艾克爾 (AMKR) 2004 Q4 法說會逐字稿

Welcome to the Amkor fourth quarter 2004 earnings release conference call.

At this time all participants are in a listen-only mode.

Following today's presentation, instructions will be given for the question-and-answer session. [Operator Instructions] As a reminder, this conference is being recorded today, Tuesday, February 8, 2005.

I would now like to turn the conference over to Jim Kim, Chairman and Chief Executive Officer.

Please go ahead, sir.

Thank you very much.

Thank you for joining us today.

I am Jim Kim, Chairman and Chief Executive Officer.

With me today is John Boruch, President and Chief Operating Officer, and Ken Joyce, Chief Financial Officer.

I'll make brief remarks.

Ken will discuss our operating results, and then John will have some closing comments.

It is our belief that the inventory correction will play out in the first half of 2005.

By most accounts, there is a modest amount of finished goods inventory in the channel.

We believe that most of the semiconductor inventory is in the form of fabricated wafers.

To be sure, our wafer banks are at all-time high, and as finished goods inventories become even more depleted, Amkor and other companies in our sector should act as the leading indicators of an upturn in the business cycle.

During 2004, we commenced a strategy expansion program, and our core strategy for 2005 and 2006 is to grow into our operational footprint.

We'll do this by leveraging our business alliance with IBM, by extracting the potential from our Unitive acquisitions, and by providing the right assembly and the test solutions for our customers; together with the highest level of service and support.

We are forecasting moving forward with these strategies to position Amkor for growth.

Ken Joyce will now review our financial results.

Thank you, Jim.

Before we discuss our financial results, I'd like to remind you that any forward-looking statements made during the course of this conference call represent the current view of management.

We refer you to today's press release which was filed with the SEC, Form 8-K, prior to this conference call; and our other filings with the SEC for information on risk factors that could cause actual results to differ materially from our current expectations.

The fourth quarter results were within our previously announced guidance range.

Fourth quarter gross margin of 16 percent was down from 18 percent in Q3 due to the combined impact of lower quarterly revenue and the absorption of costs and expenses associated with increasing our factory work force, equipping lines and qualifying new business in our new operations in Taiwan, Singapore, and at Unitive.

Over the near term we expect continued pressure on gross margin reflecting the affects of lower asset utilization, competitive price actions, and the continued absorption of factory costs as we ramp up our recent acquisitions in advance of expected growth later in the year.

Legal expenses in connection with the mold compound litigation and the final stages of our patent infringement suit against Carsem continue to run high, 5 million this quarter compared with an historical norm of less than $1 million per quarter.

We were encouraged with November's favorable verdict in the Phillips trial, which held that Amkor had no liability.

This could have positive implications for our legal expense profile going forward.

However, there is no way for us to predict how the other litigants will view their respective claims against Amkor, notwithstanding the recent court decision in our favor.

Following this favorable verdict, in the fourth quarter we reversed a reserve of $1.5 million that was provided in the first quarter related to a proposed settlement of the Phillips lawsuit.

Last week, we announced that the U.S. international trade commission has decided to conduct a full review of our patent infringement lawsuit against Carsem.

By way of background, in 2003, we filed suit against Carsem for infringement of our micro lead frame patents.

In November of 2004, the ITC's administrative law judge issued an initial determination that Carsem did not violate the Tariff Act.

Of course, we feel otherwise.

And we are pleased that the full ITC has elected to review the initial determination.

We expect the ITC to complete its investigation by March 31st of this year.

Fourth quarter capital expenditures of 38 million were focused on increasing our capabilities in wafer bumping, flip chip, and test.

We are targeting CapEx of 50 million for the first quarter of 2005 to support continued expansion of our wafer bumping, wafer level packaging, flip chip assembly, and test operations.

During the quarter, we strengthened our financial liquidity by entering into a $300 million term loan.

This loan has a single bullet payment due October, 2010 and will provide us with additional liquidity as we work through the current industry downturn.

During the fourth quarter, we made scheduled payments totaling approximately $120 million to IBM and a Chinese developer, completing our financial obligations under our strategic collaboration with IBM which we announced last May.

Here is a recap of our first quarter 2005 guidance contained in our earnings release.

Revenue should be down between 8 and 12 percent from the fourth quarter of 2004.

Gross margin should be in the range of 10 to 12 percent.

We expect the first quarter net loss in the range of $0.34 to $0.40 per share.

Now I'll turn the call over to John Boruch for some additional comments.

Thank you, Ken.

During the fourth quarter we moved forward with several key strategies.

We are integrating the Unitive operations in North Carolina and Taiwan to provide a complete turn-key solution for rapidly growing market opportunities in flip chip and wafer level packaging.

Based on indications of strong customer demand, we're expanding our capacity for wafer level packaging in North Carolina, and we will be building capacity in Asia later this year.

We are also expanding our wafer bumping capacity at Unitive Taiwan and building scale in wafer probe, flip chip assembly and final test.

These activities are geared at supporting growth opportunities in the graphics, chip set, and gaming markets.

Our recently acquired IBM test operation in Singapore is performing to our expectations and we are building up our tester base in response to increased business, not only from IBM, but other customers as well.

