A Word of Caution on NuScale Power Before You Buy the Dip

A University of Colorado Boulder study estimates NuScale SMR operating costs above $93/MWh—$17.42/MWh for fuel, $3.55/MWh variable O&M, and $72.26/MWh fixed O&M—versus historical power prices of roughly $42-$45/MWh. The non-peer-reviewed analysis suggests NuScale reactors may not recover costs over their 60-year lives even with zero construction cost. The findings add to execution concerns after NuScale's Carbon Free Power Project cost estimate rose from $4.2B in 2018 to $9.3B in 2023 before its November 2023 cancellation.

Analysis

The key valuation issue for SMR is not a near-term earnings miss but whether its addressable market can clear a bankability threshold. If all-in operating cost is structurally above wholesale power prices, project finance will require long-dated, above-market offtake contracts, regulated cost recovery, or federal subsidies; each shifts the business from scalable technology vendor toward politically dependent infrastructure developer. That would compress the revenue multiple investors assign to SMR well before commercial deployment data arrive.

Near-term downside catalysts are financing terms, revised customer economics, and any delay or scope reduction in the TVA/Romanian development pipeline. Over the next 1-3 months, the stock remains vulnerable because a non-peer-reviewed academic estimate can prompt investors to re-underwrite assumptions, while management has limited operating data with which to rebut it. The thesis is falsified by independently verified operating-cost guidance materially below the cited range, a binding power purchase agreement at an economic tariff, or an EPC/financing structure that transfers cost-overrun risk away from customers.

OKLO is not a clean long beneficiary: it faces the same fixed-cost and utilization challenge, although its fuel-cycle strategy and prospective data-center contracts could support premium power pricing unavailable to merchant generation. The more important second-order effect is that AI-linked nuclear demand increasingly favors incumbent, dispatchable assets and large-reactor life extensions—Constellation Energy (CEG) and Vistra (VST)—because they can monetize capacity now while SMR developers must still prove construction, licensing, and operations simultaneously.

Key Decisions for Investors

• Maintain/establish a 1-3 month short SMR versus long CEG: the pair isolates execution and financing risk from the broader nuclear/AI-power narrative. Size for high short-interest volatility; cover if SMR secures a binding, creditworthy offtake contract with disclosed economics or if the relative spread moves 20% against entry.
• Do not add to OKLO solely on SMR weakness; place an alert for disclosed customer power prices, fuel-cost assumptions, and plant utilization targets. A long is actionable only if contracted tariffs demonstrably exceed modeled operating plus capital-recovery costs.
• For nuclear exposure over 6-18 months, favor CEG or VST over pre-revenue SMR developers: existing generation converts tight power markets into cash flow, whereas SMR valuations require multiple unproven milestones. Reassess if power-price forward curves weaken materially or nuclear life-extension policy reverses.
• Watch TVA and Romanian milestones over the next two quarters. Any funding gap, delayed final investment decision, or customer renegotiation is a catalyst to add to SMR shorts; conversely, fixed-price EPC backing from a credible counterparty would reduce conviction.