Nuclear Stocks Are Down 47%: Is This a Buying Opportunity?

Nuclear developers are hitting their most significant technical milestones in years at the precise moment nuclear equities are cratering, and that tension is not a coincidence worth ignoring.

Consider the split screen. Oklo achieved first criticality on private land in August 2026, X-Energy’s construction permit is on track for Q1 2027, and hyperscalers have committed over 9.8 GW of nuclear capacity across 13 announced deals. Yet OKLO shares are down roughly 47% year-to-date, VanEck’s NLR ETF has shed approximately 35% from its January 2026 peak, and the Sprott Physical Uranium Trust is trading at a mid-single-digit discount to net asset value.

That divergence between operational progress and share price reality is striking enough to demand a close look at what is actually driving it. This piece gives you a structured framework for separating the noise driving this sell-off from the signal embedded in the structural build-out. After this, you will know which metrics to watch, which risks are already priced in, and which are not, before committing capital to the sector.

Why nuclear equities sold off despite a strengthening build-out story

The sell-off did not come from anything wrong with uranium. It came from everywhere else.

Over the weeks leading into late summer, three exogenous pressures converged on the sector. Broader anxiety about the pace of AI growth soured sentiment across anything tied to data-centre power demand. Public commentary from Anthropic and OpenAI leadership added to the negative technology mood. And unfounded fears around unverified Saudi Arabian mineral discovery claims, attributed to a Reuters report, gave momentum sellers one more excuse.

None of that touched the physical commodity. Uranium spot sat at US$89.70/lb as of 18 September 2026, up roughly 4.3% year-to-date, with long-term contract prices near US$96.50/lb. Utility buyers, in other words, are still paying up for future delivery.

The equities told a different story entirely. Uranium mining shares fell 3.9% over the same window, junior miners 7.43%, and developer names such as Oklo fell far harder. That gap between a rising commodity and falling equities is the single most important tell in this whole sell-off.

The uranium price-equity divergence playing out in 2026 is not unprecedented; similar disconnects emerged during the early phases of the 2003-2007 cycle before the equities caught up violently, which is precisely why understanding the mechanism behind the gap matters more than the gap itself.

Sprott framed the split plainly.

Jacob White, Director of ETF Product Management at Sprott, described the divergence as a buying opportunity driven by macro risk-off positioning rather than weakening fundamentals.

VanEck’s July 2026 “Nuclear Reset” report reached a compatible read, noting its NLR ETF had become temporarily expensive and that investors were shifting their demands from “promise” to “proof.”

Here is what that tells you. When the physical commodity holds firm while the equities fall, you are looking at a valuation reset after a period of excess, not a crack in the supply-demand picture. That distinction is the prerequisite for every decision that follows: if the cause is sentiment rather than a structural fault, the risk of holding through the volatility looks materially different.

The structural case that makes the pullback look like opportunity

Strip out the sentiment, and what remains is a demand picture most commodity sectors would envy.

Start with the buyers. As of May 2026, the four major U.S. hyperscalers had committed over 9.8 GW of nuclear capacity across 13 announced deals. These are not options or letters of intent. They are multi-decade power purchase agreements signed by counterparties with investment-grade balance sheets.

• Microsoft: an 835 MW, US$16 billion, 20-year PPA tied to the restart of Three Mile Island Unit 1.
• Google: commitments to 500 MW from Kairos Power plus a 22-year nuclear PPA with Fortum linked to Finland’s Loviisa plant.
• Amazon: investment in X-Energy and a 960 MW data centre adjacent to Talen Energy’s Susquehanna nuclear plant.
• Meta: early-2026 agreements with TerraPower, Oklo, and Vistra targeting 6-6.6 GW of capacity by 2035.

Read those terms carefully. A 15- to 20-year contract signed by a company with the capital to honour it functions as a revenue anchor, the exact instrument that finances first-of-a-kind reactor projects. For a U.S. investor, that contractual visibility substantially changes the risk calculus, because demand is no longer speculative or dependent on policy alone. It is locked in.

