Bank of America Steps Back from Moonshot AI’s Hong Kong Listing
TREE NEWS reports: Bank of America has withdrawn from negotiations to join the underwriting syndicate for a potential Hong Kong initial public offering by Chinese artificial intelligence developer Moonshot AI, the company behind the Kimi chatbot. People familiar with the matter said the U.S. lender had discussed serving as an overall coordinator for the deal but was ultimately not formally appointed. Both Bank of America and Moonshot AI declined to comment.
The listing is expected to arrive as early as the first quarter of 2027 and could raise as much as $5 billion, a figure that would place it among the largest Hong Kong offerings in recent years and among the most significant capital markets events for a Chinese AI company to date.
Why a U.S. Bank Walked Away
The retreat is less a verdict on Moonshot’s business than a reflection of the widening compliance maze facing Western banks in China-linked technology deals. U.S. financial institutions must weigh outbound investment screening, export-control sensitivities around advanced computing, and the reputational and political scrutiny attached to backing Chinese frontier AI. Underwriting a marquee AI listing puts a bank’s name on the front page of a geopolitical story it may prefer to avoid.
Hong Kong’s IPO pipeline has already tilted toward Chinese issuers returning from New York or seeking a domestic-adjacent venue, and the mandates increasingly flow to Chinese and Hong Kong underwriters with fewer cross-border constraints. That shift matters for price discovery: a syndicate without bulge-bracket U.S. participation may lean more heavily on mainland and regional institutional demand.
The Crypto and RWA Read-Through
For digital-asset markets, the signal is subtler but real. Moonshot’s core asset is compute-intensive model training, and the crypto industry has spent two years building parallel infrastructure — decentralized GPU networks, inference marketplaces, and tokenized compute credits — that competes for the same capital and talent pool. A $5 billion AI listing would benchmark public-market valuations against those private and token-based alternatives.
- Compute as an asset class: If public investors reward Moonshot’s model economics, tokenized compute and GPU-dePIN projects gain a comparable valuation anchor.
- Capital rotation: Large AI IPOs absorb institutional risk appetite that might otherwise flow into crypto venture rounds.
- Jurisdictional arbitrage: The underwriting shuffle underscores how AI, like crypto, is being carved along geopolitical lines.
What to Watch
Three things will determine how this story develops: whether Moonshot confirms a filing timeline, which banks ultimately anchor the syndicate, and whether U.S. institutions re-enter on a secondary or advisory basis. For crypto observers, the more useful question is whether the AI capital cycle continues to pull liquidity away from digital assets — or whether tokenized compute becomes the bridge that lets the two markets share the same institutional balance sheet.
