UMC 联华电子

Operator

Welcome everyone to UMC's 2011 Q2 earnings conference call. (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. Richard Yu, Head of Investor Relations at UMC.

Richard Yu

Thank you and welcome to UMC's conference call for the second 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 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 U.S. and the ROC Securities authorities. I'd now like to introduce UMC's CFO, Mr. Chi-Tung Liu to explain UMC's Q2 2011 business results.

Chi

Thank you, Richard. For the second quarter of 2011, revenue was NT$28.15 billion, a 0.1% quarter-over-quarter increase from NT$28.12 billion in Q1 '11 and a 5.4% year-over-year decrease from NT$29.75 billion in 2Q 2010. Gross margin was 23.9%. Operating margin was 11.8%. Net income was NT$3.19 billion. And earnings per ordinary share were NT$0.26. Above, there is a short summary for the results on 2Q 2011. More details are available in the quarterly report, which has been posted on our website. I would now turn the call over to Dr. Sun.

Shih

Thanks, Chi-Tung. Good morning, good afternoon and good evening, ladies and gentlemen. In Q2 2011, wafer shipments reached 1.15 million 8-inch equivalent wafers, in line with UMC's guidance. As we enter the third quarter of 2011, certain macroeconomic factors such as European and U.S. sovereign debt issues and the emerging market inflation have led to unfavorable global economic conditions. These macro uncertainties have led customers to adopt a conservative outlook. Furthermore, customers have adjusted their order patterns in order to consume elevated inventory levels due to overstocking after Japan's March earthquake. This ongoing situation has negatively impacted the worldwide semiconductor industry and will offset the high demand season traditionally seen during the second half of the year. Consequently, we anticipate UMC's business performance may reflect this weakened demand condition beyond Q3. During this period, we will further enhance our operations through active customer diversification, efficient business execution and aggressive cost control measures. We will also focus on the opportunities ahead. UMC's CapEx, which remains at US$1.8 billion for 2011, will support our advanced production ramp. UMC's R&D efforts will continue in earnest to prepare for customers' future technology requirements. We will also leverage our foundry technologies to take advantage of the fast-growing mobile communication and the computing markets with high-performance, low-power consumption requirements. UMC's volume production 40nm process is expected to reach 10% of revenue by the end of 2011. We are also smoothly developing our 28nm process and design IP platform, with the first of several 28nm customer products entering pilot production in Q3 2011. For 65nm, UMC will benefit from the second wave of new customers targeting specialty technologies including analog, high-voltage, CMOS image sensor and embedded non-volatile memory products. We believe through the leadership of our cohesive, high-caliber management team, healthy capital structure and clearly defined business strategy, UMC will emerge from the downturn stronger and better positioned to capitalize on the business opportunities in 2012 and beyond. Now, let me provide you with the guidance for the third quarter of 2011. Revenue will show a low-teen percentage range, about 10% to 12% decline, mostly driven by lower volume. Operating margin would be in the low single digit percentage range. Capacity utilization would be in the low 70% range. The consumer and the computer segments will outpace the communications sector. This concludes my comments. We're now ready for questions. Operator, please open the lines up. Thanks.

Operator

(Operator Instructions) Your first question comes from Kevin Chen from Credit Suisse.

Randy Abrams

Hi, this is Randy Abrams from Credit Suisse. The first question follows on a bit from this afternoon. On your fast-follower strategy, you've traditionally over the past couple of years invested a bit slower on R&D and CapEx to focus on profitability. Are you starting to tilt back as you look at 28nm to be more aggressive now in R&D and CapEx to start to narrow the gap again, so be less of that fast-follower and start to try to narrow the gap again with TSMC?

Shih

First of all, we are not really fast-follower if you're talking about logic technology. I think that's really the Intel. They are defining the process architecture for logic technology generation-after-generation, starting from strain engineering, silicon-germanium, high-k metal gate, et cetera. So for UMC, (inaudible) we are trying to provide our customer-driven solution in a timely fashion, also fit into our overall structure. We still try to raise our R&D spending. In the past few quarter, it's 8% or below. So we're still doing that. However, on 28nm, you're right. We are not only selecting the right technology, which is high-k metal gate (K-Labs). In addition, I mentioned in the afternoon we're also offering a very competitive 28nm Poly-SiON technology with tweaked performance, with very competitive performance and price ratio. So overall speaking, we are trying to do our best to provide timely and adequate solution for our customers.

