# AwarenessFund — AI-Supply-Chain Ticker Universe (Spec Sheet)

**Author**: capital-markets-lead (ACG / A-C-Gee)
**Date**: 2026-05-18
**Status**: DRAFT TEMPLATE — REQUIRES EXPERT REVIEW BEFORE PRODUCTION USE
**Audience**: AwarenessFund partners (Corey / Russell Korus / Apex-Jordana-Pyonair) + investor insiders + sister AiCIVs

---

## How to read this document

- 4 verticals × 8–15 tickers each = candidate **picks-and-shovels** universe
- "Pure picks-and-shovels" means we are explicit: **NO model-labs** (no Anthropic-public-proxy, no OpenAI-IPO speculation, no LLM application-layer). The fund buys the **supply chain**, not the applications.
- Every ticker has a sub-segment tag + 1-sentence "why included" + data-tier required
- This is a **candidate universe** — the back-test protocol (see `backtest-protocol-spec.md`) will rank/filter, not this document
- Data-tier column says what's needed to *fully* analyze the ticker; *all* tickers are price-trackable on free yfinance + EDGAR
- **Symbols only** — exchange suffix included for non-US listings since back-test must normalize across listings

---

## Vertical 1 — MINING (raw materials of the AI economy)

**Thesis-link**: AI hardware and grid build-out is materials-intensive. Copper (interconnect + grid), nickel (batteries + stainless), uranium (clean baseload for data centers), rare earths (motors + magnets), silver (PV + electronics), lithium (storage), gallium/germanium (chips). Mining capex cycles are 8–12 years from discovery to first pour — supply elasticity is near zero on the timescales AI demand needs.

### Candidate tickers — Mining

| # | Ticker | Name | Sub-segment | Why included (1 sentence) | Data tier |
|---|--------|------|-------------|---------------------------|-----------|
| 1 | FCX | Freeport-McMoRan | Copper majors (US) | Largest US-listed copper pure-play; copper is the single material with the tightest AI-grid bottleneck | Free (EDGAR + FMP) |
| 2 | SCCO | Southern Copper | Copper majors (LatAm) | Lowest-cost copper producer globally; long-life reserves in Peru/Mexico | Free |
| 3 | TECK | Teck Resources | Copper + zinc diversified | Pure-play copper after coal divestiture; Quebrada Blanca 2 ramp | Free |
| 4 | RIO | Rio Tinto | Diversified majors (iron + copper + lithium) | Lithium (Rincon, Jadar) + copper + iron-ore picks-and-shovels triple play | Free |
| 5 | BHP | BHP Group | Diversified majors (copper-tilt) | Largest miner by market cap; copper ~30% of EBITDA and rising | Free |
| 6 | MP | MP Materials | Rare earths (US-listed) | Only US-domestic rare-earth mine-to-magnet; DoD-aligned, dual-use AI motors | Free + EDGAR |
| 7 | LYC.AX | Lynas Rare Earths | Rare earths (ex-China) | Largest ex-China rare-earth producer; magnet supply-chain backbone | Yahoo only (AX listing) |
| 8 | CCJ | Cameco | Uranium (Tier-1) | Largest publicly traded uranium producer; baseload-for-AI-DC narrative | Free |
| 9 | URA | Global X Uranium ETF | Uranium basket (ETF) | Ticker-universe diversifier; captures juniors that fail standalone screens | Free |
| 10 | NEM | Newmont | Gold + copper | Largest gold producer; gold sub-thesis = real-asset hedge inside the fund | Free |
| 11 | ALB | Albemarle | Lithium specialty | Largest US-listed lithium producer; battery-grade chemicals | Free |
| 12 | NTR | Nutrien | Potash + nitrogen (specialty) | Specialty-materials proxy; food-as-energy-input adjacency | Free |
| 13 | SCCO.LM | Sociedad Quimica y Minera (SQM) | Lithium (Chile) | Alternative lithium exposure to ALB; brine-based, lower cost | Free |
| 14 | URNM | Sprott Uranium Miners ETF | Uranium juniors basket | Junior-uranium exposure without single-stock blow-up risk | Free |
| 15 | PICK | iShares MSCI Global Metals & Mining ETF | Diversified miners basket | Beta benchmark for the entire vertical | Free |

**Vertical 1 data-tier summary**: All 15 fully analyzable on FREE tier (EDGAR for SEC-filers, yfinance for OHLC, FRED for commodity prices via PCOPPUSDM / PURANUSDM proxies). One LatAm listing (SQM) requires yfinance .SN or ADR (SQM ADR is US-listed).

