israelinvestor.com
Tower Semiconductor, Toshiba, and Fab2
Mark Savolainen Smith 4/3/2000 02:00
Notes on the March 31 Tower Semiconductor conference call
The market was not kind to Tower Semiconductor (TSEM) last week. Opening with a bang, Tower hit $43 1/2 early Monday. Going out of the week with a whimper, however, it dipped as low as $20 5/8 on Friday afternoon.
General market weakness coupled with uncertainty over a media report of a possible secondary stock offering eclipsed a very important Tower announcement: a technology transfer agreement with Toshiba Corporation. Moreover, as the week wore on it became clear that the implications of the announcement were being misunderstood. A conference call was set-up with Tower's Co-Chief Executive Officer, Dr. Yoav Nissan-Cohen, to set the record straight.
As Dr. Nissan-Cohen put it, "From all the questions, chat discussions, and even erroneous media coverage, it seems that there are some misunderstandings and confusion regarding the exact meaning and significance of our [Toshiba] announcement."
To help our readers get to the heart of the matter, we present the following notes from Dr. Nissan-Cohen's presentation and his responses to questions on the March 31 Tower conference call:
Toshiba Makes Fab2 Possible
Dr. Nissan-Cohen emphasized that Toshiba's technology transfer of its most advanced CMOS technology to Tower was the critical first step towards making Tower's long planned fab2 a reality. "Without it [advanced CMOS technology], the new fab would be irrelevant."
The Toshiba agreement has three aspects. First, in addition to the technology transfer, Toshiba has agreed to train Tower's senior technology team in its application. The technology transfer includes process recipes, and patent licenses required for the use of Toshiba's advanced process technologies. Toshiba will also provide on-site assistance to support the transfer of the technology to Tower's new facility.
Second, Toshiba has also decided to invest in Tower. A small initial equity investment of $10 million has been committed to Tower's expansion (not fab2), but Toshiba has the option of investing "significantly more at a later stage."
Finally, Toshiba itself will become a Tower customer and has secured a capacity allocation in fab2.
Now that the the Toshiba technology transfer agreement is in place, financing can be negotiated, customers lined up, and Israeli government grants secured in order to make Tower's proposed new plant in Migdal Ha'emek a reality.
A key next step will be bringing in a "major strategic partner who will also become a major customer of the new fab."
Dr. Nissan-Cohen noted that "At this time we are in advanced negotiations to finalize an agreement with such a strategic partner. We will report on such developments as soon as appropriate."
Dr. Nissan-Cohen pegged the cost for fab2 at $1.3 billion .
Israeli Government Grants
The Toshiba agreement has substantially helped to secure Israeli government support. The president of Toshiba Semiconductor, Yasuo Morimoto, came personally to Israel to sign the agreement with Tower. During the visit he met with the most senior Israeli government officials, including Prime Minister Ehud Barak.
Dr. Yoav Nissan-Cohen reported having consistently received very positive feedback from government representatives regarding the importance of Tower's project to Israel. He added that he was very optimistic that the financial incentive package for fab2 will be approved, although no dollar figures were mentioned.
Secondary Only at a Much Higher Stock Price
Dr. Nissan-Cohen made it clear that the market's fears of an imminent and dilutive secondary are misplaced. As sensitive as the issue of the secondary is, it is probably best to let Dr. Nissan-Cohen's words speak for themselves:
"OK, I have seen so many speculations about it [the secondary] that it is really getting out of control. Basically we do not have any decision about a secondary at this stage. Obviously, we need to raise more equity for the project. And it is all very clear, we prefer having this equity coming from the financial strategic investor, who is also going to become a customer in the facility. This is what I referred to as the negotiations that are taking place right now. This is definitely the preferred way to raise equity, but we haven't excluded other possibilities. We'll see.
"You know, if the stock will be very high- and I know it is very different than what we are seeing today -- maybe eventually we will go into a secondary offering, but there is no decision about this and it is not clear at all how and when we are going to do that."
Fast track schedule for Fab2
In one of the most detailed segments of the call, the proposed fast track schedule for fab2 was outlined. Tower seems both confident and aggressive.
Tower has already interviewed three major engineering companies for the construction of fab 2. All are among the biggest engineering companies in the world. The final selection will be made "very soon" so the design process for the plant can begin.
Second quarter of 2000: final design of the plant completed.
Third quarter of 2000: Construction begins.
Third quarter of 2001: Construction completed. Equipment installation begins.
Fourth quarter of 2001: Prototype runs begin.
Early 2002: Commercial production begins.
Although not discussed, it should be noted that it can take up to a year and a half to ramp up production fully once commercial production begins.
Fab2 Design: Copy Exact to Start
In another very interesting and detailed segment of the call, Dr. Nissan-Cohen revealed Tower's thinking regarding the fab2 design itself. The plan is to minimize risk, while at the same time being prepared to maximize reward once production is established:
"Ideally we would like to do copy exact [of a Toshiba facility] at least for the logic process in order to minimize the risk...So we have to be very careful, because our engineers always like to improve, and get things more efficient and less costly, or any other advantage they see, and we would like to postpone all these improvements for a later stage and start by having a copy exact.
"At the same time there will be some engineering projects that we will have to do and this is referring to installing our specialized modules such as the microFLASH and such as the CMOS image sensor on the baseline of Toshiba technology. So it will probably start again as a copy exact of Toshiba logic process but pretty soon after we would like to be able to run our specialized modules as well."
While it is true that, as Dr. Nissan-Cohen explained in this call, "This facility [fab2] is well justified without the microFLASH," it is in all likelihood, also true, that should microFLASH (Tower's name for Saifun's proprietary N-ROM technology) prove itself, Tower could be sitting on a gold mine.
Although microFLASH has the potential to be most significant, Tower also has been finding success with another specialized technology: CMOS image sensors. The long term Tower investor should keep a close watch for developments on both these fronts.
In another important milestone last week, Tower announced that it had begun shipping the first commercial microFLASH product (to Acer Labs). Dr. Nissan-Cohen also let on that there have been, "major developments on the [Saifun] technology over the last 6 months."
The world will be listening carefully for such news in the Tower quarterly conference call in a few week
|