Assembly volumes under the IBM supply agreement should build over the next several quarters as the factory qualification process moves forward, and IBM continues to transfer business from our competitors.

In China, we have facilitized the remaining 75,000 square feet of our existing factory in Shanghai to accommodate anticipated business growth.

We have no immediate plans to facilitize the new building we acquired in the IBM transaction, however this building will accommodate our long-term expansions in China.

Our assembly unit volume rose 30 percent during 2004, driven by strong gains in micro leadframe, stack packages, chip scale BGA, System in Package, and also some legacy leadframe packages where we provide a turn-key, assembly, and strip test.

Our stack package units more than doubled in 2004 versus 2003.

More significantly, revenue from stack packages in 2004 increased nearly 2.5 times.

Our leadership in advanced stacking techniques, including logic plus memory, and logic plus logic, is creating a richer product mix for us in stack packages.

Amkor was the first outsource assembly company to stack ASIC and mobile DRAM memory chips which is becoming a critical solution for 3G cell phones.

We are the clear OSAT leader in this space today.

Also during 2005 our leadership in stacking logic plus logic should allow us to significantly increase our share of the package market for CDMA chipsets.

Our micro leadframe business continues to grow.

We shipped nearly 1 billion MLF units in 2004 compared with 750 million in 2003 and around 400 million in 2002.

We will be broadening our MLF package offering and manufacturing capability through 2005.

We're taking advantage of the increasing popularity of these packages.

When considering our capabilities and positioning in products such as stacked packages, micro leadframe, flip chip, wafer bump, wafer-level CSP, and our IBM alliance, we expect that following a down Q1 we will grow through the rest of 2005.

We believe this growth will occur even without a major market recovery.

At this point, our customers do not have good visibility into the timing of the industry recovery.

However, because most inventory seems to be concentrated in the form of fabricated wafers, we expect the industry recovery to be led by the OSAT sector.

Operator, we will now open this call to questions.

[Operator Instructions] John Pitzer, Credit Suisse First Boston.

It's actually Brian Chin for John Pitzer.

A couple of questions.

Let me lead off with this one first.

Just in terms of the die bank situation, what you're seeing right now.

It looked like the laminate revenues held up better than the leadframe side of your business.

Some of the reasons are probably apparent but is there kind of a bifurcation of, you know, fundamentals in terms of your leadframe die bank right now versus the laminate, or maybe the situation has improved on the laminate side, maybe not improved on the leadframe side?

Let me divide that in half.

Sure.

The die bank inventory has grown in laminate inventory-- in laminate packages and in leadframe intended packages.

So I don't see a big difference in where the die was built.

Our customers just built a lot of die prior to shutting off the wafer supply.

On the laminate business versus leadframe, yes our business seems to have more amplitude up and down on the lead frame side.

We think one of the factors is that the IDMs who do a lot of lead frame internally, can internalize that capacity whenever things get soft, so we tend to see a bigger bump down on leadframes.

And, reversely, as we get through the middle of a cycle, up-- on the upturn we see our leadframe business expand rapidly as their factories fill up.

On the laminate side, since so much of that is outsourced I think Amkor, along with our competitors, sees more of the total market fluctuations and it's not amplified by internalization very much.

So that's how I see it.

Part of the reason I ask that is because you say that maybe 25 percent of your customers bank die with you versus what they hold themselves, and it probably seems like there's some sort of hierarchy in terms of die bank that they need to bleed down internal versus when they access the die bank reserves that you have.

I'm wondering if you can kind of quantify how much more work-- how much more work through there is at your IDM customers in terms of their own internalized die bank.

Say it a different way.

I don't understand the question.

Basically there's two areas of die bank;

One at your customers and, I guess, one at yourselves.

I'm wondering if you can help us understand when your customers' die bank will be worked down and then when you can start to see them draw down from your die bank reserves.

Yes, that's hard for us to see.

They-- they-- they stock die in our factories on the packages that they build predominantly in our factories, because it's a safe place to put it.

They stock die in their own warehouses where they have multiple sources for packaging those die, so they tend to keep it in their own warehouse and then send it out to us or one of our competitors.

So that's how I believe that's played.

So we don't much visibility into what they're keeping at their-- at their own factories except that we believe there's a direct correlation to what we see and what our customer's doing in their own factories.

We don't see any reason why there wouldn't be.

So basically what the whole thing about die bank is that, as usual, it's no secret, they tend to shut off the wafer starts much later than the assembly starts because they build assembly starts almost to order and they build wafer starts to plan.

There's some other reasons there economically.

That's the way it's worked.

So with this cycle being, what we call, a typical mid-cycle correction, in my opinion, then-- then the assembly starts should appear earlier in the cycle than the wafer starts.

Okay.

Let me just maybe sneak in a few other quick questions, then I'll let other people get on.

In terms of your gross margin guidance, 10 to 12 percent, can you give us your utilization outlook for 1Q?

And then also your depreciation outlook for 1Q as well?