---

## Vertical 2 — ENERGY (powering the AI economy)

**Thesis-link**: US grid needs **+1 terawatt** of new generation in next 3–5 years for AI data-center demand (per industry estimates; cf. EPRI / DOE projections — see `thesis-substantiation.md` for receipts). This bottlenecks AI inference at the wall plug. Picks-and-shovels = (a) **utility owners** of generation/transmission near hyperscaler sites, (b) **independent power producers (IPPs)** with nuclear/natural-gas, (c) **grid-equipment makers** (transformers, switchgear, HVDC), (d) **uranium fuel cycle** (handled in Mining vertical above — overlap is intentional).

### Candidate tickers — Energy

| # | Ticker | Name | Sub-segment | Why included (1 sentence) | Data tier |
|---|--------|------|-------------|---------------------------|-----------|
| 1 | VST | Vistra Corp | IPP — nuclear + gas | Largest beneficiary of hyperscaler PPAs (Constellation-pattern); diversified nuclear + gas baseload | Free |
| 2 | CEG | Constellation Energy | IPP — nuclear pure-play | Largest US nuclear fleet; Three Mile Island restart catalyst (Microsoft PPA) | Free |
| 3 | NRG | NRG Energy | IPP — gas + retail | Texas/Northeast gas-baseload IPP; retail-energy hedge | Free |
| 4 | TLN | Talen Energy | IPP — nuclear (Susquehanna) | AWS-aligned hyperscaler-sale revenue; "AI-data-center" optionality | Free |
| 5 | NEE | NextEra Energy | Utility + renewables (largest) | Largest US utility + largest renewables developer; FL + national IPP arm | Free |
| 6 | DUK | Duke Energy | Utility (SE US) | Southeast US data-center corridor (Atlanta/Virginia adjacency) | Free |
| 7 | SO | Southern Co | Utility (SE US) | Vogtle 3/4 nuclear in service; Georgia DC corridor | Free |
| 8 | D | Dominion Energy | Utility (Virginia) | Virginia = 70% of US data-center inventory; load-growth pure play | Free |
| 9 | GE / GEV | GE Vernova | Grid + gas-turbine OEM | Spin from GE; gas-turbine + grid-transmission + wind triple play | Free |
| 10 | ETN | Eaton | Electrical equipment | Switchgear + data-center electrical infrastructure leader | Free |
| 11 | ABBN.SW | ABB | Electrification + automation | Largest ex-US electrical-equipment maker; HVDC + DC-grid | Yahoo + ADR (ABB) |
| 12 | SIEGY | Siemens AG ADR | Industrial automation + grid | Grid-equipment + industrial digitalization; AI-factory tooling | Free (ADR) |
| 13 | HUBB | Hubbell | Electrical components | T&D + utility-grade connectors; grid build-out beneficiary | Free |
| 14 | LNG | Cheniere Energy | LNG export (gas-feedstock) | Adjacency: AI compute → US gas demand → LNG export-economics | Free |
| 15 | URA | Global X Uranium ETF | Uranium fuel cycle (overlap) | Uranium-fuel-cycle beneficiary for nuclear-IPP narrative | Free |

**Vertical 2 data-tier summary**: All 15 analyzable on FREE tier. Heavy SEC-filer coverage (US utilities + IPPs file detailed 10-K capex disclosures via EDGAR). FRED carries `EIA-natural-gas-price` + `electricity-price-by-region` for context.

---

## Vertical 3 — CHIP-FABS + anything that supplies them

**Thesis-link**: The picks-and-shovels of AI compute are NOT the model labs and NOT even the chip designers — they are the **equipment that makes the chips** (lithography, deposition, etch, metrology) and the **specialty materials** (photoresists, gases, wafers, advanced packaging substrates). This is a 4-company oligopoly at the high end (ASML, AMAT, LRCX, KLAC) plus a Japanese duopoly in materials. Capex cycles ~24 months; AI demand is restocking faster than supply chain can absorb.

