高塔半導體 (TSEM) 2004 Q1 法說會逐字稿

Good morning welcome to this Tower Semiconductor's First Quarter Earnings 2004 Conference Call. Today's conference is being recorded. At this time, for opening remarks and introductions, I would like to turn the call over to Chairman and Chief Executive Officer of Tower Semiconductor, Mr. Carmel Vernia. Please go ahead, sir.

Thank you very much. Hello everyone and thank you for joining us for our quarterly conference call. First, I would like to apologize for the technical difficulties some of you experienced on our call on the previous quarter. We took the necessary actions and hope you will hear clear throughout this call. Before I turn the call to Amir Harel, our CFO, I would like to comment that I am very pleased with our results for the first quarter and with our outlook for the second quarter of '04. Our figures clearly demonstrate that Tower has the ability to leverage market opportunities and turn them in to a successful business. Amir Harel, our CFO, will now discuss some of the highlights of our financial results. Amir?

Hi everyone. Before we begin, I would like to remind you that some statements made during this call may be forward-looking and are subject to certain risks and uncertainties that could cause actual results to differ materially from those currently expected. These risks and uncertainties are more fully discussed in the Company's Forms F-3 and 20-F filed with the SEC on Q1 '04 and in our recent press releases on Form 6-K that have been filed with the Securities and Exchange Commission and the Israeli Securities Authority and are available through Tower's website. Tower assumes no obligation to update any such forward-looking statements.

And now, I would like to share with you some of the financial highlights of Q1 '04. Revenues were $27.2 million in Q1 ’04, 37% higher than $19.8 million in Q4 '03 and 116% higher than $12.6 million in Q1 '03. Q1 '04 was the first quarter in which Fab 2 revenues exceeded Fab 1 revenues. The net loss in Q1 '04 decreased to $38.5 million from $46 million in Q4 '03. The increase in net loss versus Q1 '03 is primarily attributed to the commencement of Fab 2 depreciation from Q3 '03. Fab 1 achieved net profit in Q1 '04 for the first time since Q4 2000, mainly as a result of increased utilization.

Depreciation and amortization was approximately $26 million in Q1 '04, in line with our guidance and is expected to further increase in future quarters and significantly impact our net loss. The cash closing balance as of March '04 was $97 million versus $61 million as of December '03. Total cash out relating Fab 2 was $44 million in Q1 '04 including $24 million equipment and $922 million project to-date as of March '04.

Cumulative loans remained at $431 million level; there was no draw downs in Q1 '04. The total balance sheet as of March '04 was $830 million versus $788 million as of December '03. As to equity, we have completed a follow-on equity offering in Q1 '04 raising $80 million gross or $75 million net. Total shareholders' equity as of March '04 was $267 million versus $229 million as of December ‘03.

The company had 65.6 million outstanding shares as of March '04. 13.9 million shares were issued in Q1 '04 of which 11.4 million shares are related to follow-on offering and 2.3 million shares are related to the fifth and final milestone payment by our strategic investors. As of March '04, the company has 16.6 options, warrants, and convertible debentures, including 6.7 million options to employees and directors.

And now to bank covenants about money raising, as a result of the agreement with the banks as amended, Tower is obligated to raise $152 million by December ’05. As of March '04 we have raised 83, mainly attributed to the follow-on offering and the balance of $59 million is due only by June and December of ’05.

Guidance for Q2 '04. The company expects Q2 '04 consolidated revenue to be in the range of $33 million and $36 million. Depreciation for Q2 '04 is expected to be in the range of $27 million and $22 million. And now back to you, Carmel.

27 to 29

Sorry, a correction, depreciation is between 27 and 29. Sorry

Thank you Amir, I will now go into a more detailed status update of the company in the various areas of activity and we will try to address questions you are likely to ask.

On the financial front, the completion of the $80 million follow-on offering during the first quarter allowed us to focus all of our efforts on executing our business plan. We are constantly evaluating our financing options towards the completion of Fab 2’s ramp up and beyond; however, we cannot report on any specific plan at this time.

