Orbit Post Sitemap
AI把换芯片的活干了，英伟达的墙就矮了
一份合作论文里，$AMD 的芯片跑出了接近对手的吞吐。
这个数怎么算的：十项推荐模型，同卡数、同批量。
MI350X 综合训练吞吐是 B200 的 101.2%。
倒推一下，差距只剩不到两个点。
容易误读的地方：换芯片难，不在硬件。
难在客户多年攒下的代码要重写、报错要重调。
这些活过去只能靠工程师一项项磨。
现在 AI 反复编译、读报错、自动补功能。
迁移成本一降，客户就敢引第二家供应商。
议价的手，也就从一边挪到了两边。
真正该盯的不是跑分，是迁移工时的报价。
#芯片股八连跌后反弹，AI交易能否修复？
#$AMD $NVDA #谷歌云业务增长82%，AI投入回报如何？ Google's winning move in this earnings report isn't about whether cloud revenue can maintain high growth, but that it is using free cash flow turning negative to bet on a variable that could rewrite valuation logic: TPU external sales.
Looking at the numbers first, cloud revenue surged, backlog orders hit a high, and profit margins are improving. But there's a hidden risk: capital expenditures doubled in Q2, directly wiping out operating cash flow, turning free cash flow negative. This burn rate would have driven most companies out of the market.
Why is the market still patient? Because Google is doing something traditional cloud providers haven't: selling its self-developed TPU hardware to external AI companies. Traditional cloud rents computing power with a profit margin ceiling; selling hardware is a one-time settlement, though it requires massive upfront capacity investment, which is the fundamental reason for negative cash flow. The business hasn't worsened; the model is shifting from renting to selling equipment.
This is the biggest expectation gap. The market is still pricing it as a cloud company, but if TPU external sales really take off, Google's valuation system will be re-evaluated as an AI hardware vendor, breaking the valuation multiple ceiling.
Don't bet on whether cloud revenue beats expectations. Focus on the earnings call to see if management dares to present specific data or guidance on TPU external sales. If they reveal it, even if cash flow looks bad, the capital will be re-priced; if they only use cloud growth to mask the burn, this earnings report is shaky. $GOOGL $xGOOGL - 枚 BTC、浮盈接近 44%——这个数字摆出来的时候，我盯着屏幕愣了两秒。 你每天画线盯盘忙进忙出，最后跑赢了一个"每天只买 1 枚"的国家？ 先把事实放稳：El Salvador 手里大约 7800 枚 BTC，均价约 55000，按现在 81700 附近算，账面浮盈接近四成。更关键的是节奏——从 2024 年 3 月起每天固定买 1 枚，持续 900 多天，IMF 不让动国家资金，他们就绕去个人捐款继续买。这不是择时，这是把买入写进了日历。 我真正在意的不是他们赚了多少，而是这件事对当下盘面的暗示。大盘在 81700 附近磨蹭，情绪偏冷、多数人账户难看，可这类"无脑定投式"持仓反而活得很好，说明现在缺的不是方向，是耐心和成本优势。衍生品这边更能说明问题：震荡区间里资金费率容易反复翻正翻负，持仓却迟迟不肯大幅撤退，多空都挤在窄区间里试探，这种结构一旦被单边触发，挤压会比趋势本身更猛。 板块强弱上也能看出偏好：BTC 作为"国家储备叙事"的锚更抗跌，ETH 跟涨但弹性一般，山寨则明显分化，只有少数有真实买盘支撑的标的能站稳，其余靠情绪脉冲。偏多的一路径是：只要 80000 上方守住On October 19, it was reported that last week, global publicly listed companies (excluding mining companies) net purchased $360 million worth of BTC, with Strive spending about $330 million, contributing over 90%. The actual buying pressure is somewhat bullish for spot demand, but since the funds are highly concentrated in one company, whether the buying pressure will spread remains to be seen. Next, attention will be on whether Strive can continue to increase its holdings and whether other listed companies will follow suit, to determine the sustainability of this round of institutional demand.
$BTC #机构资金转弱，BTC与ETH谁先迎来回流？ ONDO 来回波动这么大，小幅涨幅会不会误导盘面判断🤔
不少人只看 24 小时涨跌幅，容易忽略盘中震荡区间，这类滚动统计口径以连续滑动 24 小时窗口计算，和固定自然日统计逻辑并不相同。
1、欧易现货 ONDO/USDT 最新成交价 0.4341 USDT，24 小时涨幅 1.35%，区间最高 0.4426 USDT、最低 0.4113 USDT。
这里有个容易踩坑的点：净涨跌只对比起点和终点，高低区间代表这段行情走过的范围。就算收尾变化不大，中间也能走出明显波动，区间的上下沿，不能直接当成有效支撑阻力，该判断仅适用于当前统计时段，无法直接外推到后续行情。
当下成交额 3894769.4813552 USDT，短期盘面后续能不能守住区间，还需要持续观察。
#ONDO来回波动这么大涨跌幅有没有把你骗了 #麻吉今日加仓约302万美元ETH $ONDO 8 个交易日，SMH 从 639 跌到 540，留下 4 个缺口，一个都没补。
10 月 6 日它摸到 639.47，是 60 日最高。
第二天直接低开 11 个点，那一段空档到今天还空着。
之后又跳了三次。
一次从 601 跳到 598，一次从 563 跳到 560，一次只有 0.5 点。
四次加起来 11.6 点。
这一百点的跌幅里，一成多是跳过去的。
跳过去的意思是，那几档价格根本没有成交过。
没有换手，也没有人在这里认亏。
量在配合。
9 月 28 日到 10 月 6 日日均 5.3M 股，10 月 7 日之后日均 10.5M 股，快翻倍。
放量加跳空，说明是有人真想走，不是没人接盘。
我觉得这比均线有用。
头顶第一道坎是 560.6，就是 10 月 15 日那个缺口的下沿。
反弹到那里，面对的不是套牢盘，是一段空白。
再往上是 588、598、625，一档一段空白。
一个板块 ETF 走出四级台阶，卖的就不是某一家公司的故事了。
你更愿意等它补第一个缺口，还是等它回到 60 日低点？
$SMH
#四级台阶Strategy's recent operations show that fund allocation explains the situation better than the number of coins purchased.
As of October 18, the company holds 848,517 BTC and disclosed about $5.3 billion in USD assets.
In the same week, it spent approximately $193.1 million to repurchase STRC, but only about $22.1 million to buy 267 BTC. The repurchase amount is nearly 9 times the expenditure on buying coins.
This means Strategy is not only expanding its Bitcoin position but also actively adjusting its financing structure.
STRC is the company's issued perpetual preferred stock. Repurchasing this type of security can reduce future dividend burdens; however, the cost is a reduction in cash and liquidity impact.
