台積電 ADR (TSM) 2010 Q3 法說會逐字稿

Welcome to TSMC's third quarter 2010 results webcast conference call.

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

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

I would now like to turn the conference over to Dr.

Elizabeth Sun, TSMC's Head of Investor Relations.

Thank you.

Good morning, good afternoon and good evening, everyone.

Welcome to TSMC's third quarter 2010 conference call.

Joining us on the call are Dr.

Morris Chang, our Chairman and Chief Executive Officer, and Miss Lora Ho, our Senior Vice President and Chief Financial Officer.

The format for today's conference call will be as follows.

First, Lora will summarize our operations in the third quarter and give you our guidance for the fourth quarter.

Afterwards, TSMC's Chairman, Dr.

Chang will provide his general remarks on the business outlook and a couple key messages.

Then we will open the floor to questions.

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 results presentation.

I would like to remind all listeners that the following discussions may contain forward-looking statements that are subject to significant risks and uncertainties which could cause actual results to 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 April 15, 2010 and such other documents that 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.

And now I would like to turn the call over to Lora.

Thank you, Elizabeth.

Good morning and good evening to everyone.

Welcome to our third quarter earnings conference call.

To start with, I will walk you through the financial highlights in the third quarter.

Then I will give you the guidance for our fourth quarter.

You may also refer to the quarterly financial summary slides on our website.

May we remind you that all dollar figures are in NT dollars unless otherwise stated.

TMSC's net sales reached TWD112.3b, up 6.9% from the second quarter and up 24.8% from the same period a year ago.

Wafer shipments were 3.19m 8-inch equivalent wafers, up 9% from the prior quarter and up 30.5% from the year ago quarter.

Gross margin was 50%, representing 0.5 percentage point increase from the second quarter and a 2.3 percentage point increase from 3Q '09.

Operating margin was 38.4%, down 0.2 percentage points sequentially but up 2.8 percentage points compared with the year ago quarter.

EPS for the third quarter reached TWD1.81.

ROE was 36.5%.

Let's now take a closer look at the income statement.

Third quarter gross margin was 50%, up by 0.5 percentage points from 49.5% in the second quarter, mainly due to continued cost improvement.

Operating expense increased TWD1.6b from the second quarter, primarily due to a higher level of development activity for our 28-nanometer and 20-nanometer technologies and also higher opening expenses for our Fab 12 Phase 5.

Non-operating income increased by TWD4.3b from the second quarter, mainly reflecting the receipt of SMIC shares from the litigation settlement.

Net investment gains was TWD900m, up TWD370m from the prior quarter, mostly due to business improvements among certain invested companies.

Net margin was 41.8%, up 3.4 percentage points sequentially and up 7.8 percentage points year over year.

Now let's take a look at our revenue by application.

In the third quarter demand from communications and industrial-related applications continued to be strong as they grew 13% and 19% quarter over quarter respectively.

Consumer increased by 4% from last quarter while computer declined 7% sequentially due to weaker than seasonal demand.

Overall, revenue from communications, computer, consumer and industrial applications accounted for 44%, 25%, 14% and 17% respectively of our wafer sales in the third quarter.

By geometry, we continue to see strong adoptions from customers in our leading edge technologies.

40-nanometer and 65-nanometer continued to grow and now represented 17% and 29% of total wafer sales respectively.

These two most leading edge technologies combined were 64% -- or 46% of our total wafer sales.

The revenue mix of 90-nanometer and 0.13 micron remained stable at 14% and 12% of wafer sales respectively.

The aggregate revenue from 0.13 micron and below accounted for 72% of total wafer sales.

Now let's move on to the balance sheet.

We ended the third quarter with TWD167b in cash and short-term investments, representing a decrease of TWD39b from the prior quarter, primarily due to the cash dividend payment of TWD77b during this quarter.

Total current liabilities decreased by TWD57b, primarily due to the payment of cash dividend, partly offset by an increase in short term loans.

Account receivable days remained at 39 days.

Inventory turnover days were 47 days.

Net fixed asset turnover was 1.3 times.

On page nine, now let's proceed to cash flow.

Cash flow generated from operating activities totaled TWD64b, representing an increase of TWD15b from the second quarter, mainly due to the increase in net income and the decreased use in non-cash net working capital.

Capital expenditure was TWD42b in the third quarter.

Also just as I mentioned we paid out TWD77b of cash dividend in the quarter.

