As hardware technology stocks closely tied to AI computing power infrastructure continue to rally in China’s stock market, the STAR 50 Index (000688), which covers the country’s top local chip enterprises, has posted a staggering 64% gain this quarter—rivaling the Philadelphia Semiconductor Index, which includes Nvidia (NVDA) and Micron (MU). Surging global spending on AI computing by hyperscale cloud providers, combined with domestic policies fostering local tech champions, has jointly fueled this rally. At the same time, investor capital rotating out of A-share consumer and retail sectors has further amplified the strong performance of technology stocks.
While there is little skepticism among investors regarding the long-term narrative of China’s technology supply chain, the sharp short-term share price surge has sparked concerns among some investors. As the earnings season kicks off with preliminary announcements, tech companies now face the test of justifying this rally. A fund manager at Hangzhou Xiyan Asset remarked that market sentiment is nearing a short-term peak, with strong growth expectations for many AI hardware leaders already priced in, making it increasingly difficult to rely solely on fundamentals to drive further upside. Capital has already spilled over into every corner of the industrial chain, including sub-sectors such as electronic fabrics, copper foil, and glass substrates—a move interpreted by some as a sign of waning confidence in chasing leading names at elevated levels.
Take optical module leader Zhongji Xuchuang as an example: the stock has surged 132% this quarter, pushing its forward price-to-earnings ratio above that of the Dow Jones U.S. Telecom Equipment Index, while its share price now trades above the average analyst price target. Institutional data indicates that although analysts have slightly raised aggregate earnings forecasts for the STAR 50 Index, the magnitude of the upward revisions falls far short of the index’s gains, driving its forward P/E ratio to a new high of 69x.
Nonetheless, most mainstream institutions remain steadfastly bullish. Both Goldman Sachs and Morgan Stanley favor A-share hardware companies, citing their more pronounced “hard-tech” attributes. Goldman Sachs pointed out that Chinese household equity allocation accounts for less than 10% of total assets. With the wealth effect from real estate diminishing and deposit rates trending lower, household savings are expected to gradually shift toward the stock market, forming a “triple support” of AI earnings growth, household asset reallocation, and policy backing. Additionally, regulators’ commitment to easing listing standards for AI enterprises is viewed as a strong policy signal.
Some companies have already previewed strong earnings results—Satellite Chemical and Changchuan Technology both announced significant profit growth, sending their shares sharply higher. However, market divergence is equally evident: the Hang Seng Tech Index declined this quarter, largely due to its internet giants lagging behind their North American peers in AI application ecosystem development.
UBS emphasized that exiting now is premature, as the core fundamentals—earnings and order books—have yet to weaken. China’s AI hardware stocks are expected to deliver 80% earnings growth, with order visibility extending through the end of 2027. Goldman Sachs also believes that the global AI bull market is transitioning from the “chip purchasing” phase to the second stage of “building AI factories,” where excess returns will spread across the full-stack infrastructure landscape, including high-performance CPUs, storage, liquid cooling, optical interconnects, advanced packaging, and upstream materials.
China’s AI computing power rally is entering a critical phase where risks and opportunities coexist. Valuations are no longer cheap, and subsequent gains will need to be underpinned by real orders, technological moats, and earnings cash flow—rather than being indiscriminately driven by the grand narrative of “domestic substitution” alone. The year 2027 could become a true stress test, when intensifying competition weeds out high-beta small-cap names. However, niche leaders that control critical bottleneck nodes in the computing power chain—such as optical modules, computing chips, high-end PCBs, semiconductor equipment, and power infrastructure—still have the potential to become the most sustained core theme in China’s asset revaluation, driven by continued household capital reallocation and policy anchoring.
