STMicroelectronics

STMicroelectronics reported second-quarter revenue of $3.49 billion, up 26% year over year and above the midpoint of guidance. ST forecasts third-quarter revenue of $3.70 billion, which would be up 16% year over year but below FactSet consensus estimates of $3.79 billion.

Why it matters: ST’s third-quarter revenue forecast was below our expectations, but we’re not alarmed. The firm will have softer sales into personal electronics, or PE, due to high memory prices and is facing some tight supply in its core end markets. Overall, ST is still seeing high demand for artificial intelligence products, especially in optical components. Industrial revenue has snapped back nicely from a prior cyclical downturn, and automotive revenue is picking back up. We’re encouraged that ST lifted its 2027 AI forecast to “well above $2 billion” and believe that ST has high visibility into these revenue engagements.

The bottom line: We maintain our $72 (EUR 62) fair value estimates for narrow-moat ST. Shares sold off about 14% on the revenue miss, which we think is punitive, but investors were likely seeing a stronger forecast after a massive run-up in many semis stocks in recent weeks. Shares now appear modestly undervalued to us, and we like ST’s exposure to AI and low earth orbit, or LEO, satellites. Lower PE sales may be a near-term headwind but doesn’t affect our long-term thesis.

Coming up: ST’s third-quarter revenue forecast was modestly below our expectations, as was its gross margin forecast of 37.0%, which we attribute to lower sales levels and some stubborn manufacturing reshaping costs. Still, we think ST’s longer-term gross margin expansion potential is on track. ST forecasts fourth-quarter revenue to be $4.0 billion, an acceleration over its third-quarter forecast but still a projection that was below our prior expectations, which we again attribute to PE weakness.

BLANK PAGEOn a year-over-year basis, ST expects close to 60% growth in its consumer electronics and computer peripherals end market, which houses both its AI and LEO satellite chip revenue. This forecast is a mild disappointment to us, but ST should largely make up for it in our model as it projects 85%-90% growth in the fourth quarter.

Across other end markets, on a year-over-year basis, ST expects industrial revenue to reach about 35% and 40% growth in the third and fourth quarters, respectively. Automotive should achieve low-teens growth for both quarters, as well as for the full year. PE revenue should be down about 5% in both the third and fourth quarters, again due to lower device production due to high memory prices. We don’t believe that ST has lost meaningful chip content within devices, most importantly at Apple.

On the bright side for ST, while it has some capacity constraints in general purpose semis, the firm believes it has sufficient capacity to satisfy the booming AI optical and LEO demand in the quarters ahead. ST’s book/bill ratio is nearly 2.0 for the entire company, meaning that orders well exceed ST’s capacity to ship. ST’s AI optical products are significantly above 2.0, again pointing to a robust backlog of demand. As with every upcycle, we anticipate that ST will ramp up capacity to meet this demand in the medium term, but these metrics bode well for ST's revenue growth and a healthy pricing environment in the near term and medium term.

Our fair value estimate is $72 per share, which implies 2026 and 2027 price/earnings of 55 times and 23 times, respectively.

STMicro navigated the covid pandemic well, as a global chip shortage enabled revenue to grow 25% in 2021, 26% in 2022, and 7% in 2023. It prospered from a strong pipeline of automotive and industrial design wins, offsetting well-known softness in PCs and smartphones. However, much of this growth was cyclical, as the company experienced a 23% revenue decline in 2024 and 11% in 2025 due to a severe inventory correction and sluggish macroeconomic activity in manufacturing.

We believe that STMicro will prosper from a cyclical recovery in chip demand as business conditions improve. It should also benefit from new growth drivers in AI and LEO satellites. We model 24% revenue growth in 2026 and and 21% in 2027, ahead of our prior estimates of 21% and 16%, respectively, thanks to stronger-than-anticipated demand in AI and LEO. We model 12%-13% growth in each of the next three years thereafter, ahead of our prior estimate of 9% midcycle growth, due to our improved expectations for AI and LEO chip demand in the long run. We account for potential growth in AI data center products, partially offset by the rising threat of increased competition from Chinese chipmakers (especially in microcontrollers and power/discrete semis) at the low end of the market. We now expect STMicro to exceed its 2028 revenue target of $18.0 billion, as outlined at its November 2024 capital markets day, as AI and LEO lift it above this target. Excluding the recent boost in demand for AI and LEO products, we would not have expected STMicro to reach this goal.

We believe that STMicro’s gross margin peaked at 48% in 2023 during the global chip shortage, as it fell to 34% in 2025. We anticipate a recovery as STMicro rightsizes its business and as revenue improves, reaching a 37% gross margin in 2026 and 44% in 2028, hitting the low end of the 44%-46% target from its 2024 capital markets day.

