Silex Microsystems AB (publ) (SILEX) Earnings Call Transcript

Good morning, everybody, and welcome to the Q2 report for Silex Microsystems. And I am Edvard Kalvesten, and I'm the CEO and Founder. And here, we have also Maria Engstrom, which is our CFO. So what has happened in Q2 for Silex? We have made a successful listing at NASDAQ, and then we also raised the financing for the fab we are buying in U.S. and for growth in Sweden. We have shown a good growth and profitability, growth of 27% and profitability of 32% EBIT margin. We see a stable and growing customer demand generally in all application areas, but specifically in data centers for optical switches. We have signed an agreement to acquire a U.S. Fab, and that has been on the request from the U.S. customers. And this opened up also other new opportunities like defense sector. And finally, we continue to expand in Järfälla to build capacity also in the Sweden fab. So what company is Silex? Silex is a Swedish company with a lot of international customers. We have 57% of the sales is in U.S. We have 30% in Europe and 11% in Asia. We are a pure-play MEMS foundry. That is exactly the same as TSMC is for semiconductors. So we don't have any own designs, no own products. The customer come with the design, and we help them to translate that in a process and a product design. Silex is the world leading in terms of MEMS capability and technologies for all MEMS foundries. And finally, we work with the biggest companies in the world, supporting their innovations to translate that into MEMS innovations. So since several years, we have been the #1 global MEMS foundry. We have about 500 employees, of which around 200 engineers and technicians. And we have a high margin generally, we have been on -- we have a financial target of about 30% and had a growth of 19% in the last 12 months. So what is MEMS? MEMS is a niche within semiconductors. So we are using the same equipment to 80% as semiconductors, and then we add 20% of MEMS equipment. And in that way, you build a MEMS fab. A MEMS chip doesn't only have electrical functions. It also has physical functions. It can have membranes or beams for actuation and also groves and different kind of mechanical structures. Silex vision is to become the undisputable MEMS foundry, meaning that we -- everybody who should build a MEMS product should always come and discuss and ask Silex if we can do it. And that is exactly the same as TSMC is in semiconductors. They have more than 50% market share. Silex is around 15% market share today. So that is our long-term vision. We are present in all different kind of application areas, everything from industrial areas to the data centers with transceivers, switching and different kind of photonics, medical devices, DNA analysis, pressure sensors. We have the pressure sensor fabricated since Silex started 26 years ago. We are getting more and more into consumers in terms of audio MEMS, microphones, loud speakers, microLEDs, all the way to automotives, which includes self-driving cars, drones and different kind of -- all kind of self-driving vehicles. Some key figures for 2026. So the net sales, SEK 393 million, 27% growth since last -- Q2 last year. And for the EBIT, SEK 134 million with a 34% margin. Cash flow is in line with the EBIT margin. If we look at the net sales per geography, you can see that North America, main part is U.S., of course, has grown quite a lot since last year, and that is driven a lot by the data centers and the big tech companies. If you look per end market, you can see that all end markets are growing, somewhat bigger growth in telecom, where we have the data centers. And if you go to per product category, you can see that the development is growing more than production, and that is partly due to that we get in more new customers with IPO with the new fab in U.S., but mainly part to product mix just this quarter. You should remind yourself