全部推文
共 6436 条 · 第 177 / 215 页 · 滚到底自动加载更多
-
列举AI领域资本支出加速增长的10大证据，强调AI赛道持续高景气度与投资机遇
对于AI领域的任何人来说，很难不看好。 资本支出正在加速增长，而且是以指数级的速度。 直接流向以下几个方面： 新云服务商：$CIFR、$NBIS、$WULF、$IREN， 连接性：$ALAB、$CRDO、$CLS， 能源：$VST、$FLNC、$TE、$EOS， 半导体/晶圆厂：$NVDA、$AMD、$GOOGL、$TSM， 存储：$SNDK、$MU和$STX。 仅在过去几周，我们就看到： 1. AI曼哈顿计划——美国政府正给予顶级模型访问专有实验室数据的权限以加速研究。 2. $GOOGL在德克萨斯州投资400亿美元建设数据中心。 3. Anthropic投资500亿美元建设边缘计算基础设施以支持其Opus 4.5+模型。 4. $TSM公布创纪录的远期收入数据（AI支出）。 5. $NVDA确认创纪录的远期收入数据（AI支出，锁定2年产量）。 6. $META将2025年数据中心/AI资本支出提升至400-450亿美元，用于llama5-6。 7. 今年三次降息以加速增长并降低融资成本。 8. Dominion Energy警告AI数据中心带来大规模电力负荷激增。 9. $AVGO表示AI网络订单达到前所未有的规模。 10. 阿联酋和主权国家推进AI发展。 我们没有看到任何放缓。只有创纪录的增长。 事实上，随着Claude Opus 4.5、Gemini 3的最新模型进展，以及美国政府的新承诺，感觉我们才刚刚看到人工智能新前沿的冰山一角。 （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）: Nebius [$NBIS]是当前被低估最多的成长型公司。 它有潜力以210亿美元的市值成为下一个$GOOGL。原因很简单： 它的投资组合公司令人惊叹。 这个概念最令人难以置信的例子是$FTX公司。以下是故事： 当我们观察$META如何增长成为万亿美元公司时，不仅仅是Facebook。他们的投资组合公司Instagram、Whatsapp和其他应用使Meta主导了社交媒体领域。 $FTX在数字资产和前沿技术领域做着类似的事情。 四年前，即2021年，$FTX向一个大资产篮子投资了58亿美元。其中很大一部分投入了这三家核心公司： 1. Anthropic，持股13.56%，估值25亿美元。 2. Robinhood [$HOOD]，持股7.6%，估值85.4亿美元。 3. Solana [$SOL]，4100万+代币。 快进到今天，那将是： · Anthropic最新一轮估值3500亿美元。那部分股份价值约474亿美元。 · Robinhood现在市值超过1000亿美元。那部分股份价值约76亿美元。 · Solana现在每个代币价值131.5美元，使那部分股份远超57亿美元。 仅这三家公司就在4年内产生了超过550亿美元的价值，这甚至还不包括FTX的数百亿美元加上其他数十项投资，以及Chime、Layerzero、Aptos、Hidden Road（被$COIN收购）和加密货币的持股。 他们的投资组合公司比他们的核心业务更持久（想象一下，如果核心业务像$GOOGL搜索和YouTube一样持续增长，那将价值多少）。 $NBIS现在有着与$FTX在加密领域、$META在社交媒体领域相同的布局，但在人工智能领域拥有合法且飞速增长的核心业务。 Nebius拥有： 1. Clickhouse，28%持股，估值约70亿美元（2025年上半年为63亿美元）。 2. Avride，83%持股，估值约60亿美元（优步融资后）。 3. Toloka AI，约65%持股，估值约6.4亿美元。 4. TripleTen，100%持股，估值约3亿美元。 · Clickhouse为Anthropic、$META、$TSLA、$NET和许多财富500强公司提供支持。 · Avride是一家自动驾驶出租车机器人公司，从Yandex分拆出来，$UBER在3.75亿美元融资轮中投资以与Waymo竞争。 · Toloka是一个AI标注平台，亚马逊、微软、Anthropic和Shopify都在使用。 19.6亿美元+49.6亿美元+4.16亿美元+3亿美元=76亿美元的投资组合公司估值，这些公司的增长速度超过大多数公开成长型公司。 但如果我们看看他们以每年700%+的速度增长至70-90亿美元ARR的核心业务，拥有48亿美元现金，为$META、$MSFT、Cursor、政府和更多客户提供支持…… 这可能是它以低于90美元的最后一个月，因为今天MSCI纳入将为其带来从数亿美元到低数十亿美元的额外资金流入。如果我们看看$IREN或$CIFR等热门选择，没有任何其他数据中心成长型公司有这种类型的投资组合。 $NBIS估值仅210亿美元，市场正在忽视这个机会。
英文原文
It’s hard for anyone in the AI space not to be bullish. Capex is ramping up. Exponentially. And flowing directly down to: Neoclouds: $CIFR, $NBIS, $WULF, $IREN, Connectivity: $ALAB, $CRDO, $CLS. Energy: $VST, $FLNC, $TE, $EOSe Semi/foundries: $NVDA, $AMD, $GOOGL, $TSM Memory: $SNDK, $MU, and $STX In the past few weeks alone, we got: 1. Manhattan Project for AI - US government is giving top models access to propriety labs data for accelerating research 2. $GOOGL spending $40 on DC buildout in Texas 3. Anthropic spending $50 on EC buildout to support their Opus 4.5+ models 4. $TSM confining record forward revenue numbers (AI spend) 5. $NVDA confirming record forward revenue numbers (AI spend, 2Y production locked in) 6. $META upping 2025 DC/AI capex spend to $40-$45B for llama5-6 7. 3x rate cut this year to accelerate growth and make funding cheaper. 8. Dominion Energy warning of massive AI power load surge from AI datacenters 9. $AVGO signaling AI networking orders at unprecedented scale 10. UAE and sovereign countries pushing into AI We’re not seeing any slowdown. Only record growth. In fact, with the recent model developments from Claude Opus 4.5, Gemini 3, and now new commitment from the US government, it feels like we're just seeing the tip of the new frontier for Artificial Intelligence.
