China’s innovation revolution

For centuries, Chinese ingenuity reshaped the world. Inventions such as paper, gunpowder, silk, porcelain and the magnetic compass transformed global trade, science and culture. Today, that spirit of innovation is evident in China’s rapid progress in next-generation technologies, from autonomous vehicles to renewable energy and artificial intelligence (AI).

Companies from across the AI value chain were among the China market’s best performers in 2025, including internet firms that are integrating large language models (LLMs) into their platforms; manufacturers whose components are powering servers and data centres; and miners unearthing essential materials, such as the copper used in high-voltage cables and battery systems. Many of these firms have delivered stellar returns at a time when the domestic economy remains lacklustre and consumer-orientated companies are struggling. The result: a bifurcated market, similar to the pattern that is emerging elsewhere in Asia and around the world.¹

In 2025, performance of our China strategies reflected this divide. On the one hand, our investments in high-quality AI-related companies did well, as hyper-scalers’ capital expenditure (capex) continued to rise and AI became more embedded within business models and daily life. On the other hand, holdings in sectors exposed to domestic consumption lagged.

As technology evolves, supply chains adjust and regulation shifts, we are closely assessing how AI is reshaping China’s opportunity set – and which businesses we believe are best placed to deliver sustainable returns over the longer term. At the same time, we continue to find plenty of strong companies in non-AI sectors, many of which are innovating in quieter but no less important ways.

To understand how the rise of AI is affecting the landscape for bottom-up equity investors, it is helpful to trace the relationships between firms across the value chain. Harvard Business School professor Andy Wu offers an instructive analogy that compares AI investment to the gold rush in 19^th century America: chipmakers and other component manufacturers provide the tools; AI model-builders dig for gold; and internet platforms and other consumer-tech firms fashion the final products.²

Among the positive contributors to our performance in 2025 were suppliers of the “picks and shovels”, such as ACM Research, a company that operates in a critical niche of the semiconductor supply chain. ACM designs and manufactures high-specification wet-cleaning tools used in chipmaking, equipment that plays a vital role in maintaining yield and quality at fabrication facilities.

Chipmaking demands extraordinary precision. Microscopic contaminants or chemical residues can render silicon wafers unusable, particularly at the high-end nodes required for AI applications. ACM’s tools help address this challenge, supporting customers as they push the limits of density and complexity.

The company has seen increased demand amid the rapid development of China’s semiconductor ecosystem, which has been reinforced by the government’s desire to reduce reliance on foreign suppliers in advanced technologies. Importantly, ACM’s competitive position reflects more than cyclical demand; the quality and reliability of its equipment has enabled it to take share from both domestic and international rivals, while also expanding its overseas customer base, which includes Korean memory-chip leader SK Hynix.

Despite its strong recent growth in revenue and profits, the company’s management team remains disciplined, taking steps to expand ACM’s product portfolio, which now encompasses plating and wafer packaging, and diversify its client base. For all ACM’s attractive characteristics, however, its valuation has now reached expensive levels, and we have used the recent share price strength to trim our position.

In addition to component manufacturers, we also own companies that are using AI to enhance the services they offer. History suggests that during times of technological change it is often those firms that take advantage of new advances to improve processes, distribution models and products, rather than the builders of the underlying infrastructure, that benefit the most.³ To return to the gold-rush analogy, these are the firms that can lucratively convert the raw materials of generative AI into sparkling jewellery.

A good example is our holding Tencent, which was another key contributor to performance in 2025. Tencent has unique advantages in the AI era: a massive user base across WeChat, games and payments; rich first-party data; and integrated advertising and content systems.

Rather than rushing into the AI capex “arms race”, however, Tencent is deploying the technology to strengthen its core businesses and improve return on investment. For instance, AI enables the company to sharpen advertising targeting, personalise gaming experiences and improve content recommendations. Over time, this integration of AI should support higher productivity, stronger engagement and, in turn, more-resilient profit growth.

While AI-related companies have made positive contributions to performance, we have chosen not to invest in some of the names that the market perceives to be the biggest beneficiaries.

Alibaba is a case in point. The company’s approach to AI contrasts sharply with Tencent’s asset-light strategy: it has announced plans to spend more than US$50bn building new data centres over the next three years.⁴ While we own Alibaba in some of our China and Hong Kong portfolios, our decision to limit our exposure to the company negatively affected relative performance in 2025, as optimism around its cloud computing business lifted the share price. While we are reviewing the investment case, we continue to prefer Tencent.

Beyond technology, performance across our other holdings was mixed. Among the biggest detractors were dairy firm China Mengniu Dairy and food-delivery company Meituan, reflecting weak consumer demand. Both have been proactive in strengthening their businesses through the slump; Mengniu has improved its product mix, for example, which has brought early signs of margin recovery. Meituan, meanwhile, is facing intense competition from the likes of JD.com, which aggressively entered the food-delivery market in 2025. We believe such irrational competition is unsustainable, and that Meituan will be well placed when it abates; the company’s market capitalisation is modest when viewed in the context of its gross merchandise value (GMV) and the great potential of China’s local-services market.

