yangjun

#参议院拟9月表决CLARITY法案 U.S. Senate to vote on the CLARITY Act in September; encryption regulation faces a key window

Early on August 8, Senate Majority Leader John Thune filed a procedural motion to lock the debate-ending vote for the “Clarity Act” to September 15. This means that after failing to clear the hurdle before the August recess, the bill will be put on “pause,” then return at the September reconvening directly to the procedural voting track, becoming the final legislative window within 2026 before year-end.

The CLARITY Act is intended to establish the first federal-level market structure framework for U.S. crypto assets. Its core goal is to clarify whether digital tokens are securities or commodities, and to draw the regulatory boundary between the SEC and the CFTC. It also covers stablecoins, DeFi front-end requirements, and platform listing/registration rules. The bill passed the House in July 2025 by a bipartisan vote of 294–134, and in May 2026 it was advanced by the Senate Banking Committee by a vote of 15–9. However, the full Senate vote has repeatedly been stuck in partisan wrangling.

The September 15 event is not the final legislative vote, but the procedural threshold to end debate with “60 votes.” Republicans hold 53 seats, so they need at least 7 Democrats to break ranks. There are three major obstacles right now: first, federal officials’ ethics rules—Democrats are pushing to limit profits for the president and members of Congress from crypto projects, directly targeting businesses linked to the Trump family; second, whether “interest-like” payments for stablecoins could affect community bank deposits; and third, integrating DeFi anti–money laundering provisions with the text from the Agriculture Committee. A White House crypto adviser has warned that if there is no breakthrough before September 15, the bill’s prospects will be significantly narrowed afterward.

Market sentiment has cooled in parallel: on Polymarket, the probability of passing the law within 2026 has fallen from over 70% in May to about 14%. Even if the September motion fails, the SEC and CFTC could continue to push classification-based regulation through administrative guidance, but the “certainty-focused legislation” the industry is counting on will be delayed again. September 15 is not the endpoint, but the test of whether the United States can turn crypto regulation from enforcement-driven chaos into written law.

#伊拉克石油出口下降75% Iraq’s Oil Exports Plunge 75%: The Plight of a Single Sea Lane and a Warning for Energy Security

On August 8, 2026, Iraq’s Oil Minister, Bassem Mohammed Hudaier, publicly stated that due to the closure of the Strait of Hormuz, the country’s oil export volume has fallen by 75% compared with before the conflict. Before the war, Iraq exported about 3.4 million barrels of crude oil per day through the strait; now this “lifeline at sea” is nearly cut off. Oil tankers at the southern Basra port can no longer depart in an orderly manner, and the nation’s main fiscal artery has been dealt a severe blow.

The root cause of this cliff-like drop is the spillover of the U.S.-Israel-Iran conflict, which repeatedly disrupts navigation through the strait. Roughly 88% of the Iraqi government’s revenue depends on oil, yet the export route is highly concentrated along the Hormuz line, with few alternatives such as east–west pipelines or land-based pipeline networks. Once the strait is sealed, inventories pile up rapidly. Southern fields are forced to cut output, exposing the country’s economic vulnerability in full.

Faced with this predicament, Iraq is trying to break out on two fronts. On the one hand, it is negotiating with Iran to seek a passage exemption, though coordination has not yet taken effect. On the other hand, it is accelerating export diversification—restarting the land route at the Iraq–Syria Rabia border crossing as an emergency measure to clear stockpiles, and planning a new Basra–Feshkhabour pipeline to bypass the strait. Hudaier has clearly emphasized that “achieving diversification of oil export channels” has become an urgent priority.

Iraq’s experience reflects a shared weakness among Gulf oil producers: about one-fifth of the world’s seaborne crude oil must transit through the Strait of Hormuz. Any geopolitical shift, however small, is directly transformed into supply disruptions and oil-price volatility. In the short term, the sharp export decline will weigh on Iraq’s finances and intensify domestic hardship. In the long term, it forces a reshaping of Middle East energy infrastructure and also reminds consuming countries that energy supply chains overly concentrated in a single corridor have almost no margin for error in wartime.

#东证拟设重大业务变更再审查制度 East Securities plans to establish a system for re-examining major business changes: shifting from "once-and-for-all clearance" to "dynamic post-checks"

Recently, East Securities has signaled to the regulators through external communications that it intends to set up a "system for re-examining major business changes." Under this plan, matters such as adjustments to business scope, consolidation of subsidiaries, launch of innovative businesses, and implementation of mergers and restructuring will be included in a review framework that combines periodic and trigger-based post-checks. This move is not an isolated fix to internal controls, but a response to the bottom-line requirement in Article 122 of the Securities Law, which stipulates that brokerage firms’ major matters—including changes to business scope and mergers, divisions, or other corporate restructuring—must be approved by regulators. In essence, it extends "having obtained regulatory approval once" into "internal continuous calibration."

