不需要天才,只需押对"瓶颈"——这个框架让我从公开分享至今录得约400%升幅

分享
不需要天才,只需押对"瓶颈"——这个框架让我从公开分享至今录得约400%升幅

No Genius Required — Just Identifying the Right "Bottleneck": The Framework Behind ~400% Returns Since Public Sharing

核心观点 Core Thesis

自公开分享至今录得约400%升幅,说实话,这不是靠预测宏观,也不是靠神级选股。大部分来自两个因素:选对赛道,顺着资金流动走,再加上一点运气。今年真正赚钱的逻辑只有一个框架:哪里是 AI 扩张的物理瓶颈,钱就往哪里堆。

A cumulative return of approximately 400% since public sharing — frankly, this did not come from predicting macroeconomics or extraordinary stock selection. Most of it came from two factors: identifying the right track, and following capital flows. Plus a degree of luck. The single framework that generated real returns this year: wherever the physical bottleneck of AI expansion is, that is where capital will accumulate.


一、为什么押"瓶颈"而不是押"故事" I. Why Bet on Bottlenecks, Not Stories

故事可以被叙事驱动,可以被情绪放大,也可以在一夜之间崩塌。瓶颈不同。瓶颈是物理限制——当需求爆发而供给无法快速响应时,瓶颈资产天然获得定价权。这种定价权不需要市场共识,不需要情绪配合,只需要需求持续存在。

Stories can be driven by narrative, amplified by sentiment, and collapse overnight. Bottlenecks are different. A bottleneck is a physical constraint — when demand explodes and supply cannot respond quickly, bottleneck assets naturally acquire pricing power. This pricing power requires no market consensus, no sentiment support — only sustained demand.

今年最重要的一课不是判断对了未来,而是不断问一个问题:下一个被卡住的地方在哪里?因为资金永远先奖励瓶颈。等所有人都看见的时候,利润已经被吃掉一半了。

The most important lesson this year was not predicting the future correctly — it was continuously asking one question: where is the next point of constraint? Because capital always rewards bottlenecks first. By the time everyone sees it, half the profit has already been captured.


二、四个核心瓶颈赛道 II. Four Core Bottleneck Tracks

记忆体(Memory)— SNDK、SNDK、 SNDK、MU、$SIMO、三星、SK Hynix

当算力爆发,HBM、DRAM、NAND 就不是可选项,而是刚需。算力可以扩容,但记忆体跟不上,系统就卡死。这就是瓶颈溢价。根据 Counterpoint Research,2026年第一季度 DRAM 价格较上季度飙升80%至90%,SK Hynix 已公开表示其 HBM、DRAM 与 NAND 产能基本上已为2026年售罄。Gartner 预测2026年 DRAM 价格将上涨47%。

When computing power explodes, HBM, DRAM and NAND are not optional — they are essential. Computing capacity can be expanded, but if memory cannot keep pace, the system grinds to a halt. This is the bottleneck premium. According to Counterpoint Research, DRAM prices surged 80–90% QoQ in Q1 2026. SK Hynix has publicly stated its HBM, DRAM and NAND capacity is essentially sold out for 2026. Gartner projects DRAM prices to rise 47% in 2026.

光子学与光模块(Photonics)— LITE、LITE、 LITE、COHR、AAOI、AAOI、 AAOI、AXTI

当数据中心进入 GW 级别扩张,算力之间的连接速度成为核心问题。GPU 再强没用,光互连跟不上,延迟就是死穴。2026年超大型云服务提供商合计资本支出预计突破6,000亿美元,光互连需求随之爆炸性增长,这个赛道的瓶颈属性正在快速显现。

As data centers enter gigawatt-scale expansion, the connection speed between computing nodes becomes the core bottleneck. The most powerful GPU is useless if optical interconnects cannot keep pace — latency becomes the fatal constraint. With combined hyperscaler capex projected to exceed $600 billion in 2026, optical interconnect demand is growing explosively.

