Oklo Inc.（OKLO） · Nuclear Energy

Oklo (OKLO.US) Zen Horizon Research Report

Jun 8, 2026·~26 min read

Quick ReadPlain-language overview · read this first

Oklo is a U.S. company that builds small nuclear reactors. The report's stance is “Watch,” meaning observe it for now; it is not cheap at the current price.

It differs from ordinary nuclear-power companies: others sell reactors, while Oklo sells electricity. It builds and operates its own power plants and signs long-term power-supply contracts to sell electricity to customers, mainly targeting AI data centers with huge power demand. That sounds like it has caught the right trend, but the problem is that Oklo has to fund all the power-plant construction itself.

The most important point: the company currently has not generated a cent of revenue, and the first meaningful power-sales revenue will not arrive until 2028, yet the market values the whole company at about 10 billion dollars. It does have about 2.5 billion dollars in cash on the balance sheet and no debt, so it has ample ammunition. But that also means three quarters of the company's value rests on story and expectations, not hard cash flow.

A few things deserve the most caution. First is fuel cost: a short seller has calculated that the true price of the nuclear fuel it uses may be 5 times what the company suggests. If that is true, whether the power generation economics work becomes a major question. Second, TerraPower, a competitor on the same path and backed by Bill Gates, has already moved ahead and taken the lead. Third, company insiders have been selling shares, with cumulative cash-outs of about 369 million dollars.

On price, the current price is about 58 dollars. The report thinks it falls within a “holdable” range, but there is no downside cushion at all, and being wrong could mean a near-total loss. The report believes the ideal buy price would require a pullback below 32 dollars. For investors who have not bought yet, the current price does not offer a reason to buy.

This is only an explanation of the report, not investment advice. Stocks carry risk; invest with caution.

Lead
Oklo is an advanced nuclear fission developer founded in 2013 by two MIT nuclear engineering PhDs and taken public on the NYSE in 2024 through Sam Altman's SPAC, with its sodium-cooled fast reactor Aurora (75 MWe) and a build-own-operate model that sells power rather than reactors to AI data centers under long-term PPAs. The core thesis is real policy access and real cash, but a pre-revenue company with a Q1 2026 net loss of $33M, $2.54B of cash, and DOE reactor-pilot selection is already carrying a very long-dated option value. Report Rating Watch: an excellent story with no margin of safety at the current price.

Full report

Prices in the article are as of publication; see the valuation band above for the live price.

Research Perspective Statement

This report applies the Zen Horizon Framework to Oklo Inc. (New York Stock Exchange: OKLO, Chinese name "Aokeluo") as a third-party deep-dive study. Vertically, it reviews the company's full commercial arc from its 2013 founding, the NRC denial in 2022, its 2024 public listing through Sam Altman's SPAC, the surge to an all-time high of $193.84 in 2025, and the subsequent 70% collapse. Horizontally, it compares Oklo with the global advanced nuclear and SMR competitive landscape, including listed peers NuScale, X-energy, Centrus, BWX, and TerraPower, the private Bill Gates-backed sodium-cooled fast-reactor player that has already moved ahead on schedule. All key numbers are anchored in primary filings from the U.S. Securities and Exchange Commission (SEC), including 10-K, 10-Q, 8-K, S-1/S-4, DEF 14A proxy statements, and Form 4 insider transactions, plus the company's investor-relations materials and official disclosures from the U.S. Department of Energy (DOE) and the Nuclear Regulatory Commission (NRC). Secondary data have been cross-checked against at least two sources. Where definitions differ, they are marked with ⚠️.

Data are as of the U.S. market close on 2026-06-05 (OKLO $58.09, -11.16% that day; 2026-06-06 and 2026-06-07 were weekend non-trading days). Financial data are as of FY2026 Q1 (2026-03-31, released 2026-05-12) and FY2025 full year (2025-12-31, 10-K filed 2026-03-17). Oklo's fiscal year matches the calendar year. This is research analysis, not investment advice. Monetary figures are in U.S. dollars unless otherwise stated. The rating conclusion is independent of the site's growth scorecard.

One common correction: Oklo is listed on the New York Stock Exchange (NYSE) under the ticker OKLO and has a single class of common stock, one share one vote. Although the company uses the wording "Class A common stock," there are no super-voting rights and no Class B shares. It is therefore not the "Nasdaq" or "dual-class share structure" sometimes described loosely in secondary sources. The founders influence the company through roughly 12% economic ownership plus board seats, not through super-voting control (2026 DEF 14A proxy statement).

1. Conclusion First

Rating: Watch. Oklo is an advanced nuclear fission developer with excellent positioning, a powerful story, and very thin fundamentals. The positioning is real: it is the only company with 3 project slots in the U.S. Department of Energy's Reactor Pilot Program, covering the Aurora powerhouse, the Groves isotope test reactor, and the Pluto plutonium-fuel reactor, while other companies each received only 1 (BusinessWire 2025-10-01). Its first reactor is following the DOE authorization path, invoking the Atomic Energy Act exemption for construction "for the account of DOE" to build and operate first while bypassing the NRC commercial-licensing bottleneck, which is the institutional basis for management's "first power in 2028" claim. It has $2.54B in cash and no long-term debt (Q1 2026 10-Q). Its named customer pipeline is about 14 GW and includes data-center giants such as Meta, Switch, and Equinix. The fundamentals are thin: Oklo has zero revenue, a Q1 2026 net loss of $33.1M, no meaningful power-generation revenue before 2028, and a current market cap of $10.11B. That means roughly 75% of its market value is story and option value outside cash.

One-sentence logic: Oklo has stitched together three of the hottest narratives, AI data-center power shortages, the nuclear renaissance, and the Altman halo, while adding genuine DOE pilot-program positioning and real cash ammunition of $2.5B. But a pre-revenue company carrying a $10B market cap, with first power not expected until 2028, fuel economics targeted by short seller Kerrisdale, regulation still not fully closed, insider net selling of $369M, and TerraPower, a Gates-backed peer on the same technology route, already ahead with an NRC construction permit and active construction, creates a classic "good story, bad price, no floor" payoff structure: the downside is permanent capital loss if the option value collapses; the upside requires multiple milestones to go right over many years.

Core tension: The policy and funding position is real (DOE three-project sweep + DOE authorization path + $2.54B cash + 14 GW pipeline + first reactor already broken ground) versus fundamentals and valuation that are far apart (zero revenue supporting $10.11B, first power in 2028, no margin of safety), fuel economics questioned at 5x by Kerrisdale, regulation not closed (with the 2022 denial as prior evidence), one-way insider net selling of $369M, and TerraPower moving ahead on the same route. The result is "Watch." The story is independently coherent, but the current price offers no margin of safety. In substance, this is a pre-revenue option and momentum vehicle.

Valuation and buy zone (USD):

Scenario	Range	Key Assumptions
Bear	$18-32	Kerrisdale's fuel-economics critique proves valid / NRC delays again / LOIs do not convert / continued discounted equity issuance; market cap falls back toward $2.5B cash plus modest option value, about $3-5.5B. The downside is permanent capital loss.
Base	$48-76	First reactor broadly arrives in 2028 / part of the PPA pipeline converts into binding contracts / small isotope revenue begins; InvestingPro model at $48 versus cautious sell-side Hold targets from Goldman ($66) and Citi ($76).
Bull	$110-140	AI power narrative fully plays out / high conversion of the 14 GW pipeline / DOE pilot succeeds / multiple reactors plus isotope revenue scale; bullish sell-side targets from Wedbush ($110), Cantor ($122), and Canaccord ($125).

The current price of $58.09 sits inside the base range. It is neither deeply undervalued nor at the top of a bubble. It is the market's and sell-side's risk-weighted pricing of a "14 GW non-binding pipeline times a risk discount." But readers must understand that the tails are very wide across all three scenarios, with downside of -45% to -69% and upside of +90% to +141%. The downside is permanent loss if option value goes to zero; the upside is the compound probability of multiple milestones across years. Margin-of-safety conclusion: none. The ideal buy zone is <= $32, where EV would be close to cash plus a risk-adjusted pipeline option value, about 45% below the current price.

2. What the Company Is: From UPower to Aurora

Oklo is not a conventional nuclear company that sells reactors. To understand it, start with three differences: it sells power, not reactors; it uses a sodium-cooled fast reactor, not the mainstream light-water-reactor route; and its money comes entirely from selling stock, not selling products, because it does not yet have a product for sale.

The company was founded in 2013 by the husband-and-wife team Jacob DeWitte and Caroline DeWitte, formerly Caroline Cochran. Both are trained nuclear engineers from MIT. Jacob DeWitte earned an MIT master's degree in nuclear engineering in 2011 and a PhD in 2014 after an undergraduate degree in nuclear engineering from the University of Florida. Caroline earned an MIT master's degree in nuclear engineering after undergraduate studies at the University of Oklahoma in economics and mechanical engineering, and later served as a member of a DOE nuclear-energy advisory committee (2026 proxy executive biographies). The company was originally called UPower and entered Y Combinator in 2014. The early problem it wanted to solve was straightforward: provide power to remote locations that had to burn diesel, where fuel delivery was expensive and difficult (MIT News 2020-11).

