Chronometer’s Smith Warns Data Centers, LNG Exports Face 2028 Gas Deficit

Chronometer的Smith警告：数据中心与LNG出口将在2028年面临天然气短缺 来自AI算力和已承诺出口终端的前所未有的需求将超出美国天然气生产能力，恐将在2030年前耗尽工作储气量并引发剧烈的价格暴涨。 在本期节目中，Matthew Smith详述了一项论点，预测美国将从2028年开始面临严重的结构性天然气短缺。在预先承诺的LNG出口履约要求与迅速扩张的AI算力负荷的双重推动下，新增天然气需求将超出全美128至132 BCF/d的最大输送能力上限。在当前市场定价反映出平坦的远期曲线且普遍存在自满情绪的背景下，Smith警告称工作储气量可能在2030年前耗尽，从而导致无上限的价格暴涨、更高的居民用电成本以及地缘政治溢出效应。为避免这一危机，他主张立即转向大规模核电（AP1000）、扩大住宅太阳能并建设新的加拿大进口管道。 受AI数据中心电力需求与已承诺液化天然气（LNG）出口增长的双重拉动，预计美国自2028年起将遭遇前所未有的结构性天然气短缺。 (0:00) 美国已承诺的LNG出口能力计划在2030年前从15 BCF/d跃升至35 BCF/d，几乎将消耗该国最大增产能力（+20 BCF/d）的100%。 (4:56) 在基准情景下，高确定性的AI算力项目将在2028至2030年间增加至少5 BCF/d的天然气需求，导致国内天然气供应毫无余量。 (4:56) 由于存在具有约束力的长期合同，且全球盟友高度依赖美国的供应，削减美国LNG出口在实际和法律层面上均不可行。 (9:33) 美国天然气面临的核心问题在于输送能力与时间节奏的瓶颈，而非地下实际资源总量不足，因为核心一级（tier-1）钻探储备正在迅速耗尽。 (11:38)

Matthew Smith of Chronometer Partners argues that the simultaneous expansion of committed U.S. liquefied natural gas (LNG) exports and AI data center power demand will trigger an unprecedented natural gas supply deficit by 2028. Without rapid intervention, this imbalance risks exhausting U.S. working gas storage by 2030, creating convex price spikes that will fall heavily on American utility consumers. Structural Deficit from LNG and AI Demands (4:56) Smith models that firmly scheduled U.S. LNG export capacity will jump from roughly 15 billion cubic feet per day (BCF/d)—about 12% to 15% of the current 110–112 BCF/d of national production—to 35 BCF/d by the end of 2030. Granular modeling across the Permian, Haynesville, and Appalachian basins indicates that the country can realistically add at most 20 BCF/d of new supply over this period. Consequently, committed export obligations alone will consume virtually all incremental domestic production capability, leaving zero spare gas for growing domestic electric power loads. Layering AI compute demand onto this balanced baseline generates an immediate structural shortfall. Incorporating only high-probability "P50" data center projects—those with site control, power purchase agreements, and queue positions—adds at least 5 BCF/d of incremental gas demand by 2028–2030, with unmitigated demand potentially reaching 12 to 15 BCF/d. Mandates requiring hyperscalers to "Bring Your Own Generation" (BYOG) or deploy distributed assets, such as Bloom Energy fuel cells consuming 150 MMcf/d per gigawatt, will force data centers to compete for fuel that the current supply architecture cannot supply. Meanwhile, curtailing LNG shipments is legally and geopolitically unfeasible due to binding long-term financing and international allied dependence on U.S. gas. Physical Wellhead and Midstream Bottlenecks (18:20) The supply constraint is not an outright absence of underground resources, but a physical deliverability bottleneck driven by steep decline curves and depleting top-tier inventory. Smith argues that corporate claims of multi-decade inventory in major basins are overstated, projecting that many producers will exhaust their premium drilling locations within four to five years. Under aggressive development assumptions, aggregate U.S. gas deliverability caps out at roughly 128 to 132 BCF/d. Permian associated gas cannot easily close the gap, as the 7+ BCF/d of planned pipeline expansions through 2030 is already factored in, and drilling additional oil wells depends heavily on global crude prices. Midstream infrastructure poses an equally rigid bottleneck. Raw wellhead gas requires processing facilities to remove impurities and adjust heat content to the standard 1,010–1,030 BTU pipeline specification, requiring two- to three-year construction lead times. Gathering lines require extensive capital linked to specific well coordinates, while interstate transmission faces severe regulatory resistance. Over the past decade, the Mountain Valley Pipeline was the only major interstate natural gas pipeline completed in the United States. Storage Exhaustion and Market Complacency (22:21) Despite impending supply tightness, market participants remain complacent after fifteen years of cheap shale gas. Forward price curves remain flat in the mid-$3 range through 2030, encouraging hyperscalers to contract power without confirming physical fuel deliverability. If gas prices escalate, Smith estimates that energy could jump from roughly 10% to 20–30% of a hyperscaler's total cost of compute by 2029. Because production cannot quickly flex upward, the structural deficit will be absorbed by U.S. working gas storage, which has a total capacity of approximately 4 trillion cubic feet (TCF). Under Chronometer's model, seasonal storage draws will breach historical norms by 2028–2029 and face complete physical exhaustion by 2030. When storage buffers are depleted, natural gas pricing ceases to follow linear cost curves and enters unbounded, convex spikes. Market Beneficiaries, Nuclear Realities, and Consumer Costs (26:11) Because natural gas operates as the marginal fuel setting wholesale power prices in U.S. competitive markets, elevated gas prices will directly lift electricity prices. Upstream producers with core acreage inventory—such as Expand Energy, which holds about 70% of core Haynesville inventory, and Range Resources in Appalachia—stand to benefit substantially. Zero-marginal-cost renewable power operators, including utility-scale solar asset owners like Clearway Energy and XIFR, are similarly positioned to expand margins as existing power purchase agreements re-contract at elevated rates. Nuclear power cannot avert the 2030 bottleneck. Small modular reactors (SMRs) lack commercial manufacturing scale for the current decade, while proven gigawatt-scale designs like the Westinghouse AP1000 face operational lead times reaching 2033–2034, even though supply-chain partners like Cameco and BWX Technologies stand to gain from eventual buildouts. Consequently, Smith notes that the American consumer will ultimately bear the brunt of the gas trilemma through dramatically higher retail electricity bills. Smith's projection of an unavoidable gas bottleneck hinges on whether forward markets reprice before storage depletion forces acute physical shortages. Significant uncertainties include whether policy makers will intervene in export contracts, whether advanced non-lithium battery storage can commercialize rapidly, and whether pipeline projects to wheel trapped Canadian natural gas into the U.S. Midwest can overcome midstream permitting barriers.

Chronometer Partners投资人Matthew Smith认为，受已承诺的液化天然气（LNG）出口激增与AI数据中心电力需求的双重拉动，美国天然气市场将在2028至2030年间遭遇前所未有的结构性供给赤字，甚至面临耗尽可用库存与引发电价暴涨的系统性风险。 LNG出口与AI算力夹击下的供给天花板 (4:56) Smith指出，美国天然气日产量目前约为1100亿至1120亿立方英尺（BCF/d），受核心区块枯竭和钻井效率曲线限制，到2030年全国关键盆地累计最多只能增加约20 BCF/d的产能，总供给物理上限被压缩在1280亿至1320亿立方英尺。然而，由于长周期审批与已锁定的项目融资，美国LNG出口名义产能已确定由目前的15 BCF/d攀升至2030年的35 BCF/d，净增的20 BCF/d将全额吸收全国全部新增天然气产能。 在出口需求已挤占全部增量的基准下，AI算力负荷构成了额外的供给缺口。根据对具备土地审批、购电协议（PPA）和电网互联申请的“P50”高确定性项目的统计，AI数据中心到2028至2030年将至少带来5 BCF/d的额外天然气需求，若包含更广泛方案，该需求可能升至12至15 BCF/d。例如，单套Bloom Energy 6系列燃料电池每吉瓦（GW）即消耗1.5亿立方英尺/日（MMcf/d）天然气。即便考虑二叠纪盆地（Permian）现存规划中的逾7 BCF/d外输管道，伴生气增产也高度受制于原油钻探经济性，无法独立填补缺口。 中游基建卡脖子与2030库存枯竭危机 (18:20) 地表下储量并不等同于地表输送能力。Smith强调，新井产出的原料气必须经过脱除杂质并处理至1010至1030 BTU的管道标准，而新建天然气处理厂通常需要2至3年周期；同时，跨州长输管道受严格环境监管制约，过去十年间仅有一条大型跨州管道（Mountain Valley Pipeline）成功投运。此外，由于受制于具有法律约束力的数十亿美元长期供货协议与国际盟友的能源依赖，美国在法律与地缘政治上无法单方面切断LNG出口来保供国内。 当前金融衍生品远期曲线在2028至2030年间仍平坦维持在每百万英热单位3.5美元左右，反映出市场长达十五年页岩繁荣后的盲目乐观。Smith的模型预测，美国约4万亿立方英尺（TCF）的商业工作储气库将在2028年年中跌破季节性历史均值，并在2029年击穿历史最低位；到2030年，在现有基准不变的情况下，储气库面临彻底耗尽的风险。届时，现货气价将出现非对称的凸性暴涨，并直接传导至电价端，美国普通消费者将承担飙升的电费账单。 产业链赢家与被低估的算力成本错配 (26:11) 当天然气作为边际机组决定全网批发批发电价时，气价上涨将给零边际成本的绿电资产带来巨额超额利润。公用事业级光伏运营商（如XIFR、Clearway Energy）在PPA重签过程中拥有显著的利润扩张空间；户用光伏与储能系统也将因居民防范日间高峰电价而迎来加速普及。在上游勘探与开采（E&P）领域，掌握优质存量核心资源的头部企业处于优势地位，如持有Haynesville核心存量约70%的Expand Energy，以及阿巴拉契亚产区的Range Resources，但目前在二级市场仅以约4倍EBITDA的低估值交易。 相比之下，依赖分布式天然气发电以缩短并网等待时间（time-to-power）的模式面临严重燃料风险。超大规模云厂商推行“自带电源”（BYOG）以及Caterpillar、Bloom Energy等扩大发电机组产能的举措，正加剧局部燃料争夺。Smith测算，随着气价翻倍或数倍上涨，能源在数据中心总运营成本中的占比可能从目前的约10%飙升至2029年的20%至30%甚至40%，这将严重侵蚀低估燃料供应链复杂度的算力运营商盈利水平。 长期脱困路径与万亿美元级电网权衡 (30:08) 为了在2030年后从根本上解决基荷电力缺口，Smith主张必须回归大规模商业核电建设。小型模块化反应堆（SMR）由于缺乏规模化制造能力，在2030年前难以形成实质性产能，更具可行性的是西屋AP1000等百万千瓦级成熟压水堆设计。他建议美国联邦政府直接出资并全流程承建2至4座AP1000机组以理顺供应链并撬动民间投资，使新建机组能在2033至2034年投入运行。产业链上的Cameco（持有西屋49%股权）与BWX Technologies将成为核电复苏的直接受益者。 在政策层面，Smith还呼吁重新强力补贴居民屋顶光伏以压低日间峰值负荷，并修建自加拿大向美国中西部（MISO等电力市场）输送1至2 BCF/d的跨国输气管道，以激活加拿大丰富但受困于管道的深层低成本天然气资源。若不采取激进干预措施，美国将在维持全球LNG供应信誉、满足AI算力扩张与控制国内民生用电通胀之间陷入不可调和的“三难困境”。 该分析高度依赖于AI数据中心落地规模按期兑现、LNG出口终端全负荷投运以及上游勘探地质衰减模型的推演。若AI基础设施资本开支出现大幅收缩、美国监管层强制干预出口合同，或者非锂长时储能技术实现商业化突破，天然气供需错配爆发的时间节点与烈度可能会被重塑。

