Author: David Weng , who built a red pavilion to entertain guests.
In this recent annual report season, home appliance companies were among the few in many industries that dared to provide a detailed account of the current operating conditions faced by Chinese companies:
"In 2025, the US will impose substantial tariffs on Chinese home appliance exports. High-end Chinese home appliances exported to the US will need to reconfigure their supply chains. Regional protectionism and non-trade barriers will compound the already complex business environment for Chinese companies going global. Furthermore, the Russia-Ukraine war continues, adding another layer of crisis to the Red Sea." — Haier Smart Home Annual Report
"The storms of 2025 are more intense than anticipated, with constantly changing tariffs, global adjustments to production capacity, the unprecedented impact of the AI technology wave, and the continuous evolution and escalation of geopolitical conflicts..." — Midea Group Annual Report
Of course, the reason why the home appliance industry dares to openly write about these difficulties in its annual report is because, as the first batch of Chinese companies to go overseas and establish themselves globally, they have both the confidence and the strength to overcome the difficulties.
In fact, Midea immediately followed up with: "...Nevertheless, after weathering the storm, Midea still achieved a record high in performance."
Even its annual report this year was titled: "We always have a way."
Other industries were not so "lucky".
For example, Shenzhou International, a giant in the textile industry, devoted a lot of space to describing its problems in its annual report, but ultimately failed to achieve a turnaround.
"Currently, the global textile and apparel industry is facing multiple pressures and challenges, mainly manifested in weak growth in end-user demand, fierce market price competition, continuous rise in production costs, increased volatility in raw material prices, tense geopolitical situation, and a complex and volatile trade environment."
The industry as a whole is showing a pattern of "strong supply and weak demand". Domestic production capacity is squeezed by multiple factors such as weak recovery of domestic demand, transfer of overseas demand and rising overall costs. Some companies have adopted the business model of exchanging price for volume, and production capacity is showing a continuous trend of shifting to Southeast Asia and other regions.
At the same time, rising labor costs and fluctuations in the RMB-USD exchange rate have increased the exchange rate risk for export companies, further squeezing the industry's profit margins.
Geopolitical conflicts in the Middle East have led to significant fluctuations in international crude oil prices, which have not only increased the cost of raw materials for chemical fibers but also caused longer logistics cycles and higher transportation costs.
Uncertainty surrounding U.S. tariff policy and increased scrutiny of entrepot trade are putting sustained pressure on the market share of Chinese textile and apparel products in the U.S. — Shenzhou International Annual Report
Supply and demand imbalance, rising labor costs, exchange rate fluctuations, geopolitics, and tariff policies—these five major challenges weigh heavily on this industry leader, and will naturally also weigh on many other companies without a voice in the market.
However, they may be busy rushing to get things done, or they may feel that revealing this in their annual reports would only increase investors' concerns, so most of them have chosen to remain silent.
Of course, on the other end of the industry chain, upstream resource companies and the AI industry chain are showing a booming scene.
"Global demand for metals mining is undergoing a structural revolution, with the energy transition and the rigid pull of the computing power era becoming the 'super incremental' demand in this mining cycle." — Zijin Mining Annual Report
"Ultra-large-scale AI computing power clusters bring about a massive demand for high-speed optical modules." — Easun Annual Report
Even Haomai Technology, which started by making tire molds and gas turbine parts and seems to have nothing to do with AI, said:
"Recently, the world's three major gas turbine manufacturers released data: GE Vernova's gas turbine production capacity has been sold out until 2028, with only 10% remaining in 2029; Siemens Energy sold 194 gas turbines in 2025, almost double the number in 2024; Mitsubishi Heavy Industries plans to double its gas turbine production capacity in the next two years... The explosive growth of AI data centers remains the core driving force of this gas turbine demand revolution."
So what exactly is happening in China's industrial sector?
In last year's annual reports, most entrepreneurs discussed "how to weather the next few years," so the key word I used in this article is evolution. Evolution is action; it's about how companies move.
The keyword I want to use this year is restructuring. Restructuring is about cognition; it's about how a company thinks.
After reading over 300 annual reports this year, I realized that entrepreneurs are no longer content with simply "getting by." Some of them are beginning to articulate their previously vague judgments in clear language, attempting to describe this new world.
This event itself is a signal.
one
The first step in restructuring is the macro level. A relatively accurate definition of the current macro and industry state comes from Midea Group's annual report, which is also quite familiar to us—"K-shaped differentiation".
"...The tide of the times is rolling on, and we are all in it. The K-shaped differentiation is accelerating the reshaping of the fate of nations, industries, enterprises, and individuals." — Midea Group Annual Report
The term "K-shaped recovery" has received particular attention in recent years. Initially, scholars used it to describe the vastly different recovery paths of different industries in the United States after the pandemic. Later, it was used more in the field of distribution to describe the significant divergence in income and wealth growth trends among different social classes over the past few years.