Its flagship product, Aurora Powerhouse, is a sodium-cooled fast reactor that burns metal HALEU fuel (high-assay low-enriched uranium, 5-20% U-235). Its design directly inherits from Idaho National Laboratory's EBR-II experimental breeder reactor, which operated for 30 years from 1964 to 1994 and is regarded in the nuclear industry as a benchmark for passive safety. The first Aurora reactor is even expected to reuse metal HALEU fuel originally prepared for EBR-II (World Nuclear News). Its power rating has moved upward over time: from an early 1.5 MWe microreactor to 15 MWe, then to 50 MWe and 75 MWe on 2025 earnings calls. The current first reactor at INL is 75 MWe (ANS 2025-09).

There is also a historical scar here. In March 2020, Oklo submitted a customized combined license application (COLA) for Aurora to the NRC. It was the first advanced-reactor COLA ever accepted by the NRC. But two years later, on 2022-01-06, the NRC "denied without prejudice" the application because Oklo repeatedly failed to provide enough information on two safety-critical issues: maximum credible accident methodology and safety classification of structures, systems, and components (SSC classification). After three rounds of requests for supplemental information, the NRC still found Oklo's report "conceptual" and insufficiently descriptive on methodology, making it impossible to schedule the review (NRC press release 22-002). "Without prejudice" means the company can resubmit after filling the gaps, but this denial remains the "regulatory hard wound" repeatedly cited by shorts. A former NRC commissioner later said publicly that Oklo had "a lot of hubris."

3. Vertical Review: Development History and Capital-Market Narrative

3.1 Public Through Altman's SPAC, Then Broke Issue Price and Halved

In May 2024, Oklo went public through a reverse merger with AltC Acquisition Corp, a blank-check company sponsored by Sam Altman, CEO of OpenAI, and Michael Klein, the former Citi investment banker and founder of the Churchill SPAC series. When the transaction was announced on 2023-07-11, Oklo was valued at about $850M (Bloomberg). The merger closed on 2024-05-09, bringing in roughly $306M of gross proceeds plus $25M of customer prepayments, and OKLO began trading on the NYSE on 2024-05-10.

Yet this hot "Altman concept plus AI nuclear power" SPAC had a brutal start: first-day close of $8.09, about -51% below the $16.45 opening price. That was a typical de-SPAC selling-pressure event. The stock hovered near the $10 anchor for months afterward and briefly fell to an all-time low of $5.35 in September 2024, when the market treated it as another failed de-SPAC.

3.2 Four-Stage Narrative Arc: From $5 to $194 and Back to $58

Stage 1, dead-money broken-issue period (2024-05 to 2024-10): First-day halving, low-price drift, and few catalysts.
Stage 2, narrative ignition from 2024-11: On 2024-12-18, Oklo signed a 12 GW master power agreement with Switch, marketed as one of the largest corporate clean-power agreements ever. The order book jumped from 2.1 GW to "about 14 GW," and the stock doubled from $10 to $20-23. In the same month, short seller Kerrisdale published its short report, discussed in the risk section.
Stage 3, surge to $193.84, ending 2025-10-15: In 2025, catalysts piled up: Trump signed 4 nuclear-energy executive orders in May; Oklo completed the NRC pre-application readiness assessment in July; the first INL reactor broke ground on September 22; and on October 1, Oklo was selected for the DOE pilot program and uniquely received 3 project slots. These pushed the stock to an all-time intraday high of $193.84 on 2025-10-15, with market cap briefly exceeding $30B while the company still had zero revenue and zero operating reactors. From the post-listing low, the stock rose more than 35x, one of the most extreme momentum curves of 2025 (Trefis).
Stage 4, de-enchantment and collapse (2025-10 to 2026-06): The triggers by weight were: 1. Google Gemini 3 was released in mid-November 2025, trained throughout on more power-efficient in-house TPUs, which ignited fear that "AI may not need that much power" and directly hit Oklo's "AI power shortage to SMR necessity" logic. The stock fell 31% in November (Motley Fool). 2. Extreme valuation mean reversion plus continued dilutive issuance, with about $3.5B of stock sold over the past year to fund survival. 3. Wider Q3 and Q1 losses. 4. Wolfe Research initiated coverage on 2026-05-19 at "Hold" with a fair-value range of $51-71. The stock fell 77% to a 52-week low around 2026-03-30 at $44.88 intraday and $45.58 at the close, rebounded, then fell again to $58.09 at the 2026-06-05 close, down 70% from the peak and up 29% from the 52-week low.

This arc itself is a warning: this is a high-beta momentum stock driven by narrative and liquidity, with fundamentals not yet delivered. Over the past 12 months, it could rise 12-36x and also fall 77%, while revenue stayed at zero throughout.

4. Business Model and Moat

4.1 Build-Own-Operate: Selling Power, Not Reactors

Oklo's core differentiation is its business model. Traditional nuclear vendors such as Westinghouse and GE sell reactor hardware plus EPC services, recognize one-time revenue, and leave customers to bear ownership, operation, capital, regulatory risk, and complexity. Oklo does the reverse. It builds, owns, and operates its own power plants, turning one-time reactor sales into recurring power-sales cash flow through long-term power purchase agreements, mostly 20-year PPAs. Customers "buy power, not plants" (10-Q text: "sell power in the forms of electricity and heat directly to customers").

The benefit is a larger imagination space, long-duration cash flow, and sticky data-center customers. The downside is a double-edged sword: selling power means Oklo itself carries all construction, financing, and operating capex. Its balance sheet must be much heavier than an asset-light peer that sells reactors or licenses. This also explains why it must keep issuing stock.

4.2 Fuel Chain: HALEU on the Front End, Recycling on the Back End, a Real Position and a Real Bottleneck

Oklo has laid out both ends of the nuclear fuel chain:

Front end, HALEU procurement: Aurora burns HALEU. In 2019, Oklo obtained 5 metric tons of HALEU through a competitively awarded INL cooperation agreement, sourced from DOE EBR-II spent-fuel recycling. This was its earliest fuel position (DOE.gov). It also works with Centrus Energy, sourcing HALEU from Centrus's American Centrifuge Plant in Piketon, Ohio, the only NRC-licensed HALEU production facility in the United States. In 2026-03, the two companies signed a joint venture for HALEU deconversion (Centrus official release). But one correction matters: the 5 companies in the DOE HALEU Availability Program's first allocation in 2025-04 were TRISO-X, TerraPower, Kairos, Radiant, and Westinghouse. Oklo was not on that list. Its HALEU source is the separate 2019 agreement above, not the DOE's public allocation line.
Back end, spent-fuel recycling: In 2025-09, Oklo announced construction in Oak Ridge, Tennessee, of the first privately funded spent-fuel recycling facility in the United States, with investment of up to $1.68B. The facility would use pyroprocessing, an electrochemical dry-reprocessing method, to convert light-water-reactor spent fuel into metal fuel usable by Oklo's fast reactors. On 2026-05-26, Oklo was also selected for the DOE's surplus plutonium utilization program, converting Cold War-era weapons plutonium into advanced-reactor fuel (ANS).

The closed-loop story is attractive, but the bottleneck is hard. Russia was previously the only commercial supplier of HALEU, and the United States legislated a ban on Russian uranium imports in 2024-05. Domestic Centrus capacity is ramping slowly. By mid-2025, Centrus had cumulatively delivered just over 920 kilograms to the DOE, while Kerrisdale estimated that one initial 15 MWe reactor load alone would require about 4,750 kilograms. Near-term U.S. HALEU supply remains an order of magnitude short of commercialization demand, making fuel the most realistic constraint on the closed-loop story.

4.3 Isotopes: The Earliest Cash-Flow Source

In early 2025, Oklo acquired radioisotope company Atomic Alchemy for $28.4M, including share-price changes (primary 10-Q). Its proprietary VIPR technology can produce more than 40 isotopes, including Ac-225 for cancer therapy and Pu-238 for space RTGs. This may be Oklo's earliest revenue business. The Groves isotope test reactor in Texas targets criticality on 2026-07-04, while a radiochemistry laboratory in Idaho has already received an NRC materials license and is advancing the first commercial isotope contract. Management's wording is that first, small revenue could arrive "as soon as 2026." But the radiopharmaceutical market of about $5.5B in 2022 and the "total isotope market" of about $55.7B are two different definitions and should not be treated as one TAM.

4.4 Moat Score: 2.5/5

On balance, Oklo's moat scores 2.5/5.

The real parts: 1. the DOE pilot three-project sweep plus the DOE authorization path around the NRC bottleneck is a unique policy position among peers; 2. build-own-operate can theoretically lock in recurring cash flow; 3. $2.54B of cash is a rare funding moat among peers, with NuScale at about $1B and Nano far lower; 4. the company is positioned on both the front and back ends of the fuel chain.

