Orbit Post Sitemap
BTC has rebounded, but ETF buying has shrunk by 95%: This recovery is still missing a leg
BTC is around 82,681, ETH about 2,439, DOGE about 0.0815. All three have bounced back from their lows in the past 24 hours but have yet to reclaim their respective highs: BTC targets 83,440, ETH targets 2,478, DOGE targets 0.08234.
What really matters is not "how much it has rebounded," but whether buying momentum is being sustained. Farside data shows that the US spot BTC ETF had a net inflow of $526.7 million on October 19, but only $23.2 million on the 20th, still a net inflow but with an incremental drop of about 96%. So I don't attribute this rise directly to continuous institutional buying; ETF buying hasn't disappeared, but its marginal strength has clearly weakened.
My judgment: The short-term recovery is valid, but the trend reversal is not yet confirmed. BTC needs to stand above 83,440 and hold on a pullback, ETH must simultaneously reclaim 2,478, only then can the mainstream be considered collectively strengthening; if BTC falls back below the 24-hour opening level of 81,751 and ETH also loses 2,395, this rebound risks turning back into a decline.
For now, I won't chase the highs; I'll wait for the highs to be reclaimed or for the opening levels to be broken before making further judgments.
$BTC $ETH #BTC现货ETF连续净流入，市场反应有限 #OKX星球话题来啦 I think the new high in Ethereum staking shouldn't be rushed as a price increase signal. By the end of Q3, about 43.8 million ETH were staked, accounting for approximately 35.9% of the supply, but the yield was suppressed to about 2.63%. Spot price is around $2446 (Beijing 14:35, OKX).#特斯拉交付超预期后盈利修复成焦点
Tesla's Q3 delivery data has landed, with overall performance better than market expectations, and short-term sentiment has clearly warmed. But now the core market contradiction is no longer whether cars can be sold, but whether profitability can truly recover.
For a long time, price cuts to boost volume have been Tesla's main strategy. While volume increased, the gross margin per vehicle has been under continuous pressure. Coupled with the heavy ongoing cash burn in AI and robotics businesses like FSD, Optimus, and Dojo, this has been a key factor suppressing valuation. Delivery exceeding expectations only proves the fundamentals remain stable; it does not mean the profit inflection point has been confirmed.
Next, focus on several core points: whether automotive gross margin can stabilize and rise, whether the average selling price per vehicle has stopped falling, whether free cash flow continues to turn positive, and whether FSD commercialization progress can bring additional profit increments.
The bullish logic is: stable deliveries + declining costs, profitability is expected to gradually recover, and the long-term valuations of AI, Robotaxi, and humanoid robots will be re-priced.
The bearish logic is: the industry price war is not over, R&D investment remains huge, and even if deliveries look good, the pace of profit recovery may fall short of expectations, keeping valuations relatively high.
My view: exceeding delivery expectations is only the first hurdle; the real touchstone is the gross margin in the financial report. Before profitability data is realized, the market is more about sentiment repair. Be cautious about chasing highs; waiting for financial report verification before making decisions is safer.#Meta's Massive AI Investment Enters the Return Testing Phase Regarding Meta's earnings reports, in the last four reports, three showed declines. So this time, following the pattern, I also bet that its stock will drop after the earnings release. Last week, the Philadelphia Semiconductor Index fell for eight consecutive days, evaporating about $2 trillion, the longest streak since October 2008; it rebounded 2.4% on Monday, but some called it a "dead cat bounce." The bond market is becoming increasingly picky about AI financing. In the early hours of October 22, Meta will release its Q3 report, being the first major AI spender to report this week. The advertising side is likely not bad: market consensus revenue is $63.3 billion, up 23.4% year-over-year. But growth relies on price increases; exposure growth slowed from 19% to 14%, with unit price rising 12% for two consecutive quarters. The key metric to watch is unit price: if it falls below 9% and exposure continues to slow, it indicates the AI bidding dividend has peaked. The Q4 guidance consensus is $73.7 billion; if it falls below $72 billion, the market will view it alongside the spending curve. The real test is capital expenditure in 2027. Management did not provide guidance last quarter; sell-side consensus is about $193 billion. History is poor: last October, when the 2026 framework was significantly raised for the first time, the stock dropped 11.3% the next day; in the last four earnings reports, three showed declines. An acceptable range is within $195 billion or providing a phased path; exceeding $220 billion without new revenue guidance will likely repeat last year's scenario. The balance sheet is also under pressure: EPS consensus ranges between $6.17 and $6.76; Q2 operating margin was 31%, compared to 43% the same period last year. If a full $10 billion provision is made, GAAP margin could be compressed to 15% Oil prices surge past $110! BTC rebounds, but ETH and SOL are less optimistic?
On October 21, BTC was around $82,797, ETH about $2,447, and SOL approximately $106.23. BTC's rebound does not equal a bull market reversal; capital divergence is the key.
On October 19, the US spot BTC ETF saw a single-day net inflow of $526.7 million, but the ETH ETF had a net outflow of $103 million, and the SOL ETF also had a net outflow of $6.7 million, indicating that capital has not fully returned.
Macroeconomic risks cannot be ignored either: Middle East supply concerns pushed oil prices higher, with Brent crude briefly rising to $110.38. If energy prices remain high, inflation and Federal Reserve policy expectations may continue to suppress risk assets.
Next, watch whether BTC can hold above $82,000, whether ETH can reclaim $2,500, and whether SOL can break through $110. The rebound depends on sustainability; don’t mistake a single bullish candle for a reversal signal. $BTC #BTC财库优先股融资升温 #BTC现货ETF大额流入后转负 #BTC成交萎缩，ETF买盘能否回暖 Tesla will release its Q3 earnings tonight, and I chose NO in the prediction market.
Although Q3 deliveries exceeded expectations, the positive news has basically been priced in by the market. Compared to sales volume, I am more focused on the profit margin of the automotive business and the subsequent growth guidance.
Additionally, Tesla's stock price performance after recent earnings reports has not been very favorable, so I remain cautious this time as well.
Currently, YES is priced at 87¢, while NO is only 14¢.
Buying YES at 87¢ offers poor odds; betting on NO at 14¢ seems more worthwhile for a small position.
Of course, earnings season is highly uncertain, this is purely my personal judgment, and I am participating with a light position.
#特斯拉交付超预期后盈利修复成焦点 #NetflixQ3ProfitBeat
Netflix's Q3 earnings report exceeded expectations, with profit margins rising driven by price increases and stable user retention. However, the execution of advertising business fell short of expectations, with lowered advertising revenue guidance and weak Q4 revenue forecasts, putting pressure on valuation. Competition in streaming services is intensifying, and the price increase advantage for subscribers is reaching its peak, with growth focusing on advertising and live content. Short-term profitability remains confirmed, but the growth ceiling in the medium to long term is capped #BTC现货ETF连续净流入，市场反应有限
The capital flow and price trend are diverging; ETFs are experiencing net outflows while prices are bouncing up, indicating increased capital fragmentation.
Bitcoin ETFs saw a single-day outflow of 139 million, institutional buying is retreating, but BTC rebounded from around $80,000 to $83,000 and is now oscillating at a high level.
This rebound is not driven by new capital buying in, but by short covering pushing prices up. It's a battle of existing funds, so the price naturally rebounds. Currently, no incremental funds are entering the market, and this rebound is very likely to lead to a decline, with no guarantee of sustainability.
$BTC $ETH $ZEC $HYPE At 8 AM, I was watching 81.815, saying if it can't hold above that, it's weak. The Asian session proved that wrong—at 09:00 it really broke through, even reaching 82.345, a 24-hour high.
But it only lasted two hours.