that you can't just look at a single quarter. You should look at longer periods if you would see how that would change over time. As a market update, you can -- as I said before, you can see that the data centers with all type of optical circuit, photonic solutions is continuing to grow. You can also see in the market that defense and aerospace companies are having requests for different kind of MEMS solutions as well. And that has helped now with the U.S. establishment we are doing with the fab in U.S. And generally, you can see that we follow the same trend as semiconductor has done in the last 20 years. That is more and more foundry business. That is what TSMC has been driving in the last 20 years, taking business from the IDMs to the foundry business, and we see the same trend in MEMS. Here is the foundry ranking for 2025 in terms of revenue, and this is coming from Yole Intelligence report, newly released. And here, you can see that Silex continue to grow faster in absolute numbers than the competitors. We can also see that the foundries, the CMOS foundries, which are also doing some MEMS, TSMC and X-FAB is continuing to grow, but not as much as Silex and also the product companies with some MEMS foundry business like Teledyne and Infineon are slowing down. Generally, in this graph, you can see that Silex Microsystems and Atomica and UNT, these are the pure-play MEMS foundries. Most of the others are mixing, doing both semiconductor and MEMS or mixing doing both products and doing foundry business. So the expansion we are working with now is, firstly, the binding agreement we signed with this fab in U.S., Onsemi. So this is according to LOI we had in the prospectus. So this really enables growth within U.S. and within U.S. customers mainly. And -- but in parallel, we're also growing the fab in Sweden to be able to add around 35% additional production capacity in Sweden. So the fab in Mountain Top, Pennsylvania we acquired. It's a 200-millimeter fab, Onsemi, standard semiconductor fab, they are moving their products from that fab to other fabs, internal and external. This is an asset purchase and includes production facility, the real property, all infrastructure, manufacturing equipment, but we will also take over the 130 employees as of 1st of January 2028. Purchase price, USD 40 million, we paid USD 10 million at signing. We are planning to close that in the end of '27 -- 1st of January '28. And that means that Onsemi can continue to manufacture their legacy products during this time to make -- finish that. And in parallel, we can use them as an outsourcing partner to develop and also produce our products during this '26 and '27. But the 1st of January, we are taking over the fab, subject to approval of CFIUS for foreign investment, which we expect normally that type of approval takes around 3 to 6 months from now. So this is a recap of what we are focused on at Silex. Of course, continued focus on key end markets as we have done in the last 26 years since start. We continue to invest in innovation in different capabilities for MEMS, but also capacity. IPO is now unlocking different type of segments like defense. And the U.S. expansion open up -- even if U.S. market has been our most important market, this open up new opportunities in U.S., not only in the new segments, but also for new type of customers, which really want a U.S. fab, especially now in a geopolitical situation we are now that it changes every day. It's important to have a U.S. fab. So here, we have some time line. So we signed this binding agreement now in '26. We will invest in that fab, build -- adding the MEMS equipment and develop products and start to even produce some products in '27. In total, SEK 1 billion in '26, '27, including the purchase price, and then we will invest another SEK 200 million per year, '28 to '30. And closing again will be around 1st January, '28. Then I hand over to Maria, our CFO.