Randy Abrams

You touched a bit this afternoon on it. I guess the challenges that you've had on 40nm for a bit more mild ramp and then 28nm like the Poly-SiON or high-k metal gates. Are you starting to see customers from new applications to diversify out further on 28nm? Are you attracting some new customers there?

Shih

Actually 40nm, the ramp, if you compare to 65nm, it's not really late. Our first wave of design customers, they are moving along pretty smoothly, and the impact of economy may slow that down a little a bit. And in the meantime, we are at the incubation stage for the other 40nm products. Certainly, some of them are second source. So the design and the process of tuning and tweaking may take some time, but we will be there. I mentioned that by the end of the year around 10, so we'll ramp continuously next year. 28nm, we're grabbing opportunities. As I mentioned already, certainly 90nm is a weak node for UMC. Many customers jump from 0.13 to 65nm. We are still trying to understand closely on the development.

Operator

Your next question comes from Dan Heyler from Merrill Lynch.

Dan Heyler

I had a question on whether or not you could break out the $1.8 billion, because you haven't changed your CapEx number. So I'm wondering if you could break it out for us again what the contribution is between the advanced technology 28nm and other capacity-related spend?

Chi

8-inch is about 14%, and 12-inch including R&D is about 86%.

Dan Heyler

Could you give me the R&D number?

Chi

I think R&D is around 15% or less.

Dan Heyler

No change at all to the mix?

Chi

Not much.

Shih

For the 65nm, I mean 12-inch, we have already done for 65nm investment at this moment. So most of them at 12-inch, they are 40nm and 28nm related.

Dan Heyler

And then on terms of the revenue, you talked about 10% of revenue by yearend, but we don't know whether revenue is growing or not, as you alluded to earlier. Given the uncertainty and macro factors, I'm not sure whether you'll be getting growth in the fourth quarter, as you said this afternoon. Could you tell me in absolute terms is 40nm revenue growing in the third and fourth quarter?

Shih

We have the internal numbers of our revenue forecast. The ratio, we're still targeting 10% for 40nm. As far as the exact numbers, I think 40nm is ramping now for us.

Dan Heyler

Because obviously there has been cancellations and push on inventory, and you're ramping 40nm. So I was wondering if in spite of the inventory whether your 40nm is still growing sequentially in 3Q and 4Q. It sounds like you believe that it is.

Shih

Yes, well, that's our target. As I mentioned earlier, certain ramp speeds are related to some of the design revelations, and it's becoming a more difficult node for you and for the design and the customers to get the final version into production and to ramp to volume. We are at those final stages.

Dan Heyler

In terms of your process partners over the years, is there a different mix of your early ramp process partners at 28nm than you had at 40nm?

Shih

So we wish they have the early customers ramping with us. In the meantime, we are focusing on the more industry standard process, and that's very popular actually, for example, 28nm, high-k metal gate, even going back to the 40nm, 40G, 40LP. That really lowered the engagement barrier for many of our customers. So actually we've more partners. Early or later, it's timely I think. If it's still providing good value for them, it's good for us.

Dan Heyler

For the 28nm kind of leading initial partners, is it effectively the same group of guys that were with you at 40nm?

Shih

I mentioned earlier ones. Some earlier ones are the same. In the meantime, we are offering our own platform as well, as I mentioned, tweaked, enhanced 28nm technology platform. And also, we're working with the IP partners to provide a complete IP suite. And this is gaining momentum and customers are very interested, which provides us very, very competitive performance and price ratio for applications.

Dan Heyler

And contribution to revenue on 28nm by the yearend was what?

Shih

We don't have the 28nm. We're still working on that.

Operator

Our next question comes from Robert Lee from Jefferies.

Robert Lee

Actually I was going to ask on some of the product mix issues and also the CapEx side, but that's appears to have been asked already a little bit. But I guess with the $1.8 billion CapEx spend, to what extent are you committed? And if the macroenvironment were to take a further turn to the worst, what degree of flexibility would you have to reduce that number you saw fit at the time? And in addition, is the fact that you are not actually changing CapEx even though you face what most people would say is going to be quite a challenging quarter, should we see that as a signal that this is a temporary event impacted by a number of temporary factors which will pass in time? To what extent, do you have visibility that the business will stabilize into the fourth quarter and third quarter of next year?

