← Back
MarketsMarketWatchJul 23, 2026· 1 min read

JPMorgan Flags AI Stock Discrepancy, Citing 1990s Tech Parallel

JPMorgan warns that the current split in AI stock performance between hyperscalers and chip/infrastructure firms mirrors late 1990s tech market dynamics. The bank suggests the next few weeks are critical for these market segments.

JPMorgan strategists are drawing parallels between the current divergence in artificial intelligence (AI) stock performance and market dynamics observed in the late 1990s. The bank highlights a notable split in price action between "hyperscalers" – large-cap tech companies leading AI development and deployment – and firms primarily focused on AI-enabling chip and infrastructure technologies. Historically, such discrepancies can signal differing market expectations regarding the immediate and long-term beneficiaries of technological shifts. In the 1990s, the dot-com boom saw varying levels of success and eventual consolidation across different segments of the internet industry. JPMorgan's analysis suggests that the market is currently assigning distinct valuations and growth trajectories to the various components of the AI ecosystem. While hyperscalers, often characterized by their extensive data centers, proprietary AI models, and large customer bases, have seen robust performance, the underlying hardware and infrastructure providers have experienced a more nuanced trajectory. This divergence could reflect investor focus on companies with established revenue streams and direct control over AI application, versus those with more indirect or commoditized exposure to the technology's growth. The investment bank is cautioning that the coming weeks could be pivotal for these market segments. This period may reveal whether the current valuations are sustainable, if a convergence or further divergence will occur, and which parts of the AI value chain are poised for continued outperformance or a potential re-evaluation by investors. Monitoring these trends will be critical for understanding capital allocation within the rapidly evolving AI landscape.

Analyst's Take

While the headline points to a valuation split, the real insight lies in the potential for capital rotation. Should hyperscaler growth moderate or face increased regulatory scrutiny, a delayed but significant re-rating could occur in the underlying infrastructure, particularly specialized compute and data center REITs, as their long-term essentiality becomes undeniable, likely by late Q3 as corporate AI adoption ramps up beyond initial pilots.

Related

Source: MarketWatch