海盗鸭

In 2012, a British man put about $2,000 into Bitcoin, and then the early UK trading platform he used shut down, leaving that money effectively a bad debt in his mind. In January this year, he contacted the law firm CEL Solicitors, and by the end of May the settlement was completed; 61 bitcoins $BTC were recovered, worth about $4.5 million at current prices. The recovery did not come from cracking a wallet, but from on-chain tracing plus a legal claim: the firm and its tracing team identified an address believed to belong to users from that era, which held more than 5,500 coins, worth over $400 million at today’s prices. The firm said other old users who can prove ownership may also recover a share. To convert it, $2,000 became $4.5 million, more than 2,000 times. The significance of this case is that there are still many unclaimed coins sitting in early failed platforms, and whether they can be recovered depends on whether you have kept records proving “this is mine.”

Evernorth's second-quarter report presents a set of seemingly contradictory figures: the average daily trading volume on the order book of ledger $XRP rose to 3.57 million units, up 79% year over year (about 1.99 million in the same period last year); but the number of active accounts actually placing orders each day fell from more than 1,800 to about 1,100, down 40% year over year. With fewer participants and higher volume, the average daily trading volume per account is now nearly triple what it was before. Over the same period, the value of assets hosted on XRPL surpassed $4 billion. This is not retail users returning, but rather the user structure of the ledger undergoing a reshuffle: small accounts are exiting, while addresses used for institutional settlement and large-value matching are becoming more dominant. The upside is depth and efficiency; the downside is that when participation becomes too concentrated, if one or two big players pull out, liquidity can look very weak. Looking at a chain like $XRP , which markets itself on payments and settlement, would you care more about the number of active addresses, or the actual settlement volume?

Prediction markets have put the probability of the CLARITY Act becoming law in 2026 at 15%. Behind that number, it’s not that political sentiment has changed — it’s that the calendar has run out.

Here’s the rundown: H.R. 3633 (the Digital Asset Market CLARITY Act) passed the House in July 2025 by a vote of 294 to 134. In the Senate, Majority Leader Thune filed a cloture motion on the motion to proceed before the August recess. The vote is set for September 15 and needs 60 votes. The key detail is this: this is not the final passage vote; it only decides whether the Senate can begin formal consideration. The Senate version of the text has already been amended, and even if the Senate completes the process, it still has to go back to the House for another vote. Meanwhile, House Republican leadership removed the weeks of September 21 and September 28 from the schedule, cutting eight legislative days and collapsing the usable window to just two days. Back in February, the probability was still around 82%. On September 3, the National Sheriffs’ Association shifted its stance from opposition to neutral, one of the few positive changes.

The practical impact on related assets, ranked by importance:

First, market structure legislation is about what counts as a commodity, who regulates spot markets, and what can be listed legally. It does not determine short-term price; it determines the range of assets that can be allocated to.

Second, if the bill does not pass, regulation will continue to advance through case-by-case enforcement, and the compliance discount on altcoins and DeFi will be hard to unwind. Institutional money will keep crowding into a small number of assets that already have compliant channels — assets like $BTC and $ETH with spot ETFs will absorb most of the incremental inflows. That also explains why this cycle’s so-called alt season has never felt complete.

Third, 15% does not mean dead; it means the timeline has been pushed into the lame-duck session after the midterm elections. For traders, the real event is the 60-vote threshold on September 15 — if it doesn’t happen, don’t price it in early.

Do you think the bill will pass first, or the rally will run out first?

#CLARITYAct2026legislativeprobability15%

On September 2, the U.S. 10-year Treasury yield briefly touched 4.818%, the highest level since November 2023, and above the January 2025 peak, before easing back to the 4.77%-4.80% range.

There were three drivers: oil prices climbing above $90 to raise inflation expectations; fiscal supply pressure combined with a synchronized global bond selloff; and a repricing of the Fed path—Fed Chair Worsh's hawkish speech at Jackson Hole pushed the probability of a September rate hike from about 36% to 58%-66%, while on September 3 Governor Waller said he would lean toward holding steady if inflation continued to improve, and the probability then fell back to 48.4%. The meeting is set for September 15-16, and August CPI is the deciding vote. Please note: this round is trading the expectation of a "rate hike," not a rate cut.

