Orbit Post Sitemap
Harvard's "spaceship" according to the latest 1F data holds 430,000 shares of SPCX at the end of Q2, accounting for 1.50% of the entire US stock portfolio. But real cash can't compare to Musk's 48.4% or Alphabet's 550 million shares. Nvidia, Gafafund, and Jane Street's holdings are also distinctive. Cathie Wood's ARK holds 4.48 million shares, which were truly bought on the secondary market. After SPCX's lock-up expiration, it rebounded to 111, with room for further gains. However, note that 1F data lags by one and a half months, Musk's selling pressure is already evident, Q losses are 500 million but revenue exceeded expectations, so caution is still needed. 💰 Solana’s Kinetics (price velocity) has been negative for 284 consecutive days.
That is a long, sustained directional bleed, and has been pointed down since last November. But momentum is now on the verge of turning positive again.Harvard is truly going all in on space this time. The latest 13F digging reveals that Harvard Management held 12.93 million shares of SPCX at the end of Q2, with a market value of $2.21 billion, accounting for 52% of its entire public U.S. stock portfolio. A $4.3 billion stock, half of each stock was sold, and the largest position was not sold out
But Harvard really doesn't make the list on the big players' tables:
Musk himself 48.4%, with a market value exceeding 900 billion;
Alphabet holds 551 million shares, $94.2 billion—the $900 million invested in 2015 has multiplied a hundredfold in ten years;
Nvidia: 123 million shares, 21 billion—but this was acquired by SpaceX during its acquisition of xAI this year, not bought in the secondary market;
Gigafund holds 172 million shares worth 29 billion, and Jane Street holds 57 million shares worth 9.7 billion.
Sister Wood really bought ARK in the secondary market, with 4.48 million shares and 765 million yuan—the largest Q2 purchase.
SPCX rebounded over 20% in five days after unlocking, returning above the issue price of 135, now near 140. The average selling target price on Wall Street is $232, still 66% of the price.
But pouring cold water on the situation: 13F is lagging data from the end of June, and after a month and a half, they've already adjusted their holdings. The second batch of restrictions will be lifted on August 20, and there's another big round in December. Musk's 48.4% stake is locked in until mid-next year, so selling pressure is an open card. Q2 lost 540 million but revenue was 7.8 billion, exceeding expectations. But Q1 lost 4.28 billion, and they're still burning cash. Just take a look, don't just copy the trade.The revealing number is not just Harvard Management’s roughly 12.935M SpaceX shares, valued near $2.21B at June 30, but their reported weight: about 51.8% of its public-securities portfolio. That is not Harvard’s full asset allocation, yet it still highlights how concentrated disclosed institutional exposure can appear.
With the FT also reporting Nvidia at nearly 123M shares worth about $21B, ownership visibility is improving just as post-IPO lockup expiries could reshape liquidity. My read: unlock timing may matter more for near-term valuation volatility than the prestige of the holder list. Not advice, just analysis.
#SPCXOwnershipRevealedKey points of the article include: · An AI computing power financing platform exceeding $500 billion, but it is clearly stated that this is not orders or funds already received by NVIDIA, and cooperation is still pending final agreement · 13F disclosures of holdings in SpaceX, Intel, CoreWeave, etc. · Items to continue verifying in the August 26 earnings report NVIDIA's most important move last week: inviting financial institutions into the AI infrastructure construction site The company signed memorandums of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish an independent AI computing power financing platform, aiming to gradually mobilize over $500 billion in third-party capital. Simply put, NVIDIA hopes to make it easier for customers who need large amounts of GPUs to obtain long-term funding. ⚠️ These six institutions are not directly giving money to NVIDIA, nor are these finalized loan contracts. Official documents clearly state that these are currently memorandums of understanding, and final agreements are still pending. NVIDIA's quarterly holdings report also disclosed: holdings in SpaceX and Intel stocks Approximately $20.976 billion in SpaceX stock Approximately $29.989 billion in Intel stock ⚠️ These are quarter-end holdings disclosed last week and do not mean these stocks were all purchased last week. [August 10] Six financial institutions entered the scene Is it worth paying attention to? Yes, for two reasons. First, NVIDIA is starting to handle customer funding issues more directly. Previously, selling chips focused on product performance and$XCH Many people are still discussing when Permuto will launch its products, but they overlook something more important: Permuto's first product has actually already appeared. It is neither a Microsoft stock certificate nor a traditional tokenized security, but a brand-new trading product—svPerps. The first round of testing has ended, and the $20,000 testing reward has been distributed. Next, svPerps is preparing to open trading to eligible non-U.S. users. This indicates that it is no longer just a white paper, concept art, or regulatory document, but is moving from testing environments to real markets. But what exactly is svPerps? Many people, upon seeing "Perps," immediately think it's just another perpetual contract. Go long when BTC rises, go short when BTC falls; Nvidia makes money when it rises, and loses money when it falls. If that were all, the market would already have countless competitors, and Permuto would have no need to redesign a new product. What svPerps truly trades is not the price direction of the underlying asset, but its volatility. This is a completely different concept. Suppose a stock rises from $100 to $110 today, then drops back to $100 tomorrow. Directionally, the price has barely changed at the end, but during this period, it has experienced significant fluctuations. Traditional perpetual contract traders must judge whether it will rise or fall; once the direction is wrong, losses can still occur even if the market is highly volatile. svPerSanDisk’s rally isn’t just about momentum—it’s a valuation reset. 🚀
Investor Day highlighted three major catalysts:
•AI inference is creating structural NAND demand
•$93.9B in long-term contracts improves revenue visibility
•Higher-margin targets are forcing the market to rethink valuation
The bigger story is the shift from a cyclical storage play to an AI infrastructure thesis
That’s why capital is repricing $SNDK so aggressively.
#BTCVolumeDriesUp #SPCXOwnershipRevealed #AIInfraEarningsWatch #SPCX持股结构曝光, Harvard 13F is heavily invested
1. Real-time data
SPCX is currently priced at $140, down 0.91% intraday, with a total market value of $1.8 trillion; Harvard Fund holds 12.9351 million shares, with a market value of $2.21 billion, making it its top US stock holdings. Institutions have simultaneously allocated $BTC spot ETFs, with holdings totaling $101 million; BTC current price is 62,900, with market turnover continuing to shrink and spot ETFs seeing slight net outflows.
2. Core logic
Harvard's long-term positioning indicates institutions are optimistic about the space AI computing power track, and Musk's related narratives have indirectly boosted $DOGE market sentiment; Hashrate and on-chain asset narratives are positive for SOL and TAO. However, SPCX is highly valued, and with large lock-ups approaching, institutional profit-taking will divert risk market funds, suppressing short-term crypto gains.
3. Personal Views
For long-term stability, focus on $TAO and $SOL; Only small positions are used for short-term betting on $DOGE news rally. Currently, market liquidity is weak, overall holding light positions and waiting for signals of capital return.
These represent only personal views and do not constitute investment adviceThe crypto world is saying there will be one last drop in October,
Now, all the brothers in the crypto world who hold U and sit in the pit suffer from a common problem: they wait every day for the legendary last drop, just like waiting for matchmakers to come to the village entrance—poor food, poor sleep, and just checking the app every two minutes.
I couldn't escape either, clutching my wallet tightly. I've long memorized the little notebooks of Ethereum 1200-1500 and SOL 35-60, muttering every day: Hurry up and crash! Smash the market hard! Give me a chance to buy at the bottom! Just waiting to hit a big pit, go all-in on hoarding coins, then lie flat and slack off waiting for the bull market to take off.
