Great. We'd like to get started. Welcome back. My name is Toshiya Hari. I cover the U.S. semiconductor space here at Goldman. Very happy, very honored to kick off the PM session here with the team from Amkor Technology. To my left, we have Megan Faust, EVP, CFO, and Treasurer, and to the far left, we have Kevin Engel, EVP of Business Units. I do have a very long list of questions, but would love to keep this interactive. So to the extent you may have questions, please be prepared. I'll pause here and there. Thank you for coming. Amazing to have you here.
Good to be here.
Yeah, so I wanted to start off high level. Megan, if you could recap Q1 earnings for us. It's been about a month, since you reported results, very strong results relative to guidance. What were some of the key highlights?
Sure. So our Q1 results came in a little bit stronger than expected. We were at $1.365 billion. So while that was a deeper than seasonal decline than what we've traditionally seen overall, it was a good Q1. What drove that is a more than seasonal decline in our communications market, and that's really off of a record 2023. So communications for Amkor actually grew in 2023, despite the cycle and of all the other markets being depressed, so that was a pretty tough comp for Q1. Automotive, I would say, came in a little bit lighter than we expected, and that's continuing into Q2. The bright spot was computing. That was the only market with some sequential growth, and I'm sure we're gonna talk a lot about our 2.5D technology here today.
So those are probably the key highlights for Q1.
Okay. That's, that's great. And then in terms of your Q2 guidance, you, you expect revenue to grow about 6%, sequentially at the midpoint, which I think is a bit better than what we've seen, seasonally over the past handful of years. I know seasonality has been a little bit wacky, the past handful of years, but can you provide a little bit of context color by end market in terms of your Q2 outlook?
Sure, yeah. Q2, definitely stronger than what we would seasonally expect. All of our markets are gonna either be flat to growing, so we see that as a good signal. Year-over-year, Q2 will be essentially flat, so that's signaling for us, you know, coming out of the cycle. Primarily, what we're seeing in Q2 is starting to, I guess from an absolute dollar, the biggest growth driver will be communications, beginning to start to ramp up for that second half peak, the iOS launch that we would all expect. We're also gonna see continued growth in our computing segment, supporting the 2.5D technology for AI applications, but also high-end PCs that are converting to the Arm-based architecture. And then, with consumer and automotive probably being more flattish.
Got it. And then kind of further going in time, second half of this year, I know lots of moving parts, but you did soft guide revenue to grow about 30%, half-over-half. I guess part one is, what gives you the confidence to guide beyond the current quarter? And part two is, are there any specific design wins or content gains that, again, sort of supports that confidence?
Sure. Yeah, so we feel really good about our second-half growth story. So while automotive is coming in lighter than what we anticipated at the beginning of the year, we are seeing upsides in some of our other, markets. And specifically, we can talk about 3, programs that we see giving us confidence for that second half, 30%, half-over-half growth. The first one is communications. Second half is always a very strong seasonal ramp. We always are seeing continued content growth, despite units maybe being more mild. We're also seeing market share gains in that space, and we've seen that over the last several years. The second largest driver to that second half growth is a specific ramp in the consumer segment for IoT wearables.
So this would be a next-generation product, that's coming out at high scale, and that really can drive significant revenue. And then the third is we're putting in place, at the tail end of Q2, incremental capacity for our 2.5D technology, which will then support further growth in the compute segment in the second half. We're also seeing some recovery in the Android market, as well as the memory market, that are helping, you know, to give us confidence in that second-half story.
Got it. That's very helpful. Before we transition to questions that are a little bit more specific to the end markets, I was hoping to bring Kevin into the conversation. Kevin, I think you joined Amkor 20 years ago. It was through an acquisition. What has the journey been like for you over the past 2 decades at the company? And as the EVP of Business Units, what are your key responsibilities, and what are your near- and long-term goals?
