We Answer the Number One Question Facing Investors Right Now | WDWL

解答当前投资者面临的首要疑问 | WDWL 讨论小组探讨了美国经济抵御传统衰退的韧性背后的结构性驱动因素，重点阐述了服务业、能源效率、灵活的货币政策以及更高的赤字支出如何为企业增长提供缓冲。Jessica Rabe 详细分析了科技股相较于标普500指数大盘表现极度分化的统计数据，并指出半导体板块的高估值可能引发资金向表现落后的大盘股轮动。最后，对纳斯达克牛市进入第四年的历史走势分析表明，在没有美联储突发利率冲击的情况下，广泛的股市涨势在结构上仍能获得有力支撑。 股市的核心问题在于，在要求立即兑现收入之前，投资者还能容忍科技超大规模企业进行庞大AI资本支出多久。 (0:13) 美国经济展现出极强的抗衰退能力，在过去16年中仅有1%的时间处于衰退，而此前各15年周期中该比例为13%至14%。 (2:10) 六大关键结构性转变驱动了现代经济韧性：向服务业转型、能源效率提升、技术驱动的管理、美联储灵活的货币政策、零工经济的缓冲作用以及高基准政府支出。 (5:04) 占GDP 6%的联邦基准预算赤字为需求提供了持续托底，尽管公共债务总额已超过GDP的122%，仍支撑着经济增长。 (5:04) 结构性经济稳定性支撑了高股票估值（市盈率约20倍）、收窄的信用利差、更低的市场波动率，以及企业为巨额AI资本支出提供资金的信心。 (12:37)

DataTrek Research co-founders Nick Colas and Jessica Rabe argue that structural economic shifts have made the United States uniquely recession-resistant, justifying high equity valuations and underwriting massive corporate AI infrastructure investments. While statistical extremes in semiconductor stocks signal a near-term rotation into mega-cap laggards, historical cycle data suggests the secular tech rally can continue absent an aggressive monetary tightening shock. Structural Drivers of Macroeconomic Resilience (2:10) Nick Colas highlights that between 2010 and 2026, the U.S. economy spent roughly 1% of its time in recession—just two months during the 2020 shock—compared to 13% to 14% in preceding 15-year blocks. Despite confronting the 2011 European debt crisis, the 2018–2019 trade war, 500 basis points of Federal Reserve rate hikes, and 2023 regional bank failures, the economy avoided prolonged contractions. Colas attributes this durability to six structural shifts: an economy dominated by services rather than manufacturing (reducing inventory bullwhip risks), increased energy efficiency per dollar of GDP, agile corporate management utilizing real-time data, rapid Federal Reserve policy interventions, gig-economy employment buffers, and baseline federal budget deficits running at 6% of GDP compared to a 3% historical norm. Addressing high debt levels, Colas observes that even with federal debt surpassing 122% of GDP, 10-year Treasury yields remain comparable to 2002–2004 levels. Host Josh Brown characterizes the economy as recession-resistant rather than recession-proof, noting that future downturns are far more likely to stem from unanticipated external shocks or policy mistakes than traditional inventory-driven business cycles. Corporate Stability and the AI CapEx Cycle (12:37) Colas links structural macroeconomic stability directly to equity valuations and corporate spending behavior. Predictable economic cash flows support S&P 500 price-to-earnings multiples around 20x and keep credit spreads near their tightest historical percentiles. Because baseline earnings risks remain low, equity market volatility stays subdued, repeatedly rewarding investors for buying pullbacks. Crucially, Colas asserts that this macro predictability enables hyperscale tech firms to commit hundreds of billions of dollars to generative AI capital expenditures without fearing that an immediate recession will deplete their reserves. However, Colas points out two structural side effects: ongoing fiscal and consumer demand makes the Fed's 2% inflation target harder to reach without an economic contraction (a target mathematically rooted in the typical 2-percentage-point disinflationary drag of post-1950 recessions), while persistent asset compounding at 10% annually widens the wealth divide against non-investors. Semiconductor Extremes and Tech Sector Rotation (19:50) Jessica Rabe warns that technology sector performance reached historically unsustainable extremes in mid-2026. On June 2, the Technology Select Sector SPDR Fund (XLK) outperformed the S&P 500 by 29 percentage points over a rolling 50-day window—a 6-plus standard deviation event—and by 25 points over a 100-day window, an extreme matched in only 0.7% of trading days since 1999. In comparable historical periods, such as 1999–2000, tech subsequently underperformed the broader market by an average of 10.8 percentage points over the following 100 days. Underlying this divergence, Rabe contrasts the top five semiconductor stocks with the broader mega-caps. Sell-side analysts raised forward EPS projections for the top five chipmakers by over 30% across 90 days, pricing in 61% year-ahead earnings growth and driving forward valuations to 52.5x (+168% YTD). By comparison, the Magnificent 8 (excluding Tesla) trades at 25.9x forward earnings with 23% projected growth (+4.5% YTD). DataTrek anticipates a rotation within tech out of richly valued semiconductors into mega-cap laggards. Addressing investor fatigue toward hyperscaler CapEx, Josh Brown compares the cooling sentiment around heavy spenders like Meta and Oracle to early 1999, when spending internet service providers faltered before hardware suppliers eventually rolled over. Nasdaq Bull Market Precedents and Rate Risks (33:13) Contextualizing the Nasdaq's multi-year run (+43% in 2023, +29% in 2024, +20% in 2025, and +13.1% YTD in 2026), Rabe analyzes historical index behavior following bear market troughs. Since 1972, bull markets lasting between three and six years have occurred six out of ten times. Notably, the Nasdaq Composite has never concluded an expansion at precisely four years; closing Year 4 in positive territory has historically signaled that gains will continue into Year 5 or Year 6. Across six historical instances where the Nasdaq logged three consecutive positive years after a downturn, Year 4 delivered positive returns 67% of the time, averaging a 16.8% gain when excluding losing years. In the two historical instances where Year 4 suffered losses (1994's 3% decline and 2022's 33% decline), the contractions were directly induced by Federal Reserve rate-hiking campaigns. Rabe concludes that unless incoming Fed leadership triggers an unexpected monetary tightening shock, historical trends favor remaining invested through market dips. DataTrek concludes that while statistical extremes point toward a rotational pause between semiconductor winners and mega-cap laggards, the broader tech advance remains supported by structural macroeconomic resilience. Barring an unforeseen exogenous crisis or an aggressive monetary tightening shock that raises the cost of capital, historical market duration models suggest the equity expansion has further room to run.

DataTrek Research 联合创始人 Nick Colas 与 Jessica Rabe 认为，美国经济已展现出抵御传统商业周期衰退的结构性韧性；尽管科技板块短期内出现极端分化与均值回归压力，但在缺乏美联储严重利率冲击的前提下，整体牛市格局有望延续。 经济结构转型削弱衰退脆弱性 (2:10) Colas 指出，过去 15 年间美国经济仅有约 1%（约两个月）的时间处于衰退之中，而此前两个 15 年周期的衰退时间占比分别为 14% 和 13%。该经济体在此期间承受住了 2011 年欧债危机、2018 至 2019 年紧缩与贸易摩擦、2021 至 2022 年加息 500 个基点及多起地缘政治冲击，表现出远超传统历史模型的抗压能力。驱动这一转变的核心在于六大结构性支柱：由高库存风险的重工业向服务业转型、单位实际 GDP 能耗下降降低了石油敏感度、数字化管理与高流动性劳动力加速资源配置、美联储政策转向更为敏捷、零工经济充当就业缓冲垫，以及占 GDP 约 6% 的常态化联邦财政赤字为总需求托底。 宏观稳定对资产估值与资本开支的支撑 (12:37) 宏观经济的平稳运行为企业现金流提供了确定性，Colas 认为标普 500 约 20 倍的市盈率源自可持续的盈利增长而非盲目乐观，同时投资级与高收益信贷利差均处于历史最低水平区间。这种宏观底座消除了企业对经济骤然衰退的担忧，使科技巨头敢于向人工智能基础设施投入数千亿美元资本开支。不过这一结构也伴随着代价：持续的需求韧性导致底层通胀粘性更高，美联储若不引发经济收缩则更难将其压回目标区间，同时资产增值也进一步拉大了有资产群体与非投资者的财富差距。 科技板块极端超额收益引发轮动预期 (19:50) Jessica Rabe 分析称，截至 2026 年 6 月 2 日，标普科技板块（XLK）相对大盘的 50 天超额收益达 29 个百分点，偏离度超过 6 个标准差；100 天超额收益达 25 个百分点，超过 27 年均值（1.2 个百分点）逾 3 个标准差，在统计上属于发生概率仅 0.7% 的极端水平。历史数据表明，此类极端超额往往伴随阶段性跑输，例如 1999 至 2000 年期间类似极端读数后科技股平均跑输大盘 10.8 个百分点。近期科技板块已自高点回落 6.3%，同期标普 500 仅跌 0.60%，反映出相对表现向均值收敛的健康整理过程，且这一收敛更可能通过资金向其他板块或科技滞涨股轮动实现，而非大盘暴跌。 半导体与巨头估值分化及开支端博弈 (24:40) 基本面预测显示，过去 90 天分析师对标普五大半导体公司的今明两年 EPS 预期分别上调 36.5% 和 33.0%，隐含未来一年 EPS 增速达 61%；而 Magnificent 8（排除特斯拉）同期 EPS 仅上调 7.5% 和 4.8%，预期增速为 23%。但估值已充分体现上述预期：前五大半导体年内涨幅达 168%，预期市盈率达 52.5 倍，而排除特斯拉的 Mag 8 估值仅 25.9 倍，年内涨幅仅 4.5%。Josh Brown 提出当前市场对云厂商庞大 AI 资本开支的容忍度正在下降，类似 1999 年网络服务商开支放缓后硬件供应商最终承压的先兆；Colas 则指出当年崩盘本质源自美联储剧烈加息抬高资本成本，而非单纯的开支题材破裂，后续资金亦可能在科技内部寻找更具防御性的标的。 纳指第四年走势的历史概率指引 (33:13) 纳斯达克综合指数目前处于 2022 年大跌后的第四年上涨通道中。Rabe 梳理自 1972 年以来的数据指出，纳指在经历下跌年份后实现连续三年上涨的情况共有 6 次，其中第 4 年继续上涨的概率为 67%（4 次上涨，2 次下跌）。全部 6 次案例的第 4 年平均回报为 5.1%，若剔除 1994 年（-3%）和 2022 年（-33%）两次由美联储剧烈加息触发的亏损年份，其余 4 次的平均涨幅达 16.8%。历史上纳指从未在第四年涨势结束后立即彻底见顶，只要年内收涨，行情往往还会延续 1 至 2 年。这意味着在没有突发性货币政策利率冲击的前提下，当前调整依然属于长期牛市中的布局窗口。 当前美国宏观体系表现出抵御传统衰退的较强韧性，为 AI 基础设施投资提供了底座支持；尽管短期内半导体与科技股面临超买技术指标的休整压力，但历史周期表明，只要美联储不实施破坏性的突发加息，牛市基础依然稳固，真正的下行风险更可能来自不可预知的外部外生冲击而非内生周期拐点。