At the industry level in China, this year's K-shaped divergence has reached an astonishing level.
The most obvious next step in the K series still falls on the real estate industry. Vanke, which has completely fallen from grace, recorded a loss of nearly 100 billion RMB in 2025, and described its situation in its annual report as a "what's the use" failure:
"Due to various factors, the risks have not been completely resolved, and the business development still faces severe challenges... At the same time, the Group's liquidity pressure has further increased due to the concentrated repayment of publicly disclosed debt."
At the same time, it also honestly laid out the reasons for the losses for us:
"(1) The settlement scale of real estate development projects has decreased significantly, and the gross profit margin remains low. During the reporting period, the settlement profit of real estate development business mainly corresponds to the projects sold in 2023 and 2024 and the inventory of existing and near-existing houses to be digested in 2025. The land acquisition cost of these projects is relatively high, and the sales performance and gross profit margin are lower than the investment expectations, resulting in a significant decrease in the total settlement gross profit during the reporting period."
(2) Due to the increase in business risk exposure, new provisions for credit impairment and asset impairment were made.
(3) Some operating businesses incurred overall losses after deducting depreciation and amortization, as well as some non-core financial investments incurred losses.
(4) The prices of some large-scale asset transactions and equity transactions were lower than their book value. — Vanke A Annual Report
It's a classic case of everyone kicking someone when they're down. If you can't make money selling houses, you can blame it on paying too high land prices in previous years. But if your business operations are losing money, your financial investments are losing money, and even your assets are being sold off at below book value, then who's to blame?
The real estate market is tough, and companies in the supply chain are naturally not faring much better. For example, Red Star Macalline, the king of home improvement and building materials markets, also suffered huge losses in 2025, stating in its annual report:
"...Affected by the continued downturn in the real estate industry and the decline in demand in the home furnishing and building materials industry, the demand in the home furnishing retail market has weakened. The company has continued to retain customers by reducing or waiving rent and management income. At the same time, it has actively adjusted its strategy and product category layout to attract high-quality business formats and brands with preferential business conditions. In the early stage of expansion, it has provided rent and management fee discounts, which has had a significant impact on rental and management income, and the rent level has decreased significantly compared with previous years."
But in fact, its losses were not solely due to a decline in rental income; the more than 20 billion yuan in losses were primarily due to the following reason:
"Due to a shift in market expectations for future rental prices, the estimate of a short-term recovery in rental prices needs to be adjusted. Therefore, the company has adjusted its expected future rental income accordingly, resulting in a significant decline in the value of investment properties. The company's total fair value change loss for 2025 amounted to RMB 23.44 billion." — Meikailong Annual Report
That's how the economy works. During an industry upswing, any asset a company possesses can command a high valuation, whether it's real estate, equity, or even accounts receivable. So it doesn't just appear as a period of prosperity, but rather as a period of rapid growth.
During a downturn, not only will the main business suffer losses, but the assets on hand will also shrink, making business contraction and systemic losses inevitable.
The sluggish real estate market is also reflected in the shopping malls that accompany residential buildings. For example, the MTR Corporation seemingly mentioned it only briefly in its annual report:
"In February 2026, the Shenzhen 'Songhui' shopping mall project, with a total floor area of approximately 10,000 square meters, failed to be successfully tendered for sale."
I checked, and The Grand Central Mall is MTR's first flagship commercial project in mainland China. The residential units above it were almost all sold out seven or eight years ago. Now the company wants to sell the ground floor shops all at once, but the auction failed.
Meanwhile, local governments are also anxious about not being able to sell their assets. Although the debt reduction efforts that began at the end of 2024 reversed some market sentiment, I still see the delayed effects of debt reduction in the annual reports of Zhejiang Merchants Shanghai-Hangzhou-Ningbo REITs, and even the seemingly unrelated highway industry may be affected.
In short, for many years, the section of the Hangzhou-Hui Expressway in Lin'an District of Hangzhou has been toll-free for passenger vehicles with Zhejiang A and Zhejiang M license plates that use ETC cards. This cost was actually paid by the Lin'an District People's Government, with the aim of "improving the efficiency of the road network, alleviating urban traffic congestion, reducing travel costs, and promoting economic development."
The amount is not small each year, reaching 135 million yuan in 2025, accounting for 19.05% of the total income of REITs asset portfolios.
According to the annual report of Shanghai-Hangzhou-Ningbo REITs, "Lin'an District's general budget public revenue in 2025 was 8.542 billion yuan, a year-on-year decrease of 10.5%, and fiscal revenue is facing certain pressure."