### Candidate tickers — Chip-fab + supply chain

| # | Ticker | Name | Sub-segment | Why included (1 sentence) | Data tier |
|---|--------|------|-------------|---------------------------|-----------|
| 1 | ASML | ASML Holding | Lithography — monopoly | Only EUV-lithography supplier globally; literal AI-bottleneck monopoly | Free |
| 2 | AMAT | Applied Materials | Deposition + etch | Largest WFE (wafer-fab-equipment) maker; broadest tool portfolio | Free |
| 3 | LRCX | Lam Research | Etch + deposition | #2 in etch; memory + advanced-logic exposure | Free |
| 4 | KLAC | KLA Corp | Metrology + inspection | Process-control monopoly; advanced-node yield dependency | Free |
| 5 | TER | Teradyne | Test equipment | Semiconductor + robotics test; AI-chip QA dependency | Free |
| 6 | TSM | Taiwan Semi (ADR) | Foundry — pure-play | Largest pure-play foundry; advanced-node AI-chip manufacturer | Free (ADR) |
| 7 | INTC | Intel | IDM (US-fabs) | US fab build-out (Ohio, Arizona); CHIPS-Act primary beneficiary | Free |
| 8 | 6857.T | Advantest | Test equipment (Japan) | AI-chip test monopoly (HBM + GPU final-test) | Yahoo (.T) |
| 9 | TOELY | Tokyo Electron (ADR) | WFE (Japan) | #4 WFE globally; etch + deposition + cleaning | Free (ADR) |
| 10 | 4063.T | Shin-Etsu Chemical | Specialty materials (wafers, photoresists) | Silicon-wafer + photoresist duopoly leader | Yahoo (.T) |
| 11 | 4183.T | Mitsui Chemicals | Specialty materials (packaging) | Advanced-packaging substrate adjacency | Yahoo (.T) |
| 12 | ENTG | Entegris | Specialty materials + fluids | Wafer-handling + advanced-process fluids monopolist segments | Free |
| 13 | AEHR | Aehr Test Systems | Test (small-cap, EV/AI burn-in) | High-growth small-cap; SiC + AI-chip burn-in (high-beta exposure) | Free |
| 14 | ONTO | Onto Innovation | Metrology (small-cap) | KLAC challenger; advanced-packaging metrology | Free |
| 15 | SOXX | iShares Semiconductor ETF | Sector basket | Beta benchmark; captures designers we explicitly exclude (NVDA/AMD pulled out via overlay) | Free |

**Vertical 3 data-tier summary**: 4 Japanese tickers (6857.T, 4063.T, 4183.T, plus ADRs TOELY) require yfinance Japan-market or US-ADR access. ASML and TSM are EDGAR 20-F filers (annual only, not quarterly). Free tier covers all 15 for OHLC; fundamentals subset is FMP free (250 calls/day budget — tight).

**Explicit exclusion**: NVDA, AMD, AVGO, MRVL — these are **chip designers**, not picks-and-shovels. They are **downstream buyers of the supply chain**, not the **supply chain itself**. SOXX captures them as benchmark overlap. If partners want chip-designer exposure, that is a **separate sleeve** decision, not picks-and-shovels.

---

## Vertical 4 — AI-ENDPOINT (data-centers for training + inference; picks-and-shovels ONLY)

**Thesis-link**: AI-endpoint = the **physical buildings** + **cooling** + **networking** + **power distribution** where compute happens. **Explicitly excludes** model labs (Anthropic-proxy, OpenAI-IPO speculation, application-layer AI). Includes: data-center REITs, hyperscaler-CapEx proxies (only where capex is dominantly endpoint-spend, not chip-design margin), networking-fabric makers, cooling specialists.

