Orbit Post Sitemap
#苹果公司市值重回全球首位, surpassing Nvidia, Apple reclaims the top spot: The AI track has shifted from "selling shovels" to "selling brands." On July 27 Eastern Time, Apple's stock price hit a new all-time high, with its market value surpassing $4.95 trillion, officially reclaiming the world's top market value from Nvidia. It has been exactly 15 months since Apple last reached the top. Even more dramatic, just a month ago, Apple lost over $260 billion in market value in a single day after announcing price hikes for Macs and iPads. In just a few weeks, it completed a "V-shaped reversal," with a year-to-date increase of 24.55%, compared to Nvidia's rise of only 5.50% during the same period. The ebb and flow of the market represents a fundamental shift in AI investment logic. Over the past year, Nvidia has soared with its status as the "shovel seller" in AI chips, with its market value once surpassing $5 trillion. But now, Wall Street is beginning to worry about the sustainability of massive capital expenditures—Nvidia is reportedly negotiating financing guarantees of up to $250 billion for OpenAI's data center project, a "circular financing" model that has made investors increasingly cautious. In contrast, Apple was once questioned for its conservative AI strategy, but now it has become a safe haven due to its "asset-light AI approach." Apple doesn't spend heavily to build its own computing power, but instead relies on external models and on-device ecosystems to shift cost pressure onto consumers, which actually strengthens profit expectations. The market trend is shifting from chasing "computing infrastructure" to favoring application giants with strong brand premiums and ecosystem stickiness. Apple's July 30 earnings report⚡ My objective review: I bottom-fished SanDisk in the morning session on July 29, but the overall logic didn't hold up
1. I believe the most critical issue: We are currently in an extreme wait-and-see window before the Federal Reserve's rate decision, with unified risk avoidance across the board, and there is no bottom-fishing environment for high-volatility small-cap growth stocks.
The interest rate decision will only be announced in the early hours tonight, and the extent of the rate cut and the direction of monetary policy remain uncertain.
All funds are withdrawing from high-valuation stocks like computing power and storage, concentrating in banking and consumer defense sectors.
The brief slight rise in the morning was merely an oversold technical impulse, not a trend reversal. Entering now would mean actively taking on selling pressure.
2. My observation of the market trend: SanDisk's complete downward channel has fully opened, and the one-sided downtrend has not ended 📉
The stock price fell over 51% for the entire month of July, being cut in half and continuing a streak of decline for several consecutive days.
After every small rebound, a massive wave of shares to break even at high levels and take profits is triggered by dumping shares.
Blindly bottom-fishing and taking the knife in a downtrend will only deepen the trapped positions, with no safe bottom support.
3. My interpretation of fundamental gaps: SanDisk is the weakest stock in the storage sector, with very little AI dividend realization, and its resilience is far inferior to Micron and SK Hynix.
Most of the company's revenue comes from weak consumer-grade USB drives and regular SSDs, with demand for mobile phones and PC terminals continuing to shrink.
It cannot rely on HBM high-end memory and long-term computing power orders from cloud vendors to stabilize performance, making it the largest bubble in the sector.
When sector sentiment warms up, it leads in gains; once the market weakens, the pullback always leads the entire sector.
4. The current market pricing logic means that early bottom-fishing leaves no profit margin 💡
Currently, the secondary market is trading expectations of forward overcapacity in 2027, rather than the current flash price hikes and high corporate profits.
Even though spot storage remains tight and major companies lock in orders for years, institutions continue to cash in at high levels.
Good news comes with negative news; the slight rebound in early trading was just a temporary exit window for stranded funds, not a buying opportunity.
5. The structure of small-cap chips has significant risks, with completely uncontrollable volatility
SanDisk's circulating chips are loose, with a turnover rate hovering around 18% year-round, and short-term quantitative funds are used to buy back and cut off.
During the panic downturn, without major funds to support the market, once market sentiment turns negative again, the stock price will quickly hit new lows.
Without stabilized candlesticks or signals of shrinking volume and stopping the decline, bottom-fishing at any point carries extremely high risk of loss.
6. Two future market scenarios: bottom-fishing profits and losses are completely unequal
Even if the Fed cuts rates as scheduled, SanDisk will only experience a brief oversold rebound for a few days, and after the rebound, the downward trend will continue;
If the rate decision is hawkish and the rate cut falls short of expectations, the storage sector will enter a new round of sharp declines, and bottom-fishing positions will be deeply trapped.
The profit-loss ratio is severely imbalanced, making it a completely unprofitable trade.
7. My final practical viewpoint
Before the downtrend stabilizes and major macro events take place, waiting and waiting is always better than bottom-fishing.
For consumer flash memory stocks like SanDisk, every rebound is an opportunity to reduce positions at high prices, never a buying opportunity. $SNDK #美联储即将公布利率决议 All on-chain indicators point to the bottom of Bitcoin $BTC approaching, and the countdown to the cycle reversal has begun. Bitcoin's price follows a cyclical pattern of bull and bear cycles. At the beginning of the decline, everyone is guessing the bottom price, but once the decline stabilizes, market attention shifts to when the rebound will begin.
According to estimates, this round of major bottoms is most likely to occur in mid-August, with an upward turning point in late September, with the timing and rhythm basically matching the historical patterns of the halving cycle. Market sentiment is the core driving Bitcoin higher, and the FOMO sentiment from missing out is brewing momentum for going long.
However, during the upward movement, regulatory negative factors are inevitable, and the timing coincides with the US midterm elections. Negative news during the election cycle is likely to be temporarily shelved, and the overall external environment is favorable conditions.Guys, after reviewing the situation, the last two days of bottom-fishing will be lucky tonight. Why:
1. The leverage in Korean stocks is still very high, it's really frightening
2. Wash's speech tonight is highly volatile; guard against unreasonable rate hikes or extreme risk from the hawks
3. This bottom-fishing move was originally about fast entry and exit, without a strategic view
Specific Target Operations:
$MU The short-term range is 800-880, which is the put wall and call wall. The left side is support, the right side is resistance. I'm not too worried about breaking below 800 tonight, so try to sell near 880
$SNDK 1100 is a very thick put wall. Before the 31st, it will fluctuate around 1100. Market makers will stabilize the price, and there is rebound demand tonight. If it rises, it will likely move first between 1150-1200$AEON plunged today, mainly due to profit-taking after the previous surge combined with multiple negative factors resonating:
· Exchange delisting: BitMart announced the delisting of AEON as early as June 5, with withdrawal channels closing on August 5. The liquidity and confidence shock caused by the delisting is a continuing negative factor.
· Short-term speculation fading: Previously, AEON surged due to listing on Binance Alpha and airdrop hype. Analysis suggests this was more of a short-term event-driven spike, and after the hype fades, a value correction is inevitable.
· Airdrop and mining sell pressure: AEON was listed on multiple Launchpools with a total prize pool of 1.16 million AEON. Many "farmers" chose to dump tokens to cash out after receiving them.
· High leverage contracts: Exchanges like Bitget launched perpetual contracts with up to 20x leverage. High leverage amplifies losses during declines, accelerating market crashes.
Additionally, the AEON project itself is positioned as an "AI agent payment settlement layer." Although the narrative sounds good, it lacks actual adoption data support, making it difficult to maintain price at high levels. SK HYNIX IS CRASHING BUT AI DEMAND IS NOT THE REAL PROBLEM. 🧵
Record earnings could not stop $SKHY and $SKHYNIX from falling because expectations, leverage and ADR premium risk all broke at the same time.
1/ SK Hynix reported record Q2 revenue of KRW 79.3T and operating profit of KRW 60.5T.
But revenue and profit still missed extremely high market forecasts. In today’s AI trade, record results were not enough.
2/ Main reasons behind the decline:
• Slower-than-expected HBM4 revenue
• Concerns about future AI capex
• Profit-taking after the Nasdaq listing
• Leveraged ETF and margin-call liquidations
• Chinese memory competition
• No detailed shareholder-return plan yet
3/ Important difference:
$SKHY tracks the U.S. ADR.
$SKHYNIX tracks the Korean-listed share.
One SKHY ADR represents one-tenth of a Korean common share. At the current snapshot, SKHY was still trading at a large premium to the Korean-tracking contract.
That premium creates extra downside risk for SKHY.
4/ Key $SKHY levels:
Support: $128–$130
Next zone: $118–$122
Premium-compression risk zone: $98–$105
Resistance: $136–$140
Major resistance: $149–$155
Higher targets: $168–$170 and $194–$195
5/ Key $SKHYNIX levels:
Support: $950–$1,000
Major support: $850–$880
Resistance: $1,050–$1,100
Major resistance: $1,200–$1,280
Higher resistance: $1,400–$1,500
6/ Recovery timeline:
A relief bounce could happen within days.
A proper base may require 2–6 weeks.
A full recovery toward previous highs will likely take months and require stronger HBM4 shipments, stable AI spending and the end of forced liquidations.
This looks more like a violent leverage reset than the end of the SK Hynix AI story but recovery is not confirmed until major resistance is reclaimed.
#SKHynixRecordMiss $SKHYNIX $SKHY #美联储即将公布利率决议
⚠️The big show is on tonight! Massive volatility is on the way! Will $BTC and $ETH face a sharp dump?
The Federal Reserve interest rate decision will be announced promptly at 2 AM, followed by Chair Powell's press conference at 2:30 AM!
Here’s Dragon Lady’s judgment: the probability of a direct rate hike is actually not high. Although oil prices have recently dropped significantly, they remain at a high level overall, and inflation risks still loom overhead.
So, it’s very likely that rates will be held steady, but hawkish rhetoric will be used to suppress a price rebound—this point must be highly watched!
So how should we respond to the market under these circumstances?
Let me share a recurring historical pattern in the crypto space that has been proven true seven or eight times 😂
Before the Fed meeting, the market often preemptively speculates on expectations, causing prices to rise first; once the meeting concludes, a 1-2 week sustained correction and sell-off often follow.
Looking at the current market, the pre-meeting rally has already happened solidly, so the likelihood of history repeating itself is very high.
