Aston在下不求

Is SpaceX pricing at $135 per share? With a price difference of up to 40% compared to exchanges like Binance, is it an arbitrage or a chase opportunity?

According to the latest report from Reuters, SpaceX, led by Elon Musk, plans to raise $75 billion through an Initial Public Offering (IPO). The news indicates that SpaceX is expected to issue 555.6 million new shares, priced at $135 each, leading to a total company valuation of up to $1.75 trillion. SpaceX is set to kick off its Roadshow this week, with the issuance price potentially being finalized by June 11, followed by its listing on Nasdaq under the ticker symbol "SPCX." Major crypto exchanges have already launched related products, with current quotes around $190, showcasing a price difference of up to 40%. Is this an arbitrage or a chase opportunity? (SpaceX IPO price revealed! Priced at $135 per share, set to join the Nasdaq 100 index upon listing) SpaceX's fundraising scale and underwriter greenshoe option This time, SpaceX plans to issue approximately 555.6 million new shares, raising a total of $75 billion for company operations and expansion. Reports indicate that the company has set a maximum of 15% greenshoe option, allowing for additional share issuance when market demand is high. This fundraising is led by major investment banks like Goldman Sachs, and subsequent funding subscriptions and institutional pricing dynamics are worth continuous tracking. A share price of $135, compared to a 40% price difference on major exchanges? The news indicates that SpaceX is expected to issue 555.6 million new shares, priced at $135 each. Major crypto exchanges including Binance, Bitget, and Defi Hyperliquid have already launched related products, with current quotes around $190, resulting in a price difference of up to 40%. The IPO market is heating up; is this an arbitrage or a chase opportunity? Is a 40% price difference an arbitrage opportunity? 40% seems like a lot, but let's not forget that the IPO market always triggers human FOMO emotions. Last May, when stablecoin issuer Circle went public, it surged by 168% on the first day and skyrocketed to a historical high of $263 within two months, compared to its IPO price of $31, representing a 7.5 times increase. (Circle's glorious listing, surging 168% on the first day, will it lead to a wave of crypto IPOs?) If the estimated issuance shares and valuation pricing from the aforementioned exchanges are correct, there might still be room for a price surge? This article is not investment advice; please DYOR. This article "Is SpaceX pricing at $135 per share? With a price difference of up to 40% compared to exchanges like Binance, is it an arbitrage or a chase opportunity?" first appeared on.

Korea launches leveraged ETFs to boost trading, with Taiwan and Korea's market cap only 100 billion USD apart, experts urge to quickly buy more Taiwanese stocks

Thanks to the booming global artificial intelligence (AI) sector, the stock markets of Taiwan and South Korea, major hubs for semiconductors in Asia, have shown strong momentum this year. The combined market cap of Taiwan and Korea has officially surpassed that of the Indian stock market, ranking fifth and sixth globally. This shift highlights the concentration of global capital in markets with AI infrastructure and hardware supply chain advantages. Notably, South Korea's regulatory authorities officially approved single-stock leveraged ETFs at the end of May, further boosting the performance of large-cap stocks like Samsung Electronics and SK Hynix. Financial expert You Ting-Hao is calling for everyone to quickly buy more Taiwanese stocks to keep distance from the Koreans. The AI wave boosts Taiwan and Korea's stock markets, with their market cap exceeding India's. The demand for AI infrastructure continues to expand, driving significant growth in the stock markets of Taiwan and South Korea, which have advanced semiconductor manufacturing capabilities. Capital market data shows that the total market cap of both Taiwan and Korea has shown strong growth this year, successfully surpassing the increasingly watched Indian stock market, ranking fifth and sixth globally. South Korea's launch of leveraged ETFs in May injects momentum into the semiconductor giants. The South Korean regulators, previously concerned about protecting retail investors and avoiding high market risks, had long prohibited the listing of single-stock leveraged products. However, in recent years, day traders in South Korea have shown intense demand for investments in AI semiconductor stocks, with substantial funds flowing into leveraged semiconductor funds listed in the U.S. or the 2x leveraged ETFs for Samsung and SK Hynix listed in Hong Kong. Data indicates that funds flowing into these Hong Kong-related products have each reached approximately 1.3 billion USD this year, even surpassing similar products tracking Tesla or Microsoft. In order to prevent excessive outflow of domestic funds, the South Korean Financial Supervisory Service decided to adjust its policy and allow the issuance of local products. After the product was listed, it quickly attracted a large amount of capital seeking AI themes, significantly boosting the stock price performance of the two semiconductor giants, Samsung Electronics and SK Hynix. The influx of leveraged funds not only injected additional liquidity into the market but also became one of the key factors driving rapid growth in Korea's total market cap, narrowing the gap with and surpassing India. Market concentration rises, and potential volatility risks require continued attention. Although AI themes have brought significant market cap growth to Taiwan and Korea, the phenomenon of market capital being overly concentrated in a few semiconductor stocks has also raised concerns. Especially after the introduction of leveraged products in Korea, rebalancing trades conducted to maintain leverage ratios may further amplify market volatility during index fluctuations. Currently, the combined weight of the two semiconductor giants, Samsung Electronics and SK Hynix, in the Kospi index is close to 50%. Meanwhile, Taiwan's TSMC (2330), a national treasure, has a weight exceeding 40% in Taiwan's weighted stock price index, meaning that the overall market trends will be more easily influenced by individual company earnings reports. Investment institutions need to closely monitor related structural risks when assessing the future development of Taiwan and Korea's stock markets. The market cap gap between Taiwan and Korea is only about 100 billion USD, and You Ting-Hao urges everyone to quickly buy two more shares. The Korean Kopsi index has risen 104% so far this year, far exceeding Taiwan's weighted index of 58%. Currently, the market cap gap between Taiwan and Korea is only about 100 billion USD, and financial expert You Ting-Hao loudly calls for 'Taiwan's stocks are in crisis,' urging everyone to quickly buy two more shares to distance themselves from the Koreans. However, some netizens are calling for 'Eastern mysterious power' to come and buy Korean stocks, seemingly turning stock buying into a patriotic act. This article Korea launches leveraged ETFs to boost trading, with Taiwan and Korea's market cap only 100 billion USD apart, experts urge to quickly buy more Taiwanese stocks first appeared on.

