聯華電子 (UMC) 2011 Q3 法說會逐字稿

Welcome everyone to UMC's 2011 Q3 Earnings Conference Call.

All lines have been placed on mute to prevent background noise.

After the presentation there will be a question-and-answer session.

(Operator Instructions)

For your information this conference call is now being broadcasted live over the Internet.

A webcast replay will be available within an hour after the conference is finished.

Please visit our website, www.umc.com, under the Investor Relations Investor Events section.

I would like to introduce Mr.

Bowen Huang, Head of Investor Relations at UMC.

Mr.

Huang, you may begin.

Thank you and welcome to UMC's conference call for the third quarter of 2011.

With me today is the CEO of UMC, Dr.

Shih-Wei Sun; and the CFO, Mr.

Chi-Tung Liu.

During this conference, we may make forward-looking statements based on management's current expectations and beliefs.

These forward-looking statements are subject to a number of risks and uncertainties that could cause actual results to differ materially, including risks that may be beyond the Company's control.

For these risks, please refer to UMC's filings with the SEC in the US and the ROC Securities authorities.

I'd now like to introduce UMC's CFO, Mr.

Chi-Tung Liu to explain UMC's Q3 2011 business results.

Thank you, Bowen.

For the third quarter of 2011 revenue was TWD25.19 billion, a 10.5% quarter-over-quarter decrease from TWD28.15 billion in second quarter 2011, and a TWD22.9 year-over-year decrease from TWD32.65 billion in 3Q 2010.

Gross margin was 19.8%.

Operating margin was 6.1%.

Net income was TWD1.95 billion and earnings per ordinary share were TWD0.16.

Above is the short summary of UMC results from Q3 2011.

More details are available in the quarterly report which has been posted on our website.

I will now turn the call over the Dr.

Sun.

Thanks, Chi-Do.

Good morning, good afternoon and good evening ladies and gentlemen.

In Q3 2011 revenue was in line with UMC's guidance.

We shipped 1.03 million eight-inch equivalent wafers with ASP remaining flat.

Overall utilization rate was 74% with 40% revenue coming from 65 nanometer and below technologies.

We maintain our previous semiconductor market view as we enter the fourth quarter of 2011 with end result European and the US sovereign debt, China inflation and the limited inventory distribution and digestion visibility through the supply chain all contributing to industry uncertainty.

Accordingly semiconductor market demand continues to be weak as reflected by our customers' conservative order patterns.

During this time UMC will continue with its cautious approach.

We do anticipate UMC's rate of revenue decline to ease in Q4 and will remain operating profitable as our efforts to strengthen the Company's operating efficiency and the cost structure bear fruit.

Despite the dynamic nature of the semiconductor industry we believe that customer-driven development of advanced processes and IP platforms will form the foundation of UMC's next growth phase.

In addition to the volume production 49 nanometer technologies we have also invested heavily to develop our 28 nanometer technologies and IP platforms, including the high-k metal gate 28 HPM and the polyscion 28 HLP processes.

UMC's 28 HPM process is developed with the mainstream (inaudible) approach and is suitable for high performance mobile devices and high-speed networking products.

The 28 HLP platform delivers industry-leading performance to cost ratio through the adoption of the traditional polyscion gate stack and the proprietary performance enhancement process features.

These 28 nanometer technologies are supported with robust IP platforms developed through cross collaboration with ARM, Synopsys and our customers to create, optimize the 28 nanometer integrated solutions.

Customer 28 HLP product has entered the pilot production with 28 HPM scheduled for pilot production in mid 2012.

We are optimistic about our 28 HLP and 28 HPM since they form the duel engine that will propel UMC's advanced process growth, strengthen our future competitiveness and enhance our portfolio of the comprehensive foundry solutions available to UMC customers.

Now let me provide you with the guidance for the fourth quarter of 2011.

Wafer shipment will decrease approximately 10%.

Wafer ASP NT$ will increase approximately 5%.

Operating margin will be in the low single digit percentage range.

Capacity utilization will be in the mid to high 60% range.

The communications segment will outpace consumer and the computer segments.

UMC 2011 CapEx will remain at $1.8 billion.

That concludes my comments.

We are now ready for questions.

Operator, please open the lines up.

Thanks.

(Operator Instructions).