But the market's old sixth brother is the best at reading retail investors' minds. While everyone is eagerly waiting for a big drop, it insists on playing hide-and-seek with us $OKB
Looking back at the real drop after the FTX collapse in 2022, no one hoped to buy the dip; everyone was terrified, only frantically cutting losses to escape, with cheap chips scattered everywhere, everyone avoided them, and no one dared to reach out to pick them up. When the ultimate panic hit the moment, fear locked the mind, and the pre-written bottom-fishing plan was completely ruined.
Now the situation is completely reversed, with the entire internet reaching a consensus: there will definitely be one last drop ahead. Everyone is ready, positions are well calculated, and everyone is just waiting to pick up bargains. The main players aren't stupid; how could they actively sell off the market and give away benefits?
Even if there is one last drop, even in October, Ruoshui will think this drop should first lure the price up, get those who missed out to get in, then release huge negative news to crash the market, making retail investors afraid to get in!! Or even hand over their chips!! $BTC $ETH #以太坊 Hegotá 升级：把隐私、抗审查、扩容打包到同一次分叉
8 月 16 日以太坊基金会 Toni Wahrstätter 在 X 上把 2027 年 Hegotá 升级首批筛选名单摆出来：66 个 EIP 进了候选盘，接下来几次核心开发者会议决定最终上不上车。不像 Pectra 主打质押和账户抽象，Hegotá 想一次把三件原本各开一条线的事摞在一起：原生隐私、抗审查、扩容。EIP-8081 是这次升级的 meta EIP，所有候选都挂在它下面。
FOCIL 已经上车，是这场旗舰
FOCIL（EIP-7805，Fork-choice enforced Inclusion Lists）在 EIP-8081 里被列为 Scheduled，是这次升级最先确定入档的提案。机制是让验证者构造一个强制纳入的交易清单，列表里的交易必须被打包，提议者不能绕过去。这是一个抗审查结构，针对 MEV-Boost 主导下、提议者可能受某些地区审查要求绕开特定地址的现实。
EIP-8081 把它定为旗舰是这次升级的方向声明：以太坊要主动把抗审查从口头共识升级到协议层强制。这也解释为什么 @dataalways 在 X 上反复说 FOCIL 这一波不上、下次不知道要等多少年，它不是工程难，是政治决定。
隐私三件套：把隐私从 L2 拉回 L1
候选盘里隐私相关 EIP 有三个：EIP-8141 Frame Transactions、EIP-8250 Keyed Nonces for Frame Transactions、EIP-8272 Recent Roots for Frame Transactions。这套组合让用户能在协议层直接构造不可关联的交易，不必再依靠 L2 隐私协议或混币器。
如果三件套同时入档，$ETH 的叙事会显著变化：从过去两年 L1 做结算、隐私靠 L2 的分工，变成 L1 自己也支持隐私原语。这跟 Circle 9/16 推 ARC 隐私稳定币主网、Tether 完成四大审计一起看，行业趋势是把隐私当成基础设施，而不是边缘功能。
扩容这条还有 Quick Slots（EIP-8198）把 SLOT\_DURATION 12 秒缩到 8 秒，仍属 Proposed；以太坊传统上一次只动一件事，它被推迟到再下一档的可能性不低。
质押通胀改革：Morton's Fork 是这次最难的政治判断
EIP-8363 Tapered Issuance Burn 是这次候选盘里争议最大的一条。机制是随质押率上升、烧掉的验证者奖励比例同步上升，质押率达 50% 饱和时新发行被烧到 100%，相当于过饱和之后不再增发新 $ETH 给质押者，目的是抑制过度质押、防止 stETH 把流通量吞掉。
但这条引入一个被 HTX 和 Galaxy 称作 Morton's Fork 的两难：要么切掉质押奖励把散户验证者挤出网络走向中心化，要么保留高奖励让大机构继续吃利差也走向中心化。Solana 在平行经历同一组抉择，两条链都没找到既要又要的解。
按 EIPs 官方页这条还是 Draft，不是 Scheduled；Galaxy 8/7 报告里也明确说还没被定档或投票。把它放进 Hegotá 候选盘是一次政治试探，不是已经定的方向。
盘面：叙事热度慢慢进来，不是一次定价
$ETH 永续 8/17 11:50 报 1,903.91 美元，24h +1.13%；资金费 +0.0070% 贴零，OKX 单块 SWAP oiUsd 约 13.4 亿美元。FOCIL 入档和隐私三件套入候选的消息 8/16 出来，盘面没立刻跳，Hegotá 是 2027 年的事，市场不会把它当短期催化。24h 涨幅比 BTC 高一倍多，叙事热度是慢慢进来的。
如果你押 Hegotá 节奏，看接下来 9-12 月每次核心开发者会议：FOCIL 是否仍属 Scheduled、隐私三件套到底入档几个、EIP-8363 是升档还是继续 Draft。EIP-8363 的去留决定 $ETH 发行曲线在 2027 之前会不会动。
聊聊3 个问题。
FOCIL 把抗审查写进协议层，在你看是把以太坊的中立性坐实了，还是给监管留下新的冲突点？
原生隐私三件套如果都入档，ETH 上做隐私的 L2（Aztec、Fhenix 那一类）会不会反而被踢出叙事？
EIP-8363 Morton's Fork 你站哪边——切质押奖励防过度质押，还是保奖励防中心化？
$ETH $BTC #Ethereum #Hegotá #FOCIL #以太坊升级#Tether首次完整审计：透明度成焦点
《Tether首闯审计关 储备超债68亿》
Tether 首次完整审计落地，KPMG 出具无保留意见。
过去几年它只发季度 attestation，圈内不少人拿这个当红旗举着，我也翻过几份。这回 KPMG 把四张表全过了一遍，金库金条一根一根实物点过。
8月13日官宣。总资产1877亿对负债1836亿，146吨黄金加9.9万枚比特币躺在账上，Q2营业利润15亿美元，USDT流通量约1846亿，占稳定币市场约61%。
CEO Ardoino 放话："这是稳定币行业的决定性时刻，我们又一次证明他们错了。"
这口气咽得下去吗？Finews 提醒，公告里唯独缺了报表本身。资产负债表和利润表没随公告公开，红旗只放下了一半。
我的复核习惯就两条线。excess reserves 当血压盯，审计口径68.14亿，季报口径41.1亿，比一季度82.3亿腰斩，缓冲垫在变薄。报表公开才算闭环，这步没走完前别下重注。
审计盖章是里程碑，价格只认细节。68亿只是开头。$BTC The US-China AI “pick sides” policy isn’t just geopolitics, it’s now crypto’s most potent macro driver. BTC and ETH no longer trade on tech narratives alone, they’re priced as risk-sensitivity barometers for the AI arms race. Key Data Points: - When US-Iran tensions eased (Hormuz reopened), BTC surged +12.8% ($59,375 → $67,000) in June; ETH hit $1,974, its highest since April. Oil dropped ~7%, 10Y yields fell 42 bps, and S&P 500 rallied +3.6%. - In late July, renewed Iran hostilities spiked oil 最近看 UniSat 和 FB，我发现这 3 件事开始连起来了 👀
最近重新翻了一下 UniSat 和 Fractal 的动态，感觉现在已经不是“今天更新了什么功能”这么简单了。
有几条线，正在慢慢接到一起。
第一，FB 第一次减半越来越近。
很多人第一反应还是会想到价格，但我更想看的是减半之后，供应节奏、矿工和市场参与度会怎么变化。
毕竟这是 Fractal 上线以后第一次经历完整的减半周期。
第二，FIP-102 开始讨论 FB 和 Bitcoin 主网之间更深的连接。
重点不是多发币，而是尝试让 FB 在 Bitcoin 主网上拥有更多使用空间。
如果后面真的推进下去，FB 的定位可能也会慢慢发生变化。
第三，UniSat 做的事情越来越不像单纯的钱包。
从资产管理，到交易、数据、Fractal，再到各种 Bitcoin 原生生态入口，感觉它更像是在慢慢搭一条“用户进入 Bitcoin 生态的路”。
$BTC $XCH At 2 a.m., I recalculated $XCH's market value, and my first reaction wasn't cheapness, but absurdity. An L1 designed by BitTorrent founder Bram Cohen, featuring an independent consensus mechanism, a dedicated virtual machine, an independent asset model, and still developing its underlying protocol, currently has a circulating market cap of only about $20 million. What does this mean? Many projects without mainnet, revenue, institutional products, or even GitHub that are about to stop updating can easily exceed hundreds of millions of dollars in valuation just by labeling them AI, DePIN, or RWA. But Chia has been running for five years, with no network downtime, no code stoppage, and no development team disappearing, yet the market almost prices it as "the project is about to die." But the question is, is Chia really dead? Open GitHub, and the answer is very straightforward. In March 2026, Chia released versions 2.6.1 and 2.7.0, adding new wallet and RPC features, PoS2-related development, and significantly enhancing network attacks, mempools, and node connections. In May 2026, version 2.7.1 will continue to be released. The main repository still has a large number of pull requests, with wallets, full nodes, consensus, security, and development tools continuously updated. A project that has truly been abandoned will not continue to rewrite network defenses, optimize wallet protocols, advance the next generation of Proof of Space, or see token prices fall$BTC $ETH #BTC成交萎缩. Can ETF buying rebound 🌍? Latest international news (8/17 morning–noon) · The Fed has shifted its pricing to a dovish but remains unwavering: CME shows a 66.9% probability of holding rates unchanged in September, with a 33.1% rate hike; Goldman Sachs says a rate hike is "highly unlikely," but Walsh's written statement still sticks to the 2% inflation target. The yield on 10-year U.S. Treasuries peaked at 4.65%, while the 30-year yield fell below 5.2%. The opportunity cost of zero-yield assets has not truly decreased. Hormuz unresolved: weekend traffic plummets (5 boats on Saturday / 0 boats on Sunday vs 31 boats the week before), WTI 82.4, Brent 88.5, Middle East risk premium persists→ tail inflation + risk appetite pressure hits both sides. · US/Japan-Korea divergence: S&P edged down on 8/14 but hit a weekly high, SanDisk +7.4% storage attracted funds; Nikkei rose 0.59% early on August 17, South Korea's KOSPI rose 2.42%, with Asia-Pacific risk appetite warming but crypto desensitized. ETF bleeding: BTC spot ETFs had a net inflow of 865 million yuan from 8/3-7, 131 million yuan outflow on 8/13, 56.2 million yuan outflow on 8/14, totaling a weekly net outflow of 385 million yuan. Institutions are "reducing positions without bottom-fishing." 