GiGiZ 發財豬

Polymarket’s $200M trades allegedly suspected of insider trading

Polysights marked around 34,000 potential insider-trading transactions from August 2025 to June 2026. In the first half of this year, the suspicious amount was about $200 million. The most noticeable abnormal growth was in markets related to geopolitics and war.

The patterns are extremely consistent: newly created accounts, low-probability entries, large bets. In 57% of the related wallets, the accounts were created less than 24 hours before the trades.

Across 38 linked addresses, bets were placed in 90 geopolitics markets related to Iran and Venezuela. The win rate was 98%, for total profits of $1.6 million. All withdrawals went to the same充值 address.

A 98% win rate is statistically almost impossible to be luck. This isn’t a difference in analytical ability—it’s information asymmetry.

Yesterday I wrote about the World Cup with 31 BTC—where I said the transparency of prediction markets is selective. This dataset makes the issue clearer. On-chain activity can be traced, but after funds are mixed, the source of the money is obscured. By the time institutions like Polysights flag it, the whole cycle may already be over—months have passed.

The concentration of profits also tells a story: the top 1% of wallets take more than half of the returns. A healthy prediction market should be the “wisdom of the crowd” competing through a dispersed structure. Once it becomes this kind of setup, fundamentally it’s a small group with an information advantage extracting money from retail traders.

Polymarket has turned nearly 100 wallets over to law enforcement. Kalshi strengthened identity checks and restricted political candidates and athletes from participating in related markets. This direction is right, but the timing is clearly lagging behind the problem. Insider trading appears to have started as early as August 2025, while the data report was released in 2026—an entire one-year window.

The core value proposition of prediction markets is aggregating real information into price signals. Information advantages are reasonable, but information from public analysis and from insider information are two different things. If this line gets blurred, the purpose of market pricing itself is lost.

Regulatory gaps won’t be solved simply because platforms voluntarily cooperate with law enforcement. $200 million is only what has been flagged—how much more hasn’t been flagged is anyone’s guess.

DYOR Not investment advice

#Grayscale Launches a Worldcoin ETF Application
Grayscale has submitted an ETF registration filing for WLD to the SEC. This is another step forward after BTC, ETH, Dogecoin, SOL, and Chainlink

The direction is very clear: they are systematically packaging crypto assets into the ETF “shell,” pushing forward one by one. But with this WLD application, I think it’s the most controversial one among this batch

First is the scale issue. The current market cap of $WLD is about $1.3 billion, ranking 57th among crypto assets. By comparison, SOL’s market cap is dozens of times larger, and ETH is in an entirely different order of magnitude

Putting an ETF around an asset with a $1.3 billion market cap and limited liquidity won’t face small compliance resistance at the regulatory level. The SEC’s requirements for the depth and liquidity of the underlying assets are real hurdles. Whether WLD can get approved is, in terms of probability, not as optimistic as BTC and ETH ETF applications

Second is the controversy surrounding Worldcoin itself. Sam Altman’s endorsement is a double-edged sword @sama
OpenAI’s halo gives the project instant attention from the moment it was created, but Worldcoin’s core business is verifying human identity via iris scanning—something that has already triggered data privacy regulatory investigations in Europe and multiple markets

A biometric identification project facing compliance pressure in multiple countries now wants to enter the U.S. capital market in the form of an ETF. The dimensions of regulatory scrutiny are not just about crypto assets; it also has to clear privacy regulations

But interestingly, Grayscale daring to file this application in itself is a signal. It’s not saying the application will definitely get approved—rather, that under the current regulatory environment, institutions already believe these kinds of filings are worth trying

After the Trump administration took office, the SEC’s overall stance toward crypto clearly shifted. Grayscale believes there’s now a window of opportunity, so it has started submitting various ETF applications in a concentrated way, and WLD is only one step in that strategy

$WLD is up 2.91% today. That reaction is actually quite restrained. The market has already been “educated” by BTC and ETH regarding ETF approval expectations; it knows that the path from filing to approval is long, so it won’t blindly chase

