/demos/ai-stocks-playbook AI Stocks Playbook A visual map of the AI buildout cycle: from silicon and memory to power, critical materials, robotics, defense, and the space economy.
42 companies
3 cycle phases
12 investment themes
Thesis map, not financial advice. Scores are qualitative research ratings using moat, theme fit, maturity, balance of risk, and execution evidence; they are not buy or sell recommendations. equity autoresearch model
How the scores are built Each company is rated with a reusable equity research loop: moat 25%, theme fit 25%, maturity 20%, execution 15%, valuation risk 10%, and business risk 5%. The label then converts the score into Core, Strong, Watchlist, or Speculative.
2026-2027 Compute Scarcity AI demand accelerates and capital crowds into the bottlenecks that make intelligence run: chips, memory, racks, power, and data capacity.
AI accelerators and full-stack compute platform
GPU leadership, CUDA software gravity, networking, and system-level AI infrastructure make it the anchor name in model training and inference buildouts.
96 Core Risk: Medium
Best-in-class AI infrastructure exposure; quality is exceptional, but valuation and cycle expectations leave little room for disappointment.
AI GPUs, CPUs, and adaptive compute
A credible second-source compute platform for hyperscalers that want performance, pricing leverage, and supply diversification.
82 Strong Risk: Medium
Good asymmetric challenger if AI GPU share gains continue, with CPUs providing a base; execution versus NVIDIA remains the key test.
Networking silicon and custom AI chips
Custom ASICs, switching, and connectivity sit in the expensive middle of AI data centers where scale matters more each cycle.
91 Core Risk: Medium
High-quality pick-and-shovel name across custom AI silicon, networking, and software cash flow; less pure upside than NVIDIA, more business durability.
Data infrastructure silicon
Optical, networking, storage, and custom silicon exposure gives it leverage to the data movement bottlenecks inside AI clusters.
77 Strong Risk: High
Compelling AI data movement exposure, but more dependent on custom silicon ramps and networking cycle timing.
DRAM, NAND, and high-bandwidth memory
AI servers pull memory intensity higher, and high-bandwidth memory can improve mix and pricing when demand outruns supply.
80 Strong Risk: High
HBM demand improves the story, but memory remains cyclical; attractive when supply discipline and AI mix are both improving.
Flash storage
AI data growth increases the need for fast storage tiers, archives, and edge data capture where NAND economics matter.
60 Watchlist Risk: High
Useful AI data-storage angle, but NAND economics and corporate transition risk make it less clean than the compute leaders.
Hard drives and storage infrastructure
Exploding model, video, sensor, and enterprise data keeps high-capacity storage relevant even as compute gets the headlines.
68 Watchlist Risk: High
Data growth is real, yet storage pricing cycles can dominate fundamentals; better as a cyclical infrastructure trade than a core AI holding.
Data center power and thermal systems
AI racks are denser and hotter, making power delivery, cooling, and uptime infrastructure critical spend items.
88 Core Risk: Medium
One of the clearest AI power-and-cooling beneficiaries; strong demand visibility, with valuation and capacity execution as main risks.
AI servers and rack-scale systems
Fast design cycles and GPU server exposure can capture demand when customers need capacity deployed quickly.
58 Watchlist Risk: Very high
Huge AI server demand exposure, but governance, margin, and customer concentration risks make position sizing matter more than the narrative.
AI cloud infrastructure
Specialized AI cloud providers can benefit when GPU access, sovereign demand, and workload-specific infrastructure remain scarce.
64 Speculative Risk: Very high
Potentially powerful AI cloud scarcity play, but still proving scale, margins, and long-term customer stickiness.
Power-backed data centers
Energy access and data-center sites are becoming strategic assets as AI compute competes for grid capacity.
62 Speculative Risk: Very high
Power access is valuable, but the business mixes data-center optionality with volatile crypto-adjacent economics.
2028-2030 Power Becomes Strategy Electricity, grid capacity, copper, uranium, lithium, and strategic minerals move from commodity footnotes to board-level constraints.
Data center power and thermal systems
AI racks are denser and hotter, making power delivery, cooling, and uptime infrastructure critical spend items.
