Orbit Post Sitemap
$BTC The Federal Reserve's July 30 decision to "stop openly and hawk in secret" has clearly signaled short-term bearish pressure for the crypto market and will continue to exert pressure in the medium term. 🎯 What exactly is the "Bright Halt and Hidden Eagle"? Nominally, the Fed has kept the rate unchanged at 3.5%-3.75% for the fifth consecutive time, in line with market expectations. But the "eagle" has three aspects: Rare internal division: The decision vote was 9 to 3, with three regional Fed chairs voting against and advocating for an immediate 25 basis point rate hike. This is the first time since 2016 that three unanimous opposing votes have appeared, signaling a strong tightening policy. Hawkish Statements: Chairman Wash emphasized that the 2% inflation target "leaves no room for flexibility" and announced a substantial withdrawal from the forward-looking guidance, no longer signaling interest rate paths in advance. Soaring expectations for a September rate hike: After the meeting, the market believed the probability of keeping rates unchanged in September was only 36.8%, while the probability of a 25 basis point hike rose to 63.2%. 📉 Multidimensional Impact on the Crypto Market · Short-term price pressure but relatively resilient decline: after the decision was announced, BTC and ETH both fell about 1%. Although they fell in tandem, compared to the Dow's plunge of over 1,100 points (-2.19%) and the Nasdaq's drop of 1.74%, the decline in crypto assets was relatively smaller. Rising expectations of "liquidity drying up": For risk assets dependent on liquidity, prolonged high interest rates mean higher funding costs and suppressed risk appetite. Chairman Wash even expressed a desire to rely on spontaneous market rate increases to tighten financial conditions, which has intensified market liquidity concerns你们认可吗？BTC的底部由矿工关机价格决定，不是画线或指标。
我们一起来看看其中的逻辑：
当BTC低于挖矿成本时，矿工被迫关机→算力下降→难度下调→幸存矿工成本降低→价格见底。
这是BTC自带的底部机制，历史上每次终极底部都发生在这个链条末端。
历史上BTC底部从未跌破矿工全面关机价，2022年底部15,476时关机价12,000-14,000美元，底部在关机价上方10-20%。
2026年矿工关机价核心区间在42,000-45,000美元。
按此规律，42,000-45,000美元关机价对应底部约48,000-54,000美元。
核心判断： 矿工关机价是BTC最硬的底部锚，历史100%有效。同一天，有人认为ETH可能跌到500美元，也有人坚信它会冲上8000美元。 两个目标相差16倍，看起来像是多空双方中必有一方失去理智。 但如果把价格拆开看，就会发现他们交易的根本不是同一个世界： - 500美元对应的是宏观流动性重新收紧、比特币破位、ETH继续失去相对强度； - 8000美元对应的是比特币进入大牛市、ETH/BTC明显反转、机构资金和以太坊生态同时爆发。 所以，ETH年底到底能到多少，不能只盯着一根K线猜数字。 真正的问题是：未来五个月，市场会进入哪一种资金环境？ 截至2026年7月29日截稿，ETH价格约为1875美元，流通量约1.2068亿枚，对应流通市值约2260亿美元；同期ETH/BTC约为0.0295。 我的核心判断是： > 2026年年底，ETH最高概率区间为1800—2600美元；概率加权价格中枢约为2400美元。 弱势情况下可能回到900—1300美元；极端风险情景为500—800美元；强牛市可看3200—4500美元；6000—8000美元只适合列入极端狂热路径。 需要说明的是，2400美元是情景概率加权后的中枢，不是年底必然收盘价，更不是保证能够达到#停火48小时告吹，美伊边打边谈
$CELO
商务部回怼“脱钩”指控，措辞硬，但没加码。市场第一反应：CELO在0.06-0.07晃了一下，+1.73%，短线资金在试探，但没人敢押注。
真正的信号在风投数据：7月仅150家机构参与加密融资，2020年11月以来最低，比2022年峰值跌了87%。央行逆回购放水，但加密没沾到光。
资金怎么理解：宏观风险还在，但短期没新利空。风投数据更关键——一级市场在缩，钱从冒险转向避险。BTC稳，大盘就不会崩；ETH得跟上，风险偏好才算修复；SOL有弹性，但得看到主动买入。CELO这波反弹缺链上或叙事支撑，更多是情绪跟风。
观察条件：1. BTC放量站回62,000以上，ETH同步走强，CELO的0.07可能变短期支撑。2. 风投数据继续恶化，BTC缩量横盘，CELO的0.06会被再测一次，别急着抄底。
风险提醒：风投数据滞后，但趋势明确——资金在收缩。宏观摩擦加流动性收紧，CELO这类小币波动会放大，但方向不确定。没有确定性行情，只有概率和仓位管理。AI还在疯狂烧钱，但华尔街的耐心值已经快用完了。这大概是本轮美股七巨头财报季最真实的写照。 截至7月29日，Alphabet、微软、Meta和特斯拉已经交卷。结果很鲜明：市场不再满足于听“AI未来会怎样”的故事，而是直接翻开资产负债表，追问“钱到底赚回来没有”。目前来看，只有微软暂时经住了考验。$MSFT 微软是少数赢家。 它的逻辑很简单：有收入，有路径。第四财季营收900亿美元，Azure云增长43%，Microsoft 365 Copilot付费用户突破3000万。虽然一个季度资本开支也砸了约410亿美元，但Azure和Copilot已经让人看到了钱在往回流。市场愿意为看得见的增长买单。$GOOGL 但其他人就没这么幸运了。 Alphabet率先拉响了警报。谷歌云收入确实猛增82%，但单季资本开支飙到449亿美元，导致自由现金流转负，达到-59亿美元。更让人担忧的是，公司还把2026年全年资本开支指引上调到最高2050亿美元。云业务再强，也抵不过现金流恶化的恐慌，股价自然承压。$TSLA Meta和特斯拉则被资金挑剔得更厉害。Meta营收增长28%，广告依然能打，但成本和费用比特币深陷“慢熊”困局：加息阴影、ETF失血与“囤币信仰”的崩塌
过去一周，加密市场再次被剧烈波动笼罩。比特币价格在 7 月 28 日全线跳水中一度跌破 6.3 万美元，超 16 万人爆仓，爆仓金额高达 6.86 亿美元。截至 7 月 30 日，比特币虽暂稳于 6.4 万美元附近，但较 2025 年 10 月近 12.6 万美元的历史高点已回调约 48%。
这并非一次由单一黑天鹅事件引发的急跌，而是一场由宏观紧缩、机构撤退与叙事崩塌共振的“慢速需求衰退型熊市”。
1. 宏观“加息阴影”重压，风险资产首当其冲
市场对美联储政策转向的担忧是此轮波动的核心导火索。本周议息会议前，利率期货显示的加息概率一度升至 30% 至 35%，政策不确定性显著升温。中东冲突推高油价，进一步加剧通胀尾部风险，迫使市场重新评估“更高更久”的利率前景。
作为不产生现金流的资产，比特币在实际利率上升时机会成本激增。美元走强与国债收益率攀升，正同步压缩现货估值与杠杆多头的承载能力，使其越来越像一只高贝塔的科技股而非“数字黄金”。
2. 机构资金“急转弯”，ETF 七连流入终结
资金面的急剧逆转是本轮调整的直接推手。7 月 23 日至 24 日，美国现货比特币 ETF 合计流出超 4.65 亿美元，终结了连续七日的净流入趋势。其中仅 27 日单日便再流出超 2.25 亿美元，彻底扭转了此前一周的温和净流入态势。
机构信心的脆弱暴露无遗。尽管《CLARITY 法案》曾带来监管利好预期，但在加息担忧面前，任何短期利好都显得苍白无力。
3. 微观结构恶化：“囤币旗手”倒戈与矿企转型
更令人担忧的是行业微观结构的显著恶化。
· 最大多头“认输”：此前五年“只买不卖”的最大上市持币公司 Strategy，如今因浮亏超百亿美元而被迫松口，授权出售最多 12.5 亿美元比特币以补充流动性。其叙事冲击远大于实际抛压，彻底动摇了“囤币飞轮”的信念基石。
· 矿工转型求生：减半后挖矿成本高企（部分矿企成本达 7.8 万美元），叠加币价低迷，矿企被迫抛售比特币并转型 AI 数据中心以求生存，进一步增加了市场的潜在供给。
4. 后市展望：最后的防线与真正的反转信号
当前比特币正与 50 日移动均线（约 6.33 万美元）展开关键博弈，此处是最后的防线；若失守，6.2 万美元将面临直接威胁。
CryptoQuant 分析师警告，当前现货需求疲弱（30 天需求跌至负 17 万枚 BTC），价格仅靠衍生品杠杆支撑，结构极为脆弱；一旦出现新的现货卖压，极易引发多头强制平仓。
真正的趋势反转或许需要等待以下信号：ETF 恢复连续数周的大额净流入、加息预期见到明确顶部、以及 Strategy 等大型持币者重新成为净买家。在此之前，每一次反弹可能都只是高波动区间内的流动性修复，而非牛市的确认。#$ETH 说一下美股什么时候止跌 储存到底还能抄底不？ 昨天的尾盘开始放量下跌，从消息面看是通胀加剧，又开始讲加息预期，实际上还是美股3-6月的获利和AI投入的不确定性，市场在面对不确定性的时候就会无脑卖出。 而目前看并没有止跌的趋势，纳斯达克跌多涨少，主要靠七姐妹支撑，原来的逻辑是硬科技跌，软件涨，英伟达涨，苹果涨，谷歌涨，现在明显卖出硬科技后，资金也放弃七姐妹了 这种情况什么时候好转，这种好转的最终形态应该是英伟达开始止跌，走稳，甚至走出强趋势，这种既需要英伟达有炸裂的财报以外还需要技术突破和AI新一轮的缺口，英伟达上涨，半导体才会回暖，储存才会止跌反弹，美股才会赢来反弹。 而现在，并不需要更多关注每个人的建议，我刚买美股的时候，买完就疫情了，当时看完全站岗了山顶，事后去看，那个地方只是一个小坑 做现货和定投的好处就是能熬的住，美股的波动大，长期看，qqq仍然平均收益率20%，我并不认为AI就此一蹶不振，也不认为现在会买到纳斯达克百年最高点。 市场确实需要降温，降温后，理性的自己才能买到谷歌，才能买到SMH，当时买入持有的价格也不是为了涨一点就卖，跌一点就受不了。 太多了想在美股上一夜暴富了闪迪SNDK：本轮超级牛熊完整复盘（精简干货）