Moreover, the funding sources for the two transactions differ: funds for buying BTC mainly come from selling MSTR common stock, while repurchasing STRC uses the company's cash holdings.
This is a more complex capital management approach: increasing BTC holdings through common stock financing while using cash to repurchase preferred stock.
The company also raised the authorized limit for the digital credit securities repurchase program from $2 billion to $3 billion. However, the authorized amount does not equal actual expenditure. The latest disclosure shows about $1.16 billion remains available after the adjustment.
What needs to be observed next is not just how much more BTC it can buy, but whether cash reserves, dividend payments, and financing capacity can remain coordinated. ETH is being sandwiched: ETFs withdrew $955 million in October, but institutional buying is increasing.
Here's what I see: ETH is currently at $2,394 (intraday 10/19, Binance), with the 20-day moving average at 2545 and the 50-day moving average at 2551 pressing tightly overhead; spot ETFs have had almost daily net outflows in October, totaling $955 million. On the other hand, Bitmine just bought another 12,535 coins, bringing holdings to 6.04 million coins (about $14.6 billion), just 1% short of the "taking over 5% of Ethereum's supply" target; JPMorgan announced the launch of JPM Coin deposit tokens on Ethereum the same day.
My view: The funds withdrawing are short-term money, while the accumulation is by long-term investors. The price is stuck between support at 2378 and resistance at 2552, fluctuating back and forth, making chasing rallies or bottom-fishing unprofitable.
My plan: Do not chase; wait for direction. A daily close above 2552 turns bullish targeting 2680; breaking below 2378 first targets 2318 (low on 10/15), then down to 2150 (200-day moving average).
Which side are you on: ETF outflows signaling a top first, or Bitmine reaching 5% supply first?
$ETH $BTC $SOL
#Ethereum #EarningsSeason Macro Outlook: Oil Prices, AI Financing Risks, and Bitcoin Under Pressure This week, the market remains heavily influenced by the US-Iran conflict, with oil prices emerging as one of the biggest concerns. In September, US gasoline prices surged 9.1% month-over-month, accounting for more than half of the monthly CPI increase. Although the market generally expects the Federal Reserve to hold off on raising interest rates in October, expectations of a potential rate hike in December have already st#HBM4量产能否推动SK海力士盈利再创新高？
Q2 profits have already increased by 557%, and Q3 aims to "set a new record," with the bar set by themselves.
▪️ Q3 earnings report to be released on 10/28, focus on shipments and prices
▪️ Q2 operating profit 60.5 trillion KRW, up 557% year-on-year, a quarterly record high
▪️ Management said in July that Q3 DRAM bit shipments would increase about 10% quarter-on-quarter, prioritizing servers
▪️ HBM4 will ramp up fully in the second half of the year, with 1c nanometer DRAM scaling up simultaneously
▪️ HBM4 base chips are outsourced to TSMC's 12nm process, costing 3 to 4 times more than their own 1b process
▪️ Today SK Hynix-related stocks +7.04%, SKHY +5.43%
The disagreement is not about whether HBM4 can be mass-produced, but that the bar for "setting a new record" was raised by Q2 itself.
+557% is largely due to last year's low base; the same profit next year will automatically show a smaller year-on-year increase.
HBM4 first involves cost factors: base chips are outsourced, unit price three to four times higher, and yield ramp-up still needs discounts.
The outlook is "high probability of Q3 record high, but year-on-year growth rate will definitely decline" — conditions: DRAM contract prices hold until November; risk: yield drags shipments.说实话，我以前也把“倍数不高”当成放松警惕的理由，觉得扛一扛总能等回来。后来吃过亏才明白，仓位太大时，嘴上说能承受，真遇到波动就容易改规矩。上班已经够累了，没必要再给自己安排一份盯盘夜班。
刚看到 Core Trader 07 把 $PYTH 的多仓全部平掉，杠杆3倍，平仓金额约16.38万美元，已实现亏损17,527.69美元。这单已经结束，不能拿它当还在押反弹的信号。
我的判断是，低倍数不能替代仓位管理。按素材里的开平仓均价算，价格反向走了大约9.67%，3倍下对应的保证金损耗大约29%，低倍也会疼。再一个，同样的波动，仓位越大，亏损绝对值越大，人就越容易舍不得认错。不过不知道他的账户总资金，也不能凭这一单就断言他重仓或没设止损。
我现在坚持两条：开仓前先定这单最多亏多少，再倒推仓位，不拿倍数当安全证明；退出条件提前写下，到线就执行，不因为不甘心临时放宽。
大家有没有因为“才几倍”就放松仓位限制的经历？愿不愿意分享一条后来真正执行下来的规矩？#俄美柴油合同推进，签约后15至20天或抵美 俄美柴油运输协议落地推进，这批柴油预计签约后15-20天运抵美国，直接缓解美国东海岸柴油供给压力。
美国冬季是柴油需求旺季，库存一直偏紧，此前市场担忧柴油短缺推高能源与通胀。这批俄柴油到港，短期压制柴油现货价格，间接降低成品油带来的通胀粘性。
但要注意两层约束：第一，并非大规模长期供应，属于阶段性补给，很难彻底改变美国柴油供需格局；第二，政策层面存在不确定性，后续仍会受到制裁条款、舆论影响，存在变数。
宏观传导上，柴油价格回落有助于缓和整体能源通胀，会轻微改善市场对美联储政策预期。不过影响偏短期，油价大方向依旧要看OPEC+减产以及中东地缘风险。
总结：短期利好柴油价格回落，小幅减轻通胀压力，但属于一次性补给，难以形成长期趋势性行情。
$BTC $ETH $ZEC 摩根大通把 JPM Coin 扩展到以太坊了。
表面看，这是银行支付业务增加了一条区块链通道；往深一点看，它涉及的是传统银行存款如何进入链上金融体系。
JPM Coin 此前已经在 Base 上运行，这次扩展到以太坊，意味着机构客户可以在更多链上环境中转移代币化美元。
为什么以太坊有意义？
因为这里已经聚集了大量代币化资产和金融应用。资产可以在链上交易，但如果资金仍需要频繁回到传统系统完成结算，整个流程就很难充分发挥区块链的效率。
而 JPM Coin 还有一个关键区别：它不是普通稳定币，而是代表摩根大通美元存款的银行存款代币。
同样是锚定美元，背后的机制却不一样。稳定币与银行存款代币在发行、赎回、法律关系和风险承担方面存在差异，不能简单地把它们视为同一种资产。
对机构而言，这类产品的潜在用途也不局限于转账，还包括代币化证券结算、抵押品交付和自动化资金管理。
如果链上的资产与结算资金能够更顺畅地配合，交易流程就有机会减少等待与资金占用。
当然，技术接入不等于所有机构都能自由使用，也不意味着传统银行体系的限制会自动消失。US-Russia diesel agreement, Iran's "7-day plan" and negotiations, short-term news pressure on oil prices, but why do I insist that oil prices won't fall?