Strive's May Boost of 2,500 BTC, Saylor Selling 32 at the Same Time

U.S. asset management company Strive, Inc. (Ticker: ASST) announced in an SEC 8-K filing on 6/2 that it scooped up 2,500 BTC between May 23 and June 1, with an average buy-in price of about $74,092 per coin (including fees and expenses), totaling an investment of approximately $185 million. This 9-day accumulation period almost completely overlaps with the timeframe when Strategy (formerly MicroStrategy) sold 32 BTC, highlighting a clear divergence in the strategies of these two Bitcoin treasury firms. After this accumulation, Strive's total Bitcoin holdings rose to around 17,509 BTC. In comparison, Strategy holds 843,706 BTC, which means Strive's size is only about 2%, but its aggressive accumulation pace compared to Strategy's concurrent selling provides a concrete example of a 'trend reversal' in the market. The 9-day accumulation period completely covers the timeframe of Saylor's sell-off, and it's worth noting the timeline. Strategy disclosed in its SEC 8-K on 6/1 that it sold 32 BTC for an average price of $77,135 between May 26 and May 31, totaling about $2.5 million. Strive's accumulation period (5/23-6/1) fully encompasses this window and extends beyond Saylor's selling conclusion. In other words, while Saylor was selling 32 BTC in those 6 days, Strive was at least adding over a thousand BTC during the same time. A more detailed aspect to consider is the average price comparison. Strive's average purchase price of $74,092 is significantly lower than Saylor's selling average of $77,135, a difference of about $3,000. This indicates that Strive absorbed the chips Saylor released at a higher price range, and although the absolute scale difference is vast, it signals meaning for value-oriented investors. The strategic difference between Strive and Strategy: pure buyers vs. structured treasury. The core difference between the two companies lies in their capital structure and operational logic. Strategy has issued multiple series of preferred stocks (STRC, STRK, etc.), requiring stable dividend payments; the sale of 32 BTC on 6/1 was for paying preferred stock dividends (see market divergence analysis). Strive is currently still in a pure buyer stance, maintaining a smaller preferred stock structure, not needing to leverage BTC holdings for dividend financing. Strive has simultaneously issued 'Variable Rate Series A Perpetual Stretch Preferred Stock' (Ticker: SATA), structurally comparable to Strategy's preferred stock tools. One of the key points for market observation is whether Strive, as its preferred stock expands, will follow the same 'selling BTC to cover dividends' path as Strategy or maintain its pure buyer stance. This article, Strive's May Boost of 2,500 BTC, Saylor Selling 32 at the Same Time, first appeared in <a>...</a>.

Alphabet raises $80 billion to seize the AI market, Berkshire backs with $10 billion

Google's parent company, Alphabet, announced on Monday that it will raise $80 billion through a series of equity financing schemes to support its massive capital expenditures in AI infrastructure. This colossal financing plan includes a $40 billion At-The-Market (ATM) offering, $30 billion in underwritten stock and mandatory convertible preferred stock, along with a $10 billion investment agreement with Berkshire Hathaway Inc. Given the sheer size of the financing, the market expects this move to crowd out funds in this year's bustling IPO market. Alphabet's strategy to expand self-developed chips and AI infrastructure is at the core of this epic financing move, aimed at accelerating the development of cutting-edge AI models' essential infrastructure. The company is actively ramping up production of its in-house AI chips—Tensor Processing Units (TPUs)—to meet the strong demand for high-performance computing. In the current chip market dominated by Nvidia, Alphabet's TPUs have become a major alternative, and this financing will help it stay competitive in the tech race. Expectations for soaring capital expenditures and cash flow pressure According to CFO Anat Ashkenazi's earlier forecasts, Alphabet's capital expenditures (CapEx) in 2027 are expected to significantly exceed the budgeted $190 billion for 2026. Market analysts even estimate that next year's total capital expenditures could reach $300 billion, surpassing Alphabet's own operating cash flow. Therefore, external equity financing has become a key means to maintain its investment flexibility and financial robustness. Berkshire invests $10 billion, new CEO Abel's investment strategy The $10 billion investment from Berkshire Hathaway has become a focal point of this financing. Since Warren Buffett's retirement last year and Greg Abel taking the helm at Berkshire, the company has been actively utilizing its cash reserves of up to $397 billion. Berkshire has been steadily building its stake in Alphabet since last year, and as of the end of March this year, it holds Class A and Class C shares of Google, valued at $16.6 billion. This additional investment reflects the new management team's confidence in Alphabet's long-term growth prospects in the AI sector. (Abel's first acquisition post-transition, Berkshire expands real estate footprint with a 24% premium acquisition of TMHC) The crowding-out effect in global capital markets This historic equity transaction is expected to have a profound crowding-out effect on upcoming tech unicorn IPOs. This year, high-growth companies including SpaceX, Anthropic, and OpenAI are planning to go public, but the size of market funds is limited. When large institutional investors allocate significant funds to a mature operator like Alphabet, which has TPU growth prospects, it may directly compress the fundraising space for other startups in the IPO market. This article, Alphabet raises $80 billion to seize the AI market, Berkshire backs with $10 billion, first appeared on <a>...</a>.

Anthropic Confidentially Files S-1, $965 Billion Valuation Opens IPO Doors

AI company Anthropic announced on June 1 that it has confidentially submitted its S-1 registration statement for an initial public offering (IPO) to the U.S. Securities and Exchange Commission (SEC). Anthropic further detailed on its official X account: "Once the SEC completes its review, the company will retain the option to proceed with the IPO, with the final decision to go public depending on market conditions and other factors." This application, submitted under Rule 135 of the Securities Act of 1933, falls under the confidential filing procedure, meaning that specific details about the number of shares, pricing, or underwriters are not publicly disclosed at this stage, and the document does not constitute a direct offer to sell securities. This application only initiates the SEC review process and does not mean that Anthropic has definitively decided to go public. Context of the $965 billion valuation: On May 29, they announced the completion of a Series H fundraising round, raising $65 billion, bringing the post-money valuation to $965 billion, surpassing OpenAI as the highest-valued private AI company. The Series H was led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, with strategic investors including memory giants Samsung Electronics, SK Hynix, and Micron Technology. If Anthropic proceeds with the IPO at nearly a $1 trillion valuation, it would be one of the largest tech IPOs in history. For comparison: Alibaba raised $25 billion in its 2014 IPO, Saudi Aramco raised $29.4 billion in 2019, and Facebook raised $16 billion in 2012. As of June 1, if Anthropic releases a traditional IPO proportion of equity (5% to 10%), the fundraising size could exceed $48 billion to $97 billion, with a high likelihood of setting a record. Joining the 'trillion-dollar IPO' queue with SpaceX and OpenAI, Anthropic is one of the tech giants expected to enter the public market intensively by 2026. On May 29, SpaceX also lowered its IPO valuation to $1.8 trillion, aiming to go public as early as mid-June; OpenAI has been reported by multiple media outlets to have initiated pre-IPO preparations, with a valuation range between $500 billion and $1 trillion. Anthropic's confidential S-1 filing significantly compresses the uncertainty of the listing timeline, which is expected to further accelerate the IPO pace for similar AI companies. For the secondary market, the Anthropic IPO will present a rare direct holding opportunity for AI investors. Over the past two years, exposure to AI themes in the secondary market has mainly been through chip manufacturers like NVIDIA, TSMC, and Broadcom, as well as cloud giants like Microsoft, Alphabet, and Amazon. A direct listing by Anthropic would officially bring the asset class of 'pure AI companies' into the public market. Follow-up Observations: SEC Review Timeline, Listing Date, Valuation Gap between Series H and IPO Pricing Market observation focuses on three key points: First, the SEC's review timeline for the S-1 typically takes about 60 to 90 days from confidential filing to first public filing, followed by another 30 to 60 days for actual pricing, meaning Anthropic could potentially go public by the end of 2026; Second, market conditions, AI valuation sentiment, and the state of the U.S. stock market at the time of IPO pricing will directly impact the final valuation; Third, whether the IPO pricing will exceed the Series H valuation of $965 billion, and whether Series H investors can achieve positive returns, will be critical signals for subsequent capital flows in the AI private market. This article on Anthropic's confidential S-1 filing and $965 billion valuation opening IPO doors first appeared on <a>...</a>.