The questionable parts: 1. regulation is not closed. The 2022 denial remains, and the first reactor uses a DOE track rather than an NRC commercial license. "Reaching criticality" is not the same as "commercial power sales." 2. With zero revenue and first power in 2028, the moat has not yet converted into a dollar of profit. 3. Build-own-operate requires Oklo to bear all capex, making it an asset-heavy model. 4. Most importantly, TerraPower, backed by Bill Gates and pursuing the same sodium-cooled fast-reactor route most similar to Aurora, received an NRC construction permit in 2026-03 and began construction in Wyoming in 2026-04, moving ahead of Oklo across the board. Oklo's differentiated "advanced nuclear first mover" narrative faces direct falsification risk.

5. Vertical Financial Review: Anatomy of a Cash-Burning Machine

Oklo's financial statements read very differently from those of ordinary companies because there is no revenue line. The analysis starts with operating expenses.

Q1 2026, as of 2026-03-31 (10-Q):

Revenue $0 (pre-revenue).
Operating expenses of $51.2M, including $27.0M of R&D and $24.2M of G&A, up 187% year over year.
Net loss of -$33.1M, but 47% of that was non-cash stock-based compensation, with SBC of $15.6M. Another $21.3M of interest income from $2.5B of cash flattered the income statement. Core operating loss was actually -$30.0M.
EPS of -$0.19.
Operating cash outflow of only -$17.9M, which is the "cash actually burned" and far below the accounting net loss because SBC is added back.
But capex was -$32.8M, entering the reactor-construction phase and up sharply from $0.3M year over year.
Total quarterly cash consumption = operating $17.9M + capex $32.8M = $50.7M.

Balance sheet, as of 2026-03-31: cash and equivalents of $1.59B plus marketable securities of $0.94B, totaling $2.54B; no long-term debt; total liabilities of only $64.9M; shareholders' equity of $2.64B; accumulated deficit of $273.8M; and no going-concern concern. The 10-Q explicitly says cash is sufficient to support operations for one year from the release date.

The runway must be de-misread: If one looks only at operating cash burn, FY2026 guidance of $80-100M per year would make $2.54B last more than 25 years. That is wrong. FY2026 guidance is operating cash burn of $80-100M plus reactor-construction capex of $350-450M, or about $430-550M per year. Using true total cash use including capex, runway is about 4-5 years. Reactor construction is the large item and the real liquidity cushion.

Dilution is structural: Shares outstanding have risen from about 94M after the SPAC transaction to 173.9M (financecharts). The path was: an underwritten offering of 6.67M shares at $60 in 2025-06, netting $383M, followed by a $1.5B ATM program established in 2025-12 and largely used, raising about $1.5B cumulatively, followed by a new $1.0B ATM program on 2026-05-13 (Orrick announcement). Zero revenue plus asset-heavy build-own-operate means continuous equity financing is structural. The new $1.0B ATM means dilution will continue. Kerrisdale cites sell-side estimates that Oklo's five-year deployment plan still needs about $2.7B of incremental capital.

6. Industry and Cycle: Sitting at the Intersection of Three Hot Themes

Oklo's valuation cannot be separated from the rare moment when three tailwinds are blowing at the same time:

AI data-center power shortage: Power density per hyperscale data-center rack has surged, making electricity a hard constraint on AI compute expansion. "Nuclear power for AI" became one of the strongest themes of 2024-2025.
Nuclear renaissance plus policy tailwind: On 2025-05-23, Trump signed 4 nuclear-energy executive orders, covering NRC reform with an 18-month licensing deadline, rapid deployment of reactors on DOE and Department of Defense sites, rebuilding the fuel chain, and raising the 2050 nuclear-capacity target from 100 GW to 400 GW (DOE). Oklo is closely connected to this administration. Altman previously served as chairman, and former director Chris Wright became Energy Secretary. Oklo is one of the clearest policy beneficiaries.
SMR investment boom: Small modular reactors are viewed as a "faster, smaller, cheaper" substitute for traditional large nuclear plants, attracting heavy capital inflows.

But the sector has an unavoidable warning case: NuScale's UAMPS project. This was the closest U.S. SMR flagship project to implementation. Its target power price rose from $58/MWh to $89/MWh, even after about $4 billion of federal subsidies, while construction cost rose from $5.3B to $9.3B, up 75%. The project was ultimately canceled in 2023-11 after roughly 10 years of effort because costs lost control (IEEFA). SMR levelized cost of electricity (LCOE) is widely questioned as high, with first-of-a-kind estimates of $80-150/MWh, far above combined-cycle gas at $40-75/MWh. "SMR is too expensive" is a systemic risk for the whole sector, and Oklo is not automatically immune.

7. Horizontal Review: Competitive Comparison, Who Has Real Revenue and Who Is Pure Story

Placed back into the advanced-nuclear landscape, Oklo's valuation looks especially aggressive. The table below uses a unified 2026-06-05 closing basis, with stockanalysis as the primary source. That day was a broad selloff in the nuclear sector, so prices were depressed:

Company	Ticker	Listed / Private	Technology Route	Deployment Progress	Market Cap (2026-06-05)	TTM Revenue / Profitability
Oklo	OKLO	Listed	Sodium-cooled fast reactor Aurora + fuel recycling	First reactor in 2028 (INL, ground broken)	$10.11B	Zero revenue; net loss -$129M
NuScale Power	SMR	Listed	Light-water pressurized SMR	Only NRC-certified design; RoPower decision in 2026	$3.84B	$18.7M; net loss -$386M
X-energy	XE	Listed (2026-04 IPO)	High-temperature gas reactor + TRISO fuel	Dow customer deployment	$8.40B	$94M; net loss -$390M
Nano Nuclear	NNE	Listed	Microreactor (early stage)	No formed product	$1.23B	Zero revenue; net loss -$31M
Centrus Energy	LEU	Listed	HALEU fuel supply (not reactors)	Production started, backlog to 2040	$3.18B	$452M; profitable +$61M
BWX Technologies	BWXT	Listed	Nuclear components + naval nuclear propulsion	Mature operations	$17.04B	$3.38B; profitable +$345M
TerraPower	Private	Private (Gates-backed)	Sodium-cooled fast reactor Natrium (same route)	Construction started in 2026-04, NRC construction permit in hand	2022 = $3.8B (current not disclosed)	Not disclosed

This table says three things:

First, Oklo has the most aggressive valuation in the group. It has zero revenue and a $10.11B market cap, while NuScale, which holds the only NRC-certified design, has a market cap of only about one-third of Oklo's. Every sell-side peer valuation of Oklo uses NuScale to pressure the multiple.

Second, the biggest threat to Oklo is private TerraPower. Gates-backed TerraPower follows the same sodium-cooled fast-reactor lineage as Aurora, with Natrium at 345 MWe plus molten-salt storage, yet is clearly ahead on schedule. It received the first U.S. commercial reactor construction permit in nearly a decade in 2026-03, also the first non-light-water commercial power reactor construction approval in more than 40 years, and officially began construction in Kemmerer, Wyoming, in 2026-04 (GeekWire). Behind it are more than $1 billion of Bill Gates's personal investment and heavy-capital backing from NVIDIA, SK, and HD Hyundai. Oklo's differentiated "advanced nuclear first mover" narrative is being directly challenged by a same-route competitor that is moving faster.

Third, the companies with real revenue are the "fuel and picks-and-shovels" players. Centrus, which sells HALEU and has $452M of revenue and profit, and BWXT, which sells nuclear components and naval nuclear propulsion with $3.38B of revenue and profit, show where money is actually being made in the nuclear value chain: supply-chain chokepoints, not developers selling power stories. Oklo wants to turn the developer into the profit pool too, but that waits until 2028.

8. Current Fundamentals and Bull-Bear Debate

What is the market trading now? In one sentence: it is trading the long-dated option that "Oklo can convert 14 GW of PPA intent plus DOE pilot positioning into real power-sales cash flow during 2028-2034." The stock is extremely sensitive to narrative. When Gemini 3 was released and the market questioned whether AI was really that power-hungry, the stock fell 31%. When the DOE signaled loan support, the stock rebounded 50%.

Bull case, each point requiring evidence:

DOE pilot three-project sweep plus DOE authorization path, the only peer position that can "race ahead around the NRC."
14 GW customer pipeline plus endorsements from data-center giants Meta, Switch, and Equinix.
$2.54B cash equals ample ammunition and the strongest funding moat among peers.
Fuel recycling plus isotopes provide second and third growth curves, with isotope revenue possible as soon as 2026.
23 analysts with a consensus Buy and an average target of $88.89, or +53%.

Bear case, each point also requiring evidence:

Zero revenue supporting $10B, with first power in 2028, while bears think it may slip to 2030.
Kerrisdale's fuel-economics attack, detailed in the next section. If HALEU costs $35,000/kg rather than Oklo's assumed $7,000/kg, the commercial feasibility narrative collapses.
Most of the 14 GW pipeline is revocable, non-binding LOIs, with only Meta's 1.2 GW binding.
Continuous equity-financing dilution, including the new $1.0B ATM.
Regulation is not closed, with the 2022 denial and a former NRC commissioner calling the timeline "not credible."
TerraPower is ahead on the same route.
Insider net selling of $369M.

The disagreement itself is a signal: Target prices from 23 analysts range from a low of $14 to a high of $140, a 10x spread. That shows that for a zero-revenue company, there is no unified valuation anchor. Bull targets are essentially "narrative and pipeline conversion targets," not cash-flow anchors.

