Orbit Post Sitemap
$SNDK faced heavy selling as the 30Y Treasury yield stayed near multi-year highs. Price fell from $1,827 to $1,566 before recovering toward $1,612, near my $1,615 cost.
The issue isn’t storage demand—it’s valuation. With long-term yields elevated, investors are less willing to pay today for profits expected years ahead.
AI trade overcrowding and stretched valuations added fuel. Next, watch the 30Y yield: if it stays high, high-beta rebounds may struggle.
#XiaomiQ2Earnings The more specific stablecoin regulation becomes, the more $ETH resembles a financial highway, and $BTC resembles a safe in the digital dollar world.
The stablecoin rules related to the GENIUS Act have entered the market's view, making terms like customer identification, anti-money laundering, issuance licensing, reserve supervision, and payment stablecoin definitions increasingly important. Many people think this is news only for stablecoin issuers, affecting only USDT, USDC, banks, and payment companies. But in reality, stablecoin regulation will redefine the division of labor between BTC and ETH in on-chain finance.
$ETH is one of the most direct beneficiaries. Stablecoins are the cash layer of the on-chain world, and the Ethereum ecosystem has long carried a large amount of stablecoins, DeFi collateral, on-chain liquidation, and RWA attempts. If the digital dollar becomes more compliant, more institutions, payment companies, and financial platforms will find it easier to enter on-chain settlement. Once funds flow on-chain, smart contract platforms, settlement layers, and application layers are needed, making ETH's infrastructure value more visible.
However, stablecoin compliance will also subject the ETH ecosystem to stricter scrutiny. Previously, the on-chain world was very free: protocols ran themselves, front ends connected themselves, and users bore the risks. But once stablecoins become formal financial products, wallets, DeFi front ends, RWA issuance, cross-border payments, and custody platforms will be required to take on more responsibility. ETH's opportunities come from financialization, and so do its pressures. The more it resembles financial infrastructure, the less likely it is to grow entirely without regulation.
$BTC's logic is completely different. Stablecoins are not substitutes for BTC because stablecoins are just digital dollars. They improve the liquidity efficiency of the dollar but do not solve whether the dollar will be diluted long-term. The more successful stablecoins are, the more people enter the on-chain world, and the more funds get accustomed to staying on-chain. Users first use stablecoins for transfers, trading, and payments, then ask: besides digital dollars, what else can I hold? If I don't want to place long-term value entirely in dollar credit, is there an on-chain hard asset? This question brings BTC back to the table.
So the more compliant stablecoins become, the busier ETH gets, and the clearer BTC's role becomes. ETH is like a highway with stablecoins running on it; BTC is like the safe at the highway's end, which users realize after running for a while that they can't just hold cash. One is responsible for liquidity, the other for reserves. Stablecoins bring the dollar on-chain, ETH enables the dollar to be used on-chain, and BTC reminds the market that the on-chain world shouldn't be only dollars.
This is also why stablecoin regulation cannot be understood simply as payment news. It will expand the entry points to on-chain finance and enlarge BTC's potential user base. Without stablecoins, many people would never enter crypto; with stablecoins, they first become familiar with wallets, on-chain transfers, trading platforms, and custody services; after familiarity, BTC and ETH have the chance to become reserve assets and settlement assets respectively.
In the short term, stablecoin rules won't immediately push BTC past $64,000, nor will they instantly free ETH from $1,900. But in the long term, it may be a more important variable than single-day ETF inflows. ETFs open asset entry points; stablecoins open usage entry points. One lets people buy crypto; the other lets people use on-chain finance daily.
The bigger the digital dollar, the easier it is to see ETH's settlement layer value; the more compliant the digital dollar, the harder it is to ignore BTC's non-dollar attributes. Stablecoins are not the end; they are the cash layer of on-chain finance. Once the cash layer grows, both the road network and the safe become more valuable. ETF funds fluctuating repeatedly is not necessarily a bad thing; the real question is: does $BTC have a long-term allocation base, and does $ETH have on-chain yield buying demand?
Currently, the market's view on ETFs is too emotional. When BTC ETF inflows occur, everyone says institutions are bullish again; when BTC ETF outflows happen, everyone says institutions are fleeing. When ETH ETF shows a little movement, the market starts fantasizing about a revaluation of on-chain finance; if ETH ETF performance is flat, people say no one wants ETH. In fact, ETF funds are not believers; they are allocation funds. Allocation funds adjust positions based on interest rates, risk budgets, client redemptions, quarterly rebalancing, volatility, and macro events.
For $BTC, the real proof ETFs need to provide is whether a long-term allocation base has formed. The buying logic for BTC ETFs is clear: digital gold, non-sovereign asset, fixed supply, macro hedge, portfolio diversification. Institutions may not believe BTC will rise every day, but they might be willing to hold a small portion in their portfolios. BTC's real strength is not to have a single fund go all in, but to have more and more portfolios feel "they can't be completely without it."
Therefore, BTC ETF outflows are not scary; what is scary is if no one picks up the price after outflows. If BTC remains around $63,000 to $64,000 and does not break down despite ETF fluctuations, it indicates other buyers underneath: long-term holders, large on-chain wallets, corporate treasuries, native funds, macro allocators. Bad news failing to break support is more valuable than good news pushing out a bullish candle.
The test for $ETH ETFs is completely different. ETH ETFs are not buying a simple scarce asset but an entire set of on-chain economic assumptions. Institutions buying ETH are not just buying the second-largest coin; they are betting on staking yields, stablecoin settlements, DeFi, RWA, L2, and smart contract ecosystems. This story is richer than BTC's but also harder for traditional funds to quickly accept. For ETH ETFs to have sustained inflows, they need to prove that on-chain yields and application value are worth allocating to.
This is why ETH holding just around $1900 is not enough. BTC holding can indicate allocation demand remains; ETH must prove active buying is returning. If ETH ETFs remain weak for a long time, it means institutions are still cautious about on-chain finance; if ETH ETFs start continuous inflows, especially as staking yield mechanisms become clearer in the future, ETH's valuation logic will change significantly.
BTC and ETH in the ETF era are not the same exam. BTC tests assetization of faith; ETH tests assetization of on-chain yield. BTC ETF inflows represent traditional funds entering crypto; ETH ETF inflows represent traditional funds willing to buy on-chain finance. Only when both strengthen simultaneously can the market move from defense to expansion.
What matters most now is not how much ETF inflow occurs on a single day, but whether inflows are continuous, whether prices have support, and whether ETH starts outperforming BTC. Stable BTC ETFs mean a stronger market base; improved ETH ETFs mean risk appetite truly spreads to the on-chain economy. ETFs are not an automatic bull market button; they are a ballot box for traditional funds. BTC has already secured a relatively clear vote; ETH is still fighting for a more complex and resilient one. Latest analysis tracking Elon Musk's moves!
Recently, Musk's core actions have focused on two main directions: SpaceX aerospace propulsion and AI industry statements, directly impacting sentiment in the aerospace and storage sectors, and indirectly influencing the crypto market's related derivatives.
In aerospace, SpaceX just completed its 100th rocket launch this year, achieving the milestone of over 100 launches annually for three consecutive years. The debris from 13 Starship test flights has been recovered, with further testing planned; meanwhile, the market is speculating on a Tesla and SpaceX merger. Predictive platform data shows that the market's betting probability for a merger before the end of 2027 has reached 55%, fueling speculative trading and directly driving short-term trading sentiment for $xSPCX.
Regarding industry statements, Musk has repeatedly emphasized that storage chips are currently the biggest bottleneck for AI development. The storage demand growth driven by the AI computing power explosion far exceeds capacity expansion. This statement is the core emotional trigger behind the recent collective strength in the storage sector, boosting SanDisk and Micron, while crypto derivatives $xSNDK are simultaneously being heavily speculated on by investors. Additionally, he made an aggressive prediction that SpaceX's AI business revenue will surpass all traditional businesses like aerospace and Starlink by September this year, and that AI will account for 99% of SpaceX's valuation in five years, further reinforcing the market's long-term narrative expectations for AI computing power and the storage sector.
This article is for market review only and does not constitute any investment advice. $BTC $ETH Why is everything selling off today? 📉 The move looks less like a single shock and more like a layered, risk-off shift across markets. As of 18:48 UTC, the S&P 500 is down about 0.56%, the Nasdaq has fallen roughly 1.25%, and the Dow is holding up better at just -0.09%. The VIX ticked up from 15.19 to about 15.66. The pattern is clear: money is rotating out of high-duration tech and semiconductor names, while defensive and value sectors in the Dow remain relatively resilient. Here’s what’s drivUS Treasury 5.2% Fully Inverted DeFi Rates: Risk-Free Yields Siphon Crypto-Native Capital, How Should On-Chain Liquidity Break Through?
As the US 30-year Treasury yield stubbornly remains above 5.2%, tokenized real-world assets (RWA) like BlackRock BUIDL and Ondo provide on-chain funds with about 5% "risk-free yield" backed by the US Treasury's hard floor. However, in traditional decentralized lending protocols on Ethereum such as Aave and Compound, the native annual percentage yield (APY) for stablecoin lending has long hovered at a low 2.5% to 3.5%.
The yield inversion of up to 150 to 250 basis points has directly triggered a large-scale outward migration of crypto-native capital.
During the bull market frenzy, DeFi could easily generate inflated annual yields of 10% or more through leverage nesting and governance token subsidies; but as the market enters a deep and volatile phase, speculative leverage demand has sharply declined, exposing DeFi's lack of real economic self-sustainability.