The 09:00 surge with 920,000 volume hit 82.345 but closed at 81.84; starting from 10:00, there were four consecutive bearish candles: 81.68, 81.13, 80.74, 80.62. The combined volume of these four candles was 2.22 million, more than double the 920,000 volume of the surge candle. Up with 920,000, down with 2.22 million—this is a bull trap, not a breakout.
A real breakout leaves a step; the high of the breakout candle becomes support afterward. Here, 82.345 was only held for one candle and then completely swallowed; that price was artificially fed.
Current price is 80.80, MA20 at 80.45, treading on thin ice by about 0.35.
Short: 1H candle closes below 80.45 with volume ≥550,000 → enter on pullback between 80.60~80.75, stop loss at 81.15, target 79.37, risk-reward 2.8.
Long: 1H candle closes above 81.05 with volume ≥600,000 → enter on pullback between 80.85~80.95, stop loss at 80.40, target 82.345, risk-reward 2.9.
Don't mess around between 80.45~81.05, these 60 points are meaningless to scalp back and forth. Wait for US market open tonight to reassess risk appetite.
(Not investment advice, trade at your own risk) META is about to release its earnings report soon, and I want to share my real thoughts before the market opens. Right now, Meta is completely caught in a tug-of-war between "fiery expectations" and "cold reality." Recently, Muse AI has been incredibly strong, evolving from basic Q&A to an "executor" that can help you fill out forms, send emails, and even book services. This product strength has directly boosted Morgan Stanley's price target to $820. But don’t get too excited—I actually have a concern. Institutions are chasing the highs while simultaneously buying CDS aggressively to hedge against the debt risk of an AI bubble burst, which shows that smart money is nervous and buying insurance. The biggest pain point is still the cash burn. Meta is heading for over 130 billion in annual capital expenditures, severely draining free cash flow, and Muse’s traffic hasn’t fully converted into real revenue yet; the commercialization loop isn’t completely closed. So tonight’s earnings night, I’m focusing on two points: whether the advertising base has truly been driven by AI, and whether capital expenditures are out of control. If ad conversion rates exceed expectations and the cash burn rate stabilizes, then riding Muse’s momentum, the stock price reaching $820 is totally possible; but if capital expenditures keep soaring and monetization isn’t reflected in the results, then $820 is definitely a "poisoned milk"—the earlier gains will likely be given back with interest. In the short term, the pre-earnings positive news has already been priced in with a rally, and there’s definitely profit-taking pressure inside. Once the earnings report drops tonight, expect intense volatility.Term Structure Radar
The first-level positive basis for $BTC corresponds to a very small scale. The bid-ask spread for buying near and selling far is +0.96%, with the first-level pair size about $108; at a scale of 10,000, this level only covers about 1%. The curve has a shape, but the volume is too small.
The first-level positive basis for $ETH corresponds to a very small scale. The bid-ask spread for buying near and selling far is +1.13%, with the first-level pair size about $59; at a scale of 10,000, this level only covers about 0.59%. Capacity is capped.
For $SOL, the marked basis is positive, but the first-level bid-ask spread for buying near and selling far is inverted. Near-month annualized +1.9%, far-month annualized +0.35%, the near end is more expensive; the marked near-far difference is +0.1%, but the first-level bid-ask spread for buying near and selling far is -1.09%. The executable direction is opposite to the mark.The rebound is a correction, not a reversal
Contract-side capital inflow has fueled the rebound, but the trend reversal is still unconfirmed. The term structure remains torn: BTC and ETH have net inflows of 832 million and 257 million USD respectively in the past 24 hours, yet net outflows of 7.236 billion and 7.822 billion USD over 30 days. Short-term funds are replenishing, long-term funds have not returned, so it is still a correction.
🔸$BTC: Continuous net inflow from 1 hour to 24 hours, buying pressure dominant, short-term bullish; after inflow, beware of a spike and pullback, do not chase, wait for a pullback confirmation.
🔹$ETH: 257 million USD inflow in 24 hours, resilience is not weak; 7.822 billion USD outflow over 30 days, correction will take time, sustainability depends on capital relay.
🔹$SOL: 3.55 million USD net outflow in 24 hours, inflows fluctuate, lacking sustained signals, watch and wait.
🔸$ZEC: 48.77 million USD net inflow in 24 hours, but turns to outflow in 5 to 30 minutes, weakening support, wait for capital to return for confirmation.
In short: The rebound is measured, reversal unproven; short-term looks like correction, trend depends on long-term funds.
#BTC现货ETF连续净流入，市场反应有限
#SpaceXQ3营收大增161%，AI业务成增长主力 After lunch, I lay on the bed, originally wanting to find a direction, but the market was even more sleepy than me. So I quietly closed OKX again!
SOL is so stubborn, the only one among the top three with bears dominating in the last 24 hours, the long-short ratio is 0.9253, and the funding rate has even turned green to -0.0016%. On the eve of the upgrade, the bears are already positioned, betting on the opening at 21:20 tonight. Looking at the whole network, the 24h long-short ratio is bulls 52, bears 48—just 4 points more bulls, and they still dare to call it bullish? But everything changes in 5 minutes, BTC bulls shoot up to 73.61%, SOL also reaches 59.91%. Sentiment is counted by the minute, stance by the day, those shouting and those holding are never the same people, just get used to it.
BTC is hovering around 82899, ETH and SOL are also rising—but despite the rise, BTC holdings only increased by 0.95% in 24h. The mouth is full of bulls, but the wallet is honest.
Looking at major news: a ship was attacked near Hormuz, crude oil broke through 110 in one go. When geopolitics flare up, it's normal for long-term funds to retreat first. The whole network saw 276 million in 24h liquidations, 69% were shorts, over 70,000 people got liquidated, the largest single order was 11.96 million, a BTC short on a certain exchange. Prices rose, but all the buyers were shorts. 👀
Solana co-founder joked today that this upgrade could have been done in half the time. Alright, let's see the results tonight, I'll take a nap first, dreams have everything. 💤October 21 Morning Brief | ETF suddenly reverses to an inflow of 527 million, but only BlackRock is buying — can you guess why others are not following?
【Today's One-liner】ETFs are buying, but only BlackRock is buying. Whales are accumulating, but spot buying shows a negative premium. Market makers hold 85.5% of positions short. 82,500 is today's face value.
Today's key variables: ① Bitcoin spot ETF single-day net inflow of $527 million, a one-month high, but IBIT alone contributed $665 million, while FBTC and GBTC combined outflow was $139 million ② Whales' 60-day net accumulation changed from negative 62,600 coins to positive 5,230 coins, shifting from distribution to accumulation, but Coinbase premium remains negative at -0.1416% ③ Wintermute holds an $81.36 million position on Hyperliquid with 85.5% short positions, the largest short is a $22.61 million ETH 15x short35.9% of the $ETH supply is staked! Is this a super bullish sign or another kind of risk?
According to Bitwise data, Ethereum staking has reached a record 43.8 million ETH, about 35.9% of the total supply. In the first nine months of 2026, the newly staked amount was about 7.96 million $ETH
At first glance, this data is very friendly to bulls.
But I think this matter cannot be viewed from only one side.
First, staking does not equal permanent lock-up. Ethereum has an exit mechanism, so staked ETH cannot simply be understood as chips that will never be sold.
Second, a reduction in supply does not necessarily mean an increase in demand. If market funds continue to withdraw, even if circulating supply tightens relatively, prices can still fall.
Third, staking yields and price volatility are two different things. If ETH drops 20%, staking yields may not be able to offset the principal loss.
What bulls should really look forward to is not just more people locking up, but whether new demand can continuously exceed new selling pressure.
