Orbit Post Sitemap
#AMD财报超预期 has growth been overdrawn?
AMD's financial report is a microcosm of the AI narrative.
Revenue hit record highs, data centers doubled, profits exceeded expectations—every line of numbers tells the same story—AI demand is real, and AMD is making money from this wave.
However, the stock price fell 7% after hours.
The good news was interpreted as "not good enough." The AI narrative market has shifted from "whether there is demand" to "whether demand is growing fast enough."
AMD's Q3 guidance is $13 billion, which is $500 million above analysts' average expectations.
But aggressive investors want more, as AMD's stock price has already doubled.
The higher the valuation, the stricter the market's requirements for growth rate. This isn't just AMD's problem—it's a test faced by the entire AI hardware sector—when your valuation has already priced in the most optimistic scenario, any earnings report that is "just a little better than expected" will be penalized.
The competitive landscape of AI chips is also changing. Nvidia still holds the dominant position, while AMD is catching up. Major clients favor Nvidia. AMD's data center business is moving from zero to one, from one to ten, but the market is already at the stage of pricing from ten to a hundred. AMD's earnings report is fine, but market expectations are questionable. The biggest lesson from this financial report is: in AI narratives, doing it right isn't enough; you have to do it right beyond everyone's expectations.Morgan Stanley had analyzed SpaceX's valuation before its earnings report:
Rocket business valued at 8 yuan, Starlink at 128 yuan, and AI business at 152 yuan. Target price of $300. At the current price of 117, the market's pricing for SpaceX doesn't even cover Starlink alone. The market's underlying message is: Starlink is worth over a hundred, rockets are free, AI business is either worthless or negative equity.
The space business generated Q2 revenue of 962 million yuan but an operating loss of 542 million yuan. AI business revenue was 2.561 billion yuan, with an operating loss of 1.257 billion yuan. Of the three major businesses, only the connectivity business (Starlink) is making money. AI is burning money, rockets are burning money. Most of Starlink's earnings go to supporting AI and Starship.
Musk's net worth evaporated by $646.4 billion in just one and a half months. SpaceX's market value fell from $2.6 trillion to $1.5 trillion. But what truly determines the stock price is not Musk's wealth, but whether early employees and investors will sell after the lock-up is lifted on August 6. Their costs might be a few yuan, even a few dozen yuan, not just a few hundred. What they are waiting for is not financial reports, but liquidity windows. The window will open tomorrow. #SpaceX首份财报超预期, unlocking remains a key variable There are reports that the Trump administration plans to ban China's new optical modules, which is explosive positive news for the optical communications sector! 🧐
Simply put, optical modules are the data center's "super transmission belt" that converts electrical signals into light signals and transmits data at high speed between AI chips. Nowadays, every rack in AI data centers is packed with this stuff. Currently, the industry is rapidly upgrading from 800G to 1.6T, and this is simply not enough. Currently, in terms of market share, Chinese companies occupy half of the optical communication sector.
Think about it: if the US really wiped out China's supply chain to zero, would the demand from AI giants disappear? Not at all. Data centers still need to be built, AI training continues, and this massive order can only be poured into the hands of the remaining American companies. And the "American players" who can handle this huge fortune are very concentrated, with only these three core players:
$COHR (Coherent) & $LITE (Lumentum): These two are absolute "duopolies" in the optical and laser fields. How strong is it? Even Nvidia, in order to secure its supply chain, has invested billions of dollars in advance to sign long-term underwriting contracts with them. They control the upstream core laser chips and components; as long as the data center is still under construction, almost every optical path involves them.
$AAOI (Applied Optoelectronics): This is the purest American-made target. Not long ago, people were still questioning whether expanding by 400,000 square feet of capacity in the U.S. was too much of a gamble. At first, it seemed a bit aggressive, but now, combined with this ban, it's practically a divine prediction! With capacity released, this is definitely the biggest beneficiary of U.S. domestic manufacturing returning to shore in the future.
What's even more striking is that both policy and supply sides are fueling the fire.
Just last week, the US FCC banned China's humanoid robots and power inverters under the same national security ban, with the same logic: protecting domestic AI development and forcing manufacturing to return. Optical communications, as the throat of AI, are clearly the next domino to be toppled.
Combined with material shortages and physical production restrictions: Lumentum's CEO previously stated that the shortage of indium phosphide materials used to make laser cores could be even more absurd than memory chips. In other words, on one hand, there is a physical shortage of core raw materials; on the other, policies require the destruction of a large portion of existing global production capacity, with obvious consequences.
The current logic is extremely simple and blunt: surge in demand + policy blockade of Chinese production capacity + severe raw material shortages. These three forces have united and directly pushed all their shares into these three companies. Currently, all three have gained over 10% in pre-market trading, which is worth watching! 🧐 Fundamental Research Report $SUI / Sui (Public Chain/L1) $0.69 (24h -0.73%)
To summarize: Sui ($SUI) has an overall score of 67/100, with fundamentals meeting but some flaws. Looking at the three layers, the company team has cash reserves, the protocol network already shows signs of paid usage, and token capture has been implemented.
Project Overview: Sui (token $SUI), public chain/L1 track. Focusing on the Move-based parallel settlement public chain. Benchmarked against APT and SEI. Traditional collaboration between enterprises relies on cloud servers and contract reconciliation, which causes gas surges, TPS constraints, and frequent cross-chain bridge security incidents during high concurrency. Public chains use a unified state machine for trustless settlement, reducing reconciliation costs. Average order value is $50-500/month, with settlement required in USDC or fiat currency. Narrative-driven tracks, bear market usage cut by 60-80%. Positioning the end-to-end vertical platform. Product implementation: The protocol layer is officially operational, and the on-chain dashboard shows protocol fees accumulating, showing signs of paid usage. The latest version, testnet-v1.77.1, has been valid for 9,999 submissions in the past 90 days.
At the user level, address MAU not disclosed, DAU not disclosed, 24-hour transaction volume $110.59M, TVL $413.24M. Wallet addresses do not equal monthly active users of natural persons; large large addresses holding concentrated positions tend to overestimate the actual number of users. On the revenue side, user fees are not disclosed. Supply-side revenue is about 80-90% of user fees (distributed to LPs and nodes), protocol treasury revenue is $779.6K, and token holders buy back and burn at an annualized rate without a burn mechanism. 24-hour transaction volume is business turnover, not revenue. A company making money does not mean the protocol makes money, and protocol profits do not equal token holders making money. On the code side, 9,999 valid submissions in 90 days, 100 active contributors, latest version testnet-v1.77.1. GitHub is a Class A evidence that can be directly verified. Investment background: For company equity financing, look to PitchBook/Crunchbase (A-level); for token private and public funding, use whitepapers, release curves, and on-chain unlocked contracts (A-level); market makers and ecosystem funding are B-level and do not represent long-term holdings of tech VCs; for technical integration, look to API/SDK access evidence (B-level); strategic partnerships and logo walls are D-level. The use of NVIDIA GPUs does not equate to NVIDIA investment, and going public on exchanges does not equal strategic investment.
On the token side, total supply is 10,000,000,000.0, circulating 4,074,529,886.441529 (40.7%), FDV $6.90B, next unlock undisclosed (share of circulating undisclosed), annualized share of burn and buyback not explicitly repurchased or burned. Do you have to buy coins to use the product? Yes, strong value capture (Gas/Collateral/Service Access). Let's look together with peers (unified criteria, no cross-sector comparisons): In terms of circulating market capitalization, Sui $2.81B, APT undisclosed, SEI undisclosed. For FDV, Sui $6.90B, APT undisclosed, SEI undisclosed. In terms of annualized revenue, Sui $779.6K, APT not disclosed, SEI undisclosed. Regarding monthly active addresses or users, Sui has not disclosed, APT has not disclosed, SEI has not been disclosed. Figures are based on public data snapshots; any omissions are supplemented by official self-reports or industry standards. Valuation, market capitalization $2.81B, FDV $6.90B, P/S 3605.5x, FDV divided by revenue 8848.9x. Pessimistic outlook: $2.81B at 50-70% off, oscillating within a neutral range; optimistic: revenue doubles, burns are implemented, enterprise clients are coming in, FDV corresponds to P/S, aligns with the top companies. In the end: solid fundamentals (score 67/100). Token value capture has been implemented (buyback/burn/gas). Circulating market cap is relatively high relative to fundamentals, overdrawing expectations, and FDV is moderate. Risks to watch for: short-term large-scale unlocking and sell-off, long-term protocol revenue wiping out, token demand relying solely on incentives (once incentives are cut off, usage collapses). Next, focus on these numbers: protocol fee weekly, burn amount, active address retention, TVL/loan balance, and GitHub version release. The above is the logic and judgment of the publicly available information and does not constitute buying or selling advice. Core financial indicators deviate by more than 30%, and the conclusion needs to be reassessed.
That's all for the fundamentals; leave the rest to the market.
#基本面研报 #加密 #研究 #OKXOrbitWhile US-Iran negotiations are progressing, OPEC+ is doing something else—agreeing to increase production for the sixth consecutive month in September.
The oil price plunge is the result of two forces combined: expectations of supply disruptions are weakening, actual supply is increasing, and both sides are loosening simultaneously.
But supply is returning soon.
Iran insists on dominance over the strait passage rights, and shipping analysts believe the recovery of oil flow will be slower than official statements, and inbound vessels will need to replenish the actual volume to confirm the actual passage. The demand side is also pessimistic; China's ample inventories continue to suppress import demand. The increase from the Middle East's summer peak in power generation oil consumption is estimated at only 500,000 barrels per day.
If the Hormuz agreement is truly reached, oil prices may still have room to fall. Brent's deep back structure between near-month and far-month is clearly narrowing as premiums are clearing. After the agreement is implemented, it may shift to a Contango structure. But the agreement hasn't been signed yet, and Iran is still insisting on its dominance. If the agreement falls through in the next two days, there will be room for a rebound at the lows after the sharp drop.
The drop in oil prices below 80 is not necessarily a trend; it could just be the market pricing in something that hasn't happened yet.
