ASE Technology Holding Co., Ltd. (ASX) on Q1 2021 Results - Earnings Call Transcript
Ken Hsiang: Hello. I am Ken Hsiang, the Head of Investor Relations for ASE Technology Holdings. Welcome to our First Quarter 2021 Earnings Release. Thank you for attending our earnings presentation today. Please refer to our Safe Harbor notice on Page 2. All participants consent to having their voices and questions broadcast via participation of this event. If participants do not consent, please disconnect at this time. I would like to remind everyone that the presentation that follows may contain forward-looking statements. These forward-looking statements are subject to a high degree of risk and our actual results may differ materially.
Ken Hsiang: Now, we open the floor for Q&A.
A - Ken Hsiang: If you have a question, please raise your hand in Webex. First question is coming from Randy Abrams, Credit Suisse. Randy?
Randy Abrams: Okay. Yes. Thank you. Yeah. Actually, if I could ask the first question I tried to get down the guidance real quick, but I want to just make sure I have the right understanding. For IC ATM, so we should imply 5% sequential – about 5%, if I look back to last year in US dollar terms with slightly up gross margin. And then for EMS, I think I saw compared to the third quarter 2020 growth rate. Is that a sequential or a year-on-year for that growth? And then for operating margin I assume it's slightly below. So it's a good sequential improvement. But if you could just recap it just so we have the right assumption on those guidance metrics?
Joseph Tung: Okay. I can’t hear you.
Tien Wu: I didn't write it down.
Joseph Tung: The guidance that we provided for ATM, you're correct. In terms of ATM on the top line, we're expecting the same level of growth that we saw in previous second quarter. In terms of the gross profit margin, we're looking at a slight improvement in the quarter. For EMS, we are looking at EMS second quarter. The top line will be similar to third quarter 2020 level. And the operating margin should be slightly below full year 2020 level.
Randy Abrams: Okay. If I could follow up the -- two things on the constraints. Is there a way to think about the -- how much it is limiting you or how much behind you are on IC ATM? And is it strictly a wirebond that's still the bottleneck, or do you now have constraint on your -- like more advanced packaging the flip chip and wafer Level packaging as well?
Tien Wu: The constraint we were referring to applies to capital equipment including wirebonder also related to substrates, lead frames and other components that are required to do the final assembly. So every product is different. I mean I cannot tell you definitely it's not just a wirebonder thing. In terms of the -- how do we manage the line balance I think that's the operations job. Whenever we're missing some components or materials or try to do the line conversion, we switch back to the other assembly where we have materials in reserve. In terms of how much that limits our potential growth, it's very difficult to quantify that because the process right now is very dynamic. I won't be able to give you a quantitative number.
Randy Abrams: Okay. All right. And the second part of the guidance outlook the EMS actually picking up in second quarter. Last year did, but some years it's still down. If you could talk about the drivers for the pickup if there's some SiP projects or just existing EMS business recovering a bit earlier? And if you could give an update on the overall SiP outlook how that's now looking whether on a year-over-year or if any change versus the incremental growth you're expecting?
Joseph Tung: Okay. I think from second quarter as well as for the full year in terms of EMS, we will continue to see growth. And for this year, I think EMS will go through a typical kind of first and second half distribution of revenue. It will be similar to roughly 44-56 or 43-47 type of allocation distribution. And I think the overall growth comes both from traditional EMS as well as SiP. In terms of SiP, I think last year we went through a phenomenal growth in terms of SiP revenue. We have been entertaining many more projects and many more customers as well. And some of the new -- in terms of the new project revenues we had hit. We had about close to $400 million revenue coming from new projects and we are seeing the same kind of momentum this year. And overall I think the -- in terms of the composition of the SiP revenue, there are some mature products going through some -- gradually tapering off because of feature transition. And there is some projects that we are seeing second sourcing coming in. But on the other front, there are new projects and the -- or the newer projects that we entertain -- or we're starting to entertain from last year we're seeing them starting to also kind of expand for this year. So we're going to see decent growth in SIP overall and particularly in terms of new projects, so we're seeing still very strong momentum going forward.
