The second half of 2026 opened with a notable shift in market leadership, as AI, memory, and semiconductor trades, which previously drove market gains, experienced sharp declines. Momentum fell over 18% in two days, marking its worst two-day decline since November 2020. A basket of AI beneficiaries-versus-AI-at-risk dropped approximately 16%, and memory stocks plunged over 18% in their sharpest two-day decline in at least 12 years. AI semiconductors also saw their worst two-day move since Liberation Day. This rotation was not a market collapse, but rather a redistribution of risk away from crowded AI trades, with financials and consumer discretionary sectors outperforming, and the Dow holding firm.
Contributing to this market shift was weaker-than-expected jobs data. June payrolls increased by only 57,000, significantly below expectations, while prior months' figures were revised downwards. Although the unemployment rate fell to 4.2%, the labor force shrank by about 720,000, with a substantial portion (700,000) from workers aged 25 to 34. This softening labor market, coupled with a softer tone on inflation risks, led traders to trim expectations for further Federal Reserve rate hikes, causing two-year Treasury yields to fall around 4 basis points to 4.14%. Gold found support, and Bitcoin posted its strongest two-day run since February.
The decline in AI-related stocks reflected concerns about the sustainability of the AI build-out and whether valuations had become too stretched. Companies like SanDisk, Kioxia, Micron, SK Hynix, and Samsung saw significant drops. For instance, SanDisk fell roughly 22% from its recent high. Reports that Meta might sell excess AI compute capacity further fueled questions about whether supply could catch up with demand. While AI spending has not disappeared and hyperscaler spending remains strong, investors are now seeking delivery and tangible results rather than relying solely on narrative, heading into an earnings season with higher scrutiny and less room for disappointment.
