Orbit Post Sitemap
基本面研报 $HNT / Helium（DePIN） $3.20
概括一下：Helium（$HNT）综合评分 49/100，评级 早期项目，验证不足。 三层拆开看，公司团队 有现金储备， 协议网络 已有付费使用痕迹， 代币 捕获已落地。
先看项目：Helium（代币 $HNT），DePIN 赛道。 主打 无线网络DePIN龙头。 对标 GRASS、IOT。 传统做算力租赁的是 AWS、CoreWeave 这些巨头，按 GPU 小时计费，A100 月租金 1.2-2.5 万美元，贵且门槛高。 链上方案把算力碎片化竞价，供应商无需中心化审核，闲置 GPU 变成可用供给。 客单价 50-500 美元/月，需 USDC 或法币结算。叙事驱动型赛道，熊市使用量砍 60-80%。定位端到端垂直平台。 产品落地：协议层已正式运行，链上仪表盘显示协议手续费正在累积，已有付费使用痕迹。 最新版本 未查到，近 90 天有效提交 60 次。
用户层面，地址 MAU 未披露，DAU 未披露，24h 成交额 $80.00M，TVL 未查到。 钱包地址不等于自然人月活，大额地址集中持仓会高估真实用户量。 收入端，用户费用 未披露， 供应方收入大约是用户费用的 80-90%（归 LP 和节点）， 协议金库收入 $2.00M， 代币持有人回购销毁年化 无销毁机制。 24h 成交额是业务流水不是收入。 公司赚钱不等于协议赚钱，协议赚钱不等于代币持有人赚钱。 代码侧，90 天有效提交 60 次，活跃贡献者 25 人， 最新版本 未查到。GitHub 是 A 级证据可以直接核验。 投资背景，公司股权融资看 PitchBook/Crunchbase（A 级）， 代币私募公募看白皮书和释放曲线以及链上解锁合约（A 级）， 做市商和生态资助是 B 级不代表技术 VC 长期持仓， 技术集成看 API/SDK 接入证据（B 级）， 战略合作和 Logo 墙是 D 级。 NVIDIA GPU 被使用不等于 NVIDIA 投资，交易所上线不等于交易所战略投资。
代币侧，总量 1,300,000,000，流通 950,000,000（73.1%）， FDV $4.20B，下次解锁 2026-Q4（占流通 +3.50%）， 销毁回购年化 无明确回购销毁。用产品必须买币？部分需要，中等价值捕获（质押/折扣/治理）。 和同行放一起看（统一口径，不跨赛道乱比）： 流通市值方面，Helium $3.00B，GRASS 未披露，IOT 未披露。 FDV 方面，Helium $4.20B，GRASS 未披露，IOT 未披露。 年化收入方面，Helium $2.00M，GRASS 未披露，IOT 未披露。 月活地址或用户方面，Helium 未披露，GRASS 未披露，IOT 未披露。 数字以公开数据快照为准，部分缺失由官方自报或行业口径补。 估值，流通市值 $3.00B，FDV $4.20B， P/S 1500.0x，FDV 除以收入 2100.0x。 悲观看 $3.00B 打 5-7 折，中性区间震荡， 乐观看收入翻倍、销毁落地、企业客户进来，FDV 对应 P/S 与头部对齐。 最终判断：证据不足，叙事为主（评分 49/100）。代币价值捕获已落地（回购/销毁/Gas）。 流通市值相对基本面偏贵，透支预期，FDV 温和。 需要注意的风险：短期大额解锁砸盘、协议收入长期归零、代币需求仅靠激励（激励断即使用量崩）。 后续跟踪：协议手续费周度、销毁金额、活跃地址留存、TVL/贷款余额、GitHub 版本发布。 以上是公开信息的逻辑和判断，不构成买卖建议。核心财务指标偏离 30% 以上，结论需要重新评估。
先聊到这，有想法评论区见。
#基本面研报 #加密 #研究 #OKXOrbitThe previous research note discussed Coinbase's business boundaries. From spot trading to stablecoins, custody, staking, lending, and on-chain infrastructure, Coinbase is clearly more than just a US-compliant exchange. But just because its business is growing more doesn't mean its revenue structure has fully transformed. What really matters is to ask: After developing so many new businesses, are transaction fees still its main source of revenue? My judgment is also clear: Coinbase has broken free from its single-point dependence on crypto spot trading fees, but it has not broken free from overall dependence on trading activity and the crypto cycle. Trading fees are no longer the only source of income, but they remain the most important profit driver. 1. Coinbase was once a fee machine. Let's go back to the last bull market. In the second quarter of 2021, Coinbase's trading revenue was about $1.93 billion, while subscription and service revenue was only about $100 million What does trading revenue account for nearly 95% of net income—what does this mean? At that time, Coinbase's business model could almost be summed up in one sentence: the bull market came, users started trading frantically, and Coinbase followed to make money; The bear market arrives, trading volume drops sharply, and both revenue and profits shrink. The market decides how many people are willing to trade, and Coinbase decides how much fee it can cut from these trades. This model is strong in bull markets because new trading income doesn't require CoinbaseI actually thought Saudi Arabia had already solved this problem. After the Strait of Hormuz became a "hostage" in the Iranian standoff, they quietly devised an alternative plan: to switch to transporting crude oil via the Yanbu on the Red Sea coast, then push it through the "Petroline" pipeline, thereby completely bypassing Hormuz. At first, the plan seemed stable. Last week completely shattered that assumption.
On July 25, the Houthis directly struck the Jizan refinery and Yanbu, pushing Brent crude prices back above $100. This is not just another headline about the Middle East. Yanbu is precisely the channel that has allowed about 92% of Saudi crude oil exports to continue flowing even after Hormuz has been effectively neutralized. Now, with the Houthis incorporating the Bab al-Mandeb Strait into their maritime blockade, Saudi Arabia truly has no second open corridor that remains open. Kpler tracking data shows that Yanbu's vessel loading volume has dropped by about 30%, and some tankers have even "disappeared" signals to cover up this gap.
What I find interesting is that the cryptocurrency market still seems not to factor in this risk in its pricing. Whenever we see a shock in oil supply, this pattern repeats: inflationary pressures accumulate, the likelihood of Fed rate cuts shrinks, and risk assets are impacted. Altcoins usually react first, while BTC tends to lag behind. If this blockade continues and the extended terminals or pipelines are directly hit again, this will no longer be a simple "oil price story" but will evolve into a broader "risk aversion" event for the market.
If this upgrade continues, do you think BTC will see a pullback first, or will it be an altcoin?SK Hynix rose 17.52%, SanDisk rose 25.99%, Micron rose 18.36%, Microsoft gained 15.51%, and even the VIX Fear Index fell 17.28% in a single day. Just a few days ago, the memory chip sector, which was in turmoil, has staged a stunning reversal overnight.
Such a rebound has never been driven by a single factor. It is the result of multiple forces resonating among South Korea's market rescue, U.S. inflation cooling, Microsoft's earnings, and short squeezes. But before getting excited, there's a question to clarify: is this the start of a reversal, or a standard reprieve?
After several consecutive days of sharp declines, the Korea Exchange (KRX) has internally assessed the technical feasibility and system preparation time for temporarily banning short selling, while also reviewing the feasibility of narrowing the current 30% price limit. South Korea's Ministry of Economy and Finance also held an "Emergency Market Situation Assessment Meeting," after which it stated it would maintain the highest level of market vigilance and activate a 24-hour cross-departmental monitoring mechanism.
This rescue operation is also politically sensitive. The South Korean government had previously actively encouraged retail investors to join the AI chip stock boom led by Samsung Electronics and SK Hynix. Now, retail investors suffered huge losses within 48 hours, putting the authorities in a dilemma—regardless of public opinion, they couldn't bear it; but also afraid of interfering with market pricing.
Because of this, the measures introduced this time appear quite sincere. The anticipated short-selling ban combined with discussions of narrowing price fluctuations can significantly limit short-term selling pressure. For the market, this means at least one thing: the policy bottom has already appeared before the market bottom.My conclusion first: this round of GRVT is more like cross-platform price turnover in the early stages of launch, not that the market has already formed a unified short position. Falling prices are real, and so is the concentration of transactions; But OKX is unwinding, while Binance and Bybit are increasing positions at low levels. Continuing to chase the dip may actually mean worrying about position squeeze. Data as of 17:00 on August 2 (Beijing time). In the completed minute sample from 05:01 to 16:59 in this round, OKX spot dropped from 0.28012 to 0.26060, a decrease of 6.97%, with a turnover of about 18.476 million USDT. The most notable event occurred from 15:00 to 15:59: single-hour transactions reached 13.279 million yuan, accounting for 71.9% of the entire window, with the lowest price at 0.25081. KuCoin's peer declined about 7.4%, with similar directions and magnitudes, so a single platform's bad price can be ruled out. Before the strict window ended, multiple GRVT views had already appeared in Planet search results. Some claimed to be selling proactively, while others suspected market making or airdrop resale. The popularity is real, but these posts only show that disagreements are heating up and cannot replace causal evidence. Position structure is more useful than sentiment. From 05:05 to 17:00, OKX perpetual coin-margined open interest dropped from about 10.741 million to 10.475 million tokens, a decrease of 2.47%. At 16:00, the funding rate shifted to -0.04152%; Binance's price declines were similar, with open interest increasing by 12.8% and Bybit up 18.0%, while Bybit's funding rate remains around +0.Don't be deceived by the financial report numbers! Behind the depreciation period adjustment, the tech AI cycle is entering a critical turning point #折旧年限延至25年，微软资本开支指引下调
Most people only see the surface-level positives: capital expenditure reduction = slower cash burn, profit improvement, directly driving Microsoft stock higher.