If the $ETH staking rate continues to rise, do you feel more optimistic about the future supply-demand structure, or are you worried that market liquidity will become increasingly fragile? Let's discuss in the comments. #BTC现货ETF连续净流入，市场反应有限 #SpaceXQ3营收大增161%，AI业务成增长主力 #奈飞Q3盈利超预期，广告业务增长放缓 $SOL Morning daily chart analysis set aside, focus only on this four-hour chart in the afternoon.
At 22:00 last night, a volume of 1.84 million pushed up to 108.24, the only decent attack in 48 hours, closing at 107.70. One hour later at 23:00, with 1.41 million volume, it was hammered back down to 105.79 — volume didn't increase, and the price fully gave back the gains. Small volume going up, large volume coming down, 108 is a distribution zone, not a breakout zone.
The following sixteen hours were spent consolidating, with each high point lower than the last: 09:00 touched 107.09, 11:00 down to 106.26, 14:00 only 105.67. The lows ground down from 105.78 to 105.54, four tests of 105.5~105.8 were bought back, but the buying strength weakened each time.
Current price 105.73, MA20 at 106.07 pressing down. The center of gravity is slowly dropping: the lower boundary holds, but the top can't be lifted.
Short position: 1H candle closes below 105.54 with volume ≥250,000 → enter on rebound at 105.70~105.85, stop loss at 106.25, target 104.50, second target 103.74, risk-reward ratio 2.7.
Long position: 1H candle closes above 106.15 with volume ≥280,000 → enter on pullback at 105.95~106.05, stop loss at 105.45, target 107.09, risk-reward ratio 2.0.
Avoid trading within the 0.60 range in the middle, as back-and-forth grinding only incurs fees.
(Not investment advice, trade at your own risk) SpaceX, Conference Call Interpretation (1)
After SPCX's Q3 report, several executives including Musk said quite a bit, and I will interpret the key points.
Starship and Starlink have no major issues and can provide good cash flow.
Let's consider AI and computing power leasing together. SPCX's AI revenue is also based on existing computing power, so we only calculate efficiency on a unified basis.
The base price is set at $50 billion/GW.
Comparing Q2 and Q3, the incremental revenue from SI is $5 billion, and Q4 is expected to maintain rapid growth.
Actually, what I care most about is capital expenditure. The Q4 planned expenditure is $35-40 billion, which will roughly add 1GW of computing power. The calculation logic is that by the end of October, it will reach 2.2GW, an increase of 0.5GW compared to the end of Q3, and it is quite likely to reach 2.7GW by the end of December.
Therefore, we can infer that SPCX's capital expenditure for adding 1GW of computing power only needs less than $40 billion.
Then, speculating further, the capital expenditure needed for 10GW of computing power by 2027 might be between $350-400 billion.
Averaged monthly, that's $29.17-33.33 billion, while the Q1 2027 revenue guidance from the conference call is $33-37 billion, about $11-12.3 billion per month, so the gap will continuously consume SPCX's free cash flow. Therefore, the $40 billion borrowing plan implemented in Q4 will be a key step in executing computing power construction.BTC institutional cost: The comprehensive cost estimate for the US spot BTC ETF announced on October 14 is about $81,500, with the current coin price approximately 1.81% above this reference; this estimate has not been updated today. As of October 18, Strategy holds 848,517 BTC, with an overall average cost including fees of $75,446, and the current coin price is about 9.98% above this cost; the above ratio only reflects the price difference relative to BTC cost. In brief market view: The near-term liquidation zones above and below BTC are close in distance, while ETH is closer to the potential short liquidation zone above. When BTC pulls back, focus on the $82,100-$82,400 range and the adjacent $81,500 ETF cost reference area for support; liquidation zones or institutional costs should not be directly treated as definite support. #交易之声：你的经验值得被听到 $BTC $ETH Last night a big bullish candle pushed up to 83,335, and everyone in the groups was shouting bull return, but I, Huahua, watched the order book all afternoon and didn’t dare move a single U in my hand 🥲
Last night the computing power leader CoreWeave ($CRWV) violently pulled from a new low of 57 dollars to 65 dollars after hours, temporarily lifting the alert on computing power debt defaults. $BTC leveraged the momentum to rise from 81,200 to now 82,695, and $ETH also pulled back to 2,442.
But how was this bullish candle pulled up? Look at three painful data points:
1️⃣ Purely from short squeezes, while spot was actually being dumped. Last night the whole network blew up 171 million dollars in short positions, forcibly pushing the price up; but the total market volume actually shrank by 1.96%, and the $BTC spot net sell gap expanded from 67 million dollars to 83.6 million dollars. Contracts were partying, while spot main players were offloading along with the buy orders.
2️⃣ Leverage is full again. The whole network OI rose 3.75% to pile up to 388.6 billion dollars, and the funding rate paid by longs surged nearly 50% this week.
3️⃣ Ancient whales stirred. Coinbase just transferred in 500 BTC that had been dormant for a full 10 years (41.33 million dollars).
#10月FOMC临近，美联储暂停加息还有哪些变数？ Less than 45 minutes after the US stock market opened, $BTC surged from 81K to 83K with a strong 15-minute bullish candle. Those who were still hesitating whether to cut losses during the day are probably just slapping their thighs now 😅 Current price is about 82,800, up nearly 1.12%, but the most interesting aspect of this rally is not the price itself, but the fact that the 4-hour open interest is shrinking during the rise. What does this indicate? Shorts were collectively triggered and forced to liquidate, pushing the price upward—a classic short squeeze structure, not new funds actively entering positions. In other words, this is a pulse triggered by position structure, not a trend confirmation signal. What really needs caution is the divergence in sentiment. $ETH is rebounding the strongest, directly hitting 2,450, but the 4-hour long-short ratio has soared to 2.59, indicating extreme retail crowding, all chasing the rally. Such uniform expectations often become a playground for whales to manipulate and shake out positions. $SOL also closed with a big bullish candle at 107.2, with a 4-hour long-short ratio of 2.33; clear resistance above 108 is visible, and blindly chasing higher carries significant risk. My judgment is: this rally is driven by US stock market capital overflow combined with short squeeze, representing a passive rise rather than an active offensive. Overheated retail sentiment, open interest divergence, and unbroken key resistance—these three signals combined mean chasing higher has very low cost-effectiveness. After the short squeeze, there will likely be a shakeout and pullback; waiting for the structure to stabilize before seeking opportunities is much more rational than rushing in now. To sum up this market move in one sentence: the price was pushed up by shorts' stop-loss orders, not by bulls' confident buying—this kind of rally, most10 years at 10% annualized, 55% volatility — JPMorgan's pricing for $BTC translates to: in the long run, it outperforms the US stock market and gold, but the journey will shake you up.
JPM Asset Management has included BTC in its long-term capital market assumptions for the first time, with a baseline scenario annualized return of 10%, higher than the S&P 500 and gold. This isn’t just Jamie Dimon talking; the asset management division has actually put real money into the model. The attitude has truly changed — from "scam" to "allocation option" — over ten years.
But the 55% annualized volatility figure is even more worth noting: it means if you hold for 10 years, you will likely experience more than one 50-70% drawdown. 10% annualized sounds good, but the question is whether you can avoid panic selling when it’s down 60%. A 30% drop in US stocks is already considered a crash; a 60% drop in BTC is just a normal correction.
So the question "If you can only pick one for the next decade" is essentially not about choosing returns, but about how much volatility you can endure without running away. Those who choose BTC aren’t buying 10% annualized returns; they’re buying the belief that "I can survive 55% volatility."
