作者以霍华德·马克斯的泡沫框架解读美股调整与AI资本开支回报
作者转述霍华德·马克斯的观点称，当前市场尚未达到真正意义上的泡沫状态，因为市场持续讨论泡沫风险，投资者并未对风险“浑然不觉”；他同时认为投资者对进一步上涨存在错失恐惧，鲁莽行为信号则较为混杂。
#研报/观点
X（推特）上的财经、科技与宏观账号
X·grokAI 评分5151 X·grokAI 评分2525 美股的泡沫分为估值泡沫和盈利泡沫
美股的泡沫分为估值泡沫和盈利泡沫 估值现在根本没有泡沫，无论静态还是动态。 盈利泡沫需要向前看，在你听到别人聊泡沫时，记得多问一句，你看到未来的盈利风险在哪里？对建立自己的观点会有帮助的
X·grokAI 评分3636 挺有意思，周末很多人聊的都是黑色星期一但早上美股开盘看起来并没那么慌不择路。“黑天鹅或者黑色星期一不会在每个业务爱好者在线上不停谈论它的时候发生。但会发生你睡着觉然后醒来发现白费20%”
挺有意思，周末很多人聊的都是黑色星期一但早上美股开盘看起来并没那么慌不择路。“黑天鹅或者黑色星期一不会在每个业务爱好者在线上不停谈论它的时候发生。但会发生你睡着觉然后醒来发现白费20%” https://t.co/nRd4b6jjEe https://t.co/UmXU2Q4cu4 引用 @qinbafrank：过去七个交易日内美股的这一波抛售是如何发生的？10月8日前后开始，市场对AI高增长叙事和资本开支可持续性的质疑逐步升温，多个事件叠加形成共振： 1）10月8日左右：OpenAI营收口径差引发第一波抛售 OpenAI向投资者披露，截至9月底年化营收约500亿美元，低于市场此前流传的约7
X·grokAI 评分4343 新云资本结构与融资方式影响其风险情景范围
新云的资本结构和融资方式会影响其风险情景范围，激进融资若依赖再融资并承担全部现货风险，在利率和现货价格走势不利时，再融资可能变得灾难性。一个年化回报率为15%的两年期项目，MOIC约为1.2倍；用两年期分期偿还的15%债务融资，贷方回报倍数也是1.2倍，而用15%的两年期债券融资，回报倍数为1.3倍。分析需明确资产与负债的久期及其中隐含的再投资机会假设。
X·grokAI 评分4040 过去七个交易日美股抛售如何发生？AI资本开支与通胀担忧共振
过去七个交易日，美股因AI变现与资本开支回报受质疑、通胀和油价担忧升温而集中抛售。OpenAI年化营收口径差、Anthropic营收增速放缓及AI基础设施公司融资压力，加上CPI、PPI和布伦特原油走高，加重市场对利率与现金流的担忧。作者认为这属于高速增长叙事下的市场波动，后续关注云巨头财报能否显示收入增长和现金回报。
X·grokAI 评分2828 Nebius的投资者不能既要又要：一边声称享受杠杆化AI基础设施扩张的上行空间，一边因为Nebius有软件栈就无视共同风险。
Nebius的投资者不能既要又要：一边声称享受杠杆化AI基础设施扩张的上行空间，一边因为Nebius有软件栈就无视共同风险。 Nebius有Aether、Soperator和Token Factory。CoreWeave有SUNK、Mission Control和Weights & Biases。AWS有SageMaker和Bedrock。编排、托管推理和开发者工具在这些提供商之间都是有竞争力的产品。它们的质量很重要，但“Nebius在建云、CoreWeave是GPU数据中心OEM”这种区分站不住脚。 你可以更喜欢Nebius的软件、客户结构或融资方式，而且它的合作伙伴出资模式可能降低资本
X·宏观AI 评分99 Wrote this 2.5 years ago. Think it's still mostly right. I hope so. FWIW my personal usage of code…
Wrote this 2.5 years ago. Think it's still mostly right. I hope so. FWIW my personal usage of code peaked around 2017 when I first became deeply informed about Machine Learning and AI. But imagining and adapting to the evolving forest has never required expertise in the trees 引用 @dampedspring：The
X·宏观AI 评分2828 Probably better to see some noticeable productivity acceleration on a macro level before pitching…
Probably better to see some noticeable productivity acceleration on a macro level before pitching 5x the growth rate ever seen in the history of man in just 3yrs. https://t.co/GaR70UEw0q https://t.co/OvGP0bDwEE 引用 @scaling01：Surveyed Anthropic staff's answers imply 17.1% GDP growth in 2030 https://
X·宏观AI 评分00 Excellent book
Excellent book 引用 @gbrew24：Good time to crack this one back open. https://t.co/eslPIdGL3L
X·宏观AI 评分1111 A guy on the All In podcast is mad about Anthropic
A guy on the All In podcast is mad about Anthropic 引用 @dnapway：Ben Horowitz says Anthropic is one of the most negligent companies he’s ever seen on security "The irony on these things is wild. Anthropic is one of the most negligent companies that we've ever seen in terms of their own security." "
X·宏观AI 评分2929 This chart is from The Economist's article on "the fast-growing stock of AI-related debt."