## Background 本期视频由知名财经频道 The Compound 于 2026 年 7 月 16 日在 YouTube 平台发布。节目由 Josh Brown 主持，并邀请了 DataTrek Research 的联合创始人 Nick Colas 与 Jessica Rabe 共同探讨科技股领导力是否正在降温、极端市场行情背后的历史规律、超大规模科技股内部的轮动契机、半导体板块是否涨幅过头，以及纳斯达克迎来牛市第四年对投资者的深层含义。 ## TL;DR 节目围绕当下投资者最关心的核心命题展开：超大规模云厂商（Hyperscalers）激进的 AI 资本开支（CapEx）还能获得市场多久的宽容。Nick Colas 指出，现代美国经济因去工业化、能耗下降、高管管理水平跃升及巨额财政赤字托底，已形成了极具抗压韧性的“去周期化”低波动环境，有力支撑了当前的高估值与逢低买入心态。随后 Jessica Rabe 通过统计数据剖析了科技股相对于大盘的极端超买状态，揭示出资金正从高估值的半导体板块向性价比较高的 Mag 8 巨头进行内部机械性再平衡。最后，嘉宾们结合 1999 年互联网泡沫的历史经验指出，纳斯达克在连涨三年后进入第四年往往预示着牛市仍将延续，只要美联储不出现意外的激进紧缩，投资者就不应因恐高而过早离场。 ## Atmosphere & style 整场对谈专业深刻且充满机锋，兼具华尔街资深交易员的实战厚重感与量化研究员的数据严谨度。Josh Brown 与 Nick Colas 频频抛出古典学隐喻与 90 年代亲历泡沫破裂的“血泪史”，在风趣幽默的插科打诨中拆解复杂的宏观机制，而 Jessica Rabe 则以冷静扎实的历史统计分布形成理性平衡。聆听过程如同置身于顶尖机构闭门策略会，既有对宏观常识被打破的惊叹，又弥漫着在狂热牛市中保持清醒却绝不盲目做空的坚定与从容。 ## Bulletin - [03:00](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=180s) 美国经济在过去近 16 年中仅有略超 2 个月处于正式衰退期（仅占 1%），服务业主导转型与政府高赤字支出（占 GDP 约 6%）打破了传统工业库存周期引发衰退的链条。 - [19:50](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1190s) 科技板块相对于标普 500 指数的 50 日滚动回报率此前超过 6 个标准差，历史上同类极端超买后均面临相对跑输大盘的均值回归压力。 - [24:40](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1480s) 前五大半导体龙头前瞻市盈率高达 52.5 倍且年内大涨 168%，组合再平衡压力或将驱动资金机械性轮动至估值仅 25.9 倍的 Mag 8 巨头。 - [32:43](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1963s) 纳斯达克指数自 1970 年代以来从未在连涨四年后戛然而止；只要美联储不出现激进加息冲击，牛市有望延续至第 5 或第 6 年。 ## Key takeaways - [00:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=13s) 市场当前面临的最核心问题在于：投资者对超大规模云厂商（Hyperscalers）的高额 CapEx 开支还能保持多久的耐心与宽容度。 - [03:00](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=180s) 过去 15 年美国经济衰退时间仅占 1%，远低于此前两个 15 年周期的 14% 与 13%，传统衰退触发机制已严重失效。 - [05:04](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=304s) 经济从重工业与制造业转向低周期性的服务业，彻底瓦解了因实物库存积压向下游传导并引发裁员的经典多米诺骨牌衰退路径。 - [07:42](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=462s) 能耗密度下降、科技赋能企业管理、高流动性劳动力以及零工经济的缓冲，共同构成了现代美国经济抵御外部冲击的底层韧性。 - [07:42](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=462s) 财政赤字占 GDP 比例翻倍至 6% 提供了极具确定性的基准负荷需求，而总债务激增至 122% 的同时 10 年期美债收益率仍未被惩罚性推高。 - [12:37](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=757s) 标普 500 指数 20 倍前瞻 PE 并非盲目狂热，而是由企业约 20% 的稳健盈利增长、历史极端低位的信用利差及对无深度衰退的确定性所支撑。 - [12:37](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=757s) 宏观确定性维系了长期的“逢低买入”（Buy the dip）机制，并为超大规模云厂商激进推进大规模 AI 资本开支提供了必要的前置信任。 - [15:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=913s) 美国经济当前是“抗衰退”（recession resistant）而非“防衰退”；决策层展现出的快速救市看跌期权（如硅谷银行全额兜底）有效切断了流动性传染。 - [19:50](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1190s) 科技股相对于标普 500 的 100 日滚动超额收益达 25 个百分点（超均值 3 个标准差），历史回测显示其在未来数月将面临向均值靠拢的相对休整压力。 - [24:40](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1480s) 半导体板块年初至今平均暴涨 168%，前瞻 PE 达到 52.5 倍，由于科技主题基金持仓比例受风控约束，资金被动再平衡流向 Mag 8 的动力增强。 - [27:47](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1667s) 当前资本开支的买方（如 Meta、Oracle）股价不再因增加 CapEx 获得奖赏，出现类似 1999 年基础设施买方见顶而卖方设备商滞后冲高的背离隐忧。 - [29:52](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1792s) 1999 年末期市场狂热并未戛然而止，而是发生领导权转移，资金从面向个人的 ISP 转向 B2B 企业端应用；泡沫破裂的终极催化剂是美联储猛烈加息。 - [29:52](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1792s) 云厂商自研芯片浪潮使半导体制造设备股（Semi-Cap）相较单一芯片设计商具备更稳固的长期基本面护城河。 - [32:43](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1963s) 纳斯达克在经历大跌后连续三年上涨超 20% 属于极罕见前置形态，第四年胜率为 67%，常态涨幅在 6% 至 12% 之间。 - [32:43](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1963s) 历史上纳斯达克在牛市第四年出现亏损的唯二案例（1994 年与 2022 年）均由美联储紧缩政策直接造成，只要没有激进加息冲击，长牛周期往往能延续。 - [36:15](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=2175s) 在强劲的科技长周期中，试图精准波段操作或因恐高提前离场的机会成本极高，投资者应在波动中坚守大趋势。 ## Detailed walkthrough ### [00:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=13s) 赞助商信息与开场：当下投资者面临的最大问题 节目开场由赞助商 Federated Hermes 呈现。主持人介绍了主动型 ETF（Active ETF）如何改变当下资产配置与投资组合构建方式，为投资顾问及客户提供更高的灵活性。Federated Hermes 强调其主动型 ETF 投资组合经过严格严谨的审查与市场情景模拟推演，涵盖股票与债券市场的 12 只主动型 ETF，既可作为核心底仓，也可用于战术配置。 紧接着节目正式开始，Josh Brown 抛出了本次节目的核心命题：在当下股市中，投资者面临的最重大问题究竟是什么？Josh 明确指出，这个关键问题就是——“在投资者对资本支出（CapEx）感到厌倦并转为空头之前，他们还会给超大规模云计算厂商（Hyperscalers）多少宽容度与时间？”这关乎目前驱动美股所有利润增长的核心动力源泉。本期节目邀请到 DataTrek Research 的联合创始人、每日《Morning Briefing》的撰写者 Nick Colas 与 Jessica Rabe。Josh 强调，如果市场突然不再像过去几年那样对巨额 CapEx 投资报以热烈掌声，下半年美股的投资逻辑和行业轮动将发生彻底改变。 ### [03:00](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=180s) “这次真的有些不同”：长达16年的反常无衰退期 Nick Colas 调出首张分析幻灯片《This Time is (at least a bit) Different》（这次至少有些许不同）。这一观点是 DataTrek 过去一到两年间反复向机构与高净值客户探讨的核心逻辑，它不仅关乎巨头们的 CapEx 支出，还支撑着当下美股的估值体系与市场底层结构。 从数据与历史周期来看，过去 15 年间，美国经济处于正式衰退期的时间仅有刚刚超过 2 个月（即 2020 年新冠疫情爆发之初），仅占整个时间窗口的 1%。相比之下，在之前的 15 年中，衰退时间占比达 14%；再往前推 15 年，这一比例也高达 13%。这意味着在过去将近 16 年的时间里，除去疫情期间带星号标注的 2 个月异常值，美国经济几乎没有经历过任何实质性衰退。在 Nick 超过 30 年的从业生涯中，这种现象是极度反常的。尤其值得注意的是，过去十几年里并不缺乏足以引发严重衰退的黑天鹅与导火索： - 2011 年的希腊主权债务危机与欧债风暴； - 2015 年全球增长失速恐慌； - 2018 至 2019 年美联储的政策失误以及巨额关税与贸易争端阴云； - 2021 至 2022 年罕见的通胀飙升； - 2022 年俄乌冲突爆发引发的全球石油危机，以及美联储闪电加息 500 个基点； - 2023 年硅谷银行等美国区域性银行爆雷破产潮； - 随后接连出现的贸易政策震荡与中东战事及原油冲击。 按照 1990 年代分析汽车与强周期性工业板块的传统经济学模型，上述任何一个催化剂都足以在一夜之间将美国经济推入衰退，但经济却一次又一次顽强地挺了过来。 ### [05:04](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=304s) 经济去周期化与工业库存传导机制的瓦解 在探讨为何衰退迟迟没有到来的原因时，Nick 列举的第一大要素是美国经济结构的根本性转变：从 1970、1980 和 1990 年代重度的制造业与重工业经济，全面转向了服务业主导型经济。服务业的内在周期性远远低于传统工业制造，普通人无论经济好坏都需要理发、看病、去餐馆就餐，而换车或买房的需求则极容易被大幅推迟。 Josh Brown 特别叫停并追问了这一点背后的传导逻辑。Josh 指出，在工业和实物制造驱动的经济中，高库存积压（Inventory Overhang）往往是压垮骆驼、引发经济衰退的导火索。一旦下游停止订购零部件或整机设备，厂商就必须降价打折清库存，随后利润暴跌、被迫裁员，引发连锁的多米诺骨牌效应。而摆脱对实物库存的依赖，正是消除衰退触发机制的核心。 ```mermaid flowchart TD accTitle: 传统制造业库存周期引发衰退的链式反应 accDescr: 展示下游需求骤降如何导致库存暴增，进而引发裁员、资本开支削减并导致供应商濒临破产的恶性循环。 subgraph S1["下游渠道与需求冲击"] A["汽车经销商维持 60 天库存"] --> B["终端销量与需求突降 50%"] B --> C["有效库存被动拉长至 120 天"] end subgraph S2["上游生产与供应链休克"] C --> D["经销商停止向主机厂下单"] D --> E["整车厂与各级零部件供应商停产"] E --> F["初请失业金人数短期飙升 (如1990年从30万跃至50万)"] end subgraph S3["财务危机爆发"] F --> G["企业现金流枯竭并砍光 CapEx"] G --> H["供应商大规模逼近破产边缘 (如克莱斯勒危机)"] end ``` Nick 对此完全赞同，并结合自己在 1990 年代初跟踪汽车行业的研究经验做出了生动例证：当时汽车经销商通常在展厅备有 60 天库存，一旦伊拉克入侵科威特导致油价暴涨、需求腰斩，库存指标瞬间恶化至 120 天。经销商立刻停止向工厂订货，随之而来的减产不仅导致主机厂裁员，更以惊人的速度波及上下游所有零部件供应商。在 1990 年的衰退中，单周初请失业金人数在短短数周内从 30 万人飙升至 50 万人；所有供应商的资本开支瞬间归零，克莱斯勒（Chrysler）当时实质上已接近破产。 ### [07:42](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=462s) 现代经济韧性的底层支柱与“降神”般的财政赤字 回到演示幻灯片，Nick 进一步阐述了支撑现代美国经济抵御外部冲击的其余核心支柱： 1. **能耗密度显著下降**：美国经济对能源和石油的依赖程度大幅减轻，油价冲击不再像 1970 年代那样能直接瘫痪整个经济系统。 2. **企业管理精细化与科技赋能**：美国企业管理层利用先进科技手段优化运营效率的能力大幅提升，对生产与供需的调控更为敏锐。 3. **劳动力素质与流动性增强**：劳动力受教育程度大幅提升，跨行业再就业能力与地理流动性更强。 4. **科技赋能的零工经济（Gig Economy）**：零工经济成为就业市场的巨大缓冲垫，劳动者在失业后能迅速进入灵活用工领域以维持基本生计，避免了消费端瞬间断崖。 5. **极度敏捷的宏观调控与快速纠错**：货币与财政政策应对冲击更加迅捷，例如美联储主席鲍威尔在 2018 年第四季度犯下紧缩过度的政策失误后，看到 VIX 恐慌指数飙升至 36 且股市逼近熊市，便在 2019 年 1 月 4 日以惊人的速度火速逆转政策定调。 6. **基线政府财政支出的托底效应**：美国财政赤字占 GDP 的比重目前维持在约 6% 的高位，而 1979 年至 2010 年的历史常态水平仅为 3% 左右。庞大的政府支出为经济活动提供了坚不可摧的“基准负荷需求”（Base Load Demand）。 更为不可思议的是，虽然全美总债务占 GDP 比重从 2002 至 2004 年间的约 60% 飙升到了如今的 122%，但目前 10 年期美债收益率竟然与当时大致持平。Josh Brown 将这种奇观戏称为现代金融领域的“机械降神”（Deus Ex Machina）——就像古希腊戏剧作家在剧情陷入绝境、无法收场时，直接让阿波罗或雅典娜从天而降拯救世界一样，美国政府在以翻倍的赤字规模举债的同时，借贷成本却没有受到市场的惩罚性推高。 ### [09:53](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=593s) “天降神兵”式的财政刺激与现代化企业管理 讨论从宏观经济干预的底层支撑切入。Josh Brown 指出，美国经济近年来仿佛获得了一种持续推进的“机械降神”（deus ex machina）：每当经济遭遇危机，政府就祭出更大规模的财政支出，而且目前来看并没有为此付出更高的代价。Nick Colas 调侃这是高中古典学教育给出的经典比喻，但他强调，这种看似“无痛”的救市并非凭空产生，而是建立在美国经济的高效、稳健、智慧与超强弹性之上。这绝不是旧日那些末日论者所唱衰的“美国将变成津巴布韦”式的高赤字恶性通胀，而是一个具备系统重要性、运行极度良好的巨型现代经济体。 随后，Josh 提出了整份分析中最被低估的一点：美国现代企业的管理水平和对技术的运用能力大幅提高。他质疑道，社会普遍承认过去50年间科学技术、土木建筑和桥梁工程都取得了巨大进步，为什么偏偏不愿承认现代高管可以从前几十年的商业教训中吸取经验？相比上世纪五六十年代的高管，今天的 CEO 拥有浩瀚的商业管理文献和历史数据作为参照，能够主动规避前人踩过的坑。Nick 深表赞同并补充道，许多人看到如今上市公司 CEO 平均在任仅4年左右就黯然下台，便错误地认为现代管理层无能；但事实上，即便是底特律“三大车企”（The Big Three）的 CEO，尽管深处糟糕的行业周期，其整体素质和执行力也远胜于一两代人之前。 ```mermaid flowchart TD accTitle: 现代宏观与企业韧性机制 accDescr: 展现从现代企业管理效率与宏观弹性到稳定资本开支的传导链路。 subgraph MacroFoundation["底层支撑系统"] A["高效弹性的现代经济体"] B["管理学进化与吸收历史教训"] C["积极迅速的财政与货币干预"] end subgraph MarketOutcomes["资本市场表现"] D["企业盈利与现金流稳定"] E["低波动率 (VIX < 20) & 信用利差收窄"] F["“逢低买入” (Buy the Dip) 正向反馈"] end subgraph LongTermCycle["战略再投资"] G["大规模 AI 资本开支 (CapEx)"] end A --> D B --> D C --> D D --> E E --> F D --> G ``` ### [12:37](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=757s) 宏观稳定性如何重塑估值、资本开支与“逢低买入”心态 Nick 转向总结核心逻辑：宏观经济的长期平稳直接带来了企业盈利与现金流的稳定增长。当前正处于周期中段，但企业盈利增速依然维持在惊人的 +20% 左右。正是这种稳固性为美股高估值提供了坚实支撑——标普500指数达到20倍前瞻市盈率，根本不是盲目的“非理性繁荣”，而是市场在审视过去15年表现后形成的共识：企业不会在下一年突然遭遇破坏性衰退，因此投资者愿意支付更高溢价。与此同时，投资级与高收益债券的信用利差均处于多个周期的历史低位（约1%分位数），债市同样在为现金流的稳定性投下信任票。 在市场行为层面，长期宏观稳定性让恐慌指数（VIX）长期稳定在20的长期均值下方，即便遭遇冲击也能迅速回落。这种确定性孕育了散户乃至机构投资者深信不疑的“逢低买入”（Buy the dip）循环机制。Nick 特别强调，极其稳定的宏观环境是当前大规模 AI 资本开支（CapEx）热潮的根基所在。如果超大规模云厂商（Hyperscalers）的高管们预期未来12个月内会出现严重衰退、需要额外预留20%的备用现金，他们绝不敢如此激进地投资。Josh 指出，做这些重磅 CapEx 决策的年轻一代管理者甚至没有经历过七八十年代难以预测的滞胀周期；而 Nick 幽默地反击空头论调：“他们之所以没见过经济衰退，是因为我们根本不再发生传统衰退了。” - 标普500指数盈利在周期中段仍能保持约 20% 的强劲增长，打破了传统周期的降温规律。 - 投资级与高收益企业债利差下行至历史 1% 的极端低位分位数，显示信用市场对违约风险的极低预期。 - VIX 指数长期运行于 20 之下，市场面对短期负面冲击展现出极强的均值回归能力。 ### [15:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=913s) 贫富差距、2%通胀目标的起源与“抗衰退而非防衰退” 在乐观的宏观论据之后，Nick 提出了两个需要警惕的潜在隐患： 1. **财富鸿沟的加剧**：稳定的高股市回报率使拥有储蓄和投资能力的阶层以每年 10% 的复合收益率扩张财富（资产每7年即可翻倍），而缺乏投资现金流的底层群体则被彻底拉开差距。 2. **通胀中枢的上移**：美联储主席候选人（如即将出席汉弗莱-霍金斯听证会的凯文·沃什 Kevin Warsh）所接手的是一个对冲击具有极强免疫力的经济体，但这同时也意味着在强劲底层需求的推动下，通胀更容易超标。 在此处，Nick 揭示了一个关于“2%通胀目标”鲜为人知的历史数学逻辑：美联储设立 2% 目标的初衷，是因为从1950年代至今的历次经济衰退平均会使通胀率下拉约 2 个百分点；为了在衰退降温时不至于滑入类似日本的全面通缩，央行才将目标设定在 0% 加上 2% 的安全垫。然而，若当前经济彻底告别了深度衰退，通胀的下行就变得异常棘手。沃什的第一要务是“不要搞砸系统”，第二要务则是在不亲手制造一场衰退的前提下压低通胀。 Josh 用一个生动的比喻总结了当前经济的本质：“我们是‘抗衰退’（recession resistant），而非‘完全防衰退’（recession proof）——就像你可以戴着生活防水的手表游泳，但绝不能戴着它去深海潜水。”未来的衰退大概率不会由传统温和周期引发，而必须是某种无法预测的“未知的未知”（unknown unknowns）等外生冲击。但正如2020年疫情与2023年硅谷银行危机所证明的那样，美国决策层拥有反应迅速、毫不犹豫的“政策看跌期权”（policy put）。在2023年春季的银行业风暴中，官方几乎未经过冗长辩论或国会投票，一夜之间就事实性打破了 FDIC 25万美元的存款保险上限，直接兜底受波及机构的全部储蓄，展现了雷霆手段切断系统性危机的绝对意志。 ### [19:11](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1151s) 政策兜底的新范式与向科技板块的过渡 在讨论银行业危机应对时，嘉宾指出，不仅是美联储，以 FDIC（美国联邦存款保险公司）为代表的各大监管机构都在采取全新的兜底思维：无论如何储户不会承担损失，这种处理方式已然成为标准操作。过去人们认为绝不可能凭空出现的危机解决方案，如今却能迅速落地并发挥效用。 随后，Josh Brown 邀请 Jessica Rabe 展开后续的核心话题。Jessica 表示，鉴于时间安排，正好切入备受关注的科技股表现与 AI 交易的前景剖析。 --- ### [19:50](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1190s) 科技股极度超买与均值回归的历史警示 Jessica 回顾了他们在 6 月 8 日节目中的预警：当时科技股相对于标普 500 指数的 50 日滚动回报率达到了统计学上的极端水平。