## Background 本期节目由知名商业与投资播客《Invest Like the Best》（Colossus 旗下频道）于 2026 年 7 月 21 日发布在 RSS 及网页端。播客简介指出，嘉宾 Matthew Smith（Chronometer Partners 创始人兼 CIO）在过去 18 个月带领团队对全美几乎所有天然气井、管道与加工设施进行了详尽的底层建模，并得出一个市场主流尚未意识到的惊人结论：从 2028 年起，AI 数据中心与 LNG 出口的新增需求将彻底超出美国的生产与输送极限，甚至可能在 2030 年耗尽全美有效储气库，导致气价面临无上限且呈凸性的暴涨风险。 ## TL;DR 通过对全美天然气开采、中游管网及地面加工设施的原子级建模，Matthew Smith 指出美国的天然气供给已逼近物理天花板，在既定 LNG 出口翻倍与 AI 算力激增的双重冲击下，全美供需平衡将在 2028 年急转直下，并在 2030 年耗尽工作储气库。由于长期出口合同与地缘政治约束，美国无法轻易关停 LNG 出口，而当前市场不仅对 2028 年后的供给缺口毫无定价，甚至缺乏实物对冲与锁定意识。随着天然气作为全美边际定价燃料价格失控，普通消费者的电费账单将承受重压，科技巨头亦将面临能源成本占算力开销飙升至 20% 到 40% 的盈利危机；在此背景下，上游优质气商与公用事业级光伏将成为赢家，而大规模商用核电（如 AP1000）与户用光伏则是唯一可行的长期与防御性解药。 ## Atmosphere & style 对谈氛围严谨、深邃且充斥着一种令人清醒的紧迫感。两位对谈者如同在平静展开一场硬核的工程与金融沙盘推演，没有浮夸的末日论煽动，而是凭借详实的地质数据、管网物理参数与严密的合同逻辑层层抽丝剥茧。整期节目传递出一种强烈的反差张力——一边是资本市场对 AI 繁荣与无限廉价燃料的麻木自满，另一边则是物理世界即将撞墙的冰冷现实，宛如目睹一场正在“起初极慢、随后骤然爆发”的宏大能源危机逼近眼前。 ## Bulletin - `03:33` Chronometer Partners 历时18个月的原子级建模显示，在既定 LNG 出口爬坡与 AI 数据中心用电的双重抽水下，美国自 2028 年起将遭遇前所未有的天然气结构性短缺，全美工作储气库（约 4 Tcf）预计将在 2030 年见底枯竭。 - `04:56` 即便全美所有高品位井位全力投产，天然气日交付能力上限仅为 128 至 132 Bcf/d，而规划中的 LNG 出口到 2030 年将吃掉 35 Bcf/d，在 AI 算力介入前新增产能就已被海外出口长约基本锁死。 - `10:00` 受长达 15 至 20 年的照付不议合同及欧洲、亚洲盟友能源安全约束，美国无法轻易行政下令切断 LNG 出口，一旦储气库耗尽，Henry Hub 天然气价格上行风险将呈现无上限且高度凸性（unbounded and convex）的特征。 - `37:57` 超大规模科技巨头（Hyperscalers）仍普遍套用 3.50 美元/MMBtu 的平坦远期曲线做成本测算，未对冲 2028 年后的物理断供风险；届时能源成本占算力总营运成本的比重可能从当前的约 10% 骤升至 20%~40%。 ## Key takeaways - `03:33` 天然气已占全美发电量 40% 以上并正在取代石油成为核心战略燃料，2026-2027 年的供需平衡将在 2028 年急转直下，并在 2030 年演变为波及全美的电价危机。 - `04:56` 供需失衡的根源早在 AI 爆发前就已埋下：已获批并规划的 LNG 出口能力到 2030 年达 35 Bcf/d，而全美天然气外输增产极限最多仅约 20 Bcf/d，出口端在 AI 介入前就已吸干所有增量空间。 - `07:24` 在 P50 基准情景下，AI 算力将带来约 5 Bcf/d 的新增刚性用气需求；若计入早期概念及激进的表后发电（BTM）假设，需求缺口将在 2030 年代初扩大至 12~15 Bcf/d。 - `10:00` LNG 出口背后绑定着国际项目融资长约与地缘战略承诺（特别是欧洲对美气的依赖），单方面违约切断出口在法律与外交层面均寸步难行。 - `15:15` 天然气面临的是工业系统交付的“流量（Flow）”物理天花板，而非地质“存量（Stock）”缺乏：全美最大日交付能力极限仅为 128~132 Bcf/d，且大部分优质库存井位将在 4 至 5 年内耗尽。 - `20:00` 地面中游基础设施构成致命卡点：天然气加工厂建设周期需 2-3 年且目前产能滞后，而跨州干线管道在过去十余年实质上仅建成了一条 MVP 管道。 - `22:21` 金融市场对 2028 年以后的供给断崖极度麻木，天然气远期曲线直至 2030 年均走平，生产商甚至在现阶段主动关井限产，未进行任何前瞻性资本开支。 - `25:00` 一旦约 4 Tcf 的全美工作储气库被彻底掏空，结构性刚需将使气价呈现不可预测的非线性暴涨，过往寒潮及地缘冲突导致的 8~10 美元高价将成为常态甚至被轻易突破。 - `27:00` 由于天然气是全美电网的边际定价机组，边际气价暴涨将推高整体电价，燃料成本为零的公用事业级光伏运营商及掌控优质储层的上游气商（如 Expand Energy）将成为最大赢家。 - `30:08` SMR 小型反应堆尚处于商业化概念与科学实验阶段，规模化远水难解近渴；依托西屋 AP1000 路线的大型商业核电机组是 2030 年代唯一的确定性底荷解法。 - `34:29` 普通美国消费者将承担最终代价，被迫在民生电费飙升与工业/算力用电挤兑中买单；各州推行的“自带发电机（BYOG）”本质上只是把表外用气隐蔽化，进一步加剧气源紧张。 - `36:37` 燃气轮机与分布式发电机组制造商（如 Caterpillar旗下Solar Turbines、Bloom Energy）正逆势在周期高点扩产，很可能在 2028-2029 年遭遇“有发电机组却无气可烧”的订单断崖。 - `40:00` 二叠纪盆地的伴生气增产受制于原油价格逻辑与管道外输上限（2026-2030 年仅新增 7+ Bcf/d），现场数据中心就地发电只会减少外输气量，无法解决全国性赤字。 - `44:25` 化解危机的政策抓手在于联邦政府直接出资托底建造数座 AP1000 示范反应堆以激活供应链，同时在居民侧大力补贴屋顶分布式光伏以削减白昼尖峰负荷。 - `50:00` 当前资本市场正疯狂涌向 2035 年的远景概念，却以 4 倍 EBITDA 的超低估值冷落当下具备充沛现金流的上游气商；未来 6 个月内行业将被迫打响争夺 2028 年现货气源的“白刃战”。 ## Detailed walkthrough ### `00:00` 节目序言与重磅嘉宾介绍 本期《Invest Like the Best》的主持人 Patrick O'Shaughnessy 迎来了时隔数年再次做客的老朋友 Matthew Smith。Matthew 是 Chronometer Partners 的创始人兼首席投资官（CIO），在能源、工业、原材料、电力及公用事业等基础设施投资领域深耕已逾二十年。Patrick 评价他拥有百科全书式的行业洞见，对能源市场的理解极其深邃。在过去的 18 个月里，Matthew 带领团队完成了极具野心的自底向上式底层建模：他们几乎扫描并模拟了全美境内的每一口天然气井、每一条输气管道和每一处加工处理设施，意在摸清在 AI 算力与数据中心催生庞大电力需求的大背景下，美国能源底座的真实承压能力。 这项耗时一年半的扎实研究所得出的结论不仅与当前华尔街主流市场的共识大相径庭，甚至令人不寒而栗。Matthew 发现，自 2028 年起，美国将遭遇前所未有的天然气供应短缺；到了 2030 年，这一结构性缺口很可能彻底耗尽全美的天然气有效工作储气库（Working Gas Storage）。用他的原话来说，届时天然气价格的上涨风险将变得“无上限且呈现高度凸性（unbounded and convex）”，并不可避免地演变为一场席卷全美的电力价格与生活成本危机。 --- ### `03:33` 历时18个月的研究结论：天然气供需的历史性巨震 当 Patrick 邀请 Matthew 开门见山抛出研究的核心结论时，Matthew 明确指出：美国正不可逆转地滑向一场历史级别的天然气供应短缺，这将在全社会范围内引发剧烈连锁反应。当前天然气发电量已占全美总发电量的 40% 以上，考虑到日常发电消耗的天然气规模，它即将彻底取代石油，加冕为美国本土最重要的战略燃料。 根据 Chronometer Partners 的精确测算，虽然在 2026 至 2027 年间全美天然气供需尚能勉强维持动态平衡，但一旦时间推进到 2028 年，形势将急转直下。一方面，庞大的 AI 算力负荷被迅速接入电网，定向锁定具体发电设施的用气配额；另一方面，既定规划中的液化天然气（LNG）出口项目正开足马力向海外输送。两者双重挤压之下，充当全美天然气供需调节枢纽与安全缓冲垫的有效工作储气库将被持续吞噬。这会导致天然气价格爆发非线性的极端上涨，而普通家庭与工商业企业最先感受到重击的阵地，正是 2028 至 2030 年间一路狂飙的电费账单。 ```mermaid flowchart TD accTitle: 美国天然气供需失衡与储气库耗竭演进路径 accDescr: 展示从2026年供需平稳到2028年AI与LNG双重抽水，最终导致2030年储气库耗竭与电价暴涨的逻辑流程。 subgraph Phase1["2026-2027：勉强平衡"] S1["国内产气与管道增量"] --> B1["供需基本匹配"] D1["现有LNG出口 + 传统用电"] --> B1 end subgraph Phase2["2028：供需拐点与断层"] LNG["既定LNG出口翻倍爬坡"] --> Deficit["气源供给出现缺口"] AI["AI算力数据中心并网用气"] --> Deficit Deficit --> Drain["持续消耗地下工作储气库 (Working Storage)"] end subgraph Phase3["2030：储气耗竭与价格失控"] Drain --> Empty["有效储气库被彻底掏空"] Empty --> Crisis["天然气价格无上限飙升 (Unbounded & Convex)"] Crisis --> PowerHike["终端电力价格暴涨，演变成全美能源危机"] end Phase1 --> Phase2 --> Phase3 ``` --- ### `04:56` 早在 AI 登场之前，大局就已注定 许多人直觉地将这场能源困局归咎于 ChatGPT 和突然爆发的 AI 热潮，但 Matthew 直言：“早在 AI 算力登上历史舞台之前，命运的齿轮就已经注定（The die was cast long before AI compute came to the scene）。”回顾历史，在 2010 年页岩革命爆发之前，美国本土天然气根本无法自给自足，必须依赖进口来填补消费缺口。随后页岩气产量的爆发式增长彻底逆转了格局，不仅让本土天然气陷入极度过剩与廉价的红利期，还催生了庞大的 LNG 跨洋出口产业，其先驱便是 Cheniere Energy。 短短十余年间，全美 LNG 的名义出口能力已从零飙升至目前的约 150 亿立方英尺/日（BCF/d）。对比全美目前每日 1,100 亿至 1,120 亿立方英尺的总产量，出口量已经吃掉了本土供给能力的 12% 到 15%。随着廉价资源的持续繁荣，大量新增出口终端项目陆续获批并开工上马。截至目前，已排定日程并获得许可的 LNG 规划出口能力将在 2030 年底前达到惊人的 35 BCF/d。 - **LNG 项目审批与建设周期的刚性**：此类超大型基础设施采用项目融资（Project Financing）模式，前期的选址、环境审批与长协签署通常需提前数年敲定，目前支撑出口规模从 15 BCF/d 迈向 35 BCF/d 的主力项目早已木已成舟。 - **天然气新增产能的物理天花板**：根据 Matthew 团队对阿巴拉契亚（Appalachia）、海恩斯维尔（Haynesville）及二叠纪盆地（Permian）全部已建与规划气井、集气管线及外输能力的摸底，全美整体外输及开采增产极限最多仅为 20 BCF/d 左右。 - **AI 介入前的收支相抵**：在完全不考虑任何 AI 算力新增负荷的前提下，美国未来新增的所有天然气产能极限，几乎已被既定的 LNG 出口增量全额消耗殆尽，供给端毫无冗余空间。 --- ### `07:24` 量化 AI 的新增用气需求：从 P50 基准到极端假设 在出口端将本土新增天然气榨干的同时，AI 数据中心对“即时可用电力（Time to Power）”的极度饥渴，为本已脆弱的供需天平砸下了另一枚巨型砝码。为了满足算力爆发的供电需求，科技巨头与独立电力生产商正在部署形形色色的技术方案——从最顶层、最高效的大型公用事业级燃气联合循环电站（如 GE Vernova 的 9HA/7HA 级重型燃气轮机），一直下探到分散式的“表后发电（Behind-the-Meter）”方案，包括 Bloom Energy 的固态氧化物燃料电池（SOFC）、瓦锡兰（Wärtsilä）的大型往复式内燃发电机，以及卡特彼勒旗下 Solar Turbines 制造的小型燃气轮机。 为了准确测算真实消耗量，Matthew 的团队与专业外部机构合作，根据审批进度、购电协议（PPA）签署情况以及电网互联协议（Interconnection Agreement）的进展，对市面上流传的海量发电项目赋予了实现概率。在他们的基准情景（Base Case，即 P50，只计入具备清晰合同且概率在 50% 以上的高置信度项目）中，仅由 AI 算力驱动的天然气新增刚性需求就达到约 5 BCF/d。 | 评估情景 / 技术路线 | 实现概率及前提条件 | 对应天然气新增需求 | 关键物理参数与行业现状 | |:--- |:--- |:--- |:--- | | **基准情景 (Base Case)** | **P50 及以上**（已签署 PPA、取得部分许可、互联流程在途） | **~5 BCF/d** | 已经足以打破没有 AI 时的供需紧平衡状态 | | **激进 / 极端情景** | **P30 甚至 P0**（包含早期概念规划及所有规划中的表后发电） | **12 ~ 15 BCF/d**（到 2030 年代初） | 若缺乏有力调控，供需裂口将彻底演变成灾难 | | **燃料电池单项负荷示例** | Bloom Energy 第 6 代固态氧化物燃料电池（SOFC） | **1.5 亿立方英尺/日 / GW** (150 MMCF/d per GW) | 市场普遍预期其年产能达 2 GW 并扩产至 5 GW；若无气源置换，天然气根本无法承载此等消耗 | --- ### `09:33` 为何不能“一刀切”停止 LNG 出口？ 面对数额看似只有 5 到 15 BCF/d 的算力缺口，市场中常有一种极其幼稚的直觉论调：“既然全美面临电价暴涨、百姓民怨沸腾，为什么不直接行政下令切断海外出口？反正页岩革命前美国本来就不出口天然气，把天然气留给国内自用不就万事大吉了吗？” Patrick 将这一普遍的民间疑惑抛给 Matthew。Matthew 苦笑指出，现实远非大笔一挥关掉阀门那么简单。大型跨国 LNG 贸易深深扎根于神圣不可侵犯的**商业契约法（Contract Law）**体系之下，背后绑定着长达 15 至 20 年的严格照付不议（Take-or-Pay）采购长约与复杂的国际地缘政治结盟机制。随意撕毁出口合同不仅会在国际商业社会引爆空前的司法诉讼与违约索赔风暴，更将直接摧毁全球盟友（尤其是摆脱俄气依赖后的欧洲与部分亚洲盟友）的能源生命线。这种粗暴的民粹式方案看似诱人，在现实操作中却寸步难行。 ### `10:00` 无法轻易切断的 LNG 出口：合同约束与地缘政治锁死 面对迫在眉睫的供需矛盾，最直观的疑问是：既然国内可能缺气，为什么不直接叫停或减少液化天然气（LNG）出口？Matthew Smith 指出，这种想法忽略了商业合同与国际现实的极端复杂性。这些出口设施背后绑定着数百亿美元的项目融资协议与长期的照付不议（Take-or-Pay）合同。更关键的是，在未来数年内，美国将占到全球天然气供应量的近三分之一；无论是欧洲等关键盟友，还是签订了自由贸易协定（FTA）乃至非 FTA 的贸易伙伴，其国家能源安全都在深度依赖美国的天然气出口。单方面切断供应不仅在法律上面临巨大的违约成本，在外交层面也将引发灾难性后果。 在 Chronometer Partners 建立分析基准时，首要原则就是尊重现实约束和法律条文——“从签署的合同开始，看纸面上写了什么、允许什么、禁止什么”。如果假设合同法依旧有效，那么 LNG 码头的出口规模就几乎是一条不可逆的刚性需求曲线。理论上，除了强行切断出口，另一个缓解系统压力的杠杆是“放缓 AI 算力的扩张速度”。但面对科技巨头无法抑制的算力饥渴，放缓 AI 显然是一个极其不受欢迎且极不现实的提议。因此在基准假设中，这两大需求力量都将全速狂飙，把整个天然气供应网络推向承受极限。 主持人追问，在恶化甚至最坏的情景下，这是否意味着一场全美范围的恶性危机？Matthew 确认了这一点：如果放任这种失衡发展，天然气价格将出现极其暴力的非线性飙升，成为全国的核心经济风暴。这绝不是局部的轻微波动，而是将自上而下冲击电价、制造业与家庭账单的宏观危机。 --- ### `11:38` 地下到底有没有气？原子级模型揭示的资源真相 要搞清楚危机的根源，必须厘清一个根本问题：究竟是北美地下在物理意义上真的没有足够的天然气了，还是人类在勘探、开采、加工、储运的速度上跟不上需求？换句话说，这到底是一个“绝对储量不足”的问题，还是一个“工业系统交付能力”的问题？ Matthew 解释称，过去 16 个多月里，他的团队在“原子级”粒度上对全美几乎所有天然气资产进行了详尽建模。