Orbit Post Sitemap
350亿美元，足够吸引眼球。但Meta这次融资真正值得问的是：钱什么时候到，股东要为此让出多少权益，AI又要多赚多少才能覆盖这笔投入？ Meta在10月26日公告中启动约350亿美元的两项发行：约225亿美元A类普通股，配合远期出售协议；约125亿美元代表强制可转换优先股权益的存托股份。这不是一笔普通银行借款。 到账时间尤其容易被忽略。公告明确，公司起初预计不会从远期购买方出售普通股中收到资金，而是在未来结算远期协议时取得相应款项。因此，不能把“启动350亿美元发行”写成350亿美元已经全部进了公司账户。 计划里还有约600亿美元的未来按市价发行项目。但它尚未签订相关协议，具体时间、规模和条款仍未确定，并预计不早于2027年一季度启动。把两部分相加，可以讨论约950亿美元融资规划，不能报道成已完成募资。 扩张的支出也很大：公司预计2026年资本开支为1350亿至1450亿美元，并披露前三季度经营现金流970亿美元。前者是全年预期，后者是九个月实际值，不能简单相减，冒充全年现金缺口。 我的判断：股权融资能增加资金来源，也带来潜在稀释。判断这笔钱值不值，最终要看新增投入能否带来足够的盈利和AI’S MEMORY BOOM FACES A REALITY CHECK.
SK hynix reports Q3 earnings on October 28.
Analysts expect nearly ₩77 trillion in operating profit and margins above 78%.
But the headline isn’t the real test.
Can HBM4 shipments scale fast enough to justify those expectations as Samsung intensifies competition?
For AI markets, tomorrow’s guidance on production, pricing and demand may matter more than record profits.
#AI
#SKhynixQ3HBM4Earnings Anthropic走到了公开递交IPO文件这一步，但“申请上市”和“已经确定上市价格”，中间还有很长一段路。热榜里的估值纪录，现在还不能当成落定的结果。 10月26日递交至美国SEC的S-1文件写明，公司申请在纳斯达克全球精选市场挂牌，拟使用代码ANT。文件仍是初步招股书，发行数量和发行价格区间保留空白，并明确说明，信息不完整、可能修改，注册声明生效前不能出售所述证券。 我更关注文件里另一句：这次拟议发行既包含公司出售股份，也包含售股股东出售股份；公司不会从后者的出售中获得资金。也就是说，未来即使看到一个很大的IPO总规模，也要继续拆开：多少进入公司账户，多少属于原有股东出售。 这会影响我们理解融资的意义。公司拿到的新资金，可以支持后续经营投入；老股出售主要改变股份持有人。两者都属于发行的一部分，却不能全部算成新增的AI研发或算力预算。 我的判断：现在讨论“会不会创纪录”，不如先等能计算的条件。发行价、发行后股数、新股与老股比例、公司净募资额，缺一项，都可能改变对交易的理解。估值越高，也越需要经营增长兑现，而不是只比较谁的数字更大。 后续若修订文件补齐价格区间和发行结构，再结合经营表现客户把钱存进平台，和平台把利润赚到手，中间还有一段距离。Robinhood这次财报，我更想看这段距离有没有缩短。 公司投资者关系页面列出的第三季度电话会时间，是美东10月27日17点，对应北京时间10月28日凌晨5点。这篇先按财报前瞻讨论，不能把尚未核对的季度成绩提前写出来。 已核对的8月运营数据里，净存入资金为40亿美元，过去12个月累计741亿美元。资金进入平台，有助于扩大业务基础，但净存入不是营业收入；统计范围还包含并购业务及部分托管账户的贡献，不能全部当作原有业务自然增长。 交易数据也不是整齐向上。8月股票交易金额同比增长68%；期权合约数同比增长50%，环比却下降10%；加密交易金额175亿美元，环比增长61%，同比下降38%。同一业务，选不同的比较基准，就会讲出完全不同的故事。 而且，175亿美元里，Robinhood App占74亿美元，Bitstamp占101亿美元。两部分的客户结构、收费方式可能不同，不能把合计交易量直接乘上一个固定费率，就算作公司收入。以上均为8月初步、未经审计的运营数据，不能代替完整三季度业绩。 我的判断：增长能否延续，要看客户资金是否留下、交易最Brother Sun added another $20.32 million $PEPE?! Since 10.17, a total of $48.76 million has been transferred out 🐸
10 hours ago, the address 0x721…E5991, marked by Arkham as associated with Sun Yuchen, withdrew 4.65 trillion PEPE; currently Brother Sun holds a total of 11.8 trillion $PEPE, with an average withdrawal price of $0.000004114. Despite market fluctuations, there is still an unrealized profit of $2.085 million, but it is unclear whether this is Brother Sun's personal wallet or exchange-related funds.
Wallet address /0x72132bF09a9ee6517F3f1d997b4b3cd2115E5991The 28th will be exciting, with the Federal Reserve's policy meeting, but I don't think they will raise rates this time, especially since the election is on November 4th. There's no need to raise rates again before the election. Currently, employment and the economy are holding up. The inflation caused by oil is a bit high, but you gain some and lose some. The growth rate of U.S. oil exports is somewhat exaggerated. Inflation is inflation, but money is being made. Now even Middle Eastern and European oil has to be booked years in advance.
Recently, there have been a huge number of short positions. Then it's time for Microsoft, Google, and Amazon earnings reports. If there's any issue, tech will immediately crash on you. At this point, if their capital growth meets expectations and earnings are a bit better, shorts will be aggressively targeted.
I still believe the real risk lies after the midterm elections. The big downturn hasn't come yet; keep watching when the last two upward trends start to weaken. That will be the total collapse.
#AWS算力需求旺盛，亚马逊Q3财报有哪些看点？ 看到中东原油运输恢复，第一反应往往是：供应回来了，油价该降了吧？但恢复出口量，还没有回答炼厂要花多少钱才能把油运到手。 10月26日，Discovery Alert的一篇油市分析援引花旗数据称，9月海湾原油流量，不含伊朗，已恢复至战前水平的81%以上。这里说的是特定范围的运输恢复程度，不能扩大成整个中东已经完全恢复正常。 同一报道指出，每天超过100万桶原油通过船对船转运完成运输。转运能让货继续走，却也增加作业环节、占用船舶时间。即使出口桶数恢复，完成同样的运输量，也可能需要更多船和更长时间。 这就解释了一个容易忽略的矛盾：油在流动，物流效率却可能仍没恢复。该报道提到的海湾至中国超大型油轮运费约为每桶35美元，是特定航线的费用口径，不能当作全球平均，更不能直接加在任何一桶原油报价上。 我的判断：运输恢复可以缓解供应紧张，但要判断这部分成本压力是否消退，还得看航程与转运是否减少、运费与保险成本是否回落。出口量只是其中一项指标。 当然，高成本也不保证油价上涨。若航线逐步正常化，同时终端需求转弱，原来的风险溢价就可能被压下来。后续如果出口继续恢复，相关航线运费也持续下降，我会撤回对物流成本仍THE RISK MAP HAS A CRYPTO CORNER.
