NuScale Power Corporation (NYSE: SMR) — A Licensed Design in Search of a Buyer, Priced for a Sold-Out Order Book It Does Not Have

Independent Equity Research Sector: Industrials · Electrical Equipment / Small Modular Reactors Report date: 2026-06-20 · Price: $11.74 (2026-06-18 close) · Market cap: ~$3.8B (Class A) / ~$4.0B fully diluted · CIK: 0001822966

⚡ Claude’s Take

This is the author’s own independent opinion and general information only — not investment advice. The analysis that follows takes no position and carries no price target; the only view expressed anywhere in this article is in this block.

Verdict: AVOID at $11.74 — a deeply speculative, governance-impaired pre-revenue option, not an investment. Not-a-short (net cash + thematic squeeze risk + genuine NRC scarcity). Accumulate-only, and only as a sized-to-zero call option, near the cash-plus-credential floor in the ~$4–6 zone. Conviction: medium.

The bull story — “the only NRC-approved small modular reactor design, sitting on a fortress balance sheet, levered to an inexhaustible AI-power demand wave” — is each clause individually true and collectively misleading. After eighteen years and roughly $2 billion of cumulative spend, NuScale has never signed a single binding order to sell a reactor module. Its one flagship customer (the UAMPS Carbon Free Power Project) cancelled in November 2023 when the levelized cost ran away from it. In August 2025 the company responded not by lowering cost but by outsourcing its entire sales function to ENTRA1 Energy — an entity ~75% controlled by a businessman with a decade-long joint venture with Fluor (NuScale’s controlling shareholder, and the former employer of both the CEO and the Chairman), which NuScale does not disclose as a related party, and to which it booked a ~$507 million cash charge in 2025 against a non-binding memorandum of understanding. Meanwhile Fluor — the most informed holder on the register — converted its supervoting stake and sold ~$1.7 billion, collapsing from ~50% to ~4% in six months, dumping all the way down from $19 to $11.63. A securities-fraud suit (Truedson v. NuScale) followed in February 2026. The most-informed insider is sprinting for the door while the company funnels half a billion dollars to a Fluor-orbit agent against a handshake.

What you are buying at $11.74 is ~$2.5–3.0B of enterprise value — about $3/share of net cash and ~$8.50 of pure commercialization option value — on $31M of declining revenue (~80–97x sales), with first module delivery the filings themselves place at “2031 or beyond.” A reverse-DCF says the price embeds roughly 120+ delivered, margin-positive modules — more than the entire non-binding 72-module pipeline, fully converted and built flawlessly. That is not a margin of safety; it is a narrative. The factor tape agrees: beta ~2.3–3.0, specific vol ~77%, no value/quality/momentum loading, a twin of the clean-energy/nuclear thematic ETFs (PBW, BE, LEU), down ~78% off its October-2025 high and ~70% year-on-year — a crowded high-volatility thematic vehicle now in a falling-knife drawdown. The real NRC credential and the ~$1B of net cash keep it off the short list and put a soft floor under the equity; everything above that floor is a bet that a flawless commercialization — one that has eluded the company for eighteen years and just cost it its flagship customer and its founding shareholder — finally arrives. What flips me bullish: the first genuinely binding, economically-priced module purchase order from a creditworthy counterparty (RoPower FID with firm offtake, or a real data-center contract), proving the design converts to financeable demand. What flips me (further) bearish: a Milestone Contribution 2 cash transfer to ENTRA1, an adverse Truedson development, or RoPower slipping — any of which drains the only thing holding the floor.

Tag: “A licensed blueprint with no buyer, an exiting founder, and a sales agent collecting the receipts.”

📈 Stock Price Action — Five-Year Event Map

NuScale has completed a full bubble-bust round trip and is now grinding through a fresh falling-knife leg. From a de-SPAC reference near $10 (May 2022), the stock collapsed ~80% to an all-time low of $1.92 (Jan 2024) after its flagship customer walked, then re-rated more than 27-fold into an all-time high of $53.43 (15 Oct 2025) on the AI-power/nuclear-renaissance mania, before giving back ~78% to $11.74 today. The 52-week range is $9.16–$53.43; the stock sits ~78% below its high and ~28% above its April-2026 low, below all major moving averages (200-EMA ~$17.7). This is factual price history, not a forecast.

Cycle narrative. (1–2) The de-SPAC priced NuScale as a soon-to-commercialize reactor vendor; the November 2023 cancellation of the UAMPS Carbon Free Power Project — the company’s anchor customer, lost to ~$58→$89/MWh cost escalation — destroyed that premise and drove a ~28% layoff and the $1.92 trough. (3) Through 2024 the stock was swept up in the broader AI-electricity-demand trade that lifted the entire nuclear complex regardless of company-specific fundamentals. (4) The 2025 melt-up combined a real milestone — NRC approval of the uprated 77 MWe six-module Standard Design Approval (May-2025) — with narrative milestones: Trump’s nuclear executive orders and ENTRA1’s non-binding TVA agreement for “up to 72 modules.” Management sold ~$1.3B of stock via ATM into this spike (the one genuinely shrewd capital act). (5–6) As the 2025 financials revealed the ~$507M ENTRA1 milestone charge and the dilution, and as Fluor liquidated ~$1.7B from ~50% to ~4%, the stock round-tripped to a new falling-knife low. (7) The recent bounce is thematic, not order-driven. Each numbered move ties to a research-log entry and is cross-referenced to earnings prints, 8-K events, and the news feed.

1. Executive Summary

NuScale Power Corporation is the most credentialed and least commercial of the listed small modular reactor (SMR) developers. It is the only company with a U.S. Nuclear Regulatory Commission–approved SMR design — in fact two: the original 50 MWe design (FSER 2020) and the uprated 77 MWe, six-module “VOYGR” Standard Design Approval (SDA) granted in May 2025. No competitor has even applied for an SMR SDA. That is a genuine, hard-won regulatory asset.

It is also, on the evidence, the weakest business in the group. NuScale is a capital-light reactor design licensor and equipment supplier — not an EPC, not an owner-operator, not (yet) a manufacturer at scale. It books revenue from licensing and engineering work, which in FY2025 totaled just $31.5M and fell year-on-year. After eighteen years of development and roughly $2B of cumulative cash consumed, it has never signed a binding order to sell a reactor module. Its flagship customer — the UAMPS Carbon Free Power Project — cancelled in November 2023 when costs escalated, the definitive lived proof that an NRC-approved design does not automatically convert into a financeable plant.

