$AMZN
相关推文
按时间倒序
-
长期更偏好英伟达上游瓶颈而非NVDA本身。
回复 @darkseidzz：嗯，长期看，相比做多 $NVDA，我更喜欢你所有的上游瓶颈环节，因为这些会被重估得最多（英伟达已经是世界最大公司）。 我很确定，超大云厂商 ASIC 最终会分流 $NVDA 的需求，比如 $GOOGL 的 TPU、$AMZN 的 Trainium 项目。
英文原文
@darkseidzz hmm, i prefer all your upstream chokepoints over $NVDA long term since those will be re-rated the most (nvidia already largest company in the world) pretty sure hyperscaler ASICs would eventually siphon off $NVDA demand like $GOOGL TPU, $AMZN trainium programs. wouldn't be too
-
高盛上调四大云厂商资本开支预期，利好上游半导体供应链。
是的……我认为你们所有上游半导体供应链公司都会涨得更高。 高盛现在预计，2025到2030年四大超大云厂商 $GOOGL / $META / MSFT / $AMZN 的资本开支合计将达到5.3万亿美元。 这一数字从一季度财报后的4.5万亿美元上调。
英文原文
Yeah… I think all your upstream semi supply chain companies are going much higher. Goldman now expects a combined $5.3 trillion of capex spending for the four largest hyperscalers $GOOGL / $META / MSFT / $AMZN from 2025 to 2030. Revised up from $4.5T from Q1 earnings. https://t.co/UtkU1IU2V5
-
讨论高通与字节/亚马逊相关机会及光子学收购。
回复 @madridraptor：如果我没记错，$QCOM 是字节跳动，也可能是 $AMZN，不是 $GOOGL TPU。 不过我真的不喜欢高通对中国的敞口有多大。但他们做了一些非常重要的光子学收购，比如 Alphawave（类似 Marvell 收购 Celestial）……
英文原文
@madridraptor $QCOM was Bytedance and maybe $AMZN if i remember correctly, not $GOOGL TPU. I'm really not a fan of how much exposure Qualcomm has to China though. But they made some really important photonics acqusitions like Alphawave (similar to what marvell did with celestial)... for
-
称AAOI是当前美国市场最喜欢的光子学敞口。
$AAOI 实际上是我现在在美国市场最喜欢的光子学敞口。 去年我在28美元以低仓位做多，当时我猜它们正在与 $AMZN 和 $MSFT 认证。 业绩后约70美元时我有高信念，因为它们宣布了1.6T 和其他批量订单……
英文原文
$AAOI is actually my favorite photonics exposure in the US market right now. I went long last year with low sizing at $28, back when I guessed they were qualifying with $AMZN and $MSFT. High conviction post earnings at ~$70, when they announced 1.6T and other volume orders with
-
Ayar与Wiwynn公告对SIVE CPO机架级部署潜力重大。
Ayar 今天与 Wiwynn 的公告，对 $SIVE 在 CPO -> 机架级部署方面可能非常重要。 Wiwynn 的云客户包括 $AMZN、$META、$MSFT。 而且他们一直在谈 $GOOGL TPU 部署。 我认为，仅供参考架构来看，每个机架大约需要512+个 SuperNova 光源。 所以如果 $SIVE 是主要激光阵列供应商（我们预计如此，因为 Macom + Lumentum 已从 Ayar 网站移除）。 即使是中等规模的机架部署，也会对收入非常有意义。 这只是 $SIVE / Ayar / Wiwynn 当前机架级商业化潜力，还不会体现在收入财务中。
英文原文
Ayar’s announcement today with Wiwynn is potentially very material for $SIVE regarding CPO -> rack scale deployments. As Wiwynn cloud clients include $AMZN, $META, $MSFT. And they’ve been in talks for $GOOGL TPU deployments. I think just for some reference architectures it’s around 512+ supernova light sourc a rack. So if $SIVE is the primary laser array supplier (which we expect, given Macom + Lumentum was removed from Ayar’s site). Even modest rack deployments would be very meaningful for revenue. This is just rack scale commercialization potential right now from $SIVE / Ayar / Wiwynn, which won’t show up in revenue financials yet.
-
MRVL财报验证CPO主题和Celestial放量，间接指向SIVE激光需求。
$MRVL 财报是更广泛 CPO 主题的看多信号（也包括 $SIVE 作为可能的激光供应商）。 - “Scale-up interconnect 是 AI 基础设施中最新、最具战略意义的新兴机会之一。” CPO 主题 go brrr。 - 确认 Celestial 被 T1 超大云厂商选中用于 scale-up。 我过去发现过 Celestial 的 $AMZN 认股权证……所以很可能是 Amazon。 - 明年带 Celestial 的 scale-up 光学收入应该超过此前约1.5亿美元预期的2倍。 前瞻收入爬坡预期 go brrr。 - Celestial 团队加上 $MRVL 光学团队是“全垒打”。 Marvell 把 Celestial 视为增长向量，上游 Celestial 供应商受益。 - $MRVL 现在专注于把 Celestial 推向高量制造。 这是放量指标。 如果你不记得，此前有 OSINT 映射显示 $SIVE 直接对应 Celestial，而不是通过 $POET。 所以 Celestial 的前瞻增长，是 Sivers 激光器的放量指标。
英文原文
$MRVL earnings were a bullish indicator on the broader CPO theme (and $SIVE as the likely laser supplier). - “Scale-up interconnect represents one of the newest and most strategically important opportunities emerging in AI infrastructure.” CPO thematically go brrr - Confirmation Celestial was selected by T1 hyperscaler for scale up. I’ve found Celestial $AMZN warrants in the past)… so probably Amazon. - Scale-up optics revenue next year should be more than 2x prior ~$150M outlook with Celestial Forward revenue ramp expectations go brrr. - Celestial team plus $MRVL optics team was a “home run”. Marvell sees celestial as growth vector, upstream celestial suppliers go brr - $MRVL is now focused on bringing Celestial to high volume manufacturing. Volume ramp indicator If you don’t recall, there was OSINT mapping $SIVE directly to Celestial, not through $POET. So Celestial forward growth is a volume ramp indicator for Sivers lasers.
-
回顾AAOI/AXTI早期被FUD，如今类似CPO转折正在SIVE/Foci上重演。
当我在 $AAOI 约20-30美元做多时： 我以为 $AMZN 和 $MSFT 正在为其 ASIC 项目认证特定光收发器。 结果它更可互换、更可大规模生产。 无论如何，很高兴我关于 $LITE、$AAOI、Innolight、$COHR、$AXTI 的 thesis 表现这么好。 记住：当 $AAOI 在20-30美元时，X 上所有人都说“管理层诈骗”…… 或者在 $AXTI 12美元时说“诈骗公司”。 FUD 非常极端。 现在感觉又像似曾相识……做多下一轮 CPO 架构转变，比如 $SIVE 或 Foci？然后收到同样评论。 我们会看看我的 CPO 多头是否会像 Shunsin 对 Innolight、或 $SIVE 对 $LITE 那样兑现。 无论如何，那些名字现在都已经上涨数百到数千个百分点。
英文原文
Back when $AAOI was ~$20-30 when I went long: I thought $AMZN and $MSFT were qualifying specific optical transceivers for their ASIC programs. Turns out it’s more interchangeable/mass producible. Regardless, glad my thesis on $LITE, $AAOI, Innolight, $COHR, $AXTI played out so well. Keep in mind: everyone on X was saying “scam management” with $AAOI back at $20-30… Or “scam company” with $AXTI at $12. FUD was pretty extreme. Feels like dejavu again… going long on the next CPO architectural shift like $SIVE or Foci? And getting the same comments. We’ll see if my CPO longs play out the same way like Shunsin does with Innolight or $SIVE does with $LITE. Either way all of those are now up hundreds to thousands of percent.
-
认为只要 thesis 不破，公司就能继续成长
我不确定大家有没有意识到，只要一个 thesis 没有彻底被证伪，像 $NBIS 这种公司就可以继续成长。 看看过去 15 年的 $AMZN 或 $GOOGL 就知道。 如果你老是“减仓”，往往会先触发税负。很多回撤其实还没大到抵消你已经交掉的税。 等所谓“50% 暴跌”真的发生时，价格可能早就已经复利了几百甚至几千个百分点。 如果你需要花钱，只要仓位到了七位数、八位数甚至九位数，你完全可以借着这些资产去融资，同时继续让它们升值。 NFA，只是我个人观点。你们自己随意，但这非常、非常取决于你选的公司。 像 $IREN 这种垃圾就没法这么做。不过我确实认为 $NBIS 已经站在成为下一个 hyperscaler 的位置上了。
英文原文
Not sure if people realized this but unless a thesis completely breaks, companies like $NBIS can keep growing. Just look at $AMZN or $GOOGL over the past 15 years. If people "trim" it often triggers taxes. And a lot of corrections are typically less than those taxes paid. By the time a "50% crash happens", it's probably already compounded hundreds of even thousands of percent. If people need to pay expenses, once you hit 7-8-9 figures, you can always borrow against those assets and keep letting them appreciate. NFA, just personal opinion. You all do you, but it's highly, highly, dependent on the companies you pick. Can't do this with something trash like $IREN. But I do believe $NBIS is positioned to be the next hyperscaler.
-
把 SIVE 和 Ayar / AlChip / Amazon Trainium 链条连起来
@DanilSer33 新的 $SIVE / Ayar / AlChip / $AMZN Trainium 供应链，比最近那个 $POET 5000 万到 5 亿美元采购订单更值得看多。 市场大概率不会立刻看出 Ayar、AlChip 和 $SIVE 之间的联系，除非有人把它点出来。
英文原文
@DanilSer33 The new $SIVE / Ayar / AlChip / $AMZN trainium supply chain is more bullish than the recent $POET $50m -> $500m purchase order. Markets probably won’t see the connection between Ayar AlChip and $SIVE unless someone points it out though.
-
补充 Amazon / AlChip / Ayar 对 SIVE 的利好链条
再补一句关于很正面的 $SIVE 消息，以及 MSCI 未来两周可能带来数千万资金流入： $AMZN 刚和 AlChip 做了新的私募配售。 这大概率意味着未来 Trainium 还有设计赢单。 如果你还记得的话……AlChip 是 Ayar 的主客户。而 $SIVE 是 Ayar 的主要激光供应商。 所以对 $SIVE 的含义会非常大，它是在搭 Amazon 生态增长的顺风车。
英文原文
Just to add to the very positive $SIVE news amid MSCI likely tens of millions of inflow in 2 weeks: $AMZN has a new private placement with AlChip. Probably implying design wins with future Trainium. If you don’t remember… AlChip was Ayar’s lead customer. And $SIVE is the primary laser supplier to Ayar. So implications for $SIVE, is enormous piggybacking off of Amazon’s ecosystem growth.
-
发现 Nextronics 也是机器人供应链里被忽视的供应商
现在再想想…… Nextronics（8147）其实也是 Amazon 机器人供应链里一个很被低估的供应商。 所以 Amazon 越扩张，他们的收入也会越大（市值才约 2 亿美元，如果毛利率真能做到 38-40%，那影响应该会很明显）。 我之前只盯着 $NVDA 的 CPO 供应链，因为那是它们最大的增长向量。 但我也认为 Amazon 的机器人计划是最适合大规模生产的路线，因为每造一个内部机器人，就会降低 opex、削减 headcount、提升利润率。 而且整个生态都会受益。
英文原文
Now that I think about it more… Nextronics (8147) is a pretty undiscovered supplier to robotics supply chains like $AMZN too. So as Amazon scales up, so does their revenue (it’s only ~$200m mc so it should be material if they’re hitting 38-40% gross margins). I was only focusing on $NVDA CPO supply chains as their largest growth vector earlier. But I do think Amazon’s robotics program have the simplest route for mass production since each one they make internally lowers opex, cuts headcount, and improves profitability. And their whole ecosystem should benefit.
-
认为 Amazon 在机器人领域被严重低估
@Dog_Ziller 我其实觉得 $AMZN 的机器人业务被严重低估了。 因为它立刻就能通过自动化降低 opex 和 headcount，带来实际收益。 至于通用人形机器人，$TSLA、Unitree、Figure、Boston Dynamics、Agibot、Agility 这些名字应该都会表现不错。
英文原文
@Dog_Ziller I actually think $AMZN for robotics is extremely underrated. Since since there’s immediate practical upside to lower opex + headcount for automation. But for general purpose $TSLA, Unitree, Figure, Boston Dynamics, Agibot, Agility, and others should do well.
-
解释为什么每条帖子都会有人追问同样的问题
我不明白为什么每一条帖子下面，从 $AAOI 到 $LPK，大家都会问这种问题？ 我已经说过很多次了，$FLNC 宣布拿下两个新的 hyperscaler 合同之后，我是有仓位的。 但我确实觉得 $FLNC 很有吸引力，因为从 $GOOGL 到 $AMZN，hyperscaler 一般不会签小单。 尤其是看 $MSFT 和 $NBIS 的合同，很多内容都让人非常意外。
英文原文
Not sure why people ask these types of questions across every single stock from $AAOI to $LPK under every post? I've already said multiple times, I have positions after $FLNC earnings announced 2 new hyperscaler deals. But I do think $FLNC Is compelling since hyperscalers from $GOOGL to $AMZN don't sign small deals in general. Especially when you look at $MSFT contract with $NBIS a lot of it is a major surprise.
-
认为 FLNC 应该更高估值
我其实觉得 $FLNC 应该更值钱。 一个季度里接连拿到 2 个直接 hyperscaler 合同，这个意义非常大。 $MSFT 到 $AMZN 不会签小单。 当然市场通常会等更多实际新闻 / 采购订单数字出来，怕它最后没落实或者比预期低。 但一家公司的确不会无缘无故宣布 2 个 hyperscaler MSA，还说订单会在 Q3 到。 而且一个季度赢得多个 hyperscaler 合同，本身就是后续更多合同的领先指标，尤其是 Fluence BESS 越来越标准化的时候。 考虑到空头比例大约 27.69%，我不确定财报前的空头是否真的敢冒这个险…… 如果像 $GOOGL 这样的 hyperscaler 签下大合同，我觉得有机会来一波历史级上涨。
英文原文
I actually think $FLNC should be a lot higher. The implications of having 2 incoming direct hyperscaler contracts in 1 quarter is enormous. $MSFT to $AMZN don't sign tiny deals. Obviously markets like to wait more until actual news/purchase order numbers come out... in the off-chance it doesn't go through or lower than expected. But a company doesn't just randomly announce 2 hyperscalers MSas and an expectation of the orders to hit Q3. Also winning multiple hyperscaler deals, in a single quarter is a leading indicator for more, especially as Fluence BESS becomes standardized. Given short interest is around 27.69%, I'm not sure if pre-earnings short sellers are very comfortable to take a risk... I think there's a chance for a generational run if a hyperscaler like $GOOGL signs a massive contract.
-
介绍 Nextronics 及其在 Nvidia CPO 供应链里的位置
我只是把 Goldman Sachs 的 $NVDA 供应商报告放出来。 有一家很有意思、约 2.1 亿美元市值的公司 Nextronics（8147），我跟着 GS 之后也建了仓。 它向 Nvidia 的 CPO 供应链提供 CPO connectors 和 Cage Thermal Modules。 它们也在 $AMZN 的供应链和 humanoids 里，顺带还做按摩椅。 我只是想分享一个有意思的想法（NFA / DYOR），我之后会做 BOM 分析，但在 Nvidia CPO 项目放量时，这相对市值看起来很有意义。 当然也有多供应商 / 被设计出局的风险，但 GS 多次把这家微型供应商列在 40 亿到 100 亿美元以上公司名单里，可能是有原因的。
英文原文
Just putting it out there with the Goldman Sachs $NVDA supplier note. There’s a very interesting ~$210m MC company Nextronics (8147) that I ended up taking positions on following GS. That supplies CPO connectors and Cage Thermal Modules to Nvidia CPO supply chains. They’re also in $AMZN supply chains and Humanoids. And massage chairs too. Just thought I’d share an interesting idea (NFA/DYOR), I’ll do a BOM analysis later, but it looks very material relative to MC as Nvidia’s CPO program scales up. Of course risks are multi-sourcing/getting designed out, but there’s probably a reason why GS flagged this micro supplier multiple times among $4-10B+ companies.
-
AAOI被比作光学版Intel，借超大规模需求东风，2027年收入预测显示严重低估。
我之前在 $AAOI 市值 $6.49B 时就说过同样的话。 现在市值 $11.5B，故事还是一样的： 基本上这就是光学(Optics)版的 $INTC，他们正在将产能从激光晶圆制造扩展到美国本土组装。 他们生产的任何东西都可能被超大规模云厂商(Hyperscalers)收购。 如果看 2027 年中期预测的 $1.41B 季度收入增长（$471 x 3）与当前市值的对比，当前的财务数据其实不太重要。 以及高盛预测的 9 倍 TAM 增长曲线。
英文原文
I’ve said this before when $AAOI was $6.49B. Now it’s $11.5B and this is the same story: Basically this is the Optical version of $INTC, where they’re scaling capacity from laser fab to assembly Made in America. Anything they make is likely to be bought out by hyperscalers. Current financials don’t really mater if you look at $1.41B quarterly revenue ramp by mid 2027 projections ($471 x 3) vs current MC. And the overarching 9x TAM curve from GS projections.
-
认为 FLNC 的两份 hyperscaler 合同会压制空头
@pennycheck 对，不太明白为什么很多基金喜欢做空美国能源公司的方向性。 未来几个月内如果这两份 hyperscaler 合同落实，虽然数量还不清楚，但对他们来说大概会很吓人。这样可能会让他们逐步降低 $FLNC 的风险敞口，反而增加买盘。
英文原文
@pennycheck Yeah not sure why many funds like directionality shorting US energy companies... Having 2 hyperscaler deals convert sometime in the next few months with unknown figures is probably scary for them. Would probably make them derisk $FLNC over time, adding to buying pressure.
-
Sivers处CPO供应链前沿，910亿美元TAM蓝海刚起步，H2超级周期可期。
只是近期信息发现的回顾 + 与$SIVE的可能映射： -> $JBL 1.6T -> Lightmatter -> Ayar -> $MRVL Celestial -> Lightelligence -> $POET -> $GFS生态 -> $AMD CPO -> O-Net / Enablence -> $AAPL 硅光子学 _ -> $YSS Golden Dome/DoD -> $RTX / $ERIC (太空) -> Bae Systems -> $AEVA $JBL到Ayar供应给超大规模云商如$MSFT、$GOOGL、$AMZN、$META。 Lightelligence到O-Net可能供应给亚洲超大规模云商如腾讯、字节跳动和百度。 在此之上……根据高盛报告，CPO的整体TAM（总可寻址市场）从0增长到$91B。 而Sivers恰好是CPO领域的前沿（也从0开始）。 这绝对是高贝塔和高波动的。 但如果Win卷随着$SIVE一起增长，我看到他们明年都能成为$10B+的公司。 这只是CPO超级周期H2 2026开始之前非常早期的阶段（H1）。
英文原文
Just a recap of recent information discovery + likely mapping with $SIVE: -> $JBL 1.6T -> Lightmatter -> Ayar -> $MRVL Celestial -> Lightelligence -> $POET -> $GFS ecosystem -> $AMD CPO -> O-Net / Enablence -> $AAPL Silicon Photonics _ -> $YSS Golden Dome/DoD -> $RTX / $ERIC (Space) -> Bae Systems -> $AEVA With $JBL to Ayar feeding into hyperscalers like $MSFT, $GOOGL, $AMZN, $META. With likely Lightelligence to O-Net feeding into Asian Hyperscalers like Tencent, Bytedance, and Baidu. On top of that... the overarching TAM with CPO from the GS report goes from 0 -> $91B. And Sivers happens to be the bleeding edge for CPO (also starting from 0). This is definitely high-beta and volatile. But if Win volume ramps alongside $SIVE, I see them both becoming $10B+ companies next year. This is just extremely early on (H1) before the CPO supercycle starts H2 2026.
-
超大规模云厂商 capex 仍然很强
我已经帮你把情况盯好了。 超大规模云厂商 capex 支出的 TLDR（除了 $TSM 之外的信号）。 来自 $MSFT、$AMZN、$META 的财报： 上游半导体供应链会继续起飞。 $META：2026 财年 capex 为 1150 亿至 1350 亿美元 修正后的 Q1 财报：capex 提高到 1250 亿至 1450 亿美元。 -> 更高的零部件价格（涨价） -> 激进的 AI 基础设施支出 -> 定制模型训练 $MSFT：实际为 319 亿美元，而不是（352.9 亿美元），但原因是供应链瓶颈，而不是资源不足。 我们还在等 $AMZN 的确认，不过他们今年年初的原始 capex 预期大约是 2000 亿美元，来自 Jassy 的备忘录。 所以你可以放心，所有半导体供应链名字在下个季度大概率还是会继续起飞，因为有太多 capex 被灌进去。 然后还有新任美联储主席，他对 AI 和降息都非常积极，而 Jerome 在后面保持平稳（市场大概会喜欢这个）。
英文原文
I've monitored the situation for you. TLDR on hyperscaler capex spend (signal aside from $TSM). From $MSFT, $AMZN, $META earnings: Upstream semi supply chains go brrr. $META: 2026 FY capex $115-$135B Revised Q1 ER: $125B-$145B capex raised. -> Higher component prices (price hikes) -> Aggressive AI infra spending -> Custom Model Training $MSFT: was actually $31.9B vs ($35.29B) but due to supply chain bottlenecks, rather than lack of resources. Still waiting for confirmation around $AMZN but their original capex projection was ~$200B from Jassy's note earlier this year. So you can sleep easy, all the semi supply chain names still likely to keep going brr next quarter, since there's so much capex funneled into them. Then you get the new fed chair who is gung-ho all in on AI + rate cuts, with Jerome keeping stuff chill in the back (which markets probably like)
-
SIVE 刚过 10 亿市值
有个有意思的事实，$SIVE 刚刚突破 10 亿美元市值门槛。 所以现在已经有少数美国机构可以买了（基金章程限制） 但绝大多数机构在它登陆纳斯达克之前还是买不了。 顺带一提： 如果最后真的能服务 $JBL、$AMD、$AAPL、$AMZN、$MSFT 等超大规模云供应链，那么 10 亿美元市值对美国机构来说其实只是零钱。 你只要看看 $LWLG，靠一份测试协议都能做到 19 亿美元市值。
英文原文
Fun fact, $SIVE just crossed the $1B MC threshold. So a select few US institutions are able to buy it now (fund mandates) However, the vast vast majority still can’t until they get listed on NASDAQ. Just an FYI: $1B valuations are spare change for institutional investors in US hyperscaler supply chains if they end up powering $JBL, $AMD, $AAPL, $AMZN, $MSFT and others. Just look at $LWLG, $1.9B MC off 1 testing agreement with $TSEM.
-
DNB 为 SIVE 估值发声
赞同，也很高兴看到欧洲领先银行之一 DNB 跟我一起出来为 $SIVE 的估值辩护。 我还是认为，随着激光业务按比例扩张，$SIVE 一年内可以到 100 亿美元市值。其增长会按以下规模推进： - $AAPL Watch - $JBL 1.6T 量产 - $MRVL CPO 量产 - Ayar 量产 - $POET 量产 这取决于它们如何从认证阶段走向量产爬坡。 既然 Sivers 正在给这个领域里下一代 1.6T / CPO 玩家供应激光器（进入类似 $AMD、$NVDA、$AMZN、$MSFT 这样的供应链），那就是已经存在的玩家了，市值大约 9.9 亿美元而已。还没算 TAM 扩张或者更多合作出现。 尤其现在有了纳斯达克上市，美国机构会提前按 12 个月的前瞻来定价，而欧洲本地估值主要看过去 12 个月。 银行通常会给出很保守的目标（比如 3 年 10 倍），但我确实看到这家公司很快可能变成下一个 $LITE。 欧洲应该拥抱自己公司为超大规模云厂商供货所带来的正和增长。 因为它们的前沿公司会通过税收、经济增长和就业增长回馈本地。
英文原文
Agreed, and glad DNB, one of Europe's leading banks, went out to defend $SIVE valuations alongside me. I still think $SIVE can reach $10B MC in 1 year time as their laser growth scales proportionally to: - $AAPL Watches - $JBL 1.6T Volume - $MRVL CPO Volume - Ayar Volume - $POET Volume Depending on how their qualification plays out into volume ramp. As Sivers supply lasers to all the next generation of 1.6T/CPO players in the space (into ~ $AMD, $NVDA, $AMZN, $MSFT type supply chains). These are EXISTING players at a ~990M MC. Not even including TAM expansion or more partnerships coming up. Especially now with NASDAQ listing, US institutions are forward looking and price in ~12M ahead of time, compared local European valuations that mainly look at previous 12 months. Banks usually provide very conservative targets (eg. 3 years for a 10x), but I do see potential for this company to be the next $LITE very soon. Europe should embrace positive-sum growth of their own companies that supply to hyperscalers. As their frontier companies provide back to locals through taxes, economic growth, and job growth.
-
把 SIVE 激光供应链映射到各家客户
我现在已经对瑞典文化感到很有趣了。 当有人把 $SIVE 激光的供应链映射到下面这些名字时，他们的散户 / 记者每天都在哭： -> $MRVL CPO -> $AMD CPO -> $JBL 1.6T LRO -> $NVDA -> $AMZN、$META、$GOOGL、$MSFT -> 腾讯、百度、阿里用于 ELS。 然后他们还会哭得更厉害…… 当别人展示 $AAPL 是 Sivers 激光器的秘密超大规模云客户时（因为苹果下一代 Apple Watch 开发需要每年 5000 万台产能）。 给超大规模云厂商供货……会让本地人难受成这样，甚至想自毁吗？ 有个“瑞典专业人士”出来说 CPO 没什么新东西、也没什么特别，所以 Sivers 也没什么特别，这真的很可笑。 与此同时，$NVDA 正在给从 $MRVL 到 $LITE、$COHR 的每个 CPO 项目都砸 20 亿美元。 X 最酷的地方在于，你每天和全球各地的人互动，都会学到新东西。
英文原文
I'm amused by Swedish culture at this point. That their retail/journalists are crying every day when someone posts a supply chain mapping of $SIVE lasers to: -> $MRVL CPO -> $AMD CPO -> $JBL 1.6T LRO -> $NVDA -> $AMZN, $META, $GOOGL, $MSFT -> Tencent, Baidu, Alibaba for ELS. Then they go crying even further... When someone shows $AAPL is secret hyperscaler customer of Sivers lasers for their next-gen Apple Watch development (which requires 50,000,000 units annually). Does supplying to hyperscalers... Hurt locals's feelings that much to the point they try and self-destruct? It's hilarious that a "Swedish professional" went out and said CPO is nothing new and nothing special, so Sivers is nothing special. Meanwhile $NVDA is just funding every CPO program $2B from $MRVL to $LITE to $COHR left and right. Cool thing about X is you learn something new every day interacting with global audiences.
-
两周等权收益回顾
先放这儿…… 如果等权计算，两周回报会是 +15.02%。 在 30 只不同股票上……大多是中大盘。 1. $INTC +29.62% 2. $MRVL +40.95% 3. $TSM +4.72% 4. $COHR +18.9% 5. $RKLB +26.76% 6. $DRAM +12.29% 7. $AVGO +18.32% 8. $AMZN +9.17% 9. $ARM +36.6% 10. $TSEM -1.25% 11. $IBIT +7.68% 12. $NBIS +15.22% 13. $GOOGL +6.41% 14. $AMKR +32.25% 15. $HOOD +19.14% 16. $CRCL +17.58% 17. $META +4.9% 18. $LITE -5.28% 19. $LPTH +20.23% 20. $FN +11.54% 21. $JBL +15.45% 22. $MP +17.48% 23. $HIMS +42.53% 24. $SMTC +18.83% 25. $POWL +9.26% 26. $VPG +17.44% 27. $MOG.A -3.96% 28. $MSFT +11.44% 29. $CVX -1.47% 30. $XLU -2.29% 显然时间周期很短，但我预计这些里面很多还会继续上涨。 如果你在特定名字的下跌时点位进场，而不是一口气同时做多，结果可能会更高。 也不算差吧？
英文原文
Just putting out there... Would have been +15.02% in 2W equal-weighted return. On 30 different stocks... mostly medium-large cap. 1. $INTC +29.62% 2. $MRVL +40.95% 3. $TSM +4.72% 4. $COHR +18.9% 5. $RKLB +26.76% 6. $DRAM +12.29% 7. $AVGO +18.32% 8. $AMZN +9.17% 9. $ARM +36.6% 10. $TSEM -1.25% 11. $IBIT +7.68% 12. $NBIS +15.22% 13. $GOOGL +6.41% 14. $AMKR +32.25% 15. $HOOD +19.14% 16. $CRCL +17.58% 17. $META +4.9% 18. $LITE -5.28% 19. $LPTH +20.23% 20. $FN +11.54% 21. $JBL +15.45% 22. $MP +17.48% 23. $HIMS +42.53% 24. $SMTC +18.83% 25. $POWL +9.26% 26. $VPG +17.44% 27. $MOG.A -3.96% 28. $MSFT +11.44% 29. $CVX -1.47% 30. $XLU -2.29% Obviously short timeframe, but I expect many of these to keep going up more. And probably would have been higher if you time the drop on specific names, rather than going long all at once. Not too shabby?
-
SIVE 终局映射很夸张
$SIVE 可能在 2027-2028 年映射到的所有超大规模云厂商，光按 9 亿美元市值来看就已经夸张了。 市场根本不懂接下来会发生什么。 从推测性映射来看： > $SIVE -> $POET -> $MRVL -> 1. $AMZN（来自 Celestial 的光子 fabric 采购协议 / warrants） 2. $MSFT（maia） 3. $GOOGL（最近和 Marvell 的开发对话） $SIVE 为 Poet Starlight / 光学 interposer 提供激光，而且几天前 Poet 的 CFO 也确认他们正在给 Marvell 供货。 > $SIVE -> $POET -> “其他超大规模云供应商的 NDA” 1. 西方超大规模云厂商 > $SIVE -> $JBL（1.6T LRO）-> 1. $META（Jabil 承接了 $INTC 的硅光遗产，对应 Meta LRO 项目） 2. $NVDA（NVIDIA 可能把光模块 OEM 出去）-> $MSFT | AWS | 其他超大规模云厂商 $SIVE 是 $JBL 1.6T 光模块已确认的激光源。 > $SIVE -> Ayar（上个月为了量产爬坡融资 5 亿美元） -> 1. Alchip（联合 CPO） 2. Intel 3. GUC/Wiwynn -> $AMZN（Alchip） -> $AMD（来自 $GFS 合作的 CPO）也有可能。 $SIVE 已知是 Ayar 的激光供应商，而且 Ayar 最近把 $MTSI / $LITE 从网站上删掉了。现在只显示 $GFS + $SIVE，这很可能说明 Sivers 是主要激光供应商。 既然 $GFS x $AMD 最近合作了，那如果 AMD 走 Ayar 路线，Sivers 就可能成为其 CPO 项目的核心激光供应商。 > $SIVE -> Enablence -> O-Net（大型亚洲 OEM）-> 亚洲超大规模云厂商 1. $AVGO ELS（可能） 2. $META 和 $GOOGL ELS 3. ByteDance（可能）-> ELS 4. Tencent（可能）-> ELS 5. Alibaba（可能）-> ELS $SIVE 和 O-Net / Enablence 在 OFC 附近有 ELS 合作关系。 Sivers 激光由 Win Semi 之类的代工厂大规模生产……而且在最近的图片展示里，它们也验证在 $GFS 的 CPO 供应链中。 这不是在看 Sivers 今天靠认证收入能指引多少，而是未来每一个西方 / 亚洲超大规模云厂商在 2027、2028、2029 及以后对 CPO / 1.6T 的需求所对应的营收。 $SIVE 看起来是市场上最不被了解、但还没被真正反映出来的光子股票之一。
英文原文
All the hyperscalers $SIVE likely ends up in 2027-2028 is staggering at a $900m MC. Markets don't understand what's coming. From speculative mapping: > $SIVE -> $POET -> $MRVL -> 1. $AMZN (purchase agreement/warrants with photonic fabric from celestial) 2. $MSFT (maia) 3. $GOOGL (recent development talks with Marvell) $SIVE powers Poet Starlight/optical interposers, and Poet's CFO confirmed they're supplying to Marvell few days ago. > $SIVE -> $POET -> "NDAs other hyperscaler suppliers" 1. Western Hyperscalers > $SIVE -> $JBL (1.6T LRO)-> 1. $META (Jabil $INTC SiPH inheritance, maps to Meta LRO program) 2. $NVDA (NVIDIA possibly OEMs optical transceivers) -> $MSFT | AWS | hyperscalers $SIVE is the confirmed laser source for $JBL 1.6T optical transceivers. > $SIVE -> Ayar ($500m fundraiser last month for volume ramp) -> 1. Alchip (Joint CPO) 2. Intel 3. GUC/Wiwynn -> $AMZN (Alchip) -> $AMD (CPO from $GFS partnership) possible. $SIVE is known laser supplier to Ayar, and Ayar removed $MTSI / $LITE from their website recently. Only showing $GFS + $SIVE, likely showing Sivers was primary laser supplier. As $GFS x $AMD partnered up recently, that makes Siver a possible core laser supplier for $AMD's CPO program if they go with Ayar. > $SIVE -> Enablence -> O-Net (massive Asian OEM)-> Asian Hyperscalers 1. $AVGO ELS (possible) 2. $META and $GOOGL ELS 3. ByteDance (possible) -> ELS 4. Tencent (possible) -> ELS 5. Alibaba (possible) -> ELS $SIVE ELS partnership with O-Net/Enablence around OFC. Sivers lasers is mass produced by foundries like Win Semi... and they're validated in $GFS CPO supply chains too from their recent image presentations. It's not about what Sivers is forecasting today from qualification revenue that everyone models off of. Alpha comes from future revenue proportional to demand from every Western/Asian hyperscaler for CPO/1.6T in 2027, 2028, 2029, and onward. $SIVE looks like one of the most unknown photonic stocks on the market that's yet to come.
-
博主炫耀自己早前推荐的IQE等股票被机构买入，强调研究前瞻性。
各位，大家都曾质疑我推荐的像$IQE这样的股票。 但为什么…我发了研究报告之后，瑞银或Point72这些机构都买入了？ 不知道瑞典媒体是否在做空$SIVE，但他们意识到是谁发掘了$AXTI的价值吗？ 我非常有信心，机构可能很快会跟进，因为Sivers的激光器为$MRVL、$JBL、Ayar等公司提供动力…这些都与$GOOGL、$MSFT、$AMZN及其他超大规模企业有关联。
英文原文
Bro everyone was doubting me on names like $IQE. Yet how come… all these institutions from UBS or Point72 buy in after my thesis posts? Not sure if Swedish media bearposting $SIVE realized who found $AXTI? I’m very confident institutions will likely follow-in soon given Sivers lasers power $MRVl, $JBL, Ayar, and others… with links to $GOOGL, $MSFT, $AMZN and other hyperscalers.
-
Blue Origin发射失败导致积压，连锁延迟影响ASTS等公司未来收益增长，非无关紧要。
太空板块的$ASTS看起来正玩得开心。尽管Blue Origin遭遇了重大失败，X平台上却莫名乐观...(这对SpaceX来说是利好，顺便也利好$RKLB。)至于$ASTS，当他们的卫星还没上天时，确实会影响未来的收入确认。Blue Origin的发射大概会推迟好几个月，可能影响~400-700万美元？如果一家主要以未来增长定价的公司被推迟了，那问题就不只是一颗卫星，而是后续的一系列卫星也会跟着延后...这对未来收益增长的影响很大(因为看起来他们没能达到45颗的目标)。经过数月的调查，Blue Origin可能会被$AMZN亚马逊的Kuiper项目和国防部的有效载荷积压住。看来他们处于非常不利的境地，不得不依赖竞争对手SpaceX。这绝非一些人用那些弯弯扭扭的线图所声称的"无关紧要"，但我也不评价这是否是买入机会。
英文原文
Space sector with $ASTS looks like it’s having fun. Lot of optimism on X somehow, even after Blue Origin had a major failure... (positive for SpaceX and byproduct $RKLB.) As for $ASTS it does matter for future revenue recognition, when they dont have satellites in space. Blue Origin launches probably got delayed a ton of months, maybe like ~4-7M? If a company that is priced primarily by forward growth gets delayed. Then it's not just one satellite, but the subsequent ones ones as well... this hurts forward earnings growth a lot (since it doesn’t look like they hit 45 targets) After months of investigation Blue Origin will likely be backlogged from $AMZN Kuiper and DoD payloads too. Seems like in a pretty bad spot being dependent on their competitor SpaceX. Definitely not "immaterial" as others were claiming with their squiggly line charts, but no comment about buying opportunity.
-
SIVE 供应链里很多都很强
我一直在告诉大家…… $SIVE 供应链里有很多东西都很有吸引力，适合做多。 Win Semi（3105）一天涨 10%，然后涨停。 虽然它们显然是为 $AVGO 或 SpaceX 供货，但光子学爬坡是由 $NVDA 领头，随后是 $GOOGL、$AMZN、$MSFT。 这会让光子学成为 Win 的一个巨大增长引擎。
英文原文
I'm telling you all... Lot of stuff in the $SIVE supply chains make very compelling longs. Win Semi (3105) just goes up 10% a day then halts trading after hitting its max. Obviously they do things for $AVGO or SpaceX, but photonics ramp spearheaded by $NVDA and followed by $GOOGL, $AMZN, $MSFT Would make photonics a massive growth vector for Win.
-
SIVE 下一站像 LITE
谢谢，我之前确实说过 $SIVE 看起来像下一个 $LITE！ 去年我因为 Google TPU 看好 Lumentum，而市场通常会更喜欢 Google 生态而不是 Amazon 生态。 但如果 Apple 在 CPO 上放量，市场会重新认识这整条线。
英文原文
Thanks, I did say $SIVE looked like the next $LITE! I went long on Lumentum last year because of the Google TPU and markets tend to like Google's ecosystem a bit more than $AMZN or $MSFT. Of course, a lot can happen in the next 2 years so this is speculative but likely supply chain mapping, we'll find out what happens.
-
市场更爱确证型证据
@Gubloinvestor 我的意思是，谁都可以通过 AlChip、Marvell 之类把它映射到 Amazon，或者映射到 Microsoft，对吧？ 但市场有时候更喜欢一些更具体的“证据”，比如来自 $AMZN 的 photonic fabric 采购协议，而它又能映射回 $SIVE。
英文原文
@Gubloinvestor I mean anyone can just map it to Amazon through AlChip, Marvell, etc? Or Microsoft? But sometimes markets like more concrete“proof” like photonic fabric purchase agreements from $AMZN that maps to $SIVE
-
Amazon 的 warrant 协议会带来放量
@PhotonCap 基本上每次 $AMZN 有像 $ALAB 2024、$AAOI 2026 这种 warrant 协议时，量产订单都会起飞。
英文原文
@PhotonCap Basically every time $AMZN has these warrant agreements like $ALAB 2024, $AAOI 2026, volume orders go brrr
-
SIVE 的终端用户是谁
以防你还在想 $SIVE 的终端用户到底是谁： Sivers 激光 -> $AMZN（Marvell 的 8-K SEC 文件里有） Amazon 和 Celestial 之间有 “photonic fabric” 的采购协议，而这条链通过两跳关系映射到 Sivers 激光。 协议里 Marvell 的股份价值按 87 美元/股算，所以 Amazon 有很强的动力尽可能多买（Marvell 现在股价已经 139 美元以上了）。 这些量产订单对 $SIVE 最有帮助，因为他们体量小。 我只是想指出这一条关系。 但在 7.2 亿美元市值下…… 市场显然漏掉了 Sivers 作为未来超大规模云光学超级周期里关键激光供应商的角色。
英文原文
Just in case you were wondering who the end users of $SIVE likely were: Sivers lasers -> $AMZN (Marvell's 8-K SEC filings). Amazon has purchase agreements for "photonic fabric" from Celestial, which maps to Sivers lasers through two-hop connections. Marvell's share value under the agreement was $87/share, so Amazon is very incentivized to buy as much as they can (Marvell is trading at $139+). These volume orders help $SIVE the most out of this supply chain, given their size. This is just one-relation to Amazon, I wanted to point out. But at $720m MC... Markets missed Sivers as the critical laser supplier for the upcoming hyperscaler optical supercycle based on new architectural changes.
-
把 SIVE 与 JBL 的供货关系视作结构性转变。
这终于算是 $SIVE 的供应链映射验证了。 $JBL 真的******巨大，而 Sivers 激光现在大概率会进入 Jabil 的 hyperscaler 客户，比如 $AMZN、$GOOGL。 这是一个非常结构性的变化，如果官方确认下来，我个人会把 Sivers 估到 20 亿美元以上。
英文原文
This is finally supply chain mapping validation with $SIVE. $JBL is ****** massive and Sivers lasers now will likely go into Jabil's hyperscalers clients like $AMZN, $GOOGL. This is such a structural change and I would personally value Sivers over $2B+ after official confirmation.
-
批评只看技术图形的分析，强调基本面、催化和现金流建模。
那些收 500 美元一个月、只会画线的人，大概是我最烦的存在了……尤其是当他们还会在我的 thesis 帖子下面评论，比如 $AEHR 财报那种。 股价是由基本面和催化推动的。技术分析只是自我实现的占星术；如果 $AMZN 有一笔巨大的采购订单，没人会在乎技术面说的“超买”，真正重要的是把这会带来多少自由现金流，相对于市值要怎么建模。 $IREN 也是一样，居然还有人拿技术图去画它，而市场里明明有 60 亿美元的稀释，图表根本没把这部分算进去。
英文原文
Swiggly line drawers that charge $500 a month are possibly the greatest annoyance to me... Especially when they comment underneath my thesis posts like $AEHR earnings. Stocks move based on fundamentals and catalysts. TA is astrology self-fulfilling prophecies but if $AMZN has a massive purchase order, nobody cares about "overbought" based on TA it's about modeling in how much FCF that brings relative to MC. Same with anyone who tries drawing a TA for $IREN when there's literally $6,000,000,000 in dilution that a chart doesn't account for.
-
亚马逊因Anthropic模型基于Trainium训练消息反弹，带动MRVL、AAOI等相关股
顺便解释一下你们可能想知道的$AMZN今天为何反弹：新消息显示Anthropic的Mythos及其最新模型很可能是基于Amazon Trainium训练的。这对亚马逊生态系统来说意义重大，从$MRVL到$AAOI都受益于Anthropic。就像今年早些时候谷歌TPU生态因Gemini而反弹，包括$LITE一样。
英文原文
Just in case you’re wondering why $AMZN rallied today: New news that Anthropic Mythos and their latest models were likely trained on Amazon Trainium. This is big for Amazon’s ecosystem from $MRVL to $AAOI due to Anthropic. Just like how Google’s TPU ecosystem rallied earlier this year like $LITE from Gemini.
-
博主列出30只看好的美股并简述理由，涵盖AI半导体、太空、机器人、加密等领域
今天随机列出30只美国可投资的股票以及我喜欢它们的理由： 1. $INTC——美国晶圆代工的希望，国家安全 2. $MRVL——未来的maia芯片和CPO等附加产品推动营收规模化，业务范围太多数不过来 3. $TSM——半导体/AI的支柱 4. $COHR——业务全垂直整合 + 捕获光学周期 5. $RKLB——太空领域的终极前沿，5年后和20年后都会存在 6. $DRAM——三星/SK海力士的内存敞口 7. $AVGO——超大规模云商不喜欢英伟达GPU税 8. $AMZN——没人能竞争得过隔夜送达卫生纸这件事。机器人技术会长期降低运营费用 9. $ARM——AGI CPU在未来十年会大幅提升营收 10. $TSEM——做光相关的东西你需要晶圆代工 11. $IBIT——比特币，这点我们现在已经众所周知了 12. $NBIS——我认为它是下一个AWS。此外它们还与优步合作做自动驾驶，拥有可扩展的数据库公司、数据标注业务。几乎像一个小谷歌。 13. $GOOGL——youtube不会消失，gemini很棒。它们通过TPU实现垂直整合，用运营收入资助基础设施建设，所以我看好它。 14. $AMKR——超级工厂将在2027-2028年末投产。受益于美国制造政策 15. $HOOD——短期我不喜欢，但长期我看好罗宾汉，因为他们抓住了散户 + 有更多产品如银行等正在规模化。产品创新很疯狂。 16. $CRCL——我恰好非常喜欢稳定币，并认为它们是支付和持有的未来（取决于澄清法案的进展） 17. $META——人们不会很快停止使用Instagram、Whatsapp或其他应用。 18. $LITE——$GOOGL TPU敞口在BOM中占比较高。只要谷歌的AI计划持续运行，我认为$LITE会表现良好。 19. $LPTH——锗和中国出口管制问题将持续存在，因此美国制造的工程替代品将始终重要。 20. $FN——总得有人来组装光学产品 21. $JBL——同上，但加上英特尔硅光业务的IP，可能会成为innolight那样的公司 22. $MP——美国稀土项目极其重要，类似于$INTC涉及的国家安全风险 23. $HIMS——听我说，他们刚收购了一堆公司，股价在$19时有全球DTC渠道。做空者非常讨厌这家公司，但我认为作为逆向多头它实际上很有前景。 24. $SMTC——LRO/LPO转型 25. $POWL——美国版hammond，用于开关设备DC类瓶颈的替代方案 26. $VPG——人形机器人未来会成为主流，大概2027-2028年，这家公司做传感器。 27. $MOG.A——感觉在机器人领域、spacex供应链里到处都能看到它们 28. $MSFT——在$375这个价位，有一天我们会回头看发现这是买入机会。 29. $CVX——战争结束后油价可能会崩，但这些石油公司将变得极其重要，尤其是委内瑞拉还是个金矿。 30. $XLU——我认为降息可能会重启，我们需要电力/电网来支持AI，所以从$CEG到$NEE这些名字将始终重要。 只是顺便抛出一些除了$AAOI和$AEHR之外的想法。
英文原文
Here's a bunch of random 30 US-available random stocks I like today and why: 1. $INTC - America's hope for foundry, national security 2. $MRVL - scales rev from future maia asics and add ons like cpo, they do everything lost count 3. $TSM - backbone of semis/ai 4. $COHR - They do everything vertically integrated + captures optical cycle 5. $RKLB - the final frontier of space will be around 5 years from now and 20 years from now. 6. $DRAM - memory exposure for samsung/sk hynix 7. $AVGO - hyperscalers dont like nvidia gpu tax 8. $AMZN - nobody can compete against the overnight shipping of toilet paper. robotics will lower opex over time 9. $ARM - AGI CPUs scale revenue quite a bit over the next decade 10. $TSEM - you're going to need a foundry for light based stuff 11. $IBIT - bitcoin, we all know by now 12. $NBIS - i think it's the next AWS. Also they do self-driving cars with uber, own scaling DB companies, data labeling. It's almost like a mini Google. 13. $GOOGL - youtube is not going away, gemini is great. they're vertically integrated with TPUs and fund buildout with operating income so i like it. 14. $AMKR - super facilities coming online in late 2027-2028. benefits from made in america 15. $HOOD - i dont like short term, but long term i'm a fan of Robinhood since they captured retail + have more products like banking, etc that they're scaling up. product innovation is wild. 16. $CRCL - I happen to really like stablecoins and see them as the future for both payments/holding (depends on clarity act) 17. $META - people aren't going to stop using instagram or whatsapp, or others anytime soon. 18. $LITE - $GOOGL TPU exposure decently high part of BOM. As long as Google's AI program keeps running I think $LITE will do well. 19. $LPTH - Germanium and China export controls will always be an issue so US made engineered alternatives will always be important 20. $FN - Someone needs to assemble optical stuff 21. $JBL - same as above, but added with ip from Intel's SiPh acqusition so might end up like innolight? 22. $MP - American rare earths program is extremely important, similar to $INTC national security risks 23. $HIMS - Okay here me out they just acquired a ton of companies, and at $19 they have global DTC channel. short sellers really hate this company, but I think it's actually promising as a contrarian long 24. $SMTC - LRO/LPO transition 25. $POWL - US alternative to hammond for switchgear DC type bottleneck 26. $VPG - Humanoids will be a thing down the road maybe 2027-2028, this makes the sensors. 27. $MOG.A - Feels like i see them everywhere in robotics, to spacex supply chains 28. $MSFT - At $375, one day we'll look back and see this as a buying opportunity. 29. $CVX - oil might crash after war but these oil companies are going to be extremely important, especially when Venezulea is a goldmine. 30. $XLU - i think rate cuts might be back online, we need power/grid for AI so these names will always be improtant from $CEG to $NEE Just throwing out other thoughts aside from $AAOI and $AEHR.
-
用订单和 backlog 说明变压器 / 开关设备是数据中心的关键瓶颈。
当前瓶颈：变压器 / 开关设备。 交易思路：在 184 加元买入 Hammond（约 22 亿加元 / 15 亿美元）的多头。 它们主导的市场包括： - 变压器（干式，多年瓶颈，约占市场 23%） - 开关设备（2-3 年瓶颈） - 还生产液体型产品（5 年瓶颈，规模更大） 我个人预计其价格会像 NAND 一样上调，因为 $AMZN、$MSFT 等公司都在争夺配额。 你可能见过这类说法： “美国一半的数据中心建设被推迟或取消，增长受限于电力基础设施短缺”…… 再往下看： “为了解决短缺……加拿大、墨西哥……成了 AI 数据中心高压变压器的最大供应国”…… 猜猜谁在加拿大的 Guelph、墨西哥的 Monterrey 3 和 4，以及美国都有布局？ Hammond。 再看文章引用的，为什么超大规模客户的数据库扩张正在崩掉： “这些挫折背后的主要原因，是关键电气组件的可获得性——比如变压器、开关设备。” 机构大概都在看 Powell、Eaton 之类的公司……但他们可能不知道？ 这些公司其实会买 Hammond 的变压器再装进自己的开关设备里（“大量卖给数据中心、开关设备制造商”）。 它们在变压器市场的份额其实很大（例如干式约 23%）。 最有说服力的信号： -> 2025 年 backlog 同比增长 122%。我们可以据此推测超过 10 亿加元。 例如，公司 2025 年销售额为 8.98 亿加元，产能已经顶到上限。管理层说，2025 年第三季度末的订单价值相当于整个第三季度 backlog 的 53%。 考虑到 2025 年第四季度营收是 2.54 亿，而 backlog “翻了一倍多”，我们可以推算总 backlog 超过 10 亿加元。 另外： “去年毛利率压缩是因为墨西哥工厂建设，但预计毛利率都会上升，工厂扩建会在 2026 年第二季度转化为加速收入增长”，也就是现在。 下行风险在于原材料成本（铜、电工钢）如果再次上升，不过既然处在这种瓶颈里，他们是可以提价的。 我个人的 2026 年前瞻市盈率估计大概是 18-21 倍，2027 年不到 15 倍，因为有产能爬坡。 但如果他们能一边提价，一边接到超大规模客户的紧急订单，前瞻市盈率甚至有可能到个位数。不过那可能会和新并购混在一起。 总之它还是很便宜。 简单总结一下： $AMZN、$MSFT、$META、$GOOGL、$ORCL 的数据中心正在被变压器 / 开关设备短缺卡住。 市场似乎错过了这家市场份额很大的小公司，尽管它有 backlog 可见度，而且随着产能扩张上线，营收还在增长。 我个人觉得很有吸引力，所以建了多头仓位。 当然，这只是我的个人想法，做决定前请自己 DYOR。
英文原文
The current bottleneck: Transformers/Switchgear. Trade Idea: Long Hammond (~2.2B CAD / ~$1.5B USD) at 184 CAD. They dominate the market for: -Transformers (dry, multi year bottleneck ~23% of market), -serve to switchgear (2-3Y bottleneck) -and manufacture liquid too (5Y, larger bottleneck) I personally anticipate components price hikes like NAND, as $AMZN, $MSFT and others compete for allocation. You might have seen: “Half of US data center builds have been delayed or canceled, growth limited by shortages of power infrastructure”… Then you go further: “To address shortages… Canada, Mexico… became the biggest suppliers of high-power transformers for AI data centers to AI data centers” Guess who is in Canada (Guelph).. Mexico (Monterrey 3 and 4)… and the US? Hammond Then here’s the reason the articles cite why hyperscaler DB buildouts are falling apart: “Major reason behind these setbacks is the availability of key electrical components — such as transformers, switchgear”. Institutions are probably looking at Powell, Eaton, and others… but little do they know? Companies like these actually buy Hammond’s transformers to put inside their own switchgear (“strong sales into data centres, switchgear manufacturers") Their market share over the transformers market is actually pretty large (eg. ~23% dry). The most compelling signal: -> 122% Y/Y 2025 backlog increase. And we can infer this to be 1B+ CAD. Eg. company achieved 898m CAD in sales in 2025, capacity ceiling. Management said close of Q3 2025 orders were valued at 53% of the entire closing third-quarter backlog. Given that Q4 2025 revenue was 254 million and the backlog is "more than doubled," we can infer a total backlog value exceeding 1 billion CAD. Also: “Gross margin compression last year was due to the buildout of their Mexico facility, but both gross margins are expected to increase and the facility expansions are expectied to turn into accelerated revenue Q2 2026)” which is now. Downside is if raw material costs (copper, electrical steel) spike again, but given this bottleneck, they can price hike. Personal FWD P/E estimates would be ~18-21 for 2026, <15 for 2027 from volume ramp. But I think it’s possible to hit single digit fwd P/E if they do price hikes mixed with hyperscaler emergency orders. But that might get a little mixed with the new acquisition. Regardless still looks cheap. Just a TLDR: $AMZN, $MSFT, $META, $GOOGL, $ORCL datacenter are being bottlenecked because of a lack of transformers/switchgear. Seems like markets missed this little player with large market share, despite backlog visibility and increasing revenue from capacity expansion coming online. I personally found it pretty compelling, so I went long. Just sharing my personal thoughts, of course DYOR before making any decisions yourself.
-
消费级光子市场比AI数据中心更大，光子公司不应只盯大客户订单。
消费级激光剑产品的市场规模比AI数据中心还大。说真的不明白这些光子公司在干嘛，被$MSFT、$AMZN的超大规模数据中心订单分散了注意力。
英文原文
@tomszczypka Consumer lightsabers have a bigger TAM than AI DCs. Not sure what these photonic companies are doing, getting distracted by $MSFT, $AMZN hyperscaler orders tbh.
-
认为 AEHR 在小市值下被低估，正被多家超大客户做资格认证。
$AEHR 在大约 11 亿美元市值下看起来极有吸引力。 Aehr 开始让我想起早期的 $TER，又有点像财报前的 $AAOI。 如果把时间线和猜测客户列出来： 2 月 11 日：Sonoma 为超大客户的 AI ASIC 处理器拿到生产胜利。（大概率是 $GOOGL、$AMZN、$META） - 可能是谷歌？Aehr 收购了 Incal，而 Incal 曾被猜测会被谷歌用于 TPU。 2 月 26 日：来自 AI 主力客户的 1400 万美元。（大概率是 $AMD、$NVDA） - 这里更像是 $AMD，用于 Instinct MI300/MI400。 3 月 3 日：一台 FOX-XP 系统对应的主硅光客户。（大概率是 $INTC siph） - 很可能 $INTC 就是他们的主客户。 3 月 31 日：来自一家新的大型硅光客户的首单。（大概率是 $AVGO、$MRVL、$CSCO） - 新客户（排除了 Intel），大概是这些正转向 800G/1.6T 硅光收发器的公司之一。 （这些都属于非常保密的 BOM 级推测） 不管怎样，这条时间线对 $AEHR 来说就是在不断加码。 也许就是下一次财报。也可能再过两季。 但感觉量产订单只是时间问题。
英文原文
$AEHR looks extremely promising at ~$1.1B MC. Aehr is starting to remind me of an early $TER, mixed with pre-earnings $AAOI. If we look at the timeline and speculated customers: Feb 11th: Sonoma production win for Hyperscaler's AI ASIC processors. (likely $GOOGL, $AMZN, $META). - Probably Google? Aehr bought Incal, who was speculated to be used by Google for their TPUs. Feb 26th: $14 million from AI lead customer (likely $AMD, $NVDA) - Probably $AMD here for Instinct MI300/MI400. March 3rd: Lead silicon photonics customer for one FOX-XP system (likely $INTC siph) - Very likely $INTC has been their lead customer. March 31st: Initial order from major new silicon photonics customer (likely $AVGO, $MRVL, $CSCO ) - New customer (rules out Intel), prob one of these transitioning to 800G/1.6T silicon photonics transceivers (All speculative, very confidential BOM) Regardless. This timeline is just bottling up for $AEHR. Could be next earnings. Or two quarters from now. But feels like a matter of time before we see mass orders.
-
把 OpenAI 比作地缘武器化的主体，认为 AI capex 还会延续。
OpenAI 就是科技界的伊朗。 如果他们没资金，却还造成了那么大的连锁反应——就像伊朗把无人机送到所有邻国一样——那会非常危险…… 从 $MSFT 的算力 backlog、$ORCL Stargate、$AMD 交易、$NVDA 交易、$CRWV 交易、$AMZN 交易，到 SoftBank 交易。 再加上 Cerebras 和很多私有公司的交易。 这些都足以说明，如果它倒下，支撑这些公司估值的很大一部分也会一起崩。 OpenAI 融了 1220 亿美元，我猜这部分也在某种程度上武器化了这种连锁反应，就像伊朗在控制海峡时那样。 所以 AI capex 的这波行情大概率还会再持续至少一年。 它们是 AI 牛市的起点；如果它们现在出问题，反而会成为戳破泡沫的人。
英文原文
OpenAI is the tech equivalent of Iran. They would have caused so much contagion - like Iran sending drones to all their neighbors - with funding they didn’t have…. From $MSFT compute backlog, $ORCL Stargate, $AMD deals, $NVDA deals, $CRWV deals, $AMZN deals. SoftBank deals. Deals with private companies like Cerebras and many others. So much so, that if they went down, so would a huge part of what gave these companies their valuations. OpenAI raised a whopping $122B, my guess is partly weaponizing this contagion like what Iran is doing to the Strait. So now the AI capex spend rally is likely to last another 1 year min. They’re the first to start the AI rally; and if they go under, at this point they’ll be the one to pop the bubble.
-
反问 AAOI 直连这些巨头客户还有什么更好的客户可言。
@Jornka329996 $AAOI 直接连着 $ORCL、$MSFT、$AMZN，甚至可能还有 $META，还需要什么新客户？ 这大概已经是他们能拿到的最好客户了，哈哈。
英文原文
@Jornka329996 How is $AAOI direct to $ORCL, $MSFT, $AMZN, and maybe $META needing new clients? Those are probably the best they can get lol
-
AI数据中心供应链严重依赖亚洲，中东冲突加剧流动性风险，稀土应成美国国安优先事项
Hi Ron，这个话题分为两个不同的部分： 1. 超大规模云服务商（$ORCL、$META、$AMZN）：他们的支出超过了盈利水平（$GOOGL 是罕见的例外，而 $AAPL 相对而言支出并不多）。 因此市场担心这并没有转化为实质性的收入。 尤其是像 Oracle 这样负债累累的公司，降息是使其稍微可持续的重要因素（如果你看债务利息，数目非常庞大），而降息既能推动产能扩张，也能带来实质性的成本节省。 由于伊朗战争和原油价格上涨，原定2026年的降息预期已经落空。 雪上加霜的是，人们担心他们的 AI 数据中心建设供应链会中断。 如果你看看他们的 AI 数据中心部件产地： - 来自台湾 - 来自韩国 - 来自日本 这是一个全球化的供应链。 虽然美国看起来没问题，因为它在石油、液化天然气、氦气和其他半导体供应链所需的资源方面是绝缘的。 但我们的亚洲合作伙伴并不处于长期冲突中。 所以当你看到亚洲那些规模较小、专业化的公司时，它们正在苦苦挣扎。 这种影响会层层放大，一直传导到终端的亚马逊网络服务 AI 数据中心（成本上升、供应短缺）。 2. AI 整体而言：投资者对此非常乐观。 问题在于实现路径，而美国在表面上看起来被中国卡住了。 - 美国投入过多，而中国和其他国家正在蒸馏我们的最新模型（需要 KYC 端点管控），坐享其成。 - 随着 AI 项目的规模扩大，它正变得越来越依赖中国公司。 我认为这是因为美国缺乏稀土，这应该成为我们的头号国家安全优先事项。 这是制造新一代 AI 硬件、机器人供应链和太空探索所需的。 这些资源都由中国和俄罗斯控制。即使是领先的"西方"公司，我已发现它们仍然主要依赖中国/俄罗斯。 这是我们确保前沿项目安全最大的漏洞之一。 如果你看光子学（AI 硬件如何通过光加速），我们没有西方的供应链来满足需求。主要材料来源是中国的 $AXTI 和中国的 Vital 两家公司，然后流转到日本或其他公司，最后进入美国。 如果你看我们的人形机器人项目（我们如何将 AI 带入现实世界），$TSLA Optimus 的所有躯体都是在中国制造的。 美国供应链缺乏制造这些组件所需的稀土，因为成本太高。 3. 中东地区的冲突 近期下跌的一个重要原因是流动性问题。我们在阿联酋、中东的合作伙伴主要在为私募市场提供资金（想想 OpenAI），他们大量的支出流向 AI 市场。 他们也是从 Meta 到微软等科技七巨头最大的投资者之一。 如果他们的油田继续受到干扰，他们可能不得不从美国市场撤出流动性。 这既影响这些公司在 AI 上的支出能力，也会导致美国市场的抛售下跌。 如果有takeaway（要点）的话：请把稀土列为美国头号国家安全优先事项。特朗普总统已经投资了 $USAR、$MP 和几个单独的公司。 但还有许多更多至关重要的、尚未盈利的公司，它们对于制造 AI、自动驾驶汽车和机器人都是必需的。我们需要打破对中国/俄罗斯的依赖。
英文原文
Hi Ron, so there's two different parts to it: 1. Hyperscalers ( $ORCL, $META, $AMZN): They're spending more than they have with profit ( $GOOGL is the rare exception, and $AAPL isn't really spending much relatively). So markets are worried less worried it's translating into material revenue. Especially with companies like Oracle taking on immense amounts of debt, rate cuts were a large driving factor in why it's slightly more sustainable (if you look at debt interest, it's massive), and cutting rates drives both forward expansion and material amounts of savings. Due to the War in Iran, rising crude, former projections with cutting rates in 2026 are now gone. This is compounded that there are worries that their AI buildout supply chain gets disrupted. If you look at where their AI datacenter parts are made: - it comes from Taiwan - it comes from Korea - it comes from Japan It's global. While the US might look like it's fine, since it's insulated to Oil, LNG, helium, and others needed for the semiconductor supply chain. Our Asian partners are not in a long drawn out conflict. So when you look at companies in Asia that are more small and niche they are struggling. And this gets compounded tens of times until it reaches all the way down to the end Amazon Web Services AI datacenter (increased costs, lack of supply). 2. AI as a whole: Investors are very bullish on it. It's just how we get there, and America looks strangled by China optically. - America is spending way too much on it, with China and others distilling our latest models (need KYC endpoints), and reaping all the benefits. - As we scale our AI programs, it's becoming increasingly reliant on Chinese companies. I'd argue it's because America lacks the rare earths, which should be our #1 national security priority. This is needed to make new generations of AI hardware, robotic supply chains, and Space. They're all controlled by China and Russia. Even the leading "Western" companies, I've identified they're mainly relying on China/Russia still. And this is one of the biggest vulnerabilities in securing our frontier programs. If you look at photonics (how AI hardware is sped up by light), we no Western supply chain to fulfill our needs. It's $AXTI (in China), and Vital (in China) as the two main sources of materials needed to make them, that gets passed down to Japan, or other companies before they end up in US. If you look at our humanoid program (how we move AI to the real world), all the bodies of $TSLA Optimus are made in China. American supply chains lack the rare earths needed to make the components because it gets too expensive. 3. Disruption in the Middle East A large part of the recent drop is mainly due to liquidity. Our partners in the UAE, Middle East are largely funding private markets (think OpenAI) and a lot of their spend going to AI markets. They're also one of the biggest investors in Mag7 from Meta to Microsoft. If their oil fields continue getting disrupted they may have to pull out liquidity from US markets. This impacts both the amount of money these companies can spend on AI. As well as causing a drop in American markets from selling. If there's any takeway: Please make Rare Earths Amercia's #1 national security priority. President Trump already invested in $USAR, $MP, and a few individual names. But there are many more extremely critical, yet unprofitable companies that are needed to make AI, to self driving cars, to robotics. And we need to break our reliance on China / Russia.
-
认为即使大盘再跌，AAOI 也可能跑赢。
@Ungrate11057544 就算 $SPY、$MSFT、$AMZN 再跌 20%，我也还是觉得 $AAOI 会跑赢市场。
英文原文
@Ungrate11057544 Let's just say even if $SPY, $MSFT, $AMZN crashes another 20% I expect $AAOI to outperform the market.
-
不喜欢 Jane Street 持有和自己一样的 AAOI，因为他们会放大波动。
这就是为什么我不喜欢 Jane Street 持有和我一样的 $AAOI。 他们是做波动交易的，尤其喜欢搞散户最爱的名字，然后触发止损和恐慌。 但如果你知道自己持有什么： 一家美国公司，6.69 十亿美元市值，做的是整条收发器供应链： 从激光 -> 设计 -> 组装。 如果 $MSFT、$AMZN、$ORCL 这些公司真的会把他们能买的都买了，那对我个人来说，这个价格简直是捡漏。 对高 beta 股票来说，入场前把 conviction 和仓位管理好特别重要。 否则你就会一遍又一遍地在底部割肉、在顶部追高。
英文原文
And... this is why I don't like Jane Street owning the same stocks I do like $AAOI. They trade volatility, especially with retail favorite names and trigger stop losses/panic. If you know what you own though: $6.69B for a US company that makes the entire transceiver supply chain: From Laser -> Design -> Assembly. With likely $MSFT, $AMZN, $ORCL buying anything they can make, is a steal for me personally. With high beta stocks, it's really important to build conviction and manage sizing correctly before entering a trade. Otherwise you'll end up capitulating the bottom and buying the top over and over.
-
认为 SIVE 可能被多方盯上，也可能因持股结构保持独立。
洗澡时冒出来的一个想法…… SIVE 现在可能已经被 $QCOM、$MRVL、$AVGO、AlChip、联发科、$AMZN、$META、$MSFT 的风投部门盯上了？ 既然现在大家都公开知道了，Broadcom 只要花 3.5 亿美元就能买下 Marvell 在光子学上的咽喉点（Amazon / Microsoft 的 ASIC 项目）…… 或者 Qualcomm 也可以用同样的方式收购 $SIVE，然后在收购 Alphawave 之后把激光 IP 做垂直整合。 又或者……超大规模客户会通过持有足够多股份，想办法让 $SIVE 继续保持独立？
英文原文
It’s funny a shower thought… Probably put $SIVE on the radar for $QCOM, $MRVL, $AVGO, AlChip, Mediatek, $AMZN, $META, $MSFT venture arms now? Since everyone publicly knows now Broadcom can just go and buy Marvell’s photonics chokepoint vulnerability (Amazon/Microsoft ASIC programs) for a rounding error of $350m… Or Qualcomm can do the same buying $SIVE then vertically integrate laser IP after Alphawave aquisition. Or maybe… hyperscalers figure out a way $SIVE remains independent by owning enough shares?
-
Sivers作为稀缺独立cw激光供应商，战略价值远超当前市值估值。
我个人看到 $SIVE 下跌风险较小的原因： 像 $AVGO 这样 $MRVL 的竞争对手可以直接收购 Sivers，从而控制其近中期光子学路线图... 估值约 2.9 亿美元... 如果再便宜一点的话。 Gemini 的回应很有趣： "像 Broadcom 这样以约 2.9 亿美元收购 Sivers Photonics 来扼制 Marvell 新收购的 Celestial AI 路线图不仅仅是战略行为； 这是一场投资回报率高得离谱的供应链战争。" "花点小钱，竞争对手就能控制上游光源，把握直接的 IP 路线图，并让试图整合 Celestial AI 的 Marvell 付出数十亿美元的机会成本。" 对于 $MSFT, AlChip, $MRVL, $AMZN 等超大规模云服务商来说，投资 Sivers 来确保其共封装光学(CPO)供应链可能比让它被竞争对手收购更有利。 就像 $QLCM 与 Alphawave 针对 OpenLight 生态所做的操作一样。 我的观点是：$SIVE 作为全球少数保持独立运营的 cw 激光器供应商之一的关键瓶颈地位... 其固有的价值远超其市值。 对冲基金因「MC」和瑞典交易所上市限制无法投资从而压低估值... 对于收购方或散户投资者来说反而是一种因祸得福。
英文原文
The reason I personally see less downside risk with $SIVE: A competitor to $MRVL like $AVGO can just buy Sivers outright and their near-medium term photonics roadmap… At ~$290m… if it gets any cheaper. Gemini responses were interesting: “ A competitor like Broadcom buying Sivers Photonics for ~$290M to choke Marvell’s newly acquired Celestial AI roadmap isn't just strategic; It is supply chain warfare with an absurdly high ROI. “ “ For pocket change, a competitor could buy the upstream light source, control the immediate IP roadmap, and inflict billions of dollars worth of opportunity cost on Marvell right as they try to integrate Celestial AI “ It’s likely better for hyperscalers like $MSFT, AlChip, $MRVL, $AMZN to invest in Sivers to secure their CPO supply chain then have it being bought out by one of their competitors. Like what happened already with $QLCM and Alphawave for OpenLight. My opinion: $SIVE chokepoint as one of the few projects remaining independent cw laser suppliers in the world… Is inherently way more valuable than their marketcap. Hedge funds not being able to invest due to “MC” and Sweden listing to keep valuations depressed… is a blessing in disguise for both acquisitions or retail investors.
-
提出让 SIVE 沿光子链条扩张 TAM、借资本注入和拆分推进 ELS 模块规模化。
难道 $SIVE 就不能学 $LITE 那样，把 TAM 往 ELS 整个堆栈一路往下扩吗？ 我已经开始想起早期的 $LITE 了（Oclaro、NeoPhotonics、Cloud Light）…… 下面是我的论点/设想： Sivers 如何把大约 300-400 美元的 ELS（来自 50-100 美元的阵列）吃下来，并成长为一家 100 亿美元以上的公司。 路径大概是： -> Sivers 需要融资。 -> 拿 NRE 预付款或者 PIPE（比如 $AXTI/Northland） -> 直接告诉市场：“$AVGO 和 $LITE 会把你们定制 AI 集群的利润率卡住。” ----> 让 $META / $AMZN / $MSFT 投资注资，来支持未来 ASIC 集群的多来源供应，因为他们会害怕。 尤其是 $NVDA 已经出资 40 亿美元来保障从 $LITE 和 $COHR 的供应……再加上当前 EML 产能瓶颈带来的 PTSD。 其实比大家想的容易。 -> 把光子学业务分拆到美国（市场不喜欢给老牌公司投资）。 - - - - - > 这件事大概率会在 2026 年奏效，因为 CPO / photonics 现在是热潮，而不再是 2024 年到 2025 年初那样。 -> 筹到足够的钱后，自动化晶圆探针机，并给模块 NPI 提供资金。 -> 采用轻资本 fab-lite 测试模型：自建探针卡并内部测试，但把 capex 交给 OSAT。 -> 收购硅光封装 IP（或者继续和 O-Net、$POET 做 JDM）。 -> 保留 Glasgow 的 InP 晶圆厂，用于内部测试、迭代和原始外延。 -> 把光刻、刻蚀等环节外包给 Win Semi 和其他厂商（别自己做重资本环节）。 -> 然后 $SIVE 只要把经过测试的激光器和封装 IP 交给 $FN。 然后就成了。 你拥有了一个独立、可大规模生产、高毛利的 ELS 模块，而 ELSFP 也会成为标准。 之后他们还能做别的事，比如购买 DSP 设计并开始整合技术栈。 不再是每个 10 美元的激光芯片乘 6，做成 50 到 100 美元的激光阵列，而是变成 400 美元的可热插拔模块。 基本上，他们的起点是掌握光子学里激光层面的一个瓶颈护城河…… 然后他们就能像 $LITE 当年那样，往下游做 EML 和光模块。 TLDR： Sivers（$SIVE）在世界正向 CPO 转型的这个时点，持有一个稀缺、独立的 InP 激光瓶颈点。 全球只有少数几家独立公司能做这件事，比如 $SIVE（例如住友、$LITE、$COHR）。 但当然，他们需要资金、需要正确的愿景，才能实现 TAM 扩张。
英文原文
Can't $SIVE just pull a $LITE and TAM expansion down the ELS stack? I'm getting early stage $LITE flashbacks (Oclaro, NeoPhotonics, Cloud Light)... Here's my thesis/idea: on how Sivers can capture the ~$300-$400 ELS (from $50-100 arrays) and become a $10B+ company. How this plays out: -> Sivers is needs funding. -> Get NRE prepay or PIPE (eg. $AXTI/Northland) -> Say "Hey, $AVGO and $LITE are going to choke your custom AI cluster margins". ----> get $META / $AMZN / $MSFT capital injection for multi-source supply for next ASIC clusters cause they're scared. Esp. with $NVDA funding $4B to securing supply from $LITE and $COHR... then PTSD from the current EML capacity bottleneck. It's easier than people think. -> US spinoff of photonics arm (markets don't like investing in legacy companies). - - - - - > This will likely work in 2026, bc CPO/photonics is hot, now and not like 2024-early 2025. -> Raise enough to automate wafer probers and fund the module NPI. -> Fab-lite test model: develop probe cards and test inhouse, but push capex to OSATs. -> Buy SiPh packaging IP (or just continue JDM with O-Net and $POET) -> Keep their Glasgow InP fab for in-house testing for iteration/raw epitaxial. -> Outsource lithography, etching, etc. stuff to Win Semi + others (don't do capex heavy things yourself) -> Then just $SIVE can just hand tested lasers and packaging IP to $FN And Poof. You have a independent, mass producible, high margin, ELS module as ELSFP becomes standard. Then they can do other stuff like buying DSP designs and start consolidating the tech stack. Instead of cheap $10 laser dies x 6 for a $50-$100 laser array, you have $400 hot swappable modules. Basically their starting point is owning one of the chokepoints moat in photonics on the laser level... Then they can go downstream like what $LITE did with EML and optical transceivers. TLDR: Sivers ( $SIVE ) holds a rare, independent chokepoint in InP lasers at the exact moment the world is transitioning to CPO. There's only a few independent companies in the world that can do this like $SIVE (eg. Sumitomo, $LITE, $COHR). But of course they need the funding to achieve TAM expansion and the right vision.
-
分享小账户分散化投资组合，涵盖AI半导体、光子学、机器人、加密等领域的高增长和杠铃策略标的
快速增长赛道： $AAOI - 光收发器2027年下半年收入增长10倍 $NBIS - 2026年第四季度收入增长10倍 $ARM - 全新AI CPU带来5倍收入增长 $MRVL - 来自$MSFT Maia芯片项目2-3倍收入增长 $AVGO - 长期持有超大规模云厂商ASIC $LITE - 长期持有OCS/谷歌TPU Win Semi - 对前沿行业的代工敞口 $TSEM - 长期持有光子业务，订单积压 SK Hynix - 存储业务敞口，极端营业利润增长 同时配置一些杠铃策略，远离超大规模云厂商资本支出（除亚马逊外）： $VNP - 长期持有西部供应链的稀土 $NEO (TCX) - 机器人供应链 $AMZN - 机器人/AI降低运营成本 $CRCL - 稳定币多头 $RDDT - 高得离谱的利润 $GLD - 安全对冲 $IBIT - 2028年减半 $CVX 看涨期权 - 石油对冲 也许还有长期配置（你知道迟早会来）： $INTC/$AMKR - 美国制造供应链 $SOI - 硅光子学/共封装光学(CPO)基板 $RKLB - 对太空行业的长期看涨期权 然后挑选一两支小盘股博暴涨： $SIVE - CW激光关键点或$I QE - Landmark重组重估，这是我最喜欢的两个 还有其他我提过的，比如$AEHR用于测试或$VPG用于Optimus 我自己在$AXTI等股票上的主动管理模式与其他人应该做的完全不同风险敞口配置。 在这种宏观环境下全仓投入高贝塔标的不太是最佳选择。
英文原文
Faster compounds: $AAOI - 10x revenue ramp from optical transcivers h2 2027 $NBIS - 10x revenue ramp Q4 2026 $ARM - 5x revenue growth from their new AI CPU $MRVL - 2-3x revenue growth from $MSFT Maia Ramp. $AVGO - Long hyperscaler ASIC $LITE - Long OCS / Google TPU Win Semi - Foundry exposure to frontier industries $TSEM - Long photonics, backlogged SK Hynix - Memory exposure, extreme operating income ramp With some barbell exposure away from Hyperscaler capex aside from Amazon: $VNP - Long term rare earths for Western Supply chains $NEO (TCX) - Robotics Supply chains $AMZN - Robotics/AI cutting opex $CRCL - Stablecoin long $RDDT - Ridiculously high profit $GLD - Safe Hedge $IBIT - Halving 2028 $CVX Calls - Oil Hedge And maybe long term (you know it's coming): $INTC / $AMKR- Made in America supply chains $SOI - Silicon Photonics / CPO substrates. $RKLB - Long term call on Space industry Then pick one or two small cap moonshots: $SIVE - CW Laser Chokepoints or $IQE for Landmark rerating on restructuring were my two favorites. There's others I've mentioned like $AEHR for testing or $VPG for Optimus. How I actively manage my own stuff from $AXTI and others is a lot different risk profile than what others should do. Going full port into high-beta in this macro environment is not the best idea.
-
分享一个股票配置框架示例：分高增长股、moonshot、安全股、长期国家安全股四层配置
我个人不会仅投资一个行业以分散风险。我提到光子学如$AAOI是因为我认为它有最高的短期上涨空间。但像$NBIS这样的公司，正如Jensen准确所说的，会"照顾好你"的长期表现。也许可以找一些高增长long标的，例如：$ARM - 从新AI CPU获得5倍收入增长；$NBIS - 到第四季度实现7-9B ARR增长带来10倍收入增长；$AAOI - 从光收发器需求获得10倍收入增长；$MRVL - 从$MSFT Maia ASICs获得2-3倍收入增长。选择一两个moonshot：我提到$SIVE是我最喜欢的，但鉴于其规模小，我不会投入太多集中度。然后用一些"更安全的标的"进行杠铃配置：$AMZN长期看涨，即使通过机器人/AI削减运营成本，但其走势像蜗牛；$RDDT长期看涨，因为它利润极高且今天正在产生巨额自由现金流。也许还有一些"超长期玩家"具有深远的国家安全优势，例如：$INTC的"美国制造"；$AMKR的"美国制造"等。这只是一个虚构的例子。
英文原文
I would personally not just invest in one sector for diversification sake. I talk about photonics like $AAOI recently because I see it as highest short-term upside. But others like $NBIS as Jensen accurately said "will take care of you" long term. Maybe figure out high growth longs for example: $ARM - 5x revenue from new AI CPU $NBIS - 10x revenue to q4 $7-9B ARR ramp $AAOI - 10x revenue ramp from optical transceiver demand $MRVL - 2-3x revenue ramp from $MSFT Maia ASICs Pick one or two moonshots: I mentioned $SIVE as my favorite, but given it's small size, I wouldn't put too much concentration into them and then barbell with some "safer plays" $AMZN long term I'm bullish on even from robotics/AI cutting opex though it moves like a slug $RDDT long term I'm bullish on from on just because it's ridiculously high profit and generating massive FCF today. and maybe some "long, long term players" that have deep national security benefits eg: $INTC for Made in America $AMKR for Made in America, etc. Just a made up example.
-
阐述AI基础设施供应链中不可替代公司的投资逻辑和超额收益来源
我的投资组合也因宏观因素出现回撤。年初至今收益现在是527%。在指数下跌7%之后。 - $AAOI“崩盘”从$30 -> $100 -> $96… - $LITE“崩盘”从$330 -> $800 -> $702… - $AXTI“崩盘”从$15 -> $70 -> $60… 我不会低估伊朗战争的影响。这会对流动性和能源产生严重影响，所以我已经降低了杠杆。但如果一家公司季度收入将从$1.34亿增长到预计每季度$15.4亿，这来自美国制造的光收发器(mount optical transceiver)产能爬坡…… 或者如果一家市值$36亿的小公司拥有超大规模云商光子学建设的材料供应链…… 或者一家市值$2.9亿的小型激光供应商$SIVE为$MRVL的共封装光学(CPO)项目或Jabil光收发器提供配套…… 或者像$LITE这样的光学巨头其EML产能已售罄到2028年…… 或者$SOI这家欧洲小公司提供硅光子学/共封装光学(CPO)所需的所有衬底。 或者$IQE这家欧洲小公司拥有超出其估值数倍的隐性反应釜产能…… 或者像SK海力士这样的内存公司预计三年内创造的利润将超过其当前市值…… 在市场认识到它们对供应链的重要性过程中，股价上涨途中会有大量波动。但如果你无法绕过它们来规模化AI，或者它们已被纳入$AMZN或$MSFT的供应链中…… 也许它们会跑赢市场？
英文原文
My portfolio has drawdowns from Macro as well. YTD is now 527%. After the index crashed -7%. - $AAOI “crashed” from $30 -> $100 -> $96… - $LITE “crashed” from $330-> $800 -> $702… - $AXTI “crashed” from $15 -> $70 -> $60… I’m not underestimating the War in Iran. This has serious consequences to liquidity/energy, so I’ve winded down margin. But if a company is going from $134m quarterly revenue to projected $1.54B a quarter from Made in America optical transceiver ramps... or if the tiny $3.6B company owns the materials supply chain for the hyperscaler photonics buildout... or a tiny laser supplier at $290m in $SIVE feeds to $MRVL CPO programs or Jabil transcivers... or a optical giant like $LITE is sold out of EML capacity until 2028... or a small European company in $SOI provides all the substrates required for silicon photonics / CPO. or a small European company in $IQE has latent reactor capacity multiple times what their valued at… or a memory company like SK Hynix is projected to make more than what their current market cap is in 3 years… There’s going to be tons of volatility on the way up, as markets realize their importance to supply chains But if you can’t bypass them to scale AI. Or they’re designed into the supply chains of $AMZN or $MSFT. Maybe they tend to outperform the market?
-
认为美股科技股卖压过头，SpaceX IPO 会制造流动性真空，短期别碰期权。
$MSFT、$GOOGL 和 $META 这轮美国科技股抛售，看起来已经过度了？ Mag7 大致跌幅： Microsoft 年初至今跌 -21.9% Google 年初至今跌 -7.69% $AMZN 年初至今跌 -7.36% Meta 年初至今跌 -8.54% 更糟的是，如果 SpaceX 以 1.7 万亿美元估值在快速纳入指数后 IPO。 大型 IPO 通常会制造巨大的市场流动性真空（对 SpaceX 来说当然是好事）。 不过，我不认为市场真的相信伊朗战争会很快结束。 -> 大盘股出现巨大流动性抽水， -> SpaceX IPO 还会带来更多抽水， -> 尽管基本面在改善 = 短期别碰期权。 长期来看，我预计 Mag7 会恢复。
英文原文
The US tech selloff from $MSFT, $GOOGL and $META looks really overdone now? Mag7 roughly down: Microsoft down -21.9% YTD Google down -7.69% YTD $AMZN down -7.36 YTD Meta down -8.54% YTD What makes things worse is SpaceX IPOing at $1.7T if they get fast-index inclusion. Mega-IPOs typically serve as a massive market liquidity vacuum (good for SpaceX though) However, I don’t think markets are truly convinced War in Iran is ending anytime soon. -> Massive liquidity drain from large caps, -> more upcoming soon with SpaceX IPO -> despite improving fundamentals = don’t touch short term options. Long term I expect Mag7 to recover.
-
博主回顾从业经历，从Intel到Berkeley研究RISC-V，最终创办金融科技公司。
我现在经营一家金融科技公司。 不过我之前确实在半导体行业有一些零散的经历？ 我的第一份工作是在 $INTC，但我干了很短时间就辞职了。在UC Berkeley花了一两年时间从事RISC-V相关工作，比如RocketChip、安全飞区(secure enclaves)等。 曾在某个人工智能实验室担任研究员，所以勉强算是相关领域？ 做过各种零散的事情，比如$AMZN机器人、强化学习(RL)无人机仿真，或半导体创业公司……所以对我来说是四处漂泊，但从来无法朝九晚五。 积累的知识确实会叠加！
英文原文
I run a fintech company now. But I did have some random experiences here and there in semis? My first job was $INTC but I quit after a rlly short time. Spend a year or two working on RISC-V related stuff like RocketChip, secure enclaves etc. at UC Berkeley. Was a researcher at some AI Labs for a tiny bit so it’s kinda adjacent? Did a bunch of random stuff like $AMZN robotics RL drone simulation or semi startups… so kinda all over the place for me, but never could do a 9-5. The cumulative knowledge does add up!
-
强调 SIVE 通过 CHIPS Act 和国安项目对美国很重要，并给出更高估值判断。
如果你到现在还不知道，$SIVE 通过 CHIPS 法案对美国国家安全很重要。 而且尤其是，它还是超大规模云 AI 供应链的光源。 到目前为止，他们已经拿了大约 1160 万美元的两笔 CHIPS 法案资助： 1. 和 $ERIC 以及 Raytheon $RTX 一起做 FR3 beamformer ICs 2. 和 $BA（LSE）BAE Systems 一起做 EW 技术 CEO 还暗示 2026 年会有更多来自 CHIPS 的资金。 小盘股几乎从来拿不到美国政府这种支持。 你手里有一家市值 3.3 亿美元的公司： 它对美国半导体业务的国家安全至关重要， 也对超大规模云供应链（$MSFT、$META、$AMZN、$ORCL 等）里的光子学至关重要， 通过 $MRVL Celestial、Ayar、Jabil、ONet 等链条。 关于扩产和 capex 的问题，现在已经由 Win Semi 合作基本回答了。 至于“等待 CPO”的机会成本，现在也被 OFC 上 Jabil 1.6T 可插拔模块放量这件事回答了。 $SIVE 是光子学里最被低估的宝石之一，我个人认为它今天就应该值 20 亿美元以上。
英文原文
If you didn't know by now $SIVE is important to US national security via CHIPS act. But especially, as the light source for hyperscaler AI supply chains. They got two CHIPS acts grants ~$11.6M so far: 1. FR3 beamformer ICs with $ERIC and Raytheon $RTX. 2. EW Tech with $BA (LSE) BAE Systems With more coming funding from CHIPS in 2026 hinted by the CEO. Small caps almost never get this sort of support from the US government. You have a company sitting at $330M: That's critical to US National Security for their Semi arm. and Critical to Hyperscaler supply chains ( $MSFT, $META, $AMZN, $ORCL, etc) for Photonics through $MRVL Celestial, Ayar, Jabil, ONet, and others... The main question regarding scale vs. capex is now answered by Win Semi partnership. And the question regarding "CPO waiting" opportunity cost is now answered by the Jabil 1.6T pluggable ramp from OFC. $SIVE of the most unknown gems in the photonics space and I personally think this company should be $2B+ today.
