Real Yields Push 30-Year Treasury Yields to 20-Year Highs

实际收益率推升30年期美债收益率创20年新高 推动长期美债收益率走高的是实际利率上升而非通胀预期；若10年期美债收益率触及5%，股票估值将面临威胁。 Josh Brown 与 Nick Colas 深入探讨了为何激增的实际利率（而非上升的通胀预期）正在推高长期美债收益率，并从根本上重塑资产配置策略。他们分析了超大市值科技公司庞大的 AI 资本支出如何直接与国债争夺资金，并使科技公司的资产负债表转向固定收益风险。最后，Colas 概述了 DataTrek 的标普 500 指数公允价值模型矩阵（S&P 500 Fair Value Grid），详细阐述了三种前瞻性市场情景、实现 AGI 的战略竞赛，以及美中 AI 主导权的地缘政治利害关系。 像 SGVA ETF 这样的固定收益创新通过消除每月应税现金分红来实现内部持续复利，解决了传统超短期国债效率低下的问题。 (0:13) 30年期名义美债收益率的上涨完全是由实际利率的飙升（突破2.5%升至3.0%）所推动，而长期通胀预期依然牢牢锚定。 (2:59) 持有长期美债（TLT）已从2010年代可靠的对冲工具（年复合增长率 +7.8%）转变为2020年代投资组合的持续拖累（年复合增长率 -4.4%）。 (2:59) 实际利率走高由四大关键因素驱动：美联储量化宽松的缩减、更高的中性利率（$R^*$）、巨额联邦财政赤字，以及科技巨头发债用于 AI 资本支出。 (7:30) 超大市值科技巨头大规模发行企业债直接与国债拍卖竞争，分流了对国债的需求并推高了收益率。 (11:10)

Nick Colas and Josh Brown argue that rising real yields, rather than inflation expectations, are driving long-term Treasury rates higher and reshaping market risks as mega-cap tech companies increasingly fund massive artificial intelligence projects with debt. Surging Real Yields Reshape the Bond Market (2:59) Long-term Treasury yields have climbed toward 15- to 20-year highs, with nominal 30-year yields reaching around 5.2% to 5.3%. According to DataTrek co-founder Nick Colas, this rise is driven entirely by surging real interest rates—now pushing toward 3.0% from negative levels—while 30-year inflation expectations have stayed anchored between 1.5% and 2.5% since 2010. This dynamic has severely penalized long-duration assets: the iShares 20+ Year Treasury Bond ETF (TLT), which delivered a +7.8% annualized compound return in the 2010s, has fallen to a -4.4% annualized return across the 2020s, upending its historical role as a dependable portfolio hedge. Structural Drivers and Corporate Debt Competition (7:30) Colas identifies four catalysts pushing real rates higher: the Federal Reserve winding down quantitative easing, an economy resilient enough to suggest a higher neutral rate ($R^*$), expanding federal borrowing, and heightened corporate debt issuance. Year-to-date corporate debt issuance is on track for $1.75 trillion, running 20% to 30% higher than the previous year. Tech giants like Alphabet, Amazon, Meta, and Microsoft are tapping the debt market rather than diluting equity to fund AI infrastructure, competing directly with sovereign debt for institutional capital. In response, Colas advises allocators to keep fixed income duration under five years until signs of economic weakness cap long-end yields. Equity Thresholds and Valuation Multiples (15:23) Colas highlights 5.0% on the 10-year Treasury as the historical threshold where borrowing costs begin to squeeze equity valuations and corporate profits, while Brown notes that the speed of rate changes matters as much as the nominal level. Unlike the 1990s, when yields were falling from 17% down to 5%, yields today are climbing off zero. Despite this headwind, the S&P 500 gained approximately 12.1% year-to-date through mid-August 2026 entirely through fundamental earnings revisions—up 15% for the current year and 13% for the next—with zero contribution from price-to-earnings multiple expansion. Colas outlines three 12-month upside scenarios: a base case (+6% to +16% at a flat 20x multiple), a better case (+13% to +28% at 22x if energy costs fall), and an extreme bull case (+12% to +39% at 24x if enterprise AI revenue accelerates outside software coding. Public Equities as High-Stakes Venture Capital (224:47) Corporate capital allocation has shifted dramatically from returning cash via buybacks to funding an all-out race for Artificial General Intelligence (AGI). Colas and Brown note that hyperscaler spending is driven by the conviction that an early winner in AGI will build an insurmountable moat, noting AI benchmark advances are outstripping Moore's Law. However, this transition turns public tech giants into speculative capital vehicles. With tech representing roughly 38% of the MSCI ACWI and 30% of the S&P 500, markets are heavily exposed to whether enterprise AI generates enough cash to service newly issued debt. Furthermore, sector correlations have fallen to decade-plus lows, signaling that broad equity pricing relies on the implicit assumption that the U.S. economy faces zero risk of recession over the next five years. The central unresolved challenge is whether enterprise cash flows from artificial intelligence will materialize fast enough to validate record capital expenditures and cover rising debt service costs, especially as the U.S. and China vie for technological supremacy.

DataTrek联合创始人Nick Colas与Josh Brown指出，推高长期美债收益率的核心动力是实际利率的急剧飙升而非通胀预期，这受到美联储量化宽松退出、历史性财政赤字以及科技巨头为AI资本开支大举发债的多重驱动。与此同时，美股在完全依赖企业盈利上调支撑的背景下，正面临国债收益率逼近5%关键门槛以及大型科技股由“现金返还机器”向“高杠杆风投模式”转型的深层估值考验。 实际利率飙升主导长端国债收益率走势 (2:59) Nick Colas分析指出，30年期美债名义收益率近期攀升至5.2%至5.3%，逼近15至20年高点。根据名义收益率由通胀预期与实际利率构成的框架，自2010年以来，长端通胀预期（以30年期平准通胀率衡量）始终稳定锚定在1.5%至2.5%之间，因此推高名义收益率的全部动力来自实际利率——后者已从深度负值反弹逼近3.0%。这一转变严重冲击了长久期债券，iShares 20+年期国债ETF（TLT）的年化复合回报率从2010年代的+7.8%骤降至2020年代的-4.4%，使长端美债从传统的危机避险工具沦为投资组合的拖累。 Colas总结了推升实际利率的四大结构性支柱：美联储停止主动购债并推进资产负债表正常化；美国经济在利率急升与油价冲击下表现出异常韧性，印证了更高的中性利率（R*）；联邦借贷规模激增导致国债供给压力空前；以及科技巨头大规模发债对国债形成的直接资金竞争。基于此，Colas建议财富管理者将固定收益久期控制在5年以内，直到出现经济明显放缓的信号来压制长端实际利率。 科技巨头举债AI基础设施挤压国债需求 (11:10) Josh Brown与Nick Colas指出，2026年迄今美国企业债发行规模预计达到1.75万亿美元，同比大幅增长20%至30%。Alphabet、Amazon、Meta和Microsoft等科技巨头为了避免股权稀释，选择大量进入债务市场为人工智能基础设施融资。这种举债行为引发了与主权债务的直接竞争：对于机构买家而言，超大型科技公司拥有全球性现金流和极高的资产负债表安全性，同时还提供高于国债的信用利差，这使得固定收益资金从国债竞标分流至高评级企业债，直接压低国债价格并推高基准收益率。 Colas进一步强调，大型科技公司的资产负债表与资本配置模式已发生深刻质变。这些企业过去以30%的高净资产收益率（ROE）运行，并通过大规模股票回购和分红返还现金；如今它们转变为债务发行主体，承受债务展期风险，并将数十亿美元投入AI基础设施建设。Brown补充指出，许多高杠杆AI项目甚至通过私募信贷和私募股权的非流动性结构运作，并由超大规模云厂商提供担保，这为金融系统埋下了比2022年周期更复杂的再投资与信用风险。 5%收益率警戒线与盈利驱动的估值模型 (15:23) 针对收益率上升对美股的影响，Colas指出10年期美债收益率达到5.0%是股市的关键“震动点”（jitter point），此时借贷成本将显著压制经济扩张与企业利润。Brown则强调上升速度与绝对水平同样关键：渐进上升可被市场消化，而剧烈跳升将引发恐慌。Colas对比了历史背景，反驳了“1990年代5%收益率仍伴随大牛市”的论调，指出1990年代的5%是从保罗·沃尔克时代的近17%下行而来，构成了长达40年的资本宽松顺风，而当前环境则是收益率从0%向5%或6%逆向攀升。 在估值方面，Colas展示了DataTrek的标普500合理估值模型。截至2026年8月中旬，标普500指数年内约12.1%的涨幅完全由企业盈利预期上调驱动（当前财年上调15%，下一财年上调13%），市盈率（P/E）估值倍数反而略有收缩。针对未来12个月，Colas提出了三种情景：基准情景下，盈利持续走高且P/E维持在20倍，指数有6%至16%的上涨空间；较好情景下，地缘冲突缓解令油价降至60至65美元，推动P/E升至22倍，带来13%至28%的涨幅；最乐观情景下，AI资本开支转化为切实的企业非编程收入并推升P/E至24倍，对应12%至39%的涨幅。基于戈登增长模型，即便5%收益率抬高无风险折现率，只要企业盈利增长能够超越利率增幅，估值倍数依然具备扩张可能。 AGI军备竞赛与公共股票的风投化风险 (224:47) Colas指出，科技巨头数千亿美元的AI支出并非为了订票或基础代码自动化等渐进功能，而是为了争夺首先实现通用人工智能（AGI）的主导权。Artificialanalysis.ai数据显示，自2022年11月ChatGPT发布以来，其智能指标提升了1300%，Anthropic的Claude在过去一年内也跃升了76%，技术演进速度大幅超越摩尔定律约41%的年化复合增长率。Brown补充指出，AGI赛道具有“赢家通吃”属性，最先达成自我迭代的平台将建立无法逾越的护城河，这也是资本市场相对忽视中国开源模型的原因。 这一竞争促使成熟科技股演变为公募风险投资。科技行业已占MSCI ACWI全球指数的38%以及标普500的近30%，意味着全球近四成权益资本正在押注于高风险前沿科技。与此同时，标普500各板块相关性跌至十余年低点，市场定价隐含了未来五年美国经济“零衰退”的极高共识。从地缘政治角度看，全球市值前20大企业中有16家位于美国，Colas与Brown强调，维持AI主导地位关系到美国的全球资本吸引力与美元储备货币地位；若中国科技企业在该领域胜出，将深刻改写全球贸易与资本流动格局。 本期节目的核心分歧在于，科技巨头由高现金流回购模式向大举借债推进AI资本开支的转型，能否在宏观高利率环境下如期转化为充沛的非编程领域现金流。虽然美股凭借超预期的盈利修正抵御了实际利率上行的冲击，但长端收益率高企与美国五年“零衰退”预期构成了极为脆弱的估值前提，一旦未来AI变现放缓或经济出现衰退裂痕，缺乏估值容错空间的公开市场或将面临剧烈重定价。

## Background 本期内容由 YouTube 财经频道 The Compound 于 2026 年 8 月 24 日发布。视频节目《What Did We Learn》由 Josh Brown 主持，特邀嘉宾为 DataTrek Research 联合创始人 Nick Colas，主要探讨长端美债收益率攀升背后的真实利率驱动因素、对美股估值与科技巨头债务风险的影响，以及标普 500 指数的公允价值矩阵与未来走势。 ## TL;DR 节目深入拆解了长端美债收益率走高的底层机制，指出此轮飙升并非源于通胀预期失控，而是由于量化宽松逆转、经济高韧性推升中性利率、财政赤字庞大以及科技巨头 AI 融资发债对国债配置产生挤出效应，导致真实利率逼近 3.0%。科技巨头从以往低负债的股票回购机器转变为激进借债的重资产实体，削弱了其面对高利率环境的免死金牌，令 10 年期美债收益率 5% 成为严峻的估值警戒线。结合 DataTrek 的标普 500 公允价值矩阵，当前美股上涨完全由盈利上修驱动，若要消化 5% 以上的高利率并实现估值扩张，必须依赖油价回落、地缘局势缓和以及 AI 技术在实体产业中兑现超出无风险利率的高盈利增速。 ## Atmosphere & style 对谈整体洋溢着机构投研级别的严密逻辑与专业张力，Nick Colas 凭借三十余年的买方与卖方经验条分缕析，Josh Brown 则以幽默敏锐的提问不断将复杂的宏观数学模型具象化为通俗的市场隐喻。对话节奏紧凑扎实，既直面债务展期与科技股重资产化转型潜藏的冷酷现实，又在戈登模型与 AGI 终局博弈的宏大图景中勾勒出对科技生产力的坚定远见。整场讨论既有防范久期风险的审慎戒备，又洋溢着穿透市场迷雾的理性快感。 ## Bulletin - [03:48](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=228s) 30 年期美债名义收益率走高并非由通胀预期失控引起，通胀预期长期锚定在 1.5% 至 2.5%，核心推手是真实利率逼近 3.0%，导致长期国债 ETF（TLT）在 2020 年代以来的年化复合回报率暴跌至 -4.4%。 - [10:26](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=626s) 年初至今美国企业债发行规模达 1.75 万亿美元，Meta、亚马逊及 Alphabet 等科技巨头为资助 AI 资本开支大举长端发债，顶级信用评级的高流动性企业债直接吸走了以往配置长期国债的资金。 - [26:54](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1614s) 标普 500 指数年内约 12% 的上涨 100% 由企业盈利上修驱动，而非估值倍数扩张；但当前 20 倍以上的市盈率容错空间极窄，10 年期美债收益率逼近 5% 的心理警戒线是压制市场风险偏好的核心阻力。 - [38:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2310s) 公募股票投资正呈现“VC 风险投资化”倾向，科技巨头停止将现金流大比例回馈股东，转而豪赌通用人工智能（AGI）；MSCI 全球指数中大科技权重已近 40%，促使等权重指数成为降低单一主题杠杆暴露的对冲工具。 ## Key takeaways - [03:48](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=228s) 名义国债收益率的暴涨根源在于真实利率的结构性逆转，而非长期通胀预期的失锚。 - [03:48](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=228s) 超长端国债在过去 6 年不仅未能发挥经济重创时的对冲避险属性，反而带来了年化 -4.4% 的资本净损失。 - [07:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=450s) 美联储量化宽松结束、经济韧性推手中性利率上移、财政赤字高企以及科技企业长债扩容，是推高真实利率的四大动因。 - [07:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=450s) 在宏观经济出现实质性衰退放缓信号前，债券投资组合久期应采取防御姿态控制在 5 年以内。 - [10:26](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=626s) 科技巨头天量发债抢占高评级固收买盘，高现金流科技企业债正在挤占美国国债的市场需求。 - [12:20](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=740s) 市场对无风险资产的现实认知发生分化，高现金流全球科技公司的偿债信用品质正在与主权美债相提并论。 - [15:20](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=920s) 10 年期美债收益率 5% 是股票估值的警戒分水岭，即便升速温和有序，过高的绝对水平仍会动摇 20 倍以上的市盈率基础。 - [18:28](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1108s) 相较于 2022 年凭借低负债与削减开支自救，如今科技巨头在 AI 军备竞赛下大举借债，利率风险与再融资展期风险重返资产负债表。 - [20:56](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1256s) 巨头资本分配从提供稳定 30% ROI 回报与股东回购，转变为高杠杆承担未知的 AI 基础设施再投资风险。 - [24:25](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1465s) 1990 年代的 5% 利率环境得益于从 15% 峰值下行的宽松顺风，而当前利率是从 0% 暴拉至 5% 的收紧逆风，二者不可同日而语。 - [26:54](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1614s) 标普 500 今明两年的盈利预期分别被上调 15% 和 13%，全年的指数涨幅罕见地全额由基本面盈利上修所支撑。 - [29:00](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1740s) 在基准情景下维持 20 倍估值倍数不变，依靠盈利水龙头持续释放，标普 500 未来 12 个月仍有 6% 至 16% 的上行空间。 - [30:59](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1859s) 若原油价格大幅回落并伴随地缘政治缓和，估值倍数有望扩张至 22 倍，推动股市实现 13% 至 28% 的戴维斯双击。 - [32:16](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1936s) 根据戈登增长模型，若下游客户在广泛经济领域兑现 AI 降本增效，核心成分股高双位数盈利增速可抵消 5% 至 6% 高折现率的估值压制。 - [34:47](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2087s) 科技巨头敢于豪赌 AGI 军备竞赛并在板块间保持极低相关性，其底层建立在“未来五年绝无传统衰退”的宏观假设之上。 - [38:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2310s) MSCI ACWI 中大科技权重高达 38% 至 40%，二级公募资金被迫承受早期 VC 级风险，等权重标普 500 成为分散集中度风险的工具。 - [38:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2310s) 美元储备货币地位与高税基依赖于技术创新的全球主导权，在中美 AI 争霸中美国不能承受失去技术霸权与创新垄断的代价。 - [41:07](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2467s) 西方资本主义体系对巨额利润积累与创始人财富的激励机制，是其科技创新在商业回报维度优于竞争对手的制度根基。 ## Detailed walkthrough ### [00:13](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=13s) 赞助商信息与节目开场 本期节目由 F/m Investments 及旗下的 F/m Accumulator Ultrashort Treasury ETF（代码：SGVA）赞助。Josh Brown 首先祝贺 Alex 及 F/m 团队与 T. Rowe Price 达成合作，迈入全新发展阶段。针对超短期国债 ETF 的市场痛点，传统产品往往每月向投资者强制派发不需要的现金分红，导致投资者必须承担当期应税收入（taxable income），并在扣税后手动进行再投资。而 SGVA 的产品设计旨在规避这种不必要的应税分红，通过在 ETF 内部实现资金自动复利（compounding inside the ETF），让投资者能够一直持有超短期国债资产，并仅在自身实际需要现金时自主决定赎回，实现“为增长而建，而非为分红而设（built to grow, not distribute）”。 在节目开场中，Josh Brown 欢迎观众来到全新一期的《What Did We Learn》。随着美股财报季大部分落幕，本期将聚焦过去一周对所有投资者、资产配置机构、投资组合经理以及对冲基金而言最重要的宏观核心议题。DataTrek Research 联合创始人兼《DataTrek 每日晨报》（面向 1500 多家机构与零售客户）的主笔 Nick Colas 担任本期嘉宾（另一位联合创始人 Jessica Rabe 因身体抱恙暂未出席）。两人简短寒暄了夏末生活后，迅速切入正题。Josh 指出，股票市场参与者通常不太关心债市动态，甚至习惯认为两者走势各自独立；但眼下国债市场重新杀回市场关注的核心舞台（front burner），长端国债收益率的剧烈波动及其背后的多重叙事，已成为牵动整个资本市场神经的头等大事。 ### [03:48](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=228s) 30 年期国债收益率拆解：真实利率飙升而非通胀预期 长端美债收益率近期持续走高，尤其是 30 年期美债收益率已突破 15 至 20 年来的高位，市场对此成因充满困惑。Nick Colas 调出历史数据，对 30 年期国债收益率进行了关键的因子拆解：名义利率（如票面的 5.2%~5.3%）实际上等于“投资者对未来 30 年的年化通胀预期”加上剩余的“真实利率（real interest rates / real yields）”。 ```mermaid flowchart LR accTitle: 30年期名义国债收益率结构拆解 accDescr: 展示名义利率如何由长期稳定的通胀预期与近期大幅飙升的真实利率两部分构成。 subgraph Decomp ["名义收益率构成 (Nominal Yield ~5.2%-5.3%)"] A["长期通胀预期<br>(1.5% - 2.5%，长期极度平稳)"] B["真实利率 / 实际收益率<br>(突破上行逼近 3.0%)"] end A --> C["驱动力：近乎零贡献"] B --> D["核心驱动力：推动名义利率创新高"] ``` 数据显示，自 2010 年以来的 15、16 年间，红线代表的通胀预期基本处于 1.5% 到 2.5% 的极窄区间内波动，走势异常平稳，说明市场对于长期通胀的基本锚定从未失控。真正发生剧烈波动并促成这轮收益率暴涨的，是扣除通胀预期后的“真实利率”——它从过去典型的 2.0% 左右甚至更低水平，一路突破上行并逼近 3.0%。真实利率的结构性逆转对资产配置造成了毁灭性冲击： - **iShares 20 年期以上美国国债 ETF (TLT) 在 2010 年代表现**：年化复合增长率（CAGR）达到接近 8%（约 7.8%），作为信用层面上近乎无风险的资产，即便承担久期风险（duration risk），也成为了绝大多数机构和理财师资产配置组合中的核心定海神针。 - **TLT 在 2020 年代迄今表现**：随着真实利率从负值区间暴力拉升至极高正值区间，在将票息全部再投资的前提下，TLT 过去约 6 年间的年化复合收益率暴跌至 **-4.4%**。 - **对冲属性的失效与再思考**：投资者原本将长端国债视为“极度风险厌恶资产（risk-off asset）”，预期在股市崩盘或经济陷入重度衰退时充当强力对冲（如 2020 年疫情暴跌初期的亮眼表现）。然而，在当前通胀和利率结构下，这项避险工具反而带来了持续的资本侵蚀，迫使资金分配者开始全面反思其作为长期固定底仓的合理性。 ### [07:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=450s) 推动真实利率攀升的四大诱因与久期策略 针对真实利率为何居高不下且持续攀升，Nick Colas 归纳了四大核心驱动因素： 1. **美联储量化宽松（QE）周期的全面逆转**：在 2010 年代初以及 2020 年疫情期间，美联储通过大规模购债人为压低真实利率至深度负值以刺激投资，达成了政策目标；如今美联储进入量化紧缩与缩表常态化运作，结束了资产购买，过去被人为压制的真实利率出现报复性“解除压制（unwind）”。 2. **经济屡经冲击却无衰退，中性利率（R-Star）结构性上移**：自 2020 年以来，美国经济硬生生抗住了 2022 年激进加息、2025 年贸易政策冲击，以及 2022 年和 2025 年两度原油剧烈震荡等多重外生冲击，却始终未陷入经济衰退。这种强劲韧性证明，美联储在整个商业周期中维持的中性利率水平必然显著高于历史经验值。 3. **联邦预算赤字高企与主权信用担忧**：美国财政赤字规模持续庞大，国债天量发行带来的供应压力，以及市场对主权信用品质的边际担忧，推高了长期限国债所要求的期限溢价。 4. **AI 与科技基建驱动的企业长债发行挤占需求**：市场上涌现出大量与 AI 基础设施、数据中心及能源网络相关的企业长期借贷，众多过去很少发债或长端发债平缓的企业大举发行超长期企业债，直接分流并争夺了以往流入长期国债的市场配置资金。 基于上述判断，Nick 建议客户在投资组合的债券久期管理上采取防御姿态：在观察到美国宏观经济显著放缓之前，必须将债券久期压在 5 年以内（“keep portfolio durations pretty short, like under five years”）。Josh 对此打趣总结道，目前理财师配置固收资产时，策略应当是“造一个矮矮的垫脚凳（step stool），而不是搭一个架向 30 年期的长梯子”。唯有在经济数据明显走弱、外生冲击真正可能演化为系统性衰退事件时，拉长久期作为避险对冲的逻辑才重新成立；而在当前经济韧性十足的背景下，盲目抄底超长债依然风险极高。 ### [10:26](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=626s) 赤字扩张与 AI 发债潮：企业借贷正在吸走国债资金 经济的韧性意味着中性利率（neutral rate）理应更高，这是长期美债收益率走高的基础逻辑之一。与此同时，美国联邦财政赤字的失控进一步动摇了投资者的心理防线——年内迄今的赤字规模已经超过了上一全年的总和。尽管美债在名义上依然安全，但投资者的安全感已显著减弱。更关键的边际变量在于第四点：与人工智能（AI）资本开支相关的巨额长期企业发债，正在剧烈分流国债的需求。 年初至今，美国企业债的发行总额已达到约 1.75 万亿美元，比去年同期大幅高出 20% 到 30%。市场上充斥着科技巨头的融资动作，甲骨文（Oracle）、Meta、亚马逊（Amazon）和 Alphabet 纷纷走向资本市场。除了评级偏弱的甲骨文外，绝大多数科技巨头都拥有顶级的信用评级，不仅能以极低的利差借入巨资，而且发债相比股权融资稀释度更低。公开市场的投资者对这些企业债表现出了极度饥渴的认购热情。许多人没有意识到，高评级企业债的火爆正在直接对长期国债形成挤出效应。 - 年初至今企业债发行规模达到 1.75 万亿美元，同比增速达 20%–30%。 - Meta、亚马逊、Alphabet 与甲骨文等 AI 核心玩家均在大规模举债或发售股票。 - 绝大多数超大规模企业评级过硬，借贷利差极薄，吸筹能力极强。 ### [12:20](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=740s) 谷歌债对比美国国债：信用安全与美元底层逻辑的辩论 当把谷歌（Alphabet）的 10 年期债券和 10 年期美国国债摆在面前时，债券投资者的风险定义正在发生深刻变化。学术界虽然仍将国债定义为无风险利率，违约概率极低，但现实中的持仓体验却完全是另一回事。长期美债 ETF（TLT）在过去六年里复合回报率为 -4%，对于厌恶亏损的固收投资者而言，这是难以承受的真实资产缩水。相较之下，美国政府拥有征税权，但谷歌拥有深不见底的全球现金流和极高的企业管理水平。在抗通胀和稳健经营维度，全球化跨国巨头的现金实力完全不输主权政府。 ```mermaid flowchart TD accTitle: AI 企业发债对国债收益率的推升路径 accDescr: 展示科技巨头发债如何分流美债需求并进而推升收益率的过程 subgraph CapitalShift["资本流向分流"] A["AI 巨额资本支出需求"] --> B["科技巨头天量发债<br/>(Alphabet/Amazon/Meta)"] B --> C{"固收机构/投顾选择"} C -->|"追求更高收益与优质现金流"| D["增配高评级企业债"] C -->|"减少配置份额"| E["美债配置需求下滑"] end subgraph MarketImpact["市场传导结果"] E --> F["长期国债买盘减弱"] F --> G["美债收益率走高<br/>(压制全市场估值)"] end ``` 债券投资者的诉求非常简单纯粹，他们并不是在押注谷歌股票能否跑赢纳斯达克，而仅仅是希望"按期收回票息与本金"，并且在此期间不承担剧烈的净值波动。由于谷歌没有陆军、海军和海岸警卫队，不是主权实体，其债券提供的收益率天然会略高于同期国债，这形成了极具诱惑力的性价比。更深层次看，美元成为全球储备货币的核心底气并不仅仅是军队和税赋，而是社会利用资本进行长期创新的能力，谷歌正是资本高效配置的典范。投资经理在资产配置中每投向高等级企业债一美元，就意味着减少购买一美元的美债，这种买方需求的枯竭直接将国债收益率推向更高水平。 ### [15:20](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=920s) 利率上升对股市的阈值：5% 的警戒线与速度之争 对于股票投资者而言，利率上行的真正警报在于何时会开始扼杀实体经济并侵蚀企业盈利。从历史经验来看，10 年期美债收益率达到 5% 一直是市场公认的危险分水岭。一旦 10 年期收益率触及 5%，借贷成本的激增将抑制经济增长，企业盈利预测将迅速恶化，进而引发美股剧烈震荡。目前 10 年期美债收益率徘徊在 4.70% 至 4.72% 附近，虽然尚未破线，但这已是所有股票多头严阵以待的"触发点"（trigger point）。 电视财经评论员目前频繁鼓吹一种安抚市场的论调：只要利率上行是"有序的"（orderly），绝对水平的高低便不足为惧。事实上，整个夏季不仅长端利率在走高，收益率曲线中段（belly of the curve）也在向上漂移，但由于斜率没有出现失控的暴击，市场普遍沉浸在这种平稳的借口中。然而，这种秩序感掩盖了估值压力的本质： - **速度维度的合理性**：温和上升确实给予了权益投资者时间，去逐一评估每家公司承受借贷成本上升的能力，避免出现瞬时流动性挤兑的系统性恐慌。 - **绝对水平的刚性约束**：即便上行过程没有发生失控，当 10 年期美债收益率在 2023 年达到 5% 时，标普 500 指数依然出现了明显抽搐。在 5% 的无风险回报率面前，股票市场很难再心安理得地为标普 500 支付 20 倍以上的市盈率估值。 ### [18:28](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1108s) 2022 年的免死金牌已失效：科技巨头的发债与债务展期风险 回顾 2021 到 2022 年美联储以 70、80 年代以来未见的烈度激进加息时，美股之所以能大难不死，是因为当时拥有独特的"免死金牌"（get out of jail free card）。彼时在各大指数中权重最大的超级科技股基本没有负债需求，它们不仅不依赖信贷，反而在 2020 至 2021 年的零利率盛宴中将大量债务锁定在了极低的固定成本。因此，当 2022 年股价腰斩 30% 至 50% 时，巨头们迅速自我纠错：Meta 成为削减无 ROI 支出的典型标杆，大幅削减成本换取了盈利修复，整个市场根本不用担心标普 500 龙头的利息负担或再融资展期风险（roll risk）。 但今时不同往日，如今科技巨头已经从纯粹的"股票回购机器"逆转为"发债与二级市场增发者"。AI 军备竞赛迫使它们重新大举借债，利率风险正在重回超级大盘股的资产负债表。甲骨文正是最明显的预警信号——它不仅在巨头阵营中股价表现垫底，其债务评级更是险些被直接降至垃圾级。2022 年保护市场的低负债屏障已荡然无存，面对居高不下的长期利率，投资者若继续盲目乐观，将忽视科技巨头潜在的债务展期危机。 ### [20:56](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1256s) 科技巨头的性质嬗变：从资本回报到 AI 资本开支与债务扩张 回顾 2022 年标普 500 指数下跌 19%（创下 1928 年以来表现第 9 或第 10 糟糕的年份），市场经历了大范围的去杠杆与头寸清算。Nick Colas 强调，如今科技巨头的基本面性质已经发生了根本性转变，核心原因有两点：其一是财务杠杆的急剧变化；其二则是资金流向的转移——企业不再将大量资金返还给股东，而是倾注到人工智能（AI）基础设施中。 过去，大型科技公司被视为拥有 30% 稳定高 ROI / ROE、既能满足自身再投资需求又能通过回购和分红回馈股东的优质现金流机器。如今，除了来自债券市场的利率风险，投资者还必须承受巨大的“再投资风险”（Reinvestment Risk）。尽管从技术实践与理论前景来看，AI 具备改变世界的潜力，市场此前也愿意给予信任，但 Nick 坦言自己更担忧 AI 侧的回报能否如期兑现，因为这些巨额投资未来必须承担偿还债务、支付票息以及本金兑付的重任。如今巨头们的资产负债表与业务结构，已与三年前大相径庭。 ```mermaid flowchart LR accTitle: 科技巨头资本分配模式的转变 accDescr: 对比科技巨头从过去的股东回报模式转向当前的AI高额资本开支与举债模式 subgraph Past["过去：低杠杆与高股东回报"] A["高额自由现金流"] --> B["30% 稳健 ROI / ROE"] B --> C["股票回购与股东分红"] end subgraph Present["现在：高资本开支与债务压力"] D["AI 基础设施军备竞赛"] --> E["财务杠杆与发债扩张"] E --> F["再投资风险与债务偿付压力"] end ``` ### [22:16](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1336s) 10 年期美债收益率 5% 警戒线：超大规模企业的周期性软肋 Josh Brown 提出疑问：当前 10 年期美债收益率逼近 5% 的心理关口，是否比 2022 年或 2023 年春季硅谷银行等地区性银行倒闭时更具杀伤力？如今的市场充斥着围绕 AI 单一主题建立的重债项目，涉及大量私募信贷（Private Credit）、私募股权（Private Equity）以他人资金加杠杆运作的非流动性资产，并由超大规模云计算服务商（Hyperscalers）直接提供债务担保与负债支持。 Nick Colas 认为，虽然结构复杂化，但本质风险依然相同：这些科技巨头支撑庞大 AI 投资的底层现金流，依然对宏观经济极度敏感。无论是广告投放还是高端消费电子硬件，一旦经济陷入衰退，这些支出都会被优先削减。尽管美国经济展现出惊人的韧性——承受住了 5% 的利率、2022 年四次加息 75 个基点的冲击以及两次原油价格飙升——但收益率曲线上必然存在一个“经济断裂点”。无论这个临界点是 5% 还是 6%，一旦跨过，汽车销量骤降、房地产销售冻结、消费萎缩以及大规模失业就会接踵而至，而 5% 正是市场情绪开始集体紧绷的分水岭。 ### [24:25](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1465s) 历史对照：90 年代的 5% 利率与当下的本质差异 针对市场上部分人（包括 Josh 的同事 Ben Carlson）提出“1990 年代 10 年期美债平均收益率就是 5%，而那十年美股表现极佳、经济也仅在初期的储贷危机后便持续高歌猛进”的论调，Nick 给出了极具洞察力的历史还原：静态地看平均利率确实是 5%，但当时的 5% 是从保罗·沃尔克（Paul Volcker）在 80 年代抗击通胀时期创下的 15% 乃至 17% 峰值一路下行而来的。 资金成本从极高水平急速下降，为 90 年代经济的再繁荣提供了无与伦比的宏观顺风；而当前的情况完全相反——全球刚从长达数十年的利率下行周期、1% 甚至 0% 的超低利率环境中走出，利率正从 0 冲向 5% 并可能窥探 6%。在资本市场中，“你来自何处”往往与“你身在何处”同样关键。债券收益率虽有零利率下限的阻隔，但向上却缺乏天然天花板；若能明确向市场保证 10 年期美债收益率未来十年止步于 5%，标普 500 的估值倍数可能明天就会直接扩张两个点。 - **1990 年代宏观顺风**：10 年期美债收益率从 15%~17% 峰值回落至 5%，伴随长达十年的降息红利与经济扩张。 - **当前宏观逆风**：基准利率从 0%~1% 快速反弹至 5%，并面临进一步上探的不可测风险。 ### [26:54](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1614s) 标普 500 估值矩阵：100% 由盈利上修驱动的罕见反弹 讨论转向股市估值层面。Nick 介绍了 DataTrek 每两到三周向机构客户发布的“标普 500 指数公允价值矩阵”（Fair Value Grid）。该矩阵基于 FactSet 对今明两年的盈利共识，结合过去十年 14 到 22 倍的标准市盈率区间进行测算，并额外引入了 2000 年互联网泡沫顶峰时的 24 倍倍数以及极其乐观的 26 倍“梦想场景”。从矩阵结果看，若估值回落至 14、16 或 18 倍，即便盈利超预期，指数回报也均显示为亏损；唯有当市盈率维持在 20 倍以上时，投资者才能获得正向回报，这表明当前市场估值已相当昂贵，容错空间极其微小。 然而，截至上周五标普 500 指数年内约 12.1% 的涨幅背后，呈现出极为罕见的健康结构：今明两年的盈利预期分别被上修了 15% 和 13%，这意味着全年的股市上涨 100% 是由盈利上修推动的，估值倍数不仅没有扩张，甚至略有收缩。Nick 指出，在他从事卖方分析师以来的 30 年职业生涯中，分析师通常习惯年初将盈利预期定得很高，随后在年内不断向下修正；而今年完全颠覆了这一惯例，在科技与能源板块的强劲拉动下，盈利预期持续上调，构成了美股本轮行情的坚实底座。 | 市盈率倍数区间 | 历史代表意义 | 对应估值矩阵收益表现 | |:--- |:--- |:--- | | **14x - 18x** | 历史常规均值偏低水平 | 较当前标普 500 现价显著承压，矩阵显示全线亏损 | | **20x - 22x** | 过去 10 年估值区间上限 | 维持当前点位并创造适度正向回报的核心区间 | | **24x** | 2000 年互联网泡沫（Dot-com Bubble）顶点 | 极端乐观扩张，需极强投资者风险偏好支撑 | | **26x** | “梦想场景”（Dream the dream） | 极小概率的极致多头情景 | ### [29:00](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1740s) 未来 12 个月标普 500 的上行路径情景推演 基于当前高估值与高盈利并存的格局，Nick 详细梳理了标普 500 指数在未来 12 个月内的两种主要上行路径： 1. **情景一（基准/最可能发生）：6% 至 16% 的上行空间** 维持现有约 20 倍的市盈率估值不变，无需任何估值扩张。企业强劲的盈利增长势头持续，盈利预期的“水龙头”继续喷涌。市场可能在年底迎来一波反弹，并在下半阶段出现小幅震荡，整体平稳兑现盈利增长带来的涨幅。 2. **情景二（乐观/潜在更好）：13% 至 28% 的上行空间** 企业盈利增长依然坚挺，同时伴随地缘政治层面的破局（如美伊冲突得到化解），促使国际油价走低，从而缓解原油对通胀预期和美联储货币政策的压制。