Good afternoon. Welcome to the 27th Annual Needham Growth Conference. Our next session is going to feature a discussion with Advanced Energy. AE is a leading supplier of precision power conversion products. Many folks are well aware of what the company does. My name is Jim Ricchiuti, Senior Analyst in the Needham Equity Research Department, covering companies in the advanced technology space.
Joining me today from the company is Steve Kelley, President and CEO, Paul Oldham, CFO, Edwin Mok, in the front row, VP of Strategic Analytics and Investor Relations. Before I start with the fireside, I'm going to hand it over to you, Paul, to go through the safe harbor.
Yeah, thanks, Jim. And just a reminder that any statements we make today may be subject to a number of risk factors. We outline those in our various filings. Also, our last earnings release was back in October, so we won't be providing any updates to guidance today. Rather, talking more about the long-term strategies of the company.
Okay, great. And as I mentioned, Steve, I think most of the folks in the audience are familiar with Advanced Energy. But before we get into the discussion a little bit more deeply, it's probably worth providing a snapshot of the company and the major end markets, which we'll go into in a little bit more detail.
Yeah, sure. We're a 44-year-old company. We started in 1981. We're located in Denver, Colorado. So the roots of the company are in high-end power delivery systems for plasma chambers. So we've always been heavy into the semiconductor equipment market, and today, it's about half our business. The other half of our business is focused on high-end applications in the industrial medical markets, as well as in data center, telecom, and networking. Today, roughly 70% of our revenue comes from sole-source products. We're trying to get that to 80% over the coming years.
We had an Investor Day in November here in New York, and we outlined a plan to double our sales between now and 2030, and also improve our gross margin from 37%- 43%, and we think that's eminently achievable, and we're looking forward to executing that plan.
And thanks for sharing, again, some of those metrics that you provided at that investor event in mid-November. Any other key takeaways? I mean, there was a lot of information at that event that is worth going over for folks who may have missed it.
Yeah, so we spent two hours going through our strategy as a company, as well as within each of our business units. And the takeaway is we're leading with technology. So if you look in semiconductor, our growth will be driven by market share gains because of the new technology that we're bringing to market. So there's a couple of products that we came out with called eVerest and eVoS about a year and a half ago. And we've shipped over 250 beta units of these products since we've launched over the past 18 months.
And so many of these design wins that we're earning will ramp to production starting second half of this year, and will basically impact our market share in a positive way starting next year. That was one of the reasons we had the Investor Day in November, because we had developed strong confidence in our ability to grow share in semiconductor, based on technology. If you look in data center, again, it's a technology story. We think we have the best technology that combines power density, efficiency, reliability, all the things that matter a lot to, particularly, AI data center developers.
And so we've got a good play there that's playing out for us pretty well this year. And then industrial medical is more of a combination where it's technology as well as revised go-to-market strategy. We're introducing products at a rapid clip. We're able to customize those products for customers in the industrial medical markets. And then we've gotten a little bit smarter on how we go to market with a new website, easy to use, easy to find our products, a lot more programs with distribution, and a much more dedicated sales force.
So yeah, as we sit here today, and we're hearing this in almost every session, a lot of discussion at the conference about tariffs in China, export restrictions. Maybe turning to the semi portion of the business, can you talk to the direct, indirect impact on AE, and whether it's minimal or whether there's uncertainty as it relates to some of these new actions that we've heard about?
So maybe talk about export control first and tariffs second.
Yeah. So export control, we took our hit back in 2022 with the revised export control regulations that came out in October of 2022. So before that time, we had a low single-digit % of our revenue that was derived from Chinese semiconductor OEMs. After that regulation came out, that reduced considerably. And so the latest export control regulations, they have some direct impact, but it's not much compared to what happened in 2022. Now, from an indirect standpoint, it's yet to be seen, because obviously our customers sell into China.