Moody’s Analytics’ latest matrix examines 27 potential shocks over 12–18 months. A crypto crash ranks among the more likely risks, but its estimated damage to the wider economy is comparatively limited.
No crash probability. No price target. Just a reminder that liquidity matters more than headlines ahead of the Fed's Oct. 28 decision.
Read risk maps as warnings, not trading signals.
#Macro #CryptoTreasuryDivides 盘面安静得有点发闷，BTC 在 800 一带反复试探，上方 805 到 808 压着，下面 800 像一道不太情愿退让的防线。 这种"没死透也没活过来"的状态，是不是比单边行情更磨人？ 我盯着这段横盘看了很久，最大的感受不是方向问题，而是节奏问题。纳指盘前稍微回了点血，但 28 号 FOMC 就在眼前，结果出来前不太可能给出很强的修复，能摸到 293xx 到 295xx 已经算体面。真正让交易变难的是后面：28 号大概率插针，11 月 3 号中期选举接上，接下来两周市场运行速度会明显加快。没在旧低或旧高拿到筹码的人，更适合日内和短波段，别硬改结构，不然赚了没落袋会变成日常剧情 😂 这也是我想聊的风险管理镜头：现在不是比谁看得准，是比谁在波动放大前把仓位和止损想清楚。偏多的一边在于，800 防守还在，纳指盘前有修复意愿，说明风险偏好没有彻底熄火，一旦 FOMC 落地、选举不确定性开始被消化，BTC 有机会带着情绪回暖，ETH 和山寨才敢跟上。偏空的一边是，上方 805 到 808 没拿下之前，反弹都只能算修复，插针行情里追高最容易被打脸，杠杆和密集止损区会被反复扫。 XAU 这边也类似Will SanDisk's earnings report this quarter exceed expectations?
My judgment is: the probability that the performance figures will exceed the market consensus is relatively high, but it is not highly probable that the stock price will continue to rise after the earnings report.
There are three main factors supporting the positive performance:
NAND is still in a seller's market, with tight supply and product price increases continuing to support revenue and gross margin;
The company’s revenue, gross margin, and free cash flow all improved significantly last quarter, and operating leverage is still being released;
Data center and enterprise storage demand is strong, and AI-related high-capacity products remain the main growth drivers.
But the risks are also clear. SanDisk’s stock price has already risen significantly, and the market has priced in strong profit expectations in advance. For this type of cyclical stock, what truly determines post-earnings performance is often not "whether last quarter exceeded expectations," but rather:
Whether revenue and profit guidance for the next quarter will continue to be revised upward;
Whether the NAND price increase is starting to slow down;
Whether the non-GAAP gross margin has peaked;
Whether long-term supply agreements can improve revenue visibility;
Whether the data center business can continue to maintain rapid growth.
SanDisk is very likely to deliver an earnings report that exceeds market expectations, but the stock price performance may be "earnings beat but no stock price increase," or it may be a "slight beat with stock price adjustment."
From an investment perspective, it is not advisable to bet solely on a single quarter’s performance; it is more worthwhile to track management’s latest judgment on next quarter’s guidance, NAND price trends, and supply-demand dynamics. #闪迪存储业务高增长能否延续？ $SNDK $ETH
A bunch of people are laughing that Maji is about to blow up again, the liquidation king. I just casually checked the on-chain data, and this time it's really precarious: the 25x long position only has 68 ETH left, opened at 2313.9, liquidation price 2290.6, with just over 4,000 USDT in the account as margin. Right now ETH is at 2318, just one step away from liquidation, $28 difference.
At the same time, on-chain monitoring shows that the old OG who entered at $0.31 sold another 1500 ETH in the past 8 hours, and an ancient giant whale moved 3000 ETH to exchanges. Some are stubbornly holding on at the liquidation line, while others are offloading.
Everyone is watching to see when he will blow up, but I think this position itself isn't that important; his position size can't make waves. But someone who lost tens of millions down to a few thousand USDT and still dares to open 25x leverage is the truest thermometer of this round of sentiment. When he stops opening positions, it might be more useful than any indicator.
#本周将公布10月FOMC利率决议 AI内存的故事很热，但海力士明天交卷，订单之外还有三道题：货交了吗，利润留下了吗，现金收回来了吗？ 公司10月21日的官方邀请函确认，第三季度业绩将在10月28日韩国时间上午9点前公布，电话会在9点开始，对应北京时间上午8点。截至这篇财报前瞻查证时，不能把市场预测当成已经披露的成绩。 这次围绕HBM4，我先看产品进度。客户认证、量产、实际出货、收入确认，是不同环节。即使管理层谈到订单充足，也要追问交付时间，以及有多少已经进入当季收入。没有单独披露的地方，就保留未知，不能自行算出HBM4贡献了多少利润。 第二看利润质量。高端产品卖得更多，可能改善产品组合；但爬坡成本、良率和新增设备折旧也会影响盈利。因此，收入增长与营业利润率需要一起看。即使整体利润变好，也要分辨是否来自传统内存价格，而非全部归功于HBM4。 第三看现金。经营现金流增加了多少，资本开支又吃掉多少？账面利润增长，如果伴随更重的库存、应收账款或扩产支出，股东能拿到的现金未必同步变厚。 我的判断：这份财报最有价值的，是补齐“订单—交付—利润—现金”的证据。若公司只给出需求乐观表态，缺少可核对的交付与盈利信息，我不会据此上调对HB$BTC's latest round of decline occurred after attempting to rebound near $85,000. The report on October 26 showed Bitcoin's 24-hour trading volume was approximately $17.55 billion and noted that Bitcoin's price once dropped to $79,332. These figures are rolling data and will change over time. Please do not confuse them with subsequent exchange quotes or treat them as closing prices.
