高塔半導體 (TSEM) 2007 Q1 法說會逐字稿

Ladies and gentlemen, thank you for standing by.

Welcome to Tower Semiconductor's 2007 first quarter results conference call.

All participants are at present in a listen-only mode.

Following management's statements, instructions will be given for the question and answer session.

With us online today are Mr.

Russell Ellwanger, Tower's CEO and Mr.

Oren Shirazi, CFO.

As a reminder, this conference is being recorded May 1, 2007.

I would like to turn the conference to Noit Levi from the investor relations and communications department.

Thank you, and welcome to Tower Semiconductor's conference call for the first quarter 2007.

Our agenda for today's call is as follows.

Russell will start with remarks about the quarter's highlights.

Oren will then provide an analysis of the first quarter financial results.

After management's prepared remarks, we will begin the question and answer session.

Before we begin, I would like to remind you that some statements made during our 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 Forms 20-F and 6-K, as well as filed with the SEC and the Israeli securities authorities.

They are also available on our website, as well as in our Forms F-1 and F-3.

Tower assumes no obligations to update any such forward-looking statements.

Now, I'd like to turn the call to our CEO, Russell Ellwanger.

Please go ahead.

Thank you, Noit.

Welcome and thank you very much for joining us today.

I'm very pleased to report that the positive momentum reported during the past year continued through Q1.

Revenue in the first quarter of 2007 was $55.6 million, hitting the upper half of our guidance.

This represented an increase of approximately 55% when compared to the first quarter of 2006, against a worldwide foundry that was in decline of about 10%.

Results were driven by strong customer demand and continued high fab utilization rates.

Additionally, the first quarter represented our sixth consecutive quarter of positive EBITDA and our second consecutive quarter of positive cash flow from operations at $6 million for the first quarter.

We expect to continue to report positive EBITDA and positive cash from operations throughout 2007.

Operationally, we continue to maintain high utilization levels, having hit above 90% in both of our fabs during the quarter.

Fab 2 shipped a record number of wafers.

During the entire first quarter, we had more customer demand than we had capacity, and in response to this demand we have nearly completed the 50% capacity ramp of Fab 2 to 24,000 wafer starts per month and are increasing the fab loading as the new equipment has come in line.

All process tools are installed and most are qualified and beginning to run production.

The remaining sets of process tools are in final stages of qualification.

During the first quarter, we continued to gain traction with existing customers.

Let me take a moment to discuss a few of these.

A partnership with Sandisk for 0.13 micron fabrication is progressing as expected, and our 10-year volume agreement with Vishay Siliconix should fully ramp to 17,500 wafer-per-month in late summer, representing a significant amount of the available capacity in Fab 1.

We announced an agreement with Zoran Corporation to begin high-volume production of Zoran products on our 0.16 micron manufacturing process.

Zoran chose Tower as the manufacturer of the new products due to our longstanding relationship, the high level of customer support, as well as the cost-effective technology that we offer at Fab 2.

As a result of this agreement, Zoran is expected to become one of our top four customers in 2007.

Additionally, last quarter we mentioned Modiotek as an example of an existing partnership that drives product success through synergistic technology and accelerated time to market.

Our customer support and manufacturing teams worked closely with Modiotek to ensure smooth delivery and quality execution.

We look forward to continued working relationship with them for the coming generations of their families of products.

Forza Silicon expanded their relationship with us based on our expertise of providing high-performance CMOS image-sensor processing technology and services.

We're now designing and producing mixed signal and image-sensor chips for Forza that use our advanced process and design technology capabilities.

On top of that, we have gained several new customers that are worth mentioning.

E2V chose Tower as their supplier of choice for their CMOS image-sensor devices.

These image sensors will target a broad range of industrial and medical applications.

Tower's CMOS manufacturing capabilities in Fab 2 are a perfect fit for this customer's requirements.

We anticipate this will result in a new generation of system on chip sensors.

CopperGate will utilize Fab 2 0.18 micron process for its newest home networking modem controller.

Our state of the art manufacturing process and technical support enabled the company to implement this product in record time and to achieve excellent performance in a cost-effective manner.

