GLOBALFOUNDRIES Inc. (GFS) Earnings Call Transcript & Summary

Okay. Great. Thanks, everyone. Good morning, good afternoon, wherever you are, listening in online. I'm Chris Danely, the semiconductor analyst here at Citigroup. It's our pleasure to have one of our top picks GLOBALFOUNDRIES. Thomas Caulfield, the President and CEO, is here. Why is it our topic? We like companies with secularly increasing margins and earnings, especially given the industry-leading portfolio they have for both analog in terms of automotive and industrial and also the wireless markets. It's been a pretty remarkable turnaround story since they came to the public markets. So I want to say it's only been 1.5 year or 2 years. And the main architect of it is sitting right here to the left of me, Tom. So thanks, again, for coming, Tom. We really appreciate your time.

Thanks, Chris. October 2021, but who's counting?

Yes, exactly, almost a 2-year anniversary. You guys are fairly broad-based. So I guess maybe we could just kick it off. Give us your view on the semi industry and the various end markets, I can run them down, but how do you feel about the semi industry these days?

So let's go back to what we talked about at the beginning of the year and what we've been delivering against. We said that we have very modest growth quarter-on-quarter. We delivered that in 2Q. We gave guidance for Q3. And so we're on track to go do that. Now having said that, that's still a year that we said we'd be down in the mid- to high single digits. As we said in the third quarter call, we're probably the higher end of that down mid-single digits, and that's consistent with the rest of the industry. So how do I feel? I feel like there are segments that are really going well and will continue to go well. And some of it is because of the disruptive nature of going from an internal combustion engine to autonomous connected and electrified cars. So that market is very strong for us. It's continuing to grow this year. 2020, it was under $100 million business; last year, it grew to $375 million; this year, we've said it -- we're confirming that it's -- we're going to hit our heads on $1 billion worth of order revenue. And that's a pipeline of growth that will continue for us. Now on the other end of this, the consumer side of our business is still for all semi is still struggling. Whatever spending is going on in the consumer side, it seems to be another discretionary spend in like travel and things of that nature. High interest rates obviously with inflation are -- always give trepidation for a discretionary spending. So for us, what do I see going forward is in new strength in automotive for our business. And we will go as kind of the macro economy goes. And for me, it's what I see versus what I believe. What I believe is we've been here before. I believe inventory will eventually get worked out of the channel, at least for the first time, while we saw inventory at the OSAT and OEM level decrease in 2Q with only moderate increases in the rest of the supply chain, which is good news because we're not going to go anywhere until inventory normalizes. And so I believe that we will eventually get ourselves back into a growth industry. But what I tell you I see right now, I'm no smarter than anybody else or the economists, which is the world economy look like that will get that catalyst going for broad-based demand in our industry.

Sure. So let's dig into those end markets, especially the ones you specialize in, auto, industrial, wireless and consumer. I guess let's start with the most positive or the best one, the automotive end market. You mentioned continuing push for more EVs. I mean even hybrid vehicles have a lot more semi content. Do you -- I mean, do you see any sort of correction there? Or do you think that the content growth can power us through the automotive end market?

I can tell you we see based on the our design wins and the sockets we're supplying to. One, a lot of this business was one between 2019 and today, it just takes a while for it to ramp. And so a lot of that was for the next-generation models. It's not legacy models. So that helps the resiliency of where we're growing that $20 million in 2020 that grows to $1 billion today, it's all a concerted effort when we pivoted the company and started to focus on these end markets where our differentiation matter. A lot of it is also the different applications we play in. There's a broad range of microcontrollers from a range of applications in automobiles. We do some of the precision battery management. Our power technology is really, really differentiated. We do a number of radar chips for sensing and a host of other applications in the automobile. And so as cars become more about a connected intelligent edge device that just happens to move, the more it plays to the strength of our technology.

Great. Switching gears a little bit to another large end market for you, the industrial end market. How are you feeling about industrial? Maybe just refresh us on what you guys said on the last conference call.

Yes. In home and industrial IoT, the kind of the home consumer part of that is slower than in a way we would like, but it's being offset by strength in industrial, and we continue to see that throughout the rest of this year.