You have a question from the line of Randy Abrams of Credit Suisse.

Yes.

Good evening.

The first question I had was on the IDM revenue where it dropped from I guess a pretty big drop this quarter.

Was that in-sourcing or using some adjustments taking place for reclassification of customers just to drive that big change this quarter, and if now I guess looking forward you're seeing any stabilization from the IDMs?

So if we can separate IDMs outsourcing into the advanced leading technologies that part is intact.

There is not much of a change, but for certain IDM customers during this economic slowdown they pulled some of the legacy technologies back to their own factory.

Okay.

And I guess a question on the 28 nanometer you mentioned a bit of color in your prepared remarks about the process for polysilicon and the high-k metal gates.

What are you seeing in terms of customer interest for the two different processes, and if you were to think of timing of the inflection where we'll start to see meaningful volume ramp up for both of those processes?

So for the 28 HPM, as I mentioned in the statement it's targeted for very high performance applications.

And for 28 polyscion, as I mentioned as well we actually turbocharged the process into a better performance option and it can meet quite a bit of application so I'm quite optimistic about that as well.

So but for use as foundry service provider we are offering both processes.

And in the early stage as in the polyscion we will have a more volume user.

And as far as volume ramp, as I mentioned, we are -- today we are doing pilot and pilot production, pilot verification for the real products so it may take nine to, nine months to a year before starting ramp, which leads to the second half of next year.

Okay.

And in this environment with utilization coming down are you starting to see more price discounting crop in?

And normally I think beginning of the year there's a bit of a price drop, but what should we expect looking forward from the pricing environment?

The pricing environment is always a challenge and is a customer by customer negotiation.

So for today's environment all the customers are very careful about the demand inventory situation.

So in terms of the pricing elasticity versus volume we see a certain level of that, but not too much.

Today if they don't want it it doesn't matter how much you lower the price.

Yes.

The elasticity is weak.

Okay.

And last question I had was on the 65 nanometer where there's a good shift in that direction.

What are the applications you're seeing driving that strength and is that something you expect that to continue to grow as a percent of sales?

Or would you expect it to start shifting toward 40 nanometer?

For near today's loading is not that great and so last quarter only 74%.

Randy Abrams.

Yes.

So over speaking the 65 percentage growth mainly from the communications sector, mostly from the 3G smartphone.

Okay.

And will that continue to grow as a percent of sales or do you expect 40 to start seeing a faster ramp up now?

Our visibility is not that great, is kind of shallow, but next quarter the 65 will continue to grow quite aggressively.

Okay.

Thanks a lot.

Thank you, Randy.

(Operator Instructions).

Your next question comes from the line of Robert Lee of Jefferies.

Okay.

Thanks very much for taking my call.

Randy has done a first-class job in asking several of my questions already so I think I've only got one remaining.

Going back to 40 nanometer as a percentage of sales I think it has been static about 6% over the last few quarters.

I think previously you said you expected to reach around 10% by the yearend.

Is that still a realistic target and can you also talk about how the sort of customer list and design ins, et cetera, building both on 40 nanometer and for 28 nanometers as well please?

Thank you.

40 nanometer revenue percentage will be over 10% in the month of December, exactly as we planned out.

So there is no surprise.

So by December month it will be surpassing 10% for 40 nanometer revenue, as far as the customer base 40 nanometer today approximately 20 customers with about 60 product tape outs, over 20 in production, 28 engagement about ten customers and the ten tape outs on macro IP.

And certainly as I mentioned earlier one customer is in pilot production on real products.

Okay.

And on the 28 nanometers there are they focusing on particular segment?

Again is it in communications or across the board, FPGA?

Can you give a little bit of color how the tape out is looking in both 40 and 28 nanometer?

It is very [graphage] there for the application process and based on the ASIC controller re-channel kinds of engagement, very broad based.

Okay.

And if I may, can I ask a quick follow-up question?

Obviously there is a lot of reference to the current macro uncertainties and global macroeconomic risks within your statement.

Do you -- some companies out there or some of your customers have alluded to perhaps a bottoming process beginning to take place.

I don't know if you have a firm view or even a gut feeling as to where we might be in this current downturn.

Do you have any sort of comment or color that may be helpful?

So traditionally Q3 for foundry has been our peak season historically in the past ten years.