-
说 SIVE 因 Jabil、MRVL、Ayar、Onet 等链条而从高风险变成低风险高回报。
是的，我原帖里也公布了财务数据。2 月份的重组，让 $SIVE 变得更有吸引力。 所以我现在认为 $SIVE 已经是低风险、高回报，而不是以前那种高风险了。 在它被宣布成为 Jabil 的光源之后，它也成了我现在高信念的小盘股。 你有 $JBL 在填补 1.6T 可插拔光模块这个巨大的空缺，而不是一直等 CPO。 然后是 $POET -> $MRVL Celestial -> $AMZN、$MSFT。 再然后是 Ayar -> AiChip / 其他 -> $AMZN、$MSFT、$META 等。 最后是 O-Net -> 亚洲的 CPO 供应链。 另外，我不觉得在光子超级周期里，任何收入预测都应该是准确的，因为 TAM 扩张 / 量产放大本身就会改变一切。
英文原文
Yes in my original post, I published financials. The restructuring February, made $SIVE much more highly compelling. So, I see $SIVE being more low risk, high reward. Not high-risk anymore. And it's now my high conviction small cap after they were announced as the light source for Jabil. You have $JBL bridging the MASSIVE gap for 1.6 pluggable optical transceivers, rather than waiting for CPO. Then $POET -> $MRVL Celestial -> $AMZN, $MSFT. Then Ayar -> AiChip/others -> $AMZN, $MSFT, $META, and others. Then onet -> Asian supply chains for CPO. Also I don't think any revenue forecasts should be accurate given TAM expansion/ramp for photonics supercycles.
-
Sivers以2.5亿美元估值复制LITE/COHR模式，成为CPO/ELS时代的光源核心供应商，被严重低估。
估值约2.5亿美元的$SIVE|$SIVEF看起来是光通信半导体市场中结构性错误定价最严重的标的之一。 $LITE和$COHR目前拥有450亿美元以上的估值，主要因为它们控制了外部腔体激光器(EML)和垂直腔面发射激光器(VCSEL)的出货——这些器件流向-> InnoLight和Eoptolink，用于当前可插拔光收发器(supercycle)。 我会一直强调这个观点，直到市场理解： Sivers已经为下一个技术范式复制了这一高度盈利的Merchant Supplier（商业代工）模式： -> 共封装光学(CPO)和外部光源(ELS)。 同时还有： -> 激光器，某大型制造商如Jabil正在用于当前的1.6T升级周期。 Sivers预见到架构正在从EML转型，并成为连续波(CW)激光器的纯Play供应商。 当前周期（可插拔光模块）：LITE/COHR供应EML -> InnoLight/Eoptolink制造模块 -> $GOOGL、$META、$MSFT、$AMZN购买。 下一个周期（CPO/ELS）：Sivers供应CW DFB激光器 -> $POET、Ayar Labs -> 流向超大规模数据中心商如$AMZN、$META、$MSFT等。 但与建设大型晶圆厂、承担良率/产能爬坡风险不同： 他们转型为无晶圆厂模式，通过Win Semi外包生产大批量CW激光器，有效降低了规模扩张风险。 那为什么在光电子supercycle期间估值只有2.5亿美元？ 我的观点： 1. 没人知道这家公司。 2. 许多基金的投资限制阻止它们购买瑞典的小盘股。 然而，当你开始关注超大规模数据中心光供应链中那些不知名的上游标的时： 光源领域最顶层的公司是$SIVE。 一家2.5亿美元的小公司，置身于$MTSI、$LITE、$COHR、住友电工以及百亿至数百亿美元市值的现有巨头之间。 再说一遍…… 一家2.5亿美元的公司： 为Jabil的1.6T LRO光收发器模块提供核心光源。 Jabil是巨头。 同时还为Ayar、$POET、Enablence/$ONET以及许多其他名字（尚未公开披露）提供多元化供应。 $SIVE已成功打入超大规模数据中心商的供应链，成为未来光电子架构的核心光源供应商。 相对于未来市场空间扩张和产能爬坡来看，估值被严重低估了。
英文原文
< $SIVE | $SIVEF > at a ~$250M valuation looks like one of the most severe structural mispricing in the optical semi market. $LITE and $COHR, have $45B+ valuations today: Largely because of their control over EMLs and VCSELs that they ship to -> InnoLight and Eoptolink. For current pluggable transceiver supercycles. I will keep hammering this home until markets understand: Sivers has replicated this exact, highly lucrative merchant-supplier model for the next paradigm: -> CPO and ELS. And also: -> Lasers that a massive manufacturer like Jabil is using for the current 1.6T upgrade cycle. Sivers saw architectures shifting away from EML and became the pure play supplier for CW. Current Cycle (Pluggables): LITE/COHR supply EML -> InnoLight/Eoptolink build modules -> $GOOGL, $META, $MSFT, $AMZN buy them. Next Cycle (CPO / ELS): Sivers supplies CW DFB -> $POET, Ayar Labs -> flows to hyperscalers like $AMZN, $META, $MSFT and others. However, instead of building up mega fabs with yield/capacity ramp risk: They transitioned to an outsourced, fabless model for high-volume CW Lasers with Win Semi and effectively de-risked scaling. So why does it have a $250M valuation during the photonics supercycle? My opinion: 1. Nobody knows about it yet. 2. Many fund mandates prevents them from buying small caps in Sweden However, when you start looking at obscure upstream names in hyperscaler light supply chain: At the top for the light source sits $SIVE. A small $250M company among $MTSI, $LITE, $COHR, Sumitomo and the $10B-$40B+ incumbents. Again... A $250M company: Powers Jabil's 1.6T LRO optical transceivers modules. Jabil is massive. As well as Ayar, $POET, Enablence/ $ONET and many other names now (that are not disclosed) for diversification. $SIVE found their way to scale as the lightsource of hyperscaler supply chains for future photonic architectures. And looks incredibly undervalued, relative to forward TAM expansion and ramp.
-
用“AI 养牛”调侃 AI 造成的自动化失业会继续外溢到更多行业。
我当时在推演那家做“牛用 AI”的 20 亿美元公司会带来哪些二阶效应。 然后我突然意识到一件很残酷的事： 牧羊犬也要失业了。 AI 带来的岗位替代会波及一切。 $ORCL、$AMZN、$META 等公司都在因为 AI 自动化裁员。 现在连你家附近的狗也要失业了。 我的宠物鱼是不是也要下岗了？
英文原文
I was mapping second-order effects of the $2B company that does AI for Cows. I just realized something brutal: Shepard Dogs Are Going to Be Unemployed. AI displacement is coming after everything. $ORCL, $AMZN, $META, and others are laying off workers from AI automation. And now your local dogs will be unemployed too. Is my pet fish going to be out of a job too?
-
分析SIVE预收入仅$250M却或是超大规模数据中心未来光源，低估值蕴含巨大潜力。
今天关于 $SIVE 的一些想法，市值约 $250M： Sivers 是未来可能的共封装光学(CPO) + 激光阵列光源，为 $AMZN、$META 和 $MSFT 等超大规模数据中心提供动力。 以约 $250m 的市值... 从已确认的客户来看，估计映射如下： 1. Jabil ($JBL) LRO 光收发器（前身为 $INTC 的硅光子学业务）-> $AMZN、$META 及其他超大规模数据中心。 2. Ayar -> AIChip/GUC -> $AMZN 及其他超大规模数据中心。 3. Enablence -> O-Net -> 亚洲超大规模数据中心。 4. 以及其他未确认的客户。 人们一直在问关于今天波动性的无尽问题：如果一只股票能上涨 20%，它也能下跌 20%。 $AXTI 经常出现 +30%、-30% 这样的波动（例如财报 -30% 至 $20，现在在 $60）。 就像 AXT 一样，正如你所见，重要的是它有长期跑赢大盘的潜力。 这就是为什么在进入任何交易之前形成自己的信念很重要，这样你就不需要每天问我问题了，如果你自己做了这笔交易的话。 我个人持有长仓，我真的认为 Sivers 有潜力成为市值 $5B+ 的公司，尤其是获得 Win qualification 的话。 尤其是在今天一些分析师告诉人们卖出 $SIVE 获利之后： - 如果你看 $COHR、$LITE 和其他光源公司的估值，它们现在是 $40B+。 - $SIVE 预收入先进封装 counterparties 都是 $1B-$4B+。 - 而我们有 $SIVE 作为超大规模数据中心供应链的未来光源和 Jabil 光收发器： 市值约 $250M。 这只是我的个人意见，但将信念锚定在市值（$250M）——作为超大规模数据中心供应链的光源——比锚定在价格波动上更好... 尤其是当机构（约 14.1% 持有）能够通过震荡多数持股的股票来建仓时。 这就是为什么我早早分享了我关于 $AXTI 或 $SIVE 的想法，因为我相信散户投资者有可能抢先机构一步，抓住 AI 供应链中即将出现的瓶颈环节。 以及为什么我分享了 $SIVE 的超大规模数据中心供应链映射，这样人们可以基于未来光源增长形成自己的信念。
英文原文
My thoughts today on $SIVE, at a ~$250M MC: Sivers is the future likely CW + laser array light source powering hyperscalers from $AMZN, $META, and $MSFT. At ~$250m... From confirmed clients, est. mapping: 1. Jabil ( $JBL ) LRO Transceivers (Former SiPH $INTC) -> $AMZN, $META, and other Hyperscalers. 2. Ayar -> AIChip/GUC -> $AMZN and other hyperscalers. 3. Enablence -> O-Net -> Asian Hyperscalers 4. And other unconfirmed customers. People have been asking me endless questions about today's volatility: If a stock can go up 20% it can do up 20% too. $AXTI had those +30%, -30% movements all the time (eg. Earnings -30% to $20, now at $60). And like AXT as you've seen, what matters is it has the likely potential to outperform long term. That's why it's important to develop your own conviction before entering any trade, so you don't need to ask me questions every day if you take it yourself. I personally have a long position and I do think Sivers has the potential to be a $5B+ company with Win qualification. Especially amid some analysts post today telling people to take profits on $SIVE: - If you just look at $COHR, $LITE and other light source valuations today, they're $40B+. - $SIVE pre-revenue advanced packaging counterparties are all $1B-$4B+. - And we have $SIVE as the future light source for hyperscaler supply chains and Jabil transceivers: At ~$250M. Just my personal opinion, but it's better to anchor conviction to the marketcap ($250M) as the light source for hyperscaler supply chains, than price fluctuations... Especially when institutions (~14.1% owned) are able to to shake the tree of a majority owned stock to build a position. This is why I've shared my thoughts about $AXTI or $SIVE early since I think it's possible for retail investors to frontrun institutions for the up and coming chokepoints in AI supply chains. And why I've shared hyperscaler supply chain mapping for $SIVE so people can build their own conviction on future light-source growth.
-
认为 SIVE/SIVEF 本周大涨后仍被低估，是多家光子和云厂供应链中的关键激光源。
$SIVE <> $SIVEF 这周已经涨了 165%。 估值呢？大约 3 亿美元市值。 不过我可能是傻了，或者说 Sivers 其实就是当下光子领域最好的机会之一。 你买到的是 Jabil、Ayar、Poet（$MRVL Celestial）、O-Net 等公司的激光供应商： 而这些最终都会流向 $GOOGL、$MSFT、$AMZN、$META 的 AI 数据中心。 市值才 3 亿美元。 今天的 EML 激光供应商，比如 $LITE 到 $COHR，作为参照都在 450 亿美元以上。 这可能是未来上游光子供应链里最不被市场理解、却又至关重要的瓶颈之一。 而市场现在才刚开始给它定价。
英文原文
$SIVE <> $SIVEF is now up 165% this week. Valuation? ~$300M MC. However; either I’m dumb or Sivers is one of the best opportunities in photonics today. You get the laser supplier for Jabil, Ayar, Poet ( $MRVL Celestial ), O-Net, and others: That end up in $GOOGL, $MSFT, $AMZN, $META AI datacenters. At ~$300M. The EML laser suppliers today from $LITE to $COHR for reference are $45B+ This is one of the most undiscovered yet critical bottlenecks for future upstream photonics supply chains. That markets have only starting to price in today.
-
把 SIVE 视作 MRVL Celestial、Ayar、O-Net、Jabil 等链路的激光源，最终服务于大型云厂商。
它们字面上就是 $MRVL Celestial 通过 $POET、Ayar -> AI Chip / GUC、O-Net，乃至现在的 Jabil 所使用的光源。 所以终端客户大概率全是 $MSFT、$GOOGL、$META、$AMZN……也就是说，它们可能会成为整个超大规模云供应链，甚至亚洲供应链的未来潜在光源。 市值只有 2.8 亿美元，wtf？ 按 30 亿到 50 亿美元去看都不夸张。
英文原文
They're literally the light source for $MRVL Celestial through $POET, Ayar -> Ai Chip/GUC, O-Net, and now Jabil. So end users are probably all $MSFT, $GOOGL, $META, $AMZN... The future potential light source for the entire hyperscaler supply chain and even ones in Asia. At $280m MC, wtf? $3-5B MC for $SIVE near term is not far fetched.
-
Sivers成为Jabil 1.6T光收发器激光光源供应商，是其在CPO/硅光子供应链中的重大突破。
Sivers <$SIVE/$SIVEF>为Jabil（270亿美元市值）<$JBL>的1.6T长距离光(LRO)收发器提供光源。 作为激光光源供应商。 这无疑是$SIVE（2.8亿美元市值）公司历史上——也是今天$NVDA GTC大会上——最大的新闻…… Jabil从$AMZN、$GOOGL和$META等终端用户那里带来约300亿美元以上的收入。 以及$NVDA、Arista、$AVGO和Cisco等OEM厂商。 尤其是在收购$INTC的硅光子(SiPh)光收发器业务后，他们在该领域已是一级供应商(Tier 1)。 Sivers为Jabil提供连续波分布反馈激光器(CW DFB laser)光源，应用于硅光子学(Silicon Photonics)和共封装光学(CPO)及超大规模云服务商供应链： 对于一家市值仅2.8亿美元的公司来说，这是难以置信的，而且他们是光子学领域我见过的最不知名却最具潜力的公司。 $NVDA GTC的消息只是进一步巩固了这一投资论点。 我认为他们真的有机会成为下一个$LITE。
英文原文
Sivers < $SIVE / $SIVEF > powering Jabil ($27B MC) < $JBL > 1.6T LRO transcivers. As the laser light source. Is by far the biggest news from $NVDA GTC for Sivers ($280M MC) today... in history. Jabil brings in about $30B+ revenue from end users like $AMZN, $GOOGL, and $META. As well as OEMs like $NVDA, Arista, $AVGO, and Cisco. Especially after acquiring $INTC SiPH optical transceiver business, they're Tier 1 in the space. Sivers powering Jabil as the CW DFB laser source for Silicon Photonics and CPO and hyperscaler supply chains: At a $280M MC is incredible, and they're one of the most unknown yet highest potential company I've seen in the photonics sector so far. And the news from $NVDA GTC just cemented this thesis further. I think they genuinely have a shot at becoming the next $LITE.
-
博主认为SIVE成为Jabil光源供应商的消息将驱动SIVE股价大幅重新估值。
@mrsgayathridas 关于$IBKR的看法。$SIVE成为$JBL的光源这个新公告可能是迄今为止最大的新闻。我预计这只股票将因这个消息而获得巨大的重新估值。Jabil为$GOOGL、$META、$AMZN、$MSFT提供光收发器。他们是一个巨头的存在。
英文原文
@mrsgayathridas $IBKR. The new announcement $SIVE is the light source for $JBL is probably biggest news to date. And I expect the stock to get re-rated immensely from the news. Jabil powers $GOOGL, $META, $AMZN, $MSFT for optical transceivers. They’re a huge player.
-
Sivers被指定为Jabil 1.6T光收发器光源供应商，验证其CPO/硅光子学光源论点
Sivers < $SIVE / $SIVEF > 将为 Jabil ($JBL) 的下一代光收发器提供动力。 Sivers 被指定为 Jabil 旗舰 1.6T 模块的光源供应商…… 这可能是迄今为止最大的新闻？ 这验证了 Sivers 作为共封装光学(CPO)/硅光子学光源的下一个 $LITE 的论点。 来源：Semicap，$NVDA GTC。
英文原文
Sivers < $SIVE / $SIVEF >to power Jabil ( $JBL ) next-generation optical transceivers. The fact Sivers is named as the power source for Jabil's flagship 1.6T module… is probably the biggest news to date? This is thesis validation that Sivers is the next $LITE as the light source for CPO/Silicon Photonics. Source: Semicap, $NVDA GTC.
-
分析Sivers和IQE在硅光子学供应链中的卡位，看好其从传统业务向InP光子学的转型机会。
$SIVE/$SIVEF和$IQE。Sivers为CPO/硅光子学(SiPh)制造激光阵列和CW DFB激光器。并且已经嵌入超大规模云厂商如$AMZN和$MSFT集群的供应链，作为Captive模式如$MRVL Celestial和Merchant模式如Ayar的光源。然后他们通过晶圆(win)认证来提升产能，这样在内部扩产时不会消耗太多资本支出。$IQE是InP外延片(epiwafer)供应商，正在将legacy沉积反应器重构用于光子学供应链。他们已经是$LITE的已知供应商，另一大供应商是Landmark，估值约40亿美元（但产能潜力较小）。这基本上相当于将$IREN或$RIOT的比特币矿工转为HPC，但针对的是光子学。有相当多的债务，但这更多是在赌台湾销售能通过，以及管理层会转向InP生产以获得重新评级。所以这两家是我目前最看好的。
英文原文
$SIVE / $SIVEF and $IQE. Sivers makes laser arrays and cw dfb lasers for cpo/silicon photonics. And is embedded into hyperscaler supply chains like $AMZN and $MSFT clusters as the light source from captive models like $MRVL celestial from $POET and merchant models from Ayar. Then they’re up capacity through win qualification, so they don’t burn as much capex scaling in house. $IQE is InP epiwafer and refactoring reactors from legacy drag to photonic supply chains. They’re already a known $LITE supplier and the other biggest one is Landmark sitting around ~$4B (with less latent capacity). This is basically the equivalent of $IREN or $RIOT turning their Bitcoin miners into HPC, but for photonics. There’s a decent amount of debt, but this is more of a guess that their Taiwan sales will go through and management will pivot toward InP production for rerating. So those two are my favorites right now
-
Sivers涨幅29%达2.82亿市值，作者认为其作为CPO/硅光子学上游激光供应商，是下一个LITE，尚未被市场定价。
Sivers今天又涨了29%，市值达到2.82亿美元。我真心认为$SIVE可能是下一个$LITE。因为他们被定位为下一代超大规模算力供应链接构的光源。当scale up-scale out架构转向共封装光学(CPO)和硅光子学(Silicon Photonics)时——光子学超级周期(Photonics Supercycle)的开端。那些从$POET和Ayar购买并封装其激光阵列/cw DFB激光(连续波分布反馈激光器)的公司——估值约10亿至40亿美元以上（不包括O-Net或未披露的公司）。在当前可插拔光收发器(Pluggable Transceiver)周期中：激光供应商通常比其购买光源的封装对手方获得更高的估值溢价。所以，我确实认为$SIVE在很大程度上被市场忽视，远未反映在股价中。（该推文引用了@aleabitoreddit的推文，仅供理解语境）：$SIVE是CPO和硅光子学的上游激光供应商。他们是未来类似$COHR/$LITE的光源供应商，可能用于：- $AMZN Trainium集群- $MSFT Maia集群以及可能的其他超大规模算力商如$META MTAI和$GOOGL TPU集群。目前市值约2亿美元。关联映射（推测性）：$SIVE（光源）-> $POET（光学中介层）-> $MRVL（可能是Celestial内部供应）-> $MSFT Maia + $AMZN Trainium。$SIVE（光源）-> Ayar -> AiChip -> $AMZN Inferentia/Trainium。$SIVE（光源）-> Enablence -> O-Net -> ？亚洲超大规模算力商。持续进行中：$SIVE（光源）-> Ayar -> GUC -> ?（Google $TPU）。$SIVE（光源）-> Ayar（TeraPHY/SuperNova）-> Wiwynn（内部CPO）-> ?（$MSFT, $META历史上Wiwynn最大的客户）。由于类似$MRVL Celestial这样的内部供应模式，他们获得了免费搭乘。然而，他们确实在商业模式下与Lumentum、Coherent和$MTSI多源自ELS竞争，并与Ayar一起获胜。但无论哪种方式他们都能赢。对于大规模生产爬坡，其中很大一部分取决于正在进行的Win半导体验证，但这可能是一个重要指标。再强调一下，供应链BOM（物料清单）是极其机密的。$AMZN永远不会告诉任何人"嘿，我们用$SIVE"。但如果你把1+1+1+1+1放在一起，你可以拼凑出可能的供应商。大多数人看到"Poet Starlight"使用$SIVE，或Ayar使用$SIVE。但没有映射所有多跳关系来看它们最终去向哪里。我确实认为$SIVE是一个非常未被发现的机遇，作为未来硅光子学领域可能的迷你版$LITE，当时市值2亿美元。作为未来CPO/硅光子学规模扩展的上游激光供应商，用于cw DFB激光器和激光阵列的超大规模算力供应链接构。
英文原文
Sivers is up another 29% to a $282M marketcap today. I genuinely think $SIVE could be the next $LITE. As they’re positioned as the light source for the next generation of hyperscaler supply chains. When scale up-scale out architectures transition to CPO & Silicon photonics. For the start of next Photonics Supercycle. The companies that buy and package their laser arrays/cw dfb lasers from $POET to Ayar? Are worth ~$1B-$4B+ (not including o-net or undisclosed). With current pluggable transceiver cycles: Laser suppliers typically command higher valuation premiums than their packaging counterparties that buy the light source. So, I do think $SIVE is largely undiscovered by markets and far from being priced in.
-
用设计/组装、激光/收发器两类公司估值作比较，推导 SIVE 与 AAOI 的合理市值区间。
在设计 / 组装这一侧： Eoptolink（中国）：约 500 亿美元市值，约 53 亿美元营收 Innolight（中国）：约 840 亿美元市值，约 110 亿美元营收 $FN 这类低毛利组装公司约 200 亿美元市值，对应约 40 亿美元预期营收。 在激光 / 收发器这一侧： $LITE 约 470 亿美元市值：到 2026 年末营收将超过 30 亿美元。 然后是 $AAOI，现在月收入 3.75 亿美元左右，供应链位于激光 / 设计 / 组装这一层，目标年化营收约 45 亿美元 + 继续增长。 按这个逻辑，$AAOI 估值到 130 亿美元似乎合理，因为它还没真正到那个数字。 而 $SIVE，你只要看 CPO 的爬坡去猜就行。 通常先进封装公司（至少在当前周期里）会比它们对应的激光公司拿到更低的估值，除非它们自己也做设计。 $SIVE 给 $POET / Ayar 等公司供激光，而后者估值都在 10 亿到 30 亿美元+。 所以 $SIVE 至少也应该有 10 亿美元左右的下限。 这只是拿相似公司的估值 / 营收预测做比较，然后再做一个估算。
英文原文
For design/assembly Eoptolink (China): ~$50B MC: ~$5.3B revenue Innolight (China): ~$84B MC ~$11B revenue $FN for low margin assembly ~$20B MC, ~$4B projected revenue. For lasers/transceiver: $LITE ~$47B MC: ~$3B+ EOY 2026. Then you have $AAOI, $375M/month for laser/design/assembly supply chain, ~$4.5B ARR target + growth. Maybe ~$13B sounds reasonable since it hasn't hit those numbers yet. It's discretionary. As for $SIVE, you can just look at CPO ramp and just take a guess. Usually advanced packaging companies (at least in current cycles), have around the less valuation as their laser counterparts (Unless they do design). $SIVE supplies lasers to $POET/Ayar both of which are $1-$3B+ valuations. So $SIVE at least should be ~$1B low end. This is just comparing similar company valuation/revenue projections, then making an estimate.
-
用容量和供应链地位推演 AAOI 与 SIVE 的潜在市值上修空间，认为光学超级周期极其夸张。
我觉得市场应该这样给它们定价： $AAOI：从 93 美元到 162 美元（约 130 亿美元市值），这是在他们今天公布新产能预测之后的合理水平。 而 $SIVE 至少也应该从 7.7 美元到 38.5 美元，按约 11 亿美元市值去看？ 如果其中一个到 2027 年末扩产到约 19.7 亿美元的产能（基本上就等于营收，因为超大规模云厂商会把能买的都买掉）， 那另一个又是从 $AMZN Trainium 到 $MSFT Maia 集群的超大规模云供应链里，最可能的激光供应商？ 至少也该把未来营收增长计进去。 这还不算执行不确定性，也没给其他像 $LWLG 这类公司加上溢价（它们也已经有 11 亿美元以上市值）。 如果 Win Semi 认证了 $SIVE，而 $POET / Ayar 等继续扩产，我觉得 $SIVE 的估值从这里 20-30 倍都不是没可能？ 如果 $AAOI 真按每月 3.78 亿美元的目标跑，股价从这里 5 倍到大约 300 亿美元市值也完全可能。 这些潜在目标价足以说明光学超级周期有多夸张（像存储一样），只是最终要看每家公司执行得如何。
英文原文
I feel like markets should value: $AAOI $93 -> $162 (~$13B MC) after their new capacity projections today. and $SIVE should be valued at 7.7 -> 38.5 ($1.1B MC) at least? If one is ramping to ~$1.97B capacity EOY 2027 (which is basically revenue, since hyperscalers are buying anything they can make) Then the other is the likely laser supplier to hyperscaler supply chains from $AMZN Trainium to $MSFT Maia Clusters? At the very least, should price in forward revenue growth. This is including execution uncertainty, and without premiums assigned to others like $LWLG at $1.1b+. If Win semi qualifies $SIVE and $POET/Ayar/and others scale up. I feel like $SIVE valuation could easily 20x-30x from here? If $AAOI hits $378M/month projections, could easily 5x from here to a ~$30B MC. These possible price targets is how insane the optical supercycle is (like memory), but largely depends on how each company can execute.
-
强调自己会先用前瞻基本面和供应链终端需求建模，而不是先看图形。
我会先把前瞻性的基本面（营收、利润）建模进去，然后再让市场去给 $AAOI 这样的公司定价。 或者像 $SIVE 那样，去看供应链终端客户，比如 $AMZN 或 $MSFT 对光源的需求。 市场里大部分 alpha，本来就是这么找到的，不是靠画一条歪歪扭扭的线。
英文原文
I try to model in forward fundamentals (revenue, profit) before markets price the company in with stuff like $AAOI. Or with $SIVE, looking at supply chain end users like $AMZN or $MSFT for light sources. That's how most of the money is to be made in markets finding alpha, not drawing a swiggly line on a chart.
-
用 ASP 和产能推演 AAOI 的营收爬坡，认为市场严重低估其容量兑现速度。
$AAOI 以 64.9 亿美元市值看起来明显低估了。 如果我们今天按 ASP 和他们最新的产能预测来建模： 来自产能的营收： Q2 2026：约 3.121 亿美元 Q4 2026：约 14.1 亿美元 Q2 2027：约 15.3 亿美元 Q4 2027：约 19.7 亿美元 这就是非常夸张的爬坡（按约 34-40% 的毛利率估算）。 ASP 是按 LightCounting、Dell'Oro Group 和 Yole 的资料，以及 ELSFP 模块定价推出来的（其中最有猜测性的就是 ELSFP 模块）。另外也参考了一些卖方模型，比如 Raymond James、B. Riley、Northland Capital 和 Goldman Sachs。 大规模订单的精确合同价格并不知道，所以这部分有猜测成分。 但 Q2 的量 * ASP 估算，实际上和他们到 2027 年 Q2 每月 3.78 亿美元的目标是对得上的。 再说一次，你可能会问：这只是产能，不等于营收吧？ 但从 $AMZN 到 $MSFT 的超大规模云厂商，正在把 $LITE 到 $COHR 这些公司能做出来的任何产能都买走，而且是提前很多年就开始买。 这也包括 $AAOI 在他们之前的财报电话会上提到的部分。
英文原文
$AAOI looks very undervalued at $6.49B. If we model ASP and their newest capacity projections today: Revenue from Capacity: Q2 2026: ~$312.1M Q4 2026: ~$1.41B Q2-2027: ~$1.53B Q4-2027: ~$1.97B This is absurd ramp (off ~34-40% est. gross margins). ASP modeled off (LightCounting, Dell'Oro Group & Yole, pricing for ELSFP modules is the most speculative). And some sell-side models (from firms like Raymond James, B. Riley, Northland Capital, and Goldman Sachs). Exact contract pricing for massive volume orders is not known, so this is speculative. But the Q2 volume * ASP estimates actually align with their $378M/month target Q2-2027. Again, you might be wondering? This is capacity, doesn't translate into revenue right? Hyperscalers from $AMZN to $MSFT are buying any capacity any of these companies from $LITE to $COHR can make, years out. This includes $AAOI from their former earnings call.
-
$SIVE作为2亿美元市值的硅光子学上游激光供应商，是未被发现的机遇，供应超大规模业者CPO供应链。
$SIVE是CPO和硅光子学的上游激光供应商。 他们是未来可能像$COHR和$LITE那样的光源供应商，用于： - $AMZN Trainium集群 - $MSFT Maia集群 以及可能还有其他超大规模业者的$META MTAI和$GOOGL TPU集群。 市值约2亿美元。 关系映射（推测性）： $SIVE（光源）-> $POET（光学中介层）-> $MRVL（可能是Celestial专用）-> $MSFT Maia + $AMZN Trainium $SIVE（光源）-> Ayar -> AiChip -> $AMZN Inferentia/Trainium $SIVE（光源）-> Enablence -> O-Net -> ? 亚洲超大规模业者 正在进行中： $SIVE（光源）-> Ayar -> GUC -> ?（Google $TPU） $SIVE（光源）-> Ayar（TeraPHY/SuperNova）-> Wiwynn（专用CPO）-> ?（$MSFT、$META历史上是Wiwynn最大的客户）。 由于$MRVL Celestial等专用模式的存在，他们可以免费搭便车。 然而，他们在与Lumentum、Coherent和$MTSI的多源ELS竞争中对抗Ayar并获胜，即使在商业模式下也是如此。 但无论哪种模式他们都能获胜。 对于大规模生产爬坡，其中很大一部分取决于正在进行中的Win半导体认证，但这可能是一个重要指标。 同样，供应链BOM（物料清单）极度保密。$AMZN永远不会告诉任何人"嘿，我们使用$SIVE"。 但如果你把1+1+1+1+1加起来，你可以拼凑出可能的供应商。 大多数人都知道"Poet Starlight"使用$SIVE。或者Ayar使用$SIVE。但没有将所有多跳关系映射到最终目的地。 我真的认为$SIVE是一个极未被发现的机遇，可能成为下一个小型$LITE，用于硅光子学，在2亿美元市值下。 因为他们是未来CPO/硅光子学规模化扩展中超大规模业者供应链的上游激光供应商，cw DFB激光器和激光阵列用于扩展。
英文原文
$SIVE is the upstream laser supplier for CPO and Silicon Photonics. They're the likely $COHR / $LITE type future light source for: - $AMZN Trainium Clusters - $MSFT Maia Clusters and possibly other hyperscalers like $META MTAI and $GOOGL TPU clusters. At a ~$200M MC. Relational Mapping (speculative): $SIVE (light source) -> $POET (optical interposers) -> $MRVL (Likely Celestial Captive) -> $MSFT Maia + $AMZN Trainium. $SIVE (light source) -> Ayar -> AiChip -> $AMZN Inferentia/Trainium $SIVE (light source) -> Enablence -> O-Net -> ? Asia Hyperscalers _ Ongoing: $SIVE (light source) -> Ayar -> GUC -> ? (Google $TPU) $SIVE (light source) -> Ayar (TeraPHY/SuperNova)-> Wiwynn (captive CPO) -> ? ( $MSFT, $META historically Wiwynn's largest clients). Because of captive models like $MRVL Celestial, they get a free ride. However, they do compete multi-source ELS against Lumentum, Coherent, and $MTSI with Ayar and win anyway in merchant models. But they win either way. For high-volume production ramp up, a large part of it depends on the ongoing Win semi qualification, but this will likely be a large indicator. Again supply chain BOM is extremely confidential. $AMZN will never tell anyone "Hey, we use $SIVE ". But if you put 1+1+1+1+1 together, you can piece together the likely suppliers. Most people see "Poet Starlight" uses $SIVE. Or Ayar uses $SIVE. But don't map all the multi-hop relations to see where they end up. I do think $SIVE is an extremely undiscovered opportunity as the next possible mini $LITE for Silicon Photonics at $200m MC. As they're the likely upstream laser supplier for hyperscaler supply chains for future CPO/Silicon Photonics scale up with cw dfb lasers and scale out with laser arrays.
-
认为 AI 已引发不可逆的岗位替代，投资相关基础设施和瓶颈是对冲失业冲击的主要方式。
这个消息真的很令人心碎： $META 裁员 20% $ORCL 裁员 $AMZN 长期将有 60 万名工人被机器人和 AI 取代。 这是一种反乌托邦未来。 企业在没有人工劳动成本的情况下，仍然能赚创纪录的利润。 唯一能从中受益的办法： 把 AI 当成对冲工具去投资。 未来几年，想逃离 AI 造成的永久底层阶级，感觉最主要的办法就是这样。 AI 产生的股东权益回报，将流向股东。 而那些没投资股票、只能靠工资过活的人之间的差距会继续拉大。 这不是未来，而是已经在发生。 - Opus 4.6 已经足够替代今天大多数软件工程师。 - Waymo 已经开始在旧金山这样的地方替代出租车司机。 - 我们知道 $TSLA 的人形机器人也快来了，因为它们在中国已经普及。 这件事现在就在发生。 伊朗的扰动只是 AI 建设加速过程中的短期事件。 AI 已经到达拐点，而且似乎不可避免。 你已经看到美国就业修正接近减少 100 万，这非常惊人。 而且我们正在看到，新一代 LLM 正在由它们前一代模型自己构建，AI 正在逼近奇点（AI 递归增长）。 如果你要投资那些运行 AI 所需的算力和硬件： 数据中心 / 电力 / 电网板块： $NBIS、$XLU、$VRT、$BE 光子板块，负责扩展 AI： $LITE、$COHR、$AAOI、$TSEM 半导体板块，负责芯片： $NVDA、$TSM、$ASML、$INTC 存储板块，负责芯片： $MU、$SNDK、SK 海力士、三星 ASIC，负责超大规模云 AI 推理： $AVGO、$MRVL、Mediatek 良率 / 测试板块，确保芯片能工作： $TER、$AEHR、Advantest 以及 AI 所需的原材料或基板： $AXTI、$COPX、$SOI 还有很多其他公司，都在成为对抗大规模 AI 失业的单一、最大对冲工具。 谁掌握了算力的生产资料（瓶颈、材料、数据中心）： 谁就掌握了 AI 的未来。
英文原文
The news is pretty heartbreaking: $META 20% layoffs $ORCL layoffs $AMZN 600,000 workers long term layoffs as they get replaced by robotics and AI. This is a dystopian future. Companies end up with record profits, without the cost of human labor. The only way to benefit: Investing in AI as a hedge. The next few years feels like the main way to escape the permanent underclass, caused by AI displacement. The return on equity derived from AI will go to the shareholders. While the gap between those who live paycheck to paycheck, not invested in stocks. Will continue to grow. This is not the future. - Opus 4.6 is good enough to replace most software engineers today. - Waymo has started to replace taxi drivers in places like SF today. - We know $TSLA Humanoids are coming next as they’re widespread in China, today. This is happening now. Disruptions in Iran are only temporary to the accelerating AI buildout. AI has hit the inflection point, and looks inevitable. You’re already seeing US job revisions down close to 1 Million, which is staggering. And we’re seeing the newest LLMs be built by their previous models, as AI approaches the singularity (AI led recursive growth). Investing in where the compute and hardware needed to run the AI: From the datacenter/power/grid sector: $NBIS, $XLU, $VRT, $BE Photonics sector needed to scale AI: $LITE, $COHR, $AAOI, $TSEM Semi sector needed for the chips: $NVDA, $TSM, $ASML, $INTC Memory sector for the chips: $MU, $SNDK, SK Hynix, Samsung ASICs for hyperscaler AI inference: $AVGO, $MRVL, Mediatek Yields sector to make sure the chips work: $TER, $AEHR, Advantest Along with the raw materials or substrates needed for AI: $AXTI, $COPX, $SOI And many others become the single, largest, hedge against widespread AI displacement. Whoever owns the means of compute (bottlenecks, materials, datacenters): Owns the future of AI.
-
认为 AAOI 已经进入多家大厂路线图
@Sher_Lockknows 没有。但 $AAOI 其实不太需要那种催化，因为他们大概率已经进入了 $AMZN、$MSFT、$ORCL / $META 的路线图。
英文原文
@Sher_Lockknows No. But $AAOI doesn’t need it since they’re likely in $AMZN, $MSFT, and $ORCL / $META roadmaps.
-
认为 Nebius 是下一家 AWS 级超大规模云
最新消息：Nvidia 向 $NBIS 投资了 20 亿美元。 我确信 Nebius 终将成为下一家 Amazon 级别的超大规模云。 因为它其实是 5 家业务的组合，而且这些业务的部分总和都在以三位数同比增长： - Nebius：AI Cloud，7 到 90 亿美元 ARR，正朝着下一个 AWS 发展 - ClickHouse：支撑大多数财富 500 强公司的数据库，上一轮估值 150 亿美元 - Avride：Robotaxi 和 FSD-4 级技术，现在正和 Uber 一起实车放量 - Toloka：数据标注平台，得到了 Jeff Bezos 的资金支持，也被 $AMZN 使用 - TripleTen：教育平台，虽然没那么性感，但增长也很快 另外还有像 SWE-rebench 这类广泛使用的 benchmark 之类的 side quests。 Jensen 对 Nebius 的 $NVDA 投资表达的也是同一个意思：Nebius 站在 AI 前沿，正引领浪潮去满足不断增长的算力需求。
英文原文
Just in: Nvidia has invested $2B into $NBIS. I'm convinced Nebius is the next Amazon-level hyperscaler over time. As they're 5 companies all growing sum-of-parts triple digits Y/Y: - Nebius: $7-9B ARR from AI Cloud as the next AWS. - Clickhouse: DB powering majority of fortune 500 companies, last valuation at $15B - Avride: Robotaxi and FSD-4 level technology, now scaling up live with Uber. - Toloka: Data Labeling platform, funded by Jeff Bezos and used by $AMZN - TripleTen: Education platform, not as sexy as the rest, but still fast growing. With sidequests from the widely used benchmarks like SWE-rebench. Jensen's $NVDA investment into Nebius shares the same sentiment: Nebius is at the forefront of AI and leading the wave to meet the growing demand for compute.
-
AXTI与住友垄断InP衬底，成AI光子化关键瓶颈。
匿名者，你听了我关于 $AXTI 的论点吗？ 警告：整个 AI 行业可能会受到两家公司的瓶颈制约： 1. $AXTI（市值 7 亿美元） 2. $SMTOY（市值 317 亿美元） 这两家公司控制了全球 60–70% 以上的磷化铟(InP)衬底市场。 未来的 $NVDA、$GOOGL TPU v7 集群、$META、$MSFT、$AMZN 超大规模计算集群都需要基于磷化铟(InP)的激光器和接收器。 $AVGO、$LITE、$COHR 使用磷化铟(InP)用于 800G/1.6T 光收发器中的电吸收调制激光器(EML)、分布反馈激光器(DFB)及其他光基础设施。 如果没有磷化铟(InP)衬底，供应链就会停滞。 在查看了从 TPU 到 Maia 的物料清单(BOM)后，看起来未来的专用集成电路(ASIC)、GPU 和超大规模部署严重依赖光子学。 而这两家供应商可能会冻结全球磷化铟(InP)衬底市场，涵盖几乎所有领域： - 超大规模计算光学（TPU 集群等） - 光收发器（5G、数据） - 激光雷达(LiDAR)（Robotaxi、无人机、军事） - 光模块（互连集群） - 硅光子学激光芯片（Nvidia 未来的共封装光学(CPO)以及 Intel/Broadcom 硅光子学引擎使用磷化铟(InP)连续波激光阵列。） 由于这些公司占据了市场供应的大多数： - AXTI（估计 ~30–35%） - 住友（估计 ~30%） - JX 日本（估计 10-15%） 仅此而已。（例如，Yole 2021 年的行业报告指出“住友电工 + AXT 共同拥有‘超过 75%’的磷化铟(InP)衬底市场”） 超大规模计算/AI 正在向光子学转变，但整个 AI 行业是脆弱的。 如果 $AXTI 或 $SMTOY 中的任何一家停止供应材料，整个未来的 AI 建设就会瘫痪。更疯狂的是，一家 7 亿美元的公司可能成为这一切的中心。 随着 AI 行业向光子学转变，磷化铟(InP)衬底可能会与高带宽内存(HBM)一起成为最大的瓶颈之一。
英文原文
Hope you listened to my $AXTI thesis anon? https://t.co/d5eg0BIhlG
-
HIMS价值源于分销网络，且成功完成全球DTC并购。
@iraw000 我的观点是，$HIMS 的价值源于其分销网络，类似于 $AMZN 的配送体系，而非其销售的商品（如电脑、书籍）。此外，他们成功在全球范围内收购了主要的医疗 DTC（直接面向消费者）公司，且未遭遇监管阻碍。
英文原文
@iraw000 My argument is that $HIMS value is derived from their distribution network like $AMZN delivery. Not by the drugs they sell (eg. computers, books). And they managed to acquire the major healthcare DTC companies globally without getting blocked.
-
$AAOI获超大规模客户大额订单，基本面验证优于内部人抛售担忧。
“$AAOI 收到来自主要超大规模客户的首笔 1.6T 数据中心光收发器(Transceivers) 2 亿美元批量订单” “我们预计很快将成为美国最大的 800G 和 1.6T 光收发器产能基地” 这逐渐证实了他们盈利预测中每月 3.78 亿美元的收入预期。 人们曾对内部人抛售感到恐慌，但关键在于执行力和基本面。 尤其是当他们获得来自 $AMZN、$META 或 $GOOGL 等超大规模客户(Hyperscalers)的批量订单时。
英文原文
$AAOI Receives First $200M Volume Order of 1.6T Data Center Transceivers from Major Hyperscale Customer” “We expect that we will soon have the largest production capacity for 800G and 1.6T transceivers in the U.S” This is gradual confirmation of $378m/month revenue from their earnings projections. People were panicking over insider sales, but what matters is execution and fundamentals. Especially when they’re getting volume orders from Hyperscalers like $AMZN, $META, or $GOOGL
-
分析AAOI估值与基本面，建议结合营收和供应链地位而非仅看图表。
如果 $AAOI 股价上涨 100%，并声称明年每月营收将达 3.78 亿美元，这得益于 $META、$MSFT 和 $AMZN 购买其能生产的所有供应品。而它现在的估值为 70 亿美元。我不会只看图表，还会关注营收预测、毛利率、在供应链中的地位以及预测数据。
英文原文
If $AAOI ran up 100% and stated they're doing $378M/month revenue next year off likely $META, $MSFT, and $AMZN buying any supply they can make. And they're valued at $7B now. I wouldn't look at the chart alone, would look at revenue projections, gross margins, position in supply chains, and forecasts.
-
NVDA锁定EML产能，光子学上游材料受制于中国，COWOS/HBM/光子学成瓶颈。
$NVDA 去年锁定了大部分光模块(EML)产能，导致其他超大规模云服务商面临巨大瓶颈。然而，光子学整体是一个材料问题，最上游的材料由中国控制。这并非英伟达能单方面决定的。不过，他们通常是最早布局的，这给 $META、$MSFT、$AMZN、$GOOGL 留下了更多问题，因此本文指出共封装光学(COWOS)、高带宽内存(HBM)和光子学可能是瓶颈。
英文原文
$NVDA secured most of the EML capacity last year and caused a huge bottleneck for other hyperscalers. However, photonics as a whole is a materials problem, the most upstream materials controlled by China. Not exactly something Nvidia can decide. However, they're usually the first to things and that leaves more problems to $META, $MSFT, $AMZN, $GOOGL afterward, hence the point of the post signaling cowos, hbm, and photonics are probably the bottlenecks.
-
超大规模云厂商资本支出未放缓，光学组件产能已满，预示AI光子学瓶颈将至，建议持有。
情绪（如战争/石油/LNG）决定短期价格，基本面决定长期价格。 我看不到那些关键的超大规模云服务商（$GOOGL, $MSFT, $AMZN 等）的资本支出有任何放缓迹象。 而且，在光学组件全面扩产之前，从 $AAOI 到 $LITE 的所有产能已在涨价前满负荷运转。 当新的 AI 集群架构采用光子学进行扩展时，这就是即将出现的瓶颈的定义。 如果这些公司的基本面没有发生任何破坏性变化，可能最好持有以度过短期波动。
英文原文
Sentiment (eg. War/Oil/LNG) dictates short term prices, fundamentals dictate prices in the long run. I see zero slowdown from the hyperscaler capex that matters ( $GOOGL, $MSFT, $AMZN, etc). And before the ramp for optical all their capacity from $AAOI to $LITE are already maxed out before price hikes. This is the definition of an upcoming bottleneck when new AI cluster architectures use photonics to scale. If there's been nothing breaking the fundamentals of any of these companies, probably better to hold through short term volatility.
-
警惕与股票“结婚”，投资逻辑变化时应果断调整持仓。
大多数 X 用户都会犯与股票“结婚”的错误。 如果你的投资逻辑因 $HIMS 或 $IREN 发生实质性变化，你的持仓也应随之调整。 如果去年你对 $IREN 的看多逻辑是通过托管(Colo)业务变现 3GW 产能。 那么如果该公司： -> 在 122 亿市值下不可避免地向你稀释 60 亿美元 -> 在每次反弹中将这些股份抛向公开市场 -> 转向 GPU 业务 你的逻辑已经变了。 这很可能是一个退出以追求更具非对称性机会的时机。 如果你对 $HIMS 的逻辑是它是医疗界的亚马逊。 但如果他们被 $NVO 和美国政府起诉至绝境，退出是合理的。 但如果随着 $NVO 撤诉并达成合作，逻辑重新成立，那么再次做多也没有问题。 每个月都有许多因催化剂或基本面而改变的事情。 如果你仍在为必然发生的、针对你 $IREN 持仓的 60 亿美元新股抛售而欢呼。 而你唯一的理由就像 $AMC 投资者那样“信任管理层”。 也许现在是时候问问自己： 你是否因为与股票“结婚”而忽略了所有危险信号？
英文原文
Most people on X make the mistake of getting married to a stock. If your thesis materially changes with $HIMS or $IREN, so should your position. If the bull case with $IREN last year was monetizing 3GW capacity through colo. Then if the company: -> inevitably dilutes you $6B off a $12.2B MC -> sells those shares into the open market in every rally -> pivots to GPU offerings Your thesis has changed. And it’s likely a time to exit to pursue more asymmetric opportunities. If your thesis with $HIMS was that they’re the Amazon of healthcare. But they get sued to oblivion by $NVO and the US gov. It’s respectable to exit. But if the thesis is back online given $NVO dropping their lawsuit and partnering up, then there’s nothing wrong with going long again. A lot of things change every month with catalysts or fundamentals. If you’re still cheering on an inevitable $6B worth of new shares getting sold against your $IREN positions on every rally. And your only qualifier is “Trusting in Management” like $AMC investors. Maybe it’s a good time to ask yourself this: Are you ignoring every red flag because you’re married to the stock?
-
因巨额稀释及风险收益失衡，作者看空 $IREN 并推荐 $CIFR 和 $NBIS。
我看不到任何持有 $IREN 的 compelling case（有力理由）。 尤其是考虑到在 $128亿市值下新增 $60亿的股票稀释。 人们可能会看到市值膨胀至 $200-$250亿。 但他们的股份价值会随时间递减。 风险收益比已经不再存在。 像 $CIFR 这样的公司，通过 Fluidstack 为 $AMZN 和 $GOOGL 提供 Colo（数据中心托管）模式，提供了更具非对称性的上行空间。 而像 $NBIS 这样的公司则提供更好的多元化（Robotaxis（自动驾驶出租车）、ClickHouse），执行风险更低，且资本支出得到良好支持。 对于 $IREN，无法合理化接近市值一半的稀释并为此欢呼。 即使有新的 Hyperscaler（超大规模云服务商）交易，风险收益比也不存在。 我去年对 $IREN 看多，但已卖出。持有者经历了从： -> 3GW 容量，轻资产，Colo 模式（我曾看好） -> 为 $MSFT 购买 GPU 并为其转型的执行风险欢呼（我不看好） 到 -> 为 $60亿新股稀释欢呼，并在公开市场上被反向抛售。（AMC 接盘者领地） 如果你努力证明新的 ~50% 稀释是好事，且在你持仓被抛售时信任管理层： 抱歉告诉你，你现在加入了数据中心领域的 $AMC 等价俱乐部。 市场上有更好的多头标的。
英文原文
I see zero compelling case to hold $IREN. Especially given the new $6B share dilution at a $12.8B marketcap. People will likely see that marketcap inflate to $20-$25B. But the value of their shares decrease over time. The risk reward is just not there anymore. Companies like $CIFR offer much more asymmetrical upside given their colo model for $AMZN and $GOOGL through Fluidstack. And companies like $NBIS offer much better diversification (robotaxis, clickhouse), derisked execution, and are well supported for capex. With $IREN, there’s no way to justify being diluted close to half the market cap and cheering that on. Even with a new hyperscaler deal there’s risk reward is just not there. I was bullish $IREN last year but sold it. As holders went from: -> 3GW capacity, asset lite, Colo model (was a fan of this) -> Buying GPUs for $MSFT and cheering on execution risks from the pivot (not a fan) Into -> Cheering on $6B of new share dilution + getting sold on the open market against their positions. (AMC bagholder territory) If you’re trying hard to justify why new ~50% dilution is a good thing sold against your positions and trusting in management: Sorry to tell you, that you’re now in the $AMC equivalent club for datacenters. There are much better longs out there.
-
看好$AAOI美国制造优势，认为其将超越$LITE并获超大规模云厂商青睐。
我们已经起飞……尽管更广泛的光子学板块下跌，$AAOI 仍上涨 11.26%。 每个人都在寻找下一个从 49 美元涨到 614 美元的 $LITE。 有时候，它可能就坐在你面前？ 高确信度做多：$AAOI。 我真心认为这很容易在明年实现 3 倍增长。 英伟达资助了 $COHR，后者在马来西亚进行 800G/1.6T 的制造。 $LITE 在泰国使用 $FN 进行量产，并在泰国拥有自己的制造设施。 我会不断强调这一点，但 Applied Optoelectronics 是唯一纯正的“美国制造”光收发器标的。 再次重申，两家“美国”光公司将其外包给亚洲，而 $AAOI 多年来一直在德克萨斯州建设产能和晶圆厂。 英伟达刚刚资助了 $COHR 和 $LITE 以建立美国版本，以隔离其最关键供应链的地缘政治风险。 但猜猜谁已经建立了供应链设置，在这方面领先数年？$AAOI。 $LITE（市值 550 亿美元）2026 财年预估约 29.1 亿美元。 $AAOI（市值 71 亿美元）2027 年下半年：43.5 亿美元 ARR（年度经常性收入）。 如果管理层执行到位（且拥有约 40% 的毛利率），$AAOI 实际上将超越 $LITE 2026 财年的预测。 再说一次。$AAOI（70 亿美元市值）如果实现其预测，将超越 $LITE（550 亿美元市值）整个 2026 年的收入预测。 亚洲的 $FN，2026 年预测实际上与 AAOI 完全相同。 12.4% 的毛利率带来约 43.9 亿美元收入。而它的市值是 200 亿美元（且毛利率低得多）。 即使 $AAOI 只达到目标的 70%，其估值也很可能大幅重估，远超当前市值。 简而言之：在这些水平上很难看到 $AAOI 的下行风险，尤其是随着 3-4 家超大规模云厂商（可能是 $GOOGL、$MSFT、$AMZN）希望购买其未来几年的所有产能。而且 $GOOGL 不采用共封装光学(CPO) 路线。 $AAOI 在增长方面超越了 $CRDO、$ALAB、$LITE 等，并受益于光子学主题相对于铜缆（前两者）的优势。 只要管理层交付成果，$AAOI 仍然是 1 年期高确信度的非对称性机会。
英文原文
We've reached liftoff... $AAOI +11.26% despite the broader photonics names down. Everyone is trying to find the next $LITE that went from $49 to $614. Sometimes, it might just be sitting right in front of you? https://t.co/deBAD017ll
-
AI结构性改变需求，超大规模云服务商瓶颈下不会像亚马逊当年那样取消订单。
人们常将此次情况与多年前 $AMZN 因供过于需而取消订单引发的创伤后应激障碍(PTSD)相提并论。 人工智能在结构上改变了对许多前“大宗商品”的需求，包括GPU、内存，以及现在的光模块/激光器。 鉴于超大规模云服务商目前面临的瓶颈，我认为这次不会出现取消订单的情况。
英文原文
People often make comparisons to the $AMZN PTSD pulling orders many years ago when there was an oversupply. AI structurally changed demand for a lot of former“commodities” from the GPUs, memory, and now optical transceivers/lasers. Don’t think there will be an order pull this time around given how bottlenecked hyperscalers are
-
分析$AAOI空头逻辑与执行风险，看好其超大规模云需求及潜在10倍回报。
我看到的关于 $AAOI 的空头基本观点是：2.5亿美元的定向增发(ATM)不足以覆盖产能爬坡，他们还需要5000万至2.78亿美元（考虑到规模，这只是小钱）。 主要论点始终围绕执行力和潜在逆风： - 他们能否将台湾组装业务回流美国 - 扩大激光产能，以便不再零星使用 $COHR - 制造扩张的执行情况 - 中国厂商是否会导致毛利率压缩 - 共封装光学(CPO)的到来（$GOOGL + $AVGO 已确认其退出） 我不认为这是客户集中度问题，因为从 $AMZN、$MSFT 到 $META 的每个超大规模云服务商都在抢购他们能生产的所有供应。 但我认为风险已经计入市值(MC)，超大规模云服务商的需求加上 $NVDA 资助生态系统可能是最大的顺风+信号。 如果他们能执行到位，这多少让我想起 $RKLB 的“中子星”火箭。如果完美执行，我看到10倍的空间。如果不能，那就是我为上行空间承担的风险。
英文原文
The base case I’ve seen from $AAOI short sellers was that the $250m ATM doesn’t cover the capacity ramp and they needed $50m-$278m more (which is pennies given scale) The main argument is always around execution and potential headwinds: - whether they can reshore Taiwan assembly to US - scale up laser capacity so they don’t need to use $COHR here and there - executing on manufacturing expansion. - whether there’s margin compression from Chinese players - cpo arrival ( $GOOGL + $AVGO confirmed its way out) I don’t see it being a customer concentration problem since every hyperscaler from $AMZN, $MSFT, $META are trying to buy up any supply they can make. But I see the risk being priced into MC already and hyperscaler demand + $NVDA funding the ecosystem is probably the biggest tailwind + signal. It sorta does gives me $RKLB neutron vibes if they can execute. And if they can flawlessly I see it being a 10x. And if they don’t, that’s the risk I’m taking for the upside.
-
AAOI获三大云厂商包销产能，营收有望超越LITE，引发市场炒作。
如果 $AAOI 披露有三家超大规模云服务商客户（可能是 $META 或 $ORCL）加上 $AMZN 和 $MSFT 正在买断其所有产能。然后他们从 $COHR 激光器到设计再到组装实现端到端全流程。接着他们预测营收将超越另一家市值 550 亿美元的美国公司 $LITE。嗯，难怪会有这么多炒作。
英文原文
If $AAOI reveals there's three hyperscaler clients (maybe $META or $ORCL) + $AMZN + $MSFT) buying out any capacity they make. Then they do end-to-end from $COHR lasers to design to assembly... And then they're projecting revenue to leapfrog the other $55B US company in $LITE. Yeah, makes sense there's a lot of hype.
-
$AAOI 具备全产业链美国制造优势，有望超越同行营收，高确信度看涨。
$AAOI 令人极度兴奋。 如果他们能执行到位，明年市值有望从 70 亿美元重估至 350 亿美元甚至更高。 我将简要概述行业格局及原因： 组装环节： -> 从 $LITE / $COHR 采购激光器 -> 根据蓝图组装 -> 然后销售光收发器(transceiver)。 $FN (亚洲) -> 市值约 200 亿美元。 预计营收约 40 亿美元，毛利率 12.4%。 设计 + 组装： - 从 $LITE / $COHR 采购激光器，设计 800G 和 1.6T 产品 -> 然后销售光收发器。 中际旭创 (中国)：市值约 840 亿美元 毛利率约 46.2%，预计营收约 110 亿美元 新易盛 (中国)：市值约 500 亿美元： 预计营收约 53 亿美元 激光器： - 制造激光器出售给中际旭创 + 新易盛，或制造激光器 + 设计后交给 $FN 组装。 $LITE，市值 550 亿美元： 2026 财年预估营收约 29.1 亿美元（~40% 毛利率） （他们也做激光器以外的业务，例如 $LITE 通过 Cloudlite 进行设计 -> $FN 根据蓝图组装，但并非端到端全流程）。 （$COHR 和 $AVGO 也这样做） _ _ _ _ _ _ 整个供应链（激光芯片、设计和组装） $AAOI 市值 75 亿美元： 2027 年中点预估年化经常性收入 (ARR) 约 45 亿美元（40% 毛利率） - $AAOI 制造激光器（像 $LITE），从零开始设计（像中际旭创），然后像 $FN 一样组装： -> $AAOI 内部完成 $LITE / $COHR 的激光器制造 -> $AAOI 完成中际旭创/新易盛的光收发器设计。 -> $AAOI 完成 $FN 的组装。 这实现了全环节的利润率扩张/优化。 最棒的是，他们预计将超越 $LITE 2026 财年预估的约 29.1 亿美元营收... 通过：2027 年中点实现约 45 亿美元 ARR 当你看到 $AAOI 的 75 亿美元市值。 再看光电子供应链的每个部分，从市值 550 亿美元的 $LITE 到市值 550 亿美元的新易盛。 任何人都能看到如果执行到位，其原始、纯粹的上涨空间。 （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）: 高确信度做多：$AAOI。 我真切地认为这很容易在明年实现 3 倍增长。 英伟达资助了 $COHR，后者在马来西亚制造 800G/1.6T 产品。 $LITE 使用泰国的 $FN 进行量产，并在泰国拥有自己的制造设施。 我会不断强调这一点，但 Applied Optoelectronics ($AAOI) 是唯一纯正的美国制造光收发器标的。 再次强调，两家“美国”光电子公司将业务外包到亚洲，而 $AAOI 花了多年时间在德克萨斯州建立产能和晶圆厂。 英伟达刚刚资助了 $COHR 和 $LITE 以建立美国版本，以隔离其最关键供应链的地缘政治风险。 但猜猜谁已经建立了供应链布局，在这方面领先数年？$AAOI。 $LITE（550 亿美元市值）2026 财年预估约 29.1 亿美元 $AAOI（71 亿美元市值）2027 年下半年：43.5 亿美元 ARR。 如果管理层执行到位，$AAOI 实际上将超越 Lite 2026 年的营收预测（且毛利率约 40%）。 再说一次。$AAOI（70 亿美元市值）如果达成预测，将超越 $LITE（550 亿美元市值）整个 2026 年的营收预测。 亚洲的 $FN，2026 年预测实际上与 AAOI 完全相同。 ~43.9 亿美元营收，毛利率 12.4%。市值 200 亿美元（且毛利率低得多） 即使 $AAOI 只达成目标的 70%，其估值也很可能大幅重估，远超当前市值。 简而言之：在这些价位很难看到 $AAOI 的下行风险，尤其是 3-4 家超大规模云厂商（可能是 $GOOGL, $MSFT, $AMZN）希望购买其未来几年的所有产能。且 $GOOGL 不采用共封装光学(CPO) 路线。 $AAOI 在增长方面超越 $CRDO, $ALAB, $LITE 等，并受益于光电子主题而非铜缆（前两者）。 只要管理层交付成果，$AAOI 仍然是 1 年期高确信度的非对称性机会。
英文原文
$AAOI is extraordinarily exciting. There is a chance this re-rates to $35B+ or higher next year from $7B if they can execute. I'll give a TLDR of the landscape and a simple explanation why: Assembly: -> Lasers from $LITE / $COHR -> assembles from blueprints -> then sell the transceiver. $FN (Asia) -> ~$20B MC. ~$4B projected revenue, 12.4% gross margins. Design + Assembly: - Buys lasers from $LITE / $COHR, design the 800G and 1.6T -> then sells the transceiver. Innolight (China): ~$84B MC ~ 46.2% gross margins, ~$11B projected revenue Eoptolink (China): ~$50B MC: ~$5.3B projected revenue Lasers: - Creates the Lasers to sell to Innolight + Eoptolink or creates the lasers + design to give to $FN to assemble for them/. $LITE, $55B MC: FY 2026 est. ~$2.91B (~40% margin) (they also do more than lasers, eg. $LITE with Cloudlite does design -> $FN to assemble based on blueprints too, but not the entire process end-to-end). ( $COHR and $AVGO do this too) _ _ _ _ _ _ Entire supply chain (laser chips, design, and assembly) $AAOI $7.5B MC: Midpoint 2027 est. ~4.5B ARR (40% margin) - $AAOI makes the laser (like $LITE), designs it from ground up (like Innolight), then assembles it like ( $FN ): And it's primarily made in the USA. -> $AAOI does the $LITE / $COHR lasers in-house -> $AAOI does Innolight/Eoptolink transceiver design. -> $AAOI does $FN's assembly. This is possible margin expansion/optimization across the board. Best of all they're projected to leapfrogging $LITE's FY2026 projected ~$2.91B revenue... By doing: ~4.5B ARR mid-year 2027 When you look at $AAOI's $7.5B Marketcap. And you look at each part of the photonics supply chain from $LITE at $55B to Eoptolink at $55B. Anyone can see the raw, unadulterated upside if they execute.
-
若管理层执行到位，当前估值偏低，且云厂商产能紧张，亚马逊订单有望保留。
确实，这是一个很大的假设，且整体情况取决于管理层的执行能力。仅从预测来看，如果其实现每月3.78亿美元营收且毛利率约40%，那么在70亿美元市值下看起来被低估了。话虽如此，鉴于超大规模云服务商在未来两年的产能紧张状况，我认为$AMZN这次不会取消订单。让我们看看事态如何发展。
英文原文
True, it’s a big if and whole thing relies on management execution. Going off the projections alone if they hit $378m/month revenue off ~40% gross margins. It looks undervalued at $7B. That being said don’t think $AMZN will pull orders this time around given how capacity strained all the hyperscalers are for the next 2 years. We’ll see how things go
-
AAOI获三大云厂商抢购产能，预期已部分定价，风险在于执行。
我去年在$AAOI股价约$28时便已建仓，至今涨幅已达3倍。起初我对它并没有很高的确信度，只是认为鉴于“美国制造”供应链优势及当时超大规模云服务商的资格认证，它具备潜力。但管理层近期的指引最终让我下定决心，因此我在当前价位加仓。拥有三家超大规模云服务商（可能是$GOOGL、$AMZN、$MSFT）抢购其所有产能，这是极其看涨的信号。尤其是当它们预测2027年下半年月收入将达到3.78亿美元时。市场是前瞻性的，所以我认为这一预期现在已开始被定价。唯一的变数是执行能力，但这正是投资大多数公司所必须承担的风险。
英文原文
I was actually in $AAOI since $28-ish last year, so it's 3x'd since then. But didn't really have high conviction in it, just thought it had potential given Made in America supply chains + hyperscaler qualification at the time. But management projections recently sealed the deal, so I ended up buying more around these levels. Having 3 hyperscalers, likely $GOOGL, $AMZN, $MSFT, buying up any supply you can make is incredibly bullish. Especially when they're projecting $378M a month revenue round H2 2027. Markets are forward looking so I think it started to get priced in now. Only uncertainty is execution, but that's the risk you're taking with majority of companies.
-
看好$AAOI美国制造优势，认为其营收将超越同行且下行风险小。
高确信度做多：$AAOI。 我真心认为这明年很容易翻3倍。 英伟达资助了$COHR，后者在马来西亚进行800G/1.6T的制造。 $LITE在泰国使用FN进行量产，并在泰国拥有自己的制造工厂。 我会一直强调这一点，但Applied Optoelectronics是唯一纯正的“美国制造”光收发器标的。 再次重申，两家“美国”光通信公司将业务外包到了亚洲，而$AAOI则花了多年时间在德克萨斯州建设产能和晶圆厂。 英伟达刚刚资助了$COHR和$LITE以建立美国版本，旨在将其最关键的供应链与地缘政治风险隔离。 但猜猜谁已经建立了供应链布局，在这方面领先数年？$AAOI。 $LITE（市值550亿美元）2026财年预估营收约29.1亿美元。 $AAOI（市值71亿美元）2027年下半年：43.5亿美元年化经常性收入(ARR)。 如果管理层执行到位（且拥有约40%的毛利率），$AAOI实际上将超越$LITE 2026财年的预测。 再说一次。如果$AAOI实现其预测，它将超越$LITE（市值550亿美元）整个2026年的营收预测。 亚洲的$FN，2026年预测实际上与AAOI完全相同。 营收约43.9亿美元，毛利率12.4%。且其市值为200亿美元（毛利率低得多）。 即使$AAOI只达到目标的70%，其估值重估很可能远超当前市值。 总结：在这些价位上，$AAOI很难看到下行风险，尤其是3-4家超大规模云厂商（可能是$GOOGL、$MSFT、$AMZN）希望购买其未来数年的所有产能。且$GOOGL不采用共封装光学(CPO)路线。 $AAOI在增长方面超越了$CRDO、$ALAB、$LITE等，并受益于光子学主题相对于铜缆（前两者）的优势。 只要管理层兑现承诺，$AAOI仍是一个非对称的1年高确信度标的。
英文原文
High conviction long: $AAOI. I genuinely think this could easily be a 3x by next year. Nvidia funded $COHR, who does Malaysia manufacturing for 800G/1.6T. $LITE uses FN in Thailand for volume production, and has it's own manufacturing in Thailand. I will keep hammering this home but Applied Optoelectronics is only pure Made in America, optical transceiver play. Again, the two "American" optical companies outsourced it to Asia, while $AAOI spent the years building up capacity and fabs in Texas. Nvidia funded both $COHR and $LITE just now to build out a US-version to insulate its most critical supply chain from geopolitical risks. But guess who already has the supply chain setup and is years ahead in that regard? $AAOI. $LITE ($55B) FY 2026 est. ~$2.91B $AAOI ($7.1B MC) H2 2027: $4.35B ARR. $AAOI will actually leapfrong Lite FY 2026 projections if management executes (and with ~40% gross margins). Once again. $AAOI ($7B) will leapfrog $LITE ($55B MC) entire 2026 revenue projections if they deliver their projections. $FN over in Asia, 2026 projections are actually around the exact same as AAOI. ~4.39B revenue off 12.4% gross margins. And it's a $20B MC (with much lower margins) Even if $AAOI hits 70% of their target, it's likely to be heavily re-rated way past it's current marketcap. TLDR: Hard to see downside with $AAOI at these levels, especially with 3-4 hyperscalers (likely $GOOGL, $MSFT, $AMZN) wanting to buy up any capacity it can make for years out. And with $GOOGL not going the CPO route. $AAOI leapfrogs $CRDO, $ALAB, $LITE, and others in growth + benefits from photonics theme vs. copper (from the first two). $AAOI remains an asymmetrical 1Y high conviction as long as management delivers.
-
澄清超大规模云厂商订单无取消风险，强调美国本土供应链优势及光子学瓶颈。
超大规模云服务商提前一年下满所有产能的订单，并不意味着有取消风险？我不确定你从哪得出这个结论。 什么20%的利润率上限？他们提到的是40%。 此外，它们拥有“美国制造”的供应链，这很罕见，$LITE 和 $COHR 的组装都在亚洲进行。 但如果人们因多年前 $AMZN 取消订单而患有创伤后应激障碍(PTSD)，那也能理解。但这次应该不同，因为光子学(Photonics)是已知的瓶颈，且需求超过供应。
英文原文
Three to four hyperscaler orders for all the capacity they can make a year out doesn’t mean risk of cancellation? Not sure where you’re getting that from. What 20% margin ceiling? They cited 40%. Also they’re Made in America supply chains which is rare, $LITE and $COHR do assembly in Asia. But makes sense of people have ptsd from $AMZN order cancellation many years ago. But this time should be different since photonics is a known bottleneck and demand outstrips supply
-
AI光模块产能瓶颈持续，云厂商订单不减，风险收益比极佳。
是的，很多人对管理层过去的问题有创伤后应激障碍(PTSD)。我只是觉得，如果他们能实现甚至65%的指引，风险收益比就极其不对称。此外，我认为像$AMZN这样的超大规模云服务商这次不会削减订单，因为800G/1.6T光模块确实是瓶颈，且产能缓解遥遥无期。
英文原文
Yep lot of people have PTSD from management in the past. I just felt the risk-reward was extremely asymmetrical if they deliver even 65%% of their guidance. Also I do feel like hyperscalers like $AMZN won't pull back orders this time as 800g/1.6T it's genuinely a bottleneck with no end to capacity relief in sight.
-
看好$AAOI十倍增长潜力，视其为光子学核心持仓。
$AAOI 对我来说感觉是必须配置的共封装光学(CPO)敞口。 他们的财报简直像黑魔法一样神奇。 $LITE（市值500亿美元）：2026财年预估营收约29.1亿美元 $AAOI（市值56亿美元）2027年下半年：43.5亿美元年度经常性收入(ARR)。 他们的增长预测在2026年实际上已经超越了 $LITE。 我没开玩笑。 $AAOI 的毛利率约为40%，$LITE 凭借光电路交换(OCS)垄断地位可能达到45%。 “长期毛利率目标约为40%。” 别误会，$LITE 是一只极佳的长线股，因为其在 $GOOGL TPU 和 Google 资本支出中占据8-12%的物料清单(BOM)份额... 但 $AAOI 的营收顶层增长接近去年营收的10倍，这简直不可思议。 如果你用草稿纸数学模型计算： “管理层预测，到2027年第二季度，每月光收发器营收潜力达3.78亿美元” ~ 43亿美元目标，40%毛利率（自产），17.2亿美元毛利润，也许~6亿美元资本支出/运营支出： 净利润 ~8.96亿美元。 远期市盈率（基于2027年净利润）：~6.4倍远期市盈率，对应约900%的同比增长。 而且如果你看光子学产能爬坡，预计向2030年呈指数级增长。 从他们的管理层电话会议来看，超大规模云服务商（$GOOGL, $MSFT, $AMZN）很可能买断了所有可用的光收发器产能。 这让我想起早期的 $ALAB 或 $CRDO，所有超大规模云服务商都是他们的客户。 如果你考虑到像 $SNDK 的NAND闪存那样的价格涨幅： +15% 平均售价(ASP)涨幅：14.1亿美元净利润 - （4.1倍远期市盈率） +30% 平均售价(ASP)涨幅：19.2亿美元净利润 - （3.0倍远期市盈率） 这就是为什么我在 $AAOI 财报后如此兴奋。 甚至还没包括 $AAOI “美国制造”的光收发器供应链，这将使他们获得相对于 $FN、中际旭创(Innolight)、易普联(Eptolink)等海外厂商的溢价。 有点像 $INTC 对 $TSM 或 $MU 对 SK海力士的溢价。 如果他们的预测成真，且价格涨幅使光子学板块在800%+增长下达到3倍市盈率，回报率将是天文数字。 再次强调，这些是管理层预测，执行力才是关键。 无论如何，我将 $AAOI 作为我最大的光子学持仓之一，因为这看起来像黑魔法，也是下一个 $SNDK 的开端。 如果他们能实现预测，非零概率是我们在一年半后看到其达到 $LITE 500亿美元的估值。
英文原文
$AAOI feels like must-have photonics exposure to me. Their earnings report is genuinely black magic. $LITE ($50B MC): FY 2026 est. ~$2.91B $AAOI ($5.6B MC) H2 2027: $4.35B ARR. Their growth projections literally leapfrogged $LITE for 2026. Not even kidding. Gross margins are ~40% for $AAOI, maybe 45% for $LITE with OCS monopoly. "-term objective of gross margins to around 40%," Don't get me wrong, $LITE is an amazing long due to 8-12% BOM of $GOOGL TPU and Google capex spend... But $AAOI top line growth for revenue is close to 10X revenue growth from last year's revenue, which is just insane. And if you go off napkin math modeling: "By Q2 2027, management forecasts monthly transceiver revenue potential of $378 million" ~ $4.30B target, 40% margin (in-house), $1.72B gross profit, maybe ~$600m capex/opex: Net income ~$896M. Forward P/E (on 2027 NI): ~6.4x forward p/e off ~900%-ish Y/Y growth. And if you look at photonics ramp, it's expected to increase exponentially toward 2030. From their management call hyperscalers likely ( $GOOGL, $MSFT, $AMZN ) are buying out all available capacity for transceivers. Reminds me of a early $ALAB or $CRDO where all the hyperscalers are their customers. And if you get to likely price hikes like NAND for $SNDK: +15% ASP Price Hike: $1.41B net income - (4.1x forward p/e) +30% ASP Price Hike: $1.92B net income - (3.0x forward p/e) This is why I got so excited about $AAOI post-earnings. Not even including $AAOI "Made in America" transceiver supply chains that would give them a premium where $FN, Innolight, Eptolink, and others do theirs abroad. Kinda like $INTC to $TSM or $MU to SK Hynix premiums. The ROI for this is just astronomical if they hit their projections and price hikes give them a 3 p/e rate off 800%+ growth in photonics segments. Again, these are management projections, execution is the biggest the name of the game. Regardless, I'm making $AAOI one of my largest photonics positions because this looks like black magic and the beginning of the next $SNDK. There's a nonzero chance we might see this at a $50B $LITE valuation one and a half year from now if they can execute on their projections.
-
AAOI财报揭示上游设备与材料瓶颈，IQE等上游厂商受益显著。
来自 $AAOI 财报的二阶效应： -> 对 MBE/MOCVD 反应堆的需求极度旺盛。 它们用于制造外延晶圆(epiwafers)。 这为“制造”环节释放了两个信号： - $IQE 坐拥一座绝对的“金矿” - 资本开支周期将惠及类似 $VECO ($18.5亿) 和 Aixtron ($AIXXF $37亿) 这样的设备商。 引用：3-4 家超大规模云服务商，可能是 $GOOGL、$AMZN、$MSFT，对光收发器(transceivers)的需求已触顶。 -> 中际旭创(Innolight)、易普拓(Eptolink)、$COHR 等厂商可能也会经历一轮资本开支周期以满足需求。 这主要利好两家外延晶圆代工代工厂： -> $IQE ($1.75亿) ~100+ 台反应堆 -> 晶品半导体(Landmark) (~$36.8亿) ~27-30 台反应堆。 所以这大概就是为什么 $LITE 大量使用 $IQE，而 Landmark 营收屡创新高。$AAOI 则一直在建设其德州工厂的产能。 总结： 1. Aixtron 和 $VECO 受益于今年类似迷你版 $ASML 的资本开支周期。当 Aixtron 在2月指引中表示预计2026年“整体市场环境持续疲软”时，也许他们在低调隐藏实力(sandbagging)或订单尚未下达。话虽如此，我认为最佳敞口可能并非设备卖家。 2. $IQE 坐拥满足 $LITE 及其他超大规模云服务商对 InP 外延晶圆需求的产能金矿。他们只需成功重组，估值就能更像其台湾同行($36.5亿市值)。 3. $AAOI 的财报显示收发器制造商及其组装所需的组件/材料需求惊人，市场肯定尚未定价。 4. $AXTI 一如既往地成为所有环节的瓶颈，正如我去年所提到的。 名字越响亮(如 $LITE、$COHR)，交易可能越拥挤/已充分定价。 但当你向上游追溯时，还有大量未被市场定价的隐藏宝石，它们将从巨大的光子学扩张中受益。 这是投资者的新淘金热，但本质是瓶颈狩猎。
英文原文
Second order effects from $AAOI earnings: -> Extreme demand for MBE/MOCVD reactors. They're needed for epiwafers. This signals two things for "manufacturing": - $IQE is sitting on an absolute gold mine - capex cycle to ( $VECO $1.85B, Aixtron $AIXXF $3.7B) type players. Quote: 3-4 hyperscalers, likely $GOOGL, $AMZN, $MSFT. demand maxed out for transceivers. -> Innolight, Eptolink, $COHR, and others will probably go through a capex cycle to meet demand too. This largely benefits both outsourced foundries for epiwafers: -> $IQE ($175M) ~100+ reactors -> Landmark (~$3.68B) ~27-30 reactors. So that's probably the reason why $LITE has been heavily using $IQE, and Landmark has been hitting record revenues. $AAOI's has been building capacity from their texas fab. TLDR: 1. Aixtron, $VECO benefits from mini $ASML style capex cycle this year. When Aixtron said 2026 "continued softness in overall market environment expected" from feb guidance, maybe they're sandbagging or orders havent come in yet. That being said, best exposure imo is probably not the machine sellers. 2. $IQE is sitting on a gold mine of capacity to meet demand across $LITE and others hyperscalers for inp epiwafers. They just need to restructure succesfully, to be rerated more like their $3.65B friend in Taiwan. 3. $AAOI earnings is just insane demand for both transceiver makers and the components/materials required to assemble them, definitely not priced in yet. 4. $AXTI as always happens to be the bottleneck of everything as I mentioned last year. The more well known the name is ( $LITE, $COHR ) the likely more crowded/priced-in. But when you go more upstream, there's a ton of hidden gems markets haven't priced in yet that benefit from the massive photonics ramp. It's the new gold rush for investors, but bottleneck hunting.
-
$AAOI获三大超大规模客户订单，营收激增，基本面强劲。
$AAOI 看起来像是早期的 $SNDK。而且在未来一年，其市值很可能远超 55 亿美元。 在他们的财报电话会议上： - 3 家超大规模客户（可能是 $GOOGL、$AMZN、$MSFT） - “他们希望购买我们能生产的所有 800G 和 1.6T 光收发器”——完全售罄，就像内存一样。 - 营收增长预测超过 900%，达到 43 亿美元，而当前市值仅为 55 亿美元。 光子学领域的需求史无前例，$AAOI 刚刚证实了这一点。 而且他们随时可以开始提价，将毛利率从 40% 进一步提高。当然，存在执行风险…… 但这是一个你应该关注基本面而非图表的例子。
英文原文
$AAOI looks like an early $SNDK. And is likely going much higher than a $5.5B MC over the next year. On their earnings call: - 3 hyperscale customers (probably $GOOGL, $AMZN, $MSFT) - "they would like to buy all the transceiver we can make for 800G and 1.6". - completely sold out, like memory. - 900%+ revenue growth forecast to $4.3B off $5.5B MC. There is unprecedented demand for photonics, and $AAOI just confirmed it. And they can always begin price hikes to increase gross margins from 40%. Of course there are execution risks... But this is one example where you look at fundamentals over chart.
-
深度解析$IQE：通过出售台湾业务去债，转型InP光子学产能，类比矿企转AI。
深入解析 $IQE（市值1.79亿美元）： IQE隐藏的磷化铟(InP)期权价值 vs LandMark 35亿美元的估值。以及 $IREN / $CRWV 等“矿企”向光子学(Potonics)转型的逻辑： 此前我做过一个高层级的“淋浴思考”概览，例如 $AXTI -> $IQE -> $LITE -> $GOOGL TPUs，但这次是更深入的尽职调查(DD)。 基本上：IQE是全球最大的独立商用化合物半导体外延晶圆代工厂，按反应炉数量和物理产能计算。 然而，由于受困于低毛利的传统无线业务和短期流动性约束，其估值处于困境状态。 LandMark Optoelectronics (TPEX: 3081) 是最接近的对比标的。作为800G和1.6T光互连市场中AI InP需求的纯代理标的，LandMark市值约38亿美元，相比市值1.75亿美元的 $IQE 享有巨大溢价。 但如果深入看底层硬件，这种脱节非常有趣： LandMark的运营规模在物理上受限。据估计，他们在台湾单一校区仅运营约27-30台金属有机化学气相沉积(MOCVD)反应炉。 相比之下，IQE在全球拥有超过100台MOCVD和分子束外延(MBE)系统。 IQE光子学资产基础的潜在重置价值和结构性产能似乎远超其当前公开市场估值。 这有点像比特币矿企拥有3GW产能 vs 750MW产能，如果进行转换，前者有巨大的变现期权价值。 我们也看到了光模块瓶颈： -> 下游对光收发器(Optical Transceivers)的需求正在以前所未有的速度加速。 -> 来自 $GOOGL、$MSFT、$AMZN 等的极端需求直接向上游硬件供应链传导。 这给Innolight等光模块集成商、$COHR、$LITE 和 $AVGO 等光组件制造商，以及最终为 foundational 激光芯片生长原始外延晶圆的商用外延代工厂带来了巨大压力。 由于其他玩家触及物理产能天花板，像 $COHR 这样的垂直整合玩家已满载，超大规模云厂商和模块制造商迫切需要在 $IQE 这样的玩家中寻找替代产能。 而且... 隐藏在IQE合并报表之下的是大量的 Aixtron AIX 2800G4-TM 反应炉。 这些反应炉原生具备双能力(GaAs/InP)，可以以相对较低的成本（每台50万-150万美元）重新用于InP生产，但需要数月或一年时间来重构。当然，执行层面增加了资格认证和良率风险（类似于比特币矿企对GPU进行软件编排，如 $CRWV）。 但无论如何，IQE拥有类似 $IREN 或转型HPC的比特币矿企那样的产能。而LandMark证明了纯代理标的带来的估值溢价。 核心问题.. 释放被困价值： 虽然IQE的营收远高于LandMark，但由于4500万英镑的总债务，其股价（1.75亿美元市值）被定价为破产边缘。 但对超大规模云厂商来说，这笔债务微不足道： 其正在进行的由Lazard指导的战略审查的明确目标是：最终完成IQE台湾（传统GaAs无线业务）的出售，并利用所得资金彻底且永久地消除母公司的限制性债务结构。 再次强调，其可转换贷款票据面值2120万英镑，公司实收1800万英镑。净债务为2350万英镑。 -> 需要立即清除的债务负担：2350万英镑汇丰设施 + 2120万英镑CLN = 约4500万英镑。 假设IQE台湾单元以1亿至1.5亿英镑的高价出售（不保证），IQE在完全无债后将获得5000万至1亿英镑的盈余现金。 然而，射频GaAs目前并不“热门”，因此在困境资产出售中可能仅值5000万至6000万英镑，这足以清偿债务，但缓冲空间很小。 地缘政治转向： 一旦无债，IQE可以将其在北卡罗来纳州和威尔士等地大量闲置的制造产能转向数据中心InP外延晶圆市场。 这为光子学中最关键的瓶颈创造了一条完全资本化、纯西方的供应链，在美国和英国高度重视国内半导体基础设施之际，减少了对亚洲的依赖。 基本上，鉴于 $IQE 拥有的原始资产数量： -> 成功出售台湾业务将消除持续经营风险，清除所有债务，并让他们直接向Tier 1光模块玩家变现其6英寸InP外延晶圆技术。 这是一笔基于成功重组以释放被困价值的深度资产价值交易。且它已是超大规模云厂商光网络的知名供应商（并非科学实验项目）。 下行风险是过度稀释和重组失败。但鉴于其对西方供应链和超大规模云厂商供应链的地缘政治重要性，似乎有更多缓冲。 我个人决定将其作为巨大的潜在反转机会做多。但再次强调，这不适合所有人，风险极高。 TLDR： -> IQE被定价为困境中的射频供应商。 -> 它拥有真正具备光子学能力的硬件基础设施。 -> 如果4100万英镑总债务被移除，且管理层将资本支出重新分配给InP，股权可能大幅重估。 -> 重组+产能期权交易，风险极高但上行空间也极大。 最接近的类比是 $IREN 或 $CIFR 等将GW产能转向AI HPC的比特币矿企。他们拥有大量物理硬件（GW产能），需要资金进行转型（通过出售台湾业务或稀释）。 我乐观地认为他们能做到（像 $INTC 那样，鉴于其对西方的地缘政治重要性，有回旋余地）。 下行风险是极端稀释，这始终是一种可能性（意味着你的股权可能被清零以清偿债务或帮助重构）。 我只是发现4500万英镑总债务（占流通股本+债务的14.4%）并非不可承受，且管理层正试图通过资产出售而非股东稀释来清除债务。 只想发布更深入的拆解和这个非常二元化的**高风险**但潜在高回报交易更多风险。
英文原文
Here's a deeper look into $IQE ($179M): IQE's hidden InP optionality versus LandMark's $3.5B valuation. And the $IREN / $CRWV "miner" pivot to photonics: Before I did a high-level shower thought overview eg. $AXTI -> $IQE -> $LITE -> $GOOGL TPUs, but this is slightly more DD. Basically: IQE is the largest independent merchant compound semi epitaxial foundry in the world by reactor count and physical capacity. However, it's trading at distressed valuations because it's burdened by a low-margin legacy wireless business, and near-term liquidity constraints. LandMark Optoelectronics (TPEX: 3081) is the closest comparison. As a pure-play proxy for AI InP demand in the 800G and 1.6T optical interconnect market, LandMark commands a ~$3.8B billion market cap with large premiums in comparison to $IQE which is trading at a $175M MC. But if you look deeper at the physical hardware, the disconnect is pretty fascinating: LandMark's operational scale is physically limited. They only operate around 27 to 30 Metal-Organic Chemical Vapor Deposition (MOCVD) reactors out of a single campus in Taiwan per some estimates. IQE, by stark contrast, possesses well over 100+ MOCVD and MBE systems globally. The underlying replacement value and structural capacity of IQE’s photonics asset base looks to vastly exceeds its current public market valuation. Kind of like if a Bitcoin miner has 3GW capacity, vs 750 MW, there's large optionality to monetize it if they convert it. And we're seeing an transceiver bottleneck too: -> The downstream demand for optical transceivers is experiencing unprecedented acceleration. -> Extreme demand, from $GOOGL, $MSFT, $AMZN and others flow directly up the hardware supply chain. This puts immense pressure on transceiver integrators like Innolight, optical component manufacturers like $COHR, $LITE, and $AVGO, and ultimately, the merchant epitaxial foundries that grow the raw epiwafers required for the foundational laser chips. And since other players are hitting a physical capacity ceiling, vertically integrated players like $COHR are capped out, hyperscalers and module makers are desperate for alternative capacity in players like $IQE And.. Hidden entirely beneath IQE's consolidated corporate lines is a massive fleet of Aixtron AIX 2800G4-TM reactors. These are natively dual-capable (GaAs/InP) and can be repurposed for InP production at a relatively modest cost ($500K-$1.5M per reactor) but take few months or year to refactor. And obviously qualification and yield risk added to execution (similar to Bitcoin miners doing software orchestration to GPUs like $CRWV). But still, IQE has the capacity kinda like $IREN or Bitcoin miners that pivoted to HPC. And LandMark is proof of the valuation pure play exposure brings. The Major Question.. Unlocking Trapped Value: While IQE generates significantly higher top-line revenue than LandMark, it's priced ($175M MC) for bankruptcy because of its gross debt of 45M. But the debt looks like pennies to hyperscalers: The explicit, stated goal of their ongoing Lazard-advised strategic review is to definitively conclude the sale of IQE Taiwan (their legacy GaAs wireless business) and utilize the proceeds to completely and permanently extinguish the parent company's restrictive debt profile. Once again their convertible loan notes is a norminal face value of £21.2 million, for proceeds of £18 million for the company. Then they're net debt, £23.5 million. -> The immediate debt burden requiring clearance: £23.5M HSBC facility + £21.2M CLN = ~£45M. Assuming a highly sale price for the IQE Taiwan unit of between £100 million and £150 millio (not guaranteed), IQE would net £50 million to £100 million in surplus cash after becoming completely debt-free. However, RF GaAs is not currently "hot", so in a distressed asset sale it might only be £50M to £60M, which gives it enough room to clear debt alone and little cushion room. The Geopolitical Pivot: Once debt-free, IQE can shift its massive, currently underutilized manufacturing capacity in places like North Carolina and Wales toward the InP epiwafer market for datacenters. It creates a fully capitalized, purely Western-based supply chain for the most critical bottleneck in photonics, eliminating more dependency on Asia at a time when the US and UK are heavily prioritizing domestic semiconductor infrastructure. Basically, just given the amount of raw assets $IQE has: -> Successfully selling off their Taiwan business wipes out the going-concern risks, clears all debt, and leaves them to monetize their 6-inch inp epiwafer tech directly for the Tier 1 optical transceiver players. It's a deep asset value trade on a successful restructuring to unlock trapped value. And a currently well-known supplier for optical networking for hyperscalers (so not a science project). Downside risks are excessive dilution and failure to restructure. But given it's geopolitical importance to Western supply chains and hyperscaler supply chains, it seems to have more cushion. I personally decided to enter this long as a massive potential turnaround. But again, it's not for everyone and it's extremely high risk. TLDR: -> IQE is priced like a distressed RF supplier. -> It owns real photonics-capable infrastructure. -> If gross 41M debt is removed and management reallocates capex toward InP, the equity could rerate materially. -> Restructuring + capacity optionality trade with extreme risk but extreme upside. Closest comparison is Bitcoin miners like $IREN or $CIFR that pivot their GW capacity to AI HPC. They have a ton of physical hardware (GW capacity), and need funds to pivot (either through sale of Taiwan business or dilution). It's an optimistic trade I took they can do it (with wiggle room like $INTC given their geopolitical importance to the West). The downside is extreme dilution, which is always a possibility (meaning your equity gets wiped out to 0 to clear their debts or to help them refactor). I just found that 45M gross debt (14.4% of float + debt) wasn't the most and management was looking to clear that through asset sales rather than dilution to shareholders. Just wanted to publish deeper breakdown and more risks of this very binary **high risk**, but potentially high upside trade.
