Orbit Post Sitemap
Exchange balances continue to decline, but why do BTC and ETH still struggle to rise?
Many people equate the continuous outflow of exchange tokens with the market about to start soon.
A drop in exchange balances means tokens are withdrawn from the trading platform and transferred to personal wallets. Selling pressure is indeed reduced, but it does not mean an immediate rise.
Current situation:
$BTC large amounts of chips left the exchange and were picked up by ETFs, listed companies, and long-term whales. These chips are almost never sold off due to short-term fluctuations, so it's unlikely for a catastrophic drop to occur on pullbacks, and the bottom is very resilient.
$ETH also see exchange balances fall, but a large amount of withdrawn tokens will be transferred to staking contracts, while others will flow into DeFi lending, indirectly converting into potential selling funds in the market.
There is a key misconception here: withdrawing chips from the exchange only addresses downside risk; When the market rises, it must rely on newly added off-exchange funds actively entering and buying.
Currently, selling pressure is easing, but buying is insufficient.
Chips are locked up, only holding the bottom; To rally, new increments are needed.
Looking only at the exchange balance as a single indicator can easily misjudge market trends.Hynix leveraged ETF trading volume continues to evaporate, South Korean retail investors' deleveraging is nearing its end
A few days ago, I wrote about the retreat of South Korean retail investors from Hynix leveraged ETFs. At that time, Bloomberg's statistics on 7 SK Hynix leveraged ETFs showed that trading volume had dropped from nearly 800 million shares at the end of July to below 100 million shares.
The latest data further confirms this trend. Among them, the KODEX SK Hynix single-stock leveraged ETF had a single-day trading volume exceeding 600 million units at the end of July, but quickly fell to tens of millions of units after entering August, shrinking by more than 90% compared to the peak.
This indicates that the most aggressive round of deleveraging in the South Korean market is indeed approaching its end.
After raising the minimum cash margin at the end of July, a large number of retail investors who relied on financing to repeatedly trade single-stock leveraged ETFs have exited. Coupled with nearly 1 trillion KRW in forced liquidations over the previous two months, the leveraged positions that can still be forcibly liquidated have significantly decreased.
However, this may not be entirely good news for Hynix. Previously, these leveraged funds created forced liquidation pressure during declines and were also very important incremental buyers during rallies. Now that the trading volume has dropped from hundreds of millions to tens of millions of units, the risk of chain forced liquidations has indeed decreased, but the funds willing to chase gains and bottom-fish with high leverage have also disappeared.$LINK. $HYPE. $DOGE Sentence to death in one sentence:
$LINK: For those who really get to work, RWA+AI is a real need for utilities. No one wants to spend money, so it's perfect for a fixed deposit on the eve of a bull market.
$HYPE: Casinos distribute dividends, the fees are really burning, and buybacks are real, but big players control the market until you're sick of it. If you're timid, don't get involved, and if you do, don't shout.
$DOGE: Pure Musk's electronic pet—fly away at a call, forget and reset to zero. The consensus is retail investors, and the end is buying the share.
Right now, BTC is shrinking and playing dead. Don't chase HYPE's peak, and don't date DOGE.
LINK slowly absorbed, HYPE only used sentiment divergence points, DOGE kept 1% U and other Twitter "dog barking" moments.
Remember: LINK is an asset, HYPE is a token, DOGE is a lottery.
Don't treat lottery tickets as assets to buy, don't treat chips as faith 🥃SpaceX没有上车还能上车吗？？？
基本面层面，星链是唯一稳定盈利板块，用户持续增长；AI算力、星舰项目短期资本开支巨大，持续烧钱，二季度资本开支高达183.69亿美元，市场分歧巨大。投行目标价分化严重，最高看300美元，最低给到115美元，多空逻辑同时存在 。短期最大风险来自8月底即将到来的第一批解禁，流通筹码翻倍预期压制上行空间 。
盘面数据上，当前映射代币24小时成交额4264万美元，账户多空比1.7，多头情绪偏强，近期自低点104一带反弹至141附近，属于超跌后的修复行情，但航天赛道目前缺少新热点催化，板块整体热度偏低。
分两种思路看待上车时机：短线不要追反弹，等正股回踩关键支撑、热度降温后再博弈；波段布局，要等待星舰发射、星链新订单、AI业务营收超预期这类明确催化落地，再分批介入。映射合约不适合长期死拿，一旦正股趋势走弱，代币下跌速度会更快。现阶段没有高盈亏比的上车窗口，耐心等待更好的价格位置或者利好信号出现。
本文仅行情复盘，不构成任何投资建议#SPCX持股结构曝光，哈佛13F重仓 $BTC $ETH #AI押注受挫, Wall Street trading giants lost $15 billion in the month
Wall Street has been quite shaken this time
What does it have to do with the crypto world? Two layers.
The first layer: liquidity has been drained. The SA fund's $35 billion worth of assets was liquidated, Jane Street lost $15 billion, and all AI-related leveraged positions on Wall Street are being forced to close. This money won't flow into any risk assets in the short term, and the crypto market certainly won't get a share.
The second layer: risk appetite is fading. Even top market makers lose 15 billion yuan in a month, indicating that the risks in AI trading are greater than people think. Such losses will directly shift institutions from offense to defense, with crypto, as a highly volatile asset, bearing the brunt.
Let me share my thoughts.
The SA fund's approach is no different from the crypto world's all-in all-in betting on the same track and leveraged bets on direction. The only difference is it betting on SanDisk and Micron, while crypto betting on Bitcoin and altcoins. The underlying logic is exactly the same—get rich in the right direction, go back to zero if wrong.
Jane Street's $15 billion loss is a short-term double suppression of liquidity contraction and declining risk appetite, which is not good news for the crypto world. But in the long run, the more thoroughly Wall Street deleverages it, the healthier the funds spillover into the crypto market in the future.
The current volatility in the market is not large, which is closely related to this issue. Wall Street is deleveraging, and without incremental funds here, they can only keep grinding. So, be patient and wait; rushing won't help
$BTC $SNDK $ETH 共识机制本是技术路线之争，如今却成了税收政策的分水岭——这正是美国 PARITY Act 点燃的争议核心。 这份由两党议员推动的数字资产税收法案，允许"被动质押者"将质押奖励的纳税义务递延最多五年，直到出售时才确认收入。问题在于，修订版草案把挖矿奖励明确排除在递延资格之外。理由看似简单：PoS 验证者几乎没有经营性开支，而 PoW 矿工天然要承担电力、矿机和基础设施成本，于是被定义为"经营性主体"，不符合"被动"条件。 结果就是一个罕见的政策分裂：同样是为网络提供安全保障，ETH 验证者可以选择何时缴税，BTC 矿工却必须在奖励到账当年就按市价纳税——哪怕币还没卖、甚至还没流动性。Bitcoin Policy Institute 直接批评法案名为"平等"，实则在"挑选赢家和输家"。 这背后是两种成本结构的碰撞。$BTC 的成本来自电力、矿机和算力军备竞赛，是持续流出的真实现金；ETH 的成本来自质押资本、验证者运维和机会成本，更像一种资本占用。当税法开始识别这种差异，影响就不会止步于合规层面：矿工税负若持续偏重，会挤压 BTC 生产端的利润空间；而 PoS 获得递延待遇，等于为 $E$XAU 🟡【黄金8/17｜资金面开始站队】
我的判断：黄金短线偏多，但还没到无脑追多的时候。
现在最重要的，不是“黄金涨了多少”，而是三个信号开始同时出现：
1️⃣ 利率预期转向
上周零售销售、消费者信心等数据明显偏弱，市场对9月加息的押注从接近50%降到约30%。
美元走弱、实际利率承压，这才是本轮黄金重新站上 $4,380 附近的核心燃料。
2️⃣ 资金开始回流贵金属
美国黄金ETF资金明显改善，中国/印度ETF持续流入。
更关键的是：中国央行7月继续增持黄金，已经连续第21个月。
这意味着黄金现在不只是“交易降息”，还有一层越来越强的官方资金托底。
3️⃣ 技术面终于给出确认
黄金重新站上100日均线≈$4,387。
目前正在测试心理关口：
$4,400
如果有效突破：
→ $4,450
→ $4,504（200日均线）
反过来，如果跌回$4,387下方：
→ $4,311
→ $4,202
所以现在真正值得盯的，不是预测黄金明天涨还是跌，而是：
$4,400能不能从“阻力”变成“支撑”。
⸻
⚠️ 但这里有一个很大的隐患：
市场只是降低了“9月加息”的担忧，并没有完全消除“年内加息”的风险。
而8/27–29的Jackson Hole才是下一颗真正的雷。
如果Warsh释放明显偏鹰信号：
美元↑
美债收益率↑
降息预期↓
黄金可能迅速回吐。
所以我的短线框架：
🟢 基准：$4,320–4,450震荡偏多
🟢 突破$4,400：看$4,450→$4,504
🔴 跌破$4,387：警惕$4,311→$4,202$SanDisk's weekend saw the market waiting for the opening on Monday
Last week's SanDisk Investor Day delivered impressive results—mid-to-high double-digit revenue growth from FY2028 to 2030, gross margin 80%, operating profit margin 75%, free cash flow margin about 50%, and 100% excess cash return to shareholders.