The economics and the governance are the story. FY2025’s headline GAAP operating loss of −$689.6M is dominated by a ~$507M cash charge (“Milestone Contribution 1”) that NuScale owes ENTRA1 Energy — the entity to which, in August 2025, it handed exclusive worldwide commercialization rights, and which is ~75% controlled by a businessman with a decade-long Fluor joint venture (Fluor being NuScale’s controlling shareholder and the former employer of both CEO and Chairman). The trigger for that half-billion-dollar payment was a non-binding MOU between ENTRA1 and the Tennessee Valley Authority for “up to 72 modules.” NuScale does not disclose ENTRA1 as a related party; a securities-fraud complaint (Truedson v. NuScale, D. Or., Feb-2026) alleges exactly this. Simultaneously, Fluor sold ~$1.7B and cut its stake from ~50% to ~4% in six months, selling all the way down from $19 to $11.63.

Underneath the noise, the balance sheet is genuinely strong: ~$1.0B of liquidity, net cash, no debt, built by opportunistically selling ~$1.3B of equity into the 2025 price spike. The underlying operating burn (ex-milestone) is a manageable ~$200M/year, implying a multi-year runway — but episodic ENTRA1 milestone transfers (Milestone Contribution 2 and beyond, potentially ~$1.1B+ over the pipeline) are unquantified wild cards that each lop years off it.

At $11.74 the market assigns ~$2.5–3.0B of enterprise value (~80–97x trailing sales) to a company whose first module delivery the filings place at “2031 or beyond.” A reverse-DCF implies the price embeds the flawless delivery of more modules than the entire non-binding TVA pipeline. The factor tape reads as a crowded, high-volatility thematic vehicle in a falling-knife drawdown — not a value name, not a momentum name. The analysis below takes no position; it argues the evidence. The single labeled exception is the opinion block above.

2. Business Overview

What NuScale does. NuScale designs the NuScale Power Module (NPM) — a self-contained, factory-fabricated integral pressurized-water reactor that generates 77 MWe (uprated from the original 50 MWe). Multiple modules are assembled into a VOYGR power plant, offered in four-, six-, and twelve-module configurations (VOYGR-4 ≈ 308 MWe, VOYGR-6 ≈ 462 MWe, VOYGR-12 ≈ 924 MWe). The design’s signature engineering claim is passive safety: natural-circulation cooling that requires “no operator or computer actions, grid connection or emergency back-up power” to safely shut down, the feature underpinning the NRC’s willingness to approve a smaller emergency-planning zone.

How it makes money — and how little it makes. NuScale is, in business-model terms, a technology licensor and equipment integrator, deliberately capital-light (FY2025 capex was ~$0.5M). It does not build plants (that is Fluor’s EPC role) and does not own them (that is the utility/developer’s role). Its three intended revenue legs are (i) design licensing, (ii) long-lead-material and NPM equipment supply, fabricated through a supply chain anchored by Doosan Enerbility (forgings) and others, and (iii) life-of-plant operating/engineering services. None is yet recurring at scale. FY2025 revenue of $31.5M was essentially Romania-related licensing/FEED work; it declined from $37.0M in FY2024, and the multi-year trend is small and lumpy ($2.9M → $11.8M → $22.8M → $37.0M → $31.5M across 2021–2025). Gross margin is volatile and meaningless at this scale (36% in FY2025).

The ENTRA1 commercialization layer. The defining structural feature, established in August 2025, is that NuScale has outsourced its entire sales and commercialization function to ENTRA1 Energy under a “Partnership Milestones Agreement” (PMA). ENTRA1 holds exclusive worldwide rights to market and sell NuScale technology, with non-circumvention clauses barring NuScale from selling around it. In return NuScale pays ENTRA1 “Milestone Contributions” per module — 15% of a defined value on a non-binding MOU, 35% on a binding power-purchase agreement, 50% on a binding equipment order — even though ENTRA1 retains “sole and full discretion” over whether to buy NuScale’s modules at all, subject to a maximum-sale-price cap on NuScale’s own equipment. In effect, NuScale pays a third party for the prospect of orders while ceding pricing power and the customer relationship. (Governance implications are discussed in the Capital Allocation, Changes, and Risk sections.)

Customers and end markets. Target end markets are utility-scale baseload power, data-center/AI captive power, district heating, desalination, hydrogen, and industrial process heat. The actual customer base is thin: one funded development project — RoPower in Romania (a Nuclearelectrica/Nova Power & Gas vehicle at Doicești) — and a pipeline of non-binding agreements routed through ENTRA1 (the TVA “up to 72 modules / 6 GW” MOU being the headline). There are zero binding module purchase orders. Revenue is therefore neither recurring nor, today, commercial; it is development-stage cost reimbursement and licensing.

Verdict. A real reactor design business — capital-light, technically credentialed — wrapped in a commercialization structure that has, so far, produced reimbursable development work rather than product sales, and that now routes the customer relationship and a large slice of future economics through a single exclusive agent. It makes money the way a pre-commercial engineering firm does: by spending investor capital faster than it earns fees.

3. Industry Dynamics

The demand side is genuinely attractive — on paper. The structural case for nuclear is the strongest it has been in a generation: COP28’s pledge to triple nuclear capacity by 2050; the U.S. executive orders targeting a quadrupling of nuclear capacity (toward ~400 GW); surging data-center/AI electricity demand seeking firm, clean, 24/7 baseload; the ~92% capacity factor that makes nuclear uniquely dispatchable among low-carbon sources; and bipartisan policy support (ADVANCE Act, DOE loan/grant programs). SMRs specifically promise factory fabrication, smaller upfront capital, siting flexibility, and shorter build times than gigawatt-scale plants. If even a fraction of the announced data-center nuclear demand materializes, the addressable market is enormous.

The supply side and capital cycle are brutal — the Marathon lens. Against that demand stands a textbook capital-cycle problem. No commercial SMR operates anywhere in the West; the only operating SMRs are in China and Russia, none NRC-approvable. Meanwhile a wall of subsidized, well-capitalized competitors is racing for the same handful of first deployments: GE Hitachi’s BWRX-300 (already under construction at Ontario’s Darlington — the Western front-runner by physical progress), TerraPower’s Natrium (Bill Gates–backed, building at Kemmerer, WY), X-energy’s Xe-100 (Amazon-backed), Westinghouse AP300, Holtec SMR-300, Rolls-Royce SMR, and the advanced-fission developers (Oklo’s Aurora, Kairos, etc.). This is the Marathon signature: high announced returns and lavish policy capital attract a flood of supply ahead of demonstrated demand, “bankers lubricating the cycle,” compressing the economics for everyone before the first plant earns a dollar.