在通胀压力下降与经济风险出清的推动下，市场风险偏好提升，估值倍数有望扩张至 22 倍，实现盈利与估值双击。 ### [30:59](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1859s) 估值扩张路径：油价、地缘政治与极度乐观的牛市情景 Nick Colas 详细梳理了标普 500 指数估值倍数（P/E）向上扩张的先决条件。若想看到估值从目前水平开始实质性抬升，能源成本必须率先回落：原油价格需要回到类似年初的合理区间，比如 60 至 65 美元左右；柴油价格也应从当下每加仑 100 美元以上的水平滑落至 70 到 80 美元。当前估值倍数之所以长期停滞不前，核心原因在于市场既要消化剧烈的石油供给冲击，又要应对美联储换届带来的政策不确定性。如果油价回落、估值倍数扩张至 22 倍，并且企业盈利增速得以确认，那么标普 500 将迎来 13% 至 28% 的上行空间，这属于“较好情景”（better case）。 Nick Colas 坦言，大约在六周前，这种 22 倍估值配合盈利增长的情境其实是他的基准预测（base case）。然而，美伊冲突的持续拉长以及伴随而来的高油价，让他在短期内不再轻言十几个百分点（teens level）的丰厚回报，目前现实的涨幅区间更偏向 6% 到 16%。但如果把想象力推到极致、描绘一个“超级牛市情景”（all-out bullish dream），市场将需要“日月星辰同时对齐”——同时满足三个关键催化剂：盈利持续增长、油价大幅走低，以及科技巨头庞大的 AI 资本支出（capex）开始真正转化为实打实的科技板块利润。若科技投资的回报在未来一年内全面兑现，市场估值倍数有望一举推升至 24 倍，带来 12% 至 39% 的惊人涨幅。 - **估值修复门槛**：原油需回落至 60–65 美元/桶，柴油回落至 70–80 美元。 - **较好情景回报**：22 倍 P/E + 既定盈利增速，对应标普 500 上涨 13%–28%。 - **当前现实区间**：油价与地缘风险未除前，预计处于 6%–16% 涨幅带。 - **超级牛市情景**：24 倍 P/E + AI 投资变现，对应标普 500 上涨 12%–39%。 ### [32:16](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1936s) AI 盈利闭环与戈登增长模型下的利率迷思 Josh Brown 紧接着拆解了 AI 资本支出证明其价值的具体传导链条：今年即将迎来 Anthropic 的首次公开募股（IPO），市场需要看到 Anthropic 明确交出由 AI 驱动的强劲营收答卷；同时 Alphabet 也必须凭借 Gemini 的商业化确认这一势头。这种盈利逻辑不能仅仅停留在底层基础设施提供商（如 OpenAI、Anthropic、谷歌、AWS、微软），更关键的是必须延伸至其下游客户。如果各行各业的企业级买家在财报中持续交出盈利超预期的成绩，并明确将“AI 提效与降本”列为盈利爆发的根源，投资者才会真正意识到当前市场的估值倍数对于这场技术革命而言实际上是“太便宜了”。Nick Colas 指出，今年 AI 的爆发式需求高度集中在软件编程（coding）领域，然而非编程领域才是实体经济的绝对主体，AI 必须向更广泛的经济场景完成价值外溢。 面对 Josh Brown 提出的核心疑虑——在 10 年期美债收益率达到 5% 的高利率环境下，市场能否同时承载 22 倍乃至 24 倍的高估值？Nick Colas 结合商学院最基础的现金流折现与戈登增长模型（Gordon Growth Model）做出了有力反驳。资产定价的核心公式是 $P = C / (R - G)$，其中 $R$ 代表折现率或无风险利率，$G$ 代表增长率。华尔街的分析师往往只盯紧作为分母的利率上升，却忽视了增速 $G$ 的对冲作用：只要企业盈利的边际增速快于无风险利率的边际上行，整体估值就会自然扩张。如果标普 500 核心成分股在未来五年能保持年化 18% 左右的极高盈利增速，哪怕 10 年期美债收益率停留在 5% 甚至 6%，估值倍数依然能够提升，因为盈利增速从低两位数跃升至高两位数带来的扩张效应，完全可以覆盖高利率折现带来的压制。 ```mermaid flowchart TD accTitle: AI盈利闭环与估值扩张逻辑 accDescr: 展示从AI基础设施提供商到下游客户企业盈利爆发，最终驱动市场突破高利率压制实现估值倍数扩张的传导链条。 subgraph Infra["基础设施与模型提供商"] A["Anthropic IPO 与商业化营收"] --> C["底层算力与云生态 (AWS / 微软 / 谷歌)"] B["Alphabet / Gemini 变现确认"] --> C end subgraph Adoption["全行业客户落地渗透"] C --> D["非编程经济领域广泛采用"] D --> E["企业财报盈利超预期<br/>(明确归功于 AI 降本增效)"] end subgraph Valuation["资本市场估值逻辑"] E --> F["盈利高增长 G 显著抬升 (如年化 18%)"] F --> G["G 增长效应压倒折现率 R 上行 (抵消 5%-6% 美债利率)"] G --> H["估值倍数扩张至 22-24x"] end ``` ### [34:47](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2087s) 抢跑 AGI 终局之战与“无衰退”市场假设 在谈及 DataTrek 近期与机构客户的高频交流时，Nick Colas 强调，眼下科技行业天量资本支出的驱动逻辑，根本不是为了解决当下诸如“帮用户订机票”这类细枝末节的 AI Agent 工具，而是一场关乎通用人工智能（AGI）绝对主导权的军备竞赛。根据第三方评测平台 ArtificialAnalysis.ai 的数据，自 2022 年 11 月发布以来，ChatGPT 的智能水平已经暴增了 1300%；在过去一年内，Claude 的性能也大幅提升了 76%。AI 技术的进化速度正以每年 41% 复合增速翻倍的节奏大幅碾压半导体领域的摩尔定律（Moore's Law）。从硅谷顶尖科技播客到初创公司创始人，圈内的全部共识只有一个：谁能率先研发出能够自我迭代学习的 AGI，谁就能形成不可逆转的壁垒与垄断优势。正因如此，开源模型尽管表现抢眼，却未能削弱闭源巨头争夺 AGI 终局的战略定力。一旦 AGI 诞生，其商业变现与霸权地位将不可动摇，市场也会毫不犹豫地为其赋予极高溢价。 与这场长远技术豪赌并存的，是当前美股市场内部极度反常的低相关性现象。DataTrek 的追踪显示，标普 500 指数各板块以及个股之间的相关性已经跌至十年以上的历史低位，与此同时 VIX 波动率指数保持低迷。这种“各走各路”的市场生态源于全市场对“美国经济具备衰退免疫力”的坚定信念：既然未来看不到周期性衰退，资金便可以在医疗、金融、科技之间从容轮动选股。同样，科技巨头管理层之所以敢于通过加杠杆和消耗账面现金来激进投资，正是建立在“未来五年绝无传统衰退”的底层假设之上。在科技领袖眼中，过去 15 年经历的风雨已经证明宏观经济周期风险并不可怕，真正的终极威胁是落后于生存级别的技术范式变革（existential tech innovation）。 - **模型进化速度**：ChatGPT 问世以来智能提升 1300%，Claude 年内提升 76%，远超摩尔定律。 - **核心博弈目标**：不惜一切代价率先实现具备自我教学能力的 AGI。 - **市场相关性指标**：标普 500 板块间相关性触及 10 年以上最低点。 - **底层宏观假设**：科技巨头与美股多头一致认定未来 5 年不会发生经济衰退。 ### [38:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2310s) 公募投资 VC 化与中美技术博弈的终极外溢 美股市场的资金配置模式正在发生根本性异化：传统的公募股票投资（Public Equity Investing）正在全面沦为“风险投资化”（VC-like Investing）。过去，标普 500 头部的大市值蓝筹公司通常会将三分之一甚至一半的自由现金流通过股票回购（buybacks）和现金分红返还给二级市场股东；但如今，这些现金流几乎被全盘挪用，每赚取的一块钱都被投向了充满未知的“科学实验项目”（science project）。在 MSCI 全球指数（MSCI ACWI）中，大科技板块的权重占比已高达 38% 至 40%，意味着全球基准配置中有四成直接与这场 AI 赌局深度绑定。Nick Colas 指出，他和合伙人 Jessica 长期以来并不推崇等权重标普 500（Equal Weight S&P 500），但当下等权重指数仅包含约 14% 的科技敞口，反而成为了投资者降低对巨头资本狂欢过度杠杆暴露的有效避险工具。 面对 Josh Brown 提出的终极疑问——中美之间的 AI 争霸战是否构成了超越日常琐碎基本面的“实质性生存威胁之墙”？Nick Colas 给出了极度笃定的肯定回答。美国之所以能够长期维系全球储备货币地位并承受巨额财政赤字，根本基石在于其能够高效利用资本并引领前沿技术创新，从而不断做大整体税基。在全球市值最高的 20 家领头羊企业中，有 16 家是纯正的美国公司，其余 4 家分布在亚洲与荷兰的企业本质上也是美国科技生态圈的衍生分支。如果美国输掉 AI 军备竞赛，由中国主导的硬软件设备与智能服务将席卷全球，不仅会重构全球贸易网络，更将动摇支撑美国财富积累与全球货币体系的核心逻辑。 ### [41:07](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2467s) 创新驱动、储备货币地位与中美科技巨头的资本回报差异 Nick Colas 深入剖析了储备货币与资本市场底层逻辑的本质联系。他指出，要让全球资本坚信持有美元等储备货币是最明智的选择，核心就在于吸纳全球的资本并进行明智、理性的投资，从而不断推动技术创新。归根结底，是“人类的技术创新”在驱动着股票市场的丰厚回报，并维系着储备货币的至高地位，这正是资本市场最本质的核心。 Josh Brown 随即顺着这一逻辑进行推论，直言在当下的全球竞争中，美国显然希望由自家的 AI 科技巨头去征服全球市场，而不是中国企业。他用极其形象的对比指出：“我们希望最终胜出的是 Alphabet，而不是百度（Baidu）。” Nick 对此表示高度赞同，并从体制与资本回报的角度给出了更深层次的理由。他指出，美国的资本体系会允许这些科技公司赚取巨额利润，也允许其创始人变得更加富有；但过去几年的事实表明，中国显然并不喜欢这种资本积累的基本模式。因此在投资回报层面上，即便百度、阿里巴巴、腾讯或是 DeepSeek 等中国科技力量在技术竞争中胜出，投资者也往往难以从中赚到大钱，因为这些企业的发展和盈利规模会受到体制层面的严格约束。唯有在西方体系下，企业才被允许尽其所能地去追求利润和商业价值。 ```mermaid flowchart TD accTitle: 资本与创新驱动储备货币地位的闭环 accDescr: 展示全球资本流向、创新落地、企业巨额回报与储备货币地位之间的增强循环。 subgraph CapitalEngine["资本与创新引擎"] A["全球资本汇聚"] --> B["明智投资与持续创新"] B --> C["高额股票回报与财富效应"] end subgraph InstitutionalSupport["制度分化与最终结果"] C --> D{"制度环境与利润容忍度"} D -- "西方资本体系：允许自由赚取巨额利润" --> E["维系全球领先地位与储备货币霸权"] D -- "约束资本与财富扩张" --> F["企业盈利与投资者回报受限"] end E -.-> A ``` - **资本市场的核心动力**：Nick 强调人类的持续创新是推动长期股票回报、维持全球资本吸引力与储备货币地位的真正基石。 - **中美巨头盈利机制差异**：在西方市场经济体制下，企业创造超额利润与创始人积累财富是被制度所允许和激励的；相反，无论阿里巴巴、腾讯、百度还是 DeepSeek 取得多大的技术突破，其商业回报和利润空间往往都会受到政策与体系的隐性制约。 ### [42:00](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2520s) 节目收尾与 DataTrek 投研资源指引 随着对宏观、利率与科技地缘竞争探讨的结束，Josh Brown 对 Nick Colas 莅临节目表达了诚挚的感谢。Josh 直言，无论是听众群体还是他个人，每次与 DataTrek 交流都能学到海量高价值的干货洞见。为了让更多寻求专业视角的投资者持续跟踪他们每天发布的分析报告，Josh 特意请 Nick 介绍获取 DataTrek 每日客户报告的渠道。 Nick 介绍称，访问 DataTrek Research 官网（datatrekresearch.com）非常便捷，页面只有一个极其简洁的注册输入框，"无需信用卡，没有任何繁琐套路"（No credit card, no nonsense），即可直接订阅他们的每日研报。Josh 随后热情道别，特意嘱咐 Nick 代向身体欠佳的 Jessica Rabe 转达问候，希望她早日康复并期待他们全团队的回归，同时祝愿 Nick 享受接下来的夏日时光，期待不久后的再次连线。 ## Notable quotes > [00:13](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=13s) built to grow, not distribute — Josh Brown > [07:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=450s) keep portfolio durations pretty short, like under five years — Nick Colas > [07:30](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=450s) 造一个矮矮的垫脚凳（step stool），而不是搭一个架向 30 年期的长梯子 — Josh Brown > [26:54](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=1614s) Dream the dream — Nick Colas > [41:07](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2467s) 我们希望最终胜出的是 Alphabet，而不是百度（Baidu）。 — Josh Brown > [42:00](https://www.youtube.com/watch?v=WaKDHRCC3NE&t=2520s) No credit card, no nonsense — Nick Colas ## Mentioned Josh Brown, Nick Colas, Jessica Rabe, Alex, Ben Carlson, Paul Volcker, F/m Investments, T. Rowe Price, F/m Accumulator Ultrashort Treasury ETF (SGVA), DataTrek Research, iShares 20+ Year Treasury Bond ETF (TLT), FactSet, Alphabet / Google, Meta, Amazon / AWS, Oracle, Microsoft, Anthropic, OpenAI, Baidu, Alibaba, Tencent, DeepSeek, ChatGPT, Claude, Gemini, ArtificialAnalysis.ai, S&P 500, MSCI ACWI, VIX ## Video description On this episode of What Did We Learn, Josh Brown and Nick Colas discuss rising long-term Treasury yields, why real yields not inflation are driving the move, and what higher rates could mean for stocks. Plus, they break down S&P 500 valuations, the earnings revisions powering this year’s gains, three paths to new highs, and more insights from DataTrek’s latest research. This episode is sponsored by F/m Investments and SGVA, the F/m Accumulator Ultrashort Treasury ETF. To learn more about SGVA, visit: https://www.Fminvest.com/SGVA ►00:00 - Ad Read ►01:30 - Intro ►02:59 - Long Term Treasury Yields Just Keep Moving Higher ►26:54 - We Publish an S&P 500 Fair Value Grid Every 2-3 Weeks ►34:48 - Grab Bag of Other Recent DataTrek Client Notes/Discussions 👕 Check out The Compound shop: https://www.idontshop.com 🎙️Listen to our podcasts: TCAF: https://podcasts.thecompoundnews.com/show/TCAF/ Animal Spirits: https://podcasts.thecompoundnews.com/show/animalspirits/ Ask The Compound: https://podcasts.thecompoundnews.com/show/ATC/ The Compound Newsletter: https://www.thecompoundnews.com/subscribe Linkedin: https://www.linkedin.com/company/the-compound-media/ Instagram: https://instagram.com/thecompoundnews Twitter: https://twitter.com/thecompoundnews Tik Tok: https://www.tiktok.com/@thecompoundnews Investing involves the risk of loss. This podcast is for informational purposes only and should not be or regarded as personalized investment advice or relied upon for investment decisions. Josh Brown and Michael Batnick are employees of Ritholtz Wealth Management and may maintain positions in the securities discussed in this video. All opinions expressed by them are solely their own opinion and do not reflect the opinion of Ritholtz Wealth Management. See our disclosures here: https://ritholtzwealth.com/podcast-youtube-disclosures/ The Compound Media, Incorporated, an affiliate of Ritholtz Wealth Management, receives payment from various entities for advertisements in affiliated podcasts, blogs and emails. Inclusion of such advertisements does not constitute or imply endorsement, sponsorship or recommendation thereof, or any affiliation therewith, by the Content Creator or by Ritholtz Wealth Management or any of its employees. For additional advertisement disclaimers see here https://ritholtzwealth.com/advertising-disclaimers. #thecompound #investing #stockmarket #stocks [Source](https://www.youtube.com/watch?v=WaKDHRCC3NE)

[0:13] Josh Brown: built to grow, not distribute [7:30] Nick Colas: keep portfolio durations pretty short, like under five years [7:30] Josh Brown: 造一个矮矮的垫脚凳（step stool），而不是搭一个架向 30 年期的长梯子 [26:54] Nick Colas: Dream the dream [41:07] Josh Brown: 我们希望最终胜出的是 Alphabet，而不是百度（Baidu）。 [42:00] Nick Colas: No credit card, no nonsense

