Orbit Post Sitemap
别再盯着“人工智能多牛逼”的新闻傻乐了。你以为这些科技巨头是在搞技术突破，其实他们现在玩的是一场赤裸裸的“圈地运动”。如果说以前互联网的入口是流量，那现在AI的命根子就只有两个字：算力。而在这场决定未来十年霸权归属的烧钱大战中，哪怕是强如Meta和英伟达，兜里的现金也不够烧了。他们正在疯狂地把华尔街的资本、传统能源的基建，甚至是你家下个月要涨的网费，通通绑上这辆失控的战车。 你看新闻了吗？Meta刚宣布跟贝莱德搞了个合资公司，要在得克萨斯州那个鸟不拉屎的埃尔帕索建数据中心。这笔账怎么算的？总投资140亿美元，贝莱德出80%，Meta只占20%。这哪是什么战略合作，这分明是哪怕强如Meta，面对动辄百亿的基建黑洞也开始腿软了。具体的玩法很心机：贝莱德这只掌管十万亿美元的资本巨鳄负责掏钱买地、修房子、搞定电力和冷却系统，成了妥妥的“超级包租公”。 而Meta呢？只拿了20%的股权，却签了个长达20年的租约。说白了，Meta就是用极少的股权成本撬动了巨大的算力资源，把重资产的包袱甩给了贝莱德。 但这种操作的风险在哪儿？那140亿美元里，贝莱德的钱可不是大风刮来的，那是全球养老基金和散户的血汗看似实现芯片量产高光，英伟达GB300的本土化终究造了一场寂寞
美国芯片本土化建设的分段困局与长期产业布局
晶圆量产达成阶段性目标
台积电坐落于美国亚利桑那州的Fab21工厂，近期依靠4纳米制程工艺，完成英伟达GB300人工智能芯片首批晶圆的加工下线工作。从产业进程来看，这次晶圆量产，代表美国本土终于拥有高端AI芯片晶圆批量加工的生产条件，打破了此前本土无先进晶圆量产车间的现状。但从芯片完整生产流程来讲，光刻打磨成型的晶圆仅仅是基础半成品，内部电路结构无法直接适配服务器、算力设备的装配需求，必须经过专业化封装处理，整合线路与外部接口才能成为可用芯片成品。现阶段本土工厂只能完成前端制造工序，后续必不可少的加工步骤无法就地开展，这批全新产出的晶圆只能安排空运，运往其他区域进行二次加工，本土化制造的阶段性成果存在明显短板。
核心封装环节仍存在对外依赖
这批美国本土制造完成的GB300晶圆，运输终点设定在中国台湾地区的台积电厂区，由当地成熟产线负责CoWoS高端封装这一关键工序。结合行业调研信息能够看出，当前美国人工智能芯片整条供应链，其余生产环节都已有落地规划，唯独先进封装生产线处于空白状态，这也是整条产业链无法独立闭环运行的核心症结。我认为，单一工序的缺失，会让前期晶圆制造的本土化价值大打折扣，跨地区物料运输会拉长生产周期，叠加航空物流、仓储中转产生的额外开销，本土建厂压缩供应链成本的初衷很难在短期内实现，分段式的生产模式会持续制约本地算力芯片的出货效率。
台积电加码美国厂区全方位产能建设
为补齐封装环节的产业缺口，台积电敲定了规模庞大的美国扩建方案，计划在亚利桑那州新建两座专注先进封装的生产厂房，同步配套SoIC集成工艺生产线，项目整体规划投资金额达到2650亿美元。等到厂区投产运营之后，本地就可以独立承接CoWoS、SoIC两类主流封装业务，支撑英伟达Blackwell、Rubin新一代旗舰芯片的全套加工需求。除此之外，企业敲定了制程升级时间表，计划2028年以前，将N2两纳米、A16 1.6纳米顶尖制程技术落地美国厂区。在我看来，巨额投资拉长了回本周期，海外建厂还要适配当地用工、环保、土地各类规则，工艺落地的实际进度大概率会和纸面规划存在偏差。
下游终端组装同步推进本土化落地
上游芯片制造与封装产能规划稳步推进的同时，产业链下游的服务器组装环节，各大代工企业也开启本土化落地布局。鸿海已经在得克萨斯州休斯顿完工专属生产基地，厂房专门针对GB300人工智能服务器开展装配作业，纬创则选址达拉斯搭建独立组装流水线。上下游企业分头建厂的模式，本意是拆分产业建设压力，逐步搭建完整的本地硬件配套体系。但上下游建设节奏很难完全同步，上游封装产线投产时间未定，下游组装厂房提前完工，极易出现产能空置、原料供给脱节的问题，资源配置的协调性存在不小隐患。
产业发展走向的双向争议总结
业内对于美国这套分段打造自主芯片产业链的模式，形成了两种截然相反的看法。乐观一方觉得，随着封装新厂、高精制程、终端组装项目逐个落地，数年之后就能打通全流程生产链条，彻底告别跨区域工序协作的限制。质疑观点则提出不同意见，千亿级别的巨额投入、严苛的本土建厂环境、长期积累的封装技术门槛都是客观阻碍，纸面规划不等于落地产能，高昂的投入成本能否达成产业链自主可控的最终目标，未来很长一段时间内都无法下定确定结论。
#韩股重挫8%，长鑫首日登顶A股 $NVDA BTC was quite bearish last night.
First, they pierced through 62.8K, then crawled back to 64K. At first glance, it looked like someone was crazily hunting for goods below.
Looking at Farside, the net inflow from US spot ETFs was only $5.1 million.
Damn, putting 5.1 million in BTC is like an ambulance gas cost.
So don't rush to hype this lower shadow into a bullish uprising.
I'd rather be a short seller to take profits first.
Above is still the 64.2K–64.8K door.
Unable to recover after a counter-pull, 62.7K—63K still gets hit; If last night's low is broken again, the 61K–61.8K will be ready to take in customers.
If the bearish script is going to collapse, it needs to regain 65.3K, and don't take it down.
Then look at 66.5K again, it's still in time.
Tonight there's the Federal Reserve's rate decision, and the most disgusting part is that the statement first slashes one side, then the press conference covers the other.
In this kind of setup, I won't heavily bet on a guess.
Don't treat the 64K area as a free score question.
Wait until it regains 64.8K, or breaks through 62.8K, then make a move.
The market hasn't even opened yet, so don't rush in and donate blood first.📰 US airstrikes Iranian-affiliated armed groups in Iraq! Geopolitical storm hits, BTC $63,951 about to reverse?
Event overview
The U.S. Central Command (CENTCOM) launched military strikes against Iran-backed armed groups within Iraq, citing that these groups pose a direct threat to the United States and Saudi Arabia. Simply put, the U.S. was angered and directly struck at Iran's younger brothers. This issue isn't just about the Middle East—global risk assets will be affected, with the crypto world bearing the brunt.
In-depth analysis
Why is this news important?
Guys, this isn't just a matter of casually firing a couple of shots.
First-layer logic: The Middle East situation has always been a powder keg for global financial markets. The U.S. has directly taken action against Iranian armed groups this time, showing that the Biden administration has chosen to confront security threats head-on, squeezing the space for diplomatic negotiations. If Iran chooses retaliation, crude oil supply near the Strait of Hormuz will be affected. If oil prices soar, inflation expectations will return, and the Fed's rate cut window will have to be pushed back again.
Second layer of logic: For the crypto world, geopolitical conflicts have a two-sided impact. In the short term, market panic will cause some funds to withdraw from risk assets, causing BTC and ETH to be hit hard. But in the medium term, every geopolitical crisis will make more people realize that you need an asset that is not frozen by any government or restricted by banks. BTC was the best example during the 2022 Russia-Ukraine conflict.
Third layer of logic: Right now, BTC is at $63,951, ETH is at $1,919, and both are oscillating in a sensitive range. At times like this, any breaking news would be magnified. Once market sentiment turns to safe-haven currencies, gold and BTC often strengthen in tandem.
Impact on the market
In the short term, when the news of the air raid first emerged, the market's first reaction was most likely panic. BTC may quickly test the $63,951 level, inserting a spike to wash the market. ETH follows in a coordinated move, and the $1,919 level may not withstand the first wave of selling pressure.
But honestly, if Iran's response is merely verbal condemnation without substantial military retaliation, then this negative news is just a paper tiger. BTC may soon rebound, or even break out of its current consolidation range. Because the market will realize that the conflict hasn't escalated, panic has subsided, and previously wait-and-see funds will flow back.
Referring to the time in April 2024 when Iran launched a direct missile attack on Israel, BTC briefly plunged below $60,000, but recovered lost ground in less than a week. The impact of geopolitical conflicts on BTC is increasingly resembling a "V-shaped reversal" scenario.
ETH's recent performance has been much weaker than BTC's, rising only 1.95% in 24 hours. If this wave of risk-off sentiment passes, ETH is very likely to catch up.
Operational Approach
🎯 Influence prediction
- Currency: BTC / ETH
- Direction: Bullish 📈 (medium-term geopolitical conflict drives safe-haven capital inflows)
- Duration: BTC 12 hours / ETH 24 hours
💡 My judgment is that it's a bit overweight. There may be panic pulls in the short term, but if the support near $63,951 holds, it would be an opportunity to get in. If ETH falls below $1,919 and does not break through, you can take a light position. Key signal: Pay attention to the wording of Iran's official responses. If the Foreign Ministry only issued a statement with a few insults without the phrase "will respond with force," it was basically a false alarm and a decisive bullish position.
