供应链分析
产业链结构、上下游与瓶颈环节研究 · 共 1315 条 · 滚到底自动加载更多
-
CXMT和YMTC被西方供应链禁，对AI内存贸易无影响
@firstadopter 供参考，这两家公司均已被禁。该文章措辞具有误导性。 关于长鑫存储(CXMT)和长江存储(YMTC)的内存生产挑战内存公司的新闻。 新闻措辞有误。 > 它们已被列入西方超大规模数据中心(hyperscaler)供应链的黑名单。 > 这很可能仅限于国内市场。 当标题耸人听闻地称“中国两大内存制造商将全球供应短缺视为机遇……”时，因为它们在西方供应链中没有任何份额。 YMTC 被全面禁止（美国实体清单）- 国防部第 1260H 条清单（中国军事公司） CXMT - 国防部第 1260H 条清单。涉军清单。 美国商务部曾考虑是否将此类 DRAM 制造商等中国科技公司列入出口管制清单。但由于一些非西方供应链（如日本东京电子 Tokyo Electron 等公司）仍在使用或与其合作，因此尚未列入。 简而言之：对超大规模数据中心 AI 内存贸易没有影响。
英文原文
@firstadopter FYI they’re both banned. Article is disingenous with wording https://t.co/SLbweoaNdc
-
美系大厂转向高利润AI内存，中国厂商冲击有限，主要影响传统DRAM厂。
所以关于内存需求的那个说法是对的，但这并非文中所列公司的关注重点。 它们已基本放弃了YMTC/CXMT试图倾销的市场，转而追求利润率更高的超大规模AI基础设施建设，因此对$MU等公司影响甚微。 如果非要说有影响，那应该是打击像$NANYA和$WIN这样的传统型DRAM公司，这原本才是讨论的重点。 简而言之，$MU、$SKHYNIX和$SAMSUNG已基本清除了来自中国的市场需求（例如$MU约10%，$SKHYNIX约15-18%），并将DRAM晶圆产能转向HBM3e和HBM4。也许只有$SAMSUNG因其在移动/消费设备领域的份额而受到轻微影响。
英文原文
So that memory demand statement is true, but it's not the focus of the companies listed in the article. They've largely left the market YMTC/CXMT is trying to flood to go after the higher margin hyperscaler AI buildout, so little to no effect on companies like MIcron. If anything it would hurt legacy void type companies like Nanya + Winbond, which should have been the focus. Naunce is Micron, Sk Hynix, and Samsung have largely hollowed out their demand from China (eg. ~10% micron, ~15-18% Sk hynix), and shifted dram wafers to hbm3e and hbm4. Maybe only Samsung is slightly affected given their share with mobile/consumer devices.
-
CXMT和YMTC受制裁，不影响西方AI内存供应链。
关于长鑫存储(CXMT)和长江存储(YMTC)的内存生产挑战内存公司的新闻。 媒体的措辞是错误的。 > 它们被列入西方超大规模数据中心(hyperscaler)供应链的黑名单。 > 这很可能仅限于国内市场。 当标题耸人听闻地称“中国两大内存制造商将全球供应短缺视为机遇……”时，因为它们在西方供应链中没有任何份额。 长江存储(YMTC)被全面禁止（美国实体清单）- 国防部1260H条款清单（中国军事公司）。 长鑫存储(CXMT) - 国防部1260H条款清单。涉军清单。 美国商务部曾考虑是否将此类DRAM厂商列入出口管制清单。但尚未执行，因为一些非西方供应链仍在使用并与它们合作，例如日本东京电子(Tokyo Electron)等公司。 简而言之：对超大规模数据中心AI内存贸易没有影响。
英文原文
Regarding news of memory production by CXMT and YMTC challenging memory companies. This is worded wrong by news. > They're blacklisted in Western hyperscaler supply chains. > This is likely domestic only. Headlines are sensational when they say "China’s two major memory manufacturers view the global supply shortage as an opportunity..." because they have none in Western supply chains. YMTC is fully banned (US Entity List) - DoD Section 1260H list (Chinese Military Companies) CXMT - DoD Section 1260H list. Military-linked list. The U.S. Commerce Department was weighing whether to slap more Chinese tech firms like this DRAM player to its export restriction list. But haven't yet since some non-Western supply chains still use work with them eg. Japanese companies like Tokyo Electron. TLDR: no effect on the hyperscaler AI memory trade.
-
InP是AI基建刚需且依赖中国，垂直整合无法解决材料瓶颈。
观点中肯，但有两点误解： - 美国AI基础设施建设必须使用磷化铟(InP)，这是不可妥协的。我们近期看到“交易”案例：$NVDA GPU换取解除关键材料/原料/衬底禁令。美中虽为最大地缘政治对手，却不幸相互依赖。 - 尽管今日小幅回调，7N铟在SMM已创历史新高，而光子学(CPO)产能爬坡甚至尚未开始。你提到硅光(SiPh)替代，但无法替代InP。定制芯片可能采用硅+InP激光器的混合方案，但仍需InP。 - 需求呈指数级增长，$NVDA已锁定外延量子阱激光器(EML)产能，$MSFT Maia预计将占据两位数比例的InP供应链。我高度怀疑库存足以让所有人持续获取。 - 当然，或者他们可以通过涨价来弥补良率损失。 - 这不是垂直整合能解决的。整个供应链是物理/材料问题，因为大部分资源仅在中国可及。 超大规模云厂商可以垂直整合掉$LITE，但无法整合掉原始前驱体或材料。
英文原文
Fair points and two misunderstandings: - US requires InP for AI buildout. This is non-negotiable. And we saw that "trade", $NVDA GPUs for unblocking critical materials/feedstock/substrates bans recently. US and China unfortunately are dependent on one another despite being the biggest geopolitical adversaries. - We're already seeing 7n indium reach ATHs on SMM (depsite recent dip today), and photonics ramp hasn't even started the curve yet. You say SiPh substitution, but you can't substitute away InP. Custom chips might use hybrid w/silicon + InP laser but you still need it. - It scales up exponentially, $NVDA secured EML capacity, $MSFT maia is expected to take up double digits of inp supply chain. Highly doubt there's enough inventory for everyone to keep it accessible. - Sure, or they can price hike to make up for yield - This is not something you can vertically integrate. The entire supply chain is a physics/materials problem because majority is only accessible in China. Hyperscalers can vertically integrate away $LITE but not the raw precursors or materials.
-
SK海力士拟赴美上市，韩股高净利引美资关注
@Alisvolatprop12 三星电子(Samsung Electronics)和SK海力士(SK Hynix)在2026~2027年预计获得的净利润真的非常惊人。我听说SK海力士计划在美国股市上市，我确信许多美国投资者渴望获得韩国股票的投资敞口(Exposure)。
英文原文
@Alisvolatprop12 삼성전자와 SK하이닉스가 2026~2027년에 거둘 것으로 예상되는 순이익은 정말 엄청난 수준입니다. SK하이닉스가 미국 증시에도 상장될 예정이라고 들었는데, 많은 미국 투자자들이 한국 주식에 대한 투자 기회(익스포저)를 원하고 있을 것이라 확신합니다.
