Oklo Inc (OKLO): Comprehensive Equity Analysis

1. The Big Insight

Oklo's most valuable asset isn't its reactor—it's its fuel recycling thesis, and the market hasn't priced in how far that thesis is from validation.

Oklo's entire competitive narrative rests on a vertical integration strategy where recycled spent nuclear fuel (96,000 metric tons of U.S. stockpiles representing 200 GW of potential capacity) slashes long-term fuel costs 70-80% versus competitors locked into once-through fuel cycles (Reports 1, 3). This is what theoretically justifies the build-own-operate model, the $9 billion market cap on zero revenue, and the claim of eventual $40/MWh LCOE. Yet the Tennessee recycling facility won't operate until the 2030s (Report 3), the fuel fabrication facility at INL only received preliminary safety approval in December 2025 (Report 1), and the illustrative economics published in 2023 assumed HALEU costs of $7,000/kg when current reality is $35,000+/kg (Report 1). The fuel recycling moat is real in concept—no competitor possesses it—but it exists on a timeline that stretches well beyond the Aurora-INL demonstration reactor, which itself hasn't received NRC licensing. The market is pricing in the destination without adequately discounting the distance.

---

2. Business Model: What Is Known vs. Speculated

What is firmly established

Oklo operates a build-own-operate (BOO) model: it will design, build, own, and operate Aurora fast fission reactors, selling electricity and heat via 20-year PPAs directly to customers without requiring customer capital expenditure (Report 1). This is confirmed in SEC filings and contrasts with NuScale's licensing model (Report 3). The company has $1.18 billion in liquidity as of Q3 2025, with guided annual burn of $65-80 million, yielding approximately 15-17 months of runway at midpoint before needing additional capital—though a $3.5 billion shelf registration provides the mechanism for continued dilution (Report 5).

The announced pipeline totals ~18 GW in non-binding LOIs and master agreements (Report 1). Key disclosed customers include:
- Switch: 12 GW Master Power Agreement through 2044 (non-binding framework) (Report 1)
- Meta: 1.2 GW Ohio campus with prepayment funding Phase 1 development, first power ~2030, full capacity by 2034 (Reports 1, 4)
- Equinix: 500 MW pre-agreement with $25 million non-refundable prepayment (Report 1)
- Diamondback Energy: 50 MW 20-year PPA LOI for Permian Basin operations (Report 1)
- Prometheus/Wyoming Hyperscale: 100 MW data center (Report 1)

Ground was broken September 22, 2025 at Idaho National Laboratory for the first 75 MWe Aurora unit under the DOE Reactor Pilot Program, with a binding Siemens Energy contract for power conversion signed November 2025 (Report 1).

What remains speculative

No binding PPAs exist beyond frameworks. The entire 18 GW pipeline consists of non-binding LOIs and master agreements that require milestone conversions (Report 1). The Meta deal—the strongest commercial validation—involves prepayments for development but targets first power in 2030, four years away (Report 4). Report 1 notes that "pipeline fragility (all non-binding) means entrants need big-tech anchors to fund first-of-a-kind builds."

No confirmed OpenAI commercial relationship exists despite Sam Altman's central role. His April 2025 resignation as chairman was framed as enabling such partnerships, but no deal has materialized (Report 4). The AI power demand narrative is directionally correct—data centers may consume 9-12% of U.S. power by 2030 per IEA/Goldman estimates (Report 4)—but Oklo captures none of this revenue today and won't before 2027-2028 at the earliest.

Illustrative unit economics are unverified. The 2023 investor presentation projects $1,600/kWe capex, $105/MWh revenue, 76% cash margins, and $40/MWh LCOE with tax credits (Report 1). These figures assume HALEU at $7,000/kg when spot prices exceed $35,000/kg, a gap that materially inflates projected LCOE (Report 1). No independent audit of these economics exists.

