(Shutterstock)
A lingering sense of nervousness remains among AI investors.
A significant concern centers on whether Chinese AI developers and their low-cost models threaten to usurp US competitors with higher sunk investment costs. Following a period where Chinese companies outperformed their US counterparts, last week's trading ended with a 1.4% daily gain for the Magnificent 7 on Friday. In contrast, Chinese AI leaders experienced declines, finishing the week 8% to 10% below recent highs, largely due to fewer positive earnings surprises.
While both the United States and China have made significant strides in the AI sector, CIO believes there are compelling reasons to favor US AI companies over their Chinese counterparts, especially in the near term. This article delves into three key drivers that make US AI investments more attractive in our view: capital expenditure (capex) trends, research and development (R&D) spending, and monetization potential.
US firms’ outsized capex should drive competitive advantage. Capital expenditure (capex) is a critical metric for evaluating the growth potential of AI companies. It refers to the funds used to acquire, upgrade, and maintain assets such as property, tech infrastructure, or equipment. In the AI sector, capex is essential for training and inference of large language models (LLMs), which are the foundation of AI advancements.
US AI companies, particularly the Big 4 (Amazon, Alphabet, Microsoft, and Meta), have demonstrated a strong commitment to capex, significantly outpacing China's Big 4 (Alibaba, Baidu, ByteDance, and Tencent). In 2024, the US Big 4 spent USD 224 billion on capex, and this is expected to rise to USD 302 billion in 2025, representing a 35% growth rate. By contrast, China's Big 4 are projected to increase their capex from USD 33 billion to USD 51 billion over the same period, a 54% growth rate. However, the absolute spending in the US remains almost six times higher, providing a clear scale advantage. Scale increasingly matters as LLMs develop to become reasoning models that require significantly more computational power to fulfill the tasks asked of them.
The higher capex intensity in the US, defined as capex spending divided by revenues, stands at 20% in 2025 compared to China's 11.7%. This disparity highlights the US's commitment to maintaining a technological edge, even though it may lead to higher depreciation-related expenses in the short term.
US firms look better placed to find “the next big thing.” Research and development (R&D) spending is another crucial factor driving AI innovation. Higher R&D investment often correlates with greater pricing power and the ability to develop cutting-edge technologies. US AI companies are leading the way in R&D, both in absolute terms and relative intensity.
In 2025, the top three US cloud platforms (Microsoft, Amazon, and Alphabet) are expected to spend a combined USD 180 billion on R&D, compared to USD 35 billion by the top three Chinese cloud platforms (Alibaba, Tencent, and Baidu). This significant difference underscores the US's commitment to fostering innovation and maintaining a competitive edge in the AI sector, in our view.
R&D intensity, measured as R&D spending divided by revenues, is also higher in the US, at 13.5%, compared to 8% in China. This higher intensity is supported by the robust gross margins of US cloud platforms, which stand at around 70%, versus 50% for their Chinese counterparts. This advantage allows US companies to invest more heavily in R&D without sacrificing profitability.
Use cases and target customers suggest US profits will exceed Chinese ones. Monetization potential is the third key driver favoring US AI over China's. It refers to the ability of companies to generate revenue and profit from their AI investments. US AI companies have an advantage in this area, particularly in the highly profitable enterprise segments.
The top three US cloud platforms are projected to generate 12 times more cloud revenue than their Chinese counterparts, despite spending only 6-8 times more on cloud/AI capex. This disparity highlights the superior monetization strategies employed by US companies, which we believe are better positioned to capitalize on the growing demand for AI solutions in enterprise technology.
Moreover, the relatively low penetration of enterprise technology in China, coupled with a focus on low-cost or open-source models, limits the revenue potential for Chinese AI companies. In contrast, US firms benefit from first-mover advantages, stronger pricing power, and a larger addressable market, making them more attractive for investors seeking growth in the AI sector. We believe the chances of a relatively quick payback are higher for US AI compared to China AI. This is because we anticipate that China AI will focus largely on integrating AI into lower-margin consumer technologies where it already has dominance, such as e-commerce, gaming, and electric vehicles.
For longer-term investors, the valuation outlook for US cloud platforms remains appealing, in our estimation. Despite trading at a higher P/E ratio (23x this year’s earnings) compared to their Chinese counterparts (18x), the premium is justified by their superior scale, R&D capabilities, and monetization potential. When adjusted for free cash flows, US cloud platforms trade at a more modest 14% premium (versus around 30% P/E premium), highlighting their efficiency in generating cash. This, combined with their strong pricing power and robust growth prospects, makes US cloud platforms an attractive investment for those looking to capitalize on the long-term growth of the AI sector, in our view.
So without taking single-company views, we believe investors looking to benefit from the ongoing capex boom in AI should focus on leading AI compute names within the US market. With the US Big 4 expected to see a 35% growth in capex in 2025, and a strong 30% CAGR estimate for AI compute spending from 2024-29, the demand for AI tokens is set to remain strong.
Beyond investing in CIO’s own "AI" portfolio, investors can prioritize investments in AI semiconductors and leading cloud platforms. Tools that help investors monetize elevated volatility and buy the dip may be promising in the current environment. And we also see merit in using satellite investments in broad indexes of semiconductors as a complement to core holdings in well-diversified, actively managed investment approaches.
For frequent updates on artificial intelligence and CIO’s views in this fast-evolving corner of the market, please subscribe to our regular Intelligence Weekly publication (latest edition here.)
With thanks to Sundeep Gantori (UBS Chief Investment Office) for his contribution.