The big bank pricing BTC at 10% isn’t telling you to go all in tomorrow. It’s telling you: this asset has entered the long-term model and can no longer be classified as a "scam." $ETH $DOGE
#BTC现货ETF连续净流入，市场反应有限 JPMorgan's first BTC assumption set is notable for its restraint: a 10% base-case annualized return sits beside 55% volatility and unstable correlations. That combination matters most to institutions that need diversification to survive committee scrutiny, not simply upside.
A lower-volatility path could broaden BTC's role, but the forecast itself shows why adoption remains conditional.
#JPMorganBTCForecast SanDisk is currently at 1404, and I plan to place a short order at 1518. This is not chasing a short at the current price, but waiting for a proper rebound before taking action. Why choose 1518? From a technical perspective, this rebound has reached around 1518, which is exactly the dense trapped position during the previous downtrend consolidation, and also the place where the daily moving averages are heavily pressing down. SanDisk has been hammered down from a high level, the trend is already broken, and the current rally is purely a technical correction after being oversold. Rebounding to this position, there are many trapped chips waiting to be released above, plus selling pressure, making it harder to break through in one go than climbing to the sky. As long as the volume can't keep up, the rebound to 1518 will hit a wall. Looking at fundamentals and capital flow, the storage sector's sentiment has been very fragile recently. The DRAM progress at Changxin Technology has scared the entire sector badly. Although SanDisk mainly focuses on NAND, it is closely linked to the sector, and any disturbance causes it to fall faster than others. Moreover, in AI hardware, the market is now starting to calculate returns. As long as earnings reports or industry news slightly miss expectations, high-valued hardware stocks will be the first to suffer. Previous waves of big gains were overly optimistic, and now it's time to pay the debt. The method to place the order is simple. Place a short directly at 1518; enter when it reaches, otherwise keep waiting. Set the stop loss above 1580. For a highly volatile stock like SanDisk, enough room must be left for spikes. If volume breaks and holds above 1580, it means the short logic is completely wrong, so admit defeat and exit. Take profit in two batches: the first target is 1300, reduce half the position when reached, and the rest target 1200 or even lower. For every 50-point price drop, move the stop loss down by 30 points to lock in profits We bought SpaceX at a bottom of over 100, and added a bit at 140. The earnings report just came out, and it surged 8% after hours. Let's calculate whether its valuation is expensive now.
Revenue is 14.6 billion, a year-on-year increase of 161%. This quarter's net profit is 4.4 billion, while it was still a loss of 1 billion in the same period last year. The most impressive is the AI business (data center), which grew 830%, with revenue of 7.6 billion, surpassing Starlink for the first time.
We mentioned before that its data center leasing price is too high, and it can break even in just over a year. Anthropic and Google pay it tens of billions monthly; just these two companies pay nearly 7 billion per quarter, so such explosive growth is not surprising.
But the real money-maker is still Starlink. With 13.6 million users, operating profit is 2.4 billion, making it the company's largest profit source. The rocket business is also growing.
The problem is poor cash flow. This quarter's capital expenditure is 19.6 billion, mostly spent on data centers, resulting in significantly negative free cash flow. The current market value is about 2 trillion; based on current profits, the valuation is not cheap. If it can earn about 20 billion next year and profits rise by 2030 in the long term, the valuation will be more reasonable.
But short-term volatility will still be large, so don't rush to chase the highs.
$SPCX
#SpaceXQ3营收大增161%，AI业务成增长主力 #AI #美股 #数据中心The most difficult issue for the Fed this time might be the inflation caused by oil prices, which has started to clearly impact bills again.
The US September CPI rose 0.6% month-on-month, with gasoline prices up 9.1%, contributing more than half of the overall monthly increase. Seeing this data, interpreting “a possible pause in rate hikes in October” as inflation being no longer a problem is a bit overly optimistic.
The trouble with rising energy prices is that rate hikes can’t directly ensure safe passage for oil tankers, nor can they immediately increase refining capacity. But if costs continue to pass on to other goods and services, the Fed will find it hard to keep waiting.
This makes the subsequent policy path particularly awkward: there are reasons to observe for now, but also reasons to continue tightening later.
I’m more worried that the market only trades on a pause, without leaving room for this back-and-forth. Especially BTC and tech stocks—they get easily excited at the slightest hope of policy easing; then when inflation data comes back, valuations have to readjust again.
So I tend to look at crude oil and inflation together. If energy pressure really eases, a pause is more likely to be extended; if oil prices keep fluctuating, the central bank’s patience may also wane.
Guessing every day whether officials’ speeches lean hawkish or dovish is quite exhausting. This round, sometimes looking at gas stations and transportation costs makes it easier to understand why they are reluctant to loosen policy.
#10月FOMC临近，美联储暂停加息还有哪些变数？ JPMorgan did their own calculation and said Bitcoin could reach 385,000 in ten years.
My first reaction was, didn’t JPMorgan Asset Management CEO Jamie Dimon say that was a scam?
The three numbers they laid out are like this: first crash to 25,000, then rebound to 215,000 within ten years, and finally rise to 385,000.
But what really makes me anxious isn’t these three numbers,
the "how many people are using Bitcoin" factor in their model is only given a 5% weight.
The remaining 95% is 65% betting on continued global money printing, and 30% betting that it can take money from gold.
In other words, their reasons for being optimistic about Bitcoin have almost nothing to do with Bitcoin itself.
Oh, and there’s also a 55%.
Bitcoin can rise or fall 55% in a year, while the US stock market is only 18%. Meaning doubling or halving is not unusual.
The report itself knows this, so it didn’t even recommend you buy it, just wrote a line saying "ten-year forecasts are still too early, don’t build your portfolio based on this."
On one hand, they call it a scam, on the other, they put it on the list of best assets for the next ten years.
The guy from Bitwise said it pretty harshly: first ignore you, then laugh at you, then fight against you, and finally write you into long-term forecasts as the best performing asset of the next decade.
My own understanding is, 385,000 is a bet that the liquidity will come, not that the coin will be used.
As for whether that’s right or not, I’m not sure, but I know this is JPMorgan’s view, not mine.Elon Musk used to attract capital with imagination
Now he starts to prove himself with profits
So don't rush to call Musk invincible
AI business growth is indeed fierce
But the investment is equally shocking
Capital expenditure in the third quarter was nearly $19.6 billion
Making a lot of money
Spending is also really generous
So I think this earnings report
What really deserves attention is not the surge in revenue
But whether the AI business can continue to make money
After all, what the capital market likes most
Is companies that grow fast and can make moneyUrgent Notice
Samsung and SK Hynix upgrade NAND factories in China but are stuck at the next generation
To respond to competition from Yangtze Memory (Q2 China market share 38.5%, surpassing Samsung's 33.5%), Korean manufacturers are simultaneously upgrading their production lines in China:
• Samsung Xi'an (accounting for 40% of global NAND): from 200 layers to 286 layers, targeting a monthly output of 40,000 wafers by 2027
• SK Hynix Dalian: from 192 layers to over 240 layers, with phase two adding 50,000 wafers monthly
Bottleneck: BIS does not intend to approve expansion and upgrades in China, Dalian is only approved for 240 layers; meanwhile, Yangtze Memory has mass-produced 294 layers and plans for over 300 layers. Korean manufacturers' production lines in China are locked at the next generation, while they are pushing new factories in the US and R&D for 300+ layers.