This decline happened after three consecutive trading days of inflows and two trading days of outflows in the US stock spot ETF. Farside Investors' daily trading table showed inflows of $526.7 million on October 19, $23.2 million on October 20, and $524.8 million on October 21. Outflows were $635.6 million and $137.9 million on October 22 and 23, respectively, totaling $773.5 million in outflows. Summing the five inflows results in a net inflow of $301.2 million. Upon verification, Farside had an issuer entry missing on October 23, so the displayed total is still subject to adjustment.
Therefore, even with two consecutive days of decline, it does not necessarily mean the week will end with a net outflow. Funds ultimately still rose, but demand reversed on the day, and Bitcoin failed to hold the high level. Previous reports on three days of inflows had already indicated the market pulled back after approaching $85,000. New ETF funds alone cannot sustain Bitcoin at that price level. ⚠️⚠️⚠️
Brothers, the FOMC meeting is happening today and tomorrow, with the decision announced at 2 AM Thursday and Powell's press conference at 2:30 AM. Let me highlight the key points for you:
First, there's a 90% chance of no rate hike, which is already priced in, so no hike isn't really a positive surprise.
Second, the focus is on his stance for December. CME currently shows a 69% chance of a rate hike in December; if he even hints "December not ruled out," Bitcoin will drop to a lower level immediately.
Third, Brent crude is still above 110, the 10-year US Treasury yield is over 5%, inflation can't be contained, so Powell has no reason to be dovish.
I think the most likely scenario is: no rate hike but hawkish tone, Bitcoin will first drop sharply then rebound. Don't bet on direction early; wait until it's confirmed before making moves. Reduce positions to the minimum before Wednesday.
#本周将公布10月FOMC利率决议
#SK海力士Q3财报在即，HBM4能否提振盈利？
#谷歌云业务增长82%，AI投入回报如何？ A recent signal worth noting: long liquidations are clearly more than shorts, and the funding rate is still positive, indicating that those chasing longs are still paying, but the price can't rise. This kind of structure tends to first trigger a wave of long liquidations before any rebound. Volatility has also been squeezed from about 13% down to around 6%, tightening more and more, so the breakout will happen quickly. Therefore, I now calculate stop losses and position sizes for every trade in advance, exiting with a 1%-2% loss per trade, not holding losing positions. Staying alive is more important than guessing right once.
This is my personal opinion and does not constitute investment advice.
#TradingInsights #RiskManagement First consider the worst-case scenario: if this incoming transfer is gradually liquidated, short-term absorption might be tight. According to monitoring, an early address that subscribed to 100,000 ETH in 2015 transferred 3,000 ETH to exchanges in the past 24 hours, still holding 47,000 ETH on-chain.
⚠️ I tend to be cautious but do not equate the transfer directly with a dump: whether it has been traded is still unclear, and how the remaining holdings will be handled is also unknown. Ordinary people should not treat whale movements as trading signals; first, list how much loss you can bear per trade; I will continue to watch for actual trading evidence.【市场实时行情】$BTC 在$80,000附近窄幅整理，$ETH $2,317附近微跌，主流币在宏观窗口前保持谨慎。而AI Agent赛道这边，气氛完全是另一番景象。 👇 先把今天这两条平行线摆在一起： 📌 蚂蚁灵光2.0：10月27日发布，核心升级是跨会话记忆、长程任务调度和工具连接能力。依托独立云主机，AI在用户退出对话后继续运行，主动识别需求、发起监测、在关键节点提醒。已接入支付宝账单、会员、出行，以及钉钉、飞书、邮箱、米家等服务能力，管理范围延伸到跨平台工作流和智能家居控制。 📉 链上AI Agent代币：总市值从2025年1月的157亿美元峰值，跌到如今约36.5亿美元，跌幅近80%。Bittensor（$TAO）报$250.50，24h跌1.1%；NEAR报$4.48，跌1.5%；Fetch.ai（$FET）报$0.174，跌0.5%。 这两条线放在一起，问的是一个币圈很难绕开的问题：如果大厂已经把能管事的Agent做成了日常基础设施，链上Agent代币的溢价到底该建立在什么之上？ 答案藏在“产品”和“代币”之间的那道裂缝里。Sparsity Data的研报指出，开源Traditional financial giant Franklin Templeton has come up with a new twist: from now on, when you buy its stock ETFs, dividends won't be paid out in cash anymore; instead, the money will be used to buy $BTC for you
They just applied for two new funds with a simple strategy: over 95% of the money still buys U.S. stocks, while about 5% of the portfolio is allocated to buying BTC
Previously, when funds received stock dividends, they would reinvest them in stocks. But for these two funds, as soon as a dividend is paid, the next day the dividend money is directly used to buy BTC on the market
So traditional stock investors don't need to register on crypto exchanges; just hold this stock fund quarterly, and it will automatically "dollar-cost average" $BTC for you using dividends
Industry analysis: the fund is currently small in scale, so the amount of Bitcoin bought with dividends is limited and won't have a significant direct impact on the market
But the important thing is its strong symbolic meaning: it’s like a traditional financial stock account gradually turning into an automatic buying pipeline for the crypto world
#富兰克林邓普顿推出美股股息买BTC新型ETF 2026-10-27:
BTC and ETH both contracted with reduced volume, with the FOMC announcement about 40 hours away. Price changes are minimal, but funding rates have both risen to the 85%+ percentile over 90 days, indicating a clear increase in leverage crowding.
Continuing to wait and watch today, no early directional bets.
Compared to yesterday, the real focus is on derivatives:
BTC: Price −0.33%, funding rate percentile 74.4 → 85.6, OI +1.95%
ETH: Price −0.41%, funding rate percentile 51.5 → 86.7, OI +2.28%
Price down, OI up, high funding rates indicate new leveraged positions are accumulating.
ETH is deteriorating faster.
BTC | Current price around 80,150.
4H Bollinger Band width has compressed to the 10.5% percentile of the past 200 candles, indicating extremely low volatility.
Core range:
79,220 — 80,600
No trading as long as 4H remains within this range.
Break below 79,220: watch 78,400 → 78,000.
Break above 80,600: considered an upward breakout, but current long R:R is insufficient, so temporarily abandoned.
The biggest message from BTC is not direction but that volatility is about to release.
ETH | Weaker
Current price about 2,325, daily EMA200 around 2,319, only +0.24% above.
