Advanced Energy Industries, Inc. (AEIS) Earnings Call Transcript & Summary

Hi. Good morning, everybody. My name is Amanda Scarnati. I am one of the semiconductor analysts here at Citi. We're joined this morning by Steve Kelley, the CEO of Advanced Energy; as well as Paul Oldham, the CFO of Advanced Energy; and Edwin Mok, who is on the strategy team, I believe, at Advanced Energy. We're going to dive into some prepared questions that I have. Feel free to either click on the button to send me questions directly or you can send me an e-mail at [email protected]. And I'll be happy to ask those investor questions later on in the presentation.

Steve, let's just dive right into the semiconductor side of the equation, and it's still the biggest piece of the portfolio by far, if you kind of divide everything out. The revenue has been quite a bit lumpy over the past 4 quarters. We've been seeing quite a bit of strength in the semiconductor capital equipment space. How should we think about growth for the balance of this year -- for the balance of for 2021?

Yes. I think in semiconductor, the way to think about this year is totally supply constrained. We've had a couple of issues. One is obviously getting ICs to build our units, which we deliver to our customers. The second is we run into constraints in Malaysia where we do the final assembly and test. So the good news is as we look into Q4, we think the constraint in Malaysia goes away. Because over 80% of our employees in Malaysia will be vaccinated in the next week. So once we hit that threshold, we can come back to full participation of our workforce at the factory. That's a big plus. The second issue is parts availability. And I think we're seeing some incremental progress there. So what we've said is Q3 is going to be better than Q2 from a revenue standpoint in semiconductor. And we think the second half is going to be better than the first half. We think next year is going to be quite impressive as basically the supply catches up with demand. So we're looking forward to 2022 in the semiconductor front.

Perfect. Semi grew up about 52% in 2020, which is more than double the rate of WFE growth. Do we think that we can continue to see outgrowth versus the WFE market? Or do we expect to start seeing some under growth there?

I think when you look at that market, obviously, we outgrew the WFE by a wide margin, about 2x WFE in 2020. Now 2021, we've actually been below WFE because of our parts shortages. So we're shipping below WFE growth this year. But we expect to ship more than 1.2x WFE next year, again, as the supply of ICs catches up with the demand we see in front of us. So we think overall, we're growing fast than WFE by a significant margin, if I look over a 3-year horizon.

Another thing that we've been hearing quite a bit about, right, is that you're talking about supply constraints, you talked about potentially outgrowing WFE by 2x the market next year, and a lot of that has to do with some of this backlog that you've been having. Can you just give a little bit more details in this large increase in backlog? How much is expected to ship in 2 half? And how this gives you all this confidence in this 2x growth number for next year?

Yes. I think the demand is very sticky. As a reminder, 70% of our products of our revenue is sole sourced revenue. And so it's very sticky. It's a long life-cycle business. And that's in semiconductor, almost 100% proprietary, and we have a similar situation in industrial. So that's why I have high confidence that we can ship that backlog once we get the parts to build the units. So that's, I think, the first point. We're seeing progress, I think, on the parts side. But again, it's lumpy. We've narrowed our focus to really a few main suppliers at this point, and we've made great progress with the number of suppliers over the past, say, 4 months.

Right. Perfect. Last question on these supply constraints. And I think the elephant in the room will be gone. And then we can dive into some other stuff. So looking at the industry, some mixed messaging between what you're seeing, what peers are seeing in terms of supply chain issues. How much of this do you really think is related to the slowdown that you've been seeing in Malaysia versus maybe some supply management issues or other factors that are making your commentary different than what others might be seeing?

Yes. I think we're in good company. I mean there's a lot of companies that are constrained today by part shortages. And as a reminder, we did set revenue records in semiconductor in the first half of this year. So it's not all bad. But I think there were some issues we had that were particular to AE, particularly on the Artesyn side of the business, where we're optimizing for cost and for low inventory going into this shortage scenario. And so it took us a little bit more time, I think, to recover in those areas, particularly in areas like hyperscale and some of the computing areas.