Conclusion: The memory competition is divided into two lines — the China line defends market share, the advanced line pursues process technology. Yangtze Memory surpasses with process technology, Korean manufacturers want to use Chinese capacity to compete with Chinese rivals but are capped by policy limits. #Netflix Q3 profits exceed expectations, advertising business growth slows Netflix is currently in a transitional pain period between "on-paper financial prosperity" and "real growth slowdown." In the short term, market sentiment is pessimistic, and today will most likely be dominated by volatility or decline. The trading logic of the prediction platform in your screenshot may be more based on a bottom-fishing mentality of "the stock price has already fallen a lot, so there should be a rebound," but on the day of a financial report shock, such rebound expectations are often fragile.#奈飞Q3盈利超预期，广告业务增长放缓 In short: the core business is very strong, but the side business is lagging.
First, let's look at the fundamentals. Revenue was $12.874 billion, up 12% year-over-year, and net profit surged 44% to $3.664 billion. The most impressive is the operating margin reaching 34.6%, even higher than the company's own guidance. This shows Netflix still has very strong control over content costs and subscriber pricing, with a rock-solid foundation.
But where's the problem? The advertising business. Netflix cut its full-year 2026 ad revenue forecast from $3 billion to $2.7 billion, citing delays in launching some ad products.
Don't underestimate this $300 million cut; behind it lies a signal closely related to our macroeconomy. Who drives Netflix's ad revenue? Major brands and companies placing ads. The lowered ad forecast means companies are cutting marketing budgets and lack confidence in the future consumer market. This is essentially a microcosm of economic cooling.
This also perfectly explains why the Federal Reserve is struggling now. Economic data is weakening, but inflation remains stubbornly high, and the high interest rate sword hangs overhead, making it hard to lower rates.
Back to our big picture, the logic is very straightforward. Companies are hesitant to spend, and the consumer market is cooling, meaning the macro environment remains tight. As long as this expectation persists, risk appetite for funds won't rise, and BTC can only continue to consolidate around the 80,000 level, making it difficult to break out into a strong, one-sided rally. $BTC "Heaven and Hell Day by Day: The Storage Adjustment Isn't Over, But the Sell-off and Fundamentals Are on Different Time Scales"
Let's first look at the rhythm over the past three days: On Monday, the Philadelphia Semiconductor Index rose 1.61%, ending an eight-day losing streak; on Tuesday, US storage stocks collectively fell again, the KOSPI dropped 3.73% the same day, and SK Hynix fell 4.23%; today, Korean stocks rebounded in early trading. A different direction each day—this is not a reversal, it's a full divergence.
The reason for the sell-off lies in the long term. The trigger for this round was the news about ChangXin Memory's 4F2 architecture; in mid-October, Hynix fell over 10% in a single day, as the market fears that Chinese supply will eventually crush storage prices. But Goldman Sachs said the impact is overestimated; ChangXin is still some distance from true mass production.
Fundamentals are more recent, with a completely different tone. South Korean customs data: chip exports in the first 20 days of October rose 222.5% year-over-year, hitting $27.5 billion, a new high for the period. KeyBanc expects DRAM prices to rise 15% to 20% and NAND prices about 20% in Q4. HP and Dell have extended supply agreements to 2031, and Google is negotiating through 2035. Customers are using long-term contracts to secure goods, not cutting orders.
Therefore, the verdict is not in the market price but in the earnings reports: Hynix on October 28, Samsung on the 29th with full earnings, and SanDisk's report—all three will be released consecutively. There is also a small side note: Micron's Taiwan factory may strike on Thursday.
Conclusion: It's impossible to definitively say the adjustment is over now. When data and stock prices diverge, chasing rallies or selling off is the most costly. Wait for the guidance on the 28th before making a call.
#存储芯片股反复震荡，调整是否结束？ On October 20, Apple announced an expansion of its Made in USA program partnerships: Perryman will supply titanium materials, and Qnity will provide advanced film technology related to the camera's variable aperture system. Among these, Apple's commitment to Perryman is up to $100 million over the next three years. The most pressing question about this news is: at which specific stage of phone production do these collaborations take place, and can this be used to infer that full device assembly will move to the United States? The answer lies in the details of titanium processing and camera film technology. What does the titanium factory do, and what does the $100 million commitment mean? According to Apple's disclosure, Perryman conducts titanium melting, forging, and rolling at its Houston, Pennsylvania factory. These processes belong to material processing: from handling the metal to changing its shape and specifications, providing materials for subsequent manufacturing. This indicates that the U.S. factory undertakes specific production steps. However, material processing completion is different from component manufacturing completion and full phone assembly. Knowing only where the titanium is processed does not allow us to determine where it is later turned into components or where it is assembled with other parts into a phone. Apple also stated that the titanium used for the iPhone Duo contains at least 80% recycled content. This figure refers to the titanium material and should not be broadly interpreted as 80% recycled materials in the entire phone, nor should it be taken as the percentage of U.S. manufacturing content. The financial figure should also be understood in full context. The announcement is a "commitment of up to $100 million over the next three years": three years is the time frame, "up to" is the maximum amount, and "commitment" points to$RIVER Keep holding, just wait. No rolling positions. Slowly doing it, can only research how to do rolling positions. The big picture is still bearish.
Long-short ratio: Retail investors are frantically chasing the rally.
Binance retail long-short ratio 2.6792, OKX retail 2.56
Whale side: Whale count long-short ratio 2.9432, but whale position long-short ratio is only 1.5405.
Project background: TVL dropped from a peak of 600 million to 160 million, top three addresses control over 53%, manipulated a crash from $87 to $1 in January. This is still a highly controlled "single-player game".
$BTC $ETH #ETF持续流出，BTC反弹能否延续？ 【NEAR Supply Reduction by 36%, Token Holders Finally See a Breakthrough】
The $NEAR community is getting serious: The HSP-029 proposal plans to reduce the annual issuance rate from 2.5% to 1.6% over 24 months, a 36% cut. This reduction targets newly minted tokens, not a direct burn of existing supply. The proposal still requires governance voting and must be adopted by validators through a protocol upgrade to take effect, with voting ending on November 2.
This is indeed good news for NEAR holders. According to the proposal's calculations, the annual new issuance will significantly decrease, easing long-term dilution pressure. However, the cost is tangible: staking yields will drop accordingly, with the proposal estimating the gross staking yield to fall from about 5.27% to 3.37%. Whether smaller validators can continue to cover operational costs will also be a challenge.
NEAR is starting to seriously address token supply issues. However, a supply cut does not equal increased buying pressure, nor does it guarantee a price increase.
The next focus points are whether the proposal passes and if validators support the upgrade. If new supply decreases but ecosystem revenue and real demand do not keep pace, deflation expectations alone will not sustain a long-term price rise.Elon Musk releases major news:
SpaceX and Tesla have established a special committee to begin exploring a potential merger.
He stated at SpaceX's quarterly earnings call that the merger holds significant value, especially in jointly developing superintelligent software. However, no deal terms have been reached yet, and the merger should not be considered a done deal.
SpaceX surged 8% after hours, driven not only by the merger news but also by better-than-expected earnings.
Key changes include: AI has become the largest source of revenue.
Q3 revenue was $14.575 billion, up 161% year-over-year; net profit was $4.404 billion, compared to a loss of $1.054 billion in the same period last year.
AI revenue surpassed Starlink for the first time: Superintelligent business SI revenue was $7.56 billion, up 195% quarter-over-quarter, mainly driven by cloud computing contracts, Grok, and Cursor consolidation.
Computing power continues to expand: nominal computing capacity at quarter-end was 1.7GW, with a target of 2.5–3GW by year-end.
Capital expenditures increased: about $19.6 billion in Q3, with Q4 expected to be $35–40 billion, potentially pressuring profit margins.
The superchip factory Terafab aims to complete its first test chip by May–June 2027, with possible mass production in the second half of 2028.
