Grantham Warns AI Spending Blitz Replaces Tech Monopolies With 'Dogfight'

格兰瑟姆警告：AI支出狂潮正将科技垄断转变为“残酷混战” 人工智能领域的巨额资本支出热潮正在将以往的大型科技垄断转变为一场激烈的七方混战，并可能引发历史性的市场崩盘。 著名市场策略师 Jeremy Grantham 剖析了金融泡沫的运作机制，将当今由 AI 驱动的暴涨与南海泡沫、1929年、1972年、2000年以及2021年等历史狂热期进行了对比。他解释了领先的投机性股票往往在大盘指数见顶之前就出现崩裂的规律，并指出大型科技公司已经从彼此互不重叠的垄断格局，转向了激烈且资本密集型的 AI 军备竞赛。除金融指标外，Grantham 还警告了更广泛的文明威胁，包括 AI 带来的巨大能源压力、全球暴跌的生育率以及日益严峻的生态限制。 SpaceX 市值达到 2.7 万亿美元，而预计年营收仅 200 亿美元，市销率超过 100 倍，反映出极度狂热的投机市场情绪。 (0:00) 传奇投资人 Jeremy Grantham 认为当前市场估值处于现代史最高水平之一，并指出投机性高 Beta 动量股在整体大盘上涨时绝对值走低，是泡沫见顶的关键警示信号。 (10:00) 回顾历史，1929年、1972年、2000年及2021年的主要投机领头羊均在指数见顶前先行崩跌，随后才引发更广泛的市场深度回调。 (15:00) 科技七巨头已从过去互不干涉的护城河垄断，转变为在 AI 领域展开全力以赴的资本开支军备竞赛，行业格局变得极度血腥与激烈。 (25:00) 历史上的重大泡沫往往诞生于真正改变世界的革命性构想（如铁路和互联网），即便短期面临剧烈的投机破裂，其长期深远影响依然真实存在。 (30:00)

GMO co-founder Jeremy Grantham argues that current market euphoria around artificial intelligence and multi-trillion-dollar valuations follows the mechanics of historic super-bubbles, where massive capital spending masks late-cycle structural fragility. Identifying Market Tops Through Breadth Divergence (10:00) Grantham defines major financial bubbles as two-sigma statistical deviations above long-term price trends and notes that market peaks display a consistent structural signature. Rather than all stocks declining at once, late-stage bubbles experience an internal breakdown in market breadth where high-beta, speculative leaders roll over while broader indices continue to hit new highs. According to Grantham, this breadth deterioration has occurred four times since 1925: in 1929, when low-priced speculative equities fell nearly 40% while the S&P surged 35% into October; in 1972, when the average S&P stock fell 17% despite a 17% headline index gain; in 2000, when tech collapsed 40% as value gained 15%; and in early 2021, when meme stocks and early speculative darlings plunged 35% to 40% well before the broader market peaked. Grantham illustrates this froth using his early venture stake in QuantumScape, which surged to $131 per share in late 2020, granting a pre-revenue battery laboratory a market capitalization exceeding General Motors. Tech Monopolies Collide in AI CapEx Race (25:00) While acknowledging that modern Big Tech firms generate substantial real earnings—such as Microsoft growing revenues from $143 billion in 2020 to $218 billion in 2025, and Nvidia surging from $11.7 billion in 2019 to $130 billion in 2025—Grantham views Nvidia's trajectory as an extraordinary stroke of luck rather than an unbreachable permanent moat. He observes that competitors will inevitably attempt to design around existing chip architectures over the coming decade. Crucially, Grantham observes a structural shift among the Magnificent Seven: whereas these giants historically operated as non-overlapping monopolies in search, social media, e-commerce, and electric vehicles, they are now engaged in direct, high-stakes competition over AI. Companies are escalating capital expenditures to between $127 billion and $200 billion to avoid losing the first-mover advantage. Grantham points out that this unprecedented, sector-concentrated CapEx cycle was powerful enough to derail what he had anticipated would be a standard cyclical bear market following his late-2021 exit call. Historical Bubble Precedents and Valuation Baselines (30:00) Grantham rejects the idea that great bubbles are mere fraudulent schemes, contending instead that history's largest speculative runs are built around transformative innovations. He ranks artificial intelligence alongside the 19th-century railroad mania as humanity's two primary speculative events, viewing the dot-com bubble as modest by comparison. However, true technological transformation does not prevent severe capital destruction; Amazon lost 92% of its value following the 2000 crash before emerging to dominate retail, just as extensive rail networks survived transatlantic railroad insolvencies. Contextualizing modern valuations, Grantham notes that the 21st century has averaged an elevated price-to-earnings ratio of 23x compared to the 20th century's 15x benchmark. He considers Japan's 1989 peak at 65x earnings—which required 35 years just to regain nominal highs—the ultimate historical extreme. Grantham emphasizes that his record consists primarily of valuation warnings rather than persistent crash forecasts, highlighting his only two explicit bullish macro calls: July 1982 when the S&P 500 traded at 7x earnings, and March 2009 at the generational low of 666. Ecological Boundaries and Global Demographic Decline (53:00) Looking beyond equity multiples, Grantham highlights the compounding physical limits confronting industrial society. He emphasizes that AI development requires enormous amounts of electricity and data-center capacity, aggravating global carbon emissions and energy supply strain. With humanity already estimated to consume resources equivalent to 1.7 planets—and five planets if global consumption mirrored American levels—Grantham cautions that deploying physical autonomous humanoid robotics will drive energy demands far beyond current infrastructural capacity. Simultaneously, Grantham identifies voluntary fertility decline as an evolutionary counter-trend. Approximately 65% of nations now sit below the replacement fertility rate of 2.1 children per woman, with China near 1.0 and South Korea declining faster, while sub-Saharan Africa's rates have dropped from seven to four children per mother. Grantham suggests that while shrinking populations might eventually reduce pressure on planetary boundaries toward a more sustainable one billion people, financial markets remain fixated on quarterly metrics rather than confronting these converging demographic and civilizational limits. Grantham leaves open whether artificial intelligence will establish benevolent superintelligence or cause acute ecological disruption, but warns that market history exhibits an almost one-to-one relationship between the extreme height of speculative valuations and the severity of the eventual unwinding.

GMO联合创始人兼长期策略师Jeremy Grantham指出，当前由人工智能与巨额资本开支驱动的市场估值已具备典型超级泡沫特征；历史规律表明，投机性垃圾股与基准指数走势背离是泡沫破裂的核心前兆，而狂热的市场叙事终将面临能源供给上限与人口结构萎缩等现实物理边界的硬性约束。 泡沫破裂的结构力学与四次历史信号 (10:00) Grantham指出，超级金融泡沫在统计学上通常表现为资产价格偏离长期趋势线两个标准差（2-sigma）的罕见事件。他强调，试图在顶部精准逃顶的动量交易策略极具欺骗性，而预示市场见顶的核心先导指标是“市场广度的内部瓦解”：在超级泡沫末期，高贝塔（high-beta）的投机性垃圾股往往率先停止上涨并转为绝对下跌，而核心大盘指数则在避险资金推动下继续冲顶。自1925年以来，这一技术性背离信号仅明确出现过四次。 具体而言，1929年美股核心指数在冲向10月峰值期间大涨35%，但低价投机股指数暴跌近40%；1972年“漂亮五十”（Nifty Fifty）周期中，标普主指数上涨17%，其内部平均个股却下跌17%，直接诱发了1973至1974年实际通胀调整后跌幅达65%的大熊市；2000年互联网科技股在2月见顶并在随日数月内重挫40%，同期标普500价值股却上涨15%直至9月见顶；2021年散户投机与Meme股票（如ARK基金重仓股及无营收即达千亿美元市值的QuantumScape）在年初崩跌35%至40%，而标普500指数彼时仍在刷新历史新高。 巨头AI资本开支竞赛与护城河重塑 (25:00) 针对当前AI热潮，Grantham分析认为，尽管NVIDIA营收从2019年的117亿美元迅猛扩张至2025年的1300亿美元，但其成功很大程度上源于游戏显卡架构意外契合AI计算的巨大运气，而非牢不可破的永久护城河。历史经验表明，巨大的先发优势将在未来十年面临Google等竞争对手及全新硬件架构的技术分流。与此同时，部分新上市与私有标的估值更趋狂热，例如SpaceX在2025年预期营收仅200亿美元的背景下估值冲上2.7万亿美元，市销率超过100倍，其募股说明书中近90%的估值叙事皆依赖AI逻辑。 更为根本的结构性转变在于，以往各据一方、享有类垄断地位的“科技七巨头”（如Alphabet主导搜索、Amazon主导电商、Tesla主导电动汽车）如今打破了互不重叠的界限，全面转入同一条AI赛道。各大巨头正在开展高达1200亿至2000亿美元规模的激进资本开支竞赛，使原本温和的寡头格局转变为残酷的消耗战。Grantham指出，正是这场空前的单行业资本开支狂潮，在2022年底硬生生中断了原本由宏观周期驱动的标准熊市进程。 铁路泡沫类比与历史估值均值回归 (30:00) Grantham驳斥了“泡沫纯属欺诈”的常见误区，指出重大超级泡沫必然建立在具有深远变革意义的伟大概念之上。在他看来，纵观数百年金融史，唯有19世纪横跨大西洋的铁路泡沫在社会影响力与资本狂热度上堪与当前的AI并驾齐驱；当年铁路概念在经历毁灭性崩盘后，遗留的基础设施与实际需求最终重塑了人类文明，正如Amazon在互联网泡沫破裂暴跌92%后重新崛起一样。因此，AI技术的划时代价值与其资产定价的泡沫化破裂并不矛盾。 在估值维度上，Grantham回顾了历史上最极端且最具破坏性的泡沫——1989年的日本市场。当时日股估值从长期低于25倍市盈率暴涨至65倍（远高于美股科技泡沫顶峰的35倍），泡沫破裂后日本资产历经35年才重新触及高点。Grantham强调，历史泡沫见顶的估值高度与后续崩盘的惨烈程度及调整周期几乎呈一比一的正相关。相较于短期的利率波动或债务指标，资产价格终究会遵从向长期均值回归的数学铁律。 物理边界承载力、AI风险与人口滑坡 (53:00) 除了金融估值风险，Grantham将视野延伸至市场日常忽视的文明级物理约束。他指出，AI算力的爆炸式增长正在对现实能源电网造成前所未有的压力，物理人形机器人的落地更将成倍放大充电负荷与碳排放。当前人类维系现有生活水平已需要消耗1.7个地球的生态承载力，若按美国人均消费水平计算更需5个地球。同时，顶级科学家与诺贝尔得主在AI终局安全性上存在剧烈分歧；正如Geoffrey Hinton所述，历史上从未出现高级智慧被低级智慧长期掌控的先例，而资本目前极度偏向追求竞争性能力而非对齐安全性。 与此同时，全球人口正面临前所未有的自主性断崖式萎缩。全球已有约65%的国家总和生育率低于2.1的世代更替水平；中国生育率已降至1.0左右（意味着每30年新生人口减半，90年内萎缩至八分之一），韩国生育率更是跌至一代人减少三分之二的水平。Grantham警告，当全球人口剧降与气候生态承载力超载同时发生，投资者若仅将注意力局限于短期的市盈率高低，实则是对正在收紧的实体经济与生态大边界的严重误判。 Grantham总结认为，尽管底层前沿技术的变革性潜力毋庸置疑，但脱离盈利基本面的极端估值终将面临历史均值回归与物理现实的严格检验；投资者在面对动量狂热时，必须清醒区分技术革命的长期成果与资本市场泡沫出清的周期性代价。

## Background 本期节目由彭博社（Bloomberg）旗下的知名播客《Odd Lots》于 2026 年 6 月 18 日在 RSS 及各大网络平台发布。节目邀请了传奇资产管理机构 GMO 的联合创始人兼资深策略师杰里米·格兰瑟姆（Jeremy Grantham），围绕其新出版的回忆录探讨当下的 AI 狂热与资产泡沫特征。节目简介指出，曾精准预警 2000 年互联网泡沫的格兰瑟姆将在此拆解大科技股的历史性分水岭、泡沫破裂前夕的预警信号，以及真正令他夜不能寐的长期生存危机。 ## TL;DR 节目从近期 SpaceX 等资产脱离基本面的极端高估值切入，格兰瑟姆系统剖析了资产超级泡沫的形成与破灭规律。他指出，泡沫见顶前最典型的先兆是“投机龙头逆势下跌而大盘权重虚假冲顶”的罕见分化，并对比了 1929 年、1972 年、2000 年及 2021 年的四次历史大背离。他强调当前科技七巨头（Magnificent Seven）已从过去安逸分治的轻资产垄断，骤变为在单一 AI 赛道上疯狂消耗资本开支的殊死“笼斗”。在访谈后半段，格兰瑟姆超越了短期的市盈率讨论，警告全球电网能源危机、环境恶化以及东亚乃至全球生育率断崖式崩塌等慢变量危机，远比短期的利率决议更为致命。 ## Atmosphere & style 对谈在幽默讽刺与末世警示的交织中展开，既有专业金融圈的会心一笑，又带有直面文明危机的沉重感。两位主持人 Tracy Alloway 和 Joe Weisenthal 风格敏锐而轻松，不断用市场调侃与即时数据向格兰瑟姆抛出尖锐问题；而已届高龄的格兰瑟姆则展现出历史学家般的超脱与直率，嬉笑怒骂间将市销率百倍的狂欢比作南海泡沫笑话。听众在聆听过程中会经历从技术狂潮的燥热、到泡沫破裂规律的理性冰水，再到最后跳脱出金融投机、直面人类存续危机的巨大认知震颤。 ## Bulletin - `01:25` SpaceX 上市不到一周市值逼近 2.7 万亿美元，按 2025 年预估 200 亿美元营收计算，其市销率（P/S）突破 100 倍，反映出市场借助 AI 与太空垄断叙事进行极端投机。 - `12:30` Jeremy Grantham 指出识别超级泡沫顶部的核心先兆：在指数与权重蓝筹继续冲顶的同时，前期领涨的高贝塔投机垃圾股率先掉头逆势深跌，该背离在 1925 年以来仅触发过四次。 - `25:00` 美股“科技七巨头”（Magnificent Seven）正迎来关键分水岭，从以往各自割据垄断垂直赛道的轻资产印钞机，演变为共同扎堆 AI 赛道、年掷数千亿美元资本开支内卷互搏的重资产巨兽。 - `30:00` 历史上最重大的泡沫往往诞生于真实的划时代技术构想（如 19 世纪铁路和 2000 年互联网），AI 具备超级泡沫的一切要素，其最终也难逃估值严重透支后的大出清宿命。 - `58:15` 相比于市场过度焦虑的短期利率与市盈率，全球生育率断崖式跳水（中韩代际萎缩、全球 65% 国家低于更替线）与 AI 带来的能源生态重压才是威胁文明存续的真正深层危机。 ## Key takeaways - `01:25` SpaceX 的极高估值是当前市场投机狂热的缩影，市场强行将其包装为核心 AI 基础设施以支撑超过 100 倍的市销率，与当年的南海泡沫如出一辙。 - `08:50` Grantham 澄清“永久大空头”标签仅是出版噱头，他已移交日常投资决策，50 年投资生涯的核心观察是金融市场只有超级泡沫的酝酿与崩解才是决定性的胜负手。 - `12:30` 泡沫破裂的终极预警信号是“投机领头羊与指数逆向分化”，即前期领涨的投机垃圾股深跌 30% 到 50%，而蓝筹权重仍拉动指数冲顶，该现象在 1929、1972、2000 及 2021 年皆曾出现。 - `20:00` 英伟达从游戏显卡切入 AI 算力属于不可思议的商业天赐良机，但其拥有的是“先发优势”而非坚不可摧的“终极护城河”，未来必将面临自研芯片与技术迭代带来的残酷竞争。 - `25:00` 科技七巨头原有的温和寡头垄断已被打破，巨头们全数涌入单一 AI 赛道进行“只能活一个”的军备竞赛，由轻资产模式彻底蜕变为重资本消耗模式。 - `30:00` 伟大的金融泡沫往往建立在真实的革命性构想之上（如 19 世纪铁路革命），尽管技术最终将永久改变世界，但投资者仍需在泡沫破裂中承担如亚马逊当年回撤 92% 般的惨痛代价。 - `37:00` Grantham 强调在长达半个世纪的职业生涯中仅有两次发出彻底清仓的“弃船”呼吁（2008 年 7 月与 2021 年底），以及两次旗帜鲜明的抄底做多（1982 年与 2009 年金融海啸大底）。 - `39:00` ChatGPT 的横空出世及其激增的巨额资本开支打破了 2022 年原本节奏工整的标准去估值熊市，硬生生阻断了实体经济滑向传统温和衰退的进程。 - `41:00` 过去 40 年全口径债务占 GDP 比例激增数倍但经济增速反而放缓，证明短期利率和债务对宏观经济的驱动作用被过度夸大，投资者无需过度纠结短期降息预期。 - `45:00` 2009 年标普 500 触及 666 点低点之际，Grantham 凭借《在极度恐惧中重新入市》备忘录精准抄底，践行了估值跌入历史极端低位时无需猜测精确拐点、必须制定计划重返市场的投资法则。 - `48:35` 1989 年日本股市是所有泡沫的始祖，其市盈率曾达 65 倍且最终带来长达 35 年的失落代价，证明泡沫规模与后续均值回归的崩盘烈度及出清时长呈高度正相关。 - `55:00` AI 顶尖学术界内部存在前所未见的认知撕裂，但在现实商业博弈中，资本更倾向于烧钱开发具侵略性与竞争力的“恶魔型”AI，叠加其巨大的能源与电网消耗，正在加剧潜在系统性生存风险。 - `58:15` 全球生育率断崖式跳水是正在眼前发生却被资本市场普遍漠视的灾难，东亚等经济体代际人口萎缩速度骇人，比微小的市盈率变动更能决定人类文明的长期命运。 ## Detailed walkthrough ### `01:25` SpaceX 狂飙：从估值神话到投机情绪的蔓延 节目录制于 6 月 16 日，主持人 Tracy Alloway 和 Joe Weisenthal 开场便聊起当天占据各大财经头条的震撼消息：SpaceX 股价单日暴涨 17%，市值接连超越亚马逊等科技巨头，正势头汹汹地逼近并有望超越微软。自 SpaceX 启动 IPO 以来，短短不到一周的时间里，整个市场仿佛被彻底点燃。彭博社一向偏爱诸如“自某年以来的最大涨幅”之类的夸张统计，而就在前一天（6 月 15 日），彭博全球亿万富豪榜上前 500 位富豪的财富单日总额更是创下了历史性的飙升纪录。 这种暴涨背后呈现出的估值逻辑令人咋舌。Joe Weisenthal 指出，按 2025 年预估营收 200 亿美元计算，SpaceX 的市值已经被推升至约 2.7 万亿美元的高位。在当前阶段，投资者甚至无法为其计算传统的市盈率（P/E），因为公司几乎没有利润可言，唯有单纯的销售额；其市销率（P/S）更是直接突破了惊人的 100 倍。为了给如此疯狂的估值寻找合理性，市场开始不断炮制各种叙事，最典型的莫过于硬将其包装为“AI 概念股”——理由是其部署了大量 GPU 和数据中心，外加星链（Starlink）在卫星和太空领域拥有无可置疑的垄断地位。 Tracy Alloway 翻看彭博终端的证券描述页面（DES page）时发现了一个耐人寻味的细节：SpaceX 的公司总部赫然写着“德克萨斯州星际基地”（Starbase, Texas），那是马斯克以自身项目命名的城镇。Tracy 打趣称，光是这个听起来极具未来感的地名，“恐怕就值 2.7 万亿美元了”。尽管 SpaceX 成立多年且拥有 2.2 万名员工，不同于当年互联网泡沫早期那些仓促上市的小作坊，但在如此极端的市销率下，投资者若想在未来获取回报，唯一的赌注就是押注它在极短期内实现近乎奇迹的爆发式增长。 ```mermaid flowchart TD accTitle: SpaceX 与市场狂热估值传导链 accDescr: 展示市场如何将太空与硬件资产包装成 AI 叙事并推高市销率至极端水平的逻辑链条 subgraph Narrative ["市场叙事包装"] A["SpaceX 卫星与星链垄断"] --> B["大规模 GPU 与数据中心部署"] B --> C["被冠以'核心 AI 基础设施'概念"] end subgraph Valuation ["估值脱离基本面"] C --> D["2025年预期营收仅 200 亿美元"] D --> E["缺乏净利润支撑 (无法计算市盈率)"] E --> F["市值飙升至 2.7 万亿美元<br/>(市销率 P/S > 100x)"] end Narrative --> Valuation ``` - **暴涨速度**：上市不足一周，单日跳涨 17%，市值连续超越亚马逊与微软等老牌科技巨头。 - **财务与估值悬殊**：以 2025 年预估仅约 200 亿美元的营收支撑起 2.7 万亿美元市值，市销率（P/S）超 100 倍。 - **后续上市浪潮**：市场将 SpaceX 视为风向标，预计年内 OpenAI 和 Anthropic 等重量级 AI 独角兽也将陆续登场。 --- ### `05:00` 狂热市场下的生存之道与 Jeremy Grantham 登场 面对全市场对 AI 概念近乎失控的狂热，摆在每一位投资者面前的核心疑问是：这究竟是不是一个巨大的泡沫？如果是，身处其中的人究竟该如何自处？眼下的股市完全被极度亢奋的动量交易（Momentum-driven）主导，正如阿波罗资产管理公司首席经济学家 Torsten Sløk 在此前的现场秀中所展示的数据，AI 因子几乎已经渗透并嵌入到了指数中的每一只成分股里，普通投资者几乎避无可避。 Joe Weisenthal 分享了自己作为散户的心态。他坦言自己的个人养老金账户采用的是最枯燥、最普通的宽基指数化被动投资，尽管没有通过 2 倍杠杆重仓押注韩国存储芯片股等极端标的，但仅仅跟随标普 500 指数享受水涨船高的被动收益就已经相当可观。然而，这种淡然是受托管理人无法享有的奢侈——如果一位以跑赢基准为目标的专业基金经理在当下仅仅拿到市场平均回报，资金委托人绝不会买账。在泡沫中“如何保持理智与清醒”，成为了职业投资人最残酷的考验。 为了解答这些困惑，节目请来了资深投资家、GMO 联合创始人兼长期策略师 Jeremy Grantham。Grantham 在投资界以多次提前精准预警历史级超级泡沫而闻名，其中包括 2000 年前后的互联网泡沫（Dot-com bubble）以及 2008 年次贷危机。他近期出版了新回忆录《永久大空头的形成：短期世界中长期投资的险境》（*The Making of a Permabear: The Perils of Long-Term Investing in a Short-Term World*），其经历恰好为当下狂热的市场环境提供了一面极具冲击力的历史镜像。 --- ### `07:15` “查验数字”：历史学家的狂欢与泡沫的重演 当被问及在当前全方位被 AI 兴奋情绪包围的市场中，投资者到底该做些什么时，Jeremy Grantham 给出的回答简练而直接：“避开炒作，查验数字”（avoid the hype, check the numbers）。他笑言，光是看到 100 倍的市销率，任何具备基本常识的投资人就该心知肚明了。 从一名市场历史学家的视角来看，Grantham 认为眼下的局面堪称“非同寻常且令人兴奋”。他穷尽一生、尤其是过去 15 年潜心研究的所有宏观重大风险与系统性危机，如今几乎在同一时间节点汇聚爆发，而全球金融市场却报之以“有史以来最昂贵、估值最高的股市”。这种极端割裂的反差让他大呼过瘾。 Grantham 大胆断言，50 年后的市场历史学家在回顾今天时，必将带着惊叹与嘲弄的心态看待 SpaceX 及其上市招股书，甚至会将其当作一本荒诞小说或一则行业笑话来阅读。他将此景与 18 世纪英国著名的“南海泡沫”（South Sea Bubble）相提并论：当年南海泡沫破裂前夕，曾有骗局宣称要开展一项“利益极其巨大但眼下绝不能对外透露”的绝密项目，以此大肆卷走股民资金并逃之夭夭——在 Grantham 看来，历史并未改变，今天那些追逐天价市销率的投机资本同样是在自投罗网。 > “在 50 年后，市场历史学家在回顾 SpaceX 时，会像读小说或笑话一样去读它的招股说明书，并拿它和当年南海泡沫的故事相提并论——‘一项具有如此巨大价值的事业，以至于此时此刻绝不可公之于众’。当年那些人卷走了巨款并逃之夭夭，而那些盲目的跟风者完全是咎由自取。” —— Jeremy Grantham --- ### `08:50` 澄清“永久大空头”的标签与 GMO 的角色分工 针对 Grantham 的新书书名《永久大空头的形成》，Joe Weisenthal 提出了一个敏锐的疑问：书名中的“永久大空头”（Permabear）究竟是不是一种外界强加给他的误解？在阅读时，读者是否应该在心里给它加上讽刺的双引号？因为无论是从 Grantham 过去深谙市场历史周期的理性分析来看，还是翻看其实际创立的 GMO 基金持仓，都绝不能简单归类为死空头——GMO 既没有全仓躲进美国国债和黄金等避险避风港里，反而持有微软（Microsoft）和 Meta 等科技巨头的股票，这与世俗认知中囤积罐头准备末日降临的“空头神棍”大相径庭。 Grantham 笑着坦白，在出版确定书名时，他自己确实极力主张给“Permabear”一词打上双引号，以表明这只是外界赋予他的刻板印象，但出版商出于营销噱头的考虑否决了他的提议。随后，Grantham 澄清了自己在机构内部的职责边界： - **早已移交投资决策权**：在他年满 70 岁时，便认为应当把日常投资组合的实际决策权全面交给更年轻的合伙人与同事，目前他本人完全不参与 GMO 投资组合的日常买卖。 - **专注于研究长期生存威胁**：他现在唯一的全职任务，就是研究关乎金融市场与整个人类社会存亡的重大长期威胁，这其中核心的一环就是“超级大泡沫的形成与破灭过程”。 - **泡沫破裂才是核心胜负手**：Grantham 强调，在自己超过 50 年的投资生涯中，他始终深信，在波澜壮阔的金融史上，真正具有决定性意义的唯有“超级泡沫的酝酿与崩解”；除了这些惊心动魄的历史大拐点之外，绝大多数正常年份里，基金经理只要按部就班打卡上班、洁身自好、做好日常功课，就能获得不错的市场回报。 ### `10:00` 泡沫时期的客户管理：在狂热市场中坚守常识的代价 面对像 SpaceX 估值飙升这类令人瞠目结舌的市场狂欢，作为秉持均值回归理念的长期投资者，承受的压力往往非同寻常。格兰瑟姆（Jeremy Grantham）坦言，在重大牛市中应对客户向来极其艰难，尤其是当客户背负着至少要跑赢基准指数的考核重压，看着别人赚取暴利而自己错失良机时，那种焦虑便会达到顶峰。然而自格林斯潘时代以来，金融市场几乎就是一个泡沫接一个泡沫地轮转：从科技泡沫到房地产泡沫，再到2021年末的高估值狂欢，以及当下的 AI 与高估值行情。GMO 在几十年中历经淬炼，付出了惨痛的学费——最为人所知的是在2000年前后的互联网泡沫期间，GMO 因为过早看空并坚持防守，在短短两年零一个季度内痛失了近一半的资产管理规模（AUM）。 谈及面对躁动客户的具体做法，格兰瑟姆强调，唯一的办法就是做到极致的诚实与透明。基金管理人必须把所有事实毫无保留地摊在桌面上，彻底剥除市场上所有天花乱坠的炒作（hype），向客户深入阐述市场的运作规律、定价机制以及均值回归的历史必然性。这种沟通是一场持久战，无论是牛市还是熊市都要日复一日地进行。真正波动的并不是投资机构的事实逻辑，而是客户自身的情绪周期：在熊市中他们变得沮丧而谨慎，而在狂热的牛市中他们又兴奋得雀跃不已。 - **历史教训**：GMO 在科技股泡沫顶峰前因坚持避开投机性科技股，导致旗下客户资产规模在两年零三个月内流失了整整一半。 - **沟通策略**：彻底剔除投机叙事，用客观事实与历史定价模型向客户持续揭示底层风险与均值回归机制。 --- ### `12:30` 泡沫破裂的终极先兆：投机龙头在上涨行情中的罕见分化 许多投资者都清楚当下的市场已充满泡沫（frothy），但普遍存在一种致命的侥幸心理，认为只要顺着动量行情参与，凭借自己的聪明才智，一定能在泡沫彻底破裂、所有人涌向出口之前抢先一步套现离场。针对这种盲目自信，格兰瑟姆指出，判断泡沫破裂的时机确实极难，但他总结出了一条百年来屡试不爽、且极具洞察力的核心规律。这一指标自1925年以来在美股历史上仅仅触发过四次。 格兰瑟姆将真正的超级泡沫定义为基于资产价格统计罕见的“双西格玛（2-Sigma）事件”。当泡沫在早期迅猛膨胀时，往往由高贝塔、高波动的投机热门股领涨，单年涨幅动辄达 60% 至 80%。但当泡沫进入尾声时，一个极为反常的现象会出现：**上一年度领涨的投机垃圾股开始掉头向下，甚至在整体大盘指数依然强劲上攻的背景下逆势下跌**。这种符号反转（Sign Error）——即投机前锋未能跟上甚至逆市大跌——是市场发出的最强烈的预警信号，宛如股市在崩溃前发出的“最原始的尖叫”（the greatest primal scream from the stock market ever）。 ```mermaid flowchart TD accTitle: 泡沫末期核心分化信号机制 accDescr: 展示泡沫从领涨投机股转入末期分化再到崩盘的过程 subgraph EarlyPhase ["泡沫早期：全面狂欢"] A["极度宽松 / 概念炒作"] --> B["高贝塔投机股暴涨 (60%-80%)"] B --> C["指数突破 2-Sigma 均值线"] end subgraph DivergencePhase ["泡沫晚期：致命分化 (触发信号)"] C --> D["大盘指数与权重股继续冲顶"] D --> E["投机龙头股与垃圾股逆市暴跌 (-30%至-50%)"] E --> F["资金被迫躲入权重蓝筹以维持仓位"] end subgraph BurstPhase ["全面崩盘"] F --> G["流动性断裂，大盘全线补跌 (跌幅达 50%-65%)"] end ``` --- ### `14:40` 历史上的四次超级分化：1929、1972、2000 与 2021 格兰瑟姆用历史上仅有的四次重大背离，佐证了这一末期分化指标的威力： 1. **1929 年大崩盘前夕**： 在经历 1928 年的投机狂欢后，1929 年前十个月标普大盘继续狂飙 35% 触及历史大顶。然而，汇集了众多高波动堕落天使股票的“标普低价股指数”（S&P Low-Priced Index）在整整一年里不仅没有上涨，反而逆势深跌了近 40%。投机客在潜意识中察觉到危险，正如花旗前首席执行官查克·普林斯（Chuck Prince）那句名言：“只要音乐还在播放，你就必须继续起舞。” 但聪明的资金不再愿意和纯投机概念股跳舞，他们选择躲进可口可乐（Coca-Cola）这类盈利稳定的蓝筹股中起舞。 2. **1972 年“漂亮50”（Nifty Fifty）见顶**： 1972 年大盘指数与个股出现极端分裂：标普指数全年录得上涨 17%，然而标普成分股中的平均股票却逆势暴跌了 17%。这种撕裂随后引发了自大萧条以来最悲惨的 1973-1974 年大熊市，无论大盘股、小盘股、优质股还是垃圾股全部被腰斩，在计入严重通胀后实际跌幅接近 65%。 3. **2000 年互联网泡沫破裂**： 1998 至 1999 年狂飙猛进的科技成长股于 2000 年 2 月见顶坍塌，暴跌 40% 到 50%。但标普 500 指数的其他传统成分股不仅没有下跌，反而逆势上攻 15%，甚至在 2000 年秋季重回历史高点，与成长股形成巨大的剪刀差。随后市场迎来了惨烈的全线去估值，2001 年持续下跌，2002 年更是再跌 22%。 4. **2021 年迷因股与颠覆性科技狂潮**： 类似的历史在 2021 年再度重演。散户抱团的迷因股（meme stocks）以及凯西·伍德（Cathie Wood）旗下的高成长投资组合在 2021 年年中见顶回落，到年底已自高点重挫 35% 至 40%；然而在同一时期的后六个月里，标普 500 权重股依然一路高歌猛进，完美印证了泡沫见顶前“投机股领跌、指数权重虚假繁荣”的分化规律。 - **1929 年数据**：标普指数至10月顶峰上涨 35%，低价投机股指数全年逆势暴跌近 40%。 - **1972 年数据**：标普大盘指数上涨 17%，而成分股中位数/平均个股下跌 17%。 - **1973-1974 年熊市结果**：通胀调整后美股实际跌幅约 65%。 - **2000 年分化细节**：成长股自2月见顶暴跌超 40%，标普其他板块逆势上涨约 15% 并在秋季打平高点。 --- ### `17:30` 亲历疯狂：估值超越通用汽车的电池研发实验室 为了展现 2020 至 2021 年末期投机泡沫有多么荒谬，格兰瑟姆讲述了自己与固态电池初创公司 QuantumScape 之间戏剧性的投资经历。几年前，格兰瑟姆获得了一个早期投资机会，但投资条件是“要么全部认购，要么彻底放弃”（all or nothing）。由于该投资规模对他的家族环保基金会而言集中度过高，他只能自掏腰包以个人名义全额买下。这也是他个人名下直接持有的唯一一只股票。 随后，QuantumScape 借壳特殊目的收购公司（SPAC）上市，以每股 10 美元挂牌，格兰瑟姆的初始持仓成本仅为每股 2.5 美元，上市瞬间便实现了四倍收益。然而接下来的三个月里，市场陷入了完全脱离理智的狂热： - 到 2020 年底，该股在投机热潮中一路被爆炒至惊人的每股 131 美元。 - 在 130 美元以上的价位，这家尚未实现任何商业化的研发实验室，其总市值甚至超越了百年汽车巨头通用汽车（General Motors）。 - 格兰瑟姆感叹，这比投资 SpaceX 还要不可思议：QuantumScape 当时“不要说利润，连一分钱的营业收入都没有”，本质上只是一家希望未来能研制出固态电池的科研实验室，却在散户与投机资金的推波助澜下变成了典型的迷因股龙头。这也成为了那轮狂热投机周期中最先见顶的极值符号。 ### `20:00` 2021年的狂热与英伟达的“深层价值”反思 在探讨当前市场狂热前，主持人和杰里米·格兰瑟姆（Jeremy Grantham）回顾了2021年那个极度荒谬的年份。当时市场上充斥着类似 QuantumScape 或上市估值飙至天际的 Rivian 等投机故事，甚至比目前的市场还要疯狂。然而，转向当下的核心焦点——人工智能（AI）时，主持人提出了一个极具挑战性的论点：调出彭博终端的数据来看，英伟达（NVIDIA）在2019年的营收仅为117亿美元，而到了2025年激增至1300亿美元，短短6年内营收增长了超过10倍。如果后见之明地看，2019年买入英伟达的投资者，难道不是做成了有史以来最伟大的“深层价值投资”（deep value investing）吗？ 对此，格兰瑟姆犀利地回应：如果你有先见之明（clairvoyant），那确实是深层价值机会。英伟达无疑是一家极其出色的公司，AI 也是一项极其壮观且具有深远意义的技术突破。但他指出，英伟达是过去30年里“能想到的最不可思议的一场天赐良机”（the biggest piece of luck）。他们原本押上全部身家去做的只是打电脑游戏用的显卡，结果时代突然转向，发现这种架构恰好是运行 AI 运算最完美的芯片设计。 在格兰瑟姆看来，英伟达目前拥有的并不是坚不可摧的“终极护城河”（ultimate moat），而只是在一个巨额资金博弈中的“终极领先起跑”（ultimate head start）。这个先发优势确实带来了营收十倍级的暴涨，但也仅仅是先发优势而已。像谷歌等科技巨头正在全力追赶、自研芯片并成为直接竞争对手；在接下来的十年里，残酷的市场竞争必定到来，甚至可能出现彻底绕开当前芯片路线的新技术，而商业史中这种颠覆屡见不鲜。 - 2019年英伟达营收为117亿美元，到2025年膨胀至1300亿美元，6年增长超10倍。 - 市场普遍惊叹于英伟达的护城河，但格兰瑟姆认为其本质是凭借游戏芯片架构偶然契合 AI 浪潮，获得了难以复制但非永久的“先发优势”。 --- ### `25:00` 微软的均值回归往事与“美股七巨头”的历史性分水岭 针对科技巨头的高估值质疑，主持人反驳称，微软（Microsoft）在2020财年的营收为1430亿美元，到2025财年已达2800亿美元；全球体量最大的企业在短短5年内将坚实的实际业务翻了一番，且连续多次击败华尔街的预测，这与1999年互联网泡沫中没有盈利基础的疯狂大不相同。格兰瑟姆顺势回忆起 GMO 过去对微软的量化投资体系：当时 GMO 设立了价值流（Value Stream）和动量流（Momentum Stream），每月在全市场筛选出最吸引人的前10%股票，构建出12个独立的月度组合。 微软曾长时间位列 GMO 最便宜的价值十分位数中，并一直保持到1999年7月才退出。格兰瑟姆解释，GMO 的价值模型并不单看静态倍数，而是利用净资产收益率（ROE）的均值回归规律，推演长期的未来盈利与股息。在他们的模型中，某些特质会极大减缓均值回归的速度——如果一家公司展现出不可撼动的市场主导地位和清晰的定价权，其 ROE 衰减就会非常缓慢。微软正是典型范例：极低的波动性加上无可争议的定价权，让模型判定即使其市净率（P/B）达到9倍依然具备投资价值，而市场上很多同行在7倍时就已经在抛售了。 ```mermaid flowchart TD accTitle: 从七大分治垄断到AI全面肉搏的模式转变 accDescr: 图表对比了科技七巨头在过去各自垄断垂直领域与如今在AI赛道迎头相撞的资本与竞争模式差异。 subgraph PastMode ["过去时代：各自割据、轻资产"] direction TB P1["7个垂直生态<br/>（电商、搜索、EV等各自垄断）"] P2["温和寡头竞争<br/>（如云计算三足鼎立、监管沉睡）"] P3["高利润率、轻资产运营<br/>（Idea-Heavy, Capital-Light）"] P1 --> P2 --> P3 end subgraph PresentShift ["关键分水岭（Watershed）"] CrossOver["AI 赛道引发全面交锋<br/>（赢家通吃的一号位争夺战）"] end subgraph FutureMode ["未来时代：单一战场、重资本博弈"] direction TB F1["7大巨头全挤入同个 AI 市场"] F2["极度残酷的恶斗<br/>（单年超2000亿美元Capex军备竞赛）"] F3["重资产、现金流大消耗<br/>（Idea-Heavy, Capital-Heavy）"] F1 --> F2 --> F3 end PastMode --> CrossOver --> FutureMode ``` 然而，格兰瑟姆话锋一转，指出了当前“美股七巨头”（Magnificent Seven）正在经历一场前所未有、却无人真正重视的“历史分水岭”： 1. **回顾过去**：七家巨头各自盘踞在不同的垄断领地中——特斯拉率先占领电动车领地，亚马逊统治电商零售，谷歌称霸搜索，反垄断司法机构如同在沉睡，没人像当年老罗斯福总统（Teddy Roosevelt）拆分标准石油那样去拆解它们。它们充分享受定价权，大发横财，并且属于“重创意、轻资产”（idea-heavy, capital-light）模式。即使是在云计算领域，也是微软、亚马逊、谷歌三家公司维持着极其绅士、守规矩的温和寡头垄断（genteel oligopoly）。 2. **展望未来**：七巨头突然一致决定杀入同一个单一市场——AI。它们正在上演极度残暴的野兽争斗（dogfight），纷纷擂胸示威，炫耀谁单年砸出的2000亿美元资本支出压过了对手的1270亿美元。所有人都笃信电影《高地人》里的名言——“最终只能有一个幸存者”（There can only be one），都妄图夺取第一个建立垄断地位的先发暴利。 3. **商业模式剧变**：过去它们是轻资产印钞机，如今在 AI 战局中全部变成了疯狂吞噬现金的“重资产”（capital-heavy）巨兽。格兰瑟姆讽刺道，这七家坐拥庞大现金流的富豪公司厮杀惨烈，简直应该把擂台摆在白宫门外进行一场血腥决斗。十年后再回望，市场一定会懊悔为何当初看不清“七个安逸垄断者”与“七只恶犬同场绞杀”之间的本质区别。 --- ### `30:00` 泡沫的真正本质：从19世纪铁路狂热到当前 AI 泡沫的宿命 针对主持人关于“哪一次历史泡沫与当下最为相似”的提问，格兰瑟姆纠正了一个大众常有的偏见：人们总以为金融泡沫都是骗局（con job），但事实恰恰相反，**人类历史上真正的超级大泡沫，无一例外都建立在几十年来最伟大、最真实的技术变革构想之上**。 在格兰瑟姆的坐标系中，唯一能与当下的 AI 泡沫相提并论的，是19世纪的铁路泡沫（Railroad bubble）： - **划时代的技术颠覆**：当年的百姓坐着时速7英里的马车赶到火车站，随后登上时速60英里的火车疾驰几千英里，这完全改变了人类文明的时空概念。 - **狂热与随后的崩盘**：正因为每个人都能清晰看到铁路必将改变世界，所有人都渴望分一杯羹，疯狂地将资金砸向铁路。结果，大西洋两岸爆发了人类历史上最惨烈的大崩溃，无数投资者血本无归。 - **废墟中重生的奇迹**：但在泡沫破裂的灰烬中，铁轨依然铺在那里，蒸汽机车依然在运转，客运货运需求依然旺盛，它最终确实重塑了世界。 格兰瑟姆接着提到了互联网泡沫。作为极为强大的构想，它在1999年催生了巨大的狂热（虽然也夹杂着许多荒诞的垃圾公司）。亚马逊在1999年股价暴涨了六七倍，而在泡沫破灭时，股价骇人听闻地狂跌了92%（"check it, 92%"）。然而，亚马逊最终同样从废墟中浴火重生，接管了全球零售市场。 | 历史时期 / 资产 | 核心驱动构想 | 泡沫形成的核心逻辑 | 泡沫破裂后的结局 | |:--- |:--- |:--- |:--- | | **19世纪 铁路狂热** | 时速60英里的火车替代7英里的马车，颠覆物流与客运 | 变革极其明显，全民涌入盲目投资，两岸资本严重过剩 | 股市彻底崩盘，投资者巨亏；但铁轨留存，永久改变全球经济格局 | | **1990年代 互联网** | 全球信息互联与线上商业革命 | 真实构想伴生投机概念，亚马逊等龙头估值被推至云霄 | 泡沫破裂纳指暴跌，亚马逊从高点回撤92%；随后脱颖而出接管零售 | | **当今时代 AI** | 机器智能重构生产力与认知边界 | 规模甚至可能超越铁路；所有资本与科技巨头全部押注 | 估值完全透支未来，七巨头疯狂Capex内耗，符合经典泡沫的所有特征 | 格兰瑟姆总结了孕育一场“伟大泡沫”必须同时满足的几大核心定律： 1. **良好的宏观经济环境**：经济景气越好，泡沫吹得越大； 2. **宽松的货币与充裕的流动性**：钱来得越容易，泡沫越坚挺； 3. **一个极度耀眼、能让全人类都一眼看懂其划时代意义的宏大构想（fabulous idea）**。 当下的 AI 完美勾选了历史上所有超级泡沫的特征清单。甚至在一些私营公司（如 SpaceX 等）的融资材料中，绝大部分估值也被强行归功于 AI，即便其底层 AI 技术在现实中正被两三个竞争对手打得溃不成军，但市场在极具诱惑的故事面前从不在意客观事实。格兰瑟姆断言，AI 的规模不仅远超互联网，甚至可能超越当年的铁路狂热成为历史之最；如果有人认为当下的市场不是泡沫，“那么未来等待你的将是极其苦涩的幻灭”（you are going to be in for a bitter disappointment）。在访谈的最后，主持人将话题引向格兰瑟姆的新书，询问他如何在高压的职业生涯中引导自身强烈的竞争本能，使其免于陷入追逐短期排名的陷阱。 ### `35:00` 投资哲学：站在更长远、更宏观的角度竞争 Jeremy Grantham 谈到自己对竞争的独特理解，将其类比为网球双打比赛。在他看来，打网球双打时每一分都必须当作生死之战来拼，并要挑选同样全力以赴的搭档与对手，这样才能体验到纯粹的乐趣。在投资和职业生涯中，他也将这种好胜心倾注于“思想”之中。他直言，如果自己将主要精力放在“利润最大化”上，原本可以赚到多得多的钱，但他始终将理念置于首位。对他而言，竞争的核心在于能否在智力上胜过对手。 而他发现的最简单制胜之道，就是把时间拉得更长、视角放得更广、站位比别人退得更远。一旦这么做，就会立刻意识到这几乎不是一场公平的较量——因为市场上绝大多数人甚至根本没有尝试这么做。每个人都在紧盯眼前和近期的利益，鲜有人在身处市场的同时去探寻三五年甚至一二十年之后的格局。这种与众不同的超长周期视角，让他能够既保持极强的竞争力和乐观心态，即便“经常看错”也坦然处之。 针对主持人关于“既然现在人人都在喊 AI 存在泡沫，你作为被外界贴上‘永久看空者（perma-bear）’标签的人，该如何脱颖而出”的提问，Grantham 坦言自己其实并不在意如何与其他看空者区分。他指出，市场往往无法厘清一个关键区别：指出“某类资产估值过高、未来 20 年的回报率会远低于估值减半时的水平”只是一种基于数学规律的客观陈述，但人们一听到这种判断就会叫嚷“你又说市场要崩盘了，你永远都在看空”。Grantham 强调，当自己真正极度看空并明确建议投资者“离场逃命”时，话说得非常直截了当，而他在漫长的职业生涯中其实只做过两次这种彻底的看空呼吁。 ### `37:00` 生涯中仅有的两次彻底看空呼吁：2008 与 2021 Grantham 回顾了他生涯中仅有的两次正式发出“弃船逃生”信号的时刻： ```mermaid flowchart TD accTitle: Jeremy Grantham 历史上的两次绝对清仓看空呼吁 accDescr: 展示2008年7月和2021年底两次决定性看空报告及其后续市场剧烈下跌的走势。 subgraph Call2008["2008年7月：弃船呼吁"] A1["2008-07-15 季报：'弃船 / Sauve qui peut'"] --> A2["清仓此前唯一看好的新兴市场"] A2 --> A3["随后4个月内（7/15至11/15）<br/>新兴市场指数腰斩（-50%）"] end subgraph Call2021["2021年底：狂欢开场"] B1["2021年底季报：'Let the wild rumpus begin'"] --> B2["标普500暴跌25%，成长股跌35%，Mag 7跌40%"] B2 --> B3["债市遭遇历史最惨烈的一年"] end ``` 第一次是在 2008 年 7 月 15 日的季度信中，他直接使用了“弃船”（法语“Sauve qui peut”，意为各自逃命）的字眼，甚至引用了儿歌“别逞能，快逃跑，留得青山在不怕没柴烧”，告诫客户千万不要承担任何不必要的风险。