ECB Executive Schnabel: Stablecoins are a replica of MMFs, warning of Euro sovereignty risks

Isabel Schnabel, a member of the European Central Bank (ECB) Executive Board, spoke at the "Central Banks and the Future of Money" international conference hosted by the Bank of Korea (BoK) on June 1. Her speech was titled "From money market funds to stablecoins: lessons for central banks." She drew historical parallels between the current financial stability risks of stablecoins and the money market funds (MMFs) of the 1970s, warning that 99% of stablecoins are dollar-denominated, threatening the currency sovereignty of the Euro. Schnabel is one of the earliest officials within the ECB to publicly advocate for the push towards a digital Euro, consistently leaning towards "proactive legislation paired with public sector infrastructure." Her arguments presented at the Bank of Korea hold more weight than typical working papers or technical discussions at the staff level. Stablecoins = 1970s MMF replica: liquidity mismatch and redemption risks Schnabel compared stablecoins to the MMFs of the 1970s, pointing out their structural vulnerabilities: liabilities can be redeemed at any time, and the assets are held in less liquid instruments. Once the market loses confidence in the quality of these assets, it can trigger a run. She named two specific cases. Tether (USDT) holds reserves in assets that are "relatively illiquid and high-risk," including commodities, loans, and crypto assets; USDC, on the other hand, faced a brief decoupling during the March 2023 collapse of Silicon Valley Bank (SVB), as part of Circle's reserves were held at SVB, representing a concrete example of stablecoins carrying "bank contagion." Schnabel stated, "Due to liquidity mismatches and potential loss of trust in asset quality, stablecoins face redemption risks." 99% dollar-denominated: Stablecoins are becoming a channel for 'technological dollarization' A more geopolitically significant argument is that of currency sovereignty. Schnabel cited data showing that the global market cap of stablecoins is around $300 billion, with USDT and USDC accounting for about 90%, and the overall dollar-denominated proportion reaching 99%; in contrast, Euro-denominated stablecoins only amount to about €500 million, while 85% of crypto trading volume is priced in stablecoins (mostly dollar stablecoins). She warned, "Dollar-based stablecoins could create a new cross-border network, and dollarization will become a byproduct of technological adoption, rather than a deliberate currency choice by countries." This mechanism is particularly impactful for emerging markets and small economies, as "mass issuance of dollar stablecoins could amplify the transmission of U.S. monetary policy internationally." For the Eurozone, this will erode the long-term standing of the Euro as an international currency. Digital Euro + Pontes/Appia: Europe's countermeasure combo Schnabel reaffirmed the dual-track strategy of the Euro system: the retail digital Euro (CBDC) will retain the "option for people to use public money," while wholesale tokenized central bank money will provide a secure anchor for on-chain settlements between financial institutions. She noted that the digital Euro is also a specific tool for "European strategic autonomy," reducing reliance on non-European payment service providers. On the infrastructure side, she named two projects the ECB is advancing: Pontes (DLT settlement bridge connecting central bank settlement systems and distributed ledger platforms) and Appia (a broader tokenized financial ecosystem infrastructure). She also mentioned plans to expand TIPS (the instant payment system) and EUREP (liquidity swap mechanism) to reduce the Eurozone's dependence on external payment infrastructure. On the regulatory front, Schnabel advocated for strengthening reserve quality, liquidity requirements, transparency of composition, and redemption protection mechanisms, aligning with MiCAR's regulatory path for major stablecoin issuers with a 60% minimum reserve requirement. She also issued a principled call: the future tokenized financial system should "remain open and multi-currency," avoiding becoming a single base monopolized by the dollar. Follow-up observations: Timelines for Pontes/Appia's advancement and MiCAR implementation There are three key market follow-up observations: First, the progress of the European Parliament's deliberation on the digital Euro legislation and when it can enter the actual issuance phase; second, the timeline for Pontes and Appia to transition from conceptual design to usable infrastructure, which will directly affect the supply capacity of Euro-denominated tokenized assets; third, the enforcement strength of MiCAR against major dollar stablecoin issuers like Tether and Circle, particularly regarding the 60% minimum reserve threshold and service restrictions within the EU. Any of these conditions being realized would change the power dynamics between the Eurozone and dollar stablecoins. This article ECB Executive Schnabel: Stablecoins are a replica of MMFs, warning of Euro sovereignty risks first appeared on <a>...</a>.

Going head-to-head with Nvidia? Intel is rumored to launch AI chips by year's end, focusing on low power consumption to cut costs for businesses

According to the Financial Times, Intel is expected to roll out its latest AI chips by the end of 2026, aiming to capture market share from Nvidia and AMD. Intel's new product is said to address the power consumption and cost issues brought about by artificial intelligence (AI). This marks Intel's first major product launch in the data center sector since Lip-Bu Tan took over as CEO last March. No longer chasing ultra-high-performance chips, Intel is balancing cost and efficiency. Xeon product line director Kira Boyko stated that Intel's philosophy has shifted to helping customers find a balance between system costs and performance, rather than just selling the most powerful chips. It is expected that Intel's new products will drive server integration, reduce data center space requirements, and simultaneously increase data transfer rates. Intel plans to introduce the Crescent Island data center graphics processing unit (GPU) to customers in 2026, targeting the AI inference space, with samples expected to be shipped in the second half of the year, challenging Nvidia's dominance in the AI chip market. Next-gen GPUs ditch high-bandwidth memory. Intel has faced challenges in the AI chip market, with the Gaudi 3 accelerator positioned as a lower-cost alternative for AI training and inference, ultimately failing to change the market landscape. The costs associated with Gaudi inventory listed in Intel's financial reports for 2024 and 2025 directly reflect unmet market demand, and this history will be a challenge Intel must overcome when promoting Crescent Island. To avoid repeating past mistakes, Intel will focus on AI inference rather than frontier model training, utilizing the Xe3P architecture and LPDDR5X memory instead of the high-bandwidth memory (HBM) commonly used in high-end AI accelerators. Avoiding the HBM supply chain crisis: lowering costs with air-cooled enterprise-grade servers. Unlike Nvidia and AMD, which heavily use HBM in their top-tier GPUs, Intel will enhance product competitiveness with different technology. According to Counterpoint Research, the HBM market is currently dominated by SK Hynix, Samsung, and Micron, with a market share of up to 95%. Analysts expect HBM prices to face a surge of up to 100% by 2027. Nvidia's advantage lies in selling "complete systems," integrating GPUs, networking, software databases, and development tools; whereas Intel will focus on chips suitable for air-cooled enterprise-grade servers, avoiding HBM supply pressures while handling common inference tasks at a highly competitive cost. Will Intel's products fill the gap in the mid-to-low computing demand market? Intel's business strategy isn't to completely replace top-tier Nvidia systems from day one, but to secure a customer base that can transition workloads painlessly. Many businesses in their daily AI tasks do not require the world's most expensive accelerators; they value stable and predictable performance and ample supply. If Intel can design and manufacture chips that lower operating costs, its product prices will be more competitive compared to rivals relying on TSMC's foundries, allowing it to carve out a niche in the AI inference market. (Intel invests billions to expand EMIB packaging capacity, officially launching its foundry revival plan) This article: Going head-to-head with Nvidia? Intel is rumored to launch AI chips by year's end, focusing on low power consumption to cut costs for businesses first appeared at.

