高塔半導體 (TSEM) 2002 Q4 法說會逐字稿

Welcome Ladies and gentlemen to the Tower Semiconductor Fourth Quarter and Year-End Conference Call. At this time, I would like to inform you that this conference is being recorded and that all participants are in a listen-only mode. At the request of the company, we will open up the conference for questions and answers after the presentation. If you would like to access the replay after the call, please dial 1-800-428-6051 or if you are calling internationally, please dial 973-709-2089 with the pin number of 286732. I would now like to turn the conference over to Dr. Nissan Cohen. Please go ahead.

Thank you. Good day and welcome to Tower Semiconductor Quarterly and Yearly Conference Call. I am Yoav Nissan Cohen and with me is Amir Harel, our CFO. I will focus on general statements and I will let Amir discuss the financial results.

I am very happy to review with you what I consider to be a good year for Tower. I know we cannot be too happy about the year in which our share value dropped almost 50%, but just to keep us all in perspective, no foundry was favored in 2002. If you look at the three leading foundries TSMC, UMC, and Chartered, we see that during 2002 the respective share price declined even more by 59%, 65%, and 85%. I would like to take a moment and review the main achievements of the year.

Despite the tough business and financial conditions, we made significant progress throughout the year and laid a foundation for the future of the company. Thanks to the dedication of our people and also to the commitment of our partners, we managed to fulfill almost every goal we set ourselves for the year. I will start from discussing our Fab 1.

It is on this Fab that we generated all our business until today. We surely take pride in its performance and are proud that it was selected by Motorola as a foundry of the year for three years in a row. It was very important for us to return Fab 1 to positive cash flow, which we did with the combination of strong management focused on cost reduction and new business mostly in the image sensor area. The jewel of the crown of Fab1 is our new image sensor with the 14 million pixels, which is being used by Kodak for the most advanced, professional digital camera in the market. Sales of image sensors in Q4 climbed to 20% of our total sales and are expected to count for about 30% of our Q1 sales. We expect the CMOS image sensor products will generate additional business for Fab1 in the second half of 2003 and also down the road with an enhanced offering of our specialized technologies in Fab 2.

Looking forward, clearly our Fab 2 is the key for any significant growth in the future. As of today, we have already invested more than $600m in the construction and tuning of the new Fab. We are running prototypes for our customers for quiet some time. And we will commence limited volume production in Q2. This is somewhat later than our original plan, mostly due to the slow market conditions. Nevertheless, the Fab is now ready for production orders, based on the number of prototypes we are currently running, we see the demand storming up in the second half of the year, and therefore we continue to bring in new tools and add production capacity. Our plan is to increase the capacity from the current level of about 2,000 wafers per month to close to 10,000 wafers per month early next year.

Another crucial element for the success of Fab 2 is the technology front. We are working in parallel on a few projects. Our strategy is to partner with the industry leaders to advance our technologies. Initially, we transferred the 0.18-micron process from Toshiba. We added mixed-signal and analog options to our base technology in co-operation with IMEC(ph) in Belgium. With adaptation of our basic 0.18 micron technology to the foundry standard was done in co-operation with [inaudible]. The development of embedded FLASH option for our 0.18-micron process is currently being done in Japan in co-operation with Matushita Semiconductors. And at last, we developed our 0.15-micron process in co-operation with Motorola.

In our industry no one can develop all the technologies alone, therefore, working with leading partners is one of the keys to being successful. Our teams in Kyoto, Japan and in Austin, Texas are working on our future technologies in cooperation with these kinds of industry leaders. Further technology alone is not sufficient and we also made special efforts to enhance our design support offering. We engaged with all the leaders in the Silicon IP industry to provide libraries, compilers, and design tools to enable our customers to implement their design in the most efficient way. To list few of the names of our design partners, I can mention names such Cadence, Synopsis, Avizan(ph), Vira Arm(ph), and many other companies.

