xiaoshayu

93.52% voted in favor of Starknet “leaving Ethereum,” but STRK fell 40% in a week

A seven-day Starknet governance vote closed on October 22: around 3,200 votes, representing 93.52% of the voting power, backed “researching a transition to an independent, quantum-resistant L1.” Turnout set a record in Starknet governance history. StarkWare abstained, leaving the decision entirely to the community.

Don’t rush to read this as a “split”: this was a non-binding vote that only authorized “research,” not a “move.” Starknet is still an Ethereum L2, STRK remains its native asset, no new tokens are being issued, and the cross-chain bridge remains open as usual. An actual migration would require another binding vote. StarkWare says a draft white paper is due in November, with the goal of achieving quantum resistance by 2027.

The market has voted with its feet: STRK is currently trading around 0.0545, down about 41% in a week. The community wants direction; investors want execution—and a narrative about “looking into it” isn’t enough to support the token price.

Quantum security is a challenge for the entire industry, and Ethereum itself is replacing cryptographic components. Keep an eye on that draft white paper in November; that’s the ammunition for the next leg of the market.

$STRK $BTC

Ethereum has “delayed” its upgrade—and that’s exactly why it’s reliable

The Glamsterdam hard fork, originally scheduled to go live on the Hoodi testnet on October 26, has been postponed by the Ethereum Foundation. A new date has not been set.

The trigger was an October 13 warning from security firm ChainSecurity: two proposals in the upgrade, EIP-8037 and EIP-8038, change the gas rules and could weaken protections against “gas insufficiency griefing attacks.” They have already found 13 affected cases in real-world contracts and are working through them one by one via responsible disclosure. The plan to raise Hoodi’s gas limit to 200 million has also been put on hold, and the mainnet launch, originally scheduled for early December, is now uncertain.

Many people hear “postponement” and assume it’s bearish, but that’s backwards: the whole point of a testnet is to uncover problems early. Waiting until it reaches mainnet to fix them would be a disaster. A chain that can hit the brakes itself and address those 13 cases before moving ahead is far more reliable than one that just “ships first and asks questions later.”

In the short term, ETH is following the broader market. Whether it can hold 2,400 is key; if it fails, watch the 2,380 area. The upgrade is delayed, not canceled.

$ETH $BTC

ETFs bought $1.08 billion in three days, yet BTC still fell back to 81,500: institutional cost bases are being put to a public test

First, the money: $526.6 million on Monday, $23 million on Tuesday, and $528 million on Wednesday—for a total of $1.08 billion over three days (SoSoValue). BlackRock’s IBIT saw a single-day inflow of $665 million on Monday, its largest since January, bringing its total holdings to 814,000 BTC.

Now, the price: BTC has retreated nearly 4% from its high of 84,808, returning to around 81,500. Bitfinex’s view is straightforward: the short-term range is 81,300 to 84,400, and the average cost for ETF buyers is about 81,500—this level is now in the spotlight. If the daily close falls below 81,300, the next level to watch is the October 15 low of 79,821.

One more detail: Lookonchain spotted a whale that sold 1,917 BTC ($159 million) over 11 days, then turned around and bought 64,500 ETH ($155 million). Institutions aren’t here to catch falling knives; they’re here to buy the dip. If the bottom breaks, they’ll rebalance all the same.

The conclusion stings a little: inflows didn’t lose—price did. ETF money can buoy sentiment, but it can’t prop up the Nasdaq.

$BTC $ETH

Tesla and Nvidia are teaming up to pour $2.8 billion into something that isn’t chips, but the “packaging” for chips

Reuters, citing South Korea’s Maeil Business Newspaper, reports that Samsung Electro-Mechanics will receive an investment of 3.7 trillion won (about $2.8 billion): Tesla will contribute 1.1 trillion won and Nvidia 2.6 trillion won. The money will go toward building an FC-BGA advanced packaging substrate production line in Sejong, alongside long-term supply agreements. Note: None of the three companies has officially announced this; for now, it’s based on reporting by a reputable news outlet.

