方法论
研究框架、选股思路、思维方式 · 共 651 条 · 滚到底自动加载更多
-
不要只看 meme 标签
就在一个月前，记者和媒体还在淡化我对 $SOI 和 $RPI 的 thesis。 他们引用我的 WSB 标签，把它说成是“注定要崩的 meme 股”，却没有分析底层 thesis。 结果这两只都涨了 100%，而且把涨幅守住了。 $SIVE 现在也是同样的情况。 我从来不靠权威说话……但既然他们想走这条路： 我就很好奇……他们的叙事里为什么总把我在 Nature 之类的地方发表过、被引用上千次的基础 AI 论文背景漏掉？ 又或者为什么今天 X 上那些看多 $SIVE 的分析师里，有些人是 UC Berkeley 的 Ph.D.，而且本身就在光子学领域发文？ 反而总是英语系或非技术背景的毕业生跑去媒体上，对 CPO 超大规模云厂商认证周期和供应链映射发表最强硬的意见。 我总觉得机构最不喜欢的，就是散户投资者在 X 上真的懂自己在说什么。 一个 thesis 应该靠本身的价值生死，而不是发言人是什么权威。
英文原文
Just 1 month ago, journalists and media tried downplaying my $SOI and $RPI thesis. As “meme stocks that were set to crash” citing my WSB tag without analyzing the underlying thesis. They’re both up 100% and held their gains. Same is happening to companies like $SIVE. I never argue from authority… but since they’re going that route: I’m curious why… in their narratives they just leave out the fact I published fundamental AI papers in places like Nature with thousands of citations? Other the fact other analysts on X positive about $SIVE today have Ph.Ds from places like UC Berkeley and publish in the photonics space? It’s always the English or non-technical graduates that go out in the media and have the strongest opinion about CPO hyperscaler qualification cycles and supply chain mapping. It just feels like institutions hate it when a retail investors on X know what they’re talking about. A thesis should live and die based on merit, not the authority of who is comes from.
-
AEHR 被重新定价
我说过了，$AEHR 正在起飞。 当你把市值做到一定程度，基本上就是机构在主导价格了。 如果你找到一家很多超大规模云厂商都在用的公司……它被重估几乎是注定的？
英文原文
Told ya. $AEHR is cooking. Once you reach a certain point with MC it’s mostly institutions driving price. If you find a name that a ton of hyperscalers use… it’s feels bound to get repriced? https://t.co/AHqpgbqHJK
-
批评仅看图表的技术分析谬误，强调基本面比K线更重要。
谢谢，我只是厌倦了人们在图表上画线，或者说'已经涨了200%，所以可能不会再涨了'。这可能是投资领域最大的谬误之一，尤其是欧洲当地人可能很难理解这一点。基本面相、market expansion、revenue acceleration、客户，比图表上的数字重要得多。
英文原文
Thanks, I'm just so tired of people drawing lines on a chart or saying "it's gone up 200%, so it probably can't go up more". This is probably one of the greatest fallacies in investing, especially one European locals might have a hard time understanding. Fundamentals, market expansion, revenue acceleration, customers, are much more important than numbers on a chart.
-
光子学是可扩张性的解法
@dontbuytops 光子学就是解决可扩展性的办法。 如果我们说的是像 EML / DFB 这种光子的大规模生产，那确实有瓶颈。 这也是我看多 $SIVE 的原因之一——他们从一开始就不会自己烧太多 capex 去处理这些问题。
英文原文
@dontbuytops Photonics is the solution to scalability. If we’re talking about mass production of photonics like EML/DFB stuff, there’s bottlenecks for sure. Which is why I’m long the entire supply chain from substrates to foundries to osats .
-
两周内多数持仓股大涨，长期逻辑不变，短期入场点决定10-20%收益差异。
哇，大部分这30只我喜欢的股票两周内都涨了很多（给新来的朋友们做个回顾）顺便说一下，我对它们任何一只的长期看法都没有改变，从$MRVL、$AMD、$ARM到其他的。短期入场点位是会变的，比如从$AAOI到$AEHR，而这些点位差异会产生+10-20%的不同。我在做新布局时，分析中会花很多精力在"未被发掘"的标的上，比如Riber或$SIVE或$RPI或$IQE->等待它演绎出来。但去年关于$LITE或$NBIS或$AXTI的论点还是一样的。我不需要反复发布相同的论点，因为它已经不是新东西了。但它们之所以不再是新的，是因为市场已经验证了这些论点，正在因此对股票进行实时重新定价。
英文原文
Wow, majority of these 30 stocks I’ve liked are up a lot in just two weeks (just a recap to new folks) By the way, my long term opinion doesn’t change on any of them from $MRVL, $AMD, $ARM and others. Short term entry points do though with names like $AAOI to $AEHR. And they make the difference between +10-20%. I focus a lot about the “undiscovered” ones like Riber or $SIVE or $RPI or $IQE in analysis when I make a new entry -> wait for it to play out. But the same thesis around $LITE or $NBIS or $AXTI from last year is still the same. And I don’t need to post that same thesis multiple times, since it’s not new anymore. But the reason they’re not new is because markets have validated the thesis and are repricing the stocks live because of them.