❓ I think this wave of geopolitical conflict will drive BTC higher. Like and let me see how many people are on the bulls' side
$BTC $ETH #BTC #ETH
#地缘政治
⚠️ This does not constitute investment advice; the forecast is for reference onlyOn the first day of trading at Changxin Technology's CXMT market, the nominal trading volume of CXMT-related prediction markets on Predict Fun once approached 1 million USD
Meanwhile, in terms of cumulative trading volume, Predict Fun has almost a one-sided advantage in the CXMT market, far surpassing Polymarket
CXMT itself is a very typical Asian narrative:
- China's AI and semiconductor industry chain
- One of the most watched IPOs of the year on the A-share market
- The Chinese-speaking community naturally has higher information density and discussion heat
For these types of targets, Asian users tend to have a stronger willingness to participate and are more likely to form community spread
This is similar to the logic of some exchanges establishing TradFi perpetual competitive advantages, choosing to offer regional assets such as A-shares and Hong Kong stocks, relying on differentiated targets to build competitive advantage
Similar opportunities may exist in the forecast market. Mid-tier forecasting markets can focus on sectors where leading platforms have limited coverage and stronger regional user demand, such as China's technology industry, Asian capital markets, regional policy events, popular local companies, and industry chains
These markets may not attract global interest, but for Asian users, they are actually more attractive
CXMT is a validation. In the future, similar products like CXMT, which combine industry narrative, regional attention, and event catalysts, will continue to emerge. As a prediction market platform closer to Asian users, Predict Fun may continue to replicate this trading heat and carve out a differentiated path
#韩股重挫8%, Changxin topped the A-share market on its first day 温馨提示：以下所有观点均是个人投资心得体会，和个人身边真实案例分享，供大家交流讨论，不涉及任何投资建议，请大家别盲目跟风，盈亏自负！成年人要有自己的判断。 7月27日A股科创板迎来历史性一刻，长鑫科技正式挂牌上市，首日暴涨465.82%，总市值突破3.28万亿元，一举登顶A股市值榜首，成为国内DRAM国产化里程碑事件。 和A股场内的狂欢形成极致反差，当晚海外存储赛道全线崩盘：SK海力士美股ADR大跌7.47%，美光科技收盘下跌2.25%，主营NAND闪存的闪迪（Sandisk）暴跌超11%，整个费城半导体指数尾盘大幅走弱，全球资本集体重新定价存储行业长期竞争格局。 很多散户疑惑：长鑫如今全球DRAM份额仅8%，短期产能、高端工艺和三星、海力士仍有差距，为何能直接带崩海外存储巨头股价？ 核心答案只有一个：市场恐慌的从来不是当下的产能竞争，而是长鑫拿到天量募资后，国产DRAM扩产、技术迭代全面提速，持续三十年由美韩三家企业垄断的DRAM定价权、全球供给格局，即将被中国企业彻底改写。 一、先理清基础常识：市场暴跌标的业务划分，避免认知误区 很多普通投资者分不清DRAM与NAND闪存，误以为闪[2026-07-29] 🟡 Watching | BTC rates rebound, ETH leads, but on-chain has not followed up. Seven-day fee trend (small volatility group: BTC/ETH/CRV): BTC rebounded from 0.003% yesterday to 0.0063%, deviating +45% from the 7-day average, signaling a renewed rally of bulls; ETH simultaneously rebounded to 0.0017%, following BTC's rhythm; CRV fell back to 0.005%, just above the 7-day moving average, with both bulls and bears balanced. Large volatility group (H/OP): H remains at 0.005%, below the 7-day average of 0.0127%, slightly cool; OP turned positive from yesterday's -2.46% (warning) to +0.0041%, fully restoring extreme bearish sentiment within 24 hours. Key findings (3 items) 1. BTC rates rebounded but OI slightly declined, and increasing positions was uncertain. Rates rebounded from 0.003% to 0.006% (OKX), with a deviation of +45%, indicating a rapid recovery in bullish sentiment. However, CoinGlass's OI 24-hour was only -0.08%, nearly flat—the price increase was driven by increased long positions rather than new capital inflows. 2. ETH leads the market with a 2% increase, altcoin activity boosts ETH (+2.02%)> BTC (+0.57%), OI +1.86% increased interest, and fees rebounded in sync (OKX 0.00169%). This is closeOn July 29, storage giant $SKHY (SK Hynix) shattered the biggest recent AI negative concern in the market during its Q2 earnings call. Recent market panic: Will the popularization of efficient small-parameter AI models and major companies readjusting data center leasing strategies lead to a cooling of global AI infrastructure investment and weakening demand for computing power/storage? In response, SK Hynix gave a clear statement: the current time is not to cut AI investment, but to improve industry quality, increase efficiency, and accelerate commercialization and monetization! Core Key Logic 1. Hardware demand for lightweight AI models ≠ declining Efficient model iteration and upgrades will not reduce memory or computing infrastructure requirements. Instead, it allows the same hardware to support more users and more AI services, greatly improving device utilization and expanding AI scenario coverage. 2. User demand is exploding. After the implementation of the new generation of high-efficiency AI models, the threshold for mass adoption has lowered, user numbers and usage frequency have continued to surge, and efficiency upgrades have driven overall hardware consumption. 3. Medium- and Long-Term AI Investment Certainty Fully Leveled The company has deeply engaged with leading global clients for medium- and long-term orders, confirming: This year is adjusted to an optimized cycle, and from next year onward, AI infrastructure investment will continue to rise steadily. Market summary: The previous market concerns about "AI retreat, peak computing power, and weak storage demand" were essentially short-term market misjudgments. $The real industry trend: From extensive growth of "mindless hardware stacking" to mature growth characterized by "high efficiency, high utilization, and high monetization." SK Hynix, Micron, and $Western Digital are the core美光掉队，高端存储地位岌岌可危
英伟达下一代AI平台的HBM供货名单里，美光只拿到5%-10%的份额。
这个数字，基本等于陪跑。
全球第三大存储芯片厂商，市值刚跌破万亿美元关口，较历史高点回撤31%，股价单日再跌3%。华尔街给出的态度很直接，谨慎。
问题不是美光做不出新产品，而是做出来了，依然打不过。
HBM，高带宽内存，是当前AI训练芯片最核心的配套存储方案。
英伟达每一代GPU平台都要搭配对应规格的HBM，谁能拿到英伟达的订单份额，谁就吃到这轮AI军备竞赛最肥的一块肉。
目前这块肉的分法很明确，SK海力士吃大头，三星紧跟，美光垫底。
美光最新的HBM4 36GB 12层堆叠产品已经开始出货，适配英伟达Vera Rubin平台。
听起来没掉队，但细看参数和客户导入进度，差距就出来了。
带宽落后，堆叠工艺落后，量产爬坡节奏也慢一拍。
SK海力士的HBM产品早在上一代平台就拿下了英伟达超过50%的供货比例，技术验证周期更短，良率更稳。
三星虽然前两年在HBM上吃过亏，良率一度翻车，但砸了重金追赶，现在也稳住了第二梯队。
美光呢？5%-10%。
这不是“还在追赶”，这是被甩开了身位之后，对方还在加速。
HBM和普通内存不一样。普通DDR5你做出来，插上主板能跑就行，渠道铺货卖就完了。
但HBM是定制化供货，要跟GPU厂商深度联调，从芯片设计阶段就要对接，验证周期长达6-12个月。
这就导致一个残酷现实，落后一代容易追，落后一个验证周期就很难追。
SK海力士和英伟达的协同开发关系已经磨合了好几代产品，默契程度和响应速度不是美光短期能赶上的。
即便美光在实验室里跑出了更好的参数，客户那边的信任窗口也不会轻易打开。
技术能补课，但供应链信任关系补不了课。
HBM打不开局面，传统DRAM和NAND业务又面临周期性压力。
PC和手机终端需求波动加剧，服务器端虽然受AI拉动，但大客户越来越集中在头部几家云厂商手里，议价能力被压缩。
多重因素叠加，美光的盈利能力修复节奏明显慢于市场预期。#韩股重挫8%，长鑫首日登顶A股 $MU 🚨Is the BEAT shakeout over? The real market may just be beginning......
$BEAT
Yesterday.
The comment section is almost full of shouts:
**"BEAT is over. ”
"The main force has fled."
"This time it's really over."
**
But today.
I looked at the market again.
Suddenly realized.
The most panicked moment.
Instead, some people quietly started picking up chips.
⸻
BEAT's biggest feature these days.
Not a drop.
Rather,
I can't fall.
Every time I crashed.
There will always be funds to catch them.
Although prices fluctuated repeatedly,
However, the low point did not continue to move sharply.
Brothers.
This is where the real attention is warranted.
⸻
Now let's look at trading volume.
During the recent crash.
Released a huge amount.
And recently.
Prices gradually stabilized.
Selling pressure began to ease.
Indicate who wants to sell in the early stages.
Almost all of them have already been sold.
Those who truly remain inside the arena.
More importantly, they are waiting for the next direction to choose.
⸻
Of course.
The end of the shakeout does not mean an immediate surge.
The real launch.
There are two other signals to watch.
📈 First, volume surpasses previous highs.
Only the trading volume continued to expand.
The price has broken through a key resistance level.
Only when the bulls are truly gaining momentum.
📈 Second, it cannot be broken by a rebound.
If after the breakthrough,
Pushing back again can still hold one's ground.
This shows that the capital is taking on strong demand.
This is the most common rhythm of the main upward wave.
⸻
So.
I won't shout now:
BEAT will definitely double.
But I won't let a single adjustment happen.
and concluded it was over.
One of the things the market loves to do.
It was at the moment of everyone's most desperate moments.
Suddenly walk in the opposite direction.SanDisk (SNDK) Continuous Decline: Complete Analysis As a pure NAND flash memory stock, SanDisk fell from its June high of $2,350 to a low of $1,448 this round, with a maximum drawdown of 38% and multiple single-day drops of 10%+. This is a deep correction characterized by a sector cycle turning point + valuation bubble + multiple negative resonances, broken down in four layers: 1. Short-term direct crash trigger (single-day drop trigger) 1. Earnings guidance falls short of expectations, NAND price cuts warning. On July 24, after market closed, the company released quarterly operating guidance, clearly warning that the Q4 NAND flash spot price upward trend has peaked and is likely to weaken. Downstream consumer SSD order demand is slowing, and revenue and profit growth guidance fell short of Wall Street consensus expectations, directly triggering concentrated stop-loss losses from long sellers on the market. The company plunged 10.1% in a single day. 2. Investment banks collectively lower their forecasts, bearish on memory cycle turning point. Morgan Stanley released a major research report indicating that the AI storage price hike cycle is nearing its end, with DRAM/NAND contract prices peaking in Q4; Senior analysts lowered SanDisk's target price, and combined with continued short reports from citron and other short institutions, investors have shaken confidence. 3. Korean storage stocks crash linked to the market, sentiment spreading to US stocks. SK Hynix and Samsung Electronics' domestic stock prices plunged consecutively. South Korean regulators tightened semiconductor leveraged ETFs, causing large amounts of leveraged funds to passively close positions. Panic spread to the US storage sector, with SanDisk, a highly elastic asset, falling twice the industry average. 4. Geopolitical risks + technology sector funds as safe havens. The US-Iran conflict continues to escalate, and market risk appetite is broad各位朋友好，昨晚美股存储芯片代表闪迪直接来了个 -14.25% 的深水炸弹，美光科技也没好到哪去，跌了 -8.85%，同样的半导体芯片也是大跌。这哪里是调整，这简直是“跳水比赛”。 我们不禁会想，完了完了，美股龙头都这样了，咱们A股的存储芯片板块今天周三是不是要凉凉了？ 先别急着下定论。作为一名在市场里摸爬滚打多年的散户，我习惯透过现象看本质。咱们把情绪放一放，仔细看看这几张图，答案其实就藏在细节里。本文内容仅个人观点，不构成投资建议！ 一、周二那场“惨烈”的撤退 咱们先把时间拨回周二。 如果你盯着存储芯片的分时图看，你会发现全天几乎没有任何像样的抵抗。早盘冲高回落，午后更是单边下行，最终收了一根光头光脚的大阴线，跌幅超过5%。 这意味着什么？意味着多头在盘中彻底放弃了抵抗，空头主导了战场。 再看资金流向图，这就更有意思了。周二这一天，主力资金净流出高达 -108.36亿。你看那个绿色的柱子，从开盘到收盘，流出曲线就像滑梯一样顺滑。这说明什么？说明这不是散户恐慌抛售，而是机构、大户在坚决地“且战且退”。 但是，请注意一个细节：散户流入却有878亿。这画面太熟悉了：主力在卖，散户在接。这#Korean stocks plunge 8%, Changxin tops A-shares on debut
This drop in the Korean stock market is not just an ordinary correction.
What really makes the market nervous is that the long-standing dominance of Korean semiconductors is facing a strong challenger.
Recently, Korean stocks have sharply declined, with the semiconductor sector becoming the market focus. The South Korean economy heavily depends on technology exports, and $SAMSUNG Samsung and $SKHYNIX SK Hynix have long held key positions in the global memory chip market. However, with Changxin Technology's listing on the A-shares and the ensuing capital chase, the market is beginning to reassess the competitive landscape of the global memory industry.
Why did Korean stocks plunge?
There are three main reasons.
First, the market fears that competition in memory chips is entering a new phase.
In recent years, the global DRAM market has been dominated mainly by Samsung, SK Hynix, and $MU Micron. But Changxin Technology's strong performance on its first day of listing, with a rapid increase in market value, sends an important signal:
China's memory industry is accelerating breakthroughs, gradually moving from chip applications toward core manufacturing processes.
This means that the future global memory market may no longer be a competition among a few giants but will evolve into a more intense multi-party contest.
Second, the AI wave is redefining the importance of memory chips.
Previously, the market focused more on GPUs and computing chips, but with the rapid development of AI servers, cloud computing, and large model applications, the importance of high-performance memory is becoming increasingly prominent.
The AI era requires not only powerful computing capabilities but also faster and larger data read/write capacities.
Therefore, memory chips are transitioning from traditional cyclical industries to becoming a crucial part of AI infrastructure.
Third, the capital market is repricing hard technology companies.
Changxin Technology's enthusiastic capital reception on its first day of listing reflects the market's recognition of the growth potential of the domestic semiconductor supply chain.
Investors are focusing not just on current profits but on future technology upgrades, capacity expansion, and competitiveness in the high-end memory field.
So, what does Changxin topping the A-shares mean?
First, the global memory chip competitive landscape may change.
In the coming years, Samsung, SK Hynix, and Micron may face increased competitive pressure, and Chinese manufacturers will become important variables in the global memory market.
