Orbit Post Sitemap
ASML已经把客户的采购兴趣讲到2030年了。
三季度收入117亿欧元、毛利率57.5%，是10月14日已经公布的结果。我更在意公告的后半段：有客户表达了对2029年、甚至2030年下单的兴趣。AI投资的时间表，已经在设备供应商这里延伸得很远。
但公告用了“有兴趣”这个说法。客户预测、正式订单和最后确认收入之间，还有距离。ASML也判断，按现有计划，2027和2028年的供给不能完全满足客户需求，正在研究增加后续产能。这个判断支持需求强劲，同时也提出一个问题：现在听到的远期需求，到时候有多少能按时交付？
我不想把管理层的积极表态打折成没价值，它当然比空泛的行业口号有用。可供应商要提前投入，客户自己的项目也可能改变节奏。越远的采购意向，越需要后来不断用正式订单验证。
这份财报让我对设备需求继续保持兴趣，也让我更想区分市场已经确认的收入和仍在谈的计划。若把2030年的采购兴趣直接放进当下盈利估值，未来几年的顺利执行就被提前假定了。好消息值得高兴，具体拿什么价格去相信它，还得自己算。
#阿斯麦财报前瞻：EUV订单成关键看点 Last night, the CPI disappointment landed, and the group chat was full of voices saying "Gold is about to surge": Inflation is high, shouldn't gold go up?
Then $XAUT gold immediately crashed. A 1.65% drop, quickly hitting 4096, with a low of 4088. It only recovered a bit today,
Meanwhile, $BTC was relatively stable last night, fluctuating between 81000-83000, with a much gentler drop than gold. How come?
Breaking it down, it's one thing: Inflation exceeded expectations, the market assumes the Fed will keep high interest rates longer → US Treasury yields are pushed up → Gold holding costs rise → Funds withdraw.
Gold used to be about "inflation hedge," now it's about "fear of rate hikes." As long as the Fed remains hawkish, gold has little short-term prospects. Nothing surprising here. Who calls the shots in the market now?
The Fed. Gold, BTC, and US stocks are all trading the same thing: liquidity. When the Fed sneezes, none can escape. #美国9月CPI同比升至3.7%，核心通胀超预期 奈飞从高速拉新阶段，进入成熟盈利阶段。新增会员见顶，靠涨价+广告业务维持收入增长；利润和现金流很强，但营收增速下滑，市场担心增长空间不足。All three coins fell, each dropping for different reasons: BTC, ETH, and SOL pressure source breakdown
Falling on the same day doesn’t mean the same thing
BTC, ETH, and SOL candlesticks look similar, but the capital flows tell different stories
$BTC looks more like an emotional retreat; the market cools down, selling triggers linked orders, but no large holders are collectively exiting on-chain
Support still exists near the 400-day moving average, signs of accumulation at low levels remain
Its weakness is emotional release, not chip collapse
$ETH, on the other hand, is capital leaving; ETFs have continuous net outflows, wallet balances are migrating outward, some holders actively reducing positions
It’s not just dragged down by the market but a contraction in capital
Therefore, recovery depends more on new capital returning, which is more difficult
$SOL’s pressure is in derivatives, with perpetual contract shorts increasing
Leverage pushes the price down, but spot outflows are limited, chips haven’t changed hands on a large scale
The decline is fierce, but the source is more on the contract side.
So, the three bearish candles each have meaning:
BTC is emotion, ETH is capital, SOL is contracts
Only looking at the order of decline can easily mistake emotional fluctuations for trends
Breaking down with on-chain data clarifies who is truly selling and who is just dragged down
Knowing where the pressure comes from tells you what to be cautious about and what to ignore #美国CPI超预期升至3.7%，今晚PPI再迎考验 #BTC现货ETF再度净流出，机构买盘能否回暖？ #美俄柴油直供推进，美国考虑放宽油轮制裁 #TSMCQ3RecordProfit TSMC is going to release its latest earnings report this Thursday✨
Actually, a 50% year-over-year revenue growth is nothing new; the market already got preliminary data last week making this clear. What really holds the market's pulse are three key numbers hidden in the earnings report: gross margin, full-year performance , and next year's capital expenditure.
TSMC's preset range for Q3 gross margin is 65%-67%, but last quarter's gross margin even touched a high of 67.7%. #TSMCQ3RecordProfit TSMC's revenue has already exceeded expectations, so what is the Thursday earnings call really about?
TSMC's Q3 revenue was actually announced early, about $46.7 billion, higher than the company's own forecast of $44.6 to $45.8 billion. The Thursday earnings call isn't about revenue, but about the costs behind that revenue. 阿简不建议把今天市场的普跌简单理解为所有项目一起变坏了，有些资产只是被利率重新定价、有些资产是高估值被砍、有些只是funding太贵，多头被迫平仓，这几种情况的价格表现可能一样，处理方式却完全不同：
如果你持有的是$BTC 、$ETH 、$SOL ，先看宏观和资金；
如果你持有的是RWA、AI、DeFi，先看收入和估值；
如果你持有的是高倍杠杆仓，先看清算。
别问，问就是先把下跌原因拆开，再决定要不要恐慌Run a round of xp simulation 【2026 Midterm Elections · Control of Both Chambers】
First, my judgment: I bet on "Democrats take the House" (I estimate 90%), not optimistic about "Republicans sweep" (I estimate 9%). The basis is not the market odds, but these three points:
① General vote polls D+7 ~ D+9.6, for a president with about 30-32% approval — this combination is a very bad environment for midterm elections
② Cook Political Report on 10/9 shifted 18 House districts toward Democrats at once and estimated a net gain of 9-19 seats for Democrats
③ The real variable is the Senate: model range 54%-74% is very wide, ME/TX/OH/AK/IA all within single digits, and red state Senate polls historically overestimate Democrats
My allocation logic: House nearly locked → heavy position; Senate only small position; "sweep" which requires both chambers → light position.
By the way, an observation: the sum of the YES sell prices for the 5 outcomes of this event = 1.05 (should be 1.00), indicating the whole basket is overvalued by 5% — buying YES on any side is buying at a premium.