There are four things ordinary users can actually use:

First, treat 4.8% as the opportunity-cost benchmark. Stablecoin yields, basis arbitrage, and staking returns all need to subtract this risk-free rate first; only the remainder is the excess return you earn for taking volatility risk. If you can't beat it, there's no need to take the risk.

Second, look at what makes up the yield move. If it's driven by inflation expectations, <$BTC > is relatively resilient; if it's driven by real rates, risk assets tend to fall broadly. Watching the real rate implied by inflation-protected bonds is more useful than staring at nominal yields.

Third, don't confuse probability with conclusion. The rate-hike probability moved from 36% → 66% → 48% within a week. Using leverage before such an event is negative expected value; position management is more valuable than directional calls.

Fourth, pin the schedule to your calendar: August CPI, and the September 15-16 FOMC. On September 3-4, <$BTC > rebounded from 77,000 to above 81,000, and the trigger was the drop in rate-hike odds. This correlation is very strong right now—the first-hand news for crypto is news on rates.

How is your current position set up: for a "rate hike" or for "holding steady"?

#U.S. 10-year Treasury yield touches highest level since November 2023

Two consecutive inflow streaks were broken on the same day. Data for September 2 showed that U.S. spot Ethereum ETFs saw net outflows of $48 million, ending a 12-trading-day streak of inflows totaling $1.62 billion; spot XRP ETFs saw net outflows of $7.2 million, ending an 11-day winning streak, with about $170 million in inflows during that run. On the same day, spot Bitcoin ETFs instead recorded net inflows of $101.2 million. The cumulative net inflows into XRP spot ETFs were still about $1.68 billion, and the price at the time was around $1.36. How should we read these numbers: the significance of a single day's flow is far smaller than the continuity of the trend. The break in the winning streak shows that incremental buyers at this price level are starting to be selective, while capital simultaneously flowing back into Bitcoin is a typical sign of risk appetite contracting — first defend the leader, then look at the altcoins. What to watch next is whether the outflows form a string.
$ETH $XRP #EthereumXRPETF winning streak ends

This lawsuit is worth untangling, because it may decide what crypto perpetual contracts look like inside the United States.

Timeline: On May 29, the CFTC approved Kalshi to list Bitcoin perpetual contracts and allowed other designated contract markets (DCMs) to list similar products as futures; on June 18, CME sued the CFTC, arguing that such perpetual contracts should be classified as swaps under the Commodity Exchange Act and Dodd-Frank, and that the CFTC sidestepped the rules to give a competitor the green light; on September 2, the CFTC filed a motion to dismiss, bluntly calling it much ado about nothing.

The CFTC's argument has three layers. First, CME lacks standing: it can list perpetuals itself as a DCM, yet it publicly said customers are not asking for such products, so the alleged harm is self-inflicted; second, the data do not support injury: CME's monthly trading volume for Bitcoin and Ether futures was higher in June and August than in May — after the competitor was approved, volume did not fall, but rose; third, even if the court reclassifies perpetuals as swaps, Kalshi and others could still issue them under swap rules, so CME's alleged harm would not disappear.

Why does this matter? Futures and swaps are two different systems when it comes to the regulatory framework, clearing, tax treatment, and retail access. By classifying perpetuals as futures, the CFTC is effectively opening a more flexible compliance path for regulated U.S. markets than swaps would allow; if CME wins, the product could shrink back into a structure dominated by large exchanges plus swaps, with access barriers rising accordingly. The CFTC also invokes the legislative purpose: the Commodity Exchange Act is meant to promote responsible innovation and fair competition — using regulation to shut out new rivals is the wrong direction for an established exchange.

