台積電 ADR (TSM) 2005 Q1 法說會逐字稿

Welcome to TSMC's 1Q '05 results webcast conference call.

Today's event is chaired by Ms. Lora Ho, Vice President and Chief Financial Officer, and Dr. Rick Tsai, President and Chief Operating Officer.

This conference call is being webcast live by the TSMC website at www.TSMC.com and only in audio mode.

Your dial-in lines are also in listen-only mode.

At the conclusion of the management presentation we will be opening the floor for questions.

At that time, further instructions will be provided as to the procedure to follow if you would like ask any questions.

Please be advised for those participants who do not yet have a copy of the press release, you may download it from TSMC's website at www.TSMC.com.

Please also download the summary slides in relation to today's quarterly review presentation.

Once again, the URL is www.TSMC.com.

I would now like to turn the conference over to Dr. Elizabeth Sun, TSMC's head of investor relations, for the cautionary statements before the main presentation by Ms. Ho and Dr. Tsai.

Good morning and good evening to all participants.

This is Elizabeth Sun, head of investor relations for TSMC.

Before we begin I would like to state that management's comments about TSMC's current expectations made during this conference call are forward-looking statements subject to significant risks and uncertainties, and that actual results may differ materially from those contained in the forward-looking statements.

Information as to those factors that could cause actual results to differ materially from TSMC's forward-looking statements may be found in TSMC's annual report on Form 20-F filed with the United States Securities and Exchange Commission on May 28, 2004, and such other documents as TSMC may file with or submit to the SEC from time to time.

Except as required by law, we undertake no obligation to update any forward-looking statements whether as a result of new information, future events, or otherwise.

Now, I would like to turn the conference call over to Ms. Lora Ho, our Vice President and Chief Financial Officer.

Thank you, Elizabeth.

Welcome to TSMC's first-quarter conference call.

Due primarily to low seasonality and customers' continuing inventory digesting, our first-quarter business has declined as we expected.

However, we anticipate the digestion of inventory to be near its end.

For TSMC, the business condition is expected to broaden out in the second quarter.

I will now take you through our overall results of the quarter.

We have prepared some slides for you to look at as you listen to the call.

These slides are available on TSMC's website.

During the call I will refer to the slides by their individual numbers.

To start with, let's take a look at some highlights of our first-quarter 2005 operating results.

These figures are shown on slide number 4.

As expected, the quarter one results were not as good as the previous quarter.

Our quarter one revenue was NT$56 billion, 13% lower than that of fourth quarter last year.

Net income after-tax totaled 17 billion; (indiscernible) earnings per share reached NT$0.72.

TSMC ended the first quarter with NT$112 billion in cash and short-term investments.

Shareholders equity amounted to 415 billion.

During the first quarter, we shipped a total of 1.11 million 8-inch equivalent wafers, 9% lower than the 1.22 million pieces shipped in fourth-quarter '04.

Our overall utilization rate, which included TSMC fabs, WaferTech, SSMC, were 78% for the first quarter.

Return on equity for the quarter was 16.5%, about 6 percentage points lower than the 22.8% achieved in fourth quarter last year.

Let's now go to slide number 5.

Compared to the fourth quarter last year, revenue declined by about 13% due to low seasonality and inventory digestion.

The overall utilization rate dropped to 78% this quarter from 88% previously.

The (inaudible) were compounded by a weaker U.S. dollar against local currency.

However, our gross margin still managed to reach 38.9%, mainly due to continued new (ph) improvement for our 12-inch size (ph) and the sustained demand for advanced technology products.

Operating expenses was 9% lower than that of last quarter, but track 14 (ph) continued to increase commercial production volume.

Fewer engineering wafers was required, which caused R&D expenses to go down.

Then as for Fab 12, Phase II commenced commercial production, which stopped incurring of opening expenses, which led to lower SG&A expenditures.

During the quarter, operating expenses representing 10% of net sales compared to last quarter's 9.6%.

Net non-operating income for the quarter reduced by around NT$70 million, mainly due to higher value revaluation (ph) losses over short-term (ph) investment in marketable securities and foreign exchange losses from hedging activities, partially offset by the US$20 million payment of our settlement of lawsuit with (indiscernible) SMIC.

With low seasonality, utilization rates for our affiliates have all declined.

Overall investment losses were about NT$200 million.

Net income before taxes reached NT$16 billion; income after-tax came to 17 billion; while EPS decreased 25% to NT$0.72 per share.

You can see from slide number 6 that on a year-over-year basis with lower wafer shipments and unfavorable exchange rates our revenue decreased by 3%, while gross margin was also driven down.

Total operating expenses went up about NT$400 million, mainly due to higher R&D spending over 90-nanometer and 65-nanometer development.

As a percentage of net sales, operating expenses reached 10% higher than the 9% for Q1 '04.

Non-operating items went up around NT$300 million, mostly due to higher interest income and receipt of SMIC settlement, partly offset by higher short-term investment and valuation (ph) losses.

As opposed (ph) to our disposal at the end (ph) of our venture capital equity investment in Q1 last year, core (ph) net investment suffer a loss this quarter, as WaferTech's profit level declined due to lower utilizations, and TSMC Shanghai incurred higher expenses with heightened activity levels.

Pretax income was NT$16 billion, 26% (ph) lower than the NT$8 billion achieved in Q1 last year as a result of tax expense booked.