本轮闪迪AI存储超级牛市，是2025—2026年美股最强周期行情之一，完整高低点脉络彻底清晰！
一、本轮牛市真正启动底
大周期底部：40—45美元
分拆初期市场完全低估，无人关注AI存储逻辑，这里是整轮几十倍行情的绝对原点底。
二、主升浪确认起点
趋势突破位：200美元
突破后彻底打开上行空间，机构集体重估NAND AI价值，正式开启无脑主升浪，再也无低位机会。
三、历史终极顶
最高点：2354美元
从40美元启动，最大涨幅接近58倍！
极致情绪、极致溢价、完全透支未来几年景气度。
四、本轮暴跌最低低点
本轮回调极值：998美元
从高点2354下来，最大回撤超58%
短短一个多月，超级牛市直接腰斩！
核心总结
1. 上涨：AI存储稀缺标的 + 周期涨价 + 分拆估值重估，三重驱动走超级大牛市。
2. 下跌：涨幅透支一切 + 韩厂扩产预期 + 高位巨量获利盘兑现 + 赛道抱团瓦解。
周期永远不变：
底部无人问，顶部人声沸，极致上涨之后，必是极致杀跌。Recently, some interesting data from the US stock market showed retail investors frantically selling individual stocks, with the selling orders almost all concentrated in one place. According to statistics from Vanda Research, on Tuesday, U.S. retail investors recorded a single-day net stock sale of $213 million, marking the highest level since the pandemic crash in March 2020. But what's even more worth pondering is not how much was sold, but what was sold. Data shows that a full 88% of this wave of selling pressure is concentrated in the memory chip sector. Micron, SanDisk, Seagate, and Western Digital have become the main players being "abandoned." You have to know, these are the core stocks in the AI hardware wave—the more aggressively they rose before, the more decisively they were dumped now. $SKHYNIX Interestingly, retail investors did not completely exit the market. They sold individual stocks and immediately bought the Roundhill Memory ETF, essentially switching from picking stocks themselves to buying a basket of storage chip companies. $MU What does this mean? Simply put, retail investors are not losing confidence in AI; rather, they have become more "shrewd" or more "timid." They are still willing to stay in the storage sector, but no longer dare to put their chips on one or two companies. Behind this is a reassessment of risks. Previously, AI hardware prices rose too quickly and valuations were not cheap, causing the market to start anxious about future investment returns. Rather than betting on Micron or Western Digital's "single signing," it's better to buy ETFs to diversify risk, at least keeping pace with the sector's average gains. From a broader perspective, this phenomenon$QQQ 干跌俩点，$SOXL 杀出一条 -13.8% 的血路，半导体集体阵亡，$BTC 却只 +0.64% 淡然看着？盘面没脱钩，是钩子藏得更深了——真正的暴击还在特朗普嘴里，别急着上桌。 看数字 加密：$BTC 63,957 +0.64% $ETH 1,907 +0.08% 美股：$QQQ -2.04% $SPY -1.54% $IBIT -0.39% 汇金：$DXY -0.58% $GLD +0.46% $MU -9.3%、$SKHYNIX -10.3%、$SOXL -13.8%，AI 故事突然讲不动了，一把把 $QQQ 拖进深水。原油和霍尔木兹还在往通胀里浇油，美债收益率压得紧，Fed 没给软话，估值这口气就是喘不顺。币圈这边成交额最大的反而是 $ETH -0.1% 装死和 $BTC 勉强红着，$SOL 跟着 +0.4% 假装没事，但资金根本没敢到处浪。 细看，$ETH 连 $BTC 的尾灯都摸不着，钱全往硬的怀里缩。$QQQ 腿软成这样，$IBIT 比现货还软 -0.39%，ETF 率先露怯，现货的底是有人在撑，但不想加码。$DXY 终于松了 -0.58📍 能源通膨與 AI 現金流｜叮叮跨市場筆記 截至：2026/07/30 08:00 UTC+8 【核心觀點】：油價尾部風險與利率分歧，壓過單點科技利多。 📊 叮叮多空雷達： ├─ 🪙 Crypto：偏空 ├─ 🇺🇸 美股：偏空 ├─ 🛢️ 原油：偏多 └─ 🟡 黃金：高波動 💡 市場屬性：能源通膨尾部風險＋AI 資本支出品質檢驗 ⚡ 未來風險窗口（傳統市場 1–3 個交易日；Crypto 24–72 小時）： ├─ 🪙 現貨／中長線：暫停追價、分批觀察、保留現金 └─ ⚡ 合約／短線：僅在航運再受阻，且美債殖利率與美元同步上行時偏空；航運緩和且兩者回落則失效。需避險，避免追價與過度槓桿 💬 一句話總結：市場同時計價供應風險、Fed 分歧與 AI 現金流，方向仍看航運與利率。 ─── 【🛑 第一戰線｜中東與能源】 💥 伊朗向約旦美軍目標發射飛彈，美沙打擊伊朗支持的伊拉克武裝；伊朗另宣稱擊中並攔停荷莫茲海峽三艘油輪。 • 驅動核心：這不是已確認的供應中斷，但提高航運受阻機率；若風險延續，原油→通膨預期→殖利率的鏈條會壓縮美股與 Crypto 估值。 ─── 【🛑希捷 $STX 周二美股盘后发布的财报，单看数据几乎挑不出什么大毛病。 本想着今天开盘能逆着 $MU 的走势走一会，结果开盘后的走势直接来了个高开低走 看着还有继续下去的样子 “AI基本盘回落”这个事，难道真的没有例外么？ 希捷FY2026 Q4营收达到36.29亿美元，同比增长48.5%，较上一季度31.1亿美元环比增长约16.7% 更夸张的是，每股收益(EPS)和自由现金流都实现了三位数同比增长 而且这次增长是由主营业务带来的增长，并不是单纯依靠生产更多传统硬盘推动收入 这背后对应的是云数据中心对大容量硬盘的需求 "公司表示，云数据中心需求是本轮增长的主要驱动力" 市场需求没有发生变化，甚至还挺强劲的 管理层对下一季度也没有表现出需求减弱的迹象，收入和每股收益均给出了双位数的增长指引 但即便如此，交了这样的一份财报，也没有走出独立行情出来 过去2个月我看到市场对AI概念股的预期都是“好是不够的，得一直超超超预期的好” 这个预期值的持续有点太夸张了，目前AI板块的风险偏好仍然很弱，如此强的财报，也只能换来高开后回吐 而且半导体指数今天盘中也在继续走弱 所以盘面高开回落，还是好消息今天早上，长鑫科技的股价上演了一场“火箭发射”。开盘价52.20元，随即一根直线拉升至53.89元，涨幅3.24%。而在另一个战场Hyperliquid的CXMT合约市场，价格更是一度冲到了7.646美元，开盘一分钟内就暴拉了6个多点。 这波急速拉升，直接让一个做空的“巨鲸”在开盘仅34秒后，就彻底“爆仓”出局。 根据链上监测，这个地址为0xf91开头的巨鲸，在开盘不到一分钟的时间里，连续被强平了两轮。总计24.69万份CXMT空单被一扫而空，清算均价在7.52美元左右，总规模高达185.8万美元。这一把，他净亏了约9.54万美元。 最让人唏嘘的是，这并非一次冲动的“梭哈”。$SPCX 这笔空单其实是这位巨鲸精心策划的“杰作”。从昨天晚上开始，他就启动了一个长达10个小时的TWAP卖单策略，慢慢地、有计划地建仓做空，目标总量是25万份。到今早9点前后，空单基本布置完毕，平均成本控制在7.138美元，可以说相当稳健。 那为什么结果这么惨烈？复盘一下他的操作路线，你会发现这是一个典型的“多空双杀”剧本。$SNDK 就在前一天，也就是长鑫科技上市的第二个交易日大涨之后，这位巨鲸才平掉了手$LAB 3.1亿枚代币从狗庄的地址分发到一个新建地址上，又要开始分发出货了！Meta Q2 指引不能當成結果：Q3 營收、全年費用與 AI 資本開支三張表
財報發布後最容易被誤寫的是管理層前瞻。Meta 預期 Q3 2026 營收介乎 610 億至 640 億美元，並假設匯率對同比增長約有一個百分點逆風。這是 7 月 29 日時點的未來區間，不是 Q3 已完成收入，也不是公司保證一定落在區間內。
全年費用前瞻調整為 1,650 億至 1,690 億美元。公司表示，提高下限是把 Q2 已確認的 24 億美元法律程序費用納入全年預期。這裡有兩層：24 億美元已經進入 Q2 GAAP 結果；全年區間則仍包含未來兩季的假設。內容發布必須把已發生和前瞻分開。
資本開支前瞻收窄到 1,300 億至 1,450 億美元，先前為 1,250 億至 1,450 億美元。Q2 實際口徑是包含融資租賃本金的 310.78 億美元，上半年自由現金流 131.70 億美元。不能用全年區間除以四來填補季度，也不能因上限不變就說投資計畫沒有變化。
公司仍預期 2026 全年營業利潤高於 2025 年，但 Q2 營業利潤 187.75 億美元、同比下降 8%，營業利潤率降到 31%。全年目標能否達成，將取決於下半年廣告收入、法律與遣散費基數、AI 基礎設施折舊、人才成本及 Reality Labs 虧損。這些條件尚未發生，不能寫成既成事實。
稅率前瞻亦已調整。若稅務環境沒有變化，公司預期 2026 剩餘季度稅率為 15% 至 17%，先前為 13% 至 16%。Q2 實際有效稅率為 16%，去年同期 11%。實際稅率和未來區間可以放在同一張表比較，但不能將區間中點代入尚未發布的 EPS。
營運面還有約 8,000 名員工受 5 月裁員影響，多數預計到 Q3 末不再計入員工數。Q2 期末報告員工 75,472 人，同比下降 1%，但當中仍包括受影響人員。未來費用節省與遣散現金支付時點不一定同步，應等待正式季度表格驗證。
公司同時提醒，法律與監管事項可能對財務結果產生重大影響。這類風險沒有可安全代入模型的固定季度金額；Q2 的 24 億美元是已確認費用，未來案件結果則仍不確定。發布內容不會把本季費用直接年化，也不會因管理層列出風險就預設下一季必然再出現同額支出。
