海盗鸭

In pre-market trading, Nasdaq futures directly gapped higher. The trigger was Amazon’s after-hours earnings report for $AMZNB , in which AWS growth re-accelerated. Its cloud and AI capital expenditure guidance was raised even above market expectations. What’s interesting is that this rally wasn’t just Amazon’s— the chip sector caught fire as well. The Philadelphia Semiconductor Index futures also moved up in sync— $SMHB . The logic is straightforward: when cloud companies increase capex, the first to benefit are the upstream compute suppliers. I’ve been watching for two quarters— the market had been worried that the giants would cut spending, but Amazon essentially disproved that with the numbers. Still, keep an eye out: a market driven by an earnings catalyst and tied sector spillover often gets digested within one or two days. The risk of chasing is that you’re effectively buying someone else’s expectation getting fulfilled. Near term, sentiment is fairly warm, but don’t treat a single-quarter capex guidance as confirmation of a long-term trend—capex can be cut just as quickly as it’s raised.
#美股财报 #AI chips

The trading volume of RWA perpetual contracts on Hyperliquid and Binance is actually getting close to Bitcoin—this signal is quite easy to overlook. RWA has moved from concept hype earlier in the year to real derivative demand now, which means this track is starting to consolidate rather than just being slogans. As $BTC serves as the anchor for the entire crypto market, the fact that its volume is being diverted to new sectors is actually a sign of market maturity. I will treat the RWA trading volume share as an indicator to observe sector rotation. In the short term it can’t take pricing power from Bitcoin, but the trend is worth watching over the long run.

In Texas, criminals used encrypted ATM scams to trick people out of $57 million, and lawmakers are now considering directly banning them. Three states have already outlawed Bitcoin ATMs, and the committee chair has even said they want to “go even further.” These kinds of offline kiosks have long been a gray area: transfers are irreversible, and older people are especially easy to deceive, making them a hotspot for scams. I’m not surprised that regulators are tightening oversight, but note that this crackdown targets offline exchange channels—not the compliance of $BTC itself. This is really two separate lines: one is cleaning up scam exit routes, while the ETF and institutions are still moving in. Don’t interpret this kind of local news as a major negative for the market.

An anomalous data point: Robinhood now earns more from prediction markets than from crypto trading. Don’t underestimate this signal—the retail crowd’s speculative enthusiasm is shifting from buying coins to “betting on events,” suggesting that pure coin price fluctuations are becoming less enticing. For us, that may not necessarily be a bad thing—when exchange revenue models change, companies like Coinbase that rely on trading commissions will also need to think about a transition. $BTC ’s spot activity level in the near term may

The ETH/BTC ratio has just surged above 0.030, hitting a three-month high. A lot of people are already shouting that the alt season is coming. But the data actually pours cold water on that—at the same time, Bitcoin’s market share is also rising. What does it mean when both are going up? Money is concentrating into the mainstream coins, not being scattered into grassroots coins. The real alt season is when BTC goes sideways and funds overflow to pump smaller coins—this situation clearly isn’t that. $ETH ’s relative strength can only be seen as the leader catching up with gains; don’t treat it as a signal of a broad-based breakout. If you really want to wait for an alt season, you’ll have to see Bitcoin turn down and make room. As for $BTC here, if it doesn’t loosen up, it will be hard for altcoins to have a major run.

Bitcoin’s intraday volatility has narrowed to its tightest level in more than a month—so quiet it feels a bit abnormal. Anyone trading short-term knows this is exactly when it’s hardest to make a move; you get swept for losses back and forth. Historically, such extreme compression of low volatility is often a prelude to a volatility surge. The direction isn’t known, but the energy is clearly building. $BTC In my experience, when things like this happen, don’t rush to take a heavy position betting on direction. Wait for a confirmed breakout and then follow through. The calmer the chart looks now, the more cautious I become—volatility mean-reverts. If it stays quiet for too long, it has to be paid back.

Korean news just out says that SK Hynix’s HBM orders have again been pushed to next year. Market rumors suggest that in Q3, HBM shipments will see another step-up quarter-over-quarter. The stock rally this time was indeed driven by demand for AI compute capacity—its logic holds: as long as NVIDIA is still fighting to secure production capacity, upstream high-bandwidth memory will have buyers. But let me add one caution: HBM is now split among three players: SK Hynix, Samsung, and Micron. Micron’s $MUB last quarter guidance was also very strong, meaning this cake is being re-cut. SK Hynix currently leads in market share, but if Samsung’s yields catch up, the premium will ease—this is a risk not fully priced into its valuation right now. Whether to buy on a pullback depends on whether you believe this wave of AI capex can hold up through the year after next—demand is real, but the pace won’t be a straight line. $SKHYB is worth watching for upstream inventory inflection points; don’t just watch the stock price.
#AI芯片 # Apple chip shortages are dragging down sales expectations