If the speech signals hawkishness, the risk of a downward correction after the meeting will further increase!$DATA Currently quoted at $0.2232, down 6.88% in 24 hours, with a intraday high of $0.2434 and a low of $0.2226, a recorded amplitude of 0.0%. According to real-time OKX data, trading volume is close to zero. Single-day volatility is nearly stagnant, trading volume has completely shrunk, and this kind of bottom-selling market with no volume often signals liquidity withdrawal that is far more alarming than the price itself. On the giant screen in the financial center, the cold-toned candlesticks seem frozen in creative vision, lacking any urge to rebound. From an on-chain structural perspective, the MVRV ratio has deeply fallen to around 0.82, indicating that the vast majority of short-term holders are in floating losses. Historically, when MVRV remained below 1 with no signs of rebound, passive lock-up in chips further suppressed buying interest. The expenditure-to-output profit ratio (SOPR) recorded 0.94, with the average sold tokens shifting to losses, indicating that even in a low turnover environment, exit funds still choose to admit defeat rather than wait. If this phenomenon weakens in sync with prices, it usually lacks the signal value of a bottoming divergence and instead indicates that capitulation selling has not yet been fully released. The on-chain token distribution of URPD shows that the $0.2250 to $0.2350 range has seen dense turnover, with the current quote just below the lower boundary of this cluster. Once this level is confirmed, these chips will turn into a recent resistance wall above, and a short-term rebound to around $0.23 will face selling pressure. In terms of exchange balances, net inflows in the past 48 hours rose slightly by 0.7%. Although the absolute value is not large, considering the trading volume is almost zero, any small net inflow in any direction could lead to a one-sided market outflow. Clearly, at this moment, the bulls' strength is almost empty, with no funds to take over. Combined with candlestick patterns, $DATA hit resistance at $0.2434 and closed bearish continuously, with almost no effective buying support during the session. Volatility was compressed to an extreme state, which usually easily triggered a downward pulse breakout. There is no obvious technical support below; if the psychological level of $0.22 is breached, the probability of a slide toward $0.20 is high. For small-cap stocks like $DATA, the technical structure when trading volume dries up is often fragile, and the bearish direction is clear, but it is important to note that placing orders under liquidity traps carries extremely high risk. The only rational judgment given by the market is to avoid risk and wait and see; any left-side bottom-fishing lacks data support. This does not constitute investment advice. Tuo-ge On July 27, 2026, a Chinese internal chip company was listed on the Shanghai Stock Exchange, with its stock price soaring 466% on its first day. Within hours, SanDisk fell 12%, Micron dropped 5%, Western Digital dropped 7%, and SK Hynix's American Depositary Receipts dropped 6%. This report will tell you who CXMT is, what they do, why Wall Street is paying so much attention, and how to think about what all this might mean for the memory stocks you already hold. 1. What Happened On July 27, 2026, CXMT Co., Ltd. (Changxin Memory Technology Co., Ltd.) surged 466% on the first trading day of the STAR Market on the Shanghai Stock Exchange—from an issue price of 8.66 yuan to a closing price of 49 yuan, pushing the market value of China's largest memory chip manufacturer to the highest point on China's A-share market in a single trading day, surpassing Industrial and Commercial Bank of China to become the company with the highest market capitalization listed on the Chinese mainland stock exchange. The intraday high reached 55.03 yuan, then fell back to close at 49 yuan. The figures behind this listing are equally impressive. Before the green shoe mechanism was implemented, CXMT raised 57.92 billion yuan (about 8.6 billion USD), becoming the second largest domestic IPO in Chinese history (second only to Agricultural Bank of China's 2010 IPO of about 10 billion USD), the largest IPO in STAR Market history, and the largest IPO in Asia in 2026. Institutional subscription multiples exceed 500 times, while retail investors have high subscription multipliers Minnesota didn't become a prediction market, and this is really related to Robinhood $HOOD. Minnesota originally planned to ban prediction markets starting August 1, but before the law even took effect, a federal judge blocked it. For Robinhood $HOOD, the fastest-growing product line in company revenue is precisely the forecasting market. Robinhood did not participate in this lawsuit. Minnesota was sued by the CFTC, with Kalshi and Polymarket later joining. On July 27, the judge issued a preliminary injunction, believing they had a better chance of winning. While the lawsuit continues, the state government is temporarily unable to enforce this injunction. Why is Robinhood involved? Its event contracts are provided by Robinhood Derivatives and are traded on CFTC-regulated cooperative exchanges. Minnesota wants to ban such products with state-level gambling laws, and Robinhood is also following the path of federal regulation. The judge temporarily accepted the CFTC's position, so Robinhood naturally had one less trouble. This business is already quite large. According to Robinhood's May data, users traded a total of 3.9 billion event contracts, a 22% increase over April; Daily average was 126 million copies, an increase of 18%. Here, 3.9 billion refers to the union👀 Everyone is watching ETF headlines, but what institutions really care about is the underlying foundation.
In the past few weeks, two major events have quietly taken shape:
✅ Ripple has obtained the full MiCA CASP license, covering 30 countries across the European Economic Area, enabling it to provide compliant crypto asset services in these countries.
✅ Clearstream (a subsidiary of Deutsche Börse) has officially expanded its regulated crypto custody platform, adding XRP and XLM beyond BTC and ETH.
🎯 Most people overlook the key perspective: this is far more than just another "bullish" piece of news. This is a silent infrastructure build—custodial, compliant, and regulated settlement channels. Large financial institutions must wait until this system matures before deploying large sums of money.
📊 So the next confirmation signal worth tracking isn't FOMO on social media, but rather:
- Actual growth in institutional usage
- New banking integration and custody adoption
- Measurable asset inflows through regulated channels
The real signals are hidden there, not just momentary emotions.#EarningsObserver: Microsoft, Meta, Amazon Report Tonight
The AI big test is due tonight, and the market has no patience for stories.
Last week's Google earnings were a mirror—an 82% growth in cloud business is impressive, right? Yet, just because of an upward revision in capital expenditure guidance, the stock plunged 7% in a single day. The market signal couldn't be clearer: the AI narrative phase is over, and the era of capital efficiency scrutiny has arrived.
Here are the key points I think the three companies should focus on:
Microsoft: Azure growth is the lifeline
Market expectations are about $87.7 billion in revenue and $4.24 EPS. But the real focus is only one thing: can Azure's constant currency growth maintain the 39%-40% guidance range? Falling short means missing expectations and stock price pressure. Whether Copilot paid seats can unexpectedly surpass 30 million is a bonus. More critically, the FY2027 capital expenditure guidance—if significantly above $220 billion, free cash flow pressure will become a stock price killer.
Meta: No matter how strong ad revenue is, it can't withstand the burn rate
The advertising base is solid; Q2 ad revenue is expected to surpass Google's search ads for the first time, a historic milestone. But what really keeps the market awake is capital expenditure—Q2 free cash flow may record a negative value exceeding $1 billion, and full-year free cash flow could plunge from $43.6 billion in 2025 to less than $1.9 billion. If this time the capital expenditure ceiling is raised from $145 billion to $150 billion, the stock price will likely replay the last after-hours 7% drop.
Amazon: Whether AWS growth can break through 30% is the only variable
As the last of the four cloud giants to report, whether AWS growth can accelerate from 28% in Q1 to above 31% will directly determine market sentiment. But capital expenditure pressure is also not to be ignored—the market expects Q2 capital expenditure to rise to $48.7-$49.1 billion, and if the full-year guidance is raised above $210 billion, free cash flow will face further pressure.
My personal view: Tonight's test is not about performance but capital discipline. Whoever can clearly explain "where the money is spent and when it will break even" will survive; otherwise, Google's script will repeat, and storage will still lack expected rebound.
$SNDK $SKHYNIX $MU Russia has taken action! The earthquake in the crypto world is happening faster than expected.
Just yesterday, the Russian central bank dropped a bombshell—the country's first draft cryptocurrency regulation was officially released. This is not a test, not a rumor, but a complete set of "rules of the game" that is about to be implemented.
Why do you say this matter is more important than you think?
First, this is a "financial breakthrough" under sanctions.
On July 21, the Russian State Duma passed the "Digital Currency and Digital Rights Law"; On July 24, the EU announced its 21st round of sanctions, targeting 14 crypto companies; On July 27, the Russian central bank released a draft regulation. The timeline is so tight—every inch the West tightens, and Russia takes a step forward. This is no coincidence; it is a strategic hedging.
Second, the threshold is astonishingly high, and compliance costs are astronomical.
The draft requires all digital custodians to have a registered capital between 50 million and 250 million rubles—equivalent to nearly 3 million US dollars at most. Moreover, this capital must be highly liquid and high-credit assets. Want to use low-quality assets to make up the debt? There was no way to get started. Exchanges must develop their own rulebooks to calculate market prices and weighted average prices in real time. This is using the harsh standards of traditional finance to put a "compliance shackle" on the crypto industry.
Third, opportunities and limitations for retail investors.
Under the new framework, non-accredited investors can only purchase mainstream assets such as BTC, ETH, and USDT annually, with a cap of only $4,000. Retail investors are protected, but also restricted. Industry participants will enjoy a transition period until March 1, 2027.
What is most thought-provoking is that Russia is building a completely independent "on-chain + traditional finance" dual-track clearing system independent from SWIFT. The operating principles of digital custodian institutions are fully consistent with those of traditional securities custodians—crypto assets are being incorporated into the framework of national financial sovereignty.
On September 1, this framework will officially take effect. Time left for the market is running out.
You and I are witnesses to this upheaval.BTC is stuck around $63.4K $64K, and this is not just a chart story.
The Fed is very likely to keep interest rates at 3.5 3.75% today, meaning liquidity remains tight. Meanwhile, US-Iran tensions pushed oil up to $74.67, causing the market to worry about inflation returning.
BTC once fell from $72K to $63K when the conflict escalated. Altcoins will suffer more if the Fed continues to be hawkish.
Crypto is not a safe haven yet. Currently, it is trading like a risk asset.$BEAT Holding position keeps increasing. Every time this project team unlocks, they pull the price up once
This demon coin always pulls up before unlocking, and this time is no exception
It's already up to 50%. Will there still be room for improvement?
Now on-chain funds are starting to flow out, and at the peak, they still want to lure in a short wave
Personally, I still think it's at the top. First, on-chain funds have already started to flow out, plus selling pressure near $4. So at the current 3.8 level, you can take a light short position. Just set your stop-loss at $4
#美联储即将公布利率决议 $BTC $ETH #财报观察员: Microsoft, Meta, and Amazon will deliver their results tonight On Tuesday, the US stock market showed strong sector rotation.
The core theme is "tech stocks squeezing out water and the return of traditional blue-chip performance."
On one hand, strong Q2 earnings catalyzed significant capital inflows into Dow Jones value blue-chip stocks and traditional consumer/industrial sectors; On the other hand, the semiconductor and AI core hardware chains have experienced a sharp global correction.
Meanwhile, crude oil prices continued to retreat from their highs, further easing market concerns about inflation and the Fed's interest rate path.