And then maybe give some more commentary in terms of the pricing environment that's, I guess, pressuring the gross margins and whether there's any sort of scheduled price -- pricing concessions or just, you know, lowering of your pricing that's already built into certain contract arrangements you have with customers that are impacting that situation in 1Q.

Okay.

Well, Brian, that's a loaded question.

Let us get back and try to take it in pieces here.

First, as far as depreciation, depreciation in Q4 was around 61 million.

I would anticipate that relatively flat, up slightly in Q1.

With respect to utilization, our utilization in Q4 was in the area of 70 percent.

Our guidance for Q1 is down slightly, so I'd expect that utilization come to down modestly.

So that should be there.

As far as the pricing environment, John, would you like to --.

Yes, I'll talk to that.

We see-- we see pricing pressure because of the underutilized capacity throughout the-- throughout our sector.

But we don't consider it yet abnormal.

We think it's in the normal range.

And we see pockets of more competitive situations like PBGA, it's quite competitive, as that demand moves towards flip chip, and there's idle capacity, especially in Taiwan.

We've been saying that the last couple of quarters, we still see that.

But in the other areas we see normal-- normal-- normal pricing patterns.

Mark Bachman, Pacific Crest Securities.

Can you talk a little about your IBM deal?

Are you guys still expecting 100 million in revs from this in 2005?

We talked about the potential when we first did this transaction, 150 million, that was before the downturn.

We've lowered that expectation for the year, but our expectation is now somewhere in maybe the 125 range.

Okay.

And how should that ramp throughout the year?

It'll ramp, you know -- it starts in Q1, it'll pick up speed as we go through Q2 and Q3, okay, and then level out in Q4, I think, somewhat as we pick up--.

The first round is with IBM existing, mostly leadframe packages being transferred into our factories from our competitors, and then as we go into second quarter and third, there's the associated foundry business that we hope to pick up as we go through that part of the year.

Okay.

So then on your Q1 guidance, how much of Q1 is being enhanced then by IBM right now?

In other words, I'm trying to look at your organic decline from quarter-to-quarter.

I think -- I don't have those numbers in front of me, but it's a few percentage points.

Okay.

Todd Diffely, Merrill Lynch.

Quick question on capital spending.

Is the plan for '05 still about 100 million?

What we see now as we go into 2005 is that on the flip chip wafer level CSP, the associated probe and test, and bumping demand, it's -- we're looking quite good for us.

We think with the Unitive acquisition was very good in opening up a lot of doors.

So we spent -- we allocated about $50 million for Q1, much of that going into those areas I just mentioned.

We think the 100 million may be a little bit too low, and we don't know where the next 50 will be spent, but we think somewhere in the second half of the year the rest of the market will show up and we'll be looking at some opportunities we want to fund.

So our numbers now are more in the 150 range rather than 100 million, but all dependent on whether the market picks up or not.

Okay, but it's fair to say most of that's going to be flip chip versus wirebonder?

Yes, mostly flip chip in associated areas rather than in general capacity.

Okay.

Then one quick question on utilization rates.

You had it at 70 percent again this quarter even though unit shipments were down 15 percent.

That doesn't seem to make sense.

It's all about the mix.

So just a higher ASP mix -- or a longer --.

Higher yes, the leadframes went down, and the laminates, we didn't go down as far, so we had a better mix.

Satya Chillara.

John, on the same lines here, or for Ken, if I look at Q1 gross margin guidance Ken talked about utilization being in the same 70 or 68 kind of a level but gross margin is going down almost, what, six percentage points here.

Can you shed some light here?

How should we be thinking about gross margin versus capacity utilization versus product mix?

How should we be thinking or modeling in the future?

It depends a lot, Satya, on what the product mix is, what the pricing environment is.

Our outlook right now is we can expect the gross margin to fall a little bit in Q1 and then rise slowly over the next couple of quarters as we generate positive operating leverage from increasing revenue.

The drop in the margin has a lot to do with the reverse affects of the operating leverage, if you will.

With the drop in revenue, we have the heavy fixed cost structure and margins do tend to compress.

In addition to that, as we had discussed previously, we entered into a number of acquisitions which would have increased the Company's footprint and fixed costs which do have also the affect in the short term of dampening margins.

Okay.

Great.

The second question either is for Jim or John.

In terms of 2005, with respect to market share, with respect to profitability, how are you guys going to make the connection here in terms of, obviously you lost some market share in '04.

Would you be going after market share or is profitability the key metric that we should be looking at?

John will answer this.

This is me.

I think with all of the things we're seeing now, even though our first quarter doesn't show-- the numbers don't show it and I wouldn't expect them to; but the initiatives we took later last year we think are turning into fruit as we go through 2005.

So I'm talking about the bumping and flip chip and wafer level CSP and associated test and probe.

Those are large ASP opportunities if you do the whole turn-key.

We think we have a good leadership in technology and stacked technologies.

So that should do well for us.

We talked about penetrating the graphics market and we can tell you that we've qualified for the leading graphics makers on our various packages, and we'll see how well we execute as we go through 2005.

We talked about stacked and some other things, our MLF packages being accepted into the CDM-- CDMA arena, and that should bear well.

And those are all areas we didn't do much in, or anything in, in previous years.