The step from LOI to operating reactor has never been completed. Oklo must navigate: MOU → LOI → site evaluation → PPA → NRC licensing → construction → commissioning → operations. The company is between steps 1-3 on all projects (Reports 1, 2).

---

3. Competitive Differentiation: Genuine Moat vs. Closing Gaps

Where Oklo holds structural advantages

Fuel recycling is unique and defensible. No competitor possesses Oklo's capability to convert spent nuclear fuel into usable reactor fuel. The company has secured 5 metric tons of recycled EBR-II waste from DOE—the only company with fuel for its first commercial unit (Report 1). The March 2026 Centrus JV for HALEU deconversion co-located with the Ohio campus creates an integrated fuel supply chain that competitors must replicate from scratch (Report 2). Report 3 confirms this "slashes long-term fuel costs 70-80%" and creates "endless supply" from America's 96,000 MT stockpile.

Compact form factor suits behind-the-meter data center deployment. At 15-75 MWe per unit, Aurora matches individual data center "halls" without grid interconnection, whereas NuScale's 77 MWe modules scale to 462-924 MWe utility plants and TerraPower's Natrium targets 345-500 MWe (Report 3). For hyperscalers wanting co-located power avoiding grid queue delays, Oklo's size is right.

EBR-II heritage provides a licensing data moat. 400+ reactor-years of global operational data for similar sodium-cooled fast reactor technology shortens licensing timelines to 24-36 months versus 44+ months for genuinely novel designs (Report 1).

Where competitors hold decisive advantages

TerraPower is Oklo's greatest competitive threat (Report 3's explicit conclusion). TerraPower secured the first NRC commercial construction permit in over a decade on March 4, 2026, for its 345 MWe Natrium reactor in Wyoming—full nuclear-related construction is now authorized (Report 3). TerraPower also uses sodium-cooled fast reactor technology, operates at utility scale with integrated molten salt storage (500 MWe peak), has $1.46 billion+ in funding including Gates and NVIDIA, and signed a Meta deal for up to 8 units/2.8 GW (Report 3). Critically, TerraPower's construction permit puts it 3-5 years ahead of Oklo on the same fundamental technology pathway.

NuScale remains the only NRC-certified SMR design, with the uprated 77 MWe module approved in 2025 and Romania's RoPower reaching final investment decision for 6 modules in 2026 (Report 3). While NuScale targets a different market segment (utilities, grid-scale), its regulatory precedent is unmatched.

X-energy achieved a fuel supply chain milestone Oklo hasn't: NRC licensed two TRISO fuel fabrication facilities in February 2026—the first advanced reactor fuel factory approved—backed by $1.8 billion in funding and Amazon's 5 GW commitment (Report 3). X-energy's fuel self-sufficiency contrasts with Oklo's HALEU dependency.

The regulatory gap is widening, not narrowing. As of March 2026, Oklo has not submitted its revised Combined License Application to the NRC (Report 2). TerraPower has a construction permit. X-energy has fuel fab licenses. NuScale has design certification. Oklo has pre-application topical reports under review—a meaningful step, but categorically different from the milestones competitors have achieved.

---

4. Regulatory and Fuel Supply: Identifying the Binding Constraint

NRC Licensing Risk

The 2022 denial was unusually harsh. The NRC found Oklo repeatedly failed to provide sufficient information on maximum credible accidents, safety classification of structures/systems/components, and related methodologies across three rounds of RAIs (Report 2). Former NRC Chair Allison Macfarlane called the process "extremely frustrating," and a senior official labeled Oklo "the worst applicant the NRC has ever had" (Report 8). This was not a technicality—it reflected fundamental gaps in safety documentation.

Oklo has responded with a phased strategy: topical reports on principal design criteria (accepted September 2025 for accelerated review), quality assurance, and performance-based licensing (Report 2). Phase 1 readiness for the INL site passed in July 2025 with "no significant gaps" (Report 2). However, no revised COLA has been docketed as of March 2026, despite being planned for late 2025 (Report 2). The original Aurora docket is closed with no new documents post-2022 (Report 2).