Shih

For CapEx, it's always forward looking. As I mentioned earlier, the CapEx investment today, they are both for 40nm and 28nm already. So in the afternoon conference, I mentioned for 2012, not only 40nm will ramp, you're going to see 28nm will start to ramp very aggressively, actually steeply next year. So we're really preparing for the future. And today's business, we talk about the inventory issues of global macroeconomic concerns, we will definitely deal with it. Actually we're quite excited about future prospects in the mid to long term.

Chi

Robert, as for the flexibility we have for CapEx numbers, as we mentioned, there is NT$1.8 cash base CapEx budget, certainly we have the room to slow down the payment if we want to. However, consider our cash in hand. About US$1.5 billon, we saw the depreciation come in, and there is really no point to slow down the payment unnecessarily. So I would say the currency impact along with some payment terms may alter our final $1.8 billion number a bit, but there is no intention to change the regular payment terms for the $1.8 billion number.

Robert Lee

Can I ask you a quick follow-up question just in terms of your current visibility with the communications-related customers, because I guess that is one area, particularly on the feature phone and mainstream concept sector, where both yourselves and other suppliers out there seem to be experiencing some weakness at the moment? What visibility, if any, do you have on when that situation may stabilize, because like your own business, many of these end-customers would typically be experiencing a relatively seasonal strong second half which doesn't appears to be coming through? Now clearly, some of these customers have their own issues to deal with in addition, which are compelling their problems. But to what extent do you have any visibility that things are beginning to stabilize for those key communications customers?

Shih

Actually our Q3 revenue decline is a reflection of certain of these issues that you mentioned. But looking forward to the future, we are closely watching, but I doubt it will be too much worse.

Robert Lee

So it is a possibility that the business could reach some sort of stable stay into the fourth quarter?

Shih

That's hard to say. As I mentioned earlier in the afternoon conference, if it's just the inventory correction, it's possible. But NT amount is really the key issue or concern.

Robert Lee

And related to that, I mean I guess, everybody has been impacted by the soft macro developments that we've seen in Europe than elsewhere. Expectations in the sector are being set lower and people in general are taking quite a conceptive view on life. Therefore, to what extent is there some upside risk if we do see a little bit of seasonality coming through, if we do see consumers in the main western economies going out and spending in Christmas? Is that something that people are really betting on at the moment? And to what extent do you think it could happen and help to stabilize the business into the yearend?

Shih

It's definitely possible. But traditionally Q3 is the peak season for foundry, actually preparing for the U.S. Thanksgiving and globally the Christmas events. But this Q3 is much we've heard in the past. So as far as the outside risk, we definitely will come on that, but we need to be very cautious as well.

Operator

The next question comes from Donald Lu from Goldman Sachs.

Donald Lu

My question is a follow-up to the last question on the Q4 outlook. I agree the economy is very uncertain. Have you seen any recent order cut or indication from customers that they are going to really control their inventory in Q4, because of the economy? In terms of inventory correction, as you commented it probably mostly going to complete in Q3? And is there any sign that is showing any particular area may have lingering correction in Q4?

Chi

I guess a multiple question there. If we go back to the inventory in Q4, I can talk about UMC's Q3 for the different segments. I think for UMC the strengths have been consumer and computing area. For example, sensor area holding up pretty good; and hard disk controller, the re-channel controller chips are pretty good. Communications are still weak. That's where we are. As far as customer and inventory control is still happening. And we hope it will be all cleared up towards the end of Q3. But that depends on the demand situation. So we are watching it closely.

Donald Lu

And just in your own bookings in the last three-four weeks, are things getting more stabilized in the last few weeks or is that still continues to trend lower?

Shih

So I don't know. Let me provide you from a different angle. So the monthly revenue trend in our third quarter is going to be trend down months by months. So our Q4 starting point has to be at the low point of Q3. And we already have visibility to the end of Q4, so it's difficult for us to comment on the situation for overall Q4 outlook.

Donald Lu

My next question is more on the cash flow side. I think UMC has been have very healthy positive free cash flow for many years and this year the CapEx is higher, so that the free cash flow probably is going to be inactive. First, I want to confirm if that's the case. And secondly, going forward, is there any guidance on that, whether the company will try to maintain a balanced free cash flow or anything related to that?

Chi

To have self-sustained cash flow cycle is always what we will try to achieve. Last year and this year we spend a little bit more CapEx considering our cash position. But we also have a plenty of room from the debt financing, which we did issue in ECB earlier this year to cover the gap. And in the meantime, we also pay out pretty handsome cash dividend. So ultimately, we will continue to achieve this self-sustained cash flow cycle. But if necessary, we also have room to gear up a little bit if we want to. But this year, I think given the ECB issuance, we should be able to reach cash flow equilibrium in 2011.