The product is a commodity sold into a hostile cost environment. The output is an electron — a pure commodity. The competing technology is cheap, fast natural gas: GE Vernova’s gas-turbine backlog alone is enormous, and a combined-cycle plant can be built in ~2–3 years at a fraction of nuclear’s per-kW cost. The reference point for new nuclear economics is grim — Vogtle 3&4 came in around $16,000/kW and years late — and the entire SMR thesis is the unproven claim that factory modularity will break that curve. Until an Nth-of-a-kind plant is actually built at an economic LCOE, that remains a hypothesis. The UAMPS cancellation (CFPP target LCOE rising to ~$89/MWh before the customer walked) is the live data point arguing the curve has not yet been broken.

Regulation is both the moat and the millstone. NRC licensing is a multi-year, multi-hundred-million-dollar barrier that, once cleared, is genuinely hard to replicate — the source of whatever durable advantage exists in this industry. But it is also a cost and timing burden that pushes first revenue half a decade out and exposes developers to regulatory and political risk.

Verdict: structurally bad on free-market terms; viable only as a subsidized, regulation-protected niche with a 5–8+ year time-to-cash. The demand narrative is real but distant and contested by cheaper substitutes; the supply side is a crowded, policy-driven capital cycle in its enthusiasm phase. Returns in this industry will be policy-determined, not demand-determined, for the foreseeable future — exactly the configuration Marathon warns destroys capital.

4. Competitive Position

The one real asset: regulatory primacy. NuScale’s genuine differentiator is concrete and verifiable: it is the only SMR developer with NRC design approval, and now holds two approvals — the 50 MWe design (Final Safety Evaluation Report, 2020) and the uprated 77 MWe six-module Standard Design Approval, granted May 2025. No competitor has even submitted an SMR SDA application. In a business where the regulatory clock is the binding constraint, a multi-year head start through the NRC is a real advantage and the strongest single fact in the bull case.

But in Greenwald terms it is the weakest class of advantage — and it is depreciating. Competitive advantage must show up as customer captivity, scale economics, or a sustainable cost edge. NuScale’s edge is a supply-side/regulatory credential with none of those reinforcing characteristics: no installed base, no switching costs, no network effect, no demonstrated cost advantage (the opposite — its flagship plant got more expensive until the customer left), and no scale (it has sold nothing). A regulatory head start is a depreciating asset: GE Hitachi’s BWRX-300 is already under construction and expects its own NRC nod, and every competitor is advancing through the same pipeline. The approval buys time, not captivity.

The financial test confirms the absence of a moat. The Greenwald financial screen for a real advantage is persistent excess returns and stable market share. NuScale has deeply negative ROIC/ROE (pre-commercial), zero market share in a market that does not yet commercially exist, and declining revenue. There is no financial signature of a moat because there is no moat yet — only a credential and the hope that it converts.

The lived counter-evidence: UAMPS. The cleanest test of whether NRC approval equals durable demand already ran, and NuScale failed it. The Carbon Free Power Project — its anchor, government-supported, design-approved flagship — cancelled in November 2023 when target LCOE escalated to ~$89/MWh and the subscribing utilities walked. Regulatory approval did not make the plant financeable. That single episode is worth more than any quantity of MOU headlines.

Head-to-head. Versus OKLO (the closest listed pre-revenue twin): NuScale is better credentialed (two NRC SDAs vs Oklo’s earlier denied combined license application) and has revenue and more cash, but carries a worse commercial track record (a cancelled flagship), an exiting controlling shareholder, and the value-extracting ENTRA1 structure. Versus BWXT (profitable naval/medical/special-reactor incumbent) or GE Vernova/Constellation: those are real businesses earning real returns; NuScale is an option on a business. Versus BWRX-300: NuScale leads on paper approval, trails on physical deployment.

Verdict: durable competitive advantage absent. NuScale possesses a valuable but depreciating regulatory credential, not a moat. It is a first-mover on the licensing clock in a crowded, subsidized field, with no customer captivity and a flagship cancellation proving the credential does not, by itself, convert to financeable demand.

5. Growth History and Forward Opportunities

History: there is essentially none to speak of, by design and by misfortune. Reported revenue is a rounding error and non-linear ($2.9M → $11.8M → $22.8M → $37.0M → $31.5M, FY2021–25), composed of development-cost reimbursement, licensing, and FEED fees — not product sales. The defining historical event is negative growth: the loss of the UAMPS/CFPP project (Nov 2023) erased the only path to near-term plant revenue and forced a ~28% workforce reduction (Jan 2024). So the “growth history” is, candidly, eighteen years of pre-revenue development punctuated by the cancellation of the one project that mattered.

Forward opportunities — real but distant, contingent, and routed through ENTRA1.

• RoPower (Romania) is the most concrete: a VOYGR-6 at Doicești, with a license signed (Jul-2024) and a Nuclearelectrica investment decision approved (Feb-2026), supported by U.S. EXIM and DOE interest. This is the single most advanced customer step — but it remains pre-EPC, pre-final-investment-decision, with no firm module purchase order. It is the project to watch.
• The TVA / data-center pipeline is the headline number — “up to 72 NPMs / ~6 GW” — but it is an explicitly non-binding MOU, contracted with ENTRA1, not NuScale, and it is what triggered NuScale’s ~$507M payment out. AI/data-center captive power (the ENN, Standard Power–type deals) is the most exciting narrative opportunity and the thinnest contractual one.
• Optionality exists in district heating, hydrogen, desalination, and international markets (the design’s flexibility is real), but all are 2030s-and-beyond, all require an economic Nth-of-a-kind plant first, and all flow through the ENTRA1 economics.

The quality of any future growth is structurally capped by the ENTRA1 take. Even if the pipeline converts, ENTRA1 collects 15/35/50% milestone economics and caps NuScale’s module pricing. So NuScale bears the development cost and regulatory risk, while a substantial share of the upside — and the customer relationship — sits with its agent.

Verdict: low-quality, non-existent realized growth; high-variance, distant, contingent forward opportunity. The forward case is a genuine option on the SMR/AI-power theme, but it is unrealized, non-binding, half a decade out, and economically shared with ENTRA1. This is opportunity, not growth — and the company has been “on the cusp” of it for a very long time.

