal in the development of advanced nuclear energy systems using thorium as a sustainable fuel source.

This is a technical milestone—the first time in over 50 years that any nation has successfully operated a molten salt reactor of this type. But viewed through the lens of macro-strategy, this is not just a science experiment. It is the unveiling of a massive “strategic hedge” against China’s most glaring resource vulnerabilities.

For a nation that imports over 70% of its oil and over 80% of its uranium, energy security is a constant anxiety. The breakthrough in Wuwei offers a potential exit ramp from this dependency. (I wrote about this in the previous newsletter: “China is short on crude oil!”, where I explained how China’s vast land can be misleading, leading many to believe that the country is rich in strategic resources. In fact, many crucial resources are actually scarce. This supports the key motivations behind China’s national strategy to pivot toward EVs, sustainable energy, and generational projects such as building the world’s largest hydropower station in Tibet.)

By mastering the Thorium cycle, China is effectively unlocking a domestic energy reserve that has been sitting in its soil for eons, untouched—burning rocks that were once considered waste to power the country for thousands of years.

Subscribe now

From “Waste” to “Fuel”

In nature, Thorium and Rare Earth Elements are “twins”—they almost always coexist in minerals like Monazite.

For the rest of the world, this relationship is a curse. When Western companies mine for Rare Earths, they are forced to extract Thorium as a by-product. Because they lack a use for it, it is classified as hazardous radioactive waste, requiring expensive disposal and strict environmental permitting.

For China, however, this relationship is an industrial “cheat code.” Because China has monopolized global Rare Earth refining for decades, it has been passively accumulating massive stockpiles of Thorium. For instance, in the tailing dams of the Bayan Obo mine in Inner Mongolia, thousands of tonnes of Thorium sit waiting. (The Bayan Obo mine is the world’s largest deposit and producer of rare earth elements.)

While the US or Europe would need to permit and dig new mines to access this fuel, China simply needs to open the warehouse door. It is a pre-mined, pre-refined energy inventory that turns a decades-old environmental liability into a tier-one strategic asset.

The History of the Thorium Research and How China Did It

The concept of a molten salt reactor is not new; the technology behind the Wuwei reactor was actually pioneered in the United States. The US Oak Ridge National Laboratory (ORNL) proved the physics worked in the 1960s. In the 1960s, Alvin Weinberg, the legendary director of the ORNL in Tennessee, built and ran the Molten Salt Reactor Experiment (MSRE). It ran successfully for four years. Weinberg argued passionately that thorium was the safer, cleaner future of civil nuclear power [*].

So, why did the US kill the program in 1973?

The answer is a mix of Cold War pragmatism and technical hurdles.

Read more

China must undertake significant structural adjustments. Consumption—not investment—must become the new engine of economic growth. And within consumption, the next major wave of expansion will come from services.

To understand the issue better, we turned to CF40’s recent research.

Why CF40’s new work matters

Some of you already know this, we are collaborating with CF40, China’s pre-eminent economic thinktank, to help run their English Substack, CF40 Research:

CF40 Research

Insights on China’s policies and markets from the leading think tank and top economists in China

Four recent pieces from CF40, taken together, offer one of the clearest pictures yet of China’s service-economy challenge.

The first essay, based on A-share consumer companies, shows that China’s consumption mix remains heavily goods-driven, that essential spending has risen since 2020, and that services remain underrepresented and undercapitalized.

The second essay revalues China’s actual consumption basket using foreign price systems and finds that China’s goods consumption volumes are already close to developed-country levels. The real gaps lie in services.

The third essay looks across 60 years of global data and shows that only two types of services consistently expand as countries become wealthy: producer services and high-end social services. These are precisely the areas where China lags behind.

The fourth essay examines how advanced economies avoid stagnation. It finds that welfare systems and knowledge-intensive services are essential for balancing high savings with adequate investment and consumption. These service-heavy systems are the institutional foundation China has not yet fully built.

Below is a detailed digest of these four essays, in sequence.

What A-share companies tell us about consumption

CF40’s “Understanding China’s Household Consumption through A-Share Consumer-Sector Listed Companies” begins with a simple but revealing fact: the listed consumer universe is overwhelmingly goods-oriented. As the authors write, “A total of 581 A-share consumer companies are selected as the analytical sample, of which 521 are classified as goods companies and 60 as services companies”. This imbalance mirrors the real economy.

From the A-share perspective, goods dominate both in scale and in structural trends. The authors note that “goods consumption showed an upward trend while service consumption exhibited a downward trend from 2014 to 2022”. Services only began to recover in 2023, and even then from a small base.

One of the clearest findings is the post-2020 pivot toward essentials. The paper highlights that “the share of essential consumption has increased, while the share of discretionary consumption has declined”. Grain and food products surged and have stayed elevated, reflecting both uncertainty and the defensive nature of household spending.

A-share data also became more tightly aligned with macro indicators after 2020. The authors report that “the linear correlation between the two growth series strengthens considerably after 2020… 0.7 for 2015–2019 and 0.97 for 2020–2024”.

In the conclusion, the authors underscore the central structural issue: “China’s share of service consumption still needs to increase, and the representation of service-sector firms among listed companies remains noticeably low”.

What China really consumes

In “The Value of China’s Consumer Basket in the U.S., Japan, France, Germany, and Mexico,” CF40 researchers ask what Chinese households actually consume, not what they spend. They reprice China’s consumption basket using foreign price systems.

The results challenge the prevailing narrative of “low consumption.” As the authors show, “China’s consumption basket is valued at $24,748 in the United States, approximately 80% of the U.S. level; $6,006 in Mexico… 1.17 times that of Mexico’s level; $9,253 in Japan, 93% of Japan’s level; and $9,235 and $10,805 in France and Germany respectively, approximately 75% of the levels in France and Germany”.

Even under the conservative scenario, which assumes unmeasured categories are as low as nominal data suggests, “China’s entire consumption basket is valued at $26,857 in the United States, representing 50% of U.S. per capita consumption”.

Where does China differ? In the parts of consumption that the basket does not capture. The authors explain: “The portions not included in the consumption basket are tobacco, alcohol and beverages, daily household goods, as well as service consumption in transportation, postal services, culture and entertainment, and restaurants and hotels services”.

In the summary, the authors remind readers that these measurement adjustments do not resolve structural issues. “Re-examining China’s consumption level does not resolve the existing structural challenges in the economy”. The challenge lies not in goods consumption, which is relatively high, but in the deep underdevelopment of many service categories.