FC-BGA is a packaging substrate for AI chips—no matter how powerful a chip is, without a substrate, it can’t be mounted on a board. It’s one of the tightest bottlenecks in the current AI hardware boom, and major players are now moving upstream to lock in capacity and secure supply with advance payments.

The interesting part is the logic: when you can’t get ahead in chip design anymore, you buy into the “scarce link” in the chain. The crypto world uses the same idea: BTC’s scarcity is hard-coded, while AI scarcity is created by spending big to secure supply. One is deflationary by algorithm; the other, through capital expenditure. The more frantically money is poured into AI infrastructure, the more people listen to the digital gold narrative—after all, both bets are on the same thing: limited supply and unlimited demand.

$BTC

Tesla fell 5% in premarket trading: earnings beat expectations, but Musk personally slashed the AI budget by $5 billion

Q3 figures look decent: revenue was $28.05 billion, roughly flat (vs. $28.34 billion expected); adjusted EPS was $0.48, beating the $0.45 forecast; free cash flow was $1.53 billion, and cash on hand totaled $45.4 billion. Deliveries came to 486,000 vehicles, with an automotive gross margin of 18.1% (excluding regulatory credits).

What the market took issue with was this cut: the 2026 capital expenditure guidance was reduced from more than $25 billion to around $20 billion. Musk also dampened expectations for Optimus on the earnings call, while Cybercab is back in NHTSA’s sights (Tesla must respond by October 30). The proposed merger with SpaceX has only led to the formation of a committee, with no decision made. Shares closed at $374.58 on October 22, down 3.08%, then fell another 5% in premarket trading.

Most analysts, meanwhile, raised their price targets (Piper: $475; TD Cowen: $460), with only HSBC bearish at $157. The knock-on effects are real: the Nasdaq fell 2.18%, BTC was dragged down to around $81,500, and ETH dropped below $2,400. BTC has once again acted like a high-beta tech stock in this cycle: as AI capital spending gets repriced, crypto gets caught in the crossfire.

$BTC $ETH

China’s prosecutors seek to have 61,800 bitcoins awarded to China: the coins are now in the UK, worth $5 billion

The case centers on Qian Zhimin. From 2014 to 2017, the Tianjin Lantian Gerui company she controlled sold unlicensed wealth-management products, raising 40.378 billion yuan from 102,000 investors. It ran up a shortfall of nearly 20 billion yuan, and the money went toward repaying old debts, buying bitcoin, and purchasing homes and cars. In 2017, Qian fled to the UK, where British police seized the coins—the largest cryptocurrency seizure ever confirmed in law-enforcement history.

There were two major developments last week: on October 13, the Tianjin Second Intermediate People’s Court announced that it had accepted prosecutors’ application for confiscation on September 28. There is a six-month period for claims, during which victims can come forward to assert their rights. On October 12, the UK High Court ruled that victims’ claims would be governed by Chinese law, not British law.

But don’t get the wrong idea: the coins are still in the UK, and China’s confiscation order doesn’t mean they’ll automatically be transferred back. Ownership is far from settled. At current prices, the coins are worth about $5 billion. More than 60,000 bitcoins are in limbo, creating a long-term variable on the supply side. The short-term impact on the market is limited; over the long term, it will depend on how courts in China and the UK divide the pie.

$BTC

ETFs bought $1.08 billion in three days, while miners quietly dumped: who won this game?

This week’s ETF numbers look great: $526.6 million on Monday, $23 million on Tuesday, and $528 million on Wednesday—for total net inflows of about $1.08 billion over three days (SoSoValue), ending two straight weeks of net outflows. BlackRock’s IBIT ($235 million) and Fidelity’s FBTC ($213 million) led the way on Wednesday, accounting for 85% between them.

But there’s another side to the coin: miners. Binance Research’s weekly report noted that the amount of BTC in miner wallets has fallen by about 15% since June. In August alone, 9,566 BTC flowed out—the largest monthly reduction since July 2024.

In other words, while Wall Street is taking in supply through ETFs, miners are cashing out at higher prices. Coins are moving from the hands of “producers” to those of “allocators.” And the price? BTC has been caught in a tug-of-war between $82,000 and $84,800, with neither side gaining the upper hand.

