Point of View: The Federal Reserve Has Over-Eased, with a 65% Chance of Stagflation in the United States

June 18: Gary Marcus, an MIT cognitive science PhD and AI researcher, warned that OpenAI’s potential IPO and ongoing cash burn issues pose risks to tech stocks heavily reliant on AI computing power, delivering a bearish take on the currently red-hot AI infrastructure sector. Marcus notes that valuations for companies including Nvidia, Oracle, and CoreWeave are partially tied to expectations that OpenAI will continue its large-scale chip procurement and data center buildouts. If OpenAI’s IPO process encounters difficulties, its valuations come under pressure, or the company cuts spending amid price wars, these suppliers could face downward revisions to their revenue forecasts. OpenAI submitted a confidential S-1 filing this month to lay the groundwork for a possible IPO. Meanwhile, the market is closely monitoring three key factors: OpenAI’s steep computing costs, its competition with Anthropic, and whether enterprise clients will reduce their usage due to the high costs of implementi

June 18 marked another high-profile day for Micron Technology amid AI hardware trading, as the semiconductor manufacturer once again became a focal point for investors. Multiple Wall Street institutions have recently raised price targets for the storage chip maker, citing surging demand for AI servers that’s tightening supply dynamics for DRAM, HBM, and NAND. Deutsche Bank analyst Melissa Weathers lifted Micron’s target price from $1,000 to $1,500, implying roughly 47% upside from the stock’s prior closing level. FactSet data shows TD Cowen and Cantor Fitzgerald have also set matching $1,500 targets for the company. Micron’s shares have rallied dramatically this year: up about 266% since 2026, and a staggering 750% over the past 12 months. The market’s core bull case rests on AI data centers needing not only GPUs but also increased high-bandwidth storage, enterprise SSDs, and standard DRAM to support model training, inference, and AI agent workloads. TD Cowen goes further, arg

June 18. U.S. semiconductor equipment stocks rallied on Wednesday after Citigroup raised price targets for Applied Materials, Lam Research, and KLA while maintaining its buy ratings, citing upwardly revised AI-focused capital expenditures (capex) by large tech companies. Citi analyst Atif Malik hiked the price target for Applied Materials from $550 to $710, Lam Research from $315 to $450, and KLA from $206.40 to $290. Per Investor’s Business Daily, Applied Materials closed up 4.4% on Wednesday, hitting a new intraday high, while Lam Research also briefly touched a fresh peak. The note is viewed by markets as a signal of repricing for the global wafer fabrication equipment (WFE) sector. Citi projects that driven by investments in AI computing power, expanded storage, and advanced processes, the WFE market could grow from $145 billion in 2026 to $200 billion in 2027, and reach $250 billion by 2028. For Asian markets, this dynamic translates directly to demand for semiconductor e

June 18: According to monitoring by ai_9684xtpa, Ethereum has declined again amid Fed Chair Jerome Powell’s hawkish remarks. A large Ethereum address that entered a position in 2024 (correcting the obvious typo of "1989" – Ethereum did not exist in 1989) with 5,972 ETH is suspected to have liquidated its holdings. Roughly 9 hours ago, this address deposited ETH worth approximately $10.59 million into Binance, at an average price of about $1,774. The position was established between March 27 and 28, and the holder exited after holding for nearly three months. Based on current Ethereum pricing, the total loss on this transaction is estimated at around $1.287 million. Historical holdings and transfer records for this address indicate that the holder has mostly completed a phased exit strategy.

On June 18, SK Hynix announced it has supplied samples of its 12-layer HBM4E to key customers, according to Securities Times. This next-generation ultra-high-performance DRAM is built for artificial intelligence (AI), featuring a maximum pin speed of 16 Gbps, over 20% improved energy efficiency, and enhanced data processing capabilities to support AI training and inference.

**June 18 Update: OKX to Gradually Launch RE Trading & Derivatives** OKX announced in an official statement on June 18 that it will roll out RE (Re) trading and derivative products in phases between June 18 and June 19. RE/USDT spot trading will go live at 22:00 (UTC+8) on June 18, with pre-market contracts being converted to standard perpetual contracts within three hours after the spot trading opens.