The US-China Tech Race and What It Means for Investors

The US-China Tech Race and What It Means for Investors

The technology competition between the United States and China has moved well beyond trade tariffs and political rhetoric. It is now a full-spectrum contest spanning semiconductors, artificial intelligence, quantum computing, rare earth supply chains, energy infrastructure, and the race to wire the developing world. US policymakers have drawn explicit parallels between the current rivalry and the urgent industrial mobilisations of World War II and the Cold War, when intense scientific experimentation and technological roadblocks led the US government to organise and fund research in nuclear energy, jet aviation, and aerospace. But this time is different. The technologies at the centre of the US-China competition are primarily commercial. While they have benefited from publicly funded research and carry military and strategic significance, the markets for AI, semiconductors, and advanced manufacturing are vastly broader than anything contested in earlier eras. That distinction matters, because scaling public investment in AI or semiconductors diverts government programmes at a time when private sector funding for these technologies is already enormous. Neither country holds a decisive advantage across the board. The investment opportunities lie in understanding exactly where each country leads, where each is vulnerable, and which supply chains carry the most disruption risk. For portfolios with meaningful tech exposure, this is the defining macro theme of the decade.

Published 4 December 2025. Analysis based on institutional research available at time of writing.

The US Policy Roadmap for Tech Self-Sufficiency

The United States has assembled a layered and increasingly aggressive policy framework aimed at rebuilding domestic capacity in semiconductors, critical minerals, advanced manufacturing, and defence technology. The scale and scope of this effort have expanded significantly over the past three years, and the latest legislative and executive actions suggest the pace is accelerating rather than slowing.

The foundation remains the CHIPS and Science Act, which allocated roughly $43 billion in direct funding for semiconductor manufacturing, with additional funding channelled through the CHIPS R&D COMPETES Act passed in 2022. But the more recent One Big Beautiful Bill Act (OBBBA) has pushed the incentive structure further. OBBBA renewed and expanded the Advanced Manufacturing Investment Credit (Section 48D) from 25% to 35%, directly subsidising domestic investment in semiconductor fabrication facilities. It also preserved the Inflation Reduction Act’s Advanced Manufacturing Production Credit (Section 45X) for batteries, solar components, and wind components, keeping the incentive architecture intact for the broader clean energy and advanced manufacturing supply chain.

Beyond tax credits, OBBBA directed $2 billion to the Defence Innovation Unit for military technology development and $1 billion to the Industrial Base Fund for investments in critical mineral supply chains. The Trump Administration separately announced $10 billion in government financial stakes across ordnance, energy, semiconductor, and steel sectors, signalling a willingness to take direct equity-like positions in industries deemed strategically essential.

On the trade side, the picture has evolved in ways that matter for investors tracking specific sectors.

• The Trump Administration reduced the 20% tariff on imports from China to 10% as part of the October deal, which also paused US export controls in return for China easing rare earth export restrictions. A remaining 10% tariff on all imports, plus an additional 10% “reciprocal” tariff on certain categories, remains in place.
• The Administration launched national security investigations under Section 232 covering pharmaceuticals, critical minerals, drones, wind turbines, robotics, and industrial machinery. These investigations give the government broad authority to impose tariffs or restrictions on any sector deemed a security concern.
• The Biden Administration’s GANI Act and Diffusion Rule, finalised in 2025, would have required US chip companies to retain a sufficient portion of AI-relevant chips for domestic use. While the current Administration has not fully embraced those rules, the policy direction of ensuring domestic supply of frontier AI chips remains broadly bipartisan.

The cumulative effect of these measures is a policy environment that is pouring money into domestic production, restricting technology transfer to China, and using national security authorities to extend government reach into sectors that were previously considered purely commercial. For investors, the implication is that companies aligned with these policy priorities will benefit from sustained government support, while companies caught on the wrong side of export controls or tariff barriers face ongoing headwinds.

China’s Approach Is Comprehensive and Systematic

China’s technology strategy operates on a fundamentally different model. Where the US approach is layered across multiple pieces of legislation, executive orders, and agency actions, often with competing priorities and political constraints, China’s approach is centralised, long-term, and designed to build entire ecosystems rather than subsidise individual companies.

The institutional architecture was upgraded in 2023 when Beijing established the Central Science and Technology Commission as part of the broader “Party and State Institution Reform” and restructured the Ministry of Science and Technology to give the central government tighter control over research priorities and funding allocation. This is not cosmetic. It places technology strategy at the same institutional level as national security and economic planning, with direct oversight from the most senior levels of the Chinese Communist Party.

The 15th Five Year Plan proposal, which will guide policy through the end of the decade, identifies three areas of focus that give a clear picture of where China intends to compete.