As the Federal Reserve's FOMC meeting is underway (with the decision to be announced on Wednesday), market funds are shifting heavily to low-valuation sectors supported by earnings such as consumer and industrials on the eve of key macro moments and tech giants' earnings reports.
#美联储即将公布利率决议 BTC is stuck around $63.4K $64K, and this is not just a chart story.
The Fed is very likely to keep interest rates at 3.5 3.75% today, meaning liquidity remains tight. Meanwhile, US-Iran tensions pushed oil up to $74.67, causing the market to worry about inflation returning.
BTC once fell from $72K to $63K when the conflict escalated. Altcoins will suffer more if the Fed continues to be hawkish.
Crypto is not a safe haven yet. Currently, it is trading like a risk asset.#交易之声: Your experience deserves to be heard. Is a US stock going to be the first to rebound? Looking back, I saw that the $spcx in the US stock market broke its bottom after last night's opening, and did not follow the decline before today's open, giving the impression that a strong rebound is about to take the lead. Since its IPO, SPCX has pulled back 53% from its high. Last night's rebound after the drop indicated short-term selling pressure has exhausted, and some funds may be playing for key events on the left side ahead of time. Next month, SPCX will have two major first-time events: 1. SPCX will release its first earnings report since its IPO (Q2 earnings report on August 4). Currently, the market is divided on the true profitability of spcx's AI and space hardware (xAI, Starlink, and launch business). If the financial report can prove its top-tier profits and revenue, it would be a major positive sign, leading to a strong rebound. 2. But there is also some bad news: on August 6 (the third day after the earnings report), the IPO shares were unlocked. This is the first unlock since the IPO, and early investors will be allowed to sell on that day. This is a signal of selling pressure. But what matters most is the financial report. If the report is good, selling pressure is less; if not, the massive unlocking of chips could trigger a chain of selling pressure. Additionally, the 13th Starship launch window is approaching, and any unexpected success of Starship will be directly reflected in market performance. From a technical perspective, a breakout and reversal is a signal. During the day, you can take the lead position early. If fundamentals at night do not break below the new low and support around 110 is stable, then this is the best opportunity to position long. Enough talk, I'll hold back from bottom-fishing$ZEC （Zcash）作为本轮隐私叙事行情的核心龙头币种，前期从368美元一路冲高至588.70美元阶段高点，如今走出连续性回落行情，当前报价458.98美元，单日跌幅4.07%，7日累计下跌9.66%，即便盘面放出矿场扩建、矿企筹备上市的行业利好，依旧没能止住下跌势头。 很多交易者十分困惑，明明有利好消息加持，曾经的隐私币龙头为什么还在不断阴跌走弱？结合隐私赛道行情变化、矿工筹码行为、大盘资金流向拆解本轮下跌逻辑，同时梳理$ZEC 本身独有的币种特征。 一、本轮冲高之后持续下跌对应的市场事件背景 1. 隐私币集体炒作行情落幕，题材热度全面退潮 前段时间市场集中炒作隐私支付、链上匿名交易叙事，ZEC作为老牌合规隐私代币被资金抱团拉升，透支消化了后续上涨空间。随着市场资金流向切换到存储题材、Meme热点，隐私赛道不再成为场内资金主攻方向，没有新增增量资金接力接盘，前期获利盘开始分批止盈离场。 反观同期一众二线隐私小币率先崩盘，龙头ZEC只是下跌节奏更平缓，终究逃不过题材退潮带来的估值回调。 2. 矿企扩建矿场+上市利好落地，走出典型“利好兑现出货”行情 本次推送的消息：Zcash头"DataHunter Crypto Research Report" · July 29, 2026
Understanding the Market Through Data
📊 1. Market Overview
BTC currently at 64,466 USDT, +1.46% in 24 hours. Intraday low at 62,742, high at 64,745, a range of about $2,000, with volatility significantly increased. ETH at 1,921 USDT, +1.1% in 24 hours, slightly underperforming BTC.
Fear and Greed Index: 29 (Fear), unchanged from yesterday, remaining in the fear zone for several consecutive days.
In the past 24 hours, total liquidations across the network reached $522 million, with long liquidations at $236 million and short liquidations at $286 million. Shorts have begun to experience a certain scale of forced liquidation, indicating market sentiment is shifting from a "one-sided long liquidation" to a two-way battle between bulls and bears.
📍 2. Market Trend
BTC formed a "V-shaped reversal" pattern today. The Asian session continued weakness, with a sharp drop to 62,742 in the afternoon, marking the lowest since July 14. However, bulls quickly entered, pushing the price up over $1,700 consecutively, reaching a high of 64,745, currently consolidating near 64,400.
Daily level: Price has risen above MA7 (64,450) and MA25 (64,274), short-term moving averages reclaimed, indicating strong rebound momentum. Volume shows a clear increase corresponding to the candlestick, indicating this rebound is not a "fake rally" caused by short covering but supported by genuine buying.
Indicators: MACD on the 4-hour chart shows a golden cross signal, DIF and DEA converge in the negative zone near -82, with green bars continuously shortening. RSI(6) rises to 49.39, just crossing the neutral line, showing bullish momentum is accumulating but not yet fully dominant.
Key levels:
· Resistance above: 64,745 (today's high), 65,000 (round number), 65,800-66,000
· Support below: 63,500-63,800 (buying on dips zone), 62,742 (today's low, validated), 62,000
🌍 3. Core Logic Behind the Rebound
BTC's rapid rise from 62,742 to above 64,700 is mainly driven by:
First, the sharp drop in oil prices easing inflation concerns. Brent crude has fallen over 16% in three days, dropping from above $100 to around $88. The decline in energy prices directly reduces market worries about a secondary inflation surge, giving the Fed more room to "hold steady."
Second, the FOMC "sell the rumor, buy the fact" scenario is playing out early. The market had already priced in rate hike risks—BTC fell from 66,800 to 62,742, a drop of about 6%. When the price approached 62,700, shorts took profits while longs bought the dip, creating a resonant rebound. Some traders began to speculate in advance on a "no rate hike" outcome.
Third, the drop in U.S. Treasury yields provides relief. The 10-year Treasury yield fell from 4.71% to around 4.60%, lowering the risk-free rate and supporting risk asset valuations.
⚠️ 4. FOMC Decision Countdown: 6 Hours
At 2:00 AM Beijing time on July 30, the Fed will announce its rate decision. CME data shows a 69.5% probability of holding rates steady and a 30.5% chance of a 25bp hike.
Scenario 1: Hold rates steady + hawkish statement (highest probability)
The 16% drop in oil prices over three days provides ample reason to hold steady. However, the statement may emphasize "inflation remains above target" and "a strong labor market," keeping the option of a September hike open. The market will likely "sell the rumor, buy the fact" for a rebound, but gains will be limited.
Scenario 2: Hold rates steady + dovish statement (medium probability)
If Waller signals dovishness, such as acknowledging "energy price declines improve the inflation outlook," BTC could quickly rebound to 65,500-66,000 or higher. But given Waller's historically hawkish stance, this scenario is less likely.
Scenario 3: Unexpected 25bp rate hike (about 30% probability)
Risk assets would be heavily sold off. BTC could quickly fall back to 62,000 or even 60,000-61,000. Citi calls this "the most divisive moment since September 2024," and Castle Securities has already bet on this direction.
📝 5. Trading Framework
With less than 6 hours before the FOMC decision, the best strategy is to wait for a clear direction before acting.
Before the decision: Avoid chasing longs above 64,400. Those without positions can wait for the FOMC outcome before deciding. Long holders may consider partial profit-taking in the 64,500-65,000 range while keeping a base position for further play.
After the decision—if hold steady + hawkish: BTC may briefly surge to 65,500-66,000, but hawkish wording will limit the rebound, so consider reducing positions on the rally. If the statement is neutral to dovish, BTC could test 66,000-67,000.
After the decision—if unexpected hike: Risk assets will be heavily sold off; stay out and observe. If BTC falls to 61,000-62,000, consider scaling in gradually; this is the most cost-effective entry zone.
Not trading before a clear direction is also part of trading.
Risk warning: This article is a research note and does not constitute investment advice.
DataHunter | Understanding the Market Through DataThere is no liquidation line, and if the direction is correct, gold options can still lose money, and in severe cases, the entire premium can be reduced to zero.
Binance launched European options settled in USDT for gold and silver today, and regular users can only buy call or put contracts. Buyers don't need to worry about forced liquidation, and the maximum loss is limited to the premium paid. It sounds milder than perpetual contracts, but it simply replaces "sudden liquidation" with "void upon expiration."
A simplified example makes it clear:
Suppose the current price of gold is $4,000, and you spend 50 USDT to buy a call option with a strike price of $4,100. At expiration, gold rose to $4,120, which was indeed the right direction, but the option only had an intrinsic value of $20. After deducting the $50 premium, you still lose $30. The real breakeven point is at $4,150; gold only starts to turn a profit after rising above this level.
If gold stops at $4,090 at expiration, even if it is $90 higher than when purchased, the option will still lose its strike value and the premium paid may be lost.
Short-term options are especially prone to falling into this trap. The products launched this time mainly have one-day and one-week terms, leaving very little time for market movements to be realized. If gold prices don't rise enough or rise too late, the value of time will keep eroding. If the market is highly volatile at the time of buying, the premium is already expensive, and then implied volatility drops, possibly causing gold to rise while call option prices fall.
Therefore, before buying options, you shouldn't just judge whether gold will rise or fall; at the very least, you should look at the strike price, expiration date, premium premium, and breakeven point. They haven't even calculated "when and how much the price should rise," which is essentially still a gamble.
The risk of perpetual contracts is written in the liquidation price, while the risk of options is hidden in time and pricing. No liquidation warning does not mean losses will not occur; Sometimes it just quietly waits until it expires, turning the premium into zero.$HYPE HYPE Falls Below $55, Institutions Fall Apart — Which Side Are You On?
At $54.9, someone is taking a profit, someone is saying it's cheap.
Two on-chain moves last night: Multicoin Capital unstakes 1.97M HYPE ($108M), $4.78M has already been transferred to Coinbase Prime — bought at ~$30 via Galaxy Digital OTC five months ago, coming out close to double to $55. Selini also deposited 495K HYPE (~$26.8M) in OKX, founder Jordi Alexander explains: for HyperEVM tx fees, staking, LP, and arbitrage — not selling.
On the same day, Grayscale turned bullish.
Their research head said HYPE should be valued based on EPS — Hyperliquid is estimated to generate $1B in revenue by 2027, or $3.25–$3.75 per share in current circulation. At $55, the forward P/E is only 15–18x — still cheap compared to fintech peers of 20–40x.