88 Core Risk: Medium
One of the clearest AI power-and-cooling beneficiaries; strong demand visibility, with valuation and capacity execution as main risks.
Electrical equipment and grid modernization
AI data centers, electrification, and reshoring all require switchgear, transformers, breakers, and power management.
86 Core Risk: Medium
High-quality electrification compounder with exposure to data centers, utilities, and industrial power upgrades.
Grid engineering and infrastructure services
The grid upgrade cycle needs skilled execution, not just equipment, and Quanta sits close to that bottleneck.
84 Strong Risk: Medium
Execution-heavy grid upgrade beneficiary; attractive because bottlenecks require labor, permitting, and field capability, not just equipment.
Utility and electrical products
Utility hardening, grid reliability, and industrial electrification create durable demand for its components.
80 Strong Risk: Medium
Steady utility and electrical-products exposure makes it a durable, less flashy way to own grid modernization.
Lithium materials
Lithium remains a core input for electrification, batteries, and energy storage even through commodity downcycles.
57 Watchlist Risk: High
Strategic lithium exposure, but commodity pricing and capital discipline dominate the near-term case.
Lithium and specialty chemicals
Low-cost lithium resource exposure gives it leverage to battery supply chains and long-duration electrification demand.
61 Watchlist Risk: High
Low-cost lithium gives long-term relevance, while Chile policy, pricing, and commodity cyclicality keep risk elevated.
Domestic solar and energy supply chain
Domestic clean-energy manufacturing is strategically important if grids and AI infrastructure need more resilient supply chains.
46 Speculative Risk: Very high
Domestic energy supply-chain angle is interesting, but it needs proof of scale, economics, and market durability.
Power generation and grid technology
Gas, grid, wind, and electrification equipment make it a broad pick-and-shovel for rising electricity demand.
83 Strong Risk: Medium
Broad power-generation and grid exposure fits the AI electricity shortage theme better than most single-technology bets.
Copper mining
Copper is central to transmission, data centers, EVs, motors, and power equipment, making supply scarcity a powerful theme.
76 Strong Risk: High
Clean copper leverage to grids and electrification; quality asset base, but commodity prices will drive volatility.
Copper and critical minerals
Copper growth and diversified resource exposure can benefit from grid expansion and materials security.
72 Strong Risk: High
Good critical-minerals exposure with copper upside, though less direct and more commodity-sensitive than electrical equipment names.
Large-scale copper producer
Scale, reserves, and copper purity make it a direct expression of long-cycle electrical infrastructure demand.
74 Strong Risk: High
Direct copper exposure with scale and reserves; attractive in a copper bull case but highly sensitive to price and political risk.
Rare earth materials and magnets
Magnets are strategic for motors, robotics, defense, and electrification, and domestic supply is a national priority.
69 Speculative Risk: Very high
Strategic domestic rare-earth asset with real policy relevance, but processing, pricing, and execution risk remain high.
Critical minerals development
Early-stage critical materials exposure can compound if Western supply chains keep moving away from single-country dependence.
38 Speculative Risk: Very high
Very early critical-minerals option; upside exists, but evidence quality and financing risk make it venture-like.
Rare earth mining and magnet supply chain
Domestic rare earth and magnet capacity is strategically aligned with defense, EV, robotics, and grid priorities.
44 Speculative Risk: Very high
Domestic rare-earth magnet thesis is strong, but investability depends on execution, funding, and credible production milestones.
Rare earth and critical minerals exploration
A speculative way to express domestic critical mineral optionality if projects advance and policy support persists.
30 Speculative Risk: Very high
Tiny critical-minerals optionality; potentially explosive if projects advance, but weakest evidence quality in the basket.
Uranium and rare earth processing
Nuclear power and secure mineral processing both become more valuable as AI pushes electricity demand higher.
63 Speculative Risk: Very high
Useful uranium and rare-earth processing angle, but earnings power depends on commodity pricing and project execution.
Small modular reactor technology
If SMRs become commercially viable, modular nuclear could match the always-on power profile that AI campuses need.
52 Speculative Risk: Very high
Important nuclear technology option, but commercialization timing, customer adoption, and regulatory execution are not yet de-risked.
Advanced nuclear power
Advanced reactors are a high-upside bet on clean, firm power for data centers, industry, and remote infrastructure.