NVIDIA's N1X PC ushers in a new era, and Ming-Chi Kuo estimates 10 million units shipped in two years.

On June 1st, NVIDIA's CEO Jensen Huang delivered a keynote at the Taipei Pop Music Center, announcing that 'A new era of PC' is here. This launch coincides with hype built alongside tech giants like Microsoft and ARM, with the market closely watching the newly released N1X processor. In response to this significant industry development, TF International Securities analyst Ming-Chi Kuo released the latest supply chain survey report. The report indicates that this chip is primarily targeting heavy users with high computational demands, estimating a terminal shipment of about 10 million units over the next two years. Taiwan boasts a mature semiconductor and hardware industry cluster, and it's expected that related chip packaging and testing firms, as well as terminal OEMs, will benefit first. The N1X processor aims at high-demand computational users. NVIDIA's N1X processor integrates Arm architecture with powerful GPU technology, specifically designed for users running large language models (LLMs) locally. The chip provides excellent on-device AI computational power, striking a balance between performance and portability. Analyst Ming-Chi Kuo estimates that in the next two years, the shipment volume will be around 10 million units, still in a niche market, primarily offering heavy computational users a new option outside of the Apple Mac ecosystem. OS support is key for AI PC adoption. Currently, most AI applications on PCs still rely on cloud computing power. For instance, while the shipment of certain popular models in 2026 is expected to increase by 100%, the main reasons remain low pricing and ecological advantages. Ming-Chi Kuo emphasizes that the core value of on-device AI lies in maintaining privacy and integrating user data across applications. Whether the N1X devices can drive a large-scale upgrade wave will heavily depend on Microsoft's support for AI computational scheduling in the Windows OS and the software workflow. NVIDIA's new chip release benefits Taiwanese stocks. With the launch of the N1X processor, Taiwan's mature PC supply chain is set to gain new momentum. According to current industry intelligence and supply chain news, the N1X processor is a collaborative product developed by NVIDIA and Taiwanese IC design giant MediaTek (2454), while companies responsible for NVIDIA chip testing and packaging, like King Yuan Electronics (2449) and ASE Technology Holding (3711), are expected to benefit first from the initial chip mass production demand. In the terminal market, global PC brand giants and design OEM/ODM companies are actively preparing new machines. As hardware devices equipped with the new chip hit the market in the second half of the year, it is estimated to drive revenue growth opportunities for related Taiwanese firms. This article, "NVIDIA's N1X PC ushers in a new era, and Ming-Chi Kuo estimates 10 million units shipped in two years," first appeared on <a>...</a>.

Abel takes the helm with his first acquisition, Berkshire expands its real estate footprint, acquiring TMHC at a 24% premium

Berkshire Hathaway announced on Sunday that it will acquire U.S. homebuilder Taylor Morrison Home Corp. (TMHC) for approximately $6.8 billion in an all-cash deal. The acquisition offer is priced at $72.50 per share, which is a 24% premium over last Friday's closing price, and the transaction is expected to close in the second half of this year. This marks the first major acquisition under CEO Greg Abel's leadership. This move not only expands Berkshire's presence in the real estate market but also provides an outlet for its cash reserves of up to $397 billion. After the transaction, Taylor Morrison will become a private company, with the current management team continuing operations. Amidst a recent decline in U.S. housing starts, Berkshire's counter-cyclical investment strategy has become the focus of market attention. Abel's debut acquisition as Berkshire's new CEO
Berkshire Hathaway's cash reserves reached a record $397 billion at the end of the first quarter this year. This $6.8 billion acquisition is Abel's first major deal since taking over as CEO last year. Year-to-date, Berkshire's stock has dropped 5.6%, while the S&P 500 index has risen by 10.7% during the same period. Market investors previously anticipated that Berkshire could enhance capital efficiency through large acquisitions, providing fundamental support for its weak stock performance. TMHC's market positioning and business integration strategy
Taylor Morrison is a leading community developer and builder in the U.S., with over 350 communities across 12 states, offering financial services such as mortgages, title, and insurance. CEO Abel stated in a release that he hopes to integrate these businesses into a unified platform in the future. After the deal closes, Taylor Morrison will officially delist and become a private entity, with the existing senior management team, including CEO Sheryl Palmer, retained to continue leading the company's operations. Berkshire's counter-cyclical positioning in U.S. real estate
Berkshire has long invested in the homebuilding sector, owning Clayton Homes and holding shares in Lennar Corp. This acquisition comes at a time when new housing starts in the U.S. are slowing; the latest data shows a 2.8% decline in new residential construction in April, with single-family home starts down by 9%. Berkshire's decision to expand at this time reflects its long-term strategic positioning and capital allocation considerations in the U.S. real estate market. This article Abel takes the helm with his first acquisition, Berkshire expands its real estate footprint, acquiring TMHC at a 24% premium first appeared on <a>...</a>.

77% returns? Li Hong (1780) is going public, what should you pay attention to in the stock lottery?

Li Hong (1780) will kick off its public offering on 6/5, with an underwriting price of NT$18.88 and a latest closing price of NT$33.66, estimating a potential profit of around NT$14,000 if you hit the lottery, translating to a 77% return. What is a stock lottery? How do you participate? Li Hong (1780) operates in Taiwan's OTC biotech sector, classified under "biotechnology and medical industry," mainly focusing on the R&D and manufacturing of functional food ingredients and preventive medicine products. Its core business includes the R&D, manufacturing, and sales of various carotenoid-related products. What is a stock lottery? The official name for a stock lottery is "public offering," a mechanism where companies issue new shares to the public to raise funds. When a company undergoes an IPO or a cash increase post-listing, it is required by law to allocate a certain percentage of shares for public subscription. Investors need to apply through a brokerage trading platform, ensuring that their settlement account has enough funds and related fees by the specified deduction date. If the number of subscribers exceeds the available shares, the Taiwan Stock Exchange will conduct a random draw via computer to ensure fairness and transparency in the allocation mechanism. What are the risks of stock lotteries? Most investors participate in lotteries driven by the potential profit margin between the "underwriting price" and the "market price." To attract market subscriptions, the lead underwriter typically offers a discount on the underwriting price. However, participating in the offering involves the cost of capital being tied up. On the deduction date after the subscription window closes, the full share amount will be deducted and frozen for several trading days until the refund date for unsuccessful entries. In practical terms, investors must also consider implicit trading costs. Each participation incurs a processing fee of NT$20 and a mailing fee of NT$50 for the lottery notification; if unsuccessful, the share funds and the NT$50 postage will be refunded, meaning the NT$20 is a sunk cost. Li Hong (1780) is transitioning from OTC to listed, referencing the lower liquidity and potential price deviation of OTC prices; whether it can maintain the same high price when listed is under scrutiny, and investors must assess whether the price differential is sufficient before entering the lottery. (Yitai Industrial (7818) got trapped in the lottery, what should you pay attention to in the stock lottery?) How to participate in the stock lottery? Based on Li Hong (1780)'s underwriting price of NT$18.88 and the latest closing price of NT$33.66, a potential profit of around NT$14,000 can be estimated, with a return of 77%. However, this time, only 2,381 shares are being released for public subscription, with an estimated lottery success rate of about 1%. The lottery process is straightforward: just log into your brokerage app between 6/5 and 6/9 before 2 PM, click on stock lottery, and deposit NT$18,950. Results will be announced on 6/11. This article "77% returns? Li Hong (1780) is going public, what should you pay attention to in the stock lottery?" first appeared on .

