Mizuho Securities has lowered its price target on Intel from $109 to $92 while maintaining a Neutral rating, with analyst Vijay Rakesh acknowledging that the chipmaker's AI narrative is compelling but arguing the near-term math does not support a higher valuation.

The reduction stands out given the trajectory of Mizuho's earlier revisions. During the first half of 2026, Rakesh had repeatedly raised his Intel target, lifting it to $128 in June and again to $135 in July. The move down to $92 marks a pullback of roughly 32% from that July peak.

The bull case Rakesh is not ignoring

Rakesh identified four factors that could support the stock over time. First, demand for server CPUs driven by agentic AI is expected to accelerate through 2027. Second, CPU supply constraints are likely to persist through 2027, giving pricing power to manufacturers able to deliver. Third, Intel's advanced-packaging revenue, centered on its EMIB-T technology, could reach $3.5 billion by 2029. Fourth, PC refresh cycles have proven steadier than feared, providing a floor under Intel's client computing segment.

Intel's most recent results lend some support to this outlook. In Q2 2026, Data Center and AI segment revenue rose 24% quarter-over-quarter, driven primarily by growing server CPU demand.

Why Rakesh cut anyway

In Rakesh's view, favorable catalysts in 2027 and 2029 do not address current conditions. Margin pressure in Intel's foundry business remains a drag, and continued softness in PC demand weighs on the top line, contributing to what Rakesh describes as valuation compression across AI-related equities more broadly. That compression reflects a wider pullback in how investors value semiconductor and AI-infrastructure stocks, where analysts have grown more discriminating between companies with near-term earnings power and those whose AI payoff sits further out on the horizon.

His $92 target sits well below the Wall Street consensus of approximately $116.

Intel's competitive position in the AI arms race

Intel positions itself as a key supplier of CPUs for AI workloads while simultaneously building out its foundry services. The company is investing in advanced manufacturing nodes including 18A and 14A, along with packaging technologies such as EMIB-T, which allows Intel to integrate multiple chiplets into a single package. The $3.5 billion revenue projection for 2029 would make advanced packaging a meaningful contributor to Intel's top line. On the foundry side, the margin headwinds Rakesh flagged are a direct consequence of this strategy's early stages—a challenge amplified by the scale of established foundry rivals such as TSMC, which dominates the contract chipmaking market Intel is trying to enter.

What investors should watch

The gap between Mizuho's $92 target and the consensus estimate of $116 creates a notable divergence. Mizuho's repeated revisions throughout 2026—up to $135 in July, then down to $92 in September—underscore how volatile the underlying assumptions are. Key checkpoints ahead include Intel's foundry margin trajectory as its advanced nodes ramp, server CPU shipment trends as agentic AI demand builds, and whether advanced packaging revenue stays on the path toward the 2029 projection. The next few quarters of earnings and forward guidance will likely determine whether the consensus moves toward Mizuho's more conservative view or holds closer to $116.