Where service growth really comes from

“The Rise of the Service Sector in the Post-Industrial Era” looks across 23 economies over six decades. Its main finding is that the global rise of services is highly concentrated. As the authors write, “a very high proportion of the increase in the service sector’s share since 1965 comes from just a few subsectors”.

These subsectors fall into two groups.

First, producer services. The paper notes that these “account for about 60 percent of the rise in the service sector’s share”. These are the R&D, legal, accounting, engineering, and digital services that support a modern industrial economy.

Second, high-end social services. These include healthcare, social work, education, and other human-capital-related services. Depending on the sample, they contribute between 17 and 28 percent of service-sector expansion.

By contrast, the consumer-facing services people often associate with “service consumption” — restaurants, hotels, entertainment — remain small everywhere. They do not structurally drive service-sector growth.

The international comparison section quantifies China’s gaps. The authors write that China lags the global median in “renting and business services; health and social work; and real estate activities”. These are the very categories that expand most predictably as countries become wealthier.

The implication is direct: if China wants its service sector to rise to international norms, it must expand producer services and high-end social services, not rely on tourism or leisure alone.

How services prevent high-savings economies from stagnating

The final CF40 essay, “Rebalancing Savings and Investment in Industrialized Economies,” examines how advanced economies avoid the long-term stagnation that can arise when savings chronically exceed investment. The paper argues that much depends on households' ability to consume and firms' ability to invest in new forms of capital.

One mechanism the authors highlight is the development of new, service-heavy growth fronts. They write that advanced economies depend on “new investment opportunities in knowledge-intensive sectors” once traditional industrial returns decline. These sectors are overwhelmingly producer services: software, scientific R&D, intellectual property, and business services.

At the same time, social services and welfare systems reduce precautionary savings by lowering the household burden of healthcare, education, and old-age support. The combination of robust producer services and strong social services helps prevent persistent gaps between desired savings and actual investment.

This leads to the paper’s clearest message for China: countries at this stage “must build new institutional foundations for consumption and new frontiers for investment”. In China’s case, these frontiers align closely with the producer services and social services identified in the previous essays.

Bringing it together

Viewed together, these four CF40 essays point to a coherent conclusion.

China’s consumption challenge is not simply about confidence or stimulus. Goods consumption volumes are already high. What is small is the country’s service ecosystem: producer services that support innovation and industrial upgrading, and social services that reduce precautionary savings and support household wellbeing. The service economy is not just a symptom of development; it is the mechanism through which modern economies balance savings, investment, and household welfare.

For China to move into its next stage of growth, it must build the service infrastructure that its development level now requires.

Subscribe now

But there is a deeper layer to this story. China’s tech innovation doesn’t come out of nowhere; it flows from a massive, rigorously selected STEM talent pool. (For instance, DeepSeek’s founder and core team are local talents from Zhejiang University in Hangzhou, one of China’s most prestigious research universities.)

To truly understand the future of Chinese tech, you have to look at how that talent is made.

In this newsletter, I want to take you inside Westlake University — China’s first privately funded research university, also based in Hangzhou. During Baiguan’s very first China tour in late October, we visited Westlake University in Hangzhou, which offers a rare peek into how the country is innovating its higher education system to select and train the next generation of scientists, researchers, and world-class STEM talents.

Subscribe now

The context: How China’s broader education system works

When we talk about China’s higher education, we are usually describing a contradiction.

On the one hand, you have students studying insane hours, weekends included, with their fates sealed by a single score on the Gaokao (the national college entrance exam). There is a popular saying in China: “A single point on the Gaokao can crush thousands of competitors.“ It is brutal, stressful, and for many, results in a degree that doesn’t guarantee a job as the youth unemployment rate keeps climbing in recent years.

On the other hand, this same system produces students who dominate international math and science competitions, found groundbreaking companies, and drive technological breakthroughs globally.

How can a system be so stressful yet so effective?

All coins have two sides — the system isn’t designed to fit everyone, but it’s exceptionally good at selecting STEM talent. For quantifiable subjects, exams are simply the fairest way to give everyone — rich or poor — a shot at changing their fate. The Gaokao system is, in my opinion, less of an “education” system and more of a meritocratic selection mechanism.

However, in recent years, the system has faced a growing crisis of quantity over quality. Since 1999, China has massively expanded college enrollment. From 1998 to 2001, total higher-education enrollment nearly doubled. It relieved unemployment pressure at the time and raised overall education levels — but also created issues: too many poorly designed majors, student-to-faculty ratios ballooning, and degree inflation that left many graduates misaligned with the job market.

This backdrop—the need to build a parallel track to nurture top-tier talent—is why a case study of Westlake University matters.

Westlake University: a case study in educational innovation

It is easy to imagine that the entire Chinese higher education system works like the aforementioned exam-based, rote memory syllabus, and cutthroat machine. Westlake University is the opposite. It is small, flexible, research-driven, and deeply selective. And because it is private, it moves fast and has the freedom to choose exactly how it runs itself.

In China, where higher education is traditionally dominated by the public system, the concept of a private research university was almost unthinkable just a decade ago. Yet Westlake, founded only in 2018, is already attracting top talent globally. The proposal was made in 2015 by seven prominent scholars and entrepreneurs; it was quickly approved by Hangzhou and Zhejiang Province, and the government helped build a massive, brand-new campus with world-class labs in just a few years.

Westlake University is now home to a faculty of internationally recognized scholars, including Shi Yigong, a leading biophysicist and the university’s founder and current president.

Here is how they are doing it differently:

1. True “Liberal Education”—or rather, “teach students in accordance with their aptitude” (因材施教)

Westlake provides what the West might call a “liberal arts” education, but I prefer the Confucian concept of “因材施教 Yincai Shijiao”—teaching in accordance with the student’s aptitude.

This has long been the ideal of education in Chinese history, but due to massive college expansions, few universities today have the capacity to actually implement it. Westlake does, largely because it limits its scale to focus on quality. The university hosts only about 200 undergraduates compared to 2,700 PhD students and 300 professors, concentrating intensely on STEM and medicine rather than trying to be a comprehensive giant.

Because of this small ratio, the curriculum is highly personalized and interdisciplinary. For the first two years, undergraduates don’t even choose a major; instead, they choose a mentor. Based on mutual research interests, professors and students select each other, allowing students to explore research fields before committing. The atmosphere mirrors top US research universities, with students living in one of four residential colleges where they socialize and participate in extracurriculars with dorm-mates across different disciplines. Remarkably, for this level of education, the tuition for local students is just 6,000 RMB (approx. $845 USD) per year.