MediaTek AI Without Limits Unveils End-to-End AI Solutions

MediaTek showcased its AI solutions themed 'AI Without Limits' at this year's Computex Taipei. The exhibition included products like the supercomputer NVIDIA DGX Spark, Wi-Fi 8, 6G, automotive intelligence systems, and end applications in collaboration with NVIDIA. General Manager Chen Kuan-Chou stated that MediaTek has advantages from edge to cloud and will continue to partner with NVIDIA and global collaborators to expand investments in AI infrastructure. MediaTek and NVIDIA showcased the supercomputer 'NVIDIA DGX Spark' at this year's Computex, featuring the NVIDIA GB10 Grace Blackwell super chip. This device boasts 1 petaFLOP performance with GPU and a 20-core CPU, utilizing LPDDR5x unified memory architecture, enabling the operation of large AI models directly on the device for automated intelligent coordination tasks. The venue also featured gaming display control chips equipped with NVIDIA G-SYNC Pulsar technology. (Celebrating the global launch of the smallest AI supercomputer! Jensen Huang personally delivered DGX Spark to Musk at SpaceX) The world's first satellite call-enabled automotive chip debuted MediaTek unveiled the Dimensity cockpit flagship platform C-X1, integrating NVIDIA application technology, supporting hybrid computing at the edge and cloud, upgrading the cockpit system to an active AI intelligence agent. Also showcased was the Dimensity automotive connectivity flagship platform MT2739, the world's first automotive chip that supports 3GPP R18 5G NR-NTN satellite calls, claiming seamless video communication from anywhere on the globe. It also features MediaTek's AI communication technology (MediaTek Modem AI, MMAI) which can reduce signal switching lag by 30%, ensuring applications like video conferencing and mapping software remain unaffected when vehicles enter basements or tunnels. Smart terminal application products unveiled The event showcased several high-performance computing platforms, including smartphones and tablets supporting offline AI generation, e-readers for meeting transcription, and efficient Chromebook platforms. In the display market, they launched a gaming display control chip supporting 5K AI image quality enhancement and IoT platforms suitable for drones and industrial retail. The world's first AI main chip supporting Dolby Vision 2, MediaTek Pentonic 800, was also a highlight, integrating next-gen image engines and content intelligence technology, capable of intelligently adjusting brightness and contrast based on playback content, delivering director-level high-spec editing perspectives, with several TV products featuring this chip expected to hit the market this year. Advanced packaging optical technology performance To meet the demand for large-scale AI data center deployments, MediaTek introduced an end-to-end AI data center platform encompassing customized ASIC and XPU designs, 2.5D / 3.5D advanced packaging, and high-speed interconnect technology. Addressing the industry's trend toward silicon photonics to enhance bandwidth density, they showcased co-packaged optics (CPO) technology with bandwidth rates of up to 400 Gbps / Fiber and MicroLED optical technology for active optical cables (AOC) used in data center interconnections. This technology can be integrated into existing CMOS transceivers, offering copper-level reliability while reducing power consumption by 50%. This MicroLED optical technology will also extend to CPO and NPO technologies, helping customers improve performance per watt. The world's first AI-powered wireless communication, 'Wi-Fi 8', and 6G communication technology results displayed In the communications arena, MediaTek introduced the Filogic 8800 chipset, enhancing the connection stability of existing Wi-Fi devices through Wi-Fi 8's dynamic sub-band operation (DSO) technology, achieving a system throughput increase of up to 200%. The AI-powered MediaTek Filogic AI technology is the world's first intelligent Wi-Fi powered by AI, with a network doctor feature capable of reducing repair time to under a minute, significantly lowering operational overheads, while the energy-saving mode can reduce power consumption by 50%. Products featuring the T930 5G fixed wireless access (FWA) platform were also showcased. Radio Interoperability '6G wireless access interoperability' supports both edge AI and cloud AI, achieving excellent scheduling flexibility with high-speed transmission while ensuring low network latency and low power consumption, paving the way for future hybrid edge and cloud Agentic AI collaborative models, suitable for large-scale AI applications connecting AI with robots, humans, and services. The 6G device collaborative multi-antenna (Device Collaborative MIMO, Co-MIMO) technology can increase downlink throughput by over 60%. This article MediaTek AI Without Limits Unveils End-to-End AI Solutions first appeared on.

TWSE's Lin Hsiu-Ming in Bloomberg Interview: Bridging Taiwan's AI Industry with International Capital

Taiwan Stock Exchange Chairman Lin Hsiu-Ming and Wistron Chairman Hong Li-Ning recently sat down for a Bloomberg interview, focusing on 'Beyond the Exchange, Behind the World’s AI', highlighting Taiwan's pivotal role in the global AI supply chain. Lin noted that Taiwan boasts a complete AI ecosystem from wafer foundry to advanced packaging and system assembly. This highly consolidated geographic advantage brings a unique high-valuation ecosystem to the capital markets. Hong Li-Ning pointed out the surge in global data center demand, with Taiwan's vibrant and highly liquid financial environment demonstrating an ecosystem advantage where industry and capital markets support each other. Taiwan's AI industry holds a 'full stack' advantage. Chairman Lin emphasized that international investors should reassess the overall value of Taiwan's supply chain, as global AI infrastructure demand is poised for explosive growth. Taiwan is not only an indispensable core for AI technology and manufacturing but has also become a key hub supporting the expansion of related industries in the capital markets. The Taiwan Stock Exchange will continue to serve as a platform connecting international capital with local innovative industries, aiding listed companies in deepening investor relations and transforming technological advantages into long-term investment value. Taiwan's unique strengths in semiconductors and advanced computing hardware make it the only region in Asia with a full-stack supply chain, from chip manufacturing to end-server platforms. Lin pointed out that this high level of industry transparency and cluster effect creates a unique Ecosystem Valuation Premium, and international investors value this comprehensive AI supply chain when assessing the Taiwanese market. Taiwan has 81.44% of its stock market value concentrated in tech-related industries. As AI development enters the infrastructure deployment stage, global cloud service providers are significantly increasing capital expenditures. Hong Li-Ning noted that international giants are entrusting key hardware needs to Taiwan's supply chain, with the high liquidity of Taiwan's capital markets providing robust cross-border financing channels and funding momentum. Currently, the Taiwan Stock Exchange (TWSE) has grown to be the fifth largest stock market globally, with as much as 81.44% of its market value concentrated in tech-related industries, making Taiwan a critical hub for global AI capital pricing. Lin stated that the stock exchange is actively connecting international capital with local enterprises, deepening international investment communication channels. (NVIDIA's AI ecosystem hardware super partner: Cooling system vendor Qihong Technology invests $400 million in Vietnam) TWSE will participate in this year's COMPUTEX Taipei International Computer Show, themed 'AI Together', from June 2 to 5. During the event, the Taiwan Stock Exchange will hold a media briefing, with representatives from several listed tech companies in attendance, including Delta Electronics (2308), Nanya Technology (2408), Qihong Technology (3017), Xinxing Electronics (3037), and Wistron (3231). The briefing will delve into the role of Taiwan's capital markets and listed companies in the global AI development landscape. This article, TWSE's Lin Hsiu-Ming in Bloomberg Interview: Bridging Taiwan's AI Industry with International Capital, first appeared on.

