ON Semiconductor (NASDAQ: ON) is one of the most misunderstood stories in the power-semiconductor complex right now. The stock trades at $72.97, roughly 46% below its 52-week high of $134.92, after a brutal two-year downcycle in automotive and industrial chips gutted revenue from a 2022–2023 peak of more than $8.3 billion to $6.0 billion in fiscal 2025. GAAP earnings per share collapsed to just $0.29 in 2025 as the company absorbed more than $660 million in impairment and restructuring charges while it consolidated fabs. On the surface, that looks like a broken semiconductor name.
Look closer and the picture inverts. onsemi (the company’s brand name) just beat Q2 2026 estimates, driven by a genuinely new growth vector — silicon carbide power devices going into AI data centers — layered on top of a stabilizing automotive business that is now gaining share in China’s electric-vehicle market. Consensus forward earnings power sits at $4.50 per share, versus $1.55 on a trailing-twelve-month basis, and Wall Street’s mean price target of $110.26 implies roughly 51% upside from here. This is the classic setup of a high-quality cyclical caught at the bottom of its cycle just as a secular catalyst reignites the top line.
This article makes three core arguments. First, the automotive and industrial inventory correction that crushed ON’s revenue is largely behind it, and the incremental margin recovery off a $6.0 billion trough is powerful because onsemi deliberately shed low-margin revenue and vertically integrated its silicon carbide supply chain during the downturn. Second, the silicon carbide franchise — historically an EV story — now has a second, faster-growing leg in AI data center power delivery, where 800-volt architectures increasingly favor SiC over legacy silicon. Third, at 16x forward earnings for a business with a durable cost-and-integration moat in power semiconductors, ON is priced as a broken cyclical rather than a recovering structural grower, and that gap is the opportunity.
Below, we walk through onsemi’s business model and segment mix, size the power-semiconductor and silicon carbide market it competes in, dissect its economic moat, rebuild its financials year by year through the downcycle, run a valuation with explicit bull/base/bear scenarios, lay out the real risks, and finish with an actionable rating and exit plan. Throughout, the anchor question is simple: what is normalized earnings power worth for a company whose reported EPS is temporarily depressed by a cycle trough and one-time charges?
1. Company Overview
ON Semiconductor Corporation, which markets itself as onsemi, is a designer and manufacturer of intelligent sensing and power-management semiconductors. In plain terms, it makes the chips that convert, control, and route electrical power, and the image sensors that let machines “see.” Its products are the unglamorous but essential silicon inside electric-vehicle traction inverters, on-board chargers, industrial motor drives, solar inverters, factory automation systems, and — increasingly — the power-delivery backbone of AI server racks.
The company generates revenue through three reporting segments. The Power Solutions Group (PSG) is the largest and houses the discrete power devices, modules, and — critically — the silicon carbide (SiC) franchise that sits at the center of the growth thesis. The Analog and Mixed-Signal Group (AMG) supplies power-management ICs, connectivity, and analog components. The Intelligent Sensing Group (ISG) builds image sensors and depth-sensing products, where onsemi holds a strong, established position in automotive image sensors used for advanced driver-assistance systems (ADAS). Power Solutions is the swing factor for both revenue and margins; the automotive and industrial end markets together drive the majority of corporate revenue, which is exactly why the 2024–2025 downturn in those markets hit so hard.
onsemi’s revenue model is a blend of design wins and long-term supply agreements. In power semiconductors, once a chip is designed into a customer’s traction inverter or on-board charger, it tends to stay there for the multi-year life of the vehicle platform, generating recurring, sticky volume. The company has also signed multi-year, capacity-linked supply agreements with major automotive and industrial customers for silicon carbide — a structure that converts what would otherwise be spot demand into contracted backlog. Its customer base spans the largest global automakers and Tier-1 suppliers, and management has highlighted content gains with Chinese EV makers including Geely’s Zeekr brand and Xiaomi as evidence that its automotive franchise is taking share even where overall vehicle volumes are soft.