Meanwhile we increased another TWD19.8b in short term loans to hedge a growing proportion of the Company's US dollar position.

In sum, the ending cash balance was TWD132b, down TWD40b sequentially.

Free cash flow was TWD21.8b in this quarter.

Now let's turn to CapEx and capital expenditure.

In the third quarter due to productivity improvement at our Fab 14, we accelerated our schedule and upward revised our whole year capacity plan.

Based on our current planning 2010 overall capacity is expected to increase by 14% year over year, with 12-inch capacity up 37% year over year, compared with our last quarter's announcement of 36%.

Total installed capacity was about 2.95m wafers in the third quarter, representing a 7% increase from the prior quarter.

In terms of capital expenditure, we spent $1.32b in the third quarter.

Total capital expenditures for the past three quarters were $4.42b.

Total 2010 capital expenditure will be $5.9b as we guided last time, remains unchanged.

Now let's turn to the outlook of the fourth quarter 2010.

Based on business expectation and exchange rate assumption of TWD30.60 to $1, we expect our consolidated revenue in the fourth quarter to come in between TWD107b and TWD109b.

In terms of margins, we expect our fourth quarter gross margin to be between 48% and 50%, operating margin to be between 35.5% and 37.5%.

This concludes my remarks today.

Now I will turn the call over to Dr.

Morris Chang, our Chairman and CEO for his remarks.

Hi, everybody.

This is Morris Chang.

I will make a few comments first on third quarter, fourth quarter and the whole year this year.

The third quarter was a historical record in revenue dollars, in gross margin dollars and in net income dollars.

It was also a near historical record in gross margin percentage and operating income percentage.

If the fourth quarter were at the same exchange rate as the third quarter, then the fourth quarter guidance would be revenue TWD111.6b to TWD113.7b, which is higher than the third quarter actual.

GM, gross margin percentage would be 49.5% to 51.5%, which is also higher than the third quarter actual.

Operating profit percentage would be 37.3% to 39.3%, which would be flat from the third quarter.

All those numbers I've given you are if the fourth quarter foreign exchange rates were the same as the third quarter.

So you can see that the business really is continuing at a brisk level in the fourth quarter.

There's still considerable demand that we cannot schedule production for.

Now for the whole year, this year, in spite of the exchange rate variation in the fourth quarter, the year 2010 will be a historical record for TSMC in every important operating measure.

We also believe that we have increased our addressable market and increased our share in that addressable market.

Our growth momentum has been fueled by the following factors.

One, we have increased 40, 45-nanometer revenue dollars more than five times in 2010 over 2009.

40, 45-nanometers in the fourth quarter accounts for 20% of our total revenue, with a gross margin at a level above corporate average.

Two, 44% growth in revenue dollars in 65-nanometers over 2009.

Three, and this is perhaps new to a lot of people.

We have had a 52% growth in specialty technology revenue over 2009.

Specialty technologies include embedded Flash, CMOS images and high voltage.

All those are in the mainstream technology nodes.

I'd like to make a few comments on supply chain inventory and 2011 outlook.

Supply chain inventory has increased in the third quarter in both dollars and VOI and at the end of third quarter is estimated to be close to the normal seasonal level.

At the end of fourth quarter, supply chain inventory is forecast to be at about the normal seasonal level.

As for 2011, we are forecasting that the global semiconductor market will have a 5% growth in 2011.

We are also forecasting that the total foundry growth, foundry revenue growth will be 14% in 2011, greater than the semiconductor market growth.

As to TSMC, we believe that our momentum will continue into 2011 because we have the right technologies, we have available capacity and we have good customer relationships.

In summary, we expect to grow more strongly than the total foundry growth, which as I said is 14%.

Now I'd like to say a few words about our mission and strategy.

Our mission is simply to be the technology and capacity provider of the global logic IC industry for years to come.

We want to be the technology and capacity provider of the global logic IC industry for years to come.

Our strategy.

First, we want to extend and strengthen our technology leadership.

Our R&D expenses were $657m in '09, are estimated to be $945m in 2010, and will increase to $1.1b in 2011.

All those are in US dollars.

Our R&D capital expenditure is at $400m this year and will increase to $700m next year.

Our R&D major projects will include 28-nanometer, 20-nanometer, 14-nanometer and exploratory work on 10-nanometer and 7-nanometer.

The major projects will also include specialty technologies, some of which I have already mentioned, and back-end.

