庫力索法 (KLIC) 2004 Q4 法說會逐字稿

Good morning, ladies and gentlemen, and welcome to the Kulicke & Soffa fourth quarter 2004 earnings conference call. At this time, all participants are in a listen-only mode, and a brief question-and-answer session will follow the formal presentation. If anyone should require operator assistance, please press star zero on your telephone keypad. As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Mr. Michael Sheaffer, Director of Media and Shareholder Activities for Kulicke & Soffa.

Thank you, Megan. Good morning, everyone, and welcome to Kulicke & Soffa's conference call for its fourth quarter and fiscal year 2004 results.

An audio recording will be made of the entire conference call, including any questions or comments that participants may contribute. The audio recording will be available on the Internet for a limited time and may be accessed from the Kulicke & Soffa Web site at www.KNS.com.

The content of this conference call is owned by Kulicke & Soffa Industries and is protected by U.S. copyright law and international treaties. You may not make any recordings or other copies of this conference call. You may not reproduce, distribute, adapt, transmit, publicly display, or publicly perform the content of this conference call in all or in part without the written permission of K&S.

Today's remarks are governed by the Safe Harbor provisions of the 1995 Private Securities Litigation Reform Act. Actual results may turn out significantly better or worse than indicated by any forward-looking statements that we may make this morning. For a more complete discussion of risks associated with the operations of Kulicke & Soffa, please refer to the Company's SEC filings, especially the 10K, for the year ended September 30, 2003, and our most recent 8K.

And now it's my pleasure to introduce the host for today's call Scott Kulicke, CEO and Chairman of the Board. Scott?

Thank you, Michael. Good morning and welcome to this call.

We want to cover three topics this morning before we take your questions. The first has to do with the year just ended, the Company's achievements, and the things we need to keep working on.

The second is our view of the semiconductor cycle. And the third, the steps we are taking to maximize our performance now and in the next upturn.

Some of this year's achievements can be quantified in terms of market share. For instance, we believe our Maxum Plus bonder solidified our position as leading provider of wire bonders. Beliefs supported by DLSA's latest data.

Similarly, our bonding wire business grew at about double the rate of growth of IC units, implying strong share growth. This kind of market share growth reflects the high levels of customer satisfaction that flows from a combination of the right products and an organization that crisply executes the day to day running of the business.

Only in test did we fail to grow market share. In this part of our business, we're just now kicking off the next round of consolidations that we previously described, further lowering costs and also creating the organizational platform from which we plan to launch a series of next generation test products in 2005. I'll touch on these new products in a minute.

Another major theme for the Company in 2004 was our broad push to reduce the Company's cost base.

I'm going to turn the mike over to Maurice Carson, our CFO to summarize our financial performance.

Thank you, Scott. And good morning, everybody. Let me take you through some of the key financial indicators from this quarter.

Gross margin percent was down 500 basis points from the prior quarter. This was driven by lower factory absorption in bonders and an increase in the proportion of wire in our business. Wire, which includes the cost of gold, carries a much smaller margin percentage than bonders.

Op ex was 1.2 million better than the prior quarter, due to lower incentive accruals.

Net interest was better than the June quarter by 800k, as we received partial benefit of the redemption of our 5.25% bonds. The full effect will be seen in the December quarter when net interest will be approximately 500k.

At the same time, we took a charge of 1.9 million for the redemption of these bonds, 800k of this a non-cash charge from writing off the issuance costs from these bonds.

On to the balance sheet, cash.

Cash, restricted cash from short-term investments fell by $46 million, due to the $55 million we spent on redeeming the 5.25% bonds.

Inventories were higher by 4.5 days. 2.4 million of this was in bonders as we were unable to completely adjust to the slow down in equipment sales. A million seven was due to the setting up of two new wire customers on a consignment program, and general growth in the market share in wire.

The other item of significance on the balance sheet was the reduction of debt associated with the redemption of the bonds.

Regarding cash flow, cash flow from operations was $14.8 million, with working capital performance positive in absolute dollars but weak in days. Capital expenditures was $5.3 million, and some of the major projects that drove that were advanced bonder development, gold wire manufacturing capacity, systems to support our China manufacturing initiative, and increased capillary production.

Looking at the total year 2004 compared to 2003, there were several major cost projects that were completed. We sold our flip chip subsidiary, a business that was not profitable and did not fit into the strategic plans for the Company. In 2003, this business lost $7.9 million.

We sold our printed circuit board business, also never profitable, and we'll save 500k a year from this move. We closed an underutilized factory in France, a move that will save 800k annually.

We refinanced our high interest debt, a move that will save $13 million in interest expense in 2005 compared to 2003.