The 2024 annual report states in the same location: "The GDP of Lin'an District, Hangzhou City in 2024 was 73.29 billion yuan, and the economic development was in good condition."
Of course, the next stroke of the K-shaped curve includes not only the real estate industry chain, but also industries that are relatively sensitive to the price increases of upstream resources, while downstream competition is so fierce that the price increases cannot be passed on.
Among them, I unexpectedly saw Huichuan Technology, which has been touted as a model company for the past few years. Its annual report stated:
"The company's main upstream raw materials include copper, aluminum, silicon steel, and power semiconductors. Rising prices of commodities and power semiconductors will negatively impact the company's profitability… Increased industry competition is leading OEMs to pass on cost pressures upstream; the increased self-manufacturing ratio by OEMs will intensify competition among third-party powertrain suppliers. If industry competition further intensifies, the company's products will face the risk of price reductions, thus negatively impacting the company's performance." — Huichuan Technology Annual Report
After careful comparison, the risk warning text mentioned above did not appear in the 2024 annual report. However, in the recently released Q1 2026 report, it delivered a performance showing revenue growth but declining profits, a case of increased revenue but decreased profit.
AVIC Optoelectronic, a leading company in the defense sector, also wrote in its annual report:
"...The complex and volatile external environment and the prominent cyclical fluctuations in the industry have led to a temporary slowdown in demand in the defense sector. Meanwhile, the prices of precious metals such as gold, copper, and silver, as well as other commodities, have continued to rise, putting considerable pressure on the company's profitability." — AVIC Optoelectronics Annual Report
If even leading companies in the industrial control and military sectors are experiencing such difficulties, it's not hard to imagine how tough it must be for smaller manufacturing companies.
Beyond the manufacturing sector, the plight of other SMEs can perhaps be glimpsed from the financial reports of industrial park REITs. In fact, as a whole, nearly 90% of industrial park REITs experienced a year-on-year decline in revenue for the entire year of 2025 and the first quarter of 2026, and the proportion with an even larger quarter-on-quarter decline in the first quarter reached 75%.
"Affected by both the macroeconomic environment and changes in market supply and demand, competition in the surrounding market intensified, resulting in a year-on-year decrease in the average effective rental price per unit for the Guangming Accelerator Phase II project and Wanrong Building; the occupancy rates of Wanrong Building and Wanhai Building also decreased year-on-year." — Bosera Shekou Industrial Park REIT Q1 2026 Report
"The regional industrial park market is under overall pressure, with intensified competition. Existing and newly added R&D office properties are creating significant competitive pressure on the project; a certain internet company group did not renew its lease after its various business segments merged their offices, and another online education company did not renew its lease upon expiration due to business transformation." — CICC Hubei Science and Technology Investment Optics Valley REIT Q1 2026 Report
two
Since it's a K-shaped divergence, there's always an upward trend. Moreover, in a stagnant economy, upward movement often comes at the cost of downward movement.
The most typical contrast comes from upstream resource-based enterprises. While companies like Huichuan are troubled by rising raw material prices, Zijin Mining, which holds a large amount of mineral resources, is issuing zero-coupon convertible bonds with negative yields.
"In January 2026, the company successfully issued US$1.5 billion in zero-coupon H-share convertible bonds with a negative yield. The final conversion premium of 40% is the highest ever achieved by mining convertible bonds in the Asia-Pacific region, fully demonstrating the international capital market's high recognition of the company's market value." — Zijin Mining Annual Report
Let me briefly explain just how amazing this financing method is.
First, Zijin borrowed $1.5 billion. Because it was a zero-coupon bond, it didn't have to pay any interest when repaying the loan; it only had to repay the principal.
Secondly, there is a negative yield. That is to say, assuming the face value is 100 yuan, the company sells it at a price of 102 yuan. Not only does it not pay interest, but it only needs to repay 100 yuan when it matures.
Of course, investors aren't stupid. The reason they're willing to accept "at-the-money" loans is because they have no intention of getting their principal back; they're after the equity conversion rights.
Thirdly, the conversion premium is as high as 40%. That means, assuming the initial offering price is 40 yuan, the conversion price would be 56 yuan. Therefore, investors can only profit by exercising their conversion rights if the share price rises above 56 yuan in the future.
Regardless of whether investors will ultimately make money, this at least means that Zijin Mining's financing costs are negative at this stage—not only is there no interest, but it also earns a premium for free.
This is no longer a question of whether the capital market approves or not; it's practically investors chasing after them, feeding them money.
