TSEM Stock | Tower Semiconductor Ltd. Q3 2020 Earnings Call

Thank you, Nirit, and thank you everyone for joining our call today discussing our 2020 third quarter business and financial results. Firstly, with regards to the cyber event that we announced at the beginning of September, our IT safeguards had identified a security incident on some of our systems. We took immediate actions to prevent damage, closing our Israeli and US IT systems, hence halting those facilities. In less than a week, all factors were returned to operational capability due to the effective procedures. There was no damage to the functional quality of the work in progress, with company and customer data protected. Activities further securing the company's IT environment were put in place. The impact of this event on our operations was between 8 and 12 days of new wafer starts, and as the incident occurred during the last month of the quarter during a demand ramp, we lost multiple weeks of full fab activity levels. This impacted utilization levels for the third quarter, which I will address later in this call. Third quarter revenues were within our guidance range at $310 million, resulting in EBITDA of $79 million and net profit of $15 million. Oren Shirazi, our CFO, will provide an in-depth review of our third quarter financials later in the call. We are guiding the fourth quarter to a mid-range of $340 million, representing 10% quarter-over-quarter and 11% year-over-year growth, or an organic growth of 17% quarter-over-quarter and 14% year-over-year. Looking at our activities in our different business units, within our analog business unit, our silicon germanium optical business has grown throughout the year and is again expected to increase in the fourth quarter. Our third quarter growth was driven primarily by demand for 5G infrastructure. Additionally, we see growth in data center demand. In both markets, we build optical transceivers operating at 25 gigabits per second for 5G infrastructure and predominant at 100 gigabits per second for data centers using high-speed silicon germanium technology. Growth in silicon germanium is supported today by increased utilization of our Newport Beach facility and in the future will be further supported by increased utilization of our San Antonio facility. We continue to see a good flow of new designs in our most advanced technologies targeting 200, 400, and even 800 gigabit per second products to maintain our market as these new standards ramp. At these higher data rates, we also anticipate increased adoption of our silicon photonics platform. We are presently in low volume production at the 100 gigabit per second node with silicon photonics and expect the technology to be adopted more widely in the 400 and 800 gigabit per second transceivers. Our mobile business is experiencing surge growth both in immediate orders and very importantly in longer-term customer forecasts. This strength is broad-based and includes advanced products running in our 300mm facility in Japan and 200mm facility in Israel, and also mid-range products running in our San Antonio facility. Last quarter, we provided an estimate of year-over-year growth in this market of 10 to 15%, but based on the strength we see now, we are increasing our growth expectation for 2020 over 2019 to about 25%. Considering that 2019 was reported at over 40% year-over-year growth in RFS, these high sequential numbers can only be achieved by very strong increases in market share. Growth in this market will continue as 5G handsets are expected to proliferate over the next several years, and which handsets require substantially, namely 30 to 50%, more RF content. Our power IC business is also seeing renewed broad-based strength both in consumer and industrial products, with automotive demand now stabilized. We anticipate exceeding the expectation we had set last quarter of 20% year-over-year growth for our power IC business. This strong growth is primarily a result of market share gains due to strong technology platforms which offer industry-leading performance across a broad range of voltages and applications. This, along with our updated high voltage 200mm RESURF and SOI technologies and industry-leading 65 nanometer BCD 300mm offering, has driven the present market share growth and enabled a healthy funnel of new design activity for future increases. Our power discrete business, predominantly Tier 1 MOSFET customers, has shown a year-over-year decline that is consistent with what has been published for the discrete market. New order levels have stabilized with customer forecasts now increasing, which signals a recovery for this market. Moving to our sensors and displays business unit, first looking at non-imaging sensors, we are steadily growing our manufacturing volume of MEMS microphone products, expecting continued increase throughout the next year, while in parallel co-developing new platform offerings. We have several customer products and platform development stages, such as MEMS speakers, radiation sensors, and remote infrared thermal sensors. Our magnetic sensor TMR activity with Crocus technology is expected to ramp to mass production in the first half of next year with numerous state-of-the-art sensors. On the display front, our development program with Leia has accelerated and is expected to capture a large market share with their unique 3D micro LED technology. We're also working on micro OLED screens for the VR market. Moving to imaging sensors, we have achieved very good results and expect manufacturing to start shortly for lens-type fingerprint sensors. On time-of-flight sensor front, we also received very good results from our lead customer sensors and are moving according to the plan to production in the first half of next year. This would be our first product moving to mass production using our 300mm stacked wafer backside illumination pixel-level bonding platform. We continue to see weakness in two market segments: the X-ray dental sensor market and the industrial sensor market. However, we begin to see a rebound in customer demand forecast for industrial sensors. Third quarter 2020 utilization levels were impacted by the cyber event as mentioned. In Migdal HaEmek, Fab 1, our 6-inch factory, we had 50% utilization. Fab 2, the 8-inch factory, was at 60%. Newport Beach, California, Fab 3 was about 70%. San Antonio factory, Fab 9, was at 60%. Our TPSA 8-inch factories in Japan foundry business was at about 60% rate. In our 12-inch factory, we had a 10-point increase in layers processed versus the second quarter. We also increased our photo layer capability by 25% as a result of the previously announced capacity expansion, and hence the resultant utilization was 70%, allowing for Q4 and continued in 2021 revenue growth against a high customer demand. With that, I'd like to turn the call over to our CFO, Oren Shirazi. Oren, please go ahead.

