Oklo Funds Its Reactor Buildout by Selling Stock at Falling Prices

If you manage an artificial intelligence data center, your biggest headache isn't buying graphics chips. It is finding the electricity to run them.

Oklo wants to solve that problem by building private nuclear power plants right beside the server racks. But Oklo isn't an equipment vendor selling metal reactors off an assembly line. Oklo is a commercial landlord that plans to build, own, and run small modular reactors, metering the power directly to tech giants.

The demand from Big Tech is undeniable. Training advanced AI models consumes vast quantities of continuous, carbon-free baseload power. Meta has lined up a 1.2-gigawatt campus with Oklo in Ohio, Equinix signed an agreement for 500 megawatts with an upfront cash prepayment, and Switch inked a pact targeting up to 12 gigawatts.

Yet Oklo has never sold a single watt of commercial power. Total quarterly revenue sits at just $1.21 million, generated from technical service and machining contracts tied to recent acquisitions.

Because Oklo is trying to build an entire power grid from scratch before collecting utility bills, its share price serves as its cost of capital. When a pre-revenue developer needs billions to pour concrete, it uses an at-the-market equity facility, dripping newly printed shares into the public float to generate cash. The arithmetic of that mechanism is unforgiving.

Back in May, Oklo raised roughly $1.0 billion by issuing about 18 million shares. After exhausting that facility in September, it immediately launched another $1 billion program. But with shares falling roughly 85% from their peak to hit a fresh 52-week low, raising that same $1 billion now requires dumping about 36 million shares onto the market.

The exact same construction dollar now costs current shareholders twice the equity it did five months ago. It is like paying for a phone on an installment plan where the monthly payment doubles every time your credit score drops.

Between its 2024 debut and August, Oklo expanded its share count by roughly 52%, from about 122 million shares to over 186 million. Serial equity issuance is standard practice across advanced nuclear developers—peer NuScale Power unloaded nearly 90 million shares for $1.0 billion earlier this year before opening another program—yet Oklo is selling into a retreating market.

The engineering itself is making tangible headway. On private land in Lockhart, Texas, Oklo took its Groves test reactor from a blank dirt lot to first criticality in under twelve months, a rare speed record for American pilot projects.

The Department of Energy recently selected Oklo for its Launch Pad initiative to produce cancer-treating medical isotopes using its planned Groves-2 Forge facility. "Groves showed how quickly a new reactor could be built and brought to criticality through close collaboration between the private sector and DOE," CEO Jacob DeWitte said.

With $3.01 billion in cash and marketable securities and an annual cash burn guided between $520 million and $650 million, Oklo holds about five years of runway.

The bear case, however, ignores the press releases and targets the balance sheet. Oklo is caught in a dilution loop where falling equity forces management to issue larger blocks of stock, flooding the market and depressing the price further. Commercial power from its flagship Idaho reactor remains pegged to late 2028, and the company has not yet submitted its combined license application to the Nuclear Regulatory Commission.

Federal regulators just handed down an 18-month delay on a Virginia project by denying an emergency interconnection complaint. The pivot to medical isotopes looks like a narrative distraction without binding customer commitments, and top executives sold millions in personal shares this month instead of buying the dip. If the mid-November filing shows management sold heavy share volume near recent lows to push the share count toward 200 million, the short thesis becomes booked reality.

For the equity to stage a durable re-rate, Oklo must prove it can hit concrete operational gates before issuing away its remaining upside. Submitting an accepted NRC application, protecting its 2028 operational timeline in Idaho, or logging first isotope revenue in early 2027 would give the stock the valuation floor it needs to stop the dilution bleed.

Until then, the market is pricing this risk correctly: building capital-intensive infrastructure with a shrinking stock price guarantees that existing investors pay the bill.

What we are watching is the third-quarter report in mid-November: the share count reported on the cover page will reveal precisely how much cash management pulled from its latest equity program.

The engineering is real, but the dilution is permanent.

Disclosure: The Signal holds no position in OKLO. Positions may change. This is not financial advice.