ON Semiconductor Targets AI Power Growth Amid End-Market Shifts

What happened

ON Semiconductor is publicly repositioning toward integrated AI power solutions as its traditional automotive and industrial end markets remain weak. Management projects strong revenue and margin expansion through 2030, framing a $213 billion addressable market, but the article acknowledges investor skepticism has not dissipated. Our analysis shows ON's AI power revenue is still negligible, with the company relying heavily on a cyclical recovery in auto/industrial and ambitious SiC ramps. The stock around $60 prices in a near-perfect recovery, trading at ~77x trailing earnings and ~10x EV/EBITDA, leaving little room for error. This announcement appears more promotional than substantive, as the master report has already flagged the AI opportunity as a long-dated tailwind with no near-term revenue impact.

Implication

Investors should treat the AI power pivot as optionality, not a near-term driver, given ON's tiny revenue base in this segment and intense competition from established players. The stock's valuation already embeds a successful shift, leaving limited upside from announcement alone; any execution misstep could trigger a sharp de-rating. Monitor concrete indicators such as AI power design wins, revenue contributions from data center power solutions, and segment margin improvements over the next several quarters. The company's aggressive $6 billion buyback plan and restructuring actions may support the stock, but they also raise the risk of financial engineering if free cash flow disappoints. We would only turn constructive if gross margins sustain above 42% and auto/industrial revenue shows high-single-digit growth, as outlined in our re-assessment window.

Thesis delta

The AI power angle adds a new growth narrative but does not change our fundamental thesis. The master report already identifies AI power as a tailwind, but its contribution to revenue is currently negligible, and the company's valuation already embeds significant success. We remain on the sidelines until concrete evidence of AI power revenue and margin expansion emerges, while continuing to monitor existing risks in auto/industrial and SiC execution.