This chart is from The Economist's article on "the fast-growing stock of AI-related debt." Without a matching rise in loanable funds, persistent rate pressure will likely crowd out rate-sensitive, non-tech sectors. #economy #markets #bonds #yields @TheEconomist https://t.co/zbJNors0XI
X·宏观AI 评分1212 This is the only theory that makes sense me why VCs whine about Anthropic non stop. Like they feel…
This is the only theory that makes sense me why VCs whine about Anthropic non stop. Like they feel embarrassed by them somehow. 引用 @MattZeitlin：They’ll never forgive a bunch of grad student libs for making a company more valuable than almost anything they’ve ever funded and that all their portfolio
X·科技公司AI 评分1313 True, even SpaceX’s competitor’s FUTURE plans are behind Starship 🤣🤣
True, even SpaceX’s competitor’s FUTURE plans are behind Starship 🤣🤣 引用 @TSLAFanMtl：Ok so yeah... $SPCX has no competition. None. And that's today, before Starship becomes their mainstream rocket. Starship widens the moat *considerably* more. The competition is so far behind, they may as well be
X·财经博主AI 评分2424 Power On: Apple’s new HomeView shows the undeniable potential of a higher-end version with a…
Power On: Apple’s new HomeView shows the undeniable potential of a higher-end version with a robotic arm. https://t.co/NVMPS3cJOo
X·财经博主AI 评分2121 Apple’s iPhone Duo preorders were an absolute disaster. That may help the iPhone 18 Pro. More in…
Apple’s iPhone Duo preorders were an absolute disaster. That may help the iPhone 18 Pro. More in Power On: https://t.co/qX4Wp8R6sc
X·财经博主AI 评分2424 This is important to crystallize.
This is important to crystallize. The underlying demand trend is incredibly strong and required. Whether that means shaky business models, bad deals, expensive financing, or scarcity pricing pencils is another debate altogether. Someone will be there to pick up the pieces. But 引用 @BenBajarin：All
X·财经博主AI 评分3434 JPM’s chart puts global AI compute demand at ~450 GW in 2030 and ~530 GW in 2031. Translating that…
JPM’s chart puts global AI compute demand at ~450 GW in 2030 and ~530 GW in 2031. Translating that into capital and revenue shows aggressive these assumptions seem to be. At an illustrative $50B per fully equipped GW (low-end by the way as costs are moving up), supplying 530 GW 引用 @pequityresearch
X·财经博主AI 评分66 The problem with being too bearish here is the bears are way more retarded than the bulls 😔
The problem with being too bearish here is the bears are way more retarded than the bulls 😔 Bulls are annoyingly hyperbolic/extrapolative but at least they understand the technology
X·财经博主AI 评分4242 若这些数据属实，我大体认同 Josh 对 Anthropic 盈利能力的判断
发文者表示，若 Josh 引述的数据属实，他大体认同其判断：按 YE26 口径，约 1000 亿美元收入对应约 5 GW 算力，其中仅约 1.25 GW 服务付费客户。按付费服务算，每 GW 收入约 800 亿美元，按整体混合口径约 200 亿美元；引用内容称，Anthropic 85% 的毛利率和超过 40% 的算力利润率意味着 70%—75% 的算力用于研发。
X·财经博主AI 评分1818 Can someone tell me how a company that is targeting the biggest IPO in history at over $2 Trillion…
Can someone tell me how a company that is targeting the biggest IPO in history at over $2 Trillion can have it's bonds rated a "junk"? And if bond investors don't even view this as Investment Grade worthy, why should equity investors? https://t.co/LS7XPw12Rr
X·财经博主AI 评分1010 for all the folks that said ignore ARR, focus on revenues. we do realize that the trailing and…
for all the folks that said ignore ARR, focus on revenues. we do realize that the trailing and go-forward revenues are lower than the run-rate revenue which makes all the comparisons to ROI and otherwise... worse (right now) over time ARR and revenue should converge (unless it's
X·财经博主AI 评分1010 based. now this is the type of respectful pushback & fact-based discourse we need.
