Tower Semiconductor Ltd. (TSEM) Q2 2019 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by. Welcome to the TowerJazz Second Quarter 2019 Results Conference Call. [Operator Instructions]. As a reminder, this conference is being recorded, July 29, 2019.Joining us today are Mr. Russell Ellwanger, TowerJazz's CEO; and Mr. Oren Shirazi, CFO. I would now like to turn the call over to Mr. Noit Levi, Vice President of Investor Relations and Corporate Communications. Ms. Levi, please go ahead.

Thank you, and welcome to TowerJazz financial results conference call for the second quarter of 2019. Before we begin, I would like to remind you that some statements made during this call may be forward-looking and are subject to uncertainties and risk factors that could cause actual results to be different from those currently expected. These uncertainties and risk factors are fully disclosed in our Form 20-F, F-4, F-3 and 6-K filed with the Securities and Exchange Commission as well as filings with the Israeli Securities Authority. They are also available on our website. TowerJazz assumes no obligation to update any such forward-looking statements.Please note that the second quarter of 2019 financial results have been prepared in accordance with U.S. GAAP. The financial tables and data in today's earnings release and in this earnings call also include certain adjusted financial information that may be considered non-GAAP financial measures under Regulation G and related reporting requirements as established with the Securities and Exchange Commission. The financial tables include a full explanation of these measures, and the reconciliation of these non-GAAP measures to the GAAP financial measures.Now I'd like to turn the call to our CEO, Mr. Russell Ellwanger. Russell, please go ahead.

