The New Capital Cycle.

“History rarely rewards the tenants of a new economic system. It rewards the landlords.”

Wall Street believes it is investing in artificial intelligence. It is not. It is investing in the last phase of the AI trade. The next decade will not be won by the companies with the largest language models, the highest chatbot adoption or the most downloaded applications. It will be won by those who own the infrastructure without which artificial intelligence cannot exist.

This is where consensus is becoming dangerously complacent. Markets remain obsessed with software valuations while quietly overlooking the strategic assets that make intelligence possible: advanced semiconductors, memory, networking, energy systems, robotics, autonomous mobility, sovereign AI infrastructure and defense technologies. The debate over whether AI is expensive has become irrelevant. The real question is where capital is migrating next.

Today’s headlines — from Luxshare’s IPO and Kioxia’s inclusion in major indices to Meta’s restructuring, Europe’s AI competitiveness challenge, Thales’ acquisition of Exail, Volkswagen’s strategic crisis and the accelerating transformation of modern warfare — are not isolated events. They are different expressions of the same phenomenon. This is not another technology cycle. It is not another semiconductor cycle. It is not even another AI cycle. It is the beginning of a new capital cycle — and the Infrastructure of Intelligence is becoming the world’s most valuable asset class.

1. 01 The AI trade is moving down the stack

   For two years investors rewarded companies building AI models. The next phase will reward those supplying the physical foundations of intelligence. Luxshare’s IPO, Kioxia’s strategic importance, Unimicron’s capital raising and the expanding semiconductor ecosystem all point the same way: the bottleneck is no longer algorithms — it is industrial capacity. Memory, advanced packaging, networking, power electronics, cooling, energy, precision manufacturing. That is where pricing power is migrating.

   Investment implication The market is broadening beyond hyperscalers toward the companies that enable accelerated computing at industrial scale.

2. 02 Europe risks financing the AI revolution instead of leading it

   Europe’s macro backdrop keeps improving — inflation moderating, rates less restrictive, activity stabilizing. Yet it still lacks sufficient AI-native industrial champions. Without leadership in accelerated computing, sovereign AI infrastructure, advanced manufacturing and intelligent mobility, Europe risks becoming the customer of the next industrial revolution rather than one of its architects. Macroeconomic stability cannot compensate for technological dependency.

   Investment implication Europe’s next winners will build strategic AI infrastructure — not simply benefit from lower interest rates.

3. 03 Artificial intelligence has entered the income statement

   Meta and TikTok’s workforce reductions are not isolated restructuring announcements — they are the first visible wave of AI-native operating models. AI is moving from experimentation to execution, from productivity promises to cost structures, from pilot projects to permanent organizational redesign. Companies that compress decision cycles while expanding operating leverage will increasingly command valuation premiums.

   Investment implication Markets will begin rewarding AI execution rather than AI narratives.

4. 04 Defense has become an accelerated-computing industry

   Thales’ acquisition of Exail illustrates one of the market’s most underestimated structural shifts: defense is becoming an AI infrastructure business. Autonomous systems, AI-enabled sensors, space technologies, navigation, quantum positioning, robotics, underwater autonomy. The future battlefield is increasingly defined by software running on intelligent machines — making defense spending one of the largest long-term capital-allocation programs supporting intelligent infrastructure.

   Investment implication Defense deserves to be valued as an advanced technology platform, not merely a traditional industrial sector.

5. 05 Volkswagen is not the exception — it is the warning

   Volkswagen’s difficulties are often explained through slowing EV demand or Chinese competition. That reading is incomplete. The real challenge is structural: the automotive industry is evolving into an intelligent mobility platform where software, batteries, accelerated computing, autonomy, robotics and continuous data determine advantage. Scale without intelligence becomes fixed cost; scale without velocity becomes strategic inertia. Every legacy industry now faces the same test.

   Investment implication The market will distinguish between companies digitizing legacy operations and companies rebuilding themselves as AI-native enterprises.

Every headline today points toward the same conclusion. Capital is no longer rewarding digital transformation — it is rewarding ownership of strategic infrastructure. The market’s center of gravity is shifting from applications to architecture, from software to systems, from platforms to infrastructure, from consumption to production. This distinction will define the next decade.

The companies likely to capture the greatest economic value are those controlling the strategic assets below — this is where the next trillion dollars of enterprise value will be created.

The new scarce assets

Who owns the Infrastructure of Intelligence

• Advanced semiconductor manufacturing.
• High-bandwidth memory and storage.
• AI networking and interconnect technologies.
• Electricity generation and grid modernization.
• Data centers and AI factories.
• Industrial robotics and automation.
• Autonomous mobility systems.
• Defense and dual-use technologies.
• Sovereign AI infrastructure.

Every structural transformation creates a period during which markets continue valuing yesterday’s business models while tomorrow’s winners quietly accumulate strategic advantage. This appears to be one of those periods. Investors continue asking whether AI valuations have become excessive. They should instead ask a far more important question: who will own the Infrastructure of Intelligence?

That question — not quarterly earnings revisions, not short-term rate expectations, not temporary market volatility — will determine the dominant investment winners of the next decade. The market is still chasing artificial intelligence. The smartest capital is already accumulating the infrastructure upon which artificial intelligence depends.

Key takeaways

What to carry into the next session

• The AI trade is maturing. The Infrastructure of Intelligence trade is just beginning.
• Capital is rotating away from AI applications toward the strategic assets that make intelligence scalable.
• Memory, advanced manufacturing, networking, energy and intelligent infrastructure are becoming the new scarce assets.
• Europe’s greatest challenge is not economic growth but technological sovereignty.
• Defense has become one of the world’s fastest-growing accelerated-computing industries.
• Legacy companies will not fail because they lack scale; they will fail because they lack velocity.
• The next trillion-dollar winners will own the infrastructure that powers intelligence — not merely the software that demonstrates it.
• The market is still pricing AI as a technology story — not yet as the largest capital-allocation revolution of the twenty-first century.

That is the rotation almost nobody is pricing. That is The New Capital Cycle.