Thank you. So not much has happened over the past 20 years. No, so I joined Unitive, or Amkor, when Amkor acquired a company called Unitive, Unitive Electronics, in 2004. At that time, I was just an engineering manager. Over my course of career at Amkor, over 20 years, I transitioned into many management roles, as well as engineering roles, really focused on advanced packaging. I've really seen Amkor transition into focusing on advanced packaging and been really excited about that transition, and becoming more involved in the business unit allowed me to really get more involved and see how that was gonna transgress over time. So across that timeframe, I also did a 2.5-year expat assignment in, in Taiwan. So Taiwan is kind of my, my second home, so very exciting.
So, you know, if I look forward into the future with Amkor, you know, I joined, we were about a $2 billion company. In 2022, we were a little over $7 billion. You know, I fully anticipate that we'll continue to outpace the semiconductor market growth and excited to see how that transitions, obviously, as we all have a goal of semiconductors getting to $1 trillion around the 2030 timeframe. Exciting future ahead.
That's awesome. Thank you. And then kind of delving into, to some of the end markets, wanted to start off with communications. It's, it's roughly half your business today. It's doubled in revenue over the past several years, and that was during a timeframe in which smartphones were flat, plus or minus, so, so you guys have done really, really well. What have been the key drivers of growth over the past three, four, five years? And as you think about the forward, what are some of the, the potential key drivers?
Yeah, so I think there's a few different dynamics there. Obviously, we participate in the Android ecosystem as well as the iOS ecosystem. We tend to perform at a higher level in the premium tier phones as well. So as consumers purchase more on the premium tier side, that helps Amkor. You know, if you look at the content increases, you know, continued applications, continued functionality, those are areas that continue to help us to grow, even as the overall unit sales may not be growing. So that's a good opportunity for us, and then just market share gains in general, we continue to focus in that area.
You touched a little bit on units, but is that an area... I don't know if you guys have a strong view, but even in a flattish environment, is it fair to say that you continue to plow ahead and...
Yeah
-grow the business?
Yeah, like I said, there's opportunities there because, again, we're tending to see the consumers purchase more premium tier, which is also a higher share for us. And, you know, we're continuing to see, you know, Android, at least short term, starting to see some growth trajectories in the China space, Android recovering, and you'll have a lot of new launches coming up in the iOS ecosystem soon.
Got it. Within the context of your communications business, you have large exposure to one customer. They accounted for 28% of revenue in 2023. That was up from 21% in 2022, and then 14% in 2021. It is really nice to see you have so much success at a very demanding customer. I think it is a really good sign. That said, Megan, I am sure you can attest to this, but the market, the financial markets have mixed views toward customer concentration. As a management team, how do you manage that customer concentration? And is there a certain level where you think this is too much, or how do you think about that?
Mm-hmm. I can kick it off.
Yeah, sure.
I think one thing to remember is Amkor is engaged with all semiconductor companies.
Yeah.
So we're gonna be engaged, you know, with all of them. So it's not that we're gonna manage it from that perspective. You know, one of our strategic pillars is to engage with the technology leaders, and that one specific customer, as we all know, is quite an innovator. And so it's important for us to have deep relationships and, and key development roadmaps with those customers. The last thing I'd mention, too, is we have done a good job diversifying our portfolio with that specific customer beyond communications, so that will also help to the extent there's any shift. Anything-
Yeah, the only other thing I'd add is, I'm sure we'll talk about some of the other markets, but we see some really good growth potential in some of the other markets that could outpace the growth in communications, so some, some other opportunities.
Got it. Transitioning to automotive and industrial, a little bit over 20% of revenue in 2023. The business has, has been down about 20% on a peak-to-trough basis. Megan, I think you, you touched on it a couple of times, still perhaps a little soft. Many of your customers in the IDM space have spoken to improving bookings and, you know, customer cancellation rates normalizing and some signs of expedites, all, all signs that are positive, as we think about the cycle. Have you seen a, a bottoming, sort of, or a bottom form in your business as well, or is it a little premature to, to make that call?
Yeah.
What are you seeing in the market?
So I would say that, you know, we definitely a little bit softer. You know, if we think about the correction in general, automotive was probably the last, you know, segment to really start to see that correction. We definitely still see that inventory levels are getting worked off, you know, probably another quarter or so to, you know, maybe two quarters to burn that off. But in general, the customers feel relatively confident that that will, you know, start to transform. You know, for us, we're focused on things like our customer partnerships. We recently had an announcement with Infineon in our Portugal facility, which is helping to diversify the supply chain within the European ecosystem, so there's a lot of interest in that Portugal facility.