-
分析$NBIS自建数据中心、GPU贬值、融资稀释及过度建设四大风险。
关于风险的好问题，我也喜欢讨论下行风险。 1. 超大规模云厂商(Hyperscalers)在5年内完全自建数据中心，包括GPU（$GOOGL的TPU，$AWS的Tranium）、能源和选址。 这类似于高通(Qualcomm)与苹果的关系。苹果使用$QLCM，然后建立垂直整合——一旦完成就抛弃客户。因此在此期间建立公司自身护城河很重要（$NBIS在此期间通过Cursor、Shopify等企业客户进行全栈建设并扩展子公司。完全依赖一两个超大规模云厂商合同的公司表现不佳）。 2. $NBIS、$IREN、$CRWV和$ORCL的GPU贬值风险。 任何进行全栈服务的提供商都有更高的利润率、更高的收入和更快的爬坡速度。但这涉及在GPU上花费数十亿美元，而这些GPU在未来4年内价值大幅缩水。 我一直认为GPU多年后仍有价值，例如TPU仍在使用7-8年，2020年的$NVDA GPU仍高价转售并用于不同层级的推理（低优先级），并与新模型并行运行。 3. 信贷收紧+稀释 建设支出过多->需要更多稀释以建设和购买GPU。我们已经看到$4B可转债+$2.5B ATM发行。$NBIS应该足够，但始终存在利润率低于预期的风险（如$ORCL建设），然后公司需要更多稀释。在更难融资的市场中（如$APLD垃圾债），可能会遇到$CRWV的问题，被迫以糟糕的利率融资，侵蚀利润率。 4. 过度建设 如果LLM软件有突破，例如Claude Opus 4.5使用DeepSeek类型的轻量级推理，那么GPU利用率降低，我们会看到数据中心、$NVDA、$AMD等随AI浪潮下跌。 这些可能是$NBIS的四个主要风险。每家公司不同，对于$IREN我会指出其IaaS层的软件编排以优化利润率，但$NBIS已经掌握这一点。
英文原文
Good question about risk, I enjoy talking about the downsides too. 1. Hyperscalers completely their own datacenter buildout with GPUs (TPUs with $GOOGL, Tranium with $AWS), energy, locations in 5+ years. This is the analogy of Qualcomm with Apple. Apple just uses $QLCM, then builds it's own vertical integration -> once it's done it sacks the customer. Hence why it's important to build up a company's own moat during the time ( $NBIS is doing full-stack with its own enterprise clients like cursor, shopify, etc during this time and scaling subsidiaries. Companies fully dependent on one or two hyperscaler contracts don't do well). 2. GPU deprecation for $NBIS, $IREN, $CRWV, and $ORCL. Any provider doing full stack has higher margins, higher revenue, higher ramp. But that comes into spending billions on GPUs and the GPUs not too much value over the course of the next 4 years. I've always maintained GPUs are still valuable many years later, eg. TPUs still are used 7-8 years later, 2020 $NVDA GPUs are still resold high and are used different stacks of inference (lower prio) but run alongside newer models. 3. Credit tightening + dilution Too much spend on buildout -> needs to dilute more to build + buy GPUs. We've already seen that with the $4B convertible + $2.5B ATM offering. $NBIS should have enough, but there's always the risk margins are lower than expected eg. $ORCL buildout, then the company needs to dilute more. In a market where it's harder to raise eg. $APLD junk bonds, it might run into the $CRWV problem where they are forced to raise money with terrible interest rates cutting into the margins. 4. Overbuildout If there's a breakthrough in LLM software, for example, Claude Opus 4.5 using deepseek type lightweight inference, then less utilization of GPUs then we see data centers, $NVDA, $AMD, and others go down with the AI ship. Those are probably the four main risks for $NBIS. This is different for each company, for $IREN I would point to software orchestration for their iaas layer for margin optimization but for $NBIS they have that down.
-
日元套利交易平仓将推动资金回流，建议做多日经指数。
日本加息看似对日经平均指数不利，但由于美联储正在降息，这将导致2024年重新积累的日元套利交易(Carry Trade)崩溃。因此，在资金回流日本、推动带对冲买入的平仓(Unwind)阶段，我认为做多日经指数是明智的仓位配置。
英文原文
日本の利上げは日経平均にとって一見マイナスに思えますが、米国のFRBが利下げに動いていることもあり、2024年に再び積み上がった円キャリートレードが崩壊する要因になります。 そのため、資金が日本へ還流し、ヘッジを伴った買いが促されるアンワインド局面では、日経をロングするのは賢いポジショニングだと私も考えます。
-
借鉴FTX投资组合经验，探讨核心业务搭配高增长子公司的科技巨头成长路径。
这是一个合理的评论，因为很多人对FTX这个词有创伤后应激障碍(PTSD)，且存在大量负面联想。（我并非试图制造这种联想）。 然而，尽管他们存在会计欺诈，他们在现代任何初创企业中做对了一件事，即投资组合公司的增长/投资。当散户进行长期投资时，我们可以从中吸取一些教训。 在过去4年中，我们看到了这些回报（1000%+）的成果，这正是我发布的内容。 核心观点是，如果你拥有一家像Google Search（例如 $NBIS）这样合法的运营业务，并将其与像YouTube一样的超扩展子公司（例如Avride, Clickhouse）相结合，那就是成为下一个科技巨头的完美组合。
英文原文
That's a fair comment since a lot of people have PTSD with the word FTX and there's a lot of negative associations. (which im not trying to make). However, despite their accounting fraud, they did one thing right out any other startup in the modern era, which was portfolio company growth/investments. And there's some lessons we can take from that when retail does long term investors. We got to see the fruits of those returns (1000%+), in the past 4 years, which was what I posted about. The core point was that if you had a legitmate operational business like Google Search (eg. $NBIS) , and paired it along with Youtube-like hyper scaling subsidiaries (eg. Avride, Clickhouse), thats the perfect mix to becoming the next tech giant.