So -- and then we have our IBM alliance and the potential fruit that lay out ahead of us through 2005 and 6 with that.

So that's why I closed my comments on the call saying that, regardless of the market, unless it goes way down, we think that we will grow this year.

Now, I can't talk to market share because I don't know what our competitors are looking at, but I can tell you this, that we're quite optimistic about the period after Q1.

Ali Irani, CIBC World Markets.

Just looking at your utilization in Q1 in the low to mid-60s, how do you expect to leverage the operations as you go through the year?

Is there a potential here for incremental cost cuts?

Maybe some shutting down of capacity?

Or are we going to look at a year where it's just going to be very slow leverage growth, even with the volume gain from IBM?

Let me start it, and Ken can jump in later.

Based on a change in mix, and reapplying the resources that we already purchased, that 30 percent or so that are underutilized now in the right areas, and the power of revenue growth, with a reasonable ASP.

I mean, the models we have says that it's not that slow of a return to reasonable gross margin levels as we go through 2005, provided that we execute on the opportunities that are in front of us and we deliver our customers the product and gain the revenue out of them.

That's the big if.

Okay?

In our opinion, it's all about execution, and so if we execute, the revenue will be there and we should see a pretty good lift-off in gross margins.

Don't know exactly what that might look like through the year but it's certainly improving.

And I don't classify in my interpretation that as slow return to good growth margin areas.

Could you qualify or quantify reasonable gross margins for us?

I mean, are we looking at something in the high teens, mid teens, or the low 20s?

Yes, I'm looking at around the 20 percent area to be back to what I call a reasonable -- not good, but a reasonable gross margin.

Jesse Pichel, Piper Jaffray.

Let me ask you this question.

If all your die banks were immediately sent for assembly, how many quarters of COGS would that give you?

Wow, that's a good question.

I don't have the answer.

I think -- I'm trying to remember how many hundreds of millions of die we have in the bank.

I can't answer that.

It's a lot, okay?

And regarding those die banks.

I've never thought about it in those terms because it doesn't happen that way, okay?

It's just tough for to us get visibility there, and there's not a lot of detail given by your customers.

Regarding these die banks where is the inventory most pronounced?

What end markets?

And I do have a follow-up question.

Couple of areas.

Number one, there's some safe markets that build inventory, that's in the PLD areas.

So pretty standard product and they can stock die and get away with it.

Then you can get more die in leadframes because it seems to be less customized than the laminate.

The laminate's 100% custom, so either you hit the market window with a laminate die, or you throw it away.

You know what I mean?

Right.

So that's how it's skewed, I believe.

And then for us and our customer base, our -- the area we play the biggest in is communication, mostly wireless.

And that, unfortunately, was the weakest part of the semi market in the second half of 2004 and, consequently, I don't know this for sure, but my guess is there's a lot of wireless die in our die base right now because that market weakened relative to expectations throughout the year.

The computer side of the business seemed to remain strong but we don't play as much as we like in to that area.

We hope to in the future.

And I would guess that our computer die inventory is in the reasonable range.

At your analyst day you talked about -- or actually, you didn't, but representatives of Amkor talked about specific customers that were in flip chip qual.

And you just mentioned one end market sector where they're moving out of qual.

Have any of these customers moved into production?

And maybe you could share with us what you think some of your strengths and weaknesses have been in pursuing that end market.

And, you know, if you could talk about what -- how much CapEx did you actually spend, and then I have just one last follow-up question.

Okay.

Let me -- let me remember what you asked.

Sorry.

On the market segments, yes, I think on the graphics side we are beginning to do some ramping up in February, leading up to accelerated rates in March, April, May, June.

So that seems to be happening, and it all depends on our execution in our new factory in Taiwan, in our new acquisition in Unitive in Taiwan, also.

So they go together.

In that area also we had to take Unitive up from, I think, 4,000 wafers a month to more like 17, 18,000 wafers a month.

That CapEx has been committed but not in the factory yet and it's going to be rolling in as we go through Q2 and maybe late Q1.

I'm not quite sure on that.

So that will be coming in during those periods.

So I don't have exact numbers for flip chip and associated.

It's test, it's probers, and it's flip chip packaging, it's space buildout in our new T3 factory, it's normal production for the Unitive operation for bumping.

So it's a lot of stuff and I don't have that under brain immediately, but I would say more than half of our CapEx in Q1 and of late has been in those areas.

Stanford Nishikawa, Citigroup.

I actually have two questions.

The first one is, can you talk about the six-month forecast and/or I think in-- previously you've talked about sort of your accuracy rate or hit rate, kind of the difference between what they've been telling you and what they've been pulling?

And then the second question I have is, can you help us better understand kind of the dynamics of the, quote, business that you're winning that's justifying the change in CapEx?

Because I think the part that's a little confusing to me is that, you know, the CapEx seems to be ramping up, and you seem to be indicating that it's kind of incremental business.

And is it genuinely new business that you've won with new programs, or was this all sort of, you know, all in the same, in terms of the general movement in flip chip and wafer and that type of stuff?

Okay.

Let me start backwards.