#美伊谈判推进, oil prices fell below $80 📊 Daily Tokenized Equity Overview
Top Performers:
• $XONDS : 8.88 (+1.49%) 🔥
• $GRVT 0.26339 (+0.66%)
• $XTER : 408.16 (+0.45%)
• $xGEV: 1,025.24 (+0.32%)
• $xXLE: 58.51 (+0.10%)
Slight Pullback:
• $xADBE: 257.3 (-0.10%)
• $xVRT: 273.29 (-0.30%)
$xONDS and $GRVT lead today's board with solid green pushes while tokenized tech assets hold key range support levels!
NFA – Educational purposes only.
#EarningsRealityCheck #SpaceXBeatEstimates #AMDBeatsButDrops 油价跌破80，对加密市场来说不是直接利好，是间接利好。
逻辑链条是这样的：油价跌→通胀预期降温→美债收益率下行→加息预期回落→风险资产估值压力减轻。
8月4号这波美股大涨，核心驱动就是这条链。9月加息概率从67%回落到了56%。多空双方在加息的概率上松了半口气。
但这条链有两个脆弱环节。第一，伊朗否认直接谈判，协议没落地，油价随时可能反弹。第二，即便协议签了，油价从80回到70还是维持在80-90区间震荡，决定了加息预期能松多少。
高盛的判断更倾向于后者——协议签了也只是回到区间震荡，不是趋势反转。油价跌了，加息预期松了一点，但没松到让大饼直接起飞的程度。大饼从63,000弹到64,000，弹了1%左右，然后就没然后了。真正的反转需要油价持续低于80，或者CPI连续两份低于预期。一两次油价波动，改变不了趋势。#美伊谈判推进，油价跌破80美元 #AMD财报超预期 has growth been overdrawn?
AMD's earnings report beats expectations ≠ rise | Growth has been overdrawn, post-hours -9% is the market repaying "expectation debt"
8/4 AMD handover: Revenue $11.54 billion (YoY +50%), Non-GAAP net profit $2.76 billion (+253%), EPS 1.66 (expected 1.62%), data center $6.7 billion (+107%)—all four exceeded expectations, and Su Ma brought out Helios, EPYC, and Instinct all on the board.
And what happened? In regular trading, it closed up 7% at 518.58, but after hours, it plunged back to around 472, dropping -8.7%~-9%.
My view is straightforward: this is not a performance bombshell, but a classic pricing of "expectation overdraft."
Three points that most people overlook:
1. The stock price has been ahead of the earnings for a year: AMD has risen over 110% this year, with a PE (TTM) of about 170 times as of the close on August 4, and a 52-week high of 584.73. The market has long priced it as the "AI No. 2 player," so the earnings report only needs to be "impressive." This year's revenue only exceeded expectations by about 260 million yuan, with EPS above expectations of $0.04—a "good but not explosive" approach.
2. Q3 guidance of 13 billion vs. buyer's whisper of 14 billion: The median official guidance of +41% year-on-year is actually not low, but some aggressive funds are betting on 14 billion, not breaking below the most optimistic expectations = a signal of momentum peaking.
3. Capex 808 million (nearly 3 times year-on-year) + AI burning money accelerates: Helios rack-level mass production requires higher upfront investment, while profit margin ceilings are being suppressed by itself. Wall Street's tolerance for "AI burning money for market share" is declining.
To put it plainly: the company hasn't weakened; it's the stock price that has discounted its 2027 AI shares ahead of schedule to August 2026. Positive news realization = profit-taking withdrawal, not fundamental reversal.
• BTC/ETH and Nasdaq tech lines are strengthening in synergy. AMD's "AI sentiment anchor" has dropped 9% after hours, which will spill over to AI× DePIN (RNDR/akash/TAO classes) and XAMD tokenized markets. Don't just watch BTC's mood
• But note: AMD's data center doubling ≠ AI demand is disproven; valuations need to be repriced. If NVDA's month-end earnings report is also "good but not impressive," the entire AI hardware chain will all enter the consolidation box, with risk appetite pulling back to the top beta crypto AI coins first
• Trading: XAMD/stock derivatives will not chase earnings reports with gaps; wait for support in the 472-485 range (after-hours low + 7/28 previous lows); BTC itself relies more on macro liquidity than on AMD per note; don't equate AMD's pullback with BTC turning bearish
Conclusion: AMD is a good company, but the current price is for the "Nvidia Challenger" dream discount, which has been reset to zero. Next, let's see if Helios's Q4 volume ramp can push the PE ratio from 170 back to just over 100. Short-term unlocking + valuation double kill not yet complete, but the long-term AI computing power curve remains unbroken.8月5号盘后，闪迪要交Q4财报。
#闪迪财报前夕，HBF与存储紧缺引发热议
华尔街预期营收83.9亿美金，EPS 33刀。
公司自己的指引是77.5到82.5亿。
关键不是能不能超公司指引，是能不能超已经被市场抬高的一致预期。
真正有分量的数字是85亿以上营收、35刀以上EPS、毛利率稳住80%左右，同时上调2027财年指引。如果只超公司指引没达市场预期，照样会被资金按“不及预期”处理。营收低于81亿，或者毛利率、远期指引有任何松动，股价可能再跌15%到25%。
这轮财报本质上是AI存储叙事的一次压力测试。市场最关心的不是已经兑现的利润，是AI存储的高景气还能持续多久。上季度闪迪数据中心收入环比增长了233%，这季度得证明企业级SSD需求还在加速。期权市场定价的财报波动预期是±21%，多空两边都在押注。Nvidia, Microsoft, Amazon, and Meta continue to expand their AI data centers, causing memory shortages and prices to keep rising.
Industry insiders expect this supply-demand imbalance to continue into 2027.
Goldman Sachs clearly predicted before its earnings report that SanDisk's Q4 would be a "very strong quarter." NAND flash supply and demand remain tight, with demand on mobile, PC, and data center sides recovering simultaneously. But the stock price has fallen from a high of 2354 at the end of June, down about 25% over the past month. When prices rise, AI storage is a supercycle; when prices fall, the market starts to worry about tech giants overheating their AI spending.
Same fundamentals, two different pricing. The purpose of financial reports is not to prove that "AI storage demand exists"—this is already a consensus; what the report should prove is "how much longer AI storage demand can last." The market needs to see that the gross margin of around 80% and the AI storage shortage can still be sustained. Guidance is much more important than the current figures.
#闪迪财报前夕, HBF and storage shortages have sparked heated discussion After SanDisk's financial report, how will the storage sector move?
#闪迪财报前夕, HBF and storage shortages have sparked heated discussion
It depends on three numbers.
The first is gross margin. Last quarter it was 78.4%, about 80% of market expectations. If it holds 80%, it means pricing power still lies with the supplier. If it falls below 78%, the market will start asking, "Is oversupply coming?" The second is the guidance for fiscal year 2027. Wall Street no longer cares much about the current figure; what matters is whether it can continue to rise over the next 12 months. The third is the month-on-month growth rate of data center revenue. Last quarter, it rose 233%. Given this base, doubling again is unrealistic, but if growth drops below 50%, the market will reprice the valuation of the entire storage sector.
The current correction in storage stocks has been significant—SanDisk has fallen 25% from its peak, Micron 21%, and SK Hynix over 30%. If the earnings report exceeds expectations and guidance is raised, the storage sector may see a retaliatory rebound. If the financial report only "meets expectations" without surprises, it may "crash first, then crash." If revenue falls below 8.1 billion yuan or gross margin loosens, it could fall another 15% to 25%.On August 4th, the US stock market exploded.
The Dow rose 907 points, surpassing 54,000 points—a first in history.
The S&P 500 rose 1.79%, closing at 7,736 points, marking the first time it broke above 7,700. The Nasdaq rose 2.59%, and the Philadelphia Semiconductor Index gained 6.55%.
All three major indices hit record highs simultaneously, a trend that hasn't been seen much this year.
Why the price has risen? Five major catalysts have overlapped. Becent said that as early as Tuesday or Wednesday, an agreement with Iran could be reached to reopen the Strait of Hormuz, and oil prices would immediately fall below $80.
US Treasury yields have also fallen, with the probability of a rate hike in September dropping from 67% to 56%. Palantir's earnings report exploded, with a 13% rise in after-hours trading. Caterpillar's quarterly revenue surpassed 20 billion for the first time. Coupled with falling oil prices suppressing inflation expectations, five factors resonated on the same day.
Paul Hickey of Bespoke is right—this rally isn't driven by a single event, but by multiple positive catalysts taking over. AI hardware, industrial, financial, and healthcare are all seeing double-digit growth. US stocks are no longer just driven by tech stocks—the whole market is resonating. But Piper Sandler reminded us—risk appetite is indeed rising, but market breadth is still weak. This recovery is "good, but not quite perfect."
#标普500首次站上7700点, setting a new all-time high SpaceX这份上市后的首份季报，刚出来就直接引爆了市场。
核心数据确实能打：第二季度营收直接干到了78.14亿美元，同比增长92%。运营亏损也大幅收窄，从之前的9.7亿美元一下子缩到了1.43亿。营收和亏损改善两方面，双双超过市场预期，这个成绩单绝对算硬核了。
更关键的是财报里公布的那条消息——宣布跟英伟达合作，开发Starmind AI1卫星计算载荷。这意味着SpaceX不只是造火箭和卫星了，而是要把太空基础设施往AI方向去推进。这就把它的故事从一家航天公司，直接拉到了AI基础设施的赛道里面。
但紧随而来的，就是悬在头顶的那把剑。
8月6号，也就是明天，SpaceX首批限售股将正式进入解禁窗口。符合条件的老股东，最多可以出售20%的限售股份。
注意这个描述——潜在供应规模超过当前公众流通盘。意思就是，解禁出来的筹码可能比现在市面上能买到的还多。这才是这波行情接下来真正的压力测试点。
说到底，财报数字给了长线逻辑足够的支撑，92%的营收增速、亏损大幅收窄，加上跟英伟达在AI卫星上的深度绑定，这些决定了它的天花板在哪里。但明天解禁开闸后的抛压能不能接住，才是短期定价的关键变量。
市场现在看的是两条线：一条是SpaceX到底能不能持续赚钱，另一条是这么大一波解禁砸下来，接盘的力量够不够强。
这两件事谁能占上风，接下来一周就会见分晓。
#SpaceX首份财报超预期，解禁仍是关键变量 $SPCX $SNDK $SKHYNIX The S&P breaking above 7700 is most notable for its driving structure.