9. Valuation Analysis: Pricing a Zero-Revenue Company

9.1 Conventional Multiples All Fail

Oklo has no revenue, so P/S is not applicable because the denominator is 0. The only usable metrics are:

P/B of 3.83x, using $10.11B market cap divided by $2.64B shareholders' equity, with book value per share of $15.18.
Market cap / cash of about 4.0x, using $10.11B divided by $2.54B, which means about 75% of market cap is "story/option value" outside cash.
Cash per share of $14.58, meaning the stock trades at 4.0x cash per share.
EV of about $7.57B, market cap minus net cash. This entire $7.57B is expectation value for future commercial operation.

9.2 Market Cap / Pipeline: Looks Cheap, but Only Under a "100% Conversion" Assumption

The market's implied pricing for Oklo is $10.11B divided by 14 GW, or $722/kW. On an EV basis it is $541/kW. Compared with overnight costs of $6,000-12,000/kW for traditional new-build large nuclear plants, and more than $15,000/kW for the Vogtle project, Oklo is being valued at only about $540-720 per kW, which looks "cheap" on the surface. But this treats the entire 14 GW non-binding intent pipeline as if it converts 100%. Actual commercial capacity under construction or approved is 0, and the first reactor will not operate commercially until 2028. Put differently, the market has already discounted "the whole pipeline becomes long-term PPA cash flow" into today's stock price.

9.3 Three Valuation Frameworks, Three Answers

For a pre-revenue nuclear developer, the industry uses three mutually competing frameworks, which map neatly to three sell-side camps:

Model camp, InvestingPro fair-value model: about $48. Based on comparables plus cash flow, it called Oklo overvalued in 2025-11 when the stock was $132, and the stock later did fall to $44-58 (Investing.com).
Discounting camp, Goldman Hold at $66, Citi Neutral at $76, and Wolfe fair value of $51-71: close to the current price, implying that "the valuation is roughly reflected."
Narrative/option camp, Wedbush/Dan Ives at $110, Cantor at $122, and Canaccord at $125: Oklo is treated as "a basket of call options on nuclear commercialization." The $2.5B of cash is exercise ammunition, first-reactor approval is the exercise trigger, and extremely high volatility lifts option value.

The current price of $58.09 sits between the "model camp" at $48 and the "discounting camp" at $66-76.

9.4 Margin-of-Safety Review: None

Under the Zen Horizon Framework discipline, the current price of $58 is a large premium to the bear scenario of $18-32, leaving no margin of safety. If revenue remains zero over the next 3 years, which is the high-probability path, annualized return at the current price will depend purely on narrative-driven valuation changes, with no earnings support. This is a classic good story at a bad price, even an unverified story at a bad price. Margin-of-safety conclusion: none. A zero-revenue company waiting until 2028 for first power and carrying a $10B market cap offers no downside protection at $58. The downside is permanent capital loss if option value goes to zero.

10. Risk Analysis

10.1 Fuel Economics: Kerrisdale's "Fission Impossible" Attack, the Most Fatal Issue

On 2024-11-20, short seller Kerrisdale Capital published a report titled "Fission Impossible," attacking Oklo's unit economics. The sharpest claim was that Oklo assumed HALEU fuel cost of $7,000/kg in its calculations, while Kerrisdale argued this was "understated by 5x" and that the real figure should be $35,000/kg, with expert estimates at $30,000-40,000/kg. Recalculated on that basis, overnight capital cost for a 50 MWe plant rises from Oklo's stated $2,312/kW to $6,680/kW, and LCOE rises from $45/MWh to $94/MWh. For an unsubsidized 15 MWe first reactor, LCOE would be as high as $230/MWh. If true, Oklo's stated $40-90/MWh competitiveness range collapses and aligns with the broader industry critique that "SMR is too expensive." The report also cites a former NRC commissioner saying the "2027 timeline" is not credible and that licensing would take at least 4 years, and cites sell-side estimates that another $2.7B of capital will be needed over 5 years (Kerrisdale report).

Probability high / impact high. Observable indicators: the actual $/kg disclosed in future Oklo fuel-procurement contracts and the actual $/kW overnight cost of the first reactor.
One balancing point to treat carefully: Kerrisdale's $35,000/kg is a first-of-a-kind cost, while the Oklo plus Centrus joint venture and DOE fuel-line pilots are intended to lower cost at scale. The actual commercial price is likely somewhere between $7,000 and $35,000. The 5x figure may be an upper bound. The short case is not necessarily entirely right.

10.2 Regulation and Execution: It Has Never Built a Commercial Reactor

Aurora is first-of-a-kind. Oklo has never built or operated any commercial reactor, yet it wants to build and operate hundreds of them itself. The MCA and SSC methodology gaps behind the 2022 NRC denial are exactly the kinds of issues a new COLA substantive review must solve. "Accepted for review" is not the same as "technical gaps have been eliminated." Add the engineering history of sodium-cooled fast reactors, including Japan's Monju, which suffered a 640-kilogram sodium leak and fire in 1995 and was eventually abandoned, France's Superphénix, and Russia's BN-600, which had 27 sodium leaks and 14 sodium fires over 17 years. Sodium reacts violently with water and air, a systemic risk. Oklo's selling point is reusing the mature EBR-II design, but the first reactor is already commercial scale, and 75 MWe is much larger than EBR-II's roughly 20 MWe class. First-of-a-kind scale-up risk remains. Probability medium / impact high.

10.3 Dilution and Finance: No Revenue, Survival Through Stock Sales

Shares outstanding rose 28% TTM, a new $1.0B ATM has been established, and another $2.7B of capital may be needed over 5 years. If capital markets cool or the stock price keeps falling, discounted issuance can create a negative feedback loop of more dilution and lower per-share value. Probability high / impact medium-high.

10.4 Governance and Insiders: Net Selling of $369M, but Separate Three Things

This point is easily misread, so three things must be separated:

Insiders have indeed been one-way net sellers: Since listing, open-market insider sales total about $369M, while purchases total only about $0.25M, from two early low-price director purchases. The $369M is a transaction-by-transaction Form 4 cumulative figure, including the couple, GRAT/family trusts, and executives, not one single disclosed figure. The DeWittes sold about $313M under a 10b5-1 plan established on 2025-03-31. The CFO sold about $23.7M cumulatively. But note: the DeWittes' highest single sale price was only about $117. They did not sell at the $194 peak. The person who truly sold at $133-135 was director Michael Klein, at about $25M.
Sam Altman did not "cash out and run": Altman's stake fell from 8.2% to 4.8% because his Hydrazine fund made an in-kind distribution to outside LPs. He personally has no Form 4 open-market sale. He resigned as chairman on 2025-04-22, with the official reason being to eliminate conflicts of interest so OpenAI and others could negotiate power purchase agreements with Oklo. After resigning, he still held about 4.8% (8-K).
The 16.4 million-share "gift" in 2025-12 was family-trust movement, not selling. Do not misread it as a large-scale exit.
Overall: insider net selling is factual and large. Since insiders' cost basis is close to zero, any selling is a large cash-out. But much of it used 10b5-1 plans, and Altman was not personally selling. Probability medium / impact medium. It is a warning signal rather than proof of imminent collapse.

10.5 Customer Pipeline Quality: 14 GW Is Roughly a Marketing Narrative

The definition must be pinned down: the pipeline is about 14 GW, but most of it consists of non-binding LOIs or framework agreements that are revocable and generate zero cash today. Switch's 12 GW is explicitly a "non-binding" master power agreement. Equinix's 500 MW is an LOI plus a $25M prepayment. The only binding item is Meta's 1.2 GW from 2026-01, including a prepayment mechanism. Goldman pointed out in 2025-09 that Oklo had a 14 GW pipeline but no signed PPA. LOI-to-binding-PPA conversion rate, timing, customer concentration, and dependence on one AI data-center end market and AI capex cycle are all unproven. Probability medium / impact medium-high.

11. Catalysts and Tracking Metrics

Positive catalysts: 1. Groves isotope test reactor reaches criticality on 2026-07-04, the first DOE pilot milestone. 2. First commercial isotope revenue is realized, as soon as 2026. 3. LOIs in the 14 GW pipeline convert into binding PPAs. 4. NRC COLA substantive review progresses without major gaps. 5. AI power-demand data again exceed expectations.

Negative catalysts: 1. First-reactor timeline slips again, from 2028 to 2030. 2. A new round of discounted equity issuance. 3. Kerrisdale's fuel-economics critique is confirmed by later contracts. 4. TerraPower and other competitors remain ahead. 5. AI capex cycle cools or the "AI is getting more power-efficient" narrative returns.

Tracking dashboard, by importance:

First-reactor milestones: Groves criticality date and whether Aurora-INL commercial operation holds 2028.
Cumulative binding PPA GW: How far the number moves up from 1.2 GW with Meta is the hard indicator of pipeline quality.
Quarter-end cash plus new ATM use: Tests dilution speed and runway.
Fuel-contract $/kg: Tests whether Kerrisdale's fuel-economics critique is falsified.
Short interest, currently about 20-25% of float with days-to-cover of about 2.6: high short interest plus low days-to-cover can trigger short squeezes.
NRC COLA review status: Watch for renewed "information gaps" wording.