在此之前，GMO 唯一保持看好的是新兴市场，而 Grantham 在信中宣布改变主意，认为新兴市场也走到了尽头，必须能卖尽卖。在报告发布前不久，GMO 完成了公司历史上规模最大的一笔交易，彻底清仓了手里剩余的新兴市场资产。结果在随后的四个月里（7 月 15 日至 11 月 15 日），新兴市场指数直接腰斩，跌幅高达 50%，创下历史上主流指数最快、最惨烈的暴跌纪录之一。 第二次全面看空发生在 2021 年底，那篇季度信的标题取自绘本《野兽出没的地方》中的名句——“让狂欢开始吧（Let the wild rumpus begin）”，明确示意投资者赶快抽身避险。随后的市场演变完全印证了判断：标普 500 指数如火箭坠落般下跌了 25%，成长股重挫 35%，“科技七巨头（Mag 7）”重挫 40%，而债券市场更是遭遇了有史以来表现最差的一年。 ### `39:00` 被 ChatGPT 粗暴打断的熊市与 AI 资本开支潮 Grantham 戏称，自己原本节奏极佳、教科书般的熊市，被 ChatGPT 的横空出世“粗暴地打断了”。原本整个经济正在按照历史惯例，由于动物精神（animal spirits）的降温而平稳地滑向一场温和的浅度衰退。然而，AI 领域激增且规模庞大的资本开支（CapEx）打破了这一节奏，硬生生把实体经济原本萎靡不振的情绪重新拽了回来。尽管除科技股以外的其他标普成分股在其后 10 个月里仍在下跌，并未轻易随之起舞，但 AI 在股市和 CapEx 投资层面的狂飙突进彻底改变了游戏规则。 这种历史级别的转折在 Grantham 的职业生涯中仅见过一次。他坦承，自己根本无法预测这种事件，就像无法预知 COVID-19 疫情一样。对于崇尚历史规律的投资者来说，这种全新的“未知事物”极为棘手，因为历史没有参照坐标，必须诉诸完全不同的原则去推导。原本那是一场极其标准且容易预测的熊市——此前领涨股已经掉头向下，而蓝筹股还在勉力支撑，这种背离正是典型的见顶特征；但随后发生的事情，彻底超出了传统周期的范畴。 主持人引用 Apollo 首席经济学家 Torsten Sløk 的观点补充道，AI 交易表现出一种对传统宏观指标完全免疫的特质。即便美联储迟迟未如预期降息，企业也完全不在乎高利率，依旧大举进行资本开支，传统的衰退指标和商业周期似乎直接被这波 AI 狂潮给掀翻击碎了。 ### `41:00` 利率预测的局限与 40 年债务神话的破灭 谈及市场对利率变动的高度敏感，Grantham 直言，过去 50 年来关于利率及其走势的种种铺天盖地的预测让他感到“完全提不起劲”，他更愿意把这类分析留给别人，因为他认为这些因素的作用被普遍且严重地夸大了。 他指出，在过去半个世纪中，所谓的专家们一直在依据美国、日本以及其他主要经济体不断上升的“债务占 GDP 比率”，不断预言经济即将面临崩盘；但债务比率越走越高，预测的灾难却一次次落空，而债务依然在攀升。传统观点认为，利率的唯一功能是让举债变得更加容易，而便利的低成本债务理应推动经济增长。然而，现实数据却给出了一个巨大的悖论： - **格林斯潘时代之前的缓慢抬升**：在美联储前主席 Alan Greenspan 上任之前，全社会债务占 GDP 的比率仅仅因为金融业务复杂化而非常缓慢地小幅上升。 - **格林斯潘之后的 45 度飙升**：自格林斯潘上任起，债务占 GDP 比率以 45 度角斜率剧烈拉升，如果将全口径债务计算在内，在过去 30 年到 40 年间，债务比率翻了整整三到四倍。 - **宏观检验的矛盾结果**：在全球最大的经济体美国，用长达三四十年的时间窗口进行了一场宏观检验——债务占 GDP 比重翻了数倍，但同期的 GDP 经济增长率反而出现了下行。 如果庞大的债务真的是经济增长的真实引擎，这一历史事实在逻辑上就根本说不通。因此，Grantham 选择彻底远离这些纠结于短期利率预测的噪音。 ### `43:00` 21 世纪的估值新常态与生涯仅有的两次做多 Grantham 随后将话题拉回市场的长期估值，再次强调“指出市场高估”绝不等于“预测市场立刻崩溃”。他坦承，自 2000 年以来美股确实持续处于估值偏高的状态，GMO 也一直在公开指出这一点。如果以 20 世纪为基准，21 世纪的资本市场完全进入了一个不同的世界： - 20 世纪美股的平均市盈率（P/E）大致在 15 倍左右。 - 21 世纪以来的美股平均市盈率却常年维持在 23 倍上下，这绝非微不足道的增幅，从历史眼光来看确实长期偏贵，且当下的市场依然如此。 但与此同时，Grantham 澄清自己绝非只会看空的“永久死空头”。在他一生的投资生涯中，曾做过两次极其坚决的全面做多（bull calls）： 1. **1982 年 7 月的超级大底**：在 GMO 成立的前十年里，他们唯一被主流媒体引用的经历，是出现在一份早已停刊的行业周刊《华尔街快讯》（Wall Street Letter，当时附属于《华尔街日报》体系，主打行业内部动态）。在那篇通讯极其不起眼的文末，Grantham 抓住了公开表达观点的机会。当时标普 500 指数的市盈率仅有 7 倍，他断言“股票和债券市场即将在眼前迎来一场史无前例的超级大牛市”。随后的市场正如他所料一路狂飙，而随着估值水涨船高，GMO 在随后的漫长岁月里逐步转向谨慎。 2. **2009 年金融危机底部的翻多**：在 1982 年那次精准抄底之后，Grantham 耐心等待了数十年，直到 2008 年金融海啸导致市场在 2009 年彻底砸出深坑，他才迎来了职业生涯中第二次旗帜鲜明全面做多的历史性机遇。 ### `45:00` 在极度恐惧中重新入市：抄底2009年大崩盘的经典之作 杰里米·格兰瑟姆（Jeremy Grantham）回顾了他在投资生涯中极为罕见的一次公开抄底行动。在他长达30年的投资备忘录撰写生涯中，只写过两篇单页纸长度的信件，其中一篇被他自评为生平最佳之作——《在极度恐惧中重新入市》（*Reinvesting When Terrified*）。这篇信件之所以优秀，很大程度上正是因为其言简意赅。信中的核心逻辑非常纯粹：普通投资者绝无可能精准预测市场大底，因此根本无需去白费力气；真正需要关注的事实是，当时的股票市场估值已经降至22年来的最低点。按照GMO自有的七年期预测模型，标普500指数的年化复合回报率将达到12%，而在新兴市场和海外权益资产中，这一回报预期甚至更高。 面对恐慌至极的投资机构，格兰瑟姆在信中大声疾呼：“哪怕是一个糟糕的计划，也强过毫无计划。”机构投资者必须制定一份具体的资金回流计划并提交给投资委员会，开始按部就班地把现金重新配置回暴跌的市场中。格兰瑟姆幽默地对播客主持人调侃道，在投资界“只有落在纸面上的白纸黑字才算数”，光凭口头分析根本无足轻重，很快就会消散在以太之中。这篇短文原本第一时间提交给了《华尔街日报》，但对方拖延了四天迟迟没有回复。眼看时机紧迫，格兰瑟姆和他的公关宣传顾问决定不再等待，直接自主公开发布。正是这四天的阴差阳错，让这篇备忘录精准地在市场见底当天面世——那天标普500指数跌至历史低点666点，不及如今点位的十分之一。 - **信件名称**：《在极度恐惧中重新入市》（*Reinvesting When Terrified*），全篇仅一页纸。 - **历史估值与预期回报**：当时市场为22年来最便宜估值，标普500预期年化复合收益率达12%，新兴市场与海外市场预期收益更高。 - **发布巧合**：因《华尔街日报》拖延4天未复，最终于标普500指数触及666点最低谷当天由GMO自主发布。 ### `47:00` 散户入场与动量狂热：泡沫中的估值失效规律 主持人将话题引向当下的市场结构异动：近年来价值投资似乎持续输给动量交易，市场是否发生了某种根本性的结构断裂？大量通过Robinhood等散户交易平台涌入市场的资金，伴随着诸如交易成功时弹出庆祝动画等高度游戏化的界面，似乎让当代投资者变得极度短视，完全聚焦于短期价格波动和即时反馈，彻底背离了20世纪价值投资者的传统思维框架。 格兰瑟姆对此给出了极具历史纵深感的反驳：这绝非现代科技或散户平台特有的全新现象，而是历史上“每一次泡沫中毫无例外的经典重演”。在狂热期，投资者眼中的“关注价格”根本不是在寻找便宜的估值洼地，而是彻头彻尾的动量崇拜（Momentum）。只要资产价格飞速暴涨，大众就会陷入狂热，而底层的内在价值直接被抛诸脑后。从18世纪的南海泡沫（South Sea Bubble）、17世纪的荷兰郁金香狂热（Tulip Mania），到19世纪的英国铁路泡沫、1970年代的“漂亮50”（Nifty Fifty），乃至2000年的互联网科技泡沫，人类追逐动量的投机本性从未改变。 ```mermaid flowchart TD accTitle: 资本市场动量投机与资产泡沫形成机制 accDescr: 展示从追逐动量到估值脱钩，最终导致历史级泡沫破裂的演变过程。 subgraph Phase1["狂热启动期"] A["短期价格飞速上涨"] --> B["动量效应吸引海量资金"] B --> C["散户与机构弃绝基本面价值"] end subgraph Phase2["历史级泡沫共性"] C --> D["资产估值与基本面彻底脱钩"] D --> E["历史重演：南海泡沫/郁金香/铁路/Nifty 50/Dot-Com"] end subgraph Phase3["终局反馈"] E --> F["泡沫规模越庞大、越畸形"] F --> G["崩盘持续时间越长、出清代价越惨烈"] end ``` ### `48:35` 日本泡沫：所有泡沫的始祖与长达35年的代价 在格兰瑟姆看来，当前美股科技股的溢价虽然高昂，但若与1989年的日本市场相比根本不值一提。他将1989年的日本股市定义为“所有资产泡沫的始祖”（The mother and father of all bubbles）。彼时日经指数的市盈率一路飙升至令人瞠目结舌的65倍，即便在最资深的价值投资者噩梦中，这种估值也足以让人“半夜惊醒尖叫”。作为对比，美股在2000年互联网科技泡沫顶峰时的市盈率约为35倍（在此之前从未超过21倍），而日本在二战后市盈率从未超过25倍，却一口气被爆炒到了65倍。更荒谬的是，所罗门兄弟（Salomon Brothers）的研究团队当时在市场上四处游说，声称鉴于日本超低的债券收益率，日本股市的市盈率理应达到100倍。 ```mermaid flowchart LR accTitle: GMO在1989年日本股市泡沫期间的生存机制 accDescr: 说明GMO如何利用美国机构配置海外资产的红利抵消日本市场低配带来的业绩落后并最终安然度过泡沫。 subgraph Alloc["客户委托环境"] H["哈佛/耶鲁等首度配置海外股票"] I["客户基准锚定标普500而非全球指数"] end subgraph GMO["GMO投资策略"] J["坚决低配/回避65倍市盈率的日本股票"] K["整体海外投资组合收益大幅碾压标普500"] end subgraph Result["博弈结果"] H & I --> K J --> L["连续3年每年落后日本指数10个百分点"] K --> M["绝对收益优势巨大，客户零流失"] M --> N["日本泡沫破裂，GMO毫发无伤获利"] end ``` 幸运的是，GMO凭借独特的历史契机在这次浩劫中全身而退。当时国际资产配置刚刚在美国机构中兴起，苏格兰等欧洲老牌资管早已布局海外，而美国本土机构反应极其迟钝。GMO当时恰好在向哈佛大学、耶鲁大学等顶级捐赠基金推销它们有生以来的第一批国际股票投资组合。由于海外投资组合刚刚建立两三个月到两三年，当时没有任何客户使用国际指数来进行业绩基准对比，而是直接拿来同美股标普500指数一较高下。 虽然GMO因坚决不买昂贵的日本股票，导致其国际组合连续三年每年落后日本基准指数整整10个百分点，但由于整体海外市场相对标普500指数领先优势巨大，GMO不仅一个客户都没有流失，还牢牢稳住了市场份额。最终日本股市与当时并行的房地产泡沫轰然破裂。格兰瑟姆强调，日本泡沫给全球投资界留下了极其残酷的教训：市场为推高估值而疯狂狂欢的代价，是整整35年的漫长失落；即便经过35年日经指数重回高点，若扣除哪怕是极其温和的通胀因素，投资者依然未能真正解套。资产泡沫的体量越大、推得越高，随之而来的崩盘衰退期与破坏烈度之间的相关性就越接近于1。 - **日本泡沫顶峰估值**：1989年日本股市市盈率高达65倍（历史常态不超过25倍），所罗门兄弟分析师曾鼓吹应达到100倍。 - **泡沫破裂代价**：日经指数历经长达35年才重回前期高点，若计入通胀实际仍未完全复原。 - **均值回归铁律**：泡沫膨胀的严重程度与后续熊市清算的持续时间、惨烈程度呈几乎完全正相关（相关系数接近1）。 ### `54:15` AI能否拯救环境危机？顶尖智囊团的剧烈分歧 在谈话尾声，主持人将话题转向了格兰瑟姆近年来深耕的领域——气候变化、塑料微粒污染以及全球生态安全危机。主持人提出疑问：在当前这股席卷全球的AI热潮中，这种划时代的技术突破是否能够转化为拯救人类生态与环境危机的关键解药？ 格兰瑟姆坦言：“上帝啊，我也真希望能知道答案。”他感叹道，与以往任何科技革命不同，AI领域展现出了极其惊人、前所未见的观点对立。在常规的技术演进或宏观趋势中，人们往往能看到一种知识层级的阶梯分布——普通大众持有一种朴素观点，而处于行业金字塔顶端、掌握最前沿认知的顶尖专家们则形成另一种更为统一的高级共识。但AI却彻底打破了这一常态：在这个领域里，甚至连深耕AI长达30年的全球最顶级泰斗和诺贝尔奖得主们之间，都在发生极端激烈且互不相让的观点碰撞；无论是底层技术从业人员还是顶尖科研领袖，面对AI对人类未来的终极影响，都深陷在前所未有的巨大分歧与认知撕裂之中。 ### `55:00` 人工智能的极端分化、能源黑洞与掠夺本能 在谈及人工智能的前景时，杰里米·格兰瑟姆（Jeremy Grantham）指出，关于 AI 的未来走势市场完全不存在共识，两种极端叙事撕裂得非常剧烈。一种设想是 AI 将创造不可思议的充沛财富，人类只需要躺在海滩上享受机器人的服侍；另一种设想则是机器人更进一步，无论出于疏忽还是蓄意，彻底抹除人类。面对这种前所未有的终极复杂系统，任何人都无法预知确切走向，唯一能做的就是针对极宽维度的可能后果做预案。但眼下一个无法辩驳的事实是：AI 正在吞噬难以想象的巨量电力，并伴随着海量二氧化碳排放和沉重的环境重压。如今仅局限于笔记本电脑和服务器里的 AI 就已让电网不堪重负，如果未来还要生产数以亿计的人形机器人——每运作二十分钟就得像喝咖啡一样去充电——机器人在实体世界穿梭所消耗的能源将彻底超出当前认知，人类届时需要的能源产出将是现今全球总量的数倍之多。 人类早已在“透支度日”。格兰瑟姆引述深入研究该领域长达 30 年的资深专家结论指出，以可持续的方式维持当前全球收入水平需要 1.7 个地球的资源容量；而若要让全球所有人都过上美国人那样的生活方式，则需要整整 5 个地球。AI 在最理想的情况下或许能提供解决方案，但寄希望于此却面临着逻辑困境。正如“AI 教父”杰弗里·辛顿（Geoffrey Hinton）所言，在地球进化史上，极难找到一个智商更高的物种或文明心甘情愿被一个智力远逊于它的物种所掌控的先例。人类似乎盲目地默认未来具备自我意识的超级 AI 会心怀仁慈，然而现实中资本与科技界并未投入多少精力和金钱去研发一个“善良仁慈”的 AI，反而都在烧钱打造更强大、更具攻击性与竞争力的“恶魔型”AI。 这种机器系统天然继承了人类自身的生存法则——物竞天择、趁能抢的时候尽情掠夺、绝不为三四年后的事情忧虑。格兰瑟姆辛辣地抨击道，美国企业界与现代资本主义的运作逻辑也是一模一样的短视。人类的行为模式充满精神分裂：“我们周末在草地上陪孙辈踢足球，帮他们付高昂的学费，但一到周一就回到化工巨头或化石燃料公司上班，做着彷佛要把子孙后代全部毁灭的事。”这是所有物种刻在骨子里的掠夺本能与即时享乐冲动。然而，面对一种智力注定将碾压人类的非人智能，这种短视将演变成人类有史以来遭遇的最大生存威胁。 ```mermaid flowchart TD accTitle: 现代资本主义与AI发展面临的多重存在危机 accDescr: 展示短期掠夺机制、资源超载、人口暴跌与超级AI失控如何共同加剧人类的生存危机。 subgraph Mindset["人类掠夺性本能与短视资本"] A["短期利益驱动<br/>'抢占当下，不管三四年后'"] --> B["企业军备竞赛<br/>开发更具竞争力的恶魔型AI"] A --> C["无节制开采消耗<br/>维持美式生活需5个地球"] end subgraph Crises["现实正在恶化的三重危机"] C --> D["生态能源崩溃<br/>电网告急与气候失控"] B --> E["超级智能失控风险<br/>缺乏仁慈对齐的高阶物种"] F["全球生育率断崖式跳水<br/>中国、韩国等面临代际萎缩"] end subgraph Result["投资者的荒诞焦点"] D & E & F --> G["文明级生存问题近在咫尺"] G -.->|市场却在紧盯| H["微不足道的短期P/E估值与降息"] end ``` ### `58:15` 正在发生的现实危机：气候失控与全球生育率断崖 当全球资本市场还在为市盈率（P/E ratio）几个百分点的浮动焦虑万分时，格兰瑟姆提醒大家，真正决定命运的慢变量早已开始反噬，而且爆发的时间节点从原本以为的“几十年后”提前到了“当下”。气候异常不再是三四十年后的抽象警告，而是正在眼前演变成环境灾难；同样在眼前发生且被普遍无视的，是全球人口生产率的彻底瓦解，其恶化速度远超此前半个世纪甚至一个世纪的预测。 以东亚国家为例，中国当前的生育率已经跌落到 1 左右的极低水平。按照这种代际更替速度，人口规模每 30 年就会缩减一半，90 年内就会衰减到原有的八分之一。韩国的情况更为骇人听闻，其代际生育规模每 30 年萎缩至原有的三分之一，60 年后变为九分之一，90 年后仅剩二十七分之一。这意味着如果不发生逆转，韩国整个社会在短短一代人的时间跨度内就将面临实质性的“关门大吉”。目前全球有高达 65% 的国家生育率已经低于世代更替水平，并且这一趋势还在以肉眼可见的速度持续恶化。然而，由于人类的大脑构造天然无法对缓慢蔓延的长周期灾难产生警惕，大众和决策层对此几乎无动于衷，这种无知与对 AI 失控风险的漠视叠加在一起，正将人类推向一个充满不可测风险的奇点。 - **资源超载指标**：按现有全球生活水准可持续运转需 1.7 个地球；若复刻美式生活标准则需要 5 个地球。 - **中国人口萎缩曲线**：生育率逼近 1，代际每 30 年减半，90 年内降至初始水平的 1/8。 - **韩国人口崩溃节奏**：每 30 年代际缩减至 1/3，60 年变 1/9，90 年变 1/27。 - **全球生育现状**：全球 65% 的国家已处于世代更替水平（2.1 孩）以下。 ### `01:00:45` 人口收缩的悖论与马斯克的“火星多行星论” 听到格兰瑟姆提到“需要五个地球”，主持人乔·韦森塔尔（Joe Weisenthal）打趣道，这番话直接给埃隆·马斯克（Elon Musk）的 SpaceX 提供了最强硬的看多逻辑——因为人类如果真需要五个星球，就只能靠火箭移民火星了。特雷西·阿洛威（Tracy Alloway）也笑称自己刚想说同一个笑话。但格兰瑟姆严肃地指出，解决地球资源超载其实还有另一条截然不同的出路：让地球人口自然降到 10 亿人，而不是盲目膨胀到 80 亿、100 亿甚至 120 亿。 格兰瑟姆分析称，如果历史上某一个社会的统治阶层强行命令人们“少生孩子”，在大家本身强烈渴望生育的时代，除非动用残酷的强制暴力手段，否则断无成功的可能。然而现代人类却正在通过纯粹的自发选择走向人口剧烈缩减。这一代人拥有“成百上千个极其充分的理由”决定少生或不生孩子，这种全社会的观念转变在人类史上是绝无仅有的。但这一趋势所带来的数学结果非常残酷：如果一个文明无法保证每个女性平均孕育 2.1 个健康且受过良好教育的孩子，这个文明就在走向消亡。当下几乎所有发达国家都在这条下行曲线上狂奔；即便是被视为高生育率代表的撒哈拉以南非洲，生育率也在断崖式下跌，从以往每位母亲生育 7 个孩子暴跌至 4 个左右。虽然绝对数值依然相对较高，但其过去 50 年间丢失的新生儿降幅其实比欧洲还要剧烈。访谈结束之际，格兰瑟姆感慨地向两位主持人致谢，感谢他们愿意在节目里“拿出至少两分钟时间，讨论市盈率之外的真正严肃问题”。 ### `01:02:15` 主持人复盘：巨头内卷笼斗、历史加速与在末日面前谈估值的荒诞 访谈结束后，两位主持人意犹未尽地展开了延伸讨论。乔和特雷西回忆起播客早期录制过的各类奇特投机泡沫（例如佛罗里达地产泡沫与鲶鱼泡沫），并高度认同格兰瑟姆关于“科技七巨头”（Magnificent Seven）进入关键分水岭的判断。十年前，硅谷巨头之间的护城河与主营业务界限分明，谷歌做搜索、Meta 做社交网络；但如今，这种垂直领域的绝对垄断地位荡然无存。所有巨头都在白宫草坪上展开“贴身笼斗”，抢夺同一块蛋糕，为了在大模型和生成式 AI 赛道上不被淘汰，纷纷掷出惊人天量的高额资本支出（Capex），原先清晰的业务护城河彻底被打破。 特雷西随后感慨道，格兰瑟姆在节目尾声抛出的那些终极追问极具震撼力：人形机器人究竟是会变成海滩上端茶送水的侍从，还是会直接终结人类？人类会不会因为干脆不生孩子而自行灭亡？在这些关乎文明存续的宏大命题面前，金融市场还在喋喋不休地争论“市盈率是 20 倍还是 25 倍”或者“新任美联储理事凯文·沃什（Kevin Warsh）上任后第一次决议会不会降息”，显得荒谬又可笑——十年之后，没人会在意某次具体的利率决议。但乔敏锐地反驳指出，大型科技巨头正是通过将一切包装成“生死存亡的终极决战（existential threat）”，才得以合理化其当下的任何天量支出与估值泡沫，市场逻辑旋即又诡异地绕回了市盈率本身。 两人最后谈到了近期的“绳索历史”专题播客作为对照：人类在旧石器时代花费了近百万年，手中握着的唯一技术发明只有石斧；而在短短过去几年间，人类文明就集中迎来了 ChatGPT、GLP-1 减肥神药、电动汽车等一系列颠覆性浪潮。正如格兰瑟姆所感叹的，在 20 世纪 50 年代，可口可乐新开一家工厂就能成为头条大新闻，那时的历史进程平缓而单调。如今整个世界的运转节奏在被急速压缩，每天爆发的重大历史事件密度成倍增加，“时间本身彷佛正在加速”，在令新闻业繁荣的同时，也让跟踪这一切的人深感精神疲惫。节目在对制作团队的常规致谢中正式收尾。 ## Notable quotes > `07:15` 避开炒作，查验数字 — Jeremy Grantham > `07:15` 在 50 年后，市场历史学家在回顾 SpaceX 时，会像读小说或笑话一样去读它的招股说明书，并拿它和当年南海泡沫的故事相提并论——‘一项具有如此巨大价值的事业，以至于此时此刻绝不可公之于众’。当年那些人卷走了巨款并逃之夭夭，而那些盲目的跟风者完全是咎由自取。 — Jeremy Grantham > `12:30` 上一年度领涨的投机垃圾股开始掉头向下，甚至在整体大盘指数依然强劲上攻的背景下逆势下跌 — Jeremy Grantham > `14:40` 只要音乐还在播放，你就必须继续起舞。 — Chuck Prince > `25:00` 最终只能有一个幸存者 — Jeremy Grantham > `30:00` 人类历史上真正的超级大泡沫，无一例外都建立在几十年来最伟大、最真实的技术变革构想之上 — Jeremy Grantham > `37:00` 让狂欢开始吧 — Jeremy Grantham > `39:00` 粗暴地打断了 — Jeremy Grantham > `45:00` 哪怕是一个糟糕的计划，也强过毫无计划。 — Jeremy Grantham > `45:00` 只有落在纸面上的白纸黑字才算数 — Jeremy Grantham > `55:00` 我们周末在草地上陪孙辈踢足球，帮他们付高昂的学费，但一到周一就回到化工巨头或化石燃料公司上班，做着彷佛要把子孙后代全部毁灭的事。 — Jeremy Grantham ## Mentioned Jeremy Grantham, Tracy Alloway, Joe Weisenthal, Torsten Sløk, Cathie Wood, Chuck Prince, Geoffrey Hinton, Elon Musk, Kevin Warsh, Alan Greenspan, Teddy Roosevelt, SpaceX, GMO, Amazon, Microsoft, Meta, OpenAI, Anthropic, NVIDIA, Google, General Motors, QuantumScape, Rivian, Robinhood, Salomon Brothers, Coca-Cola, Apollo Global Management, The Making of a Permabear: The Perils of Long-Term Investing in a Short-Term World, Reinvesting When Terrified, Where the Wild Things Are, Wall Street Letter, The Wall Street Journal, Bloomberg ## Video description <p>Jeremy Grantham, co-founder and long-term strategist of GMO, has a long history of calling bubbles. As he recounts in his new memoir, <em>The Making of a Permabear: The Perils of Long-Term Investing in a Short-Term World</em>, that includes spotting the dot-com bubble of the early 2000s, which some people see as analogous to the current excitement over AI. And when it comes to today's market, there are a lot of signs of frothiness you could point to. In this episode, we speak to Grantham about how he sees markets right now, including a watershed change for Big Tech stocks, the signs he watches out for to spot when a bubble might burst, and what really keeps him up at night.</p> <p>Only Bloomberg.com subscribers can get the Odd Lots newsletter in their inbox &mdash; plus unlimited access to the site and app. Sign up at&nbsp;<a href="https://www.bloomberg.com/subscriptions/oddlots?in_source=oddlotspodcast" data-link-cookie="1" data-destination="https://www.bloomberg.com/subscriptions/oddlots?in_source=oddlotspodcast">bloomberg.com/subscriptions/oddlots</a></p><p>See <a href="https://omnystudio.com/listener">omnystudio.com/listener</a> for privacy information.</p> [Source](https://podtrac.com/pts/redirect.mp3/tracking.swap.fm/track/UVBrz8bN8aM2Xe47PEPu/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/8a94442e-5a74-4fa2-8b8d-ae27003a8d6b/60f6b6ac-8fe5-49e8-80a2-b46c01750478/audio.mp3?in_playlist=982f5071-765c-403d-969d-ae27003a8d83)

[7:15] Jeremy Grantham: 避开炒作，查验数字 [7:15] Jeremy Grantham: 在 50 年后，市场历史学家在回顾 SpaceX 时，会像读小说或笑话一样去读它的招股说明书，并拿它和当年南海泡沫的故事相提并论——‘一项具有如此巨大价值的事业，以至于此时此刻绝不可公之于众’。当年那些人卷走了巨款并逃之夭夭，而那些盲目的跟风者完全是咎由自取。 [12:30] Jeremy Grantham: 上一年度领涨的投机垃圾股开始掉头向下，甚至在整体大盘指数依然强劲上攻的背景下逆势下跌 [14:40] Chuck Prince: 只要音乐还在播放，你就必须继续起舞。 [25:00] Jeremy Grantham: 最终只能有一个幸存者 [30:00] Jeremy Grantham: 人类历史上真正的超级大泡沫，无一例外都建立在几十年来最伟大、最真实的技术变革构想之上 [37:00] Jeremy Grantham: 让狂欢开始吧 [39:00] Jeremy Grantham: 粗暴地打断了 [45:00] Jeremy Grantham: 哪怕是一个糟糕的计划，也强过毫无计划。 [45:00] Jeremy Grantham: 只有落在纸面上的白纸黑字才算数 [55:00] Jeremy Grantham: 我们周末在草地上陪孙辈踢足球，帮他们付高昂的学费，但一到周一就回到化工巨头或化石燃料公司上班，做着彷佛要把子孙后代全部毁灭的事。