The core logic of the system is "dual-track re-examination." On one hand, it will conduct an annual review of existing major businesses for compliance, risk, and capital-match alignment to prevent business drift or a disconnect between business operations and risk controls. On the other hand, it will establish trigger mechanisms: once it involves circumstances such as a proposed acquisition of 100% equity in Shanghai Securities, changes to the actual controller or major shareholders, capital adequacy constraints reaching critical thresholds, or any issues in customer funds segregation, a special re-examination will be initiated, and no further stage may proceed without passing the review.

Against the backdrop of East Securities advancing its restructuring with Shanghai Securities, while operating in an industry where capital scale is among the top ten, the substance of this system is to bind "getting bigger" with "getting real." It not only meets the rigid constraints of the CSRC’s prudent supervision and business-scope approval, but also conveys a governance posture to the market that "expansion does not equal losing control." For the industry, this is a snapshot of leading brokerages shifting from "seizing territory at full speed" to "internal-control premium." For investors, major business changes face an additional internal brake, which means the space for information asymmetry and aggressive games is compressed. Re-examination is not the enemy of efficiency; it is a safety chain that locks the franchise value of brokerages onto a compliance track.

#黄金突破下行趋势线 Gold Breaks Through the Downtrend Line: A New Window of Technical Reversal and Capital Resonance

On August 5, spot gold surged more than 4% in a single day. It strengthened and broke above the $4,200 per-ounce level, decisively penetrating the downtrend line that had been suppressing price action for months since the historical high of $5,598 in January 2026. It also regained the 20-day and 50-day moving averages at the same time, completing the key technical shift from “converging and oscillating” to “turning stronger.”

This breakout is not an isolated chart event. On the macro front, the U.S. July ADP employment report rose by only 44,000, well below expectations; the U.S. dollar index slipped below 100; and market expectations for a September Fed rate hike fell to 45%. The opportunity cost of holding gold was therefore compressed. Meanwhile, central banks in Poland and South Korea restarted gold purchases after years, and China’s central bank continued to add to its holdings—providing a structural bid to support demand. On the capital side, there is an even stronger “accelerator”: previously, CTA trend funds had accumulated a net short position according to their models. After gold broke the line, short-covering was triggered, creating a positive feedback loop—“breakout → short covering → chasing longs”—which further amplifies the elasticity via systematic buying.

In terms of levels, $4,200 has turned from resistance into the first support. If pullbacks do not break it, the breakout will be confirmed as valid. The $4,300–$4,400 range above corresponds to the 52-week moving average, the 0.382 Fibonacci retracement (around $4,333), and a historical turning zone—this is the next target area for long positions. If the daily close holds above $4,200, Deutsche Bank expects $4,700 by year-end and JPMorgan expects an average of $4,500 in the fourth quarter. Conversely, if price falls back below $4,070 (the original trend line), the “breakout” would be invalidated and gold would likely return to the $4,000–$4,200 trading range.

In the short term, RSI has entered a strong zone and the 4-hour chart is overbought, reducing the value of chasing higher prices. A better strategy is to wait for a pullback and scale into positions along the $4,200–$4,255 support band. This gold breakout reflects a triple convergence: technical clearing, macro easing, and central bank gold buying. However, key uncertainties—such as the Non-Farm Payrolls data and Middle East variables—have not been resolved. The trend has begun, but timing matters more than ever.

#SpaceX首个锁定期8月6日到期 SpaceX’s first lock-up expires on August 6: a pricing trial amid a trillion-dollar unlock wave

On August 6, 2026, U.S. Eastern Time, the first insider lock-up period following SpaceX’s listing (ticker: SPCX) officially comes to an end. This is also the second trading day after the first quarterly earnings report (released after the close on August 4) following its “largest IPO in history,” which priced in June at $135 per share. Under the stepwise unlock rules designed in the prospectus, up to about 911.5 million shares held by employees and Pre-IPO investors are allowed to trade. Based on the share price of around $108 at the time, this corresponds to a market value of about $116 billion, or 1.4 times the existing float (about 639 million shares).

Unlike the U.S. stock market’s typical “all at once after 180 days” approach, SpaceX deliberately staggers the release. On August 6, roughly 20% will be freed first, followed by about 7% each for batches spanning 70 to 135 days. After the third-quarter report, another 28% will be released, so that by early December—the 180-day mark—the float will expand to 5.33 billion shares, about 40% of total shares outstanding. If, in the first 10 trading days before the earnings release, the stock closes above $175.50 (the offering price plus 30%) on 5 days, an additional 455.8 million shares could be released—but that condition has currently been missed. Musk himself, with roughly 6.4 billion shares locked until mid-2027, is not included in the selling pressure.

The market has already voted with its feet: the stock price has fallen from the June 16 peak of $225.64 to around the $108 level, slipping below the offering price. Short positions account for roughly 30%–34% of the float. “Unlocking” doesn’t necessarily mean an automatic sell-off: early VC costs were very low and many holders have strong incentives to cash out. Still, employee tax planning and long-term believers will likely cushion any selling pressure. The real pricing power will come in the three trading days after August 6—Starlink cash flow in the earnings report, capitalized expenditures for Starship, and AI compute contract guidance will determine how much of that $116 billion “paper supply” becomes real sell orders.