Donald Lu

And going forward, would you try to maintain an operating cash flow minus CapEx to be relatively neutral, or you think next two years is good opportunity because some your competitors are having lot of turmoil at the management level?

Shih

Cash flow capability definitely is a boundary when we're considering or planning our CapEx. And as I mentioned, we certainly try to achieve the status of self-sustained from a cash flow point of view. But we also have room for gearing up if we want to. So basically, the goal is to reach this self-sustained level.

Operator

Your next question comes from Szeho Ng from BNP.

Szeho Ng

My first question is with regard to non-operating item in Q3. Please share with us what numbers we're putting out in the model.

Chi

Normally, we will have within income in the third quarter, but we also have some earnings tied to the ups and downs of the overall stock market. So I think it's safe and conservative to factor in higher than that quarter two numbers in your Q3 forecast. So we expect the number to be slightly higher than that in our quarter two number.

Szeho Ng

Any ballpark number that you'd suggest?

Chi

I'd say it's slightly higher.

Szeho Ng

Do you have any plan to convert the 65nm capacity for, let's say, 40nm. In fact that 65nm transition seems to be coming down?

Shih

It is a routine for us when we deal with no transition. For example, as we were converting a certain 0.13 90nm to 65nm. But the key is the efficiency. Usually, when you do this type of capacity conversion, you have to take care of the efficiency and you need to spend additional money to do the conversion. It's a routine for us. At the right time, we will do it.

Szeho Ng

Any other thing you can share with us? What percentage of your 65nm you plan on upgrading?

Shih

I don't have the number I can share with you today. But 65nm is doing okay. You can see the percentage of share. It's still growing.

Operator

Your next question comes from Dan Heyler from Merrill Lynch.

Dan Heyler

You talked about your Poly-SiON technology. What area of application do you anticipate the first ramp in your 28nm? Will it be performance driven or do you think it will be more in the low-power wireless sector?

Shih

Actually the first product is on a high-performance application processor.

Dan Heyler

As you look at the technology, most of the problems appear to be in low-power based on what we're hearing as well challenges there. Could you elaborate on the 28nm? Are we going to see a low-power adoption come much later down the road versus high-performance than we saw, say, in 40nm, that is the other time between performance and low-power that may come inside a bit later than normal?

Shih

At 40nm, we still have high-performance process, for example 40G, and the low-power process 40LP. The process is actually converting into one process at 28nm. That's the trend. We call that SoC process. It covers both high-performance and low-power into one process. So that's something new for 28nm. One process probably covers a broad range of applications.

Dan Heyler

So therefore you would expect the low-power ramp to come relatively shortly or in line with your high-performance ramp?

Shih

Yes, high-performance application processes are used in more mobile devices. So these devices are very high performing.

Dan Heyler

I'm comparing, say, network processors or FPGAs versus the big data apps processors. So I am kind of thinking high-performance digital versus low-power.

Shih

So it's not working over FPGA processor.

Operator

Your next question comes from Steven Pelayo from HSBC.

Steven Pelayo

I am trying to sort out UMC's relative performance taking into respect your different customer exposures and your different technology mix. I guess I'm struggling. So maybe I want to ask it the other way. Think about the movement in market exposure, you have about 50%, 60% of revenues from communications. Could you really help us understand have you done the analysis on what percentage of that is more a 3G smartphone world versus the 2G feature phone world?

Shih

I don't have the exact number. But we are exposed for both 3G and 2G today and some 3.75G even. So all kinds of customers and applications. But we are a foundry manufacturing service provider. So I guess we're exposed for all of those stuffs.

Steven Pelayo

I guess maybe I was thinking your relatively higher exposure to the feature phone world is causing maybe some of the relative underperformance to companies like TSMC.

Shih

Yes, to some extent, that's true, but we're in the transition as well.

Steven Pelayo

And maybe just an accounting question. You guys didn't give gross margin guidance this quarter, focusing more on the operating profit line. You do have depreciation growing. Could you just remind me what the target is for the full year and how you see it specifically within the cost of goods sold line in the third quarter and in the second half?

Chi

Yes, if you talk about depreciation, the full-year depreciation number will grow year-over-year, but it shouldn't be more than 5%. And for third quarter, the reason we give operating profit margins instead of gross margin is operating profit margin guidance is low single digit, and we just want to make sure that we're at this moment comfortable with this guidance. In terms of the absolute operating expenses, in absolute dollars you'll show slightly growth from the number of last quarter.