Any sense of when you think the excess inventory in the industrial IoT market is going to be cleared. I'm asking every company that you'd be a surprised the most popular answer is "we have no idea." and I say, "How will you know when things are better." They say, "We'll get purchase orders."

Look, some of us that have been around for a while, we start to look for signals that we see. So in the beginning of the downturn, you start to see push outs and you stop seeing kind of last-minute orders, right, and say, okay, this is starting to get. Then when you start to hit the steady state, what do you see? You don't see pushouts, so it is kind of planned things. So we don't see that anymore. What we're building, it's shipped to, it's agreed to. POs come in an orderly fashion, you don't send your [ kit ] squad sales team how to get POs, where are the POs. And then you start to see a little bit of the unexpected order came in. Hey, PO came in unexpectedly for 5,000 wafers, right, for a particular part. And that tells me -- that's the kind of signal to me that says, okay, we're at least normalized now. We're starting to see the indications of what looks like people starting to fill back in their business. Now the rated pace of that is the question mark, is that what we're seeing this quarter and then it slows down again. But at least that's the a little bit we're seeing. And so for me, this is about to what you want to believe versus what you see. I believe it's just a matter of time. But what I see is little things like that, that give me hope that maybe we're going to start to see this turn around. But as the Fed keeps raising interest rates, this inflation get moderated or become worse and put more uncertainty, are there more geopolitical things that start to create uncertainty for everybody. Those are the overlays that we have to factor into this. It is what I'm telling you, I'm no smarter than anybody else who sat in this chair. I have no idea other than when I see and what we plan for it.

Sure. And then switching gears to one of your larger end markets - wireless. When I talk to my industry contacts out there. And we've had several of your customers presenting, I would say the two areas that they really like about GF are your analog capabilities and especially in wireless, your RF capabilities. So maybe talk a little bit about that? And then what are your thoughts on the wireless market when you see some green shoots there?

So when I think the wireless you're talking about, I am thinking of smart mobile devices, so yes, that is a really important part of our differentiation. There isn't a single handset in the world whether it's Android based or OS based that doesn't have GF's content somewhere in that front-end module, whether it's LNA, a PA, filter switches, some combination thereof, we supply to all players, and they use that technology. Now others have leveraged it more than some in their deployment of it. The good news is even in a market where smart mobile devices is really at a low point because we play in Tier 1 phone -- disproportionately in the Tier 1 handset market and disproportionately with the non-Android-based phones, it plays to seeing less of an impact to our business. But the reason we have this business is because of our differentiation. We've recently announced -- actually, it was last week, the launch of our we call our 9SW platform, which builds on 8SW for better performance, better power. And we already have customers doing their first tape-outs to do test chips around this technology for next-generation handsets. So we keep innovating on this technology and trying to drive as much integration of these key elements of switches, filters, LNAs, PAs as much as possible and on a single technology.

Yes. You -- I mean you go back at GF a long time. I think the RF technology goes back several years.

It goes -- it was part of our IBM acquisition in 2015.

Yes. Let's talk a little bit about ASPs, maybe on the foundry side. So I noticed that in Q2, they dipped a little bit. Was that a mix issue? Can you talk about why that happened?

Look, the real view of looking at ASPs is their pricing environment changing. And the answer for GF is when you are 66%, 2/3 of your revenue single source, 90% of your design wins are single source business, there's no pricing sensitivity, right? The demand is there, it's not demand. There's no demand. And so when you see that fluctuation, it's mostly mix. We have a range of technologies from, call it, 40 mass steps to 60 to 70 mass steps. So depending on the complexity of the device with the technology node, you'll get different pricing points. And so the best measure for how we're doing is how our gross margins are, because we can have a product that's priced lower than another one to make more margin because the complexity of build of it is less, but it has a different level of differentiation. So first order, our pricing environment because of the nature of our business is very strong. You'll see some minor fluctuation in that because it's all mix related. But the key is to keep driving better margins through remixing our business not on ASP, but as much as accretion to gross margins.

Great. And then how do you see pricing going forward? Any change there?