Take, for example, the consumer retail sector. On one hand, even shopping malls built above Shenzhen subway stations are struggling to sell, while on the other hand, new business models and channels are emerging one after another. Yunnan Baiyao, in its description of its toothpaste business segment, states:
"The offline sales structure is changing. The proportion of traditional channels—hypermarkets and convenience stores—is shrinking, while emerging channels—warehouse membership stores, snack concept stores, and discount stores—are growing against the trend, becoming new engines for offline growth." —Yunnan Baiyao Annual Report
This is corroborated by the collective rise of snack retail groups. Wancheng Group's annual report states:
"During the reporting period, the company actively promoted its store network expansion plan... In terms of stores that can be included in the operating cycle, the company added 4,720 stores and reduced 602 stores, with a total of 18,314 stores at the end of the period."
Opening 4,000 stores in a year is a speed that is hard to imagine in today's mainstream retail industry.
Meanwhile, another snack giant, Mingming Very Busy, boasts a county-level coverage rate:
"As of December 31, 2025, the Group has established a store network of 21,948 stores, covering 30 provinces and all city tiers in China. Approximately 60.0% of the Group's stores are located in county towns and townships... covering 1,401 counties, with a coverage rate of approximately 75.0% in all county towns in China."
Therefore, while investors may be unwilling to spend money to buy a physical shopping mall, thousands of franchisees are willing to spend real money to allow snack giants to easily acquire tens of thousands of stores.
In addition, the direction of government spending is also quietly changing.
CAR-T therapy, a leading domestic innovative drug developer, is delighted with the new forms of government support. According to its annual report:
"Our company, through its indirect wholly-owned subsidiary Shanghai Kaixing Diagnostic Technology Co., Ltd., signed a strategic cooperation agreement with Shanghai Jingong Enterprise Development Co., Ltd., an important platform enterprise in the Bay Area High-tech Zone of Jinshan District, Shanghai. The total investment will not exceed RMB 37 million, and an advanced commercial production base for CAR-T cell products will be built in Jinshan District, Shanghai."
I looked into the specific operational model. Simply put, the 370 million yuan investment for building this new CAR-T drug production base isn't coming from the company now. Instead, the government will first handle the renovation of the factory and the purchase of facilities. After delivery, the company will pay a quarterly service fee (not exceeding 10 million yuan annually for the first 10 years, and not exceeding 18 million yuan annually from the 11th to the 20th year), and will repurchase the facilities within 20 years according to the agreement. The repurchase price will be determined based on the final investment amount, with a maximum of 370 million yuan.
In this way, as the company further stated in its annual report:
"In its early stages, the company did not need to make large capital expenditures, effectively preserving valuable cash flow for core R&D and market expansion. At the same time, the buyback mechanism ensures that the company can fully acquire asset control after long-term operation, maintaining production stability while enhancing the flexibility of asset allocation." — CARsgen Therapeutics Annual Report
This arrangement, where local governments build first and companies buy back the project within 20 years, essentially applies the old infrastructure logic to the biopharmaceutical industry. However, nowadays, not every industry can enjoy such treatment.
Therefore, I have always believed that the key to K-line divergence is not the difference in the operating states of the upper and lower sides, but the fate of complete decoupling behind them. With the support of the market, capital, and even policies, the good will get better, and the bad will get worse.
Who decides on this support? Ultimately, it comes down to whether it aligns with the new logic of the new era.
So, what is the new logic?
three
The first phrase in the "new logic" comes from Zijin Mining. As a new member aiming to enter the trillion-yuan market capitalization club by 2025, it states the following in its annual report:
"Against the backdrop of intensifying global strategic competition, key minerals have risen from mere economic factors to the focus of national security and great power rivalry... The ability to control the entire supply chain of key minerals is becoming a new yardstick for measuring a great power's right to development and industrial dominance."
This geopolitically driven competition for resources is accelerating the transformation of the global mining industry from a globalized division of labor to a more fragmented and compartmentalized system, triggering a profound structural change.
The profound transformation of the global resource governance landscape is a systemic restructuring driven by a shift from an "efficiency logic" to a "security logic." The traditional linear value chain based on globalized division of labor—"mining—exporting—overseas processing—end-product manufacturing"—is disintegrating, replaced by a new ecosystem driven by regional closed loops, localization, and green compliance. —Zijin Mining Annual Report
In fact, "security logic" is not only a term in corporate annual reports, but also a key word in the 15th Five-Year Plan.
When I returned to China in March, a friend who works in primary market investment told me that all projects related to security, from commercial aerospace to semiconductor chips, are being snapped up. A couple of days ago, I listened to the "High Energy" podcast and heard Uncle Feng say that the new National Integrated Circuit Industry Investment Fund is already investing in early-stage and small-scale projects with valuations up to 500 million yuan.
Of course, the logic of security isn't limited to future-oriented industries; it's also reflected in upstream resource sectors. With the outbreak of the US-Iran conflict, it's conceivable that this logic will be further strengthened, and competition among nations for resources and even supply chains will become increasingly fierce.