But few see through the core truth: this adjustment is not a cut in AI investment, but a compliant accounting change, a financial maneuver under AI arms race pressure, beautifying short-term reports while burying long-term risks.
1. Clarify the underlying facts of the event
Microsoft's two simultaneous adjustments:
1. Depreciation period for data centers and office buildings extended from 15 years to 25 years;
2. Fiscal year 2027 capital expenditure guidance lowered from $190 billion to $175 billion.
Management emphasized: excluding accounting rule changes, the actual scale of compute power and data center construction investment has not decreased; subsequent infrastructure investment will continue to grow.
Many traders confuse a key point:
Extending depreciation applies to building assets; GPU and server hardware still maintain their original depreciation cycles and should not be conflated.
Data center buildings have a physical lifespan of 20-30 years, so the accounting adjustment is reasonable; but the market controversy focuses on the timing of the adjustment amid ongoing massive AI investment cycles.
Plain explanation of accounting logic
Longer depreciation period → less depreciation expense allocated annually → reported operating profit is passively boosted.
Combined with many data center leases shifting from finance leases to operating leases, no longer counted in capital expenditure, causing CAPEX figures to "appear" lower.
Cash outflow remains unchanged; only the timing of cost recognition is deferred.
2. Three common market misconceptions
❌Misconception 1: Capital expenditure cut = Microsoft reducing AI investment
Wrong! The CAPEX number shrinkage is due to accounting method changes.
This quarter's single-quarter capital expenditure reached $41 billion, a significant year-over-year increase, with substantial funds still flowing to Azure compute clusters.
AI demand remains intense; Microsoft has no conditions to actively reduce compute deployment.
❌Misconception 2: Longer depreciation = permanent profitability improvement
Wrong! Depreciation is a non-cash cost.
Reported net profit looks better, but free cash flow does not improve.
Profit is just redistributed across periods, not actual cash earned.
❌Misconception 3: Industry peers following suit means the investment phase is ending
Cannot generalize.
Google and Meta previously adjusted server depreciation, mostly hardware-focused; Microsoft's adjustment this time is more about building assets.
The essential signal: sustained high investment continues to erode profits, tech giants begin proactively optimizing report expectations to reassure secondary market funds worried about "endless cash burn."
3. Core logic breakdown from bullish and bearish perspectives
✅Bullish support logic
1. Eases market anxiety: investors' biggest worry was endless compute investment suppressing margins. After adjustment, short-term earnings expectations are supported, supporting valuation;
2. Financial rules compliance, no violation risk; institutional funds are willing to accept this accounting estimate change;
3. Azure and Copilot commercialization continue to deliver, revenue growth can offset infrastructure cost pressure;
4. Cooling capital expenditure expectations dispel pessimistic pricing about "endless capital devouring cash flow."
⚠️Bearish potential risks
1. Long-term "deferred depreciation risk": if future industry cycles decline or data center utilization falls short, long-term depreciation pressure will suppress profits for years;
2. Asset attribute distinction: buildings can be used for 25 years, but internal AI chips face iteration and obsolescence every 3-5 years. Hardware updates continue, so long-term capital expenditure pressure won't disappear;
3. Demonstration effect: if major cloud providers collectively use accounting adjustments to beautify short-term reports, a unified correction later could trigger a round of collective earnings downward revisions;
4. Narrative inflection signal: when giants start relying on accounting tricks to optimize reports, it indirectly indicates the increasing difficulty of relying solely on revenue growth to absorb massive investments.攻击者没有偷走设备，没有骗到助记词，甚至没有碰过受害者的钱包。 被攻破的，是这串助记词原本应该拥有的“不可猜测性”。 截至目前，Galaxy Research通过链上交易归集发现，约1082.65 BTC在6个区块、41分钟内被转走，涉及1196个地址，按当时价格计算约7000万美元。 这仍是链上研究机构的动态估算，不是Coinkite公布的最终损失；1196个地址，也不等于1196名独立受害者。 但漏洞本身已经得到官方确认，而且风险不只停留在理论层面。 最值得警惕的是：整个攻击过程不需要让冷钱包联网。 冷钱包很冷，但种子可能从出生时就不安全 硬件钱包主要完成三件事： 生成种子、保存私钥、签署交易。 不联网能够保护后面两件事，却无法自动保证第一件事一定正确。 正常的助记词，应该像从一个大到几乎无法穷尽的号码池中随机抽出一张票。 这次的问题是：设备看起来仍在生成随机结果，实际使用的号码池却比预期小得多。 攻击者不需要进入你家，也不需要拿到设备。他只需要在自己的计算机上批量尝试候选种子，再检查对应地址是否持有BTC。 只要搜索空间足够小，所谓“永不联网的私钥”一样可能被找到。 一个存在了五Maximizing capital selectivity amid the BTC-led phase, the temperature difference between bearish stocks and rotation-expected stocks is the key variable. Now that relative strength has become clear, could the movement and withdrawal of market funds determine future returns? A key fact confirmed in the recent market is that the distribution method of liquidity has changed. It is a selective phase where capital concentrates on already verified assets, rather than a broad rise that lifts the entire market simultaneously. This means that risk appetite is not completely off, but profit opportunities are compressed into specific stocks. Looking at the current structure based on relative strength, BTC maintains its role as the market anchor and absorbs the largest share of new liquidity. ETH is solidifying its status as a trusted asset in the market with steady institutional demand inflow, and SOL has established itself as a momentum stock that reacts first when risk appetite recovers. WLD is forming its own demand base as the AI narrative continuously attracts investor interest. DOGE and ZEC are retail #财报观察员：下周四场开奖，Circle压轴
Circle will release its Q2 2026 financial results on August 5, 2026, and hold a conference call. As a leading stablecoin issuer listed on the US stock market, Circle's earnings report is an important window to observe the digital dollar ecosystem, reserve yield models, and regulatory progress.
Circle's core business model heavily depends on the circulation scale of USDC and the yield from reserve assets. The company primarily allocates the US dollar reserves backing USDC into short-term US Treasury bonds and cash equivalents to earn interest spreads, while gradually diversifying revenue through subscription services, transaction fees, and other income.
Reviewing the previous quarter's performance, Q1 2026 total revenue and reserve income combined reached $694 million, a 20% year-over-year increase. Reserve income was $653 million, up 17% year-over-year, mainly driven by a 39% growth in average USDC circulation, partially offset by a 66 basis point decline in reserve yield. Other income was $42 million, showing significant year-over-year growth. Adjusted EBITDA reached $151 million, up 24% year-over-year. USDC circulation at quarter-end was $77 billion, up 28% year-over-year; on-chain transaction volume reached $21.5 trillion, a substantial 263% year-over-year increase. USDC's share of stablecoin transaction volume has risen to approximately 63%.
Entering Q2, market consensus expectations for performance roughly range from $700 million to $740 million in total revenue (including reserve income), with EPS around $0.17 to $0.22.
We should focus on the following key dimensions:
First, the average and quarter-end USDC circulation. It was $77 billion at the end of Q1, with a phased decline in circulation observed entering Q2 and July, with recent market data around $72 billion. It is necessary to observe whether net redemptions are due to crypto market volatility, exchange fund outflows, or seasonal factors, and whether institutional custody and payment scenarios continue to support the scale.
Second, reserve yield. The current federal funds target rate range remains at 3.50% to 3.75%. Yield pressure was already evident in Q1. If the Federal Reserve maintains a high interest rate environment, it will continue to support reserve income; if market expectations shift toward rate cuts, it will directly pressure core profits.
Third, other income and platform business progress. This includes subscription and service income, transaction-related income, CPN network transaction volume, as well as new business contributions from the Arc network, tokens, and the Agent Stack aimed at the AI-native economy. These projects are key paths for the company to reduce sensitivity to interest rates.
Fourth, costs and profit margins. Operating expenses rose significantly in Q1, including factors such as post-IPO stock compensation. The market will watch for expense discipline and whether economies of scale begin to emerge.
The company has also made positive recent progress. On July 31, it obtained a limited purpose trust license from the New York Department of Financial Services, having previously received final approval from the OCC National Trust Bank, further strengthening its moat as a regulated stablecoin issuer. It also completed the acquisition of IBM's blockchain patent portfolio and continues to advance the Arc network and AI Agent infrastructure. On the macro front, the Federal Reserve held rates steady at the July meeting, with market divergence on the subsequent path. High rates benefit reserve income, but if economic data trigger rate cut expectations, that will be the biggest variable.
Key risks include the direct impact of rate declines on reserve income, competition with USDT and share battles in emerging payment scenarios, legislative details such as the Clarity Act potentially affecting revenue distribution models, crypto market volatility's linkage to USDC demand, and significant stock price fluctuations since the IPO.