I would tell you what -- how we think about it is what does our future business look like? So how do we judge that? We design -- we judge that with our design win funnel. And so when we look at whether it's first half of this year or just even the most recent quarter, we will not take design wins that are not in aggregate accretive to our long-term model because we'll never get to our long-term model, if we're not -- the future business is not going to be richer in leveraging our differentiation. And so the confidence that we build each and every quarter about where we stand against our long-term operating model is we don't need to look any further than what is our design win pipeline and what we achieved in that design win pipeline.

Yes. We had, I think, 3 or 4 of your top 10 customers here. They were all asked about foundry pricing. All of them said firmed up going forward. So you guys must be doing something right. That was -- let's talk about the positive comments.

One, I think there's a little bit about differentiation and creating value and capturing value. And the second thing is about we, as an industry, need to invest to create this capacity that doesn't exist today. That investment needs a return. If we've lived for the longest time on capacity that was fully depreciated and it served the needs of the market, that created one pricing structure. If now we need to invest, we need to make that investment and get a return. Now how we make the investment and how we eliminate inefficiencies in ecosystems? How we're going to play this game in the future to minimize those impacts? So if we make an investment, we need to mark up our pricing to comprehend that investment on top of the value we create in our differentiation of the product we ship. But the worst thing to do that is to send that through a long supply chain where everybody marks it up without adding value to it and finding creative models where we eliminate economic models with partnerships with end customers as well as near customers, fabless customers to go create the best economics for everybody is the way we're going to win.

Yes. I mean we've talked about this. The one thing that drives me bonkers is you do get certain investors that say, "Oh, GF is the same as UMC and I say UMC is a second source TSMC." you guys are almost all single sourced. And if you look at the growth in your customer base, these are leading-edge analog companies that are coming to you, largest wireless company, wireless chip company in the world is coming to U.S. because you guys have good products. So anyway, that's my plug.

You should be in this seat.

One other thing I want to talk about and you guys have been one of the pioneers in terms of these LTAs or long-term contracts. There is a big part of GF, and I think it is providing some cushion in these turbulent times. Maybe talk about how this generated? Was this customers coming to you? Is this sort of a joint venture? And how do we see LTAs holding up in these tumultuous times?

Yes. Look, if you took a step back and you were a student of this industry, you would have looked at how we ran this industry was, no one in the industry in the world would do this. One company makes an investment for capacity that's special to their customers with only one commitment, the foundry making the investment and hoping that someone would use it. No other industry would ever work that way. So what felt like unorthodox when we began this in 2021 to create long-term agreements was actually just very sound business. And it turned out a lot of this certainty we were looking for our customers wanted as well. So let's talk about what happened. We find ourselves in 2021 where the industry is way beyond in creating the capacity it needed for the demand. If any one customer had asked me for capacity, we probably could have accommodated it, but they all did. And so we needed to create a different framework by which we would operate together so that we could make the investments with as much confidence as both parties could have recognizing that if there's a downturn, these agreements would least govern how we can work together to share in whatever pain there was rather than having it disproportionate to the company that made the investment versus the one that did not make the investment. And so that was the beginning of it. And I would tell you the first LTAs we signed, we were more of the instigator. We were the ones driving for it. Hey, to make that $2 billion investment, we're going to need some certainty on that. By the time, and that was first quarter, early 2Q of 2021, by the time we hit this summer, customers were driving us aggressively for long-term agreements. Why? They saw a world where they needed more than just a surety of supply. They wanted to know they had a surety of pricing as well. They needed fixed volumes, fixed pricing, fixed duration. They couldn't plan their business. If they knew today, they were paying this, 4 months later, they come in with a PO and say, "Oh, there's a new price now, right, because the world has changed." They wanted certainty to predict their business. And a lot of their commentary was, "If I know what I have, I can plan my profitability in my business around that." If I'm -- every quarter wondering, I don't want to be maybe in the more of the memory business of arbitraging pricing depending on supply and demand. And so those LTAs have served our customers in this downturn as well as it served us because it's preserved for them the ability to sit and talk and say, how are we going to work this together? Most recently, a great customer of ours, Cirrus Logic, in their earnings call talk about how they worked with us in their LTA to shift the capacity from one product set to another in a different node. And we were able to make sure they were building products they needed to meet demand and not letting the LTA get in the way of doing the right thing. A number of our customers are sitting with us now and how do we preserve the economic value. It's not saying it was x number of dollars over 3 years becomes x number of dollars over 4 years. It's how do we make that fourth year even richer with different designs that you can design into this thing. So it just creates, again, a framework that depending on the depth and the length of the cycle, how do we get through it where we're both kept whole as much as possible rather than being disproportionate on one side versus the other. And believe it or not, the partnerships that we created, creating the LTAs are being validated in the conversations we're having now in the downturn where it is a lot harder. Do you guys have a -- sort of by the way, it's $27.5 billion of revenue to date of LTA signed with about $5 billion of customer participation in some form or the other of the CapEx required to create that capacity? So it's been very successful.

