Contrarian Codex — Cameco Q2 results (2026-JUL-31)

One-line take: Cameco's Q2 headline looks bad and, on Mart's read, isn't — the weakness is a Westinghouse year-on-year comparison artifact plus two deliberate choices. Volumes fell 18% (7.1m lbs vs 8.7m) because Cameco is choosing to sell fewer pounds into a rising market, and unit costs rose 26% mostly on heavier market purchases, product-loan revaluations and the Cigar Lake maintenance outage landing in Q2 rather than Q3. Guidance went up on price and revenue (realized C$91–96/lb from C$85–89; consolidated revenue C$3,320–3,570m) with production and delivery volumes unchanged — "a price and currency revision and not a volume one," and the higher cost guide is mostly the 1.33 → 1.35 USD assumption, not operational trouble. The structural point is the contract book: ceiling-capped older vintages hold the realized price below spot today, and Cameco's own sensitivity table shows a book held flat at $100 spot realizing ~$67/lb in 2026 but ~$88 by 2030 as those ceilings expire. On the call the term price is $97 and demand keeps showing up, new layers carry higher floors and higher ceilings with base pricing in the mid-to-high $90s, and COO Grant Isaac's sharpest line was that the market has never traded here while contracting ran this far below replacement rate — every previous cycle peak came at the back end of a contracting cycle; this one is at the front. On Westinghouse: the AP1000 pipeline is now 91 opportunities / ~105 GW, the DOE's conditional $17.5bn covers long-lead items on up to 10 reactors — and Westinghouse has confidentially filed a draft Form S-1 for a proposed IPO. Mart refuses to put a valuation on it and says plainly he would rather Westinghouse stay on Cameco's books, as one nuclear-power-and-fuel-cycle giant commanding a higher multiple.

1. Stocks & names mentioned

A single-name results piece. Contextual entities — the Inkai JV, Brookfield (as Westinghouse's co-owner and the likely driver of a monetization route), the DOE and Department of Commerce, and the V.C. Summer / Lubiatowo-Kopalino projects — appear in the talking points rather than as tabled picks.

2. Talking points

Ticker	Name	Research	View	What he said	At
CCJ	Cameco	QT · SA · STK · FA	Positive	"The quarter is a good deal stronger than it may come out if you just throw it into an AI agent and ask for a summary" — the ugly headline is a Westinghouse comparison artifact; volumes are down 18% by choice and costs up on FX (1.33→1.35) and purchase timing, while price and revenue guidance went up (realized C$91–96/lb) with volumes held. The ceiling-capped older contract vintages that cap today's realized price are exactly what unwinds in Cameco's favour: ~$67/lb realized in 2026 vs ~$88 by 2030 at a flat $100 spot.	read ↗
Westinghouse	Westinghouse Electric (private — Cameco 49% / Brookfield)	—	Positive	The forward picture "firmed": AP1000 pipeline up to 91 potential reactors / ~105 GW, the DOE's conditional $17.5bn for long-lead items on up to 10 reactors, and a Commerce partnership vesting a government participation interest at $80bn+ of US reactor investment before 2029. A confidential draft Form S-1 for a proposed IPO is now filed — no share count or range, so "there is no valuation to point at and I am not going to invent one" — but Mart's preference is explicit: "I would rather have Westinghouse stay on Cameco's book," as one nuclear-power-and-fuel-cycle giant commanding a higher multiple.	read ↗

Lower volumes were a decision, not a demand crack

Deliveries of 7.1m lbs against 8.7m a year ago, an 18% drop, tied by management to contracting discipline and lower planned 2026 deliveries — and customers can choose when in the year they take delivery, "so the quarterly pattern jumps around regardless."
Production of 3.9m lbs (Cameco's share) vs 4.6m, down 15%, is mostly timing: the annual Cigar Lake maintenance outage fell in Q2 this year rather than Q3.
Mart's filter: "The things that would actually worry me, real production trouble or a crack in demand, are not in this quarterly update."

The 26% jump in unit cost, decomposed

Uranium-segment average unit cost of sales rose 26% to CAD$70.81/lb, with three identified drivers: much heavier market purchases (2.8m lbs bought vs 0.7m a year ago), product loans revalued to the weighted-average cost of inventory each period, and the Cigar Lake outage hitting cost of sales in Q2 rather than Q3.
None of the three is an operating deterioration — they are purchase mix, an accounting revaluation and a calendar shift.

