Denison Mines (DNN): what it does, how it makes money, and why Phoenix matters

The standard $DNN story is a cheap uranium stock: a sub-$3 share that moves when the uranium price moves. That is half-right. Denison does trade with the uranium pack, but it is not a producer selling pounds from a running mine. It is a developer in the middle of building one, and almost all of its value sits on a single project that will not produce before mid-2028.

The more accurate frame: Denison is a construction story with a uranium treasury attached. The board took the final investment decision on Phoenix, an in-situ recovery mine on its 95%-owned Wheeler River project in Saskatchewan, on 2026-02-24, at an initial capital cost of C$600M. This piece walks through what Denison does, how it pays for the build before the mine earns a dollar, where it sits in the nuclear stack, and the risks underneath. Figures are as of 2026-10-02 unless noted.

QA's video teardown of Denison is embedded on the live page, /stocks/dnn, next to the price, the peer correlations and the curated thesis.

Why it matters now

DNN closed at $2.57 on 2026-10-02, about 41% under its $4.37 closing high, down 24.6% over one month and 18.4% over three months. That drop came with no company-specific news, while the build itself kept moving: on 2026-07-28 Denison announced the end of site preparation and the start of full-scale construction. A project on schedule and a falling share price can coexist when the market is repricing the commodity, not the company. That is the distinction worth drawing.

What does Denison Mines do?

Denison is a Canadian uranium developer. Its core asset is Wheeler River, 95% owned, in the Athabasca Basin of northern Saskatchewan, the same basin that hosts the world's highest-grade uranium deposits. Phoenix is the first deposit at Wheeler River to be developed.

The method is what makes Phoenix unusual. Instead of digging the ore out, in-situ recovery (ISR) pumps a mining solution through the ore body underground, dissolves the uranium, and brings the uranium-bearing solution back to the surface for processing. To contain that solution in the sandstone, Denison is building a freeze wall around the deposit: by end of July 2026, Phase 1 freeze wall installation had started and over 20% of site civil work was done, per the Q2 2026 report. ISR is common in Kazakhstan and the US, but it has never run at commercial scale in Athabasca sandstone. Phoenix would be the first.

The 2023 feasibility study plans about 56.7M lb of U3O8 from ore grading near 19%, among the highest grades anywhere. Two other pieces sit beside the project:

• 22.5% of the McClean Lake mill joint venture, operated by Orano. McClean Lake is one of the few licensed uranium mills in the basin.
• A physical uranium inventory: about 1.1M lb held at 2026-06-30 (950,000 lb physical plus 145,926 lb of McClean concentrates), which Denison sells to help fund construction.

How Denison makes money

Today, it mostly does not, in the operating sense. Reported revenue over the trailing twelve months was about $2.9M (QA data, 2026-10-02), which puts the price-to-sales ratio near 568: a number that says "pre-production", not "expensive". There is no mine revenue before Phoenix starts, guided for mid-2028.

What funds the company until then is a treasury built on three legs:

1. Cash. C$465.3M of cash and cash equivalents at 2026-06-30, nearly flat versus C$465.9M at 2025-12-31.
2. Convertible notes. US$345M of 4.25% convertible senior notes due 2031, closed 2025-08-15, with proceeds earmarked for Phoenix. They convert at 342.9355 shares per US$1,000, about US$2.92 per share.
3. Uranium sales. In Q2 2026 Denison sold 750,000 lb at US$89.17/lb, or US$66.9M, booking a realized gain of C$64.1M. Another 600,000 lb are committed from Q3 2026 to Q2 2027, 350,000 lb of them at a fixed US$95.17/lb average (US$33.3M). About 500,000 lb remain uncommitted.

Those uranium sales are booked as gains, not as operating revenue, which is why the revenue line looks so small next to the cash coming in.

The future business is the mine itself. The company's January 2026 base case puts Phoenix at an after-tax NPV8 of C$1.57B, an IRR of 73% and a payback of about 12 months at US$69 to 78/lb uranium; at US$86/lb the company estimates an NPV of C$1.94B and an IRR of 82% (capex update, 2026-01-02). These are company estimates, not QA's. On the customer side, QA's data lists Cameco (toll-milling counterparty at McClean Lake), Orano (JV operator and processing partner) and an undisclosed US utility under a 2023 long-term offtake.

Where it sits in the nuclear bubble

Denison sits in QA's Nuclear / SMR bubble as a primary member (weight 0.75) and maps onto the Nuclear theme. Its role is the fuel-supply leg: the ore at the very start of the chain.