The second main element of our strategy is capacity leadership.

Capital expenditures will be $5.9b this year as we guided a quarter ago and capital expenditures will be greater next year.

We spend capital on the basis of forecast market demand from our customers and expected a return on investment, not on speculation.

Future generations will be increasingly more capital intensive.

28-nanometer capacity is already twice as expensive as 65-nanometer capacity.

28 is two times more expensive than 65.

As to dividend, we already have one of the highest dividend yields in the technology sector.

We will maintain TWD3 per share cash dividend payout in 2011.

We will try to reward shareholders more in share price appreciation than in dividend payout.

Now we have already seen the first fruit of our strategy in the 2010 results.

We will continue this strategy in 2011 and beyond.

Now I'd like to say a few words about new technology.

First, on 28-nanometer, the 28-nanometer development is going well.

The gate last technology is proving out very well.

Customers' engagement is stronger than we had experienced in the 40, 45-nanometer node at the same stage.

Customer engagement is stronger now with the 28-nanometer than 40, 45-nanometer at the same stage.

Now I'd like to show a slide.

This slide is a scattergram of the tape-outs that we have scheduled over the next some five quarters.

Now they're all 28.

Each star is one tape-out and it means that it's one design that's based on our 28-nanometer technology that the customer has already scheduled a tape-out with us.

So there are altogether 71 tape-outs already scheduled and we are quite sure that more will come as time progresses.

So this is more tape-outs than we had with the 40, 45-nanometers at the same stage of development.

Now I also want to point out that at the 28-nanometer generation, customers' designs are very difficult to port between foundries.

They are very difficult to move from one foundry to another.

This is a new phenomenon which did not exist even in the 45, 40 generation.

Now I'd like to say a few words about specialty technologies.

We have achieved a fair amount of success in opening up our new adjustable markets, which were not reached by the foundry segment before.

Some examples are embedded Flash.

After many years of investments in developing this technology, we are now ramping up much faster than before.

We expect our MCU, microcontroller unit business, to grow 72% and our automotive business to grow 80% this year.

In the area of automotive, the requirement for quality and reliability is quite strict for the under-the-hood applications.

TSMC has gone from providing infotainment to under-the-hood for the automotive industry.

And only a few companies are qualified to provide under-the-hood products for the automotive companies.

Another example of special technology is CMOS image sensors.

TSMC's proprietary backside illumination technology, also called BSI, has the advantage of quality and form factor.

It has been popular among smartphones and compact digital still cameras.

TSMC is the first company to enter volume production in BSI.

A few comments on back-end technologies.

We are currently developing technologies in silicon interposer and 3D-IC.

While we're scaling CMOS from 40 nanometers to 28 nanometers and to 20 nanometers, etc., for improved speed, power and density, we find even more opportunity to improve system performance on PC board level.

As an example, interconnect density on a PC board is really more than 1,000 times lower than that of silicon.

That's why there is really a lot of room for improvement at the PC board level.

And that's what we are working on with our silicon interposer development and 3D-IC development.

Now a few comments on new businesses, which are LED and solar.

Our differentiation will be technology.

That's where we hope to differentiate ourselves in the LED and solar businesses.

We have more than 250 employees in new businesses now and they are still increasing.

They are very busy developing LED and solar technologies.

LED lab and fab is almost completed.

Solar fab construction will be completed in mid-2011, mid next year.

That's the solar fab construction.

The new businesses will not continue -- I'm sorry, will not contribute significant revenue in 2011, but will in 2012.

Those are all the comments I have at this time.

This concludes our prepared statements.

Operator, please open the floor to questions.

(Operator Instructions).

Our first question comes from the line of Mehdi Hosseini of Susquehanna International.

Please proceed.

Thanks for taking my question.

Dr.

Morris, going back to your commentary about the prospect of revenue growth for next year, if I just assume the mix helping with the ASP improvement, therefore I would estimate that unit or wafer shipment growth will be about 10%, 12%.

And if the CapEx is going to be higher than this year, wouldn't that suggest that your capacity would actually grow at a faster rate than your shipment for next year?

No.

The answer is no because I do not think that it's correct to assume that the average selling price will decline -- will increase.

I'm sorry.

We're assuming that it will increase.

And so therefore, we still plan to have rather full loading of our capacity.

As our capacity increases, we still plan to have rather full loading.

And I will not specify how full is rather full.

But it means pretty full loading every quarter next year, even as our capacity increases.

