The divergence between surging Chinese AI stocks and a slumping Hong Kong tech index widens, as investors chase specific "national champion" names over broad market exposure.
The divergence between surging Chinese AI stocks and a slumping Hong Kong tech index widens, as investors chase specific "national champion" names over broad market exposure.

Shares of Hong Kong-listed AI developers surged, with Knowledge Atlas Technology (Zhipu) climbing 30 percent and MiniMax advancing over 15 percent, signaling intense investor appetite for specific AI leaders even as the broader Hang Seng Tech Index languishes.
"Investors choose to express their views on the China AI story through specific names, rather than the broader index," Jason Lui, head of Asia-Pacific equity and derivative strategy at BNP Paribas, said during a recent briefing.
The gains extend a dramatic run for newly-listed AI firms, with Zhipu having surged more than sevenfold since its debut and MiniMax quadrupling from its listing price. The rally comes as Hong Kong’s IPO market raised a combined $13.3 billion in the first quarter, outpacing the Nasdaq and NYSE combined, according to London Stock Exchange Group data. In contrast, the Hang Seng Tech index was down 11.7 percent for the year as of April 30.
The performance gap highlights a strategic shift where capital is rotating from large internet platform companies into "hard tech" AI firms. Morgan Stanley estimates Zhipu and MiniMax could join the Hang Seng Tech Index as soon as June, potentially attracting $1.75 billion in inflows from passively managed funds and further cementing their role in Asia's evolving tech market.
The rally in specific AI names is part of a wider listing boom on the Hong Kong Stock Exchange. In the first quarter of 2026, 38 companies raised a combined $13.3 billion, the bourse’s busiest first quarter since 2021. The list of high-flying debutants includes AI software firm Manycore Tech, which rose 185 percent on its first day, and Lightelligence, whose shares were over 5,700 times oversubscribed.
The frenzy stands in sharp contrast to the muted reception for technology IPOs in the United States. According to data from PitchBook, the median U.S. IPO has underperformed its benchmark by a record 42 percentage points within 120 days of listing over the past year. Roughly two-thirds of companies that went public since the start of 2025 are trading below their IPO prices.
This divergence is fueled by concerns that AI poses a displacement risk to traditional software-as-a-service (SaaS) business models, a phenomenon some analysts have dubbed a "SaaSpocalypse." While U.S.-listed AI infrastructure firm CoreWeave has soared, other venture-backed listings like Figma and Klarna are down more than 65 percent from their debut prices.
“Structurally, these companies came to market at more digestible valuations relative to their growth profiles compared to U.S. tech IPOs, which have repeatedly disappointed at high entry multiples,” said Harrison Rolfes, a senior research analyst at PitchBook. He added that a "national champion premium" among investors in Asia has also bolstered demand for the Chinese firms.
The improved sentiment toward Chinese AI can be traced to the early 2025 launch of DeepSeek’s V4 model, which demonstrated capabilities that could challenge U.S. rivals like OpenAI. Analysts say the breakthrough is poised to drive demand for domestic computing power and accelerate commercial adoption, directly benefiting model developers like Zhipu and MiniMax.
DeepSeek’s V4 was optimized for domestically made chips, a key factor for reinforcing China’s technology stack under U.S. sanctions. The launch is expected to spur demand for high-performance chips from mainland producers like Hygon Information Technology and Moore Threads Technology, and benefit fab operators such as Semiconductor Manufacturing International Corp (SMIC).
For investors, the outperformance of Hong Kong-listed AI companies challenges the assumption that U.S. markets offer the best exposure to the AI boom. "The broader takeaway," Rolfes said, "is that Chinese AI has likely graduated from a risk to monitor to a market to understand.”
This article is for informational purposes only and does not constitute investment advice.