得益于极高的数据透明度，团队通过各企业控制区块的多边形轮廓（Polygon shapes）与经纬度数据，将每一口井与特定区块精准匹配，详细追踪了全美各大产气盆地的开发阶段： - **阿巴拉契亚盆地（Appalachia）**：主要由马塞勒斯页岩（Marcellus）与尤蒂卡页岩（Utica）构成，是全美纯天然气开发的核心主力； - **海恩斯维尔页岩（Haynesville）**：贴近墨西哥湾沿岸 LNG 终端的关键边际调节盆地（Swing basin）； - **二叠纪盆地（Permian）与鹰滩页岩（Eagle Ford）**：以开采石油为主的油田，天然气多作为伴生气（Associated gas）产出。 分析结果显示，从地质层面来看，地下确实存在大量天然气。如果各大勘探开发公司将其现有控制区块内的所有合理井位全部钻探投产，理论上确实能够贡献团队预测模型中每日 200 亿立方英尺（20 Bcf/d）的产量增长。然而，地质储量能被找到，并不等同于它能在需要的时间点以经济可行的方式送达地面并进入管网。 ```mermaid flowchart TD accTitle: 天然气从地表下储量到管网交付的约束路径 accDescr: 展示从地下地质资源到最终符合管道规格交付的各个递进瓶颈环节 subgraph SubGeo["地质储量（Stock）"] A["已知地层潜力: 阿巴拉契亚 / 海恩斯维尔 / 二叠纪"] --> B["企业掌控的高品位核心井位（4-5年消耗殆尽）"] end subgraph SubFlow["开采与流速瓶颈（Flow）"] B --> C{"产能交付上限: 128-132 Bcf/d"} C --> D["单井初始高递减率与老井自然衰减"] end subgraph SubSurface["地面中游瓶颈（Midstream）"] D --> E["地面处理厂: 杂质分离（硫、氮）与液态天然气分离"] E --> F["达到管道规格: 1,010 - 1,030 BTU"] F --> G["主干管网外输至 LNG 码头与 AI 数据中心"] end ``` --- ### `15:00` “流量”而非“存量”：产能交付上限与储量快速消耗 天然气市场最容易被外界误解的一点，在于混淆了“存量”（Stock）与“流量”（Flow）。地下就算有上千兆立方英尺的储量，但如果一口井一天的出气量有限，且衰减极快，那么在特定的时间窗口内，能够提供的有效“流量”就存在物理天花板。 Matthew 强调，按照目前各家天然气开采巨头的钻探节奏，大部分公司现已圈定的优质库存井位将在未来 4 到 5 年内被基本耗尽。页岩气井具有天然特征： - **初始自然递减率极陡**：新井投产第一年产量下降极快； - **老资产组合的递减趋缓**：像 EQT 和 Expand Energy（由切萨皮克能源与西南能源合并而成）这类巨头，由于其资产组合极其成熟，基础产量递减速度变慢，相比五年前维持现有产量的资本重置成本有所下降。 即便假设这些勘探开发公司在当前低迷的远期价格曲线约束下，把名下剩余的全部区块以最优方式钻探完毕，全美天然气的最大日交付能力极限也仅在 **128 到 132 Bcf/d** 之间。Matthew 直言：“我们模型采用的上限是 132 Bcf/d，因为只有达到这个数字，才能同时满足美国已经承诺的全部 LNG 出口订单。但现实中，我们倾向于实际产出会低于这个上限。” 对于业内常听到的“二叠纪和阿巴拉契亚仍有无限资源”的乐观论调，Matthew 表达了强烈质疑。许多上市公司向投资者吹嘘拥有数十年的可钻探井位库存，但在工程事实上根本站不住脚：“当人们与这些公司开会时，应当要求他们在地图上出示精确的工程井位。他们不仅要有开采潜力，还必须有地表配套设施让气体外输，并具备在财务纪律内进行投资的能力。”成熟产区已高度透明，指望未来突然发现一个颠覆供需格局的新型超级页岩盆地完全是不切实际的幻想。 - **存量衰竭速度飞快**：按当前开采强度，已知最高回报、最具经济性的井位正以惊人速度被消耗。 - **冷门盆地难以救场**：美国地质情况已被充分探明，虽然存在其他已知含气构造，但往往开采成本极高，且完全缺乏管网与处理等地面基础设施，根本无法在几年内形成有效运力。 - **历史周期的幻觉**：2010 年至今长达十余年的“天然气严重过剩”时代，在历史上只是一段特殊的插曲；此前美国长期缺气，而未来也很可能重回供不应求的常态。 --- ### `19:10` 地面之上的瓶颈：天然气加工与管网规格限制 假设即便井口能够打出足够的气体，天然气开采的第二道重大关卡立刻横亘在眼前——地面中游（Midstream）处理能力。从地下开采出来的天然气并不是直接就能烧的干燥甲烷，往往伴随着复杂的伴生物质，成为限制产量的致命瓶颈： 首先，气体必须经过分离与精炼。天然气出井时通常混杂着乙烷、丙烷、丁烷等天然气凝析液（NGLs）；在某些区块，气体中还含有必须脱除的剧毒硫化氢或高浓度的氮气；若来自二叠纪等油区，还必须配备原油分选与收集系统。商业外输管网对进入管道的气体有严格的热值与成分标准，通常要求达到 **1,010 至 1,030 BTU** 的标准规格。如果地面没有充足的天然气加工厂（Processing Plants）将多余的重烃和杂质剔除，符合规格的干气就根本无法被注入管道。 目前全美仅有少数几个盆地拥有少量的加工富余能力，而这些闲置产能很快就会被填满。面对未来数年因 LNG 和 AI 电力激增所要求的每日 20 Bcf/d 增量，地面的天然气加工产能建设远远滞后。没有这些重型地面处理装备，即便地下的石头里有再多的气体，也无法转化为终端用户能够点燃的商品。 ### `20:00` 天然气供应链瓶颈：加工厂、集输管道与跨州管网 要想实现支撑未来需求所需的天然气增产目标，基础设施的建设周期构成了一个不可忽视的现实阻碍。天然气加工厂（Processing Plants）的平均建设周期通常需要 2 到 3 年。业内非常清楚目前二叠纪盆地（Permian Basin）或阿巴拉契亚盆地（Appalachian Basin）的加工商公布了哪些投资项目，也基本能推算出 2027 至 2028 年处理原材料和液化物的输送吞吐能力。然而，即使只是要达到 Chronometer 团队模型中保守估计的 20 Bcf/d（十亿立方英尺/日）增量天然气产量，也必须立即投入大量资金建设更多的加工能力。 除了加工设施，另一个关键制约是集输管网（Gathering System）。集输管道是将天然气从井口输送至加工厂或主干管网的小口径管道。尽管市场上有些关于集输系统扩建的披露，但团队将所有这些项目按照经纬度标注在全美每一口油气井上后发现：即便为了实现全美天然气产量达到 130 Bcf/d 的一小部分增幅，整个集输系统也必须立刻、大规模地追加资本开支。 最棘手的瓶颈则是跨州天然气主干管道（Interstate Gas Pipelines）。管道本质上具备垄断属性，终端向居民输送燃气的本地分销公司（LDC）也是垄断或寡头垄断。过去 10 到 12 年间，全美实质上仅建成了一条大型跨州天然气管道——连接阿巴拉契亚与大西洋中部地区的山谷管道（Mountain Valley Pipeline, MVP）。环境审查、繁琐的许可流程以及监管体制的频繁变动，使修建跨州管道变得极其艰难。尽管现任政府一直尝试降低审批壁垒以加强跨区域调度能力、服务激增的 AI 算力电力负荷，但远水解不了近渴。 ```mermaid flowchart TD accTitle: 天然气从井口到终端的供应链瓶颈 accDescr: 展示从油气井口、集输管道、加工厂到跨州主干管网的制约环节 subgraph Wellhead["井口与集输系统"] A["油气井口 (Wellhead)"] --> B["集输小口径管道 (Gathering)"] end subgraph Midstream["加工与长输系统"] B --> C["天然气加工厂 (Processing) - 需耗时2-3年建设"] C --> D["跨州主干管道 (Interstate Pipelines) - 审批周期极长"] end subgraph Demand["下游终端负荷"] D --> E["AI 数据中心与发电厂"] D --> F["LNG 出口码头"] end ``` - **加工厂建设周期**：通常需要 2 到 3 年时间才能投产，必须“立刻、马上”开工才能赶上 2027–2028 年的负荷。 - **全美产量瓶颈**：要达到 130 Bcf/d 的总产量目标，集输端必须全面追加投资。 - **跨州管道建造成果**：过去 10 到 12 年全美仅真正落成了一条跨州管道（Mountain Valley Pipeline）。 --- ### `22:21` 系统的自满惯性与 2028–2030 年的库存枯竭危机 如果整个能源系统维持现有的惯性，没有人重视这一警告并采取实质行动，到 2030 年世界会变成什么样？当前的实际情况是，AI 算力产业链对以天然气为主要发电燃料存在极大的“路径依赖与系统惯性”。例如 SemiAnalysis 在算力供应链研究上处于领先地位，甚至精准做出了表后（BTM, Behind-The-Meter）负荷预测与发电机组的匹配模型，但所有这些研究都止步于“电从哪里来”，完全没有考量“发电用的天然气从哪里来”。 市场目前仅仅聚焦于“缺电”这一表面现象，试图通过光伏加储能、风电、核电（包括大型核反应堆与小型模块化反应堆 SMR）或天然气发电来填补缺口。但由于 2026 和 2027 年当下的天然气供应依然充裕，导致根本没有人追加天然气上游投资。头部生产商如 EQT 甚至因为认为未来的天然气更值钱，而选择在现阶段主动关井限产（shut-in）；实时钻机数量（rig count）也丝毫看不出市场对 2028 至 2030 年供应收紧有任何警惕。过去 15 年来，北美页岩气“源源不断、取之不尽”的观念早已根深蒂固，整个市场弥漫着麻木与自满。 这种自满直接反映在期货远期曲线上：从 2026、2027、2028 直到 2030 年，天然气价格曲线几乎完全走平，因为所有人依然深信资源是无限的。然而，结构性矛盾将在 2027 至 2028 年迅速显现。全美工作气有效储量（working gas storage）总容量约为 4 Tcf（万亿立方英尺），通常随冬夏高峰采气、春秋淡季注气呈现周期性波动。到 2028 年年中，全美储气库的抽采幅度将达到前所未有的水平；2029 年，库存水平将彻底跌破所有历史已知极值；到 2030 年，储气库库存恐将接近见底枯竭。 - **全美工作储气量**：总容量约为 4 Tcf，按季节在春夏注气、秋冬抽气。 - **历史性供需拐点**：2028 年初至年中将首次打破历史季节性规律，开始深幅度消耗库存。 - **远期曲线失真**：直至 2030 年的远期天然气价格曲线依然平坦，完全未定价即将到来的供给悬崖。 --- ### `25:00` 凸性爆发与非线性暴涨：二十美元气价与断供困境 一旦天然气有效库存被完全打穿，由于历史上从未出现过全美范围的永久性气荒，市场只能参考过往的极端供需失衡事件。例如俄乌冲突爆发时，因大量天然气流向欧洲，美国 Henry Hub 气价迅速飙升至 8、9、甚至 10 美元/MCF；又如 2014 年极地涡旋或 2022 年 12 月冬季寒潮，气价也曾短暂触及 6 至 10 美元。但那些全部是“暂时性冲击”，而 2028 年后面对的却是 AI 算力与 LNG 出口驱动的“结构性刚需”，与现有产能极限完全脱节。 届时供需赤字呈现出极端的“凸性（convex）与无上限（unbounded）”特征。当被问及气价是否可能暴涨至 20 美元时，Matthew Smith 表示甚至难以给出最高价格预测。在 8 到 10 美元区间，市场机制可能会迫使一部分通过现货船货（spot cargos）出口欧洲套利的 LNG 停止装船，留在美国国内，模型中虽已计入这部分现货缓冲，但仅靠现货根本无法填补缺口。若要真正平抑国内危机，必须强行截断已签署长期协议的 LNG 出口合同——但这牵涉到巨大的地缘政治与商业违约代价，比如直接切断那些依赖美国长期气源保障供电的日本公用事业机构的生命线。 - **历史极端价格参照**：俄乌冲突期间及极地寒潮期间气价曾冲至 8 到 10 美元/MCF，但均属短期事件。 - **现货 LNG 调节上限**：8 至 10 美元价位或抑制套利性现货出口，但对整体赤字而言杯水车薪。 - **长约 LNG 违约成本**：强行阻断长协出口将动摇海外盟友（如日本电力企业）的能源安全底线。 --- ### `27:00` 赢家盘点：优质上游生产商与公用事业级光伏的暴利 在如此极端的凸性局面下，资本市场的赢家与输家分化将极其剧烈。在上游天然气生产商中，最核心的受益者是 Expand Energy（由 Chesapeake 与 Southwestern 合并而成）。该公司掌控了海恩斯维尔页岩（Haynesville）约 70% 的剩余核心高品位储层，这些是地质参数最明确、单井产出极高的优质岩层。耐人寻味的是，Expand 目前正经历 CEO 更迭，股价在过去半年内因管理层空缺而大幅下跌，按远期曲线估值仅以 4 倍 EV/EBITDA 交易，市场完全没有定价未来气价飙升的可能。阿巴拉契亚地区的代表则是 Range Resources，其具备充足的增产空间与极高的资本回报率。 另一个不符合直觉的巨大赢家是公用事业级光伏（Utility-Scale Solar）。在全美各区域电力市场中，天然气发电厂始终是边际定价机组（marginal fuel），“天然气价格怎么走，全美电价就怎么走”。在整个发电调度曲线（dispatch curve）中，光伏、风电和水电的燃料边际成本几乎为零。过去十年中，除天然气外，光伏加储能占据了全美电网新增并网排队（interconnect queue）的 90%。当电价因高企的边际气价而大幅攀升时，“阳光本身的成本却分文未涨”，光伏运营商将坐享巨大的超额利润。 与此同时，户用分布式光伏（Residential Solar）也将迎来爆发。尽管自 1970 年代末以来长期享有的税收补贴正在退坡，但由于 2028 年后天然气暴涨将推高白天用电高峰（上午 10 点至下午 6 点）的居民电费，户用光伏加储能电池将成为居民对冲高额电费账单的少数手段之一，其经济自洽性首次完全无需依赖政府补贴。 | 受益板块 / 公司 | 代表标的 | 核心逻辑 | |:--- |:--- |:--- | | **海恩斯维尔天然气巨头** | Expand Energy | 掌控约 70% 核心海恩斯维尔优质井位；估值仅 4 倍 EBITDA，市场严重忽视其资产价值。 | | **阿巴拉契亚优质生产商** | Range Resources | 拥有极高质量的岩层储量，具备充沛的增产潜能与股东回报释放空间。 | | **公用事业级绿电运营商** | XPLR (XIFR, 原 NextEra Yield Co), Clearway Energy | 边际气价推高整体电价，而光伏边际成本为零；购电协议（PPA）到期重签带来无资本开支的纯利润扩张。 | | **户用光伏与家庭储能** | 户用光伏产业链企业 | 在白昼电价暴涨预期下，居民为防范电费飙升被动加装，项目在零补贴下具备极强经济性。 | ### `30:08` 破解天然气死局的唯一解：为何大核电优于小型模块化反应堆（SMR） 在探讨如何应对本十年末天然气及电力价格的“凸性暴涨”（convex situation）时，马修·史密斯（Matthew Smith）坦言，市场根本不存在立竿见影的“银弹”。除了风电与光伏，目前天然气是唯一能够支撑电网直到核电并网的灵活性过渡燃料（flex fuel）。团队在深入调研整个核电生态系统后得出一个明确结论：大规模商用核电机组是唯一切实可行的解决方案，而这一方向直接指向了西屋电气（Westinghouse）的 AP1000 技术路线。 面对主持人关于“核电站建设难道不需要耗费五年甚至更久”的疑问，史密斯直言实际耗时远不止于此。回顾过去30年，美国仅建成了两座大型核电机组——佐治亚州的沃格特尔（Vogtle）3号和4号机组；而同期南卡罗来纳州的 VC Summer 核电项目不仅彻底烂尾下马，甚至直接将公用事业公司 SCANA 推向破产边缘，最终被迫易主被收购。从福岛、切尔诺贝利到三里岛，这三十年间美国核电工程师、科学家及企业大规模流失，全行业几乎丧失了建设大型核电的“肌肉记忆”。Vogtle 3、4号机组历经约15年、耗资达到最初预算三倍的惨痛教训至今仍让市场心有余悸。 然而，如果将目光投向2030年代，天然气供应赤字在2031年和2032年只会愈演愈烈。在史密斯看来，唯一的出路是以最快速度重启大型核电建设，使其能够在2033或2034年按期并网投产。受规管的公用事业公司、超大规模云厂商（Hyperscalers）以及监管机构本应紧密协同达成共识。但现实是，大多数人不相信未来几年天然气会陷入严重短缺，他们“看不清问题，自然就会抗拒解决方案”。美国历来习惯于修建管道去解决“眼前的问题”，而不是提前五年或十年未雨绸缪；史密斯和团队正在做的，正如冰球名言所说，是提前滑向“球即将到达的位置”。 ```mermaid flowchart TD accTitle: 大型核电与天然气困境推导图 accDescr: 展示2030年代天然气缺口扩大如何倒逼大型核电成为唯一底荷解法 subgraph Problem["天然气与电网危机"] A["2030年后气荒加剧"] --> B["电价呈非线性/凸性暴涨"] B --> C["风光灵活性不足 / 缺乏跨期过渡燃料"] end subgraph NuclearPath["核电路径抉择"] C --> D{"替代方案评估"} D -- "技术与供应链未成熟" --> E["SMR 小型反应堆<br/>(数十吉瓦缺口远水难解)"] D -- "唯一可行工业化路径" --> F["大型核电机组 AP1000<br/>(目标 2033-2034 投运)"] end subgraph Beneficiaries["关键供应链受益者"] F --> G["Cameco & Brookfield<br/>(西屋电气控股方)"] F --> H["BWXT<br/>(核心堆芯设备与高美元价值量)"] end ``` --- ### `32:40` SMR 的“科学实验”困境与西屋电气 AP1000 的工业复兴 针对市场热炒的使用小型模块化反应堆（SMR）作为数据中心“表后侧”（Behind-the-Meter, BTM）供电方案的构想，史密斯给出了相对冷静的评估。他指出，目前的很多 SMR 本质上仍属于“科学实验阶段”。尽管美国核能管理委员会（NRC）与联邦政府正在积极拆除监管壁垒、加速验证技术可行性与经济性，但绝大多数 SMR 初创企业根本不具备在2030年代实现“数十吉瓦（tens of gigawatts）”规模化工业制造的能力。 相比之下，大型核电站的确定性要高得多。Vogtle 4号机组在建造效率上已较3号机组展现出显著改进。放眼全球，中国当前在建的约39座核反应堆中，有34座装机容量在1吉瓦（1,000兆瓦）以上，其中约三分之一正是基于西屋 AP1000 的技术路线开发。行业如今对如何安全、高效建造大型反应堆的认知，已远非昔日频频踩坑时可比。在清晰的投产时间表、更可控的度电成本以及现成的工业配套面前，成熟的大堆路线完胜商业化落地存疑的 SMR。 - **西屋电气（Westinghouse）的隐性价值**：布鲁克菲尔德（Brookfield）与卡梅科（Cameco）分别持有西屋电气 51% 和 49% 的股权。史密斯认为，美国政府目前正在积极促成采购承诺以保障供应链，西屋电气未来若重新拆分上市，其在 Cameco 资产包内被严重低估的价值有望得到全面重估。 - **核心军工/核电供应商 BWXT（BWX Technologies）**：作为美国海军核动力反应堆的核心供应商，BWXT 将在下一轮商用大型核电周期中深度受益，在每台 AP1000 机组中均拥有极高的“单堆美元含量（dollar content）”。 --- ### `34:29` 最大的输家：美国消费者与“自带发电机（BYOG）”的悖论 当被问及这场能源挤压中谁将蒙受最大损失时，史密斯神色凝重地指出：**最大的受害者将是美国普通消费者**。如果这套预测模型哪怕只被部分验证，远期电力曲线呈现的电价飙升都会迅速演变为一场残酷的社会政策困境：面对有限的天然气供应，政策制定者是在天然气出口赚取外汇、保障 AI 算力电力供应，还是维持居民可负担的民生电费账单之间进行权衡？ 这种不可调和的矛盾必然会引发更激烈的社会公共讨论与“邻避效应”（NIMBYism）。史密斯强调，他们绝非反技术或反 AI，“我们深知 AI 具有巨大的变革性，但它确确实实是个耗电巨兽”。 更加反讽的是当前各州政府和电网针对数据中心的解决方案。由于监管层普遍盲目相信天然气储备无穷无尽，他们给超大规模科技巨头开出的药方是“自带发电机”（Bring Your Own Generation, BYOG）——要求企业在选址时自行配置离网发电设施。然而，这种所谓的独立供电绝大多数依然是就地燃烧天然气。 > “现在每当你看到 Bloom Energy 或类似公司发布重大供电合作新闻稿时，别想别的，请在脑海里翻译成四个字：‘消耗更多天然气’（Think more gas）。” 根据各发电设备的能效换算，这些增量需求正在将原本就极度紧绷的天然气供需天平推向崩溃边缘。 --- ### `36:37` 燃气轮机制造商的周期陷阱：在顶峰盲目扩产 在资本市场上，燃气轮机与分布式柴油/天然气发电机组（gensets）制造企业在过去两年迎来了惊人的牛市行情与丰厚利润，但史密斯警告，这一行业极具“繁荣与萧条”（boom and bust）的周期宿命。本世纪初（2000年代初），全美曾掀起一股疯狂兴建天然气电厂的热潮，导致装机容量严重过剩，随后整个设备制造业陷入了长达近二十年的低迷煎熬。 戏剧性的一幕正在重演：面对眼前的电力短缺，这些制造商正计划在 2028 至 2029 年再度大规模扩张产能，而这极有可能踩在周期的悬崖边缘： - **卡特彼勒（Caterpillar）**：其旗下的 Solar Turbines 计划在2029年底前将产能翻倍，史密斯直言这几乎是“在最糟糕的时间点逆势扩产”，届时电厂运营商很可能会因燃料价格过高或气源不足而被迫推迟乃至放弃部署这些机组。 - **Bloom Energy**：近期市场聚焦于其燃料电池制造过程中涉及的稀土供应链瓶颈，但史密斯指出其面临更致命的硬约束——当其规划部署规模达到 2 吉瓦甚至更高时，在全网天然气高度紧缺的环境下，这些分布式燃料电池根本无法抢到足够的天然气分子。 史密斯预测，除非天然气上游勘探与管道输送能力出现实质性暴增，否则到了2029或2030年，继续新建大型天然气发电机组在经济上将完全失去合理性；最早在2026年，整个燃气发电设备行业的增量订单就可能出现剧烈断崖式下滑。 --- ### `37:57` 超大规模科技巨头（Hyperscalers）的算力成本盲区与 DRAM 式危机 这种全行业视而不见的供应瓶颈，与半导体行业曾经经历的短缺惊人相似。史密斯提到，他曾与往期播客嘉宾深入交流该模型，对方当场惊呼：“这听起来完全就是两年前 DRAM 内存芯片危机的翻版——**起初极其缓慢，随后骤然爆发（slowly at first and then all at once）**。”上游产能资本开支长期不足的恶果，终将以极度剧烈的方式反噬终端买家。 目前，超大规模科技巨头的成本预算体系中存在严重的盲区： - **能源预算被严重低估**：当前数据中心总体拥有成本（TCO）中，折旧费用占据绝对大头，存储与服务器硬件次之，而电力能源成本通常仅被粗略预算为总成本的 10% 左右。 - **能源成本倍增风险**：如果未来 AI 算力高度依赖天然气发电，一旦天然气结构性翻倍甚至翻两番（甚至无需极端恶劣天气触发），到2029年，电力支出占算力总成本的比重可能会骤升至 20% 到 30%，彻底颠覆云巨头的单机柜经济模型。 - **平准化度电成本（LCOE）的定价假象**：当前科技巨头在核算项目回报率时，普遍直接套用平缓贴水、远期仅在 3.50 美元/MMBtu 左右横盘的天然气远期曲线（forward curve）。这给决策层制造了一种“未来燃料无限且廉价”的虚幻安全感，使他们将全部精力扑在采购芯片、交付 Anthropic 等大模型算力上，却对底层支撑体系即将坍塌的物理现实毫无准备。 | 评估维度 | 超大规模云厂商（Hyperscalers）当前假设 | Chronometer 团队量化推演（2029-2030） | |:--- |:--- |:--- | | **远期天然气价格预设** | 锚定在 ~$3.50/MMBtu，远期曲线平坦稳定 | 结构性供应枯竭，价格上行空间无上限且呈凸性 | | **能源占算力总成本比例** | 约 10%（主要关注服务器硬件与折旧） | 预计飙升至 20% ~ 30%，成为核心成本负担 | | **分布式供电（BYOG）策略** | 部署 Bloom/燃气轮机等天然气分布式电源 | 面临“有发电机组却无气可烧”的断供窘境 | | **设备投资周期判断** | 持续追加燃气发电机组及中短期基础设施 | 2026年起增量订单恐因燃料断供而急剧停滞 | ### `40:00` 二叠纪盆地的供给局限与电池技术变量 针对天然气短缺的预测，市场上最常见的质疑就是：“二叠纪盆地（Permian Basin）拥有巨量的伴生气（Associated Gas），油价又高，为什么二叠纪的产能无法彻底解决问题？”针对这一疑问，团队早已通过深入中游管网的物理与监管瓶颈做了严密测算。当前基准模型已经计入了 2026 年至 2030 年间确定投产或在建的每天 70 亿立方英尺以上（7+ Bcf/d）的新管道运力。在现行监管审批流程和工程周期下，如果 2030 年前还有任何未宣布的管道能上线，业界不可能不知情。因此，问题不在于地底下的天然气储量，而在于有多少气体能够实际“运出盆地”，无论是送往沿海的 LNG 出口终端，还是留在得州本土及周边消纳。 更深层的经济逻辑在于伴生气的开采机制：要产出更多的二叠纪天然气，就必须有动力去开采更多的二叠纪原油。但在伊朗地缘冲突爆发前，全球原油供给实际上相当充裕，价格并不支持无节制的超额增产。如果要让二叠纪的天然气产量超出目前已规划的这 7+ Bcf/d 管网外输极限，就需要油价在极长时间内维持在极高水平，而这本身就会严重加剧普通消费者的能源成本危机。至于“在二叠纪盆地现场建立表后（Behind-the-meter）发电机组为数据中心供电”的提议，模型同样完全吸纳了这部分已宣布项目的需求——但在本地被就地烧掉的天然气，恰恰意味着无法再注入下游干线输送给其他市场，根本无法缓解整体的系统性短缺。 在技术变革方面，电池储能是唯一可能在结构上重塑天然气调峰需求的变量。目前纳离子等新型储能技术虽然备受关注，但尚未实现商业化部署；现阶段具备经济性的电池项目绝大多数仍是短时放电（Fast discharge cycle）的锂电池系统，这些已知装机已被完整纳入电力系统模型中。如果电池技术出现断代式的跨越（Step-function leap），不仅会彻底颠覆现有的行业赢家与输家格局，更将成为化解这场天然气危机的一大历史转折点。 ```mermaid flowchart TD accTitle: Permian Gas Constraints accDescr: 解释二叠纪盆地天然气受制于原油开采动机、管道外输上限以及本地表后消费对下游供应的挤压。 subgraph Permian["二叠纪盆地约束条件"] OilLink["伴生气依赖原油开采<br/>需油价长期居高才有增产动力"] PipelineLimit["外输管网物理极限<br/>2026-2030年仅新增 7+ Bcf/d"] LocalPower["表后发电机组本地消耗<br/>挤占下游外输管网气量"] end subgraph MacroImpact["宏观外溢结果"] ConsumerCost["消费者能源账单与油价双双承压"] DownstreamShortage["沿海出口与区域电网依然缺气"] end OilLink --> ConsumerCost PipelineLimit --> DownstreamShortage LocalPower --> DownstreamShortage ``` --- ### `44:25` 能源沙皇的药方：核电破局与户用光伏自救 当被问及如果拥有一天“能源沙皇”的绝对决策权会如何化解危机时，给出的首要长期解药是由联邦政府彻底托底核电产业。美国政府应当直接出资，从头到尾完整建设 2 到 4 座 AP1000 大型核反应堆。当前整个产业陷入了囚徒困境：虽然美国能源部（DOE）各机构规划了 10 到 20 座反应堆的愿景，且美国能源部贷款项目办公室（LPO）坐拥巨额资金额度（如 EDF 等项目在 2028 年底前面临大量资金窗口），但商业实体谁也不愿意充当第一个承担供应链超支风险的先驱。若联邦政府带头注资建完前几座示范堆，彻底扫清供应链壁垒，就能带动后续跟进，甚至有望像中国那样批量建造 30 座核反应堆。能源和资源的可获得性、外输通道（Egress）以及工程知识积累，正是限制美国技术生产力全面铺开的核心瓶颈。 在分布式自救层面，推动全美居民屋顶光伏普及被视作当下最紧迫的行动。虽然分布式太阳能受制于阴晴天气，无法做到全天候稳定，但它能精准对冲每天上午 10 点到下午 6 点之间的电网尖峰电价，而这几个小时恰恰构成了居民电费账单中最昂贵的部分。过去一年中，随着部分补贴政策的取消，美国住宅太阳能行业出现停滞。各州政府亟须重新审视政策法规，出台全新激励措施来重振户用光伏，让居民能在未来几年的电价暴涨潮中建立自主防御手段。 --- ### `46:52` 加拿大输气潜力与全球 LNG 供应的地缘多米诺 在北美跨境能源流动方面，加拿大作为天然气大国，通常仅在 1 月份等深冬极端时期具备向美国每日输送 110 到 120 亿立方英尺（11-12 Bcf/d）天然气的能力。但由于加拿大本身的储气库容量极其有限（仅约 1 万亿立方英尺 / 1 Tcf，大约只有美国工作储气容量的四分之一），整体管网主要按冬季调峰设计，导致美国无法全年大规模依赖加拿大进口。然而从 3 到 5 年以上的周期来看，加拿大地下拥有全北美最具经济性、储量最深厚的优质天然气资产，只是长期受困于管道运力不足。理想的跨区域方案是在中西部建设一条 1 到 2 Bcf/d 的输气主管道直通 MISO 电力市场，并以此为枢纽在 MISO、PJM、SPP 和 ERCOT 各大电网间灵活调度过网，以此满足日益激增的电力负荷，防止电费飙升拖累经济需求。 这一内生性短缺对外溢至全球地缘格局的影响尤为致命。过去几年，美国天然气出口实现了从近乎为零到每天十多 Bcf、再到如今 15 Bcf/d 的爆发式增长，未来甚至规划迈向 35 Bcf/d。在俄乌冲突切断了俄罗斯对欧供气后，欧洲乃至全球大量新建的燃气发电机组均完全寄希望于美国的 LNG 供应： - **欧洲的再平衡困境**：若美国无法如期输出 30 至 35 Bcf/d 的天然气，欧洲将被迫在高昂的再平衡成本中艰难抉择——要么承担难以承受的超高价美国现货，要么被迫回过头去妥协采用俄罗斯管道气，因为欧洲本土根本没有增产补缺的能力。 - **亚洲盟友的能源安全**：在俄乌冲突爆发前，亚洲曾是美国 LNG 的最大买家，随着中东局势及霍尔木兹海峡外围的不确定性，亚洲未来极可能重新成为最大需求方。美国若因国内短缺而被迫限制出口，将在最需要外交协同的时刻重创关键盟国。 - **政策干预的隐忧**：尽管大幅削减 LNG 出口会严重打击国际盟友信誉，但在本年代末国内电价失控的政治压力下，限制出口几乎注定会成为白宫不得不摆上台面的调控杠杆之一。 --- ### `49:26` 算力繁荣背后的资产错配与锁定气源的集体盲区 回顾长达一年半的全产业链建模分析，最令人震惊的现象莫过于企业高管的公开表态与物理系统真实承载力之间的巨大脱节。在来自资本市场的压力与获取“AI 算力租金（AI Compute Rent）”的狂热驱动下，数以百亿美元计的资金正在疯狂涌入分布式天然气发电机组（Distributed Gas Gensets）。这类机组热耗率高、发电效率低下、运维成本极重，在正常的电力工业环境中本应只充当应急备用电源，但现在却因为科技巨头急于缩短“通电等待时间（Time to power）”，被当作主力基荷电源推上前线。更荒谬的是，在四年或五年之后，这些效率低下的设备很可能因为极其昂贵的运行成本而被迫停机闲置。 更让分析团队感到不可思议的是，在这一波波耗资巨大的分布式能源狂欢中，包括一些极具声望的顶级投资机构在内，几乎没有人在决策前停下来追问一句最根本的问题：**到时候究竟有没有足够的天然气来喂饱这些发电机？气价又会涨到什么地步？** 市场上几乎看不到任何企业通过长协去锁定实物天然气气源以确保供应确定性，金融衍生品市场的对冲交易也极其匮乏。由于 2028 年天然气远期合约的流动性严重不足，现阶段的远期价格曲线尚未反映出未来的结构性短缺；但一旦公用事业公司如梦初醒，集体在 2028 年及以后的合约上开始大规模建仓对冲抢购天然气，整个能源市场的价格风暴就将瞬间爆发。 ### `50:00` 2028年的“白刃战”与远期市场的严重错配 天然气发电企业即将迎来一场前所未有的资源争夺战。Matthew Smith 直言，随着2028至2029年物理交割市场的实质性收紧，行业内将爆发一场“为了争夺2028年天然气现货供给的白刃战（a knife fight to secure natural gas physical）”。然而令人极其意外的是，迄今为止市场上几乎还没有人真正着手锁定现货。资本市场目前呈现出一种极为扭曲的割裂：海量资金正疯狂涌向那些未经检验、尚未成熟、甚至要到2035年才可能落地的概念性技术；与此同时，像 Expand Energy 这样的头部天然气生产商估值仅有4倍 EBITDA，自由现金流收益率（FCF yield）高达10%到15%中位区间。 这种定价反差源于当前金融市场的极度短视与自满。投资人紧盯眼前的夏季气温波动或个别 LNG 出口终端的短暂检修停运，完全忽视了未来几年确定会并网投产的庞大需求增量。在当前的天然气远期曲线上，这种自满情绪被体现得淋漓尽致。然而，随着时间推移，当企业在未来六个月内开始将采购窗口滚动对准2028年时，远期曲线必将迎来剧烈的向上重估。市场一方面不愿穿透短期供应充足的表象去看待天然气上游资产，另一方面却愿意为2035年完全没有事实依据的远景付出极高估值，这种跨行业假设的巨大自相矛盾令人震惊。 - **估值反差强烈**：大型天然气生产商当前仅以约4倍 EBITDA 交易，提供低至中双位数的自由现金流收益率。 - **时间窗口收窄**：市场主流视线仍停留在即期天气与短期 LNG 检修，但未来6个月内，市场将被迫开始定价2028年的实质性供需缺口。 - **现货争夺加剧**：区域性物理供给在2028至2029年急剧紧绷，现货保供将取代金融对冲成为核心矛盾。 ### `51:40` 给企业高管的警示：锁定现货与重估交易对手风险 针对那些直接或间接将能源作为关键生产要素的企业 CEO，Matthew Smith 给出了严厉的风险提示：企业在规划和投建资产时，必须彻底搞清楚未来的天然气究竟从何而来，并且必须在当下就锁死物理供应。更关键的是，管理层必须重新审视“交易对手风险（counterparty risk）”。在过去两年的 AI 繁荣期中，行业几乎从未严肃讨论过交易对手违约的可能。但在剧烈波动的商品市场中，多空力量的失衡往往会引发致命冲击。Matthew 打了个生动的比方：“想象一下，如果在12或18个月前你发现自己突然做空了半导体存储器（Memory），那会是什么下场？现在的天然气市场对我们而言就是这种感觉——而且这不需要等两年，未来六个多月就会显现。” 这种物理保供审查对于计划部署单循环燃气轮机（Simple Cycle）、联合循环燃气轮机（CCGT）的大型电厂买家至关重要，而对于燃料电池（Fuel Cells）的追捧者而言更是当头棒喝。Matthew 对大规模部署燃料电池持极其怀疑的态度，因为管道管网中根本没有足够的边际天然气能够支撑这些设备实现“全天候365天不间断”的基荷运行。在 Chronometer Partners 的基准预测模型中，燃料电池仅仅被按应急备用电源（backup gen）处理；如果科技公司执意将其作为全天候基荷电源进行大规模部署，只会把前述天然气供需错配的时间节点进一步提前，并在价格凸性上火上浇油。 ```mermaid flowchart TD accTitle: AI能源供应链与天然气约束 accDescr: 展示超大规模数据中心、燃气电厂与燃料电池在面临天然气物理供应受限时的传导瓶颈与风险。 subgraph Demand ["激增的电力需求"] A["Hyperscalers 算力中心"] --> B["自建燃机 / CCGT 电厂"] A --> C["燃料电池 (试图作为基荷)"] end subgraph Bottleneck ["物理供给瓶颈 (2028-2030)"] D["地下管网与存储能力耗尽"] E["交易对手违约与交付风险"] end subgraph Impact ["宏观反噬"] F["天然气价格脉冲 ($10+)"] G["居民电价激增与监管介入"] H["算力利润率与瓦特性能受挫"] end B --> D C -. "缺乏全天候供气".-> D D --> E E --> F F --> G F --> H ``` ### `53:15` 产业链各方的生存法则：从工程建设商到超大规模云厂商 产业链上不同生态位的企业面临着各异的生死挑战。以当前以25倍现金流高位交易的工程建设与承包（E&C）企业为例，这一估值倍数在建筑工程行业历史上处于极端高位，其核心逻辑建立在源源不断的天然气电厂建设订单上。然而到了2029至2030年，当高昂的气价和严厉的监管审查阻碍新电厂审批时，这类企业将面临订单断崖。这些工程企业的当务之急，是通过具有协同效应的兼并收购（M&A）实现业务多元化，彻底摆脱单一依赖天然气发电设备建设的被动局面。 对于超大规模云厂商（Hyperscalers）而言，算力硬件层面的存储短缺曾是众所周知的痛点，但接下来的致命打击可能来自能源成本。当云巨头本该迎来商业模式利润率飞跃的“逃逸速度（escape velocity）”时，天然气及衍生电力成本可能会猛增至其总运营成本的20%、30%甚至40%。这也意味着，“每瓦算力性能（Performance per watt）”将成为未来评估计算架构的核心指标。