Faced with 2.8% crypto-native lending that carries smart contract black swan risks on one side, and 5.0% tokenized US Treasuries with sovereign credit backing on the other, any rational institution or whale capital would not hesitate to "vote with their feet."
This institutional liquidity siphoning brings an extremely profound test of authenticity to the entire Web3 financial ecosystem:
First, pseudo-high-yield protocols relying solely on token inflation subsidies accelerate liquidation.
Liquidity pools without real business income, maintaining yield illusions only by issuing counterfeit tokens, instantly lose appeal against the hard 5% opportunity cost of US Treasuries, causing TVL to plummet.
Second, DeFi protocols are forced to fully pivot towards "real yield" and "RWA collateralization."
Leading protocols like MakerDAO (Sky Protocol) and Aave no longer resist traditional finance but actively introduce tokenized Treasuries as collateral, converting the risk-free yield of US Treasuries into stable protocol dividends. The survival path of decentralized finance is evolving from early offshore confrontation to efficient on-chain transmission of sovereign risk-free rates and interest rate arbitrage.
When off-chain risk-free yields become an unavoidable gravitational benchmark for crypto assets, DeFi native finance must prove its unique productivity in capital allocation efficiency and global settlement speed, surpassing traditional banking.
Facing the 5.2% US Treasury risk-free yield and 3% DeFi native rate inversion, in your on-chain asset allocation, would you choose to hold RWA tokenized Treasuries for steady interest, or stick with crypto-native lending awaiting market leverage recovery? Do you think this yield inversion represents traditional finance's dimensionality reduction absorption of the crypto world, or an inevitable metamorphosis towards DeFi maturity?
---
The above content represents personal views only and does not constitute any investment advice. DYOR, NFA.
#30年期美债收益率创2007年以来新高 GRVT is up +6.0% in 24 hours after a slight pullback over 4 hours. Following an impulsive surge 24 hours ago, activity has increased — volumes have grown by 2.78 times.
The price at 0.2995 has approached the nearest resistance at 0.30429. The nearest support is at 0.29106, which has not been broken yet.The culprit behind the tech stock plunge: a global bond market sell-off frenzy!
US Treasury yields have surged to the highest level since 2007, with long-term bonds in Europe, the US, and Japan all rallying, pushing global risk-free rates continuously higher.
Two key market factors:
1. Tensions in the Middle East and firm oil prices are driving inflation rebound, making high interest rates unlikely to fall in the short term.
2. AI has become a top capital vacuum! Tech giants like Google and Oracle are aggressively issuing bonds to expand computing power, with corporate bond yields reaching 7%-8%.
Market logic has completely reversed:
Government and corporate bonds offer a steady 5%+ risk-free return, causing high-valuation tech stocks to lose their capital appeal entirely, leading to continuous capital outflows. This is the core reason for the tech stock crash!
Sustained high interest rates continue to suppress market valuations, leaving the market outlook unresolved in the short term, awaiting easing of Middle East tensions and cooling of the AI financing boom.
Are tech stocks currently bottoming and stabilizing, or is this a continuation of the downtrend? Will you choose to buy the dip or wait and see?
#30年期美债收益率创2007年以来新高 #闪迪回落逾9%，存储估值分歧加剧 #OKX预言家第二季正式上线 $GOOGL $ORCL $SNDK Despite BTC's rebound, short position liquidations have emerged prominently across the market. Is this a signal that the derivatives market risk moves ahead of the spot price trend? BTC recovered to $60,000 and ETH partially recouped losses, but market position adjustments are not yet complete. The key event this week was Strategy selling 1,638 BTC. The funds were used to replenish USD reserves and redeem preferred shares, which is interpreted more as balance sheet management than abandoning the holding strategy. Although the scale itself is not large, structurally, the signal of selling by a major holder prompted leveraged positions on the buy-the-dip side to react first. On the policy front, World Liberty obtained a conditional trust bank license, and BitMine related to ETH continued buying, with holdings approaching 5% of circulating supply. However, the fact that these positive news items were not immediately reflected in prices indicates the market is focusing more on liquidation risks and position realignment than on the news itself. This adjustment's BTC is hovering around 64K, ETH hovering at 1.9K, but what's really interesting is the underlying current beneath the surface calm. Have you noticed that recently AI and US stocks have been heating up, yet the crypto world still seems to be waiting for people in place? On the surface, both Big Bing and Second Bing are waiting for the FOMC to give direction, like the calm before a storm. But while watching the market, I feel the market isn't directionless; rather, it's using "sideways movement" to digest something people don't want to openly address—risk appetite is being diverted. - The profit-making effect of the US AI sector is so strong that capital would rather chase the highs than rush back to the crypto world to buy the dip. - The support level was indeed held, but the rebound was weak; every rally felt like someone was holding its head. The core battle of this round isn't whether BTC can break above 65K, but whether the crypto world can absorb the spillover of sentiment when there's any stir in the US market. My understanding is this: now it's more like a "halftime" in a trend continuation rather than a reversal. The path for bulls is simple: hold the support, wait for the FOMC to take effect, and risk appetite to return, so funds will naturally flow back. But the short-selling risk is also obvious—if the US AI market adjusts at high levels, the crypto community will not only fail to receive overflow funds but may even be sold off as "risk assets." The strength of the sector is now clear: AI is the strongest, mainstream coins are neutral, and altcoins are weak. In this pattern, chasing the rise is the easiest way to get hit. - For those already in position, holding on is better than just messing around$SNDK Even SK Hynix's massive buyback can't save it, the bears' opportunity has arrived
SanDisk violently rebounded just after a 9% plunge overnight. Big news came out from Korea that SK Hynix approved a 40 trillion KRW buyback plan
This is actually good news for SanDisk bears
SK Hynix's stock price fell from 2.98 million to 1.5 million, nearly halving, and the leader was forced to dig into its own pockets to support the market, which is why the buyback happened!
SanDisk rose 76% in the past half month. Profit-taking piled up like a mountain, one big bearish candle can't clean it out
Even a buyback at SK Hynix's level couldn't fully lift the sector, which precisely indicates that funds are using the good news to sell off.
Tonight when the US stock market opens, there will definitely be another big pull-down #闪迪回落逾9%，存储估值分歧加剧 A bookkeeping error caused 2,500 BTC to be sent into a customer's wallet.
This sounds like a joke, but it actually went to court.
The Singapore International Commercial Court just issued a freezing order, locking Bitcoin and USDC worth about 75 million SGD. The plaintiff is the operator of a large global crypto exchange, and the defendant is one of its long-term customers. The dispute isn't about fees or contract liquidations, but a bookkeeping error made by the platform itself.
The timeline goes back to March 2020. This customer had 2,500 BTC and 2,500 BCH in a dedicated wallet on the platform, and that month he transferred the coins out. The problem was on the platform's side: due to technical reasons, the internal ledger did not record this transfer, so the wallet appeared empty in the system.
An empty wallet sat on the books for over four years without anyone noticing anything unusual.
By July 2024, the platform, relying on its ledger records, believed the customer's assets were still there, so it transferred another 2,500 BTC and 2,500 BCH to another wallet of his. The same amount, duplicated out of thin air.
The customer acted quickly. He exchanged 20 BTC for about 816,000 USDC, then between July and November 2024, transferred those along with 780 BTC to wallets outside the platform's custody.
The platform only discovered the ledger error in January 2025, then froze the customer's wallet and recovered the remaining 1,700 BTC and 2,500 BCH. The 780 BTC and 816,000 USDC that left can only be reclaimed through litigation. At the current price of over $64,000 per BTC, 780 BTC is roughly $50 million.
The court granted the plaintiff a temporary injunction prohibiting the defendant from disposing of these assets and any related gains, and also required disclosure of the funds' whereabouts. However, the judge added a restriction that the plaintiff cannot use this disclosure to seek the same injunction in other jurisdictions.
What stands out most here isn't the 780 coins, but the four years. Four years during which a top platform's internal ledger and the on-chain reality did not match, and it wasn't discovered through self-audit but only exposed when the platform sent out the coins again.
When we usually talk about security, we focus on private keys, phishing, and contract vulnerabilities. This time, the failure was in the most basic bookkeeping.
If one day the platform accidentally transfers you extra coins, would you contact customer service first or transfer them away immediately? Retail investors have left halfway, but institutions are quietly moving money on the XRP chain.
Blockworks just released the XRP Q2 report, and the numbers are a bit split.
Let's look at the ugly half first. XRP closed the quarter at $1.04, down 19.9% for the quarter, with a market cap of $65.8 billion, ranking fourth among non-stablecoin assets. Spot trading volume on centralized exchanges was $57.6 billion, a quarter-on-quarter drop of 53.5%, and perpetual contract volume also fell by 44%. On-chain activity is also quiet: total transactions were 222.4 million, down 6.5%; daily active addresses were 16,800, down 10.7%; on-chain DEX volume was 482.9 million, down 35.9%; and the average transaction fee dropped to $0.00024, declining for five consecutive quarters. Fewer people, less activity—this is a typical bear market pattern.
Now look at the other half, which is completely inconsistent. Native stablecoin supply on XRPL rose 195.4% quarter-on-quarter to $825.5 million, with Ripple's own RLUSD accounting for $676.9 million, up 257% for the quarter. Stablecoin transfer volume increased 207.5% quarter-on-quarter, reaching about $10 billion, 90% of which was moved by RLUSD. Tokenized RWA (Real World Assets) value doubled quarter-on-quarter to $4.46 billion, setting a new quarterly high, with about half of that from Justoken's energy-related JMWH.