The merger is still in early stages: both parties have established an independent directors' special committee, and deal terms have not been finalized; Musk is not participating in discussions due to conflicts of interest. "EPS Beats Expectations but Stock Price Falls: Netflix's Earnings Report Loses Out on Next Quarter"
First, an admission of error. Yesterday I said Netflix's expectations were already so low that "as long as it's not bad, the stock can rise," but the market taught a lesson overnight: this quarter not being bad is useless, the guidance for next quarter was cut, and the stock still got slammed.
This quarter's results actually have no obvious flaws: EPS $0.87, expected $0.82; revenue $12.87 billion, in line with expectations; operating margin 34.6%, higher than the company's own guidance of 33.2%; net profit up 44% year-over-year.
The problem lies entirely in the outlook. Q4 EPS guidance is only $0.67, expected $0.73; full-year ad revenue target cut from $3 billion to $2.7 billion, citing slower rollout of ad products; full-year free cash flow also lowered from $12.5 billion to $12 billion. The stock closed around 68.7 on Tuesday, then fell about 4% after hours, approaching the 52-week low near 65.
But don't write it off completely. The Q4 margin guidance of 26% includes a one-time cost of about 120 basis points from layoffs, not a deterioration of the business itself. The company also set a record in Q3 by repurchasing $4.8 billion in shares, with $22.3 billion still authorized for buybacks.
So this drop is a "guidance kill," not a "performance kill." Whether to buy near 65 depends on tonight's official opening volume. If volume picks up and the price recovers above 70 tonight, the panic is likely mostly over; if it continues to drift down, the second curve of advertising growth will have to wait another year.
$xNFLX $NFLX #奈飞Q3盈利超预期，广告业务增长放缓 [How much longer can the yen bears hold out with Japan raising rates every 3 months?]
The Bank of Japan's pace of rate hikes may really be speeding up. Former executive Eiji Maeda expects another rate hike in December, followed by another in March or April next year. The latest Reuters survey also shows that most economists expect the rate to rise to 1.50% in December. This is no longer a sudden hawkish signal; the market is already pricing in this path in advance.
Japan's problem is that a weak yen pushes up energy and import costs, and inflationary pressure forces the central bank to continue tightening. But raising rates too quickly would increase the financing burden on businesses and the government. Maeda also warns that consecutive rate hikes at meetings could lead the market to form more aggressive expectations, and the central bank may not want to take that risk.
The key is whether the yen can truly strengthen, not just how many times the former executive says rates will rise. If rate hike expectations continue to heat up and the US-Japan interest rate gap narrows, yen bears may be forced to cover; but if US rates remain high, Japan's unilateral rate hikes may not reverse the exchange rate. The Bank of Japan meeting on October 29-30 will be the next critical test.🔥Heavy selling pressure in the US spot market❗ ❗
Bitcoin on Coinbase is $115 cheaper than on Binance and OKX.
This price gap is very unusual, especially since the ETF clearly had over $500 million in purchases yesterday.
While the ETF is buying, the US spot market is quietly dumping.
Will this divergence signal ferment further?
❗Market data sharing only, does not constitute any trading advice, contract risk is extremely high
#ETF仍在流入，BTC为何下跌？ $WLD, how to put it, is a typical combination of “high-end narrative + actual operation that really cuts in.” The project launched by Sam Altman’s side claims to “prove you are a real person and fight against AI bots,” but the core method is using Orb to scan people’s irises. Especially in the early stages, it targeted low-income countries heavily, using tokens worth a few or tens of dollars as bait to exchange for immutable biometric data. Privacy controversies have never stopped from the start; regulators in Kenya, Hong Kong, and multiple places in Europe have taken turns intervening, some outright banning it, others demanding data deletion. It’s still under scrutiny now. The token economy is an even bigger flaw. At launch, it played the low circulation and high valuation game to the max, then unlocked tokens continuously like a faucet, causing the price to crash from over ten dollars at its peak to around fifty or sixty cents now, a devastating drop. The chips are highly concentrated; a few wallets can influence the market, so decentralization is basically just lip service. The governance token’s role is weak, with actual control still in the hands of the team and foundation. As for utility, after all the hype about World ID and World Chain, the truly sustainable demand and real-world scenarios supporting the coin price remain thin. It mostly relies on airdrops and narrative to hold up. Institutions occasionally buy a little, but when whales dump, the true nature is exposed. In short, $WLD looks more like a harvesting tool packaged as “the future of humanity”: eyeball data collected, tokens dumped, and the people who take the risk have to deal with it themselves. Those who like grand narratives can keep hyping it, but those putting real money in need to seriously consider how high the risks are BTC Funding Brief｜There is buying pressure in the rebound, but it is not yet time to be fully bullish
Key data points:
• Funding rate is about +0.0052%/8 hours, bulls have a slight advantage, but no extreme overcrowding has appeared.
• 24-hour long-short ratio is about 1.07, bulls about 51.7%, market is slightly bullish but the advantage is not obvious.
• Total network contract open interest is about $53.5 billion, with liquidations of about $58.9 million in the past 24 hours, approximately 82% from shorts, indicating this rally has a clear short-covering component.
• In the past 5 trading days, BTC spot ETFs have had a net inflow of about $211 million, ETF funds continue to provide support.
• Latest net inflow to exchanges is about 5,276 BTC, 7-day average net inflow about 1,889 BTC; whale net inflow to exchanges is about 12,855 BTC, indicating potential short-term selling pressure is increasing.
• Buy orders on the order book are currently thicker than sell orders, but single-layer orders may be withdrawn at any time and cannot be directly regarded as institutional accumulation.
My judgment:
This rebound is mainly driven by ETF support, short covering, and some new longs, but spot exchange inflows and whale transfers still indicate supply pressure above.
Therefore, this looks more like a rebound correction and it cannot yet be confirmed that a new uptrend has begun.
Only a break and hold above $83,400 will offer a chance to test $85,000; if it falls below $81,300 while funding rates and contract open interest remain high, beware of a bull squeeze.More people are entering $ETH than leaving, and even Bitmine continues to lock positions, so short-term selling pressure may ease. Next, it depends on whether the buying side can keep up.SpaceX revenue surged 161%, but capital inflows have become more cautious!
At 4:01 AM today, SpaceX Q3 earnings were released
showing revenue of $14.575 billion, a year-over-year increase of 161%, net profit of $4.404 billion, successfully turning a profit. AI business revenue was $7.56 billion, soaring 830% year-over-year, surpassing Starlink's $5.248 billion. (This report is indeed impressive; AI has become the biggest growth engine.)
But the most interesting thing is that after the earnings release, SPCX did not immediately spike sharply, but gradually rose after some time.
(It seems capital is becoming more cautious now; even with good news in front of them, they are not rushing in.
This does not mean the market is bearish, but that everyone is starting to focus on what’s behind the performance: AI revenue is growing so fast, but can the investment convert into sustained profits? How long can the high growth be maintained? (Good revenue is one thing, but how much profit can be made in the future is another.
So far, SPCX has risen nearly 10%. This time, I’m more concerned not about how much it has risen, but why capital did not immediately enter.
When even an earnings report that beats expectations cannot instantly ignite buying, it shows the market’s reaction to good news may be becoming cautious. Of course, one event cannot represent the entire market; good earnings don’t necessarily mean capital will immediately buy in.
Today, just waiting for $TSLA news. Hopefully, there will be good news for TSLA as well.
#SpaceXQ3营收大增161%，AI业务成增长主力 $SPCX "EPS 0.35 crushes expectations, night session rises over 8%: SpaceX results are out, but the $4.4 billion profit needs to be analyzed separately"
Q3 revenue was $14.575 billion, up 161% year-over-year, while expectations were around $12.7 billion; adjusted EPS was $0.35, versus an expected $0.14, more than doubling; net profit was $4.4 billion, the first profit since listing. The structure has also changed: AI business generated $7.56 billion, up 830% year-over-year, surpassing Starlink's $5.25 billion for the first time. More than half of this company's revenue now comes from selling computing power.