For example, SF Holding stated in its annual report:
"In an international environment characterized by escalating geopolitical conflicts, rising trade barriers, and complex multilateral relations, the global supply chain is undergoing a profound systemic reshaping, with security, resilience, and efficiency becoming core considerations for the global industrial chain layout." — SF Holding Annual Report
Safety first, resilience second, efficiency third.
The second term in the new logic comes from Haier Smart Home, and it appears after the passage about the business environment quoted at the beginning of the article:
"Looking ahead to 2026, we see that the original trade paradigm has disappeared. Tariff games have prevented the efficient integration of supply chain layouts. Geopolitical conflicts have driven considerations of supply chain security, disrupting the normal flow of supply chains. All of these factors will lead to a decline in return on investment, the elimination of less competitive companies, and a reduction in market supply, thus bringing about a new balance." — Haier Smart Home Annual Report
If "security logic" is a new logic from the perspective of the nation and the industrial chain, then "trade paradigm" is a new logic that companies that want to survive in the wave of going global must understand.
How is the new trade paradigm being restructured?
At least one of these key terms was pointed out by Universal Scientific Industrial (USI), a company in the even more competitive electronics manufacturing industry: friend shoring.
"Electronics manufacturing services providers are facing transformation. Geopolitical factors and the trend of global trade regionalization are impacting the restructuring of global supply chains. To diversify their supply chains and manage risks, customers are shifting some offshore outsourcing to nearshore or partner offshore outsourcing, leading to rapid changes in demand and orders." — Universal Scientific Industrial (USI) Annual Report
Clearly, globalization may not be completely dead yet, but everyone is starting to be selective about their friends.
The third term in the new logic comes from Foxconn Industrial Internet (FII), a company that went public as a contract manufacturer but has now transformed into a global leader in AI servers. In his letter to shareholders, the chairman wrote:
"The widespread application of generative AI has rapidly shifted computing power demands from 'general-purpose computing' to 'high-performance computing.' AI computing infrastructure has become the core foundation for driving the development of the digital economy, and the global technology industry has thus entered a new growth cycle." — Foxconn Industrial Internet Annual Report
Unlike the US, which places more emphasis on the narrative of "soft AI" giants like OpenAI and Anthropic, as well as AI capital expenditures in the secondary market, Chinese investors are currently more focused on "hard AI," such as "AI computing infrastructure."
Whether it's optical module companies "standing in the light", AI chip industry chain which surged last year led by "Hanwang" and has been relatively low-key this year but whose stock prices have actually returned to their highs, or even the power equipment sector which has the highest average increase in the industry this year, they all fall into this category.
Interestingly, in the face of absolute competition in the technological revolution, security logic and trade paradigms can still be set aside; efficiency, quality, and capability are what matter most in this field.
This is precisely why a large number of Chinese companies can still benefit from this wave of US AI infrastructure development, and some, like Foxconn Industrial Internet, can even directly jump from the smartphone industry chain to the AI hardware industry chain.
Four
So what should companies that are not in the "AI computing infrastructure" industry chain do? In order to adapt to the first two new logics, I think it is appropriate to use a term that Li Lu, the founder of Himalaya Capital, recently used to talk about the growth history of BYD: "a global company with Chinese nationality".
He said this in the interview:
The concept of a multinational corporation originated in industrialized European countries and was later realized in the United States after World War II, where businesses could develop simultaneously around the world while being based in one country.
Now, at this juncture, Chinese companies are beginning to possess such opportunities, energy, and capabilities.
The most important condition that China offers these companies is that it is a huge, single, and large market. It has a single and unified regulatory system, and it is also an open market to the world.
These four conditions combined give the champion in this market the opportunity to become a champion globally, and it must be localized and become part of the local economic development, not just an exporter.
Therefore, this era will see the emergence of global companies based in China, global companies with Chinese nationals. BYD is an important ripple in this wave, but not the only one. This process has only just begun, and there is still a very long way to go.
His statement reveals two prerequisites for becoming a "Chinese-owned global company": one is to carve out a path to success in the unified Chinese market, and the other is to become part of the local economic development after going global.
The former is easy to understand, while the latter is increasingly becoming a must-choose option for companies going global under the two new logics of security and trade paradigm.
Take the biggest challenge facing Chinese companies going global in 2025—the tariff war—as an example. Although in hindsight, due to Taco and the counterattacks from China and other countries, the tariff war seemed to have more bark than bite and did not cause much disruption to economic operations, we can still get a glimpse of the harrowing experience at the time from many companies' annual reports.