Automotive, in general, longer term, we see a lot of, you know, potential opportunities for growth, whether it come from, you know, the safety features, you know, ADAS and other applications that are continuing to grow and then proliferate into other models. And then, obviously, the transition to, to hybrid and EV, which is also creating opportunities in the power module space. So a lot of, lot of future models.
Maybe as a quick follow-up, Kevin, since we have you, the partnership that you just talked about with Infineon, can you sort of talk about how that came together and what that means for them and for you?
Yeah. So for us, you know, our facility in Portugal, we acquired that facility in 2017, so that was a nice footprint add for us. We acquired them for a wafer-level fan-out technology. Over time, you're going through COVID, and then, you know, starting to see the geographical dynamics happening, customers started to become much more interested in allocations. We've added a lot more capabilities, including flip chip, BGA, sensor applications, and other applications. We've also had a partnership announcement with GlobalFoundries in their from their Dresden facility, moving some of their assets into our Portugal site. So we're even in the phase now of increasing the floor space of that facility to allow for some of these opportunities. So with Infineon, it's more of a collaboration to build new technologies and scale in that facility over time.
Got it. Thank you. In your consumer business, again, 10% plus of revenue in 2023, it performed extremely well during the pandemic, and off of those highs, I think revenues are down, or were down about 40% in 2023. Is the inventory decline at your customers more or less done? Is it still in motion? How would you characterize kind of the cyclicality in that business?
Mm-hmm. Yeah, so for us, I'd say there's two components there. Obviously, we're this tied to consumer spending in general and the overall, you know, consumer purchasing. And then the other one is just new product launches. So if you have a high-volume product launch, then that can really drive up volumes for us. We see an exciting new product launch going in the second half of the year, so we definitely see that second half will bode well for us.
Medium to longer term, I think you peaked at $1.4 billion. Any specific applications or device types that we should be looking out for in terms of potential growth drivers within the context of consumer?
Yeah, I think it's broad. It's very broad. I mean, think of any kind of wearables, hearables, those types of applications that drive high volume, those are good opportunities for us.
Okay, got it. Saved the best for last, computing. 2.5D, obviously very, very topical. I think publicly you've said, you'd triple your, your capacity-
Yeah
... by the end of this quarter, I believe, relative to a year ago.
Correct.
Two-part question. One, what device types are you currently addressing? And number two, how does your technology differ from the likes of TSMC?
Yeah, so a few things there. So first of all, we work directly with the end customers. I'd say majority today is in the GPU space. We do have engagements with other customers as well. And interestingly, there's a nice pipeline, you know, moving forward on top of that, so, and starting to get into even different application spaces. So we've invested significantly in that space. That's a pretty exciting investment. We continue to invest even beyond this quarter, so that's an area where we're growing. And those assets were, are also fungible to some of our existing technologies, like wafer bumping and/or flip chip assembly, and to our future generation technologies, whether they're, you know, what Amkor call our SWIFT technology, which is similar to a TSMC CoWoS-R, or our S-Connect technology, which is similar to a TSMC CoWoS-L, as an example.
We see a good long pipeline, and then engagements with our end customers is pretty exciting.
Okay, that's awesome. And then a couple of follow-ups on 2.5D . You know, demand for GPUs are literally off the charts right now, and there's good visibility into 25. How much visibility do you guys have when it comes to your 2.5D business? Do we, do investors need to worry about oversupply at some point? It's probably a little too early to be worried about that, but helpful to get your input. And then part three, what are your medium to long-term aspirations in the market? Again, you've tripled capacity on a one-year look back, but as we think about the next 12 months, 24 months-
Yeah
What kind of capacity are you planning?
Yeah, thanks. So I think if we think about oversupply dynamics, yes, like you said, very early in the phase right now, so a lot of, lot of growth potential right now. Obviously, multiple suppliers, you know, increasing capacity. For Amkor we're, you know, only one of two that are supplying the full, the full flow. So that means also, processing the interposer, placing on the substrate, and placing the ASIC die and the HBM on the interposer wafer. And then also going on the substrate, which, you know, in some flows, that only on substrate portion is being performed. So we are supporting the full flow, which is exciting.