The money movers are not anonymous whales. Ondo, JPMorgan's Kinexys, Mastercard, and Ripple together completed cross-chain redemptions of U.S. Treasuries; UK's Aviva Investors then issued tokenized fund shares on XRPL; OKX launched over 280 spot trading pairs with RLUSD; Japan's Financial Services Agency recognized RLUSD as an electronic payment instrument, delegating issuance to SBI VC Trade. The ETP line also saw net inflows for three consecutive quarters, with $253.6 million entering in Q2, totaling over $1.9 billion.
The most interesting part is that the native on-chain segment is actually stuck. Lending proposals XLS-65 and XLS-66 received only 9 and 8 votes respectively out of 35 validators, falling 29 votes short of the activation threshold, so they must be revised and resubmitted in Q3. In other words, what’s really running on this chain isn’t native DeFi, but institutions moving their own assets in.
I've been staring at these two sets of data for a while. People are leaving, money is coming in; speculators are gone, business operators are arriving. Previously, we judged a chain’s success by transaction volume and address count, but this metric seems insufficient now.
So what do you think? Is a chain with decreasing daily active users but increasing institutional assets getting better, or is it changing its nature? Two addresses minted one billion tokens out of thin air, but they're only worth one million dollars
At 4:41 PM today, Paidun issued a warning. Two on-chain addresses each minted about 463 million MemeCore tokens, totaling nearly one billion tokens. These tokens were then transferred and consolidated into the same address. That address now holds this pile of chips, which is worth about $1 million at the current price.
Setting aside who did it, the numbers themselves are very abnormal. One billion tokens correspond to only $1 million, less than one cent per token. And MemeCore's M was never at this price level during its peak. So what exactly are these one billion tokens? Are they from another contract with the same name, a shadow pool with such thin liquidity that it barely trades, or was there never an intention to sell at the main token's price? Only the minter knows.
What concerns me more is the minting action itself. Two addresses, almost exactly the same amount, 463 million versus 463 million, looks like the same hand holding two keys opening two doors separately. After minting, they didn't stay put but consolidated into one wallet. This path is very familiar on-chain: first gather the tokens, then wait for a window to sell.
Placed in today's market context, this is even more interesting. Meme leaders across chains are collectively weakening; ANSEM dropped 30% in one day, CASHCAT's market cap fell below $100 million, and retail wallets are shrinking. Yet at such a time, someone on-chain is gathering one billion new tokens into one place. What do you think they are aiming for?
The most fragile part of the meme play has never been the price but the supply. The line on the chart looks good, but as long as someone holds the authority to mint tokens at any time, the so-called scarcity you buy is just a claim. Most people look at candlesticks, community heat, or whether a big influencer retweeted, but few check the contract to see who can still mint more, whether there's a cap, and who can change that cap.
So what we should really remember from this event is not the one billion tokens, but how many addresses hold the same button in the tokens we usually buy, just haven't pressed it yet.
Will you check the minting permissions of the tokens you hold, or have you never thought about this layer?Real-time Heat Ranking Data Analysis (August 19, 17:26)
The current heat ranking list is still firmly dominated by BTC and ETH in the top two positions, with their trading volumes leading by a wide margin, serving as the core foundation of market capital. The second tier of the list mainly consists of US stock-mapped derivatives, with xSNDK and xSPCX maintaining high heat. After the news of SK Hynix's massive buyback, discussions around storage-related targets have rapidly increased, and short-term speculative trading sentiment has heated up.
Most of the other popular tokens are rotating outdated meme coins, with GPS, BEAT, and $APR taking turns on the list. The hotspot shifts very quickly, with generally weak sustainability, as funds move from one place to another. Many tokens have high heat but lack matching trading volume, relying solely on community sentiment to boost attention, with insufficient incremental capital entering.
Approaching the evening Federal Reserve meeting minutes, overall trading activity has declined, with a large amount of capital choosing to temporarily observe and not daring to open positions blindly. The list intuitively reflects the current market mainline: the storage sector has become the short-term topic center but has not yet developed into a mainline trend that can drive the entire market, still in a phase of internal rotation of existing funds.
Market dynamics are provided only as a reference for review and should not be directly used as a basis for judging price movements.
This article is only a market review and does not constitute any investment advice.Bitcoin futures volume has been overtaken by a stock
Yesterday evening, if you opened the futures leaderboard, you would see a somewhat incongruous phenomenon. The top spot was no longer Bitcoin, but a storage chip stock with the ticker SNDK. Even more striking, among the top fifteen contracts by volume, ten were traditional financial assets, meaning what we usually call U.S. stocks.
SNDK is SanDisk, a storage giant spun off from Western Digital. The fact that it has perpetual futures listed on Binance is already quite novel, but after going live, its trading heat directly pushed Bitcoin's USDT perpetual futures to the back. This indicates one thing: the active crowd on crypto exchanges is no longer solely focused on the crypto world.
In recent years, the core users of platforms like Binance were basically chasing Bitcoin, Ethereum, and various meme coins. Now, with the AI and semiconductor sectors booming, storage stocks like SK Hynix and SanDisk have been heavily speculated on, and capital naturally wants an entry point. Instead of opening a U.S. stock account and dealing with time zones and thresholds, it's more convenient to open a 20x leveraged SNDK futures contract directly on Binance.
Actually, this trend didn't start overnight. Perpetual contracts for companies like Changxin Technology and Unitree Robotics have long been listed on Binance. Whenever the related stocks move in the A-share or U.S. stock markets, the futures side follows the excitement. Crypto exchanges are quietly turning into a 24/7 U.S. stock night session, with the pool of underlying assets filled with names from traditional financial markets. What users want is simply a trading interface that allows them to get in anytime, with leverage, and operates around the clock. Whether the underlying is a coin or a stock seems less important now.
Behind this is actually an identity reversal. A place originally created for trading crypto assets is gradually taking on the appearance of a stock exchange. We used to say Binance was the home court of crypto, but now half of the top listings are Wall Street assets. Those old players shouting decentralization and rushing on-chain are quietly betting on traditional market ups and downs through futures.
Interestingly, this migration isn't always comfortable. Stock contracts come with real company earnings and financial reports, and their volatility logic is completely different from meme coins. When Bitcoin stagnates and bores people, everyone rushes to speculate on storage stocks, but once the semiconductor sector collectively plunges like today, the pain under high leverage is no lighter than in crypto. Someone joked that Binance now looks more like a U.S. stock broker disguised in crypto clothing, only the risk warnings have to be read at crypto speed.
So the question arises: as crypto exchanges increasingly resemble U.S. stock night sessions, is the crypto world we hold still in its original place? Next time you open the futures page, is the ticker you’re watching really what you want to trade? Amazon quietly equipped AI with a wallet that can spend money on its own
AWS announced today that Amazon Bedrock AgentCore payments are officially fully available. The partners are clearly stated: one is Coinbase, the other is Stripe. With just a few lines of code, an AI agent can pay for calling others' APIs, MCP services, and content services by itself. The wallet uses the Coinbase and Stripe Privy system, and the settlement explicitly supports small payments in USDC.
The process is even simpler than imagined. You first fund the agent's wallet, either by credit card or directly with USDC, then give it spending authorization. It supports x402 and MPP protocols, and has added a new payment scheme called "upto," which sets no fixed price but only an upper limit; the actual cost is calculated based on the reasoning consumption. Each interaction has a single spending limit and expiration time; if exceeded, payment is refused outright. There is also an Observability panel in the backend, showing transaction audit logs, success rates, and average transaction amounts.
More importantly, this is not a demo. The related capabilities have already been integrated into CDNs like CloudFront and Cloudflare. Customers such as Anchor Browser, SpreadX, and Travala are already using it. Multiple regions are open, and developers can access it from the console, CLI, or even open-source frameworks like LangGraph.
My first reaction to this news was somewhat complex. Over the past few years, we've been debating the killer use case for stablecoins—cross-border remittances, payroll, providing accounts for those who can't open USD accounts—all discussed for six or seven years with fluctuating interest. The first truly essential scenario with instant settlement, unattended operation, and possibly only a few cents per transaction might not even involve humans as the main actors.
Humans can just use credit cards; machines cannot. If an agent wants to call a paid API at 3 a.m., it has no ID, can't wait three days for settlement, and can't call customer service. So what it needs are exactly the features that stablecoins have been praised for over the years. In July, the total market cap of stablecoins was still declining with net outflows for several months, and many said the narrative was cooling off. Yet Amazon has turned it into the default settlement layer for AI calling external services.
But this introduces a brand-new risk. Previously, wallet issues were due to private key leaks; now there's another path—what if the agent is tricked by a prompt injection? It holds the quota you gave it, and a malicious page tells it to pay to this address to unlock content—will it pay? AWS uses single transaction limits, session expiration, and audit logs as safeguards, but these only cover the scale of losses, not the error in judgment itself.
So I'm quite interested in hearing your thoughts. If tomorrow you could give your AI assistant a USDC wallet with a fixed monthly quota to buy data, rent computing power, book flights on its own, how much would you dare to give it? Looking further ahead, when machine-to-machine transfers exceed human-to-human transfers, will the blockchain's block space and fees need to be repriced for these users who never sleep? He shorted correctly on the direction but might be slowly killed by interest fees
Yushu today surged 463% on the A-share market, with a turnover exceeding 20 billion yuan in one day, and the stock price reported at 850 yuan. The market's focus is all on how much was earned from the new stock issuance, but the real dramatic scene happened on-chain.
There is an address called 0xfe4 that hit the peak on Yushu-related perpetual contracts. He shorted with 10x leverage, opening a position at an average price of $125.04, with a position value of about $728,000. The price did indeed go down later, and the unrealized profit once reached $58,000. The direction was right, and the timing of entry was quite precise.