-
英伟达财报虽超预期但股价跌，长期看AI加速，应关注云厂商资本开支。
$NVDA 远超预期，预计营收780亿美元，高于预期的720亿美元。 该股在财报发布后下跌4.82%。 这是一个有争议的观点，但在我看来，英伟达的财报是一个部分滞后指标。 最需要关注的是超大规模云服务商(Hyperscaler)的资本支出(Capex)预测以及 $TSM 的预测。 市场倾向于认为 $SNDK 等存储芯片、来自 $TSM 的半导体、电力/电网等其他板块，将主要依据上述两项指标涨跌，因为它们代表了整个人工智能(AI)行业。特别是随着从 $GOOGL 到 $AMZN 的超大规模云服务商纷纷加速构建其专用集成电路(ASIC)（除英伟达外）。 英伟达大幅超预期仅仅是已实现的确认，表明人工智能交易可能会加速。 短期流动性和做市商(MM)头寸影响周度价格，但长期来看，情况似乎将进一步加速。
英文原文
$NVDA blows away expectations, with $78B in revenue projected vs. $72B. The stock dropped 4.82% on the earnings. Controversial opinion but Nvidia earnings is a partially lagging indicator to me. The #1 thing to look out for is hyperscaler capex projections and $TSM projections. Markets like $SNDK in memory, semis from $TSM, power/grid and others are likely to rise/fall based on those two in specific as they represent the entire AI sector as a whole. Especially as hyperscalers from $GOOGL to $AMZN rush out to build their on ASICs (apart from Nvidia) Nvidia beating by a wide margin is just realized confirmation that the AI trade is likely to accelerate. Short term liquidity and MM positioning affect weekly prices, but long term, things look to accelerate even further.
-
看好AXTI光子潜力及美股科技巨头，建议做多黄金。
$MU - 最终更看好 SK Hynix，但这是更安全的“美国制造”标的。 $AXTI 还有很长的路要走，需等待1年直到极端短缺出现。可能是光子学(Photonics)板块中回报潜力最大的最爱。 $EWY - Vega扩张的尾声，隐含波动率(IV)可能会将其推高一点。但如果你买入看涨期权(Call)，应是因为 KOSPI 上涨而非博弈 IV。 做多黄金是好的，因为外国正在大量买入并抛售美国国债。不喜欢在当前价位做空比特币(Bitcoin)。 $AMZN、$MSFT、$META 在当前估值下具有吸引力。
英文原文
$MU - ended up liking SK Hynix more but safer made in America play $AXTI long way to go, 1 year wait until extreme shortage hits. Prob my favorite in photonics sector for return potential $EWY - tail end of Vega expansion, IV will likely send it a tad higher. But if you’re buying calls do it because KOSPI goes up not for IV plays. Gold longs are good because foreign countries are buying it up and selling Us treasuries. Don’t like Bitcoin puts at these levels. $AMZN, $MSFT, $META are attractive at these valuations.
-
电力成美国AI最大瓶颈，看好$XLU因降息及基建资本支出重估。
美国AI最大的瓶颈： 电力与电网容量。 最近，美国各大AI公司的CEO都反复强调这一点。 以下是从$MSFT到$AMZN的名单： 黄仁勋（$NVDA）：“在最底层：能源。中国的能源总量是我们的两倍，而我们的经济体量比他们大。这对我来说毫无道理……没有能源，你就无法发展任何新产业。”（CSIS活动） 埃隆·马斯克（$TSLA / xAI）：“价值数十亿美元的最先进AI硬件。处于闲置状态。不是因为芯片不工作。而是因为电力不足，无法运行它们。”（最近播客） 山姆·阿尔特曼（OpenAI）：“满足全球AI需求最终需要数百吉瓦的电力。谁控制了电力，谁就控制了AI价值链。”（最近采访） 安迪·贾西（$AMZN）：“[我们]最大的单一制约因素是电力。我不相信我们在几个季度内就能完全解决我们所需的需求容量。”（亚马逊财报电话会议） 萨提亚·纳德拉（$MSFT）：“我们现在面临的最大问题不是算力过剩，而是电力。也就是能否在靠近电力的地方快速完成建设。如果你做不到，你可能会有一堆无法插电使用的芯片库存。”（BG2 Pod） 每一位美国AI实验室的CEO都在告诉你美国人工智能的主要瓶颈是什么： 发电和扩大电网。 因此，我们可能会看到$XLU中无聊的公用事业和电力生产商因为以下原因被强烈重估： 1. AI推理+训练极端扩张以及对GW级数据中心的需求。 2. 超大规模云厂商向该领域投入极端资本支出。 3. 降息顺风。 最重要的是： 作为一个国家，美国必须重建其电网以赢得地缘政治军备竞赛。 （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）: 如果我要在1年内将$10万变成$100万。 我会选择：$XLU 2年期虚值看涨期权 2026年是现代历史上市场首次同时具备： - 利率下降 - AI推理+建设 通过映射，XLU有潜在~40%的涨幅（虚值期权1000%+）。 这是我的宏观论点： 1. 降息 当美联储在不引发衰退的情况下降息时，公用事业债务成本降低，机构将低收益现金转向公用事业股息。 这导致估值倍数立即扩张： 1995年：标普公用事业板块在1995年回报+31.3%，1996年再+12.1% - 累计回报~47% 2019年中周期降息：结果：XLU在那一年产生了+25.9%的总回报 标准的软着陆降息周期自然映射到25%到30%的基准回报。而我们要进入2026年的新降息周期。 2. 基础设施超级周期资本支出 基础设施资本支出为该板块带来复合盈利增长。继2000年代初之后，公用事业进入了一个巨大的资本支出周期以现代化老化的电网基础设施。 因为他们不断支出并扩大其保证费率基数，XLU在2004年回报+23.5%，2005年+16.3%，2006年+20.8%，2007年+18.4%。 然而这一次： 2026年8000多亿美元的AI建设使得2004年的电网现代化看起来像零钱。 所以你有来自#1降息的估值倍数扩张（+15%到+20%）。来自#2资本支出历史数据的每股收益增长（+18%到+20%）。仅从历史教训来看。 但2026年是历史上AI使用最独特的时刻。 仅从我自己的模型预测来看，所有以前的估计可能都因AI极端扩张而错误（例如DOE/LBNL预测）： 超大规模云厂商资本支出流入（支出）（亚马逊、微软、Meta、谷歌、甲骨文）到数据中心估计： 2024年：2200亿美元 2025年：3500亿美元 2026年：5500亿美元 2027年：8000亿美元 2028年：1.2万亿美元（4年增长：+445%） 美国数据中心电力使用： 2024年：190 TWh 2025年：280 TWh 2026年：430 TWh 2027年：650 TWh 2028年：980 TWh（4年增长：+415%） AI消耗的总美国电力百分比： 2024年：美国电网的4.5% 2025年：6.6% 2026年：8.2-10.2% 2027年：13.4-15.4% 2028年：21.3-23.3% 劳伦斯伯克利国家实验室和能源部似乎在AI使用上偏离了（他们预测到2028年约为12%） 物理电网容量需求： 2024年：18 GW 2025年：35 GW 2026年：65 GW 2027年：105 GW 2028年：160 GW 基本上你可以看到2026年到2028年是拐点，而2024-2025年是爬坡的缓慢年份。 然后是独立公司的“绝望溢价”。因为电网容量已售罄，科技巨头向公用事业支付巨额溢价以插队。例如PJM互联（弗吉尼亚“数据中心巷”），容量价格从2024年的每兆瓦日28.92美元飙升至2026/2027年令人难以置信的329.17美元。 $VST或Constellation是ETF中独立电力生产商的大权重。 纵观全局，你可以看到从2026年（现在）到2028年的极端扩张，以及用于建设基础设施的极端资本支出，与往年相比。 2026年是现代市场历史上第一次所有因素同时为无聊的电网/电力板块发力，AI是最大的顺风。 正如埃隆所引用的：“价值数十亿美元的最先进硬件。处于闲置状态。不是因为芯片不工作。而是因为电力不足，无法运行它们”。 再次强调，2026年是一个绝对的历史异常值，由于AI和做市商已将历史隐含波动率（极度平坦~14%-16%）计入虚值看涨期权。 我们看到AI推理（超出之前的测量）以及训练（根据OpenAI今天的报告）的爆发。 所以地球上最无聊的板块（电力/电网），可能会因为超大规模云厂商/政府支出用于电网改进 -> AI推理/训练的极端电力消耗 -> 降息和其他因素，成为主要反弹的起点。 这只是我的个人论点，期权带有风险并放大下行风险。这些也是我自己的预测，不确定是否会超过或低于它们。 但基本上： 2026年是一个绝对的历史异常值。 美国的新瓶颈是电力。 来自AI的极端需求，极端资本支出，降息： $XLU看起来是最好的暴露交易。 时间会证明这是否正确与否。
英文原文
The biggest United States AI bottleneck: Power and Grid Capacity. This has been echoed by the CEOs of every major US AI company recently. Here's the list from $MSFT to $AMZN: Jensen Huang ( $NVDA ): "At the lowest level: energy. China has twice the amount of energy we have as a nation, and our economy is larger than theirs. It makes no sense to me... There are no new industries you can grow without energy." (CSIS event) Elon Musk ( $TSLA / xAI): "Billions of dollars of the most advanced AI hardware ever built. Sitting dark. Not because the chips don’t work. Because there isn’t enough electricity to run them." (Recent Podcast) Sam Altman (OpenAI): "Meeting global AI demand will eventually require hundreds of gigawatts of power. Whoever controls power controls the AI value chain." (Recent Interview) Andy Jassy ( $AMZN ): "[Our] single biggest constraint is power. I don't believe that we will have fully resolved the amount of capacity we need for the demand that we have in a couple of quarters." (Amazon Earnings Call) Satya Nadella ( $MSFT ): "The biggest issue we are now having is not a compute glut, but it's power. It's sort of the ability to get the builds done fast enough close to power. So, if you can't do that, you may actually have a bunch of chips sitting in inventory that I can't plug in." (BG2 Pod) You have every single CEO of American AI labs telling you what the major bottlenecks are for US Artificial Intelligence: Generating the Energy, and expanding the Grid. So, we might see boring utilities and power producers in $XLU be strongly re-rated because of: 1. AI inference + training extreme ramp and demand for GW-scale DCs. 2. Hyperscaler extreme capex pouring into the sector. 3. Rate Cut tailwinds. And most of all. That fact that: The United States as a country has to rebuild its grid to win a geopolitical arms race.
-
超大规模云厂商负债投入AI基建，资金流向英伟达等上游供应商。
@LuffyDDK 他们真的在负债 lol 世界上最富有的公司正在为人工智能基础设施建设而负债。 尽管拥有数百亿美元的净利润，从 $AMZN 到 $GOOGL 的公司都大幅增加了资本支出(capex)： 以至于部分公司预计将在 2026 年出现净现金流出。 以下是结果及受益者： 亚马逊 ( $AMZN ) 2025: +460 亿美元 2026 (预估): +110 亿美元 Alphabet ( $GOOGL ) 2025: +803 亿美元 2026 (预估): +130 亿美元 Meta Platforms ( $META ) - 净债务(Net Debt) 2025: +229 亿美元 2026 (预估): -70 亿美元 (预计转为净债务) 微软 ( $MSFT ) 2025: +492 亿美元 2026 (预估): +590 亿美元 甲骨文 ( $ORCL ) 2025: -980 亿美元 (净债务) 2026 (预估): -1150 亿美元 微软似乎处于最安全的位置。虽然亚马逊和谷歌主要依靠运营收入为人工智能基础设施建设提供资金。 然而，巨大的现金流缓冲已消失。甲骨文和 Meta 似乎正在通过负债来推动建设，尽管 $META 实现了惊人的运营收入数字。 现在，资金流向哪里？ 运营收入预测： 1. $NVDA - 美国 1350 亿 -> 1865 亿 -> 2401 亿美元运营收入 2. 三星 - 韩国 ~302 亿 -> ~1700 亿 -> ~2267 亿美元 3. SK 海力士 - 韩国 ~327 亿 -> ~1240 亿 -> ~1610 亿美元 4. $TSM - 台湾 ~622 亿 -> ~863 亿 - ~1033 亿 – 1053 亿美元 5. $AVGO - 美国 ~415 亿 -> ~628 亿-651 亿 - ~847 亿-933 亿美元 当然，这些是基于分析师预测的粗略估计。 然而，从总体趋势来看，这看起来像是一种杠杆式赌注，即人工智能基础设施建设完成后将带来自由现金流(FCF)的分红回报。 但明显的赢家似乎是英伟达、三星、SK 海力士、博通和台积电。 随着超大规模云服务商(hyperscalers)中部分公司陷入债务，他们将资产负债表转移给这些供应商，期望从人工智能支出中获得长期投资回报率(ROI)。
英文原文
@LuffyDDK They literally are going into debt lol https://t.co/A4KDf93pYk
-
看好$XLU虚值长期期权，因AI电力需求爆发叠加降息周期带来历史性机遇。
如果我要在1年内将10万美元变成100万美元。 我会选择：$XLU 虚值(OTM) 2年期长期期权(LEAPS) 2026年是现代市场历史上首次同时出现： - 利率下降 - AI推理(AI Inference) + 基础设施建设 通过映射分析，$XLU 有潜在约40%的涨幅（虚值期权可能带来1000%+的收益）。 这是我的宏观论点： 1. 降息 当美联储在不引发衰退的情况下降息时，公用事业公司的债务成本降低，机构投资者会将低收益的现金转向公用事业股息。 这会导致估值倍数立即扩张： 1995年：标普公用事业板块(S&P Utilities)在1995年回报+31.3%，1996年再+12.1%——累计回报约47% 2019年中周期降息：结果：$XLU 在该年产生+25.9%的总回报 标准的软着陆降息周期自然映射为25%至30%的基础回报。而我们要进入2026年的新降息周期。 2. 基础设施超级周期资本支出(CapEx) 基础设施资本支出为该板块带来复合盈利增长。继2000年代初之后，公用事业公司进入大规模资本支出周期以现代化老化的电网基础设施。 由于他们不断支出并扩大其受监管的费率基数(rate base)，$XLU 在2004年回报+23.5%，2005年+16.3%，2006年+20.8%，2007年+18.4%。 然而这一次： 2026年8000多亿美元的AI建设支出，使得2004年的电网现代化看起来像零钱一样微不足道。 因此，你有来自#1降息的估值倍数扩张（+15%至+20%），以及来自#2资本支出历史数据的每股收益(EPS)增长（+18%至+20%）。仅从历史教训来看。 但2026年是历史上AI使用带来的最独特时刻。 仅从我自己的模型预测来看，由于AI极端扩张，所有以前的估计可能都是错误的（例如美国能源部/劳伦斯伯克利国家实验室的预测）： 超大规模云服务商(Hyperscaler)资本支出流入（支出）（亚马逊、微软、Meta、谷歌、甲骨文）进入数据中心估算： 2024年：2200亿美元 2025年：3500亿美元 2026年：5500亿美元 2027年：8000亿美元 2028年：1.2万亿美元（4年增长：+445%） 美国数据中心电力使用量： 2024年：190太瓦时(TWh) 2025年：280太瓦时 2026年：430太瓦时 2027年：650太瓦时 2028年：980太瓦时（4年增长：+415%） AI消耗的总美国电力百分比： 2024年：美国电网的4.5% 2025年：6.6% 2026年：8.2-10.2% 2027年：13.4-15.4% 2028年：21.3-23.3% 劳伦斯伯克利国家实验室和美国能源部似乎低估了AI使用量（他们预测到2028年约为12%） 物理电网容量需求： 2024年：18吉瓦(GW) 2025年：35吉瓦 2026年：65吉瓦 2027年：105吉瓦 2028年：160吉瓦 基本上你可以看到2026年到2028年是拐点，而2024-2025年是爬坡期的缓慢年份。 然后是独立公司的“绝望溢价”。因为电网容量已售罄，科技巨头向公用事业公司支付巨额溢价以插队。例如PJM互联电网（弗吉尼亚“数据中心巷”），容量价格从2024年的每兆瓦日28.92美元飙升至2026/2027年令人难以置信的329.17美元。 $VST 或 Constellation 作为独立电力生产商在ETF中占很大权重。 纵观全局，你可以看到从2026年（现在）到2028年的极端扩张，以及用于建设基础设施的极端资本支出，与往年相比。 2026年是现代市场历史上第一次，所有因素同时为枯燥的电网/电力板块发力，其中AI是最大的顺风。 正如埃隆·马斯克所说：“数十亿美元最先进的硬件。闲置黑暗。不是因为芯片不工作。而是因为没有足够的电力来运行它们”。 再次强调，2026年由于AI和做市商(MMs)基于历史隐含波动率(IV)（极度平坦~14%-16%）定价虚值看涨期权，是一个绝对的历史异常值。 我们看到AI推理（超出之前的测量范围）以及训练（根据OpenAI今天的报告）的爆发。 所以，地球上最无聊的板块（电力/电网），可能会因为超大规模云服务商/政府对电网改进的支出 -> AI推理/训练的极端电力消耗 -> 降息等因素，成为重大反弹的起点。 这只是我的个人论点，期权伴随风险并放大下行风险。这些也是我自己的预测，不确定是否会高于或低于它们。 但基本上： 2026年是一个绝对的历史异常值。 美国的新瓶颈是电力。 有来自AI的极端需求，极端资本支出，降息： $XLU 看起来是暴露于此的最佳交易。 时间会证明这是否正确。
英文原文
If I had to turn $100k -> $1M in 1 year. It would be: $XLU OTM 2 year leaps 2026 is the first time in modern history markets have: - falling interest rates - AI inference + buildout There's a potential ~40% for XLU (1000%+ OTM), from mapping. Here's my macro thesis: 1. Rate Cuts When the Fed cuts rates without a recession, utility debt becomes cheaper, and institutional rotates low-yielding cash to for utility dividends. This causes immediate valuation multiple expansion: 1995: The S&P Utilities sector returned +31.3% in 1995 and another +12.1% in 1996 - ~47% cumulative return 2019 Mid-Cycle Cut: Result: XLU generated a +25.9% total return in that single year Standard soft-landing rate-cut cycle naturally maps to a 25% to 30% baseline return. And we're entering a new rate cut cycle in 2026. 2. The Infrastructure Supercycle Capex Infra CapEx gives the sector compounding earnings growth. Following the early 2000s, utilities entered a massive CapEx cycle to modernize aging grid infrastructure. Because they were constantly spending and expanding their guaranteed rate base, XLU returned +23.5% in 2004, +16.3% in 2005, +20.8% in 2006, and +18.4% in 2007. However this time: The $800B+ AI buildout of 2026 makes the 2004 grid modernization look like pennies. So you have Valuation Multiple Expansion (+15% to +20%), from rate cuts from #1. EPS growth (+18% to +20%) from #2 from capex spend historically. Just from a history lesson. But 2026 is the most unique moment in history from AI usage. Just from my own model projections as all former estimates are likely wrong from extreme AI ramp (eg. DOE/LBNL projections): Hyperscaler CapEx Inflows (Spend) - (Amazon, Microsoft, Meta, Google, Oracle) into DCs est: 2024: $220 Billion 2025: $350 Billion 2026: $550 Billion 2027: $800 Billion 2028: $1.2 Trillion (Growth: +445% over 4 years) U.S. Data Center Power Usage: 2024: 190 TWh 2025: 280 TWh 2026: 430 TWh 2027: 650 TWh 2028: 980 TWh (Growth: +415% over 4 years) % of Total U.S. Electricity Consumed by AI: 2024: 4.5% of the U.S. grid 2025: 6.6% 2026: 8.2-10.2% 2027: 13.4-15.4% 2028: 21.3-23.3% Lawrence Berkeley National Laboratory and the Department of Energy seem off by AI usage (they're projecting ~12% by 2028) Physical Grid Capacity Demand: 2024: 18 GW 2025: 35 GW 2026: 65 GW 2027: 105 GW 2028: 160 GW Basically you can just see 2026 into 2028 being the inflection point whereas 2024-2025 where slower years on the ramp up. Then there's the "Desperation Premium" for independent companies. Because grid capacity is sold out, tech giants are paying massive premiums to utilities to cut the line. eg. PJM Interconnection (Virginia "Data Center Alley"), capacity prices spiked from $28.92 per MW-day in 2024 to an unfathomable $329.17 per MW-day for 2026/2027. $VST or Constellation are a large weighting in the ETF as independent power producers. Across the board, you can see the extreme ramp from 2026 (now) into 2028 compared to previous years, alongside extreme capex going into building the infrastructure. 2026 is the first time in modern market history that every single thing is firing at the same time for the boring grid/power sector with AI as the biggest tailwind. And as Elon quotes it: "Billions of dollars of the most advanced hardware. Sitting dark. Not because the chips won't work. Because there's not enough electricity to run on them". Again 2026 is an absolute historical anomaly due to AI and MMs have priced in historical IV (extremely flat ~14%-16%) for OTM calls. We're seeing an explosion in AI inference (beyond previous measurements) as well as training (per OpenAI report today). So the most boring sector on earth (power/grid), might just be the start of a major rally due to hyperscaler/gov spend into grid improvements -> extreme power consumption from AI inference/training -> rate cuts and others. This is just my personal thesis, options come with risk and magnifies downside too. These are also my own projections, no certainty if they will exceed or be lower than them. But basically: 2026 is an absolute historical anomaly. New bottleneck in the US is power. There's extreme demand from AI, extreme capex, rate cuts: $XLU looks like the best trade for exposure. Time will tell if this is right or not.
-
AXTI与住友垄断InP衬底，成AI光子化关键瓶颈。
你听劝了吗，匿名者？- $AXTI https://t.co/Pcuu4M3muO （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）： 警告：整个AI行业可能会受制于两家公司的瓶颈： 1. $AXTI（市值7亿美元） 2. $SMTOY（市值317亿美元） 这两家公司控制了全球60-70%以上的磷化铟(InP)衬底市场。 未来的$NVDA、$GOOGL TPU v7集群、$META、$MSFT、$AMZN超大规模数据中心集群都需要基于磷化铟(InP)的激光器和接收器。 $AVGO、$LITE、$COHR使用磷化铟(InP)制造用于800G/1.6T光收发器的电吸收调制激光器(EML)、分布反馈(DFB)激光器和其他光基础设施。 如果没有磷化铟(InP)衬底，供应链就会瘫痪。 在查看了从TPU到Maia的物料清单(BOM)后，看起来未来的专用集成电路(ASIC)+GPU+超大规模部署严重依赖光子学。 而这两家供应商可能会冻结全球磷化铟(InP)衬底市场，涵盖几乎所有领域： - 超大规模光模块（TPU集群等） - 光收发器（5g，数据） - 激光雷达(LiDAR)（自动驾驶出租车，无人机，军事） - 光模块（互连集群） - 硅光子学激光芯片（Nvidia未来的共封装光学(CPO)和Intel/Broadcom硅光子学(SiPh)引擎使用磷化铟(InP)连续波激光阵列。） 由于这些公司占据了市场供应的大多数： -AXTI（估计~30-35%） -住友电工(Sumitomo)（估计~30%） - JX日矿（估计10-15%） 就这些。（例如，Yole 2021年的行业报告指出“住友电工+AXT共同拥有‘超过75%’的磷化铟(InP)衬底市场”） 超大规模/AI正在向光子学发展，但整个AI行业是脆弱的。 如果$AXTI或$SMTOY中的任何一家停止供应材料，整个未来的AI建设就会瘫痪。更疯狂的是，一家7亿美元的公司可能成为这一切的中心。 随着AI行业向光子学转变，磷化铟(InP)衬底可能会与高带宽内存(HMB)一起成为最大的瓶颈之一。
英文原文
Did you listen anon? - $AXTI https://t.co/Pcuu4M3muO
-
$EWY期权IV被低估，杠杆衍生品将推高波动率，看涨期权或受益。
$EWY 看起来即将变得更加波动。 做市商可能没有正确地将跳跃期权隐含波动率(IV)定价，因为 IV 可能会扩张至 55% 以上。 有两个原因源于未实现（但即将到来）的波动性。 - 10 倍杠杆的三星/SK 海力士通过去中心化金融(DeFi) 已经出现，因为它们开始产能爬坡。 如果规模变得很大，这会在期货（如 $EWY）的清算和对冲上产生大量的未实现波动性。 $EWY 由一半的 SK 海力士/三星组成，而韩国指数本身已经非常波动。 - 三星和 SK 海力士的 2 倍杠杆单只股票 ETF 即将到来 2026 年 1 月，韩国监管机构正式宣布允许国内 2 倍杠杆单只股票 ETF。 同样，这增加了未实现的波动性。 但就目前的期权 IV 定价而言，$EWY 的波动幅度相当极端。 如果非要打个比方，它的百分比波动范围肯定大于 $MRVL（IV 为 55-56%，而 $EWY 仍为 41%）。 三星的对应物可能是 $MU（约 70%），SK 海力士更接近 $SNDK（91%），但可能在 70-80% 范围内。 而且……韩国指数基本上就是 SK 海力士和三星。 随着三星/SK 海力士开始从存储超级周期中重新评级，这种波动性可能会成为新常态。但做市商可能仍然在使用广义自回归条件异方差(GARCH) 模型或基于长期平均值（当时指数完全平坦）的波动率预测投影。 来自 10 倍杠杆衍生品的数量将仅仅为这些个股的对手方和做市商创造更多的对冲/波动性 -> 这会滞后地反映在 $EWY 的定价中。 简而言之：随着做市商未能正确定价 IV 和未实现波动性，$EWY 看涨期权可能会从 Vega 扩张中受益。
英文原文
$EWY looks like it's about to get more volatile. Market makers might not be pricing in leap option IV correctly as IV may expand past 55%. Two reasons due to unrealized (but upcoming) volatility. - 10X leveraged Samsung / Sk Hynix through Defi are here as they begin ramp-up. This creates a great amount of unrealized volatility from liquidations and hedging on futures like $EWY if this ends up becoming large. $EWY is half Sk Hynix / Samsung, and the South Korean index is already really volatile. - 2x leveraged single-stock ETFs for Samsung and SK Hynix are coming In January 2026, South Korean regulators officially announced they will permit domestic 2x leveraged single-stock ETFs. Again, this just adds to unrealized volatility. But as it stands right now regarding option IV pricing. $EWY moves pretty extreme amounts. If I had to give a comparison, it definitely moves in greater % ranges than $MRVL, which sits at 55-56% IV (while $EWY is 41% still). Samsung's equivalent is probably $MU (70%-ish), Sk Hynix is closer to $SNDK (91%) but likely in 70-80% range. And... South Korea's index is basically just Sk Hynix and Samsung. This volatility would likely be the new norm as Samsung/SK Hynix begin their re-rating from the memory supercycle. But MMs are still probably using garchr models or volatility forecasting projections against long-term averages (where index was completely flat) And the amount of leveraged derivatives from 10x will just create more hedging/volatility on these individual stocks from counterparties and market makers -> that gets priced in late to $EWY. TLDR: $EWY calls likely benefit from vega expansion as MMs aren't pricing in IV and unrealized volatility correctly.
-
建议做多$XLU LEAPS，看好AI电力需求激增及降息利好。
交易思路：做多虚值(OTM) $XLU LEAPS（2年期，2027年12月/2028年1月到期）。 受AI驱动，这感觉像是一次千载难逢的做多机会。 $XLU 集中持有 $VST / $CEG 等电力公司。 两个原因： 1. AI数据中心(DC)电力使用带来的范式转变。 2. 基于历史平均水平（自2000年代以来持平）的低隐含波动率(IV)（约14%）。 AI的电力消耗是天文数字。 这一点怎么强调都不为过。 历史上从未有过数据中心消耗如此多吉瓦(GW)电力的情况，尤其是当它们需要核反应堆的输出来训练大语言模型(LLMs)时。 这迫使 $META、$AMZN、$GOOGL 等公司签署多年协议以尽可能多地消耗电力。然而，电力仍然供不应求。 -> 因此，数万亿美元可能会涌入电网升级。 通常利率会损害该板块，但我们将进入更多降息周期，这使得该板块成为更好的做多标的。 OpenAI致国会的信恳请美国投资能源以与中国竞争。 因此，由于以下原因，这感觉像是一次十年一遇的做多机会： - AI消耗所有可用电力带来的范式转变 - 为与中国竞争而进行的数万亿美元电网升级 - 降息 以及基于历史平均水平的低隐含波动率定价。
英文原文
Trade Idea: Long OTM $XLU leaps (2 years, Dec 2027/Jan 2028). This feels like once-a-generation long due to AI. XLU has concentration in $VST / $CEG power companies. Two reasons: 1. Paradigm shift due to AI DC electricity usage. 2. Low option IV (~14%) based on historical averages (flat since 2000s). AI power usage is astronomical. This cannot be understated. Never before in history have DCs use up this much GWs in power, especially when they require outputs of nuclear reactors for training LLMs. This forces $META, $AMZN, $GOOGL, and others to sign multi-year agreements to consume as much power as possible. And yet they still don't have enough. -> So, trillions would likely be poured into grid upgrades. Usually interest rates hurt the sector but we're going into more rate cuts, so it makes the sector a much better long. OpenAI's letter to congress pleaded the US to invest in energy as well to compete vs. China. So, this feels like a once-a-decade type long due to: - paradigm shift eating up any available power from AI - trillions in grid upgrades to compete vs. China - rate cuts. And low IV pricing from historical averages.
-
利用韩国指数期权定价错误，做多$EWY看涨期权以捕捉存储周期波动率扩张。
指数下跌-2.46%。整个期权链全线飘红+13-20%。 这就是当你发现做市商在期权链中存在定价错误时会发生的情况。 话虽如此，隐含波动率(IV)回升至更合理的38-39%，但个别组件（SK海力士、三星）的波动率可能仍有几个百分点的偏差。 （引用内容翻译）： 我发布在“淋浴思考”频道的交易思路： 韩国指数波动率套利并利用布莱克-斯科尔斯模型。 $EWY 看涨期权似乎存在定价错误。 这是贝莱德旗下的韩国指数，主要由存储芯片（三星电子、SK海力士）构成。 尽管该指数被定价为普通指数的隐含波动率，但个股每日波动2-5%+，且1年涨幅达136.25%。 三星波动剧烈。SK海力士波动剧烈（例如预估65%-80%）。 但通过指数组合后的定价远低于低贝塔值的 $GOOGL (37.33%) 和 $AMZN (39.12%)，隐含波动率仅约32%。 我观察 $EWY 一段时间，它确实看起来波动很大。 至于定价，我猜测做市商(MM)基于历史平均值（5-10年）定价隐含波动率，当时韩国指数完全持平。他们预期两年后的看涨期权会回归均值。 但这种波动率应成为新常态，因为市场正在定价新的存储超级周期（例如 $TSM 从30% IV升至46.2% IV）。 看涨期权将从三星+SK海力士带动指数上涨中受益。 主要好处是你能获得 $KORU 无法提供的 Vega 扩张。 你也无法像美国个股那样获得这种期权做市商的压盘效应，因为这是韩国国家指数且期限较长。 简而言之：个别组件SK海力士+三星具有高波动性。 它们基本占指数的一半，但指数期权以低波动率定价，或许是因为基于过去5-10年的历史数据。 看涨期权将从未正确定价的 Vega 扩张中受益，因为做市商的前瞻波动率估计过于锚定于历史已实现波动率，而过去5-10年 $EWY 的波动率很低。
英文原文
Index down -2.46%. The entire option chain green +13-20%. This is what happens when when you find mispricing in option chains by market makers. That being said it’s a more respectable 38-39% IV, but maybe few percent off (SK Hynix, Samsung) individual components volatility still.
-
科技巨头负债投入AI基建，资金流向英伟达等上游供应商。
世界上最富有的公司正在为AI基础设施建设而负债。 尽管拥有数千亿美元的净利润，从 $AMZN 到 $GOOGL 的公司都大幅增加了资本支出(capex)： 以至于部分公司预计将在2026年出现净现金为负的情况。 以下是结果及受益者： 亚马逊 ( $AMZN ) 2025年：+$460亿 2026年（预估）：+$110亿 Alphabet ( $GOOGL ) 2025年：+$803亿 2026年（预估）：+$130亿 Meta Platforms ( $META ) - 净债务(Net Debt) 2025年：+$229亿 2026年（预估）：-$70亿（预计转为净债务状态） 微软 ( $MSFT ) 2025年：+$492亿 2026年（预估）：+$590亿 甲骨文 ( $ORCL ) 2025年：-$980亿（净债务） 2026年（预估）：-$1150亿 微软似乎处于最安全的位置。虽然亚马逊和谷歌主要依靠运营收入来资助AI基础设施建设。 然而，那巨大的现金流缓冲已经消失。甲骨文和Meta似乎正在通过负债来推动建设，尽管 $META 实现了惊人的运营收入数字。 现在，资金流向了哪里？ 运营收入预测： 1. $NVDA - 美国 $1350亿 -> $1865亿 -> $2401亿 运营收入 2. 三星 - 韩国 ~$302亿 -> ~$1700亿 -> ~$2267亿 3. SK海力士 - 韩国 ~$327亿 -> ~$1240亿 -> ~$1610亿 4. $TSM - 台湾 ~$622亿 -> ~$863亿 - ~$1033亿 – $1053亿 5. $AVGO - 美国 ~$415亿 -> ~$628亿-$651亿 - ~$847亿-$933亿 当然，这些是基于分析师预测的粗略估计。 然而，从总体趋势来看，这看起来像是一种杠杆式赌注，即AI基础设施建设完成后将带来自由现金流(FCF)的红利。 但明显的赢家似乎是英伟达、三星、SK海力士、博通和台积电。 随着超大规模云服务商(hyperscalers)中部分公司陷入债务，它们将资产负债表转移给这些供应商，期望从AI支出中获得长期投资回报率(ROI)。
英文原文
The richest companies in the world are going into debt for the AI buildout. Despite hundreds of billions in net income, companies from $AMZN to $GOOGL have increased capex so much: That some are projected to be net negative cash in 2026. Here are the results and who profits: Amazon ( $AMZN ) 2025: +$46.0 billion 2026 (Est): +$11.0 billion Alphabet ( $GOOGL ) 2025: +$80.3 billion 2026 (Est): +$13.0 billion Meta Platforms ( $META ) - Net Debt 2025: +$22.9 billion 2026 (Est): -$7.0 billion (Expected to swing into Net Debt) Microsoft ( $MSFT ) 2025: +$49.2 billion 2026 (Est): +$59.0 billion Oracle ( $ORCL ) 2025: -$98.0 billion (Net Debt) 2026 (Est): -$115.0 billion Microsoft appears to be in the safest position. While Amazon and Google have been largely funding the AI buildout with operating income. However, that large cashflow buffer has vanished. Oracle and Meta appear to be in debt to fuel the buildout, despite $META achieving staggering operating income numbers. Now, where does the money flow into? Operating Income Projections: 1. $NVDA - USA $135.0B -> $186.5B -> $240.1 Billion Operating Income 2. Samsung - Korea ~$30.2B -> ~$170B -> ~$226.7 Billion 3. SK Hynix - Korea ~$32.7B -> ~$124B -> ~$161.0 Billion 4. $TSM - Taiwan ~$62.2B -> ~$86.3B - ~$103.3B – $105.3 Billion 5. $AVGO - America ~$41.5B -> ~$62.8B-$65.1B - ~$84.7B-$93.3B Of course, these are rough estimates based on analyst projections. However, from the general trend, this looks like a leveraged bet that the AI buildout will pay off dividends in FCF after they're finished. But the clear winners appear to be Nvidia, Samsung, SK Hynix, Broadcom, and TSMC. This comes as hyperscalers, with some going into debt, transfer over their balance sheets to them, expecting a long term ROI from their AI spend.
-
AI 驱动下，三星和海力士营业利润增速惊人，有望在 2027 年追平或超越美股巨头。
全球最盈利公司排名（Mag 7 vs. 世界） 2025->2026->2027 年营业利润(Operating Income)预测。 #1: $NVDA (美国, 4.4T 市值) 🇺🇸 ~1350 亿 -> 1865 亿 -> 2401 亿美元 #2 三星电子 (韩国, 8200 亿市值) 🇰🇷 ~302 亿 -> ~1700 亿 -> ~2267 亿美元 #3 $MSFT (美国, 2.9T 市值) ~1285 亿 -> 1530 亿 -> 1815 亿美元 #4 $GOOGL (美国, 3.7T 市值) ~1290 亿 -> 1420 亿 -> 1730 亿美元 #4 海力士 (韩国, 4100 亿市值) ~327 亿 -> ~1240 亿 -> ~1610 亿美元 #5 $APPL (美国, 3.76T 市值) 1331 亿 -> 1460 亿 -> 1605 亿美元 #6 $AMZN (美国, 2.13T 市值) 800 亿 -> 1050 亿 -> 1365 亿美元 #7 $Meta (美国, 1.62T 市值) 833 亿 -> 970 亿 -> 1215 亿美元 #8 $TSLA (美国, 1.31T 市值) 44 亿 -> 80 亿 -> 240 亿美元 韩国 8200 亿市值的三星电子预计将在 2027 年在营业利润上追平 $NVDA。 同时，海力士预计将在 2027 年在营业利润上超越 $APPL 和 $AMZN。 主要结论是，由于人工智能(AI)的加速部署，美国超大规模云服务商和韩国股票的增长令人惊叹。
英文原文
Global ranking of the most profitable companies in the world (Mag 7 vs. World) Projections for 2025->2026->2027 (Operating Income). #1: $NVDA (USA, 4.4T MC) 🇺🇸 ~$135.0B -> $186.5B -> $240.1 Billion #2: Samsung Electronics (Korea, $820B MC) 🇰🇷 ~$30.2B -> ~$170B -> ~$226.7 Billion #3 $MSFT (USA, $2.9T MC) ~$128.5B -> 153.0B -> $181.5 Billion #4 $GOOGL (USA, $3.7T MC) ~$129.0B -> $142.0B -> $173.0B #4 Sk Hynix (Korea, $410B MC) ~$32.7B -> ~$124B -> ~$161.0 Billion #5 $APPL (USA, $3.76T MC) $133.1B -> $146.0B -> $160.5B #6 $AMZN (USA, $2.13T MC) $80.0B -> $105.0B -> $136.5B #7 $Meta (USA, $1.62T MC) $83.3 -> $97.0B -> $121.5B #8 $TSLA (USA, $1.31T MC) $4.4B -> $8.0B -> $24.0B Samsung Electronics, a $820B company in Korea is projected to catch up to $NVDA in 2027 in operating income. Meanwhile Sk Hynix is projected to overtake both $APPL and $AMZN in operating income in 2027. The main takeaway is that the growth of both US hyperscalers and South Korean equities is astounding due to Artificial Intelligence ramp.
-
警示散户勿重仓短期期权，亚马逊周权期权因股价连跌归零。
Reddit 上最受关注的故事： 一名 Robinhood 用户用积蓄购买了五位数金额的 $AMZN 周权看涨期权(Weekly Calls)。 条件是？ 股价需收复 205 美元以上，潜在收益可能超 100 万美元。 否则归零。 今天，亚马逊收于 198.8 美元。 亚马逊已连续第 9 个交易日下跌，导致其周权期权组合在本周结束时变得一文不值。 故事的教训：请停止持有短期期权。 那 2.6 万美元本可以随时间复利增长至 50 万美元，但现在它毫无价值。
英文原文
The most watched story on Reddit: A Robinhood user bought 5 figures of $AMZN weekly calls with their savings. The condition? Recover past $205 for a potential $1M+. Or go to $0. Today, Amazon finished at $198.8. Amazon has now dropped for the 9th straight session, causing their portfolio of weekly options to end the week worthless. Moral of the story: Please stop holding short-term options. That $26K could easily compound to $500K with time, but now it’s worth nothing.
-
利用$EWY期权定价错误，做多看涨期权以捕捉存储超级周期带来的维加扩张。
我发布在“淋浴思考”频道的交易思路： 韩国指数波动率套利，并利用布莱克-斯科尔斯模型(Black-Scholes models)。 $EWY 的看涨期权似乎定价错误。 这是贝莱德(Blackrock)的韩国指数，主要由存储芯片（三星电子、SK海力士）构成。 该指数每日波动2-5%以上，尽管隐含波动率(IV)定价像普通指数，但过去一年仍上涨了136.25%。 三星波动大。SK海力士波动也大（例如估算65%-80%）。 但通过指数组合后的定价远低于低贝塔值的 $GOOGL (37.33%) 和 $AMZN (39.12%)，约为32%的IV。 我观察 $EWY 一段时间，它确实看起来很波动。 至于定价，我猜测做市商(MMs)基于历史平均值（5-10年）定价IV，当时韩国指数完全持平。他们预期两年后的看涨期权会均值回归。 但这种波动应成为新常态，因为市场正在定价新的存储超级周期（例如 $TSM 从30% IV升至46.2% IV）。 看涨期权将从三星+SK海力士带动指数中受益。 主要好处是你无法从 $KORU 获得的维加(Vega)扩张。 你也无法像美国个股那样获得期权做市商的压盘效应，因为这是韩国的国家指数且期限较长。 简而言之：个别成分股SK海力士+三星具有高波动性。 它们基本上占指数的一半，但指数期权以低波动性定价，可能是因为基于5-10年的历史数据。 看涨期权将从维加扩张中受益，因为做市商的前瞻性波动率估算过于锚定于历史已实现波动率，而过去5-10年 $EWY 的历史波动率较低。
英文原文
Trade idea that I published to my shower thoughts channel: Korean Index volatility arbitrage and taking advantage of Black-Scholes models. $EWY long options seem mispriced. This is Blackrock's Korea Index, which is majority memory (Samsung Electronics, Sk Hynix). The stock swings 2-5+% a day, and is up 136.25% 1Y, despite priced like a normal index IV. Samsung is volatile. SK Hynix is volatile (eg. 65% - 80% est). But the combination of the two through the index is priced way less than both low beta $GOOGL (37.33%) and $AMZN (39.12%) at ~32% IV. I've been watching $EWY for a bit and it does look volatile. As for pricing my guess is MMs priced in IV based on historical averages (5-10 years), where the Korean index was completely flat. And were expecting calls 2 years out to revert to the mean. But this volatility should be the new norm as markets price in the new memory supercycle (eg. $TSM went from 30% IV to 46.2% IV). Long calls should benefit from both Samsung + Sk Hynix carrying the index. And the main benefit is vega expansion that you won't get from $KORU. You also can't get this option MM pinning like individual US stocks since this is Korea's national index and long term. TLDR: Individual components SK Hynix + Samsung are highly volatile. They're basically half of the index, but options in index are priced with low volatility, perhaps due to historical 5-10 year data. Long calls benefit from vega expansion that weren't priced in correctly as MM forward vol estimates are anchored too heavily on historical realized vol, which was low for $EWY over the past 5-10 years
-
LPTH是国防锗瓶颈，短期受板块拖累，长期受益于供应链转移。
$LPTH 主要是国防领域的锗瓶颈。它似乎更多是从中国供应链向美国黑钻（Black Diamond）过渡的长期受益者。 它在算法上与国防板块挂钩（近期因缺乏催化剂，从 $ONDS 到 $AVAV 均下跌）。如果美国入侵伊朗，该交易可能会再次火热。 不太可能看到像 $GOOGL 和 $AMZN 资本支出带来的半导体供应链瓶颈那样巨大的短期飙升。或者像 $AXTI 因光子学应用可能看到的瓶颈+涨价。 但从 Andruil、$NOC、$ONDS 等来看，长期应有增量收入。
英文原文
$LPTH is germanium bottleneck mainly for defense. It seems more of a long term beneficiary from transition from China supply chains to US black diamond. And it’s algorithmically tied to defense sector (which is down recently, from $ONDS to $AVAV due to a lack of catalysts). If US invades Iran the trade is probably going to be hot again. Not likely going to see that massive short term spike that semi supply chain bottlenecks do from the $GOOGL and $AMZN capex spend. Or bottleneck + price hikes that $AXTI might see from photonics usage. But long term from Andruil, $NOC, $ONDS and others should be incremental revenue
-
对比法美AI投入，指出法国资金规模远不及美巨头。
@ticker_thoughts 我确实关注了。法国2030年3000万欧元的投入，与$GOOGL和$AMZN约3800亿美元的年度资本支出(capex)相比，对法国AI领域的影响肯定微乎其微。他们并没有打破刻板印象。https://t.co/qXdcTRvmVd
英文原文
@ticker_thoughts I did. That whole 30M euros for 2030 is definitely going to make a large impact on France’s AI sector compared to $GOOGL and $AMZN ~$380B 1 year capex spend. They’re not really breaking stereotypes here. https://t.co/qXdcTRvmVd
-
AXTI因InP供应链瓶颈逻辑6周涨60%，成AI光子学关键瓶颈标的。
自我发布关于 $AXTI 的初始帖子以来，已经过去了6周。 它的股价又上涨了60%。 几乎没有人在这上面讨论这只市值仅7亿美元的小盘股。 尤其是从磷化铟(InP)供应链与光子学的角度。 现在它已成为一家14亿美元的公司，进入了机构和政府的视野。 在此期间： - 中国出口管制禁止了其竞争对手的前体材料 - 7N级磷化铟(InP)价格周周创历史新高 - 他们从Northland筹集了1亿美元以翻倍产能 - $LITE 业绩大超预期，预计光学需求将超过供应 - 所有超大规模云服务商(Hyperscaler)的资本支出均超出预期 - 特朗普签署行政令，从美国境内的外国拥有公司手中接管磷化铟(InP)供应链。 这6周里，我预计光子学在未来两年及以后将增长最快。 每只股票都有波动性（这只股票每天上下波动20%+），所以在入场交易前，重要的是建立你自己的信念。 但 $AXTI 是我从7亿美元微小市值开始的最有趣的瓶颈多头标的之一。
英文原文
It’s been 6 weeks since I made the initial $AXTI post. It’s now up another 60%. Almost nobody talked about this tiny $700M stock on X. Especially not from angles of InP supply chains with photonics. And now it’s a $1.4B company on institution and gov radars. Since then: - China export control banned precursors to their competitors - 7n InP prices hit all time highs week after week - They raised $100M from Northland to double their capacity - $LITE blows out earnings with expectations of optical demand outstripping supply - Every hyperscaler blew out expectations of capex - Trump signed executive orders to seize InP supply chains from foreign owned companies in the US. It’s been 6 weeks. And I expect photonics to ramp up the most over the next two years onwards. Every stock has volatility (this one moves 20%+ up and down daily) so it’s important to develop conviction by yourself before entering a trade. But $AXTI is of my most interesting bottleneck longs starting from a tiny $700M MC.
-
超大规模云厂商资本支出激增导致GPU供应紧张，利好新云服务商。
黄仁勋在CNBC表示：“六年前销售的GPU价格正在上涨。” GPU折旧曾是$NBIS、$IREN和$CRWV最大的担忧。 但现在，它们拥有： - 超大规模云厂商资本支出增加带来的顺风 - 全面的需求增长 - GPU折旧担忧缓解。 所有顺风同时利好新云服务商(Neoclouds)。 （引用内容：根据3Fourteen Research：2月份主要云服务的GPU可用性大幅下降。随着$META、$AMZN、$GOOGL大幅增加基础设施资本支出，主要赢家是新云服务商：$IREN、$NBIS和$CRWV是三家提供AI云服务的企业。随着Opus 4.6等新AI模型的推出，可能出现新一轮需求冲击。从B200等新一代到旧型号，需求全面增长。这是产能限制累积效应的可视化体现。）
英文原文
Jensen Huang on CNBC: "GPUs sold six years ago are increasing in price." GPU depreciation was the largest concern for $NBIS, $IREN, and $CRWV. But now, they have: - Tailwinds from increased capex spend across hyperscalers - Increased demand across the board - GPU depreciation fears eased. Every tailwind just hit at once for Neoclouds.
-
云厂商GPU供应紧缺，Neoclouds因需求冲击受益。
根据3Fourteen Research的数据：2月份主要云服务提供商的GPU可用性大幅下降。 与此同时，$META、$AMZN和$GOOGL大幅增加了基础设施资本支出。 主要的受益者是Neoclouds（新型云服务商）： $IREN、$NBIS和$CRWV是拥有AI云服务业务的三家。 随着Opus 4.6等新AI模型的发布，很可能出现了一波新的需求冲击。 这种需求增长是全方位的，涵盖了B200等新一代芯片以及旧型号。 这是产能限制累积效应的直观体现。
英文原文
According to 3Fourteen Research: GPU availability across major cloud services have dropped immensely in Feb. This comes as $META, $AMZN, $GOOGL have increased infrastructure capex spend immensely. The major victors, Neoclouds: $IREN, $NBIS, and $CRWV are the three with AI cloud offerings. There's likely been a new wave of demand shock with new AI models coming out from Opus 4.6 and others. And the demand increase is across the board with newer gens like B200, and older models. This is the cumulative effect of capacity constraints, now visualized.
-
超大规模云厂商AI资本支出激增，利好上游半导体供应链，建议坚守基本面做多。
市场正遭受AI资本支出(AI capex)冲击的洗礼。 这实质上是超大规模云服务商(hyperscalers)与外国之间的一场AI军备竞赛。 2025年至2026年间，资本支出(capex)增长如下： $GOOGL：从910亿-930亿美元增至1750亿-1850亿美元 $AMZN：从1250亿美元增至约2000亿美元 $META：从720亿美元增至1150亿-1350亿美元 $MSFT：从1100亿美元增至1500亿美元 $TSLA：从85亿美元增至200亿美元以上 当谷歌表示其年度支出（用于数据中心、TPU、基础设施）从去年的920亿美元激增至1850亿美元时，这一规模是巨大的。 $AMZN等公司的资本支出预测涵盖机器人到太空领域，情况不一，但同一家内存公司和半导体公司实际上能从中双重受益。 市场可能不欣赏$GOOGL如此巨大的支出，但上游AI供应链 certainly 乐见其成。 尤其是那些最明显受益的公司，例如： 专用集成电路(ASICs)：$AVGO（谷歌TPU、Meta MTIA等） 内存：SK海力士、三星、美光 $TSM（没有台积电你造不出“上帝”） $NVDA（训练） 每个建设项目中都有隐藏的受益者（例如$MSFT带动的$AAOI，$MRVL带动的$POET，$GOOGL带动的$LITE）。但有时，像内存或$TSM这样显而易见的“傻瓜式”标的反而就是最好的选择。 判断AI交易是否持续的最大信号是超大规模云服务商的支出： 而他们刚刚彻底超出了所有预期。 在AI建设浪潮中保持做多，基本面比宏观环境和短期流动性冲击更重要。
英文原文
Markets have been hit with an AI capex shock. This is effectively an AI arms race between hyperscalers and foreign nations. Between 2025 to 2026, capex has increased from: $GOOGL: $91B - $93B to $175B – $185B $AMZN: $125B to ~$200B $META: $72B to $115B-$135B $MSFT: $110B to $150B $TSLA: $8.5B to $20B+ When Google says they are spending $185 Billion in 1 year (on DCs, TPUs, infrastructure), from $92B the previous year, that is enormous. Capex projections like $AMZN are mixed from robotics to space, but the same memory companies and semiconductors actually benefit twice. Markets might not appreciate $GOOGL spending that much but AI supply chains upstream certainly do. Especially the most obvious ones that benefit from most such as: ASICs: $AVGO (Google TPUs, Meta MTIA, etc.) Memory: SK Hynix, Samsung, Micron $TSM (You cannot build God without TSM) $NVDA (training) There's hidden beneficiaries in each buildout (eg. $AAOI with $MSFT to $POET with $MRVL, $LITE with $GOOGL). But sometimes, the braindead obvious plays like memory or $TSM are are just the best. The biggest signal of whether the AI trade continues is hyperscaler spending: And they just blew all expectations out of the room. Stay long in the AI buildout, fundamentals matter more than macro and short term liquidity shock.
-
有人全仓做多亚马逊看涨期权，财报后爆仓。
@hopefulmetsfan 我是说有人全仓买入 $AMZN 看涨期权。但在今天亚马逊财报发布后，他们被洗盘了。https://t.co/FKa4U27ZyW
英文原文
@hopefulmetsfan I mean someone else did full port $AMZN calls. But they got wiped out after Amazon earnings today. https://t.co/FKa4U27ZyW
-
建议放弃逆向股，重仓三星和SK海力士等AI存储巨头。
老实说，光看这份预估净利润图表…… 让人不禁疑惑，为什么人们还要试图全仓押注像 $PYPL、$FIG 或 $DUOL 这样的逆向反转交易？ 我觉得，坐在像三星或 SK 海力士这样盈利且超增长的公司里“躺平”，才是最容易做的事？ 它们相比“七巨头(Mag7)”规模相对较小，但预计将在 2027 年成为全球最盈利的公司。 有时候，最好的做法就是参与这些显而易见的交易，开启“简单模式”的生活。
英文原文
Honestly just looking at the quoted net income chart… Makes you wonder why people try and full port it into contrarian turnaround plays like $PYPL, $FIG, or $DUOL I feel like sitting back and chilling in profitable, hyper-growth companies like Samsung or SK Hynix is just the easiest thing to do? They’re relatively small compared to Mag7, but projected to become the most profitable companies in the world in 2027. Sometimes the best thing to do is just join in on the no-brainer trades and live life on easy mode.
-
摩根士丹利预测2027年存储巨头利润登顶，确认存储超级周期已至。
存储超级周期已至。 据摩根士丹利(Morgan Stanley)最新报告估算： - 三星(Samsung)净利润约1630亿美元 - 海力士(SK Hynix)约1209亿美元。 这将使三星成为全球最赚钱的公司，甚至超过$NVDA和$GOOGL。 以下是前十名排名： 1. 三星电子 ~1630亿美元 (2027年) 2. Alphabet $GOOGL 1524.4亿美元 3. Apple $APPL 1330.5亿美元 4. Microsoft $MSFT 1276.5亿美元 5. 海力士 $SKHYNIX 1209亿美元 (2027年) 6. NVIDIA $NVDA 1165.1亿美元 7. Amazon $AMZN 952.2亿美元 8. Meta $META 850.9亿美元 9. 伯克希尔哈撒韦 $BRK 815亿美元 10. 摩根大通 $JPM 728.1亿美元 这是基于companiesmarketcap来源的2027年前瞻净利润与过去十二个月(TTM)净利润的对比。感谢@jukan05提供摩根士丹利报告。上述数据排除了沙特阿美(National companies)等国有企业。 来自$MU、三星和海力士的存储超级周期显然已经到来。
英文原文
The Memory Supercycle is here. New reports from Morgan Stanley est. - Samsung's net profit ~$163.0B USD - SK Hynix's ~$120.9B USD. That would make Samsung the most profitable company in the world, even more than $NVDA and $GOOGL. Here are the top 10 rankings: 1. Samsung Electronics ~$163.0B (2027) 2. Alphabet $GOOGL $152.44B 3. Apple $APPL $133.05B 4 Microsoft $MSFT $127.65B 5. Sk Hynix $120.9B (2027 6. NVIDIA $NVDA $116.51B 7. Amazon $AMZN $95.22B 8 Meta$META $85.09B 9. Berkshire Hathaway $BRK $81.5B 10. JPMorgan Chase $JPM $72.81B This is forward 2027 net income compared to TTM net income sourced by from companiesmarketcap. Credit to @jukan05 for the Morgan Stanley report. Excluding National companies like Saudi Aramco in these figures. The memory supercyle from $MU, Samsung, and SK Hynix is here clearly here.
-
发布1月25日美股评级，强烈看好AI、内存及美国供应链瓶颈股，回避高估值与稀释风险标的。
1月25日评级。欧盟关税及$INTC财报后更新。 强烈买入： $SNAP $META 三星电子 SK海力士 $MU 欣兴电子 $TSM $CRCL $AXTI $LPTH $COPX $LIT $AEHR $FORM $AMKR $AVGO $MRVL 买入： $COIN $SMCI $GOOGL $FIG $AMZN $IBIT $RDDT $TTD $HIMS $HOOD $COHR $AMBA $IREN $POET $AAOI $LASR $VPG $OSS $INTC $UMAC $ONDS $AIRO $DPRO $AVAV $BULL $ETOR VLN NBIS GLXY CIFR HUT WULF 存疑 $VELO $SKYT 回避 $UAVS $BKKT WLMT SLNH $PLTR CRWV $ORCL $BMNR $IONQ, RGTI, QBTS _ 强烈买入 Snapchat - 底部约在$7.4，我认为在此位置非常强劲。随着内存资本支出减少及内存变现进入2027年，自由现金流(FCF)增加。只需等待重估。 Meta - 营收同比增长26%极其强劲，上季度产生超$100亿自由现金流。预计下季度财报后走势将走强，此前因光学(环比EPS光学效应超700%)导致BBB抛售，现在应已消化。 三星电子 - 半导体领域的圣杯，三星同时提供高带宽内存(HBM)和代工(Fab)敞口。 SK海力士 - 内存超级周期 美光 - 内存超级周期，但有美国背景支持。 欣兴电子 - 针对HBM、IC载板、玻璃基板、CoWoS及其他所有瓶颈环节的“邪恶”长线持仓。 台积电 - 印钞机，字面意义上不会出错。 Circle - 预计降息2-3次可能会大幅损害Circle净利润，因此已被定价。但在$160亿市值时是极佳的长线标的，他们正在印钞，且应开始看到美元稳定币(USDC)的扩张。 AXTI - LPTH: 磷化铟(InP)/锗等瓶颈。将成为2026年的巨大主题。只需等待AXT的供应链中断或Lightpath的黑钻(Black Diamond)在美国制造。我认为由于产能爬坡->收入增加，下行风险极低，但类似HBM的“登月式”涨价可能存疑。 COPX - 锂：稀土/材料如铜、锂是2026年的极佳长线标的。与上述瓶颈类似，来自中国的供应链中断将导致资金流向确保供应+建设新供应链。 AEHR - 说实话，他们处于AI和机器人两个热门垂直领域。$550万索诺玛(Sonoma)订单可能与美光和碳化硅(SiC)测试有关。看起来是市值低于$10亿的极佳“登月”标的。 FORM - 由于涉及DRAM/HBM及代工/逻辑，可能在美国供应链中变得重要。良率在HBM4中尤为重要。 AMKR - 美国本土制造供应链及台积电->美国转移的极大受益者。 博通 - 财报后近期大幅回调。鉴于超大规模客户ASIC将继续爬坡（尽管有一些延迟），我认为在此位置强烈买入。 Marvell - 与博通故事相同，Marvell因微软Maia延迟传闻而抛售。只需等待2027年营收约翻倍，当市场开始定价这一点，以及在Celestial收购后，他们在互连等其他领域做得很好。 买入 Coinbase - 加密货币近期回调使Coinbase在$570亿市值下价值再次合理。我从未喜欢其交易所部分，但为贝莱德IBIT ETF提供基础设施+与Circle的USDC收入分成，赋予Coinbase相当不错的长期价值。 SMCI - 从$60+暴跌回$30+呈现了极具吸引力的机会。市场极度担忧毛利率->SMCI向海外扩张，特别是主权AI+购买低端Nvidia GPU。且SMCI在那里的毛利率应会提升。也可能因为与客户达成的交易变得“粘性”。他们的营收增长并未停滞，仍达$360亿+。 GOOGL - 此时Gemini可能会接管ChatGPT，所以我会继续做多谷歌。 Figma - 软件板块抛售为许多被重挫的标的如Figma提供了良好机会，其拥有极高的毛利率+稳健增长。 亚马逊 - 价格基本与去年持平，他们一直在增长，AWS表现良好，涉足机器人+太空低轨卫星(LEO)，看起来是未来极佳的长线标的。 比特币 - 始终是极具吸引力的长线标的 Reddit - 估值高，但毛利率极高且不会消失，因为每个人都在用Reddit。 TTD - 2025年的抛售再次呈现了极具吸引力的估值 HIMS - 说实话，在$29时对我极具吸引力，可能会再次放入强烈买入，但当然营收减速非常令人担忧。主要Alpha在于市场未定价Zava收购，仅凭庞大的客户群，他们就能从新客户中衍生大量营收。 Robinhood - 从$140抛售回$100再次为Robinhood提供了良好机会。他们不会消失，加上银行+其他新产品营收扩张，应带来积极顺风。 Coherent - 长线美国供应链，特别是光子学、InP等。 AMBA - 针对边缘AI推理用于机器人爬坡+边缘计算的“登月”长线标的。 POET - 现在基本是1/2现金，通过Celestial间接进入Marvell+超大规模客户。鉴于承销商在$7.25买入，$6.8的股价具有吸引力。 AAOI - 与微软Maia和AWS Trainium绑定的长线标的。两者都尚未真正起飞，所以只是等待游戏。 LASR - 定向能武器非常酷。我不太喜欢基本面如20%左右的营收增长，但技术实在太酷了。 VPG - 与Optimus爬坡绑定的长线标的。我们应在2026年底看到工业用例，2027年底看到消费用例，所以Optimus生产可能现在开始或Q2影响资产负债表。 OSS - 国防板块及边缘AI+$2亿合同的长线标的。 INTC - 做多美国政策，财报并未改变任何观点，只是短期价格。 UMAC - 在此水平上是美国无人机制造的极佳长线标的。 ONDS, Airo, DPRO - 与AIRO, DPRO相同，看多无人机板块。相比几周前美国入侵委内瑞拉并威胁格陵兰时，没有太多巨大的顺风，但主题上看多。 AVAV - 关于将研发类合同->长期合同的错误信息导致抛售，呈现了相当大的上行空间 BULL - 我喜欢像Robinhood, Webull等拥有大量零售用户的券商，因为一旦拥有客户群，就有无尽的变现方式。抛售回$8呈现了极具吸引力的上行空间 ETOR - 抛售过度，净利润同比高，基本50%现金，下行风险低。只需等待财报重估。他们表现也不错，AUM同比70%+，所以不明白为何这样定价。 VLN - 不再像以前那样接近1:1资产/净值，曾有一段时间他们有$1100万+投资(下跌63%)毛利率，$9300万现金，所以会更接近1.1-1.2亿 : $1.4亿市值，这说不通。话虽如此，仍有$8000万远期营收，毛利率从63%->69%，看起来重估机会相当大。市场似乎只是不喜欢与某国相关的公司如Etoro，我想 Nebius - $150亿Clickhouse估值仅显示了分部加总(SOTP)，我不惊讶他们的子公司如Avride最终会超越主营业务。话虽如此，由于$20亿+ ATM在公开市场出售，近期有卖压。随着他们在2026年底达到$70亿ARR目标，应会极速爬坡。 GLXY, CIFR, HUT, IREN, WULF - 继续做多colo及其他新云板块标的。话虽如此，大多数自2025年以来已上涨30-40%+，所以它们不再是强烈的买入，因为已被定价。但仍有很多上行空间。 存疑 VELO - 很多人问我对此的看法，因为FinX喜欢这只股票。他们有很酷的客户如SpaceX，但基本面看起来糟糕。 ~$1180万现金 + $1750万发行 vs. ~$2300万债务。他们几乎没有剩余跑道，现在买入的人可能会被稀释。 Velo是拥有像IQE(欧洲InP供应链)这样极佳客户群但基本面糟糕的完美例子。 SKYT - 它是美国本土制造供应链的极佳标的，用于量子组件或边缘等酷东西。受益于芯片法案，但营收增长非常缓慢。它比Velo是更好的投机性长线，因为基本面更好。 24%左右的低毛利率，非常低的运营利润率，显然已计入市值，但美国纯代工应是一个溢价的好故事。底线是增长不够快。 回避 UAVS - 无尽的稀释机器，超过100%的市值已给予可将对市值25%以下的股份100%+转换为零售出售的套利投资者 BKKT - $3亿ATM稀释，而市值为$5.5亿。无尽的稀释机器 沃尔玛 - 43倍市盈率，不可能。 SLNH - 前方有大量稀释。 Palantir - 担忧估值P/E Coreweave - 担忧巨额债务，$10亿+债务利息严重损害自由现金流。然后是OpenAI的分配/建设，如果考虑到Gemini正在接管OpenAI的市场份额，对其能否履行合同义务存在极度、极度的担忧。 Oracle - 可能有技术性反弹，但说真的，他们为OpenAI(如Stargate)花费了太多资本支出，像Coreweave一样，OpenAI在能否履行合同义务方面存在极度担忧 BMNR - 无尽的稀释机器为愚蠢的项目融资，如$2亿投入Mr. Beast的公司。预期长期ETH质押ETF，做空BMNR，溢价将消失，例如$2亿现金投入Mr. Beast的公司流动性很差。 IONQ, RGTI, QBTS - 量子估值非常拉伸。 _ 总体想法： 我个人保持极度做多，这只是个人想法，非投资建议(NFI)。 许多小盘股和投机性公司自1月1日以来已经重估，我不认为许多50-100%的涨幅会持续（周五我们看到这些标的中有很多获利了结）。 话虽如此，特朗普正试图进一步降息（再降息2-3次），特别是因为中期选举即将到来。 SPY上涨 = 当选几率更大。所以我会保持极度做多直到中期选举后。 话虽如此，这有助于成长、投机性公司等。但我们已经看到这在很大程度上已被定价，如我最喜欢的长线标的之一Rocketlab，季度营收$1.55亿却达到$450亿+市值，所以我开始质疑估值->将许多头寸转向更多价值型（如软件下跌或内存超级周期）。 主题上我极度看多 - AI, 内存, 半导体 - 瓶颈 - 关键材料等。 非常看多 - 美国本土制造供应链 看多 - 国防板块 并会寻找软件到社交媒体公司等的波段交易/复苏/重估机会，鉴于近期的抛售。
英文原文
Jan 25th Ratings. Post EU Tariffs and $INTC ER. Strong Buy: $SNAP $META Samsung Electronics SK Hynix $MU Unimicron $TSM $CRCL $AXTI $LPTH $COPX $LIT $AEHR $FORM $AMKR $AVGO $MRVL Buy: $COIN $SMCI $GOOGL $FIG $AMZN $IBIT $RDDT $TTD $HIMS $HOOD $COHR $AMBA $IREN $POET $AAOI $LASR $VPG $OSS $INTC $UMAC $ONDS $AIRO $DPRO $AVAV $BULL $ETOR VLN NBIS GLXY CIFR HUT WULF Questionable $VELO $SKYT Avoid $UAVS $BKKT WLMT SLNH $PLTR CRWV $ORCL $BMNR $IONQ , RGTI, QBTS _ Strong Buy Snapchat - Bottomed around $7.4, imo very strong at this level. Increased FCF from memory opex reduction and memory monetization into 2027. Just a waiting game for re-rating. Meta - 26% Y/Y revenue growth is extremely strong, produced $10B+ FCF last quarter. Expect it to pick up after next quarter earnings due to optics (700%+ Q/Q EPS optics) that caused selloff last time from BBB. Samsung Electronics - Holy grail for semis, samsung provides exposure to both hbm and foundry. SK Hynix - memory supercycle Micron - memory supercycle, but with US backing. Unimicron - unholy long for hbm, ic substrates, glass core, cowos, and all other bottlenecks. TSM - money printer, literally can't go wrong with this. Circle - 2-3x projected rate cuts would likely hurt circle net income a lot, hence why it's being priced in. But amazing long at $16B as they print money and should start seeing expansion of USDC. AXTI - LPTH: Bottlenecks for InP / Germanium, etc. Will be a huge theme going into 2026. It's just a waiting game for both supply chain disruption (in AXT) or made in America w/ black diamond in Lightpath. Low downside risk imo due to capacity ramp -> revenue increase, but moonshot HBM type price increases might be questionable. COPX - LI: Rare Earths/Materials like Copper, Lithium are great longs for 2026. Similar with bottlenecks above, supply chain disruptions from China will cause money to flow into securing supply + buildout out new supply chains. AEHR - Honestly, they sit in two different hot verticals in AI and Robotics. $5.5m Sonoma order might be linked with Micron and SiC Testing. Seems like an extremely good moonshot sub $1B MC. FORM - Likely to be important in US supply chains since they do DRAM/HBM, and Foundry/Logic. & Yield is especailyl important w/ hbm4. AMKR - extreme beneficiary of made in america us supply chains and tsm -> US AVGO - Large correction recently post earnings. Strong buy IMO at these levels given hyperscaler ASICs will continue to ramp (even though there's been some delays). MRVL - Same story with Broadcom, marvell selloff after rumors of Microsoft maia delays. It's just a waiting game for ~2x revenue in 2027 and when markets start pricing that in, and after celestial acqusition, they're doing great stuff in other segments like interconnects. Buy Coinbase - Recent correction to Crypto makes Coinbase value decent again at $57B. Was never a fan of their exchange portion, but providing infra for Blackrock IBIT etfs + USDC revenue sharing with Circle, gives Coinbase pretty good long term value. SMCI - Extreme selloff from the $60's+ back to $30's presents attractive opportunity here. Markets are extremely concerned about gross margins -> SMCI expanding overseas, especially with soverign AI + buying lower end nvda gpus. and SMCI's margins should increase over there. Also likely due to deals to become sticky w/ customers. It's not like they're dying revenue growth to $36B+. GOOGL - Gemini at this point would likely take over chatgpt, so i'd remain long google. Figma - Software selloff provides good opportunity into a lot of the hammered names like Figma which extremely high gross margins + sturdy growth Amazon - Basically same price as last year, they've been growing, AWS is doing fine, they're in robotics + space LEOs, and just seems like a great long going forward Bitcoin - Always an attractive Long Reddit - High valuations, but extremely high gross margins and not going anywhere since everyone uses reddit. TTD - Selloff from 2025 presents attractive valuations again HIMS - Honestly extremely attractive for me at $29, might be put into strong buy again, but of course revenue deceleration is very worrysome. Main alpha is that markets arent pricing in Zava acqusition and just from sheer customer base, they can derive a lot of revenue from new customers. Robinhood - Selloff from $140 back to $100 presents a good opportunity for Robinhood again. They're not going anywere, plus new product revenue expansion from banking + others, should present positive tailwinds. Coherent - Long US supply chains, esp. for photonics, inp, etc. AMBA - Moonshot long for edge AI inference for robotic ramps + edge compute. POET - Basically 1/2 cash now, backdoor into marvell + hyperscalers through celestial. Attractive upside at $6.8 given underwriters bought at $7.25 AAOI - long play tethered to msft maia and aws trainium. both of them haven't really taken off yet so it's just a waiting agme LASR - energy directed weapons are super cool. i dont quite like the fundamentals like low 20% revenue growth, but the technology is just way too cool. VPG - Long play tethered to optimus ramp. we should see industrial use cases EOY 2026 and consumer EOY 2027, so maybe optimus productions starts hitting balance sheet now or q2. OSS - Long play on defense sector and edge AI + $200m contract. INTC - long on us policy, earnings didn't really change any perspective, just short term price. UMAC - Great long play at these levels on drone manufacturing in US. ONDS, Airo, DPRO - Same with AIRO, DPRO, bullish on drone sector. There's not much of a massive tailwind compared to a few weeks ago when US was invading venezuela and threatening greenland, but thematically bullish. AVAV - selloff from misinformation about converting r&D type contracts -> long term contract presents considerable upside BULL - I do like brokerages like robinhood, webull, etc. that have a ton of retail users since there's endless ways to monetize once you own the customer base. selloff back to $8 presents attractive upsdie ETOR - selloff way overblown, high net income y/y, basically 50% cash, low downside risk. just waiting for re-rating per earnings. they're doing well too, 70%+ Y/Y AUM, so not sure why they're being priced in like this. VLN - not quite the same anymore as close to 1:1 assets/nav, at one point they had $11m+ inv (off 63%) gross margins, $93M cash, so would have been closer to 110-120m : $140m MC, which made no sense. That being said still $80m fwd revenue off 63% -> 69% gross margins, seems like considerable opportunity for re-rating. Markets just don't seem to like companies eg. Etoro related to a certain country, I guess Nebius - $15B clickhouse valuation just goes to show Sum of Parts, where I wouldn't be surpirsed if their subsidaries like Avride ended up overtaking the main business. That being said, near term selling pressure due to $2B+ ATM being sold on open market. Should ramp up extremely fast as they meet their $7B ARR target EOY 2026. GLXY, CIFR, HUT, IREN, WULF - Remain long on the colo, and other neocloud sector plays. That being said most are up 30-40%+ since 2025, so they're not exactly a strong buy anymore as they've been priced in. But lot of upside remains. Questionable VELO - Lot of people asked my opinion on this since FinX loves this stock. They have really cool customers like SpaceX, but fundamentals look terrible. ~$11.8M cash + $17.5M offering vs. ~$23M. debt. They barely have any runway left and people buying now are likely to be diluted. Velo is the perfect example of amazing customer base like IQE (EU for inP supply chain), but terrible fundamentals. SKYT - It's a great made in america supply chain company for a lot of cool stuff like quantum components or edge. Benefits from CHIP act, but very slow revenue growth. It's a lot better speculative long than Velo since it has better fundamenatls. Lower gross margins like 24%, very low operating margins, is obviously priced into MC but U.S. pure-play foundry should be a good story for premium. Bottom line are not really growing too fast though. Avoid UAVS - Endless dilution machine with over 100%+ of marketcap given over to arbitrage investors that can convert 100%+ of the shares under 25% market value to sell on retail BKKT - $300m ATM dilution right now while MC is $550m. Endless dilution machine Walmart - 43 p/e, there's no way. SLNH - Lot of dilution ahead. Palantir - Concern over valuation P/E Coreweave - Concerns over large debt, $1B+ in debt interest hurts FCF a ton. Then there's allocation/buildout for OpenAI, which has extreme, extreme concerns if they can fulfill contract obligations, especially since gemini is taking over market share of openai. Oracle - There might be technical rebound, but seriously, they've spent so much capex just for openai (eg. stargate), and like coreweave, OpenAI, which has extreme concerns over if they can fufill contracts obligations BMNR - endless dilution machine to fund silly projects like $200m into mr. beast's company. Expect long eth staking etfs, short bmnr plays, and premium to go under as $200m cash into mr. beast's company for example is not very liquid. IONQ, RGTI, QBTS - Quantum valuations are very stretched. _ Overall Thoughts: I'm personally staying extremely long, this is just personal thoughts NFI. A lot of small caps and speculative companies have already been re-rated since Jan 1st and I don't expect many of the 50-100% moves to continue (we've seen a lot of profit taking Friday on some of these names). That being said, Trump is trying to cut rates even more (another 2-3x projected), esp. since Midterms is coming up. SPY Up = better chance of getting elected. So I'm staying very long until after Midterms. That being said a lot of this helps growth, speculative companies etc. But we're already seeing this largely priced in like Rocketlab, one of my favorite longs, reaching $45B+ MC off $155m quarterly revenue, so I'm questioning valuations a bit -> pivoting a lot of positions into more value (eg. software drop or memory supercycle). Thematically I'm extremely bullish on - AI, Memory, Semis - Bottlenecks - Critical Materials, etc. Very bullish on - Made in America supply chains Bullish on - Defense Sector And would look for swing trades/recovery/re-rating for stuff like software to social media companies around now given the recent selloff.
-
看好欣兴电子，作为AI供应链关键瓶颈，受益于存储及先进封装需求。
我持有市场忽视的$MU相关头寸： 欣兴电子(Unimicron)。 这家市值160亿美元的公司是极佳的长线标的，因为它是以下领域的基础供应商： - $MU和三星的存储基板 - $NVDA的Rubin/Blackwell架构 - 台积电(TSMC)的CoWoS工艺 - 苹果（iPhone和M系列芯片） - 谷歌TPU - AMZN/超大规模云厂商的ASIC - 英特尔的ABF基板 - 购买欣兴基板的Amkor - 以及全球所有半导体。 欣兴电子是全球AI供应链中最显著的“瓶颈”之一，因为它提供面向2026-2028年的未来核心部件： - 用于HBM4/存储（DRAM/NAND）+ GPU/CPU的IC基板 - 台积电CoWoS中的“S”（基板） - 用于CoPoS的玻璃核心基板 以及当前iPhone等设备所需的所有基础部件。 它处于存储瓶颈、CoWoS瓶颈以及超大规模云厂商ASIC互连瓶颈的中心。 同时也涵盖未来周期的玻璃核心基板等。 AI围绕另一家Micron（欣兴）展开，你可以借此暴露于当今及未来最热门的瓶颈环节。
英文原文
I've hold positions in the $MU that markets sleep on: Unimicron. This one is an incredible long at $16B as they're foundational for: - $MU and Samsung memory substrates - $NVDA Rubin/Blackwell - TSMC for CoWoS process - Apple (iPhones and M-series) - Google TPUs - AMZN/Hyperscaler ASICs - Intel for ABF substrates - Amkor that buys Unimicron substrates - and all the world's semis. Unimicron is one of the most significant "choke points" in the global AI supply chain as they provide the future: - IC Substrates for HBM4 / Memory (DRAM/NAND) + GPUs/CPUs - The "S" (substrate) in TSM CoWoS - Glass Core Substrates for CoPoS for 2026-2028 and all the foundational parts for customer devices like IPhones and semiconductors used today. They're in the center of the memory bottleneck, CoWoS bottleneck, and interconnects bottleneck for Hyperscaler ASiCs. As well as the future cycles with Glass Core substrates and more. AI revolves around the other Micron and you get exposure to most popular bottlenecks like HBM today and tomorrow.
-
BKKT因ATM增发暴跌，警示散户警惕有毒融资结构。
$BKKT 盘前下跌 -23.87%，原因是新一轮 $300M 的自动取款机(ATM)增发稀释。在我看来，这是整个市场中表现最糟糕的公司之一。我为那些在 $45 时跟随某位网红(influencer)买入的散户感到难过，因为他们的投资组合现在已亏损超过 64%。这是一个无尽的循环：新闻引发的错失恐惧(FOMO) -> ATM/大规模稀释，这只会伤害散户。以下是 $BKKT 试图不断掩盖的历史：2024年2月：通过注册直接发行(registered direct offering)发行约 140 万股 + 认股权证。2025年7月：在股价约 $24 时，通过 $75M 的公开发行又发行约 675 万股 + 预融资认股权证(pre-funded warrants)，但以约 $10（60%+折扣）的价格卖给机构。9月：Mike Alfred 加入 -> 他开始每天发布关于该股的帖子。管理层在股价上涨后试图向高管发行 24% 的稀释性奖励以进行套现。今天：当市值仅为 $500m 时，$300 Million 的 ATM 被倾倒在公开市场上，由散户接盘。$BKKT 是我见过的最大的散户资本毁灭公司之一。股票本应是人人受益的正和博弈，但这看起来只是从散户口袋到高管的价值转移。网红推广后 -> ATM 或高管授予稀释。我没有任何持仓，但作为旁观者：这是为什么你应该对融资结构完全有毒的公司进行独立研究(Do Your Own Research)的完美例子。
英文原文
$BKKT is down -23.87% premarket after a new $300M ATM dilution. In my opinion, this is one of the worst companies in the entire market. I feel sorry for people who followed a certain influencer at $45 since retail portfolios are now wiped over 64%. There’s an endless cycle of news fomo -> ATM/massive dilution that only harms retail. This is the history $BKKT endlessly tries to hide: Feb 2024: Issued ~1.4M shares + warrants in a registered direct offering. Jul 2025: Issued another ~6.75M shares + pre-funded warrants in a $75M public offering, when the stock was trading around $24, but sold them to institutions at ~$10 (60%+ discount). September: Mike Alfred joins -> he starts posting about the stock every day. Management after rally tries to issue 24% dilution awards to executives to sell Today: $300 Million ATM dumped in the public market on retail, when MC is only $500m $BKKT so far is of the biggest retail capital destruction companies I’ve seen. Stocks are a positive sum game where everyone benefits, but this just looks like a value transfer from retail pockets to executives. After influencer promotion -> ATM or executive grant dilution I have zero positions but as a spectator: This is the perfect example of why you should do your own research into companies that have completely toxic financing structures
-
AAOI作为垂直整合玩家，受益于美国制造及云厂商扩张。
@yianisz 同意，$AAOI 是少数几家美国垂直整合玩家之一。显然，主要挑战在于实现规模化，但它极大地受益于“美国制造”的顺风——即 $MSFT、$AMZN 等超大规模云服务商的产能扩张。
英文原文
@yianisz Agreed, $AAOI is one of the rare US vertical integration players. Obviously with this, main challenge is getting it to scale, but it benefits a lot from Made in America tailwinds -> $MSFT, $AMZN hyperscaler ramps.
-
驳斥ONDS观点，揭露UAVS通过可转债套利及虚假宣传收割散户。
认为$ONDS的高行权价可转债性质截然不同，这种观点简直是妄想。 $UAVS将超过100%市值的控制权交给了套利投资者，他们可以在$1.23的价格（低于当前股价50%）增发价值1亿美元的股票，并直接赚取差价。 这只有在散户买入该套利头寸时才有效。正如我们所见，该公司为了诱导散户买入，无所不用其极，甚至发布诸如与$AMZN合作的虚假新闻稿。
英文原文
$ONDS convertible note at a higher strike is wildly different and this is a delusional take. $UAVS gave control to arbitrage investors of over 100% the marketcap where they can mint $100 million worth of shares at $1.23 (-50% under current prices) and just pocket the difference. This only works if retail buys into this arbitrage. And as we’ve seen the same company would do anything from putting out fake press releases like $AMZN partnerships to get retail to buy.