• “Crustal industries” including new energy technology, advanced materials, aerospace, and biotech. These are sectors where China already has significant domestic capacity and intends to push toward global leadership.
• “Future industries” including quantum computing, biocomputing, energy and nuclear technology, brain-computer interfaces, AI, 6G mobile communications, and next-generation space satellites. These are longer-term bets where China is investing heavily in basic research and early-stage commercialisation.
• Government support extends well beyond traditional tax credits and grants. Chinese companies in priority sectors receive cheap land, expedited permits, abundant energy supplies, and systematic efforts to connect upstream and downstream players to foster complete domestic supply chains. The objective is not just to support individual firms but to create self-reinforcing ecosystems that reduce dependence on foreign inputs at every stage of production.

The results are visible in the data. China’s production capacity now exceeds global demand in solar modules, lithium batteries, construction machinery, air conditioners, electric vehicles, steel, and power semiconductors. Institutional research shows that Chinese firms in priority sectors receive financial subsidies far exceeding those available to their global counterparts, including grants, income tax concessions, and below-market borrowings that are all significantly higher than OECD averages.

This is not a country playing catch-up across the board. It is a country that has achieved industrial dominance in several critical technology supply chains and is systematically extending that dominance into the sectors that will define the next two decades of economic competition.

The Semiconductor Supply Chain and Its Chokepoints

Of all the technology supply chains caught up in this competition, semiconductors are the most strategically significant and the most closely watched by investors. Understanding the structure of this supply chain is essential for evaluating the risks and opportunities.

The global semiconductor industry operates across three main business models. Integrated Device Manufacturers like Intel and Samsung design and fabricate their own chips. Fabless companies like Nvidia and AMD design chips but outsource fabrication. And pure-play foundries like TSMC and GlobalFoundries manufacture chips designed by others. Each model occupies a different position in the supply chain and faces different competitive dynamics in the US-China context.

Three chokepoints in the supply chain deserve particular attention because they represent points of concentrated leverage.

• Advanced lithography. ASML, the Dutch company, is currently the only company in the world with the capability to produce extreme ultraviolet (EUV) lithography machines, which are required to manufacture chips at the most advanced nodes below 7 nanometres. This is a genuine monopoly position that has taken decades and billions of dollars of R&D to establish. China cannot currently replicate this capability domestically.
• Electronic Design Automation (EDA) and intellectual property. The software tools used to design advanced semiconductors are dominated by three US-headquartered companies: Synopsys, Cadence, and Siemens EDA. Without access to these tools, designing chips at the leading edge is effectively impossible. The US also controls critical IP libraries that are licensed across the global chip design ecosystem.
• Wafer fabrication at advanced nodes. Actual chip manufacturing at the leading edge is concentrated in Taiwan (TSMC) and South Korea (Samsung), with the US holding a relatively small share of global fabrication capacity. This geographic concentration is one of the primary motivations behind the CHIPS Act and related incentive programmes.

The US dominates EDA, IP, and logic design, but the actual physical production of advanced chips is overwhelmingly located in East Asia. US fab capacity is projected to increase roughly 200% from 2012-2022 levels by 2032, compared to a world average increase of around 100%. That is a meaningful shift, but it will take years to materialise and the US will remain a minority player in global fabrication for the foreseeable future.

China has been building semiconductor capacity aggressively, but primarily in trailing-edge (mature) nodes. These older process technologies are less glamorous than the cutting-edge chips that power AI training, but they are essential for automotive electronics, industrial controls, telecommunications equipment, and consumer devices. China’s dominance in mature-node production gives it significant leverage over global supply chains even without access to the most advanced fabrication technology.

Why Export Controls Have Not Worked as Expected

One of the most important findings from institutional research on this topic is that US export controls on advanced chips have not been especially effective at slowing China’s AI development. This point deserves more attention than it typically gets, because it challenges a central assumption of US technology policy.

Experts studying the semiconductor supply chain note that there is no single semiconductor supply chain but rather multiple overlapping ones, each with different chokepoints and different vulnerability profiles. The export control regime has targeted the most advanced chips and the equipment needed to make them, but Chinese companies and research labs have found several workarounds.

• Using less advanced chips in larger clusters, trading energy efficiency for raw compute power.
• Architectural innovations that squeeze more performance out of available hardware.
• Aggressive optimisation of software and training pipelines to reduce the compute required for a given level of model performance.
• Domestic chip development programmes that, while still behind the leading edge, are closing the gap faster than many expected.
• Grey market procurement through intermediary countries and shell companies.

The net result is that export controls have imposed a cost on Chinese AI development but have not prevented it. China’s commercial AI adoption is actually ahead in many areas despite the chip restrictions. Less advanced chips can still power a wide range of AI applications, and the gap between what China can produce domestically and what is needed for most commercial AI use cases is narrower than often assumed. For investors, the implication is that the semiconductor supply chain remains strategically important, but the thesis that export controls will permanently hobble Chinese technology is not well supported by the evidence so far.