After Amazon’s earnings report came out, the stock jumped immediately. AWS’s cloud revenue growth accelerated again this quarter, easing a lot of the market’s concerns about “AI burning too much money and the payback still being far away.” This is actually quite important: for the past two quarters, what people were questioning was capital expenditure as a black hole. Now AWS is responding with real, cash-based revenue growth—because cloud demand is the most direct monetization channel for AI implementation. Compare Microsoft’s $MSFTB Azure and Google Cloud: all three are competing for this segment. Whoever can rent out computing power effectively and keep profit margins stable will be the first to prove that the AI numbers add up. This time, Amazon has set an example for the whole sector. But don’t get overly excited: it’s easy to speed up in a single quarter—whether it can hold steady through several consecutive quarters while capital expenditures climb is the real test. $AMZNB In the short term, sentiment is relatively warm; over the medium term, we still have to see whether gross margin can hold up.
#美股财报 # Nasdaq rebounded 2.8%, ending a six-day losing streak

The U.S. has stepped in to buy Japanese yen. This is the first time since 1998 that it has actively intervened in the FX market, breaking a decades-long tacit understanding. Bitcoin reacted immediately, dropping briefly to around 63,000. Why did crypto wobble first? Because this move directly brings back memories of unwind operations from carry trades—once the yen is forcibly pushed higher, those who borrowed cheap yen to lever up into risk assets have to pull back. Highly volatile assets like Bitcoin are often the first to get hit. The global selloff triggered by last August’s yen appreciation is still fresh in mind; this time, the market is responding reflexively, running for cover first. My view: a single intervention won’t change the trend, but the signal it sends is that central banks’ tolerance for exchange rates is declining, and the liquidity backdrop is becoming more delicate. $BTC is under near-term pressure—don’t rush to bottom-fish; first, see whether the yen continues to be pushed up consecutively.
#比特币 #gold falls back

A company that makes protective gear previously loudly announced it would pursue a reserve strategy of using $32 million to buy Bitcoin. But that 45-day backstop financing facility quietly expired on October 9—without buying a single coin, issuing any shares, or receiving any funds. This is pretty representative: after the DAT (digital asset reserves) narrative was popularized last year by MicroStrategy, a bunch of low-tier companies riding the hype issued announcements to pump their stock prices. But when it came time to deliver, it was all hot air. $BTC is fine in itself, but that kind of ragtag crowd using it to tell a story should be cleared out. Only the ones like $MSTRB that truly backed it with real money to buy hundreds of thousands of coins deserve to be called reserves.

IBM’s CEO recently said that quantum computing will realistically start impacting company revenue from 2028 to 2029, and he even drew a picture of a grand $1 trillion value by the end of this century. After that, Bitcoin’s encryption system was brought up again for discussion—there are estimates that around $4.37 billion worth of BTC faces potential quantum exposure risk. My take: don’t be scared by the headline. A practical quantum machine that can break existing encryption by 2028 is still far off, and the community has already been working on post-quantum (anti-quantum) solutions. This is the kind of narrative to keep an eye on long-term, but not to use as a reason for trading in the short term—emotional hype outweighs real threats. $BTC

The last day of July is a bit divided: BTC and ETH are pulling back, while the stock market and Asian trading hours are rising, and U.S. stock index futures are also moving up. But looking over a longer span, the CoinDesk 20 index is set to close this month with its largest monthly gain in a year. This combination of “month-end profit-taking, but a big rise over the whole month” is pretty typical—selling off near month-end to lock in profits doesn’t necessarily mean the trend has reversed. $BTC is indeed under short-term pressure, but the monthly trend for the entire crypto sector is actually quite strong, and $ETH is riding along with this rebound too. Don’t let the red candles from the last two days throw you off the timing—look at the structure of the monthly chart rather than focusing on intraday moves.

Something else happened again: a certain hardware wallet has a flaw in its random number generation. The mnemonic phrases—supposed to be “impossible to guess”—became enumerable. In 25 minutes, 594 of them, including $BTC , were swept clean, and about $38 million is gone. The most painful part of this is—it's not that your password was leaked; the wallet’s underlying random number generation was broken from the start, and users did nothing wrong. Cold wallets have long been treated as the safest option, but this time it directly blows that myth apart. Reminder: don’t blindly trust “hardware = absolutely secure.” Use multisig and distribute storage whenever you should. Security is a tightrope—during a bull market it’s the easiest to loosen, and it’s exactly when you can’t afford to loosen it.