We aggressively ramped capillary production in China. 65% of capillary revenue shipped from China in the fourth quarter of 2004 compared to 35% in the fourth quarter of 2003.

We began building cantilever probe cards in China. 21% of our cantilever shipments were from China in the fourth quarter, compared to zero in the fourth quarter of 2003.

All told, we have significantly reduced our fixed costs in absolute terms during a year with very significant revenue growth. Scott?

Thanks, Maurice.

My second subject is the semiconductor cycle. Wire bonder purchases have historically been a leading indicator of cyclical change in the semiconductor business.

This cycle, bonder orders peaked in the March quarter and have been soft since then. We expect further decline in bonder sales this quarter to the trough levels of the last few cycles.

And as we get through the holiday season, it is reasonable to expect a certain amount of softness in our other product lines as well. This view of the cycle is baked into the 115 to 130 million revenue guidance we're giving for the December quarter.

We don't know how long the industry will stay in the trough this cycle. But we do know it's only a matter of time before these conversations shift to the timing of the next recovery. And we expect that just as in past cycles, wire bonders will be the leading indicator [up].

All of this brings us to the question of what are we doing with K&S. Most of our recent conversations have centered on cost reductions. And as Maurice indicated in his summary, we made a huge amount of progress in 2004.

There is more in store for 2005, especially in our test business. We view cost management as a continuous improvement process, and as with any continuous improvement process we're always working on the next project.

These cost improvements have been the focus of much of our conversations with you in the last year, but now, we're at a point where cost improvement is just part of what we are. It's time to shift the discussion so that we have more emphasis on customers and on market share and on technology.

Even though we're near the trough of the cycle, our work with customers is ongoing, as we try to win new customers, or new applications with existing customers. For instance, we have more than a dozen wire bonder evaluations going on right now, with many more evaluations of materials and test products as well.

No matter how well our products are performing, there's always opportunity to increase share.

On the technology side of the business, we are equally busy. The next upturn will more or less coincide with the volume ramp of 90 nanometer wafer fab processing.

This will strain the performance of today's installed base of equipment and the corresponding specification level of the materials and test interface products currently qualified for production. In response, we have new products for every major part of our business in the development pipeline.

Some of these products are incremental improvements and some are completely new. That we can fund this broad-based product development effort is one of the key benefits of our improved cost structure.

These various activities, the ongoing cost reductions, the work with customers, and the new products are all part of our plan to make K&S a company that will bring superior performance over the whole semiconductor cycle.

Megan, we'd be happy to take a few questions.

Thank you, sir. Ladies and gentlemen, at this time, we will be conducting a question-and-answer session. If you do have a question, you may press star one on your telephone keypad. A confirmation tone will indicate your line is in the question queue. You may press star two if you would like to remove your question from the queue. For participants using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. Please hold one moment while we poll for questions. Our first question will be coming from Mike O'Brien of Bear Stearns.

Hi, good morning, Scott. Just a quick question. First of all on utilization rates, you know, if you can give us an idea of where they're at, and you know, what the change has been recently, continuing to trend down, finding a bottom, just a question on that?

And then just really on the growth of the wire bonder business, you know, commentary out of, you know, Siliconware, when they did their earnings calls saying, you know, we don't really know when or if we need to buy additional bonders. Obviously the "if" is probably, or a certainly an exaggeration, but it doesn't sound like they're in a big hurry any time soon. If you could comment on those, that would be great.

Okay. First, the utilization, the curbs, which by the way, I guess are on the Web site or will be on the Web site.

It will be on the Web site.

The update of curbs will be on the Web site so you can see for yourself later on. There is no appreciable trend line at this point. The data has been pretty flat. I mean it bounces up and down, week to week, so I don't think that we can call a trend yet and it's in the high 70s.

Obviously, we expect it to soften up a little bit after the holiday season is over. But we don't see anything that I'd say is more than just normal noise patterns right now.

Let me make a comment about the holiday season while I'm there. We tend to think of the holiday season as the Christmas season, which means that it's over from a manufacturing point of view. But I think that increasingly with the growing prosperity in Asia, the holiday season extended to include Chinese New Year, and it's one of the reasons why in the last year or two, you've seen the holiday season from a manufacturing point of view extend later and later into the fall.

So we are expecting to see some softness in our material shipments and in the bond reutilization numbers. As I said in my comments, that's baked into the guidance we've given you, but we don't really see the trend lines turning down in any appreciable way or in a way that I would say is more it's more than just normal noise, which means it could go up for a couple of weeks or down for a couple of weeks, but that's not enough to declare a pattern.

About bonder capacity additions, again, as I said in my opening comments, they are rapidly reaching typical trough of the cycle levels which for us is a few hundred bonders a quarter. That's just the way it gets to. That's consistent with the data you've heard from Siliconware.