On the operational front, we currently have an installed tool capacity of approximately 10,000 wafers per month and we are on track with our plans to reach tool capacity of approximately 14,000 wafers per month by the end of ’04. We see rapid improvement in our operational parameters, utilization on scheduled delivery and yields and expect to reach benchmark performance by the end of this year.

On the technology and sales and marketing front, we are continuing to implement our strategy to become a leading foundry service provider in a number of specialized areas, CMOS image sensors, embedded flash and mixed signal ICs.

The latest in our specialized CMOS image sensor is that our 0.18 micron process is production-ready and we currently have three customers in the design and prototype stage with production scheduled for early ‘05.

Joining our marketing team is Dr. Cliff Rowley (ph.), an industry veteran with many years of experience in CMOS image sensors in Motorola and in small camera technology. Cliff will lead our CIS technology offering and roadmap and we trust he will contribute to strengthen our position in this market.

In the area of embedded flash process technology, we can report that our joint development program with Matsushita in Japan is on schedule. The design of the first module, an 8-megabit flash module, has been completed and a few additional modules are now under development. Concurrently, we are transferring the process from Matsushita to Tower, we expect Matsushita to start production this year for its internal consumption and we expect Tower to start production in ‘05 providing foundry services for embedded flash design with the 0.18 micron process mode.

In the analog mixed-signal area, we started volume production of 0.18 micron mixed-signal devices as a process which was independently developed by Tower. Fab 1 also has an important new customer in its mixed-signal offerings. Alliance Semiconductor, our Fab 2 wafer partner selected Fab 1 to produce its new mixed-signal line of products.

On the development front, we engaged in a development project of our CMOS capabilities targeting the rapidly growing market of RFID.

We are pleased to report that our 0.13 micron development program with Motorola is completed successfully. We are currently transferring the process to Tower and installing the first toolsets. We expect to launch our 0.13 micron offering later this year and start 0.13 micron CMOS production in ‘05.

I would like to add one final note before I conclude. In the last conference call, we reported the signing of a $200 million MoU with Vishay Siliconix for the supply of foundry services by Fab 1 over a 7-10 years period. We expected a definitive agreement to be signed during Q1 ‘04. As it turned out, the process of finalizing the definitive agreement extended beyond the expected date; however, I am still very optimistic that this agreement is to be signed very soon and I can report that the teams from both sides already started to work together on the technology transfer process.

With this, I conclude and I would like to open the floor for questions to myself, Amir, or Doron Simon, President of Tower USA, who is also on this call with us. Thank you.

If you would like to ask a question on today’s call, you may do so by pressing "*" "1" on your touchtone telephone. Again, to ask a question, you may do so by pressing "*" "1" on your touchtone telephone. If you are on a speakerphone, please make sure you mute function is turned off to allow your signal to reach our equipment. Once again, that is "*" "1" to ask a question. We will pause a moment to assemble our roster. And we will take our first question from Steve O’Rourke with Piper Jaffray.

Good afternoon. A couple of questions here. I’m assuming as you kind of assess the market environment, and you look out towards ‘05 and we talked a little bit in the past about maybe accelerating the ramp of Fab 2 beyond 14,000 wafer starts per month, have you looked in detail into equipment deliveries and is there a picture that you can provide to us of how equipment deliveries look going out through the end of this year should you decide to do that?

This is Carmel, Steve. We haven't looked into it because right now we anticipate we won't be able to get the equipment even if we wanted to, and we don't have the financial resources to accelerate at this point in time. So, we're on schedule with our original plan to reach 14,000 this year.

Okay, and you mentioned in prepared comments about Motorola 130 nm calls in progress. Is that correct?

No, our development project with Motorola was completed, and it's being transferred to Tower right now.

When do you start taking equipment? I am assuming this is for copper backend of line, is that correct, as part of that process?

That's correct.

Okay, and when do you start bringing in equipment for copper backend of line?

