Orbit Post Sitemap

##AI巨头债券利差飙升：投资风险还是抄底良机 The more it falls, the more you buy; what you're really buying isn't cheapness, but pushing yourself into a corner. Have you noticed that some orders pile up thicker and thicker, yet the price becomes lighter and lighter? Last night, I checked the order books of SNDK and MU, and the feeling was very subtle—the buy orders densely stacked upward during the downtrend, while the sell orders were thin, like catching a knife that hasn't fallen yet. This isn't bottom fishing; it's the position deciding for someone. First, look at the facts: these two stocks have weakened continuously, but the buy order volume is several times that of the sell orders. Every time the price steps down, more people add positions, repeatedly lowering the cost basis, and the stop-loss levels keep dropping. It sounds like discipline, but it's actually giving up the retreat bit by bit. What is the market trading? Not value, but "I don't want to admit it." When one side keeps taking the goods and the other keeps selling, the price can only slide in the direction of least resistance. More importantly, the unrealized losses after adding positions don't grow linearly; every further drop multiplies the losses—the heavier the position, the thinner the margin for error. From the trend perspective, this looks more like a divergence in a downtrend continuation, not a start or a healthy continuation. In a real start phase, volume and price cooperate; here, the more the buy orders pile up, the lower the price goes—a typical aftermath of distribution. Bulls think they're diluting their cost, but they're actually handing chips to the bears. Looking at the bigger picture: after tech earnings, the Nasdaq weekly chart still fell about 3.4%, and risk appetite hasn't truly recovered. In this one-sided decline, BTC and ETH's correlation leans more toward sentiment—not active capital inflow, but passive technical rebounds after falling. Altcoins are even more sensitive, aOn the surface, it looks like Elon Musk is finally putting the brakes on the Cybertruck, but what's really loosening underneath is something else. Guess what the market is panicking about? Tesla is reallocating the Texas factory's capacity to Cybercab and plans to gradually discontinue the Cybertruck by mid-2027. This news sounds like a straightforward stop-loss. After all, the original plan was to sell 250,000 units annually, but in 2025, only about 20,000 were actually sold in the U.S., and the gap is clear. Yet, the stock price first dropped 10.77%, then bounced back 4.2% the next day, closing near $358. This back-and-forth itself shows that people aren't voting on the model but repricing the narrative.✨ What I'm watching isn't the car, but the emotional temperature within the derivatives structure. The Q3 earnings report showed revenue of about $28 billion, but operating profit dropped 56% year-over-year. The profit pressure on the automotive business is out in the open, and autonomous driving still requires continuous investment. So, the market's pricing of Tesla increasingly resembles trading a forward story rather than current cash flow. In this state, the options market is prone to two-way bets: some buy calls chasing the narrative, others buy protection against data disappointments. Implied volatility is being pulled back and forth, with FOMO and hesitation sitting at the same table. The bullish path is clear: cut slow-selling models and shift resources to Cybercab, which is an early admission of mistakes and a pivot. If subsequent urban operation data gradually materializes, risk appetite will return, and high-beta assets like BTC and ETH will likely follow the tech narrative sentiment for a while. The altcoin side is even more sensitive, as long as the autonomous driving topic... $IMX Gaming adoption requires more than blockchain integration. Players care about gameplay, ownership, accessibility, and entertainment before they care about the underlying technology. Immutable focuses on blockchain infrastructure for gaming, including digital asset ownership and game development tools. Its biggest challenge is helping games attract genuine audiences rather than relying primarily on token incentives and speculative interest.BTC and ETH stuck in a volatile squeeze: chips are lightened, but bulls are still holding on hard $BTC is repeatedly tugging between $75,000 and $78,000, while $ETH is consolidating narrowly around $2,200–$2,300. In October, both fell by 7.3% and 16.0% respectively, with market breadth sharply contracting. Price is just the surface; the core issue lies in the capital structure: · Capital is withdrawing: Both BTC and ETH spot ETFs saw net outflows, with ETH alone experiencing a $1.6 billion monthly outflow, indicating institutional demand is shifting to tactical contraction. · Leverage is reduced, but sentiment remains firm: Open interest across the market dropped to a 12-month low, leverage has basically cleared out, yet perpetual funding rates remain positive—chips have been washed out, but bulls unwilling to concede are still paying to hold on. · The path likely dips before diverging: 10x Research believes BTC could still reach $90,000 by year-end, but only after first dropping to $77,000–$79,000 to complete the final shakeout; ETH is likely weaker than BTC due to value continuously flowing to Layer 2, resulting in insufficient returns for holders. The key signal for a trend reversal is when funding rates turn negative, which would truly mean bulls have conceded and a bottom structure is established. Before that, rebounds are more likely repairs rather than reversals. Coupled with the US October non-farm payrolls exceeding expectations and rising bets on December rate hikes, short-term sentiment remains suppressed—fasten your seatbelt.#SoftBank's Heavy Bet on OpenAI: Can the Q2 Earnings Report Deliver Impressive Results? ​ Positive Highlights (Likely to Create "Impressive Paper Gains") - OpenAI's valuation remains stable or is adjusted upward → resulting in substantial unrealized investment gains, boosting the group's net profit and net asset value (NAV) significantly; ​ - Arm's resilient performance and stable dividends serve as SoftBank's cash flow ballast, offsetting the high-risk AI investments; ​ - The Japanese joint venture SB OAI Japan is deploying OpenAI services for enterprises, opening local B2B commercialization opportunities, representing a new growth point in SoftBank's own operations. Key Risks (Market's Main Concerns and Earnings Report Deductions) - OpenAI is still burning huge amounts of cash, with revenue growth but even higher computing costs, sustaining large operating losses, no stable cash returns, and equity that cannot be liquidated at will—entirely paper wealth; ​ - SoftBank has pledged OpenAI equity to secure margin loans in the tens of billions, with additional collateral clauses; any downward adjustment in OpenAI's valuation would directly impact SoftBank's net worth and trigger debt risks. CDS spreads in the bond market have widened, and rating agencies are worried about exposure risks; ​ - Intensifying competition: Anthropic, Google, and others continue to catch up, with large model price wars squeezing gross margins; meanwhile, OpenAI's complex governance structure and uncertain IPO timing mean the exit window is far off; ​ - Funding pressure: substantial future investment payments still need to be made in batches, continuously draining SoftBank's cash and increasing capital expenditure pressure #软银重金押注OpenAI，Q2财报能否交出亮眼成绩？ Core Risks (Most Market Concerns and Deduction Points in the Financial Report) - OpenAI is still burning huge amounts of cash, revenue is growing but computing power costs are even higher, resulting in sustained large operating losses, no stable cash returns, equity cannot be liquidated at any time, all just paper wealth; - SoftBank has pledged OpenAI equity to obtain a hundred-billion-level margin loan, with additional collateral clauses; if OpenAI's valuation is downgraded, it will directly impact SoftBank's net worth and trigger debt risks, CDS in the bond market has already widened, rating agencies are worried about exposure risks; - Intensified competition: Anthropic, Google, and others continue to catch up, large model price wars squeeze gross margins; meanwhile, OpenAI's governance structure is complex, IPO timing is uncertain, exit window is far off; - Funding pressure: there are still large amounts of investment funds to be delivered in batches, continuously consuming SoftBank's cash, with high capital expenditure pressure. . Financial Report Outlook - Book profit: may look very good, provided OpenAI's current valuation remains unchanged or is raised; but profit quality is weak, it is fair value change gains, not main business cash flow; - Conditions for exceeding expectations: Arm's profit exceeds expectations, management provides a clear OpenAI exit timetable, Stargate project progress is clear; - Triggers for underperformance: OpenAI valuation downgrade, management admits expanded additional investment scale, leverage/liquidity guidance turns conservative Don't get excited yet: On November 5th, the U.S. Department of Labor sent the final rule "allowing 401(k) retirement plans to buy crypto assets" to the White House for final review. This is not a positive event being realized, but the start of the pipeline. ① How big is the pool? ICI data: U.S. 401(k) assets total $10.8 trillion, total U.S. retirement assets $51.2 trillion (Q2 2026). ② How slow is the pipeline? Experts estimate: White House final review takes up to 90 days, then about 60 days after publication to take effect, so a few options may appear in 2027, and widespread adoption in 2028; ETFs and funds will come first, not direct coin holdings. ③ Who can get in? The DOL's March proposal set six criteria: performance, fees, liquidity, valuation, benchmark, complexity. Lawyers put it bluntly: $BTC clearly passes, $ETH passes, others fail on performance records or benchmarks. The key question is "can you get a reliable daily quote from a regulated wrapper?" So for small coins, this is a pseudo-positive — money will concentrate on the two or three assets with the best liquidity. Reality check: In the same week, U.S. spot Bitcoin ETFs saw a net outflow of $1.28 billion, the worst since June, turning negative for the first time in 2026; $BTC current price $75,800, down from $81,199 this week. $ETH spot ETFs had a net outflow of $193 million over two days. Policy is a slow variable, not a buy signal. This $10.8 trillion, do you see it as incremental buying power or just sentiment? #BTC现货ETF年内累计净流出约4.16亿美元 The US Bitcoin spot ETF has seen a cumulative net outflow of about $416 million this year, but this is a slight net negative after offsetting inflows and outflows, and should not be simply interpreted as a full institutional withdrawal. According to Farside data as of November 5, the total net inflow days this year amount to about $25.63 billion, while total net outflow days amount to about $26.05 billion, with a gap of only $416 million. The marginal direction is more concerning: on November 6 alone, there was a $412 million outflow; over the past five trading days, the total outflow was about $1.282 billion; the cumulative net outflow in November is about $1.28 billion, with only one trading day of net inflow so far this month. The recent outflow speed and concentration clearly deviate from the annual average. After a hotter-than-expected October nonfarm payroll, the December rate hike bet rose to about 80%, and US Treasury real yields increased, suppressing high-risk assets with zero cash flow; after the midterm elections, expectations for crypto-friendly policies cooled, prompting some funds to reduce exposure. Institutional differentiation is obvious: BlackRock's IBIT is the main source of recent outflows, with a net outflow of $1.209 billion in a single week in November, and about $403 million redeemed on November 6 alone; Fidelity's FBTC saw a counter-trend inflow of about $42.8 million, but it was insufficient to offset outflows. Price and funds form a negative feedback loop: Bitcoin fell below $75,000, ETF redemptions added extra selling pressure, and over $1.27 billion in crypto derivatives positions were liquidated simultaneously, with longs accounting for over $1.1 billion. The turning signal should be whether the ETF can return to sustained net inflows. $BTC L3 and AppChains once excited developers, claiming to give each application its own dedicated lane. But the reality is: the cars are built, but there are no cars on the road, and the funds are all fragmented into pieces. Comparing these three projects: $DYDX: The pioneer migrating from L2 to Cosmos AppChain. It gained highly customized performance at the cost of losing aggregated liquidity from the Ethereum mainnet. Due to a lack of deep market makers, trading volume has been heavily eroded by competitors like Hyperliquid, and the 75% buyback mechanism is useless due to revenue exhaustion. $ZORA: An L2/L3 focused on the creator economy. It attracted many NFT and content creators, but on-chain assets are highly homogeneous. Lacking real DeFi lending depth, relying solely on internal PVP trading cannot accumulate large external funds, making protocol revenue extremely unstable. $ARB: The king of L2. Why is it successful? Because it has a "liquidity network effect." Although its Orbit L3 chains are continuously launched, the underlying assets and settlements remain anchored on the Arbitrum mainnet, allowing funds to form a closed loop. Conclusion: DYDX is "highly customized but liquidity exhausted"; ZORA is "locally prosperous but hard to accumulate large funds"; ARB is "underlying aggregation with strong network effects." The biggest trap of L3 is using technical "sovereignty" to exchange for liquidity "islands." #Anthropic still plans November IPO, investors lower valuation expectations Anthropic's plan for a November IPO has not been withdrawn, but control over valuation has clearly shifted to investors. At the beginning of October, the market once valued it at about $2 trillion, dropping to $1.4 trillion to $1.5 trillion by mid-October. After this week's tech stock sell-off, potential investors have further lowered the discussion range to $1.2 trillion to $1.3 trillion, with some even seeing around $1 trillion. The company originally planned to start roadshows the week of November 9 and price the IPO the following week, but the probability of completing the IPO before November 30 on Kalshi has dropped from 84% to 60%. The valuation downgrade is driven by both macro and company-specific factors: the Nasdaq 100 plunged 6.4% in a single day, AI assets were heavily sold off, and market concerns about data center prospects and the return on massive AI expenditures have intensified; the prospectus shows revenue of $16.286 billion in the first half of 2026, GAAP net loss of $80.748 billion mainly due to non-cash items such as fair value changes of convertible notes caused by valuation increases, operating loss narrowed to $693 million, with a single quarter of