The DOE Reactor Pilot Program provides a partial bypass: Oklo's INL demo can pursue criticality (targeted mid-2026) under DOE authorization rather than full NRC licensing (Reports 1, 2). But commercial operations still require NRC licensing, and Report 2 estimates 36-60+ months for FOAK fast fission reactor COLA reviews even with reforms.

HALEU Fuel Supply

Centrus Energy—the sole U.S. HALEU producer—operates at approximately 900 kg/year, with plans to scale to 12 MT/year by 2029 via $2.7 billion in DOE funding (Report 2). Oklo needs 7 MT for initial loads (Report 2). The March 2026 Oklo-Centrus JV for deconversion (converting enriched uranium hexafluoride to metal for fast reactors) addresses a key bottleneck, but the domestic supply chain requires 5-10 fabricators to serve the broader advanced reactor market (Report 2).

Russia supplied 100% of commercial HALEU before sanctions, and the ban takes full effect in 2028 (Report 2). DOE projections show an 11.9 million SWU shortfall by 2030 absent a 10x production ramp (Report 2). Oklo's 5 MT of recycled EBR-II fuel provides a bridge for the first unit, but fleet-scale deployment depends on a supply chain that does not yet exist at commercial scale.

The Binding Constraint

NRC licensing is the binding near-term constraint; HALEU becomes the binding constraint at fleet scale. Oklo cannot generate revenue until Aurora-INL receives operating authorization (DOE pilot or NRC COL). The DOE pathway may accelerate the demo to 2027-2028, but every subsequent commercial unit requires NRC approval on a 3-5 year timeline (Report 2). At fleet scale, even with NRC licensing, HALEU production must increase roughly 10x from current levels to serve Oklo's pipeline—and every other advanced reactor developer simultaneously (Report 2). The Centrus JV helps but cannot single-handedly close this gap.

Reports 2 and 8 are aligned that the FOAK regulatory pathway remains the primary risk to the 2027-2028 timeline, while Report 2 identifies HALEU scale-up as the critical risk to the 2030s fleet buildout.

---

5. Valuation Framework

What the Market Is Pricing In

Oklo trades at approximately $58/share, yielding a $9.1 billion market cap and $8.2 billion enterprise value on zero trailing revenue (Report 6). This implies an infinite current EV/Revenue multiple. Against analyst consensus of $16 million in 2027 revenue, the forward EV/Revenue exceeds 500x (Report 6).

For comparison (Report 6):
| Company | Market Cap | TTM Revenue | EV/Revenue |
|---------|-----------|-------------|------------|
| Oklo | $9.1B | $0 | ∞ |
| NuScale | $3.9B | $31.5M | 85-114x |
| Cameco | ~$50B | $3.5B | 13-19x |
| UEC | $6.3B | $50M | 116x |

To justify the current $9 billion EV via DCF, Report 6 calculates that Oklo would need approximately $21 billion in revenue by 2038 at 60% margins, requiring successful deployment of 14 GW—essentially the entire current non-binding pipeline—with flawless execution. Seeking Alpha DCFs place fair value at $9-11/share in a base case, representing ~85% downside (Report 6).

Bull Case ($150-175/share)

Key assumptions: NRC licensing achieved 2027, first Aurora commercial by 2028, 1 GW deployed by 2030, fuel recycling drives 60%+ margins, AI data center demand absorbs 14+ GW by 2040. Analyst targets from Wedbush ($150), Texas Capital ($138), and BofA ($127) embed this scenario (Report 6). The thesis requires: DOE pilot success → NRC license → serial factory production → HALEU supply chain maturation → hyperscaler PPA conversions. The plutonium bridge fuel strategy could accelerate pre-HALEU deployment (Report 6).

