Nuclear Stocks That Could 10X Before 2035
2035年前可能涨10倍的核能股票
AI chips growing 170% YoY. Data center footprints doubling every 3 years. The real bottleneck isn't silicon — it's electricity. By 2030, US data centers consume 12,200 TWh/yr. Nuclear is the only zero-carbon, always-on, scalable baseload. Four connected tiers, each depending on the layer below.
AI芯片年增170%。数据中心面积每3年翻倍。真正的瓶颈不是硅——是电力。2030年美国数据中心年耗12,200 TWh。核能是唯一零碳、全天候、可扩展的基础负荷。四层相连，每层依赖下一层。
Tier 1 — Mining & Feedstock
第一层——采矿与原料
180M lb/yr demand vs 140M lb supply. The 40M lb deficit filled from shrinking inventories. New production isn't optional — it's urgent. Without feedstock, the entire fuel cycle stalls.
年需求1.8亿磅 vs 供应1.4亿磅。4000万磅缺口由缩水的库存填补。新产不是选项——是紧迫需求。没有原料，整个燃料链停滞。
Uranium Energy Corp
Uranium Energy Corp
Fully permitted, shovel-ready projects in Texas and Wyoming (acquired from Rio Tinto and Uranium One Americas at rock-bottom prices). In-situ recovery mining — inject solution underground, extract uranium-rich fluid, no rock moved. Can restart production in months, not years. Delivering directly into DOE stockpile programs. Domestic supply = national security priority.
德州和怀俄明州全面许可、随时可开项目（在价格谷底从力拓和Uranium One Americas收购）。原地浸矿——地下注入溶液，提取铀富集液，不搬岩石。数月而非数年即可重启。直接交付DOE储备项目。国内供应=国家安全优先。
NexGen Energy
NexGen Energy
Arrow deposit in Saskatchewan's Athabasca Basin — 4%+ average grade vs 0.2% global average. Production costs under $10/lb. Secured agreements with First Nations communities (critical for Canadian permitting). Arrow could produce 29M lb/yr — ~10% of global mine supply from a single site. Highest grade uranium on Earth.
萨斯喀彻温阿萨巴斯卡盆地Arrow矿床——平均品位4%+ vs 全球平均0.2%。生产成本低于$10/磅。已与原住民社区达成协议（加拿大许可关键）。Arrow可年产2900万磅——单矿占全球供给约10%。地球上最高品位铀。
Energy Fuels
Energy Fuels
Owns White Mesa Mill in Utah — only conventional uranium mill operating in the US. Licensed to process third-party uranium and handle rare earth elements — dual-use strategic asset. Already sold uranium into DOE's new uranium reserve. Positioned as primary US converter for customers avoiding Russian supply chain.
拥有犹他州White Mesa Mill——美国唯一运营的常规铀磨。许可处理第三方铀和稀土元素——双重用途战略资产。已向DOE新铀储备出售铀。定位为避开俄罗斯供应链客户的主要美国转化商。
Cameco — The Blue Chip
Cameco——蓝筹
Second largest publicly traded uranium producer globally. Cigar Lake mine: one of the world's highest grade operations. Controls ~15% of global production. Owns part of Westinghouse Electric — exposure across entire fuel cycle. At $20B+ market cap, priced for stability not explosive growth. The blue chip uranium play.
全球第二大上市铀生产商。Cigar Lake矿：世界最高品位作业之一。控制约15%全球产量。拥有西屋电气部分股权——覆盖整个燃料循环。200亿+美元市值，定价为稳定而非爆发增长。蓝筹铀标的。
Tier 2 — Fuel & Enrichment (The Bottleneck)
第二层——燃料与浓缩（瓶颈）
You can dig up all the uranium you want — until it's enriched into reactor fuel, it's just a rock. 45% of enriched uranium still comes from Russia. The companies rebuilding this capacity aren't just suppliers — they're gatekeepers. The moat is time.
你可以挖出所有铀——在浓缩成反应堆燃料之前，它只是一块石头。45%的浓缩铀仍来自俄罗斯。重建产能的公司不只是供应商——他们是守门人。护城河是时间。
Centrus Energy — The Only HALEU Producer
Centrus Energy——唯一的HALEU生产商
Only US company licensed to produce HALEU (high-assay low-enriched uranium, up to 19.75% U-235 vs standard 5%). First cascade of centrifuges running at DOE's Pike, Ohio facility. First commercial batch late 2023. Scaling to 6 metric tons HALEU/yr by 2025. Without Centrus, no SMR from Oklo, NuScale, or utilities can even start. Single point of failure = single point of leverage.
美国唯一获许生产HALEU的公司（高含量低浓缩铀，高达19.75% U-235 vs 标准5%）。首批离心机在DOE俄亥俄州Pike设施运行。2023年末首批商业批次。2025年扩展至年产6公吨HALEU。没有Centrus，Oklo、NuScale或任何公用事业的SMR都无法启动。单点故障=单点杠杆。
BWX Technologies — Fuel + Fabrication
BWX Technologies——燃料+制造
If Centrus makes enriched fuel, BWX turns it into usable reactor cores. Fabricates nuclear components for US military and private sector, including TRISO fuel (uranium kernel wrapped in carbon and ceramic — withstands extreme heat, no meltdown). Under DOE and DoD contract at Lynchburg, VA facility. Defense contracts fund R&D; commercial side just beginning — same TRISO pellets will power micro-reactors for AI data centers. Also runs component manufacturing for submarines and naval reactors. Fuel + fabrication + recurring revenue.