For the AI trillion-dollar IPOs that OpenAI and Anthropic are waiting to file for, this week for SpaceX is a dry run for the public market: when story premium collides with insiders’ exit window, where should the valuation anchor—on dreams or on free cash flow? August 6 will deliver the first set of data.

#伯克希尔股价创八个月新高 Berkshire Hathaway’s Class A and Class B shares rose together this week, both hitting their highest closing levels in eight months. Class B shares closed at $512.37, their best close since November 28 last year; Class A shares closed at $768,010, about 5.3% below the all-time closing high of $809,350, but well off their earlier lows.

The key driver behind this catch-up rally is the resonance between “defensive assets + heavyweight holdings.” As technology stocks came under pressure, funds rotated into Berkshire, a diversified defensive leader spanning insurance, railroads, and energy. Although Berkshire still lags the S&P 500 by about 8 percentage points year to date, that very gap has created room for a rebound. Barron’s said its “rally still has room to run.”

On the portfolio side, three major holdings have provided strong support this year: Apple’s stake is worth more than $70 billion, and Apple itself is up more than 13% year to date; Coca-Cola is up about 25%; Bank of America has gained more than 12%. UBS accordingly raised its Class A target price to $877,848, kept a “buy” rating, and lifted earnings expectations, citing lower-than-expected second-quarter catastrophe losses and improved BNSF railroad profitability.

The balance sheet tells a compelling story as well: as of the end of the first quarter, cash and short-term investments totaled about $397.4 billion, a record high, providing both a “war chest” and an interest-income cushion. Market estimates suggest second-quarter share repurchases may have reached $5 billion to $11 billion; if confirmed, that would be the strongest signal yet after Abel takes over, with the exact figure to be revealed in the second-quarter report on August 8.

On valuation, Class A shares trade at about 1.4 times book value, below the 1.8 times level seen in May 2025 and near the lower end of the recent range. With “nearly $400 billion in cash + a return of buybacks + strength in core consumer and financial holdings,” Berkshire is shifting from an “underperforming value discount” to a “defensive premium re-rating” — the new eight-month high is not the endpoint, but the starting point of the first repricing in the Abel era.

#对冲基金加码原油多头 Hedge funds increase bets on long oil positions: a shift in holdings amid a reassessment of geopolitical risk premium

The latest data from the U.S. Commodity Futures Trading Commission (CFTC) shows that for the week ending July 28, 2026, fund managers boosted their net long WTI crude oil positions by 21,402 lots to 108,307 lots in a single week—an increase at the fastest pace since March. Bullish sentiment rose to the highest level since mid-June. Meanwhile, net long gasoline climbed to a four-month high, diesel longs approached a five-month peak, and refined products and crude oil formed a synchronized bullish structure.

This round of adding positions is not a broad-based bet on rising global oil prices, but rather a typical differentiated wager of “overweight the U.S., underweight Brent.” In the same week, net long positions in Brent decreased slightly by 6,948 lots to 185,083 lots. The underlying logic is that three major energy arteries—the Strait of Hormuz, the Red Sea-Mandeb Strait, and the Black Sea—are simultaneously constrained, sharply increasing uncertainty over spot cargo flows in the Middle East and the North Sea. By contrast, WTI, backed by shipments through the Gulf of Mexico and pipelines within the United States, is viewed as the only “reliable supply buffer” with the freedom to flow globally. Funds are willing to pay a scarce premium for it, even giving rise to WTI-vs-Brent spread trades that go long WTI and short Brent.

The transmission chain is clear: escalation in the Iran-U.S. conflict, with transit volumes through the Strait of Hormuz falling to a three-week low, quickly flipped the market from “worrying about oversupply” to “scrambling to cover shorts.” WTI rose about 6.8% in the week, Brent hovered near the $90 level, and near-month contracts kept trading at a backwardation/negative spread, meaning maintaining long positions on a rolling basis could yield positive returns. But the risk is that geopolitical news is highly volatile and changeable. Implied volatility in options has surged. After CFTC top players concentrate their positions, any easing of strait navigation or a substantive production increase from OPEC+ could trigger an equally aggressive long liquidation and profit-taking.

Overall, hedge funds increasing long oil exposure is fundamentally about repricing “geopolitical risk plus U.S. supply certainty,” rather than asserting a rebound in demand. Going forward, managing volatility in the market may matter more than directional judgment in determining outcomes.

#韩国拟暂停可疑加密账户支付 South Korea plans to suspend payments from suspicious crypto accounts: from “post-facto asset recovery” to “freeze during the process”

On July 28, 2026, 15 people including Rep. Kim Sang-hoon of South Korea’s People Power Party submitted amendments to the Act on Specified Financial Information to the National Assembly. For the first time, the amendments clearly define “virtual asset accounts” as the “unique identifier issued by an exchange to a user,” and grant the Financial Intelligence Unit (FIU) unilateral authority to suspend payments: if an account is deemed to be involved in the transfer of assets illegally, the FIU may require the platform to stop payments for 30 days, extendable once (up to 60 days). If the platform refuses to comply, it faces a maximum fine of 100 million won. The bill takes effect six months after its announcement.