(xp simulation record, not investment advice)
#OKXProphet #EarningsSeason $SNDK 1200这位置，我盯死了！！！
早间韩股直接带崩主流和美股，美股这边存储板块也跟着破位。今天存储芯片走势已经很难看，到晚上大概率还会二次下探，而且这次是利空级别的。再加上12月加息预期又抬头，短期想快速修复反弹，基本没戏
从盘面上看，闪迪之前1500附近就是顶部，高点一个比一个低，多头已经被打服了，根本抬不起头。我个人判断美股还会继续往下砸，尤其是闪迪，1200这位置大概率会到
因此我的操作思路：求稳的话，等反弹到1365-1380区间去空，目标先看1275，终极目标1200
#美国CPI超预期升至3.7%，今晚PPI再迎考验 $BTC Non-farm night, don't mistake a sharp BTC rally for a turning point
Tonight ADP warms up first, then non-farm sets the tone; the interest rate market won't trade just one number. The market expects employment to continue cooling, but if wages rise month-over-month, risk assets might shift from a rebound to a trend.
Rather than guessing the data, I'm more concerned about a preliminary question: if non-farm is soft, who is willing to buy on the relay? As of October 12, the CME Bitcoin futures basis hasn't significantly widened, and the seven-day net increase in stablecoins is limited. Before spot funds return, any sharp rally looks more like short covering than the start of a trend.
BTC current price 83,356, ETH 2,505, SOL 107.41. With these three numbers on the table, tonight I’m focusing on three things:
1. Whether the unemployment rate and hourly wages cool down simultaneously. If only unemployment rises but wages don’t fall, the dollar and US Treasury yields may not ease, and the rebound could easily become a pulse.
2. Whether ETH/BTC and SOL/BTC can recover synchronously after the data. BTC running alone only indicates risk aversion or short covering; ETH and SOL following suit represent risk appetite spreading.
3. After the first surge, whether spot trading volume and perpetual funding rates cooperate moderately. A sharp rise in funding rates and a surge in open interest indicate crowded longs; spot volume expansion with stable basis indicates healthy turnover.
So tonight I won’t take sides early, just verify. Non-farm decides the first jump; exchange rates and capital structure decide if there’s a second leg.
$BTC $ETH $ZEC #TSMC Q3 net profit exceeds expectations, profits hit a record high The overall market is still paying off CPI debts, but DeFi old money CRV is quietly approaching its previous high. Is this a rotation of funds or just another pump and dump?
Bitcoin is hovering above 81,000, OKB has dropped 4%, yet the popular rankings see the veteran DeFi project—Curve's CRV—rising 3.4% today to 0.342. In the last 24 hours, it climbed steadily from 0.31 and is now right below the previous high of 0.344 and the upper Bollinger band at 0.346.
Honestly, there’s no new positive news today. It looks more like funds are moving off the AI and Meme short squeeze tables, rotating into deeply discounted DeFi blue chips, catching the on-chain warming as Curve’s active loans recently rebounded to $76 million, with perpetual contracts also seeing inflows. But in early October, it surged past 0.37 only to be slammed back by protocol losses and large holders. Now the RSI is at 63, standing near a resistance level.
The market isn’t unanimously bullish: OKEx spot 24-hour volume is only around 2.45 million USDT, CRV is priced across multiple platforms with OKEx just one venue, and volume hasn’t clearly expanded. The 15-minute candle still closed bearish at a high of 0.343, pulling back.
Watch for one signal: can it break above 0.346 with volume and reach the October trapped zone at 0.37? If it can’t hold and volume shrinks with a wick at the upper band, that’s another range-bound pump and dump—don’t chase at the wall. DYOR, not investment advice.
#CRV #Curve #DeFiThe more volatile the market, the more it doesn't mean ordinary investors make more money. Sometimes, the ones who truly make stable profits are those providing trading services.
Morgan Stanley's latest quarterly results show that its equity trading revenue increased by 41% year-over-year, reaching $5.8 billion. The company's overall performance exceeded analysts' expectations.
Behind this figure lies a noteworthy distinction: financial institutions can earn income through trading services, market making, and other businesses; ordinary investors, however, must predict market direction, bear price fluctuations, and pay trading costs.
Of course, rising trading revenue does not mean all income comes from retail investor fees, nor does it imply that clients as a whole are losing money.
But it at least reminds us: financial markets are not just about "who bought right and who sold wrong." Different participants make money through different mechanisms.
Some profit by correctly predicting direction, some by providing liquidity, and some by charging asset management fees. If you only focus on price movements, you might overlook who exactly you are competing against.
I've always believed that before ordinary people enter the market, understanding their source of profit is more important than first finding a magical indicator.
Do you think retail investors find it hardest to beat market direction, or trading costs, information asymmetry, and emotions?BofA's $1.2B in fines did more than lower reported EPS to $0.91: it shows how compliance can reshape earnings even as revenue grows. That matters against Morgan Stanley's IPO-linked asset influx and Goldman's trading gains, where revenue support may prove more cyclical.
The open question is whether high rates and credit quality add another burden.
#BofAQ3HitByFines #奈飞增长或放缓，广告业务成新看点 Fifteen percentage points, that's not volatility, that's the opponent on the other side of the board flipping the timer. CoreWeave's stock price dropped 15.46% in one day, and the 2032 bond yield briefly exceeded 15%—to a grandmaster, this is like a player who originally set up a Sicilian Defense suddenly realizing their central pawn chain is completely broken, and their king's flank is exposed under crossfire from the opponent's two rooks and queen.
I calculate twenty moves ahead before making a move. The market is doing the same thing now: it's calculating CoreWeave to 2032. Over sixty billion dollars in debt and lease liabilities, over ten billion in order backlog. Management says the reaction is irrational, long-term orders are enough to cover the debt. This sounds familiar in chess logic—many players say "my position has potential compensation" when at a disadvantage, but whether that compensation can be realized depends on whether you can truly mobilize your pieces to the critical squares. Orders are the pieces; delivery and computing power deployment are the moments they step out of their squares and occupy the center. If the pawns can't advance and the rooks, knights, and bishops are all crowded on the back rank, that's stagnant assets.
The AI debt situation is sliding from the opening's mindless advance into the midgame's piece exchange phase. Hot inflation is the uninvited variable that disrupted the originally smooth rhythm. When funding costs rise, all positions relying on discounted future cash flows must be reassessed. And the $xPLTR linkage shows the market is already doing cross-board portfolio calculations—not just looking at a single pawn, but counting the pressure along the entire major diagonal.