自那次发出警告以来，科技板块的表现确实落后标普 500 指数 1.3 个百分点。通过追踪 2015 年至今科技精选行业 SPDR 基金（XLK）与标普 500 指数的 50 日滚动价格回报，数据显示在 6 月 2 日，科技股在过去 50 个交易日内跑赢标普 500 指数达 29 个百分点，这是一个超过 6 个标准差（6-sigma）的罕见事件，也是整个数据集中最极端的读数。自该高点以来，科技板块下跌 6.3%，而标普 500 仅微跌 0.6%（60 个基点），科技板块相对跑输 5.6 个百分点。 为了消除短期噪音并观察跨越多个市场周期的结构性回报，Jessica 进一步展示了追溯至 1999 年的 100 日滚动超额回报数据（约 4.5 个日历月）。截至 6 月 2 日，科技股在过去 100 个交易日内跑赢大盘 25 个百分点，高出 27 年平均水平（1.2 个百分点）整整 3 个标准差以上，该情况在历史上仅发生过 0.7% 的时间。在历史上仅有的两次类似案例中： - **1999 年 12 月至 2000 年 4 月（共 41 次实例）**：科技股平均超额收益达到 30.9 个百分点，但随后的 100 个交易日中，每一次科技股都转为跑输大盘，平均落后 10.8 个百分点。 - **2023 年 5 月至 6 月（仅 3 次实例）**：随后的回调非常温和（不足 1 个百分点），但这主要是由于当时刚走出 2022 年熊市底部，且 ChatGPT 发布点燃了对生成式 AI 的初始热情。 Jessica 强调，2000 年真正击垮科技股的不是估值，而是美联储的紧缩政策——当年 2 月、3 月连续加息，5 月再度加息 50 个基点，将基准利率推向周期新高。虽然当前强劲的就业市场构成了缓冲，但新任美联储官员展现出的鹰派姿态仍是不容忽视的关键风险。从 6 月 2 日起算 100 个交易日将延续至 10 月下旬，预计科技板块的相对超额收益将向 1.2 个百分点的长期均值靠拢。虽然这种回撤可能给人带来“科技熊市”的体感，但从长期结构性叙事来看，这属于健康的休整。 Josh Brown 补充道，这种调整完全是“相对意义”上的。市场可能会出现资金流向医疗保健或金融等板块的轮动，板块轮动持续 6 个月并不意味着科技股必须暴跌 20%，仅凭相对表现滞后就能完成图表所预示的均值回归。 --- ### [24:40](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1480s) 科技股内部的机械性轮动：半导体与 Mag 8 的估值裂痕 Jessica 赞同 Josh 的观点，并进一步提出：即便在科技板块内部，也正在酝酿一场“机械性重于基本面”的内部轮动。她通过一组核心数据对比了“科技巨头八巨头”（Mag 8）与标普 500 中前五大半导体公司的分化现状： | 衡量维度 | 科技巨头八巨头 (Mag 8) | 前五大半导体龙头 (Top 5 Semis) | |:--- |:--- |:--- | | **近 90 天分析师当期 EPS 预测修正** | +7.5% | +36.5% | | **近 90 天分析师次年 EPS 预测修正** | +4.8% | +33.0% | | **未来一年隐含 EPS 预期增速** | 23% | 61%（接近前者的 3 倍） | | **前瞻市盈率 (Forward P/E)** | 25.9 倍（剔除特斯拉） | 52.5 倍（溢价近 27 倍） | | **年初至今平均股价涨幅 (YTD)** | +4.5% | +168% | Jessica 指出，分析师对盈利预测的向上修正（Earnings Revisions）是今年区分科技股赢家与输家的核心分水岭，五大半导体公司的次年预期无一例外均出现两位数上调。然而，在半导体板块年内涨幅已达三位数、估值大幅溢价的背景下，它们继续跑赢的门槛已经远高于半年前。 ```mermaid flowchart TD accTitle: 科技板块内部资金机械性轮动逻辑 accDescr: 展示半导体过度集中与估值高企如何促使科技基金被动轮动至估值较低的大盘科技股 subgraph Concentration ["集中度与估值压力"] A["半导体龙头年内暴涨 168%<br/>前瞻PE达 52.5倍"] --> B["科技投资组合集中度过高<br/>个股持仓逼近风控上限"] end subgraph Rotation ["资本再平衡路径"] B --> C["资金受指令约束必须留在科技板块"] C --> D["寻找估值偏低且基本面稳健的标的"] end subgraph Destination ["轮动承接标的"] D --> E["Mag 8 巨头<br/>前瞻PE仅 25.9倍 / 预期盈利增长 23%"] end ``` 这种轮动很大程度上是由投资组合的再平衡机制驱动的：当少数几只半导体标的暴涨后，科技主题基金在其身上的持仓集中度会过高；由于资金受投资范围限制必须留在科技板块内部，基金经理极有可能将获利了结的资金重新分散配置到估值较低、基本面依然稳健的年内落后标的中。当前悬而未决的问题在于：单凭 Mag 8 的低估值是否足以触发这一轮动，抑或是市场需要先等待二季度财报电话会议上超大规模云服务商（Hyperscalers）给出的指引。 --- ### [27:47](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1667s) 资本支出反思：1999 年网络股见顶的历史幽灵 Josh Brown 紧接着抛出了投资者当前面临的核心焦虑：进入 2026 年以来，市场对主导 2025 年的资本支出（CapEx）狂潮已不再盲目狂热。大举支出的买方（Spenders）股价不再因增加资本支出而上涨，Meta 和甲骨文（Oracle）等公司的股价甚至给管理层施加了节制开支的压力。华尔街正在用脚投票，明确表示不相信维持现有的开支节奏符合股东最佳利益。 Josh 引用了上周与摩根大通（JPMorgan）策略师 Michael Cembalest 的深度交流。Michael 指出，1999 年初互联网服务提供商（ISP）的股价最先见顶下跌，这是投资者对早期互联网基础设施建设信心动摇的全民公决；然而与此同时，作为资本支出受益者的硬件与组件供应商——如戴尔（Dell）、思科（Cisco）和英特尔（Intel）——股价却仍在惯性冲高。 这种开支买方股价率先转跌、卖方设备股滞后见顶的背离，被视作周期见顶的早期预警信号： - 购买基础设施的客户股价已不再对此产生正面反应； - 上游半导体与组件供应商仍在根据过度乐观的订单全力生产； - 最终供应商会突然发现自己如同卡通角色冲下悬崖，“低头一看，脚下是万丈深渊”。 这种对资本支出热潮退坡的恐惧，正是眼下笼罩在半导体概念股与 AI 资本支出宠儿头上的最大阴影。随后，Nick Colas 迫不及待地打断并接过话茬，直言 1999 年的历史切中了他的专业领域，因为当时他本人正处于 SAC Capital 的交易第一线。 ### [29:52](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1792s) 互联网泡沫末期的领导权更迭与美联储催化剂 Nick Colas 深入探讨了互联网泡沫破裂前夕常被忽略的关键拼图：在那个周期接近尾声时，接管市场领导权力的其实是 B2B（企业对企业）公司。当时市场上以 Commerce One 等为代表的企业开辟了利用互联网创造商业价值的全新叙事，因此当时绝非“网络服务商（ISP）开始见顶走跌，周期就立刻结束”那么显而易见；相反，投资者在积极寻找二阶、三阶甚至四阶效应。Nick 生动回忆起当时自己与 Commerce One 以及通用汽车（GM）的管理层坐在一起，探讨 B2B 将如何彻底重塑整个工业基础。当时市场的狂热能量并没有消退，而只是发生了转移。随后，正如 Jessica Rabe 所指出的，纳斯达克指数在短短数周内从高点暴跌 30% 的真正催化剂，百分之百源于美联储——市场猛然意识到资本成本急剧上升、资金获取渠道即将迅速关闭，这才是导致泡沫破裂的核心缘由。 Josh Brown 对此深表赞同，并坦言自己当年也在场内交易，做过 i2 Technologies（ITWO）和 Commerce One（CMRC）等大牛股，并在狂欢戛然而止时“输得惨烈，头都被打烂了”（"had my head handed to me"）。他指出，当时企业端普及应用（Enterprise Adoption）接棒消费端互联网，成为 2000 年至 2002 年甚至未来五年的宏大故事，导致 AOL 等老牌动量股被冷落，而 Commerce One 等新星迅速被捧上神坛。回到当前的半导体板块，Josh 指出，过去 90 天内芯片股的盈利预期上调了超过 30%，未来 90 天要再次实现这种幅度的上修难度极高，市场需要喘息空间。相比之下，Josh 对半导体制造设备股更具信心，因为各大超大规模云服务商（Hyperscalers）正在全力自研芯片，无论最终哪家芯片公司卖得好，只要有人在制造芯片，设备厂商就能持续受惠；不过，若权重前五的大票无法重现强劲的上调动力，整个板块的表现也很难维持当下的强势。 - **泡沫末期领导权转移**：从面向个人的消费类互联网服务商（如 AOL）转向重塑实体产业的 B2B 企业端应用（如 Commerce One、i2 Technologies）。 - **崩盘催化剂**：美联储加息导致资本成本与流动性骤然紧缩，引发纳斯达克在数周内从高点闪崩 30%。 - **半导体设备（Semi-Cap）逻辑**：得益于云服务巨头自研芯片的军备竞赛，只要先进制程制造需求存在，设备板块的基本面确定性就高于单纯的单一芯片设计商。 --- ### [32:43](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1963s) 纳斯达克牛市第四年：历史规律与收益率分布 Jessica Rabe 带领大家退后一步，从更宏观的历史维度审视科技股在连续上涨周期中的规律，重点分析当前市场所处的“第四年”（Year 4）。纳斯达克综合指数在经历了 2022 年暴跌 33% 之后，已经连续三年录得超过 20% 的惊人涨幅：2023 年上涨 43%、2024 年上涨 29%、2025 年上涨 20%。通过梳理自 1972 年以来的历史走势，Jessica 指出下行年份过后最常见的牛市反弹通常持续两年（共发生过 4 次），但持续 3 至 6 年的长牛更为普遍（在 10 次样本中占了 6 次），而当前市场正身处长牛阵营中。 至关重要的是，**自 1970 年代初以来，纳斯达克指数从未在刚好连续上涨四年时止步**。换言之，如果纳斯达克在第四年收涨，历史规律表明它通常还会再涨 1 到 2 年。回顾历史上 6 次出现“下跌后连涨三年”的样本，第四年录得上涨的有 4 次，下跌的有 2 次，胜率为 67%。在全部 6 次样本中，第四年的平均涨幅为温和的 5.1%（被 2022 年大跌拉低）；若剔除两次亏损年份，第四年平均回报率跳升至 16.8%，高于 13.2% 的长期平均水平。但即便在上涨样本中，也有 1998 年单年暴涨 40% 的极值拉动，其余三年涨幅均处于 6% 至 12% 之间。这说明在连续三年暴涨超 20% 后，第四年收益率低于长期平均水平是完全正常的，因为市场很难在预期极高的情况下再度超额惊喜。 ```mermaid flowchart TD accTitle: 纳斯达克下行年后的三年及第四年历史路径 accDescr: 展示纳斯达克综合指数在下行年份后连涨三年的历史样本分布及第四年的涨跌概率与后续走势 A["下行年份后连续上涨3年<br/>（1972年至今共6次）"] --> B{"第四年表现<br/>（胜率 67%）"} subgraph S1["正收益年份 (4/6 次)"] B -- "上涨 (4次)" --> C["剔除亏损后平均涨幅 +16.8%<br/>（受1998年+40%拉高）"] C --> D["常态区间: +6% 至 +12%"] D --> E["历史规律：从未在第4年止步<br/>通常再涨 1 至 2 年"] end subgraph S2["负收益年份 (2/6 次)"] B -- "下跌 (2次)" --> F["1994年 (-3%) & 2022年 (-33%)"] F --> G["共同成因：美联储强力加息冲击"] end ``` - **牛市年数分布**：1972 年以来，3 至 6 年的长牛组合出现概率高达 60%（6/10）。 - **极罕见的前置形态**：大跌后“连续三年年涨幅超 20%”在历史上此前仅发生过两次： - 第一次：1994 年下跌 3% 后，1995-1997 年均涨超 20%，随后在 1998 年（+40%）和 1999 年（+68%）继续狂飙，最终于 2000 年见顶。 - 第二次：2018 年下跌 4% 后，2019-2021 年均涨超 20%，随后在 2022 年遭遇美联储加息引发的熊市。 - **亏损年份的唯一共性**：历史上的两个负收益第四年（1994 年跌 3%、2022 年跌 33%）均是由“美联储加息冲击”一手造成。 --- ### [36:15](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=2175s) 坚守牛市：莫因提前离场而错失狂欢与结语 Jessica 总结道，纳斯达克年内迄今上涨 13.1%，略微低于长期均值，但只要不出现美联储激进加息的宏观外力冲击，历史经验表明纳斯达克的涨势有望延续至今年之后。尽管任何牛市途中都会出现波折与回调，投资者仍应将其视为逢低买入的良机。她笑言自己的观点听起来可能像个“坏掉的复读机”（"broken record"），但 1990 年代后期的历史是一面极好的镜子，生动地提醒了投资者：那些因为恐高或试图做精准波段而过早离场的人，在牛市后半场究竟会错失多么惊人的财富。 Josh Brown 赞许地回应道，即使是坏掉的复读机，也“一直在循环播放正确的神曲”。在投资世界里，资金管理的核心目标不是靠花样百出的杂耍来追求娱乐性，而是真真正正把大方向做对。他代表全场与观众向 Nick 和 Jessica 表达了真挚的谢意，称赞他们在判断市场大势上展现出非凡的远见，不仅帮助投资者稳稳拿住了这轮牛市，还始终准确指出了最关键的核心变量。最后，Josh 向观众推荐了 Nick 与 Jessica 的 YouTube 频道以及 DataTrek Research 的投研服务，并相约在夏末再次进行季度复盘。 ## Notable quotes > [00:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=13s) 在投资者对资本支出（CapEx）感到厌倦并转为空头之前，他们还会给超大规模云计算厂商（Hyperscalers）多少宽容度与时间？ — Josh Brown > [03:00](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=180s) This Time is (at least a bit) Different — Nick Colas > [12:37](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=757s) 他们之所以没见过经济衰退，是因为我们根本不再发生传统衰退了。 — Nick Colas > [15:13](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=913s) 我们是‘抗衰退’（recession resistant），而非‘完全防衰退’（recession proof）——就像你可以戴着生活防水的手表游泳，但绝不能戴着它去深海潜水。 — Josh Brown > [27:47](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1667s) 低头一看，脚下是万丈深渊 — Michael Cembalest > [29:52](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1792s) had my head handed to me — Josh Brown > [32:43](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=1963s) 自 1970 年代初以来，纳斯达克指数从未在刚好连续上涨四年时止步。 — Jessica Rabe > [36:15](https://www.youtube.com/watch?v=tE4Z1TscZgU&t=2175s) 一直在循环播放正确的神曲 — Josh Brown ## Mentioned Josh Brown, Nick Colas, Jessica Rabe, Michael Batnick, Kevin Warsh, Michael Cembalest, Federated Hermes, DataTrek Research, The Compound, JPMorgan, 克莱斯勒 (Chrysler), 通用汽车 (GM), Meta, 甲骨文 (Oracle), 戴尔 (Dell), 思科 (Cisco), 英特尔 (Intel), Commerce One, i2 Technologies, AOL, 特斯拉 (Tesla), ChatGPT, 美联储 (Federal Reserve), FDIC (美国联邦存款保险公司), 标普 500 指数 (S&P 500), 纳斯达克综合指数 (Nasdaq Composite), 科技精选行业 SPDR 基金 (XLK), VIX 恐慌指数, Mag 8, SAC Capital ## Video description On this episode of What Did We Learn, ⁠⁠Josh Brown⁠⁠, ⁠⁠Nick Colas and Jessica Rabe⁠⁠ discuss whether Tech's leadership is finally cooling off, what history says about rare market extremes, the case for a rotation within mega-cap tech, why semis may have gotten ahead of themselves, and what the Nasdaq's fourth year of a bull market could mean for investors. This episode is sponsored by Federated Hermes. Explore their full ETF lineup at ⁠⁠https://federatedhermes.com/us @NickColasandJessicaRabe Follow Nick and Jess! ►00:00 - Ad Read ►01:13 - Intro ►02:10 - This Time is (at least a bit) Different ►19:50 - Tech Performance and Earnings ►32:43 - Year 4 in a Bull Market Outcomes 👕 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=tE4Z1TscZgU)

[0:13] Josh Brown: 在投资者对资本支出（CapEx）感到厌倦并转为空头之前，他们还会给超大规模云计算厂商（Hyperscalers）多少宽容度与时间？ [3:00] Nick Colas: This Time is (at least a bit) Different [12:37] Nick Colas: 他们之所以没见过经济衰退，是因为我们根本不再发生传统衰退了。 [15:13] Josh Brown: 我们是‘抗衰退’（recession resistant），而非‘完全防衰退’（recession proof）——就像你可以戴着生活防水的手表游泳，但绝不能戴着它去深海潜水。 [27:47] Michael Cembalest: 低头一看，脚下是万丈深渊 [29:52] Josh Brown: had my head handed to me [32:43] Jessica Rabe: 自 1970 年代初以来，纳斯达克指数从未在刚好连续上涨四年时止步。 [36:15] Josh Brown: 一直在循环播放正确的神曲