[2:59] Nick Colas: Nominal Treasury Yield = Expected Inflation Rate + Real Interest Rate [11:10] Josh Brown: why would an investor buy a bond issued by Alphabet over a 10-year U.S. Treasury? [29:26] Nick Colas: Valuation = C / (R - G) [29:26] Nick Colas: Good - Most Likely [29:26] Nick Colas: Dream the Dream [17:26] Josh Brown: get out of jail free card [218:37] Nick Colas: science projects [218:37] Josh Brown: wall of worry

[0:13] This podcast is sponsored by FM [0:15] Investments and SGVA, [0:18] the FM Accumulator Ultrashort Treasury [0:22] ETF. A quick congratulations to Alex and [0:25] the FM team on joining forces with Troll [0:28] Price. We're excited to see what this [0:31] next chapter brings for FM and its [0:34] innovative approach to fixed income [0:35] investing. That innovation is [0:37] particularly relevant when you look at [0:39] how ultrashort treasury ETFs are [0:41] typically structured. Most pay out [0:43] monthly cash distributions that [0:45] investors don't need and don't want. [0:48] Those distributions come as taxable [0:50] income that must be reinvested after [0:53] taxes. Here's an ETF to solve that [0:55] problem. The FM Accumulator Ultrashort [0:58] Treasury ETF, ticker SGVA. SGVA is [1:02] structured to avoid unwanted taxable [1:05] distributions and harness the power of [1:07] compounding inside the ETF. Instead of [1:10] monthly distributions, you stay invested [1:12] in ultrashort treasuries and you choose [1:15] when to redeem based on your personal [1:17] cash needs. SGVA, built to grow, not [1:21] distribute. To learn more, visit [1:23] fminvest.com/sgva. [1:30] Well, well, well. Here we are again. How [1:33] we doing, Nick? [1:34] >> Very good. How are you? [1:35] >> Good. Let me introduce the show real [1:37] quick and we'll get we'll get down to [1:38] business. Ladies and gentlemen, welcome [1:40] to an all new edition of What Did We [1:43] Learn? On today's show, we're doing I [1:46] think the most important topic of the [1:49] past week for investors, allocators, [1:52] portfolio managers, hedge funds. This is [1:55] the thing that everyone's looking at now [1:57] that we are most of the way through [1:58] earnings season and we're very [2:01] fortunate. I have my friend Nick Cole is [2:02] here. Uh Nick along with his co-founder [2:05] Jessica Rae, who is uh who is on the [2:08] bench today for for today's episode, not [2:10] feeling great, but Nick is here. Nick is [2:13] the co-founder of Data Trek Research and [2:15] the author of Datrek's Morning Briefing [2:17] Newsletter, which goes out daily to 1500 [2:21] plus institutional and retail clients. [2:24] Nick and Jessica also have their own [2:26] awesome YouTube channel. You can find a [2:28] link to it in the description below. [2:30] Welcome back, Nick. It's so great to see [2:32] you. [2:32] >> Great to see you. [2:34] >> All right. Uh, you having a good summer? [2:35] >> Very good. How about you? [2:37] >> I This is This is a really good one for [2:39] me, too. So, um I uh I I say this every [2:43] year. Um this is the part of the summer [2:46] where it's almost over, but you almost [2:48] say to yourself, "Thank God, I'm having [2:50] too much fun. I actually need to get [2:51] back to work." So, this is like that [2:53] that week for me usually. Um but but [2:56] we're here regardless. So, I want to do [2:58] this uh I want to open this segment by [3:01] just saying um typically we talk about [3:04] the stock market. Typically what happens [3:06] in the bond market doesn't really have [3:08] much of an impact on the stock market or [3:10] the sentiment in the stock market [3:12] because people just understand it's two [3:14] different markets. Sometimes they [3:16] fluctuate for the same reason. Sometimes [3:19] they are doing their own thing [3:20] independent of each other. However, [3:23] right now the treasury bond market has [3:25] come back. Uh it's the front burner of [3:28] the market conversation because we're [3:31] seeing some volatility there. We're [3:33] seeing a lot of narratives being kicked [3:34] around about what's causing it and uh I [3:37] think it's a great way for us to open up [3:38] the conversation today. So Nick, you're [3:40] talking about long-term Treasury yields [3:43] specifically moving higher. So [3:46] >> let us know what we need to know. [3:48] >> Sure. Um I'm focusing right now on the [3:51] 30-year Treasury because that's gotten a [3:53] lot of the press. It has broken out to [3:55] plus 15 maybe 20 year highs in terms of [3:58] yield and there's a lot of confusion of [4:00] what's causing that. So, I pulled [4:01] together a chart that we can start [4:02] talking about, and this shows you [4:05] 30-year Treasury yields decomposed into [4:08] inflation expectations and then the [4:10] residual, which is called real interest [4:12] rates. The nominal rates that you see on [4:14] the screen, the 5.2 5.3% can be broken [4:18] down into how much inflation investors [4:20] expect over the next 30 years per year, [4:22] and then whatever's left over is called [4:24] the real yield. Um the red line which is [4:27] inflation expectations has been dead [4:29] flat pretty much for the last 15 16 [4:32] years. This chart goes back to 2010 and [4:34] it goes between you call it 1 and a half [4:36] and and 2%. It is uh a or 2 and a.5%. It [4:40] it is very stable. The market always [4:43] knows inflation is coming and it's very [4:45] stable. Those expectations are very [4:47] stable. What moves around a ton meaning [4:49] a ton is real rates. what you have left [4:53] over after inflation expectations and [4:56] they're are the ones that have broken [4:57] out. So instead of being call it 2% 2 [5:01] and a.5% they're now pushing up on 3% in [5:05] the current environment and that's the [5:07] reason for rates being higher and it's [5:10] super important if we can just throw [5:11] throw the chart back up for one more [5:12] second. If you were an investor in TLT, [5:15] which is a very popular longdated [5:17] treasury bond ETF out of EyesShares, you [5:20] compounded at almost 8% a year positive [5:23] through the 2010s. It was a very good [5:25] money-making investment. In the 2020s, [5:28] because real rates have exploded from [5:30] being negative to being very positive, [5:33] your kager, your compound annual growth [5:35] rate over this decade so far has been [5:38] negative 4.4%. [5:40] So, this went from being a massive [5:41] money-making trade being long the long [5:43] end of the curve to being just very [5:45] destructive to portfolio returns. And [5:48] let me entirely let me let me just make [5:51] sure everyone understands and just [5:53] double click on that real quick. Um, a [5:56] kagger of 7.8% a year on a portfolio of [6:00] long-term treasuries, which effectively [6:04] from a credit perspective are risk-free. [6:06] the risk that you're taking is duration [6:08] risk because the interest rates [6:10] fluctuate. But like to be able to earn [6:13] 8% a year from 2010 through 2019 [6:17] led to a lot of allocators including TLT [6:22] or something like it as as sort of a [6:25] permanent part of their um asset [6:28] allocation. [6:29] That piece of the puzzle, that piece of [6:31] the allocation in the last 6 years has [6:35] been compounding at -4.4%. [6:38] And that's for the riskoff asset. A lot [6:41] of people thought TLT should always be [6:44] included in allocations because when the [6:47] market's getting killed or the [6:49] economyy's in big trouble, that should [6:51] work really well to the upside as a [6:52] hedge. Y [6:54] >> your hedge your hedge for an overall [6:56] portfolio should lose money when times [6:59] are good and it and it's worked but I [7:02] think a lot of people are now looking at [7:04] that and saying maybe that's not the [7:06] exact hedge that I want going forward. [7:08] >> Yes, precisely so. And by the way, those [7:10] returns are with coupons reinvested. So [7:12] that's your total return. That's just [7:14] sitting there with the asset [7:16] reinvesting. [7:17] >> So it's been tough. And look, that that [7:19] trade even worked in 2020 when [7:20] everything was falling apart in the [7:22] pandemic. TLT was rallying. It was the [7:24] anchor for a lot of portfolios. It just [7:26] has been destructive since. [7:28] >> Yeah. Okay. Go on. [7:30] >> Okay. So, that takes care of the [7:33] picture. Now, let's talk about, you [7:34] know, what's going on with the the next [7:36] uh the next slide. And here we kind of [7:38] break down what is the actual reason for [7:40] real rates being high and climbing. And [7:42] there's, I think, four that are [7:44] important. First of all, they were very [7:45] artificially depressed by the Fed [7:47] purposefully so in the early 2010s and [7:49] 2020s 2020s because of quantitative uh [7:52] easing bond buying by the Fed to bring [7:54] real rates down. The Fed did that and it [7:56] worked fantastically well. They got real [7:58] rates super negative. That encouraged a [8:00] lot of investments. So, mission [8:01] accomplished. [8:02] >> This is the unwind. [8:04] >> This is the unwind. So, now we did the [8:06] unwind and now we have a very stable Fed [8:08] balance sheet with no bond buying. The [8:10] second, and I think a theme that we're [8:12] going to touch on a couple of times in [8:13] this conversation today, is there's been [8:15] no recessions since 2020, despite a [8:18] whole range of shocks. 