More importantly, the NBM long-term agreement. SanDisk has already signed long-term contracts with eight clients, with a maximum term of five years and guaranteed revenue of $93.9 billion. There is also a $16.5 billion performance bond as a guarantee. Simply put, it's about transforming a storage cycle stock toward a "subscription-based revenue" model.
The market's reaction last week was very direct—Investor Day rose nearly 14% that day, then another 7% the next day, closing at $1641. Goldman Sachs set a $2200 target price, JPMorgan Chase set 2250, and Cantor Fitzgerald even shouted 2900.
But one detail worth noting—US stocks were closed over the weekend, and SNDK had no new trading prices. However, xSNDK/USDT on the platform kept moving and had already surged above 1720 before the market opened Monday morning.
What does this indicate? The market is rehearsing its moves. No trading opened over the weekend, but sentiment hasn't cooled, and funds are pricing in early.
Monday's opening will be the real test.
The core logic behind this rally is: NBM long-term contracts have transformed SanDisk from a cyclical stock into a "quasi-subscription" company. But the market's acceptance of this logic ultimately depends on the performance of the underlying stock after the open—whether it continues to push upward, surges higher, then retreats and takes profits.
The $93.9 billion contract is real, and the 80% gross margin target is still just the goal. Monday's session will provide the first real answer.
#闪迪长期协议成焦点, the opening performance remains to be seen SEC把代币化股票的“绿色通道”又关了。
原定8月14日公布的代币化证券“创新豁免”再度推迟，同一天的加密规则制定会议也直接取消。Bullish跌了8%，Figure跌9%，Coinbase跌2%，Circle跌近4%。消息出来，代币化概念股集体跳水。
豁免这东西，原本是给代币化美股开合规通道的——允许链上24/7交易苹果、特斯拉这类股票，免去完整的经纪商注册流程。
为什么又关？两股力量同时施压。
白宫担心这时候放行会干扰CLARITY法案的立法进程。华尔街那边SIFMA也在反对，认为去中心化交易场所和自动做市商没法兼容现行证券规则。
CLARITY法案本身也悬着。9月15日参议院程序性投票，需要60票才能推进。但Galaxy已经将通过概率下调到10%。Polymarket上跌破20%。
白宫怕捅马蜂窝、华尔街怕规则不兼容、CLARITY法案本身也悬着——三重阻力，卡住了代币化美股这条通道。 The current cryptocurrency market shows a clear pattern of structural differentiation, with mainstream assets diverging from altcoins. Market observers point out that large-cap blue-chip coins are relatively stable, while small- and mid-cap coins are highly volatile, making a broad rally unlikely to occur in the short term. Against this backdrop, capital management strategies are shifting from one-sided bets to layered allocation. $BTC are considered market bottom positions, and unless extreme conditions occur, positions are not easily adjusted. Data shows that Bitcoin is trading well around $63,000, and the ETF funds that had been continuously flowing out have recently narrowed significantly, indicating that large funds are gradually accumulating in the bottom area. Analysts believe that if the overall market remains volatile, Bitcoin's resilience will be significantly stronger than most public chain tokens and highly volatile altcoins, and holding a base position offers a higher margin of safety. $ETH, the logic of long-term investment has not fundamentally changed. The RWA (Real-World Asset) and Layer 2 narratives still have room for imagination, and signs of capital allocation are also emerging in the options market. However, in the short term, ETH faces multiple pressures: ETF funds inflow and outflow are unstable, some off-exchange funds flow into the AI sector, $SOL continue to attract speculative sentiment, and a significant portion of funds choose to earn returns in stablecoins rather than enter the market. These factors collectively constrain ETH's short-term performance. It is worth noting that market participants currently focus more on the ETH/BTC exchange rate rather than the absolute price of ETH. This exchange rate is regarded as an important indicator of changes in risk appetite. If the exchange rate stabilizes and rebounds, it indicates improved market risk appetite, and investors can increase their ETH holdingsSandisk (SNDK) Thoughts at $1820
Personal analysis, not investment advice.
Sandisk is rallying hard at $1820, driven by booming AI‑data‑center storage demand. Large long‑term enterprise SSD contracts have made the market re‑rate it as a core AI‑play, not just a consumer‑storage stock.
Key levels
‑ Near‑term support: $1640‑1660. A decisive close below this zone weakens the current uptrend.
‑ Major resistance: $1950‑2050. Heavy profit‑taking pressure sits at this previous high zone.
Bull case
AI KV‑cache creates real incremental demand for enterprise storage. Locked‑in long‑term contracts partially smooth cyclical risks, and recent earnings results are solid.
Material risks
1. Price has priced in far‑future management targets. The 80% gross‑margin goal is for 2028‑2030, not current results. Disappointing earnings could trigger heavy profit‑taking.
2. Storage remains cyclical. Expanding production from Samsung and SK Hynix may ease NAND price strength later.
3. Extremely high‑beta stock; sharp double‑digit pullbacks can happen quickly after big runs.
Practical suggestions
‑ If holding: Use $1640‑1660 as your defensive line. Trim partial profits when approaching $1950‑2050.
‑ If not in position: Avoid chasing highs. Wait for a pullback toward support before re‑evaluating.
Monitor two key signals going forward: cloud‑capex spending trends and NAND flash pricing strength. Respect the uptrend, but do not go heavy at elevated prices.
#Sandisk #SNDK #USStock #AIStorage最近$SNDK热度登顶全市场美股映射合约，大量交易员选择开空，但集体做空不等于行情马上反转，近期空头反而连续遭到逼空爆仓。 从OKX盘面数据来看，平台内SNDK空头持仓占比高达77%，绝大多数散户交易者站在空头一侧，大家看空的逻辑高度统一：短期涨幅巨大，RSI进入超买区间，估值已经透支部分利好，赌一波正股高位回调带动代币跳水。 但残酷的现实摆在眼前，过去24小时，全网SNDK空单爆仓规模突破1233万美元，占整体爆仓总额97%，不断出现空头被洗盘的行情，多空力量出现严重撕裂。不少巨鲸也出现分歧：部分巨鲸在高位布局空单博弈回调，同时还有另一批资金借着每一次小幅回踩进场接多，资金博弈异常激烈 。 一定要分清两件事：美股正股Sandisk本身的空头筹码，和加密代币市场的空头筹码完全不是一回事。代币价格锚定美股正股走势，只要美股存储行情继续发酵，场内海量空单就会变成上涨推力，形成空头挤压行情；只有当美股正股拐头下行，这批加密市场的空单，才有机会迎来真正的盈利窗口。 现在散户空头扎堆，仅仅代表市场情绪，不能直接作为见顶信号。博弈这个位置的空单，本质是在赌主线题材行情提前结束，风险$SNDK SanDisk, this time I admit it, but I refuse to accept it.