From Day 1, Westlake has been committed to a global standard of research. All classes are taught in English, and it is mandatory for undergraduates to study abroad for a semester or a year, supported by need-based scholarships.

The most fascinating part of Westlake is how they pick these students. The process is designed to find “prodigies,” with prospective students taking Westlake’s own exam before the national Gaokao. Remarkably, 60% of their admitted students have skipped a grade. Even though they are younger than the average high school senior, their Gaokao scores are typically only about 20 points shy of the ultra-elite Tsinghua or Peking University standards.

But high scores are just the prerequisite. Finalists must attend a 3-day assessment at the university where professors might deliver a lecture on a topic the students have never heard of, then test them on the spot to gauge their learning agility rather than rote memory. Students must also deliver a full English presentation. In Westlake’s values, they want talented minds who know, from very early on, that they want to become scientists and researchers.

Share

2. Embracing curiosity and the unknown

As a research university, curiosity and a willingness to embrace the unknown are deeply rooted in Westlake’s values. You can see this in their hiring strategy: they prioritize young professors—the average age is just 42, and 85% have overseas experience. Notably, 66% were recruited through state-sponsored talent acquisition programs designed to repatriate top scholars globally, ensuring that the faculty consists of world-class researchers returning to China with fresh perspectives.

During our lab tour, the team shared a story that perfectly captures this culture. Led by Dr. Sun Licheng, the group was researching bio-membranes designed to convert greenhouse gases like carbon dioxide into useful resources. They were specifically hunting for a cost-effective alternative to electrolyte membranes, which traditionally rely on expensive noble metals.

The breakthrough arrived by accident. The team discovered a watermelon that had been left in a forgotten lab refrigerator. When they found it, the skin had already naturally detached from the fruit due to the cold.

The team got curious: could the skin be useful? They decided to experiment. They took the skin, froze it, restored it to room temperature, and tested it. It worked—the material offered efficiency and dynamic response times that rivaled expensive noble metals. The professor told us this reflects their core philosophy: “start from nature and make some artificial things from it.”

In fact, this watermelon skin research won the Pineapple Science Award—a prize established in Zhejiang in 2012 (similar to the Ig Nobel Prize) to celebrate curiosity-driven, imaginative, and humorous research. It is a telling detail: when a university is rooted in a province that actively celebrates this kind of curiosity, it is likely no coincidence that tech talents and unicorns like DeepSeek are emerging from this city.

3. The governance model

As the first private research university in China, Westlake operates with a level of freedom that public universities simply don’t have. While public universities are led by a Party Committee, Westlake operates under a Board of Trustees leadership system where the Principal takes responsibility. The board includes heavyweights like Nobel laureate Yang Chen-ning, scholars, as well as entrepreneurs.

The funding model is a unique hybrid. The state provides the “hard assets” like land and lab equipment, but the private foundation covers the day-to-day costs such as salaries and operations. This allows them to pay market rates to attract top global talent. For instance, the university would tailor one lab for the newly hired professor to make sure he or she gets all that’s needed for the research.

This governance structure also leads to highly pragmatic and comprehensive decision-making. The Board of Trustees includes renowned scholars—such as former Vice Presidents of Tsinghua and Peking University—sitting alongside major entrepreneurs, including a co-founder of Alibaba Cloud. With one-third of the trustees being entrepreneurs (who have donated over 200 million RMB), the university bridges the gap between academia and industry. This means graduates are trained not just for foundational research, but also for down-to-earth, practical entrepreneurial work. In fact, Westlake has already incubated 52 high-tech enterprises, most of which are based right here in Hangzhou.

Since it relies on a foundation rather than the state for operations, Westlake is currently receiving donations but has reimagined fundraising in a way I’ve rarely seen elsewhere in China. We heard a fascinating anecdote about their donation flexibility: naming rights can extend down to specific seats in the auditorium. Someone, for instance, sponsored Row 5, Seat 3 simply because it matched their birthday. They even accept small donations via Alipay. By lowering the barrier to entry, the university found a fun way to maintain the deep connection supporters—both in China and abroad—have with the school, regardless of the amount they could afford to give.

This kind of flexibility is unheard of in traditional Chinese universities, but it makes perfect sense in Hangzhou—a city where the entrepreneurial private economy contributes over 60% of GDP and accounts for over 90% of all business entities.

Some say Hangzhou is becoming China’s Silicon Valley, and it’s an accurate analogy. Its population is digitally savvy and entrepreneurial, while prestigious institutions like Zhejiang University and Westlake University provide talent and fuel technological development. This ecosystem has birthed Alibaba, as well as the “Six Little Dragons”—influential startups including DeepSeek, Unitree, Deep Robotics, and Game Science (creators of Black Myth: Wukong).

Subscribe now

The broader innovation in China’s talent selection

Westlake isn’t alone. China has started building new pathways to cultivate foundational research talent.

Since 2020, China’s Ministry of Education has rolled out the “Pilot Reform Program of Enrollment for Fundamental Disciplines” (强基计划). This program targets students who are exceptional in foundational fields—math, physics, chemistry, biology, history, philosophy, and ancient linguistics.

The pilot program only involves a selected group of elite universities in China, and Admission is based on a composite score: at least 85% from the Gaokao (college entrance exam) plus about 15% from a school-administered test. Once admitted, students get specialized training: smaller classes, a mentor system, and are designed to support “undergraduate → master → PhD” pathways.

By 2023, the program had evolved: some schools added more interdisciplinary and strategy-relevant majors (e.g., AI, new materials), and tightened selection thresholds.

On this front, I have never doubted China’s ability to ensure its top STEM talent is selected and trained. This is not to ignore the obvious issues: the stressful exam-based system, massive class sizes, and degree inflation remain alarming problems for the average student. But for the true high-performers, it is fair to say that China’s higher education remains an effective meritocracy.

To be fair, even the average baseline is high. The average high school student here possesses math skills that often outstrip their counterparts in other countries—a disparity that becomes evident if you compare the difficulty of China’s standardized exams against Western equivalents like the SAT subject tests or AP tests. These innovations in education act as finer filters on an already massive, capable pool.

Such a system may not be optimal for liberal arts talent, but if you are good at math and science, you will have a fair opportunity. That reality lies the foundation for the sheer volume of brilliant scientists, researchers, and engineers coming out of China.