因此，Meta 的財報後文章會固定採用「Q2 實際、Q3 指引、全年指引」三欄：實際數字判斷已完成經營，Q3 區間觀察近期假設，全年區間追蹤資本與費用方向。把三者分開，才能既呈現管理層信號，又不把 AI 投入回報、營業利潤或下一季收入提前寫成事實。$BANK 一个月前狗庄集群积累的30多个地址还是没有出货，不过可以明确的是狗庄其他集群已经出过几次货了，暴跌还会再次到来，拉高就空，直接住皇宫！昨夜最容易被忽略的，不是“美联储没有加息”，而是这次按兵不动背后的9比3。 联邦基金利率目标区间继续维持在3.50%—3.75%，但三名委员明确主张加息25个基点。政策表面静止，内部裂缝却在扩大：这意味着市场下一步定价的，不只是“何时降息”，而是通胀重新抬头时，鹰派会不会把方向盘抢回来。 BTC的反应也很诚实。OKX BTC-USDT-SWAP在本次快照中报约64,260 USDT，24小时高低区间为64,743.8—63,226.5，振幅约2.40%。消息后先下探、再急拉，但仍没有真正脱离区间。换句话说：情绪完成了一次过山车，趋势却还没完成选择。 我盯三道门： 1）64,744上方能否出现连续1小时收盘，而不是一根消息针； 2）回踩63,226附近时，成交量是否缩、低点是否抬高； 3）如果再失守63,226，先看流动性回撤，不把“第二次反弹”自动当成反转。 真正的风险不是看错一根K线，而是在政策分歧扩大、价格仍处于箱体时，把高波动误读成高确定性。研究结论很简单：突破要等确认，抄底要有失效点，仓位必须给“再来一次假动作”留空间。 #美联储维持利率区间3.50%-3.75%，表决为9比3BTC sitting at $64K with essentially no directional bias ahead of tonight's Fed decision and Big Tech earnings is not bullish or bearish on its own. It reads like a market that has already priced the hold outcome and is waiting to see whether the guidance shifts. The Hynix print matters here: record revenue that still missed expectations tells you forward multiples are the real test, not the backward numbers.
The interesting signal tonight is not the rate call itself but the dot-plot tone and whatever Microsoft and Meta say about AI capex. Crypto's correlation to risk-on positioning tends to compress before catalysts and snap sharply after. Flat BTC at this level is positioning, not conviction.
Just my read, not advice.
#OKXOrbit$LAB 拉高就空一点毛病没有，狗庄一直在分发代币！四个 LAB 项目多签地址向一个分发钱包发送了 453,000,010 LAB
随后其中 404,159,163 被分成 50 次转账分配到 37 个钱包中
仍有 48.8M 留在分发钱包中
此次分配是脚本化的。相同金额在不相关的钱包中重复出现，精确到小数点后第四位：2,182,045.4830 转入 8 个不同地址，17,292,845.1667 转入 6 个，16,889,095.0007 转入 6 个。
我检查了所有 37 个钱包的当前余额。每个钱包都正好持有收到的金额。没有流出转账。没有交易所存款，没有 DEX 出售。在此之前，这些钱包中没有任何一个持有 LAB。Strive raises BTC warehouse to 20,000 VND
Strive has just announced the purchase of an additional 79 BTC at an average price of around $65,700, bringing its total holdings to exactly 20,000 BTC. 
In the context that the market is still sideways and many funds are withdrawing capital, the fact that a public company continues to accumulate steadily has quite great psychological significance. They don't buy peaks, nor do they wait for absolute bottoms – just accumulate as planned.
Insight: The "New MicroStrategies" are learning to buy more sustainably. They do not create price waves immediately, but gradually create a psychological support layer when the price returns to the low area. This is a "slow but sure" type of cash flow, which is different from FOMO retail.AI股跌成一片，币圈也被带下水了
昨夜美股明显下跌，纳指跌1.74%，多只AI和芯片股接近跌停式下挫
原因并不复杂：前面涨得太多，财报稍微不够惊艳，资金就会先卖出锁定利润
币圈和科技股经常吸引同一批喜欢高波动的资金
当科技股大跌时，这些资金也会减少币圈仓位
目前BTC约6.37万美元，SOL约73.6美元，市场情绪偏弱
后面若美股企稳，币价会轻松一些
若科技股继续下跌，BTC即使反弹，也可能很快遇到卖盘$BTC $SOL 一份接近满分的财报，输给了市场提前写好的「超满分」答案。 撰文：Jim，麦通 MSX 编辑：Frank，麦通 MSX 利润同比暴增 557%，营业利润率达 76%，双双刷新历史纪录。 放在任何一家制造业公司身上，这都是一组近乎梦幻的梦幻数字，但 SK 海力士做到了：Q2 实现营收 79.32 万亿韩元，环比增长 51%、同比增长 257%；营业利润达到 60.54 万亿韩元，环比增长 61%、同比增长 557%，双双刷新历史纪录。 然而，财报公布后，SK 海力士盘中股价一度重挫逾 15%。 原因也并不复杂，市场此前预计其营收约为 84 万亿韩元、营业利润约为 64 万亿韩元，实际结果均略低于预期，换句话说，海力士交出的并不是一份差财报，而是一份接近满分、却没能达到市场「超满分」要求的财报。 这也是近期 AI 硬件资产共同面对的共性难题。 当市场早已知道收入会增长、价格会上涨、利润会刷新纪录之后，单纯的「业绩很好」已经不够，投资者更需要确认利润还能以多快的速度继续超预期，以及这场存储牛市究竟还能持续多久。 一、数字依然强劲，低于预期却未必意味着需求转弱 先看最重要的主营业务。 第二季度，Google's century-old bonds fell to 89 pence, Oracle's CDS soared to an 18-year high, and AI-related bond spreads widened to 121 basis points—I stared at these numbers and laughed for a long time, confirming one thing: the market is repricing the story of the "AI money printer" into the reality of an "AI money shredder."
📉 Let's first see what happened
American tech giants are borrowing money for AI at an unprecedented pace. Amazon, Alphabet, Meta, and Oracle have issued about $194 billion in bonds in the first seven months of this year, 79% more than the full-year 2025 market. Goldman Sachs estimates that the five major hyperscale cloud service providers will issue $250 billion in bonds this year and $400 billion next year.
What is the result of the supply surge? Bond prices fell, spreads soared.
The spread on 10-year ultra-large-scale cloud service provider bonds has widened to 121 basis points, while the overall spread on investment-grade corporate bonds during the same period is only 80 basis points. All maturities widened across the board: 2-4 year terms rose from 30 basis points to 40, 5-7 years from 50 to 60, and over 20 years from 108.5 to 118.
What's even more heartbreaking are the specific cases:
· Alphabet's century-old bond—a £1 billion bond issued in February maturing in 2126—fell below 90 pence for the first time, closing at 89.978 pence, with a risk premium once reaching 139.8 basis points.
· Oracle's CDS soared to 2.03 percentage points, the highest since 2008—the market believes the probability of a company defaulting within five years has risen from 8.6% to 11.8%.
· The subscription ratio collapsed—from nearly 5x in February to less than 2x in July. Amazon was oversubscribed by 3.4 times in March bond issuance, but dropped to just 1.6 times in July.
· High-grade bond funds saw weekly outflows of $7.1 billion, the largest since April 2020.
· Of the 91 AI-related bonds issued this year, 78 yielded above their initial issuance, with a median increase of about 22 basis points.
Simply put: more and more people are borrowing money, fewer are willing to borrow, and prices are getting higher.
