全部推文
共 6424 条 · 第 152 / 215 页 · 滚到底自动加载更多
-
机构算法交易二阶效应，博主致力于让供应链影响对散户更易理解。
谢谢，很多人看到了新闻，但大多数人并没有真正看到市场的一阶、二阶效应，或者无法得出可操作的见解。通常，这完全是全新的信息发现，比如关于 $AXTI 和磷化铟(InP)瓶颈，或者 $OSS 及其在美国对委内瑞拉突袭中的使用。但机构算法是基于这种映射进行交易的，我只是让供应链/关系影响对散户来说更易消化。
英文原文
Thanks, a lot of people see the news but majority don’t really see second-third order effects on markets or can come to actionable insights. Often times it’s completely novel information discovery like with $AXTI and InP bottlenecks or $OSS and their usage in the US raid in Venezulea But institutional algorithms trade off this mapping and I’m just making supply chains/relational impacts digestible for retail.
-
比特币市场平稳，市场已适应关税，建议忽略噪音继续持有。
@SukhTatla 比特币市场是一个相当不错的代理指标，价格几乎没有下跌。 我认为市场似乎不再太在意关税，因为关税已成为日常现象。 美国经济增长势头强劲，所以我个人建议继续持有并忽略噪音。
英文原文
@SukhTatla Bitcoin markets are a pretty good proxy, there’s almost no price drops. I don’t think markets seem to care much about tariffs anymore since this is a daily occurrence. US economic growth is an absolute powerhouse, so i personally would just stay long and ignore the noise
-
指出Lightpath($LPTH)有9000万积压订单，但DoD不管理其估值。
我只是在这里指出关键瓶颈以及身处其中的参与者。 不幸的是，我完全不知道美国国防部(DoD)的人心里在想什么。 我认为他们并不主动关心或管理市场给 $LPTH 定价的估值，因为他们不是对冲基金。Lightpath 确实有 9000 万美元的积压订单，所以他们正在接受长期订单。
英文原文
I'm just here to point out critical bottlenecks and the players that sit inside them. Unfortunately, I have no idea what goes on in the minds of people at DoD. I don't think they actively care or manage what valuations markets price $LPTH at since they're not a hedge fund. Lightpath does have a $90M backlog, so they are taking in longer term orders.
-
LPTH因独家锗替代技术成美军关键单一来源供应商，具极高战略价值。
$LPTH 不仅仅重要。 Lightpath 是美国国防部（战部）的关键故障点。 “锗”是 AI 领域的磷化铟(InP)，也是洛克希德“毒刺”导弹和 Anduril Ghost / $ONDS 无人机的关键材料。 锗的价格已翻三倍（2025/26年约 $4,150/公斤+）。 美国面临需求冲击，因为美国拥有量为零： 美国有零产能（是的，韩国 Zinc 在田纳西州有项目，但直到 2030 年才投产）。 美国开采锗矿石（阿拉斯加的 Red Dog 矿），但我们没有大规模将其精炼成金属的工业能力。 60-70% 的精炼锗来自中国，其余来自加拿大（Teck）、比利时（Umicore）和俄罗斯（也被禁运）。 而且中国禁止向美国军事项目出口。 从现在（2026年1月）到 2029/2030 年，实际上没有新的国内产能上线。 需求冲击： 就在供应切断的那一刻，三个领域的需求爆炸式增长： - AI 数据中心：锗是高速光纤电缆的必需品。 - 太空：它是卫星太阳能电池（如 Starlink 等）的主要材料。 - 国防（最大的）：每个 Anduril 无人机、夜视仪和导弹导引头传统上都使用锗透镜。 $LPTH 目前对五角大楼 + 美国国家安全具有生存级关键意义。你可能认为其他公司可以制造“无锗”玻璃，但这很难。 Lightpath 使用的特定材料 Black Diamond，是由美国海军研究实验室发明的。Lightpath 拥有商业化 Black Diamond 的独家许可。 如果洛克希德或雷神想要制造使用这种特定军用级锗替代品的导弹导引头，他们依法必须从 Lightpath 购买。 没有第二来源。 $LPTH 是许多先进武器的单一来源供应商。 示例 1）盟友正在囤积短程拦截器，例如“毒刺”导弹。盟友现在需要建造数千枚，但我们无法获得中国锗来制造导引头。 Lightpath 是洛克希德等公司的光学组件“单一来源”供应商。 示例 2）无人机公司制造“蓝色 UAS”（政府批准的无人机）。为了不被列入黑名单，他们必须移除中国部件。 像 $ONDS 或 $UMAC 这样的公司可能已将热成像相机供应单一来源给 Lightpath，以避免在无人机组件中使用中国供应链。 2026年（现在）：中国禁止出口。美国库存极低。 2030年（最早）：田纳西州的新“超级冶炼厂”（韩国 Zinc 交易）投产，实现国内锗精炼。 缺口：未来 4 年，美国没有国内金属供应。 这种情况看似荒谬，但由于美国关闭了大多数锌冶炼厂，我们意外地也关闭了锗供应。30年来，将矿石运往中国/加拿大比在这里处理 EPA 法规更便宜。 美国军方目前的处境是：拥有阿拉斯加的原始矿石，但必须请求加拿大或等待一家韩国公司（2030年）将其转化为导弹导引头。 人们认为 $LPTH 是某种廉价仙股。但它实际上现在是美国国家安全的中心。 Lightpath 是西方国家绕过锗瓶颈的主要方式，市值仅 $6.7 亿。
英文原文