From April to June, the S&P kept struggling around 7620, unable to go up or down.
Breakthroughs require new catalysts.
Becent's remarks were the trigger—expectations of US-Iran negotiations drove oil prices down, inflation expectations cooled, and long-term US Treasury yields fell accordingly.
With the simultaneous easing of three major constraints, the index naturally surged upward.
But Bespoke's Hickey added another point—the Nasdaq has risen more than 1% for four consecutive days. Historical experience shows that this kind of continuous buying usually represents genuine capital flowing back, not short-term rebounds.
This is not retail investors chasing, but institutions re-allocating.
Jay Woods of Freedom Capital put it more bluntly: the negative side of the war is being digested by the market, and what truly dominates the market now are corporate earnings.
Palantir's revenue grew by 93%, Caterpillar surpassed $20 billion, and AMD doubled its data center business—these are real numbers.
Oil prices fall, interest rates fall, and earnings are strong—three things happening simultaneously. The S&P reaching 7700 is the result, not the end.
#标普500首次站上7700点, setting a new all-time high US stocks hit a new high, and Bitcoin is still grinding around 63,000.
#标普500首次站上7700点, setting a new all-time high
The S&P 500 rose 1.79%, semiconductors rose 6.55%, and storage stocks surged across the board—SanDisk rose 10%, and Micron rose 7%.
Dabing has only followed by less than 1%. $BTC
The Nasdaq is already close to its previous high, and the market is still half of its all-time high.
The logic behind decoupling is clear. US stocks are pricing in "falling oil prices→ cooling inflation→ and falling interest rates," while the market is still being held back by a "56% probability of a rate hike in September."
For the same issue, US stocks interpret it as positive, while Bitcoin remains on the sidelines.
Piper Sandler said this recovery is "good, but not yet excellent."
The US-Iran agreement has not been formally signed, and Iran is still denying direct negotiations.
As long as the agreement isn't implemented, the geopolitical premium hasn't been cleared for a day.
If negotiations break down, oil prices rebound, and rate hike expectations jump back up, the foundation for this wave of US stock market highs will be loosened.
Bitcoin is now 63,000, half the price of US stocks, but it also carries twice the risk.Many people think MSTR's stock exchange for dollar reserves is "neglecting its duties," but in fact, this is precisely a top-tier open strategy aimed at a trillion-yuan market cap.
The core logic is simple: buying coins with STRC preferred shares is much more cost-effective than issuing common shares.
Issuing common shares to buy coins is a 1:1 exchange, so the amount of Bitcoin per share remains unchanged, effectively wasting all their efforts. But issuing STRC to buy coins is equivalent to getting $100 worth of Bitcoin at zero cost today, with only 12% interest paid annually.
Crunch: Bitcoin has risen 20% annualized over the long term, more than sixfold in 10 years, turning $100 into $619; And a 12% interest rate pays only $120 over 10 years. Inside, shareholders earned nearly $500 for free, without diluting their equity.
Issuing stocks now to exchange for dollar reserves is meant to solidify the balance sheet, allowing STRC to be issued at low cost over a sustained basis. When Bitcoin rises, this interest rate will be negligible compared to the gains.
To put it bluntly, it's not that they don't believe in Bitcoin, but rather that they've adopted a smarter, more sustainable approach to squeeze out the long-term compound interest of Bitcoin $BTC. #MSTR再卖1638枚比特币, scale halved #MSTR再卖1638枚比特币, scale halved
MSTR sells another 1,638 Bitcoins, halving the scale: from "hoarding maniac" to "active selling"
The Bitcoin whale Strategy (formerly MicroStrategy, MSTR), once known for its "buy only, not sell" strategy, has once again broken its promise. From July 27 to August 2, the company sold 1,638 Bitcoins, cashing out about $104.7 million, at an average price of $63,957. This sales volume is nearly "halved" compared to the 3,588 shares sold during the first share reduction at the end of June.
1. Why sell at a loss?
The average sale price was $63,957, far below the company's overall holding cost of $75,419, resulting in a paper loss of about $11,462 per coin.
The proceeds from the sale were precisely split in half:
· $52.4 million: Dividends paid on preferred stock
· $52.3 million: Repurchased discounted STRC preferred shares
The core goal is to save STRC. STRC is a perpetual preferred stock issued by Strategy, with a face value of $100, but it has long traded at a discount in the $89-92 range. CEO Phong Le clearly stated that the company's core goal is to keep STRC stable at $99 to $100 over the long term. To this end, Strategy is willing to sell coins and buy back shares at a loss to "eliminate high-interest debt."
2. A Bigger Blueprint: A $5 Billion Cash-Out Plan
This sale is just part of a larger strategic transformation.
On June 29, Strategy officially launched the "Digital Credit Capital Framework," completely breaking the unwritten rule of "never sell." In early July, the company initially authorized the sale of $1.25 billion in Bitcoin; a few weeks later, this target was quadrupled to $5 billion.
A triple use of $5 billion:
· Enhanced dollar reserves: up to $1.25 billion
· Dividends paid on preferred stock: approximately $1.76 billion
· Driving stock buybacks: an additional $2 billion
As of August 2, the company's dollar reserves had reached $4 billion. So far, only about $320 million worth of Bitcoin has been sold, achieving just 6.5% of the $5 billion target—even greater selling may be yet to come.
3. Not just selling coins: Sell stocks simultaneously
During the same period, Strategy also sold 3,011,361 shares of MSTR common stock through an ATM (Market-to-Market) offering, raising a net raise of approximately $290.6 million. $250 million injected into US dollar reserves, $28.9 million repurchased the STRC, and $11.7 million to supplement cash balances.
Longtime Bitcoin critic Peter Schiff criticized that Saylor is diluting MSTR shares while selling coins, "becoming a burden on MSTR and BTC." However, CEO Phong Le responded that the company's core goal is to increase the number of Bitcoins represented per share, and since 2020, MSTR has outperformed Bitcoin in every four-year holding cycle.
4. Market Reaction: MSTR actually rose
Interestingly, after the announcement, MSTR's stock price rebounded 1.7% that day, closing at $94.86. Analysts consider this a "relief rebound"—proactive selling of coins has addressed market concerns about the company's liquidity.
Some analysts (including Barclays and Cantor Fitzgerald) expect MSTR to break above $180 and aim for above $240. However, some viewpoints point out that MSTR is still consolidating below the lower boundary of the $100 to $180 range for 2024.
5. Summary
The strategic shift essentially represents a paradigm shift from "only buying, not selling" to "active capital management." The company still holds 842,138 Bitcoins (4% of the total Bitcoin supply), with a total holdings valued at approximately $63.5 billion.
But the narrative of "never selling" has come to an end. As whales begin selling at a loss to pay dividends, the market's long-standing narrative of belief in Bitcoin is undergoing a subtle repricing. Currently, its Bitcoin holdings have a book loss close to $11 billion. Whether further selling pressure will continue to depend on whether Bitcoin's price can rebound above $75,000—and whether STRC can return to face value.
$MSTR On August 5, several leading ETF issuers jointly submitted S-1 amendments to Solana's spot ETFs to the SEC.
Franklin Templeton, Bitwise, Fidelity, Canary Capital, CoinShares, Grayscale, and VanEck are all on the list.
Bloomberg ETF analyst James Seyffart stated directly that all Solana spot ETFs have submitted for a third revision, which is almost equivalent to locking in the approval signal.
Nate Geraci, President of NovaDius Wealth, also confirmed that the SEC is in talks with issuers regarding the terms.
The SEC has simplified the approval frameworks for Solana, XRP, and Cardano.
Unlike the lengthy tug-of-war in BTC and ETH approvals, the altcoin ETF channel is being systematically compressed in time.
CoinShares also filed an S-1 for its Solana staking ETF, aiming to list on Nasdaq.
Rex-Osprey's Solana+Staking ETF attracted $12 million on its first day, with AUM reaching $137 million.
The market demand for Solana ETFs is more real than many people expected.
The entire Solana fund system currently has about $2.4 billion in AUM, which is only 8.7% of the Ethereum ETF.
Solana's market cap is about 20% of Ethereum's, and there is still room for ETF penetration to double that gap.
If approved, once the gates for traditional financial capital open, SOL may be repriced. $SOL On August 5th, the SEC did something even more thought-provoking than the ETF approval itself.
The Jito liquid staking protocol for Ethereum Lido and Solana has been officially exempted from securities laws.
Jito is the largest staking and MEV infrastructure service provider in the Solana ecosystem, controlling a significant portion of the Solana staking market.
This SEC exemption sets a regulatory precedent for non-custodial staking. Previously, SOL was repeatedly labeled an unregistered security by the SEC in its complaint. As long as this qualification is not lifted, SOL's compliance ceiling will remain low. Jito's exemption does not directly launder SOL, but it takes the path of "liquid staking protocols without triggering securities laws." If Jito can pass, other DeFi protocols on Solana could theoretically follow the same path.
SOL is currently priced at $85.54, down 0.53% in the past 24 hours, and up 2.43% over the past 7 days. Market value is $49.2 billion, with a circulating supply of 575.6 million tokens. Spot trading volume in 24 hours was 579 million, contract trading was 10.2 billion, and leveraged trading volume was 17 times that of spot trading. This leverage ratio requires ongoing attention in highly volatile market environments.
$SOL 8月13号，Solana的TVL破了310亿美金的历史纪录。
超过1月20号创下的308亿高点，SOL当时293刀。
现在SOL只有185刀，比历史高点低了37%。
同样TVL，币价低了37%，说明链上锁的资产变多了，不是靠SOL涨价撑出来的。
USDC发行方Circle是Solana上最大的TVL驱动者，87亿美金，占了链上生态估值28%。
借贷协议Kamino、DEX聚合器Jupiter、流动性质押协议Jito贡献了107亿。
但Solana的交易量从1月峰值每周1030亿跌到了169亿，跌了84%。
Meme币退潮之后，链上手续费从5300万掉到7230万。
TVL在涨，交易量在跌，手续费在跌。链上资金在从投机性交易转向生息协议和稳定币托管。$SOL Global + SNDK Minimalist Situation of the Day
I. Macro Market Environment
1. Geopolitics: Expectations of easing US-Iran negotiations materialize, international crude oil plunges 5.7% in a single day, inflation pressure significantly cools down, US Treasury yields across all maturities decline simultaneously; the market lowers the probability of a Fed rate hike in September, overall risk appetite fully recovers.