Meanwhile:
Funding rate percentile 51.5 → 86.7
25d skew 2.47 → 4.12
Retail long-short ratio 3.39
ETH/BTC near recent lows
ETH’s structure, sentiment, and positions are weaker than BTC.
The key level to watch:
Whether the daily closes below 2,319.
If it breaks down effectively, consider:
Entry 2,310
Stop loss 2,352
Targets 2,250 / 2,200
R:R about 2.6
BTC vs ETH:
BTC is closer to a “turning point”;
ETH is closer to a “breakdown.”
If both trigger downward simultaneously, I would prioritize ETH because its daily structure is more cleanly broken.
The biggest upcoming variables:
October 29, 02:00 FOMC
October 29, 02:30 Powell press conference
So today’s strategy remains simple:
No new positions, no predictions, wait for a breakout.
Low volatility + high leverage + FOMC approaching.
A major move may be coming soon, but there is no clear direction yet. The heavy bombshell hanging over the US stock market this week: the Federal Reserve interest rate decision + Powell's press conference, GDP, and core PCE inflation data.
The market is now most looking forward to a rate cut #本周将公布10月FOMC利率决议
The 2:00 AM rate decision + 2:30 AM press conference:
The rate itself is unlikely to change much this time; the focus is on what Powell says. If he signals that inflation is easing and there is room for rate cuts later, growth stocks like the Nasdaq and storage will be directly lifted by capital; if he continues to speak tough, saying inflation is still volatile and rate cuts will be delayed, high-valuation stocks like $SOXL and $SNDK will immediately come under pressure and sell off.
The 8:30 PM GDP and core PCE:
GDP reflects the economy's heat; if GDP is too strong, the market will interpret it as the economy holding up well, so there is no rush to cut rates;
Core PCE is the Fed's most important inflation indicator! If PCE rises, rate cut expectations cool down immediately, and the US tech sector is prone to plunge; if PCE is 0.2% lower than the previous value, rate cut expectations heat up, giving growth stocks upward momentum.
SK Hynix's earnings report just came out; storage was originally in a tug-of-war for reversal, but whether it can have a big rally depends on the Federal Reserve.
Earnings are the sector's own fundamentals, while the rate decision and inflation data are the macro gatekeepers for the entire market.
Even if SK Hynix's earnings are good, as long as Powell's speech is hawkish, the rebound can easily be interrupted; conversely, as long as this round of inflation data weakens and the Fed signals dovishness, US stocks, semiconductors, and risk assets like BTC will see a resonant rally.In the futures market, even if it's called copper, different grades and delivery locations cannot be mixed for accounting. When I look at computing power contracts, I usually first check the "Delivery Specification" before discussing how large the market can be.
Arthur Hayes proposed in "The Trifecta" to establish a computing power trading market for AI agents through the Flop Network, paid with $FLOP, and to develop forward, futures, and perpetual contracts based on the spot price of computing power. This is a concept and should not be considered as already implemented.
I think this matter is more important for hedging value between computing power supply and demand than for stimulating speculative sentiment. But computing power is not a standard commodity; hardware and task compatibility affect actual output. If the contract does not match the actual rental cost, hedging could instead leave exposure. This is my speculation; the key lies in the contract details. Are there any friends who have also transitioned from traditional markets? What specifications do you think should be clearly written for computing power delivery?Is Dogecoin $DOGE worth holding long-term in this bull market???
I've always thought that the cryptocurrencies truly worth holding long-term should include both BTC and DOGE.
BTC is more like digital gold, with its core value being scarcity and store of value. But from the perspective of Satoshi Nakamoto's original concept of peer-to-peer electronic cash, DOGE aligns more closely with that direction.
Many people attack DOGE for its unlimited issuance, but they overlook a key point: a fixed issuance of about 5 billion coins per year does not mean the inflation rate remains constant forever. As the total circulating supply grows, the same amount of new issuance represents a decreasing proportion. Currently, the annual inflation rate is about 3%, and it will continue to decrease in the future.
Currency doesn't necessarily have to be absolutely deflationary. DOGE's design leans more towards circulation and payment, rather than having everyone lock up coins as scarce assets for hoarding.
DOGE itself is an independent PoW public chain, requiring miners to invest computing power and electricity to maintain the network. Its underlying mechanism is different from tokens issued directly on existing public chains. As BTC becomes increasingly financialized and more focused on value storage, it has indeed drifted from the original electronic cash vision. Coupled with the long-standing community disputes over its direction, it makes one wonder: is BTC's development path today still what Satoshi Nakamoto originally envisioned? #本周将公布10月FOMC利率决议 October FOMC in one sentence: The baseline scenario is a “hawkish pause” — no rate hike, but absolutely no rate cut, with focus on whether to leave the door open for a December hike
This ties in with your earlier look at SK Hynix HBM memory chips: AI high-growth sectors (HBM, computing chips) are very sensitive to long-term US Treasury yields, and this FOMC is a major macro variable before SKHY earnings
Time: Beijing time, October 29 at 2:00 AM decision, 2:30 AM Powell press conference; ⚠️ This meeting will not update the dot plot or SEP economic forecasts, so information is concentrated in the statement wording + press conference Q&A
1. Current market consensus (CME FedWatch)
- Probability of holding rates steady in October (3.75%-4.00%) is about 90%, with only about 10% chance of a 25bp hike, and almost no pricing for a rate cut
- The real market game is whether there will be another 25bp hike in December; futures price the probability of a December hike close to 70%
- Background: a hike was just made in September; inflation stickiness + rising oil prices are hawkish reasons; weaker nonfarm payrolls and tech credit disruptions are reasons to pause consecutive hikes
2. Three scenario simulations (corresponding to the market, including storage stocks like SK Hynix)
✅ Scenario 1: Neutral baseline (highest probability) — pause rate hikes, keep wording that “further tightening may still be possible”
Powell’s stance: no commitment to a December hike, nor closing the door on hikes, continuing data dependence. ETH 现价约 2305 美元，24 小时跌 1.37%。2300 是眼下最后一道整数防线，盘中最低已经摸到附近。更关键的是以太坊 ETF：单日净流出 6960 枚 ETH，七天累计流出 85801 枚，价值接近 2 亿美元。链上还有远古巨鲸在动，2015 年 ICO 那批地址近 24 小时又往交易所转了 3000 枚 ETH。价格在 2300 上方撑着，但卖盘是从 ETF 和巨鲸两个方向同时来的。 SOL 现价约 100.35 美元，24 小时跌 1.22%。100 这个位置，之前几次盘中穿一下还能收回来，这次不一样：链上数据显示，三个疑似同一巨鲸的钱包在过去三天转移了 482,228 枚 SOL，价值约 4917 万美元，其中 171,430 枚已经存进了交易所。不是挂单，是已经摆上货架。SOL 的清算数据也很有意思，24 小时爆仓 655 万美元里，79% 是多单。多头在被推着走，不是主动进场。 XRP 现价约 1.29 到 1.30 美元附近，24 小时看着微涨。但清算数据拆开看：XRP 的 24 小时爆仓里，87% 是多单。这意味着押涨的人在被清算，而价格勉强横住，靠的不是新$ONE is tightly controlled, continuously hitting new lows.#Robinhood交易业务增长能否延续？ Inflows look good, trading is coolTrump's approval rating is 32%, and the Democrats lead by 11 percentage points in the generic congressional ballot.