based. now this is the type of respectful pushback & fact-based discourse we need. no horse in the race but rooted in fundamentals and facts. presenting the sources. Rittenhouse doing a lot to raise the level of discourse on this topic here and elsewhere on this site. 引用 @RHouseResearch：@daniel_k
X·财经博主AI 评分2424 Thought it'd be helpful to frame this in economic output terms relative to history: Anthropic…
Thought it'd be helpful to frame this in economic output terms relative to history: Anthropic staff's survey answers imply US real GDP growth of ~18.6% in the 12 months to January 2030 (17.1% in the model's log terms). In the BEA annual record back to 1929, comparable growth https://t.co/j1RJcEaOKC
X·grokAI 评分2727 大多数关注 AI 交易的投资者没有意识到，货币和财政政策一直是支撑这笔交易的一大底层驱动力。下一轮政策刺激（policy pump）将难得多。本周在博客 App 上免费阅读：
大多数关注 AI 交易的投资者没有意识到，货币和财政政策一直是支撑这笔交易的一大底层驱动力。下一轮政策刺激（policy pump）将难得多。本周在博客 App 上免费阅读： https://t.co/LVoRhG3Yzp 【被引推文】下一轮政策刺激（The Next Policy Pump） 过去几年，市场一直受到财政和货币双重宽松的支撑，但这些利好将会逆转，这意味着任何新的刺激都将需要远为非常规的手段。 https://t.co/o3oq3TJ8lq 原文： Most investors focused on the AI trade don't get that monetary
X·grokAI 评分2222 Markets Weekly 2026年10月17日
Markets Weekly 2026年10月17日 经济扎实 AI出现裂痕 VIX飙升 https://t.co/Z2myeuyAxK 原文： Markets Weekly October 17, 2026 Economy is solid AI cracks VIX spikes https://t.co/Z2myeuyAxK
X·grok精选AI 评分7676 作者解读Circle的USDC收益分配模式与稳定币分销网络
作者转述Circle首席商务官Kash Razzaghi关于USDC“飞轮”经济模型的文章，称Circle将大部分收入用于分销伙伴分成和基础设施再投资。
推荐理由：文中将储备收益分配与交易、支付、银行及结算网络分层梳理，并列出Circle上半年收入分配数据，可了解其所述网络构成与商业模式。
X·grokAI 评分3434 他们也会持有一些东西，押注市场会因为新闻或情绪之类的把它们炒高或打压
他们也会持有一些东西，押注市场会因为新闻或情绪之类的把它们炒高或打压 归根结底，数字自己会说话，各个团队都清楚自己承担的风险 但是，是的，认为他们都有同样深入的流程并不符合事实 【被引推文】人们高估了多经理人平台（pod shop）在建仓某个标的之前所做的尽职调查 投资论点往往由短期驱动，取决于公司下个季度的财报能否超预期 这其实并不需要你了解业务的长期基本面。只需对公司是否会超出市场一致预期形成看法 PM们当然有一份偏爱的标的清单，他们倾向于在上面下重注，分析师为这些标的维护非常详细的预测模型 账簿中总有一部分保持相当的方向性，并取决于短期催化剂 原文： They also
X·宏观AI 评分1515 Can anyone see a world in which we need five ( or more if you include non US) frontier models? I…
Can anyone see a world in which we need five ( or more if you include non US) frontier models? I don't get how all can win. But not my lane. Like why would we need 5 roughly identical generalized models who leapfrog each other in capabilities instead of 2 or 3?