📊 BTC/ETH real-time market (around 13:00 on 8/17) · BTC: $62,740-$63,100, 24h ±0.1%, down about 3% weekly; The intraday range is 62,780-63,320, 63,000, shifting from support to resistance, with the 50-day moving average at 63,370/2😮 💨 A young woman born in 2005 mustered the courage to step into the cryptocurrency market on Friday, only to run straight into a high wall. Yesterday, the market performed poorly, but today the market rebounded slightly, but the heat was not enough to resolve the position difficulties in hand. She used 8x leverage, buying BICO, hoping to earn some pocket money, but unexpectedly, holding onto her position every day and restless every night, her mental toll was far heavier than she imagined. Looking back at trader data, most bears have already surrendered, while the bulls are barely holding up amid losses. This disconnect reflects the tug-of-war between speculative sentiment and genuine liquidity in the current market: some people are ramping up their bets on a rebound out of fear, while others are quietly cashing in profits during the rebound. 🍂 Two major events pulled her back to reality. First, consumer momentum is weakening, and inflation's constraints on policy easing have not been lifted. Whether marginal easing will occur in September remains highly uncertain. Second, although S&P earnings exceeded expectations, Wall Street seems unconvinced, with a somewhat conservative target level, suggesting limited room for risk assets to rise further. The market is slow to react to positive news but highly sensitive to negative news; this state often indicates accumulating some kind of sentiment swing before a market change. 📉 She also saw another data point pushed by the system: ETF buying has reversed, and BTC leveraged positions have rebounded. This seems like an interesting signal—institutional funds occasionally flow back, but leveraged funds are quietly increasing, indicating that divergence among different capital attributes in the market is widening. Some steadily build positions through spot trading, while others gamble on their own interestsWhat are $SNDK really trading? Don't just focus on AI storage shortages—the answer is hidden in the market data
When many people mention $SNDK, their first reaction is that AI storage is out of stock. This is only surface logic; market data already reveals the core of the real capital game.
First, the market is rewriting its valuation logic. SanDisk relies on long-term supply agreements to lock in large orders, turning cyclical storage business into more stable revenue expectations. Funds are competing to keep its gross margin high going forward, which is the underlying support for premium.
Second, betting on the dividends of next-generation storage technology. It partnered with SK Hynix to launch the HBF high-bandwidth flash standard, positioned between HBM and conventional NAND, adapted for AI inference scenarios, with capital betting early on whether this new standard could open up a whole new hardware track.
Third, the unique capital effect of the crypto market. Among US-mapped tokens, $SNDK contract open interest has long been among the highest, with a large amount of capital unable to directly participate in US stocks and can only enter and trade through the token, with frequent short-term capital inflows and outflows, further amplifying market volatility.
Short-term shortages can only trigger a wave of market rallies; the essence of this round is the result of valuation reshaping, technical expectations, and incremental capital in the crypto world.
This article is only a market review and does not constitute any investment advice. #BTC成交萎缩, can ETF buying rebound? #SPCX持股结构曝光, Harvard 13F heavy position #财报观察员: AI infrastructure earnings report debuts $BTC $ETH $SNDK Fewer coins available doesn't automatically mean buyers are ready to act on it.
$BTC is holding near $62.9K, $ETH sits close to $1.9K, and neither is doing much to inspire confidence right now. What's interesting is the whiplash underneath: August opened with real strength, ETFs pulling in over $750 million across the first week, only for that momentum to reverse into three straight sessions of outflows through Aug 14, including a $57.6M single-day drain.
That's not collapse — it's hesitation. Broader sentiment is sitting neutral to mildly bearish, with short-term traders staying defensive while longer-term holders remain patient. Price is still capped below the $65-66K zone, and nothing in the current data points to a forced squeeze or a clean technical breakout. This reads more like a controlled bounce off support than genuine conviction returning.
The structural story — tightening supply — hasn't gone anywhere. What's missing is the demand side actually showing up with consistency instead of one strong week followed by a retreat.
#BTCVolumeDriesUp #SPCXOwnershipRevealed #AIInfraEarningsWatch
Not financial advice.My view on this morning's spike in SNDK.
Before the European session opened, there was a sharp pull from 1630 straight up to 1775 and then an immediate drop back down. The tactic was too crude. Over the weekend, BTC was sideways, US stock futures were closed, and the perpetual contract depth was pitifully thin. In this environment, a 40+ dollar upper wick was most likely the big players using small funds to knock out the short positions clustered above to trigger stop losses, while also testing the selling pressure at the high level.
As for fundamentals, the original factory's production cut effect is indeed fermenting, and enterprise-level SSD orders are booked through Q1 next year, which everyone knows. But the problem is, this information has been circulating since the June earnings report, and the stock price should have priced it in long ago. The daily chart in the past two weeks clearly shows an ascending wedge pattern, and today's spike just hit the upper boundary, with volume not keeping up—a typical false breakout warning pattern.
My trading view is as follows:
· Chasing at 1775 is a gamble with a poor risk-reward ratio; above 1800 there is a huge volume of trapped positions waiting to be released.
· If I were to participate, I would wait for a pullback to the 1580-1600 range, which is the previous dense trading zone overlapping with the 20-day moving average, where I could try a light long position.