Randy Abrams: Okay. And I could try to do the math. I think on the 47-53 , is it the implication IC ATM you had mentioned high end of the 2x, which seems to imply U.S. dollar up high -- mid to high teens. Is EMS a similar type of growth profile factor you consolidate?
Joseph Tung: I think the overall annual growth in the EMS business will be slightly better than the ATM overall growth for the year.
Randy Abrams: Okay, great. Okay. And if I could ask a follow-up just on the -- actually two questions on the CapEx. One is the upgrade. I think if I have it right it was originally maintained at the high level you invested last year but now increasing. That's one I guess the area that relative to the prior year you're increasing? And then the second one on the wirebond at 10% to 15% year-over-year, is that more what you see as the need like based on real demand, or is it a constraint number that you would add even more if the capacity? And if I could fit a third, I'm curious the foundries -- usually they'll talk it takes a couple of years to bring up fab, so we might have a shortage for two years. Do you have a view on sustainability of the tightness? Where traditionally lead times they're stretched, but a bit shorter than building a fab. How long it looks like -- like if this looks like it may extend into next year at this stage?
Tien Wu: The wirebonder delivery right now is one year is anywhere between 40 weeks to 52 weeks. And that's the wirebonder delivery. In other words whatever wirebonders that I've ordered now, it won't be delivered until next year. So the wirebonder lead time as well as the other equipment that go with the wirebonding line also got elongated. The wirebonder demand right now is clearly above the efficiency improvement as well as the new capital equipment that we can receive this year. So right now the wirebonder is under allocation highly constrained. Previously I made a comment that for the whole year of 2021, we will see wirebond constraint. My view remains the same except that the wirebonder constraint might last a little bit longer. The back-end equipment has a shorter lead time comparing to the fab, so I will not draw comparison between one to the other. But right now the supply/demand imbalance is obvious for the whole industry, which is why many of our customers who already signed long-term agreement and we are talking about how do we collectively through the design optimization material standardization, we can collectively improve the efficiency to support them for 2021 as well as 2022.
Randy Abrams: Okay, great. That's helpful. And I guess just a follow-up. If you could clarify like the increase in CapEx, it sounds like that might be an area you could place order to get additional tools. So was there kind of a -- versus the prior framework where the new spend is directed?
Tien Wu: The CapEx right now is literally across the board. We're seeing the test equipment. We're seeing the fan-out equipment. We're seeing the bumping equipment. We're seeing the wirebond equipment. The -- almost all kinds of equipment where we're issuing CapEx. Right now we are working with our suppliers, trying to prioritize delivery schedules. In terms of the total equipment that we plan to order, we will stay at the up level -- the high level, if not exceeding last year. In terms of the actual delivery, that is what we need to do from an operational perspective. Now, why are we placing order knowing that we have such a low lead time, because our customers' development and product cycle, as well as the long-term service agreement dictates that. We're working closely with our customer to understand the demand profile long term. And all of the capital equipment expansion we'll take that into consideration.
Randy Abrams: Okay. Great. No, appreciate the color. Thank you.
Ken Hsiang: The next caller we have is Gokul from JPMorgan.
Gokul Hariharan: My first question -- yes, can you hear me?
Ken Hsiang: Yes. We can hear you.
Gokul Hariharan: All right. Thank you. My first question, could you talk a little bit more in detail about what are you seeing? What are the nature of these longer-term commitment partners? Are you talking about two to three-year fixed price, fixed contact -- fixed volume kind of orders? What does that mean for ATM pricing, margins, et cetera? And are these primarily for wirebond, or are we also seeing this spread to other areas in advanced packaging as well? Could we talk in a little bit more detail about what are the nature of these kind of longer-term committed orders that you're getting?