Mengniu and Meituan are not the only consumer-linked businesses that have disappointed of late. The weakness of consumption across the economy – which has persisted for longer than we anticipated, following the short-lived post-Covid reopening – has affected other consumer holdings, which comprise a significant proportion of our China portfolios.

Nevertheless, we remain confident that consumer spending will structurally rise over the longer term. With household bank deposits now exceeding RMB160tn (more than 100% of GDP), the issue is not so much a lack of wealth as a lack of confidence among China’s consumers. Recognising this, the government has indicated it will bring in policies to strengthen the social safety net – and thereby encourage citizens to loosen their collective purse strings – as part of its next Five-Year Plan.

For the time being, lacklustre demand and the low-interest rate environment in China is driving investors towards companies that are demonstrating strong growth – even where the sustainability of this growth looks questionable. History offers a useful reference point: during 2020-21, our China portfolios delivered strong results after we avoided popular stocks that later underperformed when growth expectations fell short, and we would expect the pattern to repeat. In the meantime, we continue, as always, to focus on fundamentals, seeking out companies with capable leaders, strong franchises, attractive earnings growth and sound balance sheets.

One reason for optimism on China is the scale and diversity of innovation occurring across the economy, with AI only one part of the picture. Official data shows overall research and development (R&D) spending exceeded RMB3.6tn (around US$500bn) in 2024, with such investment growing faster than in any other major economy.⁵ China now produces significantly more science, technology, engineering and mathematics (STEM) PhDs than the US, with 40% of all Chinese graduates earning STEM degrees.⁶

China’s technological progress is especially visible in the advanced manufacturing network that spans the cities of Guangzhou, Shenzhen and Hong Kong: this was ranked the world’s top innovation cluster in 2025, as measured by scientific publications, venture capital activity and patent applications.⁷ As author Dan Wang argues in his recent book Breakneck, companies in this cluster excel in “process knowledge”, a kind of technical and operational expertise deriving from collaboration across engineering teams and rapid feedback loops between R&D and production.⁸

Our holding Luxshare Precision is a clear example. A Guangdong-based electronics contract manufacturer (ECM), Luxshare has embedded technical knowledge at every level of the firm, enabling it to move quickly, take advantage of synergies across business lines and solve complex problems for major clients such as US tech giant Apple.

Co-founder and CEO Wang Laichun embodies the company’s ethos. Having started her career in Shenzhen as an assembly-line worker at Taiwanese manufacturing firm Hon Hai Precision Industry (also known as Foxconn), Wang started Luxshare in 2004 and has built the company through strategic investments in assembly-line automation and smart robotics, combined with a series of shrewd acquisitions.

Over time, Luxshare was able to reduce its reliance on external suppliers and sub-components and develop into a vertically integrated business. It was therefore well positioned to win major assembly contracts when Apple started to diversify its supply chain in the early 2010s, first for the iPhone, then for other gadgets such as the Apple Watch, AirPods and Vision Pro headset.

Working with the US giant to solve intricate engineering challenges has helped lift Luxshare to new heights – Wang likens the partnership to “flying with a phoenix”⁹ – but with Apple now accounting for more than 70% of its revenue and 60% of profit, Luxshare is sensibly taking steps to diversify its own business. At a recent company investor day, we were impressed by the scope of its growing product range, which includes components for high-performance computing, data-centre and automotive applications.

While entering the market for data-centre parts should enable Luxshare to capture rising AI-related demand, its automotive business looks even more promising. As electric vehicles (EV) take a larger share of car fleets around the world, the automotive supply chain is opening to new players in areas such as electric-motor assembly and the manufacture of cables and connectors, where Luxshare has expertise. Thanks to its joint venture alongside Chinese carmaker Chery Auto, and growing business from other EV clients including Tesla and BYD, we believe the auto segment could become a key growth driver for Luxshare within the next three to five years.

Risks to Luxshare include the impact of US tariffs on its export-focused model, and, over the longer term, increasing competition for Apple contracts from other regional manufacturers. We would also like to see the company simplify its rather complicated ownership structure.

However, Luxshare’s disciplined management and track record of scaling new businesses give us confidence in its prospects. While its valuation has risen over recent months, we think the company still looks attractive; with Luxshare’s evolution from contract manufacturer to diversified technology firm gathering momentum, we see potential for a mid-teens compound annual growth rate (CAGR) in earnings over the next five years.

Chinese innovation is also evident among more consumer-facing companies. Over recent years, slower economic growth has intensified competition for customers in China, forcing companies to rethink and refine their products, branding, distribution and operations.

Among these innovators is our holding Midea Group. Founded in 1968 as a maker of bottle caps, the company moved into home appliances in 1981. In recent years, Midea has entered new segments, including industrial machinery, but it remains best known for its air conditioners, washing machines, refrigerators and other household products, and ranks first or second in China across all major home-appliance categories.