#俄美柴油合同推进，签约后15至20天或抵美
US-Russia diesel agreement finalized, Russia resumes partial diesel exports, US temporarily relaxes sanctions;
Iran's 7-day ceasefire + asset unfreezing + reopening of the Hormuz Strait plan, although no substantial breakthrough, the window remains open;
Affected by these "positive" news shocks, $BZ price has retreated from recent highs,
but this does not mean oil prices will drop sharply.
My judgment is that the current situation is more like a "fake fall," the fundamental problems are not resolved, and the medium term remains relatively strong.
Basis:
The news is just an emotional shock. The flow through the Strait of Hormuz is still far below pre-war levels, some tracking data only about 40%–80%, Iran's crude oil exports are basically zero, attacks are frequent, and war risk premiums are high;
The Saudi-Houthi conflict drags down the Red Sea detour route, inventories have been significantly depleted, physical supply remains tight.
Negotiation conditions are harsh, both sides distrust each other, core risks remain unresolved.
Brent $BZ trend:
Short-term support is seen at $105–106, breaking below that targets $102–103; upper resistance at $108.5–110.
If negotiations make no substantial progress or attacks resume, it is highly likely to stabilize above $105 and then return to above $110;
Only when the Strait truly stabilizes and normalizes can a deeper correction space open up.Vitalik gave Ethereum a requirement: the worst-case scenario for proofs must be clearly calculated.
While others are competing on who can produce proofs faster,
he is asking—how much computational power does the most demanding block require.
First, some background.
The ZKsync team has achieved very fast proof generation speeds. On October 15, Vitalik publicly congratulated them.
Then he proposed three tasks to accomplish:
First, ensure the proofs align with Ethereum’s specifications.
Second, set an upper limit on the cost of proofs.
Third, make the proof system itself secure.
Why is the second point the hardest?
Ethereum uses gas to measure workload and limit resources per block.
But there has never been a way to account for how much computational power is needed to generate a proof.
So what he wants is a conversion rule: as gas increases, the proof should consume proportionally more computational power.
The tricky part lies in the phrase "worst-case scenario."
Ordinary blocks can be proven in seconds, which seems cheap.
But the rule must guard against someone deliberately submitting an especially hard-to-prove block.
And here lies a loophole.
EIP-7904 has already done calculations: at current CPU speeds, no re-pricing is needed.
But that analysis did not include the proof generation part.
So Vitalik’s worst-case test question is:
Given a machine with specified hardware and time, how much does it cost to process the hardest valid block?
No one knows the answer yet.
In short:
Ethereum’s ZK path doesn’t lack faster speeds; it lacks a guarantee of an upper bound to prevent runaway costs. Crypto Circle Financial Report: Raydium Q3 2026 Analysis Raydium remains one of the leading decentralized exchanges on Solana. Its Q3 2026 performance looks impressive at first glance, but a closer look reveals several underlying risks that investors should not ignore. Key Financial Highlights Trading Fees: Raydium generated approximately $57.1 million in Q3 trading fees, more than tripling from roughly $17 million in Q2. However, September alone accounted for around $44.6 million of the total. This morning, some friends came over to chat and asked if the AI bubble is really about to burst, and whether another crisis might happen. After looking into it, we actually feel that the market isn't that panicked this time. The 2008 situation was truly problematic, but now the issues aren't necessarily that severe. The Philadelphia Semiconductor Index's eight consecutive declines are indeed the result of many negative factors piling up. We reviewed these 8 days and summarized all the negative news below; indeed, there are various adverse factors. We've actually discussed the above points with everyone before: the numerator side reflects lowered market expectations, while the denominator side acts more like an amplifier, intensifying the market's pessimistic expectations on the numerator. Whether it's the US CPI or oil price fluctuations, the effect is similar. What everyone in tech is most worried about now is that AI cloud spending might burn through cash unsustainably. After reviewing the entire industry's status and finances, we feel it's not that dangerous. Amazon Cloud's CFO said in an interview last week that their AI servers can break even in less than three years. This was also mentioned by Amazon's CEO during the Q2 earnings call, who added that the servers can last at least 5 to 6 years. What does breaking even in three years mean? These machines are generally calculated on simple interest, with about one-third of the investment recouped each year, yielding roughly 33%. To be pessimistic, breaking even in the first three years and net profit in the last three years, assuming the equipment's residual value is zero, and calculating compound interest, the annual return is still about 24%, and even if the servers only last 5 years, it's still 20%. Being usable for 5 to 6 years is very likely reliable. Nvidia's A100, released in 2020, up toToday while browsing the news, I saw that Dogecoin's retail ecosystem has taken another step forward. Through the cooperation between MoonPay and Paxos, $DOGE has been integrated into over 6,000 merchants, and Paxos's transaction has directly placed DOGE onto the regulated channels behind PayPal and Venmo.
I stared at the words "PayPal and Venmo" for a long time.
In the past, when explaining to friends what Dogecoin could do, I could only say "it can be used for transfers." Now someone has connected the channel to the payment software used daily by tens of millions of people worldwide. Users might not even realize they're dealing with crypto; they just tap on Venmo, and DOGE is running underneath.
Some people point to the fact that there are only about ten thousand transfers on-chain daily and think the ecosystem is quiet. But I think about something else. A 2013 meme is now lying within PayPal's compliant channels. This fact itself is far more important than a few points of price increase today.
$DOGE rose 3% today, with the price still around 0.08. The path for payments is being paved step by step, and I'm not in a hurry to see the outcome.$BTC fluctuated between 80,622 and 83,157.9, rising only 0.99% in 24h, with a trading volume of 11.9 billion USD, which is not large. Central bank gold purchases are slow variables; for this to affect the crypto price, risk-hedging capital needs to overflow, which is not seen in current data. Contract open interest is 7.82 billion USD, recent three funding rates range from 0.0050% to 0.0100%, indicating bulls are not leveraging up. Options put/call volume is 1.24, higher than the open interest ratio of 0.86, with new orders buying protection; DVOL at 38.0 also shows no one is betting on large volatility. 24h long liquidations totaled 24.41 million USD, more than short liquidations at 17.4 million USD, meaning those chasing longs at the upper boundary were shaken out. My judgment is $BTC will continue to grind within the range short-term, leaning weak. Conditions to turn bullish are breaking above 83,157.9, with funding rates consistently above 0.0100%, and put/call volume returning near 0.86. Conditions to turn bearish are breaking below 80,622 with continued increase in long liquidations. Stablecoin total supply at 306.1 billion USD has not shrunk; money is on the sidelines but has not yet entered.WLD's ecological technology progress cannot stop quantum resistance destruction and high proportion of uncirculated tokens
Core summary: WLD's World ID and World Chain ecosystem are indeed continuously iterating, but two major inherent flaws cannot be resolved through ecosystem development: the token's underlying lack of native quantum resistance protection, and a large amount of tokens locked for a long time awaiting unlocking. These two risks are independent of the progress of ecosystem applications.
