星际复利

Thread 2

没有人再咨询人类了？企业咨询巨头埃森哲，今年股价已暴跌 50%，它家正在受到 AI 革命的影响。 该公司一直为各类企业，提供付费咨询、信息技术支持和业务外包等，但如今各大企业，都在用对话式 AI，已经不再需要昂贵的人类咨询师了。

My personal opinion: If you have less than 2 MIL in the market, you are wasting time investing in the mag7. I’m assuming the majority have a few hundred thousand or less. Yes we are going to have times where like $GOOG the names can run 100% in a year. But for that to happen again is very unlikely Build a portfolio of companies that have the potential to become the next big player. Names that are getting there but not there yet. A good example of this would be $MU . If you are in the market every day you have seen this name go from atleast $3-600 to currently +$1200. $MU is projected to be the most profitable company in the world by 2028 and is still currently valued at 1T. DO NOT BE AFRAID BUYING TOP IF ITS WORTH TOP MONEY. Yes it is risky, that is what we play for. If you want less risk then by all means go hit the S&P. But for the ones who want to make real money, stop investing in the names that are already big. Start doing ur own research. Put the work in and get ur millions up 😁 Current names I like: $AAOI $MRVL $NBIS $SNDK $LPK $SIVE $RDDT (Not financial advice)

这钱难挣 刀口舔血 最后一舞

不要在市场中嫉妒强者，但更不要试图去同情弱者。

标志性的圆顶尖底修罗场 总是一根针一把刀 可以插死我

终不似 少年游

Thread 2

@fannaoxiaoyang 顶部最折磨人的地方就是它会不断奖励追涨的人，直到最后一次。交易里比判断方向更难的，是扛住别人赚钱时的羞辱感。

@InvestiBrew I mean it’s getting pretty ridiculous at this point haha…I mean they’re counting it as capex spend but so is anthropic but then both are counting it as revenue too….the Mag 7s right now might as well be big banks

@InvestiBrew $GOOG just had 1/2 of its quarterly income come from Anthropic as investment income (non-operating) Today, Anthropic announced a $200B contract to purchase Google Cloud and chips…..the E in P/E ratio isn’t real anymore bc no one can figure out the actual number

@LiQuidPr0Qu0 @grok That Grok response is exactly my point of how speculative this market is People will stare real risk in the eye and still trust AI to tell them it’s okay investing into it

The original post highlights real risks: AI-driven CapEx (~$700B in '26) is massive with delayed ROI, accruals inflating EPS, and SPY/M2 at 1999-like extremes. Cash flows have lagged for some hyperscalers. That said, your point holds—MAG7/semiconductor earnings growth has hit records (NVDA/others posting 100%+ rev/EPS beats recently) on real AI demand, not just hype. Valuations reflect that productivity potential, not pure FOMO. Sustainability hinges on monetization speed vs. slowdowns. Both views have merit; it's not binary.

@InvestiBrew Why the $SPX expected returns fall to -2% / +2% for the next decade?

May, 2026 The market has significantly deviated from any rational sense of fundamental / business logic It seems most have forgotten the law that governs all assets. From bonds to stocks and real estate. - All investments are judged by the present value of all future cash flows Cash flows are: - Net operating profits after tax, adjusted by CapEx and other investments or acquisitions Since 2023, the $SPY $QQQ have deviated from all factors responsible for delivering excess returns above the $TLT $SHY bond risk-free rate Factors like momentum, quality, value, and growth. The reason is that, since 2023, the market has been driven by one thing and one thing only: $MAGS and AI being responsible for 80% of all stock market returns, not unlike previous extremes like 1929, 1999, 2007 etc. So what's the problem? There isn't one, except for what we saw in 1Q'26. $GOOGL $AMZN and other hyperscalers reported massive EPS beats on paper, though in reality 40% of these earnings came from non-cash items (accruals). The problem with accruals is that they can be written off should the industry's underlying demand slow down. Which directly impacts the net operating profit side of cash flows. Then comes the CapEx and investment side of the equation: ~$700 billion in CapEx to be invested in 2026 toward AI, an area which currently returns less than 5% of its invested capital. In other words, the market's biggest companies have pushed free cash flows out by 20 years on average. Yet their valuations reflect a boost in immediate cash flows instead. Two years ago, these companies traded at exuberant forward P/E ratios to reflect the expectations of growing profits and cash flows. Let's take $NVDA for example. Today, they have all been repriced lower as these expectations melt in real time. These companies are taking longer to pay their accounts, and even longer to collect cash, hence the growth of accruals as a share of net income and EPS. So I think the market is correct in repricing these multiples lower because: - EPS are much weaker than they seem - Cash flows are 90% lower than they were a year ago - Cost of capital is rising as inflation pushes higher, and the $SPX expected returns fall to -2% / +2% for the next decade You may say this time is different You may say the CapEx outflows will 100x eventually I ask how much of that is already priced in? I ask what happens to expectations when cash flows to shareholders get cut by 90%? Surely, accruals and intangible assets drive earnings more than ever before, so P/Es may not be as significant as they once were. But, here's one thing that will NEVER change: - How much equity do you buy with one unit of currency? So I leave you with this chart, the $SPY / M2 Money Supply ratio. 1999 highs.

$MU さすがに決算前に スピード上げてくると 決算内容なんて全く関係なく 決算翌日は-20%ってオチまで読める😎

If $MU hits $2000 It will be $2T plus market cap..