Holding short positions in my hands, I no longer want to think about costs. Now, the on-site contract has already touched above 1800.
There was no lead stock guide over the weekend, and contract prices were like kites with broken strings, supported entirely by sentiment. Order orders were scattered, and a small buy order could push the price up by a bit. I watched the candlestick close one by one, like counting my own loss progress bar.
My reason for shorting is actually quite simple—NAND capacity utilization is recovering, consumer electronics demand hasn't exploded, and no matter how compelling the story of AI storage is, it still needs to be filled by real shipments. With prices rising this much, expectations have already outpaced reality by a wide margin.
But I overlooked one thing: the pricing power of patient capital.
This round of long-term contracts is not just a simple commercial contract; it flattens and lengthens the revenue curve and also smooths out major cyclical fluctuations. Eight customers, $9.39 billion, up to five years—these figures mean SanDisk's performance fluctuations have shifted from "quarterly games" to "annual and even cross-year predictable."
The market's premium for this kind of "certainty" far exceeded my previous estimate.
What made me reflect even more was the inconspicuous statement from Investor Day: "We are no longer storage providers, but the cornerstone of data infrastructure." ”
It sounds like a slogan, but the details of the long-term contract implementation show they are indeed moving in this direction. Product customization, supply binding, and long-term customer engagement—when these three factors combine, the valuation system needs to change.
The worst part now is that the underlying stock hasn't opened yet, so all judgments can't be verified. Contract prices were pushed up over the weekend. Whether institutions are building positions in advance or short-term funds are using liquidity to amplify volatility will be clear as soon as Monday opens.
But regardless of the outcome, this single case has taught me enough:
When a company's "identity" changes, all judgment models based on historical cycles must be thrown into the trash.
I might not close the short position, but I definitely need to adjust the framework. #闪迪长期协议成焦点, opening performance remains to be verified Jensen Huang just published an article stating that NVIDIA is partnering with SB Energy to build a massive AI data center in Ohio, USA, with the first phase supplying power up to 4.25GW. OpenAI will lease this facility, using a full set of NVIDIA GPUs, CPUs, networking, and supporting software.
This time, NVIDIA is not just selling chips; they are personally handling the tough tasks of acquiring land, securing power supply, and building the data center. This shows that competition in the AI industry has extended downstream to infrastructure.
Many people think that just buying GPU graphics cards is enough. Actually, even if you get the GPUs, you need power, a completed data center, and customers willing to pay long-term for computing power. Only when this whole system runs smoothly can you make money.
Going forward, the scarcest resources in the AI industry have changed. It’s not just about competing for GPUs anymore; power, land, and data centers that can be completed and delivered on time are the real hard currency from now on.
#财报观察员：AI基建财报接力登场 周末梳理完这家存储企业的相关信息，如今已经不能简单拿传统周期股的逻辑去评判它。
八份长期供货合约落地，锁定了可观的保底营收，合约平均周期超过四年，直接锁定下一财年半数的出货量，再往后一年的供货占比更是达到三分之二。相当于把未来数年的基础收益提前确定下来。交流会上面又抛出了未来几年的盈利预期，给出很高的毛利与营业利润率目标，直接搅动市场情绪。上周单日大幅上行，次日继续冲高，短短五个交易日累计涨幅可观，开盘阶段成交额直接登顶市场前列，热度被彻底点燃。
越是行情火热，越要保持理性看待。短期连续拉升之后，积累了大量获利盘，此前也出现过内部高管减持的记录。盘前小幅上涨，价格在一千七百三十附近来回震荡，看着走势很强，但高位高开之后转头下杀套住进场资金的情况，市场早已屡见不鲜。
长期合约带来的逻辑确实很硬，相当于给波动极强的存储行业加装缓冲装置。可缓冲并不代表周期属性就此彻底消失。超高毛利目标能不能落地兑现、合作客户会不会出现履约隐患、现货产品价格后续走向，全部都是还没有答案的变数。概念故事讲得再动听，最终依旧需要财报数据来一一印证。