2. US Stock Market: Dow Jones and S&P 500 both hit new all-time closing highs, Nasdaq surges 2.59%; the market is highly structured, with capital concentrated flowing into AI hardware, storage, and optical communication sectors, while broad consumer and some internet stocks weaken.
3. Pre-market Futures: Before today's open, the three major stock index futures show slight divergence, bullish sentiment marginally contracts, entering a cautious wait-and-see phase before earnings releases.
II. Sector Performance (Semiconductors/Storage)
Philadelphia Semiconductor Index soars 6.55%, rising for four consecutive days; storage chips rally across the board: SanDisk and Intel surge over 10%, Micron and Hynix also rise sharply.
Underlying drivers: Spot and contract prices for NAND and DRAM continue to rise week-over-week; global AI data center expansion brings long-term storage demand, industry cycle reversal logic is being repriced by capital.
III. SNDK SanDisk Stock Situation
1. Price Performance: Closed at $1427.62 yesterday, intraday high $1446.62, volume-driven strong rebound, representing a robust recovery after a deep correction, but has not yet returned above the previous downtrend.
2. Core Variable: Q4 earnings report to be released after US market close today; options market prices intraday volatility at ±13.1%, the largest trigger point for the day's movement.
3. Technical Structure: Short-term strong bullish trend but dense resistance above; before earnings release, the market mainly plays on expectations, with amplified long-short divergence.
IV. Short-term Market Summary
Easing external environment + industry fundamentals support this round of storage sector rebound; all volatility ultimately depends on tonight's SNDK earnings data to determine the strength and continuation, with range-bound oscillation expected before the report.8月4号，Phantom钱包宣布收购了Solana上的AI Meme币交易平台Solsniper。
Phantom今年早些时候刚以30亿估值完成了1.5亿美金的C轮融资，红杉和Paradigm领投。
Phantom从Solana生态钱包起家，到收购Solsniper，再到年初收购NFT数据平台SimpleHash，方向很清晰，从钱包变成Solana生态的超级入口。
Solsniper会保持独立运营，说明Phantom想要的是技术能力和用户数据，不是简单的用户并表。
SOL现在是85.54美元，合约持仓51亿美金。
多空双方在等ETF的最终结果。
CoinShares、富达、Bitwise、灰度都在排队，SEC与发行方的对话在推进。等ETF获批之后再决定怎么走。$SOL Reasons for SanDisk's (SNDK) Surge Yesterday + Future Market Scenarios
1. Core Catalysts for Yesterday's (August 4 US Market) Surge
Flash Memory Summit Major News (Most Direct Trigger)
Flash Memory Summit 2026: SanDisk and SK Hynix jointly released the HBF high-bandwidth flash universal standard.
Aimed at solving the AI inference “memory wall,” targeting cloud/edge AI, integrating GPU + high-speed NAND new solutions. Market interpretation: SanDisk successfully enters a new AI storage track, opening long-term growth potential, no longer just a traditional NAND manufacturer.
Industry Fundamental News Resonance
Multiple media confirmed: Samsung, Micron, and SK Hynix's 2027 DRAM/HBM capacity has been fully locked by cloud providers; NAND spot and contract prices continue to rise in Q3; market strengthens the expectation that "storage tight supply and demand will last at least until the end of 2027."
Technical Aspect: Oversold Rebound and Capital Replenishment
SanDisk has been continuously correcting from its high point since late July, with short-term chips fully exchanged; the Philadelphia Semiconductor Index collectively strengthens, AI computing power and storage sector funds return, trend funds bottom-fishing.
Event Expectation: Pre-earnings Window Speculation
The market is speculating ahead of the upcoming earnings release, with capital expectations: enterprise SSD orders, gross margin, and long-term supply agreements (LTA) data continue to improve.
2. Next Market Scenarios: Three Possible Outcomes
[Optimistic Scenario (Bullish Logic)]
✅ Trigger Conditions: Earnings revenue, gross margin, and next quarter guidance significantly exceed expectations; management confirms NAND price increases continue and HBF commercialization timeline.
Trend: Stabilize above $1450 resistance, challenge the $1600–$1750 range upward.
Driving Logic: AI storage narrative continues to strengthen, re-pricing SanDisk from a cyclical stock to an AI growth target.
[Neutral Scenario (Most Probable)]
Trigger Conditions: Earnings data meet market expectations, no surprises or negative news.
Trend: Wide-range consolidation at high levels
Reference Range: $1300 — $1500.
Characteristics: Profit-taking digestion after a big rise, repeated tug-of-war. Storage sector has significant long-short divergence (institutional target price range $1000~$3250), lacking new catalysts makes continuous one-sided rise difficult.
[Bearish Scenario (Needs Close Attention)]
❌ Risk Triggers (Any one can cause a pullback):
Earnings guidance below expectations, newly signed order prices weaken month-over-month;
Major manufacturers signal capacity expansion, market worries about increased supply by end of 2027;
Overall US tech market pullback, high-volatility storage stocks sharply decline;
Cloud providers lower capital expenditure expectations.
Trend: Effectively break $1300 support, test $1150–$1200 range.
3. Key Indicators to Monitor Continuously
Three Core Earnings Highlights
① Data center business revenue growth rate; ② NAND average selling price (ASP) guidance; ③ Long-term supply order scale.
Industry Tracking: TrendForce monthly NAND contract prices
External Macro: Federal Reserve policy expectations, liquidity environment for tech growth stocks
Competitor Dynamics: Samsung, Kioxia capital expenditure and expansion plans
4. Brief Summary
Yesterday's rise was a multiple resonance of technical oversold rebound + HBF technology benefits + storage market optimism.
Short-term, a simple continuous surge is unlikely; the biggest variable is the upcoming earnings report.
Key Watershed: $1450 resistance, $1300 support. Holding above resistance opens upward space; breaking support ends the rebound phase and returns to adjustment.
#闪迪财报前夕，HBF与存储紧缺引发热议 #韩国杠杆ETF成交额降九成，波幅收窄 $SNDK #标普500首次站上7700点，创历史新高 真正让这币凉的，不是市场，是链上。
8月2号，链上分析师Ai姨监测到——那个“LAB疑似内幕地址”时隔三周又动了，向某交易平台充值了580万枚LAB，价值83.4万美元，均价0.1439美元。
充值完一个半小时内，价格从0.1439直接砸到0.1273，跌了11.5%。
这不是第一次。
7月12号同一地址向Aster转移了915万美元的LAB，币价当天暴跌42.5%。
每次动一下，价格就崩一次。这个地址目前还持有7470万枚LAB，价值约1011万美元。
按现在的价格算，这7470万枚还能砸出1000多万的卖盘。
而24小时现货成交量才219万——这些货够砸好几天。$LAB 链上追踪到的东西更完整——2026年4月，跟LAB团队相关的实体从团队收到了超过1.96亿枚LAB，随后分批转入交易平台充值地址，5月中旬从平台提出来分散到10个地址，休眠了两个月，7月开始往市场砸。
截至目前这个实体仍持有约8150万枚LAB。
还没完。7月14号，1623万枚LAB解锁，占最大供应量的1.6%，当时值406万美元。投资者总量1.92亿枚，目前只解锁了9460万枚，占49.27%。8月14号、9月14号、10月14号、11月14号、12月14号，每个月还有1623万枚要解锁。70.8%的代币标注“未追踪”，可能随时解锁。
团队保持沉默，社区在追查钱包动向，但没人能确认开发者是不是已经走了。一个连团队在不在都不知道的项目，你问我0.157是不是底？我哪知道。但我知道一件事——0.18的时候也有人问过同样的问题。$LAB Wakaka, SpaceX's earnings report production is super impressive, like a StarCraft movie
SpaceX Q2 earnings report submitted, business stronger than expected, but I still maintain a cautious view.
I finished reading SpaceX's first earnings report after going public first thing today, and the operating data is indeed impressive:
Q2 revenue about $7.8 billion, up 92% year-over-year, exceeding market expectations;
Operating loss significantly narrowed, from nearly $1 billion in the same period last year to about $140 million;
Starlink revenue continues rapid growth, still an important pillar of cash flow;
AI business revenue greatly increased, starting to gradually realize commercial value from the investment phase.
But the other side is also worth attention.
Capital expenditures remained high this quarter, with AI infrastructure investment accounting for the vast majority. Management also stated that large investments will continue in the coming quarters. This means the company is growing fast, but cash burn is also huge. The market will next focus more on whether the investments can continuously convert into profits.
What really deserves observation is the stock price performance.
Before the earnings release, $SPCX once surged to about $130, forcing many shorts to cover; but after the report landed, the stock price quickly fell back, dropping below $120 again, with most of the gains from the previous two days quickly given back.
This indicates the market is not denying SpaceX's fundamentals but is reassessing whether the current valuation has already priced in future growth.
The biggest variable coming up is the first batch of restricted shares unlocking on August 6.
Unlocking does not mean all shares will be sold immediately, but the circulating supply will obviously increase, and the market will need more new funds to absorb the supply. For high-valuation growth stocks, supply and demand often influence short-term stock prices more than an excellent earnings report.
SpaceX remains one of the world's most competitive aerospace and technology companies, with Starlink, Starship, and AI all having long-term potential.
However, a great company does not necessarily mean any price is worth buying.
At this stage, I am more focused on the supply pressure brought by the unlocking and whether the market funds can smoothly digest it, rather than the earnings report itself. Therefore, before the unlocking risk is fully released, I still maintain a relatively cautious attitude.
The business can continue to be bullish on, but short-term valuation still needs observation. On August 4, Ethereum spot ETFs saw a net inflow of $48.72 million.
BlackRock's staked version of ETHB saw a single-day net inflow of $5.77 million.
On August 4, Bitcoin ETFs saw a net outflow of 168 million, with the two sides moving in completely opposite directions.
Funds left the BTC ETF and then moved into the Erbing ETF. This trend started in mid-July and has lasted for nearly a month.