Tower made further strides in its design service provisions.

We completed for the first time a chip design from spec sheet to physical design to silicon verification with a lead customer and we continue to expand this work for multiple new designs.

Ultimately, this type of working partnership can lead to the development of optimal product performance metrics and could potentially accelerate the time to market.

We're excited about our future opportunities within this domain.

For Tower, this is a new business model that will not only produce a royalty revenue stream, but will continue to help differentiate Tower from our peers.

An example of the added value of our design services and support team is with Paragon.

In April, we announced that Paragon utilizes our state of the art RF-CMOS process to implement their new data communication product.

Through our combined work, Paragon developed a power amplifier booster, which attained improved transmission power and lower heat dissipation.

This booster will be used in emerging applications such as wi-fi, cellular and WiMAX.

In April, we announced that Tower's CMOS product line was chosen to manufacture many image sensors for Jet Propulsion Laboratory, JPL, a commonly known company for working with NASA in the United States.

This agreement is significant for us, because it further demonstrates our leadership in the CMOS image sensor area, especially in the high-end, high-performance arena, for which we believe to have a best-of-breed capability for rad-hardened sensors.

We also announced a partnership with Impinj, who provided the system on chip developers with their non-volatile-memory solutions using Tower's 0.13 micron technology.

Additionally, we signed a license agreement with Virage Logic allowing Virage's NOVeA NVM module to be manufactured on Tower's 0.18 micron technology.

In summary, we continue to focus on execution and are pleased with our quarterly financial results.

We also are excited by momentum and traction in our business, as evidenced by the fact that in this quarter we continued to have greater customer demand than available capacity.

We've guided Q2 revenue to be between $56 million and $60 million, representing year-over-year growth of between 26% to 35%.

Within this guidance, we have an increase of 25% in Fab 2 revenue from Q1 to Q2.

Fab 1 utilization will not be as high as previous quarters, however, we expect this to be on track in mid Q3 and Q4 to the same high utilization rates that we've had in the last year, as driven by the fulfillment of the Vishay volume agreement and several other new projects.

As stated, our Fab 2 demand continues to exceed capacity and we're evaluating opportunities to increase the same.

IDC, a leading research firm, believes that pure play foundries should see an average annual growth rate of up to 21% per year through 2010, versus a 9% growth for the entire semiconductor industry.

We believe our consistent financial results, combined with our differentiated business model, specialty processes and high-volume customer relationships, will enable us to achieve, if not exceed, the overall growth trends of the industry.

As we have discussed on previous calls, we believe we're on track to reach our next goal of achieving an annual revenue run rate of $300 million.

With that, I'll turn the call over to Oren.

Thank you, Russell, and hello, everyone.

I would now like to review the financial highlights of the first quarter in more detail.

Firstly, as you may have noticed, we added to our press release financial measures that exclude depreciation and amortization costs, as well as stock-based compensation costs.

The release, including the financial papers attached to it, presents also financial information that may be considered non-GAAP financial measures under Regulation G and related reporting requirements promulgated by the Securities and Exchange Commission as they apply to our companies.

These non-GAAP financial measures exclude, (A) deprecation and amortization expenses and (B) compensation expense in respect of options granted to directors, officers and employees.

These non-GAAP financial measures should be evaluated in conjunction with and are not a substitute for GAAP financial measures.

The table also presents the GAAP financial measures which are most comparable to the non-GAAP financial measures, as well as a reconciliation between the non-GAAP financial measures and the most comparable GAAP financial measures.

The non-GAAP financial information presented herein should not be considered in isolation from or as a supposed to for operating income, net income or loss, cash flow provided by operating, investing and financing activities or any other income or cash flow statements that are prepared in accordance with GAAP.

So, now, as you can see in those non-GAAP basis tables in the release, we moved from negative $12 million of operating loss in Q1 '05 to a positive $5 million in Q1 '06 on a non-GAAP basis and to a positive $12 million in Q1 '07.

Further, we moved from a negative $12 million gross loss in Q1 '05 to a positive $13 million gross profit in Q1 '06 and to a positive $21 million gross profit in Q1 '07.