### Candidate tickers — AI-endpoint

| # | Ticker | Name | Sub-segment | Why included (1 sentence) | Data tier |
|---|--------|------|-------------|---------------------------|-----------|
| 1 | EQIX | Equinix | DC-REIT (retail colocation) | Largest interconnection-dense colo; AI-inference-edge proxy | Free |
| 2 | DLR | Digital Realty | DC-REIT (wholesale + retail) | Wholesale hyperscale + retail; largest DC-REIT by power | Free |
| 3 | IRM | Iron Mountain | DC-REIT (growing AI-DC arm) | Data-center pivot from records-storage; emerging AI-DC pure-play | Free |
| 4 | VRT | Vertiv | DC infrastructure (power + cooling) | Pure-play DC infrastructure OEM; liquid cooling + UPS leader | Free |
| 5 | SMCI | Super Micro Computer | DC servers (AI-optimized) | AI-server assembly leader; explicitly hardware-not-model | Free (caveat: 2024 accounting drama → red flag, see thesis-substantiation) |
| 6 | ARI / ARM | Arm Holdings | CPU IP licensing | DC-CPU + AI-accelerator IP royalty model; arms-dealer-by-definition | Free (recent IPO — short history) |
| 7 | ANET | Arista Networks | DC networking (Ethernet fabric) | Hyperscaler Ethernet-switching leader; alternative to NVDA NVLink | Free |
| 8 | CSCO | Cisco | Networking (enterprise + DC) | DC-Ethernet + Silicon-One ASIC; mature picks-and-shovels | Free |
| 9 | CIEN | Ciena | Optical networking (DC-interconnect) | DC-to-DC long-haul optical; AI-cluster-interconnect | Free |
| 10 | COHR | Coherent Corp | Optical components (transceivers) | 800G/1.6T optical transceivers for AI-DC; lasers + photonics | Free |
| 11 | LITE | Lumentum | Optical components | Coherent rival in DC-optical; 100G/400G transceivers | Free |
| 12 | JCI | Johnson Controls | Building HVAC + DC cooling | Building-systems + commercial HVAC; DC-cooling adjacency | Free |
| 13 | NVT | nVent Electric | Electrical enclosures + cooling | DC-rack-cooling + electrical-enclosure pure-play | Free |
| 14 | SYM | Symbotic | Warehouse robotics (proxy for physical-AI) | Optional bridge to Phase-2 physical-AI thesis | Free |
| 15 | DTCR | Global X Data Center & Digital Infrastructure ETF | DC-REIT basket | Basket benchmark for the vertical | Free |

**Vertical 4 data-tier summary**: All 15 US-listed, EDGAR-filed, yfinance-tracked, FMP-free-tier-fundamentaled. SMCI carries an explicit RED FLAG note (2024 Hindenburg short report + DOJ probe + auditor resignation; see thesis-substantiation §AI-endpoint risks).

**Explicit exclusion across this vertical**: MSFT, GOOGL, AMZN, META — hyperscalers are mixed-margin businesses where DC-capex is one line item among many (cloud margin, ads, retail). Owning them is **NOT** picks-and-shovels. If partners want hyperscaler exposure, that is a separate sleeve.

---

## Vertical 5 — SPACE-INFRA (data centers in orbit + the infrastructure to put them there)

**Thesis-link**: Terrestrial AI-endpoints are running into hard physical ceilings — grid power, water for cooling, fiber latency budgets, real-estate adjacency to demand. Orbital and cislunar compute is on the near-term roadmap (Lonestar Holdings + Ramon Space have public DoD/NASA contracts) and Dyson-swarm-style space-based solar is the long-arc energy-collection paradigm (Caltech SSPP / NRL SSPIDR / JAXA OHISAMA / ESA SOLARIS feasibility programs). Picks-and-shovels here is: (a) **commercial launch + reusability** (SpaceX is private — only proxies are tradable), (b) **satellite manufacturers + comms-backhaul + direct-to-cell**, (c) **defense primes that hold the DoD/SDA/NASA orbital-compute and space-systems contracts** (mixed-exposure caveat — these are mostly defense betas with space optionality), (d) **rad-hardened space-grade compute + ground-station + ISL networks**, (e) **space-solar / Dyson-swarm frontier** (20+ year optionality only — no near-term deployable claim). See `thesis-substantiation.md` §5 for sourcing.