2022's aggressive [8:20] rate hikes, uh 2025's US trade policy [8:24] shock, two different oil shocks in 22 [8:26] and 25. And the inevitable conclusion [8:28] has to be that the neutral rate of [8:29] interest, the rate at which the Fed [8:31] keeps rates over a cycle, must be higher [8:34] than historically has been the case [8:36] because we've stood up to a bunch of [8:38] shocks with no noticeable impact on the [8:41] on the economy. The third is federal bud [8:43] budget deficits are remain high and [8:45] credit quality is becoming into [8:46] question. I think it's a pretty well [8:47] understood topic. And then finally, a [8:50] new topic is there's a lot of air [8:52] related long-term corporate borrowing [8:54] going on and that's pulling demand away [8:56] from treasuries because all of a sudden [8:58] where companies were kind of flat [9:01] issuing long-term debt, now you have a [9:03] lot of new players in the market issuing [9:05] a lot of long-term debt that is [9:07] competing directly with the Treasury [9:10] bonds. So, the culmination, the [9:12] combination of those factors is why [9:14] we're seeing real rates keep climbing. [9:16] And the conclusion that I've been [9:17] talking about with my clients is you got [9:19] to keep on portfolio durations pretty [9:21] short, like under five years until the [9:23] US economy weakens because that'll begin [9:25] to take away this topic of mutual rates [9:29] have to be higher and that should [9:30] stabilize real rates at the long end of [9:32] the curve. [9:33] >> So, okay. So, financial advisors [9:36] building portfolios with fixed income, [9:38] the answer to those people would be you [9:41] can ladder, but you don't want to be [9:42] laded out to 30 years. you it's it's [9:46] more a step stool than a ladder. [9:47] >> It's it's a step stool. Um the time to [9:51] get longer on duration is when the [9:53] economic data is weakening and the risk [9:56] of one of these shocks turning into an [9:58] event. That's where you might want that [10:00] longerterm duration as your portfolio [10:03] hedge. Um but but you know we don't have [10:06] any signs of that right now. It's not [10:08] that it can't start. It just there's no [10:10] reason to think that it is starting. [10:11] >> No. And we we actually have the counter [10:13] example of all these shocks since 2020 [10:16] and no recession, which in many ways is [10:18] just astounding. It explains a lot of [10:20] things in capital markets, not just [10:21] what's going on in bonds, but it really [10:23] shows up in bonds in a very pronounced [10:25] fashion. [10:26] >> Right. So to sum up, the resilience of [10:28] the economy means the neutral rate [10:30] should be higher. So that's one reason [10:32] for it climbing. And then you've got um [10:36] then you've uh put the put the slide [10:38] right back up. [10:40] Then you've got the federal budget [10:42] deficits. I think uh already year-to [10:45] date uh the deficit is already larger [10:47] than it was for all of 2025. So that's [10:50] heading in the wrong direction. Um and [10:53] that that that pulls people out of uh [10:57] feeling very safe in US bonds. They [11:00] still feel safe but maybe a little less [11:02] safe. Um the artificially depressed part [11:06] is obvious. I want to double click [11:07] though on this fourth point. AI related [11:10] long-term corporate borrowing pulling [11:12] demand away. Chart off, guys. So, [11:16] I think the number is 1.75 [11:18] trillion in corporate bond issuance this [11:21] year. And uh I think we we did a show [11:25] about this last week. I think that's [11:28] running 20 or 30% ahead of where we were [11:31] at this time in 2025. [11:33] >> And we all have seen the headlines. [11:35] We've seen Oracle and Meta and Amazon [11:38] and Alphabet. They're all out in the [11:41] market raising money. Sometimes they're [11:43] doing secondary stock offerings, but [11:45] oftent times they're doing bond [11:47] offerings as well. They should, with the [11:50] exception of Oracle. These are very [11:53] high-grade uh credits. They're able to [11:56] borrow at very low dollar amounts. It's [11:58] less dilutive than doing this through [12:00] equity. And um the the bottom line is [12:03] the the investing public is hungry for [12:06] these deals. The competition with [12:09] treasuries part I think is the leap that [12:11] a lot of people don't make very easily. [12:13] So could you explain why Alphabet [12:16] selling bonds would impact the price of [12:18] a long-term treasury? [12:20] >> Sure. As a bond investor, you got to [12:22] think about risk right now. Treasuries [12:24] have been notionally the risk-free rate [12:26] and still are from an academic [12:27] standpoint and certainly they're money [12:29] good as far as getting repaid. But I [12:31] would put it this way. If I told you I [12:33] was going to sell you either a Google [12:35] 10-year bond or a US 10-year Treasury, [12:38] which do you think will trade better or [12:41] which do you think is more notionally [12:43] secure? Because on the one hand, you [12:45] have yes, the US government ability to [12:47] tax. Great. On the other hand, you have [12:49] Google's cash flows which are profound [12:52] and in some ways Google I think I I [12:55] could argue is at least as safe as the [12:57] US government because it's a global [12:59] business and very well-managed. So as a [13:03] bond investor you're thinking okay what [13:04] is risk? We've just seen TLT compound at [13:07] negative4% over the last six years. [13:10] That's real risk. So, I think bond [13:13] investors look at this more holistically [13:14] than perhaps an equity investor think [13:16] they thinks they probably should, but [13:18] they're just looking for the right [13:19] amount of yield with relatively little [13:20] risk. Remember, bond investors are risk [13:22] averse. The best they can do is get what [13:24] they're promised back again, coupons and [13:26] principle. They're just looking to get [13:28] their coupons and get their principle [13:29] back. That's all they want. And Google's [13:31] just as likely to do that as the US [13:33] government, [13:34] >> right? They're not making a bet that [13:35] Google is going to outperform the NASDAQ [13:38] or that Google's stock will be higher. [13:41] The bet is Google will give me back my [13:44] principal and they're going to pay the [13:46] interest rate [13:47] >> and there won't be insane volatility [13:50] along the way toward that happening. The [13:52] yield is going to be higher than what a [13:54] comparable maturity treasury might [13:57] yield. And that's because Google is not [13:59] sovereign. [14:01] Doesn't have the Marines and the Coast [14:02] Guard and a navy. Uh but it's got other [14:07] attributes that maybe in times like [14:09] these are as or even more attractive um [14:13] than the safety that you feel buying a [14:15] 10-year Treasury. [14:16] >> Exactly. And and we we'll get to this in [14:18] the last section, but I I I've been [14:21] making the argument to clients that the [14:22] reason the US dollar is a reserve [14:25] currency is because of the innovation in [14:27] the US, the long-term innovation in the [14:29] US. That is what makes a reserve [14:31] currency. Not just the army and the navy [14:33] and the taxation, but what a society [14:36] does with its capital. And Google is an [14:37] exemplar of doing something clever with [14:39] your capital. [14:41] >> Right. Okay. So competition from from [14:44] Alphabet and Amazon and these other [14:47] giants selling bonds in the market. So [14:51] every dollar that goes into a high-grade [14:53] corporate bond is a dollar that does not [14:55] go into a US Treasury, [14:58] which is less demand on the buy side, [15:00] which means higher yields. Is that what [15:02] we're saying? [15:03] >> Yes. Essentially, yes. Exactly. It's [15:06] it's a trade-off every PM has to make. [15:08] Do I buy treasuries? You know, if you're [15:09] a pure treasury buyer, yes, of course, [15:11] you're just buying treasuries, but most [15:12] fixed income, most RAAS are looking at [15:15] the whole spectrum of of fixed income [15:18] and saying, where's my best bet? [15:20] >> Okay. So, now the question, so, so where [15:23] the equity market investor then goes to [15:26] is, well, what does this mean for my [15:28] stocks or what does this mean for the [15:30] the the market or the potential for, you [15:34] know, new investment coming into stocks [15:36] versus bonds? like at what point do [15:38] treasuries, you know, at what point do [15:41] higher yielding treasuries represent new [15:44] competition for, for example, dollars [15:47] coming out of a money market fund? [15:49] >> Yeah. I mean, the way I think about it [15:50] is more around the concept of at what [15:53] level do rates begin to threaten the [15:55] economy? Because that's as an equity [15:57] investor, you're worried about earnings. [15:58] You're worried about the volatility of [16:00] earnings, the strength of earnings. And [16:02] so the question is at what level do we [16:04] get rates? And it was for the most of [16:06] this this decade it's been 5% on tens. [16:09] If you get to 5% on tens, there's the [16:11] threat the economy slows down because [16:13] borrowing costs are higher, economic [16:15] growth slows and earnings become touchy. [16:18] Um we're not there yet on tens. We're at [16:19] 470 472 today. And so we're not quite [16:23] there, but I think this is if you want [16:24] sort of a trigger point, look at 10, [16:26] look at 5%. If you get there on 10ens, [16:29] then at least over the 2020s, the equity [16:32] market's begun to really jitter. Uh and [16:34] that's the level I'm looking for. [16:35] >> One of the things we're saying on TV [16:38] which uh everybody seems to like and [16:40] clap at is [16:42] uh as long as the pace [16:46] of like as long as the uh it's it's like [16:50] in other words not about the absolute [16:51] level of the rate. It's the volatility [16:55] of how quickly the rates get to that [16:56] level that's more potentially damaging. [17:00] And so I I think you would agree for the [17:02] most part this summer rates have been [17:04] rising. Now they're rising in the belly [17:06] of the curve. Um not just the long end, [17:10] but it's been orderly. And so people on [17:13] TV keep saying yes, but it's orderly. [17:16] Yes, the rates are higher, but the path [17:18] to