Quarterly letter
Q1 2026. The earnings-season quarter.
This is the first quarterly letter from Halvren. The format you are reading runs on a four-times-a-year cadence: a short note on the quarter just past, what we wrote about, what changed our minds, what we passed on, and what is on the desk for the next ninety days. Public letters live here. Position-level commentary, the standing book, and the trades in either direction go to the private letters on Substack. Both halves are written by the same person, with the same voice, against the same Canadian-operator universe.
The quarter, in one paragraph
The first quarter of 2026 was the earnings-season quarter for Canadian operators. Cameco, Canadian Natural, First Majestic, and Nutrien all reported their full-year 2025 numbers within a six-week window in February and March. The four names we cover most closely each told a slightly different story, and the dispersion between commodity-driven and operator-driven returns widened over the quarter. The standout in our reading was not the strongest tape or the loudest beat. It was Cameco closing the year with a negative net debt position. Strong commodity producers do not usually carry net cash. Cameco does, and that quietly earned it a place at the top of the desk for Q2.
What we wrote
Cameco (CCO) · Uranium
FY 2025 revenue of C$3.48B (+11%), adjusted net earnings doubled to C$627M, and 21 Mlbs of U₃O₈ produced against a backdrop of lifted contract realizations as the utilities' term book began to roll. The numbers we keep coming back to live below the headline: C$1.2B of cash against C$1.0B of total debt while the commodity is hot. Most Tier-1 producers spend a uranium uplift; Cameco saved it. Westinghouse contributed roughly C$170–200M of adjusted EBITDA on Halvren's read, and management's 2026 production guidance steps down at the midpoint, which is a discipline choice and not a capacity ceiling. We took up our prior on Westinghouse's economic participation in the AP1000 cycle. We did not take up our prior on the cycle itself; that is a decade-long question and we will not get it from a quarter.
Canadian Natural (CNQ) · Oil & Gas
Murray Edwards's playbook had its twenty-sixth consecutive year. Adjusted funds flow of C$15.5B funded roughly C$9.0B of shareholder returns: about C$4.9B in dividends, C$1.4B in buybacks, and C$2.7B of net debt reduction. The 2024 AOSP acquisition synergies showed up at the high end of guidance. The decline rate that defines this business held at the low end of the corporate range. The succession question has not gone away, but neither has Murray Edwards, and we watch the proxy circular and the named-officer purchases more closely than the press release.
First Majestic (AG) · Silver
The hardest of the four to own honestly. Revenue rose 124% on a 47% silver-price tailwind and a 50% increase in payable AgEq sold; AISC barely moved at US$21.17/AgEq oz. When the metal moves and the cost stays still, you find out who is in the right quartile. The 2025 print did not, in our reading, resolve the per-share-value question after the Gatos transaction. Honest read: the stock outran the fundamentals in Q1, the fundamentals have not yet earned the multiple, and the Mexico tax overhang is not a footnote. We left First Majestic in the “not yet” column. The Checklist works for us when it produces “not yet” as often as it produces a position.
Nutrien (NTR) · Fertilizers
Nutrien snapped back. Adjusted EBITDA of US$6.05B, with potash at record sales volumes and an attendant US$2.25B of segment EBITDA. The number that matters most for next year is not the FY 2025 print; it is the US$551M of share repurchases at meaningfully better levels than the 2022 vintage. Buybacks executed counter-cyclically are the cleanest signal of management treating the share count like an asset rather than a target. The Saskatchewan cost-curve position is real and not changing. The Retail segment quality remains the single most important multiple variable, and we got partial — not complete — clarity in the FY release. We expect to be more confident on Retail by mid-year.
What changed our minds
Coming into 2026 our prior on Westinghouse was skeptical-of-cyclicality. The 2025 EBITDA contribution and the international project pipeline (Poland, Bulgaria, Slovenia, U.S. small modular reactor licensing) moved that probability up materially. We are not buying a more constructive view of the AP1000 cycle than we held ninety days ago. Cycles take longer to play than thesis decks suggest. What we are buying is a more constructive view of Cameco's economic participation in whatever cycle plays. Those two are not the same thing, and we owe readers the distinction.
We also revised our prior on Canadian potash. The post-Belarus supply structure has not unwound the way the 2022 narrative suggested it would. Some of the displaced tonnes have come back, less of them than we modelled, and at less attractive economics for the marginal producers than we modelled. Saskatchewan's first-quartile position is more durable than we thought a year ago. That is good for Nutrien, broadly speaking, but the read does not fully translate into the multiple yet.
What we passed on
A junior Canadian gold producer with a single Quebec deposit, no positive earnings yet, and a thesis that required a higher gold price to underwrite. The presentation was excellent and the geology may genuinely be excellent. The Halvren Checklist took thirty minutes to disqualify it on Pillar I, question two: the unit economics did not work at the worst gold price of the last decade. Most names die there. That is not a bug of the screen. It is the screen.
We also looked carefully at a Canadian fintech with a credible insurance angle and an aggressive growth narrative, and concluded the per-share-value math was an opaque function of three executive estimates the company itself does not disclose. We will revisit when the disclosure improves.
The cycle, briefly
Mid-2026 is asking one small question and one large one. The small one: where does Canadian heavy oil clear if WTI sits in the high US$60s to low US$70s and AECO stays soft into a warm winter. The large one: how much of the post-2022 repricing of potash and uranium is structural and how much is narrative. We will not know the answer to the second question for years, and we are not in a hurry to. The discipline is underwriting to a Canadian-operator world that doesn't require either to resolve in our favour. The names we keep coming back to are the ones that work without the bull case. CNQ is the cleanest example. Cameco is becoming one. Nutrien is one if Retail clarifies. First Majestic is not, and might never be without a permanent reset to the silver tape.
On the desk for Q2
Three threads. First, pipelines and midstream. The infrastructure pillar of the coverage map has been on the desk since launch and is not yet on the page. We expect a full writeup in the next four to six weeks; the question we are working through is regulatory jurisdiction versus rate-base growth at the operator we are most interested in, and what the right normalized multiple looks like. Second, a name we declined two cycles ago. We are revisiting it to see whether the Checklist holds; it might not, and we will tell you either way. Third, the standing watchlist refresh: an updated by-the-numbers block on each of the four current names against Q1 2026 results when those land.
A note on the publication
Halvren launched the public site in December 2025 and the first batch of writeups in April 2026. Five thousand-word research pages, the Halvren Checklist, the working glossary, the founding memo, and this letter all live on halvrencapital.com and will stay free, indefinitely. The position-level commentary, the standing book, and the trades on either side of any name are reserved for the private letters on Substack. The split is not pricing strategy. It is the line between research that is useful to anyone and research that is useful to someone deciding what to do with their own capital. The latter requires a different relationship with the reader.
Sign-off
Thank you for reading. Halvren is a small operation by design, and the letter goes to the patient few. If you find an error in any of the public writeups, please write — corrections come in within twenty-four hours and earn a permanent footnote.
— Amirali Karimi
Halvren Capital · Vancouver, BC · April 2026
This letter is for informational and educational purposes only. It is not a recommendation, solicitation, or price call. The author may hold positions in any name discussed and may transact at any time without notice. See the Terms of Use for the full disclaimer.