Stock Markets July 6, 2026 05:55 AM

BofA Calls Chip-Sector Drop a Short-Term Reset Ahead of Renewed Rally

Bank of America says recent semiconductor weakness reflects seasonal adjustment, while cloud and AI capex should drive a multi-year recovery

By Avery Klein
Share
Twitter Reddit Facebook LinkedIn
NVDA TXN ADI CDNS SNPS

Bank of America views the semiconductor sector's recent pullback as a temporary correction rather than a structural slowdown in AI-driven demand. While the SOX index slid 11% in the third quarter after an 88% jump in the second quarter, BofA expects lower-beta names to lead near-term performance and forecasts robust cloud and AI infrastructure spending that should revive memory, compute and related supply chains through 2026 into 2027.

BofA Calls Chip-Sector Drop a Short-Term Reset Ahead of Renewed Rally
NVDA TXN ADI CDNS SNPS
Summarize with
ChatGPT Perplexity Claude Grok Gemini

Key Points

  • BofA views the recent semiconductor decline as a short-term adjustment rather than a structural drop in AI demand; the SOX index fell 11% in Q3 after an 88% gain in Q2.
  • The bank expects lower-beta semiconductor stocks - including NVDA, TXN, ADI, CDNS and SNPS - to outperform in the near term.
  • BofA projects global cloud and AI infrastructure capex will reach $1.5 trillion by 2027 (a 40% to 50% year-over-year increase), supporting renewed demand for memory, compute, semiconductor capital equipment, optics and networking.

Bank of America characterizes the semiconductor industry's recent price weakness as a transitory reset instead of a fundamental retreat in artificial intelligence demand. The brokerage noted that the Philadelphia Semiconductor Index (SOX) fell 11% in the third quarter following an 88% gain in the second quarter.

According to BofA, the current pullback is consistent with historically weak seasonal patterns for the sector. The firm recommends investors look to lower-beta semiconductor names to lead the market recovery in the near term - specifically naming NVDA, TXN, ADI, CDNS and SNPS as likely front-runners.

The bank projects a large expansion in global cloud and AI infrastructure capital expenditures, estimating spending will reach $1.5 trillion by 2027. That figure, BofA says, equates to a 40% to 50% year-over-year increase and is expected to be propelled by token expansion, broader agent adoption and infrastructure capacity constraints.

BofA also reports that hyperscale cloud operators continue to emphasize utilization and growth objectives over strategies that would prioritize depreciation-based optimization. This emphasis on scale and usage, the bank argues, supports the case for renewed demand across memory, compute, semiconductor capital equipment, optics and networking stocks as visibility into 2027 cloud budgets improves through the second half of 2026.

The emergence of Chinese open-weight large models from companies such as GLM, Kimi, DeepSeek and Qwen is another development highlighted by BofA. The bank notes these models have narrowed performance differences with U.S. frontier labs while providing lower inference costs. BofA views that dynamic as constructive for adoption; lower-cost intelligence should broaden usage and in turn raise demand for compute, memory, networking and power infrastructure.

Memory spending now represents a larger slice of cloud AI capital expenditures than in prior cycles, accounting for roughly 35% to 40% of such outlays, according to BofA. Despite the larger role memory plays in cloud AI capex, memory equities trade at about 10 times forward price-to-earnings ratios, the bank notes.

On individual coverage, BofA maintains a buy recommendation on Micron Technology (MU) with a price target of $155.

Overall, BofA frames the recent sector decline as a healthy reset that could clear the path for a resumed rally once spending visibility and supply-side dynamics align with ongoing growth in AI-driven workloads.


Summary

BofA sees the semiconductor pullback as temporary, points to seasonal weakness, expects lower-beta names to lead, forecasts $1.5 trillion in cloud and AI infrastructure capex by 2027, and highlights rising memory share of AI capex and positive implications from cheaper Chinese open-weight models.

Risks

  • Seasonality and short-term market volatility could continue to pressure semiconductor equities, impacting investor returns in the near term - relevant to the semiconductor sector and equity markets.
  • Visibility into 2027 cloud spending must materially improve through the second half of 2026 for BofA's anticipated momentum in memory, compute and related equipment stocks to materialize - this affects cloud infrastructure suppliers and capital equipment vendors.
  • If lower inference costs from emerging open-weight models do not translate into increased adoption, demand projections for compute, memory and networking infrastructure could be weaker than expected - a risk for cloud service providers and hardware suppliers.

More from Stock Markets

ZIM Shares Slide After Israeli Officials Voice Opposition to Hapag-Lloyd Deal Jul 6, 2026 Thales and Leonardo Win NATO Contract for Special Forces Communications, Shares Rise Jul 6, 2026 Seer Shares Jump 30% After Investor Presentation Highlights Growth and Cost Cuts Jul 6, 2026 Strategy shares rise as firm unveils plan to monetize Bitcoin holdings and cut funding costs Jul 6, 2026 Compass Pathways Shares Gain as Pivotal COMP006 Readout Window Arrives Jul 6, 2026