higher hasn't been shocking or the [17:22] velocity of the rates rising has been in [17:26] check. like these are the things that [17:27] you're hearing people say and everybody [17:30] likes hearing those things because it's [17:31] reassuring. What do you what do you [17:33] think about that concept that it's not [17:35] just the absolute level of rates, but [17:37] what does the path look like for us to [17:39] get there? [17:40] >> It's a fair point and it's fair because [17:42] if you move slowly, then you have time [17:44] as an equity investor to judge how those [17:46] changes in rates are affecting earnings [17:48] at your companies. If it happens [17:50] quickly, you don't have that [17:51] opportunity. it's just a big shock and [17:53] rates go to five and what do I do and [17:55] what's going to happen to my companies? [17:57] When things go slowly, the market has [17:59] time to judge and assess and look at [18:02] what's going to get hurt, look at what's [18:03] not going to get hurt. But yeah, the [18:06] rate of change does matter. Um, and so [18:08] far, yes, it has been orderly. That [18:11] being said, it was not disorderly when [18:13] it got to when the 10ens got to five and [18:15] 23, I think it was, and the market still [18:17] kind of twitched at that point. So I [18:20] think it is a matter of speed, but I [18:21] also think that there is a level where [18:23] equities say I'm not so comfortable [18:25] paying 20x for the S&P 500. [18:28] >> One of the one of the get out of jail [18:30] free cards from 2022. [18:33] So in 21 they start the rate hiking [18:35] campaign and it's way more aggressive [18:38] than anything our generation has [18:40] experienced. Um you had to have been [18:42] around in the 70s and 80s to have seen [18:45] anything like it. So that's a really [18:47] long time without as aggressive a rate [18:50] hike cycle as what we experienced in 21 [18:53] and 22. But one of the get out of jail [18:56] free cards for the stock market was that [18:59] it just so happened [19:02] the largest companies in the indices [19:05] that had the least amount of interest [19:06] rate risk because they weren't big [19:08] borrowers were the companies that we [19:11] were relying on the earnings coming [19:12] from. And so what they did was they [19:15] looked at their share prices down 30, [19:17] 40, 50%. And they committed to getting [19:21] some religion on spending. Meta is like [19:24] the poster child for this. They got over [19:26] their skis in the bubble in 2020 and [19:29] 2021, funding a lot of projects with no [19:31] ROI. They cut all that out. stock prices [19:35] recovered, but we were never really in [19:37] danger from 5% interest rates or from [19:40] the velocity of that rate rise because [19:43] the largest companies weren't big bond [19:45] issuers didn't have that roll risk. In [19:48] fact, it was the opposite. They [19:50] refinanced so much of their debt at 0% [19:54] in 2020 and 2021 that we weren't even [19:59] thinking about the indebtedness of the [20:01] Fortune 500 companies. [20:04] I think what's changed is we can't say [20:05] that anymore. These companies have [20:08] become they've gone from being buyback [20:11] companies to being companies that sell [20:14] stock on the secondary market and issue [20:16] debt. And some of them do have royal [20:18] risk. And I don't think it's an accident [20:21] that Oracle is the worst performing mega [20:25] cap stock and is also the loan mega cap [20:28] stock with a debt rating that's almost [20:31] been cut to junk. So talk a little bit [20:34] about the difference between now and [20:35] 2022 and maybe why we shouldn't be so [20:38] sanguin about it. [20:40] >> Well, okay. So in 22 tech was the [20:43] hardest hit. Um and it only bottomed [20:45] >> stock prices. [20:47] >> Stock prices. So granted, they had had a [20:49] huge run in 2021. There was a pullback [20:51] coming, but [20:52] >> you know, I think 21 2022 was I want to [20:56] say the ninth or 10th worst year for the [20:59] S&P since 1928. It was up there down 19% [21:02] on the year. It was bad. So there were a [21:04] lot of unwinds going on. Your [21:06] fundamental point, I think, is [21:07] absolutely right. [21:08] >> The nature of these companies has [21:10] changed profoundly for two reasons. The [21:12] first is financial leverage, which [21:14] you're right. It's it's much different [21:16] now. [21:17] And associated with that, you don't have [21:19] money going back to shareholders. You [21:20] have money going into AI. And that's the [21:24] second reason they're different. You now [21:25] no longer have the idea that these are [21:28] steady 30% ROI roe businesses uh that [21:32] can reinvest what they need to reinvest [21:34] and give c a lot of cash back to [21:36] shareholders. You now have a lot of [21:38] reinvestment risk on top of ro risk as [21:41] you call it from the bond market. So you [21:43] have these two simultaneous factors. the [21:46] offset is the technology that they're [21:48] investing in is, you know, both in [21:50] practice and in theory pretty compelling [21:52] and and pretty potentially [21:54] worldchanging. So, the market so far [21:56] I've been willing to give the benefit of [21:57] the doubt. I worry a lot more about the [22:00] AI side because ultimately it's going to [22:02] have to fund those those debt issues, [22:06] fund the coupons, fund the principal [22:07] repayment. But you're right, it kind of [22:10] b, you know, you put a bow on it and [22:12] it's a much different package from what [22:14] we had just three years ago. [22:16] >> Yeah. Um, is that why the 5% trigger [22:20] could really come into play here? Uh, it [22:24] sort of did in in 22 and we did have a [22:27] marketwide pullback and you're right, [22:29] like the tech stock selloff was much [22:32] more acute than the S&P. Then we had [22:35] this this hiccup in 23 with some of the [22:37] banks in the spring. We had a couple of [22:40] banks go under or be uh resolved uh is [22:43] how the regulators put it. Um but like [22:46] this time around now we have heavily [22:49] indebted projects all built around a [22:52] single theme. We've got this interplay [22:54] of private uh credit and private equity [22:58] companies funding some of these projects [23:00] with other people's money. a lot of ill [23:03] liquidity and then a lot of these [23:05] liabilities or guarantees coming [23:07] directly from the hyperscalers. Like [23:09] these are all things that didn't exist [23:11] in 2022 or at least not to the extent [23:14] that they do today. Is that why that 5% [23:17] interest rate trigger on the 10-year [23:19] bears even more uh importance [23:22] potentially than than it did then? [23:25] I think there's a lot of the same [23:27] honestly and the fundamental thing is [23:29] that ultimately the current cash flows [23:31] of the hyperscalers are still [23:33] economically sensitive. It's [23:34] advertising. It's high-end electronics. [23:37] >> It's all the things that get pulled back [23:40] in a recession. And there there we [23:42] haven't found it yet, but there must be [23:44] a breaking point for the US economy on [23:46] the yield curve somewhere. You know, we [23:48] got to five and didn't have a recession. [23:50] We got four 75 basis point rate hikes in [23:53] 22. Didn't get a recession. We've had [23:55] two oil spikes and didn't get a [23:56] recession. So there's an amazing [23:58] durability in the US economy. But there [24:00] has to be a point on the 10ens, maybe [24:03] it's five, maybe it's six, where car [24:06] sales go down a ton, home sales go down [24:08] a ton, people stop spending, you start [24:10] to get layoffs. We haven't seen where it [24:12] is, but five has been a level where the [24:15] market begins to get worried about it. [24:17] And again, you know, the the cash flows, [24:19] underlying cash flows from big tech are [24:21] still cyclical. They're still cyclical [24:23] businesses, [24:23] >> right? Okay. So, we're going to so we're [24:25] going to watch that level. The last [24:27] thing I would say about it, my colleague [24:29] Ben Carlson has written a lot about how [24:32] the average rate for the 10-year in the [24:35] 1990s was 5%. And that was a uh a decade [24:41] with [24:43] no real recession. you had like early [24:46] 90s you had the SNL crisis but like once [24:49] that was over we were pretty we were [24:51] pretty okay. Um the economy grew stock [24:55] market was one of the best stock market [24:56] decades ever. So the level itself [25:00] provided like we we can get um [25:03] accustomed to it I guess is not the [25:06] threatening thing per se. Um it's it's [25:09] about what else is going on in the [25:11] context of that level or or or maybe not [25:14] in your opinion. What do you think? [25:16] >> It is a fair point in terms of static. [25:20] Yes, the average was five. However, it [25:22] was coming down from 15 in the 80s, [25:25] right? It was coming down from the Paul [25:27] Vulker squeeze the economy to get [25:29] inflation down. So I think I want to say [25:32] 10 peaked at 17 and then by the 1990s [25:35] they were five and there was a massive [25:37] rejuvenation of economic growth because [25:39] the cost of money had come down so fast. [25:41] >> It's a big tailwind that we do not have [25:44] >> and now we have sort of the opposite. We [25:46] had a world where you made 8% on 30-year [25:49] treasuries for a decade per year and now [25:53] you're not. And now you used to have 1% [25:55] rates and now you got five going to six. [25:58] And so, you know, the where you're [26:00] coming from is is often times just [26:03] important as where you are when it comes [26:04] to capital markets. We're coming from an [26:06] era of declining rates for the better [26:09] part of 40 years