My take: Miner selling is a bearish supply-side overhang being cleared. Whether prices can rise hinges on whether ETF demand can absorb it. Around $80,000, buyers are clearly more patient than sellers. I’m not making a move unless it breaks below $80,000.

$BTC #Bitcoin

The U.S. froze $833 million in crypto assets. This wasn’t a sell-off—it was a crackdown on scammers.

At a White House cybersecurity event on October 21, the U.S. announced the results of a new enforcement effort targeting overseas crypto scams: about $833 million in crypto assets tied to criminal activity was frozen, 503 scam websites were shut down, and nearly 9,000 Americans who were being scammed were directly warned.

The crackdown targeted “pig-butchering” scam compounds in Myanmar, Cambodia, Laos, and the Philippines. Their tactics included romance scams, phishing, and ransomware. Last November, the Treasury Department formed a multi-agency Scam Center Strike Force. This operation also involved sanctions, cutting off the scam rings’ technology and funding pipelines.

This is a neutral-to-positive development for crypto: first, the targets are scammers, not the industry, which should reassure legitimate players; second, tougher enforcement lowers the psychological barrier for institutions to enter the market. Consider that the OCC just issued its first final rulebook for stablecoins—the regulatory “licensing era” and “enforcement era” are arriving at the same time.

One final reminder: if someone asks you to download an app or follow a “teacher” giving trading signals, assume it’s a scam, no matter how much it looks like an exchange promotion. Legitimate projects don’t ask you to transfer money in a private chat.

$BTC #Bitcoin

Musk wants to put two companies together—and, in the process, create the world’s sixth-largest Bitcoin treasury

On October 20, during SpaceX’s earnings call, Musk confirmed that Tesla and SpaceX had each formed an independent special committee of the board to evaluate a merger. He also offered a caveat: the committees haven’t negotiated any terms yet, and “it’s an exaggeration to treat a merger as a done deal.”

But the numbers are already there: SpaceX holds 18,712 BTC (hidden in its IPO filing for years before being revealed), and Tesla holds 11,509. Together, that’s 30,221 BTC. At the current price of just over $82,000 per coin, they’re worth about $2.5 billion—enough to break into the top six corporate Bitcoin holders worldwide, behind only a few of Strategy’s giants. Even more striking is their cost basis: the two companies paid about $1.05 billion in total, at an average price near $35,000, leaving them with more than $1.4 billion in unrealized gains.

Let’s keep things in perspective: those 30,221 BTC are a “combined balance sheet” figure, not new buying. Neither company has any plans to buy or sell in the near term, so don’t expect this to directly give Bitcoin’s price a boost. The real signal is that Musk-affiliated companies treat BTC as a long-term treasury asset—and don’t sell, even when the price falls. That kind of corporate holding—“accumulate, never sell”—is the strongest foundation in a bull market.

My own position-sizing rule is simple: as long as $80,000 holds, I’m not making a move.

$BTC #Bitcoin

The market isn’t buying UNI’s trade, but it could be DeFi’s biggest move this year

CME officially confirmed that FalconX and GSR completed the first-ever UNI block trade: privately negotiated, then cleared through an exchange. Hayden Adams said it was “the first time any DeFi token has gone through a compliant TradFi channel,” making UNI the first DeFi token to be listed on CME.

The interesting part is that the price barely reacted: UNI is hovering around $7.20, down 3% over the past 24 hours. The news was announced back in September, and the token had already rallied ahead of the listing. Now it’s a classic sell-the-news move: the good news is priced in, and holders are cashing out.

But the long-term picture is worth watching: Bitwise filed an S-1 for a spot UNI ETF in February. Based on the industry’s reading of the generic listing standards the SEC approved last September, an asset may qualify with six months of futures trading history on a CFTC-regulated market. If all goes smoothly, UNI could meet the criteria around April 2027.

DeFi’s absorption into traditional finance is a slow-moving trend, and a lagging price doesn’t mean the underlying thesis has changed. Still, the ETF is far from a done deal, so don’t rush in: at $7, you’re buying an expectation.