晚间开盘的表现会成为重要的检验节点。如果能够放量站稳一千七百位置，代表做多资金还愿意继续推高行情；一旦高开冲高之后快速回落，那么短期回调的压力就不容小觑。现阶段不会选择追高入场，安静观望盘面变化就好。$BTC $ETH $SNDK 现在盘面迟迟打不开上涨空间，核心症结其实很直白：增量资金并未进场，市场很难跑出可观的盈利机会。
大盘标的近期成交规模持续收缩，价格区间波动收窄，对应的产品资金流入力度偏弱，稳定币的存量还在不断向外流出。场外资金现阶段并没有迫切进场的意愿，多数都保持观望姿态。
反观另一个主流标的，近期更受资本青睐。七月份它对应的现货产品资金流入表现，甚至超过大盘标的。这也说明机构资金不是选择彻底退场，只是资金流向发生切换。
这点值得重点留意，当下大量流动性还滞留在海外权益市场，集中在AI、芯片存储以及贵金属这类赛道。这些板块近期行情火热，资金自然优先涌向能够快速产生收益的方向。
大盘标的当下就好比敞开大门等待来客，可资金都被别处的行情所吸引。
后续如果对应产品恢复持续性资金涌入，稳定币存量回流，同时市场成交同步放大，那么连续多月的窄幅震荡，就有可能在蓄力酝酿一轮级别较大的行情。
倘若资金依旧持续涌向海外科技相关板块，那么圈内市场很难迎来普涨行情。大概率会维持大盘横盘震荡，其余主流与热点标的轮番异动的格局。
归根结底，现阶段圈内整体赚钱效应已经比不上外部科技板块，资本天然会向着收益机会更好的领域流动。$BTC $ETH $SNDK 131万人持有代币化股票，一个月翻倍——钱在从币圈往代币化股票流，这事比BTC横盘有意思。
持有者从67万干到131万，月转账量231亿，涨了179%。NVDA代币化股票持有者16万，TSLA 9.6万，AAPL 7.4万。不是几十个人在玩，是百万人级别的资金在搬家。
但有个细节更值得琢磨——代币化股票总价值约24亿美金，月交易量却干到231亿，换手率9.4倍。传统美股的年换手率才100%出头，代币化股票一个月干了将近10倍。买来卖去，没几个人真在“持有”。
一边是持有者在翻倍，一边是SEC在踩刹车。SEC推迟了代币化证券的“创新豁免”计划，取消了加密规则会议。想在链上买美股的人越来越多，但监管框架还在路上。不过纽交所母公司ICE已经投了OKX，华尔街的通道已经打开了。
钱在流入，监管在拖延，换手率在飙升。链上买美股这扇门已经开了，但门框还在晃。 帮主有话说
SPCX这波确实是我应得的。
从110就开始陆续抄底，一路拿到150，这票中间砸到过105，能扛住没走，靠的是对基本面的判断。8月6日解禁那周，市场普遍预期暴跌，结果该跌不跌，从105拉到133，那时候就该意识到底部承接有多强。
#SPCX持股结构曝光，哈佛13F重仓
这波能到150，核心逻辑没变过。SpaceX不再只是火箭公司，星链用户翻倍到1200万，AI算力季度收入25.6亿，在手合同141亿。马斯克说五年后AI贡献99%价值，市场开始把SpaceX当AI基建标的来定价了。
哈佛持仓占13F组合51.8%，英伟达持股210亿，这些顶级机构锁仓比例高，流通盘一直偏紧。8月20日第二批解禁约7%，但真正的大头在12月初。底部已经确认了，剩下的解禁窗口只要市场能消化，底部就越来越扎实。$BTC $ETH $SNDK
150到了，你从110一路拿上来的仓位，这个位置可以考虑分批兑现一部分利润。底仓留着继续格局，等解禁消化完再决定怎么走。
大饼踏空就踏空了，闪迪1741空单止损1800继续拿着，SPCX这单底仓留着。今晚节奏好，利润拿得稳。
以上分析都具有时效性，单子必须要挂好止损，祝君好运。为什么我刚刚说$XSKHY 很强？因为它一周时间从137涨到177，涨幅30%。但这周韩国芯片在被抛，海力士明显是逆势上涨。
1. SK海力士是HBM存储龙头，跟美光直接竞品。刚刚我也说了，美光其实是不如海力士的
2. 日成交额352万，xStocks里中等偏下。韩国芯片股波动本来就大，xSKHY放大了这种波动。
3.存储板块内部走势已经分化了，不是全面上涨。而是部分强势股上涨，而且越强势涨的越高。
我的思路：现阶段还是观望为主。韩国芯片抛压还没结束，165-170追高风险大。等站稳160以上或者跌到145-150再看。【为什么闪迪这么猛？！】
1）财务曲线像垂直起飞
• 自由现金流单季70.7亿美金，全年114.8亿，负债率仅30%，账上现金47.6亿。
• 投资人日放话：未来三年资本开支后剩余现金100%返股东。
2）为什么突然这么暴利？三线共振
• AI吃存储：AI服务器单机NAND搭载量是传统的数倍，企业级SSD供不应求。
• NAND涨价飓风：TLC NAND 26Q2合约价环比+70~75%，DRAM同季+58~63%，十五年级别单季涨幅。
• 产品结构化跃迁：BiCS9 QLC、256TB企业SSD、CBA（CMOS直接键合）技术落地；推HBF高带宽闪存（补HBM和SSD之间的“内存墙”），SK海力士/三星入局标准联盟。
• NBM长期协议托底：已签协议最低保底营收938亿美元，锁到2030+，覆盖FY27预期bit产出约1/3。
$SNDK $BTC $SAMSUNG 这一轮行情走势相当出人意料，多空双方处境直接彻底反转，不少看空的头寸承受了不小的压力。
这波利好能够打动市场，核心并不单纯来源于存储芯片产品涨价，更关键的信号，是企业尝试挣脱传统周期股的标签，向着长期成长逻辑去转型。
在投资者交流会上，企业抛出了未来数年的经营指引，预期营收维持中高双位数的扩张，调整后的毛利率目标设定在80%，运营利润率看向75%，超额现金流将全部返还给到持股方。更值得留意的是，已经和八家合作方敲定最长五年的全新合作框架，整体合约规模接近93.9亿美元。
这背后释放的信号值得琢磨。
放在过去，市场最大的顾虑就是周期波动的桎梏。行业景气阶段靠着产品涨价可以赚得丰厚收益，一旦后续供给放量，产品价格回落，企业利润就会快速缩水打回原形。
而现在企业借助大批量长期合约，把下游客户需求、产能输出以及产品价格提前锁定。部分合约已经覆盖对应年度三分之二的存储产能需求，营收的可预判性得到明显提升。
所以当下支撑行情的核心逻辑，已经不只是看硬件涨价，而是AI浪潮催生的存储增量需求，叠加长期订单锁住利润空间的双重故事。
但即便利好消息密集落地，我也不会盲目进场追涨。
要知道八成毛利率、75%运营利润率的目标门槛很高，后续市场一定会拿实打实的财报数据，来检验这些目标能不能兑现。
因此正式开盘之后，我重点关注的关键点，就是高开之后价格能否稳稳守住盘面。
倘若开盘冲高的同时还能持续放出成交量，代表资金愿意接受这套全新估值体系；可一旦走出高开低走，伴随放量下挫，就意味着这份对未来的美好预期，已经被行情提前消耗完毕。
我的选择是先保持旁观，不急于高位进场博弈，静待盘面给出进一步信号。$SNDK正在重新定价！！！
今晚存储板块集体爆发，美股闪迪大幅拉升，市场正在重新给这条赛道定价，代币$SNDK也迎来估值重估窗口。
本轮上涨早已不是单纯短期消息刺激，核心来自企业端经营逻辑改变。闪迪通过长期供货协议锁定AI数据中心订单，主动管控产能扩张，把毛利率长期目标定在80%，试图拉长存储景气周期，市场不再单纯用传统周期股逻辑给它估值。
映射代币$SNDK价格锚定美股正股，但币圈自带杠杆交易属性，波动会进一步放大。最近几天它已经成为加密市场唯一持续性主线，大量资金从熄火妖币撤离，抱团涌入存储映射赛道，热度持续走高
但重定价不等于无脑看涨。一方面，存储是典型强周期行业，一旦后续厂商集体扩产，供需格局反转，估值会快速下杀；另一方面，代币走势完全绑定美股正股，如果美股高位拐头向下，代币会同步快速回调
当下盘面，短线资金博弈情绪拉满，多空分歧剧烈。想要走出持续性行情，必须美股正股继续站稳向上；一旦正股上涨乏力，赛道热度很容易快速退潮。博弈这条主线，一定要紧盯美股盘面节奏，不能脱离正股独立判断
以上仅代表盘面数据，不构成投资建议。$BTC $ETH 美股的热门板块解析：
HBM存储$SKHY 目前还有分歧，单NAND$SNDK 分歧小很多了，HBM分歧结束的涨跌幅可以参考这两天的SNDK
还有就是SKHY的市盈率快赶上工商银行，贵州茅台了，一旦市场认可HBM的超级周期，那补涨将会非常强烈，市盈率能到$LITE 的地步
这几天光模块也回调结束修复的差不多了，再有大动作得等NVDA的财报了
NVDA也量产CPO了，对LITE是利好，其他NPO LPO 光模块是利空，但现在对利空也不敏感了#闪迪长期协议成焦点，开盘表现待验证 The $SNDK expansion narrative is dead in the water.
Down >99% from ATHs, steady unlock pressure and liquidation cascades are capping every minor bounce.
Unlike $BICO,$BEAT, $ALLO,$KAITO, and $APR—which successfully absorbed liquidity to print recovery setups—$SNDK is just drifting lower with zero compression.
Catching this falling knife is high risk until real spot buyers build a clear support floor.