And last but the most important all of these efforts would not mean much without the appropriate business organization, to position the company in the market and bring the customers. We acknowledged the need to strengthen this area. To become a major force in the foundry market, we needed to reorganize the business operation and build marketing and sales force, which will be able to grow sales to $0.5b and beyond. With this in mind, we spent a long time searching for the right executive and now we are extremely pleased to announce that Harold Blomquist(ph) has joined Tower as a Senior VP for business operation and also as CEO of Tower-USA. Harold brings to Tower and an impressive record of proven success in marketing and sales of semiconductor IC and foundry products and services. He is a great addition to our team and I'm confident that he will contribute much to our success in the future.

It is clear that one of the most demanding tasks in 2002 was to obtain additional financing for Fab 2. Considering the limitations in the financial market, we were very fortunate to have the consistent support of our strategic and the financial investors. During 2002, they came forward, advance payment and even investment -- invested more than originally committed by them. As a result, we needed to modify the terms of the investment to reflect the market conditions.

At this stage, I concluded my opening statements and I would now like to turn the call over to our CFO, Amir Harel for his talk. Amir.

Hi everyone. I would like to share with you some of the financial highlights of Q4 and financial year '02. Starting with the sales, Fab 1 sales in Q4 were $11.5m versus $12.2m in the previous quarter. For the year '02, the sales were $43.7m versus $52m in '01. bFab 2 revenue was $4m in the quarter and $8m in the whole year, relating a technology deal. Consolidated sales in FY ’02 were $52m similar to those of '01. The gross operating net loss in Fab 1, while comparing the gross operating net loss in Q4, of Fab 1 to same parameters in Q4 ‘01, we noticed that although sales in Q4 were only $2.7m higher, the gross loss was $1.4m versus $5.3m and the operating gross was $2.2m versus $7.4m, and the net loss was $2.8m versus $4.8m. Such analysis reassumes our significant cost reduction strategy in Fab 1. The consolidated net loss was $13.6m versus $2.3m in Q4. The loss in Q4 includes $10.8m Fab 2 expenses net versus $4.5m in Q4 ‘01. The consolidated net loss in '02 was $51.4m versus $38.5m in previous year. The loss in '02 includes $37m Fab 2 expenses net versus $14m in '01. Cash flow from operations in Fab 1 was positive $1.9m in Q4 versus $1.4m in the previous quarter and versus $0.1m negative in Q4 '01. Cash flow from operation was positive $2.7m in FY '02 versus negative $1.4m in '01.

The consolidated cash closing balance as of December '02 was $82m versus $65m as of September '02 and versus $33m as of December '01. The total balance sheet increased to $716m in December '02, out of which over $600m relates to Fab 2, from $630m in September and $470m in December '01. The increase is primarily as a result of increase in bank loans and the profits from equity and convertible deals. The cash in amount described above were invested in Q4 primarily in Fab 2 equipment and in financial year '02 primarily in Fab 2 equipment and construction. Analyzing the cash out of Fab 2. In Q4, we have spent $79m including $48m for equipment. In year '02, we have spent $270m, including $136m in equipment. And project to-date, we have spent over $600m, including $265m in equipment.

And now to the equity. Total shareholders’ equity as of December '02 was approximately $300m. As of December '02, the company has 33.4m outstanding shares and 4.6m options including options to employees, directors and banks. Such count excludes potential shares underlying the convertible debenture deal and options granted to the wafer partners and financial investors including OTPP. In financial year '02, 18.4m shares were issued, out of which 13.2m in the fourth quarter. In Q4, we have raised $80m as follows -- $45m from [inaudible] from the Fab 2 investors, out of which $15m were recorded as free-payment and the remaining $30m as equity for 6.1m shares. And $20.5m from the rights offering deal, and $15m from a Canadian investor OTPP, for 3m shares.

And now to the recently signed Milestone five agreement. In February, we have signed an agreement with the four major investors, amending the terms of the fifth and last investment milestone. The main business terms are as follows: the investment is not conditional upon the achievement of milestone five and the requirement to bring new wafer partner in the amount of $15m. The milestone five investment will be done in two installments. The first installment 60% or $24.2m will be paid in [inaudible] after shareholders’ approval, at the share price of approximately three(ph), generating additional shares approximately 8m. The second installment 40% or $16.1m will be paid upon the later of August 1 '03 and the day Tower raised additional $22m, which should be done not later than December '03. Dollar per share for your second installment will be similar to the price of raising the $22m.