On market position, onsemi is one of the top players globally in discrete power semiconductors and image sensors, and it has invested aggressively to be a vertically integrated silicon carbide supplier — from substrate through device — which is a differentiator versus rivals that buy substrates externally. On governance and ownership, ON is a widely held large-cap with a market capitalization of $28.41 billion; the shareholder base is dominated by institutional investors typical of an S&P 500 semiconductor name, and insider ownership is modest, as is common for a company of this scale. Beta of roughly 2.0 tells you the market treats ON as a high-cyclicality stock — a fact that cuts both ways depending on where you catch the cycle.
2. Industry Analysis
2-1. Market Size & Growth Trajectory
onsemi competes primarily in the power-semiconductor market — the chips that manage the conversion and delivery of electrical energy. This is a large and structurally growing category. Industry estimates put the total power-semiconductor market in the range of $50–55 billion today, growing at a high-single-digit to low-double-digit compound annual rate as electrification spreads across vehicles, industrial systems, renewable energy, and computing. Within that, the silicon carbide sub-segment is the fastest-growing slice: SiC device revenue is broadly expected to scale from a few billion dollars today toward roughly $10 billion or more by the end of the decade, a mid-to-high-teens or faster CAGR depending on the forecaster.
The reason the market is growing is physics, not fashion. Silicon carbide is a wide-bandgap material that switches faster, tolerates higher voltages, and runs cooler and more efficiently than legacy silicon. In an electric vehicle, replacing silicon insulated-gate bipolar transistors (IGBTs) with SiC in the traction inverter improves driving range by several percentage points — an enormous benefit when battery packs cost thousands of dollars — and enables the 800-volt architectures that allow faster charging. The same efficiency advantage is now proving valuable in AI data centers, where the sheer power draw of accelerator racks is forcing a redesign of the entire power chain.
Where is the industry in its cycle? The answer is nuanced and central to the thesis. The automotive and industrial power-semiconductor market is emerging from a painful inventory correction that ran through 2024 and 2025, when customers who had double-ordered during the shortage years worked down excess stock. That correction is what compressed onsemi’s revenue. On the demand side, the underlying content-per-vehicle trend never reversed — electrification keeps adding silicon and silicon carbide to each car. Simultaneously, the AI data center power opportunity is in an early-growth, acceleration phase. So onsemi sits at the intersection of one market bottoming and normalizing and a second market inflecting upward — an unusually favorable place to be positioned as the cycle turns.
2-2. Structural Growth Drivers
Driver 1 — Electric-vehicle content growth and silicon carbide adoption. The single largest long-term driver is the rising dollar content of power semiconductors in each vehicle as the fleet electrifies and as automakers migrate from silicon to silicon carbide. A conventional internal-combustion car carries a few hundred dollars of semiconductor content; a modern EV carries multiples of that, and the power-specific portion — inverters, chargers, DC-DC converters — is precisely onsemi’s wheelhouse. Even in a year when total vehicle unit sales are flat or down, onsemi’s automotive revenue can grow because content per vehicle is rising. Management’s disclosure is instructive here: in the first half of 2026, its automotive revenue in China rose 13% year over year even as total China vehicle sales fell 4%, because content gains and new program ramps more than offset the volume softness. The company guided to China automotive silicon carbide revenue growing 60–70% for the full year as share gains and new vehicle programs ramp. This is the definition of a structural driver overwhelming a cyclical headwind.
Driver 2 — Silicon carbide in AI data center power delivery. This is the newest and, arguably, the most exciting leg because it diversifies the SiC story away from pure automotive cyclicality. As AI accelerator racks push power densities to unprecedented levels, hyperscalers are re-architecting data center power distribution toward higher-voltage systems that minimize resistive losses. Silicon carbide’s efficiency at high voltage makes it the natural switching technology for this transition. onsemi has guided that its silicon carbide revenue tied to AI data center applications is set to grow nearly 60% year over year in 2026. Crucially, this demand is far less correlated with the automotive cycle, so it both accelerates growth and smooths the revenue profile. It also validates onsemi’s decision to build vertically integrated SiC capacity during the downturn — that capacity now has a second, fast-growing customer set to absorb it.