6. Financial Quality

Reframing the headline loss (the quality-of-earnings crux). FY2025’s GAAP operating loss of −$689.6M looks catastrophic and is widely misread. The driver is general & administrative expense exploding to $609.8M from $75.9M, of which ~$507.4M is “Milestone Contribution 1” (MC1) — the payment owed to ENTRA1 under the August-2025 PMA, triggered by the non-binding TVA MOU. Critically, this is not a reversible non-cash mark (the company’s warrants were already redeemed by December 2024; the often-cited “warrant fair-value” item was a FY2024 event). MC1 is “consideration paid to a prospective customer” under ASC 606 — expensed with no asset or revenue received in return — and it is a real cash obligation ($247.5M paid in 2025, ~$259.9M accrued and largely paid out in Q1-2026 as accounts payable drained by ~$264M). So the headline loss does overstate recurring operating deterioration, but the relief is limited: the offset is a genuine half-billion-dollar cash transfer to an affiliate-adjacent agent, not a paper entry that reverses.

The clean underlying run-rate. Q1-2026 shows the true operating shape with no milestone noise: operating loss −$57.5M, net loss −$46.7M, G&A back to $24.8M. Stripping MC1, FY2025 cash operating burn was roughly $190–220M (FY2025 OCF of −$459.6M plus ~$269M of MC1 cash ≈ −$190M; Q1-2026 underlying ≈ −$50M/quarter after the AP drain). SBC is modest (~$19M). So the recurring business burns ~$200M/year — large for $31M of revenue, but not the −$690M the headline implies.

Balance sheet: the genuine strength. At 31-Mar-2026: cash $341M + short-term investments $549M + ~$120M long-term investments ≈ ~$1.0B liquidity, net cash, no debt (capital leases ~$0.7M). Current ratio ~29x. Tangible book is positive (goodwill is only ~$8.3M; intangibles ~$26M). Net cash per share is ~$3, and BVPS is $7.13 — mostly cash. On the ~$200M underlying burn this is roughly a multi-year runway with no near-term financing cliff and no going-concern flag. The caveat is large: Milestone Contribution 2 (35% tier, estimated ~$16M/module, ~$1.15B across the pipeline) and further MC1 top-ups are unquantified and could each consume a meaningful slice of the cash if the pipeline “progresses.”

Cash flow quality. Operating cash flow has been persistently and increasingly negative (−$148.6M / −$183.3M / −$108.7M / −$459.6M across FY2022–25), with FY2025 inflated by the MC1 cash and a lumpy AP swing. FCF tracks OCF (capex is negligible). There is no cash generation; the company runs on its balance sheet, which it refills by issuing equity.

Dilution. Share count rose from 134M (Q2-2025) to 323.7M Class A (Q1-2026). Decomposed: ~111M is Fluor’s Class B→A conversion (up-C plumbing, already in the fully-diluted base), ~57M is genuine new ATM equity ($1.30B raised into the $20–53 spike), and the remainder is exercises. Fully diluted (residual Class B + options/RSUs) ≈ ~340–345M. The only cash-raising dilution is the ~57M ATM shares — and they were sold near the top, which is to NuScale’s credit.

Verdict: economics do not improve with scale — because there is no scale. This is a financing-dependent, pre-commercial design shop with declining negligible revenue, a ~$200M recurring annual burn, episodic multi-hundred-million-dollar cash transfers to its sales agent, and no binding customer contract after 18 years. The redeeming features are the capital-light model and the ~$1.0B net-cash runway — itself a product of selling equity into a thematic spike. It survives on cash, not cash flow.

7. Capital Allocation

The one genuinely shrewd act: opportunistic financing. NuScale sold ~$1.3B of equity via ATM into the 2025 price spike (issuance spanning roughly $20–$53), converting a thematic bubble into a fortress balance sheet — ~$1.0B net cash, no debt, no near-term financing cliff. For a pre-revenue company, raising cheap capital when the market is euphoric and avoiding buybacks/dividends is exactly correct. Credit where due.

The marquee deployment, however, is deeply troubling. The largest single use of capital in 2025 was the ~$507M Milestone Contribution 1 paid/accrued to ENTRA1 — against a non-binding MOU, to an entity that retains sole discretion whether to buy NuScale’s product, and that caps the price NuScale can charge for its own modules. Over the full pipeline the PMA implies ~$3.4B owed to ENTRA1 (MC1 ~$507M already; MC2 ~$1.15B; MC3 ~$1.69B), payable ahead of any module revenue. This is capital allocated to acquire the appearance of demand rather than to build product, lower cost, or return capital.

The governance red flag at the center of it. ENTRA1 Energy is ~75% controlled by Wadie Joseph Habboush (Habboush Group), who has led the Fluor-Habboush International Bermuda joint venture (50/50, since 2012). Fluor is NuScale’s controlling shareholder and EPC partner, and the former employer of both CEO John Hopkins (Fluor, 1989–2012) and Chairman Alan Boeckmann (former Fluor CEO). Yet NuScale discloses only Fluor as a related party — not ENTRA1 — in both the FY2025 10-K (Note 16) and the 2026 proxy, despite paying ENTRA1 ~$507M and granting it exclusive worldwide commercialization rights. The economic substance is a related-party-style arrangement that has escaped related-party review. A securities-fraud complaint, Truedson v. NuScale (D. Or., No. 3:26-cv-00328, Feb-2026), names Hopkins, the CFO, and Fluor as controlling persons and alleges ENTRA1 was misrepresented as a credible commercialization partner (described in the complaint as a roughly three-year-old entity that had “never built, financed or operated anything”). This is, at minimum, an unresolved governance overhang of the first order.

Incentives reward activity, not value. The 2025 compensation scorecard pays for SDA milestones, work-scope execution, and headcount — and, tellingly, weights module-purchase commitments and cash revenue at 0%. There is no ROIC, no per-share, no TSR metric (textbook Marathon mis-incentive: management is paid to advance the project, not to create per-share value or sell product economically). NEOs earned ~80.5% of target for FY2025; CEO Hopkins’s summary-compensation-table pay was ~$4.08M (compensation-actually-paid, reflecting the stock collapse, was a token ~$33K). There are zero open-market (code P) insider purchases by anyone.

And the most-informed holder is leaving. Fluor converted its Class B and sold ~$1.7B+, from ~50% to ~4% in roughly six months (Feb–Apr 2026), liquidating all the way down from $19 to $11.63. Whatever the stated rationale (portfolio focus), the founding operating partner monetizing nearly its entire stake into a falling market — while the company routes half a billion dollars to that partner’s affiliate’s agent — is the single loudest capital-markets signal on the name.