Typically, the big subcontractors are the ones who, you know, they drive the upturn, they drive the downturn, we have seen a couple, some demand, some sporadic demand from some of the, a couple of the other big subcontractors, but compared to their buying patterns of six or nine months ago, it's all very anemic, and it is what you expect from cyclical behavior.

Okay. And do you expect, I mean with the move to flip chip, does that make the growth potential, or the peak for wire bonders less, when we do start to grow again?

No. There's a real interesting chart on our Web site in the Financial Section that is extrapolated from VLSI data. It talks about interconnect growth over the, projected interconnect growth over the next five years, and if you look at that, you'll see that the growth rate in wire bonded interconnects, is, I don't know, something like a 20% compound annual, between now and '08, VLSI is projecting that it's something approaching a 20% compound annual growth rate. I think that's the number, I'm recalling it from memory.

We'll put that chart on the Web site as well.

We'll put it up. But it's also part of our standard financial presentation. And the point is, there's lots of growth available in wire bonders. Obviously, it will be a roller coaster ride. But the peaks we expect will be higher each cycle, lots of growth in wire, lots of growth in capillaries.

Correspondingly, we get a pretty good look into the flip chip business through our vertical probe card sales. Vertical probe cards, Cobra probe cards sell mostly into the array market, where bond pats are organized in an array over the whole face of the chip.

And you know, the demand for Cobra-like cards has not been growing as quickly as the flip chip pundits would have you believe. So again, it all supports the idea that, yeah, flip chip's going to take some share, but the major part of the action remains with wire bonding in the semiconductor business.

Thanks, Scott.

Thank you. Our next question will be coming from John Pitzer of Credit Suisse First Boston.

Yeah, good morning. Got a couple of questions.

Hey, John.

How are you doing?

Pretty good.

Can you just, you know, when you look at your guidance for the December quarter, can you help us understand sort of the equipment sales falloff versus the materials? Is historical materials sort of flattish quarter-on-quarter on the December quarter and what's your expectation? And then remind us again what materials might do on a seasonal basis going into the March quarter.

Right now, materials actually I guess will be flat, maybe up a little bit. I'm sorry, down slightly this quarter. That's a forecast based on seasonal tail-off in the month of December. You know, it would probably be a little more tail-off in the March quarter. But you know, that's just sort of a rule of thumb based on past cycles, this is not a particularly forecast-driven data point.

But that's historical what it does. Because the material sales all track IC units and that's what you would expect IC units to do in terms of the calendar cycle.

And then Scott, can you give us a sense of how the business on the equipment side is breaking out between subcons and IDMs and is the pattern significantly different between those two groups? And I guess when you look at the subcon space, is it after Chinese New Year when you start to get some better clarity about demand trends there for next year?

First, that has been the historic pattern. In general, the subcons turn down more violently than the IDNs. They both obviously have turned down a lot.

We're seeing, you know, at the trough at the cycle, you see drips and drabs, you see a lot of people that, I don't want to call them second tier customers, because we love them all, but they're not the big names you think of when you oversimplify the marketplace.

We do a lot of work there. It's an opportunity for us to go in and target lower volume applications, applications that are a little bit away from the mainstream of the business. And of course, what's all important now is the qualifications of next generation technology that we're doing.

So there's a lot of that stuff going on. Some of it is subcon, some of it is IDNs. But the ratio is shifted a little bit more back to the IDNs right now.

And then Scott, the last question --

But that [inaudible] as I say that is not a secular shift. That is just a little, a minor cyclical variation and the long-term secular shift is clearly toward the subcons, will continue that way, and when the cycle turns they will be back buying bonders by the ton, it will seem.

And then Scott, lastly, you guys have done a great job on the breakeven or the fixed costs out of the equation for most of this year. I'm just kind of curious when you look at sort of your revenue guidance for the December quarter and the mix you expect in December, where do you think breakeven stands today?

And kind of, you talked about further cost reductions. Where could that go over the next couple of quarters then?

Well, first, the breakeven, I mean in the, I have to remind everybody of how we've been calculating it which is operating expenses divided by gross margin percentage. And with the shift in market, a shift of mix, the gross margin percentage has gone down. So it looks like breakeven went up, even though the operating expenses continue to trend in the right direction. You know, and I'm not going to, beyond that, I'm not going to give you a forecast for the December quarter, except it will be up and again, it will be up as a mixed basis. Maurice is giving me a note here which I'm trying to read as he writes it.

We're looking at operating expenses either flat or down slightly, excluding some of the costs that will come from the consolidation steps Scott talked about earlier, so there will be slight improvement in the expense basis.

