Could the latest pullback in AI stocks forms another opening for the sector? J.P. Morgan analysts think so, pointing to lower valuations, stronger investor positioning, persist AI capital spending, and improving evidence that companies can turn AI investments into revenue. The firm sees precise strength in semiconductor stocks, even because it stays more cautious about software.
AI Spending Remains a Key Driver
AI-linked stocks began 2026 firmly before concerns about infrastructure spending and near-time period returns put pressure on the wider trade. Earlier this month, the sector faced another selloff after leading AI executives raised issues about the risks related to rapid AI development.
J.P. Morgan claims that the correction has changed the setup. Investor positioning has become less crowded, valuations have dropped across a much of the AI market, and capital expenditures are anticipated to stay substantial.
The bank also anticipates earnings strength and clearer signs of AI monetization to support renewed interest.
That spending outlook matters beyond financial markets. Persist investment in data centers, accelerators, networking, memory, and other infrastructure determines how quickly organizations can amplify training and inference workloads.
Semiconductors Continue to Outpace Software
J.P. Morgan stays positive on semiconductors, citing healthy industry fundamentals, pricing growth broadening into 2027, and tight supply-demand conditions that could persist until 2028.
Market performance already displays that divide. The MSCI World Semiconductors and Semiconductor Equipment Index has gained approximately 48% in 2026, even as the MSCI World Software and Services Index has risen only 1.3%.
The bank is more cautious about software as generative and agentic AI are strenghtening competition and creating uncertainty around established business models. Instead of treating software as a straightforward bearish bet, but, J.P. Morgan sees the relative strength of semiconductors compared with software as the more notable trend.
That distinction emphasize a wider shift across the AI economy. As enterprises experiment with new models and applications, the underlying compute layer persist to benefit from growing demand regardless of individual AI products ultimately gain market share.