Thank you, Noit. Welcome to our second quarter of 2019 financial results conference call. Revenue for the second quarter of 2019 was at our guidance of $306 million, with EBITDA of $70 million and free cash flow of $28 million. Oren will provide full financial details later in the call. The second quarter was our first quarter within the new 3-year contract extension with Panasonic. We're able to offset, to a great extent, the $22 million Panasonic revenue reduction, having achieved good quarter-over-quarter organic growth of 11%. We guide further organic and total revenues growth for the third quarter.We're executing on exciting opportunities within all of our business units, many of which segue into new and large served markets. A particular interest, our 300-millimeter activities have resulted in very strong demand and forecasted excess demand, which we are now investing to meet. This investment has additional benefits that tie to substantial 200-millimeter partnership projects. I'll now review our business for the different business unit activities. In the first half of 2019, our RF mobile business experienced strong growth versus 2018, primarily as a result of strong design wins, one, during 2017 and 2018 in RF SOI for our advanced platforms, both in 200 millimeter and in 300 millimeter. 300-millimeter RF SOI has now ramped to high levels. They've executed on initial small CapEx and other projects to relieve some flow-related bottlenecks to meet present demand. Customer forecasts have grown well beyond our current 300-millimeter capacity. I will address this in a few minutes with our CapEx expansion plans.Of interest to note, our third quarter forecast presents the highest revenue quarter for RF SOI that the company has seen to date. In addition to our successful ramp of 300-millimeter RF SOI, we have introduced an industry-leading 200-millimeter platform, QT9, to address high-power switch and tuner applications. This new technology complements our digitally intensive 300-millimeter process and positions us for further growth in the strong 200-millimeter RF SOI market. We have one initial design slots in this process and anticipate ramping this new technology in late 2020 and into 2021. We expect overall growth in mobile to result from a combination of market share gains, a significant portion of 300-millimeter volume that has ramped recently, for example, are design slots we did not previously hold as well as general RF content growth in handsets due to increased adoption of sub-6 gigahertz 5G standards. And longer term, we view millimeter wave 5G to provide additional drivers for growth as this standard will enable integration of power amplifiers of our technology, first, today where most power amplifiers are built in 3 5 technologies, which do we do not serve.In the first half of 2019, our infrastructure optical silicon germanium business suffered from an industry-wide pullback, which is expected to run its course through 2019. Third quarter shipments will be lower than the second quarter, and we have not yet seen signs of rebound. In addition to inventory correction, the most recent U.S. government ban on Huawei may be further softening the short-term demand in this space. However, data traffic rates only continue to increase for all analyst reports. There's no question that data center growth will return, and hence, we look forward to enjoying and sharing with you the fruits of our high market share due to our differentiated high-end optical device performance in 2020 and beyond.Looking beyond current production, which is dominated by the 100-gigabit optical standard, we have strong design wins in 400 gigabit with new products in our highest performance silicon germanium H5 platform as well as our new silicon photonics platforms. To speed up the 400-gig design cycles for our customers, we announced this past quarter our further enhancement of our design kits to enable the co-optimization of silicon germanium and silicon photonic components in a single design environment. We anticipate that the new 400-gigabit silicon germanium and SiPho product lines, in addition to the recovery in the 100-g market, will provide strong potential growth in an area where we provide high value, and hence, are able to maintain significant ASPs. 400-gigabit products are just beginning to hit the market now and will ramp to significant portion of the market over the next few years.Our power business continues to see a good revenue stream, but, more importantly, an accelerated pipeline of design activity and more highly differentiated areas of high voltage and in 65-nanometer, 300-millimeter BCD. Our 300-millimeter 65-nanometer technology has best-in-class figures of merit for low-voltage power. During the first half of 2019, we delivered initial production volumes as well as continued to attract a large amount of prototyping activity. Decision to providing the ability to integrate large digital content, the platform provides a very low ohm resistance of low voltages, which has allowed some customers to remove their external discrete power fabs and integrate them directly on-chip, saving build-up material costs and board area, while improving our overall efficiency. The strong prototyping activity from Tier 1 customers in this technology has also contributed to our decision to expand capacity in our 300-millimeter factory, which I'll further elaborate on shortly.For higher voltage application requirements, we have won initial Tier 1 designs in our 200-millimeter, 140-volt technology we announced this past quarter. This technology targets the high volumes of automotive, data centers and industrial markets, which demand 48-volt to 125-volt operation, and is unique in that it provides a high level of integration, smaller footprint, using bulk silicon, hence, eliminating the need for expensive SOI starting material and providing an efficient, low-layer comp process flow. Also within automotive, we continue to see strong demand for a family of power management within lithium-ion battery stacks. To note, electric vehicle sales worldwide continued strong growth, driven primarily in China. We serve this exciting EV market and expect to see continual strong tailwinds with our advanced power management platforms within battery management systems. At even higher voltages, we received a Tier 1 design win for our new 200- to 300-volt power management within SOI technology.Looking into our CMOS image sensor business, our largest application and market is the industrial market. As previously discussed, we have seen a pullback, which our customers attribute to the trade war. It is starting to pick up now with new projects, many of which are targeted towards large screen display inspection using very high resolution global shutter sensors. All of our new projects are based on our state-of-the-art global shutter pixels in our 65-nanometer, 12-inch line in Uozu. We expect these projects to ramp towards the end of next year. Orders for present products are forecast by customers to recover with wafer starts beginning during the fourth quarter of this year.We have won a large face recognition sensor project for smartphones. It will be based on indirect time-of-flight technology and we'll use state-of-the-art stacking technology, utilizing our 300-millimeter, 65-nanometer platform. In parallel, for mobile applications, we are working with 3 leading fingerprint companies for under OLED and under LCD optical sensors, based upon our unique pixel technology. These projects are expected to begin to ramp in 2020, utilizing our well-established 0.18 micron, 200-millimeter CIS technology.In the high-end photography area, we're moving along with the next-generation stacking sensor project, partnering with an undisputed leader in the market, targeted to ramp in 2021. Medical and dental x-ray demand has remained stable with strong margins. We see an increased demand for large CMOS-based panels. We're now in the final qualification stages of new products with one of the leading providers. Additionally, we are fully qualified and started to ship single-die wafer-scale medical x-ray sensors on 300-millimeter with 2 additional customers planning final product tapeout in the fourth quarter of this year.Looking to the 300-millimeter capacity expansion. As a result of present capacity and forecasted demand growth of our highly differentiated 300-millimeter RF SOI, 65-nanometer BCD power management and imagers, both smallest pixel global shutter industrial imagers as well as high quantum efficiency stack imagers for facial recognition and high-end photography, we have decided to accelerate our planned expansion and to allocate $100 million to increase the capacity of our 300-millimeter Uozu fab in Japan. Equipment should begin to arrive in this center, with most to all tools expected to be qualified during the first half of 2020. This investment not only increases our 300-millimeter wafer capacity but will drive additional benefits that tie to new and large 200-millimeter partnership activities.In our TOPS business, although seeing an inventory correction pressure, mainly stated by our customers at their distribution channels, our TOPS business revenue comprised predominantly of power discretes, is stable after 2018 growth year, with continual refreshing of new developments with our largest partner and otherwise multiple new activities, which should turn into revenue in 2020.Specific to automotive, during the second quarter, we announced partnership with Lumotive, a Bill Gates-funded LIDAR startup, with successful demonstration of the first beam steering IC for automotive true solid-state LIDAR systems based on Lumotive's unique liquid crystal metasurface technology. Lumotive's complete LIDAR system will also utilize a custom silicon photomultiplier sensor using TowerJazz's cutting-edge single-photon avalanche diode technology.In addition, and very importantly, we announced an expansion of our manufacturing collaboration with our long-term successful partnership with Vishay-Siliconix for next-generation automotive platforms.In terms of utilization, Fab 1 was at 76% during the second quarter; Fab 2 utilization was 80%; Fab 3 utilization for the quarter was about 70% due to the stated decrease in silicon germanium demand for data centers; Fab 9, San Antonio facility, was at about 50% utilization, impacted by discretes. TPSCo blended utilization remained at an average of about 50%, with foundry utilization having gone up substantially, and specifically, Uozu 300-millimeter foundry up 33 points.In summary, we started the second half of the year having shown good organic growth and guiding a 6% organic growth for the third quarter, resulting in a $312 million midrange guidance. We are seeing very strong demand in 300-millimeter, the fruits of past year's developments of advanced 65-nanometer platforms in RF, power and imaging, and as stated, our investment to grow the capacity to support the increased demand with customer forecast showing continued growth throughout all 2020, and then as well for 2021 and 2022. This CapEx expansion should begin ramping into increased revenues at the end of the first half of 2020.At this time, I'll turn the call to our CFO, Mr. Oren Shirazi. Oren?