· If it breaks below 1550 directly, then this spike is a signal of a phase top, and it's time to exit.
In short: watch from above 1700, only consider entering below 1600. In this market, watching more and trading less is better than anything else—don't let a single spike throw off your rhythm.😄
$SNDK
$BTC
#消费动能转弱，9月政策仍受通胀制约 来自 Woofun‑AI 的数据显示，8 月 13 号美国比特币现货 ETF 出现了 1.311 亿美元的资金净流出，FarsideInvestors 的数据也证实了这次资金出逃。而且这不只是单日偶然现象，已经是连续第二天往外掏钱，流出的速度相比前一天还明显加快了
这也说明，短暂稳住一段时间之后，市场情绪又开始变得谨慎起来。但大家先要分清一点：这次资金跑路，更多是行情走到关键关口，机构避险落袋为安，并不代表长期上涨的大趋势已经反转。
我们拆开看看钱都从哪里撤走了。方舟投资的 shturl. 流出 5880 万美元，是这次赎回潮里面跑的最多的；紧随其后的就是富达的 FBTC.US，流出 5510 万美元；灰度的 GBTC.US 同样被赎回，流出了 3630 万美元
头部这几个老牌 ETF，就是当天资金流出最主要的来源。之所以 ARKB、FBTC 流出这么猛，很大概率是前期一路上涨之后，一部分机构选择了结获利，也有一部分资金开始重新评估短期宏观层面的风险。连续两天的流出，放大了市场的担忧情绪，短期市场流动性确实迎来了一波考验
不过有意思的来了，虽然大部分 ETF 都在被卖出，但场内资金并不是全部选择离场，分化的迹象已经非常明显
花旗美邦的 MSBT 反倒逆势拿到了 710 万美元的资金进场；灰度的 MiniBTC 更是一口气吸金 3890 万美元。这两笔进场的资金对冲掉了一部分其他产品的抛压
这里就能看出一个很关键的看涨信号：大资金并没有全部抛弃比特币赛道，只是从高价位、高费率的老产品，换到了手续费更低的新品种里面。资金只是在板块内部调仓，而不是彻底撤出整个市场。
之所以出现这种调仓行为，一方面不少机构赚了一波之后选择落袋，另一方面大家都在等着美联储给出更加清晰的货币政策消息。最近比特币一直在区间里面来回震荡，没人愿意贸然开启大行情
所以现在资金从高位老 ETF 撤离，转头布局费率更低的 MiniBTC，本质上就是资金的一次内部轮换。
短期连续流出，看着像是传统金融投资者热情降温，但单日、连续几日的资金流动只能当作短期情绪参考，不能直接判定长线行情走坏
往后我们可以多留意低费率 ETF 的资金动向，如果 MiniBTC 这类产品能够持续不断地吸引买盘，就说明机构长期布局的底盘依旧还在。接下来几天 ETF 的资金流向数据，就是检验这次到底是资金出逃，还是一次换仓洗盘最关键的证据$BTC $ETH
#BTC成交萎缩，ETF买盘能否回暖 今天上午 10x Research 连发两条：先是「$BTC 交易量萎缩、市场进入窄幅盘整」，几个小时后补一句「成交量急剧萎缩、波动区间收窄」。同一个机构一天之内喊两遍，说明这个信号他们已经观察了至少一个交易周期，不是偶发现象。 把数据拉出来看，萎缩是真的。我抓了 $BTC 最近 60 根日线：近 7 日日均成交量只有前 14-21 日的 75%，也就是缩掉了四分之一。最极端的是昨天（08-16），全天成交量 695 BTC，是近两个月单日最低的量级。同期 BIT 数据印证：$BTC 7 日隐含波动率掉到 25%，历史低位。市场不是不涨不跌，是根本没人愿意下注。 有人会问：那 ETF 买盘能不能把这潭水搅活？答案有点冷。上周 $BTC 现货 ETF 净流出 3900 万美元，主导方是 Fidelity 的 FBTC。这意味着什么：机构没有在抄底，反而在减仓——$BTC 现货 ETF 上市以来，资金流向一直是「$BTC 吃肉、$ETH 喝汤」，但过去两个月反过来了：$ETH ETF 资金流连续两个月跑赢 $BTC。$ETH 昨天（12:04）还有新闻说大户成本区在 1900-2400、现一、币种与当前现状 GPS 是 GoPlus Security 的原生代币，主打 Web3 去中心化安全基础设施，为钱包、dApp、公链提供实时链上风险检测、安全审计 API 等服务，属于加密赛道的安全基础设施标的，已上线币安等主流交易所，代币总供应量 100 亿枚。 1. 基本面现状 • 赛道属性：有实际产品落地，协议通过 API 服务产生真实收入，并非纯空气叙事币，但整体流通市值偏小，属于中小盘币种，弹性和波动都远大于主流币。 • 上涨驱动：本轮暴涨以超跌修复+山寨币轮动资金为核心动力——该币历史高点达 0.22 美元，当前价格仅为高点的 7% 左右，长期跌幅巨大；叠加近期加密市场小盘题材股普涨、安全赛道关注度回升，资金借势拉升。目前暂无重大落地级利好公告，情绪驱动属性较强。 • 潜在抛压：代币存在定期解锁规则，2026 年 7 月刚完成一轮千万美元级的代币解锁，后续仍有持续的流通量释放压力。 2. 技术面现状 • 价格强度：最新价 0.014772 USDT，24 小时涨幅 39.35%，7 日累计涨 56.93%，90 日涨幅翻倍，属于典型的短期暴力拉升行情；但大周期仍处于超跌刹不住车了，坠车只是时间问题
数据很漂亮——盈利+31%、PE从26倍降到22倍、华尔街目标7894点（还有1.4%空间），看着像健康牛市。
但真实情况是——VIX年内低位、期权全面转向看涨、人人都在抢call——波动率越低，市场越脆弱——所有人都在同一个方向下注，只要一个意外，踩踏不可避免。
盈利利好已经被定价了，消费数据在崩、通胀预期还在涨、霍尔木兹没落地——市场在最好的消息里，忽略最坏的风险。
极端行情，方向一致，刹不住车，最后的结果只有一个——坠车。 📉#标普盈利超预期，华尔街为何仍谨慎？ $SNDK What exactly are we trading? Don't just focus on "AI storage shortages"—the market's pricing logic has long been reflected in the data on the board.
Driven by Investor Day activities, $SNDK's stock price has surged recently, with concepts such as an 80% gross margin target, 2030 long-term plans, and long-term supply agreements widely promoted in the market.
However, for holding investors, these grand narratives have limited reference value. Core questions worth pondering: What expectations has the current $1,641 stock price already priced in? What signals should be focused on in future earnings reports to judge whether there is room for upward revision in the company's performance?
The core logic can be summarized in one sentence:
The market is no longer just about whether memory chip prices can rise, but whether the company can carry over the ultra-high stage profits from the fourth quarter of fiscal 2026 into long-term sustainable profitability through fiscal year 2027 and beyond.
#闪迪投资者日后股价大涨, long-term goals remain to be verified
#SPCX持股结构曝光, Harvard 13F is heavily invested
#财报观察员: AI infrastructure earnings report debuts one after another
$SNDK $ETH 📰 Lining up to show off your holdings? This is "goods looking for people"!
Harvard heavily invested, Nvidia holds, BlackRock closely followed—institutions rushed to support SpaceX—but the disclosures were June holdings, while the August unlocking had already begun, with a market value of hundreds of billions waiting for competitors.