Thank you, Edvard. So let's go to the next slide, the net sales. We delivered a strong top line growth during the second quarter with net sales of SEK 393 million. This was a growth of 27.3% year-over-year or SEK 84 million. Currency adjusted, the growth was 31.8% and the difference is due to foreign exchange effects. Approximately 60% to 70% of our net sales is invoiced in U.S. dollars. A key driver of the growth was a continued strong demand for development projects, partly following the transition to Swedish majority ownership, but also due to customer mix. Revenue from development projects increased by SEK 61 million year-over-year or 73% of the total net sales increase. North America was our largest market and remained a key growth driver. Revenue in this region increased by SEK 70 million year-over-year or 85% of total net sales increase, mainly due to strong demand for optical switching applications, as Edvard said, and that is within our telecom end market. Looking at the longer-term trend, our rolling 12 months growth ended up at 19% year-over-year and reflecting a stable growth and continued strong demand from our customers. Let's go to next slide and the profitability. We delivered a strong EBIT of SEK 134 million, more than doubling compared with same quarter last year. As a result, the EBIT margin improved to 34% from 18.1%. Excluding nonrecurring items, our adjusted EBIT reached SEK 136 million, compared with SEK 63 million last year. This corresponds to an adjusted EBIT margin of 34.5%, going up from 20.3%. The adjustments primarily relate to IPO-related costs this year and last year, both IPO-related costs and costs due to change of the ownership. We had a strong improvement in profitability, primarily driven by the higher sales, of course. The sales grew faster than our cost base, and that, of course, demonstrates the operating leverage in our business. EBIT was positively impacted by the foreign exchange effects of SEK 19 million related to revaluation of balance sheet items. Last year, the corresponding figure was negative SEK 4 million. Excluding these foreign exchange effects in both periods, the EBIT margin was still good, approximately 29% this year and 19% same period last year. Here, again, looking at the long-term trend, our rolling 12 months EBIT growth was 32%, also, of course, reflecting a stable growth and continued efficient operations. So let's go to the cash flow and financial position. We delivered a strong operating cash flow of SEK 123 million due to strong operating result, but also a positive change in working capital, mainly due to an increase in trade payables due to investments in tangible fixed assets. At the same time, we continue to invest in our new technology and capacity in our Järfälla facility and -- of course, to support future growth and meet increasing demand from our customers. These investments amounted to SEK 98 million during the quarter. Following the SEK 1 billion capital raise in connection with the IPO, we continue to maintain a strong financial position, and we ended the quarter with cash and cash equivalents of SEK 1.3 billion. And this, of course, provides us with the financial flexibility to execute on our long-term expansion strategy. In addition, during the quarter, we secured a SEK 750 million revolving credit facility, which further, of course, enhanced our financial flexibility. However, that remained undrawn at the end of the quarter. So let's summarize this with the financial results. We had the strong revenue growth of 27.3%, mainly driven by development projects and OCS applications. Our continued investments in people and production generated a cost increase of 7.1%, but were more than offset by the higher net sales level. EBIT margin ended up at 34% and adjusted EBIT margin at 34.5%, demonstrating the scalability of our business model. In addition, EBIT benefit of the SEK 19 million in unrealized foreign exchange effects, as I mentioned earlier. This was mainly due to the strong U.S. dollar versus SEK. As a result, we delivered an improvement of SEK 67 million in net earnings compared to the same quarter last year. So let's take a look at the first half, the previous slide. Yes. Thank you. Looking at that, net sales increased with SEK 125 million or 19.5%. Currency adjusted 31.1%, reflecting continued strong demand from our customers. Growth was primarily driven by continued strong demand from, again, our development projects, which increased with SEK 117 million year-over-year, following transition to major majority Swedish ownership partly. We saw continued strong demand also in the OCS applications, as mentioned, particularly in the U.S. We continue to invest in people and production capacity to support future growth. Despite this investment, the higher sales contributed to a significant improvement in profitability, again, demonstrating the scalability of the business model. The EBIT margin improved to 34.1% compared with 22.2% in the second half -- or the second half 2025. EBIT included a positive foreign exchange effect of SEK 32 million for the first half and last year, a negative of SEK 32 million. Excluding these foreign exchange effects in both periods, we still ended up at a very good EBIT level of 30% this first half and 27% in 2025. As a result, net profit, as I said, more than doubled to SEK 207 million and earnings per share after dilution increased to SEK 1.92 from SEK 1.11, first half last year. So let's go to the financial targets. And we delivered good across all these targets. And the net sales on the last 12 months reached SEK 1.5 billion, moving us closer to our target of SEK 2.5 billion by 2030. At some time, we -- our EBIT stands at 32.2% at LTM and are a bit above our medium target of more than 30% EBIT. Finally, our net debt towards EBITDA ratio was negative 1.6x, meaning being net cash positive. This reflects again our strong balance sheet and cash position. Overall, we believe the current performance demonstrates that we are executing well against our financial targets. It's also important to mention that these financial targets, as we have communicated before, do not include the U.S. expansion initiative. So before we go to questions, this is our financial calendar. Our Q3 report will be released on 4th of November and the year-end in February next year and annual report in March. And then I hand over to Q&A session.