$LPTH isn't just important. Lighpath is a critical point of failure for the US Dpt of War. "Germanium" is the AI's InP for Lockheed Stinger Missiles and Anduril Ghost / $ONDS Drones. Germanium's price tripled (~$4,150/kg+ in 2025/26). There is a demand shock since US has 0: The US has zero (Yes Korea Zinc in TN, but not until 2030). US mines germanium ore (Red Dog mine in Alaska), but we do not have the industrial capacity to refine it into metal at scale. 60-70% of refined Germanium comes from China and the rest elsewhere like Canada (Teck), Belgium (Umicore), and Russia (also banned). And China banned exports to US military programs. There is effectively no new domestic capacity coming online between now (Jan 2026) and 2029/2030. The Demand Shock: At the exact moment supply was cut, demand exploded from three sectors: - AI Data Centers: Germanium is essential for high-speed fiber optic cables. - Space: It is the primary material for satellite solar cells (Starlink, etc.). - Defense (The Big One): Every thermal Andruil drone, night vision goggle, and missile seeker head traditionally uses a Germanium lens. $LPTH Is existentially critical to the Pentagon + US National Security right now. You might think other companies can just make "germanium-free" glass, but it's hard. The specific material Lightpath uses called Black Diamond, was invented by the US Naval Research Lab. Lightpath holds the exclusive license to commercialize BLack Diamond. If Lockheed or Raytheon wants to build a missile seeker that uses this specific, military-grade germanium replacement, they legally have to buy it from Lightpath. There is no second source. $LPTH is the Sole Source supplier of many advanced weapons. Example 1) Allies are stockpiling Short Range Interceptor, eg. Stinger Missiles. Allies need to build thousands more now, but we can't get the Chinese germanium to build the seekers. Lightpath is the "sole source" supplier for the optical assembly for companies like Lockheed. Example 2) Drone companies build "Blue UAS" (government-approved drones). To stay off the blacklist, they must remove Chinese parts. Companies like $ONDS or $UMAC have likely sole-sourced their thermal camera supply to Lightpath to not use Chinese supply chains in their drone components. 2026 (Now): China bans exports. US stockpiles are critically low. 2030 (Earliest): The new "Super Smelter" in Tennessee (Korea Zinc deal) comes online to refine germanium domestically. The Gap: For the next 4 years, the US has no domestic metal. This situation seems absurd, but since the US closed most of its Zinc smelters, so we accidentally closed our Germanium supply with them. And for 30 years, it was cheaper to ship the ore to China/Canada than to deal with EPA regulations to refine it here. The US military is currently in a position where it has the raw ore in Alaska but has to ask Canada or wait for a Korean company (in 2030) to turn it into a missile seeker. People think $LPTH some cheap penny stock. But it's actually central to US national security now. Lightpath it's the main way around the Germanium chokepoint for Western countries at a small $670M MC.