The problem lies in the money he didn’t calculate. The current hourly funding rate for this contract is -0.3544%, negative, meaning shorts have to pay longs. Multiplying this rate by his position equals $2,579 per hour, about $60,000 per day.
At the time of writing, he has already paid $5,448 in funding fees. After deducting this part, the unrealized profit left is about $55,000. If this rate continues statically, in less than a day, all the money he earned will be eaten up by funding fees. Beyond that, even if the price doesn’t move, the holding cost alone is enough to push him toward liquidation. His liquidation price is $130.44.
This is the most absurd part. He can correctly judge the direction and wait for the price to drop, but as long as the sentiment gap between longs and shorts isn’t repaired, that hourly deducted amount will gradually erode his principal. The market doesn’t need to slap him in the face; it just needs to make him wait.
Binance also took action on the same day. They changed the funding rate settlement cycle for the UNITREEUSDT perpetual contract from once every 8 hours to once every 4 hours, effective from 8:15 UTC, with settlement points at 0:00, 4:00, 8:00, 12:00, 16:00, and 20:00 daily. The single settlement upper and lower limits remain ±2%. Worth mentioning is that this contract does not apply the rule of automatically shrinking to hourly settlement after the funding rate hits the upper or lower limit.
More frequent settlements mean extreme funding rates are realized faster. Previously, it was settled three times a day; now it’s six times a day, so whether a position can hold or not will be revealed sooner. Exchanges adjust this parameter not to favor anyone but to reduce their own risk and push out positions that shouldn’t remain in the market earlier.
Actually, this script is not new. The same happened when Changxin went public; before the market opened, on-chain contracts pushed expectations extremely high, funding rates were driven to extremes, and the largest long position took profits and exited after the first day’s high opening. This time, Yushu replayed the same story, but the side opposite the funding rate is now the short side.
The original intention of making an unlisted company into a perpetual contract on-chain is to allow everyone to price it in advance. But once leverage and funding fees are added to pricing, it becomes another game. You’re not just betting on how much the company is worth, but also on how many hours you can hold on.
The problem remains here. A position with the correct directional judgment and a liquidation price not yet touched may ultimately disappear because it can’t afford the interest. Do you think this way of winning and losing counts as market efficiency or market absurdity?People who have been shouting for years about $100,000 are now saying to wait another twenty months.
A person who has long been a supporter of Bitcoin on TV has changed his tune today in front of CNBC cameras. SkyBridge founder Scaramucci directly said that Bitcoin is now clearly in a bear market.
When this comes from him, it carries a different weight. Over the years, he has been one of the most willing on Wall Street to speak up for Bitcoin on camera, with positions in his fund, and when the market looked bad, he was the one to come out and soothe the market. Now he himself is the first to say the words "bear market."
What’s more concerning is the timeline he gave. He said the $100,000 target in his mind still stands, but the next catalyst might take twenty months. What does twenty months mean? Counting from now, it’s roughly spring 2028. In other words, according to him, the rest of this year plus the whole of next year will probably be a period of choppy, sideways movement.
He gave three reasons, each quite piercing.
First, the miners. We’ve all seen the activity at mining farms these past two years: Riot signed a 20-year lease with Anthropic, HIVE signed a $350 million GPU cloud contract, one after another freeing up rigs to run AI. Hash price has dropped from $63 down to just over $30, mining’s marginal profits are thinning, and renting out computing power is really lucrative. The group that was once most committed to pouring money into the network is now drifting away.
Next, where the money is going. Capital hasn’t disappeared; it’s just moved elsewhere. Recently, communication, storage, and cloud services stocks in the US market have collectively plunged, 30-year US Treasury yields hit a new high since 2007, yet even amid this turmoil, funds would rather take a hit in the AI narrative than look back at Bitcoin.
Finally, the cycle position. Whether you believe in the four-year clock or not, many still arrange their positions according to it.
What’s really interesting is what he said in the latter part. He said this round has dropped about 55% from the peak, whereas previous rounds were more like 75% to 80%. In his view, this isn’t a bad thing; it’s actually a positive signal. The price not falling further means there are more buyers at the bottom than before, a group of net buyers slowly accumulating chips for the next round.
This statement does align with some data. On Monday, Bitcoin spot ETFs saw nearly $300 million in net inflows, and on Tuesday another $189 million came in, with BlackRock’s IBIT alone taking $144 million. The price is hovering around $64,000, but money is flowing in.
There’s another detail worth pondering. He mentioned that in the past three months, Bitcoin only outperformed the S&P on about one-third of trading days, but on Monday and Tuesday this week, it outperformed the major US stock indices two days in a row. That sounds like good news, but from another angle, since US stocks were falling those days, outperforming means it fell less.
So the current situation is quite awkward. On one side, the most eloquent bulls are starting to change their tune, pushing hopes twenty months into the future. On the other side, ETF money is still steadily flowing in, the price neither crashes nor soars, just stuck there.
I want to ask, if even someone like him has to change his tune and say wait another twenty months, then those who have been holding on stubbornly—are they holding on to this price, or to their original judgment? The MU contract, which once piled up 300 million in leverage, is collapsing
The open interest of the MU-USDC perpetual contract on Hyperliquid has quietly dropped to about $138 million in the past two days, hitting the lowest point in two and a half months. Back in July this year, the open interest of this contract was close to $300 million. In just over a month, the leverage bet on it has shrunk by more than 60%. MU is a token with considerable attention in the Hyperliquid ecosystem; as soon as the perpetual contract launched, it attracted a lot of high-leverage funds. When it surged to nearly $300 million in July, it looked like a very strong consensus.
What’s more worth watching is not the number itself, but how it dropped. In the past 24 hours, MU’s price fell nearly 7% in a single day, and the liquidation records show only long positions being liquidated—no shorts were liquidated at all. The address Kvyadav1 had long positions liquidated worth $77,700; the one starting with 0x4a was liquidated for $55,900; and the 0x59 address was liquidated for $55,500 just four hours ago. All longs, no shorts.
This indicates that the previously piled-up leveraged long positions are being passively cleared out bit by bit by the market. Experienced contract traders know that this kind of structure means the market is still very uneasy. When the price drops, longs can’t hold and get forcibly liquidated first; forced liquidations create more selling pressure, which pushes the price down further, triggering the next wave of long liquidations. Without short liquidations to hedge, the entire clearing process is almost one-sided.
There’s another detail that’s a bit worrying. The current positions are actually quite concentrated. The top 1 address holds about $9.65 million, accounting for 7%; the top five combined hold about $35.92 million, or 26%; and the top ten together hold $54.36 million, nearly 40%. In other words, if someone wants to dump and liquidity is thin, the moves of a few large addresses can mess up the price badly.
Low liquidity combined with passive long liquidations is inherently the most fragile combination. Monitors also point out that under this structure, downside risk is often amplified, so in the short term, it’s necessary to closely watch the subsequent moves of the top holding addresses. This pattern of one-sided long liquidations with zero short liquidations is often not just a normal pullback but a signal of leverage moving from crowded to a stampede.
There’s a common misconception in our circle that if a coin’s top perpetual contract and open interest surge, it means strong consensus and optimistic funds. On the contrary, high open interest itself is also a source of high risk. The thicker the leverage pile, the faster the stampede comes. This round of MU is a vivid footnote to this old truth.
What we should watch most now is what those top addresses will do next. Will they continue to hold hard, or start to withdraw? With such concentrated positions, the answer might not be gentle. I just checked the market over the past few days, and honestly, I don't know whether to call it strong or weak. Maya Protocol first gave everyone a lesson. Yesterday, six vulnerabilities were exploited in successive attacks, and nearly $1.65 million in assets were stolen. As a result, CACAO crashed nearly 90%, and the total market value of the entire pool dropped by $11 million at one point. This is the most frustrating aspect of DeFi. The hacker took 1.65 million, and in the end, the market evaporated by 11 million. The remaining money didn't just disappear into thin air; it was basically the funds in the pool that were swept away along with the coin price. But right after the explosion, BTC was still hovering around $64,250, while SOL rose 2%, and ETH climbed back above $1,900. Even stranger is that in South Korea, Samsung and SK Hynix both fell by more than 7%. Chip stocks have dropped this much, yet the crypto market is still rising, at least indicating that the people buying crypto during this period are thinking differently from the traditional market funds. Policies are also being pushed forward at this time. The SEC has released a draft "Regulation Crypto," but the details haven't been fully disclosed yet, but the market is already speculating whether they want to open a token safe harbor or something similar. The CLARITY Act is still being developed. So when you look at these things together, you'll find that things are quite contradictory now. There was plenty of bad news, but the price barely responded. I actually think this is more worth watching than just a simple rise or fall. When does the market become less sensitive to negative news, it often becomes more sensitive than a few points of gains?Prediction markets rejected by the White House are being moved on-chain
At the White House crypto conference, the organizers quietly withdrew invitations for Polymarket and Kalshi. The reason was that these two prediction market platforms were considered too sensitive and unsuitable for official occasions. Yet, in the same week this issue was brewing, another clue quietly unfolded on-chain, this time starring Hyperliquid.
Hyperliquid’s core contributor Xulian has recently been soliciting feedback from the community through public channels on a testnet template called HIP-4. The name might seem unremarkable, but its purpose is significant: it aims to allow anyone to directly deploy prediction markets on Hyperliquid without needing to apply, get approval, or wait for anyone’s nod. Xulian is not the first to want to create permissionless prediction markets, but Hyperliquid, with its traffic and depth, is a different league from those smaller platforms of the past.
