LEU › analyze
LEU · Analyze
Date: 2026-09-10 | Price: $165.87 (post-mkt $167.00) | 52wk Range: $142.13–$464.25 | Market Cap (pre-raise): ~$3.31B | Sector: Energy / Uranium (Enrichment & Technical Solutions)
A portfolio-specific passage was removed from the public build.
Context. A full differential re-analysis was completed only 14 days ago (analyze-from-before-2026-08-27.md, ACCUMULATE 7.8). Q3 FY2026 does not print until ~early Nov, so no new quarterly statement exists — every TTM fundamental ties exactly to the Aug-27 figures. What moved in 14 days is entirely price and one corporate action: the stock fell 14.85% and Centrus priced a $500M equity-and-warrant offering on Sept 9. This report re-runs the full pipeline but leans on the verified Aug-27 base rather than re-deriving settled numbers.
What changed since Aug 27 (the whole delta)
| # | Item | Direction | Detail |
|---|---|---|---|
| 1 | Price −14.85% | ⚖️ cheaper | $194.78 → $165.87 on unchanged backlog/cash/capex. EV/Backlog re-compressed 0.68x → 0.57x (back near the June baseline's 0.60x). |
| 2 | $500M offering priced Sept 9 (closes ~Sep 11) | ⚠️/✅ mixed | 500K shares + 2.006M pre-funded warrants + 6.99M common warrants (struck $226.86–$362.98, 2028–31). Combo price $199.64. ~+12.6% immediate dilution (~19.95M → ~22.46M sh); ~$480M net cash added → net cash ~$1.17B. Drove today's −8.6% drop. |
| 3 | Radiant/Kaleidos HALEU offtake signed (Sept 9) | ✅ | Third SMR customer after Oklo + X-energy; includes prepayments funding capacity. Dollar value undisclosed. Overshadowed by the raise. |
| 4 | "Gain on Sale of Security" RESOLVED | ✅ quality | 10-Q read directly: it is "Investment income" — ordinary interest on the cash pile (Yahoo taxonomy mislabel), recurring 5+ quarters ($8.0M→$12.9M→$16.5M→$17.0M→$16.3M), scaling with cash. Not a one-off. But it will shrink as cash is deployed into capex (the raise refills it near-term). |
| 5 | General Matter — new HALEU entrant (missed at baseline) | ⚠️ structural | SF startup won an equal $900M DOE HALEU award (Jan 2026) to build a second domestic source at Paducah. Unlicensed/unbuilt; first output ~2034. DOE's explicit goal: end single-source dependency. |
| 6 | Evergreen re-rated 8 → 7 | ⚠️ | Independent verification reframes the moat as a policy-engineered monopoly with a visible ~2032 expiration, not an indefinitely-renewing franchise. Monopoly claim itself CONFIRMED (see Moat). |
| 7 | Jefferies initiates Hold, PT $169 (Sep 3) | ⚠️ mild | First new cautious desk; Hold count 6→7, mean target $253→$248. Still 0 Sells, still Buy-heavy. |
| 8 | Short interest building | ⚠️ | 27.9% of float, 8.2 days to cover, up from prior month — shorts pressing the pullback. |
| 9 | DOE FY2027 TS/HALEU cascade line | → unchanged | Still proposed unfunded (~$0.8B / 18% of backlog exposed); no congressional resolution. The #1 named risk, unmoved. |
| 10 | Uranium spot ~$89.49/lb | → flat | vs $87.55 — not a driver. Sector equities de-rated 25–50% off highs (UEC −50%); LEU's move is part sector, part dilution. |
Break trigger #4 check ("equity raise at <$150/share") — did NOT fire. The raise priced at $199.64 combo, well above the $150 threshold. It is dilutive but it is not the trigger event. However, the Aug-27 REFRESHED claim that "dilution risk is lower than baseline implied / prepayments reduce reliance on equity" is now partly superseded — they raised $500M in equity anyway, two weeks later.