What’s truly worth watching is the timing and rhythm of the SEC’s subsequent response, and whether Worldcoin can resolve its European regulatory disputes in the meantime. Any change in either of these will cause a clear reaction in WLD’s trajectory

I bought WLD at the top of the mountain at the peak—basically zeroed out. I even sold out my own eyes (metaphorically), chasing the hope of getting back to even one day, despite the very low safety net

#芯片股反弹 US stock short positions hit a record high

On the first day of earnings week, the semiconductor sector put two “ace cards” on the table at the same time—and these two cards point in completely opposite directions

The market is rising: Ambarella is up 6.24%, Teradyne up 3.54%, and Marvell up 3.32%. AI and chip stocks are leading the rebound—seems like sentiment is repairing

But at the same time, the percentage of S&P 500 constituent short positions relative to free-float shares has climbed to 3.79%, and Russell 3000 shorts are up to 6.3%—both have set new all-time highs

Over the past eight weeks, hedge funds have net sold US technology stocks in six of those weeks, trimming by roughly 10%, the largest pullback in more than a decade

Prices are rising, and “smart money” is on the move—this divergence is precisely the most worth pondering

The reason the shorts keep pressing is persistent concern about returns on AI investment. This worry isn’t new, but it’s being supported by an increasingly concrete number

Research from Nikkei shows that off-balance-sheet, implicit liabilities of five tech giants—Alphabet, Microsoft, Amazon, Meta, and Oracle—have grown eightfold over four years, with a total size of about $1.65 trillion, already exceeding their book liabilities

Meta alone is about $420 billion—close to three times its book liabilities. BlackRock plans to issue more than $12 billion in bonds to finance Meta’s data centers in Texas

This means the AI arms race is becoming increasingly dependent on debt financing. As long as AI revenue and returns keep running fast enough, this structure can hold. But once returns fall short of expectations, or interest-rate conditions tighten again, these off-balance-sheet debts could turn into systemic pressure points

So this week’s earnings are a real test—not whether revenue beats expectations, but whether capital expenditure guidance holds. If these companies keep ramping up AI spending while revenue growth starts to slow, the shorts’ logic will be validated. If AI monetization starts to accelerate, then today’s rebound would truly have a firm footing

Chip stocks are rising, but the fact that short positions are at an all-time high tells you something: a considerable amount of capital in the market believes this rebound won’t be able to sustain. Who’s right and who’s wrong will be answered by earnings

DYOR—not investment advice

The prediction market is growing wildly, but where the money comes from matters

During the World Cup, a single number made me look at it again and again: activity in prediction markets has already accounted for 27% of legally regulated sports betting in the US, up from just 9% at the beginning of the year.

Kalshi added 3 million new users. On Predict Fun, the World Cup champion market alone saw single-market trading volume of $320 million. This growth isn’t linear—it’s explosive.

Before last night’s World Cup final, on-chain tracking found that an entity created new accounts after withdrawing from a mixer, then heavily bought positions on Spain winning—31 BTC in total.

This exposes a fundamental contradiction in prediction markets: their transparency is selective. On-chain data is visible to everyone, but the source of funds can be obscured. Anonymous capital can quietly enter and then place massive bets on a particular outcome.

For years, traditional sports betting has struggled to properly control insider betting. Prediction markets have simply repackaged the problem using blockchain. The issue hasn’t disappeared—if anything, anonymity makes it even harder to trace.

The CFTC’s investigation into insider betting involving Kalshi’s commentators was opened earlier. The 31 BTC case from the World Cup is just another instance that brings the problem back into the spotlight.

Growth is real, and user demand is real. But the industry’s current challenge isn’t whether the market is big enough—it’s whether the regulatory framework can keep up with this pace.

Of the 3 million new users Kalshi gained, how many came seriously to participate in prediction markets, and how many were only drawn in by the World Cup traffic? How many will remain after the tournament ends won’t be clear until three months later.

Prediction markets are a genuinely valuable financial tool that turn collective intelligence into price signals—at least in theory, often more accurate than many analysts’ judgments. But “high value” and “low risk” are two different things.

The bigger this market gets, the stronger the incentives for gray-market capital. Until regulations are fully implemented, this contradiction won’t go away.