Along the same lines, we think non-GAAP operating margin peaked at 27.5% in 2023, as it dropped to 1.5% in 2025. We again anticipate a recovery and foresee STMicro getting back to 22% in 2028, just at the low end of its target of 22%-24%.

The majority of STMicro’s revenue and cash is in US dollars, which is also the currency used for financial reporting. As a European company, however, a hefty portion of its costs are in euros. At its November 2024 capital markets day, STMicro disclosed that 90% of revenue is denominated in US dollars, but 32% of cost of goods sold and 71% of operating expenses are denominated in euros. As a result, the firm is susceptible to exchange-rate fluctuations: A 1% change in the euro/dollar exchange rate may equate to an $8 million-$10 million change in operating income. Our valuation assumes a constant dollar/euro exchange rate for our five-year forecast.

We assign ST a narrow economic moat, based on intangible assets stemming from the firm's design expertise, as well as customer switching costs, as ST’s products are rarely ripped out once designed into a given electronics device.

We believe that leading broad-based chipmakers like ST (which we consider to be those well-versed in analog, mixed-signal, and microcontroller semis) benefit from favorable characteristics that lend themselves to economic moats. First, we believe that engineering talent (especially in analog) is difficult to come by, as greater emphasis is placed on digital chip improvements, and it often takes years to train up-and-coming analog engineers in the intricacies of chip designs. Thus, it is extremely difficult for startups to replicate the many years of analog expertise held by incumbents. Leading broad-based chipmakers also face stringent quality requirements in some end markets, such as the automotive industry, for example, where defects can only be tolerated as low as one part per million.

Furthermore, broad-based semis like analog and MCU parts tend to make up only a small portion of a product's bill of materials, so purchasing decisions tend to be based on performance rather than price, helping firms retain pricing power. A lower quality tolerance might be acceptable in devices with shorter useful lives, such as smartphones, PCs, and consumer gadgets, but moaty chipmakers tend to have limited exposure to these end markets.

Regarding customer switching costs, we believe that engineers loathe to swap out an analog or MCU from an existing design (again, only to save a few pennies on cost) because of the onerous re-design and re-testing costs associated with the switch. Further, customers also become accustomed to the software and development tools used to test and design a given product.

In our view, Exhibit A for our switching cost thesis is the global chip shortage of 2021 and 2022. Instead of finding untested suppliers or securing any reasonably performing chip they could get their hands on, automotive and industrial companies halted production of their gadgets until their key chip suppliers could get the proper, reliable, well-tested, well-designed parts into their customers' hands. We did not see any meaningful market share loss during the shortage, as customers again waited to receive the necessary chip, rather than re-design their products to take on a lower-quality chip instead.

In addition to these favorable moat sources, broad-based chips are trailing-edge products that don’t require the hefty investments needed by digital processors to move down Moore’s Law. In turn, broad-based chipmakers profit by having lower ongoing research and development and capital expenditure investments, which helps to contribute to healthy returns on capital for shareholders.

Further, buyers of analog semis typically don't demand smaller chips packed with more transistors, but rather, reliable products that deliver the desired accuracy and precision in power management or signal processing. Shrinking the chip might not necessarily enhance accuracy (and might even serve to reduce it), so analog chips tend to be made with lagging edge manufacturing techniques.

In discrete power semiconductors, ST competes in a fragmented market where moaty firms can carve out a cost advantage. We view these types of chips as relatively commoditized. Certainly, some firms have design expertise and reputational advantages over others, especially at higher voltages. However, Infineon and Onsemi, both of which have greater overall market share than ST, have sufficient economies of scale advantages to warrant narrow moat ratings in power semis based on cost advantage. We would not assign a cost advantage moat source to ST’s power semis business, or the business as a whole, at this time.

STMicro’s broad product portfolio enables the company to capitalize on new and emerging technologies, such as chips used in AI and LEO satellites.

Chipmakers with heavy automotive exposure, like STMicro, should profit from the secular trend toward more advanced electronics content in cars over the next few years.

If US-China trade tensions continue, Chinese chip buyers could decide to shift away from US analog and microcontroller suppliers and instead buy from neutral Europe-based firms like STMicro.

STMicro has a greater presence in certain types of lower-margin power semiconductors that, when combined with significant manufacturing capacity, may depress future returns relative to some of its US-based pure-play analog peers.

Apple is an important customer, and it would be a damaging blow if STMicro were to ever lose its content in the slate of future iPhone models.

China is supporting a growing ecosystem of broad-based chipmakers, which may lead to rising competition with STMicro in microcontrollers and power/discrete chips.