My judgment is straightforward: the bulls are now thinking while in check. They either find a brilliant sacrifice to seize the initiative and accelerate cash flow realization; or they are gradually squeezed into a position with no moves left. Historically, no game has been won by "I feel the position is good." The real decider is who can truly convert the order piece into cash and have it land in the account.
Will funding pressure spread to the same sector? To read the board, you must look at the connections between pieces. If this corner is breached, the adjacent weak pawns will inevitably collapse as well. #coreweavestockbonddropNetflix ($NFLX) is really heating up, so is the growth slowing down, and advertising becoming a key variable?!
Current price: about $70.57 (YTD down about 26%, about 44% retracement from the high)
Forward PE about 20x, analyst consensus target price about $90–96 (potential upside about 30%)
Core contradictions
- Growth slowdown: 2026 revenue guidance 13–14% (to $51–51.4 billion), Q3 guidance about 12%
- Profit resilience: operating margin target 31.5% (29.5% in 2025), strong FCF
- Advertising highlight: over $1.5 billion in 2025, 2026 target about $3 billion; monetization pace slower than expected
Key levels
- Support: $70 → $65–68
- Resistance: $71.5–72 → $75 → $80
Observation point: Q3 earnings report on October 20, focus on advertising revenue realization and subsequent guidance.
Current price already reflects growth slowdown, progress in advertising business will determine the pace of valuation recovery. For reference only, not investment advice.
#奈飞增长或放缓，广告业务成新看点 October "Uptober" turns into "Rektober"
BTC opened at 84,853 at the beginning of the month, reached a high of 86,972 on the 5th, then continuously declined. After the CPI release on the 14th, it plunged again to 80,125, rebounded to around 81,600 on the 15th, down 2.49% in 24 hours, down 2% for the week, but still up about 4.4% for the month, retracing about 32% from the historical high of 126,200 in October 2025. ETH was weaker, falling below 2,400, down 18.5% year-to-date. The total crypto market cap fell from above 3 trillion to 2.75–2.85 trillion, evaporating about 110 billion USD in 36 hours.
Three macro factors hit simultaneously: The 10-year US Treasury yield rose to a 24-year high of 5.36%, the 30-year broke through 5.73%; September CPI year-on-year was 3.7%, exceeding expectations, and the October rate hike bet surged from 21.6% to 43.7%; oil prices rose above 93 USD combined with Middle East geopolitical tensions, risk-off sentiment pressured risk assets. Insufficient new buying, BTC’s correlation with tech stocks deepened.
Capital outflow is obvious: BTC spot ETF net outflow was about 679 million USD from October 5–9, ETH ETF outflow was 542 million USD over five days; monthly liquidations exceeded 1 billion USD, with long positions accounting for over 90%, the 80,000 level was defended twice but leverage was not truly cleared.
Key outlook: Short term lacks conditions for a one-sided move, likely to range between 80,000–87,000, awaiting PPI, retail sales, and the Fed’s November meeting for direction
$BTC 地基正在开裂，而甲方还在盯着外立面的玻璃幕墙够不够亮。
美国银行这栋老牌金融大厦第三季度营收涨了3.2%到约299.6亿美元，看似楼体又加高了一层，但结构工程师一眼就能看出问题——12亿美元的罚款直接切进了承重墙，把每股收益削到0.91美元。这不是装修超支，这是地基浇筑时偷了工。反洗钱与制裁合规的罚单超过12.1亿，相当于在主体结构上钻了一排穿梁孔洞，抗震等级瞬间掉档。股价单日跌4.8%，市场不是在评价外观，是在重新核算这栋楼的剩余使用寿命。
对比一下隔壁两个工地。摩根士丹利的财富管理板块创下纪录，净新增资产2320亿，其中过半靠IPO带动——这是典型的裙楼商业面积热销，回款快、现金流漂亮。高盛的交易收益更是在给顶楼加装观景层。但问题在于：靠承重结构以外的附加面积冲业绩，能撑多久？IPO和交易收入本质上都是周期性施工队，行情一来蜂拥而至，行情一走就地解散。
真正决定项目长期估值倍数的，从来不是某个季度的精装交付，而是合规成本、信贷质量和利率环境这三根桩。高利率是持续的地下水位压力，信贷质量是土壤承载力，合规成本则是必须预埋的消防与抗震系统——省不得，还越埋越贵。
回到链上资产的联动，$xNFLX 这类映射标的本质是做传统金融的预制装配式构件，母楼一沉降，镜像结构必然同步变形。当银行业的合规地基被反复罚款掏空，当利润越来越依赖IPO和交易这种临时支撑，市场给出的估值就会出现明显的结构偏心。
恐惧与贪婪的刻度盘，本质上是建筑沉降观测仪。读数一旦跳变，说明某一根桩已经受力异常。 #bofaq3hitbyfinesThat warm high-pressure core called "advanced process" over the western Pacific just pushed the central pressure to a historic extreme — a single quarter net profit of 763.1 billion New Taiwan Dollars, up 68.7% year-on-year, revenue of 1.49 trillion, up 51% year-on-year, all solid sounding rocket data, not model-fitted illusions.
I've been watching this echo in the duty room for a long time. Its structure is very clean: the 2nm high-altitude jet just formed, contributing 8% of wafer revenue; and advanced nodes total 79%, meaning the main rainband has completely shifted north, with the old mature processes only seeing scattered edge precipitation. This is not an isolated thunderstorm cell, but a signal that an entire moisture transport belt has been thoroughly opened — the demand for AI chips is that continuous warm and moist airflow pushing from low to high latitudes, creating a sustained pressurized circulation background on the supply side.
But please note, my first rule in forecasting for twenty years: heavy rain upstream does not necessarily mean flooding downstream.
$xIREN type mapped targets are essentially a cumulonimbus cloud downstream of the main rainband. Its precipitation efficiency depends on three variables: first, whether moisture can be delivered — that is, whether the parent company's performance certainty is truly priced by the market; second, whether there is orographic lift — that is, whether capital is willing to move vertically at this position rather than drifting horizontally; third, and most critical, the local pressure field. Once the emotional dew point is suppressed very low, even the best convective clouds will be shredded by descending airflows.