This mechanism compresses the previously court- and prosecutor-led process—where freezing on-chain assets required a case filing and a court warrant—into an administrative order delivered directly to exchanges, skipping the judicial prerequisite. In effect, it installs a gate at the exit for “kimchi-plate” (retail) funds.

The backdrop is South Korea’s ongoing crackdown on anti–money laundering: In March 2026, the FIU fined Bithumb 36.8 billion won for missing KYC requirements and partially suspended operations for six months; in April, Coinone was fined 5.2 billion won for 70,000+ cases of identity verification failure and suspended new user deposits and withdrawals for three months. In the same month, the Financial Supervisory Service (FSS) tightened rules on pausing suspicious PG (payment gateway) virtual account transactions. In May, the FIU had proposed requiring that any cross-border transfers exceeding 10 million won be reported as suspicious transactions. However, because the number of exchange alert reports surged by 85 times (from 63,000 to 5.44 million), it reversed course and shifted to risk assessments by platforms. Yet the account-level freeze power has now been solidified in legislation.

For the crypto market, the costs for South Korean retail users—who often rotate funds rapidly and transfer across multiple platforms—will rise sharply. For the global signal, this points to a transition in East Asia’s crypto regulation from “taxation + licensing” to “controlling accounts + freezing payments.” After the six-month grace period ends, compliant transaction records, clean addresses, and minimizing interactions with unregistered offshore exchanges will move from “good practices” to “a must to prevent freezes.”

#IonicDigital纳斯达克首日涨26% Ionic Digital Nasdaq first-day jumps 26%: a new AI-computing power tycoon grown from the ruins of Celsius

On July 28, 2026, Ionic Digital (ticker: IOND), formed from the restructuring of Celsius Network’s bankrupt mining business, listed on the Nasdaq via a direct listing. It opened at $50, closed at $62.90, up 25.8% (around 26%) from the opening price. Its market value surged to $2.8 billion, becoming the largest direct-listing case for a U.S.-listed company since 2021.

The company has a unique origin: established in January 2024 to assume Celsius Mining’s assets, it issued about 37 million shares of Class A common stock directly to Celsius’s bankruptcy creditors. The shares soared on the first day—at its core, the move gave creditors, who had waited for two years, a piece of liquidity they could monetize. This time, it did not issue any new shares or raise new funds. JPMorgan Chase served as the financial adviser, and up to 10.8 million shares of existing stock could be resold.

What the market is buying isn’t Bitcoin mining—it’s the story of “turning mining sites into AI data centers.” Ionic converted a 234-megawatt power infrastructure in Ward County, Texas into HPC/AI data centers and leased them to Nscale. Under a 10.5-year lease, contract revenue totals $1.95 billion (potentially rising to $2.6 billion after expansion). In 2026, revenue is expected to be $190–195 million, with about 90% coming from infrastructure leasing. As of the end of March, it still held 2,815.6 BTC (about $192 million) and had zero interest-bearing debt.

After the first day, the stock fell 6.5% to $58.8 in the after-hours session, reminding the market that creditor-unwinding selling pressure, an arms race for AI data center capacity, and execution falling short of expectations are all variables hanging over the $2.8 billion valuation. But no matter what, this “bankruptcy claims → public equity” closed loop provides a scarce exit channel for capital submerged in the crypto winter.

#美国国债收益率回落 U.S. Treasury yields fall: a phased recovery amid cooling inflation and oil prices

In late July 2026, U.S. Treasury yields saw volatile declines. The 10-year benchmark retreated from above 4.70% to around 4.62%, the 2-year yield slipped to roughly 4.32%, and the 30-year yield also fell in tandem to around 5.12%. The spread between 10-year and 2-year yields narrowed to about 32 basis points, showing a classic “bull flattening” pattern.

The drivers are concentrated in two main lines. First, energy-driven deflation squeezed out the inflation premium. Signals became calmer after Iran–U.S. and over the Strait of Hormuz shipping-related developments. WTI dropped more than 7% in a single day, while Brent eased to around $86. This, combined with June CPI year-on-year falling to 3.5% and PPI month-on-month at -0.3%, led the market to quickly roll back pricing of an “oil price out of control → rate hikes resume” scenario. The probability of a July FOMC rate hike was cut from above 40% to about 10%–15%. Second, repricing of short-end expectations: the 2-year yield’s decline was slightly larger than that of the long end, indicating that the bond market was mainly digesting “no near-term rate hikes,” rather than betting on the start of an easing cycle.

In terms of transmission, falling yields temporarily eased discount-rate pressure on long-duration growth stocks (AI and semiconductors). Gold rebounded, the U.S. dollar index weakened, and emerging-market currencies and offshore Chinese tech stocks gained valuation breathing room. However, if the Middle East situation repeatedly flares up and oil rebounds, the long end is likely to give back its gains quickly.

What remains clear is that this is still a tactical rebound under a backdrop of “cooling inflation + supply constraints,” not a turning point toward easy policy. Fed Chair Powell continues to emphasize “zero tolerance for inflation.” The 30-year yield holding near 5.1% reflects that medium-term constraints from fiscal supply and term premia have not disappeared. In one sentence: the short end can exhale a bit, but the long end isn’t ready to celebrate yet.