Before the next big event arrives, whether this 27% can hold is the real stress test.

DYOR Not investment advice #预测市场

#WTI原油涨2%至84美元
The escalation between Iran and the U.S. continues, with Brent breaking above $90

At around 2:00 a.m. on July 20, the U.S. military launched a ninth consecutive round of nighttime strikes against Iran. Trump’s stated reason is to avenge three fallen American service members.

Brent crude has broken above $90. Kuwait’s oil facilities were attacked, and merchant ships in the Strait of Hormuz have continued to be targeted.

I think the logic behind oil prices isn’t just concern about supply disruptions anymore; the market is pricing a longer-term uncertainty.

The fact that we’re already on the ninth consecutive nighttime strike alone tells the story—this conflict has no clear way out. Both sides are fighting, and neither shows any sign of coming to the negotiating table.

If the Strait of Hormuz were to truly face a large-scale blockade, 20% of the world’s crude oil supply chain would be cut off in an instant. $90 is only the starting point, not the endpoint.

But gold is more interesting. Geopolitical conflict would normally be expected to push gold higher, yet the price has pulled back to below $4,000. The reason is that the surge in oil prices lifts inflation expectations, and the market starts betting that the Federal Reserve may be forced to raise rates again.

Once rate-hike expectations rise, the U.S. dollar strengthens, and gold comes under pressure. Safe-haven logic and monetary-policy logic are fighting each other—gold is caught in the middle.

This combination is not friendly for the crypto market. Oil rising lifts inflation expectations; if the Fed truly turns hawkish, tighter liquidity expectations would suppress all risk assets.

In the short term, there’s a higher probability that BTC will move along with overall risk assets. An independent行情 is unlikely to emerge in this kind of macro environment.

The variables that really need watching are whether the Strait of Hormuz will escalate from attacking merchant ships to a substantive blockade, and whether Trump will next bring Iran into a sanctions framework like the one used against Russia. Either of these becoming reality would be an event of a completely different magnitude.

At this point, all that can be done is wait. The war has reached a ninth round and is not over—suggesting that neither side is willing or able to quickly back down.

$CL $XAU $BZ

DYOR Not investment advice

#韩股KOSPI启动Sidecar半导体领跌
Han stocks fall behind by more than 4%; memory stocks continue their downturn
$SKHY $MU $NVDA.US

Last Friday, when global semiconductors crashed, South Korea was on holiday. Today, trading opened and the country “paid the debt” all at once. The KOSPI opened down 4.16% at 6,536.65 points. Samsung Electronics and SK hynix both fell more than 5%, shattering market sentiment right from the start.

Since the June 22 peak, the KOSPI has cumulatively dropped nearly 25%. Even more striking is the forward P/E ratio at 5.78x. It’s not just a new low beyond the trough of the 2008 financial crisis—it’s the lowest since 2004. This isn’t a normal correction; it’s the market pricing in deeper, structural concerns.

What the market is doing now is repricing the timing of the entire memory cycle. People have been talking about this upcycle in memory since late 2023, but demand driven by AI has concentrated in high-end HBM products. The inventory clearance for traditional DRAM and NAND hasn’t turned out as clean as expected.

The “price-hike” narrative is real, but how long can prices keep rising—and in which specific product categories can gains actually be realized—are big questions the market is now putting a giant question mark over.

Institutions are also split in a very interesting way. Goldman Sachs maintains a KOSPI 12,000 target and calls for buying on dips. UBS, however, gives a target of 9,200 and urges a shift toward defense. Both sides use the same data yet arrive at completely opposite conclusions, which shows that uncertainty here is genuinely real—no one has full confidence.

At the same time, TSMC plans to add another $100 billion in U.S. investment, for a total scale that could reach $265 billion. On the industry side, capacity expansion is underway; on the secondary market side, shares are being sold off. This divergence in itself speaks volumes. What capital markets worry about isn’t that AI demand doesn’t exist—it’s that the speed of supply expansion is far faster than demand ramping, ultimately pushing prices down.