-
上涨时更应坚持
@amandaong23 对，我觉得美国前一阵刚经历过一次小崩盘。 我一直都认为，如果一只股票在那种调整期还能上涨，那它在后面正面宏观环境里大概率会涨得更多。
英文原文
@amandaong23 We literally just had a mini crash two weeks ago. I’ve always maintained that if a stock is up during that correction period, it will probably be up a ton more during a friendlier macro climate.
-
自建模型和数据管线
@PacificPal00 我最后还是自己搭了自定义模型和数据管线。 Frontier 模型半数时间都很烂，爱胡编，除非你做的是更基础的事。
英文原文
@PacificPal00 I ended up just building my own custom model and data pipelines. The frontier models suck and hallucinate half the time unless you’re doing something more basic
-
这是基于已确认协议的映射
@crux_capital_ 这就是基于已确认协议做的供应链映射。 $GOOGL 不会在新闻稿里突然跑出来说它的上游基板供应商是 $AXTI。 我在上面的说法里会用“很可能”这样的词，哪怕对第一轮量产周期来说几乎已经是板上钉钉。
英文原文
@crux_capital_ It’s supply chain mapping off confirmed agreements. $GOOGL won’t go out and randomly say in a press releases their upstream substrate provider is $AXTI. I use words like likely in my statements above even though it’s near certain for the first production cycles
-
认证周期不能短线拆解
@crux_capital_ 如果你觉得能在很短时间内把 $POET 从 Celestial 2028 的量产爬坡里设计掉，那你就是没理解认证周期。 $SIVE 和 $POET 也是一样，但我预计这些关系大多会随着时间变成多来源结构。
英文原文
@crux_capital_ That’s misunderstanding qualification cycles if you think they can design out $POET in a short timeframe for celestial 2028 volume ramp. Same with $SIVE and $POET but I expect most relationships to be multi source over time
-
就是供应链映射
@crux_capital_ 这是供应链映射。 https://t.co/loBcX9iRNp
英文原文
@crux_capital_ It’s supply chain mapping https://t.co/loBcX9iRNp
-
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
-
批评加密圈忘了区块链原本的用途。
这感觉就像……对冲基金 / VC 突然忘了区块链本来是干什么的？ 他们居然把数字资产变成了： 代币融资 -> 散户退出流动性，循环往复。 现在，没人再在乎“治理币”或 gas 代币了，因为那只是印钞后的更大傻瓜理论。 区块链最初的意义，是尽可能减少中间商，也顺带把费用降下来。 尽管如此： 1. 3 天 ACH 通道还在……虽然我们已经有 0 成本的支付通道。 2. 巨额 interchange 还在……虽然我们已经有 0 成本的支付通道。 3. 数字资产“信用卡”只是给 Visa/Mastercard 又包了一层更磨损的外壳。 4. 储蓄账户利率还是 0.3%，而现在法律还在试图堵住和稳定币相关的收益。 数字资产原本想做到的事，除了稳定币之外，几乎都没实现。 也许你们只是盲目追下一家 $HOOD / Hyperliquid（代币化、链上股票），或者盲目跟着 Ethereum，所以表现才会这么差。 重点不是靠代币价格上涨赚钱，再把 meme 币的癫狂变现。 而是要去资助真正有颠覆性的项目，比如 Bridge（稳定币出金/入金、interchange）之类。 这个圈子里绝大多数人毫无 alpha，因为他们都忘了它最初到底是为了什么。
英文原文
It's like... Hedge Funds/VCs suddenly forgot what blockchain was ever supposed to do? They somehow turned digital assets into: Token raise -> retail exit liquidity on repeat. Now, nobody cares anymore about "governance or gas" tokens since it's greater fools theory from printed money. The original point of blockchain is removing as many middlemen as possible, and as a byproduct, the fees. Despite all of this: 1. 3 Day ACH rails are still around... Even though we have payment rails for 0 cents. 2. Massive interchange is still around... despite having payment rails at 0 cents... 3. Digital Asset "Credit Cards" are just more friction-add wrappers around Visa/Mastercard. 4. Yields are still .3% in checking accounts and now laws are trying to bank stablecoin related yields. Very little that digital assets set out to do actually got accomplished, aside from Stablecoins? Maybe you all have bad performance by just blindly following the next L2 exchange token or blindly into Ethereum. It's not about making money from token prices going up + monetizing degeneracy from meme coins. It's to be funding any legitimate disruptors, like the next $HOOD/Hyperliquid (Tokenization, on-chain equities), Bridge (stablecoin off-ramps/intercharge), etc. 0 alpha from the vast majority of folks in that space, because everyone there lost vision on what it was originally supposed to do.