Yeah, but as I said, because of mix reasons, and Maurice touched on this a little bit even in the September quarter, the gross margin goes the wrong way at this point in the cycle. And that's just the way it is, that's a cyclical issue.

You asked about where we're going to go, I'm not going to put numbers on it, but we have the next round of factory consolidations in the test business. We are continuing to move product, test product to Asia, which reduces gross margins there. So there's a whole lot of efforts that are underway that will probably take a couple of quarters to completely play out from the point of view of the income statement. I'd give you a teaser but I'm not going to put numbers on it.

Thanks, Scott.

Our next question is coming from Edward White of Lehman Brothers.

Hi, Scott. I was wondering if you could talk a little bit more on the technology side about what you started to talk about with the, you know, the transition to 90 nanometers and early work on 65 nanometers, how will that affect wire bonder, and test technology? In other words, you know, what are the issues that you have to address? And that, you know, that really represent opportunities for you as you get those transitions, you know, both on the wire bonder side and on the test side?

Okay, let's start with the wire bonder side. For the last, for most of the last decade, the trend line was the pad [paged] scaled more was linearly with feature size. And then in the last couple years that trend line was interrupted as people moved to multiple rows of bond pads, which put some changes in what you did with the wire bonder. No longer were you driving for the absolute accuracy necessary for finer and finer pad pitches, but you started to stress things like looping ability, which was really key in multiple rows of bond pads.

Looping is hard to talk about because there's no single number you can hang on it. No simple metric, the way you can talk about pad pitch. It's a more complicated set of interactions between the bonder and the materials, in the wire and the capillary.

We expect to see continuing requirements for longer lower loops, which again is a wire and a capillary issue. We also expect to see the pad pitch reduction curve start to pick up steam again.

We're starting to see customers wanting to evaluate 35 micron pad pitch capability. That goes back to requirements on the bonder itself, the ability to sustain small balls, the absolute accuracy of the bonder, but also longer and longer wires as you shrink the bond pitch. If you work it out geometry, that means the wires have to get longer in most applications. So we'll build that capability into new bonders that we'll be introducing this year.

We're also starting to introduce some wiring capillary products, we have them in the pipeline and we've actually already introduced the first wire product which has the unfortunate name of [Braidix]. It's a wire that's built to have very high reliability with very small bond pads or ball sizes, so there's that kind of effort going on.

And in general, you will see us respond with a suite of products which is to say a bonder plus a capillary plus a wire as opposed to just a bonder product, because we think that to really optimize the performance, you have to deal with all of those things as a system, and we think that's an area of increasing competitive advantage to us, because we're the only company that has the ability to optimize those three variables as a system.

On the test side, we're seeing, well, as you see pad pitch shrink, you will also see some corresponding reduction in probe needle sizes, the ability to space probe needles closer and closer together.

The other thing that you'll see is an increasing IO count, which puts a lot of pressure on the ability to build these probe cards with more and more points on them, or sockets with more and more connections on them, and the ability to get enough current through the things. You know, so there's those trend lines.

Also, in the test side, you'll see a move towards increasing parallelism. The most obvious case of that is in the memory side, where it's pretty straightforward and pretty well known. But we're also starting to see increased parallelism even in logic.

Now, it is not that bi-64, bi-128 kinds of duct counts or device under test counts, but you're seeing logic chips with a bi-4 or a bi-6 or a bi-8 configuration and that also puts a lot of stress on the probe card, and in test is one of the areas where we've got probably the most exciting new products coming out in the pipeline as we start to formally introduce them, we will be talking about them in these calls.

Okay. Finally, one quick question, relating to cost, where do you think you are now in terms of getting wire bonder costs down to, you know, down to, you know, the world standard levels? I mean I know were you close in terms of having them down as low as the lowest out there. Is there any progress made on that?

No. Well, first, because of overhead absorption reasons, all the costs go the wrong way in this part of the cycle and that's just mathematics and accounting.

Yeah.

But in terms of the underlying designed-in costs, we're basically at a hiatus with the Maxum Plus, you know, we're not doing a whole lot of work in reducing the costs of the Maxum Plus because it's nearing the end of its life. There's a little bit of improvement with its replacement that will be out for a sampling of the replacement will be in the spring, and we'll talk about that product then.

Okay. Great. Thank you.

Our next question will be coming from David Duley of Merriman.

Yes, good morning. I was wondering, when you talk about trough units, just so I can get a reference point, is that the three to 350 range? And then could you talk about --

David?

Yes.

I said a few hundred and I'm going to leave it at that.

Okay.

You know, so everybody understands, it's not just me torturing David. David is like me and many of you would like me to give our absolute unit shipment every month or every quarter, and Maurice keeps kicking me under the table and says don't talk about units at all. So I'm sort of splitting the difference and calling it a few hundred.