Regarding the CapEx, and the need for it, it's -- as I said before, it's in that flip chip bumping test probe area, a whole lot of it and, you know, when we bought Unitive we did not know exactly what -- how successful we could be.

Because we knew the potential was there, they had the right technology, we think, leading technology for what they do in the world.

And we polled our customers prior to buying Unitive and they said they liked the technology, what they knew about it.

So we thought we could unleash that potential.

And as it seems to be turning out, we are and we've got lots of opportunities, maybe more than we can handle in the short term even, but we're starting from a pretty small base.

So that CapEx we're spending is all on what I consider new business opportunities that certainly wasn't on the books six months ago.

It was potential, but that's about all and we had to go work that.

Thanks, that was helpful.

So on the area of the business cycle, what we see now is that our six-month forecasts and three-month forecasts are moving forward on the upward slant slope after three quarters of going down, basically.

Maybe starting May or June of last year somewhere, kind of curved down, and now we see a return to up, which is what we expected.

We have been -- we're waiting until after Chinese New Year in these typical cycles like this, I call it mid-term correction, that we see a acceleration in unforecasted die support to the assembly areas as people come back from that vacation and realize they have some business to service that they didn't maybe expect two weeks ago, even.

Things change fast and we are seeing a little bit of activity in the unforecasted die range which is sort of the leading leading indicator, certainly an indicator that the trough is here during Q1, and the question is the timing of the -- of what we call a pickup, and, you know, it's anybody's guess.

As early -- end of February and late as later on in Q2.

We -- we'll all find out.

So those are all good signs though that it's a matter of time, not if the upturn is coming.

In our opinion here at Amkor.

Eric Rubel (ph), Miller Tabak Roberts.

Ken, question for you.

It looks like in the financing that the Citizen payment was included in this quarter.

Could you confirm that?

And could you give us some sense of whether you think working capital will be a source or use in the first quarter?

With respect to Citizen, we have not made the final payment there.

Okay.

We are -- we're in arbitration with them right now, and as soon as that's resolved we intend to make the payment.

With respect to working capital in the Q1, we would see it as actually a use of funds in Q1.

Sundar Varadarajan.

Just a follow-up on Stanford's question.

In terms of your CapEx spend, again, it seems like the ramp is weighted towards the first quarter.

How much of that potential revenue that you're seeing from, the Unitive acquisition is actually signed contracts as opposed to an expression of interest?

I'm just concerned because last quarter-- last year, too, we had like ramp-up in the first half and the demand necessarily did not materialize and, therefore, you ended up with higher CapEx.

We thought it was going to be more kind of evenly weighted during each of the quarters this year.

Seems like again you're ramping up in the first quarter.

So how confident are you that, you know, the revenue that this CapEx is banking on is actually -- is going to come through in '05?

Yes, you know, I hate going over history.

Last year, yes, we were ramping up Q1 for a lot of what we call new packages, MLF and stacked BGA were the primary ones, in anticipation of a strong second and third quarter, a lot of it driven by the wireless people and various products, and it didn't happen.

So we -- our customers forecasted it, but it never happened, and we still have that capacity in place, and we'll really have our P & L, whenever that starts selling, we certainly hope later this year sometime.

On the -- on the flip chip side, you know, that's all been recent activity, you know, and, I'm sorry, I forgot what exact question -- repeat --.

What I'm saying is, what are the chances that we kind of face the same situation with the flip chip?

You're spending the CapEx, but then the customers -- you know, the revenue doesn't materialize in '05.

In other words, how certain are you that this revenue is going to --.

I think we've got a little bit different share, maybe a lot.

We were calling a market segment to continue to grow, last year.

This year we're counting on an existing market that's converting from wirebond to flip chip, and those designs are already done.

Okay?

It is no secret, and you can ask anywhere, that the graphics-- especially the high-end, mid-range graphics are converting to flip chip.

The gaming chips are going to flip chip, et cetera.

As speeds go up you go to flip chip, et cetera.

Wafer levels, CSP on the lower end, for a bunch of other reasons, are taking off.

So this is more a conversion from current market technology to new market technology called flip chip and wafer level CSP and that's why we're confident this market is going to sell.

It's all about execution for us now.

And by the way, on contracts, they don't exist in our business so we have the same risks as we always have if the market doesn't appear or the customer loses his market share, then we'll have to ride that horse that we jumped on.

That's the way the game is played here.

Just one follow-up actually for Ken.

Again, Ken, as we look at next year, we again thought that CapEx was going to go down, therefore this year the free cash flow is going to look a lot better than it did in '04.

With CapEx already, at the first pass, looks like you're spending 50 million more than what we first heard.

What do you think are going to be the free cash flow prospects for '05?

Is this, in fact, going to be a free cash flow year, or are you going to see a break-even to negative free cash flow in '05?

Initially in Q1 and Q2 our current outlook is -- well, clearly, in Q1 our current outlook is that-- with free cash flow would be negative based on the CapEx and the relatively modest level of sales.

As we go through the year, we see an improving environment and we believe that we would be cash flow positive in the second half of the year.

The whole year as a whole do you still expect to be positive to neutral?

Neutral to positive.

One more comment on CapEx.