Next, watch two dates: October 2, the deadline for CME to file its opposition, and whether the judge agrees to dismiss the case. Whatever the outcome, this battle over whether BTC perpetuals are futures or swaps will set a reference point for global crypto derivatives regulation. #CFTCRequestsDismissalOfCMEsPerpetualContractLawsuit

Mexico’s third-richest man called fiat inflation an “invisible tax that takes away wealth.” On August 31, Ricardo Salinas Pliego posted that fiat inflation is an invisible tax, also calling central banks’ money creation out of thin air a “fiat scam,” and directly advising followers to buy and hold Bitcoin for the long term. He is not just talking: about 70% of his portfolio is allocated to Bitcoin, and he has also publicly said that Bitcoin is better than real estate for preserving value. Statements like this spread widely, but it is important to understand the conditions under which they make sense: a 70% concentration in a single asset is backed by a balance sheet large enough that its cash flow does not depend on that investment. The same position size means something entirely different for ordinary people: he can withstand a 50% drop in paper value, while for most people a single 30% drawdown is enough to change their lives. One can agree with the judgment that money is losing value, but position sizing cannot simply copy someone else’s balance sheet. $BTC

Oracle rose 5.7% on September 3, but the reason had almost nothing to do with its business. The trigger was Fed Governor Waller hinting that he leans toward keeping rates unchanged this month. Why this is especially useful for Oracle: it has taken on a lot of debt to expand AI data centers, and every easing in financing costs flows directly to its profit and loss statement. This gain was also part of a broader rally across the software sector. The setup was also favorable because the stock had fallen about 36% over the past 12 months. Jefferies said that “the worst-case scenario may already be priced in,” citing that its cloud infrastructure business is still growing at 93% on a constant-currency basis, while market sentiment has already become extremely pessimistic. My view: rebounds of this kind, where the stock jumps as soon as rates ease, are essentially handing the company’s valuation over to Treasury yields rather than to its ability to deliver cloud orders. The next earnings report will be the real test: now that the money has been spent, can revenue and cash flow keep up? $ORCLB

The key to Dell’s earnings report is not revenue, but gross margin.

Here are the numbers first: for the quarter ended at the end of July, revenue was $47 billion, up 58% year over year and about $2.2 billion above market expectations; adjusted EPS was $7.04, up 203% year over year. AI server recognized revenue was $16.4 billion; new orders signed during the quarter were $60.9 billion; backlog was $95 billion. All three were records. Full-year revenue guidance was raised from $167 billion to about $192 billion.

Why is this worth 8% (it actually rose about 16% the next trading day)? Because over the past year, the bears’ core argument has been that “AI servers are being sold at a loss.” In the previous quarter, the ISG segment’s operating margin fell from 14.8% to 10.5%, and gross margin dropped from 21.1% to 17.8%, while HBM and DRAM prices surged by more than 50% in a single quarter. This quarter, profits tripled and guidance was raised, which effectively knocked that argument down on the spot. What got re-rated was the profitability quality of the entire AI hardware chain, not just Dell’s revenue.

The impact chain for insiders is shorter than it looks. The 95 billion in backlog will ultimately need power and data-center space, while existing grid-connected power is in the hands of miners: already in August, a listed mining company leased 191 MW at a Texas site to Anthropic. In the first half of this year, listed miners’ actual hash rate fell by about 56 EH/s, a decline of 15%, faster than the 10% drop across the network, while AI-related revenue rose 52% year over year in the same period.

This is a two-way street for $BTC : steadier miner cash flow and less pressure to sell coins are good; but as hash-rate growth gets pulled toward AI, the pricing logic for mining stocks is shifting from “coin-price beta” to “power assets,” and the two no longer move in lockstep.

Three things to watch next: the delivery pace of the 95 billion in orders, memory costs, and the supply constraints management keeps mentioning. $DELLB is already a business that has gone from selling PCs to moving power and memory.

#DellEarningsBeatStockUp8Percent

Crude oil stopped after three days of gains, which does not mean the rally is over; it means the war premium has shifted from "expansion" to "maintenance".

First, the level: WTI briefly tested $91, and settled around $91.01 on Thursday, while Brent was about $95.63, both at six-week highs. For the week, WTI rose about 9%-10% and Brent about 6.5%-7%, making it the strongest week since July. It fell during Thursday’s Asian session because there were signs that passage through the Strait of Hormuz was recovering, then the conflict pulled it back up again.