-
警告散户远离$UAVS，揭露其融资结构旨在向套利者转移财富。
警告散户投资者： 在22日的投票结果出来之前，远离 $UAVS。 我在研究无人机股票推荐时，发现了SEC文件，其中显示不透明的董事会投票决定将财富从散户转移给套利者。 这是一家无人机公司，因为被美国陆军工程兵团(USACE)、北约(KFOR)和其他武装部队使用，且最近有大量新闻报道，从而获得了散户的关注。 我原本对这个市值7500万美元的公司作为多头标的感到非常兴奋，因为他们将农业应用场景转型为无人机国防承包商。 然而，2025年11月7日的文件是一次巨大的稀释核弹。 基本上，他们在去年的G轮融资中，将整个公司卖给了套利投资者，赋予这些人以1.23美元（比当前市价低75%）的价格购买1亿美元股份的权利。 早在2021年，同一批G轮投资者散布了关于与 $AMZN 合作的虚假信息，诱使散户买入 -> 然后向这些投资者抛售股份（Bonitas Research指出该新闻是假的）。现在整个公司由Alpha Capital Anstalt等相同的套利投资者控制。 这些公司也控制着股价（可以轻松将市值推高至3亿，然后在散户看到价格上涨并买入时进行稀释）。 凭借这一条款，他们现在有效地控制了100%以上的整个公司，他们可以发行新股 -> 获利50%以上 -> 重复。我在研究无人机股票时差点就掉进这个陷阱。 简而言之：买入这只股票实际上是在向套利投资者捐款。董事会投票支持这种结构令人警觉。 股票长期来看是正和博弈，如果论点方向正确，每个人都能受益。在这种情况下，由于当前的融资结构，每个散户投资者都会输。
英文原文
Warning to retail investors: Stay away from $UAVS until the result of the vote on the 22nd. I was researching drone stock recs, and found SEC filings where the shady board voted for a wealth transfer from retail to arbitrage. It's a drone company that's been getting retail attention because it's used by: US Army USACE, NATO (KFOR), and other armed forces with a lot of news published recently. I was super excited about this potential long for a $75M company, since they pivoted their agricultural use case to drone defense contractor. However, the November 7, 2025 Filing was a massive dilution nuke. Basically they sold their entire company to arbitrage investors last year in their Series G round, giving these guys the right to buy $100M shares at $1.23 or 75% under market price current is. Back in 2021, the same Series G investors put out false information about an $AMZN partnership to get retail to buy -> dump shares on them (Bonitas Research pointed out this news was fake). The entire company is now controlled by the same arbitrage investors like Alpha Capital Anstalt. These companies control the price too (can easily make this $300M then dilute when retail sees price go up and buys in) With this clause, they have effective control of 100%+ of the entire company now where they can issue new shares -> profit 50%+ -> repeat. I almost fell for it while looking into drone stocks. TLDR: Buying this stock is effectively donating money to the arbitrage investors. The fact the board voted for this structure is alarming. Stocks positive sum game long term where everyone benefits if the thesis is directionally right. In this case, with the current financing structure, every retail investor loses.
-
CRDO因市场误读亚马逊线缆颜色大跌后反弹，作者批评市场逻辑幼稚。
$CRDO 现已回升 15.47%+。 有趣的是，仅仅因为 $AMZN 不喜欢紫色线缆，Credo 的市值就被抹去了数十亿美元。 Credo：“我们同比增长了 272%，每股收益超出预期 35%。” 市场：“是啊，但那些线缆是紫色的吗？” 有时市场很愚蠢，用幼儿园级别的逻辑进行交易。 https://t.co/L2RjbB04GM
英文原文
$CRDO is now back up 15.47%+. It's funny billions were wiped from Credo just because $AMZN doesn't like Purple. Credo: "We grew 272% Y/Y and beat earnings by 35%." Market: "Yeah, but are the cables Purple?" Sometimes markets are stupid and trade off Kindergarten logic. https://t.co/L2RjbB04GM
-
AXTI垄断InP成AI基建单点故障，市场开始关注。
$AXTI 在磷化铟(InP)领域实际上处于垄断地位，且在中国对日本实施出口管制禁令后，成为以下 AI 基础设施建设中唯一的单点故障(Single Point of Failure)环节： - 英伟达 Nvidia ($4.6T) - 微软 Microsoft ($3.59T) - Meta ($1.64T) - 亚马逊 Amazon ($2.57T) - 谷歌 Google ($3.89T) 看来市场开始意识到这一点了。https://t.co/rxGkcHYBtS
英文原文
$AXTI is effectively a monopoly for InP, and the single point of failure for the AI buildout for: - Nvidia ($4.6T) - Microsoft ($3.59T) - Meta ($1.64T) - Amazon ($2.57T) - Google ($3.89T) after China's export control ban on Japan. Looks like markets are starting to find out. https://t.co/rxGkcHYBtS
-
澄清AMZN谣言，分析AXTI光学瓶颈及铜缆长期需求。
很多人误读了“无缆化”的评论，尽管主要下跌是由关于$AMZN的“color change”（色彩变化）谣言引起的。 你精准指出了$CRDO的看空逻辑，但这我们一直都知道。我的持仓观点较为微妙，因为我看到$AXTI等公司正在对规模化扩展所需的光学组件形成瓶颈。（尤其是昨天住友电工的西方供应链因出口管制而遭受重创，这 largely 是光学建设所必需的） 我们可能会看到对铜缆及铜缆寿命延长的长期需求，尽管Credo确实拥有光DSP垂直领域业务。
英文原文
Tons of people misinterpreted the "cableless comment", though the main drop was the $AMZN color change misinformation. So you've nailed the bear case for $CRDO, but we've always known this. My position is nuanced since I see $AXTI and others bottle necking optical components required for scale out. (especially now that the Western supply chain in Sumitomo got export control nuked yesterday, largely required for optical buildout) We'll likely see long tail demand for copper and copper life extension, although Credo does have optical dsp verticals.
-
澄清CES误导，认为$CRDO错杀，因内部线缆变更及光子学布局而加仓。
我在$CRDO下跌25.7%时加仓，利用了CES上的信息误导。 1. 最初的大幅下跌是因为投资者在$AMZN数据中心发现了橙色和蓝色线缆（暗示$CRDO失去客户）。 实际情况是亚马逊仅为内部物流要求更改颜色——由Mizuho的Vijay Rakesh证实。 2. CES 2026主题演讲——第二次下跌是因为黄仁勋提到Rubin NVL72机架的“无缆”设计。 “无缆”说法特指连接GPU与NVLink交换机的内部铜背板（盲插中板）。 $CRDO的主要产品是有源电缆(AEC)，用于机架外部连接服务器与网络交换机。Nvidia的“无缆”技术仅替换Credo未生产的内部部件。 黄仁勋确实提到了Spectrum-X以太网光子学，但$CRDO也在光子学领域运营，拥有“BlueBird”（光数字信号处理器）和“ZeroFlap”光模块。还有机架顶部连接，这是Credo的主营业务（也是未来2-3年能效最高的解决方案）。 正如我多次在$AXTI中指出的，人们对转向光子学存在更大担忧，但鉴于供应短缺，行业正走向混合模式（铜+光子学）。 总之，投资者只是不了解细微差别，并非故意散布虚假信息。 因此我个人加仓了$CRDO，因为其大部分市值因无关紧要且错误的消息被抹去。
英文原文
I've added $CRDO on the 25.7% drop, taking advantage of CES misinformation. 1. Large drop initially was because investors spotted orange and blue cables in $AMZN DCs (implying $CRDO customer loss). All that happened was Amazon requested a color change for internal logistics. - confirmed by Vijay Rakesh from Mizuho. 2. CES 2026 keynote - Second drop was Jensen Huang said "Cableless" design of Rubin NVL72 rack. The "cableless" claim refers specifically to the internal copper backplane (blind-mate midplane) that connects the GPUs to the NVLink switches within a single rack. $CRDO's main product is AEC, for outside of the racks that connect the entire server rack to the network switch. Nvidia's "Cableless" Tech only replaces internal parts Credo didn't make. Jensen did mention Spectrum-X Ethernet Photonics, but $CRDO is operating in the photonics space too with "BlueBird" (Optical DSPs) and "ZeroFlap" optics. And then there's Top of the Rack Connectivity, which is Credo's main business (which is the most power efficient solution for the next 2-3y). There are larger concerns over the shift to photonics (as I've pointed out with $AXTI multiple times), but industry is moving toward a hybrid model (copper + photonics) given the supply shortage. Again as an overview, investors just did not know the nuances, not intentional disinformation. So I ended up personally adding $CRDO since a large percent of its market cap got wiped from immaterial + incorrect news.
-
材料价格创历史新高，西方公司支付溢价，超大规模厂商尚未开始大规模采购。
@itsthesquonky 我会再给它几个月时间。你已经可以看到材料价格触及历史高点(ATH)，西方公司正在支付巨额溢价。 这还是在产能爬坡(pre-ramp)之前，$MSFT、$AMZN 和 $META 甚至 barely 开始。（$NVDA 预订了产能，造成了最初的冲击）
英文原文
@itsthesquonky I’d give it a few months. You can already see material prices hitting ATHs, with Western companies paying large premiums. This is pre-ramp too, $MSFT, $AMZN, and $META barely even started. ( $NVDA booked out capacity, causing the initial shock) https://t.co/rEG5X7Co81
-
AXTI因InP衬底短缺及中美价差套利，受益于AI供应链瓶颈，具非对称上涨潜力。
磷化铟(InP)衬底短缺与“价差”套利 - $AXTI 市场研究显示，InP和光学组件正变得类似于高带宽内存(HBM)的短缺和价格狂热。 最初的瓶颈是由$NVDA垄断$COHR到$LITE的电吸收调制激光器(EML)产能造成的（TrendForce）。 以下是详细分析： 据报道，$NVDA已对EML激光芯片进行战略性产能预订，将这些特定组件的交货期推至2027年以后。 通过预订像$LITE和$COHR这样受InP限制的集成器件制造商(IDM)的成品产能，英伟达实际上消除了光学引擎的产能。 “二阶效应”：因此，许多超大规模云服务商（如$MSFT到$AMZN）可能推迟了其路线图/爬坡计划，并正在寻找产能。 光学制造商也在疯狂寻找替代来源。 由于主要IDM已被预订，超大规模云服务商正在向上游移动，直接锁定晶圆、衬底和原料产能，从而在材料层面放大需求。 上海金属市场上6N和7N级铟均处于历史高位；西方市场价格更高，代表了中国与西方市场之间的价格分化。 例如，6N级铟价格约为每公斤412.83-425.46美元。 6N级西方价格约为668.00美元（1月1日，来源：Indium Corporation，5-9公斤数量），溢价超过50%+。（与超大规模云服务商的直接现货谈判价格可能不同）。 7N级的溢价可能更高，衬底享有的溢价最高（无公开数据，但一些估计达到三位数溢价）。 然而，瓶颈现已向上游迁移至衬底本身（以及三阶效应，原料）以锁定产能。 $AXTI在InP瓶颈中或许扮演着最关键且独特的角色，即“InP供应链的40%”（$AXTI CEO ER语录）： 低成本：AXTI的工厂位于北京。他们采购材料（拥有10+家材料供应商）并自行精炼成成品衬底。 高售价：他们将成品InP衬底销往全球市场，那里的价格因短缺和西方铟价格而通胀。 套利：只要AXTI继续获得出口许可证，他们就能捕获中国低制造成本与全球高售价之间的巨大价差。 相比之下，西方公司很大程度上受到上游材料成本和西方溢价的影响。因此，$AXTI下个季度的利润率应会迅速扩张，加上Northland通过1亿美元股份出售注入的生产产能。 这源于12月下旬亚洲金属(Asian Metal)的报道，表明中国供应商明确“提高了对美出口报价”： 2025年12月19日：亚洲金属报道“铟锭供应商提高美国报价”，标志着西方定价特定分化的开始。 2025年12月24日：随后报道“铟锭供应商提高中国报价”，确认到周末定价压力已在国内和出口渠道固化。 这源于超大规模云服务商的产能垄断、交货期延长以及“AI”材料定价与商品基准的脱钩。 InP衬底“大幅涨价”的条件不仅存在，而且正在活跃。 在整个2026年，随着超大规模云服务商向上游锁定需求，InP衬底价格以及由此一步之遥的铟商品价格可能会迅速飙升。 从InP衬底+InP原料价格可能的抛物线式上涨中双向受益的公司将是$AXTI，在$16股价、约8.5亿美元市值下呈现非对称上行空间。然而，风险依然存在，因为他们严重依赖出口许可证和美中贸易关系。 TLDR，AXT： 1. 位于InP衬底和InP原料瓶颈的中心 2. 受益于分配战期间中国生产与美国超大规模云服务商溢价之间的利润率套利和衬底价格飙升。 3. 控制着西方超大规模云服务商AI建设所需衬底的主要全球现货产出份额。 超大规模云服务商看到英伟达垄断了EML产能，现在可能正在查看像$AXTI、住友或原料（Vital, $AXTI, $DOWA）这样的公司以确保其路线图。 AXT是8.5亿美元市值下最有趣的公司之一，因为它控制着超大规模云服务商AI建设的关键瓶颈，并捕获InP衬底价格飙升+利润率套利机会。
英文原文
The InP substrate shortage and "Spread" Arbitrage - $AXTI Market research is showing InP and optical components to become analogous to the HBM shortage and price craze. The initial bottleneck was caused by $NVDA monopolizing of the production capacity of EML from $COHR to $LITE (TrendForce). Here's the breakdown: $NVDA has reportedly engaged in strategic capacity reservations for EML laser chips, pushing lead times for these specific components beyond 2027. By booking out the finished goods capacity of IDMs like $LITE and $COHR (who are constrained by InP), Nvidia has effectively removed the capacity for optical engines. The "Second Order effect": As a result, many hyperscalers have likely delayed their roadmaps/ramp from $MSFT to $AMZN and are looking for capacity. Optical makers have also scrambled for alternative sources. Since the primary IDMs are booked, hyperscalers are moving upstream to secure wafers, substrate, and feedstock capacity directly, creating a effect that is amplifying demand at the materials level. Both 6N and 7N Indium are at all time highs on the Shanghai Metal Markets; prices sit higher on Western markets, representing a price bifurcation between China and Western markets. 6N for example sits ~$412.83-$425.46 per kg. 6N Western prices are ~$668.00 (Jan 1st, source: Indium Corporation at 5-9 kg quantities), over 50%+ premium. (Spot direct negotiations to hyperscalers would likely have different prices). Premiums for 7N would likely higher margin, with substrates commanding the most premium (not public data, but some estimates go into triple digit premiums) However, the bottleneck has now migrated upstream to substrate itself (and third order effect, feedstock) to secure capacity. $AXTI perhaps plays one of the most critical roles and unique roles as "40% of the InP supply chain" ( $AXTI CEO ER quote) in the InP bottleneck: Low costs: AXTI's factory is in Beijing. They source materials (own 10+ materials suppliers) and refine them themselves into finish substrates. Sell High: They sell finished InP substrates to the global market, where prices are inflated by the shortage and the Western Indium price. Arbitrage: As long as AXTI continues their export permits, they capture the massive spread between low Chinese manufacturing costs and high global selling prices. This is compared to Western companies that are largely affected by upstream material costs and Western premiums. As a result $AXTI, margins next quarter should expand rapidly on top of their production capacity funded by the $100m share sale injection by Northland. This is implicated from reports from late December reports from Asian Metal, indicating that Chinese suppliers explicitly "raised offering prices" for exports to the US: December 19, 2025: Asian Metal reported that "Indium ingot suppliers raise offering prices in United States," marking the beginning of the specific divergence in Western pricing. December 24, 2025: They subsequently reported that "Suppliers of indium ingot raise quotes in China," confirming that the pricing pressure had solidified across both domestic and export channels by the end of the week. This stems from capacity monopolization by hyperscalers, extended lead times, and a decoupling of "AI" material pricing from commodity baselines. The conditions for a "Massive Price Increase" for InP substrates are not just present, they are active. Throughout 2026, both the price of InP substrates, and by one-hop, the commodity price of Indium would likely see a rapid surge as hyperscalers go upstream to secure demand. The company that benefits both ways from the a possible parabolic rally in InP substrate + InP feedstock prices would be $AXTI, presenting an asymmetrical upside at $16, ~$850M MC. However, the risks are present as they're heavily reliant on export permits and US-China trade relations. TLDR, AXT: 1. Sits in the center of both InP substrate and InP feedstock bottlenecks 2. Benefits from margin arbitrage and substrate price surges between China production and US hyperscaler premiums during allocation battles. 3. Controls a major a major percentage of the world's merchant output for substrates required for the Western hyperscaler AI buildout. Hyperscalers saw Nvidia monopolized EML capacity and are likely looking at companies like $AXTI, Sumitomo or feedstock (Vital, $AXTI, $DOWA) to secure their roadmaps now. AXT is one of the most interesting companies at $850m given it controls the critical bottleneck to the hyperscaler AI buildout and captures both the InP substrate price surge + margin arbitrage opportunities.
-
2026年1月1日美股评级：推荐SMCI、INTC等复苏及AI基建标的，回避高估量子及零售股。
欢迎来到2026年。1月1日评级： 强烈买入： $TTD $SMCI $AIRO $INTC $HIMS $AXTI $TSM $NBIS $CIFR 三星电子 (KRX: 005930) $HUT $IREN $WULF $GLXY $TSSI $META $ETOR $CRCL 买入： $KRKNF $ONDS $GEMI $NVDA $MU $AMKR SK海力士 $SNAP $RDDT $AAOI $COHR $FISV $FLY $DJT $LITE $AMZN $MRVL $AVGO $OSS $BULL $ORCL $CRDO $ALAB 回避： $RGTI $QBTS $RGTI $BMNR $ETH $PLTR $WMT _ 简而言之的想法： TTD - 估值完全重置，年初至今下跌67%，叠加年底税务抛售。进入2026年是极佳的复苏标的。 SMCI - 仅因推迟一个季度以符合新Blackwell规格而交易得像一家困境公司。前瞻收入同比增长50%，市销率(P/S)接近0.5。税务收割后的极佳复苏标的。 AIRO - 资产负债表约1/6为现金。随着政府加速投资，无人机领域备受追捧。另一只被抛售的IPO新股。进入2026年，尤其在热门细分领域，是极佳的复苏标的。市销率约3.8倍，相比ONDS的25-30倍市销率，当然其教育等业务对利润率计算有很大干扰。 INTC - 它实际上已成为美国政府的半导体臂膀。超大规模云厂商很可能被激励（强烈施压）在有机会时优先使用Intel而非TSM、三星等。我不会赌美国政府会输。 HIMS - 进入2026年前大幅抛售。从70多美元下跌。销售/流量下降，但Zava收购/增长应在2026年带来巨大顺风。尤其是有数亿美元回购，前两个月是强劲的复苏标的。 AXTI - 之前发过相关论点。CEO称“40%的磷化铟(InP)供应链”，InP将是2026-2027年超大规模AI建设中的巨大瓶颈，直到2028年有足够时间通过工程手段绕过它。 TSM - 我引用过很多次。利润率提升。需求最大化。未来几年极好的复利增长者。 三星电子 - 受益于代工/存储。在所有顺风助力业务方面简直是金蛋。 NBIS - 极强买入，$7-90亿年度经常性收入(ARR)，它实际上是5家不同公司，每年三位数增长。管理层引用20-30%息税前利润(EBIT)率，这只是等待游戏。 CIFR, HUT, IREN, WULF, GLXY - 整个数据中心板块在Oracle/OpenAI担忧后极度超卖。OpenAI最近融资$400亿，$AMZN再投$100亿等。因此关于资本支出(Capex)的担忧已大幅降低风险。这也是多方面的，例如比特币下跌影响$CIFR资产负债表，$GLXY在加密领域。但总体而言，新云(neoclouds)板块有巨大的复苏/扩张潜力。 TSSI - 类似SMCI。递延收入=核弹级利空。税务收割后应恢复，且大量收入将被确认。 META - 财报后因一次性税务问题遭遇巨大算法抛售。他们还削减了现实实验室等部门的资本支出/运营支出，这应为2026年的每股收益(EPS)带来巨大顺风。 ETOR - 市值$28亿，坐拥$12亿现金，仍保持双位数同比增长。$1.5亿回购应是不错的顺风，且年初至今表现带来的税务收割效应应会消退。 CRCL - 稳定币论点在2026年应非常稳固。 买入 KRKNF - Anduril合作伙伴+规模。2026年可能转板，国防支出带来大量顺风。 ONDS - 收入增长极具爆炸性，到处都有新的$1000万合同。大量现金余额资助研发。市销率很高，但该领域的投机性领导者如RKLB有估值溢价。 GEMI - 通常我不喜欢交易所，但Gemini从$30+ IPO跌至$10以下。极佳的复苏标的。 NVDA - 巨大的积压订单。大家都知道Nvidia的牛市逻辑。 MU - 存储火热。 SK海力士 - 存储火热。 AMKR - 受益于“美国制造”芯片生产扩张。 SNAP - 存储运营支出削减，存储货币化增加收入，Perplexity贡献$4亿。季度收入$15亿。如果他们将所有这些转化为$10亿+自由现金流(FCF)/年，即使收入完全停止增长，也会完全重估Snap。 RDDT - 说实话，未来10+年不会消失。它是社交媒体界的Robinhood，通过新的收入货币化方式极速增长，且极其盈利。 AAOI - Amazon, MSFT ASIC规模化的互连标的。 COHR - 受益于下一代ASIC的光子学部署。 FISV - 财报后抛售过多，税务收割后是强劲的复苏标的。 FLY - SpaceX IPO带动太空领域火热。税务收割结束后应表现良好，且即将有Northrop的中期催化剂。 DJT - 我从没想过会把它放在这里，但这只是因为他们的TAE合并。 LITE - Google TPU部署带来巨大的物料清单(BOM)，估值有吸引力。Google TPU修正预估后轻微抛售，但它基本存在于每个超大规模ASIC部署中。 AMZN - 七大科技巨头中估值不过高的一个。 MRVL - 分析师误导导致抛售，进入2026年强烈买入。尤其是MSFT Maia收入翻倍时，Marvell当前收入也将随之增长。 AVGO - 像NVDA一样，随着AI基础设施部署加速，强劲的长期持有标的。 OSS - 我曾发帖推测他们是Anduril的供应商之一。但无论如何，边缘计算在2026年将很火热，其1.8亿市值呈现有吸引力的上行空间。 BULL - 类似Robinhood，拥有巨大用户群，但他们只需找出货币化方法。 Oracle - 我认为抛售过多。几个月前我将其列入回避，但从$330跌至$190后，尤其在OpenAI再融资$400亿后，再次具有吸引力。 CRDO - 极高利润率，数据中心部署所需的必要连接。 ALAB - 极高利润率，数据中心部署所需的必要连接。 回避： 有很多在“高估名单”上的东西，比如$RKLB，我喜欢但除了这些之外我不会说回避它们。 RGTI, QBTS, RGTI - 量子名称仍然高估，且可能在未来几年无法交付自由现金流。 BMNR, ETH - 如果你看过我的ETH帖子，我不太看多，因为每天ETH燃烧量只是个位数到低双位数，这简直是笑话。 PLTR - 最被高估的AI名称之一。 WMT - 这怎么是40倍市盈率？这是沃尔玛？ __ （这些基于今日价格） 简而言之： Circle, Etoro, AIRO, Klarna, Figma等IPO名称在下跌+税务收割后进入2026年呈现有吸引力的上行空间。 大量如SMCI, HIMS等过去3个月下跌约40%的名称，在税务收割+一月效应后是极好的波段/复苏交易标的。 许多如FiserV或The Trade Desk等暴跌的名称在税务收割后也是良好的复苏交易标的。 许多数据中心股票如nebius, iren, cifr, wulf, galaxy也是极好的复苏交易标的。 许多其他领域如存储、瓶颈、光子学等在2026年只是极好的长期持有标的，尽管各自都触及历史新高。 仍有相当多高估的名称，从量子到某些太空股票（如planet或rocketlab），特定AI名称如Palantir到零售股票如沃尔玛，我可能会暂时回避，直到有轻微回调。 这只是一个简而言之，如果我仅做短期交易（非长期），但欢迎提问。
英文原文
Welcome to 2026. Jan 1st ratings: Strong Buy: $TTD $SMCI $AIRO $INTC $HIMS $AXTI $TSM $NBIS $CIFR Samsung Electronics (KRX: 005930) $HUT $IREN $WULF $GLXY $TSSI $META $ETOR $CRCL Buy: $KRKNF $ONDS $GEMI $NVDA $MU $AMKR SK Hynix $SNAP $RDDT $AAOI $COHR $FISV $FLY $DJT $LITE $AMZN $MRVL $AVGO $OSS $BULL $ORCL $CRDO $ALAB Avoid: $RGTI $QBTS $RGTI $BMNR $ETH $PLTR $WMT _ TLDR thoughts: TTD - Complete valuation reset dropping 67% YTD, compounded by EOY tax sell-off. Great recovery play going into 2026. SMCI - Trades like distressed company just because they delayed revenue by 1 quarter for new blackwell specs. Forward revenue is increasing 50% Y/Y, P/S close to .5 now. Great recovery play from tax harvesting. AIRO - Roughly ~1/6th balance sheet was cash. Everyone seems to be into drones, especially with accelerated gov inevstments. Another IPO name that got sold off. Great recovery play going into 2026 with esp. hot segment. Roughly ~3.8x P/S compared to ONDS trading at 25-30 P/S, but obviously there's quite a lot of other businesses like their education sector which messed up margin calculations quite a bit. INTC - It's literally become the semi arm of the US government. Hyperscalers will likely be incentived (strongly pressured) to use Intel whatever chance it gets over TSM, Samsung, etc. I would not bet against the US government. HIMS - Huge selloff going into 2026. Down from $70's. Sales/Traffic is down, but Zava acquisition/growth should add a huge tailwind going into 2026. Esp. with few hundred mill buybacks, strong recovery play first two monts in. AXTI - Posted thesis on this earlier. CEO - "40% of Inp supply chain", InP will be a huge, huge bottleneck for hyperscaler AI buildout 2026-2027 until there's enough time to engineer around it in 2028. TSM - I've covered this quote a lot. Increasing margins. Maxed out demand. Just extremely good compounder next few years. Samsung Electronics - benefits from foundry/memory. just golden egg regarding all the tailwinds helping the buisness. NBIS - Extremely strong buy, $7-9B ARR, it's literally 5 different companies growing triple digits Y/Y. management quoted 20-30% EBIT margins, it's just a waiting agme. CIFR, HUT, IREN, WULF, GLXY - Whole datacenter space is extremely sold off after Oracle/OpenAI fears. OpenAI recently raised $40B, another $10B from $AMZN, and more. So a lot of fears regarding capex spend has been de-risked. It's multifaceted too, eg. Bitcoin drop, affects $CIFR balance sheet, $GLXY in crypto space. But generally huge recovery play/ramp for neoclouds sector. TSSI - Similar to SMCI. deferred revenue = nuke. Should recover after tax harvesting + lot of revenue gets recognized META - Huge algorithmic selloff post earnings due to one-time tax. They also cut capex/opex spend of their reality labs and other departments and this should be a huge tailwind for EPS going into 2026. ETOR - Literally sitting on $1.2B with a $2.8B marketcap and growing double digits Y/Y still. $150M buyback should be a nice tailwind, and tax harvesting from YTD performance should subside. CRCL - Same as stablecoin thesis should be really solid going into 2026 Buy KRKNF - Anduril partner+ scale. Probable uplisting in 2026, lot of tailwinds from defense spending. ONDS -pretty explosive revenue growth, new $10m contracts left and right. large cash balance to fund r&d. Pretty high p/s but there's valuation premiums for speculative leaders in the space like rklb. GEMI - So i typically dont like exchanges, but gemini got nuked from $30+ IPO sub $10. pretty solid recovery play. NVDA - Huge backlog lol. Everyone knows bull case for nvidia MU - Memory is hot SK Hynix - Memory is hot AMKR - benefits from "made in america" chip expansion in prod. SNAP - Opex Cut from memory, increase revenue from memory monetization, $400m from perplixity. $1.5B revenue/quarter. They could literally stop growing revenue complelty if they convert all of that to $1B+ FCF/year, it would re-rate snap completly. RDDT - This is not going anywhere for the next 10+ years tbh, it's like robinhood of social media, growing extremely fast from new ways to monetize revenue, and just extrmeely profitable. AAOI - interconnect play for amzn, msft asic scale up. COHR - benefits from photonics rollout for next gen asics. FISV - Nuked a bit too much post ER, strong recovery play esp. post tax-harves.t FLY - Space is hot from SpaceX IPO. Should do well given tax harvesting is over, and they have medium lift coming up with northrop. DJT - I never thought i'd put this here lol, but this is just because of their TAE merger. LITE - Large BOM from Google TPU rollout, attractive valuation. Slight selloff after Google TPU revised est. but it's basically in every single hyperscaler asic deployment. AMZN - one of the mag7 that's not overvalued MRVL - Selloff from analyst misinformation, strong buy going into 2026. Especially with msft maia revenue doubling Marvell's current revenue when it ramps up AVGO - Like NVDA just strong long, as AI infrastructure deployment ramps up OSS - I made a post speculating that they're one of andruils' suppliers. but regardless, edge computing will be hot 2026 and its 180m mc presents attracctive upside. BULL - similar to robinhood where they have a huge userbase, but they just need to figure out monetization Oracle - Sold off a bit too much imo. I put this on avoid months ago but after the from from $330 to $190, it's more attractive again esp. after openai raised another $40B CRDO -extremely high margin, necessary connectivity for dc rollout ALAB - extremely high margin, necessary connectivity for dc rollout Avoid: There's a lot of stuff on the "overvalued list" like $RKLB that i like but I wouldn't quite say avoid it either aside from these. RGTI , QBTS, RGTI - Quantum names are still overvalued and likely won't deliver fcf in the next few ytears. BMNR, ETH - if you saw my eth post, not exactly bullish since the amount of ETH burn is just single-low double digits every day, which is a joke. PLTR - one of the most overvalued ai names WMT - How is this 40 p/e? This is Walmart? __ (these are based on today's prices) TLDR: IPO names like Circle, Etoro, AIRO, Klarna, Figma, present attractive upsides post drop + tax harvesting going into 2026. Tons of names like SMCI, HIMS that dropped 40% or so past 3 months, are amazing swing/recovery trades post-tax harvest + Jan effect. Lot of the names that doom dropped like FiserV or The Trade Desk present good recovery trades too post-tax harvest. Many datacenter stocks like nebius, iren, cifr, wulf, galaxy, are amazing recovery trades too. Lot of other segments like memory, bottlenecks, photonics, and others are just great longs in 2026, despite each hitting ATHs. There's still quite a lot of overvalued names from Quantum, to certain Space stocks (eg. planet or rocketlab), specific AI names like Palantir to retail stocks like Walmart that I would probably avoid for the time being until there's a slight correction. This was a TLDR just if I'm short term trading-only (not long term) but feel free to ask questions.
-
难以直接投资机器人板块，因多为巨头部门或涉及地缘政治风险。
很难直接投资有趣的机器人公司，因为它们只是 $AMZN、$TSLA、现代 KRX: 005380（波士顿动力）以及你们那些巨头公司的部门。 其他有趣的可能像 $XPEV 这样是外国公司（但我不投资那些推进地缘政治对手军事相关 AI/机器人项目的公司）。 有些像 $RR 这样的公司，但它们远不是我正在关注的垂直领域类型。
英文原文
It's hard to get direct exposure to interesting robotics companies since they're just divisions in $AMZN, $TSLA, Hyundai KRX: 005380 (Boston Dynamics), and your massive companies. Other interesting ones might be foreign like $XPEV (but I don't invest in companies that advance geopolitical enemy AI/Robotic programs that relate to military). There's some like $RR, but they're nowhere close to the types of verticals I'm looking at.
-
2026年十大主题投资：聚焦AI供应链瓶颈、软体机器人及支付颠覆。
2026年通讯。 主题投资：演进、颠覆与瓶颈 1. 软体机器人 - 向 $TSLA、$ONDS、波士顿动力演进。 2. 硅光子(SiPh) - 磷化铟(InP)瓶颈 | $AXTI、$LITE、$GOOGL 3. 玻璃基板 - 瓶颈 | $NVDA、$INTC、$TSM 4. 资金流动 - 对 $V、Stripe、$BOA 的颠覆 5. AI云层级 - 瓶颈 | $NBIS、$IREN、$HUT 6. LLM网络安全 - 向 $CRWD、$CSCO、$MSFT 演进 7. 低轨(LEO)太空基础设施 | 向 $RKLB、SpaceX、$ASTS 演进 8. 消费者代理工作流(50步) - 对消费者劳动力的颠覆，来自Manus、$PATH Cognition 9. 分布式计算延迟 - 瓶颈 | $TSLA、$AMZN、$GOOGL 10. 铜互连寿命延长 - 瓶颈 | $NVDA (LPU/Groq)、$AMD、$INTC _这是我对从公开信息综合及瓶颈二/三阶效应来看最感兴趣的主题投资的简要概述！_ 1. 软体机器人：向机器人的演进 传统机器人（Optimus、波士顿动力）依赖逆运动学控制刚性关节。软体机器人改变了数学模型。 我们已到达硬件（Optimus、波士顿动力、Figure）与LLM（Gemini、Grok、Opus）相遇的节点，正处于大规模商业化的开端。 通过使用受章鱼触手和人类皮肤启发的材料，机器人正从齿轮转向流体性，以处理极其精细的任务，如像人手一样处理农产品，或为 $ONDS/Andruil 无人机添加类章鱼延伸以拾取极重表面。 这种演进在于跳出思维定势思考机器人能做什么。我记得7年前曾与该领域的斯坦福博士合作，AI在多年研究后开始商业化，因此该领域也是如此。 将类生物流体性添加到刚性机器人中的可能性是无限的，这只是自然演进。 大多数可能是私人公司。 2. 硅光子 - AI基础设施的瓶颈“磷化铟(InP)卡脖子点” 从Blackwell Ultra集群到Google TPU已触及上限，需要光子互连 | 共封装光学(OCS)以实现扩展。 基板：$AXTI（通过Tongmei）和住友（日本）控制全球约60-70%的InP基板市场。 材料：Vital Materials（中国）和AXT等公司控制原材料铟本身的精炼（78%+的供应链）。 如果你是美国科技巨头，你2026年的整个“AI增长故事”取决于由地缘政治对手控制的材料。 唯一可扩展的解决方案是工程绕行，要么实现芯片上光传输，同时减少90%的铟使用，要么使用微小的磷化铟薄片代替大型昂贵晶圆。 瓶颈本身有机会，如AXT、住友。或帮助解决它的公司如 $POET。 3. 玻璃基板 - 解决从 $NVDA 到其他公司的共封装光学(CPO)瓶颈 向玻璃基板的转变本质上是半导体行业对当前材料物理极限的回答。 当前芯片位于有机材料（本质上是专用塑料）制成的基板上。随着芯片变大，如Nvidia巨大的GPU封装，塑料基板会翘曲。 因此，玻璃基板正成为共封装光学(CPO)的行业标准，因为它们解决了光子学中最大的对齐问题。 美国政府已视其为必要，我们看到巨额补贴流向这些公司。 $INTC、三星电子、Absolics（SKC子公司）、DNP等是主要受益者，尤其是随着MRVL和 $AVGO（推动光学开关的玻璃）推进CPO革命。 4. 资金流动 - 对卡网络、银行、交易所和支付的颠覆 几十年来，资金转移一直是“收费公路”业务。每次刷卡，2%到3%的钱流入卡网络（Visa/Mastercard）和发卡行的口袋。 或者从交易所买卖加密货币是0.2-1%。这是历史上最有利可图、“不可杀死”的商业模式。 直到现在。2025年的“天才法案”刚刚将金钱传输许可证或银行特许状交给像 $XRP 这样的公司，赋予了他们王国钥匙。 对我来说并非理论。我恰好正在自己的初创公司与创建V / $PYPL 实时支付网络的人一起从事这项工作。 但基本上，拥有现有MTL或追求银行特许状并利用天才法案及其他技术的公司，现在可以通过在美联储和区块链之上进行结算来绕过传统百分比费用，有效地将基于百分比的费用转化为几美分。 99%的公司会这样做吗？可能不会，因为支付行业的所有利润率都将归零。但我乐意看到。 但基本上，Stripe以11亿美元收购Bridge本应是对现有公司的红色警报，表明1天ACH、 interchange模式、25美元国际转账的日子即将结束。 这扩展到许多其他相邻领域，从低费用颠覆者如 $HOOD、Mercury，一直到稳定币新银行，或制作自己稳定币的公司如 $SOFI。 5. AI云层级 - 超大规模计算瓶颈的解决方案 当超大规模云厂商被困在3-5年的电网互连队列中时，像WULF和IREN这样的矿工今天就拥有即插即用的GW级算力。 这是千载难逢的机会，超大规模云厂商将其现金牛云收入流向小公司。 这里有不同的层级，从Fluidstack、Poolside、Fireworks在GPU编排层，到IREN等公司构建的裸金属层。 然后有成为超大规模云厂商本身，如NBIS拥有物理位置、GPU、软件编排，然后为推理提供简单接口。 这是少数小公司在未来一两年成为AWS或Azure，或被收购（例如GOOGL以47亿美元收购Intersect）的机会。 像NBIS、IREN、CRWV这样的新云，以及像CIFR、WULF、HUT这样的colo玩家（以及私人部门->能源）将受益。 6. LLM网络安全 - 向现代安全和漏洞防御的演进 最近的报告（例如来自Anthropic红队）显示，高级模型如Opus（及未来版本）可以自主扫描开源智能合约，并在几分钟内识别价值数百万美元的“零日”漏洞。 含义：如果AI能在不可变的区块链合约中找到逻辑缺陷，它也能在银行的SWIFT API或电网控制软件中找到缺陷。 同样适用于KYC/AML。像Gemini Nano Banana这样的模型能够创建逼真的图像/视频，人们能够绕过许多程序。 这个领域有很多不性感但具Alpha潜力的事情，如LLM自动化SOC2/PCI DSS合规，代理坐在服务器上，持续监控日志，并自动生成审计所需的证据。 7. 低轨(LEO)太空基础设施 | 向拓展最后疆域的演进 太空是下一个大事情。这并不新鲜。（希望你懂这个笑话）。但从像 $RKLB、SpaceX这样的公司，到解决轨道拥堵或发射节奏瓶颈的公司，再到像ASTS或Starlink这样商业化基础设施的公司，在未来一年呈现许多机会。 因此，像Impulse、Blue Origin、$ASOZF到RKLB、$ASTS这样的公司将受益于整个链条。 8. 消费者代理工作流(50步) - 对消费者劳动力的颠覆，来自Manus、PATH Cognition 这一点很简单，无需解释。但在对就业+成本节约的潜在影响上显而易见。 你如何自动化商务拓展？如何自动化营销？如何自动化软件工程师？ 这超越了ChatGPT的几步回答，直接进入现实世界，AI代理可以在X上漫游，找到合适的人，发送DM，继续对话，并在一个工作流中导致销售电话。 这是“聊天机器人”时代的结束和“行动”时代的开始，取代公司以前需要的所有人。 我尚未看到任何公司大规模做到这一点。拥有这些的公共公司如META并没有呈现最佳敞口。也许是 $PATH 在公共空间。 9. 分布式计算延迟 - 解决AI计算容量紧张的瓶颈 像GOOGL Cloud、MSFT Azure这样的超大规模云厂商已达最大容量。 Elon Musk已经提出分布式计算作为解决此问题的未来（例如，拥有 $TSLA 网络为LLM推理提供计算）。 “Tesla计算云”论点很迷人，但我识别出的最大物理障碍是：推理延迟。 要生成“Token B”，模型必须先生成“Token A”。它不能同时做两者。如果你将一个巨大模型（如Grok-3）拆分到5辆不同的汽车中以适应内存，你必须为每个生成的Token在这些汽车之间发送数据。 因此，如果汽车之间的网络延迟甚至是20ms（5G的乐观估计），而你生成50个Token，你刚刚在计算时间之上添加了1秒的纯“等待时间”（延迟）。在使用NVLink的数据中心中，该等待时间以纳秒计。 同样适用于零售用户拥有的任何备用计算机、GPU等。有数十亿消费级GPU（Teslas、iPhones、游戏PC）90%的时间闲置。 解决推理的“分布式延迟”问题呈现了计算史上最大的套利机会之一。 尚未看到任何公司大规模完成此任务。也许NVIDIA Dynamo、$AKAM、TSLA正在接近。 10. 铜互连寿命延长 - 解决Nvidia和其他公司的瓶颈 既然我们不能拥有无限的InP，我们必须用现有的东西（例如铜）进行工程绕行，所以铜电缆可以做物理上说它不应该做的事，如在不损失信号的情况下跨机架传输224G信号。 行业在InP上遇到了硬性停止，美国在物理上无法开采和精炼足够的InP将数据中心中的每个链接变成光纤。 如果有任何帮助，那就是好事。例如，NVDA对Groq团队和IP的200亿美元“收购雇佣”。LPU更多是关于推理延迟/架构，但它作为副产品解决了铜寿命延长。Groq的整个架构在延迟上击败了Nvidia，因为它拒绝了光学。Groq使用“确定性”网格，依赖芯片之间的直接电气（铜）连接，避免光学开关的“抖动”和转换时间。 像 $ALAB、$CRDO、Groq，或任何能找到用铜绕过光学瓶颈方法的公司将是赢家。 _这里有从私人部门投资到公共部门的众多交易！只是今天即兴写下了我的想法，但乐意稍后详细阐述。 无论如何，我相信这些主题投资中的许多： 从投资InQ瓶颈绕行（$POET）或瓶颈本身（$AXTI）到公共部门的颠覆者（$CRCL）。 到投资铜扩展瓶颈修复（Groq）、银行特许状颠覆者（Mercury）到私人部门的演进公司（Lightmatter、Festo）。 在2026年呈现不对称上行空间。 新年快乐！
英文原文
2026 Newsletter. Thematic Investments: Evolution, Disruption, and Bottlenecks 1. Soft Robotics - Evolution to $TSLA, $ONDS, Boston Dynamics. 2. SiPh - InP Bottleneck | $AXTI, $LITE, $GOOGL 3. Glass Substrates - Bottleneck | $NVDA, $INTC, $TSM 4. Money Movement - Disruption to $V, Stripe, $BOA 5. AI Cloud Layers - Bottleneck | $NBIS, $IREN, $HUT. 6. LLM Cybersecuirty - Evolution to $CRWD, $CSCO, $MSFT 7. LEO Space Infrastructure | Evolution to $RKLB, SpaceX, $ASTS 8. Consumer Agentic Workflows (50 Step) - Disruption to the Consumer Workforce, from Manus, $PATH Cognition 9. Distributed Computing Latency - Bottleneck | $TSLA, $AMZN, $GOOGL, 10. Copper Interconnect Life Extension - Bottleneck | $NVDA (LPU/Groq), $AMD, $INTC _ This is an light overview of thematic investments I find the most interesting from a public-information synthesis perspective + second/third-order effects from bottlenecks! _ 1. Soft Robotics: The Evolution to Robotics Traditional robotics (Optimus, Boston Dynamics) relies on Inverse Kinematics to rigid joints. Soft robotics changes the math. We've met the point where hardware (Optimus, Boston Dynamics, Figure) met LLMs (Gemini, Grok, Opus), and we're at the beginning of possible widespread commercialization. By using materials inspired by octopus tentacles and human skin, robots are moving away from gears and toward fluidity to handle extremely delicate tasks like handling produce like the human hand, to picking up extremely heavy surfaces adding Octopus-like extensions to $ONDS/Andruil Drones. The evolution is thinking outside the box in terms of what robotics can do. I remember working with some Stanford PHds in this field like 7 years ago, and it just so happens AI is starting to be commercialized after many years of research. So expected, this field to be as well. Possibilities are limitless adding organism-like fluidity to rigid robotics, this is just the natural evolution. Most of these are prob private companies. _ 2. Silicon Photonics - Bottleneck of the AI Infrastructure "InP Chokepoint" Blackwell Ultra Clusters to Google TPUs have hit the upper wall and requires photonics for interconnects | OCS to scale up. The Substrates: $AXTI (via Tongmei) and Sumitomo (Japan) control roughly 60-70% of the world's InP substrate market. The Materials: Companies like Vital Materials (China) and AXT control the refining of the raw Indium itself (78%+ of supply chain). If you are a US tech giant, your entire "AI Growth Story" for 2026 depends on materials controlled by geopolitical rivals. The only scalable solution is engineering around it, either by delivering light-on-chip, while using 90% less InP or companies that use tiny slivers of Indium Phosphide instead of large, expensive wafers. There's opportunities with the bottleneck itself like AXT, Sumitomo. Or companies that help address it like $POET. _ 3. Glass Substrates - Fixing the Bottleneck for CPOs from $NVDA to others. The shift toward glass substrates is essentially the semiconductor industry’s answer to a physical wall they are hitting with current materials. Current chips sit on a substrate made of organic materials (essentially specialized plastic). As chips get larger, like Nvidia's massive GPU packages, plastic substrates warps. So, glass substrates is becoming the industry standard for Co-Packaged Optics (CPO) because they solve the single biggest problem in photonics with alignment. US Government already sees this as a necessity and we've seen huge subsidies funneling down to some of these companies. Companies like $INTC, Samsung Electronics, Absolics (SKC Subsidiary), DNP, and others are the main beneficiaries, especially as MRVL and $AVGO (driving glass for optical switches) move forward with CPO revolution. _ 4. Money Movement - The Disruption to Card Networks, Banking, Exchange, and Payments For decades, moving money has been a "toll road" business. Every time you swiped a card, 2% to 3% of that money vanished into the pockets of the Card Networks (Visa/Mastercard) and Issuing Banks. Or buying/selling crypto from an exchange would be .2-1%. It was the most profitable, "un-killable" business model in history. Until now. The "Genius Act" of 2025 just handed companies like $XRP with Money Transmitter Licenses or Banking Charters the keys to the kingdom. Not really theoretical for me. I happen to be working on this myself at my own startup with some folks who created V / $PYPL's real-time payment networks. But basically companies with existing MTLs or pursuing banking charters leveraging the Genius Act and some other tech can now bypass legacy % fees by doing settlement on top of the Federal Reserve and blockchains, effectively converting percentage-based fees into a few cents. Would 99% companies do it? Probably not since every single margin from across the payment industry would just go to 0. I'd be happy though. But basically Bridge's $1.1B acquisition by Stripe should have been a red-alarm to existing companies that days of 1-Day ACH, interchange models, $25 international transfers, are soon to be over. This extends to many other adjacents from low fee disruptions like $HOOD, Mercury all the way to Stablecoin Neobanks, or companies making their own stablecoins like $SOFI. _ 5. AI Cloud Layers - The Solution to HyperScaler compute Bottleneck While Hyperscalers are stuck in 3-5 year grid interconnection queues, miners like WULF and IREN are sitting on plug-ready GWs today This is the opportunity of a lifetime as hyperscaler funnel their cash cow Cloud revenues down to tiny companies. There's many different layers to this from Fluidstack, Poolside, Fireworks on the GPU orchestration layer, to the bare metal layer that companies like IREN are building. Then there's becoming the hyperscaler themselves like NBIS owning the physical locations, the GPU, software orchestration, and then providing simple interfaces for inference. This is the opportunity for a few small companies to become Amazon Web Service or Microsoft Azure over the next year or two, or get acquired (eg. GOOGL buying Intersect for $4.7B) Neoclouds like NBIS, IREN, CRWV, down to colo plays like CIFR, WULF, HUT (and private sectors -> Energy) stand to benefit. _ 6. LLM Cybersecurity - The Evolution to Modern Security and Vulnerability Defense Recent reports (e.g., from Anthropic's Red Team) showed that advanced models like Opus (and future iterations) could autonomously scan open-source smart contracts and identify "Zero-Day" exploits worth millions of dollars in minutes. The Implication: If an AI can find a logic flaw in a immutable Blockchain contract, it can find a flaw in a bank's SWIFT API or a power grid's control software. Same with KYC/AML. Models like Gemini Nano Banana are able to create realistic images/videos of people and people are able to get past a lot of programs. There's tons of things as an unsexy alpha in this field like LLMs automating away SOC2/PCI dss compliance to agents sitting on a server, continuously monitor logs, and auto-generate the evidence needed for auditors. 7. LEO Space Infrastructure | The Evolution to Expanding into the final frontier. Space is the next big thing. This is not anything new. (hope you got the joke). But anywhere from companies like $RKLB, SpaceX. Companies that fix orbital congestion or launch cadence bottlenecks. To companies that commercialize the infrastructure like ASTS or Starlink present many opportunities over the next year. So companies like Impulse, Blue Origin, $ASOZF to RKLB, $ASTS stand to benefit across the entire chain. 8. Consumer Agentic Workflows (50 Step) - Disruption to the Consumer Workforce, from Manus, PATH Cognition This one is simple and needs no explanation. But largely obvious in potential impact on employment + cost saving. How do you automate away business development? How do you automate away marketing? How do you automate away software engineers? This is going past few step ChatGPT answers and directly in to the real world where an AI agent can roam X, find the right people, DM someone, continue conversations, and lead to a sales call in just one workflow. This is the end of the "Chatbot" era and the beginning of the "Action" era replacing everyone previously required in a company. I haven't quite seen this done at scale yet with any company. Public companies like META that own these, don't really present the best exposure. Maybe $PATH for public space. 9. Distributed Computing Latency - Fixing the Bottleneck for AI Compute Capacity Strains Hyperscalers like GOOGL Cloud, MSFT Azure at max capacity. Elon Musk already floated distributed computing as the future of solving this issue (eg. having networks of $TSLA's providing compute for LLMs for inference). The "Tesla Compute Cloud" thesis is fascinating, but the single biggest physical barrier I've identified is: Inference Latency. Too generate "Token B," the model must first finish generating "Token A." It cannot do both at the same time. If you split a massive model (like Grok-3) across 5 different cars to fit it in memory, you have to send data between those cars for every single token generated. So, if your network latency between cars is even 20ms (optimistic for 5G), and you are generating 50 tokens, you just added 1 full second of pure "waiting time" (latency) on top of the compute time. In a data center using NVLink, that wait time is measured in nanoseconds. Same applies to any spare computer, GPU, and others owned by retail users. And there's billions of consumer GPUs (Teslas, iPhones, Gaming PCs) that sit idle 90% of the time. Solving the "distributed latency" problem for inference presents one of the single greatest arbitrage opportunity in the history of computing. Haven't really seen any companies that accomplished this at scale yet. Maybe NVIDIA Dynamo, $AKAM, TSLA, getting a little closer. 10. Copper Interconnect Life Extension - Addressing the Bottlenecks of Nvidia and Others Since we can't have infinite InP, we have to engineer around it with what we have (eg. Copper), so copper cables can do things that physics said it shouldnt like carrying 224G signals across a rack without signal loss. The industry is hitting a hard stop on InP where, US cannot physically cannot mine and refine enough InP to turn every link in a data center into fiber optics. If anything helps, then it's good. EG. NVDA's $20B "Acqui-hire" of Groq's team and IP. LPU is more about inference latency/architecture but it addresses copper life extension as a byproduct. Groq’s entire architecture beat Nvidia on latency because it rejected optics. Groq uses a "deterministic" mesh that relies on direct electrical (copper) connections between chips, avoiding the "jitter" and conversion time of optical switches. Companies like $ALAB, $CRDO, Groq, or anyone who can find ways to engineer around the optical bottleneck with copper will be a winner. _ There are tons of trades from both private sector investments to public! Just wrote up my thoughts on the fly today, but happy to elaborate later. Regardless I believe a lot of these thematic investments from: Investing in InQ Bottleneck Workarounds ( $POET ) or the bottleneck itself ( $AXTI ) to Disruptors ( $CRCL ) in the public sector. To Investing in copper extension bottleneck fixes (Groq), bank charter disruptors (Mercury) to evolutionary companies (Lightmatter, Festo) in the private sector. Present asymmetrical upside in 2026. Happy New Year!
-
驳斥MRVL被取代谣言，指出物理限制使其无法短期切换，视抛视为买入良机。
终于深入研究了 $MRVL。 我的第一反应：市场和分析员/新闻到底在抽什么？ 来自 Maia 量产的收入几乎是 $MRVL 2025 财年总收入的两倍。 看数据，这令人震惊： 富邦 (Fubon) 预计 Marvell 仅在 2027 年微软 Maia 量产中就能获得 100-120 亿美元的收入。 作为对比，Marvell 2025 财年收入为 57.7 亿美元。 无论如何看，Marvell 看起来都是博通 (Broadcom) 的强劲竞争对手（在收购 Celestial 之后），也是一个极佳的长线标的。 但是... The Information / Benchmark 发布了新闻称： - 微软正在与博通谈判，以在未来几代产品中取代 Marvell。 - 亚马逊 Trainium 3 和 4 的设计流向了台湾竞争对手 Alchip Technologies。 导致股价大幅抛售。 深入研究后，这怎么可能呢？ “Marvell Technology 股价下跌，因为 Benchmark Equity Research 下调评级， citing 失去亚马逊 AI 芯片业务。” “Benchmark 认为 Marvell 输给了 Alchip，尽管 Marvell 最近预测数据中心增长强劲。” - Trainium 3 的生产已锁定 Marvell 的设计。如果不从头设计全新芯片，你无法“切换”到 Alchip。如果要共享范围，当然可以，但措辞简直糟糕透顶。 $MRVL 已经获得了 $AMZN 整个 2026 年的 Trainium 3 订单，亚马逊为什么要取消所有订单然后选择新供应商？ 可能有不同的范围、额外的设计服务等，但所有新闻的框架听起来像是 $MRVL 被取代了。 Benchmark 自己在造成损害后（通过更多“行业讨论”）大幅收回了说法，称 Marvell 并未完全失去 Trainium 3/4，而是亚马逊增加了 Alchip 作为额外设计支持，而 Marvell 仍保留地位。 但股价已经受损。 至于 The Information 关于 $MSFT Maia 输给博通的报道，这是煽动性新闻。 摩根大通 (JP Morgan) 后来直接出来基本说这是假的，“没有 ASIC 项目份额损失”。 在微软交易上，市场完全错过了“立即取代”和“未来几代”讨论之间的细微差别。 物理规律使得从 Maia 到博通的立即切换不可能，除非有惊人的多年延迟。 现在用博通取代 Marvell 的 Maia 300 需要废弃整个 2nm 掩模集，重新设计 I/O 环，并重启 3 年的验证周期。Marvell 特定的 SerDes IP 也嵌入在芯片的 I/O 环中（重定时器和交换机通信）。如果微软目标是在 2027 年量产 Maia 300，芯片必须在 2026 年中后期进入验证阶段。这意味着物理设计今天已经冻结或接近冻结。 今天（2025 年 12 月）与 $AVGO 的新 ASIC 设计将面临以下最短时间线： - 架构与 RTL（12 个月）：逻辑定义和 IP 选择。 - 物理设计到流片（9-12 个月） - 掩模生产与晶圆厂（3-5 个月）：TSMC 2nm 周期。 - 硅后验证（6-9 个月） 总时间：~30-38 个月。 即使加快几个月，这些“与 $AVGO 的讨论”最早也要到 2028 年才能大规模部署。为了赶上 2027 年的量产时间表，微软必须出货 Marvell 的设计。 此外，微软的机架基础设施使用 Marvell 的“Alaska”重定时器和 DSP 在电缆中。 微软必须： - 重新认证机架中每根电缆和背板的信号完整性。 - 拆下主板上的 Marvell 重定时器并替换为博通版本 还要更改他们的定制液冷机架“sidekick”，这是针对 $MRVL Maia 芯片特定版图分布指定的。$AVGO 芯片会有完全不同的物理版图和热密度，他们需要重新设计机架。 所以当市场计入立即取代风险并基于 TMZ 式的讨论八卦谣言交易时，他们完全是在开玩笑，谣言称微软想探索选项。 $MRVL CEO 甚至出面表示他们手中握有明年全年预测的“采购订单”。2026 年的预测可能偏低，因为 $AMZN 和 $MSFT 的量产都面临延迟。 但如果你看长期 Q4 2026 到 2027，仅从他们的订单和 Maia 300 来看：即使仅估计 ~200 亿美元收入带来 7.72 美元每股收益，30 倍市盈率（之前交易在 35-40 倍），如果是两年后等待，股价将从 85 美元涨到 231.60 美元/股。 所以 TLDR：整个“取代”理论被极其完全地误解了，只是噪音。通常我不会为此写整篇文章，但看到 The Information 和其他新闻一次又一次出现，通过煽动性措辞传播错误信息，这简直疯了。 中期切换在物理上是不可能的，市场误解了取代。所以最近的抛售看起来是一个坚实的买入机会。 至于未来的讨论，当然任何公司都希望供应商多元化。如果 Meta 想从博通多源采购，$MRVL 也可以对 $META MTAI 未来几代做同样的事。 如果仅按单个项目 Maia 300 建模并忽略媒体噪音，$MRVL 看起来非常有前景。
英文原文
Finally took a deeper look at $MRVL. My first reaction: What are the markets and analysts/news smoking? Revenue derived from Maia ramp is literally double $MRVL's FY 2025 revenue. If we look at the numbers, it's staggering: Marvell is modeled to take in $10-$12B in revenue (Fubon) in Microsoft Maia ramp in 2027. Alone. To put that in perspective, Marvell's FY 2025 revenue was $5.77B. No matter how I look at it, Marvell looks like a strong Broadcom competitor (after Celestial acqusition) and a terrific long. But... The Information / Benchmark released news that: - Microsoft is negotiating with Broadcom to displace Marvell in future generations. - Amazon Trainium 3 and 4 designs went to Taiwanese competitor Alchip Technologies. causing a massive selloff. After looking into it more, how is this even possible? "Marvell Technology declines after Benchmark Equity Research downgrades the stock, citing a loss of Amazon’s AI chip business." "Benchmark believes Marvell lost Amazon’s Trainium 3 and 4 designs to Alchip, despite Marvell’s recent forecast of strong data-center growth." - Trainium 3 production are locked to Marvell's design. You can't "swap" to Alchip without designing an entirely new chip from scratch. If you want to share scope, sure but the wording is absolutely atrocious. $MRVL already secured Trainium 3 orders from $AMZN throughout 2026, why in the world would Amazon cancel them all and then go with a new vendor? There can be different scope, additional design services, etc. but the framing from all the news sounds like $MRVL is being displaced. Benchmark itself later walked it back materially after causing damage (via more "industry discussions") and said Marvell did not lose Trainium 3/4 outright, but that Amazon added Alchip for additional design support while Marvell retains a position. But the damage has already been done to stock price. As for The Information report on $MSFT Maia loss to Broadcom this is sensational journalism. JP Morgan later literally came out and basically said it's all false, "there has been no ASIC program share loss". On the Microsoft deal, markets completely missed the nuance between "immediate displacement" and "future generation" discussions. The physics of this make an immediate swap impossible from Maia to Broadcom without incredible multiple-year delays. To replace Marvell with Broadcom for the Maia 300 now would require scrapping the entire 2nm mask set, redesigning the I/O ring, and restarting a 3-year validation cycle. Marvell’s specific SerDes Is is also embedded in the chip’s I/O ring (retimers and switches comms). If Microsoft targets a 2027 ramp for Maia 300, the chip must be in validation phase by mid-to-late 2026. This implies the physical design is already frozen or nearing freeze today. A new ASIC design with $AVGO today (Dec 2025) would face this minimum timeline: - Architecture & RTL (12 months): Logic definition and IP selection. - Physical Design to tape ot (9-12 months) - Mask prod & Fab (3-5 months): TSMC 2nm cycle. - Post-Silicon validation (6-9 months) Total Time: ~30–38 months. Even if this is sped up by many months, these "Discussions with $AVGO " not be ready for volume deployment until 2028 at the very earliest. To hit the 2027 ramp timelines, Microsoft must ship the Marvell design. Then there's Microsoft’s rack infrastructure uses Marvell’s "Alaska" retimers and DSPs in the cables. Microsoft would have to: - Re-qualify the signal integrity of every cable and backplane in the rack. - Rip out Marvell retimers from the motherboard and replace them with Broadcom versions As well as change their custom liquid-cooling rack "sidekick", which was specified toward specific distribution of the $MRVL Maia chip’s floorplan. A $AVGO chip would have a completely different physical floorplan and thermal density and they'd need to regineer the rack. So market are completely trolling if when they price in immediate displacement risk and trade on TMZ-style discussion gossip rumors that Microsoft wanted to explore options. $MRVL CEO even went out and said they have "purchase orders in hand" for the entirety of the next year's forecast. 2026 forecasts are probably light because $AMZN and $MSFT both faced delays on their ramp. But if you look at longer term Q4 2026 into 2027, we just from their orders and Maia 300 alone: even just estimating $7.72 EPS from ~$20.0B revenue, 30x P/E (they've traded 35-40 before), would be $231.60/share from $85 if you want to wait 2 years. So TLDR: The whole “replacement” theory is extremely completely misinterpreted and is just noise. Normally, I wouldn't write a whole post on it, but seeing The Information and other news pop up again, and again, spreading misinformation through sensational wording is just crazy. There's a physical impossibility of a mid-cycle swap and market misunderstanding of displacement. So the recent sell-off looks like a solid buying opportunity. As for future discussions, of course any company would want vendor diversification. $MRVL can go do the same with $META MTAI future generations too if Meta wants to multi-source away from Broadcom. If you model just by single project Maia 300 alone and ignore media noise, $MRVL looks incredibly promising.
-
解析AI供应链各环节标的风险收益特征，推荐TSM稳健、AXTI高弹性、NBIS高成长。
感谢关注！我来简单拆解一下：我看好金融科技(Fintech)和人工智能(AI)，这是我的两个主要成长型投资板块。像 $CRCL 这样的稳定币(Stablecoins)属于金融科技领域。至于 AI 领域，我专注于供应链投资： 新云厂商(Neoclouds) -> AI 顶层消费者/超大规模云提供商。$NBIS 是我的最爱。 顺着供应链往下，谁帮助制造 $NBIS 和 $IREN 购买的 GPU？ -> 来自 $NVDA 的 GPU，部分 $AMD，以及未来可能来自 $GOOGL、$MSFT、$AMZN 的 ASIC。 谁帮助生产这些 GPU？ -> $TSM 谁帮助在数据中心扩展这些 GPU 集群？ -> $LITE、$COHR、$AAOI 用于生产扩展 GPU 集群所需互连件的原材料是什么？ -> $AXTI 它们都主题性地搭乘 AI 资本支出(CapEx)浪潮，但敞口各不相同。 我通常不会说“1”，因为它们的风险特征不同。 如果必须为复利增长选择最安全的一个：$TSM 如果必须选择 6 个月内潜在上行空间最大的一个？$AXTI 可能因材料短缺而受益，但如果中国封锁出口也可能归零。 如果必须选择 2 年内最有可能实现 5-10 倍增长并成为超大规模云厂商的一个？$NBIS。但存在稀释风险，自动取款机(ATM)发行可能导致股价停滞数月，但从 70-90 亿 ARR 和 20% EBIT 的未来前景来看，长期非对称上行空间最大。 当然，如果 AI 资本支出浪潮放缓，它们都会下跌，因此分散投资于金融科技等其他板块是明智的。
英文原文
Thanks for following along! I'll just give you a breakdown: I like Fintech and AI, those are basically my two main growth investment baskets. Stablecoins like $CRCL are in the fintech domain. As for AI domain I am just supply chain investing: Neoclouds -> Top level consumer/hyperscaler cloud providers for AI. $NBIS is my favorite. Go down the supply chain, who helps manufacture the GPUs used that $NBIS and $IREN buy? -> GPUs from $NVDA, some $AMD, and maybe ASICs from $GOOGL, $MSFT, $AMZN down the road. Who helps create those GPUs -> $TSM Who helps scale up those GPU clusters in data centers? -> $LITE, $COHR, $AAOI What are those raw materials used to produce the interconnects to scale up those GPU clusters? -> $AXTI They're all ride the AI CapEx wave thematically, but the exposure is all different. I never typically never say "1" because they all have different risk-profiles. if I had to choose the safest one for just compounding: $TSM If I had to choose the one with the highest possible upside over 6 months? $AXTI from material shock, but it could also go to 0 if China blocks exports. If I had to choose the best chance to 5-10x and become a hyperscaler over 2 years? $NBIS. But there's risks from dilution, ATMs stalling out the stock for a few months, but highest asymmetrical upside over periods of time from 7-9B arr 20% ebit in the future But of course, if AI CapEx wave slows, then they all go down, which it's good to diversify in other segments like fintech.
-
分析InP材料供应瓶颈，通过历史案例说明关键材料博弈论，预判超大规模云商将抢购InP衬底产能。
当我发WSB帖子时，我半开玩笑地提到$AXTI。但对于极端瓶颈和InP供应冲击，我可不是开玩笑的。这是一个关于关键材料博弈论的呼唤。 如果我们回顾历史上许多瓶颈案例： 1. 镝(Dysprosium)：约2300%涨幅（约$100/kg -> $2400/kg，2010-11年） 2. 氖气(Neon Gas)：约1000%-2000%涨幅（2022年） 3. 铑(Rhodium)：约$3,000/盎司 -> 约$29,000/盎司（约860%） 当前铟(In)的价格涨幅与其他材料相比如何？ 89%：（约$440/kg -> 约$832/kg）。 用于半导体的关键材料（如用于光刻激光器的氖气），因为芯片制造商如英特尔和三星在乌克兰冲突期间进行了储备，价格暴涨超过2000%。他们吸收了成本，因为氖气只占芯片价值的一小部分。 与氖气一样，磷化铟(InP)是一种关键的"赋能"材料。晶圆成本相对于AI集群成本（$100的晶圆对比$30,000的GPU）来说很小。超大规模云商宁愿吸收巨大的价格上涨，也不愿延迟部署。InP目前还没有出现这种情况，但我们很可能会看到。 传统分析师和许多AI模型不知道如何建模关键瓶颈，因为历史上很少有这样的案例：一种极其廉价的商品（如InP，TAM只有几亿美元）突然成为价值数万亿美元的AI建设中最关键的材料。 目前，超大规模云商可能已经向下游供应商Coherent/$LITE下了订单，但可能没有意识到激光器生产和产能受到上游材料短缺的制约而排满。 InP衬底（$AXTI + Sumitomo + JX生产）和磷化铟原料（$AXTI、Vital、$DOWA）都存在供应储备，需求超过供应数倍。 $NVDA最初通过锁定EML大量产能加剧了供应短缺。但短缺早在超大规模云商扩产之前就已经存在，大部分需求将在2026年底至2027年Trainium或Maia出现时涌入。 但如何绕过原始瓶颈并确保AI部署？ 直接从（$AXTI、Sumitomo）购买衬底产能，或者更进一层，购买InP原料并将其交给衬底生产商，以确保谷歌的TPU项目不会停滞。 与其他关键材料（如氖气用于光刻激光器）类似，磷化铟和InP衬底对整个AI建设更为关键： 1. Nvidia：InfiniBand / NVLink（光学 - 800G/1.6T EML光模块）- 对InP高度依赖 2. Google：Jupiter / Apollo（光学 - OCS），高度依赖 3. Microsoft：Azure Maia：高（800G DR4/FR4） 4. Meta：F16 / Artemis（800G）高 这还不包括其他Mag7公司如$AMZN或使用光子技术扩展ASIC的其他公司。 目前，分析师基于TAM对InP公司进行建模。例如，2024年InP晶圆市场总规模为$183M-$205M（Straits Research, Mordor Intelligence，2024年） - 6英寸（150mm）细分市场TAM：$65M - 预计2032年InP总TAM：$580M-$700M 现在看激光级磷化铟原料的TAM： - 多晶InP原料TAM：约$3亿 - 高纯度铟（6N/7N）仅：约$4亿 衬底数据：Mordor Intelligence（InP晶圆市场2025-2030）和Straits Research。 原料数据：Market Report Analytics（磷化铟多晶行业分析2025）。 这是一种极其便宜的电信产品，现在成为整个供应链中最珍贵的材料之一。 价格上涨3000%只会占$GOOGL TPU部署或$MSFT MAIA部署BOM值的低个位数。而仅微软2027年的扩产就会占用全部InP产能的两位数百分比。 就像2022年的英特尔或半导体行业一样，超大规模云商可能会像当时囤积氖气一样，直接购买InP衬底+InP作为"保险"。 如果"TAM"实际上只有几亿美元，但每个超大规模云商如果不确保这种材料就会面临部署延迟，那么这就变成了一场博弈论竞标战。 在一个因$100衬底而停滞的价值$200亿的TPU部署中，超大规模云商会毫不犹豫地支付$10,000（比当前价格高出100倍）。 这就是为什么当你看到InP衬底公司有$5000万订单积压（来自2024年），同样的$5000万产能对未来来说在极端情况下可能价值$5亿、$50亿甚至更多。 这是一个"尾部风险"投资，如果超大规模云商被这种材料卡住，回报是极高的，但并非没有重大的地缘政治风险和稀释。 我不知道会发生什么，但我预计2026年InP衬底和InP原料都会出现巨大瓶颈。 那么，如果像$AXTI这样的公司拥有整个InP供应链40%的份额（来源：AXT CEO），而估值只有$7亿，你会给它什么估值？
英文原文
When I made my WSB post, I'm half joking about $AXTI. But not about the extreme bottleneck and InP supply shock. This is a call on critical materials game theory. If we see many historically at bottlenecks: 1. Dysprosium: ~2,300% increase (~$100/kg -> $2400/kg 2010-11) 2. Neon Gas: ~1000%-2,000% increase (2022) 3. Rhodium: ~$3,000/oz -> ~$29,000/oz (~860%) The current Indium spike compared to others? 89% : (~$440 / kg -> ~$832 / kg). Critical materials used for semiconductors like Neon Gas (semi-grade lithography lasers), spiked over 2000% because chipmakers like Intel and Samsung had stockpiled during Ukraine conflicts. They absorbed the cost because neon is a small fraction of the chip's value. Like Neon, InP is a critical "enabler" material. The cost of the wafer is small relative to the cost of the AI cluster ($100 wafer vs $30,000 GPU). Hyperscalers would happily absorb huge price increases rather than delay deployment. We haven't seen that yet with InP, but we likely will. Traditional analysts and many AI Models do not understand how to model critical bottlenecks because there's not many examples in history where an extremely cheap commodity like InP with few hundred million TAM suddenly became the most critical material for the multi-trillion+ AI buildout. Currently, hyperscalers likely placed backlogs on downstream providers Coherent / $LITE, but likely didn't realize that laser production and capacity are backlogged because of materials shortage upstream. There is a supply storage both for InP substrates ( $AXTI + Sumitomo + JX production) as well as the Indium Phosphide feedstock ( $AXTI, Vital, $DOWA) where demand exceeds supply by multiple factors. $NVDA originally contributed to the supply shortage by locking in a large percent of capacity of EML. But shortage is already pre-hyperscaler ramp, where majority of the demand will come in late 2026 into 2027 when Trainium or Maia show up. But how do you get around original chokepoint and securing your AI deployment? Buying out substrate capacity directly from ( $AXTI, Sumitomo) or going one step lower and buying InP feedstock and passing them to substrate producers so Google's TPU program doesn't stall. And similar to how other critical materials were needed for lithography lasers. Indium Phopshide and InP substates are even more critical to the entire AI buildout: 1. Nvidia: InfiniBand / NVLink (optical - 800G/1.6T EML Transceivers) - Extreme dependency on InP 2. Google: Jupiter / Apollo (optical - OCS), Extreme dependency 3. Microsoft: Azure Maia: High (800G DR4/FR4) 4. Meta: F16 / Artemis (800G) High and this is not considering other Mag7 companies like $AMZN or other companies scaling out ASICs with photonics. Currently, analysts are modeling InP companies based on TAM. For example, the total InP Wafer Market from 2024 is $183M – $205M Straits Research, Mordor Intelligence (2024) - 6-inch (150mm) Segment TAM: $65m - Projected Total InP TAM (2032): $580M – $700M Now looking at TAM for Laser-Grade Indium Phosphide Feedstock: - Polycrystalline InP Feedstock TAM: ~$300 Million - High-Purity Indium (6N/7N) Only: ~$400 Million Substrate Data: Mordor Intelligence (InP Wafer Market 2025-2030) and Straits Research. Feedstock Data: Market Report Analytics (Indium Phosphide Polycrystalline Industry Analysis 2025). This was an extremely cheap telecom product and now it's one of the most precious materials in the entire supply chain. Increasing prices 3000% would only be low single digits of BOM value to $GOOGL TPU deployments or $MSFT MAIA deployments. And the project Microsoft ramp in 2027 alone would take up double digits of all InP capacity. Like Intel or semis back in 2022, hyperscalers will likely buy insurance directly with InP substrates + InP, similar to how semis stockpiled Neon Gas. If the "TAM" really was a few hundred million, but every single hyperscaler would face deployment delays if they don't secure this material. Then this becomes a game theory bidding war. In a $20B TPU deployment is stalled because of a $100 substrate, a hyperscaler would easily pay $10,000 (100 times current prices) for it without hesitation. That's why when you look at InP substrate companies and see $50m backlog (from 2024), that same $50m worth of capacity for future ones could be valued at $500m, $5B or more in an extreme scenario. This is a "tail-risk" investment, extremely reward if hyperscalers get bottlenecked by this material, but not without significant geopolitical risk and dilution. I don't know what's going to happen, but I do expect there to be an immense bottleneck in 2026 both in InP substrates and InP feedstock. So, if companies like $AXTI owns 40% of the entire InP supply chain (source: AXT CEO) and is valued at $700m, what value would you place on it?
-
梳理AI光子学及InP衬底供应链核心标的，涵盖从底层材料到芯片的全链条。
呃，根据我个人的研究，为了全面覆盖我看好的一切： 1. $AXTI 位于整个供应链的最底端，占据 1/3 的磷化铟(InP)衬底和 1/4 的磷化铟(InP)份额。AI 供应链竟然与这只市值 7 亿美元的股票紧密相连，这完全出乎意料。 2. $DOWA + 住友电气，用于在衬底和材料方面进行西方对冲。 3. $AAOI, $LITE 针对超大规模云服务商的芯片，一家服务于 $AMZN/$MSFT，另一家服务于其他客户但更侧重于 $TPU。 4. $MRVL + $AVGO 我认为这就足以捕捉光子学物料清单(BOM)的价值以及磷化铟(InP)供应冲击。 还有一些其他有趣的标的，比如 Landmark Opto，我正在关注。 $POET 没进这个名单，反正你可以通过 $MRVL 获得敞口，除非我漏掉了什么。
英文原文
Uhh, from my own personal research so far, to cover all grounds I’ve liked: 1. $AXTI at the very bottom of the entire supply chain + 1/3rd of InP substrates 1/4th of InP. Completely unheard of that AI supply chain is tethered to this $700m stock. 2. $DOWA + Sumitomo Electric for Western Hedge on substrate + materials. 3. $AAOI, $LITE for hyperscaler chips, one for $AMZN/ $MSFT, the other for everything but more levered to $TPU 4. $MRVL + $AVGO I think that’s kind of all you need to capture photonics BOM value + InP supply shock. There’s some other interesting ones like Landmark Opto that I’m looking at now. $POET didn’t quite make that list and you’d be getting exposure through $MRVL anyway, unless there’s something I missed
-
AI光互连依赖磷化铟，其原料78%在中国，AXTI成关键瓶颈。
“瓶颈中的瓶颈”：磷化铟(InP) $AXTI 和 $SMTOY 构成双寡头。它们控制了全球 60% 以上的磷化铟(InP)衬底，是 AI 建设的关键瓶颈。 更令人警觉的是：这依赖于磷化铟(InP)。 磷化铟(InP)是垄断市场。78% 由 $AXTI 和中国控制。 这就是“瓶颈中的瓶颈”： 磷化铟(InP)衬底市场已是日本和中国实体之间的双寡头。然而，上游激光级原料（如住友所需）的生产主要由 $AXTI（再次出现）和中国公司主导。 主要问题在于： AI 行业极度依赖光子学来实现规模化。无论收发器是“基于磷化铟”还是“基于硅”，光源始终是：磷化铟(InP)。 随着 AI 集群从 1 万颗 GPU 扩展到 10 万颗以上，“铜缆瓶颈”迫使全面转向光互连。这使得磷化铟(InP)对 AI 规模化的重要性超过几乎任何其他单一材料。 此外，激光级原料纯度必须至少达到 6N（99.9999%）。即使是微量杂质（硅、硫、锌）也不像 LED 等其他领域那样可接受。大多数研究可能混淆了激光级原料与普通磷化铟产出。 住友是磷化铟衬底市场最著名的名字。他们在高端激光衬底领域占据巨大市场份额。但贸易数据显示，他们越来越依赖来自 $DOWA 和中国公司的外部磷化铟原料。（由于自有回收和储备，他们可能仍被计入原料所有权列表中） 所以，如果认为整个西方 AI 建设因有一家日本公司作为双寡头之一而感到安全，请三思。 瓶颈的源头被 $AXTI 和中国供应商所瓶颈。 如果我们看生产，以下是深度研究估算的控制权分布： Vital Materials - 35%（现货市场） - Vital 在 2020 年收购凡亚金属交易所库存（3600 吨铟），使其掌控全球铟市场。他们利用这一优势成为最大的磷化铟多晶生产商。主导成本领先者；控制原生铟供应（凡亚库存）。 $AXTI - 25%（北京同美晶体技术） - 主要自给自足。其大部分多晶生产被自家晶体生长炉消耗以制造晶圆。但他们也向现货市场销售原材料和化合物，是混合型玩家。 他们是全球唯一从矿山、精炼厂到磷化铟衬底产出的公司，鉴于他们同时拥有原材料和磷化铟衬底控制权，这是一个绝对疯狂的瓶颈。供应自家晶圆生产；关键的非日本来源。 株洲科能新材料 - 18% - 科能实现了“磷化铟多晶量产”并拥有“高度集成的自动多晶生产能力”。他们是中国铟出口配额的关键接收方。纯合成专家；主要出口商。 然后外部： $DOWA - 12% - 西方公司中最大的供应商（是的，日本）。在采矿（秋田锌）方面根基深厚，因此是高纯度金属有机化合物和化合物的领先供应商。 “日本企业”（住友、JX）的“来源” Wafer Technology Ltd - 5% - IQE plc 的子公司。对于需要非中国原料的北约联盟供应链来说，这是关键战略资产，尽管他们目前正在进行资产清算。 InPACT（法国）- 3% - InPACT 称他们通过高压合成的水平梯度凝固(HGF)生产多晶锭，以满足（诺基亚、爱立信）和一些欧洲国防部门的要求。 世界其他地区 - 2% 这包括主要衬底制造商（Coherent、住友）、小规模研究生产以及试图打破垄断的韩国或台湾新兴玩家。 现实是，全球 78% 的激光级磷化铟原料产能位于中国。 光组件的需求已经以数倍超过供应/生产。这还是在上量之前。整个 AI 建设被两家公司瓶颈。 但 1/4 的材料来自：$AXTI。而约 1/3 的磷化铟衬底产出再次来自激光级原料。 整个 AI 建设依赖于一家随机的 7 亿美元公司，它既是瓶颈中的瓶颈，也是瓶颈本身。 扩大磷化铟衬底和激光级原料的生产既是国家安全问题。
英文原文
"Bottleneck within a Bottleneck": Indium Phosphide $AXTI | $SMTOY is a duopoly. They control 60%+ of the world's InP substrates as the bottleneck of the AI buildout. What's even more alarming? This is dependent on: Indium Phosphide. InP is a Monopoly. 78% control by $AXTI / China. Here's the bottleneck within a bottleneck: The InP substrates is already a duopoly between Japanese and Chinese entities. However, upstream production of laser-grade feedstock required by companies like Sumitomo is dominated by $AXTI (again) and Chinese companies. Here's the major issue: The AI industry is extremely reliant on photonics to scale. Whether, transceivers are "InP-based" or "Silicon-based" the photon source remains: Indium Phosphide. As AI clusters move from 10,000 to 100,000+ GPUs, the "copper bottleneck" is forcing a total shift to optical interconnects. This makes InP more critical to AI scaling than almost any other single material. On top of that, laser-grade feedstock must be at least 6N (99.9999%) pure. Even trace amounts of impurities (Silicon, Sulfur, Zinc), cannot be used like they do in other segments like LED. And most research probably confuses laser-grade feedstock with regular InP outputs. Sumitomo is arguably the most famous name in the InP substrate market. They hold a massive market share in the high-end laser substrate segment. Yet trade data suggest they are increasingly reliant on external feedstock for Indium Phosphide from companies like $DOWA and from China. (They're likely still included in feedstock ownership lists bc of captive recycling and reserves) So if the entire Western AI buildout feels safe regarding a Japanese company as a duopoly, think again. The source of the bottleneck is bottlenecked by Chinese suppliers like $AXTI and Chinese companies. If we look at production there's are estimates from deep research breakdown of estimate controls: Vital Materials - 35% (Merchant) - Vital acquired Fanya Metal Exchange inventory (3,600 tonnes of Indium) in 2020 gave them large control over the global Indium market. They have leveraged this to become the largest producer of InP polycrystal. Dominant cost leader; controls raw Indium supply (Fanya stocks). $AXTI - 25% (Beijing Tongmei Xtal Technology) - Primarily Captive. Most of their polycrystal production is consumed by their own crystal growth furnaces to make wafers. However, they also sell raw materials and compounds to the merchant market, making them a hybrid player. They're the only company in the world that goes from mines, refineries -> InP substrate output, which is an absolutely crazy bottleneck given they own both the raw materials into InP substrate controls. Feeds own wafer production; key non-Japan source. Zhuzhou Keneng New Material - 18% - Keneng has achieved "mass production of polycrystalline indium phosphide" and possesses "highly integrated automatic polycrystalline production capacity". They are a key recipient of China's export quotas for Indium. Pure-play synthesis specialist; major exporter. Then Outside: $DOWA - 12% - The biggest supplier out of any Western company (yeah Japanese). Deep roots in mining (Akita Zinc), so they're the leading supplier of high-purity metal organics and compounds. The "source" for Japan Inc. (Sumitomo, JX) Wafer Technology Ltd - 5% - A subsidiary of IQE plc. Critical strategic asset for NATO-aligned supply chains that require non-Chinese feedstock, even though they're going through asset liquidations right now. InPACT (France) - 3% - InPACT states they produce polycrystalline ingots via Horizontal Gradient Freeze (HGF) using high-pressure synthesis so they meet requirements from (Nokia, Ericsson) and some European defense sectors. Rest of World - 2% This includes major substrate makers (Coherent, Sumitomo), small-scale research production, and emerging players in Korea or Taiwan who are attempting to break the monopoly. The reality is that 78% of the global capacity for laser-grade InP feedstock is located in China. The demand for optical components already exceeds supply/production by multiple factors. This is pre-ramp too. The entire AI buildout is bottlenecked by two companies. But 1/4th of the materials comes from: $AXTI. And ~1/3rd of InP Substrate outputs again comes from laser-grade feedstock. The entire AI buildout is dependent on some random $700m company that serves as both the bottleneck of a bottleneck and the bottleneck. It's both national security issue to scale up InP production both for substrate and laser-grade feedstock.
