STMicroelectronics looks like a classic post-earnings overreaction. The selloff fixated on an ugly profit miss and a slightly light Q3 guide, while management quietly handed the market a much bigger AI-linked revenue story by lifting its data-center ambition to above $1 billion in 2026 and well above $2 billion in 2027. That is not a cosmetic tweak. It is the kind of step-change that can reframe STM from a cyclical chip name into an AI infrastructure supplier, and the market's first reaction missed that shift.
The most important number here is the data-center target reset. In early June, management moved its 2026 data-center revenue goal to about $1 billion from a prior target of nicely above $500 million, then reiterated with Q2 results that 2027 could reach well above $2 billion if demand trends hold. That matters because it gives STM a concrete AI revenue bridge instead of a vague narrative. When a semiconductor company doubles its near-term ambition in one of the market's highest-multiple end markets, that deserves more attention than a one-quarter margin wobble.
The quarter itself was not nearly as broken as the price action suggested. Q2 revenue came in at $3.49 billion, up 26.0% year over year, and management said Q4 revenue should land above $4 billion, driven mainly by AI datacenters and LEO satellite programs. Yes, Q3 guidance of $3.70 billion plus or minus 3.5% was a touch below consensus, but that is a tiny miss relative to the size of the longer-term demand upgrade. The market treated STM like the growth engine stalled when management was actually pointing to acceleration into year-end.
The setup also looks better than the headline valuation suggests. STM's trailing P/E of 102.66 looks extreme because earnings have been crushed, with EPS growth down 89.0% year over year and net margin at just 3.5%, but that is exactly why the stock should be judged on recovery potential and balance-sheet resilience instead of depressed trailing profits. The TickerSpark Score backs that up: Financial Health is a standout 96, Valuation is 60, and Momentum is 60 despite the post-earnings hit. Against ON Semiconductor, STM is the more compelling contrarian AI rerating candidate right now: ON trades at 36.36 times earnings and 5.79 times sales with revenue down 15.3%, while STM trades at a lower 3.57 times sales and now has a clearer data-center growth catalyst.
The pushback is easy to understand because the earnings quality was weak. EBITDA of $679 million missed the market's $797.7 million expectation, the company has beaten estimates only 3 of the last 8 quarters, and the latest EPS miss was brutal at 0.06 versus 0.26 expected. The stock had also already surged 95.4% year to date, so investors were primed to punish anything short of a clean beat-and-raise.
That skepticism is fair, but it still misses the bigger point. The market already knows STM's current profitability is soft; the TickerSpark Score gives Profitability just 45 and Growth only 10, so none of that is hidden. What is changing now is the mix of future revenue. If Q4 really clears $4 billion on AI datacenter demand, the stock will not be priced off today's 3.5% net margin forever. It will be priced off the path back to operating leverage.
That leaves STM looking more attractive after the drop than it did before the print. We would treat this as a recovery-and-rerating setup, not a clean quality compounder, which means the right posture is constructive but disciplined. The trigger that keeps the bull case alive is simple: management has to convert the raised AI data-center ambition into visible Q4 acceleration and better margins by the October 29 report.
If that proof shows up, the selloff will look like the market handed investors a second entry into an AI-adjacent semiconductor name. If it does not, and the soft Q3 guide turns into a broader slowdown instead of a timing issue, the thesis breaks fast. For now, we'd rather own STM than ON because STM's AI revenue narrative just got materially bigger, while the post-earnings damage already reflects a lot of the near-term ugliness.