In the short term, the pace is rough.
ETFs have had three consecutive weeks of net outflows, with an additional $4.13M released on July 27–28. Technically, the EMA50 ($58.45) and EMA200 ($62.15) form a double barrier. There are 6.93M more HYPE to unlock in the next 7 days — 3.3M of that as of tomorrow.
But the foundation is solid: Accumulated protocol revenue ~$1.15B, with buyback and burn exceeding 44M HYPE, annualized buyback exceeding $30M.
Short-term selling pressure vs. long-term value. At $54.9, which side are you on?#美联储即将公布利率决议 #财报观察员:微软 Meta亚马逊今夜交卷 #海力士业绩创纪录但不及预期, 存肨股剧烈波动 $BTC $ETH $HYPE HYPE Breaks Below $55, Institutions Diverge — Which Side Are You On?
At $54.9, some are cashing out, while others call it cheap.
Two on-chain moves last night: Multicoin Capital unstaked 1.97M HYPE ($108M), with $4.78M already moved to Coinbase Prime — bought at ~$30 via Galaxy Digital OTC five months ago, exiting near double at $55. Selini also deposited 495K HYPE (~$26.8M) to OKX, with founder Jordi Alexander clarifying: for HyperEVM tx fees, staking, LP, and arbitrage — not a sell-off.
Same day, Grayscale turned bullish.
Their research head argued HYPE should be valued on EPS — projecting Hyperliquid can generate $1B in profit by 2027, or $3.25–$3.75 per share on current circulation. At $55, that's just 15–18x forward P/E — still cheap versus fintech peers at 20–40x.
Near-term, however, looks rough.
ETFs saw net outflows for three straight weeks, with another $4.13M pulled on July 27–28. Technically, EMA50 ($58.45) and EMA200 ($62.15) form a double ceiling. Another 6.93M HYPE is set to unlock over the next 7 days — 3.3M of that tomorrow alone.
But fundamentals remain solid: Cumulative protocol revenue ~$1.15B, with buyback and burn exceeding 44M HYPE, annualized buyback over $30M.
Short-term selling pressure vs. long-term value. At $54.9, which side are you on?#美联储即将公布利率决议 #财报观察员：微软Meta亚马逊今夜交卷 #海力士业绩创纪录但不及预期，存储股剧烈波动 $BTC $ETH CORE Chain, are you doing well now?
Let me ask you first: Core DAO team, do you still watch DefiLlama yourself?
Today, July 29, 2026, I have a look for you. The data looks like this:
· TVL: $4.37 million
· 24-hour total protocol revenue: $8.64
What level is 4.37 million? A slightly popular meme project might have a liquidity mining pool deeper than this. And that $8.64 is not profit, but the total fee all protocols on the entire network receive from users today. One Layer 1 public chain generates a whole day's commercial income, which isn't enough to buy two sandwiches at the convenience store downstairs.
This isn't a crash—it's a death sentence.
Who is still using this chain?
I flipped through the on-chain records. This $8.64 is not a DefiLlama bug; it is real. It comes from some leftover exchange and staking operations, with so few trades you can count them all on two hands. Lending agreement? No one is borrowing. Derivatives? No one opened the warehouse. The whole ecosystem is like an abandoned mall—the elevator is still buzzing, but all the shops are closed, and only a vending machine in the corner occasionally drops a bottle of water, contributing to the mall's revenue.
This is called the "zombie chain," and it's the kind of zombies that don't even bother to eat their brains.
Verifier: The worst worker in history
The security of a public chain depends on validators bearing it with their lives (money). Currently, the entire network collects a total daily fee of $8.64. How much can each validator actually receive? How do they pay their electricity bills? How do you pay server rent?
Some say, "Anyway, the block reward is inflationary tokens, and validators run for that"—but those saying this probably haven't done the math. If a chain's network utilization hits rock bottom, the token loses its value anchor. Coins generated by inflation can only be used to zero out, what else can they do? The rewards validators receive are essentially a pile of paper that depreciates every day. To those who can still keep running nodes in this state, I can only say: you truly love technology. I'm crying.
Data doesn't lie, but people do escape
DefiLlama is not mistaken. I repeatedly confirmed that Core's bridge data, on-chain activity, and fee statistics all point to the same conclusion: users left, money left, and only an empty shell remained.
That is the most terrifying part. A crash isn't scary, going to zero isn't scary, what's scary is "no one left." No one scolds you, no one uses you, and not even anyone shorts you. At this point, is there still any point in discussing a chain's market value? It's like discussing how much a mummy weighed a few grams today.
To sum up the current state of the Core chain bluntly: the mainnet is alive, but the ecosystem is dead; Blocks are coming out, but where are the people?
If the team is still there, if the community is still there, step out and take a few steps. Let's take a look at how a single day of $8.64 in revenue supports a story of a "Bitcoin DeFi future."
I'm waiting.$HYPE HYPE breaks below 55, institutions are fighting—which side are you on?
HYPE at $54.9 is running while shouting about its low price.
Last night, two on-chain events occurred: Multicoin unstaked 1.97 million HYPE ($108 million), and 4.78 million transferred to Coinbase Prime—$30 cost five months ago, $55 shipped, and exited at double the price. Selini also deposited 495,000 tokens on OKX, and the founder responded: It's for HyperEVM, not for selling.
On the same day, Grayscale was called 'Bulg'.
Hyperliquid is expected to earn $1 billion in profit by 2027, with earnings per share of $3.25–3.75 based on circulating volume. A $55 equivalent to a forward PE of only 15-18 times is still considered cheap compared to 20-40 times for fintech companies.
In the short term, it really looks bad.
ETFs saw net outflows over three weeks, with another $4.13 million withdrawn on July 27-28. On the technical side, the EMA50 (58.45) and EMA200 (62.15) are both suppressed. In the next 7 days, there will be 6.93 million more unstaked tokens, and 3.3 million coins in Tomorrow Days alone.
But the fundamentals are tough: cumulative protocol revenue is $1.15 billion, buybacks and burns exceeded 44 million coins, and annualized buybacks exceed $30 million.
Short-term selling pressure vs. long-term value. $54.9, which side are you on? #美联储即将公布利率决议 #财报观察员: Microsoft, Meta, and Amazon will deliver their #海力士业绩创纪录但不及预期 tonight, with storage stocks experiencing sharp $BTC $ETH volatility "FOMC + Microsoft/Meta Earnings: One Conclusion"
The long-awaited shoe finally dropped, and the result is quite interesting:
🔹 FOMC: Maintained interest rates unchanged, no rate hike, no worse-than-expected negative news. The market's anxiety eased, and risk assets took a breather.
🔹 Microsoft: CapEx held steady at 190 billion, no further increase. The market's biggest fear of a "capital black hole" did not expand, and the AI narrative is temporarily stabilized.
🔹 Meta: CapEx maintained in the 125-145 billion range, no further upward adjustment. Meta, which previously plummeted due to AI spending, might get some relief tonight.
Conclusion: Bullish on ETH, bearish on MU.
ETH: Uncertainty eliminated, shorts covering, direction unchanged.
MU: High AI spending continues, but the market is not looking at "how much was spent," but rather "has it been earned back?" As long as this answer is not clear, the bearish logic on Micron remains.
Both positions are held, no rush. 🧐
---
The above is a personal trading record and does not constitute investment advice. The Federal Reserve decision will be announced tonight. What truly determines the direction of BTC and the US stock market is not whether interest rates are raised or not.
At 2:00 AM Beijing time on July 30, the Federal Reserve will release its latest interest rate decision, followed by a press conference hosted by Chairman Powell.
What makes this meeting special is that for the first time, the market cannot use "forward guidance" to bet in advance.
According to CME's "FedWatch," the current market pricing shows:
Maintain interest rates: 69.5%
Raise rates by 25 basis points: 30.5%
On the surface, "no rate hike" is still the most probable event, but the 30% chance of a rate hike is already much higher than normal meeting levels. More importantly, U.S. banks point out that since 1994, the Fed has never suddenly raised rates when the market's rate hike probability was below 60%.
In other words, if a rate hike really happens this time, it will be a highly impactful "hawkish surprise."
The data is conflicting.
On the side supporting rate cuts or pauses:
July consumer confidence dropped to 90.8
Employment expectations and employment sentiment are weakening
Residents' confidence in future income and consumption is declining
All these indicate the U.S. economy is cooling down.
But the hawkish side also exists:
International oil prices have recently rebounded significantly
Energy prices are pushing inflation expectations back up
Service sector inflation remains sticky
Simply put: the economy is cooling, but inflation is not dead yet.
This is exactly the headache the Fed faces now.
What the market is really watching is Powell's "next sentence."
Many focus on "whether there will be a rate hike tonight."
But historically, major market moves are often driven not by the rate number itself but by Powell's description of the future path.
Especially this time, he has removed the previously clear forward guidance.
Previously, the market could trade in advance based on "likely September moves" or "likely end-of-year moves"; now this framework is basically invalid.
Therefore, the most important thing tonight is not:
Whether the rate stays at 5.25%–5.50%
Or is raised by 25BP
But rather:
Whether the possibility of a rate hike in September remains
Whether inflation is still considered "not meeting the target"
Whether the slowdown in employment is enough to change policy direction
What does this mean for BTC?
If the following combination occurs:
Maintain rates
Powell emphasizes data dependence
No clear signal of a September rate hike
Then the market will likely interpret this as "dovish," and risk assets ($BTC, Nasdaq, AI tech stocks) may see a round of sentiment recovery.
But if:
Rates remain unchanged
While emphasizing oil prices, service sector inflation, and inflation risks remain high
Clearly stating that a September rate hike is still possible
Then the market will reprice "Higher for Longer."
This is not good news for high-valuation tech stocks and crypto assets.
My view:
I still believe:
The probability of a direct rate hike in July is low
The real key meeting is in September
Tonight is more like a "tone-setting meeting," not a "final meeting"
The biggest risk now is not that the market doesn't know the result, but that the market lacks consensus on the "future path."
And without consensus, volatility will be amplified.
Finally, the same advice:
Don't just focus on what Powell says, but watch how prices move.
Because in financial markets, prices are always more honest than words.
$BTC $AEON $SNDK
#美联储即将公布利率决议 Is a stock market crash unfolding? The global AI chip sector is caught in valuation trampling
Global AI chip assets are experiencing a fierce valuation stampede, with panic spreading rapidly across markets.
🇰🇷 The Korean market
South Korea's KOSPI index triggered circuit breakers the previous trading day, closing down 10.8%; The next day, midday trading dropped another 8.2%.