Thank you, Russell, and welcome, everyone. Thank you for joining us today. We'll start by providing the P&L highlights for the second quarter of 2019 and then discuss our balance sheet. Revenues were $306 million, reflecting 11% organic growth as compared to the first quarter of 2019. Net profit was $21 million and diluted earnings per share of $0.20 and adjusted non-GAAP earnings per share of $0.24. Cash from operating activities was $72 million in the second quarter and free cash flow was $28 million, net of $44 million investments in property and equipment. We are satisfied to see that we were able to mitigate a large portion of the Panasonic revenue and margin reduction for strong organic growth and efficiency measures.Looking at the revenue line, we see that 91% of the $22 million revenue reduction in Panasonic quarter-over-quarter was mitigated from different sources of organic revenue, which went up by a total of $20 million. Analyzing the margins by looking at the operating profit and EBITDA lines, we see that about 60% of the Panasonic revenue reduction impact on our operating profit and EBITDA lines was mitigated by the organic revenue growth and by efficiency measures we took, contributing $30 million improved profitability to these lines, excluding the Panasonic impact.Net profit for the second quarter of 2019 was $21 million or $0.20 per share on a diluted basis as compared to net profit of $26 million or $0.25 per share in the prior quarter.Before moving to the cash flow and balance sheet analysis, I would like to mention the capacity expansion plan in our 300-millimeter fab in Uozu, Japan. Following the recent substantial increase in our 300-millimeter fab utilization and forecasted customer demand exceeding our current capacity capabilities, we have announced a capacity expansion plan in order to satisfy this excess demand for our highly differentiated 300-millimeter RF SOI, 65-nanometer BCD power management and CMOS image sensors platform. Under the plan, we will allocate an amount of about $100 million to increase the capacity of our Uozu fab in Japan. The equipment tools to enable this increase are targeted to be installed during the first half of 2020 and expected to result in higher revenue and higher margin commencing already in the second half of 2020.We'll now provide the cash flow highlights for the second quarter and a balance sheet analysis as of June 30, 2019. During the second quarter of 2019, the company generated $72 million in cash from operations and invested $44 million in fixed assets net, resulting in $28 million free cash flow. Compared to the first quarter of 2019, cash from operations was $75 million and investments in fixed assets net were $42 million. Net current assets, as presented on our balance sheet, namely current assets less current liabilities, was w$808 million on June 30, 2019, resulting in a current ratio of 4.8x as compared to $784 million net current assets and a similar current ratio of 4.8x as of December 31, 2018.Short-term and long-term debt presented in our balance sheet as of June 30, 2019, have increased as compared to December 31, 2018, mainly due to the implementation of accounting standard update, ASU 2016-02, Leases effective January 9, 2019. With regards to lease, right of use assets and lease liabilities, which implementation also increased fixed assets bars. Additional details regarding ASU 2016-02 were also included in note 2Y to our annual financial statements for the year ended December 31, 2018.In addition, the first principal payment of $18 million scheduled to be paid in Q1 2020 for the Series G bonds, which we issued in 2016, was recorded as short-term debt as of June 30, 2019. Shareholders' equity as of June 30, 2019, reached a record of $1.29 billion, which is a $59 million increase as compared to December 31, 2018.Moving to elaborate on the tax line in the P&L, I would like to describe our applicable and effective all-in tax rates. Our U.S. affiliate, Jazz Semiconductor and TowerJazz Texas, which own our Newport Beach and San Antonio fabs, respectively, are taxed at 21% rate starting in 2018 following the U.S. tax reform as compared to 35% prior to that. EPS gross profits for these Japan operations are subject to an approximate 32% tax rate, and our profits in Israel for Fab 1 and Fab 2 operations was subject to a 7.5% statutory tax rate, are not expected to result in any tax payments in Israel for the foreseeable future, since we have more than $1 billion in historical NOLs still to be utilized, which can be carried forward indefinitely. Considering this, and since we have certain tax exemptions, discounts and credits, our all-in worldwide weighted average effective tax rate was 4% for the year ended 2018 and 1.4% in 2019 to date.I would like to describe now our currency hedging activities. In relation to the euro currency, we have almost 0 business in euros, hence, no exposure to the euro. In relation to the Japanese yen, since the majority portion of TPSCo's revenue is denominated in yen and the vast majority of TPSCo's costs are in yen, we have a natural hedge over most of our Japanese business and operations. In order to mitigate the remaining yen exposure, we execute 0 cost cylinder hedging transactions. These 0 cost cylinder transactions hedged currency fluctuations to be contained in a narrow range as compared to the spot exchange rate. Hence, while the yen rate against the U.S. dollar may fluctuate, the impact on our margin is limited.In addition, in relation to the Japanese yen impact on the balance sheet, we have a natural hedge on cash and loan balances since the loans and the cash are both yen-denominated. This helps to protect us from potential impacts of yen fluctuations. And lastly, in relation to the fluctuation of the Israeli shekel currency, we have 0 revenues in this currency, and while less than 10% of our costs are denominated in the Israeli currency, we also hedge a large portion of this currency risk using 0 cost cylinder transactions. A last note on our share count. As of June 30, 2019, we had 106 million outstanding ordinary shares. We no longer have any capital notes outstanding since all were converted into equity in the past. The fully diluted share count is 109 million. The difference between the outstanding and the diluted share count is comprised entirely of ESOP-related options and RSUs.And now I wish to turn the call to the operator.