### Candidate tickers — Space-infra

| # | Ticker | Name | Sub-segment | Why included (1 sentence) | Data tier |
|---|--------|------|-------------|---------------------------|-----------|
| 1 | RKLB | Rocket Lab USA | Launch + reusability (small/medium lift) | Largest publicly-traded launch operator; Electron + Neutron reusable-medium-lift in development | Free |
| 2 | IRDM | Iridium Communications | Satellite comms (LEO constellation) | Operational global LEO constellation; DoD-aligned backhaul + IoT revenue | Free |
| 3 | VSAT | Viasat | Satellite comms + ground stations | Geo-broadband + tactical milsatcom + ground-station footprint (post-Inmarsat) | Free |
| 4 | SATS | EchoStar | Satellite comms + spectrum | Hughes Network + S-band spectrum holdings; strategic-asset thesis | Free |
| 5 | ASTS | AST SpaceMobile | Direct-to-cell satellite | Only public pure-play in direct-to-handset cellular-from-space; AT&T + Verizon + Vodafone partnerships | Free (recent IPO — short history) |
| 6 | LMT | Lockheed Martin | Defense prime — space systems | Holds SDA Tranche / NASA Orion / GPS-III contracts (mixed defense beta with space optionality — NOT a pure-play) | Free |
| 7 | NOC | Northrop Grumman | Defense prime — space systems | NGI (Next Gen Interceptor), HALO lunar gateway, SDA tranches (mixed beta caveat) | Free |
| 8 | RTX | RTX Corp (Raytheon) | Defense prime — space + radar | Space radar + GPS + RL10 engines via subsidiary; mixed defense beta | Free |
| 9 | BA | Boeing | Defense + commercial space | Space Launch System + Starliner + ULA joint-venture stake; commercial-space exposure but heavy non-space drag | Free |
| 10 | MRCY | Mercury Systems | Rad-hardened compute + space electronics | Closest public pure-play in space-grade processors + secure-compute modules | Free |
| 11 | KTOS | Kratos Defense | Ground stations + small launch + space-comms | OpenSpace virtualized ground systems + Zeus rocket + space-tech components | Free |
| 12 | UFO | Procure Space ETF | Diversified space basket | Sector benchmark — ~30 holdings spanning launch/sat/defense-primes | Free |
| 13 | ARKX | ARK Space Exploration ETF | Diversified space basket (alt) | Alternative basket — Cathie Wood-tilted, more concentrated in disruptors | Free |

**Vertical 5 data-tier summary**: All 13 US-listed, EDGAR-filed, yfinance-tracked, FMP-free-tier-fundamentaled. Defense primes (LMT/NOC/RTX/BA) carry **explicit mixed-exposure flag**: their space-segment revenue is typically 10-30% of total, so basket-level space exposure is diluted — partners should view these as "space optionality" rather than pure-play. ASTS has limited price history (2021 SPAC merger) — back-test window may need haircut for ASTS coverage.

**Explicit exclusion / "honest gaps" for this vertical**:

| Excluded exposure | Why excluded | What it would unlock |
|-------------------|--------------|----------------------|
| **SpaceX (private)** | Not publicly tradable; no direct equity exposure exists for retail or fund vehicles | If SpaceX IPOs, becomes the cornerstone of this vertical. Until then: ZERO direct SpaceX exposure in this basket. |
| **TSLA as "Musk-proxy"** | TSLA is a vehicle-and-energy-storage company that happens to share a CEO with SpaceX. Owning TSLA for SpaceX exposure imports brand-discount, EV-cycle beta, and FSD-narrative volatility that have nothing to do with picks-and-shovels space-infra. | If partners want Musk-beta, that is a separate sleeve decision. |
| **MAXR (Maxar Technologies)** | Taken private by Advent International in May 2023 ($6.4B all-cash deal) — no longer tradable on US exchanges. | If Maxar successor lists publicly again, re-evaluate. |
| **ASTR (Astra Space)** | Delisted from Nasdaq in July 2024 after taking the company private; no longer tradable. | N/A unless successor-vehicle lists. |
| **Lonestar Holdings, Ramon Space** | Private orbital-compute / data-center-in-space operators with DoD-SDA / NASA contracts publicly announced 2024. No public equity exposure. | If either IPOs, would be highest-conviction pure-play orbital-compute name. Today: track contract-announcements for thesis-evidence only. |
| **Aitech, BAE Space (BAE-subsidiary), Cobham (private)** | Private or non-tradable rad-hardened-compute / space-electronics specialists. | MRCY is the closest tradable approximation. |
| **CSPP / Caelus Space Power and other space-solar concept-stage operators** | All concept-stage / feasibility-study tier; no operating revenue, mostly government-funded programs (Caltech SSPP, NRL SSPIDR, JAXA OHISAMA, ESA SOLARIS). Not investable today. | If any space-solar program reaches an operational demonstrator with private capital backing, re-evaluate. |