Tax expense booked was lower.

Our profit after-tax was down 10% than that of Q1 '04.

EPS also reduced by 10% from NT$0.80 to NT$0.72 per share.

Slide number 7 will detail a review, overview of (indiscernible) points, important items from TSMC's balance sheet and some key financial ratios.

First of all, our cash and short-term investments for the quarter was NT$111 billion, which is 7 billion lower than the previous quarter.

Accounts receivable balance also reduced from NT$27 billion to NT$22 billion, mainly reflecting lower March sales.

Accounts receivable turnover improved to 42 days.

Current liability at the end the quarter was NT$38 billion, which is NT$23 billion lower than the NT$61 billion at the end of Q3, mainly due to reduced payables to equipment vendors.

Inventorywise, our Q1 inventory came to NT$13 billion, about 1 billion lower than that of the previous quarter.

Days of inventory was 42 days compared to 41 days in Q4 last year.

The minor increase was due mainly to higher polishing (ph) of 12-inch wafer production that took place during the first-quarter '05, which carries higher value of working profits.

Current ratio for the quarter was 4.2 times.

TSMC continued to maintain a very solid balance sheet.

Slide number 8 summarizes TSMC's cash flow for the quarter.

During the first quarter, with the net profit of NT$17 billion, and depreciation and amortization of 17 billion, TSMC generated NT$36 billion cash inflow (ph).

On the investment activity front, our capital expenditure for Q1 increased to NT$38 billion.

In terms of cash investments, our short-term investments reduced by NT$1 billion in government bonds and long (ph) bonds.

Long-term investment went up by NT$4 billion, including NT$3 billion tax injection (ph) to TSMC Shanghai and further investment in corporate and government bonds of about NT$2 billion.

There was no material financial activity during the first-quarter '05.

Overall speaking (ph), our cash balance reduced by NT$5 billion for the quarter; cash and short-term investment together was down by 7 billion.

You can see on slide number 9 that our capital expenditure for the quarter was US$1.3 billion, and it grew (ph) US$1.2 billion for TSMC itself and 21 billion for TSMC Shanghai, which is representing about 84% of the total expenditure was due to build up capacity for 12-inch fab.

Let us look at some business analysis.

Slide number 10 provides a quick look at our revenue breakdown by technology.

Despite low seasonality in the quarter, TSMC continued to achieve significant growth in our domestic (inaudible) segment.

Our 90-nanometer sales also continued to grow fast (ph) .

You can see clearly from the chart that the percentage of advanced technologies has reached 45% in the first quarter, as opposed to 36% previously.

Slide number 11 provides a view by application, which did not change much from the last quarter.

Computer factory (ph) revenue contributed 34% of total revenue.

Decline rate for (indiscernible) driver (ph) and RFID (ph) segments were less the year (ph).

Communication revenue accounted for 41% of total revenues.

On the wireless front, wireless LAN (ph) and the Bluetooth insheer (ph), while telephone suffered due to low seasonality and inventory adjustments.

Wireline was also down for similar reasons.

Consumer segment took up 17% of total revenues.

We observed significant growth in digital camera and MP3 sectors, but again were outweighed by losses in the hand-held game and DVD players.

Slide number 12 shows the geographical breakdown of our revenues.

Revenue from North America, Asia-Pacific, Europe, and Japan now accounts for 79%, 80%, 6%, and 7% of our total revenue for the first quarter respectively.

Compared to last quarter, North America is (inaudible).

Slide number 13 provides a revenue breakdown by customer type.

In the first quarter, revenues generated from fabless customers accounted for 68% of total sales;

IDM 31%; while system customers accounted for 1% of our total revenues.

Now let's move to slide number 14.

This slide shows our fab utilization and ASP trends.

As expected, TSMC's utilization rate for the first quarter slipped to 78% due to low seasonality and customers' inventory adjustments.

ASP remained at the same level as a result of (indiscernible) improvements offset by price erosion.

Let's turn to slide number 15.

This slide summarizes our installed capacity by fab.

Our total capacity during the first quarter of the year (ph), 1.317 million 8-inch equivalent wafers, slightly higher than that of Q1 guidance, as our Fab 8 annual maintenance was rescheduled to second quarter.

We also estimate our Fab 12 and Fab 14 capacity to increase by 36K of 12-inch wafer in 2Q, which will bring our overall capacity for the quarter to 1.45 million 8-inch equivalent wafers.

Our capacity for year 2005 is expected to be around 5.97 million pieces of 8-inch equivalent wafers.

Lastly, slide number 16 provides a recap of the major events during the first quarter.

I will leave that for your own reference.

Based on the current business, the management's expectations for the second-quarter 2005 performance are the following.

Wafer shipments to increase by a mid to high single digits percentage point sequentially.

Overall utilization rate will be about 80%.

Gross profit margin to be in the range of 38% to 40%.

ASP to decline by a middle single digit percentage point sequentially.

This ends my presentation today.

Thank you very much.

Operator, please open the floor for questions.

(OPERATOR INSTRUCTIONS) Michael McConnell of Pacific Crest Securities.

Looking at capacity for Q2, my understanding was it was going to be up 10%.

It looks like it is going to be up somewhere around 4%, if we ex out Vanguard.

Can you talk about, was that -- did I hear this correctly?