The current situation is a typical "excellent upper-level pattern with severe low-level wind shear." On the long-term circulation map, the high-pressure ridge is strong and the hard indicators are flawless; but near the surface, panic and greed indices jump back and forth like a wind vane on a meteorological tower, and the wind shear is enough to tear any beautiful convective cell into stratiform clouds. I've seen this configuration many times — radar echoes glowing purple, but the landing zone shifts 30 kilometers, and the ground isn't even wet.
One more thing to watch out for: the stronger the advanced process high-pressure ridge, the more obvious the peripheral divergent subsidence. When an industry's profits are highly concentrated at the very upstream center, the mid and downstream will experience persistent clear sky zones, that kind of "seeing hotspots flying everywhere while your own place drizzles not a drop" drought is more tormenting than a typhoon. This is not alarmism, this is circulation conservation.
The numerical model I have shows this scenario: short-term strong convection maintained, mid-term cold air moving south along the credit spread path, whether it can cross the mountain range into this basin depends on the trading volume data from a ground observation station. The wind profile hasn't stabilized yet; any premature judgment of the landing zone is gambling probability based on experience.
The duty room fax machine is still printing maps; the next cloud image will come at dawn. The echo top has already reached 16 kilometers, surpassing the tropopause. Such a strong system has only one meteorological definition — a strong convection warning has been issued. As for who it will hit, that depends on the wind, not wishes. #TSMCQ3RecordProfit Tesla at $367, 1.1 billion quarterly net profit, 340x PE.
Selling 480,000 cars can't support a 1.45 trillion market cap; the market is buying the Robotaxi pie that hasn't started running yet.
If automotive gross margin drops again in Q3 earnings, this premium will be cut by one layer first. The whole market is looking for a bottom, but institutions are not panicking at all. I reviewed yesterday's ETF data and found something unusual.
First, let's talk about $BTC. There was a net outflow of 31.44 million yesterday. BlackRock sold 13.34 million, and Fidelity sold 18.36 million. They had just bought over 20 million the day before, but sold it all the next day. This shows institutions are just testing the waters with BTC, pulling back immediately after the CPI spike.
Next, $ETH had a net outflow of only 51,900. Grayscale sold a little, BlackRock sold a little, basically no movement. The price dropped to 2315, but institutions neither bought nor sold; they are just watching.
$SOL is even more interesting: net outflow of 12,200, with only Bitwise selling. This is the eighth consecutive day of outflows, but compared to tens of millions on previous days, today's volume is almost negligible. The main force driving the dump is no longer ETFs but long liquidations in contracts.
My conclusion: institutions are not panicking; they are waiting. The selling pressure on ETH and SOL is actually decreasing. The current drop is mainly long liquidation triggering more long liquidation. The long-to-short ratio for SOL is still 2.68; longs haven't fully exited, and the chain reaction hasn't ended.
I still have no positions. Bottom fishing in this market is a gamble. I'll wait until the liquidation data is mostly out and the market stabilizes before considering.
Key levels: BTC 80000/78000, ETH 2300/2200, SOL 100/95
#美国CPI超预期升至3.7%，今晚PPI再迎考验
#BTC成交萎缩，ETF买盘能否回暖
#ETH触及2500美元后震荡 As soon as the CPI was released, the entire market wailed; even ASML's huge profits couldn't save it. Is there still a chance for a reversal in the US stock market?
When the US September CPI data came out, the whole market immediately plunged into collective panic selling. Inflation year-over-year was 3.7%, 0.1 percentage points higher than the market expectation of 3.6%, and month-over-month it surged 0.6%, reigniting market fears that the Federal Reserve would resume rate hikes. Both the US stock and crypto markets fell sharply in response.
The S&P 500 dropped 1.11% in a single day, ending the previous record of 53 trading days without a single-day drop exceeding 1%; the Nasdaq fell 1.88%, with the semiconductor sector hit hardest, as the Philadelphia Semiconductor Index plummeted 4.3%. As early as the 8th of this month, I warned everyone that Samsung and SK Hynix's stock buyback cycles had ended, and the memory chip sector's stock prices would come under renewed pressure. I wonder if those who heeded the warning managed to exit at the top.
The lithography giant ASML delivered a quarterly report far exceeding expectations: quarterly revenue of €11.66 billion, a 55% year-over-year increase, net profit close to €5 billion, and a gross margin reaching 57.5%. All core indicators outperformed market expectations. Wall Street had hoped it could be the "firefighter" to stabilize the market.
The logic for a quick reversal in the US stock market is actually simple: as long as the Strait of Hormuz reopens, crude oil prices will fall, inflationary pressure will naturally ease, and the market can quickly get back on track. This crucial choice now lies in Trump's hands.