[5:00] Jeremy Grantham: I think in 50 years, market historians will look back and talk in awe of SpaceX and read as a kind of novel slash joke its prospectus... and compare it with the stories they tell about the South Sea bubble, you know, 'an undertaking of such enormous value that it cannot at this time be revealed.' [15:00] Jeremy Grantham: As long as the music's playing, I have to keep dancing. And that, of course, is the name of that game. But he doesn't have to dance with Pumatech... You go off the cliff in Coca-Cola. That's the ideal. [40:00] Jeremy Grantham: Reinvesting When Terrified [35:00] Jeremy Grantham: Let the Wild Rumpus Begin [35:00] Jeremy Grantham: abandon ship

[0:00] Whatever your goal, trade show giveaways, client gifts, or team gear, 4imprint has the promo products to match. With thousands of options, from apparel and drinkware to tech and totes, it's easy to find the right fit for your brand and budget, with standout choices at every price point. And with their 360-degree guarantee, you can be 4imprint certain your order will show up just right, right on time. Explore more at 4imprint.com. 4imprint. 4certain. So there's a lot of noise about AI. But time's too tight for more promises. So let's talk about results. At IBM, we work with our employees to integrate technology right into the systems they need. Now, a global workforce of 300,000 can use AI to fill their HR questions, resolving 94% of common questions. Not noise. Proof of how we can help companies get smarter by putting AI where it actually pays off. Deep in the work that moves the business. Let's create smarter business. IBM. Health care doesn always work great If you ever waited on a refill or couldn schedule an appointment you get it That the kind of stuff Optum is changing They using data and technology to integrate patient care pharmacy and everything else So health care is connected not complicated What that look like Cheaper prescriptions that are easier to get and care that looks at the whole person how you need it Optum is helping make healthcare work as one for everyone. Learn more at business.optum.com. Bloomberg Audio Studios, podcasts, radio, news. Hello and welcome to another episode of the All Thoughts Podcast. I'm Tracy Allaway. And I'm Joe Weisenthal. Joe, I have a headline for you. Go on. This is a good day for headlines. We're recording this June 16th. You have a lot of headlines you could choose from. But one in particular, I'm sure you saw this already. SpaceX extends gain to 17% set to overtake Microsoft in value? I had a feeling, I think at some point, did they also overtake Amazon? Yeah, I think so. Like all of those? Yes. I think in the last week, I mean, we were traveling the week of the SpaceX IPO, but it just captivated everyone and day after day. And there are any number of superlatives You know at Bloomberg we really like superlatives It like the largest gain since X or the biggest These are more stats than superlatives right Well I suppose But there was also one I saw I guess yesterday so June 15th was the world 500 richest people collectively added billion to their wealth It was the single biggest. The numbers get bigger over time. There are some big times for the market. A nice day for them. But the crazy thing about SpaceX is, okay, you have this company that's now worth $2.7 trillion or something on $20 billion of revenue from 2025. And I can't even do, you know, if we're talking about valuations, you can't even do a traditional price earnings ratio because there's no earnings. There's just sales. And the price to sales ratio is more than 100, something like that. You know, look, like, I don't even know where to begin. I mean, you've seen. Choose a superlative. Yeah, choose a superlative. I mean, the argument that's being made is that there's a lot of arguments being made. But people are like, oh, it's actually an AI company because of all the GPUs and data centers that they've built. They really do have and we done an episode on this an extraordinarily commanding lead in space and satellites and Starlink etc The revenue is what it is I don know But also I think it fair to say things feel a little bit speculative at the moment right You see a headline like that, 17%. That's overtaking a stalwart of the tech industry in the space of, it's been less than a week, right? Yeah. I mean, look, I mean, the other thing is, look, these companies, and we're going to get these other big IPOs later in the year with OpenAI and Anthropic, presumably. I mean, SpaceX has been around for a long time. Historically, companies came public much earlier, et cetera. So obviously, it does have 22,000 employees. It's a big company, but the revenue is what it is. And if you're looking at it from a sort of like valuation based metric, you would have to say at a very minimum, investors, you have to be banking on very rapid growth in the very short term in the coming years to expect good returns here. Right. So obviously one of the big talking points. Do you know, I'm looking at the DES page for SpaceX. Do you know where it's headquartered? No, actually. Starbase, Texas. Oh. So they have their own town in Texas that they got to name, but that is their corporate info on the DES page, Starbase, Texas. Well, that's definitely worth 2.7 trillion. [5:00] All right. The big talking point in markets is obviously valuations. All of this AI frenzy. Is it a bubble? Is it not? But even if you think that a lot of this is speculative, my big question is, what do you actually do at this point? Right. Because so much of the market has been momentum driven recently. AI enthusiasm is pretty much everywhere. So we talk a lot about weightings in the indices, benchmark indices of big tech. But also, as we discussed recently with Torsten Slock in his great presentation at our live show, the AI factor is basically embedded in pretty much every stock at this point in time. Look, the way I look at it for my own investing in my retirement, I have like sort of very boring normie index-based investing. It's like, look, I may be missing out, but these are great returns, et cetera. I'm not managing other people's money. You know, other people like that luxury. You know what I'm saying? It's like I have the luxury of being able to say these are fine returns. I'm just getting from the S&P or whatever, roughly. And that's fine. It's a great year. And if like, you know, I'm not all in on two X levered Korean memory stocks, but I'm still like these are great returns. Had I been had someone been paying me to manage their money I don know if they be happy to get market returns right now Yeah that right So this poses a bunch of interesting questions obviously Yeah The big one being are we in a bubble or not And then I guess the next biggest one being well, what do you actually do? Yeah. How do you stay sane? That's right. How do you stay sane in a giant market bubble? And we do, in fact, have the perfect guest, someone who has managed to do that over the years, more or less, I think. We're going to be speaking with Jeremy Grantham. He is the co-founder and long-term strategist at GMO, as well as the author of the new memoir, The Making of a Permabear, The Perils of Long-Term Investing in a Short-Term World, and famously, the caller, the accurate caller of many previous market bubbles, including .com, which is the parallel that everyone keeps using. So, Jeremy, thank you so much for coming on OddLots. It's a pleasure. Thank you for having me. So what do you recommend investors to at this point in time? Because it seems like there's no escape from AI enthusiasm, to put it mildly. Well, my simple advice is usually try and avoid the hype, check the numbers. And 100 times sales pretty well does the job for you It easy for kind of market historians to be having a good time These are extraordinary times It seldom been more interesting I thrilled to be alive when all the major issues that I have spent my life studying, and particularly the last 15 years, are coming to a head basically at the same time. And to find that being met by the highest priced market in history, give or take, is extraordinary. And I think in 50 years, market historians will look back and talk in awe of SpaceX and read as a kind of novel slash joke its prospectus. And compare it with the stories they tell about the South Sea bubble, you know, an undertaking of such enormous value that it cannot cannot be at this time revealed. It scooped up a lot of money and ran off with it and they deserved it. The naming of your book, The Making of a Permabear. What is it like when I read that title? Should I have permabear in air quotes as in like people perceive you to be a perma bear I voted for quotes and the publisher didn like it Interesting. Yeah, because A, I don't – I associate you with like warning about the markets can get over their skis and understanding market history, etc. etc. On the other hand, I look at GMOs holdings and positioning and I see, you know, I do not see funds that are just overwhelmingly in treasuries and gold and safe haven assets. I see ownership of Meta and Microsoft. This does not look like the portfolio of someone who I would think of as a quote, perma bear, unquote. When I was 70, I figured it was time to leave my colleagues in charge of all the day-to-day decisions. I have nothing to do with the portfolio today. My only job is to study long-term existential threats to the market and society. And that includes the making and breaking of the great bubbles, which is fine because I have always, for 50 years at least, considered the making and breaking of the great bubbles to be the only thing that really matters. The rest of the time, show up for work, keep your nose clean, you do [10:00] fine. But the forming of these spectacular bubbles and their breaking really separates the men from the boys. So you touch on this in the book, but when we're in the midst of a major bubble and people are seeing crazy returns