In terms of mechanism, war premium prices in "probability × loss," not a supply cutoff that has already happened. There are three layers supporting it: first, the inventory cushion is thin. The EIA expects U.S. commercial crude inventories to stay below the five-year low through the end of 2026, and last week they fell another 4.5 million barrels. Second, the production-card is almost played out: OPEC+’s September output increase of 188,000 barrels per day is seen as the final quota hike for 2026, and the buffer is getting thinner. Third, the cost of going through the strait. The war-risk surcharge in the Strait of Hormuz has jumped from about 0.25% of hull value to around 5%; for a $100 million tanker, one trip through the strait means insurance costs on the order of $5 million — an invisible tariff. Even if tankers still move, delivered costs still rise. For scale, the EIA estimates that in the second quarter, crude oil and petroleum liquids flowing through the strait fell to about 4.9 million barrels per day, versus a normal level of about 21.6 million barrels per day.

So the accurate meaning of "stabilizing" is this: the premium has stopped expanding, but no one dares bet it will unwind. Going forward, watch three high-frequency variables: war-risk quotes, the actual number of ships passing through the strait and their volumes, and the pass-through from oil prices into inflation data. The third is most relevant to crypto: oil staying above $90 will support inflation expectations, and the Fed meeting on September 15-16 sits right on that line.

Do you think this round’s oil-price peak was set by demand, or by insurers?

#CrudeOilStabilizesAfterThreeDaysOfGains

The technical story behind BNB this time is rooted in a business-side rollout. On September 1, Binance launched stock options covering more than 1,000 U.S. stocks and ETFs, with physical delivery, offered through its Abu Dhabi ADGM-licensed entity Nest Trading. A concurrent data point better shows the direction: on the platform, TradFi perpetual futures had about $433.4 billion in trading volume in August, roughly 15 times January’s level. On the price side, after BNB retested $680, the former resistance level turned into support, rising nearly 4% in 24 hours. It is now stuck below the $733–740 resistance zone, and some analysts believe that breaking above $740 could open the way to $825, implying roughly 15% upside. But to be clear: there is no direct cash-flow transmission between the new product and BNB’s price, and treating the two as equivalent is risky. What really matters is whether the fee scale of these new businesses and BNB’s actual use cases have increased. $BNB

ARK made a very clear portfolio rotation on August 28: selling AMD and buying Nvidia and Broadcom. Based on the closing prices that day, it sold 156,300 shares of AMD for about $72.8 million, bought 243,700 shares of Nvidia for about $53 million, and 112,800 shares of Broadcom for about $41 million. The amount sold was about $20 million more than the amount bought, which means this was not just a change in holdings, but also a modest reduction in overall chip exposure. The timing is also interesting: on the day it bought Nvidia, Nvidia’s stock had just closed down 4.6% at around $217, so it was a buy-the-dip move. How to read this adjustment: AMD’s recent rally has been sharp and its valuation has moved up quickly, while Nvidia and Broadcom have more clearly visible order pipelines. In other words, this is a shift from the upside potential of a “catch-up” name back to the certainty of existing orders. A one-week institutional rebalance should not be treated as gospel, but the direction is worth noting.
$AMDB $NVDAB $AVGOB

On September 2, U.S. spot Ethereum ETFs saw net outflows of about $48 million, ending a 12-trading-day streak of net inflows. During that streak, they attracted about $1.62 billion in total. By product, ETHA saw $53.4 million in outflows, FETH $26.2 million, and ETHE $23.5 million. On the same day, spot XRP ETFs saw net outflows of $7.2 million, ending an 11-trading-day inflow streak (about $170 million during the period), while cumulative net inflows rose to more than about $1.66 billion and net assets were about $1.44 billion. In the opposite direction, Bitcoin ETFs saw net inflows of $101.2 million that same day, after posting net outflows of $236.5 million the day before.

There are three things to remember about the mechanism.

First, ETF flows are a shadow of subscriptions and redemptions, not the full picture of buying. They include market makers’ hedges and basis trades, and the directional component has long been overestimated.

Second, the scale comparison is brutal. In August, Ethereum ETFs pulled in more than a billion dollars in net inflows, yet $ETH price still slid from above $2,500 to around $2,436. Tens of millions of dollars in daily net subscriptions are only a drop in the bucket relative to the daily turnover in spot and derivatives; pricing power sits on the side of perpetual leverage and macro expectations.

