After a 78% drop, did OKLO actually succeed?

NEXT PICK · Market Insights

After a 78% drop, did OKLO actually succeed?

Tuesday, September 8, 2026

A company with total revenue of only $1.21 million over the past twelve months, a market value of $8.06 billion, an incalculable price-to-earnings ratio, and earnings per share of minus $0.93—and on the back of this financial snapshot, it ignited its first reactor in early August and signed a 1.2 GW nuclear power park agreement with Meta at the end of August.

Even more glaring is the stock price

$43.31, up 4.94% today, but this price is nearly 78% deep from the 52-week high of $193.84, and only an 18% buffer from the 52-week low of $36.61.

Fundamentals are moving forward while prices are retreating. This divergence is not uncommon among growth stocks, but rarely has a single stock widened the gap between the two so widely—so, did the market see something first, or was it too lazy to look further?

Latest price $43.31 ▲ +4.94%

Height 52 weeks apart -77.7% ▼ High $193.84

Sales-to-market ratio About 6,600 times higher ▲ TTM revenue was $1.21 million

With revenue of 1.21 million and a market value of 8 billion, what exactly is this company selling?

Founded in 2013 and headquartered in Santa Clara, California, Oklo was renamed and listed on the New York Stock Exchange in May 2024 through a merger with AltC Acquisition Corp. Its core product is the Aurora Powerhouse, a modular small stack (SMR) designed to cover power ranging from 15 to 75 megawatts.

Let's first look at the revenue structure, because it's more honest than any narrative: Q2 2026 revenue is $1.21 million, including $800,000 in engineering and consulting services, $168,000 in manufacturing and fuel processing services, and $242,000 in other businesses. And the rolling revenue over the past twelve months is also $1.21 million—meaning nearly all of the company's annual revenue came from that one quarter.

What truly sets Oklo apart from its peers is not power generation itself, but fuel. The company is advancing both nuclear fuel recovery and fuel manufacturing technologies, converting spent fuel into reactor-usable fuel. The Aurora-INL project plans to extract 5 metric tons of high-content low-enriched uranium (HALEU) from the EBR-II reactor.

Against the backdrop of ongoing global HALEU supply chain constraints and the ongoing construction of concentrated production capacity in the U.S., this closed-loop capability means less upstream dependence and more controllable cost structures—for customers like data centers who prioritize "power continuity" over costs, whether the fuel source is self-sufficient is not insignificant.

Another difference is adjustable power. The 15 to 75 megawatt range allows Oklo to simultaneously meet the small scale needs of a single industrial facility and the grid connection needs of small utility clusters, whereas most competitors' products have relatively fixed power and cannot be used at the same time.

Rather than thinking of Oklo as a power company, think of it as a heavy equipment manufacturer that hasn't delivered its first product yet—you're not buying a bill, but a capacity option that won't be paid out until around 2030. So the question becomes straightforward: why is an option worth $8 billion?

The absurdity of a 6,600-fold price-to-sales ratio versus the reality of $3 billion in cash

If you open Oklo's valuation snapshot, you'll find that almost all commonly used tools fail: P/E ratio can't be calculated due to ongoing losses, earnings per share are negative $0.93; Common anchors like price-to-sales ratio, enterprise value multiple, and free cash flow yield are either impossible to calculate or have no reference value.

The only thing that can still be implemented is the price-to-book ratio (P/B), about 2.5 times book value—while mature nuclear operators typically only range from 0.8 to 1.5 times. As for price-to-sales ratio, $8.06 billion in market cap divided by $1.21 million in rolling revenue is about 6,600 times, which is less a valuation and more a reminder not to use it for valuation.

Cash is actually the most solid part of the whole story. At the end of Q2, the company held $3 billion in cash and securities, a figure much higher than the research report caliber, pushing back the risk of "financing gaps" by several years.

But the pace of burning money is accelerating at the same time

The full-year 2026 operating cash burn guidance is raised from $80 million to $120 million to $150 million, and the capital expenditure guidance is raised from $350 million to $450 million to $400 million to $500 million. Combined, the full-year consumption level is between $520 million and $650 million.

At this rate, $3 billion in liquidity reserves can support about four to five years. And the guidance increase itself has two sides: on one hand, blood loss accelerates; on the other, projects under construction are really moving — capital expenditure never increases because of PPTs.