🔍 Why is the market panicking?
Money always has a place to come from. When such large-scale issuances are concentrated on a single theme, the market begins to rebound in pricing.
The widening spreads reflect a structural contradiction: the same group of companies driving the AI revolution are also the main source of new bonds issued for this revolution.
The six major hyperscale enterprises are expected to spend 785 billion yuan in capital expenditure in 2026 and may approach 1 trillion USD by 2027. But operating cash flow is only 778 billion, and about one-third of its expenses must be covered by bond issuance.
What made the bond market even more anxious was that Alphabet's free cash flow last quarter was already -5.86 billion. You make money while burning it, burning more than you earn—how can creditors not panic?
The market is becoming increasingly aware that supply in all asset classes shows no signs of weakening. This isn't panic, it's math.
💎 Reverse interpretation: This is not an AI collapse, but rather AI pricing shifting from "faith" to "mathematics"
Everyone is shouting "AI bonds crashed," but what I see is something else—
This is not the bursting of the AI bubble; rather, the bond market is finally starting to settle accounts with AI.
In the past, buying AI bonds relied on faith—"these giants won't go bankrupt anyway." Now it's different. BlackRock issued $12.55 billion in bonds for Meta's Texas data center, with a yield set at 7.534%, close to junk bond levels. The investor said: If you want me to buy, that's fine, but give me enough interest.
The market did not reject AI bonds; it was simply saying: "The storytelling phase is over, now please show your cash flow." ”
This is exactly the same logic as in the crypto world. In 2021, people bought coins based on the narrative of "digital gold." After LUNA crashed in 2022, the market began to ask, "Where are your real returns?" AI bonds are now undergoing the same process—moving from "narrative premium" to "risk pricing."
Barclays put it most clearly: widening tech bond spreads have made bond valuations attractive, but market expectations for more supply may keep volatility high. Before a clear inflection point in financing demand appears, there is no need to significantly increase positions just to capture the interest rate spread.
Translated into plain language: Cheap is cheap, but it might be even cheaper.
🎯 My trading advice
First, don't rush to bottom-fish AI bonds. Barclays doesn't even recommend rushing in now. Supply hasn't peaked yet, and Microsoft hasn't returned to the bond market yet. There may still be room for the spread to widen by 10-15 basis points. If you don't catch the throwing knife well, your hand is gone.
Second, focus on the "Oracle Bone Script Issue" vs. the "Real Giants Problem." Most of the widening spreads were actually driven by Oracle Bone Script. Excluding Oracle, the credit spreads of Google, Microsoft, Meta, and Amazon are actually relatively stable. Market fear is an oracle story, not a comprehensive accusation against the entire tech giant. If you really want to build a build, keep an eye on those four brands—don't touch Oracle.
Third, wait for two signals: first, a clear turning point in financing demand (Barclays' original words); second, the growth rate of AI capital expenditure is starting to slow. Before that, cash is king.
🎬 The final sincere words
I was the man who carried 10 hits to 17, watched it hit back at 5.5 and then back to 17. Every time the market tells you "this time is different," it ultimately proves that "it's actually the same."
The current situation of AI bonds is no different from tech stocks in 2022, cryptocurrencies in 2021, or internet bonds in 2000—everyone is borrowing tomorrow's money and betting on today's dreams. The only difference is that some people win the bet, while others die before dawn.
The bond market does not lie. Stocks can rise through stories, but bonds can only be repaid with cash flow. When creditors start demanding higher interest rates, it means they start to doubt whether the story can be fulfilled.
Follow me, and I'm not teaching you how to chase gains or buy drops; I'll teach you to wait until others panic before entering to buy bargains. Give us a follow, and next time the market recites 'AI changes the world' to you, at least someone will shout in your ear—'Look at cash flow first!' Let's look at the story again! ”
---
#AI巨头债券利差飙升: Investment risk or a good opportunity to buy the dip @你的爱播Misa @币圈搅屎棍 @Wolf.Win @芦焕衡 @币圈搅屎棍 $SNDK 从OKX合约数据来看，目前除了比特币和以太坊之外，成交量已被美股、黄金、原油等传统资产大幅占据。
币圈整体交易活跃度明显下滑，并非资产归零，而是参与者显著减少。
部分资金已转向美股等传统市场进行交易。同时，OKX近期也悄然上线了较多美股现货交易品种，反映出平台正在积极拓展多元资产布局。在这样的环境下，热点轮动速度加快。
存储板块降温后，市场又会寻找新的叙事点。目前来看，RWA（真实世界资产） 已成为币圈最具持续性的功能方向之一，其将传统资产上链并实现 tokenized 的趋势，值得长期关注。
$BTC $ETH $SNDK Interpret the midnight meeting from three points, and you'll fully understand gold!
1. Short-term Pulse Market (Decision Stage)
The market's first reaction was interpreted as "dovish," with short stop-losses driving gold prices up rapidly, driven by sentiment and generally with weak sustainability.
2. The Press Conference Sets the Tone (The Real Watershed)
Walsh sends a hawkish signal: be cautious about inflation, keep rate hike options on hand. The market repriced forward rates, gold surged but then pulled back—a typical case of "decisions luring bulls, speech reversal."
3. Biggest Market Outlook: Continued Range-Bound Fluctuations
On one hand, geopolitical conflicts and central bank gold purchases provide a long-term support for gold; On the other hand, the Federal Reserve's high interest rate cycle continues, suppressing upward space for gold prices. In the short term, it is difficult to break out of a one-sided sustained trend; oscillating and consolidating the market remains the main theme.
So when it comes to defining trends, I've never blindly guessed—I've always been rational. If you rebound at a high point, don't chase long stocks—it's just a wide range of oscillations!
The early morning high of 4116 is strong resistance. If the rebound fails to break through again, it means a double top pullback. Bearish is sound!
Short at 4095-4110, guard against 4118, target 4050, break down to 4000!In August 2025, Ethereum finally broke its all-time high, reaching 4900, and the E-Guardians celebrated wildly
But there is a huge problem that few people pay attention to
Ethereum's TVL in 2025 will reach 400 billion, twice the massive bull market of 2021, making Ethereum the absolute star in 2021;
During the same period, Ethereum stablecoins reached 180 billion, a record high, twice the massive bull market of 2021
Fundamentals doubled, but prices endured for a full four years to barely move a little more than the peak four years ago
Why is it that although fundamentals have multiplied several times, prices still struggle to break historical highs?
We still use first principles to thoroughly understand this issue, which will determine whether you can lock in profits in a bull market
We all know that crypto prices are driven by narratives, which are responsible for inflating valuation multiples.
But the narrative itself has two sides: it can accelerate bull market rallies and bear market crashes
So when prices driven by narrative enter a bear market, what is killed is the valuation multiple, until it returns to near fundamentals.
For Ethereum, the most critical factor is the narrative + fundamentals double kill. Because speculation cools down in a bear market, protocol revenue also cools, and fundamentals fall accordingly. Fundamentals originally supported prices, but this floor keeps dropping, causing price drops even more
In 2022, Ethereum fell from 4900 to 880, a nearly 80% decline, and mainstream Ethereum narratives like NFTs, games, and fi were all proven off, pushing on-chain speculation to near rock bottom
Once the narrative ends, speculative capital immediately turns to chase the next high-growth story
From then on, to return to previous highs, optimistically, it would take several years to endure, plus the fundamentals multiplying several times. This is already the best outcome Ethereum with real fundamentals can achieve.
This model has been thoroughly validated in the US stock market
Intel: 20.9 billion in 2020 profit, double the 10.5 billion in 2000. The highest stock price in 2020 was $69, down from $75 in 2000.
Micron: 2020 profit of $2.69 billion, nearly 80% increase from $1.5 billion in 2000. The highest stock price in 2020 was $75, which is 20% lower than the $97 in 2000.
Cisco: Profit in 2020 was $11.2 billion, more than four times the $2.67 billion in 2000. The highest stock price in 2020 was $50, only 60% of the $82 in 2000.
Three companies, over twenty years, have higher revenues, higher profits, and healthier companies, yet their stock prices have not returned to previous highs.
Even if it later hits new highs, it was only achieved over a quarter of a century and several times the fundamental growth.
The essence is: the narrative that once hyped valuations to the skies has long since collapsed. The fundamentals slowly climbed back but couldn't recover the original multiples.