企业高管与项目合伙人必须对业务进行极端情景压力测试：如果天然气价格不再是每百万英热3.50美元，而是跳升到10美元甚至更高，商业模式能否承受？如果物理管道断供怎么办？如果公众因电价上涨而抗议、倒逼监管机构出手干预时，企业又该如何应对？对这些潜在致命变量的预判与对冲，将决定谁能在即将到来的能源瓶颈期生存下来。 - **工程建设企业（E&C）**：25倍现金流估值处于历史泡沫边缘，必须通过 M&A 拓展非气电业务以防2029年后的行业停摆。 - **超大规模云厂商（Hyperscalers）**：能源成本或占业务开销的20%至40%，直接侵蚀 AI 商业化利润空间；“每瓦算力性能”成为核心技术考量。 - **压力测试假设重构**：企业财务模型必须废除“3.50美元平稳气价”的假设，转向推演10美元以上高气价与物理断供环境下的抗风险能力。 ## Notable quotes > `04:56` The die was cast long before AI compute came to the scene. — Matthew Smith > `10:00` 从签署的合同开始，看纸面上写了什么、允许什么、禁止什么。 — Matthew Smith > `15:00` 我们模型采用的上限是 132 Bcf/d，因为只有达到这个数字，才能同时满足美国已经承诺的全部 LNG 出口订单。但现实中，我们倾向于实际产出会低于这个上限。 — Matthew Smith > `27:00` 天然气价格怎么走，全美电价就怎么走。 — Matthew Smith > `27:00` 阳光本身的成本却分文未涨。 — Matthew Smith > `34:29` 现在每当你看到 Bloom Energy 或类似公司发布重大供电合作新闻稿时，别想别的，请在脑海里翻译成四个字：‘消耗更多天然气’（Think more gas）。 — Matthew Smith > `37:57` 这听起来完全就是两年前 DRAM 内存芯片危机的翻版——起初极其缓慢，随后骤然爆发（slowly at first and then all at once）。 — Matthew Smith > `50:00` a knife fight to secure natural gas physical — Matthew Smith > `51:40` 想象一下，如果在12或18个月前你发现自己突然做空了半导体存储器（Memory），那会是什么下场？现在的天然气市场对我们而言就是这种感觉——而且这不需要等两年，未来六个多月就会显现。 — Matthew Smith ## Mentioned Matthew Smith, Patrick O'Shaughnessy, Chronometer Partners, Cheniere Energy, GE Vernova, Bloom Energy, Wärtsilä, Caterpillar, Solar Turbines, SemiAnalysis, EQT, Expand Energy, Chesapeake Energy, Southwestern Energy, Range Resources, Clearway Energy, XPLR, NextEra Yield Co, Westinghouse Electric, Cameco, Brookfield, BWXT, SCANA, Anthropic, US Department of Energy (DOE), DOE Loan Programs Office (LPO), Nuclear Regulatory Commission (NRC), Mountain Valley Pipeline (MVP), AP1000, MISO, PJM, SPP, ERCOT ## Video description My guest today is Matthew Smith. Matthew is the founder and CIO of Chronometer Partners, which invests in energy, industrials, materials, power and utilities, and related infrastructure. For the last 18 months he and his team have modeled nearly every natural gas well, pipeline, and processing asset in the United States. He's reached a conclusion most of the market doesn't share.&nbsp; Starting in 2028, AI data centers and LNG exports will need more gas than the country can produce and deliver. By his math, the US could exhaust its working natural gas storage by 2030. In his words, the upside risk to prices becomes unbounded and convex. We talk about why this was set in motion long before AI arrived, why the US can't just turn off exports, who wins and loses among producers, nuclear, solar, and the hyperscalers, and what he sees as the only long-term solution. Please enjoy my conversation with Matthew Smith. For the full show notes, transcript, and links to mentioned content, check out the episode page ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠here⁠⁠⁠⁠⁠.&nbsp; ----- In June, Matthew wrote a letter to a small group of confidants laying out the full case behind his natural gas forecast. He has allowed us to publish it. You can read the full letter here. ----- Become a Colossus member to get our quarterly print magazine and private audio experience, including exclusive profiles and early access to select episodes. Subscribe at ⁠colossus.com/subscribe⁠. ----- ⁠Ramp’s⁠ mission is to help companies manage their spend in a way that reduces expenses and frees up time for teams to work on more valuable projects. Go to⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠ramp.com/invest⁠⁠ to sign up for free and get a $250 welcome bonus. ----- Trusted by thousands of businesses, ⁠Vanta⁠ continuously monitors your security posture and streamlines audits so you can win enterprise deals and build customer trust without the traditional overhead. Invest Like the Best listeners get a special offer of $1,000 off Vanta when you go to ⁠vanta.com/invest⁠.&nbsp; ----- WorkOS⁠ is the infrastructure B2B and AI-native companies use to sell to enterprise. It covers everything enterprise security requires: SSO, SCIM, RBAC, Audit Logs, AI governance, and more. Trusted by 2,000+ fast-growing companies, including OpenAI, Anthropic, Cursor, and Vercel. ----- Rogo is the AI platform for finance. They're building agents for Wall Street that are trained to understand how bankers and investors actually do work: from diligence and modeling, to turning analysis into deliverables. To learn more, visit rogo.ai/invest. ----- ⁠Ridgeline⁠ has built a complete, real-time, modern operating system for investment managers. It handles trading, portfolio management, compliance, customer reporting, and much more through an all-in-one real-time cloud platform. Visit⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ridgeline.ai⁠. ----- Editing and post-production work for this episode was provided by The Podcast Consultant. Timestamps: (00:00:00) Welcome to Invest Like the Best (00:02:02) Episode Intro: Matt Smith (00:03:33) The Conclusion After 18 Months (00:04:56) The Die Was Cast Before AI (00:07:24) Sizing AI's Gas Demand (00:09:33) Why Not Just Stop Exporting? (00:11:38) Is the Gas Even There? (00:13:53) The Timing Problem, Not Supply (00:15:15) Flow Versus Stock (00:19:10) What Slows Gas to Market (00:22:21) If Nothing Changes by 2030 (00:26:11) Could Prices Hit Twenty Dollars? (00:27:00) Gas Producers Poised to Win (00:28:54) Utility-Scale Solar's Windfall (00:30:08) What About Nuclear? (00:32:40) SMRs (00:34:29) The US Consumer Pays (00:36:37) Turbine Makers Building Too Late (00:37:57) Are Hyperscalers Exposed Too? (00:44:25) Kickstarting the Nuclear Build (00:46:20) Put Solar on Every Roof (00:46:52) Implications for the World (00:49:26) No One's Securing Supply (00:52:57) The Challenge for Energy CEOs [Source](https://traffic.megaphone.fm/CLS1352215254.mp3)

[4:56] Matthew Smith: The die was cast long before AI compute came to the scene. [10:00] Matthew Smith: 从签署的合同开始，看纸面上写了什么、允许什么、禁止什么。 [15:00] Matthew Smith: 我们模型采用的上限是 132 Bcf/d，因为只有达到这个数字，才能同时满足美国已经承诺的全部 LNG 出口订单。但现实中，我们倾向于实际产出会低于这个上限。 [27:00] Matthew Smith: 天然气价格怎么走，全美电价就怎么走。 [27:00] Matthew Smith: 阳光本身的成本却分文未涨。 [34:29] Matthew Smith: 现在每当你看到 Bloom Energy 或类似公司发布重大供电合作新闻稿时，别想别的，请在脑海里翻译成四个字：‘消耗更多天然气’（Think more gas）。 [37:57] Matthew Smith: 这听起来完全就是两年前 DRAM 内存芯片危机的翻版——起初极其缓慢，随后骤然爆发（slowly at first and then all at once）。 [50:00] Matthew Smith: a knife fight to secure natural gas physical [51:40] Matthew Smith: 想象一下，如果在12或18个月前你发现自己突然做空了半导体存储器（Memory），那会是什么下场？现在的天然气市场对我们而言就是这种感觉——而且这不需要等两年，未来六个多月就会显现。

[3:20] Host: By his math, the US could exhaust its working natural gas storage by 2030. In his words, the upside risk to prices becomes unbounded and convex. [3:20] Host: Starting in 2028, AI data centers and LNG exports will need more gas than the country can produce and deliver. [0:00] Matthew Smith: By 2028, as data centers connect to the grid with dedicated natural gas assets and planned LNG export facilities ramp up, total demand will outpace the nation's ability to produce and deliver natural gas. [4:56] Matthew Smith: Baseline LNG export growth alone consumes virtually 100% of the country's incremental gas supply capability, leaving zero margin for new domestic power demand. [22:21] Matthew Smith: When structural deficits hit, price movements become unbounded and convex; brief past spikes to $8 or $10 per MCF during events like the Russia-Ukraine war or Polar Vortex winter storms were transitory, but the coming shortage will be permanent without immediate intervention. [37:57] Matthew Smith: Energy currently represents roughly 10% of a hyperscaler's operating cost structure... However, if gas prices double or triple structurally, energy could surge to 20% or 30% of the total cost of compute by 2029. [35:00] Matthew Smith: Every press release announcing behind-the-meter or distributed generation—such as from Bloom Energy—means more gas consumption beyond the base case. [46:52] Matthew Smith: As utilities and energy companies begin rolling forward to secure physical supply for 2028 and 2029, a dramatic 'knife fight' for physical gas will break out. [46:52] Matthew Smith: Public equities in unproven 2035 technologies or engineering and construction (E&C) firms trade at high multiples, while natural gas producers trade at depressed valuations based on a complacent forward curve.