When all valuation anchors for a company fail, what is the market actually pricing? The answer isn't the $1.21 million revenue, but the "1.2 GW launch by 2030" that hasn't happened yet, and the probability of it being delivered.

From "government verification" to "first criticality," labels need to be rewritten

Under traditional frameworks, Oklo has always been classified as a "government validation" phase: no commercially operational reactors, no formal design certification, no substantive power purchase agreements. But progress over the past few months has pushed this label forward more than half a step.

In April 2026, the U.S. Nuclear Regulatory Commission (NRC) approved the Major Design Guidelines Seminary Report for the Aurora Idaho project; The DOE Idaho office subsequently approved Aurora-INL's nuclear safety design agreement and approved a preliminary documented safety analysis in Q2; In July, the company selected Kiewit Nuclear Solutions as the main contractor for the project.

Stronger evidence comes from early August: the Groves isotope reactor achieved its first criticality, just 11 months before groundbreaking and 229 days of actual construction, setting a record for the fastest private funded reactor to reach criticality. Management thus claims the company already has a replicable deployment organization—processes, safety benchmarks, trained operators, procurement networks, and real cost data.

But the market hasn't changed its tune because of this. The latest price at $44.07 on the 50-day moving average and $64.85 on the 200-day moving average is both below these two lines, about 33% away from the 200-day moving average. When progress is moving forward while prices lag behind, is execution underestimated or time cost underestimated?

NuScale dropped from $57 to $9.7, and the entire track was repriced together

Let's first look at the most straightforward comparable listed companies. NuScale Power (NYSE: SMR) currently has a stock price of $9.70, a market capitalization of $4.17 billion, and a 52-week range of $7.21 to $57.42. It was the world's first SMR developer to receive NRC design certification, but suffered a major commercial setback in 2023 after canceling the 12 VOYGR reactor project in UAMPS, Utah.

The weight of this case lies in:

Obtaining certification does not mean receiving the order; Obtaining the order does not mean the project will not be canceled.

The private sector is no exception. TerraPower, founded by Bill Gates, has a Natrium sodium-cooled fast reactor designed for about 345 megawatts. The Wyoming demonstration project is supported by DOE and also has a fuel cycle layout, directly competing with Oklo in terms of fuel autonomy.

GE Hitachi's BWRX-300 boiling water reactor is designed for 300 MW and has secured early deployment commitments in Poland and Canada. Its parent company's engineering manufacturing capabilities and global supply chain are its most advantageous areas. X-energy's Xe-100 high-temperature gas-cooled reactor is about 80 MW, highly overlapping with Aurora, also uses fuel recovery, and has received funding from the DOE Advanced Reactor Demonstration Project. Kairos Power's Hermes molten salt cooling demonstration reactor is undergoing the NRC pre-application process.

Oklo's differentiation lies in four points: adjustable power coverage for a broader customer base, fuel closed-loop reducing upstream dependence, a public financing channel from NYSE listing (including a $1 billion ATM issuance quota), and now the latest one—it really stacks up a dot.

If the past two years have been a contest of whose roadmap is more beautiful, then NuScale's drop from $57 to $9.7 shows that the market has quietly shifted its scoring standards: now it's about who really poured the concrete.

The ceiling is high enough, but the road to the ceiling will be built until 2034

Let's start with the big picture. The global nuclear energy market is about $250 to $400 billion, with the SMR sector accounting for about 5% to 10%, or $12.5 to $40 billion; If the current slope of power demand for AI data centers continues and US nuclear policies remain favorable, the potential total accessible market (TAM) could expand to $50 to $80 billion by 2030.

Then count the portion that can end up in their own hands. In the U.S. alone, data centers have an annual demand for new baseload power of about 20 to 50 GW, with SMRs reaching about 10% to 20%, i.e., 2 to 10 GW; Based on the industry's prevailing investment threshold of 3 to 5 million USD per megawatt, Oklo's serviceable market (SAM) ranges from 6 billion to 50 billion USD.

Meanwhile, the Available Market (SOM) is no longer just a hypothesis. The agreement reached with Meta at the end of August targets a 1.2 GW advanced nuclear park in Pike County, Ohio, with preliminary construction and site surveys to begin in 2026, with Phase I expected to go online as early as 2030 and gradually expand to full capacity by 2034; Meta has also set up an upfront payment mechanism to secure fuel early and advance early milestones for Oklo.