#全球央行权衡油价逼近百美元 2026 July 23, Brent crude oil intraday broke through $100/barrel (spot price on the 24th was $100.69), the first time above 100 since late May, up about 40% in the past 20 days. The trigger was the escalation of the US-Iran conflict + Houthi attacks on Saudi oil tankers in the Red Sea, putting the Strait of Hormuz and the Bab el-Mandeb Strait “dual channels” under pressure at the same time.

The moment oil prices broke 100, the script for global central banks was rewritten—shifting collectively from “when to cut rates” to “will they raise rates again.”

Federal Reserve: At the July 28–29 policy meeting, keeping the 3.50%–3.75% range unchanged is still the base case, but CME data shows the probability of a September rate hike has surged from 53% a week earlier to 82%; even the probability of directly raising rates by 25bp next week has risen to around 35%. US June CPI rose 3.5% year-on-year, with core CPI at 2.6%. With oil adding fuel to the fire, the rate-cut narrative has basically gone out.

European Central Bank: On July 24 it stayed put (deposit rate 2.25%), but Lagarde candidly admitted that “an internal discussion about raising rates” had taken place, leaving September as an option and warning that second-round energy effects will keep eurozone inflation above 2% through the first half of 2027. Markets have already priced in two more rate hikes this year.

Bank of England: The 10-year UK gilt yield has held above 5% for nearly two decades, a record. Next week’s policy meeting is very likely to stay unchanged, but easing expectations have been cut in half.

Bank of Japan: Inflation has rebounded for the first time in three months, the 2-year government bond yield hit a 31-year high, and policymakers are sounding more relaxed about “accelerating rate hikes,” but a weaker yen continues to tie their hands.

People’s Bank of China: “China’s policy should be based on our own conditions” + stronger exchange-rate flexibility to hedge imported inflation; PPI is being hit by the oil-price pulse, but CPI transmission via domestic demand is weak. The probability of a direct rate hike this year is extremely low; the window for reserve-requirement cuts or rate cuts depends on third-quarter fiscal bond issuance pace, though external high rates are squeezing room for easing.

The essence is a dilemma: hike rates to fight inflation and risk triggering stagflation; don’t hike and allow oil prices to pass through again, which is even more troublesome. Global bond markets first “fell” in respect—10-year German bund yields broke 3.21% (highest since 2011), French bonds broke 4%, and the US 10-year moved toward 4.68%—with markets voting for “higher for longer” through yields.

Over the next three weeks, watch three things: whether the US and Iran leave room for negotiations, actual traffic through the Strait of Hormuz, and whether the July FOMC statement treats oil prices as “one-off” or “persistent” — if the latter is confirmed, the “inflation + high interest rates” pricing regime for global assets will be re-anchored.

In the Senate consolidated draft of the CLARITY Act unveiled by Lummis on July 22, #CLARITY法案拟奖励白帽黑客 2026, a previously low-profile cybersecurity provision has come to light — a proposal to establish a white-hat rewards program through a “digital asset cybersecurity coordination mechanism.” This is not about encouraging bounty hunters to break into systems at will, but about incorporating Web2’s mature bug-bounty disclosure mechanism into the text of a U.S. federal crypto market structure law for the first time.

The logic of the provision is clear: security researchers who, through authorized channels, discover and responsibly disclose vulnerabilities in exchanges, custody systems, wallets, smart contracts, cross-chain bridges, clearing and settlement, private key management, and other infrastructure, may receive rewards after verification and after sufficient time is allowed for remediation. It also clearly defines the legal boundary between “security research” and “malicious intrusion.” The idea follows former CFTC Chairman Giancarlo’s advocacy that “market resilience comes from transparent disclosure,” taking consumer protection one step further than simply preventing platforms from misappropriating customer assets — to preventing system vulnerabilities from wiping out customer assets overnight.

Why make this law now? The FTX and Celsius bankruptcies exposed the risks of “commingled ledgers,” while billions in losses from bridge and custody contract hacks over the past year have shown that relying only on companies to voluntarily offer bounties is not enough. The draft bundles customer asset segregation, bankruptcy isolation, anti-misappropriation rules, and white-hat incentives together, effectively adding a technical front-line defense to consumer protection.

But implementation still has gray areas: Will the reward pool be set by CFTC/SEC rules, or funded by exchanges? The scope of liability exemption, disclosure standards, and reward tiers have not yet been specified. The entire bill is still stuck at the Senate’s 60-vote threshold; if it fails to advance before the August recess, the white-hat provision could also be pared back in floor amendments.

If it ultimately becomes law, its significance goes beyond “hacked making money legally” — it would mark a shift in U.S. crypto regulation from “catching scammers after the fact” to “buying vulnerabilities in advance.” The roles of audit firms, insurers, and compliant custodians would all be revalued.

On July 23, #七巨头单日市值损失7970亿美元 2026 (Thursday), U.S. tech “Magnificent Seven” stocks (Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, Tesla) suffered their worst single-day selloff since the tariff storm in April 2025—erasing about $797 billion in combined market value. The Mag 7 index plunged 4.8%, the S&P 500 fell 1.21%, and the Nasdaq dropped 2.15%.