Second, the value of the AI industry chain is spreading upstream.
Previously, market funds concentrated on AI models and computing chips; in the future, storage, semiconductor equipment, servers, power, and other infrastructure areas may also gain more attention.
Third, the investment logic in the technology industry is changing.
The market is shifting from traditional industries toward hard technology fields with technological barriers and strategic industrial value.
For the crypto community, this is also worth noting.
The development of the AI industry not only drives the growth of computing power demand but also affects the entire tech ecosystem, including data centers, chip supply chains, and energy infrastructure.
The core of future market competition is not just "who has the stronger AI model" but who controls the underlying resources supporting AI development.
The plunge in Korean stocks and Changxin's rise reflect a reshuffling of the global technology industry landscape.
The competition in memory chips may have only just begun. SNDK and MU are still being dragged along.
Tomorrow SK Hynix will release its earnings report, and the South Korean KOSPI has already crashed in advance.
Interestingly, the Nasdaq now seems to ignore fundamentals and is just focusing on the KOSPI — this index is almost turning into a casino mode.
In the next two days, capital expenditure figures from Microsoft, Amazon, and Meta will also be released.
The forward PE of memory stocks has already been squeezed down to 5-6 times, which indeed looks absurd.
There is also a background: Apple wants to use Chinese memory chips on the blacklist, and Micron is pushing back hard; this issue is likely to drag on for a long time.
High capital expenditure + political friction, the pressure is not small at all.
The process of clearing leveraged long positions is not over yet, and funds are still waiting to slowly buy in at lower levels. SNDK and MU are still being dragged along.
Tomorrow SK Hynix will release its earnings report, and the South Korean KOSPI has already crashed in advance.
Interestingly, the Nasdaq now seems to ignore fundamentals and is just focusing on the KOSPI — this index is almost turning into a casino mode.
In the next two days, capital expenditure figures from Microsoft, Amazon, and Meta will also be released.
The forward PE of memory stocks has already been squeezed down to 5-6 times, which indeed looks absurd.
There is also a background: Apple wants to use Chinese memory chips on the blacklist, and Micron is pushing back hard; this issue is likely to drag on for a long time.
High capital expenditure + political friction, the pressure is not small at all.
The process of clearing leveraged long positions is not over yet, and funds are still waiting to slowly buy in at lower levels. $SKHY SK Hynix delivered a financially distinctly "poor" report:
1. Revenue was 79.3 trillion KRW, up 51% quarter-on-quarter and 256.8% year-on-year; Analysts expected 83.85 trillion won
2. Operating profit was 60.54 trillion KRW, up 61% quarter-on-quarter and 557.2% year-on-year
3. Gross margin of 83%
Impact of the financial report:
1. Hanhai Lishi only rebounded after dropping to 983. If the negative news doesn't deepen the drop, then a short squeeze is needed. Short sellers should be careful, as both the odds and win rate for short sellers are low. I placed a long order at 1010 with a stop loss of 980, but no trade
2. High-end storage capacity expansion is very rapid, with volume ramping up in the second half of the year. Mid- to low-end is a must-see, but in the medium to long term, it will definitely be negative for industrial standard component storage represented by SNDK
3. Gross margin of 83%, most likely peaked, as the entire storage industry is expanding capital expenditure and capacity
4. Further confirmed that storage stocks are cyclical. If it's a cyclical stock, don't have a big view—just buy and exit quickly, treat it like a contract, set stop-losses, find places with a high profit-loss ratio, and exit after a quick trade. But a more stable long-term strategy is to short (I'm doing rebounds in the short term).
5. This financial report further increases the probability that "Samsung Hynix's price has already peaked."
6. The subsequent financial reports are hard to beat this one. Of course, this good is based on expectations, not actual quality, but on how well it matches the price
7. Except for HBM, other product lines are expected to be quickly crushed by China. The timeliness and efficiency of Chinese people cannot be measured by the American approach
Future trading direction: An objective fact shows that Teacher Shenyu's leverage data for Korean retail investors is still at the 98th percentile of history, which remains very high. Later, we just need to wait for Koreans to keep pulling leverage to over 50%, then we can keep shorting at high levels, with a very high win rate! I've figured out the wildness of Koreans: Hynix is highly volatile, making it a great cash machine. Of course, in the short term, it's still worth buying and taking a rebound
$SKHYNIX
$SNDK SNDK and MU are still being dragged along.
Tomorrow, SK Hynix will release its financial report, and Korea's KOSPI has already crashed ahead of schedule.
Interestingly, Nasdaq now seems to ignore fundamentals and focus only on KOSPI—this index is almost turning into a casino mode.
In the next two days, we will also see capital expenditure figures from Microsoft, Amazon, and Meta.
The forward PE ratio of memory stocks has already dropped to 5-6 times, which looks truly outrageous.
There's also the background: Apple wants to use Chinese domestic storage from the blacklist, while Micron is fighting head-on. This issue is likely to drag on for a long time.
High capital expenditures + political friction mean the pressure is no small matter.
The process of liquidating leveraged bulls is not yet complete; funds are still waiting to slowly buy at lower levels.SNDK and MU are still being dragged along.
Tomorrow, SK Hynix will release its financial report, and Korea's KOSPI has already crashed ahead of schedule.
Interestingly, Nasdaq now seems to ignore fundamentals and focus only on KOSPI—this index is almost turning into a casino mode.
In the next two days, we will also see capital expenditure figures from Microsoft, Amazon, and Meta.
The forward PE ratio of memory stocks has already dropped to 5-6 times, which looks truly outrageous.
There's also the background: Apple wants to use Chinese domestic storage from the blacklist, while Micron is fighting head-on. This issue is likely to drag on for a long time.
High capital expenditures + political friction mean the pressure is no small matter.
The process of liquidating leveraged bulls is not yet complete; funds are still waiting to slowly buy at lower levels.沉寂了几年的老牌芯片巨头英特尔，业绩迎来了久违的“爆发式”增长。 日前，英特尔公布了截至2026年6月底的第二季度财报：实现营收161亿美元，同比增长25%，较分析师普遍预期的144亿美元高出近17亿美元，同比增速创2011年第三季度以来新高。调整后每股收益为42美分，远超市场预期的21美分。业绩发布后，英特尔股价在盘后交易中一度涨超13%。 “我们看到了近15年来最强劲的营收增长。”英特尔首席执行官陈立武在财报电话会上直言，这背后是产品需求的强劲持续超过公司不断增长的供应。他进一步指出，英特尔正处于AI基础设施大规模建设浪潮的核心受益位置，拥有x86 CPU、先进封装及晶圆代工三大战略资产。 当整个行业正面临“历史上最严峻的供应限制之一”时，这三张王牌如何共同构成英特尔在短缺潮中的防御与进攻逻辑？ x86 CPU的不可替代性：AI推理阶段的“大脑”需求 AI训练阶段大量使用GPU，这一点毋庸置疑。但真正决定AI应用能否大规模落地的，是推理阶段——模型部署、实时响应、任务调度，这些环节对通用计算的需求正在激增。 背后的逻辑正在发生深刻变化。随着AI从训练向推理扩展，并日益向自主智能体和Missing out on SanDisk (SNDK), which surged 53 times this year, my calm SKILL
I used to think of SanDisk just as an old marginal brand making camera memory cards and USB drives? But after its spin-off and reorganization from Western Digital in 2025, SanDisk on one hand chokes the throughput of the entire AI large model cluster, and on the other hand steadily advances in the consumer storage field.
In 2025, SanDisk officially went solo from Western Digital and relisted on Nasdaq, catching the rocket of AI enterprise storage. The latest Q3 fiscal 2026 report is terrifying: quarterly revenue soared to $5.95 billion, exceeding expectations by 25.6%; EPS per share blasted to $23.41, grinding Wall Street’s expectations into the ground!
But facing such performance and takeoff speed, having completely missed the ride, today I try to use my previously outlined investment framework to calm my restless heart and avoid rash mistakes.
1. WASP Moat Breakdown
Width: Fully capturing the dual dividends of North American cloud AI data lakes and high-speed consumer electronics. The core product ace is AI and data center enterprise-grade solid-state drives (eSSD and large-capacity high-density QLC flash arrays dedicated to data lakes).
Advantage: Joint venture wafer fab with Kioxia, while accumulating a massive flash memory physical layer and controller chip layer with an extremely large cross-patent pool.
Stickiness: Once enterprise PCIe Gen5 SSDs run training arrays, the replacement cycle is very long, creating a certain lock-in effect.
Pricing power (critical pain point ⚠️): Strong. But this is a "limited-time trial card." NAND flash inherently has commodity cycle attributes; when the whole network is out of stock, it’s a money printer, but once major fabs release excess capacity, price cuts will be ruthless.
2. AI BOM Industry Chain Positioning
In the AI infrastructure bill of materials (BOM), our focus is usually on GPUs and HBM memory. But large model throughput requires massive "high-speed training data lakes" to feed computing power. SanDisk’s enterprise large-capacity NVMe SSD arrays are exactly the underlying core that blocks "compute hunger."
3. GTM Timing and My Thoughts
Currently, SanDisk’s dynamic PE is as high as 75.9x, with the stock price crazily revalued from an unbelievable 52-week low within two years. The current valuation not only includes AI’s high growth but is also filled with the most optimistic sentiment of the commodity cycle.
Core judgment: Great industry, explosive performance, but at an extremely dangerous "cycle peak" expectation.
Current strategy: Absolutely no chasing highs! Currently in a severe overbought phase. I prefer light positions to watch, or wait for commodity sentiment to pull back, then consider absorbing high implied volatility premiums by selling long-dated PUTs in the $1600-$1750 support range.
In summary: SanDisk is a super cycle beast cloaked in AI holy robes; you can envy its huge profits, but you must also respect the gravitational pull of the cycle. Missing out on SanDisk (SNDK), which has risen 53 times annually, my calm SKILL
I used to know SanDisk, but was it just an old-fashioned edge brand making camera memory cards and USB drives? After its 2025 spin-off from Western Digital, SanDisk will both hold the throat for the entire AI large model cluster throughput and steadily pursue progress in the consumer storage sector.
In 2025, SanDisk will officially go solo from Western Digital and re-list on Nasdaq, riding the rocket of AI enterprise-level storage. The latest Q3 fiscal year 2026 financial report is nothing short of staggering: single-quarter revenue soared to $5.95 billion, beating expectations by 25.6%; EPS per share soared to $23.41, crushing Wall Street's expectations!
But faced with such performance and rapid takeoff, I never got on the board. Today, I tried to use the investment framework I had previously organized to calm my restless heart and avoid impulsiveness and mistakes.
1. WASP moat dismantling
Width: Fully absorbs the dual dividends of North American cloud giants AI data lakes and high-speed consumer electronics. The core flagship of the product is the AI and data center enterprise-grade solid-state drives (eSSD and high-capacity, high-density QLC flash arrays dedicated to data lakes).
Advantage: Joint venture wafer fab with Kioxia, while also accumulating a massive pool of flash memory cross-patent pools for both physical and control chip layers.
Stickiness: Once enterprise-grade PCIe Gen5 SSDs run through training arrays, the verification and replacement cycles are very long, resulting in a certain lock-in effect.
Pricing (Fatal Pain Point ⚠️): Strong. But this is a "limited-time trial card." NAND flash memory is inherently a cyclical commodity market. When the entire internet is out of stock, it acts like a money printer. Once major wafer fabs release excess capacity, bargaining will be ruthless.
2. Positioning the AI BOM industry chain
In the bill of materials (BOM) for AI infrastructure, our focus is on GPU and HBM memory. But in reality, large model throughput requires massive "high-speed training data lakes" to fill computing power. SanDisk's enterprise-grade large-capacity NVMe SSD arrays are precisely the core solution to block the "computing power hunger."
3. GTM timing and my approach
Currently, SanDisk's dynamic PE ratio is as high as 75.9 times, and the stock price has been wildly revalued over two years from an incredible 52-week low. Current valuations not only include AI's high growth but also the most optimistic sentiment in commodity cycles.
Core judgment: Good industry, explosive performance, but extremely dangerous "cycle peak" expectations.