operating profit in Q2, but a fixed cost pressure from a more than $430 billion compute procurement commitment over the next decade. Insiders say that if AI assets continue to be under pressure, the company may delay the issuance; Morningstar believes accepting a slightly lower pricing to complete the listing may be more pragmatic. Essentially, the IPO window is still open, but its width is not determined solely by Anthropic. #马斯克预计SpaceX年底年化营收突破1000亿美元 $xSPCX Q4 revenue is expected to grow about 70% quarter-over-quarter to nearly $25 billion, achieving an annualized revenue exceeding $100 billion by December, with the statement "Unless a natural disaster occurs, we will reach that goal." He also set the 2027 year-end annualized revenue target at over $500 billion, considering $1 trillion "not impossible" with about a 2% probability. After the announcement, SpaceX's stock closed up 8.62% on Friday at $134.27, still below the $135 IPO price in June. The source of growth has undergone a structural shift. Q3 revenue was $14.58 billion, up 161% year-over-year, equivalent to an annualized approximately $58 billion; among this, AI business revenue was $7.56 billion, soaring 830% year-over-year, accounting for 52% of total revenue, surpassing Starlink's 36% and launch business's 12% for the first time. Government contracts' share dropped from about 20% in 2025 to 16.6% in Q3, with Musk expecting it to fall below 10% this quarter. In other words, the $100 billion annualized revenue is a short-term verifiable milestone, while the $500 billion essentially remains a management vision rather than a market pricing benchmark. The focus going forward should not be on the target number itself, but whether Q4 actual revenue can reach near $25 billion, and whether the renewal pace of AI computing power contracts can support the current growth rate.The non-farm impact has been fully digested, and the crypto market has entered a phase of low-volume tug-of-war. The US added 165,000 non-farm jobs in October, significantly exceeding market expectations. The unemployment rate fell simultaneously, and wage growth picked up, indicating that the labor market remains resilient. After the data release, interest rate expectations were quickly revised, the dollar and US Treasury yields strengthened, and the crypto market first experienced a round of leverage liquidation. Long positions were concentratedly cleared in the short term, and macro negative factors followed. Currently, sentiment has clearly cooled, mainstream coins' volatility has narrowed, and bullish and bearish forces are temporarily balanced, waiting for new directional signals. BTC is quoted at 75,922, with limited intraday amplitude, roughly trading between 74,800 and 76,400 over 24 hours. Resistance is at 76,172, support at 75,496, Bollinger Bands continue to contract, and volatility is suppressed. ETH is currently around 2,241, repeatedly tugging around the middle band, with support at 2,221 and resistance at 2,244, showing no strong short-term trend. OKB is moving sideways with the broader market, currently priced at 105.5, above the short-term moving average, ranging between 103.26 and 107.36, performing relatively steadily. The impact of macro news has temporarily paused, and the market has entered a news vacuum period. Lacking incremental funds to drive it, the market is more likely to clean leverage through upper and lower spikes, forming a grinding consolidation. In the short term, it is advisable to watch more and trade less, patiently waiting for an effective breakout before following.Apollo is spending $75 billion on AI chips — the AI buildout is now private-equity scale. 🟠 BTC at $75,845, holding $75,000. Support: $75,000. Resistance: $76,371, then $79,000. Chips are the new oil — and the AI trade keeps pulling risk assets along. My take: AI infra is the macro tide; BTC holds $75K until CPI tells it where to go. Is AI spend the next bubble, or the next backbone? 👇 #Bitcoin #BTC #AI #TechStocks #OKXOrbit#SoftBank's Heavy Bet on OpenAI: Can Q2 Earnings Deliver Impressive Results? SoftBank has invested approximately $64.6 billion in OpenAI, holding about 13% equity, making it one of the largest single bets currently. This year, it completed three additional investments totaling $30 billion, with the final $10 billion tranche settled on October 1, partially funded by high-yield bonds with coupons close to 9.75%—this is not a small gamble but a bet with the balance sheet at stake. The Q2 earnings report will be released on November 10. The market's main focus is not on traditional business but on how the fair value of OpenAI's equity is recognized, whether the LTV (loan-to-value ratio) remains controllable, and the structure of cash and debt. Previous quarters saw huge investment gains due to OpenAI's valuation surge, but recently the AI sector has cooled, OpenAI's financing targets have been lowered, and its valuation has retreated from higher expectations to around $1.25 trillion. Can the paper gains continue? Masayoshi Son's style has always been "all in on the future." From Yahoo, Alibaba, to Arm, and now the full chain of OpenAI + computing power + robotics, he is betting on dominance in the AGI era. But this time the leverage is higher, the term longer, and the exit path more uncertain—OpenAI is not yet public, and data center and power investments have long cycles. Under high-interest financing, if valuations stagnate or decline, interest and market value fluctuations will squeeze simultaneously. If Q2 delivers impressive results (continued upward revision of OpenAI's fair value, strong contribution from Arm), it could temporarily ease credit concerns. $OPENAI #OKXTraderVoices If you are starting your journey in the world of cryptocurrencies today, remember this rule: never invest money you cannot afford to lose.$BTC It sounds simple, but emotions drive people to break this rule every day. When the market moves fast, panic takes over. The best way to endure and succeed in the crypto world is to build your investment portfolio gradually. Start by looking at established assets with high liquidity, instead of randomly chasing trending tokens. Tell me in the comments: what is your number one rule for risk management 比特币核心开发者Cory Fields提出了一种新颖的机制来威慑量子计算机对比特币的攻击：“量子地雷”方案。该方案要求未来试图利用量子计算机窃取比特币的攻击者，必须先抵押等量的比特币作为担保。原持有者可事先设置隐藏的所有权证明，并在资金锁定期（如30天）内使用该证明收回自己的比特币并没收窃贼的抵押品。攻击者在付出资金前，无法知道哪些持仓里埋了“地雷”。 该方案仅适用于仍持有私钥的所有者，不适用于已丢失的比特币。针对不活跃钱包，Fields还设想了一套拍卖系统：竞争者可提前在脆弱钱包上质押隐藏出价，若遭量子窃贼攻击，最高出价者的智能合约会自动触发，销毁窃贼的目标币，没收其抵押品并作为赏金转给最高出价者。 行业分析与影响 该提案凸显了加密货币行业对比特币椭圆曲线密码学（ECDSA）长期安全性的日益担忧。尽管实际的量子攻击可能还需数年甚至数十年才能实现，但理论上的风险已促使研究人员探索主动防御措施。Fields的方案在博弈论上颇为巧妙：通过让盗窃行为在经济上自杀，旨在将攻击扼杀在发生之前。 然而，该提案也面临复杂的实施挑战。它很可能需要通过软分叉或硬分叉来修改比特币协议，而隐藏的所有权证明🚨 BREAKING: Bitcoin open interest has reached the exact levels where MAJOR SHORT SQUEEZES have historically occurred. The last 3 times this happened, $BTC saw massive pumps. #USJobsBeatExpectations #软银重金押注OpenAI，Q2财报能否交出亮眼成绩？ SoftBank Group will release its FY2026 Q2 financial report on November 10, with the market focusing on the financial results of its AI sector bets. Through Vision Fund II, SoftBank has completed an additional $30 billion investment in OpenAI's third round, with a total cumulative investment of $64.6 billion, corresponding to about a 13% stake, making it one of OpenAI's most important investors. The core highlights of this financial report focus on the Vision Fund's investment gains and losses, changes in the fair value of underlying assets, and the group's financing pressure. To complete this huge capital outlay, SoftBank arranged financing tools such as foreign currency senior notes, and the large capital investment will be directly reflected in liabilities and cash flow data. Fluctuations in OpenAI's valuation will directly affect the Vision Fund's book gains or impairments, impacting SoftBank's overall net asset performance. The high volatility of AI investments is the biggest variable. If OpenAI's valuation continues to rise, it will continuously increase SoftBank's investment income; if the AI sector valuation corrects, the huge holdings may bring asset impairment risks. The market will use this financial report to judge the sustainability of SoftBank's continued AI funding and the ability of Masayoshi Son's AI investment strategy to deliver. $BTC $ETH $SNDK A rebound to 82,000 does not mean the market has reversed; what really matters is whether it can hold at that level. @张教主。 This judgment on $BTC is not blindly bullish, but rather to first acquire spot positions, then wait for the price to reach a critical point before distinguishing whether this is just a rebound or the start of a stronger trend. He repeatedly reminds: even if the direction is correct, contract positions may not survive until the payoff day. Separate the weekend sideways movement from the subsequent trend. Facing doubts about the rebound's weakness, he does not define a day's fluctuation as a failure but places the next confirmation on Monday: whether the current minor high can break through, and if it can hold after breaking through, then continue upward. Only with such strong performance will he raise expectations for the subsequent market. Until it breaks out, hope cannot replace confirmation. On a smaller scale, he keeps a scenario that makes contract bulls uncomfortable: the price may first drop to sweep liquidity again, then recover and continue to oscillate within the range. The previous false breakdown does not guarantee it won't happen again. For spot holders, this may just be volatility during holding; for those with stop losses near the lows and heavy positions, the same movement might first clear them out before continuing in the originally anticipated direction. When he mentions the stop loss near 74,300, the focus is not on deciding whether viewers should move it, but to remind that level may be swept again. This is not a reason to cancel stop losses, nor to bear losses indefinitely. On the contrary, the issue is whether this possibility was factored into the risk before opening the position.The market's last bullish hope did not survive the payrolls print. US jobs data came in far above expectations, reviving the case for tighter policy, and the dollar and Treasury yields rose together. $BTC could not even defend its psychological levels, sliding to 74,877 before stabilising near 75,700. That is the tell. Good macro data should have been a relief; instead it hardened rate expectations and squeezed the most rate-sensitive assets. Crypto is trading as a liquidity barometer, not a groTrump has been in business for decades, long handling issues of interest exchange, negotiation, and risk. He cannot be unaware of the toughness of the Iranian Revolutionary Guard, nor can he be unaware that Iran is not a country that would easily surrender after a few rounds of bombing. Currently, Trump has deployed the U.S. Army's elite 82nd Airborne Division, the U.S. Marine Corps' 31st Expeditionary Unit, the U.S. Navy SEALs, and Army Rangers to the Middle East, establishing a multi-layered military deployment including the Air Force, Navy (three carrier strike groups), Marine Corps (two amphibious landing groups), airborne troops, and special operations forces, totaling 60,000 troops! Based on the current U.S. military deployment in the Middle East and the reinforcements being sent there, it is still insufficient to overthrow the Iranian regime, but it can completely shatter Iran's economic lifeline! Oil price risks will persist for a long time, and the U.S. stock market will continue to be under pressure! If an energy shock and AI valuation adjustment happen simultaneously, that would be even more interesting!$BZ #SoftBank's Heavy Bet on OpenAI: Can Q2 Earnings Deliver Impressive Results? In a nutshell SoftBank's Q2 performance heavily depends on the fair value appreciation of its OpenAI equity; the book profit has a good chance to be strong but is unrealized and not cash; Arm is the stable foundation, while the biggest risk lies in OpenAI's ongoing massive losses, large investment exposure, liquidity, and leverage risks, making it difficult to deliver solid, sustainable impressive results. Key Points 1. Bet Overview SoftBank has committed a total investment of about $64.6 billion in OpenAI, targeting approximately 13% ownership, making it the second-largest shareholder after Microsoft; this investment is held in Vision Fund 2, and is accounted for at the latest financing round valuation ($500 billion), with valuation changes directly determining SoftBank's current investment income and being the biggest variable in the financial report. At the same time, SoftBank co-develops the Stargate computing power project with OpenAI and Oracle, representing a heavy-asset, long-term AI layout. 2. Positive Highlights (likely to create "impressive book" results) - Stable or upward OpenAI valuation → brings large unrealized investment gains, boosting group net profit and net asset value (NAV) significantly; - Arm's resilient performance and stable dividends serve as SoftBank's cash flow ballast, hedging high-risk AI investments; - The Japanese joint venture SB OAI Japan delivers OpenAI services to enterprises, opening local B2B commercialization leads, representing a new growth point in SoftBank's own operations. 3. Core Risks (market's main concern and financial report's deduction points) - OpenAI is still burning huge cash, revenueRocket Lab ($RKLB) Q3 Earnings Forecast: I choose YES, leaning towards revenue beating expectations. There are three main reasons: ① High certainty of growth in launch business. Q3 Electron completed 6 orbital launches, while Q2 had only 4 similar launches with revenue mainly recognized in the same quarter. Launch business revenue is expected to grow from $44.6 million in Q2 to about $59.8 million. ② Continuous expansion of satellite manufacturing business. Q2 space systems business revenue reached $189.5 million. The acquisitions of Mynaric and Motiv contributed full quarter revenue in Q3, combined with ongoing satellite manufacturing order deliveries, expected to push this business beyond $200 million. ③ Backlog provides support. At the end of Q2, backlog reached $2.36 billion, a 137% year-over-year increase. The company expects to recognize about 45.5% of this revenue in the next 12 months, providing strong support for future growth. Currently, the market expects Q3 revenue around $257 million, with company guidance between $250 million and $265 million. I personally expect actual revenue near $260 million, slightly exceeding market expectations. However, Q3 gross margin guidance is only 29% to 31%, with R&D investment still high, posing a risk of EPS falling short of expectations. Overall judgment: choose YES, mainly optimistic about revenue beating expectations, confidence about 60%. A flat BTC is not a calm market. It is a deaf index. Three free checks: 1) Funding: SOL 0.0046%, BTC 0.0067% - spot sells, not leverage. 