*Westlake University now operates primarily through its endowment and still relies on donor support. If you’re interested in learning more about its funding model or making a contribution, feel free to reach out to us.

Related Reads

The most immediate impression came when I took a taxi. The driver’s attitude was remarkably pleasant. In a 2023 article, I wrote about the struggles of ride-hailing drivers—oversupply, the overwhelming power of platforms, and rental companies squeezing their earnings. In Shanghai, taxi drivers often seem irritable and quick to complain. At Hongqiao Airport, if your trip is a short one, drivers can get visibly upset after waiting in line for hours for a passenger. But the drivers in Yulin appeared far more relaxed, unhurried, and content, even with fares as low as ¥5–10. The contrast with Shanghai was striking.

I’ve heard that many locals received substantial cash compensation from the government for land acquisition. They don’t necessarily need the money, but they drive to maintain a daily routine—to have somewhere to be, something to do.

Subscribe now

Yulin’s economic profile: An energy hub with pioneering social policies

Yulin’s GDP has consistently ranked second in Shaanxi Province, second only to the provincial capital Xi’an. Its economic output even surpasses that of some provincial capitals in western China, making it a crucial economic driver for Shaanxi. With its high-value energy industry and relatively small population (around 3.6 million), Yulin’s GDP per capita is among the highest in Shaanxi and nationwide, even exceeding that of some eastern coastal cities.

Boasting world-class energy and mineral resources, Yulin is often called the “Kuwait of China.” Coal is the absolute backbone of its economy—the Yulin Coalfield is one of the world’s seven largest, with massive reserves of high-quality coal (low ash, low sulfur, low phosphorus, high calorific value). The city is a core base for China’s “West-to-East Coal Transmission” strategy. It also forms a key part of the Shaanxi-Gansu-Ningxia oil and gas field, with abundant reserves of petroleum and natural gas.

Supported by its resource-driven fiscal capacity, Yulin and its affluent counties—such as Shenmu and Fugu—have rolled out social policies that exceed national standards and gained nationwide recognition:

• “Shenmu Healthcare Reform” (*): As early as 2009, Shenmu pioneered a “universal free medical care” system. Funded by local fiscal revenue, the policy covers the majority of hospitalization expenses for urban and rural residents at designated hospitals within the city. Although implemented at the municipal level, it demonstrates how resource wealth can support high-level welfare programs.

• 15 Years of Tuition-Free Education: Yulin has implemented 15 years of free education from kindergarten through high school across the entire city—6 years longer than the national 9-year compulsory education requirement—significantly reducing the financial burden on families.

*National Basic Insurance: Across China, the basic medical insurance system consists primarily of two schemes: one for urban employees and another for urban and rural residents. These schemes provide a fundamental level of coverage, but patients still bear a significant portion of costs through deductibles, co-payments, and expenses for drugs or services outside the reimbursement list. The out-of-pocket burden can be substantial, especially for major illnesses.

Share

Urban infrastructure: A train station with crystal-clear screens and only public service ads

The quality of Yulin’s urban infrastructure is immediately apparent in public facilities like its train station. Unlike the often crowded and commercially saturated stations in major hubs, Yulin Station feels spacious, well-maintained, and thoughtfully designed.

A striking detail captures this perfectly: the digital display screens within the station are exceptionally high-definition, arguably sharper than many home televisions. More notably, these screens are dedicated solely to broadcasting public service announcements (PSAs), with no commercial advertisements in sight.

Well-preserved folk culture in Yulin: prosperity makes courtesy/”仓廪实而知礼节”

If you have visited various “ancient towns” and “old streets” across China, you may have noticed that most of them feel strikingly similar. The reason lies in the fact that many of these historic streets and towns have been restored or reconstructed essentially as commercial real estate projects. To recoup investments quickly, minimize risks, and ensure profitability, developers tend to adopt the most proven and low-risk development templates—standardized architectural designs that can be rapidly replicated nationwide.

Furthermore, shops in these ancient towns are mostly operated by individual tenants whose product selection follows a simple logic: whatever sells well gets stocked. Sourcing from national wholesale markets like Yiwu offers low costs and variety, but inevitably leads to the same products being sold everywhere. Embroidered bags, wooden combs, ethnic-style scarves, and low-quality crafts become the standard fare. Similarly, snacks like stinky tofu, grilled sausages, bubble tea, and skewers—easy to make, widely popular, and highly profitable—naturally become merchants’ first choice, crowding out traditional local snacks that require complex techniques, carry unique regional characteristics, and may appeal to a smaller audience.

However, Yulin Old Street is distinctly different.

Subscribe now

Firstly, alongside tourists, the street is also frequented by locals who visit its grocery stores, dairy shops, and food suppliers. Secondly, folk culture is visibly alive and present everywhere. For instance, there is a shadow puppet (”皮影戏”) museum housing the private collection of a German enthusiast. After the puppets were damaged in the floods in Germany a few years ago and the collector unfortunately passed away, the museum has been meticulously restoring them.

To escape the rain, we stepped into a Northern Shaanxi storytelling hall (”陕北说书馆”), where entry was free. We watched a documentary about the art of storytelling and a live performance.

*The game Black Myth: Wukong masterfully incorporates the gritty, narrative strains of Northern Shaanxi Storytelling as a diegetic narrator (Let me show you a glimpse of it with this video clip)

Another example is Colored Cotton Collage Art. As the name suggests, it is a form of relief art created using colored cotton as the raw material, through techniques such as layering, attaching, and sculpting. It is not a centuries-old craft, but rather a unique folk art invented by Mr. Zhang Shiyu, a folk artist from Yulin, in the 1990s. It has been included in the Shaanxi Provincial Intangible Cultural Heritage protection list.

When we went into a small shop displaying colorful cotton collage art, an elderly gentleman welcomed us and explained that the craft wasn’t centuries old—it was actually invented by his father, and now he and his son are carrying it forward. Here, cultural heritage is not a static relic; it is vividly alive and actively present.

As I write this, my initial point was to illustrate the vastness of China and the tremendous disparities that exist between different regions—economically, fiscally, socially, culturally, and in the daily lives of their residents. While young people struggle to find jobs in first-tier cities and the middle class grapples with falling property prices and mortgage stress, there are places where people lead prosperous and tranquil lives.

However, upon closer examination of Yulin’s economic situation, I’ve come to recognize underlying risks. The main concerns are: 1) its heavy dependence on resource prices, and 2) the significant disparities within Yulin itself, between its different areas.

