Oklo's reactor milestone puts Aurora's economics in focus

The thesis

Oklo (OKLO $34.70 +0.38%) plans to build, own and operate nuclear power plants, selling electricity rather than handing customers a reactor. That model offers long-lived revenue if plants work economically. It also requires Oklo to arrange construction capital before those revenues arrive.

The Groves reactor's first criticality improves the company's credibility as a builder. The more difficult investment question is whether its commercial Aurora plants can earn enough, after construction and financing costs, to reward common shareholders. Those are separate propositions: a team can deliver a technically successful plant without creating an attractive return on the capital invested.

The strongest bull argument is that completed work, customer funding and a substantial cash balance can reinforce one another, making each subsequent project easier to deliver. The strongest bear argument is that the costs and financing required to reach that fleet remain unknown.

Groves changes the execution argument

On August 5, the Groves reactor achieved first criticality, as reported in the Q2 10-Q. Oklo's August 6 announcement describes a low-power isotope test reactor built under the Department of Energy's Reactor Pilot Program on private land. It moved from groundbreaking to a controlled, self-sustaining chain reaction in less than a year.

Management's Q2 earnings call adds useful context. Chief executive Jake DeWitte described full-scale civil construction, not a laboratory assembly. The team had to coordinate procurement, excavation, safety documentation, fuel loading and startup. Completing that sequence is stronger evidence of execution than a design or construction schedule alone.

The distinction is what the project tested. Groves is a low-power isotope test reactor, not a 75-megawatt Aurora powerhouse. It does not supply grid electricity or demonstrate a commercial power-conversion system, capacity factor or construction cost per kilowatt. The reasonable inference is narrower: Oklo has gained construction and startup experience that may help later projects. How much of that experience transfers to Aurora remains to be demonstrated.

Aurora's licensing path remains the decisive bridge

The first planned commercial unit, Aurora-INL in Idaho, is moving through a DOE route. The DOE approved its Nuclear Safety Design Agreement in March, and Oklo announced on June 11 that the agency had approved its Preliminary Documented Safety Analysis. Those documents complete the first two of five steps Oklo identifies in that authorization pathway. Management still targeted 2028 for go-live on the Q2 call and attributed part of its higher 2026 capital budget to long-lead procurement and grid-interconnection work intended to protect that schedule.

For wider deployment, the Nuclear Regulatory Commission describes the current Aurora engagement as pre-application work for liquid-metal-cooled, metal-fueled fast reactors of up to 75 megawatts electric. Oklo announced approval of its Principal Design Criteria topical report in May, allowing future applications to reference that safety framework. This is progress toward licensing, not a reactor operating license.

The history matters: the NRC denied an earlier Aurora application without prejudice in January 2022 because required information was missing. That decision was not a permanent prohibition on the design. It does explain why completed reviews deserve more weight than an expected licensing date.

A successful Aurora-INL project could supply operating evidence for later applications. It would not automatically authorize the Ohio campus or other customer projects. Investors should therefore track two distinct developments: completing the first DOE-authorized plant and establishing a repeatable NRC licensing route for broader deployment.

The 2028 target is management's schedule, not an assured revenue date. Its credibility depends on equipment delivery, fuel fabrication, grid connection and the remaining safety reviews progressing together.

Demand is real, but 13.2 gigawatts is not contracted backlog

Oklo's two largest named demand signals are of very different quality. In January, the company signed a prepayment agreement with Meta supporting a planned 1.2-gigawatt campus in Pike County, Ohio. Meta's funding is intended for fuel procurement and Phase 1 development; site work was slated to begin in 2026, with the first phase targeted as early as 2030 and the full campus by 2034.

The agreement provides a way for customer capital to arrive before electricity is delivered. That can reduce Oklo's early funding burden. The announcement does not disclose the funding amount, payment schedule, power price or project return. A funding commitment should therefore not be treated as cash already received or unrestricted money available for unrelated projects.