like 17% on SpaceX, what do you tell clients? They're missing out on these huge gains. I assume you're encouraging them to be patient to wait for that mean reversion, but how do you actually handle the pressure of having a customer, a client who is under pressure to, at a minimum, meet their benchmark? With great difficulty. It's always been difficult dealing with clients in a major bull market. Luckily, we've had quite a bit of experience because since Greenspan, we've gone from one overpriced market to the next, starting with the tech bubble and then the housing bubble. And then, in a sense, the end of 21 was a spectacular overpriced market and now this. So we've had a lot of experience and we're more careful at handling the clients now than I think we were in the tech bubble where we famously lost half our book of business in two and a quarter years Actually what does that look like in practice And I understand that you're not active day to day in the security selection, but obviously people who work for you are and they're very active and obviously in the client handling. What does that actually look like in practice? Good client management at any firm. In our case, being as honest as you can be, lay the facts on the table, make them as clear as you possibly can, try and take out 100% of the hype, and kind of engage with your client so that they understand exactly how you see the market working, why it does this, why it does that, how it's in general priced. and that's a long continuous job and it keeps going in the bull markets and the bear markets what changes really is the client's level of excitement they they become careful and miserable for a while and they excited and jumping up and down for a while but our process of trying to deliver the facts as we see them doesn really change much So I think a lot of people would probably agree that markets feel a little frothy at the moment But I think there's also a mentality. Again, it goes back to this momentum factor that's dominated in recent years. But I think there's a mentality that everyone just assumes they're going to be able to get out before everyone else. Everyone's going to head for the exits at the same time. So you enjoy the gains and then hopefully you're slightly smarter than everyone else and you manage to save yourself before the bubble actually bursts. How do you go about thinking about timing of the bursting of bubbles? What are the signs that you actually watch out for in terms of when everyone is heading for the exits all at once? Wow. I have a few rules that have worked more often than not. The one that's most interesting to me has only flashed four times since 1925. And that is that in the early phases of the bubble, that is we define a bubble as a rare two sigma event and based on the price only. And when it breaks through that let assume 1928 you have a huge move And the stocks that lead it are going up 60 80 percent for the year And then in 1929, the junkie flyers start to go down. It's not that they underperform a rising market. The market goes up 35 percent to the peak in October. They can't even get the sign right. The previous year's leaders, the spectacular movers, start to decline. And they spent the whole of 29 going down. And by the time the market broke, the S&P low-priced index, which regrettably was discontinued years ago, was down almost 40%. Low-priced index were a bunch of fallen angels with enormous volatility, very high betas. and they had had a spectacular 1928 and they started to decline. Why is that? I have a theory that this is relates to Mr. Prince's unusually honest answer. Why is he still in the bull market? As long as the music's playing, I have to keep dancing. And that, of course, is the name of that game. But he doesn't have to dance with Pumatek. [15:00] was the most advanced spec in 99, which was a hell of a good year to be the most advanced stock. And you don't have to go off the cliff in Pumatech. You go off the cliff in Coca-Cola. That's the ideal. And that's exactly what happened in 1929. People, in that case, actually Coca-Cola, they gravitated to the Coca-Colas and the radios and away from the junk, and the junk started to decline. That's an incredible signal. The greatest primal scream from the stock market ever. And nothing like that happens again. You have underperformance of high flyers, but you never have them go down in a decently rising market until, drumroll, 1972, the top of the nifty 50. The S goes up 17 The average S stock goes down 17 So I can remember the numbers forever And then we have biggest bear market since the Great Depression That was a truly miserable bear market 73 74 whether you had small cap large cap quality junk everything went down 50 percent and then went adjusted for inflation closer to 65 An absolute monster And nothing like it happens again until 2000 In the year 2000, you may remember, the growth stocks peaked horrifically in February. The rest of the S&P did not. The rest of the S&P rose about 15%. So you had a co-equal high in October. But the growth stocks were down 40%, having been as low as 50%. By then, they rallied a little. And back in September, you had the S&P as high as it was in March of 2000. Just amazing deviation between the growth stocks, who'd been making spectacular running in 98, 99, and early 2000, and the rest of the market that continued up. And then you had a very sharp break, the end of 2000, a steady break in 2001, and then to rub it in a miserable minus 22 percent in 2002. And nothing like that happens again until 2021 In 2021 you may remember the mean stocks peeled off By the middle of the year Cathie Wood and her portfolio were going down. By the end of the year, they were off 35, 40% from their peak. And yet during that same six months, the S&P powered ahead. It's really quite remarkable when they get the sign wrong. And unfortunately, I had an adventure with QuantumScape. QuantumScape turned out to be a meme stock without my knowing it because I had a huge position personally. And the reason I had a huge position is that I was offered an opportunity to invest several years earlier on an all or nothing basis. I took this big position or I had nothing. So I took a deep breath and it was much too big for our foundation which we have for the protection of the environment my grant and family and so i had to own it personally so it was the only stock that i owned it was a very big chunk and it came as a spack at four times my investment better better than a kick in the pants ten dollars a share up from two and a half and within three months it was 131 Let me just talk about 131 At the end of 2020 that was the very first stock to peak out At over 130 it was bigger than General Motors And it was a battery research lab They were going to design with any luck a solid state battery, and they still may, but they still have not got a battery on the market. So this wasn't like SpaceX. This had no sales, forget no profits, and was selling for more than General Motors. Actually, I think that is more speculative than SpaceX, which I think is a very, very high-ho. Whether you're planning a big tech event, launching a new campaign, or just stocking up on team gear, finding the right promotional products makes all the difference. 4imprint offers thousands of options, from on-trend apparel and premium drinkware to tech, totes, and giveaways, so you can find the right fit for any audience, purpose, or budget. You can customize it all, your logo, [20:00] your message, your look, and many items come with no setup charge to help you save. And if you're really watching the bottom line, you'll find standout choices at every price point so you can make a real impact while staying on budget. Plus, you'll get expert help, fast turnaround times, and their 360-degree guarantee, so you can be 4imprint certain your order will arrive on time and look exactly right. Whatever your goal, 4imprint makes it easy to find your perfect promo match. Explore the possibilities today at 4imprint.com. For imprint, for certain. Support for the show comes from public.com. If you're actively involved in your portfolio, you probably catch yourself repeating the same actions. Buying the dip, manually sweeping idle cash, putting on a hedge. On public, you can now create AI agents that handle all these tasks on your behalf. Just describe what you want to do in plain English. Like, if the VIX hits 25, buy a put option on the S&P 500. or if my cash balance goes above $20,000, move the excess into my direct index. You approve the workflow and your agent handles the rest. Monitoring the market watching for your conditions and executing your strategies exactly as defined An investing platform driven by your intent not just your clicks You can also get full read and write access to your account via the public api go to public slash market and fund your account in five minutes or less that's public.com slash market paid for by public investing brokered services by open to the public investing inc member finra and sipc advisory services by The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work. now we're helping companies get smarter by putting ai where it actually pays off deep in the work that moves the business let's create smarter business ibm i'm glad you brought back 2021 because that was a weird year wild times in a way that really was see i would argue the 2021 was actually much more wild than right now because of the proliferation of quantum scapes which was one of many stories that you could tell you could go back to the Rivian valuation at its peak or numerous others Okay, okay. All right, fine. But let's talk about AI for a second, because obviously we reach back towards memories of the dot-com bubble or 2021. But I'm looking at a chart on my terminal right now of NVIDIA revenue. So in 2019, NVIDIA had revenue of $11.7 billion. And in 2025, it had revenue of $130 billion. So the revenue has gone up, therefore, I think by over 10x in six years. We talk about, and everyone said, oh, market, very high valuations. Can't we say that the investor in NVIDIA in 2019 was getting the mother of all deep value investing, given what we now know in retrospect about what its earnings were about to do? Well, obviously, if you're clairvoyant, that was a deep value opportunity. NVIDIA obviously a brilliant company and AI by the way is a spectacular important development But Nvidia the biggest piece of luck I think I can think of in the last 30 years you have all your money in chips made for playing games And suddenly it turns out that that is absolutely the right design for AI. And you have such a running start that it may take, you tell me, five years, 10 years. It doesn't seem like the ultimate moat. It seems like the ultimate head start in a game where a head start is worth a ton of dough. It is worth a tentuple in your sales, but it's only a head start. Other people, Google, gearing up to be competitors. In the next decade, of course, there will be plenty of competition, maybe even technologies that end run the current type of chip. These things happen. History is full of them. Let me just push on that again. So I take your point about NVIDIA having this incredible mode from maybe arguably stumbled into the technology backwards because they were obviously in video games before. But like I'm looking at, say, even a Microsoft 2020 annual revenue of one hundred and forty three billion