UBS: China's AI Trading Congested but Fundamentals Strong, HKTECH Difficult to Outperform in Short-Term

UBS released a China stock strategy report on May 21, addressing the five most pressing concerns of investors recently one by one.

Overall, market sentiment is leaning towards optimism, with the strong earnings performance of hardware technology companies leading investors to largely digest the impact of the closure of the Strait of Hormuz, and AI hardware has become a consensus long position among institutional investors.

How crowded is AI trading?

The conclusion given by UBS is: crowded, but not newly so, and there has been no significant deterioration recently.

Looking at domestic public mutual funds' holdings, the proportion of the electronic sector fell slightly from a peak of 22% in the third quarter of 2025 to 20% in the first quarter of 2026, but the proportion of the telecommunications sector (also related to AI) rose to a historical high of 11%, with the combined total configuration close to a historical high, remaining essentially stable over the past three quarters.

At the individual stock level, the crowding score of hot AI stocks has risen significantly over the past 12 months but has slightly receded from the high point at the beginning of May. In terms of market activity, the transaction volume of the TMT sector accounts for about 37% of the total transaction volume of A-shares, which is below the historical peak of 43% in 2023.

Furthermore, the recent continuous outflow of foreign capital from Taiwan and South Korea's stock markets also reflects a certain degree of profit-taking by global investors in AI-related markets, and emerging market funds are still underweight in North Asia markets.

What factors could end this AI trend?

UBS admits it has no definitive answer. The correlation between Chinese and American AI technology stocks has risen significantly this year, and the global market spillover effect is the biggest downside risk UBS sees.

Several signals that investors might reassess their AI positions include: the continued upward movement of US Treasury yields (although the correlation between Chinese hardware technology stocks and US Treasury yields has historically been close to zero); a substantial slowdown in profit growth (with profit growth in the first quarter still at 64%, the momentum is strong); and the listing of large US AI language model companies, which some investors might see as a signal that the currency-ization of LLMs has peaked.

For investors concerned about the crowding of the technology sector, UBS suggests focusing on China's power equipment and energy storage sectors, as these assets can both indirectly benefit from the global construction of AI data centers and hedge against the risks of high-energy prices, and are also undervalued relative to international peers.

Will rising PPI compress corporate profits?

This is a common concern among investors, but the research from UBS gives a counter-intuitive answer: historically, rising PPI often corresponds to faster revenue and profit growth of listed companies on the A-shares, which was also the case during the Russia-Ukraine war in 2022.

Specifically, sectors such as energy and capital goods directly benefit; export-oriented enterprises often have an excuse to raise prices in an import inflation environment, with a historical correlation of 0.82; sectors with strong pricing power, such as shipping and consumer durables, also perform well; midstream manufacturing sectors such as chemicals, automobiles, and tech hardware can partially pass on costs; semiconductors have a lower correlation with PPI due to their unique profit cycle; food and beverage, and utility sectors have weak pricing power and are hit the hardest, but these sectors only account for about 10% of the CSI 300.

However, UBS also notes that there is still uncertainty about whether China can continue to move out of deflation. The high and persistent youth unemployment rate, coupled with a slowdown in enterprise wage growth in the second half of 2025, indicates a weak labor market that cannot support a broader consumer recovery in the short term.

What about HSTECH?

The Hang Seng Technology Index has fallen about 10% year-to-date, lagging the MSCI Global Index (which is close to historical highs) by about 20 percentage points.

UBS attributes this to a combination of factors: the internet, automotive, and consumer sectors together account for about 70% of the index, each facing different headwinds; the AI narrative has shifted from software (viewed as losers) to hardware (beneficiaries), suppressing the valuation of Chinese internet stocks; earnings forecasts have been cumulatively reduced by 37% over the past 12 months, primarily due to competition in food delivery among Meituan, JD, and Alibaba; the slowdown in south