Hong Kong Sees Largest Capital Raise in Five Years as Chinese AI IPOs Drive $44 Billion

The Hong Kong Stock Exchange closed the first half of 2026 with $44 billion raised through IPOs, placements, and block trades, according to data compiled by Bloomberg and published on Sunday, June 28. This figure represents a 29% increase over the same period in 2025 and marks the largest volume in five years, in a window where the secondary market in Hong Kong remains technically weak and the regulatory environment continues to pose challenges for Chinese companies listed abroad.

The explanation lies in the demand for exposure to artificial intelligence. Chinese corporate giants have led the way: Contemporary Amperex Technology (CATL), the world's largest battery manufacturer, and Victory Giant Technology Huizhou, specializing in printed circuit boards for AI servers, have executed multibillion-dollar offerings that anchored the semester. The capital that has returned to Hong Kong is, in part, money that three years ago would have gone to New York and is now facing political restrictions to get there.

The AI Lineup Ahead

Zhipu, one of China's leading developers of foundation models, is considering a multibillion-dollar listing in Hong Kong following a 2,000% valuation increase in the private market over the past twelve months, according to Bloomberg on June 24. The current open window explains the urgency: the combination of institutional demand eager for AI names and a newly heated primary market creates rare conditions for Chinese issuers to price at a premium.

For global investment banks, the trend has immediate implications. Goldman Sachs, Morgan Stanley, and JPMorgan are vying for coordination mandates in Hong Kong at a pace not seen since 2020. Revenue from equity capital markets in the Asian financial center could double in 2026 if the semester's pace continues, according to preliminary calculations from buy-side analysts cited in specialized coverage.

The Perspective in the United States

The movement also reshapes the landscape for Nasdaq. Chinese AI companies that could have sought dual listing have opted to focus on Hong Kong, and the risk for New York is structural: once the framework of sell-side coverage, market making, and indices consolidates around the Hong Kong-Singapore axis for Chinese AI names, reversing the flow becomes costly. For American funds, the only way to capture exposure is through offshore vehicles or the remaining ADRs, which adds a layer of cost and operational risk.

The contrast with the scenario outlined this same weekend by the Bank for International Settlements is striking. While the BIS warns of the risk of an "AI bubble" and circular financing between hyperscalers and labs, a significant portion of global institutional capital is accelerating, not slowing down, its allocation to the sector. The divergence between the regulatory tone and market appetite is likely to increase volatility when the next repricing occurs.

What This Means for CFOs in Brazil

For Brazilian corporate treasuries, the second-order effect comes via funding costs. The flow concentrated in AI offerings in Hong Kong raises the relative cost of capital for emerging market issuers competing for the same institutional dollar. Companies like Vale, Petrobras, and Brazilian banks aiming to access Asian investors in the second half now face competition from a pipeline of IPOs offering a much more aggressive growth narrative, albeit with questionable fundamentals.

The implication for CFOs is tactical. Delaying an issuance until January may seem prudent, but if the AI window closes abruptly, institutional flow could migrate back to defensive names and the Brazilian window improves. Timing needs to be monitored weekly, not quarterly, and the technical reading of the IPO book in Hong Kong becomes a useful leading indicator.

Why CATL and Victory Giant Matter Beyond China

The significance of CATL and Victory Giant in the semester's composition reveals a point that the aggregate numbers obscure. It is not the debut of small model labs moving the ticker; it is the capital raise of companies with heavy operational revenue, positioned in critical layers of the AI stack: energy storage for data centers in CATL’s case, and physical infrastructure for acceleration servers in the case of Victory Giant. This type of issuer is more resilient in a potential repricing than model-pure startups.

For institutional investors, this means that the AI thesis in Hong Kong rests on a foundation of real cash flow, not merely on prospective multiples. For the American regulator, it indicates that the competition for a share of AI capex has shifted from just about talent and chips to openly include the industrial infrastructure that supports the sector. The next competitive front is less about who trains the model better and more about who owns the physical skeleton that sustains it.