NuScale Power Stock Is Down 75% in 12 Months. Here's Why.

NuScale Power (SMR) shares are down ~75% as the company has not converted its certified small modular reactor design into commercial deployments. The sell-off is attributed to fading AI/data-center power enthusiasm and slower-than-expected SMR sales, with its Romania and TVA projects still pre-development and unlikely to produce concrete outcomes until the 2030s. Overall, the article frames SMR as an early-stage, high-risk long-duration bet rather than a near-term revenue story.

Analysis

SMR’s equity story remains a financing-duration trade, not a fundamentals trade: when the market stops paying for a hypothetical 2030s backlog, the multiple should compress toward other pre-revenue industrial developers rather than AI-adjacent energy winners. The key second-order effect is that capital is likely to migrate from long-dated nuclear optionality into nearer-term power solutions for data centers, which creates a relative-value tailwind for BE and, more broadly, for distributed generation and grid-interconnection names.

The biggest loser set is not just SMR shareholders but also adjacent “promise” names that depended on an AI capex narrative to shorten payback periods. If hyperscalers keep choosing modular, quicker-to-deploy solutions, SMR’s addressable demand becomes more of a utility procurement story than a tech-cycle story, and that pushes monetization farther out. That matters because the market has already started discounting the gap between certification and cash flow; another quarter or two of no commercial conversion can still drive meaningful downside.

Catalysts are binary and slow: a signed FOAK deployment, project financing, or a credible partner commitment would be the only near-term reversal, and those are months-to-years away. Absent that, the stock is vulnerable to continued derating whenever AI enthusiasm cools or rates back up, since long-duration equity claims are most sensitive to higher discount rates. Contrarian-wise, the market may be underestimating how much nuclear optionality remains in a constrained-grid world, but that upside is too remote for the current tape unless execution accelerates materially.

This looks best expressed as a relative-value short, not a standalone outright short, because the thesis is about timing and capital allocation rather than terminal failure. If BE keeps converting data-center demand while SMR stays pre-commercial, the spread should widen over 1-3 months. The thesis is falsified by a firm commercial award or project-financing milestone that moves SMR from narrative to booked revenue.

Key Decisions for Investors

• Enter a tactical pair: long BE / short SMR, sized for a 1-3 month catalyst window; thesis is nearer-term monetization winning over pre-revenue optionality, with downside on SMR if no commercial milestone appears.
• Avoid adding to SMR on weakness unless there is independently verifiable evidence of a signed deployment or project finance close; without that, treat rallies as sellable into lower liquidity.
• For existing SMR longs, set a hard risk trigger around any announced delay in the Romania/TVA pipeline or another quarter with no conversion evidence; the stock can re-rate lower quickly if the market pushes the first revenue into the 2030s.
• Watch BE and other distributed power proxies for follow-through; if they continue to outperform on data-center demand, that is confirmation the market prefers near-term power solutions over nuclear duration.