Uranium: Substack vs the street

'The next gold' priced 3 ways. The disagreement comes down to utility timing.

May 07, 2026

MarketStack Meta-Analysis | 16 posts | Apr → 7 May 2026

Not investment advice. Written & fact-checked by Claude Opus. Articles sourced on the MarketStack Terminal. Please note, Claude can make mistakes. There was such little divergence in view across Substack authors that I ran it vs the sell side.

The street is split, Substack is not.

The sell side can’t agree on where uranium is going. Bank of America is calling $135/lb. Goldman has it at roughly $91 by end-2026. Scotia is anchored at $80. One sell-side commentator called the $80–$150 spread “handwaving rather than analysis,” and that’s about right — the dispersion is the story.

Read across the Substack universe over the same window and the picture inverts. 16 posts, 11 authors, 3 weeks. Not a single bear case. Metals and Miners prints an explicit $200/lb structural target. Aurelion Research’s just-published Uranium Primer models a 2040 supply gap of 109–150 million pounds annually — a 63% shortfall against demand. Asymmetric Research traces a public sentiment trajectory of cautious → constructive → “outright bullish” across 2025. Where the street hedges, the independent analysts don’t.

This piece is a cross-check: where the two corpuses converge, where the Substack authors are running ahead, and where the sell side has facts the Substackers haven’t reckoned with.

Where they agree

The structural deficit. Both sides anchor on the same underlying picture: long mine lead times (15–20 years per Aurelion), 13 consecutive years of utility contracting below replacement rate (SEQH), and a fuel cycle hollowed out by two decades of underinvestment. Goldman’s updated supply-demand model — a 1.91 billion pound cumulative net deficit through 2045 — is the most rigorous expression of what the Substack corpus is also describing in pieces.

The AI demand thesis. Universal. Aurelion’s 165% data-center power-consumption growth by 2030 is the same trade as the sell-side citing Meta’s 7.8 GW of new nuclear agreements. Annualize This frames it as “Big Tech PPAs”; the banks frame it as “hyperscaler procurement.” Same data, different vocabulary.

Long-term contract pricing. Tight convergence: $90/lb LT high-since-2008 (sell side) ≈ SEQH’s $91.50–$93.00 range ≈ Aurelion’s $90 figure ≈ Goldman’s Cameco-linked LT midpoint of ~$100.

Spot level. Both put it at ~$86/lb in early May.

The equity hierarchy. Producers and near-term developers preferred over explorers. Cameco as core long. Kazatomprom as the supply-side proxy. UEC carrying ~80% Strong Buy with a $18.95 consensus PT on the street; SEQH lists CCJ, UEC, LEU, DNN, UUUU among core. The street and the Substacks are buying the same paper.

Where the Substacks go further

Price targets. This is the cleanest divergence. Bank of America’s $135 is the most aggressive number on the street. Metals and Miners’ $200/lb is roughly 50% above it. Aurelion notes the inflation-adjusted 2007 peak at ~$240 and uses it as a reference. Junior Mining Pro builds a thesis on STMN trading at $157/hectare versus IsoEnergy at $33,764. The Substack corpus is willing to print numbers the sell side won’t underwrite.

Cameco’s “release valve.” Sell-side framing is that utilities are patient and inventories are larger than analysts can see. Asymmetric Research is reading exactly the opposite signal out of Cameco’s Q1 results: the product-loan facility — historically the spot market’s pressure-release mechanism — is tightening, with inventory at a record low and new loan terms harder than precedent. Same Q1 print, opposite interpretation. This is the most substantive single-data-point divergence in the corpus.

HALEU concentration. Centrus Energy (LEU) is named bullishly by btyc, SEQH, and StonkChris. btyc explicitly calls it “most investable” — the only profitable HALEU player with a $3.9 billion order backlog. Sell-side fuel-cycle exposure is more diversified; the Substacks have already concentrated.

Kazakhstan as something more than a guide. The street read of Kazatomprom’s 71.5–75.4 Mlb 2026 guide is “above expectations, below caps.” The Hoot reads the underlying policy: production ceiling cut from ~85M to ~77M lbs with another 10% reduction planned for 2026, plus a 100,000-tonne strategic reserve that locks up four years of national output and removes ~5% of world primary supply from the market. The banks see a quarterly guide; The Hoot sees a structural withdrawal.

Where the sell side has facts the Substacks ignore

Equities have outrun the commodity. Uranium stocks are up roughly 40% year-to-date against what Sprott itself called a “rather muted” commodity move. None of the sixteen Substacks engages with this. The corpus treats current valuations as if the commodity has already moved, when in fact the move is in the equities. Sprott reads it as forward-looking; the sell side is more nervous.