[0:00] Ramp is the only platform built to make your finance team leaner, faster, and better, saving businesses 5% annually on average so you can stay focused on growth. Ramp customers grew revenue 3.2 times faster than the average American business. Visa, Vercel, Cursor, Stripe, Notion, 11Lab, Shopify, and 70,000 other businesses all run on Ramp. Mine does too, and so should yours. Learn more at ramp.com slash invest. Felix by Rogo is a personal finance agent that turns a single prompt into finished client-ready work using your firm's own templates, context, and standards. Send Felix an email like, take these comments and turn them for me, or update my tracker with the context of these emails, or run the ability to pay math on this buyer, and Felix sends back finished PowerPoint decks, Excel models, and sourced research. Felix works the way your team already does, delivering work quickly and accurately around the clock. Learn more at rogo.ai slash Felix. The best AI and software companies from OpenAI to Cursor to Perplexity use WorkOS to become enterprise ready overnight, not in months. Visit WorkOS.com to skip the unglamorous infrastructure work and focus on your product. Hello and welcome everyone I Patrick O and this is Invest Like the Best This show is an open exploration of markets ideas stories and strategies that will help you better invest both your time and your money If you enjoy these conversations and want to go deeper, check out Colossus, our quarterly publication with in-depth profiles of the people shaping business and investing. You can find Colossus along with all of our podcasts at colossus.com. Patrick O'Shaughnessy is the CEO of Positive Sum, all opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of Positive Sum. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum may maintain positions in the securities discussed in this podcast. To learn more, visit PSUM.VC. My guest today is Matthew Smith. He's been on the show before many years ago, and I always love talking to him about energy markets where he's worked for 20 years. He's the founder and CIO of Chronometer Partners, which invests in energy, industrials, materials, power and utilities and related infrastructure. He and his team have modeled nearly every natural gas well pipeline and processing asset in the United States He reached a conclusion that most of the market does not share Starting in 2028 AI data centers and LNG exports will need more gas than the country can produce and deliver By his math the U could exhaust its working natural gas storage by 2030 and could lead to a true energy crisis In his words, the upside risk to prices becomes unbounded and convex. We talk about why this was set in motion long before AI arrived, why the U.S. can't just turn off exports, who wins and loses among producers, nuclear, solar, and hyperscalers, and what he sees as the only long-term solution. Please enjoy my great conversation with Matthew Smith. So Matt, the last time we did this was, I think, during COVID. Kind of crazy that it's been six years. I've always loved talking to you about energy markets. You've been working in this space for 20 years. You're about as encyclopedic on this stuff as anyone I've ever met. But you've also been acutely studying the current energy situation in the U.S. rebuilding in a way that you'll describe from the well level up a picture of what's happening, especially as AI is creating all this new demand through data centers, et cetera, of what is going on over the last 18 months of concerted effort. You've reached a fascinating and somewhat scary conclusion. I'd love you to just start with a conclusion, and then we're going to talk through how you came to this conclusion who the winners might be the losers might be what to be done about it But before we get deep into all the component parts just tell us what you found after 18 months of study We are headed into a place where we see an historic deficit in natural gas supply available in the United States, which does pretend some pretty serious consequences. Natural gas, which is over 40% of U.S. power generation, is imminently going to become the most important fuel in the country. It's overtaking petroleum, given the amount that we use now for generation. Our work suggests that 2627 natural gas is appropriately supplied. But as we get into 28 and you plug in this compute and you assign gas to very specific assets as they're plugged in as well, and you continue to export LNG as we're planning to do with known projects, we start to eat into our working gas storage, which is the nexus of supply and demand in the country, I think we will come to the conclusion that the upside risk of the price of natural gas is both unbounded and convex. And so where you will feel that the most securely will be electricity prices in 28, 29, 2030, based on our work. How much of this is just attributable to data centers? [5:00] just purely we're building a lot more data centers, that's just for AI. Is it that simple or is there something else going on as well? The dye was cast long before AI compute came to the scene. If I may set the stage a little bit, U.S. gas was plentiful. Starting in about 2010, when shale started to really change, come to the scene and change things. We had been importing natural gas to satisfy consumption on top of what we produced domestically. shale started to be very productive, surprised the upside, and became this abundant force. And as natural gas became more abundant, we started to export, starting with Chenier. We've gone from that early Chenier exporting to today we're exporting about 15 BCF a day of nameplate US export capacity. Now that 15 billion cubic feet is on a base of about 110 to 112 BCF a day of natural gas production in the US. So if you think about it, it's become about 12 to 15% of the US daily ability to supply the market we're exporting. As this abundance continued, more and more facilities, projects have been announced. As of today we scheduled to export up to 35 BCF a day by the end of 2030 And in that case the dye has been mostly cast To build an LNG project you need various approvals They project financed You cite and permit many years in advance. Most of these projects that get you from 15 or 16 BCF a day to day nameplate to 35 are well on their way. And so that's the primary incremental demand driver in the country over the last 10 years and will be at least for the next five. We had moderate population growth during the tens and teens into the 2020s. We went through a period of stagnating electricity demand. You had energy efficiency and some other things driving down electricity demand while you had more demand for gas-driven generation. But it's really been in the recent past where compute has started to pull incrementally. But before that, you had LNG as the main driver demand. Now let's put those together. I just shared that we're gonna go from about 15 to 35 BCF a day of incremental LNG exports. And after evaluating every producing gas well and the entire pipeline and processing and gathering system, we have the capacity to add about 20 BCF a day of gas production even without AI compute We had sources and uses matched between our ability to deliver new natural gas from Appalachia Hainesville Permian and that which is supposed to leave the door through LNG Now I compute. There's so many different power generating ideas in order to power compute time to power that folks talk about so much. It sort of goes from the large scale, most efficient assets, which are GE for NOVA combined cycle, all the way down through the distributed generation assets, which we'll call fuel cells. We will add Wartsella Oys or Caterpillar, solar turbines. There are various local field level behind the meter assets. Those assets are also relevant. We have had to assign with an outside partner probabilities to all of this stuff. And so what we've gone about doing is we will start with our base case, which is we'll call it P50, everything with a probability of 50% or more, 50% being they have some approvals. They have usually a PPA, someone planning to buy power from them under contract. They usually have some sort of interconnection agreement or they're in process with the interconnection agreement. Those are the assets we've taken seriously in our base case. And so we'll call that the P50 level. And when you do that it about five BCF a day we think of very credible incremental natural gas demand associated with mostly AI compute Now importantly there are multiples of what we are considering seriously in our base case that have been proposed that will consume natural gas. Every solution today, a six series Bloom Energy latest gen fuel cell will take 150 million cubic feet a day of gas per gigawatt. The market has been assigning a high probability on them attaining two gigawatts a year of productivity or of manufacturing capacity. And that's likely to ramp to five gigawatts. There isn't a gas for that unless you take it from something else. In the extreme case, how high does that number get? If you start to move it down, say P30 or P0, that number can more than double and be 12 to 15 BCF a day by the early 2030s if unmitigated. One naive way to approach this is to say, like, this doesn't sound like that big of a deal, like 12 new in the extreme case. Just shut off the exports. Like, who cares? We didn't export natural gas for a long time. People domestically are not going to tolerate skyrocketing energy prices, especially when they think like the simple solution to this is just like stop shipping it out of the country. You just use it for ourselves. Why is the solution not just like shut off exports? It's more confident that contract law. There are rules. There are really good reasons. [10:00] We're exporting and these projects, you know, there's tens of billions of project financing and contracts attached to or associated with these LNG projects. And as the U.S. will be about a third of global gas supply in several years, our allies and other FTA and increasingly non-FTA countries are reliant on U.S. gas free trade agreement. So the answer is that it's both because it's a third of the global supply. That's really important for the rest of the world. And domestically, there's just contracts and investments. And it could be stopped, but it'd be very complicated. Yeah, let's step back for a moment. You have to have a starting place for a base case, which is typically starts with signed contracts. And what are the words on the page say? What's allowed or what's not allowed? When we set out to build the firm, we've had about 16 plus months to start to model almost every asset at a time at the atomic level. Along the way, there are numerous constraints and rules and regulations and contracts. When we set out to build this, it was about acknowledging those constraints for what they are assuming that contract law will be followed And then as we go through and build all of this we can flex up and down based on the choice to send less LNG out of our terminals for instance or slow AI compute growth, which is one of a solution which we're not really willing to propose because we know that there's insatiable demand. And so it's not popular to say slow AI compute growth. But to the extent that would happen, that would be another lever to reduce the pull or strain we expect in the system as the decade goes along. I'm just going to try to ask really simple questions here because so as not to minimize it, it's your view that in the bad to worst case scenarios, like this is like a full blown crisis. This is not like a small thing. This is like the story in the country. So I want to make sure like the reason we're going into all this detail is like in this scenario, it's it's really, really bad. And it's it's really bad primarily, I guess, through prices that maybe you can continue to articulate why we don't necessarily want this specific outcome and what we can do about it. But help me understand, And like underneath the United States right now or North America, there's a certain amount of gas, just like objectively. I'm trying to understand like how much of this is that we are literally going to run out of the gas that's under the ground versus it's just a problem of how quickly we can find out where it is get it out economically process it store it transmit it use it et cetera Like those seem like two separate problems like literally just how much there is and then what we can do with it and about it And so is any element of this problem like there just literally not enough of it Starting with what we'll call a resource in the ground. There's tremendous data availability. We can measure where we are in the exploitation of most of the major gas producing basins. Appalachia, which is primarily the Marcellus plus the Utica. the Haynesville, which is a key swing basin, and then, of course, the Permian. And let's just say the Eagle for these are oil-directed plays where the decision to drill and produce is driven by oil and gas as a byproduct. So in each of these plays, there are some stacked pay or zones where well penetrations output can be measured with a lot of data. What that allows us to do, when you digitize the acreage controlled by each one of these companies with polygon shapes that uses a bunch of Latin longs to drop in and associate a well with an area that's controlled, you can figure out what's left. And the reality is there is gas. And we've, as a part of our analysis produced the gas that is logically captured and can be produced from wells from existing acreage positions of all these companies And so there is gas We assuming it gets developed here That how you get to our 20 BCF a day of growth But there are other constraints It unbelievably cool that we can like literally know at this precision what is underneath the ground, often deep underneath the ground and hard to reach places. It's a technology story, right? That would be fun to tell sometime. But it doesn't sound like the actual problem is that we are literally running out of the stuff underneath the ground. We've also had a history of just finding new stuff that we didn't know existed before. So it sounds like the problem is more our ability to serve the demand in this kind of timeframe, not that we're literally going to run out of the resource over the next 20 years. So it's a little more complicated than that. We get through most of the existing captured inventory of companies in the next four or five years. And so if you think about, you bring on a new well, it has a decline rate, and each well, as it's stacked on an existing company-wide portfolio decline, and a lot of these companies' decline curves are maturing some, and so they don't have to- They're pretty steep in natural gas, right? It's steep initially in natural gas, but Expand and EQT and others have such mature portfolios, the replacement is less costly today than it would have been five years ago. So when you stack all of these wells based on existing acreage up on these companies, assuming they're going to drill optimally based on the forward curve, [15:00] which is depressed, and we'll talk about that, you get to this 132, 128, 130, 132 BCF a day of maximum deliverability. So we are assuming that all of these companies develop the rest of their acreage. But that's a flow metric, not a stock metric. That's a flow metric. What is possible when you use known well performance parameters to maximize production before you get to midstream and other surface level constraints, which we'll talk about. There's resource. We are depleting the known resource. If you were to assume prices go up meaningfully, you may unlock additional basins that are legacy known basins. We know a lot about most of the rock in the US. There are other known gas basins, but they're uneconomic. And furthermore, there is an infrastructure to really accelerate drilling and activity in those basins to solve this. The constraints are multifold. So the first constraint is the ROC. We have the ability, we think, to get to 128 to 132 BCF. I started in our highest estimate, which is 132 BCF, as a starting place, because that's how you solve the LNG exports we commuted to We will take the under on that but that where you can get to A common pushback as we have gone through this is there plenty of resource available to us in the Permian There plenty of resource in Appalachia A number of companies describe themselves as having a lot more inventory of wells to drill than we can justify with the facts. And I'll just leave it at that. But when folks meet with companies, they should ask to understand exact engineered locations on a map. Where do they have not just the ability to produce, but plans to have infrastructure on the surface to allow it to flow, for instance, and the ability within financial parameters to invest and produce. The resource in the ground, we're fairly far along in understanding it. We've accounted for all of the major productive basins in the country. And I do not think we're likely to be surprised by some new major shale find. At this point, knowledge of those things are pretty mature. To say it back, there's a lot of resource, but at this rate, we're depleting the known resources quite quickly. We are advanced in depleting the known resources, especially as we move to the next layer, which is infrastructure. Vanta automates security and compliance for over 16 fast companies like Ramp Cursor and Harvey keeping them audit around the clock It the number one agentic trust platform and it now helps companies like yours watch for the risks that show up between audits across your vendors your AI tools and your whole environment. Every new tool your team signs up for, every vendor that turns on AI features, is an opportunity for something to go wrong, and most security programs weren't built for AI's pace of growth. The Vanta agent works like a 24-7 GRC engineer in the background, finding issues, drafting fixes for you and cutting vendor assessment time by up to 50%. Whether you're a fast-growing startup or a global enterprise, Vanta helps you earn and prove trust. Invest like the best listeners get a special offer for $1,000 off at vanta.com slash invest. Ridgeline is the first end-to-end system of record with embedded AI for investment management firms, running portfolio accounting, reconciliation, reporting, trading, and compliance on one unified platform. Firms are moving off legacy technology and onto Ridgeline because of how far ahead Ridgeline's AI features are compared to anything else in investment management software, which