This week’s earnings season will be the validation window. Intel and Texas Instruments will report numbers first, and SK hynix and Samsung will follow. If guidance on capital expenditures from U.S. tech giants is cut again, memory stocks could get another hit. If results exceed expectations, that’s the basis for a rebound.

Before the data comes out, I neither dare to say this is an oversold rebound nor dare to say it’s okay to keep missing out. A forward P/E of 5.78x really looks cheap, but cheapness is never a sufficient condition for stocks to rise—cheap can always get cheaper.

Wait for the earnings, look at the data, and then make a judgment.
DYOR Not investment advice

#2026足球风潮
The 2026 World Cup comes to an end.
Spain 🇪🇸 1–0 Argentina 🇦🇷 — collectivism wins

In the 106th minute, Ferran Torres unleashes a volley into the net. It isn’t just the end of a final—it announces the handover of an era

0–0 in regular time, and both sides know this isn’t a match decided by attack. Until Enzo receives two yellow cards and is sent off, forcing Argentina to play with ten men—then the balance of fate tips. Argentina had zero shots on target in regular time. That number tells the story better than the score

Spain represents the most mature form of modern football. No single superstar, yet no real weakness either. Rodri is the heart, Yamal is the lightning, Cubarsí is the iron wall, and Unai Simón made it through the tournament with almost no mistakes. This isn’t a revival of tiki-taka—it’s an evolutionary upgrade. Possession is the means, not the end. Wait patiently for the opponent to make a mistake, then strike with a single cut

Messi scores eight to reach the final. After the match, his embrace with 19-year-old Yamal is the most moving scene of the tournament. One dynasty closes, another begins. His era was already mythic—but in the end, it’s over. Messi’s time depended on this pattern

Mbappé scores 10 to win the Golden Boot again, and the record books keep getting rewritten. Yamal becomes the first U20 player to lift both the European Championship and the World Cup. Expanding to 48 teams brings both surprises and dead weight. High-pressing combined with data-driven execution and youth development output—that’s the future main theme

Co-hosted by the US, Canada, and Mexico, at MetLife Stadium. 48 teams, 104 matches— the biggest World Cup in history. What happens on the pitch is always the purest: a group of people playing one ball, turning a whole month of the most condensed human collective emotions into reality

Congratulations to Spain 🇪🇸. Salute to Messi 🫡
See you in 2030 ⚽️

#worldcup #世界杯

#CLARITY法案仍待参议院推进
The vote is coming—can it pass this time?
The bill has been sitting in the House for a full year. This week is the last chance.

🪁
CLARITY is one of the most important U.S. digital-asset regulatory framework bills to date. After the House passed it last year, the Senate stalled it for a full year. Majority leader Thune has set this week as the voting window. Novogratz says there isn’t much disagreement on the technical side; the last step is getting past the wording of the ethical provisions—which, in the end, may still not clear.

🪁 The obstacle isn’t the technology—it’s politics
Warren’s letter forces Trump to come clean first.

Warren, a senior Democratic member of the Senate Banking Committee, sent a letter to Trump demanding that he voluntarily disclose his cryptocurrency gains and holdings from January to July by the 23rd. Since the president has financial interests in this industry, Congress can’t debate the bill fairly under partial disclosures.

Getting Trump to take a stance first can both add conditions to the bill negotiations and seize the initiative in the public narrative. If Trump refuses or drags his feet, Democrats will have reason to keep blocking it. If he cooperates, it’s also another opportunity to pressure Republicans in the court of public opinion.

On the Republican side, Hagerty’s response is equally blunt: the biggest obstacle is partisan political bargaining—not the bill itself.

🪁 The industry’s position
Novogratz has unusually called for bipartisan compromise. The crypto industry can’t wait any longer. If the bill isn’t passed before the recess, it’s likely to be a long way off. Regulatory uncertainty will continue to hang over the entire market.

Politicians want leverage; the industry wants certainty—that’s the basic structure of this game.

🪁 Impact on the market
If passed, the CLARITY bill would be a systemic positive for the crypto market: it clearly defines which assets are commodities and which are securities, opens compliance pathways for institutions, and will accelerate the entry of a large pool of sidelined capital.