-
看好SpaceX创始人埃隆推动公司市值达万亿。
@RKLBMan 埃隆正不惜一切代价试图让市值达到1万亿美元！ 我很想看到这一幕成真，他是我们这一代的爱迪生。
英文原文
@RKLBMan Elon is trying to hit $1T by any means necessary! I’d be interested in seeing it happen, the guy is the modern Edison of our generation
-
调侃特斯拉持有者期待某特定时间线
@amitisinvesting 那将是一个非常有趣的时间线。我敢肯定所有 $TSLA 的持有者都在暗自希望那会发生
英文原文
@amitisinvesting That would be one very interesting timeline. I’m sure all the $TSLA holders are secretly hoping that happens
-
市场对新消息反应平淡，恐慌情绪消退成为新常态。
@wallstengine 我喜欢市场（例如比特币）对此甚至不再有任何反应。如果是去年，我们会因这一消息下跌 3% 以上，所有人都会陷入恐慌。这就是新常态。
英文原文
@wallstengine I love how markets (eg. Bitcoin) don’t even react to this anymore. If this was last year, we’d be down 3%+ on the news and everyone would be panicking. This is the new norm
-
半导体板块分化，部分厂商因OpenAI诉讼风险受市场重估。
@8l8220656450 半导体板块整体没问题。只是某些供应商与 OpenAI 深度绑定，而许多其他公司如 $TSM 到 $MU 则相对独立。我确实看好 Elon，但我们将看到市场如何定价影响 OpenAI 关联公司的诉讼。
英文原文
@8l8220656450 Semiconductor space will be fine. Just certain vendors are heavily tethered toward OpenAI while many others like $TSM to $MU are more isolated. I do like Elon, but we'll see how markets price in the lawsuit impacting OpenAI relational companies
-
分析AI算力股与云厂商的算法关联及市场定价不确定性。
$GOOGL / Gemini 可能正在公开为这一结果欢呼。像 $CIFR 这样的公司主要通过 Fluidstack 与 Google 紧密绑定。然而，正如我们在上一轮下跌中所见，$NBIS 到 $IREN 在算法上被关联到 Oracle 和 Coreweave，无论 OpenAI 的关系影响如何。不确定市场这次会如何对此进行定价。
英文原文
$GOOGL / Gemini is probably openly cheering for this outcome. Companies like $CIFR are largely tethered to Google through Fluidstack. However, as we've seen last drop, $NBIS to $IREN were algorithmically associated with Oracle and Coreweave regardless of OpenAi relational impact. Not sure how markets price this in this time.
-
预测市场流动性低需审慎，不确定性风险及连锁反应仍存。
@CosmicSenate 这是一个很好的观点，我本应注明关于低流动性的问题。对于预测市场(Prediction Markets)的潜在结果，始终应持高度怀疑态度。无论如何，这带来了不确定性风险悬顶，且二阶/三阶连锁反应依然成立。
英文原文
@CosmicSenate This is a great point I should have noted regarding low liquidity. Prediction markets should always be taken with a large grain of salt regarding potential outcomes. Regardless, this presents an uncertainty risk overhang, and second/third order ripple effects still stand.
