STMicroelectronics N.V. (STMPA) Earnings Call Transcript

Hi, everyone. I'm Alex Duval, and I'm the Head of Europe semis, comtech and hardware research in Europe at Goldman Sachs. Delighted to host Jean-Marc Chery, CEO of STMicro; and Lorenzo Grandi, CFO. Jean-Marc and Lorenzo, welcome to the call. Thank you very much for joining us.

Thank you.

Thank you.

Great. So I think we can get started straightaway on Q&A. Be really fantastic to get your perspective on where we are in the cycle, particularly as we think about automotive semis and industrial semis. Clearly, you talked about some very robust trends near term. But as we think about how things will progress towards the second half of the year, how should we be thinking about that semi cycle and in particular, any risks in terms of inventory correction or double ordering given how strong things are looking right now?

Yes. Okay. Maybe I am thinking to elaborate it, okay, and what we have seen. It is clear that, okay, starting Q3, we have seen across all the verticals, solid booking and revenue increasing. And it was, let's say, valid for automotive late in Q3, okay, starting end of the summer, valid on industry or in Asia, okay? And it was a still soft at this point of time in Western countries, so in Europe and in America. And on personal electronics, okay, I put aside, okay, the [indiscernible], okay, but it was quite solid on personal electronics related to 2 things. First of all, the introduction of new device, okay, from the well-known Apple. And also so the stay-at-home effect, okay? Clearly, accessories, okay, smartphone, tablets, okay, computer, the demand was really strong and sustained by this demand of stay at home. Then in Q4, what we have seen. Well, we have seen, let's say, 2, 3 phenomena. Well, first of all, okay, still strong and stronger than expected on overall personal electronics. Why? Because a great success of Apple. I guess, okay, when they have done their earnings announcement, it has been proven and demonstrated. And we have seen continuous increase of the booking on microcontroller, on power, on general purpose analog to address the industrial market in Asia, but as well in EMEA and in America this time. And then what we have seen, we have seen a peak -- a strong increase of the booking in the legacy automotive. And believe me, okay, between end of November and end of December, okay, the backlog we had in Q1 2021 end of November increased 50% end of December. So we have seen a very strong increase, sudden increase of the automotive demand, legacy, not specifically electrical car, okay, silicon carbide or digitalization, like, let's say, Mobileye or so, but really the legacy, peaking in November, December, loading heavily Q1 and Q2. And now, okay, we are coming to, let's say, a still variously booking, okay, with a very good book-to-bill, where, okay, we continue, okay, to grow on all the verticals according, okay, as a seasonality, according, okay, what should be a good 2021 year. In terms of inventory, we monitor very closely, okay, all the channel, definitively. So everywhere for all our product, the situation is very -- when I say healthy, it's more than healthy, it's very lean. So there is, at this stage, no inventory buildup in the channel. Well, and as far as I know, with the signal we receive from our customers, which are on emergency, okay, like an automotive carmaker, but now big player in industrial as well, basically, okay, there is no inventory in the field because, okay, believe me, okay, the number of, let's say, customer service call we have because, okay, we are on stretch situation to supply, okay, all the demand, okay, we conclude that at this stage, the inventory is very low. So our assumption that is definitively 2021 will be, let's say, a year of solid growth for the -- let's say, the semiconductor industry, stems, okay, either by megatrends, okay, which are well known, okay, electrification, digitalization of vehicles, okay. So still the stay-at-home because we do believe that even if, okay, we will move out progressively the pandemic, the remaining lifestyle in terms of work will remain hybrid one between traveling and using digital tools. So the stay-at-home effect will last, okay, for the next, let's say, years, I have to say, 2, 3 years. Then industrial, all the initiative around automation, okay, to take lessons after the COVID because we know that industry, which are well automatized, they resist better to the pandemic than the manual one. What is around power and energy control, all the corporate sustainability, CO2 control, okay, either because your own emission, but energy control is very important. So we see all the megatrend. Then there is, okay, some specific other seasonality when you have a new device introduction like a new smartphone. But there is this peak on automotive, okay, which has been very brutal, okay, in Q4 after a restart of Q3 that the industry has to digest, and our visibility is that the industry will digest it in 2, 3 quarters, and then we will come back, okay, more normal steady growth. But clearly, our assumption that automotive is following now a kind of V-shaped recovery. And it was not just scenario, okay, 2 months ago. 2 months ago, when the scenario was technical [indiscernible], and then, okay, we will come back to 2019 level not before '22 or '23. Now it is an accelerated path. So this is the situation we are seeing, Alexander.