The numbers themselves are not fatal. What is fatal is that the Democrats have already sharpened their knives in advance.
Crypto companies associated with the Trump family—WLFI, 1789 Capital, Polymarket, Kalshi, plus Zhao Changpeng—are all preemptively hiring lawyers and lobbyists, some specifically connected to the Democratic Party.
When have you ever seen a "legal and compliant" industry collectively hire lawyers before an election?
Let's start with the $500 million for WLFI.
On January 17, 2025, four days before Trump's inauguration, Aryam Investment, owned by Sheikh Tahnoon of the Abu Dhabi royal family, signed an agreement to acquire 49% of WLFI for $500 million. Of the initial payment, $187 million went directly to Trump family entities.
Who is Tahnoon? The UAE National Security Advisor, managing assets over $1.3 trillion, nicknamed the "Spy Prince"—he has been pushing the U.S. to approve the export of tightly controlled AI chips.
A person who needs U.S. government approval to export chips gave nearly $200 million to the president's family project four days before the inauguration. The Democrats have singled out this deal as the core of their investigation.
Now about CZ's pardon.
Seven Democratic senators have already sent a letter to the Department of Justice demanding an explanation for why Trump pardoned CZ. The letter states: this pardon sends a signal to the crypto industry that "you can commit crimes with impunity."
CZ has admitted to violating the U.S. Bank Secrecy Act. Binance also helped develop stablecoins for a company partially owned by the president.
The Democrats' logic is straightforward: you pardoned a person who pleaded guilty, and his company has business dealings with your family.
But what the market really overlooks is WLFI's own data.
The governance token WLFI fell from $0.45 in August 2025 to $0.05 in June 2026, a drop of 88%. On October 20, WLFI dropped nearly 6% again, closing at $0.0538.
On the same day, WLFI's issued stablecoin USD1 saw $401 million redeemed within two weeks, including $200 million from an unidentified wallet, transferred out in batches of $300,000 to $500,000.
The stock price drop can be explained as "market volatility." But an unidentified large holder withdrawing $200 million in two weeks cannot be explained.
Bloomberg itself admits there are obstacles to the investigation: foreign evidence is inaccessible, and the DOJ may not cooperate.
But the impact of the investigation never requires a "conviction."
A subpoena, a hearing, a media leak—these are enough to overturn the valuation models of WLFI and BNB. The market fears not the "outcome" but the "process"—and this process will last for months, with every tug-of-war amplifying emotions.
The Democrats do not need a new law to crack down on crypto.
A subpoena and a hearing are enough to rewrite the valuation models of WLFI and BNB.
$BTC $WLFI $BNB Arthur Hayes proposed in "The Trifecta" to establish a computing power trading market for AI agents through the Flop Network, using FLOP tokens for payment, and to develop forwards, futures, and perpetual contracts based on spot computing power prices. He also pointed out that there is currently no unified global benchmark for computing power pricing.
I think this is an important infrastructure concept, but the short-term bullish sentiment may be overestimated. Without a unified benchmark, spot prices are hard to stabilize; when spot liquidity is insufficient, derivatives will amplify both volatility and pricing deviations, which affects ordinary users the most.
📌 Here's a leverage tip for newcomers: assuming the computing power price is 100 and you go 10x long, ignoring maintenance margin, a price drop of about 10% approaches liquidation. If the quoted price itself is not transparent, the liquidation point becomes even harder to determine. Next, I will focus on whether spot computing power can form a reliable price, then look at derivatives.How much has the prediction market actually grown this year? Someone dug up a set of data: from January 1 to October 25, the cumulative number of transactions on a certain prediction platform surged from 241,600 to 80,584,900, a 334-fold increase. It multiplied more than three hundred times.
This growth rate is not driven by a single coin pump, but by the repeated emergence of demand itself. After the 5-minute BTC price fluctuations end, there are still 15-minute US stock market ups and downs; after a Dota match, there’s LOL, and after LOL, there’s the next World Championship; football happens every week, and beyond football, there’s NBA, football, weather, interest rate cuts, earnings reports, gold, and crude oil. There are 365 days in a year, and almost every day there are new topics to bet on.
What’s more crucial is the threshold. Opening a contract requires understanding margin, liquidation, and funding rates; the prediction market only requires judging right or wrong, a binary outcome—if you win, you win. If you talk to someone who has played LOL for ten years about chains, wallets, and airdrops, they’ll lose focus in three sentences; but if you ask them who will win the championship this year, they can talk with you all night.
This might be the real entry point for the prediction market: not educating traders better, but directly turning ordinary people’s interests into liquidity.
The risks are also obvious. Odds are determined by the pool; unpopular markets have very thin depth, and a single large order can cause price slippage; settlement rules, oracle disputes, and regulatory standards are all variables. Good-looking data doesn’t mean every market can make money.