BlackRock's total net inflow into ETHB has reached $550 million. Grayscale Mini Trust saw a net inflow of 13.31 million ETH in a single day. Large sums of money are indeed moving toward the Er Bing ETF, with a clear direction.
But there's one issue worth pondering—ETFs are buying, but prices aren't rising. On August 4th, nearly 50 million yuan was bought in from the ETF, and the price was still fluctuating around 1,850. Some are entering through ETFs, others are selling in the spot market—both sides are opposed. The total ETF market value is now around 10.2 billion, accounting for 4.55% of the total market capitalization of Erbing. The market is still small and not yet capable of dominating prices.
What's even more troublesome is retail investors selling off. Data from FXStreet shows that retail investors sold about 360,000 ETH over the past week. Whales are buying, ETFs are entering, retail investors are selling—the three forces are canceling each other out around 1,850. Whoever can't hold out first will have their price go in their favor.
$ETH The total amount of Ethereum staked has reached a historic high of 41.4 million ETH, accounting for 34% of the total supply.
In just the past week, more than 1.4 million ETH have been newly staked.
The validator entry queue was still in line, and there were far more people wanting to come in than wanting to leave.
What's even more noteworthy is the whale address 0x2e80, which has withdrawn a total of 112,000 ETH from Gemini over the past three weeks, worth $208 million, all of which were staked. On August 4, another 19,000 coins were issued, valued at $35.44 million. This address is continuously buying and staking, with a very stable pace, not a short-term trade.
Meanwhile, Arthur Hayes bought another $5 million worth of ETH yesterday. He just bought near 1,700 in July, and now it's still rising at 1,850, with a clear rhythm—buy when it drops, hold it when you buy.
But L1 network activity is cooling down. While whales are accumulating frantically, activity at the on-chain base layer is declining. Big players are buying, while regular users are using less. If this divergence continues, either the whales have misjudged or the market has priced it wrong.
$ETH AMD's numbers this quarter look pretty good: revenue of $11.5 billion, up 50%, with data centers doubling to $6.7 billion. The stock price surged briefly during trading but plunged nearly 10% after hours.
The market is honest: you exceeded expectations, but I have already bought out the "better" in advance.
AMD today is no longer the company that once relies on Ryzen and gaming to turn its fortune around. Data centers account for nearly 60% of revenue, while gaming business dropped by more than 30% year-on-year, which is basically negligible. Right now, it's competing with NVIDIA for AI computing power. Helios is starting to ramp up volume, and the visibility of major client orders is rising—these are all real. Lisa Su said data centers could double again by 2027, and the market is willing to listen.
But that's where the problem lies.
The stock price has nearly tripled in one year, and valuations have already priced in many optimistic scenarios in advance. If you give a "good" financial report, people will just think, "That's just it." Although the guidance exceeded consensus, it was still far from the more aggressive figures some expected, and funds were immediately realized. Capital expenditures have risen, cash flow is suppressed, and this has also given a pretext to crash the market.
My view is quite straightforward:
There is indeed short-term overdraw. High expectations combined with high valuations have very low margin for error. Any result that is "just good but not explosive enough" will be used to smash. This after-hours decline is essentially due to poor expectations, not a sudden deterioration in fundamentals.
What about the medium to long term? The story isn't over yet. The demand for AI computing power is real, and AMD is indeed moving from a "Nvidia alternative option" to "having an independent market share." But if Helios and Instinct's growth in volume and market share don't meet the expectations implied in current pricing, they will continue to be proven wrong.
For those making trades, it's now more like betting on execution than on narrative. The narrative has already been fully priced in; the remaining question is whether it can be delivered on time and in the right amount.
In short: The fundamentals are decent, but the stock price has almost exhausted the "decent" part. Next, we'll see if it can continue to exceed expectations, rather than telling the AI story again. #AMD财报超预期 has growth been overdrawn? 以太坊研究员Justin Drake提交了一份EIP-8361草案，提出“梯度发行销毁”机制。
核心逻辑是：随着质押比例上升，逐步销毁越来越多的验证者奖励。
当质押比例达到50%（约6,025万枚ETH）时，共识层的净发行量将降至零。
目前质押率已经34%，按现在的速度，50%不是遥不可及的事。如果这个提案通过，ETH的供应曲线会被彻底改写——从“通胀资产”变成“通缩资产”的速度会大幅加快。但提案还在讨论阶段，通过时间和具体执行方式都不确定。
美联储那边，9月加息概率还在58%左右挂着，名义上没加息，但内部三票反对加息的裂痕已经公开化了。油价一跌，加息概率就降一点；美伊一谈，市场就松一口气。二饼现在1,850，比4,950的历史高点还差60%。鲸鱼在买，ETF在进，质押在涨，散户在卖，宏观在压。方向没出来之前仓位别太重，等1,900站稳或者1,830破了再说。$ETH #财报观察员：业绩喜忧参半，解禁将至！SpaceX后续怎么看？
Just finished reading SpaceX's first quarterly report since going public, and two days later the epic lock-up expiration is landing. This giant spanning aerospace, satellite broadband, and AI computing power is now fully supported by half fundamentally positive factors, with half the risk of selling pressure looming. Here's my take.
1. Breaking down the earnings report: strong positives, unavoidable pain points
Core positives: Starlink generates steady cash flow, revenue growth far exceeds expectations
Q2 total revenue was $7.81 billion, up 92% year-over-year, more than 10% above Wall Street estimates, firmly securing the profitable core business.
1. Starlink is the company's only cash cow
Quarterly revenue of $4.29 billion, operating profit of $1.66 billion, profit margin surged to 38.6%, netting $18 million daily; global subscribers reached 12 million, high-priced government, enterprise, aviation, and maritime orders doubled, large U.S. government contracts continue to land, moving toward becoming the fourth largest telecom operator in the U.S. This business has solid cash flow, and as long as overseas low-end user expansion stabilizes ARPU, profits can steadily increase.
2. Overall losses narrowed significantly, aerospace launch business steadily ramping up
Net loss narrowed from $1.01 billion last year to $541 million, nearly halving the loss; rocket launch business surged 55% quarter-over-quarter, Starship test flights continuously break technical bottlenecks, in-house satellite networking speed is industry-leading, aerospace barriers have no short-term competitors.
3. AI business losses better than pessimistic expectations
The market originally predicted AI computing power segment losses exceeding $2.4 billion, actual loss was only $1.26 billion, burn rate slower than expected, but revenue has started to grow, a small surprise.
Negative risks: capital expenditures surged, stock price plunged over 8% after hours
This is the core reason the report scared off funds: Q2 capital expenditures were $18.37 billion, up 550% year-over-year, management clearly stated that the burn scale will not shrink in the next two to three quarters.
Money mainly spent on two areas: large-scale AI data center expansion, Starship iteration + Starlink satellite mass production. Simply put: the money Starlink earns now cannot fill the investment gap in AI and aerospace; the company still has a long way to go to overall profitability.
Adding to this, the AI sector currently faces generally cautious U.S. stock funds, no longer blindly funding high valuations and high investment, slow profit models; funds are voting with their feet, and the stock price plunged immediately after the earnings release.
2. Major risk: $100 billion market value lock-up expiration on August 6, short-term selling pressure maxed out
This lock-up expiration is one of the largest single lock-up waves in U.S. stock history, key points explained:
1. Massive scale: new float > current total float
On August 6, 911.5 million shares held by internal employees and early investors will unlock, corresponding to a market value over $100 billion; after unlocking, the float will surge 140%, previously scarce post-IPO shares drove the price to a $225 high, now shares flood the market, supply and demand completely reversed.
2. Strong motivation to sell, stock price already below IPO price but early shareholders still hugely profitable
IPO price was $135, current price around $110, already below issue price; but early VCs and employee holdings cost very low, even with the stock halved, paper gains remain 10 to 50 times, strong desire to cash out. The only stabilizing factor: Musk's high-voting shares are locked until 2027, preventing large-scale sell-offs that would crash the price.
3. Shorts pre-positioned, amplifying volatility
Short positions before lock-up already account for 34% of float, over $24 billion; if concentrated selling occurs after lock-up, shorts will further amplify the downtrend, short-term price volatility will max out.
3. Personal outlook: two scenarios for short-term and mid-to-long-term trends
Short-term (1-2 weeks, lock-up digestion period) cautious, mainly pressured with volatility
Fundamental positives are unlikely to offset the selling pressure from the $100 billion lock-up, likely two scenarios:
① Weak trend: concentrated selling on lock-up day, price dips to $100-$105 range to find support;
② Volatile trend: top institutional funds absorb most selling pressure, price consolidates between $110-$130, slowly digesting panic with Starlink order positives.
For short-term trading, I won't rush to bottom-fish; I'll wait 2-3 trading days after lock-up for turnover to stabilize and no large-volume drops before looking for safer opportunities.
Mid-to-long-term (quarterly dimension): focus on two key inflection points to determine if valuation can hold
1. Positive inflection: AI business revenue rapidly climbs, Starlink profit growth outpaces capital expenditure growth, market sees overall business profitability timeline, supporting high valuation;
2. Negative inflection: AI burn rate continues exceeding expectations, Starlink overseas user growth slows, combined with multiple subsequent lock-up batches unlocking, valuation will compress further.
Long term, Starlink's satellite communication monopoly barrier and Starship's low-cost launch technology are hard assets, but current valuation overextends AI optimistic expectations, requiring performance to be realized step by step.
