Investment Thesis
Picks & Shovels in the AI Revolution
We do not invest in AI companies directly. We invest in the infrastructure that AI needs to exist — the semiconductors, power plants, data centers, grid hardware, and critical minerals that every AI model, every GPU cluster, and every hyperscaler depends on. Government funding de-risks the demand side. Physics constrains the supply side. That asymmetry is our edge.
Government-Backed Assets
Companies directly benefiting from CHIPS Act, IRA, and DOE funding. Sovereign-grade demand with multi-year visibility.
Key Drivers
Key Tickers
AI Infrastructure
Semiconductors, data centers, networking, and cooling — the physical layer AI cannot exist without.
Key Drivers
Key Tickers
Power & Energy
Nuclear, natural gas, and grid infrastructure powering AI data centers. Each 1 GW data center requires dedicated baseload.
Key Drivers
Key Tickers
Critical Minerals
Rare earths, lithium, uranium, and copper — constrained supply meeting exponential AI-driven demand.
Key Drivers
Key Tickers
Sector Conviction Map
Government-Backed Assets
Highest
AI Infrastructure
High
Power & Energy
High
Critical Minerals
Medium-High
Supply Chain Convergence
The thesis is one interconnected supply chain. Government capital triggers infrastructure buildout, which drives power demand, which accelerates mineral consumption. We are positioned across the entire chain.
Government Capital
CHIPS Act + IRA + DOE
$280B+ authorized
Infrastructure Buildout
Fabs, Data Centers, Networking
$1T+ hyperscaler capex 2024-2028
Power Demand Surge
Nuclear, Gas, Grid Upgrades
+47 GW by 2030 (est.)
Mineral Demand
Uranium, Copper, Rare Earths
Supply deficits widening
Demand Certainty
Government appropriations create multi-year demand visibility that private markets cannot. CHIPS Act milestones, IRA tax credits, and DOE loans are contractual commitments.
Supply Constraints
New fabs take 3-5 years. New mines take 7-15 years. Nuclear plants take 5-10 years. Physics and permitting create moats the market consistently underestimates.
Asymmetric Payoff
If AI scales, infrastructure is the bottleneck and our positions benefit. If AI disappoints, government funding still flows and energy demand still grows.
Position Alignment (0/0 mapped)
Thesis Validation Status
CHIPS Act Disbursement
confirming$39.6B allocated
Intel, TSMC, Samsung, Micron — preliminary terms signed
US Data Center Pipeline
confirming43.9 GW planned
Hyperscaler capex acceleration — MSFT, META, AMZN, GOOG
Uranium Spot Price
confirming$81/lb U3O8
Up 130% from 2022 lows — supply deficit structural
Nuclear Restarts
confirming3 announced
TMI Unit 1, Palisades, Hatch — more expected
AI Capex Guidance (Big 5)
confirming$260B+ FY2025
MSFT, META, AMZN, GOOG, AAPL — all guiding higher
Copper Supply Gap
watching~6Mt deficit by 2030
Electrification + AI demand vs. underinvestment in mines
Grid Interconnection Queue
watching2,600 GW backlog
FERC Order 2023 reforms — bottleneck easing slowly
GPU Supply/Demand
confirmingDemand >> Supply
Blackwell ramp — lead times extended through 2026
Warsh Doctrine — Live Signals
LOADINGLoading FRED data...
Loading execution strategy...
Actionable Strategy — Live Recommendations
AWAITING DATAMargin Reduction
Upload portfolio & balance data to generate margin plan
Tax Loss Harvesting
Upload portfolio data to identify harvesting opportunities
Options Strategy
Upload portfolio data to generate options recommendations
Portfolio Allocation
Upload portfolio data to review allocation
Thesis alignment scores and validation metrics are maintained by the Shanfeld Equities AI research process. Portfolio weights update in real-time from connected brokerage data.