NASA ETF pulls in $2.6 billion in two months as retail traders position ahead of SpaceX IPO

The Space Innovators ETF (ticker: NASA), launched by US ETF issuer Tema ETFs on March 30, 2026, crossed the $1 billion asset mark in just 37 trading days, reaching a total of $2.6 billion by last weekend. The rapid scaling of this fund is primarily due to its direct holdings in SpaceX private equity, making it a key ETF channel for retail investors to gain exposure before SpaceX's official IPO. Active ETF Design: Direct Holdings in SpaceX Private Equity NASA is an actively managed ETF with an annual net expense ratio of 0.87%, where the Tema ETFs team actively selects stocks instead of tracking a passive index. Its unique feature is the direct inclusion of SpaceX private equity in its portfolio, allowing retail investors to indirectly gain positions in SpaceX prior to its public listing through brokerage accounts. This design is atypical in the US ETF market, where most new issuances primarily cater to institutional investors. SpaceX has submitted its IPO prospectus to the US Securities and Exchange Commission (SEC) and is expected to list on Nasdaq under the ticker SPCX, aiming for a target valuation of about $1.75 trillion, which the market views as one of the largest IPOs in history. SpaceX and its underwriting team have also adopted an unusual approach by reserving a direct channel for retail investors to participate in the subscription through brokers, creating two pathways for retail exposure to SpaceX alongside the NASA ETF. Retail Rush for SpaceX: ETF Channel and Direct IPO Subscriptions Working in Tandem The speed at which NASA ETF has attracted $2.6 billion reflects the strong demand from retail investors for "private equity" assets, which were previously limited to institutions and accredited investors, ahead of the SpaceX IPO. Besides NASA, other space economy-themed ETFs like Destiny Tech100 (DXYZ) and VCX have also seen significant inflows due to expectations surrounding the SpaceX IPO. For individual investors, gaining exposure to SpaceX through an ETF bypasses the accredited investor threshold and allows for daily trading on public markets, but they must bear the active management fees associated with the ETF structure, as well as the lock-up conditions for the SpaceX positions within the ETF—after SpaceX goes public, the ETF will still need to adhere to the original private equity lock-up terms and cannot immediately sell on the public market. ETF Scale Competition and SpaceX's Potential June Listing Timeline Recent news of SpaceX lowering its IPO valuation to about $1.8 trillion, with a possible listing as early as mid-June, has accelerated fund inflows into the NASA ETF and other related ETFs. Once SpaceX completes its listing, its stock will directly enter the public market, allowing retail investors to buy SPCX directly, potentially diminishing the "pre-IPO exposure" premium of existing ETF channels and slowing the expected growth rate of ETF scales. This ETF scale competition is a concrete example of market structure innovation—actively managed ETFs are packaging private equity into publicly tradable products, allowing retail investors to take positions before major tech company IPOs. Whether NASA's success will attract more ETF issuers to follow suit will be a key observation point for the upcoming trend of private equity securitization. This article "NASA ETF pulls in $2.6 billion in two months as retail traders position ahead of SpaceX IPO" first appeared on <a>...</a>.

Claude Code Dynamic Workflows: Coordinate 1,000 Subagents in Parallel

With the launch of Claude Opus 4.8, Anthropic has rolled out a new feature in Claude Code called "Dynamic Workflows" that allows Claude Code to coordinate hundreds to thousands of parallel subagents at once. Currently, this is open in a research preview state for users on Enterprise, Team, and Max plans. Independent tech analyst Akshay Pachaar released a detailed breakdown on May 29, revealing the core differences between Dynamic Workflows and the existing subagents and agent teams multi-agent mechanisms. Core Innovation: Write Orchestration Scripts in JavaScript The key design of Dynamic Workflows is to allow Claude to no longer store the entire task plan within its own context window, but instead to "write a JavaScript orchestration script" for the background runtime to execute. After the user describes the task, Claude automatically generates a script to determine how to split the task, allocate it to how many parallel subagents, and how to verify the results, collectively referred to as the "zero orchestration burden." In terms of scale, Dynamic Workflows support 16 parallel executions, with a single workflow coordinating up to 1,000 subagents. Each subagent reports back the results after completing their tasks, and Claude performs "adversarial verification" on these results—cross-checking from different angles to confirm the output's reliability before responding to the user. Distinction from Existing Multi-Agent Mechanisms Anthropic has already provided two types of multi-agent mechanisms in Claude Code: First, Subagents are lightweight, isolated workers that report back to the parent agent, suitable for task distribution; second, Agent Teams allow multiple Claude instances to collaborate through a shared task list, but practically limited to about 5 parallel instances. Dynamic Workflows, however, significantly raises the parallel scale—from 5 to 1,000—and hands over the orchestration logic to the JavaScript scripts generated automatically by Claude. Anthropic states that this mechanism allows Claude Code combined with Opus 4.8 to achieve "codebase migrations on the scale of hundreds of thousands of lines of code," using existing test suites as acceptance criteria. Usage Recommendations and Limitations In his analysis, Akshay Pachaar provides usage recommendations: when trying it for the first time, start with "small tasks with limited scope" to calibrate token usage, enable auto mode, use ultracode settings, and initially review the execution plans generated by Claude. Enterprise administrators need to manually enable Dynamic Workflows in the backend to use it. For developers, Dynamic Workflows change the "task scale limit" of Claude Code: previously, agent workflows were limited to a dozen parallel steps, but now they can handle hundreds to thousands of parallel tasks at once, which has practical productivity implications for codebase migration, cross-service exploration, security audits, and other large-scale work. This feature is still in the research preview stage, and how it will expand to more subscription plans in the future is worth further observation. This article Claude Code Dynamic Workflows: Coordinate 1,000 Subagents in Parallel first appeared on.