-
说这些名字大概率还会继续冲，自己是在抢跑机构。
@Lee_Trades 我提到的这些名字大概率还会继续往上冲。 三位数回报看起来很多……但那不是因为我个人在追 10 倍。 这更像是在一个超级周期最早期抢在机构前面。
英文原文
@Lee_Trades All of the names I’ve mentioned are likely to keep on running. The triple digit return looks like a lot… but it’s not cause I’m personally chasing 10x. This is what it feels like to frontrun institutions at the very beginning of a supercycle
-
建议写 16 个名字的 thesis 并高集中做多。
@UntappedAlpha 给 16 个单独名字各写一篇 thesis 帖。 高集中度做多。 4 个月里让它们都涨 100-1000%+？其实很简单，说真的。
英文原文
@UntappedAlpha Write a thesis post on 16 individual names. Go long on them with high concentration. The have them all return 100-1000%+ in 4 months? Pretty simple tbh
-
股票市场是正和游戏，可从AI瓶颈中获利而非投资传统公司
随时哈哈..我不认为大多数人已经意识到股票是正和(Positive Sum)游戏，每个人都能从中受益。我也不认为他们意识到你可以从AI瓶颈中获利，而不是做多像$PYPL或$CRM这样的传统公司。
英文原文
@pennycheck Anytime lol.. I don't think majority of people realize yet stocks are positive sum, where everyone benefits. Also don't think they realize you can make profit off AI bottlenecks, instead of going long on traditional companies like $PYPL or $CRM.
-
区分自己与 Roaring Kitty 的做法。
当然尊重 Roaring Kitty，但他买 meme 股是为了逼空机构。 而我做的是指出那些基本面扎实、无论我是否发声都会继续上涨的公司。 而且我会先把它们分享给散户。
英文原文
All respect for Roaring Kitty but he buys memestocks to short squeeze institutions. What I do is point out fundamentally sound companies that would keep going up with or without me. And I point them out to retail first. FOR FREE. Instead of selling to institutions for $20k/article so they can frontrun retail investors.
-
作者解释其研究方法论：提前发现被低估的小盘股，在机构买入前将信息分享给散户
不。这种扯淡的思维模式该停止了。 我的做法是在这些股票实现100-500%+回报之前，先向散户指出它们。 美国机构如Point72或Apollo最终会把它们买下来的。 1. $IQE上涨是因为它们坐拥全球最大的InP反应器隐性代工产能，当时市值仅1亿欧元，而Landmark等公司交易在38亿美元。 它们还是$LITE的供应商，而光子学/epiwafer需求今年爆发了。 2. $SIVE上涨是因为它们与$JBL和O-Net签订了新协议。 但当我首次建仓时，它们还鲜为人知——作为$MRVL CPO项目的激光供应商。 如果我没有引起关注，像$AVGO这样的美国机构很可能会像高通收购Alphawave那样，直接收购这家公司——在openlight那边发生的事情。 那样的话瑞典散户投资者就得不到任何上涨收益了。 3. $ALRIB上涨是因为它们的财报将市盈率压低至远期26倍，尽管在MBE领域与$VECO保持双头垄断地位。 加上新的SiPH设备，以及$IQE和QD Laser（量子点）都是它们的客户。 这是我十年来最原始的信息发现——原来$MSFT Quantum是它们的买家。 你不会看到直接的超大规模云厂商前沿量子计算项目依赖一家市值不到10亿美元的公司。 4. $SOI上涨208%是因为它在硅光子学和CPO用的SOI基材上拥有不为人知的垄断地位。 这更多是信息整合加上在它们传统业务周期底部买入的时机把握。 5. $RPI上涨是因为财报和AI硬件使用量。 我只是第一个指出它的人。 我预测营收增长55%，而分析师预测14%。它们实际做到了58%。 我只是给了散户在机构有机会发现并定价之前买入的机会。 即使我不发我的分析报告，这只股票也会基于纯基本面上涨——作为一家无晶圆厂公司，你不可能用5亿美元市值做出5.11亿美元营收。 我只是给散户在机构之前获得信息发现的所有机会。 这与那些机构告诉你买指数基金或已经涨了1500%的股票——让你当退出流动性的完全不同的模式。
英文原文
No. This type of BS mindset needs to stop. What I do is point them out to retail first before the 100-500%+ returns. US institutions like Point72 or Apollo would have bought them out eventually. 1. $IQE went up because they're sitting on the most latent merchant capacity in the world for InP reactors back at a 100M euro marketcap. While companies like Landmark were trading at $3.8B. They were also the supplier to $LITE, and photonics/epiwafer demand took off this year. 2. $SIVE went up because they had new deals with $JBL and O-Net. But they were already unknown as the laser supplier to $MRVL's CPO program when I first went long. American institutions like $AVGO would have likely just bought the company directly like what Qualcomm did with Alphawave over in the openlight side of things if I didn't bring attention to it. Then Swedish retail investors wouldn't get any of the upside. 