If the bill doesn’t make it onto the full-chamber vote this week and is delayed again, near-term sentiment may be suppressed—but it won’t change the medium-term direction. The market has been pricing this uncertainty for a long time.

I think the odds of passage this week are neither very high nor very low. The real variable is Warren’s deadline on the 23rd, and how Trump responds. If the White House is willing to make a show of good faith on disclosure issues, some Democratic votes could loosen.

The worst-case scenario is that there’s no vote before the recess, the bill enters a new round of bargaining, and the market keeps waiting.

DYOR

Iran stops honoring the US-Iran memorandum; conflict may escalate further
The ink on the negotiations table is barely dry, yet bombs are already on the way
The US-Iran reconciliation memorandum signed last month has now been rendered void on both sides. On July 18, Iran’s Deputy Foreign Minister said the United States violated its commitments first, and Iran therefore stopped honoring the memorandum

The Supreme Leader’s wording was even more direct: “Trump’s signature is worthless.” Trump’s response was that he doesn’t care at all

There are no signs of cooling on the military front. The US Central Command announced that beginning this morning it will launch a new round of airstrikes on Iran. The targets are meant to weaken Iran’s ability to threaten commercial shipping through the Strait of Hormuz—also in response to Iran’s Islamic Revolutionary Guard Corps’ attack on US forces in Jordan the day before

The Strait of Hormuz is the biggest wildcard
It’s the choke point for global crude oil shipments, with about 20% of the world’s oil supply passing through this waterway each day. One of the official reasons for the US airstrikes is to protect commercial shipping here, but the strikes themselves end up increasing the risk premium for this route

In history, Iran has repeatedly threatened to blockade the Strait of Hormuz. Each time it issues such warnings, oil prices swing sharply in the short term. This escalation has reached the point where the memorandum is voided and both sides are calling each other out. The probability that the blockade threat moves from words to action is rising—but it’s still a different magnitude from an actual blockade

Crude oil is the directly affected asset, and the geopolitical risk premium will be repriced in the short term. US stock energy shares may benefit, but an overall contraction in risk appetite will weigh on technology and growth stocks

For the crypto market, this is a double-edged sword. When risk-off sentiment heats up, it typically flows first into gold and US Treasuries. But if the situation further spirals and pushes up inflation expectations, the BTC anti-inflation narrative will likely surface as well. In the short term, the more likely direction is to follow the broader decline in risk assets rather than stand out and strengthen independently

I think the breaking of the memorandum itself isn’t the end—it’s the reshuffling of negotiation leverage. When Trump says he doesn’t care, that’s negotiation language, not a real abandonment of diplomacy. But in this stage, both sides need to play even tougher cards first before they can return to the table

The risk is misjudgment. With the airstrikes escalating, Iran taking actions in the Strait of Hormuz, and the United States further upping the ante—each step carries the possibility of pushing the situation toward a point of no return. Markets hate uncertainty, and the developments over this weekend are worth keeping a close watch

#伊朗原油突破80美元

After two years of big-model frenzy, AI giants collectively turn around to fix the data foundation for their systems

The engine is installed, but the road hasn’t been repaired yet
First half: Everyone races in the exhibition hall

The AI narrative over the past two years has been remarkably uniform—bigger models, more parameters, higher benchmarks, with open-source and closed-source camps competing for the spotlight. Company playbooks have also become highly similar: buy compute, hire algorithm talent, deploy models, and assume that if the model is strong enough, deployment will naturally follow

Reality dealt a blow. The story of a large central state-owned enterprise CIO is a typical case study: 17 business systems, 9 data warehouses, and 3 clouds. Data formats are a patchwork, and lots of paper reports have never been digitized. In the end, even the most basic application—equipment fault diagnosis—couldn’t be run, because the model couldn’t even understand historical maintenance records

The bottleneck isn’t that the model isn’t capable—it’s that the data can’t be fed in
Second half: The car needs to be driven into real streets

Factories must keep running, hospitals must remain safe, and governments must stay compliant. No matter how strong the model is, if the data highway hasn’t been repaired, it can only keep circling in place

