Oklo (NYSE: OKLO) is optionality, and I want to name that plainly at the top. It is a pre-revenue advanced-nuclear company selling a genuinely compelling vision — build, own, and operate small fast-reactor “powerhouses” and sell the electricity — into a market (AI data centers desperate for firm, clean power) that has never been more receptive. But it earns no revenue today, faces a multi-year regulatory gauntlet, and its value is almost entirely in the future. That’s why my stance is Watching, not Bullish, and why the sizing is deliberately small.
Figures are illustrative, from company disclosures and my dashboard snapshot. Not advice.
The model is the interesting part
Most nuclear startups plan to sell reactors. Oklo’s pitch is different and, if it works, better: it intends to build, own, and operate its Aurora powerhouses and sell the electricity under long-term power purchase agreements. That flips the business from a lumpy, capital-intensive equipment sale into a recurring, utility-like revenue stream — contracted power over decades. It also aligns with what AI hyperscalers actually want: not to buy and run a reactor, but to sign a long-term contract for firm, clean, 24/7 power. Oklo has accumulated a large pipeline of customer interest (letters of intent and agreements) precisely because that model fits the moment.
Customer pipeline (illustrative) — interest is not yet revenue
Illustrative pipeline growth. Crucial caveat: letters of intent are demand signals, not contracted, delivered, revenue-generating megawatts. The gap between the two is the entire risk.
1 — Operators & backers
Oklo carries high-profile backing and a technical team pursuing fast-reactor designs with a genuine long-term vision (including fuel recycling as a later-stage differentiator). Sponsorship and ambition are real. But for a company like this, the operator test isn’t charisma — it’s regulatory and construction execution. Can they navigate the NRC licensing process, actually build a first-of-a-kind plant on something resembling budget and schedule, and prove the operating model? That’s unproven, which is why this pillar can’t score green yet.
2 — Primary research: respect the regulatory reality
Reading the primary story honestly: Oklo’s licensing path has not been smooth — an earlier application was denied and the company has had to refine and refile its approach. That’s not disqualifying (first-of-a-kind regulatory paths rarely are linear), but it is the defining risk. Nuclear is regulated for good reason, timelines are long, and the process is largely binary and out of the company’s control. I refuse to model this as if approval and on-time construction are foregone conclusions. The financials today are simple: no revenue, ongoing cash burn, a cash balance funding the runway toward first deployment.
| Stage | What it proves | Status |
|---|---|---|
| Site & fuel secured | Foundation for first plant | Progressing |
| NRC licensing | The binary gate | In process (refiled) |
| First Aurora built & operating | Model works in reality | Future / the moment |
| First revenue under PPA | Recurring cash begins | Future |
| Fleet deployment at scale | The full thesis | Years out |
Illustrative milestone ladder. Everything hinges on clearing the licensing gate and building the first unit. Until then, the equity is a claim on a plan.
3 — A real wave, an unproven seat
The wave is as real as it gets: AI data centers and re-industrialization need vast amounts of firm, clean, always-on power, and the grid can’t deliver it fast enough with existing sources. Advanced nuclear is one of the few technologies that could structurally answer that need. Oklo’s potential seat — owning and operating the assets and selling power — is attractive if it can execute. But the seat is unproven, competitors (other SMR and advanced-fission players) are chasing the same demand, and the winner will be decided by who actually gets built and licensed, not who has the best deck.
4 — The economics are a promise, not a print
The long-term economics could be excellent: a build-own-operate fleet selling contracted power is a high-quality, annuity-like model with strong margins once assets are operating. But that is entirely prospective. Today there is no cost curve to model from real operations — only projected costs for a first-of-a-kind plant, which history says tend to run over. The honest framing: I’m underwriting a range of futures, heavily weighted by execution and regulatory probability, not a financial model with real inputs.
5 — Asymmetry & stance
Stance: Watching. Conviction: Speculative. The asymmetry is deliberately barbell-shaped: small, defined downside against a large, contingent upside tied to a genuinely transformative model in a market that wants exactly what Oklo is selling. The stance moves toward Bullish only when execution risk starts converting to evidence — licensing progress and, ultimately, steel in the ground.
Upside case
- Licensing clears and the first Aurora gets built and operates.
- Build-own-operate PPAs convert the pipeline into recurring, decades-long revenue.
- AI/data-center power demand makes firm clean power a premium product.
- Fuel-recycling and fleet economics become a durable differentiator.
Downside case
- Regulatory approval stalls or fails — the binary risk hits.
- First-of-a-kind construction runs badly over budget/schedule.
- Repeated capital raises dilute holders before any revenue arrives.
- Competing SMR designs get to market first.
The Ovatek Lens scorecard
A classic “Watching” profile — the asymmetry is real, but two amber pillars (regulatory-dominated thesis, pre-revenue economics) keep it speculative.
What would move it up — and thesis-break criteria
Moves toward Bullish if: licensing progresses concretely and the first powerhouse moves from paper toward construction and operation.
Out / trimmed if:
- Licensing is denied again with no credible path forward.
- Construction of a first unit slips indefinitely or costs blow out.
- The pipeline of LOIs fails to convert into binding, revenue-generating agreements.
- Capital raises turn chronic and dilutive with no milestones met.
I’d rather hold a small, honest option on a transformative technology than pretend a pre-revenue nuclear startup is a sure thing. Speculative bets get speculative sizing — and defined exits.