The Q3 over Q2 has been over 10%, but Q3 we dropped around 10% so it is a big swing.

And the Q4 is not so strong either.

And now leading to Q1 next year Q1 is traditionally a weak season for foundry domain.

So I guess we will not have exactly a clear visibility because today's world is very uncertain.

It's really difficult to make too much a long-term forecast, but I think Q2 should be the good next checkpoint for all of us.

Okay.

Thanks very much.

Thank you.

And you have a question from the line Steven Pelayo of HSBC.

Yes, just a couple of follow-up questions.

I know you spoke about maintaining operating profitability, but I think we're all concerned a little bit also on the bottom line.

What does it mean for asset impairment or investment losses, things like that?

Can you give us some general idea of the ranges or what you might think the impact might be to the non-operating line in the fourth quarter?

We don't expect any major impairment loss in on the horizon or in Q4.

And however we do have some new business in the segment of solar and LED, which are also struggling profit wise.

So for Q4 we don't have final figures in mind yet about the non-operating, but there shouldn't be any major impairment loss coming in Q4.

Okay.

And I'm wondering if you could just help us understand how much capacity you have today for 40 nanometer and even for 28 nanometer, some of the initial I guess pilot capacity that you have, and then maybe think about it, I don't know, a year from today where you think that might be.

Is there any way you can help us size that?

Don't have an exact number for 40 nanometer.

We are building towards, I don't know, maybe 15K per month at a certain point next year.

I don't have an exact number.

28 is a big question.

We can -- we are still doing our planning for next year.

We may be very aggressive.

The demand is very strong so we haven't the exact number, right?

Today is mainly R&D pilot that kind of a capacity, as I mentioned and ramp is the second half of next year.

It's still well within our capacity planning cycle.

You mean for 28, but what about for 40 nanometer capacity today?

No.

I mentioned the building towards 15K per month, yes, sometime next year I guess.

Okay.

And then when I look within your segment details obviously 90 nanometer was down pretty healthy quarter-on-quarter, yet quarter micron was up.

I'm wondering if there's any details behind that, if there's anything specific that what were the relative areas of strength at quarter micron and what were the maybe strong, weakest areas at 90 and thoughts to going from that?

Yes.

90 nanometer for us I think is a weak note.

And many customers just jump from [escape] nodes to a 65 or a 55 as it's getting mature, the mass is achieved now.

So that's we were very happy to see that happening and it is no surprise.

25 I think there are lots of, as I mentioned earlier, there are lots of new engagements, lots of IDM outsourcing and analog or PMI see many new engagements into that node.

So it is for a game changes very fast, but 90 is shrinking.

That will probably will continue unless some of the traditional, until some traditional eight-inch business starts to get into that domain, for example, driver ICs.

And we see that happening also into 1980, so again it's changing now.

And my last question is just on during the afternoon you spoke a little bit about the quarterly depreciation, looking cost to get though those go up quite a bit in the fourth quarter.

And yet your operating expenses also didn't really fall much or flex down as much here in the third quarter.

So when I think about your breakeven ratio and forward it seems to me that actually could be raising a little bit.

Could you help me understand kind of a breakeven analysis to the operating line?

Our breakeven point is actually highly sensitive to a few parameters, of course mostly is loading, but also TWD20 exchange rate as well as our ASP.

In Q4 both ASP and NT$ exchange rate work in our favor so that helps to bring down the operating breakeven point a little bit.

And going forward I would say somewhere between 55 to 70 is probably our breakeven point level, but depends on the parameters I just mentioned.

I guess I'm going back a year, maybe two years ago.

Dr.

Chang at TSMC had talked about with the move to 40 nanometer and below he was going to be spending an extra 100 basis points on R&D.

I think it was going from 6% to 7% or 7% to 8%, can't seem to remember.

You seem to be talking a lot more about engaging at 28 nanometer and so I'm just curious does this suggest structurally higher R&D requirements for you guys as well?

Actually our 28 nanometer technology wise the pure development phase is done.

Now it's working with customers on implementation.

So our structure is -- will continue as where we are today.

We are focusing on the 20 nanometer development now.

We have taking out test the [eco], working with customers on 20 nanometer development.

28 is pretty much done I think.

Okay, great.

Thank you.

Thank you.