Again, I talked about additional customer roadmaps, so not only for this, you know, the primary customer everybody wants to talk about today, but in the future we see additional application spaces for 2.5D. So not really worried about overcapacity over the next, you know, six months, a year timeframe. You know, beyond that, again, we'd look at the fungibility of those assets. It's a good pipeline, as well as a fungibility into next-generation technologies.
Okay. But on the-
Yeah.
Sorry, go ahead.
And then further on, you know, thinking further on, again, you know, the whole trend towards chiplets, you know, to me is kind of driving this advanced packaging trend, and that bodes really well for us. That'll drive a lot of these newer technologies and then also increase, you know, some of the value that, you know, the tier one ASICs can bring to the market.
Okay. On future capacity additions, you haven't really said anything publicly?
Yeah, we haven't publicly given another, but we are, we are looking at expanding going forward. Given we already have scale, that can take a different form. You don't have to put in a whole new line to, to extend your throughput.
Okay, understood. I've got a bunch more, but I wanted to pause here and see if you have any questions from the audience. Okay, I'll keep going?
Yeah.
Yeah, so, shifting gears a little bit, on gross margins, probably one for you, Megan. So I think generally, your gross margins have, you know, hovered around the mid- to high teens. I know you and your team don't necessarily run the business to a gross margin percentage number. I think you're focused on dollars and free cash flow. But as you probably know, we collectively do scrutinize gross margin percentage. Given the increased focus on advanced packaging, you know, some of the things Kevin talked about, I would think you would be able to capture more value, higher value. How should we think about that? Is that something that we should look forward to 20%, 20% plus, is that achievable, too high of a bar, any thought?
Sure. So let's level set. Q1 is our low point for utilization. That was around 55%, mid-fifties utilization. Really, the way that we look at our margins is there's a great leverage in our financial model, and with our financial discipline associated with that, we're able to get really good incremental drop through as that utilization comes back. So just the second half example, where we're expecting great growth, we'll probably get up into the mid-seventies utilization, and I would expect to see that margin expand in the high teens, and we won't even have reached peaks, you know, that we saw back in 2022. So utilization is the biggest lever to our profitability.
On the second comment you made, as far as commanding more value, we do see, as far as our engagements on some of these very advanced technologies, as well as our geographic footprint, that that does enable us to, I would say, command more value. So example, the business we were just talking about with 2.5D, that does have higher than corporate average gross margins. So those will be tailwinds to the profitability, but overall, we're really focused on optimizing the use of our assets.
Okay, understood. I guess, I guess as a quick follow-up, maybe for, for Kevin, from an operational perspective, are there any sort of initiatives in place to, to improve gross margin percentages or anything that you can point to that could potentially drive, you know, margin?
No, there's always a cost dynamic, so we're always looking at managing our cost structure and two components to that, obviously, managing our material suppliers, but then also really going back to utilization. So, you know, we are a very geographically diversified company. So with that, you know, we are shifting assets around to make sure that they're getting utilized, you know, as best as possible, rather than investing in one region versus another. So I'd say overall, that utilization is the key driver.
Okay. On the competitive landscape, I think historically you've gained share very successfully so. But I think when we take a step back and look at the OSAT market, vis-a-vis some of the other subsegments that we cover in, you know, across the overall semiconductor ecosystem, it continues to be a pretty crowded market. It's very competitive. How would you characterize the competitive environment today, vis-a-vis, you know, five years ago? And if we asked you, you know, how are you different relative to the pack, how would you go about responding to that question?
So I would say there has been some consolidation, at least in the OSAT, so we can kind of talk about this in segments. So where today there's really only what we would characterize two Tier One OSATs that have full advanced packaging capability and a very broad portfolio. So it's really consolidated down over the course of the last 5-7 years. When we think about the other two edges of the OSAT space, you have you know, IDMs. They also provide their own packaging, and traditionally, there had been a spillover model to OSAT. But what we're seeing in the market today is they're choosing not to invest in the very advanced packaging as part of an IDM, you know, model. So they're looking to OSAT who can invest at scale. So that has become a tailwind for the OSAT.