Bear Case ($9-14/share)

Key assumptions: NRC delays mirror 2022 (additional RAI cycles adding 2-3 years), HALEU shortages push first commercial unit to 2029+, FOAK costs hit historical 100%+ overruns (Report 8 documents average nuclear FOAK overruns of 102.5%), data centers lock in gas/renewables/restarts before SMRs scale, continued dilution from $3.5 billion shelf erodes equity value. Report 8 notes that hyperscaler prepayments may convert to debt obligations if Oklo misses July 2026 milestones. The SPAC parallel is sharp: Report 6 documents that clean energy SPACs (Proterra, Lilium, Lordstown) trading at infinite pre-revenue multiples collapsed 90-100% on execution failures.

The Pivotal Variable

The single variable that most determines which scenario materializes is the Aurora-INL demonstration outcome in 2027-2028. A successful criticality and power generation demonstration would validate the reactor design, de-risk NRC licensing for subsequent units, convert non-binding LOIs into PPAs, and justify the capital required for fleet deployment. Failure or significant delay would likely trigger the bear case cascading effects: pipeline evaporation, dilutive capital raises at depressed prices, and competitive displacement by TerraPower and others already further in regulatory process.

---

6. Key Non-Obvious Insights

Insight 1: Oklo's Real Competitive Moat May Not Be Where the Market Thinks It Is

The consensus narrative focuses on fuel recycling and Sam Altman's AI connections. But the most immediately valuable asset is Oklo's DOE Reactor Pilot Program selection, which enables a parallel authorization pathway at INL that bypasses the standard NRC timeline (Reports 1, 2). This is why Oklo broke ground in September 2025 despite having no NRC license—it doesn't need one for the demo. Only 3 of 11 applicants were selected (Report 7). This DOE pathway creates a 1-2 year lead over competitors restricted to NRC-only routes (Report 5), but the market appears to conflate the DOE demo with commercial licensing, which still requires full NRC review for every subsequent unit (Report 2).

Insight 2: The HALEU Cost Assumption Gap Is a Hidden Valuation Risk

Oklo's entire unit economic model—the foundation for every analyst DCF—derives from a 2023 investor presentation assuming HALEU at $7,000/kg (Report 1). Actual market prices exceed $35,000/kg, a 5x gap (Report 1). This isn't a minor input sensitivity—it's the difference between a projected LCOE of ~$40/MWh (competitive with gas) and something potentially closer to $100+/MWh (uncompetitive without subsidies). The Centrus JV and fuel recycling are designed to close this gap eventually, but "eventually" means the 2030s for recycling at scale (Report 3). No analyst report cited in the research appears to stress-test this assumption adequately.

Insight 3: Meta's Prepayment Is More Important Than the Pipeline Headline

The 18 GW pipeline number is impressive but analytically misleading—Switch's 12 GW master agreement alone constitutes two-thirds of it and is non-binding (Report 1). The Meta deal is qualitatively different: it involves actual prepayments funding development (Report 4), creating non-dilutive capital flow for a pre-revenue company. This is the proof-of-concept for Oklo's BOO model—if hyperscalers will fund reactor development through prepayments, Oklo's capital intensity becomes manageable. But Meta simultaneously signed deals with Vistra (2.4 GW restarts, online 2027) and TerraPower (up to 2.8 GW) as part of its 6.6 GW nuclear strategy (Reports 3, 4, 8). Meta is hedging across the nuclear landscape, not betting exclusively on Oklo.

Insight 4: The "Worst Applicant" Label Creates Asymmetric Political Risk

Report 8 documents that a senior NRC official called Oklo "the worst applicant the NRC has ever had," while former Chair Macfarlane described the process as "extremely frustrating." This institutional memory within the NRC creates a headwind that goes beyond technical deficiencies. Senator Markey has flagged conflicts of interest regarding Energy Secretary Chris Wright's prior Oklo board service and DOE plutonium transfers (Report 8). If political winds shift—or if Oklo's DOE fast-track draws scrutiny for bypassing safety protocols—the regulatory environment could tighten rather than loosen. NPR and Bloomberg have already published critical investigations (Report 8). This is a risk that doesn't appear in Oklo's investor presentations.

