范式转移：从现金奶牛到基建狂魔
过去二十年，美国顶尖科技公司以其惊人的自由现金流（Free Cash Flow: 扣除资本支出后的净现金流）生成能力统治了市场。然而，当前的宏观环境正在发生深刻转型。长期以来，这些公司几乎不需要大规模实物投资就能创造巨额回报，将大量现金通过回购返还给投资者。但现在，随着 AI 竞赛的升温，它们开始投入天文数字般的资金建设数据中心、采购芯片和能源。这种从“轻资产、高回报”向“重资本、砖瓦式（Brick and Mortar）”建设的转变，引发了市场对未来回报率能否维持历史水平的深度疑虑。
Original English Source
Joe Weisenthal: One of the things that we've been talking about... is how the biggest, most profitable companies in America... they've switched from being throwing off tons of free cash flow to big investors spending a lot of money. Tracy Alloway: Yeah, that's right. So we've had years and years and years of big tech basically generating infinite amounts of cash... and now they're switching to actually spending some of that cash to build very expensive data centers and things like that. Joe Weisenthal: For much of the 2000s the investment was in sort of like intangible... SaaS type stuff. And now we're switching to really like brick and mortar. They're paying to build energy capacity and they're paying for actual chips and actual buildings to house a bunch of air conditioners and servers and all of that.
宏观透视：高估值背后的理性逻辑
尽管传统估值指标如 希勒市盈率（Shiller CAPE Ratio: 经周期性调整的市盈率）长期处于历史极高点，暗示市场存在严重泡沫，但明尼阿波利斯联储经济学家 Jonathan Heathcote 提出了一个不同的宏观视角。如果将分母从“盈余”换成“自由现金流”，我们会发现当前的股价相对于公司产生的实际可分配现金，并未显著偏离历史波动区间。自 1980 年以来，尽管企业盈余稳步增长，但自由现金流的增速更快。这主要归功于两个宏观杠杆：一是劳动力份额（Labor Share: 总产出中分配给劳动者的比例）的持续下滑；二是企业在很长一段时间内无需大规模投资即可维持增长，导致资本支出强度降低。
Original English Source
Jonathan Heathcote: Price-earnings ratio... has been drifting up and up... way above its historical average for a long time. But you know, there are other metrics you can look at. Another thing you can look at is you can look at prices relative to free cash flow... If you look at that ratio, the value of all the firms in the US relative to the total cash flow they're generating... it doesn't have a long-term drift. Joe Weisenthal: What you're saying is that when you look at the entire market through this lens, it just does not have the same extreme drift outside of normal ranges that you see when you look at price-earnings ratios. Jonathan Heathcote: That's right. If you look at where it was, say in 1980... and you look at it in the second quarter of 2022, that ratio of value to free cash flow is the same.
财富分配的剪刀差：劳动力与资本的博弈
企业利润分配的结构性变化是驱动股价上涨的底层动力。研究显示，自 1980 年至 2022 年，美国公司部门的工资和薪金支出占 GDP 的比例下降了约 8 个百分点，这意味着“饼”的更大份额流向了资本所有者，而非劳动者。这种劳动力份额下降（Declining Labor Share: 劳动报酬在国民收入中占比缩小）与股价上涨之间存在明显的正相关。虽然部分员工通过股票期权获得了补偿，但从宏观总量来看，财富更多地向拥有资本的少数群体集中，加剧了财富不平等（Wealth Inequality: 不同群体间资产积累的差距）。
Original English Source
Jonathan Heathcote: A larger share of the pie seems to be going to owners of firms and a smaller share as wages to workers. And obviously that ties directly into thinking about valuations. Joe Weisenthal: At the aggregate level... labor share... it's generally going down. How robust is that measure? Jonathan Heathcote: Those wages and salaries have fallen by about eight percentage points since 1980... from 1980 to 2022. So, that's eight percentage points of GDP. That's a big change.
AI 实验场：投资浪潮与社会公平的变数
AI 的兴起正在现实世界中检验这一宏观框架。一方面，AI 被视为进一步降低劳动力成本、提高生产力（Productivity: 单位投入产生的产出效率）的神器，可能促使劳动力份额进一步萎缩，从而利好资本估值。但另一方面，AI 的应用并非“免费午餐”，它需要巨额的前期投入。与 1980 年代 IT 革命初期类似，市场正在经历一场“赢家与输家”的重构。Jonathan 认为，AI 可能不仅仅挑战低端工作，甚至会威胁到高薪的知识工作者（Knowledge Workers: 以处理信息和知识为核心职责的专业人员），而护理、建筑等实操性强的工种反而表现出更强的韧性，这或许会带来一种意想不到的社会财富再平衡。
Original English Source
Jonathan Heathcote: I think you can make a case that this AI is going to reduce labor share of income further. So that would be a plus for stock values. On the other hand... they're definitely doing plenty of capital expenditure. Jonathan Heathcote: Historically... technology was going to automate kind of low-skilled jobs... And now I think that's flipped a little bit and now... the high-skilled workers are starting to get a little bit nervous that maybe it's going to be some of the knowledge workers that are going to get replaced. Joe Weisenthal: I like the version of compressing inequality where people who got paid less would make more than the one where the knowledge workers shrink.