Steven Pelayo

And the depreciation within your cost of goods sold, they grew 11% quarter-on-quarter in 2Q. What's the outlook for that particular line, so I can try to think more about your gross margins in the second half?

Chi

Because we did consume some of the inventory in our working process in the second quarter, I think for Q3 we should be back to same ratio.

Operator

The next question comes from Donald Lu from Goldman Sachs.

Donald Lu

I have a question on depreciation again. Chi-Tung, can you give us an estimate on depreciation next year, assuming your CapEx will be relatively flat or whatever number? Is there just a directional thing we can have?

Chi

Well, I can tell you from the other angle again. This year, we do have inspiring depreciation carry over the last four, five years. So we do see some of our existing facilities have significant reduction in the depreciation expenses. But that probably won't repeat again next year. So we do expect to see their depreciation increase in terms of percentage to be relatively higher than that of this year. I don't have a final number for CapEx next year. So it's difficult for us to give you a percentage right now.

Donald Lu

Do we know like what's the difference in terms of the depreciation come out of line this year relative to next year, like the difference?

Chi

The difference is quite big. We are talking about more than NT$1 billion.

Donald Lu

And another question is on the tax rate. Is there any update on that for this year or next year?

Chi

Well, I think this year and next year, we're always in the range between 8% to 10%.

Donald Lu

So it will be a relatively unstable.

Chi

Yes.

Donald Lu

Your last press release with Cypress on this embedded flash is quite interesting, and it seems like a pretty new product if I am correct. What would be the application for that thing? What kind of product would that be used on?

Shih

Well, actually this 65nm embedded flash using, we call it, SONOS technology. It's an extension from Cypress 0.13 SONOS product. It's used broadly on their PSoC product lines for all the controller, et cetera, et cetera. But now, they have been working with us on 65nm for many years. And a good thing for this, we are allowed to promote this technology to other customers as a broad-based IP and macro for broad-based applications. It's very competitive from performance, cost consideration. So we're quite excited

Donald Lu

When would you think you will see revenues coming from this technology?

Shih

I don't have the number. I think, yes, early next year probably in the case with Cypress.

Operator

The next question comes from Steven Pelayo from HSBC.

Steven Pelayo

As you guys were realigning the company for better returns over time, you actually mentioned there were a few quarters I think with 30%-plus gross margin, 10% double-digit plus ROEs. Now you got to kind of prove ourselves maybe through either mid type of correction or downturn here so. Could you help us understand really what's different today to make you whether the downturn is always better, maybe quantify a little bit more on customer diversification or breakeven levels, and maybe just the fact to help understand how UMC were whether a full back better this time?

Shih

So from a technology perspective, I mentioned, we are in the transition from many of the other 65% customers moving on to 40. But I had mentioned earlier, we are still in a transition phase doing the final design and tuning in for these things. And the 65 will have a second wave of our products. At 40, we are providing 40G, 40LP and those technologies. And the 28, I mentioned next year, we probably are ramping together with 40. So in a long run, it's quite exciting. But this, you said the mid-cycle inventory correction we just have to deal with very efficiently and transition smoothly to the next cycle.

Chi

And Steven, we do have very a cohesive management team and we're pacing ourselves. And if you look at our CapEx number, we do feel we are one of the most disciplined ones. So basically we are focusing on addressable market, which are very clear target. So I guess that's a key difference compared to the last down cycle.

Steven Pelayo

And then maybe if you could just talk a little bit about customer types as well, and I know you guys have been working a lot on diversifying your customers beyond kind of these top three of the past? Maybe remind, if again, how many are above 5% revenues remind us? I think a year or two ago, you talked about really wanting to focus more on U.S. fabless as well. I kind of want to hear how this potential downturn is different than the last one for UMC.

Shih

I think our top 10 customers is about 70% of our revenue. We have three 10% of revenue customers, and there are four 5% revenue customers. And customer diversification, as Chi-Tung mentioned, the difference is that we are really focusing on addressable customers in addressable markets in certain area. If these disconnect in timing on technology or service we can provide, we tend not to spend too much in resources, that's where we are.

Operator

There are no further questions at this time. Sir, please continue.

Chi

Thank you again for your interest in UMC. Please feel free to contact us directly, if you have additional questions. Operator, back to you for the closing remarks.