#美国CPI超预期升至3.7%，今晚PPI再迎考验 收益多一点，要付出什么？
$PENDLE 我觉得有个细节值得留意：它交易的是未来收益，未来越难判断，买卖双方越有分歧。
平台把生息资产拆成本金部分和收益部分，其中YT的收益权有期限，到期后便不再产生收益。
因此，买YT不能只看眼前收益率，还得算剩余时间里，实际能收回多少。
对PENDLE的判断，我更关注这些交易有没有持续需求。产品可以帮助用户管理收益，代币能获得多少价值，还要另外研究。
$CRV 要考虑的是时间成本。
直接锁定CRV获得veCRV后，锁定不可撤销，veCRV也不能转让，原币需要等到期才能取回。
这意味着不能只比较奖励多少，还得问自己，这段时间需不需要用钱。
行情变化时，能否调整持仓，本身就有价值。锁得久之前，要把这个代价算进去。
$JUP 我会先核对奖励条件。
它公布的活跃质押奖励方案，涉及投票活动和社区参与，并非单纯持币就能领取。
所以看到别人晒收益，先别直接套到自己身上。
参与时间、资格和规则都可能不同，过去一轮拿到多少，也不能当成下一轮的固定收入。Seeing liquidation warnings one after another in the early morning made me a bit tense. Is this drop really just because of the CPI? CPI year-on-year is 3.7%, core inflation exceeded expectations, and the market's first reaction was a rise in tightening expectations, a stronger dollar, and risk assets being directly pressed down. BTC, ETH, SOL, DOGE all plunged almost simultaneously, the total liquidations across the network quickly piled up, and longs were continuously trapped. My three long positions in SOL, DOGE, and ETH are also deeply buried; SOL fell from 111.35 to 104.23, DOGE slid from 0.09071 to 0.08152, ETH dropped from 2594.23 to 2416.51, with a total unrealized loss of 4537.93U. It’s false to say I’m not upset, but more worth recording than the discomfort is what exactly the market is trading during this volatile phase. First, the facts: CPI exceeding expectations means the rate cut path is pushed further back, real interest rates rise, and short-term funds’ first choice is to reduce high-beta assets. Crypto, as the far end of risk appetite, naturally gets attention pulled away first. But here’s an easily overlooked point — this sell-off is not entirely macro-driven; it’s more about the leverage structure reinforcing itself. Once the price breaks key levels, stop-loss and forced liquidation orders trigger automatically, and selling pressure snowballs, so you see the drop is fast and sharp, with sentiment collapsing before fundamentals. The bullish logic isn’t absent either. Although core inflation exceeded expectations, it hasn’t spiraled out of control to the extent that requires renewed rate hikes; the market’s pricing of tightening feels more like a one-time emotional reaction *Why it’s booming:*
1. *AI / HPC dominance:* High-Performance Computing = 68% of revenue, up from 57% last year and up 22% QoQ. Smartphones only 21%
2. *Advanced nodes = 79% of wafer revenue:*
- 2nm: 8% (vs 3% in Q2) — ramp is accelerating
- 3nm: 33%, 5nm: 28%, 7nm: 10%
3. *Pricing power:* Operating margin beat its own 58% forecast upper bound短线王送了会外卖，依旧是外卖补仓，狠狠掏了王兴十几刀，回来一看持仓，电动车被狗庄掏走了。
当前版本为地狱模式，CPI超预期反而迫使热钱流入美债，细心的朋友们会发现长端美债收益迎来大跳水，美债的大跳水又对各大市场的压制降低，听起来有点左右脑互搏，十年期美债对美股加密具有压制，无风险的年化5.3，为什么还要在加密美股这样高风险的市场博弈？目前美债的盈亏比十分诱人，因此在CPI出来的时候，美债先是拉高然后跳水，对黄金加密压制降低，asml的前瞻指引，今天的台积电，业绩都是非常强，进一步说明ai逻辑还在，控制风险等待反弹。$ETH
主流今天继续承压，凌晨两点美联储褐皮书进一步反应通胀将有可能继续上涨，美联储最忌讳通胀继续抬头，意味着需要持续加息才能压制通胀。$BTC
今晚还有重要数据，零售月率与PPI，PPI又称为CPI的先行指标，企业成本上涨最终会转嫁到消费端，零售月率则代表消费者的消费需求，用于观测加息是否导致经济疲软，消费端消费需求下降。$NEAR
#美国CPI超预期升至3.7%，今晚PPI再迎考验 #台积电Q3净利润超预期，盈利创历史新高 The market is still falling, but $ZK managed to generate proofs for 7,200 consecutive Ethereum blocks with just one graphics card, with 99% of the proofs completed within 10 seconds, rising about 12% against the trend. Meanwhile, $ARB and $OP, which are also in the L2 track, weakened. But does strong technology necessarily mean it's worth chasing? ZKsync Era's DeFi locked value is only about $15 million, far below ARB and OP. More importantly, the originally planned cross-chain fee mechanism has not been activated after the v31 adjustment, so token demand may not necessarily increase. However, the top ten addresses hold nearly 51%, so personally, I think the current state is more like being controlled by major holders. Therefore, I admit that ZK is stronger than its peers in the short term, but I tend to believe it is driven by a combination of technical benefits and trading capital.BTC ETH 以太坊继续下跌📉，2300是均线120的支撑位，这个位置支撑不住，那就是准备去1800
现在该多头表演了，我倒要看看多头多牛逼，能不能支撑住。宏观面美国9月CPI升至3.7%超预期，美债收益率高企，降息预期持续降温，风险资产整体承压，BTC围绕8万关口博弈，大环境偏空难有独立行情。
我2700的时候，有人说我是燃料，我2500的时候，有人说现在是洗盘，现在2300了，还有什么借口？继续讲，我看看多头是不是套树上给套的神智不清了。当前ETH日内跌超2.4%，弱势明显。图中50x永续空单入场2272，标记2369，浮亏中，高杠杆在剧烈波动里如履薄冰，多头连环强平风险与空单挤压并存。
台积电Q3营收创新高，盈利表现待揭晓，科技基本面虽强，但加密市场仍受宏观流动性主导。2300多单密集，一旦失守，下方缓冲变薄，1800或是下一站。不赌方向，轻仓防守，严控杠杆。$BTC $ETH $ZEC In the past 24 hours, approximately $975.5 million worth of contracts across 10 platforms covered by CoinGlass were liquidated, with long positions accounting for about $893.6 million, or 91.6%. For ETH alone, about $293.5 million was liquidated, with long positions making up 91.8% of the liquidated amount.
The directly affected are leveraged positions with insufficient margin. After price drops and expanding account losses, platforms will forcibly reduce or close positions. When long positions are liquidated in concentration, the sell orders from liquidations may continue to push prices down, triggering nearby positions; whether this chain reaction continues depends on whether buyers can absorb the selling pressure.
Next, attention should be paid to whether new liquidations in shorter time windows continue to increase, rather than only focusing on the 24-hour cumulative figures. The rolling window automatically removes old records, so a decrease in cumulative amounts may simply be old liquidations exiting the statistics and should not be directly interpreted as risk disappearing.
Source and images: Original CoinGlass page, captured at 15:38 Beijing time on 10/15; refresh times of various modules in the image vary slightly. Statistics are based on contract data from covered platforms, including non-crypto assets, and do not represent complete market losses. The ETH/USDT perpetual below is provided only as a market entry.
https://www.okx.com/zh-hans/trade-swap/eth-usdt-swapSame root bearish candlesticks, three different causes: Who is truly falling among BTC, ETH, and SOL?
All three coins are dropping together with similar candlestick patterns, but the reasons behind them differ.
BTC is a "sentiment cold." The market turning cold triggers a chain reaction of sell-offs, but on-chain data shows no mass exit by whales, and support remains near the 400-day moving average, with signs of accumulation at low levels still present. Its weakness is a retreat in sentiment, not a collapse of holdings.
ETH is "capital bleeding." ETFs have seen continuous net outflows, wallet balances are moving out, and some holders are actively reducing positions. It’s not just dragged down by the market but a real contraction of capital. Recovery depends on new funds returning, which is more challenging.