Not good news, but a bullish outlet 📉
#SPCX持股结构曝光, Harvard 13F is heavily invested #BTCVolumeDriesUp
Bitcoin is trading in an unusually quiet environment. Spot volume has contracted, the trading range is near a multi-month low, and implied volatility remains subdued. Bitcoin ETF demand has also weakened, while Ethereum funds have recently attracted stronger inflows relative to their size. However, institutional interest has not disappeared: UBS reportedly increased its IBIT options exposure and added spot IBIT holdings during the second quarter.
Low volume often means the market is waiting for a catalyst. A breakout without stronger spot participation may be difficult to sustain, especially if stablecoin liquidity continues leaving crypto. Ethereum’s relative strength suggests that some capital is rotating rather than exiting completely. My view is that Bitcoin needs renewed ETF inflows, stablecoin growth and clear spot buying before a durable trend can form. Until then, leverage-driven moves may reverse quickly. At Monday's open, the market's focus was not limited to cryptocurrency trading itself, but rather on the significant divergence between oil prices and Bitcoin's movement. Last week, spot Bitcoin ETFs recorded net inflows of over $800 million, but in the second week, the flow quickly reversed to net outflows. Institutional buying has not continued, indicating that large funds remain cautious at the current price level and have limited willingness to enter. Meanwhile, leveraged long positions in the futures market continue to accumulate, with open interest once rebounding above 760,000 contracts. There is a clear divergence between the withdrawal of funds from the spot market and the rise in derivatives market holdings. This structure usually reflects that although market sentiment is optimistic, the buyer's willingness to invest real money in the spot market is weakening. On the geopolitical front, new uncertainties emerged over the weekend in the Strait of Hormuz. The U.S. side claims that after defeating Iran, it will designate the strait as U.S. territory, while Iran says it has reached a provisional agreement with Oman on passage. Both sides hold their own accounts and do not acknowledge the other's control over the waters. As of now, the strait remains closed, and direct US-Iran negotiations have not resumed. As a result, international oil prices continued to rise, with Brent crude prices breaking through $88 and WTI crude also climbing above $81. Although oil prices remain high, they do not directly affect Bitcoin pricing, but their transmission path is clear: rising energy prices will drive up inflation expectations, thereby compressing the central bank's room for rate cuts. The US dollar and US Treasury yields remain strong, suppressing capital inflows into risk assets. Against this backdrop, Bitcoin's short-term performance may be weaker than the market has been before🛢️ 原油 WTI | $82.21 | 停火博弈撞上宽松共振
先上结论：原油短期结构偏多，但正卡在斐波半分位这道坎上；对BTC而言，眼下是"供给端油价回落 + 金银宽松共振"的中期利好组合，真正的尾部风险是停火破裂。
- 🥇 黄金 $4449，21日 +10.9%
- 🥈 白银 $65.63，21日 +17.1%
- 💵 DXY $99.49，21日 -1.25%
金银比 67.8，跌破70。这意味着什么？如果市场是纯恐慌避险，应该是"买金抛银"、金银比冲上80+。现在是白银涨得比黄金还猛——这不是避险，是"降息预期 → 弱美元 → 商品起飞"的宽松共振，教科书级利好BTC的宏观环境。
翻译成币圈黑话：流动性要放水了，BTC中期吃的是这碗饭。
一句话总结
原油本身偏多但别追——反弹到斐波50%(82.57)遇阻、KDJ76.9高位，站上82.57才看85.88，跌破MA60(81.56)看79.27。
中期偏暖。 弱美元+金银宽松共振+油价供给端回落三重buff，指向降息预期升温。唯一要盯的雷：停火谈判崩了→油价重新冲高→滞胀剧本→那是BTC最不想看到的环境。📉 The US stock market is too aggressive, all the funds have been drained!
Current state of the crypto market
· BTC trading volume has noticeably contracted, with volatility dropping to multi-month lows
· ETF buying is weak, and stablecoins continue to flow out
· BTC spot + ETF daily average trading volume has shrunk significantly compared to the bull market peak
ETH is relatively better—the net inflow ratio for ETFs in July was 9.4 times that of BTC. But that's just right, not that strong.
The fundamental reason is just one — the US stock market is too strong!
The S&P 500 has hit consecutive record highs, AI giants (Nvidia, OpenAI, SpaceX) have successively delivered positive news, institutional funds are all rushing into US tech stocks, and crypto has no new stories, low volatility, poor profitability—who will take the lead?
US stocks are celebrating, crypto is drying up, and before rate cuts really materialize, crypto is a game of stock. Don't expect a broad rise—wait for US stocks to catch their breath before capital may flow back 📉
#BTC成交萎缩, can ETF buying rebound? Expectations for a broad rally have been broken, and a structural market trend is now a foregone conclusion.
Abandon illusions and implement the following configuration:
Core asset: BTC$BTC
· Task: Bottom warehouse defense and countering systemic fluctuations.
· Basis: The 63,000 support level has extremely high chip density, ETF selling pressure is decreasing, and whales are accumulating shares at low levels.
· Discipline: Stay still, don't engage in short-term games.
Offensive asset: ETH
· Task: Flexible game but not triggering a strong attack for now.
· Constraints: Long-term logic remains unbroken (RWA/L2/option betting), but short-term constraints are affected by ETF volatility, AI sector drainage, SOL diversion, and stablecoin high-yield diversion.
· Key Indicator: ETH/BTC Exchange Rate.
Exchange rate battle threshold:
· Below 0.028→ Fully reduce risk exposure, reduce holdings of ETH and Sector Coins, and increase allocation to BTC+ stablecoins.
· Hold above 0.03 → confirm the return of offensive sentiment, gradually increase ETH holdings, and participate in L2/RWA hotspots with small positions.
Position Discipline:
· BTC Bottom → Long-term holding is not negotiable.
· ETH positions→ Positions are built in batches; single heavy positions are prohibited.
· Hot Positions → Trial and error for very small positions, strictly prohibited from chasing gains.
[Final Summary]
This cycle is not the same as that cycle; the logic of profiting by holding with your eyes closed has no longer worked.
#BTC成交萎缩, can ETF buying rebound?
#SPCX持股结构曝光, Harvard 13F is heavily invested
#财报观察员: AI infrastructure earnings report debuts one after another SanDisk’s 70% Rebound: Has the Market Completely Repriced Its Future? SanDisk’s recent move has been remarkable. The stock surged from roughly 1,000–1,190 to 1,775, delivering a massive rebound within just two weeks. The key catalyst behind the move was the company’s Investor Day on August 13, which appears to have significantly changed how the market values SanDisk. 1. Investor Day Changed the Valuation Narrative Management presented an ambitious long-term financial outlook: FY2028–FY2030 revenBTC ETF buying is back, but what we really need to be careful about is that people are once again believing 'it won't fall too deeply'
ETF capital inflows and leveraged positions rebound, of course, are signs of a recovery in risk appetite. But this is where BTC is most easily deceived: spot buying looks stable, but derivatives sentiment can quietly push the market to a more fragile level
I don't like to translate ETF buying orders directly as bull market confirmations
Because ETFs are the slow money entry point, while leverage is the loudspeaker of fast money. When both move in the same direction, the market will go smoothly; But if macro data, interest rate expectations, or oil price risks suddenly reverse later, the first to be forced out is usually not the ETF, but the leveraged position
So this line is more suitable as a framework for observation
It's not about asking if anyone is buying BTC, but about how much money in the buying market is willing to bear the volatility, and how much is just short-term money waiting for the market to continue feeding back
The difference is hugeThe core logic of the current market can be summed up in four words:
Liquid siphons
Funding is extremely realistic. When US stocks like AI, storage (SNDK), and gold (XAU) show overwhelming profit-making effects, capital naturally surges like a tide toward places with higher certainty. That's why BTC now feels like a restaurant with its doors open but no customers waiting for customers—shrinking volume, sideways trading, weak ETFs, and stablecoins continuing to bleed.
But note, institutions haven't left—they're just "changing the venue." In July, the capital-attracting ability of ETH spot ETFs directly surpassed $BTC, which is the best proof.