[Operator Instructions]. The first question is from Cody Acree of Loop Capital.

Congratulations on the continued progress. Oren, if we could start maybe on the CapEx schedule, your expectations on how that money is going to be spent and what should we build into our model?

I believe most of the payments, the vast majority of them will be done through 2020. I think it's reasonable to assume a linear spread over 2020. It could be that up to 5% of that or 10% will be already paid in 2019, in Q4, because it will be attributed to our facility preparation and some downpayments required. But I wouldn't think that we need to increase the CapEx model for 2019. I guess it will be within the number that we know how -- the model which is up to $45 million a quarter. I don't believe we will cross that number in '19. Definitely, 2020, we'll see this amount.

And are there any other areas of your fab infrastructure that you could increase if your demand continues to trend in the [indiscernible]?

Yes, the answer would be. I don't think it's a huge amount of CapEx that will be required. But as we move forward, the different requirements from customers, different products we have from customers sometimes requires a few tools to change the capability of the line. If you recall, when we substantially moved the new Newport Beach Fab 3 from a lot of RF SOI into adding much more silicon germanium, the CapEx spend was relatively small for the amount of silicon germanium capacity that we put in because we added a few unique tools for the capability that drove the movement of the mix. So there's strong possibility continually that we change the mix of different factories. It doesn't take an entire set of tools to do so, but it will always require a few unique tools in order to change the mix capability.

And then lastly, also for you, just on your silicon germanium ramp. Last quarter, you had said that you're getting some anecdotal inputs from your customers that demand was still soft, inventory was still being burnt, but you did expect orders to come in to Q3 for delivery in Q4. It sounds like that may not have materialized. Can you just talk about it a bit?