12. Zen Horizon Cross-Check Summary

12.1 Three Bull Points

The policy position is real and unique: DOE pilot three-project sweep plus DOE authorization path around the NRC bottleneck gives Oklo an institutional channel peers do not have in the race to build the first advanced reactor. Its first reactor broke ground in 2025-09.
The funding moat is real: $2.54B of cash, no debt, and no going-concern issue give Oklo one of the thickest funding cushions among peers and ammunition to keep advancing during narrative drawdowns.
The sector tailwind is real: AI data-center power shortages, nuclear renaissance, and Trump's nuclear executive orders are all blowing at once. The 14 GW pipeline plus Meta/Switch/Equinix endorsements show that demand-side interest is genuine.

12.2 Three Bear Points

Fundamentals and valuation are far apart: Zero revenue supports a $10.11B market cap, first power waits until 2028, about 75% of market cap is story value beyond cash, the margin of safety is zero, and the downside is permanent loss if option value goes to zero.
Fuel economics, regulation, and execution are all unverified: Kerrisdale's 5x fuel-cost critique hits the core of commercial feasibility. The 2022 NRC denial, first-of-a-kind status, and sodium-reactor engineering history all reduce confidence in "commercial operation in 2028."
The differentiated moat is being challenged: TerraPower, on the same route, already has an NRC construction permit and has begun construction, moving ahead. NuScale has the only certified design yet only about one-third of Oklo's market cap. Oklo's valuation premium lacks peer support.

12.3 Pre-Mortem: If the Stock Loses 50% in 3 Years, What Is the Script?

Script 1, fuel economics plus timeline double hit: In 2027, Oklo discloses initial HALEU procurement contracts, and the actual unit price comes close to Kerrisdale's $30,000/kg rather than $7,000/kg, making unit economics fail. At the same time, NRC commercial licensing drags into 2029 as the former commissioner suggested, pushing first-reactor commercial operation from 2028 to 2030+. Non-binding LOIs such as Switch gradually withdraw because of delays. The market reprices Oklo as "another NuScale-like cost-overrun story," and market cap falls from $10B back toward cash at $4-5B, with the stock at $25-30. A stock already down 70% from the peak would halve again.

Script 2, narrative fade plus dilution: The AI capex cycle cools in 2027, and the "AI power shortage" narrative fades, similar to the 2025-11 Gemini 3 episode but more persistent. To keep reactor construction moving, Oklo uses the $1.0B ATM plus additional financing, taking share count from 174M to 250M+. Under zero revenue and tighter liquidity, it is forced into deeply discounted issuance. Valuation and per-share value are hit together, sending the stock back to the SPAC-era $15-20 range.

12.4 Company Profile Score and Final Conclusion

【Company Profile Score】

Fundamental quality: low (zero revenue, cash burn, first power in 2028)
Growth: high (if delivered, 14 GW pipeline plus fuel recycling plus isotopes provide three curves)
Moat: medium (policy/funding position is real, but technology differentiation is being challenged by TerraPower)
Financial robustness: medium ($2.54B cash and no debt are strengths; survival still depends on continued dilution)
Management credibility: medium (MIT-trained team and strong policy positioning, but 2022 denial, repeated timeline movement, and limited commercial nuclear operating experience)
Valuation appeal: low (zero revenue supporting $10B, no margin of safety)
Risk level: high (multiple binary risks across regulation, execution, fuel economics, dilution, and competition)
Suitable investor type: high-risk speculation / thematic momentum traders; not suitable for value investors seeking margin of safety and not suitable for ordinary investors

【Investment Rating】Watch.

One-sentence investment thesis: The DOE pilot three-project sweep and $2.5B of cash are real positioning, but zero revenue supporting $10B, first power in 2028, and no margin of safety make this an option, not an investment.
Three price signals: Ideal buy <= $32, where EV is roughly close to cash plus a risk-adjusted pipeline option value and about -45% below the current price; holdable $48-76, the base scenario and where the current price sits; clearly overvalued > $140, above the upper end of the bull scenario.
Current price classification: Inside the holdable range, but with no margin of safety. For non-holders, it does not constitute a buy case.
Whether it is worth waiting for a better price: Yes. The triggers for buying would be a price back to <= $32, or at least two of the following three: first reactor reaches criticality, cumulative binding PPAs rise materially, and fuel economics falsify Kerrisdale. If those occur, the stock can be re-evaluated even at a higher price. The opportunity cost of waiting is missing a narrative-driven upside move, but for a zero-revenue name this is discipline.
Target holding period: 3-5 years or longer if purchased. This is a long-duration bet on first-reactor delivery.
Expected annualized return: Bear -45% to -69% (permanent loss) / base -17% to +31% / bull +90% to +141%. Upside and downside are severely asymmetric. Downside is permanent loss; upside is the compound probability of multiple milestones.
Maximum loss risk: Under the pre-mortem, the worst case is a return to $15-30, or another 50-74% loss from the current price, triggered by fuel economics being confirmed, timeline slippage, and narrative fade.
Signals that trigger reassessment: 1. First-reactor commercial-operation timeline slips from 2028 to 2030+. 2. Initial HALEU contract price is confirmed above $20,000/kg. 3. The company uses ATM issuance at a discount for two consecutive quarters and annual share-count growth exceeds 30%. 4. A major LOI withdrawal appears in the 14 GW pipeline. 5. Price falls back to <= $32, where the rating could be raised to "Cautious Buy."

Research Uncertainties (Known Blind Spots)

True fuel cost: Between Kerrisdale's $35,000/kg and Oklo's $7,000/kg, the real commercial unit price is unknown and depends on whether Oklo plus Centrus can scale. This is the largest single-point uncertainty in the entire valuation.
First-reactor timeline: Different sources use "late 2027 to early 2028" and "2028" differently, and criticality under the DOE authorization path is not the same as NRC commercial-licensed power sales. The actual connection between the two tracks is unknown.
FY2026/27/28 revenue consensus: Sell-side coverage is thin, and many analysts provide target prices without revenue lines. There is no concrete consensus revenue dollar figure to anchor on.
Peak market-cap definition: The all-time high of $193.84 was the intraday price on 2025-10-15, while the highest close was $174.14. The "about $30-31B" peak market cap depends on the share-count definition at the time and is not one precise single-authority number.
Class-action status: Multiple law firms began securities-fraud investigations in 2024-11, but whether any case was filed, dismissed, or settled has not been confirmed. Public information only shows the "investigation / plaintiff solicitation" stage.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.

Mentioned Tickers

SMR XE NNE LEU BWXT LTBR NVDA META EQIX

Baillie Framework · Ten Questions for Growth Investing

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

How high is its market ceiling? Is it enlarging an existing pie, or creating an entirely new market?6/10

Conclusion: Oklo's market ceiling is high, but it is not creating a new market for "electricity" out of thin air. It is trying to create a new form of supply inside the huge existing electricity and nuclear power market: replicable, small-scale 24/7 clean nuclear PPAs for large customers such as AI data centers. If Aurora can be built on schedule, at low cost, and in volume, the ceiling starts at a dozen-plus GW and can expand over the long term into a larger power infrastructure market. Today, however, it is still a pre-revenue option, not a proven market creator.

The demand side is genuinely large enough. AI and data centers are turning electricity from a back-office cost into a front-line bottleneck: the IEA's 2026 update expects global data-center electricity use to rise from about 485 TWh in 2025 to about 950 TWh in 2030, with AI data-center electricity use growing faster; DOE, citing EPRI, says U.S. data centers could account for up to about 9% of U.S. electricity generation by 2030, up from about 4% in 2023. This explains why customers such as Meta, Switch, and Equinix are willing to discuss "new nuclear": what they want is not ordinary renewable energy certificates, but clean baseload power that can supply electricity continuously.

But "the market is large" must be separated from "Oklo can capture it." Oklo's named pipeline is about 14GW, which sounds like a giant power sales platform; yet its largest piece, Switch's 12GW, is explicitly a non-binding Master Power Agreement, with individual binding PPAs to be converted only after project milestones. The firmer order is Meta: Oklo announced that the two sides support a 1.2GW nuclear energy campus in Ohio, under which Meta may prepay for power and provide early funding, with the first phase online as early as 2030 and expansion to 1.2GW by 2034. This shows that demand interest is real, but "14GW of saleable power assets" has not yet become real.

It is more like enlarging an existing pie and changing how that pie is sliced. Electricity, nuclear power, and data-center power supply are all existing markets. Oklo's novelty lies in build-own-operate: it builds, owns, and operates plants itself, selling power and heat through long-term PPAs rather than selling reactors. If this model works, it would expand advanced nuclear from "utilities procuring large nuclear plants" into a new procurement category in which "large customers use private capital to fund additional clean baseload." In other words, it is not creating entirely new end demand; it is moving nuclear power's serviceable scenarios from the traditional grid side toward AI data centers, industrial parks, defense, and remote energy-use settings.