Heavyweight semiconductor sectors led the decline: SK Hynix plunged 12.6%, and Samsung Electronics also dropped 8%.
🇺🇸 The U.S. market
The sell-off also spread to the US semiconductor sector: Micron Technology plunged 8.9%, AMD dropped 8.1%, and Applied Materials closed down 7.8%.
A deeply ironic reality: even today, even impressive financial reports cannot stop stock prices from crashing. SK Hynix has just announced its strongest single-quarter performance ever, with quarterly operating profit reaching 60.5 trillion KRW, nearly a sixfold year-on-year increase, setting a new record, yet its stock price is still relentlessly sold off by the market.
Strong performance has instead become a window for cashing in and exiting, fully reflecting that the current market is no longer paying for long-term stories, and funds are beginning to worry about the sustainability of AI capital expenditure.
For the upcoming market outlook, focus on three core variables:
1️⃣ Whether the US chip sector experienced a breakout and decline due to increased volume, confirm whether this round of adjustment is a pullback or a trend reversal
2️⃣ Can leading tech giants provide real cash flow data to prove that high AI investments can translate into real returns?
3️⃣ Signals from the Federal Reserve's policy meeting: whether to release a hawkish stance on further rate hikes will directly determine the liquidity environment for global risk assets
Disclaimer: This is for market review only and does not constitute any investment advice. Financial market volatility and risk are extremely high. Please make independent and prudent decisions.$SNDK 别急着喊反转！！！
一根长下影从991拉回1090上方，恐慌盘被快速接走，说明1000附近确实有资金进场，空头最凶的一段暂时缓了下来
可价格依旧压在下降趋势线和短期均线下方，前面的低点、高点都在持续下移，4小时结构还没有真正翻多
下杀时成交量明显放大，反弹后的量能却在收缩，代表抛压暂时减弱，主动抢筹还不够强，这里更像超跌修复，不是已经确认反转
下方1000—1020是刚抢回来的阵地，再次回踩能够缩量守住，低点不再下移，筑底才会慢慢有样子
上方先看1120附近，真正难啃的是1180—1200，这里既是前面破位区，也是大量套牢筹码等着离场的位置
消息面也没有完全转暖，AI服务器和企业级SSD需求仍在支撑NAND市场，可长江存储扩产和份额提升，让资金开始重新评估后面的竞争与价格压力，YMTC去年全球NAND份额已接近SNDK，市场担心的不是眼前需求消失，而是未来利润会不会被竞争吃掉
另一边，SNDK将在8月5日盘后公布财报，8月13日还有投资者日，市场接下来会死盯NAND价格、企业级需求和下半年指引，数据落地前这种上下几十上百点的洗盘不会少
1000附近能守住，反弹还有继续往上修复的空间
真正把1200拿回来，才说明这波暴跌不只是喘口气
现在最值钱的不是猜中991是不是底，而是等它把下降趋势线狠狠干开以后，再决定这波反弹值不值得跟到底
$MU $ETH #美联储即将公布利率决议 #海力士业绩创纪录但不及预期，存储股剧烈波动 本次美联储FOMC会议，市场基本预判维持利率不变，重点不在于加不加息，而是表态偏鹰还是偏鸽
本次属于非SEP会议，不会出新点阵图与经济预测
偏鹰：强调通胀顽固、高利率维持更久，美元和美债收益率走强，利空$BTC 、$ETH 、美股成长股
偏鸽：释放通胀降温、宽松预期升温信号，利好加密资产与科技板块
核心逻辑：维持利率不变早已被市场提前计价，真正引发大幅波动的是预期差。若措辞比预期强硬，或发布会出现意外信号，行情才会迎来大震荡#海力士业绩创纪录但不及预期，存储股剧烈波动 #美联储即将公布利率决议 $ETH $BTC AI computing power prices have not dropped at all: H100 up 25%, B200 up 28.6% within the year
Silicon Data collects over 170,000 hash rate quotes worldwide every day.
The latest data shows that since the beginning of this year, on-demand rental prices for A100, H100, H200, and B200 have all risen by 15%–30%: A100 up 20.6%, H100 up 25%, H200 up 14.4%, and B200 up 28.6%.
Analysis by U.S. Investment Network believes that the price hikes across the board—from old cards to the latest generation GPUs—indicate that AI demand is not concentrated only in a few flagship chips, but is spreading across training, inference, and customers with different budgets.
This set of data directly refutes the claim that "GPUs are already in surplus."
A large amount of hash power will still be charged at low-priced long-term contracts for 2024–2025, while spot prices continue to rise; As old contracts expire and are repriced, cloud providers' revenue and operating cash flow may be significantly revised upward.
The market worries they spend too much on AI, but the real computing power market gives the answer: they may just be temporarily too cheap.
$NVDA $MU $TSM $LITE $COHR #美股特朗普“买股神话”再次刷屏，但真正重要的不是他推荐了哪些股票。
而是：
市场正在重新定价一类资产。
最近，一段特朗普早年的采访视频突然在 X 上爆火。
视频里他说：
“我会告诉你如何赚钱。”
网友随后翻出他的“历史战绩”：
$INTC
约20美元 → 上涨近600%
$DELL
约230美元 → 上涨近90%
$MU
约740美元 → 上涨约70%
于是市场开始寻找：
他下一次关注谁？
$ALOY
$USAR
$MP
但如果只盯着“特朗普荐股”，可能错过真正的投资逻辑。
因为这些股票背后有一个共同点：
它们都踩中了美国未来几年最重要的产业方向。
过去两年，市场只有一个超级主线：
AI
资金追逐：
GPU、HBM、光模块、数据中心
但进入2026年，市场开始问一个更现实的问题：
AI最终需要什么？
需要：
芯片制造能力；
稳定能源供应；
关键矿产资源；
本土供应链。
于是资金开始寻找 AI 时代的“基础设施”。
这也是为什么：
英特尔重新获得关注；
美光成为存储周期核心资产；
稀土、核能、能源供应链开始被重新估值。
因为未来竞争，不只是模型竞争。
而是：
谁掌握算力背后的资源和制造能力。
特朗普政策一直强调：
美国制造回流。
减少关键供应链依赖。
这意味着：
未来几年，美国政策资金可能持续偏向：
半导体制造
能源安全
稀土资源
国防供应链
所以，这段旧视频今天重新传播，并不是因为特朗普“预测股票”。
而是市场突然发现：
过去被忽略的公司，正在被重新定义。
投资机会往往不是出现在故事最热的时候。
而是在：
一个旧逻辑，被新的时代重新赋予价值的时候。
当然，政策支持≠股票必涨。
最终还是要回到：
订单。
盈利。
现金流。
美股投资网认为：
AI下一阶段的赢家，可能不仅仅是卖GPU的人。
还有那些提供：
能源、材料、制造能力的人。
#美股Everyone's watching the chart. The real story is happening behind it. 👀
$BTC is hovering around $63.4K–$64K, and this isn't just a technical setup.
The Fed is widely expected to keep interest rates at 3.5–3.75% today, meaning liquidity is likely to remain tight. At the same time, rising US-Iran tensions have pushed oil to $74.67, reigniting concerns that inflation could make a comeback.
We've already seen how sensitive the market is. $BTC dropped from $72K to $63K as the conflict escalated, and if the Fed maintains a hawkish stance, altcoins could face even greater pressure.
For now, crypto isn't behaving like a safe-haven asset—it's still trading like a risk asset.
The next move may depend less on the chart and more on what comes out of the Fed today.
#DailyOrbit #FedRateDecision #BigTechEarningsNight 科技七姐妹回调反弹过后，抱团行情还能维系多久
一、本轮先跌后涨的底层逻辑⚖️
依我看，七姐妹集体回调后的反弹，纯粹是超跌情绪修复，并非全面抱团行情重启，资金抱团模式早已发生质变。$SNDK $SPCX $ETH
1. 前期大跌根源：美联储降息预期反复摇摆，高利率持续压制高估值成长股；谷歌、微软、Meta 不断加码 AI 资本开支，自由现金流持续承压，市场质疑 AI 投入兑现利润周期过长，集体抛售高位筹码，板块整体回撤超 14%。
2. 反弹驱动力：短期利空充分消化，叠加存储芯片（美光、闪迪）算力刚需走强，带动整条 AI 产业链回暖；BTC 同步企稳回升，风险资产偏好小幅回暖，资金进场抄底超跌龙头。
3. 关键变化：从前两年无脑全线抱团买入，彻底转为内部极致分化，苹果逆势走新高，微软、特斯拉大幅回撤，整齐上涨的抱团时代已经落幕新浪财经。
空一行
二、三段周期抱团持续性预判（乐观 / 悲观双情景）
✅短期（3-15 个交易日，美联储决议落地窗口期）
乐观触发条件：凌晨议息维持利率不变，措辞中性偏鸽，预留 9 月降息空间，两大鹰派委员仅投反对票不加息。
我预判：短线抱团修复行情可以延续 7-10 个交易日，纳指震荡上行，英伟达、存储龙头领涨；但板块内部分化不会消失，现金流稳健、算力硬件兑现业绩的标的涨幅更强，烧钱严重的云厂商反弹力度偏弱。
悲观触发条件：决议措辞鹰派，推迟降息甚至意外加息。
抱团反弹直接宣告结束，资金集体出逃科技高位股，再度切换至能源、公用事业防御板块，七姐妹重回阴跌磨底走势。
✅中期（1~2 个月）
乐观情景：美国通胀稳步回落，9 月降息落地；各大巨头二季度财报兑现 AI 营收增长，算力、存储订单持续饱满。
我的观点：结构性抱团会持续存在，不再是七家齐涨，资金集中扎堆英伟达、苹果、美光这类业绩落地确定性标的，AI 算力硬件成为抱团核心主线，纯软件云企业逐步被资金边缘化。
悲观情景：油价反弹推升通胀，降息不断延后；AI 资本开支投入产出不及预期。
大范围抱团彻底瓦解，七姐妹各自走独立行情，板块箱体宽幅震荡，只有存储芯片细分赛道具备独立上行机会。
✅长期（半年以上）
依我综合产业周期推演：完整全面抱团行情仅剩 2 个月左右窗口期；结构性抱团可延续至 2027 年年中。
1. 抱团收尾节点（2 个月后）：前期普涨式抱团彻底终结，资金告别一篮子配置七巨头的模式。
2. 结构性抱团存续逻辑：AI 数据中心存储缺口将维持至 2028 年，HBM、企业级闪存刚需源源不断，算力硬件赛道业绩持续兑现，资金会长期抱团存储、GPU 硬件龙头；软件端巨头则依靠自身生态缓慢增长，很难走出大级别行情。
3. 最终拐点（2027 年中旬）：日韩存储大厂新增产能集中释放，AI 硬件供需缺口收窄，AI 超级周期红利衰减，整条科技赛道抱团行情正式落幕。
空一行
三、支撑抱团延续核心利好📈
1.AI 需求不可逆：大模型私有化部署普及，全球算力扩容需求持续攀升，存储、GPU 硬件订单排期充足，业绩有实打实基本面托底。
2. 权重维稳需求：七姐妹占据标普 500 超 32% 权重，纳指牛市离不开科技巨头托底，资金不会彻底撤离该板块。
3. 流动性宽松预期：只要美联储开启首轮降息，全球资金会重新回流美股成长资产，利好科技赛道估值修复。
空一行
四、瓦解抱团的关键风险⚠️
1. 高利率周期无限拉长：降息持续推迟，高估值科技股估值不断被压缩，资金持续撤离高位龙头。
2.AI 盈利兑现不及预期：云厂商持续烧钱却无法转化稳定利润，市场 AI 叙事信心崩塌。
3. 供给端产能释放：三星、SK 海力士放开存储产能，芯片涨价周期结束，硬件赛道红利消退。
空一行
五、个人实操研判总结
我的实操思路很清晰：短线博弈本轮反弹仅限轻仓参与，不要追高；中期放弃全线抱团思路，精选存储芯片、GPU 硬件龙头布局；半年之后逐步减仓离场，切莫死守老旧全面抱团行情思维。The first step in bear market research projects may be elimination. The deBridge team may be one of the most sensitive Web3 teams. Started in the Solana ecosystem in 2023, and will develop cross-chain integration between Solana and EVM ecosystems in 2024~2025; AI tools will be developed in Q1 2026, and cross-chain payments will focus on Q2 2026...... Before HyperLiquid's boom, it integrated with HyperLiquid and HyperEVM ecosystems and quickly supported it after Robinhood's launch...... ┈➤ Suspected Lying Flat ecosystem At some point, deBridge stopped supporting many ecosystems. Including: ◆ Stablecoin-dedicated chain—Plasma, TVL$609.7 million ◆ Public chain founded in 2015—Gnosis, TVL$99.61 million ◆ Deeply tied to Bybit—Mantle, TVL$7,812 ◆ Berachain, creator of the POL consensus three-coin model, TVL $49.79 million ◆ High-speed parallel public chain—Sei, TVL $41.64 million ◆ Chain extended from Fantom—Sonic, TVL $14.14 million ◆ Self-developed chain for gaming and consumption—Flow, TVL $10.87 million ◆ Consumer chain built by the Fat Enterprise team—Abstract, TVL $10.16 million Additionally, there are BOB and Neon, TVL less than 10 million. Today, the most worth watching in the US stock market is not who is rising, but that money is quietly changing seats.