1. Fundamentals — health confirmed, price is the story
Every TTM figure ties exactly to Aug-27 (same four-quarter window). Confirmed vs vendor data this pass:
| Metric | Value (TTM) | Note |
|---|---|---|
| Revenue | $473.9M | +14% YoY at Q2; inside $450–500M FY guidance |
| Gross margin | 23.7% | vs FY25 26.2% — still compressing (SWU cost +13%, volume −23% at Q2) |
| Operating margin | 1.6% | thin during ramp |
| Net income | $48.5M | ties Yahoo netIncomeToCommon |
| Interest expense | $15.7M | flat — capitalized into construction (ASC 835-20). Baseline's bear case stays retracted. |
| OCF | −$55.0M | — |
| Capex | −$108.8M | accelerating within the window: $4.4M→$9.6M→$23.2M→$71.6M by quarter |
| FCF (quarter-sum) | −$163.8M | vendor snapshot field shows −$127.3M (~$36M gap); use the quarter-sum — Yahoo's own per-quarter FCF also sums to −$163.8M |
| Net cash (pre-raise) | +$691M | $1.869B cash − $1.178B debt; ~$1.17B post-raise |
| Debt/Assets | 46.5% | (Q2'26) vs 49.7% FY25 — modest de-lever |
| Shares | 19.95M → ~22.46M post-raise | dual-class: 19.23M Class A + 0.72M Class B (Yahoo sharesOutstanding counts Class A only) |
Vendor traps cleared this pass (all documented Playbook pitfalls, all live on this name):
- enterpriseValue $2.93B is wrong — Yahoo used non-diluted market cap; hand EV = $2.62B (pre-raise).
- enterpriseToEbitda 290x rests on a $10.1M EBITDA field vs a $72.7M quarter-sum — do not use.
- forwardPE/epsCurrentYear mix non-GAAP add-backs — unreliable, not used for the trim (see Valuation).
Verdict: health confirmed, not deteriorated. The substantive gain is data quality — the investment-income line is resolved as legitimate and recurring, removing an earnings-quality unknown. The single most important number is the price: −14.85% on identical fundamentals.
2. Moat — monopoly CONFIRMED, evergreen LOWERED
Independent competitive-timeline verification (the gap the Aug-27 pass left open):
| Competitor | Status Sep 2026 | HALEU-capable? | First commercial HALEU |
|---|---|---|---|
| Centrus | Only NRC-licensed and operating HALEU producer; demo cascade since Oct 2023; $900M DOE task order to scale to 12 MT/yr | Yes — now | Producing; scaling through decade |
| Urenco USA | LEU+ (≤10%) authorized Sep 2025, deliveries mid-2026; separate dedicated HALEUF planned | Partial (bottom-of-band, <19.75% SMRs need) | HALEUF ~2032 |
| Orano (Project Ike) | $900M DOE award (LEU only); NRC accelerated review, decision by 2027-04-30; not started | No (LEU only) | early 2030s, LEU only |
| General Matter (new) | Equal $900M DOE HALEU award Jan 2026; unlicensed, unbuilt | Intended | ~2034 |
| GLE (Silex/Cameco) | Declined HALEU; chose tails re-enrichment | Possible, not pursued | speculative |
Monopoly claim: CONFIRMED true as of Sep 2026. Nearest full-spec competitor (Urenco HALEUF) ~6yr out; nearest funded pure-play (General Matter) ~8yr out. The "5–7yr lead" holds.
Moat rating: 8.5/10 (confirmed). Evergreen: 8 → 7 (lowered). The distinction that matters: this is not a network-effect or switching-cost franchise — it is a temporary, policy-engineered monopoly with a known expiration (~2032). The same government that anchors LEU has already funded a second source (General Matter, equal dollars, explicit intent). The barrier is regulatory + construction lead time, not unreplicable capability — durable to ~2031–32, then engineered to erode by design.
Stress-test: moderately attackable, but the attacker is DOE, not a rival firm, and the clock is the only defense. Centrifuge tech is proven and available to Urenco/Orano/General Matter, so neither capital nor IP is the moat — the NRC-license-plus-construction lag is.