So the judgment on this question is: the ceiling is high, even very high, but the market capture that can currently be proven is still very low. As of Q1 2026, Oklo still had zero revenue, and its first meaningful power-generation revenue will have to wait until 2028 or later; at the current $58.09 price and about $10.11B market cap, investors are mainly buying a long-dated option on "AI power shortage + nuclear revival + conversion of the 14GW pipeline." Its Q1 is not a case of "no market," but rather "the market is enormous, while the chain from interest to contracts, from contracts to plants, and from plants to profit remains long and unproven."

Jun 8, 2026
Can its revenue at least double over the next five years? Will growth mainly be driven by volume, price, or new businesses?4/10

Conclusion: If judged mathematically from a starting point of zero revenue, Oklo's revenue can of course very likely "double" over the next five years, because the current base is 0. But under the Baillie Gifford framework's real concern, which is whether the company can move from commercial validation to sustainable high growth, the answer needs to be far more cautious: revenue is likely to appear over the next five years, but whether it can form a continuously doubling curve depends on three things: the first reactor, isotopes, and PPA conversion, not the current 14GW pipeline number itself.

As of Q1 2026, Oklo was still in the pre-revenue stage. Its Q1 2026 10-Q showed a net loss of $33.1M, cash and marketable securities of $2.5369B, and primary uses of capital including powerhouses, radioisotopes, fuel businesses, and growth plans. So "doubling" here is not the problem of a mature company going from $100M to $200M; it is the problem of moving from 0 to first commercial revenue, and then whether that can be replicated at scale.

In the sequence of revenue drivers, the earliest may not be power sales, but a new business: radioisotopes. In its 10-Q, Oklo disclosed that Atomic Alchemy's Abundantia project is expected to generate revenue from sales of purified radium and other isotopes as early as 2026; Atomic Alchemy's Groves isotope test reactor has also received DOE approval for its nuclear safety design agreement and targets criticality on 2026-07-04. This means that if there is revenue in 2026-2027, it is more likely to be "new business start-up revenue." The amount may be small, but it is important for breaking the zero-revenue state.

Growth in the core power sales business will mainly be driven by volume, not price. Oklo's model is to build, own, and operate plants itself, selling power and heat through PPAs; the company also says in its 10-Q that its main product is energy generated by Aurora powerhouses once they are operating, sold directly to customers through PPAs. The price side may be locked in by long-term contracts and is unlikely to drive high growth through price increases the way software can. What really determines the revenue slope is how many MWe/GW can come online on schedule, how many LOIs can become binding PPAs, and how capacity factors and commercial operating reliability perform.

The demand side has imagination value, but it cannot yet be treated as certain revenue. Switch's 12GW agreement is a non-binding Master Power Agreement, with individual projects still needing to sign binding agreements after milestones; Meta's 1.2GW southern Ohio project is firmer, with a prepayment mechanism and project funding support, but the company also disclosed that the first phase could come online as early as 2030 and gradually expand to 1.2GW by 2034. This helps the five-year view, but it is not immediate large-scale recognition of power sales revenue in 2026-2028.

So my judgment is: Oklo has a decent probability of revenue going "from 0 to something" over the next five years, making a nominal doubling easy; but high-quality doubling mainly depends on volume delivery, namely Aurora's first commercial operation, replication of more units, and LOIs converting into binding PPAs. The first short-term push comes from the new isotope business, while price is not the main driver. The biggest risk is that if first power slips from 2028 to 2030, or fuel costs or regulation create further problems, five-year revenue may still remain at the level of small isotope revenue and prepayments, far from a validated growth curve.

Jun 8, 2026
After five years, what will take over as the next growth engine? Does this "second curve" exist today?4/10

Conclusion: If Aurora's first reactor and data-center power sales can work five years from now, the growth engine that takes over should not merely be more small reactors, but "closed-loop nuclear fuel + isotopes". Today, however, it is more like the seed of a second curve, not a curve already validated by revenue.

The first line is fuel recycling. Oklo has announced the construction of a private nuclear fuel recycling facility of up to $1.68B in Oak Ridge. The logic is to turn spent fuel into usable fuel for Aurora, ease the HALEU bottleneck, and move itself from a "nuclear plant developer" toward a "fuel-chain controller." In 2026-05, it was also selected for the DOE surplus plutonium utilization program negotiation list, with the goal of turning U.S. surplus plutonium into advanced reactor fuel. If this succeeds, Oklo's value capture would expand from selling electricity to the fuel supply chain, and its moat would be deeper than simply building reactors.

The second line is radioisotopes. Oklo entered isotopes through the acquisition of Atomic Alchemy; the 10-Q cited in the report shows consideration of about $28.4M. This segment may generate small revenue earlier than Aurora power sales, because the Groves isotope test reactor targets criticality on 2026-07-04, and the Idaho radiochemistry laboratory is also advancing commercial contracts. Its advantage is that revenue timing may be earlier, and its applications extend beyond electricity to medical isotopes, space RTGs, and more; its weakness is that the gap between market size and Oklo's current $10B market cap is too large, so it cannot support the main narrative in the short term.

So the answer is: the second curve exists today, but only as "projects and options," not yet as a "business curve." Oklo already has the foundation of a Meta-backed advanced nuclear project of up to 1.2GW, fuel recycling plant plans, an entry into DOE fuel projects, and Atomic Alchemy isotope assets; these are real positions. At the same time, the company is still at zero revenue, with a Q1 2026 net loss of $33.1M, cash and marketable securities of $2.54B, and first power from the core Aurora still waiting until 2028. In other words, the second curve is already at the foundation stage, but it will take the first reactor, fuel costs, regulatory approvals, and the first batch of commercial revenue together to prove that it has truly become a growth engine rather than a story.

Jun 8, 2026
What is its core competitive advantage? Will this moat widen or narrow over the next three to five years?4/10

Conclusion: Oklo's core competitive advantage is not a cost advantage that has already been realized, but an early-mover window formed by policy access + fuel-chain positioning + a build-own-operate power sales model + ample cash. Together with Atomic Alchemy, it obtained three DOE Reactor Pilot Program project slots, with Aurora, Groves, and Pluto corresponding respectively to power plants, an isotope test reactor, and the plutonium fuel direction. This program aims to accelerate advanced reactor testing and criticality validation through DOE authorization pathways, and it does give Oklo policy positioning. The company's own 10-Q also states the business model clearly: Oklo does not sell reactors, but plans to build, own, and operate Aurora powerhouses itself, sell power and heat directly through PPAs, and treat fuel recycling as part of vertical integration of the fuel supply chain.

But this moat has not yet truly become a commercial moat. The reason is that the nuclear industry ultimately competes not on stories, but on licenses, fuel, construction, costs, and operating records. Oklo's Aurora combined license application in 2022 was "denied without prejudice" by the NRC. At the time, the NRC clearly said the issues involved information gaps in key areas such as potential accident descriptions, safety systems, and component classifications (NRC press release); this means its regulatory capability cannot be treated as "proven." The customer side is similar: the named pipeline is large, but much of it is still LOIs/framework agreements; Meta's 1.2GW is more concrete, but it has not yet become a portfolio of financed, built, and stably power-selling projects.

Over the next three to five years, my base-case view is: the moat will first be compressed, and will widen only after key milestones are delivered. Compression comes from two directions. First, the policy fast lane is not exclusive to Oklo. TerraPower, backed by Gates, also follows a sodium-cooled fast reactor route and had already received an NRC construction permit and started work in 2026, overtaking Oklo in progress; this weakens Oklo's narrative of being the sole advanced nuclear first mover. Second, fuel economics remain under market attack. Kerrisdale's short report questions whether Oklo's HALEU cost assumptions may be materially underestimated; that view cannot be taken directly as fact, but it does hit the most fragile part of Oklo's moat: if fuel costs and first-reactor capex run out of control, the supposedly low-cost, replicable Aurora model becomes hard to sustain.

So the more accurate judgment is: Oklo currently has a "to-be-validated regulatory/fuel/capital moat," not an already deep operating moat. If Groves or Aurora-INL advances on schedule in 2026-2028, PPAs convert into more binding contracts, HALEU/recycled-fuel costs are proven through physical delivery, and first-reactor capex does not run out of control, this moat can upgrade from a policy window into a moat of operating data and financing capability. Conversely, if the first reactor is delayed, the NRC commercial pathway gets stuck again, PPAs do not convert, or fuel costs are confirmed to be high, it will shift from "first-mover advantage" into "execution liability under a high valuation." For now, it should be viewed as unproven and under pressure.

Jun 8, 2026
If its core business is disrupted, does it have the genes for self-reinvention? How does it handle mistakes and bad news?4/10

Conclusion: Oklo has some embryo of self-reinvention, but it has not yet been proven by commercial results. Its strength is not the flexibility to "switch tracks if the product does not sell," but the ability to keep changing entry points within the very narrow nuclear track: from early UPower/remote power microreactors to the Aurora self-built, self-operated power sales model for data centers; from a single power-generation story to the HALEU fuel chain, spent-fuel recycling, and radioisotopes. For example, its 2025 acquisition of Atomic Alchemy may allow the isotope business to generate revenue earlier than power generation; it also obtained EBR-II recycled HALEU through INL, advanced fuel-service cooperation with Centrus, and planned an Oak Ridge spent-fuel recycling facility. This shows that the team is not betting on one isolated product, but making adjacent expansions around "advanced nuclear fuel + small reactors + specialized power/isotopes."