Right. On the margin side of the business, Paul, this might be a little bit more geared towards you. Margins contracted about 4.5 points from peak levels last year. How much can margins improve as we head into the December quarter? And when can we see return to margins in the high teens?

That's right. If you look at gross margins for a minute, Amanda, we were running late last year and even the first quarter very close to our long-term margin target of 40%. We were between 39% and 40%, which was about a year ahead of schedule when we laid out that plan. Unfortunately, the supply constraints have had not only an impact on revenue, but a significant impact on gross margins as well. So if you look at that [ 4 ] to 400 basis point change that we've seen and you kind of break it down, the first thing that you see is that there is higher costs on materials primarily. These are primarily for things like expedite fees, running hot lots, and trying to find parts from distributors or other uncommon places where we might pay a premium and logistics costs. The supply chains and the freight lanes are very full. And as we expedite into the material round to meet customer needs for paying more, that's the biggest part, making 200 to 300 basis points of the decline. The next thing is that we -- our factories aren't running very efficiently right now. Our factory utilization, given the supply constraints, is low. And we've made a conscious decision not to try to furlough or reduce our workforce during this environment because we do expect the parts environment to improve, and we want to have the capacity and capability in place. It's still a tight labor market in these countries, and we want to make sure we're retaining our workforce. That's another 100 to 150 basis points. And then there's some other plus and minuses that go into that. Now as we look forward, we think things will start to get a little bit better in the fourth quarter. Certainly, we expect to ship a little bit more, resolve some of our issues from an output and from getting materials out of the factory. As Steve said, Malaysia should be running a bit higher output given the COVID should be improved. But at the same time, we'll have the full impact of our -- the parts, really start to hit us in the fourth quarter. Some of those costs are sitting in inventory today. So they haven't flown through the P&L yet. As we look into next year, we don't see any reason why we can't get back to our long-term target of 40% over the course of the year. As the supply situation improves, we should be able to see more normalized procurement processes, less expediting. Some of these premiums we're paying in the near term should ameliorate, and they should -- and our factory utilization will definitely improve. Now there's going to be some higher material costs, but we're also able to pass some of that along to our customers. So when we look out a year, 18 months, we believe we should be able to get back to our original targets as we work through this sort of supply chain -- the near-term impact of the supply chain environment that we're in.

Perfect. And then on the operating margin side as well, obviously, we talked all about what's impacting the gross margin. But how can we get some improvements here on the operating margin?

Well, looking from an operating margin perspective, again, we've decided to continue to make strategic investments and other investments in the business rather than taking short-term actions to try to pull back investments in a supply-constrained environment. So as revenue growth returns and as gross margin improvement returns, we should largely be able to see that drop to the bottom line with very little increase in our fixed cost structure. Now there will be a little bit of variable costs that come through as volumes increase. But essentially, we'll be able to see operating margins improve back on the rate that we've seen historically in the 35% to 45% range. So as the supply chain environment improves, as revenues pick up, our factories get more efficient, we should be able to leverage our cost structure and see that same improvement roughly at the bottom line.

Great. Let's jump gears over to the data center side of the business. And you mentioned a little bit about some unique issues with Artesyn and on the supply side as you were sort of rightsizing that business. The data center business was down. It's been quite a bit lumpy, started rebounding a little bit in the June quarter. Could you talk about some of -- what's happening with this lumpiness? Could we get to more stability in this business? And what some of those unique issues were on the Artesyn side that has caused some of this?

Yes. I guess, first, let me just talk about the demand in that segment. It's very strong, essentially. So what we're talking about, I think, for the next year is our ability to supply to address that demand. And again, that's limited by parts availability. But we're being expedited heavily by the hyperscalers to deliver product. What we're trying to do, quite frankly, is to capture more customers and capture more applications. And we think by doing that, we can dampen some of the lumpiness that's inherent in this business to basically mitigate the lumpiness by bringing in more customers and different platforms. And this should even out the loading over the course of the year. That's our objective.

And how many customers do you have today? We've talked about maybe 4 in the past. Have you been able to increase that number?