-
InP衬底成AI光互连瓶颈，$AXTI/$SMTOY掌控命脉，2026年或致供应链危机。
“磷化铟(InP)瓶颈”：AI基础设施建设的关键瓶颈解析： $NVDA Blackwell、$META MTAI、$GOOGL TPU 和 $MSFT Maia 的产能爬坡，其未来取决于两家市值仅7亿美元的小盘股 $AXTI 和 $SMTOY。 如果无法解决 InP 问题，AI“增长”故事将在2026年终结。 原因如下： AI行业正开始向光子学迁移，以部署未来的 ASIC/GPU，因为铜互连已触及物理极限。 然而，超大规模云服务商在这样做时，却将命运押注在了 InP（磷化铟）这一通用材料上，而全球仅有少数工厂能以激光器所需的纯度生产6英寸 InP 晶圆。 以 Google 及其 TPU v7 Ironwood 项目为例： Google 使用光电路交换(Optical Circuit Switching, OCS)，简单来说就是由光构成的交换机。Pod 中的每一个 TPU 进行通信都需要基于 InP 的激光器。与 Google 合作的 $LITE 在很大程度上依赖 InP 衬底（如 AXT/住友）来制造这些激光器。 如果没有这些衬底，Google 的整个 Ironwood 项目不仅仅是“放缓”，而是会直接撞墙。 来自 $NVDA GB 系列、$AMZN Trainium、$MSFT Maia、$META MTAI 的现代 ASIC/GPU 都做出了同样的押注：光是未来的方向。 现在，问题出现了。 整个西方 AI 路线图目前都系于一家市值7亿美元的小盘股和一家单一的日本公司，它们生产了全球光子学所需的大部分 InP 衬底。 目前这是一个双寡头格局（粗略估计 AXT + 住友供应约60%），最近的估计显示约70%+的供应来自住友电气、AXT、Freiberger、JX 和 Visual Photonics Epitaxy（填补剩余缺口）。 无论如何，整个未来的 AI 供应链细如针尖： - Moomoo 研究：InP 市场处于“全球争夺”和“严重供应短缺”状态，NVIDIA GB200 的推出（扩展仍需大量 InP，不仅仅是机架内的 NVL72 通信）。 - 高速收发器的需求目前可能超过供应近两倍（LightCounting） - 看到创纪录的预订，但明确“受限于 InP 激光器的供应”（$COHR CEO Q3 财报电话会） - 麦肯锡：800G 模块存在 40% 到 60% 的缺口，1.6T 模块存在 30% 到 40% 的缺口。 鉴于需求的激增，这些报告可能理解得过于保守。即使按照微软对 Maia 爬坡的预测（据 UBS $MRVL 笔记，2027年 Maia 估计超过100万台），加上未来一年超过200万台 1.6T 收发器，这一体量如此之大，代表了全球衬底产量的双位数百分比。 AI 预期的“指数级增长”即将与关键材料生产的现实发生碰撞。因此，“Ironwood”、“MTIA”和“Maia”的爬坡不仅雄心勃勃，在当前材料限制下可能根本不可能实现。 即使 $COHR、JX 日本、住友、$AXTI 等以最大产能扩产（例如 $COHR / JX -> 6英寸 InP 晶圆产能提升4倍），他们可能仍无法满足超大规模云服务商日益增长的需求。特别是考虑到需求激增，例如仅 $NVDA 自身（GB200/GB300 修订版）的需求。 硅光子学等技术解决方案可以弥合差距，但这仍然主要需要外置 InP 激光器作为光源。TFLN 或量子点激光器还需要很多年才能成熟。 未来几年可能无法摆脱对 InP 的需求。 因此，芯片设计与材料可用性之间的错配创造了一个战略瓶颈，在供应链的最底层，极少数公司控制了大部分配额、定价和供应。当叠加美中关系紧张和出口管制的地缘政治风险时，这尤其危险。 话虽如此，以下是可能发生的事情： - 价格飙升：$AXTI、JX、住友的价格将大幅飙升 -> $LITE、$COHR、中际旭创（也会通过下游传导提高价格） - 超大规模云服务商将直接囤积材料，绕过传统组件采购，直接从 $AXTI、JX 日本、住友购买 InP 衬底库存，并直接委托给像 $LITE 这样的收发器制造商。 （例如，Meta 将绕过收发器公司，直接去找 AXT 或住友） - 超大规模云服务商将买断生产配额（就像 $NVDA 已经积极“锁定”EML 产能（在 InP 衬底上制造）那样）。 购买衬底制造商或生产配额将成为一种必要，以免被 $NVDA 或 $GOOGL 等竞争对手饿死。 随着 TPU v7 和其他超大规模云服务商在 2026-2027 年爬坡，我们可能会进入衬底的“饥饿游戏”阶段，每个超大规模云服务商都将为了资源配额而相互吞噬对方的增长。 像 $NVDA（拥有创纪录的预分配量）这样的公司可能暂时没事，但其他项目可能会面临重大延误。 思考： 1. 一些超大规模云服务商可能没事（$NVDA）。其他如 $GOOGL 和 $MSFT 需要在其他人之前买断材料和配额。 2. 行业需要加倍投入工程转型，如延长铜的使用寿命和更节省材料的方式如硅光子学(SiPh)。 3. 转向6英寸晶圆以提高良率（能缓解情况，但仍不足以满足需求） 因此，现状是，数万亿美元的 AI 扩展系于一些不起眼的7亿美元公司 $AXTI 和 $SMTOY。除非架构改变，否则 AI 似乎不可避免地会因 InP 衬底产能而触及物理天花板。 2024年的瓶颈是 GPU。2025年是 HBM。2026年，主要约束很可能是光互连，特别是驱动它们的 InP 衬底。 这已成为 AI 基础设施建设中隐藏的瓶颈。
英文原文
The "InP Chokepoint": The Bottleneck of the AI Buildout explanation: The future of $NVDA Blackwell, $META MTAI, $GOOGL TPU, and $MSFT Maia ramp is tied to: A $700M small cap $AXTI and $SMTOY. The AI "Growth" story ends in 2026 if there's no solution to InP. Here's why: The AI industry started its migrating to photonics for future ASIC/GPU deployments, because copper is hitting a physical limit. However, in doing so, hyperscalers traded the common material for InP (Indium Phosphide), when there's only a few factories capable of producing 6-inch InP wafers at the purity levels required for lasers. Let's take for example Google and their TPU v7 Ironwood program: Google uses Optical Circuit Switching (OCS), in simpler terms, switchboards made of light. For every one of those TPUs in the pod to talk, they require InP-based lasers. $LITE, which works with Google on this, largely depends on InP substrate (eg. AXT/Sumitomo) to make them. If they don't have it Google's entire Ironwood program doesn't just "slow down", it hits the wall. Modern ASICs/GPUs from $NVDA GB series, $AMZN Trainium, $MSFT Maia, $META MTAI have all made the same bet: Light is the way forward. Now, here's the issue. The entire Western AI roadmap is currently tethered to a $700M small-cap and a single Japanese company that produce majority of the world's InP substrates required for photonics. It's currently a duopoly (rough estimates majority supply ~60% between AXT + Sumitomo), with recent estimates of ~70%+ coming from Sumitomo Electric, AXT, Freiberger, JX, and Visual Photonics Epitaxy (filling in the gaps). Regardless, the entire future AI supply chain is thinner than a needle: - Moomoo Research: InP market is in a state of "global scramble" and "serious supply shortage" NVIDIA GB200 rollout (scale-out still requires tons of InP, not NVL72 within-the-rack comm). - Demand for high-speed transceivers today probably exceeds the supply by almost a factor of two (LightCounting) - Seeing record booking, but explicitly "supply-constrained by InP lasers" ( $COHR CEO Q3 ER) - McKinsey: 40% to 60% shortfall for 800G modules and a 30% to 40% shortfall for 1.6T modules. And these reports are likely understanding + very conserative given the demand ramp. Even going off Microsoft's projections on Maia ramp, ( est. 1M+ Maia by 2027 on UBS $MRVL note), with 2 million+ units of 1.6T transceivers over the next year, this volume is so large it represents a double-digit percentage of global substrate output. The projected "exponential growth" of AI is about to collide with the reality of critical material production. So, the "Ironwood", "MTIA" and "Maia" ramps aren't just ambitious, they may be impossible under current material constraints. Even if $COHR, JX Nippon, Sumitomo, $AXTI, and others, ramp up at maximum capacity (eg. $COHR / JX -> 6-inch InP wafers for 4x capacity), they still might not be able to meet the increasing demand from hyperscalers. Especailly with demand spikes occurring, eg. just for $NVDA alone (GB200/GB300 revisions). There are technical solutions like silicon photonics is one solution to bridge the gap, but this still largely requires an external InP laser as the light source. TFLN or quantum dot lasers are many many years away. There's probably no escaping the InP requirements for the next few years. So, the mismatch between chip design and material availability has created a strategic chokepoint, where if you go to the very bottom of the supply chain, very few companies control a majority of allocations, pricing, and supply. This is especially dangerous when compounded with geopolitical risks on US/China relations + export controls. That being said, here's what's probably what's going to happen: - Price Spikes: Prices from $AXTI, JX, Sumitomo will spike significantly -> $LITE, $COHR, Innolight (also increases prices from pass down) - Hyperscalers will directly stockpile materials, bypassing traditional component procurement and buying InP substrate inventory from $AXTI, JX Nippon, Sumitomo, and directly to consign to transceiver manufacturers like $LITE. (eg. Meta would bypass transceiver companies and go directly to AXT or Sumitomo) - Hyperscalers will buy out production allocation ( like $NVDA that has already aggressively "locked in" EML capacity (manufactured on InP substrates). Buying a substrate manufacturer or production allocation would become a necessity to so others like $NVDA or $GOOGL doesn't starve them out. As TPU v7 and and as other hyperscalers ramp up in 2026-2027, we will likely enter a "hunger games" phase for substrates where only each hyperscaler will be cannibalizing each other's growth for resource allocation. Companies like $NVDA (with record amounts of pre-allocation), might ramp be okay for the time being, but others programs would likely face major delays. Thoughts: 1. Some hyperscalers might be fine ( $NVDA). Others like $GOOGL and $MSFT will need to buy out materials + allocation before others do. 2. Industry needs to double down on engineering shifts like copper life extension and more material efficient ways like SiPh. 3. Move to 6-inch wafers for yields (eases things, but still not enough to meet demand) So the way things are now, the multi-trillion dollar AI scaling are tethered to some obscure $700m company $AXTI and $SMTOY. It seems inevitable that AI will hit the physical ceiling because of InP substrate capacity unless architectures change. In 2024, the bottleneck was GPUs. In 2025, it was HBM. In 2026, the primary constraint will likely be the optical interconnect, and specifically, the InP substrates that power them. This has now become the hidden bottleneck of the AI buildout.
-
AXTI和住友垄断InP衬底，或成AI光子化最大瓶颈。
警告：整个AI行业可能会受到两家公司的瓶颈制约： 1. $AXTI（市值7亿美元） 2. $SMTOY（市值317亿美元） 这两家公司控制了全球60-70%以上的磷化铟(InP)衬底市场。 未来$NVDA、$GOOGL TPU v7集群、$META、$MSFT、$AMZN等超大规模数据中心集群都需要基于InP的激光器和接收器。 $AVGO、$LITE、$COHR使用电吸收调制激光器(EML)用于800G/1.6T光模块、分布反馈(DFB)激光器和其他光基础设施。 没有InP衬底，供应链就会停滞。 在查看了从TPU到Maia的物料清单(BOM)后，看起来未来的专用集成电路(ASIC)+GPU+超大规模部署严重依赖光子学。 而这两家供应商可能会冻结全球InP衬底市场，涵盖几乎所有领域： - 超大规模光器件（TPU集群等） - 光模块（5g，数据） - 激光雷达(LiDAR)（自动驾驶出租车，无人机，军事） - 光模块（互连集群） - 硅光子学激光芯片（英伟达未来的共封装光学(CPO)和英特尔/博通硅光子学引擎使用InP连续波激光阵列。） 由于这些公司占据了市场供应的大多数： -AXTI（估计~30-35%） -住友（估计~30%） - JX日本（估计10-15%） 仅此而已。（例如，Yole 2021年的行业报告指出“住友电气+AXT共同拥有‘超过75%’的InP衬底市场”） 超大规模/AI正在向光子学发展，但整个AI行业是脆弱的。 如果$AXTI或$SMTOY中的任何一家停止供应材料，整个未来的AI建设就会瘫痪。更疯狂的是，一家7亿美元的公司可能成为这一切的中心。 随着AI行业向光子学转变，InP衬底可能会成为与高带宽内存(HBM)并列的最大瓶颈之一。
英文原文
Warning: The entire AI industry will likely be bottlenecked by two companies: 1. $AXTI ($700M) 2. $SMTOY ($31.7B) Which both control 60–70%+ of the world's InP substrates. Future $NVDA, $GOOGL TPU v7 pods, $META, $MSFT, $AMZN hyperscaler clusters require InP-based lasers and receivers. $AVGO, $LITE, $COHR use for EMLs for 800G/1.6T transceivers, DFB lasers, and other optical infra. Without InP substrates, the supply chain falters. After looking at TPU BOM to Maia BOM, it looks like future ASICs + GPUs + hyperscaler deployments are heavily reliant on photonics. And two vendors could freeze the global InP substrate market covering nearly all of: - Hyperscaler optics (TPU pods, etc) - Optical transceivers (5g, data) - LiDAR (robotaxis, drones, military) -Optical Modules (interconnect clusters) - Silicon photonics laser dies (Nvidia’s future co-packaged optics and Intel/Broadcom SiPh engines use InP CW laser arrays.) Since these companies make up majority of the market supply: -AXTI (est. ~30–35%) -Sumitomo (est.~30%) - JX Nippon (est. 10-15%) That’s it. (eg. 2021 industry note from Yole states that "Sumitomo Electric + AXT together had “more than 75%” of the InP substrate market") Hyperscalers/AI are moving toward photonics but the entire AI industry is fragile. If either $AXTI or $SMTOY stop supplying materials, the entire future AI buidlout gets crippled. It's even crazier that a $700m company could become the the center of it all. InP substrate will likely one of the biggest bottlenecks alongside HMB as the AI industry shifts to photonics.
-
解析微软Maia 300供应链，指出AAOI和MRVL为光模块核心受益者。
个人研究：$MSFT Maia 300 供应链。 根据 $MRVL 的报告，瑞银(UBS)预计2027年前 Maia 出货量将超100万颗。 预计 Maia ASIC 量产将带来超180亿美元流向供应商。 资金流向如下： 1. $MRVL - 营收80-120亿美元（~30-40%利润率），作为主要受益者。他们设计数字底层架构（SerDes，互连），从 SK Hynix 购买 HBM4 内存，支付 TSMC 的 2nm 晶圆和 CoWoS 封装费用，并将成品模块卖给微软。 当 Maia 在2026年底至2027年初开始量产时，鉴于其带来的巨大营收，Marvell 可能会迎来大幅重估。 2. 光网络架构：约30亿至40亿美元 主要受益者：Applied Optoelectronics ($AAOI)，Coherent ($COHR)，Innolight Maia 集群需要1:1或更高的光收发器与GPU比例。对于1.6T速度（Maia 300所需），每个节点的光学内容估计超过3,000美元（每个加速器约6个800G/1.6T模块）。 这里 AAOI 大获全胜：与市值700亿美元的 $MRVL 不同，$AAOI 是一家25亿美元的公司。即使只获得这块30亿美元蛋糕的20-30%（6-9亿美元），也将使其当前年化经常性收入(ARR)翻三倍，但具体份额取决于其德州磷化铟(InP)工厂的扩产情况（可能更多或更少）。 因此，Applied Optoelectronics (AAOI) 和 Marvell 的光学 DSP 业务是隐形赢家，受益于连接这些巨型芯片所需的1.6T升级周期。 像 $TSM 这样的公司会增加30-40亿美元营收，但相对于其1000多亿美元的营收来说，这只是中等体量。（仅说明 TSM 体量之大） 硅基 BOM 拆解： 1. 计算逻辑 (ASIC) - TSM + Intel Foundry/18A Marvell (设计), TSMC, Intel, ARM/Synopsys? ($2,800 - $3,500) 2. 内存 (估计 HBM4? (6-8层堆叠), 估计值) SK Hynix, Samsung ($3,360 - $4,500) 3. I/O 与缓存 Die - TSMC N5/N4 Marvell, TSMC ($400 - $600) 4. 先进封装 - CoWoS-L / Foveros Direct TSMC, Intel ($900-$1300) 5. 散热方案 - 微流道冷板 / TIM ($200-400) 6. 杂项 (大电流电感, 电容) ($150-$200) 硅基 BOM 总计：$7,860 - $10,550 这非常粗略，未涉及无源器件/基板/盖板 (Ibiden, Unimicron)，测试与组装 (Advantest, Teradyne) 等。 现在如果我们深入看光学 BOM 以及 $MSFT 在空心光纤和 OCS 方面的布局： DSP / SerDes IP: $MRVL ($800 - $1,200) SiPh 引擎 (CPO): Marvell / Intel ($600 - $900) 激光源 (ELS): $AAOI, $LITE ($300 - $600) 收发器组装: $AAOI, Innolight ($400 - $600) 光纤布线: 空心光纤 / MPO ($MSFT) ($200 - $300) 光学 BOM 总计~$2,300 - $3,600~占系统总成本的25% 我们已经确立 $LITE 和 Innolight 是光子学的两大主要玩家（在 $NVDA 和所有超大规模云厂商 ASIC 中） 然而，$AAOI 的市值仅为 $MRVL 和 $LITE 的一小部分，作为小盘股，Maia 的机会对其股票更具变革性。 另需注意，富邦研究(Fubon Research)估计 Maia300 芯片将于2026年底开始生产，约30万至40万颗，2027年增至120万至150万颗，但我们仅采用瑞银的~100万颗数据。（因此实际爬坡可能更多） 但向1.6T的转变以及对空心光纤/OCS网络高功率激光器的特定需求是一个巨大的顺风。如果 AAOI 获得 Maia 光学 BOM 的一部分，他们到2027年仅从一个客户那里就能使 ARR 翻三倍或四倍（不包括 $AMZN）。 风险：主要风险是资格认证失败。如果他们的激光器未能通过微软严格的可靠性测试，他们可能会被 $LITE 一夜之间取代。 然而，Maia 300 的量产标志着超大规模云厂商 ASIC 交易的开始，关键供应商已就位。 从2026-2027年底这一特定 ASIC 中受益最大的公司？ 看起来是 $AAOI, $MRVL, $TSM 和 SK Hynix。 警告：这是基于公开材料+深度研究+映射的推测性研究+建模。没有公开的新闻证实这些信息，不要将其视为事实。
英文原文
Personal Research: $MSFT Maia 300 supply chain. On a $MRVL report, UBS est. 1M+ Maia by 2027. An est. of $18B+ is estimated to flow down to suppliers from Maia ASIC ramp. Here's where it goes: 1. $MRVL - $8-12B revenue (~30-40% margin) as the prime beneficary. They design the digital plumbing (SerDes, interconnects), buy the HBM4 memory from SK Hynix, pay TSMC for the 2nm wafers and CoWoS packaging, and sell the finished module to Microsoft. We'll likely see a large re-rating in Marvell when it comes time for Maia ramp late 2026, early 2027, especially given how much revenue this brings in. 2. The Optical Fabric: ~$3 Billion to $4 Billion Primary Beneficiaries: Applied Optoelectronics ( $AAOI ), Coherent ( $COHR ), Innolight Maia clusters require a 1:1 or higher ratio of transceivers to GPUs. For 1.6T speeds (required for Maia 300), the optical content per node is estimated at $3,000+ (approx. 6x 800G/1.6T modules per accelerator). Here AAOI Wins Big: Unlike $MRVL (which is $70B), $AAOI is a $2.5B company. Capturing even 20-30% of this $3B pie ($600M-$900M) would triple + their current ARR, but the amount captured is dependent on their InP Texas Fab scaling (could be a lot more or less). So Applied Optoelectronics (AAOI) and Marvell's optical DSP business are the silent winners, riding the 1.6T upgrade cycle required to interconnect these massive chips. Companies like $TSM gain another $3-4B revenue, but that's medium volume compared to their $100B+ revenue. (just speaking to how much of a giant TSM is) For the silicon BOM breakdown: 1. Compute Logic (ASIC) - TSM + Intel Foundry/18A Marvell (Design), TSMC, Intel, ARM/Synopsys? ($2,800 - $3,500) 2. Memory (est. HBM4? (6-8 stacks), estimate). SK Hynix, Samsung ($3,360 - $4,500) 3. I/O & Cache Die - TSMC N5/N4 Marvell, TSMC ($400 - $600) 4. Advanced Packaging - CoWoS-L / Foveros Direct TSMC, Intel ($900-$1300) 5. Thermal Solution - Microfluidic Cold Plate / TIM ($200-400) 6. Misc (High-Current Inductors, Caps) ($150-$200) Total Silicon BOM: $7,860 - $10,550 This is really high level doesn't go into Passive/Substrate/Lid (Ibiden, Unimicron), Test & Assembly (Advantest, Teradyne), etc. Now if we look deeper at optical BOM and what $MSFT is doing with hollow core fiber and OCS: DSP / SerDes IP: $MRVL ($800 - $1,200) SiPh Engine (CPO): Marvell / Intel ($600 - $900) Laser Source (ELS): $AAOI, $LITE ($300 - $600) Transceiver Assembly: $AAOI, Innolight ($400 - $600) Fiber Cabling: Hollow Core Fiber / MPO ( $MSFT) ($200 - $300) Total Optical BOM~$2,300 - $3,600~25% of Total System Cost We've established $LITE and Innolight as the two major players in photonics already (within $NVDA and every single hyperscaler ASIC) However, $AAOI MC is a fraction of $MRVL and $LITE as a microcap, making the Maia opportunity far more transformative for its stock. Someting also to note is Fubon Research est Maia300 chip is set to start prod in late 2026 with about 300,000 to 400,000 units, increasing to 1.2 to 1.5 million units in 2027, but we're just going with the ~1M figure from UBS. (so might ramp up more) But the shift to 1.6T and the specific requirement for high-power lasers for hollow-core/OCS networks is a big tailwind. If AAOI captures a portion Maia optical BOM, they triple or quadruple their ARR from one client alone by 2027 (not including $AMZN). Risk: The primary risk is qualification failure. If their lasers fail Microsoft’s rigorous reliability testing, they could be replaced by $LITE overnight. However, Maia 300 ramp signals the start of the hyperscaler ASIC trade and key suppliers are there. The companies that benefit the most from this specific ASIC late 2026-2027? Looks like $AAOI, $MRVL, $TSM, and SK Hynix. Warning: This is speculative research + modeling given public materials + deep research + mapping. There is no public breakdown and news of this information, do not treat it as a fact.
-
某仙股若执行成功，当前定价或低估其在AI集群中的价值。
嗯，不完全是。他们有过执行失败的历史，$AMZN 多年前就将其作为客户剔除。但如果考虑到他们现在手握铜缆到光子学（CPO）的“金票”，以及内存等供应常量，且能在此次成功执行，其当前的仙股定价可能并未计入他们在每个 Trainium 和 Maia 集群中的物料清单（BOM）份额，以及可选的美国磷化铟（InP）晶圆厂。
英文原文
Well not exactly, they have a history of execution failures with $AMZN dropping them as a client many years ago. But if they can execute this time around given their new golden ticket of copper -> photonics and supply constants like memory, their current penny stock pricing probably doesn’t price in their BOM share of every Trainium and Maia cluster they’re a part of + InP fab optionally in US.
-
InP成2026新瓶颈，市场重估AAOI与LITE在超大规模云ASIC中的价值。
$AAOI 自今日发布论点以来上涨 24%，$LITE 上涨 5%。 从物料清单(BOM)分析来看，LITE（市值 270 亿美元）因光路交换(OCS)技术而向 TPU Ironwood 倾斜，同时也受益于 NVDA 及所有专用集成电路(ASIC)。 AAOI（市值 25 亿美元）则主要受益于 MSFT MAIA 的量产爬坡和 Amazon Trainium。 磷化铟(InP)就像高带宽内存(HBM)一样，将成为 2026 年的瓶颈，因为它们是这些部署中激光器使用的基础材料。 类似于美光和 SK 海力士的内存瓶颈，市场注意力可能会转向 InP 晶圆厂，例如 $AAOI，它是美国少数几家此类工厂之一（还有 COHR, Macom）。 但相比之下，$LITE 由于 Google TPU 的成功（来自 Meta 和 Anthropic 的采购订单），今年迄今(YTD)已上涨 362%，而 $AAOI 今年迄今仅上涨 7%。 我们主要看到这种情况是因为散户或媒体对 $AMZN Trainium 或 $MSFT Maia 部署缺乏关注，这些部署预计将在 2026-2027 年大规模量产。 然而，由于每个超大规模云服务商都希望降低其自有云平台的推理成本，它们很可能都会成功。 如果我们看到其他超大规模云服务商采用 OCS 以实现 TPU 达到的优化性能，鉴于 $LITE 在该特定领域的垄断地位，预计其估值将比现在进一步提升。 然而，如果我们看到 $MSFT Maia 量产（鉴于 $AAOI 可能正在为他们开发新架构），以及 $AMZN Trainium 量产（40 亿美元权证 + 采购订单），预计 $AAOI 将重新估值。 光电子学和 InP 将成为像内存一样的新瓶颈。 我们可能会看到投资流向下游玩家，如 $COHR、中际旭创(Innolight)、$LITE，以及 2026 年针对特定超大规模云服务商 ASIC 的隐藏杠杆标的如 $AAOI 这一主题。 市场目前正在奖励 Google TPU 供应链，但可能错过了其他超大规模云服务商 ASIC 的量产机会。
英文原文
$AAOI is up 24% and $LITE is 5% since my thesis today. From BOM analysis, LITE ($27B) is levered toward TPU Ironwood due to OCS but benefits from NVDA + all ASICs. AAOI ($2.5B), is levered toward MSFT MAIA ramp and Amazon Trainium. InP like HBM, will be a bottleneck for 2026 as they’re the foundational materials used for lasers in these deployments. Similar to memory bottlenecks with Micron and SK Hynix, we’ll likely see attention drawn to InP fabs, such as $AAOI, which happens to be one of the sole ones in America (COHR,Macom) But compared to $LITE that is up 362% YTD due to the success of Google’s TPU (from Meta and Anthropic purchase orders), $AAOI is only up 7% YTD. We’re largely seeing this because there’s a lack of retail or media attention on the $AMZN Trainium or $MSFT Maia deployments, which are largely expected to ramp up in 2026-2027. However they’re all likely to succeed due to each hyperscaler wanting to lower costs of inference for their own cloud platform. If we see other hyperscalers adopt OCS for optimized performance that the TPU achieved, expect $LITE to re-rate more than they have now given their monopoly in that specific segment. However, if we see $MSFT Maia ramp up (given $AAOI is likely developing a new architecture for them), and $AMZN Trainium ramp up ($4B warrant + purchase orders), expect $AAOI to rerate. Photonics and InP will be the new bottleneck like memory. We’ll likely see investments pour down stream to players like $COHR, Innolight, $LITE, and hidden levered plays on specific hyperscaler ASICs like $AAOI as a theme in 2026. The market is currently rewarding the Google TPU supply chain but might be missing other hyperscaler ASIC ramps.
-
分析LITE与AAOI在AI芯片供应链中的价值及2026年前景
是的，$LITE 的表现令人惊叹，看到它出现在每一款 ASIC 以及 $NVDA Blackwell 部署中后更是如此。 然后，当你进行物料清单(BOM)分析并建模 TPU/GPU 部署时，它即使在历史高点(ATH)看起来也被低估了。 接着是 $AAOI，它与 LITE 类似，但更侧重于 AMZN | MSFT 的 ASIC。它是美国少数几家磷化铟(InP)晶圆厂之一，其扩张规模将取决于 Trainium 和 Maia 的表现。它的市值也只有 25 亿美元。 光子学/存储领域在 2026 年将会极其疯狂。
英文原文
Yeah $LITE is incredible after seeing how it's in every single ASIC + $NVDA blackwell deployments. Then once you do BOM analysis and model TPU/GPU deployments, it looks undervalued even at ATHs. Then there's $AAOI, which is similar to LITE but levered toward AMZN | MSFT ASICs. It's one of the only InP Fabs in US and will scale up depending on how Trainium and Maia do. It's a $2.5B MC too. Photonics/Memory is going to be extremely crazy in 2026.
-
$AAOI 是 $MSFT/$AMZN ASIC 扩产的高弹性隐藏宝石，具极高不对称收益。
在 $LITE、$COHR、中际旭创等公司中，我发现 $LITE 的不对称性（asymmetrical setup）最强，因为它是 Ironwood（含共封装光学(CPO)）、Blackwell 及其他超大规模客户专用集成电路(ASIC) 的核心。然而，$AAOI 是极致的超额收益（alpha）。尤其是作为 $AMZN Trainium 和 $MSFT Maia ASIC 扩产的对冲工具，以及美国少数磷化铟(InP) 晶圆厂之一。尽管拥有 $40 亿 $AMZN 认股权证和采购协议，12 月的订单似乎确实是基于 Maia 的集群。有传闻称 $AAOI 似乎正在为 $MSFT 的硅片开发新的光互连架构。尽管超大规模客户 ASIC 炒作和光子学反弹，$AAOI 股价毫无动静，因此在 $20 亿市值下绝对是隐藏宝石。但它高度依赖于 $MSFT 和 $AMZN（鉴于其云部门，它们很可能成功并大规模部署自有 ASIC）。
英文原文
Among $LITE, $COHR, Innolight, etc. I've found $LITE to be the most asymmetrical setup since it's central to ironwood (with OCS), blackwell, and other hyperscaler ASICs. However, $AAOI is extreme alpha. Especially as a levered play on $AMZN Trainium, $MSFT Maia ASICs scale up + one of the only InP fab in US. Although they have $4B $AMZN warrants + purchase agreements, it does seem likely that Dec's orders were Maia-based clusters. There's rumors that $AAOI seems to be developing a new optical interconnect architecture specifically for $MSFT's silicon too. $AAOI hasn't moved at all, despite hyperscaler ASIC hype + photonic rallies, so it's a hidden gem for sure at $2B MC. However, it is highly levered to $MSFT and $AMZN (which will likely succeed + deploy their own ASICs at scale given their own cloud departments)
-
分析LITE、COHR光子学双寡头地位及AAOI对微软亚马逊ASIC的杠杆看涨逻辑。
我手里只有 $LITE 和 $AAOI 哈哈。但我对 $LITE 和 $COHR 的看法是，它们在光子学(Photonics)领域基本处于双寡头地位。其中 $LITE 额外受益于 Google TPU 的增长杠杆。然后是 $AAOI，它具有极高的超额收益(Alpha)潜力，且是对 $MSFT 和 $AMZN 超大规模云计算厂商专用集成电路(ASIC)的杠杆式看涨押注。目前尚未有行动，也没有大幅上涨（尽管作为光子学玩家，今年反而下跌了12%），因为市场尚未对 Amazon Trainium 或 Microsoft Maia 产生炒作热度。
英文原文
I only have $LITE and $AAOI lol. But how I'd frame it is $LITE and $COHR are basically duopolies in photonics. With $LITE extra leveraged on Google TPU growth. Then there's $AAOI, extreme alpha + leverage call on $MSFT + $AMZN hyperscaler ASICs. No action yet, no major runup (even down 12% this year despite being a photonics player) because there's no investor hype about Amazon Trainium or Microsoft Maia (yet).
-
LITE因OCS垄断地位及供应瓶颈，受益于AI巨头需求。
$LITE 正处于“刚好合适”的黄金时刻，因为所有的 $GOOGL TPU、超大规模云厂商的 ASIC 以及 $NVDA Blackwell 架构都依赖它。目前他们因需求旺盛而面临供应瓶颈（类似于美光 Micron）。据我了解，光子学领域目前呈双寡头垄断格局（$COHR 和 Lite）。然而，在光电路交换(OCS)领域——这是 Google TPU 架构为实现更高性能所采用的技术——凭借专利组合和技术优势，$LITE 处于垄断地位。如果 $MSFT 和 $AMZN 最终也采用 OCS，那对 LITE 来说将是绝佳时刻。超大规模云厂商在磷化铟(InP)晶圆厂方面的垂直整合在短期内难度极大。未来几年内不太可能有人能取代他们（也许 2027-2028 年 $AVGO 的竞争会加剧），但就目前而言，他们只是受限于供应能力。
英文原文
$LITE is in the goldilocks moment where every single GOOGL TPU + hyperscaler ASIC + NVDA blackwell depends on it. Right now they're supply constrained from demand (similar to Micron). From what I understand, photonics right now is a duopoly ( $COHR and Lite). However, for OCS, which is what Google TPU architecture uses for better performance, $LITE is a monopoly from their patent portfolio + tech. If $MSFT, $AMZN end up adopting OCS, it's just a holy moment for LITE. Hyperscaler vertical intergration of InP fabs is just way too difficult near term. Nobody will likely replace them next few years at least (maybe 2027-2028 increasing competion from $AVGO) but as of now, they're just supply constrained right now.
-
对比LITE与AAOI，看好AAOI在MSFT和AMZN ASIC中的重估潜力。
“太晚了”。 $LITE 是一家市值 260 亿美元的公司在未来 $GOOGL TPU 部署以及 $NVDA / $AMZN 等其他 GPU/ASIC 的核心位置。 它占据 Google TPU v7 8-12% 的物料清单(BOM)份额。这就像说当 $NVDA 市值达到 1 万亿美元时它“太晚了”一样。 但如果你想要另一个参与者，那就是 $AAOI。 $MSFT 正在 Maia ASIC 中使用它们，并且很可能正在与 $AAOI 创建一种新架构。 它们还与 $AMZN 拥有 40 亿美元的认股权证+购买协议。 所以这更像是一个做多 $MSFT Maia + $AMZN Trainium ASIC 的玩法，但市场尚未真正关注这两家超大规模云服务商的 ASIC。 如果有任何关于这两者的利好消息，$AAOI 将极快地重估。 尽管据报道 $META 将购买大量产品，但每个人都跳上了 $GOOGL TPU 的顺风车。
英文原文
"Late". $LITE is a $26B MC in the center of every future $GOOGL TPU deployment, $NVDA/ $AMZN / other GPU/ASICs. 8-12% BOM share of every Google TPU v7. This is like saying $NVDA was "late" when price it was a $1T MC. But if you want want another player, $AAOI. $MSFT is using them for Maia ASICs and are likely creating a new architecture with $AAOI. They also have a $4B warrant + purchase agreement with $AMZN. So it's more of a long $MSFT Maia + $AMZN Trainium ASIC play but markets haven't really cared about those two hyperscaler ASICs yet. If there's any positive news about those two $AAOI will revalue extremely fast. Everyone jumped on the $GOOGL TPU bandwagon after $META reportedly was going to buy a bunch though.
-
Lumentum凭借AI光网络垄断地位，有望随巨头资本支出增长实现市值翻倍。
$LITE 的投资逻辑：AI 领域的隐形垄断。 Lumentum 今年迄今上涨 316%，但到 2027 年可能上涨 1000%+。 美光 ($300B) 或台积电 ($1.5T) 位于每个部署的 TPU/GPU 中心。 但 $LITE 也是如此，而其市值仅为 $26B。 在谷歌的每一个 TPU 中，$LITE 为其市值创造了惊人的利润。 这是因为它是光路交换 (OCS) + 光网络的标准。 它也存在于： - $NVDA Blackwell - $AMZN Trainium - 以及其他超大规模云服务商的 ASIC 中。 Lumentum 位于光子学芯片部署的“神圣三位一体”中。 对于每笔 TPU 资本支出，$LITE 占据 8-12%。对于每个 Nvidia GPU，$LITE 占据 ~2-3%（在 InnoLight 和其他公司之间分配，所以计算有点复杂）。 仅针对 NVDA GPU 部署的 BOM 粗略估算： NVIDIA Blackwell (GB200)： HBM 内存：~50–55%（SK 海力士主导，美光，三星） 逻辑（GPU 裸片）：~25-30%（$TSM 4NP） CoWoS 封装：~13-18% $TSM 光网络：~3–5%（InnoLight, Lumentum, Coherent） PCB/电源：5% 谷歌 TPU “Ironwood” TPU v7： HBM 内存：38-42% 三星 / SK 海力士 逻辑裸片：TSM ~28-33% 设计/I.O：8-10% 联发科 光网络：10-14%（$LITE 主要，$COHR 次要） 光交换：2-4% $LITE $LITE 估计的总集群份额：~8–12% 仅供参考，谷歌的“光”BOM 份额（8–12%）是一个异常值，源于其独特的光路交换 (OCS) 垄断。 粗略估算： 到 2027 年谷歌 TPU 支出 $40B。$LITE 占据 10%（30-40% 利润率），仅来自谷歌的自由现金流超过 $1.5B，仅其主要客户就带来 17 倍市盈率。（分析师在预测 TPU 支出增长时可能严重偏离。） 这甚至不包括其在 $AMZN Trainium、$NVDA Blackwell、$MSFT Maia 和其他芯片部署中的份额。 $LITE 目前位于每个 TPU/GPU 未来芯片部署的中心，并从中分得一杯羹。 唯一的 downside 是，虽然他们现在是明确的市场领导者，但 $AVGO 和 $COHR 很可能在 2027-2028 年准备好竞争。然而…… 人们说“$26B，历史新高，为什么现在买”。 这就是原因。他们参与了未来每一个部署的 TPU/GPU/ASIC。 如果谷歌 TPU 和其他芯片支出增加，且 LITE 从每个美元支出中占据 2-3%（来自 $NVDA, $AMZN, $MSFT）或 8-12%（来自 $GOOGL），$LITE 最终很容易超过 $60B+。
英文原文
The $LITE thesis: The hidden monopoly in the AI. Lumentum is up 316% YTD, but might be 1000%+ by 2027. Micron ($300B) or TSM ($1.5T) sit in the center of every TPU/GPU deployed. But same with $LITE, but it's a $26B MC. In Every, Single, TPU from Google, $LITE makes unbelievable amounts of profit for their marketcap. That's because it's the standard for Optical Circuit Switching (OCS) + optical networking. It's also in - $NVDA Blackwell -$AMZN Trainium - and other hyperscaler ASICs. Lumentum sits in the holy trinity of every single chip deployment for photonics. And for every TPU capex spent, $LITE takes 8-12%. For every Nvidia GPU, $LITE takes ~2-3% (split between Innolight and some others, so the math gets a bit complex). But some napkin math on NVDA GPU deployments alone for BOM: NVIDIA Blackwell (GB200): HBM memory: ~50–55% (SK Hynix (Lead), Micron, Samsung) Logic (GPU Die): ~25-30% ( $TSM 4NP) CoWoS Packaging: ~13-18% $TSM Optics/Network: ~3–5% (Innolight, Lumentum, Coherent) PCB/Power: 5% For Google TPIU "Ironwood" TPU v7: HBM Memory: 38-42% Samsung / SK Hynix Logic Die: TSM ~28-33% Design/I.O: 8-10% MediaTek Optical Network: 10-14% ( $LITE (primary), $COHR secondary) Optical Switch: 2-4% $LITE $LITE est. total cluster share: ~8–12% Just an FYI, Google's "Optical" BOM share (8–12%) is an anomaly due to their unique Optical Circuit Switch (OCS) monopoly. Just for some napkin math: $40B Google TPU spend by 2027. $LITE captures 10% (30-40% margins), $1.5B+ FCF from Google alone, 17x earnings from just their primary customer. (analysts are probably extremely off with projecting TPU spend scaling). Not even including their split from $AMZN Trainium, $NVDA Blackwell, $MSFT Maia, and other chip deployments. $LITE is in the center of every single TPU/GPU future chip deployment for now and takes a cut. The only downside is they're the clear market leader now, but $AVGO and $COHR are likely set up to compete by 2027-2028. However... People say "$26B, ATH, why are you buying now". This is the reason. They're involved in every future single TPU/GPU/ASIC deployed. $LITE could end up easily over $60B+ if Google TPUs, and other chip spend ramps up and LITE takes a 2-3% (from $NVDA, $AMZN, $MSFT) or 8-12% cut (from $GOOGL) for every single dollar spent.
-
作者认为AAOI和POET虽具十倍潜力但因执行风险未去风险化，不属于高确信度标的。
对我个人而言：当一家公司在根本上已“去风险化(de-risked)”并正从该点迈向规模化时，我才视其为“高确信度(high conviction)”标的。 例如，$NBIS 在与微软达成交易后。在光子学(photonics)领域，我个人更看好 $AAOI 而非 $POET。但它们都尚未“去风险化”。 正如你提到的，Poet 受益于 Celestial（已被 $MRVL 收购）。Marvell 基本上通过 Celestial 向超大规模云厂商(hyperscalers)出售 Poet 的中介层(interposer)，用于连接 TPU/Trainium。 但这中间隔着多层环节。 $AAOI 已经拥有与超大规模云厂商的直接合同，例如与 $AMZN 合作 Trainium，与 $MSFT 合作 Maia。此外还有类似 $ALAB 与亚马逊那样的股权协议。 然而，它们实现了完全垂直整合且工厂建在德克萨斯州。如果能让整个供应链在美国实现规模化并完美执行，它们有巨大的10倍增长潜力。 但由于相比 $COHR 规模较小，$AAOI 存在巨大的执行风险，因此不属于高确信度标的。$POET 同理，目前尚无让我完全确信的去风险化因素。 不过，Poet 和 AAOI 都是具备10倍潜力的光子学标的，这也是我持仓部分 $AAOI 的原因，但它们都不属于高确信度投资。
英文原文
For me personally: A stock is "high conviction" when they are fundamentally de-risked and are going from that point to scale. For example, $NBIS post-MSFT deal. I like $AAOI much better than $POET in the photonics space personally. But neither of them are "de-risked yet". So Poet as you mentioned, benefits from Celestial, which got acquired $MRVL. Marvell basically sells Poet's interposer via Celestial to hyperscalers for connecting tpus/trainium. But this is via multiple hops away. $AAOI already had direct hyperscaler contracts like $AMZN for Trainium + $MSFT for Maia. And equity agreements like what $ALAB has for Amazon. However, they're fully vertically integrated + built in Texas. They have a pretty big potential to 10X if they can get the whole supply chain to scale in America + perfect execution. But because they're at lower scale compared to $COHR, $AAOI has huge execution risk + it's not high-conviction. Same with $POET, there's no de-risking factor yet that would make me fully convinced. However, Poet and AAOI are both photonics plays that can 10x, which is why I have some $AAOI in my portfolio, but neither are high conviction.
-
感谢 LITE 提醒，并把它和 CRDO/ALAB 的连接层逻辑并列
感谢提醒 $LITE。 我之前没有意识到它对 $NVDA Blackwell、$GOOGL TPU 和 $AMZN Trainium 芯片有多关键。 离谱的是，它竟然处在每家 hyperscaler 的单 GPU/TPU/ASIC 核心环节里（这也是为什么它是高信念标的）。 至于 $CRDO，它和 $ALAB 类似，都是高 Nvidia 式利润率、三位数同比增长，以及依赖 hyperscaler 连接需求的逻辑。
英文原文
Thanks for the $LITE shoutout. I didn't realize how instrumental it was to $NVDA blackwell, $GOOGL TPUs, and $AMZN Trainium chips. Insane how it's part the core of single GPU/TPU/ASIC from every hyperscaler (hence why it's high-conviction) As for $CRDO, similar ground to $ALAB regarding high-Nvidia like margins, triple digit Y/Y growth, and hyperscaler dependencies for connectivity.
-
认为 Nebius 基本面没变，情绪转向且收购叙事增强
像 $NBIS 这样的公司，基本面没有任何变化（除了继续改善）。变化的只是市场情绪突然切换了。 但现在，Wedbush 的 Dan Ives 这样的人突然出来说：“Nebius 可能在 2026 年被 $GOOGL、$MSFT、$AMZN 这样的 hyperscaler 收购”，这极其利多。
英文原文
Nothings fundamentally changed (aside from improving) about companies like $NBIS. Aside from the sentiment flipping the switch. But suddenly we have people like Dan Ives from Wedbush going out and saying "Nebius likely gets acquired by a hyperscaler like $GOOGL, $MSFT, $AMZN " in 2026, which is incredibly bullish.
-
OpenAI 高估值融资改善 AI 数据中心与供应链风险偏好
最新消息：OpenAI 在完成 5000 亿美元估值、100 亿美元以上的 Amazon 轮融资后，正在以 7500 亿美元估值继续融资。 $CRWV：+15.85% $NBIS：+10.28% 有了这笔交易，像 CoreWeave 和 Oracle 这样的公司，在为 OpenAI 的资本开支需求建设容量时，对手方风险在结构上降低了，因为 OpenAI 的资产负债表更强了。 像 Nebius 这种算法上和行业龙头绑定的公司，也因此上涨。 从 Rocket Lab 到 Bitcoin，高 beta 资产普遍上涨。 最近因为日元套息交易解除，加上今天大量未平仓合约到期，市场出现了极端波动。 不过，AI 交易的基本面（尤其是 Micron 神级财报显示了极强内存需求）以及 Neocloud 的基本面，仍然比以往更好。
英文原文
Latest news: OpenAI is raising funds at a $750B valuation after their $10B+ Amazon round at $500B. $CRWV: +15.85% $NBIS: +10.28% With this deal, companies like Coreweave and Oracle structurally have less counterparty risk with OpenAI's stronger balance sheet to fund capex requirements. Companies like Nebius that are algorithmically tied to the sector leaders are up as a result. High-Beta Assets from Rocketlab to Bitcoin are up across the board. There was extreme volatility recently with the Yen carry trade unwinding + large open interest expiring today. However, the fundamentals of the AI trade (especially with Micron's godlike ER showing extreme memory demand), and Neoclouds remain better than ever.
-
Amazon 投资 OpenAI 对 AI 数据中心和 ASIC 供应链的连锁影响
最新：Amazon 的 100 亿美元 OpenAI 融资，以及 AI 供应链的涟漪效应。 $AMZN 将以 5000 亿美元以上估值向 OpenAI 投资 100 亿美元以上。 为什么这是 AI 股票的重大结构性变化： 1. AI 数据中心交易去风险：（$ORCL、$CRWV、$APLD、$CORZ） 结合前面提到的 SPEED 法案，影响 Neocloud 的主要问题是： 1. 数据中心延迟和递延收入 2. 不可持续的 CapEx -> 没有 FCF 3. OpenAI 传染/积压订单风险 SPEED 法案直接处理了第 1 和第 2 点，但没有处理 OpenAI 相关的第 3 点。 影响 Oracle、CoreWeave 等最大 Neocloud/数据中心提供商的主要担忧，是它们为 OpenAI 这个对手方投入了巨额 capex，而 OpenAI 本身没有足够资金承诺这些资本开支。 但现在，随着 Amazon 新融资，第 3 点开始被解决。 - 新增 100 亿美元资金，加上 Amazon 的资产负债表支持 ChatGPT 的创造者，OpenAI 对 Oracle 和 CoreWeave 的早期承诺开始获得兜底。 - 下游影响：这直接降低了 $CRWV 和 $ORCL 的风险，因为它们在为 OpenAI 建设容量。再往下两层，依赖 CoreWeave 作为租户的 $APLD 或 $CORZ 也会被去风险。 由于 CoreWeave 和 Oracle 被视为“行业龙头”，这会立刻改变整个 Neocloud 板块的情绪，包括 $NBIS、$IREN、$CIFR、$WULF 等。 2. Hyperscaler AI capex 浪潮（$AMZN、$MRVL、Alchip）： 我们之前看到，$AVGO ASIC backlog 之后，市场开始担心 hyperscaler 支出减弱，很多相关玩家因此大跌。 但这笔交易的关键条件之一，是 OpenAI 采用 Amazon 自研 Trainium ASIC。这代表非 Nvidia 集群会激进扩张。 - 设计与 IP：直接利好 Marvell（$MRVL）和 Alchip 等 ASIC 设计伙伴。 - 定制硅集群需要大量光互连和 HBM。这会为光子（$AAOI、$LITE、$COHR）和内存（$MU、SK 海力士）创造新的 capex 超级周期。 - 代工厂，比如 $TSM。 以及更多参与 hyperscaler ASIC 建设的相关公司。 唯一输家？Nvidia（$NVDA）。 Amazon 正在成功用它庞大的资产负债表，迫使领先 LLM 摆脱对 H100/Blackwell 的依赖，同时推动整条 AI 供应链。 核心结论是，AI 交易由世界上最富有的公司资助，比如 $AMZN；OpenAI 也正在证明，它可以通过出让股权来扩大资产负债表，以满足需求。 做多 AI 板块。
英文原文
Just In: Amazon's $10B OpenAI Funding and The AI Supply Chain Ripple Effect. $AMZN is set to invest $10B+ in OpenAI at a $500B+ valuation Why this is a MASSIVE structural shift for AI stocks: 1. De-Risking the AI DC trade: ( $ORCL, $CRWV, $APLD, $CORZ ) With the SPEED Bill mentioned earlier, the main issues affecting Neoclouds were: 1. DC Delays & Deferred Revenue 2. Unsustainable CapEx → No FCF 3. OpenAI Contagion/Backlog. the Speed bill directly addresses #1 and #2. But not #3 with OpenAI. The main fears affecting the biggest Neocloud/Datanceter providers like Oracle, Coreweave was their immense capex spend for a counterparty (OpenAI) that doesn't have the funding to commit to it's capex spend. But now, #3 is starting to be addressed with the new Amazon funding. - With a fresh $10 Billion and Amazon’s balance sheet backing the creator of ChatGPT, OpenAI's early commitments to Oracle and CoreWeave are now starting to be backstopped. - Downstream Impact: This directly derisks companies like $CRWV and $ORCL, who are building capacity for OpenAI. And by two hops, companies like $APLD or $CORZ that rely on Coreweave as a tenant. As Coreweave and Oracle are seen as the "sector leaders" this immediately changes sentiment across the whole Neocloud sector from $NBIS, $IREN, $CIFR, $WULF and others as well. 2. The HyperScaler AI capex wave ( $AMZN, $MRVL, Alchip): We've seen fears after $AVGO ASIC backlog about hyperscaler spending waning. And many related players tanked on the news. However, a key condition of this deal is OpenAI’s adoption of Amazon’s proprietary Trainium ASICs. This signals an aggressive scaling of non-Nvidia clusters. - Design & IP: Direct benefit to ASIC design partners like Marvell ( $MRVL ) and Alchip. - Custom silicon clusters require massive optical interconnects and HBM. This creates a new capex supercycle for photonics ( $AAOI, $LITE, $COHR ) and memory ( $MU, SK Hynix). - Foundries such as $TSM. and many more related companies involved in the buildout of hyperscaler ASICs. The only loser? Nvidia ( $NVDA ). Amazon is successfully using its massive balance sheet to force the leading LLM to diversify away from H100/Blackwell dependence and boosting the whole AI supply chain alongside it. The main takeaway is that the AI trade is funded by the richest companies in the world, such as $AMZN, and OpenAI is showing it can scale up its balance sheet to meet requirements by trading off equity. Go long on the AI sector.
-
长文拆解 SPEED Act 对 Neocloud 和 AI 数据中心的去风险作用
刚刚，SPEED Act 在众议院取得进展。 这是今年 Neocloud 板块（$NBIS、$CRWV、$IREN）最大的单一去风险法案/事件。 下面是原因和梳理： 美国政府正准备以美国对中国的国家安全为理由，支持从 Oracle 到 Nebius 的 AI 数据中心建设。 Oracle 和 CoreWeave 最近跌了 40%+（也把 $NBIS 从 140 美元打到 79 美元、$IREN 从 80 美元打到 35 美元、$CIFR 从 24 美元打到 14 美元），核心恐惧有三点： 1. 数据中心延迟和递延收入 2. 不可持续的 CapEx -> 没有 FCF 3. OpenAI 传染/积压订单风险 SPEED Act 和美国政府干预，修复了数据中心建设延迟的空头论点，也解决了利用率滞后带来的盈利问题（利润率）。 #1 数据中心延迟和递延收入 空头 thesis：多年许可延迟（NEPA、输电）把高价值合同变成了递延收入风险。$CRWV 明确把供应商延迟作为下调指引的原因，并在财报后大跌，因为相当一部分收入被推迟到 2026 年 Q1/Q2。 如果 SPEED Act 通过，alpha 在这里： 强制速度和诉讼保护。 - SPEED Act 要求联邦环境和监管审查遵守严格且不可协商的截止日期，通常是 1-2 年。 - 诉讼保护：该法案大幅缩短针对已批准许可提起诉讼的时效，比如缩到 150 天，并指示法院即使许可被临时挑战，也要允许数据中心建设继续推进。 结果：从签约到“GPU 上架 -> 收入流入”的时间线，被压缩了，并由联邦政府在政治上去风险。 递延收入被提前确认，修复了 $CRWV、$APLD 和 Neocloud 板块此前面临的延迟、递延利润/收入问题。 #2 不可持续 CapEx -> 无法从资产变现出 FCF 空头 thesis：公司在 GPU 和建设上花费数十亿美元（$ORCL 的 capex 很巨大），但从购买 GPU 到变现之间的利用率拖累严重影响盈利能力和 FCF。 这也大幅影响 AI Cloud 供应商，因为它们缺乏足够电力来把 GPU/capex 变现。 公司因此面临巨大的减记风险，也就是利用率拖累：GPU 闲置时，折旧和通电成本的时钟仍在跑。 这个拖累对数据中心部门盈利能力影响巨大，The Information 关于 $ORCL AI 利润率极薄的报道也提到过这一点。 SPEED Act 和美国政府干预会直接降低 CapEx 风险，因为速度上的立法要求（修复点 #1）实际上保证电力基础设施会在一个确定且较短的时间线内到位。 这种确定性让 $NBIS、$CRWV 和 $IREN 可以更有信心安排数十亿美元 GPU 的采购和部署，知道资产到货后就能立即开始变现，同时也通过降低利用率拖累来加快 FCF。 这种结构性变化会流向整个行业。它会立即降低主要 AI Cloud 供应商（$AMZN、$MSFT、$ORCL）的风险，因为它们现在能更确定地保证产能；同时也会保证 Colo/Infra/Energy 提供商（$CIFR、$WULF 等）的需求，因为它们的核心业务就是提供电力容量。 关于 capex -> FCF 以及建设延迟时间线的空头论点，现在已经被 SPEED Act 直接处理。 现在美国政府准备加速 $NBIS、$CRWV 和 $IREN 这类公司，因为 AI 数据中心已经被放到美国和中国 AI 国家安全竞争的前线。 它能否在众议院通过，是每个投资者都该关注的事。但如果通过，这会是 Neocloud / AI 数据中心建设里最大的、尚未被充分讨论的顺风之一。
英文原文
Just now, the SPEED Act ADVANCES in the House. This is the single biggest de-risking bill/event for the Neocloud sector ( $NBIS, $CRWV, $IREN) this year. Here's why and a rundown: The U.S. GOVERNMENT is set to support the AI data center buildout from Oracle down to Nebius on national security grounds for US vs China. Oracle and CoreWeave recently dropped 40%+ (tanking $NBIS $140 -> $79, $IREN $80 -> 35, $CIFR, $24 -> $14 as well) on three core fears: 1. DC Delays & Deferred Revenue 2. Unsustainable CapEx → No FCF 3. OpenAI Contagion/Backlog. The Speed Act and US Government intervention fixes bear-case points for data center buildout delays and addresses utilization lag profitability issues (margins). #1 DC Delays & Deferred Revenue Bear Thesis: Multi-year permitting delays (NEPA, transmission) turned high-value contracts into deferred revenue risk. $CRWV explicitly cited vendor delays for lowering guidance and tanked on earnings shifting a large portion of revenue to from Q1 Q2 2026. The alpha if the Speed Act passes: Mandatory Speed and Litigation Shields. - The Speed Act mandates strict, non-negotiable deadlines (often 1-2 years) for federal environmental and regulatory reviews. - The Litigation Shield: The bill drastically shortens the statute of limitations for filing lawsuits against approved permits (e.g., to 150 days) and instructs courts to allow DC buildout to continue even if a permit is temporarily challenged). The Result: The timeline from contract signing to "GPUs on racks -> revenue flowing" is now compressed and politically de-risked by the Federal Government. Deferred revenue is pulled forward and fixes delays and deferred profitability/revenue that plagued $CRWV, $APLD, and the Neocloud sector. #2: Unsustainable CapEx -> No FCF from monetizing the assets Bear Thesis: Companies were spending billions on GPUs and construction ( $ORCL's capex is massive) with utilization drag (from the point of purchasing the GPUs to monetization) largely affecting profitability and FCF. This also largely affects AI Cloud vendors (lacking power to turn monetize the GPUs/capex). Again This forced companies to take a massive write-down risk due to Utilization Drag (the time the GPU sits idle while the clock runs on depreciation/power-up). This drag is HUGE for profitability on DC segments, as cited in The Information reports on $ORCL's razor-thin AI margins. The SPEED Act and the US Government intervention directly de-risks CapEx as the legislative mandate for speed (Fix #1) effectively guarantees that power infrastructure will arrive within a defined, short timeline. This certainty allows $NBIS, $CRWV, and $IREN to time the purchase and deployment of billions in GPUs with high confidence that the assets will begin monetizing immediately upon arrival as well as accelerates FCF from reducing utilization drags. This structural change flows down the entire industry. It instantly de-risks the major AI Cloud vendors ($AMZN, $MSFT, $ORCL) who can now guarantee their capacity, and it guarantees demand for the Colo/Infra/Energy providers ( $CIFR, $WULF, and others) whose core business is supplying that power capacity. The bear case on capex -> FCF + buildout delay timeline has now directly addressed with the Speed ACT. Now the US government is set accelerate companies like $NBIS, $CRWV, and $IREN as AI Datacenters is now placed on the forefront of the AI national security battle between the United States and China. Whether it passes legislation in the House is what every investor should be watching, but if does, this is one of the largest (not-talked about) tailwinds for the Neocloud /AI Decenter buildout.
-
认为 Neocloud 板块在长期合同去风险后仍被错杀
同意。 像 $CIFR 这样的公司，现在交易得仿佛 $AMZN、$GOOGL 的合同从未发生过。 在整个 Neocloud 板块里，我们看到的是一场崩盘，尽管 Mag7 合同已经让全板块去风险，并提供了长期收入可见性。 $ORCL、$CRWV 以及一些个股因为 OpenAI 敞口而下跌是有道理的，但它把板块其他公司也一起拖下去了。
英文原文
Agreed. Companies like $CIFR are trading like the $AMZN, $GOOGL deals never happened. In the whole Neocloud sector, we’re seeing a crash despite de-risking and long term revenue visibility across the board with Mag7 contracts. $ORCL, $CRWV, and some individual names made sense given OpenAI exposure but it just dragged the rest of the sector alongside it.
-
博通财报后AI板块抛售是误解造成的买入机会，新云板块中OpenAI依赖股除外
博通[$AVGO]业绩及其对AI板块的影响，如$LIITE和$NBIS： 博通的业绩"双重超预期"，营收$180.2亿(+28% Y/Y)，EPS $1.95，超出共识预期。 但AVGO下跌-11.64%，并拖累了整个AI板块。 这是买入机会吗？ 是的。 博通被视为超大规模云厂商ASIC代理增长的代表，因为亚马逊$AMZN Trainium、微软$MSFT Maia、尤其是谷歌$GOOGL TPU V7 Ironwood都通过它进行规模化部署。 而像$ALAB(-13.2%)、$CRDO(-5.11%)、$LITE(-12.23%)、$TSM(-3.71%)、$COHR(-9.25%)等公司都是TPU/ASIC建设以及博通作为公司的直接受益者。 博通下跌有三个原因，市场下跌有一个原因： 就博通而言，有一些小问题，如税率变化影响EPS模型，或因更多定制AI芯片而非更高利润率软件导致的"利润率压缩"，但这只是会计处理框架问题。（类似于$META在一次性税收后最初的下跌） 对博通和整体市场而言，是积压订单预期问题。以上所有引用的内容与ASIC市场预期增长相比都是小问题。 博通披露未来18个月$730亿的AI积压订单。而有关Anthropic和META购买价值数十亿$GOOGL TPU的传言，人们隐含预期是$800亿+。 然而，这次抛售是由算法和短期AI泡沫情绪驱动的价格错位，而非基本面破裂。 这条积压订单引用是确认订单的最低合同底线。谷歌$GOOGL、亚马逊$AMZN等公司可能会继续增加ASIC订单，而市场未能辨别这一细微差别。 分析师预期营收转化会更加前置，Q4之后积压订单应该会减少，这给出了2026年更高的可能范围$550-600亿+，而非$730亿预期中的$500亿。 TLDR：关于超大规模云厂商ASIC加速以与$NVDA依赖竞争这一论点没有改变。$AVGO和其他如$COHR、Sk Hynix、$MU、$VRT和$LITE都将受益。 这不是关于营收积压订单的最佳消息，但由于交付周期/订单周期和最低底线而被误解。 如果非要说什么，较低的超级云厂商ASIC需求对$NVDA及其生态系统是有益的，但我们也看到$CRWV、$SMCI、$NBIS和$NVDA GPU/DC计算生态系统今天都从盲目抛售中下跌超过5%，尽管存在负相关性。 这又是典型的"AI泡沫"周期因误解而再次来袭。AI股票普遍下跌10-12%的恐慌是一个绝佳的买入机会。 （该推文引用了@aleabitoreddit的推文，引用内容仅供理解语境）: 甲骨文[$ORCL]业绩及其对新云板块如$NBIS和$IREN的影响： 甲骨文报告EPS超预期且积压订单创纪录，但盘后下跌12%。 甲骨文较9月11日高点下跌39.8%，并拖累整个板块。 原因如下： 这次抛售不仅仅是对边际营收miss的反应，而是算法做空和投资者对AI资本支出周期可持续性以及该板块主要租户信用资质的不看好： OpenAI。 甲骨文宣布2026年资本支出增加$150亿至近$500亿，这与报道的与OpenAI $3000亿合作伙伴关系密不可分。 最初，OpenAI是前沿LLM，对甲骨文、Coreweave等公司有前景良好的资本支出承诺，促成了该板块的初始重新定价。 然而，随着超过$1万亿的义务以及Anthropic、Gemini、XAI等竞争对手的增加，市场严重质疑甲骨文、Coreweave等公司是否为无法从运营现金流履行其义务的租户在建设。 我们看到市场有效地发出信号：甲骨文正在为OpenAI创造一种不可持续的债务驱动型"供应商融资"，而OpenAI无法履行其承诺。 因此下跌是理性的：这次抛售是由信用风险和资本密集度的理性重新定价所驱动。 OpenAI融资担忧是合理的：OpenAI缺乏资金履行合同的假设得到其营收($130亿)与义务($600亿/年)之间明显不匹配的支撑。 信用担忧是真实的：甲骨文CDS利差扩大显示"信用事件"降级或违约的概率上升。 此外，我们看到这在新云板块引发传染效应，$NBIS从$140跌至$90s，$IREN从$80跌至$40s，$CIFR从$24s跌至$17s。 但这对$WULF、$NBIS、$IREN等新云公司来说是买入机会吗？ 是的。 这对$ORCL来说是好的买入机会吗？ 不是。 前瞻展望： $ORCL（很大一部分）、$CRWV（25%积压订单）是两个主要依赖OpenAI的公司，这一叙事可能因OpenAI的融资活动而瞬间翻转（+30%+变化）。 如果OpenAI在2026年以高估值超额认购IPO，且其新GPT模型击败Gemini/Claude，我们可以看到这种改变。 然而，许多其他公司与OpenAI无关。新云板块的原始论点是Mag7资本支出从其现金牛业务（Azure、AWS、GCP）向下流向：$NBIS、$IREN、$CIFR、$WULF等。 但随着最大玩家($ORCL、$CRWV)下跌，这些公司算法性地拖累了整个板块。 如果你看各家公司，$CIFR和$WULF由$GOOGL兜底，$IREN/$NBIS由$MSFT资助。 这些是与超大规模云厂商/Mag7的锁定合同积压订单，而非OpenAI。 这种因误解风险而导致的不理性抛售为新云板块提供了绝佳的买入机会，但不是与OpenAI相关的公司如$ORCL和$CRWV。
英文原文
Broadcom [ $AVGO ] earnings results and its effect on the AI sector like $LITE and $NBIS: Broadcom's ER was "double beat" with $18.02B revenue (+28% Y/Y) and $1.95 EPS, beating consensus. But AVGO dropped -11.64% and brought down the AI sector. Is this a buying opportunity? Yes. Broadcom is seen as a hyperscaler ASIC proxy growth as companies like $AMZN Trainium, $MSFT Maia, and most importantly $GOOGL TPU V7 Ironwood are scaled through it. And by proxy companies like $ALAB (-13.2%), $CRDO (-5.11%), $LITE (-12.23%), $TSM (-3.71%), $COHR (-9.25%), and are direct beneficiaries of the TPU/Asic buildout and Broadcom as a company. There's three reasons why Broadcom fell and one why the market fell: For Broadcom, there's minor things such as tax rate changing EPS models or "margin compression" from accounting from just more custom AI chips than higher-margin software, but this is just accounting framing. (Similar to how $META dropped initially on one-time tax post-ER) For both Broadcom general market, it was backlog expectations. Everything cited above is all minor compared to expected growth of ASIC markets. Broadcom cited $73B in AI backlog for the next 18 months. And rumors of Antrophic and META buying billions of $GOOGL TPUs, people were implicitly expecting $80B+. However, the selloff represents a dislocation in price driven by algorithms and short-term AI Bubble sentiment rather than a fundamental breakage. This backlog quote was the MINIMUM CONTRACTUAL FLOOR of confirmed orders. Companies like $GOOGL, $AMZN, will likely continue ramping up ASIC orders and the market failed to discern this nuance. Analysts are expecting revenue conversion to be more front loaded, and that there should be less backlog beyond Q4 given the cycles, which gives a higher likely range of $55-60B+ for 2026 rather than $50B expected of the $73B. TLDR: The thesis regarding hyperscaler ASIC ramp to compete vs $NVDA dependency has not changed. $AVGO and other players like $COHR, Sk Hynix, $MU, $VRT, and $LITE all stand to benefit. It's not the best news regarding the revenue backlog, but it's misunderstood due to lead-time/order cycles and minimum floors. If anything, lower hyperscaler ASIC demand is beneficial to $NVDA and their ecosystem, but we've also seen $CRWV, $SMCI, $NBIS and $NVDA GPU/DC compute ecosystem drop over 5%+ today from an indiscriminate sell-off despite inverse correlation. This is just the typical "AI Bubble" cycle hitting again from misunderstanding. The widespread panic of AI stocks dropping 10-12% is a great buying opportunity.
-
看好$SNAP因降本增效及AI收入，预计2026-27年有100%+涨幅。
$SNAP 在 7.64 美元时被极度低估。 -> Snapchat 通过转向 GCP 削减运营支出，并将其转化为收入。 -> Perplexity 交易在现有收入基础上增加了 4 亿美元股权/现金。 预计这些变化带来的净自由现金流(Net FCF)在 2026 年底至 2027 年初将达到 +6.3 亿美元（+1.9 亿运营云支出节省，+5.4 亿内存收入，-1 亿存储成本）。 CEO 出售 1000 万美元股票并不重要，尤其是当有 5 亿美元回购时。 $SNAP 不需要成为 $META，它永远不会是。 在 130 亿美元市值下，它只需要削减成本并改善来自其 15-20 亿美元季度收入（~55% 毛利率）的自由现金流。 那么即使用户增长放缓或轻微下降，它也将获得巨大的重估。 目前税务收割(tax harvesting)可能在未来两周结束。但如果你能等待整整一年，我有信心一旦计入 GCP 运营支出削减和 AI 收入流，Snap 将大幅跑赢任何指数。 进入 2026-2027 年，这里有 100%+ 的上行空间。
英文原文
$SNAP is incredibly undervalued at $7.64. -> Snapchat cutting opex costs with GCP and converting that to revenue. -> Perplexity deal adding $400m in equity/cash on top of existing revenue. Net FCF Is estimate from these changes is +$630M (+$190M opex cloud, memory revenue +540M, storage cogs -$100m) late 2026, early 2027. The CEO selling $10M in shares is not material, especially when there's $500M in buybacks. $SNAP doesn't need to be $META, it never will. At a $13B marketcap, it just needs to cut costs and improve FCF from its $1.5B-$2B quarterly revenue (~55% gross profit margins). Then it will be re-rated immensely, even if user counts slows or slightly drops. Right now tax harvesting is likely ending next two weeks. But if you can wait a whole year, I'm confident Snap will strongly outperform any index once we factor in GCP opex cuts and AI revenue streams. 100%%+ upside here going into 2026 - early 2027.
-
博主披露核心及轮动持仓，并解释转向方向性评论的原因。
核心持仓是高确信度的多头：$BTC、$RKLB、$HOOD、$NBIS、$ALAB、$TSM。可能将 $LITE 和 $CRCL 移入上述核心多头组合，但它们是我正在建仓的新头寸。然后是短期至中期混合持仓，如 $SNAP、$CIFR、$RDDT、$SMCI、$HIMS、$TE、$LTC、$KRUS、AMKR、$LITE、$FLY、$WLAC、$META、$AMZN、$TTD 以及现在的 $AAOI 等。我在短期至中期持仓之间进行大量轮动。我以前发布更多关于日内交易的内容，但最终在这里获得了太多粉丝，所以想转向方向性评论。发布头寸更新很难，因为我喜欢解释我这样做的原因！我记得在 $IREN 约 50-60 美元时卖出，结果在接下来的三周里收到了一堆恶评 lol
英文原文
Core portfolio is high conviction longs: $BTC, $RKLB, $HOOD, $NBIS, $ALAB, $TSM Probably moving $LITE and $CRCL to the core long port above, but they’re newer positions that I’m building up. Then short-mid term mix like $SNAP, $CIFR, $RDDT, $SMCI, $HIMS, $TE, $LTC, $KRUS, AMKR, $LITE, $FLY, $WLAC, $META, $AMZN, $TTD, and now $AAOI etc. I rotate between short-medium term holds A LOT. I used to post more day trading stuff but I ended up getting too many followers here, so wanted to switch to directional commentary. It’s hard to post position updates because I like to explain why I do things! I remember selling $IREN around $50-$60 or something and just got a bunch of hate comments for the next three weeks lol
-
看好AAOI中期潜力，因ASIC生态关键地位及美国制造优势，但暂小仓位以控风险。
是的，我非常非常喜欢 $LITE，但它的高估值溢价已经被市场充分定价。 我仅在 $AAOI 上持有相对较小的仓位，主要是出于执行风险的考虑：万一超大规模云厂商取消采购订单，或者面临捷普(Jabil)的竞争。 话虽如此，它目前被市场当作一家小型组件制造商来定价，但实际上它受益于微软(Microsoft) Maia 和亚马逊(Amazon) Trainium 的 AI 芯片量产，且“美国制造”是其强劲驱动力，因为它在超大规模云厂商的专用集成电路(ASIC)生态系统中扮演着关键角色。 这只是一项基于定性分析的投资，依据诸如 30-35% 的毛利率等粗略估算，以及假设其因 2026 年超大规模云厂商的放量而超越市场预期。 不过，其市值足够小，足以让我将其作为中期投资标的。
英文原文
Yeah I really, really like $LITE but it’s already priced in with a premium. I only have relatively small positions in $AAOI though just because of execution risk in case a hyperscaler drops purchase orders + competition from Jabil. That being said it’s priced as a small cap component manufacturer but has AI $MSFT Maia $AMZN trainium ramp and made in America as a great driver since it plays a critical part in the hyperscaler ASIC ecosystem. This is just a qualitative investment, off rough estimates like 30-35% margins and assumptions of beating projections due to hyperscaler ramp in 2026. The marketcap is small enough to make me an investor for medium term though.
-
解析Mag7 ASIC供应链，看好$AAOI因$MSFT散热互连修复及$AMZN订单带来的增长。
不客气！如果我们看Mag7的ASIC，$GOOGL TPU v7 Ironwood的供应商在$META交易后曾有一波上涨。 $AMZN Trainium即将像$GOOGL那样开始放量，近期超大规模云厂商$AAOI的订单很可能来自他们。 $AAOI的超额收益逻辑在于：$MSFT Maia 200的订单更多是研发支出，且他们推迟了路线图（可能由于散热和互连问题，例如连接10万颗芯片而不熔化），标准线缆无法用于Maia 200。 但这对$AAOI是利好，因为$MSFT很可能使用他们来解决这一问题，且所有这些ASIC支出将发生在2026-2027年，加上远期收入指引的大幅超预期。 但$AMZN提供了约40亿美元的安全垫以降低风险。
英文原文
NP! If we look at mag7 ASICs, $GOOGL TPU v7 Ironwood suppliers had a run after the $META deal. $AMZN Trainium is going to ramp up soon like $GOOGL and the recent hyperscaler $AAOI order was likely from them. The alpha for $AAOI is that $MSFT Maia 200 orders were more of R&D spend and they delayed their roadmap (likely due to thermal and interconnect issues eg. 100,000 chips to connect without melting) and standard cabling wasn't working for Maia 200. But this is bullish for $AAOI since $MSFT likely uses them for this fix and all of this ASIC spend will be occur in 2026-2027 + huge beat on forward revenue projections. But $AMZN is providing the nice $4B floor for de-risking.
-
分析AAOI依赖Mag7的垂直整合优势，已建仓并看好LITE。
我原本在研究一篇来自 Semianalysis 的文章中关于 $AMZN Trainium 基准测试以及亚马逊供应链的内容。鉴于一些朋友（如 @yianisz）之前曾对此发帖讨论，我决定深入挖掘。亚马逊在扩大 Trainium/Inferentia 集群时提供了约 40 亿美元的支出作为后盾，加上 $MSFT 在 ASIC 集群上的投入，随着行业向 800G 升级，我们很可能会看到更多超大规模云厂商客户。对于一家市值 25 亿美元的公司来说，如此依赖 Mag7 的情况实属罕见 lol。下行风险与积极因素并存：由于激光器在德克萨斯州内置生产，该公司实现了垂直整合（更高利润率）。我看好未来巨大的量增需求，但：关键在于执行，他们必须在不搞砸的情况下建成 800G/1.6T 光收发器。我已经全仓持有 $AAOI，出于风险管理考虑，仓位规模相对较小适中（毕竟它仍是 25 亿市值的小盘股）。我依然更看好 $LITE，但鉴于其本月已上涨 46%，很难进行补仓。
英文原文
I was actually looking into $AMZN Trainium benchmarks from a semianalysis post and at Amazon's supply chain. And some friends like @yianisz post about it earlier, so I decided to look deeper at it. But it has ~$4B spend from $AMZN as a backstop as they ramp up Trainium/Inferentia clusters, spend from $MSFT ASIC clusters, and we'll likely see more hyperscaler clients as the industry moves to 800g ramp. Mag7 dependency is incredibly rare coming from a $2.5B marketcap company lol. Downside + positive: Is vertically integrated (higher-margins), since it's lasers are builtin Texas. I see just incredible volume demand, moving forward but: the whole name of the game is execution since they actually have to build out the 800G/1.6T transceivers without messing up. I'm fully invested already in $AAOI already with relatively small-moderate position sizing for risk management (it's still a small cap with $2.5B MC). I still like $LITE a lot more, but a bit hard to cost average up on that when it's up 46% this month.
-
作者认为因拥有直接大客户及更低估值，$AAOI 优于 $POET。
我研究过 $POET，但相比 $AAOI 我更喜欢后者，因为 AAOI 已经拥有 $MSFT 和 $AMZN 等直接超大规模客户(hyperscaler customers)。$POET 旗下的 Celestial 已被 $MRVL 收购，Celestial 使用 Poet 的中介层(interposer)来连接 TPU/Trainium，而 Marvell 将其出售给超大规模客户。但这并不等同于通过直接来自 $AMZN 和 $MSFT 的大批量订单以及类似 Amazon 与 $ALAB 的股权投资协议来实现的“几步之遥”的去风险(de-risking)。Amazon 在 $AAOI 中拥有类似 $ALAB 的财务激励，但 Astera 已经是家市值 300 亿美元的公司，而 AAOI 的市值仅为 25 亿美元。
英文原文
I've looked into $POET but didn't like it as much compared to $AAOI given AAOI already has direct hyperscaler customers with $MSFT and $AMZN. $POET had Celestial that got acquired $MRVL, where Celestial used poet’s interposer to connect tpus/trainium, and Marvell sells that to hyperscalers. But it's not quite the same de-risking few hops away as direct massive volume orders from $AMZN and $MSFT and equity agreements that Amazon has like they did with $ALAB. Amazon has a financial incentive in $AAOI like $ALAB, but Astera is already a $30B company, while AAOI is a $2.5B MC.
-
看好$AAOI，因其在AI光互联中的核心地位及美国本土制造优势，认为其被严重低估。
我买入 $ALAB、$NBIS、$TSM 和 $LITE，是因为“七巨头”(Mag7) 的收入正流向这些公司。 $LITE 的独特之处在于其在 GOOGL TPU v7、AMZN Trainium v3/4 和 NVDA Blackwell 中的角色。 但我发现了一个新标的。 一家市值低于 30 亿美元的小盘股，符合我的投资逻辑： 名称 - $AAOI 一家小型光子学玩家，也是我持仓的两家光子学公司之一： 1. Lumentum ($LITE) 在每一款芯片的部署/爬坡中都占据独特地位，因为其光电路交换(Optical Circuit Switching) 技术被用于 Blackwell、Trainium 和 Ironwood，作为一种“横向扩展”(scale-across) 型技术。 无论超大规模云厂商选择 ASIC 还是 GPU，$LITE 都能胜出，因为它处于核心位置。 2. Applied Optoelectronics ($AAOI) 更多涉及“纵向扩展”(scale-out) 连接，针对 Trainium、Maia 等定制 ASIC 集群，通过 400G 和 800G 光缆及光收发器实现。 鉴于 $AAOI 在 AWS（Trainium 的大客户）和 MSFT Maia ASIC 中的角色，它也能无论何种情况都胜出。 行业正经历由向 800G 速度迁移驱动的“超级周期”，而 AAOI 正处于中心位置。 此外，AAOI 具有独特的地缘政治角度： 美国优先。 与许多无晶圆厂(fabless)并将生产外包到台湾的公司不同，Applied Optoelectronics 在德克萨斯州自行制造激光器。 美国超大规模云厂商（特别是 Amazon 和 Microsoft）正在积极减少对关键基础设施中国供应链的依赖，这有助于 AAOI 实现本土激光器制造。 我们还看到一家“主要超大规模云厂商”对其 800G 数据中心光收发器下达了另一笔巨额订单。 但 AAOI 的交易表现像一家困境公司；然而，其亚马逊权证协议隐含的收入为 2025 年 Q4 和 2026 财年的 800G 爬坡创造了非对称的风险/回报特征。鉴于其在 AI 建设中的角色及现有的超大规模云厂商合同，AAOI 似乎结构性被低估。 市场终于开始重估 $LITE，但感觉 $AAOI 才刚刚开始，鉴于其小市值以及作为超大规模云厂商 ASIC 集群关键玩家和“美国制造”的独特角度。 市场目前正在对光子学玩家进行大幅重估并赋予高溢价，但 AAOI 今年仅上涨 2.20%，似乎蓄势待发。
英文原文
I entered $ALAB, $NBIS, $TSM, and $LITE because of Mag7 funneling revenue numbers into them. Lite uniquely because of its role in GOOGL TPU v7, AMZN Trainium v3/4, and NVDA Blackwell. But there's a new one I found out about. A small cap <$3B player that fits the thesis: Name - $AAOI A small cap photonics player, and one of the two photonics players I'm invested in: 1. Lumentum is uniquely positioned in every single supply chip deployment/ramp, as the Optical Circuit Switching technology is used in Blackwell, Trainium, and Ironwood as a "scale-across" type technology. $LITE wins no matter what. Hyperscaler ASIC vs. GPU as it's in the center of it all. 2. Applied Optoelectronics is more of the "scale-out" connectivity, for custom ASIC clusters like Trainium, Maia through 400G and 800G optical cables and transceivers. $AAOI wins no matter what as well given its role with AWS as a whale client for Trainium, and with MSFT Maia ASICs. The industry is going through a "supercycle" driven by the migration to 800G speeds and AAOI is in the center of it. On top of that, AAOI plays unique geopolitical angle, America first. Unlike many other companies that are fabless and export production elsewhere to Taiwan, Applied Optoelectronics makes their own lasers in Texas. US hyperscalers (specifically Amazon and Microsoft) are aggressively reducing reliance on Chinese supply chains for critical infrastructure and that helps AAOI’s ability to manufacture lasers at home. And we've seen another huge volume order from a "major hyperscaler" on its 800G data center transceivers. But AAOI trades like a distressed company; however, the implied revenue of its Amazon warrant agreement creates an asymmetric risk/reward profile on its 800G ramp in Q4 2025 and FY2026. AAOI seems structurally undervalued, given its role in the AI buildout and existing hyperscaler contracts. The markets are finally catching up to LITE, but it feels $AAOI is yet to begin, given its small market cap size but unique angle of a critical player to hyperscaler ASIC clusters and Made in America. The market is currently re-rating heavily with photonics players and assigning a heavy premium, yet AAOI is only up 2.20% this year and seems like it's just about to begin.
-
Neocloud龙头承压，但获超大规模云商背书的个股是理想买点。
我预计Neocloud股票今天会表现不佳，因为$ORCL/$CRWV是“板块龙头”。 但这将是该板块个股的理想买入机会： 因为市场忽略了$ORCL/$CRWV高负债扩张的主要交易对手是OpenAI这一细微差别。 但该板块的其他公司如$NBIS、$CIFR、$WULF则由AAA评级的超大规模云服务商（$META/$META/$GOOGL/$AMZN）的现金牛业务提供背书。
英文原文
I expect a bad day for Neocloud stocks just because $ORCL / $CRWV are "sector leaders". But this would be an ideal buying opportunity for individual companies in the sector: Due to the market missing the nuance that $ORCL / $CRWV's debt-filled buildout is for OpenAI as the main counterparty. But other companies in the sector like $NBIS, $CIFR, $WULF are backstopped by AAA-rated Hyperscaler ( $META / $META / $GOOGL / $AMZN) cash cows.
-
AI数据中心客户资金差异大，市场误判导致板块错配机会。
AI 数据中心(AI DC)领域极其微妙。 $ORCL 正基于租户 OpenAI 的积压订单/信用状况，通过债务建设产能。 OpenAI 目前无力履行其义务，且获得资金的可能性越来越小。 $CIFR 正为 $GOOGL、$AMZN 建设产能。 谷歌/亚马逊拥有资金，且这些合同已锁定。 $NBIS 正为 $META、$MSFT 建设产能。 微软/Meta 拥有资金，且这些产能合同已锁定。 但市场目前将所有积压订单都视为来自 OpenAI（不可持续的建设），因此这是买入该板块个别组件错定价的好机会。
英文原文
The AI DC sector is extremely nuanced. $ORCL is building capacity from debt based on backlog/credit worthiness of OpenAI which is the tenant. OpenAI does not have the funds at the moment to fulfill its obligations and it's looking less likely to get it. $CIFR is building out for $GOOGL, $AMZN. Google/Amazon does have the funds and these contracts are locked in. $NBIS is building out for $META, $MSFT. Microsoft/Meta does have the funds and these contracts are locked in for capacity. But the market is currently treating all the backlog like it comes from OpenAI (unsustainable build) which is why it's a good opportunity to buy into mispricing of individual components of the sector.