Yes. You just took my next question, so I can check that off. One other one on the LTAs is, is there a limit to the sort of backlog coverage that you will have from an LTA and maybe talk about, let's just take the cycle out, but let's say, for the next, I don't know, 1.5 years or 2024 what sort of LTA coverage are we looking at or striving for that...

Yes, let's talk about, I think, the right model on that because you're always playing around the edges in that, if you are under a look. The last thing -- in the beginning, it becomes a little bit like, wow, let's get everything we -- let's get 100% certainty baked into our business. So let's -- and that's the wrong model because then you don't have the ability to remix, you have the ability to respond to opportunities to even some of the customers that maybe didn't want to commit to their complete demand needs, but wanted to know that if it came in, was there a way to solution it. So I would tell you, depending on where you are in time, so maybe 1 year out, you want 85% to 90% certainty in that business. But 3 years out, you may only want 60% coverage of your capacity because you want that ability to remix the business and to work with customers for new design wins. The worst thing you can do is to say, "I love to win this business if we may, I just want a capacity for you. And so we're constantly fine-tuning on what horizon do we want that certainty of demand vis-a-vis the capacity.

And then I think Dave mentioned on the last call that some of the LTAs get renegotiated. Like Cirrus Logic where you just shipped them. So how do you guys approach that? Is it only on volume? Is there some on price? What's the -- if you could give us some insight into the process of that.

There's a number of different levers that can be used and customers get to choose. Some of them is, how do I remix to a different corridor like that one. I think another key lever is, oh, how do I preserve the longer-term economic value. I'll do even more business but let's spread it over a longer period of time. And then in some cases, maybe some customers got really well ahead of what they thought they would need different elements took place in the marketplace, and they choose to pay underutilization fee. That's their choice, to take it as a onetime charge. And so those are the elements and it doesn't mean you do one versus the other two. Sometimes it's a combination of these. But those are the key ingredients of how to go work through this. The most important ones for us are the remixing and creating longer-term business opportunity, holding true to the economics of the original deal, not spreading the revenue over more years, but making more revenue over a longer period of time.

And it sounds like there's a number of ways to cushion the blow as well. Just on the competitive front, I mentioned UMC earlier. So hopefully, that dispels that, although probably not given how much I've flapped my gums about that over the last year. You see China launching all these boundaries out there really all over semis. How do you assess the competitive environment from them? What's the threat level for you there?

I think the common misconception is because I manufacture, I can be a foundry. Now you can't be a foundry if you don't manufacture, but it's like a fraction of what you need to do. You need to have the capabilities to actually have an enablement for customers to design your technology. So it's your process design kits, it's your standard sales, your libraries, your foundational IP, your complex IP on a platform that's built for applications that matter to customers. So buying equipment is the easy part is to have a checkbook, right, hooking them up, running them in an efficient and effective manner on technologies that matter. And so you just don't wake up one day and say, "I want to be a manufacturer, I'm going to get myself into the manufacturing business." This is something that's a long journey. And so we see a lot of that. There's a lot of desire to create manufacturing, but not necessarily a lot of what it takes to actually create and use that capacity. And for us, we will stay as long as we play in a very big market opportunity, our SAM is big and growing compared to the size of our revenue. We will pick and choose what we call priority battlegrounds, the areas of the SAM where differentiation matters that we enable our customers to win in the marketplace. And as long as we stay true to that, we're not going to get into the fray of more commoditized capacity that's being put on.