SOL is "contract internal injury." Perpetual short positions are increasing, leverage is pushing the price down, but spot outflows are limited, and holdings have not changed hands on a large scale. Although the decline is fierce, the source is mainly on the derivatives side.
Simply ranking by price drop can easily mistake emotional fluctuations for trends. Breaking down with on-chain data helps distinguish who is truly selling and who is just being dragged down. BTC awaits sentiment recovery, ETH awaits capital return, SOL awaits leverage clearing—knowing where the pressure comes from tells you what to be cautious about and what to ignore.
$BTC $ETH $SOL
#美国CPI超预期升至3.7%，今晚PPI再迎考验
#BTC现货ETF再度净流出，机构买盘能否回暖？
#交易之声：你的经验值得被听到 #台积电Q3净利润超预期，盈利创历史新高
台积电又交出了一份强财报，AI 这条线，确实有人在真赚钱。
2026 年 Q3 净利润约 7631 亿新台币，同比增长 69%，高于 LSEG 预期的 7463 亿，刷新单季纪录。
但看财报，有个细节值得拆开：同比增长 69%，不代表超预期 69%。按上述数据算，这次净利润超出预期约 2.3%。增长很强，但市场原本给的预期也很高。
所以，交易上真正要看的是：这份成绩单，还有多少没被 Price in（提前计价）？
再看前瞻指引：台积电给出的 Q4 营收区间为 **523亿—543亿美元**，高于 Q3 的 467.2 亿美元；Q3 毛利率达到 **66.4%**。这两组数据，比单独一句“盈利创新高”更值得琢磨。
我的理解是，AI 基础设施需求仍有业绩支撑，但行情能不能继续走，还要看后续盈利预期上修的幅度，以及市场愿意给多少估值。#台积电Q3净利润超预期，盈利创历史新高
美银分析显示，台积电主要客户当前市盈率较台积电溢价约45%，高于长期平均溢价22%；HPC客户的溢价更高达100%至150%，远超历史均值53%。美银认为，台积电股价过去两年的涨幅主要由盈利上调驱动，估值倍数扩张有限，未来随着技术领导地位进一步强化，股价存在重估至至少20倍NTM市盈率的空间。Macro Risk Repricing - Macro Risk Repricing
Macro risk repricing refers to the process where the market reassesses the future cash flows, discount rates, and risk premiums of assets after changes in key variables such as inflation, interest rates, employment, fiscal policy, or geopolitical factors. Its impact often extends beyond a single asset: rising interest rate expectations may lower the valuation multiples of overvalued tech stocks and reduce the appeal of assets that do not generate interest income; changes in the US dollar, bond yields, and risk appetite further influence the allocation of funds among stocks, gold, and crypto assets. Therefore, market analysis should not only focus on candlestick charts but also identify the macro pricing logic behind price changes.
Currently, SanDisk and Micron face the dual challenges of storage industry competition expectations and macro interest rate pressure; $BTC has seen a single-day inflow from ETFs but still requires sustained validation; $ETH continues to be dragged down by cumulative ETF outflows. The most important thing going forward is not to predict daily price movements but to track three changes: whether the price can hold key areas, whether fund flows can continue to improve, and whether macro data continues to increase the discounting pressure on risk assets.
#TSMC Q3 net profit exceeds expectations, profits hit a record high
#US CPI exceeds expectations, rising to 3.7%, with PPI facing another test tonight 先说结论：$SKDD 上市第二天走出教科书递进建仓——8 根 4H K 线 8 根红，量从 6.7 万堆到 60.4 万，9 倍。
10-14 08:00 上市第一根 9.54 开 9.57 收，量 6.7 万——试探。
10-14 16:00 才摸 10 块，量没起（3.7 万）。
转折 10-15 00:00：破 10.5，量 9.3 万——庄家干活。
04:00 收 11.36，量 12.6 万（+35%）。
08:00 拉到 11.88，量 60.4 万——分水岭，单根吃前 24h 全部量。
12:00 还在 11.92 横着，量缩到 47.7 万，价格没崩——典型控节奏。
关键数据：上市 32 小时 9.54→11.92（+24.9%），成交 $1700 万。不是散户堆的。
横向：BTC -2.78% ETH -5.49%，全市场 266 跌 20 涨，$SKDD 独立行情。资金是来占位的。
「8 根 4H 阳线 + 量能 9 倍递进」只在建仓期见过。回撤看 11.13，守住就强；破位警惕 10.55。
$SKDD 是 TradFi 永续里第一个走出独立行情的样本吗？Conclusion first: Because the driving logic of platform tokens is different from mainstream coins — they focus more on the platform ecosystem and less on macro factors. OKB dropped from $123.29 yesterday to a low of $120, closing at $122.92, down only 0.41% for the day, with a trading volume of $11.5 million. In comparison: BTC fell 2.1%, ETH fell 3.4%, SOL fell 2.8%, and ZEC fell 6.5%. On a day when CPI triggered a broad market decline, OKB barely dropped and was the most resilient among the Top 10. Why? The pricing logic of platform tokens (like OKB, BNB) differs from BTC/ETH. They do not directly follow macro liquidity and risk appetite but depend more on the issuing platform's ecosystem activity, buyback/burn mechanisms, and user demand. So when macro negative factors like a "hawkish CPI" hit, platform tokens usually fall less than mainstream coins. This explains why OKB only dropped 0.4% today. Looking at the structure: spot price $122.92, perpetual contracts aligned, no de-peg. Open interest about $31.63 million; funding rate +0.005%, mildly positive. The structure is healthy with no anomalies. Looking at the position: currently $122.92, below the 7-day moving average ($124.1) and 14-day moving average ($124.7), but still above the 30-day moving average ($121.58) and 60-day moving average ($116.16). The 14-day high is $143.32, 1Anomalous market reaction after CPI release: US Treasuries fall, but BTC continues to weaken
$BTC
US Treasury yields surged to 5.33% last night, then retreated to 5.11% this morning. Normally, with rates falling, BTC should rebound, but this time BTC dropped instead, currently hovering around 81,000.
This drop in Treasury yields is not due to easing inflation or weaker rate hike expectations, but rather a stock market crash driving safe-haven flows into Treasuries. The Nasdaq plunged 1.9% last night, marking the largest single-day drop in July; Asian markets followed suit, with the South Korean market falling as much as 8% intraday, and the AI sector suffering widespread heavy losses.