-
分析超大规模云厂商数据中心模式差异及潜在合作逻辑
我不会像某些特定的 $BMNR 或 $IREN 投资者那样盲目吹捧，认为街上的绿灯就意味着对公司利好。所以我倾向于认为这只是巧合。因为看起来 $AMZN、$GOOGL 倾向于偏好数据中心托管(Colo)模式（因为他们可以插入自己的 TPU、未来的 Trainium 芯片），而 $MSFT、$META（以及 99.9% 使用 $NVDA 且没有现成定制芯片的 AI 公司）则偏好 $NBIS、$IREN 类型的模式。所以 $AMZN 宣布在印第安纳州为其 AWS 数据中心部门投资 150 亿美元。但关键要注意的是，他们此前在那里已经花费了 313 亿美元，所以这并不算全新投入。你可能看到了 $NBIS 正在印第安纳波利斯附近建设 1000 多英亩的绿地数据中心。这只是一个有利于建设的区域，可能并非针对合作伙伴关系或像 Anthropic（最近承诺 400 亿美元资本支出）那样的另一笔超大规模云厂商交易。说实话，我现在可能更相信 Anthropic 的资本支出承诺而不是 OpenAI。话虽如此，也许如果 $AMZN 耗尽容量并达到与 $NVDA 的最大采购订单上限（因为他们正在插入自己的 $CIFR 芯片），他们可能会转而使用 Nebius。
英文原文
I'm not going to be a blind shill like some specific $BMNR or $IREN investors that can say a green light on the street means bullish for the company. So I would lean coincidental. Since it looks like $AMZN, $GOOGL tends to favor colo models (since they can plug in their own TPUs, Trainium chips in the future) while $MSFT, $META (and 99.9% of AI companies that use $NVDA and don't have custom chips readily built out) prefer $NBIS, $IREN type models. So $AMZN announced they're investing $15B in Indiana for their AWS DC segment. But key thing to note is they've already spent $31.3B there before so it's not exactly new. You probably saw how $NBIS is doing a 1000+ acre greenfield DC near Indianapolis. It's just a favorable area for buildout, probably not directed at a partnership and another hyperscaler deal like Antrhopic (who committed $40B in capex spend recently). I'd probably trust Antrophic more than OpenAI right now with capex spend lol. That being said, maybe if $AMZN runs out of capacity and hits max purchase order with $NVDA (since they're plugging in their own chips with $CIFR), they would use Nebius instead.
-
对比NBIS的爆发式增长与CIFR的稳健长期价值。
从我在AI增长板块的观察来看，是的，$NBIS 极其罕见，因为它前身是 Yandex。作为背景，这基本上相当于俄罗斯的百度/中国或谷歌/美国，他们将 Avride 等子公司剥离至 Nebius。至于 $CRWV 等其他公司，它们对 Monolith AI 等收购案更多是为了补充现有核心业务，而非像 $NBIS 那样实现三位数的独立同比增长。而收购 Weights and Bias 花费了17亿美元（据我了解其增长并不快）。至于其他数据中心/矿企，不幸的是比特币增长也不快，但长期来看防御性很强。因此，例如 $CIFR 通过其资产负债表的增长速度会慢得多。但从长期来看，作为股东持有它非常稳健，符合其下行风险较低的数据中心(colocation)模式和较慢的收入爬坡模型。如果你想一年内从0做到2000亿，$NBIS 是做法；如果你想用5年做到，$CIFR 是选择。
英文原文
From what I've seen in the AI growth sector, Yes, $NBIS is incredibly, incredibly rare since it was Yandex before. For context, that's basically Baidu/China | Google/US equivalent but for Russia and they spun off subsidiaries like Avride into Nebius. For stuff like $CRWV and other companies, their acquisitions like Monolith AI were more to compliment their existing core business rather than growing independently at triple digits Y/Y. Then $1.7B for Weights and Bias (that is not growing too fast from what I'm floated numbers) As for the other DC/miner companies, unfortunately Bitcoin isn't growing too fast either, but it's very defensible long term. So growth rate of $CIFR via their balance sheets for example will be a lot slower. But longer term it's really solid to have as a shareholder and goes in line with their low downside risk colo model, slower slower revenue ramp model. If you want to go from 0 to 200B in a year, $NBIS is how you do it. If you want to do that in 5 years time, $CIFR is where it's at.
-
NBIS资金充裕，建议保留高增长子公司并寻求并购以超越当前市值。
$NBIS 应该拥有足够的资金来支持建设，包括 48 亿美元现金储备加上 25 亿美元自动取款机(ATM)融资，以及从收入合同中获得的任何自由现金流(FCF)都将重新投入建设。 我认为 Nebius 管理层指的是如果出现极端的信贷紧缩，他们将出售投资组合中的公司，这在最后关头是合理的。 但他们处于健康的状态。如果我是他们，我会让它们继续增长，并寻找更多的并购(M&A)，就像 FTX 在加密领域所做的那样，或者 Meta 早期在核心运营业务之外对社交网络的投资，以建立更稳固的防御。 如果 Avride 成为一家 250 亿美元的企业，或者 Clickhouse 成为一家 800 亿美元的企业（鉴于它们的增长轨迹，这两种情况都非常可能），这些单独的业务价值将超过当前的市值。
英文原文
$NBIS should have enough to fund buildout through $4.8B cash stack plus $2.5B ATM + any fcf they get from revenue contracts gets put back into buildout. I think Nebius management were referring to if there was extreme credit tightening, they would sell off portfolio companies, which would make sense in a last ditch scenario. But they're in a healthy spot. If I were them I would let them all keep growing, and look for more M&As like what FTX did in the crypto sector, Meta early stage with social media networks alongside their core operational business for more defensibly. In the event Avride becomes a $25B business or Clickhouse becomes a $80B business (which are both extremely plausible scenarios given their growth trajectories), those individual stacks would be worth more than the current market cap.