$UNI #FirstUNIBlockTradeViaRegulatedTradFiChannelCompleted

BTC took 6 hours to break above 84,000—and only 6 hours to give it all back

Last night, BTC pulled off something both impressive and ugly: it first broke above 84,000—a level that had rejected rallies twice, on October 13 and 14—and reached a high of 84,800. Then, within 6 hours, it gave back more than 2,000 points, falling to 82,600. It’s now hovering around 82,300.

The blame lies with U.S. stocks. Tesla fell 3% after hours and cut its capital expenditure outlook; chip equipment maker Lam Research plunged 7% on a weak forecast. Nasdaq 100 futures followed, down 1.1%, bringing back concerns about whether AI spending can keep burning cash. Even Meta’s 2% post-earnings gain couldn’t lift sentiment. Add escalating tensions in the Middle East and broad declines across Asian markets, and risk assets are all losing steam.

Spot ETFs have seen net inflows for three consecutive days this week, totaling about $1.1 billion ($527 million on Monday, the biggest single-day inflow in nearly a month), but prices paid no attention—the drop was purely sentiment-driven.

Key level: 82,500. If it holds, BTC could try for 84–84.5K again; if it breaks, the next levels to watch are 81.2K–80K. The half position I bought at 81,500 and added to at 80,800 is still in the green. I’ll hold as long as 80K holds, and won’t rush to add.

$BTC #BitcoinDropsBelow$84K

ETH Touched $2,526, Then Fell Back to $2,430: Is the $2,500 Gate Made of Paper?

ETH did something pretty exciting today: it surged to an intraday high of $2,526, breaking above $2,500, then gave back almost all of its gains, with the price returning to around $2,430.

The bulls have just one card to play: a short squeeze. After $45.8 million in ETH short positions were liquidated, ETH broke above $2,450, with liquidation-driven buying pushing the price higher. But spot market outflows haven’t let up—the U.S. spot ETH ETFs saw $1.08 billion in net outflows in October, their largest monthly outflow of 2026, and the selling pressure from redemptions is there for all to see.

Staked ETH did hit a new high, though, at 43.8 million ETH. Bitmine staked another 159,000 ETH, bringing its total to 5.23 million. Long-term holders are keeping their coins locked up, but holding on doesn’t mean there’s spot demand—staking and buying the dip are two different things.

The next catalyst is the activation of the Glamsterdam upgrade on the Hoodi testnet on October 26. Technically, $2,502 is the first hurdle overhead, while $2,438 is support below. The risk of chasing the price here outweighs the potential reward, so first let’s see whether it can hold above $2,438.

$ETH #EthereumSurpasses$2500

7.79 million transactions moved just $7,722: RLUSD’s “boom” was 110 bots spamming transactions

An on-chain investigation published by Bitquery on October 19 offers a warning to anyone looking at public blockchain data: from September 19 to October 18, 110 wallets on the XRP Ledger accounted for 99.6% of RLUSD payment transactions—7.79 million transactions that transferred just $7,722 in total, averaging $0.001 each, with the funds ultimately flowing into three collection addresses.

The most extreme day was September 26: of 573,000 payments, only 388 were worth more than $1. The same group of wallets also accounted for 45.6% of all account-to-account transfers on the network that week. On the other hand, 14,600 large payments of more than $1 transferred $7.33 billion during the same period, with just 10 wallets accounting for $6.05 billion. Real money and fake activity came from completely different groups. And the cost of inflating the numbers was absurdly low: 8.6 million transactions incurred fees of just 495 XRP (about $648).

The lesson is simple: when looking at a public blockchain, don’t just count transactions—look at the money. Transaction counts can be inflated; value flows can’t. Next time you see “daily active users hit a new high on [a] chain,” first ask: is that real money?

$XRP #110 automated wallets accounted for 99.6% of RLUSD payments

Losing $4.8 Billion in Two and a Half Months: South Korean Retail Investors’ Leveraged ETP Losses Offer a Lesson for Crypto Traders

Data submitted by South Korea’s Financial Supervisory Service to lawmakers shows that from May 27 to August 14, retail investors trading single-stock leveraged ETPs through 10 brokerages lost a combined 6.313 trillion won, or about $4.8 billion. Of that, 2.32 trillion won was realized losses and 3.99 trillion won was unrealized—more than 60% of the losses haven’t been cut yet. If South Korean stocks keep weakening, the bill will only grow.