Overall, Circle's Q2 earnings will verify two things: first, the real growth resilience of USDC in institutional and compliant scenarios; second, whether the company can continue to improve profitability quality under interest rate and cost pressures while accelerating the proportion of other income. If circulation remains stable or rebounds, reserve yields meet expectations, and new businesses make substantive progress, it will strengthen the market's long-term logic for regulated digital dollar infrastructure. Conversely, if circulation continues to shrink or costs spiral out of control, short-term stock price pressure may increase. 💡 Core viewpoint: Refined defense, focus on liquidity
1. Bottom-fishing defense: Trading space for survival
Bottom-fishing near 1000, hard stop loss (999-930), easily shaken out by a hollow break. Defense should be layered:
900 points (first defense): If it falls below 972, there is a high probability of braking at 930. 900 points only reduced interest by 20%.
< 850 Strong Tie Line (Second Defense): Keep the tie below 850, completely immune to the 900-999 gap gap.
2. The harsh truth about liquidity: precise team-up
The market is extremely picky, and funds no longer rise broadly. Stocks like $BEAT are not just building momentum, but falling behind completely.
🔥 Capital inflows: $JTO, $JELLYJELLY, $OPG, $SLX, $LAB, $ALLO, $CHIP
❄️ Capital outflows: $BEAT, $EDGE, $COAI, $TRUMP, $RAVE, $SPACE, $VIRTUAL
👀 Watchlist: $MEME, $EDEN, $HUMA, $ZKP, $METIS 3.
Leading Stock Landscape Hard Map: $ETH Timing to Buy the Dip?
When a coin is everywhere, smart money has already eaten the meat. Currently, $ETH is in a left-side bottom-fishing phase, ignored by retail investors and quietly accumulating shares by institutions.
Core assets: $BTC Stabilizing liquidity and $ETH attracting institutional funds.
High-Burst Vanguard: $SOL is the strongest Beta option; $TAO + $WLD dominates the AI narrative.
Sentiment indicators: $HYPE reflects risk appetite, $DOGE + $ZEC represents retail sentiment.
4. Macro Storm: U.S. Treasuries remain high, negative news has all been exhausted
U.S. Treasury yields hit a 19-year high: liquidity continues to tighten, and funds will only band together to strengthen the narrative.
Amazon's guidance falls short of expectations, but rises 9%: classic negative news is gone. Market expectations have hit bottom, and real buying decisions determine everything.
#30年期美债收益率创19年新高
#财报观察员: Next Thursday's draw will be held, with Circle as the grand finale At 3 a.m., I stared at the liquidation map and suddenly realized this round of decline wasn't just news-based. Have you ever wondered why, for the same geopolitical news, the US stock futures haven't even moved yet, but the crypto world has already collapsed? To start with the conclusion: the most vulnerable link in the market has never been spot trading, but the leverage of derivatives piling up. At this time last night, BTC suddenly broke through 62,500 and quickly recovered. On the surface, it seemed like a sudden negative news from the US-Iran conflict, but if you look at the liquidation data, you'll find that a large number of long positions are densely accumulating in the 62,000-62,500 range. Once triggered at this level, it triggers a chain reaction of liquidations. BTC is now back above 62,500, but don't be too quick to relax. What truly deserves attention is not the price itself, but the change in open interest at this level. If open interest decreases instead of increasing during the rebound, it indicates that the rebound is driven by short covering, not by new bull entries. Market liquidity was as thin as a cicada's wing over the weekend, and any news could be magnified into a large bearish candlestick. ETH's situation is even more intriguing. After breaking below the 1850 support, the low reached 1820, but after rebounding but unable to rise above 1850, this is itself a weak signal. On the daily chart, the MA20 is showing clear suppression, with 1890 marking the dividing line between bulls and bears. On the counterfeit side, SOL's monthly chart with ten consecutive losses is already quite rare; this level of weakness can't be fixed in just a day or two. Cross-market collaboration is the highlight today. The Korean market hasn't opened yet, and US stock futures haven't reacted, but the crypto world has already fallen first. What does this indicate? This explains this77,402 $BTC from older UTXO bands has moved since the Coldcard seed phrase bug became public.
The biggest takeaway? People are not ignoring the risk. They’re acting.#美方酝酿打击伊朗能源设施, the embassy issued an evacuation warning
Trump said Iran requested a pause, and the Iranian military called this a "new lie."
On August 1, Trump posted on social media that Iran and other Middle Eastern countries had requested the U.S. to temporarily halt the attack. According to him, the parties have already formed a preliminary framework for the agreement, including the immediate and full opening of the Strait of Hormuz, as well as addressing the Iranian nuclear issue; He therefore agreed to cancel the original military operation, but provided the agreement could be finalized quickly, and Israel would make the same commitment.
However, according to Xinhua News Agency citing a report by the Mehr News Agency on August 2, an Iranian military official said that Trump's claim that "Iran demanded a halt to the attack" is a new lie. The official stated that whether the U.S. continues to act or retreats, Iranian forces remain on top alert. Previously, Iranian Foreign Minister Alagazi also warned the U.S. through intermediaries against "risky actions," stating that Iran would respond firmly to any infringement.
Currently, no third party has publicly confirmed whether Iran has directly made the request, nor has there been any sign of the agreement text, formal negotiation schedule, or arrangements for reopening the Strait of Hormuz.
But the Strait of Hormuz is the most direct verification line.
Xinhua News Agency cited data from the shipping platform stating that on July 30, the number of vessels passing through the strait dropped from 22 the previous day to 5, a decrease of 77%. Although the channel has not completely cut off flow, safety risks, insurance costs, and detour pressures have already affected crude oil transportation. If the so-called agreement framework is implemented, vessel traffic, tanker insurance fees, and Gulf crude oil exports should improve first; If the channel remains restricted and the alert status is not lifted, then the "cancellation operation" is more like a conditional pause, with the risk not truly gone.
During the current weekend, crude oil, gold, and US stocks have yet to see a new round of pricing. According to OKX spot market data, as of 14:00 Beijing time on August 2, $BTC was about 63,455 USDT, up about 0.67% in 24 hours. The crypto market has not experienced synchronized panic, but the performance of crypto assets is insufficient to represent that energy and global risk assets have accepted the ceasefire narrative.
Therefore, whether the final cooling can shift from verbal easing to actual easing depends on whether the middle party confirms the framework of the agreement, whether the U.S. continues to cancel its original actions, whether Hormuz navigation can resume, and whether a timetable for formal negotiations is provided.$MU MU fell 6% on Friday, but what I see is not a worsening demand situation.
The latest closing price was MU at $823.03.
During the session, the highest surged to $930.33, the lowest to $812, with a single-day volatility close to 15%.
What does that mean?
Those who chased at the opening could get a pullback of more than ten points that day.
Moreover, throughout July, MU's cumulative pullback was close to 29%. Previously, people were shouting about the "AI memory supercycle," but now the comment section has started discussing whether the storage bubble has burst.
But there is an interesting contradiction here.
MU's stock price is falling, but downstream customers are still complaining:
Memory is too expensive.
In Qualcomm's latest financial report, management directly mentioned that memory prices and supply issues are squeezing profits, and the earnings guidance for next quarter is also below market expectations.
This actually indirectly proves that DRAM and high-end memory are still in short supply.
If storage demand really collapses, downstream manufacturers will likely first feel price cuts, not rising costs.
Micron previously stated that all HBM capacity by 2026 has been sold out, and the next-generation HBM4 has begun mass production. The company has also signed long-term customer agreements worth approximately $22 billion.
So MU's core conflict now isn't "whether there's demand."
Instead:
With such strong demand, how much has the stock price been overdrawn in advance?
Earlier, MU's stock had grown several times in a year, with its market value once surpassing one trillion dollars. The market is no longer just buying HBM's strong sales this year; it is expected to be in demand for the coming years, with prices continuously rising, and cloud vendors constantly expanding data centers.
If any one of these signs loosens, the stock price will quickly crash into valuation.
So this drop won't be simply interpreted as the end of the storage cycle.
But I wouldn't think a 29% drop means it's definitely cheap.
Next, I mainly look at three signals:
First, whether MU can sustain support around $800;
Second, whether DRAM and HBM prices will continue to remain strong;
Third, downstream manufacturers are cutting back on procurement, or are they complaining about prices while continuing to compete for stock?
If customers keep saying it's expensive but orders don't decrease, MU's fundamental logic still holds.
But if cloud providers start delaying procurement, or storage vendors collectively ramp up capacity, the situation will be completely different.
What MU is most worth watching right now isn't how much they've dropped in a single day.
Instead, while trading demand peaks, downstream players are still struggling with memory price hikes.
In both cases, one side is always misunderstood.
This is for personal market observation only and does not constitute investment advice. DYOR.AI Volatility Check: Cloud AI is stable, but memory remains under pressure
$QQQ Trading range is the 1.4% intraday high to low, currently rebounding to an intraday high near $676, while $SPY is approaching the intraday high of $741. This divergence may still be related to pressure on domestic Chinese inventory and semiconductor equipment, while cloud AI remained relatively strong on the first day of the July 28-29 Federal Open Market Committee (FOMC) meeting.
$SOXX After experiencing a 2.9% intraday high to low drop, it remains in the intraday weak zone of $491. $MU was quoted at $817, $SNDK at $1,091, $WDC at $440, $AMD at $455, all rebounding from their lows but still well below the opening price. $SNDK fell 9.6% from the opening to the low, and $WDC dropped 8.2%.