-
马斯克起诉OpenAI或引发AI产业链连锁反应，市场已部分定价该风险。
突发：埃隆·马斯克起诉 OpenAI，索赔 1340 亿美元。 预测市场今日给出的马斯克胜诉概率为 59%。 这可能令人担忧。 如果 OpenAI 败诉： 市场可能会引发与 OpenAI 相关的 AI 生态系统连锁反应（Contagion）。 $AMD 或 $CRWV 以及像 $ORCL（Stargate 项目）到 $MSFT 这样的公司将是巨大的输家，因为它们严重依赖 OpenAI。 此外，还有涉及 Coreweave 的 $APLD 和 $CORZ 等公司的三阶效应（Third-order effects）。 $AMD 的合作伙伴如 $RIOT 也是如此。 例如： 1. 微软/甲骨文 -> 数十亿美元的投资减值，建设 + GPU 利用率滞后将造成数十亿美元损失 2. CoreWeave -> 积压订单收入面临巨大风险，利用率滞后 + 已因巨额债务利息而处于风险之中 3. APLD / CORZ -> 依赖 Coreweave 的收入，如果租户倒闭，其短期收入也会随之下降。 如果 OpenAI 被迫支付巨额赔偿金，回归严格的非营利地位，或因流动性危机破产（无法履行合同义务）： 这可能会触发许多与 OpenAI 相关的 AI 板块股票的连锁反应事件。 当然，这是最坏的情况。 我的预期是中间地带： -> 马斯克赢得审判（或在判决前迫使达成和解）。 -> OpenAI 被迫支付巨额罚款（例如 100 亿 -200 亿美元）并可能开源其旧模型 -> $MSFT 可能会介入支付罚款或重组债务，因为利害关系太大。 或者在马斯克 41% 的败诉概率下：无事发生，与 OpenAI 相关的公司将安然无恙。 然而，市场似乎已经通过甲骨文（$330 -> $195）和 Coreweave（$150 -> $101）的价格大幅调整，定价了 OpenAI 连锁反应风险带来的大量不确定性。 这起诉讼绝对是市场关注的最大事件之一。
英文原文
Breaking: Elon Musk has sued OpenAI for $134 Billion. Prediction markets have priced the odds of Elon winning at 59% as of today. This may be alarming. If OpenAI loses this: Markets may trigger a contagion across the AI ecosystem tied to OpenAI. $AMD or $CRWV and companies like $ORCL (Stargate) to $MSFT are massive losers as they're heavily levered to OpenAI. Then there's third order effects on companies as $APLD and $CORZ involved to Coreweave. Same goes with $AMD partners, such as $RIOT. For example: 1. Microsoft / Oracle -> Write down billions in investment, buildout + GPU utilization lag would cost billions 2. CoreWeave -> Massive risk to backlog revenue, utilization lag + already at risk due to massive debt interest 3. APLD / CORZ -> Depend on Coreweave for revenue, if the tenant goes down, so does their near term revenue. If OpenAI are forced to pay massive damages, return to a strict non-profit status, or declare bankruptcy due to liquidity crises (and cannot fulfill its contractual obligations): This might trigger a contagion event across many AI sector names tied to OpenAI. Of course this is the worst case scenario. My expectation is middle ground: -> Musk wins the trial (or forces a settlement just before the verdict). -> OpenAI is forced to pay a massive fine (e.g., $10B–$20B) and possibly open-source its older models -> $MSFT likely steps in to pay the fine or restructure the debt since there's too much at stake. Or in the 41% chance Musk loses: nothing happens, and companies tied to OpenAI will be fine. However it looks markets already priced in a lot of that uncertainty with OpenAI contagion risk with Oracle ($330 -> $195) and Coreweave ($150 -> $101). This lawsuit is definitely one of the biggest events markets are watching.
-
英特尔因涉国家安全获两党支持，政策利好预期持续。
@traderofstox 我认为关于 $INTC 的问题在于，确保美国本土制造能力（如《芯片法案》）是两党共识。而采取持有10%股份等方式则不然。无论如何，鉴于英特尔已成为国家安全问题，预计其支持将在任何政府任期内持续。
英文原文
@traderofstox I think the issue regarding $INTC is bipartisan ensuring American domestic capability like CHIPS act. Other ways like taking a 10% stake, not so much. Either way support is expected to continue in any administration since Intel is now a national security issue.