Verdict: mixed-to-negative and conflicted. One shrewd financing decision (the ATM) is overwhelmed by a marquee deployment that transfers cash to an undisclosed-related-party agent against non-binding orders, an incentive structure that ignores per-share value, and a controlling shareholder racing for the exit. Management has allocated capital to survive and narrate, not to build value per share.

8. Changes and Headwinds — Last Two Years

A dated two-year timeline, and what it means for the thesis:

• Nov 2023 — UAMPS/CFPP cancellation. The anchor customer walked as target LCOE escalated to ~$89/MWh. Thesis-defining negative: proved an NRC-approved design ≠ a financeable plant. Weakens.
• Jan 2024 — ~28% workforce reduction. Cost-cut response to the cancellation; rational but confirms the demand vacuum. Weakens.
• Jul 2024 — RoPower (Romania) licensing agreement. First concrete post-CFPP customer engagement. Strengthens (modestly).
• May 2025 — 77 MWe six-module SDA approved by NRC. The strongest positive of the period and the company’s single best asset; widens the regulatory lead. Strengthens.
• 2025 — AI/data-center power pivot + Trump nuclear EOs. Re-rated the stock and the narrative; mostly market/sentiment, little contracted substance. Neutral-to-positive (narrative).
• Aug 2025 — ENTRA1 Partnership Milestones Agreement. Outsourced sales to a single exclusive agent; created the ~$507M (and ~$3.4B contingent) cash drain to a Fluor-orbit, undisclosed-related-party entity. Weakens (materially).
• Sep–Oct 2025 — ENTRA1/TVA non-binding 72-module MOU; ATM-into-spike; $53.43 peak. Raised ~$1.3B cheaply (good) on a non-binding headline that cost $507M (bad). Mixed.
• Feb 2026 — RoPower (Nuclearelectrica) investment-decision approval. Most concrete customer step to date; still pre-EPC/FID, no firm module order. Strengthens (modestly).
• Feb–Apr 2026 — Fluor sell-down ~50%→~4% (~$1.7B), down to $11.63; Truedson securities suit filed. Weakens (materially).
• 2026 (ongoing) — minor execution news (Mirion/Paragon final-design contract; Virginia training center; analyst reinstatement at Neutral). Neutral.

Verdict: net weakening. The two genuine positives (the May-2025 SDA and the Feb-2026 RoPower FID) are real but incremental and still short of a binding order. They are outweighed by the cluster of negatives: a cancelled flagship, a pivot to a conflicted commercialization wrapper that drains cash, a controlling-shareholder exit, and live securities litigation. The two-year arc is the swap of a real-but-uneconomic customer for a non-binding pipeline owned by an agent — while the most-informed insider left.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss assessment. A total loss is unlikely in the near term: ~$1.0B of net cash, positive tangible book, no debt, and a saleable NRC credential put a floor under the equity. The realistic downside is not zero but a prolonged de-rating toward cash-plus-credential value as the pipeline fails to convert and milestones drain the balance sheet. The tail risk that would threaten the floor is an adverse Truedson/fraud outcome or an open-ended milestone-cash commitment with no offsetting orders.

10. Valuation Discussion (Embedded Expectations)

No conventional multiple applies. With negative earnings, EBITDA, and FCF, the only computable multiple is EV/sales, and it is extreme. Fully-diluted shares ≈ 340–345M; at $11.74, market cap ≈ $3.8B (Class A) to ~$4.0B (fully diluted). Less ~$1.0B net cash → EV ≈ $2.5–3.0B → ~80–97x trailing sales on $31.5M of declining revenue. The AZI own-history valuation read is split and must be read with care: P/S 89.8x sits at the 73rd percentile of its own (short, bubble-distorted) history, while P/B of 1.65x sits at the 3.8th percentile — the latter only “cheap” because the share price has collapsed toward a cash-heavy book ($7.13 BVPS, ~$3 of it net cash). The composite (38th percentile) blends these and is not meaningful for a pre-revenue name.

The cash floor is the only hard anchor. Net cash of ~$3/share (and book of $7.13, mostly cash) is the defensible valuation floor — but it is an eroding floor: the ~$200M/year underlying burn and episodic ENTRA1 milestone transfers consume it over time. So the floor is perhaps ~$3–5/share today and declining unless the burn is funded by new equity (dilutive) or — the bull case — by actual orders.

Embedded-expectations / reverse-DCF. Strip out the ~$3 of net cash and the market is paying ~$8.50/share, or ~$2.5–3.0B of enterprise value, for the commercialization option. What must be true to justify that? Assume NuScale captures ~$200M of revenue per 77 MWe module (its equipment-plus-license slice of a ~$1.5B/plant build) at ~35% gross margin ≈ ~$70M gross profit per module. With first delivery “2031 or beyond,” a revenue stream centered ~2035 and discounted at ~15% (PV factor ~0.28), generating ~$2.5B of present value requires roughly $8.8B of undiscounted gross profit ≈ 120–130 delivered, margin-positive modules (~10 GW, ~20 VOYGR-6 plants). That is more than the entire non-binding “up to 72 module” TVA pipeline, fully converted, built, and delivered on assumed economics — and before accounting for ENTRA1’s 15/35/50% take and module-price cap, which would push the required module count materially higher. In short: the market is underwriting flawless conversion of essentially the whole non-binding pipeline (and then some) by a company with zero binding orders in 18 years and a flagship that cancelled over cost.

Scenario analysis (3–5 year; illustrative, not forecasts).

The payoff is convex in the tail, but the distribution is more bear-weighted than OKLO’s because SMR uniquely carries a lived cancellation precedent (CFPP), an exiting controlling shareholder, and the value-extracting ENTRA1/PMA structure.

Comp set. No operating-multiple comparison is meaningful. OKLO (~$7.5B EV) is the closest pre-revenue twin — SMR is cheaper and better credentialed but worse scarred and governed. BWXT, Constellation (CEG), GE Vernova (GEV) anchor the “real nuclear economics / cheap-fast-gas substitution” contrast. Bloom Energy (BE), Centrus (LEU) are thematic/factor twins; Cameco (CCJ) is the fuel-cycle demand anchor. The honest comparison is not to a peer multiple but to a venture-stage option: NuScale is a binary bet on a single technology platform’s commercialization, priced in the public market.

Verdict (no recommendation): no margin of safety above the cash floor. The ~$2.5–3.0B enterprise value is pure, distant, contingent option value on a commercialization the company has not achieved in 18 years, shared with an agent, and contradicted by its own flagship cancellation. The defensible value is closer to the eroding cash-plus-credential floor than to the current price.