-
美联储降息后发布12月11日个股评级，重点推荐AI基建、稳定币及超跌成长股。
美联储降息25个基点后。 12月11日评级： 强烈买入： $CRCL $COIN $AMKR $CRDO $IBIT $MSTR $AMZN $SMCI $TSM $TSSI SK海力士 $SNAP 三星电子 $ALAB $META $NBIS $CIFR 买入： $KRUS $AVGO $NFLX $KRKNF $HIMS $FLY $OSS $TE $FLNC $LITE $COHR $RKLB $TTD $NVDA $CLS $GOOGL $RDDT $WULF $CRWV $IREN $GLXY $WLAC $MPWR 回避 $RGTI $PLTR $WMT $ETH $BMNR $TSLA $IONQ $ORCL $SLNH $OKLO 解释： 今天美联储如期降息25个基点。这通常会引导流动性进入成长股，并利好那些债务使用最多（以更低利率再融资）的中小盘股，例如像$NBIS和$CIFR这样的新云(Neoclouds)。 然而，这也恰逢日本加息，可能导致去年重新加载的套息交易(Carry Trade)平仓；但这是短期的，基本面>短期波动。 强烈买入评级： Circle - 大幅下跌主要由于IPO后的股份解禁。然而，降息损害了其商业模式~利息收入减少20%。 话虽如此，我们看到稳定币市场大幅增长，我个人看到大量早期风险投资（a16z, Sequoia等）涌入与稳定币相关的公司，如新银行(Neobanks)。我们应该看到所有这些都流入更多的USDC铸造，铸造量将抵消降息的影响。 Coinbase - 与Circle相同，他们在USDC方面有50%的收入分成。然而，他们还有自己的交易所，而且降息通常有助于风险资产如加密货币（尤其是比特币跌破9万美元后）。 Amkor - 受益于半导体/晶圆厂向美国制造的转移。 Credo - 过去5天下跌-16%，今天下跌8%。很好的恢复性买入，不认为数据中心建设的数据连接需求会下降。 ALAB - 与CRDO相同的论点 IBIT (比特币) - 始终是长线好标的，尤其是在$93K时 Microstrategy (MSTR) - 受益于比特币复苏，并分析了他们是否会爆仓。TLDR：不会，在需要支付利息之前（约2029年），我们还有另一次比特币减半事件。 Amazon - 今年一动不动。基本面改善，年底有助于电商部门。定制芯片、星座、Robotaxi，他们基本上什么都做，而市场尚未真正奖励他们的努力。只是感觉我们可能会在接下来的2个月看到它跑赢大盘。 SMCI - 之前发过关于这个的论点帖，惊人的恢复性买入。它因将收入积压转移到下一季度而在财报后下跌，但市场没有定价他们未来收入同比增长60%但交易在~11倍远期市盈率的事实。 TSM - 整个AI/半导体建设的骨干。我们看到关于TPU与GPU的争论，但TSM不在乎。 TSSI - 与SMCI相同的论点，依附于Dell，作为一个代理，我们看到来自IREN等供应商和其他在2026年建设数据中心的新云的巨大积压，我们应该看到这在明年实现。 SK海力士 - 显然有关于在美国市场上市升级的传闻，这应该提振流动性。此外，内存市场因AI建设而需求极高。 Snapchat - 只是被低估。$13B市值，~1B+季度收入。NA DAU较上季度下降3%，但不要为了成为下一个FB而买入。他们只需要削减GCP成本并货币化记忆功能（他们已经做了），我们应该看到明年重新评级100%+，特别是随着Perplixty交易带来的$400m+额外收入/股权。 三星电子 - 人们认为这也是内存，因为它构成了他们利润的很大一部分，但我将其视为潜在的下一个现金牛晶圆厂玩法，如TSM，作为第二大玩家吸收任何最大产能溢出。 META - 一次性税收抛售，超卖。现在我们终于看到他们创建前沿模型（Avacado，如果我记得没错的话）。所以他们可以货币化他们一直在花钱的llama开源llm努力。他们还削减了元宇宙努力，这应该是对盈利能力的巨大推动。 Nebius - 由于2500万股稀释导致短期拖累。ATM可能正在提供。话虽如此，一旦完成，由于来自其DC业务（7-9B ARR）及其4家市场未定价的子公司（同比增长100%+）的远期收入/增长，极度低估。 CIFR - 由于比特币价格（资产负债表上持有大量）导致短期下跌，但由于他们做托管(Colo)模式，不受GPU贬值争论的太大影响。此外由Google背书，并与Amazon有合同，因此从根本上降低了风险，是新云领域的顶级买入之一。 买入评级： 文字空间不够，所以给出更短的TLDR Kura Sushi - 波段交易，拉出5年图表，你会明白我的意思，每次它触底（大约现在）。这从未失败过！ Broadcom - 超大规模建设，与联发科一起对TPU至关重要 Netflix - 16%的下跌对于收购来说感觉有点不必要 KRKNF - 基本面增长良好，作为Andruil供应商的市场具有防御性。 HIMS - 股票回购计划，通常低于$40是很好的买入/波段交易。Zava收购未被定价，且仍在增长。 FLY - SpaceX $1.5T估值应该提振整个太空板块。这是2026年中型发射的玩法。 OSS - 之前对此进行过DD，潜在的Andruil供应商。否则，在这个市值下无论如何都有些低估。 TE - 少数Murican能源基础设施之一，太阳能。它可能比核能更商业化。 FLNC - 与AI建设+能源相同的论点 LITE - 现在相当过度延伸，不会追高。但长期受益于处于tpu ironwood + blackwell建设的中间。 COHR - 与Lite相同，但似乎是次要玩家。 RKLB - 可能是我最喜欢的长线。现在相当高估，但由于SpaceX的FOMO无法避免。 TTD - 之前的论点帖，仅基于远期收入数字，似乎是一个很好的恢复性玩法。 NVDA - TPU恐惧有点夸大，看看积压订单。 CLS - TPU v7生态系统买入 GOOGL - 他们像NVDA一样销售TPU，像Waymo一样增长Robotaxi市场，Gemini成功。全方位开火。 Reddit - 就像早期的Robinhood一样，只是一台印钞机。对RDDT通过FCF增长收购做了一些论点评论。否则，他们将留下来并受益于所有世代使用它（不像Snap那样早期） WULF - 类似于CIFR。重新评级可能会发生，取决于更多关于Anthropic建设的信息。 CRWV - 糟糕，糟糕的长线。良好的短期恢复性买入。 IREN - 如果他们继续购买GPU来做AI云，我不会把钱投进去，只是因为稀释。但他们可能会做托管，并且拥有大量的GW容量，所以仍然很有希望。 GLXY - 数据中心建设的受益者。 WLAC - 可能他们本月进行SPAC IPO。他们说Q4。 MPWR - TPU v7生态系统买入 回避 RGTI - 量子，没有基本面/收入支持 PLTR - 449.01B市值lol WMT - 他们每年增长4%的收入，但交易在40倍市盈率，这很疯狂。 ETH - 以太坊伟大的网络。然而，没有代币销毁，也没有收入流向代币持有者。糟糕的投资，伟大的开发者工具/生态系统。 BMNR - 以太坊代理。 TSLA - 有点脱离基本面。但这是对elon musk、大规模Robotaxi、机器人的赌注。我个人只是认为这过度承诺，但我们会看到。 IONQ - 量子，没有基本面/收入支持 ORCL - 大部分远期积压依赖于openai，如果openai在市场份额上输给claude/gemini，这使得事情极其不确定/有风险。话虽如此，现在是一个很好的恢复性买入，但长期来看有风险。 SLNH - 如果你想被他们的2.8gw管道稀释到虚无，这是要持有的股票。 OKLO - 没有像量子那样的基本面来支持目前的市值，这可能需要多年才能实现。
英文原文
Post-Fed Interest Rate 25BPS cut. December 11th ratings: Strong Buy: $CRCL $COIN $AMKR $CRDO $IBIT $MSTR $AMZN $SMCI $TSM $TSSI Sk Hynix $SNAP Samsung Electronics $ALAB $META $NBIS $CIFR Buy: $KRUS $AVGO $NFLX $KRKNF $HIMS $FLY $OSS $TE $FLNC $LITE $COHR $RKLB $TTD $NVDA $CLS $GOOGL $RDDT $WULF $CRWV $IREN $GLXY $WLAC $MPWR Avoid $RGTI $PLTR $WMT $ETH $BMNR $TSLA $IONQ $ORCL $SLNH $OKLO Explanations: Today fed cut interest rates 25BPS as expected. This usually funnel liquidity into growth stocks and benefits small-medium caps that use debt the most (refinance with lower interest rates), such as Neoclouds like $NBIS and $CIFR. However, this coincides with Japan hiking, which might lead to carry trade unwind from last year's reload; but this is short term, fundamentals > volatility short term. Strong Buy Ratings: Circle - Massive drop mainly due to share unlock post IPO. However, rate cuts hurt their business model ~20% revenue cut from interest. That being said, we're seeing a massive growth in the stablecoin market, and I'm personally seeing huge early venture capital funding (a16z, sequioa, etc). being poured into stablecoin related companies such as Neobanks. We should see all of this funnel into more USDC printing, and the printer outweigh rate cuts. Coinbase - Same as Circle, they have 50% revenue sharing in terms of USDC. However, they also have their exchange on top, and rate cuts generally help riskier assets such as crypto (especially post drop Bitcoin sub 90k) Amkor - Benefits from Made in America shift to semis/fab. Credo - Dropped -16% last 5 days, and 8% today. Great recovery buy, don't see connectivity demand dropping from DC buildout. ALAB - Same thesis as CRDO IBIT (Bitcoin) - Always a great long, especially so at $93K Microstrategy (MSTR) - Benefits from Bitcoin recovery and did an analysis whether they would get liquidated or not. TLDR: no, we have another bitcoin halving event before they need to pay off interest, which was around 2029. Amazon - Hasn't moved an inch all year. Fundamentals improving, EOY helps E-commerce division. Custom chips, constellations, robotaxis, they're basically doing everything and market hasn't really rewarded their effort yet. Just a feeling we might see this outperform next 2 months. SMCI - Did a thesis post on this earlier, amazing recovery buy. It dropped on earnings due to shifting revenue backlog to next quarter, but markets aren't pricing in the fact they're growing 60% Y/Y forward revenue but trading at ~11 forward p/e or so. TSM - Backbone of the whole AI/semi buildout. We're seeing arguments about TPU vs. GPU, but TSM doesn't care. TSSI - Same thesis with SMCI, piggybacks off of Dell, just as a proxy we're seeing massive backlog from vendors such as IREN, and other neoclouds building out DCs 2026, and we should see this come into fruition next year. Sk Hynix - Apparently there's been rumors about uplisting to US markets, which should be a boost to liquidity. Also memory markets is just incredibly high demand from AI buildout. Snapchat - Just undervalued. $13B marketcap, ~1B+ quarterly revenue. NA DAU dropped 3% from last quarter but don't buy this for being the next FB. All they need to do is cut GCP costs and monetize memories (which they did) and we should see this re-rate 100%+ next year, especially with $400m+ in added revenue/equity from the Perplixty deal Samsung Electronics - People think of this as memory as well because it makes up a large part of their profit, but i see this as a potential next cash cow foundry play like TSM, as the 2nd largest player to soak up any max capacity overflow. META - One time tax selloff, was oversold. Now we finally see them create a frontier model (Avacado) if i remember correctly. So they can monetize the llama open source llm efforts they've been just blowing money on. They also cut their metaverse efforts, which should be a huge boost in proftiability. Nebius - Short term drag due to 25m share dilution. ATM is likely being offered. That being said once this finishes, insanely undervalued due to forward revenue/growth from both its DC business (7-9B ARR), and its 4 subsidaries that the markets dont price in (growing 100%+ Y/Y) CIFR - Short term drop due to Bitcoin prices (holding a lot on balance sheet), but not really affected by GPU depreciation arguments since they do colo models. Also backstopped by google, and they have contracts with Amazon, so fundamentally disrisked and one of the top buys in neocloud secotr. Buy Ratings: Running out of text space so will give a shorter TLDR Kura Sushi - Swing trade zoom out 5 year chart and you'll see what I mean every time it bottoms (around now). This never fails! Broadcom - Hyperscaler buildout, critical to TPU alongside Mediatek Netflix - 16% drop feels a bit unwarranted for the acquisition KRKNF - Great growing fundamentals and defensible market as an andruil supplier. HIMS - Share buyback program, usually sub $40 great buy/swing trade. Zava acqusition not being priced in and it's still growing. FLY - SpaceX $1.5T valuation should boost up the whole space sector. This was a 2026 play for medium lift. OSS - DD on this earlier potential andruil supplier. Otherwise, kind of undervalued at this MC anyway. TE - One of the few Murican energy infra, Solar. It's likely more commercial than Nuclear. FLNC - Same thesis with AI buildout + energy LITE - Pretty overextended right now, wouldn't chase. But long term benefits from being in the middle of both tpu ironwood + blackwell buildout COHR - Same with Lite, but seems like a secondary player. RKLB - Probably my favorite long. Pretty overvalued right now but can't help it due to SpaceX fomo. TTD - Thesis post earlier, just based on forward revenue numbers, it seems like a great recovery play. NVDA - TPU fears are a bit overblown, just look at backlog. CLS - TPU v7 ecosystem buy GOOGL - They sell TPUs like NVDA, growing robotoaxis market like waymo, gemini succesful. Just firing on all fronts. Reddit - Just a money printer like early day Robinhood. Made some thesis comments about RDDT growing in terms of acquisitions from FCF. Otherwise, they're here to stay and benefits from all gens using it (unlike snap which is earlier) WULF - Similar to CIFR. Rerating might happen depending on more info about the Anthropic buildout. CRWV - Terrible, terrible long. Good short term recovery buy. IREN - I would not put money into this if they kept buying GPUs to do AI cloud just due to dilution. but they might do colo and they have an immense amount of GW capacity so it's still promising. GLXY - Beneficary of DC Buildout. WLAC - Possible that they're SPAC ipoing this month. They did say Q4. MPWR - TPU v7 ecosystem buy Avoid RGTI - Quantum, no fundamentals/revenue to back it up PLTR - 449.01B market cap lol WMT - They're growing like 4% revenue a year, but trading at 40 p/e which is insane. ETH - Ethereum great network. However, there's no token burn and none of the revenue goes to token holders. Terrible investment, great developer tooling/ecosystem. BMNR - Ethereum proxy. TSLA - Kind of detached from fundamentals. But it's a bet on elon musk, robotaxis at scale, robotics. I personally just see this as overpromising, but we'll see. IONQ -Quantum, no fundamentals/revenue to back it up ORCL - Most of forward backlog is dependent on openai, which makes things incredibly uncertain/risky if openai falls to claude/gemini in market share. That being said, it's a good recovery buy right now, but long term it's risky. SLNH - This is the stock to be in if you want diluted to oblivion on their 2.8gw pipeline. OKLO - no fundamentals like quantum to back up mc at this moment, this likely years out to come into fruition.
-
引用亚马逊案例佐证观点。
@bubbleboi 确实，看看 $AMZN 就知道了。https://t.co/EpkUPYB4GW
英文原文
@bubbleboi True, just look at $AMZN. https://t.co/EpkUPYB4GW
-
看好RKLB与NBIS，认为市场低估了这两家行业第二梯队的成长潜力。
市场正在错误定价地球上最大板块中的第二梯队玩家。 我对 $RKLB = $NBIS Avride 的论点： 1. 太空领域 ($RKLB) -> SpaceX 估值刚从 $350B 升至 $800B。Rocketlab 将从 $26B 增长至 SpaceX 之前的估值水平。 这只是时间问题。 2. 自动驾驶出租车 (Robotaxis) ($NBIS) Waymo 在一年内从 $45B 涨至 $200B。$NBIS Avride 将从 $6B 增长至 Waymo 之前的估值水平。 这只是时间问题，无论是一年还是四年后。 关于 $NBIS 的关键点是：Avride 只是其众多子公司之一，仅这一家公司在两年内的价值就可能超过当前的整体市值。 市场和空头都错误定价了这样一个事实：Nebius 拥有地球上增长最快、最热门的板块公司，且全部实现 100%+ 的同比增长。 做多 $RKLB，做多 $NBIS。
英文原文
The market is mispricing the #2 players in the biggest sectors on Earth. My thesis on $RKLB = $NBIS Avride 1. Space ( $RKLB ) -> SpaceX just got valued at $800B from $350B. Rocketlab will grow into SpaceX's prev valuation from $26B It's just a matter of time. 2. Robotaxis ( $NBIS ) Waymo went from $45B → $200B in 1 year. $NBIS Avride is going to grow into Waymo's previous valuation from $6B. It's just a matter of time, whether that's 1 year or 4 years from now. And here is the thing with $NBIS: Avride is just one subsidiary out of multiple, and this company alone could be worth more than the entire market cap today in 2 years. Both the market and short sellers misprice the fact that Nebius owns the fastest growing and hottest sector companies on Earth, all growing 100%+ Y/Y. Long $RKLB, Long $NBIS.
-
Avride是$NBIS核心资产，自动驾驶出租车市场潜力巨大，机构严重低估$NBIS价值。
根据 Seeking Alpha 报告与 Nuro 的基准对比，其估值现在可能约为 60 亿美元。但我会说 Avride 是 $NBIS 的皇冠明珠，而非 Clickhouse，尤其是当 Waymo 等自动驾驶出租车 (Robotaxi) 公司在短时间内估值飙升至 1000 亿美元以上时。 $TSLA 仅因推出自动驾驶出租车就增加了数千亿美元的价值（例如 Cathie Wood 表示自动驾驶出租车可能占特斯拉价值的 90%），这是一个因参与者稀缺而具有极大上行空间的市场。 基本上只有 $TSLA、$GOOGL 的 Waymo、$AMZN 的 Zoox、现代……以及正试图抵御自动驾驶出租车竞争的 $UBER。 $UBER 通过 Avride 而非与 Waymo 合作来扩大其防御能力至关重要，以避免将客户引导至竞争对手的应用程序（例如他们联合投资了 3.75 亿美元的原因）。 鉴于 $NBIS 拥有 Avride 83% 的股份，如果其规模达到约 450 亿美元市值（这是 Waymo 在 2024 年 10 月的估值）……这实际上超过了 Nebius 今天的市值，而它才刚刚在德克萨斯州启动。 像 Hedgeye 这样的基金可能没有计算 SOP 增长，并且严重低估了 Nebius。
英文原文
Probably around $6B now as per the seeking alpha report benchmarked vs Nuro. But I'd say Avride is crown jewel of $NBIS, not clickhouse, especially when robotaxi companies like Waymo scaled to a $100B+ valuation in a short time. $TSLA gained hundreds of billions in value just due robotaxi launch (eg. Cathie Wood said robotaxis could be 90% of tesla's value) and it's a market with extreme upside due to lack of players. There's basically just $TSLA, $GOOGL Waymo, $AMZN Zoox, Hyundai... and then $UBER which is trying to defend themselves against robotaxi competition. $UBER has the upmost importance to scale its defensibility with Avride instead of partnering with Waymo -> funneling customers into the other app. (eg. why they joint invested $375M). Given how $NBIS owns 83% of avride, if it scales to ~$45B MC, which was Waymo's valuation BACK in October 2024.... That's literally more than Nebius's marketcap today lol, and it just launched in Texas. Funds like Hedgeye probably aren't accounting for SOP growth and are vastly underestimating Nebius.
-
反驳Hedgeye做空逻辑，对比Coreweave指出Nebius财务更健康且业务多元。
做空机构 @Hedgeye 正在做空 Nebius $NBIS。股价现为 $97.8。 他们的观点？Nebius 是 Coreweave 2.0。 Coreweave [ $CRWV ]： - 客户多元化程度低，主要是超大规模云服务商，约四分之一的积压订单来自 OpenAI。 - 剩余现金 18 亿美元。 - 由于 8-10% 利率的有毒融资，每年债务利息超过 13 亿美元。 Nebius 也是如此吗？ Nebius [ $NBIS ]： - 在 Neocloud 市场中拥有最高的客户多元化，包括 $META、$MSFT、政府、$SHOP、Mistral、$NOW 等。 - 剩余现金超过 48 亿美元。 - 票据结构约 40%+ 价外(OTM)，年利率约 2%，每年利息支出约 7660 万美元。 - 拥有对 $UBER 至关重要的 FSD 4 级 Robotaxi 自动驾驶汽车业务板块（最后一轮融资 $UBER 投入 3.75 亿美元），以对抗 Waymo。 - AI 训练板块由贝索斯资助，并被 $AMZN、Anthropic 等使用。 - 持有子公司 28% 的股份，该子公司被 $TSLA、$NET、Tiktok、$META 等大多数上市公司使用。 - 4 家子公司同比增长 100%，伴随核心运营业务增长，Tripleten/Toloka 每年增加净收入。 这是同一家公司吗。
英文原文
A short seller firm @Hedgeye is now shorting Nebius $NBIS. The stock price is now $97.8. Their claim? Nebius is Coreweave 2.0. Coreweave [ $CRWV ]: - Small diversification in customers, mainly hyperscalers, ~quarter of their backlog is OpenAI. - $1.8B in cash left - $1.3B+/yearly in debt interest from toxic financing at 8-10% interest. Same thing as Nebius right? Nebius [ $NBIS ]: - Highest diversification in the Neocloud market of customers from $META, $MSFT, governments, $SHOP, Mistral, $NOW, and many others. - $4.8B+ in cash left - ~$76.6M/yearly from note structure 40%+ OTM at ~2% interest. - FSD level 4 Robotaxi self driving car segment that is of critical importance to $UBER (last round $375M with $UBER) to compete vs Waymo. - AI training segments funded by Bezos and used by $AMZN, Anthropic, and others - 28% in subsidiary that most of the public companies from $TSLA, $NET, Tiktok, $META and so on use - 4 subsidiaries growing 100% Y/Y alongside their core operational business, with Tripleten/Toloka adding to net income every year. Same company.
-
反驳做空 $NBIS 观点，对比其与 $CRWV 在财务及业务上的巨大差异。
做空 $NBIS 是个好主意，因为你认为它和 $CRWV 一样，不错的论点！ $CRWV -> 每年 13 亿+美元的债务利息，拥有来自 OpenAI 的大量积压订单，剩余现金 18 亿美元。 $NBIS -> 可转换票据结构，利息不会严重侵蚀盈利能力；客户基础多元化以实现高利用率（Shopify、Cursor、政府机构、Meta/MSFT 等超大规模云服务商）；拥有 48 亿+美元现金用于建设；Robotaxi 自动驾驶汽车板块被 $UBER 使用并正在扩展以与 Waymo 竞争；AI 训练板块由 $AMZN、Anthropic 等资助并使用；持有 28% 的子公司股份，该子公司被大多数上市公司使用，所有 4 家子公司同比增长 100%——市值 240 亿美元。 当然和 Coreweave 是同一家公司。
英文原文
Great idea to short $NBIS because you think it's the same $CRWV, nice thesis! $CRWV -> $1.3B+ yearly debt interest, large backlog from OpenAI, $1.8B in cash left. $NBIS -> Convertible notes structure where interest doesn't heavily cut into profitability, diversified customer base for high utilization (shopify, cursor, governments, hyperscalers like meta/msft), $4.8B+ in cash for buildout, Robotaxi self driving car segments used by $UBER and scaling up to compete vs Waymo, AI training segments funded by and used by $AMZN, Anthropic, and others, 28% in subsidiary that most of the public companies use, all 4 subsidiaries growing 100% Y/Y - $24B MC. Same company as Coreweave for sure.
-
分析NBIS、IREN、CIFR及TE的投资逻辑与风险，并点评用户高波动组合。
从我的角度来看： $NBIS - 拥有最高的非对称上行潜力，因为除了核心的运营数据中心业务外，他们还有4家类似Robotaxi的子公司，同比增长超过100%。这只是一场等待市场定价部分价值（例如$UBER因Avride自动驾驶汽车大涨，但持有Avride的$NBIS却无反应）的游戏。 我非常看好Robotaxi和Avride。我不明白市场为何没看到这个机会（也许我们只是太早了，不过据我上次检查，机构持仓已从38%累积至流通盘的52%以上，所以有些人可能已经意识到了）。 $IREN - 潜在上行空间最大，但风险也是所有标的中最大的。仅为了达成与$MSFT的交易，他们就不得不稀释股权并支付58亿美元以上的硬件费用，且为了加速建设，每兆瓦成本高达200万美元。然而，这仅仅是为了变现一小部分产能，他们选择的是AI云路线，而非像$CIFR那样购买GPU做托管(Colo)。如果$IREN继续提供GPU服务，我会作为投资者极度担忧，因为执行存在不确定性，且变现“3GW管道”需要极其巨大的稀释。也许再等几个月会更好，因为短期投资者受损最重。 $CIFR - 可能是新云(Neocloud)板块中最安全的，因为他们有$AMZN和$GOOGL背书，资产负债表上有大量比特币以应对下一次减半事件，且虽然营收增长较慢，但通过超大规模云厂商合约获得了高毛利的托管(Colo)收入可见性。 $TE - 能源是个不错的博弈方向。 你的投资组合波动性极大哈哈（可能回撤50%），但如果你出身于WSB Reddit、加密货币和期权交易圈，那也没问题。
英文原文
Okay so from my perspective $NBIS - Highest asymmetrical upside possible because they have 4 subsidaries like robotaxis growing 100%+ Y/Y alongside their core operational DC business. It's just a waiting game until market prices in parts (eg. $UBER rose a ton from self driving cars with Avride, but $NBIS had no movement but owns Avride). I really really like robotaxis and Avride. I don't know how markets aren't seeing this opportunity (maybe we're just early, institutions accumulated over 52% of the float last i checked though from 38%, so some are probably aware). $IREN - highest possible upside, but biggest risk of them all. Just for their $MSFT deal they had to dilute and pay for $5.8B+ worth of hardware just to do the deal and ended up spending up $2m/mw in costs just for expedited buildout. However, that was just to monetize a small amount of capacity and they're going the AI cloud route instead of colo buying GPUS, unlike $CIFR. I'd be extremely worried as an investor if they kept doing GPU offerings because of execution uncertainty and extreme, extreme amount of dilution to monetize that "3 GW pipeline". Maybe it would be better off waiting another few months because near term investors are harmed the most. $CIFR - probably safest one in the neocloud sector since they are backstopped by $AMZN and $GOOGL, large amounts of bitcoin on balance sheet for next halving event, and are doing slower revenue growth but higher margin colo offerings with revenue visibility from hyperscaler deals $TE - energy is a good play. Your portfolio is incredibly volatile lol (can swing -50%) but it's fine if you're born from the depths of wsb reddit, crypto, and options.
-
博主认为七大科技股当前估值具备较高概率优势。
@Scepticus16 就我而言，从当前估值来看，$meta、$amzn、$googl、$nvda、$appl、$msft、$tsla 的概率较大。
英文原文
@Scepticus16 Prob $meta, $amzn, $googl, $nvda, $appl, $msft, $tsla for me in terms of current valuations
-
AI算力需求指数级增长抵消GPU迭代贬值，NVDA客户优质，非泡沫崩盘。
答案很微妙。 主要看两个因素： 1. GPU 变得更节能。 2. 大语言模型(LLM) 在容量/能效上更高效。 在 LLM 方面，我们看到像 DeepSeek 这类模型在处理不需要高精度的任务（如回答烹饪食谱或知识库查询）时极其高效。 然而……随着计算力的增加，准确率（尤其是复杂研究问题）也在提升。Elon 和 Magnificent Seven 意识到了这一点，所以他们正在扫货市场上的所有 GPU 以创造超级智能。这也是为什么 Anthropic 和 Google 正在建设耗资 400 多亿美元的数据中心，用于运行需要更多算力进行批判性思维（如 Genesis 任务）的更高级 Opus 和 Gemini 模型。 在 GPU 方面，每一代新 GPU（例如 H100 -> B200）在能效和每美元性能上都有显著提升。例如，Blackwell B200 是 Hopper H100 的 30 倍。 如果基于这个假设，那么到 2027/2028 年，市场上将出现大量过时的低效 H100 和 B200，导致二手 GPU 市场崩盘。 但是：这是假设我们没有看到对新 AI 能力的指数级需求（我们很可能会看到，且正在发生）。正因为这种指数级需求，今天旧模型（如 7 年前的 TPU 和 2020 年的 GPU）仍被用于低优先级的推理任务。 $NVDA 的订单已积压数年，人们正在购买 $AMD 的 GPU 和 $GOOGL 的 TPU 来构建任何新增产能。 至于思科类比，思科的客户是互联网泡沫时期无盈利能力的公司。$NVDA 的客户是 $META、$AMZN、$GOOGL、$MSFT，这些是世界上最盈利的公司。所以最坏的情况我们可能看到回调，而不是互联网泡沫式的崩盘。
英文原文
Answer is nuanced. So two factors: 1. GPUs get more power efficient. 2. LLMs get more capacity/power efficient. For the LLMs case, we're seeing that on deepseek type models be extremely efficient on stuff that don't require much accuracy. Basic stuff like responding to questions about cooking recipes, or knowledge-base stuff. However... accuracy increases, especially with complex research questions, scaled with compute. And people like Elon + mag7 realize this, which is why they're just buying up all the GPUs on the market to create superintelligence. And why antrhopic/google is building $40B+ datacenters for more advanced opus and gemini models that require more compute for critical thinking (eg. Genesis Mission) For the GPUs case, every new generation of GPU (e.g., H100 -> B200) offers dramatic improvements in power efficiency and performance per dollar. eg. Blackwell B200 is 30x than the Hopper H100. If we go off that assumption, then there would be a massive useless supply of less-efficient H100s and B200s in 2027/2028 creating a used GPU market crash. HOWEVER: This is if we don't see an exponential demand for new AI capabilities (which we likely will, and what we're seeing now). Because of this exponential demand, TODAY, older models are still used (eg. TPUs from 7 years ago and GPUs from 2020), for lower inference task in lower priority inference tasks. $NVDA is backlogged for years and people are buying GPUs from $AMD /TPUs from $GOOGL to build out any new capacity. As for Cisco analogy, Cisco's customers were .com bubble companies with no profitability. $NVDA's customers are $META, $AMZN, $GOOGL, $MSFT the most profitable companies in the world. So worst case scenario we might see a correction, not a .com bubble crash.
-
解析$NBIS旗下Avride与Toloka的高增长潜力，指出市场低估其子公司价值。
$NBIS 即将超越 $CRWV，原因如下： 在 $TSLA、$AMZN、$GOOGL 和现代汽车中，Nebius [ $NBIS ] 拥有美国仅有的 4 家 FSD-LVL 4 级自动驾驶出租车公司之一：Avride。 人们还不知道的是： $NBIS 拥有 Toloka，其同比增长 140%。 Toloka 被 Anthropic、Hugging Face、$AMZN、$MSFT、ServiceNow 和 $SHOP 用于训练人工智能模型，在杰夫·贝佐斯风险投资部门的最新一轮融资中，Toloka 估值约为 7 亿至 12 亿美元。 Nebius 拥有多数股权（约 60%），但市场错误定价了其子公司与其核心业务一同以极高速率增长的事实。 虽然不如 Avride 或 Clickhouse 具有主导地位，但鉴于其增长率，Toloka 代表了 Nebius 投资论点中的重要组成部分。 我们最近看到 Scale AI（估值 290 亿美元）被 $META 收购（技术上约 49%），以此了解市场规模，并且由于利益冲突担忧，$GOOGL 和 OpenAI 的客户流失，这增加了 Toloka 的价值。据报道，Meta 收购后，Google 和 OpenAI 切断了与 Scale 的联系。仅 Google 在 2025 年为 Scale 预算了 2 亿美元。即使捕获其中 10-20% 的转移支出，也将实质性提升 Toloka 的收入。 管理层指引 2025 年收入为 50-70 亿美元。牛市情景假设他们达到 70 亿美元以上（上限或超过）。鉴于他们在 2024 年增长了 140%，如果企业交易转化，这并非不合理。计算如下：70 亿美元 × 25 倍 = 175 亿美元。若为 80 亿美元且乘数为 25-30 倍（若势头异常强劲），则接近 20-25 亿美元。 但鉴于其增长率，我们可能会看到： 基准情景（12-15 倍乘数）：7-10 亿美元（当前）-> 9-13 亿美元（1 年后）。 牛市情景 20-25 倍：12-17.5 亿美元（当前）-> 15-25 亿美元（1 年后）。 2024 年收入同比增长 140% 至约 2640 万美元，2025 年指引为 5-7 亿美元（同比增长 90-165%）。在一般牛市情景下，我们可能在 1 年内看到额外约 14 亿美元的子公司价值添加到分部加总估值中。 多年后，如果公司和市场增长，该价值可能单独带来 30 亿、60 亿、90 亿美元。 简而言之，市场仍在将 Nebius 的子公司估值为缓慢增长、商品化的业务，而非以三位数速率高价值复利的业务。真正的超额收益来自投资者理解市场错误定价了 $NBIS 这一篮子超高速增长公司。 $NBIS 拥有多个三位数增长的子公司，受益于并行价值创造：每项业务独立扩展，但其进展强化了母公司的估值。
英文原文
$NBIS is about to leapfrog $CRWV, here's why: Among $TSLA, $AMZN, $GOOGL, and Hyundai, Nebius [ $NBIS ] owns 1 of only 4 FSD-LVL 4 US Robotaxi companies: Avride. What people also don't know is this: $NBIS owns Toloka, growing 140% Y/Y. Used by Anthropic, Hugging Face, $AMZN, $MSFT, Servicenow, and $SHOP for training artificial intelligence models, Toklo is likely valued around $700M-$1.2B in its latest round by Jeff Bezos's venture arm. Nebius owns a majority stake (roughly ~60%), but the market is mispricing how their subsidiary companies grows at extreme rates alongside their core business. While not as dominant as Avride or Clickhouse, Toloka represents a meaningful piece the Nebius investment thesis given their growth rate. We’ve seen Scale AI ( $29B ) get acquired by $META recently (technically ~49%) to get a sense of market size, and we’ve seen customers from $GOOGL and OpenAI defect due to conflict of interest concerns, given added value to Toloka. Google and OpenAI reportedly cut ties with Scale after Meta's acquisition. Google alone had $200M budgeted for Scale in 2025. Even capturing 10-20% of that displaced spend would materially boost Toloka's revenue. Management guided $50-70M for 2025. The bull case assumes they hit $70M+ (the upper bound or exceed it). Given they grew 140% in 2024, this isn't unreasonable if enterprise deals convert. The math: $70M × 25x = $1.75B. At $80M with 25-30x (if momentum is exceptionally strong), you get toward $2-2.5B. But given their growth rate we’ll likely see: Base case (12-15x multiple): $700M-$1B (currently) -> $900M-$1.3B 1Y forward. Bull Case 20-25x: $1.2-1.75B (currently) -> $1.5-2.5B 1Y forward Revenue grew 140% year-over-year in 2024 to approximately $26.4 million, with 2025 guidance of $50-70 million (90-165% Y/Y growth). And with a general bull-case sceanrio, we might see an additional ~$1.4B in subsidiary value added to Sum of Parts in 1 year time. In multiple years time, that value might bring in $3B, $6B, $9B alone if the both the company and market grows. In short, the market is still valuing Nebius’s subsidiaries as if they were slow-moving, commoditized businesses instead of high-value compounding at triple-digit rates. The true alpha from investors understanding that the market is mispricing $NBIS basket of hyper-growth companies. $NBIS with multiple triple-digit-growth subsidiaries benefits from parallel value creation: each business scales independently, but their progress reinforces the parent company's valuation.
-
分析Robotaxi竞争格局：Waymo领先，Uber联手Avride和WeRide应对巨头威胁。
欢迎反驳关于 $TSLA 是 FSD SAE 2级自动驾驶的观点。特斯拉被归类为2级，是因为其 Robotaxi（自动驾驶出租车）始终有人远程监控。埃隆正在追求一个更宏大的计划，即采用摄像头+非地理围栏解决方案，所以我做的并非简单的同类比较。以下是三点看法：1. 如果他能制造出可重复使用的大型太空火箭，我毫不怀疑他最终能在地球上实现他的计划。2. Waymo 明显领先于其他所有竞争者。3. 我的观点是，来自 $GOOGL、$TSLA、$AMZN 等万亿美元市值公司的个位数 Robotaxi 玩家正在与 $UBER 竞争。Uber 视此为威胁，因此正与 $NBIS 的 Avride（美国）、WeRide（中东、中国）合作，以在明年扩大规模。
英文原文
Feel free to dispute the point that $TSLA is FSD SAE level 2 automation. Tesla is classified as level 2, because in their robotaxis, you have people remotely behind the wheel at all times. Elon is pursuing a more ambitious plan with cameras + non-geofenced solution so it's not an apples to apples comparison I'm making. That being three things: 1. I have no doubt Elon can achieve what he's planning in due time back in planet earth if he's able to create large reusable rockets that go into space. 2. Waymo is clearly ahead of all other players. 3. Point I'm making is there's single digit robotaxi players that all come from trillion dollar company exposure in $GOOGL, $TSLA, $AMZN that compete vs $UBER. Uber sees this as a threat so they're working with $NBIS Avride (United States), WeRide (Middle East, China) to scale it up over the next year.
-
Nebius核心业务高增，其L4级Robotaxi子公司Avride被市场低估。
Nebius [ $NBIS ] 的核心业务简直疯狂。年收入(ARR)在一年内实现700%以上的同比增长，达到70-90亿美元以上。 更疯狂的是这一点。 其子公司可以超越主营业务 🤯 原因如下： $GOOGL Waymo、$TSLA、$AMZN Zoox、Motional、$NBIS Avride 是目前仅有的5家美国完全自动驾驶(FSD) Robotaxi 公司。 是的，$NBIS Avride 是其中之一。 - 它是4家拥有L4级完全自动驾驶（与Waymo同级）的公司之一。 - 它是3家实现L4级FSD商业化的公司之一（Motional仍处于研发阶段）。 作为参考，Tesla [ $TSLA ] 的Robotaxi目前仅为L2级（需要人工监督）。 人们希望购买Robotaxi公司，因为Cathie Wood所说：“Robotaxi可能占 $TSLA 价值的90%）。 因此，他们通过 $3.6T 的 $GOOGL 和 $1.4T 的 $TSLA 等万亿美元公司来买入该板块的敞口。 但有一家L4级FSD Robotaxi公司即将通过 $UBER 与Waymo竞争： 那就是Avride，一家估值60亿美元的子公司，$NBIS 持有其83%的股份，而市场完全忽略了它。Avride 是 该60亿美元估值基于Seeking Alpha分析师报告（2025年8月），将Avride与Nuro的E轮融资估值进行对标。仅在2025年1月，其估值估计为34亿美元，短短6个月内几乎翻倍（反映了Avride与Uber和现代合作的势头）。 支持当前估值的关键近期进展： - 2017年：Avride由Yandex自动驾驶集团创立 - 2017-2024年：研发... - 2024年10月：与Uber达成多年期配送机器人和Robotaxi合作伙伴关系 - 2025年3月：与现代签署L4级自动驾驶汽车联合开发谅解备忘录(MOU) - 2025年10月：获得Uber和Nebius高达3.75亿美元的战略投资 - 2025年底：在德克萨斯州达拉斯通过Uber推出Robotaxi 我们现在看到7年的研发开始商业化爬坡（像Waymo一样），今年通过 $UBER 实现。 你可以以 $NBIS 210亿美元市值的零头价格，获得这家超高速增长的L4级FSD Robotaxi公司。🚗 鉴于Avride同比增长三位数，且我们看到该公司在7年的资本支出投入研发后终于开始商业化，它在2年内估值远超Nebius当前市值的可能性非零。 我们曾看到Waymo从450亿美元市值（2024年10月，融资56亿美元）增长到2000亿美元以上（DA Davison建议的估值）。 Avride目前仅为60亿美元，是 $NBIS 市值的一小部分。但它正处于与 $UBER 合作的增长临界点，就像Waymo最初在旧金山起步但现在无处不在一样。 Nebius ( $NBIS ) 是市场中被误解的成长型公司。但作为投资者，尽早布局下一代公司才能获得最高回报
英文原文
Nebius [ $NBIS ] core business is insanity. 700%+ Y/Y growth to $7-9B+ ARR in 1Y. What's more insane is this. Their subsidiary can outgrow their main business 🤯 Here's why: $GOOGL Waymo, $TSLA, $AMZN Zoox, Motional, $NBIS Avride are the only 5 FSD US Robotaxi companies right now. Yes $NBIS Avride is one of the 5. - And 1 of 4 with FSD Level 4 - fully driverless, the same level as Waymo. - And 1 of 3 that are commercialized with FSD level 4 (Motional is R&D phase) Tesla [ $TSLA ] robotaxis for perspective is only at Level 2 (requires human oversight). People want to buy robotaxi companies because as Cathie Wood puts it “Robotaxis could represent 90% of $TSLA ‘s value). So they buy exposure to the segment through trillion dollar companies like Waymo through the $3.6T $GOOGL and $1.4T $TSLA. But there's a Level 4 FSD robotaxi company that is just about to compete with Waymo through $UBER: That name is Avride, a $6 billion subsidiary company that $NBIS owns 83% of, and one that the market completely has missed. Avride is an The $6B valuation was based off a Seeking Alpha analyst report (August 2025) that benchmarked Avride against Nuro's Series E valuation. Just in Jan 2025 it was estimated to be valued at $3.4 billion, growing almost double in just 6 months time (reflecting Avride's momentum with Uber and Hyundai). Key recent developments supporting the current valuation: - 2017: Avride founded though Yandex Self Driving Group - 2017-2024: Development... - October 2024: Multi-year Uber partnership for delivery robots and robotaxis - March 2025: Hyundai MOU for co-development of Level 4 autonomous vehicles - October 2025: Up to $375 million in strategic investment from Uber and Nebius - By EOY 2025: Dallas Texas Robotaxi Launch with Uber We’re now seeing 7 years of R&D starting commercialization ramp (like Waymo), through $UBER this year. and you can get this hyper scaling self-driving FSD level 4 robotaxi company. As discounted spare change of $NBIS at a $21B marketcap. 🚗 Seeing how Avride is growing triple digits Y/Y and we’re seeing the company finally commercialize after 7 years of capex into research, there’s a nonzero chance it becomes valued way more than the Nebius current market cap in 2 years. We've seen Waymo grow from a $45B marketcap (October 2024, $5.6B raised) to over $200B+ (DA Davison suggested valuation). Avride is now only at $6B and a small change of $NBIS's marketcap. But it's at the precipice of growth with $UBER, just like how Waymo started out in SF but now it's everywhere. Nebius ( $NBIS ) is the misunderstood growth company in the market. But being early to the next generation companies is where the highest returns are made as an investor
-
看好传统云商，CRWV或成新云代表，板块成员间接受益。
鉴于数据的重要性，概率上更倾向于像 $ORCL 这样的联邦云(FedRamp)提供商以及像 $AMZN、$GOOGL 和 $MSFT 这样的传统超大规模云服务商(Hyperscalers)。 $CRWV 可能是最有可能成为新云(Neocloud)代表的公司，但他们仍处于应用阶段。$NBIS 的可能性不大。 话虽如此，新云板块的其他成员也是间接受益者。
英文原文
Prob just fedramp providers like $ORCL and traditional hyperscalers like $AMZN, $GOOGL, and $MSFT given how critical the data is. $CRWV is probably the most likely Neocloud but they’re still in application phase. $NBIS is unlikely. That being said, other Neocloud sector members are indirect beneficiaries.
-
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.
-
深度解析9只个股基本面与目标价，指导散户建立独立估值模型。
基于权益排名表： 以下是对每只股票的深入分析，以及我如何重新调整投资组合以利用市场重置： · $NBIS 现价 $92，目标价 $400 / 1年 · $RKLB 现价 $43，目标价 $500 / 5年 · $CRCL 现价 $72，目标价 $150 / 8个月 · $ALAB 现价 $143.4，目标价 $250 / 6个月 · $SNAP 现价 $8.1，目标价 $22 / 1年 · $CIFR 现价 $14.8，目标价 $28 / 6个月 · $RDDT 现价 $185，目标价 $275 / 8个月 · $SMCI 现价 $34，目标价 $55 / 6个月 · $HIMS 现价 $35，目标价 $60 / 6个月 此顺序基于发布时的持仓集中度权重，以及基于现有信息对中型市值（$50亿+）板块的内部目标价推测。 以下是每只股票及目标价时间框架的深入拆解，以及“定性”理由： 1. Nebius ($NBIS)：市值 $230亿。极度低估且与基本面脱节。 $70-90亿远期年度经常性收入(ARR)，20-30%息税摊销前利润(EBIT)，来自 Shopify、埃森哲、Cursor、外国政府的企业合同，以及来自 Meta 和微软的超大规模客户合同，为 Nebius 提供了收入可见性。拥有 $48亿+现金，使其免受影响数据中心的信贷紧缩影响。预计 2026 年签约容量达 2.5 GW，可与许多其他公司（如 $IREN 的 2.8 GW）媲美，并击败许多关于容量/电力的论点。由于其许多投资组合公司支持 Tesla 和 Anthropic 等公司，它也具有更高的增长潜力（想想 $MSFT 及其投资组合公司带来的长期防御性）。 此外，随着 $NVDA 第四季度业绩爆发，Jensen 澄清了反对 GPU 折旧的论点，这有助于提升数据中心板块情绪。 1年目标价 $400，基于远期收入/利润率估值达 $1000亿+。 2. Rocketlab ($RKLB)：市值 $220亿。短期高估，长期潜力低估。 Rocketlab 是我与比特币并列的最高确信度5年长线标的。在太空领域，并非赢家通吃，我维持 $3500-5000亿长期目标价以匹配 SpaceX 的最新估值/能力。 目前它被高估。但从纯技术角度看，构建可重复使用火箭构成了惊人且具防御性的护城河，我们正处于其端到端太空产品大规模商业化（可能在 ~2028 年）的早期阶段。 然而，市场正在计入 Flatlite 商业化（如 Starlink）和中程有效载荷（SpaceX Falcon 9）的远期增长。市场也计入远期增长，但对于 Rocketlab 而言，关键在于未来有多远。这始终是一个坚实的买入机会，取决于你对公司执行的耐心程度。 3. Circle ($CRCL) - 市值 $160亿，低估。 对于 Circle，自其市值 $500亿以来我就看空，建议做空 Circle，做多 Coinbase，因为 $COIN 与 Circle 有 50% 的收入分成。 此前因流通股本数字和财报后/12月2日巨大的内部人锁定期导致抛售（类似 $BULL）而被高估。流通股本动态很重要，像 Cathie Wood 这样的 ETF 经理似乎不理解（因此我的警告）。 但现在我们达到了合理的估值水平。我预计 USDC 商业化将继续，鉴于数字资产市场的监管重点，我看到 $CRCL 将接管 Tether 的大部分市值。 话虽如此，一旦我们看到内部人股份重新分配给机构和长期持有者，计入稳定币交易量增长，它完全值得 $300亿+的市值定价。 4. Astera Labs ($ALAB) - 市值 $220亿，估值合理 ALAB 是我中期高确信度选股之一，因为 Mag7 采用其连接技术用于数据中心建设。 增长极高，利润率类似 $NVDA，约为 ~74%，最新估算：$2.3亿/季度（同比增长 101%）。我的论点是，如果 Mag7 依赖某家公司（$NVDA 用于 GPU，NBIS、IREN、CIFR 用于 DC AI 云建设），该公司将连续几个季度大幅超出预期，我们正看到这一点。 尽管全面超出业绩预期，Astera 最近从 $250 抛售回 $140 区间，这提供了一个良好的买入机会。 我维持中期目标价 $250，以在 NVDA 业绩后及 Anthropic $400亿 DC 到 $GOOGL 在德克萨斯州 $500亿 DC + 连接需求创纪录的数据中心建设后恢复。 5. Snapchat ($SNAP) 市值 $130亿，低估。 $SNAP 是我最不喜欢的股票和 CEO 之一（抱歉 Evan）。 然而，我无法反驳基本面的变化。我最近论点帖子的 TLDR 是，他们正在削减来自存储 10 年前记忆/视频的巨大运营支出膨胀，如果你查看他们的 GCP 托管费用，这会侵蚀利润率。 现在他们既减少了该 OPex 成本，又从中增加了收入。我们还有与 Perplexity 的 AI 交易，增加了 $4亿+的额外收入流，如 RDDT。 然而，短期内由于相对于 AI 公司的表现不佳，它正遭受税务收割。在 2026 年 Q1，我预计市场将开始大幅计入新的基本面，该公司将大幅超出预期。 话虽如此，随着市场计入新动态，我预计从此处 1 年内有超过 200%+ 的上涨空间。 5. CIFR ($CIFR) - 市值 $50亿，低估 $CIFR 是 Neocloud 板块中我最喜欢的第二只股票。据记忆，其资产负债表上持有大量比特币，并受到 BTC 价格从 $120k 跌至 $90k 的实质性影响。 然而，我预计随着级联保证金清算结束且机构以低价买入比特币，加密资产价格将在几个月内恢复。 Nebius 位居榜首，因为它拥有完整的 AI-云价值链，具有更高的收入潜力和更强的回报，尽管这迫使他们处理编排、软件和 GPU 生命周期风险，而不是坚持托管。 然而，$CIFR 避免了整个风险面，并拥有 AMZN 和 GOOGL 的支持作为长期收入锚点。它也免受 GPU 采购、管理和折旧的影响。 对于 CIFR 的经济模型，我们获得了一个基于超大规模客户空间、电力和冷却的高利润率、年金结构。经风险调整后，它是该组中最安全的名称之一。但权衡是上涨空间受限。像 10 年、15 年这样的长期租赁减缓了收入爬坡，并相对于从 $1.45亿季度收入到一年内 $21亿的全栈 Neocloud 运营商（如 NBIS）削弱了回报。 话虽如此，一旦市场计入 $AMZN、$GOOGL Fluidstack 收入且比特币价格恢复，我维持 1 个月后的 $28 目标价。 6. Reddit ($RDDT) - 估值适中 来自 WSB 子版块 Wendy's 垃圾桶的我，自然对这个平台有偏见。 然而，Reddit 最初从 $270 的抛售是由于对 ChatGPT 引用的担忧，这是不重要的。现在，最新数据显示引用回来了，但 Reddit 的价格仍停留在 $185（远低于该数字）+ 部分由于宏观因素。 Reddit 是最不臃肿、高利润的社交媒体公司之一。由于年轻和年长受众的网络效应的长期防御性（与 Snap 9亿+ MAU 主要为年轻一代相比），它将长存。 我预计 RDDT 将通过收购（如 $HOOD 交易所）扩大货币化途径，得益于其巨大的自由现金流(FCF)和盈利能力，或者像 Facebook 最初收购 WhatsApp、Instagram、构建 Messenger 一样。这是一只低风险、高增长的股票，因此我维持 8 个月后的 $275 目标价。 7. SMCI ($SMCI) - 低估，市值 $200亿。 $200亿市值是个笑话。无话可说。他们正在实现 $50亿季度收入（当然利润率较低）。然而，市场正在计入公司收入下降。 SMCI 将大部分积压订单推迟到 2026 年 Q2，这与 Neoclouds 到 Mag7 客户的许多数据中心建设相一致。 他们预计明年收入同比增长 50%+，至少 $360亿收入，但鉴于 NVDA 业绩爆发带来的数据中心建设，我预计服务器机架公司如 $DELL 和 SMCI 将在 2026 年 Q2 表现优异。 这就是为什么我利用当前季度的收入滞后延迟，并分配 6 个月后的 $55 目标价。 8. Hims and Her Health ($HIMS) - 低估（市值 $80亿） 个人而言，我仅将 HIMS 用于短期交易突破。我一直不长期持有高于 $50 的股票。 然而，回到 $35，它重置了今年大部分的增长，但收入同比增长 49% 至 $5亿，并产生大量自由现金流。 最被低估的叙事是 Zava 收购。这为 HIMS 平台增加了 130万+用户，并使公司能够扩展到欧盟市场。 类似于 META 收购 Instagram 等公司，扩大基础+货币化，我预计 HIMS 将对 Zava 做同样的事情 + 市场正在计入当前的 Zava 估算数字。 这可能是我信心最低的股票，尤其是 CEO 在离开后出售股份，在 $70 时的 SS 帖子让我味道不好 👀。 但话虽如此，这是一个在 6 个月时间内反弹至 $60 的好机会。 希望你喜欢我的观点。有很多关于价格的帖子，但我试图留下更定性的拆解（+ 部分定量，但省略很多技术内容以便阅读），以帮助散户建立自己的确信度和理解。 建立理解对于创建你自己的内部估值模型很重要，而不是盲目跟随 FinX 发帖人 + 当股价暂时下跌时投降。 如果你留下你的投资组合+集中度，我很乐意讨论更多。
英文原文
Based on the equity ranking table: Here's a deeper analysis of each stock, alongside how I reposition my portfolio to capitalize on the market reset: · $NBIS at $92, PT $400 / 1Y · $RKLB at $43, PT $500 / 5Y · $CRCL at $72, PT $150 / 8M · $ALAB at $143.4, PT $250 / 6M · $SNAP at $8.1, PT $22 / 1Y · $CIFR at $14.8, PT $28 / 6M · $RDDT at $185, PT $275 / 8M · $SMCI at $34, PT $55 / 6M · $HIMS at $35, PT $60 / 6M This is in order of concentration weighting from when posted and internal PT speculation based on existing information for mid-cap ($5B+) sections. Here’s a deeper breakdown on each one and PT timeframe, and a “qualitative”why: 1. Nebius ( $NBIS ): $23B marketcap. Incredibly undervalued and detached from fundamentals. $7-9B forward ARR, 20-30% EBIT, enterprise contracts from Shopify, Accenture, Cursor, foreign governments and hyperscaler contracts from Meta and Microsoft give Nebius revenue visibility. With $4.8B+ in cash, it's isolated from credit tightening affecting data centers. With 2.5 GW expected capacity contracted 2026, it rivals many others eg. $IREN at 2.8 GW, and defeats many of the capacity/power arguments. With many portfolio companies powering companies like Tesla and Anthropic, it also has higher growth potential (think $MSFT with its portfolio companies for longer defensibility). We also had stellar $NVDA earnings going into Q4 with their blowout, Jensen clarifying arguments against GPU depreciation, which helps with DC sector sentiment. $400 1 year price target, $100B+ valuation given forward revenue/margins. 2. Rocketlab ( $RKLB ): $22B marketcap. Overvalued current term, undervalued long term potential. Rocketlab is my highest conviction 5Y long alongside Bitcoin. With Space, it's not winner takes all, and I've maintained $350-500B long term PT to match SpaceX’s most recent valuation/capabilities. As of now, it's overvalued. But it's an incredible + defensible moat from purely a technological standpoint building reusable rockets and we're early in terms of commercialization of their end-to-end space products at scale (likely ~2028). However, we're pricing in forward growth with Flatlite commericalization (eg. Starlink), and medium-lift payloads (SpaceX Falcon 9). The market prices in forward growth as well but it’s more about how long in the future with Rocketlab. It's always a solid buy, depending on how patient you are with company execution. 3. Circle ( $CRCL ) - $16B marketcap, undervalued. With Circle, I've been bear posting it since it was a $50B marketcap, saying short Circle, long Coinbase, given $COIN has 50% revenue sharing with Circle. It was overvalued due to float numbers and massive insider lockups 2-3 days after earnings/Dec 2nd led to a sell-off (like $BULL). Float dynamics matter a lot that ETF managers like Cathie Wood seem to not understand (hence my warnings). But now we're reaching respectable valuation numbers. I expect USDC commercialization to continue and given a regulatory focus in the digital asset market, I see $CRCL taking over a lot of Tether's marketcap. That being said, it's well deserving of a $30B+ marketcap pricing in stablecoin volume growth once we start seeing insider shares redistributed to institutions and long term holders. 4. Astera Labs ( $ALAB ) - $22B marketcap, reasonable valuation ALAB was one of my mid-term high conviction picks, due to Mag7 adoption of connectivity for datacenter buildout. Incredibly high growth and $NVDA-like margins sitting at ~74%, latest er: $230m/q (101% Y/Y growth). My thesis was that if Mag7 is dependent on a company ($NVDA for GPUs) ( NBIS, IREN, CIFR for DC AI cloud buildout), the company will blow away expections quarter after quarter, and we're seeing this. There's been a recent sell-off on Astera from $250 back to $140 marks, depsite beating earning expectations across the board and this presents a good buying opportunity. I maintain a medium term PT $250 for recovery after NVDA earnings and record-high DC buildout from Antrophic's $40B DC to $GOOGL's $50B DC in Texas + connectivity demand. 5. Snapchat ( $SNAP ) $13B marketcap, undervalued. $SNAP is one of my least favorite stocks and CEO's (sorry Evan). However, I can't argue with fundamental changes. A TLDR of my most recent thesis post was that they're cutting their massive opex bloat from memories/videos stored 10 years ago and if you look into their GCP hosting fees, it's cutting in margins. Now they're both reducing that OPex cost and increasing revenue from that. We also have AI deals with perplexity adding $400m+ additional revenue streams like RDDT. However, short term it's suffering from tax-harvesting due to underperformance this year relative to AI companies. In 2026 Q1, I expect the market to start pricing in the new fundamentals Hard. and for this company to beat expectation soundly. That being said I expect over a 200%+ upside 1Y from here with the market pricing in the new dynamics. 5. CIFR ( $CIFR ) - Undervalued at $5B marketcap $CIFR is my second favorite stock in the Neocloud sector. From memory, it holds a lot of Bitcoin on its balance sheet and is materially affected by the selloff in BTC prices from $120k to $90k. However I expect crypto asset prices to recover in a few months once cascading margin liqudations finish and instituions buy-in Bitcoin at low prices. Nebius is top because it owns the full AI-cloud value chain for higher revenue potential and stronger returns, even though it forces them to handle orchestration, software, and GPU lifecycle risk instead of sticking to colocation. However, $CIFR because it avoids that entire risk surface and has backing from AMZN and GOOGL for long term revenue anchors. It also stays insulated from GPU procurement, management, and depreciation. For CIFR's economics we get a a high-margin, annuity structure built on space, power, and cooling for hyperscalers. Risk-adjusted, it’s one of the safest names in the group. But the trade-off is capped upside Long leases like 10Y, 15Y slow the revenue ramp and mute the payoff relative to full-stack Neocloud operators like NBIS that go from $145m quarterly revenue to $2.1B in a year. That being said I maintain a $28 PT in 1 month once market prices in $AMZN, $GOOGL Fluidstack revenue and Bitcoin prices recover. 6. Reddit ( $RDDT ) - Moderate valuation Coming from the Wendy's dumpsters on WSB subreddit, I am naturally biased toward this platform. However, the initial sell-off of Reddit at $270 was due to fears over ChatGPT citations, which was immaterial. Now, recent data shows that citations are back, but Reddit's price still sits at $185 (way below that number) + partly due to macro. Reddit is one of the least bloated, highly profitable social media companies. And it's here to stay due to long term defensibility of the network effect of both younger + older audiences (compared to Snap 900m+ MAU of mostly younger generation). I expect RDDT to scale up monetization avenues through acquisitions like $HOOD (exchanges) due to their massive FCF and profitability or how Facebook originally acquired WhatsApp, Instagram, built out messenger. It's a low-risk, high growth stock, which is why I maintain a $275 PT in 8 months. 7. SMCI ( $SMCI ) - Undervalued, $20B marketcap. $20B marketcap is a joke. Nothing else to say. They're doing $5B quarterly revenue (off lower-margins for sure). However, market is pricing in the company revenue dropping. SMCI quoted majority of the backlog delay to Q2 2026, which aligns with a lot of the DC buildout from Neoclouds to Mag7 customers. They expect revenue to grow 50%+ Y/Y next year, with at least $36 billion revenue, but judging from DC buildout from blowout NVDA earnings, I expect server rack companies like $DELL and SMCI to outperform Q2 2026. This is why I'm taking advantage of revenue lag delays from the current quarter and assigning a $55 PT in 6 months time. 8. Hims and Her Health ( $HIMS) - Undervalued ( $8B marketcap) Personally, I've used HIMS just for short term trading breakouts. And I've been one to not long-term hold the stock above $50. However, back at $35, it's reset most of the year's growth but grew revenue 49% Y/Y to $500m and is producing a good amount of FCF. The most under-priced narrative is the Zava acquisition. This adds 1.3M+ users to the HIMS platform and allows the company to expand to the EU market. Similar to how META acquires companies like Instagram, grows its base + monetizes, I expect HIMS to do the same with Zava + market is pricing in current est. Zava numbers. It's probably my least confident stock out of the bunch, especially leaving me with a bad taste with the CEO selling shares after leaving 👀 on SS posts back at $70. But that being said it's a great rebound opportunity to $60 in a 6 month timeframe. Hope you enjoyed my perspective. There's a lot of x at price posts, but I try to leave a more qualitative breakdown (+ part quantitative but leave out a lot of technical for easier reading) to help retail develop their own conviction and understanding. Building understanding is important to create internal valuation models yourself rather than blindly following along FinX posters + capitulating when stock prices temporarily drop. Happy to discuss more if you drop your own portfolio + concentrations.
-
看好美股基本面改善，逢低买入，若业绩不及预期则退出。
@platochi 的观点也很有道理。当然没有唯一正确的答案，但我核心的逻辑是基本面正在改善（前瞻盈利、利润率、增长），从 $GOOG、$AMZN 等超大规模云服务商(hyperscalers)到 $HOOD、$RKLB 等成长股，全线受益。如果公司开始像 $FISV 那样不及预期并下调指引（当然可能没那么夸张 lol），那我就会考虑退出市场，而不是逢低买入。
英文原文
@platochi has a valid thesis too. There's not one correct one for sure, but my core reasoning is that there's improving fundamentals (forward earnings, margins, growth), across the board from hyperscalers like $GOOG, $AMZN down to growth stocks from $HOOD, to $RKLB. If companies started missing earnings and revising downward like $FISV, (maybe not to that extent lol) then I would be looking to exit the market instead of buying dips.
-
对比NBIS全栈优势与CIFR低风险低上限，看好两者但首选NBIS。
感谢 @accounting_ds 总结我的观点。明天我会发布更深入的全面分析。 $NBIS 和 $CIFR 是我在 Neoclouder（新型云服务商）板块中最看好的两只股票。不过 CIFR 的地位略低于 Nebius。 Nebius 位居榜首，因为它拥有完整的 AI-cloud（人工智能云）价值链，具备更高的收入潜力和更强的回报，尽管这迫使它们承担编排、软件以及 GPU lifecycle（GPU 生命周期）风险，而非仅局限于 colocation（托管服务）。 我看好 $CIFR 是因为它规避了上述所有风险敞口，并拥有 $AMZN 和 $GOOGL 的支持作为长期收入锚点。它还免受 GPU 采购、管理和折旧的影响。 就 CIFR 的经济模型而言，我们得到的是基于为 hyperscalers（超大规模云厂商）提供空间、电力和冷却的高利润率、年金式结构。经风险调整后，它是该组别中最安全的标的之一。但权衡之下，其上行空间有限。 10年、15年的长期租约减缓了收入增长步伐，并削弱了相对于 NBIS 等全栈 Neocloud 运营商的回报，后者能在一年内将季度收入从 1.45 亿美元提升至 21 亿美元。
英文原文
Appreciate, @accounting_ds for summarizing my opinion. I'll post a deeper dive into everything tomorrow. $NBIS, $CIFR are my favorite two in the Neoclouder sector. CIFR sits right below Nebius though. Nebius is top because it owns the full AI-cloud value chain for higher revenue potential and stronger returns, even though it forces them to handle orchestration, software, and GPU lifecycle risk instead of sticking to colocation. I like $CIFR because it avoids that entire risk surface and has backing from $AMZN and $GOOGL for long term revenue anchors. It also stays insulated from GPU procurement, management, and depreciation. For CIFR's economics we get a a high-margin, annuity structure built on space, power, and cooling for hyperscalers. Risk-adjusted, it’s one of the safest names in the group. But the trade-off is capped upside Long leases like 10Y, 15Y slow the revenue ramp and mute the payoff relative to full-stack Neocloud operators like NBIS that go from $145m quarterly revenue to $2.1B in a year.
-
TSM盈利是AI核心指标，超大规模云厂商建设不受NVDA财报影响。
我的观点是，$TSM 的未来盈利（forward earnings）是迄今为止人工智能交易（AI trade）最大的指标，因为它涵盖了从 $GOOGL TPU 生产到 $AMZN，以及 $NVDA、$AMD 等所有超大规模云服务商（hyperscalers），且我们已看到其盈利和利润率大幅超出预期。 即使 $NVDA 的未来营收被大幅下调（我对此表示怀疑），也不会使 $MSFT 到 $IREN、$AMZN 到 $CIFR 等公司已签订的多年度算力（compute）积压订单失效。 我们已看到 Anthropic 与 $MSFT Azure 的算力协议（这将惠及 $NBIS、$IREN 和 $CRWV 等），$GOOGL 近日建设超 400 亿美元的数据中心（此前已与 $WULF、$CIFR/FluidStack 签署托管设施（colo）协议），$NVDA 的财报不会改变超大规模云服务商的建设步伐。 但它确实对整个 AI 交易产生重大影响，并直接影响 $NVDA GPU 板块（例如重度依赖 $NVDA 的 AI 云厂商）。
英文原文
My opinion is that $TSM forward earnings was the biggest indicator of the AI trade so far since they span from all hyperscalers such as $GOOGL TPU prod to $AMZN, as well as $NVDA, $AMD, and we've already seen them blow away earnings + margins. Even if $NVDA forward revenue is quoted heavily downward (which I doubt), it won't invalidate multi-year contracted compute backlog from $MSFT to companies like $IREN or $AMZN to $CIFR made already. We're already seeing Anthropic x $MSFT compute deals with Azure today (which flows down to $NBIS, $IREN and $CRWV others), $GOOGL build out a $40B+ datacenter the other day (they've made colo deals with $WULF, $CIFR / fluidstack previously), and $NVDA earnings won't the change the hyperscaler buildout. But it does have a large impact on the overall AI trade as well + directly affect $NVDA GPU parts of the sector (eg. $NVDA heavy AI clouds).
-
NBIS暴跌后机构逆势增持，基本面强劲，维持$400目标价。
Nebius ($NBIS) 上月暴跌 -33.61%，市值跌至 210 亿美元。 尽管如此，最新的 13F 文件数据显示 $NBIS 的机构持股比例上升： · 38.36% → 44.6% 🟢 (+6.24%) Fintel 的数据则更接近 ~46.3%。 Nebius 是一家市值 210 亿美元的全栈新云(Neocloud)服务商： · 远期年度经常性收入(Forward ARR) 达 70-90 亿美元（同比增长 700%+），EBIT 利润率 30%，为 $MSFT Azure、$META | 从 Cursor 到 Shopify 的企业 | 以及政府提供动力。 · 现金超 47 亿美元，投资组合公司净资产价值(NAV) 超 70 亿美元，这些公司支撑着 Cloudflare、Tiktok、Tesla、Netflix 和 Anthropic。 正如上月所料，机构开始意识到这种一代人一次的增长机遇，因为新云(Neocloud)是 AI 数据中心建设及 Mag7 AI 算力的核心。 Nebius 的机构持股比例一直较低 (<28%)，这源于其非典型的 IPO 方式。对于高价值公司，机构持股通常在 60-80% 之间。 因此，我们看到机械性抛压 + 散户投降 + 保证金清算 -> 通过暗池/大宗交易 -> 机构在低位吸筹。 这种吸筹模式并未改变，我们之前在 Robinhood (S-st) 上也见过同样的故事，其机构持股现已达 74.62%，此前股价从 $20 涨至 $150。 非理性的恐惧和投降是机构从散户手中吸筹的方式。近期的下跌是股价变化，而非基本面恶化。$NBIS 的 1 年目标价仍为 $400。 当远期增长支撑市值预测时，这只是时间和执行力的问题。
英文原文
Nebius ( $NBIS ) crashed -33.61% last month, falling to a $21B market cap. Despite this, new 13F filing data shows $NBIS institutional ownership increase: · 38.36% → 44.6% 🟢 (+6.24%) with Fintel placing it closer to ~46.3%. Nebius is a $21B MC full-stack Neocloud with: · $7-9B forward ARR (700%+ Y/Y), 30% EBIT margins, powering $MSFT Azure, $META | enterprises from Cursor to Shopify | and Governments. · $4.7B+ Cash, $7B+ NAV in portfolio companies that power Cloudflare, Tiktok, Tesla, Netflix and Antrophic. As expected from last month, institutions are starting to realize this once-a-generation growth, as Neoclouds are the center of the AI DC buildout and Mag7 AI compute. The institutional ownership of Nebius was always low <28%, due to how their unusual IPO. With high-value companies, ownership typically ranges from 60-80%. So, we're seeing mechanical flows down + retail capitulation + margin liquidation- > dark pools/block trades -> for institutions to accumulate at lower prices. This type of accumulation pattern has not changed and we've seen this same story before with Robinhood ( S-st ), now at 74.62% ownership, before their rally from $20 to $150. Irrational fear and capitulation is how institutions accumulate off shareholders. The recent drop is a change in stock price, not in fundamentals. The 1Y price target on $NBIS remains $400. When forward growth backs up any market cap projections, it’s a matter of time and execution.
-
AI股因信贷收紧恐慌抛售，建议转向有Mag7合同且低债的新云龙头。
我们正目睹 AI + 数据中心(DC) 股票实时崩盘，$IREN 本周下跌 37.88%，$NBIS 本月下跌 35.27%。 恐慌源于信贷收紧、高息债务、OpenAI 传染效应、宏观因素，以及最重要的：非理性恐惧。 以下是实时发生的情况： 1. 信贷收紧 - 像 $APLD 这样的公司未能成功发行债券（认购不足），最终发行了 23.5 亿美元的高收益债务融资——彭博社 11 月 14 日报道。债券以 97 美分的价格折价出售，收益率约为 10%。 该债券交易的风险主要与 Applied Digital 对 $CRWV 的依赖有关，后者占其合同收入的大部分。随着建设阶段机构关注度增加（例如 Burry 关于 GPU 折旧的做空报告），数据中心更难筹集资金。 融资条件收紧影响了整个通过发行债务为自身建设融资的生态系统。 2. 高息债务 - 由于信贷收紧，$CRWV 为扩张而承担的劣质债务导致每年超过 10 亿美元的损失。这侵蚀了利润率和自由现金流(FCF)。 同样，$APLD 和其他以不良收益率筹集资金的公司也面临同样的未来问题，我们看到相关公司因风险管理而出售。 3. OpenAI 承诺的 1 万亿美元+ 资金没有余额支撑 - OpenAI 承诺超过 1 万亿美元的资本支出(capex)，其中 224 亿美元+ 流向 $CRWV，3000 亿美元+ 流向 $ORCL，他们正在为此建设产能。由于 OpenAI 手头没有余额并采取避险策略，市场现在对这些承诺持怀疑态度。 现在，我们看到这种建设从 Coreweave 等公司向下传染到 $CORZ 和 $APLD，这些公司在建设期间可能依赖 $CRWV 的未来收入。 4. $CRWV 是 AI 基础设施（以及部分 $ORCL）的行业领导者。当行业领导者抛售时，通常其他公司会跟随。 5. 12 月降息概率下降： 除了高息债务外，我们看到 12 月降息概率降至约 53% 的抛硬币水平。较高的利率使得像 $CRWV 这样背负利息债务的公司再融资以增强 FCF 变得困难。 _ 细微差别与重新定位： 我们看到全面抛售。然而，Mag7（$META, $GOOGL, $AMZN, $MSFT）的资本支出和合同已确认，且盈利能力极强。 有一种简单的方法可以安全重新定位，以避免未来收入增长 + 债务周期的问题。 1. 配置与 Mag7 有合同可见性的新云(Neoclouds)： - $NBIS - 190 亿美元 $MSFT 交易，30 亿美元来自 $META，广泛的企业支持 - $CIFR - 55 亿美元与 $AMZN 的 Colo 交易，30 亿美元与 $GOOGL 的交易 - $WULF - 37 亿美元+ 与 $GOOGL 的交易等。 - $IREN - 90 亿美元与 $MSFT 的交易 我们可能会看到更多像 $SLNH, $CLSK 等投机性容量公司下跌，并整合到具有可见企业合同（未来收入下行风险低）的新云中。 2. 配置无/极低息债务且资产负债表强劲的新云： 像 $NBIS 这样的公司因拥有 47.9 亿美元的可转换票据现金及现金等价物，而与当前信贷市场波动隔离。或者像 $CIFR 这样拥有 12 亿美元现金的公司。 3. 避免与 OpenAI 有直接合同而非 Mag7 的公司。 _ 新云是 Mag7 AI 算力的骨干。这一论点没有改变，我们可能会看到一些规模达到 1000 亿美元+ 的公司成为下一个 AWS。 然而，我们看到由 $CRWV, $ORCL 和 OpenAI + 信贷收紧（部分由于 Burry 关于 GPU 折旧的错误机构做空报告）引起的广泛、不分青红皂白的抛售。 对于像 $NBIS 这样拥有 70-90 亿美元远期年度经常性收入(ARR) 和极强资产负债表的公司，这种抛售完全不合理。但对于其他投机性公司，这是一次必要的修正。 一句好话是“散户最后卖出，最后买回”。 如果你认为你可以卖出并在两周后回来，错过抛售后的一个或两个巨大反弹日可能会让你失去大部分恢复收益。 机构可能正在通过暗池/大宗交易购买更好的名字，例如 $NBIS，在这些清洗和保证金清算之后。我们还看到机构所有权从 38%-> 43-46%，尽管股价下跌。 利用这次机会购买好名字，并持有度过这些波动时期。 （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）: 13F 新云申报的机构流向评估： · Nebius ( $NBIS ): 🟢 强烈正面 (7/10) · WULF ( $WULF ): 🟢 高度正面 (8.5/10) · IREN ( $IREN ): 🔴 非常负面 (3/10) · CIFR ( $CIFR ): 🟢 高度正面 (8.0/10) · Coreweave ( $CRWV ): 🟡 中性 (5.5/10) · Cleanspark ( $CLSK ): 🟢 高度正面 (9.0/10) _ TLDR 摘要更新： $NBIS · 机构所有权整体定量增长强劲，由坚实的长期机构买家以及量化基金和对冲基金的高活跃度驱动。 $WULF · 机构基础结构稳定且安全，以 Vanguard 和 BlackRock 等大型纯多头资产管理人为特征，他们对股票持有高信念。 $IREN · 所有权由高频交易员、做市商和量化基金（例如 Jane Street, Citadel）主导。这种结构被视为负面，因为它表明缺乏稳定的短期持有压力和长期机构缺乏信念。 $CIFR · 优秀的机构设置，以纳入主要指数（被动需求）和主动基金的强烈信念购买为标志，Alyeska 的大幅增持尤为突出。 $CRWV · 高交易量表明市场兴趣浓厚，但所有权目前高度集中，做市商存在显著，导致中性评分，暗示流向的风险/回报平衡。 $CLSK · 被认为是该组中最好的机构结构，结合了广泛支持的被动指数纳入和激进的、高信念的主动积累，表明定性风险最低。 _ 评论： 例如，“看多 - Jane Street + Citadel 拥有 x 的 $IREN ”（提示：这不是正面的）。尤其是当长期持有者如 FMR (Fidelity)：212 万股，大幅削减持仓时。 基金类型很重要，例如： 被动（最正面） Vanguard, BlackRock 指数, State Street, Geode 等 多头（正面） Fidelity, T. Rowe, Wellington, BIT 等 对冲基金（中性到正面） Alyeska, Coatue, Millennium 等 量化/做市商（负面） Jane Street, Citadel, Susquehanna, SIG, Two Sigma 等 这是一个主观框架，用于仅从本季度评分流向有多正面（而非整体所有权）。
英文原文
We're seeing a live crash in AI + DC stocks right now with $IREN down 37.88% this week to $NBIS down 35.27% this month. Panic stems from Credit Tightening, High Interest Debt, OpenAI contagion, Macro, and most importantly: Irrational Fear. Here's what happening real time: 1. Credit Tightening - Companies like $APLD failed to sell bonds (under subscribed) and ended up issuing $2.35B in high yield debt offerings - Bloomberg Nov 14th. The bonds were sold at 97 cents on the dollar, a significant discount, with a yield of approximately 10%. The bond deal's risk is primarily tied to Applied Digital's reliance on $CRWV, which accounts for a substantial portion of its contracted earnings. More institutional cation (eg. Burry short report on GPU depreciation) around the buildout makes it harder for DC's to raise funds. Tightening funding conditions affect the entire ecosystem that issue debt to fund their own buildout. 2. High Interest Debt - Due to credit tightening, bad debt, which $CRWV has to fund expansion, causes $1B+ in losses a year. This cuts into margins and FCF. Likewise, $APLD and others that raise funds with bad yields face the same future issues and we're seeing a selloff in related companies for risk-management. 3. $1T+ in OpenAI promised funding without balances - Over $1 trillion+ in capex was promised by OpenAI going to $22.4B+ $CRWV and $300B+ to $ORCL, which they are building capacity for. The market is now skeptical of these promises from OpenAI due to not having the balances on hand and taking a risk-off approach. And now we're seeing contagion on that buildout from companies like Coreweave spread down to $CORZ and $APLD during this buildout that might rely on $CRWV for future revenue. 4. $CRWV is a sector leaders of AI infrastructure (and partially $ORCL). When the sector leader sell-off, usually others follow suit. 5. Rate Cut odds in December Fall: On top of the high interest debt, we're seeing rate cut odds in December drop to a coinflip ~53%. Having higher interest rates make it trouble for interest-debt burdened companies like $CRWV to refinance for stronger FCF. _ THE NUANCE and Repositioning: We're seeing a broad sell-off across the board. However, Capex and contracts from Mag7 ( $META, $GOOGL, $AMZN, $MSFT ) are confirmed with their extreme profitability. There's an easy way to safely reposition to avoid issues with future revenue growth + debt cycles. 1. Position into Neoclouds with contract visibility with Mag7: - $NBIS - $19B $MSFT deal, $3B from $META, broad enterprise support - $CIFR - $5.5B Colo deal with $AMZN, $3B deal with $GOOGL - $WULF - $3.7B+ deals with $GOOGL and more. - $IREN - $9B deal with $MSFT We will likely see more capacity speculative companies like $SLNH, $CLSK, and others fall off and consolidation into Neoclouds with visible enterprise contract (low downside risk of future revenue). 2. Position into Neoclouds with no/extremely low interest debt and high balance sheets: Companies like $NBIS are isolated from current credit market volatility as they have $4.79 billion in cash and cash equivalents from convertible notes. Or companies like $CIFR with $1.2B in cash. 3. Avoid companies with direct contracts with OpenAI that's not Mag7. _ Neoclouds are the backbone of Mag7 AI compute. This thesis has not changed and we'll likely see some scale to $100B+ companies in the future as the next AWS. However, we're seeing a widespread, indiscriminate sell-off caused by $CRWV, $ORCL, and OpenAI + credit tightening (partially due to incorrect institutional short reports from Burry on GPU depreciation). These types of sell-offs are completely not warranted for companies like $NBIS with $7-9B forward ARR and extremely strong balance sheets. But for other speculative companies, it's a well-needed correction. A good quote is “retail is last to sell, last to buy back”. If you think you can sell and come back two weeks later, missing one or two huge rebound days after a selloff can cost you most of the recovery gains. Institutions are probably buying the better names right now eg. $NBIS through dark pools/block trades now after these types of washouts and margin liquidations. We've also seen institutional ownership go from 38%-> 43-46% despite share prices dropping. Use this as a buying opportunity for good names, and hold through these periods of volatility.
-
新云板块因恐慌抛售与基本面脱节，建议持有以捕捉反弹。
是的，像 $CIFR 这样的新云(Neoclouds)股票在盘前走势简直惨不忍睹，而且这还是在其与 $GOOGL 的交易价格之前，盘前已下跌超过 10%。 这也假装 $AMZN 的托管中心(colocation)交易从未发生过。正如你提到的 $NBIS 的情况，正逼近其与 $MSFT 的交易前水平。由于市场崩盘+恐慌性抛售，我们看到整个板块与基本面完全脱节。 目前，我们可能正在目睹新云(Neoclouds)因保证金级联清算(margin cascading liquidations)+短期期权到期导致的机械性下行资金流。我虽然没有使用任何保证金，但鉴于 $IBKR 等经纪商的保证金使用率创历史新高，可能整个板块都存在保证金清算。 但在抛售后的反弹中错过一两个巨大的反弹日，可能会让你失去最大的恢复收益，这就是为什么在波动时期持有更好。我仍然持有 $NBIS 和其他股票，因为我对它们的基本面有信心。
英文原文
Yeah premarket on Neoclouds like $CIFR is just disgusting to look at and it's pre $GOOGL deal price with the 10%+ drop premarket. And that's pretending $AMZN colo deal never happened too. Similar thing as you mentioned with $NBIS, approaching pre $MSFT deal. We're seeing a complete detachment from fundamentals across the board due the market crash + panic selling. Right now, we're likely seeing margin cascading liquidations + mechanical flows downward with short term option expiration on Neoclouds. I'm not using any margin but there's probably margin liquidations across the board given record high usage on brokerages like $IBKR. But missing one or two huge rebound days after a selloff can cost the biggest recovery gains, which is why it’s better to hold through periods of volatility. I'm still holding $NBIS and others since im confident in their fundamentals.
-
高贝塔AI股因宏观恐慌错杀，基本面强劲者终将迎来价值回归。
我们正目睹高贝塔值AI股票板块的崩盘，部分个股在1个月内下跌30-45%+，今日再跌10%+。 这包括直接受益于“七巨头”资本支出的公司： - Nebius ($NBIS)：$META 30亿美元交易 / 80亿美元远期年度经常性收入(Forward ARR)。 - TeraWulf ($WULF)：Anthropic数据中心(<500亿美元) + 与$GOOGL的合资光纤网络(FS)。 - Iren ($IREN)：$MSFT 90亿美元高性能计算(HPC)。 - Cipher ($CIFR)：与$AMZN签署15年55亿美元合同。 这些公司在过去一个月里改善了基本面。 但市场仍在恐慌性抛售从$CRDO (-10.33% 1D)到$CIFR (-7.28% 1D)的股票，尽管它们拥有创纪录的远期收入增长。 这主要是由宏观担忧（政府停摆、12月降息、公司债利差等）和虚假（但微妙）的机构观点（如Burry关于GPU折旧的说法）驱动的，而非基于个股基本面。 这清洗了AI投机性板块（如能源股$OKLO，1个月跌37%）中的泡沫，但高利润率的增长型股票也受牵连。 我之前说过，这让我想起$HOOD从$65跌至$28后反弹至$150，或$ALAB在所有财报超预期后从$100跌至$50再反弹至$245。 当市场充满恐惧，尤其是像现在这样的“AI泡沫”恐慌时，股价有时会与基本面脱节。 但当远期增长能支撑市值预期时，这只是时间和执行力的问题。
英文原文
We're seeing a sector crash in high-beta AI stocks, with names down 30-45%+ in 1M & 10%+ today. This includes direct beneficiaries of Mag7 Capex : - Nebius ( $NBIS ): $3B $META deal / $8B forward ARR. - TeraWulf ( $WULF ): <$50B Antrophic DC + JV FS/ $GOOGL - Iren ( $IREN ): $9B $MSFT HPC - Cipher ( $CIFR ): 15Y $5.5 billion w/ $AMZN. that improved their fundamentals in this past month alone. But markets are still panic selling names from $CRDO (-10.33% 1D) to $CIFR (-7.28% 1D) despite record forward revenue growth. This is driven over fears from macro (shutdown, Dec rate cut, corporate bond spreads, etc) and false (but nuanced) institutional claims eg. Burry on GPU depreciation, rather than individual fundamentals. This wipes away froth from the speculative parts of the AI sector such as $OKLO (energy, -37% 1M), but higher-margin growth names get caught together with it. I've said this before, but this reminds me of the time $HOOD sold off to $28 from $65 then pulled off a rally to $150. Or when $ALAB sold off from $100 to $50 off all earnings beats, then rallied to $245. Sometimes stock prices get detached from fundamentals when there’s fear in the markets, especially with “AI bubble” fears like what's happening now. But when the forward growth backs up any market cap projections, it’s just a matter of time and execution.
-
新云财报解读及反驳伯里GPU贬值论，强调新云逻辑强劲。
今日新云(Neocloud)财报 + 为何迈克尔·伯里(Michael Burry)错了： $WULF - 财报超预期(ER)表现积极，来自520兆瓦(HPC)算力的高性能计算(HPC)容量总合同价值(TCV)达170亿美元。 在德州成立240兆瓦合资企业(JV)用于数据中心(DC)扩张 + 50亿美元融资。 不像$CIFR与AWS的财报超预期那样特别，只是确认（如交易融资）和站点开发。 $CRWV - 因第四季度远期收入较低，财报超预期表现短期看空。 556亿美元积压订单（其中224亿美元来自OpenAI）。小幅盈利超预期13.6亿 vs 12.9亿（同比129%），EBITDA利润率（61%）。 股价因数据中心延迟下跌，这损害了2025财年（50.5-51.5亿 vs 共识52.9亿）。运营利润率下调，正如我预测的那样，因为债务融资产生的利息（~10亿+利息/年）侵蚀了利润率，而其他新云没有同样的问题。 话虽如此，这是一次性问题，鉴于市场是前瞻性的，我们可能会看到任何抛售在下个季度后被买回。 _ 至于迈克尔·伯里关于GPU贬值的争论：他完全错了，但在低估这一点上部分正确。 他的普遍主张：芯片在2-3年产品周期上的资本支出(Capex)不应导致使用寿命的延长 -> 在谷歌的案例中，他们7年前的TPU正以100%的利用率运行。重复一遍，是7年前。 Amin Vahdat表示，谷歌目前生产中有七代TPU硬件，且“七到八年前的TPU拥有100%的利用率”。 “谷歌称TPU需求超过供应，声称8年前的硬件迭代‘100%利用率’” - 数据中心动态 -> 在英伟达(Nvidia)的案例中，A100（PCIe和SXM4变体）仍在运行。它们于2020年推出。现在是2025年。旧型号保持其价值并仍然提供收益。它们不会仅仅因为是补充新模型的附加组件就被核销。 英伟达的A100可以划分为隔离实例以最大化利用率，旧一代加速器仍用于推理或低优先级任务进行货币化，而不是像他所声称的那样被核销。它们实际上具有残值，5年后仍能高价转售。 他在保质期问题上完全错了，但在某些超大规模云服务商如何拉伸利用率概念以证明激进资本支出方面可能是正确的。 这是一个微妙的论点，但新云逻辑依然完好且比以往任何时候都更强。
英文原文
Neocloud Earnings Today + Why Michael Burry is Wrong: $WULF - ER was positive, $17B TCV from 520 MW HPC capacity. 240 MW JV in Texas for DC expansion + $5B financing. Nothing special like $CIFR ER with AWS, just conf (eg. deal finances) and site dev. $CRWV - ER was bearish near term due to lower forward revenue Q4. $55.6B backlog ($22.4B from OpenAI). Slight earnings beat 1.36B vs. 1.29B (129% Y/Y), EBITDA margins (61%). Stock dropped on datacenter delay, this hurts FY 2025 (505-5.15B vs. 5.29B consensus). Operating margins was lowered, which I predicted due to debt financing with interest (~1B+ interest/year) that cuts into margins, while other Neoclouds don't have that same issue. That being said this was a one-off issue, and we'll likely see any sell-off bought back after next quarter since markets are forward looking. _ As for Michael Burry arguing on GPU deprecation: he's completely wrong, but partially right about understating it. His general claim: capex of chips on a 2-3 yr product cycle should not result in the extension of useful lives -> in Google's case, their TPUs from 7 years ago are being run at 100% utilization lol. Repeat 7 years ago. Amin Vahdat, said that Google currently has seven generations of its TPU hardware in production, and that the “seven- and eight-year-old TPUs have 100 % utilization.” "Google says TPU demand is outstripping supply, claims 8yr old hardware iterations have “100% utilization” -Data Center Dynamics -> in Nvidia's case A100 (PCIe and SXM4 variants), are still in operation. They were launched in 2020. It's 2025 now. Older models keep their value and still deliver equity. They don't just get nuked since it's an add-on to compliment newer models. Nvidia’s A100 can be partitioned into isolated instances to maximize utilization, and older-gen accelerators are still monetized for inference or lower-prio tasks rather than getting nuked as what he claims. They actually have residual value and still re-sale for a lot 5 years later. He's completely wrong here on shelf-life, but might be correct in how some hyperscalers stretch the idea of utilization to justify aggressive capex. It's a nuanced argument but Neocloud thesis remains in tact and stronger than ever.
