Orbit Post Sitemap
Supply side eases VS demand side delivers answers, I think this comparison is more useful than just saying "exchange outflows are bullish." CryptoQuant analyst CoinNiel pointed out that as of October 27 (UTC), Bitcoin exchanges shifted from a net inflow of 7,278 coins the previous day to a net outflow of 12,702 coins, but the evidence still supports a neutral stance.
I prefer to see it as an improvement on the supply side, not a confirmation on the demand side. Although both 7-day and 30-day periods show net outflows, there were 4 days of net inflows in the past 7 days. A one-day reversal is quite eye-catching, but looking over a longer period, the direction is still fluctuating and cannot be directly interpreted as sustained reluctance to sell.
On the demand side, the funding rate is +0.00762%, higher than the 7-day and 30-day averages; open interest contracts increased by only 0.56% over 7 days. My understanding is that bulls have a tendency but have not yet taken obvious expanded positions to support it; sentiment and incremental volume cannot be equated.
Looking at long-term holders, LTH-SOPR dropped from 1.1657 to 1.0124, which only indicates that profit margins when selling have narrowed, not that less is being sold. I also tend to imagine "earning less" as "not selling," but this step must be resisted.
So my judgment is: supply improvement is meaningful, but whether demand can take over is the next question. If you had to prioritize waiting for one signal, which side would you stand on: the supply evidence of continuous net outflows, or the demand evidence of sustained spot buying?Putting the exchange fee on Pyth, I wanted to see the exception handling branch as soon as I read this. 💻 The input for this review is: Yield Basis disclosure fee bypass path, proposed WETH fix plan, intended to add deposit debt ratio check, with the related vote scheduled to end on October 29.
My judgment is that whether the fee rules are reliable is more important than adding another charging logic. Hooking to Pyth means the fee calculation introduces an external dependency. If the price updates are not timely, whether the charging can remain consistent is the potential bug I worry about; this is my speculation, as the material did not explain the exception handling details.
If I were to make changes, I would first add handling rules for price exceptions, then test whether normal exchanges and boundary operations follow the same billing logic. I don't want to consider acceptance passed just by running the normal path; next, I will watch whether the plan clearly writes these boundary conditions.This meeting might end up worse than predicted.
It's the same familiar pattern and formula; just a different timing, and people think this time is different, but their hands don't listen and rush in again.
It pulled up to 81800, and those who can't control their hands probably rushed in to buy more, thinking there would be a big rebound. But the result was a pull-up followed by a shakeout. Many believe this is a shakeout, and indeed it is, but how long will it last? How far down will it shake out? You can't guess. The big trend has already reversed, and the hope for going north is very slim. This time, they didn't even dare to release the dot plot; it feels like a big game is being played, and the outcome might be even worse.
News can only play a supporting role. Why is it still going south if there's no rate hike this time? Shouldn't it be going north? Many are confused. Let me tell you clearly: the big trend of the market has changed, and the significance of the news is no longer great; it’s more like a smokescreen to assist capital outflows. The results are out, but don't mistake the rate hold for a positive signal; this is a hawkish trap. Two rate hike votes have directly revealed the bottom cards, yet the market is still self-deceiving, thinking that no rise means dovishness. I'm here to make it clear: Cook and Harker voted for a rate hike, indicating the Fed is close to splitting internally over how aggressively to fight inflation. Waller gave no forward guidance, plainly telling you everything depends on the data. Now with such a big split on December hikes, and Q3 GDP and PCE about to drop, on-exchange funds lack the confidence to hold firm; in this zero-sum game, whoever catches the flying knife dies.
Look at $BTC, the big coin; a slight dip means funds are voting with their feet. ETFs are all short-term players, quick in and out, no long-term liquidity. The overhead supply is heavy; until the macro thunder lands, only sideways movement is possible. Hold long-term spot firmly; don't chase highs in the short term. If you short for the short term, keep your stop loss tight; don't stubbornly hold and lose money without a stop loss.
Look at $ETH, the softest follower. The ecosystem has no breakout points, and funds lack support. When the market weakens, it falls fastest. Don't mistake sideways trading for stabilization; when BTC tests support, ETH's drop will be amplified. Control your hands; don't catch flying knives.
Look at $XAUT, gold, the most stable hard asset. High yields combined with rate hike divergence are driving safe-haven funds here, with central bank buying providing a floor. Once economic data bombs, gold will run independently. Don't let short-term sentiment trap your long-term logic.
Don't listen to misleading news; no dovish turn means a slow, painful cut. Tonight's GDP and PCE are the real tests. Control your hands; don't bet on a one-sided move. Surviving this meat grinder and preserving your principal is more important than anything. #美联储维持利率不变，12月加息分歧升温 Google Cloud up 82%, this money has little to do with the crypto circle. Tracing the 82% figure down, the realization is in the cloud and chip chain—Meta just raised its capital expenditure guidance, SK Hynix's quarterly revenue exceeded 100 trillion Korean won, the money is circulating between computing power and storage, and for now, none of it flows into token pricing.
1. Revenue has indeed picked up, but the giants spend money first and earn later; capital expenditure is still increasing, and returns will take a few quarters to show up in the profit statement. 2. On the crypto side, connecting to the AI narrative, what we see so far is leverage: that 25x $ETH long position from Maji was liquidated, leaving the account with only $5,357. No new incremental funds are seen; all added is position size. 3. On the spot side, as discussed before, the ETF direction flips daily, with a net inflow of $210 million on Friday and a net outflow of $130 million on Monday; buying is not decisive.
My conditions for $ETH here are: ETF net inflow for two consecutive days and price holding above the 60-day moving average, only then will I acknowledge real support on this chain; if it breaks the previous low and funding rates turn negative, consider AI money has bypassed the crypto circle.
#谷歌云业务增长82%，AI投入回报如何？ #$ZEC Finally waited for you, luckily I didn't give up.
I am firmly bearish, expecting it to go below 1000. The NU7 upgrade benefits have already been realized in advance. October's trading volume is significantly lower than September's peak, and the upward momentum is insufficient. It hasn't been able to rise above 1300 for a while.
The EU AMLR 2027 will restrict privacy coins from listing, Grayscale ETF approval is uncertain, and regulatory pressure continues. The DeFi ecosystem has less than $2 million locked, with weak real-world application scenarios. Adding these up, it’s impossible to stay above 1000 $BTC $ETH The October FOMC is about to be announced.
The market expects the rate to remain at 4.00%, and the announced value is very likely to be 4%. The real variable lies in Powell's post-meeting speech, not the rate number itself.
Scenario 1: Maintain 4%, Powell leans dovish — no rate hike, and signals room for future rate cuts.
Impact: The dollar and US Treasury yields weaken, BTC and ETH may experience short-term spikes, but beware of a pullback after the surge.
Scenario 2: Maintain 4%, Powell leans hawkish — no rate hike, but emphasizes persistent inflation, delaying rate cuts, with none considered this year.
Impact: Most likely to trigger a "good news sell-off," expectations dashed, capital outflows, BTC and ETH fall in sync with Hynix, sharp spikes that easily trigger stop-losses.
Scenario 3: Rate hike to 4.25%.
The market has almost not priced this in; if it happens, it will directly hit risk assets, with the crypto sector and storage stocks under collective pressure, hard to withstand in the short term.
This time it's a double impact of the decision and the press conference: the decision comes first, followed by the speech. The two phases often move in opposite directions; a rally followed by a plunge is not uncommon. Manage your positions and set stop-losses properly.
#10月FOMC利率决议即将公布 Two documents placed side by side have a very special flavor.