By age group, investors aged 40 to 49 lost the most, with 165,200 accounts recording a combined 2.05 trillion won in unrealized losses. Investors aged 70 and over had the biggest losses per account, averaging about 29.94 million won each.

Leverage works the same way in stocks and crypto: when you get the direction wrong, time is your enemy, not your friend. Crypto traders using futures should take a look in the mirror, too—chasing rallies and selling off in a one-way market with leverage means paying the same tuition as the middle-aged guys in Seoul. The data also offers another warning: with unrealized losses accounting for 63%, most people are still “playing dead.” The real wave of capitulation may not have arrived yet. Don’t look for certainty amid volatility. Survive first.

$BTC #SouthKoreanRetailInvestorsLose6.3TrillionWonOnSingle-StockLeveragedETPs

Last night’s U.S. stock earnings “reveal”: Meta posted record revenue and Tesla’s cash flow went through the roof, but after-hours trading went both ways

The two biggest earnings reports of the season landed at the same time last night. Meta reported third-quarter revenue of $64.6 billion, up 26% year over year, and EPS of $3.83 (versus the $3.45 market estimate). It raised its fourth-quarter revenue guidance to $73–76 billion and announced a 2-for-1 stock split. The cost: Reality Labs lost another $4.13 billion, and its full-year AI capital expenditure guidance is $135–145 billion. Shares rose about 3% after hours, as Wall Street is, for now, willing to pay for its AI spending spree.

Tesla’s revenue of $28.05 billion and adjusted EPS of $0.48 both beat expectations. The standout was free cash flow of $1.53 billion—versus market expectations for a $4.99 billion deficit—turning positive outright. Gross margin rose from 16.8% to 19.6%; FSD subscriptions were up 67% year over year to 1.74 million; and autonomous robotaxi miles increased eightfold quarter over quarter to nearly 2 million. Optimus has also started work on the mass-production line in Fremont. But after initially surging after hours, the stock pulled back, down about 3.3% as of 8 p.m. ET.

The Nasdaq closed up 1.05% last night. This morning, BTC held above 84,000, the Fear & Greed Index returned to greed territory at 68, and spot BTC ETFs posted net inflows for a second straight day. AI is still burning cash, but BTC caught a bit of the risk-asset momentum. Now the question is whether that momentum can last.

$BTC #StockEarningsSeason

The stablecoin “licensing era” is here: the U.S. OCC has issued its first final rule, effective January 18, 2027

On October 21, the U.S. Office of the Comptroller of the Currency (OCC) issued its final rule implementing the GENIUS Act for stablecoins, becoming the first federal regulator to complete rulemaking. The rule takes effect alongside the Act and was finalized after receiving around 328 industry comments.

There are three key points. First, “hard standards”: reserves must be fully backed by highly liquid assets, redeemable at par, and audited annually. Foreign issuers and some custody services are also brought under regulation. Second, redemption rules have changed: redemptions must be completed within two business days. The rule drops the March proposal’s provision for an automatic seven-day extension during a “redemption surge”—many commenters said that could fuel a run. Instead, firms must notify the OCC within 24 hours, and the OCC will decide whether to grant an extension. Third, the ban on yield has been eased: the Act already prohibits issuers from paying interest to holders, and the proposal added a presumption that third-party yield payments were illegal. The final version removes that presumption and opts for case-by-case review. The current offerings cited by CoinDesk will all have to meet this standard: PayPal offers 4% on PYUSD, Coinbase 3.75% on USDC, and Kraken 4.5% on USDG.

What this means for holders: stablecoins are becoming a fully licensed business, but strategies that use yield to retain customers will face tighter restrictions. The Fed’s September draft is still in the works—the game is only just getting started.