-
质疑某两家公司仅为NBIS用户而非被投企业
@Dream5728363541 那两家公司难道只是 $NBIS 的用户，而不是其投资组合中的被投企业吗？
英文原文
@Dream5728363541 Aren’t those two just users of $NBIS, not portfolio company investments?
-
博主回复称暂无确切标的，仅分享最接近的参考链接。
@OperationATM 我还没有，但这已经是我目前最接近的了 https://t.co/LDgBku35dG
英文原文
@OperationATM I don’t yet but this is the closest thing I have https://t.co/LDgBku35dG
-
认为市场低估了$NBIS及其子公司Avride的超大规模扩张潜力。
我觉得市场也错过了 $NBIS 的故事。Jim Cramer 就是一个典型的例子。 我认为我们只是处于早期阶段，但最高的回报往往来自于此。尤其是当市场折价子公司，且未将其随核心业务超大规模扩张(Hyperscaling)的因素计入定价时。 存在一种遥远的可能性，即其中一家子公司（Avride）的规模会超过主营业务。
英文原文
I sort of think the market missed the $NBIS story too. Jim Cramer for one is the perfect example. Think we're just early but that's where the highest returns come from. Especially when the market discounts subsidiaries and doesn't price in that they're hyperscaling along the core business. There's a distant possibility where one of them (Avride), becomes larger than the main business.
-
Nebius因AI投资组合及核心业务高增长被严重低估，有望成下一个GOOGL。
Nebius [ $NBIS ] 是目前被低估程度最高的成长型公司。它有望成为下一个市值达 210 亿美元的 $GOOGL。原因很简单：其投资组合公司令人惊叹。最极端的例子是 $FTX。回顾 $META 如何成长为万亿级巨头，不仅靠 Facebook，更靠 Instagram、Whatsapp 等投资组合公司主导了社交媒体格局。$FTX 在数字资产和前沿技术领域做了类似布局。2021 年，$FTX 投入 58 亿美元，重仓三家核心公司：1. Anthropic（13.56%，估值 25 亿）；2. Robinhood [ $HOOD ]（7.6%，估值 85.4 亿）；3. Solana [ $SOL ]（4100 万+代币）。如今，Anthropic 最新估值 3500 亿，该持股价值约 474 亿；Robinhood 市值超 1000 亿，持股价值约 76 亿；Solana 代币 131.5 美元，持股价值超 57 亿。仅这三家公司四年增值超 550 亿，还不包括 FTX 其他数十亿投资及 Chime、Layerzero、Aptos、Hidden Road（被 $COIN 收购）等。其投资组合价值超越了核心业务（试想若核心业务如 $GOOGL 搜索般增长该多好）。$NBIS 现在拥有与 $FTX（加密）、$META（社交）相同的结构，但在人工智能领域，且核心业务真实且快速增长。Nebius 持有：1. Clickhouse（28%，估值约 70 亿，2025H1 营收 6.3 亿）；2. Avride（83%，估值约 60 亿，Uber 领投后）；3. Toloka AI（~65%，估值约 6.4 亿）；4. TripleTen（100%，估值约 3 亿）。Clickhouse 服务于 Anthropic、$META、$TSLA、$NET 等；Avride 是自动驾驶出租车公司，$UBER 投资 3.75 亿以竞争 Waymo；Toloka 是 AI 标注平台，被 Amazon、Microsoft 等使用。投资组合估值合计 76 亿，增速快于多数公开成长股。而其核心业务同比增长 700%+，ARR 达 70-90 亿，现金 48 亿，服务于 $META、$MSFT 等。这可能是其被纳入 MSCI 指数前最后一个月低于 90 美元的机会，将获数亿至十亿流入。相比 $IREN 或 $CIFR，$NBIS 是唯一拥有此类投资组合的数据中心成长股。$NBIS 仅估值 210 亿，市场正在忽视这一机会。
英文原文
Nebius [ $NBIS ] is the most undervalued growth company right now. And it has the potential to become the next $GOOGL at a $21B market cap. There's one simple reason: It's portfolio companies are mindblowing. The most incredible example on this concept is the company $FTX. Here's the story: When we look at how $META grew into a $1T+ company, it wasn't just Facebook. It was their portfolio companies Instagram, Whatsapp, and others made Meta dominate the social media landscape. $FTX was doing something similar, but in digital assets and frontier technologies. Four years ago, in 2021, $FTX invested $5.8B into a large basket of assets. And put a large chunk into these three core companies: 1. Anthropic, 13.56% at a $2.5 Billion valuation. 2. Robinhood [ $HOOD ] 7.6% at a $8.54B valuation 3. Solana [ $SOL ], 41M+ tokens. Fast forward to today, that would have been: · Anthropic in it's latest round is worth $350B. That stake would have been worth ~$47.4B. · Robinhood is now worth over $100B. That stake would be worth ~$7.6B. · Solana is now worth $131.5 per token, making the stake well over $5.7B. Those three companies alone generated well over $55B+ in value in 4 years time, and this is not even including FTX's tens of billions of dollars + hundreds of other investments + holdings in Chime, Layerzero, Aptos, Hidden Road (bought by $COIN), and crypto. Their portfolio companies outlasted their core business (and imagine, how much it would have been worth if the core business kept scaling like $GOOGL search did alongside Youtube). $NBIS now has the same setup as $FTX did in crypto, $META in social media, but in artificial intelligence with a legitimate and incredibly rapidly growing core business. Nebius owns: 1. Clickhouse, 28% at a ~$7B valuation ($6.3 H1 2025) 2. Avride, 83% at a ~$6B valuation. (post Uber raise) 3. Toloka AI, ~65% at a ~$640 million valuation 4. TripleTen, 100% at a ~$300m valuation. · Clickhouse powers Anthropic, $META, $TSLA, $NET, and many fortune 500 companies. · Avride is a self-driving car robotaxi company, spun out of Yandex that $UBER invested in a $375M round in to compete with Waymo. · Toloka is a AI labeling platform that Amazon, Microsoft, Anthropic, and Shopify uses. $1.96B + $4.96B + $416M + $300m = $7.6B valuation in portfolio companies that are growing faster than most public growth companies. But if we look at their core business that is scaling to 700%+ Y/Y to $7-9B ARR, with $4.8B in cash, powering $META, $MSFT, Cursor, governments, and many more... This might be the last month it's under $90 before it receives hundreds of millions to low billions of extra inflows from MSCI inclusion today. There's no other datacenter growth company that has this type of portfolio if we look at crowd favorites like $IREN or $CIFR. $NBIS is only valued at $21B and the market is sleeping on this opportunity.