$SNDK #CryptoRevenueVsBTC 加密市场恐惧贪婪指数（8月17日22:00）
当前恐惧贪婪指数38，处于恐惧区间，尚未抵达20以下的极度恐惧位置，市场整体避险心态明确，但还没有出现集体恐慌割肉的投降行情。
指数由波动率、市场动量、社交媒体情绪、BTC市占率、谷歌搜索热度等多维度数据加权算出，本质是市场散户情绪温度计。
当下指数长期徘徊在30‑40，折射出很清晰的盘面现状：散户信心不足，不敢主动追高，资金纷纷从高位妖币出逃，回流BTC、ETH避险，BTC主导占比持续抬升。今天只有美股映射赛道走出持续性行情，其余热点大多只是游资短期脉冲，散户不敢长期接力，这正是恐惧情绪最直观的体现
历史规律里，极度恐慌往往容易打出阶段底部，极端贪婪常常预示行情见顶。现在只是普通恐惧，说明观望情绪浓厚，还没到情绪极致释放的时候。如果指数进一步回落至20以下，才会迎来大规模恐慌抛盘；反之，指数向上突破60进入贪婪区间，就要警惕市场过热后的回调风险
结合当下盘面，目前存量资金博弈为主，增量资金迟迟没有进场。操作上不要单纯依靠指数抄底，需要同步结合ETF资金流向、盘面爆仓数据一起判断
本文仅行情复盘，不构成任何投资建议。高盛硬刚鹰派噪音：美联储 9 月绝不可能加息，市场到底在恐慌什么？
高盛在最新的宏观研报中斩钉截铁地下了定论，美联储在即将到来的 9 月议息会议上绝对不会加息。高盛更直言不讳地指出，当前利率衍生品市场和风险资产对美联储紧缩路径的定价，明显「过度偏鹰」了。
为什么华尔街最精明的大行敢在各路官员放狠话的时候，如此笃定地站在鹰派预期的对立面？
答案藏在冷冰冰的底层宏观数据里。
过去几周，因为地缘原油反弹的脉冲和个别鹰派官员的强硬表态，市场开始自己吓自己，甚至一度把重新加息的离谱预期重新搬上了谈判桌。但翻开真实的账本，通胀指标中的核心 CPI 与 PPI 已经连续数月处于放缓通道，劳动力市场的招聘动能也在实质性降温。
在通胀退烧、名义利率维持在限制性高位的当下，美国的实际利率早已经紧得像一把铁钳。
在实体经济和中小企业融资已经明显感受到紧缩痛感的前提下，美联储根本没有任何宏观动因在 9 月去火上浇油。高盛认为，当前市场给出的高息恐慌溢价，纯粹是对美联储例行口头预期管理的过度反应。
这种「预期被过度压制」的状态，对加密市场来说反倒是一个潜在的估值弹簧。
回顾过去一个月比特币在 6 万多美元区间的窒息横盘与成交萎缩，很大程度上就是因为宏观利率预期的反复摇摆，导致华尔街配置型大资金不敢贸然在现货端大举建仓。
一旦 9 月议息会议靴子落地，确认按兵不动并重新校准未来的宽松节奏，此前在衍生品市场上堆积的过度做空头寸和避险筹码，势必会迎来一波凶猛的预期回补。
高盛的这份研报，戳破了笼罩在风险资产头顶的人造乌云。在流动性最沉闷的枯水期，不要被盘面制造的恐慌噪音震下牌桌，紧盯实际利率见顶转向的底层大势，往往比每天纠结官员的一两句狠话有用得多。
你认为美联储 9 月会如高盛所言按兵不动、甚至给年底降息铺路，还是会有出乎意料的鹰派黑天鹅？
---
以上内容仅代表个人观点，不构成任何投资建议。DYOR，NFA。
#标普盈利超预期，华尔街为何仍谨慎？ $BTC 四天前记录的收益是 219.64%，虽然周末两天的比特币基本处于休眠状态，不过和前一个四天的收益率比也差不多，那说明这也就是大概的平均水平了
不过前提还是希望近期不要出现黑天鹅就好了。
总感觉最近币圈是真的没什么活力也没什么压力资金是不是都跑去 AI 和存储了
#闪迪长期协议成焦点，开盘表现待验证 The important distinction is between cryptographic safety and user safety. SafePal says roughly 39,798 customers had names, contact details, shipping addresses and purchase histories exposed through an order-tracking plugin, while wallet systems, seed phrases, private keys and card data were unaffected.
That limits direct wallet compromise, but it does not make the incident low-risk. Real order and address details can make phishing about device issues, refunds or firmware updates unusually convincing, as user reports in May illustrate. The fix closes the flaw; sustained skepticism toward personalized support messages remains the stronger defense. Not advice, just analysis.
#SafePalOrderDataLeak#BTC成交萎缩, can ETF buying rebound?
"Shrinking volume" is not a positive sign; first look at whether the selling or buying orders disappear.
From August 10 to 14, US spot BTC ETFs saw a net outflow of about $385 million, while ETH ETFs saw only about $3 million in outflows during the same period; As of August 17, BTC's 24-hour trading volume was about 35% lower than the 30-day average, and IBIT's trading volume on the 14th was also about 31% lower. This is not the disappearance of selling pressure, but rather that both buyers and sellers are waiting.
Right now, it's more like a 'liquidity vacuum under capital rotation'—it can't be called accumulation complete. ETH is relatively strong, which means funds are picking odds, but it hasn't taken over the main upward bar yet.
My BTC core position is unmoved, new funds are only put into one-third of the planned position, not chasing ETH. The remaining two-thirds of ETFs have net inflows for three consecutive trading days, BTC volume stabilizes at $64,000; If the 30-day change in total stablecoin market cap also turns positive, add a second round. Only one day of inflow, at most a pulse.
If big money really comes back, I predict BTC will go first, ETH will accelerate later. The first round of funds usually starts with BTC, the most liquid BTC, and when $ETH/$BTC strengthens in sync, then consider switching incremental positions to ETH.
Shrinking volume isn't scary; no one wants to break the gap with real money, which is what deserves caution.苹果的故事，不能只看财报。
至少我自己是这么觉得。
因为它最终卖的是产品。
一台手机、一块手表、一副耳机，到底有没有让人觉得值得买，用户最清楚。
所以如果未来苹果靠AI重新起来，我不会因为市场突然喊得很热，就马上下结论。
我会去看真实使用。
大家是不是愿意换机？
是不是开始离不开新的功能？
是不是愿意为了AI多花钱？
这些东西比发布会上的一句话实在多了。
苹果以前最厉害的地方，就是它能把复杂的技术变成普通人一看就会用的东西。
如果AI也能做到这一点，我觉得它还有机会。
但如果只是功能很多、概念很大，用户实际没什么感觉，那我不会太兴奋。
因为科技公司最容易犯的错误，就是自己觉得产品特别厉害。
用户却无感。
最后还是市场说了算。
所以我现在看苹果，不会每天猜它下一季度卖多少。
我更想等它下一次真正让我产生“这个东西我也想要”的感觉。
如果那一天真的来了，我反而觉得市场可能会重新认识苹果。苹果现在的争议挺大。
有人觉得它创新不够。
有人觉得AI落后。
还有人觉得iPhone增长空间有限。
这些问题我都承认。
但我一直觉得，苹果还有一张很大的牌没真正打出来：
它手里有用户。
这是很多公司最羡慕的东西。
一个AI产品再厉害，没有入口也是麻烦。
苹果不一样。
它本来就在用户每天的生活里。