Tien Wu: The long-term service agreement depends on customer. Each one is different. I will not comment in detail on that. The comment that I can make is, the service agreement right now covers more than just the wirebonder. It was wirebonder the second half of 2020 and now we're spreading into flip chip and the other areas, because we do see a general constraint of the assembly capacity. In terms of the pricing environment, the pricing environment remains friendly. But as you know, we do not do tactical pricing. What we're doing right now is, we're working closely with our customer to reflect the raw material and the other component pricing increase we took that into account. And we're also working closely with customers on long-term service agreement in terms of total demand, the capacity we need to build on behalf of their demand. The only area that I would like to comment is, there are specific sectors where we have a super hot run, as well as the expedited product requirement. And normally we will have the expedite fee to apply for those particular cases. But in general, the price environment remains friendly. And I believe that condition will at least applies to the whole year of 2021 if not longer.
Gokul Hariharan: Thank you. Thank you, Tien Wu. If I may also ask about chiplets and the 2.5D, 3D packaging, clearly a lot of your compute customers, especially seem to be talking about this in a very aggressive fashion. Could we refresh what is ASE's view on this area? When we think about the CapEx increase, are we allocating some of that CapEx to your fan-out as well as 2.5D packaging efforts as well, or most of it is still going to come a little bit later?
Tien Wu: Well, I think, you can see from the general -- the mega trend, you understand that for the 2.5D, the 3D, or the chiplet, whichever architecture you're referring to, there has been growing acceptance, as well as growing demand from all regions on all application sectors. So ASE has been developing with our key customer for those architecture. So that has been in place for quite some time. Now, the 2020 and the 2021 scenario, as we're in right now, has modified the situation a little bit in the sense that because we have long-term service agreements with all of our key customers, the chiplet, the 2.5D, the fan-out also becomes a strategic development requirement, as part of that overall long-term service agreement. So in other words, as we're going through, better delivery cycle with our key customers, we are expected to do more development with them, trying to further improve the efficiency and the performance from an architectural standpoint, as well as from a process point. I'm not sure, the exact question that you're asking, but I believe, those are the answers that I can offer you today. Thank you
Gokul Hariharan: Got it. Thanks, Tien
Tien Wu: Are we all set there, Gokul?
Gokul Hariharan: Yes. Well’ I’ll go back into the queue. Thanks.
Ken Hsiang: The next question will be coming from Bruce Lu, Goldman Sachs. Bruce, are you on the line?
Bruce Lu: Yes. Yes. Thank you for taking my question. A very great result. Can you give us a little bit more color in terms of 2021 ATM overall? I look at like the first quarter revenue was very strong. Do you expect the same year-on-year performance for the whole year? And also, for the profitability, the gross margin seems to improve more than 2.5 percentage points already, but the company was guiding only that 2.5 percentage operating margin improvement. I would be greedy to ask for a little bit more because of the operating level. Thank you.
Joseph Tung: I think, the overall growth momentum in terms of ATM, remains to be strong. And in the previous earnings call, we were saying that we will be -- our overall growth will be two times of the logic market growth. And that -- I think that principle remains and although we're seeing the overall industry is kind of -- growth is kind of stepping up, but we're -- because of the -- some of the capacity constraint, we're now saying that, our overall growth should be at the high end which we actually said in the last call, we're saying the -- we're expecting the growth to be anywhere from 10% to 20%. And we are seeing that the growth likely to be reaching the high end of the range. In terms of profitability, I think for the growth level, we'll continue to see sequential growth in our gross margin, as we continue to enlarge our overall operation, also improving the efficiency that we have. And for the whole year, we are very confident that we will be reaching the mid-20% level. The operating margin improvement that you mentioned is really on the consolidated basis on holdco level that we're projecting 2.5% to 3% improvement now. So, when it translates to ATM, of course, the improvement will be higher because, we're setting the -- we're only setting the EMS operating margin target to be 4% for the year.