Rare Earths and the Supply Chain That Keeps Defence Planners Awake

If semiconductors are the most watched supply chain, rare earths are the most concerning. The numbers tell the story. China holds 49% of global rare earth reserves, controls 69% of global mining, dominates 92% of global refining (particularly for heavy rare earths), and produces 98% of the world’s permanent magnets. These magnets are critical components in electric vehicle motors, wind turbine generators, missile guidance systems, satellite components, robotics, and a long list of other advanced applications.

The only other major player at the refining stage is Lynas, an Australian company operating in Malaysia, which accounts for roughly 4% of global refined production. The United States accounts for just 1% of global rare earth refining. Mountain Pass in California, operated by MP Materials, is the only active US rare earth mine, and it primarily produces light rare earths rather than the heavy rare earths that are most critical for permanent magnet manufacturing.

The concentration risk is severe and getting worse. China recently banned the export of technology for extracting, separating, and processing rare earths, cutting off a potential pathway for other countries to build refining capacity using Chinese know-how. Without that technology transfer, countries attempting to build domestic refining face the additional challenge of developing process knowledge from scratch.

One of the most striking assessments in the institutional research comes from Jack Lifton, a rare earth industry expert. His view is blunt. “The institutional knowledge necessary to make rare earth permanent magnets, especially the specialty magnets, is all but gone. I know all the people left in America with experience making such magnets, and you could fit them into a handful of phone booths.” He goes on to describe the US government’s current plan to achieve rare earth self-sufficiency as “simply unattainable in the short run and will not succeed.”

The US has no commercial-scale facility for heavy rare earth separation. Building one is not just an engineering challenge but a human capital challenge. The expertise needed to operate rare earth refining and magnet manufacturing at commercial scale requires years of accumulated know-how that cannot be replicated through capital investment alone.

For investors, rare earth mining and refining outside China represents an area where we expect to see significant capital allocation over the coming years. The economics are challenging because Chinese producers benefit from decades of scale and process optimisation, but the national security imperative is likely to override pure economic considerations in many Western countries. Companies positioned in this space that can achieve commercially viable production will benefit from both government support and growing strategic demand.

China’s Advantages Beyond Manufacturing

The competitive dynamics extend well beyond semiconductors and rare earths. China holds several structural advantages that are often underappreciated in Western analysis.

• Global supply chains and economic ties with the Global South. China is now a larger trading partner than the United States for over 140 countries. That economic weight translates into political influence, standard-setting power, and market access that the US cannot easily replicate.
• Telecommunications infrastructure. Huawei operates in over 170 countries, establishing Chinese telecom technology standards across much of the developing world. The installed base of Chinese networking equipment creates path dependencies that are difficult to unwind, even where Western governments raise security concerns.
• International standards organisations. China has called for deeper participation in global standards bodies and has been steadily increasing its presence and influence in the organisations that set technical standards for everything from 5G to AI to quantum computing. Standards may sound bureaucratic, but they shape markets for decades.
• Energy advantage in AI. Abundant domestically produced renewable energy, particularly cheap solar power, gives China an edge in energy-intensive AI and data computing. Training and running large AI models is extraordinarily energy-intensive, and China has strategically located data centre clusters in provinces with abundant cheap electricity. Yunnan, Sichuan, Guizhou, and Inner Mongolia all offer energy costs well below what is available to American data centres.
• Overcapacity as a deliberate strategy. China’s tendency to rapidly scale up production capacity following breakthroughs is often characterised as a problem in Western analysis. But it can also be understood as a feature of the Chinese model. Flooding global markets with cheap production drives down prices, crushes competitors who cannot match the cost structure, and establishes dominant market positions. It can lead to wasteful investment, but in sectors with strong learning curves and network effects, the first mover with scale often wins.

US data centres are already facing binding constraints in several key markets. Utilities in Northern Virginia, which hosts the largest concentration of data centres in the world, are struggling to meet demand growth. Permitting timelines for new generation and transmission capacity stretch out five to ten years in many jurisdictions. Some hyperscale operators have resorted to signing power purchase agreements with nuclear plants or developing their own on-site generation, but these are expensive solutions that add to the cost structure of American AI development. As models grow larger and more compute-intensive, the cost and availability of power will become an increasingly important competitive factor. This is a structural advantage for China that compounds over time.

Two Strategic Models and Their Weaknesses

Neither approach is inherently superior, and understanding the weaknesses of each model is just as important as understanding its strengths.

The US model is more flexible and market-driven, which tends to produce genuine innovation at the frontier. American universities and research labs still attract the best talent globally. The venture capital ecosystem funds riskier and more creative research than state-directed investment programmes typically support. And the depth of US capital markets allows successful companies to scale rapidly once they have proven product-market fit.

