A paradox with no easy exit: some of China’s best companies — technically strong, reliably profitable — spend years failing to make their holders money. Pinduoduo’s algorithms move merchandise without pause, a cold and stable selling machine, yet the stock stays pinned to the floor. The roots of the Chinese listings have moved back to Hong Kong, southbound money buys every day, and still one jolt on Wall Street gaps the market lower at the next open. The companies are Chinese. The pricing is someone else’s.
The Front Lines of Pricing Power Are Segmented
Start with the most counterintuitive fact of the capital-flow ledger: in 2025 and into Q1 2026, southbound money’s quarterly net buying at one point exceeded 220 billion Hong Kong dollars, a historic record. With that much money, why hasn’t pricing power changed hands? Because the front is segmented.
On the high-dividend segment, mainland money has already won. State enterprises like China Mobile, ICBC, and CNOOC — undervalued and heavy payers — have been bought through Stock Connect until mainland funds hold over half the shares, and Wall Street has been pushed out of the pricing seat entirely. The technology-and-internet segment is another war: Tencent, Meituan, Alibaba, Pinduoduo were originally sponsored by Wall Street’s top banks, and long-only funds and sovereign vehicles still hold over forty percent of the float. Southbound buying often functions as mere backstop; the moment a large foreign fund clears out for macro reasons, the price caves. More money does not equal more power — it depends which segment the money stands on.
The Night Session Writes the Morning’s Price
One more mechanism binds Hong Kong to New York. Alibaba, JD, and their peers list in Hong Kong, but enormous ADR volume still trades in the US. The same share converts freely across both markets, and arbitrage enforces the link: daytime Beijing time, Hong Kong trades — Stock Connect and Asia-Pacific money; late-night Beijing time, New York trades BABA, PDD, BIDU. Wall Street’s panic over tariffs, its skepticism about AI monetization, is written into the ADR close that same night, and the arbitrage machinery carries the price untouched into Hong Kong’s next morning. The Chinese listings are twins wired to a heartbeat monitor in New York. While the derivatives and short positions remain on Wall Street’s books, Hong Kong cannot unilaterally cut the cable.
The Discount Is the Settlement Sheet
Pinduoduo’s triple discount now reads legibly. No buybacks, no dividends — cash reserved as winter rations and ammunition — which Wall Street reads as disrespect for capital, so long-only holders walk. The sustainability of the AI-driven low-price miracle carries a question mark: supply-chain squeezing has limits, and Western lobbyists push to close the de minimis exemption that Temu rides. Heaviest of all is the geopolitical label: many foreign institutions do not doubt the company’s quality; their compliance mandates simply say cannot buy, must sell. Good technology and fat profits no longer purchase a high multiple by themselves.
The counterweight is forming, and it has a shape. Middle Eastern sovereign capital is accelerating its return to Hong Kong, ignoring Wall Street’s baton, hunting industrial synergies in hard tech and new energy. Tencent and Xiaomi spend enormous cash on daily buybacks and cancellations — whatever the foreigners dump, the companies absorb with real money. The dividend segment has already been retaken; technology remains contested. The handover will be long and grinding.
For a holder, the tug-of-war yields one clean selection rule: profits alone cannot buy a share price, but the act of returning cash to shareholders can. Standing on names where foreign exit is complete and the company keeps cancelling its own shares means standing on the windward side of the handover. Otherwise, however thick the profit, you are only guarding trophies for someone else. The boundary, drawn as always: every company named above — PDD, Tencent, Meituan, Alibaba and their peers — is a sample of capital behaviour, not a recommendation list. What this piece offers is the coordinate frame of a pricing-rights handover: no stock picks, no price targets, no call on this week’s tape. What a name is worth and when to act are two other questions entirely.