$QQQ rebounded 1.3% from the low, but the semiconductor and memory sectors have yet to recover similarly significant losses.#亚马逊半年给 OpenAI 砸了 287 亿美元
亚马逊的 10-Q 文件刚曝出来：上半年给 OpenAI C 轮优先股投了 287 亿美元，Q2 单季度 137 亿。另外还给 Anthropic 投了 13 亿可转债。
这不是小数目。287 亿够买半个币安了。
放在加密视角看，四大云厂现在两三家都在 all-in AI，传统科技资本被 AI 基建吸走，短期对链上资金是抽血。但长期"AI×Crypto"叙事有真实资本背书。
你最看好 AI 还是 Crypto？评论区站队。
#亚马逊 #OpenAI #AI基建Hassett says "it's hard to push for rate hikes"—is a weaker dollar good for crypto?
White House National Economic Council Director Hassett stated on July 31: Based on current data, it is difficult to push for a rate hike.
This is the clearest recent signal of "no increase." U.S. initial jobless claims were 197,000, the lowest since April; But GDP is weak, and the dollar is under pressure. The Fed kept rates unchanged at its July meeting, leaving much room for uncertainty for Chairman Walsh's next steps.
No rate hikes = weak US dollar = risk assets have some breathing room. But this logic has a premise: inflation should not rebound.
Poll: What do you think the Fed will do in the next three months?
A. Continue to hold your position
B. Unexpected rate hikes to curb inflation
C. Cut interest rates first
#美联储 #加息 #哈塞特 #美元Circle's earnings report before the market on August 5th. #财报观察员：下周四场开奖，Circle压轴
CRCL stock price has dropped from the historical high of 299 in June 2025 to around 60 now. A decline of 80%.
A company with nearly 7 billion in annual revenue and 694 million profit in Q1 now has a market cap of less than 20 billion.
The problem is the market no longer pays a premium for the "first stablecoin stock" label; they want to see USDC grow again.
Q1 USDC circulation was 77 billion. By July 27th, it had dropped to 72.3 billion. Nearly 5 billion less in two months. Circle's core business model is simple—USDC reserves buy U.S. Treasuries to earn interest, which is income; the larger the USDC circulation, the more they earn. Circulation is falling, interest income is shrinking. Q1 reserve interest income was 653 million, accounting for 94% of total revenue. If Q2 circulation continues to decline, even with high interest rates, revenue will be hard to maintain. The Hong Kong Stock Connect AI infrastructure index was launched, and the AI narrative in A-shares spilled over
On July 31, Hang Seng Indexes Company launched the Hang Seng Hong Kong Stock Connect AI Infrastructure Index. Covering suppliers of semiconductors, electronic equipment, infrastructure, and hardware resources, all can be traded via the Hong Kong Stock Connect.
This move is clear: Hong Kong stocks are competing for the overflowing funds from A-share AI narratives. Previously, Changxin topped the A-share market and the Korean stock market plunged 8%, both with the same batch of funds moving back and forth within the AI industry chain.
Significance for crypto: AI hardware stocks and crypto AI narratives share the same "AI demand" storyline. When traditional markets start making index products for AI infrastructure, it shows that this narrative has become institutionalized.
Have you ever bought Hong Kong Stock Connect stocks? Will you participate in this kind of AI index fund?
#港股通 #AI基础设施 #恒生指数Circle has obtained a national trust bank license.
This matter is far more important than short-term earnings reports.
With this license, Circle can hold client assets on its own without relying on third-party banking partners.
Previously, Circle's USDC reserves were scattered across multiple banks, resulting in high compliance costs and complex processes.
With your own trust bank, reserve fund management can be completed directly within your own system, with lower costs and higher efficiency.
This could generate about $20 million in additional revenue annually.
On July 10, it also purchased IBM's blockchain patent portfolio, adding over 680 patent families. The direction is clear—Circle is transforming from a "USDC issuer" to a "blockchain infrastructure company." CRCL's stock price logic has changed. Previously, it was "the larger the USDC circulation, the more profitable"; now, "the more complete the USDC ecosystem, the more valuable it is." Competition in the stablecoin sector is intensifying—Open USD is jointly issued by over 140 institutions. Circle's moat is not USDC itself, but the infrastructure behind it. What is truly worth watching in the financial report is not the specific USDC circulating supply figures, but the progress of trust banks. #财报观察员: Next Thursday's draw will be held, with Circle as the grand finale Swan Treasury 私钥泄露，62.5 万美元 100 倍折扣被搬空
BNBChain 上去中心化资管协议 Swan Treasury 出事了。
私钥泄露。攻击者伪造有效签名，以 100 倍折扣买 STY，再用 PancakeSwap 闪电贷砸盘，最终套走约 62.5 万 USDT。
链上记录很清晰：每个 ecrecover 都解析到同一个硬编码签名者，说明密钥本身泄露了，不是签名逻辑有漏洞。
这个案子值得注意的点：私钥泄露不是小概率事件，尤其在 DeFi 协议里。很多团队把私钥硬编码在合约或服务器上，一旦被反编译就裸奔。
你在 DeFi 协议里存过钱吗？选协议的时候有没有查过它的私钥管理方案？
#SwanTreasury #私钥泄露 #BNBChain #DeFi安全CRCL's current valuation is less than 20% of its historical peak.
What is the market pricing in? It is pricing in the continuous decline in USDC circulation, intensified stablecoin competition, and Circle's resilience during the crypto bear market.
However, several data points may be underestimated. On-chain USDC transaction volume in Q1 reached 21.5 trillion, a year-over-year increase of 263%. While circulation is decreasing, transaction volume is surging, indicating that USDC's use cases are expanding beyond just being a reserve asset. Newly minted coins in Q1 totaled 73 billion, and after bottoming out in February, they climbed back to 77 billion by the end of the quarter amid geopolitical tensions. Circle's recovery capability might be stronger than the market expects. The company estimates that for every 1 percentage point increase in interest rates, reserve income increases by approximately $773 million over 12 months. Although distribution costs also rise by about $384 million, net earnings remain positive. The Iran war pushed short-term interest rates higher, partially offsetting the impact of declining USDC balances. Circulation is down, but each dollar earns more.
CRCL fell from 299 to 60, a drop that already factors in significant pessimism. If the Q2 earnings report shows USDC circulation stabilizing and trust licenses progressing smoothly, the rebound potential could be greater than the market anticipates. If circulation continues to decline, 60 might not be the bottom. The key is the direction and trend, not the numbers themselves. #财报观察员：下周四场开奖，Circle压轴 AI Tape Update: Seagate and SK hynix have completely reversed their pre-market movements.
$SOXX $488, down 5.6% at 8:20 AM (Central Time). $WDC shifted from a pre-market gain of 2.1% to a decline of 4.9%. $SNDK fell 15.2%. $MU fell 9.3%. $NVDA Basically flat.
Seagate instructs first-quarter fiscal year revenue of $4.1 billion with a float of $100 million ± non-GAAP earnings per share of $7.30 and a float of $0.20 ±, but this guidance only covers HDDs. SK hynix's second-quarter revenue grew 257% year-on-year to 7.93 trillion KRW, with an operating margin of 76%. Although it did not meet high expectations, the market is still prioritizing the highest beta value of memory and long-term stocks.The SPCX restricted share unlocking design is very complex.
The first tier unlocked 20% in two trading days after the earnings report, which is 911.5 million shares.
The second tier: if the closing price is above 175.5 yuan on 5 out of the 10 consecutive trading days before the earnings report, an additional 10% is unlocked, about 455.8 million shares.
SPCX is currently 108 yuan, nearly 40% short of 175.5, making it almost impossible to trigger this condition. Tier 3, with an additional 7% to be unlocked gradually.
Musk is the largest holder, with his net worth evaporating by over $600 billion in just over a month. He himself has never sold a single share, but the market is trading "he can start selling anytime." Bears are betting that insiders will start selling after the lock-up is lifted. Good earnings may cause bears to be squeezed; Poor earnings report, combined with selling pressure from the lock-up lift, could push the price directly above 100. Both sides are betting on this.
#SPCX首份财报将公布, the $100 billion ban is about to be lifted 上周美股最后收得还行，Amazon 大涨，AI 和云计算这条线继续撑住市场。
但 Apple 下跌也很明显，科技股内部不是全面强。
这对币圈的影响就是：风险偏好有修复，但还不够顺。
BTC 现在没完全跟上美股，说明币圈内部还有自己的压力。
我今天会看两个信号：
BTC 能不能收回 63000。
ETH 会不会比 BTC 更抗跌。
如果美股强，BTC还弱，那周末山寨别乱追。SPCX's current short bet is already a high-stakes gamble.
219.3 million shares shorted, with a notional value of $24.6 billion, even larger than Tesla's short position.
The shorts' logic is clear — a flood of unlocks worth hundreds of billions will crush the valuation.
The bulls' logic is also clear — if the earnings report surprises, especially if AI and Starlink progress exceeds expectations, the shorts will be squeezed hard.
Morgan Stanley stated plainly in their report: what truly influences market sentiment may not be the financial figures, but management's comments on the future development of AI, Starlink, and Starship. The Cursor acquisition is expected to close later this quarter, so the Q2 earnings won't disclose specific financial contributions. Investors want to hear about the Starship development timeline, AI computing power deployment speed, and progress on the Grok and Cursor models. Morgan Stanley estimates that the enterprise AI business accounts for 15.2 billion in SpaceX's valuation, more than half. The market may be underestimating or overestimating the value of the AI business.