-
识别AI供应链中工程变通与瓶颈环节的投资机会
@crux_capital_ 写得很好！我相信所有参与者都能从中受益，无论是像 $POET 和 $ALMU 这样的工程变通方案(Engineering Workarounds)，还是像 $AXTI 这样的瓶颈环节本身。通过识别正确的参与者，存在大量机会。
英文原文
@crux_capital_ Great writeup! I'm sure all players benefit both in engineering workarounds like $POET and $ALMU as well as the bottleneck itself like $AXTI. Lot of opportunities by identifying the right players.
-
回顾2019年市场对特斯拉保持领先地位的普遍误判。
@allen_drewe @zephyr_z9 市场？尤其是2019年它们进军中国时。事后诸葛亮，但大多数人认为 $TSLA 会保持其领先地位。 https://t.co/wyWecRjzkW
英文原文
@allen_drewe @zephyr_z9 The markets? esp. when they went to China in 2019. Hindsight bias, but majority of people thought $TSLA would maintain its leadership. https://t.co/wyWecRjzkW
-
类比特斯拉入华催生比亚迪，担忧台积电赴美建厂可能意外培育出美国本土芯片竞争对手。
我的意思是，$TSM 在美国建厂让我隐约想起当年 $TSLA 去中国生产电动汽车（EV）的情况，不过是反向的。当时大家都认为 $TSLA 通过在中国建厂会成为市场领导者 -> 结果 $BYD 作为副产品崛起，成为了中国本土电动汽车的领导者。我确信会有某种防火墙，更先进的制程节点仍留在台湾，但对于 $INTC 来说，谁也不敢保证。
英文原文
I mean $TSM building in US vaguely reminds me of the time $TSLA went to China for EVs but in reverse. Everyone thought $TSLA would be the market leader by building in China -> then $BYD popped out as China's domestic EV leader as a byproduct. I'm sure there's some sort of firewall with more advanced nodes staying in Taiwan, but you never know with $INTC.
-
看好三星、台积电和英特尔，认为2026年三者均将大幅跑赢大盘。
我持有三星、$TSM 和 $INTC 的敞口。它们就像火、水和草系初始宝可梦，选哪个都不会错。- 三星：在存储和代工领域拥有高上行空间，正在承接台积电的溢出订单。- $TSM：拥有惊人的盈利能力/增长，且被广泛使用，产能已满载。- $INTC：基本上由美国政府兜底下行风险，一旦完成建设，相对于 $TSM 有巨大的上行空间。我相信它们在2026年都会大幅跑赢大盘。
英文原文
I have Samsung, $TSM, and $INTC exposure. They're like Fire, Water, and Grass starter Pokemon, can't go wrong with any. - Samsung you have high upside with memory + foundry capturing overflow from TSM. - $TSM, has insane profitability/growth + everyone uses them and they're maxed out. - $INTC is basically US gov capping downside risk + large upside to $TSM once they finish buildout. I'm sure they'll all strong outperform 2026.
-
基于美国政策补贴和高确信度，看好Intel长期追赶台积电的7倍回报潜力。
$INTC 基本上就是半导体界的 $RKLB（对应 SpaceX），也是我最大的持仓之一。 众所周知，随着 Nvidia 到 Apple 等巨头的支持，它终将赶上 TSM，成为美国政府事实上的半导体臂膀。 这只是时间问题。 在 $2340 亿美元市值（$TSM 为 1.5 万亿美元）下，仅凭等待就能获得约 7 倍回报。 无论是 2 年还是 10 年，我坚信 Intel 是市场中罕见的机会：你可以投资白宫，以及用无限印钞机补贴现金消耗型晶圆厂开发。 半导体市场并非赢家通吃，总可寻址市场 (TAM) 将持续增长，$TSM 是很好的长线标的，但当 Intel 产能可用时，从 Google 到 Nvidia 的美国超大规模云服务商将承受使用 Intel 的压力。 美国不会让 Intel 失败，这是基于美国政策的高确信度长线投资。
英文原文
$INTC is basically the $RKLB to SpaceX and is one of my biggest positions. Everyone knows it will catch up to TSM as the de-facto semi arm of the US government with Nvidia to Apple backing it. It’s more of a matter of time. At a $234B MC ( $TSM is $1.5T), ~7X return just for waiting. Whether it’s in 2 years or 10 years, I strongly believe Intel is one of the rare opportunities in the market where you can invest in the White House and infinite money printer subsidizing the cash burning foundry development. It’s not winner takes all, the TAM of semis will keep going up, $TSM is a great long, but US hyperscalers from Google to Nvidia would be pressured to use Intel when available. United States won’t let Intel fail, this is a high conviction long on US policy.