US Treasury Secretary Bessent: The US has seized $1 billion in crypto assets linked to Iran, "we just outright grabbed the wallets"

US Treasury Secretary Scott Bessent publicly stated on May 29, 2026, during the Reagan National Economic Forum in Simi Valley, California, that the US government has seized approximately $1 billion in crypto assets associated with the Iranian regime, bluntly saying, "we have just outright grabbed the wallets." In an interview with Fox Business Network host Larry Kudlow, Bessent added, "Some of them might be inputting passwords right now, but they have no idea their wallets have been grabbed." This revelation marks the cumulative result of the US's "Operation Economic Fury" against Iran to date. From $500 million to $1 billion: The $1 billion figure represents an updated tally of the cumulative results of "Operation Economic Fury" since early 2026. In April 2026, the US Treasury had frozen $344 million in USDT on the Tron blockchain, with the previously announced total seizure amounting to about $500 million, which has now doubled to $1 billion as disclosed by Bessent, reflecting a two-month spike. Bessent stated that the seizure operation targets the digital assets of the Islamic Revolutionary Guard Corps (IRGC) and Iranian regime officials, forming part of the US's overall economic pressure strategy to cut off Iran's external funding, oil revenues, and sanction evasion channels. He noted that the US is collaborating with European allies to track down villas, real estate, offshore accounts, and other assets controlled by Iranian officials and the IRGC abroad. The technical positioning of "outright grabbed the wallets" has sparked discussion in the crypto community. On a blockchain technical level, "grabbing the wallets" means that the US has gained control of the private keys to the target wallets, allowing them to directly transfer the balances. Specific methods may include coordinating with stablecoin issuers (such as Tether) to implement blacklist freezes before transferring, or executing seizures through exchanges by complying with KYC connections. The Treasury has not publicly disclosed all the on-chain transaction hashes related to the seizures. Bessent framed the seizure actions as "taking back the money stolen from the Iranian people," against the backdrop of President Trump's "maximum pressure" policy towards Iran. This revelation coincides with a critical moment in negotiations regarding the US-Iran ceasefire agreement under the Trump administration. Visibility signals for the crypto industry: This revelation highlights three things for the crypto industry. First, US law enforcement's capability in blockchain investigations and asset seizures has reached an industrial scale; the $1 billion figure is not a one-off seizure but rather the sum of multiple actions. Second, collaboration with stablecoin issuers (especially Tether, which has previously worked with OFAC and the Justice Department to freeze addresses) has become a critical infrastructure for executing sanctions. Third, the intertwining of geopolitical sanctions and cryptocurrency asset tracking is accelerating, and US pressure on digital assets linked to Iran, North Korea, and Russia is expected to expand further. This article US Treasury Secretary Bessent: The US has seized $1 billion in crypto assets linked to Iran, "we just outright grabbed the wallets" first appeared on <a>...</a>.

This video is about the US stock #美股2026
Come on, hurry up and buy!
It's only going to hit new highs!
Little do they know, Wall Street is watching with binoculars #華爾街陰謀

Arca's Chief Investment Officer Jeff Dorman Warns: MicroStrategy Situation Out of Control, Someone Will Fall Within Four Months

Jeff Dorman, Chief Investment Officer of digital asset management company Arca, recently issued a stern warning regarding MicroStrategy's (MSTR) financial situation, stating that it has "lost control." He pointed out that this company, regarded as the largest corporate holder of Bitcoin, is simultaneously putting common stock shareholders, preferred stock investors, and Bitcoin holders in an unprecedented collective predicament. He warned that this crisis could lead to a liquidation moment within four months. Dorman: MicroStrategy's Heavy Bet on Bitcoin Misfired, Preferred Shares Become the Biggest Pressure Dorman noted that MicroStrategy's decision to issue preferred shares (STRC) presupposed a position: "Bitcoin is about to surge significantly, and in the future, selling just a small amount of holdings will easily cover up to 11.5% annualized dividends." However, Bitcoin has not soared as expected; in fact, its current price is nearly down 30% compared to a year ago. This has turned the approximately $15 billion of preferred shares into the biggest pressure point. The annual fixed dividend exceeding $1.5 billion weighs on an increasingly deteriorating balance sheet, and market doubts about the company's ability to truly meet these obligations are rapidly increasing. In fact, as early as the beginning of May this year, when MicroStrategy released its Q1 financial report, Saylor unprecedentedly hinted that the company might sell a small amount of Bitcoin to cover preferred stock dividends, causing panic in the market and prompting some investors to reassess the feasibility of this flywheel model. Raised $2 billion but Burned Through Lifeline, Dorman Critiques Decision Logic as Absurd In the face of liquidity pressure, MicroStrategy raised $2 billion in cash through a new stock issuance, temporarily providing about two years of dividend buffer space and alleviating the most urgent market concerns about default. However, what Dorman finds hard to understand is that the company immediately used this capital to buy back zero-coupon convertible bonds maturing in 2029 at a discount, instead of keeping it as a liquidity moat for dividend guarantees. Dorman wrote: "Why use the only cash on hand to pay off a 0% interest debt?" He pointed out that while this buyback operation slightly improved the balance sheet, it consumed critical resources that should have protected the company's future payment capacity, effectively accelerating the timetable for financial pressure to explode. Warning Signals in the Bitcoin Market, Is MicroStrategy Holding the Line Alone? Previously, Bloomberg data showed that MicroStrategy has accumulated over 170,000 Bitcoins this year alone, estimated to account for about 70% of the overall market buying pressure, far exceeding the total output of around 62,000 Bitcoins from global miners during the same period. Analysts pointed out that the funds from the U.S. Bitcoin spot ETF, which played a significant role in the 2024 bull market, have significantly retreated due to the disappearance of arbitrage premiums; retail investors in South Korea are fleeing the crypto market due to strong performances from local semiconductor stocks; and miners are generally shifting to sell immediately upon mining, reallocating funds to AI infrastructure. Now, with a market cap of $1.5 trillion, the demand for this asset has contracted from a previously diverse buying structure to nearly just MicroStrategy acting as a "Bitcoin machine" holding the line alone. The operation of this machine hinges on Saylor's ability to continuously raise funds in the capital market. Once this financing chain breaks, the market could face a domino effect. (ETF Power Dwindling, Retail Exodus, Is Bitcoin Price Only Supported by MicroStrategy?) Countdown to Liquidation in Four Months, Collapse May Affect All Parties Dorman's warning at this juncture is particularly heavy. He pointed out that this is the first time in history that all stakeholders of MicroStrategy, including common stock shareholders, preferred stock investors, and the vast number of Bitcoin holders, are simultaneously exposed to downside risk, with no one able to remain unscathed. He candidly admitted that Saylor might indeed have undisclosed capital market plans, such as refinancing debt with longer-term instruments, and those who have historically underestimated Saylor often pay the price. However, Saylor himself has publicly ruled out the possibility of restarting convertible bonds, significantly narrowing the space for potential solutions. Before this, if Bitcoin's price cannot rebound significantly, Dorman believes the most likely outcome is being forced to sell Bitcoin to cope with financial pressure. Such a sell-off, if it occurs in a continuously weakening price environment, would simultaneously drag down both Bitcoin's price and MSTR's stock price, creating a double whammy. Dorman left a heavy remark: Within four months, someone is bound to be seriously hurt and exit. This article Arca's Chief Investment Officer Jeff Dorman Warns: MicroStrategy Situation Out of Control, Someone Will Fall Within Four Months first appeared on .