2) ETH/BTC 0.0295, ETH the only green cap: rotation, not exit. 3) Turnover per 1%: ZEC $7.5M, PEPE $2.8M - 14x less money. My read: rotation. I'd flip if funding passes 0.05%. Rotation or distribution - which leg? Crypto is high-risk. Personal observation only, not investment advice. DYOR. $SOL $BTC $PEPE #Solana #OKXOrbit #Crypto #AltcoinsWhy is HYPE Treasury Company on Nasdaq applying for a US futures brokerage license? Hyperliquid Strategies' subsidiary HypeStrat Markets is applying for an FCM license through the NFA/CFTC route: if approved, it can handle HIP-3 licensed perpetual accounts for US clients—the blockchain remains a neutral infrastructure, and compliance is layered on the account side. The application was submitted around August and is still pending; the company itself stated "the timing and outcome are uncertain." A note from the quarterly report: revenue is about $13.3 million, with the majority of net profit still coming from HYPE unrealized gains—the application form and the ledger are two different things. Confidence level: medium-high | company disclosure + media reports. Application ≠ approval, and certainly does not mean US users can open positions tomorrow. Are you more focused on the license progress or on whether anyone is actually deploying HIP-3? $BTC #RocketLab财报前瞻：Neutron火箭进展如何？ #RocketLab财报前瞻：Neutron火箭进展如何？ $RKLB is getting closer to its post-market earnings report on November 9, and the feeling of holding your breath waiting for the landing is truly a unique experience for aerospace stocks 🚀 The topic views have surged to 404,800, and the attention is visibly rising, but market sentiment remains calm, even leaning pessimistic: only 27% of voters believe the earnings will beat expectations, while a full 73% expect it to miss; trading volume has reached 391,500, with intense long-short battles, yet the stock price is barely moving, up just +0.07%, almost flat. After following RKLB for a while, you get this "tug-of-war without movement" feeling. Electron launch business is a steady foundation, with revenue guidance pinned at 250–265 million, which everyone understands, making it hard to deliver big surprises. What everyone truly cares about is Neutron. Trading this stock has never been about current revenue, but betting on whether that reusable rocket can be delivered on time in the future. The most frustrating experience is this: obviously everyone is discussing it and watching it, but the price remains stagnant. Everyone is holding their breath waiting for the earnings report moment. As soon as any good news about Neutron comes out, the calm market can instantly explode; but if progress is delayed, the already pessimistic market sentiment immediately weighs down the price. $BTC Market Analysis Conclusion: Currently, there is insufficient strong rebound signal. The recovery after the main control line break is not straightforward. There are two response strategies for the weekend! Scenario 1: Stabilize above 76360 in the next two days, bullish towards 77330-78420 (30 min / 1H MA250). Scenario 2: Weak oscillation over the weekend fails to break through 76350, or dips to new lows at 73348-72680 (daily MA250) before rising again.ETH and SOL are also flowing out; is market risk spreading? This time, it's not only BTC ETFs showing negative numbers; ETH and SOL products have not turned positive either. In the past two trading days, BTC spot ETFs have had a combined net outflow of about $842.1 million. Let's first look at $ETH: around $2,243 at 12:00 UTC on November 7, 24h +0.03%; perpetual open interest across three exchanges is about $8.6 billion. The restrained price reminds us not to just listen to macro slogans. Next, $SOL: about $89.93, 24h -0.47%; aggregated open interest daily change -5.2%. High elasticity has not been fully released. Another comparison is $BTC: about $75,917, 24h -0.21%; perpetual open interest across three exchanges is about $15.2 billion. The market's support still needs spot trading confirmation. Finally, putting them together, the cross-asset simultaneous outflow indicates institutional channels are cautious, but price elasticity and capital scale differ among coins. If you could only wait for one signal, would you want BTC funds to stop bleeding, or all three types of funds to turn positive together? #BTC现货ETF年内累计净流出约4.16亿美元 #美国10月非农超预期，12月加息押注升温 #美股暴跌后科技股反弹，纳指周线仍跌约3.4% In Q3 2026, $TRX processed 1.08 billion transactions, a year-on-year increase of 31.19%. TRON's transaction volume in the third quarter continued to maintain historic highs. • Total transaction volume in Q3 2026: 1.08 billion transactions, up 31.19% year-on-year • Average processing speed this quarter reached 135.76 TPS • The fourth quarter is expected to maintain this record, with 364.4 million transactions in the past 30 days. The steady growth in transaction volume indicates effective utilization of the TRON network, especially in the stablecoin transfer sector. However, an increase in transaction volume does not necessarily mean a rise in TRX price, as Tron’s recent weekly purchases of TRX have been worth less than $50,000, indicating quite limited funding demand. I believe the TRX price is likely to remain sideways, following the overall market trend. I only see this as a positive signal for the platform itself, but I do not dare to assert that this will be a driver for price increases. People wonder whether such a large user network is enough to push up the price of TRX, or if its price will still depend on BTC.“AI meme” has once again been pushed to the front by ACT, and I almost thought the entire Solana sentiment chain was heating up. But when I flipped to PUMP, I realized I was imagining things too quickly. Let's first look at the prices. BTC is almost flat, SOL dropped 0.79%, while ACT rose 14.86%, with a volatility of 25.59%. However, it is still about 7.46% away from its high point, so this baton isn't held very steadily. Next, let's look at the funds. ACT's trading volume is about 1.74 million USDT, PUMP's is about 6.33 million USDT, but the latter dropped 10.15%. More active trading doesn't mean stronger prices; even within the same ecosystem chain, they didn't move forward together. Finally, the funds connect to that term: AI meme. For now, I lean towards it being "short-term noise," not because I find the term unattractive, but because ACT's performance alone isn't enough for me to treat this local anomaly as a narrative comeback. My brain that chases hot terms shouldn't rush to add drama. Do you stand by "the keyword is still passing the baton," or "it's just short-term noise"? Leaving aside ACT's recent gains, how many more rounds of discussion can the term "AI meme" sustain? #OKX星球话题来啦 $ACT $PUMP Today I did a long-term return comparison between $XAU and $QQQ, and the results are really quite interesting! Both are types of assets, but just different purchase times, and the final results differ so much! The first chart shows holding QQQ continuously from its inception in 1999 until now: Gold has accumulated an increase of about 1321%, QQQ about 1211%, surprisingly close, with gold slightly ahead! The second chart, if unlucky, buying right at the 2000 internet bubble peak: Gold rose about 1313%, but QQQ only 472%. This time gold really left QQQ far behind! The third chart, changing the time, buying from the top of gold’s previous bull market in 2011: QQQ rose about 1135%, but gold only about 120%! This time it’s completely reversed, QQQ left gold far behind. After seeing these three charts, my biggest takeaway is: sometimes what you buy is important, but when you buy is equally important! Buying even the best assets at high valuations may require a long time to digest the bubble. Conversely, buying during market downturns can lead to completely different long-term returns. Of course, these are just price appreciation comparisons on the charts, without uniformly accounting for dividend reinvestment and other factors. Both gold and QQQ have their own long-term investment value and have moments that make investors question their decisions.