Yeah, thank you. No, gross profit and our model will be the same, 50-55% like you mentioned. We don't expect any impact from the cyber event on the margins or anything in Q4. It's already behind us.

As I believe stated during the script, in the area of handsets in specific, our RFSOI, we had a very, very substantial increase in revenue percent 2019 over 2018, and that was above 40%. And we're projecting that at the end of this year it will be an additional 25%. And that's certainly much bigger than the market itself grew. So the only way that that can be explained is market share. And all indicators are that handset sales themselves will be very good next year, and on top of that, a higher percentage, almost a doubling of 5G handsets, to where the RF content is somewhere between 40-60% higher than in 4G and lower-end models. So I think we're in an extremely good position to continue to maintain organic growth and strong organic growth in that area. The silicon germanium for 5G infrastructure as well as data center appears that it will stay strong in the next year. Customer forecasts look good. And the power management appears to continue to be at its present rates or maybe higher. The big thing is I believe that the organic growth that we've had, and it was good organic growth, was with the headwind of having a decrease in the power discretes and a decrease in industrial and dental sensors. So the growth that we've had, obviously the numbers we say on organic has to first make up for the deficit in those other two markets, and then it pops up well beyond it. With the power discrete market right now looking that it will be coming back, and certainly the industrial sensors, we have all indication that that's going up. That's customer forecast and as well really backed by market research. We would see that the other businesses that have been strong appear to be going stronger, and that that headwind will be returning to some degree against 2020 results. So I believe that the organic growth for next year will be on a much stronger trajectory than it is this year, although this year I think it's been very good.

On a net, I think that the cyber event actually had a negative impact on Q4, not a positive. Certainly, whatever amount of a few millions of dollars that we didn't ship in Q3 because of the cyber event, we're able to ship in Q4. But missing 8 to 12 days of starts, that's all Q4 revenue. And additionally, without getting too much into the mud of the details, when you have an event at the end of a quarter to where we, by design and fortunately by design, were able to or needed to stop the operations of facilities, you have to move all the substrates to safe end inventory points that are within a few fabs. So you have a lot of the WIP that you've manually moved and you build up bottlenecks at those stations where the line, then when you start it back up, is nonlinear. So where I had stated in the script that we had multiple weeks of the fab not working or the fabs not working to their optimal activity levels, it's because of where you put the fab to shut them down in a safe position and it is then no longer linear. Now on top of that, you have the last month of the quarter and many customers that really need to get their products, so you're now putting a huge amount of energy in the back end of the line that would have been in the front end of the line in order to make up for a disruption in the activities. So another negative impact on the upcoming quarter. So that all being stated, the 340 guidance on a net is negatively impacted because of the event, not positively. And had there been no event, we would have anticipated that the guidance would have been higher. So it's really a very strong demand which I think is a strong attribute to what we've done with our customers, having good products, being in the right markets. Now in the net, I think we're in good shape, but the direct thing, I wouldn't want someone to think that a 340 guidance for Q4 is a push-out of Q3. It's really not the case at all.

So I believe you said Japan, but it's not Japan, it's San Antonio. Sorry, excuse me. Silicon germanium in Newport Beach and in San Antonio. San Antonio is well along on two different platforms for qualification, what we call our H2 and our H3, and in very good shape there. I'm not at this point 100% sure of how much will be run in San Antonio next year. We have the ability to probably increase another 20% on capacity through the capacity that's being qualified in San Antonio. And it's really just a question of the customers driving the qualifications quick enough to be there. What had happened that slowed down the San Antonio qualification, and not the internal qualifications of the platform but the customer qualifications of their parts, was that we brought it up throughout 2019. But at the end of 2019, as you know, the demand was very down for silicon germanium, so customers didn't have a lot of urgency to ensure enough capacity for the future. Newport Beach in their minds was more than sufficient. And at this time, there's multiple customers that are very desirous and excited to qualify additional factory within our fleet to ensure not just hitting forecast but to ensure the ability to supply to short-term upsides.