Subscribe now

Underlying risks: Over-reliance on the energy sector and uneven development

Yulin’s economic vitality is measured by the rhythm of a single heartbeat: the price of coal. This energy-rich region lives and breathes by the fluctuations of its primary resource—where fiscal health and personal prosperity rise and fall with the market’s every tremor.

Yulin’s economic trajectory has been closely tied to fluctuations in coal prices. In 2021, as prices surged to historic highs, the city’s GDP and local fiscal revenue grew markedly—with fiscal revenue jumping 44.6% year-on-year—while household incomes also rose rapidly. In 2022, although coal prices retreated from their peak, they remained elevated, allowing fiscal and income growth to continue, albeit at a more moderate pace. By 2023, as coal prices settled into a more rational range, growth in GDP and fiscal revenue slowed further.

Yulin also faces pronounced internal imbalances. At the regional level, resource-rich northern counties such as Shenmu and Fugu—which contribute the bulk of municipal fiscal revenue—enjoy significantly higher per capita incomes than agricultural southern counties. At the sectoral level, industries like energy extraction and chemicals generate far higher incomes than agriculture or traditional services, creating a stark “energy-rich, rest-poor” divide. This dual disparity, both geographic and industrial, highlights the typical “internal inequality” of a resource-dependent economy.

Related Reads

China has also made functional advances in technologies such as drones, extra-large unmanned underwater vehicles, and other unmanned aerial systems. Unlike a decade ago, the parade demonstrated China’s growing ability to organize coordinated military operations, integrating multiple units to act coherently.

Because China has enjoyed decades of peace since WWII, most of its weaponry has not been battle-tested, making it difficult to fully assess their performance for potential global buyers. That said, Chinese military technology is gaining increasing attention internationally. For example, Pakistan’s use of Chinese-supplied Chengdu J-10C fighter jets, reportedly shooting down multiple french-made aircraft in a conflict with India in May this year, marked one of the first significant combat tests for Chinese-made equipment against Western systems.

While rising geopolitical tensions cannot be ignored and conflict is not desirable, it is now worthwhile to set aside political debates and “China threat” narratives to objectively examine the current state of China’s defense capabilities and its position in the global arms trade.

In today’s newsletter, we translate a highly informative overview addressing exactly this question: where does China rank in global arms exports? The original content comes from a YouTube video essay by 巫师财经, whose career began as a banking analyst at China International Capital Corporation (CICC) before co-founding his own fund, and who now produces finance and business content. (For fluent Chinese speakers, the original video essay can be viewed on YouTube here.)

Subscribe now

Below is our translation of the video transcript. Some narration scripts and charts have been abbreviated or reformatted for clarity.

Military trade behind the parade: where does China rank in global arms exports?

How competitive are Chinese weapons?

How well do Chinese weapons sell globally? How much revenue do they generate? And who are the buyers?

Over the past decade, China accounted for about 6% of global arms exports, ranking fourth. The United States, with roughly 40%, remains the undisputed global powerhouse of weapons exports. Russia ranks second at 15%, though that largely reflects legacy sales.

Viewed in isolation, the trends are striking. Russia’s share, once 21% in 2015, has collapsed in recent years, leaving it on par with China. The U.S., by contrast, has seen a steady climb, reaching 47% of the global market in 2024—nearly half of the world’s arms trade. China, meanwhile, quietly gained ground: between 2021 and 2023 its exports exceeded Russia’s, peaking at 10% in 2023.

In absolute terms, China’s 10% that year placed it second globally.

The turning point came around 2012–2014, when China shifted from being a major importer to a net exporter.

Imports, in fact, have collapsed. China’s share of global imports fell from 7.4% to just 0.2% last year, dropping its ranking from the global top 10 to around 50th.

What lies behind the business numbers? It’s strength.

China’s place in global military power rankings

GlobalFirepower’s 2024 index, covering 145 countries, places the U.S. first, Russia second, and China third. India, South Korea, the U.K., France, and Japan occupy the next five slots.

The ranking, known as “PwrIndx,” aggregates more than 60 factors—from manpower and equipment to resources, fiscal capacity, and geography. A lower score signals greater strength. By this measure, China and Russia are tied at 0.0788. But as with all indices, methodology is political. Western institutions carry their own biases. If the U.S. Navy were publishing the report, China might be portrayed as poised to strike America’s shores—an argument designed to justify bigger budgets. GlobalFirepower, if anything, tilts the other way.

Even so, the pattern aligns with export data: the U.S. dominates in a league of its own; China and Russia make up the second tier; the rest trail far behind. India, though ranked fourth, lags significantly behind the top three.

If you look into China’s specific score composition, the ones holding things back are mostly related to energy consumption, debt issues, and long borders with other countries. When it comes to actual equipment numbers and quality, the scores are very high. Essentially, it all comes down to how the weights are assigned.

But maybe it’s just that after years of doing the dirty work in investment banking, I tend to assume others are always thinking a bit schemingly. At CICC, we used “Dealogic” to cut the data until we ranked first. If total IPOs didn’t put us on top, we narrowed it to state-owned IPOs, or to overseas SOE deals. If that still didn’t work, we built a weighted scoring system that gave larger transactions higher multipliers. In the end, the numbers delivered the headline we wanted: CICC was No. 1. Clients never cared about the methodology—only the headline result. By the same token, military indices should be treated as indicators of trends, not absolute truths. As Nietzsche put it, “There are no facts, only interpretations.”

Does this mean China’s ranking is artificially deflated? In practice, its defense industry has already advanced considerably. Arms import data tell the story: between 2020 and 2024, the world’s top arms importers were Ukraine (8.8%, it’s at war, after all), followed by India (8.3%), Qatar (6.8%), and Saudi Arabia (6.8%). Their primary suppliers were the U.S., Germany, and Poland.

Among the top 20 arms importers, almost all sourced from the U.S. and Europe, with only occasional purchases from Russia. Pakistan was the lone exception: 81% of its weapons came from China. This shows that Chinese equipment has yet to gain broad global recognition, or that the current global landscape limits the reach of China’s defense exports. After all, arms trade carries strong political significance, and the diplomatic environment remains relatively challenging.

For now, China cannot break into Western markets. Beyond Pakistan, the question remains: where else can Chinese arms find buyers?

Subscribe now

Who is buying China’s weapons?