Ethereum has real fundamentals, Cisco and Intel are world-class companies, and if they end up like this, then expect even less from other things
Back to the present, we need to consider that in future bull markets, the optimal principle for allowing altcoins to cash in is the best principle determined by first principles
Even leading companies with income and buybacks, once trapped at a high point, only two outcomes remain: either that peak never returns, or they endure for several years before breaking even.昨晚市场被一个消息打懵了。 美联储FOMC会议虽然选择维持利率不变，但9-3的投票结果让市场重新紧张起来。 三位联储官员公开支持加息，这是近十年来少见的情况。 原本市场还在交易“降息预期”，结果美联储直接提醒： 通胀目标没有改变，政策不会因为市场情绪提前转向。 于是风险资产开始承压。 美股道指暴跌超过1100点，纳指进入调整。 BTC也从高位回落，跌至64000美元附近。 恐慌贪婪指数重新回到28，市场情绪明显降温。 但问题来了： 这到底是趋势反转，还是一次正常洗盘？ 从资金逻辑来看，目前市场其实处于一个矛盾阶段。 一方面： 宏观压力还在。 高利率、 美元流动性、 地缘风险， 都限制着风险资产上涨空间。 另一方面： 加密行业内部正在发生一些长期变化。 比如CLARITY法案虽然出现争议，但美国建立加密监管框架的大方向并没有改变。 过去几年机构最大的顾虑就是： “不知道规则在哪里。” 如果未来监管逐渐明确，真正受益的可能不是短期热点币，而是BTC、ETH以及合规基础设施。 另外一个值得关注的信号： 传统金融正在越来越深入进入Web3。 Robinhood最新数据中，预测市场收入首次超过加美联储7月没加息，但市场反而慌了——长期美债收益率飙升，到底在怕什么？🧵👇
1/ 鲍威尔（现Warsh）用了13次「家庭斗争」这个词，暗示内部分歧巨大。
2/ 他说「通胀是一种选择」，翻译成人话就是：如果通胀再起，我们会加息。
3/ 进口价格意外上涨，中国商品价格创2008年以来新高——关税的代价开始显现。
4/ 批发价格却下跌了0.3%，因为汽油跌了。但剔除能源后的核心CPI依然顽固。
5/ 市场定价显示9月加息概率升至45%。华尔街认为「加息在路上」。
6/ 特朗普的全球关税还在加码，新加坡都紧急收紧货币政策了。
7/ 总结：美联储在玩火，通胀没死，加息只是时间问题。投资者别太乐观。转发让更多人看到。HYPE 这波回落，别只看成普通解押砸盘
它更像一次「叙事拥挤度测试」
Hyperliquid 的基本盘很强，交易量、用户黏性、永续生态都能打。但当市场看到大额 HYPE 即将解除质押，尤其单日释放规模被反复传播，价格压力就会先于真实卖盘出现。很多人不是等别人卖了才跑，而是看到别人可能卖，就先按下卖出键
这就是高预期项目的尴尬
好故事会吸引更多持有人，也会让每一次解锁都变成压力事件。大家都相信长期价值，但短期谁也不想给别人接筹码
我觉得 HYPE 真正要证明的，不是能不能讲出更大的愿景
而是平台收入、回购、生态增长，能不能消化越来越具体的供应压力
#HYPE遭大额解押减持，一周回落10% 停火 48 小时告吹，油价这条线又回到最烦的状态
不是单边上涨，也不是风险解除，而是每一条谈判消息都能把交易员来回抽一巴掌
美伊边打边谈，本质上让原油重新变成「政治期权」。只要霍尔木兹、红海、油轮保险、导弹拦截这些词还在新闻里，油价就不可能只按库存和需求走。它会多一层风险溢价，多一层航运成本，再多一层通胀预期
对加密市场来说，这种油价扰动很讨厌
油价上去，降息空间被压；油价急跌，又可能代表市场在赌冲突缓和。结果 BTC、ETH 不光要看链上和 ETF，还得盯中东新闻
这就是宏观行情最恶心的地方
你明明买的是币，最后却在看油轮往哪开
#停火48小时告吹，美伊边打边谈 苹果重新拿回全球市值第一，我觉得最刺眼的不是它赢了英伟达
而是它用一种很「不 AI 泡沫」的方式赢了
过去一年，市场一直奖励谁更敢买 GPU、建数据中心、烧资本开支。谷歌、Meta、微软、亚马逊都在拼命证明自己不会输掉 AI 军备竞赛。苹果反而慢半拍，AI 叙事不够炸，发布节奏也被骂保守
但到了资本市场算账的时候，这种保守突然变成优点
苹果没有把现金流压到极限，没有把未来押在一堆数据中心债务上。它更像是在等 AI 真正变成设备体验和服务收入，而不是先把账单堆满
市场现在不是不爱 AI
是开始讨厌「为了 AI 不计成本」
#苹果公司市值重回全球首位，超越英伟达 1. FOMC维持利率不变，但3票支持加息，市场解读明显偏鹰 美联储将联邦基金利率目标区间维持在 3.50%—3.75%，符合多数预期；但决议仅以 9票赞成、3票反对通过，三名委员主张加息25个基点，较6月会议的全票通过明显转鹰。美联储同时维持准备金充裕的政策框架。 机会与趋势： 利率没有实际上调，若后续油价和通胀回落，市场仍可能重新交易年末宽松；黄金在决议后明显走强，反映部分资金同时押注政策不确定性和地缘避险。 风险与预警： 三张加息票说明美联储内部对通胀风险的担忧正在上升。若能源价格持续反弹，市场可能继续上调“下一步是加息而非降息”的概率，压制科技股、BTC及其他长久期资产。 关键数据变化： 现货黄金一度上涨约 1.9%至4,102美元/盎司；美国长债收益率在偏鹰决议后上行，风险资产同步承压。 ⸻ 2. 美股：FOMC与AI硬件抛售共振，纳指、标普显著下挫 美东7月29日，标普500下跌 1.5%至7,316.15点，纳指下跌 1.7%至24,442.94点，道指下跌 2.2%至51,594.14点。AI芯片股仍是主要拖累，费城半导体指数已较6月高点回撤接近 30%。 机会与趋势：#美联储维持利率区间3.50%-3.75%, with a vote of 9 to 3 #韩国拟限制散户杠杆ETF敞口并设投资上限
Is a global financial crisis coming?
1/ Storage continues to collapse, and the Korean stock market remains in a circuit breaker. Many people treat it as a joke, thinking it's because Korea's local leverage is too high. But if you look back at the history of global financial crises over the past thirty years, you'll find a pattern: in every major crisis, South Korea is always the first to fall.
2/ In the 2020 pandemic stock market crash, three weeks before the four U.S. stock market circuit breakers, South Korea's KOSPI had already fallen 35%.
Two months before Lehman's bankruptcy in 2008, South Korea was already short of US dollars.
Before the Nasdaq crash in 2000, Samsung Hynix's expectations were revised downward, and South Korea's semiconductor industry peaked first. During the 1997 Asian financial crisis, South Korea was the first core economy to be breached.
3/ This is no coincidence. South Korea's capital market is almost fully open, with foreign ownership consistently exceeding 30%. Samsung and SK Hynix are among the most liquid targets globally. With free capital inflow and outflow, ample undertakings, and even large sales can be quickly executed.
4/ So South Korea has become a "reserve cash pool" for global capital. Western institutions usually earn profits in Korea, but when local liquidity tightens, margin runs tight, or debt matures, their first reaction is: sell overseas holdings and withdraw money back to the home market to put out the fire.
5/ Clear priorities: protect the local area first, then abandon the periphery; Sell first those with good liquidity, then move on to those that are hard to cash out. This has little to do with South Korea's economic performance or whether the stock market has a bubble; it's purely a capital instinct to protect itself.
6/ This time, the trigger for South Korea is the semiconductor bubble overlay leverage. On average, each person nationwide has two stock accounts, and one out of every three trades is a margin financing position. Once foreign capital withdraws, domestic leveraged markets face consecutive blowouts, and circuit breakers cannot be stopped. Since the beginning of the year, there have been 35 programmatic circuit breakers and 5 market-wide circuit breakers, already breaking the 2008 record.
7/ But South Korea's problem is not Korea's own. It is a warning sign of tightening global liquidity. When global capital began to draw blood from overseas, South Korea was the first bleeding point, then the blood spread layer by layer along the capital and industrial chains.
8/ The four historical crises each triggered different triggers, but the underlying logic was the same: liquidity gaps first appeared domestically in Europe and America, capital withdrew from Korea, Korea collapsed first, then spread to Asia-Pacific, commodities, emerging markets, and finally back to Europe and America.
9/ Will this turn into a global financial crisis? The key variable is not South Korea, but the United States. In 2020, the Federal Reserve suppressed the crisis with unlimited easing and zero interest rates. What about this time? If the Fed can cut rates and inject liquidity, the market may be held back just like in 2020. If the Fed continues to raise interest rates or delays rescue, the real crisis may only be beginning.
10/ So my judgment: South Korea's circuit breaker is a warning, not a conclusion. Whether a financial crisis will happen depends on whether the Fed still has ammunition and whether it is willing to fight. Neither of these matters is currently certain.
11/ For ordinary people, the most important thing at such times is not to predict a crisis, but to control their positions. Never go all in, and never add leverage. Always keep some cash, because real opportunities for wealth often come at the most panicked times.
12/ My approach: 60% of my core positions are in the S&P 500 and Nasdaq 100, holding them for the long term. The remaining 40% is cash or short-term bonds, especially when the index drops 15%, 30%, or 40%, adding to the position. It's not about bottom-fishing, but about following the plan.
13/ Historical data shows that the Nasdaq 100 and S&P 500 hit new highs after every major crisis. 1987, 2000, 2008, 2020, 2022—without exception. A crisis is not the enemy of long-term investors, but an opportunity.
14/ So I am not afraid of crisis. What I fear is that when a crisis comes, I won't have cash to add to my position. What I fear even more is when a crisis comes—I panic and cut my flesh, handing over my bloody chips to someone else.