-
澄清标准化数据与一次性计算导致的报告差异
@adenois 正如我提到的，经标准化调整后，其平均值更接近 10 亿美元。但 2025 年的报告数据为 25.4 亿美元，这很可能是由于一次性计算所致。
英文原文
@adenois Normalized, it’s closer to $1B on average, which I mentioned. But reports for 2025 were $2.54B likely due to one time calculations https://t.co/tEPNctm0ZH
-
存储超级周期盈利强劲，AI需求推动价格结构性上涨。
$SNDK - 即使在上涨16%之后，其2027年的远期市盈率(Forward P/E)仍是个位数。 这就是存储超级周期(Memory Supercycle)的盈利能力，涉及美光($MU)、SK海力士、三星等公司。 AI需求+价格上调看起来像是新的结构性常态。
英文原文
$SNDK - Even after a 16% increase, the forward 2027 p/e is still single digits. That’s how profitable the memory supercycle is with $MU, SK Hynix, Samsung, and others The AI demand + price hikes looks like the new structural norm. https://t.co/6RG14zqEpr
-
对比存储巨头远期市盈率，指出若周期延续，行业利润或创历史新高。
存储公司概览（2月2日）。 2026/2027年远期市盈率(Forward P/E)预估： 美光(Micron, $MU) 年初至今(YTD) +31.5%： ~10.4倍 (2026年) 闪迪(Sandisk, $SNDK) | 年初至今(YTD) +109.36%： ~14.6倍 (2026年) ~6.3倍 (2027年) 三星电子(Samsung Electronics) | 年初至今(YTD) +17.04%： ~5.1倍 (2026年) ~4.3倍 (2027年) SK海力士(SK Hynix) | 年初至今(YTD) +22.6%： ~4.3倍 (2026年) ~3.4倍 (2027年) 数据源自：三星/SK海力士来自摩根士丹利(Morgan Stanley)研报；闪迪来自贝尔斯登(Bernstein)；美光来自汇丰(HSBC)报告。 闪迪/三星/SK海力士的数据最为最新。 美光的报告日期为1月23日。但今日最新数据显示“美光提议服务器DRAM(DRAM)环比(QoQ)增长115–125%”（来源dramexchange）。 鉴于此前传闻三星NAND价格上涨100%以上：新计算可能将美光的远期市盈率降至~6.9倍-7.3倍。 原始计算使用的是之前的分析师报告。 存储公司预期的净利润令人难以置信。但显然市场在三星/SK海力士的上涨中定价了一些怀疑情绪。 然而，如果存储周期持续至2027-2028年之后，这些公司可能会成为历史上最盈利的企业。
英文原文
Memory Company Overview (February 2nd). Forward P/E Estimates for 2026/2027: Micron ( $MU ) +31.5% YTD: ~10.4x (2026) Sandisk ( $SNDK ) | +109.36% YTD: ~14.6x (2026) ~6.3x (2027) Samsung Electronics | + 17.04% YTD: ~5.1x (2026) ~4.3x (2027) SK Hynix | +22.6% YTD: ~4.3x (2026) ~3.4x (2027) Numbers were derived from Morgan Stanley's research note for Samsung/SK Hynix. Bernstein for Sandisk. HSBC report for Micron. Sandisk/Samsung/SK Hyinx are most up to date. Micron's report was January 23. But recent updates today show "Micron Proposed A 115–125% QoQ Increase for server DRAM" (dramexchange). With rumors that NAND prices increased 100%+ from Samsung the other week: new calculations would may bring Micron to ~6.9x-7.3x forward P/E. Original calculations were using prior analyst reports. The net income memory companies are expected to bring in is absurd. But clearly the market is pricing in some skepticism with the rally (with Samsung/SK Hynix). However if the memory cycle continues past 2027-2028, these companies could be the most profitable in history.
-
看好AI前景，认为SK海力士等净利润预测荒谬，短期波动将快速消化。
@Sandeman52 看到晚了——我同意。我对人工智能(AI)依然极度看好，关于三星(Samsung)到SK海力士(SK Hynix)的一些净利润预测简直荒谬。这大概是短期波动，一旦做空净头寸的机构在其他板块重新建立多头头寸，这些波动将极快地被消化。
英文原文
@Sandeman52 Saw this late - and agreed. I remain extremely bullish on AI, some of the net income projections on Samsung to SK Hynix are ridiculous. Probably short term volatiality that will get eaten up extremely fast once institutions that were short net reposition long on other sectors
-
CME保证金上调引发白银抛售，预计机构调仓后价格将大幅反弹。
我最大的猜测是，下跌很可能源于保证金上调的机制。但我希望我猜错了。 我们已看到芝加哥商品交易所(CME)提高保证金，例如白银从11%升至15%-16.5%+（今日生效），我猜测机构正利用此机会清除更多杠杆。 周五，市场目睹了白银的人为抛售，很可能是由于金银银行和基金在空头头寸上陷入亏损。 交易中的纸白银价格与实物需求（用于AI+工业）之间存在巨大的价格分化。 但一旦抛售结束，机构利用纸面清算进行净多头重新配置，我们很可能会看到大幅反弹，因为新任美联储主席引发的更广泛抛售更多是表象而非政策实质改变。 只是不知道时间框架。
英文原文
My best guess is likely down due to mechanics of margin hikes. But I hope I'm wrong. We've seen CME increase margins eg. silver 11% -> 15%-16.5%+ values (taking effect today) and my guess is institutions use this opportunity to wipe out more leverage. On Friday, markets saw manufactured selloffs on Silver likely due to bullion banks and funds being underwater on shorts. There's a huge bifurcation in prices between Paper Silver being traded and the physical demand (used for AI + industrial). But once that sell-off is done and institutions use paper liquidations to net-reposition long, we'll likely see a wide recovery since a lot of what happened regarding broader selloff from the new Fed chair is optics rather than policy change. Just don't know timeframe.
