← Analysis page  ·  Ben Finegold hub  ·  Research hub

Actionable insights — Investing in the nuclear reactor supply chain

The repeatable process behind the picks: not which nuclear names Ocean Wall likes, but how it maps a reactor into a supply-chain screen, removes design risk, finds the bottleneck parts, and picks jurisdictions and cap sizes on the fuel side. Written so it can be rerun on another reactor, another build-out, or the next conference.
2026-SEP-16 · Jimmy Connor (YouTube), WNA Symposium · Ben Finegold — Ocean Wall · ▶ Watch · full analysis · transcript
How to read this page: each insight is a method — a mapping exercise, a filter, a screen, a relative-value rule — with the boxed line showing how Finegold applied it in this 10-minute interview and a "watch for" list for re-running it. Timestamps deep-link into the video.

00:52 1. Name the phase of the theme before choosing the vehicle

The repeatable method
  1. Split a multi-year theme into legs: supply-led (the input commodity re-prices) and demand-led (the end-asset actually gets built).
  2. Judge which leg is next by the gate between them. Here that gate is policy and rhetoric turning into projects reaching final investment decision (FID).
  3. Rotate part of the exposure from the first-leg vehicles (fuel) into the second-leg ones (builders and suppliers) without abandoning the first.
Here: seven years of the nuclear-fuel trade "has been a great trade," but "the next leg… is going to be demand-led" (00:52). The last 6 years were rhetoric, policy and financing, and the next 5 are about getting projects to FID (01:49).
Watch for

01:19 2. Map the bill of materials, then score every supplier

The repeatable method
  1. Pick the reference asset most likely to be built at scale (here the AP1000) and size its pipeline in units.
  2. Break it into parts and services: EPC contractor, pressure vessel, steam turbine, compressors, software. List every company that can supply each.
  3. Score each supplier on a fixed rubric. Ocean Wall uses up to 25 points, weighting deliverability (can it actually deliver on time?) and supply-chain localization (is it inside the buyer's allied or domestic chain?).
  4. Rank to the few names whose score and exposure to the pipeline are both high.
Here: an AP1000 pipeline of 91 reactors (from Cameco's call) and 144 companies screened (02:17). The output was Westinghouse via CCJ, 034020.KS and BWXT.
Watch for

03:26 3. Filter out binary technology risk — own the picks and shovels

The repeatable method
  1. When several competing designs are racing (large reactors vs many SMR designs), don't bet on the winner.
  2. Ask of each supplier: does it get paid whichever design wins? Keep only those that do.
  3. Prefer suppliers with a sticky base business already in hand (government or defense contracts) so the new cycle is upside, not survival.
Here: "we don't want companies that are taking this binary tech risk" (03:26). BWXT: 70% US-government revenue, and "whoever wins the SMR race, BWXT is going to be a winner, too" (04:32).
Watch for

05:39 4. Screen for the long-lead-time "golden screws"

The repeatable method
  1. Tabulate the procurement lead time for each component. The parts with multi-year waits are the build-out's binding constraint.
  2. For each long-lead part, count the qualified suppliers inside the allied chain. One or two means pricing power.
  3. Check the price: already-discovered bottlenecks (sold-out order books) show up as rich multiples, so look for the constrained supplier the market hasn't priced that way yet.
Here: the reactor pressure vessel is the classic long-lead part, and 034020.KS is "the only allied" forger of large RPVs (04:06). The already-priced version: ENR.DE and GEV at 70–80x earnings with order books sold out five years (05:59).
Watch for

08:37 5. Split a global deficit by region — trade the short jurisdiction

The repeatable method
  1. Start from the headline global supply–demand gap for the commodity.
  2. Split it by bloc. Which side already has secured supply, and which is short?
  3. Find the allied jurisdiction that must fill the short side's gap. Its producers and explorers carry a strategic premium the global number hides.
Here: a ~50m lb/yr deficit that is "not geographically equal. The East is far better coverage than the West," so "Canada has to play an outsized role." His takeaway trade: "Canada, Canada, Canada" (08:37), specifically Athabasca Basin Uranium juniors.
Watch for

09:02 6. Majors vs juniors: rotate down the cap scale into consolidation

The repeatable method
  1. Hold the quality major for core exposure, but test its relative price against the juniors, not its absolute merit.
  2. When the major is fully priced and the juniors have "fallen under the radar," tilt new money to the juniors.
  3. Use the last cycle as a template. If it ended in a consolidation wave, expect the majors to buy the juniors again, which is the payoff route.
Here: "I'm a shareholder of Cameco, but relatively priced is not cheap" (09:02). He sees "relatively better value in the smaller names" and "a wave of consolidation like we saw in the previous cycle" (09:24).
Watch for

06:22 7. Benchmark the West's cost and speed gap against the fastest builder

The repeatable method
  1. Anchor on the fastest and cheapest builder's per-unit cost and build time. Here China builds reactors for under $6bn.
  2. List the drivers of the gap (regulation, NIMBYism, subsidies to FID, labor cost and skills, how many engineers are in government) and track which the West is actually closing.
  3. Treat a closing driver, such as licensing reform or government first-loss capital, as a catalyst for the Western build-out and its suppliers.
Here: EDF is seconding engineers to China (06:22). He points to "Breakneck" (engineering state vs lawyerly society) and US licensing already "much, much more efficient" (05:39). UK SMR development capital needs government to fund the first $300–500m (06:55).
Watch for

Methods distilled from the public YouTube video (Jimmy Connor, WNA Symposium, London). Not investment advice.