电力与电网(Power & Grid)— $XLU

算力是用电怪兽。数据中心每扩一轮,电网和公用事业都会被动受益。AI 是数字革命,但底层是能源革命。这个瓶颈的特殊之处在于它不受单一公司风险影响,而是整个基础设施层面的系统性受益。

Computing power consumes enormous amounts of electricity. Every round of data center expansion passively benefits power grids and utilities. AI is a digital revolution, but its foundation is an energy revolution. The special characteristic of this bottleneck is that it is not exposed to single-company risk — it is a systemic beneficiary at the entire infrastructure level.

先进封装(Advanced Packaging)— AMKR、AMKR、 AMKR、ONTO、CAMT、CAMT、 CAMT、KLIC、$FORM

HBM、Chiplet、先进封装,是算力真正落地的最后一公里。很多人盯着 $NVDA,但真正卡产能的是封装环节。每生产1比特 HBM 所消耗的晶圆产能是生产1比特 DDR5 的三倍,先进封装的产能瓶颈直接决定了 AI 算力的实际部署速度。

HBM, Chiplet and advanced packaging are the last mile where computing power actually reaches deployment. Many focus on $NVDA, but the real capacity constraint is at the packaging stage. Producing 1 bit of HBM consumes approximately three times the wafer capacity of producing 1 bit of DDR5 — the advanced packaging bottleneck directly determines the actual deployment speed of AI computing power.


三、今年学到最重要的一课 III. The Most Important Lesson This Year

不要去做"逆向拯救"——赌网络安全反转、赌 SaaS 修复、赌情绪底。市场的钱在哪里?在哪个环节最缺?顺着资金流动走,比聪明更重要。

Do not attempt "contrarian rescue" — betting on cybersecurity reversals, SaaS recovery, or sentiment bottoms. Where is the market's capital? Which part of the chain is most constrained? Following capital flows matters more than being clever.

Bridgewater 在单季度集中加仓近20亿美元,选择的正是 NVDA、NVDA、 NVDA、GOOGL、MSFT、MSFT、 MSFT、AMZN——全部站在 AI 基础设施与应用分发的核心瓶颈位置。机构的配置逻辑与瓶颈框架高度吻合,这不是巧合。

Bridgewater concentrated nearly $2 billion in additions in a single quarter, choosing $NVDA, $GOOGL, $MSFT and $AMZN — all occupying core bottleneck positions in AI infrastructure and application distribution. The alignment between institutional allocation logic and the bottleneck framework is not a coincidence.


四、核心结论 IV. Core Conclusion

真正改变收益曲线的,不是判断对未来,而是不断问一个问题:下一个被卡住的地方在哪里?瓶颈资产不需要市场情绪配合,不需要宏观环境完美,只需要需求持续存在而供给无法快速响应。这个特性,使它们在 AI 扩张周期中具备天然的定价权优势。

What truly changes a return curve is not predicting the future correctly — it is continuously asking one question: where is the next point of constraint? Bottleneck assets do not require market sentiment support or a perfect macro environment. They only require sustained demand that supply cannot quickly address. This characteristic gives them a natural pricing power advantage throughout the AI expansion cycle.

如果你对上述分析感兴趣,想了解这些瓶颈赛道如何具体落地到个别标的的买入逻辑,财富俱乐部的深度研究报告与交易机会报告将提供更完整的分析框架——包括具体建仓价位区间、三个阶段止盈目标、关键催化剂时间节点、风险评估,以及触发重新评估持仓的具体条件。配合我们18套量化交易系统的实时买入卖出信号,在关键转折点提前一步捕捉抄底与逃顶机会,系统性提升每一笔交易的胜率。

订阅会员可在研究报告及交易机会栏目获取完整内容。

If you found this analysis useful and want to understand how these bottleneck tracks translate into specific entry logic for individual positions, the Wealth Club's deep research reports and trading opportunity reports provide a complete analytical framework — including specific entry price ranges, three-stage profit targets, key catalyst timelines, risk assessment, and the precise conditions that would trigger a position reassessment. Combined with the real-time buy and sell signals from our 18 quantitative trading systems, you gain an early edge at every critical turning point — systematically improving your win rate on every trade.

Full content is available to subscribing members in the Research Reports and Trading Opportunities sections.