-
指出LLM初始错误时，集成方法无法纠正自信的错误答案。
@platochi 我关注到了！我上面漏掉的另一点是，没有任何大语言模型(LLM)能在一开始就给出正确答案（即使你个人有时能解释其中的细微差别）。 因此，如果它们都自信地给出了错误答案，采用顾问/集成(ensemble)式的方法也无济于事。
英文原文
@platochi I’m tuned in! Another point i missed above was that no LLM gets it right in the first place (even if you personally explain the nuances sometimes). So having a counsel/ensemble type approach wouldn’t do anything if they all get it confidently wrong
-
构建金融数据接入LLM成本高昂，散户难以获得机构级Alpha生成能力。
嗯，这种方法需要花费数百万美元来构建定制化的金融数据管道（Financial Data Pipelines）接入到大语言模型（LLM）中（我不想花这笔钱）。 这大概是对 @demishassabis、@sama、@elonmusk、@DarioAmodei、@ylecun、@arthurmensch、@aidangomez 或者拥有资源的初创公司来说的一个挑战。 这绝对是一个价值十亿美元的问题：如何让散户也能通过提示词（Prompt）获得与 Jane Street 算法同等的阿尔法（Alpha）生成能力。 我可以开源一个轻量级版本，就像上面的 LLM 委员会那样，仅基于我个人的领域特定判断，但肯定无法做到深入。有趣的事实是，$NVDA 过去曾试图招募我领导他们的人工智能团队之一，所以也许如果我现在没忙于其他事情的话……
英文原文
Uh this approach would cost millions into building custom financial data pipelines into the LLM (which I don't want to spend). Prob a challenge for @demishassabis, @sama, @elonmusk, @DarioAmodei, @ylecun, @arthurmensch, @aidangomez or a startup with the resources. Definitely a billion dollar question for alpha generation to make the same Jane Street algorithm capability available to retail to prompt. I could open source a lightweight version like the LLM council above just based on my own domain-specific judgment, but won't be anywhere close to being in-depth. Fun fact, $NVDA tried recruiting me to lead one of their AI teams in the past so mybe if i wasn't working on other things right now
-
分析11月市场下跌原因：政府停摆、Binance清算、套利交易解除等多因素叠加
谢谢！我想在发帖前先了解一下情况，因为我最初以为是跟降息预期有关、政府停摆导致私人系统流动性被抽走，或者可能是某种算法因素。 政府停摆确实起了很大作用，加上25个基点的降息，但它们不应该导致股票/加密货币下跌30%以上（且重新开放后几乎没有反弹），因为美联储并没有收紧政策，我们刚经历了2次降息，而且很多公司的基本面正在改善。 Binance清算只是加密货币最初的个殊催化剂，但美联储降息而日本加息导致套利交易在流动性重新加载中解除，同时叠加其他并发因素，这可能就是我们现在在这里的原因。
英文原文
Thanks! I wanted to get a grasp of the situation before posting, since originally i was thinking it was it had something to do with rate cut odds, aftermath of gov shutdown draining some private system liquidity, or maybe something algorithmic initially. Gov shutdown does play a large role + .25 bps but they shouldn’t affect stocks/crypto 30%+ (and little recovery after reopening) since the fed isn’t tightening, we just got 2 rate cuts, and fundamentals of a lot of companies are improving. Binance liquidation was just the initial idiosyncratic catalysts fr crypto but fed cutting while Japan hikes causes carry trade unwind from the reload amid the other concurrent factors is prob why we’re here now.
-
博主回复网友，称赞其想法不错。
@sak864208 好主意
英文原文
@sak864208 Good idea
-
高贝塔股常先于指数下跌，是市场风向标。
历史上确实如此，今年也不例外。在2月至4月期间，我们曾看到高贝塔值(高波动性)股票如 $HOOD 大幅回调，成长股及前沿指数也出现大幅下跌。 这并非总是如此，但指数通常最后才发生变动，而成长股/高贝塔值股票通常是煤矿中的金丝雀(市场风向标)。
英文原文
So it actually happened like that historically, and this year too. Before we saw high beta like $HOOD sell off back from Feb to April, and growth stocks frontran index on the massive drop. This isn’t always the case but indexes are usually last to move and growth/high beta stocks are usually a good canary in the coal mine.