And just two follow-ups.

How would the mix of CapEx look like into next year?

And I'm not asking for a specific number.

But as LED and solar starts to become material, how would the mix of CapEx change?

And then, the second follow-up question.

As second-tier memory manufacturers, like Powerchip increase their mix of foundry services, do you think -- or what do you think the pricing pressure would be for the trailing edge?

First question, the new businesses, capital expenditure.

It will not be very significant next year.

It will be perhaps a single-digit percent of our total capital expenditure next year.

And the second question, will the second-tier foundries affect our trailing edge prices?

Well, second-tier foundries have been in existence all along, not the same ones.

But second-tier foundries have been really present all along.

And indeed, they have had an impact on our trailing edge prices.

But I don't see that next year will be any different.

I don't see that next year will be -- will have -- we will see any weight or impact from these second-tier foundries.

Great.

Thank you very much.

Thank you.

Our next question comes from the line of Donald Lu of Goulden Sachs.

Please proceed.

Good evening, Chairman, Lora and Elizabeth.

My first question is on 28-nanometer high-k metal gate tape-out.

Can you -- I remember that TSMC was planning to make announcement in September about starting the 28-nanometer high-k metal gate service to customers.

What is the progress there?

Have the tape-outs started yet for high-k metal gate?

And what is the yield, or anything you can share with us?

Donald, our 28-nanometer tape-out progressed well.

And our first devices -- first technology, which is a low power, we have completed the process and we are seeing customers adopting it.

And the rest of three technologies are on schedule.

Well actually, the scattergram -- I think I understand your question, now, Donald, when you were referring to September.

Well, I think in the scattergram I showed, there are quite a few stars in the fourth quarter this year.

So yes, the tape-outs have started.

And in fact, they are actively going on right now.

So the high-k metal gate tape-out is starting in Q4 this year.

Yes.

Yes.

Okay, great.

And when are we going to start to see a 1% revenue from that technology?

Second half of next year, 1% to 2%, second half of next year, 1% to 2% of our total revenue.

1% to 2%, great.

My second question is on the new business.

I remember, Chairman, I think maybe a year or two ago commented that the new business would be TWD2b in revenue or more by, I think it's 2012.

(Multiple speakers).

Oh, no, no, no.

I said no such thing, Donald.

No?

Oh, I'm sorry.

Is there any quantitative goal that we can factor in, that in the next five years, what kind of contribution the new business will be?

Well, as I said in my prepared comments, it would not -- new businesses will not contribute any significant revenue next year, but will in 2012.

And by that, I mean perhaps a few hundred million dollars in 2012.

A few hundred.

Great, thank you.

And my last question is about depreciation for the next year.

Given the CapEx we are planning for this year and next year, how much would that increase by?

I'll let Lora answer that.

Lora?

Donald, we have not finalized our 2011 CapEx plan.

So I can just give you a range of our current estimation that the depreciation for 2011 may increase in the range of 25% to 30% over 2010.

However, even with the increase in depreciation, we believe our structural profitability, that is the SGM that Chairman was talking about.

Standard gross margin.

Yes.

Standard gross margin will not deteriorate from this year.

What is the -- what gross margin will not change?

Standard gross margin.

That means the margin at 85% utilization.

Will be 40% or 45%?

No, will be similar to this year.

Oh, got it.

So for next year, if we have utilization at 85%, it should be similar to your last gross margin when utilization was a similar level?

Yes.

The point is it's [regardless of] increasing depreciation, we still can maintain similar level of standard gross margin as this year.

Got it.

Great.

Thank you very much.

Our next question comes from Randy Abrams of Credit Suisse.

Please proceed.

Yes, hi.

Good evening.

I want to see if you could talk -- you mentioned this afternoon about some of the specialty processes ramping up.

I want to see if you're considering, on some of your older 300-millimeter, perhaps on the -- even the new mega-fabs, whether analog is starting to make sense.

We have TI starting to pitch 300 off a depreciated fab they bought.

But I want to see if it's starting to make sense for your business.

Well, we have made a pretty careful analysis.

And we believe that the cheapest way, the most economical way, I should say, to produce specialty technologies, which are all really in the mainstream nodes, the old technology nodes, 0.11 micron and up, 0.1190 and 0.25, 0.35, 0.5 microns, so old technology.

The most economical way to produce these specialty technology things is still with used old 8-inch equipment and not with 12-inch equipment, even used 12-inch equipment.