1. Real ecological technology progress of WLD
1. World ID (human identity proof): iris recognition + zero-knowledge proof ZK, covering a large number of global users, achieving "proof of being a real person without revealing personal identity information"; the project upgraded to SMPC multiparty computation, deleting original iris encoding to strengthen biometric privacy protection.
2. World Chain (OP Stack Ethereum L2): specially built for the real human identity network, World ID users enjoy priority block usage rights; continuously upgrading block lists, GKR proof system, opening ProveKit development tools to facilitate various applications to access human identity verification.
3. Application scenarios: wallets, cross-border transfers, DeFi applications, Visa co-branded payment cards, with the ecosystem application map continuously expanding.
👉 However: these technological advances focus on "human identity verification, privacy computing, L2 performance" and do not solve the underlying cryptographic security issues of token transactions. World ID's SMPC quantum security only protects iris biometric information, not WLD token wallet assets. WLD transfer signatures still use ECDSA-secp256k1 elliptic curve, and World ID's ZK proofs use BN254 curve; neither can resist future Shor quantum algorithm attacks.
The project currently has no independent post-quantum upgrade roadmap and can only passively follow the OP Stack Ethereum L2 ecosystem's unified upgrades; the initiative is not in WLD's own hands. Once fault-tolerant quantum computers mature, not only will WLD assets be at risk of theft, but World ID identity proofs could even be forged, directly shaking the core foundation of "distinguishing humans from AI robots."
2. High proportion of uncirculated tokens, long-term supply selling pressure risk
WLD total supply capped at 1 billion tokens, token distribution: 75% community, 25% team + early investors + reserves.
- Currently, circulating supply is less than half, with a large amount of tokens locked in linear unlocking schedules spanning over a decade; team and investor stakes are continuously released daily linearly until 2028, and community share unlocking even extends to 2038.
- Although the project has extended lock-up periods and reduced daily unlocking rates, the existing uncirculated stakes are a fixed fact. Whenever the market recovers, low-cost locked tokens have the motivation to sell and realize profits, long-term suppressing the price upside.
- Even if the ecosystem prospers, it can only increase demand but cannot directly eliminate the selling pressure from continuously increasing token supply. Whether the incremental demand brought by the ecosystem can sustainably cover the long-term unlocking supply is highly uncertain.
3. Dialectical view: these two risks are not immediate destruction but long-term looming threats
1. Quantum risk: a medium- to long-term future threat, not an immediate outbreak. Google's magic technology accelerates quantum hardware development, compressing the arrival time of Q-Day. When mature quantum devices land, WLD's entire cryptographic system will be exposed to vulnerabilities, and it has no autonomous upgrade plan for quantum resistance. No matter how rich the ecosystem applications are, underlying cryptographic vulnerabilities will not automatically be fixed.
2. Uncirculated token risk: a continuous supply pressure spanning over a decade. Even if ecosystem users continue to grow, once market sentiment weakens, the continuously unlocking low-cost tokens will keep impacting the market.
4. Horizontal comparison reference
STRK plans to upgrade to an independent L1 in 2027, aiming for full-chain quantum signature switching to proactively solve quantum risks from the bottom layer; but STRK also faces large monthly unlocking selling pressure.
WLD, on the other hand: has strong ecological identity technology, but the token's underlying cryptography lacks a quantum resistance solution, combined with the dual pressure of long-term large uncirculated tokens. Ecosystem development can only improve fundamentals but cannot eliminate these two major hidden risks.The continuous downward line on the cardiac monitor for eight days finally bounced slightly in the early hours today. The Nasdaq futures briefly rose nearly 1%, resembling the violent chest jerk after an external defibrillator shock—don't rush to declare revival; first, see if it's sinus rhythm or a precursor to ventricular fibrillation. The SOX index has fallen for eight consecutive days; this is not ordinary arrhythmia, but a persistent coronary artery spasm, with AI storage and optical communications being the leads showing the earliest ST-segment elevation. After TSMC delivered a record-breaking Q3 earnings report, its ADR dropped accordingly—this contradiction is all too familiar on the operating table: the donor heart's function indicators are perfect, but the recipient's blood pressure just won't rise—the problem isn't the graft itself, but the perfusion pressure, with the aortic valve of valuation already calcified and stenotic. The debate over AI return cycles is like chronic rejection, chip equipment demand is myocardial perfusion scanning, and the upcoming tech earnings reports are like intraoperative transesophageal echocardiography; every capital expenditure data point is a key cross-section for assessing ventricular wall motion. The partial uptick in some chip and optical communication stocks last night is at best reactive hyperemia after ischemia-reperfusion injury, with transient capillary bed dilation, not resolving the organic stenosis of proximal vessels. There is only one true criterion: whether trading volume can show a meaningful decline like myocardial enzyme profiles, and whether the revenue guidance of leading stocks has sufficient collateral circulation compensatory capacity. Without these, the rebound is reperfusion arrhythmia again; behind the beautiful QRS complexes on the monitor lies a dying myocardium giving a false sense of security. #aichipstocksrebound Xiao T's Investment Research Self-Study Room (2) What are VOL and OI, and how to predict future possibilities through price, VOL, and OI? VOL (Volume) is the total number of contracts traded within a period. "Price is direction, volume is the fuel of the market," used to judge whether price rises or falls involve real capital participation, the strength of the trend, true or false breakouts, and the degree of long-short divergence; it is an auxiliary verification indicator. OI (Open Interest) is a unique indicator in the contract market, referring to the total number of all open long and short contracts currently in the market. How should the combination of VOL + OI + price be viewed? This is the most practical "Volume-Position-Price Framework" in contract technical analysis. Essentially, it judges whether the market movement is driven by new capital inflow or by old positions closing— the sustainability of these two is vastly different and is the core tool to identify false breakouts and fake market moves. ❗Technical indicators are only for market reference and do not constitute any investment advice. 1. Four combinations on the upside 1. Price rising + VOL increasing + OI rising — Healthy main upward wave Pattern: Price steadily hits new highs, volume exceeds the average volume line, and open interest continuously rises in sync. Essence: New long capital actively enters to open positions, while new shorts enter to take the other side; both longs and shorts increase, but longs dominate. The market is driven by incremental capital. Common scenarios: Early trend start, upward breakout after sideways moving average convergence, mid-stage of the main upward wave. Follow-up: The strongest trend continuation, the core confirmation signal of a "true breakout"; as long as OI does not turn down and volume is clear Today's rebound was led by DeFi.