Why are the memory giants investing in Anthropic? A $10 billion order could be the turning point for Samsung's foundry

Anthropic has completed the largest AI financing round in history, with Micron, Samsung, and SK Hynix stepping in as strategic partners. The industry interprets that the collaboration between Samsung and Anthropic will extend from memory supply to chip foundry, bringing a crucial turnaround for Samsung's long-struggling foundry division. Anthropic's valuation has reached $965 billion, officially surpassing OpenAI. On the 28th, Anthropic announced the completion of a Series H funding round, raising $65 billion, bringing the post-money valuation to $965 billion, officially exceeding OpenAI's $852 billion, making it the highest-valued AI startup globally, just a step away from the $1 trillion mark. This round was co-led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, with Amazon continuing to invest $5 billion as part of its previously committed $15 billion cloud collaboration. (Anthropic raises $65 billion, valuation of $965 billion surpassing OpenAI) The three major memory manufacturers participated in the investment as 'strategic infrastructure partners'. Notably, Micron, Samsung Electronics, and SK Hynix, the three leading global memory manufacturers, have invested simultaneously as 'strategic infrastructure partners'. Anthropic emphasized in its statement that the technologies of these three companies play a critical role in the global supply of memory, storage, and logic chips, stating that 'collaborating with these partners will significantly help us reliably scale compute power to meet customer demands.' Among them, Anthropic specifically mentioned 'logic chips'. The manufacturing of logic chips falls under the foundry business, and among the three investors, only Samsung Electronics has an independent foundry division, while Micron and SK Hynix do not have this business. According to the Korea News Agency, the collaboration between Samsung and Anthropic will not only be on the memory supply level but may also further extend to AI chip foundry manufacturing. Samsung's foundry is expected to secure the production orders for the AI chips used in Anthropic's flagship model, Claude. If the collaboration is confirmed, Samsung Electronics will have the opportunity to take on the AI chip production orders for Claude. As a leading AI model company globally, securing its foundry orders holds significant strategic importance for Samsung. In fact, Samsung's foundry has recently landed heavyweight clients, including Tesla's next-gen AI chips 'AI5' and 'AI6', as well as NVIDIA's inference-specific language processing chip 'Groq3', all produced by Samsung's foundry. The supply contract for the image sensor of Apple's new iPhone model next year has also been finalized. If Anthropic's chip foundry orders are added, Samsung's client lineup will be further strengthened. (Samsung Chairman Lee Jae-yong's low-key visit to Taiwan, rumored to be negotiating with MediaTek for major foundry orders) Currently, Samsung's foundry ranks second in the global foundry market, with a market share of about 7.2%, but compared to the leading TSMC's 69.9%, the gap is still significant, and the company's foundry business has been in the red for several consecutive years. The market is paying attention to whether it can return to profitability next year. Samsung is expanding its AI ecosystem, comprehensively laying out its semiconductor supply chain. Industry insiders analyze that Samsung's investment in Anthropic is not just a financial maneuver but a comprehensive strategic positioning that spans 'AI semiconductors, memory, foundry, and AI infrastructure'. Samsung's differentiation strategy focuses on integrating high bandwidth memory (HBM), advanced process foundry, and advanced packaging technologies into a 'one-stop solution', attempting to establish an irreplaceable comprehensive competitiveness within the AI supply chain. As the AI market evolves from mere memory procurement to encompass chip design, manufacturing, packaging, and system optimization, Samsung, with capabilities across the entire semiconductor industry chain, is actively seeking to establish deep-binding relationships with key players in the AI era. Industry insiders point out: 'Samsung's investment is a signal indicating that it is seriously expanding strategic relationships with key enterprises in the AI era, and the market anticipates whether Samsung's foundry can seize new opportunities amid the AI market expansion wave.' Memory suppliers transforming into AI infrastructure partners. The collective investment of the three major memory manufacturers in Anthropic symbolizes the evolving relationship between the semiconductor industry and AI model companies, transitioning from pure component supply and demand to deep strategic symbiosis. For Samsung Electronics, successfully securing Anthropic's foundry orders will not only help stop the bleeding in its foundry business but will also solidify its core position in the next generation of AI infrastructure. This article on why the memory giants are investing in Anthropic? A $10 billion order could be the turning point for Samsung's foundry first appeared in.

Grayscale's HYPG ETF is looking to buy 2 million HYPE tokens, totaling about $115 million.

Asset management company Grayscale filed the fourth S-1 amendment for the "Grayscale Hyperliquid Staking ETF" with the SEC on May 28, 2026. They revealed that they purchased the initial seed shares at $100 on April 22, 2026, and plan to increase their position with an additional 20,000 shares at $25 each, totaling $500,000. They are also negotiating with a private investment firm named Hyper Holdings Global LP to acquire around 2 million HYPE tokens for the seed round, which, at the current HYPE price, equates to approximately $115 million. Both parties are still in negotiations, and no final terms have been set. This trust was originally named "Grayscale HYPE ETF" but was renamed after amendments to the Delaware statutory trust law on May 26. HYPG will be listed on Nasdaq, with Anchorage Digital acting as the custodian. The amendment also disclosed the ETF's operational structure: the sponsor is Grayscale Investments Sponsors, LLC, the trustee is CSC Delaware Trust Company, the custodian is Anchorage Digital Bank N.A., and the advisor is West Capital Advisors LLC. The updated trust name includes "Staking," indicating that Grayscale is also applying to include staking rewards in the design. The final structure and whether staking will be approved are still pending SEC feedback. Competition in the Hyperliquid ETF space is heating up, with the current HYPE spot ETF market already featuring Bitwise BHYP (listed on NYSE, total assets $45.73 million) and 21Shares THYP (listed on Nasdaq, total assets $45.75 million), bringing the total net assets in the HYPE ETF category to about $118 million; on May 27, the HYPE ETF category saw a net inflow of $3.4 million. If Grayscale can successfully implement the 2 million HYPE seed position, HYPG will be able to match the scale of the existing two ETFs combined at launch. HYPE is the native token of the decentralized perpetual futures exchange Hyperliquid, which saw its price hit an all-time high of $64.63 on May 26, before currently fluctuating in the $56 to $62 range. SEC scrutiny on DEX-based ETF review standards is a focus. This marks the fourth S-1 revision for HYPG, and the market is keen on whether the SEC will apply the same review standards to tokens like Hyperliquid's "DEX-based token ETFs" as it does to existing spot ETFs, and whether the staking reward design can achieve breakthroughs in this instance. This article about Grayscale's HYPG ETF seeking to purchase 2 million HYPE tokens, totaling about $115 million, first appeared on .