But the quality of this reinvention must be discounted. All of Oklo's new directions still depend heavily on the same set of key variables: nuclear regulation, HALEU supply, nuclear engineering execution, capital-market financing, and customers' willingness to sign long-term power purchases. If the core Aurora power sales model is disrupted, for example if the NRC's 2022 rejection of its application due to insufficient maximum credible accident methodology and safety classification information remains a problem, Kerrisdale's challenge to HALEU costs of $7,000/kg vs $35,000/kg is confirmed, or TerraPower on a similar route continues to lead, isotopes and fuel recycling may not be enough to support the current 14GW data-center power sales narrative. They look more like second and third options than cash-flow engines already able to replace the core business.

Its handling of mistakes and bad news is "tactically capable of rerouting, but strategically not yet self-proving." On the positive side, after the NRC rejection in 2022, the company did not stall. It continued to pursue the DOE authorization pathway and secured three DOE Reactor Pilot Program project slots: Aurora, Groves, and Pluto. This reflects an ability to reconstruct the regulatory path. It also did not rely only on the power-plant story, but added fuel, isotopes, and recycling, trying to turn weak points into vertical positions.

The negative side is that bad news has not been truly "closed." The NRC commercial licensing issue is still not closed; using the DOE track for the first reactor does not mean commercial power sales licensing risk has disappeared. Q1 2026 still showed zero revenue, a net loss of $33.1M, and cash and marketable securities of $2.54B, meaning the company is mostly using ample cash to buy time. The new $1.0B ATM and ongoing dilution show that its response to financing risk is still to sell stock to extend life, not to repair itself through operating cash flow. Insider net selling of about $369M also weakens the sense that insiders are sharing bad news over the long term.

So Oklo's self-reinvention capability can be summarized as follows: the engineering team has the ability to expand adjacently and reroute through regulatory channels, and the capital market has given it the cash time needed for reinvention, but it has not yet proven that it can turn a major mistake into a profitable new business. Under the Baillie Gifford framework, this is not a rigid company, but neither is it an excellent company that has already demonstrated antifragile evolution.

Jun 8, 2026
Does management, especially the founders, have a long-term view and deeply aligned interests with the company? Are they willing to sacrifice current profits for five to ten years from now?5/10

Conclusion: Oklo's management clearly has long-termism and technical founder alignment, but this is not a case of "flawless founder-shareholders." Jacob DeWitte and Caroline DeWitte have taken the company from 2013 all the way to the DOE pilot, SPAC listing, and first-reactor progress, and both have nuclear engineering backgrounds. The SEC proxy shows Jacob as CEO/chairman and Caroline as COO/director. The two still disclose about 21.16M shares and 12.16% overlapping beneficial ownership through personal holdings, GRATs, and family trusts. This figure should not be simply added up, but it is enough to show substantial economic exposure (2026 DEF 14A). This kind of alignment is stronger than in an ordinary professional-manager company, and it also explains why the company is willing to pursue a decade-long build-own-operate wager.

There is evidence of a long-term view. MIT's early coverage of Oklo shows that DeWitte framed the problem early as "redoing nuclear commercialization in a startup way" and emphasized full-stack innovation across technology, regulation, manufacturing, business model, and financing model (MIT News). The company still chooses to build, own, and operate plants itself, rather than sell reactors and collect one-time license fees. In substance, it is sacrificing short-term revenue recognition in exchange for future 20-year PPA duration cash flows.

But the phrase "willing to sacrifice current profits" needs to be understood with a discount in Oklo's case: it has no profits to sacrifice today; it is a pre-revenue company burning capital. In Q1 2026, the company still had zero revenue, a net loss of $33.1M, R&D expense of $27.0M, G&A of $24.2M, SBC of $15.6M, and cash and marketable securities on the balance sheet that came mainly from financing rather than operations (Q1 2026 10-Q). This shows that the team is indeed directing resources toward reactors, fuel, and isotope businesses years in the future, but the cost is mainly borne by outside shareholders through dilution.

The governance deductions are also clear. The report's item-by-item accumulation of Form 4 filings indicates insider net selling of about $369M since listing, with about $313M by the DeWitte couple; SEC filings also confirm that Jacob and Caroline each established 10b5-1 trading plans on 2025-03-31, under which each may sell up to 600,000 shares (Q1 2025 10-Q). This does not mean an improvised exit, and they still retain large holdings, but it materially discounts the quality of "deep alignment." Sam Altman stepped down as chairman in 2025-04, with Jacob taking over. The company said this was to continue exploring strategic collaborations with AI companies, including potential OpenAI collaboration (Oklo announcement); this may remove conflicts for the business, but it also weakens the market's perception of Altman's direct governance endorsement.

So the Q6 answer for Oklo is: the founders have a strong long-term view, real technical mission, and real remaining equity alignment, and they are willing to commit years of cash flow to the first reactor after 2028, fuel recycling, and isotope businesses; but one-way insider net selling, ongoing ATM dilution, Altman's board exit, and the lack of a proven commercial nuclear operating record make it more like a "highly aligned but highly controversial founder team," not yet something that can be treated as a top-tier owner-operator.

Jun 8, 2026
If it disappeared tomorrow, how much would customers miss it? Is its growth sustainable and not dependent on harming society or regulation?4/10

Conclusion: If Oklo disappeared tomorrow, customers would regret losing a high-risk but scarce "future clean baseload power option," but they would not immediately feel an operating outage-level pain. The reason is straightforward: Oklo still has zero revenue and no commercially operating reactor, and core Aurora power sales will have to wait until 2028 or later. What customers have today is mainly the possibility of future PPAs, not electricity they are already using.

What customers would miss is the supply format, not an existing service. AI data centers genuinely lack stable, low-carbon, 24/7 power at scale. DOE, citing EPRI, says the share of U.S. data-center electricity consumption could rise from about 4% in 2023 to up to about 9% by 2030 (DOE). Oklo's appeal is precisely to build, own, and operate plants itself and sell power and heat directly to customers (Q1 2026 10-Q). If it works, customers receive long-term clean baseload power rather than intermittent renewable energy certificates.

But customer dependence is still very low today. Switch's 12GW agreement is explicitly a non-binding master power agreement, with specific projects still requiring later milestones and binding agreements (Oklo announcement); Meta's 1.2GW is more concrete, but the first phase is also not expected to come online until 2030 and then expand gradually (Oklo/Meta announcement). So if Oklo disappeared tomorrow, Meta, Switch, and Equinix would lose a potential supplier and bargaining option, but they would not immediately halt operations the way they would if they lost AWS, TSMC, or a critical grid node.

The social value of its growth model is positive: if Aurora's cost, safety, and fuel loop are proven, it increases clean baseload supply for data centers, industrial users, and defense settings, and in theory does not rely on harming users, squeezing labor, or exploiting regulatory arbitrage to grow. It is also trying to fill gaps in the nuclear fuel chain through HALEU, fuel recycling, and isotope businesses, which has more system value than simply saying "build more reactors."

The real issue is: its sustainability must be proven by regulation and physical engineering, not by a narrative of bypassing regulation. Oklo was "denied without prejudice" by the NRC in 2022 for reasons involving maximum credible accident methodology and insufficient safety classification information (NRC). Taking the first reactor through a DOE authorization pathway can accelerate the pilot, but it does not mean commercial licensing, safety case, fuel economics, and public acceptance are all closed. If nuclear growth is to be sustainable over the long term, it has to respect regulation more than ordinary industries do, not treat regulation as an obstacle merely to be routed around.

So the judgment on this question is: Oklo's "future missability" for customers may be high because 24/7 clean electricity is truly scarce; but its "indispensability today" is low because it has not yet delivered commercial power. Its growth direction is not inherently harmful to society and may even have public value; but only after safety permits, HALEU costs, spent-fuel handling, and first-reactor operating data are empirically proven can this growth be called truly sustainable.

Jun 8, 2026
What are the unit economics of this business, such as gross margin and incremental returns? Do they improve or worsen with scale? Where does the money it earns go?2/10

Conclusion: Oklo does not yet have verifiable unit economics. It is not a company with "high gross margin that the market has failed to notice," but a heavy-asset option where "durable cash flows may exist once built, but a large amount of capital must be burned before every dollar of revenue." Q1 2026 still showed revenue of $0, a net loss of $33.1M, and operating expenses of $51.2M, so gross margin, mature plant IRR, and incremental ROIC have no actual record yet. What can be quantified is only cash consumption: operating cash outflow of $17.9M in the same quarter, capex for deployed facilities of $32.8M, and company expectations of $80-100M of operating cash use and $350-450M of investing cash use in 2026.

The theoretical good business lies in this: Oklo does not sell reactors, but uses build-own-operate to sell power and heat to customers through PPAs, a model the company says can create recurring revenue. If Aurora can be replicated in standardized form, capacity factors are high, PPA prices are locked in, and fuel costs are controllable, then mature plants would look like infrastructure assets: large upfront capital expenditure, followed by recovery through 20-year power sales cash flows, with larger scale potentially improving unit costs through engineering learning curves, supply-chain bargaining power, and fuel recycling.