Bruce Lu: Okay. Thank you. Can I track down a little bit for the ATM gross margin improvement in the first quarter? The gross margin was improved by 180 bps, compared to the previous quarter, but the earlier guidance for the first quarter of GPM for ATM is flat. So, where are the positive surprise coming from? And can you also give us a rank in terms of the gross margin amount like different business in ATM, such as wirebonder testing, flip chip SIP, what was the rank for the gross margin now?
Joseph Tung: We don't give out separate gross margins in our different business. But as a whole I think the improvement that we see -- that we saw in the gross profit margin in this quarter largely coming from -- first of all the higher-than-expected revenue that we can generate in the quarter. The other is really -- we have a higher test revenue as well. In terms of the percentage of our overall revenue test percentage is higher than expected. So that leads to a better-than-expected gross margin performance in the quarter.
Tien Wu: Just one additional comment. The -- when we offered the guidance last year we were planning for the seasonality. In other words the communications sector will go through the typical Q1 and Q2 seasonality. At the time, we were planning on some of the equipment dealing with the communications sector might not be fully utilized as well as the test equipment, as well as the assembly or the SIP equipment. What has transpired in the first quarter was because of the loading situation and strong demand we were able to collectively cooperate with our customers trying to utilize some of the idle equipment that would have been underutilized otherwise. And that as a result improved the revenue stream, as well as the gross profit. That is just another comment. We believe the similar thing will permeate into Q2. In other words what is unique about 2021 years we sort of removed from the typical communication sector seasonality in Q1 and Q2. Now having said that in Q3 and Q4, we will go through the reverse part of the seasonality. So right now from the operation execution wise we have to work very hard to secure all the supply, all the equipment and all the necessary resources to make sure that after clear Q1 execution, we exit Q2 well and we can deal with the second half including the ATM, as well as SIP, as well as the traditional seasonality of the communication sector. So overall I think 2021 will be a very exciting year. Now in terms of the supply situation the cover substrates, covers lead frame ABF substrate capital equipment that you're all very familiar with we will try to give you a quarter-to-quarter update. As of today, we have done a decent Q1 because all of the factor that Ken Joseph and I have just outlined, we are optimistic about Q2 and we're quite excited about the second half of 2021. I mean there is some headwind, the NT dollar that we have not discussed much about it the other type of constraints and of course the general global political situation as well as the pandemic. But with all of this considered, we have given you the best guidance that we believe is pertinent to the current -- the uncertainty in scenario. I hope that clarifies a lot of the questions about Q1 and also the outlook for Q2 and the second half.
Bruce Lu: Thank you. Can I ask a question about the EMS? I mean the guidance for the second quarter for EMS grew about like 10%, despite the typical low seasonality. How much is due to the Asteelflash acquisition? And also the consumer segment in EMS in first quarter seems to go down -- went down a lot. Do you see a rebound in the second quarter as well for the consumer segment in EMS?
Joseph Tung: Second quarter compared to first quarter, we're seeing that both the -- I think really it's the organic growth that we're going to see largely because of the sequential growth is mainly from the supply chain security continuity. I think a lot of the customers are -- because in the whole value chain there is -- there's constraints in terms of component and some of the material supply. And I think some of the customers are really bumping up their inventory hoping to get the -- get their safety stock go up. So we're seeing an uptick in second quarter, which is not a very -- not a typical second quarter performance, but that's what we're seeing now for our EMS business in second quarter.
Bruce Lu: So the whole year -- how do we see the whole year revenue growth for the EMS because nowadays seasonality doesn't really seem to be a good reference anymore?
Joseph Tung: Like I said the -- in terms of the overall EMS business, we're expecting kind of a 44-56 kind of split between the first and second half. So I think the -- going into the second half, we will see new products coming on screen and we're seeing more product launching and then we'll go back to the typical seasonality seeing a stronger -- much stronger uptick in the second half.