A few casual words
Market divergence is huge now; some think the halved stock price plus lock-up negative is the worst, planning to gradually accumulate; many funds think digesting $100 billion selling pressure will take at least half a month, preferring to watch and not touch. What do you think after this lock-up, will SPCX first hit new lows or will funds directly bottom-fish for a rebound? AMD财报超预期，盘后却跌超9%为什么？业绩满分，预期通胀，这就是AI芯片股的残酷游戏
核心矛盾在于“预期太高”：
1. Q3指引“不够炸”：公司预计Q3营收约130亿美元，高于分析师平均预期的125亿，但部分买方已看高至135-140亿。当股价年内已涨超140%，市场要的是“惊喜”而非“合格”。
2. 资本支出暴增：Q2资本支出达8.08亿美元，是市场预期2.986亿美元的近三倍，引发市场对利润和现金流的担忧。
3. 电话会未能打消疑虑：当分析师追问Q3增速为何低于Q2时，CFO未能给出明确乐观答复。
今晚看闪迪！
#AMD财报超预期，增长已被透支？
$SNDK $AMD $BTC 开局一张图，剩下全靠编？不，老高今天跟你唠点实在的。
SpaceX首份财报出来了，营收78.14亿，同比暴增92%；EBITDA 35亿，比预期高75%；AI亏损收窄超预期。数据漂亮不？漂亮。
然后盘后跌了近9%。
为啥？业绩好有个屁用，明天千亿解禁压顶。8月6日9.115亿股解禁，按110刀算超1000亿美金。资金不跑等着挨刀？
---
老高的心路历程：从“卧槽牛逼”到“卧槽快跑”
说实话，刚看到财报的时候，老高是真觉得要起飞了。营收78亿，市场预期才69亿；AI收入同比暴增247%；星链用户翻倍到1200万。星链独撑利润16.6亿，AI还在烧钱但亏损在收窄。
马斯克还说2030年要干到万亿营收。听着热血沸腾对吧？
但老高冷静下来一看——资本支出183.7亿，其中AI占了158亿。而且高管说了，三四季度资本支出跟二季度差不多。这烧钱速度，比老高吃火锅还猛。
更狠的是明天解禁。1000亿美金的新筹码涌入市场。这特么不是利空，这是海啸。
交易方向：短期别接飞刀，中期盯着两个信号
短期，观望为主。解禁≠全部卖出，创始人锁仓到2027年，但早期投资者和员工才是主要抛压。股价已经比135刀的IPO发行价跌了快20%。愿意在这个位置卖的人有多少？没人知道。老高不建议你去赌这个变量。
策略上：如果你想参与，等解禁落地、抛压释放完再看。支撑位看100-110区间，跌破别抄，企稳再说。
另外注意传导——闪迪财报在即。存储板块近期涨幅已经很大了，闪迪盘前涨超10%。如果闪迪资本开支指引也超预期，很可能复刻SpaceX的走势。财报前仓位别太重。
交易心得：老高用真金白银换来的三条铁律
第一，财报好≠股价涨。 市场看的是“预期差”和“接下来会怎样”。资本开支比预期高，比营收超预期更致命。记住这句话：市场奖励的是惊喜，惩罚的是惊吓。
第二，解禁面前，基本面靠边站。 再好的业绩，也扛不住千亿筹码砸盘。短期筹码结构比长期叙事重要一万倍。别跟供应量作对。
第三，别跟风，别接飞刀。 Palantir涨30%那是人家业绩兑现了。AMD跌9%那是指引没满足激进预期。科技股内部正在剧烈分化。你得看清楚自己手里拿的是什么。
老高说完了。你细品。该怂的时候怂，该干的时候不含糊。
$SPCX $SNDK $AMD
#财报观察员：业绩喜忧参半，解禁将至！SpaceX后续怎么看？
#SpaceX首份财报超预期，解禁仍是关键变量
#AMD财报超预期，增长已被透支？ Filecoin 昨晚在 X 上把两件事放在一起：它引用产业追踪说，2026 年专用工业 SSD 的交付周期可能达到 20 周以上；同时又称 Filecoin 存储已经部署运行，没有同样的等待时间。这个对比有传播力，但不能直接当成“硬件短缺已被网络存储解决”。
第一，TrendForce 7 月的 NAND 更新确实把 2026 年的紧张归因于 AI 相关需求上升和扩产受限，并预计新增供给要到 2027 年下半年才更明显。第二，Filecoin 3 月公布的 Onchain Cloud 主网上线数据是 49.41 TiB、478 个活跃数据集和 81 个付款钱包，说明产品已经运行，但这只是项目方披露的早期基线，不能替代客户侧的可用性、延迟、检索成功率或合规评估。
我的判断：SSD 交付周期描述的是特定硬件，Filecoin 提供的是由多个存储提供者承载的数据服务，两者解决的约束并不相同。真正值得追问的不是“有没有容量”，而是指定数据能否按 SLA 可靠写入、取回、复制和长期保存。当前可以确认供应链压力与 Filecoin 产品上线，不能确认它已成为工业客户的即插即用替代方案。Trump administration plans to ban Chinese optical modules: A-share "Yi Zhongtian" is shocked, but "decoupling" is not so easy
1. Overview of the Incident
On August 4, Reuters cited four sources reporting that the Trump administration is drafting a ban to prohibit imports of new Chinese data center modules, mainly targeting optical modules (optical transceivers). The FCC is formulating related measures, and U.S. officials hope to announce and implement them within 2026, citing the aim of "preventing Chinese companies from stealing data or embedding malware."
However, insiders also emphasized that the FCC may still modify or suspend the restriction. The Chinese Embassy in the United States has responded, urging the U.S. side to "stop smearing Chinese companies" and warning that all necessary measures will be taken.
2. Market Reaction: A-shares are shocked, while U.S. stocks are on a reverse frenzy
On the A-share market, the optical communication sector plummeted collectively in the morning session of August 5:
· Zhongji Xuchuang: Fell over 13% at one point, narrowed the closing loss to about 7%
· New Yisheng: Fell over 10% at one point, narrowed the closing loss to about 4%
· Tianfu Communication: Fell over 6% at one point, narrowed its closing loss to about 3%
· Liante Technology: Down over 11%, Cambridge Technology approaches limit-down
The direct reason for the plunge: these companies are highly dependent on the U.S. market—over 90% of Zhongji Xuchuang's revenue comes from overseas, while over 96% of Xinyisheng's revenue is overseas.
On the US side, however, the opposite trend reversed—AAOI surged 19%, Coherent and Lumentum both surged 15% intraday, and Corning rose nearly 8%. The market logic is clear: if Chinese optical modules are banned, orders will flow to domestic U.S. suppliers.
3. In-depth Analysis: Why Is "Decoupling" Not So Easy?
1. The probability of implementation is extremely low
CITIC Securities bluntly stated that the final result is "all talk and little action," and CICC also stated that "in the medium to long term, it's highly likely that there will still be more talk than action." The core logic is:
Chinese manufacturers hold absolute global dominance: 7 out of the world's top ten optical module manufacturers are Chinese companies, and 800G/1.6T high-speed optical modules account for as much as 70% of shipments by Chinese manufacturers. North American domestic manufacturers have obvious shortcomings in both production capacity and technological advancement.
The substitution gap cannot be bridged in the short term: long certification cycles for optical modules, slow ramp-up in mass production yield, and delivery stability requiring years of validation. North American companies may not be able to fill the gap for years, and the progress of AI infrastructure in the US will be directly slowed down.
U.S. cloud providers strongly oppose this: the ban will drive up costs for cloud providers like Amazon. When tariffs are added in 2025, optical modules will be included in the exemption list due to strong opposition from U.S. cloud providers.
2. Even if implemented, the impact is controllable
Chinese manufacturers have long been prepared: since 2022, companies like Zhongji Xuchuang and Xinyisheng have already established production capacity in Thailand, and from 2025, they will further advance production lines in the United States. The supply chain's "physical address" has long been iterationed, and policies have made it difficult to truly block delivery.
The ban has an "exemption" design: the FCC plans to exempt many non-Chinese suppliers. Even if you land, you may still have many exemption situations.
3. Deep binding of China-US industrial chains
Optical modules are essentially optoelectronic signal conversion devices and do not store business data. Upstream core components such as DSP chips and analog chips mostly come from American suppliers such as Marvell and Broadcom. A one-size-fits-all restriction only leads to mutual destruction.
4. Summary
This wave of "ban panic" on optical modules is essentially another round of emotional shock amid the US-China tech rivalry. In the short term, market panic is understandable—the related companies' revenue from the U.S. is indeed too high. But a calm analysis reveals that China's global dominance in optical modules cannot be replaced in the short term, and the economic cost of the ban is equally enormous for the U.S. itself.
For investors, after panic sell-offs, it's important to distinguish between emotional shocks and fundamental deterioration—the former is an opportunity, the latter is a risk. At least for now, this seems more like a "ghost story" amplified by emotions, rather than a fundamental reversal of industry logic.
$MU Strategy is selling coins again.
From July 27 to August 2, the company sold 1,638 BTC at an average price of $63,957, cashing out approximately $105 million.
Open interest dropped from 843,775 to 842,138. This is the third coin sale in 2026.
Six consecutive weeks without buying coins, selling coins twice in a row—the first time selling 3,588 tokens (216 million), and this time selling 1,638 coins (105 million).
The scale is shrinking, but the frequency is accelerating. Cash reserves have grown from 3.2 billion to 4 billion.
mNAV has long since fallen below 1, and the cycle of buying coins through premium financing has ended.
Strategy is shifting from a "Bitcoin hoarder" to a "Bitcoin capital manager."
Bitcoin is now around 64,000, half the price of its historical high. But with US Treasury yields above 5% right in front of us, institutions don't come, retail investors dare not move, miners are transforming, and big players are re-rotating positions. Don't hold too much position before the direction emerges; it's not too late to wait until 64,250 is above before watching. Wait for the direction to emerge before following it—better than blindly rushing in the fog.