A Japanese Bitcoin industry company announced a capital increase of 840 million yen, about 5.07 million USD, directed to Fulgur and ALC, with the agreement specifying full delivery in physical Bitcoin. The shares exchanged are not for money, but for coins.
The traditional approach is to issue shares for fiat currency, then buy coins through exchanges, incurring underwriting fees and spot price slippage. This time, all three steps are cut out; Bitcoin is transferred directly from the investor's cold wallet to the project's routing channel.
Where does the confidence come from? This company is building enterprise-level Lightning Network payment infrastructure, with a product called aurora. For the Lightning Network to operate, it inherently requires a large amount of spot coins to serve as routing liquidity reserves. Investors have idle coins, the project side urgently needs coins, so they exchange one-to-one, welding assets and business together.
On the other side is Robinhood's Q3 report. Total transaction revenue rose 19%, options and event contracts were very impressive, but crypto revenue fell 55% year-over-year, down to only 120 million USD.
With coin prices hanging high and intraday volatility suppressed very low, retail investors feel making quick money is no longer interesting and turn to options and prediction markets for higher leverage. Retail investors don't distrust, they just find it too slow.
Thus, the control has changed hands. Robinhood's crypto trading volume shrinks, highlighting that Strategy, Strive, BitMine buy coins in hundreds of millions of dollars. Supporting liquidity are dozens of corporate treasuries, ETFs, and sovereign funds.
Therefore, $BTC pricing is driven by balance sheets, not sentiment.This morning, the biggest gainer in CoinGecko's hot search was Initia (INIT). As of 10:00 AM Beijing time on 10/29, INIT was priced at $0.222, up 34.7% in 24h, 73% in 7 days, with a 24h trading volume of about $86.5 million, nearly double the market cap of $46.8 million.
News: On 10/23, Initia was exploited due to a MoveVM vulnerability, resulting in about $2.7 million in assets being cross-chain transferred; the mainnet and rollup were all suspended. On 10/28, the foundation announced funding to compensate affected iUSD collateral at a 1:1 ratio (about 1.08 million USDC, 105 ETH, 1.4 million AUSD), and the fix plan is under audit.
My view: This is a typical "bad news priced in + foundation backstop" rebound, with a rise of over 70% from the post-hack low. However, the chain has not restarted yet, with no specific date, and circulating supply accounts for only about 21% of total supply. Turnover exceeding market cap indicates chips are quite loose. The real catalysts ahead are the audit results and restart timing; a smooth restart will restore confidence. Until then, it’s mostly an emotional game.
$INIT Wow, this trade really burns the heart. I was scared when I opened my eyes this morning and saw a loss of 200u. I didn't check the order while squeezing into the subway just now, but when I got to the company and opened my phone, the $BTC short position was actually profitable.
It seems that inflation in 🇺🇸 is indeed very serious. Actually, I study macroeconomics, and to some extent, I believe inflation is related to its domestic bubble. When they say their economy is doing well, it's both true and false. If the output value is insufficient to support economic development, it will actually keep increasing the living costs for ordinary people.
I think 🇺🇸 is afraid of a repeat of the previous internet bubble or the 2008 financial crisis, so this time they are determined to maintain a strong stance to kill valuations and bubbles.
I also shared my views at 1 a.m., which are pretty much what I thought: not raising interest rates isn't necessarily a good thing! Even though they didn't raise rates, the overall Fed stance is still hawkish. I said before that not raising rates isn't necessarily good because rate hikes are inevitable. Not raising rates only pushes the market's rate hike expectations further out. So crypto $BTC will just keep waiting.
This feels like a prisoner's dilemma and a paradox. Both raising and not raising rates are unfriendly to crypto $BTC; the difference is just short-term pain versus long-term pain. Raising rates is short-term pain, not raising rates is long-term pain—it's torture!
#美联储维持利率不变，12月加息分歧升温 The correlation chart of Bitcoin and gold on the 4-hour timeframe sends a clear warning signal to buyers. These two assets, traditionally seen as stores of value, have both faced significant selling pressure since the end of September, moving almost in sync. $BTC $XAU
The core issue lies in interest rates. Both gold and Bitcoin are non-yielding assets, so when expectations of monetary tightening rise, large amounts of capital immediately withdraw to seek safer bond yields.
The market reflects extremely heavy pressure on the macroeconomic front:
Since the end of September, both Bitcoin and gold prices have dropped by about 7%.
- Gold performance in the comparison chart: -10.70%
- Bitcoin performance in the comparison chart: -11.56%
- Market probability of a rate hike in December: 72%
If bulls insist on bottom-fishing too early, they currently face very high risks. When the probability of a rate hike reaches such a high level, market liquidity will continue to dry up. The only bright spot is that there is now an opportunity to accumulate stocks for the next cycle at reasonable prices, but this is definitely not the time to use leverage. $DOOD The top ten addresses hold 91.26% of the total tokens in the entire network, with the largest wallet alone holding nearly half.
This extremely monopolized single-player market is pumped up by trending on hot searches, purely to create liquidity for the whales to unload their positions.
In just a few days, it fell back to the initial rise zone along the same path. The occasional bottom-fishing buy orders appearing on the intraday chart look like support, but in reality, the big holders casually dumping some tokens is a stampede-style crash competition.
The current market is like a dull knife cutting flesh; those occasional pulse rebounds are just the trapped positions that haven't died out yet gasping their last breath.
#DOOD
$DOOD
👇Google, as a major AI investment giant, has earnings below expectations and capital expenditures have slowed down, which is very likely a landmark event signaling the burst of this AI bubble. Previously, the market thought the Ant IPO would trigger the burst, but when the market reaches consensus, perhaps that day has come earlier. $BTC $ANTHROPIC $GOOGL Why can BTC reach tens of thousands of US dollars? Many people never fully understand it in their lifetime
Many people just can't figure it out: Bitcoin is just a string of code, invisible and intangible, so why can it rise to tens of thousands of US dollars per coin? Does it really have value?
Before explaining, brothers need to understand a basic fact: all fiat currencies share a common flaw—they can be continuously issued, constantly diluting the purchasing power of ordinary people's assets.
But Bitcoin's code locks its total supply, forever capped at 21 million coins, with production halving every four years, giving it a unique digital scarcity.
At the same time, it is decentralized, with nodes distributed worldwide, its ledger public and immutable, and no one can arbitrarily freeze or confiscate it.
Because of this, from early retail participation, it has gradually become an alternative reserve asset recognized by institutions and multiple countries.
Many people think BTC is expensive because they still view it as a toy.
The essence is not that Bitcoin is too expensive, but that the purchasing power of continuously issued fiat currency is steadily declining.
Its core value lies in scarcity and inflation resistance, providing an asset choice not controlled by any single entity.
Those who don't understand the fundamentals will only be shaken out by volatility; those who grasp the logic can hold onto the trend The Federal Reserve's October interest rate decision has been announced, with the benchmark rate held steady at 3.75%—4.00%. Although it seems like the "boot has dropped," cracks are actually hidden beneath the surface. The vote was 10 to 2, with two members advocating for further rate hikes, making internal divisions public, and the statement reiterated that inflation remains elevated. After the decision was released, $BTC briefly surged above $81,700 but then gave back some gains, failing to hold at higher levels; $ETH similarly rebounded moderately to around $2,391, also showing a pattern of rising then falling back.