The immediate trigger was two earnings reports that tore open fears of an “AI money pit.” Alphabet’s second-quarter capital expenditures surged to $45 billion, and its full-year guidance was raised to as high as $205 billion. Free cash flow turned negative for the first time since its IPO, sending the stock down 7.13% and wiping out more than $290 billion in a single day; Tesla beat revenue expectations, but profits and EPS fell far short, and Musk bluntly said 2026 would be a “big capex year.” The stock plunged 14.52%, erasing about $200 billion in market value. None of the other five escaped: Amazon -4.57%, Meta -3.36%, Microsoft -2.24%, Nvidia -1.56%, Apple -1.30%.

On the macro side, the market was squeezed by the double blow of “oil above $100 + a resurgence in rate hikes.” As the U.S.-Iran conflict escalated and the Houthis attacked Red Sea tankers, Brent crude broke above $100, 10-year U.S. Treasury yields climbed past 4.7%, and the market pushed the probability of a September Fed rate hike from 68% to 80%, with high-valuation, long-duration tech stocks hit first.

This $797 billion loss was not a normal pullback, but a repricing of the market’s clock for “AI investment versus returns”: over the past three years, valuations were expanded on the back of a narrative; now it is time to deliver profits. The Magnificent Seven have already fallen 11% from their May highs, with roughly $2 trillion erased in total, but the AI infrastructure cycle has not reversed. It looks more like a trust run in the middle of a super bull market than the end of the story.

On the night of July 23, #原油突破100美元 2026, Brent crude September futures surged more than 6%, breaking above the $100 per barrel mark for the first time since May 22, while WTI simultaneously climbed to around $91. The supply panic triggered by the Middle East’s “dual-strait linkage” was fully ignited.

The trigger was straightforward: after the Houthi armed group announced a maritime blockade on Saudi Arabia, it struck the Saudi tanker "Ensaliya" sailing in the Red Sea with missiles and drones at dawn, sharply escalating risks in the Bab el-Mandeb Strait. At the same time, the Iran-U.S. conflict continued to intensify. Trump declared that "attacking ships will be charged to Iran," and U.S. forces carried out consecutive nighttime strikes on Iranian facilities. With both the Strait of Hormuz and the Bab el-Mandeb Strait under strain, Asian buyers have already begun discussing detour plans around Africa with Saudi Aramco.

The global transmission chain tightened instantly: the U.S. national average gasoline price exceeded $4 per gallon, the 10-year Treasury yield rose to a year-to-date high, and markets priced in the probability of a Federal Reserve rate hike in September jumping from 68% to 80%. The Nasdaq fell nearly 2%, Tesla dropped more than 12%, while energy and storage stocks rallied against the trend. The European Central Bank remained on hold but adopted a more hawkish tone, and imported inflation resurfaced.

For China, the impact is "controllable but structurally differentiated": CF40 estimates that if oil rises from $70 to $100, China’s domestic PPI peak would rise by about 1 percentage point, while the impact on industrial output and CPI would be limited, and RMB assets would show safe-haven resilience. In A-shares and Hong Kong stocks, the typical pattern is "upstream benefits, midstream and downstream face pressure"—the three major oil companies, oilfield services, coal, oil transportation, and the substitution logic of new energy are favored, while airlines (with fuel costs accounting for 30%+), logistics, downstream chemicals, and high-valuation tech stocks are hit from both rising rates and cost compression.

$100 is not the endpoint, but the starting point of risk-premium repricing. If the Strait of Hormuz were to be materially disrupted, Goldman Sachs sees Brent at $120+ in Q4, and RBC’s extreme scenario would challenge $146. But the baseline scenario remains "high-level volatility plus tail risks," with the key question being whether the U.S. and Iran leave room for negotiations over the next two weeks. For individuals, the clearest reminder of oil breaking above $100 is that travel costs, courier fees, and prices of chemical consumer goods will all quietly be rewritten within the quarter.

#香港存储概念股走强 Hong Kong storage concept stocks have continued to strengthen recently. This is not simply speculative fund-rotation; rather, it is the result of a triple logic convergence: an “explosion in AI compute demand + rising storage prices + the capitalization of domestic storage” .

In the early trading session on July 22, Hong Kong’s storage supply-chain stocks rose in tandem with A-share chip stocks. Nanfang 2x Long Hai Force sold (07709.HK) jumped by nearly 15%, Nanfang 2x Long Samsung Electronics (07747.HK) rose by more than 10%, GigaDevice (03986.HK) climbed over 3%, and Ruentex Technology (06809.HK) gained nearly 2%. Semiconductor Manufacturing International (00981.HK) and Huahong Semiconductor (01347.HK), among other foundry and interface-chip segments, also moved up in sync.