手机、电脑、耳机、手表，全是现成的入口。
所以如果哪一天苹果真的把AI做得特别自然，用户甚至感觉不到自己在“使用AI”，而只是觉得设备突然聪明了很多，那事情就有意思了。
但我不会提前给它画大饼。
因为发布会说什么都不重要。
真正重要的是：
大家用了以后会不会真的离不开。
如果用了两天就觉得一般，那故事结束。
如果用了一个月以后发现已经习惯，那苹果可能真的抓到东西了。
所以我现在对苹果的态度挺简单。
不急着看空。
也不急着吹。
等它把牌打出来。
因为真正值得投资的惊喜，永远来自结果，不是发布会上的想象。$BTC
现在市场最不缺的就是故事。
AI来了。
机器人来了。
新技术来了。
一家公司只要把这些词往公告里放一点，市场马上就有人开始研究。
微软不太一样。
它根本不需要每天讲故事。
Office本来就有人用。
云服务本来就有人买。
企业客户也一直在。
AI现在进来以后，更像是在原来的生意上继续加东西。
这点我很喜欢。
因为你不用赌“AI到底会不会成功”。
AI如果成功，微软多赚。
AI如果发展慢一点，微软原来的业务还是在那里。
这种公司，我会更愿意慢慢看。
当然，我也不是说微软现在什么价格都可以买。
恰恰相反，越是市场认可度高的公司，价格越容易提前透支。
所以有时候我看好一个公司，却不买，也很正常。
因为我买的是股票，不是公司荣誉证书。
公司再优秀，买贵了照样难受。
这句话其实特别简单。
但真正做到很难。
尤其是你看着它天天涨的时候，很容易觉得：
“再不上车就没机会了。”
其实不用急。
好公司，不会只给你一天时间。兄弟们，爆仓了，马上就被爆仓了。
闪迪又开始暴涨了，今天最高冲到1816，涨幅一度接近9%，股价创7月14日以来新高。
我的空单全仓浮亏125%、逐仓浮亏105%，强平价就在眼前，账户里的U快归零了。
还没冲上1万U，却被美股干归零了，盯着盘面脑子嗡嗡的，真想给自己两巴掌。
可是我看了一下K线，4小时级别已经干到89了，严重超买。
MACD高位顶背离，短期冲高动能明显衰减。连续五个交易日上涨累计涨幅约35%日线多头结构确实完好但短期追高风险极大。
这个位置放在山寨里，第二天就是一根大阴线伺候。
再看基本面——Q4营收89.7亿美元暴增372%、毛利率84.6%创历史新高，但下季度营收指引中值105.5亿美元，低于市场预期的111.6亿美元。
花旗把目标价从2500砍到2100，富国从1620砍到1400，杰富瑞从3000直接砍到1750。
业绩炸裂，财报后照样崩了10%。CNBC名嘴克莱默说闪迪7.7倍市盈率交易，不是因为市场认为低估，而是担心高利润难以持续。
存储芯片历史上就是暴涨暴跌的循环——缺货→涨价→扩产→过剩→崩盘。
这波涨得越猛，周期拐点来的时候摔得越狠。
闪迪还能继续拉吗？能，美股没有什么不可能。
但我反而非常警惕这波上涨，这种走法让我想起BICO和CAP。
只是我也补不了仓了，U都快没了，只能闭眼扛着等回调。
一个玩山寨的人，跑来空美股被套成这样，不亏谁亏？以后老老实实做我的山寨猎手。
$BTC
$ETH
$SNDK
#闪迪长期协议成焦点，开盘表现待验证 #BTC沉睡供应创新高，稀缺性再受关注
BTC沉睡供应创历史新高，表面看是稀缺性在强化，但换个角度想——如果越来越多人只是“拿着不动”，那活跃的筹码只会越来越少，价格对新增需求的敏感度会越来越高。
CryptoQuant分析师Darkfost 8月16日发文，超过10年未移动的BTC已达356万枚，占流通供应量17.7%，创历史新高。过去30天又有超过1.4万枚BTC加入长期休眠类别。
不止是“丢失”供应在涨。K33 Research数据显示，长期持有者（至少155天未移动）已控制79%的流通供应量，同样创历史新高。截至6月6日，沉睡至少两年的旧币中，仅218,421枚被重新激活，2012年以来最低。2024年同期这个数字是118万枚——对比一下就知道差距有多大。
CZ也在8月15日发帖：BTC已挖出超2007万枚，仅剩约4.4%供应量待挖，估计现有BTC中10-20%已丢失或无法找回。
不管是ETF回流还是机构配置——都会在更薄的订单簿上产生更大的价格冲击。稀缺性是长期叙事，但真正决定短期方向的，是边际供给和边际需求的博弈。这两股力量现在正在一个越来越窄的区间里角力。$SNDK 苹果差钱吗？
当然不差。
用户多不多？
也很多。
生态强不强？
也很强。
所以苹果现在真正缺的，我觉得是一点东西：
让大家重新兴奋起来的感觉。
因为一个公司如果一直很优秀，但每年都只是“比去年好一点”，时间久了，市场也会有点麻木。
苹果过去最厉害的时候，大家期待的不是多卖几百万台。
而是它又拿出了一个完全不一样的东西。
所以现在AI出来以后，我反而觉得这是苹果的一次机会。
不是因为AI有多神奇。
而是苹果终于有一个重新讲新故事的机会。
如果它能把AI真正做进自己的生态，让用户觉得：
“这个功能我每天都在用。”
那苹果可能重新焕发一点活力。
但如果最后只是多几个功能，多几个按钮，那我觉得没什么意思。
所以我现在对苹果的判断特别简单：
别听故事。
等产品。
东西好不好，用的人最清楚。
苹果有没有第二次高光，不是分析师说了算。
是用户用完以后会不会舍不得关掉。$SNDK Summary of why the market opened with a big rally
@玩的就是实盘 Quality trader at Jiuzong
1. Direct catalytic messaging
Wedbush reiterated his positive rating and set a target price of $2,000. Institutional bullish sentiment stimulated short-term bull funds to enter the market, creating a rapid rally.
2. Sector-driven collectiveness
Micron, SK Hynix, and Kioxia surged across the board, with AI storage super-cyclical sentiment heating up collectively, sector funds clustering together, driving Sandisk to rise in tandem.
3. Speculation in supply and demand fundamentals
The market speculates on the ongoing shortage of NAND flash memory and the rigid demand for AI server storage, combined with the company's long-term contract orders locking in profits, with funds rushing in early to push prices higher.我以前选股票，特别喜欢看主线。
AI是不是风口。
新能源是不是风口。
科技是不是热门。
后来慢慢发现，一个公司真正让我放心的，反而是它不完全依赖一个故事。
什么意思？
最好有自己的基本盘。
哪怕AI暂时没那么疯狂，它原来的业务也还能赚钱。
这种公司，我会更愿意长期看。
因为资本市场最怕的是：
故事没了，公司也就没了。
但如果底下本来就是一个不错的生意，上面再叠一个新的增长点，那就完全不一样。
微软就是比较典型的思路。
原来的业务就够大，现在AI再往里面塞。
这种逻辑我会觉得踏实一点。
反过来，有些公司只要风口一过去，市场立刻不知道该怎么给估值。
这种我就比较谨慎。
因为你买的可能不是公司，而是一段情绪。
所以现在看美股，我越来越喜欢问：
“如果这个热点明天消失，它还能不能赚钱？”
如果答案是能，我会多研究。
如果答案是不知道，那我一般不会太激动。
市场热点天天换。
真正稳定的生意，没那么容易换。《我现在看美股和比特币，有个最大的变化：不会那么容易被消息带着跑了》
以前看到一个重大消息，恨不得马上改观点。
利好？
看多。
利空？
看空。
后来发现市场根本没这么简单。
很多时候，消息出来之前，价格已经提前动了。
消息真正落地的时候，反而开始兑现。
有时候明明是大利好，股票就是不涨。
为什么？
因为市场早就知道了。
所以现在看到消息，我第一反应不是“这对市场是好还是坏”。
而是：
市场之前有没有预期到？
如果所有人都已经知道，那这个消息的价值可能没那么大。
反过来，一个很普通的消息，如果完全超出市场预期，反而可能让价格动得很厉害。
这就是我现在越来越喜欢看的“预期差”。
美股如此。
比特币也一样。
所以以后看到什么重大新闻，我可能还是会看。
但不会立刻冲进去。
先看看价格怎么反应。
如果消息很好，结果价格没动，我会特别关注。
如果消息不怎么样，价格却特别强，那也值得关注。
因为市场有时候会告诉你：
“你看到的东西，不一定是我真正交易的东西。”
这句话我现在越来越信。
真正懂市场，可能不是消息看得最多。
而是知道消息出来以后，应该看价格。#财报观察员: AI infrastructure earnings report debuts one after another
"After the Bloodbath, NVIDIA Watches on August 26"
Three weeks ago on that Tuesday, SK Hynix fell 13% in a single day, Nikkei dropped 4%, and Kioxia dropped 18%, with the semiconductor sector flipped by the same leverage. The trigger is the same question: when will AI's money turn into profit?