-
分析稀土算法误卖后的中性前景及CRCL在QT与降息矛盾下的长期看涨逻辑。
稀土板块从 $MP 到 $USAR 的抛售是由于路透社发布虚假信息后撤回导致的算法错误。然而，就像比特币最初因币安错误导致的清算一样，损害已经造成。不幸的是，我看到 Warsh 对 $CRCL 总体看空。这非常微妙，因为这两件事相互矛盾：缩减资产负债表（量化紧缩 QT）是“紧缩”，而降息是“宽松”。但如果美联储降息（利好加密货币/AI增长）且长期国债收益率因美联储资产负债表缩小而保持高位，Circle 可以继续从其储备中获得高利息，同时更广泛的加密货币市场反弹。（这是“金发姑娘”完美情景）。否则，QT 和降息只会严重伤害 Circle。尽管如此，USDC 的成交量/供应量仍保持在 700 亿美元以上的历史新高，我短期保持看空中性，但在 149 亿美元市值下长期非常看好。回到稀土，本质上它是中性的。就像海外的三星/台积电一样，尽管可能有一些逆风。最终，美国的 AI 增长政策离不开它们或稀土，所以它们是供应链中的必需品。从根本上说，在那样一次重大崩盘后（即使是假的），恢复需要更长的时间，但也提供了良好的逢低买入机会。但它们年初至今（YTD）已经涨了很多，这就是为什么未来可能会有进一步的波动（例如 $USAR 年初至今上涨 58%）。
英文原文
Rare earth sector from $MP to $USAR sell-off was an algorithmic mistake due to Reuters publishing false information -> retracting it later. However, like Bitcoin's initial liquidation from Binance errors, the damage has been done. I see Warsh is unfortunately net bearish $CRCL. So it's extremely nuanced since these two things contradict each other: shrinking the balance sheet (Quantitative Tightening) is "tightening," while cutting rates is "easing. But if the Fed cuts (bullish for crypto/AI growth) and long-term Treasury yields stay elevated because of a smaller Fed balance sheet, Circle can continue to earn high interest on its reserves while the broader crypto market rallies. (this is goldilocks scenario). Otherwise QT and rate cuts just hurts Circle a lot. But despite this, USDC volume/supply remains ATHs above $70B+ and I remain bearish-neutral short term, very bullish long term at $14.9B MC. Going back to rare earths, it's essentially neutral. Like Samsung/TSM in foreign countries, even though there might be some headwinds. Ultimately US policy with AI growth cannot function without them + or rare earths so they're a necessity in the supply chain. Fundamentally after a major crash like that (even if it's fake) recovery takes longer but presents good dip opportunities. But they're still up a ton YTD, which is why there might be further volatility ahead (eg. $USAR 58% YTD).
-
白银暴跌引发连锁清算，建议转向AI及高现金流标的防御。
市场正经历清算级联效应。白银的崩盘现已蔓延至加密货币及美/外股市。具体情况如下： 以下是预期走势： - $BMNR（加密货币） - $RKLB（高贝塔值） - $SNDK（AI） - 至三星（外国股票）。 “Warsh”美联储主席提名是引发抛售的初始触发因素，市场将其视为“鹰派”->量化紧缩（Quantitative Tightening）。然而，这是一个误解，因为美联储主席很可能与特朗普的政策保持一致，且由于AI因素，其近期立场偏向短期鸽派+降息。 然而，抛售的技术性原因是CME+交易所控制强制白银保证金清算。随着白银日内暴跌33%，机构被迫清算其他标的并进行对冲。因此，我们看到了： 1. 恐惧传染——当避险金属如此暴跌时，会在其他板块引发恐慌。 2. 避险情绪——投资者恐慌性抛售“风险”资产和股票/转向美元和国债。 为了进行防御性操作： - 最好将投机性标的重新配置为符合新政策的自由现金流（FCF）/盈利型多头头寸。 - 如 $GOOGL、$NVDA 跌至 $MU、$TSM、$JPM、三星等标的将受益最大。 - 从 $PYPL 到 $SNAP 等已处于低位（具有强劲预期自由现金流）且被进一步抛售的股票，呈现出不错的反弹上行空间。 尤其是鉴于美联储主席预计对从AI到银行业的许多板块持看涨态度，且由AI增长+生产力推动的降息预期强烈。 警告： - 更多不产生巨大自由现金流的投机性小盘股（从 $ONDS 到 $RKLB）可能因与高贝塔板块抛售的相关性而面临更大风险。 - 像 $BMNR 这样持有非流动性资产（例如Mr. Beast公司2亿美元）的杠杆基金，以及像 $QBTS 或 $RGTI 这样的投机性标的，可能会最终看到重置/清零。 - 像 $JD 到 $BABA 这样的外国市场标的或像 $MELI 这样的新兴市场标的可能会受到流动性流失的影响。 当然，日内交易者可能会在高贝塔标的的反弹时机上大显身手（例如，如果 $ETH 闪崩12%至$2.1K -> 恢复至$2.3K）。 话虽如此，这并不是说“卖出高贝塔”。 鉴于以太坊从$3k+跌至$2.18k，这只是对目前在高贝塔板块使用保证金的人发出的警告： 如果高贝塔股票出现持续抛售，风险相当大。（我个人将以太坊作为代理指标）。 这只是个人市场观点，但总体而言，随着中期选举临近+更多预期的降息+财报表现创历史新高（例如 $SNDK 的爆发），对市场保持极度看涨是好的。 市场在中期选举前看到绿色V型复苏只是时间问题。 基本面没有改变，但表象和短期流动性改变了。
英文原文
Markets are seeing liquidation cascades. Silver's crash is now extending into other markets like Crypto and US/Foreign stocks. Here's what's happening: And here's what to expect from: - $BMNR (Crypto) - $RKLB (High-Beta) - $SNDK (AI) - to Samsung (Foreign). The "Warsh" Fed Chair nomination was the initial trigger that caused the selloff as markets viewed him as a "Hawk" -> Quantitative Tightening. However, this is a mistake as the fed chair is likely aligned with Trump's policies, and his recent stance is dovish short term + rate cuts, due to AI. However, the technical reason for the selloff was CME + Exchange controls forcing margin liquidations on Silver. As silver crashed 33% intra-day, institutions are forced to liquidate other names and hedge. So, we're seeing both: 1. Fear Contagion - when a safe haven metal plummets this much, this causes a panic across other sectors. 2. Flight to Safety - investors panic-sell "risky" assets and stocks / move to U.S. Dollar and Treasury Bonds. To play defensive: - It's best to reposition speculative names into FCF/profit generating long positions aligning with these new policies. - Names like $GOOGL, $NVDA down to $MU, $TSM, $JPM, Samsung, and others stand to benefit the most. - Stocks that are already at lows (with strong expected FCF) from $PYPL to $SNAP that are being sold off even more present decent recovery upside. Especially since the Fed chair is expected to be bullish for many sectors from AI to Banking, with rate-cuts fueled by AI growth + productivity. For a warning: - More speculative small cap names (that don't generate massive FCF) from $ONDS to $RKLB may be more at risk due to correlation to high-beta sector selloffs. - Leveraged funds like $BMNR with iliquid assets (eg. $200M in Mr. Beast's company) to speculative names like $QBTS or $RGTI may finally see a reset/wipeout. - Foreign market names like $JD to $BABA or emerging market names like $MELI may be impacted from a liquidity drain. But of course, day-traders may have a field day timing rebounds on high-beta names (eg. if $ETH flash crashes 12% to $2.1K -> recovery to $2.3K). That being said this is not saying "Sell High-Beta". This is just a warning to people with margin on high-beta sectors now that given Ethereum's flush from $3k+ down to $2.18k: There's considerable risk if there's an extended selloff on high-beta stocks. (I've been personally looking at Ethereum as a proxy). This is just personal market opinion, but generally as midterms come up + more expected rate cuts + earnings coming out higher than ever (eg. $SNDK's blowout), it's good to remain extremely bullish on the market. And it's just a matter of time before markets see a green V recovery before midterms. Fundamentals haven't changed but optics have and short-term liquidity have.