Okay. And well, since you won't give us units, would you let us know what pricing is?

As we said in the price release, pricing has been flat. There has been, I mean you all, customers always want to you lower the price but we have been able to hold our prices about steady. You see a little bit movement based around customer mix but in terms of the underlying prices they are absolutely flat right now and we don't expect to see any real change in that until we get into the qualification and negotiation around the next generation of bonders, which as I said in the last question, will be in the spring.

Okay. And --

So we're not seeing ASP pressure.

Okay. And when you, earlier in the your prepared comments you talked about share gains. Could you give us an estimate of where you think you're at there, and you know, how the lower cost bonder helped you to gain share perhaps?

The lower cost bonder helped a little bit. But the real activity was with the increased penetration in some of the big subcontractors, the ones that we didn't already have 100% of, and increased penetration in some of the less big subcontractors that you don't normally think about. The VLSI data I believe has been published.

It's preliminary. It's not out yet.

Oh, sorry, I was told by Christian that it was out. Okay. I've got two different, we see the VSLI numbers before they're public and some people are telling me they've published and other people have told me they haven't published yet, so I'm not allowed to talk about them until they've been published except to say over the course of the last cycle according to VLSI we gained some share and I'll have to wait till we confirm that VLSI has in fact released those numbers and I apologize.

Thank you.

Our next question will be coming from Bill Lu of Piper Jaffray.

Yeah, hi there, good morning, guys.

Hey, Bill.

You know, I hope I'm doing my math right but it looks like margins in the [declune] business were actually up sequentially?

Maurice is checking his cheat sheet but I can't imagine.

No, down slightly.

Down slightly, yeah.

I'm calculating about 44% so that's not correct?

No.

What time is it out there, Bill?

Okay, I'm sorry, I'll double check my math then. Hey --

They were down a little bit. But I mean not an appreciable amount.

Okay. And then along the same lines, in the test business, I think Scott you just said that moving to Asia is going to cost a bunch [inaudible] near-term, I'm assuming it's just because of the transitional expenses and such. How is that going to play out?

I must have misspoken. No, moving to Asia should increase our margins. We clearly have a lower cost base on a probe card built in China than we have on a probe card built in California or Europe. So if I gave that other impression, than I'm, I'll give you my excuse which is I just got off a red eye from Arizona so I'm a little fuzzy myself this morning. So, you know, clearly, the moves to China will reduce our cost base and improve our gross margin structure in test.

Okay. That makes a little more sense.

Okay. Thank you for catching me on that.

And then just one more thing on test. Scott, it sounds like you've got some new things coming on board. I know you don't want to talk about them just yet, they're not released yet, but are all the development efforts on track now?

Maurice is saying that yes, say yes, say yes. Some of them are absolutely on track. There's a couple that are, I mean only a couple weeks late. And where specific technical items or supply chain items but they're substantially on track.

That's one of the things I was doing in Arizona yesterday. And we're really excited about where we think we can go with test. You know, there's a guy out there who's got a real dominant market share position in terms of form factor that we think is vulnerable and we've got some products aimed right at the sweet spot of his business.

And then on the package side, which is a part of the business that people don't really understand very well, we see some developments that over the long-term could make our socket business maybe the biggest part of K&S. And I'm thinking a five-year time frame. So we think there's just huge opportunity for us there. And we're really redoubling our efforts. But in terms of short-term execution issues and cost issues and longer term market development and product development efforts.

What's really driving the socket business?

Well, if you sit back and look at all of sockets which include [varnon] sockets and production sockets as well as test sockets, it's a real big business. And the underlying issue there is how do you make electrical contact to the part that you're under test or holding or whatever you're doing with it.

You may have seen a press release a few weeks ago about some technology we acquired. This is technology that allows us to build very cheap, very high quality, in terms of contact resistance and very high performance in terms of inductance connections. And we think that over time, we can proliferate that capability into a range of applications that will have many hundreds of millions of dollars of served available market.

Thanks a lot for the thorough answer.

Okay.

Our next question will be coming from Tim Arcuri of Smith Barney.

Hi, this is Gary for Tim.

Hi, Gary.

How are you? Help me get a handle here on your guidance. Basically you're saying that wire bonder sales will be dropping more severely than anything else and the rest of the business will be kind of uniformly declining in the next quarter?

Well first, wire bonders always drop more severely than anything else in this part of the cycle, just like they explode upwardly more dramatically than anything else in the other side of the cycle and it's why traders love the stock so much, why traders are sitting around scratching their head and saying is it time to start to get in again or not.

But yeah, wire bonders will be off. But as we reach the trough, and the rate of decline of wire bonders will actually slow down.