For example, Giant Star Technology, a leading tool company whose sales in the Americas account for more than 60% of its total sales, stated in its annual report:
"In 2025, affected by the sharp fluctuations in the international trade environment, the company's overall production and delivery schedule was significantly impacted in the second quarter. Subsequently, overseas production capacity returned to normal, but domestic production capacity was still negatively affected by tariffs. At the same time, the additional tariff costs pushed up the industry's average selling price, resulting in a significant decline in industry sales."
Another example is Anker Innovations, whose overseas revenue accounts for more than 70% of its total revenue:
“Since 2025, global tariff policies have continued to be adjusted, trade frictions have intensified, and major economies have imposed tariffs on each other, putting pressure on the operating costs and market access of multinational companies.”
These are all companies I was worried about in my article last year, but in the end, it seems that their performance has generally achieved a smooth transition. Behind this, "becoming part of the local economic development" has played a significant role.
So, how do we become part of the local economic development? Or, in other words, how do we achieve true localization? I would like to break it down into three levels.
The first and most obvious layer is manufacturing localization. This year's annual reports from numerous companies expanding overseas discussed the activation and development of their overseas manufacturing bases, which I won't list here. However, Shenzhou International, which began establishing overseas production capacity in the last era, serves as a reminder to those who follow:
"The main reasons for the decline in gross profit margin during the year were: 1) an increase in the average salary of employees during the year, and the fact that the new garment factory in Cambodia was still in the process of improving its operational efficiency after production, which led to an increase in the proportion of labor costs in sales revenue; and 2) the impact of changes in import tariff policies on products sold to the US market, resulting in the sharing of some tariff costs with customers." — Shenzhou International Annual Report
Will overseas labor costs increase? Possibly. Will efficiency be lower than domestically? Most likely. Will tariffs affect overseas operations? Yes, they will.
Therefore, exporting production capacity does not guarantee success. If overseas labor is more expensive, efficiency is lower, and tariffs cannot be avoided, then what exactly does localization mean?
This brings us to the second level: the localization of supply chains and industrial clusters.
Last year, I introduced how SF Holding described its efforts to help companies go global in its annual report. This year, it has provided a more detailed explanation of the overseas expansion needs of certain industries, such as the tea beverage industry:
"In terms of catering brands going global, the company has successfully empowered many leading Chinese tea beverage brands, supporting the smooth opening and operation of hundreds of stores in overseas regions such as Singapore, Malaysia, Canada, Australia, the United Kingdom, and Germany."
The company provides training to its clients' domestic suppliers on product labeling and customs clearance compliance requirements in overseas destination countries to ensure that exported materials comply with local regulations; it integrates the export needs of multiple clients to improve container load rates and shorten delivery times; it provides end-to-end supply chain services from China to overseas destinations, including first-leg, trunk, and last-mile delivery; and it provides on-the-ground operational services such as warehousing and distribution in the destination countries. —SF Holding Annual Report
In fact, it is not only logistics and supply chain companies like SF Express that support Chinese companies in upgrading from "exporting production capacity" to "globalizing the supply chain," but also many upstream and downstream companies in industrial clusters.
Last year we discussed how industrial clusters have helped Chinese companies evolve; this year, let's look at how they have helped create "Chinese-owned global companies."
For example, Yutong Bus's overseas revenue has reached 60% in just a few years. It mentioned this in its annual report:
"Thanks to the improvement of vehicle technology and manufacturing processes, the first-mover advantage in new energy technology, and the comprehensive strength of a complete new energy supply chain, China's bus market share is rapidly increasing."
However, for Chinese new energy vehicles to gain a foothold overseas, simply establishing vehicle manufacturing plants abroad is not enough. Starting from the very upstream lithium mines:
"The company's Goulamina spodumene project Phase I was officially put into production in 2025 and is currently in the capacity ramp-up phase, which is an important manifestation of the accelerated entry of African lithium resources into the global market." — Ganfeng Lithium Annual Report
To the leading power battery manufacturer:
"As the construction and operation of the company's overseas bases have gradually matured, and strategic cooperation with overseas customers has deepened, the company's overseas market share and delivery capabilities have steadily increased during the reporting period... The company's after-sales service network covers 75 countries or regions and approximately 1,200 after-sales service stations." — CATL Annual Report
Then there are the component manufacturers and its own supply chain:
"In terms of long-term strategy, the company will adhere to the principle of 'optimal localization cost,' accelerate the localization of the supply chain, focus on cultivating Chinese suppliers and alternative supply chains with a global footprint, promote multi-source supply of key materials, and build a more resilient supply chain system." — Joyson Electronics
Only when industrial clusters are localized will "becoming part of the local economic development" no longer be just empty talk. And only by becoming part of the local economic development can they be constrained by security logic and trade paradigms as little as possible.
Some may question why, since multinational corporations in Europe and America did not achieve industrial cluster localization in the past, yet they still thrived, there are such high requirements in China.