-
分析套利交易平仓规模及宏观因素对高贝塔资产的影响
我对2024年套利交易平仓(carry trade unwind)的尸检报告进行了更深入的分析，其中快速资金（对冲基金、自营交易台）的短期资本约为2000亿至3000多亿美元（约占5000亿总额的40-60%左右）。然而，过去一年内这些敞口可能已重新建立，因此类似但规模较小的平仓可能再次发生。日本加息+美联储降息导致已重新建立的套利交易出现平仓。当你提到加密货币溢出到高贝塔(high beta)股票时，如果没有宏观逆风，币安清算导致的比特币跌破9万美元很可能已被买回至10万美元（尤其是考虑到三次降息），所以情况并非如此。
英文原文
So had a deeper breakdown on the 2024 autopsy report from carry trade unwind but figure was ~$200B – $300B+ish in fast money (hedge funds, prop desks) short term capital (which is ~40-60ish of total of $500B) However it's likely been reloaded in terms of exposure in the past year so a similar but less amount might be unwinding again. Japan hiking rates + Fed cutting leads to that carry trade unwind in terms of what got reloaded, When you mention crypto spilling into high beta stocks, if there weren't any macro headwinds, the Bitcoin drop sub $90k from binance liqudations probably would have been bought up to $100k already (esp with 3x rate cut) so that's probably not the case.
-
博主分享偶遇皮卡丘的趣事以安慰粉丝。
@soulbiri1 如果这能让你感觉好受些，我上个月在那里看到了一只胖皮卡丘。https://t.co/svjUkrGf4u
英文原文
@soulbiri1 If it makes you feel any better, I saw this fat Pikachu there the other month. https://t.co/svjUkrGf4u
-
调侃博主阅读速度之快，比日元套利交易规模更神秘。
@itsthesquonky 另外，@Neat_Lama 如何在前 10 秒内阅读这些回复，这比日元套利交易(Yen carry trade)的规模更是一个神秘的问题。
英文原文
@itsthesquonky also how does @Neat_Lama read these responses in the first 10 seconds is the more mysterious question compared to the size of the Yen carry trade.
-
分析2024年表外去杠杆规模，认为当前正重新加杠杆而非触及结构性资金。
大多数事情发生在表外，所以我们无从知晓。 回到2024年，去杠杆(unwind)规模约为2000亿至3000亿美元以上。包括140亿美元期货（CFTC及BIS数据）、1600亿美元对冲基金外汇远期合约（来源：BIS公报第90期），以及约1000亿美元（假设投机性的非银行金融机构(NBFI)贷款部分）。 2万亿美元以上来自GPIF（1.6万亿美元）等结构性资金，1万亿美元以上来自银行贷款等。 2024年已经清理了大量去杠杆，但未触及结构性部分。 但我的看法是，我们很可能自2024年以来正在重新加杠杆(reload)，并看到那约2000亿美元的去杠杆规模再次被清理，而非触及结构性资金。
英文原文
Most things happen off the books so we won't know. So back in 2024 unwind was about $200B – $300B+. $14 Billion in futures (CFTC & BIS data), $160 Billion from Hedge Fund FX Forwards, Source: (BIS Bulletin #90), + ~$100B (Portion of NBFI loans assumed speculative). $2T+ is structural money from GPIF ($1.6T), $1T+ from Bank lending, etc. 2024 cleared a lot of the unwind already but didn't touch the structural part. But my take is that we've likely seeing a reload since 2024 and seeing that ~$200B-ish figure be cleared out again rather than touching structural money.
-
博主复盘整体盈利，详述NBIS、META等持仓盈亏及FLY长期逻辑。
总体来看我盈利颇丰。我刚才指的是上个月，当时像 $NBIS 这样的许多收益都被抹平了。 目前我在 $NBIS 上的持仓成本平均在 90 多美元，这让我感到意外。 $META 在最近几次看涨期权（option calls）交易中的成本均价为 640 美元，但由于我做了看涨期权，该头寸目前亏损六位数。 但我曾在 3000 美元时买入 $BTC，在 14 美元左右买入 $RKLB，在 20 美元时买入 $HOOD。 关于 $FLY，正如我之前的论点所述，这是基于 2026/2027 年中型有效载荷（medium-lift payloads）催化剂的投资。无论短期股价如何波动，都不会改变中型载荷是否研发成功的事实。 如果一年后他们最终未能研发出中型载荷且股价崩盘，你可以嘲笑我。
英文原文
Overall I'm up a lot. I was just referring to this past month where many of my gains like $NBIS were wiped out. I'm below cost average right now on $NBIS in the $90s, which was surprising. $META was $640 cost average on recent calls but I'm down 6 figures on that position since I did option calls. But I bought stuff like $BTC at $3k, $RKLB around $14 and $HOOD at $20. On $FLY as I said before in my thesis, it was a 2026/2027 catalyst play on medium-lift payloads. Whatever the stock price short term doesn't really change whether medium-lift developed. You can laugh at it if in a year if they end up failing to develop medium lift and stock price craters.
-
宏观分化加剧，但股票基本面优于宏观逆风。
这是一个价值百万美元的问题，也说明了在全球市场高度互联的背景下，宏观环境有多么复杂。通常，三次降息是极其看涨的，往往会导致次年 $SPY 和 $QQQ 指数获得两位数涨幅。然而……此时日本央行(BOJ)正在加息。但如果美联储降息而日本央行加息，就会出现分化：美元融资成本更低，日元融资成本更高 -> 巨大的利差变化，这可能会加速日元套息交易(Yen Carry Trades)的平仓。我不知道确切答案。但我推测，我们可能已经看到了这些恐慌情绪的最坏阶段，许多高贝塔(High-Beta)股票在日本央行加息前提前下跌了40%。而且抛售导致许多像 $NBIS 或 $META 这样的股票脱离了基本面。所以有一点是肯定的：股票基本面 > 宏观逆风。
英文原文
So that is a million dollar question and speaks how to incredibly complex macro is given how interconnected markets are globally. Normally triple rate cut is extraordinarily bullish and normally leads to double digit $SPY $QQQ index gains the following year. However... this comes at a time where BOJ is raising rates. But If the Fed cuts and BOJ raises, you get a divergence: cheaper USD funding, more expensive JPY funding -> big differential changes, which could accelerate unwinding of yen carry trades. I don't know the answer. But my speculation is we've likely seen the worst of these fears play around where many high-beta stocks dropped 40% preemptively in advance of BOJ rate hike. And the sell-off has led many stocks like $NBIS or $META to be detached from fundamentals. So one thing is for sure though: fundamentals of stocks > macro headwinds.