The market has shifted from listening to stories to flipping through reports. SanDisk handed over its first results on August 13, with eight long-term contracts locking in $93.9 billion in guaranteed income, a gross margin target of 80%, and the stock price rose 13% that day, rebounding 55% from its low. Google raised its capital expenditure to 205 billion, but instead fell by 5.4%, marking the first time its single-quarter cash flow turned negative. Same bonus, two scripts.
On August 26, NVIDIA closed the market, and I marked the date on my calendar. Last quarter's revenue was 81.6 billion, up 85% year-on-year, with the market waiting for it to raise its guidance above 100 billion. Whether the gross margin can hold at 70% is directly questioned whether this round of burning cash yielded returns. Nvidia plans to guarantee $250 billion in funding for OpenAI. Cramer said this reminds him of the telecom bubble in the early 21st century—once the guarantee chain broke, pricing power loosened.
Retail investors don't need to gamble on financial reports; they should prepare three rulers and other matching tables. Nvidia's gross margin and guidance, after long-term contracts are locked in prices, will NAND spot prices still rise? Are there new holes in the guarantee chain? Only when the three readings match can the market have a solid foundation.
Earnings reports are like a magic mirror; before the reveal on August 26, every minute of rebound is just a rough draft for the market. $BTC 《微软这几年真正的变化，我觉得不是AI，而是它越来越像基础设施了》
微软现在给我的感觉，已经不太像单纯的软件公司了。
以前我们想到微软，就是Windows和Office。
后来是云。
现在又多了AI。
东西越来越多，但它们之间其实是连起来的。
企业本来就在用微软的产品。
现在AI来了，微软可以直接往原来的系统里面加。
这个优势挺大。
因为它不用从零教育客户。
客户本来就在那里。
你只需要让客户觉得：
“这个东西确实有用，那我就多付一点钱。”
这就是微软比较舒服的地方。
当然，市场已经看到了这个优势。
所以股票也不便宜。
这又回到了最开始那个问题：
好公司和好价格，永远不是一个东西。
微软如果未来继续涨，我不会意外。
但如果涨得特别快，我也不会因为喜欢它就追。
我宁愿等。
因为真正长期看好一家公司，并不意味着任何价格都值得买。
你可以喜欢它。
也可以耐心。
市场不会因为你少买一次微软，就不给你别的机会。
我更希望自己买进去以后，可以踏踏实实拿。
而不是因为成本太高，稍微一跌就开始怀疑人生。《英伟达现在最难的事情，可能是让所有人继续满意》
英伟达现在已经强到一个什么程度？
不是市场觉得它好。
而是市场已经习惯它好。
这两件事情差别很大。
以前它业绩出来，大家觉得超预期，股价就很容易有反应。
现在可能业绩还是特别漂亮，但大家已经提前猜到了。
那市场就会开始要求更多。
“还能不能更好？”
“明年还能不能继续这么快？”
“AI需求会不会继续爆？”
这就是大公司走到高预期以后最麻烦的地方。
你不是要证明自己优秀。
你是要不断证明自己比市场想象中还优秀。
这件事不可能永远做到。
所以如果未来英伟达出现几次业绩很好、股价却没怎么涨的情况，我反而会认真看。
因为那可能说明市场开始从“惊喜”进入“消化”。
不一定是坏事。
甚至可能是很正常的过程。
我还是看AI。
也还是看英伟达。
但我现在不会把“好公司”直接等同于“好股票”。
公司能不能赚钱是一回事。
市场现在愿意给它多高的价格，是另外一回事。
投资做久了，最后绕不过这两个问题。$BTC sideways for a month, $ETH has stopped moving: the quieter Crypto is, the less likely the options market is to sleep
The biggest feature of recent spot markets isn't price fluctuations, but lack of volatility. According to Deribit's weekly report, short-term implied volatility for BTC and ETH both hit their 2026 lows in August, with BTC stuck between $63,000 and $65,000, ETH hovering around $1,900, making people sleepy.
But the options market hasn't been stunned. The 25 Delta Risk Reverse for BTC and ETH is still a somewhat out-of-the-money put option, with quiet prices and funds willing to spend money to guard against sudden drops. BTC data is the most typical: the 30-day actual volatility has dropped to 21.8%, the lowest since October 2025, but the 30-day implied volatility remains at 36.35%, meaning the market's price for future volatility is about two-thirds higher than recent actual volatility. One-week options are even more exaggerated, with actual volatility around 16% and implied volatility about 29%, a gap close to the largest level in a year.
ETH is similar, entering a low-volatility zone for the year, with a bias still leaning toward bearish protection. What's more frustrating is that ETH spot has actually been more resilient than BTC over the past 30 days, but derivatives players still haven't fully believed in this round of strength.
So the real danger isn't the daily crash, but everyone getting used to "not moving." Low volatility forces people to sell volatility, leverage their assets, and improve capital utilization. Market trends often emerge when this structure is suddenly brokenDiscount rate stratification: Different assets have vastly different sensitivities to interest rates
I've talked about discount rates before, but from a more nuanced perspective: different assets are completely different in sensitivity to changes in discount rates.
We use the concept of "duration" to measure it. The longer the duration, the more sensitive it is to interest rates.
BTC's "duration" is extremely long—its cash flow (if any) is in the distant future, or even none, making it an "ultimate narrative" asset. If interest rates rise even slightly, its valuation will be sharply lowered. That's why BTC fell even worse than tech stocks during the 2022 rate hike cycle.
Stablecoins, on the other hand, have very short "durations"—they peg to the US dollar and do not generate interest (unless you deposit into the protocol), so changes in interest rates have almost no direct impact on them. Even when interest rates rise, holding stablecoins in interest-bearing protocols can earn higher risk-free returns.
This explains why, during rate hike cycles, funds flow from BTC to stablecoin yield-bearing protocols—not because they don't love BTC, but because the "duration matching" of funds is changing: when rates rise, people prefer to hold short-duration assets until rates fall again.