CoinDesk: DeFi index +0.81%, Arbitrum led with +6.7%, returning near $0.20. Catalyst: Arbitrum Foundation's stablecoin growth plan, 2 million ARB subsidies for the top 10 on-chain protocols holding $10 million USDG.
Uniswap +3.2%, Lido +2.5%, Ethena +1.6%, Pump.fun +7.6%; BTC only +0.5% the same day, around $82,500.
Futures 24-hour volume +51% to $146 billion, open interest only +1.25%; Deribit's top five option contracts by volume that day were all Bitcoin put options.
I think this DeFi rally is driven by subsidies and sentiment hype; real demand hasn't returned. The cost-performance of chasing highs is very low, so I will keep observing and not chase. Invalidation level: ARB holds above 0.20 and USDG real locked-in capital comes in, or BTC breaks out with volume above 83,300–84,300, then this judgment is void.
Altcoin season index 56/100, last week's high was 63.
Do you think this DeFi rebound is just the start of rotation, or will it fade once subsidies stop?
$ARB $UNI $BTC
#DeFi #cryptocurrency $BCH Today CME launched futures for BCH, but the market didn't even make a splash. I've said this twice already, but I have to keep emphasizing it. Last time I said launching futures doesn't mean someone is buying; now it looks even more awkward—the coin price was still down last week, and this week it climbed back just a bit, with volume only about 70% of usual. At the moment new futures go live, volume actually shrank. What does this mean? It means institutions don't care about cash-settled contracts since they don't have to touch the spot market. So who really uses it? Those who hold coins but don't want to sell and want to lock in prices—they just open short positions. Miners, project treasuries, they're the first to use this tool. So this news only provides the market with more hedging options, not something to pump the price. The coin price is stuck around 280, looking stable from the outside but actually weak inside. $BCH #NetflixQ3AdsAndMargins Netflix (NFLX) 2026 Q3 Bearish Report
Q3 revenue guidance is $12.86 billion, a 12% year-over-year increase, appearing solid on the surface, but the quality of growth is deteriorating. Global viewing time in the first half of 2026 increased by only 2%, with average viewing time per person actually declining by about 5%. Growth relies on continuous price hikes rather than engagement expansion, and the price increase space is nearing its ceiling.
The competitive landscape is structurally reversing. In July, YouTube's share of U.S. TV viewing was 14.2%, while Netflix's was only 7.8%; YouTube has already surpassed Netflix in daily viewing time. Netflix is even considering a "zero-threshold + ad-supported" fully free plan, signaling a weakening of its paid model competitiveness.
The advertising "second engine" has cracks. Q2 growth near 80% still fell short of expectations. Free FAST apps already account for 19.1% of U.S. TV viewership, causing an oversupply of ad inventory, putting pressure simultaneously on Netflix's unit pricing and fill rates.
Sustainability of margin improvement is questionable. Q3 operating margin is expected at 33.2%, with year-over-year improvement mainly due to a low base from a one-time Brazilian tax expense last year. Just before the earnings report, about 1,000 employees (6% of staff) were suddenly laid off, the largest scale since 2022, indicating operational pressure at this timing.
Huber Research unusually downgraded its rating from "Overweight" by two levels to "Underweight," with a target price of $70; Wells Fargo gave $57. When price hike-driven growth hits a ceiling, the advertising narrative fails to deliver, and cost cuts become the main source of profit margin, valuation lacks a margin of safety.
#Netflix growth may slow, advertising business becomes a new focus #QualcommEarnings
Qualcomm has reported earnings 12 times, beating expectations 11 times, but the stock only rose 3 times the next day. This time, I'm betting on a rise.
First, why does it fall despite beating expectations: My understanding is that the market doesn't focus on how much was earned this quarter, but on the guidance for the next quarter. The beats have long been priced in.
This time I'm betting on a rise, gambling that expectations are already low:
① Profit threshold cut by 9% over 90 days: EPS forecast dropped from $2.37 to $2.16
② Revenue forecast $10.1 billion, 10% less than the same period last year; analysts have already factored in the bad news
③ Stock price at 176, down 29% from the May high of 249
④ Options only bet on about 6.8% volatility, similar to the average 6.7% over the past 12 times; the market isn't betting on a big move
Risks upfront: Historical win rate is against this, only 3 rises out of 12 times. A deep drop doesn't guarantee a rebound; last April, it was down −34% from the high and still fell 8.9% after earnings. The most recent rise was in April this year, +15%.
Earnings released after US market close on 11-04, see you then!Crypto Circle Financial Report Analysis: Raydium Q3 2026
Raydium is one of the most important decentralized exchanges on Solana. The Q3 quarterly report looks impressive on the surface, but there are significant hidden risks in the details
Trading Fees: Reached $57.1 million in Q3, more than tripling from about $17 million in Q2. September alone contributed approximately $44.6 million
Trading Volume: About $16.1 billion, a 17% quarter-over-quarter increase, but 62% concentrated in September
RAY: Rose from $0.63 to $1.91, a quarterly increase of 204%
Buybacks: Approximately $6.2 million spent on RAY buybacks in Q3, a significant quarter-over-quarter increase
LaunchLab Upgrade: In August, external platforms were allowed to open their own launchpads on Raydium
StonkFun Migration: On September 5, a launchpad pairing new tokens with tokenized stocks migrated all issuances to Raydium's LaunchLab
Main Issues:
Growth heavily depends on a single partner. Historically, in similar situations, LaunchLab fees dropped sharply by 79% the month after the partner cooled down
Early October data shows daily fees are about 43% lower than in September
July and August were weak themselves, with almost all the quarter's highlights concentrated in September I really can't suppress the anger in my heart. Crypto, in the past two years, every market cycle has relied on liquidity spilled over from other markets. Last year, there was a big rally because of liquidity coming from the gold market; this big rally is due to liquidity coming from tech stocks. Crypto doesn't create its own wealth; it can only barely rise once by relying on handouts from other markets. This is literally like a stray dog on the roadside! If someone feels pity, they throw a piece of sausage; if no one cares, it's just a stray dog on the street. —————————————————— Today's market rebound vividly shows the true nature of crypto. Tech stocks in the US and South Korea are both rebounding, but what is crypto doing? It's freaking falling! Looking at this comparison chart, it's obvious that the trends are almost opposite. I can only say crypto is eating leftovers; what future can eating leftovers have? I now believe it's better to play the US stock market. Why? Because the US stock market has ups and downs. Those mega-cap stocks, after surging up sharply, will fall a lot; they won't let both shorts and longs be wiped out. But crypto is not like that. Crypto's rises and falls are very one-sided: either it kills all the shorts or kills all the longs. In this situation, making money is as hard as climbing to the sky. But the US stock market, after falling a lot, will have a very good rebound; after rising a lot, there will also be a considerable pullback. The US stock market's way of cutting retail investors is not as ruthless as crypto's. —————————————————— IZcash decision tomorrow: 22,000 old ZEC may be permanently frozen
Zcash's NU7 upgrade proposal (ZIP 2003) will make the final go/no-go decision tomorrow (October 20). If approved, it is targeted to activate on November 5, and the old v4 transaction format will be disabled.