$XAUT is a gold token issued by Tether, with 1 token = 1 ounce of physical gold; OKX's gold perpetual is called $XAU. This week, gold has been the most stable asset overall. What happened: The election results came in, the Republicans lost Congress, and the market bought into a "divided government" — the US dollar fell, yields dropped, and gold surged 2.8% on Wednesday alone, jumping from 3,945 straight to 4,056; on Thursday, it further climbed to 4,152, fueled by safe-haven demand amid the US stock market crash. $BTC struggled around 75,000 during the same period, with only gold fully playing the safe-haven role. The cost was borne by the shorts: this week, short liquidations totaled about $66 million, while longs only saw $3 million; those who tried to top out were forced out. But leverage has mostly been burned off: positions have dropped by over 30%, funding rates are near zero, and now spot prices are setting the tone. Currently, COMEX is closed for the weekend, and OKX's 4,077 price is just a shadow price, with less than a million traded in an hour, so don't interpret weekend fluctuations as signals. The key verification point is Monday: how COMEX opens relative to 4,077 and where it closes in the first hour. Holding above 4,152 points to 4,200; if it first dips, 4,028 and 4,000 are the touchstones for this breakout—breaking below them would indicate the election rally is set to retrace.#BTC现货ETF年内累计净流出约4.16亿美元 $BTC The current market is basically a half-dead stalemate. The selling pressure has indeed been mostly exhausted, the short momentum is fading, but the big money off-exchange is completely dormant, with no sign of buying activity at all. Deleveraging is basically over, but the macro environment and institutional funds remain bearish, with no confidence to push the market up. The strength contrast is very clear, $ETH is now a soft target, much weaker than BTC, and there is a cascading liquidation point hanging overhead. If that triggers, it will definitely cause a chain stampede. Looking at structural data, the 30-day range is between 74,888 and 84,898, and the current price is stuck rubbing the lower boundary intensely. The most fatal signal is extremely low volume; Binance’s ten high-liquidity samples only traded just over 2.2 billion, shrinking nearly 40%; Ethereum is even worse, shrinking over 40%. What does this data mean? There are indeed fewer stop-loss cuts, but no one dares to step in to catch the falling knife. Based on the current framework, the overall direction is still bearish, in a downtrend cycle. At times like this, total position size should definitely not exceed 30%, single trade risk strictly capped at 0.3%, and absolutely no adding long positions against the trend. Low-volume slow decline is the most exhausting; don’t guess the bottom blindly, wait for right-side signals before making a move. Are you currently sitting out watching, or holding half your position and toughing it out? #美国10月非农超预期，12月加息押注升温 #美股暴跌后科技股反弹，纳指周线仍跌约3.4% #英伟达冲击千亿美元季度营收纪录 $NVDA Nvidia's earnings exceed expectations, but political discounts are rewriting valuation anchors Nvidia's FY2027 Q3 revenue guidance midpoint is $108 billion, about 2.7% higher than Bloomberg's consensus estimate of $105.155 billion, and does not yet include China data center revenue, so the revenue side is indeed above expectations However, the Non-GAAP gross margin guidance drops to 74%, with Q4 expected to further slide to a low range of 71%–72%. The signal of "revenue growth without profit growth" has already appeared, which is the key reason the market is not buying it Where is Nvidia's bottom? On November 5, it plunged 9.26% to close at $181.60, the lowest since April 2026, retreating more than 25% from the October high of $243.37 Technical support is around $175–180, while the 52-week low points to $160–165 The midterm elections have ended, with a big win for the Democrats. After the midterms, the Democrats took control of both houses, and senior officials quickly sent signals targeting data center construction and AI expansion Texas, a key state accounting for 27% of the planned data center capacity in the US, also elected a Democratic governor, and Nvidia fell 8% that day Morgan Stanley's optimistic scenario is $350, corresponding to about 22x PE in 2027, assuming regulation remains at the statement level, Rubin's ramp-up goes smoothly, and gross margin stabilizes The pessimistic scenario is $160, corresponding to about 13x PE, roughly coinciding with the technical low of $165. Note the risks Liquidation data cools off overnight: from 1.256 billion to 355 million 24-hour total network liquidation dropped to $355 million, with 99,000 people liquidated; the previous day was $1.256 billion and 238,000 people, shrinking by about two-thirds. The largest single liquidation was on Binance ETH, $2.87 million, significantly reduced from yesterday's $10.93 million. Structural shift observed. The 24-hour window still mainly long positions, accounting for 73.2%; but the short-term trend has changed: in the last hour, liquidations totaled $5.74 million, with shorts at $4.7 million; in 4 hours, liquidations reached $13.55 million, shorts at $9.15 million. BTC rebounded near 74,762; after breaking below 75,000 yesterday, short sellers were liquidated on the rebound. Coin differentiation. BTC liquidations at $108 million, longs accounting for 81.4%, still mainly long position clearings; ETH liquidations at $60.23 million. ZEC and SOXS are in the red, shorts liquidated, small coins rebounded simultaneously. Total volume dropped from 1.2 billion to 350 million, indicating the most intense long liquidation phase has passed. However, short-term short liquidations seem more like a corrective rebound due to thin weekend liquidity, so a trend reversal cannot yet be confirmed. "Volatility is not the bottom, it's chips holding firm (￣▽￣;)" BTC is repeatedly tugging between $75,000 and $78,000, ETH is narrowly consolidating between $2,200 and $2,300; in October, both fell by 7.3% and 16% respectively, with market breadth clearly shrinking. On the surface, prices are sideways, but in reality, the capital structure is conflicted: BTC and ETH spot ETFs are both seeing net outflows, ETH lost $1.6 billion in a single month, and institutional demand is shifting toward tactical contraction (・_・;) More subtly, the total open interest in the market has hit a 12-month low, leverage is basically cleared out, yet perpetual funding rates remain positive—longs are being washed out on one side, but still paying a "bullish fee" on the other. Chips are leaving, but stubbornness remains (｡•́︿•̀｡) Going forward, it looks more like "probing the bottom first, then diverging," rather than flooring the gas pedal for a breakout. 