3. $ALRIB went up because their earnings sent their P/E down to fwd 26, despite holding a duopoly in the MBE category with $VECO. This combined with new SiPH equipment, as well as $IQE + QD Laser (for quantum dot) being their customers. This was combined from raw information discovery of the decade that $MSFT Quantum was their buyer. You don't see direct hyperscaler frontier programs in quantum computing dependant on some <$1B MC company. 4. $SOI is up 208% because it has an unknown monopoly over SOI substrates for silicon photonics and CPO. This was more information synthesis combined with timing the bottom of their legacy cycle. 5. $RPI went up because of earnings and AI hardware usage. I was just the very first person to point it out. I projected 55% revenue growth compared to 14% from analysts. They did 58%. I just gave retail the chance to buy it before institutions. The stock would have gone up off of pure fundamentals without me posting my thesis because you don't do $511m in revenue off a $500m MC as a fabless company. I'm just giving retail the all the information discovery before institutions have a chance to find it and price it in. This is a completely different model than the same institutions telling you to buy index funds or stocks that already went up 1500% so you're exit liquidity.
-
说明公司增长对整个经济是正和的。
但这对他们整个经济来说是正和效应啊，而且会把整个本地人口都带起来？ 如果公司拿到更多资金，他们就能雇更多本地人做研发，capex 还会继续向上游供应商传导（而这些供应商也会受益于融资）。 然后如果利润更多，交的税也更多…… 那就会有更多钱流进社会福利项目或者基础设施。 我真的不理解，公司增长本来就是正和的，它们不是在吸当地人的血来建立财富。 这又不是 $SNAP。
英文原文
It’s a positive sum effect for their entire economy though and brings up the whole local population together? If companies have more funding, they can hire more locals for R&D, the capex goes upstream to all their suppliers (who also benefit from funding). Then more taxes are paid if they make more profits… So more money goes into social welfare programs or infrastructure. I genuinely don’t understand, company growth is positive sum, they’re not leeching off the population to build wealth. This isn’t $SNAP.
-
抱怨别人限制自己分享免费 DD。
@thetakill 哥们，我就不能分享一点免费的思路型 DD，而不被人告诉我应该发什么吗。
英文原文
@thetakill Bro can I just share a free thought process DD without someone telling me what I should post
-
说对高税的办法就是赚得更多。
@miraclemaster07 对，税确实高。不过我的理念就是把短期资本利得税赚回来。 如果你赚了 20 倍，又被征 50% 的税，你最后还是能赚到原始投入的 10 倍。 如果你一直拿着同一笔多头，而它只涨 20%，那反而更差。
英文原文
@miraclemaster07 Yeah unfortunately taxes are high. But my philosophy is just outearn short term capital gains tax. If you make 20 times and get taxed 50% you still make 10 times your original investment. If you stay in the same long and it just goes up 20% that's worse.