The breakthrough point next is very clear:
If the AI hype in the US stock market cools down and capital flows back, the current sideways movement will be the buildup period for an epic big move;
If funds continue to be tightly locked in tech stocks, don't expect a full takeoff. BTC will keep playing dead, while ETH and hotspots will rotate locally. Respect the market and go with the flow.
#BTC成交萎缩, can ETF buying rebound? Crypto valuation shifts toward revenue: how should BTC be priced?
I've been staring at this debate for several days, and the more I look, the more I feel it hits the crux of this cycle.
#加密估值转向收入, how is BTC priced?
In the past, when people priced coins, they would just tell a narrative, a story, and set the market dream rate.
AI Agents shouting slogans can reach hundreds of millions, or meme jokes and their market value skyrockets.
But that playstyle has been falling apart recently.
Now the market is forcing every project to produce real things—cash flow, income, and burn speed.
I looked at several recent events, and they all seem to be trending this way.
Goldman Sachs acquired Neos, shifting the focus of crypto ETFs to earnings competition.
Even Tether has conducted its first full audit, with transparency becoming its selling point.
Simply put, money is shifting from "telling stories" to "calculating income."
What about Bitcoin? It doesn't have an income statement, so how is it priced?
My view is that it follows a different logic.
BTC's pricing anchor has never been revenue, but credit, scarcity, and the "leading narrative" itself.
While the whole market is settling scores, it actually highlights BTC's scarcity—impossible to keep track of and unable to settle accounts.
You can't figure out how much it's worth because it's inherently recalculable.
So this round of valuation shifts is a short-term positive news for copycats that can be seen in financial statements, while in the long run, it actually backs BTC's trust.
#加密估值转向收入, how is BTC priced? "Why did the S&P hit new highs and the market play dead?"
Just look at the two pallets facing each other and you'll understand
S&P closed at 7,786 on Friday, year-end target at 7,894, just over 1%. Earnings are real. Q2 year-on-year +31%. Full-year forecast raised from 15% to 27%. P/E ratio dropped from 26 to below 22. Making money by making money and making up for valuations is the only way to reach new highs
Consumer spending is not matching July's retail sales -0.6, confidence dropped from 55 to 51, inflation expectations are still at 4.3, and it's hard to increase or lower in September
AI is even stranger, the numbers look good, stock price doesn't give face, Lumentum's revenue doubled, Cisco's AI orders reached 9.3 billion, AMD just issued its largest $4.75 billion bond in history, Nvidia raised OpenAI's guarantee from $250 billion to less than $120 billion, Jane Street lost about $15 billion in July, the first monthly loss in a decade, half of Harvard's 13F public position is under SPCX, money hasn't left, changed to bond issuance and slashed collateral, no IPO
Bitcoin is now at 63,500, last October it was 126,200, started the year at 90,875, down 30% now. On the last 12 weekly charts, 11 closed between 6.0 and 66,000. ETH has rebounded 62% from 4,958 to 1,902.
For Bitcoin to move, wait until interest rates ease or if someone actually buys spot stocks. It's not the same as whether the S&P will close to a new high tonight
#标普盈利超预期, why is Wall Street still cautious?
#BTC成交萎缩, can ETF buying rebound? HYPE | Revenue shrinkage and selling pressure keep combined: How much longer can the buyback narrative last?
The real change was that on 8/14, on-chain monitoring saw a whale (after redeeming 2.886 million tokens from staking at the end of July) transfer 923,743 HYPE (about $53.02 million) to Coinbase Prime and FalconX, selling about 1.956 million (about $110 million) over two weeks, with a combined profit of about $109 million, still holding about 969,000 tokens. Combined with HyperLabs' 8/10 sales of 400,000 tokens, core contributors unlocking 10 million tokens on 8/6, and the P0 tracking item scanned on 8/12 (14-day net selling verification), new on-chain evidence was obtained.
Supporting evidence:
On 8/14, 924,000 tokens were transferred on-chain (Lookonchain monitored and redirected, original on-chain verifiable).
Protocol revenue for the first four weeks of Q3 was about $45 million, with quarterly revenue at about $150 million on a pace-based basis, marking the fourth consecutive quarter of decline (about $202 million in Q2'26, compared to a peak of $357 million in Q3'25, down about 43%).
The market share of external HIP-3 developers rose from about 2% at the beginning of the year to nearly 50%, and developer revenue share increased from 6% to 18% within one year; Platform retention revenue and buybacks (about 97% of transaction fees are on buybacks) weakened simultaneously.当K线失去波动，聪明钱正在干什么？亚盘午盘的极度平静下，$BTC在6.3万美元上方维持区间，散户在无聊中等待方向，但水面下的机构动作正在重塑行业的底层逻辑。今天不聊技术面，我们透过最新行业重磅新闻，拆解横盘期背后的范式转移。 ══════════════ 📌 【宏观与情绪背景】 📌 【$BTC 价格与市占率】$63,445 | 24h +0.51% | 市占率 58.67% 📌 【全市场总市值】$2.159万亿 | 24h +0.42% 📌 【恐惧贪婪指数】31 | 状态：Fear 在恐贪指数徘徊于“恐惧”区间的横盘期，$BTC市占率高达58.67%，山寨币流动性枯竭。这并非单纯的技术性休整，而是资金在等待合规与机构化落地的“发令枪”。 ══════════════ 📰 【事件一：3倍 $BTC ETF穿上合规外衣，交易所退居后台】 据深潮TechFlow报道，3倍比特币ETF即将推出，最危险的加密杠杆正从离岸搬进美国证券账户；同时，传统银行正把加密交易塞进自家App，交易所正在退到金融系统后台。 💡 深度解析： 这不是简单的产品创新，而是定价权与分发渠道的彻底交接。当银行Just now, $SNDK surged after hours. At this point, some might wonder if the US stock market is about to return to a bull market. I don't think so; I believe this round of US stock rally is just a bullish catalyst. Why? Because Japan's rate hikes could burst the current bubble in the US stock market at any time. The previous joint intervention by the U.S. and Japan in exchange rates proves that the U.S. is also concerned about this issue. —————————————————— Let's take a closer look at the impact of yen rate hikes on US stocks. For a long time, there has been a stable interest rate gap between Japan and the United States. Currently, Japan's policy rate is 1.00%, the Federal Reserve's rate is 3.50%–3.75%, the 10-year US Treasury yield is about 4.686%, and the 10-year Japanese Treasury yield is about 2.846%. In other words, the US-Japan policy spread is about 250–275 basis points, and the 10-year Treasury yield spread is about 184 basis points. Because of the interest rate differential, a stable arbitrage channel has emerged. Institutions in Japan absorbed the yen at about 1% cost→ converted into US dollars in the foreign exchange market→ buying 10-year US Treasuries with yields of about 4.686%. This is a very stable arbitrage path. —————————————————— In the past, the Bank of Japan maintained zero interest rates for a long time. The arbitrage model I mentioned earlier had very high profits, so the scale was very large. But times have changed. Now that the Bank of Japan has raised interest rates to 1%, arbitrage has become problematic. Although the nominal interest rate spread remains hugeBTC|The reason for outflows has disappeared, but the reason for re-flows has not yet appeared
The real change was that the SEC's vote on the Reg Crypto rule proposal on the morning of August 14 was temporarily canceled and no new date was announced, causing the regulatory negative news originally priced by institutional funds to fail; Subsequently, BTC spot ETF funds shifted from continuous outflows to moderate returns.
Supporting evidence:
The SEC's originally scheduled committee meeting on 8/14 and Reg Crypto voted to cancel or postpone it, with multiple sources aligning (KuCoin, dev.to, Sina Finance), with no official new date.