As I stated, the Q3 revenue for silicon germanium will be lower than Q2, and we have not yet seen signs of a rebound in the market. We have heard customers talk about a rebound, and we continue to, but we've not seen it as far as purchase orders coming in to the same level that we had been having in Q3 and Q4 of last year. So I don't know, really, much more to say on that. I do think, and I'll go beyond thinking, I know that the Huawei ban did have an impact. We were able to mitigate the impact and that's our Q2 guidance was hit to the midpoint. I think it was a bit better than analysts' expectation for the quarter, given everything that was happening in the quarter.But Q2 was impacted by, specifically, the ban with Huawei, with customers having canceled or delayed shipments, canceled POs and delaying shipments, and we're going to see that in Q3 as well, I mean that we know at the moment. However, as stated during the script, there is no question that data traffic continues to grow. There's no question that the devices that we serve are absolutely required for that growth. And we enjoy a very high market share within that segment and a market share that I have not seen any reason to believe that we're losing anything in. Our customers do not believe that we lost anything in it. Maybe, in some cases, to the opposite.So again, I could not state, as the CEO right now, that we see orders coming in that would show that Q4 is a rebound. To the opposite, we do not, and that is what I did state. But it is certainly not a question of if data center growth will begin, it's just a question of when it will begin. And I'm quite convinced that the pullback that we see cannot last throughout much more than this year, and we should start seeing POs that would impact revenues in 2020. Again, the optical space are many layers, so we're pretty much within 1, 1.5 months of start dates for Q4 revenue.

The next question is from Quang Le of Crédit Suisse.

The first would be regarding your 300 millimeters fab in Uozu. Obviously, you are now expanding the capacity and you expect the revenue to already come in, in the second half of 2020. May I ask what is the extent of revenues you're expecting? And what was the utilization of 300 millimeters fab in the last quarter?

We expect the revenue, depending on mix, to be somewhere between an annual $70 million to $100 million increase. As far as the specific Uozu factory utilization, that we have not given. We haven't given it from the start, and we give a blended utilization of the TPSCo factories. To get too specific on any single factory, I think, is maybe giving too much data specifically on our major partner there, which is Panasonic. But we've always, from day 1, as we reported utilization, given a blended utilization. What we did say was that the foundry utilization was up 33 points in the second quarter for the Uozu factory.

I see. Can I ask about your organic revenue? You said it grew 11% quarter-on-quarter in Q2 and you project it to grow 6% quarter-on-quarter in Q3. On a year-on-year basis, how much is that?

I'll look at that, one second, I don't have it off the top of my head. One second here. We'll calculate that and we'll get back to the answer before the end of the call. So maybe we can just move on.

Sure. Can I also ask you about your silicon germanium revenues then? Because you still gave these numbers, such as Q2, I believe, it was like $40 million, you guided at least last quarter. Can I ask what was the actual revenues for silicon germanium, and how much you are expecting in Q3?

We actually didn't guide for it. We give end-of-year breakdown of the different end markets that we serve, but I don't mind giving you the number. So one second here. So your question was Q2 and what will you expect for Q3?

Correct, yes.

Q2 total would have been somewhere about $37 million. And Q3 looks now to be somewhere about $32 million.

$32 million, got it. And my final question would be just on, again, on the capacity extensions. Obviously, you are seeing the growth in your RF SOI and your 65-nanometer BCD power management and then CMOS image sensor. If I were to give percentage-wise of the main drivers, what percentage would you put for each of these 3, if you have any indication on that?

I am not 100% sure I understand the question. Could you rephrase it?

Yes. So basically, you are adding capacity for your products, right? One of them is 300-millimeters RF SOI; the second one is your power management, 65-nanometer BCD; and the third one would be your CMOS image sensor platform, right? So of, let's say, this $100 million spend that you're going to increase the capacity of each of these, what percentage would you allocate to each of these products? Or is there are any allocation?

Our desire is not to allocate anybody within the 300-millimeter but rather to, at this point, increase the capacity and then continue, as stated, the 2021 and 2022 goes continual increase well beyond what we need to have in 2020, which itself is beyond the capacity we have now. So we're looking at multiple tracks to continually increase the 300-millimeter capacity. As far as the most immediate need is RF SOI and the highest amount of volume that's required in the immediate term is RF SOI. The highest mix of volume right now is RF SOI. As far as the very strong prototyping and demand in the number of customers, it's power management, the 65-nanometer BCD. And as far as the stickiness and projects that are really breakthrough-type projects but not the same type of volumes is the CMOS image sensor. Although the CMOS image sensor is many, many layers, and hence, very strong selling prices of wafers and all of these platforms being highly differentiated, demand very good margins.So I think, hopefully, that answers the question. By volume demand, the highest volume, short-term, medium-term and, I believe, long-term is RF SOI, as that's the type of market that it serves. The power management is an area that we've really got through with very significant breakthrough technology, but there's a lot of prototyping activity. And a big win -- a very big win that we have had, that's still growing, obviously, that will demand quite a bit of capacity. At image sensing, most of the capacity growth there is indeed at the end of 2020 and 2021, 2022.

