ON Semiconductor

ON Semiconductor hosted its 2026 analyst day, reiterating its growth strategy across automotive, industrial, and data center power chips and updating its revenue growth target to 12%-14%, excluding the pending Synaptics acquisition.

Why it matters: We think Onsemi can benefit from increased power requirements across its end markets. AI data centers are a promising new opportunity, and we see power efficiency as a critical enabler of model performance. Management unveiled the Embedded Power Platform, or EPP, which is wafer-level packaging for higher power density and smaller size. This would combine multiple chips in one package, and we see it as the latest extension of the firm's strong power chip design. Onsemi officially called for doubling AI revenue to over $1 billion in 2027 (after more than doubling in 2026) and a $2.5 billion AI revenue target in 2030. This broadly agrees with our model, and we believe there's potential upside to this growth through 2030.

The bottom line: We maintain our $100 fair value estimate for narrow-moat Onsemi. We remain slightly below management's long-term margin goals, but in line with the low end of its growth target. Shares sank 9% intraday during the event and remain undervalued to us. Onsemi shares now sit 50% below their peak in late June 2026, with the market initially selling off on the Synaptics deal and now seeming to sour on Onsemi's competitive position in AI. We saw Synaptics as value-neutral and expect a solid AI position despite a crowded field of competitors. If management hits its targets, we believe there's meaningful upside to shares today. We want to see more progress on gross margin to gain confidence in the 53% target, and we believe the market feels similarly after the firm fell below management's "floor" margin in 2025.

Specifically, management is calling for 12% to 14% growth, with automotive growing 9%, industrial growing 10%, and AI data center growing more than 50%. This is above management's prior 10% to 12% target from the 2023 investor day, which makes sense given the high growth now coming from AI. Non-GAAP gross margin's target of 53% was steady from prior guidance, but management actually lowered its non-GAAP operating margin target to 38%, from 40%, to reflect higher organic investments.

Our model comes in at 12% growth through 2030, with non-GAAP gross margin hitting 50% in 2030. This would still represent a record gross margin for the firm, and meaningful expansion from recent levels around 40%. We project 35% non-GAAP operating margin in 2030, with the gap to management's guidance coming entirely at the gross margin line. We aren't too concerned with the cut to operating margin guidance, given we already held healthy skepticism to the prior target. In short, we're skeptical of the level of consistent content growth in automotive and industrial, as we are about the firm's ability to increase like-for-like margins via new products at higher prices. Still, we believe our model implies a strong business, and certainly upside to what market prices are currently implying.

The EPP was the main news of the day, and likens to peer Monolithic Power's integrated architectures, effectively being able to put multiple disparate chips into one package. We like this approach, but are skeptical as to how much this will truly raise the growth profile of the core automotive and industrial business, rather than act as a defense against competitors pursuing similar strategies.

Our fair value estimate is $100 per share, which implies a 2026 adjusted price/earnings ratio of 30 times and an enterprise value/sales ratio of 6 times. The greatest drivers to our valuation are core automotive and industrial growth alongside burgeoning AI growth.

We expect 12% compound annual sales growth for Onsemi through 2030, excluding the pending acquisition of Synaptics. We anticipate the start of a cyclical rebound in 2026 for automotive and industrial chips after two down years, but expect a more meaningful inflection until 2027. Over these next two years, we expect Onsemi to outperform the growth of its underlying end markets as customers return to ordering for true demand. We also expect Onsemi’s budding AI data center business to supplement its core automotive and industrial chip growth as demand skyrockets for data center power.

We think Onsemi will benefit from stronger secular demand in the longer term as it increasingly aligns its portfolio to attractive secular trends like electric vehicles, renewable energy, and AI infrastructure. We forecast 6% growth for Onsemi’s automotive revenues. In automotive, Onsemi’s chips are taking share in high-voltage electric vehicles, while its image sensors win designs in autonomous driving and safety applications. In industrial, we expect 5% growth through 2030 as the firm’s high-power chips are winning designs in renewable energy infrastructure.

We expect Onsemi’s AI data center exposure to quickly rise from roughly $250 million in 2025 to above $2 billion in 2030. Management has guided to more than $500 million in AI revenues in 2026, and we expect this figure to double again past $1 billion in 2027 as data centers adopt 800-volt rack architectures. We still expect Onsemi to remain concentrated in automotive and industrial markets, but believe data centers present a new, promising growth opportunity on top.

We forecast meaningful gross margin expansion over the next three years, with Onsemi benefiting from a cyclical rebound and a streamlined manufacturing footprint. We model non-GAAP gross margin reaching 50% in 2030, compared with a cyclically low 38% in 2025. Commensurately, we anticipate higher operating margins. We forecast Onsemi to continue investing heavily in research and development but foresee some leverage on the other operating expense lines. Including gross margin expansion, we think non-GAAP operating margin can ascend to the mid-30s by 2030. We don’t model Onsemi to hit its ambitious targets of 53% non-GAAP gross margin and 40% non-GAAP operating margin in the next five years. If it does, there would be upside to our valuation. These targets appear optimistic to us given the need for continued investment against competitors, especially in the data center.