Moving to the non-GAAP margins analysis, all of the above represent a shift from negative 52% operational margins in Q1 '05 to positive 14% in Q1 '06 and to positive 21% in Q1 '07.

Further, we moved from negative 17% gross margin in Q1 '05 to positive 35% gross margin in Q1 '06 and to positive 38% in Q1 '07.

During the first quarter of 2007, 72% of revenue was to public customers and 28% to IBM.

This is compared to 67% and 33% in 2006, respectively.

Regionally, revenue from U.S.

customers represents 79%, compared to 70% in the fourth quarter of 2006.

Revenue from Asia-Pacific represents 7% of revenue, compared to 14% last year, and Israel and European customers made up for the remaining 14% of revenues.

Turning to the balance sheet, we finished the quarter with $44 million of cash, as compared to $33 million in the end of March '06.

Shareholders' equity was positive at $122 million at the end of the first quarter of '07, compared to a negative number in the end of March '06.

Our current ratio, which is our current assets divided by our current liabilities, as of the end of the third quarter of 2007, was 1.5, versus 1.3 in the end of the first quarter of 2006, further illustrating our improved financial situation.

Our total balance sheet was a positive at $689 million at the end of Q1 '07, compared to $636 million as of the end of March '06.

EBITDA continued to be positive for the sixth consecutive quarter, and on top of all mentioned before, we achieved $5.6 million of positive cash flow from operations in Q1 '07, as compared to a negative $12 million in Q1 '06.

This is the second time in a row we have this positive cash flow from operations.

Now, we will take your questions.

Operator?

Thank you, sir.

(OPERATOR INSTRUCTIONS)

The first question is from Beni Dekel of Clal Finance.

Please go ahead.

Hello, everyone.

Hey, Beni.

Hi.

First, I want to congratulate you on your good results to the first quarter of 2007.

We are looking forward for your second quarter.

My question regards of the capacity of the Fab 2 factory and I want to know when you should finish the Fab 2 ramp to be 24,000 wafers per month?

That will occur within the second quarter, within the quarter we're sitting in now.

During the second quarter?

Yes.

Okay, another question that I want to ask is after you have finished this investment, when you should do another investment in Fab 2 to increase the capacity?

As I mentioned in the previous script, that we're looking at several opportunities at present to increase the capacity at the fab.

It's a very strong positive that we continue to have greater customer demand than we have capability within the fab to produce the wafers and that the customers are coming to us with more and more products and, again, greater demand for the existing products.

So we're looking very seriously at different opportunities to increase the capacity under conditions that would allow us to bring in equipment at reasonable cost, if not at very good cost, and produce value for the customers and for our shareholders.

Lately you have raised about $29 million, is it for this reason?

Yes, we did raise money to add capacity to the fab.

Okay, thank you.

The next question is from Ramesh Misra of C.E.

Unterberg, Towbin.

Please go ahead.

Okay, good afternoon, gentlemen.

My first question was in regards to seasonality.

Well, obviously, it's a good sign that you have almost escaped seasonality, in fact growing in Q1.

Can you tell us why or what are some of the reasons that you've been able to skip seasonality altogether?

Well, to begin with, we have a reasonable percentage of our revenue is on contract or take-or-pay agreements, so that certainly modulates to some extent against seasonality.

The other part is that we have multiple segments that we participate in and within those segments several very specialized products.

So we have the core CMOS product line, we have the image sensor and the NVM product lines, the mixed signal RF product lines, and within each of these lines, again, customers that we're very partnered with that have specialty applications.

So I think the fact that we're not overly weighted in any specific segment certainly helps us quite a bit to avoid, or at least to modulate cyclicality that one would see if they were very strong, for example, in just very advanced logic that's going into PCs.

We do have customers within the PC area, but we're not overly weighted in any single area.

So I believe that that's the biggest reason why we're able to avoid it.

If you look at the fact that we play in multiple segments, we play in many specialty areas that are themselves maybe not so seasonal and, in addition, we have got a reasonable business model on take or pay agreements with several large customers.

Okay, thanks for that, Russell.