55 Speculative Risk: Very high
Strong narrative around firm power for data centers, but still a long-duration regulatory and commercialization bet.
2030+ Physical AI Goes Global The application layer shifts from chat and software into robots, autonomous mobility, defense systems, and space-based infrastructure.
EVs, autonomy, batteries, and humanoid robotics
Tesla combines real-world data, manufacturing, batteries, autonomy, and robotics into one option on physical AI.
72 Strong Risk: High
Unique physical-AI optionality across autonomy, batteries, and robotics, offset by valuation, execution, and governance risk.
Warehouse robotics and automation
Labor scarcity and logistics complexity make automated warehouses one of the clearest near-term robotics markets.
67 Watchlist Risk: High
Warehouse automation has a clear ROI story, but customer concentration and deployment timing can make results uneven.
Enterprise automation software
As AI agents enter operations, automation platforms can become control layers for repetitive digital work.
55 Watchlist Risk: High
Enterprise automation is relevant to AI agents, but the company must prove it can convert AI into durable growth.
EVs, autonomy, batteries, and humanoid robotics
Tesla combines real-world data, manufacturing, batteries, autonomy, and robotics into one option on physical AI.
72 Strong Risk: High
Unique physical-AI optionality across autonomy, batteries, and robotics, offset by valuation, execution, and governance risk.
Electric vertical takeoff aircraft
Autonomous and electric mobility could expand beyond cars if certification, battery density, and unit economics improve.
42 Speculative Risk: Very high
Interesting mobility option with large upside if certification and unit economics work; still pre-scale and binary-risk heavy.
Electric air taxi platform
A leading eVTOL contender with aircraft, operations, and certification progress tied to the next mobility layer.
48 Speculative Risk: Very high
One of the better eVTOL execution stories, but commercial adoption and certification remain major open questions.
Defense systems and aerospace
Defense budgets increasingly prioritize missiles, space, autonomy, and secure systems where incumbents retain deep moats.
78 Strong Risk: Medium
Durable defense cash flows and advanced systems exposure; slower growth but stronger evidence quality than most speculative AI-adjacent names.
Defense, space, and strategic systems
Space, command systems, and advanced defense programs align with a more autonomous and contested security environment.
77 Strong Risk: Medium
Space, strategic systems, and defense modernization exposure make it a resilient compounder, though not a high-beta AI play.
Drones, defense electronics, and hypersonics support
Smaller autonomous systems and lower-cost defense platforms fit the shift from exquisite assets to scalable fleets.
65 Speculative Risk: High
Good exposure to lower-cost autonomous defense systems, with upside tied to program wins and margin execution.
Uncrewed aircraft and loitering systems
Battlefield demand has validated drones as consumable infrastructure, and AVAV is a pure-play leader.
73 Strong Risk: High
Pure-play drone and loitering-munitions exposure with validated demand, but defense procurement timing can be lumpy.
Launch and space systems
Vertical integration across launch, spacecraft, and components creates leverage to a broader space infrastructure buildout.
66 Speculative Risk: High
Best-in-class small space-infrastructure contender, but launch cadence, Neutron execution, and profitability remain key hurdles.
Satellite-to-phone broadband
Direct-to-device satellite connectivity could unlock a massive market if deployment, spectrum, and carrier economics work.
54 Speculative Risk: Very high
Massive direct-to-device upside if deployment works, but capital needs, launch execution, and technical risk are substantial.
Lunar services and space infrastructure
Government and commercial lunar infrastructure spending can create a new services layer beyond traditional launch.
47 Speculative Risk: Very high
Lunar services optionality is real but still government-program-driven, episodic, and difficult to value.
Earth observation data
Persistent satellite imagery becomes more useful as AI improves extraction, monitoring, and operational decision-making.
58 Watchlist Risk: High
Earth-observation data should benefit from AI analytics, but growth, margins, and customer conversion need improvement.
Real-time geospatial intelligence
Demand for fast satellite intelligence can grow with defense, supply chain, disaster, and security monitoring use cases.
49 Speculative Risk: Very high
Real-time geospatial intelligence is strategically relevant, but scale and profitability are not yet proven.