It's a very big question. I appreciate the question. It's a good question. On the 200mm side, I believe that we have enough footprint to increase organically whatever is needed. If it's adding a few tools, if even expanding a site, there's much that we can do as far as taking gray area into white area should the needs be there. And that was one reason that we extended and expanded the silicon germanium offering from beyond Newport Beach into San Antonio. So we have a big enough 200mm footprint that we can grow, I believe, within the 200mm demand that we'd have over the next years. The 300mm is a different story. And we have done, I think, substantial growth for the 300mm factory in Japan. As stated, it's a very good position to be in right now that we left Q3 after adding photo capacity at a 70% utilization level of a factory that would have been beyond the 85% model that we have had we not added the photo capacity. So we have room to continue to grow Q4, Q1, Q2 of next year, maybe even Q3 of next year. But you're correct, at come Q3, Q4, we'll be hitting somewhat of a limit against an ever-increasing demand. So that must be done inorganically. And how does the, if you will, chess board look on inorganic activities at the moment? For us, I think it looks very good.

I believe our share in silicon germanium is sitting somewhere about 60-65%. I think that's where we sit. The data center, certainly in the past years, there were still 4x10 that were being sold. It's now really almost fully on our side 4x25s, and we are selling higher end at not at very high volumes now, but it'll be going. But I believe what happened in data center is there really was just an over-inventory of parts for whatever reason. And it's very clear that data centers must continue to grow. The amount of information traveling all the time is huge. So I think it really was just there was for whatever reason an over-buy, and even with our own customers an over-buy, but from the data centers themselves that happened at the end of 2018, beginning of 2019, that caused a slowdown of that market. And maybe as well there was a delay in some of the computing activities that they needed to have because of releases of certain companies' parts. But with the release of the new computing capabilities, then everything tied in and it just came back to a normal growth rate.

That's all 2021. I said in forecast, I didn't say in orders. So it's customer forecasts that are coming up, and that would be for the earliest the end of Q4 starts. But yeah, so it would be snapping back in 2021.

Competitive response, I think that the competition that we have within that isn't desirous to dump the market in pricing. So I think that the competition really deals on performance, and that's a place that it should be. I think that our gains are because of performance. And performance always has multiple things. It's the specific technical performance of what you're shipping as well as your ability to really meet customer needs and to tie into customers, to be aligned on their roadmaps, and to be able to serve them very quickly if they have upside demands to be able to get them there. So it's technical performance, but as I'll always state, once you have a good customer, you have to keep your roadmap alive and going with them technically. But the best salesperson you can have is a silicon wafer that ships on time, a good yield. So the performance part isn't just technical, it's the relationship and if they need something to be able to give it and to move that way. And I believe that we've been able to show that with our big customers across the board. So the abilities there, but the biggest growth is our most advanced platforms. It's 300mm.

We're certainly expecting to maintain. And if I look at forecasts, everything is, revenue is still going up quite substantially. So I'm not sure how much of that is related to the markets going up. At this point, what I can state very clearly is that the revenue percentage gain that we had in 2019 over 2018 and 2020 over 2019 is much bigger than the market grew. So that can only be share. I don't yet know what the market growth will be exactly next year. From what we're reading, it would appear that from customer forecasts we'd still be gaining substantial share next year. But until you really see how many units are being sold and what is the content of those units, it's a little bit hard for me right at this point to talk about gaining market share. But I can certainly state that the market share that we have gained will be driving a substantial revenue increase next year.

Yeah, I honestly wish I could give you an exact date, but if I could give you an exact date, it would have already needed to be publicly released. So there's always things we're looking at and working on. The specific question I was asked was what does the landscape look like for deals, and for us, I think it looks very good. But when and how it will be a deal, if a deal does complete, we would have to announce. As stated previously, we always have the ability to increase organically within the footprint that we have, even in Japan. However, at this point, it might be more financially viable to do something inorganically than to continue organically in Japan. But the ability to go organically in Japan still sits as an option. So when I state that we'd be seeing a Q3, most likely Q4, to where we would enter into Q1 2022 maybe with an over-demand of the capability that we now have, that gives enough time to make a decision to grow organically if we needed to in Japan. But the indication and what we would prefer to do at this point is to do something on an inorganic basis with substantially more capacity and nominally possibly a better ROI scenario than continuing at this point to do the next node of growth in Japan. I wish I could give you more granularity, but obviously it's nothing that we're talking about is at a point that it's publicly released.