Current strategy: Firmly do not chase highs! Currently, it is in a severely overbought phase. I prefer to hold a light position and wait for commodity sentiment to pull back, then consider selling PUT in a forward position to absorb high hidden wave premium funds within the $1600-$1750 support range.
In short: SanDisk is a super-cyclical beast clad in AI holy armor. You may envy its huge profits, but you must also respect the gravity of cycles.$SOL **SOL 7/29 Pre-Market Overview**
📊 **Current price $73-74**, after a 24-hour drop from a high of $77 to a low of $72.86, there was a slight rebound. ATH $295, now down **75%**, with a 60% year-over-year decline.
🔻 **The technical side is a complete mess. ** All EMAs are overhead: 20-day $76.85, 50-day $76.79, 100-day $80.99, 200-day $94.82. 4-hour RSI 35.57, not oversold but weak momentum. The daily RSI reading has reached 29, close to being oversold. ADX is only at 11.54, a weak trend, and is prone to a false breakout and stop-loss sweep.
📉 **Clearing maps is dangerous. ** Near $72.50, there is a large accumulation of long liquidations, and at $74, there are short liquidations. The price is stuck in the middle, and any breakout direction will trigger a chain reaction. $70 is a structural watershed—holding it can still target $77-80; if it falls below it, target $65 or even $60 (June low).
📈 **But the fundamentals are the best. ** Morgan Stanley just launched the SOL spot ETF on July 28, with the lowest management fees. Q2 tokenized asset trading volume was $5.8 billion**, up 114% quarter-on-quarter. Circle just minted 250 million USDC on Solana. Mubadala (Abu Dhabi sovereign wealth fund) has brought private equity funds onto the Solana chain. On-chain prices are booming, while coin prices are lying flat.
⚠️ **Tonight's FOMC meeting is key. ** CME pricing remains at 62-65%, with a 25bp hike at 35-38%. No rate hike + dovish bias→ SOL has a chance to reach $75-77; no rate hike + hawkish → might target $72.50; unexpected rate hike → $70 is highly unlikely to hold.
🎯 **Levels:** Support at $73 → $72.50 → **$70 (Life-and-Death Line)** → at $67 → $65; Resistance at $74 → $75 → $76.79-85 (EMA concentration zone) → **$80-81 (100-day EMA + liquidation wall)**. Only when the daily price closes above $80 can you say the trend is reversing; if it doesn't reach the close, don't rush to buy in.
🔥 In short: On-chain data is a bull market, K-lines are bear markets. ETFs are a long-term positive sign, but the short-term FOMC has the final say. If it doesn't break $70, just hold it and watch; if it does, don't hesitate.Missing out on SanDisk (SNDK), which has risen 53 times annually, my calm SKILL
I used to know SanDisk, but was it just an old-fashioned edge brand making camera memory cards and USB drives? After its 2025 spin-off from Western Digital, SanDisk will both hold the throat for the entire AI large model cluster throughput and steadily pursue progress in the consumer storage sector.
In 2025, SanDisk will officially go solo from Western Digital and re-list on Nasdaq, riding the rocket of AI enterprise-level storage. The latest Q3 fiscal year 2026 financial report is nothing short of staggering: single-quarter revenue soared to $5.95 billion, beating expectations by 25.6%; EPS per share soared to $23.41, crushing Wall Street's expectations!
But faced with such performance and rapid takeoff, I never got on the board. Today, I tried to use the investment framework I had previously organized to calm my restless heart and avoid impulsiveness and mistakes.
1. WASP moat dismantling
Width: Fully absorbs the dual dividends of North American cloud giants AI data lakes and high-speed consumer electronics. The core flagship of the product is the AI and data center enterprise-grade solid-state drives (eSSD and high-capacity, high-density QLC flash arrays dedicated to data lakes).
Advantage: Joint venture wafer fab with Kioxia, while also accumulating a massive pool of flash memory cross-patent pools for both physical and control chip layers.
Stickiness: Once enterprise-grade PCIe Gen5 SSDs run through training arrays, the verification and replacement cycles are very long, resulting in a certain lock-in effect.
Pricing (Fatal Pain Point ⚠️): Strong. But this is a "limited-time trial card." NAND flash memory is inherently a cyclical commodity market. When the entire internet is out of stock, it acts like a money printer. Once major wafer fabs release excess capacity, bargaining will be ruthless.
2. Positioning the AI BOM industry chain
In the bill of materials (BOM) for AI infrastructure, our focus is on GPU and HBM memory. But in reality, large model throughput requires massive "high-speed training data lakes" to fill computing power. SanDisk's enterprise-grade large-capacity NVMe SSD arrays are precisely the core solution to block the "computing power hunger."
3. GTM timing and my approach
Currently, SanDisk's dynamic PE ratio is as high as 75.9 times, and the stock price has been wildly revalued over two years from an incredible 52-week low. Current valuations not only include AI's high growth but also the most optimistic sentiment in commodity cycles.
Core judgment: Good industry, explosive performance, but extremely dangerous "cycle peak" expectations.
Current strategy: Firmly do not chase highs! Currently, it is in a severely overbought phase. I prefer to hold a light position and wait for commodity sentiment to pull back, then consider selling PUT in a forward position to absorb high hidden wave premium funds within the $1600-$1750 support range.
In short: SanDisk is a super-cyclical beast clad in AI holy armor. You may envy its huge profits, but you must also respect the gravity of cycles.Missing out on SanDisk (SNDK), which surged 53 times this year, my calm SKILL
I used to think of SanDisk just as an old marginal brand making camera memory cards and USB drives? But after its spin-off and reorganization from Western Digital in 2025, SanDisk on one hand chokes the throughput of the entire AI large model cluster, and on the other hand steadily advances in the consumer storage field.
In 2025, SanDisk officially went solo from Western Digital and relisted on Nasdaq, catching the rocket of AI enterprise storage. The latest Q3 fiscal 2026 report is terrifying: quarterly revenue soared to $5.95 billion, exceeding expectations by 25.6%; EPS per share blasted to $23.41, grinding Wall Street’s expectations into the ground!
But facing such performance and takeoff speed, having completely missed the ride, today I try to use my previously outlined investment framework to calm my restless heart and avoid rash mistakes.
1. WASP Moat Breakdown
Width: Fully capturing the dual dividends of North American cloud AI data lakes and high-speed consumer electronics. The core product ace is AI and data center enterprise-grade solid-state drives (eSSD and large-capacity high-density QLC flash arrays dedicated to data lakes).
Advantage: Joint venture wafer fab with Kioxia, while accumulating a massive flash memory physical layer and controller chip layer with an extremely large cross-patent pool.
Stickiness: Once enterprise PCIe Gen5 SSDs run training arrays, the replacement cycle is very long, creating a certain lock-in effect.
Pricing power (critical pain point ⚠️): Strong. But this is a "limited-time trial card." NAND flash inherently has commodity cycle attributes; when the whole network is out of stock, it’s a money printer, but once major fabs release excess capacity, price cuts will be ruthless.
2. AI BOM Industry Chain Positioning
In the AI infrastructure bill of materials (BOM), our focus is usually on GPUs and HBM memory. But large model throughput requires massive "high-speed training data lakes" to feed computing power. SanDisk’s enterprise large-capacity NVMe SSD arrays are exactly the underlying core that blocks "compute hunger."
3. GTM Timing and My Thoughts
Currently, SanDisk’s dynamic PE is as high as 75.9x, with the stock price crazily revalued from an unbelievable 52-week low within two years. The current valuation not only includes AI’s high growth but is also filled with the most optimistic sentiment of the commodity cycle.
Core judgment: Great industry, explosive performance, but at an extremely dangerous "cycle peak" expectation.
Current strategy: Absolutely no chasing highs! Currently in a severe overbought phase. I prefer light positions to watch, or wait for commodity sentiment to pull back, then consider absorbing high implied volatility premiums by selling long-dated PUTs in the $1600-$1750 support range.
In summary: SanDisk is a super cycle beast cloaked in AI holy robes; you can envy its huge profits, but you must also respect the gravitational pull of the cycle. SK Hynix falls short of expectations, but the market rises instead of falling
The main reason is that the Korean won appreciated by 7% in a single month, and the US dollar selling storage to exchange for Korean won has dropped by 7% compared to last month
If you interpret it this way, this is actually a great opportunity to bottom-fish
Maybe August is a good opportunity for a rebound?
$SKHYNIX
$SNDK ☀️ "Uncle's Morning Report"｜2026.07.29 (Wednesday)
📌 Asia-Pacific markets panic sell-off across the board, BTC falls below 64000, final wait before FOMC decision
Today, let's first see what happened around the world.👇
━━━━━━━━━━━━━━━━━━
🌍 Overnight Summary
On Tuesday, Asia-Pacific markets collectively experienced a panic sell-off, South Korea's stock index plunged triggering a circuit breaker, China's ChiNext index dropped 7.35% in a single day; overnight U.S. stocks showed clear divergence, software leaders held up while hardware chips continued to adjust. Oil prices retreated to digest geopolitical premiums, BTC remained under pressure trading below 64000. Global funds actively reduced positions, everyone is quietly awaiting the FOMC interest rate decision early Thursday morning.
□□ U.S. Stocks Overnight｜Sector Divergence, AI Hardware Continues Valuation Adjustment
The three major U.S. indices closed mixed overnight: Dow +0.51% at 52210.21, S&P 500 +0.02% at 7413.55, Nasdaq -0.18% at 24930.69.
Clear market layering: Google up over 2%, Microsoft nearly 2%, Apple over 1%; Nvidia dropped nearly 5%, Tesla down over 1%, Philadelphia Semiconductor Index fell 2.23%.
👉 Uncle's Observation
U.S. stocks did not collapse in sync with Asia-Pacific; funds showed clear selection: institutions kept buying AI software stocks while continuously selling AI hardware chips. This divergence indicates the U.S. market fundamentals are not deteriorating, but previous gains were excessive and are now undergoing valuation digestion. Tonight's after-hours earnings from Microsoft and Meta will test AI profitability realization.
🪙 Crypto｜BTC Under Pressure Below 64000, Bulls Weak Before FOMC
On Tuesday, dragged by Asia-Pacific panic sentiment, BTC dipped to a low of 62100 USD, currently trading around 63900 USD, continuously below the 64000 resistance level.
Key market change: the 63000-64000 range has turned into a short-term resistance zone. Although there was some bottom-fishing after intraday lows, BTC never effectively broke above 64000, fully indicating bulls are unwilling to actively counterattack before the Fed decision.
Derivatives risk warning: market long-short structure is unbalanced; a confirmed break below 62000 USD could trigger large-scale chain liquidations. The market remains extremely cautious before the FOMC announcement.
👉 Uncle's Observation
Yesterday's BTC movement was a combined effect of Asia-Pacific panic selling and bulls reducing positions early. 64000 is the short-term resistance, 62000 the first key support; if U.S. stocks weaken further pre-market tonight, support may be at risk. Before the decision, short-term rebounds should be considered as bear traps.
□□ A Shares｜Pre-market Reminder: Emotional Recovery Difficult
Yesterday A shares declined all day: Shanghai Composite -1.16% at 3813.31, Shenzhen Component -4.52%, ChiNext -7.35%, STAR 50 -6.33%. AI computing power and storage chips led the decline.
Today's core focus: can Asia-Pacific tech panic sentiment stabilize? If external sentiment fails to recover, tech growth will likely remain under pressure; funds may continue shifting to defensive sectors like utilities and banks.
👉 Uncle's Observation
ChiNext's 7.35% plunge is an emotional stampede, reflecting panic resonance in Asia-Pacific AI hardware sector. Short-term panic sentiment is not fully released; avoid rushing to bottom-fish for a reversal.
□□ Hong Kong Stocks｜Pre-market Reminder: Watch 24850 Support Validity
Hong Kong stocks opened low and fell yesterday: Hang Seng Index closed down about 1.4% near 24850, Hang Seng Tech Index dropped over 2.5%. Internet and semiconductor sectors under full pressure.
👉 Uncle's Observation
Hong Kong's decline was relatively smaller in Asia-Pacific, not due to strong resilience but deeper prior adjustments. Today's focus is on 24850 support and changes in southbound capital flows.