Your next question comes from the line of Donald Lu of Goldman Sachs.

Hey.

Good evening.

A couple of questions, first is in terms of end demand I think you mentioned that the smartphone demand in emerging countries has been the driver in Q4.

I just want to see whether there is any other areas are seeing sequential growth.

That's my first question.

Yes.

Other than the 3G cell phone that's definitely strong.

Other tablet is an area certain strength as well.

Okay.

Some e-book related tablets and some other things, yes, but it's a tough environment today in general.

Sure.

Yes, my second question is on CapEx.

I think the equipment are getting more and more expensive to 28 nanometer, 40 nanometer, et cetera.

What kind of CapEx should we expect for next year given your ramping both nodes?

Yes.

We will guide you is the numbers next quarter, but we are doing the planning now.

So you are exactly right and the demand may require additional CapEx to satisfy, especially 28 nanometers is getting more expensive.

If we start to ramp next, second half of next year we need to sink in more CapEx.

Is there any initial thought -- I thought in the afternoon there were some comments on the CapEx would be a certain percentage of depreciation or something like that?

Not really.

No.

We didn't mention about that.

Of course internally we have several guidelines from our financial structure point of view.

And the ideal situation is we will try to finance all the CapEx through internally generated cash.

So of course we still have some capacity for debt issuance so we will have some flexibilities.

I think broad guideline will be we try to finance all the CapEx through internally generated cash.

Also I would like to mention for this 28 nanometer CapEx it's very expensive, so we will not just build ahead and then look for customers or the CapEx.

We are working very hard to line them up with customer demand.

Sure.

So any guidance for depreciation for this year and next year?

This year will be low single digits, higher than that of last year, but next year we are expecting to see somewhere around 10% increase compared to that of this year.

I see.

Yes, my -- another question I have is looking back to your different nodes I think at 90, which you commented is a relatively weak node, or a 65, at 90 nanometer I think it as a percentage of revenue it maxed at around 31% of total revenue.

At 65 it last quarter was 34% of total revenue.

If we use this as a guide what do you think your 40 nanometer revenue at the peak would that still be above 30% of total revenue?

Would that be less?

That's a very good question.

Actually the industry landscape is steering into a territory very uncertain.

And certain customers are very aggressive on 28 as well.

So it's hard to predict so maybe it's hard to say.

We will provide our best service to a customer.

The customers are changing their minds all the time also.

It's a very fast-changing dynamics.

Oh, the good thing for us is for the 40 nanometer and the certain for example 28 polyscion the capacities are somewhat convertible.

Oh, you mean between 40 and 28, or 65?

Polyscion, yes.

Oh, polyscion.

The capacity for that is transferable, okay.

Yes.

Okay.

Yes, thank you very much.

Thank you.

(Operator Instructions).

Your next question comes from the line of Mahesh Sanganeria of RBC Capital Markets.

Thank you very much, another question on the capacity for the 40 nanometer and 28 nanometer.

Can you use some of your 90 nanometer and 65 nanometer capacity to convert them to 40 and 28 nanometer, and if the case how much cheaper that is to convert rather than add new wafers?

That is very difficult.

Even 65 converting to 40 nanometer it's the change of emersion scanner.

That's not convertible.

So it's even 65 converting to 40 is very expensive, not very efficient I would like to say.

But considering that you have so much capacity of unused capacity of 90 and 65 I'm just thinking maybe at least a certain percent, maybe 50% of the equipment you could reuse or you will not try that?

Actually lots of equipments are still common.

However ours, as I mentioned earlier, even 90, 65 we will always a second wave of specialty technology users to come and use those technologies.

That's the best way.

Conversion probably is not that great.

Even 90 I mentioned, even some of the high voltage driver ICs are getting to 90 and 80 nanometer for the HD720 drivers.

So even 65 there are lots of new stuff, yes, embedded Flash, CMOS in the sensor, et cetera, et cetera.

So we will always balance the conversion and finding the new specialty technology users for the legacy node.

That is nothing new from six-inch, eight inch this is always happening.

And the considering that your utilization is in mid 60s and probably Q1 stays in the similar kind of range, is it fair to assume that your CapEx in the next year will be more second half loaded and still on the first half still will be pretty weak?

Oh, no, no.