10x Research believes BTC still has a chance to reach $90,000 by year-end, but the premise is first to test $77,000–$79,000 for a final shakeout; ETH is likely weaker than BTC because value continues to flow to Layer 2, and holders' returns feel insufficient (￣▽￣)ゞ The real turning signal is not a single rebound bullish candle, but when funding rates turn negative: longs admit defeat, and the bottom structure is established. Before that, rebounds are mostly repairs, not reversals. Don't mistake a bounce for a reversal ヾ(≧▽≦*)o $BTC $ETH $ZEC #美国10月非农超预期，12月加息押注升温 BTC / ETH Today's Analysis Both are in a state of "selling pressure exhaustion but no buying interest yet," with macro and institutional funds still bearish, and derivatives deleveraging basically completed. According to your v4.8.5 framework, the outlook is bearish with 65% confidence, and the cycle is positioned in a downtrend—corresponding to a total position cap of 30%, single trade risk of 0.3%, and prohibition of counter-trend long additions. ETH is clearly weaker than BTC, with a cascading liquidation point looming overhead. Structural Assessment: The 30-day range is 74,888–84,898, currently hugging the lower boundary. The contraction in trading volume is the most critical information—Binance's ten high-liquidity spot samples combined trading volume is 2.211 billion USDT, down 37.23% week-over-week; BTC volume is 1.304 billion (−35.60%), ETH 470 million (−44.31%); eight out of ten samples are below their own 30-day median. The volume contraction to this extent during the decline indeed indicates selling pressure exhaustion, but there is no evidence of any new buying interest. ETH long positions are approaching the liquidation line. Is Brother Maji a qualified contract trader?❗️ Brother Maji's ETH long positions are in danger again: liquidation price at 2,203, current price around 2,240, less than 2% away from liquidation. In the past 4 days, he sold 15 NFTs and lost $500,000, only managing to raise $223,000 to add margin, which was burned through in just a few days. Papertrade shows a small loss of $909,000, and Hyperliquid accumulates a huge loss of $37.4 million. Once a whale holding 150 Bored Apes, now he survives by selling monkeys, then continues to open new positions. My answer is very clear: he is not❗️ He looks more like a gambler with whale capital than a contract trader. Looking at his operations, the problems are consistent: First, almost exclusively long positions. ETH, HYPE, PUMP, BTC are all longs. The crypto cycle is volatile; betting only on the upside is like handing your fate over to liquidity. Second, no active stop-loss. With 25x leverage, entry at 2,263, liquidation at 2,227, less than $40 in between. His stop-loss is forced liquidation. Third, he relies on margin top-ups to endure losses. Normal traders reduce positions and admit mistakes; he chooses to sell NFTs to survive. He once profited $44.84 million, now he has lost it all and is down $37.4 million. The core of a qualified contract trader is not always being right, but surviving when wrong. What Brother Maji lacks is not direction, but position management, stop-loss discipline, and switching between long and short. This is not a trader's liquidation; it's a wealthy person proving in the most expensive way that no matter how big the capital, it can't withstand gambler-style position openings #软银重金押注OpenAI，Q2财报能否交出亮眼成绩？ SoftBank's Q2 financial report indeed looks impressive on paper, with profits largely coming from the unrealized gains in OpenAI equity valuation, representing "paper wealth" rather than stable operating cash flow. The heavy bet on OpenAI is a high-risk gamble: in the short term, it relies on valuation increases to beautify the financial statements, but it requires continuous large investments afterward. Coupled with OpenAI's slow profit realization, intensified industry competition, and high valuation volatility, once OpenAI's valuation corrects, the performance will sharply retract. The financial report can only be considered a temporary report card and does not mean this investment is safely cashed out; its sustainability is questionable. $BTC 重回 75,000 美元上方，$ETH 却滑到 2,226，$SOL 死守 90 一线——这三个数字摆在一起，本身就是一道裂缝。更耐人寻味的是，盘口后台的买入声浪正变得越来越密集，而价格却没有给出对应的爆发力。这种"呼声高、动作虚"的组合，往往比暴跌更值得警惕。 我不急着站队，理由不在价格图上，而在一个更诚实的指标：资金费率。 真正的超卖反转长什么样？费率被砸进深度负值，空头挤得水泄不通，只有在这种持仓结构下，反弹才有逼空燃料。可眼下费率依然轻描淡写地微微为正——既没有恐慌，也没有投降。这意味着市场根本没有经历一次像样的清洗，多头没被赶下车，空头也没堆积到可以被收割的程度。 换句话说，当前更像是一段缩量下跌途中的喘息，而不是底部的画像。$BTC 站上 75,000 并不等于结构修复，$ETH 跌到 2,226 也不等于恐慌抛售，$SOL 守住 90 更像是惯性而非承接。三者走势分化，恰恰说明资金没有形成统一方向，这种分歧本身就是弱信号。 我的核心判断可以浓缩成一句话：没有负费率做燃料的反弹，只是噪音，不是趋势。买入从来不是比谁手快，而是比谁更能等。燃料没加满，我宁愿站在场外看Average transaction price behind the top spot For ONE at $0.002258, I divide the 24-hour trading volume by the transaction amount, resulting in an average price of about $0.001934. Not everyone made a profit. This is not the holding cost. Some sold early, some bought late; I leave the average price in the review section. $ONE #美国10月非农超预期，12月加息押注升温 After the US FinCEN issued a special measures order against MEXC Global, the funding channels responded first. Some users reported that when withdrawing from MEXC to platforms like OKX, Binance, Gate, etc., deposits were intercepted or delayed. The special measures target "primary money laundering concerns" institutions, meaning US financial institutions and compliant exchanges need to conduct extra scrutiny on funds related to MEXC, or even cut off directly. The judgment is somewhat negative for MEXC and its platform token MX. The key issue is not the fines, but the funding channels: once mainstream exchanges collectively tighten deposit risk controls, MEXC's deposit and withdrawal efficiency and user experience will be impacted first, easily causing withdrawal queues and asset discounts. Two scenarios to observe: if it is just isolated community cases, the main effect is emotional disturbance; if mainstream platforms publicly restrict, the pressure on MX and the MEXC ecosystem will be more obvious. Are you more concerned about whether MEXC withdrawals are experiencing queues, limits, or delays in arrival, or whether OKX, Binance, etc. have issued official announcements? Source: Wu Shuo"USDC is lively, but that doesn't mean $BTC is taking off" Circle's Q3 revenue was about $865 million, with on-chain USDC transaction volume soaring about 49% year-over-year, showing the stablecoin sector is indeed bustling. But over the weekend, $BTC was still hovering around 75,980, and the capital flow is even more striking: the US spot BTC ETF saw a net outflow of about $412 million on November 6, with BlackRock's IBIT alone withdrawing about $403 million, marking the second consecutive day of net outflow. The 24-hour range was roughly 74,890 to 76,380, with funding rates nearly zero. The rise in USDC volume indicates there is no shortage of USD-denominated liquidity on exchanges; however, with ETFs still pulling real money out, the price has to hold on its own. With this kind of volume over the weekend, I won't chase strength. Short-term, treat it as a range: support is at 75,500–74,890, resistance at 76,380; if it can't hold, don't rush to chase; if it breaks 74,890, reduce positions and observe, with a target of 76,800. Better to keep positions small and wait for liquidity to return on Monday before discussing trends. Don't mistake the stablecoin activity for a signal that BTC is about to take off; first watch how the resistance levels behave. $BTC $ETH #CircleQ3营收超预期，USDC链上交易量增长49% #美国10月非农超预期，12月加息押注升温 #美股暴跌后科技股反弹，纳指周线仍跌约3.4% This is only a personal opinion and does not constitute investment advice.🔥 I no longer want to predict BTC's lowest point because the trades that truly make money don't necessarily have to buy at the lowest price! 📉 The four-year bull and bear cycle has indeed shown some patterns in BTC's past. From 2023 to 2025, BTC's performance relatively aligns with the cycle rhythm, but altcoins have diverged early, with many coins reaching their phase highs in the first half of 2024. 