is why I believe that firms that come out ahead in the AI era will be the ones running on Ridgeline's unified platform If you serious about your firm AI strategy Ridgeline should be part of that conversation You can request a demo at ridgeline There a bigger longer term question about is this just a parenthetical period of time between 2010 and now when we were awash in gas? Prior to that, we really weren't. And maybe after this, we really won't be. Let's talk now about, okay, we get gas out of the ground. It still has to go get processed and transmitted and used. What are the most important rate limiters in that part of the equation? In some cases, it's processing. The natural gas flows to the surface with natural gas liquids embedded therein. In some cases, there's sulfur or nitrogen that has to be dealt with. In some cases, it comes with oil, and so you have to have surface level infrastructure to produce the oil, which is different. Gas primarily has to be produced into a pipeline system, and there's a certain spec on regular pipelines that's 1,010, 1,030 BTU is the spec. So you have to remove enough of these other hydrocarbons to get it to pipeline spec to be able to produce it into the system to be consumed by folks downstream. So processing is the first major constraint. There are a couple of basins with a little bit of extra processing. We'll fill it up pretty quickly. We do not have processing yet to get to our assumed 20 [20:00] production target that's necessary, but processing would be something it takes two or three years to build at the midpoint. We generally know what processing investments are being made in projects have been announced by Permian processors or by Appalachian processors. We know where the materials and liquids handling throughput capacity will be in 27, 28 at least. You really would need to in the near term, even to get to our 20 BCF of incremental gas production, we're willing to estimate, you need to have more processing built. Build stuff now. Point one, but now imminently. Gathering small diameter pipes, gathering is what takes it from the wellhead to processing our pipeline system. There is a fair amount of disclosure around processing systems being expanded and built. And we would posit we have put all those on a map on top of every one of these wells at their latin longs in order to grow even a fraction of where we must have natural gas production go in the US. Gathering has to be invested in very materially over the next, imminently to get to the place where we can achieve 130 beats a day of production in the US. The last one is interstate gas pipeline system. And this is where I come back to your answer on LNG Lots of rules and regulations around these things Pipelines are monopolies for the most part Local distribution companies that deliver gas to your stove those are monopolies or oligopolies. In the last 10 or 12 years, we've really built one interstate gas pipeline. That was Mountain Valley Pipeline connecting Appalachia to Mid-Atlantic. The various environmental permitting, the regime changes, been made very difficult to build interstate gas pipelines. This administration, this is an apolitical economy, this administration has been trying to reduce the barriers to building interstate gas pipes. We've started to see some more progress to that end. There is an urgency to build more connectivity, to wheel gas around the country, to serve this incremental AI compute load. Before we keep going through this sequence here, can you just say what you think the state will be? let's assume that like there's roughly inertia in the system and like nobody listens to this. Nobody does anything. A lot of people listen to this and have ideas. But what is going to happen in the world? What will the state of the world be like in 2030 if none of this starts getting addressed sooner than later? What's your best guess as to what it looks like? There is a tremendous inertia around natural gas being the primary fuel to power AI Semi for instance leader in many respects they done excellent work on everything up to the power source And they behind the meter BTM load forecasts, generation and load forecasts where they match. They do everything up to the point where they don't assess where the gas will come from. And so the market has been focused on understanding the power shortage and trying to solve that. Generation generally, power generation, which could be solar and batteries, wind, nuclear, whether large scale or small modular reactors, SMRs, or natural gas. Natural gas is well supplied today, 26 and 27. And the result is that nobody is investing in gas. In fact, EQT is shutting in natural gas right now because they think it'll be more valuable later. The rig count, the things that we can see real time to figure out if the market is onto this tightness in 2020 and 2030, it's not apparent today. And so it's perpetuating this view that most Americans have, which is there's plenty of natural gas, because for 15 years, it's all we've been taught. And so there's a complacency that developed And we think that complacency is going to take us right up to the point where it too late So we do think the die has been cast where gas which is currently at going out to $26, $27, $28, the curve is flat. $29, $20, $30, the curve is flat because people believe the gas is abundant. That's despite all these AI compute announcements, despite what all of the companies are doing for their investments. Gas has lulled everybody to sleep. But what happens is these structural things start to fall in place in 27, 28, and we start to draw meaningfully in the middle of 28, early 28 on the gas system like we've never drawn before. And as we look at 2020 and 2030, we start to cut into the U.S. working gas storage, which is about four TCF total gas storage. There's kind of a range of high and low for that storage seasonally as we draw in summer and winter and then build in the shorter months in the spring and fall. When you get to the middle of 2028, we start to break very materially below in a historical way where gas available in storage has ever been before. And by 2029, we drop below all known historical storage evidence. And by 2030, we get pretty close to where we think satirist parasites, [25:00] us, gas storage looks very, very low. And at that point in time, because it's not happened before, we're forced to look at where has gas price gone during shortages. Well, we can look at Russia, Ukraine. Gas went to $8, $9, $10 in MCF because we send a lot more externally to Europe. We've seen various weather anomalies, polar vortex in 14, December of 22. And those prices have gone to $6 are $8 or $10, but those have been transitory. And what we're talking about are structural drivers of demand against a known possible production of gas, and they don't match up. You pull in a very historic way, starting in 28, to the point where the deficit gets really convex and unbounded. Meaning that gas prices could be 20 or something like this? I would hesitate to even put a price target on it, but at $8 or $10, we think you potentially shut off some of the US exports, their spot cargos, and they are leaving the border to capture uplift in Europe or elsewhere Those spot cargos may not be lifted and that gas is left in the system And we tried to account for that in our model but the spot cargos alone can solve this You would have to get into shutting off contracted cargos leaving our border via LNG to really start to mitigate some of this. And it's hard for us to count on the choice to shut off contracted cargos where there's some Japanese utility counterparty who has counted on it for its provision of electricity. In crazy convex outcomes like this, Can you tick through who you think the biggest winners and losers are? There are some clear natural gas producer winners. Expand Energy is probably at the top of that list. They probably control 70% of remaining core Haynesville Wells. The very closely known parameters of rock where we know it to be very productive. And so Expand, we think is far and away the biggest winner. Uniquely, Expand is CEO list right now. There was some turnover in the year. They're going through a search. The stock has plummeted over the last six months as a part of that search, and it's trading at four times EBITDA on a forward curve where no one believes what I'm telling you to be the case, even though we think modeling the facts gets you to a much higher gas price The stock has dropped The assets have not changed It has some of the highest quality rock in the country Highest quality upstream company in Appalachia is probably Range Range has significant room to grow production and materially grow returns to investors. So those would be the upstream companies. It's not going to be obvious in the first pass through this equation for most, but natural gas sets as the marginal fuel for the next in line power generating asset in each power market. As natural gas goes, power prices go in the country. And so if you think about the dispatch curve of different generating assets in the country, there are some where the fuel is free. That'd be solar, to less extent wind, hydro. Well, solar assets, which are growing meaningfully, they've been 90% of the interconnect key with batteries in the last 10 years in terms of new assets coming on other than gas. Solar assets stand to benefit from a windfall where electricity prices are going up because the marginal plant taking fuel where the price is increasing is rising while the sun costs the same. We think there are some companies positioned very well for margin expansion for no incremental capital costs XPLR tickers XIFR formerly NextEra Yield Co which is an interesting set of assets They have a windfall coming in the latter part of the decade because they mark their PPAs to market at much higher values without any capex. Clearway Energy would be another one, similar circumstance. And so solar assets at the utility scale especially stand to win. Maybe more interestingly, as it relates to some of our discussions in the past, residential solar, which has been suffering from really the first removal of tax incentives to install residential solar since like the late 70s. Residential solar assets are really one of the only ways to protect yourself from what's going to happen during the time of 10 a.m. to 6 p.m. once gas gets really tight in the electricity markets, what you pay for electricity at your house. So we think residential solar grows exponentially from here, even without tax incentives. For the first time, it's very economic with where electricity prices are likely to go to install solar, especially when it accompanies batteries, which make the electricity much more available around the clock. I think it's important to say you're an investor, like you have money behind this work. Beyond those two categories, are there any other like surprising winners, do you think, in all of this? Like what about nuclear? What about like Westinghouse or places like this? What we're talking about is a complex dynamic system where there'll [30:00] Choices to consume electricity are not at different times. And as I go through this, I want to make sure I acknowledge that there's no silver bullet solution for what I described as a convex situation with natural gas and therefore electricity prices as the decade closes. And there's no bridge fuel other than solar and wind because currently natural gas is the only flex fuel to get us to when we can bring on nuclear. We've spent a fair amount of time as well in the nuclear ecosystem. To us, large-scale nukes are the only solution that makes sense, which point us primarily to the 181000 Westinghouse units. Don't those take like five years to build or something like that? More than that, but at least they have a story past. And very brief history, we've built two nuclear reactor units in 30 years in the U.S., Vogel III and IV. Around that time, we also tried to build one in South Carolina called VC Summer, another nuclear project at the time, nearly bankrupted Scana, which was later pushed into the arms and the project was shut down. The muscle memory from trying to build large-scale nuke, especially in the back of Fukushima in 2011 Chernobyl through my island for three decades nuclear engineers and scientists and companies moved away from nuclear And then the Vogel 3 and 4 experiment where it costs three times as much and took 15 years I think from birth to commercial service. That's the recent memory of these nuclear units. But if you go out to the 2030s, what I'm describing in terms of gas deficit only gets worse in 31, 32 and beyond. And so in our mind, the only viable solution is to build large-scale nuclear as fast as possible, which would mean it needs to come on in 2033 or 2034, which is as soon as it can come on. Regulated utilities, hyperscalers, regulators should all align around that goal. But because people don't really believe that gas is in short supply as the decade goes along. They don't believe in the problem. They don't like the solution. They need to be convinced of the problem. And the country has a history of building pipelines to solve problems that exist today, not problems that will exist in five or 10 years. And so we're trying to get out ahead and see where the puck's going. And where it's going is we are going to need large-scale nuclear by 2033, 2034. You don't think SMRs can be a solution where you use smaller reactors to power individual data centers behind the meter and this just never touches the system Many of the SMRs are still science experiments The NRC and the U government are actually doing a fair number of things to break down the barriers to bringing those and to see if they work or not and what the cost will be and whether they can be scaled or not But many of these SMR companies are not set up to manufacture and truly scale for the solution that's needed to solve this problem, which is tens of gigawatts as you go into the 2030s. That points us to these AP1000. So since Vogel 4 came on and Vogel 4 experienced very material improvements from Vogel 3 in Georgia. Today, China is building 39 plus or minus nuclear reactors. 34 of them are one gigawatt plus. I think a third of those are modeled after the AP1000. We know a lot more today about building large-scale nukes than we did when these mistakes were made. So large-scale to us where it can be commercialized on a known timeline and where the costs are probably better than where we don't even know if we can scale the businesses yet in terms of SMRs. Large scale versus small probably wins in our mind. The two companies most levered to that would be Cameco, which owns 49 percent, Brookfield 51 percent. You'll probably find that the U.S. government, I think, agrees with what I'm describing. They seem to really be lining up and trying to facilitate commitments and early procurement which will de some of the supply chain which will help put timelines on this And when the Westinghouse comes public and it deeply undervalued within Cameco today So that an interesting one BWXT which is a super interesting company they're the primary supplier of nuclear for the U.S. Navy. They significantly benefit from the coming nuclear cycle as well and lots of dollar content in the APM 1000s. Who are the big losers, do you think, in this future? Well, sadly, the biggest losers of this would be the U.S. consumer. To the point where you take what I'm saying, and if we're even partially right, electricity prices rise, which you can see some of on the forward curves in these different markets. As electricity prices rise, you start to think about the trade-off. Are we going to export natural gas to foreign buyers? Are we going to use it for AI compute? Or are we going to try to keep consumer electricity price bills? It's an awful trade-off. I think it will probably start to contribute more to the public dialogue, the nimbyism that we're seeing already pop up in some places. We think AI is tremendously transformational. We're not anti-AI, but it consumes a lot of power. And we need to really focus on the 2030 to 2035 period. And the U.S. consumer is probably going to pay the bill in the meantime. Please note that most of the time. [35:00] So the solutions being proposed by the government are to consume more gas because everybody believes it's plentiful. Bring your own generator or generation. BYOG is a thing today. That's what the hyperscalers are being asked to do to site their data center in a certain Latin long. Well, that means more gas, not less. So every time you read a press release from Bloom or from... Think more gas. Think more gas. And you can use the energy efficiency of each one of those units and understand exactly how much more incremental gas beyond the base case that I just shared is dangerously tight. Another loser, and I want to be respectful, but some of the biggest winners so far, at least in the stock market, have been the manufacturers of gas turbines or distributed power gen sets. And when you think about those companies, it's been somewhat boom and bust. In the early 2000s, there was a boom to build as many gas plants as we could. The capacity was overbuilt and the industry really languished for a long time until now. And you've had just tremendous profitability and equity returns come from these companies over the last two years. But as you look at 2028 2029 most of them are adding more capacity again just like they did in the early 2000s What kinds of companies are these Caterpillar is I think they doubling their solar turbine capacity between now and the end of 29 which I would judge is just at the exact wrong time when people may be questioning whether they even want to deploy those assets because the gas is much more expensive than they planned. Bloom Energy, they've been topical recently because of other things that folks are talking about, the rare earths they use in their manufacture, for instance. But for us, we don't think that Bloom Energy's assets at two gigawatts or more will be able to get natural gas in competition with all of the other assets that are being deployed that will consume gas given the scarcity that we see. So those are two manufacturers of distributed generation or BTM generation that we think are probably more poorly positioned than investors appreciate. It really doesn't make sense to us beyond 2029, 2030 to build large-scale natural gas generation until we ramp up production meaningfully and you can make sure we have the security of deliverability of supply of gas that's consistent with our model. We