Arthur Hayes: I bet that Bitcoin will surpass $110,000 before retracing to $76,500

According to Bitget market data, most US space concept stocks advanced at the opening of US stock trading. Specifically: SpaceX (SPCX) slipped 0.75% to $151.8; Virgin Galactic (SPCE) gained 8%; AST SpaceMobile (ASTS) rose 3.45%; Rocket Lab (RKLB) added 4%; Redwire (RDW) gained 1.9%.

According to Bitget market data, US stocks opened with broad declines in optical communication concept stocks, including: Pure Photonics ETF (FOTO) fell 5.7%; Marvell Technology (MRVL) — a stock NVIDIA CEO Jensen Huang previously touted as the next trillion-dollar market cap company — dropped 6.1%; AAOI (Applied Optoelectronics) fell 7.1%; LITE (Lumentum Holdings) dropped 8.86%; COHR (Coherent Corp.) declined 8.4%; and CIEN (Ciena Corporation) fell 4.75%.

According to Bitget market data, the three major US stock indexes all fell at opening: the Dow Jones Industrial Average dropped 0.44%, the S&P 500 declined 0.67%, and the Nasdaq Composite fell 1.1%. The semiconductor and storage sectors saw broad declines, with individual stocks performing as follows: NVIDIA (NVDA) dropped 1.56%; Intel (INTC) fell 3.5%; Broadcom (AVGO) declined 2.5%; Qualcomm (QCOM) fell 0.4%; Seagate Technology (STX) dropped 5.78%; Western Digital (WDC) fell 7%; SanDisk (SNDK) declined 7.5%; Micron Technology (MU) dropped 5.4%.

Aave founder Stani has stated that the protocol is expanding its market from crypto assets to all asset classes via securities-collateralized lending. Brokers like Robinhood and Charles Schwab usually retain 50% to 85% of stock lending fees, returning only a small portion to their users. The global securities lending market is approximately $4.6 trillion in size, generating around $350 billion in annual revenue, most of which is captured by brokers. Tokenized stocks, through Aave V4, can return the full lending revenue directly to users, offering advantages including real-time transparency, dynamic pricing, no re-collateralization, and no intermediary fee deductions.

Earlier reports indicated that Federal Reserve Chair Kevin Warsh has selected two veteran central bank economists as advisors: Daniel Covitz, one of three deputy directors in the Research and Statistics Division, and Erik Enstrom, senior deputy director in the Monetary Affairs Division. Both are long-time Fed veterans with nearly 30 years of experience, deeply familiar with the Federal Reserve’s operations. Last week, Warsh also announced the establishment of five task forces to review the central bank’s communication practices, data analysis, and portfolio management, noting that these groups would be composed of external experts, with support from internal Fed subject-matter specialists. Covitz regularly prepared materials for Warsh’s speeches during Warsh’s tenure as a Fed governor from 2006 to 2011, with research focusing on financial stability and credit markets. Enstrom specializes in monetary policy and financial market analysis. Last year, he developed a model to assess the probability of various economic scenarios, estimating that by mid-2025, the risk of a combination of high inflation and weak growth had risen, replacing the earlier "soft landing" outlook. In February this year, the two collaborated on research explaining why long-term Treasury yields rose even as the central bank cut interest rates, attributing the phenomenon to investors demanding higher compensation for risks from adverse supply shocks and swelling federal deficits. The study also found no evidence that markets had lost confidence in the Fed’s ability to keep inflation near its 2% target.

Serenity tweeted about the phenomenon of U.S. companies’ heavy reliance on Chinese AI models, citing a UBS report that many enterprises have begun routing simple tasks to cheaper Chinese open-source models. Some teams spend up to $35,000 monthly on tokens—200% over budget—putting pressure on high-end AI model revenue growth. Serenity deemed the UBS report accurate, attributing the trend to a capitalist dilemma: markets naturally gravitate toward the cheapest option, and Chinese models like DeepSeek are significantly cheaper than those from Gemini, OpenAI, and Anthropic. Serenity also stated that the Trump administration’s earlier pause on access to Fable/Mythos was the right move, as repeated distillation of top-tier models poses enormous risks, and models approaching ASI-level should have higher access barriers. The expert noted that the U.S. needs two key actions: further develop models specialized in low-cost inference, and impose bank-grade identity authentication for AI model access.