Driver 3 — Industrial electrification and energy infrastructure. Beyond autos and data centers, the broad electrification of industrial systems — factory automation, robotics, solar and energy-storage inverters, medical equipment, and grid infrastructure — provides a long-duration demand tailwind for power management and sensing. This end market is more fragmented and less headline-grabbing than EVs or AI, but it is large, diversified, and tends to carry attractive margins because designs are sticky and price competition is less intense than in consumer categories. As industrial customers work through their own inventory normalization, this segment provides a second cyclical recovery lever on top of the automotive rebound, and its long-term growth is anchored to the same efficiency-and-electrification megatrend.
2-3. Competitive Landscape
The power-semiconductor and silicon carbide arena is competitive, populated by large, well-capitalized incumbents. The table below frames onsemi against the peers investors most often compare it to. Figures are approximate and drawn from public data; they are meant to show relative positioning, not to serve as precise valuation inputs.
Company (Ticker) Approx. Revenue (TTM) Approx. Operating Margin Approx. Market Cap Primary Moat ON Semiconductor (ON) ~$6.2B ~18% (TTM) ~$28B Vertically integrated SiC, EV/industrial design wins Infineon (IFNNY) Larger Mid-teens–20% Much larger Scale leadership in power, breadth STMicroelectronics (STM) Larger Compressed recently Larger Broad analog/power, SiC scale Texas Instruments (TXN) Larger High (30%+) Much larger Analog cost/scale, internal manufacturing Wolfspeed / pure-play SiC Smaller Negative/pressured Smaller SiC substrate technology
Approximate figures for relative comparison only; not for valuation.
onsemi’s differentiated position rests on two things. First, it is one of the few players that vertically integrated silicon carbide from the substrate up, which — assuming yields and utilization normalize — gives it cost and supply-security advantages that non-integrated competitors lack, particularly against pure-play SiC suppliers that have struggled with profitability. Second, its dual franchise in both power and a strong automotive image-sensor business gives it a broad ADAS and vehicle-electrification design footprint that is hard to displace once locked into a platform. Against giants like Infineon and STMicroelectronics, onsemi is not the largest, but it does not need to be; it needs to defend and grow its share in the SiC and automotive-power niches where it has invested most heavily. Against a pure analog cost-leader like Texas Instruments, onsemi competes on application-specific power and sensing content rather than head-to-head on commodity analog.
3. Economic Moat Analysis
Moat Type 1: Switching Costs and Design-Win Lock-In
The strongest element of onsemi’s moat is the switching cost embedded in automotive and industrial design wins. When a power semiconductor is designed into a vehicle’s traction inverter or an industrial motor drive, it is qualified through a long, expensive validation process governed by stringent reliability and safety standards. Automotive-grade qualification alone can take one to two years and substantial engineering investment on both sides. Once a part is designed in, it typically remains for the entire production life of the platform — often five to seven years or more — because re-qualifying an alternative part mid-cycle is costly, risky, and rarely worth it for the customer. This is why onsemi’s content gains with Chinese EV makers matter so much: each new design win is not a one-time sale but an annuity that compounds as that platform ramps to volume. The stickiness is visible in the data — automotive revenue rising while unit sales fall is only possible when incumbents hold their sockets and add content per vehicle.
Moat Type 2: Cost Advantage Through Vertical Integration
onsemi’s second moat is a manufacturing cost-and-supply advantage built through vertical integration in silicon carbide. Rather than buying SiC substrates on the open market — where supply is tight and pricing is volatile — onsemi invested through the downturn to control its own substrate-to-device supply chain. In a wide-bandgap market where substrate availability has historically been a bottleneck, owning that step provides both cost leverage and supply security that competitors dependent on external substrate vendors cannot easily match. This integration is precisely why the company took restructuring and impairment charges in 2025: it was consolidating and rationalizing its manufacturing footprint to concentrate production in its most efficient, integrated fabs. The payoff is operating leverage — as utilization rises off the trough, incremental revenue drops through to margin at a high rate because the fixed-cost base has already been absorbed and optimized.
