STM vs. ON: Which EV Semiconductor Stock Is More Compelling?

STMicroelectronics (STM) highlighted with stronger EV-semiconductor momentum: Q2 automotive revenue +16% YoY and total Q2 revenue +26.1% YoY to $3.49B; gross margin improved 130 bps to 34.8% and non-GAAP EPS was 31 cents vs a loss a year earlier. STM’s SiC revenues are guided to grow at a double-digit rate in 2026, while ON Semiconductor (ON) showed improving China EV traction (automotive revenues +13% in 1H 2026) and Q2 free cash flow of $425M plus $332M in buybacks, but noted near-term constraints from prioritizing AI data-center demand. Zacks consensus calls for STM 2026 revenue +21.8% and EPS +145.3% to $1.30, versus ON revenue +9.5% and EPS +37.5%, supporting STM as the comparatively more compelling EV semiconductor exposure.

Analysis

The key competitive dynamic is that auto-semiconductor growth is increasingly about content per vehicle, not unit growth. That favors suppliers with broader portfolios and sticky design wins, but it also means capacity allocation matters: when a name like ON prioritizes higher-return AI demand, the near-term loser is not just auto revenue but customer trust, which can quietly open share for NXP, Infineon, and even STM on follow-on platforms.

STM’s setup is better structurally, but the market may be underpricing execution risk from its manufacturing transition. SiC and analog node changes can create a 1-2 quarter earnings air pocket where backlog visibility looks strong while gross margin and free cash flow lag; that is exactly when consensus tends to extrapolate too much into the out-year model. The bigger second-order effect is that temporary supply friction in auto semis often accelerates dual-sourcing, which can make post-transition share recovery slower than management guidance implies.

The contrarian view is that ON may be the cleaner risk/reward despite the article’s preference for STM. ON already has higher margin and cash return power, and its AI exposure is a margin tailwind, not a distraction, so the market could eventually pay up for a more balanced growth profile if auto re-accelerates once supply is normalized. Over 1-3 months the catalyst is earnings/guidance revision; over 6-18 months the test is whether STM’s transition converts into sustained margin expansion or just deferred execution costs. A miss on STM gross margin or a continued auto under-allocation at ON would falsify the current thesis quickly.

Key Decisions for Investors

• Initiate a relative-value pair: long ON / short STM, sized market-neutral, on any strength in STM into the next earnings window. Thesis: ON’s margin quality and capital returns should outperform while STM’s transition costs and supply disruptions keep estimates vulnerable; target 10-15% relative outperformance over 3-6 months.
• For outright exposure to auto semis, prefer ON on pullbacks rather than chasing STM after its strong run. Risk/reward is better if ON can hold gross margin near the 40% area while auto recovers; stop if auto revenue reacceleration fails to show up by the next quarter.
• Use SOXX as a hedge if taking a directional long in ON or STM. The sector beta is still tied to broader semiconductor de-risking, so a market-neutral construction is preferable until the next print confirms whether the auto recovery is real or just backlog conversion.
• Watchlist trigger: if STM reports another quarter of margin pressure from manufacturing reshaping without a clear inflection in SiC conversion, fade the stock on rallies. That would support a tactical short or an options overlay, with a falsifier being gross margin expansion ahead of schedule.