-
AAOI作为垂直整合玩家，受益于美国制造及云厂商扩张。
@yianisz 同意，$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, $AMZN hyperscaler ramps.
-
确认InP供应链属国安优先，短期加剧ATI瓶颈，AXT略偏多。
好问题，我当时没把1+1联系起来，错过了对$AXTI的单跳关联影响。我的观点是，短期/中期基本无影响，但由于有6个月的空窗期，对AXT略微偏多。Emcore相比之下非常小，但美国政府正试图掌控上游控制的任何微小份额。这证实了美国政府现在将磷化铟(InP)供应链视为国家安全优先事项，并验证了我之前提出的InP瓶颈论点。TLDR：行政令接管InP供应链资产是为了长期摆脱对中国的依赖，但短期/中期$ATI的瓶颈甚至更严重。
英文原文
Good question, I didn't put 1+1 together at the time and missed the one-hop relational impact with $AXTI. My take is basically, no impact short/medium term, but leans slightly positive for AXT since there's a 6 month limbo. Emcore is very tiny in comparison, but US gov is trying to gain control of any little % of the upstream control. This was just affirmation that US government now views InP supply chains as a national security priority and is validating the InP bottlenek thesis I put out earlier. TLDR: EOs seizing InP supply chain assets was a move to regain independence from China long term, but short-medium term bottleneck is even more severe with $ATI
-
博主坚持投资论点，对多只AI供应链股进行长期持有与加仓。
@truncateit 我确实相信我的投资论点，所以我会对像 $AIRO 和 $LPTH 这样的标的进行成本平均加仓。但如果你好奇的话，是的，我仍然持有 $OSS、$AIRO、$LPTH、$SSYS、$AXTI、$VLN 以及其他我做过尽职调查(DD)的股票，因为它们旨在作为1年期的持仓，而不是12小时的短线交易。https://t.co/uNvk2q5uYe
英文原文
@truncateit I actually believe in my thesis so I cost average up on stuff like $AIRO and $LPTH. But if you're curious yes I still have $OSS, $AIRO, $LPTH, $SSYS, $AXTI, $VLN, and others I do DD on since they're meant to be 1 year holds, not 12 hour ones. https://t.co/uNvk2q5uYe
-
博主解释通过供应链研究提供新信息，并承认因仓促分析而发布更正。
感谢大家。我致力于通过供应链映射和推理，尽可能多地产出新颖的信息。这样散户就能凭借掌握他人尚未知晓的信息，在机构之前更早地布局不同的股票或板块。但这次我仓促进行了分析，导致论点出错，因此不得不发布更正。
英文原文
Appreciate it. I try to produce as much novel information as possible by doing supply chain mapping and inference. This way retail can be earlier to institutions on different stocks/segments by knowing something others don't yet. But I rushed the analysis on this one and messed up the thesis, so had to write a correction.
-
AXTI获买入评级及政策利好，验证其在AI基建中的关键地位。
这是一个中性进展。Craig-Hallum 对 $AXTI 给出 $26 目标价“买入”评级，以及特朗普关于磷化铟(InP)供应控制的行政令，只是再次确认了投资逻辑。大多数人一个月前在 $14 时认为 AXT 只是一只被炒作后抛售的 $600m 仙股，因为当时我是少数提及它的人。这类后续新闻进一步验证了 InP 和 AXT 在 AI 基础设施建设中扮演的关键角色。
英文原文
It's a neutral development. Craig-Hallum's $AXTI $26 PT "Buy" analyst report and Trump's executive order for InP supply control is just reaffirming the thesis. Majority of people thought AXT was just a pump/dump $600m penny stock a month ago at $14 since I was one of the only people to mention it. Follow-up news like this just validates how critical of a role both InP and AXT plays in the AI buildout.