🛢️ Geopolitics and Oil Prices｜Brent Falls Below 85 USD, Geopolitical Premium Quickly Fades
Brent crude retreated to 84.84 USD/barrel, WTI at 82.38 USD/barrel. Expectations of a U.S.-Iran ceasefire combined with global demand concerns caused oil prices to fall over 15 USD from the 100 USD mark.
👉 Uncle's Observation
Oil prices are falling rapidly, geopolitical risk premiums continue to fade. Theoretically easing global inflation pressure and reducing urgency for Fed rate hikes, but short-term benefits are completely overshadowed by market panic.
🎯 Today's Key Variables
1️⃣ Microsoft and Meta after-hours earnings — can AI commercialization results reverse pessimism in hardware sector?
2️⃣ FOMC interest rate decision (Beijing time Thursday 2:00 AM): market 99.5% probability of rate hold; statement wording impact greater than rate hike or not;
3️⃣ Can Asia-Pacific panic sentiment stabilize? Watch A shares and Hong Kong stocks' capital absorption strength.
💎 Uncle's View
Yesterday's global market theme was very clear: Asia-Pacific stocks sharply sold off, BTC pressured down, funds fully contracted awaiting Fed decision.
South Korea's stock index plunged 10.84% in one day, ChiNext dropped 7.35%, essentially not due to sudden fundamental deterioration but panic-driven stampede; Asia-Pacific AI hardware sector is undergoing a severe clearing.
Before the FOMC decision, short-term asset movements lack long-term guidance significance. Until direction is clear, stay cautious, control positions, and wait for signals early Thursday morning. #美联储周四凌晨公布利率决议
━━━━━━━━━━━━━━━━━━
🌍 Watch global markets, follow the money flow.
🧠 Faster News. Better Decisions.
⚠️ Personal observation, not investment advice.兄弟们，昨晚全球资本市场这剧本，编剧都不敢这么写。原油又崩了，三天跌掉十几美元，WTI直接干到79块多。黄金也没扛住，从今年1月高点下来跌了快三成。最惨的是存储芯片和AI硬件，闪迪一个月腰斩，美光、SK海力士跌了快9个点，费城半导体指数两天干没4.5%。可另一边呢？道指涨了537点，苹果盘中市值摸到5万亿美金，中概股金龙指数涨了1个多点。这哪是同一个市场啊，简直是在演《双城记》。 原油三连杀，这是要回到解放前？ 先唠唠油。昨天WTI原油又跌了4.21%，收在79.13美元，布伦特也跌了3.54%，收82.83美元。算上前面两天，这一轮三连跌直接把油价从90多砸到了70多，创了近三年最大三日跌幅。 为啥跌这么狠？中东那边消停了。以色列和伊朗的紧张局势暂时缓和，市场觉得供应端不会出大乱子。再加上OPEC+那边传出要暂停增产的消息，本来该涨的反而跌了，说明市场交易的是"战争溢价"出清，不是基本面恶化。 但有意思的是，油价大跌对股市到底是好事还是坏事？教科书说油价跌=通胀降=加息概率降=股市涨。昨晚美股也确实涨了，但只涨了一半——道指大涨1.03%，纳指却跌了0.22%。钱没往科技股跑，跑去传7月29日，国际原油市场经历了一场情绪与资金的剧烈震动。 WTI原油期货在多重利空的共振下，不仅轻易击穿了80美元/桶的心理防线，更是在盘中创下近三个月以来的新低。与此同时，布伦特原油自上周中触及的阶段性高位起算，短短三个交易日内累计暴跌约16%，其下挫的烈度堪比2020年负油价危机后的单次冲击。然而，就在国际盘面一片“深绿”之际，国内成品油价格调整却进入读秒阶段——截至今日，本轮计价周期第8个工作日，机构监测到的原油变化率依然在高位坚挺，对应汽柴油上调幅度锁定在690元/吨附近。一边是国际原油的“断崖跳水”，另一边是国内油价板上钉钉的大涨，二者之间巨大的“温差”从何而来？油市后续又将何去何从？本文将从行情回溯、暴跌驱动、国内调价机制以及后市展望四个维度，对7月29日的原油走势进行深度解析。 亚欧交易时段，市场的恐慌情绪并未因超卖而衰减。WTI原油主力合约在隔夜跌破82美元后加速下行，今日午盘后空头再度发力，一举将价格砸至79.20美元附近，日内跌幅一度扩大至4%以上。这是自今年5月初以来，美油首次明确收于80美元下方。技术形态上，周线、日线均出现了“断头铡刀”式的破位，MACD死叉张[Banks Pressure to Tighten Income Incentives, Stablecoin Policy Expectations Remain Cautious]
Stablecoin policy expectations are relatively cautious, and the banking industry is calling for expanded restrictions on yield-based arrangements, meaning that even if legislation continues to leave room for payment stablecoins to develop, commercial designs around holding returns may face stricter boundaries. This letter is merely a lobbying move rather than a legal decision, so the direction has not yet turned into certainty bearish, but related expectations should not be prematurely traded into broad positive news.
134 U.S. bank association officials and bank executives jointly called on the Senate to amend Section 10404 before the CLARITY Act is finalized, strengthening limits on stablecoin interest and yield payments, and incorporating incentives, incentives, or other arrangements that provide similar economic benefits into constraints. The signatories believe that if stablecoins can attract and retain balances with rewards similar to the interest, it could weaken the deposit base supporting local loans. What truly matters is that the debate has shifted from whether to allow payment stablecoins to how they can compete for balances.
From a funding logic perspective, returns or rewards are direct tools for retaining balances; Once the scope of restrictions is expanded, the path to scale through economic incentives narrows, and product growth depends more on actual needs such as payments and settlements. For banks, this helps reduce uncertainty in deposit outflows; For the stablecoin ecosystem, it means competition between user retention and funding costs may be repriced, rather than simply being divided by regulatory approval.
Afterwards, it will depend on whether the Senate text adopts the call for expanded restrictions and to what extent "rewards and incentives" will be defined; If the provisions remain narrow, current tightening expectations may fall back.
The above is only a personal opinion sharing and does not constitute any investment advice. The market changes rapidly; trading profits and losses is borne by yourself.[Trade.xyz Stabilize Sentiment First After Full Losses, DeFi Contract Trading Remains on the Sidelines]
Regarding Trade.xyz and related DeFi contract trading narratives, the short-term caution is maintained. The platform's proposal for full compensation and accelerated reforms helps reduce losses and build trust, but is not enough to immediately offset the risk discount left by abnormal liquidations. For users who passively trigger liquidation, the most important thing is not whether the system "operates as designed," but whether the pricing mechanism can prevent similar tail end situations from happening again.
The event occurred at 23:01 UTC on July 27: the token price of SK Hynix quickly dropped from $1127.9 to $917.25, leading to a large number of long positions being liquidated. The platform stated that this price corresponds to a real transaction scraped by multiple independent data providers, sourced from Korea's main pre-market market, with oracles synchronizing external prices according to established specifications; At the same time, it announced full compensation and accelerated pricing mechanism reforms. This means the dispute is not about whether the transaction actually existed, but whether a transaction at an external venue would have too much impact on contract liquidation if included in the marking price.
Compensation reduces the pressure on this loss to spread to users, and also shows that the platform prioritizes market integrity over rule interpretation; However, from the product structure perspective, compensation only addresses one-time losses, and the marking price's dependence on external transaction synchronization determines the future risk premium. If the reform plan remains only a statement, traders will still retain higher liquidity and a trust discount for extreme markup prices.
Afterwards, it depends on whether the compensation is fully implemented and whether the platform can publish and implement verifiable pricing adjustments; It is especially important to observe whether the marking price and liquidation mechanism will still be equally amplified when extreme external transactions occur.
The above is just a personal opinion sharing and does not constitute any investment advice. The market changes rapidly, and trading profits and losses are borne by the buyer.[HYPE Release Supply Pressured, Cautious Short-Term]
HYPE is cautious in the short term, and the unstaking shares are creating real potential selling pressure. At this stage, it is not suitable to assume good news solely based on large amounts of staking funds. Over the past week, the coin price has dropped by 10%, and new large-scale collateral releases have been launched at this time, leaving supply-side expectations uncleared.
Most importantly, Multicoin Capital had 1.97 million HYPE, about $108 million, which was withdrawn after the seven-day staking period was unstaked; of these, 86,000 (about $4.78 million) had been transferred to Coinbase Prime. The previous week, the institution had transferred 395,000 HYPE, about $37 million, into Coinbase. Transferring into Prime does not mean you have already sold in the spot market, but it at least converts this portion of tokens from a locked state into liquid tokens that can be executed, distributed, or sold, and the market naturally takes the pressure first.
The core here is not the direction of single transfers, but changes in chip structure: after large holders redeem from staking, the circulating supply increases, and the volume of buyers to take on also rises. If funds are transferred into Prime only as a reserve for reducing positions, short-term valuations will continue to bear discounts; Conversely, if the scale of outflows is limited and prices can stabilize, it indicates that the market has the ability to absorb new supply.
Afterwards, it depends on whether the remaining staked tokens continue to enter the trading venue and whether the actual transaction pressure after transferring can be absorbed; Before these signals are clear, "not yet sold" should not be taken as the risk eliminated.
The above is just a personal opinion sharing and does not constitute any investment advice. The market changes rapidly, and trading profits and losses are borne by the buyer.意外，昨天尽管原油和美元都下跌，但本应上涨的黄金却下跌50美元
周二黄金价格下跌，接近当前支撑位及约4000美元的双底水平。截至美国东部时间下午4时55分，黄金期货下跌50美元，跌幅1.22%，至4029美元。
自三月初以来，黄金价格受到原油价格上涨引发的通胀压力加剧的影响。高油价促使交易者预计到9月至少会加息一次，这与此前今年降息的预期相反。
油价上涨叠加美联储主席凯文·沃什在首次主持FOMC会议时略显鹰派的表态，令市场对周三央行决策的预期仍存在分歧。约三分之二的市场参与者预计利率将维持在3.50%至3.75%不变，而约三分之一的期货交易者则预计加息25个基点，使联邦基金利率升至3.75%至4.00%。
对利率上升的预期推动美元指数升至14个月来的最高水平，直接给金价带来压力。过去五个月里，油价上涨与美元走强的双重因素使黄金价格持续处于下跌三角形的熊市格局中。在此期间，美元指数累计上涨了约3.8%。
尽管周二原油和美元指数均下跌，黄金价格仍继续走低，很可能是因为交易者已将周三美联储会议结束时加息的可能性计入了价格。
自2月28日中东冲突引发通胀预期以来，黄金一直未能突破这一下降三角形形态。自冲突爆发以来，黄金价格已下跌约22%，这与传统上受益于地缘政治不确定性的金属表现相悖。相反，黄金主要受到加息预期、实际收益率上升以及美元走强的推动。在利率开始回落或市场开始预期更低利率之前，这种趋势不太可能改变。
周三的利率决议可能对贵金属价格具有决定性影响，目前金价已接近其下降三角形形态的顶点。若美联储加息，金价很可能跌破约3900美元的关键支撑位。即使利率保持不变，如果市场认为美联储的指引偏鹰派，支撑位仍可能面临压力。相反，若美联储释放出更偏鸽派的信号，则黄金有望在五个月来首次突破下行阻力线。无论哪种情况，黄金正逼近一个由不断走低的高点和双底构成的形态顶点，为美联储决策后可能出现的重大突破奠定了基础。
-------------------------------------------------------------------------------------------------------------------------
大家对黄金最新走势有何看法，欢迎评论区评论。
作者:金银大王说金银大宗商品贵金属评论员、国际财经作者 ;
第一时间带来最新最全国内外贵金属金融资讯、深度洞察和最新趋势;
有问题可在评论区提问，尽量解答
#创作打卡赢现金#停火预期兑现，WTI原油期货单日跌8.68% $CL Why was SanDisk able to drop 15% in one day?