The CapEx is really long term and for example 28 nanometer has nothing to do with near term-term economic fluctuations.

We are quite committed to the so put in the customers on the long term, especially the leading edge.

And so you expect Q4 -- I guess you guided Q4 CapEx to be about TWD700 million.

Is that correct?

Yes.

Well our full-year budget is still TWD1.8 billion, but the actual number may alter according to the final payment and as well as the NT$/US$ exchange rate, but it won't be much difference.

Okay.

Thank you.

Thank you.

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

Thank you.

Every quarter I like to ask you guys about customer concentration.

Could you talk a little bit about I don't know maybe the percentage of revenues from your top five customers, let's say, in the third quarter versus the second quarter, and maybe how that would track going forward?

Or alternatively maybe you ask it another way, how many customers are greater than 5% in the third quarter versus the second quarter?

Could you help me try to quantify that a little bit?

Steve, we can give you the top ten revenue is about mid 70s in Q4.

It increased on high 50s in Q3.

Okay.

I was trying to get it a little bit more narrowly defined than that.

Okay, my next question is just --

(Inaudible - multiple speakers).

The top six or top seven are all over 5%.

Oh.

And was there a change in that, or a --?

Not much.

Of course Q4 will be one or two more.

Okay.

All right, and then there's a lot of questions about how big the 40 nanometer node will be and can it get it back to a third of your revenues.

Help us to understand when we went in the prior downturns as the technology node was ramping, does ultimately the pricing curve that you thought was going to be happening for that node shift lower and thus you don't monetize as much from that particular node?

I'm thinking for example SMIC for example put a lot of capacity on 65 nanometer and now of course they're facing probably much more price pressure than when they originally thought putting that capacity in.

Is this an issue or something we should be thinking about for you at 40 nanometer, or maybe put that in perspective historically for us?

Actually our 40 nanometer capacity expansion we are very disciplined and careful.

And as I mentioned for this advanced node in the past we at a certain stage we are building capacity ahead of customers' demand, but now we are very carefully aligning with customers on their demand.

And so I don't think that would be the situation.

Also as I mentioned 40 nanometer capacity and the 28 certain the polyscion they are in some way they are convertible.

They are both emersion scanner, similar [stray] engineering technologies, so we feel not too bad, but again the key is aligning with customers' demand, get the timing right, get it released timely and not too early, not too late and with a disciplined CapEx deployment, a rational expansion.

That's really what we are trying hard every day now.

The last question is you sound pretty excited about some of the early activity of your customer engagements happening at 40 and 28 nanometer in particular.

We don't get as close a look at what's going on in Samsung and Global Foundries as they're kind of hidden in private companies or bigger entities.

It's interesting your increased confidence despite the fact that it seems like the competitive environment should only be getting more difficult at that node.

Can you help us maybe frame this competitive landscape today in light of those entrants?

First of all I don't know either.

They are not just -- well, I guess some companies are not public companies.

And it's a private equity.

So we don't know much information either, but for us we are really actually for our execution delivery we are really trying to deliver what we committed to customers.

And it has been very good.

And we are earning our business on this new node very steadily.

So we are certain we read some news about some of their challenges.

So I think it's a very good opportunity if we can continue to deliver what we committed and earn our business with our technology delivery, and production and ramp.

Okay, guys.

Well, thank you.

It appears you guys are doing a pretty job in your cycle-to-cycle management this time, so congrats.

Take care.

Thank you.

Thank you, Steve.

Your next question comes from the line of Donald Lu of Goldman Sachs.

Yes.

I just wanted to follow up what would be the 12-inch and eight-inch utilization rate in Q4?

It's about the same for our guidance.

I see.

It's similar.

Okay.

So in --

Sorry?

So in Q3 the eight-inch utilization was higher and then in Q4 they are similar so the ASP actually is improving.

Is that the idea?

To some extent, yes, yes.

Okay, great.

Thank you.

Thank you.

And there are no further questions.

I would like to turn the call over to Mr.

Liu for closing remarks.

Thank you again for your interest in UMC.

Please feel free to contact us directly if you have any additional questions and, operator, back to you.

Thank you for your participation in UMC's conference.

There will be a webcast replay within an hour.

Please visit www.umc.com under the Investor Relations Investor Events section.

You may disconnect.

Goodbye.