Currently, there are sixteen template records on the testnet, with about twelve latest templates after removing old versions, covering seven types of market structures. Some can open markets settled by expiration price, some can open price-triggered markets, and there are even sports binary options and win-draw-lose types of gameplay. From stock prices to sports games, from central bank interest rate decisions to geopolitical events, theoretically, any of these can be made into an on-chain bet that anyone can create and anyone can bet on.
But this door hasn’t truly opened yet. The full process requires community feedback, validator review, and on-chain voting. Only templates that pass these steps will be written into the mainnet for deployers to use. However, the mainnet’s outcomeTemplates and deployer list are still empty, and permissionless templates have not officially launched. In other words, the rules are being tested, and the community is waiting.
The biggest difference between this mechanism and traditional prediction markets lies in one word: permissionless. No matter how popular Polymarket is, it still faces regulatory and regional restrictions; no matter how compliant Kalshi is, it cannot escape being removed from official lists. Hyperliquid takes a different path by handing over the power to open markets directly to code and the community.
Interestingly, Hyperliquid originally made its name by turning US stock indices into perpetual contracts. Now, extending this idea to prediction markets means that on-chain trading is not just about asset prices but also about trading any event that hasn’t happened yet. Whether the market is willing to pay for this depends on when the mainnet truly opens and whether templates can pass validator review.
But the problem is also obvious. Prediction markets naturally operate in a regulatory gray area; the more permissionless they are, the easier they can be used to bet on sensitive events or even as tools for information manipulation. Polymarket was targeted by regulators precisely because its bets crossed that line. Hyperliquid opens the door wide, bringing excitement but also attracting controversy. Some in the community are already speculating on what the first batch will be once the mainnet opens. Some bet on elections, some on token unlocks, and some even want to turn influencer scandals into bets. The heat is never lacking; what’s missing are boundaries.
We are all guessing what the next prediction market to be moved on-chain will be. The next election, the next interest rate decision, or some black swan event that hasn’t happened yet. When the threshold to open a market is so low that anyone can just click, the real concern might not be that no one plays, but that too many play and it becomes uncontrollable. When regulators try to shut things down, they probably never imagined the chain would fix its own gate. But once this gate is open, it’s hard to close it.Various meme coin leaders collectively plunged overnight
Last night and this morning, the hottest batch of meme coins in the crypto circle suddenly turned sour together. According to GMGN market data, ANSEM, the leader in the Solana ecosystem, has crashed from yesterday's high point, dropping more than 30%, with its market cap now down to around $227 million. On the BSC side, MarsCoin, which was once hyped to the sky, also broke below the platform level it had maintained for days, falling 12% in one day. The worst hit is CASHCAT on the Robinhood chain, which directly fell below the $100 million mark, now only at $89.37 million, down 14.61% in 24 hours. Three leaders in one sector all dropping at the same time like this is no longer just a problem with individual projects.
Interestingly, just a few days ago, these names were the hottest wealth codes in various groups. When the meme coin from Niulai Movie just launched on BSC, the issuer didn’t earn a penny from the main coin; the real winners were professional market makers and bots, while retail investors took the losses. ANSEM rode on the halo of a namesake KOL, CASHCAT leveraged the Robinhood listing narrative, and MarsCoin was repeatedly pumped on the coin-stock concept. Back then, everyone was looking for the next 100x, but now, the leaders are the first to buckle.
This collective plunge is not an isolated event. Bitcoin’s volatility has been suppressed these past two days, but retail demand has surged to the highest in nearly two years. Analyst Darkfrost directly pointed out the risk of a local peak. Money is still flowing in, but prices are starting to fall; this divergence is the easiest way to crush those chasing highs. More importantly, meme coins have cyclical attributes—they never look at fundamentals and are purely driven by sentiment relay. Once the top projects can’t hold, it’s not hard to imagine what fate awaits the smaller meme coins behind them.
Looking at a longer timeline, this meme craze has a clear difference from the last round. Previously, retail investors used real money to chase dog coins, but now there are platform tools: Pump.fun for issuing coins, GMGN for copy trading, and KOLs shouting orders form a production line. The previous article about selling shovels mentioned that platforms earn tens of millions of dollars monthly while retail investors lose money on the secondary market; platforms and market makers make guaranteed profits. Now that the leaders are plunging, it’s like another blow to this playbook. When the traffic recedes, the first to be thrown out are always the most marginal chips, and even the big brothers are starting to wobble—can the little guys hold steady?
The question is for you: is this a short-term correction, or the visible end of this meme frenzy? Those still holding these coins, are you prepared to hold on to the end, or cut your losses and exit early? $SNDK storage cycle is currently in the "post-cycle" phase — price increases continue, but the pace has clearly slowed. DRAM and NAND inventories remain at historically low levels of only 2-5 weeks, and SK Hynix even warns of the "most severe storage shortage" by 2027.
AI relies on high-speed storage, but not only HBM. GPUs determine how fast AI runs, while storage determines how far AI can go. The massive data generated by large model inference requires a layered combination of HBM (extreme speed), high-performance SSDs (mid-speed large capacity), and regular hard drives (cold data archiving).
AI demand is spreading from GPUs to the entire storage ecosystem, with long-term large orders locking in shipments for the coming years.
Consumer electronics have already started to buckle under price hikes; after the most intense price increase phase passes, stock prices may peak early.
Fundamentals are still rising, but the market is already trading on "how much longer can it keep rising."
From the current public moves by storage companies, locking in long-term contract orders, buying back shares, and committing to return cash flow to shareholders essentially means the good news has been fully priced in. Is the altcoin season really "gone forever"? Let's uncover the truth by looking at the data
Comparing the total market cap of altcoins at the peak of the 2021 bull market with the 2025 cycle, you'll find a disruptive change in market structure:
1.) On the surface, the cake looks just as big, but after deducting "stablecoins," it actually shrinks. The 2021 TOTAL2 (excluding BTC) peak was about $1.72 trillion; in 2025, it’s still around $1.5 to $1.7 trillion. Seems unchanged? But stablecoins were only about $150 billion in 2021, and have ballooned to around $280 billion in 2025. After removing stablecoins, the actual funds left for altcoins in 2025 are actually less than in 2021.
2.) The cake hasn’t grown, but the number of "mouths" eating it has increased by tens of thousands. In 2021, this cake was divided among thousands of tokens; in 2025, the same size or even smaller cake is being crazily diluted by tens of thousands of Meme coins and a massive influx of new public chains with terrifying unlock volumes (Sui, Aptos, etc.). This explains why the total market cap hasn’t collapsed, but the prices of the vast majority of altcoins you hold have dropped by 80%.
3.) 2021 was a global liquidity flood, with retail funds buying BTC, overflowing into ETH, and then wildly rotating into altcoins, lifting everything. In 2025, Wall Street spot ETFs have become the absolute main force.
An absolute "altcoin season" may be hard to come by again, but localized bull markets never die out. $CFG currently shows continuous signs of capital outflow, with the core trigger still being the unknown variables brought by token-for-equity swaps.
Many people think that a reduced total token supply after equity swaps is positive, but there is actually another risk hidden here. Once a large number of tokens are exchanged for equity, the circulating supply in the market will directly decrease, causing liquidity to shrink rapidly. When liquidity dries up, whether the market goes up or down, the trend will become extremely volatile. Even a slight selling pressure can trigger large fluctuations, maximizing uncertainty. This is also a major reason why funds have recently chosen to exit and wait and see. #闪迪回落逾9%，存储估值分歧加剧 Analysis of the Strait of Hormuz Situation
The Strait of Hormuz handles nearly 20% of the world's crude oil maritime trade and is a vital energy lifeline globally. Currently, after the window for US-Iran negotiations has closed, both sides remain deadlocked. Commercial shipping volume through the channel remains low for an extended period, with many oil tankers opting to detour to avoid risks, significantly raising shipping insurance costs. Iran has set tough negotiation conditions such as lifting sanctions, while the US is unwilling to compromise, making a quick resolution unlikely in the short term. Geopolitical risks continue to hang over the market.
The first level of transmission is crude oil: if the conflict escalates and the shipping route is further obstructed, oil prices will surge rapidly, pushing up global inflation expectations. The market will again delay expectations for Federal Reserve rate cuts, which is negative for risk assets. If the situation remains as is without direct clashes, oil prices will likely stay in a high-level range-bound pattern, making a strong one-sided trend difficult.
The second level of transmission affects gold and crypto: when the conflict heats up quickly, gold, as a safe-haven asset, tends to strengthen in the short term; risk assets like BTC will initially face pressure and only gradually show an inflation-hedging narrative once inflation expectations are fully ignited. If the situation eases, safe-haven funds will quickly withdraw, causing gold to fall and relieving short-term pressure on the crypto market.
The market has gradually adapted to the current low shipping volume through the channel. Without sudden attacks, extreme market moves are unlikely, but it remains a potential black swan. Should sudden military friction occur, market volatility will spike instantly, and the crypto market will experience severe turbulence.
This article is for market review purposes only and does not constitute any investment advice. #成品油价差破百，能源通胀会否回升 $BTC $ETH $MU Micron 952 has returned to this position again. Among the three storage giants, SanDisk is the strongest, Hynix is second, and Micron is the weakest — this ranking can be seen from the magnitude of this pullback. SanDisk pulled back from 1800 to 1580, 12%; Hynix from 1269 to 1062, 16%; Micron from 1036 to 917, this magnitude is similar to SanDisk's, but it rose less during the previous rally. 😅
SAR=982.78 is pressing overhead, EMA21=966.94 has just been broken downward, EMA55=949.28 is supporting from below. The price is exactly on the 55-day moving average, this position is very similar to the structure when FIL pulled back to EMA55 before. KDJ's K=25.39, D=23.95, J=28.28, the sign of a low-level golden cross has appeared — K line crossing above D line, although the magnitude is not large, the direction is correct. RSI6=43.14, not oversold, but already far from the previous panic zone.