Most AI companies rely on borrowing for expansion, facing huge pressure in a high-interest-rate environment. BTC is classified by the market as a tech risk asset and is being sold off alongside; computing power mining companies saw declines up to 13% last night. Therefore, this Treasury yield decline is not positive for BTC, but rather signals rising market panic.
Crude oil remains under pressure, with Brent holding near $105 amid ongoing geopolitical conflicts. The International Energy Agency is discussing releasing 100 million barrels from emergency reserves, which is currently the only variable that could ease oil prices.
Previously, I said that a drop in Treasury yields below 5.2% would signal BTC's rise, but this time the logic is different, and I am not bullish. A true upward chain requires oil prices to fall, inflation to ease, and Treasury yields to decline in sequence.
Before that happens, BTC will continue to be pressured alongside tech stocks, with the 80,000 level as the next support to watch.
#USCPI unexpectedly rose to 3.7%, with PPI facing another test tonight September CPI year-on-year 3.7%, core year-on-year 2.6%, both exceeding expectations. Once the data was released, US Treasury yields surged, the probability of an October rate hike shot above 40%, and tonight there are still PPI and retail data, the test is not over yet!
Looking again at $BTC. After the CPI release, BTC was hit down to around 80,000, bounced back but couldn't break above 81,000. ETF net outflow in a single day was 261 million, with Fidelity alone pulling out 153 million, wiping out the previous day's inflows. Institutions chose to reduce exposure after the key data landed, unwilling to hold firm at this level. 80,000 is the last short-term defense line, the moving averages above are pressing down hard, if PPI again exceeds expectations, a breakdown would trigger a bull stampede.
Looking at $ETH. ETH dropped to around 2,300, among the largest declines in the mainstream. ETF net outflow in a single day was 130 million, none of the ten products had net inflows, BlackRock pulled out nearly 110 million. It is now the preferred cash-out machine for institutional retreat; when BTC weakens, ETH falls the hardest.
Finally, looking at $XAUT gold. Gold broke below 4,100, hitting a low of 4,073, then sharply bounced back above 4,150 in early trading today. Why the rebound? The ones running are futures speculative longs, while gold ETFs and central banks have been buying at the bottom. Central banks have increased holdings for 23 consecutive months, the scale is completely different.
These three assets cannot be viewed with the same logic. For BTC, watch if 80,000 can hold; for ETH, watch when ETFs stop bleeding; gold is the only one with solid logic supporting the bottom. The data just landed, hold spot base positions firmly, unload all short-term leverage. #美国CPI超预期升至3.7%，今晚PPI再迎考验 CPI Shock Night: More Fright Than Delight (｡•́︿•̀｡)
September's US CPI brought no surprises, only shocks. Overall year-on-year at 3.7%, higher than the expected 3.6%; month-on-month 0.6%, core year-on-year 2.6%, month-on-month 0.3%, also exceeding expectations. A nearly 9% monthly surge in gasoline prices was a driver, but excluding energy, the core is still pushing up, indicating everyday price hikes haven't stopped. (╯°□°）╯︵ ┻━┻
The Fed is most troubled: the core can't be contained. The market immediately changed the script, with the probability of a rate hike at the end of October jumping from about 20% to nearly 50%. The dollar and US Treasury yields both rose, pressuring risk assets. $BTC data dipped near $82,000, down 1%-2%, but no panic selling, showing restraint. (￣▽￣)
More importantly, the day before the CPI release, the US spot Bitcoin ETF saw net inflows of about $175 million. Institutions didn't withdraw; instead, they continued accumulating. (•̀ᴗ•́)و
Short-term macro sentiment will stir prices, but the long-term institutional allocation logic remains unchanged. One day of bad news isn't scary; as long as funds keep flowing in, there will be buyers at the bottom. (๑•̀ㅂ•́)و✧
$BTC $ETH $ZEC
#美国CPI超预期升至3.7%，今晚PPI再迎考验
#台积电Q3营收创新高，盈利表现待揭晓
#交易之声：你的经验值得被听到 I am the mid-term intelligence brother
Following up on the previous $ETH positive news, currently there is considerable "potential challenge" selling pressure, also giving everyone a heads-up.
Although the ETH ETF has just been approved, it previously fell for 10 consecutive days with an outflow of 783 million, and in October a total of 182,000 coins entered exchanges. The exit queue hit a new 10-month high, with over 1 million ETH waiting to be unlocked, and the price fell below the 2617 long-short line.
It has fallen 39% in the past year, underperforming $BTC, and the on-chain economic scale is also less than TRON.
More worrying is the loosening of old chips: a nine-year-old holder deposited 6,369 coins (15.95 million USD) into Kraken, and dormant wallets transferred 48,000 coins into Bitfinex. The selling pressure is real.
But from a mid-term perspective, institutional compliant accumulation is still ongoing (Bitmine is only 77,000 short of 5%). Macro disturbances are just shakeouts, don’t get washed out by panic, keep an eye on the chips!
$SOL
#US September CPI rose to 3.7% year-on-year, core inflation exceeded expectations $SOL no longer wants to rely solely on Meme for survival; on-chain US stock tokens are also starting to capture the market
Solana's on-chain US stock token trading volume has reclaimed the weekly top spot, accounting for about 31%.
Other chains are also fiercely competing:
SOL: 31%
BNB Chain: 26%
Robinhood Chain: 25%
Base: 18%
SOL is temporarily leading, but BNB and Robinhood are already neck and neck competitors; no one can easily dominate the entire market.
SOL is trying to break away from relying purely on Meme-driven traffic and expand trading scenarios to traditional financial assets like US stocks. However, high trading volume does not necessarily mean SOL will rise. How much transaction fees and real SOL demand US stock token trading can generate still needs to be observed.
Meme is responsible for generating hype, US stocks for broadening scenarios. If Solana can truly develop both sides of trading, its value should not be measured solely by Meme mania.$TRUTH This coin has no retail investors taking the bag. I found a bot doing wash trading, which means the previous pump was also driven by this tactic, but retail investors didn't follow the trend, so it was dumped.On October 13, Circle explained the redemption arrangement for cirBTC: each token is backed by 1 native BTC, but only eligible Circle Mint customers can redeem directly from Circle; other holders must trade through third-party markets.