But the track direction has already emerged. From 241,000 transactions to 80 million, this is not a prediction, it’s something that has already happened.A trader's lament over a 10x leveraged $DOOD short and a $2,500 loss captures something more useful than a single bad trade: the mechanics of a crowded short into a market that refuses to cooperate. Negative funding rates, the detail buried in the complaint, are the tell. When shorts outnumber longs heavily enough that they pay to hold positions, the market is leaning one way, and the cost of staying there compounds daily. The $ETH position tells the same story in slower motion. Two short entrieA banking license can't buy a single price surge: What to really watch after Ripple gets final OCC approval
On October 26, 2026, the U.S. Office of the Comptroller of the Currency (OCC) gave final approval to Ripple National Trust Bank, allowing it to manage RLUSD reserve assets and serve as collateral custodian. This marks the first time a crypto company has truly obtained federal-level banking entity status, with industry significance outweighing price impact. This article breaks down the four layers of transmission and three points of attenuation from "license" to "token price" in this news, combined with the current BTC oscillation between $80,000 and $87,000 and the market environment where "Uptober" has failed for the second consecutive year. It provides a verifiable observation checklist for the next 1–4 quarters and three scenario simulations to help turn a "positive" into a trackable and falsifiable basis for judgment. #SK海力士Q3财报在即，HBM4能否提振盈利？ $XRP #Meta启动350亿美元融资，支持AI基建
Meta officially launches a $35 billion equity financing round, with all funds directed towards AI computing power, data centers, and other infrastructure. Chip manufacturers like Nvidia and AMD also plan to participate in the subscription. On the surface, this signals continued investment in the AI sector, but the market has two completely different interpretations.
Positive perspective: The AI arms race shows no signs of stopping. Massive capital expenditures will continue to drive demand for HBM and GPUs, sustaining the semiconductor industry's prosperity, boosting overall risk appetite, and indirectly benefiting crypto assets.
However, the risks are more concerning: Choosing equity issuance for financing indicates that AI's cash burn far exceeds its own cash flow capacity. Large-scale equity issuance will cause dilution, and the enormous financing volume will siphon funds from the market. If AI monetization falls short of expectations later, high-tech stocks at elevated levels are prone to concentrated corrections.
My view: Short-term sentiment is bullish, but avoid blindly chasing highs. High AI capital expenditure growth does not equal performance realization; the biggest risk in this cycle is that returns on computing power investments may fall short of expectations.
Contract strategy: The news is positive but should only be used as a sentiment reference; do not heavily chase the upside. Focus on observing the sustainability of the Nasdaq and chip sectors; if rallies weaken, consider light short positions. Only if funds continue flowing into the AI sector should you consider trend-following long positions, with strict stop-losses throughout.📉 Downward: First test 79,377.7, 79,000 will hold before the meeting.
79,220 is short by 158, 81,070 short by 220, 79,377.7 short by 344, the invalidation line hasn't been touched once.
No one got liquidated: trading volume up 56.37%, liquidations only up 6.29%, open interest 1.17%.
Money is rotating. After showing off 7346% profit, Xuan reduced position to a few thousand dollars, short positions also ready to close.
Veterans are reducing, retail long-short ratio is 69% long (Dr. Bi Kevin's post). Price is stuck.
Perpetual premium -0.0349% discount, but the rate is stuck at 0.01000% benchmark, for two periods.
Discount still at the top, indicating the rate is supported by the benchmark interest rate, not by longs competing.
Today's range is 832 dollars, narrowest in four days, upper bound 80,849.9 is the ceiling.
If 1H close stands above 81,200 or falls below 79,000, consider me wrong, until 10-29 02:00.
Contrary evidence: 10-26 21:00 that 718.5 contracts, previous six bars averaged 130.
It hit 79,455.8 and was bought back, someone is defending the lower bound, not trying to push up.
Before the meeting no one cares about rise or fall, only who pays the rent. Personal review, not advice.
Do you think it will break the lower bound first, or start the meeting directly?
$BTC #OctoberFOMCRateDecisionToBeAnnouncedThisWeekBTC near 80,300, ETH near 2,310. Over the weekend, BTC dipped to 79,278 before rebounding but encountered resistance above 80,500.
Spot buying is weak. Stablecoin reserves dropped to $60.6 billion, a one-year low; Coinbase premium turned negative, indicating a lack of incremental funds. Buy orders near 79,000 are mostly futures orders rather than spot support.
Liquidations continue to favor the bulls. In 24 hours, total network liquidations reached 175 million, with long positions accounting for 106 million. BTC long liquidations were 30.04 million, ETH long liquidations 23.31 million, bulls are still bleeding.
ETF divergence. BTC ETF weekly net inflow was 301 million, but a single-day outflow of 169 million; ETH had a single-day outflow of 16.07 million and a seven-day outflow of 198 million, funds still favor BTC.
Macro: Probability of FOMC holding rates steady is 90.6%, with a 67.5% chance of a rate hike in December. Daily RSI is about 40, approaching oversold, but the weekly trend remains bearish.
Key levels:
BTC support at 79,000-79,500 below, resistance at 80,500-81,000 above.
ETH defense at 2,290-2,300 below, pressure at 2,330-2,345 above.
In short: Spot buying is insufficient, bulls continue to take hits, and breakthroughs are unlikely before the FOMC. Do not chase shorts or bottom-fish; wait for BTC to stabilize at 79,000-79,500 or to break above 80,500 with volume before considering light long positions. Position control is paramount. $BTC and gold both dropped 7%! The so-called safe-haven myth was slapped down by the market this time?
The most counterintuitive scene of 2026 has appeared: Bitcoin and gold have surprisingly become among the worst-performing major assets!
According to the current data, $BTC has fallen 7.3% year-to-date, and gold has dropped 7.5%.
One is called digital gold, the other is regarded as a traditional safe-haven asset, yet both got hit together.
This at least sounds a warning bell for the market: don’t treat asset labels as a price guarantee.
Looking further at $BTC’s own capital flow.
On October 22 and 23, US spot BTC ETFs saw net outflows for two consecutive days, totaling about $773.5 million. The supply from long-term holders decreased by 50,017 coins, while ETF purchases were only about 3,700 coins.
Meanwhile, Strategy holds 848,517 BTC but paused additional purchases after two consecutive weeks of buying.
Putting these signals together, what the market really needs to be wary of is not how much it fell on a certain day, but whether new buying can keep up when selling pressure emerges.
#本周将公布10月FOMC利率决议 #SK海力士Q3财报在即，HBM4能否提振盈利？ #中东原油运输恢复，油价为何仍居高位？ #本周将公布10月FOMC利率决议
This week, the FOMC (October 27–28) is about 90% likely to hold rates steady, so the decision itself is basically a known factor.
The real variable is not whether to raise rates, but the wording of the statement and the press conference by Waller: inflation remains high, and the Federal Reserve must convince the market that the “pause in October” does not mean “tightening is over,” while maintaining the credibility of a rate hike window in December.
For risk assets, the decision is already priced in, but the tail risk of communication missteps is not fully priced.