$BTC 亚太市场率先传来暖意，日韩股市双双高开，给新的一周开了个好头。 日本方面，日经225指数开盘后迅速拉升，涨幅扩大至2%，显示出较强的买盘意愿。韩国市场表现更为激进，KOSPI指数开盘即上涨3.8%，强势收复6600点整数关口。$SKHYNIX 值得注意的是，这已经是该指数近期连续第三个交易日尝试上攻这一关键位置，今天的突破似乎为市场注入了一剂强心针。 盘面最亮眼的当属半导体产业链。权重股SK海力士一马当先，涨幅超过7%，领跑全场；三星电子也不甘示弱，大涨5%。这两家存储芯片巨头联手走强，很大程度上撑起了大盘的涨幅。 有市场分析认为，这可能与近期全球AI算力需求持续升温、存储芯片价格预期企稳有关，资金开始重新回流到这一前期调整充分的赛道。$SNDK 不过，细看今天的上涨，消息面上并没有出现特别爆炸性的单点利好。这轮反弹更像是前期超跌后的一次情绪修复，加上进入8月，部分机构资金开始布局三季度行情，趁着消息面相对平稳，选择在权重股上回补仓位。 一天的上涨还不能说明太多问题。日经指数上方仍有均线压力，KOSPI指数能否站稳6600点也需要明天继续观察。对于普通投资者来说，这种普涨行情下最忌Bitcoin spot ETFs saw a net inflow of $211 million yesterday: BlackRock IBIT alone took $170 million, institutions are "buying up"
1. Data core
On August 4, spot Bitcoin ETFs saw a total net inflow of $211 million, ending several days of capital swings:
· BlackRock IBIT: Net inflow of $170 million, historical total net inflow of $60.763 billion
· Fidelity FBTC: Net inflow of $19.5752 million, historical total net inflow of $9.972 billion
· Total net asset value: $78.258 billion, ETF net asset ratio of 6.06%
· Historical cumulative net inflow: $51.706 billion
2. Data Interpretation: IBIT is "Dominating," and institutions are pricing their positions
1. BlackRock IBIT Reaches $170 Million in a Single Day, Accounting for Over 80%
Of the total 211 million inflows, IBIT alone accounted for 170 million (80.6%). This indicates that traditional institutional funds mainly enter the Bitcoin market through BlackRock, the "largest channel," while other ETF products have relatively limited appeal. BlackRock IBIT's total historical inflows have reached $60.763 billion, making it the absolute dominant among the 10 spot Bitcoin ETFs.
2. After continuous outflow, the first obvious backflow occurs
In the previous week, spot Bitcoin ETFs saw a net outflow of $423 million, with $178 million outflowing on July 31 alone. Yesterday's net inflow of $211 million marked the beginning of institutional funds "buying up" again. BlackRock ETF inflows are seen as a more stable long-term allocation demand rather than short-term speculative funds.
3. Contrast: ETH ETFs are still bleeding
On the same day, Ethereum spot ETFs saw a net outflow of $11.4178 million, forming a sharp contrast with BTC ETFs. Funds are shifting from ETH to BTC, reflecting that institutions are more inclined to hold "digital gold" Bitcoin amid macro uncertainty over the more volatile Ethereum.
3. Significance for the market
1. Institutions continue to accumulate shares in the $62,000-64,000 range
Yesterday's $211 million inflow was no coincidence—on July 30, IBIT also saw a single-day inflow of $126 million. Institutions are continuously buying in the $62,000-$64,000 range, providing a solid demand base for Bitcoin's price.
2. Forming a "counterparty" with retail sell-offs
Previous on-chain data showed retail investors had net outflows of about 360,000 ETH over the past week, while Bitcoin retail investors continued to sell off. The $211 million net inflow from ETFs coincided with the panic selling by retail investors, forming a counter-market trade—chips were shifting from retail to institutions.
3. With a net asset ratio of 6.06%, ETFs are becoming a major force in Bitcoin pricing
The total net asset value of Bitcoin spot ETFs has reached $78.258 billion, accounting for 6.06% of Bitcoin's total market capitalization. This ratio continues to rise, indicating that ETF capital flows are increasingly influencing BTC prices.
4. Summary
Yesterday's net ETF inflow of $211 million was the latest evidence of institutions "buying up" in the $62,000-$64,000 range. BlackRock IBIT led with $170 million in inflows, indicating a trend of traditional institutional funds continuously allocating Bitcoin through the largest ETF channel.
This stands in stark contrast to panic selling by retail investors and the continued outflows of ETH ETFs—the market is undergoing a massive capital migration "from retail to institution, from offcoins to Bitcoin." As long as the ETF maintains its net inflow trend, Bitcoin's bottom support in the $62,000-$64,000 range remains solid.
$ETH Bitcoin fluctuated around 64,200 today, rising just over 1% in 24 hours.
Yesterday morning it was just over 62,300, but by the end of the night it had come back nearly 2,000 dollars.
The trigger was oil prices.
Brent crude plunged nearly 5% yesterday to $83.77. Reports of progress in negotiations between the US and Iran have emerged, and the market interprets this as a cooling of geopolitical risks.
As oil prices fell, inflation expectations cooled, and US Treasury yields also fell—the 10-year yield fell to 4.613%, and the 30-year yield fell from a high of 5.25%.
CME data shows that the probability of a rate hike in September has dropped from over 80% to 58.4%.
The market is even more aggressive, directly betting that there will be no rate hikes in September.
Technically, Bitcoin is trading at the end of a converging triangle between 63,000 and 64,200. If the volume increases and the above level holds above 64,250, it will directly open up short squeeze space toward 65,500 to 66,000. Below 63,000 is the first line of defense; if broken, it may look for support near 62,000. I personally won't heavily bet on the direction at this level; I'll wait until it really breaks through 64,250. Don't overdo your position. Before the direction is determined, it's better to watch more and move less than to move recklessly. $BTC On August 4, spot Bitcoin ETFs saw a net outflow of $168 million.
The day before, August 3rd, there was still a net inflow of 223 million yuan, with nearly 400 million yuan reversed in one day.
BlackRock IBIT saw an inflow of 246 million the previous day, and on August 4, Grayscale's GBTC saw a outflow of 69.11 million.
Big money is reallocating positions, not a systematic retreat. But the problem is, when the ETF direction is unstable, it's hard for Bitcoin to break out of a trending rally. The opposite direction for two consecutive days indicates that the institution itself is hesitating. At this level, both long and short positions are tough. Don't overload your position before the direction is released.
$BTC A new variable emerged on the macro side—the US and Japan joined forces to intervene in the forex market.
Japanese Finance Minister Satsuki Katayama officially announced that the two countries have taken joint action in the foreign exchange market, buying yen and selling US dollars.
This is the first joint intervention by the US and Japan in the yen exchange rate since 1998.
The U.S. Treasury also withdrew $77 billion from the supplementary reserve account to support the yen. This means that liquidity in U.S. Treasury bonds is being drained. If pressure on U.S. Treasury supply continues to increase, global funding costs may rise further. The 30-year Treasury yield once surged to 5.25%, and although it has now fallen back to 5.17%, it remains at a high level. Bitcoin's drop from 65,000 to 63,000 is not due to fundamental issues, but rather macro pressure. $BTC In early August, Bitcoin hovered around 63,700, but mining costs for listed mining companies had already reached nearly $80,000.
Every coin you mine loses money, averaging about $19,000 per coin.
But miners' stock prices are actually rising—the market no longer sees them as "miners," but as "AI infrastructure providers" pricing them.
CoreWeave, Anthropic, and Microsoft have already signed computing power orders worth tens to hundreds of billions with mining companies. Computing power is flowing from the Bitcoin network to AI data centers. Mining difficulty has dropped from 156 trillion in November 2025 to 126.23 trillion, a decline of 19.9%, the third largest drop since the advent of professional mining machines.
Miners are selling, but not out of panic—they've found a more profitable business than mining.
$BTC Strategy又卖币了。
7月27日到8月2日，公司以均价63,957美元卖出了1,638枚BTC，套现约1.05亿美元。
持仓从843,775枚降到842,138枚。这是2026年第三次卖币。
连续六周没买币，连续两次卖币——第一次卖了3,588枚（2.16亿），这次卖了1,638枚（1.05亿）。
规模在缩小，但频率在加快。现金储备从32亿干到了40亿。
mNAV早就跌破1了，靠溢价融资买币的循环已经断了。
Strategy正在从“比特币囤积者”变成“比特币资本管理者”。
大饼现在64,000附近，比历史高点便宜了一半。但5%以上的美债收益率摆在面前，机构不来，散户不敢动，矿工在转型，大户在换仓。方向没出来之前仓位别太重，64,250站上去再看不迟。等方向出来了再跟，比在迷雾里瞎冲强。
$BTC $SPCX made a profit, but the stock price fell, all the money was burned on AI
The performance looks good, right? Then the stock price dropped 7% after hours.
The reason lies in one number: capital expenditure. In Q2, 18.4 billion was spent, of which 15.8 billion was invested in AI infrastructure. The market expected 13.2 billion, exceeding by more than 5 billion. More profit, more burn. The Starlink business is a cash cow, with 4.29 billion in revenue and users increasing to 12 million, but ARPU dropped from $99 two years ago to $66. More users, lower unit price, behind the scale is intensified competition.
Two days before the earnings report, the stock price had already fallen to 108, halving from the high of 225. Short positions piled up to 34%, betting $24.6 billion, more than Tesla. Even more intense, on August 6, 911 million shares will be unlocked, with over $100 billion worth of chips flooding the market. Now both bulls and bears are waiting, each betting the other will collapse first. Some institutions say the company is not worth this high, and the AI business might be priced at zero. Bloomberg analysts say the market actually can't see clearly and doesn't know what valuation to give.
A company, earnings beat expectations, stock price falls. Made money, all burned.
#OKX星球话题来啦
#财报观察员：业绩喜忧参半，解禁将至！SpaceX后续怎么看？
#SpaceX首份财报超预期，解禁仍是关键变量 ETH Community Snapshot: Speed 2.43x, Short Window Shows Slightly Bullish Bias
ETH popularity should be viewed in two halves: one is how many people are talking, the other is the direction of the conversation.
OKX Onchain OS recorded 53 mentions of ETH in one hour at 06:00 on August 5 (China time) in the official snapshot, including 43 on X and 10 in news; totaling 524 mentions in 24 hours.
The latest one-hour speed is 2.43 times the 24-hour hourly average, meaning about 143% higher than the 24-hour hourly average, overall classified as "significantly accelerated." This describes the attention rhythm but cannot replace price, volume, or capital flow data.
In terms of sentiment, one hour shows 30% bullish, 21% bearish, and about 49% neutral, so currently "slightly bullish dominant." The 24-hour corresponding ratio is 34% bullish, 18% bearish; whether the short window is diverging from the long window is more meaningful than looking at one percentage alone.
What I care most about here is actually the denominator: only 53 mentions. A few concentrated discussions can significantly rewrite the proportions; retweets, quotes, and news restatements may all be about the same event. Bullish or bearish can be reported as is, but should not be casually translated as how much capital has established positions in the same direction.
Currently, ETH's source structure is "mainly driven by X." If X mentions increase first and news remains low, it looks like the community is spreading first; if news also increases simultaneously, it only means more verifiable material is available, and details still need to be confirmed from original announcements by foundations, protocols, regulators, or exchanges.
