Summary

  • Oklo says it completed the ATM programme launched in May by selling 17,971,448 shares for approximately $1 billion gross. That produces a derived average gross price of about $55.64 a share; it is an execution record, not a valuation guarantee.
  • The replacement programme also has a $1 billion ceiling, but no minimum amount. Oklo chooses whether to sell, the amount and a minimum price; agents can suspend. No September shares or proceeds are established by the source record.
  • At the prospectus’s illustrative $42.57 price, full use would require 23,490,721 shares. The issuer calculates $22.47 of immediate dilution in net tangible book value for a purchaser at that assumed price, while warning that actual price and volume are unknown.
  • Oklo already reported $3.006 billion of cash and marketable debt securities at 30 June. The new facility is therefore better read as flexible corporate capital for a business with large future deployment needs, not as proof that a reactor, fuel facility or customer project is financed.

The same ceiling describes two different states

The September 11 Form 8-K contains a rare clean handoff. Oklo terminated the equity distribution agreement signed on May 13 at the close of business on September 10. There was no termination penalty. By then the company had sold 17,971,448 common shares for gross proceeds of approximately $1 billion. On September 11 it signed a new agreement authorising sales of up to another $1 billion.

The first figure is backward-looking. It has a share count and a completed gross total. The second is forward-looking. It has neither. It is the maximum capacity of a sales channel, not a receivable, a bank commitment or cash held for Oklo. Because the programme has no minimum offering amount, even a perfectly valid agreement can end with anything from no sales to full use.

That distinction matters more here than in a conventional underwritten offering. An ATM does not fix one issue price and one block of shares at launch. Oklo can send transaction notices that specify how much stock to sell and the minimum acceptable price. The ten agents use reasonable efforts under accepted instructions. Sales can take place on the NYSE, through market makers, over the counter, in blocks or negotiated transactions. Oklo or an agent can suspend the process.

In other words, the programme is an option held by the issuer, exercised in pieces against market liquidity. Its dollar ceiling stays still while the number of shares needed to reach it moves.

The completed programme records a changing price

The completed May programme offers two observation points. At June 30, Oklo had sold 10,712,054 shares under it for $680.371 million gross and $670.047 million net. The quarterly filing reported an average net price of $63.51 per share. By September 10, the programme total had risen to 17,971,448 shares and approximately $1 billion gross.

The full-programme gross arithmetic is about $55.64 per share: $1 billion divided by 17,971,448. Subtracting the June totals from the completion totals leaves approximately $319.629 million gross across 7,259,394 shares after June 30, or about $44.03 per share. Those are derived averages from rounded filed amounts. They do not reveal daily sale dates, individual prices or which agent executed which block.

They do reveal the mechanism. Later dollars required more shares than earlier dollars because the implied later average was lower. The company still reached the same gross ceiling, but the ownership units exchanged for the final third of the cash were cheaper. This is why “another $1 billion” cannot be translated into a fixed dilution number before sales occur.

The new prospectus makes the sensitivity explicit. It uses the September 9 NYSE closing price of $42.57 as an illustration. At that price, $1 billion gross corresponds to 23,490,721 shares. That is 5.5 million more shares than the completed May programme required. It is not a forecast: the stock can be sold at prevailing, related or negotiated prices, and Oklo can decline to sell below its stated minimum.

A commission is observable; demand is not guaranteed

Oklo will pay the agents up to 1.5% of the gross sale price on stock sold through them. Full use at the maximum rate would imply as much as $15 million of commission, before the separately described expenses, but neither full use nor the maximum rate is certain. The prospectus estimates about $0.3 million of offering expenses excluding commissions and expense reimbursements. It also provides for specified legal-cost and quarterly reimbursements.

The agents’ names make the channel look heavily underwritten: Goldman Sachs, BofA, Citi, JPMorgan, Morgan Stanley, Barclays, Cantor Fitzgerald, Guggenheim, Canaccord Genuity and B. Riley. Yet their role does not turn the ceiling into a purchase commitment. They act as agents using reasonable efforts, and may sometimes act as principals under a separate terms agreement. The prospectus expressly says there is no assurance that Oklo will sell any shares or fully use the agreement.

The correct operating receipt is therefore not the number of banks. It is the periodic bridge showing shares sold, gross proceeds, commissions and net proceeds. Without that bridge, the programme is distribution infrastructure.

Oklo entered September with a large corporate pool

At June 30, Oklo reported $1.645 billion of cash and cash equivalents, $820.454 million of current marketable debt securities and $541.131 million of non-current marketable debt securities. Together they were $3.006 billion, excluding restricted cash. Current liabilities were $52.017 million and total liabilities $84.325 million.

Those balances were transformed by equity issuance. During the first half, the December 2025 ATM issued 12,376,352 shares for $1.200 billion gross, while the May programme issued 10,712,054 shares for $680.371 million gross. Combined, the two programmes provided $1.852 billion net and accounted for nearly all of the $1.851 billion of net financing cash flow. Outstanding shares rose from 160,514,103 at year-end to 185,090,155 at June 30, although options, restricted units and other activity also belong in the share bridge.

The capital pool is much larger than current operating revenue. Oklo reported $1.210 million of first-half revenue, an $81.601 million net loss, a $124.166 million operating loss and $65.459 million of operating cash use. It also spent $126.901 million on property, plant and equipment, including spending and prepayments associated with planned facilities. Acquisitions and other investments used further cash, while a large amount was moved into marketable securities.

This is not a simple runway comparison. Oklo says significant continuing expenditure will be required for powerhouses, fuel recycling, fuel fabrication, fuel supply and radioisotopes. A developer that intends to design, build, own and operate assets needs capital well before a power sale begins. But the gap between corporate liquidity and operating revenue makes allocation evidence more important, not less.

Broad proceeds preserve authority, not project certainty

The new prospectus says proceeds may be used for general corporate purposes, working capital, capital expenditure and potential future investments. Management has broad discretion. It does not assign percentages to an Aurora powerhouse, the Idaho work, a fuel fabrication line, the Tennessee recycling plan, an acquisition or a radioisotope project.

That breadth is useful to the issuer. It allows Oklo to fund opportunities and constraints that emerge in a development portfolio. It is also a limit on what an outsider can conclude. A dollar sold through the ATM is not project finance: it has no project collateral, construction draw conditions, completion tests, offtake covenant or ring-fenced use.

Oklo’s own filing describes an emerging market with no commercial project operating. It reports deployment and regulatory milestones, non-binding customer letters and a master power agreement, but those records do not make the September ATM proceeds contracted project revenue. Nor does the shelf registration itself provide customers, licences, fuel or construction capacity.

The prospectus measures one form of dilution

At the illustrative $42.57 sale price, the prospectus calculates that full issuance would lift adjusted net tangible book value from $17.34 to $20.10 per share while producing immediate dilution of $22.47 per share for a new buyer at that assumed price. This is an accounting comparison between purchase price and adjusted tangible book value. It is not a prediction of the stock price, a loss forecast or the complete economic experience of an existing holder.

Ownership dilution is simpler but still price-dependent: issuing more shares reduces each existing share’s percentage claim unless the holder also buys. Whether that exchange creates or destroys long-term value depends on what Oklo receives net and what management builds or acquires with it. The ATM can strengthen a balance sheet while diluting ownership; those statements are not contradictory.

The useful discipline is to keep three denominators separate. First is dollars of gross capacity. Second is shares actually issued. Third is productive or strategic output purchased with net proceeds. Oklo has supplied a complete receipt for the old programme, a capacity document for the new one, and only broad categories for future allocation.

Sources