Centrus生产浓缩燃料，BWX将其制成可用反应堆堆芯。为美军和私营部门制造核组件，包括TRISO燃料（铀芯包裹碳和陶瓷——耐极端高温，不熔毁）。在弗吉尼亚Lynchburg设施执行DOE和国防部合同。国防合同资助研发；商业端刚开始——同样TRISO颗粒将为AI数据中心微型反应堆供能。也制造潜艇和海军反应堆组件。燃料+制造+经常性收入。
Tier 3 — Innovation Reactors (Moonshots)
第三层——创新反应堆（登月项目）
High risk, extreme upside. Burning through cash, waiting for regulatory approval. If they execute: 5–10×. If they fail: could go to zero.
高风险、极端上行。烧现金等审批。成功：5-10倍。失败：可能归零。
Oklo — Nuclear as a Service
Oklo——核能即服务
Micro-reactor company backed by OpenAI's Sam Altman. Aurora: 15 MW continuous, enough for a large AI data center. Runs 10 years without refueling. No moving parts, no water-cooled core, minimal waste. Don't sell reactors — own and operate them like mini power plants. Targeting private customers, hyperscalers wanting 24/7 clean power. Conditional site use permit from DOE at Idaho National Lab. First operational reactor targeted 2027. Up 500%+ YTD.
OpenAI Sam Altman支持的微反应堆公司。Aurora：15MW持续功率，足够大型AI数据中心。10年不换燃料。无活动部件、无水冷堆芯、极少废物。不卖反应堆——像小型电站一样自持运营。面向私营客户和要24/7清洁能源的超大规模。DOE爱达荷国家实验室有条件用地许可。2027年首台目标运行。年内涨500%+。
NuScale — First & Only NRC-Approved SMR
NuScale——首个也是唯一NRC批准的SMR
In 2023, became the first and only SMR design approved by the US Nuclear Regulatory Commission. That's a massive regulatory moat. 77 MW per module, cluster up to a dozen for nearly a GW. Factory-produced, shipped by rail, installed in under 4 years (vs 10–15 for traditional). First project Voyager with Utah municipal power, backed by DOE. Exploring Eastern Europe deployments. Every SMR needs HALEU fuel — Tier 2 and Tier 3 move in lockstep.
2023年成为首个也是唯一获美国核管理委员会批准的SMR设计。巨大的监管护城河。每模块77MW，可集群至12个近1GW。工厂生产、铁路运输、4年内安装（vs传统10-15年）。首个项目Voyager与犹他市政电力合作，DOE支持。探索东欧部署。每个SMR都需要HALEU——第2层和第3层同步运行。
Tier 4 — Operating Utilities (Stability Anchor)
第四层——运营公用事业（稳定锚）
Established companies with decades of nuclear operations, extended reactor licenses, and now hyperscaler partnerships. Cash flow positive, dividend paying, expanding into SMRs.
拥有数十年核运营、延长反应堆许可、现与超大规模合作的公司。现金流正、分红、扩展SMR。
Dominion Energy — The Amazon Deal
Dominion Energy——亚马逊协议
3 major nuclear plants across Virginia and Connecticut. Reactor licenses extended into the 2060s. Just signed deal with Amazon to explore deploying SMRs in Virginia — first hyperscaler-utility nuclear partnership on record. Cash flow positive, dividend paying, already nuclear-heavy. The stability anchor baseload you can bank on.
弗吉尼亚和康涅狄格3座主要核电站。反应堆许可延至2060年代。刚与亚马逊签约在弗吉尼亚部署SMR——首个超大规模-公用事业核能合作记录。现金流正、分红、已是核能重仓。可依赖的稳定锚基础负荷。
Duke Energy — 10% of US Nuclear
Duke Energy——美国核电10%
11 nuclear units across 6 sites in the Carolinas, 10,000+ MW generation — roughly 1/10 of the entire US nuclear fleet. Extended Harris Nuclear Station through 2050s. 2025 resource plan includes new reactors and SMRs for data center/AI demand across the Southeast. Regulated, reliable, ready to scale nuclear again. The income engine.
卡罗来纳6个场地11台核电机组，1万+MW发电——约占全美核电1/10。Harris核电站延至2050年代。2025年资源规划包括新反应堆和SMR应对东南部数据中心/AI需求。受监管、可靠、准备再次扩核。收入引擎。
$100 Risk-Weighted Allocation
100美元风险加权配置
Not equal-weighted — risk-weighted. Most capital where probability of success is highest. Meaningful exposure to the bottleneck (fuel). Enough in moonshots for outsized gains.
不是等权——是风险加权。最多资本投在成功概率最高的地方。对瓶颈（燃料）有意义敞口。登月项目足以捕获超额收益。
10-Year Outlook
10年展望
Mining: 2–3× from higher uranium prices. Fuel/Enrichment: 3–5× (tightest bottleneck = biggest multiples). Innovation Reactors: 5–10× if execute, zero if fail. Utilities: 2–3× + dividends. $100 → $300–$600 in 10 years.
矿采：铀价上涨2-3倍。燃料/浓缩：3-5倍（最紧瓶颈=最大倍数）。创新反应堆：成功5-10倍，失败归零。公用事业：2-3倍+分红。$100→$300-$600 10年。