The "hold steady" stance did not bring substantive easing. Short-term U.S. Treasury yields plunged, while long-term yields continued to rise, with the yield curve sending conflicting signals and uncertainty about the interest rate path increasing rather than decreasing. The crypto market's rise and fall precisely indicate that the rebound was more about emotional recovery after short-term uncertainty was removed, rather than a reversal in liquidity expectations. Although Bitcoin spot ETFs saw inflows, leveraged funds did not expand in tandem, and medium- to long-term options remain bearish. Against the backdrop of persistent inflationary pressure and the possibility of further rate hikes not being ruled out, crypto assets' sensitivity to interest rate expectations will only grow. After the rise and fall, the real test may just be beginning.
#美联储维持利率不变，12月加息分歧升温 Last night the Fed did not raise interest rates, but it also did not provide a truly dovish signal for Bitcoin.
The rate was maintained at 3.75%–4.00%, yet two officials explicitly called for a 25 basis point hike.
The yield on the US 10-year Treasury rose back to 5.05%, further increasing market expectations for a rate hike in December.
Meanwhile, BTC ETFs confirmed net inflows for two consecutive trading days, indicating institutional funds are trying to resume allocation.
This forms the most important current market contradiction: marginal improvement in institutional funds, but the macro financial environment remains tight.
If tonight's PCE inflation is below expectations, the 10Y yield falls back, and ETF inflows continue, then BTC has a chance to retest 83K–85K.
Conversely, if inflation continues to exceed expectations, the probability of a December rate hike rises further, and ETF buying cannot be sustained, BTC may again test 80K.
What really deserves attention today is not that the Fed has paused rate hikes, but whether the latest US inflation data can change the market's judgment on the December policy path.
Do you think the Fed will raise rates by another 25 basis points in December, or continue to pause? #微软Azure增长43%，AI商业化能否提速？ Didn't stay up late to watch the FOMC and Waller's press conference. After waking up, I quickly checked various indicators:
1. The 2-year/5-year US Treasury yields dropped significantly after Waller's speech.
2. The December CME rate hike probability is 78.4%, not much different from before the meeting, which is somewhat disappointing.
3. Oil prices had a bearish candle at 3:30, which might be the main reason for the late rally in the US stock market.
4. Microsoft rose due to after-hours earnings, while Google plunged 10%.
5. Waller's speech itself didn't have much new content, still the usual talk about reducing guidance and information disclosure. However, the institutional move starting next year to gradually reduce the number of rate votes in the eight FOMC meetings is quite important. I don't see this as a strong signal; it only happens when there are fewer cards left to play, pushing responsibility to the market to reduce their own embarrassment. In a strong power-up cycle, those with more influence wouldn't try to hide themselves. This move to conceal weakness is the real reason.
6. There were still two dissenting votes in this rate decision—Cook and Harker. This is the Fed's latest way in nearly half a year, after scrapping forward guidance and dot plots, to express attitude and break consensus. The message this time is still relatively hawkish.
7. The best thing said in this press conference was Waller mentioning that amid global turmoil, inflation expectations are moving in the right direction.
Overall, this press conference was neither impressive nor disappointing; the market's rise or fall will mainly depend on corporate earnings.$SKHYNIX opened high and fell low, made a small profit on a short position this morning
Those who opened high didn't have a good outcome $80,389 这个数字跳出来的时候，我正盯着清算热图发呆。 80K 这条线，今天到底是在护盘，还是在给追多的人挖坑？ 盘面其实很诚实。BTC 夜里摸到 79,506 又爬回 80,389，日内微涨 0.17%，看着像没事人，但 24 小时爆掉 1.25 亿美元，54% 是多头。这个比例说明杠杆多头在被一点点挤出去，不是踩踏，更像钝刀子慢慢放血。ETF 那边同日净流出 1.1 亿，市值挂在 1.61 万亿附近，价格没崩，钱却在悄悄撤。 我自己的仓位昨天就减了一档，不是看空，是这种 Fed 前夜最容易被假动作骗。今天 Fed 开会、周三 28 号出结果，市场给 90% 概率按兵不动，可油价 102、股市走弱，风险偏好本来就在收着。真正被提前计价的不是"不加息"，而是"不加息但语气偏鹰"——这个预期差还没人认真交易。 偏多的路径很清楚：80K 守住，收盘站上 80,844，空头回补能把价格推到 82K，衍生品结构会从防守转成进攻，山寨才敢跟。但我更在意另一面：如果 80K 只是靠几笔大单撑住，资金费率没真正转负，那反弹就是给空头送流动性。第二层传导是，BTC 在这里磨得越久，ETH 和山寨October 29: 💥💥💥💥💥 Macro analysis, BTC market analysis!
1. The Federal Reserve maintained the interest rate at 3.75%—4.00% with a 10 to 2 vote. Harker and Cook advocated a 25 basis point rate hike, stating the economy is strongly expanding and inflation remains high. Waller did not provide a December roadmap; according to Bloomberg, bond traders are betting about a 90% chance of a rate hike in December, and pausing rate hikes does not mean a shift to easing.
2. Tonight continues to bring macro data tests. At 20:30 Beijing time, the US Q3 GDP preliminary value and September PCE will be released, and at 21:15 the European Central Bank interest rate decision will be announced. If growth is strong and inflation exceeds expectations, further rate hike expectations may intensify, posing volatility risks for BTC and high-valuation tech stocks.
[Bitcoin Market Analysis]
BTC was around $81,191 at 07:01, up 0.38% in 24 hours, with a 24-hour high of $81,823.
The conditional observation range below is $80,400—$80,800; a break below $79,800 on the 4-hour level sets the risk condition, neither of which can be regarded as confirmed technical support. Further down, $76,500 corresponds to the average cost of short-term holders cited in the material, and whether it can form support remains to be seen.
On the upside, first focus on whether $81,823 can be broken through and confirmed by spot trading; the larger resistance reference is $84,000—$85,000, a range that has blocked upward movement four times this year. 2026.10.29
The news has finally landed. Although the FOMC has kept interest rates unchanged, the probability of a rate hike in December has risen to 86%. Two officials voted against the rate hike, indicating that the macro hawkish tail risk has not been eliminated. BTC implied volatility has dropped to an eight-month low of 36%, but the 30-day option skew remains at 14%, with put options continuously trading at a premium. The market is pricing in downside risk with real money. The Fear and Greed Index is at 50, neutral, providing no extreme contrarian signals. However, yesterday's rebound still failed to break through, and momentum is extremely weak. The real environment may still need to find a bottom. Yesterday's resonant rebound was insufficient, and most cycles are at the top today, with only the daily chart in a rebound cycle. Currently, it is the end of the month, and rebound space is limited.
BTC: Shorting on rallies is preferred over buying dips. The leverage structure is unfavorable to bulls. Long liquidations are about $196 million, while shorts are only $19.47 million. The 10:1 tilt indicates that bulls using high leverage to bet on a rebound near 80000 have been largely cleared out. Coupled with most cycles being at highs today, there is a risk of resonant decline. But also note the fuel above: there is a BTC short position of about $31.5 million near 81683, less than 1% away from the liquidation price. FOMC sentiment is relatively warm, and short squeezes may extend to 82500–83000. There is support below but it is also fragile; once broken with volume, it will accelerate downward.
Support: 79000-79500, 78500, 77000
Resistance: 81500-82000, 82500-83000
ETH: Prioritize shorting at highs, assist with buying at lows. 2362 is the 100-day EMA, which acted as support earlier this month but has turned into resistance after being broken. Yesterday it was tested several times but fell back. Currently, except for the daily chart, other lower cycles are at high or near-high levels. Continuing to push up risks forming a high-level resonance. 2310-2320 is also a support node and is relatively close. This is a dense convergence zone of indicators: on-chain realized price at 2310, 200-day EMA at 2315, and around 2310 is the average cost line of on-chain holders—holding this means holders tend to be reluctant sellers; a valid break below may trigger stop-loss releases near breakeven, possibly causing accelerated downside.