Q1 was financial buying, not utility demand. The Sprott Physical Uranium Trust and AI-narrative vehicles drove Q1 spot — not utility contracting. Asymmetric implies the utility wave is imminent; the corpus does not acknowledge that the price action so far has been financial-buyer-led, and that buying capacity is finite.

Q1 volatility. Spot peaked at $101.41 before the geopolitical pullback to $86. The Substacks treat the current level as a stable base. The 15-dollar round trip is invisible in the corpus.

The 1.91 billion pound cumulative deficit number. Goldman’s modelling is more rigorous than any single Substack figure. Aurelion’s 2040 annual deficit is the only directly comparable output, and it’s a different metric.

The actual disagreement: utility timing

If you strip the cross-check to one sentence, it’s this. The Substacks and the banks are looking at the same utility patience and reading opposite signs.

The sell-side honest read is right thesis, wrong timing. Desks have been early since 2023. Utilities are unaccountably patient. Hidden inventory buffers may be larger than anyone can see. Equity valuations have priced a move the physical hasn’t delivered.

Asymmetric Research is making the inverse argument. Patience is the catalyst. Ex-China inventories will fall below two years of forward coverage by the end of the three-year window. Historical contracting precedent (2005–2010) ran 120% of reactor requirements; current contracting runs ~60%. When utilities re-enter, the demand swing is 40 Mlb/yr above replacement — 100+ Mlb above 2025 levels. The longer they wait, the more violent the leg.

Same observation, opposite signs. The trade dispute is timing, and the Substacks have committed to a sooner answer.

What it means for SMR equities

The two universes drift apart fastest on the next-generation reactor names.

The sell side carries SMR/OKLO/NNE through its broader nuclear coverage with relatively even-handed ratings. The Substacks are sharper. btyc explicitly flags OKLO insider liquidation in May 2026 and non-binding or partial customer commitments. 90s.pm.investing prices NuScale at a P/S of ~98x against execution risk that has already wiped out the TVA/ENTRA1 program, delayed Romania to 2034, and lost Fluor as primary contractor. X-Energy gets the most generous read — “ARM for the AI era” — but with a flag against the October 2026 lockup expiry.

If you want a single takeaway: the Substack universe is more bullish than the street on the commodity and more disciplined than the street on the speculative equities.

The honest read

The Substack universe is, collectively, taking the more aggressive, sooner, higher-target side of the trade than the banks. That’s either alpha — these analysts are smelling the next leg before the desks finish modelling it — or it’s positioning bias. Eleven independent authors writing for readers who are already long uranium is a sample with a known prior. No bear voices appear. Equities-ahead-of-physical is the awkward fact, and the corpus simply doesn’t address it.

Both can be true. The Substacks may be early and the desks may be conservative and the equity valuations may be premature. None of those are mutually exclusive.

The signal worth pricing isn’t whether to be long uranium — both universes agree on that. It’s the gap. When sixteen independent voices are running 50% above the most bullish bulge-bracket target with no acknowledgement of the equities-physical disconnect, the asymmetry isn’t in the commodity. It’s in the consensus.

The editorial position MarketStack readers should take away is the same one Asymmetric, The Hoot, and Aurelion are converging on — but with the utility-patience caveat the sell side won’t let go of, kept clearly in view. The discipline is keeping both halves of it in your head at once.

Visit MarketStack The Edit · MarketStack Terminal

MarketStack is free today. But if you value my work, you can pledge for a future subscription. MarketStack is an independent, anonymous publication summarising publicly available commentary and views from across financial media. Nothing here constitutes financial advice or a recommendation to buy, sell or hold any security. All views are a synthesis of public information. Past performance is not a guide to future results. This publication is not authorised or regulated by the Financial Conduct Authority. The author writes anonymously in a personal capacity.

Sources

Uranium Primer: The Nuclear Renaissance Powering AI — Aurelion Research

Photonic Sensors for In-Core Nuclear Monitoring — SEQH Capital Research

Daily Nuclear & Uranium Market Recap — SEQH Capital Research

THE COMING URANIUM SQUEEZE: The 85M lb Deficit, the A.I. Demand Shock, the Hollowed-Out U.S. Production Crisis & Why Uranium Equities Will Massively Outperform! — Metals and Miners

X-Energy (XE) — 我們畫樹 — 90s.pm.investing

NuScale Power (SMR) — 我們畫樹 — 90s.pm.investing

Oklo Inc. (OKLO) — 我們畫樹 — 90s.pm.investing

JMP Microcap Uranium Holding: Drills Turn Next Week, 3km From World’s Highest-Grade Uranium Discovery — juniorminingpro

Nuclear Thematic — Annualize This

新一代核能公司盤點報告 (Next-Generation Nuclear Energy Company Inventory Report) — btyc

Uranium — Release Valve Running Dry? — Asymmetric Research