On the driver side, the most core factor is the continuous widening of supply-demand gaps. The amount of DRAM搭载 in an AI server single unit is 8–10 times that of a traditional server, while NAND demand is over 3 times. Original equipment manufacturers are allocating more than 70% of their advanced production capacity to HBM, tightening supply for general-purpose DRAM and NAND. Adata has warned that DRAM contract prices will rise again by 20%–30% in Q3 2026, while NAND will increase by 35%–40%. Data from JPMorgan shows that in May, DRAM prices rose about 14% month-on-month and flash memory rose about 26%. Major original manufacturers such as Micron have already locked prices via long-term agreements to around 2028.

Secondly, the capitalization of domestic storage has fueled market sentiment. The launch of CXMT’s (ChangXin) Science and Technology Innovation Board IPO and Yangtze Memory’s progress toward listing have driven repricing of related names: GigaDevice (second globally in NOR Flash; benefiting from an upturn in ASP for niche DRAM) and Ruentex Technology (a leading DDR5/HBM memory-interface chip company) have both been revalued by investors. Net southbound fund inflows exceeded HK$45 billion in a single week, with both domestic and international capital reinforcing each other as they rushed to buy hard-tech stocks in Hong Kong.

It should be noted that the sector’s short-term rally has already priced in part of the optimistic expectations. In early July, there were technical profit-taking pullbacks with sharp daily declines. In addition, rising prices transmitting positively to the consumer end may suppress demand. Looking at the long term, a storage super-cycle will most likely run through 2026, with tight supply-demand balance continuing into 2027–2028; however, volatility is expected to be significantly amplified.

#韩股KOSPI因科技股抛售下跌 Korean Composite Stock Price Index (KOSPI) has recently faced significant selling pressure. In the morning session, it briefly fell by more than 4%. The key trigger was a chain reaction caused by technology stocks having an excessively high weighting. Samsung Electronics and SK Hynix together account for about 60% of the KOSPI index’s weight; both slid by more than 5% in the morning, directly dragging the broader market down.

This decline is the result of multiple factors converging:

- Reassessment of earnings expectations: The market is concerned that AI compute demand may slow down temporarily, and that price increases for memory chips such as HBM have fallen short of expectations. Profit guidance from major players like SK Hynix came in below consensus, prompting profit-taking after “good news had already been priced in.”
- High-leverage liquidation: Many Korean retail investors use margin financing and leveraged ETFs. When the index falls, it triggers forced selling, creating a negative feedback loop of “selling more as it drops,” and even led to algorithmic circuit breakers being triggered multiple times.
- Foreign capital pullback: Amid global high interest rates and valuation adjustments, foreign investors—such as U.S. funds—have continued net selling. In early July alone, foreign investors net sold more than KRW 120 trillion, further tightening liquidity.

Although the Bank of Korea stressed that the fundamental outlook for semiconductors—supply should not exceed demand—has not changed, and SK Hynix’s chairman said that AI demand will double, deleveraging and sentiment-driven selling still dominate the market in the short term. Subsequently, the index’s decline narrowed to about 0.6%, suggesting that at lower levels, investors are competing to bet on an oversold rebound. However, adjustment pressure on high-tech stocks in the medium term remains.

#布伦特原油涨4.6% Brent crude oil surged 4.6% in a single day, closing at about $88.10 per barrel—its highest level in more than a month. This strong rebound is mainly driven by a sharp escalation of Middle East geopolitical risks.

The core catalyst is the escalation of the U.S.-Iran conflict and increased risks to navigation in the Strait of Hormuz. The market is concerned that oil shipping through the Persian Gulf could be disrupted (about one-fifth of the world’s crude oil passes through this bottleneck). This concern is compounded by military actions such as Iran’s attacks on facilities in neighboring Gulf states and consecutive U.S. airstrikes, forcing crude prices to quickly price in a hefty “war risk premium.”

The knock-on effects have spread to the macro level:

- Inflation worries: Soaring energy costs have reignited global inflation expectations, which may push the Federal Reserve and other major central banks in the U.S. and Europe to delay rate cuts and keep interest rates high for longer;
- Market differentiation: U.S. stocks in energy and defense/aviation-defense sectors benefit, while high-oil-consumption industries such as airlines and logistics, as well as technology stocks, face pressure;
- Spillover to people’s livelihoods: In China, the window for domestic refined oil price adjustments faces upward pressure, and logistics and travel costs are likely to rise accordingly.

In the short term, crude oil price trends are completely tied to developments in the Middle East. If sea lanes are effectively blocked or oil production facilities are continuously targeted, Brent could test levels above $90. If signals of easing tensions emerge, the unwinding of the risk premium could also trigger a sharp pullback, and market volatility would be significantly amplified.

#SK海力士三星海外市场下跌 Recently, South Korea’s storage-chip duopoly—SK hynix and Samsung Electronics—has suffered a significant slump in overseas markets. On July 13, 2026, SK hynix fell 15.37% in a single day, retracing nearly 40% from its June peak. On the same day, Samsung Electronics dropped 10.7%, with a cumulative retracement of more than 30%. The selloff also triggered a trading halt for the Korean KOSPI index, putting pressure on the global semiconductor sector.

