(These are the market notes on today’s action by Mike Santoli, CNBC’s Senior Markets Commentator.) They say bull markets peak on good news, but not today. OpenAI’s latest gargantuan commitment for computing capacity via AMD chips – one that will be largely financed through a 10% equity position in AMD that was granted to Open AI — again punished anyone who bet that the AI hype was fully priced in. The brute force of the $1 trillion-plus in infrastructure commitments over the next several years has been too much for the immediate skepticism about the unfunded promises and overlapping spender/vendor relationships to offset. Indeed, the story of the market this year can be framed as trillions in capital and spending power being unleashed by the private sector and governments (Germany with a half-trillion-euro military tab, Japan shifting to fiscal expansion with a leadership switch). That said, the action Monday was respectable but uneven, with AMD adding some $70 billion in market value to contribute a third of the S & P 500’s roughly half-percent gain on the day, while Nvidia shed more than $40 billion as the search for scarce processing resources widens. More than half of NYSE stocks were lower on the day. The way this is all manifesting now is in a stock market led by high-volatility/low-quality stocks alongside a rush for gold and bitcoin as stores of value acting like momentum vehicles. This is both typical of strong, maturing bull markets and evidence of some extreme behavior that can generate eventual instability. Here’s the S & P 500 Quality ETF – companies with strong financials, clean balance sheets and durable profit margins – relative to the broader S & P 500, matching a ten-year low in relative performance. The broad debate seems to be between those saying “AI is a bubble just like in 1999” and others arguing “AI is a bubble in the making just like 1996.” The first thing to emphasize is that nothing is preordained about market cycles — this one may well never get as wild, lucrative and dangerous as the one 25+ years ago. In terms of market appreciation alone, the Nasdaq Composite over the past three years has returned 28% annualized. That’s the same annual rate of return it generated in the three years ending Dec. 31, 1998. It went on to surge 85% in 1999 and another 20% three months into 2000. Then a 30-month, 75% crash. A buyer of the Nasdaq on the last day of 1998 was still underwater six years later, similar to anyone who bought ARK Innovation ETF (ARKK) between October 2020 and the end of 2021. Once again, the big-cap indexes appear more controlled and muted than the action under the hood. We’ve recently seen an extreme high in the volatility of individual stocks relative to the indexes, a regime of “dispersion” that is supportive while it lasts, but can’t last forever. Speculative sub-themes continue to rip, with drones and EV helicopters and small nuclear reactors and the BUZZ meme-stock ETF and crypto proxies lathered up again. Call-option volumes are setting new records almost daily. (Although, perhaps notably, Robinhood shares are down small in a strong tech tape after a 300% year-to-date run, worth monitoring). No news on official government data remains fine with the markets, which are free to key off corporate-AI enthusiasm and the global markets’ apparent message of a reflation/reacceleration dynamic taking hold. Copper in a strong rally, global yields higher . The announced acquisition of Comerica by Fifth Third Bancorp was a welcome confirmation that some consolidation in this tier of the banking industry is starting. But the terms would seem to place a ceiling on the prices and valuations of such institutions well below their pre-SVB Financial-failure levels from early 2022.
Santoli’s Monday wrap-up: Bull markets peak on good news, but not today
Oct 6, 2025