On-chain MU contract open interest has dropped to about $138 million, this data is very critical — indicating that leveraged players are accelerating their exit. The decline in open interest means previous longs are being passively cut and are handing over chips. When leverage is mostly cleaned out, it often signals the emergence of a phase bottom.
But the problem is — Micron as a target is inherently less attractive than SanDisk with its AI storage narrative. In the same storage sector, capital is willing to pay a higher premium for SanDisk, while it is relatively cautious on Micron. Will Micron hold 950? I choose to place an order around 940 to buy, with a stop loss at 910; if it reaches, I buy, if not, I keep watching. Comment below, who do you think among the three storage giants has the most potential to hit a new high first? Is it SanDisk? Hynix? Or do you think Micron will come from behind? Show your position and speak up, no objections accepted. 🔥SK Hynix 40 trillion KRW buyback, a solid major positive!!!
This full 40 trillion KRW buyback and cancellation is the largest share cancellation action in South Korea's history, sending two core signals that strongly catalyze the entire storage sector.
First, it proves that the HBM business is highly profitable, the company holds massive cash flow, and management believes the current stock price is seriously undervalued. They are willing to put up huge amounts of real money to support the stock, directly boosting market confidence in the storage leader's profit sustainability and easing recent market fears of a valuation peak in the sector. At the same time, the company raised its shareholder return target, planning to return over 50% of free cash flow to shareholders from 2025 to 2027, which will raise the sector's valuation logic in the long term.
Second, it emotionally drives the entire storage sector. When the US market opens tonight, this positive news will likely spread to SanDisk, Micron, and Western Digital, bringing short-term emotional pulses. The crypto derivative $xSNDK will also follow the sentiment of the US stock underlying shares. But it should be clear that buybacks are a capital-level positive and will not change the spot supply-demand rhythm of storage chips, nor rewrite the industry's long-term fundamentals.
It is worth noting that the Korean stock market still fell nearly 10% that day, indicating huge divergence among on-site funds. Positive news does not equal a one-sided rise; the sustainability of the pulse market still depends on whether the AI storage shortage narrative can continue to ferment. Market dynamics are only for review reference and cannot be directly used as a basis for judging price rises or falls.
This article is only a market review and does not constitute any investment advice. $BTC $ETH $SNDK Yushi worthy of it? The first humanoid robot stock?
UBTech dropped 80% after going public. Tell me, how is Yushi's story any different from theirs?
Can't say, right?
UBTech, the first humanoid robot stock on the Hong Kong market. At the time of listing, it was also "AI revolution," "embodied intelligence," "trillion-dollar track." Its market cap once surged to 100 billion. Then it dropped 80%.
The reason for the drop is clear now: revenue growth couldn't keep up with valuation expectations. 80% of companies in the robotics industry are losing money, large-scale commercial delivery hasn't been successful, and factories aren't buying.
Which of these problems does Yushi Technology's story solve?
A 60% gross margin is indeed higher than peers. It did turn profitable in 2025. But Q1 profits have already been halved. The company itself expects non-recurring net profit to decline 6% to 22% year-over-year in the second half. And you want to enter at a 219x PE.
More importantly, the float. At Yushi's IPO, the circulating shares accounted for only 7.44%. When Changxin Technology went public, the float was 6.73%, and it surged 466% on the first day, with a market cap reaching 3.3 trillion.
But have you thought about it? A low float amplifies both the rise and the fall. Zhipu's float is less than 4%, and a turnover of 31.1 billion drove a 400 billion market cap change. After the lock-up expired, it dropped 50% in three days.
You enjoy the new stock for three happy days.
On the fourth day, you become the one locked inside waiting for the lock-up to end and the sell-off to hit.SanDisk fell more than 9%, storage valuation divergence intensifies
This decline, in my view, should not be simply understood as the AI storage logic being disproven; it is more like after a surge, the market begins to re-examine the valuation.
SNDK closed down about 9% on Tuesday at $1625.78, while storage-related stocks like Micron and Seagate also weakened simultaneously, indicating this is not an issue unique to SanDisk but a profit-taking and valuation re-pricing across the entire storage sector. 
More importantly, SanDisk just presented very optimistic long-term targets at its investor day:
• FY2028-FY2030 revenue expected to maintain mid-to-high double-digit growth
• Adjusted gross margin target around 80%
• Free cash flow margin target around 50%
• Approximately $93.9 billion in long-term business agreements signed
• Long-term agreements expected to cover about two-thirds of bit output in FY2028. 
So the real debate in the market now is no longer:
"Does SanDisk have AI demand?"
But rather:
"How much of these great expectations have already been priced into the stock?"
Why does the stock price fall when good news comes out?
Because SNDK rose too fast earlier.
After the investor day, the market had a significant rally, with a cumulative increase over the past year exceeding 3000%. 
Under these circumstances, the trading logic of capital changes:
Before:
Earnings beat expectations → Buy
Now:
Earnings beat expectations → Not enough
Long-term targets are good → Still not enough
Must continue to exceed market expectations → Only then can the stock price continue to rise.
So this decline is essentially:
Fundamentals remain strong, but valuation tolerance is starting to decrease.
There is another factor to be cautious about:
The market is beginning to worry about the sustainability of AI capital expenditures and long-term financing models.
Recent discussions about large tech companies' AI investments, off-balance-sheet commitments, and high capital expenditures have heated up, putting valuation pressure on the AI hardware sector. Meanwhile, the sustained high yields on long-term U.S. Treasury bonds also suppress high-valuation growth stocks. 
This creates a short-term contradiction:
AI demand remains strong
VS
AI valuations are already expensive
Therefore, storage stocks are not entering a "fundamentals deteriorating" phase now, but rather:
Fundamentals and valuations are seeking a new balance.
What about SNDK going forward?
I would treat this decline as a very important stress test.
If next:
It falls → volume gradually shrinks → capital re-enters → SNDK regains key levels
That means this is just profit-taking, and after high-level chips change hands, the long-term trend remains healthy.
But if:
The rebound is weak → every rally is sold off → volume expands breaking important previous support → Micron, Hynix, and SanDisk weaken simultaneously
Then we need to be cautious that this storage rally is moving from an "uptrend" into a "high-valuation digestion" phase.
The core sentence:
This 9% drop in SNDK, I would not interpret as "the AI storage bull market ending."
A more accurate understanding is:
The investor day told SanDisk's long-term story too well, pushing market expectations to a very high level. Now the stock price is starting to test whether the company's real profit growth in the coming years can catch up with such a high valuation.
So next, don't just focus on how much it falls in one day.
What really matters is whether there is capital support after the drop.
If fundamentals don't change, long-term orders don't change, AI storage demand doesn't change, and the stock price can find support again after a significant pullback, then this may instead become a high-level turnover.
But if fundamentals start to show actual deterioration in orders, prices, or gross margins, then the nature is completely different.
The biggest risk now is not "no demand for storage," but "the market has already priced in the best-case scenario for the next few years." $BTC #闪迪回落逾9%，存储估值分歧加剧 A brief chat about SanDisk $SNDK. Don't get carried away and lose money on SanDisk.
The earnings report just beat expectations, with Q4 revenue at $8.97 billion, a 51% quarter-over-quarter increase, and Non-GAAP EPS of $39.25; the company's guidance for the next quarter is also stronger than market expectations, plus an additional $14 billion buyback, bringing the total remaining buyback authorization to $15.5 billion. Along with the long-term goals discussed at the investor day, projecting mid-to-high double-digit revenue growth and sustained high gross margins from 2028 to 2030, the market is buying into the logic of “AI storage + NAND market boom + long-term contract locked demand.”
The sentiment is typically both attractive and risky. A few days ago, due to the guidance and analysts raising expectations, the stock price surged sharply, reaching around $1800 on Monday, with a very exaggerated year-to-date gain. But on Tuesday, it dropped sharply along with AI hardware/semiconductors, indicating tight positioning; everyone knows the logic is good, but once market risk appetite cools, the sell-off can be swift. Simply put: the mid-to-long-term story remains, but short-term sentiment is unstable.
From a technical perspective, I think now is not a position to chase blindly. After the previous rally, there was a significant pullback, then a recent rebound, but it was quickly pushed down from a key level, indicating considerable selling pressure above. In the short term, watch the 1600/1500 area; if it breaks down with volume, further consolidation may be needed. First, look at the 1780-1800 area; only if it can stabilize there again can it be considered repaired, then reassess.
Personally, I prefer to observe, buy on dips, and avoid chasing highs. #闪迪回落逾9%，存储估值分歧加剧 The successful recovery of Zhuque-3, can it surpass SpaceX?
Today someone asked me if the recovery of China's Zhuque-3 is a major negative for $SPCX.
First of all, congratulations on the breakthrough and advancement of China's aerospace industry, but there is still a big gap compared to spcx, so don't even mention any impact.
First, if SpaceX had not boldly made this attempt proving the feasibility of this path, playing a pioneering and demonstrative role in the industry's technical route, China's aerospace industry would not have achieved such good results. So SpaceX is currently leading in technology.
Second, there is still a gap between Zhuque and Falcon. Falcon recovers, inspects, and relaunches, while currently Zhuque has only achieved full recovery.
Third, in terms of size, there is still a difference between the two. Also, SpaceX's technology is already very mature and commercialized, while Zhuque has just passed the hardest stage from zero to one.
