Franklin Templeton Trims AI Positions, Shifts to Overweight Chinese Internet Stocks
Alina Collins
Franklin Templeton's lead EM fund manager — overseeing $14 billion+ — has cut his AI supply-chain overweight to neutral and shifted into Alibaba, Tencent, and other Chinese internet platforms, arguing that chip stocks are overpriced while the companies actually monetizing AI remain undervalued.
Why is a fund beating 92% of peers cutting its AI bet?
Chetan Sehgal's Templeton Emerging Markets fund has returned 31% this year, beating 92% of peers, per Bloomberg data.
The memory-chip stocks that drove those gains are no longer cheap. His active overweight on the AI supply chain — meaning he deliberately held more than the benchmark — has been cut sharply to neutral.
This means → the manager who profited most from the AI hardware trade is signaling that the "easy money" phase is over.
Where is the money going — and why Chinese internet?
Sehgal is now overweight Alibaba, Tencent, and South Korea's Naver. His logic: chipmakers' output ultimately flows to internet companies that build AI models and run cloud businesses.
In plain terms = the shovel-sellers have had their run; now it's the turn of the people actually digging for gold — and internet platforms are those people.
His words: "We are overweight the internet companies that are actually developing these models and delivering them to customers."
Is the market confirming this call?
SK Hynix, the poster stock for this year's AI rally, has fallen 30% since July after Korean regulators cracked down on retail speculation.
Meanwhile Alibaba and Tencent have been climbing. The Hang Seng Tech Index is up 8% this month; Alibaba alone has bounced more than 20%.
This reflects a rotation from hardware supply chain to internet platforms — not a cooling of the AI theme itself.
How does Chinese internet spend on AI differently from the US?
Sehgal notes that Chinese internet companies' AI spending is far smaller than their US and Korean peers. Their strategy leans toward embedding AI into existing platforms, not building massive new infrastructure from scratch.
In plain terms = US firms spend hundreds of billions on data centers first and figure out revenue later; Chinese firms plug AI into e-commerce, social, and cloud services they already run — and monetize immediately.
Sehgal sees this restraint as an advantage: "Even accounting for their committed spending increases, valuations remain reasonable."
What is the deeper logic behind this rotation?
Global capex on AI infrastructure now runs into the trillions of dollars. The market is increasingly questioning whether returns can justify that scale of investment.
This means → once the certainty around the "build-out phase" starts to erode, capital naturally flows toward companies that can already turn AI into revenue — and that is the deeper driver of the chip-to-platform rotation.
The key signal for ordinary investors is not "AI is over." It is that the center of gravity in AI investing is shifting from selling the hardware to using it.
Content is for reference only, not financial advice.