B.Riley cuts ON Semiconductor stock price target on revised outlook

B.Riley lowered its ON Semiconductor price target to $97 from $105 but maintained a Buy rating, implying approximately 39% upside. Q3 fiscal 2026 results slightly beat consensus, including adjusted EPS of $0.90 versus $0.87 and revenue of $1.72 billion versus $1.70 billion; AI data-center sales grew at twice B.Riley’s forecast, and the company raised its 2026 AI data-center sales target by more than 20% to $620 million. Onsemi also announced a $600 million Ferric voltage-regulator deal, including $175 million in earn-outs, and lifted its 2030 AI data-center segment target by more than $500 million to $3 billion, while flagging a softer near-term automotive revenue outlook.

Analysis

The investable signal is a potential change in ON Semiconductor’s earnings mix, not the headline growth target: data-center power demand could become a meaningful offset to weaker automotive demand and improve the market’s view of its growth durability. But the path from design wins and announced deal value to recognized revenue, attractive margins, and cash generation is the key underwriting gap. GPU/XPU power architectures may expand the addressable market while also inviting competition from incumbent power-management suppliers and customer second-sourcing, which could limit pricing and concentrate execution risk. The referenced transaction economics also warrant scrutiny: verify closing conditions, earn-out triggers, integration costs, and whether management’s targets include acquired revenue. Near term, automotive softness can still weigh on results before data-center ramps scale; over 6–18 months, successful conversion could support a higher-quality multiple, while slippage would expose the stock to a growth-expectations reset. The analyst’s lower target despite higher estimates is a useful reminder that bullish forecasts are not equivalent to a near-term catalyst. The company’s targets and broker valuation are claims to test against reported segment revenue, margins, and cash conversion—not independently verified outcomes.

Key Decisions for Investors

• Consider a measured long ON / short SOXX relative-value position rather than an outright high-conviction bet: it expresses company-specific data-center execution while hedging broad semiconductor beta. Build only after reported data-center revenue and margins confirm conversion; risk is that automotive weakness or deal dilution overwhelms the new growth stream.
• Over the next 1–3 months, track the next guidance update for data-center revenue versus the company’s stated ramp, automotive revenue trends, and any disclosure on the Ferric transaction’s closing, earn-outs, and integration economics. Do not capitalize announced deal value as recurring sales before those details are verified.
• Falsify the long thesis if data-center targets are cut or delayed, segment growth fails to translate into improving earnings/cash generation, or automotive weakness broadens enough to offset the ramp. Conversely, sustained data-center outperformance with margin evidence would justify adding exposure; absent that proof, the reported targets alone do not warrant chasing a sharp price move.