NVIDIA lost $238 billion in market value over two trading days, roughly 4.7 times its quarterly operating cash flow. The trigger was not the earnings report; last quarter, it had revenue of 81.6 billion, with data centers accounting for 92%, showing a very strong fundamental performance. What really made the market uneasy was another matter: NVIDIA reportedly plans to guarantee up to $250 billion in financing for OpenAI's data center project in Ohio.
This is the core issue. When the shovel sellers start providing loan guarantees for the gold diggers, the market instinctively asks: Is the demand in this business real orders backed by actual money, or just a cycle of moving money from one hand to the other? The sustainability of AI capital expenditure is being seriously reconsidered for the first time in half a year.
So the funds made a very honest move: withdrawing from chips and hiding in software. Today, enterprise software companies like Zscaler, Shopify, and PTC are rising, driven by falling treasury yields and renewed bets on the AI investment cycle. Chips are dragging the Nasdaq 100 down, while the Dow Jones is supported by old economy names like Boeing and Coca-Cola, rising 1%. The same market, two different worlds.
Adding the China factor makes it even clearer. With ChangXin Memory going public and rumored process breakthroughs, storage and semiconductors are collectively under pressure, with SK Hynix and Samsung crashing first in the Asian session. The market’s patience for the "AI + chip" combination is clearly not as strong as it was six months ago.
My view is that this is not the end of the AI narrative, but its first growing pain shifting from "faith-based pricing" to "cash flow-based pricing." In the previous phase, anyone could rise; in the next phase, differentiation begins: those with real revenue will be recognized again, while those relying on hype to support valuations will be slowly drained.
Apple quietly reached a $5 trillion market cap these days, surpassing NVIDIA to become the world's largest company, which in some way underscores this logic. In market turmoil, investors instinctively gravitate toward places with higher certainty.
The wind has changed. At times like this, more important than chasing gains is understanding which side of the story the stocks you hold are on. SK Hynix's Crash Warning: The Most Perfect Narrative Often Hurts the Deepest
The most vigilant aspect of SK Hynix's recent brutal decline is that the most talkative stories in the market often hurt investors the most.
Previously, the market's logic was nearly flawless: Jensen Huang personally wrote "Please Make More" on HBM4E wafers, publicly stating that the storage shortage could last for years.
HBM is the most bottleneck-neck core component in the AI computing power chain, and SK Hynix is one of NVIDIA's core suppliers. Coupled with its low P/E valuation, the fundamental story seems almost unflawed.
Major investment banks have all made a big push for bullish sentiment: Nomura raised its target price to 4 million KRW, while Shinwan and Korea Investment & Securities set targets of 3.8 million KRW, with the market's highest target expectations reaching as high as 5.3 million KRW.
As of now, among the 37 analysts covering SK Hynix, 36 have given a Buy rating, with the market consensus target price still as high as 3.409 million KRW. Even though the stock price has already collapsed, the target price set by institutions still has more than double the potential for upside compared to the current price.
Solid performance, low valuations, and highly certain AI demand are, in most people's minds, at best a temporary correction, and absolutely unlikely to cause a short-term halving. But the reality far exceeded everyone's expectations.
On June 25, SK Hynix hit an intraday high of 2.987 million won, setting a new all-time high; on July 29, it hit an intraday low of 1.246 million won.
In just 34 trading days, the maximum drawdown reached 58.3%.
Let's make a direct comparison: Bitcoin has retreated from its all-time high of about $126,300 in October 2025 to a low of around $60,000 on February 6, 2026, with the entire downturn lasting 123 days and a maximum drop of 52.5%.
In other words, SK Hynix completed a much larger round of decline in less than one-third of Bitcoin's downturn cycle.
While everyone is immersed in the narrative of the supercycle, the impact of cycle reversals is far faster than imagined.
Disclaimer: The above is only a review of market phenomena and does not constitute any investment advice. The stock market carries high risk, so please make independent investment decisions. $BTC $SNDK $SKHYNIX $BTC and $ETH rose 1.58% and 2.02% respectively before today's pre-market opening. Market sentiment appears to be warming, but the storage sector has revealed deeper issues. U.S. storage stocks all fell pre-market sharply, with Micron, SanDisk, and Western Digital all down about 4%. Intel dropped 3.2%, having already dropped over 7% the previous trading day. SK Hynix ADR and SanDisk continued to hit new lows. Meanwhile, Changxin Memory surged 466% on its first day in Shanghai, directly pushing competition expectations into global valuations. The signal is clear: strong demand does not mean stock prices won't fall; once new variables emerge on the supply side, the most expensive expectations will be the first to settle. Mapping to the crypto market, storage track coins such as $FIL, $STORJ, and $AR face similar logic. Despite the rebound in $BTC and $ETH, storage coins have experienced an average weekly correction of about 15%, with some projects dropping more than 20%. We see that newly launched decentralized storage protocols or modular storage solutions are rapidly capturing market share. The computing power growth of traditional leader $FIL has slowed for three consecutive months, while the ecosystem expansion of new projects like $AR has brought higher valuation expectations. Yesterday, news about a new storage project receiving $50 million investment from a top institution directly caused $FIL to plunge 4.7% within 30 minutes. This is the supply-side variable: new players enter the market, and old expectations begin to be cleared. Don't be fooled by the short-term rebound in the overall market. The rally in $BTC and $ETH is mostly driven by macro sentiment#美联储即将公布利率决议
The crucial Fed rate decision is imminent overnight, how should bulls and bears respond to the tug-of-war in the market?
1. Core conflict in the current market ⚖️
In my view, the Fed decision at 2 AM is the most uncertain rate meeting in the past six months. Various messages keep pulling in different directions, and capital is fully cautious.
1. Interest rate pricing divergence: CME data shows a 69.5% probability of holding rates steady, 30.5% probability of a 25 basis point hike; Bank of America’s historical review shows the Fed has never raised rates when the hike probability was below 60% since 1994. An unexpected hike would be unprecedented, but the risk cannot be completely ruled out.
2. Economic data is a two-way game: US July consumer confidence declined and employment sentiment weakened, solidly supporting dovish easing logic; however, Middle East conflicts pushing oil prices up and inflation risks returning provide hawkish tightening policy justification.
3. Rules completely rewritten: With Waller’s appointment, forward guidance was scrapped, invalidating the market’s previous fixed framework for interpreting policy. The verbal tone of the entire press conference will become the sole pricing basis for all risk assets in US stocks and crypto going forward.
4. Institutional consensus forecast: TD Securities’ baseline result is no rate change, with two hawkish members voting against a hike, internal policy disagreements will further increase market volatility.
2. Two market scenarios (with trigger conditions)
✅ Optimistic scenario (70% probability, market baseline expectation)
Trigger: Maintain current rates, neutral to dovish tone in the press conference, no emphasis on sustained anti-inflation, subtly leaving room for a rate cut in September.
My forecast: USD and US Treasury yields quickly fall, Nasdaq and Philadelphia Semiconductor Index start an oversold rebound; storage chips like Micron and SanDisk that had deep prior pullbacks rebound strongest; BTC stabilizes above $65,000, US tech stocks and crypto assets both recover.
❌ Pessimistic scenario (30% probability)
Trigger: Unexpected 25 basis point hike, or no hike but consistently hawkish tone emphasizing oil-driven inflation rebound and no rate cuts this year.
My view: A black swan event, Nasdaq plunges sharply, storage sector valuation slashed again; BTC breaks key support at $62,500, global risk assets collectively flee, short-term correction space significantly expands.
3. Special impact points of this meeting 💡
1. After canceling forward guidance, every round of economic data will trigger violent market swings. Volatility back and forth in US tech, storage sectors, and Bitcoin will become the norm, with fewer large one-sided moves.