It would not be economical.

So our plan is actually to expand our capacity facility in Shanghai, in China.

Now we have an 8-inch fab in Shanghai right now, which has the capacity of roughly, what, 40,000 wafers per month?

49,000 right now, about 50,000.

And we are planning to expand it.

And that's all 8-inch.

49,000 right now.

We are planning to expand it to 110,000, all 8-inch equipment.

And I would say that all of it will be to produce -- all the expansion will be to produce the specialty technologies, rather than doing it in a 12-inch plant.

Okay.

And the follow-up question, is the timeline then for that expansion, that's for next year, to get to 110,000?

The second question I had was a follow-up to the 28-nanometer tape-out and just your sense, at this stage.

Are most customers adopting the option to stay on the traditional polysilicon gate?

Or are you seeing a number go to high-k metal gate?

And at this stage, is there a big difference in the yield ramp up for the 28-nanometer variations?

Many customers are going to the high-k metal gate.

As for the timeline for the Shanghai fab expansion --

Randy, it will expand in two years' timeframe, not just one year.

Okay, thank you.

Our next question comes from the line of Steven Pelayo of HSBC.

Please proceed.

Okay.

I was a little surprised with the depreciation growing 25% to 30% this year.

I guess I had been modeling closer to 15%.

Could you quantify how much depreciation, on a quarterly basis, will be in cost of goods sold in the fourth quarter and the first quarter?

I guess I'm concerned that you'll see a big step-up in Q1 depreciation and cost of goods sold, and that it may actually have an impact on margins, if revenues are just flattish.

Steven, for the increase of depreciation for next year, I think it will moderately increase quarter by quarter.

I have seen an around 10% increase in first quarter versus fourth quarter this year.

Okay, excellent.

And then a question for Chairman.

A year ago -- maybe it was a year and a half ago -- you really talked about wanting to implement a new strategy at TSMC where you partner with the customers and you prove to them that you always have capacity available.

You then prevent spillover going on to second-tier foundries out there.

You then admitted that the cycle just came back too quickly and you couldn't really adopt that strategy.

It sounds like you're trying to install capacity kind of almost at the level, to remain fully loaded.

So is the strategy changed?

Do you ideally want to run TSMC at 90% utilization so there is always upside available for customers?

Or do we now think we're targeting 100%, to balance it as close to 100% as possible?

You are exactly right.

We did adopt -- we did start to adopt the strategy that I described a year ago, except that we had to catch up.

We were 100% -- more than 100% loaded last year.

And we have been more than 100% loaded this year.

And next year, we will still be 100% loaded.

But we don't really want to spend TWD10b of capital all in one year.

So yes, we are still on that strategy that I expressed a year ago.

And I think that there will be a time, maybe in 2012, when we will ready capacity.

We will have reserve capacity.

But I think that even though we are 100% loaded now and plan to be 100% loaded next year, I think that our pace of adding capacity has been faster than our second-tier competitors.

And therefore, I think, even though we don't have any reserve capacity -- we will not have any reserve capacity in the near future -- I think that certainly less will spill out to the second-tier foundries than they did.

And just a follow up on that.

A couple of years from today when you have caught up and built the reserve capacity, do you think the Company will still be able to do 50% plus gross margins if we're building in that buffer?

Well, I cannot see 50% gross margin, but I think greater than 20% ROE is an unalterable benchmark.

Excellent, thank you.

Our next question comes from Steven Goulden of Nevsky.

Please proceed.

Hello there.

Thanks for taking my question.

Congratulations on your results.

I was listening to the call this morning and you said that you're expecting going forward, obviously, foundry growth of around 14%.

TSMC could be a bit higher than that.

And you also said that having targeted 20% plus ROE from now on for the next five years, you're really looking at 10% profit before tax growth annually.

Now if we've got, say, 15% plus shipment growth - I think you mean shipments by that - and 10% PBT, then am I right in assuming that -- well, let's say that margins are relatively stable going forward, you're looking at maybe 5%, 6%, 7% blended ASP drop per year.

Is that a sort of sensible way of thinking about it?

Look, you may be right, but I don't think about these things this way.

Okay.

So I don't think about -- well, I look at average price, I don't look at average price.

I look at the price by node very, very carefully, the future price development by node, by node of technology, 65, 40 and 28 and all those stuff, and also the more mature technology.

I look at them by node.