That that was entirely due to a rescheduling of your annual fab maintenance, is that correct?

You're asking is there a capacity increase in the second quarter?

Correct.

The increase of second quarter is due to our extension of that (technical difficulty) Fab 14.

It is not due to a reschedule (indiscernible).

That reschedule only had 7/10 (ph) impact which is the shifting from first quarter to second quarter.

My understanding (technical difficulty) looking at your capacity in Q2, it was going be up 10%.

I was just curious if that is still holding; or looking at Vanguard, is that still up 10%?

That is true.

Second-quarter capacity will increase by approximately 10%.

Okay.

Looking on the pricing side, the pricing came down in Q2.

Can you talk about what is driving that?

Is it just more a mix shift to the consumer and away from some of the PCs and wireless?

Is that is what is going on there?

Well, (indiscernible).

For second quarter we have two factors so that you have pricing book guidance (ph).

One is the pricing erosion, which is (ph) greater than the normal rate due to the market situation.

The second factor is the mix.

That is the high-end technology, advanced technology ratio for the second quarter is a little behind, because of the 90-nanometer win (ph) has some minor delays because of the customer design issues.

So the combination of the two factors contribute to the mix change.

A mid to single digit change and decline (ph) of the ASPs for (ph) second quarter.

Okay.

Looking at the end-markets, can you kind of talk about briefly, a quick overview, just to what's going to be looking strong for Q2 and what is looking a little bit on the weaker side?

For Q2, we still have -- the wireless side is still somewhat weak, in particular mobile phones.

On the other hand, wireline is beginning to recover at a pretty nice clip.

Computer is kind of flattish, with some increase such as chipsets, LCD monitors.

But the rapids (ph) and the carbon-slide (ph) has a minor decline.

Consumer has a fairly significant increase from second quarter, mainly from DVD player and the set-top boxes, DV, digital TV categories.

Okay, thank you very much.

Mark FitzGerald with Banc of America.

Did I hear you correctly in terms of capacity by the end of the year is 4.9 million wafer starts on an annual basis?

No, the total capacity for this year will be 5.97 million, which is about a 25% increase from last year.

Okay.

Then you said that capacity would grow 10% in the second quarter.

Yes.

So basically, if you do the math here, that would mean you have to have a flattish low (ph) growth in capacity in the second half.

Is that correct?

That is not true.

If you are doing 1.5 million more wafer starts in the second quarter, if you just annualize that, that would put you over 6 million wafer starts for the year.

I am not sure where is the math falling here.

I am talking about 5.97 year-over-year.

So I don't think you can add percentage growth for every quarter to achieve that number.

Mark, you cannot use second quarter alone.

Basically, our first-quarter capacity did not go up.

I saw that.

Just if you just fall out where you are in terms of wafer starts today, that alone -- and just assume that is going to be the number for the next two quarters -- come over 6 million wafer starts.

So I don't -- including the first quarter.

Something is not hanging here together in terms of math; or I am doing it wrong.

Slightly wrong.

If you use second quarter, which we have just said is 1.45 million, if you use that multiplied by 3, that is your assumption, that is Q2, Q3, and Q4.

Then you add 1.32 million that is in Q1.

You do not have 5.97 million wafers for '05.

That would suggest then that capacity in the second quarter is not going to grow 10%, if it's 1.45.

Mark, I really think you can do the math yourself.

We have already said the whole year is 5.7 million; and you already know the first quarter and the second quarter.

You can calculate how much you will get in the second half.

Okay.

Ali Irani of CIBC World Markets.

I am hoping you can give us an update on your CapEx plans for the full year, and how you see those outlays.

Are we looking at a short-term pause in the second quarter and third quarter spending, for digestion?

And then a reacceleration?

Could you comment on the linearity, please?

You're asking about CapEx.

I think we have said in -- three months ago (indiscernible) it was pretty much front-end loaded.

We have guided a total CapEx for the full year would be in the range of 2.5 to 2.7; and I have just said first quarter we spent 1.3 billion, just about half of the total year's spending.

So it is not linear.

At this point are you maintaining your guidance for CapEx at 2.5 to 2.7?

Correct.

You are?

Okay.

Thank you very much.

Timothy Arcuri of Citigroup.

Back on that question about capacity, is that 5.97, does that include the non-TSMC-owned capacity?

Or is that only TSMC-owned capacity?

That includes TSMC, WaferTech, SSMC.

But it does not include Singba (ph).

Okay, great.

I guess I am asking, even though there is not going to be much capacity added during the back half of the year, that really doesn't have any sort of an implication in terms of what you're going to order, in terms of new tools.

All that is just CapEx, and that reflects what you have already ordered.

So if you kind of look more philosophically as to what your CapEx plans are, haven't you said before that you're going to spend more linearly through the cycle?

So you're going to ramp CapEx way up and way down.

Is that still the plan?

(indiscernible) asking a long-term question with a short-term question.

In this case (ph) we are not going to (indiscernible) down.

We are talking about a year-on-year basis.

You're talking about a quarter; actually it totally depends on the work we did at the end (ph). (multiple speakers)

Of course.

I am asking more kind of a --

I'm sorry?

I am just asking kind of more of a multiyear question.

So if you look out into kind of 2006 and 2007, do you expect CapEx to be spent in a more linear fashion?

Year-over-year, our thinking is fairly (ph) -- we do not expect to have a big fluctuation year-over-year.