15/ Korean circuit breakers are a wake-up call, but not a way to clear out your stock. It reminds you to check your positions, control leverage, and keep cash. The real winners are not those who predict the crisis, but those who can hold onto their chips and dare to act during the crisis.FOMC 9:3 维持利率，但三张加息票才是真正的炸弹
昨天美联储的戏，比剧本精彩。
决策本身： 9:3 维持利率在 3.50%-3.75%，符合预期。
真正信号： 3 张加息反对票——克利夫兰、明尼阿波利斯、达拉斯联储主席全投了加息。2016 年以来第一次三票一致 dissent。这不是"温和暂停"，这是内部已经分裂。
沃什发布会金句： "没有软通胀目标。" 翻译成人话：通胀没搞定之前，别指望我松手。
BTC 反应： 决议公布后稳守 64,400，沃什讲话后跌破 64k，现在 63,800 附近震荡。跌不多，但也不反弹——这才是最让人难受的走势。
我看到的信号：
CME FedWatch： 9 月前加息概率已飙到 63%。不加息到 9 月的概率只有 36%。
油在飙： 布伦特逼近 90 美元，中东局势火上浇油。能源价格传导通胀，美联储更难松口。
10 年美债收益率 4.67%——资金往那里跑，风险资产就得勒紧裤腰带。
今天值得盯的：
英国央行利率决议（同一天）
核心 PCE 数据发布
WEMIX 被黑后继续观察
我的想法：
短期空头叙事占优，这没毛病。但鲍威尔也好沃什也罢，加息周期已经到尾巴了。3 张票喊加息，恰恰说明再加的空间不多了——真有余地的话，9:3 就不会是维持利率。
对现货玩家来说，64k 以下的 BTC，是机会还是陷阱？
评论区聊聊你们的看法。我发现了一个规律，每次微软财报超预期，加密圈都要跟着嗨一把
不是因为散户买MSFT，而是MSFT的数据是AI叙事最大的基本面支撑。
昨晚微软Q4营收900亿，Azure增长43%，盘后直接+8.8%。纳德拉说Copilot要变超级应用了——整合聊天、编程、AI代理三合一。这是微软历史上第一次把全部AI能力打包成一个产品输出。
然后你猜怎么着。
英特尔居然也破天荒地向一家叫Rosaic Labs的初创公司开放了Atom处理器代码。这在英特尔历史上非常罕见——x86架构的核心技术外泄，以前Intel宁可让业务烂在手里也不松手。陈立武上台后，策略真的变了。
AI基建的狂热还没有降温，但边际信号已经出现——微软是第一个宣布小幅下调2027财年资本开支的大科技公司。不是不投了，是投的速度要降一降。
另一方面，千禧管理正在募资200亿美元押注AI行情延续。历史上还从来没有一家对冲基金募过这么多钱。说明最精明的钱仍然在往这个方向迁移。
所以我的判断是，AI叙事的长期方向没有变，但中短期可能出现"财报分化"——有收入支撑的AI公司（微软、LRCX、英伟达）继续涨，纯概念的AI币（SKYA
But when I checked, the market value was 128 million, and in 24 hours it rose to $1.29.
M is a small-cap coin suspected of being targeted by institutional investors, rising 66% but without any fundamental news to support it. Since June, I've seen more than 20 of these coins—the script is always the same: Zhuang La, retail investors, and → chase after → endless supply.
The truly interesting one is BEAT, +14%, with a market cap of 1.18 billion. This BEAT rally is not a bearish rally, but a rebound with volume and backing. Audiera has recently made new moves on the product side, and with the overall recovery of the DeFi sector, it's not the same as the 'unconditional price pump' at the end of last year.
Then guess what.
On the decline chart, HASH -31% was cut in half, and DEXE -10.7% also took a hit. From 100% in June to now a 10% pullback, this is a normal profit-taking in the DeFi sector. PEANUT -14.9% is pretty bad—a small meme targeted by bears, with no moat.
According to BlockBeats data, the altcoin resistance index is Hold, indicating that the current state is neither "comprehensive risk aversion" nor "full long"—funds are voting very precisely.
So my judgment is that the divergence between price movements and declines will continue to intensify. M's 66% is the result of capital games, not value discoveries, so it's not recommended to chase them. BEAT pullbacks are worth watching; a drop below 2.4 could trigger a technical rebound window.
Looking through today's plate, there are a few interesting points:
#
In short: Although the Federal Reserve keeps interest rates unchanged, three internal rate hike proposals have appeared, and BTC and ETH ETFs continue to flow out; Mainstream coin prices are temporarily stable, and the market remains in a phase of low sentiment and weak incremental volume. 📈 Today's Market As of July 30, 10:46 (HKT): BTC: $64,264 (+0.76%) ETH: $1,916.89 (+0.50%) SOL: $74.01 (+0.95%) Total Crypto Market Cap: $2.272 trillion (+0.36%) BTC Market Share: 56.45% Fear and Greed Index: 28 (Fear) BTC, ETH, and SOL all recorded slight gains in the past 24 hours, However, the total market capitalization only grew by 0.36%, and the Fear and Greed Index instead dropped from 29 to 28. This means: a temporary price stabilization does not mean market confidence has recovered. BTC's market share continues to stay above 56%, indicating that funds are still concentrated in mainstream assets and have not fully spread to high-beta altcoins. Among mainstream coins, XMR rose about 3.0%, showing the strongest performance; GRAM fell about 2.9%, the weakest performer. The current market remains a structural rotation, not a broad-based rally. 🔥 Today's Hot Topics (1) The Federal Reserve maintains interest rates, but three members advocate for a rate hike. The FOMC maintained the federal funds rate target range at 3.50% to 3.75% with a vote of 9 in favor and 3 against. Notably, Hammac is noteworthy近期特朗普相关加密叙事持续降温，成为本周市场重要风格切换信号。此前市场持续博弈美国加密合规、比特币战略储备、WLFI生态利好，题材炒作一度带动相关代币脉冲式拉升。 但随着美国大选周期临近、政策落地时间延后，市场逐步意识到：短期难有实质性政策落地，所有利好均已提前透支。特朗普系代币从趋势上涨，转为高位震荡出货，题材资金持续撤离。 同时白宫加密监管空窗期延续，既无强力利空打压，也无超预期利好落地，政策面彻底进入平淡期。币圈告别“消息驱动行情”，全面回归链上数据、赛道基本面、资金轮动逻辑。 名人题材退潮后，市场开始优胜劣汰：纯炒作、无落地、无生态的空气题材持续阴跌，有真实业务、有链上数据、有机构持仓的蓝筹赛道开始持续修复。RWA、DeFi、BTC生态持续走强，就是市场风格切换的最好证明。 后续行情不再靠消息刺激暴涨，而是靠真实基本面修复走趋势。短期题材炒作热度彻底终结，市场进入稳健结构性修复阶段。$ETH 看懂RWA就绕不开一个终极问题：
如今美元稳定币、代币化美股美债扎堆登陆以太坊，并且这个趋势越来越快。
大众普遍认为，以太坊蹭上华尔街红利，离开了美国金融，叙事直接崩塌。
但有没有另一种可能性？
美国想要把美元资产无孔不入输送全世界，革新老旧的跨境清算体系，迫切需要一条全球公认、流动性充足的可编程公链。眼下最优解，就是以太坊。
依附是双向的。
短期以太坊依赖华尔街资金；长期，美元向外扩张的布局，同样离不开以太坊基础设施。
所以，你觉得：以太坊需要美国，还是美国金融更需要以太坊？早上被KOSPI的新闻刷屏了，SK海力士跌了7.57%
业绩创纪录，但不够好。这就是现在AI赛道的残酷逻辑——你赚的钱破纪录了，但因为华尔街预期更高，你的股价就得跌。
SK海力士Q2的HBM出货量环比增长38%，DRAM ASP也在涨，怎么看都是亮瞎眼的财报。但市场就是嫌不够——因为英伟达的Hopper Next传说要改内存接口，海力士能不能吃到下一波存疑。
然后你猜怎么着。
三星电子也跟着跌了2.64%，KOSPI开盘涨的现在全跌回去了。但另一边，Lam Research（LRCX）盘后+6.58%，业绩和指引双双超预期，下季度营收最高直奔85亿。
存储股内部在剧烈分化——有真技术的涨，靠预期的跌。加密这边呢？BTC 63,695完全没跟韩国盘，继续走自己的独立行情。大摩推出了ETH和SOL的ETP，机构在认真配置。
所以我的判断是，AI硬件赛道的分歧正在加大而不是缩小。存储股的波动还没结束，但加密市场的"脱挂钩"趋势是长期的、结构性的，不是今天跌了7%明天就涨回的那种。
顺便留意一下最近的动态，有几个方向值得看看：
#海力士业绩创纪录但不及预期，存储股剧烈波动 动
海力士HBM出货环比增38%是非常solid的数据，但市场预期已经把"AI每季度翻倍"写进了定价里。一旦增速边际放缓，估值就会压缩。加密AI币同理——不是跌了就是机会，看清楚是错杀还是杀逻辑。
#摩根士丹利推出ETH和SOL的现货ETP
在存储股巨震、科技股抛售的背景下推ETP，大摩不是临时起意。ETP是机构配置的工具化一步，不是短线炒作工具。持续关注未来两个月ETP的净认购量，那才是真金白银。
#美联储维持利率区间3.50%-3.75%，表决为9比3
FOMC第五次按兵不动，沃什强调通胀目标不动摇。但9月加息概率不升反降。市场在说：我信你能鹰，但我不信你会真加息。这种信任对风险资产来说是一种底部支撑。
$BTC $ETH $SOL #美股 #宏观我真的会谢，上周还在纠结要不要买HYPE，还好管住手了
HYPE这一周跌了10%，而且不是插针，是慢慢跌的那种。
一看链上数据，大额解押减持在砸盘。解锁的币被分批转到交易所，每天砸一点，不给你反弹的机会。这种"清算式阴跌"比单日暴跌更恶心——你想抄底？明天还有下一批解锁在等你。
然后你猜怎么着。
这边HYPE在跌，那边美伊又在打。美军确认对伊朗动手了，霍尔木兹航运全线告急。但BTC动都不动，63,695收盘，24小时振幅不到3%。
我现在越来越觉得，市场正在发生一种"分裂"——单个币种自己走自己的，大盘也自己走自己的。HYPE被解锁砸，跟大盘偏不偏没有半毛钱关系。美伊打仗，跟BTC也没有半毛钱关系。
从BlockBeats的情绪指标看，整体流动性5Buy、5Hold、1Caution——偏多但不是那种"冲啊"的偏多，是"可以买但别追"的偏多。
所以我的判断是，HYPE短期内还得承压，这一批解锁必须等量消化完才能做多。喜欢HYPE的等链上解锁压力解除后的第二个放量阳线再上，别在阴跌中接飞刀。
再顺带看看最近大家都在聊啥：
#HYPE遭大额解押减持，一周回落10%
大额解锁+做市商分批出货的组合，HYPE的抛压结构跟2025年初ARB解锁的走法很像。不存在"提前消化"这个说法——解锁就是解锁，币价跌到解锁完成率出现拐点之前不要猜底。
#停火48小时告吹，美伊边打边谈
美军两周密集空袭方案已拟定，但市场反应几乎为零。不是市场不理性，是地缘脉冲在BTC上的边际效用已经降到历史最低。如果霍尔木兹持续断航推高油价，那才是真正影响通胀→利率路径的传导链。
#美联储维持利率区间3.50%-3.75%，表决为9比3
沃什最近几次讲话一次比一次鹰，每次说"必要时加息"BTC就弹一下——市场已经把鹰派表述当利好了。鸽派不涨鹰派涨，说明流动性预期已经触底。
$BTC $HYPE $SOL #宏观 #美联储My best friend asked me yesterday whether BTC and US stocks are still related
I said we used to have a relationship, but in the past six months...... It felt like we were about to break up.