So if I can complete the answer, Quang, for the year-over-year organic growth, so for Q2 actual, historical year-over-year organic is flat positive 1%. And for Q3 guidance, per the midpoint, year-over-year will be an additional 12% positive increase.

The next question is from Rajvindra Gill of Needham & Company.

Question on the gross margin profile as we kind of sift through Panasonic and we kind of offset Panasonic impact. How do we kind of think about the gross margin drivers as we go into calendar '20? You'll be benefiting from a better mix shift, the Panasonic headwinds will be behind you. I just wanted to get a sense of how do we think about gross margins next year? And do we have kind of maybe a long-term picture of what the margins would look like in a more normalized level?

Yes. I think you'll see already the Q2 '19 P&L, for example, the gross margin is almost 17.5% And the EBITDA margin, I think this Q2 '19, forms a very nice baseline from where we will, of course, improve it significantly. And of significantly, incrementally significant because what Russell mentioned before was the ramp is one big driver for growth. And the margins, of course, are -- the incremental margins are, of course, better than 17.5% there, much better. And for 2020, considering the expectations for SiGe recovery and data center recovery, the margins there are even better even than the Uozu fab margins incrementally over the variables. So of course, it will improve. And if you want to model it, I would assume that any additional dollar that you assume in the revenue will result in 50% to 55% blended additional gross profit.

Okay, got it. And in terms of, Russell, the strong growth you're seeing in RF mobile, you mentioned kind of strong design wins that you won back in '17-2018. Just wanted to get a sense in terms of any more details in that segment. Is it if you were to kind of weight it in terms of what's driving the growth this year, is it more market share gains, RF content in the actual phones? And how do we think about 5G being a major catalyst for smartphones next year in terms of number of frequency bands? Higher RF content as a catalyst next year?

So I believe that the present growth that we're seeing is predominantly market share, and it's lots that we did not have before. So I believe it is market share. The continual belief in the growth is both continued market share and very much what you say about 5G. As you look at the sub-6 gigahertz standards, as you referred many more bands which require more switches, more antennas which require more switches and more LNAs. With the LNAs, you have 2 succinct types of LNAs that are made. You have the RF SOI combined switched LNA, to where the 65-nanometer has very strong advantages because of the ability for digital shrink, and you have as well the standalone LNA banks that done in silicon germanium. I believe we are benefiting from both presently and we'll continue to benefit from both. So -- and then, the longer term as really the millimeter wave comes in, I believe I had mentioned it in the script, that at that point, we believe there will be a very strong use of the silicon germanium for the PAs that are now not being done in silicon germanium.

And housekeeping question Oren on the tax rate. You mentioned 4% last year, 1.4% this year. Given the Israeli situation, no taxes in Israel, how do we think about the long-term tax rate for 2020? Is it around 1.4% now? Or is it -- will it be 1.4%, or will it be different?

It's a good question. I think for a model, we should assume something like a 3% because it all depends on the specific region where the growth will come from. And it was on SiGe, so it's higher because it's in Newport Beach. If it was on Fab 1, Fab 2, it's lower, but considering all the NOLs that we have and the tax structure, I think 3% is conservative enough. It may even be lower.

The next question is from Richard Shannon of Craig-Hallum.

I guess I got a few of them here. In the third quarter guidance for sales, it wasn't clear to me whether your RF HPA grouping, whether was that going to grow faster or slower than the kind of midpoint of growth from second to third. I know that the optical may be a little soft, RF SOI offsetting it. Any way you can help us out, thinking about that group, how well that will grow?

I did state that the third quarter would be the highest revenue that the company has ever seen for RF SOI. So I think you could assume that there will all be strong growth there. As far as what the exact numbers are, it's a strong double digit. I mean, well, I think it's strong. It's -- what, hold on a second. Just one second here. About 22% growth.

22%, okay. That's excellent. Pretty helpful. A quick follow-up on the topic of silicon germanium. I think you talked about the 100- and 400-gig transitions going on here. Russell, I just want to make sure that your position that you expect in 400-gig is going to be at least as good from a share point of view as you did in the 100, is that fair to think?

I think it's fair to think that, that's our thinking. It's very hard to forecast a market share that's going to hit its peak maybe 2 years from now. But I see no reason it should stay the same. We have a very strong market share presently. The customer base that we have is working with us on the advanced platforms that will serve the 400 gig. I mean it's probably going to be a 4 by 100. So I believe that, that's the case. Again, I don't like really forecasting a market share, but I see no reason that the market share should be disrupted.