Beyond differences in culture and systems, another crucial point is that multinational corporations in the past thrived in an environment of free trade and relatively stable geopolitics. This was true for Japanese trading companies and the automotive industry, as well as American multinational consumer companies. Chinese companies, however, currently face a completely different environment.
Therefore, while both are moving towards internationalization, the path for those with "Chinese citizenship" will inevitably be more difficult.
But that's not all. The third layer of localization is research and development, which is something even newer.
The most typical industry is, of course, biopharmaceuticals:
"We have built a large global clinical team of approximately 3,800 people across six continents, enabling us to conduct clinical trials with minimal reliance on CROs." — BeiGene Annual Report
"We will fully leverage our unique 'dual-engine' R&D advantages—combining China's efficient clinical resources and execution capabilities with a global development network." —Innovent Biologics Annual Report
Of course, the pharmaceutical industry has its own unique characteristics. In order to sell drugs overseas, clinical data must be internationalized to a considerable extent. Therefore, the pharmaceutical industry inherently has a need for localized research and development.
At the same time, the localization of R&D, stimulated by the new logic, also includes the electronics manufacturing industry, which previously emphasized "outsourcing to other countries":
"Currently, the US's imposition of tariffs on multiple regions around the world will lead to increased costs for the electronics manufacturing services industry, and the business and competitive environment will become more complex. Electronic manufacturing service providers are also actively transforming and upgrading to play a more important role in the supply chain."
The cooperation and integration between downstream customers and upstream electronic manufacturing service providers are deepening. Consumer electronics brands, cloud service providers and other brand manufacturers are no longer just order placers, but are more deeply involved in the production process of electronic manufacturing service providers, including technology research and development, production planning and quality control.
Similarly, electronics manufacturing services providers are no longer simply passively producing according to orders, but actively participating in the product planning and design of brand owners, providing professional technical and process advice, thereby forming a closer partnership between the two parties. — Universal Scientific Industrial (USI) Annual Report
Actually, I felt quite emotional when I was quoting these contents.
When I first started this series in 2019, the overseas expansion I saw was mostly sales and production capacity. Supply chain expansion was still in its infancy, and R&D expansion was nowhere to be found.
However, in just a few years, the number of "Chinese global companies" has increased significantly. Their three-stage localization approach can better adapt to the new trade paradigm and follow the shift from efficiency logic to security logic, thereby reconstructing the overseas expansion model of Chinese enterprises.
"...Deglobalization does not lead to isolation, but rather to deeper localization." — Mindray Medical Annual Report
five
Of course, I'm not suggesting that going global is the only path for Chinese companies in the current environment. In fact, the last piece of inspiration regarding restructuring comes from Ningwang. Its annual report begins by stating:
"CATL has always been answering one question through its actions: How to build long-term value with certainty in an uncertain world?"
So, what does long-term value mean for this trillion-dollar new energy giant?
First, “A company’s true competitiveness lies not only in its short-term performance, but also in whether it has established a system that can create value in a long-term, stable and sustainable manner, and promote the common progress of the company, the industry and even society.”
As a company that started by manufacturing batteries, CATL believes that this system encompasses both the depth of technological innovation and the quality of products. It went on to say:
"Making batteries is easy, but making good batteries is difficult. In a highly scaled industry, any minor defect can be magnified infinitely by time and space. CATL has always regarded quality as its lifeline, and its key quality standards are orders of magnitude ahead of the industry average."
A leading company regarded by the market as a pillar of the new energy industry did not write "AI empowerment" or "full-stack ecosystem" in its annual report, but rather "minor defects will be magnified infinitely by time and space". This precisely illustrates that one of the winning logics of Chinese manufacturing is to perfect the physical details, which is something that many Chinese manufacturing industries caught in the involution of competition have to sacrifice.
However, focusing solely on quality is not enough to reconstruct long-term value, CATL stated in its annual report:
"Looking to the future, rather than replicating existing successful experiences, CATL is more committed to expanding the boundaries of new energy and driving the industry from 'partial breakthroughs' to 'overall incremental growth'."
What is incremental? It means that batteries are no longer regarded as "components of transportation or energy storage devices", but as "basic units that support the buffering, stabilization and dispatch of energy systems".
"Throughout the history of human civilization, technological revolutions have always been accompanied by energy revolutions, and energy efficiency is a key variable in the leaps of civilization... New energy is no longer a cyclical investment, but a long-term, systemic infrastructure." — CATL Annual Report
The reason why the word "infrastructure" appears frequently in this year's articles is because I have found that at this stage, the more people move towards AI, online, and virtual economic development, the more they realize how important physical infrastructure is.