-
提供SK海力士与三星数据供事实核查
@stefano_kerope 这是 SK 海力士(SK Hynix) 和三星(Samsung) 的数据，供想核实事实的人参考。 https://t.co/v24Y93epWT
英文原文
@stefano_kerope These were the figures for Sk Hynix and Samsung for people that wanted to fact-check. https://t.co/v24Y93epWT
-
摩根士丹利预测2027年存储巨头利润登顶，确认存储超级周期已至。
存储超级周期已至。 据摩根士丹利(Morgan Stanley)最新报告估算： - 三星(Samsung)净利润约1630亿美元 - 海力士(SK Hynix)约1209亿美元。 这将使三星成为全球最赚钱的公司，甚至超过$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 $MU, Samsung, and SK Hynix is here clearly here.
-
金银暴跌致短期流动性流失，但美联储主席看多AI，中期前景乐观。
我认为白银/黄金暴跌会导致全面（比特币、以太坊、$SPY、中盘股等）的短期流动性流失。新任美联储主席可能是你见过的最大的AI多头之一，他主张降息以引导AI增长+生产力提升。因此，一旦白银/黄金抛售被买回，中期前景真的非常棒。
英文原文
I see it as a short term liquidity drain across the board (Bitcoin, Ethereum, SPY, mid-cap, etc) from Silver/Gold crash. The new Fed Chair is probably one of the biggest AI bulls you will ever see and is arguing for rate cuts to funnel AI growth + productivity. So really great medium term once the silver/gold selloff gets bought back.
-
贵金属暴跌引发流动性真空，短期利空但中长期利好AI科技股。
周五白银下跌28%，黄金下跌8.35%，估计蒸发了约7.34万亿美元。这造成了巨大的流动性真空，导致比特币、科技股和小盘股等其他资产被抛售，以满足追加保证金通知或筹集现金。考虑到商品交易所(CME)可能在2日再次提高几个百分点的保证金要求，这很可能是一次短期（例如24-48小时）的清洗。然而，中长期来看这实际上是看涨的。强势美元使人工智能(AI)公司能够以更低的成本在全球进行采购，而沃什(Warsh)的“AI生产力繁荣”论点允许短期降息，这将推动高增长科技股。
英文原文
Silver dropped 28% and Gold dropped 8.35% on Friday, est. ~$7.34T wiped out. This creates a large liquidity vacuum where other assets from Bitcoin, tech, to small cap gets sold off to meet margin calls or raise cash. Is this likely a short term (eg. 24-48) hour flush, with CME raising margin requirements by another few percent on the 2nd. However, medium-long term this is actually bullish. A strong dollar makes the AI companies the ability to buys globally cheaper, and Warsh’s "AI Productivity Boom" thesis allows short term rate cuts, which fuel high-growth tech.
-
解析新任美联储主席 Warsh 政策对 AI、金属、加密及全球股市的差异化影响。
Kevin Warsh 是下一任美联储主席。 市场可能会误以为他是一只“鹰派”。 但他 2026 年的实际立场是微妙的。 以下是他的政策及其对市场的影响： 1. AI/半导体（$NVDA, $MU）：极度看涨 2. 金属（白银、黄金）：极度看跌 3. 加密货币（$BTC, $CRCL）：悖论式看涨 4. 银行与金融（$JPM, $BOA）：看涨 5. 住房与房地产：混合/不确定 6. 可再生能源：看跌 7. 小盘股（$RUT）：看涨 8. 外国股票（日本、韩国）：具有韧性 - 新兴市场（EM）：极度看跌 - 中国与香港：看跌 - 欧洲（$VGK, $EZU）：谨慎 1. AI/半导体（从 Nvidia 到 Micron）：极度看涨 Warsh 是 AI 多头。 在 2025 年底，他认为 AI 是一种强大的抗通胀力量。他相信 AI 驱动的生产力提升将使经济在不会引发通胀的情况下快速增长。 这种“生产力繁荣”为他提供了智力上的“掩护”，即使经济保持强劲，他也支持降息。（《美联储破碎的领导层》，2025 年 11 月 16 日《华尔街日报》） 这与他早期立场大相径庭，当时市场预计他会是僵化的通胀鹰派（希望利率更高的人）。 他倡导降息并希望加速 AI 发展。 2. 金属（白银、黄金）：极度看跌 投资者使用黄金作为对冲弱势美元和“印钞”的工具。因为 Warsh 希望缩减资产负债表并关闭“印钞机”，持有黄金的主要理由正在减弱。强势美元使得金属对国际买家来说更昂贵。 话虽如此，白银日内 33% 的下跌主要是由其他因素造成的，例如保证金变动引发的连锁清算，尽管新任美联储主席可能起到了次要作用。 3. 加密货币（$BTC, $CRCL）：悖论式看涨 他曾著名地表示：“如果你不到 40 岁，比特币就是你的新黄金。”他将比特币视为合法的价值储存手段，并认为这是从实物金属向代际转变。 他将区块链视为“最新、最酷的软件”，并相信美国必须在此领域保持领先，以在全球竞争中保持经济竞争力。 然而；“悖论”：为何价格下跌： 市场意识到，虽然 Warsh 希望降低利率，但他也希望美联储资产负债表更小。 投资者害怕我们进入“降息但不量化宽松（QE）”的时代。你可能会得到更便宜的贷款，但不会得到通常将 $BTC 推向历史新高的巨大“资金墙”。 所以我们有一个人对加密货币技术看涨，但他的货币纪律可能会损害短期流动性。 4. 银行与金融：看涨 由于他在摩根士丹利的经历以及对“使命蔓延”的直言不讳的批评，Warsh 是银行业界的宠儿。预计他将回滚复杂的银行资本要求（如巴塞尔协议 III）。分析师认为这对区域性和小盘银行是一个重大利好，因为它释放了用于贷款的资本。 5. 住房与房地产：混合 他希望激进地降低联邦基金利率。这将立即降低可调利率抵押贷款（ARMs）和建筑贷款的成本。 然而，看跌的情况是 Warsh 强烈反对美联储持有 2 万亿美元的抵押贷款支持证券（MBS）。许多经济学家警告，这可能会推高 30 年期固定抵押贷款利率（可能升至 7% 或 8%），即使美联储正在降低其他利率。 6. 可再生能源：看跌 他打算让美联储退出全球气候团体（如绿色金融系统网络）并结束银行的“气候压力测试”。 在杰罗姆·鲍威尔任内，美联储鼓励银行在贷款中考虑气候风险。Warsh 希望结束这种做法，这实际上消除了使绿色项目更容易从主要银行获得优惠贷款条款的“监管推动”。 7. 小盘股 Warsh 明确表示，他希望美联储关注经济的“真正驱动力”，即中小企业和企业家，而不是华尔街的“娇生惯养的王子”。 预计 Warsh 将领导大幅回滚复杂的银行资本要求。这对小盘股强烈看涨。他打算通过减轻从事大多数小企业贷款的小型和区域性银行的监管负担，扩大小企业获得资本的途径。 8. 外国股票 预计 Warsh 将在受益于强劲美国经济和易受强势美元及全球流动性收紧影响的国家之间造成分裂。 日本/韩国（三星、SK 海力士等）：日本和韩国“没问题”，因为它们拥有 Kevin Warsh 认为将拯救美国经济的 AI 和机器人交易的物理瓶颈。 通常，强势美元对外国股票不利，但对于日本和韩国来说，这是一种竞争武器： - 出口提振：由于他们的大多数 AI 和机器人合同以美元计价，强势美元意味着他们的收入（换算回日元等）大幅膨胀。 - 对美国更便宜：Warsh 的“强势美元”政策使得日本机器人和韩国芯片对美国公司来说更便宜。这加速了 Warsh 想要的“生产力繁荣”，同时增加了这些外国科技巨头的利润。 中国：强势美元给人民币带来压力，使中国人民银行（中国央行）更难降息以刺激其 struggling 的经济。 新兴市场：强势美元使得新兴国家偿还美元计价债务的成本更高。 欧洲：美元复苏可能会压低欧元，这有助于欧洲出口，但增加了其能源进口成本。 _ 周五，随着白银/黄金暴跌，市场大幅抛售，对冲操作抽走了系统中的流动性。 市场可能会将 Warsh 误认为是历史上的鹰派。 然而，最近的声明显示他短期内偏向鸽派，并支持由 AI 加速的降息。 市场目前正在定价同时降息和缩减资产负债表的可能性，但总体而言，预计从 AI 到小盘成长的许多交易将继续进行。
英文原文
Kevin Warsh is the next Federal Reserve Chair. Markets may confuse him as a "Hawk". His actual stance in 2026 is nuanced. Here's his policies and how they affect the markets: 1. AI/Semis ( $NVDA, $MU): Extremely Bullish 2. Metals (Silver, Gold): Extreme Bearish 3. Crypto ( $BTC, $CRCL ): Paradoxically bullish 4. Banking & Financials ( $JPM, $BOA ): Bullish 5. Housing & Real Estate: Mixed/Uncertain 6. Renewable Energy: Bearish 7. Small-Caps ( $RUT ) : Bullish 8. Foreign Stocks (Japan, Korea): Resilient - Emerging Markets (EM): Extremely Bearish - China & Hong Kong: Bearish - Europe ( $VGK, $EZU): Cautious 1. AI/Semis ( Nvidia to Micron ): Extremely Bullish Warsh is an AI Bull. In late 2025, he argued that AI is a powerful dis-inflationary force. He believes AI-driven productivity gains will allow the economy to grow rapidly without triggering inflation. This "productivity boom" gives him the intellectual "cover" to support rate cuts even if the economy remains strong. (The Federal Reserve’s Broken Leadership, November 16, 2025 WSJ) This is much different than his earlier stances where markets expected him to be a rigid inflation hawk (someone who wants higher rates). He is advocating for cuts and wants to accelerate AI development. 