Onsemi remains vulnerable to cyclicality. We are bullish about its pursuit of differentiated socket share in automotive and industrial applications, which we see as stickier than some of its legacy products. Still, these markets are prone to supply and demand dynamics for semiconductors; we point to the volatility in results between 2021 and 2025 as a prime example.

We assign a narrow Morningstar Economic Moat Rating to Onsemi, driven by intangible assets and switching costs in its power chip and sensor portfolios. We believe these businesses will, more likely than not, combine to earn excess returns on invested capital over the next 10 years.

We believe Onsemi holds moaty intangible assets across its power, analog, and image sensor chip portfolios. In power chips like MOSFETs and IGBTs, we observe that Onsemi has pivoted its portfolio toward higher voltages, where it can extract more differentiation, and away from more commoditized, lower-voltage transistors. This has resulted in higher profit margins across cyclical peaks, troughs, and midcycle than historically. We believe Onsemi and Infineon are the two differentiated transistor vendors operating at the top end of the market, with a long tail of smaller, less differentiated vendors serving the majority of the market. To us, differentiation in power chips comes via better performance (power efficiency) and reliability. This differentiation results in higher market share and higher profit margins versus downmarket competitors.

Alongside MOSFETs and IGBTs, we like Onsemi’s power management chip portfolio, which offers more differentiation and focuses on voltage regulation into compute chips and other subsystems. Similarly, this portfolio focuses on high voltages for higher differentiation and includes difficult-to-replicate analog chip content. It is our view that analog chip talent is scarce, and hiring qualified analog engineers is extremely difficult, as it takes years to train a new analog engineer on the intricacies of a firm’s specific designs. Thus, a strong bench of analog talent that results in strong design prowess and market share is incredibly difficult to replicate. Onsemi is one of a handful of firms at the top of the power management market that possess this talent.

In image sensors, we see Onsemi as a leader in automotive applications, resulting from differentiated intangible assets. We observe a clear top-three market-share leaders in automotive image sensing, with Sony and Omnivision as peers. Automotive image sensors sell into lidar, radar, ADAS, and other autonomous driving applications—they are the eyes of safety systems in a vehicle. Onsemi has built a strong foundation with sensors that use fewer, larger pixels to perform exceptionally well in low-light environments and amid optical interference. These characteristics make them ideal for ADAS applications where there isn’t room for failure in any environment. We think Onsemi has a first-mover advantage in the automotive market in particular, having developed technology specifically designed for ADAS applications for years rather than focusing on the cutthroat smartphone market and then transitioning to automotive, as Sony and Omnivision have. Onsemi’s prowess in low-light and high-interference image sensor design has resulted in it holding one of the largest shares in the automotive market, particularly in Western OEMs.

Onsemi also benefits from high customer switching costs, resulting from a concentration in the automotive and industrial markets. In our view, these markets offer risk-averse OEMs mission-critical applications and high costs to failure. Automotive applications make up the majority of Onsemi’s revenue, with it selling predominantly into drivetrain and safety systems: EV traction inverters, onboard charging, ADAS and autonomous driving, and digital cockpits. There is no room for failure in these systems—product dysfunction could result in human injury or death. As such, we see extremely high standards from customers, and once a supplier has proved 100% reliability, it becomes difficult to switch away. Additionally, once Onsemi has won a socket in a car, the model lifecycle typically spans five to seven years, creating a secure stream of revenue. Mid-lifecycle supplier changes are extremely rare and only occur due to product failure. Finally, we observe an incumbency advantage for Onsemi across model lifecycles, resulting from deep engineer-to-engineer partnerships with its customers and advanced knowledge of future model design needs. We observe similar switching costs in industrial applications, with markets like aerospace and defense offering even higher mission-criticality and longer product lifecycles spanning multiple decades. Together, we see more than 80% of Onsemi’s revenue coming from automotive and industrial applications that exhibit these strong switching costs.

To us, we believe intangible assets in power chip and image sensor design enable Onsemi to win new business and build new customer relationships, while switching costs make it sticky in designs it’s already won. Combined, these lead to strong returns on invested capital over the course of a semiconductor cycle—something we expect to endure over the next 10 years.

Onsemi is focusing its portfolio on the automotive and industrial markets, which we think will expand margins and create stickier customer relationships.

We think Onsemi will continue to outgrow its underlying markets by selling greater dollar content into applications like electric vehicles and renewable energy applications, which also helps partially offset its vulnerability to market cycles.

Onsemi’s margin profile is significantly structurally improved, helping profitability and economic returns across demand cycles.

Onsemi is prone to market cyclicality that can negatively affect its sales and margin profile.

Onsemi faces high competition in silicon carbide chips from well-capitalized peers in a dynamic market. We also expect some commoditization of SiC chipmaking over time.

We think management faces considerable execution risk in maintaining its lofty margin targets, especially as its footprint consolidation has lagged other power competitors.