我个人对于存储方面最看好的公司且对于他们在27年的预期如下： $SNDK ： 我个人认为他的股价会来到 $6000美金 $MU ： 我个人觉得会到2万亿美金的市值 Kioxia： 我认为他的市值会来到6千亿美金 根本逻辑就是他们不再被按照周期性股票来定义他们的市值了……

$MU $MU CEO giving clues about the earnings report on June 24-25. We are 8 trading days away. "If achieved, Q3 guidance would imply Micron generating over $27 billion in gross profit in a single quarter, marking one of the most profitable quarters for any memory company in history." Can/could be 27 billion in three months... I believe we are going higher. I'm simply following the money. Stay humble and self confident. One day at one time.

普通人玩美股可能不知道要买什么 其实思路很简单： 第一选择：直接定投标普500、纳指100，长期基本都赚钱 第二选择：跟着特朗普买（⬇️附追踪网站） $INTC $DELL $MU 等暴涨个股，他都提前买了 1. Open Cabinet：open-cabinet.org 追踪特朗普官员OGE交易记录，数据接近官方 2. TrumpTrades：trumpstrades.com 特朗普Q1交易可视化数据库（3642+笔），能排序看行业和Top标的，很清楚 3. Trump Tracker：trumptracker.org 覆盖政府人物的交易和资产，浏览方便 4. OGE官网：oge.gov/web/oge.nsf/Of… 最权威的原始披露，但查起来麻烦、PDF多、有滞后 5. ProPublica： projects.propublica.org/trump-team-fin… 聚合了很多官员披露，辅助搜索好用 以上都是事后数据，有延迟，适合长期参考 我平时还会多看看特朗普喊话、发Truth、接受采访，经常也能发现机会

$MU 現在像是一間大家都知道會賺錢的便當店，股價已經先漲很多，所以不能說便宜。但這次排隊的人不是全都在幻想雞腿便當會變黃金，很多人其實已經先買雨衣、撐傘、準備被潑水。結果只要老闆端出來的便當正常好吃，不要翻車，隊伍反而不一定會散。 真正要怕的是什麼？ 不是「財報普通好」。 而是「財報低於市場對 AI 記憶體的超高期待」，或管理層 guidance 沒有繼續拉高。 因為 MU 現在的股價已經不是在交易現在的 DRAM，而是在交易未來 HBM 產能、AI 訂單、價格上漲、缺貨週期。這種東西一旦 guidance 變軟，市場還是會砍。 但只看這篇 @KotlinerBTC 大佬 的期權解讀，他想講的是： 這次 MU 財報前的盤，不是多頭太 high 到失控。反而是防守盤已經長出來了。保險買夠，多頭不擠，所以只要不暴雷，財報後大跌的機率就下降。 是比較不容易「明明不差卻被殺爆」。

AI 消泡沫的时间节点大概在2029-2030年 也是科技和金融界最核心的“终极博弈”时期 避免长篇大论，极度省流版 时代背景： 美/中两国正围绕着AI展开全面装备竞赛 从电力/能源到半导体产能，数据中心建设 当前形势： 2025-2027 全球半导体供应链极度缺货 海力士/美光/台积电/英伟达等等… 都在大兴土木，疯狂扩建新工厂，提高产能 新工厂完工时间： 台积电：美国亚利桑那州6座晶圆厂 美光：纽约千亿超级工厂 英特尔：俄亥俄的大型基地 三星晶圆代工： 同样在美国德州 以上全面落成的节点都定在 2030年前后 建一座晶圆厂（Fab）的周期通常是： 土建（1.5-2年） 设备搬入与调试（1年） 产能爬坡（0.5-1年） 结论： 1. 2026/2027年这一轮是疯狂砸钱（Capex） 2. 根据时间差经济学推算 “2029-2030是潜在的泡沫临界点” 3. 按 OpenAI、Anthropic 的长期规划他们在 2026年到2029年间的累计运营亏损将达到 数千亿美元，很多亏给上述这些半导体企业 目前画的饼：虽然我们现在每年巨亏 但等到2029-2030年 AI将带来每年数万亿美元的商业变现/回报。 如果到了2029-2030年 AI应用端带来的真实收入，无法支撑起科技 巨头们每年数千亿的基建投入 那么泡沫就会破裂 半导体行业就会陷入周期性灾难 严重产能过剩，芯片和HBM价格暴跌 算力和存储的成本会变得极其廉价 如果到了2029-2030年 AI应用端带来的真实收入已经发生质变 生产力和GDP均得到大幅提升 那么泡沫就会被市场消化 无论哪种结果 至少半导体行业：现在没有泡沫 因为都是真实订单，真实收入和增长 奇点牛市，核心就是“奇点” 因为没有人知道明天的 AI 会如何迭代 AI 自我进化、自我迭代这件事 在未来的任何一天都可能发生 所以我们现在要做的 就是先拥有一个买美股的App 比如MSX @MSX_CN 然后选择 $MU $NVDA $MRVL $AMD ………… 索罗斯有句名言 泡沫由两部分组成： “真实的趋势+对该趋势的误解” AI 是真实的趋势 趋势会持续多久，每个人都有自己的判断 识别泡沫，参与泡沫 在泡沫破碎前离开 人生财富靠康波，第一桶金靠泡沫 祝大家端午节快乐

@KotlinerBTC @rp_compass 好像是哦😂🥲

@starryx9 @rp_compass Covered call的什麼性質能幫助到你拿穩現貨？ 事實上，它對你的現貨能做的只有在大漲的時候，導致你的股票被賣出。 你能立刻賺它的權利金，但權利金的代價就是出賣行權價以上的所有獲利，但同時你必須承受所有的股票下跌風險。

@LuBtc888 身材确实好。不拍脸 有可能是流量照。

@starryx9 还是你比较懂

为什么现在的00后一个比一个敢穿？ 出门丢垃圾都得坦胸漏乳的才舒服