print-icon
print-icon
Add ZeroHedge as a preferred source on Google

UBS Identifies "The Next Leg Of The AI Trade"

Tyler Durden's Photo
by Tyler Durden
Authored...

Catherine Gordon, UBS's energy and resources expert, highlighted "AI Infrastructure Constraints Emerge As Next Leg Of The Trade" in a Friday note to clients. The theme closely aligns with our "Owning The Bottlenecks" note, which centered on companies positioned to benefit from supply constraints in the critical materials space that feed not only the AI infrastructure buildout but also the rearmament supercycle.

Gordon's note outlined that traders are shifting their focus to the power, data-center capacity, and financing needed to sustain the artificial intelligence buildout, as infrastructure constraints emerge as a potential driver of the next leg of the trade.

Gordon explained:

The AI narrative is increasingly shifting from model development toward the infrastructure needed to support demand, with investors focusing on whether enough power, data center capacity and financing can be deployed to sustain the buildout over the coming years.

While questions around OpenAI revenue estimates briefly introduced some uncertainty into the AI trade, the broader market takeaway remains intact: demand is not the primary concern. Instead, attention is gravitating toward the supply side of the equation, particularly power availability, which is emerging as a critical constraint. Recent long-term power purchase agreements by Alphabet and Amazon have reinforced that theme and helped keep infrastructure names at the center of the discussion.

Sentiment across the AI complex remains somewhat fragile amid ongoing debates around capex sustainability and infrastructure funding, but industry feedback continues to point to a constructive demand backdrop.

Conversations at Data Center World Expo in Singapore were largely focused on the ability of the supply chain to ramp capacity quickly enough rather than any meaningful deterioration in end demand, reinforcing the view that execution and infrastructure deployment are becoming the key variables for the next phase of the AI trade.

Gordon's view of the shifting AI trade, focusing on the building blocks that make the AI data-center buildout possible, aligns with a theme we laid out in late September: "owning the bottlenecks."

We recently noted that the Trump administration is accelerating efforts to rebuild conflict-free supply chains and reduce dependence on China for critical materials. The point here is to own the producers outside China that can deliver conflict-free critical materials today to supply power grids, reindustrialization, data centers, and the rearmament cycle, such as rare earths and industrial metals, including tungsten. 

This comes as resource nationalism, sparked by China, shows no signs of stopping anytime soon.

0