The core driver of this decline is a mismatch between expectations and financial realities, intensified by capital flows. Earnings previews indicate that SK hynix’s operating profit in Q2 will be 60.4 trillion won. Although it represents a year-over-year surge, it comes in below market expectations of 65 trillion won. Because HBM is covered by long-term contract pricing, its average price increase is weaker than that of the spot market. As a result, profit upside fails to match that of peers, triggering profit-taking under the logic of “good news has been fully priced in.”

Macro and competitive factors have also come under pressure. Growing expectations of interest-rate hikes by the Bank of Korea weigh on high-valuation growth stocks. Meanwhile, foreign institutional investors systematically rebalanced their holdings and exited the AI “consensus trade.” At the same time, Chinese vendors are accelerating their rise: Yangtze Memory (NAND market share of 13%) and CXMT (DRAM market share of 7.7%). They continue to divert orders in mature process nodes and mid-to-low-end markets by leveraging cost advantages and supply-chain security, forcing Korean firms to shift toward high-end HBM—while the general-market share faces long-term erosion.

Although the logic of the AI supercycle remains intact, the market is bringing forward its pricing of concerns about capacity oversupply after 2028 and a potential peak in the cycle. In the short term, this looks like a technical correction driven by earnings and leverage. In the long run, it is a snapshot of the global memory industry reshaping from a “US-Korea duopoly” toward “multipolar competition.”

#科技股拖累美股走低 On July 13 in U.S. Eastern Time, the three major U.S. stock indexes all closed lower, with tech stocks becoming the core force dragging the broader market down. At the close, the Dow Jones Industrial Average fell 138.37 points, or 0.26%, to 52,498.64; the S&P 500 fell 60.05 points, or 0.79%, to 7,515.34; and the Nasdaq Composite plunged 408.43 points, or 1.55%, to 25,873.18.

Large-cap tech stocks were mixed: Microsoft, Amazon, and Apple posted slight gains, while Tesla and Nvidia fell more than 3%, and Google and Meta dropped more than 1%. Chip stocks faced concentrated selling, with the Philadelphia Semiconductor Index tumbling 4.78%, making it the main area hit in this round of correction. Arm fell more than 7%, Intel dropped more than 6%, and AMD and Micron Technology declined more than 4%; storage and optical communication sectors also slumped in sync, with SanDisk plunging more than 12%, SK Hynix ADR falling more than 9%, and Astera Labs dropping more than 12%.

This sharp sell-off in tech stocks was mainly driven by the dual pressure of geopolitical conflict and rate-hike expectations. On the one hand, the escalation of the U.S.-Iran conflict sent international oil prices soaring, with WTI crude futures closing up 9.42% and Brent crude up 9.59%, significantly hurting market risk appetite; on the other hand, the surge in oil prices intensified inflation concerns, and Federal Reserve Governor Waller sent a hawkish signal, saying that if core inflation pressure persists, interest rates may need to be raised in the near term, causing market expectations for rate hikes to rise rapidly.

Against the backdrop of rising interest rates and geopolitical turmoil, high-valuation growth stocks such as AI and semiconductors, which had accumulated substantial profits earlier, faced profit-taking. The market has shifted from the "growth narrative" to a stricter focus on "earnings validation," and volatility in the tech sector may continue to expand in the short term.

#韩国7月强制平仓达3442亿韩元 In July, South Korea’s forced liquidation volume reached KRW 344.2 billion, reflecting a credit-market crisis in the Korean stock market driven by highly leveraged trading. According to data from the Korea Financial Investment Association, as of July 9, the cumulative amount of forced liquidations for the month had already reached KRW 344.2 billion. Of this, forced liquidations on July 9 alone totaled as much as KRW 142.2 billion—an almost fivefold surge from the previous day—setting a new one-month high.

The direct trigger for this wave of margin calls and “head-cutting” liquidations was panic-driven plunges in the South Korean stock market. On July 13, the KOSPI index closed down 8.95%, breaking through the 7,000-point threshold and triggering the seventh circuit breaker event of the year. Semiconductor heavyweight stocks—Samsung Electronics and SK Hynix—fell sharply by 10.7% and 15.37%, respectively. The rapid drop in share prices caused retail investors’ margin accounts to fall below required collateral maintenance ratios, prompting brokers to carry out large-scale forced liquidations. This created a negative feedback loop of “falling prices—deleveraging—falling again.”

The deeper cause lies in the overheated leverage risks that had built up in the market. South Korea previously launched multiple 2x leveraged ETFs tracking semiconductor giants, attracting large numbers of younger retail investors to enter at high levels with leverage; some even used household loans to invest. As expectations for AI-related semiconductor earnings were scaled back, the “daily rebalancing” mechanism of leveraged ETFs turned into passive selling during the decline, amplifying the stampede effect. Statistics show that in June, more than 1.2 million accounts across the market had already touched margin call thresholds; around 300,000 accounts saw their principal wiped out, and people aged 20 to 30 accounted for more than 60%.

Because forced liquidation data has a two-trading-day lag, the clearing pressure stemming from the nearly 9% plunge on July 13 has not been fully released yet. The liquidation scale likely to be disclosed next will probably rise further, and the market’s ongoing pain from deleveraging is expected to continue.