You know, and as I said earlier, we expect a little bit of softness in some of the other lines of business. Certainly nothing like you see in the bonder business. And that's normal cyclical behavior.

Okay. And then you basically talked about materials, gold wire, capillaries, declining somewhat in the fourth quarter, December quarter, sorry. And yet, ASE guided to plus 5% revenue growth in December. Are we seeing material sales kind of now being kind of constrained by incremental wire bonder adds?

No, well, partly that, partly, it's partly that, offset some by market share increases, especially in the wire business. But I wouldn't take the ASE data as a proxy for the whole market, which is the other thing you did in that question.

ASE has, to their credit, has taken a lot of share in the last year. Those guys are really on a roll. And they plan to continue to take share. So I think you're seeing, I think what you'll see when you compare ASE to the other guys, is in fact they're growing when other guys are probably seeing some cyclical decline already.

Okay. Thanks a lot, Scott.

Our next question will be coming from Larry Borgman of EKN.

Good morning, Scott.

Hey, Larry.

I would, two things. One, on the 90 nanometer wafers, which of your bonders that are out there now, how many of them can handle that? And secondly, this seems like an unusually short cycle. I wondered if you --

Yeah, we noticed that as well.

I'd like to get your perspective on what caused it, the downturn, and is it just, you know, a hiccup sort of inventory type correction, or is there some other more fundamental demand factors?

The fact that the 90 nanometer issue, what we're seeing from our customers is that, let me step back. The industry talks about 90 nanometer, which has to do with basic, how the chip is designed, kind of issues. That is a little bit different than the bond pad design rules.

They can be decoupled. And you can do 90 nanometer parts with 45 micron pad pitch or 40 micron pad pitch or 35 micron pad pitch, or single row or two row staggered or three row staggered pad arrangements. They're really not exactly correlated.

And first generation 90 nanometer parts are generally being built with existing bond pad design rules, which for most customers are about the 40, or for the most aggressive customers, are about the 40 micron pad pitch area. We're starting to see the first people say no, let's see how much more we can skinny that down.

But you can change pad pitches in the middle of a wafer fab node. Yes, it requires some relay out of some of the masks, not necessarily all the masks, depending on the chip. So the kind of correlation you're implying doesn't exactly hold.

So right now, the most, most of the 90 nanometer parts we're seeing are in fact built with the same pad pitch rules or design pad, design guidelines for pads, that they had in 130 nanometer. Clearly, if people can change the design pads, the design rules for pads, they will, and that's some of the evaluation work I alluded to earlier today.

And I forget you, oh, your second question had to do with the cycle. You know, Larry, you and I have been around long enough that we've been having these conversations about the cycle for the last 20 years, and all I know is that I don't really understand why the cycles work the way they work, I'm always surprised at the downturn. We were surprised this time. So I'm not even going to pretend to be able to answer your question except to say yes, this was the sharpest upturn and the shortest time at the top of any cycle I can remember.

And I haven't a clue what that means for the downturn. Except that I know whether it's a long or a short downturn or a steep or a shallow or a real deep trough or a shallow trough, and there's opinions both ways on Wall Street. There will be in the next upturn and what we have to do is focus on being ready for that upturn.

Okay. Thanks.

Larry, Maurice wants to --

Something back to Bill's question here, Scott, I just want to come back and confirm something, okay? What Bill's looking at is on the press release which is total equipment business, which did have better margins, and I was answering the question for Bob on the--

Okay.

So if you're still out there, Bill, you're right from the press release and I wasn't answering to the whole equipment market. And when we make that distinction, because we do sell, in relatively small numbers, some other kinds of bonders which are not part of the ball bonder number, plus spares and service and related stuff. And that's the segment number you have.

Maurice was looking at a particular cheat sheet he has, which were simply ball bonders. So you're right and Maurice is the one who's fuzzy this morning. So okay, Megan, next question?

Our next question will be coming from Tom Diffely of Merrill Lynch.

Good morning. Could you please discuss the cash flow expectations for the next few quarters?

The cash flow expectations?

Cash flow from operations?

Maurice, why don't you handle that one?

We don't, just like we don't guide to the EPS we don't guide to the cash flow. We can say though that Cap Ex will remain relatively flat year-on-year. Maybe a slight increase with some of the new products, but relatively flat. And it will simply be driven by working capital, which we think we have our hands around, and our earnings. So it shouldn't be any big changes to your model. If you're modeling earnings correctly, everything else will flow through pretty close to this year.

Okay. Thank you.

Megan, do we have another question?

Yes, sir. Our next question will be coming from Robert Jaworski of Jefferies & Company.

Good morning. What was D&A this quarter?