财富俱乐部 / Wealth Club

祝你投资顺遂,生活丰盛。

May your investments prosper and your life flourish.

阅读更多

恭喜会员!NVDA 170→240,会员收益约41% $170→$240, Members Up ~41%

恭喜会员!NVDA 170→240,会员收益约41% $170→$240, Members Up ~41%

🎉 恭喜我们3月在 $NVDA 约170美元附近完成布局、并一直持有到现在的会员!今天英伟达再次刷新历史新高,盘中最高来到约 240美元。从我们当时约170美元附近的建仓位置计算,到今天最高240美元附近,阶段涨幅约41%! 这一段上涨幅度已经相当可观。更重要的是,这不是股价涨上去之后才回头说“我们早就看好”。当时的建仓时间、交易逻辑、量化信号以及后续持仓过程,全部都有完整记录,会员可以在平台随时回看、核对。 🎉 Congratulations to all our members who built positions in $NVDA around $170 back in March and have continued holding until now! Nvidia has once again reached a new all-time

lock-1 By Wealth Club
CBRS 暴涨9%背后:指标早已出现 Buy 信号,随后 Breakout 确认趋势反转 韩国

CBRS 暴涨9%背后:指标早已出现 Buy 信号,随后 Breakout 确认趋势反转 韩国

CBRS 今天单日大涨约9%,而从图上看,这次反弹并不是毫无预兆。 我们的交易指标在股价仍处于低位震荡时,已经率先出现 Buy 买入信号;随后价格企稳,短线趋势开始转强,并进一步出现 Breakout 突破信号。 从当时大约160美元附近的低位,到现在重新回到183美元附近,这一段反弹已经相当明显。 尤其值得注意的是,前面股价处于下跌趋势时,指标连续给出 Sell 卖出信号,有效避开了从200美元上方一路回落的主要跌幅;而当趋势真正开始扭转之后,Buy信号才重新出现。 这正是我一直使用这套指标的原因: 它不是去猜明天涨还是跌,而是帮助我们在趋势发生变化时,更快识别: 什么时候风险正在增加, 什么时候卖压开始减弱, 什么时候出现真正的买入信号, 以及什么时候突破开始得到确认。 CBRS 今天因为Sam Altman重新确认与OpenAI的合作关系大涨,但消息只是催化剂,真正重要的是价格结构在消息之前已经开始改善。 对于经常做波段、趋势交易,或者不想整天盯盘的投资者来说,这种 Buy / Sell + Breakout 信号最大的价值,就是把复杂的K线变化直接转换成更清晰的交易

By Wealth Club
CBRS 投资备忘录 Investment Memo

CBRS 投资备忘录 Investment Memo

AI推理时代的第二条算力主线:从"GPU替代者"走向高速Inference基础设施平台 The Second Compute Track of the AI Inference Era: From "GPU Replacement" to a High-Speed Inference Infrastructure Platform 一、执行摘要 1. Executive Summary Cerebras Systems是目前公开市场中最值得研究的非传统AI芯片架构公司之一。与NVIDIA、AMD依赖大量GPU通过高速互联组成集群不同,Cerebras从底层重新设计AI计算架构,将接近整块晶圆做成一颗超大型Wafer-Scale Engine(晶圆级处理器),试图直接解决传统GPU在大模型推理过程中最核心的两个问题:Memory Bandwidth(内存带宽)和Latency(延迟)。 Cerebras Systems is

lock-1 By Wealth Club
VST Platforms投资备忘录 Investment Memo

VST Platforms投资备忘录 Investment Memo

Vistra Corp. 是美国最大的独立电力生产及零售电力公司之一,核心资产横跨天然气、核电、煤电、太阳能、电池储能以及电力零售业务。过去市场主要将Vistra视为典型的独立发电商,其估值长期受到电价周期、天然气价格以及资本密集属性限制。但随着美国AI数据中心建设进入第二阶段,Vistra的投资属性正在发生明显变化。 Vistra Corp. is one of the largest independent power producers and retail electricity companies in the United States, with core assets spanning natural gas, nuclear, coal, solar, battery storage and retail electricity. In the past, the market

lock-1 By Wealth Club