Before the earnings report, SPCX's stock price had fallen for several consecutive days, closing at 108.37 on July 31, with an intraday low of 107.01, down 20% from the IPO price of 135, and down 52% from the all-time high of 225.64. Two days after the earnings report comes the unlock worth hundreds of billions — if the earnings disappoint, the selling pressure post-unlock could push the price below 100. If the earnings exceed expectations, shorts may be forced to cover. This earnings report is not for reading, but for betting on the direction. #SPCX首份财报将公布，千亿美元解禁在即 8月1号，30年期美债收益率首次突破5.2%，收在5.23%附近，创2007年以来最高。
这不是美联储加息导致的——7月FOMC刚按兵不动，3.5%-3.75%没变。
短端没动，长端狂飙，收益率曲线剧烈陡峭化。
2年期4.23%，10年期4.67%，30年期5.21%。长端比短端高了将近100个基点。
市场在用脚投票——不相信通胀能被压住。圣路易斯联储总裁穆萨勒姆8月1号直接说了：“30年期美债收益率被推到5.2%以上的抛售潮，是对美联储信誉的警告”。他随即加入哈马克、卡什卡利、洛根的行列，成为第四位公开支持立即加息的官员。9月加息概率已经干到80%以上。沃什撤掉了前瞻指引，市场用5.2%的长债利率回应他——你不给路标，我们就自己定价，而且定得比你想的更狠。
比特币从65,000跌到63,000附近，不是基本面出了问题，是宏观在挤压所有风险资产的水分。美债收益率每涨1%，风险资产的估值就要重算一次。#30年期美债收益率创19年新高 The "bond guardians" are back.
The last time this term was widely discussed was in 1994.
At that time, the market did not believe the Fed could curb inflation, so it frantically sold off long-term bonds, pushing yields up to force the Fed to raise rates.
In August 2026, the same script is being repeated.
Former PIMCO CEO Mohamed El-Erian said a few days ago: "The Fed has lost control over long-term interest rates." "This is no exaggeration—Walsh can set short-term rates, but long-term rates are decided by the market. When the 30-year yield surged to 5.2% without short-term rates moving, the Fed had effectively lost control over the yield curve.
What's even more worth considering is the performance of the US dollar. In theory, a surge in U.S. Treasury yields should have boosted the dollar, but this week the dollar actually weakened. The traditional interest rate logic has failed—the market is not trading "the US economy is strong, so interest rates are high," but rather "the Fed can't suppress inflation, so bonds are worthless." The direct impact path for Bitcoin is clear—a risk-free yield above 5% means the opportunity cost of holding Bitcoin has become so high that it's hard to ignore. If there really is a rate hike in September, the market may have to test further downside. The key is August CPI—if the data beats expectations again, long-term yields could surge directly to 5.5%. #30年期美债收益率创19年新高 📉 $ETH This wave isn't a sell-off, it's a discouragement
The $ETH needle that was just chased in is probably slapping the thigh now. But I want to say: don't panic, this bearish candlestick isn't that scary.
Why? Look at the capacity.
🔥 During the recent surge in trading: trading volume skyrocketed, and those who were impulsive to buy long or afraid of missing out all rushed in—a massive volume bar.
📉 Now falling and moving sideways: Volume has not continued to expand; instead, it has contracted.
What does this indicate? It's not the main players who are dumping and selling off. After that rally, the following funds didn't keep up, and no one wanted to buy at the high level. Seeing no one to carry the profitable stock, they slowly dropped their own price.
This is a typical case of a "shrinking volume pullback."
It is not a beast-like escape, but a wise moment after the passion has subsided. The price spike was too fierce, and I used up all my strength. Now I need to catch my breath.
---#以太坊主网十一周年: Eleven years of uninterrupted operation and ecological achievements
🧠 What should we watch next?
· If the price continues to consolidate sideways with shrinking volume: it indicates selling pressure is drying up, and there is still a chance to wash it out.
· If the volume breaks below key support: that is the real signal to be cautious about.
At this level, shrinking volume and pullbacks are not scary. The scary part is panicking and cutting losses during shrinking volume, then chasing in when volume surges—that's the classic scenario of chasing rallies and cutting losses.
#机构逆势加仓: SharpLink increased its holdings by nearly 40,000 ETH #以太坊验证者退出队列已降至零 收益率曲线陡峭化这件事本身，比5.2%这个数字更值得琢磨。
正常加息周期里，曲线是走平的——美联储加息拉高短端，市场预期经济降温，长端上不去。
现在的情况完全反过来——短端没动，长端狂飙。这叫“熊市陡峭化”，通常出现在市场开始怀疑央行信誉的时候。
1994年出现过一次，2022年也出现过一次，两次之后美联储都被迫加息了。
资深美联储观察家Jon Hilsenrath在记者会后写道：沃什的模糊立场正在引发一场“信誉危机”。这说法很重。美联储历史上每次面临信誉危机，最终都要靠更鹰派的行动来重建市场信任。
如果8月CPI再次超预期，30年期美债收益率可能直接冲5.5%甚至6%。到那时候，所有依赖长期借贷的资产——从房地产到科技股到比特币——都会被重新定价。不要只看比特币今天跌了多少，要看美债收益率的方向。它才是真正的总开关。
#30年期美债收益率创19年新高 The Digital Asset PARITY Act Is Moving. Here's What It Actually Means.
US lawmakers are pushing the Digital Asset PARITY Act ahead of the August recess, and the CFTC's 12-month crypto sprint wraps up around the same time. This is the most concentrated period of US crypto regulatory activity in years.
The PARITY Act is about jurisdiction: specifically, which assets fall under CFTC oversight versus the SEC. That might sound like a technocratic debate, but it has real practical consequences. Assets classified as commodities get significantly different regulatory treatment than securities, and right now most tokens sit in a legally contested grey zone.
Goldman Sachs' survey puts it directly: 35% of institutions cite regulatory uncertainty as their biggest blocker, and 32% say clarity is the top catalyst that would bring them in. This isn't abstract policy, it's directly tied to whether institutional capital flows into crypto this cycle or waits for the next one.
The comparison to Europe and Asia is worth making. MiCA and Singapore's MAS stablecoin framework have given traditional finance players clear rails to operate on. The US being behind that curve has real costs in terms of where activity and capital choose to locate.
Does the PARITY Act change how you're thinking about the US market this cycle?
Share your thoughts in the comments 👇BTC Dominance at 60%. Here's Why It's the Most Misread Chart in Crypto.
BTC dominance at 60.3% is widely read as bad news for alts. On the surface that makes sense: if Bitcoin holds more of total market cap, less flows elsewhere.
But dominance at 60% in a declining total market is a completely different story than dominance at 60% in a growing one. Right now, broader market pressure means high dominance often reflects capital retreating to perceived safety, not a permanent structural shift away from alts.
What historical cycles actually show: dominance tends to peak, then rotate hard. In early 2021, BTC dominance was near 70%. By May it had dropped to 40%. That rotation didn't happen in a straight line, but when it moved, it moved fast and made generational alt returns in weeks.
The more interesting question is what narrative breaks dominance lower when it finally turns. In 2021 it was DeFi and NFTs. In 2024, Layer 2s. In 2026 it's shaping up to be AI protocols and RWA. Different stories, same mechanic: capital chases whatever feels like the next thing.
What do you think is the catalyst that breaks BTC dominance lower this cycle?
Share your thoughts in the comments 👇DeFi Had 121 Hacks in 2026. When Does This Become an Existential Problem?
121 exploits. $942 million lost. And we're only seven months into the year.
Q2 was the worst quarter on record for DeFi security incidents: 85 attacks, roughly $775 million drained. That's not a rough patch, it's a pattern. At these loss rates, the most important questions in DeFi aren't about yields or tokenomics, they're about whether the security infrastructure is keeping pace with the capital flowing in.
The frustrating thing is that most of these exploits aren't novel. Reentrancy attacks, oracle manipulation, access control failures: these are well-documented vectors with known mitigations. The gap isn't knowledge, it's execution. Teams ship fast, audits lag, and the incentive to move quickly still outweighs the incentive to move safely for too many protocols.
For users, the calculus is shifting. High TVL used to be shorthand for safe. Now you have to dig deeper: audit quality, team track record, insurance coverage, and whether a protocol has survived real attack conditions. Not the most exciting research, but probably the most important one right now.
At what point does this security track record start routing serious long-term capital away from DeFi permanently?
Share your thoughts in the comments 👇76% of Institutions Plan to Expand Crypto Exposure. The Market Doesn't Show It Yet.
The numbers coming out of institutional surveys are striking. 76% of global investors plan to expand digital asset exposure this year. Nearly 60% expect to allocate over 5% of AUM to crypto. Goldman Sachs calls regulatory reform the single biggest catalyst for the space.
And yet the market is down. DeFi TVL is bleeding. ETF outflows ran eight weeks straight before a modest recovery. So how do you square it?
Part of the answer is timing. "Plans to expand exposure" is different from capital actually deployed. Regulatory clarity in the US is still incomplete, and 35% of institutions cite uncertainty as their main blocker. Goldman's own data shows 32% would enter if clarity improved. The capital is waiting on conditions, not conviction.
The other part: institutional money is patient. When they say they're expanding exposure, they mean over a 12 to 24 month horizon. What the surveys really tell you is where the structural bid will come from when conditions align. Macro, regulation, and on-chain fundamentals all have to cooperate.
We're not there yet. But the directional signal is real.