We have not seen significant changes in our forecast from those customers. So my first answer is I don't see major impacts from the latest export control regulations. We haven't seen them yet. On the tariff side, we've closed two factories in China over the past two years. We're closing our last factory, the third factory, mid-year in 2025. So we'll be totally out of China by July of this year. And a large part of that is due to our customers not wanting us to manufacture in China. So we've moved production to the Philippines and to Mexico.
And so for customers who want products coming out of North America, we've got the Mexico factory. For customers who prefer Southeast Asia, we've got Thailand, we've got Malaysia, and the Philippines. Now, if there are tariffs in Mexico, we have the ability to build any product that we build in Mexico at our factory in the Philippines. So we don't think tariffs are going to have much of an impact on us.
I wanted to also go back to the semi part of the business, only because a couple of your customers were talking about in early December, what their reaction to the new rules, and I'm not going to go into their detailed reactions, but one of the customers suggested that they expect growth in WFE in 2025. I'm just curious what your reaction was to that and what?
I hope they're right.
You hope they're right.
We're not WFE prognosticators. But what we have pledged is that we'll grow faster in WFE. And we think that because we're selling into markets within WFE that are growing faster, such as etch and deposition. We have a market share game plan with eVoS, Everest, and NavEx that's progressing quite nicely. Then we've got other products we sell into semiconductor.
We're number one in the high voltage area. We have a thriving service business. We're gaining share in system power and control and other types of products that sell into the wafer fab. So I think between those three things, we can grow faster. In particular, we committed to growing 30% faster in WFE over time.
Those three factors are the key elements.
Yeah, yeah. Those are the key elements.
Yeah, okay. On the new product front, it sounds like versus earlier product launches, this is fairly encouraging, fair to say, and how should we think about that potential market share gains? It'll become evident to the outside world in 2026?
Yeah, I think you'll see the market share gains clearly in 2026 and 2027, but you start to see things develop in the second half of this year. So we think we're in a good position to grow as the leading edge nodes begin to go to production, so this is 2nm and below, and it's also leading edge NAND and DRAM processes. Now, there's been some delays in some of these processes. Their ramp to volume has been delayed.
That actually works in our favor because it gives our customers more time to work out how they're going to use our technology in their latest processes, so it's actually working in our favor. And we think we'll see share gains over the next five years. I think in semiconductor, it's always going to be incremental. It's not going to be a big leap. But we're in good shape to gain share in dielectric etch and then add to our leading share in conductor etch and deposition.
What is it that you would say really differentiates yourself with these products and has already led to some pretty encouraging beta shipments?
Yeah, so there's a lot of inertia, I think, in semiconductor fab equipment. Customers will not make a change unless they absolutely have to, essentially. And so what happened was as they got to this point where they're building these high-aspect-ratio contacts, it was very difficult for them to maintain throughput and yield with existing technology. And so they needed something different.
And that's why the eVerest and eVoS products were very well timed, because it's giving our customers significantly better technology that allows them to preserve their throughput and yield in the wafer fab. And in the wafer fab business, those are the two key metrics. And that's what's driving the acceptance of our technology, because there's a clear need to do something different to maintain their performance metrics.
A lot of attention on the new products on the semi side of the business, but I think it's also worth spending a moment on some of the products as it relates to the other end markets, I&M and data center computing.
Yeah, so maybe we start with data center. We haven't gone into great detail about exactly what we're doing for data center customers. What I can say is that we changed our strategy a few years ago. Instead of chasing market share, we chose to focus on sole source opportunities and dual source opportunities where we're the clear preference. And that worked out well for us, because the biggest issue with data center is the low margins, or it was the low margins. And so we've increased the margins substantially. So now it's a very good business for the company.
Even though it doesn't average us up, it's still a lot closer to corporate average than it used to be. So we're very happy with the business. And it goes back to having a solid engineering team that can work quickly. I talked about some of the differentiation and efficiency and power density and reliability. But our ability to spin new designs quickly is critical, because the cadence of GPU introductions from NVIDIA is about once per year.
And so we're finding our customers need a new power solution once every 12 months or so. And so that requires us to work very closely with the customer earlier in the cycle and to reuse technologies as much as we can. And so that's basically how we win in data center. And I think we've done very well in the current upcycle.