Wafer partners may convert the pre-payments resulting form milestone four currently totaling $13.2m, which will be unutilized by December '05 into Tower shares, based on then market price. If the wafer partners convert credits into over 5% of Tower’s ten outstanding share capital, there will be a rights offering to enable all shareholders to participate based on the price of share at which the credits are converted. The agreement is subject to the approval of shareholders’ meeting, banks, and the Israeli investment center. The company believes that such approvals are probable.

I would like to remind you that some statements made during this call may be forward-looking and the actual results may be different from what you've currently expected. The risks factor which may influence of our results are listed in our most recent registration Form S2 as filed with the SEC and RST(ph). Yoav.

This basically ends our opening statements and we are now ready to answer your questions.

Thank you gentleman. The questions and answer session will begin at this time. If you are using a speakerphone, please pickup the handset before pressing any numbers. If you are within the United States please press "star" "1" on your pushbutton telephone. If you are calling internationally or from Israel, you may also press "star" "1" to ask your question. If you wish to withdraw that question please press "star" "2". Your questions will be taken in the order that they are received. Gentleman, please standby for your first question. Our first question in queue comes from Ali Irani with CIBC World Markets. Please state your question.

Yes. Good morning gentleman. A number of questions for you. First I was hoping you can repeat the operating performances related to Fab 1 and Fab 2. It was just running a little fast for me? And secondly, I was hoping you could give us an idea in addition to the [ICMOS] of the different products and the different line which you were at in the fourth quarter and for the year-end in Fab 1, and where you see that migrating in the first quarter?

Okay. Let me give you some -- good morning Ali.

Good morning.

Let me give you some information about our Fab 1. Basically, we've seen more weaker start in Q4 as compared to Q3. Most of the increase in the business is coming from the CMOS image sensor and in particular in the 14 million[mega]pixel image sensor, our sales were divided more as to close to 30% was done different non-volatile memory business, 20% as I said before, 20% went to the CMOS image sensor.

This is as compared to 10% of our business in Q3. And about 16% was done in mixed-signal area and the rest is done in standard CMOS. So the most -- the fastest increasing segment of our business right now in Fab 1 is with CMOS image sensor. Geographically, I can tell you that 70% of our sales went to the U.S. it is a typical number, about 10% in the Far East, and about 20% it went to Europe. I don’t have the exact division in terms of the geometry, but the reason major change relative to what we were before still 0.35 micron is a relatively small volume and then the 0.5 and 0.6 are similar. I guess that the other thing that which I should mention now is that we were able -- even though sales in terms of revenue were a little bit less than Q3 in Fab 1 by about half a million less. Then we managed to increase the positive cash from operations from the facility, it was about $1.3m -- $1.4m in Q3 and close to $1.9m in Q4. So I think that right now the Fab is running on a totally stable conditions and we expect the next quarter, Q1, to pretty much look similar to Q4.

At this point your visibility in Q1 points to flat orders, I guess, you are saying and I am wondering what kind of visibility you're beginning to get in wafer orders for the second quarters as well?

Well, it is still too early to really, you know, predict the second quarter. This is typically what we are most -- where we have most of the difficulties is to predict the next quarter, I mean not the coming one but the one after this. So I can say quite [inaudible] that this one is flat but I would be reluctant at this stage to say exactly what we are going to see in the next quarter, but I don't expect anything dramatic okay, which means that it may be a smaller decline, it may be a small [inaudible] fledge, but nothing should be dramatic. We don't see any -- at least at this stage, we have no indications to anything which is dramatic.

And just as a matter of housekeeping, could you breakout first the headcount at present associated with Fab 1 and Fab 2 at your end and I was hoping if you could also give us the depreciation figures for the fourth quarter?

It is Amir. The depreciation for Fab 1 was $4.2m, for Fab 1 and for Fab 2 currently no depreciating yet until this Fab will be operational, fully operational accounting wise and that's when the Fab 2 depreciation will start rolling in.

It is a good question related to the depreciation because it is an accounting question of a when exactly you start depreciation of facility and we will do it when it is going to be considered fully operational in terms of some level of production and some level of customers and at this point we will only then start depreciating the facility. In terms of headcount, we are currently at the proximity of about 1,200 employees and out of which about 750 are already in Fab 2 projects or Fab 2 related projects and about 440 employees are currently employed in Fab 1.