Is institutional adoption a catalyst for this cycle or the next one?
Share your thoughts in the comments 👇The market is diverting trading funds between two types of chip stocks: on one side tracking AI order fulfillment $MRVL, on the other betting on process $INTC that can turn things around. This divergence reflects the framework game between high valuation momentum and low-expectations reversals. If the yield rate of 18A breaks through and external foundry contracts are signed, the revaluation logic will dominate the upward trend; If AI customers slow down in pulling goods, momentum premiums will shrink. If macro discount rates fluctuate sharply, both valuation anchors will fail. The next step is to observe the actual CapEx spending of major companies.
#亚马逊向OpenAI投500亿美元: Bet or bubble #HYPE再遭亿元解押, Japanese companies enter for the first time #财报观察员: Next Thursday's lottery draw, Circle will be the finaleExchange Stablecoin Reserves Drained $2.3B. Here's Why That Matters.
Major crypto exchanges shed roughly $2.3 billion in stablecoin reserves over the past 30 days. That's the dry powder retail and leveraged traders use to buy spot Bitcoin on centralized platforms.
Stablecoin reserves are one of those on-chain signals that doesn't get enough attention until it does. When reserves build, capital is sitting on the sideline ready to deploy. When they drain, it means one of two things: people withdrew to go elsewhere, or they already spent it and there's less buying pressure left behind.
The timing is what makes this interesting. Bitcoin is sitting below its 20-day moving average. ETF inflows are returning but modest. If stablecoin reserves are thin and institutional buying is cautious, the market is missing a key demand driver right when it needs one.
This doesn't guarantee price falls from here. Markets can recover with lower liquidity if catalysts are strong enough. But watching whether reserves rebuild over the next 2 to 3 weeks will tell you a lot about whether a real recovery is forming or if we're just bouncing in a low-energy environment.
What's your read: is dry powder genuinely running low, or is capital quietly repositioning off-exchange?
Share your thoughts in the comments 👇#折旧年限延至25年, Microsoft's capital expenditure guidance was lowered
Microsoft extended the depreciation period for data centers/office buildings from 15 years to 25 years, and cut its FY2027 capital expenditure guidance from 190 billion to 175 billion, causing the stock price to jump 8% after hours.
But don't be fooled by the title—this isn't AI cooling down, it's accounting maneuvering.
My breakdown method
• Substantial investment: CFO Amy Hood said that, aside from accounting standards, the 2026 calendar year Capex forecast remains unchanged, with over 50 billion still needed for FY27 Q1, and full-year Capex continuing to grow year-on-year.
• Money for Location: Term Changes → More data center leases shift from "financial leasing" to "operating leases" → The former goes to Capex, the latter into operating cash flow. Capex reports show Capex is 15 billion less, but rent is still paid and GPUs are still bought.
• Depreciation thinning: 25-year amortization is lower than 15-year amortization, making the income statement look better, but the positive impact on 2027 operating profit is "minimal."
• Hidden bet: A building's physical lifespan of 25 years is fine, but GPU/switches replace every 3–5 years. Stacking the shell at 25 years and the chip inside at 6 years (Microsoft changed this in 2022), the depreciation of the shell and chip is mismatched, and there may be hidden impairment pressure when technology is scrapped in the future.
What this means for the crypto community
The market interpreted the "Microsoft Capex cut" as a mix of three narratives: "AI infrastructure turning point, computing power surplus, and macro dovish turn," with the market once using it as a catalyst for risk appetite. But the real computing power buildup hasn't stopped, so it's not a direct boon for BTC. For AI concept coins, DePIN, RNDR/TAO, and other "hash rate narratives," it's actually more neutral and positive: big companies haven't removed their ladders, just changed their accounting methods.
What we really need to look at is not the 175 billion figure, but three follow-ups:
1. Can Azure maintain 40%+ growth next quarter (this time: 41%–43%)
2. Will the $329.1 billion in operating lease commitments gradually erode free cash flow year by year?
3. Will Google, Meta, and Amazon follow suit with the "years + lease reclassification" strategy—if they follow, it will beautify the entire industry's profit statement; if not, Microsoft will benefit from individual EPS
Conclusion: Microsoft offers a cure for "AI money-burning anxiety," not a signal that the "AI cycle is over." Short-term sentiment can be traded, but in the medium term, don't treat accounting changes as fundamental turning points. For BTC, you should watch ETF flows and real interest rates, and not use Microsoft's earnings as a steering wheel.SPCX earnings countdown: Is unlocking hundreds of billions a trap a trap or an opportunity?
SpaceX has two major events coming next week, so anyone trading crypto or stock should keep a close watch.
The first quarterly financial report was released after the market closed on August 4, followed by the expiration of the massive volume of locked positions on August 6, with over 900 million shares surging out—a scale even more exaggerated than the existing circulating shares. The market has already backed down, with the stock price slipping from 225 all the way to 108, a drop of more than half, almost to the ankles.
But in the same play, the audience below each had their own thoughts.
Short-term traders only have eyes on the "bearish needle" from August 6th, fearing too many chips the market won't be able to swallow, so they simply exit first. Long-term funds are waiting for the earnings report, especially concerned about whether Starlink can present a reliable profit logic. If the fundamentals are strong enough, unlocking the stock actually leads to a shakeout—everything that should be sold is sold, the buyer gets the shares, switches hands, and the later stages are actually less heavy.
To put it bluntly, in the short term it's a thunderbolt; in the long run, it's a gift—this is where the disagreement lies.
This scenario is practically a daily replay in the crypto world: many projects crash before unlocking and panic, but when the day they actually unlock, they actually stop falling and rebound. Why? The sword hanging over the head has finally fallen; uncertainty turns into certainty, and only then does off-exchange capital dare to take the lead. SpaceX has already halved its price in advance. Once the selling pressure on August 6th is fully released, those who want to quit will withdraw, and those who want to stay will most likely be long-term investors willing to keep up.
Next week, let's look at two key points: the August 4th earnings report—will Starlink's profit model be truly clear? After the lock-up was lifted on August 6, whether the market took up any shares would mean there was sufficient turnover; if no one did, it would continue to fluctuate and bottom grinding.
$BTC $ETH $SNDK
#SPCX首份财报将公布, the $100 billion ban is about to be lifted #30年期美债收益率创19年新高
As a seasoned trader, facing the shocking macro upheaval of the current **30-year U.S. Treasury yield surging above 5.2% (the highest level in nearly 20 years since the 2007 global financial crisis)**, we must set aside local noise and face the intense fission of the global macro base.
The 30-year U.S. Treasury is known as the "crown of the global asset pricing anchor," representing long-term inflation expectations, fiscal credit, and the ultimate risk-free rate. The loss of this core indicator is far from a simple "interest rate fluctuation," but rather a wake-up call for a global macro liquidity restructuring.
Below, from the perspective of a seasoned trader, I will deeply analyze the essence of the "30-year US Treasury record" and comprehensively forecast the upcoming trends in **stocks and cryptocurrencies (Crypto)**.
1. In-depth Core Analysis: What is the essence behind the 30-year US Treasury yield hitting a 19-year high?
The "Term Premium" Spiral Out of Control and Fiscal Credit Crisis: Short-term rates are constrained by the Federal Reserve's policy rate (although the Fed has recently held rates steady, the short-term side has slightly declined), while the 30-year long-term yield riot shows the market is "voting with its feet" over the U.S. government's massive fiscal deficit and unrestrained bond issuance. Investors demand higher risk compensation to hold long-term bonds, known as "Fiscal Dominance."
Deeply rooted expectations of a "long-term inflation disease": Despite the earlier noise of a brief cooling in nonfarm payroll data, core inflation indicators like PCE remain sticky above 4% (well above the 2% target). The market no longer believes the Federal Reserve can easily suppress long-term inflation, and long-term bonds are being ruthlessly sold off.
The yield curve "non-linear steepening (Bear Steepen)": The long side surges, while the short side steeps sharply due to rate cut expectations. This is a major blow to traditional financial institutions, pension/insurance funds relying on long-term fixed income assets, and the real estate market (such as the simultaneous surge in the US 30-year mortgage rate).
2. Upcoming panoramic forecast of global major asset trends
Considering the current high long-term bond yields, the performance divergence of AI tech stocks, and the microstructure of digital currencies, the market outlook for the coming period is as follows:
1. Equity Market: The "Song of Ice and Fire" of Valuation Killing and Structural Differentiation
High-valuation tech and growth stocks (Nasdaq): Under increasing pressure. The denominator side of the DCF (discounted cash flow) model is being relentlessly eroded by 30-year risk-free yields above 5.2%. High-valuation AI concept stocks and nonprofit tech stocks, which have previously relied on "macro stories and long-term expectations," will face ongoing valuation compression.
Traditional Cyclical and Dividend Sectors: Defensive Aspects Emerge. Funds will further withdraw from high-risk, high-leverage sectors, flowing into traditional giants with strong cash flow and high dividends or energy sectors. However, due to the overall high risk-free interest rate, it is difficult for the stock market to experience a smooth, one-sided bull market; the market tends to swing widely or even temporarily decline to seek valuation safety cushions.
Specific targets (such as Micron MU, Microsoft MSFT, etc.): Microsoft is a cash flow giant with extremely strong resilience, so any pullback is a long-term opportunity; High-beta semiconductors like Micron, on the other hand, will experience intense high-level shakeouts along with the broader market as the market becomes more selective about capital expenditure (CapEx) returns.
