Why NuScale's Nuclear Bet Might Be Running Ahead of Reality

NuScale Power (SMR) just became the poster child for the AI-powered nuclear energy boom—and the market got way too excited about it.

On paper, things look incredible. The company scored the only NRC-certified small modular reactor (SMR) design in the U.S., just inked a massive partnership with ENTRA1 Energy (positioning them for up to $25 billion in investment under the new U.S.-Japan framework), and announced another landmark deal with Tennessee Valley Authority (TVA) for 6 gigawatts of new capacity. With $5 billion in valuation and the entire datacenter industry desperate for baseload power, you’d think SMR would be printing money.

Then Q3 hit. And reality checked the hype.

The Numbers Don’t Match the Narrative

NuScale reported a loss of $1.85 per share—way worse than the consensus estimate of -$0.11 and the year-ago loss of -$0.18. Analysts immediately slashed the 2025 EPS forecast from -$0.50 to -$1.64 (that’s a 490% deterioration). Revenue came in at $8.24M, up 1,635% year-over-year, but missed estimates by 25.7%. The kicker: operating expenses exploded 1,213.5% to $541.15 million, with general & administrative costs jumping 2,950.5%.

In other words, they’re burning cash faster as they scale—classic pre-revenue biotech energy problems.

The Real Problem: Timeline vs. Hype

Here’s what nobody’s talking about loud enough: even the most optimistic estimates put these reactors 5+ years away from commercial operation. The market’s already priced in a savior narrative, but NuScale is still in the engineering phase. They have $692.1M in cash (up from $420.7M last quarter), which sounds healthy until you do the math—at current burn rates, that runway gets interesting fast.

The biggest red flag? Fluor (FLR), NuScale’s largest shareholder with 39% of the company, just announced plans to monetize their 111M+ shares. That’s 500M+ shares hitting the market eventually. Activist investor Starboard Value pressured them into it, and sure enough, the stock tanked from $53 to $33 in October. It’s recovered slightly, but the selling pressure is real.

The Bull Case Isn’t Wrong, Just Premature

Don’t get me wrong—the industry thesis is solid. Data centers are consuming power like never before, traditional grids can’t keep up, and nuclear is the only baseload option that checks all boxes (clean, reliable, scales fast). NuScale has the regulatory advantage, the partnerships, and the capital. They will probably be a major player.

But investors are confusing “good long-term story” with “buy now at any price.” The stock’s trading on 2029-2030 revenue assumptions today, while the company’s hemorrhaging money on development costs in 2025.

What Needs to Happen

For SMR to go from speculative to investable, you need to see:

1. Operating expense stabilization (that 2,950% G&A jump is unsustainable)
2. A few quarters of narrowing losses
3. TVA and ENTRA1 deals actually breaking ground (not just announced)
4. Proof that the construction timeline is holding

Right now, it’s a fascinating company with a real moat—but the valuation assumes a perfect execution story in an industry that historically doesn’t execute perfectly. Smart play might be waiting for the Fluor liquidation to finish, then reassess when the stock actually reflects the 5-year development timeline baked in.