Bruce Lu: Okay. Let me try to squeeze one more question. I mean, we do see a somehow different production iteration rate especially the wirebonder between AC and a lot of Chinese OSAT makers. I mean most of the companies in OSAT are having like extremely high iteration rate. But China is high, but not -- it's like a step down comparing to most of the Taiwanese guys. Can you let us know what happened and why is that?
Tien Wu: Well, I think the supply security applies somehow into that scenario. In other words, if a supplier that has a longer working relationship who can cooperate not only the assembly complexity for you the quality, who can also secure better component molding compound as well as the lead frame and substrate supply. I believe that plays into the fact why people tend to place more order even though it's on the allocation mode just to prefer to work with the ASE or our peers in Taiwan. I can't really comment on the China OSAT because I'm sure you know their scenario better than I do. But I think the product complexity, the product security and the geopolitical sentiment might play, might not play into their decision. That you have to talk to the customers.
Bruce Lu: Tien, don't get me wrong. I'm fully agreed that ASE should be -- should have a much stronger customer demand. But your customers are actually buying or like are in serious shortage. I mean they need to grab like the dying people they grab whatever they have, right? So the gap seems to be a bit larger than I expected. So that's why I tried to get some color.
Ken Hsiang: That's precisely the point. I think when things are tight, I think most of the customer will look for the safest bands. And given our scale and given the leverage that we have in terms of sourcing, capital equipment as well as materials. I think we are much safer bet to our customers. That's one front. And the -- I think the other one is that given the geopolitical situation, I think there is a growing concern on the longer-term or mid to longer-term sustainability of some of the Chinese players. I think it does play into the current situation a bit.
Bruce Lu: I see. Understand. Thank you. I’ll go back in the queue. Thank you.
Ken Hsiang: All right. Thank you. Next question is coming from Roland Shu of Citi. Roland?
Roland Shu: Hi. Thanks a lot for taking my question. And just one quick question on the CapEx. So I think for today, I don't hear you update your total CapEx spending plan this year. So how is the CapEx spend being planned this year?
Joseph Tung: I think Ken briefly mentioned that this year, we are expecting to spend roughly 10%, 15% more than we spent last year in terms of equipment CapEx. And this is really to support the surge in demand that we're seeing now. And although we are -- we kind of brought up the overall CapEx spending amount this year, but in terms of actual spending it really depends on the delivery that we will have. But nonetheless, I think that's the current situation. We are upping our CapEx. And in terms of distribution, I think, out of the total spending roughly 65% would be for assembly, we're up 21% for tests, 12% for EMS and the rest for our material. I think that will be the distribution for this year.
Roland Shu: Understood. So it sounds like I think that the total number should be somewhere around $1.9 billion to $2 billion you are planning right?
Joseph Tung: Yes. Last year we spent close to $1.7 billion. And this year…
Roland Shu: Yes and 10% to 15% higher.
Joseph Tung: Yes. Correct.
Roland Shu: Yes. Okay. Cool. Thank you. And then probably 80%, 85% for assembly and testing. So our question is when you look at TSMC's CapEx spending on advanced back-end and their mark making is going to be around $3 billion this year. So this actually is much higher than your total spending about $2 billion this year. But does it mean that you have to raise CapEx spending significantly going forward if you are going -- also doing -- trying to do more advanced packaging business?