Great. Thank you.

Thank you.

Okay. As a reminder ladies and gentlemen, if you are calling within United States, please press "star" "1" on your push button phone to ask a question. Also if you are calling from Israel, you may also press "star" "1" on your phone to ask a question. Our next question again comes from Ali Irani. Please state your question.

Yes. Good morning gentlemen. It gives me an opportunity to follow-up. Could you give us an idea in adding(ph) equivalence how many wafer starts you are running in the fourth quarter, just to put that in perspective with the Fab 2 contribution going through next year? And looking at some of the other products, I was hoping you could spend some time on the embedded Micro FLASH, obviously that's a significant opportunity for growth growing into 2003 and I'm wondering where you are progressing in terms of qualifications with customers, both with [inaudible] and other potential customers.

Okay, in terms of for wafer starts, we are running the Fab 1 at the order of about 8,000 wafers per month in 6 inch wafers. The embedded FLASH, is a -- it is a great opportunity to the company, but I have to also say that it is business which is not easy to develop because every product that you bring up requires adjustments and it requires a adjustment to the specific customer and needs and requirements. So, it takes a lot of effort from our design support team to verify that there is a good match between the products and between the needs of the customers.

We are very hopeful, when it comes to the projects that we are currently running together with Matsushita in Japan, specifically we are now designing an 8 mega-pixel embedded FLASH module which many customers in the world are interested to get and you practically don't see these kinds of technology offering in the market. At least, we are not aware of any one who is currently offering this kind of embedded FLASH technology of high density FLASH and embedded in high density logic, so we believe that in the future, it can become very important for us, but it is going to take some time. I mean, we cannot hope to see this business, let’s say ramping up in '03. This is a business for '04 and not for '03. In '03 we will need mostly to continue to focus on either standard product or mixed signal products and we do expect to see earlier, some of the earlier CMOS image sensor in Fab 2, but which is also a development project which is likely to take more time. So our strategy remains to start from the-- what we call the twin vanilla technology introduced then watched the line with running this kind of standard technologies and at the same time developed the specialized technologies and [rent] these technologies only at a later stage.

And in terms of signing up incremental customers for Fab 2 and prototyping that you are mentioning. Can you give us an update of how that is progressing.

Right now most of what we do, in fact we have on top of the wafer partners, we have -- right now we work with two additional customers on prototyping, and there are few more on the pipeline. I think that the '03 is going to be year of focus on what I call for a long time already, a friendly customer. And this is where we are going to actually bring the Fab to production, bring the Fab to substantial number of wafer starts, and also, mostly in '04 you'll see the ramp. Lets say there is --starting Q4 ‘03 and then '04, you will start to feel ramping related to the other customers that we either are bringing up now or prototyping for now, or we are going to prototypes throughout the year.

We are very, very excited about Harold Bloomquist who has joined us -- I mean we have already seen a major change in the enhancement of our sales strategy and getting more opportunities. There will be more news coming up in the next few weeks about more activities that we are doing with that respect. Just to position Harold where he came from, Harold has a -- I would say that probably the most important -- the major experience that he had was at AMI and he practically took AMI from about $70m sales to the range of $400m sales, leaving the company [S-words] at the beginning only in the sales, and later on in all of business operations and product lines. So he is -- from the experience in AMI, AMI is both operating on a foundry model and on an [A-SEC] model so he's very experienced when it comes not only to foundry, but to the issues related to the design support and how you create more opportunities for -- in particular also for Fab 1, not just for Fab 2, and we really feel that this kind of seasoned semi-conductor professional was needed in the company, and we are extremely hopeful about the impact of Harold's activities during this year.

Could you also, looking out for next year give us an idea of where you would expect the 10,000 wafer starts in Fab 2 to be split in terms of line widths and one last question on share count. Could you give us an update of where you stood at the end of the quarter and where you should be going this quarter.