2. Korean storage leaders Samsung and SK Hynix, supported by domestic industrial policies, will show stronger resilience than US-listed storage firms like Micron and SanDisk, with divergence widening during the rate tightening cycle.
3. Short-term market moves will be completely driven by the tone of the statements; overnight volatility will be extremely intense, making heavy overnight bets unsuitable.
4. Personal practical final view
I recommend a wait-and-see approach at this stage, avoid preemptive bets on bulls or bears; after the decision, trade along with the market trend. Even if an optimistic recovery occurs, it will only be a short-term oversold rebound, not a mid-term trend reversal; if hawkish negative triggers occur, the subsequent bottoming cycle will be prolonged.
#美联储即将公布利率决议 #财报观察员：微软Meta亚马逊今夜交卷 Overnight Finance 1. Overall Market Overview On July 27, the US AI hardware sector declined significantly, with the Philadelphia Semiconductor Index closing down 2.2%, SanDisk plunging 11%, and SK Hynix down 7.5%; Selling sentiment spread to Asian markets, with the South Korean Composite Index dropping over 8% intraday and triggering circuit breakers, plunging 10.8% for the day, while the Nikkei 225 also closed down 4%. The essence of this round of decline is that global capital is reassessing the credit and supply landscape of the tech industry. The real demand for AI has not weakened substantially, and funds are shifting towards Hong Kong stocks for risk aversion and allocation. 2. The underlying logic behind the decline in AI hardware in the US stock market and the rise in CDS among overseas cloud providers 1. Sharp upward in capital expenditure expectations fuels credit anxiety. Google raises its 2026 capital expenditure guidance and increases AI infrastructure investment in 2027, driving overall market increases in spending forecasts for the four major cloud service providers. Total estimated expenditures for 2026 and 2027 are raised to $731.9 billion and $950.2 billion, respectively. The market has realized that the AI investment cycle and scale have exceeded previous estimates, and have begun to worry about tech companies' increased leverage, cash flow pressure, and monetization efficiency lagging behind spending growth, leading to widening CDS spreads among major cloud providers. 2. Nvidia's closed-loop financing model amplifies market risk aversion. Nvidia provides financing and guarantee services to downstream customers, encouraging them to use funds to purchase its own chips, forming a highly bound credit closed loop; Combined with Nvidia's $500 billion cooperation with SK Hynix and negotiations with OpenAI for up to $250 billion in financing guarantees, the market fears the supply chainAs for why I think SK Hynix is a stock worth buying for a rebound, I have a simple intuition: once the whole world knows the name, they'll link it to the topic of AI storage, and the capital will come back to speculate on. This is the power of brand recognition and narrative momentum to influence financial markets. When investors seek exposure to topics they understand, they tend to favor the names they are most familiar with. This is quite similar to assets like Bitcoin, GameStop, and Dogecoin, which experienced leverage-driven speculative frenzy, leading to large-scale forced liquidations and position liquidations. Looking at the overall picture, these assets have also experienced a decent second rally (a strong rebound, though not always returning to historical highs). Not to mention, if you study SK Hynix's history (return on invested capital, operating profit margin, free cash flow, etc.), it is indeed a well-managed storage company. Its outstanding manufacturing capabilities, technological leadership, and prudent capital allocation strategy have enabled it to maintain a competitive edge amid intense industry cycles. Despite the current market sentiment toward the company, I firmly believe its fundamental quality will eventually regain market recognition. When market sentiment stabilizes and investors begin to reassess the industry, companies like SK hynix, with strong fundamentals and high brand recognition, are likely to become the top picks when capital returns. This aligns with the pattern we've seen in other assets that have experienced similar liquidation events—well-known projects with strong fundamentals tend to be solidSanDisk's stock price has fluctuated repeatedly—how much longer can the AI chip dividend last?
1. Reasons 💥 for the current bottom layer of the volatile market
In my view, SanDisk's ongoing tug-of-war and frequent swings are not due to a decline in AI storage demand, but rather a market driven by the tug-of-war among three types of capital:
$BTC $ETH $SNDK
1. Significant pressure to realize profit-taking positions in the early stages: Year-to-date, overall gains have exceeded 570%. After hitting a high of $2,354 in June, the market has pulled back, with the largest drop exceeding 53% in just over a month. Short-term funds have made substantial profits, but at the slightest disturbance, they all flee en masse, with single-day drops of 14%. Such extreme market events have repeatedly occurred.
2. Significant internal capital divergence within the sector: Currently, funds are favored by Samsung, SK Hynix, and Micron, which have HBM memory businesses. SanDisk is only deeply engaged in the NAND flash sector, without high-end memory business to hedge volatility. The market is most elastic and the pullback far exceeds peers, with its volatility amplified.
3. Macro liquidity influences sentiment: Fed rate cut expectations fluctuate repeatedly, with daily fluctuations in US dollar and Treasury yields, combined with BTC and ETH moving in sync, causing risk asset appetite to fluctuate and directly intensify SanDisk's intraday volatility.
Leaving only one line empty
2. Forecast of Dividend Persistence in the Third Cycle (Opsimistic / Pessimistic Dual Scenarios)
✅ Short-term (3-15 trading days)
Optimistic trigger conditions: The Federal Reserve signals a clear September rate cut, BTC holds above $65,000, and cloud providers announce additional AI computing capital expenditures
I predict this period will be a window for oversold recovery. Short-term AI demand hasn't declined, and the company holds large long-term supply contracts locking in revenue. SanDisk will rebound to the $1250-$1380 range, with volatility still the main theme, making it hard to see a one-sided rally.
Pessimistic trigger: Fed hawkish rhetoric and delayed rate cuts, BTC falls below key support at 62,000
The market will continue its bearish decline and bottom-grinding trend, with repeated tests below the support near $1020. The dividend is temporarily dormant, with only short-term opportunities and no medium-term investment value.
✅ Midterm (1-2 months)
Optimistic scenario: Global NAND spot and contract prices continue to rise, with SanDisk's Q4 earnings guidance significantly raised
My view: The AI dividend still remains firmly in effect, and the dividend cycle will at least run through the entire stage. Major cloud vendors continue to deploy AI inference clusters, data storage demand steadily rises monthly, SanDisk's enterprise-grade SSD order backlogs are full, gross margins remain stable above 75%, stock prices fluctuate upward, continuously pushing up the bottom range, and dividends are realized in a volatile upward pattern.
Pessimistic scenario: Samsung and SK Hynix have slightly eased NAND capacity expansion, slowing the pace of storage price increases
Dividends will weaken marginally, and stock prices will enter a high, wide box range, compressing upside potential and only following sector pulse rallies, making it difficult to produce independent major rallies, greatly increasing the difficulty of realizing profits.
✅ Long-term (over six months)
Based on my industry data analysis, the core dividend of AI storage can last until the second half of 2027, after which a turning point will occur:
1. Full Dividend Survival Phase (remaining 10-12 months): AI large model iteration, widespread adoption of on-device AI and automotive AI, continuous doubling of flash memory per computing device, slow ramp-up of wafer equipment and advanced processes on the supply side, and supply-demand gap that cannot be quickly filled. SanDisk continues to earn high profits through enterprise-grade flash memory, and the long-term upward trend will not reverse.
2. Dividend fading turning point (Q4 2027): Japanese and Korean storage giants will concentrate new capacity, NAND supply will be significantly relaxed, the cycle of memory chip price hikes will end, the industry will return to traditional cyclical fluctuations, and the excess dividends brought by AI will come to a complete close.
Leaving only one line empty
3. Support dividends continue core positive 📈 news
1. Irreversible Demand: In the AI era, data centers have replaced smartphones as the largest consumer market for NAND flash memory. A single AI server consumes 3-8 times more flash memory than traditional servers, and the rigid need for massive model training and data retention will not disappear in the short term.
2. Long-term contract orders lock in profits: SanDisk signs multi-year long-term supply agreements with leading cloud providers like Microsoft and Amazon, locking in product prices in advance. Even if market sentiment fluctuates, the company's revenue and gross margin will not plummet.
3. Technological Iteration Opens New Space: HBF storage samples will be sent out by the end of 2026, mass production will begin in 2027, CXL high-speed interfaces will be fully implemented, continuously strengthening AI storage technology barriers and exploring new profit growth points.
Leaving only one line empty
4. Core risks ⚠️ that can interrupt dividends
1. The high interest rate cycle is infinitely extended: The Federal Reserve has delayed rate cuts, tech companies have cut back on computing capital spending, and the expansion of AI storage demand has slowed, forcing the pace of dividend payments to be delayed.
2. Uncontrolled Expansion by Peers: Samsung and SK Hynix are rapidly releasing NAND capacity to capture market share, narrowing the supply-demand gap and causing flash memory prices to turn downward.
3. Valuation bubble correction pressure: SanDisk's current P/E ratio is relatively high. If performance growth cannot keep pace with stock price increases, it will lead to sustained valuation digestion, prolonging the oscillating downward cycle.