Tien Wu : Well, the comment about the TSMC CapEx for the back-end. And I think the first clarification I would like to make is you really cannot make a -- it's not a direct comparison. The CapEx that they're referring to for their back-end versus the CapEx we are referring to for the assembly equipment are very, very different in nature. If you really go back to the CapEx the equipment list and you understand that. So there is no direct comparison that I can draw between that levels. Now the second comment is the -- if our customers demand us to engage in the type of configuration or the architectural design where our customer are designing with foundry suppliers then we are obligated to work with our customer to come up with our proposal. Our proposal could be in the form of what the foundry is using, but it might not be in the form, and it's up to the customers how to design their packaging the architecture with us. The -- so that is a hypothetical question. Right now we are engaging with several customer, trying to explore our end of the proposal. Now in some form, there will be overlap. But in the majority of the form it will be quite different, right? So I don't believe our CapEx will be at the foundry level, because in nature it is extremely different. Also in terms of -- the variety is also quite different, because we're not dealing with the large high-volume capacity for a few customers. The process and the architectural design we need to come up with has to be able to fit into the application segment for a basket or a number of customer interchangeably and that is the key difference between the business model versus the CapEx.
Roland Shu: Okay. So, Tien, correct me if I'm wrong. So, you are pretty much meaning TSMC is building its technology and proprietary technology for their customers. And this technology probably is different from the platform technology you are doing for customers product across the board. Am I reading you right on that?
Tien Wu: Yes, that I believe -- well, again, it's not what the ASE wants; it's what ASE customers want.
Roland Shu: Understood. Okay. Thank you. And I think a follow-up question is for -- in the past you said for every $1 CapEx you invest on assembly it probably will generate about NT$1 new revenue in the first year and more than NT$1 new revenue generation for testing for the first year. So, for the CapEx you are spending now and because of this capacity tightening, are you still going to generate a similar return for the CapEx or investment you invest in your new capacity now?
Ken Hsiang: Well, I think as a rule of thumb NT$1 of investment in packaging we can generate about NT$1.20 of revenue on any -- and for tests NT$1 of investment can generate about NT$0.50 of revenue. So, from a blended point perspective, I think NT$1 of -- what we're seeing is in terms of assembly and test, NT$1 of investment should generate NT$1 of revenue for us -- a blended revenue for us to make it an economically viable investment. I think the -- from a different angle to look at TSMC's investment into back-end, that's totally different scenario.
Roland Shu: Yes. I know it's very different. So, in general, you said for the back-end you probably will have -- for every NT$1 investment, you probably will generate NT$1 return. So, this is still the thumb rule still valid, right?
Ken Hsiang: That's still the rule, yes.
Roland Shu: Okay. Cool. Okay, thank you. Thanks for taking my question. Thank you.
Ken Hsiang: The next caller we have is Szeho Ng. Szeho, you're working nine to six?
Szeho Ng: Hey, I'm in Hong Kong. Yes. I have two questions for you guys. Anyway a good result. First one regarding the wirebond delivery, I recall that in Q1 you got roughly a thousand hundred wirebonders, right? So, I basically want to note the wirebonder delivery schedule for the rest of this year.
Ken Hsiang: I think currently we're looking at 3,500 to 4,000 bonders for this year addition. And we're expecting full delivery by maybe October-November timeframe. It will be delivered progressively. And I think full delivery were expected by October-November timeframe.
Szeho Ng: Okay. Got you. Okay. All right. And then the other one is on the housekeeping. What is the utilization for your wirebonder business and testing back in Q1? Where rough number would be fine?
Ken Hsiang: I think in terms of packaging, we're around 85% and for tests around 80%.
Szeho Ng: That's go at peak utilization right I believe?
Ken Hsiang: Yes, that's pretty much -- we're pretty much maxed out.
Szeho Ng: Okay. Good. And last one on the dividend policy, any update compared with three months ago?
Ken Hsiang: No, I think we've announced it already. It's going to be NT$4.2 this year and the payout ratio is roughly 65%.
Szeho Ng: Okay. And that will be the rule of thumb for the future -- at least, for the near future, right?
Ken Hsiang: Yes. Yes.
Szeho Ng: Okay, great. Congratulations on a good results.
Ken Hsiang: Thank you. We don't have any additional hand raisers in the queue.
Ken Hsiang: Okay. Thank you for attending our first quarter earnings release. See you next time.
Tien Wu: Thank you.
Ken Hsiang: Thank you.