US spot ETFs had a cumulative net outflow of $1.26 billion in October, close to the historical monthly outflow record, about six times the net outflow scale of BTC ETFs in the same period. The liquidation zone below is concentrated at 2270–2290, with the next layer of long liquidation pools at 2220–2250. The short liquidation zones above are at 2390–2400 and 2430–2440, but both are above immediate resistance. Without breaking resistance, liquidation is out of the question.
Support: 2310, 2270-2290, 2220-2250
Resistance: 2365, 2390-2400, 2430-2448$AMZN Massive CapEx vs. FCF: Amazon's 2026 capital spending plan has skyrocketed to $220 billion (up from $200 billion), which pushed its trailing-12-month free cash flow (FCF) into a negative -$7.6 billion outflow. Any stabilization here will surprise the market.
• The Prime Day Shift Impact: Because Prime Day moved entirely to Q2 this year, Amazon's Q3 revenue growth forecast (9% - 12%) looks artificially slower; without this shift, growth would be 400 basis points higher.
#AmazonQ3AWSOutlook 🔍$757 million in assets, all new money? Don't rush to conclusions. SoSoValue data shows that on October 28, Eastern US time, the US $ZEC spot ETF had a net inflow of $2.0735 million, with a cumulative net inflow of $211 million.
I focus more on subscriptions rather than scale: net asset value is also affected by coin price. Here's my speculation: if you only look at asset size, you might overestimate new demand.
Next, keep an eye on scale changes: do you think the follow-up will mainly rely on net subscriptions or coin price movements? Let's see who guesses closer.What came to my mind after reading this is: "Visible" VS "Catchable"—they are fundamentally two different things. I usually like to check on-chain records, but transparency of information does not mean the market can easily digest it.
According to Ember monitoring, the Hyperliquid development team HyperLabs applied to redeem 179,706 $HYPE tokens, valued at 14.49 million USD, expected to arrive on the evening of November 4th. The report mentioned that previously the team’s unlocking process involved exiting staking, distributing to members, and then conducting OTC trades through Flowdesk.
My judgment is that this time more attention should be paid to the quality of the takeover rather than the disclosure itself. The path being public can reduce suspicion but does not explain the transaction conditions; if it still goes through OTC, the direct pressure on the order book might be smaller, but if the acquirer then sells on the market, the pressure is just postponed—this is my speculation. When facing such team unlocks, do you value transparency of the path more, or the OTC takeover conditions?#谷歌云业务增长82%，AI投入回报如何？ The earnings report after the market close on October 28th presented a very clear signal: the market no longer prices a unified “AI narrative” uniformly, but instead uses free cash flow (FCF) as a yardstick to individually screen “whose capital expenditures are being covered by whose operating cash flow.” Microsoft and Alphabet, with two completely opposite cash flow data sets, tore the after-hours trend into two directions.
📊 After-hours core earnings snapshot
Below are the earnings and market reactions of major companies after the market close on October 28th (and some pre-market references):
#谷歌云业务增长82%，AI投入回报如何？
· Microsoft ($MSFT): Revenue of $92.28 billion, EPS $4.89, both exceeding expectations; Azure revenue grew 48%, FCF reached $18.6 billion (expected only $8.54 billion). After-hours rose nearly 6% at one point, finally closing up about 4%-5%.
· Alphabet ($GOOGL): Revenue $127.6 billion (slightly above expectations), but EPS only $2.12 (far below the expected $3.02); capital expenditures surged to $51.355 billion, FCF recorded -$9.065 billion (negative for the second consecutive quarter). After-hours plunged over 8%-10%.
· ServiceNow (NOW): Adjusted EPS $1.00 (below the expected $1.02), revenue $4.11 billion; Q4 revenue guidance $425-$426 million, below the expected $436 million. After-hours dropped 16%.
· KLA Corp (KLAC): FQ1 EPS $1.22, revenue $4.14 billion, both exceeding expectations; FQ2 revenue guidance $4.7 billion (well above the expected $4.487 billion), EPS $1.44 (above expectations). After-hours rose 3%.
· Chipotle (CMG): Revenue $3.30 billion (above expectations), EPS $0.27 (slightly below expectations), reaffirmed full-year guidance. After-hours rose 2%.
· Carvana (CVNA): Revenue $8.04 billion (well above the expected $7.67 billion), EPS $0.39 (below expectations), but strong revenue overshadowed profit flaws. After-hours rose 6%.
· Teladoc (TDOC): Adjusted EPS loss $0.15 (better than expected loss of $0.22), revenue $605.1 million (above expectations). After-hours rose 8%.
· Seagate Technology ($STX): FQ1 revenue $4.23 billion (up 61% YoY), EPS $7.68, both exceeding expectations; FQ2 guidance also significantly above expectations. But after-hours fell over 8% at one point, finally closing down about 5.5%, a typical "good news sell-off."
· NXP Semiconductors (NXPI): Q3 revenue $3.79 billion (above expectations), but Q4 revenue guidance midpoint $3.85 billion (below the expected $3.87 billion), adjusted operating profit guidance also below expectations. After-hours plunged nearly 7%-9%.
---
🔍 Core divergence: Microsoft's “cash flow closed loop” vs. Alphabet's “cash flow gap”
Microsoft: AI capital expenditure provides positive, verifiable closed-loop evidence for the first time.
Microsoft Azure revenue grew 48% to $33.17 billion, capital expenditure $45.2 billion, even below the previous guidance of $50 billion. The most critical signal is free cash flow of $18.6 billion, far exceeding market expectations of $8.54 billion. Microsoft is the only hyperscale cloud provider expected to maintain positive cash flow next year. This means the chain of “AI capital expenditure → revenue → cash flow” has formed a verifiable positive cycle in Microsoft.
Alphabet: The same tool becomes a negation tool.
Alphabet's capital expenditure increased about 114% YoY to $51.355 billion, but operating cash flow dropped from $48.414 billion last year to $42.29 billion. Free cash flow recorded -$9.065 billion, negative for the second consecutive quarter, worsening from last quarter's -$5.855 billion, nearly $6.8 billion more outflow than analysts expected (-$2.27 billion). An $8.7 billion equity investment loss further dragged down net profit, causing EPS to fall far short of expectations.
Together, these two earnings reports form a complete signal: the market no longer prices based on “whether AI demand exists,” but screens individually based on “whose capital expenditure is covered by whose operating cash flow.” Microsoft proves cash flow can be a confirmation tool for narrative repair, while Alphabet proves the same tool can become a negation tool under another name.
💎 The “good news sell-off” trap in storage and semiconductors
Seagate's earnings and guidance fully exceeded expectations, but still fell after hours. The core reason is that the previous price increase has fully priced in the AI storage boom, so when “good data” lands, it triggers profit-taking. KLA's “beat-and-raise” received positive feedback, rising 3% after hours, indicating the market is still willing to pay a premium for certainty on the semiconductor equipment side.
NXP's decline exposes another risk: Q3's “good” cannot cover Q4's “weak” guidance. When the recovery pace of automotive and industrial semiconductors falls short of expectations, the market quickly lowers the pricing for the entire cycle.
