Uranium Mining Stocks
Investment Analysis and Picks

By the time they restarted, spot had tripled. At the 2025 Investor Day the company said their long-term contract ceilings now reach $140 to $150. The ore they refused to mine at $22 per pound.

I have a friend who runs a small commodity fund and when I described the McArthur River shutdown to him his reaction was that no board of directors would approve this in any other commodity. And he's right. BHP would never mothball a Tier-1 iron ore mine to buy spot iron ore and deliver into contracts. The idea is absurd in iron ore. It's absurd in copper. In uranium, Cameco did it and it was probably the best capital allocation decision in mining in the past decade.

I'm going to talk about Cameco a lot in this piece. More than is probably balanced. But Cameco is genuinely the only uranium company I feel like I understand deeply enough to have conviction in, and I'd rather be honest about that than pretend I have equally strong views on eight different stocks.

U3O8 spot in 2025 went from $63 to $83 and back. Round trip to nowhere. The buy side of this market is Sprott's physical trust, a handful of trading desks, utilities plugging short-term gaps. That's it. A single trade of a few hundred thousand pounds can move the quote. Think about what that means for anyone trying to use the spot price as a real-time signal. It's not a signal. It's an echo.

I want to tell you something about uranium spot that took me an embarrassingly long time to figure out. The settlement cycle for a spot trade is not hours or days. It's weeks. Sometimes months. Export permits, transport permits, receiver qualification audits. So when you see "uranium spot price up 3% today" on a terminal screen, what you're actually seeing is the price of a transaction that may have been negotiated two or three weeks ago and is only now hitting the published indices. For a long time I tried to read short-term direction from spot price moves and trade around them. It didn't work and I couldn't figure out why. The answer was that the published price was stale by the time I was looking at it. I'm a little embarrassed it took me as long as it did to realize this.

Long-term contract prices through 2025 climbed from $80 to $86, above spot the entire year. By early 2026, spot pushed past $100 and overtook term. Backwardation. This has preceded every major uranium bull cycle I can find in the historical data. 2005. Briefly in early 2024. Again in early 2026.

Uranium equities were up about 40% on average in 2025 per Sprott's year-end numbers while the commodity price did nothing. A lot of people found this confusing. The stocks price off the forward contract curve, not spot. A producer with uncommitted production capacity gets bid up as the market's estimate of future contract prices rises.

This is the part of the uranium story that I find most fascinating and that I think most investment analyses get wrong or skip entirely.

Uranium fuel is something like 5% to 8% of a nuclear power plant's total operating cost. I've seen the exact breakdown in the NEA/IAEA "Projected Costs of Generating Electricity" reports but I can't remember the exact number. Let's call it 6% as a rough midpoint. At a natural gas plant, fuel is 60-70% of operating expenses. The gas procurement team lives and dies by the price they pay. At a nuclear plant, 6% of opex. Nobody's career has ever been made or ended by the timing of a uranium purchase. So the procurement people procrastinate. Why wouldn't they? If your boss doesn't care when you buy the uranium as long as the reactor has fuel, you'll put it off.

The secondary supply that used to paper over this gap is gone. The US-Russia HEU deal ended in 2013. WNA's published figures say it was feeding about 24 million pounds per year into the market. When it ended that was it. Sprott's Jacob White, in what I believe was the January 2026 market update, estimated utility working inventories hitting minimum comfort thresholds by late 2026 or early 2027.

In 2006-2007, a bunch of utility procurement managers who had all been deferring realized at roughly the same time that they actually needed to secure uranium. Spot went from the mid-forties to $136. I sometimes wonder what the internal conversations at those utilities sounded like in February 2007 when spot was at $75 and rising fast and the procurement manager had to explain to the CFO why they hadn't locked in supply six months earlier when it was $45. Must have been ugly.

The current setup has a similar shape. Maybe worse because in 2006 there was still secondary supply entering the market and today there isn't.

I keep coming back to this company.

The three-legged supply model: own production, inventories, spot market purchases. When uranium is cheap, reduce mine output, buy on spot, deliver into contracts. When uranium is expensive, crank up the mines. Toggle back and forth. Nobody else in uranium has the asset base and contract book to run this. Kazatomprom produces more pounds but operates under different constraints. The juniors don't have contracts.

Wall Street consensus: 2025 earnings up 96% year over year, 2026 another 55%.

If uranium goes sideways for two years, Westinghouse backstops. The stock probably treads water. If uranium runs, the contract book participates because the ceilings are high enough. The asymmetry here is what I like. The stock has already had a big run and some of the bull case is priced in. I still think it's the best name in the sector.

The part of uranium stock analysis I've spent the most time on and that I think matters most.

Floor and ceiling clauses in supply contracts determine how much of a uranium price rally a company actually participates in. If the ceiling is $100 and uranium goes to $150, you get $100 per pound delivered. Zero above that. Your upside is capped.

This information is in the filings. Annual report MD&A. Quarterly supplements. I have gone through Ur-Energy's 10-Q from Q2 2025 and on page 14 of the management discussion section they lay out that 23% of base commitments from 2026 through 2033 are market-price-linked, rising to about 30% if customers exercise full optional quantities. The rest has floors and ceilings. From this you can build something resembling a cash flow model at different uranium prices. The gap between URG's value at $80 uranium and at $120 uranium is smaller than you'd guess if you were just applying a generic "leverage to uranium" mental model. The market-linked 23-30% participates in the upside. The rest doesn't, or not fully.