Actions across the global industrial supply chain are starting to converge. The same large model, when run “naked” on raw enterprise data and without a complete data engineering system, shows a cliff-like accuracy gap. The difference isn’t in the model’s “brain”—it’s whether it can actually “eat” the right ingredients

That’s how a new architecture emerges: on top are the model and capabilities; below are the data engineering, permission auditing, and governance strategies. Models and data are no longer just upstream and downstream in a production line—they become partners that feed each other

China adds another layer of difficulty
Manufacturing supply chains are longer, compliance requirements are stricter, there is more unstructured data, systems are more fragmented, and data definitions vary more widely. Bridging the gap from general intelligence to industry-specific intelligence is hard—but the underlying support is precisely the end-to-end data infrastructure

This isn’t about giving AI a bigger brain; it’s about first fixing the nervous system

What’s truly scarce
This shift to fix the data foundation is, at its core, a correction of cognition. The scalable value of AI doesn’t depend on how many points a particular model scores today. It depends on whether data can be continuously supplied at high quality, whether the system can be continuously governed with trustworthiness, and whether the engineering can continuously close the loop with real deployments

China doesn’t lack the 101st open-source large model. What’s missing is cleaning the “data mess” first, and simmering out high-quality “data soup”

Once this works end-to-end, AI will move from hot news to a tool on the workbench

DYOR Not investment advice #AI

The US-China AI large model IPO wave: valuation bubble or a historic pricing moment?

A batch of AI companies that have never turned a profit is causing global capital markets to price their future

Two trillion-level US market stories

Anthropic and OpenAI filed their S-1s almost simultaneously, with a combined valuation of about $1.8 trillion. OpenAI’s $852 billion is built on monetized GPT revenue—still far from profitability, but with a real revenue foundation. Anthropic’s $965 billion bets more on the “safe AI” narrative, with commercial maturity about one level lower. With the two companies going public around the same time, the funds will directly compete, putting pressure on both offer prices

China’s three different paths

DeepSeek chose the A-share market, with a $71 billion valuation—unprecedented in the A-share context. The choice itself is a strategic move to avoid USD fundraising and geopolitical risks. “The Dark Side of the Moon” is in talks for a Hong Kong listing, with a valuation of 20 billion—while Hong Kong can likely digest that without much issue, the liquidity discount is real

The core contradiction

Pricing is for the market landscape five years from now, but no one today knows who will still be alive then. The price war for large-model APIs hasn’t ended yet. Inference costs have fallen by more than 100 times in two years. Whether a leading position can be converted into a moat is the question these firms must answer after going public

Among this group, the one most worth long-term tracking is DeepSeek—the only Chinese team that has ever posed a genuine technical challenge to OpenAI

DYOR, not investment advice

#Meta拟向Anthropic出租算力最多100亿美元

World Cup Final Countdown: Argentina vs Spain

In the 2026 World Cup final, the defending champions take on Europe’s powerhouse—who will lift the trophy?

Argentina 🇦🇷

When Messi is on the pitch, it brings not just skill, but a psychological structure for the entire team at crucial moments. This has already been proven in Qatar and the 2021 Copa América. Lautaro Martínez’s efficiency inside the box is outstanding, while De Paul and Mac Allister control the tempo in midfield—this is a well-structured style. Their defense is relatively weak, and any gaps created by high pressing could be dangerous.

Spain 🇪🇸

The attacking thrust from both flanks—Yamal and Nico Williams—was already at the very top level in the last European Championship. Pedri’s ball control is key, and Morata’s handling of drop-offs in the penalty area is mature. Spain’s position-based football emphasizes high possession, and their depth off the bench is stronger than Argentina’s, making their advantage more obvious in a long possession battle. The weakness is that, historically, their style has had trouble against low-block defenses combined with counter-attacking systems.

The Key Contradiction

This match is a clash of two philosophies. Spain wants to control the ball and build attacks, while Argentina will try to compress space and create opportunities. Once Argentina produces a high-quality counterattack and scores first, the rhythm of the game will change completely. If Spain can dominate in the first half, their stamina and substitution depth will matter more in the second.

The chances of extra time—and even penalties—are quite high. In recent major tournaments, Argentina has been consistent in penalty shootouts during knockout rounds, and the goalkeeper’s performance could provide an extra boost.