I'm sorry, what was what?

Depreciation and amortization?

Oh, Maurice?

It doesn't come out to the detailed forecast, but I can give it to you real quick here. Depreciation and amortization was 7.4 million bucks.

Okay. Thank you. And then you said that ASP didn't change quarter-over-quarter. What is the --

That was bonder ASP.

I understand. Okay. So what is the long-term trend in ASP for bonders like year-over-year? In general?

You know, I'm not sure that there is a meaningful long-term trend. Bonder ASPs rose through the, for the decade up to the late '90s, in the 2000-ish downturn, it took a step function down, and has been about flat since then.

Okay. Thank you very much.

Our next question will be coming from Andrew Schopick of Nutmeg Securities.

Thank you and good morning.

Hey, Andy, are you doing?

Fine, thank you, Scott. And I still think you should be very happy with the initiatives and steps you've taken to effect the cost reductions in the business. The dividends will come later.

Yeah, I just, I mean from an operational point of view, I am genetically impatient and wish our guys could do it faster but, you know, it's always a tension between doing it faster and doing it right and not screwing up and losing the recipe along the way. I think our team has done a pretty good job in maintaining that balance.

If I may I have several questions I'd like to direct to Maurice.

Sure.

First, on the EITF ruling regarding the convertible debt, have you presented your results consistent with that proposed ruling whereby --

Yes.

Yes?

CoCo, we don't have CoCo in our, either one of our bonds so no change in our EPS.

Excellent. 10% customers, can you comment on them for the quarter and the year?

You know, I don't have that here in front of me, Andrew. We're just working it up for the K, ASE will still be at the top for the quarter, and for the year, same as last year, but the rest of the, that will be the only one, I believe.

Okay.

I would have thought maybe Intel.

Intel might squeeze in there, Scott, but last year, they didn't. It will be close for Intel, Andy.

For the quarter or for the year?

For the year.

For the year.

For the year.

Okay. Now accrual for taxes. Can you comment on what's happening there? And any guidance with respect to the tax provision you would expect to incur in the new fiscal year?

The tax provision should remain relatively the same. We still have NOL's for U.S., for the U.S. use and we'll continue to accrue for our foreign subsidiaries, primarily Singapore and Israel at about the same rate.

The provision in the fourth quarter looks like it is about 34%.

Yeah. I think that's probably -- The only real change will come from the, that we talked about last time was, yeah, okay, good point. Scott just pointed out we also have to accrue for AMT, which continues to hurt us on our tax accrual. And then we did get the California piece that we talked about last quarter, with the NOLs being disallowed for California for a while.

So should we be using a 30%-plus type of assumption? Rate assumption?

Yes. 30%. But, right around 30%, yes.

Okay. I want to verify a couple of numbers with you Maurice. Cash flows from operations for the year, I'm coming up with about 72 million. I had hoped you'd present us with a cash flow statement. I made that appeal last time but no luck on the press release. Just wonder if you can verify cash flow op for all of fiscal '04?

I believe that you're correct on that number. And I know we're working on it, we'd like to get it in the piece, in the press release, but you know, we got to work it through and verify we can do it right. But we think that number is about correct.

Okay. A few more for you. On the, in the noncash charges, I can see that you've got this detailed in the press release with respect to the depreciation and amortization. The depreciation of fixed assets, that looks like it's about 21.6 million. I'm really trying to get at all of the other noncash charges. I think you made reference to 800,000 of the noncash charge in connection with the redemption of the convertible debt. But I'd like to know --

For the quarter.

Yeah, for the quarter.

Yes.

And for the year, other than depreciation and intangible assets amortization, what other noncash charges were there?

Yeah, and you may, we may want to do this offline, but I'll take you through some of them. Specifically, all the debt that we retired --

I'm just looking for a number.

Okay.

If you can give me a number. Consolidated number.

No, I'm sorry, I don't have that in front of me.

Okay. I can always come back to you with that one then.

Okay.

Also on the $600,000 reversal in connection with the 3 million severance charge, where does that $600,000 reversal occur? SG&A?

SG&A.

Uh-huh. Okay. And on the accounts receivable allowances for the year, I can see that they were down from the prior period, I just wonder if you can walk us through what was recovered in which particular periods the allowances dropped from 5.9 million to 3.6 million, at 9/30/04?

No, this is the allowances against AP, did you say or AR?

AR.

Yeah, no, this was just normal kind of businesses. AR came down and the reserves came down with it. There was no specific customers added or subtracted to our specific reserve list.

Yeah, our policy is that that's driven as a function of AR, so it, it mostly --

It really mostly really passed through AR. It was driven by a percentage that passed through AR which just tracks along normally.

