2026-01-02 · 个股论点

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。

涉及标的： $AAOI $AIRO $AKAM $ALAB $AMKR $AMZN $ASOZF $ASTS $AVGO $AXTI $BMNR $BULL $CIFR $COHR $CRCL $CRDO $CRWV $DJT $ETH $ETOR $FISV $FLY $GEMI $GLXY $GOOGL $HIMS $HOOD $HUT $INTC $IREN $KRKNF $LITE $META $MRVL $MU $NBIS $NVDA $ONDS $ORCL $OSS $PATH $PLTR $POET $PYPL $QBTS $RDDT $RGTI $RKLB $SMCI $SNAP $SOFI $TSLA $TSM $TSSI $TTD $WMT $WULF $XRP

中文翻译

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.

在 X 上查看原推 ↗