Hyperscaler nuclear commitments have accelerated faster in 2026 than most utility analysts projected, with the contractual structures now extending beyond simple PPAs into co-location arrangements, direct equity stakes in developers, and fuel supply pre-purchases that bind tech balance sheets to the uranium supply chain for decades.

Global demand beyond the U.S. hyperscaler layer

The demand base does not stop at American data centres, and that geographic spread matters for reducing single-region exposure.

China and India are pursuing aggressive new reactor construction programmes, while numerous other nations extend the operational lives of existing plants. This is structural demand driven by decarbonisation goals and energy security mandates, entirely insulated from any AI-sector slowdown.

The World Nuclear Association (WNA) projects global uranium demand rising 28% from 2023 to 2030, with scenarios pointing to global nuclear capacity roughly doubling by 2040. That doubling would require doubling annual uranium requirements, which grounds the long-term price case in physical reactor demand rather than equity sentiment. When the demand thesis rests on reactors under construction across multiple continents, it becomes far harder to dismiss the current equity weakness as anything but a sentiment episode.

What the milestones actually tell you about execution risk

This is where the analysis has to hold two ideas at once: the milestones are genuinely meaningful, and they are not commercial deployment.

Oklo achieved first criticality at its Groves Isotope Test Reactor in Lockhart, Texas, at 9:19 p.m. ET on 5 August 2026, becoming the fifth DOE-authorised reactor to go critical that summer and the first to do so on private land. The stock, meanwhile, traded near a 52-week low of US$34.38 and closed at US$40.18 on 21 September 2026, still down roughly 47% year-to-date.

X-Energy submitted a construction permit application with Dow for its Seadrift, Texas project on 31 March 2025. The NRC completed its review ahead of schedule, issuing an Environmental Assessment with a Finding of No Significant Impact (FONSI) in May 2026, with the final permit anticipated in Q1 2027. There is no publicly traded X-Energy equity.

These are real regulatory de-risking events. They reduce binary risk. What they do not do is prove near-term commercial readiness, and the timeline gap is substantial.

SMR technology fundamentals explain why the gap between first criticality and commercial operation is measured in years rather than months: novel coolant chemistry, first-of-a-kind licensing pathways, and modular factory fabrication supply chains each introduce dependencies that sequential project planning cannot compress.

The Information Technology and Innovation Foundation (ITIF) and the IAEA place realistic commercial nuclear infrastructure development at 7-15 years, well beyond the 2-4 year schedules some vendors advertise.

To size a position sensibly, it helps to map where each name sits on the path from lab to fleet:

1. Regulatory authorisation and licensing progress
2. First criticality and low-power testing (Oklo is here)
3. Construction permit issuance (X-Energy expects this in Q1 2027)
4. Full construction and commissioning
5. Commercial operation and revenue generation

Both companies sit in the early stages. That is meaningful progress, but for a U.S. investor it shifts the profile from “unproven technology” toward “de-risked development stage,” not to “revenue-generating operator.” The distance between those last two is measured in years and capital, not quarters.

Sector-wide execution risks beyond the milestone headlines

Beyond individual companies, four risk categories weigh on the entire SMR cohort.

Licensing uncertainty remains active, because NRC reviews for novel coolants and fuels are still being tested. Cost and financing pose a harder problem: the EFI Foundation and Hunton Andrews Kurth have stressed that high weighted average cost of capital and documented first-of-a-kind cost overruns make small modular reactors difficult to fund without offtake guarantees. Supply chains for non-traditional fuels remain immature, and waste management pathways are unresolved.

NuScale’s cancelled Utah project stands as the most proximate cautionary data point, a reminder that cost-overrun risk, not technology, remains the primary barrier to new commitments. Conflating a regulatory milestone with commercial arrival is precisely how investors get trapped in developer equities that run ahead of their fundamentals.

History as a guide: what the 2003-2007 uranium cycle teaches about positioning

The last uranium bull market offers the cleanest mirror for reading this one, and its lesson cuts both ways.

Between 2003 and 2007, uranium climbed from roughly US$10/lb to a peak near US$136-137/lb, before collapsing.