2. Metals (Silver, Gold): Extreme Bearish Investors use gold as a hedge against a weak dollar and "money printing." Because Warsh wants to shrink the balance sheet and turn off the "printing press," the primary reason for holding gold is diminishing. A stronger U.S. Dollar is making metals more expensive for international buyers. That being said the 33% intraday silver drop was mainly from other factors such as cascading liqudation from margin changes, though the new Fed chair likely played a minor role. 3. Crypto ( $BTC, $CRCL ): Paradoxically bullish He famously stated, "If you're under 40, Bitcoin is your new gold." He views Bitcoin as a legitimate store of value and a generational shift away from physical metals. He views the blockchain as "the newest and coolest software" and believes the U.S. must lead in this space to remain economically competitive against global rivals. However; The "Paradox": Why Prices are Dropping: The market is realizing that while Warsh wants lower interest rates, he also wants a smaller Fed balance sheet. Investors are terrified that we are entering an era of "Rate Cuts without QE." You might get cheaper loans, but you won't get the massive "wall of money" that usually sends $BTC to all-time highs. So we have a guy bullish on the technology of crypto, but his monetary discipline might hurt short-term liquidity. 4. Banking & Financials: Bullish Warsh is a favorite of the banking sector due to his experience at Morgan Stanley and his vocal criticism of "mission creep." He is expected to roll back complex bank capital requirements (like Basel III). Analysts see this as a major win for regional and small-cap banks, as it frees up capital for lending. 5. Housing & Real Estate: Mixed He wants to cut the Federal Funds Rate aggressively. This would immediately lower the cost of Adjustable-Rate Mortgages (ARMs) and construction loans. However, the bear case is that Warsh is a fierce opponent of the Fed owning $2 trillion in Mortgage-Backed Securities (MBS). Many economists warn this could push the 30-year fixed mortgage rate higher (potentially toward 7% or 8%) even as the Fed is cutting other interest rates. 6. Renewable Energy: Bearish He intends to withdraw the Fed from global climate groups (like the Network for Greening the Financial System) and end "climate stress tests" for banks. Under Jerome Powell, the Fed encouraged banks to consider climate risks in their lending. Warsh wants to end this, which effectively removes the "regulatory nudge" that made it easier for green projects to get favorable loan terms from major banks. 7. Small-Caps Warsh has explicitly stated that he wants the Federal Reserve to focus on the "true drivers of the economy", small businesses and entrepreneurs, rather than just the "pampered princes" of Wall Street. Warsh is expected to lead a significant rollback of complex banking capital requirements. This is strongly bullish for small caps. He intends to broaden access to capital for small firms by reducing the regulatory burden on the small and regional banks that do the majority of small-business lending. 8. Foreign Stocks Warsh is expected to createa a divide between countries that benefit from a strong U.S. economy and those that are vulnerable to a stronger U.S. Dollar and tighter global liquidity. Japan/Korea (Samsung, SK Hynix, etc): Japan and Korea are "fine" because they own the physical bottlenecks of the AI and robotics trades that Kevin Warsh believes will save the U.S. economy. Usually, a strong USD is bad for foreign stocks, but for Japan and Korea, it’s a competitive weapon: - Export Boost: Since most of their AI and robotics contracts are priced in USD, a stronger dollar means their revenue (when converted back to Yen, etc.) is massively inflated. - Cheaper for the U.S.: Warsh’s "Strong Dollar" policy makes Japanese robots and Korean chips cheaper for American companies to buy. This accelerates the "Productivity Boom" Warsh wants while padding the profits of these foreign tech giants. China: A stronger dollar puts pressure on the Yuan, making it harder for the PBoC (China's central bank) to cut their own rates to stimulate their struggling economy. Emerging Markets: A stronger U.S. Dollar makes it much more expensive for emerging countries to service their dollar-denominated debt. Europe: Dollar recovery could push the Euro lower, which helps European exports but increases their energy import costs. _ On Friday, markets sold off sharply on Silver/Gold crashing, and hedging pulled liquidity out of the system. Markets might confuse Warsh as a historical hawk. However, recent statements show he's near term dovish and supports lower rates, accelerated by AI. Markets are currently pricing in the possibility of simultaneous rate cuts and balance sheet reductions but generally, many trades from AI to small cap growth are expected to continue.