Leaning toward Argentina to take it in the end, but it won’t be an easy 90 minutes.

DYOR Tournament analysis is for entertainment purposes only
#世界杯2026

#hyperliquid跌10.28%
$HYPE Today’s decline is clearly greater than the contemporaneous pullback in BTC and ETH
The leading player in the DEX derivatives sector—why does it fall harder than others when the overall market pulls back?

Hyperliquid’s fundamentals haven’t really changed. It is currently the decentralized perpetual futures trading exchange with the largest trading volume. On-chain data has continued to grow, and user stickiness is stronger than most DEXs. But the price of the HYPE token includes a large amount of expectations for future growth and protocol fees. This type of valuation structure is especially fragile when market sentiment weakens.

There’s also a structural factor: a significant portion of HYPE’s circulating supply comes from airdrops. Addresses with extremely low cost basis have little pressure to reduce positions when the price falls—they can sell at any time. This provides ample “ammunition” for the downturn.

Today’s drop also has broader context: overall risk appetite in the market is contracting. The semiconductor sector has entered a technical bear market, renewed hawkish signals from the Fed have reappeared, and macro pressures are feeding into the crypto market. In this kind of environment, the market tends to shed assets with higher risk premium first. HYPE and smaller tokens in the SOL ecosystem are on the shortlist for the first wave of punishment.

But one point should be made clearly: a single-day drop of 10% is not new for HYPE. Historically, it has seen multiple single-day swings of 20% or more, and in the end they did not affect the medium-term trend.

The real question is whether this drop is supported by any fundamental changes—such as security incidents at the protocol level, abnormal liquidity, or unusual activity by large holders—or whether it’s simply the market-wide transmission of sentiment. Based on the information available so far, it looks more like the latter.

A sentiment-driven selloff usually recovers faster than a fundamentals-driven one, but the prerequisite is that the broader market stabilizes.

DYOR Not investment advice

#spacex星舰测试取消股价跌4% #spacex空头持仓升至流通盘29%
Just a few weeks after a company’s IPO, short positions jumped from 5% to 29%—what are they really betting on? $SPCXB

Yesterday, SpaceX’s stock price fell to near its $135 IPO offering price, with short sellers accelerating their entry. Data from S3 Partners shows that currently about 185 million shares are being shorted, accounting for 29% of publicly traded shares, with short interest of roughly $25 billion. Three weeks ago, the figure was only 40 million shares—about 5% to 7% of the float. Since July, the stock has fallen by roughly 20%; on Wednesday it even dipped briefly below the IPO offer price

At the same time, the 13th Starship test flight was canceled, and the stock fell another 4%. The market appears to be expressing a clear view that the IPO was priced too high

When SpaceX went public, the narrative was about an interplanetary internet plus rocket commercialization—so the valuation reflected a large amount of expectations for future cash flows. The problem is that at the time of the IPO, only about 5% of total shares were publicly available for trading, meaning the float is extremely small, and the price can be easily pushed up by sentiment

Now, the timetable for the end of lockups is approaching. Around 11% of shares will become eligible for sale when the Q2 earnings are released, and afterward there are additional batches of roughly 4% released at intervals—this is the release pressure that short sellers are betting on

Elon Musk holds about 42% of the shares, locked until June 2027—this is a stabilizer, but other early investors and employees’ chips won’t wait indefinitely

The cancellation of Starship tests is another variable. The commercial value of Starship isn’t just a technology demonstration—it’s the core narrative support for SpaceX’s entire reusable rocket business model. Every delay discounts that narrative

I believe a 29% short ratio is an extreme figure for any stock, indicating that institutional disagreement about this valuation is already very significant. This isn’t retail investors making a bet—professional capital is systematically shorting a newly listed company. In this situation, the stock’s direction depends largely on the results of the next few Starship tests and, when the unlock window arrives, what the major shareholders actually do

The fundamentals haven’t changed; what’s changed is how much the market is willing to pay for this story

DYOR Not investment advice

Semiconductor selloff escalates, entering a technical bear market
TSMC turns in its strongest-ever quarterly report—so why did it become the fuse for a breakdown?

