2 Nuclear Stocks Crashed by About 50% in 2026. Here's the 1 Thing That Would Bring Them Back.

NuScale Power shares are down 48% year-to-date in 2026 and Oklo has fallen 51%, as investors question whether non-binding SMR customer agreements will convert into funded projects. The sector's key catalyst is a binding power-purchase agreement: NuScale aims to secure one with Tennessee Valley Authority by year-end for a proposed 6-gigawatt facility, which would be the world's largest SMR deployment if built. Despite Bank of America's estimate of a $10 trillion nuclear opportunity over the next 25 years, driven partly by AI data-center electricity demand, execution, financing and project-cost risks remain substantial.

Analysis

The relevant repricing event is not another data-center power announcement but bankable contracting: a long-dated PPA with a creditworthy offtaker, indexed economics, and evidence of project financing. Until then, SMR and OKLO remain duration-sensitive development equities whose valuation is driven by probability-of-completion rather than near-term earnings; further equity issuance is the likely funding bridge if commitments slip. A signed agreement alone is insufficient unless it discloses price, escalation, interconnection status, construction responsibility, and termination protections.

Near term, the selloff can persist because prospective customers can meet incremental load with gas, renewables-plus-storage, grid upgrades, and demand-response before first-of-a-kind reactors are operating. The second-order beneficiary is existing dispatchable generation: CEG and VST monetize tight power markets today, while uranium exposure through CCJ/URNM captures nuclear build enthusiasm without single-project execution risk. Nuclear suppliers such as BWXT offer a more defensible route to SMR spending, since engineering and component revenue can precede commercial fleet deployment.

The contrarian case is that a credible utility-backed PPA could cause a violent rerating given depressed positioning, but this is a binary catalyst rather than an investable base case. Over the next 1-3 months, monitor definitive-contract filings, DOE loan support, NRC licensing milestones, and announced equity/debt raises; over 6-18 months, cost-per-MW and financing terms matter more than headline backlog. The thesis is falsified by repeated timing slippage, non-recourse financing unavailable at contracted power prices, or a customer opting for alternative capacity instead.

Key Decisions for Investors

• Avoid directional long exposure to SMR and OKLO ahead of binding-contract disclosure; treat them as event-driven watchlist names rather than core AI-power exposure over the next 1-3 months.
• Express the power-demand theme through long CEG or VST versus short SMR in a 6-12 month pair: incumbents have operating cash flow and immediate wholesale-power sensitivity, while SMR retains financing and execution risk. Reassess if SMR secures a fully financed, creditworthy PPA with disclosed economics.
• For nuclear upside with lower project-concentration risk, accumulate BWXT or CCJ/URNM on sector weakness over 6-18 months; risk is reactor delays reducing the timing of orders and uranium demand, rather than a single developer failing to convert its pipeline.
• Set an event alert for SEC-filed PPA terms, committed construction financing, and a named investment-grade offtaker for SMR or OKLO. Only initiate a catalyst long after confirming contracted capacity, price escalation, and funding sources; otherwise any rally is likely multiple expansion unsupported by cash-flow visibility.