2. Digital Currency (Crypto): The growing pain period of liquidity "blood drain" and extreme divergence
The natural opposite of macro liquidity: Cryptocurrencies are essentially amplifiers of global marginal liquidity (M2). When the 30-year Treasury yield soared to the extremely attractive risk-free yield of 5.2%, the OPPORTUNITY COST of global risk-on capital surged. A large amount of hot money will choose to buy US Treasuries and earn high-yield bonds rather than flood into the high-risk crypto market.
The divergence between Bitcoin (BTC) and altcoins:
Thanks to the institutional asset allocation attributes brought by spot ETFs, BTC can still demonstrate strong resilience and resistance at key support levels (such as the $62,000-$63,000 range), becoming a safe haven for capital clusters.
Altcoins will face catastrophic liquidity exhaustion. Against the backdrop of high risk-free rates, mid-to-tail tokens lacking real ecosystem vitality and relying solely on narrative support will continue to be ruthlessly sold off and deeply deleveraged.
3. Trading strategies of experienced traders
Faced with the macro shift of "30-year US Treasury Treasury hitting a 19-year high," traders absolutely cannot blindly bottom-fish high-risk assets on the left side:
Reducing overall leverage (De-leveraging): In a macro environment with high risk-free rates and tight liquidity, high leverage is a poison in the capital market. Whether it's stocks or cryptocurrency contracts, leverage must be minimized, or simply return to spot trading.
Embracing "cash flow" and "core assets":
Equities: Only buy leading companies with monopolistic business models and abundant free cash flow (such as Microsoft), and maintain respect for cyclical stocks and high-valuation concept stocks.
On the crypto side: strategically contracting the front line, heavily holding BTC or highly liquid blue chips, and abandoning futile fantasies about weak altcoins.
Alert to the "systemic risk" alert in the fixed income market: The uncontrollable surge in long-term U.S. Treasury yields often triggers secondary disasters (such as liquidity runs by certain highly leveraged hedge funds, overseas central banks, or regional banks). If a similar event occurs, the market will experience a momentary panic spike (Selling Climax), which will be the real moment for long-term funds to pick up blood-stained chips in batches. Altcoin Season Isn't Coming. A Different Version Already Arrived.
The classic altcoin season playbook: BTC pumps, dominance drops, and everything rallies together. That's not 2026. BTC dominance is holding near 60.3%. The broad "everything rally" hasn't materialized. And most alts are quietly bleeding.
But capital is moving. Just differently.
AI-integrated protocols, Real World Assets, and DePIN projects are pulling real flows while most of the market sits flat. The XRP and Litecoin ETF approvals earlier this year changed the dynamic too: institutional players now selectively add altcoin exposure, which means specific narratives get institutional-grade bids while everything else gets left behind.
This cycle may be more sophisticated and more brutal than previous ones. If you're waiting for a rising tide to lift your bags, that tide may not be coming. The action is in specific sectors, and they move fast within a narrow window.
The hard question isn't which alts will run. It's whether you have the conviction and the information edge to be in the right ones before the narrative goes mainstream.
What sectors are you actually paying attention to right now?
Share your thoughts in the comments 👇#折旧年限延至25年, Microsoft's capital expenditure guidance was lowered
Many people interpret it at first glance: Microsoft is cutting AI investment, cooling demand for computing power. But this is a typical surface misinterpretation.
Microsoft extended the depreciation period for data centers and office buildings from 15 years to 25 years, and many lease projects shifted from financial leasing to operating leases, no longer included in capital expenditures on financial statements.
Management made it clear: excluding accounting adjustments, the actual scale of infrastructure investment has not contracted, and investment in fiscal year 2027 will continue to grow.
Market pricing logic shift: bidding farewell to the mindless computing power era
Funds no longer simply chase "spending to expand capacity," but have begun to focus on cash flow and investment returns. Microsoft's stock price surge is essentially due to AI targets with commercial payout capabilities and controllable capital consumption.
The industrial chain is experiencing structural differentiation
Short-term sentiment suppresses upstream hardware such as storage and servers; Cloud services and AI application tracks have relatively strong advantages. If Google and Amazon follow suit and adjust accounting policies, the computing power sector is expected to remain under pressure.
My interpretation: It cannot be simply taken as a negative hashrate signal, but two risks must be watched for:
First, if multiple giants follow suit and adjust depreciation rules, the market will continue to question the authenticity of financial profit reports;
Second, capital preference shifted from upstream hardware sectors, intensifying internal rotation in the Nasdaq's growth sector, indirectly affecting risk asset sentiment.
Mapping the crypto market:
AI narrative differentiation continues to deepen. Volatility in related stocks relying on hash power speculation has increased; Main theme opportunities rely more on macro liquidity rather than purely driven by AI sentiment.
Continued market tracking: Capital expenditure statements from other tech giants and sustained demand from Azure customers.Bitcoin ETFs Have $51.3B in Net Inflows. Why Does the Market Feel So Weak?
Cumulative net flows into US spot Bitcoin ETFs are still positive at $51.3 billion as of late July. Two consecutive weeks of inflows after an eight-week outflow streak. On paper, a recovery signal.
But here's the thing: those eight weeks shed over $8 billion. The $273M that came back in the last two weeks is recovering roughly 3% of what left. Institutional money came in hard early and proved willing to leave quickly when conditions shifted.
BlackRock's IBIT alone holds $67B in AUM, a staggering concentration in one product. That's a lot of Bitcoin moving to the rhythm of institutional sentiment, rate expectations, and macro risk appetite. More ETF adoption isn't a guaranteed bid, it's a sentiment-sensitive one.
The dynamic worth watching: ETF inflows are trickling back while stablecoin reserves at major exchanges have drained $2.3B. That's a gap. Institutional buying returning while retail dry powder thins out. One of those trends has to give.
Is the ETF recovery enough to move the needle, or does the market need retail back in the picture first?
Share your thoughts in the comments 👇DeFi Has Lost 39% of Its TVL in 2026. What's Actually Going On?
DeFi TVL is sitting at $70 billion, down from $115 billion in January. A 39% drawdown in under eight months, and it's not a clean story with one villain.
A few things are happening at once. The broad market correction pulled capital out. But the real gut punch has been security. There have been 121 hacks across 2026 totaling $942 million in losses. Q2 alone saw 85 incidents.
At some point that kind of track record starts to affect how people think about protocol risk at a structural level.
The chain-level data is telling too. Ethereum's DeFi base is down 43% YTD, Arbitrum down 55%. But TRON is up roughly 5% on stablecoin settlement activity, and Hyperliquid is up 7% on perpetuals. Chains with real utility in specific use cases are holding ground while general-purpose ecosystems bleed.
The uncomfortable question: is this a cycle problem or a structural one? Because if DeFi's security track record doesn't improve, the next bull run might route capital around it rather than through it.
What's your read on DeFi's recovery timeline?
Share your thoughts in the comments 👇#财报观察员: Next Thursday's lottery results will be held, with Circle as the final. Circle's final earnings report is not what the market is truly waiting for, but whether the stablecoin narrative can continue to upgrade.
Next week, the market will see a round of intensive earnings reports, and what I'm most focused on is not traditional tech giants, but Circle.
The reason is simple.
Its financial report no longer belongs to just one company, but answers a key question for the entire stablecoin sector:
Is stablecoin a cyclical business, or the next-generation financial infrastructure?
Over the past year, market perceptions of stablecoins have changed significantly.
More and more cross-border payments, on-chain settlements, and institutional fund management are adopting stablecoins; The regulatory framework is also gradually improving, and stablecoins are moving from internal tools in the crypto industry to the traditional financial system.
Therefore, what truly deserves attention in Circle's financial report is not whether revenue slightly beats expectations, but three core indicators:
* Whether USDC's circulating scale continues to grow;
* Whether real-world usage scenarios such as payment and settlement continue to expand;
* Is management more optimistic about future regulation and commercialization?
Many people like to think of financial reports as an exam.
But the capital market is more like judging whether a company's growth curve has changed.
If Circle proves that stablecoin demand no longer relies solely on a crypto bull market but is experiencing sustained growth in payments and financial infrastructure, then the market's valuation logic may change further.
This is also one point I have always emphasized:
What truly drives long-term market momentum is not a flashy financial report, but the market beginning to believe that a new business model has entered the realization stage.
As trading progressed, I became less and less predictable whether the financial report would be positive or negative.
What I focus on more is:
After the earnings report is released, will funds continue to raise expectations for the industry over the next three to five years?
Because quarterly profits determine today's profits, while long-term expectations determine valuations.
For Circle, next week's release is not just about earnings, but also an important test of whether stablecoin narratives can continue to gain global capital recognition.
If funds continue to vote with real money, then what is traded in the market will no longer be a single company, but the future value of the entire stablecoin ecosystem.Morgan Stanley's latest 120-page research report has refueled the faith of fellow villagers
First, Morgan Stanley characterized this pullback as a technical element: crowded positions were digested, leveraged trading was trampling, momentum factors reversed, and non-fundamental logic was disrupted
Morgan Stanley transforms into a great scholar debating the classics, with a 120-page research report shattering market panic over AI
Market Concern 1: Companies start restricting token usage—has AI demand peaked?