Okay. Fair enough. And I want to follow up on the topic of gross margins, looking out a few quarters here. Obviously, since we had the reset of Panasonic's, I want to get a sense of where gross margins could go as you get more fuller capacity? I think one other way to compare this would be looking back at your historicals from 2016 to '17, where you hit the kind of peak near-term gross margins at 26% range. If we exclude the buildup of capacity in the Uozu fab, is there any reason why we can't get to that or approach that as you get to more full utilization with your current capacity?

I think that there's no reason at all that we shouldn't approach it or exceed it.

Okay. Fair enough. Last question for me, kind of the M&A opportunity as you talked about from time to time in the last several quarters. But, a, I wanted to get your thoughts on appetite there. And b, what do the opportunities look like, both either from a Maxim or Panasonic type of opportunity, or something else like a flat-out acquisition of a company?

The appetite is very good. The financial position of the company is very good to fill the appetite, and I think there's some very good opportunities that are being looked and are at various positions in an acquisition funnel, if you would have it. Certainly, the model that has been very nice for us and I think has been a good win-win was a Panasonic model of a Maxim-type model. I mean, both are very good models. I think that there's opportunities out there of similar models without going into more detail, but I think that's a very good way to be going about growth, that both parties can benefit from that minimize the upfront investment and maximize the financials of both partners.So I think that, again, as I state, there's appetite and there's opportunity. It's difficult because you can never forecast closure of an M&A, but I think there's appetite, opportunity and interest on both sides for things of that order. There's also still a strong opportunity for greenfields in certain geographic areas where you could set up a model that you're going to co-invest some amount of money at the onset, I'm talking specific 300-millimeter at this point, grow at capacity and not take a responsibility for a P&L until you reach a point that you really can be P&L positive, so it's not a burden on the entire company during the first 4 to 6 years of any greenfield as you're driving it to build the capacity, build the utilization. After you build the capacity and reach a P&L positive and maybe even, more importantly, before you get to the P&L positive, to reach a cash flow positive.So I think that those models exist as well, and I believe that there's interest to pursue those from both us and the other sites. So both that models look very, very real right now. As far as outright buying a company, there's certain acquisitions that we'd be looking at for buying a company. But it would not be necessarily the capacity type of an activity that we're looking at here. That would be for incremental capability moving into a new served market. But for an outright acquisition, I can honestly say that there's nothing on the workbench right now that is a full-out acquisition for a large capacity. The models that we would be going after is to gain capability with capacity in an area that we don't serve right now on an outright acquisition. So I think that's a fairly detailed answer, Richard. Does that suffice or...

That more than suffices.

The next question as from Mark Lipacis of Jefferies.

I was hoping, Russell, you might be able to characterize your competitive situation, competitive positioning on the -- in the RF SOI market and the silicon germanium market. And maybe as part of that, if you could talk about what you think the capacity available is in the market in those 2 areas.

Okay. As far as the silicon germanium, our competitiveness, I believe, we're obviously highly competitive. In the optical side, I believe that we sit somewhere around 60% market share, so competitiveness is quite strong. As far as the capacity in the market you're asking or what is the capacity question?

Yes. The capacity in the market, I'm trying to understand, to the extent that when we do see orders come back, to what extent is there capacity in the market outside of TowerJazz, either at your customer's or at your competitor's? Trying to understand what kind of upside you guys could see, should you see the orders come back.

In the area of silicon germanium, we don't have a single customer that has internal capacity. For the area optical, I believe that that's probably 98% correct. I can think of one customer that does have some internal capacity, but they're a very small customer. But for the most part, every one that we serve with regard to the optical, what we make there, the TIAs, the CDR and the laser drivers, I don't believe any has internal capacity. So there's no leveraging of one of our customers right now, keeping something inside and just doing an overflow to us as a foundry supplier. The entire market comes to us for whatever they have, and for the most part, we are the sole supplier for most of the customers in that market segment. As far as outside of us, the other players are fairly well-known. If we have 60% market share, we'd stay to what our capacity is. So that's pretty straightforward. I don't really want to get into talk about who our competitors are, I think it's not really our place to stay. But there's predominantly only two other people that's served into that market. One other is a foundry and the other is an IDM that has its own products.

The next question is from Lisa Thompson of Zacks Investment Research.

I have just two questions. The first is, is there some way to quantify perhaps what percentage of your revenues go to products in the 5G market in 2019, and how that might change in 2020, 2021?

I suppose there is a way to try to quantify it. I don't have the answer for you. I'm not sure that we would really get the answer from our customers who, for the most part, are not the phone makers themselves. So I don't have a specific answer for you how much is right now going into content that's sold as a 5G capability. But certainly, that is only going to increase the more advanced platforms that we sell. It's predominantly set to be able to be used in 5G standards. But I couldn't give you an answer off the top of my head. It's not necessarily an easy answer to get.