Place it by position in that chain. Miners and developers like Denison, Cameco ($CCJ) and NexGen ($NXE) produce or will produce uranium. Enrichers like Centrus ($LEU) turn it into reactor fuel. Reactor developers like Oklo ($OKLO) and NuScale ($SMR) build the machines that will burn it.

The correlation table confirms where the market files it. DNN's tightest correlations in the QA universe (as of 2026-10-02) run to NexGen (0.89), Cameco (0.83), Uranium Energy ($UEC, 0.83) and Energy Fuels ($UUUU, 0.81), all uranium names. The reactor developers come in lower: Oklo 0.68, NuScale 0.67, Centrus 0.67, Nano Nuclear ($NNE) 0.66. In other words, $DNN trades as a uranium pack member first and a nuclear-revival name second. NexGen is the closest comparable: another Athabasca developer with a single flagship project, also pre-production.

The numbers

QA's backend has no consensus rating or Street target on file for DNN as of 2026-10-02, so this table carries none. The useful comparison is the treasury against the bill: C$465.3M of cash plus the uncommitted uranium book against a C$600M initial capex, part of which has already been spent since March.

The bull case

• Funded and underway. Construction licence and FID in February 2026, construction since March, full-scale construction announced 2026-07-28, first production guided for mid-2028.
• Grade. Ore near 19% U3O8 is among the highest anywhere, which drives the company's estimated payback of about 12 months.
• Low-footprint mining. ISR means no open pit and no conventional underground mine, which the company says lowers capital and operating cost against a conventional build.
• A treasury that hedges itself. Selling physical pounds into a higher price funds the build; 350,000 lb are already locked at US$95.17/lb on average.
• Jurisdiction. Saskatchewan, with benefit agreements signed with Métis Nation-Saskatchewan and with three First Nations and four municipalities.

The bear case

• The bill grew. C$600M against C$419.4M in the 2023 study, or C$500.5M in 2026 dollars: about 20% higher even after inflation, with a C$65M contingency (about 12.5%). A first-of-its-kind mine can slip again.
• Unproven method in this geology. ISR has never run at commercial scale in Athabasca sandstone. Freeze wall performance and recovery rates are the open questions.
• One project, no mine revenue until mid-2028. The value rests on Phoenix and on the uranium price, and the uranium treasury shrinks as it funds the build.
• Dilution overhang. The notes convert at about US$2.92 per share, roughly 118M new shares on US$345M, about 13% of today's share count. The stock traded above that price earlier this year.
• The tape. At $2.57 the stock sits under the trailing 50-day ($3.08) and 200-day ($3.39) averages cited in QA's thesis on 2026-10-01, and moves with the uranium pack regardless of project news.

How to access

Denison trades on NYSE American as DNN and in Toronto as DML. To trade it from a US-retail account alongside the rest of the uranium and nuclear names, see /stack/ibkr.

QA's data has no ETF holdings on file for DNN as of 2026-10-02. Third-party holdings data lists it among the positions of uranium miner funds such as the Sprott Uranium Miners ETF ($URNM) and the Global X Uranium ETF ($URA); weights move, so check each fund's own holdings file before relying on a number.

The full teardown video is on /stocks/dnn. Subscribing to the digest on that page sends the Denison Teardown Dossier free; the paid version is at /shop/dnn-dossier and is included in /pro, along with bubble-correlation shifts and rule-based alerts on $DNN.

What to watch

• Next earnings: Q3 2026 results, not yet dated by the company. Last year's Q3 came out on 2025-11-07. The read is construction progress against the C$600M budget and the contingency drawn so far.
• Freeze wall and wellfield milestones. Phase 1 freeze wall completion is the step that tests the method in this geology.
• The uranium book. The pace of uncommitted pound sales and the price realized, since that is how the build is paid for between now and 2028.
• The conversion price. Where the stock sits against the US$2.92 conversion price decides whether the notes stay debt or become shares.
• Observable reference levels. QA's auto-computed retracements of the trailing 52-week $2.20 to $4.43 range sit at $2.68 (0.786) and $3.05 (0.618), both above the 2026-10-02 close; the $2.20 low sits about 14% below it. These are observable reference levels, not trade levels or targets.
• Bubble-level shift. If DNN stops tracking NexGen and Cameco and starts trading on its own construction milestones, the structural read changes.

Live data on this ticker: /stocks/dnn. Price, ETF holdings, bubble correlation, curated levels, bot positions.

Bubble context: /bubbles/nuclear-smr. The cluster this name belongs to and how it's moving.

QuantAbundance is educational research. Nothing here is investment advice. See /disclosures.