But with China's mature supply chain and productivity, once the difficulties are overcome, mass production will be as simple as copy and paste. Once the technology is verified feasible, commercialization will accelerate ridiculously fast, but catching up indeed still requires time.
Since they dare to open source, it proves they have a trump card. This time it will not affect SPCX; on the contrary, it will generate heat for this narrative, proving the feasibility of this track. As the leader in this track, SPCX will be the focus of capital attention.
This pursuit by Zhuque-3 actually promotes global commercial aerospace competition along this route #贝莱德重申BTC仍具配置价值 8月19日，被称作A股人形机器人第一股的宇树科技正式登陆科创板，开盘直接冲到1100元，较150.8元的发行价暴涨629.44%，开盘瞬间总市值飙升至4449亿元，中一签最高浮盈超47万元，创下今年新股开盘涨幅新纪录。 疯狂的行情背后，市场的核心疑问已经摆在台面上：短期情绪炒出来的千亿市值，最终靠什么业绩去消化？ 一、为什么开盘直接暴涨超6倍 1、赛道稀缺溢价，A股唯一人形整机标的 在此之前，A股机器人板块大多集中在减速器、伺服电机等上游零部件。真正完成量产、已经实现盈利的人形机器人整机上市公司，宇树是第一家。 资金买的不单单是当下的利润，更是押注万亿人形机器人赛道的入场门票，直接把远期产业预期提前打到了股价里面。 2、基本面确实拿到了阶段性成果 公司已经走出一条“机器狗打底，人形机器人突围”的成长路径： • 四足机器狗全球累计出货超3.3万台，全球市占率大约60%，提供稳定现金流基本盘 • 2025年人形机器人G1全年出货5500台，人形业务收入占比51.78%，已经超过四足机器人，成为第一收入来源，毛利率超过61% • 2025年营收16.99亿，扣非净利润接近6亿元，是全球为数不【Storage stocks finally remember to unite the masses】
SNDK stated at the investor conference: 100% of excess cash flow will be returned to shareholders, mainly through buybacks. Currently, the board has authorized a stock buyback limit expanded to $15.5 billion.
SKHY promises: at least 50% of free cash flow will be used for shareholder returns, with a 40 trillion KRW buyback and cancellation, equivalent to buying back 3.3% of shares within 3 months, an epic positive development.
What about MU? They definitely want to release some good news too, but due to chip legislation restrictions, buybacks can only start in December.
Not worried about scarcity but about inequality~ It seems like a moral issue, but it’s actually about understanding the underlying rules.
$SNDK $SKHY $MU 今晚2点，美联储7月FOMC会议纪要即将公布，这是本周市场最关键的风险事件。 7月FOMC会议以9比3维持利率在3.50%至3.75%不变，连续第五次按兵不动。三名地区联储主席投下反对票主张加息25个基点，为2016年9月以来首次三票同时反对。美联储主席沃什在会后明确拒绝将本次决定称为“暂停”，称“这只是故事的开始”。他同时强调美联储不存在软性通胀目标，2%是唯一目标。 之后经济数据已全面走弱。7月非农仅增5.7万人，CPI同比降至3.4%，市场对9月加息的定价已从会后的65%至80%大幅回落至目前约30%至40%。路透调查中约九成经济学家预计9月维持利率不变。 今晚的核心，就是看纪要揭示的鹰派阵营规模有多大。 如果纪要显示，多数委员认同通胀回落、不急于加息，9月加息概率将继续被压低，BTC有望站稳65000并向上挑战。 如果纪要显示，内部对通胀黏性的担忧更深、更多人接近支持加息，结合30年期美债收益率已飙至2007年以来最高，以及布伦特原油因美伊停火协议到期未达成而重回94美元的现实，35%的加息概率可能重新回升，BTC可能回测62000甚至60000。 总之，在沃什就任后，大幅压缩$SNDK $MU storage chips collectively surged pre-market, catalyzed by a major positive from SK Hynix: a 40 trillion KRW buyback and cancellation + S&P upgrade to A-, with a commitment to return more than half of free cash flow, driving renewed capital interest in the storage sector.
The core logic behind this rally is clear: AI computing power boosts demand for HBM and enterprise storage, major companies increase shareholder returns, reshaping market valuation perceptions of the storage cycle, leading to synchronized gains for SanDisk and Micron.
However, caution is warranted: the sector has seen significant gains recently, bullish narratives have been fermenting for some time, and positive news often triggers profit-taking. Under extremely unanimous bullish sentiment, a long-short squeeze could occur at any time. Respect the trend and strictly maintain risk control boundaries in trading.
#USStocks #StorageChips #AISemiconductors #闪迪回落逾9%，存储估值分歧加剧 #宇树科技科创板首日开盘暴涨629%，高估值如何兑现？ #宇树科技科创板首日开盘暴涨629%，高估值如何兑现？
#成品油价差破百，能源通胀会否回升
#闪迪回落逾9%，存储估值分歧加剧
· Yushu Technology surged 629%: The A-share STAR Market's new energy/robotics wealth creation myth, but this is entirely a capital game, overdrawing short-term sentiment. As someone in the crypto circle, just watch and don't let this kind of “windfall” disrupt your rhythm.
· Refined oil price spread breaks 100: This is a macro-level hidden concern. If energy prices continue to rise, inflation will rebound, and expectations for Federal Reserve rate cuts will be further delayed, which is a continuous pressure on all high-volatility speculative markets (including crypto).
· SanDisk fell over 9%: The storage chip leader's sharp drop indicates the market's aversion to high-valuation tech stocks is still spreading. If U.S. tech stocks continue to correct, the crypto market will hardly remain unaffected.
· Do not rush to bottom-fish; continue to stay out and rest. The current trend is a “gradual decline” rather than a “sharp drop,” and actively catching falling knives carries great risk.
· Keep a close eye on Bitcoin (BTC) stabilization signals. No need to focus on individual altcoin movements; only when BTC stops shrinking volume in a downtrend and is accompanied by a clear inflow of ETF funds will I consider re-entering.
· This period is a good time to stay out and review your leverage and timing issues, patiently waiting for a macro-level turnaround. Staying alive is more important than anything else. Xiaomi’s latest Q2 financial report paints a more interesting picture than a simple “growth or decline” headline. The three major engines are moving at different speeds: Autos are accelerating. Phones are facing margin pressure. AIoT remains the stable cash-flow pillar. 🚗 Auto: Scale Is Starting to Work The SU7 series delivered 104,200 vehicles in Q2, crossing the 100,000-unit quarterly milestone. More importantly, auto gross margin reached 20.1%, while losses narrowed from roughly ¥3.1B in Q1 SK Hynix's massive 40 trillion won buyback announcement was a new catalyst for sentiment in the storage sector today, but it was not the core driver of the previous round of aggressive surges. Let's start with the timeline: Previously, the main rally in SanDisk's SNDK, Micron MU, and Western Digital started much earlier than this buyback news. The core logic of that rally was the market repricing the long-term shortage of AI storage, with companies actively controlling production to protect profits, institutions using growth stock logic to revalue the track, driven by industry fundamentals. This buyback is a newly announced announcement today, and is more of a short-term emotional stimulus. This buyback sends two key signals: First, SK Hynix holds massive cash flow, confirming strong profitability in its HBM business. Management is confident in the industry's long-term outlook and willing to allocate massive funds to reward shareholders, providing positive support for the overall memory sector's profit expectations; Second, large-scale buybacks and cancellations directly signal to the market that current stock prices are undervalued. In the short term, this will boost risk appetite across the entire storage sector. Tonight's US market opening could easily trigger short-term sentiment fluctuations among SanDisk, Micron, and Western Digital. But it's important to distinguish that buybacks are positive for stock prices and won't change the underlying industry logic of memory chip supply and demand, or the pace of price increases. The crypto derivative $xSNDK is merely a token of sentiment and has no equity interests; it only follows the sentiment fluctuations of US stock stocks. The sustainability of the news and the pulse market will ultimately depend on whether the industry narrative can sustain its momentum. On the risk side, attention should also be paid to SK during today's Korean stock trading sessionJeonbuk Bank’s reported move to use Ripple Payments for near-real-time corporate cross-border settlement is more significant as a test of institutional payment infrastructure than as a direct XRP signal. It would be Korea’s first regional-bank adoption and Ripple’s third Korean institutional deal this year, following work with KBank and Kyobo Life across custody and stablecoin payments.
The measured read: distribution is broadening, but token demand remains unproven. Until disclosures identify the settlement asset, launch timing and transaction scale, stablecoins or fiat could capture the flows without creating meaningful XRP liquidity demand. Not advice, just analysis.
#JeonbukBankAdoptsRippleBTC 15-minute cycle bullish outlook
Current price 64424. After a surge to 65036, there was a round of pullback, with a low retracement to 64016 followed by a bottom test and short-term rebound. The 15-minute level shows signals of recovery and strengthening.
Short-term bullish on BTC, but there is still strong trapped selling pressure above. The larger timeframe has not fully turned strong yet; this is a rebound repair, not a direct new round of sharp rally. Core of SEC New Regulation (Regulation Crypto Assets):
· Startup Exemption: Raise up to $5 million within 4 years, no financial reports required, public fundraising allowed, no resale restrictions on tokens.
· Financing Exemption: Two tiers (within 12 months $20 million / $75 million), Tier 2 requires audit.
· Safe Harbor: After fulfilling commitments, SEC recognizes tokens as no longer securities.
Significance: After FTX, new asset creation has dried up; the proposal aims to restart the "zero to one" incubation channel. Over 95% of projects may fail, RWA becomes mainstream, but native innovation needs to recover.