I think you can say that.

Okay, great.

Thank you.

I guess the second question I have is when do you think that you will have 65-nanometer capacity online?

When do you think that you will start the first wafers for 65?

For 65, we expect to have some wafer output early -- end of the year, early next year time frame.

That will be a pretty small volume of prototyping, engineering.

But production volume probably will not be up until mid (ph) 2006.

Great, thank you.

Mehdi Hosseini of Friedman, Billings, Ramsey.

I have two questions.

First I am still confused regarding your guidance on utilization rates.

Your capacity is growing by 10% in Q2, but your wafer shipment is up single digit to high digit.

But you are guiding to a utilization rate up by 2 points.

Can you help me understand or reconcile the difference?

Then my second question has to do with your IDM-based customers.

Based on your conversations with such customers, when would you expect them to come back to you and increase their mix of outsourcing?

Thank you.

Entering (ph) our second quarter (indiscernible) 80%, about a 2% increase from first quarter, and we also add roughly 10% capacity.

Our definition of utilization on wafer output (ph); so wafer output to revenue there is some difference that has always happened in the past.

I think the (indiscernible) is not much the difference for this high (ph); so nothing unusual that we have been seeing in operations.

I can't quite --

IDMs.

Oh, IDM customers.

I think basically we have a third world base of IDM customers, and some of them are scaling back more than the others.

We do not expect basically our IDM business ratio to fluctuate much from quarter-to-quarter.

Thank you.

Shailesh Jaitly.

Firstly I just wanted to reconcile your guidance on wireless, because by and large the customers -- particularly the fabless ones -- are seeing that inventory is normalizing.

Industry as a whole is doing pretty well, looking at the results across the board.

So what is the reason for your relatively neutral guidance on the wireless segment for Q2?

Hello?

You are asking the wireless.

We assume national quarter (ph) mobile phone launch (ph) (inaudible).

We also understand the announcement from the mobile phone makers has been pretty positive.

I think there probably are several reasons, but I can only speculate.

I felt (ph) 1 billion -- actually the inventory levels at the mobile phone chips makers are still quite high from the second half last year until first quarter this year.

So they didn't -- and I think the inventory is still being burned off. but it should be down to a normal level probably this quarter.

So seriously (ph) I think the second quarter is probably the transition time for the mobile phone chipmakers, and I think we can expect growth to come back in the second half, the quarter (ph) time frame.

Okay.

This weakness that you're seeing, is it more confined towards your IDM segment?

Or is it also from your fabless customers?

I think it is pretty universal, it is pretty universal.

Okay.

Your product mix has the (indiscernible) pricing to be relatively stable.

But if you were just to compare like with like nodes, particularly 0.25 micron and 0.18 micron, what kind of pricing declines did you see in 1Q?

And what do you expect going forward?

We did mention already in the afternoon that the 0.18 micron price erosion is greater compared to the 0.25 micron, but we cannot comment on the specific numbers though.

Is it in double-digits?

I am sorry;

I cannot comment on that.

Okay, if you could just help quantify what the utilization levels are on these two nodes.

We do not quantify on each technology node utilization.

We have said second-quarter overall utilization will be 80% (indiscernible).

Okay. (multiple speakers) I think you were saying something, Rick.

I was saying we can comment on the second quarter.

If the 12 will be (indiscernible) we definitely expect the overall utilization to come back up.

(multiple speakers) significant (indiscernible).

Some color on your 90-nanometer ramp, please, as to how it is going.

Where do we expect by end of the year, what portion would be 90 nanometer?

90 nanometer will account for double-digits more than 10% starting from third quarter.

What kind of volumes are you seeing right now?

First quarter was 4% of total revenue.

Okay, thank you.

Dan Heyler of Merrill Lynch.

I had a question on the stock option expensing which will be taking place from -- it hits your peers in the U.S. in the second half.

Is that going to -- I guess two parts of the question.

Is that going to affect the way that you compensate your employees?

Would you continue to use a fairly high contribution from bonuses and options?

That is the first part.

The second part of the question would relate more to the reporting of that.

So far you reconcile to GAAP on an annual basis.

I am wondering if you have any plans to do any kind pro forma quarterly numbers so we can compare numbers on a more timely basis?

Thanks.

Stock option expensing has been delayed for six months.

So you will start to expense by January 1, 2006, instead of June 2005.

This is the recent announcement. (indiscernible) we have very little, almost -- very little stock option, which is for our subsidiary in North America and (indiscernible).

Those expensing will (indiscernible).

So (indiscernible) numbers.

I guess does mean that you're going to pretty much follow the calendar for the U.S. schedule, albeit that it has been pushed out?

Would you tend to follow that from your reporting standpoint, and move to a quarterly reporting of the -- I would say U.S.

GAAP, including your bonuses?

For our U.S.

GAAP reporting we definitely will follow U.S. GAAP.

But the problem is our (indiscernible) has not announced that stock options will be expensed.

So when they are announcing, we will also follow that; and we will (indiscernible) we have -- there's no plans from the government at this moment.

So you will stick to your annual reconciliation to U.S.

GAAP, then?

Your annual versus quarterly?

Yes, on an annual basis we would have to do U.S.

GAAP anyway.

So you'll see our annual number will show -- when we account there for our GAAP to U.S.