The first toolset will be installed this year. I assume you are not asking exactly what quarter because I am not going to be able to answer you exactly.

Okay, but the second half of the year is when you begin to take equipment and then transfer that?

Yes.

Okay, and utilization of Fab 1?

We have not supplied this figure in the past, but I can say that it's very strongly on the rise.

Okay, and do you -- I’m assuming there is probably some nominal investment required to take that utilization up significantly from where it's been , will that have an impact on any thing going forward?

The only investment we expect to do in Fab 1 at least according to our current plans is related to the Vishay Siliconix view, which requires some specific equipment which we don't have currently in our Fab. Other than that, we don't plan to make any substantial investment in Fab 1.

Okay. Thank you.

Once again, that is "*" "1" for questions. We will take our next question from Darice Liu with CE Underberg, Towbin.

Good evening everyone. Carmel, can you elaborate a little bit more on your CMOS design pipeline activity?

You are asking a difficult question because we usually don’t like to talk about our pipeline and also our customers don’t really appreciate it. So, I will only relate to it in very general terms. We see a lot more activity now than we did a few months ago. I guess, part of it is related to the fact that there is shortage of capacity in other places, all of it is related to the fact that people have more confidence in Tower now than they did six or nine months ago.

Okay. Just on [your grander] scheme, you had mentioned in your last conference call that you expected a new customer to come in in later half of '04. I was wondering if that's still on track and what type of customer penetration rate you’re seeing in terms of other guys?

I am not sure I understand what you're quoting me about in the last conference call. What did I say in the last conference call?

I think there was some sort of commentary that you were meeting informally with possibly a new customer that might come in in the later half of '04?

Are you talking, perhaps, about wafer partners, a customers' investor? Because we are talking to many different customers and we have had quite a number of new customers in different stages of engagement, but not a specific single customer that I recall mentioning.

Okay. Might be my error in that. Can you -- in terms of ramping up Fab 2, can you talk about your yield rates there?

Yeah, I can give a feel for the yield. The yields are product-related, but I can say that on the high -- one of our high-runner product's the current yield is around 88% and we expect it to reach 92% in a few months and this is the maximum that this product can reach. Does this give you a feel for where we are?

Okay. And then just a couple of housekeeping questions. Can you break out headcount? And the shares outstanding wasn't in the press release.

The shares outstanding as of March -- hi, come here. The shares outstanding as of March ‘04 were 65.6 million.

And then headcount?

Sorry?

Headcount?

Headcount is in the range of 1,200.

Thank you guys.

We will take our next question from Quinn Bolton with Needham and Company.

Hi guys. A couple of questions on the Fab 2, can you give us a sense of the utilization and specifically of the 10,000 installed wafer starts per month, how much of that is production versus how much you're still, you know, how may engineering wafers you are still running? Then I have got a few follow-up questions.

Quinn this is Carmel. At this time, we are not offering the exact utilization figure because we are in a very rapid growth mode and we think that this information a) will not help you to do the analysis and b) might harm us if our competitors find out so much information about us. But I can tell you a number of things. First of all, if you would like to calculate then what I think you should take into consideration that from the moment we have an installed capacity say 10,000 at this moment, it takes -- this is when we have the start, so the outs can be only roughly 2 months later and from that you can reduce a certain percentage which is between 10-15% which is the difference between the tool capacity and the actual potential maximum potential Fab capacity and from this you should take out also roughly 700 wafers which go for engineering purposes every month and that’s the maximum possible output of this Fab.

Sorry, when you said that -- can you just [go] for the 10 -15% number that's the difference between the actual starts versus maximum theoretical installed capacity?

Yeah, when you calculate the tool capacity that you purchase you cannot utilize the tools at 100% utilization because then your delivery times are infinite. In order to be at the delivery times which are acceptable the utilization rate is between 85-90%. This is what most Fabs will call 100% is really 85-90% of the theoretical tool ability.