If we weren't quite sure that we were going to sell that capacity out in 2005 we wouldn't be putting in place for all these new things.

If it wasn't for the flip chip and associated activity of that -- in that area, and just a general market that we were servicing, we'd be spending substantially less CapEx in Q1 and Q2 this year, way before the upturn to occur.

So I look at this as good news for Amkor that we have things to serve that look like major opportunities in the near term.

And when the market, general market, returns, then it'll be -- another layer of growth on top of what we're doing now.

Jeff Harlib, Lehman Brothers.

Just the areas you're spending on: wafer level, CSP, flip chip, wafer bumping; can you talk about how much of your business that currently comprises and how much it should comprise by, let's say, over the next year?

And how are the margins generally on that business?

And then I have some follow-ups.

Yes.

Wafer level CSP is very, very small.

We have a very small operation in Korea, prior to the Unitive North Carolina acquisition, and that business is just starting to ramp up now as we go through late -- into Q1 and Q2.

So it's a very minor portion.

On the flip chip area, we -- maybe it's about 5 percent of our total, maybe on a last quarter basis, something like that.

And, you know, I can't give you exact numbers, but when we're spending, maybe $100 million in that area, you know that we're looking for more revenue than CapEx.

So well over $100 million of business opportunities are expected from that, and we think we have enough opportunities in that area where we're picking the better ones, such that we believe we'll get a pretty good gross margin out of those transactions.

Okay.

Average or --.

Above average.

All right.

And just -- does this mean some of the advanced capacity you added last year may not, you know, be used?

Because the products are transitioning, you know, out of wirebonders into flip chip and --.

No, we think there'll be substantial demand in the future for general high end wirebonders which we purchased last year, and diebonders, for fine placement stacked CSPs and the like.

We've put a lot of MLF capacity in place and we're adding capability there for different kind of package structures and size of the package.

So we see very little -- very little-- very low probability that those assets won't be utilized sometime in 2005.

Timothy Arcuri, Smith Barney.

Actually, John, I'm still somewhat confused about the IBM business.

About three months ago when we talked about IBM we were talking about a number that admittedly has come down from 150 down to 120.

But at the time we were talking about kind of a front half loaded revenue structure, now we're talking about an extremely back half loaded revenue structure.

So if I were to kind of superimpose the industry's decline on the decline in the-- or on the apparent decline in the IBM business, it seems like the IBM business basically went from being front half loaded to being extremely back half loaded.

So why more of a change here in the IBM business relative to the other pieces of the business?

Well, I don't recall us saying it would be back loaded.

I thought that we said it would continue to be -- consecutively loaded-- incrementally loaded as we-- and it would take us to Q3 to get through most of that loading until we reached the IBM-- pure IBM foundry activity, and then that was purely dependent on how successful IBM was, and we were with IBM's end customer.

So I guess we have different recollections of how that was played out.

If you just look at the math, you start bringing more and more activities, you don't bring all that capacity in, all that demand in on the first week of January.

You have to lay it into these factories and they don't want to penalize their existing suppliers too badly, too quickly, so it's a planned incrementally increasing rate through the first half for the existing IBM business and then the foundry business that's related-- and some other things come on in late Q2 and Q3, and Q4.

So that's always been my recollection.

Maybe we didn't explain it very well.

And although there is -- we had expected more in the first half because that's what-- the numbers IBM gave us, but it now looks because of the market slowdown they'll be a little bit less.

Okay.

Maybe I had that wrong, then.

I guess maybe as a second question, then, if I look historically at utilization and I look at, you know, that relative to gross margins, there's been a very close correlation historically.

And, you know, it seems like if you look in the numbers and there's been some questions asked about this, but if you look in -- what you printed for the December utilization and what you're guiding for March utilization, relative to your-- relative to your margins, it seems like there's kind of a lot of variability in Q4 and Q1.

Is mix becoming kind of a bigger factor in the business or, you know, does this not hold anymore that utilization correlates with gross margin, or what am I missing here?

There are a lot of factors that come into play there, Tim.

Everything from the mix to the timing on the capital expenditures are going in and the rest, but that being said, I think there is an ongoing correlation.

I think there could be spikes from quarter-to-quarter, but I think if you look at it over a period of four to six quarters you'll see the correlation.

I think in addition to that, we've always had a good deal of operating leverage in the model, and right now we're in a period where the sales are declining, and with that decline, you do get the reduced gross margins.

In addition to that, as I say, we have layered in some additional fixed costs with the acquisitions that we brought on board, which we believe are going to provide us, in the long-term, very good returns.

But in the short-term, it does cause some compression on margin.

Yes, let me just add to that.

As we speak, we're bringing on a lot of people in North Carolina, a lot of people in Taiwan, to train-- ahead of what we see as a very steep production curve as we go out of Q1 into Q2.

So that just costs nothing but money because the revenue isn't there yet.

So this is the price of quick expansion.

We've elected to do it, and it's going to penalize us for a little while in the short-term.

Krishna Rangarajan, CRT Capital.

I have a question on incremental gross margins.

I believe previously when you were at lower levels of utilization you've indicated that material content's about 35 percent or so of sales, allowing the contribution margin-- incremental contribution margin to sometimes be as high as 60, 65 percent.