GAAP, you will see the stock option expense.

Right.

Okay.

The final question would be in relation to your target for 90 nanometer.

It sounds as though, unless the growth was in -- first half was wireless related and low power related.

Could you comment on what do you think the key drivers are, not only in the second half, but going out into next year?

You talked about wireline picking up, I imagine that is one source of potential growth.

But in addition to your traditional markets such as FPGAs and graphics, are there new markets that you guys are able to enter as a result of your good progress on that front?

I think with 90 nanometer the drivers, at least for early next year, will remain those (indiscernible) that you just mentioned. (indiscernible) graphics, based on chips, FPGAs.

We also have consumer electronics applications such as the DVD players, recorders, and game consoles.

The game console graphics chip is moving to 90-nanometers (indiscernible) quite soon, quite soon.

Then we are seeing probably a few second phase (ph) customers' designs coming in the second half this year and the next half, probably early next year.

The other applications.

Great, thank you.

Donald Lu at Goldman Sachs.

Can you just give us some housekeeping guidance on operating expenses, and the tax rate for Q2 and the remainder of the year?

Okay, operating expenses compared to revenue is quite stable.

I think you can assume it's always been 9 to 10% range.

That is operating expense.

You are asking about the tax expense; we don't calculate on a quarterly basis, but I can tell you our annual (indiscernible).

For your model you can use 13% for 2005.

That is based on operating profits?

No, income before tax, but deducted investment income.

Without investment income.

Okay.

Will that change?

Will you expect that to change substantially in '06 if the government passes a minimum tax law?

We don't anticipate we are going to have a significant change.

Of course every year we will look at our tax exemption ratio.

Depending on the utilization year-over-year it may have some difference.

But we do not anticipate a big change.

Great, thanks.

Shekhar Pramanick of Moors & Cabot.

A couple of questions.

The chairman of the afternoon session said that he expects semiconductor revenues to be flat year-over-year.

Should we read this comment being TSMC, given usually it has outperformed the semiconductor industry, at least for the whole year basis, you should be a little bit better than flat?

Also your thoughts, how do you see the utilizations trend in the second half of the year?

(indiscernible) said that our view on the 2005 semiconductor is minus 1% or minus 2%, eventually.

Actually we did not change our view from six months ago.

TSMC used to outperform semiconductor by a few points; and we hope this year we still can achieve that.

It is typical.

If would be more difficult this year for the foundry industry as a whole, because inventory build-up during the second half of last year, particularly with the fabless customers.

So we can tell, from the first-quarter numbers that already recorded and the guidance we are giving now for the second quarter, that (indiscernible) quite a bit of our customers revenue is still from their inventory at hand in the first half.

Of course we do expect a fairlyl strong pickup in the second half.

And we also expect to outperform the industry.

But this is -- we are still quite cautious from that point of view.

So the comment on the utilization rate?

You are asking utilization rate for the second quarter?

No, the second half.

What is your current view?

We are not in a position to guide the second half yet, but if the industry will recover from the second half we believe the utilization will be higher than 80%.

Thank you.

Andrew Tay (ph) of Boyer Allan.

I know you don't give out as a piece of official guidance, but could you give a feel for how your revenue mix will look by the different geometries in the second quarter?

We don't provide revenue breakdown by geometries. (indiscernible) our first quarter the (indiscernible) accounts for 45%; second quarter this number may not grow.

By the second half we are going to see quite some growth coming from (inaudible).

Right.

You are saying that the mix should not look that different in the second quarter from the first?

I didn't say that.

I just said that second quarter our advanced geometry will not (ph) grow.

Will not grow.

Right.

That is 13 micron and below.

Yes.

Okay.

Earlier, when people asked, your very first question was regarding pricing.

And you said that duration has been greater than normal due to the market situation.

We have talked about seasonality, but could you expand on that qualitatively?

What is the market situation that you're referring to specifically, other than seasonality and so forth?

Actually, we kind of commented on that earlier, through the inventory digestion point of view.

The whole industry (technical difficulty) from again, based just on foundry point of view, what we are going through really is the customers' inventory digestion.

We are creating capacity utilization -- or I should say low capacity utilization during the past two quarters.

That has created the pretty severe pricing environment, in particular with the 0.18 micron and 0.25 (ph) micron.

I think as we also said, we do believe the inventory is being burned off, down to quite a lower (ph) level, and in some cases lower than the same period last year.

So we (inaudible) a normal second half.

Okay.

I think there was some questions about tying your various statistics together on utilization and capacity.

Clearly the capacity that you quote is not the capacity -- the denominator that you use for calculating your utilization rate off the wafer ship.

Could you just tell me what adjustment you make to that number to arrive at the 78% on the 111 3000 wafers?

That is the wafer output divided by capacity.

Wafer output divided by capacity?

That get you the utilization rate, okay?

But the wafer shipment, you know, output -- some product has to go to beta (ph) testing.

Right.

So some come back (ph) between the two.

So the denominator is the one that you quote; it's just that the wafers output is different from the wafers shipped.

That is the difference.

Yes.

Okay.

Can I just leave with one question on this point?

I hate to sort of labor this point, but am I right and should I infer from what you're saying regarding your mix of business and the pricing environment that, because your mix shift is not necessarily improving, in the second quarter you are just feeling the full force of like for like price declines by various geometries?