I was shocked when I checked the data myself—the six-month rolling correlation between BTC and the Nasdaq had dropped below 0.3. Previously, when the Nasdaq fell 1%, BTC was desperate to drop 3%. Now Nasdaq futures are up 1%, but BTC is still hovering around 64K.
Then guess what.
At this node, Morgan Stanley launched spot ETPs for ETH and SOL.
You read that right—it's this juncture—the Fed just said 'raise rates if necessary,' geopolitical tensions are still raging, and tech stocks are being sold off. Morgan Stanley deliberately chose this moment to launch an ETP. What does this mean? It shows that the institutional allocation logic and retail investors are completely different systems. Retail investors fear rate hikes, war, and earnings reports; institutions focus on beta desensitization and the line where alternative asset allocation shifts from 0.5% to 5%.
From a technical perspective, BTC 63,695 has stabilized above 63K, with a 24-hour amplitude of less than 3%. 64.7K is short-term resistance, while 63K is strong support. ETH 1,901 holds above 1,900 but has not broken through the neckline of 1,940, while SOL 73.5 consolidates narrowly. All three major coins are shrinking volume, awaiting the Federal Reserve's rate decision.
So my judgment is that BTC's independent trend is forming but not yet confirmed. In the short term, wait and see the Fed's decision; only chase after a break above 64K.
There are a few other hot topics worth discussing today:
#比特币与纳指相关性大幅下降: Independence or Illusion
The decline in correlation is both true and fake. The real reason is that the BTC holder structure has changed—institutions, treasuries, and pension funds have come in. The fake is because liquidity is still shallow—when there is a positive correlation, BTC falls even faster. The key is whether the next round of ETF inflows can stabilize.
#摩根士丹利推出ETH和SOL的现货ETP
Morgan Stanley's launch of ETPs shows that traditional financial institutions' understanding of crypto has moved beyond the "can or can't understand" stage and has reached the stage of "how to allocate funds." ETPs cost more than buying coins directly, but institutional counter allocation needs are real.
#美联储维持利率区间3.50%-3.75%, with a vote of 9 to 3
Walsh's hawkish statement says the market has already priced in 80%. The real variable isn't whether rates will be raised, but the dot plot and Walsh's forward-looking statements on inflation. If Walsh says "keep interest rates high for the long term," risk assets will be under short-term pressure.
$BTC $ETH $SOL #宏观 #美联储隔夜全球金融市场经历了一场标准的“流动性压力测试”：美联储公布最新利率决议与政策倾向，美股科技股高位砸盘，韩股受半导体巨头拖累剧烈震荡，黄金则逆势拉升；而 BTC 依然死死锚定在 63,800 - 64,200 美元 这一窄幅区间内。
表面看似各走各的，但如果拉开视角，当前全球市场正在经历一场“高估值套现”与“流动性重新定价”的深层博弈。
以下是今日盘面与宏观局势的深度逻辑拆解：
一、 宏观与资金链：美/韩科技股巨震，究竟发生了什么？
美联储决议的底层信号：美联储维持利率不动的背后，彻底锁定了“加息周期终结，降息只是时间问题”的宏观基调。但市场之所以剧震，是因为降息预期的博弈已经从“会不会降”转向了“经济是软着陆还是衰退”。
科技股（美股/韩股）高位杀估值：美股 AI 链条与韩股半导体权重的高位回调，本质上不是宏观流动性枯竭，而是预期的阶段性错配。前期计入过多的宽松与业绩预期，在决议落地后引发了大型机构的“利好出尽（Sell the news）”和季度调仓。这部分从科技股流出的资金，并没有逃离市场，而是在寻找新的防守与高弹性落脚点。
二、 资产属性博弈：黄金破顶 vs BTC 的“定价权转换”
黄金做“信用对冲”：黄金的强势拉升，定价的是“降息周期开启后法币购买力贬值”以及“全球央行去美元化”的确定性趋势，是绝对的避险与抗通胀属性。
BTC 的“双重属性”测试：过去每逢美股科技股杀跌，BTC 往往会作为高 Beta 风险资产遭遇连带清仓。但这次科技股剧震，BTC 却没有出现跟跌破位，而是在 6.4 万美元 展现出极强韧性。 这说明：BTC 正在从单纯跟随美股科技股的“风险资产”，向兼具防守（数字黄金）与高弹性（流动性虹吸）的复合型资产过渡。 美股流出的边际资金，有一部分已经开始在加密 market 寻求沉淀。
三、 BTC 技术与结构：63,500 美元为什么是多头的“生死命门”？
筹码密集的铁底（63,500 - 63,800 美元）：从此前 57,800 美元突破 W 底以来，6.35 万 - 6.4 万是本轮机构换手最充分的成本区。隔夜在宏观外部剧震的冲击下，该位置依然毫发无损，直接证实了场内现货筹码锁仓极其死板，绝非散户在盲目博弈。
多头防守转进攻的锚点（65,000 美元 / 日线 EMA50）：宏观靴子落地后，大盘的变盘节点就在这 1-2 天。上方的首要目标是重新收复 65,000 美元（日线 50日EMA）。一旦放量站稳 6.5 万，将正式宣告本轮“突破后的 Backtest（回踩确认）”彻底闭环，主升浪将重新指向上方的 66,500 - 68,000 美元。
四、 本周战术与实操指南
现货策略：绝对定力，抱紧底仓我们在 5.8 万 - 5.9 万确立的低位核心底仓，在这个位置毫无动摇的理由。宏观大浪洗去的是高杠杆浮躁资金，留下来的是稳固的现货底座。现货死锁，不做无意义的波段。
合约策略：以退为进，右侧确认
低吸试探： 鉴于 63,500 美元支撑已历经宏观巨震检验，若短线仍有针对 63,300 - 63,800 美元 区间的探底插针，依然是风险极高的低倍顺势接多点（止损坚决挂 62,800 美元）。
右侧追多： 拒绝盲目猜顶试空。只有当盘面放量突破并日线站稳 65,000 美元 时，才是追加右侧仓位、博弈下一波主升浪的终极信号。$BTC $ETH $SNDK $545 $XMETA, after a sharp drop, would you dare to buy it?
Let's look at the market first: this is not an ordinary pullback, but a long bearish candlestick that pierced through sentiment.
I checked OKX spot data on 2026-07-30 at 10:23:54 Beijing time. In $XMETA 24 hours, it rose from 596.32 to 544.59, with a high of 600.03 and a low of 519.72, a decline of 8.67%, and a trading volume of about 2.7926 million USDT. Note that at this point, US stocks are no longer in normal trading hours. OKX's tokenized US stocks trade 24/7, and prices will follow after-hours valuations and crypto market sentiment, which may deviate from the rhythm of traditional markets after opening.
The first contradiction: the drop was fierce, but not a reckless dumping without volume. The 24h VWAP is at 548.80, and the current price is already below VWAP, indicating that most turnover costs today are overhead. If you want to recover in the short term, don't rush to call for a reversal; at the very least, it should rebound above 548-550, otherwise every rebound could easily turn into trapped positions reducing positions above.
The second contradiction: technically oversold, but the order is not very thick. The 1H RSI14 is only 27.92, MA7 is at 541.32, MA20 is at 575.71, indicating short-term recovery needs; The bid spread is $0.22, about 0.04%. The 0.5% deep buy is about 65,000 USDT, the sell is about 45,600 USDT, and the 1% depth is only 72,700 and 53,300. In other words, once the direction is driven by large orders, slippage becomes obvious. Don't treat oversold as a safety cushion.