The 12-gigawatt Switch announcement is broader and less firm. Oklo's own December 2024 release calls the Master Power Agreement non-binding and says individual binding agreements are expected as project milestones are reached. The framework runs through 2044, which makes it evidence of long-term demand but not a commitment to buy 12 gigawatts on fixed terms.

The two announcements sum to 13.2 gigawatts, but that arithmetic is not a backlog calculation. It combines a specified campus with a much broader framework and says nothing about the probability, timing or profitability of individual plants.

Meta's willingness to support development is nevertheless important: a customer can help fund a project before it produces electricity. The next commercial test is whether that early commitment develops into financeable power terms. Price escalation, minimum purchases, termination rights and responsibility for cost overruns will matter more to shareholder returns than another aggregate capacity announcement.

More than $3 billion of liquidity came with a larger denominator

The June 2026 10-Q reported $1.645 billion of cash and equivalents, $820 million of current marketable debt securities and $541 million of non-current marketable debt securities. Together, those balances were $3.006 billion. This excludes $16.9 million of separately reported restricted cash. Total liabilities were $84.3 million; the balance sheet included $700,000 of long-term debt, so describing Oklo as entirely debt-free would be inaccurate.

That balance provides substantial near-term funding flexibility. On the Q2 call, management guided to $120 million–$150 million of operating cash use and $400 million–$500 million of property-and-equipment spending in 2026. Comparing liquidity with one year's guidance is not a fleet-funding estimate: future construction, fuel facilities and acquisitions can change the spending rate substantially.

The financing source is central to the per-share thesis. Shares outstanding rose from 160.5 million at December 31, 2025 to 185.1 million at June 30, an increase of 15.3% in six months. During the first half, Oklo sold 23.1 million shares through two at-the-market programs for $1.852 billion of net proceeds. Most of the share-count increase came from those offerings; options, awards and other equity activity account for the difference.

The May ATM continued after quarter-end and was fully used by September. Oklo's September 11 filing says that program ultimately sold 17,971,448 shares for approximately $1.0 billion of gross proceeds. Of that total, 10,712,054 shares and $680.371 million of gross proceeds were already reported through June. The subsequent portion was therefore approximately 7.26 million shares and $319.6 million gross. Adding the whole $1 billion to June liquidity would double-count most of it and ignore later spending.

Oklo then opened a new ATM for up to another $1.0 billion. An ATM, or at-the-market offering, lets a company sell shares over time rather than in one fixed-price transaction. The authorization is neither a completed sale nor cash already in the bank.

The distinction matters for both camps. Bulls cannot count an unused offering as funded cash; bears cannot count every new share as value destroyed. New investors contribute capital, and what management earns on that capital determines the eventual per-share outcome.

Early revenue is not a proxy for the power business

The Q2 10-Q reports $1.21 million of revenue, its only revenue in the first half. Revenue came from engineering and consulting, manufacturing and fabrication, and other activities. None was Aurora electricity revenue.

Q2 operating loss was $73.2 million and net loss was $48.5 million. First-half operating cash use was $65.5 million. These figures describe the cost of building the platform, not a steady-state margin for its proposed plants.

Manufacturing, fuel and isotope operations may eventually add revenue or protect construction schedules. They also compete for capital. As those businesses grow, separate disclosure of their spending and returns will help investors distinguish useful integration from an expanding collection of funding obligations. A revenue multiple based on today's small service business cannot resolve that question.

What the financing math can—and cannot—tell us

The share count must be updated before attaching a valuation to a stock price. June's 185.09 million shares plus the subsequently disclosed May-program issuance produces a mechanical bridge to 192.35 million shares. This is not a current or fully diluted share count: it excludes other post-June equity changes, options, awards and any sales under the new ATM.

At an illustrative $40 share price, that bridged count implies $7.69 billion of basic equity value. The example is a sensitivity, not a quoted market capitalization or price target. June liquidity cannot simply be subtracted to produce a September enterprise value, because subsequent financing and spending changed the cash balance.

The following example shows what another $1 billion would mean at $40 per share.