Third, why did $ETH , $XRP flow out while $BTC flowed in on the same day? When the probability of rate cuts is being repriced, marginal funds first withdraw from the least liquid leg, then return to the deepest pool. This is not a narrative switch, but a risk-budget switch, and the order is almost always the same each cycle.

Next, watch three signals: whether the 3-day rolling net inflow turns negative, whether ETHA sees consecutive redemptions, and whether $ETH can hold $2,400. An end to a one-day inflow streak is basically noise; what really matters is outflows continuing into the second and third day—that is when it becomes a trend.

Do you read this outflow as repositioning, or as retreat?

#EthereumXRPETFstreakEnds

$SOL The fact that SOL fell below $100 collided with a governance vote that had just passed on-chain, and it’s worth unpacking.

First, the data: in August, SOL rose about 44%, reaching around $110.5 on August 26; on September 2, it fell below the $100 round-number level, briefly dipping to $97–98, with a 24-hour drop of 3%–3.8%, nearly 10% below the monthly high. Meanwhile, Bitcoin was fluctuating between $77,000 and $80,000, which means this drop underperformed the broader market.

What’s unusual is that the fundamentals were actually improving. On August 28, Solana passed the SGP-0002 "double disinflation" proposal, raising the annual disinflation rate from 15% to 30%. It is expected to reduce issuance by about 18.9 million SOL over the next six years, equivalent to 2.6% of supply under the old plan. The vote result: 67% in favor, 25.16% against, 7.84% abstaining; staked participation was 60.7%, with 176.29 million votes for versus 66.19 million against—the threshold was 66.67%, so it passed by just 0.33 percentage points. Fund flows were also not weak: on September 1, U.S. spot SOL ETFs saw net inflows of about $10.19 million, marking the 11th straight trading day of inflows.

Tighter supply, inflows of capital, yet the price still falls—on the surface it looks contradictory, but it isn’t. The issuance cut is a slow-moving variable spread over six years, with an annualized impact of less than half a percentage point, so the short-term effect on valuation is almost imperceptible; ETF subscriptions of around $10 million a day are also just a drop in the bucket relative to SOL’s tens of billions in daily turnover. In the short term, pricing power is still in the hands of leverage, and the direct trigger for this decline was the concentrated liquidation of long positions from late August to early September. And the vote passing by just 0.33 percentage points also shows something else: this network still lacks broad consensus on how value is distributed, and disagreements will resurface during implementation.

Going forward, I’ll be watching three things: whether $100 can be reclaimed, whether support around $97 holds, and whether ETF inflows continue uninterrupted. Fundamentals improved first, prices fell later—would you see that as an opportunity or a warning?

#Solana falls more than 3%

Kuwait has been taking hits for two straight days; let’s first make the confirmed parts clear.

On September 2, Iran’s Revolutionary Guard said it carried out a combined missile and drone strike on Kuwait’s Ali al-Salem base, targeting the command center and residence of a U.S. military commander, and also claimed to have struck U.S. facilities in Jordan and Iraq. On September 3, Iran’s military said it had again hit Kuwait’s Ahmad al-Jaber base and a U.S. base in the United Arab Emirates. Kuwait’s General Staff publicly stated that air defense systems were responding to incoming missiles and drones; the Kuwaiti government condemned the attacks for two consecutive days, saying they were a blatant violation of its sovereignty; a residential area in the capital province caught fire after being hit by a drone, and later reports said the situation had been brought under control. The Iranian side said it had caused U.S. casualties, but that is a unilateral Iranian claim and should be subject to official confirmation.

The point of disagreement is actually very clear. Those who are bullish on risk assets believe spillover is manageable, passage through the strait is being restored, and historically this kind of tit-for-tat retaliation rarely truly changes the supply curve—in fact, by the next day risk assets had already recovered their losses, and $BTC rebounded from 77,000 to above 81,000. The bearish side believes the chain is not really “war -> buy crypto as a safe haven,” but rather “war -> oil prices -> inflation expectations -> the Fed”: oil above $90 and the 10-year U.S. Treasury yield touching 4.818% intraday on September 2, its highest since November 2023, are proof of that chain. When discount rates rise, $BTC is high beta, not a safe-haven asset.