SanDisk recently closed at $1,278, down 11% in a single day. However, the official announcement has not yet been made of major negative news; the real test is the August 5th earnings report
Last quarter, SanDisk's revenue was $5.95 billion, with a gross margin of 78.4%. Its data center business grew 233% quarter-on-quarter, showing strong fundamentals. The problem is that the stock price has risen too much ahead of schedule. The market is not worried about whether it will make money, but how long it can hold up with a gross margin close to 80%.
Changxin Memory's IPO was just the trigger. Changxin mainly focuses on DRAM, SanDisk mainly on NAND, and they are not direct competitors, but the rise of China's storage capacity will indeed prompt the market to reassess the entire industry.
In the short term, the first concern is whether the 1220–1250 USD range can hold. On August 5, focus on gross margin, data center revenue, and 2027 order guidance.
If it can hold on, this time it feels more like a valuation cooling; If it can't hold on, the market may have already started to peak its trading and storage cycle.
$SNDK Good morning, today's macro must-read
The market has entered a "wait-and-see mode" ahead of the Federal Reserve decision. BTC is oscillating narrowly around $64,000, with ETH leading gains at +1.74%. Derivatives signals show leveraged longs are retreating — open interest dropped -0.17% in the past 24 hours, and the funding rate fell from yesterday's high of +0.82% to +0.01%, indicating cooling momentum in chasing rallies. Liquidation data confirms this: longs were liquidated for $74.7 million in 24 hours, 3.7 times that of shorts. Overall, the market is in a 🟡 consolidation pattern, with direction depending on tomorrow morning's Fed rate decision.
🟡 Direction: Consolidation, confidence: medium — funding rate returning to neutral + long liquidation pressure → short-term lack of one-sided momentum, awaiting macro catalyst.
🌐 Market environment
Liquidity: The US Dollar Index at $101.43 is basically flat, 10-year US Treasury yield dropped to 4.60% (-0.80%), VIX fell to $18.21 (-2.46%), indicating a relatively loose macro environment and lower funding costs.
US stock sentiment: S&P 500 (SPY) +0.24%, but Nasdaq (QQQ) -0.97%, tech stocks under pressure, capital rotating from growth stocks to defensive sectors, risk appetite diverging.
Impact on BTC: Improved macro liquidity is positive, but weakness in tech stocks may weigh on risk asset sentiment. BTC's correlation with US stocks has weakened and is more waiting for its own catalyst.
1-4 week outlook: If the Fed signals dovishness (rising expectations of rate cuts), BTC could break above $65,000; but if the decision is hawkish or hints at delayed cuts, BTC may retest support at $62,000.
Good luck to us!
$BTC $QQQ
#美联储周四凌晨公布利率决议 Guys, ZIL rose 18.33% today, current price $0.002949, trading from around 0.0025 to 0.003 over the week. The core of the rally is the recovery of hacker theft panic + funding rate -2% bearish squeeze + technical demand zone support, a triple resonance. But the divergence is huge: deposits and withdrawals remain frozen, and once locked tokens resume, they may be released collectively. With a turnover of only 35 million, the rise occurring under locked liquidity, and the selling pressure after recovery is equally unpredictable. Key price levels: Resistance $0.00304-$0.00306, support at $0.00296 (stop loss), a break below $0.00280. Watch the stop loss level at 0.00296. Before deposits and withdrawals resume, every rally may be a short-term game rather than a trend reversal. Personal market view analysis and market information compilation, not investment advice. $BTC $ETH $ZIL #财报观察员: OKX MasterClass Premieres Tonight, Helping You Understand the Financial Reports of Four Tech Giants. #停火预期兑现, WTI crude oil futures fell 8.68% in a single day, #英伟达拟为OpenAI提供2500亿美元担保 ALLO Short Selling Profit Logic + Personal Market Sentiment (Original by Bo Shen, Like and Share)
Risk Warning: Virtual currency contracts are not protected by domestic laws, and high leverage can easily cause significant losses. The content is for market analysis only and does not constitute trading advice.
The current price of ALLO remains at $0.32. The highest point in this rally was $0.54, with a drop of over 40% from the peak. There are clear signs of major players unloading at high levels. Every rebound now is an excellent shorting opportunity. After understanding the market pattern, I have no impulse to chase longs; instead, I wait for the decline to realize profits with a very firm mindset.
The core profit logic for short selling is divided into three layers. First, the market funds have completed high-level distribution, with $0.54 forming a permanent heavy trapped position. The previous rally relied entirely on contract leverage speculation, with a 24-hour contract turnover of 145 million USDT, while spot turnover was only 14 million. The rise lacks real spot support; during the slow rise phase, retail investors blindly chased longs, with a net long-short ratio of 1.31 across the network. However, large holders continuously placed short orders to cash out between $0.48 and $0.54. After the peak, a rapid plunge occurred, deeply trapping many longs at high levels. Subsequent rebounds will inevitably be accompanied by selling pressure from position unwinding, completely locking the upside space.
Second, the token’s fundamental long-term selling pressure cannot be alleviated. ALLO has a total supply of 1 billion tokens, with only 20% currently circulating. 31.05% is private placement tokens, and 17.5% is team tokens with unlocking schedules continuing until 2028, resulting in a fixed large monthly token inflow into the market. Early investors’ cost was only $0.1087, so the current price still has room for doubling to exit. The project’s decentralized AI narrative has landed poorly, on-chain activity continues to decline, and there is no revenue to support token valuation. This rally was purely short-term AI-themed sentiment speculation. After the hype fades, valuation will continue to return to a reasonable range.
Third, the technical bearish trend is fully established. The $0.54 high formed a double top resistance. The current price at $0.32 continues to be pressured by the 20-day and 50-day moving averages. The daily RSI has dropped continuously from the overbought zone at 85, and every small rebound sees RSI hit 55 before turning down. Volume and price show a standard bearish structure: volume shrinks on rebounds and expands on declines, with a complete descending channel and continuously lower highs. Practically, one can build short positions in batches at resistance around $0.34-$0.35, with a stop loss at $0.365, first take profit at $0.28, and a mid-term target of $0.23.
My personal trading mood is exceptionally clear. The community is all fantasizing about returning to the $0.54 high, blindly bottom-fishing and adding longs, completely ignoring the trapped positions at the top and unlocking selling pressure. Bubble speculation markets are always a few major players harvesting retail followers. Seeing through the major players’ pattern of first pumping to lure longs and then dumping to harvest, I position short against the trend waiting for the drop. This kind of forecasted market and steady profit-taking from pullbacks is far more satisfying than blindly chasing rallies. Overall, the bearish momentum is sufficient, and shorting on rebounds is currently the optimal strategy.Yesterday, I told everyone that tonight's market revealing the market will only result in 'maintaining the same level.' Today I want to update an important change—the wind has shifted these past few days, and they've become quite urgent. What happened: The probability of a rate hike suddenly surged. Let's clarify the timing first. The Federal Reserve's interest rate decision will be announced today (29th) at 2 PM Eastern Time, which is 2 AM Beijing time tomorrow (30th). Warsh will hold a press conference half an hour later. So, when you read this article today, the results haven't even been released yet, and we're still in the tense hours before the resolution. But the real change is: the probability of a rate hike surged in just one week. According to CME FedWatch, the probability of a rate hike jumped from 10.7% on July 15 to about 38% on July 24—this is the fastest expected repricing in recent times. Although maintaining the rate unchanged remains mainstream (about 62% to 75%), the market's fear of a "rate hike" has clearly grown significantly. Why is that? You Tong fought together. The trigger wasn't the crypto world, but the Middle East. Renewed tensions between the US and Iran, cracks in the previous ceasefire, and oil prices have been pushed higher. When oil rises, inflation worries return—and when inflation rises, the Fed's reason to raise interest rates becomes harder. The June CPI was cooling (falling from 4.2% in May to 3.5%), which originally gave the Federal Reserve room to hold steady, but once oil prices moved up, that room narrowed. Even Citadel Securities expects the Federal Reserve to raise interest rates, although JPMorgan maintains its stance and forecasts itThe intense short squeeze of BEAT and KAITO has completed a concentrated liquidation of shorts, but the current price already incorporates the dual momentum of long positions and short liquidations. What remains unpriced is the interplay between funding rate reversal, position volume reset, and liquidity depletion.
What is the market repricing?
Priced in: BEAT surged from 2.19 to 4.59, an increase of about 110%, and KAITO rose from 1.1667 to 1.2010. The violent long squeeze has fully realized the short-term short squeeze potential. Unpriced variables: whether the funding rate will turn negative due to overcrowding, whether position volume will sharply drop due to liquidations, and whether risk appetite will reverse around the Fed rate decision.
Fact check:
- BEAT: Entry price 3.26009, current mark price 4.56106, unrealized loss 91,393.84 USDT, return rate -56.79%, contract value 320,419.026 USDT.
- KAITO: Entry price 1.1667, current mark price 1.2010, unrealized loss 9,261 USDT, return rate -11.37%, contract value 324,270 USDT.
- Time window: The Fed rate decision is early Thursday morning; macro risk has not yet been released.
Structural changes:
- Funding behavior: A BEAT short position of about 300,000 USDT still added to the position after a -56.79% unrealized loss, indicating this is an extreme conviction short rather than a quant or hedge position. This position itself becomes a "fuel pool" for the long squeeze— as long as this short is not closed, longs have the incentive to keep pushing the price up to trigger its liquidation.
- Transmission logic:
- The synchronized rally of BEAT and KAITO suggests capital is concentrating attacks on low-liquidity, high-short-ratio altcoins, a typical structural short squeeze strategy.
- Impact on BTC/ETH: Such strategies usually do not directly transmit to major coins, but if BEAT/KAITO funding rates rise above 100% annualized due to continuous rallies, arbitrageurs might sell spot and buy futures, indirectly suppressing ETH/BTC spot buying.
- Altcoin risk appetite: Successful short squeezes attract momentum longs, but once shorts are fully liquidated or capitulate, the rally logic ends, and capital may quickly exit, causing BEAT/KAITO prices to crash.
Bullish path and conditions:
- If the short position is liquidated near the mark price of 4.56 (liquidation price about 5.1-5.3), BEAT could surge above 5.5, driving KAITO to rally synchronously to 1.3-1.4.
- Conditions: funding rate remains controlled (annualized <200%), short position volume does not sharply decline, and macro events (Fed) do not trigger risk asset sell-offs.
Bearish risks and conditions:
- If the short position is voluntarily closed or liquidated, longs lose their target, and prices may quickly fall back to the 3.2-3.5 range, a drop of 30-40%.
- Tail risk: A hawkish Fed statement causes a broad risk asset sell-off; BEAT/KAITO, due to fragile liquidity, could experience a "stampede" style crash, with declines possibly exceeding 50%.
- Invalidating signals: BEAT position volume drops more than 30% within 24 hours, funding rate turns negative, or mark price falls below 3.0.
Conclusion: The short squeeze of BEAT and KAITO is nearing its end. The current price reflects a "hunt" by longs targeting specific short positions rather than fundamental improvement. Watch signals: whether BEAT position volume remains high, whether funding rate turns positive, and whether the short position closes. If position volume declines or funding rate turns negative, it should be seen as a signal of longs exiting.
Risk warning: Extreme short squeezes are usually followed by sharp reversals; pay attention to position management and liquidation risk.
$BEAT $KAITO #FedDon't talk to me about faith, tonight we first watch Powell's expression: AI's "cleansing" is only halfway through
Honestly, the current market is a bit nauseating to watch.
Nvidia is once again sluggish before the market opens, dropping 2 or 3 points no longer matters; what matters is it’s like a wet rag stuck on the Nasdaq’s face, suffocating everyone. Some say Samsung and SK Hynix crashed today because of the "Changxin Memory IPO" scare? Come on, don’t give such poetic reasons for the drop. Simply put, the money ran out, and the story can’t be spun anymore.
Look at the two "memory giants" from Korea: Samsung dropped over 8%, Hynix directly fell 10%, wiping out nearly 30% in less than a month. You think it’s a problem with HBM (High Bandwidth Memory)? No, HBM is still desirable, but the phone and PC sectors are a complete mess. Institutions are very smart now; they’re starting to do the math: can the premium you earn from AI cover the losses from the general DRAM (Dynamic Random Access Memory) price drops? If the math doesn’t add up, they’ll just sell off out of respect.