-
高确信度组合本月回撤，预期12月降息后反弹创新高。
@itsdsha 只能说，对于我高确信度组合中的 $ALAB、$HOOD、$IBIT、$META、$NBIS、$RKLB 和 $TSM 来说，这一个月确实不好过。预计随着12月降息落地，一个月后情况会好转。大多数短期保证金/期权交易者已被清算，而此类事件往往推动市场创出新高 https://t.co/YCzLLaArKa
英文原文
@itsdsha Let's just say it hasn't been a fun month for my high conviction port of $ALAB, $HOOD, $IBIT, $META, $NBIS, $RKLB, and $TSM for sure. Should be fine in a month come Dec rate cut. Most short term margin/option traders got liquidated, and events like these lead to higher highs https://t.co/YCzLLaArKa
-
宏观因素致抛售，预计周一反转并抢跑12月降息。
所以我提到的四个宏观要点是此次抛售的更大原因。上周五有大量期权头寸到期，我猜测在 $NVDA 财报后，这纯粹是由算法驱动的（而且至今仍说不通）。然而，导致更长期抛售的主要原因包括：由币安(Binance)引发的连锁加密货币保证金清算、降息概率变化、AI信贷压力以及日元套利交易(Yen carry trade)。我猜测市场将在本周一反转，并在市场迫使美联储(Fed)降息后重新定价并抢跑12月降息（且在期权到期后）。不过，我不确定日元套利交易的抛售会在多大程度上抵消这一影响。
英文原文
So the four macro points I talked about was the larger reason for the selloff. There was a huge amount of option interest expiring last Friday, and my guess post $NVDA earnings was that it was purely algorithmic driven (and still doesn't make much sense). The larger overhang about the cascading crypto margin liqudations caused from Binance, Rate cut odds changing, AI credit stress, and yen carry trade are the main reasons behind the longer sell-off though. My guess is market reversal this Monday and repricing in + frontrunning December ratecut after markets bulled Fed into cutting (and after option expirations. However, I don't know how much possible Yen carry trade selloff offsets this.
-
博主回应杜罗夫，认为其言论增加了法国政府参与暗杀的可能性。
@durov 我通常从不评论政治，但你关于法国政府参与查理·柯克(Charlie Kirk)暗杀事件的评论，确实大大增加了这种可能性的可信度。
英文原文
@durov I normally never comment on politics but having you comment about French Gov's involvement in Charlie Kirk's assassination just lent a lot of weight to the possibility.
-
MSTR/BMNR 表现糟糕，ETH 高位做多风险大且网络用量不等于价格。
@a_little_tailor 拉大时间周期看，$MSTR 的情况甚至更糟。$BMNR 本月也遭重创，下跌 -47.1%。 为做多者感到难过，$ETH 在 $4k 以上做多非常糟糕，在 $4-5k 的顶部跟随 @fundstrat 而不是在 $1.5k 时跟随，这很不幸。 人们往往将 ETH 网络使用量与价格混为一谈。https://t.co/zdO3e8YsTC
英文原文
@a_little_tailor Zooming out is even worse for $MSTR. $BMNR also got nuked -47.1% this month. Feel bad for folks that went long, $ETH is a terrible long above $4k and following @fundstrat at the $4-5K top than at $1.5K is unfortunate. People tend to conflate ETH network usage with price. https://t.co/zdO3e8YsTC
-
强调投资应追求收益与权重的双重最大化。
@RJCcapital 正确的座右铭是： 最大化收益，最大化收益。 （市场，权重） https://t.co/e5oKDsYcz7
英文原文
@RJCcapital The correct motto is: Maximize Gains, Maximize Gains. (markets, weight) https://t.co/e5oKDsYcz7
-
四大流动性冲击引发高杠杆资产暴跌，但AI巨头基本面未变。
市场刚刚经历了近代史上最严重的去杠杆冲击。 高贝塔资产正在崩溃： • $NBIS, $IREN, AI 股票从高点下跌约 40%+ • $MSTR, $BTC 在过去一个月暴跌 30-40%+ • 加密货币市值蒸发超过 1.2 万亿美元 这不正常，这是针对市场上增长最快且杠杆最高的板块的强制平仓。 以下是同时发生的四个流动性流失因素： 1. 加密货币前兆冲击（2025 年 10 月 – Binance 故障） • Binance 的定价错误导致 USDe 跌至 0.65 美元，在 24 小时内触发了超过 190 亿美元的强制清算 • 使用 USDe/wBETH/BNSOL 的高杠杆头寸（25x–50x）引发了全行业连锁保证金清算的连锁反应 • 该事件使加密货币结构脆弱，波动性现在威胁到流入 $MSTR/国债股票超过 70 亿美元的 MSCI 资金，存在 BTC/NAV 错配和强制出售比特币的风险 2. 美联储政策不确定性（降息鞭打效应） • 在美联储发出矛盾信号后两天内，市场对 12 月降息预期从 97% → 35% → 70%+ • 这种波动起到了隐性紧缩的作用，迫使杠杆基金和算法（在 $NVDA 财报后看到的情况）提前去杠杆 • 政策模糊性提高了全球风险，将美国的不确定性传导至全球融资市场的广泛抛售 3. AI 信贷压力（投机性债务破裂） • AI 建设需要 3.5 万亿美元的外部融资，促使公司大量进入债务市场 • Google + XAI 在 Similarweb 上的使用量提升引发了人们对 OpenAI 和循环融资的担忧，以及 $1T+ 的资本支出流向 $ORCL, $CRWV, $AMD 等公司，而这些公司并没有必要的资本。 • Applied Digital [ $APLD ] 23.5 亿美元垃圾债券（评级 B+）暴露了交易对手和集中度风险，由于需要更多债务来资助建设，其股票及相关数据中心股票下跌。 • 资本市场开始区分拥有真实现金流的公司和依赖投机性债务及 OpenAI 合同的公司，惩罚了如 $ORCL 和 $CRWV 等标的。 4. 日元套利交易平仓（催化剂） • 日本央行加息至 0.5% 及正常化缩小了美日利差，挤压了 80 万亿日元（约 5000 亿美元）的杠杆头寸 • 日本国债收益率上升引发回流资金，给美国收益率带来压力并抽干全球流动性 • 强制平仓导致投资者出售美国科技股和加密货币以偿还日元贷款，加剧了避险情绪的连锁反应 底线： 所有四个冲击都从市场的同一角落抽干了流动性——高贝塔、高杠杆资产，导致成长型科技和加密货币剧烈去杠杆，并增加了进一步强制抛售的风险。 过度的杠杆、脆弱的资产负债表和集中度风险决定了哪些资产跌幅最大。 美联储开启新一轮降息周期是一个“拐点”，投资者需要判断这种宽松是预防性措施还是对更严重衰退的反应。然而有一件事是确定的： 宏观冲击重置了估值，但并没有改变从 $NBIS 到 $META 的企业基本面。
英文原文