ETF capital flows: 8/12 Net outflow $61.16 million (SoSoValue, IBIT -14.34 million); 8/13 Net outflow $61.1 million (Farside, FBTC -46.8 million leading); 8/15 Net inflow $11.11 million; 8/16 Net inflow $36.01 million (FBTC +61.3 million, IBIT +20.4 million, ARKB +13.4 million, BITB +12 million, EZBC +1.7 million, GBTC -72.9 million).
Macroeconomic side (within 150 words): US retail sales unexpectedly declined, rate hike expectations weakened; Coinbase BTC premium negative for 90 consecutive days, miner open interest declined, and spot demand signals are weak.当加密资产与传统股票在同一个账户里纠缠，一条简单的行情变动，就可能把人的情绪推向失控的边缘。最近，有人在社交平台上讲述了自己的操作困境，叙事不长，但信息量不小，尤其是那种从自信到慌张的心理转变，很值得复盘。 这则帖子的核心场景并不复杂。主人公以OKB作为抵押物，借入资金，去做空了SanDisk，一支传统美股存储巨头。听起来像是一套跨市场的对冲打法，加密资产做抵押，传统股票做标的，逻辑上确实有想象空间。但市场的运行恰恰不照顾任何人的个人剧本。就在帖子发出的那一刻，OKB的价格在下跌，而SanDisk的股价却还在上涨，两头相夹，浮亏已经来到18000美元。 这笔亏损本身不算天文数字，但真正刺痛人的，是资金结构出了问题。用OKB抵押借贷，等于把资产的安全边界交给了抵押物的价格波动。OKB一旦下跌，抵押率就会走高，轻则面临追加保证金的压力，重则可能触发强制平仓。而另一边，做空SanDisk的头寸正在承受相反的走势压力。两种资产的相关性并不强，本意或许是想分散风险，但实际操作中，这种组合反而让账户承受了两份波动叠加的压力。 从市场心理的角度看，帖子里的情绪演变很有代表性。一开始写“哈哈”，语气还$ETH Subtle shift in institutional sentiment | DWF Labs: By fund size proportion, ETH spot ETFs have outflowed BTC in capital flows since June
On August 17, DWF Labs published statistics on X: Based on the fund's own management scale, since June, ETH spot ETFs have significantly outperformed BTC spot ETFs.
📊 Core Data:
- June: ETH ETF net outflows accounted for 4.65% of fund size; BTC ETF net outflows accounted for 8.09% of fund size
During the pullback phase, ETH products face less redemption pressure and stronger resistance to selling pressure.
- July: ETH ETF net inflows accounted for 3.19% of fund size; BTC ETF net inflows accounted for 0.34% of fund size
ETH's relative inflow rate is 9.4 times that of BTC.
⚠️ Key distinction: it is a relative lead in share, not an absolute dollar surpass. BTC ETFs have a larger market, with total inflows into USD still higher, but the tilt of new allocations is shifting toward ETH.
An interesting point of view:
DWF Labs' judgment in May was still that overall institutional interest in ETH was low, with funds continuing to weaken. In just the past two or three months, clear signs of a shift in capital flow have appeared in recent weeks.
How to interpret this signal
1. Institutions are no longer solely betting on BTC, but are now rebalancing internal crypto positions, with ETH staking yields and DeFi narratives regaining institutional attention.
2. There is a time lag between capital signals and market trends. Currently, BTC and ETH are still in a fluctuating bottom, and the capital bias has not yet fully reflected in coin prices or the ETH/BTC ratio.
3. Only two-month phased data cannot directly define the long-term trend; subsequent consecutive months of data are needed for confirmation.
Constraints need to be considered
ETH option ETF approval postponed to November lacks the backing of options tools; Fed rate cut expectations and regulatory uncertainty continue to dampen institutional willingness to increase holdings.
$BTC $ETH
#ETF资金观察今天最重要的变化是 BTC 资金面完成一轮"监管避险流出—取消投票后温和回流"的小循环，但回流成色不足：SEC 8/14 临时取消 Reg Crypto 投票后，BTC 现货 ETF 从 8/12–8/13 连续净流出约 1.2 亿美元转为 8/15–8/16 合计回流约 4,700 万美元；8/16 单日 3,601 万，其中灰度 GBTC 流出 7,290 万抵消过半。利空延后而非取消，监管真空未解除，机构回流仍是试探性的。
第二件事是 HYPE"收入缩水与抛压连续"的剪刀差获新链上证据：8/14 巨鲸解押后再度转出 92.4 万枚（约 5,300 万美元），两周累计卖出约 196 万枚；Q3 前四周协议收入约 4,500 万美元，按节奏将连续第四个季度下滑，回购支撑同步减弱。今天 11 点 46 分，OKX 新闻流弹了一条「ETH 现货 ETF 上周净流出 226 万美元」，BlackRock 主导。十二点零四分又来一条「分析师：ETH 大户平均成本集中在 1900 到 2400」。同一天，08 点 22 分链上数据里，有鲸鱼从 Kraken 提走了 5300 枚 $ETH，按当时价格大约 998 万美元——通常这种规模的提币，目的不是卖，是去质押或者进冷钱包。 两条新闻、一笔链上数据，方向看着相反，但拼到一起才看得懂图。 把 $ETH 想成一块油田，「ETF 净流出」是机构在井口用管道抽水卖，「鲸鱼提币」是有人把抽水机拆了装进自己仓库。前者短期对价格是拖累，后者中长期是在悄悄降低市场流通筹码。同一天发生这种事，往往不是巧合：ETF 那头流出来的量级太轻（226 万美元，分到七天每天才 32 万），远小于一级鲸鱼一笔提币；机构的小动作和鲸鱼的大动作在用不同的逻辑运作。 把这层拆开看，今天 1900 这个位置就值得多看几眼。$ETH 现价 1904 站在所有均线上方一丢丢（MA5 1887、MA10 1892、MA20 1888），三条线缠得很紧——这意味着ETH数据三：大账户的成本
第一篇我们分析了投资者行为，第二篇又拆解了筹码结构，但对于一个主流币种这还不够完整；今天我们再来聊聊——“成本”。
我把持有大于100枚ETH的群体都拉出来看，他们的平均成本分别如下：
1、持有100-1k：$1,900；
2、持有1k-10k：$2,000；
3、持有10k-100k：$2,100；
4、持有>100k：$2,400；
也就是说，ETH的高净值人群和鲸鱼群体组成了一个成本带，就在$1,900-$2,400之间。
在牛市中，价格会正向偏离该成本带；反之，在熊市中也会负向偏离。偏离越严重，均值回归的需求越强烈。
所以，假设我们在ETH价格低于成本带时开始定投，回到成本带时停止，能在风险可控的前提下，确保不错过周期的红利。
如果再结合LTH-NUPL —— 当指标进入红色信号区（LTH-NUPL < 0）,意味着LTH的情绪已处于极端绝望状态。
从过往10年的数据来看，该策略的胜率几乎是99.99%。
在昨天的评论区，看到有小伙伴说：“ETH还能回到$2,700？”。
我看到也忍不住笑了...... 真的是爱之深，痛之切！
本轮周期的以太坊因种种原因，确实让太多人失望不已。对它的未来没有确定性预期。
但也不至于如此悲观，尤其是现在。$7,200的ETH或许难看到，但$2,700的ETH，我并不怀疑。BTC shows an advantage in ETF demand, while ETH shows a close flow in institutional inflows. Both assets are in the stage before trend confirmation, so what criteria will the market use to determine direction? As of August 17, BTC is around $63K and ETH is below $1.9K, with neither asset confirming a new trend. The key issue is the direction of funds. BTC has seen solid capital inflows centered on ETF demand, while ETH has shown relatively sluggish price responses despite continuous institutional inflows. This is not simply bearish, but signals that funds are allocating different roles to the two assets. - Capital flow: BTC ETFs maintain continuous net inflows and absorb defensive demand, while ETH is receiving institutional funds but has low price elasticity. - Price Levels: BTC at 63K and ETH below 1.9K are neutral for both assets, indicating that movement within the range is more dominant than directionality. - Momentum Basis: BTC at 65K $CORE 从“数字黄金”到“生息资产”：CORE机构版上线，对比特币的长期价值与短期局限 ⚠️风险提示：本文仅为行业信息交流，不构成任何投资建议。 近期CORE推出面向专业资本的机构解决方案，主打合规BTC质押、lstBTC流动性服务，定向对接托管机构、资管公司与家族办公室。客观拆解这条消息的长线价值与短期预期。 长期积极逻辑 1、直击机构核心痛点：大量机构持仓BTC长期存放冷钱包，缺少合规渠道产生收益。CORE联动BitGo、Hex Trust等头部托管服务商，资产无需转出托管体系，依托时间锁质押实现BTC生息，无需跨链封装WBTC。成熟的生息方案，有望提升传统资本配置比特币的意愿。 2、完善BTCFi叙事体系。长久以来比特币更多作为数字储值资产，金融应用场景偏少。机构工具落地后，BTC可参与质押、借贷、流动性凭证发行，进一步拓宽比特币在传统金融领域的接受度。 3、优化筹码结构。机构持有者不再只能依靠低买高卖获利，稳定质押收益会促使长线资金减少短线抛售，中长期有望缓解现货抛压。 需要理性看待的短期制约 1、机构商务落地存在漫长周期。风控审核、系统对接、资金策略调#标普盈利超预期，华尔街为何仍谨慎？
