STMicroelectronics (STM) guides Q3 revenue below consensus even as 2027 data centre target crosses $2bn

STMicroelectronics N.V. (NYSE: STM; Euronext Paris: STMPA; Euronext Milan: STMMI) shares fell as much as 17% during Thursday’s trading session, the sharpest intraday decline in about a year, after the Franco-Italian chipmaker guided third-quarter 2026 revenue below Wall Street estimates and delivered second-quarter core profit that missed consensus. The company simultaneously raised its data centre revenue ambition to above $1 billion for 2026 and well above $2 billion for 2027, reflecting continued strength in artificial intelligence infrastructure demand. The result exposed the central tension in the STMicroelectronics investment case: a rapidly scaling AI data centre franchise is running alongside a still-uneven recovery in the group’s automotive and industrial businesses, and near-term margins have not yet caught up with the top-line story. Even after Thursday’s fall, the stock remains up more than 110% year to date, which sets a high bar for any incremental catalyst.

Why did STMicroelectronics shares fall despite the raised data centre forecast for 2026 and 2027?

The sell-off reflected a mismatch between what management delivered and what a stock up more than 110% year to date had already priced in. STMicroelectronics guided third-quarter revenue to $3.70 billion, plus or minus 3.5%, against an LSEG analyst average of $3.72 billion and a Bloomberg-compiled average closer to $3.79 billion. On second-quarter earnings before interest, taxes, depreciation and amortisation, the company reported $679 million, materially below the roughly $797.7 million consensus figure cited by Reuters. Revenue for the quarter did beat expectations, but the profit shortfall was the number investors focused on, because it directly challenges the thesis that the semiconductor cycle recovery is translating quickly into operating leverage.

J.P. Morgan analysts summed up the market reaction by noting that the company had “not done enough to take the stock up significantly,” referencing the fact that a preliminary update in June had already primed investors for stronger numbers. Jefferies attributed part of the third-quarter guidance shortfall to a slower iPhone 18 ramp affecting mixed-signal and imaging content, while flagging that stronger gross margin guidance and the fourth-quarter outlook still pointed to a better-than-expected 2027. The reaction, therefore, is best read as expectations resetting rather than as a fundamental rejection of the AI story.

How does the third quarter revenue guidance compare with the accelerating fourth quarter outlook management is signalling?

Third-quarter revenue at the midpoint would represent growth of about 16% year on year, a meaningful improvement over the trough the company reported through late 2024 and early 2025. Management indicated that revenue growth will accelerate further in the fourth quarter as more orders come through for chips serving satellite communications and data centre customers. That sequencing is important. It suggests the June convertible bond raise and the capacity ramp announced earlier this year are beginning to translate into shippable product, not just booked demand, and it puts the fourth quarter in a position to test whether management’s stronger 2027 framing is credible.

For investors, the practical read is that the third quarter is a transition quarter rather than an inflection quarter. The near-term revenue print will not, on its own, validate the AI thesis. The fourth-quarter delivery, the exit-rate gross margin and any commentary on 2027 booking coverage during that release will do far more work in resetting sentiment, positively or otherwise.

What does the second quarter EBITDA shortfall reveal about margin recovery in the automotive and industrial chip businesses?

The EBITDA miss is the more informative disclosure inside the release. STMicroelectronics’ automotive and industrial exposures have been dragging on utilisation and gross margin for several quarters, and the second-quarter print suggests unused capacity charges and product mix are still working against the reported profit line. In its first-quarter release earlier this year, the company had itself flagged around 100 basis points of gross margin drag from unused capacity, and the second-quarter EBITDA gap indicates that headwind has not fully dissipated.

That matters for two reasons. First, the AI data centre franchise, while high-growth, is not yet large enough on its own to offset softness in core end markets on the profit line. Second, gross margin is the metric analysts have been waiting on to justify further multiple expansion. J.P. Morgan’s comment that investors still needed to see more evidence of gross margin improvement to become substantially more bullish captures the buy-side calculus. The revenue mix is shifting in the right direction, but the profitability mix has not yet reset.

How significant is the upgraded data centre revenue ambition relative to the group’s overall 2027 growth story?

The revised targets sharpen a picture that had been building since STMicroelectronics broke out AI-related revenue guidance for the first time in April 2026. The company had raised its 2026 data centre ambition in June from “nicely above $500 million” to about $1 billion, with 2027 seen as roughly double that. Thursday’s update lifted the 2027 anchor from “double” to “well above $2 billion” and firmed the 2026 figure at above $1 billion.