Okay so the 10,000k is a theoretical tool capability in the Fab and we should just [characterize] that by the 10 -15% to get to the kind of the effective capacity and then we back it up to 700 -- sorry 700,000 engineering, or roughly 700,000 engineering wafers for the month.

Right and then remember -- 700 per month and then remember that there is a roughly 2 month delay between [starts and outs].

Okay, great that’s very helpful. One last clarification on that, do you think you will continue to run 700k engineering wafers through the balance of the year as you -- I mean it sounds like the yields especially in the high runner parts were already up to fairly higher levels; do you need to run that number of engineering starts through the balance of the year?

Yeah , you know when I founded that specializes in specialized technology and this require more engineering wafers than in general, more R&D than the standard team of foundry. So may be it will not be at 700 but it will be higher than average.

Ok great and then for Amir, Amir can you just sort give us an update on the cash flow or the EBITDA numbers for 2Q and as you look out to hitting a 14,000 wafer start per month number by the end of the year where you think the EBITDA numbers can go?

As to this quarter the EBITDA of Fab 1 is positive as it was in the last 8 quarters. The EBITDA of Fab 2 as you surely know is yet negative and so is the consolidated EBITDA at that point. We do hope to reach EBITDA on a consolidated basis by the, at positive figure by the fourth quarter of this year.

Okay and I am sorry did you give the specific number for the consolidated EBITDA for the first quarter?

No, we haven't given a specific number but our specific target is to reach a positive EBITDA on a consolidated basis in Q4 '04.

Okay, great. Thanks guys.

Thank you.

We'll take our next question from Ollie O'Ronny with CIBC World Markets.

Yes, good morning, gentlemen. Congratulations on hitting the upside. Clearly your progress as a lot of your revenues now exceed those quarterly revenues of your Malaysian competitors, so you are well on track to ramp up Fab 2 within the industry competitive landscape and, you know, the way I look at your revenues you are getting spill over from the full Fabs in Taiwan, lifting with faster [relation] in Fab 1 and probably driving the upside that you’re reporting in your guidance for Fab 2. What do you need to do at this point in terms of recruiting higher quality or higher, longer customers and are you starting on planning 2005 plans to secure equipment delivery in fast expansion?

Ollie, Hi this is Carmel. On your first question, I would say the following our strategy is to be a specialty product foundry. So initially we are accepting into our Fab customers with plan [CMOS] requirements but going forward we expect a more and more of our product portfolio to be in specialized areas. This is where the margins are better; this is where we can be much more competitive. So, this is that kind of change that you will see throughout the next quarters and in fact the next years. Regarding '05 yes, it is in our growing board. We are, we are evaluating it all the time and we are taking steps to ensure that we have the supply of equipment for '05.

When you look at some of the customers like Motorola and the [process] technology coming in, obviously there are specialized process that are being brewed inside the large IDM’s as well and some of these emerging specialized processes that are in these applications do hit volume opportunities. Could you give us an update on where you stand with IDM customers, large IDM customers, you have had historic ties with National Semi and Motorola of course and you [cited FP] in prior conference calls, could you give us an idea of where your penetration of these IDM's could go, or is going right now?

There is nothing new I can report at this time Ollie, Motorola and National are still very significant customers in our Fab 1--

Are you talking to new IDM's though?

Yes we are.

Or Fab 2?

Yes, we are.

Would you expect any new customer announcements in 2004 on the IDM front or is this something for 2005?

That's a difficult question.

But my -- the reason I asked these questions is you know clearly utilization at the Asian foundries is running close to 100% and at the same time there is a strategy by many if not most of the IDM's, now to go to multiple foundry sources in terms of their own outsourcing supply to limit the rest, and it would just seem that you have the capacity at the right place at the right time, you have your process technologies that are broadening, you know, down to the 0.13 micron node and some of these embedded flash applications and all this should combine for more IDM's wanting to speak to [tower] at this point. Am I wrong on that?