twenty twenty five annual revenue of two. [25:00] 180 billion. So we see the biggest companies in the world doubling real business in the span of five years, which does not feel to me anything like what we were seeing. And I have fond memories of 1999 myself. That's when I got personally interested in markets. These are gigantic businesses still putting up huge growth numbers that lap Wall Street forecast year after year. yeah and for the record when microsoft first appeared in a tradable portfolio yeah for gmo it was in our value stream we had a value stream and a momentum stream and we only bought the most attractive 10 and we did that each month for 12 months so we had 12 little portfolios Each one the best 10% for each month of the year. And Microsoft entered the cheapest value decile and stayed there until July of 99. And that because our value model looked at long term future earnings and dividends projected as best we could And we did that through projecting return on equity And we did that by looking at how return on equity regressed to the mean and which factors affected the rate Certain factors slowed it down. Market domination, price setting that you could prove, slowed it down way, way down. And Microsoft was the perfect example of this. It had very, very low volatility. It was clearly overwhelmingly the price setter. And consequently, our model said it was worth nine times book, not seven times book, but you were selling it. So if you have a good value model, you can buy these things, you can see them on occasions coming along. And yes, it's done spectacularly well and has had then particularly a much better moat to me than nvidia has today but if you will allow me a minute here please if you look at the mag 7 you look backwards in history and you can say with a pretty clear conscience each one dominates their seven different niches They have near monopolies on a global basis Even Tesla has a jumpstart the biggest and the best for a long time in EVs And you have Amazon beginning to dominate retail, Google research, et cetera, et cetera. Seven decent monopolies dominating the world. Justice Department, et cetera, perfectly sound asleep. No one is interested in pulling a Teddy Roosevelt. They're not going to jump in and slash and burn and divide Exxon into seven different pieces. They're letting these things grow and fix and set their pricing and make tons and tons of money. And then you look forward. Starting from today, does it look anything like that? Doesn't it look like seven companies deciding they're all in the same market, AI, that moving the most powerfully with the greatest investment is dominant? Are they not seen here beating their chest and saying my 200 billion in investment in a single year is bigger than your 127 yeah boo They know how much gets paid off to the first mover who grabs the market They all want to be the first. There can only be one, as they say in the movie. And there's seven of them fighting it out. It could be a very messy, blood-curdling game. I suggested that they have it outside the White House. What a comparison. Just imagine 10 years from now looking back and saying you couldn't see the difference between seven easy monopolies and a dogfight of seven vicious rich companies. huge cash flow huge understanding of the virtues of being dominant all deciding at the same time to fight out in one market and you could say yes there was the cloud what about that and the cloud was a nice well-behaved oligopoly three of them genteelly deciding to compete in genteel ways exactly the right thing to do if you find yourself in that position and clearly not the approach that is [30:00] being adopted this time. We have seen huge investment. You look back, they're idea-heavy, capital-light. You look forward, let's hope they're idea-heavy, but they are capital-heavy this time. It's like a watershed in almost everything that matters between the past and the future. And nobody seems to be talking about it in that way. And I don't get it. Well, can I just ask, related to this point, the dot-com bubble has come up a number of times in this conversation. Do you actually have a preferred historical analogy for the situation that we're facing now? because we've been through technological revolutions associated with speculative manias before, ranging from, I guess, what I would say are pretty real ones like the railroad bubble and the internet to kind of crazier ones like we're all going to go deep sea diving. Yeah, totally. And get rich that way. You are a connoisseur of historical financial market bubbles. Which one is most similar to the current period that we're in? Ah so people tend to think a bubble oh it has to be somehow a con job And it exactly the opposite The great bubbles are the biggest ideas for decades So the only one as big as AI is possibly the railroads. Of course, everybody could see that the railroads were going to change the world. You arrived at the railway station in a horse and buggy, for heaven's sake. and you went seven miles an hour, and then you got on a train traveling at 60 miles an hour and went a couple of thousand miles. I mean, it was utterly revolutionary. So what happened? Everybody could see that the railroads were going to change the world, which they did. Everybody wanted to have a piece of it, and they could. Everybody put their money in it, And you had the biggest bust on both sides of the Atlantic that you could imagine. And everybody lost their money in railroads and out of the ashes. The tracks were still there. The locomotives were still there. The demand was still there. And it changed the world. And then you fast forward to the Internet. Powerful idea Clearly by the way accompanied by a lot of silly stuff as well but underneath it a very powerful idea So you had Amazon go up six or seven times in 99 And when the market broke it went down famously, infamously, 92%, check it, 92%. And then it rose from the ashes, just like the railroads and inherited the retail market more or less. To have a great bubble, you have to have decent economic times. The better off, the better the bubble. You have to have easy money, the better and easier, et cetera, the better the bubble. And you have to have a fabulous idea. And you have to have it so obviously be important that everybody can see it. Now, they're very, very rare events, aren't they? This one is as big as anything but the railroads. I am not even prepared to say it isn't bigger than the railroads. It may be, but they're the two super champs. Besides them, I think the internet is a bit of a piker, but they're the two colossal ones. If ever there was a massive idea that will change is already changing the world it AI Does anyone not know that I think everybody knows it Does anyone want to put their money in it? I think SpaceX, et cetera, gives us a pretty good idea. 90% of the value, if you read the prospectus, is based on AI, even though that particular AI seems to be having its bottom kicked by two or three others as we sit. but wouldn't let facts get in the way of a really good story. And this is an absolute classic. It checks everything off one after another, which haven't been checked off many times in history. So this is it. If you think this is not a bubble, you are going to be in for a bitter disappointment. In your book, you talk about how competitive you were and growing up and wanting to play all different types of games. How do you channel competitiveness in a productive manner, career-wise, so that it doesn't hinder you or it doesn't make you chase performance or it doesn't make you worry about one year's performance versus another competitor? How do you make the competitiveness instinct be a good thing? I think – [35:00] Try and bring it to bear on a few areas that matter to you. Starting with, obviously, the most important, playing a decent game of doubles in tennis. And every point has to be played as if your life depends on it. And you pick partners and opposition who do the same. And you have a wonderfully good time. And then you look around for other things. And for me, it was ideas. And I could have made a lot more money if I focused on profit maximizing. But for me, the idea was the dominant principle. And the idea of being competitive was everything hinges on trying to outthink the enemy. the easiest way for me, it always seemed was to be longer and wider and more comprehensive and stand further back than the other guy. And what you quickly realize when you do that is that no one else is even trying. So this is not a fair fight. Everybody is focused on the near term. And if you want to profit maximize that not a bad idea And very few people are attempting to be in the market and simultaneously asking questions that are several years out and even to some extent a decade or two out So it been very easy for me to be both competitive and cheerful and often wrong. Since we're talking about career development now, I suppose, is it important when you're a perma bear or more accurately, when you're perceived to be a perma bear, to distinguish yourself in some way from other bears who are out there. Because again, at this particular moment in time, there are a number of high profile commentators who would say that AI is a bubble. So how do you actually stand out from, I guess, the bubble calling crowd? Yeah, I've no idea. I have only made two unmitigated bullish calls. The market has a really hard time telling the difference between, hey, this is overpriced. This is going to make you less money over the next 20 years than it would do if it was half price. They're just kind of mathematical realities. And because you say that they oh you said the market was going to collapse You have been bearish forever Now when I want to be really bearish and recommend you get out of the market I say so And I only done that twice On July the 15th, 2008, I wrote a quarterly letter which actually said abandon ship, Sovkeeper, the French equivalent, and actually quoted the nursery rhyme, don't be brave, run away, live to fight another day. do not take any risk you don't have to take. We all have restrictions on how much we can get out of the risk-taking business, but do not take anything you don't have to. Okay, that was pretty clear. And the last thing that we had been bullish about was emerging markets. And I said, I've changed our mind. We think this is the end of the line. Sell any emerging that you can. and we did the biggest trade that we had ever done, getting rid of the last of our emerging, I must say, shortly before we published the letter. What year? Sorry, which year was that? 2008, right? Oh, the emergence? 