ON (ON) Q1 2026 Earnings Call Transcript

ON Semiconductor reported Q1 revenue of $1.51 billion, up 5% year over year and above guidance, while non-GAAP EPS of $0.64 also beat the midpoint. Gross margin expanded to 38.5% for a third straight quarter, AI data center revenue jumped more than 30% sequentially and is expected to double in 2026, and automotive revenue returned to year-over-year growth after seven quarters of declines. Management guided Q2 revenue to $1.535 billion-$1.635 billion with EPS of $0.65-$0.77 and said gross margin should continue to expand, supported by mix improvement, utilization gains and ongoing share repurchases.

Analysis

ON is transitioning from a cyclical recovery story into a mix-shift story, and that matters more than the headline beat. The market is likely still underestimating how much AI power and zonal content can offset the lingering drag from legacy exits: management effectively telegraphed that the 2026 revenue bridge is no longer dependent on a clean macro rebound, but on higher-margin sockets that should compound into 2027-2028. That is a better setup than a pure semiconductor beta trade because the incremental dollars are landing in businesses with structurally higher gross margins and tighter customer design-in lock-in.

The underappreciated second-order effect is pricing power. Lead times are extending, some technologies are already allocation-constrained, and management is now passing through input-cost inflation; that combination usually shows up with a lag, so the visible gross-margin step-up may be more pronounced in 2H26 than the street models. If utilization keeps drifting toward the low-80s while exits roll off, ON has multiple operating levers moving in the same direction; the asymmetry is that margin upside can arrive faster than consensus expects, while the downside from mix/price is delayed.

The most important contrarian angle is that the AI opportunity is still being modeled too narrowly as a small revenue line, when in reality it is a portfolio-wide attach rate story across the rack-to-grid stack. That creates a broader competitive moat than point-product investors may appreciate, and it also means ON can win even if it is not the primary silicon supplier in a given AI program. The real risk is not demand, but execution: if hyperscaler or EV ramps slip, the stock could de-rate on disappointment because the market is already paying for a recovery plus secular growth mix.

Net: this is bullish, but the better trade may be to own ON as a levered beneficiary of a 2H26 margin inflection rather than chase the move after the call. If the company delivers another quarter of sequential margin expansion and exits stay on schedule, the multiple should expand; if AI growth merely normalizes from very strong growth, the stock can still work because the incremental margin from Treo/GaN/SiC is the real earnings engine.

Key Decisions for Investors

• Long ON into 2H26 results with a 3-6 month horizon; thesis is that gross margin inflection will outpace revenue revisions as utilization, pricing, and mix all move higher. Risk/reward favors owning strength on pullbacks rather than breakout chasing.
• Pair trade: long ON / short a basket of slower-margin-recovery analog peers (e.g., TXN, ADI) over the next 1-2 quarters. ON should get credit for secular mix shift and AI attach, while the short leg is more exposed to a slower cyclical rebound.
• Use call spreads rather than stock if entering after a post-earnings move: buy ON 6-12 month call spreads to capture a multiple re-rating from a second-half margin step-up while limiting downside if AI growth merely normalizes.
• For event-driven traders, sell puts on ON only on weakness toward support after the next guidance update; the strongest fundamental catalyst is not the quarter itself but evidence that 2H pricing is flowing through to margins.
• Watch NIO and other China EV proxies as secondary confirmation, but do not pair ON with them; ON’s thesis is more about content gain and power architecture than broad EV unit growth, so the cleaner hedge is against analog duration risk, not auto beta.