Moat Durability Assessment
Will this moat hold over five to ten years? On balance, yes, but with real caveats. The switching-cost moat is durable by nature — automotive design cycles are long, and the shift to electrification only deepens the reliance on power semiconductors. The vertical-integration cost moat is more contingent: it depends on onsemi executing on yields and utilization, and on silicon carbide remaining the material of choice versus alternatives like gallium nitride (GaN) in certain voltage ranges. The most credible threat to the moat is aggressive pricing and capacity expansion from larger, better-capitalized rivals — Infineon and STMicroelectronics both have significant SiC ambitions — which could compress the pricing umbrella. The counterargument is that onsemi’s integrated cost position and entrenched design wins give it room to compete on price while protecting margin better than non-integrated peers, and the emergence of AI data center demand gives its capacity a second, less price-sensitive outlet. The moat is real and defensible, but it is a cost-and-relationship moat that must be continuously earned through execution, not a permanent monopoly.
4. Financial Analysis
onsemi’s financial statements tell the story of a cyclical peak, a sharp downcycle, and the early innings of a recovery. The table below lays out the trajectory using reported figures.
Fiscal Year Revenue Operating Income Net Income Diluted EPS 2022 $8,326.2M $2,764.7M $1,902.2M $4.24 2023 $8,253.0M $2,613.6M $2,183.7M $4.89 2024 $7,082.3M $1,901.6M $1,572.8M $3.63 2025 $5,995.4M $751.1M $121.0M $0.29 TTM (current) ~$6,200M — $630.2M $1.55
The arc is unmistakable. Revenue peaked around $8.3 billion in 2022–2023 at the height of the post-shortage automotive and industrial demand surge, when diluted EPS reached $4.24 and $4.89 respectively. Then the inventory correction hit: revenue fell to $7.08 billion in 2024 and to $5.995 billion in 2025 — a roughly 28% top-line decline from peak to trough. The 2025 net income figure of $121 million and $0.29 diluted EPS look catastrophic in isolation, but they are heavily distorted by more than $660 million of unusual items — impairment of capital assets and restructuring charges tied to the fab consolidation and SiC footprint rationalization. Strip those out and the picture is far less dire: 2025 operating income was still $751 million and normalized income was closer to $750 million. In other words, the GAAP trough overstates the deterioration in the underlying earnings engine.
The recovery signal shows up in the trailing-twelve-month figures. TTM revenue has stabilized at roughly $6.2 billion, TTM net income has rebounded to $630.2 million, and TTM diluted EPS is back to $1.55. Growth metrics confirm the inflection: EPS is up about 36% this year, quarter-over-quarter sales grew roughly 9%, and quarter-over-quarter EPS grew roughly 37%. The Q2 2026 beat — powered by AI data center demand, a richer product mix, and higher fab utilization — is the operational evidence that the trough is behind the company.
On profitability, the TTM margin structure is a 37.1% gross margin, an 18.2% operating margin, and a 10.2% net margin, with return on equity of 8.3% and return on assets of 4.7%. These are trough-adjacent margins — well below the 30%-plus operating margins onsemi posted at the 2022–2023 peak — which is exactly the point: the operating leverage to the upside is substantial as utilization and mix normalize. The critical operating metrics to watch are silicon carbide revenue growth (guided at 60–70% for China automotive and nearly 60% for AI data center in 2026) and fab utilization, since both drive the margin recovery.
On the balance sheet, onsemi carries a debt-to-equity ratio of 0.65, which is moderate and manageable for a company generating meaningful operating cash flow even at the trough. The forward earnings estimate of $4.50 per share — versus $1.55 trailing — encapsulates the market’s expectation of a strong margin and revenue recovery. The path from here is not a turnaround from losses (the company remained profitable throughout) but a margin-expansion and revenue-normalization story: as the automotive inventory correction fully clears, industrial demand recovers, and silicon carbide scales across both EV and AI data center end markets, revenue re-approaches prior peaks while the leaner, integrated cost structure converts more of that revenue into profit.
5. Valuation
The central valuation question for onsemi is what multiple to pay for normalized forward earnings, because trailing GAAP earnings are depressed by the cycle trough and one-time charges. Using trailing EPS of $1.55, the stock trades at a P/E of about 47x ($72.97 ÷ $1.55 ≈ 47.1, consistent with the reported 46.99) — a number that is meaningless in isolation because it capitalizes trough earnings. The relevant lens is forward earnings: on consensus forward EPS of $4.50, ON trades at a forward P/E of just 16.2x ($72.97 ÷ $4.50 ≈ 16.2, consistent with the reported 16.22). That is the number that matters.