This is the easy-to-overlook aspect of wrapped BTC: although the name includes BTC, the exit options are not necessarily the same. Having an underlying asset and having direct redemption rights are two different things.#TSMC Q3 net profit exceeds expectations, profit hits record high TSMC is not a "cyclical rebound," but the "tax collector" of AI computing power. No matter how Nvidia, Apple, AMD, and Google compete in AI, in the end, they all have to pay the "manufacturing tax" on TSMC's production lines. Who has design capability is not important; who can secure 2nm/3nm capacity is what matters.
So the real signal of this earnings report is not "TSMC is very profitable," but that AI infrastructure has not peaked yet, and the moat of advanced process technology is actually getting deeper. TSMC's bargaining power, yield, packaging, and scale are no longer the logic of an ordinary foundry, but rather "the foundational infrastructure stock in semiconductors."
In the short term, the valuation is not cheap; in the long term:
Having TSMC ≠ guaranteed win, but without TSMC = basically out of the AI chip game #TSMC Q3 net profit exceeds expectations, profits hit a record high
The rigid demand for advanced process AI chips has allowed TSMC to break free from the traditional semiconductor industry's cyclical fluctuations, achieving 11 consecutive quarters of profit growth. A net profit margin of 51.1% far surpasses the vast majority of global tech companies, proving that the scarcity of high-end advanced manufacturing has transformed into strong pricing power and profitability.#TSMCQ3RecordProfit TSMC's revenue has already exceeded expectations, so what is the Thursday earnings call really about?
TSMC's Q3 revenue was actually announced early, about $46.7 billion, higher than the company's own forecast of $44.6 to $45.8 billion. The Thursday earnings call isn't about revenue, but about the costs behind that revenue.
The first hurdle is gross margin. The Q3 guidance is 65% to 67%, with a midpoint of 66%, which is 1.M $ASML $NVDA
#TeslaQ3ProfitAndMargins $BTC is bearish. The current price is 81,051.2. I judge that the intraday low of 80,188 will not hold, and it will make new lows next. Long positions liquidated in the last 24 hours amounted to 86.02 million USD, more than twice the short positions. However, the number of short liquidations is actually higher, with 2,486 trades compared to only 1,298 for longs. The spike to 84,117.6 squeezed out a batch of small short positions, while the drop afterward forced out large long leverage. The main ones squeezed out were longs; shorts were basically unharmed. Options put/call volume is 1.05, higher than the open interest ratio of 0.91, indicating that new trades that day leaned bearish. DVOL at 39.0 is not high; the market has not priced in a panic-driven drop yet, so there is room to fall further. Funding rates have turned negative for three consecutive periods, but this is just background information and not a reason by itself. The total stablecoin supply of 306.7 billion USD is stock and does not mean someone is buying today. The chart shows moving averages in a bearish alignment and price hugging previous lows, consistent with this judgment. The condition to turn bullish: reclaim 84,117.6, which would invalidate this bearish view. #TSMC Q3 Net Profit Exceeds Expectations, Earnings Hit Record High TSMC's revenue has already surpassed targets, so what will Thursday's earnings call focus on?
TSMC's third-quarter revenue was actually announced earlier, about $46.7 billion, higher than the company's own forecast of $44.6 to $45.8 billion. Thursday's earnings call will not focus on revenue, but on the costs behind that revenue.Core CPI at 2.6%, 0.1 points higher than expected
Core CPI came out at 2.6%.
The expectation was originally 2.5%.
How this number is calculated:
It excludes oil prices and vegetable prices.
What remains are rent, services, and such.
This part can't come down, so interest rates can't come down.
Who fears this number the most:
Project teams fear it the most.
If interest rates don't ease, new money won't flow into altcoins.
$ETH fell first because it is closest to the funds.
Looking further down the line.
$BTC is still more than 3% away from 80,000.
US Treasury yields have already risen to 5.27%.
Money is flowing into bonds, not into coins.
Altcoins are being drained, not because someone is targeting anyone.
It's because the faucet hasn't been turned on.
#StandardChartered opens BTC and ETH trading, expanding banking crypto business
#$ETH $BTC # TSMC Q3 Net Profit Exceeds Expectations, Earnings Hit Record High
TSMC's Q3 earnings in one sentence: AI computing power demand pushes the foundry leader to new heights, advanced processes become an absolute cash machine.
Explosive performance: Revenue of 1.49 trillion TWD, up 50.9% year-on-year; net profit of 763.1 billion, up 68.7% year-on-year, profit growth outpaces revenue — high value-added capacity fully utilized, profitability efficiency hits a new high.
Structural disruption: Advanced processes (7nm and below) account for 79% of wafer revenue, 2nm just started mass production and already takes 8% share. Traditional mature processes are squeezed to the sidelines, technological barriers directly translate into pricing power.
Underlying logic: Growth almost entirely driven by AI chips. Nvidia, AMD, and cloud providers' self-developed ASICs are all lined up on TSMC's production lines; it is the most guaranteed winner selling "shovels" in this AI wave.
Outlook:
• For the industry: Confirms that AI hardware demand remains high, supporting the logic of rising equipment/test/material supply chains;
• For TSMC: 2nm ramp-up plus AI orders sustain short- to mid-term high prosperity;
• Risks: Huge capital expenditure, geopolitical issues, and capacity allocation are long-term variables.
Investment perspective: This is a "prosperity confirmation," not "expectation speculation," but the stock price already includes optimistic expectations, so chasing highs depends on your own position and risk tolerance. #AMD Earnings Preview: HeliosAI System Commercialization Progress
$SOL This chart has something very awkward.
It hit the trending searches on major exchanges, discussion is at its peak, yet the price has been dropping from near the weekly low, down about 5% in just over twenty hours. The hype is hype, the market is the market, and these two things are completely separated right now.
What excites everyone is that it can appear domestically, can run ads, and the upcoming new consensus upgrade claims to reduce transaction confirmation to near-instant levels. Sounds impressive. But look at the price—the spot holders are running, volume is about the same as usual, meaning it doesn't take much volume to push the price down, and no one is buying at the bottom.
This is what I call abnormal: a "positive" event paired with a "declining" price means the market has already priced in this good news. Now those buying are the ones catching the falling knife, not the ones profiting. $SOL