That move, approximately 1,801% over roughly 6.5 years, was followed by a violent reversal. A structurally justified bull market still overshot and broke down.

Today’s spot price of US$89.70/lb, having briefly touched US$100.25/lb in January 2026 before settling, sits in a different context. Several features distinguish the current cycle from the earlier one:

• Contractual demand from hyperscalers, versus a more speculative demand base in the 2000s.
• Active reactor construction across multiple jurisdictions rather than concentrated speculation.
• Policy continuity, with DOE support persisting across both U.S. administrations.

Those differences make the structural entry point look stronger. They do not make it invulnerable.

Some commentators, in Kitco and Seeking Alpha pieces noted as perspective rather than consensus, argue that SMR developer equities can behave like long-dated options on unproven technology. The implication is a value-trap risk: a prolonged deployment delay or a high-profile cost overrun could reprice the entire developer cohort even if uranium demand stays intact.

Here is the read to take. The 2003-2007 cycle tells you that a structurally correct thesis does not protect you from a sentiment-driven drawdown of 50-80%. Before treating this pullback as an automatic entry signal, you need to decide whether your time horizon and position sizing can survive that scenario.

Separating signal from noise: a framework for the nuclear entry decision

The sector offers two distinct investable layers, and treating them as one is the most common mistake.

The physical layer is uranium exposure through vehicles such as the Sprott Physical Uranium Trust (SPUT), currently trading at a discount of -4.95% to -7.24% to net asset value. The developer layer is milestone-driven equity such as Oklo, with catalysts but no current revenue. Their risk-return profiles diverge sharply.

For a U.S. investor, the physical layer at a discount to NAV offers a cleaner, more contained entry. Developer equities offer higher potential reward, but their dependence on execution timelines means they behave more like staged bets than commodity trades.

The signals worth watching fall into two groups. These would confirm the accumulation thesis:

• SPUT’s discount narrowing toward par.
• The NLR ETF reclaiming its January 2026 starting level.
• X-Energy’s construction permit issued on schedule in Q1 2027.
• Oklo progressing from criticality toward commercial licensing.

These would invalidate it, suggesting the pullback is early evidence of a structural problem rather than noise:

• Uranium spot falling and holding below US$80/lb.
• A major hyperscaler PPA renegotiated or cancelled.
• A high-profile first-of-a-kind cost overrun repudiating the SMR cost narrative.

The long-term contract price near US$96.50/lb, sitting above spot, currently signals continued utility buyer commitment. That single data point tilts the near-term evidence toward the sentiment reading rather than the structural one.

What the evidence actually supports right now

Strip away the noise, and a clear-eyed verdict emerges.

The pullback carries the characteristics of a sentiment-driven dislocation, not a fundamental breakdown. The three-pillar bull case, converging milestones, contracted hyperscaler demand, and stable physical uranium pricing, remains intact. But the recovery timeline depends on the X-Energy permit, NRC processing pace, and macro conditions that stay genuinely uncertain.

That makes the investor profile the deciding variable. This looks like an opportunity for a long-horizon, milestone-aware investor whose position sizing can absorb a 50%+ drawdown. It looks like a trap for a short-horizon investor whose thesis depends on near-term revenue and who cannot withstand further SMR sector repricing.

The honest conclusion: the sector has earned a closer look from investors with the right time horizon and risk tolerance, weighed against the counter-case of first-of-a-kind cost history, regulatory timing, and developer-equity optionality risk. It has not earned a conviction buy from anyone whose thesis depends on near-term commercial deployment at the prices developers are currently marketing.

The next 6-18 months are the evidence window. The sector will either produce the “proof” VanEck describes, the X-Energy permit, Oklo’s commercial licensing progress, and hyperscaler PPAs translating into physical construction, or it will not. That evidence, not today’s share price, is what determines whether this was an accumulation window.

For investors ready to move from thesis to portfolio construction, our dedicated guide to nuclear investment frameworks covers how to weight physical commodity exposure against developer equity, stage-gate position sizing across licensing milestones, and the specific trigger conditions that justify adding to each layer.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors, and forward-looking statements are speculative and subject to change based on market developments and company performance.