Sixteen times forward earnings is inexpensive for a power-semiconductor franchise with a vertically integrated silicon carbide moat and two secular growth drivers (EV content and AI data center). Quality analog and power peers such as Texas Instruments and Analog Devices routinely command mid-20s to low-30s forward multiples, and even accounting for onsemi’s higher cyclicality and lower current margins, a recovering ON should not trade at half the peer multiple once the market gains confidence the cycle has turned.
PER-based fair value. Applying a target forward multiple to consensus forward EPS of $4.50:
– A conservative base-case multiple of 23x on $4.50 yields a fair value of approximately $104 (23 × $4.50 = $103.50), roughly 42% above the current price and broadly in line with the analyst mean target of $110.26.
– A bull-case multiple of 26x — justified if the AI data center silicon carbide ramp accelerates and margins recover faster toward prior-peak levels — yields approximately $117 (26 × $4.50).
– A bear-case anchored on the lower FY2026 consensus earnings of roughly $3.13 per share at a de-rated 15x multiple yields approximately $47 (15 × $3.13), close to the 52-week low of $44.56, reflecting a scenario where the recovery stalls and the multiple compresses back to trough levels.
This spread — roughly $47 on the downside to $117 on the upside, with a base case around $104 — frames the risk/reward. From $72.97, the base case offers about 42% upside while the bear case implies about 36% downside, an asymmetry that skews favorably if you believe the cyclical recovery and silicon carbide growth are intact.
Versus consensus. The Wall Street mean target of $110.26 (from roughly 25 covering analysts, with a “buy” consensus) implies about 51% upside and sits just above our base case. We broadly agree with the direction but would anchor to a slightly more conservative base of ~$104 to leave margin for cyclical slippage, treating the $110 consensus as the upper end of a reasonable base range rather than a point estimate. The disagreement is minor: both our work and the Street see meaningful upside from a trough valuation.
Scenario summary. Bull $117 / Base $104 / Bear $47. The base and bull cases depend on the silicon carbide growth guidance materializing and margins recovering off the trough; the bear case depends on a stalled recovery and multiple compression. Given the operational evidence in the Q2 2026 beat and the guided SiC growth rates, we weight the base/bull outcomes more heavily than the bear.
6. Risk Factors
Risk 1 — Automotive and industrial cyclicality (and beta). onsemi is a high-beta cyclical (beta ~2.0), and its revenue is heavily concentrated in automotive and industrial end markets that move with the broader economy and the EV adoption curve. If the automotive inventory correction proves deeper or longer than expected, or if a macro slowdown suppresses industrial capital spending, revenue could stagnate below the $6 billion level and the margin recovery embedded in the $4.50 forward EPS would not materialize. The 2024–2025 downcycle is a fresh reminder of how sharply this business can contract — revenue fell 28% peak to trough. Investors buying the recovery thesis are, by definition, taking a view that the cycle has bottomed; if that timing is wrong, the stock’s high beta means it can fall quickly, as the drop from $134.92 to the mid-$40s during the downturn demonstrated. This is the single most important risk to underwrite.
Risk 2 — Silicon carbide competition and pricing pressure. The silicon carbide opportunity that anchors the growth thesis is also attracting aggressive investment from larger, better-capitalized rivals including Infineon and STMicroelectronics, as well as pure-play substrate specialists. A wave of new SiC capacity coming online across the industry could compress pricing faster than onsemi can offset with volume and cost improvements, eroding the margin recovery. There is also technology risk: gallium nitride (GaN) is encroaching on some voltage ranges where SiC competes, and any shift in the preferred wide-bandgap material for specific applications could strand capacity. onsemi’s vertical integration partially insulates it on cost, but a price war in SiC would still pressure the very margins the bull case depends on. Because so much of onsemi’s growth capital has been directed at SiC, adverse competitive dynamics there would hit disproportionately.
Risk 3 — China concentration and geopolitical exposure. A meaningful and growing portion of onsemi’s automotive silicon carbide momentum is tied to Chinese EV makers, with management explicitly citing 60–70% China automotive SiC growth and content wins with Chinese OEMs. This concentration is a double-edged sword. It is a genuine growth engine today, but it exposes onsemi to Chinese EV demand cyclicality, intense local price competition, the rise of domestic Chinese power-semiconductor suppliers that could displace foreign incumbents over time, and escalating US–China trade and export-control tensions. Tariffs, export restrictions on advanced semiconductors or equipment, or a policy push toward domestic Chinese sourcing could each impair a growth channel that the market is currently rewarding. Investors should size this exposure deliberately rather than assume the China SiC ramp is a straight line.