Insight 5: The Timeline Collision Between Nuclear SMRs and Data Center Urgency

Reports 4 and 8 converge on a critical finding: hyperscalers are not waiting for SMRs. Natural gas provides 40%+ of current data center power, with 252 GW of gas pipeline capacity in development (97 GW data-center-specific) deployable in 1-2 years (Report 8). Nuclear restarts—Three Mile Island for Microsoft (2027), Palisades (2026), Duane Arnold for Google (2029)—deliver GW-scale clean power years before any SMR (Report 8). Solar+storage at $25/MWh is the cheapest option where intermittency is acceptable (Report 8). The risk is not that nuclear loses the long-term argument for data centers—it likely doesn't—but that by the time Aurora units are operating at scale (early 2030s), the most acute demand may have been met by faster alternatives, reducing willingness to pay premium prices for SMR power.

---

7. Watch Out For

• Prepayment reversion risk: Report 6 notes that Meta/Equinix prepayments could convert to debt obligations if Oklo misses July 2026 milestones. This would flip non-dilutive funding into a liability on a company with no revenue.

• Insider selling patterns: Founders DeWitte and Cochran have sold approximately $152 million in stock through 10b5-1 plans since lockup expiration (Report 7). While framed as routine diversification, the volume is notable for a pre-revenue company whose thesis depends on founders' long-term conviction.

• FOAK base rates: Report 8 documents that no first-of-a-kind advanced reactor in the West has completed on schedule or budget, with average overruns of 102.5% and 35-month delays. Oklo's 2023 investor deck projecting $1,600/kWe capex for a 15 MWe unit sits at the extreme optimistic end of a distribution where $6,000-10,000/kWe is the FOAK norm (Report 8).

• Dilution trajectory: Shares outstanding grew 13% YTD through Q3 2025, and the $3.5 billion shelf registration (including $1.5 billion ATM) signals continued issuance. Report 5 estimates $1-2 billion in additional capital needed by 2030 for fuel recycling and fleet deployment, with total requirements potentially reaching $14 billion by the 2040s.

• Research report conflicts: Report 1 cites $1.2 billion cash (Q3 2025), while Report 6 references $922 million (Yahoo Finance mrq estimate)—likely reflecting different measurement dates or inclusion criteria. Report 5 breaks this into $410 million cash + $774 million marketable securities, suggesting the higher figure is correct but includes less-liquid instruments. Reports 3 and 6 offer somewhat different market cap figures for peers, reflecting volatile trading. On timeline, Report 1 targets "late 2027/early 2028" for Aurora-INL commercial ops, while Report 8 suggests "2028+" is more realistic given FOAK history.

---

8. Questions to Explore

1. What exactly does Meta's prepayment cover, and what milestone triggers reversion? The specific dollar amount, contractual structure, and failure conditions are not publicly disclosed in the research—yet this is arguably the most important near-term financial variable.

2. Has Oklo updated its unit economics since 2023? The $7,000/kg HALEU assumption is two years old and 5x below market. If internal models have been revised, the change would materially alter every downstream valuation metric.

3. What does the DOE Reactor Pilot Program authorization actually permit? The distinction between DOE-authorized criticality and NRC-licensed commercial operations is blurred in Oklo's public communications. Clarity on whether INL-generated power can be sold (Report 1 notes "no grid sales at INL initially") would bound the revenue timeline.

4. How does TerraPower's construction permit affect Oklo's hyperscaler conversations? With an identical coolant technology, larger output, and regulatory lead, TerraPower may be capturing deals that Oklo's pipeline assumes it will convert. The overlap is most acute at Meta, which has signed with both companies.

5. What is Oklo's actual NRC COLA submission timeline? The Phase 1 submission was planned for late 2025 but as of March 2026 remains listed as "upcoming" (Report 2). Each quarter of delay compresses the path to the 2027-2028 commercial target, which already depends on an optimistic 24-36 month review.