Affected are the old Sprout pools: about 22,309 ZEC (as of October 17) are inside. Once v4 is disabled, these coins cannot be spent. Note, they are not burned; the gate is just closed—the coins remain on the ledger but cannot be transferred out. Only those with Sprout balances left in old wallets from 2016–2018 are affected; exchange users are unaffected.
ZEC has been hot on Orbit for 4 days, with all the talk about price, but no one has yet accounted for the technical side. Tomorrow we will see the decision result.OP 30 days +66%.
a/ On 10/13, Unichain production chain was handed over to OP custody, mainnet launched on 10/29
b/ Upgrade 20 passed: 50% of Superchain revenue used to buy back OP
c/ OP Stack accounts for 66.8% of Ethereum L2 transaction volume
On the other hand
· Buyback only starts in February 2027, no spending so far
· Token down −97% from peak
· Base left in February taking away 70%+ of sequencer revenue
Chain is growing long, but the token hasn’t kept up.WLD has no quantum resistance capabilities at all in terms of quantum security! Summary of WLD (Worldcoin) in one sentence:
The WLD token itself (on-chain transfers, wallet signatures) natively lacks quantum resistance. Only the World ID iris biometric offline multiparty computation AMPC module partially uses quantum-secure cryptography; however, WLD token transaction signatures still use the traditional ETH/ECDSA secp256k1 elliptic curve cryptography, facing the same quantum cracking risks as BTC and ETH.
1. Splitting into two cryptographic systems makes it easy to distinguish
1. WLD token transfers (World Chain, OP Stack Ethereum L2)
- WLD is an ERC20 token, World Chain belongs to Ethereum L2, wallet transaction signatures use ECDSA-secp256k1, the same cryptography as Bitcoin and Ethereum wallets.
- In the future, when fault-tolerant quantum computers mature, Shor's algorithm can derive private keys from exposed public keys to steal tokens. This part has no built-in post-quantum signature.
- Upgrades can only rely on the overall post-quantum upgrade path of OP Stack (Optimism superchain). Worldcoin itself has no independent post-quantum transformation plan and must upgrade along with the Ethereum ecosystem, with no clear implementation timeline.
2. World ID identity system (the iris recognition set)
- World ID uses zero-knowledge proof ZK to prove real human identity; its underlying Groth16 (BN254 elliptic curve) is also not quantum-resistant;
- But its iris comparison AMPC privacy module uses quantum-secure multiparty computation SMPC, specifically protecting iris biometric data.
👉Key point: This quantum resistance only protects iris privacy, not the security of on-chain WLD token assets. Many people easily confuse these two points.
2. Horizontal comparison with STRK
- STRK: plans to upgrade to an independent L1 in 2027, replacing the entire chain with post-quantum signatures, aiming for native quantum resistance in token transactions;
- WLD: underlying cryptography for token transactions is not quantum-resistant, only the biometric privacy module has quantum security, token asset protection fully inherits Ethereum's quantum risk.
3. Core summary
✅ Worldcoin is not completely without quantum security technology, but it is limited to the iris identity privacy module.
❌ WLD token transfers and wallet asset signatures do not have built-in quantum resistance, sharing the same generation of cryptography as ETH and BTC, and will face threats from quantum computers in the long term.
📌 To achieve quantum resistance at the WLD token layer, it can only wait for the OP Stack superchain to complete the post-quantum cryptography migration, which is a long-term plan with no current implementation.ETF is selling, which is a different matter from whether the price falls or not
$BTC is now 81,139, $ETH 2,378.
This week, spot ETF net outflow is 228 million, with $ETH at 319 million.
Where does the money come from:
The money buying ETFs goes through the subscription and redemption channel, not on the order book.
Redemption requires selling coins to cash out, and the sell orders fall on the spot market.
How is this number calculated:
$BTC net outflow is 228 million, calculated at eighty-one thousand, about 2,800 coins.
These coins are not sold by someone bearish, but the product is shrinking.
Institutions are withdrawing, leverage is clearing, volume is shrinking, three directions pressing simultaneously.
If 81,800 cannot be reclaimed, any rebound is just a rebound.
Next level to watch is 80,000.
#机构资金转弱，BTC与ETH谁先迎来回流？
#$BTC $ETH #Netflix growth may slow, advertising business becomes a new highlight, bullish$BTC $ETH #Institutional funds weaken, who will see the first inflow, BTC or ETH?
Withdrawal has slowed, but the money hasn't truly returned yet. As of October 16, 2026, US BTC spot ETFs have seen net outflows for two consecutive weeks, and ETH for three consecutive weeks; from October 12 to 16, net outflows were $228.4 million and $320.1 million respectively, narrowing by 66.4% and 41% compared to the previous week.
After the Fed rate hike in September, inflation rose again, so patience is still needed to verify fund inflows. BTC outflows have narrowed faster, providing a clue that funds may stabilize first, but moving from reduced outflows to sustained inflows is still one step away. ETFs do not represent all institutions but offer a trackable clue on fund flows.
In the next 3–5 trading days, the focus is on whether the rolling 5-day net flow can turn positive and continue, and whether multiple products can see inflows together, to avoid mistaking a large inflow in a single fund as a broad recovery. For ETH, it is also important to see if ETHA redemptions ease and if ETH/BTC can strengthen. If ETH meets these conditions first, the judgment that BTC stabilizes first should be overturned.Both BTC and ETH are experiencing capital outflows, so why can't the same logic be applied to analyze their market trends?
One of the most common mistakes in market analysis is treating Bitcoin and Ethereum as two completely synchronized assets.
Last week's ETF data is quite worth pondering. From October 12 to 16, the U.S. spot BTC ETF saw a net outflow of about $228 million, while the ETH ETF had a net outflow of about $320 million. Looking at the numbers alone, ETH's outflow amount is even larger.
But we cannot directly conclude that ETH is weaker than BTC here. The fund sizes, market depth, and capital structures of the two are different, so comparing absolute amounts is not entirely fair.
What I care more about is: when institutional funds withdraw in phases, is there other buying interest in the market to take over? If ETFs continue to see outflows but spot prices remain supported, it indicates possible divergence among different capital groups; if capital outflows, volume changes, and price weakness occur simultaneously, the market needs to reassess demand.
Additionally, ETF data itself has a release lag, so you can't use last week's numbers to explain every candlestick today.