-
特朗普行政令强制剥离InP资产，验证供应链瓶颈，利好AXTI等西方替代厂商。
行政令：“特朗普总统命令HieFo剥离磷化铟(Indium Phosphide, InP)资产” 关于1月2日行政令的后续报道显示，特朗普总统针对的是HieFo旗下Emcore Corp.的InP晶圆制造业务。 由于$AXTI和住友控制着约80%的供应链（Fastmarket Research数据），通过强制出售，美国可能确保InP供应链继续服务于美国国防承包商和AI半导体。 最初关于2日的报道对“芯片设计和制造资产”的表述较为宽泛，对InP的提及较少。 但现在很清楚，InP供应链已成为美国政府为了维持AI建设而面临的“无声紧急状态”。 如果美国总统签署行政令以获取InP供应链，这进一步验证了InP瓶颈论点。 我们可能会看到第三阶效应： -> 美国开始加速从$POET（光学中介层）到$ALMU（量子点）的工程替代方案。 -> InP瓶颈在短期内加剧，利好$AXTI（InP供应链+衬底）。 -> 可能影响$LITE及光学供应商的供应链中断（分析师报告对此有争议，涉及库存问题）。 -> 在Sumitomo/JX被出口管制切断对华材料供应后，西方替代品如$COHR变得更为关键。
英文原文
Executive Order: "President Trump orders HieFo to divest indium phosphide assets" A follow-up report on the Jan 2nd EO shows that President Trump targeted HieFo's InP wafer fabrication operations of Emcore Corp. As $AXTI and Sumitomo control ~80% of the supply chain (Fastmarket Research), by forcing the sale, the US likely ensures the InP supply chain remains available for US defense contractors and AI semiconductors. The original reporting on the 2nd was broad regarding the "chip design and fabrication assets" with light references to InP. But it's now clear this is now InP supply chains is becoming a silent emergency for the US government so the AI buildout continues. This is further validation of the InP bottleneck thesis if the US President is now signing executive orders to gain access to the InP supply chain. We'll likely see third order effects where: -> US starts rushes engineering alternatives from $POET (optical interposers) to $ALMU (quantum dot) -> InP bottleneck increase near term with $AXTI (InP supply chains + substrates) -> Potential influence supply chain disruption of $LITE and optical suppliers (this is debated from analyst reports over stockpiled) -> Western alternatives like $COHR becoming more critical after Sumitomo/JX are blocked off China material supply from export controls.
-
指出对WULF和APLD的关注多为套利而非真实做空。
@Mr_Derivatives 提醒一下其他人，其中很多细节很微妙。对于像 $WULF 或 $APLD 这样的数据中心公司，大量的关注仅仅是为了对冲/可转换票据套利，而非真正的做空。
英文原文
@Mr_Derivatives Just a heads up to the others, a lot of it is nuanced. With data centers like $WULF or $APLD, a lot of interest is just hedging / convertible note arbitrate and not actual shorts
-
Citron推断VGP为特斯拉Optimus供应商，作者澄清属非正式映射。
@ifreebirdie 是的，忘了包含它们，Citron 推断 $VGP 是 $TSLA Optimus 的供应商。 我也再次像最初的 $SSYS 文章那样措辞有些强烈，它们都是非正式的映射。很多事情并非公开信息。https://t.co/LBWGWbnySF
英文原文
@ifreebirdie Yeah forgot to include them, $VGP inferred to be $TSLA Optimus supplier per Citron. I also worded things a bit strongly again like the original $SSYS piece, they're all casual mapping. A lot of things aren't public information. https://t.co/LBWGWbnySF
-
因AI供应链机会成本高而不看好Adobe等，认为逆趋势难且非其偏好。