What were the DSOs in the quarter?

DSO was 38, I believe. I'll give you the exact number. It was actually, and I mentioned in my comments we were disappointed with the days of our working capital all the way across so DSO was 68, up from 65 in the prior quarter.

Okay. And also, with respect to inventories, how comfortable are you with your current reserve position, given the softness that you're seeing, and given the uptick that we have seen in the inventories this year on the balance sheet, any general comment there?

We're actually very comfortable. We spent a lot of time on that in the last couple of months to verify against not only inventories but any commitments and we're very comfortable with our position as it stands today.

Okay.

Given the forecast, given the various forecasts going out for the next year.

Well, Scott, I guess when the upturn comes, we'll all be surprised together.

Yeah. It usually works that way.

Thank you very much.

Thanks, Andrew.

Okay, Megan, another question or are we done?

We still have two more questions to in the queue.

Okay. Let's take them.

We have Krishnan Ranajin of CRT Capital.

Good morning. A couple of questions here, Scott. One is with respect to revenue guidance for the December quarter. You're almost, you're a little over halfway through the quarter, and you know, your backlog position at $60 million, and looking back at say the December '01 quarter which was the previous trough, you roughly shipped twice your backlog in that quarter. So why isn't your revenue guidance range any tighter than it is? And perhaps you can also comment about what kind of expectations would lead you to the high end of your guidance?

You have to understand that in an awful lot of our business, the majority of our businesses turns business. The probe card, the test socket business, is all turns business. Probe cards in particular, where you have customers asking for turn-around times, as often as short as a week.

The materials business is all turns business. An increasing part of our materials business is out of consignment inventories and customer sites. So the effective cycle time for that is measured in hours. And it decreases our ability to effectively forecast revenue to the kind of nearest million dollar levels you guys would like. And it's the nature of the industry. And I'm not going to apologize for it.

But certainly, it is no different from what it was --

Yes, as a matter of fact, as a matter of fact, it is. We have a lot more business, a lot more consignment inventory out there. In general, all our cycle times are shorter. Bonder cycle times are shorter. And as your cycle times go down, your ability to forecast gets worse. And as a part of delivering customer satisfaction and gaining share, we have driven our cycle times down in every line of our business.

All right. My second question has to do with expectations for --

And by the way, driving cycle times down is a good thing.

Sure. With respect to equipment sales, is there any precedent for sequential increases in bonder shipments in the March quarter? And also---

Yeah, a year ago.

I meant coming out of a trough.

I don't know. I'd have to go back and look at past cycles.

Okay.

You know, I, but let me say one of the things I've learned the hard way, is every cycle is a little different.

All right. Thank you.

Okay. Megan, last question, please.

Our next question is coming from Steven Pelayo of SCP.

Hey, Scott.

Hey, Steve how are you doing?

I'm doing well. These equipment gross margins, this 44.5% up about 300 bips on a 50% drop in wire, in equipment revenues, can we talk a little bit about your nonball bonder equipment business? Is this kind of a one-time thing or are you just selling a lot of wedge bonders?

No, no, it's mostly that spares are a bigger part of the mix and you make good margin on spares.

I've got to go back to 1996 on my model for you guys to have that high of a gross margin in your equipment business. So, I don't know, I'm just trying to figure out, can I continue to model this and just imagine what the upside could be once you get some volume going?

As the volume goes up, because the mix will shift, to some extent you may actually see the margin come down a little bit before the overhead absorption pick picks up and then it goes back the other way. You may get some counter-intuitive behavior there, as did you this quarter.

Okay. But right now, we kind of expect that service in nonball and other business to be remain at those levels and we can think of this rich of a gross margin?

You know, you're going to see a little bit more decline. You might see a little bit of pressure. But it's going to be what we consider noise level. I mean you guys try and forecast this to levels, to degrees of accuracy that we think are just part of normal business fluctuation.

All right. Fair enough, Scott. Thanks.

Okay. Thank you all and Mike, I think you have a few closing comments.

Yes, I do. Thank you, Scott. I would like to announce some upcoming events which will be Web cast. We will be participating at the Credit Suisse First Boston 2004 Technology Conference at 4:00 p.m. Eastern time on December 1st in Scottsdale, Arizona. And also, at the Lehman Brothers T4 Technology Conference in San Francisco, California, on December 9th, and that's at noon Eastern time.

This concludes today's Kulicke & Soffa conference call. As we announced at the start of the call, an audio recording has been made of the entire conference call including questions or comments that participants may have contributed. The audio recording will be available on the Internet for a limited time, as well as the slides that we discussed today, and may be accessed on the K&S Web site at www.KNS.com. Thank you and have a great day.

Thank you for your participation in today's teleconference.