So the question arises: if BTC and ETH continue to diverge, which do you think better reflects the true attitude of institutional funds? Is it by looking at ETF net flows, or by observing whether the spot market can absorb selling pressure? #机构资金转弱，BTC与ETH谁先迎来回流？ $BTC $ETH At night, Bitcoin surged from 80,600 to above 83,000, then a few bearish candles pushed it back down to 82,200. The rise and fall pattern is very clear.
Three major pressures: September CPI year-on-year at 3.7%, PPI also exceeding expectations, pushing back rate cut expectations; spot ETFs have seen net outflows for the second consecutive week, about $228 million last week; sentiment remains in the greed zone.
Volume above 83,000 did not keep up, no new money stepping in. The spike at 80,600 looks more like a panic washout, the pullback is just short covering.
Now 82,200 is stuck in the middle, no chasing the rally, waiting for a pullback to stabilize before making a move. $BTC$ETH This drop looks more like a short squeeze rather than the start of a major crash.
First, look at the positions; they are more useful than the price. If a big drop were really coming, shorts would keep adding positions during the decline, and open interest should move upward; this time it’s dropping, meaning leverage is being cleared out. The liquidation structure matches this: 97% of the positions liquidated in the last hour were longs, while short liquidations are almost negligible. This cut is just a leverage purge.
But the idea that it will pump right after the purge doesn’t hold either. ETFs have seen a net outflow of one billion over three weeks, which is real selling pressure that leverage alone can’t explain. Every rebound to 2,450 to 2,480 is pushed back down; there is real money selling at the top. A single needle is enough to kill leverage, but lifting prices requires new money, and new money hasn’t arrived yet.
The bullish case is based on structure. The price dipped to 2,350 but was pulled back, and the lower channel boundary wasn’t broken—this indicates that what was crushed were leveraged positions, not spot holders running away. After clearing some floating supply, the same buying power can drive a larger rally later.
This bullish talk is about structure; the timing isn’t here yet. There are only two reliable signals: reclaiming 2,450 with volume, or ETFs stopping outflows. Until then, don’t mistake a short squeeze for an imminent rally. BlackRock has made a big move again. Just now, a wallet related to an ETF transferred $143 million worth of BTC and ETH to Coinbase Prime, including 883 BTC valued at $72.64 million and 28,957 ETH valued at $70.4 million. The amounts on both sides are almost evenly split. When suddenly a blockchain transfer of over $100 million appears, no wonder the market is getting nervous again.
However, I’ve noticed an interesting phenomenon in the crypto space: whenever the keywords BlackRock, Coinbase Prime, and large transfers appear together, people in the comments immediately shout that institutions are about to dump. The problem is, transferring into an exchange doesn’t mean an immediate sell, especially since Coinbase Prime itself handles institutional custody and settlement services. ETF subscriptions and redemptions, account reallocations, and liquidity arrangements could all involve such transfers. You can’t judge what BlackRock intends just based on a single on-chain record.
But the timing is indeed sensitive now. BTC just lost 82,200 on the weekly chart, and ETH has been capped at 2,494. The market is already lacking confidence. If ETF funds continue to flow out and spot selling pressure adds up, BTC’s 80,000 and 78,300 levels will need to be watched carefully, and ETH might retest around 2,380.
Currently, I’m still bullish, but I won’t blindly turn bullish just because institutions are increasing holdings, nor will I rush to short just because of a transfer.$INJ
INJ still falls nearly 2%, why does an improved range position not equal a trend reversal?
In the morning market snapshot, the spot price dropped about 1.96% in 24 hours, with a range of 7.803—8.832, and the price is located at about 68% of the range.
Negative returns and a relatively high range position can appear simultaneously, reflecting partial recovery from the low but the starting point has not yet been surpassed. Financial applications require real usage and fee support, and cannot be proven solely by how far the price is from the low point to justify a business revaluation.
If 8.832 is reclaimed and a higher low is formed, the repair will be more complete; if 7.803 is lost, a reassessment is needed. Assets that lag during a market rebound should not be preset as the next leader just because they have fallen a lot before.On Monday as soon as the US stock market opened, the Nasdaq jumped 1% directly, with AI and chip stocks collectively rebounding, and the screen was full of calls saying "the bottom is in." I haven't moved a single one of my crypto short positions in this wave.
Why panic when it doesn't rise, or rush to close when it doesn't improve? Because this is just an emotional rebound at Monday's open, not a trend reversal. The AI deleveraging line is being pressed down bit by bit by financing costs and debt pricing; it won't reverse just because of a few bullish candles on Monday morning. $BTC is still grinding around 82,000, and $ETH can't even hold 2,400 — if a reversal were really coming, the weakest leg would be the first to strengthen, but that's not happening now.
When trading trend positions, the biggest fear isn't holding on, but being shaken out by a single rebound. My rule is simple: if the framework isn't broken, the position stays; I only admit I'm wrong when the market itself tells me so.
Are you guys trying to catch this rebound now, or sitting tight like me?8 major altcoins, 7 of them have BTC exchange rates in the lower half of the 100-day range.
DOGE is only at 9%, XRP 16%, ADA 24%, ETH 27%.
Only AVAX barely stands at 51%.
BTC dropped only 2.6% in 20 days, altcoins dropped 10% to 15%, which seems like just a bit more.
But when converted to BTC exchange rates, it's a different story.
Exchange rates tell the real truth; the USD price is just a packaged facade.
XRP dropped 13.8% against BTC in 20 days, LINK even worse, down nearly 20%.
This is not just a bigger drop, it's continuous bleeding.
A big drop invites bottom-fishing, but bleeding means no one wants to catch it.
ETH dropped 9.4% against BTC in 60 days, and another 9.7% in 20 days, continuously declining over two months.
Every rebound is a sharp altcoin bounce, followed by an even faster fall.
I think the reason is simple: money is flowing out of altcoins into Bitcoin, not rotating within altcoins.
Altcoins' USD prices hold up because BTC supports the overall market.
Once BTC stalls, altcoins will continue to fall.
Now DOGE against BTC is only at 9% of the range, just one step away from its floor.
Breaking below means a new low, and that’s when real bleeding starts.
I’d rather hold Bitcoin and wait than catch altcoins’ falling knives at this point.
$XRP
#AltcoinsAreBleeding#Netflix growth may slow, advertising business becomes new focus How many Bitcoins $BTC do all publicly listed companies in the world hold in total?
Just checked the data: a total of 1,286,722 Bitcoins are held. The top 100 hold 1,284,004, and all publicly listed companies outside the top 100 combined hold only 2,718.
Among these top 100 companies: MicroStrategy alone holds 848,517 Bitcoins. This accounts for 66% of the top 100.
And 66% of all publicly listed companies.