2008. Oh, okay, got it. And may I say that following that, in four months, the emerging market halved. I think it was the biggest sharpest decline in the history of any major index From July the 15th to November the 15th And actually slightly before that the whole index halved And the other bearish one was at the end of 2021, where the quarterly letter was called, let the wild rumpus begin. right that meant now get your tail out avoid the market i'm happy to say s&p went down like a rocket ship minus 25 growth stocks down 35 mag 7 down 40 and the bond market had the worst year in the history of the bond market and then as i also like to say my nice bear market was rudely interrupted by chat gpt and the economy that was doing its usual thing of gracefully moving into a mild recession because animal spirits were going down was also changed by massive and increasing capex spending on ai which dragged kicking and screaming the animal spirits of the rest of the economy they didn't change easily by the way yeah the s&p the rest of it went down for another 10 months, but they kept going so powerfully in the market and so [40:00] powerfully in the CapEx business that they changed the game. That's only happened once in history. And I don't know how to predict things like that any more than COVID. And new things are a pain. There aren't happily many of them, but they're the two most interesting ones in my career. COVID was novel. How do you treat novelty if you're a historian? You don't. You have to work it out on other principles. And this AI interruption of what was a perfectly ordinary, and I thought predictable bear market because it flagged my great discrepancy between the market leaders going down as the blue chips continued up. How do you do that? I don't know. Yeah, this is really striking. We recently, we had Torsten Slock at one of our events talking about this sort of imperviousness of the AI trade to what traditionally we'd call macro. So what you described, you're like, okay, here comes the expected, as you said, probably would have been a shallow recession. And now we see this investment this capital expenditure this completely it does not seem like the companies care at all about the fact that the Fed hasn cut rates as expected et cetera These classically macro indicators the recession business cycle et cetera had just seemed to be blown out of the water Yeah, blown out of the water by the AI trade. Yeah. And I must say that overwhelming interest rate and interest rate predictions has left me totally cold for the last 50 years. I leave that to other people. I think it's, in general, wildly exaggerated. I've lived in a world where for 50 years, the increasing debt to GDP ratio of the US economy, the Japanese economy, and every other economy has been predicting imminent collapse. And the ratio has gotten higher and higher, and then it's predicted double collapse. And it still keeps rising. The only function of interest rates is that it makes debt easier to acquire. And the function of easy acquired debt is that it helps the economy. One little problem. You go back to Alan Greenspan, you find that before he gets there, there is a very, very slow increase in debt to GDP ratio just because the financial business is becoming more complicated and then after him it rises at 45 degrees and it goes from a small fraction of GDP If you throw in all debt, it triples and quadruples. And it does it over 30 years. You have biggest economy in the world, 30 years to test what happened, quadrupling of the debt to GDP, and the growth rate goes down. So how can debt be a real mover of growth rate when you've had that wonderful macro test seen from looking back over 40 years? Huge increase in debt, decrease in GDP growth rate. Very strange. So I leave all that stuff way alone. If you will let me back up, I was saying too clear, get out of the market calls. plenty of the markets overpriced. The market has been overpriced since 2000. I admit it. It's been overpriced. And we have said so the whole time, because looking back at the 20th century the 21st century has been overpriced They used to sell at 15 times earnings We have been selling at 23 times earnings That is not a small fraction of an increase It has been a different world the 21st century. But based on history, it's been overpriced. And of course, it still is. But I have made two bull calls in my life. The only time for the first 10 years we got quoted was in something called the Wall Street Letter, Long Deceased. I think it was attached to the Wall Street Journal. And it was a weekly kind of gossip thing about the industry. and there hidden in the tail end actually of that letter is my first opportunity to quote and it's July 82 and the PE of the S&P is seven times and I say I think we're close to an unprecedented rally in both the stock and the bond market and I've always been thrilled to I'm sure you give people copies of this newsletter. And then the market shoots up and we become more careful for a long, long time. And then finally the market comes down in 09. [45:00] miracle that only occurs once every two lifetimes. We published a letter, one pager. Only two of those were done in my career of 30 years letter writing. It's called Reinvesting When Terrified and I think is the best thing I wrote mainly because it was short. And it just said, you won't call the bottom of the market. Don't bother with that. Don't even try it. Just concentrate on the fact that the market is cheaper than it's been for 22 years. Even on our seven-year forecast, you're dealing with 12% a year compounded returns in the S&P, equivalent to a higher numbers in emerging and foreign equity. Get together a plan. Take it to your committee. Any plan is better than no plan. You have got to start recycling your money back into the market. And the good news is, as far as I'm concerned, it only counts if you wrote it. Yeah. Saying it is too peripheral. It gets washed away into the ether. But I wrote it and we sent it to the Wall Street Journal who didn get back And day by day four days passed until my advisor on propaganda and I decided to hell with this Let post it ourselves And because of that delay we posted it the day the market hit it slow. 666 on the S&P 500, less than one tenth of where it is today. My God, this has been a bull market. Thank you for reminding two podcasters that it only counts if you wrote it. No, it's good. But to be fair, we do write some stuff. So there is that. I'm joking. I sympathize with you. Can I just go back to you were talking about how investors seem to have, to some extent, become more comfortable with higher price to earnings ratios now versus, say, for much of the last century. When it comes to value investing, we all know that value has been losing recently to momentum. Does it feel at all to you that something has structurally broken in the sense that investors are much more focused on price nowadays They much more focused on short gains rather than longer returns And at the same time you had a lot of retail money flow into the market courtesy of new platforms Robinhood And whenever I think about Robinhood I think about clicking buttons. And remember, they used to have the animation that would celebrate if you placed a trade. That's a lot of new money coming into the market that potentially thinks differently to the way investors for much of the 1900s actually thought. I think in every bubble, it gets very much like this. And you said much more focused on price, not in the sense that they're looking for bargains, much more focused on momentum. Right. That's what I mean. Yes. Price rises rapidly and they like it. And the value is irrelevant. You know, this is what happened in the South Sea bubble. And it's what happened in tulips and what happened in the railroads and it's what happens in the nifty 50 and the tech bubble. It's what always happens. This is not remarkable, by the way. This isn't even spectacularly overpriced compared to Japan. Japan is the mother and father of all bubbles. In 1989, it sold for 65 times earnings. And if that doesn make a value manager wake up in the middle of the night screaming once in a while nothing will Because we went up finally to 35 times earnings in the tech bubble never having been over 21 That's a pretty spectacular jump. But Japan had never sold over 25 times earnings and went up to 65. And I'm happy to say we survived that quite well through good luck. And the good luck was that international investing had only just come in. And we were selling people their first international portfolios that they had ever had, including Harvard and Yale. And that is quite remarkable, by the way, how slow the U.S. was. Scotland and so on had been doing this foreign investing for a long, long time, but it was not fashionable in the U.S. Because of that, no one was comparing international with an international index. It was only two or three years or two or three months that they had had an international portfolio. they were comparing it with the S&P because all their competitors were still in the S&P. And the novelty there was betting international against the S&P. It was winning. The international was so far ahead that we could underperform because of Japan. And we underperformed by 10 points a year for three years and we lost [50:00] No business at all. And we had a decent market share. And then, of course, Japan broke. The lesson from Japan is pretty clear. The biggest bubble in the history of the stock market, of an important stock market, second only to the land bubble in Japan of coincident, more or less coincident timing. and what was the price you paid for having it go from 25 to 30 to 40 to 50 to 60 65 and the solomon brothers team went around at 60 to 65 explaining that the bond rate in japan was so low it should be a hundred times i'm not kidding you i'm not kidding you what is the price you paid a last 20 years, really, not 10 years. 35 years have to go by before you get back to a high. And I don't think even that is adjusted for the modest inflation that they had. I mean, you want to have a bigger bubble and a better bubble? Go ahead. Just be advised that the correlation with a longer and worse decline is pretty well won. Thank you Whether you planning a big tech event launching a new campaign or just stocking up on team gear finding the right promotional products makes all the difference 4imprint offers thousands of options, from on-trend apparel and premium drinkware to tech, totes, and giveaways, so you can find the right fit for any audience, purpose, or budget. You can customize it all, your logo, your message, your look, and many items come with no setup charge to help you save. And if you're really watching the bottom line, you'll find standout choices at every price point, so you can make a real impact while staying on budget. Plus, you'll get expert help, fast turnaround times, and their 360-degree guarantee. So you can be 4imprint certain your order will arrive on time and look exactly right. Whatever your goal, 4imprint makes it easy to find your perfect promo match. Explore the possibilities today at 4imprint.com. 4imprint. 