Anthropic IPO and Nvidia’s PC Chip Push Markets to New Highs—Here’s What’s Really Going On

Nvidia unveils the RTX Spark PC superchip. At Computex in Taipei, Nvidia CEO Jensen Huang introduced the RTX Spark (code-named N1X), the company's first consumer system-on-chip that pairs a 20-core Arm CPU with a Blackwell GPU packing 6,144 CUDA cores—the equivalent of a desktop RTX 5070 . Developed in partnership with MediaTek and built on TSMC's 3-nanometer process, the chip is designed for premium AI laptops and will ship in the second half of 2026 in models from Microsoft Surface, Dell XPS, HP EliteBook, ASUS, Lenovo, and MSI . Nvidia's move directly challenges Qualcomm, Intel, AMD, and Apple's M-series processors, and expands the company's total addressable market well beyond its data-center GPU stronghold. The announcement sent Nvidia shares up 6.3% and Arm Holdings up 15.6% on Monday .

By Tuesday morning, the chip sector remained red-hot with outsized gains for Marvell and Broadcom, though U.S. futures edged lower as some profit-taking set in . The rally was narrowly tech-led—enthusiasm for AI broadly lifted technology shares, European equities advanced, oil retreated from war-driven highs, and bond yields declined .

Geopolitical risk: Iran tensions that tech overwhelmed

A significant geopolitical overhang threatened to derail the rally. Iran suspended indirect negotiations with the United States in protest of Israeli military actions in Lebanon, which sent oil prices higher on Monday . Charles Schwab characterized the situation as "the tech rally out-muscling troubling war news" and an "epic clash between rising oil prices and bullish chip sentiment" .

Oil prices later retreated on Tuesday, easing some pressure on equities . But the risk remains live. Schwab also noted that Iran's announcement it plans to completely close the Strait of Hormuz could shape trading in the days ahead . For now, AI enthusiasm is winning the tug-of-war, but the geopolitical backdrop has not been resolved.

The concentration problem: what Goldman Sachs and Morgan Stanley are warning

In the weeks before the rally, Goldman Sachs and Morgan Stanley both published explicit warnings that the market's narrow leadership is dangerous. The very AI stocks driving the record highs are also the source of the market's fragility.

Goldman Sachs, in a May 19 global strategy note, pointed out that the S&P 500 had returned approximately 10% year-to-date in 2026, but technology, media, and telecom (TMT) sectors accounted for 85% of that return . The bank warned that investors are underestimating concentration risk and that this level of narrow leadership masks weakness in the broader market.

Morgan Stanley's Daniel Skelly, head of market research and strategy at Morgan Stanley Wealth Management, went further on May 29. He advised investors to pivot AI infrastructure names from "buy" to "hold," warning the AI trade could cool in a setup "similar to the mid-1990s" . His specific concern was concentration in semiconductors, IT hardware, and power—the frontline beneficiaries of AI spending . Skelly wrote that the extent of the recent rally means it may now be better to hold positions than to add to them.

Goldman Sachs separately acknowledged in a May 6 investor note that elevated concentration and narrow market leadership create "greater sensitivity to earnings disappointment and the increased probability of a disorderly market correction" . The note highlighted that the Magnificent 7 stocks had already begun underperforming the broader market in 2026.

Adding a systemic dimension, S&P Global published a report on May 4 flagging that hedge fund prime-brokerage financing is now concentrated across just four banks—Goldman Sachs, Morgan Stanley, Barclays, and BNP Paribas—at leverage multiples not seen since before the 2008 financial crisis . Gross leverage across the industry has climbed to approximately eight times NAV, up from five times a decade ago .

Collectively, these warnings paint a picture of a market riding high on AI momentum but built on a narrow, leveraged foundation. The rally on June 1-2 was real and powerful, but it amplified the very risks the banks had just flagged.

AI crypto tokens and ETF flows: an evidence gap

Despite the user's question about the contrasting performance of AI-linked crypto tokens versus persistent Bitcoin and Ethereum ETF outflows, specific data for June 1-2, 2026 was not captured in the available sources. This part of the picture remains an evidence gap. The general pattern observable through spring 2026 is that AI-themed crypto tokens—such as Render, Bittensor, and Near Protocol—have periodically rallied in sympathy with tech and AI news, while U.S. spot Bitcoin and Ethereum ETFs have seen persistent institutional outflows amid geopolitical uncertainty and rising interest rates. However, that broad trend cannot be confirmed with June 1-2 data from the provided sources.

What this means for investors

The rally on balance is an AI-driven, narrow, momentum-fueled advance that is overriding geopolitical risk but sitting on a concentrated base that both Goldman Sachs and Morgan Stanley have publicly flagged as fragile . The twin catalysts—Anthropic's IPO path and Nvidia's PC chip—are genuinely significant and will likely shape market narratives for months. But the warnings from the banks are not theoretical: they are rooted in data showing that a handful of stocks are carrying the entire market, and that the leverage behind them is historically high.

Any earnings disappointment in the Magnificent 7 or a shift in AI sentiment could trigger precisely the kind of disorderly correction that Goldman Sachs described as an elevated probability . The market is priced for AI perfection. The question is how long that perfection holds.