We've talked about having 4 customers. I think we have a fifth one in our sights, buts we have not captured that yet. But I think we should have 5 by next year.

Perfect.

Amanda, just to be very clear, we're talking about 4 hyperscale customers in the data center computing. We actually have more enterprise computing customers as well.

Correct. Yes. I meant the 4 large hyperscaler customers. How are you thinking about sort of the revenue trajectory of this business in the second half? Right now, I'm modeling somewhat of a 20%-ish decline in revenue for 2021. Does it seem somewhat reasonable? Or could you see a bigger shift in the second half that would correct some of that decline?

Amanda, I think it really comes down to parts availability and which parts we're able to procure, and the timing of that. We comment a little bit on semi. Outside of semi, there's a lot of shared parts that go across the various platforms. And so it just depends on the timing and the resolution of those. But as we look forward over time, we'd expect the supply to catch up to the demand and for us to see good growth in this market.

So is backlog then pretty strong then in the data center business?

Yes.

Very strong.

Can you talk a little bit about what your differentiators are and how you're able to grow that hyperscaler business versus others in the space?

Yes. So our focus is on highly engineered solutions. And for the hyperscalers, there's a lot of interest in efficiency, obviously, using that much electricity. Every additional percentage point of efficiency translates into big dollars. The second is into density. As we've been into artificial intelligence, the number of processors this year, data intensity per rack doubles -- sometimes more than doubles. And you're trying to squeeze more and more power into a small space. So that's really where we excel. We do a similar thing in other markets, and we basically translate that into the data center market. So as we move forward, what we're trying to do is differentiate. We don't want to be one of 2 or 3 suppliers supplying a box into an application. We'd like to be the only supplier that's capable of doing that.

Perfect. And when you brought up, there's multiple enterprise customers as well in this space. Can you talk about the opportunities in data center outside of hyperscales, maybe in the enterprise space and other spaces as well. Paul -- Steve?

Yes. It's very similar. Obviously, the other customers in this space have the same careabouts: efficiency, power density. Reliability is very important, obviously, for many of these applications. And so I think we're doing well in those markets where we're providing high value-added solutions. These are solutions that work in hostile environments reliably over long periods of time.

Let's shift over to the Telecom & Networking space. The space, I think, has been hit quite a little bit more near term. We've seen some near-term pressure. Is this really just timing of infrastructure spend? Is this also a parts issue? Or is this more on the demand side being able to actually go in and build out those infrastructures?

Yes. I think in Telecom & Networking, there's a lot of moving parts. We did take a portfolio optimization action towards the end of last year, where we basically did last time buy a number of low-margin products. So that hit the revenue going into Q1. What we're seeing now though is definitely an uptick in demand, primarily due to 5G. And so we have some strong design wins in that area, 5G base stations in particular. And so as you start to see the momentum build in the U.S. and Europe, that's basically increasing the demand on AE. Ultimately, our revenue numbers in Telecom & Networking are being constrained, and for the same reasons we're waiting for parts from some key suppliers.

And again, is this another market that could be down double digits this year? Or is there an opportunity with strong backlog to potentially drive higher growth in the second half, pending the ability of supply?

Yes, it's the same answer. This market, again, is impacted by supply constraints. And again, there's a lot of shared parts and shared platforms between some of these embedded power products. And so it will depend on the pace of the part recovery. Now in the longer term, we don't expect as fast a growth out of Telecom & Networking as the other markets. I think this will be sort of a single-digit grower. We've said it's sort of GDP-ish growth in the infrastructure side. Having said that, 5G is a macro driver not only in this part of our business, which is more infrastructure related. But it does drive a lot of activity across all our markets. Obviously, semi is having a big 5G influence with the new smaller chips. Industrial Medical is being influenced by 5G between IIoT. And you also see it in data center with more demand for data as 5G takes -- continues to take hold. So 5G is a significant driver for our business overall.

Seems like the consistent theme here is supply constraints, setting up potentially for a really big 2022. Is there any risk on that side? Say, supply comes back in full force, is there any risk that some demand falls off that you lose some of the potential backlog, if competitors are able to manage the supply chain a little bit better. And this is across the board in all of the segments.