In this respect I have two sub questions.

One, given the change in the mix away from leadframe to more of substrate-based packages, first of all, what does it do to the material content in-- your average material content as a percent of sales?

And second, given this mix shift as well and the fact that you're bringing in additional people to man the bumping lines and so on, how should we be thinking about incremental gross margins going forward, assuming your revenues, you know, start posting some sequential improvements beyond Q1?

Very good question.

Actually, I think in the short term you could lack at the historical model, and you're right, with respect to materials cost, probably in the range-- 35 would probably be a little on the low side.

It would probably be somewhere 35 to 45 giving you a 55 -- from 55, 60 percent contribution margin.

And that historically has been where we're at.

And to the larger extent that's where our mix is at right now, but as we move more and more towards the flip chip model, the total turn-key and bump, you absolutely will have a higher material cost so the material cost as a percent of the selling price will go up.

But the other aspect of that is you're also going to have a higher ASP.

The gross margin in terms of percentage will probably be somewhat less but, in terms of absolute dollars, a higher amount going forward.

So I would say, once again, for the short term I think the historical model is fine.

As we go forward, you will see a shift.

But we'll see that as we start to develop that business more, as John had indicated.

Right now the flip chip bumping is a very, very small part of our total business, but it will change going forward.

I had a follow-up question as well.

Given the fact that you're ramping up capacity in certain areas, as your mix is shifting, how does that ramp up in overhead, labor, what have you, dilute the incremental gross margins?

How should we be modeling that?

That's hard to judge.

It depends on how fast the ramp builds, and that'll depend on where we see the sales going from quarter-to-quarter.

Quite frankly, historically I think if you look over a period of six to seven quarters incremental gross margin probably in the range of like $0.50 on the dollar of sales, probably right now probably more in the range of $0.25.

Yes, just to reminder that this P & L, the gross margin hit is only in the short term, because when we acquired Unitive in both North Carolina and Taiwan, they had, you know, sizable organization -- overhead structures relative to the revenue curve because they were in development phase of their existence.

So we won't have to add to their overhead structure nearly at the rate of revenue growth as we go through 2005 and 6.

So even though we have a little problem here with -- indirect and direct labor buildup in anticipation of the production curve ramp-up; once we get through that, then it reverses itself, and overhead gets better as we go forward, as revenue climbs.

Medhi Hosseini, Friedman, Billings, Ramsey.

John, I have a question regarding your comment.

You referred to the kind of a slow down or digestion period from the past couple of quarters as a mid-cycle correction.

So if we were to use that as-- looking into '05 and '06, it doesn't get us that excited regarding the revenue growth prospects in '06.

Am I reading too much into this?

Yes, I think you are.

Our view and the model we're working by is that based on the curves that we already see, if the curve that stopped going down is now going up, as far as the customer's forecast continues up, as we expect it will; then it is a pretty classical mid-cycle correction.

And it does not last just, you know, three quarters.

It lasts typically much longer than that.

So we'll also see-- so we're going to play by the historical data and say it's going to go through 2006 too to some extent.

Sure.

But who knows?

I'm no sooth, just like you.

Right, right.

But you've got to make a call, you've got to call it one way or another.

You can't sit back and not make a decision.

Absolutely.

Nobody has a good crystal ball.

Regarding the outsourcing trend that was started sometime in 2000, 2001, help us understand to what extent that trend is moving forward.

Are we seeing the marginal benefit of outsourcing diminishing?

In other words, is outsourcing slowing down?

The big part of outsourcing, has it already been done?

No, I think it's just beginning as we've been talking about and, you know, we do see some of the IDMs still selling off their factories.

We saw ASC and NEC transaction here mid last year.

We can't talk about it but we see a number of our customers, small and larger, looking to sell some of their facilities and get rid of them, some of them completely; depending on who we're talking about.

So there aren't too many anxious buyers out there right now but, you know, that's still happening.

We see that the fabless are still outpacing the people with -- assemblyless, I should say, are still outpacing growth of the people with their own factories so far, although 2004 wasn't that much of a big gap.

And we especially see, in the area of wafer level CSP and flip chipping that very few IDMs have the capability, and the ones that have it have older technology and they're looking to outsource.

So as that market gains more and more traction and we think it'll grow very, very fast in the next five years, then that will help the acceleration from in-house to outsourcing just by the math.

So we're still -- beating the drum that outsourcing is going to continue to happen in a big way.

Howard Brazinski, EMI Capital Markets.

Can you talk about any potential debt repurchases or convert repurchases?

We continue to monitor our capital structure but we have nothing in the pipeline right now in that regard.

Eric Tobin, Bank of America Securities.

Could you talk about your expectations for SG&A through '05 with the variability in the legal expenses as well as the corporate overhead that's being brought on?

And, secondly, what your feelings are on the Unitive earn-out.

In the last Q you said it would be 0 to 2 million.

As this business is looking more and more attractive, do you expect that to change?

Well, let me address the SG&A first, if I may.

SG&A right now, we have it modeled out kind of flat, with a small increase in certain areas.