[2:10] Nick Colas: This time is (at least a bit) different. [4:10] Nick Colas: The economy has proven far more resilient than historical frameworks predict. [11:20] Josh Brown: High deficit spending paired with stable borrowing costs functions like a slow-rolling 'deus ex machina' for economic growth. [13:30] Nick Colas: Institutional and retail investors are rewarded for buying market dips because they have structural confidence that the real economy will not fall apart. [16:40] Nick Colas: Job one is avoiding ruining the current setup, and job two is figuring out how to bring inflation down without inadvertently triggering a downturn. [17:20] Josh Brown: The U.S. economy is currently recession-resistant, though not recession-proof—like a water-resistant watch that is fine for swimming but shouldn't be taken scuba diving. [18:40] Josh Brown: Agencies across the government have demonstrated a readiness to deploy unprecedented solutions instantly when systemic risks appear. [20:40] Jessica Rabe: S&P 500 tech sector exceeded the S&P 500 by 29 percentage points over a rolling 50-day window—a 6-plus standard deviation event. [23:40] Josh Brown: Relative underperformance does not necessitate a severe nominal sell-off; sectors like healthcare or financials could simply take the lead for six months to satisfy mean reversion. [35:50] Jessica Rabe: Barring an unexpected Fed rate shock, history suggests the Nasdaq rally should continue past this year, making pullbacks continued buying opportunities. [36:40] Josh Brown: Investors do not need endless variety for entertainment; they need to get the big calls right.

[0:13] This episode is brought [music] to you [0:14] by Federated Hermes. Active ETFs are [0:16] changing the way portfolios are built, [0:18] giving advisors more [music] flexibility [0:20] for their clients, but not all ETFs are [0:22] built the same. Federated Hermes puts [0:24] the investments in their active ETFs [0:26] [music] [0:26] through a ruthless vetting process, [0:28] gaming out a wide range of market [0:30] scenarios so only the [music] strongest [0:32] survive. The result? A suite of 12 [0:35] active ETFs spanning the full stock and [0:37] bond market. Whether [music] you use [0:39] them as core building blocks or tactical [0:40] allocations, you'll get the strategies [0:42] you want [music] in a convenient ETF [0:44] wrapper. Simply put, Federated Hermes [0:47] has the active ETFs to help you build [0:49] portfolios [music] designed to last [0:51] because they've been vetted for it. [0:52] Explore the full lineup at [0:53] federatedhermes.com/us. [0:56] [music] ETFs are subject to risk and may [0:58] lose value. Federated Securities Corp [1:00] distributor, before investing, carefully [1:02] consider the fund's investment [1:03] objectives, risks, [music] charges, and [1:05] expenses. Read this and more information [1:07] in the prospectus or summary prospectus [1:09] available [music] at [1:10] federatedhermes.com/us. [1:13] Welcome back. It's an all new edition of [1:15] what did we learn? On today's show, [1:17] we're going to answer one of the biggest [1:20] questions facing the stock market today. [1:22] How much more time will investors give [1:25] the hyperscalers before they turn [1:27] [music] negative on capex spending? [1:30] You guys, I actually think this is the [1:32] question [1:33] cuz this is where all the earnings [1:34] growth is coming from. [1:36] Uh okay, I'm here with Nick Colas and [1:38] Jessica Rabe, my friends and the [1:40] co-founders of DataTrek Research and the [1:43] authors of DataTrek's Morning Briefing [1:45] newsletter, which goes out daily to over [1:48] [music] 1,500 [1:50] institutional and retail clients. Nick [1:52] and Jessica also have their own YouTube [1:55] channel, which you can find a link to in [1:57] the description below. Guys, welcome [1:59] back. Somehow it's halfway through the [2:00] summer. Hope you're enjoying yourselves. [2:03] >> So far, so good. [2:05] >> All right. [2:05] >> having us back. [2:06] >> Yeah, no. Al- always always my pleasure [2:09] and a treat for the audience. So, Nick, [2:11] we're going to start with you. [2:13] Um [2:14] I guess the the the headline is this [2:17] time is different, at least a little [2:19] bit. [2:20] Um but this framing of [2:23] uh this being the biggest question [2:26] facing investors, I really think this is [2:28] the key to the second half. [2:30] If we if we think that all of a sudden [2:33] CapEx announcements and actual spending [2:37] are not going to be greeted with the [2:38] same amount of enthusiasm as they have [2:41] been over the last couple of years, it [2:43] changes an awful lot about what we think [2:46] will work in the stock market and and [2:48] what we think may not work. You You [2:51] broadly agree with that idea, I think. [2:53] >> Absolutely. [2:54] >> Okay. [2:55] >> Yeah, couldn't say it better myself. [2:57] >> All right. So, tell us tell us [2:58] uh tell us what we need to know. [3:00] >> Okay. So, let's uh pop up the first [3:02] slide cuz this is kind of a three-point [3:04] discussion and it really goes back to [3:06] something we've been talking about with [3:07] clients for the better part of 1 to 2 [3:09] years now. And this is really [3:12] underpinning not just the CapEx [3:13] question, but literally every single [3:15] important part of the market, including [3:16] valuations. So, let's just dig in right [3:18] into it. [3:19] The title of this first slide is this [3:21] time is at least a little bit different. [3:23] And the framing here is over the last 15 [3:25] years, we've had a recession in the US [3:28] for just over 2 months, 1% of the time [3:30] during the pandemic crisis. [3:32] In the prior 15 years, we had recession [3:35] 14% of the time. In the 15 years before [3:37] that, it was 13% of the time. So, we [3:40] have had literally no recession for the [3:43] better part of 16 years now, and that is [3:46] highly unusual. Aside from 2 months in [3:48] the pandemic, which we'll put an [3:49] asterisk on, it's been a remarkable long [3:52] string of growth and it's not like we [3:54] didn't have a lot of reasons for the [3:56] economy to go into recession. 2011 Greek [3:59] debt crisis, 2015 global growth scare, [4:02] 2018-19 first year of a Fed policy [4:04] mistake and then you had tremendous [4:06] tariff and trade uncertainty. 21-22 an [4:09] inflation surge, 22 again the oil price [4:13] spike from the Russia-Ukraine war and [4:15] 500 basis points of Fed rate hikes. 2023 [4:18] regional bank failures, 25 and 26 huge [4:21] trade policy shock last year and an [4:24] equally huge oil shock and Mideast war [4:26] this year. I've been doing this 30 plus [4:29] years. I can tell you any one of those [4:31] would have snapped us into recession [4:33] literally overnight [4:34] over the any one of those catalysts and [4:37] yet we didn't have a recession. [4:39] And so the big takeaway is something [4:42] feels different and I covered the autos. [4:44] I covered cyclicals in the 1990s and I [4:46] was acutely aware of recessions. I [4:48] studied recessions. We looked at it from [4:50] an industrial standpoint. And this [4:52] period feels very anomalous to someone [4:54] like me who's been doing this such a [4:55] long time that that recession framing [4:57] kind of stopped working and the question [4:59] is why. So let's pop up the second [5:01] second presentation slide. [5:04] There's a lot of possible explanations [5:06] for this and I'll just run through I [5:08] think what is the most likely five or [5:11] six and they combine up to probably a [5:13] pretty good answer. The first is we have [5:15] a very services-based economy in the US [5:17] much less cyclical than the old [5:18] manufacturing economy that we had in the [5:21] '70s, '80s, and '90s. We transitioned to [5:23] services. Services are less cyclical. [5:25] People need to have their hair cut and [5:27] have need healthcare and go to [5:29] restaurants much more than they do need [5:32] to buy a car or a house. [5:34] Uh secondly, the US economy has become a [5:36] lot less energy intensive. Oil shocks [5:38] >> wait, wait, Nick, can we back up on that [5:39] first one? [5:41] Chart chart off for just a moment. Let [5:42] me ask a follow-up question. [5:44] >> Yeah. [5:45] >> It is absolutely true [5:47] that a services based economy is less [5:49] cyclical simply because the overhang of [5:53] high inventories [5:55] in an industrial more production based [5:58] economy is the thing that tips you into [6:01] recession when [6:03] when people stop ordering more parts or [6:06] more finished equipment or whatever it [6:07] is they start discounting what they have [6:11] profits fall employment falls there's [6:13] like a whole daisy chain of things that [6:16] flow from that if we're less if we're [6:18] less reliant on physical sort of [6:22] inventories [6:23] it takes away one of the key drivers of [6:27] what starts a recession in the first [6:28] place do I have that is that the right [6:30] cause and effect [6:32] >> It is and I'll give you a little little [6:33] uh sort of auto framing for that. So the [6:35] typical dealer keeps 60 days inventory [6:37] on the lot because they know the [6:38] customers want to come in and buy a car [6:40] right away. So you get 60 days inventory [6:42] at a certain selling rate. That selling [6:44] rate goes down by 50% all of a sudden [6:46] you have 120 days of inventory and you [6:48] stop ordering from the factory because [6:50] your dealer lot is already full and now [6:52] over full given the level of demand that [6:55] reduction in production means immediate [6:56] layoffs at the automotive level not just [6:58] at the assemblers but all the parts [7:00] companies all the suppliers around them [7:03] and it cascades extremely quickly. I [7:05] can't in [7:06] in the 1990 recession you saw initial [7:08] claims go from 300 to 500 a week in a [7:10] matter of weeks after Iraq invaded [7:13] Kuwait and oil prices spiked it's an [7:15] immediate effect they used to be [7:16] suppliers [7:16] >> Suppliers don't wait they don't they [7:18] don't wait to see they don't wait to see [7:20] ah maybe this is just a dip they say we [7:23] have too many people [7:24] >> Yes we have too many people we're [7:25] spending too much on capex. We're we're [7:27] literally we don't have the cash to [7:29] spend on capex. Every auto supplier I [7:31] covered in the early 90s was close to [7:32] bankrupt Chrysler was essentially [7:34] bankrupt all because of an oil price [7:36] spike that was it that was the whole [7:38] story it was amazing [7:40] >> Okay let's go back to the slide [7:41] >> Okay [7:42] So, less energy intensive economy, these [7:44] oil shocks cause recessions less [7:46] frequently. Now, my personal theory is [7:49] that US companies are also better [7:50] managed. They use technology more [7:52] effectively and more efficiently. It's [7:54] just a better managed system. And on top [7:56] of that, US workers are now more [7:57] educated, better educated than in past [7:59] decades. They have greater mobility. So, [8:01] if they lose a job, they're more likely [8:03] to find a new job. [8:05] At a more macro level, fiscal and [8:06] monetary policy has become very [8:08] responsive to shocks. And the latter, [8:10] monetary policy corrects really quickly. [8:12] So, Powell made a huge policy mistake in [8:14] Q4 2018. He reversed course literally [8:17] January 4th, 2019, because he saw the [8:20] VIX go to 36 and the stock market do a [8:23] an immediate bear market. He knew he was [8:24] wrong. Um so, that's another one. [8:28] And we have a tech-enabled gig economy [8:30] that acts like a buffer, a bit of a [8:31] buffer now for the labor force. If you [8:32] lose your job, you can get a gig job [8:34] until you find your next full-time job. [8:37] And then finally, and I think a lot of [8:38] folks watching this will be waiting for [8:40] this point. So, let's give it to them. [8:42] US government spending has created a lot [8:44] of incremental baseline demand. Deficits [8:46] to GDP run at 6% now. They ran at 3% [8:49] from 1979 to 2010. So, there is more [8:52] government spending providing a base [8:53] load for the US economy. And that's an [8:56] important feature. I would, however, add [8:58] this has had no effect on interest [8:59] rates. 10-year yields right now are the [9:02] same as they were in 20 2002, 2003, [9:05] 2004, when deficits were 60% of GDP or [9:09] budgeted the entire debt load was 60% of [9:12] GDP versus 122% now. So, it's not like [9:15] the market's making us pay more. [9:16] >> It's sort of like a magic trick. [9:18] We we are spending at twice the level in [9:22] terms of deficit to GDP, and yet the [9:25] rate at which the government can borrow [9:28] is unch. [9:29] And that and that it it's a [9:33] I guess they they call it a a deus ex [9:35] machina. [9:37] So, when the when the ancient Greek [9:38] playwrights had difficulty coming up [9:40] with an ending, [9:41] >> Yes. [9:42] >> they said, "Oh, and then [9:44] >> Apollo comes down. [9:45] >> Right. Athena Athena pops out, and you [9:49] know, saves the day." And it's like, [9:51] "All right. So, so we So, we've sort of [9:53] had this like slow-rolling deus ex [9:55] machina in the form of problem in the [9:58] economy? No worries. More government [10:01] spending, and we're able to pull it off [10:03] without a higher cost." And that we [10:05] don't know if and when that changes, but [10:06] that's that I think that's that's a big [10:09] one, even though you saved it for last. [10:11] >> Yeah, and I would say, very fair point [10:13] about the deus ex machina ending. An [10:15] excellent high school classics education [10:17] going on there. That's where I learned [10:19] it, too. Um [10:20] but I would say that it is it is [10:22] predicated on all the prior points on [10:24] that bullet on that chart. [10:26] >> Okay. [10:26] >> It is predicated on an efficient [10:28] economy, a strong economy, an [10:29] intelligent economy, a flexible economy. [10:32] It isn't just, "Oh, we're going to [10:34] become Zimbabwe," which was the old [10:35] thing that people used to say about high [10:37] deficits. This is a rare, robust, large, [10:40] systematically important economy, and it [10:42] runs pretty well. [10:43] >> One last follow-up question. US [10:45] companies are better managed and use [10:47] technology more effectively. [10:49] In my opinion, [10:51] of all the things on your list, this is [10:52] the most underappreciated point. [10:56] Like, we we look at science and [10:59] technology and um all of these areas [11:03] where there've been advances over the [11:05] last 50 years, and it's just a given [11:08] that we like sort of agree [11:10] things have gotten better [11:12] in how we build buildings, how we build [11:15] infrastructure, and bridge it. Why can't [11:17] we agree that [11:19] executives today have had the ability to [11:22] learn from the lessons of executives in [11:25] prior decades [11:27] and not make the same mistakes. Why [11:29] can't we agree that the science of [11:31] management, even if even if you think [11:33] it's a quasi science, like the [11:36] executives in the 1950s, 60s, 70s didn't [11:39] have the same literature to learn from [11:42] that the executives of the 2020s have. [11:45] And they can see things um that were not [11:48] smart to do and then they don't do them. [11:50] They make other mistakes. [11:51] >> Yeah. [11:52] >> And they'll make new mistakes. [11:54] >> I think people conflate the fact that [11:56] the average CEO is on a job for like 4 [11:58] years before they're fired with the idea [12:00] that management isn't any good. In fact, [12:03] management is quite strong and I as I I [12:05] agree with you. I would argue that it is [12:07] better than it was 20 years ago. I see [12:09] it just in covering industrial [12:10] companies. It's better. The CEOs of the [12:12] big three are better now. They still [12:14] face a horrible industry, but they're [12:16] better than the old ones. Um I think [12:19] it's just people get confused when they [12:20] see all the CEO got fired, he must have [12:22] sucked. Therefore, management sucks and [12:23] it's not that way. [12:24] >> On average, they are better than their [12:26] counterparts of a generation or two ago. [12:29] >> Yes. [12:30] >> And and part of that is because they've [12:32] been able to learn from the past. [12:34] >> Yes. And and embrace that knowledge. [12:37] >> Okay. [12:37] >> Final slide, why all this matters cuz [12:39] this is obviously the linchpin to the [12:41] whole discussion and it feeds directly [12:42] back up to your cap ex point at the [12:44] beginning, Josh. [12:45] What this means for investors, markets, [12:47] and policy makers. The most important [12:49] thing is a steady economy equals steady [12:51] earnings and cash flow growth. That's [12:53] the way it works. Um so we have very [12:55] stable earnings growth. We have very [12:57] earnings growth right now, plus 20% in [12:59] the middle of the cycle, which is [13:00] amazing. Which supports high valuations. [13:03] That's just why the S&P is at 20 times [13:05] earnings. It is not a function of some [13:07] irrational exuberance. It is a function [13:09] of the market looking at the last 15 [13:11] years and saying earnings are pretty [13:12] steady. We can pay more for them because [13:14] we're not going to be disappointed next [13:15] year with a big recession. [13:16] It also depresses corporate credit [13:18] spreads. So current investment grade [13:20] spreads and high yield spreads are at [13:22] multi-cycle lows. They're on like the 1% [13:24] tile. So the bond market is also saying [13:27] cash flows are are more stable. [13:29] Secondly, it feeds long-term volatility [13:31] that's below average. The VIX [13:32] consistently trades below 20, which is [13:34] long-run average, and it goes there very [13:36] quickly after a shock because this [13:38] underlying bid based on a stable economy [13:41] and stable earnings supports stock [13:43] prices. It also creates this buy the dip [13:46] mentality feedback loop that we see [13:48] among investors, not just retail, but [13:50] also institutional. You can buy the dip [13:52] if you have confidence the economy's [13:54] going to stay okay. You can't buy the [13:56] dip if you don't, and that's why buy the [13:57] dip has become such a mantra in the last [13:59] 