Looking at China’s top ten buyers over the past ten years, these are the countries that have genuinely put down their money. Some might say, ‘Except for Saudi Arabia, aren’t these all relatively poor countries?’ Even so, this is already a solid achievement, achieved through China leveraging two major factors: ‘defense balancing’ and ‘opportunistic openings.’ And that’s on top of the fact that our equipment offers higher cost-performance, comes without political strings attached, and often includes favorable financing—later you’ll see why this matters.

Over a ten-year period—not just five-year—Pakistan accounts for half, ranking first. Bangladesh is second at 11%, Thailand fifth at 4%, and Serbia sixth at 3%, all showing signs of seeking defense balance. Beyond surface-level friendship, this is a key factor allowing us to expand into these markets. The other factor, of course, is taking advantage of ‘opportunistic openings’ (Baiguan: Beijing’s readiness to fill supply gaps when traditional suppliers falter.)

Thailand is a case in point. Beyond its traditionally friendly ties with China, it had already purchased a range of Chinese equipment. But Thailand also sourced heavily from Ukraine, especially tanks and armored vehicles. When the Russia-Ukraine war broke out, China stepped in to fill the gap, and exports to Thailand increased further. According to The Diplomat, between 2016 and 2022 Thailand imported $394 million worth of Chinese arms—well above the $207 million it purchased from the U.S. during the same period.

Myanmar, which accounted for around 5% of China’s exports over the past decade, has become the fourth-largest customer. Once subject to a Western arms embargo, Myanmar began expanding foreign procurement after sanctions eased in 2010. China seized the opening, offering comprehensive arms sales to meet its growing defense needs.

Further up the list, Algeria stands out as China’s third-largest customer—and a reminder that Africa has become a critical battleground in the global arms trade. Between 2020 and 2024, Russia supplied about 21% of Africa’s weapons imports, China 18%, and the U.S. 16%. Algeria is the continent’s heavyweight buyer, at one point accounting for more than one-third of Africa’s total arms imports. Historically, over 60% of Algeria’s purchases came from Russia. But as the war in Ukraine consumed Russia’s defense industry, supply gaps opened. China quickly moved to fill them, and today it has become Algeria’s second-largest supplier, with a 14% market share.

Nigeria, ranked eighth, highlights China’s growing footprint in Africa. Between 2020 and 2024, China was the largest arms supplier in West Africa, with a 26% share, followed by France (14%), Russia and Turkey (11% each), and the U.S. (4.6%). But China’s role in Africa goes beyond sales figures.

Africa’s political landscape is defined by instability and recurring civil wars, where holding power often requires genuine military strength. Against this backdrop, military education has long been a pillar of China’s influence. The Nanjing Army Command College trained numerous African leaders, including Namibia’s former president Sam Nujoma, Tanzania’s former president Jakaya Kikwete, the Democratic Republic of Congo’s former president Laurent Kabila, Eritrea’s president Isaias Afwerki, and Guinea-Bissau’s former president João Vieira.

Many senior and mid-level officers from countries like Uganda and South Sudan have also studied at China’s Shijiazhuang Army Command College—now the Joint Operations College of the National Defense University—including both current and retired officers. Don’t underestimate these African officers trained in Chinese military academies; they could very well be the children of their country’s ruling elite, a general, or even a future president

In sum, China’s arms exports to Africa benefit from competitive pricing and flexible financing, reinforced by its deep economic and infrastructure ties across the continent. Just as important, Chinese weapons come with no political strings attached and a policy of non-interference—an approach that resonates strongly in politically sensitive West African states. By contrast, U.S. and Western suppliers often impose conditions linked to political reform and human rights, while charging higher premiums. Russia, meanwhile, is consumed by the war in Ukraine. Against this backdrop, China’s long-term positioning in Africa is beginning to translate into a clear competitive advantage.

In fact, as a ‘latecomer’ in the arms market, China has been finding opportunities to expand amid territories long dominated by the U.S. and Russia. International observers often mock that we only export to small countries—but isn’t that obvious? The two major powers have already divided up the big markets, and which weapons a country buys is primarily a political signal of alignment. The markets we can penetrate now are basically what we can extract from the pockets of these two big players—a gift of timing, geography, and strategy.

A broader point is that the focus isn’t just on who we sell to, but on actually getting our weapons into the field so they can be demonstrated in action. While we don’t fight proxy wars, seeing our equipment used in real operations is crucial. Publicly reported examples—like Egypt’s counterterrorism operations in Sinai (March 2017), the Iraqi government’s strikes against ISIS (2014), the Yemen Civil War (2016–2021), and the Libya Civil War (2019–2021)—all show Chinese drones in action. These smaller conflicts, which didn’t make headlines, verified the practical utility of our drones, gradually earning international acceptance and purchases. After all, people were initially skeptical of Chinese equipment, and this year Pakistan helped trigger a new surge of interest in our products.

Subscribe now

China’s “Deepseek Moment” in the global arms trade

This moment can be called a ‘Deepseek’ moment for China’s arms trade—In the early hours of May 7, India and Pakistan fought an aerial battle lasting more than an hour. It was India’s deepest strike into Pakistani territory since the 1971 war. Despite India’s superior overall strength and its French-built 4.5-generation fighters, Pakistan’s air force—equipped with China’s J-10CE, the export version of the J-10C—emerged with a decisive advantage. Within minutes, the J-10CE shot down incoming Indian jets “like a phantom,” sparking a flood of commentary praising the aircraft’s prowess.

“Afterward, many people started saying how impressive our J-10 is, claiming: ‘Even though it’s older equipment, it’s so powerful, effortlessly taking down India’s French fighters and showcasing China’s might.’ But that’s not quite right. Let’s skip the hype and look at the facts.

The full picture is this: the J-10CE, together with PL-15 beyond-visual-range missiles and supported by the ZDK-03 early warning aircraft, forms a three-dimensional combat system covering early warning, targeting, and strike. It establishes an integrated land-and-air defense zone with roughly a 200-kilometer radius. You can think of the J-10CE as a semi-stealth missile launcher with its radar turned off—the missiles themselves are guided by the early warning aircraft, striking targets beyond visual range unexpectedly and significantly reducing the risk of enemy detection. That’s why I describe it as acting like a phantom.

Reports on the number of downed aircraft vary. Pakistan claimed it had shot down five: three French Rafales, one Russian-built Su-30MKI, and one MiG-29UPG. The tally may not seem high, but in a contest of systems against systems, the result was decisive. From the Indian perspective, their fighters were struck “before they even realized it,” leaving them in a state of confusion. Some accounts suggest India’s Rafales were primarily configured for ground-attack missions and not fully prepared for air combat. More critically, India’s air force operates a patchwork of Russian, French, and U.S. equipment, leading to poor data-link integration and a systemic disadvantage. Had Pakistan pressed further, it might have inflicted greater losses.