Can you put it in terms that you've talked about with semi in terms of the way you feel? Can you grow? How much faster can you grow than the market in this area?
Yeah, again, it goes back to our strategy in data center is different than it is in semi or industrial medical. In semi and industrial medical, we're trying to grow as fast as we can, because the margins are very good and the products are by and large sole source. So you can't go wrong. In data center, the issue is if you chase market share, you could end up going down some pretty low margin niches.
And so we're very selective in data center. And what we said in analyst day was that we're going to maintain this strategy of selectivity, but we still expect to grow by $100 million a year by the end of the planned period, which is 2030. So we think we can still grow and maintain our pricing discipline.
Let's turn to INM. That's been an area that's been a little surprisingly weak, frankly, but we're not hearing it just from you. We're hearing it from so many other companies in the space, and I think maybe we all underestimated how much inventory was out there and how long this destocking was going to take. Talk to us a little bit about where we are there.
Yeah, I think first you go back to the supply chain crisis. And unfortunately, the industrial medical customers were last in line. So remember, there are a lot of customers who were waiting for products. And semiconductor equipment makers were first in line. Then you had enterprise and networking data center customers. And then at the end, you've got industrial medical customers, many of whom buy through distribution. And so by the time they got their products from us and from other vendors, there was a disconnect between the actual demand and their supply.
So it created a lot of inventory issues at the end customer as well as at distributors. And so we cleared our delinquencies in Q3 of 2023. And in Q4 of 2023, we went into the correction for industrial medical. That was five quarters ago. So we're still correcting. But we think looking at our distributor metrics, which is important because half our business in industrial medical goes through distributors. The inventory is coming down steadily. Resales after weak Q1 of 2024 popped back up again.
We think we reached the crossover point sometime first half of this year, where we go from a shrinking business to a growing business again. But again, it's going to take us six quarters to work our way out of this problem. But I think the good news is we cleared our delinquencies in Q3 of 2023, so we're out of it this half. I think for others who didn't clear their delinquencies until 2024, it's going to take them longer to recover, essentially.
Does it vary what you're seeing in terms of the channel industrial versus medical? Are you seeing any areas that are performing a little better that give you the confidence about the turn more broadly in the INM?
Yeah, I think it's hard to see much difference between industrial and medical because there are inventory pockets in both sectors, but I would say there are certainly strong pockets. We see a lot of activity in factory automation, and again, we're focused on the high end of the industrial medical market. This is generally for customers who need some customization, so they'll take one of our standard products and we'll customize it for them, so factory automation, robotics in many different applications, horticulture, test and measurement.
Again, a lot of high end industrial and medical type of applications, and so the nice thing about that market is there are thousands of customers, and there's always going to be some customers who are doing pretty well, despite the fact the market is flattish.
Again, going back to the Everest event, you also shared with us some targets about the opportunity to add incremental revenue in that part of the business looking out to 2030. It might be worth reminding people what that number looks like and what drives that growth.
Yeah, so what we talked about was that in industrial medical, part of it's organic, part of it's inorganic. And so we expect to make acquisitions that will drive an additional $500 million in revenue over the next five years, essentially. And most of that will be for industrial medical applications, because that's a highly fragmented market where we think we can do a partial roll up. So that's part of the story. The other part's the organic growth. And I could say we're probably winning our unfair share of new opportunities.
That's a combination of focus from our internal sales teams and applications teams, a lot more new products, and a much better website. So we invested a lot of money in a new website that brought us into the 21st century. Customers can go there, figure out quickly what they could buy from us, and they can actually buy things off our website. It's interesting. We've only had the website up and running for a little more than a year, and we've generated 600 new customers and 150 new confirmed design wins. We're pretty happy with that payoff, because those are very low touch opportunities.
Not as much time spent on telecom and networking at the analyst day. How would you talk to us about plans for that business?