When you look at the foundry side, we really see what we offer to the foundries as complementary. So the foundries may be focused on a specific packaging technology in order to have the most cutting-edge technology and bring that to market. And then once that's at scale, that tends to move into the OSAT domain. So today, we would characterize, you know, the competitive landscape. It's competitive, but we would call it rational, with respect to what we're seeing both in specific OSATs and then how the dynamics work amongst the other players.
Got it. And as a follow-up to that, Megan, you talked about how the market has consolidated over the past 5-7 years, and I think you've played a role in that process as well. As you look forward, are you willing to play the role of consolidator? And if you are, what characteristics or capabilities would you all kind of look for in a potential acquisition?
Yeah, so we've really targeted our acquisitions to ensure a very specific accretive value. We would look at some sort of technology play, and that was the acquisition we had back in 2017, which allowed us to bring to market that technology, accelerate that by 18 months. We've also looked at regionally, where could we acquire in order to expand our footprint, and Japan is a great example of that, where that acquisition really, you know, I would say, accelerated our number one position in the automotive space, but also gave us a fantastic footprint in Japan. And last, the other area we look at is potentially partnering with our customers, where we've acquired their back-end facilities, and our Malaysia facility is an example of that. So while we're not, you know, acquisitive every year, it has to make sense.
They tend to fall into one of those three spaces. We're not gonna acquire just scale, that isn't something that we see as being accretive. We have a great footprint and the ability to scale that. We don't need incremental wire bond scale, which is really where a lot of the tail OSAT space is today.
Got it. You announced last November that you plan to invest $2 billion in building out a facility, a leading-edge facility in Arizona. I think you publicly shared that Apple will be your first and largest customer. I think they had a press release out on the same day. A multi-part question for you. Can you provide details around, A, the timing of completion? What your expectations are from a government support perspective, whether operations in the U.S. will be as cost-competitive as some of your international sites? And to the extent capacity in the U.S. is more expensive, are customers willing to pay more to have access to Arizona, vis-a-vis, say, Korea or Japan?
Yeah, yeah, sure.
I think that's four parts.
Yeah, that's a lot of parts. I can-
Try to capture all of them. Refresh my memory, I probably lose one. So the land that we acquired is about 15 minutes from the TSMC location, so it's in Peoria, Arizona. So it's a very good location for us. You know, from a timeline perspective, you know, we're looking at a 2- to 3-year cycle, so basically, we'd be looking at production in the 2027 timeframe. Yes, there are challenges in the U.S. related to cost. You know, for us, we're looking at that from a couple different dimensions. Number one, you know, state and federal incentives, which are critical to success, so I think that's an area that we're very focused on. Then we also have the items within our control, so things like higher levels of automation. This will be an advanced packaging site, which does bode well to more automation.
On top of that, we're working with our customers to try to look at a more level loading profile, so that, you know, we won't see the typical seasonality that we would see in the market, but more kind of preferentially load in the U.S. and then overflow into our Asia facilities. So those are some of the key dynamics there. And then working with the customers, you know, obviously, this is not only for those two customers, but for multiple customers across multiple markets, you know, including communications, automotive, as well as high-performance computing. So a lot of different aspects going on there.
We feel that over time, we can make this sustainable business, you know, with that help from the government to kinda, you know, boost to get us up into that high volume scale quickly.
To the extent capacity in the U.S. is more expensive than, say, Portugal or Korea, or Japan, you know, your ability to capture higher value, is that maybe it's a little bit early to-
It's early.
But-
I think what we're seeing, though, is there's definitely engagement throughout the supply chain to see how we can, together, work together to bring that supply chain to the U.S. So we're happy to see the level of engagement on all sides.
Okay, understood. Again, I'll pause here. Any questions from the audience? Keep going. On technology, we talked a little bit about 2.5D, which is, you know, I wouldn't say it's just getting started, but it's, you know, starting to get the attention that maybe it deserved a while back. Are there any other technology inflections, whether it be on the packaging side or on the testing side, that we all should be monitoring going forward?