Markets just suffered their worst deleveraging shock in recent history. High-beta assets are collapsing: • $NBIS, $IREN, AI stocks are down ~40%+ from peaks • $MSTR, $BTC plunged 30-40%+ in the past month • Crypto erased over $1.2 trillion in value This wasn’t normal, it was a forced unwind across the markets with the most growth and leverage. Here's the four liquidity drains hitting at once: 1. Crypto precursor shock (Oct 2025 – Binance failure) • A pricing error on Binance sent USDe to $0.65, triggering $19B+ in forced liquidations in 24 hours • Highly leveraged positions (25x–50x) using USDe/wBETH/BNSOL cascaded into a chain reaction of cascading margin liquidations across the industry • The event left crypto structurally fragile, and volatility now threatens $7B+ MSCI inflows into $MSTR/treasury stocks, risking BTC/NAV mispricing and forced Bitcoin sales 2. Fed policy uncertainty (rate cut whiplash) • Markets priced a December cut at 97% → 35% → 70%+ in two days after conflicting Fed signals • This volatility acted as a stealth tightening, forcing leveraged funds and algorithms (seen post $NVDA earnings) to deleverage pre-emptively • Policy ambiguity raised global risk, transmitting U.S. uncertainty into broad selling across global funding markets 3. AI credit stress (speculative debt cracks) • AI build-out requires $3.5T in external financing, pushing companies heavily into debt markets • Google + XAI raise in usage per similarweb raised concerns about OpenAI and circular financing, alongside how $1T+ in capex spend going into $ORCL, $CRWV, $AMD, and others without having the ncessary capital. • Applied Digital [ $APLD ] $2.35B junk bond (rated B+) exposed counterparty and concentration risk, sending its stock and related data center stocks down due to the need of more debt to fund buildout. • Capital markets began differentiating between firms with real cash flow and those reliant on speculative debt and OpenAI contracts, punishing names such as $ORCL and $CRWV. 4. Yen carry trade unwind (the catalyst) • BoJ rate hikes to 0.5% and normalization narrowed the U.S.–Japan rate gap, squeezing ¥80T (~$500B) in leveraged positions • Rising JGB yields triggered repatriation flows, pressuring U.S. yields and draining global liquidity • Forced unwinds led investors to sell U.S. tech and crypto to repay yen loans, amplifying the risk-off cascade Bottom line: All four shocks drained liquidity from the same corner of the market, high-beta, leveraged assets, driving a violent unwind in growth tech and crypto and raising the risk of further forced selling. Excessive leverage, fragile balance sheets, and concentration risk determined which assets crashed the hardest. The beginning of a new rate-cutting cycle by the Fed is an "inflection point" as investors need to determine whether the easing is a preventative measure or a reaction to a more serious downturn. However one thing is for certain: The macro shock reset valuations, but it didn’t change the fundamentals of businesses from $NBIS to $META.
-
博主吐槽平台误判机制，拟删除推文。
@JustUrAvgStonk 哈哈，我大概会删掉这条并换个内容发，因为每个正常人都被标记成了机器人，而唯一没被标记的却是个机器人。https://t.co/xoxJfE0aRo
英文原文
@JustUrAvgStonk lol prob gonna take this down and post something else since every normal person got flagged as a bot and the only person who didn’t was a bot. https://t.co/xoxJfE0aRo
-
CRWV 虽发行条款优，但高债高息致利润受损，评级为 F。
$CRWV 完全配得上 F 级评级。如果从发行条款(offering point)的角度看，它确实是一流的。但从资产负债表(balance sheet)来看，简直是一场噩梦，尤其是背负着 140 多亿美元的债务，加上为 OpenAI 建设所承担的 10 多亿美元债务利息，而 OpenAI 能否偿还尚不确定。此外，他们仅剩 19 亿美元现金。大量利润被用于支付债务利息，这严重损害了利润率(margins)（相比之下，其他通过可转换票据(convertible note)融资的新兴云厂商(neoclouds)没有这种利息负担）。
英文原文
$CRWV is well deserving of F tier. If we go from an offering point of view, then it’s top of the line. Balance sheet wise, it’s a nightmare, especially with the $14B+ in debt plus the $1B+ debt interest for the buildout for OpenAi which may or may not be able to pay. They also have $1.9B cash left. A lot of the profit goes off to paying debt interest which hurts margins a lot (compared to other neoclouds that don’t have that interest since they did convertible note).