个人而言，我并不看好！这并非因为估值（即使打折后）问题，主要是机会成本，因为国防（如无人机）以及从存储到材料的AI供应链都有太多顺风因素。 我认为未来十年很难再看到像今年这样的一年，随机商品材料如磷化铟(Indium Phosphide)竟成为AI/光子学关键材料，且总地址市场(TAM)增长4000%+。 我认为如果分析师能比其他人更早识别出这些机会，这就好比现代淘金热。 话虽如此，像$ADBE、$CRM等公司可能会看到20%的反弹（你可以用期权来加杠杆），但很难逆趋势而行，这也不是我的菜。
英文原文
Personally, not really! It’s not due to valuations (even if discounted) but mainly opportunity cost since there’s so many tailwinds with defense like drones as well as AI supply chains from memory to materials. I don’t think we’ll see another year like this in a decade where random commodity materials like Indium Phopshide end up becoming critical for AI/photonics and TAM goes up 4000%+. I think if analysts are able to identify them before others due, it’s like the modern gold rush. That being said sure companies like $ADBE, $CRM and others might see a 20% rebound (and you can play options for leverage) but harder to fight the trend + not my cup of tea.
-
软件板块恐慌性抛售中，$FIG等公司因AI增效仍具价值。
是的，这次抛售确实令人印象深刻。但由于软件板块正遭受无差别的抛售，恐惧中蕴藏着大量机会。即便有AI，我确信人们仍会使用像$FIG（3个月跌幅53.3%）这样的公司，AI只是让它们的工作效率更高。其他一些公司更有投资逻辑，关键在于在广泛抛售中识别出哪些受影响、哪些未受影响。
英文原文
Yeah the selloff is impressive. But there’s a ton of opportunities on fear since software is being sold off indiscriminately. Even with AI, I’m pretty sure people will still use companies like $FIG (-53.3% 3M), it just makes them more productive on there. Others make more sense, just need to identify what’s impacted and what’s not during the broad selloff
-
列举近期持仓标的，表示将持有头寸一年以等待价值重估。
是的？最近发布的帖子包括：用于无人机的 $AIRO、用于边缘计算的 $OSS、用于机器人的 $SSYS、用于人工智能(AI)瓶颈的 $AXTI、用于国防瓶颈的 $LPTH 以及存在定价偏差的 $VLN。自重新评级(re-rating)不会在一两天或一两周内发生，可能需要数月才能显现，因此我将持有我的头寸一年。
英文原文
Yes? $AIRO for drones, $OSS for edge, $SSYS for robotics, $AXTI for AI bottlenecks, $LPTH for defense bottlenecks and $VLN for mispricing were the most recent posts. I’m holding my share positions for a year since re-rating doesn’t happen in a day or week. Might take months to play out.
-
重申NBIS和VLN的投资逻辑，认为市值将回归基本面。
我的投资逻辑应该能自行验证，无需我发帖。我对 $NBIS 在 2026 年的观点与去年完全一致。 $VLN 也是如此，目前市值约 2 亿美元，对应 8000 万美元远期营收，毛利率 63%，基本处于 1.2 亿美元的估值水平。 一旦做市商停止因今日全市场期权大规模到期而进行的操纵小动作，市值应该会回归基本面。
英文原文
Thesis should play out by itself without me posting. My thesis with $NBIS for 2026 hasn't changed at all since last year. Same with $VLN, it's basically sitting on $120m off $200m MC right now doing $80m forward revenue, 63% gross margins. MC should catch up to fundamentals once market makers stop with their shenanigans from mass OI expiry today across the board.
-
看好自杀式无人机板块，认为市值将追随技术演进。
@Nerd_0ne 从 $AVAV 到 $AIRO 的自杀式无人机真的酷毙了。我认为市值最终会跟随技术路线，就像 $RKLB 那样！
英文原文
@Nerd_0ne Kamikaze drones from $AVAV to $AIRO are really cool. I think the marketcaps follow the tech eventually like $RKLB!