Inkai — production ahead, but the cash lands a year later

Inkai produced 2.8m lbs (100% basis) in the quarter and 5.3m for the half, both ahead of last year, still targeting 10.4m lbs for the full year. Cameco's purchase allocation is 4.2m lbs, of which only 0.8m came through in H1 — "the flow is weighted to the back end."
Because Cameco equity-accounts the stake, it books its share of Inkai production as a purchase at a 5% discount to spot; the real benefit shows up separately in equity earnings and only becomes cash when Inkai declares a dividend.
The lag in practice: a $124m dividend net of withholdings arrived in April tied to Inkai's 2025 performance; the benefit of 2026 output becomes cash in 2027.

Guidance up — on price and currency, not volume

Realized uranium price lifted to C$91.00–96.00/lb from C$85.00–89.00; uranium revenue to C$2,700–2,910m from C$2,540–2,730m; fuel services to C$610–650m; consolidated revenue to C$3,320–3,570m from C$3,130–3,370m.
Drivers: higher year-to-date uranium prices plus a stronger USD, with the full-year exchange assumption moved to 1.35 from 1.33.
The same dollar cuts the other way, so the unit cost-of-sales guide rose to C$63.00–67.50/lb (from C$61.50–65.00) and fuel services to C$33.50–35.80/kgU. Production guidance held and deliveries stay at 29–32m lbs — "a price and currency revision and not a volume one."
Westinghouse's 2026 adjusted-EBITDA share outlook is unchanged at $370–430m, with a weak first half and stronger fourth quarter — "exactly how the year was framed from the start, so the first-half loss is on script."

The contract book: ceilings today, upside in the back years

Two things hold the realized number below spot right now: Cameco choosing to sell fewer pounds, and ceiling prices on older contract vintages. Both are "features of the contract book that unwind in Cameco's favor as the older vintages roll off."
The company's own sensitivity table makes it concrete: a portfolio held flat at $100 spot realizes roughly $67/lb in 2026 but closer to $88 by 2030 as the ceiling-capped contracts expire — "the upside is loaded into the back years."
Balance sheet reads fine "once you pull the CRA remittances back out of the cash flow line."

The Q&A tell — demand shows up while the price climbs

Management "sounded readier to commit volume again," and interest keeps coming in with the term price at $97 this month. That combination — utility demand arriving while price rises — "is exactly what the higher-for-longer narrative needs."
New layers carry higher floors and higher ceilings with base pricing in the mid-to-high $90s; Mart expects "a good deal more to go in around the triple digit region as the book fills out," pulling Cameco's realized price up with it.

Grant Isaac's front-of-the-cycle argument

The sharpest point on the call, from COO Grant Isaac: uranium has never traded at these levels while contracting ran this far below replacement rate. "We are not even at replacement rate contracting yet and the price keeps climbing anyway."
His read is that the path from here is likely triple digits, and that every prior time the market reached a point like this it was already on the back end of a contracting cycle — "whereas this time it is on the front end, so there is far more still to come."
Supply reinforces it: brownfield ounces depleting, greenfield projects sliding sideways or backwards, and the restarts that were supposed to fill the gap running into trouble. Mart will pick this up with Isaac in person at the upcoming WNA symposium.

Westinghouse — a pipeline that hasn't started converting

91 identified AP1000 opportunities and up to 105 GW, but the read for the rest of the year is patience: 51 units still in origination, 11 in front-end engineering, 4 on early services contracts, and only a handful into long-lead ordering and construction (the potential V.C. Summer restart, Poland's Lubiatowo-Kopalino).
The line Mart is watching is the backlog trend: New Plants backlog was just $0.8bn at the end of last year against $13.2bn in Operating Plants — "plenty of room for it to build as opportunities convert into firm orders," and he expects it to pick up speed and magnitude quickly once it starts.
Near-term catalysts are the conditional DOE commitments turning into definitive financing documents and the DOE naming which utilities get the money, expected across H2 as each loan closes. On whether the aid pulls development forward, management said things are on schedule and they are being prudent with what they publish — "which to my eye leaves more room for a positive surprise than a negative one."

The IPO filing — direction of travel, no valuation

Westinghouse has confidentially submitted a draft Form S-1 to the SEC for a proposed IPO of its common stock; share count and price range are unset and it is subject to market conditions, "so there is no valuation to point at and I am not going to invent one."
It fits Brookfield's pattern of finding a route to monetize an asset it has built up, and it rhymes with the partnership terms where the government stake and the option to force a listing come alive if a Westinghouse IPO is valued at $30bn or more before 2029.
Mart's own preference is the opposite of a spin-out: keep Westinghouse on Cameco's books, because "one nuclear power and fuel cycle giant" should command a higher multiple for Cameco. What a listing would do to how the stake is carried is not knowable yet — "I am leaving it there and watching it."