📌 Interest rate background: constraints have not exited
The 10-year US Treasury yield rose to about 5.03% after hours, the 30-year rose to 5.37%. The FOMC maintained rates unchanged by 10 to 2 votes, with two dissenters advocating a 25 basis point hike; the statement acknowledged economic activity as “strong” and inflation as “still elevated.” The silence of the rate anchor does not equal concession; the further rise in long-term yields means valuation denominator pressure on high-valuation tech stocks remains unresolved, and the after-hours divergence occurred under this constraint.
In summary: Microsoft bought a “verification passed” ticket for the AI narrative with $18.6 billion in free cash flow; Alphabet was blocked outside with a -$9.065 billion cash flow gap. And Seagate's “drop” and KLA's “rise” show that even in the same direction (AI hardware), the market only wants to pay for “certainty not yet fully priced.” Pause in rate hikes is not a positive signal; BTC surged then fell back
The Federal Reserve maintained rates at 3.75%—4.00%, but this time it was not dovish. The vote was 10:2, with two members calling for an immediate 25 basis point hike. The statement continued to emphasize a strong economy and elevated inflation, with market expectations for another rate hike in December rising to about 86%.
Macro market performance is mixed:
The S&P 500 rose 0.3%, the dollar briefly fell about 0.2%, but the 10-year US Treasury yield remains near 5.05%, and WTI crude oil is close to $99.90. The stock market holding up does not mean liquidity has eased; high interest rates and energy inflation pressures remain.
BTC peaked at $81,674, then retreated to around $80,446. Short term support remains at $80,000, with resistance still at $81,600—$82,000. Without volume-driven stabilization above this range, I do not consider the rebound a confirmed breakout.
ETH returned to around $2,328, with a larger decline than BTC, indicating weak support. Resistance is first seen at $2,350—$2,380; a drop below $2,300 could lead to further testing of previous lows.
This pause has long been priced in; December is the main event. BTC and ETH are currently viewed as rebounding, not for chasing longs, maintaining a cautious stance.$SKHYNIX Q3 exchange loss reached 11.6 trillion KRW, with demand on PC and smartphone ends still weak. More importantly, Hynix itself plans a "significant increase" in capital expenditure in 2027 — the elastic investment rhythm on the supply side is the most classic "self-termination" mechanism of the memory cycle, only currently temporarily postponed by LTA volume lock. For contract positions, this means the near-month bullish structure has fundamental support, but the far-month contracts should not be linearly extrapolated; the "capacity release rhythm" needs to be used as a rolling evaluation anchor. #三星Q3利润预增近8倍，完整财报即将揭晓 In this AI computing power market cycle, the most expensive thing isn't the chips, but the words "already delivered."
Nscale founder Josh Payne recently spoke firmly: despite market doubts, the company is still proceeding with its IPO as planned, targeting a valuation of $30 billion. Supporting this story is the claim of having signed contracts worth $103 billion.
But looking at it from another angle, the situation feels different. As of August 2026, Nscale has actually delivered only 55 megawatts of computing power. Projects have been delayed, there have been safety violations, operations are struggling, and Microsoft has withdrawn from one of its data center projects in the UK.
On one side, there are hundred-billion-dollar contracts; on the other, double-digit megawatt deliveries—a gap of two orders of magnitude. This isn't an accounting issue, but an industry rhythm problem: AI infrastructure in the past two years has been selling "future computing power," with customers paying upfront to lock in capacity, while the actual racks and power only come online two to three years later. When projects are delayed, contracts turn from assets into liabilities.
In the crypto world, the most directly impacted is the decentralized computing power sector. Their model is almost the same as Nscale's: selling future computing power first, then delivering slowly. The industry's loudest player has been proven to have "contracts far exceeding capacity," causing investors to instinctively doubt similar assets. Assets like $TAO and $RENDER will face a round of emotional valuation cuts in the short term.
But looking at the long term, this is a good thing. Those with actual racks, customers, and the ability to run at full capacity will survive; those who only draw roadmaps will be eliminated. #三星Q3利润预增近8倍，完整财报即将揭晓
"Outperforming expectations" should be put in parentheses: 107.4 trillion is higher than LSEG's 106.1 trillion, but lower than Bloomberg analysts' average of 108.7 trillion. Using the same financial report, the conclusion reverses depending on whose expectations you use.
So tonight, regarding the departmental data, what I care more about is this: Is the profit really driven by HBM growth, or is it propped up by DRAM price increases?
#三星电子 #HBM #存储芯片Diva Staking announced it will cease operations and will no longer launch the mainnet. Early staking vault deposits can be withdrawn through Enzyme. At first glance, this seems like bad news for the liquid staking sector.
On the other hand, I prefer to see it as a project endurance alert: funds were already depleted by mid-2024, and the development enthusiasm can't support audit expenses; there is still an exit path for the deposits, so ceasing operations should not be equated directly with users losing all their funds. I would prioritize funding endurance and exit mechanisms over the launch vision.#La growth of Microsoft Azure is 43%, can AI commercialization accelerate? 43%, one might think that Azure's AI story is already in the realization phase. But by putting several data points side by side, my judgment is more complex: the revenue side is accelerating, penetration is still rising, and it is the time lag between the two that is crucial.
Azure's quarterly growth has increased from about 35% to 43%, with annual revenue exceedingDon't rush to interpret Monday's ETF net inflow as "funds returning"; this is where it's easiest to misread. Can a single IBIT return really represent a warming of institutional sentiment? I monitored the market all morning, and the most direct impression is: money has indeed come back a bit, but the way it returned is very selective. Monday's BTC ETF net inflow was $110.3 million, almost entirely completed by IBIT alone, with no obvious follow-up from other institutions. This detail is more important than the total amount; it indicates a single-point probe rather than a full-scale replenishment. It's still too early to conclude whether last Thursday and Friday's net outflows were just short-term rotation. Looking at the broader market, BTC bounced back after repeatedly testing around 80,000, with short-term support holding for now. On the 1-hour chart, the next key levels to watch are 80,600 and 81,400 to see if they can be taken and held. If these are effectively broken through, there is a chance to extend to 82,000 and 82,900. Total crypto funds haven't changed much compared to Sunday, and core holdings haven't significantly withdrawn, indicating that panic selling is not heavy, but the willingness to chase prices is also not strong. On the bullish side: ETFs have turned positive again, on-chain holdings are stable, and BTC key levels have rebounded, showing a window for risk appetite recovery. On the bearish side: inflow sources are too concentrated, indicating internal institutional disagreement; narrative fatigue persists, altcoins have limited follow-through, and FOMO sentiment hasn't truly ignited. Once macro pressure rises again, this kind of one-legged inflow can easily be repriced. My judgment is that fund preference is shifting from "indiscriminate retreat" to "selective trial positions," but it hasn't reached full-scale offense yet. Let's watch first About 654.5 ETH is still in two pre-launch vaults, which is the detail I care most about in this message. Diva canceled the mainnet launch, notified early stakers to withdraw, and did not provide an alternative DIVA distribution plan. In the broader context of fund allocation, what concerns me more is: are those who withdrew willing to continue waiting for the next project?
My speculation is that if funds are more sensitive to waiting periods and realization risks, then this kind of pre-launch staking must offer stronger credibility; otherwise, funds might prefer to remain liquid rather than occupy a position prematurely. The issue is not only about reward expectations but also about how many options are given up during the waiting period.
I tend to see this as a deduction in trust for early project financing rather than evidence of a lack of funds in the entire staking sector. A single project shutdown does not support such a broad conclusion. What everyone cares about most now is whether the opportunity cost of holding coins due to interest rates or the macro risk appetite affects early staking funds more.#谷歌云业务增长82%，AI投入回报如何？
Google Cloud doubles but plunges nearly 10%: AI money-burning race, the market starts calculating the return on investment
Google's Q3 marks an industry watershed: cloud business revenue hits $30.7 billion, doubling directly, profit margin rises to 36.2%, backlog orders at $561 billion, AI computing power demand genuinely has buyers; total revenue $127.6 billion, up 25% year-on-year, seemingly all lines are red, but after-hours it plunged nearly 10%.