But the real show was the earnings call. Musk said that reaching $1 trillion in revenue by 2028 "is no longer impossible," and personally confirmed that SpaceX and Tesla are discussing a merger, with analysts assigning a 95% probability within two years. Q1 2027 revenue guidance is $33 billion to $37 billion, while expectations were only $22 billion. The stock price rose over 8% in the night session, and $SPCX perpetuals on Hyperliquid surged to 161.7.
Of the $4.4 billion net profit, only $2.8 billion came from operations; about $1 billion was interest and other income, including $390 million in taxes. Starlink added 1.6 million new users in the quarter, still below the expected 2.29 million. Don't forget the looming overhang: about 1.3 billion shares will unlock two days after the earnings report, and another 329 million shares on October 24.
So the real test will be when the US market opens tonight: this 8% night gain has to hold up against this batch of unlocked shares first.
$SPCX $xSPCX
#SpaceXQ3营收大增161%，AI业务成增长主力 🔍"150 milliseconds final confirmation" — is this an achieved result or a milestone to be met with the upgrade? When I see this message, I first focus on the word "target" and hold off on drawing conclusions.
According to CoinDesk, Solana's Alpenglow upgrade plan went live on the mainnet Wednesday, moving validator voting off-chain and then packaging it into certificates, aiming to reduce transaction final confirmation time from about 6 seconds to approximately 150 milliseconds, roughly 40 times faster. This is a target metric, not a tested result confirmed by data.
An interesting detail here: the change is to the final confirmation mechanism and should not be directly interpreted as a 40x increase in transaction throughput. I believe this is more important for applications relying on fast settlement than for short-term sentiment around $SOL. Shorter confirmation wait times can reduce users' uncertainty about settlement delays; however, the change in voting method also requires observation of stability under real network conditions.
Here is my speculation: if the mainnet can stably approach this target after launch, then payment and on-chain transaction scenarios will have more reason to optimize their products around it, rather than just using speed figures for marketing.
I will continue to monitor actual final confirmation latency and whether stability can be maintained under network congestion. Which evidence would you prefer to see first: daily latency data or performance under high load? Let's watch the follow-up together and see whose speculation comes closer.To be honest, I used to see messages like "all-weather, more underlying assets" and immediately think the market was about to take off, focusing first on the opportunity without asking if there was enough liquidity. Later, I realized that being able to trade and being able to execute at the desired price are separated by the entire market depth.
This time, Ripple's broker Ripple Prime has opened 7×24 hour trading for institutional clients with swap contracts linked to over 100 stock, commodity, and currency markets, and they can use the USD stablecoin RLUSD as collateral; these contracts track the perpetual contracts listed on Trade.xyz on Hyperliquid. On the other hand, HSBC believes the market depth is still insufficient.
I think this is progress in trading channels and collateral usage, but in the short term, it looks more like infrastructure expansion and does not mean the related markets are mature yet. All-weather trading allows institutions to manage exposure more flexibly; using RLUSD as collateral also adds another settlement option. But when depth is insufficient, volatility or large orders may make execution more difficult. The most common mistake ordinary people make is mistaking "always tradable" for "always easy to enter and exit."
I now set two rules for myself: when seeing a new product, first look at the actual liquidity and how risk is transmitted; before understanding the contract mechanism, collateral, and execution conditions, don't rush to follow just because the headline is exciting. Have you ever fallen into the trap of mistaking "new features" for "stable opportunities"? If you look at these stock swaps, which piece of information would you confirm first, and would you be willing to share one of your own rules?"Money is back: ETF single-day inflow of 527 million sets a near one-month high, 83,000 trapped positions are being eaten"
Yesterday it was said this rebound lacked money support, but today's data proved otherwise. On Monday, the US BTC spot ETF net inflow was $527 million, the largest single-day inflow in nearly a month. And it was almost all carried by BlackRock IBIT alone: a single-day purchase of $665 million, more than the entire market's net inflow. Fidelity outflowed $83.6 million, Grayscale outflowed $54.9 million, and the rest basically remained unchanged. The money is back, but only through one door.
The hole from the first half of October has basically been filled, with the cumulative net outflow for the whole month down to about $88 million. On the price side, BTC hit a high of 83,439 today, currently around 82,900, chewing through the trapped positions above 83,000. ETF holders' average cost is about 84,358 (Bitfinex standard), and the higher the price goes, the thicker the trapped positions to be unlocked. Whether it can hold above 83,000 at close is much more important than just touching it intraday.
There are also two risks. First, ETH spot ETF had an outflow of $103 million on the same day, marking four consecutive days of outflows. This wave is a targeted buy only for BTC, and risk appetite has not fully returned. Second, the options market is still scrambling for downside protection. Tesla's earnings report tonight—if the US stock market turns sour, the crypto space will hardly remain unaffected.
The "money" hurdle for the rebound has been cleared; the next test is whether 83,000 can hold. If it holds, the rebound may turn into a reversal; if not, it remains a box between 80,000 and 83,500. The lower boundary to watch is 80,800.
#ETF持续流出，BTC反弹能否延续？ $BTC Gold $XAU's recent performance feels a bit conflicted to me 😇
The Middle East situation hasn't shown obvious easing, oil prices are still fluctuating at high levels, so risk aversion demand should be quite strong;
But last night, after gold surged, it fell back and then oscillated repeatedly, never managing to sustain a continuous rise.
The problem might still lie with inflation:
High oil prices mean energy costs continue to transmit to the economy, making it harder for the Federal Reserve to ease monetary policy.
Reuters' latest economist survey even shows the market generally expects the Fed to pause action this month, then continue raising rates...
On one side is the risk aversion demand brought by war, on the other is the pressure of interest rates possibly staying high for a long time.
What's more interesting is that last night the dollar and US Treasury yields fell at one point, yet gold didn't truly strengthen.
I think the biggest problem gold faces now is that the support from geopolitical risk can hardly fully offset the suppression from interest rate expectations anymore.
#10月FOMC临近，美联储暂停加息还有哪些变数？ 12-hour long and short data is laid out here, but I've been staring at the screen for a while, and the more I look, the more something feels off. Some numbers are meant to deceive the eyes; the real knife is hidden beneath these layers.
---
The 52% bullish bias in $BTC is a disguised face
82,795, +0.26%, long positions account for 52.07% — at a glance, "bulls slightly leading," it seems very stable.
But if you look under the derivatives table: Coinbase's $BTC balance has dropped to 845,400 coins, falling into the bottom 2% range of observations over the past year. The US buy-side premium is shrinking, but Binance Futures funding rate has surged to a high of 0.0092. Spot positions are withdrawing, leverage is piling up. This is not "bullish bias," this is bulls borrowing money to prop up the scene.
Bitfinex has made it clear: leveraged longs are rebuilding, $79,821 is a potential forced liquidation trigger point, repeating the structure seen before the two October liquidations. Huobi analysts also poured cold water on the same day — before rate hikes, $BTC is very likely to remain locked between 80,000 and 84,000, and if a direction must be chosen, the probability of testing 80,000 downward is higher.
So this 52% "bulls leading" actually reads the opposite to me: bulls haven't lost, but they're already gasping. 52% is not confidence, it's the last breath holding on.
$ETH's 50-50 split is the harshest meat grinder in the room
Longs 5.04 billion, shorts 5.01 billion, ratio 50.12% to 49.88%. The difference of 30 million is negligible at Ethereum's scale.