-
感冒简评：META/NBIS错杀强买，警示IREN/ETH，宏观看多AI基建。
11月4日 - 个股分析 + 宏观 清仓抛售：$META • $NBIS • $IBIT 强烈买入：$RDDT • $RKLB • $WLAC • $CIFR • $LTC • $SOL • $CORZ 存储：南亚科 $2408，SK海力士 $HXSCL，美光 $MU 买入：$AMZN • $DELL • $SMCI • $ALAB • $CRDO • $TSM • $AMD • $FLNC • $TE 警示：$IREN • $ETH • $BMNR • $CRCL 量子计算：$IONQ • $RGTI • $QBTS 通常我会逐一列表，列出更多的买卖建议并解释每一个原因，但我感冒了。 所以今天只做极简版，懒得研究更多股票。喝着汤，只关注几只股票并给出更广泛的概述。（如果有人问，我会帮忙解释。） $META： 再次强烈、强烈、强烈建议在财报大跌15%+及盘后下跌时买入。 营收超预期。每股收益(EPS)超预期。远期盈利完好。 如果计入一次性税收，EPS为$7.25对$6.67，但显示为$1.05（大幅不及预期），导致算法立即抛售。随后的抛售中，机械性资金流通常会将价格推得更低，但趋势通常在1-2周内反转。 利用这一点。 像“Meta在AI上花费太多”这样的叙事可能是噪音。每家Mag7公司都在增加AI资本支出，除非你从云/ASIC角度争论，否则Meta并没有做任何异常的事。有时真正的原因更愚蠢，比如算法未能识别一次性税收的标准化。 当低贝塔值的Mag7因非实质性原因下跌16%+时，有很大的获利潜力。 警示类别摘要： 不是叫卖，只是给出公平警示，让人们自行决定。 量子计算：仍无收入证明价格飙升的合理性。某些数据中心股票，即使回调后，也有远期营收+自由现金流支撑估值。 $ETH：高于$3K不是强买入（自$4.8K以来立场相同）。 $BMNR：ETH国库策略；以溢价追踪ETH价格。除非有国有化潜力（如$MSTR），否则国库策略不好。 $CRCL：下个月有巨大、巨大、巨大的股份解禁。 $IREN：$MSFT合同在标题上看起来很棒（$9B+营收，超大规模云厂商交易），但实际上利润率很差。 交易不是纯托管；$IREN吸收了~$5.8B的GPU+辅助设备资本支出。 毛利率：最好只有30%出头，自由现金流盈利能力可能降至个位数。相比之下，$NBIS的交易结构更好（$11.6M/MW/yr vs $9.7M/MW/yr）。 交易验证了超大规模云厂商将与IREN签约（正面），但交易本身很糟糕。不是卖出，只是警示那些因标题而FOMO的人。 如果你想FOMO进什么，$CIFR（AWS合同）更有意义。 _ 宏观观点： 市场一直在抛售高贝塔值股票，可能只是正常回调。 逆风： 鲍威尔暗示第三次降息不确定 -> 重新定价。 政府停摆可能延长（短期逆风）。历史上，重新开放后回调会迅速反转。 事件市场定价11月14日为结束日期（还有9天）。 如果像$NBIS（-7.9%）这样拥有惊人远期营收支撑的名字没有实质性变化，大跌时是极好的买入机会。还有像$RDDT这样发布了好财报却跌了7%的股票，也是强烈的买入机会。 泡沫在美联储收紧时破裂。我们已经有了两次降息，还有一次很可能。不要跟随那个预测了过去1次崩溃中5000次的《富爸爸穷爸爸》家伙。 AI ≠ 泡沫 yet - 由Mag7财报、真实营收和利润支撑。但如果OpenAI继续在没有资金的情况下过度承诺合同，以后可能开始像泡沫（1-2年后的投影风险）。 摘要：随着降息后的宽松，市场可能趋势向上。 数据中心建设保持完整（Mag7财报显示巨大的AI资本支出）。 回调中保持强劲的板块：存储 • 能源 • 半导体 • 连接 • 新云 保持多头，在这些回调中建立仓位
英文原文
November 4th - Stock Analysis + Macro Fire Sale: $META • $NBIS • $IBIT Strong Buys: $RDDT • $RKLB • $WLAC • $CIFR • $LTC • $SOL • $CORZ Memory: Nanya $2408, SK Hynix $HXSCL, $MU Buy: $AMZN • $DELL • $SMCI • $ALAB • $CRDO • $TSM • $AMD • $FLNC • $TE Warnings: $IREN • $ETH • $BMNR • $CRCLQuantum: $IONQ • $RGTI • $QBTS Normally I'd go down the list, have more buy/sells, and talk about every single explanation but I caught a cold. So just doing an extremely light version today and didn’t feel like researching more stocks. Just focusing on a few with a broader overview while I drink some soup. (Will help explain if someone asks though.) $META: Once again, Strong, Strong, Strong Buy on the 15%+ earnings drop and post-ER drop. Beat on Revenue. Beat on EPS. Forward earnings intact. If you factor in one-time tax, EPS was $7.25 to $6.67, but it showed as $1.05 (huge miss), which led to algorithmic sell-offs immediately after. Following sell-offs, mechanical flows often push prices lower, but trends usually reverse in 1–2 weeks. Take advantage of this. Narratives like “Meta spending too much on AI” are likely noise. Every Mag7 company is increasing AI capex, Meta isn’t doing anything unusual unless you argue from a cloud/ASIC standpoint. Sometimes the real reason is dumber like algorithms not recognizing one-time tax normalization. Lot of profit potential when lower-beta Mag7 drops 16%+ on nothing material. Warnings Category TLDR: Not calling sells, just giving fair warnings so people can decide for themselves. Quantum: Still no revenue to justify price surges. Certain data center stocks, even after corrections, have forward revenue + FCF to back valuations. $ETH: Not a strong buy above $3K (same stance since $4.8K). $BMNR: ETH treasury play; tracks ETH price with a premium. Treasury plays aren’t good unless potential for nationalization (eg. $MSTR). $CRCL: Massive, massive, massive share unlock coming next month. $IREN: The $MSFT contract looks great in headlines ($9B+ revenue, Hyperscaler deal), but in reality, margins are poor. Deal isn’t pure colo; $IREN absorbs ~$5.8B capex in GPUs + ancillary equipment. Gross margins: low 30s at best, can drop to single-digit FCF profitability. By comparison, $NBIS had a structurally better deal ($11.6M/MW/yr vs $9.7M/MW/yr). The deal validated that hyperscalers will sign with IREN (positive), but the deal itself was terrible.Not a sell, just a warning for people FOMOing the headline. If you want to FOMO into something, $CIFR (AWS contract) makes more sense. _ Macro View: Markets have been selling off high-beta stocks, likely just a normal correction. Headwinds: Powell signaling a third rate cut isn’t certain -> repricing. Gov shutdown likely extends (short-term headwind). Historically, corrections reverse quickly after reopen. Event markets price Nov 14th as end date (9 days out). If there’s no material change for names like $NBIS (-7.9%), backed by insane forward revenue, great buy on the dip. Also stuff like $RDDT that posted great ER but dropped 7.% also strong buying opportunity. Bubbles pop under Fed tightening. We've had two rate cuts, and another still likely. Don’t follow the Rich Dad Poor Dad dude who predicted 5000 of the last 1 crashes. AI ≠ bubble yet - backed by Mag7 earnings, real revenue, and profits. But if OpenAI keeps over-promising contracts without funds, it could start resembling one later (projection risk 1–2 years out). TLDR: Markets likely trend higher as easing follows rate cuts. Data center buildout remains intact (Mag7 earnings show huge AI capex). Sectors that stay strong on corrections: Memory • Energy • Semi • Connectivity • Neoclouds Stay long and build positions on these dips
-
Neocloud获微软AWS大额订单，算力需求旺盛，板块极度看涨。
Neocloud 名单已更新： $IREN - IREN 与微软达成 97 亿美元 GPU 云合同，包含 20% 预付款，将在五年内部署 NVIDIA GB300。 $CIFR - CIFR 与亚马逊云科技 (AWS) 达成 55 亿美元、15 年的租赁协议。 随着超大规模云服务商 (Hyperscalers) 的新合同涌入，Neocloud 板块变得极度看涨，这显示了对算力 (Compute) 的无限需求。 我在原始论点中预计 $ORCL、$META 和 $AMZN 会达成交易，看来 AWS 已与 $CIFR 入场，预计与其他 Neocloud 的交易也会接踵而至。$META 最近也与 $CRWV 签署了 140 亿美元的交易，从其不断扩大的 AI 资本支出 (Capex) 来看，我们可能会看到更多与 Neocloud 的交易。 这对整个板块来说都是极度看涨的。 我附上了我的个人梯队排名 (Tierlist) 供娱乐，但随着有更多时间分析交易和利润率，数据中心/矿工板块可能会迎来全面重评。
英文原文
The Neocloud List got updated: $IREN - IREN secured a $9.7B GPU cloud contract with Microsoft, including 20% upfront prepayment, to deploy NVIDIA GB300s over five years. $CIFR - CIFR secured a $5.5 billion, 15-year lease agreement with Amazon Web Services (AWS). The Neocloud sector just got insanely bullish with new contracts coming in from hyperscalers and this just shows insatiable demand for compute. I expected $ORCL, $META, and $AMZN to make deals in the original thesis and looks like AWS entered the fray with $CIFR, with likely more deals incoming with other Neoclouds. $META also signed a $14B deal with $CRWV recently and from their ER scaling AI Capex, we'll likely see other deals with Neoclouds. This is insanely bullish across the whole sector. I attached my personal tierlist for fun, but we'll likely see a re-ratings in data centers/miners across the board when there's more time to analyze deals and margins.
-
推荐年末满仓MINT组合（META/IBIT/NBIS/TSM），利用错杀与高非对称回报。
只有四只股票： 🍬 MINT 拥有极其高、非对称的回报。 意味着低资本风险，高上行回报。 这仅适用于2025年11月和12月。 就是它... MINT 关于为何在2025年底满仓MINT的解释： M = $META I = $IBIT N = $NBIS T = $TSM 将这些神圣的成长型公司纳入你的投资组合，你可以让呼吸更清新，从而吸引Sydney Sweeney： Mag7。加密货币。AI数据中心。半导体。 $META - Meta ___________ 愚蠢的抛售。 营收超预期。每股收益超预期。强劲指引。 当你看到营收512.4亿美元（同比+26.25%），净利润27.1亿美元（同比-82.73%）时，纸面上看起来极其糟糕。 🟢 营收：512.4亿美元 vs. 预期494.1亿美元 剔除一次性税收影响 → 🟢 每股收益：7.25美元 vs. 预期6.67美元 包含一次性税收影响 → 🔴 每股收益：1.05美元 vs. 预期6.67美元 我猜测最初的抛售是由算法驱动的，随后由机械性资金流引起。许多高频交易(HFT)和新闻解析系统会对直接从数据源（彭博、路透等）拉取的头条每股收益数字做出反应。 如果报告的GAAP每股收益（1.05美元）与共识预期（6.67美元）并列显示，系统会将其标记为大幅不及预期。显然这是推测性的，但许多模型并未编程以归一化处理一次性税收费用（这是新的），也不会知道去查看实际超预期的调整后每股收益（7.25美元）。 现在新闻声称抛售是由于AI资本支出惊喜所致。如果接受这个说法，因“AI资本支出增加”而下跌15%+纯粹是叙事，毫无意义，因为$MSFT、$GOOGL、$AMZN、$TSLA以及每一个Mag7都在增加资本支出。 $META做同样的事却遭到抛售说不通，除非人们认为它只是进入一个没有投资回报的黑洞（市场不同意，因为AI基本上在推动$SPY）。 一旦周末有时间让人们消化实际发生的情况，恢复将是极强的买入机会。误读+愚蠢的叙事可能导致短期过度反应（1-2个交易日），通常会被逆转。 $IBIT - 比特币 ___________ 它是比特币。目前位于11万美元。 随着美国政府的支持，以及持续的货币通胀+资金流向新一代人，没有比BTC更好的通胀对冲工具了。 Polymarket仍然定价比特币在2个月内上涨40%至13万美元，即使它上涨几个百分点，你也可以在加密货币和底层资产上获得上行空间。 $NBIS - Nebius ___________ 核心业务估值：310-365亿美元 分部加总（现金、资产、投资组合）：~106亿美元（削减40% = ~63亿美元） 基准情景估值：~390亿美元 → 即使零增长也有高上行空间 核心业务有望在FY26-27实现~50亿美元年度经常性收入(ARR)，60-70%毛利率，30%息税折旧摊销前利润(EBITDA)目标。 通过再达成一个超大规模云厂商协议+SMB扩张，实现80-120亿美元ARR可能证明~1000亿美元估值合理。 无有毒债务，高GPU利用率，全栈架构 → 更高的经营杠杆。 你可以看我引用的要点比较，以便你对$NBIS vs $IREN做出自己的判断。 然而，我将始终维持$NBIS拥有包括$IREN到$CIFR在内的最高非对称回报，因为现有的超大规模云厂商协议、企业客户（Shopify, Accenture）、政府客户、1GW容量、来自全栈的高毛利率等。 这并不是说来自$WULF的Fluidstack + Google协议以及其他容量方面没有高上行空间。 Nebius的下行风险最低。而且它真的、真的有高上行空间。 $TSM ______________ 它是AI交易的中心。 他们已经指引利润率增加——这是天文数字般的。 营收：331亿美元 vs. 315亿美元预期。 每股收益：2.92美元 vs. 2.59美元预期。 指引：322-334亿美元 vs. 320亿美元预期。 他们超收了营收指引，利润率增加，他们印钱的速度真是天文数字。 财报前交易在305-310美元，现在回到300美元。 这是一家万亿+美元的公司，同比增长30%+且利润率提升——到底怎么回事。 我之前在$GOOGL 145美元和$HOOD 20美元时说过同样的话。他们印钱。 没必要辩论$ALAB vs $CRDO，$IREN vs $NBIS，$AMD vs $NVDA。 $TSM是整个AI建设周期的中心，所以这只是等待游戏。 _ 就是它$MINT。 享受歌曲的预告，有点炸。🎶
英文原文
There’s only Four stocks: 🍬 MINT That have ABSURDLY HIGH, asymmetrical return. Meaning Low Capital Risk. High Upside Return. This is only for Nov + Dec 2025. Here it is... MINT An explanation of why to full port MINT for end of year 2025: M = $META I = $IBIT N = $NBIS T = $TSM With these four holy growth companies in your portfolio, you can make your breath smell better so you can pull Sydney Sweeney: Mag7. Crypto. AI Data Center. Semiconductor. $META - Meta ___________ Stupid sell-off. Revenue Beat. EPS Beat. Strong guidance. It looks extremely bad on paper when you look at Revenue $51.24B (+26.25% Y/Y), Net income $2.71B (-82.73% Y/Y). 🟢 Revenue: $51.24B vs. est. $49.41B Without one-off tax → 🟢 EPS: $7.25 vs. est. $6.67 With one-off tax → 🔴 EPS: $1.05 vs. est. $6.67 My guess was the initial sell-off was algorithmically driven, then caused by mechanical flows. Many high-frequency trading (HFT) and news-parsing systems react to headline EPS numbers pulled directly from data feeds (Bloomberg, Reuters, etc.). If the reported GAAP EPS ($1.05) is printed alongside a consensus of $6.67, the system flags it as a massive miss. Obviously this is speculative, but many models aren't programmed to normalize for one-time tax charges (which is new) and won’t know to look at adjusted EPS ($7.25), which actually beat. Now news are claiming the sell-off was due to AI capex surprise. If we go with this, -15%+ on “AI capex spending increasing” is just pure narratives and means absolutely jack since $MSFT, $GOOGL, $AMZN, $TSLA, and every single Mag7 is increasing on capex. $META doing the same and selling off doesn’t make sense, unless people believe it’s just going into a black hole with no ROI (which markets disagree since AI is basically running $SPY). Extremely strong buy on recovery once there’s time over the weekend for people to digest what actually happened. Misreads + stupid narratives can cause short-term overreactions (1–2 sessions), often reversed. $IBIT - Bitcoin ___________ It’s Bitcoin.And sitting at $110k. Bitcoin is here to stay with US government support, and with all the ongoing currency inflation + flow of money going to newer generation, there’s no better hedge against inflation than BTC. Polymarket still prices Bitcoin 40% to $130k in 2 months’ time, even if it moves up a few percent you can have upside on CCs and upside on the underlying asset. $NBIS - Nebius ___________ Core business valuation: $31–36.5B Sum-of-parts (cash, assets, portfolio): ~$10.6B (slashed 40% = ~$6.3B) Base case valuation: ~$39B → high upside even with zero growth Core business on track for ~$5B ARR (FY26–27), 60–70% gross margins, and 30% EBITDA target. Hitting $8–12B ARR (via one more hyperscaler deal + SMB expansion) could justify ~$100B valuation. No toxic debt, high GPU utilization, full-stack architecture → higher operating leverage. You can look at the bullet point comparison that I quoted so you can make your own judgement on $NBIS vs $IREN for example. However, I’ll always maintain $NBIS has the highest asymmetrical return over anything including $IREN to $CIFR due to existing hyperscaler deals, enterprise clients (Shopify, Accenture), Government Clients, 1GW capacity, high gross margins from full-stack, and many others. That’s not saying there’s not high upside from $WULF from Fluidstack + Google deals, and others from capacity. The downside risk on Nebius is the lowest. And it has really, really high upside. $TSM ______________ It’s the CENTER. OF. THE. AI. TRADE. They’ve already guided INCREASING MARGINS - which is astronomical. Revenue: $33.1B vs. $31.5B est. Earnings per Share: $2.92 vs. $2.59 est. Guidance: $32.2B – $33.4B vs. $32.0B est. They’ve beat revenue guidance, increasing margins, and it’s just astronomical how much money they’re printing. It was trading at $305–310 pre-earnings, now back at $300. It’s a $1T+ company that grew 30%+ Y/Y with their margins — what the actual. I said this about the money printer $GOOGL back at $145 and $HOOD back at $20. They print money. There's no need to debate $ALAB vs $CRDO, $IREN vs $NBIS, $AMD vs $NVDA. $TSM the center of the whole AI buildout, so it’s just a waiting game. _ There is it $MINT. Enjoy a teaser of the song, kinda a banger. 🎶
-
AI工具在分析Neoclouds时存在缺陷，作者更倾向手动整合信息并采用估值建模。
我发现像 @AskPerplexity、Grok 等 AI 工具在研究 Neoclouds（新型云服务商）和每日新闻频发的板块时，结果往往极不准确。它们会遗漏大量细节，例如 $NBIS 的白皮书中有关于与 $CRWV 等对比的利用率信息，这些数据对 AI 不可见（暗示了 GPU 利用率带来的毛利率增长）。当 Meta 与 $CRWV 达成 140 亿美元交易，且上周 $AMZN、$MSFT 等发布财报显示 AI 资本支出增加时，AI 可能会忽略那些未直接提及 $NBIS 但暗示超大规模云厂商交易增加的顺风因素。或者几天前的额外降息提振了涉及债务的 Neoclouds 的前瞻性盈利。亦或是关于 Clickhouse 等子公司或投资组合公司的新信息，AI 可能无法将其串联起来。FinX 散户在将这些信息拼凑成前瞻性增长方面做得更好，鉴于 AI/Neocloud 的总可寻址市场(TAM)具有高度投机性和快速增长，很难建模。这也是我采用估值方法的原因（在我之前关于 $NBIS 如何达到 1000 亿美元市值的帖子中），但如果看 $UPWK 等更标准的业务，我的估值建模方式则不同。
英文原文
I've actually found AI tools like @AskPerplexity, Grok and others to be extremely off when looking at Neoclouds and sectors with new news every day. They miss a lot of details, eg. with $NBIS you have utilization information from Whitepapers comparing Nebius to $CRWV and others that aren't publicly viewable to AI (which insinuated growing gross margins from GPU utilization). When Meta struck a $14B deal with $CRWV and you have earnings report last week from $AMZN, $MSFT, and others you have increased capex spend into AI, and they might miss tailwinds that don't mention $NBIS directly with increased changes of more hyperscaler deals. Or when there's an additional rate cut a few days ago that boosts forward earnings especially with Neoclouds that involve debt. Or when there's new information about subsidiaries or portfolio companies like Clickhouse that AI might put 1+1 together with. FinX retail does a lot better job with piecing all this information together into forward growth and AI/Neocloud TAM is really hard to model given how speculative and rapidly growing it is. That's kinda why I go with valuation approaches (in my previous post on how you get to $100B MC for $NBIS), but if you look at $UPWK and more standard businesses. I do valuation modelling differently
-
看好全栈新云$NBIS利润率优势，减持矿工股。
市场开始定价： 像 $NBIS 这样的 AWS 式全栈新云（Neocloud）与 $CIFR、$BITF、$CLSK 等矿工之间的分化。 这就是我卖出 $IREN、$WULF 等并集中持仓 Nebius 的原因。 为什么？ 💹 利润率 > ⚡️ 产能。 就在今天： Nebius 的 CTO 表示： “市场上几乎没有真正的‘新云’。所谓的新云不过是一堆裸金属服务器，上面堆砌了大量第三方软件。” 耸人听闻？完全不是——如果你认真对待“云”这个概念，而不仅仅是堆砌硬件盒子。 Danila 解释了裸金属运营商很快就会发现 Nebius 已经在大规模层面解决的那些挑战。 这正是我指出 $NBIS 拥有最高非对称上行潜力的原因，因为矿工在规模化时可能会在利润率上挣扎。 我们还看到 Nebius 联合创始人 Roman Chernin 在《福布斯》的文章中总结，AI 经济需要一种新的、全栈的、AI 原生基础设施，而不是拼凑起来的系统： “你可以从一个供应商购买机架，将它们连接起来……但从长远来看，经济性、灵活性和速度取决于你是否控制自己的全栈。” “如果你控制全栈，你就控制产品、性能和经济效益。” _ 这并不意味着像 $WYFI、$APLD、$WULF 等其他高性能计算（HPC）公司没有增长空间，新云板块才刚刚开始上涨（尤其是当我们看到 $META 或 $AMZN 的财报时）。 然而，在投机阶段，我们已经为矿工的产能建设定价——大多数已经上涨了 50–200%+。 但谈到执行层面，那些控制全栈的人将在利润率上获胜。 我们看到 $CRWV 亲自体会到了这一点，花费数十亿美元进行软件收购，利用率仍低于 $NBIS（根据白皮书）。 我们也看到一家市值 8000 亿美元的超大规模云厂商（Oracle）未能建立其全栈，毛利率仅为 14%。 如果 Oracle 都无法完全整合其软件，CoreWeave 仍在花费数年时间尝试这样做，小型矿工如何转型？ “仅 Type-1”运营商的论点将艰难地发现规模化时的利润率压缩是什么样子的。 正如《福布斯》文章总结： “归根结底，如果你只是连接盒子或在无法控制的基础设施上构建服务，你在规模和效率的游戏里将受到限制。” $NBIS 今天就已经拥有全栈。
英文原文
Markets are starting to price in: A divergence between AWS-style full-stack Neoclouds like $NBIS and miners such as $CIFR, $BITF, $CLSK, and others. This is the reason I sold off $IREN $WULF and others, and consolidated into Nebius. Why? 💹 Margins > ⚡️Capacity. Just today: The CTO of Nebius stated: “Almost no Neocloud in the market is actually a cloud. As in a bunch of baremetal with a ton of third-party software slapped on top.” Provocative? Absolutely not — if you take the term “cloud” seriously, not just as slapping on a bunch of hardware boxes. Danila explained how baremetal operators will soon discover the challenges Nebius has already solved at scale. And that’s exactly why I stated $NBIS has the highest asymmetrical upside, since miners may struggle with margins at scale. We’ve also seen a Forbes article by Nebius Co-Founder Roman Chernin, summarize how the AI economy requires a new, full-stack, AI-native infrastructure, not the stitched-together systems: “You can buy racks from one supplier, cable them together... but in the long term, economics, flexibility and speed depend on you controlling your own stack.” “If you control the stack, you control the product, the performance, and the economics.” _ This doesn’t mean there’s no room for other HPC companies like $WYFI, $APLD, $WULF, and others to grow, and the Neocloud segment is just beginning its run-up (especially when we look at $META or $AMZN earnings). However, during the speculation phase, we’ve already priced in capacity buildout with miners - most are already up 50–200%+. But when it comes to execution, those who control the full stack will win on margins. We’ve seen $CRWV figure that out firsthand, spending billions on software acquisitions and still maintaining lower utilization than $NBIS (as per the whitepaper). We’ve also seen how an $800B hyperscaler (Oracle) failed to build out its stack, running at 14% gross margins. If Oracle couldn’t fully integrate its software, and CoreWeave is still spending years trying to do so, how will small miners pivot? Arguments from “Type-1 only” operators will find out the hard way what margin compression at scale looks like. As the Forbes article concludes: “At the end of the day, if you’re just cabling the boxes or building service on top of infrastructure you don’t control, you’re limited in the game of scale and efficiency.” $NBIS already has that full stack today.
-
认为NBIS核心业务被低估，机构持仓将大幅上升。
抛开子公司不谈，仅看核心业务，我认为 $NBIS 被低估了。Nebius 在短短1年内从1亿美元 ARR（年度经常性收入）增长到10亿-15亿美元 ARR，这种增长速度令人震惊。这是小公司赋能 $GOOGL Cloud、$MSFT Azure、$AMZN Web Services 等基础设施的世代性机遇，而这些是 Mag7（美股七巨头）中最盈利的板块。散户并未看到这一机会。机构肯定看到了，我不惊讶于 Nebius 的机构持股比例从38%在6个月内升至65%。
英文原文
Subsidiaries aside and just the core business, I think $NBIS is undervalued. It is mindblowing how Nebius is growing from $100m ARR to $1B-1.5B in just 1Y. And this is a generational chance for small companies to power the infrastructure of $GOOGL Cloud, $MSFT Azure, $AMZN Web Services and others, the most profitable segments of Mag7. Retail doesn't see this opportunity. Institutions definitely do and I would not be surprised if Nebius was 65% owned by institutions from 38% in 6 months time.
-
建议做多NBIS/WLAC及TSM/AMZN看涨期权，认为当前是最佳做多时机。
@capybara_830 说实话，$NBIS、$WLAC，然后将 $NVO 拆分并买入 6 个月后的 $TSM 和 $AMZN 看涨期权，听起来太完美了。目前已有两次降息预期，且市场刚经历回调，正进入季节性表现最佳的月份。如果现在有任何时机做多，那就是现在。
英文原文
@capybara_830 Honestly $NBIS, $WLAC, then splitting $NVO into $TSM + $AMZN calls 6 months out sounds beautiful. There’s 2x rate cuts and a recent correction going into the best seasonal months. Now’s the best time to go long if any.
-
博主发布降息周交易清单，强烈看好TSM及Neocloud板块，建议做多。
10月20日，重要的降息交易周。 个人想法和解释： 🛝 = 波段交易(Swing Trade) 🐈 = 催化剂交易(Catalyst Trade) 🎇 = 2026年交易，已进行税务收割(Tax Harvested) 清仓卖出(Fire Sale) 🔥 $NBIS 强烈买入(Strong Buy) $TSM $AMKR $WLAC $AMZN $LTC 🐈 $RDDT $HIMS 🛝 $IBIT $ALAB $CRDO $SMCI $FLY 🎇 $SNAP 🎇 $ETOR 🎇 $LULU 🎇 买入(Buy) $AMD $HOOD $RBRK $UNH $TGT 🐈 $IREN 🐈 $WYFI $WULF $CIFR $SLNH $BITF $GLXY $FLNC $MU （跳过“持有”，因为我过去提到的其他股票，既然没有变化，大概就继续持有）。 卖出(Sell) $ETH $BMNR $PL $BLSKY $RGTI $OKLO $IONQ $QBTS _ 宏观方面，距离降息（约97%概率）还有9天。市场处于恐惧模式。这是建立多头头寸且不减仓的理想时机。 清仓卖出 _ $NBIS - 无需多言，我仍维持2026年牛市情景下$400的目标价，基于40-60亿美元+的前瞻性收入及约60-75%的毛利率，以及另一份可能的超大规模云厂商合同（如 $META）。 周五下跌10%+是机械式对冲和做市商(MM)钉住价格所致。尽管波动剧烈，价格仍卡在$113.5。预计周一空头对冲将平仓（鉴于做市商买入看跌期权并做空看涨期权 -> 临近到期时大量做空），价格将回升。我在下跌时买入六位数的看涨期权，因为基本面没有实质性变化。 强烈买入 TSM - 天哪，请务必将其纳入投资组合。这是一台印钞机，每年营收增长38-40%的同时毛利率还在提升，简直不可思议。财报大超预期后股价回调，这是我生命中最容易的多头机会之一。 AMKR - 我尚未持有，但计划因TSM在亚利桑那州的参与以及其作为美国供应链大伙伴的潜力而加仓（随着美国推动TSM向美国晶圆厂+制造转移）。 WLAC - Neocloud SPAC IPO，上行空间巨大。我最近常提到这个，它可能是估值最好的Neocloud之一，且已有不错的利润率（并非来自矿工转型，后者不确定性稍大）。他们与Fluidstack合作，我预计若获得Mag7合同，估值将重估500%+。 AMZN - $213太离谱了哈哈。我不明白在牛市中它为何年初至今下跌-3%。 LTC - 受加密货币清算和政府停摆延迟ETF影响。现在是买入并等待ETF获批的好时机。 RDDT - 跌至$190是很好的回调。我以为$200是底部，结果跌得更低。关于ChatGPT较少引用它的新闻引发了大幅抛售，我认为这影响甚微。 HIMS - CEO减持导致下跌14%+。所有者经常卖股，对公司基本面影响不大，仅影响短期情绪。预计会反弹。 IBIT - 比特币$108k是很好的入场点，它已在$110k-$120k之间震荡一段时间，低于此价位通常很好。 ALAB - 上次我说过，因新竞争对手新闻而抛售过度。它已在市场上与AVGO竞争哈哈，拥有类似NVDA的利润率，同比增长数百%，Mag7在数据中心建设中正在使用它。 CRDO - 逻辑与ALAB类似，随Astera抛售但幅度稍小。 SMCI - 应因明年55%+的营收增长而重估。我早先怀疑这些预测，但随着数据中心增长，看起来变得现实了。 FLY - 这是中程有效载荷的博弈。人们怀疑Fly的执行能力，但NOC联合开发中程载荷消除了很多风险（并在与Falcon9竞争时可能重估500%+）。 SNAP - 我在早期的深度研究(DD)帖子中计算过Snap记忆功能的变现能力，目前完全未定价。它季度营收$13亿+，市值仅$130亿哈哈，通过增加营收+降低Google OPEX成本获得的自由现金流(FCF)非常惊人。 ETOR - 大部分为现金，以IBKR的速度增长，受税务收割影响。 LULU - 受税务收割+Alo、Vuori等竞争影响。但季节性应该不错，且现在市盈率极低。 买入 AMD - ChatGPT下AMD订单，ORCL建设AMD数据中心。随着成为$NVDA潜在竞争对手，预计明年将重估。仍认为Nvidia将主导，但鉴于其4.5T市值，即使AMD只占一小部分份额，也有很大追赶空间。 HOOD - 10%+回调后看起来好多了。可能复刻PLTR的走势。 RBRK - 之前做过深度研究，作为网络安全公司，在下跌中看起来更好，该领域倍数极低。只需削减营销，客户粘性高。 UNH - 美国医疗很烂但不会消失。认为Warren等人知道这一点。 TGT 🐈 - 下月分红，大股息股票。我认为现在是大量建仓的好时机。 IREN 🐈 - 巨大的GW，预计获得Mag7或类似交易。 WYFI - 任何Neocloud都是买入（例如见关于Mag7将收入导向这些小型10亿-50亿公司的论点）。 WULF - Neocloud博弈 CIFR - Neocloud博弈 SLNH - Neocloud博弈 BITF - Neocloud博弈 GLXY - Neocloud衍生品博弈 FLNC - Neocloud能源博弈 MU - 中国风险解除，内存在那里有巨大市场，内存也可能在数据中心建设中重估。 _ 卖出 ETH - 不喜欢$4k+的以太坊 BMNR - 如果我不喜欢这些水平的以太坊，持有国库公司也没意义 PL - 低营收，太空股（极高估值） BLSKY - 低营收，太空股（极高估值） RGTI - 量子泡沫 OKLO - 核能泡沫 IONQ - 量子泡沫 QBTS - 量子泡沫 _ 快速宏观提示： -> 9天后降息，概率~97%。预期抢跑，做多。就这些。
英文原文
October 20th, Important Rate Cut Trading Week. Personal thoughts and explanations: 🛝 = Swing Trade 🐈 = Catalyst Trade 🎇 = 2026 Trade, Tax Harvested Fire Sale 🔥 $NBIS Strong Buy $TSM $AMKR $WLAC $AMZN $LTC 🐈 $RDDT $HIMS 🛝 $IBIT $ALAB $CRDO $SMCI $FLY 🎇 $SNAP 🎇 $ETOR 🎇 $LULU 🎇 Buy $AMD $HOOD $RBRK $UNH $TGT 🐈 $IREN 🐈 $WYFI $WULF $CIFR $SLNH $BITF $GLXY $FLNC $MU (Skipping Hold, since any other stock I've mentioned in the past, it probably just hold it since nothing's changed). Sell $ETH $BMNR $PL $BLSKY $RGTI $OKLO $IONQ $QBTS _ So macro wise, we are 9 days away from (~97% or so rate cut). Market is in fear mode. This is the ideal time to go long and not cut positions. Fire Sale _ $NBIS - Needs no explanation, I still maintain $400 PT on a bull case 2026 due to 4-6B+ forward revenue off ~60-75% gross margins, and another likely hyperscaler contract (eg. $META) What happened on the 10%+ drop on Friday was mechanical hedging and MM Pinning. You can see this with the price stuck at $113.5, despite any volatility. I'd expect short hedges to unwind Monday (given MMs bought puts and were short calls -> heavy short into expiration) and price to go back up. I ended up buying 6 figures worth of calls on the drop as there was no material changes. Strong Buy TSM - Holy crap, please have this in your portfolio. This is a money printer, and scaling your revenue by 38-40% every year WHILE increasing gross margins is just insane. It dipped as well after smashing earnings so it's one of the easiest longs in my life. AMKR - I don't have this in my portfolio yet but will be looking to add due to TSM's involvement in Arizona and potential to be a big partner in the US supply chain (as America tries to push TSM toward US fab + manufacturing). WLAC - Neocloud SPAC IPO, large upside. I talk about this a lot recently, but it's probably one of the best valued Neoclouds out there, and already has great profit margins (not a pivot from miners, where it's a bit more uncertain). They work with Fluidstack, and I'd expect a 500%+ re-rating on top of a Mag7 contract. AMZN - $213 is insane lol. I have no clue how this is down -3% YTD during a bull market. LTC - Affected by crypto liquidations and government shutdown delaying ETFs. Great time to buy and just wait for ETF to be approved. RDDT - Great dip to $190. I thought $200 would be a bottom but ended up going lower. The news about ChatGPT citing it less caused a large sell-off which I think was very immaterial. HIMS - 14%+ drop off CEO share sale. Owners sell shares all the time, it doesn't really affect the fundamentals of the company much, just short term sentiment. I'd expect it to rebound. IBIT - Bitcoin $108k great entry point, it's been swinging between $110k - $120k for awhile so anything under is usually great. ALAB - I said this last time but it sold off way too much from news of a new competitor. It's already competing vs AVGO in the market lol, NVDA-like margins, growing hundreds of percent Y/Y, Mag7 using them in data center buildout. CRDO - Similar thesis to ALAB, sold off alongside Astera but a bit less. SMCI - Should get re-rated for 55%+ or so revenue growth into next year. I doubted the projections earlier but with the data center growth, it's looking realistic. FLY - This was a medium lift payload play. People doubt fly's execution but NOC co-developing medium lift takes a lot of risk off the table (and possible re-rating it 500%+ when it competes vs falcon9) SNAP - Did the math on Snap monetization of memories in an earlier DD post and it's completely not priced in yet. It's doing $1.3B+ quarterly revenue on a $13B market cap lol, and the amount FCF they would get from increasing their revenue + lowering Google OPEX costs is insane. ETOR - Majority cash, growing at IBKR rates, suffering from tax harvesting LULU - Suffering from tax harvesting + competition from Alo, Vuori, etc. But seasonally should be good, and extremely low p/e now. Buy AMD - ChatGPT putting in AMD orders, ORCL building out AMD data centers. Likely going to get a re-rating in the next year as a potential $NVDA competitor. Still think Nvidia will dominate but with it's 4.5T marketcap, AMD has a lot to catch up on even if it takes a small percent share. HOOD - Looking at a lot better after the 10%+ correction. Could pull a PLTR RBRK - Did DD on this earlier, looks better on the drop as a cybersecurity company really low multiples in the space. Just needs to cut back on marketing, customers sticky. UNH - Healthcare is sht in America but not going anywhere. Think Warren and the others know this TGT 🐈 - Dividend next moth, big dividend stock. Around now is a good time to load up IMO IREN 🐈 - Huge GW, expect mag7 or similar deal. WYFI - Any neocloud is a buy (eg. see thesis on mag7 funneling revenue down toward these small 1B-5B companies) WULF - neocloud play CIFR - neocloud play SLNH - neocloud play BITF - neocloud play GLXY - neocloud derivative play FLNC - neocloud energy play MU - China derisked, memory had a huge market there, memory also likely going to get re-rating in tdata center buildout _ Sell ETH - Not a fan of Ethereum at $4k+ BMNR - If I don't like Ethereum at these levels, no point of holding treasury companies PL - Low revenue, space stock (extremely high valuation) BLSKY -Low revenue, space stock (extremely high valuation) RGTI - Quantum bubble OKLO - Nuclear bubble IONQ -Quantum bubble QBTS - Quantum bubble _ Quick macro heads up: -> Rate cut in 9 days ~97% odds. Frontrunning expected, go long. That's all.
-
博主复盘今年亏损持仓，总结税务收割、迷因股及期权交易教训。
周末反思 ☁️ 我想列出今年所有亏损的持仓+经验教训。 ___ 1. $ETOR - ($62 → $38.6, 定投 $48, -19.58%) 2. $VIRT - (~$38 → $33, -45%) 3. $SNAP - ($8.2 → $7.69, -6.2%) (新) 4. $SG - ($8.2 → $7.32, -10.73%) 5. $GRRR - ($20.5 → $16.99, -17.12%) 6. $FLY - ($30.2 → $26.5, -12.25%) (新) 7. $LTC - ($113 → $91.1, -19.3%) 8. $OPAD - ($4.61 → $2.94, -36.2%) 9. $CRM - (~$250 → ~$243, ~30%) 10. $AMZN - ($218 → $213) (新) 11. $RDDT - ($202 → $195) (新) 12. $WULF - ($14.5 → $14) (新) 我发了很多股票，很高兴只有约7只短期看错（抱歉如果漏了一两只），还有约5只没踩准绝对底部。 我仍然认为它们会在2026年大涨，输家不多！ 个人反思 1. $ETOR: 基本面很好的股票。即使下跌~20%，我相信它会在2026年恢复。1/3现金，以盈透证券(IBKR)的速度增长。它受税务收割(tax-harvesting)效应影响（10月/11月），因为对冲基金卖出并在12月回购。 经验教训：无，我会再做一次。持有至2026年。为税务收割小幅减仓。 2. $VIRT: 如果我买看涨期权，可能不应该公开发布，因为这会影响对冲资金流。临近两次降息（对股市利好）时减仓进行税务收割。对冲有效但损失不少。 经验教训：对冲操作保密。仍然认为方向正确，只是想税务收割。 3. $SNAP: 在$8.2左右发布，当时他们将记忆功能转化为收入。这是2026年的布局，不担心短期税务收割。 经验教训：无。持有至2026年。 4. $SG: 说实话，我只是喜欢沙拉🥗。今年早些时候是$35+。我认为它最终会恢复，只是等待游戏，因为现在约1倍市销率(p/s)？ 经验教训：无。中期持有。 5. $GRRR: 调查其$14亿合同后迁移至$WLAC，看起来可疑。被$3.8亿市值对比$14亿收入潜力所诱。 经验教训：相信对可疑公司的直觉。如果属实可能5倍，但我不信任。可能不该进入。（已减仓） 6. $FLY: 下跌~12%，仍然是2026年中期升力布局。最近发布，时机可能更好。 经验教训：无，也许税务收割时机更好。 7. $LTC: 下跌~20%。加密货币被清算+政府停摆延迟ETF批准所重创。 经验教训：不要杠杆炒币。政府停摆延迟ETF批准哈哈。 8. $OPAD: 之前卖出，但当前价格-36%。我冲进去的迷因股(meme stock)。 经验教训：不要碰迷因股。 9. $CRM: 因短期期权+错过财报日损失~30%。波段交易搞砸了。 经验教训：始终检查财报日，只是一次性错误，以前没犯过。 10. $AMZN: 它是亚马逊。有时下跌1%。 经验教训：无。刚发布不久。 11. $RDDT: 以为$200是底部，进一步下跌。平均成本降至$190。 经验教训：无。刚发布不久。 12. $WULF: 因为所有新云(neoclouds)都在涨而买入。$14.5买入，跌至$14。 经验教训：不要盲目追高(ape)。时机可能更好。 _ 税务收割总结 除了新持仓($AMZN, $RDDT, $WULF)和提到的中期布局如$FLY（持有至2026年）外，大部分输家为年底税务收割而减仓。 TLDR: - 年底不要重仓税务收割股（如$LULU, $ETOR, $SNAP）。如果择时，等12月。 - 停止交易迷因股如$OPAD。 - 当SEA相关公司看起来可疑时相信直觉($GRRR)。 - 期权资金流影响流动性较差的股票（如$VIRT）。3月看涨期权可能仍会大涨，其他人买入可能影响行情，只是想税务收割。 其他交易如$AMZN, $RDDT, $WULF非常新，也许时机更好，但没踩准绝对底部。 _ 总体而言，目前还好！我数不清有多少100%+的收益...输家数量仍然很少，无论是数量还是仓位规模。
英文原文
Weekend Reflections ☁️ I just wanted to post every position I'm down on this year + lessons learned. ___ 1. $ETOR - ($62 → $38.6, DCA $48, -19.58%) 2. $VIRT - (~$38 → $33, -45%) 3. $SNAP - ($8.2 → $7.69, -6.2%) (new) 4. $SG - ($8.2 → $7.32, -10.73%) 5. $GRRR - ($20.5 → $16.99, -17.12%) 6. $FLY - ($30.2 → $26.5, -12.25%) (new) 7. $LTC - ($113 → $91.1, -19.3%) 8. $OPAD - ($4.61 → $2.94, -36.2%) 9. $CRM - (~$250 → ~$243, ~30%) 10. $AMZN - ($218 → $213) (new) 11. $RDDT - ($202 → $195) (new) 12. $WULF - ($14.5 → $14) (new) I post a ton of stocks, and I'm happy that there are only about 7 I've gotten wrong short-term (sorry if i miss one or two by accident) and about 5 that I didn't time the absolute bottom correctly. I still think they'll print in 2026, not too many losers though! Individual Reflections 1. $ETOR: Great stock fundamentally. Even down ~20%, I believe this will recover in 2026. 1/3 cash, growing at IBKR rates. It suffers from tax-harvesting effects (October/November) as hedge funds sell and rebuy in December. Lesson learned: None, Id do the same again. Holding for 2026. Trimmed slightly for tax harvest. 2. $VIRT: If I buy calls, probably shouldn't post publicly since it affects hedging flow. Trimmed for tax harvesting since we're nearing two more rate cuts (bullish for markets). Hedging worked but lost quite a bit. Lesson learned: Keep hedges to myself. Still think I'm directionally right, just wanted to tax harvest. 3. $SNAP: Posted around $8.2 when they converted memories to revenue. This is a 2026 play, not worried about short-term tax harvesting. Lesson learned: None. Holding for 2026. 4. $SG: Honestly, I just like the salad 🥗. It was $35+ earlier this year. I think it'll recover eventually, just a waiting game cause it's like 1 p/s now? Lesson learned: None. Medium-term hold. 5. $GRRR: Migrated into $WLAC after investigating their $1.4B contract more, seemed sus. Got baited by the $380M market cap vs. $1.4B revenue potential. Lesson learned: Trust my gut on suspicious companies. Could 5x if legit, but I don't trust it. Probably shouldn’t have entered. (Trimmed) 6. $FLY: Down ~12%, still a 2026 medium-lift play. Posted recently, could’ve timed better. Lesson learned: None, maybe better tax-harvest timing. 7. $LTC: Down ~20%. Crypto got nuked by liquidations + government shutdown delays ETF approvals. Lesson learned: Don’t margin crypto. Government shutdowns delay ETF approvals lol. 8. $OPAD: Sold earlier, but current prices are -36%. Meme stock I aped into. Lesson learned: Don’t touch meme stocks. 9. $CRM: ~30% loss due to short-dated options + missing earnings date. Swing trade gone wrong. Lesson learned: Always check earnings dates, was just a on-off mistake, dont think I've ever made this mistake before. 10. $AMZN: It’s Amazon. Moves down 1% sometimes. Lesson learned: None. Just posted recently. 11. $RDDT: Thought $200 was bottom, dipped further. Averaged down to $190. Lesson learned: None. Just posted recently. 12. $WULF: Bought because all neoclouds were going up. Bought at $14.5, dipped to $14. Lesson learned: Don't ape into stocks. Could've timed better. _ Tax Harvesting Summary Trimmed most losers for EOY tax harvesting, except newer positions ($AMZN, $RDDT, $WULF) and medium-term plays like $FLY (holding through 2026) that I mentioned. TLDR: - Don't go hard into tax-harvested stocks end of year (e.g. $LULU, $ETOR, $SNAP). Wait for December if timing. - Stop trading meme-stocks like $OPAD. - Trust my gut when SEA-related companies seem suspicious ($GRRR). - Option flows affect less-liquid stocks (e.g. $VIRT). March calls could still print so other people buying might affect things, just wanted to tax harvest tho. Other trades like $AMZN, $RDDT, $WULF are very recent, maybe could've timed them better, but get the exact bottom. _ Overall, so far so good! I've lost count of how many 100%+ returns I've had... and the number of losers is still small, both for count and position size.
-
警示短期期权风险，看好AMZN隐含波动率回归潜力。
这就是为什么人们不应该做短期期权！（我看到很多WSB帖子讨论$TSM的周权，为了财报）。方向是对的，但像这样的随机事件会抹平看涨期权。不过，我认为$AMZN看起来最具吸引力，主要是因为隐含波动率（IV）回归的潜力，我们看看市场能否保持开放。
英文原文
This is why people shouldn’t do short term options! (I saw a lot of WSB posts about $TSM weeklies for earnings). It’s correct directionally but random events like this happen to wipe out calls. But yeah I think $AMZN looks the most tempting just cause of IV return potential, we’ll see what holds market open.
-
降息在即，建议利用回调做多NBIS、TSM等高确信度AI及科技股。
距离降息还有12天（概率94.1%）。做市商(MMs)喜欢利用这样的时机清洗期权链(option chain)。 就我个人关注的杠杆交易(leveraged trades)关键价位而言： $NBIS - $105-$110 $TSM - $170-$175 $BTC - ~<$100K $RKLB - $35-$40 $ALAB - $100-$120 _ 特别点名亚马逊，尽管市场反弹，其年初至今(YTD)仍下跌4.6%。 $AMZN - $200-$210（现在$211买入看涨期权(calls)极佳，让我想起$GOOGL在史诗级反弹前的$145。基本上现在看起来不错，但像Google一样，它跌破了人们预期的水平，所以甚至可能触及$200） 我对高确信度(high conviction)标的使用保证金+期权没问题，这只是我个人的风险承受能力，非财务建议(NFA)。 现在是做多高确信度标的的最佳时机。
英文原文
We are 12 days away from a rate cut (94.1% odds). MMs enjoy times like these to cleanse the option chain. In terms of levels I’m personally looking out for larger sized leveraged trades: $NBIS - $105-$110 $TSM - $170-$175 $BTC - ~<$100K $RKlB - $35-$40 $ALAB - $100-$120 _ Special shoutout to Amazon, which is down 4.6% YTD despite the rally. $AMZN - $200-$210 ($211 is extremely good now for calls, reminds me of $GOOGL at $145 before the epic rally. Basically around now seems good, but like Google it dropped below levels people expected so might even touch $200) I’m fine with margin + options on higher conviction stuff, this is just my personal risk tolerance, NFA. These times are the best to go long on high conviction.
-
区域银行恐慌是买入AI基建及新云厂商的良机，优选有真实营收支撑的标的。
宏观与回调思考 10月16日： 市场动态繁多，例如 $GLD 创历史新高，但这涉及美元的另一条复杂线索。今日股市主要下跌源于区域银行的冲击，因此我将聚焦于此。 以下是极佳的买入机会： 拥有确认的 Mag7 合同的新云厂商(Neoclouds) - $NBIS (MSFT) $CIFR (GOOGL), $WULF (GOOGL) 等。 AI 基础设施建设(AI Buildout) - 半导体如 $TSM（强烈买入，这类在爆发式财报后似乎总是回调以清洗看涨期权）。我们已有爆发式财报数据及增长的利润率+远期营收，$TSM 是我见过最简单的做多标的。TSM 的印钞能力也支持这一建设论点。 - 能源如 $FLNC 可能因数据中心建设+AI 使用带来的重估而继续上涨。 加密货币 - $BTC $107k, $LTC $90 都是极好的买入点，但我对 $CRCL 等会稍加谨慎。人们可能对 2023 年 Silvergate 倒闭及 USDC 脱钩仍有创伤后应激障碍(PTSD)。 某些股票急需修正： - 量子计算 - 机器人/无人机 - 关键矿产（仅特定标的可能优质） 这些公司零营收却因炒作呈抛物线增长。 区别在于，新云厂商+AI 基础设施建设是基于执行确认的营收。它们虽都是国家安全风险，但 AI 建设有全球最富有的超大规模云厂商背书。 “$AMZN, $META, $MSFT, $GOGL 等今年资本支出预计达 3200 亿。”- BI “$META, OpenAI 和 $ORCL 宣布未来几年将在数据中心投入超 1 万亿美元。”- BI 具体发生的情况： 市场今日因区域银行担忧下跌，但可能被夸大。最新恐慌源于 Zion/Western Alliance 的借款人欺诈。我猜测是个别银行问题，而非会导致信贷紧缩的系统性银行崩溃。 两周后降息概率接近 98%。这是全面必要的修正（以清除泡沫）+杠杆交易+期权流向。利用此次修正做多。 这仍是整个 AI 建设的最早期阶段，我预计 $NBIS 为例，1 年内 ARR 将增至 45 亿-70 亿+，市值从当前增至 600 亿+。 再次强调，新云厂商的炒作源于远期营收增长（基本面），而量子/部分无人机公司的炒作基于承诺。现在是扩大拥有疯狂远期营收增长（将导致市值大幅重估）公司多头的完美时机，而非仅如量子计算般的行业叙事/投机。
英文原文
Macro + Dip thoughts Oct 16th: There's a ton going on, eg. $GLD hitting ATHs, but that's a whole different rabbit-hole on the USD. Main stock drop today was due to a shock from regional banks, so I'll focus on that. It's a great buying opportunity on: Neoclouds with confirmed Mag7 contracts - $NBIS (MSFT) $CIFR (GOOGL), $WULF (GOOGL) etc. AI Buildout - Semis such as $TSM (Screaming buy it always seem to dips like this after blowout earnings to wipe off calls). We already have blowout earnings numbers and increasing margins + forward revenue, $TSM is just the easiest long I've ever seen. TSM printing money also supports this builtout thesis. - Energy such as $FLNC will likely continue going up because of re-rating bc of data center buildout + AI use. Crypto - $BTC $107k, $LTC $90 all amazing buys, I'd be a tad more cautious of $CRCL and others though. People probably have PTSD from 2023 silvergate, etc. going down and USDC depegging. It's a much needed correction for certain stocks: - Quantum - Robotics/Drones - Critical Minerals (only specific ones are likely good) That have 0 revenue and are increasing parabolically off hype. _ The difference is that Neoclouds + AI Buildout is confirmed revenue based on execution. They're all national security risks but AI Buildout has the wealthiest hyperscalers in the world are backstopping it. " $AMZN, $META, $MSFT, $GOGL, and others could spend an estimated 320B on capex this year". - BI " $META, OpenAI, and $ORCL, have announced plans to spend more than $1 Trillion on Data Centers in the next several years"- BI _ In terms of what happened in specific: Markets dipped today on regional banking fears, but it's likely overblown. The latest scare was due to Zion/Western Alliance mainly due to borrower fraud. My guess is that it's individual banks, not systemic banking collapses that would have credit tightening. We have a near-confirmed rate cut (close to 98%) in 2 weeks time. This is a correction needed across the board (to wipe away the froth) + leverage traders + option flow. use this correction to go long. This is still one the earliest parts of the whole AI buildout, I do expect $NBIS as an example to scale up to 4.5B-7B+ ARR in 1 year and $60B+ marketcap from here. Again the hype built into Neoclouds are because of forward revenue growth (fundamentals), while the hype into Quantum/some drone companies, etc. are based around promises. It's a perfect time to scale up longs from companies with INSANE FORWARD REVENUE GROWTH (which will cause large re-ratings to marketcaps) not just industry narrative/speculation like Quantum.
-
博主基于宏观利好更新美股AI/半导体/能源等板块的强烈买入、买入及持有名单。
大豆/植物油崩盘，个人思考与解释： 强烈买入 $ALAB $CRDO $NBIS $WLAC $LTC $TSM $BTC （+ 同上次税务收割股） $AMZN $SMCI _ 买入 $AMD $FLNC $SEI $BZAI $NKLR $IREN $WULF $CIFR $CRWV $BITF $WYFI $SLNH $BITF $RBRK $GLXY $GRAB $SEA $META $TGT $SNAP $MU $RKLB $FLY $UNH 持有 $MP $HOOD $EOSE $NVDA $GOOGL $DFLI $SOFI $VIRT $RR $AVGO $BE $ASTS （已达推文股票代码上限，其余同上次帖子，Quantum或Oklo仍建议卖出） _ 强烈买入 ALAB - 数据中心建设的重要组成部分，拥有类似英伟达(NVDA)的利润率，客户包括Mag7。已有来自博通(AVGO)的竞争者，真的不认为Arista会构成竞争威胁。 CRDO - 与ALAB同样的抛售，之前觉得两者都略有高估，但现在回到修正区间，适合补仓。 NBIS - $400目标价牛市情景。宏观顺风来自政府重新开放+10月底降息预期进入财报季，短期前景看好。利好众多（如Meta x CRWV，因此有更多Mag7客户的潜力），分部表现良好，例如Clickhouse，季度收入从$1亿扩展到$15亿+非常惊人。合同已锁定，只是公司执行的问题。 WLAC - 之前在$13时写过投资逻辑。即使在$14.5也很强，因为它可以轻松重估100%+。 LTC - 受杠杆交易者和政府停摆影响。预计停摆将持续一段时间，主要买入理由是ETF获批。但无论如何，低于$100都是很好的买入点，因为最终会获批（~95%概率）。 TSM - 天哪。如果是美国公司这将是$3万亿市值的公司，利润率惊人，对于其规模而言增长率惊人。每篇关于OpenAI X (**Sydney Sweeney合作)或AMD建设/NVDA建设的帖子中，TSM都是核心，即使以历史高点买入，也轻松成为$2万亿+公司（目前约$1.5万亿）。 BTC - $112K是好的入场点。黄金不断创历史新高，基本面没什么变化，只是最近大量清算。 （+ 同上次税务收割股） AMZN - 我真不知道为什么年初至今还在跌。我觉得亚马逊不需要太多解释，但仍在增长（例如AWS积压订单巨大，仍增长24%，当然不如ORCL、GCP等），但考虑到年底季节性和2月前的上涨，现在可能是抄底的最佳时机。AMZN今天触及$213-215是补仓的好机会，因为短期波段交易通常浮动在$218-$227，但长期我预计它会追赶其他Mag7。 SMCI - 被低估。市场关注短期表现，Charles引用的55%+同比前瞻收入增长无人相信+未实现的积压订单。但现在随着所有数据中心建设，这开始说得通了。因此应在接下来两次财报中重估。 _ 买入 AMD - OpenAI x AMD，Oracle使用AMD建设，这么多交易，如果它确实是$NVDA的强劲竞争对手，将重估为潜在$1万亿+公司。我不认为赢家通吃，可以看到$NVDA $4.5万亿+市值和$AMD $3500亿市值，所以我们可以看到大幅拉升（OpenAI通常是前沿模型领导者，如果Sam说他们可以使用AMD芯片+Elon说它对中小模型有益，可能意味着积极信号） FLNC - AI消耗后能源强劲重估，很好的买入。 SEI - AI消耗后能源强劲重估，很好的买入。 BZAI - 别人做过这家公司的深度研究，仅因板块和向边缘计算转移（例如机器人将很热）。由于低市值和类似公司的上涨可能表现良好。 NKLR - 像$OKLO这样的核能股已经起飞，这只是跟随梯队。 IREN - 无需介绍，巨大的GW算力，只是还没有宣布Mag7交易，但随时可能到来 -> 强劲重估。不是强烈买入的唯一原因是因为不完全确信矿工能像CRWV那样转型并保持高利润率（例如$ORCL打击文章），但我们会看到。 WULF - GOOGL积压订单，另外$3.6+亿左右的资金帮助很大。 CIFR - X上有很多关于未来产能和强劲重估的信息。一直喜欢这家公司，因为它是NBIS-lite。你可以买任何Neocloud，因为板块潜力巨大，Mag7将收入注入。 CRWV - 因为债务不像其他人那样喜欢，但由于植物油修正，$134（低于META交易宣布时）是更好的买入点 BITF - Neocloud类别相同 WYFI - Neocloud类别相同 BITF - Neocloud类别相同 GLXY - Neocloud类别相同，有助于其建设 RBRK - 做过深度研究，中期网络安全板块很好的买入，他们只需缩减营销，然后看起来有更多自由现金流，因为他们将大部分运营支出花在营销上。 GRAB - 基本面很好，-6.56%修正再次买入 SEA - 东南亚的AMZN，很多人使用。仅因客户群+变现潜力就买入。基本面$50亿+收入38% Y/Y增长也很好。 META - 我真的很不喜欢他们在AI上的昂贵资本支出，因为他们并没有像ChatGPT那样推出前沿模型，谁知道Zuck在做什么。但除此之外，一个月下跌7.3%，回到$700支撑位，可能在这里买入以追赶是个好主意。 TGT - 下个月股息是好的催化剂。 SNAP - Jenners回来了（有助于人气），他们将前记忆运营支出转为收入，这可能会导致明年巨大的重估。只是受税务收割影响，否则现在会是强烈买入。通常税务收割事件在12月完成。 MU - 现在中国恐惧稍微减弱，MU因为建设中的内存使用而成为更强的买入。 RKLB - 中子，金色穹顶合同，很多催化剂 FLY - 中等提升 UNH - 不受大豆影响的医疗保健股，但有修正。机构发布持仓后可能会上涨（例如Warren可能买了更多） 随机想法 基本上任何不是Oklo的成长/风险股都很好，因为我们有 -> 10月底降息 -> 政府重新开放（可能在10月底或11月初） 进入 -> 12月降息。 -> 中期选举（对股票看涨） 通常市场崩盘发生在紧缩而非宽松时。你的愚蠢量子泡沫可能会再持续3-12个月。如果你做空，可能等到明年2月。 无论如何，这是风险偏好的好时机，特别是搭乘Neoclouds -> 相关板块（例如能源）-> 相关公司（例如smci, tsm等）的趋势。 我半开玩笑地说大豆，因为它可能签署了升级紧张局势，但我可能会看到明年前的上涨。另外我可以写很多关于每一个的内容，但这很耗时，但我会不时发布关于随机股票如$RBRK的投资逻辑帖子。 太空/机器人/能源/量子/AI/半导体/关键垂直领域是目前最顶级的，不要对抗动量。我可以认为某些东西被高估了（例如一些关键材料，因为与Neoclouds相比仍然具有投机性，后者基于Mag7的执行有保证的收入），但我不会在降息时做空它。 只是个人想法，非财务建议
英文原文
The Great Soybean/Seed Oil Crash, personal thoughts and explanations: Strong Buy $ALAB $CRDO $NBIS $WLAC $LTC $TSM $BTC (+ same as tax harvest stocks last time) $AMZN $SMCI _ Buy $AMD $FLNC $SEI $BZAI $NKLR $IREN $WULF $CIFR $CRWV $BITF $WYFI $SLNH $BITF $RBRK $GLXY $GRAB $SEA $META $TGT $SNAP $MU $RKLB $FLY $UNH Hold $MP $HOOD $EOSE $NVDA $GOOGL $DFLI $SOFI $VIRT $RR $AVGO $BE $ASTS (Hit the ticker maximum but everything else from last post, still sell on Quantum or Oklo) _ Strong Buys ALAB - Huge part of datacenter buildout, NVDA like margins, Mag7 customers. Already had competitors from AVGO, really don't think Arista would be a competitive threat. CRDO - Same sell-off as ALAB, thought they were both kind of overvalued before, but now they're back in correction territory so good to stock up. NBIS - $400 PT bull case. We have macro tailwind from government re-opening + rate cut EOM october into earnings, so short term looks promising. Lot of things going for it (eg. meta x crwv, so there's potential for more mag7 clients), sum of parts doing well, eg. clickhouse, and scaling rev from $100m to $1.5B+ a quarter is insane. there's already contracts locked in its just a matter of company execution. WLAC - Wrote a thesis about this earlier at $13. Even at $14.5 strong because it can re-rate 100%+ easily. LTC - Affected by leverage traders and government shutdown. The shutdown is predicted to last awhile and the main reason to buy was the ETF getting approved. But a great buy sub <$100 anyway, because it will get approved in due time (~95% chance). TSM - Holy crap. This would be a $3T company if this were a US company, insane profit margins, insane growth rate for their size. And every post you see about OpenAi X (**sydney sweeney partnership) or AMD buildout/NVDA buildout. TSM is the center of it all and would easily be a $2T+ company (from here at ~$1.5T), even if buying at ATHs. BTC - $112K good entry point. Goldt keeps hitting ATH, nothing really changed fundamentally, just lot of liquidations recently (+ same as tax harvest stocks last time) AMZN - I really don't know how it's still down YTD. I don't think Amazon needs much explaining but still growing (eg. AWS backlog massive, still going like 24% but not as much as ORCL, GCP and others obviously), but with EOY seasonality and runup to Feb, now is probably the best chance to catch the bottom. AMZN hitting $213-215 today was a good chance to stock up since it usually floats between $218-$227 if you're short term swing trading but long term I'd expect it to catchup to other mag7. SMCI - Underrated. Markets were looking short term performance, and Charles was quoting like 55%+ Y/Y forward revenue growth which nobody believed + backlog that didnt get realized yet. But now with all the data center buildouts, now it's kinda making sense. So should re-rate in the next two earnings. _ Buy AMD - So many deals from OpenAI x AMD, oracle building out with AMD, this is going to re-rate to a potential $1T+ company if it's actually a strong competitive to $NVDA. I don't think it's winner takes all and you can see a $4.5T+ market cap size with NVDA and some $350B marketcap size with AMD, so we can see a large ramp up (OpenAI is usually the leader in frontier models and if Sam says they can use AMD chips + elon said its' good for small-medium weight models, prboably means something positive) FLNC - Strong re-rate on energy after AI consumption, great buy. SEI - Strong re-rate on energy after AI consumption, great buy. BZAI - Someone else did a DD on this company, just cause of sector and shift to edge compute (eg. Robotics goign to be hot). Because of low MC and runup of similar companies could turn out well. NKLR - Nuclear stocks like $OKLO have been taking off, this is just follow the lader. IREN - Needs no introduction, huge GW compute capacity just no announced mag7 deals yet but could come anytime -> strong re-rate. Only reason not a strong buy is because not fully convinced miners can pivot like CRWV and maintain great margins (eg. $ORCL hit piece) but we'll see. WULF - GOOGL backlog, another $3.6+ or so in funding helps a lot. CIFR - Lot of info on X about future capacity and strong re-rating. Always liked this company because it was NBIS-lite. You can probably buy any Neocloud and it will go up because the sector is incredibly high potential with Mag7 funneling revenue. CRWV - Didn't like this as much as others because of debt but because of the seed oil correction much better buy point at $134 (below when META deal was announced) BITF - Same in Neocloud category WYFI - Same in Neocloud category BITF - Same in Neocloud category GLXY - Same in Neocloud category, helps with their buildout RBRK - Did a DD on this, great buy for cybersecurity sector in mid term, they just need to scale back marketing and then it looks like they have a lot more FCF because they're spending most OPEX on marketing. GRAB - Great fundamentally, -6.56% correction good to buy again SEA - AMZN in SEA, tons of people use them. Just a buy just because of costumer base + monetization potenetial. Fundamentally growing $5B+ rev 38% Y/Y is also great. META - I really don't like all their expensive capex on AI since they're not really putting out fronteir models like ChatGPT with it, who knows what Zuck is doing. But that aside, down 7.3% over the month, going to $700 support, probably a good buy around here to play catchup. TGT - Dividend next month good catalyst. SNAP - The Jenners are coming back (helps with popularity), they're shifting former memory opex to revenue, and this will probably cause a HUGE rerating next year. Just suffers from tax harvesting otherwise would be a strong buy rn. Usually tax harvesting events are kinda done in December. MU - Now that China fears are kinda less intense, MU is a lot stronger buy just cause of memory use on buildout. RKLB - Neutron, golden dome contracts, lot of cataylsts FLY - Medium lift UNH - Healthcare stock not affected by soybeans but had a correction. Would likely go up one instituions post their ports (eg. warren likely bought more) Random thoughts Basically any growth/risk stock that's not named Oklo is great because we have -> Rate Cut end of month October -> Government re-opening sometime (likely around end of Oct or early Nov) Into -> Rate Cut December. -> Midterms (Bullish for stocks) Usually market crashes happen when there's tightening not easing. And your stupid quantum bubbles would likely continue for another 3-12 months afterward. If you're short, then probably wait till next Feb. Anyway, this is a great time for risk-on, and specially riding trends with neoclouds -> affiliated sectors (eg. energy) -> affiliated companies (eg. smci, tsm, etc). I half joke-about soybeans because it likely signed escalating tensions, but I'd probably see a run-up into next year. Also I could write up a lot about each one but it's pretty time consuming but I'll put on a thesis post about random ones eg. $RBRK, from time to time. Space/robotics/energy/quantum/ai/semi/critical top verticals right now, don't fight against momentum. I can think something is overvalued (eg. some critical materials bc. it's still spectulative compared to neoclouds that kinda have guaranteed rev based on execution from mag7) but I wouldn't short it into rate cuts. Just personal thoughts, NFA
-
分析$RBRK基本面与估值，认为其是网络安全板块好买点，但非最佳机会。
所以人们几乎每隔一条帖子就问起 $RBRK，我知道它在 X 上很火。 为了让大家别再问（我也在 $82 买入了一些 Rubrik），我最终去研究了一下它。 我的研究 TLDR（太长不看版）： - 网络安全公司（该行业估值倍数极高，参考 $CRWD 或 $NET） - 80% 的毛利率（很棒） - ARR（年度经常性收入）超 10 亿美元，同比增长 40-50%（很棒） （对比 NET，ARR 约 22 亿+，同比增长 22%，市值几乎是其 5 倍） - 运营支出 (OpEx)：60-75% 的收入用于营销。这是一个巨大的积极信号。 运营营销支出虽好，但短期看起来很难看，例如 $HOOD 给客户 3% 转账奖励时，这会伤害短期财报，因为具有误导性且没多少人做拆解；但长期来看，当削减支出时，客户粘性高，这对盈利能力帮助很大。 自由现金流 (FCF) 为正，但运营支出分解主要是营销，这是好事，不同于 Snapchat 的谷歌云运营支出。 - 客户基础多元化，像 Cloudflare（如高盛、百事、埃森哲等）。 缺点： - 资产负债表不是最好，约 11 亿美元债务用于资助收购。至少是为了收购。 - 不喜欢其远期收入数字放缓了 20-25%，相比之下 $NBIS 明年增长 700% 或更多。显然不公平比较，但这就是为什么我更看好 Neoclouds（新云基础设施）。 _ 看起来是一个不错的中期持有标的，计划很简单 -> 扩大客户群 -> 缩减营销 + 高粘性客户群 -> 赶上网络安全行业估值倍数并拥有更高的 FCF。 通常这类高毛利率（如 $HOOD 在 $18 时，增长 ~50% y/y）且实现盈利的成长型公司，重估 (re-rate) 幅度最大。 $RBRK 只需在未来减少营销支出，突然就会因为其粘性客户群拥有大量 FCF。 TLDR：网络安全板块的好买点，但其他地方有更好的机会。
英文原文
So people keep asking me about $RBRK almost every other post and I know it's really popular on X. I ended up looking into it so people stop asking (and added some Rubrik to my portfolio at $82). TLDR my own research: - Cybersecurity company (industry trades at extremely high multiples, look at $CRWD or $NET) - 80% gross margins (great) - $1B+ ARR, grew 40-50% Y/Y (great) (comparison to NET, ~2.2B+ ARR, growing 22% Y/Y, almost 5x the MC) - OpEx spend: 60-75% of revenue goes to marketing. This is a huge positive. OpEx marketing spend is great but it looks really bad short term eg. $HOOD when they give customers 3% to transfer, it hurts short term in earnings reports bc it's deceptive and not many people do the breakdown but long term when they cut back on spend, customers are sticky and this helps a lot with profitability. FCF was positive, but breakdown of opex expenses was mainly marketing, which is a good thing, unlike Snapchat google cloud opex. - Diversified client base like Cloudflare (eg. goldman, pepsi, accenture, etc). Downsides: - Balance sheet not the best, $1.1B debt or so to fund acquisitions. At least it's about acquisitions. - Don't like how their forward revenue numbers slowed down 20-25% compared to something like $NBIS growing like 700% or something more for next year. Obviously unfair comparison, but that's why I liked Neoclouds more. _ Looks like a good mid term hold with a pretty simple plan -> scale customer base -> scale back marketing + sticky base -> catch up to industry multiples in cybersecurity and hv higher FCF. Usually these types of growth companies with high gross margins (eg. $HOOD back at $18, growing ~50 y/y) that turn profitable, re-rate the hardest. $RBRK just gotta spend less on marketing down the road and suddenly they have a lot of fcf with their sticky customer base. TLDR: Great buy for cybersecurity sector, better opportunities elsewhere.
-
基于周五大跌，列出AI/加密/太空股买卖清单，看好Neocloud及算力基建。
基于周五收盘（SPY跌3.6%）的思考与解释 强烈买入 $IBIT $LTC $WLAC $NBIS $MP $TSM （针对明年） $ETOR $DKNG $SNAP 买入 $UPWK $CRDO $ALAB $AMZN $META $UNH $SG $TGT $BULL $FLY $CIFR $WULF $IREN $GLXY $SMCI $DELL $MRVL 持有 $RKLB $HOOD $RBRK $MU $HOOD $GRAB $MARA $RIOT $NVO $RR $ELOSE $FLNC $SEI $PLTR 卖出 $CRCL $ETH $BMNR $PL $BKSY 强烈卖出 $RGTI $OKLO $IONQ $QBTS $QUBT _ 解释： IBIT - 跌至10.4万美元，比特币需求来自机构，关税担忧被夸大，并引发了历史上最大的清算事件之一。在我看来，清算后是购买加密货币的最佳时机。Polymarket仍定价年底有55%几率达到13万美元，但无论如何，比特币长期来看总是好的买入标的。 LTC - 单日下跌24%。使用10倍杠杆的人可能在抢先交易ETF并在交易所被清算。这可能是我见过的在98美元以下买入的最佳时机，因为政府停摆结束后ETF可能会获批。 WLAC - 由于与Fluidstack（帮助WULF和CIFR获得GOOGL背书的公司）的合作，以6亿美元估值成为有史以来最好的Neocloud SPAC IPO 1000%机会之一，而这两家公司估值为40-70亿美元。作为背景，他们将在Q4 IPO，所以你可能需要等待约2个月。 NBIS - 我持有信心最强的买入标的，明年任何Neocloud中都有400美元的激进目标价。 MP - 稀土的国家安全风险（+其他从勘探到锂电池的稀有材料股票表现可能会很好）。 TSM - 所有AI基础设施的骨干 _ 强烈买入的税务收割股票池 （可能需要等到明年） ETOR - 在38美元处严重超卖，可能是正在进行的税务收割事件叠加亏损。 DKNG - 本质上是一只成长股。鉴于市值，营收数据惊人，但他们处于年度低点-15%。可能受不受宏观影响的税务收割影响。 SNAP - 如果你读过我的论点，他们130亿美元的市值对应13亿美元的季度营收……我相信在降低运营支出+通过货币化GCP存储增加营收后，他们明年会重新评级。这只是市场何时定价的问题，但这需要1年以上。后投资者没有耐心。再次，可能受至今表现导致的年底税务收割影响。 _ 买入 UPWK - 昨日下跌4.5%，基本面极佳，~8亿同比营收，70%~或80%~的毛利率在增长，2亿回购，22亿市值。可能受罗素指数抛售影响。 CRDO - 数据中心交易，Mag7使用它们。 ALAB - 数据中心交易，Mag7使用它们。 AMZN - 虽然受到100%中国关税的实质性影响，但他们不会失去任何东西，因为成本会转嫁给商户或客户。 META - 可能因中国关税导致广告收入减少（例如，像Temu这样的中国供应商可能不会购买广告位），但仅基于-6%的1个月低点+作为落后于亚马逊的Mag7成员，我更看好它。 UNH - 医疗保健不太受100%中国关税或稀土影响。 SG - 下跌可能由于表现不佳+税务收割，但它几乎是1倍市销率（P/S）哈哈。 TGT - 我会将其列为强烈买入，因为股息催化剂+刚刚触及5年低点，但尚未对中国关税对该股票的影响做足够的研究。 BULL - 以与HOOD相似的速率增长，其基本面对于市值来说并不完全出色，但你会基于零售客户群和未来货币化潜力进行投资。 FLY - 我之前做过DD，但中型有效载荷很可能在2027年与诺斯罗普合作成功，这需要大量的耐心以换取潜在的1000%回报。 CIFR - Neocloud交易，始终看涨。像Meta X CRWV一样，更多细节可能很快公布。 WULF - Neocloud交易看涨。 IREN - Neocloud交易看涨。他们的融资轮次价格如果我记得没错的话是其股价的70%+，这是一个看涨信号（例如NBIS和CIFR在融资轮次后都达到了这一点）。 GLXY - Neocloud交易的一部分。 MRVL - 市场尚未充分定价其像NVDA一样56%的同比增长。 SMCI - 数据中心/星门建设 DELL - 数据中心/星门建设 _ 卖出 CRCL - 我会不断重复这一点，除非Circle能展示其能在利息收入之外货币化USDC，否则COIN优于Circle。因为COIN实际上获得50%的收入分成，再加上其平台上100%的收入。 ETH - 我会不断重复这一点，但我个人不会在3000美元以上买入。所以即使它下跌16%（你可以尝试波段交易，反弹至4000美元+），也始终有继续级联下跌至3000美元以下的风险，我已经数不清历史上发生过多少次了。 BMNR - 基本面与ETH挂钩。 PL - 太空股票已经上涨很多，我认为像RGTI和其他股票一样，基本面并未反映其市值。 BKSY - 太空股票已经上涨很多，我认为像RGTI和其他股票一样，基本面并未反映其市值。 强烈卖出 RGTI - 市值与估值极度脱节 OKLO - 市值与估值极度脱节 IONQ - 市值与估值脱节 QBTS - 市值与估值脱节 QUBT - 市值与估值脱节 显然存在中美贸易战的宏观压力，但这都不是新信息（中国稀土出口管制已为人所知一段时间+美国已经对中国商品征收关税） 我们正走向定价为70%的两次更多降息，以及大幅修正以清洗高估的泡沫并清算杠杆（例如加密货币），这将帮助市场走向更高的高点。我将始终推荐股票，因为对于好公司，如果你等待足够长的时间，它们可能会跑赢（但如果出现随机的宏观风险，短期可能会跑输）。 中期来看，我对Neoclouds在未来一年AI快速建设期间最为看好 -> 机器人/太空可能是之后的下一个前沿。 （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）: Neocloud论点：超大规模资本支出漏斗 为什么我要投入150万美元+到Neoclouds，以及为什么这可能是200-300%+的回报。 🔹 类别 Mag7合同：$CRWV, $NBIS ✅, $WULF, $CIFR ✅ 带算力：$IREN ✅, $BITF 投机性：$WYFI, $GRRR ✅, $SLNH 转向HPC的矿工：$RIOT, $MARA, $CLSK, $HUT 论点： Mag7的AI算力紧张，这是由$NVDA设计导致的。 原本流向AWS、MSFT Azure、Google Cloud用于传统算力的数万亿美元资本支出，现在当他们无法处理来自Anthropic、OpenAI、Gemini等的新AI负载时，将流入Neoclouds。 这是一个十年一遇的机会，类似于使$NVDA成为4万亿美元公司的GPU军备竞赛，关于谁将为未来5-10年的AWS/Azure等提供基础设施。 NBIS（来自MSFT的170亿）、CIFR / WULF（来自GOOGL的30亿）、CRWV（由NVDA背书）都在以百分之几百的速度扩张（NBIS从1.5亿季度营收到可能的15亿+），毛利率为60-80%。 这种营收增长在历史上几乎闻所未闻。这主要是因为最富有的超大规模公司将资本支出注入小公司。 NVDA / TSM (2022->): 超大规模公司的GPU CRDO / ALAB (2024 ->): 超大规模公司获胜 -> 抛物线增长。 NBIS/CIFR/IREN等 (2025 - ) AWS/Azure等 -> 来自AI算力的抛物线增长 这就是如何获得百分之几百的回报，而不是在Paypal上进行价值投资。动量驾驭下一代公司。 所以看空论点通常涉及 - 执行风险（以前更投机，现在像NBIS这样的公司有40亿+来执行） 人们总是担心执行，但微软或谷歌不会在没有自己尽职调查的情况下签署如此大的5-10年合同。 - 高利率（主要看你CRWV），这就是为什么NBIS、CIFR和其他公司有潜在的惊人回报。 你有40亿+的资金用于$NBIS在138美元+每股（当它是107美元时）。以及$CIFR在16美元+每股的资金当它是11美元时。高于当前价格的融资是一个看涨信号。 - GPU折旧（有效的担忧，但它几乎像石油，即使是旧型号也保持价值并仍然提供权益）。 - 估值（我认为我们才刚刚开始。如果NBIS明年扩展到60亿营收75%毛利率），260亿市值非常小。 - NVDA可能推出自己的GPU即服务并直接竞争。目前这些Neoclouds是NVDA防止对Azure/AWS等集中风险的答案。 - 自定义超大规模芯片如TPU、Trainium。但可能还需要几年，因为他们仍在乞求NVDA算力并签署了5-10年合同。 无论如何，像Nebius这样的Neoclouds相对于远期营收/毛利率确实被低估了。 我们仍然非常早期。确保像特朗普选举时的Crypto/TSLA或OpenAI发布时的NVDA一样驾驭Neocloud浪潮。 当然这是高度投机的，我不会建议全仓YOLO，但为Google/MSFT等AI工作负载提供动力的这些5-200亿Neoclouds的风险回报是值得的。（交易时间范围：8个月-1年。） 这是2025-2026年最好的非对称AI基础设施交易。
英文原文
Based Friday Market Close (-3.6% SPY day), Thoughts and Explanations Strong Buy $IBIT $LTC $WLAC $NBIS $MP $TSM (For Next Year) $ETOR $DKNG $SNAP Buy $UPWK $CRDO $ALAB $AMZN $META $UNH $SG $TGT $BULL $FLY $CIFR $WULF $IREN $GLXY $SMCI $DELL $MRVL Hold $RKLB $HOOD $RBRK $MU $HOOD $GRAB $MARA $RIOT $NVO $RR $ELOSE $FLNC $SEI $PLTR Sell $CRCL $ETH $BMNR $PL $BKSY Strong Sell $RGTI $OKLO $IONQ $QBTS $QUBT _ Explanations: IBIT - Dumped to $104k, Bitcoin demand has been institutional, tariff fears overblown, and caused one of the biggest liquidation events in history. IMO post-liquidation is the best time to buy crypto. Polymarket still pricing in 55% chance to $130K EOY, but either way Bitcoin is always a good buy long term. LTC - Down 24% in one day. People on 10X margin were likely frontrunning ETF and got liquidated one exchanges. Probably the best time I've seen to buy sub $98 because ETF will likely get approved when government shutdown stops. WLAC - One of the best Neocloud SPAC IPO 1000% opportunities ever at $600m valuation because of their partnership with Fluidstack (the ones that helps WULF + CIFR get backstopped by GOOGL), and both of those are $4-7B. For context they're IPOing Q4, so you might need to wait ~2 months. NBIS - Strongest conviction buy I've had, $400 aggressive PT for next year out of any Neocloud. MP - National security risk for rare earths (+other rare materials stocks from exploration to Lithium batteries will likely perform well). TSM - Backbone of all AI Infrastructure _ Strong Buy Tax Harvesting Bucket of Stocks (Might need to wait for next year) ETOR - Way too oversold at $38, likely tax harvesting event going on compounding losses. DKNG - Fundamentally a growth stock. Revenue numbers are insane given market cap but they're -15% yearly low. Likely compounded by tax harvesting unaffected by Macro. SNAP - If you read my thesis, they're doing 1.3B quarterly revenue off a 13B marketcap... I'm convinced they will re-rate next year after lowering opex + increasing revenue from monetizing GCP storage. It's just a matter of when the market will price this in, but this is 1Y+ out. Post investors aren't patient. Again likely affected by EOY tax harvesting due to performance so far. _ Buys UPWK - Down 4.5% from yesterday, amazing fundamentals, ~800m y/y rev off 70%~ or 80~ gross margins thats growing, 200m buybacks, 2.2B market cap. Likely affected by Russell selloff. CRDO - Data Center Trade, Mag7 uses them. ALAB - Data Center Trade, Mag7 uses them. AMZN - It's materially affected by 100% Chinese tariffs but they won't lose anything, since it gets passed onto the merchant or customers. META - Probably less ad revenue by Chinese tariffs (eg. Chinese vendors like Temu, might not buy adspace), buy just based on -6% 1m low + Mag7 laggard with Amazon, I like it more. UNH - Healthcare not really impacted 100% Chinese Tariff or rare earths. SG - It's down likely due to bad performance + tax harvesting but it's almost 1 P/S lol. TGT - I'd put it Strong Buy due to dividend catalyst + 5Y low just now, but haven't dont enough research on the effect on Chinese tariffs on the stock yet. BULL - Growing at similar rates asHOOD, their fundamentals aren't exactly great for the market cap but you would invest based on retail customer base and the future potential for monetization. FLY - I did a DD on this earlier but Medium Lift Payload likely to succeed with Northrop in 2027, it takes a TON of patience for a potential 1000% moonshot. CIFR - Neocloud trade, always bullish. Like Meta X CRWV, more details likely to come soon. WULF - Neocloud trade bullish. IREN - Neocloud trade bullish. Their funding round was at some number 70%+ of their stock price if I remember correctly, which is a bullish tell (eg. NBIS and CIFR both hit that after their fundraising round). GLXY - Part of Neocloud trade. MRVL - It's not really priced in that it's growing 56% Y/Y like NVDA. SMCI - Data center/stargate buildout DELL -Data center/stargate buildout _ Sell CRCL - I will keep repeating this but COIN is better than Circle unless Circle can show they can monetize USDC outside from interest income. Because COIN literally gets 50% revenue sharing and on top of that 100% of the revenue on its platform. ETH - I will keep repeating this but I will not personally buy above $3000. So even if it dips 16% (you can take the chance of a swing trade, back up to $4k+), there's always the chance it keeps cascading down below $3k and I've lost track of how many times this has happened in history. BMNR - Tied with ETH fundamentally. PL - Space stocks have gone up a lot, I think with stuff like RGTI and others, fundamentals haven't mirrored it's marketcap. BKSY - Space stocks have gone up a lot, I think with stuff like RGTI and others, fundamentals haven't mirrored it's marketcap. Strong Sell RGTI - market cap extremely disconnected from valuation OKLO - market cap extremely disconnected from valuation IONQ - market cap disconnected from valuation QBTS - market cap disconnected from valuation -market cap disconnected from valuation There's obviously a macro overhead with China x USA trade wars going on, but none of this was new information (Chinese rare earth export controls have been known for awhile + US has already been tarrifing Chinese goods) We're heading into 2x more rate cuts priced in at 70% and large corrections to cleanse the overvalued froth and liquidate margins (eg. Crypto), will help the market go to higher highs. I will always recommend shares because with good companies, if you wait long enough they will likely outperform (but short term underperform if there's random macro risk). Mid term I'm the most bullish on Neoclouds for the next year during the rapid AI buildout -> Robotics/Space likely next frontier after that.
-
分享受关税影响较小的个股建仓及宏观恐慌下的逢低买入机会。
关于持仓变动： 新建 $DKNG 仓位 - $32.7（看起来估值不错，下跌7%） 买入 $UNH - $356（下跌3%） 买入 $AMZN - $217.8（下跌4%） 感觉这些股票受关税影响最小，尤其是 Draftkings 哈哈。 另外，由于宏观担忧因特朗普的一条推文突然浮现，$SPY 下跌1.65%提供了绝佳的逢低买入机会。 举几个例子： $AMZN - 下跌3.7% $AMD - 下跌8.27% $HIMS - 下跌8% $SNAP - 下跌5.1% $META - 下跌3.07% $WLAC, $NBIS 始终适合定投(DCA)。
英文原文
In terms of position changes: Started new $DKNG position - $32.7 (seems pretty good value, 7% drop) Bought $UNH - $356 (3% drop) Bought $AMZN - $217.8 (4% drop) Just felt these were the least unaffected by tariffs, especially Draftkings lol.
-
宏观恐慌提供买入机会，重点阐述$FLY与诺斯罗普合作开发中型火箭的高赔率逻辑。
此外，由于特朗普的一条推文导致宏观恐慌突然重现，$SPY 下跌 1.65% 是一个绝佳的逢低买入机会。 举几个例子： $AMZN - 下跌 3.7% $AMD - 下跌 8.27% $HIMS - 下跌 8% $SNAP - 下跌 5.1% $META - 下跌 3.07% $WLAC 和 $NBIS 始终适合定投(DCA)。 （引用内容）： 自从我在 Reddit 发帖后，很多人询问我的 $FLY 投资逻辑，如下： Firefly 是一家像 $RKLB 这样的火箭公司，正在开发可重复使用的中型运载火箭。 目前唯一存在的只有 SpaceX。 他们目标在 2026-2027 年实现，如果成功，公司市值将重估至 300-700 亿美元，即从当前（44 亿市值）上涨 1000-2000%。 更多细节： 300-700 亿市值的估算基于 $RKLB 的倍数，且仅假设中型运载火箭业务成功。（我忽略了其他业务板块和当前超过 10 亿美元的收入积压订单）。 在中型运载火箭业务上：预计年收入约 9-13 亿美元，在可行的发射节奏（8-12 次）下毛利率为 30-40%。猎鹰系列(Falcon-class)的定价支持这一预期。 所以现在唯一的问题是，FLY 能成功吗？ 在 Firefly 公司最近一次发射失败后，股价下跌 50%，市场已经计入了未来失败的特征。空头指出 Alpha 火箭的失败代表了执行风险。然而，他们完全错误地定价了这种风险…… 他们应该问的问题是，诺斯罗普(Northrop)和 Firefly 能否共同成功？ 这是美国最大的航空航天国防承包商之一诺斯罗普与一家市值仅 40 亿美元的小型发射公司之间的联合开发(CO-DEVELOPMENT)。 诺斯罗普此前确实有自己的中型运载火箭，但不得不终止。诺斯罗普之前的发射器 Antares 230 在俄罗斯 2022 年入侵乌克兰之前是有效的，随后供应链被切断。 然而，诺斯罗普面临生存问题：它无法再制造自己的第一级火箭。因此，为了节省时间，他们寻找一家拥有现成推进系统和结构的美国合作伙伴：Firefly。 -> 动机：Firefly 的 Miranda 发动机提供了即插即用的替代方案，可立即构建中型运载火箭。 现在你拥有一枚将在 2026-2027 年由 Firefly 货币化的中型运载火箭（因为诺斯罗普持有该公司股份），只需使其可重复使用，这是最难的部分。 但诺斯罗普不仅仅是“帮忙”。它在各个方面结构性地降低了风险。诺斯罗普的系统加上 Firefly 的新太空发动机，占据了稀缺且高需求的护城河（美国可重复使用中型运载），已通过 TacRS 和超过 10 亿美元的积压订单以及美国“金穹顶”防御系统催化剂得到验证。 因此，如果 Firefly 在中型运载火箭上成功并建立巨大护城河成为下一个 SpaceX，购买 Firefly 的风险回报是不错的。 （我的投资组合中也有 $RKLB，我只是认为由于中型运载火箭已与诺斯罗普存在，成功实现可重复使用中型运载的机会高于正常水平）。 显然，Rocketlab 实现可重复使用中型运载的成功概率更高，但这已反映在其 320 亿美元以上的市值中，而 Fly 的市值仅为 44 亿。 我认为 $FLY 在当前估值下是一个有趣的 10 倍潜力股，前提是它能在未来一两年内成功。而且我只是真的很喜欢火箭。 我还让 Grok 计算了概率，因为 XAI 由火箭人(Rocketman)本人拥有。 输出： 到 2027 年底，Rocket Lab 有 90%+ 的机会实现可重复使用中型运载飞行（可能在 2026 年中），得益于其先发优势、经过验证的执行力和内部控制。 Firefly/诺斯罗普达到 75-80% 的概率，得益于诺斯罗普的遗产，但受限于 Firefly 较新的运营记录。额外的一年缩小了差距——如果 Antares 330 按时飞行且 Miranda 发动机规模化，Firefly 可能在 2027 年底匹配或超过 Neutron 的节奏。 如果两者都成功（这是一个巨大的假设），RKLB 的市值可能重估至 600 亿美元，$FLY 可能重估至 400 亿美元，仅仅是因为太空 TAM 因国家安全风险和企业部门建设而迅速增加，可能只需再等一年。
英文原文
Also since macro fears randomly appeared today with a Trump tweet, it’s an amazing dip buy opportunity on $SPY 1.65% drop. Just some examples: $AMZN - down 3.7% $AMD - down 8.27% $HIMS - down 8% $SNAP - down 5.1% $META - down 3.07% $WLAC, $NBIS always good to DCA.