The 24-hour source background is 421 mentions on X and 103 in news. Comparing this with the one-hour 43 and 10 mentions shows whether the new round of discussion has switched communication channels. #EarningsRealityCheck #SpaceXBeatEstimates #AMDBeatsButDrops In the first three steps of the game, your mnemonic phrase has already been sifted out by a weak random number sieve—this isn't speculation on the edge of the chessboard, but the first ruthless first step of the Coldcard loophole in a real game.
The subsequent moves were far more terrifying than the 594 BTC that was swept away that night. Galaxy Research Director Alex Thorn marked the number of attackers as at least 15, but what I saw was not fifteen skirmishers, but a full line of mobile forces lurking along the open line. A victim clue with less than 1 BTC leads to 126 addresses and about 12 BTC stolen paths. This is the classic chess battle of pawns: you think the baron being cut off is already miserable, but the real fatal force is those small soldiers quietly advancing along the edge. By the time you come to your senses, they have already stood behind the baseline.
Last week's heavy blow worth hundreds of millions of dollars was like a direct blow piercing the king's wing. But today's increments reveal an even longer nightmare—the attack surface is split into fifteen lines, each move like players who don't communicate with each other, yet each holds the same deciphered game manual. This shows that the value of the loophole has been fully understood by the entire community; anyone can execute a rear wing discard piece on your weak slot in front of the king.
Charles Guillemet steers the topic into the age of artificial intelligence, which in my view represents a fundamental change in the type of chess player. In the past, we guarded against living human chess players; now, we face machines that extract patterns from every low-entropy move. It doesn't deliver flashy combo kills, but rather acts like a tireless endgame arsenal, calculating every possible move in advance, then using the most ordinary yet toughest moves to capture every pawn.
The hardware wallet industry's subsequent random number re-audits, firmware construction, and mnemonic generation processes are essentially the forced post-redemption phase in the middle game. The old styles are no longer reliable; every step must be re-validated. I have seen too many self-proclaimed fortresses die in a short, untested transposition.
Every random demotion is like a long castling by a player on the verge of time control—on the surface, the king hides in a corner, but in reality, the entire rear wing pawn is pinned down by the opponent's manual move. The grandmaster doesn't ask what was captured in that move, but what remains in the game after twenty moves. A true master doesn't judge by the loudness of a single move, but by the retreat deep within the formation that will never be blocked. The most dangerous word on the board is "Almost"—weak random numbers mean personally telling your opponent: "My king's wing's gate is open, please go ahead."
Weak random numbers are never a disaster; they are a surrender program you wrote before the start of the game.$LLY 💊LLY Eli Lilly｜GLP-1 money printer + China AI pharmaceutical paying the bill, one chart to see through the industry chain before the 8/5 earnings report
📅 Key date: Q2 earnings report before market on 2026/8/5
Consensus revenue $20.26B, EPS $6.71; 2026 full-year EPS consensus revised down from $36.59 to $34.91
Last quarter (Q1) revenue beat, EPS surprise +21.1%, four consecutive quarters exceeding expectations
🔴 Upstream｜China AI pharmaceutical becomes the largest external payer
🧠 Insilico Medicine × Eli Lilly: $2.75B AI drug discovery (Pharma.AI-driven multi-target)
🧬 Jingtai Technology × Eli Lilly: $345M bispecific + AI antibody platform licensing
Logic: Eli Lilly outsources "early discovery" to Chinese AI platforms, saving time but not money, with upfront payments + milestone-linked pipelines
🟡 Core｜Tirzepatide dual brand drives revenue
Mounjaro (glucose lowering) Q2 consensus $8.94B
Zepbound (weight loss) Q2 consensus $4.72B
💊 Foundayo (oral orforglipron) just launched in April, first sales disclosed this quarter, determining valuation ceiling
Tirzepatide in China has obtained 3 indications (T2DM/weight loss/OSA sleep apnea)
🔵 Downstream｜Mainly US + China volume growth + 40+ countries launched
US consumes over 60% of global GLP-1 demand, China volume grows post-National Reimbursement Drug List negotiation regardless of price
📊 Order amount ranking (disclosed BD)
Tirzepatide endogenous sales » Insilico Medicine $2.75B > Jingtai $345M
🎯 Institutional perspective
Consensus anchor: around $1160 = BASE scenario line (growth slowdown + Novo Nordisk pricing counterattack)
Earnings day volatility amplified: beat targets $1250+, miss targets $1050 pullback
Risks: annual drug price cuts 10-15%, oral GLP-1 capacity ramp-up below expectations, PE still above 40x+
LLY is not a "weight loss drug company," it uses China AI pharmaceutical to accelerate R&D speed, uses Tirzepatide dual brand to capture global metabolic disease dividends, and uses oral Foundayo to take the next baton as the industry chain hub. On 8/5, focus is on: whether Zepbound accelerates quarter-over-quarter, how much Foundayo sells in its first quarter, and full-year guidance movements. Recently, many coins have experienced significant surges. Many of these have previously surged, such as $BICO $HOME. Just now, I saw $BIO on the gainers list. In my impression, this coin experienced a very significant surge in April. So now I'm wondering if this coin will surge again. At present, I don't see very clear data to support my view. So, for now, I don't want to go in there. —————————————————— Let's look at its contract data. It's clear that in the early stages of $BIO's rise, there was indeed capital going long. But when the price rises significantly, a lot of short-selling funds increase in the market. Let's look at its longer-term data. Observing the three stages of increase in open interest further confirms my point. In the early stages of the rise, some were long, but after the price increased, a large number of short positions were introduced. I think there is a hint that it's about to start now, but there are still very clear signals. Let's compare $BIO's previous upward phase with its current rally. It can be seen that in the early stages of the previous upward phase, it was a type of slow oscillation. If this time is similar to last time, the current gains will gradually fall back. My current idea is to wait for a pullback and then consider whether to go long. ————————————————在零下十五度的冰原潜伏据点贴地趴了整整四十六个小时，呼出的热气在热成像瞄准镜的目镜上凝结成薄霜。
极度安静的夜色里，最危险的往往不是枪声，而是远方弹道轨迹发生的微小偏转。SpaceX 这头体量庞大的深空重兽，在公开亮出第一份体温测试表时，打出了一发让全场死寂的满环——二季度营收飙升至78.1亿美元，同比暴涨92%，营运亏损直接从9.7亿美元死死绞杀到1.43亿美元。在我的测距仪里，这不是普通的账面反弹，这是重装推进器在克服引力前最后一段窒息般的修正。原本以为这头猎物还在因巨额研发消耗而持续失血，但数据表明，它的止血带已经扎得比谁都紧，弹药库的补给速度远超所有暗处哨兵的预估。
真正让我扣紧防弹手套的，是瞄准镜边缘划过的那道绿光。它们与 NVIDIA 敲定了 Starmind 智算一号卫星计算载荷的联合架构。把星载高算力装进低轨卫星，就等于给原本只能垂直俯冲的重型弹头，装上了能够实时修正弹道的微型矢量喷头。深空轨道算力网的成型，让整个战场的规则从地面拉扯瞬间上升为轨道的绝对制空权。
但经验丰富的猎手从来不会被猎物表面的光芒蒙蔽双眼。在距离目标1200码的侧风修正盘上，8月6日那个刻度已经被我用红漆死死标记。那是第一批解禁锁仓期的窗口，最高达20%的限售股解禁许可，就像战壕里积攒了许久的散装弹药箱被瞬间打开。那些最早潜伏在土坑里的早期获利盘，随时可能在解禁声响起的同时，倾泻出遮天蔽日的抛售火力。横风会在那天达到峰值，获利盘的可持续性不是靠几句空洞的口号支撑的，暴风雪来临前的回撤压力，足够把盲目冲锋的菜鸟撕成碎片。
此刻在美股衍生战术板上，$XAVGO 的信号波形正沿着 SpaceX 的推进脉冲剧烈震荡。菜鸟看到92%的暴涨数据就会血脉宬张地按压扳机，而真正的射手只会把呼吸节奏降到最低，在风速仪的微弱抖动中，静静观察那20%抛压弹雨究竟会落在哪个弹坑。
在横风未定、第一波解禁流弹横飞的修罗场里，盲目暴露位置就是对职业纪律的背叛。The first thing I did in the morning was to check the $SPCX's movement.
In short: the bearish mindset remains unchanged, continuing to look for short-selling opportunities at high levels.
The previous rise was actually not surprising. The news from Ma Lao and the heavy short profit-taking led to a short squeeze, quickly pushing prices to around 130. This rebound indeed exceeded many people's expectations.
But the key is still to see sustainability.
After the earnings report was released, market sentiment cooled, and the stock price fell back to around 120 or even lower.
Many people think that a good financial report is a big positive, but don't forget, the market has never looked at revenue and stories alone; it also includes cash flow, investment costs, and future expectations.
If the positive news is realized and funds choose to withdraw, it means "the news lands, the market ends."
At present, 130 has become a short-term strong resistance level. If even the 120 level cannot be effectively held, it will become increasingly difficult to break upward in the future.
In addition, there is still a large amount of pressure from unlocked shares.
What does it mean for the current scale to enter the market with over 900 million potential circulating tokens?
With increased supply, capital pressure intensifies, making it difficult for stock prices to remain high.
Looking at this trend, it has risen from around 104 to 130, a significant gain in a short period.
Many low-priced chips have already been cashed out during the rise, and those who remain chasing the highs have a rising cost. Once market sentiment weakens and the buying power becomes insufficient, adjustment pressure will naturally be released.
The bullish sentiment in the market remains strong, but the real problem is:
How much new funding is still willing to take on?
So my personal judgment is:
If funds cannot continue to support after the release of unlocking pressure, it is highly likely that the market will continue to seek support in the short term. Areas around 100 or even lower levels require close attention.
Of course, the market never operates in a single track.
I didn't wait for a drop to analyze later; from the moment it fell below 110, my bearish views have been openly expressed. This rebound was also expected, and even the high levels actually provided better room for positioning.
Next, let's leave it to market validation.
Patiently wait and see who will have the last laugh in this bullish and bearish battle.
#SpaceX首份财报超预期, unlocking remains a key variable