Last night, the U.S. stock market closed. The Philadelphia Semiconductor Index fell 4.3% in a single day, down 22% from the mid-June peak, establishing a technical bear market. SK Hynix’s ADR dropped more than 13%, SanDisk fell over 12%, and Micron’s pullback from its all-time high is now over 30%. The selloff didn’t stop in the U.S. market either. Today, Japan’s Kioxia (KeiXia) hit the daily limit-down intraday, down 15.55%; its market cap was cut in half versus the June peak. The Nikkei 225 briefly fell more than 4%.

What triggered this round of selling was TSMC’s earnings report: a 67.7% gross margin, and an upgraded full-year capital expenditure guidance to $60–64 billion. Each figure on its own looks strong—but the market’s reaction was to sell.

TSMC is aggressively expanding capacity. Where is all the money going? Into AI chip production capacity. The question is: can demand for AI chips actually absorb this round of疯狂（crazy） capital spending? Or is this a self-reinforcing overheating driven by the supply side, ultimately ending with demand falling short of expectations? The stronger TSMC is, the more acute this question becomes—that’s the real reason the market is selling.

Market-flow data further confirms this. Retail investors net sold $125 million of SanDisk last week. The total trading value in individual stocks by retail investors rose to a record high of $370 billion. This isn’t normal profit-taking—it’s capital systematically exiting tech stocks. Morgan Stanley expects the growth rate of DRAM contract prices to peak in Q4 2026, meaning the valuation re-pricing still isn’t finished. This selloff may not be over yet.

I think this selloff isn’t short-term sentiment fluctuation. It’s the market resetting the pricing of the AI investment return cycle. Over the past two and a half years, the semiconductor sector’s rally was priced on the assumption of unlimited AI demand. Now people are starting to ask: where is the ceiling? Once that question is taken seriously, the valuation logic changes.

A technical bear market is the result, not the beginning.

DYOR, not investment advice

$SKHYB
$MU
$NVDAB

#亚洲股市连续第二日下跌

Logan calls for a rate hike, but the market is still pricing July unchanged
Who’s really lying—Fed officials or the market?

Today, Logan publicly stated that the Federal Reserve should raise rates to combat high inflation, and also hinted that she might vote against a hike at the late-July FOMC meeting. This is the first Fed official since Wash took office to openly call for a rate hike. Her remarks lay groundwork for uncertainty ahead of the July FOMC—not just idle talk.

But the market didn’t move at all. CME data shows an 88.8% probability that rates will remain unchanged in July. The market simply treated her comments as background noise.

This divergence is definitely worth watching. Market pricing is based on data, not officials’ words. June CPI came in at 3.5%, below expectations, and core inflation is also easing. Judging by the numbers, inflation really does seem to be cooling.

Logan said that the June data isn’t enough for her to be confident that inflation is on a path back to 2%. But the 2% goal was never something that can be achieved within a single quarter. The market’s logic is that as long as the direction is right, you don’t need to wait until it’s fully back on target to act.

Vice Chair Jefferson’s remarks are even more interesting. He leaves openings on both sides: saying, “If inflation can’t slow down quickly, we may need to reconsider,” while also saying, “The current policy stance is in a good place.” This wording is classic in central-bank communications—an accommodating tone that still hides hawkish intent. It both avoids letting the market reprice aggressively and preserves the option of further hikes.

I don’t think Logan’s comments this time are mere performance. Even a single dissenting vote doesn’t mean rate hikes are a done deal. The FOMC operates by majority vote, and hawks within the committee are still in the minority right now. The real stress test is what data comes out before the late-July meeting. If retail sales or core PCE again come in above expectations, market pricing would quickly tilt toward the hawkish camp—that would be the true turning point.

At this stage, officials are probing the market with words, while the market is pushing back on officials through pricing. There won’t be an answer to this game of give-and-take before the end of July.

For the crypto market, the 88.8% “no rate hike” probability is the underlying logic supporting risk-asset sentiment right now. If this number starts to fall, the $BTC $63K support won’t be as solid.

DYOR (not investment advice)