Funded. Profitable. Accelerating.
Against this tsunami, day-to-day market noise is ripples on the surface. We invest in the hardest-to-replicate physical bottlenecks that power the AI infrastructure buildout — and we are disciplined about any company whose stock price already assumes a smooth, uninterrupted expansion.
Hyperscalers, sovereigns, and defense establishments are committing trillions of dollars to AI infrastructure. These are not discretionary programs — they are strategic necessities that compound on themselves.
The artificial intelligence infrastructure buildout represents the largest single capital expenditure cycle in modern industrial history. Hyperscalers have committed hundreds of billions annually — not as an option, but as a competitive requirement with no exit. The industrialization of intelligence has begun.
Against this tsunami, daily volatility is ripples on the surface. We repeatedly see new all-time highs led by AI infrastructure companies providing the scarce physical inputs that feed the buildout. The trend is the signal — not the noise around it.
Every layer of the AI stack faces physical constraints that capital cannot resolve overnight. The bottleneck moves — from power to packaging to photonics to materials — but it never disappears. Identifying where it lives today, and where it moves next, is the investment opportunity.
The best holdings will not simply be "AI beneficiaries." They are companies tied to the physical constraints that cannot be resolved with software or capital alone.
What We Seek
What We Avoid
Every holding maps to one of five physical bottleneck themes that span the AI infrastructure value chain from raw materials to compute delivery. The portfolio also holds a small number of opportunistic positions — companies making a sudden shift into the AI supply chain or with significant structural AI exposure — outside of the five core pillars.
Power delivery, grid transmission, thermal management, data center construction, and utility-scale infrastructure. The AI buildout is first and foremost a power story.
Fiber cable, optical transceivers, photonic integrated circuits, and connectors. The data center spine that moves model traffic at the speed of light.
Wafer fabrication, EDA tools, process equipment, advanced packaging, and specialist logic. The geometry of AI compute increasingly lives in the packaging layer, not just the die.
Specialty process chemicals, photoresists, CMP slurries, substrates, and advanced polymers. Foundational to every wafer ever made — the least-discussed constraint in the stack.
Servers, hyperscale networking, storage, switching, and systems integrators. The hardware layer that assembles compute, memory, and interconnect into deployable AI infrastructure.
The portfolio is constructed with a set of principles designed to maximize participation in the structural AI buildout while maintaining discipline on valuation and position concentration.
Each position is assigned equal weight regardless of the size of the investment. Capital from any investor is deployed equally across all active positions at the time of entry. This eliminates size bias and forces systematic exposure to smaller, faster-moving parts of the supply chain.
Pure equity conviction in the structural AI infrastructure buildout. No short positions, no margin, no derivatives. The buildout has multi-year duration — cyclical hedges would reduce the return profile without a commensurate reduction in fundamental risk.
Five value-chain pillars define the portfolio. Selection is based on physical scarcity and bottleneck position — not quantitative screens, factor exposures, or market capitalization weighting. Every holding must earn its place through bottleneck logic.
Positions are trimmed on strength when valuation catches up to the thesis. New capital is deployed into underrepresented bottlenecks on weakness. The portfolio is not buy-and-hold — it reflects a continuously updated view of where the constraint lives in the stack.
Presvic LLC is the manager of the FourthTurning Global Capital Fund, a concentrated long-only equity portfolio focused on the physical infrastructure of artificial intelligence.
The fund's name reflects the Strauss–Howe generational theory of cyclical history — the Fourth Turning thesis identifies periods of institutional upheaval and accelerated capital reallocation that recur approximately every eighty years. We believe the AI infrastructure buildout is the defining capital expenditure cycle of the current turning: a multi-decade transformation funded by real earnings, not speculative capital.
We invest globally across the full AI infrastructure value chain — from specialty process chemicals to fiber transceivers to power transformers — targeting the companies that supply the physical inputs the buildout cannot proceed without. The portfolio holds positions in companies listed on U.S., European, Japanese, Korean, and Indian exchanges.
The fund is managed by Jay Plourde, Portfolio Manager and Managing Member of Presvic LLC, based in New York.
Presvic LLC is a New York-based investment management firm. The FourthTurning Global Capital Fund is a long-only, unlevered global equity portfolio targeting physical bottlenecks in the AI infrastructure supply chain. The fund is open to qualified investors.
The FourthTurning Global Capital Fund is available to qualified investors. For strategy materials, investor documentation, or to schedule a conversation, please reach out directly.
jplourde@presvic.com