5-10 years because of the stability. [14:02] Now, getting to the back the CapEx [14:04] point, this is super underappreciated. A [14:06] stable macro environment does allow for [14:08] much heavier capital investment among [14:10] public companies and private investors, [14:13] and that is the entire source of the [14:14] current AI CapEx cycle. We would not be [14:17] investing this much in AI if the [14:19] hyperscalers look at their businesses, [14:21] which are all cyclical, right? They all [14:22] rely on the economy, and said, "Oh, we [14:24] have to budget an incremental 20% cash [14:26] cuz there's going to be a downturn in [14:28] the next 12 [14:28] >> They're not right. They're not thinking [14:30] the way [14:31] um the CEO of an industrial corporation [14:33] may have been thinking 25 years ago. [14:36] It's totally different mentality. [14:39] Um they're they're looking at a [14:40] situation where yes, there are still [14:43] going to be ups and downs, [14:45] but not the unpredictability of the [14:47] '70s, the '80s. It's a It's just a And [14:50] let's be honest, many of these people [14:52] weren't even alive then who are making [14:54] these CapEx decisions. [14:55] >> That's true. And you know, that's the [14:57] bear case like, "Oh, they haven't seen a [14:58] recession." Like, okay, fine, but there [15:01] hasn't been one. [15:03] And that's the more important point. [15:05] >> [laughter] [15:06] >> They haven't seen one because we stopped [15:07] having them. [15:09] >> Uh so, back back to the slide to finish [15:11] up this thought. [15:12] Um [15:13] the two cautious points. The first is [15:15] strong equity returns obviously widen [15:16] the wealth gap, which is a huge topic [15:18] right now. If you are fortunate enough [15:20] to have saved a lot, earned a lot, saved [15:22] a lot, and invested wisely, you're [15:24] compounding reliably at 10% a year, [15:27] you're doubling over 7 years. Anybody [15:29] who can't invest, doesn't have the cash [15:31] flow to invest, doesn't have that [15:33] compounding, and the wealth gap [15:34] increases. And the final point, which is [15:37] kind of where I started my thought [15:39] process creating these slides, cuz I was [15:40] thinking about Kevin Warsh giving [15:41] testimony this week to Congress as first [15:43] Humphrey Hawkins. [15:45] He inherits an economy with an amazing [15:47] proven resilience against shocks, [15:49] but also one that is prone to creating a [15:51] lot of inflation, more than the Fed's [15:53] target, because underlying demand stays [15:55] strong. The easiest way to get inflation [15:57] down is to have a recession. It always [15:58] happens. It's why we have a 2% inflation [16:01] target in the first place, because [16:02] typically a recession causes a 2% [16:04] 2-point decline in inflation. That's the [16:06] 2% number. [16:07] >> Is that right? [16:08] >> Yeah, that's [16:08] >> That's where that comes That's where [16:09] that comes from. [16:10] >> That's where that comes from. [16:11] >> Okay. [16:12] >> And um and the desire not to be Japan, [16:14] not to have a Japan [16:15] >> Cuz I always thought it came I I always [16:17] thought it came from well, 3% would be [16:19] too much, [16:22] but 1% would be too little. [16:24] >> So, two? [16:25] >> So, two. [16:26] >> [laughter] [16:27] >> It's good to know that there's there's [16:29] more to it than that. [16:30] >> Yeah, I did I've done the math a bunch [16:32] of times for our clients, and you know, [16:33] you go back to every recession back to [16:35] the '50s, and you get about 2-point [16:36] decline, more in a bad recession, less [16:38] in a you you you know, one, but 2% on [16:40] average, right? So, you add two to zero, [16:42] you get two. Yeah. [16:43] >> Okay. [16:43] >> Um so, [clears throat] one final look at [16:44] that slide, just to finish this up. [16:48] So, Kevin Warsh inherits an amazing [16:50] system. Um [16:52] his job one has to be don't screw it up. [16:54] His job two has to be figure out how to [16:56] get inflation down without without [16:57] actually pushing so hard you do create a [16:59] recession. [17:00] So, the bottom line here is this time is [17:02] truly different, measurably different in [17:04] many good ways. It's helped a lot of [17:06] people, but it doesn't make them more [17:08] predictable in all ways. And so, it's [17:10] not like, oh, this time is different [17:11] means that's we're just flying into a [17:14] bunch of denial. What it means is that [17:16] it's different, but it's not more [17:18] predictable. [17:19] >> Okay, is the right way to sum that idea [17:22] up [17:23] that we are recession resistant, not [17:27] recession proof? [17:29] So, you can go swimming with a water [17:31] resistant watch, you shouldn't go scuba [17:33] diving. [17:34] Um, and at a certain point, there will [17:38] be an exogenous shock that does tip us [17:40] all the way over. [17:41] Um, that's the unpredictability. [17:44] But, like almost by definition, it'll be [17:47] an unknown unknown, [17:49] and it's probably not going to be the [17:51] type of thing that we used to say is [17:54] consistent with sort of like a plain [17:56] vanilla recession from the past, which [17:58] we seem sort of impervious to. [18:01] >> So, is that fair? [18:02] >> That is fair, and I think the market [18:04] also thinks that policy makers will step [18:06] in extremely quickly if there is a [18:08] shock, as they did in 2020. [18:10] Um, monetary policy, fiscal policy. [18:13] There is a very strong policy put a [18:15] proven policy put, and the US policy [18:18] makers have a very long track record, an [18:20] increasing track record of doing it very [18:22] aggressively and very quickly. [18:24] >> Yeah, we had a we had a uh [18:26] we had a uh a rehearsal [18:29] in we had like a fire drill in 2023. [18:32] They just changed the law. [18:33] They didn't even vote on it. They just [18:35] we had one day, we had an FDIC limit of [18:38] $250,000 [18:40] for a deposit account, [18:42] and then the next day it was unlimited. [18:44] And there was no discussion. We just [18:46] policy makers came in and said, "What's [18:49] the problem? There are five banks where [18:52] people have way too much money [18:53] deposited, and there are run on those [18:56] banks. Okay, here's the solution. [18:58] All of those banks are fine. All of [19:00] those depositors are fine, and there is [19:02] no FDIC limit. There may be a stated [19:04] limit, but we're going to put those [19:06] banks through a process, [19:08] um, and they'll be insolvent, but the [19:11] depositors will not be. And that just [19:13] became [clears throat] what it is. [19:15] >> Yeah. [19:15] >> And you know, it's not the Fed or not [19:18] just the Fed. That's basically the FDIC. [19:22] And so every agency is thinking this [19:24] way. And so there are solutions to [19:27] problems that we never before thought [19:29] could just spring up, but then they do. [19:31] >> Yeah. [19:32] Exactly right. [19:33] >> Okay. [19:34] >> All right. [19:35] >> Very very helpful. Jessica, what's your [19:37] what's your take on this idea? [19:40] >> [snorts] [19:40] >> Uh [19:41] for my next section? [19:43] >> Well, just generally speaking, you [19:47] If you if you think we should move, [19:49] let's move. [19:50] >> Sure. Yeah, let's launch into the the [19:53] next section. Just looking at time here. [19:55] Okay, so last time we were on on June [19:57] 8th, we showed that tech had just [20:00] outperformed the S&P 500 over a 50-day [20:03] weight window to a statistically extreme [20:05] degree. And we flagged that as a warning [20:08] sign for the audience. And that was [20:10] right. Since then tech has under [20:12] Thanks. Since then tech has [20:14] underperformed the S&P by 1.3 percentage [20:17] points since we were on. So today we [20:19] thought we'd update that chart and then [20:21] talk about what we expect for the AI [20:23] trade in the in the back half of this [20:26] year. So just starting with that updated [20:28] chart which shows rolling 50-day price [20:30] returns between the S&P 500 tech sector [20:33] using the XLK ETF as our proxy and the [20:36] S&P from 2015 to the present. When the [20:39] blue line's above the x-axis, tech [20:40] outperformed the S&P by the points shown [20:43] on the y-axis. So getting straight into [20:45] it, you could see on the right of the [20:47] chart that tech beat the S&P by 29 [20:49] percentage points over the prior 50 [20:52] trading days [20:53] on June 2nd, which was over a six [20:55] standard deviation event and the most [20:57] extreme reading in this data set by a [20:59] wide margin. And since then the tech [21:02] sector is down 6.3% [21:04] versus a loss of 60 basis points for the [21:07] S&P lagging by a total of 5.6 points [21:10] again since it got to that extreme on [21:12] June 2nd. So, [snorts] but we also think [21:15] it's a constructive to look at tech's [21:17] 100-day returns versus the S&P. We also [21:20] have that chart all the way back to [21:22] 1999. So, that's about 4 and 1/2 [21:24] calendar months, so it's