11. Variant Perception

Consensus belief. NuScale is the de-risked SMR leader — the only NRC-approved design — riding the unstoppable AI-electricity-demand megatrend, sitting on a fortress balance sheet, and the ~78% drawdown is a gift entry into the clean-firm-power supercycle.

Strongest bull case. Firm, clean, 24/7 power demand from AI/data centers is effectively inexhaustible and policy-supercharged; NuScale’s NRC head-start (two SDAs, no competitor even applied) is a multi-year moat in the one dimension that gates the industry; ~$1.0B net cash funds the wait to ~2031; ENTRA1 outsources sales to an entity that is the demand aggregator; and if even a handful of VOYGR plants get built, the equity is worth multiples of today. In a theme this large, optionality on the credentialed leader is cheap at a $2.5B EV.

Strongest bear case (the stronger argument on current evidence). Eighteen years, ~$2B spent, zero binding module orders; the flagship (UAMPS) cancelled over cost; sales outsourced to ENTRA1, to which NuScale pays ~$507M (and ~$3.4B contingent) against non-binding MOUs, with ENTRA1 — an undisclosed related party at the center of a securities suit — holding sole discretion and a price cap; Fluor, the controlling shareholder and most-informed holder, dumped ~$1.7B from ~50% to ~4% selling all the way down; first delivery “2031 or beyond”; a crowded subsidized field where BWRX-300 is already under construction; and compensation that pays for activity, not orders or per-share value. The valuation embeds more delivered modules than the entire non-binding pipeline. This is a narrative priced as a near-certainty.

The 3–5 assumptions that matter most:

1. MOU → binding-order conversion. Does any non-binding pipeline item (TVA, data-center) become a firm, economically-priced module order from a creditworthy counterparty? Falsify bull: RoPower or TVA lapses without a firm order. Falsify bear: a binding, economic PPA/equipment order is signed.
2. Plant financeability at an economic LCOE. Can an Nth-of-a-kind VOYGR be built at a cost that clears against gas? Falsify bull: a second cost-driven cancellation. Falsify bear: RoPower reaches FID with firm offtake at a competitive LCOE.
3. ENTRA1 value extraction / governance. Does the PMA bleed the economics and does Truedson surface worse facts? Falsify bull: an MC2 cash transfer or adverse legal finding. Falsify bear: ENTRA1 delivers a real, signed customer and the related-party concern proves immaterial.
4. Durability of the NRC head-start. Does the lead hold as BWRX-300 and others clear the NRC? Falsify bull: a competitor approval + deployment. Falsify bear: NuScale converts its lead into the first Western SMR built.
5. Timing vs substitutes. Does cheap gas (and existing nuclear uprates/restarts) absorb the AI demand before SMRs can scale? Falsify bull: gas/existing nuclear soak up the load through the early 2030s. Falsify bear: utilities commit to SMRs for firm clean capacity this decade.

The factor read (positioning evidence). FactorsToday/AZI place SMR as a high-beta (2.3–3.0), high-specific-vol (~77%) thematic vehicle with no value, quality, momentum, or growth loading — its variance is explained by the market, a utilities-sector tilt, and custom “speculative theme” baskets, and it twins the clean-energy/nuclear ETFs (PBW, BE, LEU, APLD, QCLN, ARKQ). With y1 −70%, m6 −44%, and a −87.5% lifetime max drawdown, the tape is a crowded thematic trade now in a falling-knife drawdown — the market is actively repricing the option down, not bottom-fishing value or chasing momentum. That is consistent with the bear’s read: consensus is offsides to the extent it treats a re-rating drawdown as a value entry rather than as a thematic option being marked toward its eroding floor.

12. Fact vs. Interpretation Table

13. Open Questions

1. ENTRA1 ownership/control and related-party status. What is the precise ownership and governance of ENTRA1, and why is it not a disclosed related party given the Fluor-Habboush JV, the ex-Fluor CEO/Chairman, and the ~$507M paid? (Central to the Truedson suit.)
2. Milestone Contribution 2/3 timing and size. When do MC2 (~$16M/module) and MC3 obligations trigger, and what is the realistic cash drain over the next 3 years against the ~$1.0B balance?
3. RoPower path to a firm order. Will the Romania project convert the Feb-2026 investment decision into a binding module purchase order and an economic LCOE — and on what timeline?
4. Real underlying burn at scale. As pre-construction activity ramps, does the ~$200M/year recurring burn hold, or does it climb toward the FY2025 elevated level?
5. Truedson outcome. What facts does discovery surface about ENTRA1, and what is the litigation/settlement exposure?
6. Fluor’s residual ~4% and EPC commitment. Does Fluor exit entirely, and does its reduced stake change its commitment as EPC partner?
7. Competitive timing. When does BWRX-300 (or another) achieve its NRC approval and first deployment, and how does that reprice NuScale’s head start?

14. What Must Be True

Bull case — what must be true, and its falsification test. NuScale’s NRC-approved design must convert, within ~3–5 years, into binding, economically-priced module orders from creditworthy counterparties (RoPower FID with firm offtake, and/or a real data-center PPA), demonstrating that the credential produces financeable demand; the ENTRA1 structure must deliver signed customers without bleeding the economics or surfacing governance damage; and the ~$1.0B balance must fund the wait without ruinous dilution. Falsification test: if, by year-end 2027, NuScale still has no binding module purchase order and RoPower has not reached a financed FID — or if a Milestone Contribution 2 cash transfer occurs with no offsetting firm order — the bull thesis is broken.

Bear case — what must be true, and its falsification test. The SMR economics must remain uncompetitive against cheap gas and existing nuclear through the early 2030s; the non-binding pipeline must fail to convert to firm economic orders; the ENTRA1 structure must continue extracting cash against MOUs; and the equity must de-rate toward its eroding cash-plus-credential floor as the thematic bid fades. Falsification test: if NuScale signs a binding, economically-priced module order from a creditworthy buyer and RoPower reaches a financed FID at a competitive LCOE, the bear thesis (no financeable demand) is falsified and the option re-rates.

15. Source Appendix

See the Source Appendix below for the full citation list. Primary sources: NuScale FY2025 Form 10-K (filed 2026-02-26), Q1-2026 Form 10-Q, FY2024 Form 10-K, 2026 DEF 14A proxy, Form 4 insider filings (Fluor and officers, Nov-2025–Apr-2026), and the 8-K material-event corpus (SEC EDGAR, CIK 0001822966). Quantitative data: ROIC.ai (financials/EV/ratios), AZI price CSV and valuation-percentile feed, FactorsToday factor model. Industry/peer context: independent analyses of peers OKLO, Bloom Energy, and Cameco; the Truedson v. NuScale complaint (D. Or., No. 3:26-cv-00328); NEI Magazine and trade-press coverage of the ENTRA1 litigation; NRC public records on the 77 MWe SDA. Every non-obvious fact in the memo traces to a dated entry in the research log.