How far will this round of liquidation go? I’ll be blunt—it’s far from over.
Many people compare this to the 2007 internet bubble, but I think there’s no need to go that far back; just look at the 2022 wave, Meta dropped over 60%, Nvidia over 50%. The current AI leaders? They’re nowhere near that level yet. The biggest bomb tonight isn’t who fell, but that the earnings reports and the Fed meeting on the 29th collide. SK Hynix, Microsoft, Meta will speak, and Powell will speak too. These four at one table is a "Feast at Hongmen".
If Microsoft’s cloud growth dares to fall below 30%, or Powell utters "higher for longer" (maintaining high interest rates for longer), then when you wake up tomorrow morning, what you’ll see won’t be a "correction," but a "stampede." Those brothers who leveraged 3x chasing AI will probably lose sleep tonight.
As for "Will I play storage in the next life?"—that question really hits hard.
Yes, of course I will. But I won’t play "faith," only "cycles."
Storage is inherently the most speculative business. When it rises, it makes you feel like you can reach the sky; when it falls, it makes you feel like you’re going bankrupt. What stage are we at now? Just rolling down from the mountain top, not yet at the mid-slope. You ask me where the bottom is? Wait until no one in the market talks about "HBM penetration rate" as a fancy term anymore, wait until rumors of "production cuts and maintenance" start circulating in Samsung’s factories, then bend down and pick up the bleeding chips.
The current strategy is one word: endure. Or even more timid: watch.
Don’t go against the trend, and don’t listen to those influencers shouting about a "golden pit." Golden pits are filled with other people’s corpses; how do you know you’re not the corpse? Tonight, just quietly grab a small stool, watch how Microsoft smooths things over, watch how Powell changes his face. After this storm passes, we’ll talk again about "Is AI a revolution or a scam"—because only those who survive have the right to talk about faith!"DataHunter ETF Capital Flow Research Report" · July 28, 2026
Understanding the market with data
📋 Summary of this issue
On July 27 (Monday, Eastern Time), the US spot Bitcoin ETF recorded a net outflow for the third consecutive trading day, with a single-day outflow of $11.64 million. Ethereum spot ETFs continued their net inflow trend, with a single-day net inflow of $9.23 million. BlackRock became the core driver of the capital divergence between the two major ETFs—IBIT saw $8.82 million in outflows, ETHA $11.75 million in inflows.
📊 1. Bitcoin ETF: Three consecutive days of net outflows, BlackRock IBIT becomes the main force
On July 27, U.S. spot Bitcoin ETFs recorded a net outflow of $11.6439 million, marking the third consecutive trading day of net outflows following July 23 and 24.
Key Data:
· Net outflow for the day: $11.6439 million
· Total net asset value: $78.713 billion, ETF net asset ratio reaches 6.04% of Bitcoin's total market capitalization
· Historical cumulative net inflow: $51.374 billion
By institution:
· BlackRock IBIT: Single day net outflow $8.82 million, historical total net inflow $60.386 billion
· Fidelity FBTC: single-day net outflow of $2.8237 million, historical total net inflow of $10.002 billion
· Other Bitcoin ETFs saw zero capital inflow that day
Bitcoin ETFs have recorded net outflows for three consecutive days, sharply contrasting with the previous seven days of net inflows and nearly $1 billion in cumulative inflows. Last week (July 20-24), the weekly net inflow of Bitcoin spot ETFs was only about $33.9 million, far from enough to offset the significant losses in the previous two months.
💰 2. Ethereum ETF: BlackRock's ETHA holds the spotlight, with a single-day net inflow of $9.23 million
In stark contrast to Bitcoin ETFs, spot Ethereum ETFs recorded a net inflow of $9.23 million that day.
Key Data:
· Net inflow for the day: $9.23 million
· Total net asset value: $10.65 billion, ETF net asset ratio reaches 4.53% of Ethereum's total market capitalization
· Historical cumulative net inflow: $11.19 billion
By institution:
· BlackRock ETHA: A single-day net inflow of $11.75 million, the only Ethereum ETF to achieve net inflows that day, with a historical total net inflow of $11.42 billion
· Invesco QETH: Net outflow of $2.52 million for the day
· Other Ethereum ETFs (such as Fidelity FETH, Grayscale Mini Ethereum Trust, Bitwise ETHW, etc.) all had zero capital flows that day
Last week (July 20-24), Ethereum spot ETFs saw a weekly net inflow of $103.8 million, about three times that of Bitcoin spot ETFs ($33.9 million), leading Bitcoin-related products in weekly capital flows for the second consecutive week.
🔄 3. Capital Divergence: BlackRock's "left hand switching to right hand" pattern continues
The continued divergence in fund flows between Bitcoin and Ethereum ETFs has become the most prominent feature of institutional funds recently.
Inside BlackRock, there is a "left hand to right hand" phenomenon. On July 27, BlackRock Bitcoin ETF (IBIT) saw a net outflow of $8.82 million, while Ethereum ETF (ETHA) saw a net inflow of $11.75 million. The two funds are moving in completely opposite directions—institutional clients are withdrawing from IBIT and switching to allocate ETHA. Last week (July 20-24), IBIT had a net outflow of $95.5 million, while ETHA had a net inflow of $96.3 million over the same period, showing an almost identical outflow.
Ethereum ETFs have seen net inflows for three consecutive weeks. Since the end of eight consecutive weeks of outflows for both products earlier this month, Ethereum ETFs have maintained net inflows for three consecutive weeks, with inflow scale consistently outpacing Bitcoin ETFs.
Analysts point out that IBIT, as a highly liquid product mainly used by institutions to increase exposure or hedge, continues to see outflows indicating that institutions are actively reducing short-term Bitcoin exposure. Meanwhile, ETHA's continued fundraising shows that institutional funds are also seeking relative value opportunities for Ethereum.
📝 4. Summary
The ETF fund flows on July 27 sent out two key signals:
First, outflows from Bitcoin ETFs continue, but the scale is narrowing. For three consecutive days of net outflows (about $225 million on July 23, $240 million on July 24, and $11.64 million on July 27), the outflow rate has clearly slowed.
Second, the capital rotation between BTC and ETH continues to deepen. Ethereum ETFs have seen net inflows for three consecutive weeks, with IBIT outflows from BlackRock almost synchronized with ETHA inflows—this is not an exit from crypto assets, but a relative value reallocation between two major targets.
Key variables:
1. FOMC Decision (Early morning July 30, Beijing time) — If a hawkish signal is released, ETF funds may come under pressure again
2. Oil Price Direction — Brent has fallen from $100 to around $88, and whether it can continue to decline will affect inflation expectations and institutional risk appetite
3. Can ETH's relative strength be sustained—If the ETH/BTC ratio continues to rise, capital rotation may accelerate further
DataHunter | Understanding the market with dataHas AI crashed? Chip stocks plunged into a bear market overnight, but what truly crashed was actually the market's pricing logic!
Don't rush to say the AI bubble has burst yet.
Last night, the US semiconductor sector witnessed a "massacre."
The Philadelphia Semiconductor Index plunged more than 5% in a single day, pulling back more than 20% from its June high, officially entering a technical bear market.
SanDisk fell nearly 14%, SK Hynix dropped nearly 10%, Western Digital, AMD, ASML, Micron, Intel, and Nvidia ...... Almost none survived.
But what really deserves attention is not the decline, but the reasons for the decline.
Because while the stock price was plummeting, TSMC's profits surged by 77%, ASML raised its full-year guidance, Samsung's profits soared 19 times, and Intel delivered its strongest growth rate in over a decade.
While the performance kept improving, the stock price kept falling sharply.
This shows that the market is no longer killing on fundamentals, but on expectations.
There are actually only three reasons.
First, the market is shifting from "telling stories" to "looking at profits."
In the past two years, whenever AI is involved, even those who spend money like crazy have paid the price.
Now things are different.
Investors began to ask one question:
When exactly will AI make money?
When capital starts settling accounts, high valuations naturally take the lead in the slash.
Second, crowded trading has started to stamp on.
Almost all institutions are in the same boat—going long on chips.
Once someone exits first, quantitative stop-losses and leveraged liquidations are triggered simultaneously, causing funds to trample and the decline to be magnified infinitely.
Many times, it's not the company that falls, but the positions.
Third, the market is starting to debate cycles rather than demand.
Some believe storage prices have peaked, while others think inventories remain at historic lows.
The same data, different interpretations, ultimately escalated into an emotional game.
The more panicked, the easier it is to kill valuations.
But there is one detail that most people overlook.
Nvidia is still investing $1.5 billion to secure advanced packaging capacity ahead of time.
If AI demand really collapses, why would it still rush to seize capacity in advance?
The answer is simple:
Demand hasn't changed; what has changed is the way the market sets prices.
AI is not finished; it has simply entered the next phase.
In the past, whoever had GPU, HBM, or optical modules could see price increases.
In the future, whoever can truly turn AI investment into revenue and profit will continue to rise.
Therefore, this round of adjustments feels more like a screening process rather than a final closure.
For A-share investors, what's more important is not how much US stocks have fallen.
What you really need to ask is:
Is your company growing through emotional upswings or by performance?
In a bull market, profits are made by expectations.
Navigating cycles, what you earn is always real profit. ##$SNDK $BTC #$ETH Federal Reserve meeting on Thursday: No rate hike? No rate cut either?
Mainstream institutions mostly predict a rate cut in September!
For the first time, the market is uncertain before the rate decision—the probability of a rate hike is still 33.7%. It's no longer Powell's "giving advance notice" style; since Waller took office, the forward guidance has been shut down, do whatever you want.
Data is indeed cooling down: June CPI year-on-year at 3.5%, nonfarm payrolls slowing, and revisions downward for the previous two months. There’s no reason to rush a hike. CICC and Barclays both bet on "no change."
But trouble lies ahead: oil prices are pushed up by Middle East tensions, inflation is still far from 2%. More importantly, the market has already priced in over a 50% chance of a rate hike in September. Not moving this time doesn’t mean the risk is gone.
The focus is not on the outcome, but on the details:
· Will 2-3 officials dissent (supporting a rate hike)?
· If even Fed governors support a hike, that means the "core team is panicking."
· Will the statement mention energy and inflation resilience? Will Waller be tough in the press conference?
The harshest scenario: Castle Securities says the Fed might "surprise with a rate hike" tonight—better to flex muscles now than wait until September to suppress inflation expectations.
In short, this "pause" is very likely a hawkish pause. Don’t treat it as dovish good news and get slapped in the face. 🦅When the concrete curing period for load-bearing walls is forcibly terminated early, the load-stress curve of the entire building will collapse instantly.
Last night, crude oil prices recorded their largest single-day drop since 2020—WTI fell over 8%, and Brent plunged from above $100 to near $88. This is not a simple market correction; it is the plastic deformation of the foundation material caused by external impacts. Trump ordered a halt to bombing Iran, a "design change order," where load-bearing nodes originally labeled "sustained high pressure" on construction plans were suddenly required to reduce loads. The previous 13-day air raid cycle had already pushed the stress test value of the energy foundation up to a high of $93.83. Now, the blueprint on the policy blueprint has suddenly changed, predicting the market to price a ceasefire before August 31 at 75%—like structural engineers suddenly thinning the reinforcement layer at the mid-span of a floor slab, forcing the entire macroeconomic framework to undergo a new internal force analysis.
When we do architectural design, the biggest taboo is to carry out large-span cantilever construction when the "load concentration" is unstable. The price of crude oil determines the concrete grade of all assets—whether C30 or C60 determines the cross-sectional size of the beams and columns. Now that the C60 has been replaced by the C30, all load-bearing walls in the energy supply chain must be re-checked for yield strength. The US stock token $XMU acts like a prestressed beam in this structure, requiring high-grade concrete support. When the upper energy load drops sharply, its bending moment diagram deforms dramatically. Those retail investors who only look at project white papers or market value curves are like outsiders staring at architectural renderings—can a beautiful facade really cover up uneven settlement of the foundation? No, it is not allowed.