#OCC发布GENIUS法案首个最终规则 $USDT

ZEC Jumps 11% in a Day, Touches $1,260: This Privacy-Coin Rally Isn’t Hype—The Upgrade Timeline Is Set

ZEC was the standout across the market today, gaining more than 11% in a single day and briefly touching $1,260, leading all major coins. XMR followed, rising about 5% to $620, as privacy coins bounced back across the board.

The catalyst is concrete: Zcash developers have officially set the mainnet activation height for the NU7 upgrade, scheduled for November. NU7 cuts block times from 75 seconds to 25 seconds, making payment confirmations three times faster. It also directs 60% of transaction fees into a reserve fund, to be released in installments over time—delivering both a speed narrative and a deflationary one. NU7 has been running on testnet since early October, so this isn’t just a promise.

Money is actually flowing in: ZEC futures open interest rose 24% in a day, suggesting new positions are betting on the move rather than existing positions simply changing hands. But note that the annualized funding rate has already reached 10%, and sellers have been persistently hitting bids during the rebound. In this setup, chasing the rally can leave you squeezed from both sides.

The way I see it: $1,260 is a sentiment-driven high, and with funding this expensive, chasing now doesn’t look wise. If you’re genuinely bullish on the privacy narrative, consider waiting for a pullback to $1,100–$1,150. If it breaks below $1,000, admit you’re wrong and cut the position. Coins like ZEC have never played by the rules when it comes to volatility, so don’t get carried away with your position size.

$ZEC

Solana gets a new heart tonight at 22:50: confirmations drop from 12.8 seconds to 0.15 seconds, but one metric will look “scary” tomorrow

Alpenglow activates on mainnet tonight at 22:50: after 18 months of development by Anza, it will hot-switch to Votor within a single slot, ditch TowerBFT, and cut finality from 12.8 seconds to around 150 milliseconds.

A heads-up first: don’t panic if Solana’s TPS drops from 3,000 to a few hundred tomorrow—the transactions being removed are votes, which account for about 75% of block activity. The numbers will look smaller, but usable block space will nearly triple.

Validators get a real boost: annual voting fees of 240–360 SOL are cut by more than 98%, and the break-even staking threshold drops from 4,850 to 450 SOL. The risk: Firedancer is absent, leaving the network reliant on Agave-based clients. Keep a close eye on the transition period.

SOL is currently at 106–107, down 10% in two weeks and still 64% below its peak of 293. The positive news isn’t priced in. On-chain activity isn’t dead: stablecoins are at around $16.2 billion, daily DEX volume at around $2 billion, and TVL at around $6 billion.

My position (purely fictional): I hypothetically bought some around 101.5 on October 18. I’m holding with about 5% in unrealized gains, will admit I’m wrong if it breaks below 98, and am targeting 115–120.

$SOL

After the U.S. market closes tonight, Meta and Tesla report earnings at the same time: Will AI keep burning cash? That could determine whether BTC can break above 84,000

After the U.S. market closes tonight (early tomorrow morning Beijing time), Meta and Tesla will both release their third-quarter earnings. Meta is expected to report EPS of $6.58 and revenue of $63.3 billion, with investors focused on ad growth versus AI capital spending. Tesla is expected to report EPS of $0.45 and revenue of $27.9 billion, with attention on gross margins and progress on Robotaxi.

Why crypto traders need to pay attention: Yesterday, SpaceX and Supermicro reported better-than-expected results, sending Nasdaq futures up 1% and BTC to a one-week high of 83,500. AI capital spending is the lifeline of risk assets—if Meta keeps ramping up capex, the AI narrative gets another lease on life, giving BTC the confidence to test 84,000. Analysts have drawn a line in the sand: BTC needs to hold above 84,000 to be ready for the FOMC meeting on October 27–28.

On the other hand, if either company gives a cautious outlook, tech stocks could pull back, and 83,500 will likely prove to be a false breakout. The Fear & Greed Index is at a neutral 53, and the market is waiting for direction.

Strategy: Don’t chase prices before the earnings come out. BTC will likely range between 82,000 and 84,000. A break above 84,000 on strong volume would be the real signal; if it falls below 82,000, watch 80,000.

$BTC #EarningsSeason