-
博主调整ETF持仓：大幅加仓AMD及新购多只AI/能源股，减仓部分股票以进行税务亏损收割。
哇，我在这个ETF上操作得极其激进，一切都涨得**很多。 不管怎样，以下是一些我会做的公司调整及解释： $AMD - +10%（重新评级） $FLY - +3%（新增） $WLAC - +1%（新增） $MU - +1%（新增） $FLNC - .5%（新增） $SEI - .5%（新增） $DFLI - - .25%（新增） _ 减仓（进行税务亏损收割 Tax harvesting） $ORCL $LULU $META $UPWK $ETOR $SNAP 其他小盘股 _ 以下是解释 $AMD - 刚刚获得了1000多亿美元的前瞻性收入，他们的季度收入几乎一夜之间翻倍，而且市场甚至还没有定价。极强的买入信号 $FLY - 44亿美元估值，从事类似$RKLB的小型至中型发射任务。让我想起RocketLab刚起步的时候，风险回报比很好。 $WLAC - Neocloud IPO，估值6亿美元（很低），EBITDA毛利率75%+，营收同比增长250%，很可能由Mag7（七大科技巨头）提供保底支持 $MU - 鉴于Stargate + OpenAI所需的基础设施量，内存需求旺盛 $FLNC + $SEI - 能源板块（高风险高回报） $DLFI - 电池板块（高风险高回报） 减仓 $ORCL - 在GPU建设方面遇到困难，这正好显示了超大规模云服务商和像$NBIS这样的Neocloud之间的护城河。直接购买Neocloud有更高的上行空间。 $LULU, $META, $UPWK, $ETOR, $SNAP. - 涨幅不大，我们要尽可能激进地持有赢家，并在完成税务亏损收割后回购，例如换成$AMD，以及获得更高回报的三次降息收益。 将在年底完成税务亏损收割后回购。
英文原文
Wow I cooked insanely hard with this ETF, everything is up a **** ton. Anyway, some company changes I'd make + explanations: $AMD - +10% (rerate) $FLY - +3% (new) $WLAC - +1% (new) $MU - +1% (new) $FLNC - .5% (new) $SEI - .5% (new) $DFLI - - .25% (new) _ Trim (Tax harvesting) $ORCL $LULU $META $UPWK $ETOR $SNAP Misc small caps _ Here's an explanation $AMD - Just got $100B+ in forward revenue lol, they almost doubled their quarterly revenue overnight and hasn't even been priced in yet. Extremely strong buy $FLY - $4.4B valuation doing small-medium lift launches like $RKLB. Reminds me of RocketLab when they first started, risk-reward is good. $WLAC - Neocloud IPO at $600m valuation (low) for something doing 75%+ EBITDA gross margin + 250% rev from last year + likely backstopped by Mag7 $MU - Memory in demand given the amount of infra required by Stargate + OpenAI $FLNC + $SEI - Energy Play (high risk high reward) $DLFI - Battery Play (high risk high reward) Trim $ORCL - Having trouble with GPU buildout, just goes to show the moat between hyperscalers and Neoclouds like $NBIS. Higher upside just buying Neoclouds. $LULU, $META, $UPWK, $ETOR, $SNAP. - Hasn't gone up much, we're playing as aggressive as possible with winners and will buy back once tax harvesting is done, eg. swap for $AMD, and higher return triple rate cut returns. Will buy back near EOY once tax harvesting is done.
-
博主分享了一份包含多只股票权重的投资组合分散化示例。
@DigestingX 我在这里写了一个投资组合分散化的例子：https://t.co/R7eMeeqR29 （引用内容：我收到了很多关于股票权重的提问。我会构建的投资组合是：30% $NBIS，6% AMZN，5.5% TSM，5% BTC，5% LULU，4% UNH，4% $RKLB，4% LTC，3% ORCL，3% TGT，3% GRAB，2% $IREN，2% META，2% HOOD，2% HIMS，2% AMD，2% NVO，1.5% CRDO，1% BITF，1% ASTS，1% SG，1% UPWK，1% MP，1% FOUR，1% ETOR，1% INTC，1% COIN，1% SMCI，1% MRVL，1% DAVE，0.5% DLO，0.5% MELI，0.5% SNAP，0.5% CRWV，0.2% ONDS，0.2% NFE，0.2% TSSI，0.2% BKKT，0.2% GRRR）
英文原文
@DigestingX I wrote an example of portfolio diversification here https://t.co/R7eMeeqR29
-
AWS增速虽不及竞品，但潜在合作预期支撑股价，跌幅不合理。
@AdityaInvests90 @pepemoonboy 是的，25% 依然不错，只是不如 Oracle 的百分之几百或 GCP 的 38% 那么亮眼。 只需与 OpenAI 和 AWS 达成一次握手，叙事逻辑就会改变。 但这仍然不足以解释 $AMZN 年初至今下跌 0.42% 的表现。
英文原文
@AdityaInvests90 @pepemoonboy Yeah 25% is still good, just not as good as Oracles 100s of percent or GCP 38%. Just takes a handshake with OpenAi and AWS for the narrative to change. Still doesn’t warrant $AMZN -.42% YTD.
-
看好亚马逊因降本增效及AWS订单，认为Meta资本支出效率低。
@AdityaInvests90 哎呀，我也上了 $AMZN 的炒作列车，我会和 @pepemoonboy 来点竞争。 我只是觉得 Meta 这个季度在 AI 上烧了太多资本支出(capex)，却没什么像样的成果。 亚马逊有太多利好，比如机器人降低运营支出(opex)以及 AWS 的积压订单(backlog)。
英文原文
@AdityaInvests90 Oof I’m on the $AMZN hype train, I’ll have a little competition with @pepemoonboy. I just felt like Meta blew too much capex on AI this quarter with nothing much to show. Amazon has too much going for it like robotics cutting opex and AWS backlog.
-
分析NBIS、AMZN、SNAP等个股机会，看好AI基建及降息预期。
10月6日周一收盘思考： - $NBIS 是极好的逢低买入机会。早盘上涨5.78%后回落2.38%。从 $IREN 到 $CIFR 的所有其他新云(Neoclouds)公司都守住了4%-14%+的涨幅。Nebius可能受期权资金流影响，预计很快会补涨，我维持 $225 的目标价(PT)。 - $AMZN, $META 是两只在未来2-3个月应跑赢大盘并补涨的七巨头(Mag7)成员。尤其是亚马逊。 - $SNAP, $RDDT 是两个不错的反弹标的。Snapchat尤其因为收入变现模式的改变。如果你有耐心持有股票一两年，我预期会有50%+的回报，只是取决于市场何时将其定价。并非每个人都有耐心，且将资金用于 $SNAP 而非新云(Neoclouds)的机会成本可能不划算。 关于Reddit，我一直认为ChatGPT引用是其暴跌29%的借口，所以我已买入。 - $SPRB 吸引了所有人的注意。我预计它将从7500万美元市值继续上涨至1.5亿-2亿美元，但这就像玩俄罗斯轮盘赌，通常在重大事件后2-3天就会发生稀释。 - 像 $RKLB 这样的股票，只需持有哈哈。即使它是我最高信用的5年期长线标的，也确实高估了，但此时它可能会像 $PLTR 一样表现。 - $AMD 与 OpenAI 的交易对半导体行业极度利好。我预计 $TSM, $ASML, 能源股和新云(Neoclouds)将从AI基础设施建设中受益。主要的负面因素是 $CRWV，因为其对 $NVDA 的依赖，以及显然的NVDA本身，但新云(Neoclouds)并未锁定单一玩家，且已锁定5-10年以上的合同。 这仅对 $NVDA 的护城河理念造成微小影响，但尚无实质性影响。 我个人认为AMD可能会像 $ORCL 一样在反弹后回调，然后当市场开始计入远期收入时，像 $AVGO 一样表现。 话说回来，我不知道OpenAI从哪弄来这么多钱，承诺给Oracle、AMD等这些十亿或百亿美元的交易，如果它们估值只有5000亿的话哈哈。 - 黄金每天创历史新高，仅仅表明 $BTC 始终是好的买入标的，即使在 $123k，如果它最终成为对抗通胀的对冲工具。其市值约为黄金的1/10。 - $LTC 仍然是很好的买入标的，因为ETF获批。由于政府停摆，人们只是忘记了它尚未发生，但最终应该会获批。 - $VIRT 在 $32.5 是极好的买入，我会在这个区间进行成本平均（抱歉如果你在 $36 买了看涨期权，我的仓位下跌了约35%）。但再次强调，这是对VIX的非对称对冲（VIX隐含波动率很高用于对冲，VIRT被低估~6.3倍远期市盈率，有回购和低隐含波动率），所以即使仓位下跌，你的其他股票应该会上涨以平衡。 - 仍在研究其他受益于基础设施建设的能源股、小盘股如 $EOSE、存储如 $MU 等，这些是粉丝推荐的。我尽量不在自己充分了解前谈论太多。 - 如果你使用杠杆或做多，现在是时候了，直到一月。三次降息，市场可能正在提前交易十月的降息。
英文原文
Monday October 6th Market Close Thoughts: - $NBIS extremely good dip buy. Down 2.38% after rising 5.78% in the morning. All other Neoclouds from $IREN to $CIFR held their 4%-14%+ gains. Nebius likely influenced by option flow, should play catchup soon and I stand by $225 PT. - $AMZN, $META two Mag7 that should outperform next 2-3 months and play catchup with the rest. Especially Amazon. - $SNAP, $RDDT two good recovery plays. Snapchat especially because of the revenue monetization changes. If you have the patience for shares for a year or two, I'd expect a 50%+ return, just whenever the market wants to price it in. Not everyone has patience and opportunity cost using the funds in $SNAP instead of Neoclouds might not be worth. Reddit I've maintained that the citations from ChatGPT is a BS reason for a 29% sell-off so I bought into it. - $SPRB caught everyone's attention. I do expect it to keep rising to a $150-$200m marketcap from $75m but it's like playing Russian Roulette, usually dilution happens 2-3 days after a major event. - Stuff like $RKLB, just need to hold lol. It's genuinely overvalued even if it's highest conviction 5Y long but at this point it might pull a $PLTR. - $AMD x OpenAI deal heavily bullish for semi industry. I expected $TSM, $ASML, energy stocks and Neoclouds to get a boost from AI infra buildout. Main negative ones were $CRWV, because of $NVDA dependencies and obviously NVDA, but Neoclouds aren't locked into one player, and they already have 5-10+ year contracts locked in. It just puts a tiny dent in the $NVDA moat idea but nothing material yet. I personally think AMD might pull an $ORCL where it dips past rally, and then ends up pulling an $AVGO when markets start pricing in forward revenue. Then again, I don't know where OpenAI is getting all this money to promise Oracle, AMD, etc. all these ten or hundred billion dollar deals if they're valued at 500B lol. - Gold rallying to ATH every day just signals that $BTC is always a good buy, even at $123k, if it ends up becoming a hedge against inflation. It's close to 1/10th the market-cap. - $LTC still a great buy because of ETF approval. There's the government shutdown so people just forgot it hasn't happened yet, but should get approved eventually. - $VIRT great buy at $32.5, I'd cost average around this range (sorry if you bought calls at $36, my positions are down 35% or so). But again it's an asymmetrical hedge to VIX (VIX IV very high for hedging, VIRT is undervalued ~6.3 forward p/e with buybacks an low IV), so even if positions are down, your other stocks should go up to balance it out. - Still looking into other beneficiaries of buildouts from energy stocks, small caps like $EOSE, memory like $MU, etc. that followers recommended. I try not to talk about something much until I'm informed myself. - If you're on leverage or going long, now is the time to do it until January. 3x rate cut, market probably frontrunning Oct rate cut now.
-
建议用投资组合保证金交易期权，低波标的可用杠杆，高波标的限1.5倍。
@1995bearjew 关于撰写看跌期权(put options)，我建议使用保证金(margin)交易低贝塔(low beta)标的如 $AMZN，但对于波动性更大的标的，杠杆率绝不超过1.5倍。 为此请使用投资组合保证金(portfolio margin)。
英文原文
@1995bearjew For writing put options yeah I recommend using margin for low beta stuff like $AMZN but never going above 1.5x for more volatile stuff. Use portfolio margin for this
-
分享激进期权卖出策略，强调底部择时与波动率管理，警示高风险。
我写了一个期权卖出波段交易策略，上周用100万美元本金在5天内实现了2万美元利润，年化收益率(Y/Y)超过183%。 我总是提前给出确切头寸，而不是事后诸葛亮。 $NBIS +$5.52K $HIMS +$1.427k $CIFR +$5.239K $RKLB +$3.8K $TGT +$1.3k $AMZN +$1.22K $IBIT +$947.86 $META +$869 对于100万美元的组合，这相当于每周约2%的被动复利，年化回报183%。 所有行权价均归零作废，权利金被全额收取。 即使股票下跌，如$HIMS（今日跌5.32%，本周跌6.5%），你仍有100%的利润。 这是波段交易者能够极好地抄底时采用的策略。 如果本周我再写看跌期权(Put)，我会动态调整行权价，并跳过一些之前写过的股票，比如因价格上涨而跳过的$RKLB。 再次强调，这适用于极其高级的交易者，这不是典型的“写看跌期权赚钱”策略，因为你需要主动进行底部择时（就像我今天在Reddit上发帖尝试寻找$200的局部底部）。不是在随机时间写期权。 如果你不知道如何很好地抄底，写看跌期权会放大亏损。 无论如何，这是许多高级交易者如何通过写期权复利净资产的一个例子。 （引用内容）： 激进地基于100万美元写期权将是： 5天内+$20.24K，每周2.024%，年化回报183.48%。 使用100万美元现金，盈透证券(IBKR)组合保证金示例： 85股 $NBIS $96 PUT (+$5.52K权利金) ($809K保证金占用) 55股 $HIMS $49.5 PUT (+$1.427k) ($270k) 250股 $CIFR $10 PUT (+$5.239K) - ($253k) 80股 RKLB $42 PUT (+3.8K) ($332K) 35股 TGT $85 PUT (+$1.3k) ($296k) 35股 AMZN $207.5 PUT (+$1.22K) ($725K) 50股 IBIT $59 PUT (+$947.86) - ($293k) 5股 META $712.5 PUT (+$869) ($335K) 100万美元现金，331万美元保证金。如果我把现金留到下周，我会这样做。 这是对每只股票进行底部择时，并基于基本面、宏观时机、事件和波动率预测本周不会触及的行权价。 此外，如果你在财报周这样做，回报率可能会高于183%哈哈，这可能是任何事件中利润最高的。但再次强调，超级危险，这是我个人会做的，也许只是当作乐趣阅读并在模拟账户中尝试，不要实盘。 另外，我也是凭记忆估算保证金维持要求（例如，鉴于META、TGT、AMZN等贝塔值较低，你可以使用更多杠杆）。 你必须先择时底部，然后写看跌期权。例如，AMZN不太可能跌破$210，BTC大约在$107k附近见底，RKLB短期内不会低于$43，NBIS有$100支撑，TGT见底，META不太可能一周跌5%+等。 所以希望这对活跃的波段交易者和高级交易者有点启发，如果你是只交易指数的新手，不要尝试这个。 _ 所以只是随机想法 1. 不要在你不愿意在该价位买入的股票上写看跌期权。不要受OKLO或QBTS高权利金的诱惑，因为突然下跌后它们可能几年都无法恢复。你需要已经知道什么是伟大的多头标的。 ^^^^重要*** ^^^^，请不要在随机 penny stocks 或投机性东西上写期权。只写你愿意买入并持有的东西，因为写看跌期权基本上意味着你最终会在那些价位买入它。 2. 在真正的好股票上高隐含波动率(IV)时重复操作，例如，如果HOOD IV达到90%或RKLB IV是90%或NBIS IV是90%，因为一旦行情停滞，IV最终会下降到60%或类似水平。 3. 不要对极高贝塔值的股票过度使用保证金，通常1.3-1.5倍保证金对于像NBIS或RKLB这样的股票更安全。高贝塔值使用2-3倍保证金是危险的。 这就是为什么~1.5倍保证金对于NBIS、CIFR、HIMS、RKLB等是合适的，以防除一只外其他全部下跌。 然后对于低贝塔值股票如META、IBIT、TGT、AMZN，你可以放心使用保证金，因为即使下跌那么多也不会对组合造成太大影响。 4. 学习隐含波动率并知道它为什么变化。如果你只是重复这样做，但在财报周卖出东西，而像TTD这样的股票暴跌40%，你就麻烦了。再次强调，这仅适用于非财报周。 5. 如果你真的想玩得安全，选择一个更低的行权价。比如$CIFR $7.5可能不会触及，但$10行权价有小概率触及。但我本来就想在$10买入+5.2K权利金，所以我选择了那个行权价。 6. 还需要知道任何重大宏观事件+风险水平。所以凭记忆，可能有关于美国政府停摆的负面消息，美国衰退几率增加，Polymarket定价三次降息从65%降至56%等。 另一方面，一些催化剂，比如AMZN在8-9日有Prime Day，所以它可能会表现更好并价格上涨，所以可能更好在之后写期权。所以如果风险太大，我可能会等到10月3日才开始卖看跌期权，而不是本周。 _ 一般来说粗略的经验法则 IV - <30%，不太值得，不像Blackrock、SPY那样波动不大，除非有PPI或其他事件 IV - 30-45%，通常是像MSFT、GOOGL、AMZN这样的科技股。在你用于高贝塔值的1.5倍之上额外使用保证金来做这些是很好的。 45%-65%，通常是30-60%年化增长类型的公司，如MRVL、Coin等。有时它们被严重错误定价，比如COIN/HOOD的IV有时不值得，鉴于它们的波动幅度。 65%-100%，通常是你更有趣的散户股票，如RKLB、NBIS，这是非常好的甜蜜点，因为它们可能会在下跌时反弹，如果你知道如何择时底部+增加几个百分点，很可能获得100%+的权利金收益。 100%+，哈哈危险区（如果你卖出几天到期的期权）。像OPEN、OKLO、财报这样的东西。可能有原因。如果你知道为什么，比如NBIS因MSFT交易上涨40%，我会在当时以200% IV卖出$85看跌期权，因为从根本上说我会在那个价位买入。 如果NBIS等东西在1-2天内IV升至100%+，这对期权卖方实际上是好事。如果是一周后，嗯，可能有什么东西会极度波动。 _ 如果我要分解个别股票 例如AMZN，作为波段交易者，我会在$210水平附近买入看涨期权，不太可能跌破$207.5（所以盈亏平衡点是$207.15，即-5.96%），所以你要确保选择一个它永远不会触及的水平。你也知道IV+贝塔值（波动方式）相对较低，所以你可以基于它不会触及来调整你的保证金。 例如高贝塔值，85股 NBIS $96 PUT (+$5.52K权利金) ($809K) IV是92%，这对期权卖方来说太棒了。你获得更多权利金，而且你不真的期望它跌破$100。IBKR不这样做，所以我推荐像Robinhood这样的东西来看盈亏平衡点，所以这只股票需要下跌11.3%才能盈亏平衡。 _ 再次强调，我只推荐给风险承受能力较高的活跃波段交易者，否则坚持持有超过一年的股票。 这也是为了激进复利的期权卖出， 而不是战略性地用于定投(DCA)头寸（例如，如果我想在$120买入AMZN，并且我认为我可以以更好的位置获得它，你可以在当前行权价写看跌期权，而不是低得多的价格）。 此外，除非有像特朗普关税这样绝对砸盘市场的黑天鹅事件，否则这几乎每周都有效。所以那周的看跌期权卖方被摧毁了，你可能只需要接受微小的损失并重新开始。 当我有闲置现金时我会这样做，因为我更多是突破交易者类型。 这只是我个人的交易风格和思维流程，再次强调非常危险，即使你有一些经验。这是超级高级的，我很多前量化交易员+买方同事都有现金，并在1000万美元以上做这种期权卖出风格，我只是展示它是如何做的以及背后的思维过程。 随意问随机问题，我会帮助解释。
英文原文
I wrote a Option Sell Swing Trade strategy that would realized $20k profit in 5 days with $1M last week for 183%+ Y/Y. I always give exact positions ahead of time, not retroactively. $NBIS +$5.52K $HIMS +$1.427k $CIFR +$5.239K $RKLB +3.8K $TGT +$1.3k $AMZN +$1.22K $IBIT +$947.86 $META +$869 With a $1M portfolio on it would be ~2% week passive compound for 183% Y/Y return. Every strike expired worthless and the premium would be collected. Even when stocks declined such as $HIMS (down 5.32% today), and 6.5% this week, you would still have 100% profit. This is a strategy people do when you're able to bottom time extremely well as a swing trader. If I were to write puts again for this week, I would dynamically change the strikes, and skip out on some stocks that I wrote earlier like $RKLB due to an increase in price. Again this works for extremely advanced traders, this is not your typical "write put, make money" type strategy since you're actively bottom timing (like how I posted with Reddit today, trying to time $200 local bottom). Not just writing options at random times. If you don't know how to bottom time well, writing puts will magnify loss. Regardless, this is an example of how a lot of advanced traders write options to compound net worth.
-
分享基于图表的波段交易策略，强调结合基本面与宏观分析的重要性。
我如何用图表进行波段交易，第二部分：短期（几周）+ 短中期（几个月）。 长期部分我会在另一篇帖子中讨论。 例1：$RKLB。仅凭直觉，在$40支撑位买入似乎是个好选择，$54卖出。通常下半部分（虚线）如$44也是不错的买入点，因为风险回报比良好，但你不会买到绝对底部。 你可以对像Rocketlab这样的股票这样做，因为长期来看，如果你持有足够久，即使它跌破$40，鉴于其基本面是强力买入（尽管目前略微高估），它很可能会恢复。 例2：$AMZN - 现在用正股买入是很好的选择。如果它跌破$210，你可以使用LEAPS（长期期权）。例如低于$200时，使用期限较短的看涨期权。 _ 了解基本面、宏观环境以及催化剂是否实质性影响也非常重要。如果未来收入大幅下降或行业利润率压缩，这些线条毫无意义。 很多时候它们因更非理性的因素下跌，例如GOOGL与苹果搜索之争，或者整体市场SPY下跌，但在这些情况下，如果没有实质性差异且公司持续增长，它们通常会再次上涨。再次强调，因人而异，在抄底时机上你需要分析图表以外的更多因素。这只是我所做工作的一部分。
英文原文
How I do swing trading with charts, Part 2: Short term (few week) + Short-Medium Term (few month). I'll do long term ones in another post. Ex 1: $RKLB. Just going off feels on this, seems like a great buy at $40 support, and $54 sell. Usually lower half (dotted line) like $44 is a good buy too cause risk-reward is good but u wont get the actual bottom. You can do it with stuff like Rocketlab since long term if you hold enough, even if it dips past $40 it will likely recover since it's a strong buy fundamentally albeit a tiny overvalued now. Example 2: $AMZN - Great buy with shares now. If it ever dips past $210, you can do leaps. Lower than $200 for example, shorter dated calls. _ Knowing fundamentals, macro, and whether catalysts are material or not is really also important. These lines mean nothing if forward revenue falls a lot or industry margins compresses. Lot of time they drop on more irrational things eg. GOOGL with Apple search, or maybe overall market SPY dipping but in those cases they usually rise up again if there's no difference + company keeps growing. Again different for everyone, you need to analyze more than the charts when timing bottoms. This is just part of what I do.
-
博主分享6只个股的短期交易策略，强调经验直觉胜过耐心口号。
最近 X 上流传着很多鼓舞人心的交易者心态，比如： “会有回报的。要有耐心。”全是废话。 交易者会考虑板块动量、催化剂、估值、回调、宏观、隐含波动率(IV)、期权流向等。 以下是我对各种股票短期交易的心态： 1. $NBIS - $111.91，尽管当天上涨了 1.53%，但由于 Meta 给予其 140 亿美元合同，CRWV 上涨了 12%。 这通常对所有新云厂商都是利好。它曾飙升至 $117（我可能仍会持有），但随后回落至 $111，可能是由于未平仓合约过多，但我们可能会继续看到上涨行情。因此我会利用这段时间进行定投(DCA)并买入看涨期权/股票，即使它已经上涨了 1.53%。 这不是“真正的下跌”，更像是上涨过程中的回调。 2. $HIMS - $56.4 下跌 4.67%，通常人们只是盲目抄底，但这实际上是由重大因素引起的，即特朗普推出了面向消费者的政府药品网站。随着股价升至 $60，空头比率降至 33%。 这次下跌可能会被用于空头回补。我在 $46 支撑位买入，但在反弹后不久卖出，因为我感觉它还会进一步下跌。但我个人更喜欢底部入场点，所以可能更接近 $50。 我记得 AMZN 推出竞争对手时，HIMS 暴跌 20% 然后再次上涨，我预计特朗普的计划中期也会如此，但近期它是一个逆风因素。 3. $RDDT - $228，下跌 5.45%，没有新闻。可能只是估值担忧。我们看到类似的增长股如 ALAB、CRED 有随机的 20% 回调。许多软件/社交股票如 SNAP 因非重大新闻下跌 8.1%。修正很健康，股票不会一直上涨，我个人更倾向于等待再次回到 $100+，而不是 $200+，但鉴于 RDDT 在随机日子有更大的 5-8% 回调，根据历史经验，6-7% 的下跌是日内的好买点，你可能会看到它恢复，但我们可能会看到许多增长股在 11/12 月的大涨前经历更大的修正，所以可能不是真正的底部。 我现在不看图表 RSI，只是基于过去一两年每天观察股票 + IV 的经验感觉来做。 4. AMZN - 没有重大宏观新闻，可能是 10 月 1 日的政府停摆可能会引起指数恐慌，但这没什么大不了的。它下跌了 1.35%，所以我会买入，因为这是摊低成本的好时机。 5. Klarna - $36，下跌 5.3%。有时你只是凭直觉行事。低于 IPO 价格，没有重大新闻。大多数 IPO 都下跌了，比如 Gemini 等。如果你想建立长期头寸，我会在这个水平买入。 6. TSM - $277，我有罪地进行了 $273-$279 之间的波段交易，所以我每次跌至 $273 就买入，在 $277-$279 卖出以获取 2% 的纯股票利润。到目前为止，我用股票做了大约 2 次。如果跌破 $273，我就定投，如果进一步下跌就切换到看涨期权。 没有绝对的对错方式，每个人都有自己的方法。（另外抱歉 CRM，糟糕的财报搞错了那个，如果进一步下跌我可能会摊平成本）。 但一般来说，这只是我浏览每只股票清单时的想法。再次强调，每个人想法不同，我只是想写下我的思考过程，如果对他有帮助的话。
英文原文
Lot of inspirational trader mindsets going around X lately like: "It will pay off. Be Patient". All BS. Traders consider sector momentum, catalysts, valuation, pullbacks, macro, IV, option flows, etc. Here's my mindset for short term trading for various stocks: 1. $NBIS - $111.91, even though it's up 1.53% on the day, CRWV is up 12% off Meta gives them a $14B contract. So usually it's bullish for all neoclouds. It spiked to $117 ( i probably would have still held) but pulled back to $111 likely from too much open interest, but we'll likely keep seeing a rally upward. So I'd use this time to DCA and buy calls/shares even if it's up 1.53% Not "truly a dip" but it's more of a dip during a rally. 2. $HIMS - $56.4 Down 4.67%, usually people just blindly buy the dip but this was actually caused from something material, which was Trump launching a direct to consumer GOV drug website. Short interest decreased back to 33% on the rise to $60. This dip will likely be used for short covering. I did buy $46 support but sold shortly on a bounce after I just felt like it would go down more. But I just personally prefer bottom entry points so that's probably closer to $50. I still remember AMZN launching a competitor, HIMS crashed 20% then rose again, I'd expect the same with Trump's program mid term but near term it's a headwind. 3. $RDDT - $228, down 5.45%, no news. Just probably valuation concerns. We saw similar growth stocks like ALAB, CRED, have random 20% pullbacks. Lot of software/social stocks like SNAP down 8.1% off non-material news. Correction is healthy, stocks don't just keep going up, I'd prefer to wait in the $100+ again, rather than $200+ (just personally), but it's actually a better buy than the rest, given RDDT has larger 5-8% pullbacks on random days, just from historical experiences so 6-7% drop is a good buy intra-day and you'd likely see it recover but we might see a lot of growth stocks have a larger correction into massive rally Nov/Dec so might not be an actual bottom. I don't really look at chart RSI nowadays, just do this based on feelings from experience looking at the stock + IV every day for the past year or two. 4. AMZN - No major macro news, prob government shutdown Oct 1st that might cause some panic for index but it's pretty immaterial. It dropped, 1.35% so I'd buy since it' a good time to cost average. 5. Klarna - $36, 5.3% drop. Sometimes you just go off gut feeling. Below IPO price, no major news. Most IPOs were down like Gemini, etc. If you wanted to build a long term position I'd buy at this level. 6. TSM - $277, I've been guilty of swing trading between $273-$279, so I just buy every drop to $273 and sell at $277-$279 for 2% profit purely with shares. So far I've done this ~2 times with shares. If it drops past $273, I'd just DCA and then if it drops further switch to calls. There's no True or False way to do this, everyone kind of has their own approach. (also sorry about CRM, bad earnings got that one wrong, I'll probably cost avg if ti declines further). But generally this is just what I'm thinking about when I go down the list of every single stock. Once again, everyone thinks differently, I just wanted to write down how I think if it's helpful to others.
-
博主发布周一收盘观点，列出多只美股的买卖评级及具体逻辑。
周一收盘思考： 极强买入 $NBIS $ETOR $LTC $VIRT 买入 $AMZN $SMCI $TGT $CRM $TSM $CRDO $SG $CIFR $LULU $SLNH $ORCL $MSTR $RIOT $MARA 持有 $IREN $HIMS $RKLB $PYPL $MRVL $IBIT $UPWK $GRAB $ALAB $ASTS $SOFI $NVDA $NVO 卖出 $HOOD $TSLA $RDDT $CRCL $PLTR $BMNR 强烈卖出 $OKLO $QBTS $IONQ _ 欢迎持不同意见，但这只是我的个人看法。 极强买入理由 - 买入约7万美元的Virtu看涨期权，隐含波动率28%，远期市盈率仅6.6倍，被低估。 - 在通往200美元的路上，每次回调都定投NBIS。 - ETOR在39美元时严重被低估。我不明白它怎么跌到那的。如果我没记错，市值33亿，现金储备超7亿，复利增长类似IBKR而非HOOD/BULL，但股价却直线跌破IPO价。 - LTC ETF批准在即（3-4天内），概率95%。除非被拒，现在是绝佳买点。 买入理由 - 今天买入5万美元以上Amazon看涨期权，回调至219美元以上后复苏前景看好。受益于10月至1月的年底季节性效应。10月8日Prime Day。可能再次回调，因此适合定投而非极强买入。 - SMCI仍预测55%的远期营收增长，季度营收超50亿，有点被低估。 - TGT下个月分红。有些Target事件，但我觉得不如Amazon Prime Day重要。 - CRM图表上看刚触底，基本面没太大变化。 - TSM在273美元更好，虽然总是好买点，但不像250美元以下那样极具吸引力。 - CRDO/ALAB，两者大幅回调。更像是修正而非崩盘，因此再次成为不错的买点。 - SG，不知道。我只是喜欢他们的沙拉，考虑到不久前交易价40美元，8美元的风险回报比不错。 - CIFR，GOOGL背书，现在看执行。我会逢低买入，但今天大涨了。 - LULU受益于10月至1月的年底季节性假日购物。 - SLNH，据说X上有关于波浪的讨论。市值很小约1亿美元，风险回报比似乎可以。 - ORCL，以140亿美元折扣估值持有TikTok美国大量股份，且来自OpenAI/MSFT的远期营收众多。就像AVGO一样，财报后可能回调，然后一两个月后迎来猛烈反弹。 - MSTR，比特币10月表现良好。被做空，所以溢价率可能在1.4x-1.5x，相比炒作时的2x。 - RIOT/MARA转向HPC，所以我比之前更喜欢它们。 持有股票没什么变化 - Hood，我个人做日内交易，所以如果我认为在单日上涨12.27%的130美元以上卖出是好的，别介意。 - TSLA，脱离基本面的信仰股。 - RDDT，我在100美元时持有大量，现在240美元或450亿市值不会买，所以可能会卖出/三角套利。 - CRCL，直接买Coinbase。 - PLTR，脱离基本面的信仰股，大部分利润只是利息收入。 - BMNR，如果你想买就买ETH，但ETH在4000美元以上也是强烈卖出。 强烈卖出 任何营收极少但市值100-200亿以上的股票我觉得很可笑。如果你从8美元持有OKLO到116美元，向你致敬。
英文原文
Monday Market Close Thoughts: Extremely Strong Buy $NBIS $ETOR $LTC $VIRT Buy $AMZN $SMCI $TGT $CRM $TSM $CRDO $SG $CIFR $LULU $SLNH $ORCL $MSTR $RIOT $MARA Hold $IREN $HIMS $RKLB $PYPL $MRVL $IBIT $UPWK $GRAB $ALAB $ASTS $SOFI $NVDA $NVO Sell $HOOD $TSLA $RDDT $CRCL $PLTR $BMNR Strong Sell $OKLO $QBTS $IONQ _ Feel free to disagree but these are just my thoughts Strong Buy Explanations - Bought ~$70K of Virtu calls, 28% IV and just 6.6 forward p/e is undervalued. - Always DCA NBIS on the road to $200 on every dip. -ETOR is just way too undervalued at $39 imo. I don't even know how it hit that. If I remember correctly $700M+ cash pile on a 3.3B market cap, compounding similar rate to IBKR instead of HOOD/BULL but just straight line down below IPO price. - LTC ETF approval in 3-4 days with 95% odds. Great buy now unless it gets rejected ofc. Buy Explanations - Bought $50k+ Amazon calls today, looks more promising for recovery on the dip to $219+. Benefits from end of year seasonality from Oct - > Jan. Prime Day Oct 8th. Could dip again which is why it's good to DCA and not an extremely strong buy. - SMCI still projecting 55% forward revenue growth and it's kinda undervalued doing 5B+ quarterly revenue lol - TGT dividend in another month. There's some Target event but don't really think it matters as much as Amazon prime day. - CRM just bottoming chart wise, fundamentals not really changed - TSM better at $273, it's always a good buy but not a screaming buy like sub $250 - CRDO/ALAB, both dipped a lot. More of a correction rather than crash, which is why it's a decent buy agian. - SG, idk. I just like their salad and think risk reward at $8 is good considering they were trading $40 not too long ago. - CIFR, GOOGL backstopped now just execution. I'd buy on dips but today was a big rally - LULU benefits from Oct -> Jan end of year seasonality with holiday shopping. - SLNH, apparently waves have been going around X. Pretty small $100m marketcap or so, risk reward seems okay. - ORCL, they're a large shareholder of TikTok US at a discounted 14B valuation and have tons of forward rev from OpenAI/MSFT. It's one of those things where it probably dips after earnings like AVGO then pulls off a face ripping rally a month or two later. - MSTR, Bitcoin does well in Oct. Been shorted so Nav prem is probably around 1.4x-1.5x compared to 2x like during hype waves -RIOT/MARA pivoted to HPC so I like them more than before For hold stocks nothing really changed - Hood, I personally day trade so don't be offended if I think it's a good sell $130+ on a 12.27% increase day. - TSLA, cult stock detached from fundamentals - RDDT, I had a lot back at $100 wouldn't buy at $240 or 45B marketcap now so would probably sell/tri. - CRCL, just buy Coinbase instead - PLTR, cult stock detached from fundamentals, large part of their profit is just interest income - BMNR, just buy ETH if you want but ETH is a strong sell at $4k+ Strong Sell Anything carrying barely any rev with 10-20B+ marketcap I think is amusing . Props to you if you held OKLO from $8 to $116 though.
-
警示MSTR杠杆风险，看好AMZN/META波段，回避RIVN/TSLA。
呃，我不会建议粉丝在内容提供商(CSP)上使用2倍杠杆ETF，尤其是$MSTR，它本身波动性已经足够大。如果遭遇波动率衰减导致的强制平仓，情况可能会变得很糟，尽管我确实认为MSTR正在触底。 我同意大多数观点，比如AMZN或META，通常我会在$BULL跌至13美元左右时进行波段交易。 由于基本面原因，我不会碰$RIVN或$TSLA。
英文原文
Uh I would not tell followers about CSPs on 2x leveraged ETFs especially on $MSTR, which is volatile enough. If they get assigned with volatility decay it might get messy, though I do agree MSTR is bottoming. I'd agree on most though like AMZN or META, usually I swing trade $BULL low $13. I wouldn't touch $RIVN or $TSLA cause of fundamentals.
-
建议通过虚值期权和组合保证金降低行权风险，追求每周1%的可持续收益。
关键在于，如果 $AMZN 触及 207.5 的溢价位，或 $NBIS 触及 96 的溢价位，你应该求着被行权（assigned），这些也是你可以买入看涨期权（calls）的价位。我的建议是选择更虚值（OTM）的合约并使用组合保证金（portfolio margin），而不是像 $AMZN 215 那样选择更接近行权价的合约，从而降低每周被行权的风险。除了 HIMX、CIFR 和可能的 META，我提到的这几个行权价被触及的可能性较小；如果真触及了，你应持有股票进行波段交易（swing trade）。但正如你指出的，每周约 1% 的收益对大多数人来说更可持续。
英文原文
The thing is if $AMZN ever hits $207.5 - premium or $NBIS ever hits $96 - premium you should beg to get assigned and those are levels you might buy calls on. Going more OTM and using portfolio margin is my suggestion instead of going closer to strike like $215 on $AMZN and having higher risk of assignment every week. My strikes are less likely aside from hims/CIFR/ possibly meta up there are the three I'd say, and if they do hit, you want to hold the stock for a swing trade. But yeah as you pointed out 1% a week or so is a lot more sustainable for most people
-
批评付费课程，指出高IV低贝塔期权卖出是复利关键。
@zenovaire 谢谢！是的，我只是看到随机的人说“卖出 $AMZN 看跌期权”并购买每年约 399 美元的课程，有点受够了。但通过高隐含波动率(IV)股票 + 低贝塔值(Beta)的保证金期权卖出策略，确实是许多人实现数百万美元复利增长的方式。这确实需要相当多的知识。
英文原文
@zenovaire Thanks! Yeah I just see random people saying "Sell $AMZN puts" and buy a course for like $399 a year and kinda got fed up. But playing higher IV stocks + low beta on margin option selling really is how a lot of people compound millions. It does take a quite of knowledge to do
-
ALAB因覆盖五大科技巨头且高增长，有望复制英伟达路径实现千倍涨幅。
@regard_capital 我在这里写了一点关于 $ALAB 的内容 https://t.co/T28ztuSkgY $ALAB 之所以起飞，是因为 Astera 是唯一一家系统性暴露于五大科技巨头(Mag7)中5家的小盘股。 1. ALAB 的客户群字面意义上就是 Mag7：$GOOGL、$MSFT、$NVDA、$META 和 $AMZN。 2. 他们的增长率惊人，同比增长约 150%+，毛利率为 76%（高于 NVDA，是半导体市场中最高的之一）。 3. 正如 $GOOGL 财报所示，AI 支出的资本支出(Capex)仍在增加。 NVDA 从 1500 亿市值起步 -> 4 万亿+，因为超大规模数据中心需要 GPU。我们正处于 ALAB 在 AI 供应链中产生这种效应的起点，而其市值仅为 200 亿。 如果他们执行得当且 AI 基础设施建设继续，我们可能会看到下一个 1000% 的涨幅。
英文原文
@regard_capital I wrote a tiny bit about $ALAB here https://t.co/T28ztuSkgY
-
分享激进卖出高IV看跌期权的策略、保证金管理及风险警示。
激进地卖出期权（Writing Options）$100万组合如下： 5天收益+$20.24K，周回报2.024%，年化回报183.48%。 以$100万现金、IBKR组合保证金为例： 85股 $NBIS $96看跌期权(+$5.52K权利金) ($809K保证金) 55股 $HIMS $49.5看跌期权 (+$1.427K) ($270k) 250股 $CIFR $10看跌期权 (+$5.239K) - ($253k) 80股 RKLB $42看跌期权 (+3.8K) ($332K) 35股 TGT $85看跌期权 (+$1.3k) ($296k) 35股 AMZN $207.5看跌期权 (+$1.22K), ($725K) 50股 IBIT $59看跌期权 (+$947.86) - ($293k) 5股 META $712.5看跌期权 (+$869) ($335K) $100万现金，331万保证金。如果下周持有现金，我会这么做。 这是对每只股票进行底部择时(Bottom Timing)，并基于基本面、宏观时机、事件和波动率预测本周不会触及的行权价。 如果在财报周做，回报可能高于183%，这通常是最有利可图的事件。但再次强调，超级危险，这只是我个人会做的，仅供娱乐阅读或在模拟账户尝试，不要实盘。 此外，我是凭记忆估算保证金维持要求（例如，由于META、TGT、AMZN等贝塔值低，你可以加更多杠杆）。 你必须择时底部然后卖出看跌期权。例如：AMZN不太可能跌破$210，BTC在$107k左右见底，RKLB短期不会低于$43，NBIS有$100支撑，TGT见底，META不太可能一周跌5%+等。 希望这对活跃的波段交易者(Swing Traders)和高级交易者有点启发，如果你是只交易指数的新手，不要尝试。 _ 一些随机想法： 1. 不要在你不愿意在该价位买入的股票上卖出看跌期权。不要因OKLO或QBTS的高权利金而诱惑，因为它们可能在突然下跌后多年无法恢复。你需要先知道什么是优质的多头标的。 ^^^^重要*** ^^^^，请不要在随机仙股或投机标的上卖出期权。只卖你愿意买入并持有的标的，因为卖出看跌期权意味着你最终会在该价位买入。 2. 在高隐含波动率(Implied Volatility, IV)的真正好股票上重复操作，例如如果HOOD IV达到90%或RKLB/NBIS IV是90%，因为一旦行情停滞，IV最终会降至60%左右。 3. 不要对极高贝塔股票过度使用保证金，通常1.3-1.5x保证金对NBIS或RKLB这类股票更安全。高贝塔股2-3x杠杆很危险。 这就是为什么~1.5x保证金对NBIS、CIFR、HIMS、RKLB等是合理的（约1664倍杠杆效应），以防除一只外其他都跌。 然后对META、IBIT、TGT、AMZN等低贝塔股票，你可以放心使用保证金，因为即使大跌也不会严重损害投资组合。 4. 学习隐含波动率并知道它为何变化。如果你只是重复操作但在财报周卖出，而像TTD这样股票暴跌40%，你就麻烦了。这仅适用于非财报周。 5. 如果想更安全，选择低一档的行权价。比如$CIFR $7.5可能不会触及，但$10行权价有小概率触及。但我本来就想在$10买入+5.2K权利金，所以我选了那个行权价。 6. 还需要了解重大宏观事件+风险水平。凭记忆，可能有关于美国政府停摆的负面消息，美国衰退几率增加，Polymarket定价三次降息概率从65%降至56%等。 另一方面，像AMZN在8-9日有Prime Day，价格可能上涨，所以可能更好在之后卖出期权。如果风险太大，我可能会等到10月3日再开始卖出看跌期权。 _ 一般粗略经验法则： IV <30%：不太值得，像Blackrock、SPY除非有PPI等事件否则波动不大。 IV 30-45%：通常是MSFT、GOOGL、AMZN等科技股。除了用于高贝塔股的1.5x外，用额外保证金做这些很好。 45%-65%：通常是MRVL、Coin等30-60%年增长率的公司。有时它们定价错误，如COIN/HOOD的IV有时不值，鉴于它们的波动幅度。 65%-100%：通常是RKLB、NBIS等更有趣的散户股，这是很好的甜蜜点，因为它们可能在回调时反弹，如果你知道如何择时底部+加点缓冲，很可能获得100%+的权利金收益。 100%+：哈哈危险区（如果卖出几天到期的期权）。像OPEN、OKLO、财报。可能有原因。如果你知道原因，比如NBIS因MSFT交易上涨40%，我会在200% IV时卖出$85看跌期权，因为从根本上我会在那个价位买入。 如果NBIS等股票在1-2天到期时IV升至100%+，这对期权卖方其实是好事。如果是一周到期，嗯，可能有什么极度波动的事。 _ 如果我要分解个别案例： 以AMZN为例，作为波段交易者，我会在$210附近买入看涨期权，不太可能跌破$207.5（所以盈亏平衡点是$207.15，即-5.96%），所以你要确保选择一个它永远不会触及的水平。你也知道IV+贝塔（波动方式）相对较低，所以可以根据它不会触及来调整保证金。 对于高贝塔股，例如85股 NBIS $96看跌期权，(+5.52K权利金) ($809K) IV是92%，这对期权卖方太棒了。你获得更多权利金，而且你不认为它会跌破$100。IBKR不提供此功能，所以我推荐Robinhood查看盈亏平衡点，这只股票需要跌11.3%才能盈亏平衡。 _ 再次强调，我只推荐给风险承受能力较高的活跃波段交易者，否则坚持持有超过一年的股票。 这也是为了激进复利期权卖出， 而不是战略性地用于定投(DCA)建仓（例如，如果我想在$120买入AMZN，并认为能以更好价格获得，你可以卖出当前行权价的看跌期权，而不是低得多的行权价）。 此外，这几乎每周都有效，**除非**发生像特朗普关税那样彻底击垮市场的黑天鹅事件。那周的看跌期权卖方会被摧毁，你可能需要接受微小损失并重新开始。 当我有闲置现金时我会这么做，因为我更偏向突破交易者(Breakout Trader)类型。 这只是我个人的交易风格和思维流程，再次强调非常危险，即使你有一些经验。这是超级高级的，我很多前量化交易员+买方同事都有现金，并在$1000万+上做这种期权卖出风格，我只是展示怎么做以及背后的思维过程。 欢迎随意提问，我会帮忙解释。
英文原文
Aggressively writing options off $1m would be: +$20.24K in 5 days, 2.024% a week, 183.48% y/y return. With a $1M cash, IBKR portfolio margin example: 85 $NBIS $96 PUT, (+$5.52K premium) ($809K) 55 $HIMS $49.5 PUT (+$1.427k) ($270k) 250 $CIFR $10 Put (+$5.239K) - ($253k) 80 RKLB $42 PUT (+3.8K) ($332K) 35 TGT $85, (+$1.3k) $296k) 35 AMZN $207.5, (+$1.22K), ($725K) 50 IBIT $59 PUT (+$947.86) - ($293k) 5 META $712.5 PUT (+$869) ($335K) $1M cash, 3.31M margin. This is just something I would do if I kept cash for the next week. This is bottom timing on every stock, and predicting strikes that would not hit for this week based on fundamentals, macro timing, events, and volatility. Also the return would probably be higher than 183% if you did it earnings week lol, which is probably the most profitable out of any event. BUT AGAIN SUPER DANGEROUS, this is something I'd personally do, maybe just read for fun and try on paper accounts but not live. Also, I'm also going off the top of my head with margin maintenance (eg. you can leverage more with META, TGT, AMZN, etc. given they're low beta). You have to time bottoms then write the puts. eg. AMZN not likely hitting sub $210, BTC sort of bottoming $107k around now, RKLB not going below $43 near term, NBIS $100 support, TGT bototming, META not likely to drop 5%+ a week, etc. So hopefully this is a bit informative to active SWING TRADERS and advanced traders, if you're a newcomer with just indexes, do not try this. _ So just random thoughts 1. Do not write puts on stocks you're not comfortable buying at those levels. Don't get tempted by high premiums on OKLO or QBTS cause those could just never recover for years on a sudden drop. You need to know what a great long is already. ^^^^IMPORTANT*** ^^^^, please do not write options on random penny stocks or speculative stuff. Only stuff you're fine buying and holding since writing puts kinda means you would buy it at those levels anyway. 2. REPEAT with high IV on REALLY good stocks, eg. if HOOD IV reaches 90% or RKLB IV is 90% or NBIS IV is 90%, cause eventually IV decreases to 60% or something once things stall out. 3. Do not over-margin extremely high beta stocks, usually 1.3-1.5x margin is safer for stuff like NBIS or RKLB. 2-3x on high beta is dangerous. That's why ~1.5x margin is fine 1,664 on NBIS, CIFR, HIMS, RKLB, etc in case all of them drop aside from one. Then low beta stocks like META, IBIT, TGT, AMZN you;'re fine margining since it wouldn't really dent the portfolio much if it drops that much. 4. LEARN implied volatility and know WHY it changes. If you just do this on repeat but sell stuff on earnings week and something like TTD crashes 40%, you're in trouble. Again this only applies non-earnings week. 5. If you really want to play safe, do one strike lower. Like $CIFR $7.5 will probably not hit, but $10 strike has a small chance. But I'd want to buy it at $10 anyway + the 5.2K prem, so I chose that strike. 6. Also need to know any major macro events + risk levels. So off the top of my mind, there's probably going to be negative news about US GOV shutdown, increased chances of US recession, polymarket pricing down triple rate cut 65% -> 56%, etc. On the other hand, some catalysts like stuff like AMZN has prime day on 8th-9th so it's likely to do better around and increase in price so probably better to write options later. So I might just wait until Oct 3rd, to start selling puts instead of this week if it's too risky. _ Generally rough rule of thumb IV - <30%, not exactly worth it, doesn't really move too much like blackrock, SPY unless there's like PPI or some other event IV - 30-45%, usually tech stocks like MSFT, GOOGL, AMZN, etc. It's good to do these with extra margin on top of your 1.5x that you use for higher beta. 45%-65%, usually more 30-60% y/y growth type companies like MRVL, Coin, etc. Sometimes they're really mispriced like COIN/HOOD IV is not worth sometimes given how much they move. 65%-100%, usually your more fun retail stocks like RKLB, NBIS, it's really good sweet spot since they'll likely bounce on dips and if you know how to time bottoms + add a few percent off, it's likely 100%+ premium gain. 100%+, lol danger zone (if you're selling few days out). stuff like OPEN, OKLO, earnings. Probably a reason for it. It's good if you know WHY like NBIS increasing 40% off MSFT deal, i'd sell $85 puts back then at like 200% IV because fundamentally I'd buy at those levels. IV goes to 100%+ on stuff like NBIS if there's one or two days out and that's actually a good thing for option sellers. If it's a week out then uhh something might be extremely volatile. _ If I had to breakdown individual ones AMZN for example, as a swing trader I'd would buy calls around $210 levels, unlikely to drop past $207.5 (so breakeven is $207.15 which is -5.96%), so you want to make sure you choose a level it never hits. You also know IV + beta (how things fluctuates) is relatively low so you can change your margin based on it never hitting. For high beta for example, 85 NBIS $96 PUT, (+$5.52K premium) ($809K) IV is 92% which is so nice for option selling. You get more premium, and you don't really erxpect it to dip below $100 either. IBKR doesn't do this so I'd recommend stuff like Robinhood to see breakeven, so on this stock would need to drop 11.3% for breakeven. _ Again I'd only recommend this if you're an active swing trader with higher risk tolerance, otherwise stick with stocks that you just hold over a year. This is also for aggressive compounding option selling, not using it strategically to DCA into positions, (eg. if I wanted to buy AMZN at $120, and I think I could get it at a better position, you can write PUTS at current strike instead of way lower). Also, this kinda always works every week **UNLESS** there's a black swan event like Trump tarrifs that absolutely tank the market. So PUT sellers that week kind of got wrecked, and you might need to just take a tiny loss and restart. I do this when I have spare cash on the side since I'm more of a breakout trader type. This is just my personal trading style and flow of thoughts, again VERY DANGEROUS, even if you have some experience. This is super advanced, a lot of former quant traders + buyside colleagues of mine have cash and do this option selling style off $10m+, I'm just kinda showing how it's done and what the thought process is behind it. Feel free to ask random questions and I'll help explain.
-
作者打破只买正股规则，在特定黑天鹅或底部信号下转向期权策略。
我在这里打破了自己关于只买正股而非看涨期权(call)的经验法则——32%的隐含波动率(IV)股票，但通常如果发生黑天鹅事件导致无关紧要的下跌，例如 $IREN 因摩根大通(JPM)糟糕的目标价下跌14%。或者 $CIFR 下跌17%后仅因 $GOOGL 的支撑而下跌5%，这时我会转向看涨期权。通常如果我认为触底了，或者认为像 $TSM $AMZN 这样的股票会在几个月内上涨，且我在定投(DCA)，且IV没有偏离32%太多且不至于糟糕，我会考虑使用期权。当然也有例外，比如 $IBIT 高IV时跳入周度备兑看涨期权(covered calls)，但那是另一种策略。不过，抄底时机极难把握。
英文原文
I broke my own rule of thumb here about shares only vs. call -> 32% IV stock but usually if something black swans off something immaterial eg. $IREN 14% drop off JPM bs price target. Or $CIFR 17% drop into 5% drop off $GOOGL backstopping it, then I switch to calls. Usually if I think it's a bottom or if I think a stock would go up like $TSM $AMZN over a few months and I'm DCA and IV isn't way off 32% and horrendous I'd consider it. There's exceptions obviously like $IBIT with higher IV leaps into weekly covered calls but that's another strategy. Bottom timing is extremely hard though
-
分享期权卖方策略经验：选对券商、小户用保证金、精通IV与财报风险。
补充一下关于学习期权卖方策略的讨论： 1. 券商选择极其重要，大资金账户切勿使用 $HOOD。他们声称免费，但可能是期权交易中最差的券商，因为他们将订单路由给做市商(MMs)，并通过糟糕的成交执行来赚取你的钱。除非你在Mag7或SPY上卖出低点差的现金担保看涨期权(Cash-Secured Calls, CCs)或现金担保看跌期权(Cash-Secured Puts, CSPs)，否则请使用 $IBKR 进行订单执行，即使你以最低价格卖出，他们也能提供最好的成交。看似损失几美元，但每份合约0.2-0.4美元的执行差异能带来数千美元的差别。 2. 如果是小账户，我建议用保证金(Margin)代替2倍杠杆ETF，因为后者存在波动率衰减(Volatility Decay)。（你可以谷歌搜索为什么不要长期持有2倍杠杆ETF）。使用保证金有清算风险，但仍优于杠杆ETF。帖子中应该已经提到过这一点。 3. 学习股票的隐含波动率(Implied Volatility, IV)。例如，$HOOD 通常每天波动不超过4-6%。$NBIS 通常每天波动不超过7%。$TSM 很少出现单日4%以上的波动。如果你在 $NBIS 上卖出4天到期的CCs，例如6*4，卖出约$130的24%+虚值(Out-of-The-Money, OTM)期权（当前股价$107）通常是明智且盈利的。如果你卖出约$110的CCs，你是在要求股票被行权(Call Away)，这通常是你试图卖出股票时的策略。了解催化剂(Catalysts)也很重要。如果有财报(Earnings)发布，原始IV不再适用，因为股价可能大幅波动。财报期间的CCs策略完全不同。 4. CSP策略很好。如果你想用小账户大量操作，使用IBKR的投资组合保证金(Portfolio Margin)，这样你可以在很多股票上卖出大量深度虚值的看跌期权(Puts)。最大风险是你不知道自己在做什么，或者出现像今年早些时候全球突然加征20%关税导致市场崩盘这样的黑天鹅事件。我是从大量经验的角度说的，但新手可能会亏很多钱。IV规则同样适用。如果 $AMZN 通常每周波动+-5%，你可以卖出7% OTM的看跌期权，很可能每周获利。财报策略又是另一回事。解释起来很花时间，但也极其盈利。很多人持有$1M+资金，通过“更安全”的CSP或定投(DCA)每周被动赚取$10k。每季度财报周能赚大钱，除非股票像 $TTD 那样暴跌40%。但总体而言，如果你想复利增长财富+现有资产，学习这些很重要。 期权卖方策略是极佳的“房地产”（指卖出CCs或极深度虚值看跌期权），前提是你懂行且操作于 $IIBIT、$HOOD 等正规资产。否则风险巨大。总体上是给交易者很好的建议。只是我基于个人经验的一些想法和细微差别。
英文原文
So just to add to the discussion on learning option selling: 1. BROKERAGE IS EXTREMELY IMPORTANT, NEVER USE $HOOD IF YOU HAVE A LARGE PORT. They claim it's free but it's probably one of the worst for option trading since they route your order to MMs and that's how they make money off of you by screwing your fills. Unless you're selling CC's CSP on Mag7/SPY with low spread, then use $IBKR for order execution since they give you the best fill even if you sell for the lowest price. It looks like you're losing a few dollars but .2 -.4 per contract on execution makes thousands of dollars of a difference. 2. If you're a smaller account, I'd recommend margin instead of 2x leveraged ETFs because of volatility decay. (you can use google to learn why you don't hold 2x leveraged ETFs for more than a short time). You have liquidation risks if you use margin but it's still better than leveraged ETFs. Think they pointed it out already in the post 3. Learn IV of stocks. Eg. Hood probably doesn't go up and down more than 4-6% a day normally. NBIS probably doesn't move more than 7% a day. TSM doesn't really move 4%+ a day. If you're selling CCs 4 days out like on NBIS, 6*4 for example, selling 24%+ OTM for like $130 if it's $107 is usually smart and profitable. If you sell CCs for like $110, you're asking for your stock to be called away and it's usually a strategy when you try and sell it. It's also important to learn catalysts. If there's an earnings coming up the original IV doesn't apply anymore since it could swing a lot higher or lower. Earnings CCs is completely different. 4. CSP is good. If you want to do a lot of this with a small account use IBKR portfolio margin, that way you can write a sell a ton of puts way OTM on a lot of stocks. Biggest risk is either you don't know what you're doing, or there's some black swan event like random 20% tarrifs on the world crashing the market earlier on in the year. Again I'm saying this from the perspective of tons of experience, but a newcomer might lose a lot of money. Same rule for IV applies. If AMZN is usually +-5% a week, you can sell PUTS 7% OTM and likely profit every week. Earnings again is a whole new strategy. it will probably take me a whole lot of time to explain this but it's extremely profitable too. A lot of people just sit on $1M+ and make like $10k a week passively doing "safer" CSPs or DCAing into positions. And earnings week every quarter is an insane amount of money. Unless a stock gets nuked like how TTD tanked 40%. But generally important to learn if you want to compound your wealth + existing assets. Option selling is great real estate (for selling CCs, or extremely OTM put selling) ONLY if you know what you're doing + do it on legitimate assets like IIBIT, HOOD. Otherwise there's tons of risks involved. Generally great advice overall for traders to learn. Just my thoughts and nuances to the lesson from personal experience.
-
分享今日多只股票及加密货币的短期交易与长期持仓逻辑。
除了ORCL、$NBIS、$IREN、$CIFR之外，今天这些是一些有趣的（短期）投资组合新增/交易和入场点，如果你好奇的话，附上我的一点点理由： GEMI ~$23.9 - 有趣的短期波段交易，随时波动+-~5%。不过从市值(Market Cap)来看被高估了。 AMZN ~$219.78 - 为4个月的时间框架进行定投(DCA)。 Figma ~$52.65 - 之前跌了4-5%，预计它会恢复。利润率在10亿美元营收基础上高达90%或类似的离谱数字。有趣的短期交易。 RKLB ~ $46.29 - 在$43 -> 53之间有趣的波段交易。 CRDO ~ $142.21 - 只是觉得它可能在10月反弹，定投(DCA)。 BTC ~ $109.5k - 不太在意+-5k的波动，这是为了长期持仓。 KLAR ~ $39.95 - 有趣的波段交易。 ETOR ~ $40.95 - 难以置信的低估lol。我可能会持有这东西$1M，而价格就保持不变。 当然，还在ORCL、CIFR、NBIS、IREN等股票上卖出了虚值(Out of the Money)的周权备兑看涨期权(Covered Calls)。
英文原文
Apart from ORCL, $NBIS, $IREN, $CIFR today these were some fun (short term) portfolio adds/trades and entries if you were curious + tiny bit of my reasoning: GEMI ~$23.9 - fun short term swing trade +- ~5% all the time. Overvalued though MC wise. AMZN ~$219.78 - DCA for 4 month timeframe. Figma ~$52.65 - was down like 4-5%, just expected it to recover. Profit margins are like 90% or something stupid off 1B revenue. Fun short term trade RKLB ~ $46.29 - fun swing trade between $43 -> 53. CRDO ~ $142.21 - Just thought it might rebound in Oct, DCA BTC ~ $109.5k - don't really care about +-5k too much, this is for long term positions. KLAR ~ $39.95 - fun swing trade ETOR ~ $40.95 - incredibly undervalued lol. I might end up with $1M of this thing and it will just stay the same price. And of course sold weekly covered calls on ORCL, CIFR, NBIS, IREN, etc... for way out of money strikes.
-
分享9/25市场观点、定投计划及详细投资组合权重。
9月25日的每日思考+市场下跌分析，如果你喜欢我的见解： 1. 三次降息（Triple rate cut）的概率从65%降至56%。这影响很大，因为人们正在提前交易（front-running）降息预期。无论如何，任何降息通常都会带来大量资金流入，因此对几个月后的市场总体看涨。鲍威尔关于市场估值过高的观点在某些股票上确实成立。Oklo、Quantum等严重高估，但我从不做空。甚至像RKLB这样我喜欢的股票也严重高估。但由于通胀，场外资金太多，除了股票、房地产和BTC外别无他物可持。三次降息意味着他们希望“抢椅子游戏”（musical chairs）再持续8个月。我会开始担心明年夏天。 2. 市场过去两天下跌，我会利用这个机会在$Z下跌15%后定投（DCA），或在$NBIS下跌5%后买入AMZN。我仍需研究$CIFR，因此在下跌17.5%后还不能完全有信心推荐（暂时），如果我喜欢它，我可能会隔夜买入。还有很多有趣的标的可以波段交易，比如RKLB。我仍在等待下个月TGT的六位数看涨期权收益，因为11月的股息。大家都在加载AMZN的看涨期权，但像GOOGL一样，它可能会跌至$210->$200的水平，人们放弃后开始一波愚蠢的反弹。这就是为什么我说如果跌更多就换成看涨期权。 3. 大量税务亏损收割（tax harvesting）正在进行。如果你有耐心等4个月，像ETOR、TGT、LULU这样被低估的公司可能会恢复，但显然除非你使用杠杆+期权，否则不会获得600%+的收益。对于耐心的投资者来说，现在是囤货的好时机。 4. 我一直主张在市场放弃时（例如以太坊$1600）在低位买入，而当Bitmine出现新叙事时，在$4k+卖出是好的时机。即使跌至$3.5k我也不会抄底。我对此有一个完整的论点，但留到另一天再说。这与NBIS不同，后者投机性较低，比如一个170亿美元的合同流入一家市值250亿美元的公司，这只是执行+等待的问题。仍在等待LTC，小市值，市场仍定价90%的ETF获批。几个月后LTC可能会有一个新的Bitmine，凭借小市值可以大幅反弹。 5. 我从未见过除了“ thirst traps”（诱惑照）之外，书签数超过点赞数的帖子，所以你们一定喜欢我的投资组合权重！你已经看到我在这里做日内交易：https://t.co/AjTB69Na65，但每次改变头寸都进行喊单有点烦人，所以我更喜欢只发布一般性见解+论点。 6. 市场定价的是远期收入，即使你今天看到NBIS、TSM等下跌，它们基本面很好，可能会继续上涨。
英文原文
So daily thoughts on Sept 25th + market drop if you like my insights: 1. 3x rate cut went from 65% to 56% from data today. This is a lot more material, since people are front-running rate cuts now. Either way, any rate cut usually lead to large inflows so it's generally bullish for markets months out. Powell's thoughts about market being overvalued holds kind of true for certain stocks. Oklo, Quantum, etc. way too overvalued but never short. Even stuff i love like RKLB, really overvalued. But there's too much money flowing on sidelines, nothing else to hold other than stocks, real estate, btc, because of all the inflation. Triple rate cut implies they want to keep musical chairs running for another 8 moths. I'd start to worry around Summer next year. 2. Market droplast two days, I'd use the opportunity to DCA into $Z after 15% drop, AMZN, or $NBIS after 5%. I still need to research $CIFR so can't really full conviction recommend it after a 17.5% drop (yet), if I like it I'd probably buy overnight. Lot of fun things to swing trade like RKLB on the side. I'm still waiting next month for 6 figures in TGT calls, cause of Nov dividend. Everyone's loading AMZN calls now, but like GOOGL it will might drop to levels like $210 -> $200, where people give up then start some stupid rally. That's kinda why I said shares into calls if it drops more. 3. Lot of tax harvesting taking place. If you have the patience to wait 4 months, lot of undervalued companies like ETOR, TGT, LULU, will likely recover but obviously won't net 600%+ gains unless you do leverage + options. Great time to stock up if you're a patient investor. 4. I've always maintained you should buy stuff at the lows when market gives up on it (eg. Ethereum $1600), and when there's a new narrative with Bitmine, it's a good time to sell at $4k+. I wouldn't buy the dip even if it drops to $3.5k. I have a whole thesis on this but I'll save this for another day. This is only different if it's less speculative like NBIS, like a literal $17B contract flowing into a 25B marketcap company and it's just a matter of execution + waiting. Still waiting for LTC, small marketcap, market still pricing in 90% etf approval. There's likely going to be a new BItmine for Litcoin in a few months, and with a small MC can rally quite a bit. 5. I've never seen a post get more bookmarks than likes other than thirst traps, so you all must like my portfolio weighting! You've already seen me day trade here: https://t.co/AjTB69Na65 but it's a little annoying for me to do call-outs every time I change positions so I'd prefer to just post general insights + thesis. 6. Market prices in forward revenue, even if you see stuff like NBIS, TSM and stuff dropping today, they're great fundamentally and will likely keep going up.
-
建议定投$AMZN，暂持现有仓位观望。
@__visionxry__ $AMZN 现在是定投(DCA)的好标的。 非常想买看涨期权(calls)，但除非出现单日跌幅达-4%的情况，否则我就继续持有现有仓位一周。
英文原文
@__visionxry__ $AMZN is a great DCA buy right now. Super tempted to buy calls but just sitting in current positions for a week unless there's some -4% day.
-
博主展示1000万高风险迷因组合，探讨极致收益与归零风险。
很多粉丝来自WSB（华尔街梗），没有风险过滤机制。如果我要用1000万美元为Reddit构建一个“迷因组合”以尝试最大化利润： 期权 $NBIS 3月20日 $130看涨期权 500万美元 AMZN 3月20日 $225看涨期权 200万美元 RUT 12月19日 $2400看涨期权 50万美元 2倍杠杆 $IREN 100万美元 $CIFR 100万美元 HIMS 50万美元 显然不推荐这样做，但如果我今天真的想尝试赚取1000%的收益，我会按当前价格这样操作。 如果你有1000万，好奇你会做什么风险博弈来换取豪宅还是温迪的垃圾桶。
英文原文
Lot of followers were from WSB with no risk-filter. If I had to build a meme-port for Reddit with $10m to try and maximize profit: Options $5M $NBIS Mar 20, $130 Call $2M AMZN Mar 20 $225 Call $500k RUT Dec 19 $2400 Call 2x Leverage $1M $IREN $1M $CIFR $500k HIMS Obviously not recommending this, but if I legitimately wanted to try and make 1000% I'd do this as of today's prices. If you had 10M, curious what risk-play you would do for a mansion vs. Wendys Dumpster.
-
博主调侃为抄底CIFR、AMZN和NBIS不惜代价。
@pepemoonboy 为了逢低买入 $CIFR、$AMZN 和 $NBIS，捐出一两个肾都是值得的。我是基于亲身经历说的。
英文原文
@pepemoonboy Donating a kidney or two is worth it to buy the dip on $CIFR, $AMZN, and $NBIS. Speaking from experience here.
-
博主分享其包含NBIS等39只标的的具体投资组合权重配置。
我收到了很多关于股票权重的提问。我会构建的投资组合如下： 30% $NBIS 6% AMZN 5.5% TSM 5% BTC 5% LULU 4% UNH 4% $RKLB 4% LTC 3% ORCL 3% TGT 3% GRAB 2% $IREN 2% META 2% HOOD 2% HIMS 2% AMD 2% NVO 1.5% CRDO 1% BITF 1% ASTS 1% SG 1% UPWK 1% MP 1% FOUR 1% ETOR 1% INTC 1% COIN 1% SMCI 1% MRVL 1% DAVE .5% DLO .5% MELI .5% SNAP .5% CRWV .2% ONDS .2% NFE .2% TSSI .2% BKKT .2% GRRR
英文原文
I've been getting a lot of stock weighting questions. A portfolio I'd build would be: 30% $NBIS 6% AMZN 5.5% TSM 5% BTC 5% LULU 4% UNH 4% $RKLB 4% LTC 3% ORCL 3% TGT 3% GRAB 2% $IREN 2% META 2% HOOD 2% HIMS 2% AMD 2% NVO 1.5% CRDO 1% BITF 1% ASTS 1% SG 1% UPWK 1% MP 1% FOUR 1% ETOR 1% INTC 1% COIN 1% SMCI 1% MRVL 1% DAVE .5% DLO .5% MELI .5% SNAP .5% CRWV .2% ONDS .2% NFE .2% TSSI .2% BKKT .2% GRRR
-
博主分享核心持仓与主动交易策略，反思若全仓操作收益或超6000%。
除非隐含波动率(IV)低于33%，否则我不做期权交易，比如 $GOOGL 或 $TSM，每笔都获利 $100K+。 通常我会持有一批高确信度的核心股票，如 $HOOD、$RKLB、$IBIT、$NBIS，并持有固定时间。 还有一篮子随机标的，如 $GRAB、$AMZN 等。 然后我会预留大量资金用于主动交易，这些我通常在 X 上发布。 有太多暴涨机会值得思考，比如减半前的 MSTR、Hood、SMCI、HIMS 等。我几乎抓住了每一次突破，但出于风险管理，只投入了组合中较小比例的资金。 如果每次我都全仓押注，哈哈，我的收益可能已经涨了 6000%+。
英文原文
I don't do options unless it's <33% IV things, such as $GOOGL or $TSM, both $100K+ profit each on. Usually I hold a core portfolio of high conviction stocks like $HOOD, $RKLB, $IBIT, $NBIS for a set amount of time. A basket of random stuff like $GRAB, $AMZN, etc. Then I set aside a good amount for active trading, which I usually post on X. Just too many run-ups to think about, like MSTR pre-halving, Hood, SMCI, HIMS, etc. I kind of caught every breakout but only put a smaller percentage of my portfolio into that for risk management. I'd probably be up 6000%+ if I full ported every time lol.
-
建议$220买入AMZN正股，并预测科技巨头市值目标。
如果你做正股，$AMZN 在 $220 是个很好的买入点。 我会等到 $211-216 再入场短期看涨期权。今天的美联储会议没有影响三次降息的概率。 关于科技巨头，我的预测是： $TSM -> $1.8T $META -> $2.3T $AMZN -> $3T https://t.co/5KwV6XZWxK
英文原文
$AMZN is an great buy here at $220 if you're doing shares. I'd wait for $211-216 before entering shorter term calls. Fed meeting today didn't affect odds of 3x rate cut. In terms of tech giants my guess is: $TSM -> $1.8T $META -> $2.3T $AMZN -> $3T https://t.co/5KwV6XZWxK
-
Polymarket维持三次降息预期，建议借机买入回调的AMZN。
@mattia030339 @AnaCintra24 这并没有改变 Polymarket（预测市场平台）的定价，仍然在计入 65% 的三次降息概率。我会利用这个机会买入像 $AMZN 这样下跌了 2.7% 的股票。
英文原文
@mattia030339 @AnaCintra24 It didn't change Polymarket, still pricing in 65% chance of triple rate cut. I'd use this to opportunity to buy stuff like $AMZN that dropped 2.7%
-
作者基于英伟达支持、微软大单及宏观降息，将NBIS仓位扩至百万美元，目标价225美元。
关于我将 $NBIS 仓位扩大至 100 万美元以上且目标价定为 225 美元的原因思考： 核心逻辑 ______ 当微软 (MSFT)、谷歌 (GOOGL) 和 Mag7 巨头成为你的客户时（就像 ALAB 或 CRED，我在它们股价低于 100 美元时重仓，或者在 $NVDA、AVGO 或 TSM 起飞之前），公司的未来增长可能会以百分之几百的速度爆发。 历史上几乎没有哪家公司像这样被 Mag7 巨头所依赖。 此外，当英伟达 (NVDA) 也有充分动机推高你的股价时…… NBIS 成为更强的买入标的，是我未来 1 年时间内信心最高的选择。 护城河 + 对英伟达至关重要的 GPU 租赁商 _______ 英伟达看到其他超大规模云厂商如 AMZN、GOOGL 正在构建自己的芯片并减少依赖。 对策？构建 GPU 租赁商（如 CRWV），通过持股和融资，使 Azure/Cloud/AWS 无法取代英伟达。 GPU 的获取越来越多地通过这些租赁商进行，正如我们在 OpenAI 身上看到的，他们不再拥有来自微软的算力，不得不转向谷歌。 谷歌的算力也不够，因此转向 WULF 等公司。微软转向 CRWV、NBIS 等公司，大型科技公司的算力需求向下游这些 GPU 提供商流动。 英伟达有充分理由推高 CoreWeave 和 NBIS 等 GPU 租赁商的估值，以对冲超大规模云厂商自研芯片的风险。 如果没有算力，数十亿美元将流向 NBIS，而英伟达在背后支撑他们。 宏观顺风 _______ Polymarket 已定价 3 次降息。 如果你看过我的另一篇帖子，十年才有一次的三次降息（对流入市场的流动性极其有利）+ 年底季节性因素，这对股票是积极的。 此外，NBIS 市值相对较小，且依赖融资/债务增长。当利息负担下降时，融资成本降低，扩张看起来更可行。 NBIS 的估值 heavily 基于未来收益，随着降息预期，来自微软和其他客户的数十亿美元收入预期将被上调。在昨天的美联储会议后，我们即将看到剧烈的重估。 公司对比 ________ 如果让我选择 NBIS -> CRWV -> WULF -> 其他，权衡潜力与投机性。 CRWV - 英伟达明确支持，市值已达 630 亿美元+。可能涨得更高，但考虑到 NBIS 拥有微软合同、类似的积压订单等，而市值仅为前者的一半多，这就是我选择 NBIS 的原因。 $IREN, BITF, RIOT, $GRRR 等 - 比特币矿工转型算力。我喜欢它们，但它们还没有像微软那样提供去风险化保障。 WULV - 我喜欢它们因为谷歌支持，但 NBIS 的可转债条款和相对于潜力的市值更好。 基本面 ______ 市值：247.8 亿美元（对比 CRWV 的 610 亿美元，后者今年仍下跌 30%），现在拥有类似的收入积压订单，轻松达到相同市值（每股 243 美元+）的潜力很大。 到 2026 年，计入微软交易后，NBIS 总收入可能达到 50-60 亿美元，毛利率强劲（≈70%）。 在 2027-2028 年，如果基础业务（非微软部分）也激进增长，总收入可能达到 80-100 亿美元+。所有这些都伴随着 ~60-75% 的毛利率。 这很疯狂，但主要是因为英伟达支持这些 GPU 公司。基于估值，NBIS 比 CRWV 更具吸引力，且比没有微软背书给予 170 亿美元的 IREN 等其他公司投机性更低。 杂项思考 ___ 他们已经筹集了 41 亿美元+，并为 170 亿美元的微软合同确保了资本支出。这是最大的去风险化事件，也是为什么我在 90 美元+ 投资 NBIS，而不是在 50 美元时投资更投机的标的。 可能还有其他来自超大规模云厂商的合同即将到来。我认为轻松路径是到 140 美元，然后可能因可转债转股下跌，再涨到 200 美元。 可能还有我没提到的细节，如几年后的利润率压缩、运营收入折旧风险等。 但这正是我投资决策的核心（宏观、基本面、行业增长、护城河+激励措施），与波段交易相比，这也是为什么 NBIS 是未来一年高信心标的的原因。
英文原文
Thoughts on why I’m scaling my $NBIS position to $1M+ with a $225 PT: Thesis ______ When MSFT, GOOGL, and Mag7 are your customers like ALAB or CRED (when I took large positions in both sub $100) or before $NVDA or AVGO or TSM took off, a company's forward growth would likely explode in the 100%'s of percent. There are barely any companies in history where Mag7 are dependent on them. On top of that, when NVDA has every incentive to pump your stock as well.... NBIS becomes a stronger buy and has my highest conviction for 1Y timeframe. Moat + GPU lenders Critical Importance to NVDA _______ NVDA sees other hyperscalers like AMZN, GOOGL building out their own chips and reducing reliance. Answer? Build out GPU lenders eg. CRWV, by taking a stake and financing them, so Azure/Cloud/AWS can't replace NVDA. Access to GPUs becomes increasingly through these lenders and as we see with OpenAI, they no longer have the compute from MSFT and have to go to GOOGL. GOOGL also doesn't have enough compute so they go to WULF and others. MSFT goes to CRWV, NBIS and others, and big tech compute flows downstream to these GPU providers. NVIDIA has every reason to inflate GPU lenders like CoreWeave and NBIS to hedge against hyperscaler custom chips. And without compute, billions of dollars will flow to NBIS with NVDA propping them up. Macro Tailwind _______ 3x rate cut priced into Polymarket. If you saw my other post, triple rate cuts only happen once a decade (extremely positive for liquidity flowing into markets) + combined with end of year seasonality, which are positive for equities. On top of that, NBIS is a relatively smaller market cap, and relies on financing/debt to grow. When their interest burden drops, financing gets cheaper, and expansion looks more viable. NBIS is heavily valued on future earnings, and with rate cut projections, far our earnings of billions in rev from MSFT and other clients get marked UP. We're about to see a sharp re-rating after Fed meeting yesterday. Company Comparisons ________ If I had to choose NBIS -> CRWV -> WULF ->others weighing on potential to speculation tradeoffs. CRWV - NVDA clearly backing them, already large $63B+ MC. Could go way higher but growth potential with MSFT contract with NBIS, similar backlog, etc. at more than half the marketcap is why I chose them. $IREN, BITF, RIOT, $GRRR etc. - BTC miners pivoting to compute. I like them but they don't have the same de-risking as MSFT yet. WULF - I do like them since Google is backing them but NBIS conversion terms and MC relative to potential are better. Fundamentals ______ MC: 24.78B (When you compare to CRWV at 61B, which was still down 30% this year), now with similar revenue backlog, it has an easy potential to go to the same marketcap ($243+ a share). By 2026, with Microsoft deal accounted for, NBIS could be doing $5-6B total revenue with strong gross margins (≈70%). In 2027-2028, total revenue could reach $8-10B+, especially if base business (non-Microsoft) also grows aggressively. All with ~60-75% gross margins. This is insane but mainly due to NVDA supporting these GPU companies. Based on valuation, NBIS is more attractive than CRWV and less speculative than IREN or others without MSFT backing giving them 17B. Misc Thoughts ___ They already raised 4.1B+ and secured capex for the $17B MSFT contract. This was the biggest de-risking event and why I'd invest in NBIS at $90+ compared to something more speculative back at $50. There's likely other contracts from hyperscalers coming their way. I'd say easy path to $140, then maybe drop from convertible note, then up to $200. There's probably more specifics that I didn't get to like margin compression in a few years operating income depreciation risk etc. But this was the core of my investment thesis (macro, fundamentals, industry growth, moat + incentives) when I make investment decisions vs. swing trading and why NBIS is high conviction over the next year.
-
分享9月精选15只高潜力股票及理想入场点。
我9月份精选的15只由催化剂驱动的股票，按截至2025财年的潜力排序，附解释及理想入场点： 1. $NBIS - 微软(MSFT)合同超170亿美元，稀释基本完成 < $100（现价）。相比 $IREN、$WULF 或 $BTBT 我更看好这只。 2. $HIMS - 42%的做空比例。来自欧盟的新增客户超120万。若有意外消息可能涨至$100 < $45 3. $LTC - ETF即将推出 + LTC储备 < $120（现价） 4. $RKLB - 中子号火箭将于2026年初发射 < $42 5. $TSM (看涨期权) - 市值持续增长至1.8万亿+，资本支出惊人。< $265（现价） 6. $ETOR - 随着 $HOOD、$IBKR 等金融科技股起飞，该股被低估。在当前价位下，我首选它而非 $DLO 或 $DAVE < $48（现价） 7. $LULU - 受假日消费提振，短期反弹至年底，超卖状态 < $165（现价） 7. $MRVL - 健康的45-55%同比增长，财报后遭不公平惩罚 < $70（现价） 8. $SG - 值得投资以博取反弹。去年股价$45。 < $9（现价） 9. $TSSI - 服务器机架（如 $SMCI、$DELL）正在追赶AI热潮的其余部分。 < $13.5 10. $ASTS - 像 $OPEN 一样的“邪教”级股票 < $38.5 11. $CRDO - 追赶 $ALAB 至380亿美元市值 < $125 12. $RUM - Charlie Kirk及保守派人士有望推动平台参与度至年底。 < $7.25 13. $AMZN (看涨期权) - AWS... 跟随 $ORCL 及其他数据中心财报表现 < $215 14. $SMCI - 2026年远期营收330亿美元。市值240亿美元哈哈。只要一份好财报就能起飞。$45（现价） 15. $IBIT (1年期LEAPS) - 鉴于美元超发，纯比特币敞口。不要 $MSTR 或 $BMNR，只要纯资产ETF < $112.5k 我还有什么遗漏的吗？
英文原文
My top 15 catalyst driven stocks from September sorted by potential to FY 2025, explanation, and ideal entry points: 1. $NBIS - 17B+ MSFT contract, dilution mostly finished < $100 (now). I like this over $IREN, $WULF, or $BTBT 2. $HIMS - 42% Short interest. 1.2M+ customers coming in from EU. Could up to $100 on surprise news < $45 3. $LTC - ETF launch soon + LTC reserves < $120 (now) 4. $RKLB - Neutron Launch early 2026 < $42 5. $TSM (calls)- Keep growing to 1.8T+ MC, capex spend insane.< $265 (now) 6. $ETOR - Undervalued with fintechl ike $HOOD $IBKR taking off. I'd prefer this over $DLO or $DAVE at current prices < $48 (now) 7. $LULU - Short term bounce to year end from holiday consumer spending, oversold < $165 (now) 7. $MRVL - Healthy 45-55% Y/Y, unfairly punished after earnings < $70 (now) 8. $SG - Worth the investment for bounceback. $45 last year. < $9 (now) 9. $TSSI - Server racks like $SMCI, $DELL catching up to rest of AI boom. < $13.5 10. $ASTS - Cult like stock like $OPEN < $38.5 11. $CRDO - Catchup to $ALAB at $38B MC < $125 12. $RUM - Charlie Kirk + conservatives likely to drive engagement to platform EOY. < $7.25 13. $AMZN (calls)- AWS... Follow $ORCL and other datacenter earnings < $215 14. $SMCI - 33B forward revenue 2026. 24B market cap lol. One good earnings and it pops off. $45 (now) 15. $IBIT (1Y leaps) - Pure BTC given USD printing. No $MSTR or $BMNR, just pure asset ETF < $112.5k Anything else I'm missing?
-
建议买入垄断且低估的台积电，因其他AI股均依赖其代工。
如果你在纠结 $ORCL 对比 $AMZN | $GOOGL，$AMD | $AVGO 对比 $NVDA，$NBIS 对比 $IREN 对比 $WYFI 对比 $CRWV 这很简单。直接买入 $TSM，上述所有公司都依赖于它。 估值偏低且拥有垄断地位，基于未来增长 + 边缘推理(Edge Inference)的资本支出，市值有望从 1.3 万亿美元 -> 1.8 万亿美元。
英文原文
If you're debating $ORCL vs. $AMZN | $GOOGL, $AMD | $AVGO vs. $NVDA, $NBIS vs. $IREN vs. $WYFI vs. $CRWV It's easy. Just buy $TSM, everything above is dependent on it. Undervalued and a monopoly, will likely run from $1.3T -> $1.8T MC based on forward growth + capex from ERs.
-
ALAB因覆盖五大科技巨头且高增长高毛利，有望复制英伟达路径实现千倍涨幅。
$ALAB 之所以起飞，是因为 Astera 是唯一一家系统性暴露于五大科技巨头（Mag7）中5家的小盘股。 1. ALAB 的客户群字面意义上就是 Mag7：$GOOGL、$MSFT、$NVDA、$META 和 $AMZN。 2. 他们的增长率令人惊叹，同比增长 ~150%+，毛利率为 76%（高于 NVDA，是半导体市场中最高的之一）。 3. 正如 $GOOGL 财报所示，AI 支出的资本支出仍在增加。 NVDA 从 1500 亿市值起步 -> 4 万亿+，因为超大规模数据中心需要 GPU。我们正处于 ALAB 在 AI 供应链中产生这种效应的起点，而其市值仅为 200 亿。 如果他们执行得当且 AI 基础设施建设持续，我们可能会看到下一个 1000% 的涨幅。
英文原文
The reason why $ALAB is taking off is because Astera is the only small cap company in existence with systemic exposure to 5 of the Mag7. 1. ALAB's customer base is literally Mag 7: $GOOGL, $MSFT, $NVDA, $META, and $AMZN. 2. They're growing at astounding rates ~150%+ y/y, with 76% margins (with higher margins than NVDA, one of the highest in the semi market). 3. Capex is still increasing for AI spend, as seen with $GOOGL earnings. NVDA started at a 150B market cap -> 4 TRILLION+ because hyperscalers needed GPUs. We're the start of this effect with ALAB for the AI supply chain, and it's starting at a small 20B market cap. We could see this being the next 1000% if they execute properly and AI infrastructure buildout continues.
-
分享GOOGL盈利及回购利好，看多科技股年底表现
关于我发布的WSB图表： 在$GOOGL上，已实现利润为101,428美元+，还有大量未实现利润，因为我预计如果科技股反弹持续，GOOGL年底将触及220美元+（$META $APPL $AMZN $MSFT $NVDA $SPOT）。 财报表现完全爆表，且他们仍有700多亿美元的股票回购计划 https://t.co/G6jrKAyumE
英文原文
For WSB charts I posted: On $GOOGL, realized profit was $101,428+, and lots of unrealized profit since I expect GOOGL to hit $220+ EOY if tech rally continues ( $META $APPL $AMZN $MSFT $NVDA $SPOT) Earnings were a complete blowout and they still have 70B+ in buybacks https://t.co/G6jrKAyumE