But the US model has real vulnerabilities.

• It is slow to scale manufacturing. The CHIPS Act was signed in 2022 and the first subsidised fabs will not produce chips in volume until 2027 or later.
• It is politically fragmented. Policies change with administrations, and companies making decade-long investment decisions need more certainty than US politics typically provides.
• It underinvests in the “boring” parts of supply chains. Refining, materials processing, and mature-node semiconductor manufacturing do not attract venture capital or generate the kind of margins that excite Wall Street, but they are essential infrastructure.
• Permitting and regulatory processes are slow. Building a new semiconductor fab, a rare earth refinery, or a power plant in the US takes significantly longer than in China, adding years and billions to project timelines.

The Chinese model has its own set of weaknesses.

• Capital misallocation. State-directed investment does not always go to the best projects. The history of Chinese industrial policy includes significant examples of wasted investment in sectors that never achieved commercial viability.
• Political interference in commercial decisions. As the role of the Communist Party in the private sector has expanded under Xi Jinping, foreign investors have become more cautious about the predictability of the business environment.
• Quantity over quality in research output. China produces more scientific papers than any other country, but citation rates and breakthrough discoveries still lag the US and Europe in many fields.
• Dependence on foreign technology at the frontier. Despite rapid progress, China still relies on foreign tools, equipment, and IP for the most advanced chip designs and fabrication processes.

The competition between these two models will play out over decades, and there will be periods where each appears to have the upper hand. The investment landscape will shift accordingly.

What This Means for Portfolios

The investment implications of the US-China tech race are more nuanced than simply picking a side. We think several themes deserve attention in portfolio construction.

• Semiconductor and equipment stocks have already largely priced in the current US-China dynamics. The biggest gains in names like ASML, TSMC, and Nvidia have reflected both AI demand growth and the geopolitical premium from supply chain concerns. We do not think the easy money has been made here, but the sector remains essential to own for long-term technology exposure.
• Western rare earth mining and refining companies are potential beneficiaries of the diversification push. The economics are challenging, but government support and strategic demand create tailwinds that may prove durable. This is a space where patient capital could be well rewarded if even a handful of non-Chinese projects achieve commercial viability.
• Diversification away from single-country tech supply chain exposure is increasingly important. Portfolios that are heavily concentrated in US mega-cap tech names carry implicit risk if the competitive balance shifts, if policy escalation disrupts supply chains, or if regulatory action in either country changes the landscape for specific companies.
• Asia technology companies outside China may be among the biggest beneficiaries of the tech race as it continues. Companies in Taiwan, South Korea, Japan, and Southeast Asia that serve both markets could see sustained demand regardless of how the rivalry evolves. These companies sit at the nexus of both supply chains and benefit from being essential to both sides.
• The energy-AI nexus is an emerging theme worth watching. Companies that can provide reliable, low-cost power to data centres will be in high demand globally. The geography of AI compute may shift toward regions with the most favourable energy economics, which has implications for infrastructure investment, utility earnings, and real estate values in data centre corridors.
• Policy shifts remain the biggest near-term risk factor. Export controls, tariff changes, rare earth restrictions, and new subsidy programmes can all move markets rapidly. Staying close to the policy timeline and understanding which sectors are in the crosshairs is essential for managing risk in technology-heavy portfolios.
• The Global South digital infrastructure build-out represents a long-duration growth opportunity. Whether the builders are Chinese or Western or a mix, the underlying demand for telecommunications, data centre, and payment infrastructure across Africa, Southeast Asia, and Latin America is structural and large.

The Bottom Line

The US-China tech race is not a short-term trade. It is a structural shift in the global economy that will play out over decades and affect virtually every sector along the way. The historical parallels with the Cold War are instructive but imperfect. The technologies at stake are commercial, the supply chains are globally interconnected, and the economic interdependence between the two countries is far deeper than anything that existed between the US and the Soviet Union. That makes the competition harder to manage, harder to predict, and richer in investment implications.

For investors, the key is to avoid oversimplified narratives about one country “winning” and instead focus on the specific supply chains, technologies, and geographies where the competition creates both risk and opportunity. The semiconductor chokepoints, the rare earth vulnerability, the energy dynamics of AI, and the battle for the Global South are all distinct theatres of competition with distinct investment implications. The companies and countries that can secure their technology supply chains while maintaining access to global markets will be the long-term winners. The rest will find themselves squeezed from both sides, paying more for inputs and losing market access at the same time. We think portfolios need to be positioned for this reality rather than for a world where one side achieves clean dominance.

If you would like to discuss how global macro trends might affect your portfolio, request a callback or call us on 1300 889 603.