APPENDIX A — Standard Diligence Questionnaire — NuScale Power Corporation (NYSE: SMR)

Supplemental to the research memo. Answers grounded in the research log; Fact / Interpretation / Assumption labeled where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions: (1) Is there any binding order, or only MOUs? (Answer: only MOUs/non-binding agreements; zero binding module orders.) (2) What exactly is ENTRA1, and why is NuScale paying it ~$507M against a non-binding TVA agreement? (3) Why is Fluor — the controlling shareholder — selling its entire stake? (4) What is the real cash burn once the ENTRA1 milestone noise is stripped out, and how long is the runway? (5) Is the May-2025 NRC SDA a durable moat or a depreciating head start as BWRX-300 advances? (6) How do SMR economics clear against cheap gas given the UAMPS cancellation over cost? These are precisely the memo’s focal points.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Neither — there are no earnings. The company is pre-commercial with persistent, large operating losses (−$689.6M GAAP FY2025, ~−$200M underlying ex-milestone). (Fact.)

Driven by the external environment or internal actions? Both, adversely: external (UAMPS cost escalation, competitive/subsidy dynamics) and internal (the ENTRA1 PMA structure that booked the ~$507M charge). (Interpretation.)

How stable are revenues? Unstable and negligible: $2.9M → $11.8M → $22.8M → $37.0M → $31.5M (2021–25), composed of development reimbursement/licensing, not product sales; FY2025 declined year-on-year. (Fact.)

Outlook for products/services? Binary and distant. First NPM delivery is “2031 or beyond” per the 10-K; the outlook depends entirely on converting non-binding pipeline (RoPower, TVA/data-center) into firm, economic orders. (Fact/Interpretation.)

How big will this market be — growing, shrinking, domestic or international? Potentially very large and growing if the AI-power/nuclear-renaissance demand and SMR cost-down both materialize (global, with RoPower international and TVA domestic) — but unproven commercially and contested by cheaper gas. (Interpretation.)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More: a wall of subsidized SMR/advanced-reactor entrants (GE Hitachi BWRX-300 already building, TerraPower, X-energy, Westinghouse, Holtec, Rolls-Royce, Oklo). (Fact/Interpretation.)

How profitable is the business (ROIC, ROE)? Deeply negative; pre-commercial. No meaningful ROIC/ROE. (Fact.)

How profitable is the industry — competitors, barriers to entry? No commercial Western SMR exists; the profit pool is prospective and policy-dependent. The barrier to entry (NRC licensing, multi-year/multi-hundred-million-dollar) is high and is NuScale’s one real edge — but it is a credential, not customer captivity. (Fact/Interpretation.)

Can the business be easily understood? The technology and model are understandable; the economics and governance (ENTRA1 PMA, up-C, TRA, milestone accounting) are deliberately complex and obscure the cash flows. (Interpretation.)

Can it be undermined by foreign low-cost labor? Not the core risk; the competitive threat is other (often subsidized) reactor technologies and cheap gas, plus Chinese/Russian SMRs in non-Western markets. (Interpretation.)

Do brands matter? Marginally — “first NRC-approved” is a credibility marker — but utility procurement turns on LCOE, financeability, and delivery risk, not brand. (Interpretation.)

Nature of competition? Technology/regulatory race to first economic deployment; winner-take-most dynamics around the first financeable Western SMR plant. (Interpretation.)

Customers’ switching costs? N/A — there are no commercial customers yet; once a plant is built, life-of-plant service lock-in could emerge, but that is prospective. (Interpretation.)

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The NRC design approvals (internally generated, largely expensed) are the key off-balance-sheet value; the $507M MC1 was expensed with no asset recognized. (Fact/Interpretation.)

Off-balance-sheet liabilities? The PMA milestone obligations to ENTRA1 (MC2 ~$1.15B, MC3 ~$1.69B across the full pipeline) are contingent and not yet on the balance sheet; the Tax Receivable Agreement (85% of tax benefits to legacy equityholders) carries no recorded liability today (no taxable income). (Fact.)

How conservative is the accounting? The ENTRA1 milestone is expensed immediately under ASC 606 (conservative in that direction), but the broader presentation — booking a $507M cash transfer to an agent as the trigger of headline losses, the up-C/NCI complexity — obscures the underlying run-rate. (Interpretation.)

How CapEx-hungry is the business? Not at all for NuScale — it is capital-light (capex ~$0.5M); the plants are enormously capital-hungry, but that capital sits with utilities/developers, not NuScale. (Fact.)

Capital Allocation & Management

How much FCF does the business generate; how is it used; philosophy? None — FCF is materially negative (≈ −$200M/yr underlying). The company consumes capital (R&D, the ENTRA1 milestone, RoPower) funded by equity issuance. (Fact.)

Significant acquisitions recently? None. (Fact.)

Buying back shares? No (correctly, for a pre-revenue burner). (Fact.)

Issuing large amounts of new shares to insiders? Routine SBC/RSU grants (~$19M SBC); the large share-count increase is ATM equity ($1.3B) + Fluor’s Class B→A conversion, not insider grants. (Fact.)

Compensation policy of directors/management? Metrics reward SDA milestones, execution, and headcount — no ROIC/per-share/TSR; module orders and cash revenue weighted 0%. CEO Hopkins SCT comp ~$4.08M (FY2025). A Marathon-style mis-incentive: paid to advance the project, not to create per-share value. (Fact/Interpretation.)

Motivations of management? Project advancement and survival; the conflicted ENTRA1 structure (Fluor-orbit) and the controlling-shareholder exit raise alignment questions. (Interpretation.)

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — U.S. C-corp common stock (1099), NYSE-listed, in an “Up-C” structure (Class A public; Class B vote-only LLC units). (Fact.)

Dividend policy? None (appropriate). (Fact.)

How profitable is the business? Unprofitable; pre-commercial. (Fact.)

Is net income diverging from cash from operations? Yes — FY2025 net loss (−$355.8M attributable / −$664M including NCI) vs OCF −$459.6M, distorted by the ENTRA1 milestone cash and a lumpy AP swing; the clean Q1-2026 run-rate (~−$50M/qtr) is the honest read. (Fact.)