二季度标普成分股盈利大幅超出市场预期，但华尔街机构并没有顺势上调激进目标，整体态度偏向保守，业绩亮眼和预期克制形成明显反差。
本轮盈利高度集中在AI算力龙头，指数大部分利润由少数巨头贡献，多数企业改善有限，市场结构分化严重。同时不少收益来自股权投资等非主营业务，并非全部来自经营主业。
机构顾虑主要有三点：一是通胀粘性仍存，降息节奏不确定，高利率环境压制估值进一步扩张；二是大厂AI资本开支持续加大，巨额投入能否持续转化为利润，还需要后续财报验证；三是指数已经处在历史高位，一旦AI景气边际回落，存在回调风险。
映射加密市场：美股盈利强代表整体风险偏好尚可，但机构保守意味着增量资金有限，很难大量外溢到币圈。美股新高不等于BTC必然大涨，只能作为情绪参考。
个人观点：美股现在靠企业盈利托举行情，而不是估值泡沫驱动。加密盘面更多看ETF资金流向与自身现货买盘，不要直接照搬美股走势做交易决策。ETH Data 3: Cost of Large Accounts
In the first part, we analyzed investor behavior, and in the second, we broke down the structure of the token, but for a mainstream coin, this is still not complete enough; Today, let's talk about "cost" again.
I analyzed groups holding more than 100 ETH, and their average costs are as follows:
1. Holding 100-1k: $1,900;
2. Holding 1k-10k: $2,000;
3. Holding 10k-100k: $2,100;
4. Holding > 100k: $2,400;
In other words, the high-net-worth group and whale groups form a cost belt for ETH, typically between $1,900 and $2,400.
In a bull market, prices deviate positively from this cost band; Conversely, in a bear market, there is also a negative deviation. The more severe the deviation, the stronger the demand for mean reversion.
So, suppose we start regular investment when ETH falls below the cost zone and stop when it returns, ensuring that we don't miss the dividends of the cycle while risk is controllable.
If combined with LTH-NUPL — when the indicator enters the red signal zone (LTH-NUPL < 0), it means LTH sentiment is in an extremely desperate state.
Looking at data from the past 10 years, this strategy has a win rate of nearly 99.99%.
In yesterday's comment section, I saw someone say: "Can ETH really get back to $2,700?" ”。
I couldn't help but laugh when I saw it...... Truly, the depth of love and the pain are so intense!
For various reasons, Ethereum in this cycle has indeed disappointed many people. There is no certainty about its future.
But it's not to be so pessimistic, especially now. $7,200 ETH may be hard to see, but $2,700 ETH—I don't doubt it.Stablecoins become more like banks, and $BTC are more like true off-chain assets
The more stablecoin regulation advances, the more it resembles financial institution liabilities. Customer identification, anti-money laundering, reserve requirements, licensing issuance, regulatory filing—these factors make stablecoins safer, more compliant, and easier for institutions to accept. All sounds good, but it also means stablecoins are becoming less like free assets and more like on-chain extensions of the dollar system.
This is easy to understand. Stablecoins are not essentially anti-dollar; they are digital dollars. They rely on reserve assets, bank accounts, short-term debt, issuer credit, and regulatory licensing. Users use stablecoins to make it easier to use the dollar, not to escape from it. The more compliant stablecoins are, the more they are incorporated into the existing financial order; The more included they are, the less likely they are to take on the role of "off-system assets."
$BTC Here's where the difference comes in. It has no issuer, no reserve account, no licensing entity, and no customer identification process built into the protocol. You can buy and sell it through a compliant platform, but the protocol itself isn't a financial institution's liability. This attribute might not be so obvious in normal times, since most people only care about price; But as stablecoins become more banked, $BTC's non-bank nature becomes clearer.
This is not to say stablecoins are bad. On the contrary, stablecoins may be among the most successful applications in the crypto world. They solve issues such as transfers, transactions, global dollar flows, and on-chain cash layers. Without stablecoins, liquidity in the crypto market would be much worse. The problem is that the success of stablecoins and the success of $BTC come from different needs. One needs stability, the other needs scarcity; One requires compliant flow, the other needs non-sovereign reserves.
In the future, on-chain finance may have a clear hierarchy: stablecoins handle payments and cash, RWAs handle yields and real-world asset mapping, $BTC are responsible for long-term scarce reserves. Once this structure is formed, $BTC doesn't need to be the most commonly used payment tool; it only needs to be the hardest hard asset layer to replace.
So today, when looking at stablecoin regulation, don't just look at the impact on USDC, USDT, or payment companies. Instead, look at how it reshapes the division of on-chain assets. As digital dollars increasingly resemble banking products, the reason for $BTC's existence becomes clearer: it's not a more usable dollar, but an option beyond the dollar.
Stablecoins bring crypto into the financial system, $BTC allow crypto to retain imagination outside the system. Without the former, crypto is hard to scale; Without the latter, crypto may be left with only a faster dollar network. 今天中午 12 点 27 分，OKX 新闻流弹了一条短消息——Hyperliquid AQA 预计每年贡献 2 亿美元回购 $HYPE 。 数字有点吓人。把 Hyperliquid 想成一个大菜市场——HYPE 就是这个菜市场的「招牌」，现在市场管理方说：每年从摊位费里抽 2 亿美元出来，专门砸自己的招牌。这事儿在过去两年的加密圈子里并不常见，传统的回购多见于股票和传统 DeFi。 HYPE 自己确实接得住。链上摆着的：累计已经销毁 4771 万枚代币，总营收 12.4 亿美元，Hyperliquid 的原生永续合约是主要贡献者。8 月 16 号还有个数据：RWA（现实世界资产）独立用户群，80% 以上的新用户是从别处迁过来的。这意味着 HYPE 不只是交易所在烧钱，它本身在变成一个「结算层」——多空双方把交易押到这里，市场方再拿交易费回来买自己的代币销毁。 钱是真的，价也真涨了。问题是涨到这儿还能不能进。 先看今天的状态。$HYPE 现价 58.9 附近，24 小时涨 3.4%，是今天主流币里最显眼的那只。MACD 的红柱今天扩大到了 +1.54，RSI6 跳到 68.7——这些都