Mizuho raises ON Semiconductor stock price target on AI growth

Mizuho raised its price target on ON Semiconductor to $120 from $70 while keeping an Outperform rating, implying about 18% upside from the current $102.04 share price. The firm highlighted June-quarter guidance of $1.59 billion revenue, 5% sequential growth, improving gross margins, and AI data center revenue expected to exceed $500 million in calendar 2026, more than doubling year over year. First-quarter 2026 results also beat estimates, with EPS of $0.64 vs. $0.61 expected and revenue of $1.51 billion vs. $1.49 billion consensus.

Analysis

The market is starting to price ON less like a cyclical analog name and more like a structurally improving AI/EV power play. The key second-order effect is that higher confidence in mix, utilization, and margin expansion will force the street to lift terminal earnings assumptions, not just near-term estimates — that matters because the stock is already near prior highs, so upside now depends on multiple persistence rather than simple earnings beats.

The obvious winners are suppliers that benefit from a higher-quality semiconductor capex cycle and from ON’s improving foundry/fab utilization; the less obvious loser is any adjacent power/discrete supplier competing for AI data center sockets if ON’s Treo platform proves sticky. In autos, the market is still underappreciating that flat unit demand does not preclude EPS acceleration if China EV content and margin mix keep improving; this creates a divergence between headline automotive names and content-rich suppliers.

The main risk is timing: the next quarter is likely less about upside surprise and more about whether management can sustain guidance into the September analyst day. If utilization stalls or AI revenue ramps slower than the 2026 framing, the stock can de-rate quickly because expectations have shifted from recovery to execution. A pullback would likely come from multiple compression, not earnings collapse, which makes options and relative-value expressions cleaner than outright directional longs.

Contrarian view: consensus may be too focused on the target-price upgrade and not enough on how much of the 2027 story is already being discounted in a stock near highs. If AI revenue becomes a smaller share than investors expect, or if gross-margin expansion proves back-end loaded, the market could punish the name despite continued fundamental progress. That makes this more attractive as a buy-the-dip or pair trade than as a chase-the-breakout long.

Key Decisions for Investors

• Stay long ON but use a staged entry on any 5-8% post-earnings or pre-analyst-day pullback; risk/reward is better there than buying highs because the upside is now dependent on multiple expansion continuing through 2H26.
• Pair trade: long ON / short a weaker automotive semiconductor or general analog peer with less AI content exposure over the next 3-6 months; ON should outperform if investors keep paying for margin mix and AI optionality.
• Buy ON Sep/Dec call spreads instead of outright stock if chasing momentum; capped downside makes sense because the near-term catalyst path is positive but already well-owned.
• Monitor for a relative short if ON trades above the new target ahead of analyst day without evidence of a higher 2027 growth bridge; at that point the asymmetry shifts toward disappointment risk.
• For a thematic basket, overweight ON versus broad semis until the market re-rates AI edge exposure in industrial/auto power chips; this is a cleaner way to express the thesis than owning pure-play AI names at richer multiples.