But then that situation will reverse in Q3 as you have an improvement in the mix, so that the ASP decline that you are seeing in Q2 is something that is temporary.

Yes, that's right.

I think both looking forward for Q3 the price erosion will moderate, and we expect to go back to the normal rate.

In addition, with the product mix improvement, we do expect to have firmed up ASPs for the year, for the coming quarter.

That's for (ph) quarters.

Thank you.

Thanks very much.

Sunil Gupta of Morgan Stanley.

I have a housekeeping question on your revenue mix for Q1.

Could you talk about what percentage of revenue was 90 nanometers?

4%.

4%, okay.

Second, if I could ask about operating expenses, which earlier you mentioned it would be 9 to 10% of revenue for Q2.

Could you help us break that into R&D and SG&A?

Around 50 to 60% R&D; and 40% G&A, roughly speaking.

All right.

Finally, in terms of the depreciation number, I noticed two different numbers in your reports today.

One is $16.7 billion, which is what I think you quoted in the text.

But if I look at the cash-flow statement, it is a much higher number in depreciation and amortization.

I was just wondering what is the difference here and what am I missing?

The cash flow was including depreciation and amortization both.

The other one is just depreciation?

Yes.

All right, that's it.

Thank you.

Pranab Kumar Sarmah of the Daiwa Institute of Research.

Rick, I was just following up your one comment.

You mentioned like some of the 90-nanometer shipment has been delayed from second quarter.

Could you give us some color in what area those delays were?

Well, basically 90 nanometer production by itself is moving very well, very smoothly.

Our yield and the productivity have been actually better than we expected.

The shipments of these (indiscernible) is probably the best performance we have had during the past several generations.

For the second quarter, we do have a couple of customers' designs running into some delay.

But I cannot really comment on the application (multiple speakers) as to which.

So it is a couple of customers, not one customer?

Let me see; a couple, yes.

Fair enough.

Again, the third quarter we should see a fairly major pickup in the third quarter.

One more on debt.

Today I think your Chairman has given a very nice idea about the group A and group B type of customers and their inventory ideas.

Could you give us some idea like how many customers you have in the group A that is 60% of your revenue; and how many customers are there in the group B that is 20% of your revenue?

We do not normally comment on the number of customers representing how many percentage of our revenue.

Not a lot.

Okay.

In first quarter, could you give us some idea like how many of your business was from non-wafer-related business?

That is from MOSFET (indiscernible) and other products.

And what is the profitability of that business?

It is less than 10% from non-wafer business.

Profitability of that business was more than -- was better than your group average, or --?

We cannot talk about that.

We only talk about combined margins.

Okay, thank you very much.

Definitely less than 10%.

Fayad Abbasi of Prudential Equity Group.

I had just a question on the gross margin.

Looking at the gross margin guidance you're giving for Q2, it is about flat, plus or minus 1%.

We have a slight increase in the utilization rates, so the wafer output increase given the capacity add is pretty strong.

But I am just wondering if you could comment about in what scenario would you have f the gross margins actually decline from the 39 to about a 38%?

What would happen to get to the lower end?

Would it just be an ASP driven phenomenon?

Or maybe you could give a little color on that.

Actually we did not say that second-quarter margin would decline.

Actually in our guidance it is 38 to 40% range; 10 was our guidance of first quarter.

But if you could maybe comment, what would drive to the higher end or the lower end of that range?

That would be helpful.

I think the two factors that have made impact to the margin rate, number one is utilization rate.

Number two for the second quarter is probably the exchange rate.

So I think with the 2 point spread, it is a pretty tight range.

All right, thank you.

Robert Maire of Needham & Company.

If you could give us a little bit more granularity as to the inventory that you see is still remaining out there, of die bank versus finished goods, and any particular concentrations where the inventory might take a little bit longer.

You said we were almost near the end.

Is it all gone by the end of the June quarter?

Or is it all gone already, or what?

Could you give us a little more granularity as to the remaining inventory issues that are out there?

From a days of inventory point of view, actually, we take data from public announcements.

We also have some models to look at, to kind of see why (ph) we believe a major customer's inventory days will be.

What I can say is most of our customers' inventory we believe are being consumed to the normal level by actually probably a March-April time frame.

Other than the wireless, the mobile phone customers, there may be still some; we are not very sure, but there may be still some.

So we are still guiding the (indiscernible) for the second quarter (indiscernible).

Other than that I think (multiple speakers) in good shape.

I am not clear on the answer.

I am still not understanding it.

You're suggesting that the inventory is gone as of the end of April, or there is still some?

I am confused.

Inventory has gone down to the normal levels.

So we are down to the normal novel already, is that what I am hearing?

That's what we are.

Sometime in the second quarter that is what we believe from our models.

So we're not down (multiple speakers) --

Other than the wireless segment.

So everything but the wireless segment will be down to normal levels somewhere in the middle of the second quarter; is that correct?

Yes, that is what we are assuming.

What we believe.

Okay.

One other question is a follow-up on a previous question.

An earlier question in terms of capacity and CapEx and the number of potential wafer starts you have available for capacity.

Is that a net number or a gross number?

In other words, is some capacity being taken off-line that would impact your total capacity?

Or maybe you could clarify that a little bit.

We do not take off any capacity off-line.

Actually we are actually increasing our capacity second quarter.

So all the older 8-inch fabs will remain at current capacity levels.