If the market interprets the statement as “December might not move either,” the repricing of rate expectations will first impact U.S. Treasuries and the dollar, then transmit to tech stocks and crypto assets;
Conversely, if the December expectations are successfully maintained, the short term might actually see a “boot dropping” window.
The focus is on Waller’s words, not the rate numbers. The decision is a known factor; communication is the source of volatility. @OKX星球 @OKX 产品精选 $BTC #Can SanDisk's storage business sustain its high growth?
Old Mo straightforwardly says: The volume certainty is very strong, but the price elasticity is locked by long-term contracts.
Supporting the growth is the NBM long-term contract. Eight agreements lock in a guaranteed minimum revenue of $93.9 billion, covering more than half of the shipments through 2027, with an 80% gross margin guaranteed by a floor price. As long as AI inference demand for flash memory doesn't collapse, this base is basically stable.
But the cracks are here too. The long-term contracts lock both volume and price, so if the spot price continues to surge, SanDisk won't benefit from excess profits. Coupled with market concerns about the cycle peaking and sentiment pressure from players like Changxin, the stock price is unlikely to break out of a pure cyclical steep rise in the short term.
Old Mo summarizes: SanDisk is shifting from a "pure cyclical game" to "cash flow pricing." There is a solid safety cushion on the downside, but don't expect another crazy surge in the short term. At this position, patiently waiting for a pullback and stabilization is much safer than chasing highs.
What do you think about SanDisk's long-term contract model? Let's discuss in the comments. $BTC $ETH $SNDK #SK海力士Q3财报在即，HBM4能否提振盈利？ SK海力士Q3财报前瞻简析（配图话题：HBM4能不能提振盈利）
In short: The main profit driver in Q3 is still HBM3E; HBM4's contribution this quarter is limited, serving more as a growth engine for Q4 and next year; Q3 performance flexibility comes from general DRAM/NAND price increases plus fulfillment of existing HBM long-term contracts.
Market poll in the chart shows 83% betting Yes (earnings beat expectations), 18% betting No, reflecting a generally optimistic consensus among investors.
✅ Positive factors (supporting profits, likely to exceed expectations):
1. Traditional storage price increase benefits realized
TrendForce estimates Q3 DRAM contract prices up 13%~18% QoQ, NAND up 10%~15% QoQ; AI inference drives stronger eSSD demand, with ordinary server and PC storage prices rising simultaneously, representing confirmed revenue growth in Q3 without waiting for HBM4 volume ramp-up.
2. HBM3E remains a cash cow
HBM long-term contracts lock prices and volumes, with high unit prices and very high gross margins; SK Hynix still leads in HBM market share, and HBM3E continues stable shipments in Q3, forming the profit base.
3. HBM4 yield has met standards but is in early ramp-up
HBM4 began small-batch sample deliveries in Q2, with yields approaching the mature HBM3E level; Exchange wallet $BTC has recently surged by 70,000 coins. Are the spot whales about to dump?
Many people were scared by this chart showing a large inflow of BTC into exchanges recently.
In less than a month, over 70,000 BTC spot coins flowed in, coinciding with BTC fluctuating and declining above 80,000. Could a major correction really be starting?
Today, I saw more detailed data from other bloggers (Chart 2).
The light-colored area represents the cumulative inflow of BTC by retail investors into exchanges over the past 30 days.
The dark blue area shows the data of whale inflows of BTC into exchanges.
It’s clear that retail investors have been continuously depositing coins into exchanges this past month.
However, the amount of whale inflows is shrinking and decreasing.
In other words, this wave of large inflows into exchanges is mostly driven by retail investors, not whales.
Whales have not been transferring BTC from cold wallets to exchanges in bulk, preparing for spot selling.
Therefore, the main selling pressure causing the recent decline likely comes from retail investors transferring spot coins to sell + contract leverage stop-loss orders, not whales selling spot.
As long as this situation continues, it’s unlikely to see a spot crash with no support and a cliff-like plunge.
Thus, this ongoing BTC decline looks more like a new round of chip swapping, with retail investors taking profits and whales remaining restrained and accumulating.
But this only indicates that the risk of whales chain-selling spot coins is currently low, temporarily ruling out the worst-case scenario of a panic spot sell-off.
For the market to rebound, signals of reduced retail selling pressure and whales withdrawing coins to accumulate are still needed. On October 27, the $BTC 80,000 level was repeatedly contested. What's truly interesting is the position divergence:
The long position ratio across all market accounts is 69.5%, the long position ratio among large accounts is 70.5%, and the long position ratio held by large accounts is 66.0%!
But the smart money's long position ratio is only 21.7%, with a buy ratio of 49.3%!
To put it plainly: retail investors have a high long position ratio, but smart money is not following.
For long entry recommendations, place orders around 78,700. The liquidation intensity is nearly 900 million dollars. If you get a chance to catch it in the short term, there will be a rebound. #Anthropic officially files for IPO, potentially setting a new valuation record
"Anthropic IPO Valuation Breaks Records"
Anthropic has officially submitted its IPO application, aiming to break valuation records.
On the surface, the Claude model's usage is surging, with Amazon and Google each investing over ten billion dollars. However, behind closed doors, the costs of computing power, cutting-edge chips, and cloud cluster rentals are consuming huge amounts of cash every quarter. Subscription revenue alone at this stage is far from sufficient.
The management has chosen to rush to the public market primarily to replenish R&D resources. The most critical upcoming focus will be the quarterly gross profit and computing power expenses disclosed in the first public prospectus. $ANTHROPIC officially files for IPOBTC holds at 80,000, better not chase before the FOMC night.
OKX current price is about 80,006, 24h high 80,850, low 79,378, with a fluctuation of less than a thousand points.
This morning, the liquidation zone reported about $174 million liquidated across the network in 24h, with long positions accounting for about 60.7%.
Farside: US spot BTC ETF had a net outflow of about $636 million on 10/22, and another outflow of about $138 million on 10/23; the FOMC meeting is today and tomorrow.
I think this is a typical pre-meeting low-volume stalemate, not a breakout signal.
Institutions just had outflows, the price is grinding near 80,000, chasing hard has an average success rate.
What I will do: observe and not chase. Wait for a firm hold and recovery above 80,850; if it breaks below 79,300, this 80,000 defense is considered failed.
Are you more worried about a dump after the meeting, or will you wait for the low volume to end before making a move?
