Should You Buy NuScale Power While It's Below $20?

The market is treating SMR like a binary science project, but the more important setup is option value on grid bottlenecks rather than on near-term power sales. If AI buildout keeps pulling load closer to the data center edge, the first winners are not necessarily the reactor vendors — it’s the utility-equipment stack, EPC contractors, switchgear, cable, cooling, and gas peakers that monetize the interim need for firm power while SMR deployments slog through permitting and financing.

The underappreciated risk is that SMR economics are still being validated in public, which means every delay compounds the cost-of-capital problem. In infrastructure, a first-of-kind project tends to define the entire category’s valuation range; a weak first contract or a higher-than-expected delivered cost could compress the multiple far more than investors expect, especially if rates stay elevated and the market continues to punish long-duration, pre-revenue assets.

The contrarian view is that the current selloff may already reflect most of the execution risk, but not enough of the strategic scarcity premium. If a credible anchor customer appears over the next 6-18 months, the stock can re-rate sharply because the market will begin capitalizing a multi-year backlog rather than a prototype; if not, dilution and financing risk dominate and the equity likely behaves like a call option with negative theta. This creates a cleaner relative-value opportunity in the broader power-capex ecosystem than in SMR itself.