Like Luxshare, Midea has exceptional manufacturing capabilities. For example, on a visit to the company’s headquarters in Shunde, near Shenzhen, we saw a plant that makes cavity magnetrons used in microwaves. Midea has a leading global market share in these components, with 90% of its output used in-house. We were impressed by the efficiency and sophistication of the automated production line.

As well as its manufacturing knowhow, Midea’s success is also down to product innovation and clever positioning in the market. After rapid expansion in the early 2000s, the company became known for mediocre mass-market gadgets, so it pivoted to high-end, design-led products sold under its proprietary brands, such as the COLMO range of AI-powered smart appliances. In creating this range and others, Midea has drawn on digital product-planning platforms that link customer data, product design and manufacturing, enabling it to respond more quickly to shifts in demand. Development cycles shortened by 30% between 2019 and 2024.

While Midea is diversifying its product range – it recently unveiled a six-axis robot, designed for industrial tasks – home appliances remain at the core of its business. We are consistently impressed with its management’s execution in this area, as it has defended its market leadership despite intense price competition from rival Xiaomi. Midea also sees scope to grow in overseas markets.

There may be headwinds in the short term. The government’s trade-in policy (a programme of subsidies designed to boost consumption) has distorted demand among home-appliances firms, pulling forward sales, and normalisation may affect Midea’s earnings over the coming months. Overall, however, we see the company as one of China’s most resilient and cash-generative franchises. With a dividend yield of more than 5% and a payout ratio of 68%,¹⁰ it is proving its ability to return capital to shareholders.

Another innovator in the portfolio is H World, a leading hotel-chain operator with the potential to consolidate the market. Under its franchising model, franchisees are responsible for capital investments in real estate and pay a recurring fee for H World’s branding and marketing (in many cases, this is a “manachised” arrangement, with H World also providing a hotel manager).

H World’s digital capabilities give it a key advantage over its rivals. Technology has been deployed throughout the business, from the self-service kiosks in its hotels to its app-based membership scheme. With 300 million members, the company is able to collect a wealth of data on travel patterns and preferences, enabling it to improve services and better target its marketing efforts. In turn, customers benefit from loyalty points that encourage repeat stays. Around 74% of total H World bookings come from members.¹¹

The company has admitted missteps in recent years, including a poorly timed acquisition of Germany’s Deutsche Hospitality in 2020; it is also trying to divest from some legacy leased and owned properties to become more of a pure asset-light operator. Looking ahead, revenue per available room (RevPar) has now turned positive after a downturn, and over the longer term we believe the company has a long runway of growth as it taps into increasing travel demand and upgraded consumer preferences in China. The underlying metrics are good, with H World posting strong free cash flow and return on equity (ROE) expected to exceed 34% for 2025.¹²

As this letter has illustrated, China’s innovation story extends far beyond a narrow group of AI-related names. Across technology, advanced manufacturing and consumer sectors, we see companies applying new processes, products and business models to strengthen their competitive positions.

Our China portfolios are built to capture this breadth. Rather than chasing a single theme, we focus on bottom-up opportunities in businesses led by capable management teams, with robust governance, strong franchises and the ability to compound value over time. These are the kinds of companies that we believe would  stay resilient, regardless of how the AI boom ultimately plays out.

Looking ahead, we see encouraging signs that China’s equity market is entering a more constructive phase. Corporate governance standards continue to improve, with greater transparency and accountability among listed companies, while minority shareholder protections are being strengthened. These developments should help create a more predictable investment environment, one that rewards discipline and long-term thinking.

Greater policy clarity is also emerging. The recent easing in US-China trade tensions, along with Beijing’s outline for its next Five-Year Plan, has helped reduce some of the uncertainty that weighed on sentiment. At the same time, government efforts to address “involution” – the cycle of excessive competition and price-cutting – should foster a more rational corporate landscape, allowing stronger companies to translate their advantages into higher margins and, ultimately, durable shareholder returns.

As always, we thank you for your continued support.

References

¹ See also “Beyond the AI rally”, FSSA, November 2025. 

² “Should the US be worried about an AI bubble?”, The Harvard Gazette, December 2025.

³ “On AI: The age of extremes”, FSSA, October 2025.

⁴ “Alibaba to invest more than $53bn in data centres over next three years”, Reuters, February 2025.

⁵ National Bureau of Statistics, January 2025; “R&D spending growth slows in OECD, surges in China”, OECD, March 2025.

⁶ “China’s universities outpace US peers amid tech competition”, FDI Intelligence, April 2025.

⁷ World Intellectual Property Organization ranking of global innovation clusters, 2025.

⁸ Dan Wang, Breakneck: China’s Quest to Engineer the Future (W.W. Norton, 2025).

⁹ “Luxshare Precision's Wang Laichun: To fly with the phoenix, one must be a magnificent bird”, China Entrepreneur Magazine, April 2024.

¹⁰ As of September 30, 2025.

¹¹ As of September 30, 2025.

¹² FactSet consensus estimates.

Source: All company data herein retrieved from company annual reports or other such investor reports. Financial metrics and valuations are from FactSet and Bloomberg. As at 31 December 2025 or otherwise noted.