But warfare isn’t just about fighting; there are also political and social considerations. I’m not a current affairs commentator and won’t discuss India-Pakistan or China-U.S.-Russia relations. I just want to highlight that this demonstrates the practical combat capability of China’s ‘detect-and-strike’ system. During those days of sustained operations, the successful interceptions by our HQ-9 air defense system further showcased breakthroughs in China’s system-level warfare capabilities.

For years, China’s defense exports were dismissed by many major importers—partly for political reasons, and partly because of an enduring perception that Chinese systems were cheap and low quality. The recent event, however, was a turning point—a Deepseek moment—demonstrating that China can deliver not only advanced hardware but also a full-spectrum military solution. Purchasing the system effectively creates a 200-kilometer defensive bubble: within that radius, anything that enters risks being destroyed.

Unsurprisingly, the event drew global attention. Within a month of the clash, Indonesia’s deputy defense minister publicly confirmed the country was evaluating procurement of the J-10. The irony is striking: Indonesia had already signed a deal for 42 French Rafales and was in talks for U.S. fighters. Yet after watching Rafales shot down, it was natural for Jakarta to consider China’s system—more capable and more cost-effective—as an alternative.

According to Spain’s Defensa News, Colombia is also evaluating a potential switch to China’s J-10, with reports citing a base unit price of around $40 million and concessional financing on offer. China has already tabled a formal proposal for the sale of at least 24 J-10CE fighters. In the same month, Bangladesh was also reported to be advancing plans to procure 16 J-10CEs.

At roughly $40 million apiece, a 24-aircraft package approaches $1 billion in value—and that figure covers only the jets themselves, excluding associated systems and support. The case highlights how fighter aircraft, sold together with their integrated operational packages, have become China’s most commercially viable and premium-priced defense platform.

What does China sell the most?

So which category of Chinese equipment sells the most? Many might guess missiles—thanks to jokes about “missiles sprouting in the desert”—or perhaps drones, often portrayed as China’s high-volume, cost-effective exports. In reality, the answer is engines. Over the past decade, engines accounted for 51% of China’s arms export value, followed by aircraft platforms at 29%. Missiles and air-defense systems each made up just 6%, with all other categories barely registering.

For missiles, the leader is not China but the United States. In fact, aircraft are also the top source of U.S. export revenue, representing 58% of its sales, with missiles in second place at 19%.

Who buys America’s weapons?

Since we’ve mentioned the U.S., it’s worth pulling their numbers too. After all, critics often point out that China’s defense industry relies too heavily on Pakistan—a situation that in investment banking terms would be called “overexposure to a single client.” The concern is valid, but the U.S. faces a similar pattern.

America’s largest customer is Saudi Arabia, which accounted for 17% of its arms exports over the past decade. Other buyers—such as Australia, Japan, Qatar, and South Korea—pale in comparison, each representing only a fraction of Saudi Arabia’s purchases. The difference is that Washington’s top client share is not as disproportionately high as Beijing’s.

Between 2015 and 2024, the largest recipient of U.S. arms exports was Saudi Arabia (17.0%), followed by Australia (7.7%), Japan (7.3%), Qatar (6.0%), South Korea (5.4%), and Ukraine (5.1%). Other major destinations included the United Kingdom (4.3%), the United Arab Emirates (4.2%), Israel (3.5%), and Norway (2.9%). Together, these top ten markets accounted for a significant share of total U.S. arms exports over the period.

Where do China and the U.S. import their weapons from?

What about the other side of the equation—do China and the U.S. themselves import weapons, and if so, from where?

For the U.S., imports are negligible, a tiny fraction of its already massive exports. Its purchases mainly come from European partners such as the U.K., France, the Netherlands, and Germany.

Between 2015 and 2024, the leading suppliers of U.S. arms imports were the United Kingdom (18.0%, USD 1,302 million), France (13.0%, USD 911 million), the Netherlands (12.0%, USD 851 million), and Germany (12.0%, USD 823 million). Other major sources included Israel (10.0%, USD 726 million), Italy (7.0%, USD 499 million), Norway (5.6%, USD 399 million), Sweden (5.0%, USD 356 million), Australia (4.2%, USD 300 million), and Jordan (3.8%, USD 271 million).

China’s pattern looks different. Russia has been its largest supplier, followed by France, which primarily provides high-end components such as aircraft engines. The third source may be more surprising: Ukraine.

Between 2015 and 2024, the primary supplier of China’s arms imports was Russia (74.0%, USD 7,277 million), followed by France (11.0%, USD 1,110 million) and Ukraine (8.4%, USD 833 million). Other notable suppliers included Switzerland (2.0%, USD 195 million), the United Kingdom (1.9%, USD 190 million), Germany (1.8%, USD 176 million), and Uzbekistan (1.0%, USD 102 million).

China-Ukraine relations have historically been decent, with Ukraine serving as a crucial partner for China’s defense industry thanks to its inherited Soviet-era heavy industrial base. Naval propulsion was particularly critical: China imported Ukrainian gas turbines such as the UGT-25000 to power its 051C and 052C/D destroyers, filling a gap in domestic large-scale marine engines. In aviation, Ukraine’s Motor Sich supplied jet engines for the L-15 advanced trainer and provided helicopter engine support. The most emblematic case was in 1998, when China purchased the unfinished Soviet aircraft carrier Varyag from Ukraine. After years of modification, it eventually became China’s first aircraft carrier, the Liaoning. Other transfers included design data for the An-225 transport aircraft and technology related to air-to-air missiles.

At the time, Ukrainian arms sales to China carried fewer political restrictions, while Ukraine still retained significant technical strengths in fields China urgently needed, such as propulsion and shipbuilding. Meanwhile, Ukraine’s post-Soviet economy was chronically weak and in desperate need of cash. Once an industrial powerhouse within the Soviet system, it found itself adrift after the collapse, effectively selling off assets to survive. For China, securing this “great-power inheritance” at relatively low cost proved a bargain—allowing it to rapidly close critical gaps in aircraft carriers, marine engines, and related technologies.

Subscribe now

Targeted imports, domestic breakthroughs

After 2010, China’s approach to weapons imports shifted toward small-batch, specialized procurement—purchases aimed less at filling arsenals and more at acquiring critical technologies. Buying fighter aircraft, for instance, was often about studying their engines; importing an air-defense system also meant dissecting its radar technology. China has continued to import certain core subsystems abroad to bridge gaps in domestic manufacturing.