Yes. So let me take a back to back question, Amanda. So going back to my earlier statement, about 70% of our business is proprietary. And so it takes quite a bit for people to change to a new platform. So I think that is pretty well insulated. The 30% that isn't, most of that is hyperscaler and computing-type business. And yes, is there some potential we lose some of our backlog there if we can't perform? Of course. But I think from what I can see, most of our competitors are also being impacted by the parts shortages and have their own issues with COVID. So I think we're all going to be struggling to the finish line here.

Okay. Let's jump over then to the industrial markets. You're in sort of a very wide array of markets, where seasonality is sort of very different. Proportion of sales is very different in each of the opportunities, whether it's medical, architectural glass, industrial processes, food manufacturing, which I think is -- the most strikingly odd one to me is the food manufacturing. But can you talk a little bit about the seasonality of this business, and what you expect? I assume there's going to be some supply issues as well impacting this business. But what your expectations are for the second half of this year?

Yes. Maybe some comments first about the industrial business. I know it covers a broad spectrum of applications. But it's a very interesting business for us because it's largely proprietary. It's mostly small and medium-sized customers. We're asking for us to adjust the standard products to meet their particular needs. And again, the same issues, efficiency, reliability and power density. But each application has a slightly different spin. And so our teams in Asia, our design teams, are very good. They're working with the customer, and come up with a solution that makes sense. The nice thing about these applications is they tend to be very sticky, very long life-cycle applications. And so it's similar to the semiconductor business in a way. Because, again, you're looking at proprietary long life-cycle applications. And so that's why we're trying to grow it. And so we basically doubled the number of engineers working on industrial-type applications. So we expect to see steady growth in industrial in the coming years. I think in the short term, as you mentioned, we're constrained by parts. But I think once we get through this short-term issue with the parts shortage, we have very strong ambitions for this business.

Let's switch gears a little bit. There's been a few management changes over the last year. First up, obviously, Steve, you joined, what, about 6 months ago as CEO. You recently hired a new COO. Can you talk about any sort of changes versus expectations, any mandates that the Board has put in place that maybe would have led to hiring the COO or have changed in your role over the last 6 months?

Yes. So I don't think the Board has given any mandates. Basically, I came in and took a look at the team, and I have been hiring people to strengthen the company. And so we actually started by splitting the business unit. So I have business unit managers now who are driving the semiconductor segment. We call it [ Plasma Power ]. I have a manager driving high volume, which includes hyperscalers, Telecom & Networking and computing. And then I've got a third manager driving industrial. The reason I did that was to bring focus to each area, and each area is a different strategy and a different list of careabouts. And so I'm happy with the way that's proceeding. On the operations side, I wanted to elevate our game, essentially. So I brought in Eduardo Bernal. He actually started today. And he's going to be based in Malaysia and in Singapore. Singapore is his headquarter office and will also spend a fair amount of time in Malaysia, which is our primary factory for [ Plasma Power. ] I've known Eduardo for many years. We worked together at TI. And I also had some dealings with him when I was CEO of Amkor. Eduardo is one of the best operation guys in the semiconductor business. And he also has experience running factories that build subsystems. At TI, he ran their subsystem factory in Mexico as well as the wafer assembly site there. In his last job with TI, he was running their factories in Malaysia, the back-end factories. At NXP, he ran 5 of their back end -- he had responsibility of the entire back end, which included 5 factories in Asia. So he's a big reason, I think, we're going to improve our operations in the coming years.

Last week, you refinanced your debt. Can you provide a little bit more about the details and the rationale behind this refinancing? And does it have any impact on your capital allocation plans going forward?