and now on the up [26:11] again. [26:12] >> That's a great right that's a that's a [26:15] great addendum to that idea. Yes, we've [26:18] had decades at 5% where the stock market [26:21] did well, the consumer did well, the [26:23] economy did well. However, that's a [26:26] tailwind from the rates having come down [26:28] from 15 to 5. This is not that. This is [26:32] coming from 0 to 5. And where does it [26:34] stop? We don't know yet, [26:35] >> right? And that's the weird thing about [26:37] bonds. You know, you don't know where [26:38] you you yields probably don't go [26:41] negative like they did in Europe, but [26:42] they have no natural cap. And that's the [26:44] concern. There's no knowing. Look, I [26:46] think if if we could guarantee a [26:48] listener, hey, that the 10's going to [26:50] stop at five for the rest of the decade. [26:52] Multiples would expand by two points [26:53] tomorrow. [26:54] >> Okay, let's do some stock market stuff. [26:56] >> Okay, [26:58] so every couple of weeks we put up a [27:00] grid for our clients to look at current [27:02] S&P valuations. And I'll just pop that [27:04] up on the screen now. It's kind of an [27:05] eye chart, so I'm just going to hit the [27:07] highlights and then we've got a talking [27:08] point chart behind it. What it does is [27:11] look at this year and next year fact set [27:13] earnings consensus and then puts the [27:15] standard multiples on the range over the [27:17] last 10 years has been 14 to call it 22 [27:21] and then we add a 24 multiple because [27:23] that was the peak from the dot bubble [27:25] and we add a 26 just to dream the dream [27:28] and say okay if multiples really expand [27:31] and the upshot is for people listening [27:33] I'll just summarize it most of this grid [27:35] shows losses if you're trading at 14 16 [27:38] 18 times on consensus or even above [27:41] consensus numbers, you're below where [27:43] the S&P is trading today. You only get [27:46] to real payoffs at 20 plus times [27:48] earnings. So, the underlying message [27:51] here is the market's still fairly richly [27:53] valued and has not a lot of room for [27:55] error if earnings miss, if investor [27:57] confidence begins to fade. That's the [27:59] basic message. The other thing that's [28:01] important to know is that the S&P is up [28:03] like 12% year to date. I think 12.1 as [28:05] of Friday. Earnings revisions for this [28:07] year and next year are up 15 and up 13%. [28:11] So the entire move for the S&P this year [28:13] has been earnings revisions. It has been [28:14] no PE expansion at all. As a matter of [28:16] fact, you had a little PE contraction. A [28:18] little bit of that is like the kooky [28:19] oneoff earnings growth at Amazon and [28:21] Alphabet and Q2 because of the um [28:24] markups on SpaceX. But even looking at [28:26] 2027 earnings up, you know, call it I [28:29] say 13%. [28:31] You know, we're we're basically that's [28:33] what we're that's why the S&P is up. [28:34] It's just been very strong earning [28:36] revisions. Let me tell you, I was a cell [28:37] analyst in the 90s. I've watched this [28:39] data for 30 years. Analysts never raise [28:42] numbers during a year. It just never [28:44] happens. They start high and trim. They [28:46] start high and trim. That has not been [28:48] this year. It's been an amazing year for [28:50] earnings growth because of tech and [28:52] because of energy. And that's why the [28:54] S&P has rallied the way it has. It's [28:56] been 100% earnings, which is super [28:58] unusual. [29:00] So now you have a couple of scenarios [29:02] game planned out to show people and this [29:06] is really I think this is really the [29:08] most critical like if you want to have a [29:11] a happy ending to this year which you [29:14] and Jessica have taught me over the [29:16] years still the most likely outcome [29:18] >> but you're showing us now the different [29:21] pathways that this could potentially [29:23] take. Why don't we put that on screen [29:25] while Nick explains? [29:26] >> Okay so three three scenarios for the [29:29] S&P. These are the upside scenarios. The [29:30] downside kind of you can, you know, [29:32] there's a thousand different ways to get [29:33] there. We don't discuss it. I think it's [29:35] well understood. The most likely [29:37] scenario for the S&P up 6 to 16%. And [29:40] that just comes from corporate earnings [29:41] growth remaining strong and estimates [29:43] keep increasing. They've been doing that [29:44] all year. Estimates should continue to [29:47] rise. Multiples stay flat at around 20. [29:49] And you get six to 16% upside again [29:52] based on that grid data that I showed [29:53] you. [29:54] >> I'm sorry, Nick. From here from here six [29:56] to 16% or total [29:58] >> from here over the next 12 months [30:01] >> next 12 months. Okay, great. [30:03] >> So this will take us through uh 27. So [30:06] maybe a rally into year end a little [30:08] sluggishness in the first half of this [30:10] year. [30:10] >> So so no expansion like uh we just [30:13] maintain the multiple we're at but the [30:15] corporate earnings growth that we've [30:17] been experiencing the spigot doesn't [30:19] shut off. It continues. [30:20] >> Exactly. And [30:21] >> I think everybody watching this would be [30:23] thrilled with that. Uh yes considering [30:26] all the worries out there and [30:27] considering we have not seen a penny of [30:28] multiple expansion that's fine. [30:31] >> Okay. [30:33] >> So scenario two possible. This is like [30:34] good better best. This is better 13 to [30:37] 28% upside. We get earnings growth and a [30:39] resolution to the US Iran conflict and [30:41] therefore loyal lower oil prices and [30:44] multiples go to 22. The anchor idea here [30:46] is one of the big overhangs on the [30:49] market is the effect of oil on the [30:51] economy and the effect of oil prices on [30:53] inflation and Fed policy. If we can get [30:56] oil prices to come back down not to 5560 [30:59] like they started the year but let's [31:00] call it 60 65 something reasonable. [31:02] Let's get diesel back down to you know [31:04] 80 70 from 100 plus where it is now. [31:08] Then multiples can begin to increase. [31:09] Multiples have been flat because we've [31:11] had a huge oil shock and we have a new [31:13] Fed share. So if we get multiples to 22 [31:16] and the earnings growth exists, then 13 [31:18] to 28's the upside level. That's your [31:21] better case. [31:22] >> That would be awesome. But that is not [31:24] your that is not that is not your most [31:27] likely. That's just a way that this [31:29] could get even better than it's already [31:30] been. [31:31] >> Yeah. Honestly, this was my base case [31:33] until about six weeks ago. Um, and the [31:36] longer the US Iran conflict goes on, the [31:39] higher oil prices go, the less I'm [31:41] willing to believe that you got, you [31:43] know, it's teen level kind of returns [31:46] available. I think we're in the six to [31:47] 16 band. Now, let's let's let's dream [31:49] the dream. Let's let's go all out [31:51] bullish. Okay, we get our earnings [31:54] growth and we get our loyal prices and [31:56] AI capex begins to show its value in [31:58] tech earnings. This is the other big [32:00] fundamental overhang. When are these [32:02] returns going to show up? If we get them [32:04] in the next year, you can get to a 24 [32:06] multiple and then you're talking 12 to [32:08] 39% upside. Much better. But you need [32:11] all three things to go right. The sun, [32:13] the moon, the stars, they all have to [32:14] align. [32:16] >> Yeah. So, right. You need like you need [32:20] um Anthropics IPO will be this year. [32:23] >> Yep. [32:24] you need them to really come out and be [32:29] talking about like the the revenue [32:32] generation that they're doing from AI [32:34] and then you need Alphabet to confirm [32:36] that with Gemini and then you like you [32:39] need to hear from the providers and then [32:42] you need the companies that are their [32:44] customers simultaneously [32:46] to be showing the results of their own [32:49] AI spending cuz that that's that's who's [32:51] spending with open AI and anthrop ropic [32:54] and Google and uh the AWS environment [32:58] and Microsoft. So you need the customers [33:01] to be putting up continued earnings uh [33:04] surprises and specifically citing AI as [33:07] the source of those surprises and then I [33:10] think people will say the multiple in [33:12] this market is too cheap for the [33:14] revolution that we're in. [33:15] >> Yeah, [33:15] >> but that's exactly [33:17] >> a lot has to go right for that to Okay, [33:19] >> a lot has to go right. Look, AI has [33:20] exploded this year for one reason. It's [33:22] coding. That is why AI demand is where [33:25] it is and that's fine. But the [33:27] non-coding side is the vast majority of [33:30] the economy and it's got to start [33:31] working in that arena as well. Um, [33:35] is there a push and pull between the [33:38] idea of higher rates and a 5% tenure and [33:41] us getting to a 22 or a 24 multiple? It [33:45] seems like both of those things probably [33:46] would would not occur simultaneously, [33:48] although I guess they could. They [33:50] absolutely could. And here's a super [33:53] important point that one of the first [33:55] things they teach you in business school [33:56] about discounting cash flows is it's the [33:59] cash flow that you're getting divided by [34:01] the risk-free rate. You know, the R [34:03] minus a growth rate. The minus the [34:06] growth rate is the super important part [34:08] because the higher the growth rate is, [34:09] the valuation can be higher. It's C over [34:12] R minus G for the nerds in the audience. [34:15] People always forget that the G is super [34:18] important. And