Yeah, so for me, I think I've touched on already, but the trend towards chiplets, I think, is still very important. You know, and that's gonna continue to drive these more advanced packages. Like Megan said, you know, the IDMs are focused on some of their legacy-type package types and pushing more of the advanced packaging, looking to the OSATs to help there. So I think that bodes well. So I think there's a lot of opportunity in chiplet structures. The other one that's a little bit further out there is just the trend towards electrification. You know, we're gonna have, obviously, EVs, and if we think about even data centers and GPUs, the amount of energy consumption there.
Energy usage, energy storage, energy efficiency, to me, those are all gonna be things that are gonna be very important in the future.
Understood. I'm hopping around here, but on capital allocation-
Mm-hmm.
Megan, we talked a little bit about acquisitions, but outside of acquisitions, what are your priorities? How should we think about, you know, your intentions going forward?
Sure. So our capital allocation priorities have actually shifted over the last several years. Now that we've met our target debt to EBITDA leverage of 1.5 times, that's really put our, you know, focus on reducing debt a bit in the back seat. And now we have announced a shareholder return framework. So as that focus on debt has kind of moved to the back seat, we've moved that shareholder return more front and center. So the framework that we've established is returning 40%-50% of our free cash flow to shareholders over time. So coming back to the priorities, number 1, obviously, investing in the business. We tend to operate on a capital intensity of low teens.
We've proven that that allows us to have very significant growth, both in capacity expansion, but also what I would frame in the technology development and in the R&D space. Being able to grow nearly $1 billion three years in a row from 2020 to 2022 at that intensity showed us that was the right optimal investment stage. So investing in the business is the priority. We talked a little bit about acquisitions. I would also throw in there what we would characterize as strategic investments. While they may still be viewed as organic, our greenfield in Vietnam was, you know, a project that we would put under that strategic, you know, spend.
That's opened last quarter, and we're expecting to see high volume manufacturing out of that facility this year, and then the US facility would fall into that strategic investment. On the debt side, even though we're not looking to reduce debt, we do look to optimize our cost of debt and our maturity profile, so there's always work and focus there. Then, capping out with the shareholder return, we've primarily used regular dividends for that shareholder return. We have increased those annually since inception, and we would expect to continue to do so in the future. The other tools we could use for shareholder return could be special dividends or share buybacks.
So Amkor traditionally has not had significant. We haven't had share buybacks in many years, and when we did, they were quite minor, and that's really been a function of not wanting to put pressure on float, given our shareholder structure.
Understood. You mentioned capital intensity in the low teens. As you look forward over the next couple of years and all the technology inflections that, Kevin talked about, is there any upward pressure or downward pressure? Is still low teens the right level, as you think about your, your capital intensity profile?
You know, low teens seems to be the place where we can accommodate those technology shift chains. Maybe year to year, there could be some increases, but over time, we see that as, as a good investment. To the extent there needed to be something, we wouldn't constrain that. We do have the liquidity to support that level of growth.
Understood. Megan, you were just telling me how you've been busy with all these meetings. Anything about the Amkor story or anything about the markets you play in, that you feel like we collectively miss or underappreciate?
Mm-hmm.
throughout these conversations that you're having?
Sure. So I think the area that, that we are really excited about and are starting to see better understood is our differentiation with respect to our geographical footprint. So we've talked about that in various stages here, but from Amkor's perspective, compared to our competitors, it's a significant differentiator, where we have at-scale operations in jurisdictions that are not only attractive to our customers, whether they're wanting to de-risk their profiles, but also, as you can see, many regions are investing in the semiconductor supply chain, wanting to regionalize those and bring those forward. Kevin mentioned Europe as an example. We're seeing that investment in Japan, recently announced, seeing that in Korea. So that's something that, for us, is, is quite exciting, that we're already so well diversified and being able to then support those regions and wanting to build out those supply chains.
Great. Hopefully, that was comprehensive. Anything else that you'd like to touch on, or?
No, I don't think anything.
No, I think that's good.
Okay, great. Well, thank you so much for coming.
Okay.
Appreciate it.
Great. Thank you.
Good luck.