could see orders slow very meaningfully for natural gas generating assets, even at large scale, as 26 progresses. And those could be some of the losers would be It just may not make sense to use gas for power generation for new or incremental assets after a certain point It seems like sort of like this whole memory shortage thing that we going through right now that hyperscalers might also be in trouble here if this is a key input to what they doing. Do you think that's a big problem for them? So as we've been socializing this a little bit, trying to learn more and have people poke holes, I've shared this with one of your recent guests. He listened to this. Well, this sounds like DRAM two years ago, slowly at first and then all at once. The lack of investment in capacity expansion is going to come up to bite us. And I think that's where the analog starts. When we think about the way this plays out and other analogs, that's probably the best one. And when we think about the cost of the hyperscalers, right now, energy is budgeted to be about 10% of their cost. Depreciation is the highest memory and other things factor into that as well. But energy's total cost of energy is supposed to be about 10%. If you plug in all of this compute and it's gas powered and we think gas could double or triple structurally, even without weather, it could end up being 20 or 30 percent of the cost of compute by 2029. We do think it becomes a much more material issue. Now the levelized cost of energy LCOE as it referred to takes into account CapEx it takes into account cost of fuel Everybody who are making decisions in this moment are using the forward curve for natural gas which is flat a little backward a little conteguable mostly flat out to the 2030s in the mid threes. That is a very attractive, low cost fuel for the hyperscalers to commit to when they're focused on solving everything else. Like, how do I get compute in place to manifest in this revenue growth in Entropic or elsewhere. For us, we're just focused on modeling objectively. When you plug in this computer, this computer, and this PowerGen source here and there, how exactly does it pull on the system of companies that we focus on? If you're forced to play devil's advocate in all of this and come up with a set of circumstances such that this is all much ado about nothing, and we're sitting here in 2030 and gas costs three What do you think is the most likely reason? Is it data center power requirements are much lower because we make performance breakthroughs or AI demand isn't what we think it's going to be? Like, what is this most sensitive to such that might be wrong? So after we did most of our work, we went on a bit of a listening tour to target conversations with who we think are maybe the subject matter experts in that thing. So energy storage or hyperscaler compute deployment and energy consumption [40:00] And the common pushbacks, which we've spent a lot of time understanding, are Permian oil play. Oil is high. Permian has lots of gas in the ground associated with it. Why can't Permian productivity just fix problems? So our base case model already accounts for the seven plus billion cubic feet a day of pipelines that are already being built or developed that come on between 26 and 2030. If there were a new gas pipeline that would come on between now and 2030, we would know about it because of the regulatory processes and the time it takes to build these pipes. So we've mitigated the risk of being surprised by the Permian by moving into the midstream to understand the bottlenecks constraints. So beyond the deliverability of the resource or the gas in the ground itself, how much can actually get to market and either leave via LNG export terminals or be consumed in Texas or nearby. We've already included that in our base case model. And so that will be one of the pushbacks is there's plenty of gas in the Permian. But I would posit this may be controversial. There was plenty of oil in the world before the Iran conflict surfaced. Eventually there will be oil a plenty again That why it was a a barrel before the Iran conflict In order to make more Permian natural gas you also have to be incentivized to make more Permian oil And those incentives didn't exist until Iran. In order to produce a lot more Permian gas than even these seven-plus-BCFA pipelines being built that we're already modeling, you'd need much, much higher oil for longer, which only exacerbates this consumer crisis that we are concerned about. So we don't think the Permian solves the problem. And then the other one is, well, you can locate a bunch of behind the meter local Permian power generation, which is happening, but we are modeling what has been announced and proposed. And if it's going to consume local Permian gas, that means it's not going to make it into the pipeline downstream. We can accommodate that with our model. In any case, we always try to think about technologies that can disrupt, change structurally the need and consumption of natural gas. And so it often leads us to focus on battery technologies. There's sodium and other battery technologies that are currently not commercial, but folks are getting a bit more enthusiastic about. The vast majority of economic battery deployment today is lithium ion It has a fairly fast discharge cycle Those are being deployed in earnest across the system And yes we are also modeling known battery deployments as a part of modeling this generating system across all fuel types. A step function battery technology change could be something that would affect you, but that would affect some of these pieces that I described, winners and losers, in meaningful ways. It would be a watershed moment that I would welcome because it would solve a problem that we're pretty concerned about. If you were a czar for a day and you had to decide everything that gets started to solve this problem, what are all the things that you would do to most solve and mitigate this? If I were the U.S. government, I would find a way to build entirely from beginning to end two to four AP1000 nuclear reactors. That would de-risk the supply chain. It would invite in and really open up doors to folks who want to see somebody do it first before they do it. With the hopes that we get more China-like, build 30 of them. Currently, there are 10 to 20 envisioned by the U.S. government through different groups in terms of nuclear reactions come on. But no one wants to be first I think we close to a few stepping forward But if the U just said hey we have billion to spend at the loan program office now the EDF it billion I think to spend by the end of 28 We need to build four nukes with this. That would de-risk this materially. And I think you would see a jumpstart the nuclear equation. Energy and resource availability, egress, and knowledge, we think are the biggest bottlenecks to productivity and deploying all this incredible technology that America has really been the leader of developing. When we set up our firm and started building a team, things are happening pretty quickly in AI, especially with regard to power. So we didn't set out to understand the now. It's important to understand now, but we set out to understand where the puck is going. and where it's going looks like it will meaningfully diminish growth if not dealt with. And so that's why it brings me to the nuclear solution as maybe the most viable long term. My comments on solar are probably the most important thing that I would do. I think everybody should get a solar system on their houses. It won't be perfect. It will deliver electricity when there's sun out. It may not deliver electricity when it's cloudy, but it's a way to protect yourself from very high peak power prices from 10 a.m. until 6 p.m., which is the biggest part [45:00] bill. I would incentivize people to study your state's rules and try to put incentives in place to really, really start to grow residential solar faster than what's been kind of a stagnating industry the last year after some incentives removed. That would be one place to be reinvigorating solar incentives because they're going to be needed in a few years. Can we build like a giant pipeline from Canada or something, try to tap our neighbors to help us solve this problem? Canada is an interesting partner of ours. They have the capability of delivering about 11 or 12 BCF, usually in January, periodically. And there are meaningful pipelines from Canada to the US, but largely it's seasonal and it helps us solve winter, but otherwise they're net imports most of the year. But we're not really set up to take from Canada year round. There are a few reasons for that. One is Canada has limited storage. It's about silver TCF, about a fourth of the US storage. If I think about this on a three to five year plus basis, Canada by far has the deepest and richest resource of economic gas in the ground, but it's been trapped behind pipe. I would build a one to two BCF a day at least pipe into the U Midwest the MISO power market and then wheel it around MISO PJM SPP ERCOT and try to satisfy this demand because I really don want to see demand slow I really don't want to see consumers' bills go up. What are the implications of this for the rest of the world? Well, the U.S. has become the leading provider of natural gas with our exports going from virtually zero to starting in the teens, and now we're at 15 BC every day. That'll be 35. Many countries in the world are building gas generating assets that are dependent on our delivery of that gas to them. And so lots of political conflict related tensions or bottlenecks today or Russia, Ukraine. Russia used to be one of the biggest deliverers of gas to Europe, for instance, and then we filled the gap to the extent we cannot deliver our 30 to 35 or more BCF a day of gas to the global consumer, rebalancing will be required. You probably impact Europe meaningfully, and they're left with a trade-off of taking Russian gas or much, much higher cost U.S. gas because they can't produce it there domestically themselves enough to satisfy their need Asia is a large consumer the largest consumer until Russia Ukraine of U natural gas and they probably will be at some point the largest consumer again with some of the outages in the Middle East We potentially hurt important allies at a time where we really want them to be allies if we can't send them the gas they need. And so it is pretty important that we don't curtail LNG, although that will certainly be one of the levers as we go out to the late decade that we will be forced to think about to deal with rising electricity prices in the country. Anything that we haven't talked about that has surprised you in this year and a half long analysis of trying to understand state of things and where we're going? Obviously, we've covered the big conclusion, which is scary and hard to deal with, even if we start acting now. Anything else that surprised you either in your work or in people's reaction to it as you've started to share it? really for the last two or three years, the phenomenon where CEOs, CFOs of companies, all of whom have really been asked by their investors or their products lend themselves well to deploying products to capture AI compute rent. What CEOs have said versus what is possible from the system I think is an interesting study that will happen over time The amount of capital made available to companies to make investments that are really short in the context of our work like the incremental distributed natural gas genset. It's an inefficient, high heat rate or high cost, inefficient asset that really should only serve as backup generation in any context outside of this fast time to power setting where AI compute needs the power now. Over four or five years, those assets may not even run. And so you've had tens of billions flow into these distributed power assets, all of which are short and will consume natural gas. And we really haven't seen anybody, including firms we really respect, question at any point whether there will be enough gas and what the cost will be when the time comes. It's been surprising that enough folks haven't put pen to paper to then start contracting gas to make sure they have supply certainty. We haven't seen more financial contracting. 28 is somewhat illiquid. This is why we really haven't seen the forward curve move. And we think that's where the action starts. As soon as utilities turn the page and start to really hedge or buy gas in 28 and... [50:00] We start to see all these natural gas generating companies start to think about securing supply. You're going to start to see a knife fight to secure natural gas physical in 28 like we really haven't seen before. And it's been surprising we haven't really seen any of this yet because 28, 29, the physical market tightens materially, depending on where you are. And the amount of money that's gone into unproven, untested, and the amount of capital being raised for things that really may not happen until 2035, maybe. That's been surprising, especially as it relates to you got expand trading at four times EBITDA, low to mid-teens, free cash flow yield on a gasp forward curve that is complacent to all of the objective things that we already know are likely to get plugged in. people are not really willing to look past summer heat or a slight outage in an LNG facility right now. But in six months, we start to see these companies roll forward to look at 28. You can start to see the forward curve really move up materially. And investors are not willing to look past near term appropriately supplied gas market but they willing to pay for something in 2035 that is totally untested or unproven It been surprising the assumptions and the inconsistency across sectors and industries we follow. So maybe in closing, what would be like the healthy challenge to pose to anyone out there whose business has as an input directly or indirectly energy prices? How would you encourage those CEOs, what questions do they ask of themselves or their business? Make sure when your assets are deployed that you understand exactly what the source of your natural gas will be. Make sure you have physical supply locked up and that you understand your counterparties and what will likely be very meaningful counterparty risk in two or three years. And counterparty risk isn't something we've really talked about during the last couple of years in the AI boom. But when it comes to parties being long and short, something that is moving a lot, imagine being short memory a year ago or 18 months ago and finding out all of a sudden you're short memory. That is what this natural gas market looks like to us, not two years out, but six plus months out And making sure you understand the physical provisioning of gas for your assets is important for the hyperscalers and for the buyers of simple cycle and CCGT large plants but especially for fuel cells We are very cynical whether you can deploy fuel cells at scale because there isn't the gas in the system to power those 24-7, 365. And so therefore, we treat them in our base case that I described as backup gen. To the extent you were to deploy fuel cells as baseload, Jen, that only pulls forward and is additive to the convexity that I described. CEOs and partners on projects, whether you are the E&C company trading at 25 times cash flow, which is a historically high multiple for an engineering construction firm, and your main business is building natural gas plants, and we may not be able to build or deploy more gas plants at a certain point in 2029, 2030, because gas is much more expensive and you may have regulators asking questions, the focus for you should be on how do you do a creative M&A to backfill and diversify your business so that you not entirely beholden to natural gas generating asset build For hyperscalers I know memory has been a pain point Natural gas could be 20 30 or 40 percent of their cost of doing business at a time when they're supposed to be reaching escape velocity with profitability. Performance per watt is probably a compute metric that we're going to care more and more about. I would say that the questions or the challenges for each industry of companies is a little bit different, but it's all focused on making sure you're managing risk and you understand exactly when your plans play out, how it can go wrong, which in this case means what if gas is not 350, but 10 or more? What happens if physical gas is questioned? What happens when consumers and therefore regulators start to ask questions? Matt, I love talking about the energy system with you. This was an especially fun one on the back of so much of your work. So fascinating and interesting. I hope as the U.S. has been very good at doing historically that lots of people listen and start to imagine solutions and also create the right amount of urgency to get those solutions in place and that we emerge from this more resilient, more capable, more efficient, all these things. Thanks so much for your time. If you enjoyed this episode, visit Colossus.com. You'll find every episode of this podcast complete with hand edited transcripts. You can also subscribe to [55:00] Colossus, our quarterly print, digital, and private audio publication featuring in-depth profiles of the founders, investors, and companies that we admire most. Learn more at colossus.com slash subscribe. You know how small advantages compound over time? That's true in investing and just as true in how you run your company. Your spending system is your capital allocation strategy. Ramp makes it smarter by default. Better data, better decisions, better economics over time. See how at ramp.com slash invest. As your business grows, Vanta scales with you, automating compliance and giving you a single source of truth for security and risk. Learn more at vanta.com slash invest. The best AI and software companies from OpenAI to Cursor to Perplexity use WorkOS to become enterprise ready overnight, not in months. Visit WorkOS.com to skip the unglamorous infrastructure work and focus on your product. Ridgeline is redefining asset management technology as a true partner, not just a software vendor. They've helped firms 5x in scale, enabling faster growth, smarter operations, and a competitive edge. Visit RidgelineApps.com to see what they can unlock for your firm. Every investment firm is unique and generic AI doesn't understand your process rogo does it's an ai platform built specifically for wall street connected to your data understanding your process and producing real outputs check them out at rogo.ai invest