In its annual report, Weichai Power stated that backup power for data centers is a new infrastructure, Haomai Technology stated that gas turbines are a new infrastructure, and CATL stated that batteries are a new infrastructure.
These traditional manufacturing industries are all rebranding themselves as the fundamental units of the AI era. It may seem like they're just following the trend, but upon closer examination, it's clear that they're doing just that.
In fact, although it is the same wave of AI, the capital markets in China and the United States are choosing the part that they can digest.
Because China's advantages in the entire industrial chain of energy, minerals, and manufacturing are real, and the United States' advantages in application, software, and model ecosystem are also real, the capital market's preference is an extension of their respective comparative advantages under the new wave.
This is precisely why AI is being "physicalized" in China. For many manufacturing industries, securing a place in the energy revolution and the physical construction of AI as long-term value increments is a logical next step.
six
However, the price of physicalizing technological revolutions is that imagination is stifled.
If China's AI story remains forever stuck on "how much electricity we can supply to data centers," then we will undoubtedly repeat the mistakes of the smartphone era—the most lucrative profits in the industry chain went to Apple and other software platform and application companies that could conduct global business, while our applications, apart from dominating the domestic market, could only desperately try to get a small share of the "soup" from the manufacturing process.
Therefore, in my view, the restructuring brought about by these new terms, such as K-type differentiation, security logic, trade paradigm, "Chinese global companies" and "infrastructure in the AI era," is still insufficient.
We hope to see more DeepSeek, more BeiGene, and more companies and products that can lead this technological revolution from the R&D stage. Only in this way can we truly "build certain long-term value in an uncertain world".
So what's preventing breakthroughs in R&D and application? Is it a problem with the system, capital, or talent, or path dependence?
A passage in a foreign commentary I recently read might be closest to the truth:
"...In addition to lacking the most advanced chips, China also faces many structural constraints—a weak domestic market with a capital market much smaller than that of the United States, and capital controls that limit its attractiveness to mainstream foreign investors."
While Chinese scientists have risen to the ranks of the world's top scientists, many of them are still trained at American universities. China attracts relatively fewer foreign researchers and engineers, while Silicon Valley draws on a global talent pool, including top talent from China.
A significant portion of venture capital in China's technology sector originates from the United States, making it vulnerable to US restrictions. Simultaneously, increasing exclusion from Western public markets due to security concerns further limits its access to the most lucrative capital channels, despite China's likely dominant position in developing country markets.
Each of the points mentioned here is not a problem that can be solved in three to five years. In fact, we need many other things besides time, such as disruptive innovation and the restructuring of the entire technology research and development mechanism.
Let me give another example regarding the return and recruitment of R&D talent. In recent years, I've noticed that governments and universities at all levels in China have become more "silent." They prefer to work quietly behind the scenes rather than go into the media or explain to the public what they're doing.
Of course, we can interpret this as "doing great things quietly," but to a large extent, it's also because people are afraid of not being able to grasp the current direction of public opinion, and "the more you say, the more mistakes you make," so it's better to say nothing.
But if China's universities and government remain "silent" like many listed companies during annual report season, failing to actively discuss how they support the integration of industry, academia, and research, and how R&D talent can obtain better opportunities in China, perhaps top talent can be recruited in a targeted manner, but how will a wider range of mid- to high-level talent know about China's advantages?
This is something I've truly experienced during my year-long academic visit to the United States: many talented people are hesitant to return to China because of a lack of transparency regarding China's development model.
seven
Finally, perhaps it would be appropriate to conclude this lengthy article by quoting a passage from Midea Group's annual report:
"The history of corporate changes over the past two centuries illustrates a simple truth: corporate failure or decline is common, the momentum of success accumulates into the weight of a fall, and the vast majority of companies eventually become mediocre."
The ultimate capability of a company that can last is not growth, but the ability to constantly restart or regenerate itself.
Today, Midea must dare to question the paths and methods we were once familiar with, dare to break our ingrained habits and ways of thinking, and dare to change the mechanisms and systems that are already in operation today and have been proven successful in the past.
In 2026, the world will continue to move forward amidst turmoil and change, and we believe that common sense is our way of coping. Common sense is like air; the higher you go, the thinner it becomes.
The price of prosperity is perpetual instability; destruction breeds growth; stability is merely an illusion.
From this passage, I see the perpetual anxiety and unease of Chinese enterprises, which may be inherited from the anxiety of the Chinese people.
On the one hand, it causes individual fatigue and a desire to escape; on the other hand, it has become a driving force for the continuous growth of Chinese industries.
Hopefully, when we look back next year, we will have more answers to the restructuring of the industry, and that original innovation can bring profit margins, not just revenue growth.
"We still have a long way to go."
Yes, but you see,
"We've come this far." — Midea Group Annual Report