Yes. Let me take that one, Amanda. So essentially what we did is we were able to extend our current debt deal or debt arrangement for an additional 2 years. In addition, we were able to keep the same terms that we had. And if you recall, when we put this debt originally in place 2 years ago, the financing terms were very attractive at LIBOR plus 75. So we think this is good housekeeping to be able to maintain very low cost debt. We did re-up the amounts. As you remember, we originally borrowed $350 million. So as part of this agreement, we increased that to $400 million, which will add about $85 million to our balance sheet today. We also increased our undrawn line of credit from $150 million to $200 million. So it just gives us additional flexibility as a company both to grow the business for our share buyback and other programs, to take advantage of this very low rate environment and effectively extend the already very good debt agreement that we had in place. And from a capital allocation perspective, there's no change in our overall capital allocation strategy. Clearly, our first goal is to grow the company. But we do have an opportunistic share repurchase plan. That's been a long-standing strategy where we've taken advantage of changes in kind of the market or the stock price to be opportunistic in repurchasing the stock. Overall, our goal with that is to offset dilution. But as you've seen in -- from time to time, we've been able to actually reduce the share count sum as well. And then finally, we do have our dividend, which we started in the first quarter of this year, which is a way to return small but predictable amount of cash to our shareholders.

Perfect. If we look at the different pieces of the portfolio, right, you have semis, data center, telecom and industrial. Excluding supply chain issues, excluding what's happening in that side of the equation, where do you see sort of the biggest opportunity over the next 12 months? And this could be a function of where the backlog sits as well as we look into the early part of 2022.

Yes. I think, Amanda, if I take a look at our 3 big businesses, like semiconductor, Industrial & Medical and then high volume, our objective in semiconductor and Industrial & Medical is to grow those businesses as fast as we can because they have attractive margins. They have long life-cycle products, and the customers appreciate what we could do from an engineering standpoint. So that's where we're placing most of our bets from a development standpoint. In the high-volume space, what we're trying to do is optimize. So we're trying to move more into a proprietary position. We have a number of proprietary design wins in that area, but we want to increase that. So we'll be shifting our resources more towards proprietary opportunities and less on the second or third source opportunities.

With this shift towards sort of business unit managers, does this create a better opportunity for synergies between different product lines? Does it create a more siloed approach or a more sort of unique approach?

I think it's a good team. So a part of my criteria picking business unit managers is assessing whether they can play well with their peers, right? And all 3 of business unit managers that I've hired have a good track record of basically following corporate objectives and not just their business unit objectives. So I like what I see so far. I also have a new CTO, Randy Heckman, whose job is to make sure we're exploiting all of our synergies across the company. And so we have a lot of dialogue between the engineering groups at Advanced Energy. And we're also focused on exploiting synergies with our recent acquisitions, so that we get the most -- from a technology standpoint, of our synergies across the company.

Advanced Energy has talked in the past about an opportunity to sort of cross-sell some products from one market to the next, bringing in new opportunities in either the semiconductor market, taking some more maybe the rugged power into that side of the market and vice versa. Can you talk a little bit about how some of those opportunities have shifted over the last 12 months?

Yes. Yes. We've seen some of that take place in semiconductor, where traditionally we've been strong in the [ Plasma Power ] around the chamber. And now we're getting design wins on some of the products out of Artesyn and the more embedded power parts of the machine. And we're also seeing a fair amount of synergy across the company as we move engineers around the company. There are certain groups optimized for cost, other groups optimized for performance. And sometimes, the best is trying to combine the 2. And so we're seeing that in semiconductors, where we're leveraging some of the cost consciousness of Artesyn to reduce the cost of our platform moving forward and we could pass that benefit on to our customers.

Perfect. We're almost out of time here. Did you have any final thoughts that you wanted to share before we wrap it up?

No. I think you've covered most of it, Amanda. I appreciate the thoroughness today. Thank you.

Yes. I think I would just add, Amanda, just as you've heard today, we continue to have a strong position in our markets, and the demand is very healthy. And we think that as the supply chain improves, that sets us up -- you commented on this early to, do we think a strong 2022. There's no silver bullet. There's not a big snapback, but we do believe we can grow faster than the markets in 2022. And I think that what's happening today underscores our long-term thesis and the demand growth in our market. So we look forward to continuing to see the business grow over time.

Perfect. Thank you so much. Enjoy the rest of your day. Thank you.

Thank you.