So, other than the take or pay agreements, with your relationships with many of your customers deepening further, are you able to get better visibility from these customers?

Are they more open to kind of sharing with you their roadmap and product plans, say, a little longer than they've done so in the past?

Ramesh, I think that's a very good question that could be answered in several ways.

To begin with, one of the areas that we believe as a foundry that Tower differentiates and now we've heard from multiple of our customers that they feel very pleased about, is the fact of the culture of the company being a high-security culture, meaning that our customers feel very comfortable to share with us two, three, four generations down the road what their roadmap is so that we can be prepared and work with them both on process innovation, as well as on the device design side to make sure we have the IPs and libraries in place.

Multiple of our customers are somewhat leery to do that in other geographic sectors of the world for the fact of IP leakage from what they're working on.

So I think, as a very generic statement, we maybe have more visibility into customer roadmaps than some other foundries would, just for the fact that the customers feel very secure to work with us very closely on the roadmap.

Now, on the second part, as our customer base has expanded, as we've brought on multiple large customers we're delivering more than 1,000 wafers a month to, their ability to forecast becomes stronger, because they have multiple products that we're making for them and much greater demand on those products.

So I think our visibility over several quarters has improved as well as our relationship with larger customers has improved.

Not as a relationship with them, but as we've brought on larger customers and developed relationships with the larger customers.

If you look at Tower several years ago, to where there are only two customers that were shipping more than 1,000 wafers a month to, from a whole grouping of small customers, it's very difficult to secure a forecast when the customers themselves don't have a whole lot of leeway moving products to different areas.

So I think in both instances, as far as the customers being willing to share with us their roadmap, that's definitely the case.

As far as us having good visibility, at least as good as our customers' visibility, certainly we have close cooperation, with a lot of work within forecasting, and I think within Tower we've developed a very good group that works with our customers both on the sales side and internally on the fab planning.

Okay, that's very helpful, Russell.

In regards to capacity, so clearly you're now in a situation where you're increasingly capacity constrained.

So I guess with the new equipment that you brought online and are in the final stages of qualifying, is it reasonable to expect that basically, at least in Fab 2, wafer capacity will probably plateau at this 24,000 per-month level, at least for the foreseeable future, up to I guess you deploy some of your recently raised money to that?

That depends how you define foreseeable future, but certainly, Ramesh, for the Q2, Q3 and the beginning of Q4, it's off the equipment that we're installing now.

As I mentioned, we've not announced something about a further increase in capacity.

We've evaluating it.

We have customer demand asking for it and good customers.

So there's a possibility that we could see some increase in capacity at the end of Q4 and beginning of Q1, but that would be off of doing some interesting things at present.

But if you were to define foreseeable future, the Q2, Q3, Q4 revenue, it would be off of the attainment of the 24,000 wafer starts capability, as well as the utilization of Fab 1.

Okay, and then one final question I guess, and then I'll jump off, and this question is related to product roadmap, both on the geometry side and also on the product category side.

So, clearly, you have a very, very strong capability in the image sensor space.

You have a solid presence in the mixed signal ,wireless mixed signal space.

What are some of the other areas where you expect to and have been expanding your capabilities?

And if you can talk beyond that in terms of roadmap either in terms of geometry or in terms of product (inaudible), that would be also better, might get at a better understanding future drivers.

Thanks very much.

So one of our large focuses now within the analog space is the whole area of power management.

We've had a press release in Q4 with regards to the development and functioning of LG mouse transistors, but we're driving strong within the power management area and have I think some very interesting projects going on within that area.

So that's our biggest focus within that product group at the moment.

Now, certainly, we've announced before our work with the leader in the RFID space for gen one, which as far as being the sole producer of the gen two and the RFID, that being Alien Technologies, and we see the cooperation there growing as well.

So within the RF mixed signal, the standard mixed signal applications we stay in, and those are a very good area for Tower.

Within mixed signal RF, again, the RFID, we're pretty strong within the wireless LAN space with the Pharos.

We've talked about that and several other customers that we've maybe not been able to press release with.

In the image sensor, we have a variety of applications that we drive from cell phone cameras and quite a bit of activity within the area of medical imaging and a lot of focus within the medical imaging.