**On the Dyson-swarm horizon**: Frame this exclusively as **20+ year optionality**. The Caltech SSPP June 2023 in-orbit beaming demonstration was watts-class, not kilowatts. ESA SOLARIS is a feasibility-study program, not a build program. Treat the Dyson-swarm narrative as a *direction* the universe is moving (collect energy where it's abundant), not a near-term investable thesis. If partners want exposure, the closest tradable proxies are the defense primes (NOC, LMT) and the launch-cost-curve compressor (RKLB) — both already in the basket on other grounds. **Do NOT base any back-test weighting on space-solar deployment timelines.**

---

## Cross-vertical universe summary

| Vertical | # tickers | All on free tier? | Foreign listings (need ADR or yfinance) | ETF beta-benchmark |
|----------|-----------|-------------------|------------------------------------------|---------------------|
| Mining | 15 | Yes | LYC.AX (Lynas Australia), SQM-Chile (US-ADR exists) | PICK |
| Energy | 15 | Yes | ABBN (ADR=ABB), SIEMENS (ADR=SIEGY) | XLU (not listed here — see sp-comparative-spec) |
| Chip-fab | 15 | Yes | Advantest 6857.T, Shin-Etsu 4063.T, Mitsui 4183.T, TOELY-ADR | SOXX |
| AI-endpoint | 15 | Yes | (none) | DTCR |
| Space-infra | 13 | Yes | (none — all US-listed) | UFO + ARKX (dual baskets) |
| **TOTAL** | **73 candidate** | **Yes** | **5 foreign-listed (manageable)** | **5 baskets (6 incl. dual UFO/ARKX)** |

---

## Data-tier escalation path (if partners approve premium MCPs)

| If we get... | What unlocks |
|--------------|--------------|
| **FMP Premium** ($150/mo) | 70K-ticker universe (expand from 60 to ~200 candidates); full transcript history; institutional consensus estimates |
| **Polygon paid** ($29–79/mo) | Real-time + minute-bar history; options data for portfolio hedging; 10+ years OHLC |
| **Financial Datasets MCP** (~$40/mo) | Structured 10-K/10-Q deep-dive; segment-level revenue history; capex disaggregation |
| **Alpha Vantage** ($49.99/mo) | Technicals + earnings surprises for momentum-overlay strategies |

**Recommendation for back-test (see backtest-protocol-spec.md)**: Run v1 on **free tier**. If v1 produces material outperformance signal vs S&P over 24 months, partners get budget-approval substrate for upgrading to FMP Premium + Polygon paid (~$229/mo total) for v2.

---

## Methodology notes

- **No price targets, no ratings**: this universe is a **candidate set** for back-test ranking, not a "buy list"
- **Survivorship-bias caveat**: 60-ticker universe is curated from today's market — true back-test should include **delisted tickers** for the 24-month window (research-lead's ingestion-variations spec will address; SMCI volatility, any acquired names like Atlas Air subsidiaries, etc.)
- **Vintage**: all ticker selection as of 2026-05-18. Refresh quarterly.
- **Sources**: ticker selection cross-checked against (a) public industry-classification taxonomies (GICS sub-industries), (b) AwarenessFund partner conversation themes (terawatt-grid, mining-bottlenecks, chip-fab-cycles, AI-DC capex-bifurcation), (c) absence of model-lab / application-layer names per Corey's explicit "picks-and-shovels only" framing.

---

**Status**: SPEC SHEET v1 — ready for partner review + alignment with research-lead's ingestion-variations spec when it lands.

**Next**: back-test protocol spec (`backtest-protocol-spec.md`) consumes this universe.