China’s AI hardware export boom offers hedge against Wall Street’s “dangerously circular” trade: deVere CEO
China’s new wave of AI hardware exports offers investors a hedge against Wall Street’s “dangerously circular” artificial intelligence trade, affirms the CEO of one of the world’s largest independent financial advisory and asset management organisations.
Nigel Green of deVere Group comments as fresh analysis maps out a new generation of Chinese exporters carrying a combined addressable market of $12bn to $212bn by 2030, a shift that gives investors a genuine alternative to the self-financing loops he has been warning about for weeks.
“For weeks I’ve been flagging that too much of the AI trade in the US is financing itself, with the same capital moving between a tight circle of suppliers, lenders and customers and getting booked as growth every time it changes hands.
“A supplier invests in a customer, that customer spends the investment buying the supplier’s own hardware, and the identical dollars get counted as fresh revenue at every stop along the loop,” he says.
“China’s export wave looks nothing like that. It is hardware shipped to real ports and installed on real factory floors overseas, demand that shows up in customs data rather than in vendor-financing footnotes.”
He describes China as entering “a third global export era”, moving on from the solar panels, batteries and electric vehicles that defined the last decade toward AI-enabled industrial hardware, data-centre equipment and automated production systems.
With US trade restrictions still unresolved, Chinese manufacturers across these sectors are routing growth through Europe and Southeast Asia instead.
Industrial automation and robotics stand out as sectors where corporate execution, not macro tailwinds, decides the winners.
Two listed Chinese automation specialists illustrate the range on offer. One Shenzhen-listed group carries a buy rating and a price target implying upside above 50%, with its addressable market outside China projected to more than triple from $163bn as Europe becomes its central overseas battleground.
A Hong Kong-listed robotics manufacturer, rated neutral, is expected to triple its Southeast Asian market share to around 10% by 2030 against a regional opportunity worth $20bn.
“Those two firms sit at very different stages of the same trade,” explains the deVere CEO. “The Southeast Asian push still needs deeper service coverage on the ground, while the European operation already has real scale behind it. Execution speed is what separates them, not a weaker currency.”
The scale of China’s hardware momentum is visible beyond the balance sheets. One Chinese humanoid robot maker shipped more than 5,500 units last year, compared with roughly 150 units each from its leading American rivals.
Beijing hosts the World Robot Conference from August 19 to 23, with exhibitor numbers up 36% year on year to more than 300 and over 150 products launching on-site.
“Volumes like that are the proof.
“An entire supply chain is moving up the value curve at a pace Western competitors are still underestimating, and this week in Beijing will show it in real time.”
Nigel Green argues the diversification case is now urgent for portfolios overweight the most circularly financed pockets of the American AI trade.
“Investors who have been sitting entirely inside the US AI complex are exposed to financing structures I have already flagged as fragile,” he says.
“Adding exposure tied to genuine overseas order books, across robotics, automation and industrial hardware, gives a portfolio real distance from that circularity. Every AI-heavy portfolio needs that distance before the next repricing arrives.”
He concludes that the coming week’s robotics showcase in Beijing will offer the clearest signal yet of which Chinese exporters are converting momentum into contracted, overseas revenue.
“Watch the order books coming out of Beijing this week closely.
“They’ll tell investors more about where real AI demand sits than another earnings call stuffed with financing arrangements ever could.”