Super helpful. And really appreciate you mapping out those megatrends on a multiyear basis. I guess one thing that we've heard a lot about is obviously the shortages related to semiconductors that go into automotive. I wondered if you could give a bit more color on sort of what your thoughts are there and sort of what it means for ST, maybe from a pricing and margin perspective. And then also how long it could take for the industry to be able to catch up a little bit in terms of the foundries and so forth. What's your sort of perspective on that?

Clearly, okay, when you have this, let's say, sudden acceleration on top of the situation where the supply chain was pretty loaded, and basically, before Q3, we designed our investment plan for 2021 and our peers as well, okay, and the foundry as well. But then, okay, as a matter of fact, it is clear that the capacity, which has been planned for Q4 and planned for the first half of 2021 are not important enough to -- let's say, to fulfill the requested demand by all the verticals. People are reacting, okay. But I guess everybody remember the announcement of TSMC, and they surprise a little bit the market by impact on CapEx, $25 billion to $28 billion. So it is a way to react, okay, to this demand. We have done the same at our scale, definitively. So we increased above our model. We disclosed our model for the time being because the strategic initiative is in the range of $1.6 billion to $1.7 billion. We have increased, okay, $300 million more, so $1.8 billion to $2 billion. It is a reaction, okay, of this demand. But this delta in CapEx will pay back only in H2 this year. The lead time of equipment are basically 6 to 9 months. So we will have the payback only in H2. So that's the reason why we do believe that from now up to end of Q2, the industry will have to manage at a lean inventory level. We will have to manage, okay, maybe sometime through allocation, okay? This is something the industry was used to do many years ago. So we will have to manage through allocation for a while. And most likely, okay, Q3, Q4, the situation will go back to normal. And again, for ST, this is exactly the situation we are facing. So when we elaborate our, let's say, what we call the sales and operating plan constraints, it is well above our capacity and plant capacity. So we have to decide, okay, how we allocate by vertical. We try to act consistently with our strategy, consistently with our customer engagement. We do not want, okay, to change our -- we don't want to be opportunistic. We want to continue to address automotive, industrial, personal electronics consistently, but in a way that we have to manage allocation. Well, about price, okay, clearly, there is inflation on cost because, okay, the supply chain is under siege. Yes, okay, we ask our customers price increase, but let's say, widely in one shot, okay, and not playing opportunistically. Fairly, okay? So this is what we are doing. And I do believe that when I will communicate in April, our revenue for the year and the perspective also for the gross margin. I expect, okay, 2021 to be an accelerated path versus, okay, what we told to the market in December at our Capital Market Day exercise.

Great. That's super helpful. So sort of allocation in lean inventory for a couple of quarters and then some more normalization in the back part of the year. I guess one of those sort of exciting drivers, I guess, in terms of growth is clearly automotive for STMicro. And just thinking about the year ahead, I wondered if you could decompose a little bit the different aspects that drive the growth. We've obviously got the recovery in the number of auto units being sold. Then there's potentially pricing, which could come on top. And then you've, of course, got some of these content drivers like electrification but also ADAS. So I wondered if you could talk a little bit to those and help us just think about how they drive the growth this year and the extent to which one or other of them could have a bit of an upside surprise.