Using the same per-megawatt scale, the construction scale for this single project alone is estimated to be in the $3.6 to $6 billion range. The ceiling has never been a problem; the problem is that going from $1.21 million in revenue in 2026 to full production in 2034 will require eight years, regulation, fuel, and investor patience.

The demand-side supercycle collides with the supply-side's just-proven execution

The bullish logic has changed fundamentally over the past six months — it is no longer just "AI power shortages so nuclear power is good," but now has verifiable engineering and contract vouchers:

▲ Bull Case

Projects can speed up, but valuations are overdrawn for the world ten years from now

The bears' rebuttals are also well-documented, and most have nothing to do with "whether the company is good," only about "whether the price is expensive."

▼ Bear Case

Three threads determine the next step: fuel, regulation, and orders

In the next 12 to 18 months, Oklo's stock price will likely be dominated by three types of events: whichever type will land first, and that type will gain pricing power:

•	Aurora-INL First Stack Loading and Startup: DOE authorizes initial core processing at its own fuel manufacturing facility, with progress directly determining whether commercial operation targets from late 2027 to early 2028 can be met, with a positive impact
•	Meta's Ohio project preliminary construction and site survey began: to progress within 2026, with the pace of advance payments being a key observation point and a positive impact
•	The second data center-level substantial power purchase agreement: if implemented between Q4 2026 and 2027, it will turn Meta's orders from isolated cases into patterns, with significant impact
•	Quarterly revenue a surge in volume: Whether the $1.21 million base can be raised in Q3 and Q4 2026 is a direct test of whether the "commercialization icebreaker" narrative can continue, with a neutral to positive impact
•	Further implementation of HALEU supply and fuel recovery capacity: Global supply chains remain tight, and any long-term agreement reduces the greatest single point of failure risk, with high impact

A 78% drop in one year, a turning point is a missed date

Looking at the timeline, Oklo's price range over the past 12 months has been $36.61 to $193.84—a difference of more than five times between the highs and lows, which in itself shows how sharply divided the market is over it.

In the first half of 2026, the stock price fell about 27%; In July alone, it dropped another 26%, and the yearly decline widened to around 46%. The direct trigger for the sell-off was not performance, but the company missing the DOE's July 4 threshold deadline—a date that forced the entire story's valuation model from "narrative-driven" to "delivery-driven."

The second half is interesting. The critical threshold was indeed reached in early August, Meta's 1.2 GW agreement was finalized at the end of August, but after a brief surge, the stock price fell back again. Now, $43.31 is only about 18% above the 52-week low, still below the 50-day and 200-day moving averages.

Looking back, the market's judgment that "AI data centers will pay for nuclear energy" was correct, and Meta's contract is proof; The mistake was the timeline—investors paid for 2030 capacity in 2027 in 2025, but reality made up the gap with a year-long decline. Today's 4.94% increase was accompanied by $12.7 million in bullish large orders, more like momentum funds chasing a name with a story, showing no signs of fundamentals being repriced.

⚠️ Risk Notice • Valuation Anchored: P/E ratio is about 6600 times, and any delay could trigger a deep pullback • Delivery risk: The DOE has missed its July 4 critical deadline, and there is already a precedent for project delays • Accelerated cash consumption: Full-year operating and capital expenditure guidance both raised to $520 million to $650 million • Single point of failure in fuel: HALEU supply chain is tight, and commercial-scale supply sources remain limited • Revenue realized extremely late: Meta's first phase will only launch in 2030, with no large-scale power generation revenue in between • Sector Discount: NuScale fell to $9.70, with the overall valuation center of the SMR sector still moving downward

🟡 Neutral The first critical and Meta major order has been fulfilled, and the 6600x price-to-sales ratio remains unanchored.

💬 Discussion The project is being delivered, but valuations remain unflat, so it's not advisable to chase highs. Post real-time buy and sell positions on OKLO to the official WeChat account.

Data source

•	Source: NextPick real-time snapshot + company quarterly earnings and investor materials + SEC disclosures + NRC/DOE public announcements + mainstream financial media coverage.