Revenue-stream durability: LEU segment (commercial SWU + Oklo/X-energy/Radiant offtakes, $2.4B definitized) durable; Technical Solutions segment (DOE HALEU cascade, ~$0.8B) policy-dependent — the FY2027 appropriations exposure, confirmed in Centrus's own risk disclosures.
Demand-timing caveat: SMR commercialization is now looking like a 2030s story (TerraPower Natrium slipping to 2030+; a former NRC commissioner called 2027 SMR timelines "not credible"). This pushes out when the $4.5B backlog converts to cash — but slower SMR delivery also buys Centrus more runway before competition matters.
3. Valuation — EV/Backlog, re-run post-raise
Forward-P/E anchors remain inapplicable (investment-income line + non-GAAP add-backs make the EPS untrustworthy; Graham √(22.5×$1.90×$42.45)=$42.45 vs $165.87 = 3.9x, prices a stable industrial not a pre-scale strategic monopoly). Primary method is EV/Backlog, as at baseline.
Post-raise inputs: ~22.46M shares, net cash ~$1.17B, backlog $4.5B (Radiant adds an undisclosed amount). - EV/Backlog now ≈ 0.57x (EV ~$2.55B / $4.5B) — cheaper than Aug-27's 0.68x, still below the 0.8–1.5x long-cycle-infrastructure peer range.
| Multiple | Implied price/sh (post-raise) |
|---|---|
| 0.6x | ~$172 |
| 0.8x | ~$212 |
| 1.0x | ~$253 |
| 1.3x (bull) | ~$313 |
Dilution haircuts per-share values ~4–5% vs the Aug-27 figures (~20M sh), partly offset by the extra $480M cash.
| Scenario | Driver | Fair value |
|---|---|---|
| Bear | TS line stays unfunded, capex overruns, backlog haircut to ~$4.0B, 0.5–0.6x | $135–165 |
| Base | LEU-side execution on schedule, TS resolves neutrally, 0.8–1.0x | $210–255 |
| Bull | Saudi optionality / further SMR offtakes lift backlog >$5B, re-rate 1.2–1.5x | $300–385 |
Fair value: $175–270 (base-weighted; low end pulled down by TS risk + dilution). Analyst mean target $248 sits inside it.
Entry: $150–175 (shifted down from the Aug-27 $170–195 band to reflect dilution + the sector de-rating). Current $165.87 is inside the add zone — the weakness the Aug-27 report said to wait for has arrived. Strong buy <$150 (near 52wk low $142).
Trim: 1.3x EV/Backlog (~$310–315/sh post-raise), NOT a forward-P/E multiple. Basis per CLAUDE.md's pre-earnings-scale exception: forward EPS mixes non-GAAP add-backs with the (now-understood but volatile) investment-income line, so a P/E trim would anchor to a number this report won't vouch for. EV/Backlog is what the thesis is actually built on (backlog > market cap) and self-updates as backlog/cash move.
4. Synthesis — the two real debates
Debate A — is the $500M raise bullish or bearish? Both, and the net is mildly positive for a construction-phase cash-burner. Bearish: ~12.6% dilution, priced below the 52wk high, and the timing (two weeks after saying prepayments reduced equity reliance) reads as opportunistic-at-best. Bullish, and weightier here: it retires the funding-overhang risk entirely — FY2026–27 capex ($350–500M guide) is now fully covered without touching the operating cash story, and the sharpest tail risk for a company burning −$164M FCF/yr is a forced raise into weakness, which is now off the table. For a name whose whole thesis is "survive the toll-bridge construction years to reach the backlog," pre-funding the bridge is worth more than the dilution costs. Resolution: net positive, but it caps how much conviction can rise — dilution is a permanent per-share cost.