Path: Startup Exemption → Financing Exemption → Safe Harbor → Free circulation. If the Clarity Act passes, it will resolve secondary market regulation; otherwise, it remains a gray area. The proposal provides a regulatory path, but whether it can reduce speculation is uncertain.Analysis of Major Asian Market Closing Data and Cryptocurrency Market Correlation
Today's Asia-Pacific stock markets closed with clear divergence. South Korea's KOSPI plunged 5.80%, with the semiconductor sector sharply declining; storage leaders Samsung and SK Hynix fell heavily, directly impacting sentiment in the AI storage sector. The Nikkei 225 slightly rose 0.11%, with significant volatility, while the yen exchange rate remained stable. The Shanghai Composite Index edged up 0.12%, showing weak and volatile movement. The Hang Seng Index rose slightly by 0.09%, but the Hang Seng Tech Index dropped sharply by 1.21%, reflecting severe internal divergence in technology stocks, with Xiaomi rising and Baidu and others falling significantly.
Overall, Asian stock markets did not form a unified risk appetite. The sharp drop in the Korean market was mainly due to semiconductor valuation cuts, which in the short term transmitted sentiment to storage derivatives like $xSNDK; other indices showed limited volatility, making it difficult to drive the crypto market in a distinct direction. The direct correlation between the crypto market and Asia-Pacific stock markets is weak, and crypto funds are currently not focused on the Asian session stock markets.
The market's core focus remains on the upcoming Federal Reserve meeting minutes tonight. After the Asian session closes, funds have moved further into a wait-and-see mode, with short-term capital temporarily reducing active positions, awaiting macroeconomic news to determine direction. Market movements should only be used as a reference for review and not directly as a basis for judging price rises or falls.
This article is for market review only and does not constitute any investment advice. $BTC $ETH $OKB $SNDK SanDisk Market Review
Overnight late session saw a slight surge to lure buyers, creating a false impression of continued strength in the market, but today's early session reversed sharply to weakness.
The open near 1677 quickly faced pressure and dropped, hitting a low at the 1600 level. The sentiment boost driven by yesterday's institutional research report target price of 2000 completely failed, short-term funds collectively took profits, volume expanded on the early session decline, and bearish selling pressure was released in concentration.
Core logic behind this rapid correction
1. On the macro side, US Treasury yields continue to rise, suppressing valuations across the US tech growth sector; the storage sector collectively weakened, directly dragging SanDisk down in sync.
2. The volume throughout yesterday's rally continuously shrank; it was purely short-term speculation driven by research report sentiment, with no long-term incremental funds supporting. After the sentiment faded, momentum-driven bulls quickly stampeded to exit.
3. The stock price had already risen continuously earlier, fully pricing in the market expectation of AI-driven storage price increases, accumulating a large amount of profit-taking chips at high levels. The market is extremely fragile and any slight disturbance can trigger concentrated selling.
Key short-term ranges
Resistance at 1683 above is strong and difficult to hold above in the short term;
Support at 1600 below is the first defense line; if volume expands and it breaks below in the early session, it will further test the 1595 cost entry zone, opening more room for bears.
Positioning mindset sharing
Previously, the entire network was unanimously bullish and many were bearish on my short positions, but now the market itself has given the answer.
#闪迪回落逾9%，存储估值分歧加剧 $SNDK $ETH rises to around $1920: Why is "price increase + OI decrease" worth studying?
$BTC, $SNDK
ETH is currently around $1,920, not far from the psychological resistance at $2,000. A phenomenon in the recent derivatives market worth reviewing is that ETH open interest had previously declined, but the funding rate remained positive, indicating that leveraged funds did not form a consistent new long position.
When trading contracts, you can look at price and OI together:
Price up + OI up: new positions entering, trend may be strengthened.
Price up + OI down: more likely driven by short covering, not necessarily indicating strong new buying.
Price down + OI surging: new shorts increasing, but beware of potential short squeezes later.
So if ETH next pushes toward $2,000, I would prefer to see increased volume and moderate OI growth, rather than price being pushed up solely by short stop-losses.
Many people only study candlesticks but overlook that the contract market has another layer of "position structure."
The same 5% rise, whether bought from spot or forced from short liquidations, has completely different subsequent trading value.
#交易之声：你的经验值得被听到 #贝莱德重申BTC仍具配置价值 Sideways for five weeks, what is the market digesting?

Brothers, BTC has been grinding in the range of 63,000 to 65,500 for a full five weeks. Today it continues to hover around 64,000, neither rising nor falling.

But if you only look at the price, you might miss a key change.

The US stock market is rising, BTC is not following, the two markets are decoupling.

Yesterday the US stock market had another broad rally. Nvidia rose over 4%, with a market cap surpassing 3 trillion dollars. The S&P 500 rose for the eighth consecutive trading day, setting the longest winning streak in nine months. If this were six months ago, BTC would have surged along.

But this time it didn’t. BTC barely moved, not even a decent rebound.

Why? Because the driver behind the US stock market’s rise is the single logic of AI, which can ignore interest rates and macro factors and run its own independent trend. But BTC can’t. The core force driving BTC’s rise has always been liquidity expectations.

When the US stock market rises on AI, the spillover effect BTC can get is very limited. This doesn’t mean BTC has weakened; it means the pricing logic of the two markets is diverging—US stocks are telling the AI story, BTC is waiting for real money to enter.
 Just now, SKHY from Hynix and SNDK from SanDisk surged sharply. Did you guys not react in time? The shorts must be buzzing in their heads. Don't worry, let's analyze it!
A violent rebound of nearly 10% within an hour! This round of storage stock rebound was directly catalyzed by SKHY.
The company plans to spend about 40 trillion KRW to repurchase and cancel approximately 24.07 million shares, accounting for about 3.3% of the total shares, and at the same time promises to use at least 50% of the cumulative free cash flow from 2025 to 2027 for shareholder returns.
SKHY dares to expand production while simultaneously deploying huge funds for buybacks, indicating that management believes HBM and DRAM can continue to generate cash flow.
In other words, the storage market is at least not as bad as the market previously feared, and funds then spread this logic to MU, WDC, and SNDK.
Next, we only look at the strength and weakness boundaries of the two stocks respectively:
SKHY targets $150.
If it holds and then retakes $165 to $170, the target is $180, then $190 to $195 previous highs; breaking below $149 to $150 means the buyback can only buffer the decline and is not enough to reverse market expectations.
SNDK targets $1650 and $1750.
Breaking through $1750 gives a chance to test $1800 to $1830; with volume and steady hold above $1830, the main uptrend is considered restored.
Conversely, breaking below $1650 indicates this rise is still more of an oversold rebound, and losing $1600 means a second pullback should be guarded against. $SNDK $SKHY #财报观察员：小米Q2财报出炉，是汽车救场还是手机拖后腿？
Xiaomi's latest Q2 financial report is fully released, showing three core businesses with distinctly different development trends, with a clear divide between strong and weak.
First, the brightest spot is the smart car business, which has completely become the group's strongest growth engine.
The SU7 series delivered over 104,200 units in Q2, officially stabilizing at the 100,000-unit level per quarter, with the business gross margin rising to 20.1%. Most importantly, losses have significantly narrowed, from 3.1 billion in Q1 down to 2.06 billion.
It is clear that the scale effect has fully taken hold, and the automotive business has completely left the blind money-burning phase, entering a critical period of loss reduction and profit ramp-up. The annual delivery target of 300,000-350,000 units is steadily advancing, with growth certainty maximized.
Next, the core smartphone business shows a typical pattern of volume decline and price increase, with mixed feelings.
Against the backdrop of an overall industry downturn, shipments have declined somewhat, but the high-end strategy has been effective, with the average selling price of models reaching a historic high.
However, there are also significant weaknesses: due to upstream memory chip price increases, the overall device cost remains high, directly suppressing the smartphone business gross margin, and short-term profit pressure is evident.
Finally, the AIoT smart home business remains Xiaomi's most stable foundation.
Riding on the consumption rebound from the 618 shopping festival, IoT segment revenue surged significantly quarter-over-quarter, with demand for major appliances and smart home products continuously recovering.
This business line maintains a stable gross margin around 20%, with expectations for all three lines to perform well. $XIAOMI Tonight, the global financial markets welcome the most important day of the year. In the early hours of August 20 Beijing time, two major events occurred simultaneously: 1️⃣ The U.S. Treasury auctioned $16 billion in 20-year Treasury bonds 2️⃣ The Federal Reserve released minutes from its July meeting. Any one of these two events alone would be enough to cause intense market volatility. And when they appear on the same day, their influence can be amplified by each other. 📊 Why is tonight so important? Currently, the U.S. Treasury market is experiencing the fiercest sell-off in decades. The yield on the U.S. 30-year Treasury note touched 5.327% intraday on Tuesday, the highest since June 2007. The 10-year yield rose to 4.747%, the highest since January 2025. This is no small number. The last time long-term bond yields soared to this level was before the subprime crisis erupted. Meanwhile, U.S. stocks have fallen for three consecutive trading days, with the S&P 500, Nasdaq, and Dow Jones all under pressure. The scenario the market worries about most is: weak auctions + hawkish minutes reinforcing each other on the same day, pushing the entire yield curve upward and spreading to tech stocks, emerging markets, and high-leverage trading. 📊 What does it mean for BTC/ETH? If the auction results are strong, the minutes lean dovish, U.S. Treasury yields fall, risk assets rebound collectively, BTC could directly surge to 65,000-66,000, and ETH break through 1,950. If the auction results are weak and the minutes are hawkish, US Treasury yields continue to soar, risk assets will come under pressure, and BTC may pull back to 62,000-63,000, with ETH pulling back