That being said, we believe that we have built into there the legal costs running at the current rate which are higher than our historical norm.

As we said this past quarter, we spent 5 million.

We're hopeful that some of those litigation costs and expenses are going to start to decrease but, for right now, we don't know that for certain and we have it built into our model.

So the other thing is we will try to, as we always do, focus on decreasing G&A but right now we have it modeled out kind of flat.

With respect to the Unitive acquisitions, there are earn-outs actually in both of the acquisitions, Unitive in North Carolina and Taiwan.

Right now our estimate is, is that due to the downturn the earn-outs will be very modest.

Avi Bennis, J.P. Morgan.

Two questions.

The first is kind of a follow-up on the last one.

Looks like your midpoint of gross profit is down about $30 million.

Then if we're taking SG&A and all that flat does that mean EBITDA is going to be down about $30 million in the first quarter?

Second question is, just on your balance sheet and your liquidity right now, do you still feel comfortable that you'll be able to meet the maturities coming up, '06, '07?

I missed the first part, but I'll go to the second part.

We're very comfortable that we're going to have sufficient liquidity to meet the maturities of the '06 converts.

There could also possibly be some type of a refi on that convert.

We always look at capital structure but no firm plans at this point in time.

You went a little bit quick on the first couple of questions with respect to EBITDA, so could you go through that again for me?

Sure.

It's basically that if gross profit is down about 30 million at the midpoint of guidance is that the range that we should expect for EBITDA as well?

It sounds like from your last answer about SG&A being relatively flat that that seemed like a safe assumption.

And you're saying what on the gross profit?

Down how much?

About $30 million on the midpoint of guidance.

Yes.

Okay.

And then just back to the capital structure --.

But that won't all be down to EBITDA level because part of that's -- okay, yes it would.

Okay, I take that back.

Okay.

And then on the capital structure, you said you obviously view opportunities.

It's true, though, that you're not-- you can't use cash on hand today to buy back those sub notes before maturity, correct?

That's correct.

Okay.

So when you say revisit your capital structure, you mean either doing a different kind of debt transaction, maybe going to them with a new security or maybe even issuing equity?

There are lots of possibilities, and you just mentioned two of them, yes.

Jordan Teramo, Mackey Shields (ph).

Yes, just a follow-up.

I mean I guess-- you seem to say that -- confidently that you're comfortably able being able to meet '06 maturity.

At the same time you've shown zero prudence in terms of your capital spending last year and now increasing the forecast this year.

At the same time given kind of how numbers are looking, numbers, EBITDA, free cash flow is going to be down big again this year.

And, you know, obviously on this call you're like, oh, things are looking good, things are looking good, but last year you completely missed the ball.

So I guess my point is you haven't earned any trust for at least me to sit back and say, oh, they're comfy '06 maturity, okay, I'm comfy too.

So I guess I need to hear from you more concrete plans or details given you have 233 million due in 15 months.

Well, we can't tell anything with certainty, but I can tell you this.

We do have plans, and the plans that we have in front of us are for improvement in the second half of the year.

So, you know, I can't tell you with any certainty, but certainly we believe that we have a road map and a plan that indicates to us that we will be able to meet those maturities.

Cecil Alipat, Chonnam Asset Management (ph).

I just wanted to go over just the acquisition payments again.

What exactly do you have coming due in terms of either earn-outs or related to your acquisitions?

The Unitive is not going to be that much this year.

Is there any other payments we should be thinking about?

There are a couple of payments.

In the first quarter we have a payment of 18 million due on the Fixa (ph) building that we acquired in Taiwan.

In the third quarter we have an installment payment on Unitive of 16 million.

And in the fourth quarter there's -- we have an option to buy out the remaining 40 percent of Unitive should we so desire, and right now that's our intent.

That would be another 18 million.

Okay.

Now, when you're saying you expect to be either free cash flow positive or neutral for the year are you including these payments?

Absolutely, yes.

So then I guess what I'm trying to do, in terms of just building my model and, you know, we have guidance for the first quarter and some guidance in terms of how we should think about the second quarter.

It would indicate, you're saying a reasonable gross margin to assume is something in the 20 percent range.

Are you looking to exceed that for the second half of the year in order to be able to be free cash flow positive, and how should we think about it?

I apologize for beating a dead horse on the gross margin issue.

That's okay.

We're not giving specific guidance for the balance of the year, but you're absolutely right, the margins would to have go above the 20 percent in the second half of the year to-- basically meet our plans, and that's -- we don't think that that's an unachievable goal at this point in time.

But it all depends on the market.

Thank you.

And that concludes today's question-and-answer session.

I would like to turn the conference back to management for any concluding comments.

John here.

All I've got to say is we see some rays of light coming over the horizon.

We believe it's sunshine and not the freight train.

So we'll find out as we go through this quarter and talk to you guys at the earnings call in Q2.

Take care.

Thanks.

Ladies and gentlemen, that does conclude the Amkor fourth quarter 2004 earnings release conference.

If you'd like to listen to a replay of today's conference, you may dial 303-590-3000 using passcode 11015635#.

Thank you again for your participation on today's conference and you may now disconnect.