JPMorgan: AI Memory Stock Correction Driven by Sentiment, Not Fundamentals; Q2 Earnings to Validate Super Cycle

According to JPMorgan Chase's June 24 research report, the recent volatility in AI memory stocks reflects market sentiment and trading dynamics rather than deteriorating fundamentals, with the true test coming in second-quarter earnings. The bank maintains its view that the AI memory super cycle remains intact.

JPMorgan identified three key validation metrics: long-term supply agreements (LTA) progress, cloud service providers' capital expenditure (CAPEX) trends, and HBM supply constraints. The bank noted that CSP CAPEX allocation to AI memory is projected to rise from under 20% in 2022 to 52% in 2026, indicating memory is becoming a critical bottleneck for AI infrastructure expansion. JPMorgan expects major LTA deals to accelerate in H2 2026, driven by hyperscalers and cloud giants securing multi-year memory supplies.