Okay, I'll -- with respect to the -- when we get to the 10,000 wafer starts which currently seems to be at some point earlier in 2004 based on our current plan. Then, I would guess that still all of it will be at 0.18 micron and we are still not going to actually do production at that stage. So, probably right now the way it seems the first 10,000 wafer starts will be done at 0.18 micron. And when we look at the market, we practically see that the market has relatively slowed down the transition to the newer technologies with respect for mass production of volume products. And you see more and more companies stay in the older technology node because they can get more efficient production and more efficient and high yield and cost effective products. So, we don’t see problem, we still that many of the current designs in the logic area, even companies in areas like image processing are still making new designs in 0.18 micron and we feel that the 0.18 micron node is going to become a very important technology node for many years to come. In fact, we have seen even reports that predict much longer life to 0.25 micron technology than what people have anticipated before, but I don’t know about that. But definitely for the 0.18, we see long, long future.

As to the equity -- it is Amir -- we ended up the quarter [or] the year with 43.4m outstanding shares and 4.6m options. And, as I mentioned from the first installment, which is supposed to happen in Q1, very soon, additional 8.1m shares will be added. So as of the end of the quarter and that’s what probably how Q1 will end up.

Thank you very much.

Thank you.

Thank you. Our next question in queue comes from Amir Avivi (ph.) with Oso Reeds (ph.). Please state your question.

Hi Everybody. I wanted to ask about your goal to produce 10,000 wafers at the end of 2004. Can you be more specific about how are you going to do this? This is first question. [You state] also, in the press release, that you are encouraged by the market of semiconductors to the 2003-2004, how exactly -- what is -- first of all what is this production of the market and why you’re so encouraged about it?

Well, let’s say. Thank you Amir for the question. Basically when we say that we are encouraged, we are encouraged by reading the forecast that all of us read of research companies such as Semicor and Dataquest and other companies and basically what these company’s predict is that the market is going to grow and in particular, people are waiting to the first half of the year to be over with respect to any kind of conflicts and wars and any other things of this kind that is a barrier to high growth in the world economy and in the markets. So, the general feeling right now is that the semiconductor industry is like a string waiting to get loose and what’s going to help that -- I mean, first of all there are great products in the market, great products in wireless, the celular phone business is doing good and also the other areas. I would say that surely with the exception of a wide communication infrastructure, all these other areas are doing fine and consumer business, in particular, consumer products are doing quiet well and are expected to do very well.

Now, you couple that with the fact there is a reduction in worldwide capacity. So, ‘01 because of the inertia after '99 and 2000, there was still a significant additional capacity in the market. ’02, you saw the [verifial] of this trend and you actually saw reduction of production capacity and production capacity going down and investment in new technologies and new Fabs has gone significantly down to the below yearly average, which means that with the normal growth of demand in our market and the reduction of total production capacity, at some point you are going to hit the equilibrium, which in our industry takes about micro second before you move from a buyer’s market to seller’s market and a lack of the capacity. And this is typically when market starts to be going up, you start to see prices going up and things of this kind. So, what -- we based our optimism relative to the market is not just from our assumptions, it is somewhat we are reading in the practical consensus estimates of the market research companies. And I think that it makes a lot of sense even though nobody can either guarantee this kind of results.

Now, we are planning to get at some point, I think that early next year is more appropriate with respect to the timing of the 10,000 wafers per month and this is going to be still best. Most of this production will come from the people that we are already working with. What I call the friendly customers that we are now have prototypes for them -- still prototyping for them, starting to prototype for them. So these are -- the loading for this 10,000 wafers is mostly things that we know. It is not just general ideas that are coming from the forecast or the [automatic] forecast about where the market is going but also some activities that we currently have [inaudible].

Great. Thank you very much.

Thank you.

At this time, gentleman, there are no further questions. I would now turn the conference back to Dr. Nissan Cohen. Please go ahead sir.

Okay. Thank you very much for participating in our conference call. We apologize for the short notice but once we concluded all the work, we issued the notice about the conference call and we apologize for the very short notice, and we hope the 2003 will be turnaround year for the industry and also for Tower and we look forward to continue to share with you our progress in the future. Have a nice day all of you.

Thank you ladies and gentleman. That concludes our conference for today. Thank you all for participating and have a great day. All participants may now disconnect.