Clearly, where we are not so surprised and we will grow in 2021 as expected, it is in the field of, let's say, silicon carbide. So I disclosed the plan between $450 million to $500 million, and this year, okay, with newcomer entering on top of our legacy customer. And here, basically, it's absolutely not a surprise, there is no, let's say, cut or increase taken by surprise. So we continue to execute properly. And we see the transformation of this adoption of silicon carbide going at right speed. Basically, it's very similar on the partnership we have with Mobileye, so on ADAS. We are supplying according their demand. They have built an adequate level of inventory so to protect them from the fluctuation. And so this is going at the right speed. Well, where we see clearly difference, it is on the legacy. And on the legacy, we see 2 things. We see, let's say, some content increase and, let's say, technology change. So we move from, let's say -- we see system electronic changing. So with a more sophisticated microcontroller, more sophisticated ASIC driver, okay, or radar or this kind of stuff. And hear from our Tier 1. And based on the increased, let's say, sales and vehicle to produce, plus inventory buildup because clearly, the reason why, okay, also the industry is facing this line down announced everywhere. It's because, okay, on the top of the supply chain, so carmaker in Tier 1, looks like the inventory was basically at 0. And we know that the supply chain is very complex and to manage plant the all inventories is very difficult. So clearly, on this automotive industry, there is the megatrend running at expected speed and driving the growth of ST. So you see about USD 500 million on silicon carbide. It is half of our target of USD 1 billion. So I am very happy with that. Newcomer coming, I am very happy with that. Number of program increasing, okay, as well. ADAS running at the speed we expected. But then we have this new scenario, where, clearly, on the number of vehicles, we come back, okay, to 2019, the number of vehicle basically next year. And this year, the perception we have is equivalent. Why? Because there is 88 million vehicle to produce, plus inventory to build up to come back at a normal level. And next year, okay, we will come back to the normal trend. So this is what is driving our growth on automotive. Well, and then after, there is some sporadic event, but I do not consider it as a trend. Yes, okay, look at the microcontroller, as an example, for automotive. Now it's clear that the fact we have our own fab enable us to react faster than some competitors, which are using foundry because in foundry, okay, of course, they are in competition with the consumer electronics, okay, which is very, very strong at foundry level. And us, okay, we have always the capability to allocate, to reduce inventory and so on and so forth. So yes, we have seen some sporadic increase on microcontroller, as an example, in order, okay, to compensate, okay competition failure. But I do not consider it as a structural trend, okay? I consider it more as a, let's say, sporadic point because of the supply chain situation.

That's great. And you touched on silicon carbide, where, obviously, ST has been very much a front-runner in terms of the automotive side of things. I wondered if you could talk a little bit to how you feel about your position there on auto with silicon carbide versus where you were maybe a year ago. I'm thinking about things like maybe the reliability, the efficiency of your product. And then also, clearly, a part of the story on silicon carbide that's interesting is around things like renewables and even sort of in a factory context in industrial. So I'd be curious how you think your position is there, again, relative to where you were, say, a year ago on silicon carbide.