Yeah, you know, telecom and networking for us is not as exciting. We spent a lot of time the last three years pruning commodity business, particularly networking. And so we're left with a business that's roughly $20-$25 million a quarter. We think we can maintain it there at good margins. And we could satisfy the telecom and networking customer base with technology from other parts of the company. But we just don't see a lot of growth there. So I would say that's a secondary market for us.
Paul, this may be some questions for you, but one of the things I think that has stood out for investors is that despite the weak environment, the company's been making progress on the gross margin front. Talk to us about operationally what has occurred, because there have been some significant changes and improvements.
Yeah, that's right. And I think at our analyst day, we were able to articulate a roadmap that gets us to gross margins, not just to 40%, but over 40%, up to 43% or better. There's three things that drive that. The first is we're making a significant effort to consolidate our factory footprint. This was started about a year and a half ago, but ultimately we'll move from about 15 sites down to five large sites that will be able to scale and provide us redundancy across the world.
It'll also provide us common systems and common metrics to drive quality and other things that will actually just, they'll be better factories than we have today. That consolidation, we've already started to see the benefit of it. In the third quarter, we saw 100 basis points of gross margin improvement, largely attributed to just lower cost as we move through that process. Our next milestone there will be to close our last factory in China. It's a substantial factory. And that will get us a lot of the additional improvement from a cost perspective. We expect to close that by the middle of this year.
Overall, the improvement from cost reduction is probably 200-250 basis points starting from the beginning of calendar year 2024. The second driver of gross margins is what Steve's been talking about. We have a large platform of new products coming. Those products add more value. They come at higher pricing. And we believe they'll continue to improve our portfolio. In addition, as we have more sole-source wins in industrial and medical and semiconductor, that will also increase the mix of our products. That's another 200-300 basis points of margin improvement.
Now, we haven't seen much of that yet, but it's coming as we move through this year in the second half, certainly into 2026, as these new products start to gain traction. The third factor is just volume. As volumes improve, given the leverage we have in the model, we believe we'll see about 100 basis points of gross margin improvement for every $50 million of incremental revenue. If you play that out to the top of our model, that's another 350 basis points of margins that comes from volume, just leveraging the other improvements that we've made.
We're pretty excited because this model gets us at kind of targeted revenues in the low 40s, maybe even a little bit higher than that. That as we look forward and markets go through their natural cycles and revenues cycle back down a little bit, we should structurally be able to keep gross margins above 40%. That's a significant part of the value creation of the company. It's getting margins really where they should be and demonstrating both the efficiency of our structure and the technology that's behind our products.
But again, maybe it's worth noting, if we assume kind of the status quo in semi and INM, and hopefully we continue to see decent, reasonable growth in data center, it sounds like you're suggesting there's still some headwind here for gross margins without the volume component.
Yeah, if you exclude the volume component and just say we say at roughly the same level revenues, call it around $400 million to be simple about it, we should be able to get to about 40% gross margin just on our own over the next 12 months-18 months because of the improvements we're making in the cost structure and as the new products come out. You could think about it that any volume increases is just kind of a gravy, right? Just improvement that will naturally get as business increases. Again, our goal is to be over 40% structurally, both in up markets and in down markets.
And again, going back to some of the information you shared with us back at the analyst day, if we think about a blue sky situation, how could we think about gross margins with some of the market share gains? Just maybe frame for us on kind of a peak type margin it could be?
If you think about gross margins and you layer in the volume benefits of the market growth and the opportunities that we have for share gain, which we think are very reasonable from an assumption perspective, we should see gross margins between 43% and 44%. That's what we illustrated, and again, I think that delivers significant value to our shareholders and reflects the underlying value in the company.
How do we think about OpEx just in this environment, both short and long term? I mean, you guys have done a nice job holding down expenses. How do we think about that as maybe as things change from a top line standpoint?
Yeah, it's a good question. I think the first thing is despite the downturn we've been in in our markets for the last two years, we didn't reduce R&D spending. In fact, we increased it and we focused it on some of these priorities. That's allowed us to get these new products and platforms to market at the time when our customers also have bandwidth and time to go through this qualification process. I think it's very important as we've continued to invest in the right things through the downturn.