Debate B — does cheaper price + resolved funding outweigh dilution + evergreen downgrade + a funded second source? Fundamentals says the business is unchanged and 15% cheaper (buy signal). Moat says the durability window is now dated (~2032) and DOE is actively funding its end. Sentiment says mood cooled but on a liquidity event, not fundamentals, with shorts pressing. Resolution: the near-term risk/reward improved (cheaper entry, funding secured, third offtake, monopoly independently confirmed), while the long-term durability story got marginally worse (evergreen 8→7, General Matter). For a satellite/speculative position held for the construction-to-cash-flow arc through the late 2020s — before ~2032 competition matters — the near-term improvement dominates. Hence ACCUMULATE, but conviction holds rather than rises.
Verdict: [7.6/10] — ACCUMULATE on this weakness
(was 7.8 on Aug 27 — a hair lower, not higher, despite the cheaper price)
The core thesis is intact and independently re-confirmed: sole US-licensed and operating HALEU producer, DOE-anchored $900M contract, backlog ($4.5B) exceeding market cap, now a third SMR offtake (Radiant). The 15% price drop is a liquidity/dilution event, not fundamental deterioration — EV/Backlog re-cheapened to 0.57x, and the $500M raise removes the single sharpest tail risk (a forced raise into weakness) for a name burning −$164M FCF during construction. Against that, two genuine structural haircuts the Aug-27 pass didn't fully have: ~12.6% dilution (a permanent per-share cost) and a dated moat — DOE has funded General Matter as an equal-dollar second source and Urenco is self-funding a HALEU line, so evergreen drops 8→7 and the monopoly now carries a visible ~2032 expiration.
Conviction holds ~flat (a notch down) because the cheaper price is captured in the entry zone, not the conviction number, and the dilution + second-source funding offset it. The verdict shifts from Aug-27's "HOLD here, add on weakness" to ACCUMULATE — because the weakness arrived and the funding overhang is gone.
A portfolio-specific passage was removed from the public build.
Break triggers (carried; #4 reworded): 1. DOE cancels or materially reduces the Piketon HALEU Production Contract. 2. Piketon cost overrun >30% without a commensurate DOE backstop (untested — self-reported on-schedule only). 3. Backlog falls below $2.5B for two consecutive quarters. 4. Reworded: a further equity raise at <$150/share, OR the common warrants (struck $227–363) start converting in size and stack material dilution on top of the Sept raise. 5. FY2027 federal budget finalizes with the Technical Solutions/HALEU cascade line unfunded and no replacement signed by year-end 2026. 6. NEW watch (not yet a break): General Matter or Urenco reaches an NRC construction/operating license — the first hard signal the ~2032 monopoly-expiration clock is running on schedule rather than slipping.
Upgrade conditions: TS budget line restored or replaced; TTM FCF burn narrows as Piketon starts contributing revenue; a fourth SMR/utility offtake; the raise's proceeds visibly accelerate the Oak Ridge/Piketon build without further dilution.
What this pass did NOT test
- Exact size/accounting of the Radiant prepayment and the Sept offering's final net proceeds (using ~$480M estimate; confirm at 10-Q).
- Piketon construction detail beyond management's earnings-call statements (on-schedule, first centrifuge 2026) — still single-sourced.
- Whether Russia's retaliatory export ban lapsed or renewed — framed originally to run "until 2026"; a reversal is a low-probability, high-impact tail risk to the whole ghost-demand thesis. Flag for next pass.
- SWU forward contract pricing — only spot refreshed (~$89.49/lb).
Sources: python .mcp/fin.py LEU, Yahoo Finance MCP (stock info, quarterly + annual income/balance/cashflow, recommendations, insider transactions, holder info), Centrus Q2 FY2026 10-Q (SEC EDGAR, filed 2026-08-01, fetched directly this pass), Centrus/PRNewswire Sept 9 offering pricing + Radiant offtake releases, SEC 424B5 prospectus, World Nuclear News (Orano Project Ike, Urenco HALEUF, General Matter Paducah), ANS Nuclear Newswire, NRC HALEU pages, Motley Fool Q2 2026 transcript, Arms Control Association (Saudi 123 pact), metalcharts.org (uranium spot), 24/7 Wall St / XTB (sector de-rating). Aug-27 baseline: analyze-from-before-2026-08-27.md.