long enough to [21:26] smooth out daily noise and consider [21:29] structural returns across several market [21:31] cycles. [21:32] So, you'll see also on the right side of [21:34] this chart that tech outperformed the [21:36] S&P by 25 25 points over the prior 100 [21:40] trading days again on June 2nd. And that [21:43] was over a three standard That was over [21:45] three standard deviations above the [21:47] 27-year average of 1.2 points. And it's [21:51] only happened 0.7% [21:53] of the time [21:55] over this time frame. So, it's very rare [21:58] and we can look at the last two readings [22:00] we saw [22:01] to see to see kind of help frame what [22:04] could happen next. So, the first was 41 [22:06] instances from December 1999 through [22:09] April 2000. Tech's average [22:11] outperformance reached 30.9 points. In [22:14] every single time, the following 100 [22:17] days saw tech underperform and by an [22:20] average of 10.8 points. And then the [22:22] second was during May and June 2023, [22:26] a much smaller sample at just three [22:27] instances and the pullback was mild [22:29] under one point, but that's because it [22:31] was right off the 2022 bear market lows [22:35] and just as excitement around AI was [22:37] taking hold with the launch of ChatGPT. [22:40] So, overall we think the the lesson here [22:43] is that what actually broke the broke [22:46] the back of tech in 2000 wasn't [22:48] valuation, it was the Fed. So, we had [22:50] sequential hikes in February, March of [22:53] that year, then another 50 basis point [22:55] hike in May and that pushed policy rates [22:58] to new cycle highs. And [snorts] today, [23:00] of course, new Fed Chair Warsh has [23:02] struck a notably hawkish tone. In 2022, [23:05] already showed us how brutally a hiking [23:07] cycle can reprice high multiple growth [23:10] names. Um of course, we do have a solid [23:12] labor market uh that remains the [23:15] offsetting factor to that, but we [23:17] obviously need to flag it as a key risk [23:20] here. So, overall, we we do [23:23] [clears throat] remain long-term bulls [23:24] on US large-cap tech, but at these [23:27] statistical extremes, like we flagged [23:28] last month, believing in further uh [23:30] strong near-term gains, we think is [23:32] betting against us 27 years of history. [23:36] So, [23:37] uh the way we frame it is 100 trading [23:39] days from that June uh from June 2nd, [23:42] when tech hit those extremes, takes us [23:44] to about late October. So, we think it [23:47] is reasonable uh to expect tech's [23:49] relative return to pull back closer to [23:51] its longer-run average of 1.2 points [23:54] over this period, and that may feel like [23:57] a tech bear market, but we do think it [23:59] is a healthy pause in uh in the in a [24:02] longer secular story. [24:04] >> And it's well and it's relative. [24:06] >> Yeah, that's a [24:07] good point. [24:07] >> get right, you could get into a market [24:09] environment where [24:11] health care and financials, which are [24:13] two pretty big sectors, obviously not as [24:15] big as tech, but um all of a sudden, [24:18] people just have a preference for those [24:20] stocks for 6 months. [24:22] Doesn't mean tech has to fall 20%. [24:25] But, you could just see like relative [24:27] underperformance, and it would satisfy [24:30] the the mean reversion that that chart [24:32] that you showed implies. [24:34] >> That's an excellent point. Uh yes, it's [24:36] on a relative basis. [24:37] >> good at this. [24:38] >> Really? You [24:39] >> [laughter] [24:39] >> are. [24:40] Um and that actually leads into my my [24:42] next point very well. Um so, in the [24:44] meantime, yeah, we do we agree. We think [24:47] there will be rotation within also tech [24:49] uh from a mechanical more so than a [24:51] fundamental perspective. So, if you just [24:54] throw back up that graphic, thank you. [24:56] To so, to uh yeah, so to set up this [24:58] discussion this graphic compares the Mag [25:01] 8 and the S&P 500's top five semi stocks [25:04] by their weightings. [25:05] Uh sell-side analysts 90-day earnings [25:08] estimate revisions, expected earnings [25:10] growth valuations, and year-to-date [25:12] returns. Um so just go through it pretty [25:15] quickly here. Over the past 90 days, the [25:18] Mag 8's current and next year EPS [25:21] estimates increased by an average of 7 [25:23] and a half 7 and a half and 4.8%. But [25:27] for the top five semi names, they're up [25:29] an average of 36 and a half and 33%. So [25:33] that's nearly five and seven times more. [25:36] And this was not just one or two names [25:37] carrying the groups. All five of the [25:39] largest semis saw double-digit upward [25:42] revisions to next year estimates. And [25:45] that momentum's also showing up in [25:47] earnings growth expectations. The Mag [25:49] 8's implied EPS growth over the next [25:51] year averages 23%. [25:53] For semis, it's the average is 61%. So [25:57] nearly triple. [25:58] And then naturally that's introduced a [26:00] valuation premium for most of the semi [26:02] names. So excluding Tesla, the Mag 8 [26:05] trade at 25.25.9 [26:07] times forward earnings. Semis average 52 [26:10] and a half times. And that's an almost [26:12] 27 point premium over the Mag 8. [26:16] >> Yeah. [26:17] >> Yeah. [26:17] >> And the and the stock price is already [26:19] reflect all of this, too. So like the [26:21] five semi names are up an average of [26:23] 168% year-to-date. The Mag 8 is up just [26:26] 4 and a half percent uh year-to-date. So [26:30] our takeaway here is that earnings [26:31] revisions have been the entire story [26:34] this year, the single thing separating [26:36] winners and losers inside tech. But [26:39] after triple-digit advances this year [26:41] for all of the S&P's top five semi [26:43] names, the bar for them to keep [26:45] outperforming is just [26:47] far higher than it was six months ago. [26:50] So and semis valuations, like we just [26:52] showed, now sit well above the the mag [26:54] eight. So, we do think the logical call [26:57] here is to expect the second half of [26:58] 2026 to see tech's year-to-date laggards [27:02] play some catch-up. Once again, this is [27:04] more mechanical than fundamental, but [27:06] the more a handful of names run, the [27:08] more concentrated any tech portfolio [27:10] becomes in them, and the more likely [27:12] that money needs to stay in tech, [27:15] the more likely that money needs that [27:17] money that Sorry, the more likely that [27:19] the money that needs to stay in tech um [27:22] starts spreading into other names with [27:23] lower valuations and decent [27:25] fundamentals. And we do think the mag [27:26] eight's collectively lower multiple [27:29] combined with still [27:30] solid expected earnings growth is the [27:33] obvious place to go. The only question [27:35] is whether evaluation alone is enough of [27:37] a catalyst to cause this rotation, or if [27:41] or if investors first want to hear what [27:43] the hyperscalers have to say on Q2 [27:46] earnings calls. [27:47] >> All right. So, so the obvious question [27:49] here then, [27:51] um and this gets back to the original [27:52] question, [27:54] we actually have seen less enthusiasm [27:57] for the types of CapEx spending that was [28:00] the dominant story in 2025 [28:03] so far throughout 2026. The spend that [28:06] the stock prices of the spenders [28:09] are not reacting to the upside, and in [28:11] many cases like Meta and Oracle, [28:15] um we're starting to see [28:17] some limitations [28:19] being being considered based on based on [28:22] stock price alone. [28:24] Um these compa- these companies are [28:26] being told by Wall Street, "We're not [28:28] convinced that continuing at this pace [28:31] is in our best interest, and we're [28:33] selling our shares." [28:35] Um so now you have this separation, and [28:38] I I spoke with Michael Cembalest from JP [28:40] Morgan about this last week. [28:42] He was pointing out that in early 1999, [28:45] the internet service provider stock [28:47] started going down. [28:49] Which was sort of a referendum on [28:52] how how how confident investors were [28:55] in the build-out of the original [28:57] internet. But while that was taking [28:59] place, the suppliers [29:01] the the the beneficiaries of the CapEx [29:04] those stocks kept going up. That's your [29:06] Dell computers, your Ciscos, your [29:08] Intels. [29:10] Um and [29:11] the way he thinks about it is that was [29:13] the early warning sign when the share [29:15] prices of the spenders are no longer [29:17] reacting positively [29:19] it's only a matter of time before the [29:21] component suppliers realize that they've [29:24] run off the cliff and they look down and [29:26] they see and they see nothing but a mile [29:29] below their feet. Like I think that is [29:31] the thing most people [29:33] are afraid of for the semi stocks and [29:35] the AI CapEx darlings. Um you guys [29:39] probably have a view on that. It's a [29:41] little bit outside the scope of what [29:42] we're talking about today, but [29:43] >> Well, I mean [29:44] >> I mean they they just said [29:46] >> Can I just jump in for one second? [29:47] >> Yeah, please. [29:48] >> The the '99 example is like straight up [29:50] my wheelhouse cuz I I was trading at SAC [29:52] those stocks at the time. [29:54] There's a missing piece to that analysis [29:56] and that is that it was the B2B [29:58] companies that took over leadership at [29:59] the very end of that cycle. So, the [30:01] Commerce Ones of the world um that had a [30:04] whole different way of playing, you [30:06] know, the internet and the value of the [30:08] internet. So, it was not immediately [30:11] clear like, oh, well, the the ISPs are [30:13] rolling over therefore the cycle is over [30:15] and that's an early warning sign. No, it [30:17] was investors looking at second and [30:18] third and fourth order effects. [30:20] I remember vividly like sitting with the [30:23] guys at Commerce One, the guys at GM [30:25] talking about what B2B was going to do [30:27] for the entire industrial base. So, it [30:30] wasn't that the energy in any way [30:31] diminished, honestly. It was the energy [30:33] shifted. And as Jessica said that the [30:36] the real catalyst for that implosion, [30:39] and we talked about this on the last [30:40] show, the Naz was down 30% over the [30:43] course of a couple of weeks from the [30:44] highs. The cause of that implosion was [30:46] 110% what Jessica said, it was the Fed. [30:49] It was the the realization like, "Oh my [30:51] god, the cost of capital The cost of [30:54] capital and the access to capital was [30:56] going to go away very quickly. And that [30:59] was really the cause. So, I take I take [31:00] Seamus' point, but I would just say [31:02] like, having lived through it, it's only [31:04] a piece of the story. [31:06] >> Okay. I think that's a really important [31:07] distinction. And and uh I I was trading, [31:11] too, and I remember all those stocks, [31:13] ITWO and CMRC and I was in them. I got [31:17] [laughter] [31:17] I I I [clears throat] had my head handed [31:19] to me when the party stopped, too, just [31:21] like everyone else. [31:22] Um but but you're right, there was a new [31:24] story that took over from the consumer [31:26] internet. And all of a sudden AOL was no [31:29] longer [31:30] uh a momentum name, but [31:33] uh Commerce One was. And it was like a [31:35] new phase for the internet bull market. [31:37] >> Yeah, that was the story for 20 or for [31:39] 2000, 2001, and 2002. That was supposed [31:42] to be the next, you know, 5-year cycle [31:44] was in it was uh enterprise adoption. [31:46] >> I think to your point I think to your [31:48] point though, Josh, on semis is this [31:50] like, what are the odds over the next 90 [31:53] days we're going to have or over the [31:55] past that we're going to have another [31:56] earnings revisions of plus 30% over the [31:58] past 90 days for semis like once again. [32:01] Uh that's a high bar, so think some some [32:04] breathing room. [32:05] >> I have more conviction in the semi [32:07] capital equipment stocks [32:10] just because there's such a huge [32:12] concerted effort within the hyperscalers [32:16] to build their own chips. [32:17] And of course, that's capital equipment [32:20] business. It almost doesn't matter who's [32:22] selling chips at that point for that [32:24] group so long as someone is. Who's [32:26] making chips, I should say. [32:28] Um I understand though, if the big five [32:32] don't don't see the same uh vigor of [32:35] upward revisions, those stocks will not [32:37] be acting as well as they do today Uh [32:40] regardless. [32:42] Right. Okay. [32:44] >> Uh [32:44] >> [snorts] [32:44] >> I wanted though to just for my last [32:46] section, I think this is a good time to [32:48] just take a step back um [32:50] and look at the longer the longer arc [32:52] for tech um specifically what history [32:55] says happens in year four of a run of [32:58] consecutive annual gains like the one [33:00] we're in right now cuz we're now in year [33:01] four. The Nasdaq just had three straight [33:04] years of gains of 20% or more. So uh 43% [33:08] in 2023, 29% in 2024, and 20% in 2025. [33:13] And these came after a rough 2022 of [33:15] course when the comp fell 33% [33:18] um so I just wanted to go into kind of [33:19] what what history [33:21] uh says happens after a down year [33:24] because it's it's pretty constructive. [33:26] Um the Nasdaq's most common bull run [33:29] last two years after down run down year [33:31] which has happened four times since [33:33] 1972. But three to six-year runs [33:36] combined are actually more common [33:38] happening six out of 10 instances and [33:40] we're currently in this camp so that's [33:41] in keeping with history. But um [33:44] importantly, the Nasdaq has never [33:46] stopped rallying at exactly four [33:48] straight years since the early 70s. So [33:51] if the comp is up this year, history [33:53] says it should rally another one to two [33:55] years. [33:56] Uh and then for as for what exactly hap [33:59] what as for what year four actually [34:01] looks like in these sequences since [34:03] that's the year we're currently in, uh [34:05] we have a couple here too and uh yeah [34:07] thank you that's perfect. The next [34:09] graphic um since 1972, the Nasdaq strung [34:12] together three straight up years after a [34:14] down year six times. Four of those six [34:17] times year four was also a gain and two [34:19] times it was a loss. So the the odds are [34:22] 67% for a fourth year of gains. [34:25] Uh the average return across all six [34:27] years is a modest 5.1% but that's skewed [34:30] lower by 2022's bear market. If you [34:32] strip out the two losing years, the [34:35] average year for gain jumps to 16.8% [34:38] along the above the long run average of [34:40] 13% 13.2% [34:43] but um that also is itself skewed by [34:45] 1998's blowout 40% with the other three [34:48] ranging from 6 to 12%. So sorry, a lot [34:51] of numbers there but the takeaway here [34:53] is that below a below average [34:56] uh gain in your four is actually the [34:57] historical norm and that's because it's [35:00] really hard to surprise the market into [35:01] another 20% [35:03] plus year for three straight straight [35:06] years. [35:07] Uh so the comp is up 13.1% year-to-date [35:10] so it's running just below average and I [35:13] think it's worth noting um [35:15] that [snorts] both losing years share [35:16] the same root cause and it's a recurring [35:18] theme in this episode of Fed rate shock. [35:21] So 1994's 3% pullback and 2022's 33% [35:26] decline both came from the Fed hiking [35:28] rates and the comp's current setup, [35:32] three straight years of 20% plus year [35:34] gains after a down year has only [35:36] happened twice before. Um so the first [35:40] was after 1994's 3% decline, then you [35:44] had 1995, 1996 and 1997 all delivered [35:48] 20% plus years. 98 99 of course kept [35:52] going a 40 and 68% before the dot-com [35:56] finally arrived in 2000. The second was [35:59] after 2018's 4% decline, you had 2019, [36:03] 2020 and 2021 all delivered 20% plus [36:06] year [36:07] uh 20% plus years, then 2022 brought the [36:11] Fed driven bear market. So again um so [36:15] the takeaway here is that history says [36:17] the Nasdaq should keep rallying beyond [36:19] this year barring of course a Fed rate [36:21] shock. Uh there will be more pullbacks [36:24] like in any bull market but we do [36:26] continue to treat them as buying [36:27] opportunities. [36:29] Uh we probably sound like a broken [36:31] record, but we do think the '90s [36:33] comparison is a useful reminder that uh [36:36] it's a useful reminder of how much money [36:38] was left on the table by investors who [36:40] sold who sold too early. [36:42] >> Yeah, a broken record, but continually [36:44] playing the right song. [36:46] And that's [36:47] the name the name of the game of what [36:49] we're all trying to do with our money is [36:51] not uh be endlessly entertained by [36:53] variety, but to actually get get things [36:56] right. And uh [36:57] so far you guys have been incredibly [36:59] prescient, and you've kept us in this [37:01] market, and you've repeatedly told us [37:04] the important things to watch for. And [37:07] uh I just I want to I want to tell you [37:09] how much we and the audience appreciate [37:10] it. So, thank you so much. [37:13] >> Thank you so much. We love We love [37:14] coming on. [37:16] >> All right. So, guys, once again, if you [37:18] want to follow Nick and Jessica's own [37:19] video channel in on YouTube, there's a [37:22] link in the show notes below. And we [37:24] hope that you check out [37:26] datatrekresearch.com, [37:28] and you can be on their subscription [37:30] list as well, just like I am. Thank you [37:32] so much, Nick and Jess. We appreciate [37:34] it. We'll check in with you soon, [37:35] hopefully at the end of the summer. Uh [37:37] in the meanwhile, enjoy. Thank you guys [37:39] for watching. Thank you for listening. [37:41] Have a great day. [37:43] >> [music] [37:50] [music]