You don't anticipate --?

Yes, you're right.

Okay.

Thank you.

Michael McConnell of Pacific Crest Securities.

Just a quick follow-up.

I am on the net non-operating income line in the investment income.

What is that (indiscernible) going to look like for Q2?

Can you speak a little louder?

Sure, sorry.

The net non-operating income line and the investment income line for Q2, what is your estimate for that for Q2?

Second quarter, I can tell you net income is still negative because the information (ph) is not to (ph) high levels.

Non-operating income won't be too much different from first quarter.

So on the 5.97 that you had guided to for capacity for 2005, that is exclusive of Vanguard?

Is that correct?

Correct.

So what was Vanguard's capacity in Q1 and Q2?

Can you comment on that, please?

Roughly speaking, Vanguard's capacity for TSMC was roughly 30K per month.

30K per month?

Thank you very much.

John Pitzer of Credit Suisse First Boston.

First when you look into Q2, supply is growing faster than shipments, but utilization rates are creeping up a little bit.

Does that imply that inventory is being rebuilt at some of your customers?

Then I have a follow-up.

Actually, it (inaudible) not.

(indiscernible) Our utilization is wafer out (ph); and there is some inventory we are holding back in.

So in second quarter the two numbers, (indiscernible) number and the (indiscernible) actually.

The only thing I can see right now is there are maybe some inventories in the back-end; which is kind of normal for (indiscernible).

Secondly, guys, if you look at your second-half capacity growth or supply growth, you are sort of implying with the numbers you have given out maybe about 15, 16% half on half supply growth.

If you go back and look at the 10-year, your output typically grows by about 14 or 15% half on half; which would implied a very atypical, sort of shallow utilization rate recovery.

Typically utilization rates tend to recover a lot more quickly.

If not quantitatively maybe qualitatively, could you talk a little bit about the rate of the recovery in utilizations you see as you move throughout the year?

Basically, I think what you're seeing and what we are describing is that 2005 is a flattish year overall for the industry and for the foundry as well.

So we are seeing a fairly gradual recovery of the utilization from quarter-to-quarter.

This is not a very sharp recovery.

It is not a very strong growth year for the semiconductor industry.

Great, thanks guys.

Mehdi Hosseini of Friedman, Billings, Ramsey.

Just a quick follow-up.

If you could help me understand the mix utilization rate between your leading edge 130 nanometer and below, and sort of a trailing edge 250, 180 nanometer?

For advanced technology because we are building capacity to meet the demand, so the utilization rates for those advanced technologies is higher.

For those material (ph) technology surface (ph), capacity is already there.

So we think you only count on the increase in demand since (indiscernible).

What I can tell you is that advanced technology (indiscernible) material technology.

Would it be fair to say that the differences could be as much as 15 points?

I cannot give you a quantitative indication for that.

I am just trying to understand.

You are talking about a gradual recovery in utilization rate.

Is that basically hinging on how fast you are able to fill the underutilized capacity, the trailing edge?

Yes. (indiscernible) you are correct in stating that.

To that extent, I am sorry, to that extent with a kind of typical seasonally driven year, that trailing edge capacity pretty much comes into maybe pricing pressure.

Because your competitors are also dealing with the same situation.

So ultimately maybe a pricing discount would stimulate demand and we may see ASP pressure throughout the year.

We are trying to understand your question actually.

You're talking about the pricing pressure? (indiscernible)

I am trying to understand the dynamics driving the gradual recovery in utilization rate.

I agree that the high-end, (indiscernible) is there.

But the extent of the recovery traditionally pretty much hinges on how quickly you could fill the trailing edge capacity, where most of the excess capacity is sitting on.

In an environment where most of your competitors are dealing with the same situation, I see that gradual recovery in utilization rate pretty much driving driven by pricing discounts, unless the overall economy was to pick up at a faster pace.

As we said earlier, we do expect the pricing erosion rate to moderate in the second half, to go back to the normal level in the second half.

We do not plan to use the effect of pricing theme (ph) to -- for our mature (ph) technology side.

I think we will -- the utilization will grow at the rate that we can have a very good financial performance in the second half.

Thank you.

Minish Squale (ph) of CREF (ph) Investments.

You said that you would be adding about 10% capacity this quarter; and it is probably reasonable to expect that all the addition will be at the leading edge process technology, that is, 130 nanometers or below.

The adding capacity is (inaudible).

If you are asking where we added capacity, it is mainly on advanced technology.

Right.

So if you are guiding for flat utilization rate, that means that you're also banking on improvement in utilization rate of your trailing edge process technology on a sequential basis?

Actually the technology (indiscernible) different.

We do see improvement on some mature technology utilization.

Okay, thank you.

Dan Heyler of Merrill Lynch.

Just a quick follow-up, guys.

First of all, when Chairman Chang as well as Lora, you guys have talked about building capacity here to meet demand in the second and third quarter.

Your 10% in the second and 7% or so in the third.

I am wondering how much confidence, or what specifically gives you confidence that those orders, particularly in the third quarter, do materialize?

A, is it a product launch at that point has to take place for a given customer?

Or is it migration from 0.18 and other products moving forward?

What exactly is giving you the confidence there?

That would be helpful.

Thanks.

Basically, the third-quarter demand forecast is all based on the already taped out products from major customers, and I would say the volumes taped out for that.