This assumes the new program is fully sold at one price and ignores other share changes. Holding the $40 price constant makes the added equity value equal the gross cash raised; a price decline is not mechanically required. What matters afterward is how effectively Oklo converts that capital into project returns.

That is financing arithmetic, not evidence that $40 is an attractive purchase price. A business valuation would need to connect the equity value to cash flows shareholders can eventually receive. It cannot stop at counting megawatts or listing milestones.

For each project, that means estimating electricity sales, operating and fuel costs, construction spending, financing obligations and Oklo's retained ownership. Cash flows must then be adjusted for when they arrive and the chance that the project never reaches operation. Corporate overhead and future parent-level funding belong in the calculation too. Customer prepayments can reduce the initial funding burden, but their contractual obligations must also be reflected; they are not free additional value.

The most important missing input is the first plant's full economics. Management said on the Q2 call that it was still narrowing Aurora-INL's cost estimate with Kiewit rather than providing a complete cost guide. Without a credible budget, power terms and funding structure, a numerical price target would depend heavily on assumptions the disclosures do not yet establish.

Milestones remain useful because they can improve those assumptions. They are not a substitute for valuation. The present evidence supports a stronger execution assessment; it does not, by itself, show that the market has underpriced the resulting business.

The bull–bear divide is the financing of the next plants

The bullish outcome is a learning curve that investors can observe. A completed Aurora would give lenders and customers operating evidence, while reusable engineering, supplier relationships and trained personnel could make subsequent plants less costly or uncertain. Customer funding and project-level capital could then carry more of the expansion, allowing Oklo to retain meaningful economics without repeatedly issuing large amounts of parent equity.

That outcome does not require dilution to disappear. It requires each financing round to fund assets whose value exceeds the capital committed, with enough of that value retained for common shareholders. A growing share count can coexist with a better investment if the underlying business improves sufficiently.

The bearish outcome is a company that makes technical progress but cannot repeat projects on attractive terms. First-of-a-kind overruns may strain a budget before lenders will finance the next plant. Customers may demand protections that shift construction risk back to Oklo. Raising project capital may also require sharing more ownership or cash flow than the headline capacity suggests.

Fuel remains a separate constraint. The June Centrus announcement was a letter of intent for future fuel supply, anticipating a definitive agreement. It was not delivered fuel. Supply arrangements, fabrication capacity and authorization must fit each project's commissioning schedule; pursuing several possible sources does not ensure any one arrives on time.

Delays compound these risks because they can increase costs while pushing revenue further out. The key distinction is not optimism versus pessimism about nuclear power. It is whether the eventual plants can repay their capital providers and leave attractive returns for Oklo's shareholders.

What would change the assessment

1. A costed, financeable first plant. An Aurora-INL budget, funding structure and credible operating assumptions would enable a project-return analysis. Cost escalation without corresponding commercial protection would weaken it.
2. A schedule supported by completed work. Safety approvals, equipment delivery, fuel readiness and grid connection should support the 2028 target. Slippage across those dependencies would matter more than an unchanged headline date.
3. Contracts that define returns. Binding power terms and customer contributions would improve the assessment; additional non-binding capacity would not provide the same evidence.
4. A repeatable deployment model. Progress through NRC licensing, clearer costs for later plants and funding that preserves Oklo's economics would support a fleet valuation. Persistent reliance on parent equity without those improvements would undermine it.

The conclusion is constructive on execution and unresolved on valuation. Groves deserves credit for completed work. The next major investment evidence should be a credible relationship between Aurora's construction cost, contracted revenue and funding—not simply another reactor milestone.

Until that relationship is clearer, the case for owning OKLO rests on expectations about future project economics, not demonstrated power-business returns. That may appeal to investors prepared to underwrite substantial development risk, but better execution alone does not establish a margin of safety or justify a buy recommendation. The most useful next disclosure would make those returns calculable.

---

Research date: September 30, 2026. Reported figures come from the linked SEC filings; project and contract descriptions come from the linked company and regulator disclosures. Management's expectations remain forecasts. The $40 financing example is an illustrative calculation, not a price target. This article is for information only and is not investment advice.