There is an even harder layer of disagreement: if the conflict escalates to directly hitting Gulf energy facilities, then it shifts from an emotional shock to a supply shock, and in a stagflation scenario stocks, bonds, and crypto all get hit. That is the only tail risk worth hedging in advance.

Watch three things: whether energy facilities are directly hit, whether Kuwait and the UAE escalate their response, and the war-risk insurance premium curve. This time, do you treat $BTC as a safe haven or as a risk asset?

#Iran missile drone attack Kuwait base

The debate around the $80,000 line is really just one sentence: was it pushed up by demand, or squeezed up by shorts?

First, look at the numbers. In August, $BTC rose about 25%, the strongest single month of 2026, and the first August in five years to close higher. U.S. spot Bitcoin ETFs saw $3.52 billion in net inflows in August, their strongest month of 2026, while July saw only $172 million; cumulative net inflows rose to $54.92 billion. But the start of September brought a reality check: on September 1, net outflows were $236.5 million, and the price fell back to around $77,000, repeatedly rejected above $80,000. The tone shifted again on September 3-4: after Waller’s remarks, the probability of a September rate hike dropped from 60% to 48.4%, $BTC rose more than 5% in a single day and closed around $81,272, topping $82,000 intraday, more than $415 million in short positions were liquidated, and the total crypto market cap increased by about $112 billion to roughly $2.7 trillion.

The bulls’ case: $3.5 billion of monthly ETF inflows are real incremental buying, and once macro conditions ease, the level breaks; $80,000 is just a function of rate-hike probability.

The bears’ case: August delivered $3.5 billion and a 25% gain, yet price only made it back to around $80,000 — meaning selling pressure above this level absorbed the best funding conditions. And in this rebound, spot ETF daily inflows were still only in the $100 million range, while short liquidations were more than $400 million, making the structure look more like a squeeze than allocation.

My view: a level pushed up by a squeeze must be validated by continued spot net buying, otherwise it is just providing liquidity to sellers. Three things to watch: whether the $82,000 resistance can turn into support, the August CPI and the September 15-16 FOMC outcome, and whether $80,000 can hold on days without a macro tailwind.

Which side are you on? Is $80,000 the ceiling, or the new floor?

#BitcoinStrugglesToBreak80000

OpenAI released GPT-6 Astra on September 3, first to a small group of enterprise customers, then gradually to subscribers and API users.

As for benchmarks, there wasn’t much surprise: in third-party evaluations, it was only 2 points higher than the previous generation in the highest reasoning tier, while API pricing jumped from $4/$20 per million tokens to $10/$50, a straight 2.5x increase. The context window was raised to 1.05 million tokens.

What’s really worth discussing is in the safety documentation: OpenAI itself admits this is the first model to reach the “critical” cybersecurity tier in its preparedness framework — if given enough tools and permissions, it can uncover vulnerabilities no one had found before. So this time, it’s lock the door first, then ship it; capability and control are sold as a package.

On one side, there’s a 2.5x price hike with only limited gains; on the other, the vendor is proactively saying, “this thing is too offensive.” Do you think this is a real leap forward, or just a footnote to justify the price increase?