Morgan Stanley did the math. A single AI call saves an average of $55 in labor costs, while using an agent to complete an enterprise-level task costs only $2 to $5, with ROI exceeding tenfold. For a tool with a tenfold ROI, whether a company has a budget limit or not doesn't really matter; the one that doesn't use it will be eliminated by the one that does
The net profit margin for Blackwell architecture data centers is about 58%, rising to 80% by Rubin and 90% by Feynman. Cloud providers can completely push token prices down by 75% while keeping profit margins unchanged, leaving significant room for price reductions and not losing a single cent in profit
Market Concern 2: With the low cost of large models in China, is the trillion yuan capital expenditure by American giants going down the drain?
Morgan Stanley responded with an old 19th-century story. After Watt improved the steam engine, coal efficiency increased significantly, but coal consumption in Britain soared. The reason was simple: steam engines became cheaper, so they were deployed to more factories and mines. Economics calls this the Jevons paradox. The same logic applies to AI computing power: the higher the efficiency, the lower the cost, the more feasible scenarios there are, and the greater the total consumption
There is a set of data that gives a sense of the scale of the supply-demand gap. Google executives recently stated that the company may need to double its computing power every six months, which is a 1,000-fold growth over five years. Meanwhile, NVIDIA is shipping at its highest growth rate in history, with cumulative deliveries over five years less than 10% of Google's individual demand
Market Concern 3: Will the constraints of the physical world choke the neck?
Morgan Stanley summarizes the supply-side bottlenecks in three Ps: People, shortage of skilled workers; Power, grid queue for 5 to 7 years; Politics, states have begun to revoke tax incentives for data centers, and federal legislation is also being pushed to introduce electricity surcharges for data centers
These constraints do exist, but Morgan Stanley sees them as speed bumps, not roadblocks. The truly underestimated opportunity lies in power time arbitrage: grid connection takes 7 years, but natural gas turbines can supply power in 1 to 3 years, and Bitcoin mining farms already have 10 to 19 GW of grid-connected capacity that can be directly converted for data center use. Whoever can provide "physical space with electricity available" in the short term is the target with the largest market pricing deviation
Even after accounting for all alternatives, there is still a net gap of about 1GW in a neutral scenario, expanding to 11GW in a pessimistic scenario. What cloud providers lack most now is not money, but places with electricity$BTC The Federal Reserve has paused rate hikes continuously, yet three officials insisted on raising rates, making policy differences obvious. CITIC Securities judges that high-valuation sectors in the US stock market are under short-term pressure, while the long-term logic for the AI sector remains unchanged. However, liquidity tightening pressure is now real.
Mapping to the crypto market, BTC is still stuck in the $62,000-66,000 range, with trading volume shrinking by more than half from its peak. Combining on-chain data, many whales have been continuously increasing their holdings in ETH and BTC recently, concentrating their holdings, yet the market remains slow to move, and funds are watching the Federal Reserve's stance.
Additionally, with the temporary easing of Middle East conflicts and the fading of geopolitical safe-haven dividends, the previously supporting factors for coin prices are gradually weakening. At this stage, the market's only hope is the Jackson Hole meeting in August and the September policy meeting.
As long as the Fed signals a somewhat accommodative approach, major holders are very likely to start a rally; If the hawkish stance continues, the bottoming consolidation will persist for a long time.
Do you think the Fed will loosen liquidity in September? Will Bitcoin break out of its consolidation range ahead of schedule?MRVL and INTC Trade Analysis
$MRVL is essentially a high Beta AI Infrastructure trade, with core exposure coming from customized XPU, optical communications, switches, and data center interconnects. Its bullish thesis is not only based on Hyperscaler CapEx expansion but more critically on whether Design Wins can successfully convert into mass production revenue, and the operating leverage brought by an increased AI Revenue Mix. As long as the market remains in a profit expectation upgrade cycle, MRVL has the potential to maintain Forward Multiple Expansion; however, this type of long-duration growth stock is extremely sensitive to discount rates and expectation gaps. If order slope, customer pull-in pace, or management guidance fall below market Whisper Numbers, rapid Multiple Compression can easily occur.
In trading, $XMRVL is more suitable to be observed within the AI CapEx and earnings momentum framework. It is not a low-valuation reversal but a typical Momentum/Revision Trade, where the stock price needs continuous upward revisions of EPS, order visibility, and new project ramp-ups to sustain the trend. The main risks are customer concentration, delays in mass production schedules, and overly crowded market positioning. When sell-side consensus expectations are too high and Call Skew is too steep, even earnings beats can lead to a spike and subsequent pullback due to positive news being priced in and Implied Volatility Crush.
$INTC, on the other hand, is a completely different trade. It is not purely an AI growth stock but a Turnaround/Special Situation with high Optionality. The market’s core trade is not just the recovery in PC and server demand but also whether 18A yield, external foundry customers, advanced packaging, Foundry loss convergence, and CapEx normalization can drive a valuation re-rating. If process nodes and customer validations start to materialize, INTC has the chance to be revalued from an inefficient IDM to a strategically valuable advanced manufacturing asset in the U.S., resulting in a Sum-of-the-Parts Rerating.
However, the issue with $INTC is that this trade requires a longer Catalyst Path. Continued Foundry cash burn, pressured free cash flow, and limited external customer revenue all indicate the market will not immediately assign a high valuation based on single-quarter data improvements. Its upside comes from very low expectations and positive convexity, while downside risks stem from execution delays and CapEx drag. In other words, INTC’s Reward-to-Risk may be high, but realizing returns takes time and is susceptible to policy subsidies, process progress, and asset restructuring news.
Overall, MRVL is a Momentum Trade buying into earnings upgrades, AI orders, and trend continuation, characterized by high growth, high valuation, and high sensitivity; INTC is a Mean Reversion Trade buying into process turnaround, foundry optionality, and asset revaluation, characterized by low expectations, high volatility, and a long verification period. MRVL is suitable for trend-following longs but requires caution against crowded trade breakdowns; INTC is suitable for phased accumulation during extreme market pessimism but should not mistake short-term rebounds for a completed fundamental turnaround. Michael Saylor may dump $5B of Bitcoin
headline is doing exactly what it was designed to do: scare you before you read the the full info about it..
basically,
Strategy has permission to sell up to $5B worth of BTC under its current plans.
that number comes from adding together:
- up to $1.25B to refill its cash reserve
- around $1.76B a year in dividends and interest
- up to $2B for stock buybacks
the key word is up to.
the buybacks are optional.
they can be paused, reduced or never used. the yearly payments also don’t need to be funded all at once.
Strategy already keeps a separate cash reserve for these payments and has been topping it up by selling MSTR and preferred shares.
As of today, that reserve held $2.55B.
they have sold BTC before. around the end of June and start of July, Strategy sold 3,588 BTC to pay preferred-stock distributions and refill the reserve
so yes, BTC sales are now one of the tools available to them.
but there isn’t a fixed $5B sell order waiting to hit the market.
the real danger begins if this becomes a repeated pattern:
- MSTR fundraising slows down
- the cash reserve keeps falling
- Strategy starts selling BTC every week
that would create steady supply hitting the market.
until then, the $5B figure is simply the maximum ceiling being passed around like a confirmed dump.
the two things I’d watch from here:
- how much cash Strategy still has
- and whether BTC sales start happening repeatedly
sure the headline gives you the biggest possible number but the weekly filings will tell us whether any real pressure is actually building
lets see..
#EarningsWeekAhead
#USStepsInForYen Earnings Analyst: Amazon's guidance falls short of expectations, yet stock price rises 9%
Amazon's earnings report this time did not meet market expectations, but its stock price actually rose by nearly 9%.
On the surface, it seems a bit contradictory, but on closer inspection, it makes sense. People already felt the overall environment was bad, and the company's conservative expectations were actually reassuring, showing that management wasn't making empty promises.
Growth in cloud and advertising has stabilized, and retail efficiency is gradually improving.
The market now cares more about whether the market can withstand the pressure, rather than how impressive the short-term numbers are.
So when the guidance is a bit weaker, the stock price actually rises.$LAB Brothers, stop bragging, let's analyze carefully. If you want to play the lab, read carefully.
Scenario 1: Unlock 12% of the total supply of 120 million coins in mid-August. At current prices, it's about $20 million, which isn't much. If the project team wants to keep harvesting, they'll wait for the day it unlocks and buy it back. It all depends on whether they buy this batch of goods. Compared to current prices, this batch isn't very large. Recently, both the project team has released and the high-priced goods, and unlocking buybacks are minor issues for them. To rally the market, you have to consume this batch of chips; daily unlocking is negligible for the current price.
Second scenario: Unlock in mid-August. The group still thinking about a surge in prices has faith but doesn't want to sell, then it's troublesome. In that case, a rebound won't be expected in the near term. For those who queue trades, the only option is to hold a prolonged battle. Only when the market fluctuates and all shares are sold out can we see a rebound.
Scenario Three: The project team keeps issuing positive news, but prices are falling. The worst outcome is a frenzied harvesting rush, and the price is cut all at once. So in the next few days, we can see another halving, or rather, it will keep falling and halving to scare off the unlocking of chips first.
Summary: If you don't have any orders, you can be patient; there's no need to go long now. Results will be unlocked in mid-August. Personally, I think it should be halved. Wait until these unlocked chips are used up before rebounding. Because I'm not very familiar with what happens if this coin is unlocked in large amounts every month. Get ready to reset to zero.
Brothers, if you have good ideas, feel free to discuss them together. Let's share our own analysis and insights.