Understanding the concept of "duration" gives you why BTC crashes when raising interest rates, while stablecoins actually benefit. The market is not "fleeing to safety," but "shortening duration."SPCX（SpaceX）不按常理出牌的核心原因在于其独特的“供需结构”与“市场博弈”逻辑。 这并非传统的价值驱动型股票，而是处于极低流通盘（约11.4%） 与高频期权博弈的特殊阶段。其走势往往由“空头回补”和“期权Gamma挤压”主导，而非单纯的基本面消息。
为什么常规指标失效？
常规技术分析基于“资金理性”，但SPCX当前的核心变量是博弈：
- 流通盘稀缺导致的“逼空”：IPO流通盘极小，解禁被视为“利空出尽”而非“利空”。例如8月6日20%解禁时，市场预期抛压导致空头大量建仓，结果因实际抛压有限，触发了教科书级别的逼空行情（股价大涨）。
- 期权市场的“杠杆效应”：期权市场的剧烈波动（隐含波动率高达71%）会引发做市商的被动对冲。当价格快速波动时，做市商为了对冲风险被迫买入或卖出股票，这种行为会放大价格的跳跃性，导致单日巨震（如单日跌13.61%或涨15%），常规指标无法捕捉这种非线性变化。
8月20日（本周三）7%解禁影响预判
针对你提到的“这周20号7%解锁股”，虽然规模小于上次的20%，但市场逻辑已变，重复上次“跌后反弹”的难度加大：
- 规模对比：上次（8月6日）解禁约9.12亿股（流通盘增1.43倍），本次约3.2亿股（约3.5倍日均成交量），理论冲击更小。
- 市场预期差已消失：上次大涨是因为“所有人都看跌，结果没跌”（预期差）。经过上次的教训，市场对本次解禁的恐慌情绪减弱，空头仓位可能远不如上次拥挤。缺乏大量空头作为“对手盘”，再次上演大规模逼空的基础已不存在。
未来走势推演：三种剧本
根据当前市场结构，本周及解禁后的走势主要取决于内部人抛售意愿与资金承接力的博弈：
1.乐观剧本（复刻逼空）：
- 条件：内部人惜售，且市场仍有大量空头未平仓。
- 走势：类似8月初，解禁日不跌反涨，触发空头踩踏回补，目标价看向160美元以上。但该剧本概率已降低。
2.中性剧本（震荡消化）：
- 条件：出现温和抛售，但承接资金（如看好AI转型的机构）入场吸纳。
- 走势：解禁后小幅回调或横盘震荡，波动率逐渐收敛，等待下一个催化剂（如AI业务进展）。这是目前概率最高的情况。
3.悲观剧本（情绪退潮）：
- 条件：内部人集中抛售，且市场对“AI故事”审美疲劳。
- 走势：抛压超出市场承接能力，股价下探前期低点，可能测试100-108美元区间。
交易与观察建议
- 盯紧期权链：关注关键价位（如125、140美元）的期权持仓变化，这往往比K线图更能提前反映主力意图。
- 降低预期：不要赌“一定反弹”。鉴于上次的逼空已经释放了部分情绪，本次更可能是震荡市，高抛低吸比死扛更适合当前环境。
- 关注资金面：如果解禁当日成交量未显著放大且股价稳住，说明抛压不大，可逢低关注；若放量下跌，则需警惕情绪退潮。#闪迪长期协议成焦点, the opening performance remains to be seen
SanDisk SanDisk, absolutely unbeatable! It's just too much. The roles of bulls and bears have completely switched, and they've become the strict father of the Air Force
With this wave of positive news for SanDisk$SNDK, I think what the market really buys isn't that NAND is about to rise in price again, but that it's finally trying to shift its focus from 'storage cycle stocks' to 'long-term growth stocks.'
Investor Day directly sets targets for FY2028-FY2030, with revenue maintaining mid-to-high double-digit growth, adjusted gross margin of about 80%, operating margin about 75%, and 100% excess cash returned to shareholders. More importantly, eight clients have signed new business model agreements lasting up to five years, with total contract value of about $9.39 billion.
What does this mean?
Back when SanDisk was price-raised, people feared that this year's NAND price hikes would make a fortune, but when supply picks up next year, prices drop and profits will drop back to square one.
Now, it is using long-term contracts to lock in customer demand, capacity, and prices in advance. Some agreements already cover about two-thirds of FY2028's bit demand, significantly increasing revenue certainty.《英伟达真正让我担心的，不是跌，而是以后涨不动》
如果有一天英伟达开始跌，我其实不会特别意外。
涨了这么多，调整本来就正常。
我反而比较怕另一种情况：
消息一直很好，业绩也不错，但股价就是不怎么涨。
这才是我真正想观察的。
因为这说明市场开始消化掉它的预期了。
以前业绩超过预期，股价马上拉。
以后如果业绩还是很好，但市场只是“哦，知道了”，那就说明资金的胃口变了。
一家公司到了这种阶段，其实很考验后面的增长。
不是说它不好。
只是大家已经太熟悉它好了。
你拿一个已经家喻户晓的故事，再讲一遍，效果肯定没有第一次那么大。
所以我现在看英伟达，反而更关注它每次上涨以后市场的反应。
利好出来涨多少。
涨完能不能保持。
出现利空以后跌多深。
这些东西慢慢积累起来，就能看出市场到底还认不认可。
我还是看好AI产业。
但AI很好，不代表任何一家AI公司都能一直涨。
市场最后还是会挑。
而且越到后面，挑得越狠。
所以现在对英伟达，我的态度就是：
尊重趋势。
但别迷信涨幅。《美股如果又创新高，我反而想看看市场是不是开始累了》
指数创新高当然是好事。
但我现在看到新高，已经不会像以前一样特别兴奋了。
因为涨到高位以后，最重要的不是还能不能创新高，而是上涨有没有越来越费劲。
比如以前涨一点，半导体一大片跟。
现在如果变成只有几只大票在撑，其他股票完全没反应，我就会觉得市场有点累。
这不代表马上要跌。
只是说明资金开始挑了。
而且现在大家都知道AI是主线，知道科技股是核心。
这种时候再往上，就需要更多真金白银进来。
不是靠几个新闻标题就能一直推。
所以我现在看纳指，特别喜欢看一个细节：
每次回调以后，资金回来得快不快。
如果跌一点马上有人买，我觉得没什么。
如果每次反弹都比上一次弱，那就需要留神。
其实市场很多时候不会突然告诉你“我要见顶了”。
它只是慢慢变得不那么强。
以前很容易涨。
后来需要一个大消息才涨。
再后来，好消息出来也涨不动。
这些东西都是过程。
所以创新高没什么可怕。
真正要小心的，是上涨开始变得越来越勉强。在今天 GPS 上涨了 50%+ 后，OKX Ventures 在 1 小时前把 4861.1 万枚 GPS ($75 万) 给转进了 Binance。
OKX Ventures 是 GPS 的投资机构，这些代币来源于投资解锁归属。
至于为啥 OKX 的币要到 Binance 去卖，因为 OKX 没上这个代币的现货。
地址：0x78DB576873F032A7C4749b214A8AF6966B3Be239 Sai lầm lớn nhất của nhà giao dịch không phải là mua khi thị trường lao dốc, mà là mua theo cơn sốt khi mọi người đã vào lệnh từ trước. 📉 Chỉ cần vài cây nến xanh hiện lên, Crypto Twitter lại lập tức gọi tên mùa altseason. Nhưng thị trường đang kể một câu chuyện hoàn toàn khác. 🧐 Đây không phải một đợt tăng trưởng rộng khắp. Đây là sự chọn lọc thanh khoản. 💧 Tiền không còn chảy vào mọi thứ như trước. Nó chỉ tập trung vào những dự án có nhu cầu thực sự, khối lượng giao dịch mạnh, sự tham gia cHYPE has been quiet on the timeline, but are all the big players increasing their positions?
Kevin Warsh, former boss of Federal Reserve Chair, disclosed a new position: a Nasdaq-listed company ticker PURR whose main business is hoarding HYPE.
This former boss is Druckenmiller. He has been doing macro trading for 45 years, with no year losing money. Warsh was a partner in his family office for 15 years, earning $10.2 million annually, with over $100 million personally deposited.
He only fully stepped down before Warsh was sworn in as Federal Reserve Chairman in May this year. Senator Warren directly asked at the hearing, "Who gave you this $100 million?"
This is the person who manages a $5.2 billion portfolio. Last quarter, he did two things at once: he sold out all three chip giants—Broadcom, Intel, and Micron—without keeping a single share.
It has held $264 million in BlackRock BTC spot ETFs, HYPE Treasury company PURR with $23 million, over 4 million shares of mining company Bitdeer, and 750,000 shares of mining company Riot Platforms.
Chip cuts completely, crypto is fully deployed. From spot ETFs to DeFi chains to mining companies, one line.
There have been noticeably fewer people talking about $HYPE on the timeline recently. It surged from $25 at the start of the year to $77 in June, and now back to $57.
Hyperliquid's protocol revenue fell from 357 million yuan in Q3 last year to 202 million yuan in Q2 this year, a 43% decline over four quarters. Even its core buyback mechanism shrank from 316 million yuan in a single quarter to 149 million yuan.
Then in May, Bitwise, 21Shares, and Grayscale listed HYPE spot ETFs. In June, PURR was included in the Russell 2000 index. Now it appears in Druckenmiller's 13F.
The timeline is quieting, but are the big players all increasing their positions?