Yes. And let's take the other side of that, a question that I get asked a lot and that pops up as the geopolitical issue of Taiwan. You guys don't have anything in Taiwan. You're all U.S, Dresden and Singapore. Do you ever see customers calling you up saying, "Hey, we're a little nervous about this Taiwan situation. Maybe we should give you guys a try."

So let's go a little bit back to your China question, and then I'll fast forward too. So dynamic in China. China is putting a lot of capacity on. Why? It's For China Built In China, right? Again, they're still going to have to figure out how to make something to build. One of the fastest-moving engagements we have are fabless companies in China and in Taiwan, wanting sources out of Taiwan and China. They want to be global fabless companies. And therefore, they can't worry about a geopolitical issue or a sanction of some sort, keeping them from shipping to their global customers. In fact, global customers, right, would be hesitant to bet their products if it can only be built in a region that someday could have -- and so these customers are working with us to say, maybe enough for the Made In China For China, how about Made In GLOBALFOUNDRIES for Rest Of World. And so we're seeing, in part of the response to these geopolitical concerns, right, these fabless companies coming to us. And we've already, to date, have one design win in process, another one already in first articles of product shipping and yielding and starting to think about how that will ramp. And again, it's for these customers, these fabless customers of ours to be able to ship worldwide to their customers.

Yes. And to dovetail on that, we had the CHIPS Act folks here presenting this morning, and they talked about wanting to increase the scale of U.S. manufacturing. Something tells me you guys might have been a critical element of crafting the CHIPS Act. Any insights you can share there on what the CHIPS Act could potentially mean to [indiscernible]?

Yes. So I think this is the bigger point. It's not just the CHIPS Act in the U.S. It's the -- it's the European version.

We'll get to that one, too.

Yes, so look, it comes to a fundamental issue with the industry and maybe a little bit of the fundamental value proposition that GF was first conceived. We can see we call GLOBALFOUNDRIES and the tagline for the company was it's the world's first truly global foundry, right, to create supply chain resiliency and it was met with, who cares. It was kind of an idea ahead of its time. But it did make a lot of sense. Forget geopolitical issues, there's geological issues or any kind of issues, anytime something so important that semiconductors to the world economy are so concentrated in one part of the world. A lot of risk goes in single points of failure independent of what's the origin of that risk. And so when chip crisis like 2021 happens and the world becomes a lot more aware of what a semiconductor is and how important it is. And then learning where it's all built, it starts to create, okay, we have an issue we need to do something about. And with the U.S. and Europe is trying to do is create through, I call them, co-investments, incentives is to create the economic conditions by which manufacturing in the U.S. can be globally competitive with the rest of the world. So that's the essence that -- what is GF going to do? There's -- we will continue to leverage the fact that we have a global footprint, and we will continue to grow that global footprint where we already have existing fab sites. So we have a program in Singapore. It started, first pilot on the ground, summer of 2021, first tool moved in June of 2022. Next week, we'll be in Singapore, celebrating first wafer shipped out of the expansion we started in 2021 out of that facility. Over the last couple of years, we've invested over $800 million in our Fab 8 facility in upstate New York to create capacity there. And over the last number of years, we've taken Dresden from like 350,000 wafers data and 50,000 wafers. And we will continue to do that. In the U.S. when we decide to grow that footprint, a clear part of making sure the economics work are the incentives that come from the government. So yes, we are a part of that program. Remember, 2 parts, ITC, 25% tax credit in the chips, which is a 5% to 15% coverage. And so it's important to get both of those, but not lose sight that the ITC already exists, you can already apply for credits against that if you spent CapEx since last August. And so it's an important part of taking a desire and a will U.S. government. I want more semiconductor manufacturer in the U.S. and turning it into what's required to create the right economics to attract that. It's a lot easier for companies that already exist in the U.S., especially if they already have facilities to expand than it is to go greenfield because the economics of greenfield expansion are really, really expensive.

