−13.0%
since Oct 7
The idea
I think humanoid robots move from demos to a real industry within five years, with Tesla leading the way. Even if Tesla builds most of the robot itself, it still needs magnets, chips, sensors, and connectors from others, plus power if millions of these things ever run at once. So alongside Tesla, I own the suppliers it can't easily replace, the rare-earth magnet makers, and utilities positioned for that demand, with a small slice in riskier bets on batteries and perception tech. I keep some cash on hand to use when the inevitable volatility creates better entry points.
What it owns
- The take centers on Tesla's Optimus program becoming a real product line within five years, requiring continuous credible milestones and sustained funding. · Tesla, Inc.The core bet on Optimus becoming a real product line rather than a sideshow, contingent on credible milestones and sustained in-house development.25%
- Even vertically integrated robot makers must source specialized components like sensors, semiconductors, and connectors from established suppliers. · Allegro MicroSystemsProduces magnetic sensor chips that tell robot joints exactly where they are, needed in bulk for systems with thirty-plus actuators., Taiwan Semiconductor ManufacturingAlmost certainly manufactures the chips inside any robot's brain, whether designed by Tesla or rivals, offering exposure to robot buildout with minimal company-specific risk., Amphenol CorporationSupplies the connectors, cables, and sensors that pack dense into every robot, benefiting from robot and automation demand regardless of which maker wins.25%
- Early-stage or niche technology bets that could capture meaningful robot-economy revenue if development timelines and customer adoption align. · Navitas SemiconductorGallium nitride power chips could win sockets in robot chargers and power systems where efficiency and compact size matter most., Ouster, Inc.Produces digital lidar sensors that map three-dimensional space, betting that robots moving through real-world environments need this technology even if Tesla does not., QuantumScape CorporationDeveloping solid-state batteries that are lighter and denser than conventional cells, with the bet that they power a working humanoid through a full shift within five years.15%
- Humanoid robots need dozens of small electric motors powered by rare-earth magnets, creating demand for domestic US production capacity. · MP Materials Corp.Makes rare-earth magnets that dozens of robot motors require, positioned as the dominant US producer if robot demand pulls meaningfully on magnet supply., USA Rare Earth, Inc.Building a domestic rare-earth mine-to-magnet supply chain for robot makers who might need alternatives if China tightens exports, though still in early stages.14%
- Millions of always-on robots plus AI training behind them will drive sustained growth in electricity demand, favoring clean baseload power providers. · Constellation EnergyOwns nuclear generation best positioned to supply the clean, round-the-clock electricity that millions of always-on robots and AI training will demand., Vistra Corp.A generator with nuclear and gas capacity that profits when electricity demand outpaces supply, holding firm if robot and AI buildout keeps power prices elevated.11%
- Cash10%