Basically some of these products are being qualified and certified during the second quarter.

We're not talking about any really new products taped out in this quarter, for the (indiscernible) quarter (indiscernible).

However, we do -- those products taped down in this quarter will of course contribute for the fourth-quarter ramp of the 90-nanometer product (ph).

Okay.

The reason why I am asking the question is in previous upturns you guys have taken a fairly cautious stance to some orders.

You have tried to scrutinize orders to try to ensure that you were not overbuilding; and in fact, turning some orders away, which has improved your utilization here on the bottom.

But I am wondering, now I don't hear you necessarily hear that kind of scrutiny.

In fact your capacity growth is pretty significant.

So I am wondering why the difference in terms of the approach here.

Actually, Dan, there must be some -- well, basically I think we are still scrutinizing very carefully of the demand forecast with our customers and our sales.

I think that is not changed.

We however did not -- I don't believe we have -- turned away orders. (multiple speakers)

We did on the other hand -- it is also TSMC's strong competitive edge, that is, our ability to generate and feel the demand of the most advanced technologies from our customers.

We have been progressive in getting those demands, the most advanced technology demands.

We have been very aggressive in getting those demands and to maintain our market share in the most advanced technologies.

We have not changed from that point of view.

Right.

Okay.

Just to make sure that I have my numbers right, I know your capacity does not include Vanguard.

I just want to confirm that your shipment guidance doesn't include any there from Vanguard either.

And that the (multiple speakers) -- ?

Yes, shipments include Vanguard.

Shipment does.

So part of the reason for the difference in this utilization number versus shipments over capacity, part of that shipment is not only the back end but it also to some extent has to do with the Vanguard shipments?

The delta of Vanguard.

And the delta Vanguard (ph).

Okay.

(inaudible).

Right.

Could you also -- what exactly is the Vanguard shipment number then, for the first quarter and second quarter?

We don't differentiate that.

I can tell you Vanguard capacity is approximately 30K per month.

All right.

And that is relatively stable, I imagine, first and second?

Yes.

Okay, thank you.

Operator, due to the time difference, we will only allow for two more questions.

Andrew Tay of Boyer Allan.

In your statement you were referring to the impact of the currency movements.

Could you just expand on your hedging strategies?

Because most of our revenue was denominated in U.S. dollars, and we're taking a very conservative hedging strategy.

That is to take (ph) the whole balance sheet.

But long (indiscernible) we have foreign currency (indiscernible) we hedge the position.

Right.

You generated some foreign currency exchange losses in your non-operating income.

Yes.

That happens sometimes and (ph) the currency movement is more rigorous in a month; the hedging costs tend to be higher.

On the other side, we use -- we kind of consider that the cost of doing business.

But because we have (indiscernible) so we can avoid big fluctuations in the P&L if we (multiple speakers).

So are you saying -- I may be misunderstanding.

Do you hedge the effect on the revenue front at all?

We don't hedge revenue.

We (multiple speakers).

It is just elsewhere on the balance sheet, yes.

Okay.

I know from the conversation we seem to be having or that is evolving is one of perhaps not fantastic pricing and a sort of relatively mild recovery.

I just wanted to revisit if I might very quickly your cost position.

Because you've made a point in previous presentations about the advantages of the runoff in depreciation in your older fabs.

Would you be able to just quantify the size of that and when that is likely to be manifest, if not already?

Sorry.

I don't have it percentagewise, but I can tell you our Feb 3, Fab 5, and Fab 7 have almost fully depreciated.

The only remaining fab that is not fully depreciated are 6 and 8. (inaudible).

Of course (indiscernible) under depreciation.

The 3, 5, and 7, were they fully depreciated?

Were they more or less fully depreciated during this year already?

(indiscernible) 2 and 3 have been fully depreciated; but Fab 5, the majority has been depreciated.

And part of that 8 (ph) is --

Will 6 and 8 get down to near zero depreciation sometime this year?

This year and next year and the year before last year, are three years now we are enjoying most for most of the depreciation fabs.

Sorry.

This year, next year, and the year after?

The year before.

Okay, so you have already started benefiting from that.

Okay, thank you.

Gavin Duffy of A.G.

Edwards and Sons.

Just have two quick questions.

What are the average leadtimes you're quoting customers now?

And (indiscernible) I have a second question.

Will you repeat your question?

We did not hear you clearly.

I am sorry.

The current, the average lead times that you're quoting customers now, have something -- if you were quoting a customer say 12 weeks in Q3 last year, what would kind of be the lead times now?

If you use -- because of course we have different lead times with different technologies; but if you use 12 weeks as the reference for third quarter last year, I would say we are probably -- let me think, probably 20% (indiscernible).

Okay, fair enough.

15% is kind of the ballpark number.

Good.

The second question is, when do you expect to have, I guess, Fab 12 B up running at kind of production levels?

They are now Phase II, recall.

They are now in production for actually -- (indiscernible) our 90-nanometer production is now being run in the Fab 12 Phase II.

They are running.

All right, thank you very much.

Okay.

Ms. Ho, there are no more questions at this time.

Thank you very much for your participation, and we're looking forward to see you next quarter.

Bye-bye.

Before we conclude TSMC's 1Q '05 results webcast conference call today, please be advised that the replay of the conference call will only be accessible through TSMC's website at www.TSMC.com.

Thank you all.