From the area of core CMOS itself, we have a large focus within the company on IDM transfers and IDM activities, which we think is a very strong focus and, again, something that Tower can actually differentiate in, and that being on customers that right now might find themselves manufacturing in high-cost regions with their own internal fabrication but have proprietary process flows, seem to be very willing to transfer the technology flow to Tower, again, with a strong assurance that there won't be IP leakage from their process flow, and the ability of us then to bring up their specific process flow here and continue to work in it.

So although the area of IDM transfers is not any signal segment, per se, it's an area that we're focused on and one that I think we've been nicely successful with this, so that would be the Vishay Siliconix, the International Rectifier type activities, and several other activities that we're pursuing that have not been press released, or we've not been approved to press release.

Those would be the areas that are of our present focus.

Outside of what you would see now within medium-size or large customers, we really are focused on the area of power management and completing our product portfolio within the power management arena.

Okay, thanks very much, Russell.

That's very helpful, and congratulations on the growth in Q1.

The next question is from [Thomas Schwartz] of Unterberg.

Please go ahead.

Hi, guys, congratulations on a good quarter.

I'm sorry to follow Ramesh.

I wanted to just follow up a little bit on the question that the gentleman from Clal asked about the timing of when you fill out the capacity to 24,000 wafer starts per month, and then also ask you kind of what comes after that and maybe in terms of how many additional wafer starts will cost you how many dollars in capital spending in order to get there?

Okay, the first part of the question, to get to the 24,000 wafer starts, that area, plus, minus a few wafers, that we will have the capacity in store within Q2.

Now, once you have the capacity in store, you then have to start building up, driving, now running the customer wafers, and then you have the cycle time of the fab to be shipping the wafers.

So we ship the entitlement in a revenue basis of that capacity coming in line the second half of Q3, and then in the full measure in Q4.

So does that answer that part of the question?

Yes, so Q4 is the first fully ramped quarter at 24,000?

Yes.

Got it, yes, that does.

Okay, and that's 24,000 starts, that is not 24,000 shipments.

That's 24,000 starts.

Within that is obviously engineering wafers and projects.

Yes, thank you.

Now, for the second part of the question, you had asked I think two questions, and if you asked one, we would see it and how much would we be looking at and what would the price be.

Yes.

So we have within our roadmap to move to 90 nanometer or to include 90 nanometer within the second half of 2008, so the 0.13, so 130 nanometer, 90-nanometer tools are, if you look within the new tool market, they're very expensive tools.

We're focused now on defining working, acting opportunities to where we could take on these tools at significant discount by trying to get them in a used market, but there's not a lot of these tools available in a used market, so one has to be very careful how they go about it.

Our model for the next capacity within the fab would be to try to bring added capacity on for less than or about $10 million per 1,000 wafer starts, and that would be a very aggressive model, but that's a model we think is a good model for us to have.

That would allow a very quick return on investment off of the incremental investment.

So, Tom, did that answer your question?

Yes.

That helps.

Does that $10 million per 1,000 wafer starts presume some kind of mix between newly purchased tools or new tools and used tools?

It would presume a very definite mix between new and used and that being very, very few new tools.

Right, that's helpful.

Thank you.

Thank you.

(OPERATOR INSTRUCTIONS)

There are no further questions at this time.

Thank you very much.

Sorry, Mr.

Ellwanger, would you like to conclude the call?

Sure.

So, again, we thank you very much for participating in the conference call today.

We plan to be at the Unterberg Emerging Growth Opportunities Conference in July and would hope that those of you that are calling in from the U.S.

and maybe from other countries would have the opportunity to go there.

We'd love to see you there, to talk with you.

But, again, we will be there and, again, thank you for participating, thank you for your support.

We're happy with the results of the quarter.

We're very pleased that we're able to guide for an eighth consecutive revenue increase for Q2, and we see the remainder of the year, Q3 and Q4, being very positive.

So thank you again and look forward to seeing you in July.

Thank you, this concludes Tower Semiconductor's first quarter of 2007 results conference call.

Thank you for your participation.

You may go ahead and disconnect.