But the more important point today is the other side: its incremental returns are extremely sensitive to several variables, especially capex/kW, HALEU fuel prices, first-reactor delays, financing costs, and actual PPA power prices. Kerrisdale's short calculation argues that if HALEU is not the $7,000/kg assumed by Oklo but closer to $35,000/kg, the overnight cost and LCOE of a 50 MWe plant would be revised materially upward. That view may not ultimately be correct, but it hits the vital point in Oklo's unit economics. In other words, Oklo's profitability is not "not yet released"; it is "not yet proven to exist."

Whether unit economics improve or worsen with scale depends on what it proves first. If the first reactor arrives on schedule, fuel costs decline, PPAs convert, and financing costs are controlled, scale can spread FOAK costs and improve unit economics; but if it expands before validation, scale will only amplify capital expenditure and dilution pressure. The company has $2.54B of cash and marketable securities and no long-term debt, which is an advantage; but the newly established ATM equity program of up to $1.0B shows that "using shareholder capital first to build assets" remains a core funding source, not self-rolling through operating cash flow.

There is not much to say yet about where the money it earns goes, because it has not earned operating revenue. At this stage, funds mainly go to three areas: R&D and engineering teams, advancement of the Aurora/isotope/fuel businesses, and construction of heavy-asset facilities. Q1 2026 R&D was $27.0M, G&A was $24.2M, and stock-based compensation was $15.6M; it will also need to invest in powerhouses, radioisotopes, fuel businesses, and growth plans. Fuel recycling is especially capital intensive: the company has announced an Oak Ridge private nuclear fuel recycling facility with investment of up to $1.68B. So if this business succeeds, the money it earns will most likely be rolled into new plants, the fuel chain, and isotope capacity rather than returned to shareholders quickly.

Overall, Oklo's unit economics can only be described as "to be validated" for now: the upside is a long-duration, replicable nuclear infrastructure platform with strong cash flow; the downside is a heavy-asset developer that burns more money as it scales after fuel costs and capex run out of control. The current evidence is closer to the early stage of the latter than to a compounding machine that has already proven high incremental returns.

Jun 8, 2026
What conditions must all hold for it to rise fivefold in ten years? Are those conditions realistic? What expectations are implied by today's share price?2/10

Conclusion: Using this question's unified reference price, the 2026-06-05 closing price of $58.09, a fivefold rise over ten years means about $290/share; based roughly on a market cap of about $10.11B, that means $50B+ of equity value. But given that Oklo still has zero revenue and will need to rely on ATM/project financing to keep building reactors, if share count keeps diluting, the market cap truly needed ten years from now may not be $50B, but $60B or even higher. This outcome requires more than "the first reactor generates power"; it requires Oklo to move from a pre-revenue option company into a multi-GW nuclear asset operator.

For the stock to rise fivefold, at least several things need to happen at the same time. First, Aurora's first reactor must reach substantive commercial operation around 2028, and the denied without prejudice decision made by the NRC in 2022 due to information gaps must not become a long-term licensing overhang; the DOE authorization path can help it run ahead, but it cannot replace the regulatory credibility needed for large-scale commercial replication. Second, HALEU fuel, first-reactor capex, and O&M costs must prove that Oklo's power sales cost can be attractive to data-center customers; if the fuel-cost view challenged by Kerrisdale is close to reality, the commercial model would be directly compressed. Third, the 14GW pipeline must move from story to contract: Meta's 1.2GW project needs to advance, and Switch's 12GW non-binding master agreement also needs to convert in batches into binding PPAs rather than remaining a marketing pipeline.

Fourth, financing must not dilute away shareholder returns. In its Q1 2026 10-Q, Oklo still had zero revenue, a net loss of $33.1M, and cash and marketable securities of about $2.54B; this money provides a 4-5 year runway including capex, but build-own-operate is a heavy-asset model and will require further financing. The company also set up an ATM program of up to $1.0B in 2026-05, so a fivefold share price requires not only project success, but also financing costs, equity dilution, and project returns staying under control.

These conditions are not completely unrealistic, but they form a multi-delivery scenario with little room for error. Oklo has real strengths: DOE pilot positioning, cash ammunition, AI data-center power demand, and customer anchors such as Meta. But a fivefold return requires it to clear six gates at once: regulation, fuel, engineering, customer conversion, capital expenditure, and dilution. If any one of these slips, valuation may be marked down from "nuclear platform" back to "high-risk developer."

Today's $58.09 price already implies quite a few good outcomes: the market is roughly betting that the 2028 first reactor can advance, some PPAs will convert into binding contracts, fuel costs will not completely disprove commercial viability, and financing dilution will remain manageable. It does not price in full conversion of all 14GW as a certainty, otherwise the share price would be closer to or above the optimistic range; but it is also not a cheap failure option, because most of the roughly $10B market cap is already future story value beyond cash. In other words, a fivefold return requires a "super-bull case," while today's share price has already prepaid for a "neutral-to-optimistic case."

Jun 8, 2026
Why has the market not realized all this yet? Is it that investors do not understand it, look down on it, or cannot look far enough? What will become the "narrative inflection point"?2/10

Conclusion: OKLO is not a case where "the market has not realized it yet." Rather, the market has already realized more than half of it and is demanding physical evidence for the remaining half. AI power shortages, nuclear revival, the triple DOE pilot win, the Meta/Switch pipeline, the fuel loop, and the isotope second curve are not obscure secrets. The real disagreement is whether these narratives can move from policy access and letters of intent into financeable, buildable, power-selling cash-flow assets with working unit economics.

So it is not a typical "looked down on" case. As of the report's reference date, 2026-06-05, OKLO closed at $58.09 with a market cap of about $10.11B, while the company still had zero revenue; Q1 2026 filings showed cash and marketable securities of about $2.54B, a net loss of $33.1M, and operating cash outflow of $17.9M (Q1 2026 10-Q). In other words, the market has already given it a market cap of roughly 4 times cash and about 75% story value beyond cash. A company that is truly looked down on would not carry a ten-billion-dollar market cap at the zero-revenue stage.

Nor is it simply "not understood." The market understands what it is selling: not reactors, but build-own-operate, with future 24/7 clean power sold through PPAs. The market also understands that demand is attractive: Meta's 1.2GW Ohio project does provide a prepayment and early development funding mechanism, but the first phase is still only targeted to come online in 2030, with expansion to 1.2GW by 2034 (Oklo-Meta announcement); Switch's 12GW is even more explicitly described as a non-binding Master Power Agreement (Oklo-Switch announcement). So the market is not failing to understand 14GW; it is unwilling to treat all 14GW as locked-in revenue.

The more accurate formulation is that the market "can look far, but does not dare to draw a straight line that far." The reason is that every link in OKLO's key chain remains unclosed. On regulation, the NRC in 2022 denied the Aurora application without prejudice due to insufficient key information (NRC announcement); on fuel, Kerrisdale's challenge of HALEU costs at $7,000/kg vs $35,000/kg hits LCOE directly; on capital, the company set up an ATM equity financing plan of up to $1.0B in 2026-05 (Orrick announcement); on competition, TerraPower on a similar route is already ahead in permitting and construction timing. The market's skepticism is not short-termism; it is asking Oklo to prove that this is not another story of "nuclear is beautiful, costs run out of control, schedules slip, and shareholders get diluted."

Positive narrative inflection points will be very concrete, not another sweeping vision. First, the Groves isotope test reactor reaches criticality on schedule and brings the first small commercial revenue, breaking the zero-revenue state first. Second, Aurora-INL clears successive gates in fuel, construction, and DOE/NRC coordination, turning 2028 first power from a slogan into engineering progress. Third, more binding PPAs appear beyond Meta, rather than LOIs or non-binding frameworks. Fourth, HALEU procurement, recycled fuel, or Centrus cooperation provides real costs that can suppress Kerrisdale's extreme short case. Fifth, financing no longer depends on discounted ATM survival funding, with project-level capital, customer prepayments, or government loans bearing more of the capital expenditure.

Negative inflection points are just as clear: the first reactor slips from 2028 toward 2030; first fuel contracts show costs close to the short-seller view; pipelines such as Switch/Equinix fail to convert into binding agreements; continued ATM use keeps expanding the share count; or AI data-center electricity demand cools while competitors such as TerraPower keep running ahead. At that point, the market would reprice OKLO from an "AI nuclear option" into a "high-dilution, long-cycle nuclear developer with unproven unit economics."

Therefore, the answer to question 10 is: the market is not unaware of OKLO's upside; it has already paid a fairly high option premium for that upside. What is missing is not comprehension, but verifiable milestones. The true narrative inflection point will not come from "everyone finally believing nuclear power is important," but from "Oklo proving that it can turn policy positioning, customer interest, and fuel-chain layout into financeable, replicable, profitable power assets."

Jun 8, 2026

Ask about this report

Members can ask about this report; once answered it appears under "Reader Q&A" on this page. You can also highlight a passage in the text to ask about it directly.