#GPT6Astra

The divide in the market over the attack on the Kuwait base is not about whether the strike happened, but about the gap between what was claimed and what was confirmed. After the U.S. launched a new round of airstrikes on targets of the Iranian Revolutionary Guard, the Revolutionary Guard claimed it had used missiles and drones to strike the U.S. command area at Kuwait's Ali Al Salem Air Base, saying that “multiple U.S. personnel were killed,” that a drone hangar was burned down, and that the attack was carried out in sync with actions against U.S. bases in Jordan, Bahrain, and Erbil. The next day, Iran's military announced strikes on the communications systems, equipment depot, and hangars at Kuwait's Ahmad Al Jaber Base, as well as the radar at the Al Minhad Base in the UAE. But the Kuwaiti government only acknowledged that air defenses fired in the early hours to intercept the attack, and reported no casualties or property damage. The U.S. also did not confirm any personnel losses at any of the bases mentioned in the two rounds of claims. That gap in messaging is the source of volatility: the bullish camp on oil prices takes Iran's claimed results as evidence of escalation, believing that retaliation is spreading from Iraq and Jordan to Kuwait and the UAE in the Gulf, and that a cutoff in the Strait of Hormuz is only a matter of time; the bearish camp points out that Iran's retaliation this time targeted U.S. bases rather than shipping, while on Tuesday the U.S. escorted 40 merchant ships and about 18 million barrels of crude passed through the strait, with the White House saying traffic had already returned close to pre-war levels. The supply-cut narrative lacks data support, and the risk premium will eventually fade. Both sides have a point; the key is which signal gets confirmed first. What really needs close attention are three things: whether the U.S. releases a casualty assessment and escalates its response, which would be the clearest trigger for escalation; actual traffic volume through the Strait of Hormuz and tanker insurance premiums, which are more honest than any statement; and whether Iran shifts its focus toward Israel or Saudi oil facilities. If none of these three signals move, the war premium in oil prices is just hanging there; if any one is confirmed, risk assets will need to be repriced. #IranMissileDroneAttackKuwaitBase

Shenzhen school uniforms have made the trending list, and netizens are all shouting 'so good'—what’s good is not the looks, but the way it works: all primary and secondary schools in the city use a uniform blue-and-white sports style. Schools do not take part in design, bidding, or sales; parents buy them themselves from supermarkets, stationery stores, or even delivery platforms. The cheap ones cost less than 20 yuan, and there’s no need to buy them again when moving up to the next school or transferring.

In many places, the school-uniform ecosystem is quite different: each school has its own style, the school handles bidding, the price is high and the quality is poor, and there is still room for利益输送. Shenzhen’s approach is to reduce the power involved: the education bureau only sets standards and supervises quality. Each uniform has a regulatory code for traceability, more than a hundred companies compete on the same stage, and parents vote with their feet.

Some people complain that a unified style suppresses individuality. That’s only half true—school uniforms are first and foremost clothes, and only secondarily a symbol. When school uniforms are tied to an interest chain, what children wear is actually the parents’ frustration. Do you think this homework of 'unified standards + open market + strict supervision' can be copied by other cities? How much does a set of school uniforms cost where you live?#深圳校服

CFTC asks the court to dismiss CME's lawsuit. This case is the best example for understanding the direction of U.S. crypto derivatives regulation. Here's the background: On May 29 this year, the CFTC approved Kalshi to list Bitcoin perpetual contracts and classified them as “futures,” while also stating that other designated contract markets (DCMs) could likewise list similar products — effectively opening a regulatory door for “crypto perpetuals = futures” for the first time; on June 18, CME sued the CFTC in federal court in Washington, D.C., with the core argument that perpetual contracts should be classified as “swaps” under the Commodity Exchange Act and the Dodd-Frank Act, and that the CFTC sidestepped the regulatory classification in order to approve them, effectively allowing a new competitor into its own retail futures turf. On September 2, the CFTC filed a motion to dismiss, with layered arguments: first, CME lacks standing — as a DCM itself, it could already list perpetuals, so the alleged competitive harm is “self-inflicted,” and CME had publicly said customers were not asking for such products; second, the data does not support the claim of harm — CME's $BTC - and $ETH -related futures trading volumes in both June and August were higher than in May, the month the approval was granted; third, even if perpetuals were reclassified as swaps, platforms like Kalshi could still continue offering them under the new classification, so CME's alleged harm still would not hold up; the CFTC also added that this lawsuit turns the Commodity Exchange Act's legislative purpose of “encouraging innovation and fair competition” on its head. Why does this matter? The classification dispute determines which channel U.S. crypto perpetuals will take: if they remain classified as futures, more DCMs will follow; if they are ruled swaps, they will have to enter a stricter swap regulatory framework. Next, watch for CME's deadline to file its opposition on October 2, and the court's view on the standing hurdle. #CFTCrequestsdismissalofCMEperpetuallawsuit