7. Conclusion & Exit Plan
Investment rating: Buy. ON Semiconductor offers an attractive risk/reward for investors willing to underwrite a cyclical recovery in power semiconductors with a genuine secular kicker in silicon carbide. The stock is priced at 16x forward earnings — roughly trough-cycle economics — while the underlying business shows clear evidence of inflecting: a Q2 2026 earnings beat, silicon carbide guidance of 60–70% growth in China automotive and nearly 60% in AI data center, and TTM earnings already recovering to $1.55 from a GAAP trough distorted by one-time charges. The vertically integrated SiC moat and sticky automotive design wins give the recovery durability. We stop short of Strong Buy only because of the high beta and the real possibility that the cyclical timing is early.
Entry price range. The current price of $72.97 is a reasonable entry given roughly 42% upside to a $104 base case. We would view any pullback into the $60–$70 range — which could occur on a soft automotive data point or broad semiconductor volatility — as an opportunity to build a fuller position, since it would widen the margin of safety toward the bear-case floor near $47 while leaving the recovery thesis intact.
Exit conditions:
– Target achieved: Trim the position as the stock approaches the $104 base-case target; take further profit toward the $110–$117 range where it meets or exceeds consensus and the bull-case multiple.
– Fundamental break: Sell if the recovery thesis breaks — specifically if operating margins fail to expand off the ~18% TTM level for two or more consecutive quarters despite the SiC ramp, if silicon carbide revenue growth decelerates materially below guidance, or if revenue slips back below the $6 billion trough. Any of these would invalidate the normalization-and-leverage core of the thesis.
– Time-based: Reassess in 6–9 months, or immediately upon the next two quarterly reports, to confirm the automotive inventory correction has fully cleared and the AI data center SiC ramp is tracking to plan.
Summary table:
Item Detail Company ON Semiconductor Corporation (ON) Current Price $72.97 Target Price $104 (base case) Upside ~42% (base) / ~51% to consensus $110 Rating Buy Key Thesis Trough-cycle power-semi at 16x forward earnings with silicon carbide growth in both EV and AI data center Main Risk Automotive/industrial cyclicality — recovery timing may be early
—
This article is for informational purposes only and does not constitute investment advice. All data sourced from public filings, analyst reports, and news as of the publication date. Invest at your own discretion.
This content is general investment information provided to an indefinite/unspecified audience by a quasi-investment advisory business registered under Korea’s Financial Investment Services and Capital Markets Act, and is not personalized 1:1 investment advice tailored to any individual investor. This analysis is for informational purposes only and is not a solicitation to invest. All investment decisions and their consequences rest solely with the investor. The estimates and assumptions in this report are as of the writing date (2026-08-31) and may not materialize depending on market conditions and geopolitical variables. Financial data used reflects sources such as company filings and analyst consensus, and the scenarios and price targets represent the author’s conservative assessment. All investments carry the risk of principal loss, and past performance or analytical track record does not guarantee future results. As of the writing date, the author does not hold a position in this stock. The author’s holdings and positions may change without prior notice depending on market conditions.
함께 읽으면 좋은 글
- Parsons Golden Dome Missile Defense Exposure: Why a 45% Drawdown Left an 18% Upside Setup at 14x Forward Earnings
- Abercrombie & Fitch Q2 2026 Earnings: Why 15% Operating Margins and a 12x Multiple Reframe the ANF Turnaround Story
- Shift4 Payments Forward P/E of 7 Meets 34% Revenue Growth: Why Wall Street Upgraded FOUR After a 50% Selloff
- GoDaddy Airo AI Monetization vs a 14.75% Free Cash Flow Yield: Why GDDY’s 36% Drawdown Looks Overdone in 2026
- Ross Stores Off-Price Comparable Sales Streak: Back-to-Back 10% Comps and Why the Cost Moat Points to ~12% Base-Case Upside