#机构资金转弱，BTC与ETH谁先迎来回流？ #Institutional funds weaken, who will see the first inflow, BTC or ETH?
Crypto Newsflash: Strategy increased its BTC holdings by 267 last week, preferred stock repurchase authorization raised to $3 billion
The latest 8-K filing submitted to the SEC today and data disclosed by Michael Saylor reveal that behind this seemingly routine weekly report lies a major tactical shift in Saylor's capital operations focus
Last week, only $22.1 million was spent to increase BTC holdings by 267 coins. The total holdings now reach 848,517 BTC, with a total cost of about $64.02 billion, and an average holding price of $75,446 per coin
During the same period, $193.1 million was spent to repurchase and cancel about 1.9388 million shares of its own STRC
The board officially raised the authorization limit for the "Digital Credit Securities Repurchase Program" from $2 billion to $3 billion
The money spent on repurchasing preferred stock is nearly 9 times that spent on buying Bitcoin💵
Saylor's core belief has never been just hoarding coins, but maximizing BTC yield and capital structure arbitrage,
When MSTR common stock premium is extremely high, the optimal solution is to print shares to buy BTC;
But when bond market volatility causes its own digital credit securities to trade at a discount or high yield in the secondary market, repurchasing its own preferred stock at a dollar discount is equivalent to directly achieving risk-free net asset value enhancement.$CT has been positioned from 0.3 to just above 0.4, with gains exceeding 25%. It seems profitable, but actually hides a short squeeze risk.
This rally is mainly driven by short covering, combined with expected airdrop speculation, causing trading volume to surge sharply in a short time. However, once the short liquidation wave ends and buying momentum falters, the price may quickly fall back.
What’s more concerning is that the top ten holders control about 86% of the supply, indicating highly concentrated holdings and significant selling pressure risk.
Therefore, I chose to take half the profits first and set stop-losses on the remaining position. For such highly volatile coins, locking in profits and managing risk in time is more important than greedily chasing the final leg of the rally.
#BTCETHInflowsWatch White House Crypto Chief Leaves, Where Should Retail Investors Put Their Money? On October 14, Patrick Witt, Executive Director of the White House President's Digital Assets Advisory Committee, left after his secondment period ended. This means that the core White House team responsible for coordinating U.S. crypto policy currently has no publicly confirmed staff. This is not an isolated incident. In March this year, David Sacks, head of crypto and AI affairs, resigned; in July, Deputy Director Harry Jung left. Witt is the third key White House crypto official to leave this year. Behind the staff turnover is the blockade of the "CLARITY Act" he led in the Senate — a procedural vote on September 15 failed to advance by 49 to 50. The disagreement centers on how to handle stablecoin yields and how to address President Trump's own significant crypto asset conflicts of interest. For the market, the direct impact of this event is limited. Witt's departure itself will not immediately affect prices, but it confirms a more critical fact: legislative catalysts will be postponed at least until after the November 3 midterm elections, or even later. Before the Federal Reserve meeting on October 28, the market will be driven by U.S. Treasury yields, oil prices, and ETF fund flows, rather than regulatory news. For retail investors, this provides a clear direction: do not bet your positions on expectations of "regulatory benefits." The White House crypto policy team is effectively empty, and the legislative path is temporarily blocked. In the coming period, market pricing power will return more to macro data. Manage your positions and keep an eye on U.S. Treasury yields and The storage sector collectively surged tonight, with SK Hynix on OKEx directly hitting 143, approaching a new high for the period. Leading the charge was SanDisk, which doubled with a 15% increase; the entire storage chain is moving.
The driver is all new catalysts today: South Korean Finance Minister Hong Nam-ki made statements at the IMF annual meeting over the weekend, saying the semiconductor boom will last at least until 2027, with a large supply-demand gap; Samsung is reported to be using TSMC's 3nm process to make next-generation custom HBM base dies for Nvidia, and Nvidia also requires Hynix and Micron to unify their processes; the South Korean government today also released 2 trillion KRW to support Hynix's supply chain. These three events combined have triggered a capital rush.
Looking at the market: OKEx xSKHY current price is 142.81, 24h range 132-143.25, MA5/30/120 bullish, super trend at 141.17 below, RSI 62 not overheated. My judgment: HBM leader plus policy endorsement plus earnings expectations (Hynix earnings on 10/28), this wave is a fundamental resonance, not pure sentiment, but the 143 resistance line has been tested for several days; whether it breaks is key.
Watch two signals: a volume breakout above 143 is a new step up; if it doesn't hold, wait for a pullback to 138.5 support; also watch for continuity of HBM supply chain news before the 10/28 earnings.
#SKHynix #Storage #HBM
Not investment advice, DYOR $BTC $ETH $SOL Is it like replaying the August script?
The recent rhythm of BTC, ETH, and SOL indeed shows shadows of the bull market kickoff in August: prolonged narrow-range consolidation grinding the bottom, bulls' patience being worn down, shorts gradually accumulating positions, and a chip structure similar to that time. BTC first stabilizes the key support, ETH then shows signs of stabilization, and SOL, as a high Beta asset, has the greatest elasticity and often leads once the market starts.
There are also logical similarities: macro expectations combined with capital rotation. If BTC breaks through key resistance, it easily triggers concentrated short liquidations, forming a short squeeze rally. In August, shorts were crowded, and after the breakout, a chain of liquidations rapidly pushed the market higher.
But history does not repeat exactly. The core driver in August was the liquidity boost from US Treasury bonds, which is not entirely the same this round; similar patterns do not guarantee replication. The focus next is whether spot buying can take over and whether SOL can show independent strength, which will determine the sustainability of the rebound.
#US CPI and PPI both exceed expectations, will there be another rate hike within the year? #Institutional funds weaken, who will see inflows first, BTC or ETH? #SpaceX's first full quarterly report after going public will be revealed The second batch of OKX earnings season predictions has been posted
Those who participated in the Battle Royale event should ensure they have participated in the predictions for these 4 markets
⬛️ For Meta, I chose Yes, for the following reasons
The Meta threshold is $6.31
▪️ MarketBeat's aggregated analyst EPS expectation is about $6.58 🔺
▪️ FXEmpire's earnings forecast page shows an EPS consistent with the threshold
In terms of business, Q3 revenue is expected to be about $63.3 billion, with an estimated year-over-year growth of 24% and advertising business still supporting growth
Based on the above, I decided on Yes
⬛️ For SPCX, I chose No, for the following reasons
The SPCX threshold is $0.17
▪️ TipRanks' earnings preview gives an EPS expectation of about $0.14 🔻
▪️ Investors' reported forecast range is about $0.14 to $0.23
Overall, there is a clear divergence in SPCX's expected distribution, and the $0.17 threshold is not a low bar, already higher than many analysts' forecasts
Based on the above, I decided on No