The weekend market is quite split: Bitcoin keeps grinding around the 63K level—neither breaking upward nor collapsing. But PUMP and PI, two altcoins, flip and deliver double-digit gains. Clearly, money is picking short-term setups during the lull in mainstream coins’ consolidation, chasing the elasticity of small-cap, high-volatility names. This kind of structure is actually very typical: when the broader market doesn’t move, hot money won’t stay idle and instead goes after smaller, more volatile targets to make a quick play. The problem is that this sort of行情 comes fast and leaves even faster—most who chase in will likely end up taking the last baton. $BTC being stuck at the 63K level is crucial: if it breaks above, the altcoins can follow and share the meal; if it breaks down, the ones that are surging today will likely fall even harder. My preference is still to first look at the direction of the majors. Altcoins’ independent swings can be played in the early phase of a bear-market rally and the start of a bull run, but at this “stuck-in-the-middle” spot, volatility is the biggest trap.

Coldcard’s estimate of losses from this theft has been revised upward again—after Galaxy’s analysis report was released, the figures were raised directly from the initial reporting to $70 million. Hardware cold wallets have long been treated as the last bastion of self-custody, so this kind of incident is a significant blow to the “not your keys not your coins” narrative. The detail that’s most worth pondering is that the loss amount has been continually revised upward, suggesting the attack surface is broader than initially thought—perhaps it wasn’t a single-point vulnerability, but instead impacted a batch of devices or some firmware component. This has limited impact on the price of $BTC itself; $70 million isn’t large in the overall market, but it carries an implicit hit to market confidence: retail users already have mixed feelings about self-custody, and this news will only push more of them back to exchanges. Security is always the most easily overlooked link in crypto—and when things blow up, it’s also the most deadly. Every cold-wallet holder should take a moment to re-check their firmware version.

The last earnings report before Cook hands over the CEO scepter—Apple’s stock plunged after hours. The biggest drawback was a snag in the supply chain. The management’s guidance for the next quarter was clearly worse than what the market expected, and chip and component shortages were mentioned repeatedly. Interestingly, the market showed little reaction to the emotion of “the end of the Cook era.” What actually drove the sell-off were solid fundamentals: iPhone supply schedules were disrupted and hardware gross margins came under pressure. Looking upstream, as Apple’s largest contract manufacturer, TSMC’s guidance is essentially a forward-looking signal: the production volume of A-series chips is directly tied to shipments in the next two quarters. What Cook has been most praised for over the years is supply chain management, and yet his farewell battle stumbled right here—there’s a certain irony. In the short term $AAPLB sentiment won’t be great, but I’d rather focus on how this affects the entire industry chain—because the real scale isn’t confined to Apple alone. It’s in upstream production changes like $TSMB , which are the leading indicators.

Coldcard hardware wallets are being continuously drained; a private key vulnerability caused by weak randomness has already siphoned off $38 million. The official message urges people to move funds immediately. This is a brutal blow to the belief that “cold wallets are the safest”—hardware isn’t absolute security; if the random number generation is flawed, even the coldest wallet is just paper-thin. Large holders, especially those with $BTC , should take this seriously, and anyone from the affected batch should move promptly. When prices are rising, the safety string is most likely to slack—this is precisely the time it should be tightened.

A hardware wallet vulnerability wipes out $38 million in BTC in just 25 minutes—so-called “seeds that can’t be guessed,” because the random number generation is flawed, directly turn into something that can be enumerated. This is the most painful part: you think a cold wallet is the safest, but the security is actually rotten at the entropy source. This is a reminder to everyone holding $BTC : self-custody isn’t absolutely secure—your device and firmware trust assumptions themselves can collapse. In the short term it may not affect the overall market, but for users of specific wallet brands it’s real panic. Check immediately whether you’re using a batch that’s been affected.

Cook gets another interview with the same old “the Apple story will continue to unfold.” The phrasing is pretty, but these vague, optimistic words often show up when growth is least fresh. Apple’s biggest problem right now isn’t that hardware can’t be sold—it’s that the AI narrative has fallen a whole step behind: over at Google, Gemini has already been pushed into search and the Android ecosystem, while Apple Intelligence is still squeezing toothpaste. If you truly believe “the story will continue to unfold,” you first need to see whether the next generation of products can deliver reasons good enough to make people pay. $GOOGLB is actually worth watching more when it comes to real-world deployment speed in AI—whichever story gets delivered first, that’s where the money will go.

Coinbase’s earnings this quarter again missed expectations. While Wall Street generally blames the weak crypto market, there’s a big disagreement about “when will it recover?” This is actually easy to understand: an exchange’s revenue is essentially a mirror of market sentiment. When prices are cold and trading volumes shrink, performance naturally looks worse. So when you look at $COINB , don’t just look at it by itself—you should first check whether the trading activity in $BTC has returned. If volume picks up, the exchange’s days will get better. In this kind of volume-contracted, choppy market, profit pressure is the norm. Instead of obsessing over quarterly numbers, it’s better to watch for signals of capital returning to the broader market.