AI data center power scarcity is priced as time-to-delivery, not megawatts
GPU deployment cycles outrun grid interconnection, making calendar time the binding constraint
New GPU generations land in roughly eighteen months while U.S. generation interconnection to commercial operation averages about sixty-one months, leaving a multi-year gap capital cannot buy away. Hyperscalers can contract power and buy chips but cannot jump interconnection queues, turbine slots, or long-lead transformers, so AI sites increasingly pay for when electrons become usable.
| Instrument | Side | Target | Reason |
|---|---|---|---|
| GEV | Long | We believe vendors that shorten the path from fuel to usable grid capacity capture more of AI power scarcity than merchants selling energy under regulated price caps, because the bottleneck is queue position and deliverable equipment, not willingness to build another gigawatt in theory. | |
| CAT | Long | We believe on-site and fast-ramp generation and backup power become strategic substitutes when interconnection timelines stretch past GPU refresh cycles, lifting demand for engines and power systems that can ship on industrial rather than utility-calendar schedules. |
Losses along the plant-to-GPU voltage chain make efficiency and conversion infrastructure critical
Power falls from transmission voltages near seven hundred sixty-five kilovolts to sub-one-volt logic through repeated conversions, and a large share of input energy is spent on conversion, distribution, and cooling before it reaches compute. At gigawatt scale even small percentage losses equal hundreds of megawatts, so the investable problem is the leaky pipe, not only empty reservoirs.
| Instrument | Side | Target | Reason |
|---|---|---|---|
| VRT | Long | We believe suppliers of rack-to-facility power distribution and thermal management benefit disproportionately when every megawatt of AI load forces simultaneous upgrades to electrical rooms and cooling loops to fight compounding heat from conversion losses. | |
| ETN | Long | We believe switchgear, protection, and building-level distribution gear see sustained pricing power when megawatt racks force redesign of upstream electrical architecture rather than incremental copper additions at legacy voltages. |
Megawatt racks force a coordinated shift to eight-hundred-volt DC, favoring power-semiconductor and component ecosystems
Legacy forty-eight-volt rack distribution breaks at megawatt density because current and copper losses explode; eight-hundred-volt DC cuts amperage for the same power and was adopted because electric-vehicle supply chains already matured SiC connectors and busbars. Changing rack voltage ripples through building distribution, protection, and board-level regulators, so lead-time winners are component vendors, not a single voltage tweak.
| Instrument | Side | Target | Reason |
|---|---|---|---|
| MPWR | Long | We believe high-efficiency board and rack power IC vendors gain share as AI factories redesign the last centimeters of the voltage waterfall and pay for parts that reduce heat at every conversion stage. | |
| NVTS | Long | We believe GaN and SiC switch suppliers can re-rate ahead of near-term revenue when hyperscalers reserve option value on eight-hundred-volt designs, because the investable signal is architectural commitment and delivery slots, not last quarter’s shipment tally alone. |
Pure-play electricity merchants lag when scarcity is expressed through equipment lead times and chain position
Tight grids and rising AI load should lift merchant generators, yet names focused on selling electrons often re-rated less than equipment and infrastructure suppliers that accelerate deliverability, cut losses, or sit adjacent to the GPU. Markets are sorting AI power by function in the chain—time, leakage, and proximity to compute—not by a single thematic label.
| Instrument | Side | Target | Reason |
|---|---|---|---|
| BE | Long | We believe on-site solid-oxide and distributed power platforms capture urgency premia when grid timelines cannot match AI build schedules, even where regulated energy prices hit scarcity ceilings that cap merchant upside. |
Themes
The content on this page is for informational purposes only and does not constitute financial advice. Stoquate is not a licensed financial advisor. Always conduct your own research and consult a qualified professional before making any investment decisions.