You're probably one of the better guys to ask. Do you have any idea when we might get to see some clarity on the CHIPS Act funds getting dispersed? Do you think first half of next year is a reasonable expectation for where we could see some announcements?

I think this ties back to the 3 ingredients that will make the CHIPS Act successful. The first one is broad-based distribution of funding. That's why the ITC works in conjunction so that, that $37 billion of the $54 billion that goes to manufacturing has a big impact. That means it has to start with manufacturers of the wall wafers producing in the U.S., right? And it means it goes through essential chips like GF creates. It goes to single-digit nanometer that other companies make. It goes to packaging so that you have an end-to-end holistic supply chain. So one, make sure that the money is not too concentrated or you didn't fix the supply chain, you fixed one link in the chain. The second thing is probably the most important; public partnership, public-private partnerships will always come under a high degree of scrutiny because you're using other people's money to have payment. Build an empty fab and you will see public outrage. So I think some of the timing you're seeing is timed to a market recovery where we need that capability to add. And if we go too fast and build empty factories, all the right intentions. We'll come into a lot of criticism. So I think the timing is just fine right now. If we were doing this in 2021, right, when there was clamoring, they would be a lot more aggressive in getting these programs signed up and some capacity being built. But right now, we're in an industry that's not suffering from an undersupply. It's more of a demand story. And then, of course, the last ingredient is making sure we have the talent and doing the right thing in the talent pipeline. So I think [ it is a clear reminder that ] her team being very thoughtful, making sure we can spend that kind of money that comes from taxpayers that it's done in a thoughtful way, and no one looks back and criticizes how they did it. And I think a lot of the timing has to do with demand than it does with the application price.

That makes a lot of sense. And then you touched on the, I guess, the European CHIPS Act. Anything you can share? You guys do have a pretty large facility in Dresden, the opportunity to pick up some.

Well, it's something we've already announced and it comes back to this idea of, we will never do greenfield. We don't have to. We've spent 10 years creating a global footprint that the others will try to replicate now. Our program with a partnership with ST in Crolles, France. That was part of the European CHIPS Act, it was well funded by the French government to create the types of capacity the central chips we make in partnership in Crolles. Again, that's a program that's in place and the timing at which we will create that capacity is all about certainty of demand for our customers. But knowing that, that program is there for us to go take advantage of. And the same thing will happen when we need to create new capacity in Dresden, Germany, the [indiscernible] local government, the German government's fully behind GLOBALFOUNDRIES. We've been a great partner for over 30 years now.

Yes. So let's tie that into something you said earlier about the long-term margin and growth targets. Maybe just give us a sort of a rundown of the capacity expansion plans at GF with your various locations and maybe the time lines to achieve the margin and the growth targets? And has this correction done anything to alter those timelines?

Yes. I think the conviction around getting there is only higher now. And I think part of that comes from is our margins have held steady even though we're utilized, call it, between 80% and 85%, right. And that's a big deal because for every 5 points of utilization, a point or 2 of margin. So scale is really important for all businesses. Our capacity in 2020 was 2 million wafers, 300-millimeter equivalents. By the time we get to 2024, that will go to 3 million wafers. And so that buys us a lot of headroom for revenue growth. You talked about our ASPs before, right? Rough number was $3,000, right? So you take $3,000 times $3 million, there's $9 billion of revenue off of where we are today and you put another 10% of non-wafer revenue that's kind of in our mix. So we have already in the investments we've made a pretty good growth trajectory. We're not having to spend a lot of CapEx in the balance of this year going into next year. And then from there, we will plan our business is how do we go beyond that $10 billion or so of revenue that we've already invested on. We'll do that in a very thoughtful manner; certainty, durability and profitability. Certainty of demand, durable markets where we have differentiation; and the profitability is what's the economic model that creates the right return on those investments for GF.

By the way, when do you think we can expect to see the first products from the Crolles partnership with ST? And will it be on a certain area or a certain end market or certain products? Or is it across the...

No, it's 22FDX technology platform. And I think this is something that's a couple of years out.

Great. All right. And with that, we're out of time. Thanks again, Tom.

I didn't get to talk about AI.