Every market forecast probability acts as a template for support; a 75% ceasefire probability means the support system has been partially unloaded. The current futures curve is the construction log—the delivery monthly price spread is narrowing, which is the real data of the foundation rebound. All long positions at $93.83 now have hydraulic jacks installed at the bottom of the main beam, suddenly removing the oil pump, causing the beam bottom deflection to instantly exceed limits. Brent's sliding from 100 to 88 is a warning sign that the structural elastic stage is transitioning to a plastic stage.
True architecture is not just about blueprints, but also about construction deployment. When the Foreign Ministry's negotiation venue starts to be painted with rust-proof paint, and when the military's cruise missile stockpile is marked as "detachable," you should realize: the temporary supports of this energy skyscraper are being dismantled. As for whether the $XMU beam can safely transfer loads, it depends on its own reinforcement ratio—the transaction volume of the underlying protocol, the constraint stiffness of the staking structure, and whether the development team has reserved expansion joints to handle such sudden displacements. #CeasefireHitsCrude 刚刚，不及预期 VS 深夜利好
存储巨头业绩不及预期，光模块巨头深夜利好。
今天早上，SK海力士发布二季度业绩，营收、利润双双不及预期。
SK海力士怕是欲哭无泪啊，又给了空头砸盘的理由。
二季度营业利润60.54万亿韩元、营收79万亿韩元，虽创下分季度历史新高，却双双低于市场一致预期的64.22万亿韩元、84万亿韩元。
有一种空头设置的陷井，叫：不及预期！虽然很多人都不知道，预期从何而来。
不及预期之后，又会产生“新的预期”，循环往复。
与之对应的是光模块巨头中际旭创深夜放出重磅组合拳。
董事长提议40亿-80亿元回购股份，同步披露投资者关系活动记录，直面市场传言，几条信息含金量十足。
1、40亿-80亿元回购提议，为光模块行业有史以来最大规模的回购计划。虽主要用于股权激励，但明确约定三年未使用完毕的股份将注销，也是真金白银的信心表态。
2、正面击碎1.6T光模块降价传闻。公司明确1.6T产品ASP远高于网传低价，行业不存在恶性价格战。当前原材料紧缺、交付紧张，仅有少数企业具备大规模供货能力，企业定价稳健，毛利率能够稳住。
3、订单能见度远超市场想象。现有客户订单基本覆盖2026全年，不少订单已经锁定至2027年，交付计划细化到月度。充沛在手订单。
4、长期产品路线清晰。800G、1.6T、2.4T以及NPO新品需求确定性高，增长速度可观，重点客户已经给出2028年新品采购指引，空间值得期待。
说到底，外围业绩不及预期还是情绪面的冲击，改变的只是估值预期，不是行业趋势。科技领域波动向来剧烈，但能穿越周期的，永远是手里有订单、业绩能兑现的公司。#韩股重挫8%，长鑫首日登顶A股 $SKHYNIX Rebound ≠ reversal, $ETH surged 4%, $QQQ was dazzlingly green, and the market was waiting—whoever showed weakness first would set today's tone.
Look at the numbers
$BTC 65,283 +1.45% $ETH 1,952 +4.14%
$QQQ -1.12% $SPY +0.10% $IBIT -0.82%
$DXY -0.15% $GLD +0.10%
Hormuz and crude oil are still adding variables to inflation expectations, while the shadow of US Treasury yields and Fed tightening continues to weigh on valuations. The dollar is not a backdrop; a simple adjustment of the exchange rate line can disrupt the rhythm of $QQQ$SPY. Today, it's not surprising if any switch gets touched on this plate.
$ETH is clearly more elastic than $BTC, short-term risk appetite is rising, but $QQQ is sinking downward, and money is shrinking into defense. $IBIT Weaker than spot $BTC, a weakness in ETFs means the spot market isn't that strong; $DXY Only when risk assets can breathe a sigh of relief can they catch their breath, but once tightened, they quickly turn hostile; $GLD Still quietly rising, haven't fully withdrawn safe-haven funds, don't be fooled by the surface buzz.$BTC Originally, the market was betting on expectations of a US-Iran diplomatic easing, but the situation suddenly escalated into tough military action! The U.S. military, together with Saudi Arabia, launched precision airstrikes directly targeting Iranian proxy forces. On one hand, tonight the Federal Reserve is making a major interest rate decision, while on the other, tensions in the Middle East have resurfaced. With both crude oil inflation expectations and risk appetite being disturbed, the crypto market is prone to severe volatility—beware of repeated waves of trading in the news! 1. Original Information Compilation | Bijie Network News The U.S. Central Command issued a statement saying that on July 28 local time, the U.S. military and Saudi Arabian armed forces carried out precision strikes in Iraq, targeting pro-Iranian militias commanded by the Islamic Revolutionary Guard Corps (IRGC) that had previously attacked U.S. forces and Saudi energy infrastructure. In the past 72 hours, Iran's Islamic Revolutionary Guard Corps commanded more than 30 drone air strikes. This joint airstrike was a strong response, with warplanes striking multiple armed logistics and weapons bases in eastern Iraq. The U.S. side warns: Iran's Islamic Revolutionary Guard Corps and its proxies must cease their attacks, or the U.S. will take further military action. 2. Breaking down ✅ the core market transmission logic and interpreting the positive crude oil scenario. This joint airstrike marks an increase in conflict intensity. The market is concerned that Iran will retaliate, targeting Gulf energy facilities and disrupting shipping in the Strait of Hormuz. Once risk aversion heats up, the crude oil geopolitical risk premium will rise again. ⚠️ Chain macro pressure: rising oil prices→ rising expectations of imported inflation→ strengthening expectations for delayed rate cuts, which will suppress highly elastic risk assets like BTC and ETH in the medium to long term. Why is SanDisk still able to drop 15% in one day?
SanDisk recently closed at $1278, down 11% in a single day. However, there has been no major negative announcement from the company so far; the real test will be the earnings report on August 5.
Last quarter, SanDisk's revenue was $5.95 billion, with a gross margin of 78.4%. The data center business grew 233% quarter-over-quarter, indicating strong fundamentals. The problem is that the stock price had risen too much in advance, and the market is now worried not about whether it is profitable, but how long it can sustain a gross margin close to 80%.
The IPO of Changxin Storage is just the trigger. Changxin mainly produces DRAM, while SanDisk focuses on NAND, so they are not direct competitors. However, the rise of China's storage capacity will indeed cause the market to reassess the entire industry.
In the short term, watch whether the $1220–$1250 range can hold. On August 5, key points to watch are gross margin, data center revenue, and the 2027 order guidance.
If it holds, this looks more like a valuation cooldown; if it doesn't, the market may have already started pricing in the peak of the storage cycle.
$SNDK $BOT After the three leading platforms launched intensively, spot liquidity expanded significantly. The core issue now is whether the buying pressure generated by liquidity injection can absorb the profit-taking pressure from the concentrated chip turnover.
$BOT In early July, it became the first leading platform to open spot trading, changing the previous judgment of lacking mainstream spot depth. On July 9, the second platform opened its trading channel, changing the judgment that the liquidity carrying cap for a single port was reversed. On July 10, the third platform followed suit, completing the expansion of three trading venues within 10 days, changing the market's judgment of the chip turnover cycle and the continuity of buying orders.
The primary driving force on the market is the reduced slippage and net capital inflow from deepening order books on exchanges, followed by the growth in copy trading demand driven by the recovery in market volatility, which is RoboStrategy's automated strategy track. The expansion of trading venues has transformed the accumulated attention into spot buying traffic.
The trigger for the upward scenario is that the three platforms continue to deepen their spot trading and maintain stable turnover rates. If incremental funds continue to absorb profit-taking from initial launches, the low slippage environment will attract more quantitative copy trading funds to enter; This script failed signal: a rapid contraction in spot trading volume and a severe gap in thin buy orders.
The trigger for the downward scenario is that after positive news from exchange listings is realized, profit-taking positions are concentrated in sell-off, while the automation strategy tool sector loses enthusiasm, leading to insufficient buying demand. If the buying depth cannot absorb the outflow, the price will face downward pressure; This script failed signal: large buy orders are continuously placed in key support ranges to support the market.
The overall simulation fails if spot liquidity at the three trading venues drops sharply after concentrated injection, or market volatility narrowing leads to an overall stagnation in strategy copy trading demand.
The most important variable to watch in the next seven days is the depth distribution of buy orders in spot order books across the three platforms and the persistence of trading volume after chip turnover.
#美联储周四凌晨公布利率决议 #以太坊验证者退出队列已降至零 #Storj Labs files for Chapter 11 bankruptcy restructuring, STORJ plungesMicron Technology (MU) Bearish Logic and Technical Analysis (Original by Boshen, as of 2026.7.29)
Risk warning: The following is based solely on publicly available market and industry data and objective inference, and does not constitute any short-selling or investment advice. The volatility in US semiconductor stocks is extremely high, so caution is needed regarding multiple risks.
In the past month (6.29-7.28), Micron fell from $1,145.28 to $820.53, a cumulative decline of 27.54% over the period. This round of AI storage bull market signals a temporary peak, indicating a strong medium- to long-term bearish outlook. From a fundamental perspective, the company's current gross margin of 84.9% is at a historic high in the storage industry. The strong cyclical nature of storage means high profitability cannot be sustained. Samsung and SK Hynix continue to expand production, and combined with the release of domestic Changxin Storage capacity, supply and demand will ease rapidly in 2027, with significant downward revisions in storage prices and corporate earnings. At the same time, leading cloud providers' AI capital expenditures are approaching their cash flow ceilings, downstream procurement demand growth is slowing, and combined with geographic restrictions on business in China, performance expectations have been loosened early; Recently, institutions have bought a large amount of put options, with internal executives continuing net selling, and the risk aversion of funds is evident.
Technically, the bearish structure is fully established. The moving average system is in a bearish alignment, with the stock price under continuous pressure on the 20-day and 50-day moving averages. In the short term, a rebound that touches the moving averages immediately triggers heavy selling pressure. In July, multiple rebounds to around $950 all fell back quickly, forming a strong resistance zone between $954 and $1012. On the indicator side, the daily MACD maintains a death cross, the green bars continue to expand, and bearish momentum remains intact; The 14-day RSI fell to 29, approaching oversold, but after each slight rebound, the RSI only rebounded to 55 before turning downward, showing no bullish reversal momentum. Volume and price continue to diverge, with shrinking trading volume during the rebound phase. On the day of decline, volume surged and the price plunged. On July 28, the single-day turnover was $50 billion, with a sharp drop of 8.85%, confirming that selling pressure has not been cleared.
The candlestick pattern has broken out of a standard downward channel, with the high continuing downward. The June all-time high of $1255 formed a long-term top, and each small rebound serves as a window for funds to exit the market. If the short-term support level of $800 is breached, the lower target is $737, with more than 10% downside potential. Considering the three signals of cycle inflection, capital flight, and bearish technical patterns, the current rebound presents a good short-selling opportunity. If the rebound reaches the $910-950 resistance zone, short positions can be positioned, with stop-losses set above $990 to avoid short-term oversold recovery and rebound risks.SK海力士这份财报真是把“预期管理”玩明白了。
营业利润同比暴增5倍多，净利润更是翻了12倍，这放在任何行业都是炸裂的成绩单。但市场居然不买账，股价还跌了？
核心逻辑其实很反直觉：大家以为HBM是功臣，其实是通用DRAM在疯狂涨价。而SK海力士为了绑定英伟达等大厂，签了3-5年的长协（LTA），锁死了大部分出货价格。
结果就是，当现货市场DRAM和NAND价格飙升时，它只能眼睁睁看着别人吃肉，自己因为长协限制，吃不到这波涨价的最大红利。用短期的价格弹性换取长期的确定性，这在上涨周期里确实容易让资本市场感到“不过瘾”。
不过话说回来，CEO说存储紧张要持续到2030年，HBM4也已经量产。这种“细水长流”的确定性，或许比短期暴涨更考验持有者的耐心吧。
#SK海力士二季度业绩#韩股重挫8%，长鑫首日登顶A股 $SKHYNIX