4certain. Support for this show comes from public.com. If you're actively involved in your portfolio, you probably catch yourself repeating the same actions. buying the dip manually sweeping idle cash putting on a hedge On public you can now create AI agents that handle all these tasks on your behalf Just describe what you want to do in plain English Like, if the VIX hits 25, buy a put option on the S&P 500. Or, if my cash balance goes above $20,000, move the excess into my direct index. You approve of the workflow, and your agent handles the rest. Monitoring the market, watching for your conditions, and executing your strategies exactly as defined. An investing platform driven by your intent, not just your clicks. You can also get full read and write access to your account via the public API. Go to public.com slash market and fund your account in five minutes or less. That's public.com slash market. Paid for by Public Investing. Brokered services by Open to the Public Investing, Inc., member FINRA, and SIPC. Advisory services by Public Advisors, LLC, SEC Registered Advisor. Complete disclosures available at public.com slash disclosures. The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we reduced costs by millions slash repetitive tasks and freed thousands of hours for strategic work Now we helping companies get smarter by putting AI where it actually pays off deep in the work that moves the business. Let's create smarter business, IBM. You know, we could talk for another hour just on international, all these things, but I have one last question. You're known for having a lot of personal sort of like other interests besides investing particularly related to the environment, climate change you talk about. You wrote a letter, I believe, last year about plastics and other forms of dangers to the environment. Do you have any optimism at all that AI in particular will be of service to humanity in tackling some of these concerns that you have about ecology and so forth? god i wish i knew it the spectacular thing about ai is the degree of difference of opinion you know often you find that the rank and file have one view and the hot shots who know the most have a different view but this is not like that this is you have nobel prize winners who disagree violently you have real experts who've studied it for 30 years who disagree violently yeah you have the rank and file with as much experience as they could have who disagreed [55:00] violently. There is simply no agreement on the future of AI. It will either make us all incredibly rich, we'll sit on the beach and be served by robots, or the robots will go one step further and get rid of us inadvertently or deliberately. This is not bad. This is the ultimate complexity that one has ever heard. And you cannot possibly know what is going to happen. You can only plan for a wide range of outcomes. But we know for a fact that it chews up enormous amounts of electricity. We know for a fact that that is associated with an awful lot of carbon dioxide production and real pressure on the environment. So we start knowing that it will be tough. And by By the way, you make robots every 20 minutes. These humanoid robots have to go off, take a coffee break and plug themselves in. Yeah, they will run through energy like we have no idea. We can hardly support the energy demands of current AI confined to your laptops The energy demand of having machines running around will dwarf that beyond recognition We will have to have multiples of the global energy production that we have now We are simply living beyond our means. The real experts who studied for 30 years say we need 1.7 planets to maintain the current level of income in a sustainable way. and if we want to live like Americans we need five planets and AI in the best of all possible worlds might help address this but it's hard to imagine AI becoming self-aware and being better at everything than we are it's hard to think of an example as Jeffrey Hinton would say where a smarter civilization, a smarter species, has been dominated by a comparatively stupid species. We somehow implicitly rely on their benevolence. We are not spending that much time and money trying to design a benevolent AI. We are spending money trying to design a more powerful competitive devil type of AI It inherited our style You know, humans have been the survival of the fittest. Grab what you can, why you can. Don't worry too much about three or four years from now. I find that exactly the same in corporate America and capitalism, by the way. we don't act as if we value our grandchildren we play soccer with them at the weekend as i like to say and we help pay the school fees but then we go back to work for a chemical company or fossil fuel company and act as if we mean to kill them all it's a strange nature except it's the same as every other species on the planet grab what you can live for today and here we are doing the same it's something we all recognize is a bigger danger than anything we've ever met before living with a with a another intelligence yeah that is going to be inevitably much more than we are and we are left worrying about pees when when the survival of our when our strangely our climate is going to hell not as we used to think in 20 30 40 years but now Our baby production is going to hell, not as we used to think in 50 years or 100 years, but now. China is producing fertility rate of one, a baby production that every 30 years has, that in 90 years is an eighth. And Korea is a third of its baby production each 30 years, a ninth in 60 years, a 27th, i.e. they're out of business in a single lifetime unless it changes. and it has been changing, but it's been changing steadily for the worst. 65% of all countries are below replacement and quite a few, like China, are way, way down towards one. And nobody cares. We are not programmed to worry about long-term slow burning problems and they're all coming to bear together and they're compounded by our ignorance or lack of concern about the risks of AI. This was going to be a very exciting time. [60:00] exciting slash terrifying also if we need five planets i think you just made the ball case for spacex i was just gonna make the same joke oh my god yeah i shouldn't have let you i shouldn't have let you go first there's another way to resolve that though and that is to have a billion people and not eight or ten or twelve and the interesting thing is if you asked a society to please have fewer children. When they wanted more, you wouldn't have a prayer unless you used force. But we are going to have a dramatic sustained drop in our population by sheer luck. We are the first generation in history who are deciding to have, for perfectly good reasons, a hundred good reasons, we've decided to have fewer children. And if we keep doing this, we go out of business. It's quite simple. If you don't have 2.1 healthy, well-educated children, you're on your way out. And almost nobody does in the developed world. And even in sub-Saharan Africa, baby production is falling like a stone It just falling from a very high level falling from seven babies per mother to four They have lost more babies over the last 50 years than Europe has It just that they lost them from a much higher level All right, we're going to have to leave it there. But Jeremy Grantham, thank you so much for coming on All Thoughts. Well, thank you for allowing me at least two minutes to talk about serious stuff. Of course, something other than P.E. ratios. Yeah, anytime. And you're always welcome back. Thank you so much for coming on the podcast. Thank you. Joe, I love talking about historical bubbles. Yeah, me too. And I guess I should shout out some of our really old episodes at this point on very esoteric bubbles. Yeah. Like the Florida land bubble. No, they're always fun. the catfish bubble i did think the point about the change in the mag 7 stocks yeah this sort of watershed moment this idea of a cage fight yeah on the white house lawn and i guess a change in corporate strategy where everyone is really tackling the same area of business yeah that was interesting no it super interesting like 10 years ago you could draw a very clear line between what Google business was and say Meta business right You can do that the same degree when both of them Meta they not right at the edge but they trying to they want to be in the game as like a model maker They're also spending, both spending enormous amounts of money on capital expenditures and so forth. Like they really are like no longer the sort of dominance of their verticals. verticals but i have to say like his conversation there at the end and he's like well will the robots be our butlers on the beach or will they accidentally kill us or will they purposely kill us or will humanity extinct ourselves because we're stopping having babies like yes it sort of does make you like why are we wasting time talking about p.e ratios like when these are like the big questions that we're like right up against yeah like why are we talking about p.e ratios ever How do I prep my portfolio for robot extinction? Very important. Oh, it's like, is Kevin Warsh going to cut at his first interest rate? Like, that will not very likely be a particularly important question or moment 10 years from now. Like, that's probably not what's anything will hinge on that in the grand scheme of things. That true But again this sort of goes back to the big tech argument But if you couch everything in existential terms then you can justify anything Totally Right Which is what we seeing right now in big tech I realize I just naturally went from talking about the extinction of humanity back to big tech valuation. So, yeah, I apologize for that. You know, it's interesting. You know, we did that episode about the history of rope recently. and you know in that book he and he made the point on the podcast that you know we went for about a million years maybe not i don't know if it's humans but maybe right before humans well there was literally one invention and that was the hand axe and then you think about like in the last few years alone between chad gpt and glp ones and evs etc and then jeremy was making that point like history was a little bit boring you know the 1950s like oh coke opened the new factory and that was news. And what that means is literally, and other people have said this, time is speeding up. There are just more events per day happening. I'm tired, Joe. I'm tired. I'm tired. But it's good for the news business. It's not so good for our producers. All right. On that note, shall we leave it there? Let's leave it there. All right. This has been another episode of the All Thoughts Podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway. And I'm Joe Weisenthal. You can follow me at The Stalwart. You can follow our producers, our tireless [65:00] Carmen Rodriguez at Carmen Armand Dashiell Bennett at Dashbot, Kale Brooks at Kale Brooks, and Kevin Lozano at Kevin Lloyd Lozano. And for more OddLots content, go to Bloomberg.com slash OddLots for the daily newsletter and all of our episodes. And you can chat about all these topics 24-7 in our Discord, discord.gg slash OddLots. And if you enjoy OddLots, if you like it when we talk about P.E. ratios alongside the extinction of humanity, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad-free. All you need to do is find the Bloomberg channel for Apple Podcasts and follow the instructions there. Thanks for listening. Thank you 360 degree guarantee you can be 4imprint certain your order will show up just right right on time Explore more at 4imprint.com. 4imprint. 4certain. When you're running a business, the best days are the ones where priorities stay on track. For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges. At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering, and claims experience. Learn more at thehartford.com slash risk mitigation. Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. These days, it seems like AI agents are just about everywhere you turn, every field and every function. But without identity, you can't trust they'll serve your business instead of jeopardizing it. Fortunately, Okta helps you get identity right by securing your AI agent's identities, giving you a single layer of control, a single standard of trust. So whether an AI agent supports a single user or your entire enterprise, with Okta, you'll turn risk into opportunity, Secure every agent. Secure any agent. Okta secures AI.