-
列举五大行操纵白银历史，暗示2026年暴跌或为再次操纵。
5家主要银行因操纵白银市场被刑事起诉。 这些罚款总额超过12亿美元，罪名是“通过幌骗和价格操纵来操纵白银市场”。 周五白银暴跌28%，引发数万亿美元损失。 以下是指控详情： 1. 摩根大通（$JPM，9.2亿美元，2020年）：JPM向美国司法部承认在白银市场操纵中行为不当。 2. 加拿大丰业银行（$BNS，1.275亿美元，2020年）：美国司法部指控其“在白银市场中进行欺诈性和操纵性交易”。 3. 汇丰银行（$HSBC，7660万美元，2018-2023年）：商品期货交易委员会（CFTC）对其2011-2020年的幌骗行为进行执法。 4. 德意志银行（$DB，7550万美元，2016-2021年）：涉及白银定盘价结算及CFTC对1999-2014年操纵行为的行动。 5. 摩根士丹利（$MS，150万美元，2019年）：CFTC对其2013-2014年幌骗行为的民事处罚。 来源：Navnoor Bawa。 在2008-2016年间，这五家主要银行通过幌骗、价格操纵和基准操纵系统性地操纵贵金属市场。 监管机构发现了他们。罚款和定罪跨越了2016-2025年。 这是历史上白银市场被操纵并被掩盖的实质性证据。 而且，随着2026年当前的白银暴跌，市场可能正在再次目睹这一幕。
英文原文
5 Major Banks were criminally charged for manipulating Silver markets. These fines totaled $1.2B+ for "manipulation of silver markets through spoofing and price rigging". Silver crashed 28% on Friday, triggering trillions of losses Here were the charges: 1. JPMorgan ($920M , 2020): JPM admitted wrongdoing for market manipulation of Silver to the US DoJ. 2. Scotiabank ($127.5M, 2020): US DoJ - "Fraudulent and manipulative trading practices in the markets for silver" 3. HSBC ($76.6M, 2018–2023): CFTC enforcement for 2011–2020 spoofing 4. Deutsche Bank ($75.5M, 2016–2021): Silver fix settlement and CFTC actions for 1999–2014 rigging 5. Morgan Stanley ($1.5M, 2019): CFTC civil penalty for 2013–2014 spoofing Credit: Navnoor Bawa. Between 2008–2016 five of these major banks systematically manipulated precious metals markets through spoofing, price rigging, and benchmark manipulation. Regulators caught them. Fines and convictions spanned 2016–2025. This was concrete evidence of market manipulation of silver historically that was swept under the rug. And, it's possible markets are witnessing it again with the current silver crash of 2026.
-
确认坍塌后果，提示钽供应瓶颈对AI硬件供应链的潜在影响。
是的，这确实令人痛心。初步的坍塌报告发生在29日，但后果的规模刚刚得到确认。关于二级效应，许多投资者并没有深入追溯晶圆厂 $TSM 或像 $NVDA 这样的芯片制造商所使用的原材料上游，但在企业级固态硬盘(SSD)到GPU的钽(tantalum)应用中，替代品（如铌(Niobium)）非常有限。所以只是想分享这条新闻以及一些供大家调查的潜在行动项。
英文原文
Yes this is incredibly sad. Initial collapse reports happened on the 29th, but scale of the aftermath was confirmed just now. In regards to second order effects, many investors don't really go upstream to the raw materials that fabs $TSM or chipmakers like $NVDA use but there's limited substitutes (eg. Niobium) for tantalum in enterprise ssds to gpus. So just wanted to share the news and some potential action items for people to investigate.
-
Metropolitan Capital Bank倒闭系高风险投资所致，属孤立事件，警惕$MCB错杀机会。
今年第一家美国银行刚刚倒闭： Metropolitan Capital Bank & Trust 已被联邦存款保险公司(FDIC)关闭并接管。 以下是可能发生的事情以及市场可能的反应： 这家芝加哥银行按资产规模($2.61亿)属于社区银行，但以精品投资银行的风险水平运营。 他们涉足三个领域： - 科技赋能体育 - 数字媒体 - 小众房地产建设 从公开信息来看，这两个波动性行业存在高风险集中。 特别是由老虎伍兹创立的 TMRW Sports。该银行不仅促成股权交易，还以此作为贷款抵押。 通过提供“有担保的股票贷款”和“净资产值(NAV)支持的融资”，银行允许投资者借钱买入特殊目的载体(SP.V)。 由于 TGL 项目下跌且投资缺乏流动性，这些贷款的抵押品可能变得一文不值。借款人面临流动性约束，可能会违约。 该银行还为数字媒体行业和小众房地产建设领域的客户融资“战略附加收购”。 该银行在建设项目“缺口融资”方面的参与可能同样具有毒性。随着 2024-2025 年建筑行业面临困难，MCB&T 的“创造性”缺口贷款可能在停滞或止赎的项目上遭受全部损失。 片段显示 MCB&T 持有 4300 万美元的联邦住房贷款银行(FHLB)预付款，而资产总额为 2.61 亿美元（约 16.5% 的资产负债表由 FHLB 借款资助）。 一系列糟糕的投资/贷款可能导致银行倒闭。 _ 评论： 你可能会看到关于银行倒闭和白银交易的头条新闻。 是的，社区银行被设计为保守机构，但这特定银行承担了高风险的风投债务模式。 这不是银行挤兑，只是一个运作糟糕的杠杆投资基金。 但这里要注意： $MCB - Metropolitan Bank Holding Corp 不是 Metropolitan Capital Bank。 超额收益在于，如果出现算法抛售或做空者混淆两者，这将是一个良好的买入机会。 $KRE - 如果板块因“银行倒闭”头条新闻大幅抛售，看起来是买入机会。 该银行在社区银行资产负债表上运行“赌场”模式，但这是孤立事件。 关注小型伊利诺伊州/中西部银行如 CRE Loans 的商业房地产敞口或无保险存款（>50%）可能受到影响。 然而，这看起来是一家极小的地区性银行且为孤立事件，因此市场机会不多。
英文原文
The first US Bank of the year has just collapsed: Metropolitan Capital Bank & Trust has been shut down and placed into receivership by the FDIC. Here's what likely happened and how markets might react: The Chicago bank was a community bank by asset size ($261M), but operated with the risk of a boutique investment bank. They operated in three segments: - Tech-Infused Sports - Digital Media - Niche Real Estate Construction From public information, we're able to see high concentration of risk in these two volatile sectors. Namely, TMRW Sports, the venture founded by Tiger Woods. The bank didn't just facilitate the equity; it lent against it. By providing "secured stock loans" and "NAV-backed financing", the bank allowed investors to borrow money to buy into the SPV. Since the TGL venture dropped and the investment was illiquid, the collateral for these loans likely became worthless. The borrowers, facing liquidity constraints, would default on the loans. The bank also financed "strategic add-on acquisitions" for clients in the digital media industry and Niche Real Estate Construction. The bank’s involvement in "gap financing" for construction projects was likely equally toxic. MCB&T’s "creative" gap loans likely suffered total losses on projects that stalled or were foreclosed as construction sectors in 2024-2025 faced difficulty. Snippet reveals that MCB&T held $43 million in FHLB advances against $261 million in assets (~16.5% of the balance sheet was funded by FHLB borrowing). A series of terrible investments/loans likely caused the bank to collapse. _ Commentary: You'll probably see headlines regarding bank collapses and silver trades. Yes, community banks are designed to be conservative institutions but they this bank in specific took on a high-risk venture debt models. This wasn't a bank run, just a terribly run leveraged investment fund. But here's what to look out for: $MCB - Metropolitan Bank Holding Corp is not Metropolitan Capital Bank. The alpha is that if there's an algorithmic sell-off or short sellers mistaking the two, this would be a good buying opportunity. $KRE - Looks like a buy if the sector sells off largely on "bank failure" headlines. The bank ran a "casino" model on a community bank balance sheet, but this is isolated. Look for commercial real estate exposure from small-cap Illinois/Midwest banks like CRE Loans or uninsured deposits ( > 50%) that might be affected. However, this looks to be an extremely small regional bank and isolated incident so not much market opportunity here.
