Goldman Sachs: The AI investment wave will continue to dominate the S&P 500’s second-quarter earnings season, but investors are starting to focus on return on investment.

According to Bitget market data, South Korea’s KOSPI index rose 2.00% intraday. Samsung Electronics gained 4.49%, and SK Hynix rose 1.98%.

ECB sources stated that while a rate hike could be delayed, it remains on the central bank’s agenda. The sharp drop in oil prices has eased pressure on the ECB to raise rates in July, boosting the likelihood of a September rate hike. (Jinshi)

Goldman Sachs' Timothy Moe and John Kwon noted that a 1 percentage point rise in the combined weight of Samsung and SK Hynix in South Korea's stock index could prompt foreign investors to pull around $2 billion from the South Korean market, as the U.S. Investment Company Act requires portfolios to meet diversification thresholds. Goldman Sachs also stated that a surge in capital into leveraged ETFs, combined with rising options trading and retail margin trading, has created a structural environment where daily price volatility far outpaces what corporate fundamentals can support. South Korea's asset management growth since last year has stemmed mainly from investment gains rather than new inflows. As valuations rise, institutional investors' mechanical exposure to market volatility has also increased—a trend often linked to hedging strategies. This means even a mild market correction could trigger a cascade of forced selling.

25-year-old Wall Street AI stock prodigy Leopold Aschenbrenner’s Situational Awareness LP disclosed in its latest 13G filing that the fund holds a 19.9% stake in SharonAI Holdings (SHAZ). The filing lists Leopold Aschenbrenner and Carl Shulman as filers. SharonAI Holdings Inc is a high-performance computing (HPC) firm that deploys large-scale energy and computing infrastructure and manages assets for the U.S. energy market and infrastructure. Per Bitget market data, SHAZ rose 14.07% in U.S. after-hours trading, currently trading at $92.4.

US-based sensor and power semiconductor solutions provider Allegro MicroSystems (ALGM) rose 14.7% on Monday, closing at $66.37 and hitting an intraday record high of $67.41. Mizuho Securities analyst Vijay Rakesh maintained the stock’s "Outperform" rating and raised its price target from $54 to $67, citing the company’s gains from surging demand for current sensors, fan drivers, and power semiconductors driven by the boom in AI servers. TD Cowen also named Allegro one of the "best picks" among small- and mid-cap stocks, setting a $70 price target. The firm notes that automotive electrification, autonomous driving, and AI data centers are all long-term growth drivers for the company, with its data center business currently accounting for around 14% of Allegro’s revenue. These rating adjustments signal Wall Street is seeking AI beneficiaries beyond Nvidia and Micron. AI data centers require not only GPUs and high-bandwidth storage, but also power supply, cooling, sensing, connectivity, and control systems. As capital expenditures from cloud vendors translate into actual construction orders, more small and mid-sized hardware firms are being included in AI-related deal pipelines. Analysts also point out that second-tier AI-related stocks typically have lower liquidity and higher valuation elasticity, making their share prices more sensitive to broker rating and price target adjustments. Allegro’s rally reflects market optimism about the expansion of the AI hardware supply chain, and also shows investors are willing to pay a higher premium for companies that are "indirect beneficiaries of AI".

T. Rowe Price portfolio specialist Amanda Ng noted in a report that the next phase of AI-related investment opportunities may lie further upstream in the supply chain, rather than the highly hyped AI platforms or chipmakers. She pointed out that areas like advanced packaging, semiconductor substrates, and high-end printed circuit boards (PCBs) are growing increasingly critical. While these components account for a relatively small share of the total bill of materials (BOM) required for AI, they could still become key bottlenecks. She also added that even a modest price increase for these products could significantly boost manufacturers' profits while remaining affordable for end customers.