Memory Shortage, $1T Capex Forecast Fail to Halt AI Sell-Off
Shares of leading AI hardware firms Nvidia and Micron have slipped in July despite projections that hyperscaler capital expenditures will surge to $1 trillion by 2027 and a persistent memory chip shortage.
Shares of Nvidia and Micron have declined in July as investors second-guess the longevity of artificial intelligence demand. The pullback hits two of the year's strongest performers, even as both companies signal that order books and supply constraints point to sustained growth.
Micron has seen its stock retreat over recent weeks as the market fixates on the possibility that AI demand could plateau earlier than expected. This skepticism overlooks the current supply dynamics in the memory chip market. A severe shortage has driven prices higher, a trend that Wall Street analysts expect will fuel 80% earnings growth for Micron in its fiscal 2027, which begins in September.
Company management has explicitly told investors that this undersupply will persist well beyond 2027. Even with new production capacity coming online, it will take time for Micron and its competitors to catch up with demand. The broader AI infrastructure build-out is projected to keep accelerating through 2030, suggesting the supply deficit will not resolve quickly.
Nvidia faces a similar disconnect between its forward guidance and its current stock price. As the dominant supplier of AI computing GPUs, Nvidia integrates Micron's memory into its hardware.
The company recently outlined its expectation that hyperscaler data center capital expenditures will climb to $1 trillion in 2027, rising from $650 billion in 2026. Nvidia indicated that it likely already has most of its 2027 product orders booked within its system. Hyperscalers are securing computing chips in advance to ensure availability once their broader data center infrastructure is finished.
This gives Nvidia direct insight into the sector's trajectory, yet the broader market has not fully priced in this visibility. The divergence between July's stock price action and corporate fundamentals highlights a shift in market psychology.
Investors are actively discounting the long-term projections of the very companies building the AI infrastructure. For market participants, this creates a tangible gap between current valuations and the medium-term earnings power indicated by a $1 trillion capital expenditure cycle.