Against a group revenue base that would sit in the range of roughly $14 billion to $15 billion in a mid-cycle year, a $2 billion data centre business represents a mid-teens contribution. That is a meaningful mix shift from a business that was effectively pre-revenue in AI infrastructure two years ago, and it moves STMicroelectronics closer to the diversified analog and power-semiconductor peers that have benefited most from the AI infrastructure build-out. The scale, however, remains dependent on execution against customer engagements that management has referenced only in general terms. Investors have not yet been given the customer-by-customer visibility that would allow the $2 billion figure to be modelled with high confidence, and Thursday’s reaction suggests they are unwilling to pay for it in full until that visibility improves.

What role does the 1.5 billion dollar convertible bond issued in June play in financing the AI infrastructure expansion?

STMicroelectronics issued approximately $1.5 billion in convertible bonds in June 2026 while the share price was elevated, a move that gave the company additional liquidity at a favourable cost of capital but introduced conditional dilution into the equity story. The convertible structure means that the company has retained cash flexibility to finance the data centre capacity ramp without immediately expanding the share count, provided the conversion features are not triggered.

For the near-term investment case, the practical implications are twofold. The company now has cash headroom to invest into silicon carbide, edge AI and data centre product families through the current phase of capital expenditure without stressing the balance sheet. At the same time, any further leg higher in the share price increases the probability that the convertibles move into the money, which would eventually add to the diluted share count. The right way to frame the June raise is as a funded, but not free, path to scaling the AI infrastructure business.

How is competitive pressure from Texas Instruments and analog peers shaping investor expectations for the chip cycle recovery?

The reaction to STMicroelectronics’ release cannot be read in isolation. On Wednesday, Texas Instruments Inc. delivered a sales forecast that topped estimates, citing strength in industrial, data centre and automotive demand. Analog Devices Inc., NXP Semiconductors N.V., Microchip Technology Inc. and Marvell Technology, Inc. sit in adjacent parts of the same recovery, and each print recalibrates how the market judges the peer group.

Texas Instruments’ guidance effectively raised the bar for STMicroelectronics. When a comparable analog and mixed-signal manufacturer signals that its industrial and automotive customers are increasing orders, the market becomes less tolerant of any peer that guides below expectations on those same end markets. The read-across is that STMicroelectronics may be lagging the recovery cadence in some segments even as it leads in AI data centre positioning through its silicon photonics and power-management portfolios. Closing that perceived gap in the coming quarters is now a specific and observable test for management.

What execution risks could delay STMicroelectronics from converting AI data centre demand into sustained gross margin expansion?

Several execution risks remain relevant. Capacity ramp-up on advanced silicon photonics, high-voltage power products and specialised analog nodes is capital-intensive, and any timing gap between capacity coming online and revenue recognition would repeat the unused capacity charge pattern that weighed on second-quarter margins. Customer concentration in the AI data centre business, while not fully disclosed at product level, is likely elevated at this stage of the ramp, meaning any single hyperscaler or systems partner deferring orders would move the revenue line disproportionately.

There is also a mix risk. If the data centre business scales but the automotive and industrial recoveries remain uneven, the gross margin trajectory will disappoint even if headline revenue grows. Silicon carbide, in particular, has been a swing factor for the group’s margin structure, and continued pricing pressure in electric vehicle end markets could offset gains from higher-margin AI content. Finally, the macro and policy backdrop, including US-China trade posture and European industrial demand, remains a variable that neither guidance nor forward orders can fully insulate against.

The realistic base case is that STMicroelectronics has the products, the customers and now the funding to deliver on the raised data centre targets, but the path from the third-quarter guide to the “well above $2 billion” 2027 anchor will run through several quarters of execution that need to reconcile revenue growth with margin repair.

Key takeaways for investors and analysts covering STMicroelectronics after the Q3 guidance reset

• Third-quarter revenue guidance of $3.70 billion at midpoint fell below the $3.72 billion to $3.79 billion consensus range, driving a sharp intraday reaction.
• Second-quarter EBITDA of $679 million missed the roughly $797.7 million consensus and highlighted continued margin drag from automotive and industrial end markets.
• The data centre revenue ambition was raised to above $1 billion for 2026 and well above $2 billion for 2027, sharpening the AI infrastructure story.
• The stock remains up more than 110% year to date, which materially raised the expectations bar going into the release.
• Management guided to accelerating growth in the fourth quarter, positioning that print as the more decisive test of the recovery cadence.
• Peer signals from Texas Instruments, including its recent beat and raise, have tightened the standard by which analog and mixed-signal recoveries are judged.
• The June 2026 issuance of approximately $1.5 billion in convertible bonds provides funding for the capacity ramp but introduces conditional dilution.
• Broker commentary, including J.P. Morgan and Jefferies, points to gross margin evidence as the specific catalyst needed for a further rerating.
• Slower iPhone 18 ramp was cited by Jefferies as a possible contributor to the third-quarter guidance shortfall, adding a discrete near-term variable.
• The next measurable proof points are fourth-quarter revenue delivery, exit-rate gross margin and updated 2027 commentary at the year-end release.