Right, right. Do you think it increases as a percent or just dollar-wise?

Both. The content, as we talked about a little bit earlier, the content goes up for the first implementation of the 5G, which is the sub-6 gigahertz because of more bands and more switches and more antennas, hence more switches in LNAs. So content definitely goes up. So that's definitely the case. Does the dollar amount go up per device that you sell? That has a double type of an answer. As we get into -- as we move to 65-nanometer, we were able to shrink digital content. So one of the benefits for the customer is for an LNA to be able to have a smaller area and make more devices per unit area than you could at a larger 0.18-micron, for example. So if we're able to add more value to the customer, we're going to be getting more dollars per unit silicon, which really makes sense. So our unit content capability also drives up the selling price, as well every technology node that you do, because it gives greater capability, you reset and go to the top of the curve over time. Any platform has pricing reduction year-over-year.So one of the reasons that it makes very good sense to be in areas that requires constant innovation is that something that you released 3 years, four years ago, will have had some amount of selling price erosion, but the new platform goes at least to the top of the previous curve, if not higher. And typically, it goes higher if you can add more value by integrating more devices onto the same unit area of silicon. So both is the case. I mean the percent of the mix that's going into this 5G, obviously, as 5G standard starts coming up, you're going to be selling less LTE, so that percentage goes up. The dollar value goes up for what you're selling and certainly the content is going up because of demand for more switches and more LNAs.

Okay. My last question is, could you just give us an update on what's happening in China, if anything?

This is specific to the project in Nanjing? The 200-millimeter project is the question, or just in general?

The project.

Okay. So this -- we have not released anything. And from our standpoint, there's not been a major movement, meaning a next milestone having been achieved or anything of the sort. The government is very involved, and we're still involved in helping the government move forward with different investment groups. Actually, I think there's been a lot of excitement there recently and some, possibly, very viable routes that the fab would be built. But as stated earlier, we are not willing to be the primary investor in this factory. That was never the model nor the intent, to be a partner with the government and to, again, do the things that we had stated at the onset when we had first released, moving forward. We're still very involved and very interested to do that, and in the very big picture, maybe some progress would be made. In the very immediate future, there's nothing that has been signed nor any milestone that's been achieved. So it's nothing that we would talk about is that there is no material event that would impact short-term, midterm or long-term revenues.

There are no further questions at this time. Mr. Ellwanger, would you like to make some concluding statement?

Oh, yes. To begin with, I thank everyone for participation. I thank everyone that asked questions, I think they were very good questions and enjoyed them. We have several events that are coming up very shortly. There's the Oppenheimer 22nd Annual Technology Internet Communication Conference in Boston at the Four Seasons Hotel on August 6, 2019. Dr. Racanelli, the Senior Vice President, General Manager of the Analog IC business unit, will be there giving a presentation in one-on-one meetings, welcome anyone to come to that. There's the Jefferies Semi, Hardware, Communications Infrastructure Summit, August 27, 28 at the Ritz-Carlton, Chicago. Dr. Racanelli would also be attending that. Anyone that is in that area is very welcome to sign up for one-on-one meetings. And then there's the 20th Credit Suisse Asian Technology Conference, September 4 at the Grand Hyatt in Taipei, set aside for many one-on-one meetings. I'll be attending that.So that what's in the next two months, and very happy to meet any and all of you with these different conferences for a face-to-face. As always stated, any questions that you would have or further communications, clarifications or just general discussions on vision and our view of what's going on and how we're moving forward, please contact Ms. Noit Levi, and we'll be very happy to set up calls whenever you would wish.Just a summary statement. There's been many, many good things happening in the company. The questions about gross margins are very interesting questions. I believe that in 2020, we'll see, for us specifically, a very strong rebound and as utilization rates go up across the board as well as higher-value platforms get implemented, that has a very, very strong impact on margins, and we look forward to all of that happening through 2020. I think the activities that we've been able to do throughout this year, in spite of several actions that have been outside of anybody's control, I think had been very good. We're very pleased to have made up for the bulk of the reduction in the Panasonic revenues and believe that over the next few quarters, we'll continue to make up for that on revenue and on a full margin basis. So with that, I really thank everyone. And again, our excitement is there, and we look forward to continue to update. Thank you, very, very much.

Thank you. This concludes the TowerJazz second quarter 2019 results conference call. Thank you for your participation. You may go ahead and disconnect.