No, you are absolutely right in your analysis but you asked me a very specific question, if there is going to be an announcement design win announcement with IDM this year, on that I cannot be very clear on a yes or a no. We are talking to number of IDM's who ever and when it will materialized I don't know.

Okay, and in terms of your long-term plan at this point, are you looking to fill the nominal capacity of Fab 2 in the 2005 time frame is that part of the discussions you are having?

No, the current plan calls for completing the capacity of Fab 2 by the end of '06. We have a backup plan, I shouldn't call it a backup plan. We have an alternative plan of accelerating the ramp up and with this accelerated plan we can ramp it up to its full capacity by the end of '05 but for this, we need substantially more funding than we currently have and that option is still under consideration.

So, its really not a question of demand at this point its more a question of your own resources?

I would say at this point in time you are correct. Yes.

Great. Thank you very much.

Thank you, Ollie.

We will take a follow-up from Steve O'Rourke.

Hi, has your outlook on the semiconductor cycle changed over this past quarter in any way?

No, can you give me a set of four possible answers so I can mark it.

Just based upon all the input you have gotten over the three months from your customers and kind of looking at things, are you any more positively or any more negatively biased on say, the length of the cycle or how things trend out through '04 into '05. I know it’s a kind of very qualitative type question but…

And it's also very subjective right. I can tell you from my perspective I am much more optimistic than a quarter ago and it has two sides first of all I see much more demand from customers and secondly I see that Tower can perform and execute and that’s very reassuring.

Fair enough and a more specific question here ASP trends, can you give us some commentary on ASP trends?

No, the theory of economics says that when supplies is short the prices go up. We don’t see a very strong trend in this direction and may be we are not pushing that much in this direction because our first priority is to fill the Fab and we will worry about pricing later, but what we can say is that there is somewhat less of a pressure to reduce prices than before, and occasionally we see some higher price from newer quotations. Not on existing business which fully [indiscernible]. But again, I think the important message here is that our top priority is to fill the line just like, when you are -- have an airline, you buy the airplanes and then you want them to fly full of passengers.

Okay, and one other question are you seeing any new competitive entrants in your image-sensor market?

We don't see any new entrant but there are quite a number of them out there already. So, I cannot say that we are in solitude, on the Embedded Flash we see very little competition in the foundry market but on the CMOS image sensor we see quite a number of foundries that at least claim that they have an offering.

Okay, thank you.

Now, we will take a follow-up from Quinn Bolton.

Carmel, you had mentioned the current plan on Fab 2 is to complete [capacity] by the end of '06 but that you could accelerate that to the end of '05. You know, if the financing was available if you looked at, you know, if you make the assumptions that financing is available and really just looked at your ability to fill the Fab in terms of new customers or ramping business with the existing customers, do you think you could get to effectively a full utilization of the 33,000 wafer start Fab by the end of '05, given your current customer activity and engagements with new customers? Thanks.

You know, it's a big challenge, I am not going to, I am not going to tell you that, that it's in the bag but I think it's challenging but it's achievable.

And do you think that would be you know could you comment as to what percent might come from the existing, either the strategic partners or exiting customers versus how much of that would have to be filled by new customers at Tower including non Fab 1 customers?

Existing customers, you mean the ones we currently have or the wafer partners? Its not the same thing.

If I just sort of lump the existing customers and strategic partners into one category, you know, what percent of total starts could they represent versus what percentage would have to be filled by totally new customers to Tower that would need to ramp over the next 12-18 months?

I would say but this is again -- it’s a rough guess from my gut feeling that two-thirds of the customers that would fill the full Fab 33,000 Wafers per month are already working with us and one-third will have to acquire new customers.

Okay. Great, thank you

Once again that is "*" "1" to ask a question, "*" "1" to ask a question. Mr. Vernia, there appears to be no further questions at this time I would like to turn the call over to you, sir

Okay guys, thank you very much. I hope the connection was good this time and looking forward to meet you again in a quarter or so. Bye, bye.

And this does conclude today's conference call. At this time you may disconnect.