But this "no difference" is the deadliest.
Binance's 8 largest Ethereum liquidation pools are all hanging above the current price. The nearest is at 2422, only 0.03% above current price; the largest is at 2447, just over 1% higher. In plain terms: the leverage structure has already laid the fuel, just waiting for a needle to poke upward, and shorts will be queued to pay up.
But on the other side? ETF outflows in October are approaching 1 billion, whales are adding short positions on Hyperliquid, daily moving averages remain bearish. Funding rates are still positive, longs are still paying to hold positions, but Binance's active selling accounts for 55%, and the "long" at the account level is not being supported by active buying.
On paper it's a stalemate, but at the core both sides have knives at each other's throats; whoever moves first will bleed first. $ETH is not "long-short balanced" now, it's "longs and shorts holding each other hostage." The longer this structure lasts sideways, the louder the break will be.
$SOL is the most ironic: your data says shorts dominate, but the market just buried the shorts
In your 12-hour snapshot, $SOL shorts account for 53.68%, price 105.79, down 0.29%. By this number, shorts are indeed in the lead.
But just before and after this data was captured, $SOL surged from 103.82 to 108.22 in one go, with $6.55 million in short liquidations within an hour, and $7.47 million in $SOL short liquidations across the network in 24 hours, more than twice the longs. Shorts have a high share? Yes, on paper the share is high, but the market just told shorts with a bullish candle: high share doesn't mean you win, it just means you are many, and where there are many, liquidations are more satisfying.
Alpenglow upgrade is counting down, Pando SOL ETF is landing on the Hong Kong Stock Exchange, these two events combined make a short squeeze almost inevitable. The "shorts dominate" you see now is likely an expired snapshot — at the moment the data was captured shorts were indeed more, but the market has no obligation to wait for you to finish screenshotting before moving.
What really weighs overhead, none of these three coins have priced in
$BTC has been grinding narrowly between 82,000 and 84,000 for two weeks; Wintermute's exact words: 82,500 is the test, a decisive break below points to just over 70,000. $ETH's stalemate and $SOL's short squeeze are all bouncing inside $BTC's cage. And the key to the cage is not in the candlesticks, but in the FOMC on October 28.
US Treasury yields are still hanging above 5%, oil prices hovering near 110, CPI climbing from 3.4% to 3.7%. ETF single-day inflow of 526 million did give a breath, but compared to interest rate expectations, that money can't move the whole market.
So the divergence among these three coins now is essentially one thing: funds are doing final position adjustments before the FOMC. $BTC bulls are holding hard, $ETH longs and shorts are locked, $SOL shorts are being cleared. This is not a trend, it's a clearing of the table before a turning point.
To be blunt
52% bullish bias, 50% stalemate, 53% shorts leading — these numbers look good, like three faces with clear attitudes. But derivatives data never tells you "who is right," it only tells you "who hasn't run yet."
$BTC leveraged longs haven't run, $ETH longs and shorts can't escape, $SOL shorts have just been carried away. When the table flips on the 28th, whether your "share advantage" can withstand a needle is the real truth.
---
Risk reminder: The above is data analysis and personal interpretation, not trading advice. With the FOMC approaching, the probability of spikes and traps is rising, don't get carried away with leverage.
$BTC $ETH $SOL #Arc主网上线首日数据出炉 #就业数据密集公布，沃什政策立场受检验 #ETF持续流出，BTC反弹能否延续？ The 30-year US Treasury yield is 5.48%, a level not seen since 2002. For the stock market, a bond yield above 5% makes stocks less attractive; for the crypto space, holding BTC yields no interest, while holding US Treasuries yields 5.48%, so capital naturally flows to bonds. The Fed minutes themselves admit that "long-term yields are rising faster than the FOMC expected"—meaning the market is tightening on its own, possibly more aggressively than Fed rate hikes. In this environment, don't expect the crypto market and US stocks to rally independently; they are facing the same adversary.The 12-hour liquidation map has been fully revealed: $ZEC short positions were completely wiped out, $BTC shorts were heavily covered, while $ETH longs were falling one after another. The same market, three different scripts, with leveraged funds being "precisely harvested" by coin.
$ZEC: Shorts were carried out tonight.
Out of the total 6.76 million in liquidations, shorts accounted for 6.21 million, longs only 550,000, a ratio close to 11:1. This was not ordinary stop-loss, but a short squeeze. Previously, $ZEC dropped from 1690 to 1048, shorts developed a muscle memory of "adding shorts on every rebound," stacking positions thicker and thicker — short positions once reached 173 million U, with an 82% profit ratio. When the NU7 upgrade evaluation day coincided with an oversold rebound, buying surged in, and shorts didn’t even have a chance to cover, they were swept away in one wave. On the chart, $490,000 worth of shorts were liquidated in one hour; the main force used the positive news to execute a textbook short squeeze.
Shorts lost not to fundamentals, but to overcrowding.
$BTC: Shorts were lifted by the combined forces of US stocks and oil prices.
Out of 3.58 million total liquidations, shorts were 2.87 million, longs only 710,000, a very clear pattern of short liquidation. The driver was not internal to crypto — US stocks beat expectations in Q3 earnings, Dow rose 1.2%, oil price fell from above 100 to around 96, risk appetite warmed up, and $BTC was directly bought from 8.13 to 8.34, forming a bullish engulfing pattern, forcing shorts to cover. Coinglass data shows that in just one hour, $125 million worth of $BTC shorts were liquidated, open interest rose to $53.8 billion, with short covering and new longs pushing the price up.
But the 8.41 area is previous M-top resistance; after shorts are cleared, those chasing longs will face the same resistance.
$ETH: The most painful one.
Total liquidations of 6.27 million, longs 3.62 million, shorts 2.65 million, with longs liquidated nearly 1 million more than shorts. $ETH’s problem is the exact opposite of $ZEC and $BTC: it’s not that shorts are overcrowded, but longs are heavily leveraged. Previously, a whale opened a $187 million long at max leverage, with 4,890 $ETH liquidated at $2,356, position value $11.74 million, hitting exactly the early morning low. Coinglass’s liquidation map is even clearer: below 2,333 there is $1.023 billion long liquidation pressure, while above 2,572 short liquidation pressure is only $553 million.
This means $ETH’s long leverage is far more crowded than shorts. Although the short-term 4-hour bounced from 2,394 to 2,451, within the same hour shorts liquidated 1.06 million, longs only 70,000, so the rebound itself is creating new wounds for shorts. Both longs and shorts are getting hit; $ETH became the most typical "meat grinder" in these 12 hours.
The macro backdrop hasn’t changed; what changed is which side the funds choose to tear open.
The 10-year US Treasury yield remains above 5%, the FOMC meeting is next week, and overall valuation pressure on risk assets hasn’t eased. But funds haven’t fully retreated; they are making choices: $BTC and $ZEC shorts are more crowded, so they hit shorts first; $ETH longs are more leveraged, so they sweep longs down. The same macro environment, different position structures, determine completely opposite liquidation directions.
$ZEC shorts being cleared doesn’t mean $ZEC will reverse trend; $BTC short covering doesn’t mean 8.4 will be easily surpassed; $ETH long liquidations mean leverage below is lightening, but only if the dense liquidation zone at 2,333 isn’t truly broken.
The harshest part of this market is: it doesn’t kill direction, it kills overcrowding.
Which direction’s position do you hold now? $ZEC short, $BTC long, or $ETH long? Comment your coin and direction, and I’ll help you see which side the overcrowding stands on.
#ETF持续流出，BTC反弹能否延续？ #SpaceXQ3营收大增161%，AI业务成增长主力 #奈飞Q3盈利超预期，广告业务增长放缓