Now -- basically, okay, I have 3 reference point, okay, for silicon carbide I share with you each time I communicate. This is more our revenue. Then, okay, it's a breakdown of our revenue. But okay, here, I cannot be too much accurate because it is disclosing information on customer that I must avoid. Then the second KPI is the number of program awarded and number of customers engage. And then the third point is about our initiative with Norstel, okay, on the raw material, okay? So this is basically the 3, let's say, major point I share with you. And why? Because, okay, I disclosed an ambition, we say we want to be, in 2025, a player on silicon -- major player on silicon carbide inside the electrification of vehicle with $1 billion revenue. We want to have 30% market share we assess. We do believe that 30% market share is sustainable. We don't play for more. I guess I am right when I see the base of automotive to have 100% share in 2025 on the automotive market is not a lot of peaceful. So -- and then, okay, and we want to be well balanced between industrial and automotive market. And last but not the least, we want to supply 40% internally with our silicon carbide raw material. So where we are at this stage. Revenue, I disclosed, so $450 million, $500 million. Okay, basically, legacy customer plus newcomer. Newcomer are about 20% [ or more. ] So this is a demonstration that, okay, all the programs awarded and won during the past few years are starting to pay back on top of the legacy customer. Number of programs engaged and customer based, 62 programs awarded, 66 customer, well spread, half half, industrial and automotive. In automotive, well spread across all the geographies, so America, Europe and Asia. Of course, the main volume will be automotive because, okay, we know that industrial applications are more spread, fragmented, but we like, okay, this kind of market. So here as well, on these metrics, okay, we are, let's say, on time. And believe me, okay, we can predict, okay, the revenue based on this program awarded and our assumption in term of electrical cars and so on and so forth. So I can forecast that we are well on track, okay, to achieve this $1 billion. Then on the raw material. Raw material, we -- at the end of 2020, we are using, okay, from Norstel, okay, facilities and fab that we have completely reworked already some material in production. And our raw material for Norstel is marginal definitively, but the -- let's say, the performance of our raw material in terms of number of wafer per ingot and in terms of performance of our device enabled by this raw material is at the level of the state-of-the-art reference, means Cree. Then we continue to cooperate, okay, with Cree definitely, and we see Crystal, okay, we continue to cooperate with them. Our ambition will be to start by end of 2022, early 2023, mass production in our new plant fab on raw material in Catania, okay? This is our plan. And to grow, okay, from, let's say, 2022 end to 2025 up to a maximum of 40% of our internal need. Most likely, we will certainly start, okay, this raw material still in 150-millimeter, but also equipment acquired will be 200-millimeter compatible. And when we will consider it is the right time to convert, we will convert. So you see when I monitor all the key data point, so raw material, number of programs, balance in terms of customer base and revenue and newcomer entry. Honestly, okay, we are really on track to deliver our strategic objective at 5 years. And now, then, okay, we see competition coming. So Cree is there, okay, [indiscernible] is there. [indiscernible], okay, now is coming on as well. We know Microchip with Microsemi is quite active as well. But it's normal, okay? We never pretend, okay, that we will, let's say, be the dominant of the market. And also, we said that we attack the market with a mix-and-match strategy, both silicon carbide MOSFET, but GaN MOSFET as well and IGBT. So IGBT, this year, we will grow up, okay? We do believe we can increase by, let's say, at least 50% of our revenue on IGBT. And GaN will start to come in 2 years from now. So this is where we are. And every quarter, I communicate consistently on it.

That's super helpful and really appreciate those proof points as you move forward that you shared with us. And I guess just on this point about IGBT, it'd be really great to get your perspective there as well. Clearly, you've been a front-runner on silicon carbide. So sort of given the efficiency advantages on silicon carbide, what is the advantage of leaning in on IGBT? And then secondly, sort of how do you feel about your progress ramping up there? And then I've got a follow-up on that as well.

On the electrical car, first of all, okay, you will have, let's say, a different kind of architecture, if you have, let's say, 2 electrical engine or 4 electrical engine, 2 inverter, main inverter or front, back and so on and so forth. So since the beginning, okay, we understood that because of -- because our customer also ask us to be capable to offer a mix-and-match solution. So it's point number one. Then also in terms of business model, okay, you will have business model Tesla like. You will have business model, okay, legacy, okay, with a carmaker Tier 1 and Tier 2 like us. And you will have a new business model, okay, like for maker, okay, you will have EMS, okay, designer of a car, EMS, and you will have semiconductor, okay, supplying EMS, who will supply platform, okay, standard platform to carmaker and mainly, okay, in China. We know this kind of initiative. And here, you must be, let's say, capable to provide in a very flexible manner full solution, okay? So silicon carbide MOSFET, mix-and-match with IGBT or full IGBT, and then later on, okay, GaN. As we want to play broadly and we want to play as a leader, and we want to continue to grow organically, okay, we address everything. So we have developed our IGBT. Now we do believe that, okay, we closed the gap in terms of technology, in terms of product. We have developed a strong strategy module. So now we demonstrated our capability to develop custom design module, Tesla like. We have developed more application-specific standard. So what we call ACEPACK, and we have internal capability and external capacity, okay, with [indiscernible]. We are developing strategic agreement with module makers in order, okay, to define standard and address the market. And sometimes, okay, we provide, okay, a [indiscernible]. So you see we have the full spectrum of business model on technology portfolio. Why? Because consistently, we want to be a broad range player, and now it is paying back.

That's a fascinating perspective on how these platforms will evolve. Really appreciate that. If we sort of stay on this topic of China, which you touched on, I'd just be quite curious about how you think things could evolve in terms of competition. We see more news flow about IGBT and MOSFET competition coming out of some of those Chinese players. And I'm just curious to what degree that's a challenge that you'll need to meet, particularly given some of this localization trend that we see in the region.