The third contradiction: it's weak, but not the weakest in the field. On the side, low-liquidity stocks like XVRT, XHIMS, and XONDS have fallen even more; Among high-volume stocks, $XMETA ranked among the top decliners, while $XSOXL also fell about 6.68%, while $XSKHY and $XGOOGL remained in the red. This is more like a partial retreat among major tech heavyweights, not all tokenized US stocks crashing together.
At key levels, I would see it this way: below 519.7 is the low for this round, 520-525 is the first support; If 519 is broken down by increased volume, the short-term structure will fail. The next step is to look for integer bits near 500. Look at 548-550 above, then talk about recovery after it breaks; Further up is 575-576, which is near the 1H MA20, where the real pressure is real.
Short-term traders: Only watch if volume can shrink and stabilize above 520. Return above 548 before considering a rebound to continue. Falling below 519 lowers expectations.
Swing traders: Wait for 1H to recover above 575 before this sharp drop is more than just a technical rebound; Before it was reclaimed, it looked more like a weak fix.
Medium- to long-term observation: Don't rush to draw conclusions with a one-day K-line; focus on OKX's after-hours price and whether the next normal US stock trading session can converge. If the spread continues to widen, it indicates that 24/7 liquidity is still pricing in risk.
#XMETA #代币化美股 #OKX #RWA #美股Ethereum Is Moving Quietly—But the On-Chain Signals Are Getting Louder.
While ETH is trading around $1,940, the real story isn't the price action—it's what's happening on the blockchain.
🐋 A major whale recently transferred 28,500 ETH (worth over $55 million) from exchanges into self-custody. Large withdrawals like this often suggest investors are positioning for the long term rather than preparing to sell.
📊 Ethereum's network continues to strengthen, with active wallets and long-term holders steadily increasing even as price remains stuck in a tight range. Quiet accumulation has often preceded major market moves in past cycles.
🏦 Traders are also watching upcoming macro events, especially signals from the Federal Reserve. A more supportive outlook on interest rates could provide fresh momentum for both Ethereum and the broader crypto market.
When prices stay flat but network activity keeps growing, it often means the fundamentals are improving before the market fully recognizes them.
The charts may look calm today—but Ethereum's foundation continues to get stronger.
#ETH #Ethereum #Crypto #Bitcoin #Altcoins #OnChain #Blockchain #Fed #OKX撰文： 编译：AididiaoJP，Foresight News 熊市最狠的不是价格下跌，而是惩罚两样东西：糟糕的投资策略和杠杆。 我的目标很简单——给出一套可重复执行的实操手册，根植于链上真实数据和历史周期，帮你保住本金，用稳定收益积累选择权，再在有利位置进入风险市场。 历史告诉我们什么 2018 年是极端清算风险的典型：比特币从约 1.9 万美元跌至约 3600 美元，回撤超过 80%。在底部买入的人，一年后回报约 +192%。 2022 年则是治理与信贷危机：比特币从近 6.9 万美元见顶，最低跌至约 1.65 万美元，回撤约 77%。FTX 崩盘引发的传染效应，导致约 120 万枚 BTC 出现已实现亏损。底部买入者一年回报约 +127%。 2025–26 周期完全不同：峰值约 12.48 万美元，低点 5.8115 万美元，回撤 53.43%——有记录以来最小的一次。截至目前，本轮周期已运行约 267 天。处于亏损状态的供应量达 72.4%（约 1074 万枚 BTC 水下），MVRV Z-Score 低至约 0.24。 我的解读是：结构性风险低于 2022 年（系统性传染更During a break in lipstick, she glanced at her phone and almost smeared the lipstick on her nose
The US and Iran are fighting again, and this time it's not just empty talk.
A 48-hour ceasefire was just agreed upon before it fell through, as the U.S. military confirmed it had taken action against Iran. WOJ said the two-week intensive airstrike plan is already on the table, just waiting for one person to sign.
Then guess what.
BTC 63,695, down 0.18%.
A missile flew overhead, yet BTC didn't even drop by $300. Geopolitical shocks are really getting smaller and smaller on BTC now.
Some say this is a good thing—BTC is desensitizing, maturing, and decoupling from gold. Others say it's not a good thing—it means the market has become numb to everything, and when everyone feels "can't be smashed," that's when you should be truly alert.
According to the data, the six-month correlation between BTC and the Nasdaq has dropped below 0.3. Previously, the Nasdaq fell 1% and BTC fell 3%; now the Nasdaq is down 2%, and BTC may not budge at all. Morgan Stanley also launched spot ETPs for ETH and SOL at this point, showing that institutions are seriously allocation—not just speculating and then exiting.
So my judgment is that, in the medium to long term, BTC's logic of independent trading is strengthening, but in the short term, don't treat 'desensitization' as a 'bottom.' Before the geographic clearance, the 64K-65K box will be held up for a while.
Next, let's take a quick look at the latest hot topics and chat casually:
#停火48小时告吹, the US and Iran negotiated while fighting
As the US and Iran continued talking, the market had become completely desensitized. But the suspension of shipping in Hormuz remains unresolved. If the oil supply chain continues to be disrupted, the risk of energy inflation will return, which is what the Fed truly cannot tolerate. In the short term, BTC can be left unfollowed; in the medium term, logical chains must not be abandoned.
#比特币与纳指相关性大幅下降: Independence or Illusion
Morgan Stanley's push against the trend by promoting ETH/SOL spot ETP shows that institutions are truly treating crypto as an independent allocation, rather than the Nasdaq's beta. If correlation remains below 0.3, the entire crypto valuation model will be rewritten—no longer priced as a tech stock, but as an alternative asset.
#美联储维持利率区间3.50%-3.75%, with a vote of 9 to 3
Wash's hawkish stance—keeping inflation at 2% and raising rates if necessary. However, the probability of a rate hike in September has actually decreased, indicating that the market has already priced in on the "most hawkish moment." The real risk is not in the rate decision itself, but in whether new wording has emerged from Wash.
$BTC $ETH #宏观 #地缘 #美联储群里已经开始炸了7月28号我看$ETH还在2700晃一堆人骂它是稳定币骂完29号一根针拉到3200我盯盘盯到凌晨三点眼睛酸到看不清数字手放在键盘上没敢动结果今天醒来一看以太坊基金会直接把SEAL 911的联合创始人拉进董事会了这不是随便加个人这是Pascal Caversaccio专门做隐私安全的狠角色四人的核心董事会突然扩成五人信号太明显了隐私赛道要被推到前面$ETH突然不再是那个慢吞吞的老公链了它开始认真搞基础设施里的隐私层不是喊口号是真金白银砸进去的那种姐妹说这个节点像不像2019年的$BNB那时候很多人也觉得晚了但回头看那才是真正的起点我自己今早追了一点仓位不大毕竟3200这个位置压力不小但完全空仓我更怕怕半夜再来一根针直接补不回来聊到这儿我突然想到最近圈子里天天吵的那个事以太坊基金会这次拉人进董事会说白了就是在回应前段时间DeFi圈对MEV攻击的集体愤怒不是发公告道歉是直接用人事布局告诉社区我们在做事隐私跟安全以前是锦上添花的东西现在变成刚需了未来三个月我觉得零知识证明赛道还会继续热这波不只是炒概念是底层叙事真的在转姐妹们稳住别追高但也别裸空$ETH这波节奏有点东西 有同感的美联储按兵不动，但并不鸽派。
本次会议以 9:3 的投票结果，将利率维持在 3.50%—3.75%。值得注意的是，三名反对者都支持加息25个基点，说明美联储内部对通胀的担忧正在升温。
市场焦点也从“7月是否加息”，转向了：
9月会不会补上这一刀？
对市场而言：
BTC短期获得喘息，但流动性并未真正宽松
山寨币仍容易受到高利率压制
美债收益率若继续上升，高估值科技股也将承压
只有通胀降温、美元和美债收益率同步回落，风险资产才可能迎来持续反弹
一句话总结：
美联储暂停了加息，但没有结束紧缩。市场躲过了今天，接下来开始交易9月。$BTC
2026年7月28日，谷歌和Semnova几乎同时炸场。 谷歌搞了个Frozen V2，把Gemini模型的整个架构和关键权重像刻电路一样烧录进硅片，相当于把软件直接物化成硬件，推理能效暴增6到10倍。代价是模型一改版芯片就变电子垃圾，专为搜索、YouTube这类万年不变的超大规模业务定制。Semnova的SN50则是另一条路：用可重构数据流技术，把千亿大模型推理塞进一张标准PCIe风冷加速卡，功耗300多瓦，插进旧服务器就能跑，液冷机柜根本不需要。
两个产品撕开了同一个残酷真相：传统GPU和TPU再牛，数据也得在处理器和HBM内存之间搬运，这“内存墙”堵得又慢又费油。谷歌釜底抽薪，让权重直接住进计算单元，消灭了搬运过程；Semnova则像活字印刷，推理时临时拼出硬件通路，能跑各种开源模型。
产业链的冲击立刻显现。ASIC定制设计服务、Chiplet先进封装、高效散热方案突然成了比光模块还香的饽饽。
但要警惕利空赛道。传统高端GPU的分销商和服务器组装商将面临替代压力——当推理走向极致专用化和风冷化，囤英伟达卡的老模式恐遭降维打击。更狠的是，片内集成度飙升让芯片间通信需求锐减，