[Operator Instructions] The first question comes from Simon Granath at ABG.

Edvard and Maria, congrats on the strong results. While I know that you do not want to comment on specific customer projects, I'm still very keen trying to better understand the durability of the strength in different end markets. So specifically, can you give any color on the demand you are seeing in OCS and Life Science? Is it fair to assume that it's driven by an increased number of customers and on the specific customer projects?

If we start with the OCS market, absolutely, it's driven by increased demand from new customers, so that we see several new customers coming in here in Q2. And they have a window now to get into the OCS for the data centers. It's built a lot of new data centers, not only the initial big Mag7 company, but also the others are looking into building their own data centers. So we see a continued growth there, and they open up a lot of opportunities for new type of companies delivering that type of components. For Life Science, I would say it's more stable and the growth we had in Q2 compared to Q2 last year is rather more product-specific and the current product mix and what happens just with that -- those products -- those customers we have in medical. But generally, it's a strong growth in all application areas. It's not only the data centers. All our application areas are growing this quarter.

And as a follow-up, given that you are now also attracting more customers in the OCS space, are you even -- would it be fair to say that you're even seeing an acceleration in demand?

The new customers are typically development customers. And then we know -- and you know also, Simon, not all of these customers will succeed. We have churn rate of the customers. But some will succeed and some will not. But what we have -- as a leading MEMS foundry for OCS, we absolutely have a big chance to keep most of the winners in that market.

Sounds fair. And on to the gross margin, it's coming down a bit here, which in earnest is in line with your own guidance. Going forward and given the mix you are seeing, is this a more normalized level, i.e., at around 83%? Or do you see any trend shifts as we move into H2 and 2027?

We don't give any forecast on gross margin, but -- and it's too short period to draw any conclusions on a single quarter. But generally, you can say that this -- what we have now is in line what we have seen historically and what we will see going forward.

Perfect. And just a final question from me. Development revenues remain at historically very high levels. And as you point out, there are new customers who are helping those volumes. But I know that you have previously highlighted some bottlenecks for growing this further. Is this something that you might ease going forward?

We are hiring a lot -- mainly it's engineering, which is our limitation today, and it's been the last year. And it takes time to train new engineers to get up to speed. Now we are quite big. We have around 200 engineers at Silex. So we have a faster way to train new engineers. So -- but that is our current bottleneck, and it will continue to be that, but we have just hired 25 new engineers for -- to be trained, and we moved to the U.S. fab to transfer MEMS to that fab. So we are very active in hiring. And that -- actually, it seems we are more attractive, and we get very good talent when we're hiring people now when we are a public company.

The question next comes from Ruben Devos at Kepler Cheuvreux.

Yes. I just had a few on the U.S. fab. I think with the press release last week was mentioned that you would have about USD 40 million purchase, but that's only a small part of the SEK 1.6 billion you expect to invest by 2030. So I was curious whether you could walk us through how much of that is sort of conversion in equipment and sort of when depreciation starts to impact the P&L. Just trying to wrap our head around sort of the D&A and how much of a drag that could be before the revenue comes in.

Very rough. As you said, we are now planning around SEK 1.6 billion until 2030. And of this SEK 1.6 billion, SEK 400 million very roughly is the purchase. And then we have the building, and adding clean room space will take some or the rest of the SEK 1.2 billion, but the main part will be new equipment spread out over this period, starting in the end of '26 and being finished in 2030.

Okay. That's very helpful already. And then I mean, you talked about the EBIT breakeven by 2030 and then sort of the -- by 2034, you would have Järfälla-like margins. I mean we could sort of work our way in terms of what sales contribution that might be associated with, right? But I was...

What we [ helped ] you with it was that both sales and profit margins will be similar to Sweden fab 2025 numbers. So also for the sales, you should look at the revenue 2025 for the Sweden fab.

Okay. Okay. So let's say, SEK 1.5 billion, right, sales by 2034, hypothetically, how do you think about the phasing? Because obviously, you've got 130 of engineers coming in. You have some high D&A initially, like what percentage of that SEK 1.5 billion, let's say, could you already see in the late 20s, right, in the late 2020s, SEK 400 million, SEK 300 million, 20%, something like how do you see that phasing pan out?

We haven't communicated that. But before -- in '26 and '27, we don't take the ownership of the fab. So we don't take all the cost. And that means that we also, of course, not get full availability of the fab. So we will do development and also start production in '27. So that will generate the first revenue chunk. But from '28 and going forward, then we really will ramp up in the fab. I just should correct you somewhat. We didn't do SEK 1.5 billion in 2025 for Silex. I don't have exact numbers.

SEK 1.385 billion.

SEK 1.385 billion.

Yes, just rounded it. Okay. Well, then maybe a final question on the U.S. fab still. I think Mountain Top is around 3,000 square meters of clean room in use and roughly 2,000 more available. So basically, that SEK 1.6 billion plan, it already accounts largely for filling that second tranche, that 9,000 square meter additional? Or how should we think about the capital commitment there?

I remember right. We have around 3,000 there now, and then we have the potential to add another 12,000 square meters. So that is for the MEMS equipment. So the number of square meters for clean room will not be a limitation in that fab. We have very huge capacity there if needed. But that will require more MEMS equipment as well. In terms of the standard CMOS equipment, that will take us far, but we need to add MEMS equipment to match that CMOS equipment.