$BTC $ETH $SOL
#OctoberFOMCRateDecisionToBeAnnouncedThisWeek #MiddleEastOilTransportResumesWhyAreOilPricesStillHigh?今日判断 今天有两条进展值得记下：代币化美股支持X Layer充提，预测市场也拿到了更多传统机构合作。 我对X Layer实际应用的判断更积极了，但目前还没有新增OKB质押需求得到确认。ETF数据也没齐，机构买盘是否恢复，先不下结论。 1、ETF：数据没齐，暂不判断资金回流 10月26日现货ETF（交易所交易基金）数据尚未归集完整，IBIT、FBTC、ETHA等主要基金仍缺数据。 此前23日，BTC净流出1.379亿美元，ETH小幅流入；SOL当日数据也未归集完整。目前还不足以判断三类资产的机构买盘是否转强。 2、代币化美股分批开放，X Layer承接资产流转 10月26日起，xMARA、xMCD、xMRK等首批代币化股票按计划分批进入OKX现货市场，支持通过X Layer充值、提现。 这给X Layer带来了具体的资产使用场景。现货撮合在OKX完成，X Layer承接的是充提和链上转移需求。 这项业务进展值得重视，但尚未确认这些现货交易必须质押OKB，或交易手续费会回流OKB。 Exchange OS仍在公测，外部部署和具体质押规则没有新公布，原定第三季度开放部署仍标注为“即将推出”The hottest coin today is $Q (Quack AI), ranked first in CoinGecko's hot search. As of 10:02 Beijing time, Q is priced at $0.1975, up 254% in 24h, with a trading volume of about $270 million and a market cap reaching approximately $930 million. It just hit a new all-time high of 0.2184 at 08:36 this morning.
This round of Q has been very volatile: after a surge in mid-October, it dropped 57% over six days until the 25th, then yesterday it quickly bounced back from around 0.057. So far, no new official positive news from the project team has been found; the narrative remains AI Agent payments (x402 gas-free signature payments).
My view: this kind of movement looks more like a battle between capital and contracts rather than a fundamental change. The trading volume is nearly 30% of the market cap, indicating very high turnover; about 241 million tokens will unlock on November 2. Rather than chasing the rally, it's better to understand the rhythm. I will focus on the support after pullbacks and the changes in holdings before and after the unlock.This week's tech stock earnings reports are a major test; the market is no longer focused on profits but mainly on future expenditures. ANT will be listed in 2 weeks. $BTC strong demand + strong absorption, the only drawback is not during the US market hours, but I think it's very, very good. I added to my position at 79600 📅 October 27, 2026 · On the eve of the Federal Reserve decision
🚨 The FOMC meeting opens today,
BTC is currently at $80,033, down 0.74% in 24 hours, ETH down 1.49%, SOL down 1.38%. In the past 24 hours, the total network liquidation reached $135 million, with long positions accounting for $91.74 million — leveraged longs are rushing to exit ahead of the Fed decision. Market expectations for a rate cut have been slashed to less than 10%, making a "no change" almost a sure bet; the real variable is every word from Waller at the press conference.
Strategy has paused BTC purchases for the first time in five weeks, with $772 million flowing out of US spot ETFs in two days. The two most stable marginal buyers have both stopped.
The 80,000 level was lost before the decision; the real direction will have to wait for the press conference early Thursday Beijing time. #Strategy财报前瞻：BTC账面收益大幅反弹 #SK海力士Q3财报在即，HBM4能否提振盈利？ Order Book Strength Ranking
5-minute median slippage, estimated by order book
$IRYS large buy cost decreases, sell cost increases. Simulated slippage for 100k scale buy/sell changes from 1.70%/1.38% to 1.35%/1.55%. The rising side shows reduced volume, the falling side remains stable.
$DOOD large buy cost increases, small buy cost almost unchanged. Simulated 100k scale buy slippage rises from 1.12% to 1.33%. 10k buy remains about 0.16%. When cost worsens, adjust order size, not directional view.
$CAP slippage is higher for the same buy amount. Simulated 10k scale buy/sell slippage is 0.23%/0.16%, 100k is 1.28%/0.56%. Large orders on the expensive side need to be split.Last night, the OCC issued the final license to Ripple's National Trust Bank. This is not a commercial bank that can accept deposits and make loans, but a federally regulated trust bank: managing RLUSD reserves, acting as collateral trustee, and providing institutional digital asset custody. The official press release states that RLUSD has a market value of about $2.5 billion, and the issuance side is still managed by Standard Custody, regulated by the State of New York. CoinGecko spot XRP is about $1.30 (snapshot around 09:58 Beijing time).
It is more likely a compliance foundation being completed, not an immediate doubling of payment volume. The opposing view will focus on the trust bank not accepting deposits or making loans, and the license itself does not bring transaction fees. But once the federal custody shell is established, institutions will have one less regulatory identity question when purchasing.
Let's first observe the three weeks after opening: whether the custody client list publicly shows a few verifiable names. If the list is blank, or RLUSD's market value remains almost unchanged compared to before opening, then this is just a regulatory progress update, not a sign that funds have already entered on a large scale.
$XRP
#Ripple #TrustBank Bitcoin is currently stuck at 79,000, just one step away from 80,000, but that step just can't be crossed. Every time the price tries to push up, the trading volume doesn't keep up; the momentum to break through is clearly insufficient.
There's a detail on-chain: whales have been accumulating coins offline recently, not dumping them on exchanges, so selling pressure is indeed low. But the problem is that the existing chips are just being exchanged back and forth among insiders. No matter how much turnover there is, it can't push a breakthrough because no new money is coming in. An integer barrier like 80,000 can't be sustained by old money just passing coins around.
ETFs saw a net outflow of $630 million on Thursday alone, and during this week only BlackRock is still supporting net inflows—one player can't carry the whole load. This is the most direct signal from the demand side, indicating that incremental funds are on the sidelines.
Adding to that, the Federal Reserve is very likely to hold steady on interest rates, and no one wants to add positions before the decision is announced. With both macro and capital flows weak, I don't see 80,000 being reached in the short term.
If a real bullish turn is to happen, it's better to wait until the interest rate decision is out and ETFs resume net inflows. Chasing longs at this level isn't cost-effective.
The range between 79,000 and 80,000 will likely keep grinding back and forth for a while; this kind of low-volume sideways movement is very frustrating. If it breaks below 77,000 afterward, the downward momentum will accelerate—don't rush to bottom-fish.