As a result of that, we don't need another step function up in R&D as we grow going forward because we've continued to make that investment increase it over time. So as revenues grow, we should start to get actually return on that R&D. Secondly, when you look at what we brought to market, these new platforms, they're actually much more efficient in terms of our ability to drive derivatives into the market off of these standard platforms. In fact, we think they're about five times more efficient.
We can turn a derivative five times faster than we used to be able to under our more single product strategy that we had historically. So we should get much more efficiency R&D. From an SG&A perspective, we actually reduced it during the downturn. Peak SG&A to trough was down about 12%. That's grown back a little bit, and we'll see some of that increase as revenues grow. There's a little bit of variable cost, but overall, we expect to increase revenue at no more or increase expenses at no more than about half the rate of revenue growth.
This gives us good leverage between gross margin improvement and leveraging our fixed cost structure. We expect to see 35%-40%, what we call incremental operating margins. In other words, for every incremental revenue dollar, we should see 35%-45% drop to the bottom line.
Do you see, looking at INM and data center, talk to us about the R&D synergies? Is that where they're more obvious R&D synergies?
I think there are synergies across the company. So if I look at our latest, let's say, eVerest and eVoS products in our semiconductor group, they use technologies from across the company. And so that was how one of the reasons we got to market so quickly is we didn't have to invent a lot of new stuff. And so that's key for semiconductor. The other key is working across geographical boundaries. And so the design of these new eVoS and eVerest products didn't just take place in Colorado, right? It also included engineers in Massachusetts, in Minnesota, in California, and in Korea.
So our ability of our teams to work together to get things done and to not reinvent technologies we already have, those are pretty important factors in time to market. And so I also think it's helped motivate a lot of our engineers, because we can get things done quicker, the customers are happier, and they get to interact with people around the world. And it's fun.
The other part that you alluded to of the 2030 target is that inorganic piece. And you highlighted INM being the obvious area where you would look for acquisition opportunities. Talk to us about what it is you might be looking at. Obviously, you can't be very specific. You recently did a small technology acquisition. There are also bigger things out there as well. What would you look for in, say, both in INM?
Yeah, I think more of the same. I think if you look at industrial medical, it's a fragmented market. And there are a lot of smaller players where customers would be in favor of Advanced Energy taking them over, because there are just too many players today in the industrial medical space. It also makes sense because there are some synergies there in the sales side, in the overhead, in manufacturing. So most of these companies are selling products that are pretty well protected. They're also sole source products, long life cycles.
So it fits pretty well into our business. And so we'd like to stay in our lane. Industrial medical is where most of the opportunity is for consolidation. In addition, we're going to make further technology tuck-in acquisitions. The Airity acquisition we made in July has worked out very well for the company. We'll continue to look for companies like that that could augment our existing technology base.
How would you characterize valuations that you're seeing out there? It's always a tough question. Everybody who's on the M&A side usually says they're too high.
Yeah, you know I think we've gone through a period of turbulence, because a lot of companies were dealing with supply chain shortages for two, three years. And they felt that that led to underperformance and therefore undervaluation. Now you're in a period where demand isn't as strong. And there are similar feelings, right? At some point, though, I think the sellers will realize this is who I am and this is how much I'm worth. And I think the buyers and sellers will come together.
Now, hopefully that happens this year. But we're patient. We're not going to overpay for acquisitions. We think we've got a pretty good handle on the market and on the strengths and weaknesses of our targets in the market. And we'll pay a fair price, but not much more than that.
It might be worth reminding people of the balance sheet and your ability to do different types of M&A.
Yeah, so we have a strong balance sheet. We've worked hard to build that over time. Obviously, we generate cash in both good markets and bad. But today we have about $650 million of cash. We have net cash, which a lot of companies don't have. In addition, this year we paid down our term A loan and converted it or rolled it over into a revolving line of credit. We're able to keep the same terms, same attractive terms from that term loan on the revolving line of credit and increase it a bit. So we have $600 million that's available immediately as needed at very attractive financial terms.