And I want to ensure that every node will have a return on assets which then translate to return on equity which, when you add up all these return on equity, you get greater than 20%.

20%, as I said, is an unalterable minimum.

And actually that hasn't changed since day one of our Company.

But we always manage to deliver more than that, most of the years we managed to deliver more than that.

So when I say that 20% is unalterable goal, I don't mean that we're going to retreat into that level you know.

But it's always nice to set a minimum benchmark and then also set other criteria.

And the other criteria, the new criteria that we have set now is the earnings per year growth, earnings per share, profit before tax per year growth.

That, as you said, is the 10% now.

And so what I'm saying is that I think we'll achieve 10% per year in the next five years, 10% per year profit before tax growth.

And I think that -- I also maintain the 20% minimum ROE, but certainly my expectation is that we'll probably exceed that.

Excellent, thanks.

And also you talked about how the 28-nanometer porting technology between different foundries is a lot tougher than previous generations.

Are you able to develop -- are you able to talk a little bit more on that at all, or is that perhaps getting a little bit too technical?

Well, it doesn't have to be very technical.

All it means, well, first of all, each technology is going to be sufficiently different so that when a customer designs to one technology he will really have to do a pretty extensive redesign if he wants to use another company's technology.

And this, of course, in the past I think this was somewhat so in 40, 45.

I think that it was pretty hard for the customers to second source, let's say, from another foundry.

It was already in the 40, 45 -- it is already in the 40, 45 generation.

But in the 28 generation it's almost, I think, in order of magnitude, harder than in the 40, 45 generation.

Great.

Okay, thank you very much.

Yes, thank you.

Our next question comes from the line of Pranab Sarmah of Daiwa.

Please proceed.

Hi.

Good afternoon.

Could you give me a little bit of color, the automotive type of products, how much percentage of your revenue it is contributing now for 3Q?

How many percent of revenue?

Contributing from what, Pranab?

Automotive applications, how many -- how much percentage of the revenue is coming from automotive-related product?

Pranab, it's single digit, but this volume is continuing to grow in the coming periods.

Could you expect to reach about 3% to 5% revenue somewhere middle of next year?

It's already 3%.

So it's going to be bigger than that next year.

Okay.

That's great, thank you.

And next one is we have seen like a lot of smartphone related products are coming up, and for second-tier chip makers they might need some sort of a memory integration capability.

How you're going to address that issue.

Hello?

We're still here.

We're just [cogitating] -- first I wanted your question to be repeated because I didn't understand -- I didn't hear it, I should say.

There's some noise on the line, I wonder what that is.

I'm sorry.

Go ahead.

Actually our approach to this memory plus CPU arrangement is the silicon interposer and the 3D-IC.

That's our approach.

So that's one of the new technologies, both those in the same -- in the back-end category.

And we are working on that.

Would you like to include some third party out there or you will be doing everything by yourselves?

Technology is going to be developed by ourselves.

Okay, got it.

Okay, thank you very much.

Thank you.

Our next question comes from the line of Guenther Hollfelder of UniCredit.

Please proceed.

Hi.

Thanks for taking my question.

I had one related to your new business areas, to LED.

There are right now memory players like Samsung already in the market and, of course, also the established players.

Do you have a different approach here, for example, regarding manufacturing for the product to be successful in the LED market here?

Thanks.

Well, we are still in the process -- as I said, our differentiation will be technology, in both LED and in solar.

So we're still at this point busily trying to develop the technology.

And you know very well that for solar there are different approaches, and we have actually chosen what we want to do and we are developing that.

It's a thin film thing and it's a SiC thing.

And for LED there are various technologies, not just in the emitter but also in various types of packaging, etc.

And so we're developing those, and also epitaxy and so on.

So we're developing those.

But the goal, or our strategy I should say, is to have a stronger technology which means higher performance perhaps or lower cost or both and that's what we're shooting for.

And there's -- we don't have a final product yet, therefore we're not going to have any significant revenue even next year.

But our plan is to have a significant revenue in 2012.

So just a follow up then.

I remember, I think, your press release regarding LED, you were, earlier this year you were talking about maybe adding a second fab or a second module here already for LEDs.

Is there already a decision's been made here in this regard?

I do not believe that we have a press release about second module.

We did have a press release of the first module which is near complete, as Chairman has said.

Okay.

Thank you very much.

Thank you.

Our next question comes from the line of Satya Kumar of Credit Suisse.

Please proceed.

Yes hi.

Thanks.

