EUR 850.00 For Business Accounts Only

Morningstar | STM Updated Star Rating from 09 Jan 2019

We attended STMicro's product demonstrations at CES and remain encouraged with the company's product breadth and diversity, especially in the industrial and automotive end markets. Although ST is an important supplier into Apple's iPhone, and Apple's revenue warning earlier this month may cause near-term hiccups for ST, we think the recent sell-off in ST's shares (down about 25% the past 3 months, including a 12% drop the day of Apple's warning) is overdone. We continue to foresee nice long-term growth opportunities for STMicro in 32 bit microcontrollers and imaging products like Time of Flight sensors aimed at industrial and Internet of Things applications. Although we still assign ST a no-moat rating, we continue to like the firm's focus on non-smartphone businesses and structurally higher profitability. We will maintain our fair value estimates of EUR 19 for ST's European shares and $22 for the firm's U.S. ADR and continue to view shares as undervalued.

Perhaps the best news for ST this week came from its subsequent press release that it has reached a multi-year agreement to buy up to $250 million of Silicon Carbide (SiC) wafers from Cree's Wolfspeed business. We view SiC-based power semiconductors as an exponential growth opportunity for ST in the years ahead, as ST is the main supplier of these parts into Tesla's electric vehicles today. ST is still targeting 30% of a $3 billion SiC semiconductor market by 2025, as compared with about $100 million in revenue today and virtually zero a couple of years ago, and we think the firm's aspirations are achievable with future EV auto production. The deal with Cree is especially encouraging as ST's rival in SiC, Infineon, previously struck a similar wafer supply agreement with Cree. While ST has other undisclosed SiC wafer suppliers, Cree appears to be the industry leader, and we view the deal as reducing the risk of ST facing supply constraints down the line.

Although we recognize that not every product demonstrated by ST, or any firm for that matter, will immediately bloom into a billion-dollar business, we were intrigued by a couple of new applications for the firm's existing products that may aid ST in future industrial and IoT applications. First, ST demonstrated artificial intelligence applications based on the firm's low power 32 bit microcontrollers (MCU), rather than traditional, higher horsepower applications processors. Whereas leading AI processor vendors, like Nvidia and Intel, are tackling the most complex of AI workloads, ST believes it is finding customer interest for embedded, low power, single use AI workloads to be run on MCUs. For further insight into AI processors, please refer to our August 2017 Technology Observer, "Accelerator: Rise of the Machine (and Deep) Learning Phenomenon."

One example ST showcased was the use of a 32 bit MCU specifically trained for food recognition so that a microwave or oven, for example, could identify the inserted food and set the proper timer accordingly. We note that Whirlpool announced a smart countertop oven at CES that uses a built-in camera and object recognition algorithm to identify food. While the chip vendors used by Whirlpool were unannounced, and it is possible that ST's AI advancements in MCUs might be matched by its tech-savvy rivals like Texas Instruments, Infineon, Microchip, and others, we think these types of AI advancements for MCUs bode well for the MCU industry as a whole. All in all, we anticipate that ST's MCU business will continue to see healthy demand for a whole host of applications in the long term.

ST's demonstrations also give us confidence in the sustainability of its recent success in Time of Flight sensors, most notably used in Apple's iPhone 3D sensing modules. The company is gearing current and older sensors toward industrial and Internet of Things applications, such as outfitting a robotic vacuum so it can detect various objects on the floor. Meanwhile, we suspect that ST is continuing to push the needle on tackling even more complex proximity sensing challenges longer term, such as "World Facing" sensing for augmented reality and virtual reality applications.