So we do have a good balance sheet. We have the ability to tap that to do M&A. And that's how we typically think about it. Now, in the long run, when you think about how we would think about our debt ratios, we'd like to keep our net debt ratio at about three times EBITDA or lower. We would be willing to go a little bit above that depending on the acquisition, because we know we generate a lot of cash and we could just shift and focus on deleveraging immediately if there was a right opportunity that needed a little bit more than that.
Any questions from anybody in the audience, if you don't mind?
New products that you've been excited about, Everest, eVoS, et cetera. Is there any way to kind of qualify for us on the outside? Maybe how excited you are, what you're seeing at this level versus typical? Is there anything that you can share along those lines that might be helpful for us?
I would say we're pretty excited, right? And really that drove our decision to have the analyst day in November here. I had developed a confidence in our projection that allowed us to go and talk to the wider world about our forecast, right? And so our growth forecast in semiconductor is largely driven by Everest, eVoS, NavEx, and other new products that we're coming out with, so our objective really is to win every major slot out there at the leading edge node. And that's in memory and logic.
Now, are we going to win every one? Probably not, but we're going to win most of them. And we're going to see that happen this year. And that's going to set the table for a very nice gain in market share over the coming three to five years. So we're very happy about that. There's a lot of work. These platforms are best in class, but then we have to work with each customer to basically customize those platforms to their needs.
Sometimes we have to do multiple customizations because they have multiple customers. But it pays off. It pays off because we're selling a product that brings more value to the customer, that will bring more profit margin to us, and ultimately help our customers' customer improve their throughput and their yield in the fab. So I think we're in a pretty good spot when it comes to our new products in semiconductor.
I think one data point we've shared externally is that normally when you're going through this process, you sell qualification units or beta units to the customers. Normally, with a new platform launch being somewhere between a year and 18 months into this process, we would probably have between 30 and 50 of these qualification units in the field. We've talked publicly that we expect to ship over 250 by the end of last year. So that's a significantly higher number of qualification units than we would normally see this early.
And it demonstrates to us the customer pull, not only for their own evaluation, but to have that many, they must already be getting into fabs to get qualified at the fab. So I think that's one of the data points we've shared that supports our enthusiasm around these new products. They're getting a lot of looks across a number of customers and a number of applications.
When you acquire in the INM space, can you do more than just solve that seller's problem? Can you reduce this Byzantine level of rules for payment acceptance and all the things each one of these guys apparently has different criteria leading to a whole headache for the purchaser?
Oh, is that right?
I don't know.
Okay. So the question was more, how can we harmonize terms and conditions? Yeah. So I think the good news in industrial medical is half the business goes through distribution for us. And for others in the space, it could be more than half the business. And so the distributors do a pretty good job of telling their terms and conditions to the customer.
So the question really boils down to the direct business and can we harmonize that? And I think we're doing a pretty good job there today with our existing customer base. And we'd apply the same set of rules to any acquired customers. Many times we share the customer with the acquisition candidate. And so we just fold it into AE's terms and conditions.
And if you think about it, we're going to be one of the larger companies in this space. And anybody we acquire that's smaller than us, the ultimate customer gets to consolidate customers and they get to consolidate terms and conditions. And they're not dealing with the small customer and having to worry about credit or other issues. Now they can deal with a larger company that's got the staying power that we've demonstrated over time.
Yeah, and that's a key point. A lot of the customers, the smaller competitors, have a concern, because they typically buy these products for a very long time, 10, 20 years. And they know if we bring those products into our factories, we'll support them.
And we have more leverage with our suppliers. So for some of these hard-to-get components, we have the volumes that some of our smaller competitors don't have. And then the quality performance is pretty important too. So we have pretty high quality standards at AE that are implemented in our factories and that also helps the customer value proposition.
I think we have time for one more. I think we're going to leave it there.
Thank you, Jim.
Great, thanks.
Appreciate it. Yeah.
Thanks, everybody.