The core reason the market votes with its feet is that growth concerns have overshadowed highlights. First, the burn rate far exceeds expectations: quarterly capital expenditure $51.4 billion, free cash flow turns negative $9.1 billion, management also forecasts continued significant investment in 2027, with full-year expenditure guidance as high as $195-205 billion. Second, the basic business shows red lights: search ad growth only 12%, far below expectations, with ChatGPT diverting demand, Meta snatching high-value advertisers, and AI query monetization path unclear, three pressures appear simultaneously.
Net profit down 25% year-on-year, although dragged down by $8.7 billion equity investment losses, the market's real anxiety is: AI on one hand brings incremental cloud business, on the other hand erodes the cash cow foundation of search advertising.After-hours plunged nearly 10% in one go, rushing in now to buy the dip? First, distinguish whether the drop is due to expectations or the bottom.
Here's what I see: Alphabet ($GOOGL) closed Wednesday at $324.89 (+1.34%), with a daily high of $326.41 and a low of $322.66; after-hours earnings showed revenue around $127.6 billion and adjusted EPS of $2.12, both below expectations, free cash flow about -$9.07 billion, and capital expenditures remain high. After-hours price briefly touched about $287.33, closing around $292.50, roughly 10% down from the daily close. Meanwhile, Microsoft’s Azure revenue grew 48%, and after-hours trading actually strengthened, making the contrast even sharper.
My view: This is a one-time pricing in of "AI bills burning too hot and the advertising core business failing to meet expectations," not a signal that "the drop is enough to buy." Treating the after-hours gap down as a bottom-fishing window is the easiest way to catch a falling knife—Microsoft proves some can spin expenses into returns, but Google hasn’t made that case yet.
What I’ll do: Stand firm and observe, not catch this knife; if the stock falls below the after-hours low of $287.33 and fails to recover, the narrative of "dumped and ready to pick up" is invalid for now. Wait for Thursday’s regular trading to digest the gap, then reconsider my stance.
Would you treat a "nearly 10% earnings-driven drop" as an invitation to buy the dip, or first as a risk repricing?
$GOOGL $MSFT $NVDA
#GoogleEarningsMiss #AIcapitalexpenditureconcerns
#GoogleCloudBusinessGrowth82%,HowIsAIInvestmentReturn?Bitcoin returns to 81,000, but the real signal lies in Ethereum's cracks
Bitcoin has climbed back above $81,000, and Ethereum has returned above $2,300. On the surface, this looks like a typical "Fed holds rates steady → risk assets breathe a sigh of relief" rebound. But breaking down the structure, BTC's recovery is more solid than the price suggests, while ETH's follow-up rally actually exposes the true fragility of this market cycle.
First, look at BTC: the rebound is decent, but the driving force comes from shorts covering, not new money.
On October 28, the Federal Reserve voted 10-2 to keep the benchmark interest rate at 3.75%–4.00%, with two officials dissenting in favor of a rate hike. The implied probability of a December rate hike has risen from a low to over 80%. The decision itself only removed short-term uncertainty; the real price momentum came from the derivatives market: in the past 24 hours, $43.31 million worth of Bitcoin shorts were liquidated, accounting for 75.3% of BTC liquidations; total market short liquidations were about $110 million, or roughly 66%. Shorts were forced to cover, providing the most direct fuel for the rebound.
More noteworthy is the leverage structure. While BTC price rose, open interest actually decreased by 0.1%; during ETH's rise, open interest grew only 1.7%. Prices are up, but leverage hasn't expanded in sync, indicating this rebound isn't driven by newly opened long leverage positions. The short-term chip structure is actually cleaner than a "leveraged bull" scenario. Options data also confirm this: on Deribit, BTC one-week skew compressed from about 10% to below 1%, showing a clear retreat in short-term downside protection demand. However, in 1-month, 3-month, and 6-month options, investors still prefer holding puts, and the term spread is telling—short-term panic has eased, but medium-term caution remains.
Ethereum's problem is that it shares the same macro story but not the same capital story.
ETH briefly dipped below $2,300 earlier this week before recovering above $2,350, but the driving force differs fundamentally from BTC. Bitcoin spot ETFs saw net inflows of about $176 million on October 27, marking the second consecutive day of inflows, with BlackRock's IBIT contributing $151 million in a single day. Ethereum spot ETFs, however, showed a different picture: net outflows of about $60 million on October 27, with cumulative outflows in October reaching $1.32 billion, heading toward the heaviest monthly redemption since November 2025. On one side, institutional funds continue accumulating near 80,000; on the other, ETH products are being consistently redeemed. The divergence in capital flows is even more striking than the price divergence.
Supply-side pressure is also present. Wallets associated with the Ethereum Foundation recently sold 7,594 ETH (about $17.8 million), and another dormant 1,500 ETH holding, silent for nine years, was sold. These are not ordinary profit-taking moves—the Foundation-related address reductions carry signaling significance, and the awakening of early holders raises questions about whether "old chips are beginning a systematic exit." Ethereum's tug-of-war around $2,300 essentially reflects a fragile balance between spot selling pressure and ETF bleeding, a balance not yet broken by upward forces.
Institutional divergence falls precisely between "bottom" and "still some way to go."
Galaxy Digital research head Alex Thorn is relatively optimistic, believing Bitcoin may have bottomed, noting the 50-week moving average has risen for five consecutive weeks. Historically, after falling below this average for over 20 weeks and then reclaiming it, BTC has never dropped more than 8.4% within a year from the recovery point. Morgan Stanley's Mike Wilson also believes most of the market adjustment is complete. On the other hand, Citi, Standard Chartered, and Bernstein have recently lowered their year-end Bitcoin price targets: Standard Chartered from 150,000 to 100,000, Bernstein to 125,000 while pushing the 150,000 target to mid-2027, and Citi's 12-month target is 113,000. Optimists focus on structure and cycle position; cautious voices emphasize macro constraints and slow regulatory progress—these views are not contradictory, but rather reflect that the "bottom zone" and "immediate trend up" are separated by a distance requiring sustained spot buying validation.
What really needs attention next is not whether 81,000 can hold, but December.
The Fed's "hold steady" this time has been re-priced by the interest rate futures market as a "skip" rather than an "end." If a December rate hike occurs, the increased borrowing costs will directly suppress high-duration risk assets like crypto; if tech stocks lose momentum after earnings season, BTC's risk appetite borrowed from tech stock correlation will also fade. Ethereum additionally needs to answer when ETF outflows and Foundation-related selling will show clear convergence signals. Until then, ETH's rebound is better seen as a follow-up repair rather than an independent reversal.
BTC at 81,000 and ETH at 2,300—prices were lifted by the same hand on the same macro night. But BTC was supported by short covering and ETF buying, while ETH was propped up more by technical buying after overselling. The same night, two different qualities. For readers, distinguishing this is more useful than remembering a few price points.
#美联储维持利率不变，12月加息分歧升温 $BTC Google, you really are damn it. You caused my XP to reset to zero!!!
AI is also involved
Cloud business is growing too
But when the earnings report came out
The stock price slapped back?
