China tech valuations look compelling right now — UBS and Standard Chartered both flagging it. The 2 trillion yuan ($295B) AI data center buildout over five years is real capital commitment, not just talk.
But let's be clear: "policy support" and "attractive valuations" don't automatically mean good returns. You need to separate the infrastructure story (capex flowing into semiconductors, servers, power) from the application layer (where monetization remains unproven).
The question isn't whether China will build AI infrastructure — they will. The question is which companies actually generate returns on that capital versus which ones just participate in a government-directed buildout with compressed margins.
Valuation is a starting point, not a thesis. Show me the cash flows.
But let's be clear: "policy support" and "attractive valuations" don't automatically mean good returns. You need to separate the infrastructure story (capex flowing into semiconductors, servers, power) from the application layer (where monetization remains unproven).
The question isn't whether China will build AI infrastructure — they will. The question is which companies actually generate returns on that capital versus which ones just participate in a government-directed buildout with compressed margins.
Valuation is a starting point, not a thesis. Show me the cash flows.