-
媒体掩盖交易所规则导致白银崩盘真相
白银崩盘的报道是我见过最不诚实的。从《福布斯》到《华尔街日报》等主流媒体，都将白银28%以上的暴跌归咎于新任美联储主席Kevin Warsh。但他们忽略了真正的原因：- 芝商所(CME)白银维持保证金上调5次以上，并改为百分比基准；- 金银商银行（如摩根大通、多伦多道明银行等）面临无限亏损；- 交易所因溢价高于上海黄金交易所(SHFE)合约而暂停交易。媒体系统性地掩盖了这一事实：白银是被交易所规则砸盘的。当前的媒体报道正在掩盖那些为避免机构从未持有的“纸白银”无限亏损而设立的“不公平规则”。
英文原文
The Silver crash has the most dishonest reporting I've ever seen. From mainstream media like Forbes to WSJ. Everyone pins Silver's 28%+ crash on the new Fed chair Kevin Warsh. But they fail to report the true cause: - CME maintenance margin on Silver hiking 5+ times -> changing to percentage base - Bullion banks (JPMorgan, TD, etc) were facing infinite losses - Exchange halts due to premiums over SHFE contracts. The media is systematically burying this fact: Silver was crashed by exchange rules. Current media reporting is covering up the "unfair rules" that were put in place to avoid infinite losses from the "Paper Trade" Silver the institutions never had.
-
CME保证金新规导致白银合约抵押金飙升，迫使交易员抛售并引发螺旋下跌。
白银故事的残酷教训是：交易所制定所有规则。芝加哥商品交易所(CME)将保证金改为基于百分比的9%（随后加息至11%或12%以上）。在旧制度下，5000美元的保证金保持不变。在新制度下，当白银今天触及120美元时，一份合约所需的抵押品会不断增加，飙升至超过54000美元。交易员基本上被迫在上涨过程中因要求变化而卖出。而在下跌过程中，局势则陷入螺旋式恶化。
英文原文
Brutal moral of the story with Silver is: Exchanges set all the rules. CME moved to a 9% percentage-based margin (then 11% or 12%+ with hikes). In the old system, a $5,000 margin stayed $5,000. Under the new system, when silver hit $120 today, the required collateral for one contract would keep increasing and ballooned to over $54,000. Traders were basically forced to sell on the way up as requirements changed. And on the way down, it spiraled.
-
CME保证金调整与中国流动性陷阱引发$SLV杠杆清算暴跌。
白银今日收盘下跌超过28.54%。 以下是白银/$SLV 今日暴跌的原因： 1月13日：芝加哥商品交易所(CME)将保证金从固定美元制改为基于百分比的制度。 这使得抵押品要求随合约价值缩放，实际上限制了价格上涨时的杠杆率。 维持单个纽约商品交易所(COMEX)合约所需的资本随之增加，导致即使微小的价格下跌也会触发巨额追加保证金通知。 1月27日：CME本周两次提高维持保证金比例，以确保在极端波动中拥有“充足的抵押品覆盖”。 这迫使杠杆头寸平仓多头或追加大量资金。 九天内五次上调保证金，形成了潜在的“弹簧式”抛压。 今日：西方市场关注美联储，但新任美联储主席的影响可能有限，这不过是噪音。 定价错配发生在亚洲市场。瑞银国银白银期货基金的交易价格较上海期货交易所(SHFE)合约溢价36-64%。这是中国主要的白银敞口来源。 1月30日，深圳证券交易所对国银白银上市开放式基金(LOF)实施紧急全天停牌。 这一停牌为中国机构和散户投资者制造了“流动性陷阱”。由于无法清算国内持仓，这些参与者被迫抛售$SLV和COMEX期货以筹集现金或对冲敞口。 总结： 2026年1月30日$SLV的暴跌并非白银基本面价值的失败，而是决定价格发现的“纸面游戏”的失败。 CME反复提高保证金要求以及中国的流动性陷阱导致了级联式的保证金清算，引发了杠杆头寸的抛售。 美联储主席等其他事件可能早已知晓，看起来只是关于实际发生情况的“叙事噪音”。 今天是“纸面游戏”的失败，用于交易的杠杆被交易所规则系统性地清除。
英文原文
Silver ended the day down over 28.54% Here's why Silver / $SLV crashed today: Jan 13th: CME shifted from fixed-dollar margin to percentage based margin. This scaled collat requirements with contract value, effectively capping leverage as it goes higher. The capital required to maintain a single COMEX contract increased in tandem, creating an environment where even minor price drops would trigger massive margin calls. Jan 27th: CME had increased the maintenance margin percentage twice this week to ensure "adequate collateral coverage" amid extreme volatility. This forced leveraged positions to liquidate their long positions or post substantial additional capital. There were five margin hikes within nine days that created a "coiled spring" of potential selling pressure. Today: Western markets focused on the Federal Reserve, but the new Fed chair likely did not play much of an impact as this is just noise. Pricing dislocations happened in Asian markets. UBS SDIC Silver Futures Fund traded at 36-64% premiums over SHFE contracts. And this was the main source of silver exposure in China. On January 30, the Shenzhen Stock Exchange implemented an emergency full-day trading halt for the SDIC Silver LOF. This suspension created a "liquidity trap" for Chinese institutional and retail traders. Unable to liquidate their domestic holdings, these participants were forced to dump $SLV and COMEX futures to raise cash or hedge their exposure. TLDR: The $SLV crash of January 30, 2026 today was not a failure of silver's fundamental value, but a failure of the "paper game" that dictated price discovery. CME hiking margin requirements repeatedly and the China liquidity trap led to cascading margin liquidations that caused selloffs of leveraged positions. Other events such as the Fed chair was likely known awhile, looks to be "narrative noise" regarding what actually happened. Today was a "Paper Game" failure and leverage used to trade it was systematically wiped out by exchange rules.