But thinking carefully
This matter is not that simple
Cloud business grew 82%
Sounds really good
But the problem is
The hotter AI gets, the more money Google has to throw in
Data centers need to be built
Chips need to be bought
Computing power must keep expanding
Revenue has gone up
But investment is also continuously increasing
What the market really worries about
Is not that Google can't make money now
But how much profit can be recovered in the future
from all the money thrown in
What's more interesting is
AI on one hand brings new growth to Google
On the other hand, it may change the way search makes money
Before, when users searched for things
Google made money from ads
$GOOGL #Strategy财报前瞻：BTC账面收益大幅反弹 Let's calculate, this message still can't determine how much efficiency has improved: the material doesn't provide transaction volume, fees, or time to receipt. Peer is testing support for Alipay payment in $USDC transactions on Base, where buyers pay individual sellers in RMB.
① Transaction volume is undisclosed. Having an additional payment path cannot be directly counted as liquidity growth.
② Fees are undisclosed. Being able to pay in RMB does not mean the cost of currency exchange is lower.
③ The feature is still in testing, and Peer clearly has no official cooperation with Alipay. Supporting a payment method and having endorsement from the payment provider are two different things.
My judgment is that this is progress at the entry level, but not enough to prove an improved trading experience. When I see this kind of news, I usually separate "can use" from "easy to use" in my assessment. When you evaluate such entry points, do you focus more on transaction volume, actual fees, or time to receipt? Feel free to add your own metrics as well.$BLUR Blur will directly disconnect the main site frontend on the 12th of next month, and Blend lending will also stop matching new orders.
Back then, everyone rushed for it because of that interface and the royalty war frenzy.
Now the monthly trading volume has shrunk to less than 7 million, and this protocol originally had no fee dividends; now they are simply cutting off even the traffic entry.
The market has directly smashed through the last bit of the bottom support zone, it's unbearable to watch.
Don't always think that after a 99% drop you can bottom-fish for a rebound; the main business has been unplugged, holding onto a bunch of dead narratives as treasures is purely fueling the shorts.
#BLUR
$BLUR
👇#微软Azure增长43%，AI商业化能否提速？ Just looking at the 43% figure, one might think Azure's AI story has already entered the fulfillment phase. But putting several data points together, my judgment is more complex: revenue is accelerating, penetration is still climbing, and the time lag between the two is the key.
Azure's quarterly growth rate rose from about 35% to 43%, with annual revenue surpassing $100 billion for the first time. Capital expenditure for the quarter was $45.2 billion, up 30% year-over-year, with an annual guidance of about $175 billion. Copilot paid seats increased from 20 million to 30 million, a 50% quarter-over-quarter increase, but only account for about 6.6% of M365 commercial seats, whose core seat growth has slowed to 5%–6%.
This means AI's contribution to Azure is real, but Copilot's penetration among existing office users is still a story at the 6% level. AI service gross margin is between 25% and 30%, lower than traditional cloud business. As AI revenue share rises, Azure's gross margin is expected to decline from about 60% in fiscal 2024 to around 52%.
The 43% answers "Has Azure become faster because of AI?" The 6.6% answers "Has AI truly become productivity for everyone?" The former is already validated; the latter determines the slope of commercialization acceleration. If the next stage penetration rate remains in the single digits, the capital expenditure payback cycle will be extended, and Azure's growth will eventually face a smaller denominator.A thought-provoking investment question: If the stock market no longer delivers the high returns of the past decade in the next ten years, can you still stick to your original strategy?
This set of historical data breaks down the S&P 500's real annualized returns into different periods:
• 1928–1948: 0.6%
• 1949–1968: 12.7%
• 1969–1984: 0.5%
• 1985–1999: 15.1%
• 2000–2012: -0.8%
• 2013–2025: 11.8%
The key insight is not that bull markets will definitely end at a certain point, but that long-term average returns do not mean every generation of investors will have the same experience.
Some catch high growth periods, while others may face low or even negative returns over a decade or more.
Therefore, "buying the dip" can be a strategy, but it is not a guaranteed formula for profit.
Investment plans also need to consider purchase valuation, investment horizon, cash flow, and the ability to withstand drawdowns.
Don’t just ask how much $BTC or U.S. stocks can still rise; also ask yourself: if they don’t rise for years, can I still hold on?#谷歌云业务增长82%，AI投入回报如何？ The return on AI investment in Google Cloud has already appeared on the cloud business profit side, but it has not yet closed the loop on the group's cash flow side.
Profit-side returns have appeared: Google Cloud revenue is growing rapidly, operating profit and profit margin have significantly improved, indicating that AI infrastructure, TPU sales, and Gemini-related demand are being converted into incremental profits, not just burning money to gain scale.
The cash flow side cost is heavy: capital expenditures have risen sharply, the group's free cash flow is under pressure and even negative, and operating cash flow growth temporarily cannot keep up with capital expenditure growth.
Comprehensive judgment: Google Cloud's own unit economic model is improving, and the signal that AI investment "can make money" has strengthened; however, the group's overall "investment to return" net amount has not yet turned positive. Seeing this trade, what comes to mind is: big money willing to bet on a direction and the entire market starting to avoid risk are two different things. I usually like to view a single trade within the context of liquidity first, and I don't rush to treat one wallet as a macro indicator.
This time, Core Trader 97 (0x940acf…7e639) opened a short position on $HYPE with 10x leverage, with a nominal position size of about $512,500. This is a monitoring signal for a new or increased position. The outcome is not yet known, but what can be confirmed is that it increased its short exposure.
My judgment is that this trade is more suitable as a clue to risk appetite observation and is not enough to prove that funds are fully retreating. The reason is straightforward: nominal position size does not equal invested principal, and without seeing asset allocation outside the wallet, just looking at this leg alone, it's unclear whether it is betting on a decline or hedging other long positions.
According to my speculation, if macro liquidity tightens and the market's willingness to take risks decreases, then this kind of short expression might get broader fund behavior resonance; if it is just a judgment on a single asset, then using it to explain the whole market is going too far. Another constraint is leverage: 10x will amplify the pressure of adverse fluctuations, and a large position does not mean it can withstand more.
I tend to first observe whether risk appetite weakens synchronously, reminding myself not to interpret a single significant trade as the answer for the entire market. Today it was confirmed that DTCC (the core depository clearing organization of the U.S. securities market) officially launched securities tokenization services, which is more worthy of long-term record than typical project collaborations. I have greater confidence in the long-term development of RWA (Real World Assets), but the initial beneficiaries may not necessarily be existing public chain tokens. In the short term, I remain cautious: capital improvement is mainly in BTC, ETH and SOL have not kept up, and the Federal Reserve has not given a clear easing signal. 1. ETF: BTC has had inflows for two consecutive days, with buying still mainly from IBIT. October 27 spot ETF (Exchange-Traded Fund) data: * BTC: net inflow of $176 million. * ETH: net outflow of $60.2 million. * SOL: net outflow of $19.2 million. BTC's total net inflow over two days is about $286 million. This time FBTC also had subscriptions, but the majority still came from IBIT. BTC capital continues to improve, while redemption pressure on ETH and SOL has not eased. 2. DTCC personally doing tokenization, competition begins to approach the core backend of the securities market. DTCC's DTC officially launched tokenization services, with 20 institutions signing access agreements, including JPMorgan Chase, Bank of America, Citibank, Morgan Stanley, and Fidelity. Qualified participants can convert DTC-custodied securities into on-chain digital certificates, transferable between approved wallets. The first batch connects to Canton and DTCC's private Besu network, with plans to gradually add other networks starting in 2027. What I value more is the core backend of the U.S. securities market's... #谷歌云业务增长82%，AI投入回报如何？
If you only look at the 82%, you might think this is just an acceleration of the cloud business; but after reviewing the full Q3 data, my judgment has changed: the demand for AI is real, the cost is real, and the cost is shifting from "implicit capital expenditure" to "explicit cash flow wounds."