Leaving only one line empty
5. Personal Practical Summary and Views
My practical approach is clear: in the short term, it's only suitable for light positions to play back and forth with oscillating price differences; In the medium term, on major drops, they will be positioned in batches to capitalize on the dividends of the AI storage price increase cycle; There is about a year left in the long-term holding window. In the second half of 2027, gradually take profits and exit. Do not blindly hold positions at high points for a long time. #美联储即将公布利率决议 #财报观察员: Microsoft, Meta, and Amazon deliver tonight. #海力士业绩创纪录但不及预期, storage stocks experienced sharp fluctuations #That's why the entire market crashed today—in just a few hours, the global market evaporated by over $1.4 trillion. The stock market was heavily sell-off, the metals market was hit hard, and cryptocurrencies followed with a significant decline. On the surface, this seems like another random decline day without any specific catalyst. However, reality is far more complex and enlightening than this. This is a domino effect that starts in one place and spreads everywhere. The initial trigger was an AI panic, which triggered a whole chain of events. The world's busiest deal has suddenly come under scrutiny, as reports indicate that China has made a major breakthrough in domestically produced DUV lithography technology. This has raised concerns that the West's long-term chip leadership may not be as secure as everyone imagines. Meanwhile, new headlines about AI financing have added further doubts about the sustainability of the AI boom. Reports indicate that NVIDIA is discussing hundreds of billions of dollars in funding for OpenAI's infrastructure, raising questions about the cost structure of the AI ecosystem. Meanwhile, the four major AI players are expected to spend about $700 billion on AI capital expenditure this year, representing a staggering investment and raising concerns about returns. For months, the question has been, "How big can AI become?" Today, it has turned into "Who is actually paying for all this?" "The narrative shifting from growth to profitability represents a significant shift in market sentiment and could have far-reaching effects on AI-related stocks. Once the semiconductor sectorThe biggest recent change in global capital markets is the continued weakness of high-level tech sectors in U.S. stocks, especially memory chips, AI hardware, and tech heavyweight stocks, which have pulled back continuously, directly damaging all global risk asset sentiment and causing the crypto market to enter a linked adjustment cycle. Yesterday, the Korean stock market plunged dramatically, and today the US tech sector remained weak, with global high-leverage assets collectively deflating bubbles. Core storage stocks such as SanDisk and Samsung Memory continued to plunge, with market concerns about AI storage demand peaking and overcapacity, confirming a turning point in cyclical stocks. This logic directly extends to the AI sector and storage concept coins in the crypto world, resulting in weak rebounds and continued pressure on WLD and AI series tokens. The capital market is always interconnected; U.S. stocks are driven by tech valuations, and crypto-themed coins never see major market moves. Currently, U.S. stock funds are withdrawing from growth tech stocks and shifting toward conservative defensive sectors. Global liquidity preference is declining, and naturally, the crypto world lacks incremental vitality. Currently, the crypto market is completely following the external market: the external market is stable, while Bitcoin is volatile; Overseas markets are falling, while altcoins are accelerating their pullback. Currently, there is no standalone bull market; all rebounds are oversold repairs. At this stage, investors should completely abandon the get-rich-quick mentality, prioritize avoiding high-priced themes and high-priced AI coins, focus on undervalued and oversold blue chips, and patiently wait for U.S. stocks to stabilize and risk sentiment to recover. $ETH Today, trading in the AI storage sector suffered a heavy blow, with several semiconductor stocks facing significant selling pressure. SanDisk ($SNDK), Micron ($MU), Western Digital ($WDC), Seagate ($STX), and SK Hynix ($SKHY) all experienced significant declines as China's CXMT sparked market concerns about new supply entering the market and disrupting the current supply-demand balance. The current debate on Wall Street is whether analysts and investors are confusing commercial memory chips with HBM. These are fundamentally different products with different competitive dynamics, yet the market seems to treat them equally. This confusion may create opportunities for those who can distinguish between these different segments. Concerns about CXMT's potential impact on the storage market may be exaggerated, as HBM remains a technologically advanced product that is difficult to manufacture at scale. However, the market is not always rational in the short term, and fear may drive prices lower regardless of fundamentals. The AI storage sector has long been one of the strongest performing sectors in the semiconductor industry, and the current pullback may offer buying opportunities for investors with a long-term investment perspective. However, patience remains important, as selling pressure may persist in the short term.Many people can't tell the difference between RWA wealth management and RWA concept coins, but Sister Long will explain everything in one go 👇
The real benefits of RWA: Idle USDT in bear markets has a place to earn 4-5% USD returns.
In the past, during bear markets, you either had to lie on exchanges for meager interest or gamble on high returns through DeFi mining, bearing contract risks and token crashes.
Now, products like $PLUME and $ONDO can directly buy on-chain bond credit for stable returns, without having to chase hot topics everywhere.
⚠️ But! Doing RWA wealth management and buying their platform tokens are two completely different things!
Many RWA projects make a fortune, but when you buy their governance tokens, you don't get a cent in interest or management fee dividends.
Profits go to the project team's treasury, while token prices rely entirely on market storytelling and hype. No matter how well the business develops, the coin will still fall if it should fall.
Simply put,
👉 Depositing U into products to earn interest is called participating in RWA;
👉 Buying coins like ONDO or PLUME is just hyping up the narrative, not holding RWA assets.
Only when tokens can truly be allocated to the bottom of cash flow can they be considered genuine RWA assets; most of these still fall short.#美联储即将公布利率决议 At 2 AM Beijing time tomorrow, the Federal Reserve is set to announce its interest rate decision. This time is special as it will be the first press conference since Waller took office.
The current CME data is quite interesting: there's a 69.5% probability of keeping rates unchanged and a 30.5% chance of a 25 basis point hike. According to Bank of America, since 1994, whenever the market's rate hike probability is below 60%, the Fed basically doesn't move. But if they do hike this time, it would definitely be an unprecedented exception. TD Securities predicts a compromise: rates remain unchanged, but Harker and Logan will cast two dissenting votes in favor of a rate hike.
The supporting data is currently pulling in both directions. The July Consumer Confidence Index dropped to 90.8, and employment data looks weak, which supports the dovish side. On the other hand, oil prices surged past $100 due to missile attacks, pushing inflation expectations higher and bolstering the hawkish argument.
The most frustrating part is that Waller has canceled forward guidance. This means the old formulas and frameworks used to interpret Fed statements no longer apply. The market needs to rebuild a new set of logic to understand them.
So tonight's focus is not just on the decision itself, but more importantly on the wording of the statement and the signals he will send during the press conference. Bitcoin and Ethereum are basically waiting with low volume now, and everyone is waiting for the Fed to release the new "interpretation manual" first. My strategy is simple: do nothing and wait until they clarify this framework before making any moves. [SK Hynix Financial Report]
1. Why are short-term figures said to fall short of expectations?
Looking at Q2 data, Hynix's revenue and operating profit both fell short of buyers' expectations, about 6% lower.
This magnitude is not disastrous, but for some institutional funds trading based on "quarterly earnings revisions," it is enough to trigger short-term reductions.
This is also why the market's initial reaction after the earnings report came out was to be cautious.
2. Why did the performance fall short of expectations?
The core reason is not poor demand, but that the price increase has not been fully realized.
In Q2, SK Hynix's DRAM composite ASP rose about 30% quarter-on-quarter, which is a significant increase but lower than the spot price increase for DRAM during the same period.
Why hasn't SK Hynix fully benefited from spot price increases?
There are mainly two reasons:
First, HBM accounts for a high proportion in SK Hynix's product structure, which typically uses annual protocol pricing and does not fluctuate rapidly with spot prices like ordinary commodity DRAM.
So, although spot DRAM prices have surged sharply, SK Hynix has not benefited as fully as Samsung.
Second, SK Hynix has raised prices relatively mildly for long-term contract customers, and LTA long-term contract price adjustments are not as aggressive as Samsung's.
Therefore, essentially, it's not that demand is a problem, but that the pace of price fulfillment is falling short of market expectations.
3. Why does the Q3 guidance appear conservative?
Management did not provide direct guidance on third-quarter revenue and profit, only outlook for shipments:
DRAM bit shipments grew about 10% quarter-on-quarter, while NAND bit shipments grew by low single digits.
This statement sounds rather conservative.
But the essence is not weak demand, but limited production capacity.
Management mentioned that short-term supply-demand balance is unlikely to improve significantly. At the same time, HBM4 will continue to ramp up in the second half of the year, but HBM4 consumes even greater wafer capacity, further intensifying capacity constraints.
Therefore, the "conservatism" here is more about supply constraints rather than demand shortages.
4. When will the tight supply and demand for storage ease?
From management's statements, it is unlikely that the supply-demand gap will significantly improve in the short term.
Because demand isn't growing from a single point now, but from multiple lines pulling simultaneously:
1. HBM demand continues to grow;
2. Expansion of server DRAM demand;
3. AI eSSD demand is rising simultaneously;
4. HBM4 volume expansion further occupies wafer capacity.
This means the storage industry is still in a supply tight cycle, rather than demand having peaked. #SKHynixRecordMiss #SKHYNIXPerpsCrash #BigTechEarningsNight If you're buying every green candle right now, you're probably reading the market wrong. The market looks strong on the surface, but under the hood, it's telling a very different story. This isn't broad-based strength. It's a liquidity rotation. A handful of coins are outperforming, creating the illusion that everything is recovering. Meanwhile, capital is becoming more concentrated, not more widespread. The biggest clue? 📊 Open interest is cooling while trading volume remains relatively steadyAmazon AI capital expenditures cannot be fully attributed to AWS: retail, logistics, and cloud require cross-verification
Amazon has officially scheduled its Q2 2026 earnings call for July 30, with the IR page yet to show this quarter's results. The market typically classifies all company infrastructure spending as AWS or generative AI capital expenditures, but Amazon's property and equipment simultaneously serve data centers, fulfillment centers, transportation networks, offices, and other operations. Without precise allocation from the company, one cannot categorize it independently.
The net property and equipment at the end of Q1 increased from $357.025 billion at the end of 2025 to $397.458 billion. This reflects asset scale expansion but cannot be used solely to determine whether the added capacity belongs to AWS or retail. The official 10-Q notes on property, equipment, leases, and commitments can supplement asset types, but if capital expenditures are still not broken down by segment, this article will maintain its limitations.
On the return side, three segments can be used for cross-verification. Q1 AWS revenue was $37.587 billion with operating income of $14.161 billion; North America revenue was $104.143 billion with operating income of $8.267 billion; International revenue was $39.789 billion with operating income of $1.424 billion. If data center investments mainly support AWS, subsequent capacity, revenue, and segment profit should gradually reflect this; logistics investments are more likely to show through delivery speed, unit costs, and retail profit margins.
Cash flow also has timing differences. For the twelve months ending Q1, operating cash flow was $148.531 billion, net purchases of property and equipment were $147.299 billion, and the company's free cash flow was $1.232 billion. This is a twelve-month cumulative figure, not just Q1 alone, and the entire difference cannot be attributed to AWS. Equipment finance leases and debt financing also need separate verification.
The demand narrative must also be layered. Management talks about AI models, chips, Bedrock, or enterprise demand as product and demand signals; only AWS's official revenue, operating income, remaining contracts, and cash flow can enter financial judgment. Retail-side robots and generative AI tools may first improve efficiency and do not necessarily generate independent revenue.
The Q2 results release will present asset formation, segment revenue, segment operating income, and cash recovery side by side to avoid drawing conclusions from a single capital expenditure total for all businesses. If the company provides an official new breakdown, the new data will be used; if not, AWS's share will not be estimated. Before results are released, the company's forward-looking statements, external supply chain figures, or analyst estimates will not be treated as realized Amazon expenditures.
Asset utilization also requires observation over consecutive quarters. New data centers or fulfillment centers may initially bring depreciation and fixed costs during ramp-up, with revenue and efficiency reflecting later; a single quarter's profit margin decline cannot automatically be judged as investment failure. Conversely, segment profit improvement may also come from pricing, mix, or cost control, not solely from new assets. If the company does not disclose capacity utilization, this article will not estimate it independently.