X·宏观AI 评分1010 This, from the Financial Times, highlights a core reality of today's geo-economic era:
This, from the Financial Times, highlights a core reality of today's geo-economic era: Multi-causal relationships between economics, domestic politics, and geopolitics. #economy #markets #Geopolitics #geoeconomics @FT https://t.co/UK3CPCu5e9
X·宏观AI 评分2121 Instead of dooming how about "The Bull Case for equities for Q4"
Instead of dooming how about "The Bull Case for equities for Q4" I am short equities and long bonds at the moment so I don't need to wonder about the bear case. I don't need doom porn or crash memes for black monday. I need to consider where I am wrong. I'll break it down
X·宏观AI 评分1919 聊聊CRWV的真实压力以及对能都扛过此次危机？一周前觉得高Beta容易📉的其中一个原因是市场对债务的担心：博通给anthropic和openai担保了近千亿美金，spacex计划借债500亿美金买芯片…
聊聊CRWV的真实压力以及对能都扛过此次危机？一周前觉得高Beta容易📉的其中一个原因是市场对债务的担心：博通给anthropic和openai担保了近千亿美金，spacex计划借债500亿美金买芯片。当时Crwv的cds暴涨也有了苗头 这周完全成了市场最担心的话题也是大跌的导火索之一。 引用 @qinbafrank：再看上周五推文聊的几个思路，这两天：1）伊朗袭击油轮、胡塞沙特对轰 ，油价不稳住市场也没稳住； https://t.co/5aoaaS2gMx
X·宏观AI 评分1111 去看一下这周的ALL-IN，前面20分钟，JASON聊到老马发推“YIKES”，加入CHAMATH关于COREWEAVE的评论时，他发出的奸笑，我认为他肯定做空了。
去看一下这周的ALL-IN，前面20分钟，JASON聊到老马发推“YIKES”，加入CHAMATH关于COREWEAVE的评论时，他发出的奸笑，我认为他肯定做空了。 然后视频开始后他团队做的DECK聊加息对于COREWEAVE成本的影响，“因为一开始的UNSECURED
X·宏观AI 评分00 Tbh idk how this episode came about. David told us he’s an Odd Lots listener. And I’m always down…
Tbh idk how this episode came about. David told us he’s an Odd Lots listener. And I’m always down to talk with a fellow musician. https://t.co/kfDSS4U2zQ 引用 @mean_field_zane：@TheStalwart @tracyalloway How’d you pull this off.
X·宏观AI 评分4040 Circle首席商务官谈USDC“飞轮”经济模型
Circle首席商务官Kash Razzaghi撰文阐述USDC的“飞轮”经济模型：大部分收益与伙伴共享以推动采用，Circle将剩余部分再投资于基础设施、合规与扩展能力。2026 年上半年，USDC 链上结算额超过 36。
X·科技公司AI 评分44 Yeah
Yeah 引用 @strodav：@cb_doge BS. His politics are not dark and people who get inflamed are idiots. IMHO, the best way to describe his politics is common sense. Elon single-handedly saved the first amendment in America and he started the process that is identifying at least 1/2 $trillion /year in was
X·科技公司AI 评分11 Since fate maximizes for ironic entertainment, it was inevitable that the “Roadster” would be a…
Since fate maximizes for ironic entertainment, it was inevitable that the “Roadster” would be a flying car
X·grokAI 评分1616 约6%的长期预算赤字，再加上到期债务以更高利率再融资，这笔账是这样算的：
约6%的长期预算赤字，再加上到期债务以更高利率再融资，这笔账是这样算的： #经济 #市场 #债务 #收益率 https://t.co/PhUJwQvf37 原文： The math of a protracted budget deficit of around 6% PLUS the refinancing of maturing debt at higher rates: #economy #markets #debt #yields https://t.co/PhUJwQvf37
X·grokAI 评分2929 过去一周市场出现了逆流：尽管通胀数据偏热，债券收益率仍然下降；与此同时，随着科技股遭抛售、轮动加速，股市出现剧烈波动和分化。
过去一周市场出现了逆流：尽管通胀数据偏热，债券收益率仍然下降；与此同时，随着科技股遭抛售、轮动加速，股市出现剧烈波动和分化。 与此同时，从油价到地缘经济紧张，更广泛的地缘经济压力依然存在。 随着我们进入经济数据和企业财报密集的一周，“经济韧性”这一主题面临一次重要的实证检验。 详情请通过以下任一链接阅读我关于全球经济和市场的周报： https://t.co/Ak7dVy7Lnn https://t.co/eKnKFvHluE 原文： This past week brought market counter-currents: bond yields fell despite hotter
X·grokAI 评分4949 算力短缺与资本周期错配加剧市场波动
市场追逐算力短缺，推动资本承诺加速，但竞争和商品化压缩产品利润率，利率与地缘政治冲击也可能增加融资难度。投入依据是当下的稀缺经济学，回报却取决于多年后的现金流；算力和基础设施可能结构性供给不足，建设融资的资本复杂性也可能过度扩张。
X·宏观AI 评分3333 社媒真的太多关于新云融资的FUD了。
社媒真的太多关于新云融资的FUD了。 DDTL4.0，COREWEAVE最新的融资，成本SOFR+2.25%，按现在的联邦利率就是6.13% META 2032年到期的债务YIELD是5.82%