Hook
Bank of America just slashed ON Semiconductor’s target price by 15%. The company’s quarterly profits are up 40% year-over-year. The code doesn’t lie—but the market’s narrative is rewriting the signal. Every rug pull has a pre-written script, and this one reads like a classic liquidity diversion: pump the headline earnings, then quietly reprice the underlying decay. Tracing the alpha through the noise of consensus, I found that the real story isn’t about revenue growth—it’s about the structural geometry of capital allocation in a sector that’s starving for differentiation.
Context
ON Semiconductor is a classic power semiconductor IDM, riding the dual narratives of automotive electrification and AI data center power density. Its SiC (silicon carbide) vertical integration—from substrate to module—is often touted as a moat. But the semiconductor industry is a spectrum of narratives, not a switch of binary outcomes. The same dynamics that inflated DeFi TVL during the 2021 bull run are now inflating ON’s profit line: lagging indicators that mask forward-looking compression. The Bank of America downgrade is not a bearish anomaly; it’s a red-team analysis of the company’s future free cash flow, a signal that the market is shifting from “story” to “structural math.”
Core: The Arithmetic of Narrative Decay
Let’s deconstruct the profit surge. ON’s gross margin peaked at 49% in 2022 and has since slid to ~45%. The culprit? Capacity underutilization and new fab depreciation—specifically the 300mm East Fishkill ramp and the SiC wafer fab in Vermont. These are capital-intensive bets that only pay off when utilization exceeds 85%. In a market where automotive and industrial inventories are still being digested, that threshold is a mirage. The 40% profit jump is a lagging artifact of prior backlog releases, not a signal of organic demand acceleration.
Now contrast this with the SiC price war. SiC MOSFET prices have dropped 30% year-over-year as Chinese competitors (BYD Semiconductor, Starpower) flood the market. ON’s vertical integration gives it a cost advantage, but not a pricing moat. The geometry of the market is shifting: the value is moving from “we make the best SiC” to “we can survive the price compression.” That’s a fundamental narrative shift, and the market is pricing it in via the target price cut.
My own experience auditing the 2017 Ethereum whitepaper taught me that when a network’s gas cost model has a hidden inconsistency, the market eventually finds it. The same applies here: the inconsistency is between ON’s profit growth (which is real) and its capex-to-revenue ratio (which is rising). At 10–15% capex/revenue, ON is spending more per dollar of sales than traditional analog IDMs. This is fine in a bull cycle, but it becomes a structural drag when the demand curve flattens. The code doesn’t excuse flawed fundamentals.
Contrarian: The Bull Case That’s Already Priced In
The mainstream narrative says: “AI data centers need more power, EVs need more SiC, so ON is a long-term winner.” That’s true, but it’s also a consensus trade. The contrarian angle is that the market is overestimating the speed of adoption. AI server power architectures are migrating to 48V/800V, but the volume ramp is 2027+, not 2025. EV penetration in the West is slowing due to charging infrastructure gaps. Meanwhile, the inventory glut in industrial power MOSFETs is still unwinding. The real risk is that ON’s margin expansion is a one-time event from prior cycle bookings, and the next 12 months will see a sequential compression.
Furthermore, the CHIPS Act subsidies—up to $1.5 billion—are a double-edged sword. They reduce ON’s capital expenditure risk, but they also incentivize capacity expansion that exacerbates the industry’s supply glut. The subsidy is a safety net, not a growth catalyst. Once the market realizes that the subsidy is a “debt” to future earnings (via overcapacity), the narrative will pivot again.
Takeaway: The Next Narrative Phase
The next narrative for power semiconductors isn’t “AI+EV” anymore—it’s “capital efficiency.” Investors will start valuing companies on return on invested capital (ROIC) rather than revenue growth. The same shift happened in crypto when the market moved from “total value locked” to “fee revenue per user.” ON’s future depends on its ability to keep SiC pricing above its fully loaded cost, even as the market commoditizes. If they can’t, the profit surge will be remembered as a prelude to a structural correction. The alpha lies in watching the capex-to-free-cash-flow ratio, not the earnings release. Decentralization is a spectrum, but capital allocation is a binary—you either waste it or you don’t.