💥 Next, BTC has three possibilities: breaking below $57,800 to set a new historical low; returning near $60,000 for a second bottom before ending the bear market; or directly reversing in the current area to start a new round of rally. I won't rule out any of these three scenarios. 🧠 I don't need to guess the lowest point; I just need to wait for the price to offer a good enough opportunity. Without a margin of safety, I won't buy; without clear trading conditions, I will stay out of the market. 💰 Serious opportunities are worth seeking only after a big drop, but a deeper drop doesn't necessarily mean it's the bottom. We also need to see if selling pressure is exhausted, if the price stabilizes, and if the market truly has capital support. 🔥 Tonight, patience is still key. Missing out is not scary; the scariest thing is buying at the riskiest point just to avoid missing out. Are you willing to miss the rally in order to wait for a margin of safety? #美国10月非农超预期，12月加息押注升温 🔥In this BTC bear market cycle, the truly worthwhile opportunity to wait for may not have appeared yet! 📉 According to the historical pattern of the four-year cycle, Bitcoin's bull and bear transitions usually follow a certain rhythm. From 2023 to 2025, Bitcoin's cycle performance is relatively clear, but altcoins have not fully synchronized; many altcoins actually peaked in the first half of 2024. 💥 What the market is most concerned about now is where BTC will complete its final drop? Currently, I mainly consider three possibilities: first, breaking below the historical low of $57,800 to create a new low; second, returning to just above $60,000 for a second dip before ending the bear market; third, the current area is the bottom, and the market directly makes a big V-shaped reversal. ⚠️ All three scenarios are possible, and no one can guarantee in advance which one will definitely happen. The four-year cycle can be used as a reference but should not be treated as an exact price prediction. 💰 So my strategy is simple: don't buy if it doesn't drop, only act on a big drop. Without enough safety margin, even if the price rises before your eyes, do not chase. 🧠 Tonight, I will continue to wait patiently. Do you think $57,800 will be the last line of defense, or will this cycle see a deeper correction? #美国10月非农超预期，12月加息押注升温 The Nasdaq just caught its breath, but AI coins showed two different moods On Friday, the Nasdaq rebounded 1.1%, but it still fell 3.4% for the whole week. This kind of market easily misleads people into thinking that tech risk appetite has returned, but the AI sector in crypto is not so uniform. $RENDER was about $1.587 at 09:10 UTC on November 7, up 3.41% in 24 hours; perpetual open interest on seven exchanges was about $36.63 million, up 4.95% daily. Price rising with open interest increasing at least indicates rising participation, but new positions are not necessarily all longs. $NEAR was about $3.709 at 10:15 UTC in the Hyperliquid snapshot, down 0.80% in 24 hours, with open interest around $159.5 million and a 1-hour rate of +0.0004%. Funds are still present, but the price hasn’t kept up with the tech stock rebound. RENDER is more active, NEAR is temporarily lagging; the two data sources have different standards, so only compare direction, not absolute open interest to judge strength. If tech stocks rise again next week, will you chase the leading RENDER, or keep an eye on the temporarily lagging NEAR? #美股暴跌后科技股反弹，纳指周线仍跌约3.4% Don't mistake institutional entry snapshots for your holding confidence I once regarded ETFs as "stabilizers": since Wall Street is involved, the market should be more resilient. But capital has both entry and exit points; inflow and outflow always go through the same door. On November 5, BTC spot ETFs had a net outflow of about $430 million, ETH spot ETFs had a net outflow of about $99.9 million, totaling approximately $530 million. More strikingly, after just returning the day before, the next day they turned around and exited again. This shows institutions are not "anchors" that only buy and never sell; they also rebalance, take profits, and hedge. This made me rethink "institutional entry": having bought doesn't mean they will hold forever; recognizing the market doesn't mean they agree with your purchase price. Institutions focus on allocation and timing, while you bear your own cost basis. These are not the same. In the future, when you see "institutions bottom-fishing," first check three things: the date of the data, whether outflows continue, and if they keep adding positions afterward. Don't take a single entry snapshot as a long-term commitment. They may have already swiped their card and left, while you're still taking photos excitedly at the door.Short sellers also need to guard against rebounds When storage stocks keep weakening, short sellers tend to mistake the decline for a tailwind. The more it falls, the bolder they get adding positions; the more it rises, the more uneasy they become. For stocks like $SNDK, what really troubles shorts isn’t the gradual decline, but the sharp rally triggered suddenly by bottom-fishing funds. Once the price rebounds, the floating losses on short positions get magnified. Leveraged positions become even more passive; when margin tightens, the system stops you out before you can even analyze the structure. Shorts fear not continued drops, but a sudden bullish rebound candle cutting into the downtrend. Therefore, stop-loss isn’t admitting defeat, it’s leaving yourself a fallback. Proactively reducing positions before a rebound may feel painful, but it’s much cheaper than holding on until the peak and getting forcibly liquidated. After cutting positions, you can wait for the next opportunity; stubbornly holding on might cost you both your principal and control. Short selling isn’t about who can endure more, but who can survive longer.NVIDIA's revenue potential in the Chinese market indeed exists But there is one data point worth noting: NVIDIA disclosed in its quarterly filing ending July 26, 2026, that shipments of the licensed H200 represent only a portion of the approved quantity, and the related shipment revenue accounts for less than 1% of the data center revenue for that quarter This indicates that there is still a long way to go between policy approval and actual commercial revenue Therefore, one cannot directly infer that NVIDIA is about to receive tens of billions of dollars in new orders just because the number of approved companies has increased I prefer to understand it as a long-term market access opportunity rather than a short-term performance explosion catalyst For China's AI industry: short-term to supplement computing power, long-term to compete for an independent ecosystem Here is a contradiction: Chinese AI companies need NVIDIA's high-performance GPUs, but China also needs to reduce dependence on advanced American chips If the H200 is widely introduced in China, it may improve the supply of computing power for some model training and alleviate enterprises' demand for high-performance computing resources But if the NVIDIA ecosystem continues to be used long-term, it may also increase dependence on CUDA, software tools, and the supply chain Therefore, what China's AI chip industry truly needs to break through is not just single-chip performance, but: Chip manufacturing capability High-bandwidth memory supply High-speed interconnect Cluster computing efficiency Software and developer ecosystem This is a competition from chips to the entire computing infrastructure