← Analysis page  ·  Contrarian Codex hub  ·  Research hub

Actionable insights — WNA report 2026

The repeatable analysis behind the report: how to read contracting volume against price, how to take apart an inventory "cushion," how to check a supplier's reported capacity against what it delivers, how to put a size on a contract with no published volume, and how to rebuild a position you trimmed.
2026-SEP-14 · Contrarian Codex · annual WNA conference report · read ↗ PDF · full analysis
How to read this page: each insight is a method Mart actually ran in this report, with the boxed line showing how it played out. (Written source, so no video timestamps.)

1. Read thin contracting volume together with the price, not by itself

The repeatable method
  1. Take the benchmark "healthy" volume year people cite and remove any one-off deals from it before using it as a baseline.
  2. Re-state the replacement bar in today's terms (current annual reactor requirements), not the old figure.
  3. Plot volume and term price over the same years. Volume below replacement while price rises = buyers are deferring into a market that is tightening anyway. That is a stored-up demand wave, not weak demand.
Here: 2023's ~160m lb included the ~40m lb Energoatom deal; 2024 ~110m and 2025 ~116m sat well under a ~200m lb requirement while term climbed to ~$97, so he expects an "outsized" price effect when volumes return.
Watch for

2. Break an inventory "cushion" down by where it came from and when it is needed

The repeatable method
  1. Ask where the build came from: panic double-buying and cheap contract options (upflex) are pounds headed for a reactor anyway, not surplus.
  2. Split the forward window into years. Check coverage for each period instead of treating a decade as one block.
  3. Ask whether the holders would actually sell. Strategic and sovereign stock does not come back to market.
Here: 229m lb of US+EU utility inventory covers ~2027, but 52% of the next decade of US demand (186m lb through 2035) is uncovered, and buyers asked would not part with their stock.
Watch for

3. Compare licensed capacity with the pounds actually delivered

The repeatable method
  1. When a demand event (a government RFI, a large tender) asks for supply, list the companies that could plausibly fill it.
  2. For each, compare the capacity they talk about with the pounds they recently delivered, and check whether the ore feeding that capacity exists and is permitted.
  3. Favor the operator already producing at scale; give no credit for presentation capacity.
Here: for the NNSA's ~4m lb/yr ask, UEC has licensed sites but delivered <70,000 lb in a quarter; UUUU made ~2.1m lb and owns White Mesa, so he backs Energy Fuels and adds to it.
Watch for

4. Put a rough size on a confidential contract by dividing its value by a reference price

The repeatable method
  1. Take the published headline contract value and divide by a disclosed reference price (e.g. a producer's own realized/reference price).
  2. Give a range for the price path, and say which way the result is biased (a higher contract price or a longer term shrinks the annual volume).
  3. Compare the total with the buyer's own published requirement to see whether it is covering its needs or building a reserve.
Here: Kazatomprom's >$4bn India contract at Cameco's $86.95 implies ~46m lb (call it 40–50m); with Cameco's ~22m lb, India bought 62–72m lb against a stated 23.4m lb 9-year need, 2–3x. He labels the estimate his own and "large but soft."
Watch for

5. Test the biggest assumed supply source against a delay scenario

The repeatable method
  1. Identify the single project the market's supply model depends on most.
  2. Run a realistic slip (later start, lower first-phase output) and compare the lost pounds with the forecast deficit.
  3. Ask in conversations whether buyers have a plan for that case; if they do not, treat it as upside price risk.
Here: NXE Arrow at ~16m lb by 2033–34 instead of ~30m lb by 2030–31 would be "a massive shockwave" that many buyers are "woefully unprepared for."
Watch for

6. Trim into spikes, then rebuild core names with cash when sentiment turns bad

The repeatable method
  1. Trim a slice of core holdings into sentiment spikes.
  2. When sentiment falls back toward depression while the fundamental data (term, spot floor, demand) keeps improving, put a set share of cash (a third) back into the same cornerstones.
  3. Keep any leveraged bolt-on (options) small and stated as speculative.
Here: DNN sold 20% at ~$4.10 in January, adds back 10% at $2.85; UUUU trimmed in the mid-$20s in Q1, restored to full, plus Jan-27 $15 calls at $1.30.
Watch for

Methods distilled from the Contrarian Codex WNA report (subscriber PDF). Not investment advice.