-
说自己不搞高价门槛，因为觉得市场里就能赚回来。
@SosaKayser666 我很清楚很多人甚至愿意每月付 30 美元。 但我还是把最低价定在 1 美元，因为我相信我的想法足够好，我完全可以靠市场里的收益赚回来。 而且我不认同把信息 / 标的搞成门槛。
英文原文
@SosaKayser666 I’m well aware a lot of people would even pay $30/month. But I charge $1 as bare minimum since I believe my ideas are good enough that I can make all my returns in the market. And I don’t believe in gatekeeping information/names.
-
说自己是在尽力帮助散户，分享信息发现。
我是真的在尽力帮散户投资者！ 股票是正和游戏，所以我相信最好的帮助方式就是把信息发现 / 综合整理分享出来。 是啊……很多收费给自己选股的人，半数时候根本不知道自己在说什么。
英文原文
I’m genuinely trying my best to help out retail investors! Stocks are positive sum so I do believe the best benefit is just distributing any information discovery/synthesis. Yeah… lot of the people charging money for their picks don’t know what they’re talking about half the time.
-
认可对方关于数据源的纠正。
@iamfabian 谢谢，如果真是这样，这就是一个重要的纠正，我会再深入看看。 Digitimes 通常算是比较靠谱的来源。 https://t.co/yDUu8XFWtF
英文原文
@iamfabian Thanks this is an important correction then if true, will look into it more. Digitimes is typically one of the more reputable sources https://t.co/yDUu8XFWtF
-
说会做一个长期 CPO 受益者 ETF。
@aiden1969 @afasano23 对，我很快会做一个长期 1 年期、只包含纯 CPO 受益者的 ETF，然后发出来。 只是研究真的很耗时间。
英文原文
@aiden1969 @afasano23 Yeah I’ll make a long term 1Y ETF on pure play CPO beneficiaries soon and post it It just takes a ton of time for research
-
认为 2026 下半年是向更纯粹 CPO 名字轮动的好时点。
在那个会远远持续到 2028 年以后的光子超级周期里？不是。 不过这个板块里，确实可能会在一些名字之间轮动，比如从 pluggables 轮到更多 CPO 名字…… 2026 上半年到下半年，可能是转向更多纯玩法名字的不错轮动时点，因为放量会在下半年后段到来。
英文原文
During the optical supercycle that lasts way past 2028+? No. There’s likely going to be rotations in certain names in the sector eg. Pluggables into more CPO names though… 2026 h1 into h2 is probably a decent rotation timing to more pure play names coming up since volume ramp latter part h2
-
说这是一个用小预算扰动 AI 建设的教育演练。
@BrokeBoyCap 这只是一个教育性演练，看看如何用“柠檬水摊预算”去扰动整个 AI buildout。 不是看多建议，也不是什么投资推荐。
英文原文
@BrokeBoyCap This is just an educational exercise on how one might disrupt the entire AI buildout with a lemonade stand budget. Not a long recommendation or anything.
-
说自己只是分享思路。
我只是一路分享我的思考过程而已！ 只是越来越烦那些 2000 多美元的付费课程把大家割得很惨。或者散户被机构和做市商割得很惨。 股票是正和游戏，如果这些 ticker 付费墙卖家真的相信自己的分析……他们其实应该直接在市场里赚钱，而不是试图为此收费，对吧？
英文原文
I’m just sharing my thought process along the way! Just got frustrating seeing everyone shafted by $2K+ paywalled courses. Or retail shafted by institutions and MMs. Stocks are positive sum, and if these ticker paywall sellers actually believed in their analysis… they would just make money in the markets rather than trying to charge for it?
-
说光子和瓶颈的 TAM 预估都被低估了。
@Blinklebloop 光子和各种瓶颈的 TAM 预测都大错特错。 就像分析师说 $AXTI 的 InP 衬底 TAM 只有 1.25 亿美元一样。
英文原文
@Blinklebloop TAM projections are all way off for photonics and chokepoints/bottlenecks. The equivalent of analysts saying $AXTI InP substrate TAM is $125m.
-
说自己帮助散户跑赢机构是目标。
@itsavgsizeok 很高兴听到！帮助散户投资者跑赢机构就是我的目标，哈哈。
英文原文
@itsavgsizeok Glad to hear it! Helping retail investors beat institutions is my goal lol
-
说机构做 DD 往往要几周，散户可以立即抢跑。
@tomhou101 对，他们通常要花几周做尽调，才会买入流通盘的几个百分点。散户则可以立刻抢跑他们。
英文原文
@tomhou101 Yeah they usually take a few weeks of DD to buy a few percent of the float. Retail can frontrun them immediately.