-
光子学建设规模惊人，上游前驱体材料或成瓶颈。
@PhotonCap 绝对没错，当你将光子学(CPO)的建设与$ LITE等公司的预期增长进行映射时，人们就能直观地看到其规模有多么惊人。 但他们忽略了使这一切成为可能的上游前驱体材料，这也是我预计它将成为主要瓶颈的原因。
英文原文
@PhotonCap Absolutely, when you map the photonics buildout and projected growth with $LITE and others, people can just see how staggering it is. But they miss upstream precursors that make it all happen, which is why I’m expecting it to be a major bottleneck.
-
7N铟价创新高，AI需求致上游供应瓶颈显现，涨价在即。
7N铟价格在SMM上刚刚创下历史新高，且毫无放缓迹象。 $AXTI 的瓶颈正在实时上演。 西方超大规模云服务商似乎正在吸收用于AI建设的几乎所有磷化铟(InP)衬底和前驱体供应。 就像 $SNDK 和 $MU 一样，对于像AXT这样的上游供应商来说，这只是一场等待涨价（类似NAND）的博弈。 供应冲击似乎不可避免。
英文原文
7N Indium prices have just hit ATHs on SMM with zero signs of slowing down. The $AXTI bottleneck is playing out real time. Western hyperscalers appear to be soaking up every bit of the InP substrates + precursors supply for the AI buildout. Like $SNDK and $MU, it's just a waiting game for the upstream suppliers like AXT to increase prices (like NAND). The supply shock feels inevitable.
-
SanDisk聚焦AI致传统存储真空，群联与美光科有望填补。
我持有的、在 $SNDK 财报后有趣的受益者： 群联电子 (TPE: 8299) 和 美光科 (TPE: 2337)。 类似于南亚科 (1年涨幅 1,035.96%) 填补了 $MU、SK 海力士和三星留下的传统 DRAM 空白： 群联电子在闪存 (库存) 方面对 SanDisk 做同样的事。美光科则通过 MLC NAND (拥有晶圆厂) 填补空白。 所以这是另一个“传统空白”策略。 随着 $SNDK 将其 NAND 晶圆转向企业级 SSD 以聚焦 AI，群联/美光科看起来要填补这一真空。 特别是当 SK 海力士和三星停止生产 MLC NAND 以聚焦 HBM 时，这两家公司将成为传统存储领域的“最后幸存者”。 群联电子恰好拥有 2024/2025 年大量低成本 NAND 库存。 美光科正在增加 MLC NAND 产量以填补三星/SanDisk 转型留下的空白。 所以对于 SanDisk 留下的“南亚科 2.0”： 群联 = 巨大库存的价格套利 美光科 = MLC NAND
英文原文
Interesting beneficiaries post $SNDK earnings that I own: Phison (TPE: 8299) and Macronix (TPE: 2337). Similar to Nayna (1,035.96% 1Y) that filled legacy dram void from $MU, Sk Hynix, and Samsung: Phison does the same with Sandisk with flash (inventory). And Macronix with MLC NAND (owns wafer fabs). So another "legacy void" play. As $SNDK pivots their NAND wafers toward enterprise SSDs to focus on AI, Phison/Macronix looks to fill the vacuum. Esp. when Sk Hynix and Samsung stop making MLC NAND to focus on HBM, the "Last Man Standing" for legacy storage would be those two. Phison happens to be sitting on a massive stockpile of lower-cost NAND from 2024/2025. And Macronix are upping their MLC NAND production to fill the void from the Samsung/SanDisk pivot. So for Nanya v2 left by Sandisk: Phison = Massive stockpile price abritrage Macronix = MLC NAND
-
作者认为当前属低风险博弈，预期铜-金刚石或AlSiC混合方案，但可能过早。
@PhotonCap 没问题！这对我来说只是那种风险较低、但一旦奏效就大赚的博弈类型之一。 我原本预期会看到一些铜-金刚石或铝硅碳化物(AlSiC)的混合方案，但也许我有点太早了。
英文原文
@PhotonCap No problem! This is just one of those lower risk but if it works it works type plays for me. I would expect to see some copper-diamond or AlSiC mix but maybe I’m a bit early
-
市场需计入风险，某公司疑似有Figure等人形机器人客户。
@ShortsHoward 这就是市场需要计入的风险。没有任何保证。不过他们确实有多个人形机器人客户，另一个据推测是 Figure。不过我也不完全确定。
英文原文
@ShortsHoward That’s the risk markets need to price in. No assurance there They do have multiple humanoid customers though, the other is assumed to be Figure. Not completely sure though
-
马斯克宣布转产Optimus，$VPG作为核心供应链标的，潜在营收规模巨大。
马斯克在$TSLA财报电话会议中表示：“我们将用一条年产能高达100万台Optimus机器人的生产线，取代弗里蒙特工厂的Model S和Model X生产线。” 针对$VPG（市值5.89亿美元）做一些粗略估算： 100万台 * 850美元中位数（CEO在财报中引用的每台机器人500-1200美元区间）= 短期年营收8.5亿美元。 如果按马斯克1000万台的目标推算，年营收可达85亿美元…… 毛利率在50%到55%之间。 对于这样一家小公司来说，数据相当有趣。
英文原文
Elon Musk from $TSLA earnings: "We are replacing the Model S and Model X production line in Fremont with a line capable of producing up to 1 million Optimus units per year" Just some napkin math on $VPG (a $589M company): 1M * $850 midpoint ($500-$1.2k per robot quoted by CEO from ER) = $850m revenue/year near term. If you scale that to Elon's 10m target, $8.5B revenue/year... Off 50% to 55% gross margins. Pretty interesting figures for such a small company.
-
推荐南亚科、华邦等亚洲存储股作为AI存储巨头替代标的。
其他一些亚洲存储多头思路，如南亚科（如果你不想投 $MU、三星、SK 海力士这类巨头）： - 华邦电子（DRAM/Flash）1年涨幅847% - Macronix（ROM/NOR Flash）1年涨幅389% - 群联电子（Kioxia $SNDK 的SSD控制器）1年涨幅383% 其他有趣的还有东京电子，作为NAND/DRAM设备商，是间接受益者。
英文原文
Some other Asia memory long ideas like Nanya (if you don’t want $MU, Samsung, SK Hynix types) - Winbond (dram/flash) up 847% 1Y - Macronix (ROM/ NOR flash) up 389% 1Y - Phison Electronics (SSD controllers for Kioxia $SNDK) up 383% 1Y Some other interesting ones are Tokyo Electron for NAND/DRAM equipment for indirect beneficiaries