if you're growing faster, [34:21] if incremental growth is outpacing [34:23] incremental increases in the risk-free [34:25] rates or the discount rate, then your [34:27] valuation will expand. And that's [34:29] exactly the dynamic here. If you can [34:30] show earnings growth is, let's just pick [34:33] a number, 18% for the next 5 years. I [34:36] don't care if the 10 is at five or 6%. I [34:39] honestly don't because the expansion [34:40] from low teens to high teens more than [34:43] offsets the increase in the discount [34:44] rate from higher risk-free rates. [34:47] Okay, we're going to do a a quick grab [34:50] bag to close things out. Um why don't [34:53] you why don't you take us through these [34:55] uh client discussions that you're [34:56] having. [34:57] >> Sure. Um a couple of things. The first [34:59] is let's remember that the AI investment [35:02] that we're seeing is not about where AI [35:04] stands today. It is about what AI can do [35:07] in the future. And I just pulled some [35:09] basic data from a great website called [35:10] artificial int analysis.ai. [35:13] And it shows you the speed, the [35:15] intelligence of various models over [35:17] time. Since November 22 when it [35:19] launched, Chat GPT's intelligence is up [35:21] 1300%. [35:23] In the last year, clause has improved by [35:24] 76%. AI is dramatically exceeding [35:27] Moore's law, which is a double over two [35:29] years or 41% keer. The race is about [35:33] getting to artificial general [35:35] intelligence, a much more all-inclusive [35:38] kind of intelligence that can teach [35:40] itself. This is what I listen to a ton [35:43] of Silicon Valley podcasts. Any founder, [35:46] anybody an expert in this space, I want [35:47] to hear what they have to say. This is [35:49] the conversation. It is not about, hey, [35:51] AI can help you code. It's strictly [35:53] about getting to AGI first or second. [35:56] That's the race. That's why this money [35:57] is getting spent. It's not to figure out [35:59] AI agents or anything else. It is about [36:01] getting to this end goal. And that's why [36:04] the money is getting spent. [36:06] >> So that's the underlying that's that's [36:08] the story. It is not about an agent that [36:10] can help me pick airline tickets. [36:12] >> Right? So if one platform gets to AGI [36:17] meaningfully faster than another, the [36:20] advantages that acrue to that platform [36:23] might make it so that it can compound [36:25] those advantages and be impossible to [36:28] compete with. [36:29] >> Exactly. And and this is why and that's [36:31] why the Chinese open source models [36:34] haven't super affected investor [36:36] sentiment where you think they would [36:37] because they are very good and they are [36:39] very they are fast. [36:41] >> It's because it's it's that's not the [36:43] the ultimate goal and equity investors [36:46] need to understand this very clearly. [36:47] What you're investing in right now is [36:49] 100% about getting to AGI as soon as [36:52] possible. That's the game. [36:54] >> Okay. And then once you're there, [36:57] what happens? the hierarchy becomes a [37:00] little bit more set in stone. [37:01] >> The hierarch and you can actually no one [37:04] will doubt that that's worth money. [37:06] People will pay for that. [37:07] >> Okay. [37:08] >> So the the monetization story becomes [37:09] very different. [37:10] >> So that's one thing. The second thing is [37:13] much more prosaic. S&P sector and stock [37:15] correlations are extremely low like [37:17] crazy low. We measure them sector- wise [37:19] they're at 10 plus year lows. And that's [37:22] very unusual. And the reason for that is [37:24] this notion of a recession proof US [37:26] economy. If the rec if there's no [37:28] recession in the in the wings and nobody [37:30] I think thinks there is then [37:32] correlations can be low because [37:33] investors can pick and choose. I want to [37:34] be long healthcare. I want to be long [37:35] financials. I want to be long tech. [37:37] That's right. Correlations fall apart [37:38] and volatility at the aggregate level [37:40] the VIX level is very low. And by the [37:43] way AI you know every hyperscaler [37:45] implicitly believes the same thing [37:47] because if you were worried about a [37:49] recession you would not be plowing all [37:50] this money and leveraging up your [37:52] balance sheets. [37:53] >> No way. [37:53] >> Just to invest in something. So yeah, [37:55] >> the implicit [37:57] >> assumption of this market and the [37:58] investment cycle that we're in is that [38:00] there is no recession for the next 5 [38:01] years. Period. Full stop. And if you [38:04] want to be super long, you have to [38:05] believe that 100%. [38:07] Yeah, I do think it's interesting that [38:10] um the tech founders and the CEOs at the [38:13] hyperscalers have looked at the last 15 [38:16] years and everything we've shaken off [38:19] and they've concluded there are risks [38:22] but the risk to look for is not a [38:25] vanilla business cycle correction like [38:28] that is not that is not going to be the [38:30] thing. [38:30] >> The risk is existential tech innovation. [38:33] >> Okay. Um, public equity investing is now [38:37] a lot like VC investing. What do you [38:38] mean by that? [38:39] >> Yeah, this is something that we we talk [38:40] a lot about with clients. You know, [38:42] where we used to own a bunch of [38:43] companies at the top of the S&P stack [38:45] that spun off like a third or a half of [38:47] their cash flows back to investors and [38:48] buybacks and dividends, that money is [38:50] gone. We're back to a VC kind of [38:53] investment cycle where every single [38:55] dollar these companies make is going to [38:57] a science project. 38% of MCI global all [39:01] country world is tech and big tech. So [39:04] basically 40% of the global index [39:07] >> is in this science project [39:10] >> and the only the only way you can really [39:12] escape it and as you know Jessica and I [39:14] have not been fans of equal weight S&P [39:16] for a long time. I see a point of it now [39:19] equal weight S&P is 14% tech. You still [39:22] own these companies in some smaller size [39:25] but you're not as leveraged as you are [39:27] as the market cap weighted indexes. So [39:29] that's the story there. [39:32] Um, and then the AI race between US and [39:34] China. I know you think a lot about this [39:36] and you read a lot about this. Is this [39:39] the is this the real existential wall of [39:43] worry uh dominating threat that's that's [39:47] out there beyond all the stuff that we [39:49] talk about on a day-to-day basis? Is [39:51] this really what it all boils down to? [39:54] >> It does. It does. Um, you know, for all [39:56] the reasons we've discussed, this is the [39:58] next big technological innovation. The [40:00] reason the US is a reserve currency is [40:02] again we talked about this at the top [40:04] but it's because this country does smart [40:06] things with its capital. So you can run [40:08] a deficit if you're innovating quickly [40:10] and your tax base will expand as a [40:12] result. If you lose the AI race that [40:15] begins to go away. That's it. If you [40:18] lose the AI race, [40:22] China's equipment and services will then [40:25] proliferate around the world. And [40:29] China's ability to trade with other [40:31] countries and sell AI products to other [40:35] countries can increase to the point [40:38] where it literally affects like currency [40:41] flows around the world. Is that what [40:43] we're saying? Um I put [40:46] >> what is the externality of that? [40:48] >> I put it a little differently. 16 of the [40:50] top most valuable global companies by [40:52] market cap are US companies. They're all [40:54] the ones you know. The other four are in [40:57] Asia and in the Netherlands and are [41:01] basically offshoots of US tech. So the [41:04] way you acrew wealth as a society, the [41:06] way you convince people that your [41:07] currency is the smartest thing to own is [41:10] by taking the globe's capital, the [41:12] world's capital, and investing it [41:14] sensibly and creating innovation. [41:16] Ultimately, innovation drives equity [41:18] returns and reserve currency status. [41:20] It's all human innovation. That's the [41:22] core of capital markets. [41:24] >> Okay? So we want our AI giants to be the [41:28] ones that conquer the world and not [41:30] China's. We want Alphabet to win and not [41:33] by do. [41:34] >> Yes. And you know for another reason the [41:36] the American system will allow those [41:38] companies to earn a massive amount of [41:40] money um and their founders to get even [41:42] richer and China has shown us they do [41:44] not like that that fundamental approach. [41:46] So you won't even be able to make a lot [41:48] of money if BYU wins or Baba wins or [41:50] 10-centent wins or deepseek wins. Those [41:54] will be constrained. Uh only in the [41:56] western system will you allow companies [41:57] to make the money they can. [42:00] >> Okay. Got it. Well, Nick, I want to say [42:02] thank you so much for stopping by. Uh we [42:04] always learn so the audience always [42:06] learns so much from you guys. I [42:07] personally learn so much from you guys. [42:10] Let's give people uh the URL where they [42:13] can learn more from Datar and get more [42:16] of your insights, which again are coming [42:17] out every day of the week. Where where [42:19] do they go? [42:20] >> Uh data research.com. Super simple, just [42:23] a simple signup box. No credit card, no [42:25] nonsense. [42:26] >> All right. You're the man. Please tell [42:27] Jessica to feel better. We miss her. and [42:29] uh have an awesome rest of your summer. [42:31] We'll check in with you guys soon. [42:33] >> Very good. Thank you.