The market capitalization of the new players in the U.S. nuclear power sector has evaporated by over $30 billion! SMR represents the future of nuclear energy, but the $2.1 billion short-seller profits have sounded the alarm on valuations and delivery.

According to statistics from data provider S3 Partners reported by local media on Tuesday, short sellers have gained over $2 billion in paper profits over the past year by shorting three leading firms in small modular reactor (SMR) technologyNuScale Power (SMR.US), Nano Nuclear Energy (NNE.US), and Oklo (OKLO.US). Once hot stocks amid an unprecedented AI infrastructure boom, their prices have plummeted over the past year, resulting in a total market capitalization loss of more than $30 billion for the three companies. The substantial profits for short sellers undoubtedly pose a significant short-term negative sentiment and valuation challenge for emerging leaders like Oklo in the nuclear power sector, but this does not equate to a denial of the long-term positive growth logic brought about by the AI infrastructure boom. With short sellers profiting approximately $2.1 billion, combined with around 18% of the float shares of Oklo and NuScale still being borrowed by shorts, it highlights a market shift from "AI needs efficient and clean nuclear power = all nuclear startups should be overvalued" to scrutinizing licensing processes, delivery timelines, and the economics of each SMR. From the perspective of foundational energy engineering, nuclear power supply, especially via the SMR technology route, is indeed one of the most strategically valuable long-term power sources for the AI data centers currently under large-scale construction. High-density AI GPU superclusters demand uninterrupted power supply with a very high load factor, accompanied by strict energy quality and environmental cleanliness characteristics. Nuclear power possesses over a 90% typical capacity factor, extremely high fuel energy density, very low operational carbon emissions, and asset lifespans of decades, significantly reducing data centers' dependence on weather, natural gas pipelines, and long-distance power transmission. Compared to traditional gigawatt-scale nuclear power, the SMR technology route theoretically enables deployment at a lower individual project capital threshold, modularly scaling according to the data center expansion from tens of megawatts to hundreds of megawatts, while being built closer to load centers through factory prefabrication and passive safety features. However, these cost advantages only materialize once the "Nth unit of the same type" enters the mass replication phase; the first unit may, in fact, be the most expensive. With short sellers bankrolling more than $2 billion, the valuation myth of SMR is sharply targeted by bearish actions. Adam Stein, Director of Nuclear Innovation at the Breakthrough Institute, stated on Tuesday that after the "textbook-like super-hype cycle" the industry experienced last year, "the prices and valuations of these stocks have been exaggerated due to speculation." In addition to the S3 Partners showing that bullish activities against the three leading SMR players have made more than $2 billion, another data agency, S&P Global, revealed market intelligence showing that NuScale and Oklo currently have as much as 18% of their float shares in borrowing statusthis is a typical alternative indicator to measure the intensity of short-selling activities, while Nano has nearly 30% of its float shares in borrowing status. In the coming weeks, a key test of investor risk appetite in the nuclear sector is expected to arrive, as U.S. nuclear firms Holtec International and Westinghouse, both of which have small modular reactor businesses, are anticipated to list their stocks in the public market. Large tech giants such as Meta, Alphabet Inc. Class C, Microsoft Corporation, and Amazon.com, Inc. are increasingly turning to still-developing small modular reactor technologies to meet the massive demand for efficient and clean power for data centers in the future. The Trump administration has also publicly supported the nuclear sector, pledging to reduce cumbersome regulations and investing billions in building new reactors and restarting those abandoned old reactors. Under the current and long-term global trend of low-carbon and thorough decarbonization, nuclear energy has become the preferred energy source for tech giants like Amazon.com, Inc., Alphabet Inc. Class C, and Microsoft Corporation due to its efficiency and stability. This energy, which is both clean and stable, promises to provide massive data centers with 24-hour uninterrupted power support. Therefore, the current support from global politicians and tech companies for nuclear energy and nuclear power plants may be stronger than at any time since the 1970s. However, there is still uncertainty about when these not fully validated emerging reactor technologies will be capable of large-scale delivery; some analysts have also expressed concerns about the shortage of high-assay low-enriched uranium (HALEU) supplythis is a special nuclear fuel required for small modular reactors. In response to recent media reports, a representative from Nano stated via email that fluctuations in the company's stock price do not adequately reflect whether its underlying business is successful, and disputed the description of a "hype cycle"; Oklo indicated that the company has made "several concrete and positive advances" over the past year and expects its Aurora reactor to begin efficient commercial operations by 2028. At the end of the AI cycle is power, and at the end of valuation is delivery! The market is shifting from chasing the SMR concept to verifying project progress, orders, and cash flow. Meta, the parent company of Facebook and Instagram, has signed a long-term agreement for up to 6.6 gigawatts with nuclear power companies, including a collaboration with Oklo to develop a project in Ohio with a capacity of up to 1.2 gigawatts, with the first phase planned for operation as early as 2030, demonstrating that the demand from tech giants is real; however, demand contracts do not equate to reactors being approved and delivered on time. The intensity of short-selling is fierce, driven by a severe mismatch between the certainty of nuclear energy demand and the timeline for profit realization for SMR companies. The U.S. currently has no commercially operating SMR; while Oklo has made significant design and testing progress, the Aurora commercial reactor still requires completing the comprehensive licensing, construction, fuel qualification, and grid connection processes, with almost no room for delay in its 2028 target; furthermore, the first-of-a-kind units for these three leading small modular reactor technology firms have yet to prove their mass cost, schedule, and capacity factors. In the second quarter, Oklo generated only about $1.21 million in revenue, with a net loss of around $48.54 million; despite increasing cash and securities to approximately $3.01 billion through share issuance, this also led to equity dilution. More critically, most advanced reactors rely on high-assay low-enriched uranium (HALEU), for which the U.S. Department of Energy admits that domestic supply is still insufficient. With constraints on fuel, nuclear-grade components, regulatory talent, and construction capability, a one-year project delay can significantly lower the present value of long-term cash flow. Thus, the short-sellers' logic focuses on "the market has priced in extreme valuations for dozens of SMR reactors too early," rather than the fundamental outlook of nuclear power. SMRs represent a crucial future technological route for nuclear power, but the high valuation bubble for emerging nuclear companies like Oklo prior to their plummet more closely resembles the volatile risk of venture capital options rather than mature utility assets. Oklo's $3 billion short-term asset liquidity and its collaboration with Meta for SMR deployment significantly reduce short-term financing risks but have not eliminated the risks related to technology implementation, regulation, and construction; assessing Oklo's true value cannot focus solely on orders with tech giants and the AI infrastructure process but must sequentially validate five milestonescommercial licenses, deliverable HALEU fuel, binding power purchase agreements with prepayments or credit support, first unit capital costs and timelines, and the actual grid connection of the first Aurora unit.