I should do this exercise for every uranium producer. I've only done it thoroughly for Cameco and URG. It's time-consuming because the disclosures aren't standardized. Some companies give you a lot of detail. Others give you almost nothing.

This bugs me and I haven't been able to resolve it.

Cameco's contract ceilings: $140 to $150 per pound. Published estimates of the incentive price for new greenfield uranium mines, from consultants like UxC: $80 to $120 per pound. Spot in early 2026: above $100.

If the incentive price is $80 to $120 and spot is already above $100, why aren't more projects advancing? Some of the answer is permitting (greenfield mines in Canada take seven to fifteen years from discovery to production, and Canadian environmental assessment has gotten procedurally heavier, I estimate maybe two to three extra years versus a decade ago based on tracking NexGen and Denison's timelines against their original projections). Some is financing (try raising C$4 billion in project finance for a uranium mine with zero revenue, which is NexGen's capex estimate from the DFS). But I think part of the answer is that the published incentive price estimates are wrong. Or rather, they were calculated under cost and interest rate assumptions that no longer hold. The real incentive price in 2026, with current construction costs, current interest rates, current contingency requirements that banks impose, is probably above $120 and maybe above $140.

Sprott Physical Uranium Trust has been the largest marginal buyer in uranium spot for several years running. It now holds, I don't know the exact current figure but it's tens of millions of pounds, enough that forced selling would overwhelm a spot market where the daily bid side is maybe a few hundred thousand pounds.

This scares me. Not as an abstract risk factor. As a specific scenario I've thought through and can't find an answer for.

If SPUT faces heavy fund redemptions, it has to sell physical uranium into a market with essentially no depth. The spot price doesn't decline 10%. It gaps down 30%, 40%, who knows. And this has nothing to do with uranium supply and demand. Reactors are still running. Utilities still need fuel. The fundamental picture is unchanged. It's a financial structure event hitting a market with no cushion.

I don't know how to hedge this. You can't short physical uranium. Shorting uranium equities is expensive and the timing is unknowable. The only answer I've come up with is that my uranium position has to be sized for this scenario, meaning small enough that a 40-50% drawdown doesn't force a sale. Which constrains the position size considerably. Which is frustrating when you have conviction in the fundamental thesis.

The thing that makes this different from normal commodity downside risk is that in copper or oil, a 40% drop in price eventually triggers supply cuts, demand response, and the market self-corrects. In uranium, if SPUT dumps physical material, the supply cuts don't happen because the spot price crash is financial, not fundamental. Mines don't shut down because Sprott had redemptions. So the spot price could crash and stay crashed for months until the selling pressure exhausts itself, even though nothing in the physical market has changed. The people who get shaken out during that window lose money despite being right about everything.

I spend more time thinking about this than about any other risk in uranium. More than policy reversal (Germany, Vogtle cost overruns, Japan missing restart timelines, the usual list). More than uranium's gap-risk characteristics (no futures market, long flat periods then sudden moves, Cameco dropped 40% from April 2022 to early 2024 on essentially no fundamental change). The SPUT scenario is the one where I could be completely right about the thesis and still get destroyed if I'm not sized correctly.

The demand case. I'm going to blow through this quickly because I think it's well understood and not where analytical edge exists.

Reactor life extensions: most certain demand source, nearly every US reactor pursuing renewal to 60 or 80 years, refurbishment capital committed, essentially irreversible. New builds: WNA projects 69,000 tonnes (2025) going to above 150,000 (2040), China building a lot. SMRs: no commercially operating unit anywhere, NuScale's CFPP cancelled on cost overruns in 2023, I don't think this matters before the 2030s. Data centers: Microsoft and Meta signed 20-year PPAs with Constellation, long transmission chain to actual uranium purchase, real but overhyped.

NexGen (NXE). Arrow is unambiguously world-class geology. Grades above 20% U3O8 in parts. My issue is the DFS capex above C$4 billion. That's an enormous financing for a zero-revenue company. I've tried to model different structures. Equity at current prices, streaming, project debt, combinations. The dilution range is wide. During the 2022-2024 correction NexGen fell roughly two to three times harder than Cameco. The Arrow deposit is magnificent. Whether that magnificence accrues to current shareholders depends entirely on financing terms that haven't been set yet.

Kazatomprom (KAP). Biggest, cheapest. Production cuts are involuntary, acid supply constraints. Small allocation.

Ur-Energy (URG). Already discussed the contract structure above. Production cost at $50.89/lb in Q2 2025, down from $62.06 prior quarter. One of the few companies in the sector with a repeatable track record of actually producing uranium at a profit.

UEC, Denison, Paladin, Boss. I'm going to be straight with you. I don't know these well enough to say anything useful that isn't just me rephrasing their investor presentations. UEC is a US ISR portfolio built through acquisitions. Denison is trying ISR in the Athabasca Basin, which would be a big deal if it works. Paladin and Boss are restart plays in Namibia and Australia. I've read the overviews. I haven't read the feasibility studies. I haven't modeled the contract books. If I say more than this I'm making stuff up.

ETFs: URNM, URNJ.

Grade decline. I check this annually in NI 43-101 technical reports. Reserve grade tables. Ore grades fall as mining progresses, more rock per pound, higher unit costs. Companies don't talk about this. It's in the appendices.

Supply side stuff I've already folded into the utility procurement and Cameco sections above. Kazatomprom acid constraints. Greenfield timelines of 7 to 15 years. Canadian permitting getting procedurally heavier. I don't want to repeat myself.