Risks & Downside

What factors would cause the stock to decline? Failure to sign binding orders; RoPower slippage; an MC2 cash drain; adverse Truedson developments; further Fluor selling; a competitor NRC approval/deployment; a thematic/factor unwind (beta ~2.3–3.0). (Interpretation.)

Risk of a catastrophic loss? Elevated operationally (commercialization may never reach economic scale) but mitigated financially by ~$1.0B net cash, positive tangible book, and a saleable credential. (Interpretation.)

Chance of a total loss? Low-to-moderate near term given net cash and no debt; the realistic downside is a de-rate toward an eroding cash-plus-credential floor, not zero — absent a long unfunded burn or an adverse fraud finding. (Interpretation.)

Recent News & Events

Has the business environment changed recently? Yes, materially: the Aug-2025 ENTRA1 PMA pivot, the Oct-2025 thematic peak and ATM raise, the Feb-2026 RoPower investment decision, the Feb–Apr-2026 Fluor sell-down, and the Truedson securities suit. (Fact.)

Significant acquisitions? None. (Fact.)

Change in accounting policies? The ENTRA1 milestone (ASC 606 “consideration paid to a customer”) is the notable new treatment. (Fact.)

Recent changes — new markets, facilities, management? New: ENTRA1 commercialization model; RoPower (Romania) progress; Virginia training center; Mirion/Paragon final-design contract; ~28% workforce reduction (Jan-2024) following the UAMPS cancellation. (Fact.)

APPENDIX B — Source Appendix — NuScale Power Corporation (NYSE: SMR)

Report date 2026-06-20. Primary sources first. Every non-obvious memo claim traces to a dated research-log entry and to a source below.

Primary — SEC filings (EDGAR, CIK 0001822966)

• NuScale FY2025 Form 10-K (filed 2026-02-26; period 2025-12-31). Business description, NPM/VOYGR designs, NRC SDA status, competition, risk factors, MD&A, Notes 9/16/17 (ENTRA1 Partnership Milestones Agreement, Milestone Contribution 1 ~$507.4M, related parties, TRA), revenue $31.5M, operating loss −$689.6M. https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001822966
• NuScale Q1-2026 Form 10-Q (period 2026-03-31). Clean run-rate (operating loss −$57.5M, G&A $24.8M), balance sheet (~$1.0B liquidity, net cash, no debt), AP drain (~$264M) paying accrued MC1, share count 323.7M Class A.
• NuScale FY2024 Form 10-K (filed 2025-03-03; period 2024-12-31). Trend baseline; warrant redemption (Dec-2024); FY2024 warrant fair-value item (+$223.0M).
• NuScale 2026 DEF 14A proxy. Compensation structure and metrics (SDA milestone/execution/headcount; module orders & cash revenue weighted 0%; no ROIC/per-share/TSR), CEO Hopkins SCT ~$4.08M, say-on-pay, board composition, Fluor relationship, up-C/TRA, share authorization (662M) and Nov-2025 Exchange Agreement.
• Form 4 insider filings (Nov-2025–Apr-2026). Fluor Class B→A conversion (110.9M, Nov-2025) and sales: 71.0M @ $19.05 (~$1.35B, 2026-02-13), 13.5M @ $12.07 (2026-04-09), 12.9M @ $11.63 (2026-04-15); residual ~13.5M (~4%). Officer activity: option-exercise/sell-to-cover/RSU only; zero code-P open-market purchases.
• 8-K material-event corpus (2024–2026). NRC 77 MWe SDA approval (May-2025); ENTRA1 PMA (Aug-2025); ENTRA1/TVA non-binding MOU (Sep-2025); RoPower/Nuclearelectrica investment decision (Feb-2026); ATM equity program; earnings releases.

Primary — regulatory

• U.S. NRC public records: NuScale 50 MWe design Final Safety Evaluation Report (2020); 77 MWe six-module Standard Design Approval (May 2025). https://www.nrc.gov
• Truedson v. NuScale Power Corp., U.S. District Court, District of Oregon, No. 3:26-cv-00328 (filed Feb-2026). Securities-fraud complaint naming Hopkins, the CFO, and Fluor; allegations regarding ENTRA1 misrepresentation.

Quantitative data services

• ROIC.ai — income statement, balance sheet, cash flow, enterprise value, ratios (FY2021–25 annual; Q2-2025–Q1-2026 quarterly). Reconciled to the 10-K/10-Q (primary).
• AZI (azitrading.com) — daily price/OHLCV CSV (full history; de-SPAC ~$10 → $1.92 low → $53.43 high → $11.74); valuation-index own-history percentiles (P/S 89.8x / 73rd pctile; P/B 1.65x / 3.8th pctile; composite 38th); news feed.
• FactorsToday (factorstoday.com) — factor loadings (Market +2.28, Utilities +1.11, LowVol −1.85, Liquidity −1.77, no value/quality/momentum/growth), leaderboard (y1 −70%, m6 −44%, maxDD −87.5%), beta 2.3–3.0, specific vol ~77%, related stocks (PBW/APLD/HYDR/BE/LEU/QCLN/ARKQ).

Industry / peer / trade-press context

• Peer/industry cross-reads: OKLO (2026-06-13, closest pre-revenue SMR peer), BE (Bloom Energy, 2026-06-10, factor twin), CCJ (Cameco, 2026-06-20, nuclear fuel-cycle demand anchor).
• NEI Magazine — coverage of the ENTRA1/NuScale litigation. https://www.neimagazine.com/news/nuscale-faces-entra1-lawsuits/
• Habboush Group / Fluor-Habboush International JV public records (ENTRA1 ownership ~75% W.J. Habboush; 50/50 Bermuda JV since 2012).
• Competitive landscape: public disclosures of GE Hitachi BWRX-300 (Darlington construction), TerraPower Natrium (Kemmerer), X-energy Xe-100, Westinghouse AP300, Holtec SMR-300, Rolls-Royce SMR, Oklo Aurora.
• Sector framework (dated third-party industry primers — used for value-chain/structure only, not current data): published electric-utilities/power industry primers (2006–2009 vintage).
• Cost reference points: UAMPS/CFPP target LCOE escalation (~$58→$89/MWh, cancelled Nov-2023); Vogtle 3&4 (~$16,000/kW) as the new-nuclear cost benchmark.

Management commentary (earnings calls, investor presentations) was treated as hypothesis and validated against filings, financials, and external evidence per the research standard. No internal/ownership inference is implied by any cited artifact.