Methodology
Why these 45 tickers. Why these 7 themes.
The dashboard tracks the AI / robotics / quantum trade end to end: the chips, the software that runs on them, the power and cooling that gates them, the memory that constrains them, and the funds whose 13F filings tell you who's positioned for what. This page explains how the universe was constructed and what changes.
Tickers
7 themes · 10 funds
The Bluechip 45 Index
Equal-weighted · rebased to 1,000 · 90-day lookback
The chart on the home page labelled Bluechip 45 is a synthetic equal-weighted index of every ticker in this universe. Each constituent gets the same weight (1/45). The index is rebased to 1,000 at the start of the 90-day lookback so the movement is readable as a percent change from that baseline.
Equal-weighting is intentional. A market-cap-weighted index would effectively be NVIDIA and a few mega-caps; that’s already what the Nasdaq-100 measures. The Bluechip 45 is meant to surface breadth: when small-caps (quantum names, robotics pure-plays, SMR power plays) rally or sell off together, you see it here before you see it in the cap-weighted indices.
Read the index as a breadth gauge for the trade, not as a tradeable benchmark. When the line diverges from the Nasdaq-100, that’s the signal: either the cohort is leading the broader market or the broader market is leaving it behind.
Selection Criteria
1 · Theme alignment
Every ticker must fit one of the 7 themes cleanly. Diversified conglomerates that touch AI tangentially (e.g. Honeywell, Siemens) don’t make the cut. The litmus test: would removing this company materially change the picture of the trade in its theme?
2 · Narrative relevance
The universe favors names that show up in fund letters, earnings-call discussion, and policy debate. This is why speculative quantum pure-plays (IONQ, RGTI) are in even though their revenue is small: they move the narrative.
3 · Tradability
US-listed equities only. Minimum ~$50M market cap to ensure actual liquidity. ADRs included (TSM, ASML, ARM). No private comparables, no derivatives, no synthetic exposures.
The 7 Themes
AI Compute
Every dollar of model intelligence flows through silicon. AI compute is the gating constraint and the largest capex line in the trade. The companies designing, fabricating, and networking the chips that train and serve frontier models. The incumbent (NVIDIA) and the credible challengers (AMD, Broadcom custom silicon, Cerebras for inference) sit on top of a stack that runs back through TSMC and ASML. When the trade has been wrong, it's been wrong here first.
AI Software & Hyperscale
Hyperscalers and model labs are the demand side of the compute equation, and the distribution channel for AI capability. Microsoft, Alphabet, Amazon, and Meta each operate model labs, custom silicon programs, and AI-monetized product surfaces in parallel. Oracle is the late entrant betting the company on AI cloud. Salesforce and Palantir are the enterprise-AI ops layer. The trade here is leverage on the compute cycle, not exposure to it.
Robotics & Autonomy
Embodied AI is the next surface area. Tesla's Optimus + FSD are the most ambitious public bet on general-purpose humanoid robotics. Intuitive Surgical is the proof that vertical robotics can build a moat. Industrial automation (ABB, Symbotic) is the boring trade that compounds. The signal here is the lag. Software AI moved first; physical AI is moving now, and the capex implications are larger.
Quantum Computing
Quantum is pre-revenue, multi-modality, and binary. IBM is the only at-scale incumbent. IonQ (trapped-ion), Rigetti (superconducting), and D-Wave (annealing) are pure-plays on different physical architectures, any of which could be the winning bet, or none of them. The trade is asymmetric: small position, long horizon, large multiples on success.
Power Generation
Data center capex went vertical in 2024 and the grid couldn't keep up. Independent power producers (Vistra, Constellation) sold out their dispatchable generation to hyperscalers. NextEra is the largest renewables operator at the scale AI demands. GE Vernova builds the gas turbines that bridge the gap. SMR plays (NuScale, Oklo) are the speculative tail. Power is increasingly the rate-limiter on model training, not silicon.
Cooling & Data Center Infra
Once the chips and the power are in place, you still need to keep them cool, move data between them, and own the building. Vertiv and Modine ride the liquid cooling cycle. Arista and Cisco wire the racks together. Digital Realty and Equinix own the dirt. This is the picks-and-shovels of the picks-and-shovels. Less narrative, more compounding revenue.
Memory & Semi Equipment
High-bandwidth memory (HBM) became the constraint AI chips couldn't grow past. Micron rode that wave from a commodity trade to a structural one. Western Digital and Seagate ride AI's storage demand on the cold side. KLA and Lam Research sell the equipment that makes all of it possible. When AI compute is constrained, it's almost always constrained here.
The 10 Tracked Funds
13F holdings · refreshed daily from SEC EDGAR
Each fund’s most recent 13F-HR filing is pulled from SEC EDGAR and intersected with our 45 tickers. The mix spans index giants (BlackRock, Vanguard), legacy value (Berkshire), tech-concentrated hedge funds (Tiger, Coatue, Whale Rock, Light Street), narrative leaders (Altimeter, Lone Pine, TCI, Greenoaks, D1), long-duration tech (Baillie Gifford), macro (Soros, Bridgewater), and quants (Renaissance, Citadel, Two Sigma).
13F filings lag ~45 days from quarter end. The current snapshot is Q1 2026; next refresh after the Aug 14 filing deadline.
How the Universe Changes
Editorial · not algorithmic
The universe is curated. New tickers get added when they pass all three selection criteria above and the existing cohort doesn’t already represent the same thesis. New themes get added only if the trade fundamentally requires them (cooling was added when liquid-cooling became a gating constraint; SMRs were added when hyperscalers started funding nuclear restarts).
Suggestions are open. The best filter on a thesis is people pitching the names that prove or disprove it. If you think the universe is missing a clean expression of the trade, send it to @BluechipsAI with a one-line thesis.