By the end of next year, we either have to do something inorganically or make the decision to do another organic investment in Japan. And well before the end of next year.

In both 8 and 12-inch foundry, we haven't seen really pricing pressure in 12-inch. However, that being stated, we've been in a ramp position in 12-inch for the past two years, and that's always a good thing, introducing new technologies and new platforms. On 8-inch itself, and as a company blended ASP, we've been able to maintain a stable average selling price per layer for multiple years now. There's always with a good customer over time some year-over-year reduction on previous platforms, but the ability that we have to maintain the ASP is by putting out new platforms that reset the price curve to a higher level, maybe the same level as the previous one or even higher. And that's a benefit to both, the new platform has greater capability and sometimes efficiency so that you can get more chips per wafer. So it's always a win-win, and I think that that's the best way to go. We have never yet raised prices with any customer, and I don't believe we ever will raise prices for customers because of a capacity crunch. I think that that's a very, very good way to lose customers and lose the customer trust. The same as we had one supplier, a big supplier for starting material that unilaterally raise prices to us, and we've done everything we can to eliminate the entire supply from them. So I think when you have somebody in a very crunch position to use that to hurt their business model is not the right way to go. So no, we don't raise prices, and I don't believe that we ever will, at least not as long as I'm here. But we have been able to, through partnership and developing new platforms, maintain a very good ASP throughout the company for multiple years.

I'll do that at the end of the year. I mean, I think we've been pretty granular as to how much growth there are on the different areas. And so maybe off of last year you could calculate that, but we really just give by product group once a year on the actual breakdown, and we give it very detailed. So when we release Q4 in the end of the year, we'll give all of those numbers.

I believe to the extent that we're going to forecast 2021 in this call tonight, that's a Q3 release call. We stated that we see the areas that have grown this year growing at the same rate or stronger next year, and we see the headwinds coming back to some extent. So I don't think there's any reason that the organic numbers we've had this year couldn't be maintained or increased for next year.

Certainly, thank you very much. Well, I thank everybody again for your interest in the company, for participating in the call. I'm grateful for the analysts that cover us and very good questions today, so I appreciate all of it. Thank you very, very much. In summary, the industry is experiencing a strong surge in RF demand as the handset market is rebounding from the COVID-19 slowdown and new models are ramping and 5G adoption is beginning and accelerating. Unlike 2020, 2021 is expected to be a strong growth year in overall handset unit sales with even stronger adoption of 5G. 5G-enabled handsets are anticipated to nearly double in 2021 versus 2020. If we consider the RF content increase in the 5G handsets, the RF supply growth should be well beyond a forecasted high handset market growth. And as I believe stated well because of the questions that were asked, or at least the questions were good, our share gain over the past two years should put us in a very good position to enjoy both the RF content increase on top of the market rebound for handsets as a whole. Power markets, both IC and discrete, are expected to see growth in 2021, in part due to recovery from COVID-19 slowdown in handsets as well as infrastructure, automotive, and industrial markets. Analysts anticipate a 10% rebound in automotive sales in 2021 after a down year in 2020, and it is expected that electrical vehicles, which include more of our content, will grow at an even faster rate. We have experienced a slowdown due to the COVID-19 pandemic situation for dental X-ray and due to geopolitical situation for industrial sensors market. We see signs of recovery in the industrial sensor market and expect this market to rebound in 2021 based on our customers' inputs and strongly supported by market research. Considering the very high customer forecasts in what they're demanding in 2021, backed by market research, we look forward to a strong 2021. And by strong, I mean a very strong demand and growth. So that really does more or less conclude my comments, other than stating we started this week our 2020 Annual Technology Global Symposium online event. The online event consists of a series of webinars presenting the company's premier analog technologies, manufacturing solutions, and recent developments across all of our platforms. Similar to our traditional TGS event format, the webinars are presented by Tower's leading team of experts with the added benefit of direct and easier access to interact online with the presenters and learn more about the company's platforms and offerings. All information about the timing and the specific topics of the seminars are on our website. Sign-up information is on our website, so there's very strong attendance at these, and the fact that it's virtual, it can be as big as you wish. So everyone is invited to look at these and sign up. In addition, on November 30th, 2020, Dr. Marco Rinaldi and Dr. Eli Sturm will represent the company at the Credit Suisse virtual conference. With that, I say thank you very much. Wish everybody really health, peace, security, happiness as we head into the holiday season. Thank you very much.