I was wondering if you could quantify your spending, not the dollar level but the percentage of spending by technology node for this year and next year.

And if you could also provide that for 200-millimeter and 300-millimeter?

Thanks.

Well, let me just answer this.

Almost I think -- as far as this year is concerned I think about 80%, 85% is on 65 and 40, 45.

Now next year we'll be adding 28.

So next year whatever the amount will be, I think it will be 85% on 28, 45 and 65.

Now you also have to take into account that the R&D capital, which is included in the total capital numbers, the $5.9b number for this year, and that's TWD400m this year and will be TWD700m next year.

So that additional percentage, it's included in the total, but the percentage that I just quoted to you were capacity percentages.

That does not include the R&D capital.

Okay, okay.

Did I make myself clear?

Yes.

I just wanted to -- it sounds like you still spend, excluding the R&D portion, 80%, 85% on the 28, 65 nanometer next year, consistent with 80% to 85% on 65 and 40 this year.

I think that's correct, yes.

Okay.

But you also mentioned that you're planning to increase the capacity in China for the 200-millimeter.

That sounds like more trailing edge for next year.

So is that not represented in the increase in percentage?

No, that's in the other 15%.

Okay, understood.

Earlier on I think you mentioned that the foundry industry will grow 14%, that's faster than the 5% for the semiconductor industry.

I understand that you're planning to grow faster with share gains.

Could you quantify that increased growth for foundry, how that's splitting up between growth -- faster growth for your fabless customers versus increased level of IDM outsourcing?

Well, as far as our percentage revenue from each category is concerned, I think that the percentage hasn't changed much over the years.

We have always derived around 75% from fabless customers.

Now the fabless portion, the fabless industry has grown faster than the total, than the semiconductor industry total.

And actually the delineation between fabless and IDM has started to blur a few years ago.

Most companies, most companies, former IDMs, have declared themselves to be fab-light, which means that they're really not building any capacity for the advanced technology, for the leading edge technology any more.

So, it's all kind of blurry now to talk about fabless and IDM.

But the key point however is that we are enlarging the addressable market.

The enlargement comes from the faster expansion of the fabless and also comes from the increased IDM outsourcing to us.

So our growth comes both from us enlarging the served or addressable, available market and from enlarging our share in that market.

Got it, thank you.

So larger share in an ever expanding market, that's what we're talking about.

That's helpful.

Thank you.

All right.

Operator, in the interest of time we will only accommodate two more callers' questions.

Our next question comes from the line of Mehdi Hosseini of Susquehanna International.

Please proceed.

Thanks.

One follow-up question, going back to your customers' inventory correction and Q4 back to the normal level, does that suggest that the shutdown days during the Christmas time and the Chinese holiday period is going to be typical to a normal seasonal, or should we assume differently?

It will pretty much on normal seasonal.

So in other words the impact to utilization rate should not be much.

As Chairman said, we expect to be fully loaded in the upcoming quarters.

Thank you.

Our next question comes from [Hessan Ganaria] of RBC Capital Markets.

Please proceed.

Yes, thank you.

I have a couple of questions regarding your strategy for the trailing edge.

Number one, if you can provide some more color in terms of your analysis, considering that TI thinks that 300 millimeter is better for trailing edge.

Is it -- could the differential be because of the location in US versus China.

And the second thing, I was under the impression that you're building a 300-millimeter trailing edge fab, Fab 15 in Taiwan.

Is that plan changed for expanding the China fab?

Let's see.

The question is -- yes, we're not planning to use our new 12-inch fab, which is the fab in Taichung fab we call it Fab 15, we're not planning to use it for the trailing edge specialty technologies products -- wafers.

As I said earlier, we did a very careful analysis of this and we found that most economical way to produce specialty technologies -- and of course special technologies will eventually graduate to 12-inch technology too, 12-inch advanced technology too.

But as of now the most economical way to produce specialty technologies is still 8-inch, old used 8-inch equipment which is available on the market.

And we've already got a fair amount of it and we will continue to get more of it.

That old -- all the old equipment, we're planning to expand the Shanghai fab that I just mentioned with old equipment we will be purchasing outside.

Okay.

Thank you very much.

This concludes our Q&A session.

Thank you very much for your participation.

We look forward to talk to you next quarter.

Thank you, thank you.

Bye.

Good night.

Before we conclude TSMC's third quarter 2010 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.

You may now disconnect.

Have a great day.