Bank of America cut onsemi's target price. The stock didn't crash. That's the first red flag. The market shrugged, calling it a minor adjustment. I don't shrug. I dissect.
Here's the narrative: onsemi posts profit growth, yet analysts trim the price target. Contradiction? No. It's a classic signal of forward-looking pessimism masked by backward-looking results. The profit is a lagging indicator. The target price is a leading one. The gap between them is the structural flaw.
Let me be clear: I'm not a semiconductor analyst. I'm a crypto security audit partner who spent 27 years watching blockchain projects die from the same disease—narrative inflation masking technical debt. When I see Bank of America downgrade onsemi while the company reports record earnings, I smell a pattern. The exploit wasn't a hack. It was a margin call on the AI-and-electrification story.
Context: The IDM Hype Cycle
onsemi is an Integrated Device Manufacturer—designs, fabricates, and packages power semiconductors. It's not a 3nm digital logic player. It's a mature-node powerhouse: MOSFETs, IGBTs, SiC MOSFETs, and super-junction devices. The bull case rests on two pillars: electric vehicles and AI data center power density. Both are real. Both are long-term. But the market is pricing in a linear ramp, and semiconductor history is never linear.
In 2020-2022, power semiconductors enjoyed a supply-constrained boom. Then came the inventory glut. By 2024, automotive and industrial customers were bleeding down stock. onsemi's gross margin dropped from 49% to 45%. The company blamed the cycle. I blame the structure.
Core: The Systematic Teardown
I see three structural vulnerabilities that the target price downgrade is flagging—not explicitly, but through the lens of earnings revision risk.
- SiC Price War: The Liquidity Fragmentation of Power Semiconductors
Silicon carbide (SiC) is onsemi's crown jewel. The company vertically integrated SiC substrate production via GT Advanced Technologies. That's a moat. But moats don't prevent price wars. China's SiC fabs are ramping. Infineon and ST are expanding. The result: SiC MOSFET ASPs are falling faster than volume growth. Revenue is a mirror, not a vault. When price drops 20% and volume grows 15%, you're losing ground. The market is pricing that in.
Standardization fails when it ignores human chaos. The SiC industry is standardizing around 1200V and 650V devices. But the chaos of oversupply and competing procurement strategies—especially from Chinese OEMs—destroys pricing discipline. onsemi's vertical integration protects margins less than the bulls think because the substrate cost advantage is shrinking as competitors improve yields.
- Capex Overhang: The Layer2 Scaling Trap
onsemi is spending billions on 300mm wafer capacity in East Fishkill, New York, and SiC fab expansion in Vermont. The CHIPS Act subsidizes part of it. But the revenue-to-capex ratio is 10-15%, similar to analog IDMs, not 30%+ like TSMC. That looks conservative. But the real cost is not the capex—it's the depreciation. New fabs add depreciation for 5-7 years. If capacity utilization stays below 85%, those depreciation charges compress free cash flow. Logic is binary; trust is a spectrum. The market trusts the long-term story but discounts the near-term EPS hit.
This is exactly what I've seen in L2 rollups: projects scale capacity before demand, then burn cash on idle validators. The bulls say "prepare for the next cycle." The bears say "you didn't lose to a competitor, you lost to your own cost structure." onsemi's capex is a bet on 2027 demand. The target price downgrade is a bet that 2025 earnings will disappoint.
- Customer Concentration: The Oracle Manipulation Vector
onsemi derives ~50% of revenue from automotive. That's a single exposure vector. In crypto, we call it a liquidity pool with one dominant LP. When that LP withdraws (i.e., automakers cut orders), the protocol bleeds. In 2024, automotive semiconductor orders weakened. onsemi's inventory days increased. The company blamed the macro. I blame the business model.
In code, silence is the loudest vulnerability. In financial statements, the silence is the lack of revenue diversification. The "AI and industrial" segments are growing, but from a small base. The power management for AI data centers is a high-growth niche, but it's still a small fraction of total revenue. The market is right to ask: what happens when EV adoption slows further?
Contrarian: What the Bulls Got Right
I'm not here to bury onsemi. The bulls have a valid thesis. The vertical integration in SiC is real—they control substrate, epitaxy, device fabrication, and module packaging. That's a structural advantage. The CHIPS Act subsidy reduces the risk of overinvestment. The long-term demand for power semiconductors driven by electrification and AI computing is undeniable. The blockchain remembers, but the auditors forget. The market remembers the narrative, but the analysts focus on the numbers.
What the bulls missed is the speed of price erosion in SiC and the duration of the inventory correction. They assumed a V-shaped recovery. The data shows a U-shaped, possibly L-shaped, normalization. The target price downgrade is not a rejection of the thesis—it's a recalibration of the timing. When the narrative shifts from "AI tailwind" to "earnings revision," who will be left holding the bag?
Takeaway: The Accountability Call
onsemi is not a bad company. It's a good company in a cyclical industry with a structural pricing problem. The target price downgrade is a wake-up call for anyone who treats "AI and electrification" as a perpetual growth engine. The market is pricing in the next 12 months, not the next 12 years. The question is: will the earnings revision validate the downgrade, or will the structural defensiveness of an IDM absorb the shock?
Based on my audit experience, I've seen this pattern before. The project that looks solid on the whitepaper but fragile on the financial sheet. The team that blames the market for their margin compression. The investors who buy the narrative and ignore the margin call. In blockchain, code is law until someone finds the edge case. In semiconductors, the balance sheet is truth until the depreciation catches up.
onsemi's true test is not whether it survives—it will. The test is whether it can maintain pricing power in a commoditizing SiC market. If not, the target price will be the first of many downward revisions. The blockchain remembers. The market will too.