Q3 Google Cloud revenue was $30.7 billion, a year-over-year increase of 103%, with an operating margin rising from 23.7% in the same period last year to 36.2%, nearly doubling operating profit. The cloud business has proven it is not just "burning money to gain scale"—AI computing power utilization and enterprise customer penetration are indeed releasing operating leverage. Backlog orders stand at $561 billion, with over 55% expected to be recognized in the next 24 months, providing high visibility on the demand side.
However, in the same financial report, free cash flow was negative $9.1 billion, negative for the second consecutive quarter; capital expenditures were $51.3 billion, doubling year-over-year, and expected to "significantly increase" by 2027.
More subtly, although TPU external sales contributed revenue, its operating margin is about 20%, lower than the overall cloud business margin of 36%, and the larger the scale, the more it dilutes the overall margin. This means Google Cloud's high profit margin is not guaranteed to scale linearly.
My conclusion is: the 82% or 103% answers the question "Is there demand for AI?" while the negative cash flow answers "Is this business profitable at this stage?" Both can be true simultaneously without contradiction—the market was willing to price the former in the past, but now it is starting to demand a timeline for the latter.The core of Waller's press conference is just one word: wait. The Federal Reserve is holding interest rates steady at 3.75%–4.00%, but Waller explained very clearly "why wait." First, the economy is not bad, so there is no reason to rush. GDP growth was revised up from 1.8% to 2.4%, the third quarter may exceed 3%, corporate investment is heading toward double digits, unemployment is lower than a year ago, and initial jobless claims remain at decades-low levels. Employment and growth have not triggered any red flags, so the Fed does not feel pressured to cut rates. Second, inflation has been stuck for nearly six years, but Waller does not want to make decisions based on the "calendar." He clearly said that lowering inflation does not depend on "a certain day" or "a single data point." The implication is: don’t infer the Fed’s actions from one meeting or one CPI figure; they are watching trends, not isolated points. Third, AI investment is also affected by interest rates; don’t think the tech boom is unrelated to monetary policy. When asked if AI is insensitive to interest rates, Waller directly denied it: it’s important, these sectors cannot escape the influence of interest rates. This is a reminder for crypto and tech assets—no matter how strong the AI narrative, the denominator side (interest rates) is not exempt. Fourth, what’s really worth noting is not this rate decision, but the change in meeting arrangements. Starting in 2027, the FOMC will hold one meeting each year dedicated to discussing monetary policy strategy without voting on rates or issuing policy statements; in 2028, this will increase to two meetings. Waller’s exact words were that time will be set aside to consider the "left side of the decimal point" issue. To translate: the short-term interest rate path may be "monitored" more frequently.BTC only rose by 0.02%, but CSPR surged 39.42%, making today's contrast quite striking.
The current price is about 0.0029, with a 24h high of 0.003072 and a low of 0.002053, an amplitude close to 49.63%—meaning it covered almost half the intraday range. This movement isn’t a gradual lift but more like a sudden push followed by a relay of momentum from followers.
An interesting detail: the 24h trading volume is only about 784,000 USDT. It rose by 40%, but the volume didn’t keep up. Small-cap coins have light liquidity; a few slightly larger orders can skew the K-line slope, making it look lively, but the actual turnover of chips is limited. So this wave feels more like emotion-driven rather than a trend built by sustained capital accumulation.
Meanwhile, the main market is quiet; BTC’s amplitude is only 1.77%, basically grinding in place. Funds aren’t flowing into the mainstream but are diving into small caps, a structure that’s been common recently. Whether CSPR’s narrative has been revived or it’s just a pulse caused by low liquidity, I can’t tell for now. For a coin with nearly 50% amplitude, most who chase it probably know what they’re gambling on.
What do you think about this kind of "exaggerated price increase but average trading volume" market? Does it look more like someone is genuinely building a position, or is it just that the order book is too thin? #三星Q3利润预增近8倍，完整财报即将揭晓
The core of Samsung's full Q3 earnings report is not the total profit of 107.4 trillion KRW, but the profit structure differentiation: DS operating profit exceeds 110 trillion KRW, while DX suffers a loss of over 1 trillion KRW.
HBM4 is the benchmark of value: sales increased more than threefold quarter-on-quarter, with HBM expected to account for over 60% in the second half of the year, and a yield rate of about 80%; SK Hynix's shipments fell short of expectations due to certification delays. The conference call needs to confirm whether this is a competitor's mistake or Samsung's structural establishment.
The price increase slope is also shifting gears: Q3 DRAM and NAND average prices rose about 17% and 18% quarter-on-quarter, respectively, significantly narrowing compared to the first half of the year. If DS profit margin continues to expand, it indicates that HBM4 and combination upgrades are replacing price increases; if they fall simultaneously, the cyclical component is higher. The DX recovery path is equally critical; smartphones are being hit by high-priced storage and Q4 production has been cut. Whether management can provide price increase or combination adjustment plans is more important than the loss figures.
Valuation focus is on the conference call: about 30 trillion KRW dividends and Q4 price guidance will determine whether the 4x PE is an overkill or a reasonable discount. In short: 107 trillion proves the intensity of AI storage demand, but the earnings report must answer how much is structural upgrade from HBM4 and how much is residual heat from the price increase tail.Dollar-cost averaging Bitcoin 20 u on day 123
The Federal Reserve's interest rate meeting went as expected with no changes, but the result caused a big stir on Google. Just now, I saw many people going crazy trying to bottom-fish and go all-in on Google. I can only shake my head; it’s clear they haven’t been in the US stock market long and know very little—like newborn calves unafraid of tigers.
1. In 2022, Netflix plunged 22% in a single day, prompting well-known investor Bill Ackman to continuously buy the dip at an average cost of $400. However, due to ongoing user losses and declining performance, Netflix dropped to a low of $165. Three months after bottom-fishing, Ackman cut his losses at an average price of $225, losing $400 million.
2. Mr. Munger bottom-fished Alibaba, initially buying at $240 with an average cost of $181. Alibaba then dropped to a low of $61. Mr. Munger significantly reduced his position near $75, admitting his mistake with losses exceeding 45%.
3. Meta reached a high of $338.54 at the start of 2022 but fell to $88.91 by year-end, a decline of 74%.
There are countless other examples. I hope the above data can warn friends who are impulsively trying to bottom-fish: you are investors, not gamblers. Always remember Peter Lynch’s advice:
"Don’t try to catch a falling knife. Wait until it falls, sticks in the ground, wobbles, and stops moving before picking it up." Regarding this kind of adjustment, my initial calculation is: is having a bit more turnover space worth bearing more volatility? My judgment is that this is more valuable for those with a clear repayment source, while for those wanting to borrow to the max and leverage up, the odds may not improve.
Aave has raised the borrowing limits for Bitcoin and Ether collateral on V3 and V4. Borrowers of WBTC and cbBTC in Core can get about 11% more borrowing capacity with the same collateral; however, if borrowing up to the new limit, the buffer against price drops will shrink.
If it's just for short-term turnover, the increased limit may reduce the need to add more collateral, so flexibility has value. But if used to increase risk exposure, I speculate the chance of forced liquidation during a downturn also rises. The increase in limit changes the options available but does not improve the borrower's repayment ability, which is the dividing line in my expected value calculation.
Faced with this roughly 11% extra limit, do you value more the odds of keeping enough buffer, or the odds of improving capital utilization?