AP300 — early, but real

A scaled-down AP1000 carrying over the same proven technology, passive safety and regulatory foundation, aimed at customers constrained by grid size, industrial demand, site limits or financing capacity.
About 80% of the supply chain is shared with the AP1000, "which is why they expect a robust supply chain behind it rather than having to build one from scratch." More than 30 units already in origination, first commercial operation targeted mid-to-late 2030s.
Standing caveat repeated throughout: none of the AP1000 or AP300 story is 2026 cash flow.

Housekeeping

The macro framework update ran in yesterday's newsletter and is the recommended companion read for putting prevailing price action into perspective.
A portfolio update is planned for next week — more macro data, and a decision on whether to raise cash or deploy some of it into key holdings.

3. In plain English

CCJ — Cameco Positive

Cameco mines uranium and sells it to utilities under long-term contracts, and it also owns 49% of Westinghouse, the company that designs and services nuclear reactors. This quarter's headline numbers looked weak, and Mart's argument is that almost none of the weakness is what it appears to be. Sales volumes fell 18% because Cameco chose to sell fewer pounds — it is holding back supply while prices rise rather than locking in today's price. Production fell because the annual maintenance shutdown at Cigar Lake happened in this quarter instead of the next one. Costs jumped 26% mostly because Cameco bought a lot more uranium on the open market this quarter (2.8m lbs vs 0.7m a year ago) and because a stronger US dollar makes everything it buys in dollars more expensive while its mining costs are in Canadian dollars. And a chunk of the year-on-year "decline" is simply Westinghouse being compared against a much better prior-year period.

The forward-looking numbers moved the other way: Cameco raised what it expects to be paid per pound (to C$91–96, from C$85–89) and raised its revenue guidance, while leaving production and delivery volumes untouched. In plain terms, it expects the same amount of uranium to fetch more money.

The heart of the thesis is the contract book. Many of Cameco's older contracts have a "ceiling" — a maximum price the utility will pay no matter how high the market goes — so even with uranium near record term prices, Cameco's realized price is held down. Those old contracts expire over the next few years and get replaced with newer ones written at much higher levels. Cameco publishes a table showing exactly this: if the market price simply sat at $100 a pound and never moved, Cameco would realize about $67/lb in 2026 but roughly $88/lb by 2030, purely because the cheap old contracts roll off. So today's disappointing realized price is not a problem with the business — it is a countdown timer running in the shareholder's favour.

Westinghouse — Westinghouse Electric Positive

Westinghouse is privately held — Cameco owns 49% and Brookfield the rest — and it is where the "new reactors get built" part of the story sits. Its flagship product is the AP1000, a large reactor design, and its pipeline is now framed as up to 91 potential reactors (~105 gigawatts). But a pipeline is not an order book: 51 of those are still at the earliest "origination" stage, 11 in engineering design, 4 on small early-services contracts, and only a handful have reached the stage where real long-lead equipment gets ordered. The number that would prove the pipeline is converting is the New Plants backlog, which was only $0.8bn at the end of last year against $13.2bn in the existing plant-servicing business — so there is enormous room for it to grow, and Mart's read for the rest of this year is simply patience.

What could accelerate it is government money: the US Department of Energy has made a conditional $17.5bn financing commitment covering the expensive long-lead components for up to 10 reactors, and a Department of Commerce arrangement gives the government a participation interest if it puts at least $80bn into US reactor projects before 2029. "Conditional" is the operative word — the thing to watch is those commitments turning into signed financing documents and the DOE naming which utilities receive the money, expected through the second half.

The news in this quarter is that Westinghouse has quietly filed a draft registration statement (a Form S-1) with the SEC for a possible stock-market listing of its own. It is confidential, with no share count or price range set, so nobody can put a number on it — and Mart deliberately refuses to invent one. His view is contrarian to the usual "unlock the value" reflex: he would rather Westinghouse stayed inside Cameco, because a single company that both mines the fuel and builds the reactors should be worth a higher multiple than the two pieces separately. He also notes the terms only get interesting at scale — the government stake and the right to force a listing kick in if a Westinghouse IPO is valued at $30bn or more before 2029.

Also on the roadmap is the AP300, a smaller version of the AP1000 for customers who cannot take a full-size reactor — same technology, same safety design, same regulatory approvals, and about 80% of the same supply chain, which is why it does not need a new industrial base built from scratch. More than 30 units are in early origination, with the first one running in the mid-to-late 2030s. As Mart repeats throughout: none of this is 2026 cash flow — it is why the stake is valued the way it is, not what it earns today.

Analysis distilled from the Contrarian Codex results update (PDF linked above). For personal study — not investment advice. Source material © Contrarian Codex / "Mart".