Today, the balance has flipped: China’s exports are moving up the value chain, from low-cost, high-volume equipment to advanced, high-margin systems. In 2023, six Chinese companies ranked among the world’s top 20 defense firms, with the highest reaching No. 8. The global league tables are now dominated almost entirely by U.S. and Chinese firms, with only occasional entries from Russia and Europe.

One important distinction remains. For leading U.S. defense contractors, more than two-thirds—and in some cases over 90%—of revenue comes from weapons sales. By contrast, for Chinese firms the figure is far lower, typically 20–30%.

The relationship between U.S. defense contractors and the government is unusually tight. These companies depend heavily on government contracts, while US officials often rely on the industry for financial and career incentives. Legislators push policy and budgets; defense agencies define requirements and procurement processes.

Companies, through lobbying and personnel movement—especially the U.S.’s unique ‘revolving door’ practice—ensure contract awards and ongoing purchases, even involving key corporate personnel in policy-making itself.

Much of this cannot be spelled out openly, yet over decades it has built a stable ecosystem: America’s second-largest capital complex after finance, one sustained by constant conflict abroad. Under this model of “perpetual war,” it is no surprise that more than 90% of U.S. defense firms’ reported profits stem from weapons sales.

China’s defense industry, by contrast, is constrained by international politics but structured differently at home. Through a strategy of military-civil fusion, most Chinese defense companies also develop civilian applications—in electronics, machinery, and related fields—giving them more diversified revenue streams.

Although China’s defense industry has developed rapidly, its technological level still lags far behind that of the U.S., and the profit margins on Chinese equipment are much lower than those of American defense firms. Winning the current market through cost-effectiveness, lack of political strings, and long-term strategic planning is already a significant achievement.

The recent “Deepseek moment” may mark a turning point, sparking a new wave of demand. While the overall scale of our arms trade still falls far short of challenging the U.S., the limited markets we have tapped provide more Chinese equipment with real-world operational opportunities. In today’s volatile global environment, the test is whether Chinese equipment can move beyond its old reputation for being “cheaply made” to one of being “affordable and reliable.”

Subscribe now

Related Reads

Baiguan Youtube

This past weekend, I went to see the 1st World Humanoid Robot Games held in Beijing.

The event welcomed 279 teams from 16 countries, with more than 500 humanoid robot “athletes.” Among them were China’s domestic “superstars,” such as Unitree (the Hangzhou startup whose robots danced at China’s Spring Festival Gala) and Tiangong (the robot that won first place in the humanoid robot marathon back in April, also held in Beijing).

Today, I want to share some of the best (and funniest) moments! (Some of the GIFs were taken by me at Sunday’s event, and others were kindly created by 量子位, an excellent publication on WeChat that focused on AI and technology trends.)

It was a 3-day competition featuring a wide range of events: 3v3 football, 5v5 football, kickboxing, and track-and-field sports such as high jump, long jump, 4x100 relay, 100m/400m dash, 1500m run, obstacle race, and even gymnastics—the setup was very much like a human Olympics.

Unitree won a total of 7 gold medals, including the 100m obstacle race, which tested humanoid robots’ ability to run, jump, balance, and identify barriers.

“奥林1队 (Olympics Team 1)” won the kickboxing final, which was actually remote-controlled by award-winning athlete Xue Ming, who played for China's national volleyball team.

The much-anticipated 5v5 soccer tournament was won by Tsinghua University. All teams used Booster Robotics T1 hardware, competing by optimizing algorithms. Booster Robotics—founded in 2023, also by Tsinghua alumni—designed humanoid robots specifically for developers and researchers. The T1 model became the only humanoid robot showcased at NVIDIA’s Spring Festival event.

The soccer matches were fully autonomous.

Subscribe now

A highlight: the game rules encouraged autonomous control. For instance, in track and field, robots got a 20% reduction in running time if they were fully autonomous. Tiangong Ultra won the 100m dash under full autonomous mode (the team engineer had to chase it hard to keep up!).

But of course, penalties were applied to those that “hit and run.”

Another highlight: beyond competitive sports, the Games also featured real-world scenario competitions—from hotel service and cleaning to medicine sorting and industrial material handling. These tasks required both accuracy and speed.

For instance, the medicine-sorting task required the robot to first read the prescription (fortunately, it was computer-printed, not handwritten by doctors) and then retrieve the correct medicine from the shelf.

The Games welcomed 192 university teams and 88 enterprise teams, with prize money of 10,000, 30,000, and 50,000 RMB awarded to the winners of scenario-based competitions. Teams used a variety of hardware, but impressively, all robots competing were autonomous.

There were plenty of funny moments. Everybody had a good laugh.

“Sir falls a lot”

“Sir hit my engineer”

“Human ambulance”

A robot (funded by Chia Tai Group) that took forever to finish the 4x100m relay race—but when it finally crossed the finish line, the whole audience clapped.

“Cutie McSmarty Bots” getting their medals

Robots hip-hop dancing,

walking in a fashion show (with clothes designed by local students from Beijing),

and even performing a Terracotta Army drum show.

Because it’s summer break in China, the audience was full of kids. I often overheard parents telling them: “Fun, right? Study hard and grow up to be like these engineers!” Even during vacation, Chinese parents never miss the chance for a “gentle push”—but I liked the fact that in China’s culture, engineering and tech aren’t considered nerdy—they’re seen as something worth celebrating with a fun event like this! (At one point, the commentator invited the entire audience to chant "Hello, world" :) )

Subscribe now

Robots fell a lot in the competition—but that’s half the fun! I really do hope to see this turn into an annual (or even Olympic-style) global event. The next decades will be the race of brilliant minds—What could be cooler than a worldwide showcase of such progress?

Some people complained that many of the competitions were “useless” (like dancing robots). But I think it’s awesome that China is finally encouraging these so-called “useless” pursuits by hosting a full-fledged Games. Innovation is born from generous funding, relentless trial and error, and—equally important—the curiosity and joy of doing many "useless" things just for fun. Great inventions can't be planned!

Below are the winners of the major competitions and their corresponding hardware (chart organized by the Goldman Sachs team).

And if you’d like to further understand the investment implications behind the humanoid robot value chain, you can read my previous newsletter on the topic:

I hope you enjoyed this newsletter as much as I enjoyed watching the Games!

Related Reads

Baiguan Podcast

Baiguan Youtube